Provided for non-commercial research and education use ... · Ayako Miyazaki, Makoto Nagai, Hiroshi Tsunemitsu..... 189 Validation of an assay for quantification of alpha-amylase

The Canadian Journal of Veterinary Research is published by Canadian Veterinary Medical Association. The attached copy is

furnished to readers for personal, non-commercial research and education use. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party Web sites are prohibited. Those who require further

information regarding reprints or archiving and manuscript policies are encouraged to contact [email protected].

Provided for non-commercial research and education use. Not for reproduction, distribution or commercial use.

Canadian Journal of Veterinary Research Revue Canadienne de Recherche Vétérinaire

Canadian Journal of Veterinary Research Revue Canadienne de Recherche Vétérinaire

ArticlesDetection of Mycobacterium avium subspecies paratuberculosis in tie-stall dairy herds using a standardized environmental sampling technique and targeted pooled samplesJuan C. Arango-Sabogal, Geneviève Côté, Julie Paré, Olivia Labrecque, Jean-Philippe Roy, Sébastien Buczinski, Elizabeth Doré, Julie H. Fairbrother, Nathalie Bissonnette, Vincent Wellemans, Gilles Fecteau . . . . . . . . . . . . . . . . . . . . . . . . . .175

Association between thermal environment and Salmonella in fecal samples from dairy cattle in midwestern United StatesTasha Likavec, Alda F.A. Pires, Julie A. Funk. . . . . . . . . . . . . . . . . . . . . . . . . . .183

Reinfection of adult cattle with rotavirus B during repeated outbreaks of epidemic diarrhea Michiko Hayashi, Toshiaki Murakami, Yoshizumi Kuroda, Hikaru Takai, Hisahiro Ide, Ainani Awang, Tohru Suzuki, Ayako Miyazaki, Makoto Nagai, Hiroshi Tsunemitsu. . . . . . . . . . . . . . . . . . . . . 189

Validation of an assay for quantification of alpha-amylase in saliva of sheepMaria Fuentes-Rubio, Francisco Fuentes, Julio Otal, Alberto Quiles, María Luisa Hevia. . . . . . . . . . . . . . . . . . . . . . .197

Effects of the –791(C➞T) mutation in the promoter for tumor necrosis factor alpha on gene expression and resistance of Large White pigs to enterotoxigenic Escherichia coli F18Ying Liu, Chaohui Dai, Li Sun, Guoqiang Zhu, Shenglong Wu, Wenbin Bao . . . . . . . . . . . . . . . . . . . . . . . . . . 203

Effects of inhibitors of vascular endothelial growth factor receptor 2 and downstream pathways of receptor tyrosine kinases involving phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin or mitogen-activated protein kinase in canine hemangiosarcoma cell linesMami Adachi, Yuki Hoshino, Yusuke Izumi, Hiroki Sakai, Satoshi Takagi . . . . . . . . . . . . . 209

J u l y / J u i l l e t 2 0 1 6 , V o l . 8 0 N o . 3table of contents continued on back cover/la table des matières se poursuit sur la couverture arrière

FOR PERSONAL USE ONLY

Published quarterly by the Canadian Veterinary Medical AssociationPublication trimestrielle de l’Association canadienne des médecins vétérinaires

C a n a d i a n J o u r n a l o fVe t e r i n a r y R e s e a r c h

Formerly CANADIAN JOURNAL OF COMPARATIVE MEDICINE

The journal publishes the results of original research in veterinary and comparative medicine. Manuscripts must be as concise as pos-sible, and the research described must represent a significant con-tribution to knowledge in veterinary medicine. Full-length papers, short communications, and review papers are welcome. All manu-scripts will be reviewed for scientific content and editorial accuracy.

Manuscripts must conform to the Instructions for Authors found on the CVMA website. Consultation of a recent issue of the journal to ensure that the manuscript conforms to current style is also recom-mended. Please use the website (www.canadianveterinarians.net) to submit articles. A publication charge of $90 per published page for CVMA Members and $130 per published page for Non-members, plus a cost of $15 per table, $20 per illustration, and $150 per color illustration is payable by contributors. Instructions for ordering reprints are sent with the galley proofs. Color illustrations will only be published if the author(s) agree to pay the extra cost.

R e v u e c a n a d i e n n e d er e c h e r c h e v é t é r i n a i r e

Remplace la REVUE CANADIENNE DE MÉDECINE COMPARÉE

La revue publie les résultats de recherches originales en médecine vétérinaire et comparative. Les manuscrits doivent être aussi brefs que possible et la recherche décrite doit apporter une contribution impor-tante à la médecine vétérinaire. Les exposés complets, les communi-cations brèves et les comptes rendus sont bienvenus. Nous véri fierons le fonds scientifique et la forme de tous les manuscrits.

Les manuscrits doivent être conformes aux directives aux auteurs qui se trouvent sur le site Web de l’ACMV. La consultation d’un récent numéro de la revue est aussi profitable, afin de s’assurer qu’un manuscrit se conforme au style courant. Veuillez utiliser le site Web (www.veterinairesaucanada.net) pour soumettre les articles. Les auteurs doivent acquitter des frais de 90 $ par page publiée pour les membres de l’ACMV et 130 $ par page publiée pour les non-membres, 15 $ pour chaque tableau, 20 $ pour chaque illustration, et 150 $ pour chaque illustration en couleur. Un formulaire de commande de tirés à part accompagne l’en voi des épreuves d’un manuscrit. Les illustrations en couleurs ne seront publiées que si les auteurs en assument les frais supplémentaires.

Established in 1937 ISSN 0830-9000

Address all correspondence to/Adresser toute correspondance à :Canadian Journal of Veterinary Research Revue canadienne de recherche vétérinaire339, rue Booth Street, Ottawa, Ontario K1R 7K1Tel./Tél. : (613) 236-1162 — Fax/Télécopieur : (613) 236-9681e-mail/Courriel : [email protected]© Canadian Veterinary Medical Association 2016

L’Association canadienne des médecins vétérinaires 2016

Typesetting/Typographie AN Design Communications www.an-design.ca

Canadian Journal of Veterinary ResearchRevue canadienne de recherche vétérinaireEditor — Rédacteur Éva Nagy, Guelph, OntarioAssociate Editor — Rédacteur adjoint Deborah M. Haines, Saskatoon, SaskatchewanAssistant Editors — Assistants à la rédaction Serge Messier, Saint-Hyacinthe (Québec)Managing Editor — Directrice de la rédaction Heather Broughton, Ottawa, OntarioAssistant Managing Editor — Directrice adjointe à la rédaction Stella Wheatley, Ottawa, OntarioAdvertising Manager — Gérante de la publicité Laima Laffitte, Wendover, OntarioEditorial Coordinator/Coordonnatrice de la rédaction Linda Chow, Ottawa, Ontario

Abstracted and/or indexed in:/Résumé et/ou catalogué dans :Biological Abstracts Index VeterinariusChemical Abstracts Nutrition AbstractsCurrent Contents (Agric. Section) Science Citation IndexEMBASE/Excerpta Medica Veterinary BulletinIndex Medicus Derwent Veterinary Drug File

Subscription/Abonnement 2016Personal Subscription/ $165 $ (1 GST/HST as Abonnement personnel applicable; TPS/TVH en (Canada) vigueur)Personal Subscription/ $185 $ US/É.-U. Abonnement personnel (Foreign/Étranger)Institutional Subscription/ $215 $ (1 In Canada, GST/HST Abonnement institutionnel as applicable; Au Canada, TPS/TVH selon le cas)


Article

2016;80:175–182 The Canadian Journal of Veterinary Research 175

Detection of Mycobacterium avium subspecies paratuberculosis in tie-stall dairy herds using a standardized environmental

sampling technique and targeted pooled samplesJuan C. Arango-Sabogal, Geneviève Côté, Julie Paré, Olivia Labrecque, Jean-Philippe Roy,

Sébastien Buczinski, Elizabeth Doré, Julie H. Fairbrother, Nathalie Bissonnette, Vincent Wellemans, Gilles Fecteau

A b s t r a c tMycobacterium avium ssp. paratuberculosis (MAP) is the etiologic agent of Johne’s disease, a chronic contagious enteritis of ruminants that causes major economic losses. Several studies, most involving large free-stall herds, have found environmental sampling to be a suitable method for detecting MAP-infected herds. In eastern Canada, where small tie-stall herds are predominant, certain conditions and management practices may influence the survival and transmission of MAP and recovery (isolation). Our objective was to estimate the performance of a standardized environmental and targeted pooled sampling technique for the detection of MAP-infected tie-stall dairy herds. Twenty-four farms (19 MAP-infected and 5 non-infected) were enrolled, but only 20 were visited twice in the same year, to collect 7 environmental samples and 2 pooled samples (sick cows and cows with poor body condition). Concurrent individual sampling of all adult cows in the herds was also carried out. Isolation of MAP was achieved using the MGIT Para TB culture media and the BACTEC 960 detection system. Overall, MAP was isolated in 7% of the environmental cultures. The sensitivity of the environmental culture was 44% [95% confidence interval (CI): 20% to 70%] when combining results from 2 different herd visits and 32% (95% CI: 13% to 57%) when results from only 1 random herd visit were used. The best sampling strategy was to combine samples from the manure pit, gutter, sick cows, and cows with poor body condition. The standardized environmental sampling technique and the targeted pooled samples presented in this study is an alternative sampling strategy to costly individual cultures for detecting MAP-infected tie-stall dairies. Repeated samplings may improve the detection of MAP-infected herds.

R é s u m éMycobacterium avium ssp. paratuberculosis (MAP) est l’agent étiologique de la maladie de Johne, une entérite chronique contagieuse des ruminants et responsable d’importantes pertes économiques. Plusieurs études, la plupart réalisées dans des grands troupeaux en stabulation libre, ont démontré que la technique de culture de prélèvements de l’environnement est appropriée pour la détection des troupeaux infectés par MAP. Dans l’est du Canada où prédominent les petits troupeaux en stabulation entravée, certaines conditions et pratiques de régie pourraient avoir un impact sur la survie, la transmission et l’isolement de MAP. Notre objectif était d’estimer la performance d’une technique standardisée de culture de prélèvements de l’environnement combinée à des échantillons groupés ciblés pour la détection des troupeaux laitiers en stabulation entravée infectés par MAP. Vingt-quatre troupeaux (19 infectés et 5 non infectés) ont été enrôlés, mais seulement 20 troupeaux ont été visités 2 fois dans la même année pour y prélever 7 échantillons de l’environnement et 2 échantillons groupés (vaches malades et vaches maigres). Des échantillons individuels de toutes les vaches dans le troupeau ont été également prélevés. L’isolement de MAP a été réalisé en utilisant le milieu de culture MGIT ParaTB et le système de détection BACTEC 960. Globalement, MAP a été isolée dans 7 % des cultures de l’environnement. La sensibilité de la technique était de 44 % (IC 95 % : 20 % à 70 %) en combinant le résultat des 2 visites et de 32 % (IC 95 % : 13 % à 57 %) en utilisant aléatoirement le résultat d’une seule visite. La meilleure stratégie d’échantillonnage était la combinaison des échantillons de la fosse, de l’écureur, du groupe de vaches malades et du groupe de vaches maigres. La technique standardisée de prélèvements de l’environnement combinée aux échantillons groupés ciblés présentée dans cette étude est une alternative économique à la culture individuelle pour détecter des troupeaux laitiers infectés par MAP. La répétition des prélèvements pourrait contribuer à améliorer la détection des troupeaux infectés par MAP.

(Traduit par les auteurs)

Département de sciences cliniques, Faculté de médecine vétérinaire, Université de Montréal, Saint-Hyacinthe, Québec J2S 8H5 (Arango-Sabogal, Roy, Buczinski, Doré, Wellemans, Fecteau); Direction générale des laboratoires d’expertise, Ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec, Québec G1P 4S8 (Côté); Agence canadienne d’inspection des aliments, Saint-Hyacinthe, Québec J2S 7C6 (Paré); Laboratoire d’épidémiosurveillance animale du Québec, Ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec, Saint-Hyacinthe, Québec J2S 7X9 (Labrecque, Fairbrother); Dairy and Swine Research and Development Center, Agriculture and Agri-Food Canada, Sherbrooke, Québec J1M 0C8 (Bissonnette).

Address all correspondence to Dr. Gilles Fecteau; telephone: (450) 773-8521, ext. 8337; fax: (450) 778-8102; e-mail: [email protected]

Received July 23, 2015. Accepted March 24, 2016.


176 The Canadian Journal of Veterinary Research 2000;64:0–00

I n t r o d u c t i o nJohne’s disease is an incurable, chronic, and contagious enteritis

of ruminants caused by Mycobacterium avium ssp. paratuberculosis (MAP). The disease causes significant economic losses related to reduced milk production, premature culling, increased replacement costs, and decreased slaughtered carcass weight (1–3). This intracel-lular bacterium, which invades the immune cells of the gastrointes-tinal tract, has also been linked to Crohn’s disease in humans (4–6). Recent studies have led to increased concern about the zoonotic potential of MAP (7–9). Fecal-oral contamination is the main route of MAP transmission (10) and contact between calves and the feces of adult cows is the most important risk factor (11). Young calves are the most susceptible to MAP infection (12). They are prone to becoming infected by ingesting colostrum or milk from infected animals (13) or contaminated water or food (10). As excretion and clinical signs are observed at an older age (14), shedder cows are the main infectious source of environmental contamination. The ability of MAP to survive in the environment for up to 11 mo contributes to the perpetuation of infection in dairy herds (15,16).

Environmental sampling is one of the testing procedures recom-mended for control programs to assess MAP status in dairy herds (17). The evidence suggests that environmental sampling is a cost-effective method for determining infection status in previously untested dairy herds (18). The technique is simple, less expensive than individual tests, and does not require handling of individual animals. Some studies have compared the performance of environ-mental culture (EC) with that of individual fecal culture (IFC) and enzyme-linked immunosorbent assay (ELISA) (individual milk and serum samples) for detecting MAP-infected herds (19–22). Other studies have evaluated the correlation between EC and within-herd prevalence (WHP) based on IFC (23–25). In 1 study conducted on California dairy farms, no significant difference was observed among the 3 testing methods (EC, IFC, and ELISA) in terms of the propor-tions of herds correctly identified as infected (19). Another study found a highly significant relationship between EC and IFC (20). The sensitivity of EC for detecting MAP infection at the herd level has been estimated at between 40% and 81% (21,23,24). Specificity has been estimated to be close to 99% (25).

Environmental sampling has been evaluated mostly in large free-stall dairy herds in the United States (19–21,23) and recently in western and Atlantic Canada (22,25). Eastern Canada (the region east of Manitoba) is home to about 50% of Canada’s dairy herds, most of which are small tie-stall herds. In the province of Quebec, the average herd size is 57 cows per farm and 92% of the herds are housed in tie-stall barns (26). Because of specific management prac-tices and conditions in this part of the country, this area provides an interesting regional data set for evaluating the environmental sampling technique. Manure management practices in tie-stall barns differ from those in the large free-stall facilities typical in the United States and western Canada. These characteristics, combined with eastern Canada’s humid continental climate, may influence the sur-vival, transmission, and recovery of MAP in various environmental sampling areas.

The purpose of this study was to estimate the performance of a standardized environmental and targeted pooled sampling technique

for identifying MAP-infected tie-stall dairy herds. Secondary objec-tives were to describe the distribution of MAP in the environment of tie-stall dairy herds in Quebec and to find the best sampling strategy for detecting MAP-infected tie-stall dairy herds.

M a t e r i a l s a n d m e t h o d s

Study design and sample sizeA cross-sectional study was designed to evaluate a standardized

environmental and targeted pooled sampling method in tie-stall dairy herds. The source and target populations were, respectively, the dairy herds enrolled in the Quebec Voluntary Paratuberculosis Prevention and Control Program (QVPPCP) and Quebec dairies. A convenience sample of 24 tie-stall dairy herds was purposively selected based on historical MAP status. That sample included 19 MAP-infected herds (see case definition in next paragraph) and 5 non-infected herds. Additional inclusion criteria were the owner’s willingness to participate, tie-stall configuration, regular veterinary herd health visits, access to electronic records, and no drastic changes in the farm system, e.g., a change to free stall, in the year before the study began.

Case definitionA herd was considered infected for the purpose of our analysis

if MAP was cultured from at least 1 sample (IFC or EC) during the 24 mo before the study began or during the study period itself. The specificity of the bacteriologic culture was assumed to be 100% (17). A herd was considered negative if it had 2 negative results with EC (sampled in a 12- to 18-mo interval) and no clinical animals (persis-tent diarrhea and loss of body weight and normal appetite) during the 24 mo before the study began.

Sample collectionInitially, 20 herds were visited in summer 2011 (June 20 to

August 23). These herds were visited again in fall 2011 (October 3 to November 24) and 4 additional herds were enrolled, for a total of 24 herds. The samples were analyzed at the Laboratoire d’épidémiosurveillance animale du Québec in Saint-Hyacinthe, Quebec. Upon reception and within 24 h of collection, the fecal samples were stored at 280°C until they were analyzed.

Environmental samples — A set of 7 environmental samples and 2 pooled samples was collected from sick cows and cows with poor body condition by 2 members of the research group during each herd visit using a standardized technique (Table I). The sampling area was documented with photographs and videos in order to standardize the procedure throughout the study and record the precise sampling sites for each farm.

The 7 environmental samples came from 4 sites on each farm: 3 locations (gutter, manure pit, and heifers’ area) were sampled twice and the 4th location (the boots of the farm owner or the sampler) was sampled once. A composite sample of about 20 g of manure from each site was made for the 4 locations.

Additionally, 2 pooled samples were collected from 3 to 5 cows purposively chosen from 2 categories: sick cows and cows with body condition scores (BCSs) lower than 3. For the sick cow group, the



owner selected cows that were affected by any disease, but had never been diagnosed as positive for paratuberculosis. For the group with poor body condition, the sampler chose cows with a BCS lower than 3 (using a scale of 1 to 5). An individual fecal sample of about 20 g was taken from the rectum of each selected cow, using a single-use veterinary glove without lubricant. At the farm, equivalent amounts of feces were gathered from each individual sample into their respec-tive pools and homogenized as described previously (27). Briefly, the feces were mixed with a wooden tongue depressor by means of 10 vertical stirs from the bottom to the top, followed by 10 clockwise stirs, and 10 counter-clockwise stirs. Duplicates from each pool were stored in 2 plastic containers for transport to the laboratory.

Individual fecal and blood samples — From each herd, all cows older than 24 mo that had calved at least once were tested during each visit. An individual fecal sample of about 20 g was taken using a single-use veterinary glove without lubricant. Also, a single blood sample per cow was collected from the coccygeal vein in an 8-mL vacutainer tube without anticoagulant (Becton, Dickinson, Mississauga, Ontario). Blood samples were centrifuged and aliquots of serum were stored at 220°C until ELISA analysis.

MAP cultureEnvironmental and individual fecal samples were processed fol-

lowing the manufacturer’s recommendations (Becton, Dickinson, Sparks, Maryland, USA) and the US Department of Agriculture (USDA) (28). The MAP was isolated using the MGIT Para TB culture media and the BACTEC 960 detection system (Becton, Dickinson) at the Laboratoire d’épidémiosurveillance animale du Québec in Saint-Hyacinthe, Quebec, which is a USDA-certified laboratory.

An initial 3-day decontamination was carried out on the samples. Initially, 2 6 0.2 g of feces was diluted into 17.5 mL of sterile distilled water and allowed to settle at room temperature for 30 min. Then 2.5 mL of the supernatant was transferred aseptically to a 50-mL tube with 2.5 mL of 15% yeast extract and 0.2 mL of 10% sodium pyruvate. This solution was mixed briefly and incubated for 90 min at 36 6 1°C. For each fecal sample preparation, 0.3 mL of sterile 5% malachite green solution was added to a solution of 25 mL of sterile half-strength brain heart infusion (BHI) medium and 0.9% hexadecylpyridinium chloride (HPC). Finally, all 5.2 mL of the feces-germination mix was added to the BHI-HPC solution to complete a 30 mL decontamination suspension, which was vortexed briefly and incubated overnight (18 to 24 h) at 36 6 1°C.

The next day, this decontaminated suspension was centrifuged for 30 min at 900 3 g. The supernatant was gently poured off. Then, 1 mL of an antibiotic brew (vancomycin at 100 mg/mL, nalidixic acid at 100 mg/mL, and potency-adjusted amphotericin B at 25 mg/mL) was added to the pellet. The suspension was incubated overnight (18 to 24 h) at 36 6 1°C. Also, 1.5 mL of an additive cocktail was added to each MGIT Para TB culture tube (Becton, Dickinson). The additive cocktail contained Para TB supplement (bovine albumin, catalase, casein, oleic acid; Becton, Dickinson), egg yolk enrichment, sterile water, and antimicrobials (2.5% vancomycin, 2.5% nalidixic acid, and 1% amphotericin B). These tubes were stored in a safety cabinet at room temperature for 18 to 24 h.

On the third day of the fecal sample processing, the concen-trated specimen suspension was mixed by swirling and 0.1 mL was inoculated into the MGIT Para TB culture tubes before they were introduced into the BACTEC 960 detection device for incuba-tion at 37°C for a maximum of 56 d. The additive cocktail used for the environmental samples processed in 2011 included a higher concentration of nalidixic acid (10 times more). From January 2012 onward, a single additive cocktail, which included an additional 1% ceftriaxone, was used for processing all the samples, both individual and environmental.

The tubes that gave a positive signal before 42 d were always put back in the device for further incubation. The tubes flagged as positive between 42 to 56 d of incubation were incubated for an additional 72 h at 36 6 1°C and an acid-fast bacilli stain was done using the TB Fluorescent Stain Kit M (Fisher Scientific, Ottawa, Ontario). Positive samples were confirmed by real-time polymerase chain reaction (PCR) [TaqMan MAP (Johne’s) Reagents; Applied Biosystems, Foster City, California, USA]. The results of previous tests were interpreted according to USDA recommendations (28). Samples were identified as MAP-positive if they were flagged in the system and confirmed by both the acid-fast bacilli stain and the real-time PCR.

ELISASera were processed using the IDEXX Pourquier MAP antibody

test kit (IDEXX Laboratories, Westbrook, Maine, USA) according to the manufacturer’s instructions. Optical density (OD) values were transformed into sample-to-positive (S/P) ratios as described previ-ously (29). Samples with an S/P ratio of 55% or greater were con-sidered positive. As suggested previously (30), a herd was declared ELISA-positive if the serum within-herd prevalence (WHP) was 2%

Table I. Environmental and targeted pooled samples for identifying tie-stall dairy herds infected with MAP

Environmental and targeted Number pooled of samplessamples Description per visitManure pit Samples taken more than 10 cm deep 2

Gutter At the end of the barn but before the 2 manure pit

Heifers’ pen Composite samples from 4 different 2 surfaces at the site

Boots Samples scraped from the soles of 1 boots at the end of each visit but before going to the manure pit

Sick cows Pool of 3 to 5 cows affected by any 1 disease but never having tested positive for paratuberculosis

Cows with Pool of 3 to 5 cows with a body 1 poor body condition score lower than condition 3 on a scale of 1 to 5

Total 9



or greater, given that the specificity of ELISA relative to fecal culture has been estimated at 98% to 99% (31–33).

Statistical analysisStatistical analyses were carried out with the SAS software

(Version 9.3; SAS Institute, Cary, North Carolina, USA). Descriptive statistics for the individual and environmental samples were con-ducted to characterize the distribution of positive results. The Wilcoxon rank test was used to compare the WHP by ELISA and IFC and the number of positive ECs between the 2 visits. The percent-age and confidence interval of infected herds detected by the tests used in the study were calculated for a single visit and for 2 visits to assess the impact of repetitive samplings. For the herds visited twice, 1 sampling was randomly chosen for the assessment of a single visit. The association between the number of positive ECs and fecal WHP was evaluated with the Chi-square test. This test was also used to compare the number of infected herds confirmed by each diagnostic test. The percentage of positive EC, IFC, and ELISA samples per visit was compared using the Z test. All results were considered significant if P < 0.05.

Re s u l t s

Herd characteristicsThe 24 herds were located in 4 regions of Quebec, Canada

(Bas-Saint-Laurent, Capitale-Nationale, Montérégie, and Centre-du-Québec). Median herd size was 59 lactating cows (95% CI: 48 to 65), ranging from 30 to 211 cows. In 16 herds, the cows were exclu-sively Holstein, in 1 herd the cows were exclusively Jersey, and in 7 herds, more than 1 breed was also present (Holstein and Jersey in 6 herds and Holstein and Brown Swiss in 1 herd). The mean age of the cows sampled at the beginning of the study was 4.5 y (2 to 14 y). The apparent WHP ranged from 0% to 28% for IFC and from 0% to 31% for ELISA. The number of positive EC samples per herd ranged from 0 to 7. The proportion of culled animals during the study for the herds that were sampled twice was 13% on average (2% to 27%).

When combining sampling results from both seasons, out of the 24 enrolled herds, 17 MAP-infected herds (according to the initial

MAP status) were found positive by at least 1 of the detection meth-ods used (Figure 1). According to the initial MAP status, 5 MAP-infected herds were found positive by all 3 tests. Although 7 herds were found positive only by ELISA, 1 of these herds was presumed to be non-infected based on our case definition.

Among the MAP-infected herds according to the initial MAP status (n = 19), 17 were detected as positive during the study by at least 1 of the detection methods used (Table II). Overall, the number of infected herds confirmed during either visit was 8 out of 19 for EC, 9 out of 19 for IFC, and 16 out of 19 for ELISA. More infected herds were confirmed by ELISA than by the other tests (P = 0.01). The different combinations of test results for a single visit or for both visits of MAP-infected herds are presented in Table III. When the results from 2 different visits were combined, the percentage of infected herds detected by environmental sampling was 44% (95% CI: 20% to 70%) and 32% (95% CI: 13% to 57%) when the results of only 1 random herd visit were considered, although the difference between 1 and 2 herd visits was not significant (P = 0.5). Among infected herds, more herds were found positive at both samplings based on individual tests (IFC and ELISA) compared to EC (Table III). At the individual level, 14 cows were IFC-positive at both samplings. Six cows that were IFC-negative in the summer were found to be IFC-positive in the fall.

Individual samplesA total of 3100 samples was tested from 1844 adult cows sampled

[summer only (n = 172), fall only (n = 416), and both seasons (n = 1256)]. The MAP pathogen was cultured from 21 IFC samples in the summer (1.5%; 95% CI: 0.9% to 2.2%) and from 24 IFC samples in the fall (1.4%; 95% CI: 0.9% to 2.1%). In total, 35 cows (2.5%; 95% CI: 1.7% to 3.4%) were seropositive in the summer and 49 in the fall (2.9%; 95% CI: 2.2% to 3.9%) (Table II). The proportion of positive samples per visit was not significantly different for IFC (P = 0.8) or ELISA (P = 0.4).

Environmental samplesOverall, MAP was recovered from 29 out of 392 environmental

cultures (EC) carried out during the study (7%; 95% CI: 5% to 11%) from 8 positive farms. In the summer, MAP was cultured from 12 out of 177 ECs (7%; 95% CI: 4% to 12%) from 5 infected herds. In the fall, 17 out of 215 ECs were positive (8%; 95% CI: 5% to 12%) from 5 infected herds. We did not observe a difference between the percentages of positive environmental samples per sampling period (P = 0.8).

Environmental sites — The MAP pathogen was isolated from 14% of the boot samples, 11% of the sick cow group samples, 9% of the manure pit samples, 8% of the gutter samples, 5% of the samples from the group with low BCSs, and 1% of the samples from the heifers’ area. There were MAP-infected herds identified by the manure pit, gutter, and sick cow sites. From the gutter alone, 3 MAP-infected herds were detected (Table IV). If only 2 sites were sampled, the best sampling option was to combine the samples from the manure pit with either the samples from the sick cow group or the boots because 6 MAP-infected herds were detected with each combination. In order to detect all the herds found positive by EC during the study, a combination of a minimum of 4 sites was

2 15

37

Environmental and targeted pooled culture

Serum-ELISA

Individual fecal culture

Figure 1. Of 24 herds, identification of those herds positive for MAP at either sampling using 3 detection methods.



required (the manure pit, the sick cows, the cows with low BCSs, and either the gutter or the boots). When MAP was cultured from the boots, there was at least one other positive environmental sample on the farm.

Consecutive negative EC results — Among herds that tested negative by EC at both samplings, the odds of not detecting a cow shedding MAP tended to be greater [odds ratio (OR) = 5.4; 95% CI: 0.9 to 38.2; P = 0.06] compared to herds found positive by EC in at least 1 of 2 samplings.

D i s c u s s i o nWhen the results of 2 visits were combined, the sensitivity of the

standardized environmental and targeted pooled sampling technique proposed in this paper was within the range of the values reported in the literature (23–25). Recently, 2 Canadian studies reported a higher sensitivity of environmental sampling (68% to 71%) when it was conducted quarterly (22,25). It is expected that repeated samplings

may increase the capacity of the environmental sampling technique to detect infected herds. On the other hand, herds with 2 consecutive negative EC results were more likely to have no cows shedding MAP. Even if a negative EC result does not guarantee that the farm is not infected, that result may indicate a negative or a low-prevalence herd (17). Repeated negative samplings may increase confidence that the farm has a very low prevalence or is MAP-negative.

One study did not find positive ECs when within-herd prevalence (WHP) was 2% or less (34). In the present study, positive ECs were observed in herds without positive IFCs, as previously reported (21,25). Intermittent fecal shedding may explain the absence of con-current positive IFCs in the infected herds that tested positive by EC (35). Another hypothesis is that shedder cows had been culled before the herd visit and MAP remained in the environment of the farm. One study suggested that EC may be a measure of the persistency of MAP on farms even if no individual cows are positive by concur-rent IFC, indicating that some environmental contamination remains despite the reduction in prevalence (34). Because of the bacterium’s

Table II. Herd characteristics and WHP of MAP (estimated using IFC or serum ELISA) and number of positive environmental and targeted pooled cultures for 2 sampling seasons in 24 tie-stall dairy herds in Quebec

Summer Fall Number Herd WHPb (%) Number of Herd WHPb (%) Number of of animalsStatusa size IFCc ELISAd positive EC-TPSe size IFCc ELISAd positive EC-TPSe sampled twice1 179 1.1 0.6 2 178 0.6 1.7 0 1651 38 0 0 1 45 0 2.2 0 371 52 3.6 3.8 2 55 3.6 3.6 6 441 35 28.6 31.4 6e 45 20.0 17.8 7 291 97 0 1.0 1 100 2.0 3.0 0 941 76 3.9 3.9 0 79 3.8 6.3 2 671 52 0 0 0 52 0 3.8 1 381 NS NS NS NS 72 5.6 2.8 1 —1 45 0 0 0 48 2.1 2.1 0 361 60 3.3 3.3 0 57 0 1.7 0 531 210 0 1.4 0 211 0 4.2 0 1901 64 0 3.1 0 55 0 0 0 441 41 0 9.8 0 43 0 9.3 0 361 63 0 3.2 0 62 0 1.6 0 541 39 0 2.6 0 39 0 0 0 311 NS NS NS NS 30 0 3.3 0 —1 42 0 0 0 46 0 0 0 411 NS NS NS NS 46 0 0 0 —1 42 4.8 4.8 0 44 4.5 2.3 0 392 90 0 0 0e 92 0 0 0e 872 71 0 1.4 0 73 0 1.3 0 602 NS NS NS NS 72 0 1.4 0 —2 69 0 0 0 65 0 1.5 0 582 63 0 0 0 63 0 3.2 0 53a Status: (1) Infected (2) Not infected.b Within herd prevalence.c Individual fecal culture.d Herds were considered positive if ELISA WHP was $ 2% (bold characters).e Environmental and targeted pooled cultures; 7 environmental and 2 targeted pooled samples (TPS) were taken during all herd visits except 3.NS — Not sampled.



ability to survive in the environment for up to 11 mo under optimal conditions, EC has the potential to detect MAP in herds even after infected animals have been culled or in the presence of intermittent shedding. This is a strong advantage of any technique used at any particular point in time.

Overall, no statistically significant difference was observed between the proportion of infected herds identified by EC, IFC, or ELISA. The absence of a statistically significant difference, however, could be a consequence of the low power of the study. In the present study, ELISA misclassified 1 non-infected herd as positive, according to our case definition. This herd may actually be an infected herd that our case definition failed to classify or a non-infected herd that ELISA misclassified due to a lack of specificity.

The choice of detection method depends on the objectives estab-lished by the owners or veterinarians. The environmental culture

(EC) is the most cost-effective option to determine MAP-herd status (17,18). In the context of the QVPPCP, where the main objective is to detect high-prevalence herds (those herds with the most important economic losses), EC is the most appropriate sampling alternative. Screening a whole herd with individual tests (IFC or ELISA) is more invasive, expensive, and time-consuming than EC. Individual tests are more suitable for identifying infected animals within a MAP-infected herd. At the herd level, our study indicates that all tests (EC, IFC, and ELISA) give different results at different points of time. The demographic changes in the herds could explain such variability.

In the present study, MAP was cultured from the manure pit and the gutter, which is where manure accumulates in tie-stall farms. These locations are traditionally chosen as sampling sites because they have been proven to have a high sensitivity for detecting MAP-infected free-stall herds (19–21,36). The MAP pathogen was also

Table III. Combinations of test results for either a single visit or 2 visits to detect MAP in tie-stall dairy herds in Quebec

Number of Sampling option Number of positive herdsb Sensitivity (CI)samplings (Number of infected herdsa) EC-TPS IFC ELISA EC-TPS IFC ELISASingle Summer (16) 5 6 9 31 (11 to 59) 38 (15 to 65) 56 (30 to 80)sampling Fall (19) 5 8 12 26 (9 to 51) 42 (20 to 67) 63 (38 to 84) One random samplingc (19) 6 7 9 32 (13 to 57) 37 (16 to 62) 47 (24 to 71)

Two Summer or falle (16) 7 8 14 44 (20 to 70) 50 (25 to 75) 88 (62 to 98)samplingsd Summer and fallf (16) 2 5 5 13 (2 to 38) 31 (11 to 59) 31 (11 to 59)a Infected herd: MAP was cultured from at least 1 sample (IFC or EC) for 30 mo including the duration of the study period.b Herds were considered positive through environmental culture and targeted pooled sampling (EC-TPS) strategy and individual fecal culture (IFC) if at least 1 positive sample was obtained. For ELISA, herds were considered positive if within-herd prevalence (WHP) was $ 2%.c Nineteen infected herds had available results in at least 1 sampling season. From the herds visited twice, 1 sampling was randomly chosen.d Sixteen infected herds had available test results for both seasons.e Herds meeting the criteria for a positive herd according to each diagnostic test, either in summer or fall.f Herds meeting the criteria for a positive herd according to each diagnostic test, in both summer and fall.

Table IV. Distribution of MAP-positive sites on 8 farms tested by environmental and targeted pooled sampling

Summer Fall Number Within-herd Number Within-herdEC-TPS Number of positive samples by siteb of positive prevalence of positive prevalencepositive herdsa MP G H B S P EC-TPSa IFCc ELISA EC-TPSa IFCc ELISAA 0 1 0 1 0 0 2 1.1 0.6 0 0.6 1.7B 1 0 0 0 0 0 1 0 0 0 0 2.2Cd 2 2 0 2 2 0 2 3.6 3.8 6 3.6 3.6Dd 4 4 1 2 1 1 6 28.6 31.4 7 20.0 17.8E 0 0 0 0 0 1 1 0 1.0 0 2.0 3.0F 0 0 0 1 1 0 0 3.9 3.9 2 3.9 6.3G 1 0 0 0 0 0 0 0 0 1 0 3.8H 0 0 0 0 1 0 NS NS NS 1 5.6 2.8

Number of herds 4 3 1 4 4 2 detected by sitea Positive herds by environmental culture and targeted pooled sampling (EC-TPS) strategy in either summer or fall.b MP — Manure pit; G — gutter; H — heifers’ area; B — boots; S — sick cows group; P — group of cows with poor body condition.c IFC — Individual fecal culture.d Positive farms both in summer and fall.NS — Not sampled.



cultured from samples collected from boots, the sick cow group, cows with low BCS, and the heifers’ area. While boot sampling was previously found to be a sensitive technique for detecting high-prevalence herds (37), it needs to be evaluated in low-prevalence herds as these authors suggest. Another study suggested includ-ing boots as an additional sample in the environmental sampling strategy (38). Although boot sampling (either from the owner or a researcher) is interesting, it is both recommended and expected that owners disinfect and/or change boots as they move from one area on the farm to another. This would be particularly important when different age groups are visited. As for researchers, the same rule should apply so they cannot become a potential risk of dissemination of a pathogen within a herd.

Sampling cows with low BCS was previously suggested for screening beef cattle herds for MAP (17). The pools proposed in our study may be collected by the veterinarian during a herd health visit with minimal additional animal handling. In contrast, a sample from the heifers’ area does not seem to increase the sen-sitivity of the technique, as MAP was cultured from only 1 sample throughout the study. This finding demonstrates that young ani-mals may be exposed to and shed MAP, as suggested in a previous study (39). This positive sample was collected from the herd with the highest prevalence (fecal WHP, ELISA WHP, and EC preva-lence). Our study suggests that a combination of samples from the manure pit or the gutter, the sick cows, and/or the cows with low BCS may be an effective strategy to detect MAP-infected tie-stall dairy herds.

It has been suggested that the sensitivity of environmental sam-pling is expected to be higher in high-prevalence herds (14). Another study found that sensitivity may be close to 100% even in moderate-prevalence herds (when the WHP is 8% or greater) (5). Some factors may have affected the sensitivity of the environmental sampling technique in our study. The low-prevalence herds included were purposively selected from the QVPPCP list because of the owner’s willingness to participate in research projects. It can be assumed that these producers are more aware of bovine paratuberculosis than producers who did not participate in the program. The exposure of QVPPCP herds to several years of veterinarian recommendations in order to control MAP infections may have contributed to the lower MAP prevalence. Additionally, in our study several herds that were initially considered to be infected turned out to be either negative or very low-prevalence herds. Possible reasons for this inconsistency may have been the culling of cows (either for paratuberculosis or other reasons) or the delay between the positive diagnosis classifying the herd as infected and the beginning of the study. If at the time of the positive test, the herd had low prevalence and biosecurity mea-sures had been introduced to reduce transmission (as is supposed to be done for herds enrolled in the QVPPCP), it is very likely that the WHP decreased or at least remained at the same level. Although the sensitivity of EC tended to increase when 2 samplings were considered instead of 1, our sample size did not allow us to observe a significant difference.

In conclusion, the proposed standardized environmental and targeted pooled sampling technique was a useful diagnostic method for detecting the MAP-infected tie-stall dairy herds in this study. This inexpensive and non-invasive method detected mainly high-

prevalence herds, but it did detect low-prevalence herds as well. Repeated samplings may increase sensitivity for detecting low-prevalence herds and confidence in a negative result. Our sampling strategy proposes that new sample types, compared to current methods, be included to improve MAP detection.

A c k n o w l e d g m e n t sThe authors thank the producers for participating in this study

and the veterinarians for helping with the recruitment process. This project was funded by Novalait Inc.; Agriculture and Agri-Food Canada; Ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec; and Fonds de recherche du Québec-nature et technologies.

Re f e r e n c e s 1. Kudahl AB, Nielsen SS. Effect of paratuberculosis on slaughter

weight and slaughter value of dairy cows. J Dairy Sci 2009; 92:4340–4346.

2. Hendrick SH, Kelton DF, Leslie KE, Lissemore KD, Archambault M, Duffield TF. Effect of paratuberculosis on cull-ing, milk production, and milk quality in dairy herds. J Am Vet Med Ass 2005;227:1302–1308.

3. Lombard JE, Garry FB, McCluskey BJ, Wagner BA. Risk of removal and effects on milk production associated with para-tuberculosis status in dairy cows. J Am Vet Med Assoc 2005; 227:1975–1981.

4. Hermon-Taylor J. Protagonist. Mycobacterium avium subspe-cies paratuberculosis is a cause of Crohn’s disease. Gut 2001;49: 755–756.

5. Scanu AM, Bull TJ, Cannas S, et al. Mycobacterium avium sub-species paratuberculosis infection in cases of irritable bowel syndrome and comparison with Crohn’s disease and Johne’s disease: Common neural and immune pathogenicities. J Clin Microbiol 2007;45:3883–3890.

6. Waddell LA, Rajic A, Sargeant J, et al. The zoonotic potential of Mycobacterium avium ssp. paratuberculosis: A systematic review. Can J Public Health 2008;99:145–155.

7. Momotani E, Ozaki H, Hori M, et al. Mycobacterium avium subsp. paratuberculosis lipophilic antigen causes Crohn’s disease-type necrotizing colitis in mice. Springerplus 2012;1:47.

8. Momotani E, Romona NM, Yoshihara K, et al. Molecular patho-genesis of bovine paratuberculosis and human inflammatory bowel diseases. Vet Immunol Immunopathol 2012;148:55–68.

9. Naser SA, Sagramsingh SR, Naser AS, Thanigachalam S. Mycobacterium avium subspecies paratuberculosis causes Crohn’s disease in some inflammatory bowel disease patients. World J Gastroenterol 2014;20:7403–7415.

10. Manning EJB, Collins MT. Epidemiology of paratuberculosis. In: Behr MA, Collins DM, eds. Paratuberculosis: Organism, Disease, Control. Wallingford: CAB International, 2010:22–28.

11. Doré E, Paré J, Côté G, et al. Risk factors associated with transmission of Mycobacterium avium subsp. paratuberculosis to calves within dairy herd: A systematic review. J Vet Intern Med 2012;26:32–45.



12. Windsor PA, Whittington RJ. Evidence for age susceptibility of cattle to Johne’s disease. Vet J 2010;184:37–44.

13. Streeter RN, Hoffsis GF, Bech-Nielsen S, Shulaw WP, Rings DM. Isolation of Mycobacterium paratuberculosis from colostrum and milk of subclinically infected cows. Am J Vet Res 1995;56: 1322–1324.

14. Nielsen SS, Ersboll AK. Age at occurrence of Mycobacterium avium subspecies paratuberculosis in naturally infected dairy cows. J Dairy Sci 2006;89:4557–4566.

15. Eisenberg SW, Nielen M, Santema W, Houwers DJ, Heederik D, Koets AP. Detection of spatial and temporal spread of Mycobacterium avium subsp. paratuberculosis in the environment of a cattle farm through bio-aerosols. Vet Microbiol 2010;143: 284–292.

16. Whittington RJ, Marshall DJ, Nicholls PJ, Marsh IB, Reddacliff LA. Survival and dormancy of Mycobacterium avium subsp. paratuberculosis in the environment. Appl Environ Microbiol 2004;70:2989–3004.

17. Collins MT, Gardner IA, Garry FB, Roussel AJ, Wells SJ. Consensus recommendations on diagnostic testing for the detection of paratuberculosis in cattle in the United States. J Am Vet Med Assoc 2006;229:1912–1919.

18. Sweeney RW, Collins MT, Koets AP, McGuirk SM, Roussel AJ. Paratuberculosis (Johne’s disease) in cattle and other susceptible species. J Vet Intern Med 2012;26:1239–1250.

19. Berghaus RD, Farver TB, Anderson RJ, Jaravata CC, Gardner IA. Environmental sampling for detection of Mycobacterium avium ssp. paratuberculosis on large California dairies. J Dairy Sci 2006;89:963–970.

20. Raizman EA, Wells SJ, Godden SM, et al. The distribution of Mycobacterium avium ssp. paratuberculosis in the environment sur-rounding Minnesota dairy farms. J Dairy Sci 2004;87:2959–2966.

21. Lombard JE, Wagner BA, Smith RL, et al. Evaluation of environ-mental sampling and culture to determine Mycobacterium avium subspecies paratuberculosis distribution and herd infection status on US dairy operations. J Dairy Sci 2006;89:4163–4171.

22. Wolf R, Barkema HW, De Buck J, et al. High herd-level preva-lence of Mycobacterium avium subspecies paratuberculosis in Western Canadian dairy farms, based on environmental sam-pling. J Dairy Sci 2014;97:6250–6259.

23. Pillars RB, Grooms DL, Woltanski JA, Blair E. Prevalence of Michigan dairy herds infected with Mycobacterium avium subspe-cies paratuberculosis as determined by environmental sampling. Prev Vet Med 2009;89:191–196.

24. Smith RL, Schukken YH, Pradhan AK, et al. Environmental contamination with Mycobacterium avium subsp. paratuberculosis in endemically infected dairy herds. Prev Vet Med 2011;102:1–9.

25. Lavers CJ, McKenna SL, Dohoo IR, Barkema HW, Keefe GP. Evaluation of environmental fecal culture for Mycobacterium avium subspecies paratuberculosis detection in dairy herds and association with apparent within-herd prevalence. Can Vet J 2013;54:1053–1060.

26. Nombre moyen de vaches laitières par ferme et par province au 1er janvier, Canada. Groupe AGÉCO. Canada [updated 2013 June 27]. Available from: http://groupeageco.ca/fr/pdf/stat/PCL3.pdf Last accessed May 18, 2016.

27. Aly SS, Anderson RJ, Whitlock RH, et al. Reliability of envi-ronmental sampling to quantify Mycobacterium avium subspe-cies paratuberculosis on California free-stall dairies. J Dairy Sci 2009;92:3634–3642.

28. Harris B, Robbe-Austerman S, Dykema P, Fett K, Stuber T, Stokes K. Laboratory methods for isolation and identification of Mycobacterium avium ssp. paratuberculosis (Johne’s Disease). Ames: USDA National Veterinary Services Laboratories, 2009.

29. Collins MT. Interpretation of a commercial bovine paratubercu-losis enzyme-linked immunosorbent assay by using likelihood ratios. Clin Diagn Lab Immunol 2002;9:1367–1371.

30. Lombard JE, Byrem TM, Wagner BA, McCluskey BJ. Comparison of milk and serum enzyme-linked immunosorbent assays for diagnosis of Mycobacterium avium subspecies paratuberculosis infection in dairy cattle. J Vet Diagn Invest 2006;18:448–458.

31. Collins MT, Sockett DC, Ridge S, Cox JC. Evaluation of a com-mercial enzyme-linked immunosorbent assay for Johne’s disease. J Clin Microbiol 1991;29:272–276.

32. Collins MT, Wells SJ, Petrini KR, Collins JE, Schultz RD, Whitlock RH. Evaluation of five antibody detection tests for diagnosis of bovine paratuberculosis. Clin Diagn Lab Immunol 2005;12:685–692.

33. Sweeney RW, Whitlock RH, Buckley CL, Spencer PA. Evaluation of a commercial enzyme-linked immunosorbent assay for the diagnosis of paratuberculosis in dairy cattle. J Vet Diagn Invest 1995;7:488–493.

34. Pillars RB, Grooms DL, Kaneene JB. Longitudinal study of the distribution of Mycobacterium avium subsp. paratuberculosis in the environment of dairy herds in the Michigan Johne’s disease control demonstration herd project. Can Vet J 2009;50:1039–1046.

35. Nielsen SS. Transitions in diagnostic tests used for detection of Mycobacterium avium subsp. paratuberculosis infections in cattle. Vet Microbiol 2008;132:274–282.

36. Wolf R, Barkema HW, De Buck J, Orsel K. Sampling location, herd size, and season influence Mycobacterium avium ssp. para-tuberculosis environmental culture results. J Dairy Sci 2015;98: 275–287.

37. Eisenberg T, Wolter W, Lenz M, Schlez K, Zschock M. Boot swabs to collect environmental samples from common locations in dairy herds for Mycobacterium avium ssp. paratuberculosis (MAP) detection. J Dairy Res 2013;80:485–489.

38. Donat K, Hahn N, Eisenberg T, et al. Within-herd prevalence thresholds for the detection of Mycobacterium avium subspecies paratuberculosis-positive dairy herds using boot swabs and liquid manure samples. Epidemiol Infect 2016;144:413–424.

39. Wolf R, Orsel K, De Buck J, Barkema HW. Calves shedding Mycobacterium avium subspecies paratuberculosis are common on infected dairy farms. Vet Res 2015;46:71.


Article


Association between thermal environment and Salmonella in fecal samples from dairy cattle in midwestern United States

Tasha Likavec, Alda F.A. Pires, Julie A. Funk

A b s t r a c tThe objective of this study was to describe the association between thermal measures in the barn environment (pen temperature and humidity) and fecal shedding of Salmonella in dairy cattle. A repeated cross-sectional study was conducted within a commercial dairy herd located in the midwestern United States. Five pooled fecal samples were collected monthly from each pen for 9 mo and submitted for microbiological culture. Negative binomial regression methods were used to test the association [incidence rate ratio (IRR)] between Salmonella pen status (the count of Salmonella-positive pools) and thermal environmental parameters [average temperature and temperature humidity index (THI)] for 3 time periods (48 h, 72 h, and 1 wk) before fecal sampling. Salmonella was cultured from 10.8% [39/360; 95% confidence interval (CI): 7.8% to 14.5%] of pooled samples. The highest proportion of positive pools occurred in August. The IRR ranged from 1.26 (95% CI: 1.15 to 1.39, THI 1 wk) to 4.5 (95% CI: 2.13 to 9.51, heat exposure 1 wk) across all thermal parameters and lag time periods measured. For example, the incidence rate of Salmonella-positive pools increased by 54% for every 5°C increment in average temperature (IRR = 1.54; 95% CI: 1.29 to 1.85) and 29% for every 5-unit increase in THI (IRR = 1.29; 95% CI: 1.16 to 1.42) during the 72 h before sampling. The incidence rate ratio for pens exposed to higher temperatures (. 25°C) was 4.5 times (95% CI: 2.13 to 9.51) the incidence rate ratio for pens exposed to temperatures , 25°C in the 72 h before sampling. Likewise, the incidence rate ratio for pens exposed to THI . 70 was 4.23 times greater (95% CI: 2.1 to 8.28) than when the THI was , 70 in the 72 h before sampling. An association was found between the thermal environment and Salmonella shedding in dairy cattle. Further research is warranted in order to fully understand the component risks associated with the summer season and increased Salmonella shedding.

R é s u m éL’objectif de cette étude est de décrire l’association entre les mesures thermiques de l’étable (température et humidité des enclos) et l’excrétion fécale de Salmonella chez les bovins laitiers. Une étude transversale répétée a été réalisée dans un troupeau laitier commercial situé dans la région du Midwest des États-Unis. 5 échantillons composites de fèces ont été récoltés de chaque stalle d’une manière mensuelle pendant neuf mois puis soumis pour culture microbiologique. Des méthodes de régression binomiale négative ont été utilisées pour tester l’association (ratio de taux d’incidence, IRR) entre la présence de Salmonella dans les enclos (nombre d’échantillons positifs à Salmonella) et les paramètres environnementaux [température moyenne, index humidité température (THI) pour 3 périodes (48 h, 72 h, 1 semaine)]. Salmonella a été cultivée de 10,8 % (39/360; I.C. 95 % 7,8 %–14,5 %) des échantillons composites. La plus grande proportion d’échantillons positifs ont été collectés durant le mois d’août. Le rapport des taux d’incidence (IRR) a varié de 1,26 (IIR = 1,26; I.C. à 95 % 1,15 à 1,39) à 4,5 (IRR = 4,5; I.C. à 95 % 2,13 à 9,51) pour tous les paramètres thermiques et les des périodes étudiés. Par exemple, l’augmentation d’échantillons Salmonella positifs est de 54 % par incrément de cinq °C de température moyenne (IRR = 1,54; I.C. à 95 % 1,29 à 1,85) et de 29 % pour chaque augmentation de cinq unités de THI (IRR = 1,29; IC à 95 % 1,16 à 1,42) dans les 72 heures avant l’échantillonnage. Le ratio de taux d’incidence pour les enclos exposés à de hautes temperatures (. 25 °C) était 4,5 fois (I.C. à 95 % 2,13 à 9,51) supérieur au ratio de taux d’incidence des enclos exposés à des temperatures , 25 °C. De même, Le ratio de taux d’incidence des enclos exposés à des THI . 70 est de 4,23 (I.C. à 95 % 2,16–8,28) superieur a ceux dont le THI est , 70 dans les 72 heures précédant l’échantionnage. Une association a été trouvée entre les mesures thermiques et l’excrétion de Salmonella chez les bovins laitiers. Plus de recherches sont requises pour comprendre entièrement les risques associés à la saison estivale et à l’excrétion de Salmonella.

(Traduit par les autuers)

Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Michigan State University, East Lansing, Michigan 48824, USA.

Dr. Likavec’s current address is Department of Large Animal Clinical Sciences, College of Veterinary Medicine & Biomedical Sciences, Texas A&M University, College Station, Texas 77845, USA.

Address all correspondence to Dr. Alda F.A. Pires; telephone: (530) 754-9855; fax: (530) 752-7181; e-mail: [email protected]

Dr. Pires’ current address is Department of Population Health and Reproduction, School of Veterinary Medicine, University of California, 1089 Veterinary Medicine Drive, Davis, California 95616, USA.

Conflict of interest statement All authors declare that there is no conflict of interest involved in their participation in this study.

Received October 21, 2015. Accepted February 10, 2016.



I n t r o d u c t i o nSalmonellosis is a zoonotic disease of significant clinical concern

to both cattle and humans (1,2). Subclinical shedding of Salmonella on farms is common (2,3). Several foodborne outbreaks of salmonellosis have been linked to ground beef (4) and unpasteurized dairy prod-ucts (5). Identification of on-farm interventions to control Salmonella shedding may therefore mitigate disease in both cattle and humans.

Increased Salmonella shedding has been reported in cattle during the summer months (3,6–9). Thermal stress is one component of the potential causal pathway for the association between season and Salmonella shedding. There are limited investigations of the effect of heat stress on Salmonella in dairy cattle (10–12). These studies were limited by the study design, i.e., cross-sectional, and the approach to measurement of the thermal environment. In addition, by sam-pling the same animal in the coolest and hottest period of the day, assuming an exceptionally short effect of thermal stress (10,11), or by an absence of measurement of the thermal environment (12), the effect of heat stress on Salmonella shedding was not clear. The overall objective of this observational study was to describe the association between thermal measures in the barn environment (pen tempera-ture and humidity) and fecal shedding of Salmonella in dairy cattle.

M a t e r i a l s a n d m e t h o d sA repeated cross-sectional study was conducted in a commercial

dairy herd (~3000 milking cows) located in the midwestern United States. The animals were housed in 2-row, free-stall pens with sand bedding. The outcome of interest (count of Salmonella-positive pools) was sampled by pen, each of which held cows at different lactation phases. Sampling was conducted by lactation phase in order to con-trol for the well described differential risk for Salmonella shedding by lactation phase (8,10,13). The sampled pens were: i) close-up cows (2 to 3 wk before calving); ii) pre-calving heifers (2 to 3 wk before calving); iii) fresh/hospital pen (cows and heifers up to 1 wk post-calving and animals under antibiotic therapy); iv) heifers 30 d in milk (DIM) (average of DIM); v) cows 24 DIM; vi) early-to-mid lactation heifers (150 DIM); vii) early-to-mid lactation cows (80 DIM); and viii) late-lactation cows and heifers (214 DIM).

Each pen was sampled monthly for 9 mo (from July 2010 to March 2011). During each period, 5 pools of 25 g of pooled fecal samples were collected from each pen. Each pool (25 g of feces/pool) consisted of five 5-g fecal samples collected from the fresh manure on pen floors, e.g., separate fresh fecal pats. Forty pools (8 pens times 5 pools per pen) were collected each month, resulting in a total sample size of 360 pools. Each pooled sample was col-lected using new gloves, placed into labeled specimen containers (VWR International, Radnor, Pennsylvania, USA), and transported in a cooler with frozen ice packs. Samples were mixed with a sterile spatula, stored at 4°C overnight, and processed the following day.

Samples were sent to the Diagnostic Center for Population and Animal Health at Michigan State University for microbiological cul-ture using standard methods described elsewhere (14). Briefly, fecal samples (25 g) were inoculated into 225 mL of tetrathionate broth (TTB; Becton Dickinson, Sparks, Maryland, USA) and incubated at 37°C for 48 h. After incubation, an aliquot (100 mL) of the fecal-TTB

solution was inoculated into 10 mL of rappaport-vassiliadis broth (RV; Becton Dickinson) and incubated at 42°C for 24 h. The RV broth was plated onto xylose-lysine-tergitol-4 agar selective agar plates (XLT4; Thermo Fisher Scientific, Lenexa, Kansas, USA) and incubated at 37°C overnight. Suspect Salmonella colonies from the XLT4 were screened using triple sugar iron agar slants and urea agar slants (Becton Dickinson). Salmonella-suspect colonies were screened using Salmonella Poly O Antisera agglutination (Becton Dickinson).

Temperature and humidity were recorded every 5 min, 24 h/d for the entire study period using commercially available data loggers (Hobo U23; Onset HOBO Data Loggers, Bourne, Massachusetts, USA). Two data loggers were placed in both extremities of each pen (8 pens corresponding to the 8 lactation phases and a total of 16 data loggers), fixed to a post above cow level and protected from direct sunlight. Data were manually downloaded monthly from the data loggers. A macro of commercially available software (Excel 2007; Microsoft, Redmond, Washington, USA) was used to process temperature and humidity data and to calculate the environmental parameters for each pen for the measurement period. The average values from the 2 data loggers were estimated for each pen. For all parameters, average values were calculated for every pen for 3 time periods before the time of fecal sampling (lag times of 48 h, 72 h, and 1 wk).

The following thermal environmental parameters were calculated: i) average temperature; ii) average temperature humidity index (THI) (15); iii) heat exposure, defined as any time in each respective lag time that the temperature in the pen was above the upper critical temperature of the thermal neutral zone (TNZ) (. 25°C) for adult dairy cows (16); and iv) heat index exposure defined as any time that the THI in the pen was above 70 (17). The temperature humidity index used in this study was:

THI = 0.63twb 1 1.17tdb 1 32

where: tdb and twb are the dry and wet bulb temperatures of the ambient air in °C (15).

All statistical analyses were carried out using a commercially available statistics package (SAS 9.3; SAS Institute, Cary, North Carolina, USA). To describe the frequency of positive pools, Salmonella prevalence (proportion of positive samples/tested) and respective 95% confidence intervals (CI) were estimated. Logistic models [generalized estimated equations (GEE); PROC GENMOD; SAS 9.3; SAS Institute], which accounted for non-independence of samples within pen, were used to estimate adjusted-prevalence for each production type, e.g., pen, and month, e.g., collection period, and to compare the proportion of Salmonella-positive pools among production phase types and months. The Bonferroni correction was used for multiple comparisons.

Negative binomial regression models were used to test the asso-ciation between thermal environmental variables and Salmonella pen status (the count of Salmonella-positive pools at each collection period). Correlations between the independent thermal environ-mental variables were assessed based on Pearson’s and Spearman’s coefficients depending on whether the normality condition was met. If the value of the correlation statistic between 2 independent variables was equal to or greater than 0.8 at P # 0.05, independent models were built for each environmental variable for each time



period before a collection event (48 h, 72 h, and 1 wk). The number of positive pools per pen and visit (within pen) was the outcome of interest. Robust standard errors for the negative binomial coef-ficients were estimated by using repeated statement (GEE; PROC GENMOD; SAS 9.3; SAS Institute). Incidence rate ratios (IRRs) and their respective 95% CI were estimated for each main exposure of interest, i.e., thermal environmental parameter (17,18). The IRR was defined by the percent change in positive pools for each 5°C increase for average temperature and for each 5-unit THI change. For the heat and heat index exposure variables, the IRR reflected exposure

(yes/no) to a temperature greater than the upper limit of the TNZ (25°C) and THI . 70, respectively. A P , 0.05 was considered sta-tistically significant.

Re s u l t sOverall, Salmonella was cultured from 10.8% [39/360, standard

error (SE) = 1.64; 95% CI: 7.8% to 14.5%] of pooled samples. The pro-portion of collection periods with at least 1 positive pool was highest for the pen of pre-calving heifers (66.7%) and lowest for the pens

Table I. Summary statistics for thermal environmental parameters before sample collection in free-stall barns from July 2010 to March 2011

Thermal environmental parameters July August September October November December January February MarchAverage temperature (°C)48 h Mean (SEM)a 23.3 (0.10) 24.5 (0.07) 19.5 (0.08) 14.2 (0.04) 3.4 (0.06) 0.6 (0.21) 22.1 (0.16) 25.6 (0.17) 21.7 (0.19) Minimum 22.9 24.2 19.2 14.1 3.2 23.0 22.8 26.2 22.4 Maximum 23.8 24.9 19.8 14.4 3.8 1.6 21.3 24.6 20.8

72 h Mean (SEM)a 23.4 (0.09) 24.3 (0.07) 21.0 (0.07) 15.0 (0.04) 4.0 (0.05) 0.2 (0.20) 2.4 (0.17) 26.1 (0.13) 20.5 (0.16) Minimum 23.1 24.0 20.7 14.9 3.7 20.6 1.8 26.5 21.0 Maximum 24.0 24.6 21.4 15.2 4.05 1.3 3.1 25.4 0.3

1 wk Mean (SEM)a 25.0 (0.08) 23.9 (0.07) 24.0 (0.05) 13.4 (0.04) 4.7 (0.03) 2.7 (0.36) 1.7 (0.21) 25.4 (0.12) 0 (0.16) Minimum 24.8 23.6 23.9 13.3 4.6 2.0 1.1 25.8 20.5 Maximum 25.5 24.2 24.3 13.6 4.9 3.6 2.6 24.8 0.9

Temperature humidity index48 h Mean (SEM)a 72.0 (0.16) 74.6 (0.10) 68.6 (0.13) 56.9 (0.15) 38.2 (0.10) 32.0 (0.36) 31.9 (0.33) 25.0 (0.20) 29.4 (0.29) Minimum 71.5 74.2 68.1 56.4 37.6 30.8 30.3 24.1 28.5 Maximum 72.9 75.1 69.2 57.9 38.5 33.9 33.6 25.8 31.1

72 h Mean (SEM)a 72.0 (0.15) 72.4 (0.10) 70.5 (0.11) 57.1 (0.15) 38.4 (0.10) 31.9 (0.33) 38.0 (0.32) 24.0 (0.15) 30.2 (0.29) Minimum 71.4 72 70.0 56.7 37.8 30.8 36.6 23.4 29.3 Maximum 72.8 72.8 71 58.0 38.8 33.6 39.4 24.7 31.7

1 wk Mean (SEM)a 75.2 (0.13) 73.0 (0.12) 73.2 (0.10) 54.9 (0.12) 39.7 (0.10) 36.0 (0.30) 34.9 (0.36) 25.1 (0.14) 31.7 (0.30) Minimum 74.6 72.6 72.8 54.6 39.4 35.1 33.8 24.8 30.8 Maximum 75.9 73.6 73.7 55.6 40.4 37.4 36.5 26.0 33.4

Proportion of 7.5d 47.5c 12.5d 15.5d 7.5d 0.00 0.00 5d 2.5d Salmonella- (2.4 to 20.8) (32.7 to 62.7) (5.3 to 26.7) (6.9 to 30.0) (2.4 to 20.8) (1.2 to 17.9) (3.5 to 15.7) positive pools (%) (95% CI)b

a Standard error of the mean (SEM).b Proportion of positive fecal pools (n = 40) per collection period, adjusted for clustering (within pen) using generalized estimated equations (GEEs). Different superscript letters (c and d) indicate a significant difference (P , 0.05) of proportion of positive fecal pool samples, after Bonferroni correction.



of early-to-mid lactation heifers and late-lactation cows and heifers (11.1%), followed by the fresh/hospital pen (55.6 %), close-up cows pen and heifers 30 DIM pen (33.3%), and cows 24 DIM pen (22.2%).

Within pen, the adjusted proportion of Salmonella-positive pools per pen over the study period ranged from 0% to 22% (n = 45 pools per pen): i) close-up cows (8.9%; 95% CI: 3.4% to 21.4%); ii) pre-calving heifers (22.2%; 95% CI: 12.4% to 36.6%); iii) fresh/hospital pen (22.2%; 95% CI: 12.4% to 36.6%); iv) heifers 30 DIM (13.3%; 95% CI: 6.1% to 26.7%); v) cows 24 DIM (15.6%; 95% CI: 7.6% to 29.2%); vi) early-to-mid lactation heifers (2.2%; 95% CI: 3.1% to 14.2%); vii) early-to-mid lactation cows (0%); and viii) late-lactation cows and heifers (2.2%; 95% CI: 3.1% to 14.2%).

Within the collection period, the adjusted proportion of Salmonella-positive pools per month (n = 40 pools/mo) ranged from 0% to 47.5% (Table I). Although there was a numerical difference in the overall proportion of positive samples by lactation phase, there was no sta-tistical difference (P . 0.05). The highest proportion of positive pools occurred in August. There was a significant difference in the propor-tion of Salmonella-positive pools between August and the months of July through November, February, and March (P , 0.05) (Table I). All other months were not statistically different from each other.

Associations were evaluated using multiple univariable models, rather than multivariable models due to correlations among the

independent variables. The thermal environmental variables inves-tigated in this study were significantly associated with the risk of a pool being Salmonella-positive. Separate models for each thermal exposure (total of 12 models) are shown in Table II. The incidence rate ratio (IRR) ranged from 1.26 (95% CI: 1.15 to 1.39) to 4.5 (95% CI: 2.13 to 9.51) across all thermal parameters and lag time periods measured. For the continuous independent variables (average hourly temperature and THI), the IRR is interpreted based on each 5-unit increase in the thermal parameter. For example, the incidence rate ratio of Salmonella-positive pools increased 54% for every 5°C incre-ment in average temperature (IRR = 1.54; 95% CI: 1.29 to 1.85) and 29% for every 5-unit increase in THI (IRR = 1.29; 95% CI: 1.16 to 1.42) during the 72 h before sampling. The incidence rate ratio for pens exposed to heat (temperature . 25°C) was 4.5 times (95% CI: 2.13 to 9.51) the incidence rate for pens exposed to temperatures , 25°C during the 72 h before sampling. Likewise, the incidence rate for pens exposed to THI . 70 is 4.23 times greater (95% CI: 2.16 to 8.28) than when the THI was , 70 in the 72 h before sampling.

D i s c u s s i o nThis pilot study found an association between the thermal envi-

ronment and Salmonella shedding in dairy cattle. Previous studies

Table II. Estimates of negative binomial regression coefficients (b), standard errors (SE), incidence rate ratio (IRR), and 95% confidence intervals (CIs) for the models of association between thermal environmental parameters and Salmonella pen status (number of positive pools per pen and collection period) in a free-stall dairy cattle herd, sampled from July 2010 to March 2011

Models (thermal environmental parameters) Beta SE IRR 95% CI P-valueAverage temperaturea

48 h 0.44 0.09 1.55 1.30 to 1.84 , 0.0001 72 h 0.44 0.09 1.54 1.29 to 1.85 , 0.0001 1 wk 0.40 0.08 1.49 1.26 to 1.76 , 0.0001

Temperature humidity indexb

48 h 0.25 0.05 1.29 1.16 to 1.42 , 0.0001 72 h 0.25 0.05 1.29 1.16 to 1.42 , 0.0001 1 wk 0.23 0.05 1.26 1.15 to 1.39 , 0.0001

Heat exposurec

48 h 1.36 0.36 3.88 1.9 to 7.93 , 0.0001 72 h 1.5 0.38 4.5 2.13 to 9.51 , 0.0001 1 wk 1.5 0.38 4.5 2.13 to 9.51 , 0.0001

Temperature humidity index exposured

48 h 1.44 0.34 4.23 2.16 to 8.28 , 0.0001 72 h 1.44 0.34 4.23 2.16 to 8.28 , 0.0001 1 wk 1.44 0.46 4.23 1.72 to 10.4 , 0.0001a 5°C increase in temperature.b Increase of 5 in temperature humidity index.c Reference temperature # 25°C.d Reference temperature humidity index less than or equal to 70.



have described a seasonal pattern of Salmonella shedding in dairy cattle, with the highest prevalence in summer (3,6–9). Season is a proxy measure of many potential risks in the causal pathway for Salmonella shedding, e.g., meteorological factors and changes in diet and farm management practices (7). Although the thermal environment is frequently assumed to be a major component of this seasonal effect for Salmonella shedding, few studies have evaluated thermal risk factors and pathogen shedding in cattle (10,11). In both previous studies, no association was reported between exposure to heat stress and Salmonella, but 100% of the cows were shedding Salmonella during the study (11). Also, no difference was found in heat stress during sampling times (10), which limits evaluation of the exposure. In addition, the design of the present study differs in terms of frequency of sampling (monthly sampling versus twice-daily sampling), recording of thermal parameters (over the entire period versus at the time of sampling), thermal parameters measured (THI and average temperature versus THI only), lag times investigated, length of the study (several seasons versus 2 replicates within 2 wk or 1 mo), and the type of sample collected (pooled versus individual).

The present study is limited as it represents only 1 farm, in 1 U.S. state over a 9-month period, although this farm is very typical of large dairy farms in the upper midwest. The pooled sample collec-tion limited our ability to include other potential risk factors and confounders in the model beyond stratifying for lactation phase. Despite these limitations, this study extends the work of the previ-ous research and provides support for further evaluation of the association between thermal stress and Salmonella shedding at the individual cow level.

When interpreting these results, it is critical to understand that the association with thermal environmental parameters may be a component of a complex causal pathway. The associations with the thermal environment, like the association with season, may be a proxy measure of other component risks for Salmonella shedding. Despite this, the study herein provides sufficient evidence to warrant further investigation of the role of thermal environment on the causal pathway for Salmonella shedding in dairy cattle. Further research is recommended to investigate the role of the thermal environment on Salmonella shedding in a multi-herd study, with an effort to define the component causes of Salmonella shedding that may be a compo-nent, or represented, by the risk factors of season and/or thermal environment. The consistent finding that season is a risk factor for Salmonella shedding in cattle necessitates further studies in order to understand the potential mechanisms for mitigating this effect.

It is important to note that the environmental parameters, in addition to their potential physiological effects on animals, may also affect the concentration and survival of Salmonella in the farm environment. In the present study, only pooled fecal samples were collected, although Salmonella can also be isolated in other ecological niches, such as feed, water, bedding, manure pit and spreaders, air, wildlife, such as rodents, birds, and insects, and the milking parlor. Seasonal variation of Salmonella has been reported in environmental samples (5,7,9). Further research is warranted in order to understand the survival of Salmonella in the environment, as well as increased susceptibility in the host, secondary to thermal exposures. Although this study progressed over 3 main seasons, which correspond to high Salmonella shedding, it did not take into account other possible

risk factors associated with seasonal variation that might implicate seasonal variability, such as changes in feed, presence of wildlife, and contamination of feed, soil, and water (5,7,9).

If the thermal environment is a component cause of the effect of seasonality, cooling strategies may be beneficial for managing Salmonella. The literature is not consistent, however, with regard to the effect of the use of cooling strategies on Salmonella shedding in cattle. The use of sprinklers in beef feedlots (20) and in the feedbunk in lactating dairy cattle farms (12) was not associated with Salmonella prevalence. Salmonella prevalence decreased significantly, however, when sprinklers were used in holding pens before milking (12). Conversely, in a multi-state study, the use of sprinklers or misters was associated with Salmonella-positive herds (21). This variability may be explained by other herd management strategies related to seasonal variation. The effect of those strategies on decreasing heat stress and foodborne pathogens in dairy cattle should be further investigated. Abatement of thermal stress might be advantageous, not only in terms of production and animal health and welfare (22), but also for decreasing foodborne pathogens and improving public health protection. Further research is required to evaluate thermal environmental parameters and other risk factors that may be a com-ponent of the causal pathway associated with the summer season and increased Salmonella shedding in dairy cattle.

A c k n o w l e d g m e n t sThe authors thank the College of Veterinary Medicine Endowed

Research Funds (College of Veterinary Medicine, Michigan State University) for funding, Michael Garrod and Roderick Manuzon for technical assistance, Joseph Sullivan and Joel Sparks for sample collection, and Louis M. Neuder, William Raphael, and dairy farm staff for technical assistance.

Re f e r e n c e s 1. Cummings KJ, Warnick LD, Alexander KA, et al. The incidence

of salmonellosis among dairy herds in the northeastern United States. J Dairy Sci 2009;92:3766–3774.

2. Scallan E, Hoekstra RM, Angulo FJ, et al. Foodborne illness acquired in the United States — Major pathogens. Emerg Infect Dis 2011;17:7–15.

3. Loneragan GH, Thomson DU, McCarthy RM, et al. Salmonella diversity and burden in cows on and culled from dairy farms in the Texas High Plains. Foodborne Pathog Dis 2012;9:549–555.

4. Schneider JL, White PL, Weiss J, et al. Multistate outbreak of multidrug-resistant Salmonella Newport Infections associated with ground beef, October to December 2007. J Food Prot 2011; 74:1315–1319.

5. Van Kessel JS, Sonnier J, Zhao S, Karns JS. Antimicrobial resis-tance of Salmonella enterica isolates from bulk tank milk and milk filters in the United States. J Food Prot 2013;76:18–25.

6. Edrington TS, Hume ME, Looper ML, et al. Variation in the faecal shedding of Salmonella and E. coli O157:H7 in lactating dairy cattle and examination of Salmonella genotypes using pulsed-field gel electrophoresis. Lett Appl Microbiol 2004;38: 366–372.



7. Edrington TS, Ross TT, Callaway TR, et al. Investigation into the seasonal salmonellosis in lactating dairy cattle. Epidemiol Infect 2008;136:381–390.

8. Fossler CP, Wells SJ, Kaneene JB, et al. Herd-level factors asso-ciated with isolation of Salmonella in a multi-state study of conventional and organic dairy farms I. Salmonella shedding in cows. Prev Vet Med 2005;70:257–277.

9. Pangloli P, Dje Y, Ahmed O, et al. Seasonal incidence and molecu-lar characterization of Salmonella from dairy cows, calves, and farm environment. Foodborne Pathog Dis 2008;5:87–96.

10. Fitzgerald AC, Edrington TS, Looper ML, et al. Antimicrobial susceptibility and factors affecting the shedding of E. coli O157:H7 and Salmonella in dairy cattle. Lett Appl Microbiol 2003;37:392–398.

11. Edrington TS, Schultz CL, Genovese KJ, et al. Examination of heat stress and stage of lactation (early versus late) on fecal shed-ding of E. coli O157:H7 and Salmonella in dairy cattle. Foodborne Pathog Dis 2004;1:114–119.

12. Edrington TS, Carter BH, Friend TH, et al. Influence of sprin-klers, used to alleviate heat stress, on faecal shedding of E. coli O157:H7 and Salmonella and antimicrobial susceptibility of Salmonella and Enterococcus in lactating dairy cattle. Lett Appl Microbiol 2009;48:738–743.

13. Warnick LD, Kaneene JB, Ruegg PL, et al. Evaluation of herd sampling for Salmonella isolation on midwest and northeast US dairy farms. Prev Vet Med 2003;60:195–206.

14. Pires AF, Funk JA, Bolin CA. Longitudinal study of Salmonella shedding in naturally infected finishing pigs. Epidemiol Infect 2013;141:1928–1936.

15. Lucas EM, Randall JM, Meneses JF. Potential for evaporative cooling during heat stress periods in pig production in Portugal (Alentejo). J Agric Eng Res 2000;76:363–371.

16. Kadzere CT, Murphy MR, Silanikove N, Maltze E. Heat stress in lactating dairy cows: A review. Livest Prod Sci 2002;77:59–91.

17. St-Pierre NR, Cobanov B, Schnitkey G. Economic losses from heat stress by US livestock industries. J Dairy Sci 2003;86: E52–E77.

18. Dohoo I, Martin W, Stryhn H. Veterinary Epidemiologic Research, 2nd ed. Charlottetown, Prince Edward Island: AVC Inc. 2010.

19. SAS Data Analysis Examples Negative Binomial Regression. Institute for Digital Research and Education, UCLA. Available from: www.ats.ucla.edu . stat . sas . dae Last accessed April 10, 2016.

20. Morrow JL, Mitloehner FM, Johnson AK, et al. Effect of water sprinkling on incidence of zoonotic pathogens in feedlot cattle. J Anim Sci 2005;83:1959–1966.

21. Habing GG, Lombard JE, Kopral CA, Dargatz DA, Kaneene JB. Farm-level associations with the shedding of Salmonella and antimicrobial-resistant Salmonella in U.S. dairy cattle. Foodborne Pathog Dis 2012;9:815–821.

22. Collier RJ, Dahl GE, VanBaale MJ. Major advances associated with environmental effects on dairy cattle. J Dairy Sci 2006;89: 1244–1253.


Article


I n t r o d u c t i o nRotaviruses (RVs) are members of the family Reoviridae and cause

severe diarrhea in humans and animals (1). Rotaviruses are classified into 8 species (A to H) based on the genetic property of their inner capsid protein VP6 (2). The outer capsid proteins of rotaviruses (VP7 and VP4) are defined by the number of G and P genotypes, respectively, based on differences in gene sequence (1,3). Rotavirus A (RVA), rotavirus B (RVB), rotavirus C (RVC), and rotavirus H (RVH) are known to infect both humans and animals (1,2,4). While the

epidemiology and protective immunity of RVA infection have been intensively studied (1,5), relatively little is known about non-A rotavirus infections.

Rotavirus Bs (RVBs) have been detected in humans and vari-ous animals, including cattle (1,6–8). Unlike RVA, RVB in humans primarily causes diarrhea in adults, as well as in children (9–11). In cattle, RVB has been associated with epidemic diarrhea in adults (12,13) and a marked decrease in milk production was observed in dairy cows in 1 study (14). Our serological surveys indicated the common occurrence of RVB infection in cattle (15). Triggers for

Reinfection of adult cattle with rotavirus B during repeated outbreaks of epidemic diarrhea

Michiko Hayashi, Toshiaki Murakami, Yoshizumi Kuroda, Hikaru Takai, Hisahiro Ide, Ainani Awang, Tohru Suzuki, Ayako Miyazaki, Makoto Nagai, Hiroshi Tsunemitsu

A b s t r a c tRotavirus B (RVB) infection in cattle is poorly understood. The objective of this study was to describe the epidemiological features of repeated outbreaks of epidemic diarrhea due to RVB infection in adult cattle on a large dairy farm complex in Japan. In October 2002, approximately 550 adult cows and approximately 450 in February 2005 had acute watery diarrhea at several farms on the complex. Four months before the first outbreak, RVB antibody-positive rates at subsequently affected farms were significantly lower than at non-affected farms (30% to 32% versus 61% to 67%). During the acute phase of both outbreaks, RVB antibody-positive rates in diarrheal cows tested were as low as 15% to 26%. Most of the farms affected in the second outbreak were also involved in the first outbreak. Some adult cows with RVB diarrhea in the first outbreak showed not only RVB seroresponse, but also RVB shedding in the second outbreak, although none of these cows developed diarrhea. Nucleotide sequences of the VP7 and VP4 genes revealed a close relationship between RVB strains in both outbreaks. Taken together, these results indicate that outbreaks of epidemic RVB diarrhea in adult cows might be influenced by herd immunity and could occur repeatedly at the same farms over several years. To our knowledge, this is the first report on repeated RVB infections in the same cattle.

R é s u m éL’infection par le rotavirus B (RVB) chez les bovins est peu comprise. L’objectif de la présente étude était de décrire les caractéristiques épidémiologiques de poussées de cas répétées de diarrhée épidémique dues à une infection par le RVB chez des bovins adultes dans un grand complexe laitier au Japon. En octobre 2002, environ 550 vaches adultes et environ 450 en février 2005 présentaient une diarrhée aqueuse aigüe dans plusieurs fermes sur le complexe. Quatre mois avant le premier épisode, les taux d’anticorps anti-RVB dans les fermes subséquemment affectées étaient significativement plus faibles que dans les fermes non-affectées (30 % à 32 % vs 61 % à 67 %). Pendant la phase aigüe des deux épidémies, les taux d’anticorps anti-RVB chez les vaches diarrhéiques testées étaient aussi bas que 15 % à 26 %. La plupart des fermes affectées dans la deuxième épidémie étaient également impliquées dans la première épidémie. Quelques vaches adultes avec une diarrhée à RVB dans la première épidémie avaient non seulement une réponse sérologique positive envers le RVB, mais excrétaient également le RVB durant la deuxième épidémie, bien qu’aucun de ces vaches ne présenta de diarrhée. Les séquences nucléotidiques des gènes VP7 et VP4 ont révélé une proche parenté entre les deux souches de RVB des deux épidémies. Pris globalement, ces résultats indiquent que les épisodes de diarrhée épidémique causée par RVB chez des vaches adultes peuvent être influencés par l’immunité du troupeau et peuvent survenir de manière répétée sur une même ferme pendant plusieurs années. Selon nous, il s’agirait du premier rapport de cas d’infections à répétition par le RVB chez les mêmes bovins.

(Traduit par Docteur Serge Messier)

Ishikawa Nanbu Livestock Hygiene Service Center, Kanazawa, Ishikawa 9203101, Japan (Hayashi, Takai, Ide); Ishikawa Hokubu Livestock Hygiene Service Center, Nanao, Ishikawa 9292126, Japan (Murakami, Kuroda); Viral Disease and Epidemiology Research Division, National Institute of Animal Health, Tsukuba, Ibaraki 3050856, Japan (Awang, Suzuki, Miyazaki); Research and Education Center for Prevention of Global Infectious Diseases of Animals, Tokyo University of Agriculture and Technology, Fuchu, Tokyo 1838509, Japan (Nagai); Dairy Hygiene Research Division, Hokkaido Research Station, National Institute of Animal Health, Sapporo 0620045, Japan (Tsunemitsu); Veterinary Research Institute, Ipoh, Perak 31400, Malaysia (Awang); United Graduate School of Veterinary Sciences, Gifu University, Gifu 5011193, Japan (Nagai, Tsunemitsu).

Address all correspondence to Dr. Hiroshi Tsunemitsu; telephone: 181-11-851-2123; fax: 181-11-853-0767; e-mail: [email protected]

Received October 5, 2015. Accepted March 21, 2016.



epidemic outbreaks of RVB diarrhea remain unclear, however, as do circumstances that lead to subsequent RVB infection in adult cattle.

In the present study, we report on the epidemiological features of repeated outbreaks of epidemic diarrhea in adult cattle caused by RVB infection on a large dairy farm complex in Japan. Furthermore, we describe evidence of repeated RVB infection in the same cattle.


Overview of farmsThe large dairy farm complex was composed of 7 zones, with

each zone consisting of 4 farm compartments (Figure 1). Each zone was 200 m apart and the compartments within the zones were 15 m apart. In 2002, 17 farms (Farms A to Q) had approximately 1400 adult cows (23 to 275/farm) and 200 growing cattle (0 to 60/farm). In 2005, 16 farms (Farm C had closed) had approximately 1500 adult cows (48 to 285/farm) and 320 growing cattle (7 to 46/farm). The same veterinarian provided medical care and artificial insemination at all farms, with the exception of farms A and F. Average renewal rates of adult milking cows per year were approximately 30% to 40%. Approximately 2/3 of replacement heifers were purchased from other dairies.

Clinical samplesDuring the first outbreak in 2002, fecal samples were collected

from 20 diarrheal adult cows (aged over 24 mo) at the following

affected farms: Farm E (n = 10), Farm D (n = 5), Farm J (n = 3), and Farm L (n = 2). During the second outbreak in 2005, fecal samples were collected from 28 diarrheal and 18 normal adult cows at the following affected farms: Farm E (n = 19), Farm D (n = 12), Farm L (n = 5), Farm Q (n = 5), and Farm G (n = 5). At least 2 or more diar-rheal samples were collected at these farms. These feces were sampled within 4 d after the first finding of diarrhea at each farm. Serum samples were collected from these cows at the same time as fecal sampling at the acute phase and again 19 to 28 d later at the convalescent phase. During both outbreaks, overlapping sampling was conducted in 7 of these cows. Serum samples were also collected from adult cows 4 mo before the first outbreak in 2002 during a general health examination at the following farms: Farm E (n = 66), Farm D (n = 69), Farm A (n = 118), and Farm O (n = 71). Fecal samples were subjected to reverse transcription polymerase chain reaction (RT-PCR) and polyacrylamide gel electrophoresis (PAGE) of rotavirus double-stranded ribonucleic acid (dsRNA). Fecal samples were also tested for Salmonella species using a standard technique and Coccidium and Cryptosporidium species were checked by a sucrose flotation method.

RT-PCRViral RNA was extracted from fecal suspensions using TRIzol LS

Reagent (Invitrogen, Carlsbad, California, USA) in accordance with the manufacturer’s instructions. Reverse transcription polymerase chain reaction (RT-PCR) assays were conducted using the OneStep RT-PCR Kit (QIAGEN, Valencia, California, USA) to detect the

Figure 1. Plan of large dairy farm complex in 2002 and in 2005, with capital letters indicating farm names. Farms with an outbreak of epidemic rota-virus B (RVB) diarrhea are shaded.



following genes: RVB (VP7 gene) (16), RVA (VP7 gene) (17), RVC (VP7 gene) (18), bovine torovirus (BToV) (N gene) (19), bovine coro-navirus (BCV) (N gene) (20), and bovine viral diarrhea virus (BVDV) (59 non-structural coding region) (21).

PAGE with viral dsRNAPolyacrylamide gel electrophoresis (PAGE) of dsRNA extracted

from diarrheal fecal samples was carried out with 7.5% precast gels (e-PAGEL; Atto, Tokyo, Japan) (22). The gels were stained with the Silver Stain Plus Kit (Bio-Rad, Hercules, California, USA) in accor-dance with the manufacturer’s instructions.

Virus antibody testsBovine RVB antibody was detected by enzyme-linked immuno-

sorbent assay (ELISA) with recombinant baculovirus-expressed VP6 protein, as previously described (15). The RVB antibody titers were expressed as net optical density (OD), which was calculated as the OD from antigen-coated wells minus the OD from non-antigen-coated wells. A serum was considered RVB antibody-positive if net OD was greater than 0.2. Seroresponse was defined as an increase in net OD of 0.2 or more in paired sera. As previously described, virus neutralization tests were conducted for BVDV type 1 and bovine adenovirus type 3 (BAdV-3), using paired sera from affected and normal cows (23). Antibody titers against BCV and adenovirus type 7 (BAdV-7) were also determined by hemagglutination inhibition tests (24). Seroresponse was defined as a 4-fold or greater increase in paired serum antibody titers to the examined virus. Laboratory

personnel were blinded to the disease status of the cattle when con-ducting the virus antibody tests.

Sequence analysis of the VP7 and VP4 genes of RVB

The full length VP7 gene of RVB from at least 2 fecal samples at each of 4 affected farms in 2002 and 5 affected farms in 2005 was produced by RT-PCR using the OneStep RT-PCR Kit (QIAGEN) with 59 and 39 end primers (13) and the PCR products were sequenced directly by cycle sequencing with an auto sequencer (ABI PRISM 3100; Life Technologies, Foster City, California, USA). The 59- terminal region of the VP4 gene of RVB from 1 fecal sample per farm at 4 affected farms in 2002 and 3 affected farms in 2005 was amplified by RT-PCR with a pair of primers: 1F (59-GGTATTTAATCACTAGGC-39) and 1295R (59-GGATTCAAACTGTTGTCAACTGG-39). Reverse transcription polymerase chain reaction (RT-PCR) was carried out using the PrimeScript One Step RT-PCR Kit Version 2 (Takara, Shiga, Japan). The products were cloned into pCR2.1 TOPO vector (Life Technologies) and sequenced by cycle sequencing. Sequence data were aligned using the Clustal W method and phylogenetic trees were generated using the MegAlign program (Version 11.2.1) of Lasergene software (DNASTAR, Madison, Wisconsin, USA).

Statistical analysisSeroprevalence of RVB among farms or cows was statistically

analyzed by the Chi-square test or Fisher’s exact test using Ekuseru-Toukei 2012 (Social Survey Research Information, Tokyo, Japan).

Nucleotide sequence accession numbersThe newly determined sequences have been deposited in the

DDBJ nucleotide (nt) sequence database and assigned the follow-ing accession numbers: D-2002 (VP7 gene, LC005523 and VP4 gene, LC005532); E-2002 (VP7 gene, LC005524 and VP4 gene, LC005533); J-2002 (VP7 gene, LC005525 and VP4 gene, LC005534); L-2002 (VP7 gene, LC005526 and VP4 gene, LC005535); D-2005 (VP7 gene, LC005527); E-2005 (VP7 gene, LC005528); G-2005 (VP7 gene, LC005529 and VP4 gene, LC005536); K-2005 (VP7 gene, LC005530 and VP4 gene, LC005537); and L-2005 (VP7 gene, LC005531 and VP4 gene, LC005538).

Re s u l t s

Outbreaks of epidemic diarrhea in adult cowsIn October 2002, epidemic diarrhea was first observed in some

lactating adult cows at Farm G and then spread within 2 wk to approximately 550 out of 660 adult cows (83%) at 9 (Farms D, E, F, G, J, K, L, P, and Q) of 17 farms in a large dairy farm complex (Figure 1). Diarrhea spread through each farm within several days. Notably, 4 replacement heifers had been introduced at Farm G 2 d before the outbreak. Unfortunately, fecal samples had not been col-lected from these heifers because they did not show diarrhea at the time of outbreak.

In March 2005, another outbreak of epidemic diarrhea occurred in approximately 450 out of 700 adult cows (64%) at 9 (D, E, G, J, K, L, M, P, and Q) of 16 farms in the same complex (Figure 1). Diarrhea

Figure 2. Electrophoretic migration patterns of viral ribonucleic acids (RNAs) from feces positive for rotavirus B (RVB) by reverse transcription polymerase chain reaction (RT-PCR) in 2002 and 2005. Lanes indicate feces from each farm as follows: E (Farm E), D (Farm D), L (Farm L), K (Farm K), G (Farm G), and Q (Farm Q). E and D were collected in 2002 and L, K, G, and Q in 2005. The rotavirus A (RVA) lane indicates refer-ence RVA strain OSU. Numbers and arrows indicate genome segments of rotaviruses. Characteristic migration patterns of double-stranded RNA (dsRNA) for RVB (pattern 4-2-2-3) were observed in lanes E, D, K, and G.



spread within several days at each individual farm and had affected these farms within 2 wk. No heifers had been newly introduced to any farms just before the outbreak. Eight of these 9 farms had also been hit by the previous outbreak in 2002.

Several common epidemiological features were observed between the outbreaks in 2002 and 2005. For example, affected cows had watery and brownish diarrhea, but not bloody diarrhea or fever, and each cow recovered 3 to 5 d after the onset of diarrhea without clinical treatment. Milk production decreased by an average of approximately 10%. During these outbreaks of adult cow diarrhea, no clinical signs, including diarrhea, were observed in the majority of growing cattle and calves. There were no significant differences between affected and non-affected farms in the mean number of adult cows in both outbreaks and in the mean annual renewal rates of milking cows from 2002 to 2005.

Fecal examinationSixteen of 20 diarrheal fecal samples at 4 farms affected by the

2002 outbreak and 22 of the 28 diarrheal fecal samples at 5 farms affected by the 2005 outbreak were positive for RVB as determined by RT-PCR. In addition, 8 of 18 normal fecal samples collected at the same time as diarrheal samples at these affected farms in 2005 were also positive for RVB by RT-PCR. All diarrheal fecal samples were negative by RT-PCR for RVA, RVC, BCV, BoTV, and BVDV, however, and were also negative for Salmonella, Coccidium, and Cryptosporidium species.

Figure 3. Detection of rotavirus B (RVB) antibody by enzyme-linked immunosorbent assay (ELISA) with paired sera diluted 1:100 from adult cows with or without diarrhea in acute and convalescent phases during outbreaks of epidemic diarrhea in 2002 (a) and 2005 (b, c). Values above the dotted line [optical density (OD) = 0.2] indicate positive results for RVB antibody. Seroresponse was defined as an increase in OD of $ 0.2 in paired sera.

Table I. Prevalence of rotavirus B (RVB) antibody detected by enzyme-linked immunosorbent assay (ELISA) in sera of adult cows at farms 4 mo before an outbreak of epidemic RVB diarrhea in 2002

Detection of RVB antibody Number SampleFarm positive number (%)Da 21 66 31.8A

Ea 21 69 30.4A

Ab 79 118 66.9B

Ob 43 71 60.6B

a Outbreak of epidemic RVB diarrhea occurred 4 mo later.b Outbreak of epidemic RVB diarrhea did not occur 4 mo later.A,B Values with different superscripts are significantly different (P , 0.01).



PAGE with viral dsRNAOf 14 fecal samples analyzed, 5 showed the characteristic migra-

tion patterns of dsRNA for RVB (pattern 4-2-2-3), based solely on PAGE analysis (Figure 2), and all were positive for RVB by RT-PCR. Notably, the migration pattern in 2002 was similar but not identical to that in 2005, with the spaces between RNA segments 5 and 6 and 7 and 8 in 2005 slightly wider than those in 2002 (Figure 2).

Virus antibody testsExamination of paired sera from 20 affected cows from the out-

break in 2002 and 28 from 2005 showed no significant changes in antibody titers against BCV, BVDV type 1, BAdV-3, or BAdV-7. In contrast, a significant seroresponse ($ 0.2 OD increase) to RVB was detected using ELISA in 16 of those 20 affected cows (80%) in 2002 and in 24 of the 28 cows (86%) in 2005 (Figures 3a and 3b). Taking into account these findings and the results of fecal examina-tion, bovine RVB infection was considered to be the cause of both outbreaks of epidemic diarrhea. Interestingly, paired sera from 13 of 18 normal cows (72%) at the affected farms also exhibited a significant seroresponse to RVB (Figure 3c).

Inter- and intra-farm prevalence of RVB antibodies

Four months before the outbreak in 2002, RVB antibody-positive rates determined by ELISA in 135 adult cows at later affected Farms D

and E were 32% and 30%, respectively. These rates were signifi-cantly lower than those determined by ELISA in 189 cows at later non-affected Farms A and O (67% and 61%, respectively; P , 0.01) (Table I). During the acute phase of the 2002 outbreak, the antibody-positive rate in 20 diarrheal cows at Farms D, E, J, and L was as low as 15% (Figure 3a). During the acute phase of the 2005 outbreak, the antibody-positive rate in 28 diarrheal cows (Farms D, E, G, L, and Q) was 26%, which was significantly lower than that in 18 non-diarrheal cows at 67% (Farms D and E) (P , 0.01) (Figures 3b and 3c).

Responses to RVB in the same cattleFour of 7 cows (57%) affected by RVB diarrhea in the 2002 out-

break exhibited RVB shedding in feces again in the 2005 outbreak, as determined by RT-PCR (Figure 4). Furthermore, 5 of these cows (71%) exhibited significant antibody-response with paired sera in the 2005 outbreak as determined by ELISA. It is notable, however, that none of these cows had diarrhea in the 2005 outbreak (Figure 4).

Sequence analysis of VP7 and VP4 genes of RVBThe nt sequences of VP7 genes of RVB from the same farms were

identical at each outbreak. Therefore, VP7 gene sequences from 9 RVB strains (1 strain per farm in each outbreak) were indicated and designated as D-2002, E-2002, J-2002, L-2002, D-2005, E-2005, G-2005, K-2005, and L-2005. The VP7 gene sequences of these strains were 816 nts in length and most closely related to that of the bovine RVB Nemuro strain detected in Japan in 1997. When compared with

Figure 4. Rotavirus B (RVB) antibody, incidence of diarrhea, and shedding of RVB in the same cows (n = 7) in 2002 and 2005. RVB antibody was detected by enzyme-linked immunosorbent assay (ELISA) with paired sera diluted 1:100 in acute and convalescent phases in outbreaks of epidemic diarrhea. Values above the dotted line [optical density (OD) = 0.2] indicate positive results for RVB antibody. Seroresponse was defined as an increase in OD of $ 0.2 in paired sera in each outbreak. Shedding of RVB was detected by reverse transcription polymerase chain reaction (RT-PCR) (1).



the present strains, all VP7 genes examined were identical in the 2002 outbreak and similar to those in the 2005 outbreak. VP7 genes differed by 5 to 6 nts in the 2002 and 2005 strains, however, which reflected their distinct positions in the phylogenetic tree (Figure 5a).

The 59-terminal region of VP4 gene sequences from 7 RVB strains was determined and designated as D-2002, E-2002, J-2002, L-2002, G-2005, K-2005, and L-2005 and 1295 nt sequences of these strains were compared to those of published bovine and human RVB strains. Although the VP4 genes in viruses in the present study were found to be more closely related to those of bovine RVB strains detected in India (RUBV226, RUBV282, and DB176) than to human RVB strains, the nt sequence identities of the present and Indian bovine strains were relatively low (79% to 80%). When VP4 genes among the strains in the present study were compared, the strains were similar to each other, but as with the VP7 genes, the strains of 2002 were distinct from those of 2005 in the phylogenetic tree (Figure 5b).

D i s c u s s i o nViral diarrhea in adult cattle is caused by several viruses, includ-

ing RVB (12–14). During the routine diagnosis of diarrhea in adult cattle, however, RVB infection is not typically considered due to the shortage of information about its clinical importance. Therefore, the epidemiology of RVB diarrhea in adult cattle still remains unclear. In this study, we observed that outbreaks of epidemic diarrhea with decreases in milk production caused by RVB infection in adult cows could occur repeatedly at the same farms just like epidemic diarrhea in adult cows caused by BCV infection (25). In addition, the epidemic diarrheas were observed mostly in adult cows, but not in growing cattle and calves, which is consistent with other reports (13,14).

In the present study, RVB antibody-positive rates at later affected farms were significantly lower than at later non-affected farms. In addition, the antibody-positive rates in affected cows were relatively

Figure 5. Phylogenetic trees for the VP7 gene (a) and the 59-terminal region of VP4 gene (b) of bovine and human rotavirus B (RVB). The trees were generated using the MegAlign program of the Lasergene software (DNASTAR) on the basis of 772–816 bp of VP7 and 1265–1295 bp of VP4. The length of each pair of branches represents the distance between sequence pairs and the units at the bottom of the tree indicate the number of substitution events. Bootstrap values greater than 700 in 1000 pseudoreplicates are shown as percentages. The accession numbers of the nt sequences used for the tree construction are as follows: Nemuro, AB016818 (VP7 gene); MN10-1, JQ288103 (VP7 gene); ATI, U84472 (VP7 gene); WD653, U84141 (VP7 gene); Mebus, U84473 (VP7 gene); DB101, AY158155 (VP7 gene); DB180, AF529214 (VP7 gene); DB176, AF531910 (VP7 gene) and GQ358710 (VP4 gene); CAL, AF184083 (VP7 gene) and AF184084 (VP4 gene); Bang373, NC_021542 (VP7 gene) and NC_021543 (VP4 gene); WH-1, AY539856 (VP7 gene) and AY539857 (VP4 gene), ADRV, M33872 (VP7 gene) and M91434 (VP4 gene); IC-008, GU377216 (VP4 gene); and NIV-094456, JN009774 (VP4 gene). Bovine RVB strains from the present study are indicated by boldface type and their accession numbers are given in the text.

RVB/Cattle-wt/JPN/G-2005/2005/G3PX RVB/Cattle-wt/JPN/K-2005/2005/G3PX RVB/Cattle-wt/JPN/D-2005/2005/G3PX RVB/Cattle-wt/JPN/L-2005/2005/G3PX RVB/Cattle-wt/JPN/E-2005/2005/G3PXRVB/Cattle-wt/JPN/D-2002/2002/G3PXRVB/Cattle-wt/JPN/L-2002/2002/G3PXRVB/Cattle-wt/JPN/E-2002/2002/G3PXRVB/Cattle-wt/JPN/J-2002/2002/G3PX

RVB/Cattle-wt/JPN/Nemuro/1997/G3PX RVB/Cattle-wt/USA/MN10-1/2010/G3PX RVB/Cattle-wt/USA/ATI/19XX/G3PX RVB/Cattle-wt/USA/WD653/19XX/G3PX RVB/Cattle-wt/USA/Mebus/19XX/G3PX RVB/Cattle-wt/IND/DB101/2001/G5PX RVB/Cattle-wt/IND/DB180/2001/G5PX RVB/Cattle-wt/IND/DB176/2001/G5PX RVB/Human-wt/IND/CAL/1998/G2PX RVB/Human-wt/BGD/Bang373/2000/G5PXRVB/Human-wt/CHN/WH-1/2002/G5PX RVB/Human-wt/CHN/ADRV/1986/G5PX

RVB/Cattle-wt/IND/RUBV226/2004/G5PX RVB/Cattle-wt/IND/RUBV282/2005/G5PX RVB/Cattle-wt/IND/DB176/2001/G5PX RVB/Human-wt/IND/IC-008/2008/G2PX RVB/Human-wt/IND/NIV-094456/2009/G2PX RVB/Human-wt/IND/CAL/1998/G2PXRVB/Human-wt/BGD/Bang373/2000/G2PXRVB/Human-wt/CHN/ADRV/1986/G2PXRVB/Human-wt/CHN/WH-1/2002/G2PX

RVB/Cattle-wt/JPN/D-2002/2002/G3PX RVB/Cattle-wt/JPN/L-2002/2002/G3PX RVB/Cattle-wt/JPN/E-2002/2002/G3PX RVB/Cattle-wt/JPN/J-2002/2002/G3PXRVB/Cattle-wt/JPN/K-2005/2005/G3PXRVB/Cattle-wt/JPN/L-2005/2005/G3PXRVB/Cattle-wt/JPN/G-2005/2005/G3PX



low. In human RVA infections, the presence of serum antibody was a marker for protection against RVA infection and disease (26–28). Taken together, the outbreaks of epidemic RVB diarrhea in adult cows in this study might be associated with the antibody prevalence at each farm, i.e., herd immunity (29,30). Although the reasons for the observed inter-farm differences in prevalence of RVB antibody are unknown, RVB subclinical infection or sporadic diarrhea might have occurred at some farms with high seroprevalence. Notably, uni-versal mass vaccination of children for human RVA has resulted in herd immunity, which reduces prevalence of RVA-associated disease even in older children who were not vaccinated (31).

The second outbreak in 2005 occurred at most of the farms that had been affected by the first outbreak in 2002. Although the reasons for this are not known, we consider that it does not contradict the theory of herd immunity for the following reason. On each farm of the complex, approximately 30% to 40% of milking cows were replaced annually and 2/3 of replacement heifers were purchased from outside. This means that, by 2004, almost half of milking cows might not have experienced the 2002 outbreak at all farms. Given the results of RVB VP7 and VP4 gene sequences, we consider it more likely that the strains spread among farms during each outbreak than that each strain persisted at each farm from the first outbreak and caused the second outbreak.

In humans, primary RVA infection does not usually lead to perma-nent immunity and reinfection can occur at any age (1,32), although subsequent infection is generally less severe (33,34). In addition, a cohort study of human RVA infection indicated that protection against homotypic reinfection appeared to last 2 y (35). In contrast, little information is available about immunity to RVB infection. The present study demonstrated that some adult cows became reinfected with RVB strains, in which VP4 and VP7 genes were closely related, at an interval of approximately 2 1/2 y. These cows did not exhibit diarrhea during the second outbreak, however, despite observations of RVB shedding. These results indicate that adult cows can be rein-fected with RVB, but that subsequent infection might be less severe, which appears to be very similar to RVA infection. This information might help to control RVB diarrhea in humans and animals.

In conclusion, bovine RVB can repeatedly cause epidemic diarrhea with a decrease in milk production in adult cows at the same farms and the same adult cows can develop RVB infection repeatedly. A simple test to detect bovine RVB is being developed for further epidemiological investigation.

A c k n o w l e d g m e n tThe authors thank Ms. Nachiko Hattori for her technical assistance.

Re f e r e n c e s1. Estes MK, Kapikian AZ. Rotaviruses and their replication. In:

Fields BN, Knipe DM, Howley PM, eds. Fields Virology. 5th ed. Philadelphia, Pennsylvania: Lippincott, Williams & Wilkins, 2007:1917–1974.

2. Matthijnssens J, Otto PH, Ciarlet M, Desselberger U, Van Ranst M, Johne R. VP6-sequence-based cutoff values as a criterion for rotavirus species demarcation. Arch Virol 2012;157:1177–1182.

3. Matthijnssens J, Ciarlet M, McDonald SM, et al. Uniformity of rotavirus strain nomenclature proposed by the Rotavirus Classification Working Group (RCWG). Arch Virol 2011;156: 1397–1413.

4. Wakuda M, Ide T, Sasaki J, Komoto S, Ishii J, Sanekata T. Porcine rotavirus closely related to novel group of human rotaviruses. Emerg Infect Dis 2011;17:1491–1493.

5. Angel J, Franco MA, Greenberg HB. Rotavirus immune responses and correlates of protection. Curr Opin Virol 2012;2: 419–425.

6. Bridger JC. Non-group A rotaviruses. In: Kapikian AZ, ed. Viral Infections of the Gastrointestinal Tract. 2nd ed. New York, New York: Marcel Dekker, 1994:369–407.

7. Ghosh S, Varghese V, Sinha M, Kobayashi N, Naik TN. Evidence for interstate transmission and increase in prevalence of bovine group B rotavirus strains with a novel VP7 genotype among diarrhoeic calves in Eastern and Northern states of India. Epidemiol Infect 2007;135:1324–1330.

8. Saif LJ. Nongroup A rotaviruses. In: Saif LJ, Theil KW, eds. Viral Diarrheas of Man and Animals. Boca Raton, Florida: CRC Press, 1990:73–95.

9. Fang ZY, Ye Q, Ho MS, et al. Investigation of an outbreak of adult diarrhea rotavirus in China. J Infect Dis 1989;160:948–953.

10. Hung T, Chen GM, Wang CG, et al. Waterborne outbreak of rota-virus diarrhoea in adults in China caused by a novel rotavirus. Lancet 1984;1:1139–1142.

11. Lahon A, Maniya NH, Tambe GU, et al. Group B rotavirus infec-tion in patients with acute gastroenteritis from India: 1994–1995 and 2004–2010. Epidemiol Infect 2013;141:969–975.

12. Chang KO, Parawani AV, Smith D, Saif LJ. Detection of group B rotaviruses in fecal samples from diarrheic calves and adult cows and characterization of their VP7 genes. J Clin Microbiol 1997;35:2107–2110.

13. Tsunemitsu H, Morita D, Takaku H, Nishimori T, Imai K, Saif LJ. First detection of bovine group B rotavirus in Japan and sequence of its VP7 gene. Arch Virol 1999;144:805–815.

14. Hayashi M, Nagai M, Hayakawa Y, Takeuchi K, Tsunemitsu H. Outbreak of diarrhoea and milk drop in cows infected with bovine group B rotavirus. Vet Rec 2001;15:331–332.

15. Tsunemitsu H, Kamiyama M, Kawashima K, et al. Molecular characterization of the major capsid protein VP6 of bovine group B rotavirus and its use in seroepidemiology. J Gen Virol 2005;86:2569–2575.

16. Chinsangaram J, Akita GY, Osburn BI. Detection of bovine group B rotaviruses in feces by polymerase chain reaction. J Vet Diagn Invest 1994;6:302–307.

17. Gouvea V, Glass RI, Woods P, et al. Polymerase chain reaction amplification and typing of rotavirus nucleic acid from stool specimens. J Clin Microbiol 1990;28:276–282.

18. Tssunemitsu H, Jiang B, Saif LJ. Sequence comparison of the VP7 gene encoding the outer capsid glycoprotein among animal and human group C rotaviruses. Arch Virol 1996;141:705–713.

19. Smits SL, Lavazza A, Matiz K, Horzinek MC, Koopmans MP, de Groot RJ. Phylogenetic and evolutionary relationships among torovirus field variants: Evidence for multiple intertypic recom-bination events. J Virol 200;77:9567–9577.



20. Tsunemitsu H, Smith DR, Saif LJ. Experimental inoculation of adult dairy cows with bovine coronavirus and detection of coronavirus in feces by RT-PCR. Arch Virol 1999;144:167–175.

21. Vilcek S, Herring AJ, Herring JA, Nettleton PF, Lowings JP, Paton DJ. Pestiviruses isolated from pigs, cattle and sheep can be allocated into at least three genogroups using polymerase chain reaction and restriction endonuclease analysis. Arch Virol 1994;136:309–323.

22. Herring AJ, Inglis NF, Ojeh CK, Snodgrass DR, Menzies JD. Rapid diagnosis of rotavirus infection by direct detection of viral nucleic acid in silver-stained polyacrylamide gels. J Clin Microbiol 1982;16:473–477.

23. Xue W, Ellis J, Mattick D, Smith L, Brady R, Trigo E. Immuno-genicity of a modified-live virus vaccine against bovine viral diarrhea virus types 1 and 2, infectious bovine rhinotracheitis virus, bovine parainfluenza-3 virus, and bovine respiratory syncytial virus when administered intranasally in young calves. Vaccine 2010;28:3784–3792.

24. Sato K, Inaba Y, Kurogi E, et al. Hemagglutination by calf diar-rhea coronavirus. Vet Microbiol 1977;2:83–87.

25. Clark MA. Bovine coronavirus. Br Vet J 1993;149:51–70.26. Jiang B, Gentsch JR, Glass RI. The role of serum antibodies in

the protection against rotavirus disease: An overview. Clin Infect Dis 2002;34:1351–1361.

27. Premkumar P, Lopman B, Ramani S, et al. Association of serum antibodies with protection against rotavirus infection and dis-ease in South Indian children. Vaccine 2014;32:A55–61.

28. Velázquez FR, Matson DO, Guerrero ML, et al. Serum antibody as a marker of protection against natural rotavirus infection and disease. J Infect Dis 2000;182:1602–1609.

29. Fine PE. Herd immunity: History, theory, practice. Epidemiol Rev 1993;15:265–302.

30. Tråvén M, Sundberg J, Larsson B, Niskanen R. Winter dysen-tery diagnosed by farmers in dairy herds in central Sweden: Incidence, clinical signs and protective immunity. Vet Rec 1993; 133:315–318.

31. Lopman BA, Curns AT, Yen C, Parashar UD. Infant rotavirus vaccination may provide indirect protection to older children and adults in the United States. J Infect Dis 2011;204:980–986.

32. Mata L, Simhon A, Urrutia JJ, Kronmal RA, Fernández R, García B. Epidemiology of rotaviruses in a cohort of 45 Guatamalan Mayan Indian children observed from birth to the age of three years. J Infect Dis 1983;148:452–461.

33. Fischer TK, Valentiner-Branth P, Steinsland H, et al. Protective immunity after natural rotavirus infection: A community cohort study of newborn children in Guinea-Bissau, West Africa. J Infect Dis 2002;186:593–597.

34. Velázquez FR, Matson DO, Calva JJ, et al. Rotavirus infections in infants as protection against subsequent infections. N Engl J Med 1996;335:1022–1028.

35. Bernstein DI, Sander DS, Smith VE, Schiff GM, Ward RL. Protection from rotavirus reinfection: 2-year prospective study. J Infect Dis 1991;164:277–283.


Article


I n t r o d u c t i o nThe breeding of small ruminants focuses primarily on producing

animals for the meat or milk industry. Good selection of animals, as well as the most appropriate management practices, minimize potential problems that may arise in daily practice for farm animals as well as for operators (1). Reducing the level of stress for animals during handling increases productivity (2).

Many detrimental effects of handling stressors on the perfor-mance and health of animals are likely due to fear (3). It is therefore essential to identify and understand the many factors that cause fear and the resultant stress in animals in order to ensure their welfare. Traditionally, ethological observation was the method most often used to determine whether or not an animal was stressed. These

observations were supplemented by analyzing specific markers in blood samples, such as cortisol, that rise under stress (3). Today, however, noninvasive methods such as saliva sampling are used to assess stress because blood sampling is stressful in itself (4).

Several stress studies have measured serum cortisol in sheep during various handling processes (3,5,6) in order to develop appropriate handling systems and ensure high levels of animal wel-fare (7). Evaluating cortisol assessment is useful only as a marker of activation of hypothalamic-pituitary-adrenocortical (HPA) axis (8), however, and other biomarkers are necessary to provide informa-tion about the activation of the autonomic nervous system (ANS). Salivary alpha-amylase is a possible candidate as this enzyme is considered a biomarker of ANS activity in humans (9), as well as in animals such as pigs (10,12).

Validation of an assay for quantification of alpha-amylase in saliva of sheep

Maria Fuentes-Rubio, Francisco Fuentes, Julio Otal, Alberto Quiles, María Luisa Hevia

A b s t r a c tThe objective of this study was to develop a time-resolved immunofluorometric assay (TR-IFMA) for quantification of salivary alpha-amylase in sheep. For that purpose, after the design of the assay, an analytical and a clinical validation were carried out. The analytical validation of the assay showed intra- and inter-assay coefficients of variation (CVs) of 6.1% and 10.57%, respectively and an analytical limit of detection of 0.09 ng/mL. The assay also demonstrated a high level of accuracy, as determined by linearity under dilution. For clinical validation, a model of acute stress testing was conducted to determine whether expected significant changes in alpha-amylase were picked up in the newly developed assay. In that model, 11 sheep were immobilized and confronted with a sheepdog to induce stress. Saliva samples were obtained before stress induction and 15, 30, and 60 min afterwards. Salivary cortisol was measured as a reference of stress level. The results of TR-IFMA showed a significant increase (P , 0.01) in the concentration of alpha-amylase in saliva after stress induction. The assay developed in this study could be used to measure salivary alpha-amylase in the saliva of sheep and this enzyme could be a possible noninvasive biomarker of stress in sheep.

R é s u m éL’objectif de la présente étude était de développer un test immunofluorométrique en temps résolu (TIMF-TR) pour la quantification de l’alpha-amylase salivaire chez le mouton. À cette fin, suite au design du test, une validation analytique et clinique fut effectuée. La validation analytique du test a montré des coefficients de variation (CV) intra- et inter-tests de 6,1 % et 10,57 %, respectivement, et une limite de détection analytique de 0,09 ng/mL. Le test a également montré un haut niveau de précision, tel que déterminé par la linéarité suite aux dilutions. Pour la validation clinique, un modèle de test de stress aigu a été mené afin de déterminer si des changements significatifs attendus de l’alpha-amylase étaient détectés dans le nouveau test développé. Dans ce modèle, 11 moutons étaient immobilisés et confrontés avec un chien de berger afin d’induire le stress. Des échantillons de salive ont été obtenus avant l’induction du stress et 15, 30, et 60 min par la suite. Le cortisol salivaire a été mesuré à titre d’indicateur de référence du stress. Les résultats du TIMF-TR ont montré une augmentation significative (P , 0,01) de la concentration d’alpha-amylase dans la salive après l’induction du stress. Le test développé au cours de cette étude pourrait être utilisé afin de mesurer l’alpha-amylase salivaire dans la salive de mouton et cet enzyme pourrait être un biomarqueur non-invasif du stress chez le mouton.


Animal Science Department, School of Veterinary Medicine, University of Murcia, Murcia, Spain.

Address all correspondence to Dr. María Luisa Hevia; telephone: 34-868-88-4746; fax: 34-868-88-4147; e-mail: [email protected]

Received April 21, 2015. Accepted September 14, 2015.



The literature that the authors found on the role of salivary alpha-amylase in sheep was not very encouraging, with several authors stating that salivary alpha-amylase does not exist in sheep (13–15). However, given the research career of authors in salivary alpha-amylase (9–11), it was decided to continue the trial. In this study, an immunofluorometric assay was developed and validated for measuring salivary alpha-amylase in sheep and to evaluate its pos-sible role as a biomarker of stress using a stress-induction protocol.


AnimalsAnimals used in the study were from the Veterinary Experimental

Farm Unit of the University of Murcia in southeast Spain. Sheep were housed in a flock and belonged to the Montesina breed, which is in danger of extinction in Spain and is therefore officially protected. Eleven healthy sheep, 3 y of age, were randomly selected from the flock. Sheep were given access to straw ad libitum and rationed access to a nutritionally balanced commercial diet. Water was continuously available.

Analysis of salivary cortisolCortisol was analyzed with an immunoassay system (IMMULITE

1000; Siemens Health Diagnostics, Deerfield, Illinois, USA) that was validated for measuring cortisol in the saliva of sheep at the laboratory of the Animal Medicine and Surgery Department (data not shown).

Time-resolved immunofluorometric assay to determine salivary alpha-amylase

The assay was designed as a non-competitive, indirect, sandwich method based on the anti-alpha-amylase polyclonal antibody, biotin-labelled, as capture reagent and the anti-alpha-amylase polyclonal antibody, Eu31-chelates-labelled, as detector. The concentration of alpha-amylase in the samples was determined through a standard curve. The concentration range of the standard curve was obtained by serial dilutions of a known initial amount of alpha-amylase in an assay buffer (DELFIA Assay Buffer; PerkinElmer, Turku, Finland).

The procedure was as follows: wells of Streptavidin-coated plates (DELFIA; PerkinElmer) were coated for 1 h at room temperature with 300 ng of anti-alpha-amylase polyclonal antibody, biotin-labelled, in 200 mL of assay buffer (DELFIA; PerkinElmer) per well with gentle shaking. After coating, the wells were washed 4 times with wash solution using a DELFIA 1296-026 automatic plate washer (PerkinElmer). In the second step, wells were incubated with saliva samples diluted (1:4) in 200 mL of assay buffer (DELFIA) per well with gentle shaking. After coating, the wells were washed 4 times with wash solution using a DELFIA 1296-026 automatic plate washer (PerkinElmer). In the third step, wells were incubated with 300 ng of anti-alpha-amylase polyclonal antibody, Eu31-chelates-labelled, in 200 mL of assay buffer (DELFIA) per well with gentle shaking. Finally, after washing, 200 mL of enhancement solution (DELFIA) was added per well and incubated for 5 min with gentle shaking to allow Eu31 to form fluorescent chelates. The enhanced fluorescence, proportional to the quantity of alpha-amylase in

the sample, was measured in a VICTOR 1420 multilabel counter (PerkinElmer).

Analytical validationIn general, to validate the methods, which will be described below,

precision, accuracy, and detection limit were evaluated. Six pools of saliva were used for intra-assay precision studies and another 6 pools were used for inter-assay precision studies. Each pool was formed with the saliva of 4 different sheep, mixing the same volume of each sample. All the determinations were carried out in duplicate. Intra-assay precision was calculated by measuring each of the pools selected 6 times, in the same analytical series. Inter-assay precision was calculated by measuring each of the pools once a day for 6 dif-ferent days. The samples were stored in separate vials (aliquots) to avoid the possible effect of repetitive thawing and freezing.

Since to the authors’ knowledge, no reference material or stan-dards for ovine alpha-amylase are commercially available, accuracy was indirectly estimated in the study by linearity assays. For linear-ity under dilution, 1 sample with a high concentration of alpha-amylase was analyzed in triplicate. The samples were diluted in assay buffer at 1:4, 1:8, and 1:16 of the initial samples. The average of the 3 initial measurements (1:2 dilution) was used to calculate the expected value for each dilution. The detection limit was defined as the lowest concentration of analyte that could be distinguished from a specimen of zero value. It was calculated based on data from 12 replicate determinations of the zero standard (assay buffer) as mean value plus 2 standard deviations.

Clinical validationThe stress-induction protocol used in this study was adapted from

Cook and Jacobson (16). Briefly, sheep were subjected to acute stress in the form of confronting a sheepdog for 1 min while the sheep remained in a confined space and protected from the sheepdog, which was tied up and barking. This model of stress was selected because these sheep had never been in contact with a sheepdog before and it has been demonstrated that the presence of a dog clearly elevates salivary cortisol (16). Although the sheep exhibited considerable stress behaviorally in the presence of the dog, these parameters such as nervousness, stiffness of limbs, etc were not taken into account since the confinement of the animals changed the natural ethogram of sheep in warning and threat conditions.

The sampling time (1 min) was modified from the original Cook and Jacobson protocol, during which the sheep were working with dogs for 7 to 10 min. This seemed excessive for our sampling condi-tions as the sheep remained in a confined space in our study and a long sampling time could have caused excessively high stress in animals, with the potential of serious biological consequences.

Samples were collected from 8:00 to 11:00 o’clock in the morning. Saliva samples were obtained before the experimental procedure (basal sample TB) and after 15 min (T15), 30 min (T30), and 60 min (T60) of the experimental stress episode. The choice of sampling times was based on previous evidence on the response of alpha- amylase in the saliva of other animal species (10,11). Salivary samples were obtained by introducing a small sponge into the sheep’s mouth for at least 1 min with the help of a metal rod as previously described for other species (10). Afterwards, the sponges



were inserted into a specific tube (Salivette; Sarstedt, Numbrecht, Germany) and kept refrigerated (4°C) until arrival at the laboratory. The tubes were centrifuged for 10 min at 3000 3 g and samples were collected and stored at 220°C until analysis.

All procedures involving animals were approved by the Murcia University Ethics Committee (CEEA 84/2014) and followed the recommendations of the European Parliament and the Council of September 22, 2010 on the protection of animals used for scientific purposes (2010/63/EU).

Statistical analysisFor the validation study, arithmetic means, medians, and coef-

ficients of variation (CVs) were calculated by descriptive sta-tistical procedures. Linearity under dilution was accomplished by linear regression analysis comparing the measured levels of alpha-amylase with the expected values. For the stress-induction study, the Kolmogorov-Smirnov test was conducted to assess the normality of data, giving a nonparametric distribution. Data were then log-transformed and 1-way analysis of variance (ANOVA) of repeated measures and Tukey post HOC test were used for statisti-cal processing. All statistical analyses were done using a statistical package (GraphPad Prism Version 5.0; GraphPad Software, La Jolla, California, USA) and a spreadsheet (Microsoft Excel 2010; Microsoft, Redmond, Washington, USA). The significance level was set at P , 0.05.

Re s u l t s

Analytical validation: Validation of the TR-IFMA for determining salivary alpha-amylase

Results of the precision study are shown in Table I. The intra-assay coefficient of variations (CVs) were 6.1% on average for pools containing high alpha-amylase concentration, 6.55% on average for pools containing medium alpha-amylase concentration, and 10.57% for pools with low enzyme concentration. The inter-assay CVs were 13.55% on average for pools containing high alpha-amylase concentration, 19.3% on average for pools containing medium alpha-amylase concentration, and 30.8% for pools with low enzyme

concentration. The linear regression analysis provided a correla-tion coefficient of 0.965 in pools with high levels of alpha-amylase (Figure 1). The analytical limit of detection was 0.09 mg/dL.

Clinical validation: Results of stress inductionThree of the 11 sheep did not show increases in cortisol levels in

half of the baseline levels. Statistical analysis of salivary cortisol levels showed that levels increased significantly between the values of time TB and T15 (P , 0.001), between TB and T30 (P , 0.0001), between T15 and T60 (P , 0.01), and between T30 and T60 (P , 0.01) (Figure 2). Salivary alpha-amylase increased significantly between the values of time TB and T15 (P , 0.01). The Tukey post HOC test showed a P-value of 0.0091 (Figure 3).

D i s c u s s i o nIt is first necessary to discuss the major discrepancy found with

other authors who claimed that there was no salivary alpha-amylase in sheep (13–15). The TR-IFMA presented in this study is an assay that quantifies the amount of salivary alpha-amylase. The assay is both more sensitive and more specific than other types such as enzymatic assays (15). It has been shown that enzymatic assays have pH variations or other conditions such as enzymatic hydrolysis that can affect the enzymatic activity (17,18). Moreover, the use of an enzyme assay already established for 1 species for another distinct

Table I. Intra- and inter-assay coefficients of variation (CVs) obtained for saliva pools with high, medium, and low alpha-amylase concentration

Intra-assay Inter-assay Mean (SD) CV (%) Average Mean (SD) CV (%) AverageHigh concentration (ng/mL) Pool 1 2.43 (0.04) 1.6 6.1 2.47 (0.28) 11.6 13.55 Pool 2 2.84 (0.3) 10.6 3.1 (0.48) 15.5

Medium concentration (ng/mL) Pool 1 0.72 (0.07) 9.6 6.55 0.83 (0.23) 27.2 19.3 Pool 2 0.81 (0.02) 3.5 0.84 (0.09) 11.4

Low concentration (ng/mL) Pool 1 0.15 (0.01) 12.9 10.57 0.37 (0.13) 36.1 30.8 Pool 2 0.4 (0.03) 8.25 0.55 (0.14) 25.5SD — standard deviation.

Figure 1. Linearity under dilution of an ovine saliva pool with high alpha-amylase concentration.



species must be approved beforehand. This has apparently not been done in other studies consulted (13) that rule out the presence of salivary amylase in sheep based on an old citation (15). In addition, our personal experience in using salivary proteomics (19) allows us to state that it is difficult to locate spots of proteins in very low con-centration, as in the case of salivary alpha-amylase in sheep, which agrees with the results obtained in other studies (14) not identified this enzyme in the salivary proteome sheep, in our view due to a low concentration of the protein and not its absence in sheep. For these reasons, we believe that this study is valid and reliable despite previous evidence to the contrary in the literature.

On the other hand, the results of studies to determine the level of stress in farm animals during routine handling are often highly vari-able and are difficult to interpret from an animal welfare standpoint (2). The use of saliva samples has many benefits when developing studies of stress in animals. This kind of sampling is noninvasive and takes the animal’s welfare into account, in addition to being simple to carry out, inexpensive, and repeatable (10). This easy method of sampling is strongly supported in the literature on veterinary medicine (20).

The stress model in this study uses fear and novelty as the princi-pal factors to cause stress in the sheep. Fear may be a major psycho-logical stressor in extensively handled cattle (2) and novelty is also a very strong stressor (21), especially when an animal is suddenly confronted with something new (2). Cortisol (salivary and serum) is the reference biomarker to assess this level of stress, with the con-centration of salivary cortisol being a suitable marker comparable to the concentration of serum cortisol (22). Although some authors state that cortisol levels are highly variable and absolute comparisons should not be made between studies (2), the results of salivary cor-tisol after stress induction by confrontation with a sheepdog in this study were in accordance with the results of previous studies (16). Cortisol concentration is influenced by factors other than physical or psychological stress, however, such as environmental condi-

tions and metabolic factors (23). For this reason, the use of salivary alpha-amylase was studied as an alternative. The studies of ovine alpha-amylase have been developed with plasma samples or with genetic techniques, in reference to milk traits (24) or dietary effects (25,26). An immunofluorometric method (TR-IFMA) was chosen because of its advantages over other methods such as enzymatic methods, which are simpler to carry out. These advantages include high sensitivity, no interference by hemolysis (27), or a high dynamic range (28). This allows substances present in small quantities to be measured in different organic fluids, such as saliva (27). Purified human salivary amylase was used to produce the antibodies and also as a standard because purified ovine amylase is not available commercially to the authors’ knowledge. Previously, our group vali-dated another TR-IFMA with similar assay features for determining salivary alpha-amylase in horses (11).

The results of our validation of immunofluorometric assay suggest that the method could be adequate for measuring alpha-amylase in saliva samples. This is due to the acceptance of a CV lower than 20% for analytical determinations (29). Only the inter-assay CV of pools with low enzyme concentration exceeds this value. For that reason, the results of samples with low concentration of salivary alpha-amylase should be considered with caution. Since the limit of detection of this assay for salivary amylase was very low, all saliva samples could be measured with this method because all concentra-tions obtained in saliva samples in this study were higher than the limit of detection. This low limit of detection shows a high sensitiv-ity of the assay. As far as we know, there are no positive studies for salivary alpha-amylase in sheep with which to compare these results.

The variability of results for salivary amylase obtained during the experience of stress must be kept in mind. It is possible to find values up to 80 times greater among animals, which makes it dif-ficult to obtain meaningful statistical analysis. This inter-individual variability is because the responses of the autonomic nervous system (ANS) are subject to the individual specificity phenomenon (30).

Figure 3. Salivary alpha amylase concentration in sheep stressed by confrontation with sheepdog. The first sample was taken before the immobilization of animals (TB) and the remaining samples were taken at 15, 30, and 60 min after stress induction (T15, T30, and T60, respec-tively). Horizontal lines indicate median values for each time point. Error bars indicate 2.5 and 97.5 percentiles. a = P , 0.01 with TB.

Figure 2. Salivary cortisol concentration in sheep stressed by confronta-tion with sheepdog. The first sample was taken before the immobilization of animals (TB) and the remaining samples were taken at 15, 30, and 60 min after stress induction (T15, T30, and T60, respectively). Horizontal lines indicate median values for each time point. Error bars indicate 2.5 and 97.5 percentiles. a = P , 0.001 with TB; b = P , 0.0001 with TB; c = P , 0.01 with T15; and d = P , 0.01 with T30.



This factor could explain why some animals did not experience an increase in alpha-amylase levels after stress was induced.

Considering the variations in salivary amylase throughout the different time points, a statistically significant increase in alpha-amylase concentration was obtained with TR-IFMA before (TB) and after stress induction (T15). It could be concluded that the observed increases were the result of stress induction caused by the presence of the dog in the same pen as the sheep. Although there are factors that could influence amylase levels, the origin of the observed varia-tions should not be questioned because it has been demonstrated that amylase is independent of salivary flow rate (31). On the other hand, although the release of this enzyme is regulated by a diurnal circadian rhythm (32), sampling was carried out in such a short time that this would not have affected it.

This study described and validated a new assay for measuring alpha-amylase in the saliva of sheep. The results obtained could be important in terms of the possibility of using salivary alpha-amylase as a noninvasive biomarker of activation of the autonomic nervous system (ANS). We are aware, however, that the small sample size is a limitation and that this study could therefore be considered a preliminary study. Further studies of salivary alpha-amylase and the factors that could influence the enzyme are required in order to evaluate the use of this protein (and its quantification) as a stress biomarker in sheep so that it could be included in a multi-analytes panel to accurately determine the stress state of sheep.

A c k n o w l e d g m e n t sThe authors thank Arthur Bustamante and Santiago Zuñiga for

assistance with sampling and F. Tecles and J. Cerón for technical support. This study was supported by a personal grant from the Ministry of Education and Science and by the Spanish Ministry of Education and Science (AGL 2009-08509).

Re f e r e n c e s1. Le Neindre P, Boivin X, Boissy A. Handling of extensively kept

animals. Appl Anim Behav Sci 1996;49:73–81.2. Grandin T. Assessment of stress during handling and transport.

J Anim Sci 1997;75:249–257.3. Yardimci M, Sahin EH, Cetingul IS, Bayram I, Aslan R, Sengor E.

Stress responses to comparative handling procedures in sheep. Animal 2013;7:143–150.

4. Merlot E, Mounier AM, Prunier A. Endocrine response of gilts to various common stressors: A comparison of indicators and methods of analysis. Physiol Behav 2011;102:259–265.

5. Broom DM, Goode JA, Hall SJ, Lloyd DM, Parrott RF. Hormonal and physiological effects of a 15 hour road journey in sheep: Comparison with the responses to loading, handling and pen-ning in the absence of transport. Br Vet J 1996;152:593–604.

6. Baldock NM, Sibly RM. Effects of handling and transportation on the heart rate and behaviour of sheep. Appl Anim Behav Sci 1990;28:15–39.

7. Goddard P, Waterhouse T, Dwyer C, Stott A. The perception of the welfare of sheep in extensive systems. Small Rumin Res 2006;62:215–225.

8. Chiu SK, Collier CP, Clark AF, Wynn-Edwards KE. Salivary cor-tisol on ROCHE Elecsys immunoassay system: Pilot biological variation studies. Clin Biochem 2003;36:211–214.

9. Tecles F, Fuentes-Rubio M, Tvarijonaviciute A, Martínez-Subiela S, Fatjó J, Cerón JJ. Assessment of stress associated with an oral public speech in veterinary students by salivary biomarkers. J Vet Med Educ 2014;41:37–43.

10. Fuentes M, Tecles F, Gutiérrez A, Otal J, Martínez-Subiela S, Cerón JJ. Validation of an automated method for salivary alpha-amylase measurements in pigs (Sus scrofa domesticus) and its application as a stress biomarker. J Vet Diagn Invest 2011;23:282–287.

11. Fuentes-Rubio M, Fuentes F, Otal J, et al. Measurements of sali-vary alpha-amylase in horse: Comparison of 2 different assays. J Vet Behav 2015;10:122–127.

12. Muneta Y, Minagawa Y, Nakane T, Shibahara T, Yoshikawa T, Omata Y. Interleukin-18 expression in pig salivary glands and salivary content changes during acute immobilization stress. Stress 2011;14:549–556.

13. Behringer V, Deschner T, Möstl E, Selzer D, Hohmann G. Stress affects salivary alpha-amylase activity in bonobos. Physiol Behav 2012;105:476–482.

14. Lamy E, da Costa G, Santos R, et al. Sheep and goat saliva pro-teome analysis: A useful tool for ingestive behavior research? Physiol Behav 2009;98:393–401.

15. Chauncey HH, Henrigues BL, Tanzer JM. Comparative enzyme activity of saliva from the sheep, hog, dog, rabbit, rat, and human. Arch Oral Biol 1963;8:615–627.

16. Cook CJ, Jacobson LH. Salivary cortisol as an indicator of stress in sheep (Ovis ovis). N Z Vet J 1995;43:248.

17. Turner MD, Ship JA. Salivary secretion in health and disease. In: Won DT, ed. Salivary Diagnostics. Ames, Iowa:Wiley-Blackwell, 2008:60–69.

18. Schmidt-Nielsen B. The pH in parotid and mandibular saliva. Acta Physiol Scand 1946;11:104–110.

19. Fuentes-Rubio M, Cerón JJ, de Torre C, Escribano D, Gutiérrez AM, Tecles F. Porcine salivary analysis by 2-dimensional gel electrophoresis in 3 models of acute stress: A pilot study. Can J Vet Res 2014;78:127–132.

20. Prickett JR, Zimmerman JJ. The development of oral fluid-based diagnostics and applications in veterinary medicine. Anim Health Res Rev 2010;5:1–10.

21. Dantzer R, Mormede P. Stress in farm animals: A need for reevaluation. J Anim Sci 1983;57:6.

22. Yates DT, Ross TT, Hallford DM, Yates LJ, Wesley RL. Technical note: Comparison of salivary and serum cortisol concentrations after adrenocorticotropic hormone challenge in ewes. J Anim Sci 2010;88:599–603.

23. Jaskulke S, Manteuffel G. No apparent effect of an experimental narrow confinement on heart activity and cortisol in domestic pigs. Animal 2011;5:433–438.

24. Calvo JH, Marcos S, Beattie AE, González C, Jurado JJ, Serrano M. Ovine alpha-amylase genes: Isolation, linkage mapping and association analysis with milk traits. Anim Genet 2004;35: 329–332.



25. Swanson KC, Matthews JC, Matthews AD, Howell JA, Richards CJ, Harmon DL. Dietary carbohydrate source and energy intake influence the expression of pancreatic alpha-amylase in lambs. J Nutr 2000;130:2157–2165.

26. Ohbo M, Katoh K, Sasaki Y. Effects of saturated fatty acids on amylase release from exocrine pancreatic segments of sheep, rats, hamsters, field voles and mice. J Comp Physiol B 1996;166: 305–309.

27. Parra MD, García AJ, Bernal LJ, Landete-Castillejos T, Cerón JJ. Progesterone determination in Iberian red deer by time-resolved fluorometry: An alternative method to RIA. J Exp Zool A Comp Exp Biol 2004;301:472–476.

28. Tarkkinen P, Palenius T, Lövgren T. Ultrarapid, ultrasensitive one-step kinetic immunoassay for C-reactive protein (CRP) in whole blood samples: Measurement of the entire CRP concen-tration range with a single sample dilution. Clin Chem 2002; 48:269–277.

29. Guidance for Industry. Bioanalytical Method Validation 2001. US Department of Health and Human Services [homepage on the Internet]. Available from: http://www.fda.gov/down loads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm070107.pdf Last accessed January 7, 2016.

30. Granger DA, Kivlighan KT, el-Sheikh M, Gordis EB, Stroud LR. Salivary alpha-amylase in biobehavioral research: Recent devel-opments and applications. Ann N Y Acad Sci 2007;1098:122–144.

31. Rohleder N, Wolf JM, Maldonado EF, Kirschbaum C. The psy-chosocial stress-induced increase in salivary alpha-amylase is independent of saliva flow rate. Psychophysiology 2006;43: 645–652.

32. Nater UM, Rohleder N, Schlotz W, Ehlert U, Kirschbaum C. Determinants of the diurnal course of salivary alpha-amylase. Psychoneuroendocrinology 2007;32:392–401.


Article


I n t r o d u c t i o nTumor necrosis factor alpha (TNF-a), a cellular immune factor that

is mainly secreted by macrophages, monocytes, and lymphocytes, stimulates T-cells to release cytokines [e.g., interleukin 1b (IL-1b), IL-2, and IL-6], which are involved in immune reactions (1,2). In the medical field, the study of TNF-a has been focused mostly on the promoter region, where point mutations are associated with either resistance or susceptibility to disease. For example, TNF-a 308(G→A) is associated with increased risk of premature birth, vit-iligo, and duodenal ulcers (3–5), TNF-a 238(G→A) with increased risk of hepatitis C (6), and TNF-a 204(C→T) with a reduced risk of severe acute respiratory syndrome (7). However, in studies about mutations in the promoter region of porcine TNF-a, the mutation g.6464(C→T) was found to be associated with back fat accumulation

in Large White pigs (8), and the polymorphic loci were 791 base pairs upstream from the transcription initiation site. In addition, TNF-a is a cytokine that plays opposing roles in the context of infectious disease pathogenesis. Many studies have found that upregulated expression of TNF-a contributes to improving immune response and resistance to infection in pigs (9–11). The promoter, an integral upstream regulatory region of a gene, controls the initiation of gene transcription as well as the level of gene expression.

In this study TNF-a mRNA expression in 11 tissues (heart, liver, spleen, lung, kidney, stomach, muscle, thymus, lymph node, duo-denum, and jejunum) was analyzed in 8 piglets resistant to entero-toxigenic Escherichia coli (ETEC) F18 and 8 ETEC F18-susceptible piglets of the Large White breed. The technique of polymerase chain reaction single-strand conformation polymorphism (PCR-SSCP) was used to detect −791(C→T) mutations in the TNF-a promoter and to

Effects of the −791(C➞T) mutation in the promoter for tumor necrosis factor alpha on gene expression and resistance of Large White pigs

to enterotoxigenic Escherichia coli F18Ying Liu, Chaohui Dai, Li Sun, Guoqiang Zhu, Shenglong Wu, Wenbin Bao

A b s t r a c tTumor necrosis factor alpha (TNF-a) plays an important role in the immune system. In this study, TNF-a expression was analyzed in 11 tissues of 8 piglets resistant to enterotoxigenic Escherichia coli (ETEC) F18 and 8 ETEC F18-susceptible piglets from the Large White breed. The expression levels of TNF-a were high in immune organs (spleen, lung, thymus, and lymph nodes). The levels were higher in ETEC F18-resistant piglets than in ETEC F18-susceptible piglets, with significant differences in spleen, kidney, thymus, lymph node, and duodenum (P , 0.05). The mutation TNF-a −791(C➝T) and 3 genotypes (CC, CT, and TT) were identified. The TNF-a expression levels in the spleen, kidney, lymph nodes, and duodenum were significantly higher in the TT pigs than in the CC pigs (P , 0.05). Thus, TNF-a −791(C➝T) has significant effects on mRNA expression and may regulate ETEC F18 resistance of weaning piglets. Therefore, the −791(C➝T) mutation of the TNF-a gene could be considered an important potential genetic marker of ETEC F18 resistance.

R é s u m éLe facteur alpha nécrosant des tumeurs (TNF-a) joue un rôle important dans le système immunitaire. Dans la présente étude, l’expression de TNF-a a été analysée dans 11 tissus provenant de huit porcelets résistants à une souche d’Escherichia coli entérotoxinogène (ETEC) F18 et huit porcelets sensibles à une souche ETEC F18 de race Large White. Les degrés d’expression de TNF-a étaient élevés dans les organes immunitaires (rate, poumon, thymus, et nœuds lymphatiques). Les niveaux étaient plus élevés chez les porcelets résistants à la souche ETEC F18 que chez les porcelets sensibles à la souche ETEC F18, avec des différences significatives dans la rate, rein, thymus, nœud lymphatique, et duodénum (P , 0,05). La mutation TNF-a –791(C➝T) et 3 génotypes (CC, CT, et TT) ont été identifiés. Les niveaux d’expression de TNF-a dans la rate, rein, nœuds lymphatiques, et duodénum étaient significativement plus élevés dans les porcs TT que dans les porcs CC (P , 0,05). Ainsi, TNF-a –791(C➝T) avait des effets significatifs sur l’expression d’ARNm et pourrait réguler la résistance envers ETEC F18 chez des porcelets au sevrage. Ainsi, la mutation –791(C➝T) du gène du TNF-a pourrait être considérée comme un important marqueur génétique potentiel de la résistance envers les ETEC F18.


Department of College of Animal Science and Technology, Key Laboratory for Animal Genetics, Breeding, Reproduction and Molecular Design of Jiangsu Province, Yangzhou University, Yangzhou, Jiangsu Province, People’s Republic of China (Liu, Dai, Sun, Wu, Bao); Department of College of Veterinary Medicine, Yangzhou University, Yangzhou, Jiangsu Province, People’s Republic of China (Zhu).

Ying Liu and Chaohui Dai contributed equally to this study.

Address all correspondence to Professor Wenbin Bao; telephone: 86-514-87979316; fax: 86-514-87350440; e-mail: [email protected]

Received July 12, 2015. Accepted September 28, 2015.



determine the mRNA expression levels for TNF-a in order to assess the feasibility of using the −791(C . T) mutation as a genetic marker and provide some theoretical and experimental basis for disease-resistance breeding based on the TNF-a gene.


Experimental materials and sample collectionThe animal study proposal was approved by the Institutional

Animal Care and Use Committee (IACUC) of the Yangzhou University Animal Experiments Ethics Committee with the per-mit number SYXK(Su) IACUC 2012-0029. All piglet experimental procedures were done in accordance with the Regulations for the Administration of Affairs Concerning Experimental Animals approved by the State Council of the People’s Republic of China.

In previous studies we developed a V-secretion system and a receptor-binding assay for identifying resistance and sensitiv-ity to ETEC F18 (12–14). In the present study we conducted the F18-adhesion test in 50 Large White piglets from 8 families at Kangle Farming Company (Changzhou, Jiangsu) that had almost the same weight at birth and at 28 d of age, around the time of weaning, when they are most vulnerable to ETEC F18 infection. The piglets had been allowed access to food and water ad libitum and kept in the same conditions until humanely euthanized at age 28 d. The adhesion test was done according to the method by Liu et al (12). When a large amount of adherence by F18ab-expressing fimbriae of the standard ETEC strain to intestinal epithelial cells was displayed the piglets were identified as E. coli F18-susceptible (Figure 1A), whereas when no adherence was displayed the piglets were identified as E. coli F18-resistant (Figure 1B).

For 8 pairs of ETEC F18-susceptible and ETEC F18-resistant were full-sib piglets, tissue samples from 11 organs (heart, liver, spleen, lung, kidney, stomach, muscle, thymus, lymph node, duodenum, and jejunum) were collected, immersed in liquid nitrogen, and stored at 270°C.

Approximately 1.0 g of ear tissue was collected from 201 Large White sows from the 8 families and placed in 1.5-mL Eppendorf

tubes. Genomic DNA was extracted by means of a modified phenol and chloroform protocol (15). The purity and concentration of the genomic DNA were assessed in a NanoDrop-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, Delaware, USA). The genomic DNA was diluted to 100 ng/mL and stored at 220°C.

Primer designPrimers for PCR and real-time PCR were designed according to the

TNF-a gene sequence obtained from the GenBank database (acces-sion no. X54859; National Center for Biotechnology Information, Bethesda, Maryland, USA). The primers (Table I) were synthesized by Shanghai Invitrogen Biotechnology (Shanghai, China). The genes for glyceraldehyde 3-phosphate dehydrogenase, TATA box-binding protein 1, and b-actin were used as internal controls.

Analysis by PCR-SSCPThe reaction mixture (20 mL) contained 100 ng of DNA, 5 pmoL

of each primer, 10 mL of PCR Master Mix (Tiangen Biotech, Beijing,

Figure 1. A — large numbers of enterotoxigenic Escherichia coli (ETEC) F18 adhering to an intestinal epithelial cell of a susceptible Large White piglet. B — almost no bacterial adhesion in an ETEC F18-resistant piglet. Oil immersion lens; scale bar = 20 mm.

Table I. Primers for polymerase chain reaction single-strand conformation polymorphism (PCR-SSCP) and real-time PCR

Gene Method Sequence (59→39)TNF-a PCR-SSCP F: GCCCGCCATGGTGGGTTTGT R: TGATTTCCGAACAGGGCTCAGGTA

TNF-a Real-time PCR F: CGCCCACGTTGTAGCCAATGT R: CAGATAGTCGGGCAGGTTGATCTC

GAPDH Real-time PCR F: ACATCATCCCTGCTTCTACTGG R: CTCGGACGCCTGCTTCAC

TBP1 Real-time PCR F: ACATCATCCCTGCTTCTACCGG R: CTCGGACGCCTGCTTCAC

ACTB Real-time PCR F: TGGCGCCCAGCACGATGAAG R: GATGGAGGGGCCGGACTCGTTNF-a — gene for tumor necrosis factor alpha; GAPDH, TBP1, and ACTB — genes for glyceraldehyde 3-phosphate dehydrogenase, TATA box-binding protein 1, and b-actin, used as internal controls; F — forward; R — reverse.



China), and sterilized distilled water. The PCR protocol consisted of 1 cycle at 95°C for 5 min, followed by 30 cycles at 94°C for 40 s, 63°C for 40 s, and 72°C for 45 s, and an extension step at 72°C for 10 min. The PCR products were confirmed by electrophoresis in 1% agarose gel stained with ethidium bromide. Then the SSCP of the PCR products was detected by polyacrylamide gel electrophoresis. Loading buffer (7 mL) was mixed with 15 mL of PCR product, which was denatured at 98°C for 15 min and incubated in ice for 10 min. The denatured ice-cold mixtures were loaded onto a 12% non-denaturing polyacrylamide gel (Acr:Bis = 29:1). Electrophoresis was done at 130 V overnight, and then the products were silver-stained. According to the PCR-SSCP results, PCR products of different homozygotes were sequenced in an ABI PRISM 377 automatic DNA sequencer (Applied Biosystems, Foster City, California, USA). The PCR products from genotypes CC and TT were sequenced by Sangon Biotech Company, Shanghai, China, and the results were analyzed with the use of DNAMAN 5.0 software (Lynnon Biosoft, USA).

Total RNA extraction and real-time PCRTotal RNA was extracted from the tissue samples (50 to 100 mg)

with the use of Trizol (TaKaRa Biotechnology, Dalian, China). Precipitated RNA was dissolved in 20 mL of RNase-free water. Qualitative and quantitative measurements of RNA were assessed by agarose gel electrophoresis and the ChemiDoc XRS1 Bio-imaging system (Bio-Rad Laboratories, Hercules, California, USA), respectively. The concentration of total RNA was measured by the NanoDrop-1000 spectrophotometer.

The 10-mL reaction mixture for cDNA synthesis contained 2 mL of 53 PrimerScript buffer, 0.5 mL of PrimerScript RT Enzyme Mix I, 0.5 mL of oligo-dT (a short sequence of deoxythymine nucleotides),

0.5 mL of random hexamers, 500 ng of total RNA, and RNase-free H2O. The reaction was done at 37°C for 15 min and then 5 s at 85°C.

Real-time PCR amplification was done with a 20-mL reaction mix-ture containing 1 mL of cDNA (100 to 500 ng), 0.4 mL of each forward and reverse primer (10 mM each), 0.4 mL of 503 ROX Reference Dye II, 10 mL of 23 SYBR Green real-time PCR Master Mix, and 7.8 mL of double-distilled water. The PCR protocol included 1 cycle at 95°C for 15 s, followed by 40 cycles at 95°C for 5 s and 62°C for 34 s. The dissociation curve was analyzed after amplification. A melting temperature (Tm) peak at 85 6 0.8°C was used to determine the speci-ficity of the PCR amplification. Each sample was analyzed 3 times.

The mRNA expression of TNF-a in different tissues was analyzed by real-time PCR. We first analyzed the expression levels of TNF-a mRNA in 11 tissues from Large White piglets resistant or susceptible to enterotoxigenic Escherichia coli (ETEC) F18. Then other 201 Large White sows were analyzed by PCR-SSCP, and grouped according to genotype, we chose 6 pigs from each genotype to analyze the expression levels of TNF-a mRNA in the 11 tissues according to piglet genotype.

Statistical analysesGene frequencies and genotype were calculated by the Hardy–

Weinberg equilibrium principle: p = P 1 H/2, q = Q 1 H/2, x2 = Sd2/e; d = e − o, which is the difference between the predicted and obtained values. The variables p and q represent allele frequencies at certain positions.

The 22DDCt method (16) was used to analyze the real-time PCR results:

Table II. Expression levels of TNF-a mRNA in 11 tissues from Large White piglets resistant or susceptible to enterotoxigenic Escherichia coli (ETEC) F18

Expression level (mean 6 standard deviation) Difference Tissue Resistant (n = 8) Susceptible (n = 8) multiplesHeart 1.076 6 0.435 0.970 6 0.284 1.109Liver 9.266 6 2.590 8.602 6 2.489 1.077Spleen 146.270 6 20.99a 85.808 6 18.127b 1.709Lung 72.300 6 21.083 63.981 6 18.791 1.130Kidney 17.521 6 3.718a 8.650 6 3.029b 2.025Stomach 11.920 6 3.338 6.405 6 2.458 1.861Muscle 0.246 6 0.040 0.197 6 0.068 1.253Thymus 76.252 6 8.353a 36.127 6 10.276b 2.111Lymph node 24.811 6 4.313a 12.806 6 5.657b 1.937Duodenum 8.882 6 2.015a 5.241 6 0.878b 1.695Jejunum 7.154 6 1.835 5.217 6 1.547 1.371DDCt = [average cycle threshold (Ct) value of the target gene in the tested group − average Ct value of the housekeeping gene in the tested group] − (average Ct value of the control gene in the control group − average Ct value of the housekeeping gene in the control group).a,b Within a row, the means without a common superscript differ significantly (P , 0.05). Figure 2. Result of polymerase chain reaction single-strand conforma-

tion polymorphism of the gene fragment of TNF-a. Lane 1 represents genotype CC, lane 2 genotype CT, and lane 3 genotype TT.



�DDCt = (average cycle threshold [Ct] value of the target gene in the tested group 2 average Ct value of the housekeeping gene in the tested group) 2 (average Ct value of the control gene in the control group 2 average Ct value of the housekeeping gene in the control group).

Statistical analyses were done with SPSS software, version 15.0 (SPSS, Chicago, Illinois, USA). Data were expressed as mean 6 stan-dard deviation. Student’s t-test was used to determine differences in mRNA expression, or immune index.

Re s u l t sThe TNF-a mRNA expression levels in the spleen, lung, thymus,

and lymph node tissues were high in both the ETEC F18-resistant and the ETEC F18-susceptible animals (Table II). The expression levels in the duodenum, immune tissues (spleen, thymus, and lymph node) and kidney were significantly higher (P , 0.05) in the resistant piglets than in the susceptible piglets.

According to the results of the 1% agarose gel electrophoresis the length of the amplified fragment was in agreement with the pre-dicted fragment length. The SSCP analysis revealed 3 genotypes: CC, CT, and TT (Figure 2). With a sequence for TNF-a identical to that in GenBank (accession no. X54859), the CC genotype was defined as the wild type. In comparison, the TT genotype had a C/T substitu-tion mutation at nucleotide −791, located in the promoter (Figure 3).

The frequencies of genotypes CC, CT, and TT were 0.0857, 0.5571, and 0.3571, respectively; T was the dominant allele. The results revealed that TNF-a −791(C→T) was consistent with the Hardy–Weinberg equilibrium (Table III).

When grouped according to genotype (Table IV), the TNF-a expression levels in the spleen, kidney, lymph nodes, and duode-num were significantly lower in the CC pigs than in the TT pigs (P , 0.05), and the trend in expression level was CC , CT , TT.

D i s c u s s i o nThe expression of TNF-a, which depends on environmental

stressors, triggers inflammation and apoptosis, and it prevents

bacterial proliferation (17). The real-time PCR results in this study revealed high expression levels of TNF-a mRNA in the immune system (in spleen, thymus, and lymph nodes) and in the lungs. The immune system confers resistance against infections; immune cells generate several factors that are distributed to various organs of the body. As a vital organ in the respiratory system, the lungs frequently communicate with the outside environment. Neutrophils and T-cells produce high levels of immune factors in response to various external stimuli; however, this does not occur in the lungs (18). The high levels of TNF-a mRNA expression in the immune system were attributed to the biofunctions of TNF-a.

It has been reported that TNF-a can promote dendritic cell dif-ferentiation (19,20). In piglets infected with porcine reproductive and respiratory syndrome virus (PRRSV), TNF-a inhibited the rep-lication of PRRSV (21). Furthermore, lipopolysaccharides enhance TNF-a transcription and translation (22). In our study, the TNF-a expression levels in the ETEC F18-resistant piglets were higher than those in the ETEC F18-susceptible piglets. Additionally, there were significant differences in the TNF-a expression levels in the spleen, kidney, lymph nodes, and duodenum between piglets with the TT and CC genotypes. The duodenum, which is the first intestinal segment, is vulnerable to bacterial infections. In addition to its role in digestion and absorption of nutrients, the intestine has immune and endocrine functions (23,24). However, the intestinal immune function depends on the structural integrity and function of the intestinal epithelial cells. With infection by pathogenic bacteria, the intestinal levels of IL-8, IL-1, and TNF-a increase; thus, both the immune system and the absorption of nutrients are affected (25). Therefore, intestinal TNF-a expression levels are closely linked to health: high expression levels improve the resistance to several pathogenic bacteria.

In addition to its role in the immune system, TNF-a plays an important role in the development of skeletal muscle via proteolytic digestion of the protein ubiquitin and the mitogen-activated protein kinase signal transduction pathway (26,27). As well, TNF-a may be involved in pregnancy via follicle development and ovulation, cor-pus luteum formation and regression, cyclic endometrium function through the secretion of prostaglandin (PG) F2 alpha and PGE2 (28), and embryo implantation and immune suppression (29).

Figure 3. Sequencing peak chart of the CC and TT genotypes. Because the sequence was identical to that in GenBank (accession no. X54859), the CC genotype was defined as the wild type. In comparison, the TT genotype had a C/T substitution mutation at nucleotide −791, located in the promoter.



This study showed that the TT genotype of TNF-a −791(C→T) may contribute to higher levels of TNF-a expression in intestinal tissue (duodenum) and immune tissues (spleen and lymph nodes) than the CT and CC genotypes, and these higher levels could play an important role in conferring ETEC F18 resistance. Therefore, this mutation could be considered as a potential genetic marker for ETEC F18 resistance in piglets of the Large White breed, and we should further assess this possibility in larger populations and several generations.

A c k n o w l e d g m e n t sThis study was funded by the National Natural Science Funds

(grants 31172183 and 31372285), the Genetically Modified Organisms Technology Major Project (2014ZX0800601B), and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

Re f e r e n c e s1. Bradley JR. TNF-mediated inflammatory disease. J Pathol 2008;

214:149–160.2. Aggarwal BB. Signaling pathways of the TNF superfamily:

A double-edged sword. Nat Rev Immunol 2003;3:745–756.3. Menon R, Merialdi M, Betrán AP, et al. Analysis of association

between maternal tumor necrosis factor-a promoter polymor-phism (–308), tumor necrosis factor concentration, and preterm birth. Am J Obstet Gynecol 2006;195:1240–1248.

4. Namian AM, Shahbaz S, Salmanpoor R. Association of interferon- gamma and tumor necrosis factor alpha polymorphisms with

susceptibility to vitiligo in Iranian patients. Arch Dermatol Res 2009;301:21–25.

5. Lanas A, Garcia-González MA, Santolaria S, et al. TNF and LTA gene polymorphisms reveal different risk in gastric and duode-nal ulcer patients. Genes Immun 2001;2:415–421.

6. Höhler T, Kruger A, Gerken G, Schneider PM, Meyer zum Buschenfelde KH, Rittner C. Tumor necrosis factor alpha pro-moter polymorphism at position −238 is associated with chronic active hepatitis C infection. J Med Virol 1998;54:173–177.

7. Wei MT, Han YY, He L, et al. Study on the roles of TNF-a gene polymorphism at the promoter region in the susceptibility and symptoms of SARS coronavirus infection [in Chinese]. Acta Acad Med CPAF 2009;18:169–174.

8. Szydlowski M, Buszka A, Mackowski M, Lechniak D, Switonski M. Polymorphism of genes encoding cytokines IL6 and TNF is associated with pig fatness. Livest Sci 2011;136: 150–156.

9. Pomorska-Mól M, Markowska-Daniel I, Kwit K, et al. Immune and inflammatory response in pigs during acute influenza caused by H1N1 swine influenza virus. Arch Virol 2014;159: 2605–2614.

10. Gabner S, Worliczek HL, Witter K, Meyer FR, Gerner W, Joachim A. Immune response to Cystoisospora suis in piglets: Local and systemic changes in T-cell subsets and selected mRNA transcripts in the small intestine. Parasite Immunol 2014; 36:277–291.

11. Vicenova M, Nechvatalova K, Chlebova K, et al. Evaluation of in vitro and in vivo anti-inflammatory activity of biologically active phospholipids with anti-neoplastic potential in porcine model. BMC Complement Altern Med 2014;14:339.

Table IV. Expression levels of TNF-a mRNA in the 11 tissues according to piglet genotype

Expression level (mean 6 standard deviation), genotype (number of piglets)Tissue CC (6) CT (6) TT (6)Heart 1.295 6 0.443 0.903 6 0.299 0.975 6 0.353Liver 9.610 6 1.630 8.631 6 3.302 9.186 6 3.154Spleen 79.424 6 8.548a 127.241 6 48.479a,b 141.208 6 10.077b

Lung 72.835 6 20.146 66.006 6 22.501 69.802 6 24.143Kidney 7.770 6 1.856a 14.653 6 7.830a,b 16.398 6 2.531b

Stomach 7.447 6 2.274 8.321 6 5.696 11.535 6 3.819Muscle 0.243 6 0.051 0.184 6 0.080 0.247 6 0.048Thymus 15.380 6 6.988 15.190 6 6.297 25.360 6 5.419Lymph node 40.819 6 5.283a 58.688 6 21.960a,b 86.849 6 19.840b

Duodenum 5.373 6 1.275a 6.122 6 1.240a,b 9.6535 6 2.185b

Jejunum 6.205 6 1.542 5.674 6 2.738 6.858 6 1.967a,b Within a row, the means without a common superscript differ significantly (P , 0.05).

Table III. Genotypic and allelic frequencies of the TNF-a −791(C→T) mutation in the piglets

Frequency (number of samples)Sample Genotype Allelesize CC CT TT C T x2 valuea

201 0.254 (51) 0.522 (105) 0.224 (45) 0.515 0.485 0.420a x2

0.05 (1) = 3.84; x20.01 (1) = 6.63.



12. Liu L, Wang J, Zhao QH, et al. Genetic variation in exon 10 of the BPI gene is associated with Escherichia coli F18 susceptibility in Sutai piglets. Gene 2013;523:70–75.

13. Bao WB, Ye L, Pan ZY, et al. Microarray analysis of differential gene expression in sensitive and resistant pig to Escherichia coli F18. Anim Genet 2012;43:525–534. Epub 2011 Nov 7.

14. Wu SL, Yuan ZW, Ju HP, et al. Study on genetically susceptible piglets of small intestinal epithelium receptors to pathogenic F18 fimbrial Escherichia coli adhesion in vitro. Chin J Vet Sci 2006; 26:622–625.

15. Sambrook J, Russell DW. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press, 2001; Chapter 6: Protocol 1.

16. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2−DDCt method. Methods 2001;25:402–408.

17. Schmitz H, Fromm M, Bentzel CJ, et al. Tumor necrosis factor-alpha (TNF-a) regulates the epithelial barrier in the human intestinal cell line HT-29/B6. J Cell Sci 1999;112:137–146.

18. Raz E. Organ-specific regulation of innate immunity. Nat Immunol 2007;8:3–4.

19. Chomarat P, Dantin C, Bennett L, Banchereau J, Palucka AK. TNF skews monocyte differentiation from macrophages to dendritic cells. J Immunol 2003;171:2262–2269.

20. Clay H, Volkman HE, Ramakrishnan L. Tumor necrosis factor signaling mediates resistance to mycobacteria by inhibiting bac-terial growth and macrophage death. Immunity 2008;29:283–294.

21. Zheng WH, Du HL, Lin ZQ, Shan Y, Wang XN. Differentially expressed genes and gene function enrichment analysis in the

process of porcine alveolar macrophages infected with PRRSV [in Chinese]. Genomics Appl Biol 2011;29:1039–1046.

22. Feng LL, Zhang LP, Zhang GP, et al. In vitro effects of lipopoly-saccharide on the expression of IL-1 and TNF-a in porcine alveo-lar macrophage [in Chinese]. J Henan Agric Sci 2009;3:106–109.

23. Harari Y, Weisbrodt NW, Moody FG. Ileal mucosal response to bacterial toxin challenge. J Trauma 2000;49:306–313.

24. Kiyono H, Kweon MN, Hiroi T, Takahashi I. The mucosal immune system: From specialized immune defense to inflam-mation and allergy. Acta Odontol Scand 2001;59:145–153.

25. McKay DM, Baird AW. Cytokine regulation of epithelial perme-ability and ion transport. Gut 1999;44:283–289.

26. Li YP, Lecker SH, Chen Y, Waddell ID, Goldberg AL, Reid MB. TNF-a increases ubiquitin-conjugating activity in skeletal muscle by up-regulating UbcH2/E220k. FASEB J 2003;17:1048–1057.

27. Saini A, Al-Shanti N, Stewart CEH. Waste management — Cytokines, growth factors and cachexia. Cytokine Growth Factor Rev 2006;17:475–486. Epub 2006 Nov 22. Erratum in: Cytokine Growth Factor Rev 2007;18:345.

28. Calleja-Agius J, Muttukrishna S, Jauniaux E. Role of TNF-a in human female reproduction. Exp Rev Endocrinol Metab 2009; 4:273–282.

29. Jana B, Koszykowska M, Andronowska A. The influence of interleukin (IL)-1b and IL-6 and tumour necrosis factor-a on prostaglandin secretion from porcine myometrium during the first third of pregnancy. Acta Vet Brno 2010;79:559–569.


Article


I n t r o d u c t i o nCanine hemangiosarcoma (HSA) is a progressive, highly meta-

static malignant neoplasm that affects dogs. The spleen, liver, heart, and lung are the most common primary or metastatic sites (1). The 1-year survival rate is less than 10%; the median survival time was 19 to 86 d in a group treated by surgery alone and 179 d in a group treated with a combination of surgery and chemotherapy (2,3). One study demonstrated an increase in median survival time to 273 d with the addition of immunotherapy to standard chemotherapy (4). However, effective chemotherapy to prolong survival time in canine HSA is still required.

Receptor tyrosine kinases (RTKs) are often activated aberrantly in a range of human cancers, such as HSA and non-small-cell lung

cancer (5,6). The downstream RTK pathways involving phosphati-dylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (m-TOR) and mitogen-activated protein kinase (MAPK) are consid-ered to represent the main oncogenic signaling pathways in human hematologic malignant disease (7). Imatinib and dasatinib, which are inhibitors of the RTKs, c-kit, and platelet-derived growth factor receptor 1(PDGFR-1), reduced the viability of canine subcutane-ous and renal HSA cell lines (8). The growth of primary murine endothelial cell lines and canine visceral, cutaneous, and cardiac HSA cell lines was also inhibited by a PI3K inhibitor, LY294002 (9). An inhibitor of MAPK/extracellular signal-regulated kinase (ERK) (MEK), PD325901, significantly decreased tumor growth in canine cutaneous and cardiac HSA xenografts (9). Together, these previous studies showed that the inhibitors of RTKs, the PI3K/Akt/m-TOR

Effects of inhibitors of vascular endothelial growth factor receptor 2 and downstream pathways of receptor tyrosine kinases involving

phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin or mitogen-activated protein kinase in canine hemangiosarcoma cell lines

Mami Adachi, Yuki Hoshino, Yusuke Izumi, Hiroki Sakai, Satoshi Takagi

A b s t r a c tCanine hemangiosarcoma (HSA) is a progressive malignant neoplasm with no current effective treatment. Previous studies showed that receptor tyrosine kinases and molecules within their downstream pathways involving phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (m-TOR) or mitogen-activated protein kinase (MAPK) were overexpressed in canine, human, and murine tumors, including HSA. The present study investigated the effects of inhibitors of these pathways in canine splenic and hepatic HSA cell lines using assays of cell viability and apoptosis. Inhibitors of the MAPK pathway did not affect canine HSA cell viability. However, cell viability was significantly reduced by exposure to inhibitors of vascular endothelial growth factor receptor 2 and the PI3K/Akt/m-TOR pathway; these inhibitors also induced apoptosis in these cell lines. These results suggest that these inhibitors reduce the proliferation of canine HSA cells by inducing apoptosis. Further study of these inhibitors, using xenograft mouse models of canine HSA, are warranted to explore their potential for clinical application.

R é s u m éL’hémangiosarcome canin (HS) est un néoplasme malin progressif sans traitement efficace actuel. Des études antérieures ont montré que les récepteurs à activité tyrosine kinase et les molécules dans la voie en aval impliquant la phospatidylinositol 3-kinase (PI3K)/Akt/cible mammalienne de rapamycine (m-TOR) ou la protéine kinase activée par mitogène (PKAM) étaient surexprimées dans les tumeurs canine, humaine, et murine, incluant HS. La présente étude visait à examiner les effets d’inhibiteurs de ces voies dans des lignées cellulaires spléniques et hépatiques de HS en utilisant des essais de viabilité cellulaire et d’apoptose. Les inhibiteurs de la voie PKAM n’ont pas affecté la viabilité de cellules d’HS canines. Toutefois, la viabilité cellulaire était réduite de manière significative suite à l’exposition à des inhibiteurs des récepteurs 2 du facteur de croissance de l’endothélium vasculaire et de la voie PI3K/Akt/m-TOR; ces inhibiteurs ont également induit l’apoptose dans ces lignées cellulaires. Ces résultats suggèrent que ces inhibiteurs réduisent la prolifération de cellules HS canines en induisant l’apoptose. Des études additionnelles de ces inhibiteurs, à l’aide de modèles murins de xénogreffes de HS canins, sont requises afin d’explorer leur potentiel pour une application clinique.


Laboratory of Advanced Veterinary Medicine, Department of Veterinary Clinical Sciences, Hokkaido University, Hokkaido, Japan (Adachi, Izumi, Takagi); Veterinary Teaching Hospital, Hokkaido University, Nishi 9-chome, Kita 18-zyo, Kita-ku, Sapporo-shi, Hokkaido, 060-0818, Japan (Hoshino, Takagi); Laboratory of Veterinary Pathology, Department of Applied Veterinary Medicine, Gifu University, Gifu, Japan (Sakai).

Address all correspondence to Dr. Satoshi Takagi; telephone and fax: 181 11 706 5100; e-mail: [email protected]



pathway, and MEK were effective in human HSA cell lines and in canine HSA. However, the effects of inhibitors of all these pathways have not been reported for canine HSA.

In canine HSA, previous immunohistochemical studies found that these tumors of the spleen expressed vascular endothelial growth factor (VEGF) A and VEGF receptor 2 (VEGFR-2) (10). Although VEGFR-2 was expressed in most of the HSA cell lines, cell proliferation was not stimulated by human VEGF. A previ-ous study of a canine HSA cell line also showed that prolifera-tion was not stimulated by VEGF and similar growth factors (11). Because canine HSA cell lines express both VEGF and VEGFR-2, their lack of response to VEGF may reflect receptor saturation by VEGF in an autocrine or paracrine manner. Previous studies have indicated that canine HSA cells secrete high levels of VEGF and that autocrine or paracrine secretion of this growth factor by HSA cells promoted their proliferation. Western blot testing showed that the levels of phosphorylated Akt, m-TOR, and eukaryotic initiation factor 4E (eIF4E)-binding protein 1 (4E-BP1) were higher in splenic, hepatic, and renal HSA cell lines than in normal endo-thelial cells (12). Both eIF4E and 4E-BP1 operate downstream of the Akt/m-TOR pathway. Overexpression of phosphorylated Akt, m-TOR, eIF4E, and 4E-BP1 was observed immunohistochemically in dermal HSA, and eIF4E showed stronger expression in dermal HSA cells than in normal canine endothelial cells (13). A previous study using immunoblotting found that cellular isolates of cardiac HSA showed a predominance of ERK2 over ERK1 phosphoryla-tion (9). Despite these studies, it was unknown whether therapy targeting these RTK pathways would be effective in canine HSA, with the exception of treatment with imatinib, dasatinib, LY294002, and PD325901.

The present study aimed to determine which stage of all RTK pathways, RTKs, the PI3K/Akt/m-TOR pathway, and the MAPK pathway obstructed the proliferation of canine HSA cell lines by means of assays of cell viability and apoptosis to identify potential molecular targets for the treatment of canine HSA.


Cell lines and reagentsTwo canine HSA cell lines, Ud6 from spleen and JuA1 from

liver (14), were used in this study. The cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM; Life Technologies Japan, Tokyo, Japan) with added glucose (1000 mg/L), sodium pyruvate (110 mg/L), 25 mM HEPES [4-(2-hydroxyethyl)-1-pipera-zineethanesulfonic acid], 4 mM l-glutamine, 10% fetal bovine serum (Nichirei Biosciences, Tokyo, Japan), penicillin, and streptomycin (Wako, Osaka, Japan). All of the following inhibitors were provided by LC Laboratories (Woburn, Massachusetts, USA): the VEGFR inhibitor foretinib; the VEGFR, PDGFR, and fibroblast growth factor receptor inhibitor intedanib; the VEGFR, PDGFR, and c-kit inhibitors pazopanib and vatalanib; the PDGFR and c-kit inhibitor tandutinib; the PI3K inhibitors GDC-0941, LY294002, and wortman-nin; the Akt inhibitor enzastaurin; the mTOR inhibitor everolimus; the MEK inhibitors PD184352, PD325901, and trametinib; and the MAPK inhibitor U0126.

Assay of cell viabilityFor a screening assay, HSA cells (5 3 103 cells per well) were

seeded in 96-well plates with 200 mL of DMEM and 10% fetal bovine serum and incubated overnight at 37°C in 5% CO2. Each of the 14 tested inhibitors was dissolved in 0.1% dimethyl sulfoxide (DMSO) and added to the wells to achieve final concentrations of 1 nM, 100 nM, and 10 mM. For all assays, control cells were exposed to 0.1% DMSO per well. The cells were incubated with these compounds for 24 h. Cell viability was then assessed by means of the water-soluble tetrazolium salts-8 (WST-8) assay, done in accordance with the manufacturer’s specifications (Premix WST-8 Cell Proliferation Assay System; Dojindo, Kumamoto, Japan). Cell counting kit-8 solution (10 mL) was added to each well and the plate incubated for 1 h. Absorbance was measured at a wavelength of 450 nm with a microplate reader (Multiskan FC; Thermo Scientific, Kanagawa, Japan), and cell viability was calculated as a percentage of the control cell viability (mean absorbance of treated wells/mean absorbance of control wells) 3 100. Each inhibitor concentration was analyzed in triplicate, and the experiment was repeated 3 times with each cell line.

The inhibitors that produced a significant difference in cell viability were selected for further investigation. To assess the time dependence of the effects of these inhibitors, cells were seeded in 96-well plates, incubated overnight, and then incubated with 10 mM of each inhibitor for 6, 12, 24, 48, or 72 h. Cell viability was then determined as described above. Control cells were incubated for the same periods. Each inhibitor was analyzed in triplicate, and the experiment was repeated 3 times with each cell line.

Detection of apoptosis by annexin V-biotin assayFor this assay HSA cells (1 3 105 cells per well) were seeded in

6-well plates with 500 mL of DMEM and 10% fetal bovine serum and incubated overnight at 37°C in 5% CO2. Foretinib, intedanib, pazopanib, GDC-0941, enzastaurin, or everolimus was added to the wells to achieve final concentrations of 100 nM, 1 mM, or 10 mM, and the plates were incubated for an additional 12 h. Control wells without inhibitors were also incubated for 12 h. Annexin V-biotin and propidium iodide (5 mL each) were added to each well and the plates incubated at room temperature for 5 min in the dark. Subsequently, 5 mg/mL of fluorescein isothiocyanate (FITC)–streptavidin (Southern Biotech, Birmingham, Alabama, USA) was added to each well, and the plates were incubated for 15 min. The cells were then collected by centrifugation (375 3 g for 5 min at room temperature) and analyzed by flow cytometry (BD FACSVerse; BD Biosciences, San Jose, California, USA) at an excita-tion wavelength of 488 nm, with use of an FITC signal detector. Cells were also placed on 96-well plates and observed by fluorescence microscopy.

Statistical analysisData were expressed as the mean and standard deviation and ana-

lyzed with a test of normality or analysis of variance. In the case of normal distribution and equal variance, Student’s t-test was used to compare the viability of treated cells with that of control cells. When normal distribution or equal variance (or both) was not proven,



the Wilcoxon test was used. All statistical analyses were done with JMP pro (SAS Institute Japan, Tokyo, Japan). P-values , 0.05 were considered statistically significant.

Re s u l t sAmong the RTK inhibitors (Figure 1) the screening assay showed

significantly reduced viability of both canine HSA cell lines (Ud6 and JuA1) after 24 h of incubation with 10 mM foretinib or 10 mM intedanib compared with control exposure. All tested concentra-tions of intedanib and pazopanib produced significant reductions in Ud6 cell viability compared with control exposure; the JuA1 cell line was less sensitive to these compounds. No consistent significant differences in cell viability were observed when cells were incubated with vatalanib or tandutinib.

Among the PI3K/Akt/mTOR pathway inhibitors (Figure 2) the screening assay showed significantly reduced viability of both canine HSA cell lines after 24 h of incubation with GDC-0941 (100 nM and

10 mM), enzastaurin (100 nM and 10 mM), and everolimus (1 nM and 100 nM) compared with control exposure; the effect of 10 mM everolimus was not significant in the Ud6 cell line. No consistent significant differences in cell viability were observed when cells were incubated with LY294002 or wortmannin.

No consistent significant differences in cell viability were observed in the screening assay when the 2 canine HSA cell lines were incu-bated with the MAPK pathway inhibitors for 24 h compared with control exposure (Figure 3).

The time dependence of these effects was investigated for up to 72 h with the use of 10 mM of each inhibitor. Among the RTK inhibi-tors (Figure 4) the strongest effects were observed after incubation of both cell lines with foretinib for 72 h, incubation of Ud6 cells with intedanib for 24 h, incubation of JuA1 cells with intedanib for 72 h, and incubation of Ud6 cells with pazopanib for 24 h. However, pazo-panib showed no significant effects in the JuA1 cell line. Among the PI3K/Akt/mTOR pathway inhibitors (Figure 5) the strongest effects were observed after incubation of Ud6 cells with GDC-0941 for 12 h,

Figure 1. Mean cell viability 6 standard deviation (SD), determined by the water-soluble tetrazolium salts-8 (WST-8) assay, for Ud6 and JuA1 cell lines of canine hemangiosarcoma (HSA) incubated with the indicated inhibitors of receptor tyrosine kinases (RTKs) at various concentrations for 24 h. Viability was expressed as a percentage of the viability of control cells exposed to 0.1% dimethyl sulfoxide (DMSO). Asterisks indicate a significant difference for the viability of control cells (100%) at a P-value , 0.05 in Student’s t-test or the Wilcoxon test.

140

120

100

80

60

40

20

0

1 nM-Ud6 100 nM-Ud6 10 mM-Ud6 1 nM-JuA1 100 nM-JuA1 10 mM-JuA1

Cel

l via

bilit

y (%

)

RTK inhibitors

Foretinib Intedanib Pazopanib Vatalanib Tandutinib

Figure 2. Mean cell viability 6 SD, determined and expressed in the same way, for the 2 cell lines incu-bated with the indicated inhibitors of downstream RTK pathways involving phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (m-TOR) at various concentrations for 24 h. Asterisks indicate a significant difference for the viability of control cells (100%) at a P-value , 0.05 in Student’s t-test or the Wilcoxon test.

140

120

100

80

60

40

20

0


Cel

l via

bilit

y (%

)

PI3K/Akt/m-TOR inhibitors

GDC-0941 LY294002 Wortmannin Enzastaurin Everolimus



incubation of JuA1 cells with GDC-0941 for 48 h, and incubation of Ud6 cells with everolimus for 24 h. However, no significant effects were observed in either cell line incubated with enzastaurin or in JuA1 cells incubated with everolimus.

Increases in annexin V-biotin signals, indicating apoptosis, were detected by flow cytometry after incubation of the Ud6 and JuA1 cell lines with 100 nM, 1 mM, or 10 mM foretinib, intedanib, GDC-0941, enzastaurin, or everolimus and 100 nM or 1 mM pazopanib (Figures 6

Figure 4. Mean cell viability 6 SD, determined and expressed in the same way, for the 2 cell lines incubated with 10 mM of the indicated RTK inhibitors for the indicated times. Asterisks indicate a significant differ-ence for the viability of control cells (100%) at a P-value , 0.05 in Student’s t-test or the Wilcoxon test.

140

120

100

80

60

40

20

0

6 h-Ud6 12 h-Ud6 24 h-Ud6 48 h-Ud6 72 h-Ud6

6 h-JuA1 12 h-JuA1 24 h-JuA1 48 h-JuA1 72 h-JuA1

Cel

l via

bilit

y (%

)

RTK inhibitors, ~10 mM

Foretinib Intedanib Pazopanib

Figure 3. Mean cell viability 6 SD, determined and expressed in the same way, for the 2 cell lines incu-bated with the indicated inhibitors of downstream RTK pathways involving mitogen-activated protein kinase (MAPK) at various concentrations for 24 h.

140

120

100

80

60

40

20

0


Cel

l via

bilit

y (%

)

MAPK pathway inhibitors

PD184352 PD325901 Trametinib U0126

Figure 5. Mean cell viability 6 SD, determined and expressed in the same way, for the 2 cell lines incubated with 10 mM of the indicated PI3K/Akt/m-TOR inhibitors for the indicated times. Asterisks indicate a significant difference for the viability of control cells (100%) at a P-value , 0.05 in Student’s t-test or the Wilcoxon test.

140

120

100

80

60

40

20

0

6 h-Ud6 12 h-Ud6 24 h-Ud6 48 h-Ud6 72 h-Ud6

6 h-JuA1 12 h-JuA1 24 h-JuA1 48 h-JuA1 72 h-JuA1

Cel

l via

bilit

y (%

)

PI3K/Akt/m-TOR inhibitors, 10 mM

GDC-0941 Enzastaurin Everolimus



and 7). In addition, fluorescence microscopy identified annexin V-biotin activity in both canine HSA cell lines exposed to these inhibitors at concentrations of 100 nM, 1 mM, or 10 mM; Figure 8 shows these findings for enzastaurin.

D i s c u s s i o nA previous study found that 100 nM and 1 mM were noncytotoxic

concentrations of foretinib with exposure of 24 or 48 h in human pancreatic tumor cell lines (15). Pazopanib was effective at a concen-tration of 70 mM for the clinical control or reduction of human cuta-neous and scalp HSA without side effects (16–18). Pharmacokinetic studies found that 1 mM intedanib inhibited tumor xenograft growth in mouse and rat models of human head and neck small-cell car-cinoma and renal cancer without weight loss (19). In the present study 1 nM, 100 nM, and 10 mM of these VEGFR-2 inhibitors, with the exception of foretinib, caused apoptosis and reduced growth of canine splenic HSA without adverse effects.

A previous study showed that VEGFR-2 inhibition decreased the burden of murine renal cell and colon carcinoma by preventing vascularization and growth of micrometastases in vivo rather than by preventing the establishment of micrometastases (20). Among HSA cell lines, Ud2, derived from the same source tissue as Ud6, showed the most rapid growth, and JuB2 cells, derived from the same source tissue as JuA1 cells, showed the slowest growth (12). The results herein suggest that Ud6 canine splenic HSA cells may

show more vascularization and metastasis than JuA1 cells and that VEGFR-2-targeted therapy may be effective for canine splenic HSA. In the time-dependence test the strongest effects were produced by foretinib in both cell lines and by intedanib in the JuA1 cells after 72 h; the results suggest that growth of JuA1 cells may be inhibited after incubation for more than 24 h.

No consistent significant differences in cell viability were observed in the present study when cells were incubated with up to 10 mM vatalanib or tandutinib. These drugs inhibit PDGFR-1 and c-kit. The percentage of cells expressing c-kit increased in sphere cells from canine subcutaneous HSA cell lines; flow cytometry found that one-third of these cell lines showed no c-kit expression (21) and that only 15% (5 of 34) specimens of human HSA were positive for c-kit (22). In canine subcutaneous and renal HSA, imatinib and dasatinib inhibited c-kit, PDGFR-1, and the non-RTK Src; these inhibitors also blocked non-RTK-mediated Src phosphorylation (8). This previous study suggested that these inhibitors reduced the proliferation of canine HSA by controlling Src rather than by controlling c-kit or PDGFR-1. In the present study, at concentrations of 1 nM, 100 nM, and 10 mM, the c-kit and PDGFR-1 inhibitors did not inhibit prolif-eration of canine HSA cells. A previous study demonstrated that the 50% lethal concentration of vatalanib in human chronic lymphocytic leukemia cells was 48.4 mM (23). Vatalanib inhibited VEGFR-2 at concentrations of 39 nM and 30 nM. At higher concentrations other tyrosine kinases, such as PDGFR-2 and c-kit, were also inhibited (24,25). The present findings show that a higher dose of vatalanib

Figure 6. Annexin V-biotin signals as detected by flow cytometry after incubation of the 2 cell lines with foretinib or intedanib (100 nM, 1 mM, or 10 mM) or pazopanib (100 nM or 1 mM) for 12 h. Black lines indicate untreated (control) cells and grey shaded area the cells treated with inhibitors.

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

Foretinib Ud6 12 h 100 nM Foretinib Ud6 12 h 1 mM Foretinib Ud6 12 h 10 mM

1.0

0.5

0.00 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

Intedanib Ud6 12 h 100 nM Intedanib Ud6 12 h 1 mM Intedanib Ud6 12 h 10 mM

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

Pazopanib JuA1 12 h 100 nM Pazopanib JuA1 12 h 1 mM



may be necessary to block the growth of canine HSA. Alternatively, PDGFR-2 and c-kit may be inhibited more strongly than VEGFR-2 in canine splenic or hepatic HSA cell lines exposed to 100 nM or 10 mM vatalanib. In addition, the present findings suggest that growth inhi-bition may be observed during longer incubation periods, similar to the findings with foretinib. The present study was limited to the examination of 2 canine HSA cell lines and had restricted concen-tration ranges and exposure periods. Consistent with the vatalanib findings, a previous study demonstrated that the use of toceranib after doxorubicin chemotherapy did not improve either the disease-free interval or the overall survival in dogs with splenic HSA (26). These dogs were considered free of metastatic disease, although they may have had micrometastases. Since VEGFR-2 inhibition prevented vascularization and growth of micrometastases in vivo (20), delaying this treatment until after doxorubicin therapy rather than administering the agents concurrently may have diminished the therapeutic effect of VEGFR-2 inhibition.

In vitro, previous studies have demonstrated that the PI3K inhibi-tor GDC-0941 at a concentration of 0.35 to 1.2 mM and the m-TOR inhibitor everolimus at concentrations of 10 and 100 nM strongly inhibited human malignant T-cell proliferation, with minimal cyto-toxic effects (27). Pharmacokinetic studies revealed that the growth of human colon carcinoma and glioblastoma xenografts in mice was significantly suppressed by treatment with the Akt inhibitor enza-staurin at a concentration of 3 mM (28). The present findings suggest that these PI3K/Akt/m-TOR pathway inhibitors at concentrations of 1 and 100 nM could induce apoptosis and reduce the growth of

canine HSA cells without causing adverse effects. In canine splenic HSA cells (Ud2 and Ud6), both the MAPK and PI3K/Akt/m-TOR pathways were constitutively phosphorylated. The RTKs are well-known activators of the MAPK and PI3K/Akt/m-TOR pathways, and mutations of RTKs in cancer lead to constitutive activation of these pathways (29,30). Western blot testing showed that the levels of phosphorylated m-TOR were higher in canine splenic and hepatic

Figure 7. Annexin V-biotin signals as detected by flow cytometry after incubation of the 2 cell lines with GDC-0941, enzastaurin, or everolimus (100 nM, 1 mM, or 10 mM) for 12 h. Black lines indicate untreated (control) cells and grey shaded area the cells treated with inhibitors.

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 0 102 103 104 105

Nor

mal

ized

%

FITC-A

1.0

0.5

0.0 2104 0 104 105

Nor

mal

ized

%

FITC-A

GDC JuA1 12 h 100 nM GDC JuA1 12 h 1 mM GDC JuA1 12 h 10 mM

Enzastaurin Ud6 12 h 100 nM Enzastaurin Ud6 12 h 1 mM Enzastaurin Ud6 12 h 10 mM

Everolimus Ud6 12 h 100 nM Everolimus Ud6 12 h 1 mM Everolimus Ud6 12 h 10 mM

Figure 8. Annexin V-biotin signals as visualized by fluorescence micro-scopy in untreated (control) cells (A) and JuA1 cells incubated with 10 mM enzastaurin for 24 h (B) or 48 h (C). Original magnification 3400.



HSA cell lines than in normal endothelial cells (12). The PI3K/Akt pathway is downstream of VEGFR-2, whereas m-TOR is downstream of Akt. Therefore, the results of the present study suggest that VEGFR-2 and PI3K inhibitors may provide effective treatment of Ud6 canine splenic HSA, whereas m-TOR inhibitors may be effective against JuA1 hepatic HSA.

Canine visceral, cutaneous, and cardiac HSA cell lines are suscep-tible to growth inhibition by 8.8 to 29.7 mM LY294002 (9). In the pres-ent study GDC-0941 produced the strongest effects in JuA1 cells after 48 h. Similarly, growth inhibition may be observed in cells exposed to LY294002 and wortmannin for periods longer than 24 h. The present findings suggest that LY294002 and wortmannin may not be effective in canine splenic or hepatic HSA or that the inhibitor concentrations (up to 10 mM) or incubation times (24 h) may have been inadequate. In addition, the present findings suggest that there may be differ-ences between canine HSA cell lines derived from different tissues. This study was limited because it focused on 2 such cell lines and used limited inhibitor concentrations and exposure periods.

In previous studies the PI3K/Akt/m-TOR and MAPK pathways were phosphorylated in Ud2 and Ud6 canine splenic HSA cells (29,30). Plasma PD325901 at concentrations of 100 and 173 mM was administered in 15-mg doses twice daily, without side effects, to patients with advanced non-small-cell lung cancer (31). The results of the present study suggest that the MAPK pathway alone may not control canine splenic or hepatic HSA proliferation or that the inhibitor concentrations (up to 10 mM) were inadequate.

A human pancreatic cancer cell line exhibited resistance to the immunotoxin SS1(dsFv)-PE38 (SS1P) when used alone at a half- maximal inhibitory concentration (IC50) of more than 100 mM; how-ever, when SS1P was used in combination with enzastaurin the IC50 was less than 10 mM (32). Another study reported on the same syn-ergistic effects of everolimus and an insulin receptor substrate-1/2 inhibitor, NT157 (33). Thus, use of an RTK pathway inhibitor such as vatalanib, LY294002, or wortmannin with other drugs may increase their efficacy, reduce the concentration required to treat canine HSA, and prevent adverse effects caused by high inhibi-tor doses. In addition, the safety and efficacy of using pegylated liposome- encapsulated doxorubicin in adjuvant monotherapy was demonstrated in dogs with splenic HSA after splenectomy, as compared with free doxorubicin (34). The development of drug delivery systems such as liposomes is an important breakthrough in improving tumor targeting, producing sustained release, and reducing the toxicity of chemotherapeutic drugs (35). The use of liposome-encapsulated RTK pathway inhibitors might reduce the amount of inhibitor required to produce therapeutic effects and prevent adverse effects caused by high doses of these inhibitors.

In conclusion, the present study showed that inhibitors of VEGFR-2 and PI3K/Akt/m-TOR pathways reduced the prolifera-tion of canine HSA cells by inducing apoptosis. These inhibitors warrant further study using xenograft mouse models of canine HSA to explore their potential for clinical application.

A c k n o w l e d g m e n tThe authors wish to acknowledge Dr. Satoru. Konnai, associ-

ate professor of the Laboratory of Infectious Disease, Hokkaido

University, Sapporo, Japan, for providing experimental equipment and technical advice on this study.

Re f e r e n c e s1. Brown NO, Patnaik AK, MacEwen EG. Canine hemangiosar-

coma: Retrospective analysis of 104 cases. J Am Vet Med Assoc 1985;186:56–58.

2. Sorenmo KU, Jeglum KA, Helfand SC. Chemotherapy of canine splenic hemangiosarcoma with doxorubicin and cyclophospha-mide. J Vet Intern Med 1993;7:370–376.

3. Wood CA, Moore AS, Gliatto JM, Ablin LA, Berg RJ, Rand WM. Prognosis for dogs with stage I or II splenic hemangiosarcoma treated by splenectomy alone: 32 cases (1991–1993). J Am Anim Hosp Assoc 1998;34:417–421.

4. Vail DM, MacEwen EG, Kurzman ID, et al. Liposome-encapsulated muramyl tripeptide phosphatidylethanolamine adjuvant immuno-therapy for splenic hemangiosarcoma in the dog: A random-ized multi-institutional clinical trial. Clin Cancer Res 1995;1: 1165–1170.

5. Li S, Chen S, Jiang Y, Liu J, Yang X, Quan S. Synergistic interac-tion between MEK inhibitor and gefitinib in EGFR-TKI-resistant human lung cancer cells. Oncol Lett 2015;10:2652–2656.

6. Liu L, Kakiuchi-Kiyota S, Arnold LL, Johansson SL, Wert D, Cohen SM. Pathogenesis of human hemangiosarcomas and hemangiomas. Hum Pathol 2013;44:2302–2311.

7. Matsumura I, Mizuki M, Kanakura Y. Roles for deregulated receptor tyrosine kinases and their downstream signaling mol-ecules in hematologic malignancies. Cancer Sci 2008;99:479–485.

8. Dickerson EB, Marley K, Edris W, et al. Imatinib and dasatinib inhibit hemangiosarcoma and implicate PDGFR-b and Src in tumor growth. Transl Oncol 2013;6:158–168.

9. Andersen NJ, Nickoloff BJ, Dykema KJ, et al. Pharmacologic inhibition of MEK signaling prevents growth of canine heman-giosarcoma. Mol Cancer Ther 2013;12:1701–1714.

10. Göritz M, Muller K, Krastel D, et al. Canine splenic haemangio-sarcoma: Influence of metastases, chemotherapy and growth pat-tern on post-splenectomy survival and expression of angiogenic factors. J Comp Pathol 2013;149:30–39.

11. Thamm DH, Dickerson EB, Akhtar N, et al. Biological and molecular characterization of a canine hemangiosarcoma-derived cell line. Res Vet Sci 2006;81:76–86. Epub 2005 Oct 25.

12. Murai A, Asa SA, Kodama A, Hirata A, Yanai T, Sakai H. Constitutive phosphorylation of the mTORC2/Akt/4E-BP1 pathway in newly derived canine hemangiosarcoma cell lines. BMC Vet Res 2012;8:128–141.

13. Murai A, Abou Asa S, Kodama A, Sakai H, Hirata A, Yanai T. Immunohistochemical analysis of the Akt/mTOR/4E-BP1 signal-ling pathway in canine haemangiomas and haemangiosarcomas. J Comp Pathol 2012;147:430–440.

14. Kodama A, Sakai H, Matsuura S, et al. Establishment of canine hemangiosarcoma xenograft models expressing endothelial growth factors, their receptors, and angiogenesis-associated homeobox genes. BMC Cancer 2009;9:363–374.

15. Chen HM, Tsai CH, Hung WC. Foretinib inhibits angiogen-esis, lymphangiogenesis and tumor growth of pancreatic



cancer in vivo by decreasing VEGFR-2/3 and TIE-2 signaling. Oncotarget 2015;6:14940–14952.

16. Fujita M, Endo Y, Fujisawa A, Tanioka M, Kabashima K, Miyachi Y. Pazopanib: An alternative in taxane-resistant cuta-neous angiosarcoma. Eur J Dermatol 2014;24:267–268.

17. Tomita H, Koike Y, Asai M, et al. Angiosarcoma of the scalp suc-cessfully treated with pazopanib. J Am Acad Dermatol 2014;70: 19–21.

18. Tan AR, Dowlati A, Stein MN, et al. Phase I study of weekly paclitaxel in combination with pazopanib and lapatinib in advanced solid malignancies. Br J Cancer 2014;110:2647–2654.

19. Hilberg F, Roth GJ, Krssak M, et al. BIBF 1120: Triple angiokinase inhibitor with sustained receptor blockade and good antitumor efficacy. Cancer Res 2008;68:4774–4782.

20. Lee YJ, Karl DL, Maduekwe UN, et al. Differential effects of VEGFR-1 and VEGFR-2 inhibition on tumor metastases based on host organ environment. Cancer Res 2010;70:8357–8367.

21. Gorden BH, Kim JH, Sarver AL, et al. Identification of three molecular and functional subtypes in canine hemangiosarcoma through gene expression profiling and progenitor cell charac-terization. Am J Pathol 2014;184:985–995.

22. Yonemori K, Tsuta K, Ando M, et al. Contrasting prognostic implications of platelet-derived growth factor receptor-b and vascular endothelial growth factor receptor-2 in patients with angiosarcoma. Ann Surg Oncol 2011;18:2841–2850.

23. Paesler J, Gehrke I, Gandhirajan RK, et al. The vascular endo-thelial growth factor receptor tyrosine kinase inhibitors vata-lanib and pazopanib potently induce apoptosis in chronic lymphocytic leukemia cells in vitro and in vivo. Clin Cancer Res 2010;16:3390–3398.

24. Mross K. Inhibition of the VEGF receptor system with tyrosine kinase inhibitors. Angiogenesis inhibition in oncology. Pharm Unserer Zeit 2008;37:404–415.

25. Lin B, Podar K, Gupta D, et al. The vascular endothelial growth factor receptor tyrosine kinase inhibitor PTK787/ZK222584 inhibits growth and migration of multiple myeloma cells in the bone marrow microenvironment. Cancer Res 2002;62: 5019–5026.

26. Gardner HL, London CA, Portela RA, et al. Maintenance therapy with toceranib following doxorubicin-based chemotherapy for canine splenic hemangiosarcoma. BMC Vet Res 2015;11:131–139.

27. Witzig TE, Reeder C, Han JJ, et al. The mTORC1 inhibitor everolimus has antitumor activity in vitro and produces tumor responses in patients with relapsed T-cell lymphoma. Blood 2015;126:328–335.

28. Graff JR, McNulty AM, Hanna KR, et al. The protein kinase Cbeta-selective inhibitor, Enzastaurin (LY317615.HCl), sup-presses signaling through the AKT pathway, induces apoptosis, and suppresses growth of human colon cancer and glioblastoma xenografts. Cancer Res 2005;65:7462–7469.

29. Azzariti A, Porcelli L, Mangia A, et al. Irradiation-induced angio-sarcoma and anti-angiogenic therapy: A therapeutic hope? Exp Cell Res 2014;321:240–247.

30. Chappell WH, Steelman LS, Long JM, et al. Ras/Raf/MEK/ERK and PI3K/PTEN/Akt/mTOR inhibitors: Rationale and impor-tance to inhibiting these pathways in human health. Oncotarget 2011;2:135–164.

31. Haura EB, Ricart AD, Larson TG, et al. A phase II study of PD-0325901, an oral MEK inhibitor, in previously treated patients with advanced non-small cell lung cancer. Clin Cancer Res 2010;16:2450–2457.

32. Mattoo AR, Pastan I, Fitzgerald D. Combination treatments with the PKC inhibitor, enzastaurin, enhance the cytotoxicity of the anti-mesothelin immunotoxin, SS1P. PLoS One 2013;8:e75576.

33. Garofalo C, Capristo M, Mancarella C, Reunevi H, Picci P, Scotlandi K. Preclinical effectiveness of selective inhibitor of IRS-1/2 NT157 in osteosarcoma cell lines. Front Endocrinol 2015;6:74–82.

34. Teske E, Rutteman GR, Kirpenstein J, Hirschberger J. A random-ized controlled study into the efficacy and toxicity of pegylated liposome encapsulated doxorubicin as an adjuvant therapy in dogs with splenic haemangiosarcoma. Vet Comp Oncol 2011;9: 283–289.

35. Gabizon A, Shmeeda H, Barenholz Y. Pharmacokinetics of pegylated liposomal doxorubicin: Review of animal and human studies. Clin Pharmacokinet 2003;42:419–436.


Article


I n t r o d u c t i o nIn humans, obesity and Type-II diabetes are associated with a

pro-inflammatory state (1–3). Obesity is caused primarily by accu-mulation of white adipose tissue (WAT), an important endocrine organ that secretes proteins known as adipokines, which regulate metabolism, coagulation, and inflammation (4). Key inflammatory

adipokines secreted by WAT include the cytokines interleukin-6 (IL-6) and tumor necrosis factor alpha (TNFa) and the acute phase reactants, serum amyloid A (SAA) and C-reactive protein (CRP) (5–8). Circulating inflammatory cytokines, primarily TNFa, perpetu-ate the inflammatory state by activating intracellular stress kinases in tissues (9), with subsequent transcription of inflammatory cytokines and perpetuation of the pro-inflammatory state. Additional key

Relationship of skeletal muscle inflammation with obesity and obesity-associated hyperinsulinemia in horsesHeidi E. Banse, Todd C. Holbrook, Nicholas Frank, Dianne McFarlane

A b s t r a c tLocal (skeletal muscle and adipose) and systemic inflammation are implicated in the development of obesity-associated insulin resistance in humans. In horses, obesity is neither strongly nor consistently associated with systemic inflammation. The role of skeletal muscle inflammation in the development of insulin dysregulation (insulin resistance or hyperinsulinemia) remains to be determined. We hypothesized that skeletal muscle inflammation is related to obesity-associated hyperinsulinemia in horses. Thirty-five light-breed horses with body condition scores (BCSs) of 3/9 to 9/9 were studied, including 7 obese, normoinsulinemic (BCS $ 7, resting serum insulin , 30 mIU/mL) and 6 obese, hyperinsulinemic (resting serum insulin $ 30 mIU/mL) horses. Inflammatory biomarkers were evaluated in skeletal muscle biopsies and plasma. Relationships between markers of inflammation and BCS were evaluated. To assess the role of inflammation in obesity-associated hyperinsulinemia, markers of inflammation were compared among lean or ideal, normoinsulinemic (L-NI); obese, normoinsulinemic (O-NI); and obese, hyperinsulinemic (O-HI) horses. Skeletal muscle and plasma tumor necrosis factor alpha (TNFa) concentrations were negatively correlated with BCS. When comparing inflammatory markers among groups, skeletal muscle TNFa was lower in the O-HI group than in the O-NI or L-NI groups. In horses, neither skeletal muscle nor systemic inflammation appears to be positively related to obesity or obesity-associated hyperinsulinemia.

R é s u m éL’inflammation locale (muscle squelettique et tissu adipeux) et systémique sont impliquées dans le développement de la résistance à l’insuline associée à l’obésité chez l’humain. Chez les chevaux, l’obésité n’est pas fortement ou de manière constante associée avec l’inflammation systémique. Le rôle de l’inflammation des muscles squelettiques dans le développement de la dérégulation de l’insuline (résistance à l’insuline ou hyper-insulinémie) reste à être déterminé. Nous avons émis l’hypothèse que chez les chevaux l’inflammation des muscles squelettiques est reliée à l’hyper-insulinémie associée à l’obésité. Trente-cinq chevaux de race légère avec des pointages de condition corporelle (PCCs) variant de 3/9 à 9/9 ont été étudiés, incluant sept chevaux obèses, normo-insulinémique (PCC $ 7, insuline sérique au repos , 30 mUI/mL) et six chevaux obèses, hyper-insulinémique (insuline sérique au repos $ 30 mUI/mL). Les biomarqueurs de l’inflammation ont été évalués dans des biopsies de muscles squelettiques et le plasma. Les relations entre les marqueurs de l’inflammation et le PCC ont été évaluées. Pour apprécier le rôle de l’inflammation dans l’hyper-insulinémie associée à l’obésité, les marqueurs de l’inflammation ont été comparés parmi les chevaux élancés ou idéal, normo-insulinémique (L-NI); les chevaux obèses, normo-insulinémique (O-NI); et les chevaux obèses, hyper-insulinémique (O-HI). Les concentrations du facteur nécrosant des tumeurs alpha (TNFa) étaient corrélées négativement avec le PCC. Lors de la comparaison des marqueurs de l’inflammation entre les groupes, la concentration de TNFa dans les muscles squelettiques était plus basse dans le groupe O-HI que dans les groupes O-NI ou L-NI. Chez les chevaux, ni l’inflammation systémique ou celle des muscles squelettiques ne semblent reliées positivement à l’obésité ou à l’hyper-insulinémie associée à l’obésité.


Department of Physiological Sciences (Banse, McFarlane) and Department of Veterinary Clinical Sciences (Holbrook), Center for Veterinary Health Sciences, Oklahoma State University, Stillwater, Oklahoma 74078, USA; Department of Clinical Sciences, Cummings School of Veterinary Medicine, Tufts University, North Grafton, Massachusetts 01536, USA (Frank).

Address all correspondence to Dr. Heidi Banse; telephone (403) 210-6494; fax: (403) 220-3929; e-mail: [email protected]

Dr. Banse’s current address is Department of Veterinary Clinical and Diagnostic Sciences, Faculty of Veterinary Medicine, University of Calgary, 3280 Hospital Drive NW, Calgary, Alberta T2N 4Z6.

Received September 3, 2015. Accepted February 16, 2016.



cytokines that modulate the inflammatory response include IL-6 and IL-10. Interleukin-6 is an important mediator of the hepatic acute phase response and an adipokine and myokine involved in insulin signaling (10,11). Interleukin-10 is primarily an anti-inflammatory cytokine that counters the effect of IL-6 and TNFa.

Adrenocorticotropic hormone (ACTH) has both direct and indi-rect anti-inflammatory actions and has been used to treat several inflammatory conditions, including dermatomyositis, polymyositis, gout, and other forms of crystalline arthritis (12–14). Although his-torically the anti-inflammatory activity of ACTH has been primarily attributed to stimulation of glucocorticoid release (indirect actions), anti-inflammatory activity of ACTH in induced crystalline arthritis is maintained in adrenalectomized rats, which suggests a direct anti-inflammatory effect associated with binding to melanocortin receptors (15). Furthermore, in horses, ACTH is released from both the anterior and intermediate lobe of the pituitary along with other anti-inflammatory peptides, including alpha-Melanocyte-stimulating hormone (a-MSH) and beta-endorphin. As the concentration of ACTH parallels that of other pituitary-derived hormones, it can be used as a surrogate marker for pituitary anti-inflammatory hormone activity (16,17).

In contrast to studies in humans and in mice, research into obesity in horses has not demonstrated a consistent association between systemic inflammation and obesity. Initial studies found that obe-sity in horses was correlated with systemic inflammation (18,19), but these findings were confounded by failure to control for age in the obese population surveyed. In ponies with historical laminitis, circulating TNFa concentrations were not correlated with obesity

(20). Furthermore, there were no differences in systemic markers of inflammation in equids fed to promote obesity with either a high fat or high glucose diet, compared to equids fed to maintain weight (21).

The role of systemic inflammation in equine hyperinsulinemia is similarly unclear. A previous study of hyperinsulinemic obese horses demonstrated a trend toward decreased circulating TNFa and decreased inflammatory cytokine expression when compared with lean controls, but no change in CRP (22). Interleukin-1b, IL-6, and TNFa plasma concentrations were not correlated with obesity or plasma insulin concentrations in another study, although SAA cor-related with insulin concentrations and weakly with BCS (23). Horses with insulin dysregulation have a more prolonged upregulation of circulating inflammatory cytokine gene expression in response to lipopolysaccharide (LPS) infusion than control horses, which sug-gests a pro-inflammatory state (24).

Despite multiple investigations into the relationship between systemic inflammation, obesity, and insulin regulation in horses, knowledge of tissue inflammation is limited. Inflammatory cytokine gene expression in adipose tissue of insulin-resistant horses was not significantly different than that of insulin-sensitive controls (25). The protein content of TNFa was increased in visceral adipose, but not in skeletal muscle or subcutaneous adipose of insulin-resistant horses compared to insulin-sensitive controls (26). Notably, in both of these studies, horses were stratified solely on the basis of dynamic insulin-sensitivity testing and were similar with respect to BCS.

We hypothesized that skeletal muscle inflammation is related to obesity and obesity-associated hyperinsulinemia in horses. To test this hypothesis, relationships between body condition score (as

Table I. Population characteristics by group, comparing lean or ideal, normoinsulinemic (L-NI), overweight, obese, normoinsulinemic (O-NI), and obese, hyperinsulinemic (O-HI) horses

Lean or ideal, NI Overweight Obese, NI Obese, HI (n = 17) (n = 5) (n = 7) (n = 6) P-valuec

Breed QH/Paint (n = 13) QH/Paint (n = 3) QH/Paint (n = 6) Paso fino (n = 2) TB (n = 4) Azteca (n = 1) MFT (n = 1) Morgan (n = 1) Arab (n = 1) Azteca (n = 1) QH (n = 1) TWH (n = 1)

Agea 15 6 7 16 6 5 11 6 5 15 6 4 P = 0.30

Gender Gelding (n = 11) Mare (n = 5) Gelding (n = 4) Gelding (n = 3) Mare (n = 6) Mare (n = 3) Mare (n = 3)

BCSb 5 (4 to 5) 6 8 (7 to 8) 8.5 (8 to 9)

Insulinb (mIU/mL) 6 (5 to 14) 13 (12 to 356) 14 (5 to 8) 195 (72 to 315) P , 0.0001

ACTHb (pg/mL) 35 (28 to 61) 28 (16 to 99) 36 (19 to 43) 37 (19 to 68) P = 0.22

TNFab (pg/mL) 1137 (431 to 2166) 260 (40 to 1440) 471 (40 to 1414) 209 (98 to 874) P = 0.08

SAAb (ng/mL) 35 (30 to 54) 31 (28 to 52) 54 (40 to 1322) 27 (26 to 29) P = 0.91(n = 25)QH — Quarter Horse; TB — Thoroughbred; MFT — Missouri fox trotter; TWH — Tennessee walking horse.a Age is expressed as mean 6 SD (standard deviation).b Body condition score (BCS), insulin, adrenocorticotropic hormone (ACTH), tumor necrosis factor alpha (TNFa), and serum amyloid A (SAA) are expressed as median and interquartile range (IQR).c Reported P-values are for differences among groups.



an indicator of obesity) and markers of local (skeletal muscle) and systemic inflammation were explored, including skeletal muscle gene expression of TNFa, IL-1, and IL-6 and protein content of TNFa, and circulating concentrations of TNFa, SAA, and ACTH, a pituitary hormone with anti-inflammatory activity. Furthermore, markers of inflammation were compared among lean or ideal, nor-moinsulinemic (L-NI); obese, normoinsulinemic (O-NI); and obese, hyperinsulinemic horses (O-HI).


Sample populationAdult horses donated to Oklahoma State University or the

University of Tennessee were included in the study. Body condi-tion score (BCS) was assessed in all animals (N = 35) by experienced observers, as previously described (23). All horses were considered to be free of systemic disease (aside from endocrine or metabolic disease) on the basis of physical examination. Blood samples and skeletal muscle biopsies were collected from all horses. All horses (N = 35, Table I) were included in correlation analysis evaluating relationships between body condition score and inflammation. In order to determine if obesity-associated inflammation has a role in hyperinsulinemia, inflammatory markers in obese animals (BCS $ 7) with hyperinsulinemia (O-HI) were compared to those of obese animals with normoinsulinemia (O-NI). Normoinsulinemic lean or ideal animals (BCS # 5; L-NI) were included as controls (see Table I).

Horses were not fed grain within 12 h of sample collection. Blood samples were collected between 9 am and 12 pm into tubes contain-ing no anticoagulant and into tubes containing ethylenediamine tetra-acetic acid (EDTA) and EDTA tubes were immediately placed on ice. All samples were centrifuged at 1200 3 g for 10 min within 30 min of collection and stored at 280°C until analysis.

Semimembranosus muscle biopsies were collected antemortem (n =15) for horses that were not euthanized or within 15 min after euthanasia with pentobarbital (n = 20) using an open-biopsy tech-nique as previously described (24). Muscle biopsy samples were flash frozen in liquid nitrogen and stored at 280°C until analysis. All samples were analyzed within 4 y of collection. Plasma and tissue samples have been demonstrated to maintain stability of ribonucleic acid (RNA) and protein for at least 5 y when stored at 270°C to 280°C (27,28). All samples were obtained in accordance with the institution’s Animal Care and Use Committee.

Hormone analysisConcentrations of serum insulin (Coat-A-Count; Siemens,

Tarrytown, New York, USA) were measured by radioimmunoassay and plasma ACTH concentrations were analyzed by chemilumi-nescent immunoassay (Immulite 1000; Siemens). All assays were previously validated for use in horses (29,30).

Muscle TNFa proteinMuscle sample homogenates were prepared in phosphate-buffered

saline (PBS) (~50 mg in 1 mL PBS) using a tissue homogenizer (Fisher Scientific, Pittsburg, Pennsylvania, USA). Homogenates were centri-fuged at 1000 3 g for 10 min. Supernatant protein concentration was quantified using a commercially available assay (Bio-Rad, Hercules, California, USA). Skeletal muscle TNFa was evaluated by an equine-specific enzyme-linked immunosorbent assay (ELISA) (Pierce, Rockford, Illinois, USA) that has previously been validated for use in horses (19). To validate the ELISA for use in skeletal muscle, a pooled sample with low TNFa concentration was spiked with a high TNFa standard. The pooled sample was then mixed with the spiked high sample at varying proportions and linearity was evaluated (r2 = 0.95, P = 0.005). Percent recovery was determined by spiking a pooled low homogenate sample with reconstituted standard at concentrations ranging from 62.5 to 1000 pg/mL. Recovery [mean 6 SD (standard deviation)] was 80 6 13%. Samples were analyzed in duplicate. Inter-assay coefficient of variation was 11.2% and intra-assay coefficient of variation was 8.4%.

Gene expressionTotal RNA was extracted from approximately 30 mg of muscle

tissue, using TRIzol extraction (Invitrogen, Eugene, Oregon, USA). The integrity of RNA was assessed using agarose gel electrophoresis. For quantitative polymerase chain reaction (qPCR), total RNA was treated with DNAse (Ambion, Crawley, Texas, USA) and comple-mentary deoxyribonucleic acid (cDNA) was transcribed according to the manufacturer ’s directions (Life Technologies, Carlsbad, California, USA). Equine-specific primers (Supplementary Table I) were designed with Primer3 (primer3.sourceforge.net) from pub-lished equine sequence data (www.ncbi.nlm.nih.gov/nuccore) and used to amplify TNFa, IL-10, and IL-6 messenger RNA (mRNA) using Beta actin (b-actin) and glyceraldehyde-3-phosphate dehydro-genase (GAPDH) as housekeeping genes (Table I). It was determined that b-actin and GAPDH were the most stable housekeeping genes

Supplementary Table I. Primer sequences used for gene expression analysis

Primer Forward (59–39) Reverse (59–39) Functionb-actin agtactccgtatggatcggcg ccggactcgtcgtactcctg Housekeeping geneGAPDH aagtggatattgtcgccatcact aacttgccatgggtggaatc Housekeeping geneTNFa agcccatgttgtagcaaacc aaggctcttgatggcagaga Pro-inflammatoryIL-6 ggatttcctgcagttcagcc ccggactcgtcgtactcctg Pro-inflammatoryIL-10 gatctcccaaatcccatcca ggagagaggtaccacagggttt Anti-inflammatoryGAPDH — Glyceraldehyde-3-phosphate dehydrogenase; TNFa — Tumor necrosis factor alpha; IL-6 — Interleuken-6; IL-10 — Interleuken-10.



based on analysis with a commercially available software program (geNorm; Biogazelle, Zwijnaarde, Belgium). The geometric mean of both housekeeping genes was used to create a normalization factor that was applied to each gene to determine relative expression (RE).

Systemic inflammatory biomarkersSerum amyloid A (TriDelta, Maynooth, County Kildare, Ireland)

and TNFa (Pierce, Rockford, Illinois, USA) were measured in dupli-cate in plasma using commercially available ELISAs as previously described (19,31).

Statistical analysisStatistical analysis was carried out using commercially available

software (SPSS, Armonk, New York, USA). Continuous variables were checked for normality using a Pearson d’Agostino normality test.

A Spearman’s rank correlation coefficient was used to determine the correlation between BCS and markers of inflammation. All horses (N = 35) were included in correlation analysis. Measurement of skeletal muscle inflammation included skeletal muscle gene expression of TNFa, IL-6, and IL-10 and protein expression of TNFa. Systemic inflammation was assessed by measuring TNFa and SAA. Adrenocorticotropic hormone (ACTH) was also included because pituitary hormones may have an anti-inflammatory influence in horses.

Population characteristics were compared among groups of horses [lean or ideal, normoinsulinemic (L-NI); obese, normoinsulinemic (O-NI); and obese, hyperinsulinemic (O-HI)]. All hormones and inflammatory biomarkers were log-transformed for normality. Tukey’s method was used to identify 2 SAA outliers [1744 pg/mL (O-NI) and 4000 pg/mL (L-NI] and 1 TNFa skeletal muscle gene expression (L-NI) outlier, all of which were removed from analysis. Age and insulin concentration were compared among groups using analysis of variance. Inflammatory markers [ACTH, serum amyloid A, TNFa protein expression (plasma and skeletal muscle), and skel-etal muscle gene expression (IL-6, IL-10, and TNFa)] were compared

among groups using analysis of covariance, with group as a factor and age as a covariate. When an interaction between group or age and an inflammatory marker was identified, a general linear model was used to characterize the interaction. For any marker that was different across groups, a Bonferroni’s post-hoc test was carried out to identify which groups were different. Significance for all variables was interpreted to exist at P , 0.05.

Re s u l t sPopulation characteristics of all horses (N = 35) included in the

correlation analysis are presented in Table I. Correlation analysis revealed significant negative associations between BCS and both circulating TNFa (r = 20.41, P = 0.02) and skeletal muscle TNFa (r = 20.37, P = 0.03; Figure 1).

In order to determine how the presence of hyperinsulinemia altered the relationship between inflammation and obesity, further analysis was carried out among O-HI, O-NI, and L-NI horses. As expected based on inclusion criteria, insulin concentration was increased in O-HI horses compared to O-NI or L-NI horses. It was also found that insulin concentration was increased in the O-NI group compared to L-NI horses (P = 0.02). There were no differences in age among groups.

With regard to markers of systemic inflammation, there were no significant differences among groups in plasma SAA (P = 0.91), TNFa (P = 0.08), or ACTH (P = 0.22) concentrations (Figure 2). There was no influence of age or age-and-group interaction on any systemic marker. When evaluating skeletal muscle inflammation, there was a significant difference among groups with respect to skeletal muscle concentrations of TNFa (P = 0.006), with the O-HI horses having significantly lower concentrations compared to O-NI (P = 0.02) and L-NI horses (P = 0.006) (Figure 3). There was also a significant positive impact of age (P , 0.001) and an age-and-group interaction (P = 0.02) on concentrations of TNFa in skeletal muscle (Table II, Figure 4). When skeletal muscle gene expression was assessed, there were no significant differences among groups for expression of TNFa

Figure 1. Relationship between body condition score (BCS) and A — circulating tumor necrosis factor alpha (TNFa) (r = 20.41, P = 0.02) and B — skel-etal muscle TNFa (r = 20.37, P = 0.03).

A B



(P = 0.81), IL-6 (P = 0.44), or IL-10 (P = 0.14) (Figure 3). Skeletal muscle gene expression markers were not influenced by age or the age-and-group interaction.

D i s c u s s i o nNo positive associations between skeletal muscle or systemic

inflammation and obesity were identified in this study. Furthermore, there were no significant increases in systemic or skeletal muscle inflammatory markers in O-HI compared to O-NI or L-NI horses. These findings suggest that neither skeletal muscle nor systemic inflammation is positively associated with obesity or obesity-associated hyperinsulinemia in horses.

Inflammation is considered to be a central component of obesity- associated insulin resistance in humans, with systemic and local (adipose and skeletal muscle) inflammation reported (32–34). There appear to be differences with respect to inflammation between metabolically healthy and unhealthy obese humans, with increased levels of inflammatory markers in metabolically unhealthy humans compared to healthy obese humans (35). Similar criteria for metabolically unhealthy obesity remain to be established for horses.

Hyperinsulinemia is a risk factor for laminitis (5), a painful condi-tion that is an important concern for equine welfare. Therefore, in addition to looking at relationships between inflammation and obe-sity, we also chose to evaluate hyperinsulinemic obesity separately from normoinsulinemic obesity, since there may be mechanistic differences between these 2 groups.

Previous studies have yielded conflicting data with respect to the relationship between systemic inflammatory markers and body condition score or insulin concentration, with one study reporting positive associations between inflammatory markers and BCS or serum insulin concentration (23) and another study reporting no relationship between obesity and inflammation (36). Furthermore, induced weight gain in equids did not alter systemic SAA, TNFa, or adiponectin (21,37).

In our study, we found decreased systemic TNFa concentrations in association with obesity, with no change in systemic SAA or ACTH concentrations. Holbrook et al (22) previously documented a trend towards a decrease in circulating TNFa and decreased (IL-1, IL-6) or unchanged (TNFa) cytokine gene expression of peripheral blood mononuclear cells (PBMC) in obese hyperinsulinemic horses compared to controls. As TNFa promotes transcription of proinflammatory cytokines and chemokines that are important for leu-kocyte migration and immune response (38), the decrease in circulat-ing TNFa and PBMC expression of inflammatory cytokines observed in obese horses may have implications for immune function.

Skeletal muscle gene transcription of pro- or anti-inflammatory cytokines was not altered with obesity in the horses of this study, while skeletal muscle TNFa protein expression was negatively associated with obesity. These findings indicate that skeletal muscle inflammation is not associated with obesity in horses as it is in humans (32,39). In addition, skeletal muscle inflammation was lower in O-HI compared to O-NI or L-NI groups, which suggests that skeletal muscle inflammation was not related to obesity-associated hyperinsulinemia.

Figure 2. Comparisons of systemic inflammatory markers among groups [lean or ideal, normoinsulinemic (L-NI); obese, normoinsulinemic (O-NI); obese, hyperinsulinemic (O-HI)]. A — SAA; B — TNFa; and C — ACTH. Bars = mean 6 standard deviation. There were no significant differences among groups in any systemic inflammatory marker.

A

B

C



As age is a potential confounder in evaluating inflammatory status, age was compared among groups and no significant dif-ferences were found. When examining both systemic and skeletal muscle inflammatory markers among groups, only skeletal muscle

TNFa concentration was influenced by age, with skeletal muscle TNFa increasing with age. Previous studies have demonstrated an increase in systemic inflammation in aged, obese horses compared to aged controls (18) and an inverse relationship between systemic

Figure 3. Comparisons of skeletal muscle inflammatory markers among groups. A — Tumor necrosis factor alpha (TNFa) concentration; B — TNFa gene expression; C — IL-6 gene expression, and D — IL-10 gene expression. Bars = mean 6 standard deviation. Different superscripts (a,b) indicate significant difference among groups (P , 0.05). Only TNFa concentration differed among groups.

A

C

B

D

Figure 4. Interaction of age and group on concentration of tumor necrosis factor alpha (TNFa) in skeletal muscle.

Table II. Parameter estimates for group, age, and group-by-age interaction

BIntercept 1.171O-HI 21.556O-NI 20.579L-NI 0Age 0.018O-HI 3 age 0.081O-NI 3 age 0.044L-NI 3 age 0B — Estimated regression coefficient. O-HI — Obese, hyperinsulinemic; O-NI — Obese, normoinsulinemic; L-NI — Lean, normoinsulinemic.



TNFa and insulin sensitivity in aged mares (19). Based on the findings of the present study, however, it appears that obesity and hyperinsulinemia independent of age do not contribute to a systemic or local pro-inflammatory phenotype.

Skeletal muscle inflammation was evaluated as a potential mecha-nism for hyperinsulinemia, as previous research in horses and other species has demonstrated that inflammation may lead to impaired insulin signaling in skeletal muscle, which may result in insulin resis-tance and subsequent hyperinsulinemia (26,32,40). In this study, we found a significant decrease in skeletal muscle TNFa protein concen-tration with obesity and obesity-associated hyperinsulinemia. Only the semi-membranosus muscle was evaluated, however, and these findings may not be consistent across other muscle groups. Previous studies have suggested that inflammatory state in healthy human skeletal muscle (41) and in rat skeletal muscle after dietary-induced obesity (42) may differ depending on the type of muscle fiber.

Skeletal muscle insulin resistance was not directly evaluated in this study and there could be other organs that contribute to hyperinsulinemia, including the insulin-sensitive tissues, liver and adipose, as well as the GI tract and pancreas. Previous research has suggested that adipose inflammation within select depots may be associated with insulin resistance in the horse (26), and it is possible that local inflammation could influence adipose insulin signaling and systemic insulin regulation without resulting in systemic or skeletal muscle inflammation. Furthermore, recent research suggests that the enteroinsulinar axis may play an important role in equine insulin dysregulation (43) and that tissue insulin resistance may be a secondary event.

In conclusion, despite what has been reported in other species, we did not find a positive relationship between skeletal muscle or systemic inflammation and obesity or obesity-associated hyperin-sulinemia in horses. These findings suggest that skeletal muscle inflammation is not a key mechanism of obesity-associated hyper-insulinemia. Additional work is needed to determine the site and mechanism of hyperinsulinemia in obese horses.

A c k n o w l e d g m e n t sThe authors acknowledge Kim Hill for technical assistance and

Grace Kwong for statistical assistance. Funding was provided by the American Quarter Horse Foundation and the Oklahoma State University Research Advisory Committee.

Re f e r e n c e s 1. Tantiwong P, Shanmugasundaram K, Monroy A, et al. NF-kB

activity in muscle from obese and type 2 diabetic subjects under basal and exercise-stimulated conditions. Am J Physiol Endocrinol Metab 2010;299:E794–E801.

2. Plomgaard P, Nielsen AR, Fischer CP, et al. Associations between insulin resistance and TNF-alpha in plasma, skeletal muscle and adipose tissue in humans with and without type 2 diabetes. Diabetologia 2007;50:2562–2571.

3. Bastard JP, Maachi M, Lagathu C, et al. Recent advances in the relationship between obesity, inflammation, and insulin resis-tance. Eur Cytokine Netw 2006;17:4–12.

4. Trayhurn P, Wood IS. Adipokines: Inflammation and the pleio-tropic role of white adipose tissue. Br J Nutr 2004;92:347–355.

5. Yang RZ, Lee MJ, Hu H, et al. Acute-phase serum amyloid A: An inflammatory adipokine and potential link between obesity and its metabolic complications. PLoS Med 2006;3:e287.

6. Park HS, Park JY, Yu R. Relationship of obesity and visceral adiposity with serum concentrations of CRP, TNF-alpha and IL-6. Diabetes Res Clin Pract 2005;69:29.

7. Calabro P, Chang DW, Willerson JT, Yeh ET. Release of C-reactive protein in response to inflammatory cytokines by human adi-pocytes: Linking obesity to vascular inflammation. J Am Coll Cardiol 2005;46:1112–1113.

8. Anty R, Bekri S, Luciani N, et al. The inflammatory C-reactive protein is increased in both liver and adipose tissue in severely obese patients independently from metabolic syndrome, Type 2 diabetes, and NASH. Am J Gastroenterol 2006;101:1824–1833.

9. Guilherme A, Virbasius JV, Puri V, Czech MP. Adipocyte dysfunc-tions linking obesity to insulin resistance and type 2 diabetes. Nat Rev Mol Cell Biol 2008;9:367–377.

10. Yudkin JS, Kumari M, Humphries SE, Mohamed-Ali V. Inflammation, obesity, stress and coronary heart disease: Is interleukin-6 the link? Atherosclerosis 2000;148:209–214.

11. Trayhurn P, Drevon CA, Eckel J. Secreted proteins from adipose tissue and skeletal muscle-adipokines, myokines and adipose/muscle cross-talk. Arch Physiol Biochem 2011;117:47–56.

12. Daoussis D, Antonopoulos I, Yiannopoulos G, Andonopoulos AP. ACTH as first line treatment for acute calcium pyrophosphate crystal arthritis in 14 hospitalized patients. Joint Bone Spine 2014;81:98–100.

13. Ritter J, Kerr LD, Valeriano-Marcet J, Spiera H. ACTH revisited: Effective treatment for acute crystal induced synovitis in patients with multiple medical problems. J Rheumatol 1994;21:696–699.

14. Levine T. Treating refractory dermatomyositis or polymyositis with adrenocorticotropic hormone gel: A retrospective case series. Drug Des Devel Ther 2012;6:133.

15. Getting SJ, Christian HC, Flower RJ, Perretti M. Activation of melanocortin type 3 receptor as a molecular mechanism for adrenocorticotropic hormone efficacy in gouty arthritis. Arthritis Rheum 2002;46:2765–2775.

16. Beech J, Boston RC, McFarlane D, Lindborg S. Evaluation of plasma ACTH, alpha-melanocyte-stimulating hormone, and insulin concentrations during various photoperiods in clini-cally normal horses and ponies and those with pituitary pars intermedia dysfunction. J Am Vet Med Assoc 2009;235:715–722.

17. McGowan T, Pinchbeck G, McGowan C. Evaluation of basal plasma a-melanocyte-stimulating hormone and adrenocortico-trophic hormone concentrations for the diagnosis of pituitary pars intermedia dysfunction from a population of aged horses. Eq Vet J 2013;45:66–73.

18. Adams AA, Katepalli MP, Kohler K, et al. Effect of body con-dition, body weight and adiposity on inflammatory cyto-kine responses in old horses. Vet Immunol Immunopathol 2009;127:286–294.

19. Vick MM, Adams AA, Murphy B, et al. Relationships among inflammatory cytokines, obesity, and insulin sensitivity in the horse. J Anim Sci 2007;85:1144–1155.



20. Treiber K, Carter R, Gay L, Williams C, Geor R. Inflammatory and redox status of ponies with a history of pasture-associated laminitis. Vet Immunol Immunopathol 2009;129:216–220.

21. Bamford NJ, Potter SJ, Harris PA, Bailey SR. Effect of increased adiposity on insulin sensitivity and adipokine concentrations in horses and ponies fed a high fat diet, with or without a once daily high glycaemic meal. Equine Vet J 2016;48:368–373.

22. Holbrook TC, Tipton T, McFarlane D. Neutrophil and cytokine dysregulation in hyperinsulinemic obese horses. Vet Immunol Immunopath 2012;145:283–289.

23. Suagee J, Corl B, Crisman M, Pleasant R, Thatcher C, Geor R. Relationships between body condition score and plasma inflam-matory cytokines, insulin, and lipids in a mixed population of light-breed horses. J Vet Intern Med 2013;27:157–163.

24. Tadros EM, Frank N, Donnell RL. Effects of equine metabolic syndrome on inflammatory responses of horses to intravenous lipopolysaccharide infusion. Am J Vet Res 2013;74:1010–1019.

25. Burns TA, Geor RJ, Mudge MC, McCutcheon LJ, Hinchcliff KW, Belknap JK. Proinflammatory cytokine and chemokine gene expression profiles in subcutaneous and visceral adipose tis-sue depots of insulin-resistant and insulin-sensitive light breed horses. J Vet Intern Med 2010;24:932–939.

26. Waller A, Huettner L, Kohler K, Lacombe V. Novel link between inflammation and impaired glucose transport during equine insulin resistance. Vet Immunol Immunopathol 2012;149:208–215.

27. Lewis MR, Callas PW, Jenny NS, Tracy RP. Longitudinal stability of coagulation, fibrinolysis, and inflammation factors in stored plasma samples. Thromb Haemost 2001;86:1495–1500.

28. Andreasson A, Kiss NB, Juhlin CC, Höög A. Long-term stor-age of endocrine tissues at 280°C does not adversely affect RNA quality or overall histomorphology. Biopreserv Biobank 2013;11:366–370.

29. Freestone JF, Wolfsheimer KJ, Kamerling SG, Church G, Hamra J, Bagwell C. Exercise induced hormonal and metabolic changes in Thoroughbred horses: Effects of conditioning and acepromazine. Eq Vet J 2010;23:219–223.

30. Perkins GA, Lamb S, Erb HN, Schanbacher B, Nydam DV, Divers TJ. Plasma adrenocorticotropin (ACTH) concentrations and clinical response in horses treated for equine Cushing’s disease with cyproheptadine or pergolide. Eq Vet J 2002;34: 679–685.

31. Pollock P, Prendergast M, Schumacher J, Bellenger C. Effects of surgery on the acute phase response in clinically normal and diseased horses. Vet Rec 2005;156:538–542.

32. Saghizadeh M, Ong JM, Garvey WT, Henry RR, Kern PA. The expression of TNF alpha by human muscle. Relationship to insulin resistance. J Clin Invest 1996;97:1111.

33. Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM. Increased adipose tissue expression of tumor necrosis factor-alpha in human obesity and insulin resistance. J Clin Invest 1995;95:2409.

34. Spranger J, Kroke A, Möhlig M, et al. Inflammatory cytokines and the risk to develop type 2 diabetes: Results of the prospec-tive population-based European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam Study. Diabetes 2003;52:812–817.

35. Phillips CM, Perry IJ. Does inflammation determine metabolic health status in obese and nonobese adults? J Clin Endocrin Metab 2013;98:E1610–E1619.

36. Wray H, Elliott J, Bailey S, Harris P, Menzies-Gow N. Plasma concentrations of inflammatory markers in previously laminitic ponies. Eq Vet J 2014;46:317–321.

37. Suagee J, Burk A, Quinn R, Hartsock T, Douglass L. Effects of diet and weight gain on circulating tumour necrosis factor-a concentrations in Thoroughbred geldings. J Anim Physiol Anim Nutr (Berl) 2011;95:161–170.

38. Roach DR, Bean AG, Demangel C, France MP, Briscoe H, Britton WJ. TNF regulates chemokine induction essential for cell recruitment, granuloma formation, and clearance of myco-bacterial infection. J Immunol 2002;168:4620–4627.

39. Plomgaard P, Nielsen AR, Fischer CP, et al. Associations between insulin resistance and TNF-alpha in plasma, skeletal muscle and adipose tissue in humans with and without type 2 diabetes. Diabetologia 2007;50:2562–2571.

40. de Alvaro C, Teruel T, Hernandez R, Lorenzo M. Tumor necrosis factor alpha produces insulin resistance in skeletal muscle by activation of inhibitor kappa B kinase in a p38 MAPK-dependent manner. J Biol Chem 2004;279:17070–17078.

41. Plomgaard P, Penkowa M, Pedersen BK. Fiber type specific expression of TNF-alpha, IL-6 and IL-18 in human skeletal muscles. Exerc Immunol Rev 2005;11:53–63.

42. Bhatt BA, Dube JJ, Dedousis N, Reider JA, O’Doherty RM. Diet-induced obesity and acute hyperlipidemia reduce IkBa levels in rat skeletal muscle in a fiber-type dependent manner. Am J Physiol Regul Integr and Comp Physiol 2006;290:R233–R240.

43. de Laat MA, McGree JM, Sillence MN. Equine hyperinsulinemia: Investigation of the enteroinsular axis during insulin dysregula-tion. Am J Physiol Endocrinol and Metab 2016;310:E61–E72.


Article


I n t r o d u c t i o nSampling of mammary tissue in dairy cows through biopsy has

long been considered a difficult and even an uncertain task due to its effects on milk production and mammary health (1). In studies with dairy cows, Latin-square designs are widely used mainly because they provide a rapid and powerful response to treatments (2). When biopsies of the mammary gland are taken repeatedly in studies such as Latin-square experiments, however, bleeding and mastitis may

occur that will adversely affect milk production in subsequent peri-ods of the experiment and alter the response to treatments. As the amount of tissue obtained from biopsies limits the number of labora-tory analyses that can be done, a technique is needed that provides large-core samples and minimizes damage to the mammary gland. An adequate technique of biopsy may reduce variation in results and secondary complications to animals.

Over the years, a number of methods of mammary gland biopsy have been used in dairy cows (1,3,4). One of the most popular

A new technique for repeated biopsies of the mammary gland in dairy cows allotted to Latin-square design studies

Luciano S. de Lima, Eric Martineau, Francilaine E. De Marchi, Marie-France Palin, Geraldo T. dos Santos, Hélène V. Petit

A b s t r a c tThe objective of this study was to develop a technique for carrying out repeated biopsies of the mammary gland of lactating dairy cows that provides enough material to monitor enzyme activities and gene expression in mammary secretory tissue. A total of 16 Holstein cows were subjected to 4 mammary biopsies each at 3-week intervals for a total of 64 biopsies. A 0.75-cm incision was made through the skin and subcutaneous tissue of the mammary gland and a trocar and cannula were inserted using a circular motion. The trocar was withdrawn and a syringe was plugged into the base of the cannula to create a vacuum for sampling mammary tissue. To reduce bleeding, hand pressure was put on the surgery site after biopsy and skin closure and ice was applied for at least 2 h after the biopsy using a cow bra. The entire procedure took an average of 25 min. Two attempts were usually enough to obtain 800 mg of tissue. Visual examination of milk samples 10 d after the biopsy indicated no trace of blood, except in samples from 2 cows. All wounds healed without infection and subcutaneous hematomas resorbed within 7 d. There was no incidence of mastitis throughout the lactation. This technique provides a new tool for biopsy of the mammary gland repeated at short intervals with the main effect being a decrease in milk production. Although secondary complications leading to illness or death are always a risk with any procedure, this biopsy technique was carried out without complications to the health of animals and with no incidence of mastitis during the lactation.

R é s u m éCette étude a été conduite avec l’objectif de décrire une technique pour laquelle les biopsies de la glande mammaire des vaches laitières en lactation sont répétées. Un total de 16 vaches Holstein ont été soumises chacune à 4 biopsies de la glande mammaire à un intervalle de 3 semaines pour un total de 64 biopsies. Une incision de 0,75 cm a été faite à travers la peau et le tissu sous-cutané de la glande mammaire, et un trocart et une canule ont été insérés en utilisant un mouvement circulaire. Le trocart a été retiré et une seringue a été attachée à la base de la canule pour créer un vacuum afin d’échantillonner le tissu mammaire. Afin de réduire le saignement, une pression manuelle a été appliquée sur le site de la chirurgie après la biopsie et la suture de l’incision de la peau, et de la glace a été appliquée pour au moins 2 h après la biopsie en utilisant une brassière pour vache. La procédure entière a exigé une moyenne de 25 min et deux essais ont habituellement été suffisants pour obtenir 800 mg de tissu. Un examen visuel des échantillons de lait n’ont indiqué aucune présence de sang 10 jours après la biopsie sauf pour deux vaches. Les plaies ont toutes guéries sans infection, et les hématomes sous-cutanés se sont résorbés à l’intérieur d’une période de 7 jours. Il n’y a eu aucune incidence de mammite durant la lactation. Cette technique décrit un nouvel outil de biopsie de la glande mammaire répété à de courts intervalles où l’effet principal a été une baisse de la production laitière. Bien que les complications secondaires entrainant la maladie ou la mort soient toujours un risque avec toute procédure, cette technique de biopsie a été faite sans complications pour la santé des animaux et il n’y a eu aucune incidence de mammite durant la lactation.

(Traduit par les auteurs)

Departamento de Zootecnia, Universidade Estadual de Maringá, Maringá, PR 87020-900, Brazil (Lima, Marchi, Santos); Clinique Vétérinaire de Coaticook, Coaticook, Québec J1A 1P9 (Martineau); Dairy and Swine Research and Development Centre, Agriculture and Agri-Food Canada, 2000 College Street, Sherbrooke, Quebec J1M 0C8 (Palin, Petit).

Address all correspondence to Dr. Hélène V. Petit; telephone: 819-780-7210; fax: 819-564-5507; e-mail: [email protected]

Received October 6, 2015. Accepted February 22, 2016.



techniques is that developed by Farr et al (1), in which a core of mammary secretory tissue is extracted using a stainless steel cannula with a retractable blade at the cutting edge and a slow-speed electric motor to rotate the biopsy instrument. Our team has previously used this procedure in a few Latin-square experiments and has noticed trauma to the mammary gland and excessive bleeding. Indeed, VanKlompenberg et al (3) have reported that the drill-operated instrument proposed in this earlier study (1) may lead to excessive blood loss. As this may cause discomfort and health problems for cows subjected to repeated sampling of the mammary gland, such as those carried out in Latin-square experiments, the objective of this study was to develop a biopsy technique that can be used fre-quently on lactating dairy cows and that provides enough material to monitor enzyme activities and gene expression in mammary secretory tissue.


AnimalsAll animals were cared for in accordance with the guidelines of

the Canadian Council on Animal Care (5) and all biopsy procedures were approved by the local Animal Care Committee. Individual observations of 16 Holstein cows from 2 separate lactating cow-feeding trials carried out as replicated 4 3 4 Latin-square design were used. The trials were conducted in the same year with different treatment diets and using different cows. Details on dietary treat-ments and experimental procedures of experiments 1 and 2 have been published previously (6,7, respectively).

Briefly, the experimental diets were based on corn silage and grass silage. In experiment 1, the effects of dietary flax meal and abomasal infusion of flax oil were evaluated, while in experiment 2, the effects of dietary flax meal and abomasal infusion of sunflower oil were assessed. Cows ranged from 35 to 105 d in lactation and 2 to 4 d in parity. They were housed in individual stalls with free access to water and were fed twice a day for ad libitum intake [100 g/kg body weight (BW) of refusals as fed]. Cows were milked twice daily and milk production was recorded at every milking.

Preparation, sedation, and pain managementAll cows were submitted to the California Mastitis Test (Dairy

Research Products, Ancaster, Ontario) before each biopsy to estimate somatic cell counts and to detect the presence of mastitis. On the day before the biopsy, udders were clipped to facilitate cleaning and aseptic procedures. On the day of the biopsy, cows were placed in a restraining cage and mildly sedated with an injection of 10 mg of xylazine (Zoetis Canada, Kirkland, Quebec) in the coccygeal vein. A 10-cm2 piece of skin on the upper portion of the udder hindquarter was washed and aseptically prepared 3 times with 70% alcohol and 2% chlorhexidine acetate solution (Hibitane; Wyeth Animal Health, Guelph, Ontario). Before the last washing with alcohol and Hibitane, the biopsy site was anesthetized by injecting 3 mL (1 mL intrader-mal and 2 mL subcutaneous) of lidocaine hydrochloride (HCL) 2% (Bimeda-MTC Animal Health, Cambridge, Ontario).

Design of biopsy instrumentThe biopsy instrument consisted of 3 main pieces: a trocar, a can-

nula, and a 30-mL syringe used to create a vacuum (Figure 1A). The trocar was made according to the design of Hughes (8), which was originally intended for liver biopsy in cattle. Both trocar and cannula were made of stainless steel. The cannula was 31 cm long, with an outer diameter of 9.5 mm and an inner diameter of 8 mm. The trocar was 34 cm long with a diameter that fit snugly into the cannula. The base of the trocar had a knurled end to allow a better grip with the fingers while carrying out the procedure. It is worth noting that the biopsy instrument described in the present experiment is not the

Figure 1. A — The 3 pieces of the biopsy instrument; and B — The base of the 3.8-cm needle used to plug the syringe into the base of the cannula. The needle itself was removed and only the base was used to connect to the 30-mL syringe with slip tip.

Figure 2. Ice applied to the biopsy site using a cow bra.



standard one used to obtain a percutaneous liver specimen in clinical cases where a Tru-Cut biopsy needle is commonly used.

Biopsy of mammary tissueAll biopsy instruments were autoclaved before use. Biopsies were

carried out 6 h after the morning milking as previously described (1). Mammary secretory tissue was obtained from the junction of the upper and middle third of the hindquarters. The left and right hind-quarters of the udder were alternated from one experimental period to another and biopsies were taken at least 10 cm away from the first site when a quarter was used for a second time. The mammary gland of cows was examined by ultrasonography using a Concept/MCV Veterinary Ultrasound Scanner equipped with a linear array 5 MHz probe (Tokyo Keiki, Tokyo, Japan) before skin incision to avoid any large subcutaneous blood vessels.

A 0.75-cm incision was then made through the skin and subcu-taneous tissue with a scalpel. The trocar and cannula were inserted through the gland capsule using a circular motion. The tip of the cannula was beveled inward to form a sharp cutting edge and to allow easy penetration of the mammary gland. The trocar was then withdrawn and the cannula was driven deep enough to reach into the mammary parenchyma, while applying a circular motion with the hands to properly cut a core of mammary tissue and fill up the cannula. A 30-mL slip-tip syringe (Becton Dickinson, Rutherford, New Jersey, USA) was then plugged into the base of the cannula using a 3.8-cm needle (Figure 1B). The needle itself was removed in order to use only the base to connect to the 30-mL slip-tip syringe. Mammary secretory tissue samples were drawn into the cannula by way of a vacuum created by the syringe by rotating the cannula in a circular motion. In order to detach the distal part of the biopsy and leave it free, the cannula was then removed from the mammary gland, again using a circular motion.

Cores of 800 mg (wet weight of mammary tissue) were required in order to conduct enzyme activities and gene expression analyses. If not enough mammary tissue was obtained in the first attempt, another core was obtained immediately after, using the same 0.75-cm incision. It was always attempted to avoid reaching the abdomen and to cut at different angles, with the first cut at a lower angle than the second one. Hand pressure was applied to the surgery site between attempts, after biopsy, and after skin closure to control bleeding. The skin incision was closed with a cruciate polydioxanone monofila-ment synthetic absorbable suture (PDS II 2-0 CP-1; Ethicon, Cornelia, Georgia, USA).

Postoperative careA chemical bandage (Aluspray; Neogen, Lexington, Kentucky,

USA) was applied to the skin of the mammary gland to protect the incision from infections. To reduce bleeding, ice was applied to the incision site for at least 2 h after the biopsy using a cow bra (Figure 2). An intramuscular dose (30 mL) of penicillin (Pen-aqueous; Agripharm Products, Westlake, Texas, USA) was given immediately after the biopsy and twice a day for the next 4 d to prevent wound infection, following the procedure used for large animals at the Faculté de Médecine Vétérinaire in St-Hyacinthe, Quebec. Within 2 h of the biopsy, cows were hand-milked to remove intramammary blood clots. Cows were hand-stripped as required at each milking

over the next 4 to 7 d until all blood clots were removed. Body tem-perature was monitored once daily for 8 d after the biopsy. The skin sutures were removed 7 to 10 d after the biopsy.

Statistical analysisData on milk production, i.e., before and after biopsy and decrease

in milk production after biopsy, were analyzed as repeated measure-ments using the MIXED procedure of SAS (SAS 2000; SAS Institute, Cary, North Carolina, USA) and covariance structures were modeled separately for each variable. Main sources of variation were square, period, and biopsy considered as fixed effects and cow within square as a random effect. When the 2 experiments were analyzed together, the experiment was considered as a fixed effect. Results were reported as least squares means with standard error of the mean (SEM). Significant differences were set at P # 0.05.

Re s u l t sUltrasonography of the mammary gland was carried out in the

biopsies of the first 8 cows and revealed that, despite the presence of large vessels, there was little damage from the present biopsy technique. It was therefore decided to discontinue the practice for the rest of the biopsies. Each of the 16 cows was subjected to 4 biopsies every 3 wk for a total of 64 mammary biopsies. Because the cows were mildly sedated, they were calm and easy to handle throughout the procedure, which saved time between biopsies carried out on different animals on the same day. The entire procedure took an average of 25 min from sedation to skin closure.

After the biopsies, no cows had to be removed from the experi-ments due to complications and none experienced fever as deter-mined by daily monitoring of body temperature. Although there was no control group without a biopsy to compare the effects of the surgi-cal procedure on health status and feed intake, cows were healthy and dry matter intake was slightly reduced from 30.5 to 28.3 kg/d (average of the 4 d before and after the biopsy, respectively).

Using the biopsy instrument, cores of 800 mg (wet weight of mam-mary secretory tissue) could be obtained from 1 biopsy, although most samples ranged from 200 to 700 mg. As many as 5 attempts were made during a given session to obtain an adequate amount of sample, although an average of 2.43 6 0.84 attempts were enough to obtain the 800 mg required.

Most cows had some bleeding within 24 h of the biopsy. Although no blood loss counts were done and the precise amount of blood lost could therefore not be determined, most cows lost only a few milliliters of blood. Careful hand-stripping was enough to remove clotted blood from the glands. Most glands were free of blood clots within the first 24 h of the biopsy and no anti-bleeding drugs were required. No difference in bleeding was observed between the first and following biopsies. Visual examination revealed the presence of blood in milk for most cows up to 6.11 6 0.78 d after the biopsy. The same 2 cows showed some residual blood in milk by day 10 post-biopsy in all 4 experimental periods.

All biopsy wounds healed without infection and subcutaneous hematomas resorbed within 7 d. Although no clinical mastitis was observed for any of the biopsied cows throughout the remainder of lactation, the biopsy procedure decreased (P , 0.0001) milk



production for each of the 4 periods (Table I). The average milk pro-duction decreased for 4 d after the biopsy compared to the 4 d before the biopsy (P , 0.0001) (Table II). This decrease in milk production after the biopsy averaged 2.8 kg/d in experiment 1 and this average decrease was greater (P , 0.001) in experiment 2 at 5.5 kg/d.

D i s c u s s i o nThis study developed a different technique than the one devel-

oped by Farr et al (1). In this last study, a core of secretory tissue was extracted using an electric-driven, rotating stainless steel can-nula with a retractable blade at the cutting edge, which may lead to excessive bleeding (3). Conversely, in the present experiment, the trocar was inserted through the gland capsule with a gentle, manually driven circular motion and samples were withdrawn using the vacuum created by a syringe. The trocar used was origi-nally intended for liver biopsy in cattle, although different from the standard instrument used to obtain percutaneous liver specimen in clinical cases where a tru-cut biopsy needle is commonly used.

Only 2 animals showed some residual blood in milk by day 10 post-biopsy, which suggests that these 2 cows had abnormal clot formation or coagulation time. There was no actual blood in the milk, however, and all cows had completely healed by the next biopsy 21 d later. Use of this biopsy instrument, applying pressure between and after biopsies, and putting ice on the incision site for at least 2 h after the biopsy likely helped to avoid excessive bleeding and accelerated healing of the mammary gland.

Samples of mammary secretory tissue obtained after 2 biopsies were usually considered large enough to carry out all tests required to determine enzyme activities and gene expression. Moreover, sampling once every 21 d allowed detection of statistical differences in enzyme activity and gene expression among cows subjected to various treatments assigned to Latin-square designs (6,7).

As already stated, milk production decreased after biopsies, with a greater decrease after the second experiment. Overall, milk produc-tion was higher in experiment 2 than in experiment 1 [45.5 compared to 32.6 kg/d, respectively (6,7)], which suggests that high-producing cows are more affected by biopsies. Nutrition projects were con-ducted in parallel using the same cows and it is worth noting that milk production was not affected by the dietary treatments in either experiment (6,7).

The decrease in average milk production after the biopsies is in agreement with the results of Farr et al (1) who reported lower milk production for 6.5 d after biopsy of the mammary gland when dairy cows were subjected to a procedure using an electric rotating cannula with a retractable blade at the cutting edge. Moreover, a similar decrease in milk production has been reported after 2 biop-sies carried out 30 d apart by Oxender et al (4) who used scissors to carry out biopsies of the mammary gland in dairy cows. In general, these results suggest that lower milk production is normal immedi-ately after a biopsy due to the trauma and stress of the procedure. The mammary gland has many blood vessels and blood supply is

Table I. Descriptive data of milk production of Holstein cows subjected to 4 biopsies of the mammary gland at 21-day intervals

Period 1 Period 2 Period 3 Period 4 (First biopsy) (Second biopsy) (Third biopsy) (Fourth biopsy)Experiment 1 (n = 8) Milk production (kg/d) Pre-biopsy (4-day average) 36.6 33.1 30.5 25.6 Post-biopsy (4-day average) 34.5 28.1 27.4 23.1 Decrease 22.1 25.0 23.1 22.5

Experiment 2 (n = 8) Milk production (kg/d) Pre-biopsy (4-day average) 47.4 41.8 41.5 38.9 Post-biopsy (4-day average) 40.8 37.8 35.0 30.1 Decrease 26.5 24.0 26.5 28.8

Experiments 1 and 2 (n = 16) Milk production (kg/d) Pre-biopsy (4-day average) 42.0 37.4 36.0 32.3 Post-biopsy (4-day average) 37.7 32.9 31.2 26.6 Decrease 24.3 24.5 24.8 25.7

Table II. Average milk production before and after biopsy of the mammary gland in Holstein cows

4-day average (kg/d) Before After biopsy biopsy SEM P-valueExperiment 1 and 2 (n = 64) 36.9 32.1 1.19 , 0.0001 1 (n = 32) 31.5 28.3 0.67 , 0.0001 2 (n = 32) 42.4 35.9 0.90 , 0.0001SEM — Standard error of mean.



extremely important for its function (9). Internal bleeding and clot formation could therefore impair blood supply to alveolar cells and obstruct milk ducts, thus lowering milk production.

Milk production decreased over time. Our results suggest that the trauma and stress of the repeated biopsies carried out over time likely contributed to lower milk production from the first to the fourth biopsy. Although the biopsy is clearly the most important factor that affected milk production in our experiment, it is well-known that several factors such as intake of dry matter and lactation stage also affect milk production. Indeed, the intake of dry matter decreased after each biopsy, which was likely due to the trauma and stress of the procedure. Nutritional studies were conducted at the same time as the present study and the use of non-steroidal anti-inflammatory drugs may have affected the response of measure-ments such as gene expression to feeding treatments. As a result, cows received only antibiotic treatment and no pain management. The decrease in intake of dry matter immediately after the biopsy could have contributed to decreased milk production as it is related to the amount of nutrients ingested by animals.

Lactation stage combined with other factors already discussed could also have influenced milk production. Indeed, there was a 63-day interval between the first and fourth biopsy and cows ranged from 35 to 105 d in milk, which meant that some cows were on the decreasing part of the lactation curve (10). It is well-established that apoptotic death of secretory cells in the mammary gland accounts for the decline in milk yield that follows peak production in dairy cows (11,12). Natural death of mammary secretory cells could therefore be at least partially responsible for the observed decline in total milk yield. This is supported by results reported by Capuco et al (13) who observed that milk yield declined by 23% from day 90 (peak of lactation) to day 240 of lactation, combined with a decline in epithelial mammary deoxyribonucleic acid (DNA) in multiparous Holstein cows milked twice a day. Similarly, Pollott (14) observed a decline of almost 11% in milk production of Holstein cows from peak of lactation (day 35) to day 90.

In conclusion, this biopsy technique successfully allows tissue sampling of the mammary gland to be repeated every 3 wk in lactating dairy cows allotted to Latin-square designs, with the aim of providing mammary secretory tissue samples large enough to monitor enzyme activity and gene expression. The procedure took an average of 25 min and ultrasonography was not required. Although secondary complications leading to illness or death are always a risk with any procedure, this biopsy technique was carried out without complications to the health of animals and there was no incidence of mastitis during the lactation.

A c k n o w l e d g m e n t sThis study was funded by Agriculture and Agri-Food Canada.

The authors gratefully acknowledge the staff of the Dairy and Swine Research and Development Centre for their contribution to the pres-ent study. Special thanks to Danielle Beaudry for technical assistance.

Luciano S. de Lima and Francilaine E. De Marchi were recipients of a studentship and Geraldo T. dos Santos was the recipient of a fel-lowship from the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq-Brazil).

Re f e r e n c e s 1.� Farr VC, Stelwagen K, Cate LR, Molenaar AJ, McFadden TB,

Davis SR. An improved method for the routine biopsy of bovine mammary tissue. J Dairy Sci 1996;79:543–549.

2.� Robinson PH, Wiseman J, Udén P, Mateos G. Some experimental design and statistical criteria for analysis of studies in manu-scripts submitted for consideration for publication. Anim Feed Sci Technol 2006;129:1–11.

3.� VanKlompenberg MK, McMicking HF, Hovey RC. Technical note: A vacuum-assisted approach for biopsying the mammary glands of various species. J Dairy Sci 2012;95:243–246.

4.� Oxender WD, Askew EW, Benson JD, Emery RS. Biopsy of liver, adipose tissue and mammary gland of lactating cows. J Dairy Sci 1971;54:286–288.

5.� Canadian Council on Animal Care. Guide to the Care and Use of Farm Animals in Research, Teaching and Testing, Ottawa: CCAC, 2009.

6.� Lima LS, Palin MF, Santos GT, Benchaar C, Petit HV. Effects of supplementation of flax meal and flax oil on mammary gene expression and activity of antioxidant enzymes in mam-mary tissue, plasma and erythrocytes of dairy cows. Livest Sci 2015;176:196–204.

7.� De Marchi FE, Palin M-F, dos Santos GT, Lima LS, Benchaar C, Petit HV. Flax meal supplementation on the activity of anti-oxidant enzymes and the expression of oxidative stress- and lipogenic-related genes in dairy cows infused with sunflower oil in the abomasum. Anim Feed Sci Technol 2015;199:41–50.

8.� Hughes JP. A simplified instrument for obtaining liver biopsies in cattle. Am J Vet Res 1962;23:1111–1113.

9.� Akers RM, Denbow DM. Anatomy and Physiology of Domestic Animals. Ames, Iowa: Blackwell Publishing, 2008.

10.� Wood PDP. Algebraic model of the lactation curve in cattle. Nature 1967;216:164–165.

11.� Knight CH, Wilde CJ. Mammary cell changes during pregnancy and lactation. Livest Prod Sci 1993;35:3–19.

12.� Wilde CJ, Addey CV, Li P, Fernig DG. Programmed cell death in bovine mammary tissue during lactation and involution. Exp Physiol 1997;82:943–953.

13.� Capuco AV, Wood DL, Baldwin R, McLeod K, Paape MJ. Mammary cell number, proliferation, and apoptosis during a bovine lactation: Relation to milk production and effect of bST. J Dairy Sci 2001;84:2177–2187.

14.� Pollott GE. Short communication: Do Holstein lactations of varied lengths have different characteristics? J Dairy Sci 2011; 94:6173–6180.


Article


I n t r o d u c t i o nOrthopedic infections in horses are life-threatening emergencies

that may affect the animal’s athletic future or require euthanasia. Intravenous regional limb perfusion (IVRLP) is commonly used to treat orthopedic infections because it delivers high antimicrobial concentrations above the minimum inhibitory concentration (MIC) for common equine bacterial isolates in the synovial, osseous, and

soft-tissue structures of the distal limb, while limiting systemic side effects and cost (129). Compared with systemic antimicrobial administration, IVRLP results in higher regional concentration and reduces the risk of systemic toxicity that leads to renal and gastro-intestinal side effects (1).

Common pathogens cultured from orthopedic infections in horses include Enterobacteriaceae, streptococci, staphylococci, and Pseudomonas spp. (10,11). Most of the studies on regional limb

Pharmacokinetics of a combination of amikacin sulfate and penicillin G sodium for intravenous regional limb perfusion in adult horses

Jorge E. Nieto, Jan Trela, Scott D. Stanley, Sawsan Yamout, Jack R. Snyder

A b s t r a c tThe aim of this study was to determine the pharmacokinetics of amikacin and penicillin G sodium when administered in combination as an intravenous regional limb perfusion (IVRLP) to horses. Seven healthy adult horses underwent an IVRLP in the cephalic vein with 2 g of amikacin sulfate and 10 mill IU of penicillin G sodium diluted to 60 mL in 0.9% saline. A pneumatic tourniquet set at 450 mmHg was left in place for 30 min. Synovial fluid was collected from the metacarpophalangeal joint 35 min and 2, 6, 12, and 24 h after infusion of the antimicrobials. Concentrations of amikacin and penicillin in synovial fluid were quantitated by liquid chromatography tandem-mass spectrometry analysis. Therapeutic concentrations of amikacin and penicillin for equine-susceptible pathogens were achieved in the synovial fluid. Maximum synovial concentrations (Cmax) (mean 6 SE) for amikacin and penicillin were 132 6 33 mg/mL and 8474 6 5710 ng/mL, respectively. Only 3 horses had detectable levels of penicillin at 6 h and 1 at the 12 h sample. The combination of amikacin with penicillin G sodium via IVDLP resulted in reported therapeutic concentrations of both antibiotics in the synovial fluid. The Cmax:MIC (minimum inhibitory concentration) ratio for amikacin was 8:1 and Time . MIC for penicillin was 6 h. At 24 h, the mean concentration of amikacin was still above 4 mg/mL. Terminal elimination rate constants (T1/2 lambdaz) were 13.6 h and 2.8 h for amikacin and penicillin, respectively. The use of IVDLP with penicillin may therefore not be practical as rapid clearance of penicillin from the synovial fluid requires frequent perfusions to maintain acceptable therapeutic concentrations.

R é s u m éL’objectif de la présente étude était de déterminer la pharmacocinétique de l’amikacine et de la pénicilline G sodique lorsqu’administrées en combinaison par perfusion intraveineuse régionale d’un membre (PIVRM) à des chevaux. Sept chevaux adultes ont reçu une PIVRM dans la veine céphalique avec 2 g de sulfate d’amikacine et 10 millions d’UI de pénicilline G sodique dilués dans 60 mL de saline 0,9 %. Un tourniquet pneumatique réglé à 450 mmHg a été laissé en place pour 30 min. Du liquide synovial a été récolté de l’articulation métacarpo-phalangienne 35 min, 2, 6, 12, et 24 h après l’infusion des antimicrobiens. Les concentrations d’amikacine et de pénicilline dans le liquide synovial furent mesurées par spectrométrie de masse en tandem avec la chromatographie en phase liquide. Les concentrations thérapeutiques d’amikacine et de pénicilline pour des agents pathogènes équins sensibles ont été atteintes dans le liquide synovial. Les concentrations synoviales maximales (Cmax) [moyenne 6 écart-type (EC)] pour l’amikacine et la pénicilline étaient de 132 6 33 mg/mL et 8474 6 5710 ng/mL, respectivement. Seulement 3 chevaux avaient des quantités détectables de pénicilline à 6 h et un seul pour l’échantillon de 12 h. La combinaison d’amikacine et de pénicilline G sodique via PIVRM a permis de rapporter des concentrations thérapeutiques des deux antibiotiques dans le liquide synovial. Le ratio Cmax-CMI (concentration minimale inhibitrice) pour l’amikacine était de 8:1 et la période de Temps . CMI pour la pénicilline était de 6 h. À 24 h, la concentration moyenne d’amikacine était toujours supérieure à 4 mg/mL. Les constantes du taux d’élimination terminal (T1/2 lambdaz) étaient 13,6 h et 2,8 h pour l’amikacine et la pénicilline, respectivement. L’utilisation de PIVRM avec la pénicilline ne serait ainsi pas pratique étant donné que la clairance rapide de la pénicilline à partir du liquide synovial requière des perfusions fréquentes pour maintenir des concentrations thérapeutiques acceptables.


Department of Surgery and Radiological Sciences (Nieto, Trela, Yamout, Snyder) and K.L. Maddy Equine Analytical Chemistry Laboratory (Stanley), School of Veterinary Medicine, University of California, Davis, California 95616, USA.

Address all correspondence to Dr. Jorge E. Nieto; telephone: (530) 752-9773; fax: (530) 752-9815; e-mail: [email protected]

Received November 2, 2015. Accepted February 26, 2016.



perfusion in horses have used aminoglycosides that provide good coverage for Gram-negative bacteria and Staphylococcus, but lack activity against other Gram-positive bacteria. Combinations of a beta-lactam with an aminoglycoside are commonly used systemi-cally since they provide broad-spectrum coverage and have good drug synergy (4). At our hospital, a combination of systemic procaine penicillin and gentamicin is used as a first option for orthopedic infections until sensitivity results are obtained. Aminoglycosides used with intravenous regional limb perfusion (IVRLP), which has been reported in research studies, are the most common antimicro-bials used regionally in clinical cases with orthopedic infections (1,427,12215). The use of intraosseous or IVRLP with penicillin has only been reported in clinical cases (3,6,9,16,17).

When combining beta-lactams with aminoglycosides, both in-vitro and in-vivo drug interaction has been reported (15,18220). The interaction depends on medium, temperature, time, pH, and con-centration, which results in lower antimicrobial concentration of both antibiotics and can still occur in body fluids, such as urine. The reported mechanism of interaction involves a nucleophilic open-ing of the beta-lactam ring and a reaction with an amino group of the aminoglycoside to form an inactive amide (18,21). An in-vitro study comparing the interaction of 6 beta-lactam antibiotics with 5 aminoglycosides found that amikacin was the aminoglycoside least inactivated by all beta-lactams, retaining its activity with minor changes at 48 h. The combination of penicillin G and amikacin was the most stable of all combinations evaluated (21).

Aminoglycosides are commonly used to treat serious enterococ-cal, mycobacterial, staphylococcus, and Gram-negative bacterial infections. Among the aminoglycosides, amikacin is useful for gentamicin-resistant, Gram-negative pathogen infections and is one of the most commonly used antibiotics in IVRLP due to its concentration-dependent action (6). Penicillin is still one of the most commonly used beta-lactam antibiotics in veterinary medicine. Since penicillin is considered time-dependent in its activity, the time (T) of drug concentration above the MIC (T . MIC) is important to clini-cal success (22,23). Unlike aminoglucosides, beta-lactam antibiotics exhibit little concentration-dependent killing. Penicillin has only post-antibiotic effect for staphylococci and animal data suggest that levels need to exceed the MIC for 90% of the dosing interval against gram-negative bacilli and streptococci, but only 50% to 60% for staphylococci in neutropenic animals. In nonneutropenic animals, however, the T . MIC can be reduced to 25% to 30% (23).

Bacterial killing with penicillin depends on time and when admin-istered via IVRLP is likely to result in local therapeutic concentra-tions of the antimicrobial for a longer time than is possible with systemic administration. It has been hypothetized that the use of time-dependent antimicrobials for IVRLP can be justified because it is likely to result in therapeutic concentrations of the antimicrobial in infected ischemic tissues for a longer time than is possible with systemic administration (6). In addition, after the tourniquet is released, the high antimicrobial concentrations in the surrounding tissues may serve as a depot, producing continuous diffusion from the surrounding tissues to the synovial structures (24,25).

The purpose of this study was to determine the pharmacokinetics of amikacin and penicillin G sodium when administered in combina-tion as an IVRLP to healthy adult horses.

M a t e r i a l s a n d m e t h o d sAdult horses, 4 geldings and 3 females, with an average age of

16 y (5 to 18 y) and an average weight of 520 kg (500 to 603 kg) were used in the study. Breeds included 2 Thoroughbreds, 4 Quarter horses, and 1 Standardbred. Horses were healthy based on physical examination and had no signs of lameness or musculoskeletal injury.

The protocol was approved by the University of California Institutional Animal Care and Use Committee. Horses were sedated with detomidine hydrochloride [0.02 mg/kg body weight (BW)], intravenously (IV) and butorphanol tartrate (0.02 mg/kg BW, IV), administered via a 14-gauge (ga) IV catheter placed in a jugular vein. If horses required additional sedation, detomidine hydrochloride (0.004 mg/kg BW IV) was administered. Effort was made to main-tain adequate sedation to prevent limb movement until tourniquet removal. A pneumatic tourniquet with a 10.5-cm cuff was placed at the mid-to-proximal antebrachium and insuflated at a pressure of 450 mmHg. On a randomly selected front limb, a 24-mm long, 20-ga IV catheter was aseptically placed into the cephalic vein approxi-mately 10 cm proximal to the accessory carpal bone. Two grams of amikacin sulfate (Amikacin; Teva Pharmaceuticals, Sellersville, Pennsylvania, USA) and 10 mill IU of penicillin (Penicillin G Sodium; Sandoz, Princeton, New Jersey, USA) were diluted to a total volume of 60 mL with saline (0.9% Sodium Chloride; Baxter Healthcare, Deerfield, Illinois, USA) immediately before infusion and slowly injected into the cephalic vein over 1.5 min. The tourniquet was removed 30 min after injection.

Synovial fluid was collected from all 7 horses within 5 min of tourniquet removal (35 min) and 2, 6, 12, and 24 h after the antimi-crobials were administered. Synoviocentesis was conducted using aseptic techniques from the metacarpophalangeal joint using a lateral approach through the collateral proximal sesamoidean ligament as previously described (26). Synovial fluid (1 mL) was collected and immediately centrifuged at 1700 3 g for 5 min and frozen at 280°C until analysis. Horses were evaluated for lameness, joint swelling, and phlebitis every 24 h for 3 d after the IVRLP.

Measurement of amikacin and penicillin concentrations in synovial fluid

Concentrations of amikacin and penicillin were quantitated in horse synovial fluid by liquid-chromatography-tandem mass spectrometry analysis, using modifications of previously published methods (27,28). Tobramycin and penicillin V were used as the internal standard for amikacin and penicillin analyses, respectively. For amikacin analysis, plasma calibrators were prepared by diluting the working standard solutions with drug-free synovial fluid col-lected from horses to concentrations ranging from 1.0 to 600 mg/mL. Synovial fluid calibrators for penicillin analysis were similarly prepared to concentrations ranging from 10 to 100 000 ng/mL. Fresh calibration curves were prepared for each quantitative assay. In addition, quality control samples, prepared at concentrations within the standard curve, were included with each sample set as an additional check of accuracy.

The response for both amikacin and penicillin was linear and gave correlation coefficients (R2) of 0.99 or better. For amikacin analysis, the accuracy (percentage of nominal concentration) and precision



(percentage relative to standard deviation) were 93% and 108% and 3% and 4% at 8.0 and 80.0 mg/mL, respectively. For penicillin analysis, accuracy and precision were 98%, 102%, and 98% and 3%, 3%, and 1% for 150, 8000, and 80 000 ng/mL, respectively. Accuracy and precision for both assays were considered acceptable based on US Food and Drug Administration (FDA) guidelines for bioanalyti-

cal method validation. The technique was optimized to provide a minimum limit of quantification (LOQ) of 1 mg/mL and a limit of detection (LOD) of 0.1 mg/mL for amikacin and an LOQ of 10 ng/mL and LOD of 0.5 ng/mL for penicillin.

Pharmacokinetic analysisNonlinear least square regression was carried out on the amikacin

concentration in synovial fluid versus time data using commer-cially available software and non-compartmental analysis (Phoenix WinNonlin Version 6.2; Pharsight, Carey, North Carolina, USA). Due to the limited detection time, naïve pooling of datum points was used to combine data from different horses at each time point before pharmacokinetic analysis of penicillin concentration in syno-vial fluid. Non-compartmental analysis for sparse data was used to determine the pharmacokinetic parameters for penicillin. The area under the curve (AUC) for both amikacin and penicillin was calcu-lated using the log-linear trapezoidal rule.

Re s u l t sNo lameness, joint effusion, or obvious phlebitis was observed

in any of the horses. The concentrations of amikacin sulfate and penicillin G sodium in synovial fluid at different time points are

Figure 1. Concentration of amikacin (A) and penicillin (B) in synovial fluid of 7 horses after distal limb perfusion with 2 g of amikacin sulfate and 10 mill IU of penicillin G sodium. Samples were collected from the metacarpophalangeal joint at 5 min and 2, 6, 12, and 24 h after the tourniquet was removed.

Horse 1Horse 2Horse 3Horse 4Horse 5Horse 6Horse 7

1000

100

10

10 6 12 18 24

Time (h)

Syn

ovia

l flu

id (a

mik

acin

) (m

g/m

L)

Horse 1Horse 2Horse 3Horse 4Horse 5Horse 6Horse 7

0 6 12 18 24

100 000

10 000

1000

100

10

1

Time (h)

Syn

ovia

l flu

id [p

enic

illin

] (ng

/mL)

Table I. Individual pharmacokinetic parameters describing the disposition kinetics of amikacin in synovial fluid after intravenous perfusion of amikacin sulfate (2 g) and penicillin sodium (10 mill IU) in 7 horses. All parameters were generated using non-compartmental analysis

Pharmacokinetic parameter Horse 1 Horse 2 Horse 3 Horse 4 Horse 5 Horse 6 Horse 7 Mean 6 SECmax (mg/mL)a 83.0 142.7 102.4 208.8 13.1 95.2 279.4 132.1 6 33.3Tmax (h)b 0.58 0.58 0.58 0.58 0.58 0.58 0.58 0.58 6 0.0AUClast (h*mg/mL)c 286.6 384.0 510.0 489.0 113.3 249.1 697.1 389.9 6 73.3T1/2 lambdaz (h)d 6.65 15.9 8.66 11.7 26.0 11.9 14.3 13.6 6 2.4a Maximum synovial concentration.b Time of maximal synovial concentration.c Area under the curve to the last time point collected.d Terminal elimination rate constant.

Table II. Average pharmacokinetic parameters describing the disposition kinetics of penicillin in synovial fluid after intravenous perfusion of penicillin G sodium (10 mill IU) and amikacin sulfate (2 g) in 7 horses. All parameters were generated using non-compartmental analysis for sparse data

Pharmacokinetic parameter Mean 6 SECmax (ng/mL)a 8474 6 5710Tmax (h)b 0.58AUClast (h*ng/mL)c 9247 6 5738T1/2 lambdaz (h)d 2.82a Maximum synovial concentration.b Time of maximal synovial concentration.c Area under the curve to the last time point collected.d Terminal elimination rate constant.

A B



presented in Figures 1A and 1B, respectively. Detectable concentra-tions of amikacin were obtained in synovial fluid of all horses at all time points. Penicillin was detected in the synovial fluid of 3 horses at 6 h, in only 1 horse at 12 h, and was not detected in any horses at 24 h. Pharmacokinetic variables are shown in Tables I and II. No clinically detectable adverse effects were identified.

D i s c u s s i o nIt was found that, while a combination of amikacin sulfate and

penicillin G sodium yield therapeutic concentrations of amikacin and penicillin in the synovial fluid, penicillin concentrations were maintained for only a short period of time (only 3/7 horses had detectable levels at 6 h and only 1 at 12 h). It was therefore concluded that IVRLP with this antibiotic combination may not be practical for treating clinical cases of sepsis.

As amikacin is a concentration-dependent antimicrobial, the rate and extent of bacterial killing is associated with high-peak concen-tration (Cmax) and MIC ratio. For the treatment of susceptible bac-teria, the Cmax:MIC ratio should be 8:1 to 10:1 to maximize its effect (29,30). It has been reported that concentrations of amikacin from 1 to 4 mg/mL and from 2 to 16 mg/mL have an MIC90 for Escherichia coli and Pseudomonas aeruginosa respectively (31). Using the Cmax:MIC ratio of 8:1 synovial concentrations obtained in this study, 2 g of ami-kacin achieved therapeutic concentrations for E. coli and P. aeruginosa. In addition, at the 24-hour time point, the mean concentration of amikacin was still above 4 mg/mL MIC, which indicates that daily IVRLP with amikacin is adequate. In addition, aminoglycosides have a post-antibiotic effect (the continued suppression of bacterial growth after limited exposure of the organism to an antibiotic). The post-antibiotic effect of amikacin against staphylococci has been reported to be 5 to 10 h for clinically achievable concentrations (32). Another study (33) reported a mean post-antibiotic effect of amikacin of 3.43 h in equine isolates of methicillin-resistant Staphylococcus aureus.

Although the optimal dose and frequency of amikacin for IVRLP require further investigation, it seems that the dose used in this study provides antimicrobial coverage for at least 24 h. Reported synovial concentrations in studies using IVRLP with amikacin are highly variable, probably due in part to different methodologies, includ-ing dose of antimicrobial, type of tournique, volume and speed of infusion, vein and joints used, and amikacin determination assay. Similar to our results, however, other studies have also found wide ranges in concentration of amikacin in synovial fluids. One of the reasons we used 2 g of amikacin in this study was in part due to these wide ranges found in previous studies. Clinicians need to be aware that, even when the mean value from some studies showed good MIC concentrations, some horses may achieve low antimicro-bial concentrations.

Some studies in horses have evaluated amikacin as an IVDLP using a methodology similar to the present study: horses standing under sedation, same amikacin dose, perfusion into the cephalic vein, tourniquet in the antebraquieium, and collecting fluid from the metacarpophalangeal joint (14,34). Maximum concentrations of amikacin in synovial fluid in those studies (277 and 50 mg/mL) compare with concentrations obtained in this study (132 mg/mL). One of these studies also found a very similar terminal elimination

rate constant [12.9 h in the study by Kelmer et al (34) and 13.6 h in the present study].

Two IV penicillin G salts are commercially available, sodium and potassium. Due to more accessible price, penicillin G potassium is more commonly used at our hospital. As a million IU of penicillin G potassium contain 1.7 mEq of potassium ion, we therefore chose penicillin G sodium for this study as well as to prevent the possibil-ity of hyperkalemia at the moment of tourniquet release. Doses of 1 to 10 mill IU of penicillin G potassium have been used in IVRLP in clinical cases in horses (6). Since no previous studies of IVRLP have been conducted with penicillin, we selected a high dose (10 mill IU) for this study because of the time-dependent characteristics of penicillin on the basis that a high dose could increase the half-life of the antibiotic.

Peak serum and synovial fluid concentration in healthy mares after an intramuscular (IM) injection of aqueous procaine penicillin G (22 000 IU/kg BW) was 1.42 mg/mL and 0.62 mg/mL, respectively (35). Furthermore, the mean concentration of penicillin in synovial fluid peaked at 4 h and decreased to 0.5 mg/mL and 0.23 mg/mL at 12 h and 24 h, respectively. The MIC of penicillin G in the horse for Streptococcus equui and Streptococcus zooepidemicus has been reported to be 0.002 to 0.08 mg/mL and 0.06 to 0.25 mg/mL for Corynebacterium pseudotuberculosis (35). In-vitro killing-curve studies of bacteria have shown that maximum killing is usually achieved at 3 to 4 times the MIC when using beta-lactam antibiotics (23).

Using the MIC from the previously mentioned studies, we achieved therapeutic levels of penicillin in the synovial fluid by using 10 mill IU in our IVRLP. Although our synovial fluid levels were higher than those achieved by parenteral administration of IM procaine penicillin, they dropped to 0.2 mg/mL by 2 h and were not detectable in 60% of our horses by 6 h. Concentrations of synovial and peritoneal penicillin parallel serum and penicillin leave the joint as readily as it enters in relation to serum concentrations (35). A previous study found that an intraperitoneal penicillin injection was absorbed and excreted less rapidly when supplemented by parenteral injections (36). It appears that intraarticular penicillin would equilibrate rapidly with serum, quickly depleting the synovial concentration because of the large volume of serum with which it is equilibrating. It is not known whether parenteral administration of penicillin in addition to the IVRLP would reduce the steep reduction of penicillin concentrations in the synovial fluid. The physiochemical property between aminoglycoside and beta-lactum has been reported (18,21). One of the horses in this study had a low concentration of amikacin and penicillin in the synovial fluid and a different horse with the lowest concentration of synovial penicillin had a very high concentration of amikacin.

A limitation of the study is that we did not collect blood to measure antibiotic concentrations before tourniquet removal to determine the effectiveness of the tourniquet. It is possible that the tourniquet failed in the horse with low concentrations of both antibiotics. The physical change of the penicillin molecule to form metabolites and decretive by-products occurs due to the exposure to plasma esterase present in the surrounding fluid. In addition, penicillins have a labile beta-lactam ring that has pronounced sus-ceptibility to various nucleophiles, acid-base reagents, metal ions, oxidizing agents, or even solvents such as water and alcohol (37).



Since we only measured penicillin and not its metabolites or break-down products, it is not possible to determine whether the penicillin degraded in the horse with high amikacin and low penicillin levels.

We collected synovial fluid by arthrocentesis from the same joint on 5 occasions within a 24-hour period. Although the fluid was clear in all the initial samples, the fluid became serosanguinous in subsequent samples from some horses. It is possible that hemorrhage or inflammation as a result of repeated arthrocentesis may have decreased the concentration of antibiotics and potentially short-ened its half-life. Although placing an in-dwelling small catheter could have prevented hemorrhage from repeated arthrocentesis, the inflammation induced by the catheter could also have affected concentrations of antimicrobials.

Although there are several studies showing positive results of synovial sepsis with the use of antimicrobials by IVRLP (1,4,6,8,10), all pharmacokinetic studies have been conducted in healthy animals (2,5,7,12,15,28,38). Septic arthritis can cause thrombosis of synovial vessels and necrosis of the synovial membrane, which may limit the delivery of systemically administered antibiotics. In addition, changes in vascular permeability induced by synovitis may increase the rate of antimicrobials entering or exiting the joint. A study of experimentally induced septic arthritis showed better outcome in horses receiving IVRLP with gentamicin than when the drug was administered intravenously (9). In that study, the perfused joints had lower nucleated cell counts and terminal bacterial cultures of synovial fluid and synovial membranes yielded negative results in 50% of joints, compared with 100% of joints treated only by systemic antimicrobials. A study of experimentally induced synovitis by lipo-polysacharide injection found a shorter Tmax and higher Cmax after IVRLP with amikacin compared to normal joints (38). Clinicians need to be aware, however, that the effect or the degree of antimicrobial diffusion and clearance in naturally inflamed and/or septic synovial structures when using IVRLP is not known.

Although we obtained acceptable levels of amikacin and penicil-lin in the synovial fluid, the levels of penicillin were present only for a short time. We used a high dose of penicillin in an effort to maintain antibiotic accumulation in the tissue outside the vascular space (depot phenomenon), but levels of antibiotics were detectable in only 40% of the horses in this study at 6 h. It is unknown if higher or long-lasting concentrations can be obtain if IVRLP is done using only penicillin as a single antibiotic. When penicillin G sodium is used by the IV route in horses, it is recommended that it be admin-istered every 6 h (39). Our results indicate that if penicillin G sodium is used as IVRLP, it may also need to be administered at similar intervals. As the use of IVRLP in horses is painful and patients must be sedated, it is impractical to use IVRLP several times a day. Since the time that penicillin is above the MIC of the organism is the best predictor of bacterial killing and clinical efficacy, then the T . MIC can be maximized by administering the drug by continuous infu-sion (23). When high regional antibiotic concentration of amikacin and a beta-lactam antibiotic are desired, an alternative would be to carry out IVRLP with amikacin and administer the time-dependent antibiotic using a continuous delivery system (40).

Since the interaction between aminoglycosides and beta-lactam antibiotics is also concentration-dependent, it is not known if simi-lar results would be obtained by using different concentrations of

antibiotics than those used in this study. Regardless of the cause of the rapid reduction of penicillin from the synovial fluid (in-vivo interaction between amikacin and penicillin or rapid diffusion of penicillin from the synovial structure to systemic circulation), it is concluded that IVRLP with the combination used in this study is not recommended in clinical cases.

A c k n o w l e d g m e n t sSupported by the University of California-Davis Comparative

Gastrointestinal Laboratory by a gift from Mick and Sabrina Hellman.

Re f e r e n c e s 1. Cruz AM, Rubio-Martinez L, Dowling T. New antimicrobials,

systemic distribution, and local methods of antimicrobial deliv-ery in horses. Vet Clin North Am Equine Pract 2006;22:297–322, vii-viii.

2. Errico JA, Trumble TN, Bueno AC, Davis JL, Brown MP. Comparison of two indirect techniques for local delivery of a high dose of an antimicrobial in the distal portion of forelimbs of horses. Am J Vet Res 2008;69:334–342.

3. Kettner NU, Parker JE, Watrous BJ. Intraosseous regional perfu-sion for treatment of septic physitis in a two-week-old foal. J Am Vet Med Assoc 2003;222:346–350, 316.

4. Lugo J, Gaughan EM. Septic arthritis, tenosynovitis, and infec-tions of hoof structures. Vet Clin North Am Equine Pract 2006; 22:363–388, viii.

5. Murphey ED, Santschi EM, Papich MG. Regional intravenous perfusion of the distal limb of horses with amikacin sulfate. J Vet Pharmacol Ther 1999;22:68–71.

6. Rubio-Martinez LM, Cruz AM. Antimicrobial regional limb perfusion in horses. J Am Vet Med Assoc 2006;228:706–712, 655.

7. Scheuch BC, Van Hoogmoed LM, Wilson WD, et al. Comparison of intraosseous or intravenous infusion for delivery of amikacin sulfate to the tibiotarsal joint of horses. Am J Vet Res 2002;63: 374–380.

8. Stewart AA, Goodrich LR, Byron CR, Evans RB, Stewart MC. Antimicrobial delivery by intrasynovial catheterisation with systemic administration for equine synovial trauma and sepsis. Aust Vet J 2010;88:115–123.

9. Whithair KJ, Bowersock TL, Blevins WE, Fessler JF, White MR, Van Sickle DC. Regional limb perfusion for antibiotic treatment of experimentally induced septic arthritis. Vet Surg 1992;21: 367–373.

10. Schneider RK, Bramlage LR, Moore RM, Mecklenburg LM, Kohn CW, Gabel AA. A retrospective study of 192 horses affected with septic arthritis/tenosynovitis. Equine Vet J 1992;24:436–442.

11. Snyder JR, Pascoe JR, Hirsh DC. Antimicrobial susceptibility of microorganisms isolated from equine orthopedic patients. Vet Surg 1987;16:197–201.

12. Alkabes SB, Adams SB, Moore GE, Alkabes KC. Comparison of two tourniquets and determination of amikacin sulfate concentrations after metacarpophalangeal joint lavage per-



formed simultaneously with intravenous regional limb perfusion in horses. Am J Vet Res 2011;72:613–619.

13. Mattson S, Boure L, Pearce S, Hurtig M, Burger J, Black W. Intraosseous gentamicin perfusion of the distal metacarpus in standing horses. Vet Surg 2004;33:180–186.

14. Sole A, Nieto JE, Aristizabal FA, Snyder JR. Effect of emptying the vasculature before performing regional limb perfusion with amikacin in horses. Equine Vet J 2015.

15. Zantingh AJ, Schwark WS, Fubini SL, Watts AE. Accumula tion of amikacin in synovial fluid after regional limb perfusion of amikacin sulfate alone and in combination with ticarcillin/clavulanate in horses. Vet Surg 2014;43:282–288.

16. Palmer SE, Hogan PM. How to perform regional limb perfusion in the standing horse. Proceedings 45th Annu Conv Ann AAoc Equine Pract Year: 124–127.

17. Santschi EM, Adams SB, Murphey ED. How to perform equine intravenous digital perfusion. 44th Annu Conv Am Assoc Equine Pract Year: 198–201.

18. Wallace SM, Chan LY. In vitro interaction of aminoglycosides with beta-lactam penicillins. Antimicrob Agents Chemother 1985;28:274–281.

19. Tindula RJ, Ambrose PJ, Harralson AF. Aminoglycoside inac-tivation by penicillins and cephalosporins and its impact on drug-level monitoring. Drug Intell Clin Pharm 1983;17:906–908.

20. Holt HA, Broughall JM, McCarthy M, Reeves DS. Interactions between aminoglycoside antibiotics and carbenicillin or ticarillin. Infection 1976;4:107–109.

21. Riff LJ, Thomason JL. Comparative aminoglycoside inactivation by beta-lactam antibiotics. Effects of a cephalosporin and six penicillins on five aminoglycosides. J Antibiot (Tokyo) 1982;35: 850–857.

22. Papich MG, Riviere JE. B-lactam antibiotics: Penicillins, cepha-losporins, and related drugs. In: Papich MG, Riviere JE, eds. Veterinary Pharmacology & Therapeutics. 9th ed. Vol. 1. United Sates: Wiley-Blackwell, 2009:865–893.

23. Turnidge JD. The pharmacodynamics of beta-lactams. Clin Infect Dis 1998;27:10–22.

24. Mattson SE, Pearce SG, Boure LP, Dobson H, Hurtig MB, Black WD. Comparison of intraosseous and intravenous infu-sion of technetium Tc 99m pertechnate in the distal portion of forelimbs in standing horses by use of scintigraphic imaging. Am J Vet Res 2005;66:1267–1272.

25. Finsterbusch A, Argaman M, Sacks T. Bone and joint perfusion with antibiotics in the treatment of experimental staphylococcal infection in rabbits. J Bone Joint Surg Am 1970;52:1424–1432.

26. Moyer W, Schumacher J. A guide to equine joint injection and regional anesthesia. Yardley, OA: Veterinary Learning Systems, 2007.

27. Uboh CE, Soma LR, Luo Y, et al. Pharmacokinetics of penicillin G procaine versus penicillin G potassium and procaine hydrochlo-ride in horses. Am J Vet Res 2000;61:811–815.

28. Pinto N, Schumacher J, Taintor J, Degraves F, Duran S, Boothe D. Pharmacokinetics of amikacin in plasma and selected body fluids of healthy horses after a single intravenous dose. Equine Vet J 2011;43:112–116.

29. Lacy MK, Nicolau DP, Nightingale CH, Quintiliani R. The pharmacodynamics of aminoglycosides. Clin Infect Dis 1998;27: 23–27.

30. Moore RD, Lietman PS, Smith CR. Clinical response to amino-glycoside therapy: Importance of the ratio of peak concentration to minimal inhibitory concentration. J Infect Dis 1987;155:93–99.

31. Zhanel GG, DeCorby M, Nichol KA, et al. Antimicrobial sus-ceptibility of 3931 organisms isolated from intensive care units in Canada: Canadian National Intensive Care Unit Study, 2005/2006. Diagn Microbiol Infect Dis 2008;62:67–80.

32. Isaksson B, Maller R, Nilsson LE, Nilsson M. Postantibiotic effect of aminoglycosides on staphylococci. J Antimicrob Chemother 1993;32:215–222.

33. Caron JP, Bolin CA, Hauptman JG, Johnston KA. Minimum inhibitory concentration and postantibiotic effect of amikacin for equine isolates of methicillin-resistant Staphylococcus aureus in vitro. Vet Surg 2009;38:664–669.

34. Kelmer G, Bell GC, Martin-Jimenez T, et al. Evaluation of regional limb perfusion with amikacin using the saphenous, cephalic, and palmar digital veins in standing horses. J Vet Pharmacol Ther 2013;36:236–240.

35. Stover SM, Brown MP, Kelly RH, Farver TB, Knight HD. Aqueous procaine penicillin G in the horse: Serum, synovial, peritoneal, and urine concentrations after single-dose intramus-cular administration. Am J Vet Res 1981;42:629–631.

36. Cooke JV, Goldering D. The concentrations of penicillin in various body fluids during penicillin therapy. J Am Med Assoc 1945;127:80–87.

37. Deshpande AD, Baheti KG, Chatterjee NR. Degradation of Beta-lactam antibiotics. Current Science 2004;87:1684–1695.

38. Beccar-Varela AM, Epstein KL, White CL. Effect of experimentally induced synovitis on amikacin concentrations after intravenous regional limb perfusion. Vet Surg 2011;40:891–897.

39. Knottenbelt DC. Saunders Equine Frormulary. The Netherlands: Elsevier 2006.

40. Lescun TB, Vasey JR, Ward MP, Adams SB. Treatment with continuous intrasynovial antimicrobial infusion for septic syno-vitis in horses: 31 cases (2000–2003). J Am Vet Med Assoc 2006; 228:1922–1929.


Article


I n t r o d u c t i o nAcute pancreatitis is an inflammatory process of the pancreas that

frequently involves peripancreatic tissues and remote organ sys-tems (1) and has high morbidity and mortality rates in both human and veterinary patients. The mortality rate for humans with acute pancreatitis has been reported as just under 10% (2,3), but in severe cases it is as high as 20% to 30% (4,5). The mortality rate for dogs with acute pancreatitis ranges from 27% to 58% (6–8).

The pathophysiological process of acute pancreatitis consists of activation of pancreatic enzymes within acinar cells, release of these

enzymes into the interstitium, autodigestion of the pancreas, and release of the enzymes and other factors into the circulation, which results in multiple organ dysfunction (9–13).

Dogs with acute pancreatitis generally present with a sudden onset of anorexia, depression, abdominal pain, and vomiting (14). However, the findings on clinical examination vary considerably with the severity and stage of the pancreatitis and the degrees of associated dehydration and shock (8). Mild acute pancreatitis does not cause multisystem organ failure or a complicated recovery, whereas severe acute pancreatitis causes multisystem organ fail-ure or development of severe complications (1). The severity of

Changes in gene expression of tumor necrosis factor alpha and interleukin 6 in a canine model of caerulein-induced pancreatitis

Ruhui Song, Dohyeon Yu, Jinho Park

A b s t r a c tAcute pancreatitis is an inflammatory process that frequently involves peripancreatic tissues and remote organ systems. It has high morbidity and mortality rates in both human and veterinary patients. The severity of pancreatitis is generally determined by events that occur after acinar cell injury in the pancreas, resulting in elevated levels of various proinflammatory mediators, such as interleukin (IL) 1b and 6, as well as tumor necrosis factor alpha (TNF-a). When these mediators are excessively released into the systemic circulation, severe pancreatitis occurs with systemic complications. This pathophysiological process is similar to that of sepsis; thus, there are many striking clinical similarities between patients with septic shock and those with severe acute pancreatitis. We induced acute pancreatitis using caerulein in dogs and measured the change in the gene expression of proinflammatory cytokines. The levels of TNF-a and IL-6 mRNA peaked at 3 h, at twice the baseline levels, and the serum concentrations of amylase and lipase also increased. Histopathological examination revealed severe hyperemia of the pancreas and hyperemia in the duodenal villi and the hepatic sinusoid. Thus, pancreatitis can be considered an appropriate model to better understand the development of naturally occurring sepsis and to assist in the effective treatment and management of septic patients.

R é s u m éLa pancréatite aigüe est un processus inflammatoire qui implique fréquemment les tissus péri-pancréatiques et des systèmes organiques éloignés. Elle a des taux de morbidité et de mortalité élevés autant chez les humains que chez les animaux. La sévérité de la pancréatite est généralement déterminée par des évènements qui se produisent suite à des dommages aux cellules acinaires dans le pancréas, et qui induisent des niveaux élevés de différents médiateurs pro-inflammatoires, tels que l’interleukine (IL) 1b et 6, ainsi que le facteur nécrosant des tumeurs alpha (TNFa). Lorsque ces médiateurs sont libérés de manière excessive dans la circulation systémique, une pancréatite sévère se produit avec des complications systémiques. Ce processus pathophysiologique est similaire à celui d’un sepsis; donc, il y a plusieurs similarités cliniques entre des patients avec un choc septique et ceux avec une pancréatite aigüe sévère. Nous avons induit une pancréatite aigüe en utilisant de la caeruléine chez des chiens et avons mesuré le changement dans l’expression des gènes des cytokines pro-inflammatoires. Les niveaux d’ARNm de TNFa et d’IL-6 ont culminé après 3 h, atteignant le double des niveaux de base, et les concentrations sériques d’amylase et de lipase augmentèrent également. Un examen histopathologique a révélé une hyperémie sévère du pancréas et une hyperémie dans les villosités duodénales et les sinusoïdes hépatiques. Ainsi, la pancréatite peut être considérée un modèle approprié pour mieux comprendre le développement d’un sepsis naturel et aider dans le traitement efficace et la gestion de patients septiques.


Department of Veterinary Internal Medicine, College of Veterinary Medicine, Chonbuk National University, Jeonju, Jeonbuk 54896, Korea (Song, Park); Department of Veterinary Laboratory Medicine, College of Veterinary Medicine, Chonnam National University, Gwangju, 61186, Korea (Yu).

Ruhui Song and Dohyeon Yu contributed equally to this work.

Address all correspondence to Dr. Jinho Park; telephone: 182-63-850-0949; fax: 182-63-850-0910; e-mail: [email protected]

Received August 28, 2015. Accepted March 1, 2016.



pancreatitis is generally determined by the events that occur after acinar cell injury, when various proinflammatory mediators, such as interleukin (IL) 1b and 6, as well as tumor necrosis factor alpha (TNF-a), are produced (15,16). When these mediators are excessively released into the systemic circulation, severe pancreatitis occurs, with systemic complications. This pathophysiological process is similar to that of sepsis; thus, there are many striking clinical similarities between patients with septic shock and those with severe acute pancreatitis (17–19).

In the present study, we used caerulein to induce acute pancre-atitis in dogs. We examined the pancreas and adjacent organs histo-pathologically, measured the serum amylase and lipase levels and the gene expression of proinflammatory mediators, and evaluated the suitability of this pancreatitis model as a model of septic shock.


AnimalsSix healthy adult beagles weighing 7 to 8 kg each were hospital-

ized and fasted for 12 h before the study; they were provided with water during the fasting period. Eight hours after the first infusion of caerulein the dogs were fed a commercial diet and provided with tap water. The study was approved by the Committee on Bioethics of Chonbuk National University, Jeonju, Korea.

Induction of acute pancreatitisFour dogs received caerulein (Sigma–Aldrich, St. Louis, Missouri,

USA) twice intravenously at a dose of 10 mg/kg body weight (BW) (20,21), with a 1-hour interval between infusions. The other 2 dogs were used as controls and were given 2 intravenous infusions, 1 h apart, of normal saline at a dose of 1 mL/kg BW. Body temperature, pulse rate, and respiratory rate were measured before the start of the injections and at 3, 6, 12, 24, and 48 h after the start. The dogs were euthanized at the end of the examination period, and necropsy was done immediately.

Figure 1. Changes in serum amylase (a) and lipase (b) concentrations in 4 dogs exposed to caerulein (black bars) and 2 control dogs (grey bars). Data presented as mean 6 standard deviation.

Time (h)

Amylase

Time (h)

Lipase

Mea

n va

lue

(U/L

)

Mea

n va

lue

(U/L

)

7000

6000

5000

4000

3000

2000

1000

0

2000

1800

1600

1400

1200

1000

800

600

400

200

0 0 3 6 12 24 48 0 3 6 12 24 48

Figure 2. Gross features of the canine abdominal cavity. Redness of the pancreas (white arrows) represents caerulein-induced hyperemia.

Figure 3. Severe hyperemia of the pancreas in a dog exposed to caerulein. Necrosis of the pancreatic acinar cells was not observed. Hematoxylin and eosin (H&E); 3100.



Sample collectionBlood (6 mL per dog divided into each of the 2 types of tube) was

collected from the jugular vein into tubes treated with potassium ethylene diamine tetraacetic acid and plastic vacuum-filled tubes at time 0 (baseline; before the first infusion) and 3, 6, 12, 24, and 48 h after the first infusion. Serum was immediately separated by centrifugation of the blood at 875 3 g for 5 min and kept frozen at −70°C until needed.

Hematologic and biochemical analysesA complete blood (cell) count was done at each time point by

means of an automatic impedance cell counter. The serum levels of alkaline phosphatase, alanine transaminase, amylase, lipase, choles-terol, urea nitrogen, creatinine, glucose, phosphate, bilirubin, total protein, and albumin were measured.

Separation of canine peripheral blood mononuclear cells (PBMCs)

Centrifugation with Histopaque 1077 (Sigma–Aldrich) at 450 3 g for 45 min was used to isolate the PBMCs. Erythrocytes were lysed with an 83% ammonium chloride solution (pH 7.2) and washed twice with phosphate-buffered saline. Cells were counted manually with use of the trypan blue viability test. Cell purity was determined by a conventional Diff-Quik method after Cytospin centrifugation (Shandon Cytospin, Thermo Scientific, Waltham, Massachusetts, USA) at 250 3 g for 5 min. Viability was always more than 90%, and PBMCs constituted 95% to 98% of the cells.

Quantification of cytokine gene expressionIsolated PBMCs were immediately placed in Buffer RLT (Qiagen,

Hilden, Germany) and stored at −70°C. Total RNA was isolated with use of the RNeasy Mini Kit (Qiagen) with Qiacube (Qiagen). All RNA samples were treated with RQ1 RNase-free DNase (Promega, Madison, Wisconsin, USA) to remove any genomic DNA. The RNA concentration and purity of all samples were measured with a spec-

trophotometer (Epoch; Bio-Tek Instruments, Winooski, Vermont, USA). A 260:280 nm absorbance ratio of 1.8:2.0 was regarded as indicating pure RNA. Using the manufacturer’s protocol, we syn-thesized cDNA with the ImProm-II Reverse Transcription System (Promega) and stored it at −20°C until needed.

Real-time polymerase chain reaction (RT-PCR) was done with gene-specific primers for canine TNF-a and IL-6. The 20-mL reac-tion solution contained 300 nM of each primer, 1 mL of cDNA, and 10 mL of 23 iQ SYBR Green Supermix (Bio-Rad, Hercules, California, USA). The CFX384 RT-PCR detection system (Bio-Rad) was used to quantify cytokine mRNA in the PBMCs. Samples were heated at 95°C for 3 min for i-Taq DNA polymerase activation, and then they underwent 40 cycles of denaturation at 95°C for 15 s, annealing for 15 s, and extension at 72°C for 15 s. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a housekeeping gene. All samples, controls, and standards were run in duplicate. Data collec-tion and analysis were done with CFX Manager Software, version 1.0 (Bio-Rad). Relative quantification was analyzed by the 2−DDCt method. Sample values were averaged and calculated in relation to the quan-tity of GAPDH. These normalized values were used to calculate the expression of a given sequence relative to the control.

Statistical analysisThe results are expressed as mean 6 standard deviation (SD) or

mean 6 standard error (SE). The significance of differences was evaluated by means of Student’s t-test. A P-value of less than 0.05 was accepted as statistically significant. The statistical analyses were done with the use of SPSS software, version 18.0 (SPSS, Chicago, Illinois, USA).

Re s u l t sWithin 1 h after the first dose of caerulein the dogs exhibited slight

weakness and began having diarrhea. However, vomiting was not observed during the experiment. There were no significant changes in body temperature, heart rate, respiratory rate, or hematologic

Figure 4. Severe hyperemia in the blood vessels of the duodenal villi in a dog exposed to caerulein. H&E; 3100.

Figure 5. Hyperemia in the hepatic sinusoid in a dog exposed to caeru-lein. H&E; 3100.



values within the study group. Of the various serum enzymes examined, only amylase and lipase, which are commonly used as markers of acute pancreatitis (8,22), showed significant changes over time (Figure 1). Canine pancreatic lipase immunoreactivity was measured at 3 h after the first dose of caerulein, with a commercial kit, and positive results were observed.

On gross examination of the organs from the caerulein group, we observed redness of the pancreas and hemorrhagic inflammation of the small intestine (Figure 2). Histologic examination of the pan-creas revealed severe hyperemia (Figure 3). Severe hyperemia was also found in the duodenal villi and the hepatic sinusoid (Figures 4 and 5, respectively).

Caerulein-induced changes in cytokine mRNA abundance are shown in Figure 6. The levels of TNF-a (P , 0.05) and IL-6 mRNA peaked at 3 h, at twice the baseline levels, and then rapidly declined by 6 h.

D i s c u s s i o nCaerulein is a 10-peptide molecule that structurally resembles gas-

trin and the C-terminal octapeptide of cholecystokinin. It has vari-ous biologic functions, such as stimulating gallbladder contraction, gastric acid secretion, pancreatic enzyme secretion, and hepatic bile flow in a number of species, including humans (23–26). Caerulein can stimulate pancreatic acinar cells to excrete a large amount of digestive enzymes and pancreatic fluid, resulting in mild edematous pancreatitis characterized by a high serum amylase concentration, interstitial edema, leukocyte infiltration, and vacuolation of the acinar cells (20,27,28). It has been used successfully to induce acute pancreatitis in various animals (11,21,29–32).

The dogs in the present study that were treated with caerulein showed distinct signs of mild acute pancreatitis, including higher serum concentrations of amylase and lipase compared with the con-trol group at 3 and 6 h after the first infusion of caerulein. These are the factors primarily used to diagnose pancreatitis, so we assumed that caerulein damaged the canine pancreas through increases in these serum amylase and lipase levels. In addition, microscopic

examination showed severe hyperemia in the pancreas and adjacent organs (duodenum and liver) of the dogs treated with caerulein, indicating significant acute pancreatitis and inflammation. Also, the gene expression of TNF-a and IL-6 peaked 3 h after the first infusion of caerulein, which is similar to the results of our previous endotoxemia experiment (33).

It is unknown how the initial insult to the pancreas generates the inflammatory reaction, but it has been proposed that the pancreatic tissue is capable of producing a range of proinflammatory cytokines. The pancreas can produce large quantities of kinins, which might also play a key role in the inflammatory cascade (34). In a number of human medical and experimental studies (35–37) the concentra-tion of TNF has been elevated in severe acute pancreatitis, which correlates with the outcome of our study. Also, the IL-6 level is of major prognostic significance in human acute pancreatitis (10). In addition, IL-6 plays an important role is the induction of hepatic synthesis of acute-phase proteins such as C-reactive protein (CRP) (38,39), and there is a close relationship between production and serum concentrations of IL-6 and CRP (40). In particular, IL-6 is a very sensitive predictor of the severity of illness 24 h after the onset of acute pancreatitis (41). However, there have been only a few stud-ies of these cytokines in canine acute pancreatitis.

There are many striking clinical and physiological similarities between patients with septic shock and those with severe acute pancreatitis (17–19). The complications that develop in the most critically ill patients are very similar, and evidence suggests that the proinflammatory cascade is activated in the same way in each of these patient groups (42). In addition, according to the changing patterns of TNF-a and IL-6 in this study, it is possible that this pan-creatitis model could be used as a model of septic shock. However, in future studies it will be necessary to supplement the caerulein-induction model since the acute pancreatitis was mild in this study. Ding, Li, and Jin (20) induced pancreatitis in mice with a combina-tion of caerulein and lipopolysaccharide (LPS). The pancreas was so severely damaged that it resulted in an inflammatory reaction in the entire body and systemic organ dysfunction. Ding, Li, and Jin (20) also reported that the model was almost stable when the

Figure 6. Gene expression of tumor necrosis factor alpha (TNF-a) (a) and interleukin (IL)-6 (b) in the 4 dogs exposed to caerulein (triangles) and the 2 control dogs (circles). The mRNA expression peaked 3 h after the first infusion of caerulein. Data presented as mean 6 standard error. *P , 0.05 compared with the control mean.

Time (h) Time (h)

TN

F-a

mR

NA

exp

ress

ion

IL-6

mR

NA

exp

ress

ion

3.0

2.5

2.0

1.5

1.0

0.5

0.0

3.0

2.5

2.0

1.5

1.0

0.5

0.0 0 3 6 12 24 48 0 3 6 12 24 48



experiment was duplicated. Furthermore, the pancreatic injury was much more severe, with a predominance of necrosis, after prolonged and sustained exposure to caerulein. According to Jacob et al (43) caerulein administration for 4 h in mice caused acute pancreatitis with apoptosis and significantly milder systemic injury than 8 injec-tions of caerulein. We demonstrated that caerulein injection resulted in mild acute pancreatitis in dogs and confirmed that the changes in several proinflammatory cytokines are similar to those in sepsis (21,44). Further studies are needed to evaluate the cause of more severe pancreatitis in dogs by modifying the method of inducing pancreatitis. None-the-less, our results aid in the development of pancreatitis models that can be used to further study sepsis in dogs.

A c k n o w l e d g m e n tThis research was supported by the Basic Science Research

Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education (grant NRF-2013R1A1A2057479).

Re f e r e n c e s 1. Windsor JA, Hammodat H. Metabolic management of severe

acute pancreatitis. World J Surg 2000;24:664–672. 2. Kim YT. [Medical management of acute pancreatitis and com-

plications]. Korean J Gastroenterol 2005;46:339–344. 3. Larvin M. Circulating mediators in acute pancreatitis as predic-

tors of severity. Scand J Gastroenterol Suppl 1996;219:16–19. 4. Banks PA. Modern concepts in pancreatitis. Mt Sinai J Med 1993;

60:170–174. 5. Rau B, Uhl W, Buchler MW, Beger HG. Surgical treatment of

infected necrosis. World J Surg 1997;21:155–161. 6. Maxie MG, Jubb KVF. Pathology of Domestic Animals. 5th ed.

Edinburgh, UK: Elsevier Saunders, 2007:389–408. 7. Ruaux CG, Atwell R. General practice attitudes to the treat-

ment of spontaneous canine acute pancreatitis. Aust Vet Pract 1998;28:67–74.

8. Mansfield C. Acute pancreatitis in dogs: Advances in under-standing, diagnostics, and treatment. Top Companion Anim Med 2012;27:123–132. e-pub 2012 May 30.

9. Steer ML, Meldolesi J. Pathogenesis of acute pancreatitis. Annu Rev Med 1988;39:95–105.

10. Frossard JL, Hadengue A, Pastor CM. New serum markers for the detection of severe acute pancreatitis in humans. Am J Respir Crit Care Med 2001;164:162–170.

11. Su KH, Cuthbertson C, Christophi C. Review of experimental animal models of acute pancreatitis. HPB (Oxford) 2006;8: 264–286.

12. Yadav D, Agarwal N, Pitchumoni CS. A critical evaluation of laboratory tests in acute pancreatitis. Am J Gastroenterol 2002;97:1309–1318.

13. Ruben DS, Scorpio DG, Gabrielson KL, Simon BW, Buscaglia JM. Refinement of canine pancreatitis model: Inducing pancreatitis by using endoscopic retrograde cholangiopancreatography. Comp Med 2009;59:78–82.

14. Hess RS, Saunders HM, Van Winkle TJ, Shofer FS, Washabau RJ. Clinical, clinicopathologic, radiographic, and ultrasonographic

abnormalities in dogs with fatal acute pancreatitis: 70 cases (1986–1995). J Am Vet Med Assoc 1998;213:665–670.

15. Norman J. The role of cytokines in the pathogenesis of acute pancreatitis. Am J Surg 1998;175:76–83.

16. Lane JS, Todd KE, Gloor B, et al. Platelet activating factor antago-nism reduces the systemic inflammatory response in a murine model of acute pancreatitis. J Surg Res 2001;99:365–370.

17. Beger HG, Bittner R, Buchler M, Hess W, Schmitz JE. Hemo-dynamic data pattern in patients with acute pancreatitis. Gastroenterology 1986;90:74–79.

18. Ito K, Ramirez-Schon G, Shah PM, Agarwal N, Delguercio LR, Reynolds BM. Myocardial function in acute pancreatitis. Ann Surg 1981;194:85–88.

19. Shoemaker WC, Appel PL, Kram HB, Bishop MH, Abraham E. Temporal hemodynamic and oxygen transport patterns in medi-cal patients. Septic shock. Chest 1993;104:1529–1536.

20. Ding SP, Li JC, Jin C. A mouse model of severe acute pancre-atitis induced with caerulein and lipopolysaccharide. World J Gastroenterol 2003;9:584–589.

21. Renner IG, Wisner JR, Jr. Ceruletide-induced acute pancreatitis in the dog and its amelioration by exogenous secretin. Int J Pancreatol 1986;1:39–49.

22. Van den Bossche I, Paepe D, Daminet S. Acute pancreatitis in dogs and cats: Pathogenesis, clinical signs and clinicopathologic findings. Vlaams Diergeneeskd Tijdschr 2010;79:13–22.

23. Erspamer V, Bertaccini G, De Caro G, Endean R, Impicciatore M. Pharmacological actions of caerulein. Experientia 1967;23:702–703.

24. Bertaccini G, Braibanti T, Uva F. Cholecystokinetic activity of the new peptide caerulein in man. Gastroenterology 1969;56:862–867.

25. Bertaccini G, De Caro G, Endean R, Erspamer V, Impicciatore M. The action of caerulein on pancreatic secretion of the dog and biliary secretion of the dog and the rat. Br J Pharmacol 1969;37:185–197.

26. Erspamer V. Progress report: Caerulein. Gut 1970;11:79–87.27. Frossard JL, Bhagat L, Lee HS, et al. Both thermal and non- thermal

stress protect against caerulein induced pancreatitis and prevent trypsinogen activation in the pancreas. Gut 2002;50:78–83.

28. Wagner AC, Mazzucchelli L, Miller M, Camoratto AM, Göke B. CEP-1347 inhibits caerulein-induced rat pancreatic JNK acti-vation and ameliorates caerulein pancreatitis. Am J Physiol Gastrointest Liver Physiol 2000;278:G165–172.

29. Lampel M, Kern HF. Acute interstitial pancreatitis in the rat induced by excessive doses of a pancreatic secretagogue. Virchows Archiv A Pathol Anat Histol 1977;373:97–117.

30. Watanabe O, Baccino FM, Steer ML, Meldolesi J. Supramaximal caerulein stimulation and ultrastructure of rat pancreatic aci-nar cell: Early morphological changes during development of experimental pancreatitis. Am J Physiol 1984;246:G457–G467.

31. Niederau C, Ferrell LD, Grendell JH. Caerulein-induced acute necrotizing pancreatitis in mice: Protective effects of proglumide, benzotript, and secretin. Gastroenterology 1985;88:1192–1204.

32. Chan YC, Leung PS. Acute pancreatitis: Animal models and recent advances in basic research. Pancreas 2007;34:1–14.

33. Song R, Kim J, Yu D, Park C, Park J. Kinetics of IL-6 and TNF-alpha changes in a canine model of sepsis induced by endotoxin. Vet Immunol Immunopathol 2012;146:143–149.



34. Orstavik TB, Brandtzaeg P, Nustad K, Pierce JV. Immunohis-tochemical localization of kallikrein in human pancreas and salivary glands. J Histochem Cytochem 1980;28:557–562.

35. Exley AR, Leese T, Holliday MP, Swann RA, Cohen J. Endotox-aemia and serum tumour necrosis factor as prognostic markers in severe acute pancreatitis. Gut 1992;33:1126–1128.

36. Grewal HP, Kotb M, el Din AM, et al. Induction of tumor necrosis factor in severe acute pancreatitis and its subsequent reduction after hepatic passage. Surgery 1994;115:213–221.

37. Norman JG, Franz MG, Fink GS, et al. Decreased mortality of severe acute pancreatitis after proximal cytokine blockade. Ann Surg 1995;221:625–631; discussion 631–624.

38. Mayer AD, McMahon MJ, Bowen M, Cooper EH. C reactive pro-tein: An aid to assessment and monitoring of acute pancreatitis. J Clin Pathol 1984;37:207–211.

39. Wilson C, Heads A, Shenkin A, Imrie CW. C-reactive protein, antiproteases and complement factors as objective markers of severity in acute pancreatitis. Br J Surg 1989;76:177–181.

40. Leser HG, Gross V, Scheibenbogen C, et al. Elevation of serum interleukin-6 concentration precedes acute-phase response and reflects severity in acute pancreatitis. Gastroenterology 1991;101:782–785.

41. Heath DI, Cruickshank A, Gudgeon M, Jehanli A, Shenkin A, Imrie CW. Role of interleukin-6 in mediating the acute phase protein response and potential as an early means of severity assessment in acute pancreatitis. Gut 1993;34:41–45.

42. Wilson PG, Manji M, Neoptolemos JP. Acute pancreatitis as a model of sepsis. J Antimicrob Chemother 1998;41 Suppl A:51–63.

43. Jacob TG, Raghav R, Kumar A, Garg PK, Roy TS. Duration of injury correlates with necrosis in caerulein-induced experimental acute pancreatitis: Implications for pathophysiology. Int J Exp Pathol 2014;95:199–208.

44. Jung WS, Chae YS, Kim DY, et al. Gardenia jasminoides protects against cerulein-induced acute pancreatitis. World J Gastroenterol 2008;14:6188–6194.


Short Communication Communication brève


Despite several investigations, the role of type A Clostridium per-fringens in enterocolitis in young foals is still unclear (1–9). A novel toxin, NetF, in C. perfringens has been identified recently in type A strains that cause necrotizing enterocolitis in neonatal foals (10). However, its prevalence in foals is not well characterized (10). The purpose of this study was to determine the presence and prevalence of netF-positive C. perfringens in healthy foals in Ontario.

Fecal samples from foals at different ages were obtained from 8 different horse-breeding farms (2 Standardbred, 6 Thoroughbred) across southwestern Ontario. Fecal samples were collected at serial intervals from the rectum of foals starting at birth until the foal was aged 4 mo. The collected fecal samples were stored at 270°C until processing. A total of 88 foals were tested for C. perfringens, compris-ing 5 to 15 foals from each farm at the first collection, repeated again if a fecal sample (n = 49) was available for testing. The foals sampled were aged , 1 wk (n = 24), 1 to 2 wk (n = 31), 2 to 4 wk (n = 19), 1 to 2 mo (n = 47), and 2 to 4 mo (n = 14). Sampling numbers ranged from 5 to 15 foals per farm, median 11. Six of the 137 fecal samples cultured were from foals recorded as diarrheic; the remainder were from foals with formed feces.

To isolate C. perfringens, 0.5 g of thawed foal feces was diluted in 4.5 mL of phosphate-buffered saline (PBS) and 10-fold serial dilutions (101 to 105) were prepared. One hundred microliters (100 mL) of each dilution was spread onto separate Shahadi Ferguson Perfringens (SFP) agar plates containing 10% egg yolk (Difco, Sparks, Maryland, USA). The plates were incubated overnight at 37°C in an anaerobic chamber. The typical morphology of C. perfringens on SFP contain-ing egg yolk is a colony with a black center with a surrounding opaque zone of lecithinase activity. The bacterial colonies with this appearance were quantified from each foal fecal sample. For each positive sample, 5 individual C. perfringens colonies were isolated and streaked out onto separate blood agar plates [Trypticase Soy agar (Difco), with 5% sheep blood], incubated at 37°C in an anaerobic chamber for 24 h, and confirmed to have double zone of hemolysis typical of C. perfringens as well as being Gram-positive large short rods typical of C. perfringens. The DNA was extracted using a matrix (InstaGene Matrix; Bio-Rad Laboratories, Mississauga, Ontario) according to the manufacturer’s protocol.

A multiplex polymerase chain reaction (MPCR) was developed in this study using specific primers for the cytotoxic genes cpa

Prevalence of netF-positive Clostridium perfringens in foals in southwestern Ontario

Abigail Finley, Iman Mehdizadeh Gohari, Valeria R. Parreira, Miranda Abrahams, Henry R. Staempfli, John F. Prescott

A b s t r a c tNetF-producing Clostridium perfringens have recently been identified as a cause of necrotizing enteritis in neonatal foals, but little is known about its prevalence in clinically normal foals. Foals (n = 88) ranging in age from , 1 wk to 2 to 4 mo (median age 2 to 4 wk) on 8 horse-breeding farms in Ontario were examined on 1 or 2 occasions for the presence of C. perfringens. Of the foals that tested positive, 5 isolates (n = 675) were examined for the netF and enterotoxin (cpe) genes. Colonization by C. perfringens was most marked in foals , 1 wk of age [4.85 6 2.70 log10 colony-forming units (CFU)] and declined markedly over time (1.23 6 1.06 log10 CFU at 1 to 2 mo of age). Only 2 isolates possessed the cpe gene and none possessed netF. We concluded that netF-positive C. perfringens does not colonize young foals with any detectable frequency in Ontario and this organism is not likely to be adapted to the intestine of the horse.

R é s u m éLes isolats de Clostridium perfringens producteurs de NetF ont récemment été identifiés comme une cause d’entérite nécrotique chez les poulains nouveau-nés, mais peu de choses sont connues sur leur prévalence chez des poulains cliniquement normaux. Des poulains (n = 88) variant en âge entre , 1 semaine jusqu’à 2 à 4 mois (âge médian 2 à 4 semaines) provenant de 8 fermes d’élevage en Ontario ont été examinés à 1 ou 2 occasions pour la présence de C. perfringens. Des poulains qui se sont avérés positifs, 5 isolats (n = 675) ont été examinés pour la présence des gènes netF et de l’entérotoxine (cpe). La colonisation par C. perfringens était la plus marquée chez les poulains âgés de , 1 semaine [4,85 6 2,70 log10 unités formatrices de colonies (UFC)] et diminuait de façon marquée en fonction du temps (1,23 6 1,06 log10 UFC à 1 à 2 mois d’âge). Uniquement deux isolats possédaient le gène cpe et aucun ne possédait le gène netF. Nous avons conclu que les isolats de C. perfringens net-positif ne colonisent pas les jeunes poulains avec une fréquence détectable en Ontario et que ce microorganisme est peu susceptible de s’adapter à l’intestin du cheval.


Department of Pathobiology (Finley, Mehdizadeh Gohari, Parreira, Prescott), Department of Clinical Studies, University of Guelph, Guelph, Ontario (Abrahams, Staempfli).

Address all correspondence to Dr. John F. Prescott; telephone: (519) 824-4120, ext. 54716; fax: (519) 824-5939; e-mail: [email protected]

Received October 1, 2015. Accepted January 18, 2016.



(alpha toxin), cpe (enterotoxin), and netF (necrotic enteritis toxin F). Primers for these genes have been described previously (10). The PCR amplifications were done in a 25 mL total volume containing the following: 5 mL of template DNA; 13 PCR buffer (New England BioLabs, Pickering, Ontario); 3.5 mM of MgCl2; 0.2 mM deoxynucleo-tide triphosphate mixture; 2.5 units of TaqDNA polymerase; and 12.5 mM of each primer. The PCR program for the MPCR was: 94°C for 3 min, 30 cycles of 94°C for 30 s, 45°C for 30 s, extension at 72°C for 1 min, and finally, 72°C for 5 min. A 100 bp molecular size ladder (GeneDirex; Froggabio, North York, Ontario) and a positive control isolate (JP838) that possessed these 3 genes was run with each PCR.

Foals aged , 1 wk had the largest numbers of C. perfringens in their feces, which progressively declined with increasing age (Figure 1). The limit of detection was 100 CFU/g of feces. Foals on all 8 farms were found to be positive for C. perfringens. The multiplex PCR developed for this study reliably identified the presence of cpa, cpe, and netF (Figure 2); the presence of cpa was used to confirm that the colonies recovered on SFP media were C. perfringens. A total of 675 cpa-positive isolates from the 135 samples were examined by PCR; only 2 were positive for cpe and none were positive for netF.

Necrotizing enteritis in foals caused by netF-positive C. perfringens is most commonly a disease of neonatal foals (10), but little is known about its prevalence. Mehdizadeh Gohari et al (10) identified this type in 6.8% of 58 adult horses with undifferentiated severe entero-colitis but in none of the 11 foals with undifferentiated diarrheal illness. The source of the organism may be environmental since it has also been strongly associated with severe canine hemorrhagic gastroenteritis (10). Because netF-associated necrotizing enteritis disease is associated with neonatal foals, we examined its prevalence in the feces of young foals collected serially over time but we failed to identify any isolates among 675 isolates from the 88 foals.

One study of fecal shedding by clinically normal foals also identi-fied the higher prevalence of C. perfringens in young (3-day-old) foals

than in older (1- to 2-month-old) foals (6), and our study confirms and expands this norm (Figure 1). The pattern of greater colonization in younger animals and marked decline with age is well-recognized (11,12). Earlier studies (4,6) of C. perfringens in feces of clinically normal foals also identified the rare presence (0% or 2.1% of isolates) of the enterotoxin gene, cpe, that was also observed in this study. A

Figure 1. The mean numbers [colony-forming units (CFU) with standard deviation] of Clostridium perfringens in the feces of each foal age grouping in foals on horse-breeding farms in south-western Ontario.

Age of foals

Ave

rage

log 10

CFU

/g

8

7

6

5

4

3

2

1

0

21

� , 1 week 1–2 weeks 2–4 weeks 1–2 months 2–4 months

Figure 2. Multiplex polymerase chain reaction (PCR) developed for simulta-neous detection of cpa, cpe, and netF. Lane 1 — molecular size ladder; lane 2 — negative control; lane 3 — positive control (JP 838). The largest band in lane 3 is netF, the middle band is cpa, and the smallest band is cpe.



discrepancy has been noted between the frequent detection of CPE in the feces of foals (4,13) and the uncommon presence of the cpe gene in isolates (4).

Most studies on the relationship between C. perfringens and foals have focused on diarrheal illness (1,2,7–9), including the well- recognized association with type C infection (2). Management practices associated with the occurrence of enterocolitis attributed to C. perfringens have been identified (5). Isolates of C. perfringens carrying the cpe gene are more common in diarrheic than in clinically normal foals (4), but these have not been examined for the netF gene that is always found in netF-positive isolates (10). The 2 cpe-positive isolates identified in this study were both negative for netF.

In conclusion, netF-positive C. perfringens were not found in young foals with any detectable frequency in Ontario and this organism is not likely to be adapted to the intestine of the horse. It seems unlikely that this type of C. perfringens is adapted to the intestine of the horse. It will be useful to confirm this conclusion by investigating foals in other geographic areas.

A c k n o w l e d g m e n t sThe authors gratefully acknowledge funding from the Gryphon-

LAAIR of the Ontario Ministry of Agriculture, Food and Rural Affairs, and from the Natural Sciences and Engineering Research Council of Canada. Abigail Finley was a participant in the Summer Leadership and Research Program of the Ontario Veterinary College.

Re f e r e n c e s1. Kanoe M, Inoue S, Abe T, et al. Isolation of Clostridium perfringens

from foals. Microbios 1990;64:153–158.2. East LM, Savage CJ, Traub-Dargatz JL, et al. Enterocolitis asso-

ciated with Clostridium perfringens infection in neonatal foals: 54 cases (1988–1997). J Am Vet Med Assoc 1998;212:1751–1756.

3. Netherwood T, Wood JLN, Mumford JA, et al. Molecular analysis of the virulence determinants of Clostridium perfringens associ-ated with foal diarrhea. Vet J 1998;155:289–294.

4. Netherwood T, Binns N, Townsend H, et al. The Clostridium perfringens enterotoxin from equine isolates: Its character-ization, sequence and role in foal diarrhea. Epidemiol Infect 1998;120:193–200.

5. East LM, Dargatz DA, Traub-Dargatz JL, et al. Foaling-management practices associated with the occurrence of entero-colitis attributed to Clostridium perfringens infections in the equine neonate. Prev Vet Med 2000;46:61–74.

6. Tillotson K, Traub-Dargatz JL, Dickinson CE, et al. Population-based study of fecal shedding of Clostridium perfringens in broodmares and foals. J Am Vet Med Assoc 2002;220:342–348.

7. Frederick J, Giguère S, Sanchez LC. Infectious agents detected in the feces of diarrheic foals: A retrospective study of 233 cases (2003–2008). J Vet Intern Med 2009;23:1254–1260.

8. Silva ROS, Ribeiro MG, Palhares MS, et al. Detection of A/B toxin and isolation of Clostridium difficile and Clostridium per-fringens from foals. Equine Vet J 2013;45:671–675.

9. Mehdizadeh Gohari I, Arroyo L, MacInnes JI, Timoney JF, Parreira VT, Prescott JF. Characterization of Clostridium perfrin-gens in the feces of adult horses and foals with acute enterocoli-tis. Can J Vet Res 2015;78:1–7.

10. Mehdizadeh Gohari I, Parreira VR, Nowell VJ, Nicholson VM, Oliphant K, Prescott JF. A novel pore-forming toxin in type A Clostridium perfringens is associated with both fatal canine hae-morrhagic gastroenteritis and fatal foal necrotizing enterocolitis. PLoS One 2015; DOI:10.1371/journal.pone.0122684.

11. Smith HW. The development of the flora of the alimentary tract in young animals. J Pathol Bacteriol 1965;90:495–513.

12. Chan G, Abdolvahab F, Soltes G, et al. The epidemiology of Clostridium perfringens type A on Ontario swine farms, with spe-cial reference to cpb2-positive isolates. BMC Vet Res 2012;8:156.

13. Weese JS, Staempfli HR, Prescott JF. A prospective study of the roles of Clostridium difficile and Clostridium perfringens in equine diarrhea. Equine Vet J 2001;33:403–409.




Enterotoxigenic Escherichia coli (ETEC)-related colibacillosis in piglets is a cause of major economic losses as a result of either piglet death in the case of acute ETEC infections or poor weight gain in surviving piglets (1,2). K88 (F4), K99 (F5), F6 (FasA), and F41 antigens are important neonatal piglet ETEC fimbrial adhes-ins (1). Considerable effort has been invested in the development of vaccines against ETEC strains, albeit with limited success (3,4). Mucosal immunity is needed to prevent this enteric infection. Oral vaccination has been successful in raising levels of protective secre-tory immunoglobulin A (IgA) on the intestinal surface (5); however, gastric digestion of vaccines before priming of the immune system is a big hurdle in the development of oral vaccines (5). Bacterial ghosts are induced by the controlled expression of X174 lysis gene E. E-mediated lysis of bacteria results in the formation of empty bacte-rial cell envelopes (6,7). Ghost vaccines have the same cell surface composition as their living counterparts. They display all surface components in a natural nondenatured form and are able to induce a strong mucosal immune response (6,7). Because bacterial ghosts

can avoid gastric digestion like their live bacteria, oral immunization with the ghost effectively protects against various enteric pathogenic infections (8,9). Recently, the bacterial ghost system has been used as a vaccine delivery system providing the required adjuvant activity without the need for further additions (10,11). The intrinsic adjuvant properties of bacterial ghosts enhance systemic, mucosal, and cellular immunity to the target antigens (10).

After the Salmonella ghosts carrying K88ab, K88ac, K99, FasA, and F41 antigens of ETEC were constructed, pregnant sows were inoculated via various immunization routes with the ghost cells in previous study (12). The results showed that oral immuniza-tion of sows with the ghost may effectively protect their offspring against E. coli colibacillosis. In this study, pregnant sows were orally inoculated with various doses of the ghosts to determine the optimal conditions for protection by the ghost. Immune responses induced via oral inoculation were examined for pregnant sows and their piglets. We also evaluated the efficacy of the ghost vaccine candidate for protection against experimental colibacillosis in neonatal piglets.

Protective efficacy by various doses of Salmonella ghost vaccine candidate carrying enterotoxigenic Escherichia coli fimbrial antigen

against neonatal piglet colibacillosisJin Hur, John Hwa Lee

A b s t r a c tHumoral immune responses and protective efficacy by various doses of Salmonella ghost cells carrying enterotoxigenic Escherichia coli (ETEC) fimbrial antigens for protection against piglet colibacillosis were studied. All groups were orally primed and boosted at 11 and 14 wk of pregnancy, respectively. Group A sows were inoculated with phosphate-buffered saline (PBS), and groups B, C, and D sows were immunized with 2 3 109, 2 3 1010, and 2 3 1011 ghost cells, respectively. Serum immunoglobulin (Ig) G, and colostrum IgG and IgA levels of groups C and D sows were significantly higher than those of group A sows. In addition, serum IgG and IgA levels in group C and D piglets were significantly increased compared to those of group A piglets. After challenge with wild-type ETEC, diarrhea and mortality were not observed in group C and D piglets, while diarrhea was observed in 88.9% and 58.8% of groups A and B piglets, respectively, and 16.7% mortality was observed in group A piglets. These findings indicate that oral immunization of sows with 2 3 1010 or 1011 ghost cells can effectively protect their offspring from colibacillosis.

R é s u m éLes réponses immunitaires humorales et l’efficacité protectrice de doses variées de cellules fantômes de Salmonella transportant des antigènes fimbriaires d’Escherichia coli entérotoxinogène (ETEC) pour protéger des porcelets contre la colibacillose ont été étudiées. Tous les groupes ont reçu une dose initiale et un rappel par voie orale à 11 et 14 sem de gestation, respectivement. Les truies du groupe A furent inoculées avec de la saline tamponnée (PBS), et les groupes B, C, et D immunisés avec 2 3 109, 2 3 1010, et 2 3 1011 cellules fantômes, respectivement. Les niveaux d’immunoglobulines (Ig) G sériques, ainsi que des IgG et IgA du colostrum des truies des groupes C et D étaient significativement supérieurs à ceux des truies du groupe A. De plus, les niveaux des IgG et IgA sériques des porcelets des groupes C et D étaient significativement augmentés comparativement à ceux des porcelets du groupe A. Après infection défi avec une souche sauvage d’ETEC, aucune diarrhée et mortalité ne furent observées chez les porcelets des groupes C et D, alors que de la diarrhée fut observée chez 88,9 % et 58,8 % des porcelets des groupes A et B, respectivement, et 16,7 % de mortalité fut notée chez les porcelets du groupe A. Ces résultats indiquent que l’immunisation des truies avec 2 3 1010 ou 1011 de cellules fantômes peut protéger efficacement les petits de la portée contre la colibacillose.


College of Veterinary Medicine and Bio-Safety Research Institute, Chonbuk National University, Jeonju, South Korea 561-756.

Address all correspondence to Dr. John Hwa Lee; telephone: 182-63-850-0940; fax: 182-63-850-0910; e-mail: [email protected]

Received August 10, 2015. Accepted March 12, 2016.



Pregnant Large Yorkshire sows (N = 12) and their suckling piglets (N = 78) were used for this study. All sows had their first piglets in this study. From the sows it was attempted to isolate ETEC and Salmonella species through rectal swabs, according to previously described methods (12,13) on weeks 9, 11, and 14 of pregnancy, on the day of farrowing, and on days 3, 5, 7, 14, and 21 after farrow-ing. These bacteria were not detected from the rectal swabs of all sows. All the sows were vaccinated according to the vaccination program of the pig farm except for the commercial ETEC vaccine. Approximately 10 d prior to expected farrowing, the pregnant sows were transferred to individual pens with a farrowing crate with a heat lamp for their offspring. They were fed a commercially available feedstuff for pregnant sows without antibiotics or other growth promoters (13). All piglets were fostered with their dam for 28 d after birth. Isolating ETEC and Salmonella species from rectal swabs of the piglets was attempted according to previously described methods (12,13) on days 1, 3, and 5 after birth. These bacteria were not detected from the rectal swabs of all piglets dur-ing these periods. The animal experiments were conducted with the approval of the Chonbuk National University Animal Ethics Committee in accordance with the guidelines of the Korean Council on Animal Care (CBU 2012–0017). All bacterial strains used for this study were described previously (12). JOL1285 for K88ab, JOL1286 for K88ac, JOL1287 for K99, JOL1288 for FasA, and JOL1289 for F41 were used as vaccine strains. Wild-type JOL489, JOL564, and JOL599 E. coli isolates from diarrheic piglets were used as the challenge strains (12,13). The regulatory E lysis ghost cassette inserted into the pYA3342 plasmid containing a pBR origin, a multiple cloning site (MCS), and the asd gene were previously described (14–16). Purified recombinant K88ab, K88ac, K99, FasA, and F41 fimbrial proteins were prepared as previously described (13,17), and were used as antigens for enzyme-linked immunosorbent assay (ELISA). To construct Salmonella ghost cells producing recombinant K88ab, K88ac, K99, FasA, and F41 fimbrial antigens on their envelope, the genes encoding these fimbriae were amplified by polymerase chain reaction (PCR) using each fimbria-specific primer pair described in the previous study (12). The amplified genes were individually cloned into the regulatory E lysis ghost plasmid. The ghost cells were prepared as previously described (13,14). Briefly, a single colony of each strain was individually inoculated into 200 mL LB broth con-taining 0.2% L-arabinose, and the cultures were incubated at 28°C to reach an optical density (OD) of 0.3 to 0.4 at 600 nm. The cells were washed twice using fresh LB broth without L-arabinose, and resuspended in 200 mL LB broth without L-arabinose. Subsequently, the temperature was increased to 42°C.

A total of 12 pregnant sows were divided equally into 4 groups. All groups were orally primed and orally boosted at 11 and 14 wk of pregnancy, respectively. In group A, sows were inoculated with 10 mL PBS as a control, while group B, C, and D sows were primed and boosted with 2 3 109, 2 3 1010, and 2 3 1011 ghost cells in 10 mL PBS (the mixture containing 4 3 108, 4 3 109, and 4 3 1010 cells of each of the 5 ghost types), respectively. Blood samples were taken at week 11 of pregnancy [prior to prime immunization, this week was represented as week 0 post-prime immunization (PPI) in this study], at week 14 of pregnancy (prior to the booster, this week was represented as week 3 PPI) and on the day of farrowing (prior Ta

ble

I. C

linic

al s

igns

in s

uckl

ing

pigl

ets

post

-cha

lleng

e

D

ays

afte

r cha

lleng

e

3 4

5 6

7 10

14

Sow

Nu

mbe

r Nu

mbe

r %

of d

iarrh

ea

% o

f dia

rrhea

%

of d

iarrh

ea

% o

f dia

rrhea

%

of d

iarrh

ea

% o

f dia

rrhea

%

of d

iarrh

eagr

oupa

of s

ows

of p

igle

ts

M

Mo

S D

M

Mo

S D

M

Mo

S D

M

Mo

S D

M

Mo

S D

M

Mo

S D

M

Mo

S D

A A1

4

50

0 0

0 75

25

0

0 75

25

0

0 10

0 0

0 0

100

0 0

0 50

0

0 0

0 0

0 0

A2

6

66.7

16

.7

0 0

50

33.3

16

.7

0 50

16

.7

33.3

0

66.7

0

16.7

16

.7

66.7

0

0 16

.7

33.3

0

0 0

0 0

0 0

A3

8

25

0 0

0 50

12

.5

0 0

50

12.5

0

0 50

0

12.5

0

37.5

0

0 12

.5

12.5

0

0 0

0 0

0 0

B B1

4

0 0

0 0

0 0

0 0

25

0 0

0 25

0

0 0

50

0 0

0 0

0 0

0 0

0 0

0

B2

4 0

0 0

0 0

0 0

0 25

25

0

0 75

0

0 0

75

0 0

0 0

0 0

0 0

0 0

0

B3

9 0

0 0

0 0

0 0

0 22

.2

0 0

0 44

.4

0 0

0 22

.2

0 0

0 0

0 0

0 0

0 0

0C

C1

3 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0

C2

7 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0

C3

10

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

D D1

5

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

D2

8

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

D3

10

0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0a

The

sows

of g

roup

A w

ere

adm

inist

ered

with

PBS

as

a co

ntro

l; gr

oup

B we

re im

mun

ized

with

2 3

109

cells

of t

he g

host

; gro

up C

wer

e im

mun

ized

with

2 3

1010

cel

ls of

the

ghos

t, an

d gr

oup

D we

re

imm

unize

d wi

th 2

3 1

011 c

ells

of th

e gh

ost.

D —

Dea

th; M

— m

ild; M

o —

mod

erat

e; S

— s

ever

e.



to challenge, this time was represented as week 6 PPI). Colostrum samples were collected from the sows on the day of farrowing and blood samples were collected from all suckling piglets on day 4 post-birth. All samples were stored at 270°C until use. To investigate the fimbria-specific IgG and IgA concentrations in serum and colostrum, standard ELISA was performed using the pig IgG or IgA ELISA quantitation kit according to the manufacturer’s instructions (Bethyl Lab, Montgomery, Texas, USA). Wild-type E. coli isolates, JOL489, JOL564, and JOL599, were prepared for the challenge and all piglets were challenged with these strains according to previously described methods (13,14). All 5-day-old piglets were orally challenged with a 3.0-mL total mixture containing 1 3 109 CFU of each challenge strain on the same day that the challenge strains were prepared (Table 1). Results from ELISA are expressed as median 6 confidence intervals (CI). The Mann-Whitney U-test was used to determine the signifi-cant differences in serum and colostrum antibody titers between the vaccinated and non-vaccinated groups. SPSS 16.0 (SPSS, Chicago,

Illinois, USA) was used to perform all analyses, and P # 0.05 was regarded as statistically significant.

Antibody responses against each fimbrial antigen in the sera and colostrum from the immunized pregnant sows are presented in Figure 1. Serum IgG and colostrum IgA titers against all the individual fimbrial antigens were significantly increased in group C and D sows compared to those in group A sows at week 6 PPI (P # 0.05). In addition, colostrum IgG titers in all immunized group sows were significantly increased compared to group A sows on the day of farrowing (P # 0.05). As shown in Figure 2, serum IgG and IgA titers of group C and D piglets were significantly higher than those of control piglets on day 4 after birth (P , 0.01), while serum IgG and IgA titers in group B piglets were only slightly higher than those seen in the control.

Challenge strain-induced diarrhea was confirmed by isolating the challenge strain from the rectal swab according to the methods described in previous studies (12,13). Group C and D piglets did

Figure 1. Immune responses against the recombinant K88ab, K88ac, K99, FasA, and CF41 antigens in pregnant sows. (A) Serum IgG (mg/mL) titers; (B) colostrum IgG (mg/mL) and colostrum sIgA (mg/mL). Group A () PBS control; group B () oral prime and booster with 2 3 109 ghost cells; group C (), oral prime and booster with 2 3 1010 ghost cells; and group D (X) oral prime and booster with 2 3 1011 ghost cells. Data are the medians of all sows in each group and error bars show the confidence intervals (CIs). Asterisks indicate a significant difference between the titers of the groups immunized with the ghost (*P # 0.05) and those of the control group.

20

15

10

5

0

15

10

5

0

15

10

5

0

20

15

10

5

0

15

10

5

0 0 3 6

K88ab K88ac K99 FasA F41 K88ab K88ac K99 FasA F41

0 3 6

0 3 615

10

5

0

1.2

0.8

0.4

0.0

Ser

um K

88ab

IgG

(m

g/m

L)

Ser

um K

88ac

IgG

(m

g/m

L)

Ser

um K

99 Ig

G

(mg/

mL)

Ser

um F

asA

IgG

(m

g/m

L)

Ser

um F

41 Ig

G

(mg/

mL)

Col

ostr

um Ig

G (m

g/m

L)

Weeks post-immunization Weeks post-immunization

Weeks post-immunization

Col

ostr

um Ig

A (m

g/m

L)



not exhibit clinical signs of disease such as diarrhea up to 14 d after the challenge (Table 1). In contrast, diarrhea was observed in 16 of 18 piglets of group A beginning on day 3 after the challenge, and 3 died due to severe diarrhea. Diarrhea was also observed in 10 of 17 group B piglets.

Several maternal formulations of live, killed, and subunit vaccines have been used to control ETEC infection in piglets (17,18), and are administered in late pregnancy (19). Among these formulations, formalin-inactivated formulations are the most commonly avail-able vaccines against ETEC infection (19,20). However, formalin-inactivated formulations can affect the physiochemical and structural properties of the surface antigens (6,21). Therefore, the development of inactivated vaccines that are both safe and capable of inducing a specific and efficient immune response against bacterial infec-tions is important. Secretory IgA transported and secreted across the mucosal epithelium into the lumen can inhibit attachment by microorganisms and/or neutralize exotoxins (22,23). This defense mechanism against mucosal bacterial infection depends on the clearance of the pathogens from the gut, and mucosal immunity by systemic immune production is necessary for effective vaccination against E. coli colibacillosis (24). The mucosal response is stimulated by oral delivery of the vaccine antigen to a mucosal inductive site such as Peyer’s patches in the gut (11,25). The Salmonella ghost deliv-ery system can be used as an antigen delivery vehicle in mucosal inductive tissues (8,11).

Salmonella ghost cells with ETEC K88ab, K88ac, K99, FasA, and F41 fimbrial antigens in their cell envelope were constructed in a previous study (12). In this study, the pregnant sows were orally primed and boosted with different doses of the ghost cells in order to optimize effective immunization strategy against neonatal coli-bacillosis. The serum IgG, and colostrum IgA and IgG levels from group C and D sows were significantly elevated compared to those of group A sows. In addition, the serum IgG and IgA levels in piglets from groups C and D were significantly higher than those from group A. To evaluate the efficiency of maternally transferred immunity, the piglets in this study were challenged with the virulent E. coli strains isolated from diarrheic piglets. Previously reported

studies assessing immunity to ETEC infections in the offspring of sows vaccinated with formulations containing purified fimbriae have revealed instances of diarrhea and mortality in the suckling piglets of vaccinated sows (17,26). In this study, however, very minor clini-cal symptoms were observed in the offspring from group C and D sows after challenge, while diarrhea was observed in approximately 90% and 60% of piglets from groups A and B, respectively. In addi-tion, approximately 17% mortality was observed in group A piglets. These findings indicate that oral immunization of pregnant sows with 2 3 1010 or 2 3 1011 Salmonella ghost cells produced effective protection in their offspring against colibacillosis caused by ETEC. Collectively, the immunization with as low as 2 3 1010 ghost cells can optimally induce systemic and mucosal immunity and can effectively protect from the colibacillosis.

A c k n o w l e d g m e n tThis work was supported by the National Research Foundation

of Korea (NRF) grant funded by the Korea government (MISP) (No. 2015R1A2A1A14001011). We declare that there are no potential conflicts of interest for all authors.

Re f e r e n c e s 1. Chen X, Gao S, Jiao X, Liu XF. Prevalence of serogroups and

virulence factors of Escherichia coli strains isolated from pigs with postweaning diarrhoea in eastern China. Vet Microbiol 2004;103:13–20.

2. Vu-Khac H, Holoda E, Pilipcinec E. Distribution of virulence genes in Escherichia coli strains isolated from diarrhoeic piglets in the Slovak Republic. J Vet Med B Infect Dis Vet Public Health 2004;51:343–347.

3. Melkebeek V, Sonck E, Verdonck F, Goddeeris BM, Cox E. Optimized FaeG expression and a thermolabile enterotoxin DNA adjuvant enhance priming of an intestinal immune response by an FaeG DNA vaccine in pigs. Clin Vaccine Immunol 2007;14: 28–35.

Figure 2. Immune responses to fimbrial antigens in neonatal piglets from each group of immunized sows. Serum IgG (mg/mL) and serum IgA (mg/mL). Data are the medians of all suckling piglets in each group and error bars show the CIs. Group A () PBS control; group B () oral prime and booster with 2 3 109 ghost cells; group C () oral prime and booster with 2 3 1010 ghost cells; and group D (3) oral prime and booster with 2 3 1011 ghost cells. Asterisks indicate a significant difference between the titers of the groups immunized with the ghost (*P , 0.05) and those of the control group.

40

30

20

10

0

30

20

10

0

K88ab K88ac K99 FasA F41 K88ab K88ac K99 FasA F41

Ser

um Ig

G (m

g/m

L)

Ser

um Ig

A (m

g/m

L)



4. Vandamme K, Melkebeek V, Cox E, Remon JP, Vervaet C. Adjuvant effect of Gantrez(R)AN nanoparticles during oral vaccination of piglets against F41 enterotoxigenic Escherichia coli. Vet Immunol Immunopathol 2011;139:148–155.

5. Melkebeek V, Goddeeris BM, Cox E. ETEC vaccination in pigs. Vet Immunol Immunopathol 2013;152:37–42.

6. Jalava K, Hensel A, Szostak M, Resch S, Lubitz W. Bacterial ghosts as vaccine candidates for veterinary applications. J Control Release 2002;85:17–25.

7. Jalava K, Eko FO, Riedmann E, Lubitz W. Bacterial ghosts as car-rier and targeting systems for mucosal antigen delivery. Expert Rev Vaccines 2003;2:45–51.

8. Mayr UB, Haller C, Haidinger W, et al. Bacterial ghosts as an oral vaccine: A single dose of Escherichia coli O157:H7 bacterial ghosts protects mice against lethal challenge. Infect Immun 2005; 73:4810–4817.

9. Cai K, Zhang Y, Yang B, Chen S. Yersinia enterocolitica ghost with msbB mutation provides protection and reduces proinflamma-tory cytokines in mice. Vaccine 2013;31:334–340.

10. Mayr UB, Walcher P, Azimpour C, Riedmann E, Haller C, Lubitz W. Bacterial ghosts as antigen delivery vehicles. Adv Drug Deliv Rev 2005;57:1381–1391.

11. Tabrizi CA, Walcher P, Mayr UB, et al. Bacterial ghosts — biologi-cal particles as delivery systems for antigens, nucleic acids and drugs. Curr Opin Biotechnol 2004;15:530–537.

12. Hur J, Lee JH. A new enterotoxigenic Escherichia coli vaccine candidate constructed using a Salmonella ghost delivery system: Comparative evaluation with a commercial vaccine for neona-tal piglet colibacillosis. Vet Immunol Immunopathol 2015;164: 101–109.

13. Hur J, Lee JH. Protection against neonatal Escherichia coli diar-rhea by vaccination of sows with a novel multivalent vaccine candidate expressing E. coli adhesins associated with neonatal pig colibacillosis. Res Vet Sci 2013;94:198–204.

14. Kang HY, Srinivasan J, Curtis R III. Immune response to recom-binant pneumococcal PspA antigen delivered by live attenuated Salmonella enterica serovar Typhimurium vaccine. Infect Immun 2002;70:1739–1749.

15. Arora A, Rinehart D, Szabo G, Tamm LK. Refolded outer mem-brane protein A of Escherichia coli forms ion channels with two conductance states in planar lipid bilayers. J Biol Chem 2000; 275:1594–1600.

16. Singh SP, Williams YU, Miller S, Nikaido H. The-C-terminal domain of Salmonella enterica serovar Typhimurium OmpA is an immunodominant antigen in mice but appears to be only partially exposed on the bacterial cell surface. Infect Immun 2003;71:3937–3946.

17. Hur J, Lee JH. Immune responses to new vaccine candidates constructed by a live attenuated Salmonella Typhimurium delivery system expressing Escherichia coli F4, F5, F6, F41 and intimin adhesin antigens in a murine model. J Vet Med Sci 2011;73: 1265–1273.

18. Ruan X, Zhang W. Oral immunization of a live attenuated Escherichia coli strain expressing a holotoxin-structured adhesin- toxoid fusion (1FaeG-FedF-LTA2:5LTb) protected young pigs against enterotoxigenic E. coli (ETEC) infection. Vaccine 2013;31:1458–1463.

19. Nagy B, Fekete PZ. Enterotoxigenic Escherichia coli in veterinary medicine. Int J Med Microbiol 2005;295:443–454.

20. Riising HJ, Murmans M, Witvliet M. Protection against neonatal Escherichia coli diarrhoea in pigs by vaccination of sows with a new vaccine that contains purified enterotoxic E. coli virulence factors F4ac, F4ab, F5 and F6 fimbrial antigens and heat-labile E. coli enterotoxin (LT) toxoid. J Vet Med B Infect Dis Vet Public Health. 2005;52:296–300.

21. Huter V, Hensel A, Brand E, Lubitz W. Improved protection against lung colonization by Actinobacillus pleuropneumoniae ghosts: Characterization of a genetically inactivated vaccine. J Biotechnol 2000;83:161–172.

22. MacDonald TT. The mucosal immune system. Parasite Immunol 2003;25:235–246.

23. Holmgren J, Czerkinsky C. Mucosal immunity and vaccines. Nat Med 2005;11:S45–S53.

24. Haesebrouck F, Pasmans F, Chiers K, Maes D, Ducatelle R, Decostere A. Efficacy of vaccines against bacterial diseases in swine: What can we expect? Vet Microbiol 2004;100:255–268.

25. Lawson LB, Norton EB, Clements JD. Defending the mucosa: Adjuvant and carrier formulations for mucosal immunity. Curr Opin Immunol 2011;23:414–420.

26. Nagy LK, Mackenzie T, Painter KR. Protection of the nursing pig against experimentally induced enteric colibacillosis by vaccina-tion of dam with fimbrial antigens of E coli (K88, K99 and 987P). Vet Rec 1985;117:408–413.





Buprenorphine hydrochloride, a partial m-opioid receptor agonist, is one of the most commonly used analgesics in laboratory animal medicine (1). Its simple formulation (Buprenorphine HCl) has an activity of approximately 8 h, which is due to its slow rate of drug disassociation from the receptor, making it a long-acting agent. It also has less respiratory and cardiovascular side effects than most opioids, which makes it a good therapeutic choice (2). Currently, only a few pharmaceutical opioid formulations with long-acting effects provide an appropriate analgesia for several days, seeing as their use prevents repeated injections and handling-associated stress. A common alternative to treat pain in a sustained manner is the use of transdermal systems such as fentanyl and buprenorphine patches (3–4). However, on many occasions, these patches do not properly

adhere to the animals’ skin and therefore, animals do not receive an appropriate analgesia. Other problems may occur, such as fallen patches that are ingested by animals and represent a toxic risk (5–6).

The pharmacokinetics and pharmacodynamics of sustained release (SR) buprenorphine after a single subcutaneous injection has been evaluated in mice (7–9), rats (10,11), cats (12), and dogs (13). Effective buprenorphine blood levels are observed between 12 and 24 h post-injection in mice and up to 72 h post-injection in rats. In dogs, the effective buprenorphine plasma concentration is shown to be greater than 0.6 ng/mL after a single subcutaneous administration and this value is detected in the plasma for 72 h post-injection. In pigs, a threshold of 0.1 ng/mL buprenorphine plasma concentration is considered therapeutic and SR buprenorphine is allowed for an

Plasma concentrations of buprenorphine following a single subcutaneous administration of a sustained release formulation

of buprenorphine in sheepChiara Zullian, Pablo Lema, Melissa Lavoie, Aurore Dodelet-Devillers, Francis Beaudry, Pascal Vachon

A b s t r a c tThe goal of the present study was to evaluate the potential use of slow release buprenorphine in sheep. Twelve adult female sheep (6 Dorset and 6 Suffolk, 12 months of age) were used for this project and were divided into 2 experimental groups (n = 6/group comprising 3 Dorset and 3 Suffolk sheep). Sustained release (SR) buprenorphine was administered subcutaneously in the scapular region at a concentration of 0.1 mg/kg body weight (BW) for group 1 and of 0.05 mg/kg BW for group 2. Following blood collections at selected time points, plasma concentrations of buprenorphine was performed by tandem liquid chromatograph-mass spectrometry. Mean buprenorphine concentration was above 0.1 ng/mL at 48 h up to 192 h post-injection for group 1 and it was above 0.1 ng/mL at 48 h up to 72 h post-injection for group 2. In conclusion, a long lasting potential analgesic plasma level of buprenorphine is attained following a single subcutaneous injection of 0.1 mg/kg BW of SR buprenorphine in sheep. However the effective analgesic plasma threshold still needs to be determined in sheep.

R é s u m éL’objectif de la présente étude était d’évaluer l’utilisation potentielle de buprénorphine à relâchement lent (RL) chez le mouton. Douze brebis adultes (6 Dorset et 6 Suffolk, 12 mois d’âge) ont été utilisées pour ce projet et ont été réparties en deux groupes expérimentaux (n = 6/groupe, 3 Dorset et 3 Suffolk). De la buprénorphine à relâchement continu a été administrée par voie sous-cutanée dans la région scapulaire à une concentration de 0,1 mg/kg de poids corporel (PC) pour le groupe 1 et à 0,05 mg/kg de PC pour le groupe 2. Suite à des prélèvements sanguins à des moments sélectionnés, les concentrations plasmatiques de buprénorphine ont été déterminées par spectrométrie de masse en tandem avec la chromatographie en phase liquide. La concentration moyenne de buprénorphine était supérieure à 0,1 ng/mL après 48 h et jusqu’à 192 h post-injection pour le groupe 1, et était supérieure à 0,1 ng/mL après 48 et jusqu’à 72 h post-injection pour le groupe 2. En conclusion, un niveau plasmatique prolongé de buprénorphine avec un potentiel analgésique est atteint suite à une injection sous-cutanée unique de 0,1 mg/kg de PC de buprénorphine RL chez le mouton. Toutefois, le seuil plasmatique analgésique réel demeure encore à être déterminé chez le mouton.


Département de Biomédecine Vétérinaire, Faculté de Médecine Vétérinaire, Université de Montréal, Saint-Hyacinthe, Quebec J2S 2M2 (Zullian, Dodelet-Devillers, Beaudry, Vachon); AccelLab Preclinical Research, Boisbriand, Quebec J7H 1N8 (Lema, Lavoie).

Address all correspondence to Dr. Pascal Vachon; telephone: 514-343-6111, ext. 8294; e-mail: [email protected]

Received September 18, 2015. Accepted January 18, 2016.



estimated 264 6 32 h of analgesia (4). In macaques, a therapeutic plasma concentration of SR buprenorphine is detected for at least 5 d, with an effective analgesic plasma threshold of 0.1 ng/mL (14). These findings suggest a long duration of efficacy in all these species, when a single dose of SR buprenorphine is administered subcutaneously.

Buprenorphine is metabolized to its main metabolite, norbupreno-phine, by the CYP3A liver enzyme. This metabolite is a m-, d-, and k-opioid and nociceptin receptor agonist (15,16); however, it has poor brain penetrability and few analgesic properties in mice (17). In vitro, both buprenorphine and norbuprenorphine are metabolized by the same metabolic liver pathways, and they may each interfere with the degradation of the other molecule (18). The analysis of both buprenorphine and norbuprenorphine, therefore, is important in any pharmacokinetic study.

The goal of the present study was to measure the plasma concen-trations of buprenorphine and its main metabolite, norbuprenor-phine, following a single administration of a subcutaneous SR buprenorphine formulation in sheep. A theoretical effective concen-tration of 0.1 ng/mL was used for this analysis (13,14), which will need to be validated with a pharmacodynamics study in sheep. To our knowledge, no pharmacokinetic data relating to the administra-tion of such a formulation to sheep is presently available. A longer effective duration — when compared to the simple formulation of buprenorphine hydrochloride — would have a beneficial effect on sheep welfare in experimental procedures such us orthopedic surger-ies that induce long-lasting pain (19).

Twelve female sheep (6 Dorset and 6 Suffolk, 12 months of age) were used for this project. Animals were acclimated for 3 mo in the facilities of AccelLAB (AccelLAB Preclinical Research, Boisbriand, Quebec), where they were divided into 2 experimental groups (n = 6/group), each represented by 3 Dorsets and 3 Suffolk sheep. Sheep weighed 68.6 6 22.8 kg and 63.2 6 19.2 kg in groups 1 and 2, respectively. Animals were purchased from 2 different sources (Research Flock, University of Guelph, Ontario and Pozzi Ranch, California, USA) and were free of Q-fever, chlamydia, Johne’s dis-ease, caseous lymphadenitis, Meadi Visna, and scrapie. The sheep

were fed daily with Certified Diet (Harlan Teklad Ruminant diet 7060C; Harlan Laboratories, Madison, Wisconsin, USA) and had access to municipal water and certified hay ad libitum. Environmental conditions were maintained within accepted limits for temperature (16°C to 22°C) and relative humidity (40% to 70%). Lighting was maintained in a 12 h dark:12 h light cycle. Sheep were housed on a concrete floor, covered with wood chip bedding that was changed at least once every 3 wk and feces were removed daily. The experi-mental protocol was reviewed and approved by the AccelLAB’s Institutional Animal Care and Use Committee and was in compliance with Canadian Council on Animal Care regulations.

Sustained release buprenorphine (3 mg/mL; SR Veterinary Technologies, Windsor, Colorado, USA) was administered sub-cutaneously in the subscapular region. The area was shaved and cleaned with alcohol prior to injection. Group 1 received 0.1 mg/kg body weight (BW) and Group 2 received 0.05 mg/kg BW of SR buprenorphine. Blood samples were collected (1 mL/time point) in 2 mL K3EDTA tubes (Newton, North Carolina, USA) at selected time points (Predose, 1, 4, 8, 24, 48, 72, 96, 168, and 336 h) from the jugular vein. They were kept on ice pending centrifugation (3200 3 g for 10 min, at room temperature) which was performed within 30 min of collection. Afterwards, plasma was harvested and transferred in identified plastic tubes which were placed on dry ice pending storage at –80°C.

Using protein precipitation as a sample preparation technique, buprenorphine and norbuprenorphine were extracted from sheep plasma. A 200 mL volume of internal standard solution (0.5 ng/mL of d4-buprenorphine and d3-norbuprenorphine in methanol) was added to an aliquot of 100 mL of plasma sample. The sample was vortexed for approximately 5 s and let still for a period of 10 min, then centri-fuged at 12 000 3 g for 10 min. The supernatant was transferred to an injection vial for online Solid-Phase Extraction (SPE) and HPLC-MS analysis. On-line SPE and HPLC were performed using Thermo UltiMate 3000 Rapid Separation UHPLC system, a Thermo Accela pump, a 6-port valve, and 2 HPLC columns. A 150 mL volume of the sample was injected onto the Thermo Hypersil GOLD 20 3 2.1 mm, 12 mm loading column in an aqueous mobile phase (e.g., 5:95:0.1; acetonitrile:water:formic acid). After 1 min, the 6-port valve was switched enabling the loading column to be backflushed onto the analytical column (Thermo BioBasic 50 3 2.1 mm, 5 mm). The chromatography was achieved using a linear gradient at a flow rate of 300 mL/min (i.e., 5:95:0.1 to 80:20:0.1 in 4 min) and maintained 1 min. The mobile phase composition ratio was reverted at the initial conditions (5:95:0.1) and the column was allowed to re-equilibrate for 5 min. Mass spectrometer detection was performed in positive ion mode and operating in full scan high-resolution/accurate-mass mode using a Thermo Q-Exactive Orbitrap Mass Spectrometer (San José, California, USA).

All pharmacokinetic parameters were calculated using WinNonLin 5.2 (Pharsight Corporation, Mountain View, California, USA) using noncompartmental methods (20). The elimination rate constant (kel) was calculated using the last 3 measured plasma concentrations for each animal [linear regression fit (R2) . 0.9] and a terminal elimina-tion half-life (T1/2) was calculated using 0.693/kel. The area under the curve from time 0 to the last measurable concentration (AUC0-t) was calculated using the linear trapezoidal rule and with the last

Figure 1. Semi-logarithmic graph of the concentration-time profiles (mean 6 SD) of plasma buprenorphine after a single subcutaneous administration of SR buprenorphine in sheep with either 0.1 mg/kg BW (n = 6) or 0.05 mg/kg BW (n = 6).SC — subcutaneous; SD — standard deviation.



measured plasma concentration. The clearance was calculated by dividing the dose by the AUC0-t.

Motor and feeding behaviors appeared normal following the administration of SR buprenorphine. A slight dilation of the super-ficial vessels of the ears and sclera were observed in all animals from both groups at 48 h post-injection. This was still present in 66% of the animals at 72 h, while at 96 h post-injection, the vessels’ dilation was still present in 33% of the animals from group 2 but no longer present in group 1. A mild skin reaction (swelling and erythema) was seen in only one animal of each group at 24 h and 48 h post-injection.

Buprenorphine plasma concentrations after a single subcutaneous injection of 0.1 mg/kg BW (group 1) and 0.05 mg/kg BW (group 2) SR Buprenorphine are shown in Figure 1. Mean buprenorphine con-centrations were above the 0.1 ng/mL analgesic threshold starting at 48 h and up to 192 h post-injection for animals in group 1 and from 48 h up to 72 h for animals in group 2. For group 1, mean Cmax and Tmax were 0.28 6 0.6 ng/mL and 96 h, respectively. The mean clear-ance and T1/2 were 0.05 6 0.017 mL/h and 45.8 6 23.4 h, respectively. For group 2, mean Cmax and Tmax were 0.14 6 0.04 ng/mL and 72 h, respectively. The mean clearance and T1/2 were 0.014 6 0.013 mL/h and 23.8 6 13.6 h, respectively. Norbuprenorphine was not detected in any of the samples.

Buprenorphine’s wide safety margin and long duration of action are likely the most significant characteristics that determine its widespread use in veterinary medicine. Sheep are widely used in experimental surgical models and very often require adequate control of pain, especially following orthopedic procedures. There are several ways to manage post-surgical pain, depending on the research protocol and the animal model chosen; osmotic subcu-taneous mini pumps may be an option to assure a constant and reliable drug delivery. One other option is repeated administration of analgesics. Moreover, the use of patches may constitute a good alternative to chronic injections; however, the toxicity that may be associated with the ingestion of such a formulation may represent a variable to be avoided, as reported in the literature (5–6). The simple formulation of buprenorphine hydrochloride — currently available and commonly used in sheep — has to be administered every 6 to 8 h to assure a good level of analgesia, thus requiring personnel efforts and potentially inducing significant stress as well as forcing the animals to move, which could be contraindicated in some protocols.

To predict the clinical management of pain, a theoretical effective buprenorphine concentration of 0.1 ng/kg BW was used, based on published data on different species (4,15). The subcutaneous administration of 0.1 mg/kg BW of SR buprenorphine provided mean plasma concentrations above the selected theoretical limit, starting at 48 h post-injection and up to 192 h post-injection. For our research objectives, the injection of 0.05 mg/kg BW did not provide sufficient effective drug exposure to be selected as a therapeutic option. This is justified by the longer delay to attain a 0.1 mg/kg BW plasma concentration (48 h post-administration) and the short effective exposure (48 to 72 h).

As for side effects at the injection site, mild skin reactions were previously observed in some species after a subcutaneous admin-istration of SR buprenorphine, but resolved spontaneously with time and had no major effects on the animals’ health (4). In the literature, the presence of nonpainful subcutaneous nodules have

also been reported and were described as not visually obvious; these resolved within 1 month of the initial injection. Furthermore, a nodule observed in the cervical region of a dog who received a single subcutaneous injection of SR buprenorphine, was removed and then analyzed for histopathology; the lesion was described as a pyogranuloma with central vacuoles containing sparse grey material and consistent with injection site material (13). The variability of the reactions reported may be explained in part by the lipophilic nature of buprenorphine and the diversity in adiposity in the subjects tested; such differences may alter the absorption and the metabolism of the compound, due to a possible accumulation in intracellular or subcutaneous fat.

The mass spectrometry method did not allow for any detection of norbuprenorphine, which is associated with potential clinical effects such as analgesia, respiratory depression, and sedation. Since no data are currently available on the pharmacokinetics of buprenorphine and its metabolites in sheep, it can only be speculated that the low concentrations of norbuprenorphine are most likely not contributing to the analgesia in this species.

In conclusion, this study shows a long-lasting potential analgesic plasma level of buprenorphine, after a single subcutaneous injec-tion of 0.1 mg/kg BW of SR buprenorphine in sheep starting 2 d after the drug administration and lasting for 5 d. To make full use of SR buprenorphine and depending on the length of analgesia required, treatments could start 2 d before surgery to obtain an analgesic effect immediately after the surgery, or if SR buprenor-phine is administered on the day of surgery, another analgesic drug could be given to provide analgesia for the first 2 d following surgery, providing a longer duration of analgesia post-operatively. However, before these results can be applied in a clinical setting, the effective analgesic plasma threshold — determined in other species as 0.1 ng/mL — still needs to be determined in sheep.

A c k n o w l e d g m e n t sThis study was supported by AccelLab and funded in by the Fond

de recherche pour la médecine des animaux de laboratoire (Pascal Vachon) and the Fond du centenaire de la faculté de médecine vété-rinaire. We thank Thermo Fisher Scientific for providing a generous access to a Q-Exactive Orbitrap Mass Spectrometer.

Re f e r e n c e s1. Roughan JV, Flecknell PA. Buprenorphine: A reappraisal of its

antinociceptive effects and therapeutic use in alleviating post-operative pain in animals. Lab Anim 2002;36:322–343.

2. Fish RE, Brown MJ, Danneman PJ, Karas AZ. Anesthesia and Analgesia of Laboratory Animals. 2nd ed. San Diego, California: Academic Press, 2011.

3. Malavasi LM, Augustsson H, Jensen-Waern M, Nyman G. The effect of transdermal delivery of fentanyl on activity in growing pigs. Acta Vet Scand 2005;46:149–157.

4. Thiede AJ, Garcia KD, Stolarik DeAnne F, Ma J, Jenkins GJ, Nunamaker EA. Pharmacokinetics of sustained-relaase and transdermal buprenorphine in Göttingen minipigs (Sus scrofa domestica). J Am Assoc Lab Anim Sci 2014;53:692–699.



5. Schmiedt CW, Bjorling DE. Accidental prehension and suspected transmucosal or oral absorption of fentanyl from a transdermal patch in a dog. Vet Anaesth Analg 2007;34:70–73.

6. Deschamps JY, Gaulier JM, Podevin G, Cherel Y, Ferry N, Roux FA. Fatal overdose after ingestion of a transdermal fen-tanyl patch in two non-human primates. Vet Anaesth Analg 2012;39:653–656.

7. Carbone ET, Lindstrom KE, Diep S, Carbone L. Duration of action of sustained-release buprenorphine in 2 strains of mice. J Am Assoc Lab Anim Sci 2012;51:815–819.

8. Kendall LV, Hansen RJ, Dorsey K, Kang S, Lunghofer PJ, Gustafson DL. Pharmacokinetics of sustained-release analgesics in mice. J Am Assoc Lab Anim Sci 2014;53:478–484.

9. Healy JR, Tonkin JL, Kamarec SR, et al. Evaluation of an improved sustained-release buprenorphine formulation for use in mice. Am J Vet Res 2014;75:619–625.

10. Foley PL, Liang H, Crichlow AR. Evaluation of a sustained-release formulation of buprenorphine for analgesia in rats. J Am Assoc Lab Anim Sci 2011;50:198–204.

11. Chum HH, Jampachairsri K, McKeon GP, Yeomans DC, Pacharinsak C, Felt SA. Antinociceptive effects of sustained-release buprenorphine in a model on incisional pain in rats (Rattus norvegicus). J Am Assoc Lab Anim Sci 2014;53:193–197.

12. Catbagan DL, Quimby JM, Mama KR, Rychel JK, Mich PM. Comparison of the efficacy and adverse effects of sustained-release buprenorphine hydrochloride following subcutaneous administration and buprenorphine hydrochloride following oral transmucosal administration in cats undergoing ovariohyster-ectomy. Am J Vet Res 2011;72:461–466.

13. Nunamaker EA, Stolarik DF, Ma J, Wilsey A, Jenkins GJ, Medina CL. Clinical efficacy of sustained-release buprenor-

phine with meloxicam for postoperative analgesia in Beagle dogs undergoing ovariohysterectomy. J Am Assoc Lab Anim Sci 2014;53:494–501.

14. Nunamaker EA, Halliday LC, Moody DE, Fang WB, Lindeblad M, Fortman JD. Pharmacokinetic of 2 formulations of buprenor-phine in macaques (Macaca mulata and Macaca fascicularis). J Am Assoc Lab Anim Sci 2013;52:48–56.

15. Yassen A, Kan J, Olofsen E, Suidgeest E, Dahan A, Danhof M. Pharmacokinetic-pharmacodynamic modeling of the respiratory depressant effect of norbuprenorphine in rats. J Pharmacol Exp Ther 2007;321:598–607.

16. Huang P, Kehner GB, Cowan A, Liu-Chen LY. Comparison of pharmacological activities of buprenorphine and norbuprenor-phine: Norbuprenorphine is a potent opioid agonist. J Pharmacol Exper Ther 2001;297:688–695.

17. Brown SM, Campbell SD, Crafford A, Regina KJ, Holtzman MJ, Kharasch ED. P-glycoprotein is a major determinant of norbu-prenorphine brain exposure and antinociception. J Pharmacol Exp Ther 2012;343:53–61.

18. Brown SM, Holtzman M, Kim T, Kharasch ED. Buprenorphine metabolites, buprenorphine-3-glucuronide and norbuprenor-phine-3-glucuronide, are biologically active. Anesthesiology 2011;115:1251–1260.

19. Martini L, Fini M, Giavaresi G, Giardino R. Sheep model in orthopedic research: A literature review. Comp Med 2001;51: 292–299.

20. Rowland M, Towzer TN. Clinical Pharmacokinetics: Concepts and Application. 4th ed. Philadelphia, Pennsylvania: Lippincott, Williams and Wilkins, 2010:367–389.


CVMA CONVENTION welcomes everyone "from away"

CVMA CONVENTION welcomes everyone "from away"


Produits et services Supplies and services

RENOUVELEZ VOTRE ABONNEMENT

DÈS MAINTENANT!

RENEW YOUR SUBSCRIPTION

NOW!

Four issues of the Canadian Journal of Veterinary Research for 2016 (online)

Quatre numéros de la Revue canadienne de recherche vétérinaire pour 2016

Subscription/Abonnement 2016Personal Subscription/ $165 $ (1 GST/HST as Abonnement personnel applicable; TPS/TVH en (Canada) vigueur)

Personal Subscription/ $185 $ US/É.-U. Abonnement personnel (Foreign/Étranger)

Institutional Subscription/ $215 $ (1 In Canada, GST/HST Abonnement institutionnel as applicable; Au Canada, TPS/TVH selon le cas)

Contact/Communiquez avec : Linda Chow Editorial Coordinator/Coordonnatrice de la rédactionCanadian Journal of Veterinary ResearchRevue canadienne de recherche vétérinaire339 rue Booth Street Ottawa, Ontario K1R 7K1Canada

Tel./Tél. : (613) 236-1162, x117Fax/Télécopieur : (613) 236-9681E-mail/Courriel : [email protected]

w w w . c a n a d i a n v e t e r i n a r i a n s . n e t

w w w . v e t e r i n a i r e s a u c a n a d a . n e t

Tables of contentsComing events

Article abstractsClassified advertising

Subscription informationInstructions for authors

Tables des matières Événements à venirRésumés d’articles Annonces classées Information sur l’abonnement Directives à l’intention des auteurs

Canadian Journal of Veterinary Research

available online now!

Articles are now easily accessible over the Internet at

PubMed Centralwww.pubmedcentral.com

Reaching Canada’s VeterinariansGet your message into

The Canadian Veterinary JournalFor more information contact:Laima LaffitteAdvertising ManagerTel.: (613) 673-2659Fax: (613) 673-2462e-mail: [email protected]


Relationship of skeletal muscle inflammation with obesity and obesity-associated hyperinsulinemia in horsesHeidi E. Banse, Todd C. Holbrook, Nicholas Frank, Dianne McFarlane . . . . . . . . .217

A new technique for repeated biopsies of the mammary gland in dairy cows allotted to Latin-square design studies Luciano S. de Lima, Eric Martineau, Francilaine E. De Marchi, Marie-France Palin, Geraldo T. dos Santos, Hélène V. Petit . . . . . . . . . . . . . . . . . . . . . . . . .225

Pharmacokinetics of combination of amikacin sulfate and penicillin G sodium for intravenous regional limb perfusion in adult horsesJorge E. Nieto, Jan Trela, Scott D. Stanley, Sawsan Yamout, Jack R. Snyder. . . . . . . . . . . 230

Changes in gene expression of tumor necrosis factor alpha and interleukin 6 in a canine model of caerulein-induced pancreatitis Ruhui Song, Dohyeon Yu, Jinho Park . . . . . . 236

Short Communications/ Communications brèvesPrevalence of netF-positive Clostridium perfringens in foals in southwestern Ontario Abigail Finley, Iman Mehdizadeh Gohari, Valeria R. Parreira, Miranda Abrahams, Henry R. Staempfli, John F. Prescott . . . . . . . 242

Protective efficacy by various doses of Salmonella ghost vaccine candidate carrying enterotoxigenic Escherichia coli fimbrial antigen against neonatal piglet colibacillosisJin Hur, John Hwa Lee . . . . . . . . . . . . . . . . . . 245

Plasma concentrations of buprenorphine following a single subcutaneous administration of a sustained release formulation of buprenorphine in sheep Chiara Zullian, Pablo Lema, Melissa Lavoie, Aurore Dodelet-Devillers, Francis Beaudry, Pascal Vachon. . . . . . . . . . . . . . . . . . . . . . . . . 250


Documents

Provided for non-commercial research and education use ... · Ayako Miyazaki, Makoto Nagai, Hiroshi Tsunemitsu..... 189 Validation of an assay for quantification of alpha-amylase