Mouse Models of Chronic Lung Infection WithPseudomonas aeruginosa: Models for the Study of

Cystic Fibrosis

Peter K. Stotland, MSC, Danuta Radzioch, PhD, and Mary M. Stevenson, PhD*

Summary. The discovery of the CFTR gene in 1989 has lead to rapid progress in understandingthe molecular basis of cystic fibrosis (CF) and the biological properties of the cystic fibrosistransmembrane conductance regulator (CFTR) protein. However, more than 10 years later,recurrent lung infections with Pseudomonas aeruginosa, which lead to chronic lung disease andeventual respiratory failure, remain the major cause of morbidity and mortality among CF pa-tients. A distinguishing feature of lung disease in CF is an exaggerated and persistent inflam-matory response, characterized by the accumulation of excessive numbers of neutrophils anddysregulated cytokine production. The events leading to the establishment of lung infection withP. aeruginosa, especially the inflammatory and immunological events, and the relation betweenthe CF defect and infection, remain largely undefined. Progress in this area has been hamperedby the lack of a suitable animal model. An exciting achievement in the past few years has beenthe development of a number of variants of CFTR-deficient mice which exhibit defective cAMP-mediated Cl− conductance and have a range of clinical phenotypes from mild to severe.

In parallel, a model of chronic P. aeruginosa lung infection has been established in geneticallyand immunologically well-defined inbred mouse strains which differ in susceptibility to this in-fection in the lung. BALB/c mice are resistant, while DBA/2 mice are extremely susceptible, withhigh mortality within 3 days of infection. C57BL/6 and A/J mice are relatively susceptible andexperience low mortality. Furthermore, the bacterial load correlates with the magnitude andquality of the inflammatory response in the infected lungs of BALB/c and C57BL/6 mice. Al-though results of infection studies in CFTR-deficient mice have been variable, C57BL/6-Cftrm1UNC /Cftrm1UNC knockout mice compared to littermate control mice are highly susceptibleto chronic P. aeruginosa infection in the lung.

The availability of CFTR knockout mice and non-CF inbred mice differing in susceptibility tochronic P. aeruginosa infection offers useful tools for progress in understanding the genesis ofchronic P. aeruginosa infection and the ensuing inflammation in the CF lung, as well as therelation between the CF defect and infection. Information generated from these studies willprovide the rationale for the development of novel immunomodulatory measures capable ofameliorating or modulating the chronic inflammation associated with CF lung disease. PediatrPulmonol. 2000;30:413–424. © 2000 Wiley-Liss, Inc.

Key words: cystic fibrosis; lung infection; Pseudomonas aeruginosa; animal models;inflammation; cytokines

INTRODUCTION

Cystic fibrosis (CF) is the most common autosomal-recessive disorder affecting Caucasian populations, withan incidence of approximately 1 in 2,500 live births,1 andit is even more prevalent in certain geographic areas,such as the Saguenay-Lac St. Jean area in the Province ofQuebec (Canada).2 CF was identified as a clinical entityin 1936 by Fanconi et al. and was further characterized 2years later by Andersen.3 Studies in the late 1980s led tothe identification of the cystic fibrosis transmembrane

McGill Centre for the Study of Host Resistance, Montreal GeneralHospital Research Institute and Department of Medicine, McGill Uni-versity, Montreal, Quebec, Canada.

Grant sponsor: Fonds de la Recherche en Sante´ du Quebec; Grantsponsor: Fonds pour la Formation de Chercheurs et l’Aide a` la Re-cherche; Grant sponsor: Canadian Cystic Fibrosis Foundation.

*Correspondence to: Mary M. Stevenson, Ph.D., Montreal GeneralHospital Research Institute, 1650 Cedar Ave., Montreal, Quebec H3G1A4, Canada. E-mail: mcev@musica.mcgill.ca

Received 6 October 1999; Accepted 13 March 2000.

Pediatric Pulmonology 30:413–424 (2000)

State of the Art

© 2000 Wiley-Liss, Inc.

conductance regulator (CFTR) gene on human chromo-some 7, which encodes a transmembrane glycoprotein of1,480 amino acid residues that functions as a chloridechannel regulated by cyclic AMP.4 Over 900 mutationsin CFTR have been identified since the isolation of thegene. The most common mutation isDF508, a 3 base-pair deletion in exon 10, resulting in deletion of phenyl-alanine.

Although abnormalities of pancreatic function werethe first manifestations of the disease to be described, CFis also characterized by thick mucus secretions from epi-thelia, resulting in plugging of pancreatic ducts and air-ways, and abnormal sweat chloride levels in affectedindividuals.1,5 Indeed, a diagnosis of CF is confirmed bythe finding of greater than 60 mmol/L of chloride in thesweat. Other manifestations of the disease include maleinfertility due to obstructive azoospermia, malabsorptionof gut contents due to pancreatic insufficiency, and neo-natal meconium ileus or intestinal blockage.

The major cause of morbidity and mortality among CFpatients is chronic endobronchial infection withPseudo-monas aeruginosa,an opportunistic Gram-negative bac-terium.6–8 CF is associated with the development of se-vere, progressive bronchiectasis. The exact role ofPseudomonasinfection in the progression of lung diseaseis unclear. It is now widely recognized that host factorsproduced during chronic inflammation play a major rolein the etiology of lung disease in CF.9,10 Initially, thelower respiratory tract of CF infants gets colonized withorganisms such asStaphylococcus aureusand Hae-mophilus influenzae.The incidence of these infectionswanes, whereas chronic infections with mucoid strains ofP. aeruginosatend to persist in most patients. An exag-gerated and persistent, predominately neutrophilic in-flammatory response toP. aeruginosainfection is con-sidered a distinguishing feature of lung disease in CFpatients.9,10 Recent evidence suggests that dysregulatedcytokine production by various cell types in the lung maybe the underlying basis of the excessive inflammatoryresponse in the CF lung rather than infection per se.11 Ithas been demonstrated that the levels of both neutrophilsand the neutrophil chemotactic cytokine interleukin(IL)-8 are significantly elevated in bronchoalveolar la-

vage fluids (BALF) from infants with CF compared tocontrols, even in the absence of detectable infection withCF related pathogens, includingP. aeruginosa.12,13Stud-ies by Armstrong et al.,14 however, demonstrated that innewly diagnosed CF infants under the age of 6 months,those with no detectable infection exhibit BAL profilescomparable with control subjects. Furthermore, increasesin inflammatory markers correlate with the developmentand persistence of infection among older children, andBAL inflammatory profiles decrease in the absence of, orwith clearance of infection. Together, these findingsdemonstrate the difficulty of determining the cause ofinflammation in the absence of current infection in youngpatients with CF. Despite the controversy concerning therelation between infection and inflammation, the impor-tance of inflammation in the pathogenesis of CF lungdisease is underscored by the finding that chronic useof high-dose ibuprofen is associated with a significantslowing in the rate of decline of lung function in childrenwith CF.15

Observations such as these have led to the speculationof a possible link between the basic defect in CF andinflammation.11 A number of other hypotheses to explainthe link between a defect atCFTR and chronic lunginfection and excessive inflammation have also been pro-posed. Matsui et al.16 proposed that CF epithelia hyper-absorb water, which leads to mucus and ciliary dysfunc-tion in the airways. Based on in vitro findings, Zabner etal.17 proposed that killing of bacteria in the CF lung isdiminished due to high salt concentrations, which inac-tivate antimicrobial molecules called defensins. How-ever, this theory has been questioned due to controversyconcerning salt levels in CF airway surface fluid. Alter-natively, Imundo et al.18 argued that binding of bacterialpathogens to airway cells or secretions is enhanced due toelevations in the density of receptors, such as asialogan-glioside-1, while Pier et al.19 proposed that CFTR servesas a binding receptor forP. aeruginosa.According to thislatter hypothesis, defective clearance of bacteria in theCF lung is due to decreased amounts of apical CFTR.

Although there is a strong association between the CFgenotype and the pancreatic phenotype, CF patients car-rying the same mutations in theCFTR gene display aremarkable heterogeneity in the severity of lung dis-ease.20,21The basis for this dissociation between the CFgenotype and the pulmonary phenotype is unknown, butthe presence of either modifier genes exclusive of theCFTR locus, or environmental factors, has been sug-gested.21–23

Following identification of theCFTRgene, rapid prog-ress has been made in understanding the molecular basisof CF and the biological properties of the CFTR protein.However, the events leading to the establishment ofchronic lung infection withP. aeruginosa,especially theinflammatory and immunological events and the relation

Abbreviations

BALF Bronchoalveolar lavage fluidCF Cystic fibrosisCFTR Cystic fibrosis transmembrane conductance regulatorIFN-g Interferon-gIL InterleukinMIP Macrophage inflammatory proteinNO Nitric oxideNRAMP-1 Natural resistance-associated macrophage protein 1TNF-a Tumor necrosis factor-aSPF Specific pathogen-free

414 Stotland et al.

between the CF defect and infection, remain largely un-defined. The lack of a suitable animal model has ham-pered progress in this area. The development ofCFTRknockout mice suggests the possible usefulness of amouse model to investigate these issues. Although CFmouse models do not entirely replicate human disease,several prominent features of CF occur in these trans-genic animals. In this review, we focus on the develop-ment of a model of chronic bronchopulmonaryP. aeru-ginosa infection in genetically and immunologicallywell-defined inbred mice, and the development and char-acterization of CF mice. We summarize the current in-formation generated in studies of infections using bothnormal andCFTR knockout mice. We also discuss theusefulness of the mouse model to dissect the role ofcytokines in regulating the inflammatory response in theP. aeruginosa-infected lung.

MOUSE MODELS OF CHRONIC LUNG INFECTIONWITH P. AERUGINOSA

Both acute and chronic models of lungP. aeruginosainfection have been established in several animal speciesincluding rats, guinea pigs, hamsters, mice, mink, sheep,rabbits, and monkeys.24 Several chronic models of lungP. aeruginosainfection in animals are presented in Ta-ble 1. P. aeruginosainfection in normal animals, par-ticularly mice, mimics some of the histopathology seenin the infected CF lung, but, as will be described, impor-tant differences are apparent between mice and humans.It is evident from a survey of the literature of acute vs.chronic lungP. aeruginosainfections that, in order toachieve a chronic infection lasting more than 1 month,the bacteria must be inoculated intratracheally with animmobilizing agent such as agar, agarose, or seaweedalginate. Although in most studies of chronic models,P.aeruginosawas enmeshed in an immobilizing agent, astudy by Nacucchio et al.25 showed that administration ofbacteria in the presence of agar beads produced localinfection in the lungs of rats, similar to that described forinfection induced by the administration of agar beadscontaining bacteria. Otherwise, infection induced by

aerosol exposure to bacteria or by intratracheal adminis-tration of bacteria alone in normal, immunosufficient ani-mals is acute, and bacteria are rapidly cleared from thelung in about 24–48 hr.24 Similarly, immunocompetentindividuals do not develop lung infection withP. aeru-ginosa,a normally commensal organism.

The first experimental model of chronicP. aeruginosalung infection was developed by Cash et al.,26 who usednormal rats infected intratracheally with an inoculum ofbacteria embedded in agar beads. The bacterial load wasnot cleared and proliferated within 3 days, and high num-bers of bacteria were still present at 35 days after infec-tion. Histological examination of infected lungs revealedlesions resembling CF lung disease, including goblet-cellhyperplasia, focal necrosis, and acute and chronic in-flammatory cell infiltration. Pedersen et al.27 also usedthe rat to establish chronicP. aeruginosalung infection,but immobilized the bacteria in seaweed alginate. Anumber of investigators, including Cash et al.26 and Ped-ersen et al.,27 have utilized the rat model to investigatethe protective efficacy ofP. aeruginosacomponents aspotential vaccines against chronic infection and to deter-mine the role of humoral and cellular immune responsesin chronic lung infection.24,28–32Similar to observationsin children with CF, local instillation ofP. aeruginosainone lung results in an inflammatory response in the con-tralateral lung.33 Chronic lungP. aeruginosainfectionhas also been achieved in guinea pigs,34 cats,35 and rhe-sus monkeys.36,37 Similar to the rat model, chronic pro-liferation of P. aeruginosain the lung with neutrophilicinflammation and histological damage similar to that ob-served in the lungs of CF patients were observed to occurin these species as well.

Another model of chronicP. aeruginosainfection hasbeen established in mice. Initially, outbred Swiss CD-1mice were infected intratracheally with mucoidP. aeru-ginosaenmeshed in agarose beads.38 The bacteria pro-liferated in the lungs, and a chronic infection lastingthrough 21 days was established with histopathologicalchanges, including bronchiectasis, similar to those foundin the lungs of CF patients. Instillation of sterile agarbeads alone into the mouse lung caused no local inflam-matory reaction. We have used this model in inbredmouse strains infected via the trachea with a mucoidstrain of P. aeruginosaisolated from a CF patient.39,40

Following infection with doses of bacteria ranging be-tween 104–105 cfu, resistance to infection was observedto vary among inbred mouse strains. Based on mortalityand bacterial load, BALB/c mice were found to be resis-tant and DBA/2 mice were identified as the most suscep-tible strain. Less than 2% of BALB/c mice and approxi-mately 40% of DBA/2 mice succumbed to infectionwithin 3 days. C57BL/6 and A/J mice were found to berelatively susceptible, but only a small percentage ofthese mice succumbed.

TABLE 1—Animal Models of Chronic P. aeruginosaLung Infection

Animal host Immobilizing agent References

Rats Agar Cash et al.26

Rats Seaweed alginate Pedersen et al.27

Guinea pigs Agar Pennington et al.34

Cats Agarose Thomassen et al.35

Rhesus monkeys Agar Cheung et al.36,37

Outbred mice Agarose Starke et al.38

Inbred mice Agar Morissette et al.39

Stevenson et al.40

P. aeruginosa Lung Infection in Mice 415

To further investigate the phenotypic expression ofsusceptibility to chronicP. aeruginosalung infection ininbred mice, studies were performed in our laboratories,using BALB/c mice as aP. aeruginosa-resistant host andC57BL/6 mice as a susceptible host based on the avail-ability of congenic inbred strains and recombinant inbredstrains derived from these two inbred mouse strains.Characterization of the course of infection in this straincombination demonstrated that intratracheal infectionwith a single dose of 105 cfu P. aeruginosa-impregnatedagar beads resulted in the establishment of a prolonged,severe endobronchial infection lasting more than 4 weeksin the majority of susceptible C57BL/6 mice.41,42In con-trast, among resistant BALB/c mice, 67% of the animalscleared the infection by 4 weeks. The bacterial load in thelungs of C57BL/6 mice was significantly higher than thatof BALB/c mice through 3 weeks postinfection. An ex-cessive, predominately neutrophilic inflammatory infil-trate with tissue damage similar to that observed in theinfected CF lung, and acute and organizing pneumonia aswell as atelectasis, were apparent in the lungs of suscep-tible C57BL/6 mice (Fig. 1). Chronic, granulomatous in-flammation, consisting predominately of macrophageswith little or no tissue damage, was observed in the lungsof resistant BALB/c mice. Instillation of sterile agarbeads alone in either mouse strain caused only a mild andtransient granulomatous inflammation. These differencesin lung histopathology inP. aeruginosa-infectedC57BL/6 vs. BALB/c mice suggest possible differencesin the type of T helper cell-mediated immune responseinduced in these hosts.43 However, the nature of the Thelper cell responses inP. aeruginosa-infected resistantand susceptible mice remains to be defined.

In order to determine whether intrinsic differences ininflammatory as well as immune responses may explainthe inherent resistance and susceptibility of BALB/c andC57BL/6 strains of mice, we performed further studies.Gosselin et al.44 observed that within 3 hr of infection,higher levels of tumor necrosis factor-a (TNF-a) mRNAexpression and protein secretion occurred in the lungs ofresistant BALB/c mice infected withP. aeruginosacom-pared to susceptible DBA/2, C57BL/6, and A/J mousestrains. Expression of mRNA for other inflammatory cy-tokines and chemokines, including IL-1a, IL-1b, andmacrophage inflammatory protein (MIP)-1a, was up-regulated to a similar extent in all mouse strains exam-ined. In addition, more efficient nitric oxide (NO) pro-duction in the first 24 hr postinfection appeared to exerta protective role duringP. aeruginosalung infection.Interestingly, studies in CF mice by Kelly and Drumm45

revealed that airway NO production and iNOS mRNAexpression are significantly decreased compared to nor-mal, non-CF mice. It should be noted that these micewere not infected withP. aeruginosa.In vitro studies bythe same laboratory demonstrated that endogenously pro-

duced NO may have a role in preventing the initiation ofbacterial infections in the lung. These findings are con-sistent with observations that exhaled NO levels are re-duced,46 and that iNOS mRNA is markedly reduced inbronchial epithelial cells of CF airways.45,47

Later in infection, alveolar macrophages from suscep-tible C57BL/6 mice produce significantly higher levelsof NO spontaneously and lower LPS or heat-killedP.aeruginosastimulated TNF-a levels, compared toBALB/c mice.42 Furthermore, lung interstitial T cellsfrom day 6 infected susceptible C57BL/6 mice had sig-nificantly reduced proliferative responses in vitro to heat-killed P. aeruginosaas specific antigen, as well as to themitogen concanavalin A, compared to interstitial T cellsfrom resistant BALB/c mice.40 Consistent with sponta-neous production of high levels of NO by alveolar mac-rophages fromP. aeruginosa-infected C57BL/6 mice,suppression of proliferation by T cells from the lungs ofthese mice was found to be mediated by adherent lungcells via the production of NO and prostaglandins. Simi-lar to CF patients, susceptible C57BL/6 mice were foundto have increased serum levels ofP. aeruginosa-specificantibodies.40 Taken together, our results, using a geneti-cally well-defined non-CF host, support the hypothesisthat a gene or genes independent ofCFTR regulate thevariation in susceptibility toP. aeruginosainfection inthe lung, and that the phenotypic expression may be re-lated to the inflammatory and subsequent acquired im-mune responses. This view is consistent with studies inCF patients carryingDF508 mutations, and it demon-strated an association between deficient variants of thehighly polymorphic major serine proteinase,a1-antitrypsin, regulated by a locus on human chromosome14q and earlier onset ofPseudomonaslung infection,higher total IgG, andP. aeruginosa-specific serum anti-bodies.48 It is of interest to note that second to CF,a1-antitrypsin deficiency is the most common lethal disorderamong Caucasian populations.48 Mutations at this geneare associated with lowa1-antitrypsin levels in serumand the risk of development of lung and liver disease.49

Study of chronicP. aeruginosalung infection in ge-netically well-defined inbred mouse strains offers a sig-nificant advantage over other animal models in terms ofthe availability of experimental tools, including mono-clonal antibodies against important inflammatory andimmunologic molecules, recombinant cytokines, andtransgenic and gene knockout mice, includingCFTRknockout mice. Of importance in the identification of thegenetic regulation of resistance to chronic lungP. aeru-ginosainfection, there is currently a worldwide effort tosequence the entire mouse genome. Mouse models haveproven extremely useful for determining the chromosom-al location, physical mapping, and cloning of a candidategene, and the subsequent discovery and cloning of the

416 Stotland et al.

syntenic human gene. For host resistance to infection,this has been most successfully demonstrated for the mu-rineNatural resistance-associated macrophage protein1(Nramp1) gene50 and its human homologue,NRAMP1,51

which control resistance to infection with a diverse spec-trum of intracellular microorganisms. Inbred mousestrains, identified as either resistant or susceptible to

chronicP. aeruginosabronchopulmonary infection in thestudies described above, will no doubt be useful in theidentification of genes other thanCFTR that are respon-sible for regulating host resistance to this infection in CFpatients, as well as in defining the role of inflammatoryand immune responses in host protection to this infec-tion.

Fig. 1. Lungs from representative normal (a), sterile agar bead im-plant (b), day 7 infected C57BL/6 (c) and BALB/c (d), and day 14infected C57BL/6 (e) and BALB/c (f) mice. Mice were inoculatedintratracheally with sterile agar beads 7 days previously or with 1–2× 106 cfu P. aeruginosa trapped in agar beads and killed at days 7and 14 postinfection. In b, arrows indicate sterile agar beads in distalairway surrounded by mild, granulomatous-like inflammation. In c,

the arrow indicates an airway filled with exudate consisting predomi-nately of neutrophils. In d, arrows indicate airways with mild, chronicgranulomatous inflammation. In e, the arrow indicates the persis-tence of acute inflammation in airway. In f, the arrow indicates anairway with little or no inflammation (hematoxylin and eosin; a, b, e,f, original magnification ×100; c, d, ×200). Scale bar = 100 µm.Reprinted with permission of Tam et al.41

P. aeruginosa Lung Infection in Mice 417

CF MOUSE MODELS

The development of CF mouse models represents amilestone in CF research. Within 3 years of the isolationof the CFTR gene, the first mouse models of CF weregenerated.CFTRknockout mice were created using genetargeting in embryonic stem cells to disrupt the endog-enousCFTRgene. A number of variants have been cre-ated to date and their characteristics are summarized inTable 2. It is important to point out that all the mutantmice discussed are characterized by defective cAMP-mediated Cl− ion conductance, but there is a range ofclinical phenotypes from mild to severe in terms of vi-ability and pathology. This latter observation may berelated to variation among the mutant mice in either thelevel of expression of functional CTFR protein or the siteof the mutation within the endogenous mouseCFTRgene. Similar to CF in humans, growth of theCFTRknockout mice is slow.

The first successful generation of a mouse model ofCF came from Snouwaert et al.52 and Clarke et al.53 atthe University of North Carolina. The endogenousCFTRgene in murine embryonic stem-cell lines was targeted sothat the resulting disruptedCFTRgene contained a stopcodon in the coding sequence of exon 10. The targetedcell lines were introduced into early mouse embryos, andthe resulting mice were used to generate heterozygotes,which were crossed to produce homozygous offspring,designatedCFTRm1UNCknockout mice. TheCFTRm1UNC

knockout mice expressed two copies of the defectivegene and lacked functional CFTR. Breeding of thesemice occurred in the expected Mendelian ratio, and therewere no reproductive abnormalities. Although theseCFTR knockout mice had pathological changes in theupper respiratory tract, there were no signs of either in-flammation or bacterial infection in the lungs. However,

the mice developed severe bowel disease, which led tointestinal obstruction, perforation, fatal peritonitis, anddeath of all but one mouse before 40 days. Subsequentstudies revealed that housing these mice under sterileconditions and the use of a nutrient-rich liquid diet couldprevent intestinal obstruction and significantly prolongsurvival.54

About the same time asCFTRm1UNC knockout micewere produced, another strain ofCFTR knockout micewas generated by insertional mutagenesis targeted toexon 10 of theCFTRgene by Dorin et al.55 This mousestrain, designatedCFTRm1HGV, produces about 10% nor-mal CFTR. Approximately 90% of these animals sur-vived and exhibited important hallmarks of human dis-ease, including abnormalities of the colon and vasdeferens. Intestinal obstruction was not apparent andthere were no overt histological abnormalities of the pan-creas. Only one mutant mouse displayed mild pulmonaryhistopathology. Ratcliff et al.56 created another strain ofCFTRmutant mice, using a strategy similar to the NorthCarolina group, by replacement mutation in exon 10. Thephenotype of these mice, designatedCFTRCAM knockoutmice, is very similar to that of theCFTRm1UNCknockoutmice, with high mortality and intestinal obstruction. Inaddition, there was pathology in the lacrimal glands and,in a subset of the mice, there was blockage of pancreaticducts.

Researchers in Texas and Iowa have generated a CFmouse, designatedCFTRm1Bay, by duplication of exon 3in the mouseCFTRgene.57 These mice exhibit a severephenotype with high mortality and intestinal obstruction,but no overt lung disease. Models for the CFDF508mutation have also been generated by introducing thismutation into the endogenous mouseCFTR gene. TheDF508 knockout mice generated by Van Doorninck et

TABLE 2—Mouse Models of Cystic Fibrosis 1

Knockout designation CFTR mutation Phenotype

CFTRm1UNC (52, 53) Stop codon in exon 10; null mutation Severe phenotype with intestinal obstruction and high mortality;no lung disease

CFTRm1HGU (55) Insertional mutagenesis in exon 10;10% CFTR

Mild chromosome with some intestinal blockage; minor pathologyin lungs of one mouse

CFTRCAM (56) Stop codon in exon 10; null mutation Severe phenotype with intestinal obstruction and high mortality;pathology in lacrimal gland and pancreas of some mice; nolung disease

CFTRm1BAY(57) Duplication of exon 3; null mutation Severe phenotype with intestinal obstruction and high mortality;no lung disease

CFTRDF508ROT(58) DF508 mutation; normal CFTR Mild phenotype with mild intestinal pathology and viability; nolung disease

CFTRDF508CAM (59) DF508 mutation; 15% CFTR Severe phenotype with intestinal obstruction and 35% mortality;no lung disease; no pancreatic disease

G551D (60) Missense mutation in exon 11; 4% CFTR Moderate phenotype with reduced incidence of intestinal blockageand 67% survival; no lung disease

CFTRm1HSC(23) Disruption of exon 1; null mutation Severe phenotype with intestinal obstruction and high mortality; nolung disease

1Reference numbers indicated in parentheses.

418 Stotland et al.

al.58 are viable, and they do not show severe diseasessuch as meconium ileus. These CF mice, however, ex-hibit mild pathology with electrophysiological abnor-malities in nasal epithelium, gallbladder, and intestine,and display histological abnormalities in the intestine.The mild pathology of these mutant mice may be ex-plained by possible residualDF508 CFTR activity.DF508 knockout mice were also generated by Colledgeet al.59 In contrast to the null mutants produced by Rat-cliff et al.56 which experienced 80% mortality, theseDF508 mutants exhibited approximately 65% survival.These mutant animals were observed to have pathologysuch as intestinal obstruction and excessive mucus accu-mulation in the small and large intestines, as well aselectrophysiological changes consistent with the CF phe-notype. Similar to other CF knockout mice, no patho-logical changes were apparent in the lungs of any of theDF508−/− mice examined.

A CF mutant mouse carrying the human G551Dmutation in theCFTR gene has also been generated.60

Similar to CF patients with the G551D mutation, thesemutant animals have a reduced incidence of intestinalblockage. Among mutant mice maintained under specificpathogen-free conditions (SPF), approximately 30% diedof intestinal obstruction. In contrast, there was only 27%survival among mutant mice housed in a conventionalanimal facility. This observation suggests the importanceof housing conditions in the viability of CF knockoutmice. Pathology was apparent in the small intestine ofsurviving mutant animals; there was no evidence of ei-ther pancreatic or lung disease. Although the CFTR-related chloride transport was significantly reduced inthese mice, it was further shown that these CF animalsretained approximately 4% residual activity. This level isintermediate between that ofCFTRm1UNCreplacement ornull mutant and theCFTRm1HGU insertional mutant,which retains about 10% residual CFTR activity.

Another strain ofCFTR-deficient mice, designatedCFTRm1HSC knockout mice, has been generated byRozmahel et al.23 by disruption of exon 1 of theCFTRgene. Initially, these mice were generated on a mixedgenetic background and displayed a severe phenotypewith intestinal obstruction and only 30% survival. In sub-sequent studies, the original founder mouse was crossedwith different inbred strains to generate F1 mice of dif-ferent genetic backgrounds, and the heterozygous F1mice were intercrossed to produce homozygousCFTRm1HSC/CFTRm1HSCknockout mice. Experiments inthese mice demonstrated a difference in the duration ofsurvival and the severity of intestinal disease betweenCFTRknockout mice on different genetic backgrounds.A modifier locus was mapped to the proximal region ofmouse chromosome 7. These results suggest that at leastone genetic locus unlinked toCFTRcan modify the se-verity of CF disease in mice. Furthermore, electrophysi-

ological studies in mice with prolonged survival revealedpartial correction of Cl− and Na+ ion transport abnor-malities in these animals, which may influence the se-verity of the phenotype. These results also suggest thatthere is an alternative Cl− channel in mice, exclusive ofCFTR, that can partially compensate for the mutatedCFTRgene.

Based on these observations, Kent et al.61 hypoth-esized that the mixed genetic background of the originalCFTRm1UNC-deficient strain may influence the develop-ment of lung pathology. In order to investigate thispossibility, heterozygote mice from the originalCFTRm1UNC-deficient colony were backcrossed onto theC57BL/6 background for 18 generations to ensure 100%homozygosity for C57BL/6 strain alleles, as determinedby microsatellite typing. The novel congenic strain wasdesignated the C57BL/6-CFTRm1UNC/CFTRm1UNC

knockout mouse. Although no infectious pathogens com-mon to CF are apparent in the airways, these mice werefound to develop spontaneous and progressive lung dis-ease, with features consistent with obstructive small air-way disease, including the accumulation of neutrophils inthe interstitial regions of the lung at age 1 month.

LUNG INFECTION STUDIES IN CF MICE

A limited number of infection studies have been car-ried out on the variousCFTR knockout mice. Speciesdifferences between humans and rodents may raisespeculation concerning the validity of these studies in CFmice. For example,CFTR is expressed predominately inthe submucosal glands in man, while mice have a relativepaucity of these glands. Sodium hyperabsorption is char-acteristic of the nasal tissue ofCFTRknocknot mice, andof the nasal and tracheal tissues in CF patients. The tra-cheal epithelium of CF mice, however, is hypoabsorbing.Nevertheless, CF mouse models of infection, particularlyunder controlled conditions not possible in the clinicalsetting, should prove useful in unraveling the relationshipbetween infection and inflammation in the CF lung.Snouwaert et al.62 reported that mixed genetic back-ground CFTRm1UNC-deficient mice compared toCFTRheterozygous mice did not differ in their ability to clearsingle or repetitive intranasal infection withS. aureus.Investigators from the same laboratory recently reportedthat induction of allergic airway disease induced goblet-cell hyperplasia, increased mucin gene expression, andincreased production of mucus in both normal and CFmice, but did not lead to chronic lung disease inCFTR-deficient mice following repetitive, intratracheal instilla-tion of P. aeruginosaor S. aureus.63

Using CFTRm1HGU knockout mice, Davidson et al.64

reported higher bacterial loads and more pronounced dis-ease in the lungs of deficient compared to littermate con-trols after repeated aerosol challenges withS. aureus

P. aeruginosa Lung Infection in Mice 419

and Burkholderia cepacia,a bacterium found with in-creasing frequency in CF patients. Histological studies inS. aureus-infected mice showed pathological changes inthe lungs of both CF and wild-type mice, but the inci-dence of goblet-cell hyperplasia, mucus retention, andbronchiolitis was significantly higher in CF than wild-type mice. A distinct pathology was apparent in miceinfected with B. cepacia.Histological changes in thelungs of infected non-CF mice were milder and morefocal than those in CF mice. The latter mice had in-creased pneumonia and mucus retention, destruction ofsmall airways, and extensive edema.

Two groups reported findings using the bead model toinduce chronicP. aeruginosalung infection in CF mice.Van Heeckeren et al.65 usedCFTRm1UNCknockout miceand observed that mortality among CF mice was signifi-cantly greater than normal animals by 10 days after in-tratracheal infection with 6 × 104 cfu of a mucoid isolateof P. aeruginosaenmeshed in agarose beads. At 3 dayspostinfection, no significant difference in bacterial bur-den was observed betweenCFTR-deficient and wild-typelittermates, while significantly higher concentrations ofthe inflammatory cytokines and chemokines, TNF-a,MIP-2, and KC/N51, were found in the BALF of knock-out compared to control mice. It should be pointed outthat a unilateral method of infection, in which only theright lung was instilled with bacteria, was employed forthis study, and that the genetic background of theCFTR-deficient mice was not indicated.

Our laboratories used C57BL/6-CFTRm1UNC/CFTRm1UNC knockout mice backcrossed to C57BL/6mice and housed under SPF conditions.66 After thefourth backcross, the mice were confirmed to be 100%homozygous for the C57BL/6 background by microsat-ellite typing. These were the same animals used by Kentet al.54 in the studies described above, except that ani-mals derived from backcross generations 10–12 wereused and the mice were housed in an SPF facility. Underthese conditions, we have observed that C57BL/6-CFTRm1UNC/CFTRm1UNC knockout mice in backcrossgeneration 18 developed spontaneous lung disease (un-published observations). This finding suggests that ge-netic and not environmental factors are responsible forthe development of the previously described spontaneouslung disease.54

In our study, a significantly higher mortality was ob-served within 6 days, as well as a significantly higherbacterial burden in the lungs of C57BL/6-CFTRm1UNC/CFTRm1UNCknockout mice compared to littermate con-trol mice following bilateral, intratracheal infection with105 cfu mucoid P. aeruginosain agar beads.66 Therewere no remarkable differences in lung histopathologybetweenCFTR-deficient and wild-type mice, and bothgenotypes developed severe bronchopneumonia. Further-more, the magnitude of the neutrophil inflammatory re-

sponse in the BALF was similar inCFTR-deficient ver-sus CFTR wild-type mice with the same geneticbackground, suggesting that the accumulation of neutro-phils in the lung may be regulated by a gene or genesmapping outside ofCFTR, as suggested by our earlierstudies.41,42 We conclude that the generation of thesedata in CF mice supports the concept of the usefulness ofCFTR knockout mice to study lung infections withpathogens important in CF patients.

ROLE OF CYTOKINES IN REGULATINGINFLAMMATION IN THE CF AIRWAY

The process of eradicatingP. aeruginosafrom thelung likely represents a complex network of cytokinesand effector cells, with interplay between innate and im-mune mechanisms of host defense.67 Observations madefrom studies of BALF collected from CF patients and ininitial studies inP. aeruginosa-infected CF and non-CFmice suggest that airway damage likely results from thesum of acute and chronic episodes of infection and in-flammation. The regulatory mechanisms involved in con-trolling this process are likely to be very finely tuned but,paradoxically, redundant. Excess production of eitherpro- or anti-inflammatory cytokines by airway cells maycompromise the benefits of the inflammatory response.

IL-8 has been implicated in the pathogenesis of a num-ber of inflammatory diseases, including CF. Several re-cent studies suggest an important contribution of the in-flammatory process in the development of pathologicalchanges in the lung. Immunohistochemical analysis ofthe CF bronchial submucosal glands in patients homo-zygous for theDF508 deletion expressed elevated levelsof the endogenous chemokine interleukin IL-8, but notcytokines like IL-1b and IL-6.68 In addition, the expo-sure of CF and non-CF bronchial gland cells to an el-evated extracellular Cl− concentration substantially in-creased the release of IL-8.68 These studies alsodemonstrated that dexamethasone was able to inhibit therelease of IL-8 by control cells, but not by CF glandcells.68 These findings indicate that control of IL-8 re-lease by CF submucosal gland cells is dysregulated.

It was shown recently that cultured CF bronchial glandcells express activated NFkB due to a lack of cytosolicIkBa expression.69 These investigators also demon-strated that the isoflavone genistein was able to reduceboth constitutive andPseudomonasinfection-inducedexpression of IL-8 by inducing IkBa in these cells.69

Overall, these results suggest one of the possible mecha-nisms that explains chronic lung inflammation in CF pa-tients. Consistent with these findings, it was reported thatCF patients continued to have elevated IL-8 levels inbronchoalveolar lavages, even after lung transplanta-tion.70 Interestingly, these patients were also found to

420 Stotland et al.

have lower levels of IL-10 in their lungs compared tonon-CF transplantation patients.

The anti-inflammatory cytokine IL-10 has been pro-posed to play an essential role in maintaining homeosta-sis of the inflammatory process in CF airways. IL-10suppresses the production of many key pro-inflammatorycytokines, such as interferon-g (IFN-g), TNF-a, andother chemokines.71–73 Bonfield et al.74 demonstratedthat BALF from CF patients contained significantlyhigher levels of the pro-inflammatory cytokines TNF-a,IL-1, and IL-8, and significantly less IL-10, than BALFfrom healthy control subjects. Interestingly, intracellularcytokine staining revealed that a significantly higher per-centage of alveolar macrophages from CF patients pro-duced more TNF-a, IL-1, IL-6, IL-8, and IL-10 than cellsfrom healthy controls.

In a subsequent study by these investigators, it wasdemonstrated that bronchial epithelial cells from healthyairways constitutively produced IL-10, while epithelialcell production of this cytokine was defective in thechronically infected CF lung.75 Constitutive productionof IL-10 by normal epithelia may suppress macrophageas well as neutrophil effector functions, including accu-mulation, production of cytokines, enzymes, and oxygenradicals as well as other tissue-destructive molecules, andmacrophage antigen presentation, leading to aberrant im-mune responses.71 Thus, in addition to alveolar macro-phages, the bronchial epithelium may play an importantrole in regulating the local inflammatory and immuneresponses in the CF lung. In addition, CD4+ T cells fromCF patients have also been shown to secrete significantlyless IL-10 after polyclonal activation in comparison tosimilarly stimulated cells from healthy control patients.76

On the other hand, Noah et al.77 failed to observe dimin-ished levels of IL-10 in BALF of young CF patients. Thebasis of the discrepancy between the observations ofBonfield et al.74 and Noah et al.77 may be due to knowntechnical difficulties in accurately measuring the concen-trations of soluble molecules in respiratory fluids.78,79

Thus, whether or not IL-10 plays a role in regulatinginflammation in the CF lung remains unclear.

Clinically relevant information has been derived frommouse models. These observations indicate that IL-10plays an important role in infection with a variety ofpathogens, including bacteria, in addition to regulatinginflammation.80 Treatment in vivo with recombinant IL-10 has been shown to protect mice from LPS-inducedshock associated with infection with Gram-negative bac-teria. Lethality in this model is considered to be due tooverproduction of IFN-g, TNF-a, IL-1, and IL-6. Treat-ment of mice with anti-IL-10 polyclonal antibody beforeadministration of LPS resulted in substantial increases inserum TNF-a and MIP-2 and significantly increasedmortality. In contrast, IL-10 production during some in-fections can be detrimental. For example, anti-IL-10

treatment has been shown to increase survival in a mu-rine model ofKlebsiella pneumoniaeinfection.81 IL-10production has also been observed to be detrimental inlung infection withStreptococcus pneumoniaein mice.82

The importance of IL-10 in these models of infection hasbeen attributed to the ability of this cytokine to maintainan appropriate balance between pro- and anti-inflammatory cytokines. When IL-10 administration re-sults in protection to the host, it is generally thought thatthe effect of IL-10 is due to the downregulation of septicshock-inducing cytokines, such as IFN-g and TNF-a.Conversely, when IL-10 is detrimental to the host, it hasbeen proposed that the downregulation of important pro-inflammatory cytokines prevents the generation of anappropriate immune response necessary for completeclearance of invading pathogens.

Using a model ofP. aeruginosapneumonia, Sawa etal.83 observed a significant increase in IL-10 productionin the lungs of BALB/c mice infected with cytotoxicstrains ofP. aeruginosa.A model of acuteP. aeruginosainfection that resulted in significant mortality within 48hr was used in this study. Treatment with recombinantIL-10 decreased lung injury and increased survival. Astudy by Yu et al.,84 using an acute model of repeatedaerosol exposure toP. aeruginosa,demonstrated in-creased pathology in IL-10 gene knockout mice bred ona C57BL/6 background relative to wild-type C57BL/6mice. Studies are in progress in our laboratories to ad-dress the role of IL-10 in the mouse model of chroniclung infection induced by intratracheal inoculation withP. aeruginosaenmeshed in agar beads.

In addition to IL-10, other anti-inflammatory cyto-kines such as IL-4, IL-11, and the IL-4-related cytokineIL-13 may play a role in regulating excessive inflamma-tion in theP. aeruginosa-infected lung and in promotingthe development of protective immune mechanisms. IL-4is a pleiotropic cytokine, which is produced primarily bythe Th2 subset of T lymphocytes and mast cells; it in-duces the differentiation of precursor CD4+ T cells toTh2 cells and regulates immunoglobulin (Ig) classswitching to IgE.85 Because of its ability to inhibit thedifferentiation of Th1 cells, IL-4 may have potent anti-inflammatory properties. IL-11 is also a pleiotropic cy-tokine with hematopoietic and thrombopoietic activi-ties; it has been found to be effective in several animalmodels of acute and chronic inflammation.86 The anti-inflammatory properties of IL-11 are likely due to itsability to downregulate the production of pro-in-flammatory cytokines and NO. IL-13, which is closelyrelated to IL-4 and binds the IL-4a receptor, is producedby the Th2 cells.87,88 IL-13 has similar immunoregula-tory functions to IL-4 and was recently shown to havepotent anti-inflammatory properties in animal models ofallergic responses in the lung. Although studies address-ing the anti-inflammatory role of these cytokines in the

P. aeruginosa Lung Infection in Mice 421

P. aeruginosa-infected lung have not been performed inthe mouse model of chronic lung infection, recent studiesin the acute infection model demonstrated enhancedclearance ofP. aeruginosafrom the lungs of transgenicmice selectively expressing IL-4 in the airway epithe-lium.89 Interestingly, enhanced levels of IL-4 were asso-ciated with acute neutrophilic infiltration into the lungs,suggesting that modulation of the quality of the inflam-matory response rather than the quantity of inflammatorycells may be an important consideration in the infectedCF lung.

CONCLUSIONS AND FUTURE DIRECTIONS

This review has demonstrated the usefulness of mousemodels of chronicP. aeruginosalung infection for de-fining the role of theCFTRgene as well as backgroundmodifier genes in host susceptibility to this infection.Overall, studies from mouse models, using geneticallyand immunologically well-defined inbred strains, as wellas studies in CF patients, support the concept of the roleof local inflammation in both protection and pathogen-esis of bronchopulmonary infections with this bacterium.Mouse models provide useful tools for identifying thenetwork of important pro-inflammatory and anti-inflammatory cytokines and the complex interactionsamong these molecules in the genesis of chronicP. ae-ruginosainfection in the CF lung. Studies are warrantedto define the role of the anti-inflammatory cytokines IL-4, IL-10, IL-11, and IL-13 in maintaining the balancebetween protective and deleterious inflammatory re-sponses during chronicP. aeruginosalung infection. Ad-ditional studies, particularly those usingCFTRknockoutmice, are required to define the molecular and cellularbases of a possible defect in cytokine regulation in asso-ciation with CF lung disease. A better understanding ofthe role of anti-inflammatory mediators in chronicP.aeruginosalung infection will provide a rationale for thedevelopment of novel immunomodulatory measures ca-pable of ameliorating or modulating the chronic inflam-mation associated with CF lung disease.

ACKNOWLEDGMENTS

We thank Dr. David Eidelman for critical reading ofthe manuscript and helpful advice. P.K.S. was supportedby a studentship from Fonds de la Recherche en Sante´ duQuebec and Fonds pour la Formation de Chercheurs etl’Aide a la Recherche (FRSQ-FCAR, Que´bec). Researchin the laboratories of D.R. and M.M.S. is supported bythe Canadian Cystic Fibrosis Foundation.

REFERENCES

1. Davis PB, Drumm M, Konstan MW. Cystic fibrosis. Am J RespirCrit Care Med 1996;154:1229–1256.

2. Debraekeleer M, Mari C, Verlingue C, Allard C, Leblanc JP,Simard F, Aubin G, Ferec C. Clinical features of cystic fibrosispatients with rare genotypes in Saguenay Lac-Saint-Jean (Quebec,Canada). Ann Genet (Paris) 1997;40:205–208.

3. Andersen DH. Cystic fibrosis of the pancreas and its relation toceliac disease: a clinical and pathologic study. Am J Dis Child1938;56:344–399.

4. Davidson DJ, Porteus DJ. The genetics of cystic fibrosis lungdisease. Thorax 1998;53:389–397.

5. Welsh JM, Tsui LC, Boat RF, Beadet AL. Cystic fibrosis. In:Scriver C, Beadet AL, Sly WS, Valle D, editors. The metabolicand molecular bases of inherited disease. New York: McGraw-Hill; 1995. p 3799–3876.

6. Kerem E. The role ofPseudomonas aeruginosain the pathogen-esis of lung disease in cystic fibrosis—more questions than an-swers. Pediatr Pulmonol [Suppl] 1997;16:265–266.

7. Cripps AW, Dunkley ML, Clancy RC, Kyd J. Pulmonary immu-nity to Pseudomonas aeruginosa.Immunol Cell Biol 1995;73:418–424.

8. Koch C, Høiby N. Pathogenesis of cystic fibrosis. Lancet 1993;341:1065–1069.

9. Berger M. Inflammation in the lung in cystic fibrosis: a viciouscycle that does more harm than good? Clin Rev Allergy 1991;9:119–142.

10. Berger M, Konstan MW. Immunopathogenesis of cystic fibrosislung disease. In: Yankaskas J, Knowles M, editors. Cystic fibrosisin adults. Philadelphia: Lippincott-Raven; 1999. p 115–143.

11. Konstan MW, Berger M. Current understanding of the inflamma-tory process in cystic fibrosis: onset and etiology. Pediatr Pul-monol 1997;24:137–142.

12. Khan TA, Wagener JS, Bost T, Martinez J, Accurso FJ, RichesDW. Early pulmonary inflammation in infants with cystic fibrosis.Am J Respir Crit Care Med 1995;151:1075–1082.

13. Armstrong DS, Grimwood K, Olinsky A, Phelan PD. Lower re-spiratory infection and inflammation in infants with newly diag-nosed cystic fibrosis. Br Med J [Clin Res] 1995;310:1571–1573.

14. Armstrong DS, Grimwwod K, Carlin JB, Carzino R, Gutierrez JP,Hull J, Olinsky A, Phelan EA, Robertson CF, Phelan PD. Lowerairway inflammation in infants and young children with cysticfibrosis. Am J Respir Crit Care Med 1997;156:1197–1204.

15. Konstan MW, Byard PJ, Hoppel CH, Davis PB. Effect of high-dose ibuprofen in patients with cystic fibrosis. N Engl J Med1995;332:848–854.

16. Matsui H, Grubb BR, Tarran R, Randell SH, Gatzy JT, Davis CW,Boucher RC. Evidence for periciliar liquid layer depletion, notabnormal ion composition, in the pathogenesis of cystic fibrosisairway disease. Cell 1998;95:1005–1015.

17. Zabner J, Smith J, Karp P, Widdicombe J, Welsh M. Loss ofCFTR chloride channels alters salt absorption by cystic fibrosisairway epithelial cells in vitro. Mol Cell 1998;2:397–403.

18. Imundo L, Barasch J, Prince A, Al-Awqati Q. Cystic fibrosisepithelial cells have a receptor for pathogenic bacteria on theirapical surface. Proc Natl Acad Sci USA 1995;92:3019–3023.

19. Pier GB, Grout M, Zaidi TS, Olsen JC, Johnson LG, YankaskasJR, Goldberg JB. Role of mutant CFTR in hypersusceptibility ofcystic fibrosis patients to lung infections. Science 1996;271:64–67.

20. Santis G, Osborne L, Knight RA, Hodson ME. Independent ge-netic determinants of pancreatic and pulmonary status in cysticfibrosis. Lancet 1990;336:1081–1084.

21. Hamosh A, Corey M. Correlation between genotype and pheno-type in patients with cystic fibrosis. The cystic fibrosis genotype-phenotype consortium. N Engl J Med 1993;329:1308–1313.

22. Santis G, Osborne L, Knight R, Smith M, Davison R, Hodson M.Genotype-phenotype relationship in cystic fibrosis: results from

422 Stotland et al.

the study of monozygotic and dizygotic twins with cystic fibrosis.Pediatr Pulmonol [Suppl] 1992;14:239.

23. Rozmahel R, Wilschanski M, Matin A, Plyte S, Oliver M, Auer-bach W, Moore A, Forstner J, Durie P, Nadeau J, Bear C, Tsui LC.Modulation of disease severity in cystic fibrosis transmembraneconductance regulator deficient mice by a secondary genetic fac-tor. Nat Genet 1996;12:280–287.

24. Johansen HK. Potential of preventingPseudomonas aeruginosalung infections in cystic fibrosis patients: experimental studies inanimals. Acta Pathol Microbiol Immunol Scand 1996;104:5–42.

25. Nacucchio MC, Cerquetti MC, Meiss RP, Sordelli DO. Role ofagar beads in the pathogenicity ofPseudomonas aeruginosain therat respiratory tract. Pediatr Res 1984;18:295–296.

26. Cash HA, Woods DE, McCullough B, Johanson WE, Bass JA. Arat model of chronic respiratory infection withPseudomonas ae-ruginosa.Am Rev Respir Dis 1979;119:453–459.

27. Pedersen SS, Shand GH, Hansen BL, Hansen GN. Induction ofexperimental chronicPseudomonas aeruginosalung infectionwith P. aeruginosaentrapped in alginate microspheres. ActaPathol Microbiol Scand 1990;98:203–211.

28. Klinger JD, Cash HA, Wood RE, Miler JJ. Protective immuniza-tion against chronicPseudomonas aeruginosapulmonary infec-tion in rats. Infect Immun 1983;39:1377–1384.

29. Woods DE, Byron LE. Studies on the ability of alginate to act asa protective immunogen against infection withPseudomonas ae-ruginosain animals. J Infect Dis 1985;151:581–588.

30. Johansen HK, Espersen F, Pedersen SS, Hougen HP, Rygaard J,Høiby N. ChronicPseudomonas aeruginosalung infection in nor-mal and athymic rats. Acta Pathol Microbiol Immunol Scand1993;101:207–225.

31. Johansen HK, Hougen HP, Rygaard J, Høiby N. Interferon-gamma (IFN-g) treatment decreases the inflammatory response inchronic Pseudomonas aeruginosapneumonia in rats. Clin ExpImmunol 1996;103:212–218.

32. Pier GB, Small GJ, Warren HB. Protection against mucoidPseu-domonas aeruginosain rodent models of endobronchial infection.Science 1990;249:537–540.

33. Terashima T, Matsubara H, Nakama M, Sakamaki F, Waki Y,Soejima K, Tasaka S, Nakamura H, Sayama K, Ishizaka A, Ka-nazawa M. Local Pseudomonas instillation induces contralaterallung injury and plasma cytokines. Am J Respir Crit Care Med1996;153:1600–1605.

34. Pennington JE, Hickey WF, Blackwood LL, Arnaut MA. Activeimmunization with lipopolysaccharidePseudomonasantigen forchronic Pseudomonasbronchopneumonia in guinea pigs. J ClinInvest 1981;68:1140–1148.

35. Thomassen MJ, Klinger JD, Winnie GB, Wood RE, Burtner C,Tomashefski JF, Horowitz JG, Tandler B. Pulmonary cellular re-sponse to chronic irritation and chronicPseudomonas aeruginosapneumonia in cats. Infect Immun 1984;45:741–747.

36. Cheung AT, Moss RB, Leong AB, Novick WJ. ChronicPseudo-monas aeruginosaendobronchitis in rhesus monkeys. I. Effects ofpentoxifylline on neutrophil influx. J Med Primatol 1992;21:357–362.

37. Cheung AT, Moss RB, Kurland G, Leong AB, Novick WJ.ChronicPseudomonas aeruginosaendobronchitis in rhesus mon-keys. II. A histopathologic analysis. J Med Primatol 1993;22:257–262.

38. Starke JR, Edwards MS, Langston C, Baker CJ. A mouse modelof chronic pulmonary infection withPseudomonas aeruginosaandPseudomonas cepacia.Pediatr Res 1987;22:698–702.

39. Morissette C, Skamene E, Gervais F. Endobronchial inflammationfollowing Pseudomonas aeruginosainfection in resistant and sus-ceptible strains of mice. Infect Immun 1995;63:1718–1724.

40. Stevenson MM, Kondratieva TK, Apt AS, Tam MF, Skamene E.

In vitro and in vivo T cell responses in mice during bronchopul-monary infection with mucoidPseudomonas aeruginosa.ClinExp Immunol 1995;99:98–105.

41. Tam MF, Snipes GJ, Stevenson MM. Characterization of chronicbronchopulmonaryPseudomonas aeruginosainfection in resistantand susceptible inbred mice. Am J Respir Cell Mol Biol 1999;20:710–719.

42. Sapru K, Stotland PK, Stevenson MM. Quantitative and qualita-tive differences in bronchopulmonary inflammatory cells inPseu-domonas aeruginosa-resistant and susceptible mice. Clin Exp Im-munol 1999;115:103–109.

43. O’Garra A. Cytokines induce the development of functionallyheterogenous T helper cell subsets. Immunity 1998;8:275–283.

44. Gosselin D, DeSanctis J, Boule M, Skamene E, Matouk C, Rad-zioch D. Role of tumor necrosis factor alpha in innate resistanceto mouse pulmonary infection withPseudomonas aeruginosa.In-fect Immun 1995;63:3272–3278.

45. Kelly TJ, Drumm ML. Inducible nitric oxide synthase expressionis reduced in cystic fibrosis murine and human airway epithelialcells. J Clin Invest 1998;102:1200–1207.

46. Grasemann H, Michler E, Wallot M, Ratjen F. Decreased concen-tration of exhaled nitric oxide (NO) in patients with cystic fibrosis.Pediatr Pulmonol 1997;24:173–177.

47. Meng Q, Springall DR, Bishop AE, Morgan K, Evans TJ, HabibS, Gruentert DC, Gyi KM, Hodson ME, Yacoug TJ, Polak JM.Lack of inducible nitric oxide synthase in bronchial epithelium: apossible mechanism of susceptibility in infection with cystic fi-brosis. J Pathol 1998;184:323–331.

48. Doring G, Krogh-Johanssen H, Weidinger S, Høiby N. Allotypesof a1-antitrypsin in patients with cystic fibrosis, homozygous andheterozygous for DeltaF508. Pediatr Pulmonol 1994;18:3–7.

49. Crystal RG.a1-antitrypsin deficiency, emphysema, and liver dis-ease. Genetic basis and strategies for therapy. J Clin Invest 1990;85:1343–1352.

50. Vidal SM, Malo D, Vogan K, Skamene E, Gros P. Natural resis-tance to infection with intracellular parasites: isolation of a can-didate forBcg.Cell 1993;73:469–485.

51. Cellier MG, Govoni G, Vidal S, Kwan T, Groulx N, Liu J,Sanchez F, Skamene E, Schurr E, Gros P. Human natural resis-tance-associated macrophage protein, cDNA cloning, chromo-somal mapping, genomic organization and tissue specific expres-sion. J Exp Med 1994;180:1741–1752.

52. Snouwaert JN, Brigman KK, Latour AM, Malouf NN, BoucherRC, Smithies O, Koller BH. An animal model for cystic fibrosismade by gene targeting. Science 1992;257:1083–1088.

53. Clarke LL, Grubb BR, Gabriel SE, Smithies O, Koller BH,Boucher RC. Defective epithelial chloride transport in a gene-targeted mouse model of cystic fibrosis. Science 1992;257:1125–1128.

54. Kent G, Oliver M, Foskett JK, Frndova H, Durie P, Forstner J,Forstner GG, Riordan JR, Percy D, Buchwald M. Phenotypicabnormalities in long-term surviving cystic fibrosis mice. PediatrRes 1996;40:223–241.

55. Dorin JR, Dickinson P, Alton EW, Smith SN, Geddes DM,Stevenson BJ, Kimber WL, Fleming S, Clarke AR, Hooper ML,Anderson L, Beddington RSP, Porteous DJ. Cystic fibrosis in themouse by targeted insertional mutagenesis. Nature 1992;359:211–215.

56. Ratcliff R, Evans MJ, Cuthbert AW, MacVinish LJ, Foster D,Anderson JR, Colledge WH. Production of a severe cystic fibrosismutation in mice by gene targeting. Nat Genet 1993;4:35–41.

57. O’Neal WK, Hasty P, McCray PB, Casey B, Rivera-Pe´rez J,Welsh MJ, Beaudet AL, Bradly A. A severe phenotype in micewith a duplication of exon 3 in the cystic fibrosis locus. Hum MolGenet 1993;2:1561–1569.

P. aeruginosa Lung Infection in Mice 423

58. van Doorninck JH, French PJ, Verbeek E, Peters RH, Morreau H,Bijman J, Scholte BJ. A mouse model for the cystic fibrosisDF508 mutation. EMBO J 14;1995:4403–4411.

59. Colledge WH, Abella BS, Southern KW, Ratcliff R, Jiang C,Cheng SH, MacVinish LJ, Anderson JR, Cuthbert AW, Evans MJ.Generation and characterization of aDF508 cystic fibrosis mousemodel. Nat Genet 10:1995:445–451.

60. Delany SJ, Alton EW, Smith SN, Lunn DP, Farley R, LovelockPK, Thomson SA, Hume DA, Lamb D, Porteous DJ, Doring JR,Wainwright BJ. Cystic fibrosis mice carrying the missense muta-tion G551D replicate human genotype-phenotype correlations.EMBO J 15;1996:955–963.

61. Kent G, Iles R, Bear CE, Huan LJ, Griesenbach U, McKerlie C,Frndova H, Ackerly C, Gosselin D, Radzioch D, O’Brodovich H,Tsui LC, Buchwald M, Tanswell AK. Lung disease in mice withcystic fibrosis. J Clin Invest 1997;100:3060–3069.

62. Snouwaert JN, Brigman KK, Latour AM, Iraj E, Schwab U, Gil-mour MI, Koller BH. A murine model for cystic fibrosis. Am JRespir Crit Care Med 1995;151:59–64.

63. Cressman VL, Hicks EM, Funkhouser WK, Backlund DC, KollerBH. The relationship of chronic mucin secretion to airway diseasein normal and CFTR-deficient mice. Am J Respir Cell Mol Biol1998;19:853–866.

64. Davidson DJ, Dorin JR, McLachlan G, Ranaldi V, Lamb D,Doherty C, Govan J, Porteous DJ. Lung disease in the cysticfibrosis mouse exposed to bacterial pathogens. Nat Genet 1995;9:351–357.

65. van Heeckeran A, Walenga R, Konstan MW, Bonfield T, DavisPB, Ferkol T. Excessive inflammatory response of cystic fibrosismice to bronchopulmonary infection withP. aeruginosa.J ClinInvest 1997;100:2810–2815.

66. Gosselin D, Stevenson MM, Cowley EA, Griesenbach U, Eidel-man DH, Boule´ M, Tam MF, Kent G, Skamene E, Tsui LC,Radzioch D. Impaired ability ofCftr knockout mice to controllung infection withP. aeruginosa.Am J Respir Crit Care Med1998;152:2111–2118.

67. Bals R, Weiner DJ, Wilson JM. The innate immune system incystic fibrosis lung disease. J Clin Invest 1999;103:303–307.

68. Tabary O, Zahm JM, Hinnrasky J, Couetil JP, Cornillet P, Gue-nounou M, Gaillard D, Puchelle E, Jacquot J. Selective up-regulation of chemokine IL-8 expression in cystic fibrosis bron-chial gland cells in vivo and in vitro. Am J Pathol 1998;153:921–930.

69. Tabary O, Escotte S, Couetil JP, Hubert D, Dusser D, Puchelle E,Jacquot J. Genestein inhibits constitutive and inducible NFkappaBactivation and decreases IL-8 production by human cystic fibrosisbronchial gland cells. Am J Pathol 1999;155:473–481.

70. Dosanjh AK, Elashoff D, Robbins RC. The bronchoalveolar la-vage fluid of cystic fibrosis lung transplant recipients demon-strates increased IL-8 and elastase and decreased IL-10. J Inter-feron Cytok Res 1998;18:851–854.

71. Moore KW, O’Garra A, de Waal Malefyt R, Vieira P, MosmannTR. Interleukin 10. Annu Rev Immunol 1993;11:165–190.

72. Tebo JM, Kim HS, Gao J, Armstrong DA, Hamilton TA. Inter-leukin-10 suppresses IL-10 gene transcription by inhibiting theproduction of class I interferon. Blood 1998;92:4742–4749.

73. Kim HS, Armstrong D, Hamilton TA, Tebo JM. IL-10 suppressesLPS-induced KC mRNA expression via a translation-dependentdecrease in mRNA stability. J Leukocyte Biol 1998;64:33–39.

74. Bonfield TL, Panuska JR, Konstan MW, Hilliard KA, Hilliard JB,Ghnaim H, Berger M. Inflammatory cytokines in cystic fibrosislungs. Am J Respir Crit Care Med 1995;152:2111–2118.

75. Bonfield TL, Konstan MW, Burfeind P, Panuska JR, Hilliard JB,Berger M. Normal bronchial epithelial cells constitutively producethe anti-inflammatory cytokine interleukin-10, which is down-regulated in cystic fibrosis. Am J Respir Cell Mol Biol 1995;13:257–261.

76. Moss RB, Bocian RC, Hsu YP, Dong HY, Kemna M, Wei T,Gardner P. Reduced IL-10 secretion by CD4+ T lymphocytesexpressing mutant cystic fibrosis transmembrane conductanceregulator (CFTR). Clin Exp Immunol 1996;106:374–388.

77. Noah TL, Black HR, Cheng PW, Wood RE, Leigh MW. Nasaland bronchoalveolar lavage fluid cytokines in early cystic fibrosis.J Infect Dis 1997;175:638–647.

78. von Wichert PK, Joseph K, Mu¨ller B, Franck WM. Bronchoal-veolar lavage: quantitation of intraalveolar fluid? Am Rev RespirDis 1993;147:148–152.

79. Reynolds HY. Bronchoalveolar lavage. Am Rev Respir Dis 1987;135:250–263.

80. Rennick DM, Fort MM, Davidson NJ. Studies with IL-10 −/−mice: an overview. J Leukocyte Biol 1997;61:389–396.

81. Greenberger MJ, Strieter RM, Kunkel SL, Danforth JM, GoodmanRE, Standiford TJ. Neutralization of IL-10 increases survival in amurine model ofKlebsiella pneumonia. J Immunol 1995;155:722–729.

82. van der Poll T, Marchant A, Keogh CV, Goldman M, Lowry SF.Interleukin-10 impairs host defense in murine pneumococcalpneumonia. J Infect Dis 1996;174:994–1000.

83. Sawa T, Corry DB, Gropper MA, Ohara M, Kurahashi K, Wiener-Kronish JP. IL-10 improves lung injury and survival inPseudo-monas aeruginosapneumonia. J Immunol 1997;159:2858–2866.

84. Yu H, Hanes M, Chrisp CE, Boucher JC, Deretic V. Microbialpathogenesis in cystic fibrosis—pulmonary clearance of mucoidPseudomonas aeruginosaand inflammation in a mouse model ofrepeated respiratory challenge. Infect Immun 1998;66:280–288.

85. Hart PH, Vitti GF, Burgess DR, Whitty GA, Piccoli DS, HamiltonTA. Potential anti-inflammatory effects of interleukin 4: suppres-sion of human monocyte tumor necrosis factora, interleukin 1,and prostaglandin E2. Proc Natl Acad Sci USA 1989;86:3803–3807.

86. Trepicchio WL, Bozza M, Pedneault G, Dorner AJ. Recombinanthuman IL-11 attenuates the inflammatory response through down-regulation of proinflammatory cytokine release and nitric oxideproduction. J Immunol 1996;157:3627–3634.

87. Minty A, Chalon P, Derocq JM, Dumont X, Guillemot JC,Kaghad M, Labit C, Leplatois P, Liauzan P, Miloux B, Minty C,Cassellas P, Loison G, Lupker J, Shire D, Ferrara P, Caput D.Interleukin 13 is a new human lymphokine regulating inflamma-tory and immune responses. Nature 1993;362:248–250.

88. Watson ML, White AM, Campbell EM, Smith AW, Uddin J,Yoshimura T, Westwick J. Anti-inflammatory actions of interleu-kin 13: suppression of tumor necrosis factor-a and antigen-induced leukocyte accumulation in the guinea pig lung. Am JRespir Cell Mol Biol 1999;20:1007–1012.

89. Jain-Vora S, LeVine AM, Chroneos Z, Ross GF, Hull WM, Whit-sett JA. Interleukin-4 enhances pulmonary clearance ofPseudo-monas aeruginosa.Infect Immun 1998;66:4229–4236.

424 Stotland et al.