~ Pergamon Comp. lmmun. Microbiol. Infect. Dis. Vol. 17, No. I, pp. 1-13, 1994 Copyright © 1994 Elsevier Science Ltd

0147-9571(93)E0005-Z Printed in Great Britain. All rights reserved 0147-9571/94 $6.00 + 0.00

L Y M P H O C Y T E C O L O N Y F O R M A T I O N BY A L E U K E M I C SHEEP I N F E C T E D WITH BOVINE

L E U K E M I A VIRUS

REGINALD JOHNSON*, JOHN B. KANEENE t a n d JAMES W. LLOYD

Department of Clinical Sciences, College of Veterinary Medicine, E. Lansing, MI 48824-1316, U.S.A.

(Rece&ed for publication 25 October 1993)

Abstract--The objective of this study was to investigate the effects of bovine leukemia virus (BLV) infection in sheep. A prospective study of the serologic, hematologic, and histologic changes of sheep infected with BLV was conducted. Antibodies to BLV were detectable in the sheep 3 weeks after exposure to blood from an infected cow and persisted during a 120 week examination period, whereas all control sheep remained seronegative. There were no statistically significant differences between the leucocyte counts, lymphocyte counts, and lymphocyte percentages of the infected and control sheep during the first 120 weeks of this study. However, one sheep did develop a leukopenia and lymphopenia 95 weeks after it became infected and died of histologically-confirmed lympho- sarcoma 10 days later. A lymphocyte colony assay was used to study the effects of BLV infection on colony formation by sheep lymphocytes in vitro. There was no significant difference in the number of lymphocyte colonies formed by BLV infected and control sheep. Nor was there a significant difference in the number of colonies formed by lymphocytes from the BLV infected sheep, when the autologous sheep serum was replaced with either pooled serum from the infected sheep or with pooled serum from the control sheep. BLV infection in aleukemic sheep does not appear to have an adverse affect on colony formation by lymphocytes in vitro.

Key words: Bovine leukemia virus, enzootic bovine leukosis, human T cell leukemia/lymphoma virus I, human T cell leukemia/lymphoma virus II, adult T cell leukemia, retroviruses, leukemia, lymphoma, lymphocyte colony assay, animal models of human disease, immunoassays.

R6sum6--L'objectif de ce travail est dYtudier les effets de l'infection par le virus de la leucemie bovine (LB) chez le mouton. Une 6tude prospective sur les modifications s6rologiques, h6ma- tologiques et histologiques sur le mouton infect6 avec le LB a 6t6 conduite. Les anticorps anti LB sont d&ectables chez le mouton, 3 semaines apr6s contamination avec du sang infect6 de vache, et ont persist6 durant les 120 semaines de l'exp6rimentation alors que les moutons t6moins sont rest6s seron6gatifs. I1 n'y a pas de diff6rence statistiquement significative entre le nombre de leucocytes, de lymphocytes et le pourcentage de lymphocytes des moutons infect6s et non infect6s pendant les 120 premi6res semaines de l'&ude. Cependant, un mouton a d6velopp6 une leucop6nie et une lymphop6nie, 95 semaines apr6s l'infection, et l'analyse histologique a confirm6 un lymphosarcome 10 jours apr6s. Le dosage des colonies de lymphocytes a 6t6 utilis6 pour 6tudier les effets de l'infection par LB sur la formatin de colonies de lymphocytes in vitro chez le mouton.

*Current address: Department of Veterinary Pathobiology, Purdue University, 1027 Lynn Hall, W. Lafayette, IN 47907-1027, U.S.A.

tAuthor for correspondence at: Population Medicine Center, A-109 Veterinary Clinical Center, Michigan State University, E. Lansing, MI 48824-1316, U.S.A.

Abbreviations used: agar gel immunodiffusion, AGID; adult T cell leukemia, ATL; autologous sheep serum, ASS; bovine leukemia virus, BLV; bovine syncytial virus, BSV; control pooled sheep serum, CPSS; enzootic bovine leukosis, EBL; human T cell leukemia/lymphoma virus I, HTLV-I; interleukin 2, IL-2; interleukin 2 receptor, IL2-R; infected pooled sheep serum, IPPS; protein 24, p24; protein 51, p51; pooled sheep serum, PSS; peripheral blood lymphocytes, PBL; persistent lymphocytosis, PL; Roswell Park Memorial Institute, RPMI.

REGINALD JOHNSON e t al.

I1 n'y a pas de diff6rence significative entre le nombre de colonies de lymphocytes form6es chez les moutons infect6s et les moutons t6moins. II n'y a pas de diff6rence significative dans le nombre de colonies de lymphocytes form6es, avec les moutons infect6s par LB, quand le serum autologue de mouton est remplac6 par un autre pool de serums de mouton infect6 ou avec un pool de serums venant des moutons t6moins. L'infection par LB chez les moutons leuc6miques appara~t ainsi ne pas avoir d'effet sur la formtaion de colonies de lymphocytes in vitro.

Mots-clefs: Virus de la leucose bovine, leucose bovine enzootique, leuc6mie humaine fi cellule T/lympone virus I, leuc6mie humaine ~ cellule T/lymphone virus II, leuc6mie ~ eellule T chez l'adulte, r6trovirus, leuc6mie, lymphome, dosage de colonies de lymphocytes~ modules animaux des maladies humaines, dosages immunologiques.

INTRODUCTION

Bovine leukemia virus (BLV) is an exogenous type C retrovirus that can cause abnormal proliferation of lymphoid cells in cattle [1]. The types of lymphoid proliferations include persistent lymphocytosis, which is nonpathogenic, and lymphoma and leukemia, both of which are highly fatal malignancies [2]. Human T cell leukemia virus I (HTLV-I), whose viral target cells is the T lymphocyte, is a pathogenic retrovirus of humans that is associated with adult T cell leukemia (ATL). Only B lymphocytes were thought to be infected by BLV [3]. However, the BLV provirus has been found in purified T lymphocyte populations using a highly specific BLV-DNA probe [4]. BLV and HTLV-I have strikingly similar biochemical and biological properties [5]. The 3' parts of their proviruses differ from other retroviruses in that a sequence called " p X " is located between the env gene and the 3' long terminal repeat. The pX region of the genome encodes a nuclear transcriptional activator of protein, tax, that activates the expression of genes directed by the viral long terminal repeat sequences [6].

The clinical similarities of the diseases caused by HTLV-1 and BLV, the presence of the pX region in their genomes, and the similar amino acid sequence homology between the two viruses suggest that BLV infection in animals may be a useful model for studying HTLV-! infection in humans [7]. The cow is the natural host for BLV, but experimental studies have shown that sheep also are susceptible to BLV. There are several advantages to using sheep rather than cattle as the host species in experimental models of BLV infection. Sheep, like cattle, develop persistent anti-BLV antibodies after exposure to BLV [8]. Some sheep infected with BLV also develop lymphocytosis, and lymphocytosis usually is an indication of impending leukemia and/or lymphoma [9]. The latent period for lymphosarcoma in infected sheep may be as short as 15 months. Most cattle do not develop lymphoproliferative disease prior to 48 months of age, if at all.

Expression of the HTLV-I tax gene is thought to stimulate the expression of the genes for interleukin 2 (IL-2) and the interleukin 2 receptor (IL-2R) [10, 11]. An autocrine mechanism of tumor cell proliferation by HTLV-I that involves IL-2 and IL-2R has been demonstrated in vi tro. The important role of IL-2 and IL-2R in the proliferation of lymphocyte colonies by normal cells has been demonstrated by the enhancement of clonal growth of lymphocytes by lymphocyte-derived IL-2 and recombinant IL-2 and by the inhibition o f T lymphocyte colony formation by a monoclonal antibody to the IL-2R. IL-2 concentrations in serum and plasma in humans usually exceed the in vi tro concentrations [12-14]. The infection of T lymphocytes by BLV, the role of the tax of HTLV-I in the expression of the genes for IL-2 and IL-2R by human lymphocytes, and the role of IL-2

Lymphocyte colony formation by aleukemic sheep 3

in colony formation by lymphocytes suggested to us that the lymphocyte colony assay may be a useful tool to study potential similarities in the mechanisms of cell proliferation induced by BLV and HTLV-I by using lymphocytes from BLV-infected sheep. We established a colony of BLV-infected sheep to study the effects of BLV infection on colony formation by sheep lymphocytes. Two hypotheses were tested: (1) BLV infection in sheep would result in increased colony formation by lymphocytes in vitro, and (2) serum from BLV-infected sheep would cause increased colony formation by lymphocytes from BLV-infected sheep, when compared to serum from noninfected sheep. The results of a study of this type could provide further evidence of the usefulness of BLV infection in sheep as an animal model for studying the mechanism of cell proliferation induced by human and animal leukemogenic retroviruses.

MATERIALS AND METHODS

Serologic tests for B L V and B S V antibodies

Serologic tests were used to identify the cows that would be appropriate donors of BLV and to identify the sheep that would be appropriate recipients of the virus. The agar gel immunodiffusion (AGID) test was used to examine the cow's sera for antibodies to bovine leukemia virus and to bovine syncytial virus (BSV) in order to identify appropriate blood donors that could be used to infect the sheep. The AGID test for antibodies to BSV was done by Dr Janice M. Miller of the National Animal Disease Center, Ames, Iowa. The AGID test for antibodies to BLV was done with a commercially available kit (Leukassay B, Pitmann-Moore, Washington Crossing, N.J.). The glycoprotein (gp51) and the core protein (p24) antigens of BLV are used in this diagnostic test. 10 ml of whole blood from each animal was collected into coagulant tubes (Becton-Dickinson) and was allowed to clot overnight at 26°C. The clots were removed, the remaining erythrocytes were pelleted by centrifugation at 400g for 10 min, and the serum was collected and stored at - 3 0 ° C until further use.

Selection of the donor cattle

A standard procedure was developed to identify bovine blood donors that could be used to infect the sheep with BLV without exposing them to BSV, another lymphocytotrophic retrovirus that infects cattle naturally. BLV seropositive and seronegative cattle, from which the two bovine blood donors were eventually chosen, were identified by screening adult cattle from the Michigan State University dairy herd for antibodies to BLV. The antibody prevalence to BLV in this herd of cattle had been between 70 and 95% during the previous 6 years. Several cattle whose initial AGID test results for BLV were negative were housed in isolation for 30 days, and the AGID test was repeated at the end of the isolation period to confirm that they had remained free of BLV antibodies. These cattle were then serotested to determine if they also were free of BSV antibodies. Several cattle whose initial AGID test for BLV was positive also were housed in isolation for a 30-day period. These cattle then were serotested for BSV antibodies. The BSV antibody-negative cattle whose initial AGID test for BLV was positive also were isolated for an additional 30 days, and the test for BSV antibodies was repeated at the end of the isolation period to confirm that they had remained free of BSV antibodies. All cattle that were BSV seropositive were eliminated from further consideration as blood donors.

4 REGINALD JOHNSON el al.

Selection of recipient sheep

14 clinically normal 8-month-old mixed-breed ewe lambs were obtained from the Michigan State University sheep farm. The ewes were serotested for antibodies to both BSV and BLV, housed in isolation from all other animals for 30 days, and serotested again at the end of the isolation period. The same A G I D tests were used for both the sheep and the cattle.

Infection of the sheep with B L V

The sheep were randomly assigned to a "control" group (N = 6) or to an "infected" group (N = 8). Each group of sheep was housed in open lots 0.5 km apart during the remainder of the study. The procedure used to infect the sheep is described in Fig. 1. A single donor cow that was seronegative for both BLV and BSV was selected to provide the blood that was used for the sheep assigned to the control group, whereas a single, BSV seronegative, BLV seropositive donor cow was selected to provide the blood that was used for the sheep that were assigned to the infected group. Blood samples were collected from the donor cows as well as all sheep immediately prior to the injection of blood from the donor cows to confirm that the serologic status of each animal had not changed from the previous findings. Blood from the two donor cows was collected into evacuated bottles containing acid citrate dextrose. RPMI 1640 tissue culture medium at 38°C was added to

BLV Neg. BLV Pos. BLV Neg. BSV Neg. BSV Neg. BSV Neg.

: PL Pos. :: . . . . . . . . . . . J

l Serum

Negative

w::;v, eTo:,;;':'n Anticoag. blood

I. WBC Count 2. Diff. Count

Randomization

Resuspend blood in ~ / " ~ Co~ntr ° RPMI 1640~ 57*C I Infected I L _J

Fig. 1. The procedure used to infect sheep with BLV.

Lymphocyte colony formation by aleukemic sheep 5

5.0 ml of blood at a ratio of 2:1, mixed by inversion, and equal volumes were injected intramuscularly into each sheep at 3 different sites. The sheep that were assigned to the control group were injected with 5.0 ml of diluted blood from the donor cow that was seronegative for both BSV and BLV. The sheep that were assigned to the infected group were injected with 5.0 ml of blood from the donor that was seropositive for BLV and seronegative for BSV.

Prospective evaluations of the sheep

AGID tests, leucocyte counts and differentials, and physical examination of the sheep were done on the day of infection and at 4-week intervals during a 27-month period. Leukocytes were counted manually (Unopette, Becton-Dickinson, Rutherford, N.J.). Blood smears were stained with hematoxylin/eosin using an automated slide stainer. Differential counts were done by counting at least 100-200 leukocytes using an oil immersion objective.

Persistence of BL V infection in sheep

A biological assay was used to demonstrate the persistence of BLV infection in the sheep that seroconverted and had persisting anti-BLV antibodies. 5 ml of blood was collected from three of the eight sheep that had been seropositive for 32 months after they were originally infected with blood from the infected donor cow. The blood from each sheep was then injected into one of three additional BLV seronegative sheep by using the same protocol that had been used to expose the first eight sheep. Serum samples were collected from these three sheep beginning 2 weeks after exposure, and the AGID test also was used to detect antibodies to BLV in the serum from these sheep.

Examination of the tumors

The sheep that died was necropsied and examined for tumors. The tissues were fixed in 10% formalin, embedded in paraffin, sectioned, stained with hematoxylin and eosin, and examined via light microscopy.

Lymphocyte colony assay

The details of the procedure that was used to standardize the assay for sheep lymphocytes has been described [15]. Thus, only the minor changes in the assay will be presented. Autologous sheep serum (ASS) was used in the experiments to test the effect of BLV infection on colony formation by lymphocytes from infected or control sheep. Pooled sheep serum (PSS) was used in experiments to test the effect of the source of serum on colony formation by lymphocytes from the infected sheep. To prepare the PSS, equal volumes of serum from sheep in the infected group was mixed to form one pool of serum (IPPS). Likewise, serum from sheep in the control group was mixed to form another pool (CPSS). All sera were filtered through a 0.20 micron syringe filter (Millpore Prod., Bedford, Mass.), heat-inactivated at 56°C for 30 rain and frozen immediately in 1.5 ml aliquots at -30°C. The ASS sera had been stored between 6 and 18 days prior to being used in the colony assay, and the PSS had been stored between 5 and 18 days. The preparation of concanavalin A, the lymphocyte mitogen, and collection of the peripheral blood mono- nuclear cells also has been described.

6 REGINALD JOHNSON et al.

Statistical analysis of the data

The general linear models analysis of variance (ANOVA) for an unbalanced design was used to evaluate differences between the leucocyte counts, lymphocyte counts, and lymphocyte percentages of infected and control sheep. The samples were collected from the sheep during the 18 successive examination periods at 4-week intervals. An unbalanced design was used because one ewe in the infected group was killed by a predator 60 weeks after infection with BLV, and one ewe in this same group died of lymphosarcoma 96 weeks after infection with BLV.

A nested ANOVA model was used to test the hypothesis that the number of colonies formed by lymphocytes from BLV infected sheep would be different from the number formed by lymphocytes from the control sheep [16]. An average of 10 measurements per sheep per group was made. A crossed ANOVA model was used to test the hypothesis that the number of colonies formed by lymphocytes of the infected sheep, whose cells were cultured in IPSS, would be different from the number of colonies formed by lymphocytes from these same sheep, if CPSS was used as the source of serum instead of IPSS. An average of 5 measurements was made of the cells of each sheep treated with each type of serum, The data were analyzed using the SAS computer program [17].

RESULTS

BSV and BLV antibodies before exposure

12 cows that were known to be seropositive for BLV based upon the results of serial A G I D tests of cattle in this herd were retested at the start of this study, and all were still seropositive. Only one of the 12 BLV seropositive cows was BSV seronegative. This cow served as the blood donor for the eight sheep that were assigned to the group of sheep that were infected with BLV. All three of the BLV seronegative cows that were tested for BSV antibodies were seronegative. None of the sheep had antibodies either to BLV prior to exposure to blood from the donor cow. None of the sheep developed antibodies to BSV.

Lymphocyte counts of the donor cows

The lymphocyte count of the BLV infected donor was 7950//~1, whereas that of the BLV free donor was 6500//tl. Thus, each sheep in the infected group received 19.9 x 106 lymphocytes, and each sheep in the control group received 16.25 z 106 lymphocytes. Blood from a BLV donor with persistent lymphocytosis was not used because none of the BLV seropositive cows had persistent lymphocytosis.

Prevalence of BL V antibodies in sheep after exposure

All eight sheep that were exposed to blood from the BLV seropositive donor had developed antibodies to BLV within 3 weeks of exposure. The A G I D test results were always strongly positive during the entire 27-month observation period. None of the six sheep in the control group developed antibodies to BLV during the study.

Blood cell counts of the sheep

There were no statistically significant differences between the leucocyte counts ( P = 0 . 1 4 3 ) , lymphocyte counts ( P = 0 . 7 0 ) (Fig. 2), and lymphocytes percentages (P = 0.74), of the infected sheep and the control sheep during the 18 different examination periods in which the values were determined. Both the lymphocyte count (519/~tl) (Fig. 3)

Lymphocyte colony formation by aleukemic sheep

Number of Lymphs/~l 8000

7000

6000

5000

4000

3000

2000

1 0 0 0 ~ _ ~ _ i t I t ~ J I t ~ ~ ~

-6 0 6 12 18 24 30 36 42 46 54 60 66 72 78 64 90 96 102108

T i m e (weeks)

[] Control (n=6) ~ Infected (n-8)

Fig. 2. Comparison of peripheral blood lymphocyte counts between sheep infected with BLV and control sheep. * = exposure to BLV.

and the lymphocyte percentage (17%) had decreased dramatically in one ewe 10 days prior to her death due to lymphosarcoma. There was no cytologic evidence of malignant cells in the peripheral blood based upon the microscopic examination of peripheral blood smears examined 10 days preceding the death of this ewe.

Biological assay for persistence of BL V infection

All three BLV seronegative sheep that were exposed to blood from the three sheep in the infected group that had been seropositive for 32 months also seroconverted. Two of the three sheep had seroconverted by 4 weeks after exposure, and the third sheep had seroconverted by 6 weeks after exposure. This provided evidence that the presence of anti-BLV antibodies in sheep in the infected group was indicative of a persistent infection.

Tumors in the sheep

Only one of the ewes in the infected group developed tumors. This ewe died suddenly 96 weeks after exposure to the virus. Tumors between 3.0 and 70.0 mm in diameter were located in most of the abdominal organs including the liver (Fig. 4), kidney, ovary, stomach, diaphragm, and mesenteric lymph nodes. There was no external evidence of tumors in this ewe. Histologically, the highly cellular and compact tumors found in the kidney were comprised of multiple and sometimes single large nodules that compressed the surrounding normal tissue (Fig. 5). The tumor cells were round to ovoid with distinct cytoplasmic borders, and hyperplastic nuclei within the cells contained two to three large nucleoli. There were as many as 10 mitotic figures per 40 × field. The normal lymphoid follicular architecture of the lymph nodes and spleen was completely destroyed due to the proliferation of large lymphoblastic cells. Approximately one-half of the normal par- enchyma in the liver, kidney, adrenal glands, and uterus was destroyed.

8 REGINALD JOHNSON et aL

Number of Lymphs/lxl 8000 I- * Exposure

L / ~ ,~ I" No da ta

6ooo

4000 t

2000 P

r -6 0 6 t2 16 24 30 36 42 46 ~ . 60 66 72 78 64 90 96102t08

T i m e ( w e e k s )

o C o n t r o l (n -6 ) ~ L e u k o t i c (n - l )

Fig. 3. Comparison of peripheral blood lymphocyte counts between one sheep with lymphosarcoma and control sheep. * = exposure to BLV.

Lymphocyte colonies

The colonies became visible by day 3 of culture, and the maximum number of colonies was attained around day 5 or day 6. The colonies contained several hundred to several thousand cells (Fig. 5). There was no statistically significant difference (P = 0.51) between the number of colonies formed by lymphocytes from the infected sheep and the control sheep (Fig. 6). The lymphocytes from one sheep in the infected group failed to form any colonies, although the cells were cultured several times. Nonmitogen-stimulated lympho- cytes also failed to produce lymphocyte colonies. There was no statistically significant difference (P = 0.052) between the number of colonies formed by lymphocytes from the BLV infected sheep when these lymphocytes were cultured in IPPS or CPSS (Fig. 7).

N u m b e r of Co lon ies 1200

1000

800

600

400

200

0 B L V I n f e c t e d

N = 6 Infected N = 6 Control p = 0 . 0 5 1

C o n t r o l

Fig. 7. Comparison of colony formation by lymphocytes from sheep infected with BLV and control sheep. N = 6; P = 0.51.

Fig. 4. Infiltration of tumors into the liver of a sheep infected with BLV that developed lymphosarcoma.

Fig. 5. Neoplastic lymphocytes disrupting the normal parenchyma of the kidney of a sheep infected with BLV that developed lymphosarcoma (40 x ).

Fig. 6. Lymphocyte colonies in a 5-day-old soft-agar culture of normal sheep lymphocytes stimulated with 33 pg/ml of CON A. 200 x. Each colony consisted of several hundred to several

thousand cells.

l0

Lymphocyte colony formation by aleukemic sheep 11

Number of Colonies 1400

1200 I

i

1000

80O

600

400

20O

0 Control PSS

P S S = Poo led sheep s e r u m N = 6 p = 0 .052

Infected PSS

Fig. 8. Comparison of colony formation by lymphocytes that were cultured either with IPSS or with CPSS. The lymphocytes were taken from sheep infected with BLV. N = 6; P = 0.62.

DISCUSSION

The susceptibility of sheep to infection by BLV has been documented. The blood used to infect the sheep was always obtained from a bovine donor, a donor that may have been infected with BLV as well as BSV. BSV is an exogenous lymphocytotrophic retrovirus of cows that induces formation of syncytia when cocultured with bovine embryonic spleen, embryonic lung, and ovine embryonic spleen cells [18, 19]. Antigens of BSV are also expressed in cultured bovine lymphocytes. We were concerned that the presence of BSV antigens in lymphocytes may cause false positive tests for BLV in some in vitro assays such as the syncytium induction assay and other cellular immunoassays, both of which have frequently been used to detect the infection of lymphocytes by BLV. Thus, the procedure that we used to infect the sheep in our study with BLV was designed to prevent coinfection of our sheep with BSV. We tested 12 BLV seropositive cows for antibodies to BSV, and only one was seronegative. Thus, contamination of bovine anti-BLV antisera with BSV antibodies should be considered when immunoassays are being used to detect BLV antigens in lymphocytes. Since BSV does not cause natural infections in sheep, we concluded that sheep that have been infected with BLV would be more suitable donors of the virus for long-term studies than cattle would be.

We initially sought a BLV donor with persistent lymphocytosis because in vitro studies have suggested that cows with persistent lymphocytosis could be more infectious than other cows [20]. The in vivo infectivity of cows with and without PL has been confirmed by showing that only 926 lymphocytes from a donor cow with PL and a high titer to gp51 were infectious to sheep, but 325,000 lymphocytes from a cow without PL and a low titer to gp51 were required to infect sheep [21]. We succeeded at infecting all of our sheep by using blood from a donor without PL, but a relatively large number of lymphocytes were used. Thus, a BLV donor with PL is not a prerequisite for the successful infection of sheep.

The appearance of BLV gp51 antibodies in the serum of our sheep as early as 3 weeks after exposure is consistent with others' findings [22-24]. We were concerned that the infection in our sheep may not have been persistent, because our blood donor did not have

12 REGINALD JOHNSON et al.

PL. Thus, a biological assay was conducted to determine if the persistence of the antibodies in the sheep also was an indication of a persistent BLV infection. All three sheep that were exposed to blood from three sheep in our seropositive flock did seroconvert, providing evidence that the infection was persistent during and beyond the period in which the lymphocyte colony assays were completed.

The peripheral blood lymphocyte numbers in sheep infected with BLV may vary. Lymphocytosis, leukemia, lymphopoenia, and normal lymphocyte counts have been reported in other studies, the duration of which was 30-46 months [8, 22, 25, 26]. Increased peripheral blood lymphocyte counts have been detected by others 2-24 weeks before the death of sheep that developed tumors. The only ewe in our study that had developed tumors at the time that these results were reported was slightly lymphopenic 10 days before her death. Additional changes in the other leucocyte parameters indicated that the hematologic response of that ewe was consistent with suppression of the output of cells by the bone marrow. The pathologic effects of BLV infection on the bone marrow of sheep have not been reported.

HTLV-I infected cells and HTLV-I virions have been used to induce the proliferation of T colony-forming cells in the absence of exogenous IL-2 and accessory cells [27]. HTLV-I particles were treated with HTLV-I antisera against the envelope and core proteins of the virus in order to inhibit binding of the virus to lymphocytes. Since colony formation was subsequently decreased, it was concluded that virus particles were directly involved in colony formation, and that this involvement may have an accessory cell function that triggers the autocrine secretion of IL-2. Direct activation of resting T lymphocytes, induction of IL-2 secretion and IL-2R expression by HTLV-I was later shown by the same investigators [27]. The formation of colonies was directly attributed to these events. We were not able to demonstrate differences in lymphocyte colony formation in the experiments that we conducted. The lymphocyte colony assays were done after our ewes had been infected between 102 and 108 weeks. None of the ewes whose lymphocytes were used in the colony assay had developed lymphocytosis or leukemia at this time. One ewe had developed lymphosarcoma and lymphopoenia. Lymphocytes from this sheep had not been cultured at the time of its death, because the procedures for our lymphocyte colony assay had not been standardized. Our inability to demonstrate abnormalities in lymphocyte colony proliferation may be explained by the absence of lymphocytosis, leukemia and/or lymphocytosis in our sheep. Experiments using lympho- cytes from sheep with lymphocytosis, leukemia or lymphoma should be carried out to further evaluate the hypotheses proposed as a means of evaluating potential similarities in the pathogenic effects of BLV and HTLV-I .

Acknowledgements--Funds were provided by a National Institutes of Health Training Grant No. GM-07995-07 for R. Johnson and by a Biomedical Research Support Grant from Michigan State University. The authors thank Dr J. M. Miller and the dissertation committee for advice, and S. Stark, Y. Bryant, M. Hoffmann, M. E. Shea and K. Sayles-Winsky for technical assistance.

REFERENCES

1. Miller J. M., Miller L. D., Olsen C. and Gilette K. G. Virus-like particles in phytohemagglutinin-stimulated lymphocyte cultures with reference to bovine lymphosarcoma. J, Natn. Cancer Inst. 43, 1297-1305 (1969).

2. Ferrer J. F. Bovine lymphosarcoma. Adv. Vet. Sei. Comp. Med. 24, 1~68 (1980). 3. Paul P. S., Pomeroy K. A., Johnson D. W., Muscoplat C. C., Handwerger B. S., Soper F. F. and Sorensen

Lymphocyte colony formation by aleukemic sheep 13

D. K. Evidence for replication of bovine leukemia virus in the B. lymphocyte. Am. J. Vet. Res. 38, 873-876 (1977).

4. Williams D. L., Barta O. T. and Amborski G. R. Molecular studies of T lymphocytes from cattle infected with bovine leukemia virus. Vet. Immunol. Immunopathol. 19, 307-323 (1988).

5. Sagata N., Yasanuga T. and Ohishi K. Comparison of the entire genomes of bovine leukemia virus and human T-cell leukemia virus and characterization of their open reading frames. EMBO J. 3, 3231 3237 (1984).

6. Felber B. K., Paskalis H., Kleiman-Ewing C., Wong-Staal F. and Pavlakis G. N. The pX protein of HTLV I is a transcriptional activator of its long terminal repeats. Science 229, 675~78 (1985).

7. Burny A., Bruck, C., Couez D., Deschamps J., Ghysdael J., Kettmann R., Mammerickx M., Marbaix G. and Portetelle D. Enzootic bovine leukemia: its relevance as a model system for human T cell leukemia. In Human T Cell Leukemia/Lymphoma Virus (Edited by Gallo R. C., Essex M. and Gross L.), pp. 17 24. Cold Spring Harbor, New York (1984).

8. Rogers R. J., Chung Y. S. and Dimmock C. K. Development of lymphosarcoma in sheep after inoculation with bovine leucosis virus. Aust. Vet. J. 61, 196-198 (1984).

9. Djilali S., Parodi A. L., Levy D. and Cockerell G. L. Development of leukemia and lymphosarcoma induced by bovine leukemia virus in sheep: a hematopathological study. Leukemia 1, 777-781 (1987).

10. Greene W. C., Leonard W. J., Wano Y., Svetlik P. B., Peffer N. J., Sodroski J. G., Rosen C. A., Goh W. C. and Haseltine W. A. Trans-activator gene of HTLV II induces IL 2-receptor and IL 2 cellular gene expression. Science 232, 877-880 (1986).

11. Greene W. C., Leonard W. J., Wano Y., Svetlik P. B., Peffer N. J., Sodroski J. G., Rosen C. A., Goh W. C. and Haseltine W. A. Trans-activator gene of HTLV II: interpretation. Science 235, 1073 (1987).

12. Winkelstein A., Simon P. L., Wood D., Machen L. L., Shadduck R. K. and Waheed A. The effect of interleukins and other soluble factor on T lymphocyte colony formation. Immunology 58, 173-177 (1986).

13. Jourdan M., Commes T. and Klein B. Control of human T colony formation by interleukin 2. Immunology 54, 249 253 (1985).

14. Cornaby A., Simpson M. A., Vann Rice R., Dempsey R. A., Madras P. N. and Monaco A. P. Interleukin 2 production in plasma and urine, plasma interleukin 2 receptor levels, and urine cytology as a means of monitoring renal allograft recipients. Transplant Proc. 20 (1 Suppl.), 108-110 (1988).

15. Johnson R,, Kaneene J. B. and Lloyd J. W. A lymphocyte colony assay for culture of sheep lymphocytes. Comp. Immunol. Microbiol. Infect. Dis. 17, 15-27 (1994).

16. Neter J., Wasserman W. and Kutner M. H. Applied Linear Statistical Models, 2rid edn, pp. 963-984. Irwin, Homewood (1985).

17. SAS User's Guide. Basics. Version 5 Edition. SAS Institute Inc,, Cary, N.C. 18. Bouillant A. M. P. and Ruckerbauer G. M. Isolation of bovine syncytial virus from lymphocytes recovered

from fluids used to flush uterus and oviducts of superovulated cattle. Can. J. Comp. Med. 48, 332-334 (1984). 19. Marshall R. G. and Gillette K. G. Isolation of bovine syncytial virus from mesenteric lymph nodes of cattle.

Am. J. Vet. Res. 44, 1366-1369 (1983). 20. Itohara S., Terui S. and Mizuno Y. Infectivities of bovine leukemia virus in peripheral blood lymphocytes

from naturally infected cattle and their relation to persistent lymphocytosis and antibody titers. Jpn J. Vet. Sci. 47, 807-810 (1985).

21. Mammerickx M., Portelle D., De Clercq K. and Burny A. Experimental transmission of enzootic bovine leukosis to cattle, sheep and goats: infectious doses of blood and incubation period of the disease. Leuk. Res. 11, 353-358 (1987).

22. Cockerell G. L., Parodi A-L. and Levy D. Immunocompetence of sheep experimentally infected with bovine leukemia virus. Vet. lmmunol. Immunopathol. 13, 189~02 (1986).

23. Walker A. F., Lumsden J. H. and Stirtzinger T. Neutrophil function in sheep experimentally infected with bovine leukemia virus. Vet. Immunol. Immunopathol. 14, 67 76 (1987).

24. Stirtzinger T., Valli V. E. O. and Miller J. M. The role of virus dose in experimental bovine leukemia virus infection in sheep. Can. J. Vet. Res. 52, 222-228 (1988).

25. Dimmock C. K., Rogers R. J., Chung Y. S., McKenzie A. R. and Waugh P. D. Differences in the lymphoproliferative response of cattle and sheep to bovine leucosis virus infection. Vet. Immunol. lmmuno- pathol, l l , 325-331 (1986).

26. Olson C., Miller L. D., Miller J. M., Hoss H. E. Brief communication: transmission of lymphosarcoma from cattle to sheep. J. Natn. Cancer Inst. 49, 1463-1467 (1972).

27. Duc Dodon M. and Gazzolo L. Loss of interleukin 2 requirement for the generation of T colonies defines an early event of human T lymphotrophic virus type I infection. Blood 69, 12-17 (1987).