Adhesion molecules and lymphocyte subsets in milk and blood of periparturient Holstein cows

Veterinary Immunology and Immunopathology 69 (1999) 23±32

Adhesion molecules and lymphocyte subsets in milk and blood of periparturient Holstein cows Corinne Van Kampen*, Bonnie A. Mallard, B.N. Wilkie Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, Ont., Canada N1G 2W1 Received 10 September 1998; received in revised form 8 March 1999; accepted 16 March 1999

Abstract Migration of leukocytes into the mammary gland is an essential element of resistance to infection which is likely influenced by expression of adhesion molecules. The contribution of subsets to mammary gland resistance remains unclear. Mononuclear cells from milk and blood of dairy cows were examined for variation in CD4‡, CD8‡, and WC1‡ (Workshop Cluster 1; marker for gd T cells) lymphocyte phenotypes and expression of LFA-1 and L-selectin at several time points during the periparturient period and at Week 16 of lactation. Proportions of CD4‡ T cells were higher (p  0.05) in blood than milk at all times between Week 0 and Week 16 relative to calving; the inverse was true of CD8‡ cells. Expression of L-selectin was lower (p  0.05) on CD4‡ cells and higher on CD8‡ cells from milk. The WC1‡ subset was more frequent in blood than in milk except at calving when the opposite was true. After calving, proportions of L-selectin‡WC1‡ cells decreased steadily to Week 16. Expression of LFA-1 was examined on mononuclear cell populations and found to be lower on milk cells and did not vary over time. We conclude that proportions of T cells subsets differ significantly between blood and milk, particularly around calving. Corresponding variations in L-selectin expression may indicate a role for this molecule in regulating the movement of CD8‡ and WC1‡ T cells into the bovine mammary gland. # 1999 Elsevier Science B.V. All rights reserved. Keywords: Bovine; Mammary gland; Adhesion; Subsets; Lymphocyte

1. Introduction Mammary gland defence is influenced by the composition and functional ability of resident and incoming cells, some of which have been shown to be functionally and numerically suboptimal during the peripartum period (Yang et al., 1988; Kehrli et al., 1989a, * Corresponding author. Tel.: +1-519-824-4120; fax: +1-519-767-0809; e-mail: [email protected] 0165-2427/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 9 9 ) 0 0 0 3 4 - 3

24

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

b). Migration of these cells may be influenced by bovine adhesion molecule expression such as L-selectin and LFA-1. Furthermore, colostrum may be an important source of lymphocyte effector cells for the neonatal calf (Reidel-Caspari and Schmidt, 1991a, b). Taylor et al. (1994) found that CD8‡ T cells are predominant in milk as compared to blood (Taylor et al., 1994), and gamma/delta T cells have been implicated in mucosal defence (Sordillo et al., 1997) and are also present in the bovine mammary gland (Shafer-Weaver et al., 1996). Circulating leukocyte populations are a primary source of cells that transmigrate into the gland of healthy and infected animals. Adhesion molecules are likely to be one group of molecules that regulate the entry of leukocytes into the mammary gland. There are numerous examples of the importance of adhesion molecules in host defence (Bevilacqua, 1993; Gilbert et al., 1993). Previous studies have described some of the phenotypic variation in T cell subset markers, seen in blood and/or milk during the peripartum period; however, few studies address the issue of how the regulation of adhesion molecules may affect cell trafficking of lymphocyte subsets during this period (Shafer-Weaver et al., 1996). L-selectin and LFA-1 (leukocyte functional antigen -1) are two of the most crucial molecules that regulate the `tether' and `trigger' phases of the multistep adhesion cascade (reviewed by Springer 1994). L-selectin is constitutively expressed on all naive T lymphocytes and a subset of memory T cells (Picker et al., 1990; Mackay et al., 1992), as well as other leukocytes. LFA1 is also constitutively expressed on most leukocytes and serves to mediate the triggering phase of adhesion by undergoing upregulation and conformational changes in the cell membrane (Figdor et al., 1990). This study tests the hypothesis that certain subsets of T cells express L-selectin and/or LFA-1 differently between blood and milk compartments. If true, this may begin to explain at least one of the reasons for the differences in subset proportions between the two compartments. 2. Materials and methods 2.1. Animals and sample collection 40 ml of peripheral blood was collected from the tail vein into EDTA from each of ten periparturient Holstein dairy cows at weeks ÿ 3, 0, 3, 6, and 16 relative to expected calving dates. Milk samples were collected at 48 h, and 3, 6, and 16 weeks postpartum. Mononuclear cells were collected from blood by Ficoll-Hypaque density gradient centrifugation (Sigma, St. Louis, MO) followed by hypotonic lysis of red blood cells and several washes with PBS as described previously (Van Kampen and Mallard, 1997). Milk was centrifuged (20 min, 400  g) followed by removal of fat and whey, leaving a loose cell pellet that was resuspended and washed several times with PBS (pH 7.2), followed by Ficoll-Hypaque enrichment. Viability of milk cells was 60±70% and peripheral blood cells, >95%, based upon trypan blue exclusion. Cell concentrations were adjusted to 20  106 cells mlÿ1 with PBS (0.1% azide). 2.2. Immunostaining for flow cytometry Monoclonal antibodies specific for several lymphocyte surface markers or adhesion markers were kindly provided by Dr. J. Naessens of ILRI (International Livestock

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

25

Table 1 Primary antibodies used for immunstaining of bovine blood and milk mononuclear cells Antibody

Isotype

Specificitya

Cell typesb

Supplier

IL-A11 IL-A105 IL-A29 MM1A DREG56 IL-A99

IgG2a IgG2a IgG1 IgG1 IgG1 IgG1

BoCD4 BoCD8 BoWC1 BoCD3 L-selectin BoCD11a

Th, some M Tc, some NK gd T pan T T, N, M T, B, N, M

ILRI ILRI ILRI VMRDd Jutilae ILRI

a

As defined by Vet. Immunol. Immuopathol. 39(1±3): 27±48 (1993) for bovine (Bo) surface markers. T ˆ T lymphocyte; B ˆ B lymphocyte; NK ˆ natural killer cells; M ˆ monocytes; N ˆ neutrophils. ILRI ˆ International Livestock Research Institute, Kabete, Kenya. d VMRD ˆ Veterinary Medical Research and Development, Pullman, WA. e Jutila ˆ Mark Jutila, Montana State University, Bozeman, MT. b c

Research Institute, Kabete, Kenya), Dr. M. Jutila (Montana State University, Bozeman, MT) and Dr. C. Howard of IAH (Institutes of Health, Compton, England), or purchased from VMRD (Pullman, Washington)(Table 1). All primary antibodies were used at 1 : 1000 except CD3 (MM1A) which was used at a 1 : 50 dilution. Secondary antibodies were fluorescein isothiocyanate (FITC)-labelled goat anti-mouse IgG1 (H ‡ L) (Cedarlane Laboratories, Hornby, Ontario), and biotinylated goat anti-mouse IgG2a (Cedarlane) diluted 1 : 200 in PBS (0.1% azide). Streptavidin-phycoerythrin (Cedarlane) was used at 5 ml per 106 cells, as a third step with the biotinylated antibody. Mouse IgG1 and IgG2a (Mandel Scientific, Guelph, Ontario) were used as isotype controls, diluted 1 : 500 in PBS. Dual color immunostaining was performed to ascertain proportions of CD4‡, CD8‡ and WC1‡ T (CD3‡) cells, and L-selectin expression on these subsets. Single color immunostaining quantified proportions of LFA-1‡ cells. Cells were fixed in 1% paraformaldehyde and refrigerated until examined by flow cytometry. 2.3. Flow cytometric analyses A flow cytometer (Becton Dickinson) was used to acquire all subset data and analyzed using LYSYS II software (Becton Dickinson). Ten thousand cells were analyzed from blood samples and five thousand from milk samples. Lymphocytes were gated out from other populations based upon their forward and side scatter characteristics. Dot plot diagrams were generated for CD3‡CD4‡, CD3‡CD8‡, CD3‡WC1‡, L-selectin‡CD4‡, ‡ ‡ ‡ ‡ L-selectin CD8 , L-selectin WC1 and LFA-1‡ cells in blood and milk samples. Regions of background fluorescence were established with isotype controls. Data was recorded and analyzed as percent positive cells at each time point. Two-dimensional scatter plots (FSC vs. SSC) were generated for each sample and the lymphocyte gate was designated as R1 based upon scatter characteristics. Dot plots of FL1 versus FL2 with appropriate quadrant markers based upon isotype control dot plots, were then used to determine the proportions of CD3‡/CD4‡ or CD8‡ or WC1‡ T cell subsets and L-selectin‡CD4‡ or CD8‡ or WC1‡ subsets. Single color histograms were used to quantify the proportions of LFA-1‡ mononuclear cells.

26

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

2.4. Statistical analysis Normality of data distribution was confirmed using the Univariate procedure of the Statistical Analysis System (SAS) (Helwig and Council, 1984). A general linear model (GLM) procedure was used to generate least-squares analysis of variance (ANOVA) of blood data and milk data. The GLM model included fixed effects of week (relative to calving), and sample (milk vs. blood), and a week by sample interaction term, to determine any significant differences between milk and blood subsets at each time point. Differences in lymphocyte subset proportions between weeks and sample type by week were reported using Least square means (Ls means) derived from the ANOVA. Significance in variation between lymphocyte subsets and milk and blood was accepted at p  0.05. 3. Results 3.1. Variation in T cell subsets Proportions of CD4‡ T cells were significantly higher in blood than milk from Week 0 to Week 16 of sampling (Fig. 1A). T cells of the CD8‡ phenotype were significantly lower in blood during the same period (Fig. 1B), and WC1‡ cells were significantly higher in blood than milk except at calving during which time the inverse was true (Fig. 1C). Comparisons over time revealed a significant decline in blood CD8‡ and WC1‡ subsets at calving but an increase in blood CD4‡ T cells at calving. Variations of these subsets in milk were not as pronounced; however, CD4‡ proportions were lowest and CD8‡ and WC1‡ were highest in milk at calving. 3.2. Variation in LFA-1‡ mononuclear cells Proportions of LFA-1 ‡ mononuclear cells were higher (p  0.05) in blood than milk, but there was no significant variation in LFA-1 expression over time (Table 2). 3.3. Variation of L-selectin expression on T cell subsets Blood and milk mononuclear cells were L-selectin‡ and L-selectinÿ, for each T cell subset, in two experiments in which the proportions of cells in each subpopulation remained the same. The sum of L-selectin‡ (Fig. 2A, B, C) and L-selectinÿ (Table 3) proportions of each subset at each time point was approximately equal to the total proportions of each subset measured in milk and blood (Fig. 1A,B,C). Variation patterns in T cell subset proportions during the peripartum period (Fig. 1) were similar to the pattern of L-selectin expression on T cell subsets during the same period (Fig. 2) Proportions of L-selectin‡CD4‡ cells in milk were consistently and significantly lower than the corresponding subset in blood at all time points (Fig. 2A). Proportions of this subpopulation in milk were lowest at calving; whereas, L-selectin‡CD4‡ cells in blood were lowest at Week 6 (peak lactation). The peak in CD3‡CD4‡ proportions in blood and milk corresponds (Fig. 1A) with the highest proportions of L-selectin‡CD4‡ subsets.

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32 27

Fig. 1. Least square means of bovine (A) CD4‡; (B) CD8‡, and (C) WC1‡ T cell subset proportions in blood (solid bars) and milk (hatched bars) during the peripartum period (Week 0 represents calving) and later lactation. Significant variation in cell proportions over time are represented by different lower case letters for blood and upper case letters for milk. Differences between milk and blood values are significant (p  0.05) at all time points.

28

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

Table 2 Least square means of percent LFA-1‡ mononuclear cells in blood and milk of peripartum Holstein cows Time (weeks)a

% LFA-1‡ mononuclear cells (SD)b

ÿ3 0 3 6 16 a b

Blood

Milk

60.5  2.7 59.11  2.6 59.74  2.6 60.0  2.7 63.6  2.7

ND 33.22  1.81 36.99  1.94 37.34  1.94 35.38  1.94

Week relative to expected calving date. Significant differences (p  0.05) at each time point between blood and milk.

CD4‡ cells were detected at lower proportions than the L-selectin‡ subset at all time points in milk and all but Week 6 in blood (Table 3). A significantly higher proportion of L-selectin‡CD8‡ T cells were detected in milk than in blood (Fig. 2B). The pattern of variation over time coincided with the CD3‡CD8‡ changes noted earlier (Fig. 1B). The L-selectin‡CD8‡ subset in blood varied over time with significantly lower proportions from calving to Week 6. The L-selectinÿCD8‡ proportions in blood were approximately equivalent to the L-selectin‡CD8‡ population at all time points; whereas, the proportion of CD8‡ cells expressing L-selectin in milk was 2±3 times higher than L-selectinÿCD8‡ proportions at all time points. ‡ ‡ L-selectin WC1 cells in blood only changed significantly between Week 6 and Week 16 (Fig. 2C). In milk, the L-selectin‡WC1‡ subset steadily declined from its highest proportion at calving to Week 16. Patterns of variation in this WC1‡ subset over time in blood were similar to the CD3‡WC1‡ pattern. The variation in milk L-selectin‡WC1‡ subsets did not coincide with the CD3‡WC1‡ pattern (Fig. 1C). Proportions of LselectinÿWC1‡ T cells were lowest at calving in blood and milk and lower than proportions of L-selectin‡WC1‡ cells except at Week 16. L-selectin

ÿ

4. Discussion Comparisons of lymphocyte subsets in milk with proportions found in peripheral blood, in conjunction with information about adhesion molecule expression on these Table 3 Least square means of percent L-selectin negative T cell subsets in blood and milk during the peripartum period of dairy cows Time (weeks)

ÿ3 0 3 6 16

Ls means of L-selectinÿ cells in blood

Ls means of L-selectinÿ cells in milk

CD4‡

CD8‡

WC1‡a

CD4‡

CD8‡

WC1‡

11.9  2.2 16.8  2.1 17.8  2.1 19.3  2.2 11.9  2.2

7.8  1.03 3.63  0.96 6.81  0.96 8.78  1.02 6.1  1.02

12.7  1.6 3.94  1.47 8.15  1.37 10.2  1.46 7.16  1.46

ND 4.1  0.91 6.04  0.97 6.28  0.97 7.1  0.97

ND 8.57  1.29 7.02  1.38 9.31  1.38 11.8  1.39

ND 3.1  0.77 4.49  0.83 8.7  0.83 9.24  0.83

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32 29

Fig. 2. Least square means of blood (solid bars) and milk (hatched bars) mononuclear cells expressing L-selectin and (A) CD4; (B) CD8, or (C) WC1 during the peripartum period (Week 0 represents calving) and later lactation. Significant variation in proportions over time are represented with different lower case letters in blood and upper case letters in milk. Significant differences (p  0.05) between milk and blood are indicated with an *.

30

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

cells, may provide clues about the importance of and trafficking of particular subsets to the mammary gland. Furthermore, variations in expression of L-selectin could indicate a regulatory role of this adhesion molecule in lymphocyte migration into the mammary gland since it has a key role in initiating the multistep adhesion cascade (Springer, 1994). In this study milk and blood T cell subsets were compared at various time points from 3 weeks prior to calving to Week 16 of lactation. Expression of LFA-1 on mononuclear cells and L-selectin expression on several T cell subsets from blood and milk was measured. There was a higher proportion of CD4‡ T cells in blood than milk; while CD8‡ T cells comprised a much higher proportion in milk. This supports earlier studies that indicated CD4 : CD8 ratios were reversed in milk and blood ( Van Kampen and Mallard, 1997; Park et al., 1992), and that CD8‡ T cells were the predominant subset in mammary gland secretions (Taylor et al., 1994). The significance of this finding has not been defined, however given the role of CD8‡ lymphocytes as cytotoxic or suppressor cells (Inoue et al., 1993), it is likely that they would serve as modulators of the immune response in the mammary gland and possibly in the intestine of the immunologically naive neonatal calf. Proportions of L-selectin‡CD4‡ cells in milk were much lower than in blood, particularly around calving, which may suggest less trafficking of CD4‡ cells to the mammary gland at this time. Conversely, L-selectin‡CD8‡ cell proportions were greater than L-selectinÿCD8‡ cells in milk, indicating that the prevalence of CD8‡ cells in the mammary gland may be regulated to some degree by expression of L-selectin. Higher proportions of WC1‡ cells in milk occurred only at calving, suggesting an influence of parturition on the numbers of gamma delta T cells in the mammary gland. In addition, lower proportions of WC1‡ cells in milk versus blood at all other time points may suggest that WC1‡ cells do not traffic preferentially to the mammary gland, except at parturition. The proportion of WC1‡ cells expressing L-selectin was highest at calving, further supporting this. We previously observed that proportions of WC1‡ in blood dropped from highest to lowest between Week 3 and Week 0 (calving)(Van Kampen and Mallard, 1997). L-selectin was expressed on a lower proportion of blood WC1‡ T cells than in milk at calving indicating that the higher proportions of WC1‡ cells in milk at parturition may have been a function of L-selectin expression. Furthermore, the shedding of L-selectin into the surrounding environment and plasma after upregulation as reported in some species (Walcheck et al., 1992; Schleiffenbaum et al., 1992), may account for some of the variation in proportions of L-selectin‡ cells. The L-selectin negative populations might represent subsets of memory T cells since only a proportion of memory T cells express L-selectin and all naive T cells express L-selectin (Picker et al., 1990). However, Howard et al. (1992) found that proportions of bovine CD4‡ memory T cells that do not express L-selectin is much higher than the 10±20% found in the present study whereas the proportions of CD8‡ memory T cells that do not express L-selectin is similar to that reported by Howard et al. (1992). In addition, all WC1‡ T cells in the circulation were L-selectin‡ whereas the present study found much lower proportions of WC1‡L-selectin‡ T cells. The significance of L-selectinÿ T cells in milk and blood of the peripartum cow is unknown. Changes in cell populations in the milk, may not reflect mammary gland cellularity nor phenotypes of cells that have migrated into the gland. For example, Shafer-Weaver et al.

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

31

(1996) found WC1‡ cells were significantly lower around calving (equivalent to Week 0 in the current study) than in mid-lactation (approximately Week 16) in the mammary gland parenchyma; whereas, the current study indicated significantly higher WC1‡ cells in the milk at calving than Week 16. However, cells that enter the gland secretions may still be one of the most important in neonatal immunity, since Reidel-Caspari and Schmidt (1991a, b) have shown that colostral leukocytes can cross the intestine of the neonate and modulate host defense. Lymphocytes in milk are functional (Oksenberg et al., 1985; Sordillo et al., 1991; Torre et al., 1992) and since the neonatal calf is immunologically naive, immune cells such as naive and memory lymphocytes that cross the intestine will be important for immune surveillance of the calf until its own system is more developed. Calves fed milk that contained colostral cells showed increased blastogenic responses to mitogen and increased uptake of bacteria than calves fed milk that was depleted of leukocytes (Reidel-Caspari and Schmidt, 1991a, b). Expression of LFA-1 on mononuclear cells was at lower proportions in milk than blood and did not vary significantly over time in either compartment. This may reflect minimal effects of parturition on LFA-1 expression and perhaps a less important role for LFA-1 in regulating trafficking of these cells into the gland. Shafer-Weaver et al. (1996) also found no difference in LFA-1 expression in blood between two different points in lactation, and proportions of LFA-1‡ cells were lower in mammary gland parenchyma than peripheral blood. In bovine milk CD8‡ and WC1‡ T cells predominate as a proportion of the lymphocyte population, compared to blood during the peripartum period. This may reflect a regulatory role for L-selectin in trafficking of both cell subsets to the mammary gland. Regulation of the mechanisms involved in trafficking of specific lymphocyte subsets into the mammary gland such as differential adhesion molecule expression, may have a significant role in the resolution of mammary gland infections and neonatal immunity.

References Bevilacqua, M.P., 1993. Endothelial-leukocyte adhesion molecules. Ann. Rev. Immunol. 11, 767±804. Figdor, G.G., Van Kooyk, Y., Keizer, G.D., 1990. On the mode of action of LFA-1. Immunology Today 11, 277±280. Gilbert, R.O., Rebhun, W.C., Kim, C.A., Kehrli Jr., M.E., Shuster, D.E., Ackermann, M.R., 1993. Clinical manifestations of leukocyte adhesion deficiency in cattle: 14 cases (1977±1991). JAVMA 202, 445±449. Howard, C.J., Sopp, P., Parson, K.R., 1992. L-selectin expression differentiates T cells isolated from different lymphoid tissues in cattle but does not correlate with memory. Immunology 77, 228±234. Inoue, T., Asano, Y., Matsuoka, S., Furutani-Seiki iM, M., Aizawa, S., Nishimura, H., Shirai, T., Tada, T., 1993. Distinction of mouse CD8‡ suppressor effector T cell clones from cytotoxic T cell clones by cytokine production and CD45 isoforms. J. Immunol. 150, 2121±2126. Kehrli, M.E., Nonnecke, B.J., Roth, J.A., 1989a. Alterations in bovine peripheral blood neutrophil function during the peripartum period. Am. J. Vet. Res. 50:207±214. Kehrli, M.E., Nonnecke, B.J., Roth, J.A., 1989b. Alterations in bovine peripheral blood lymphocyte function during the peripartum period. Am. J. Vet. Res. 50:215±220. Mackay, C.R., Marston, W.L., Dudler, L., Spertini, O., Tedder, T.F., Hein, W.R., 1992. Tissue-specific migration pathways by phenotypically distinct subpopulations of memory T cells. Eur. J. Immunol. 22, 887±895.

32

C. Van Kampen et al. / Veterinary Immunology and Immunopathology 69 (1999) 23±32

Oksenberg, J.R., Persitz, E., Brautbar, C., 1985. Cellular immunity in human milk. Am. J. Rep. Immunol. Microbiol. 8, 125±129. Park, Y.H., Fox, L.K., Hamilton, M.J., Davis, W.C., 1992. Bovine mononuclear leukocyte subpopulation in peripheral blood and mammary gland secretions during lactation. J. Dairy Sci. 75, 998±1006. Picker, L.J., Terstappen, L.W., Rott, L.S., Streeter, P.R., Stein, H., Butcher, E.C., 1990. Differential expression of homing-associated adhesion molecules by T cell subsets in man. J. Immunol. 145, 3247±3255. Reidel-Caspari, G., Schmidt, F.W., 1991a. The influence of colostral leukocytes on the immune system of the neonatal calf. I. Effects on lymphocyte responses. Dtsch. Tieraerztl. Wochenschr. 98:102±107. Reidel-Caspari, G., Schmidt, F.W., 1991b. The influence of colostral leukocytes on the immune system of the neonatal calf. II. Effects on phagocytosis. Dtsch. Tieraerztl. Wochenschr. 98:330±334. Shafer-Weaver, K.A., Pighetti, G.M., Sordillo, L.M., 1996. Diminished mammary gland lymphocyte functions parallel shifts in trafficking patterns during the postpartum period. Proc. Soc. Exp. Biol. Med. 212, 271±279. Schleiffenbaum, B., Spertini, O., Tedder, T.F., 1992. Soluble L-selectin is present in human plasma at high levels and retains functional activity. J. Cell Biol. 119, 229±238. Sordillo, L.M., Campos, M., Babiuk, L.A., 1991. Antibacterial activity of bovine mammary gland lymphocytes following treatment with interleukin-2. J. Dairy Sci. 74, 3370±3374. Sordillo, L.M., Shafer-Weaver, K., DeRosa, D., 1997. Immunobiology of the mammary gland. J. Dairy Sci. 80, 1851±1865. Springer, T.A., 1994. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 76, 301±314. Taylor, B.C., Dellinger, J.D., Cullor, J.S., Stott, J.L., 1994. Bovine milk lymphocytes display the phenotype of memory T cells and are predominantly CD8‡. Cell. Immunol. 156, 245±253. Torre, P.M., Konur, P.K., Oliver, S.P., 1992. Proliferative response of mammary gland mononuclear cells to recombinant bovine interleukin-2. Vet. Immunol. Immunopathol. 32, 351±355. Van Kampen, C., Mallard, B.A., 1997. Effects of peripartum stress and disease on bovine lymphocyte subsets. Vet. Immunol. Immunopathol. 59(1,2), 79±92. Walcheck, B., White, M., Kurk, S., Kishimoto, T.K., Jutila, M.A., 1992. Characterization of the bovine peripheral lymph node receptor: a lectin cell adhesion molecule (LECAM). Eur. J. Immunol. 22, 469±476. Yang, T.J., Mather, J.F., Rabinovsky, E.D., 1988. Changes in subpopulation of lymphocytes in peripheral blood, and supramammary and prescapular lymph nodes in cows with mastitis and normal cows. Vet. Immunol. Immunopathol. 18, 279±285.