Eur. J. Immunol. 1992. 22: 1713-1718
Jan NaessensV and Diana J. L. Williams International Laboratory for Research on Animal Diseases, Nairobi
CD5+ B cells in trypanosome-infected cattle
1713
Characterization and measurement of CD5+ B cells in normal and Trypanosoma congolense-infected cattle CD5+ B cells in cattle are present in peripheral blood and spleen, but not in lymph nodes, tonsils or Peyer's patches. Compared to classical B cells, they express similar levels of B cell surface markers, but have higher levels of surface IgM. We failed to find evidence for IgD on bovine B lymphocytes. The CD5+ B cells expressed CDllb (Mac-1). Another small subpopulation of B cells carried C D l l b but not CD5. In cattle infected with Trypanosornu congolense, a dramatic increase in the percentage of CD5+ B cells in blood and spleen was observed.This increase occurred 7-10 days after parasites were first detected in the blood and correlated with the increase in serum IgM and the increase in the absolute number of B cells that is typical to trypanosome-infected animals. The increase in B cells was found to be due mainly to the expansion of the CD5+ B cell subpopulation. The cause of the amplification of the CD5+B cells and their possible involvement in the production of autoantibodies and non-parasite-specific antibodies which have been described in trypanosome-infected animals are discussed.
1 Introduction Infections with African trypanosomes lead to hematological changes such as anemia and thrombocytopenia [l] and to immune defects, including a reduction in the capacity of T cells to become activated and proliferate [2] and an excessive activation of the humoral component of the immune system. High concentrations of IgM were found in the serum of mice [3, 41, man [5, 61, rabbits [7] and cattle [X-9, 101 infected with various species of trypanosomes. An enormous increase in the number of non-parasite specific IgM-secreting cells was observed in spleens from Trypanosomu brucei brucei-infected mice. High levels of spleen cells secreting antibodies to SRBC, TNP, pneumococcal polysaccharide (SIII), chicken y-globulin and FITC were detected, suggesting the occurence of a polyclonal B cell activation during the infection [3, 11, 121. Increases of T15 idiotype-bearing antibodies, which have activity for phosphoryl choline, were also recorded [13]. Less than 10 % of Ig secreted by spleen cells from 7: b. brucei-infected mice could be absorbed by homologous parasites [14]. A similar increase in the number of nonspecific antibodysecreting cells was observed in athymic nulnu [15, 111 and thymectomized mice [16], suggesting that these cells were produced by a T cell-independent mechanism. Some antibodies in trypansome-infected hosts have been demonstrated to have activity for autoantigens. Antibodies to tissue antigens were described in infected rabbits [17, 181. Rheumatoid factor-like substances [19], and antibodies against thymocytes, single-stranded (ss) DNA and
[I 103011 Supported by the Belgian government (ABOS). Correspondence: Jan Naessens, ILRAD, PO Box 30709, Nairobi, Kenya
0 VCH Verlagsgesellschaft mbH, D-6940 Wcinheim, 1992
bromelain-treated mouse RBC were found in mice [ll]. Antibodies against RBC and platelets were described in cattle [20]. These properties, a preference for antibodies of the IgM class and activity for autologous antigens, have been associated with antibodies from the Ly-l+ or CD5+ subpopulation of B cells (reviewed in [21, 221). This subset of B cells differs from conventional CD5- B cells by a number of criteria including their surface phenotype, tissue distribution and antibody specificity. They possess the usual B cell markers, but are IgMhlghIgDIOW.Human leukemic CD5+cells have been shown to express C D l l b [23]. Inmice they occur mainly in the peritoneal cavity, are rare in the spleen and undetectable in blood and lymph nodes [24,25]. In man they constitute 10%-30% of the B cells in the peripheral blood [26, 271. It has not yet been established whether they constitute a different lineage with different precursors [28], or whether they are a maturational stage in B cell differentiation, because evidence exists that CD5 can be acquired on the cell surface of normal B cells [29]. CD5+ B cells are responsible for most of the normal serum IgM [24,25,30] and for the response to the thymusindependent antigen a( 1-3) dextran [30]. They do not appear to be involved in a secondary antigen-specific response [31]. In addition, the CD5' B cells produce polyreactive antibodies and antibodies with specificity for autologous antigens [31, 321. Correlations were found between increased numbers of CD5+ B cells and certain autoimmune diseases [33] and cell sorting experiments proved that the CD5+ B cells were responsible for the production of auto-antibodies [26, 271. It is thought that their primary role is to act in homeostasis of the humoral immune system and as a first line defense mechanism when an antigen-specific system is not functioning [21, 22, 341. We have characterized the CD5+ B cell population in cattle and monitored their numbers during 7: congolense infections, to determine whether they could be involved in production of high IgM titers and autoantibodies,which are characteristic of these infections.
+
OO14-298O/92/0707-1713$3.50 .25/O
1714
J. Naessens and D. J. L. Williams
Eur. J. Immunol. 1992. 22: 1713-1718
2 Materials and methods
2.4 Immunofluorescence
2.1 Infection of cattle
Fluorescence staining of cells was done as described previously [44]. About lo6 cells were incubated for 20 min with diluted ascites fluid (SOO- to 5000-fold, depending on the concentration of antibody) at 4°C in 0.05% sodium azide and were then washed twice. FITC-labeled antimouse-Ig (Southern Biotechnology Inc., Birmingham, AL) was added and after a 20-min incubation at 4”C, the cells were washed and fixed in 2 % formaldehyde. For two-color and three-color analysis, mAb were used of different Ig isotypes, and FITC-, PE- (Southern Biotechnology) and biotin-labeled (Amersham Int., Amersham, GB) mouse isotype-specific antibodies were used in a second step. Streptavidin-allophycocyanin(Becton Dickinson) was used in a third step. mAb used in the triple stains for CDS/CDllb/Ig were CC17 or IL-A67 (IgGl), IL-A6 (IgM) and IL-AS8 (IgG2,). Flow cytometry was performed on a FACStar Plus dual laser system. An Argon-ion laser (model 2025, Spectra-Physics, Montafn View, CA) produces the 488-nm beam to excite fluorescein and PE. A dye laser (model 37SB, Spectra-Physics), pumped by an argonion laser (model 164, Spectra-Physics), produced the 630-nm beam which excited the APC. The emission was measured between 630-660 nm for FITC, 575-600 nm for PE and 660-680 nm for APC.
Two-year-old Boran (Bos indicus) and N’Dama (Bos taurus) cattle were infected by bites from five infected Glossina-morsitans centralis, which were infected by feeding on a goat with 7: congolense clone ILNat 3.1 as described [3S]. When their packed cell volume dropped below 1.5 YO,the animals were treated with 7 mg/kg of the trypanocidal drug diminanzene aceturate (Berenil; Hoechst, Frankfurt, FRG). In another experiment, 6month-old Boran calves from the ILRAD breeding herd were infected with Tcongolense clone ILNat 3.1 and were killed sequentially at weekly intervals after infection. Lymphoid organs were removed to analyze the percentages of differrent lymphocyte populations. Blood samples were collected at weekly intervals for routine parasitological and haematological analysis as described before [3S]. 2.2 Preparation of cells
To prepare spleen, lymph node and Peyer’s patch cells, a slice from each organ was minced and the cells were freed by gentle grinding with a glass pestle. The cell suspension was transferred to a 10-ml tube where debris was allowed to sediment. The cell suspension was then transferred to a clean tube and washed twice in Alsever’s solution. To prepare peripheral blood mononuclear cells (PBMC). Whole blood from the jugular vein was drawn into an equal volume of Alsever’s solution and was layered onto FicollPaque (Pharmacia LKB Biotechnology, Uppsala, Sweden). After centrifuging at 1200 x g for 20min, the interface containing the PBMC was lifted off and washed twice in Alsever’s solution. In the later stages of trypanosome infection, contaminating RBC were removed using NH4Cl lysis buffer. 2.3 mAb
All mAb that were used in these studies, except for IL-A6, were tested and discussed at the first Workshop on Bovine and Ovine Leukocyte Antigens, held in Hannover, July 1989 ( Vet. Immunol. Immunopathol., 1991,vol27). CDS is present in cattle in at least two allelic forms [36]. One allele is prevalent in Bos taurus cattle and is detected by mAb CC17 (IgG1) [37]. A second allele, which is expressed in Bos indicus cattle, is detected by mAb IL-A67 (IgGI) [36]. Two mAb were produced against bovine CDllb: mAb IL-A 15 (IgC1) [38]and mAb IL-A6 (IgM),which detects an antigen with the same cell distribution and MW, and which can be inhibited by mAb IL-A1S. mAb to bovine B cell antigens have been described (391and include mAb to WC3 (IL-A6S, IgGzn or CC21, IgGI), which is an antigen that resembles CD21[40]. mAb to bovine Ig light chain (ILAS8, IgGz,) and IgM (IL-A30, IgGl, and IL-ASO, IgG2,) have been described [41,42]. mAb to CD2 (IL-A42, ILA-43), CD4 (IL-All, IL-A12), CD8 (IL-ASl), WC1 (an antigen specific for y b T cells and recognized by IL-A29, IgGl or CClS, IgG2,) [43], and myeloid antigens (IL-A24, IL-A16) have all been described previously (Vet. Irnmunol. Imrnunopathol., 1991, vol. 27).
3 Results 3.1 Cell surface antigens 3.1.1 Fluorometry
Typical examples of two- and three-color stains of cattle peripheral blood lymphocytes are shown in Fig. 1. All T cells, defined by CD2, expressed the CDS antigen strongly (Fig. 1, top left), while the y6 T cells (Fig. 1, top middle), recognized by a mAb to WC1 [43], and some B cells (Fig. 1,top right) expressed CDS weakly.The CDS+ B cells had high amounts of surface IgM. A triple stain for CDS, IgM and C D l l b on lymphocytes taken from a trypanosome-infected cow is represented in Fig. 1 (bottom). The triple stain reveals that some B cells express C D l l b on their surface (Fig. 1, bottom left) and most of the C D l l b + cells are also CD5+ (Fig. 1, bottom middle). By selecting only the IgM+ cells from total PBMC (Fig. 1, bottom right), three different B cell subsets could be distinguished: one which is CDS- and CDllb- and corresponds to the classical B cells, a second which is CDS+ and C D l l b + and corresponds to the Lyl+ or CDS+ B cells described in mouse and man, and a small third fraction which is CDS- but C D l l b + . The same patterns and subpopulations were observed in non-infected animals, although the proportions of CDS+ and C D l l b + B lymphocytes were smaller.The forward scatter of CDS+ and CDSB cells overlapped but was higher for the CDS+ B cells (Fig. 2). 3.1.2 Immunoglobulins
The amount of IgM expressed on the surface of classical CDS- B cells ranged from low to high (Fig. 1, top right). In
Eur. J. Immunol. 1992. 22: 1713-1718
CDS+ B cells in trypanosome-infected cattle
1715
3.1.3 Other B cell antigens Four B cell-specific antigens were analyzed: WC3 [40], WC5, a B cell antigen of 220 kDa and the B cell antigen recognized by mAb VPM30 [39]. No differences in the expression of these B cell antigens were observed between the CD5+ and CD5- B cells (Fig. 4).
3.2 Organ distribution Table 1shows the percentage of CD5' and CDllb+ B cells in the different lymphoid organs of healthy animals. The
M Figure f. Two-color stains (top) and a three-color stain (bottom) of bovine PBMC. In the two-color stains, the expression of CDS is measured on CD2+ T cells (top left), y/6 T cells (top middle) and Ig+ B cells (top right) from peripheral blood of a normal animal. In the three-color analysis, cells from an animal with a high percentage of CDS+ B cells were stained for Ig, C D l l b and CDS and are represented in the three bottom panels. C D l l b is co-expressed on some Ig+ (bottom left panel) and CD5+ cells (bottom middle). When the same cells are gated for B cells (Ig+) only, one observes a subpopulation of B cells which co-express CDS and C D l l b and a another smaller subpopulation which is C D l l b + but CD5(bottom right). All CDS+ B cells expressed C D l l b .
A
B
C
D
20092-
68-
4630-
n
14-
i CD5*B
(I)
200
1000
200
1000
Forward scatter (Channel number)
Figure2. Comparison of the forward and side (90") scatter distribution of CD5- (left) and CDS+ (right) B cells. The same window was drawn in the two panels to aid comparison.
contrast, the expression of IgM was high on all CD5+ B cells. Neither polyspecific Ab nor mAb against bovine IgD have been produced, despite at least 10 fusions in our laboratory alone using purified B cells (resting and activated) as an antigen source, which yielded several mAb t o Ig and B cell markers. To determine whether IgD was present on the surface of bovine B cells, we performed several immune precipitations using mAb to IgM (IL-A5O) and Ig light chains (IL-A58). Two bands corresponding to the p heavy and the Ig light chain were precipitated with both antibodies. and no second heavy chain could be seen with IL-A58 (Fig. 3). To rule out the possibility that a bovine b chain exists with the same molecular weight as the p chain, we preabsorbed the lymphocyte lysate with anti-IgM and used the anti-Ig light chain antibody (IL-A58) to precipitate any remaining Ig. mAb IL-A58 did not precipitate an additional molecule (Fig. 3), proving that only IgM is detectable on the surface of bovine B cells.This analysis was performed using lymphocytes from spleen, lymph node and blood.
Figure 3. Immune precipitation of radioiodinated surface immunoglobulins from bovine B lymphocytes with an antibody to p heavy chain (IL-A5O) and an antibody to light chain (IL-A58).The precipitated Ig were analyzed by SDS-PAGE and exposed for a long time to reveal the weakly labeled light chains. Precipitations: lane A: IL-ASO; lane B: IL-ASS; lane C: IL-ASO, after preabsorption with IL-ASO; lane D: IL-ASS, after preabsorption with IL-ASO. Lane M contains molecular mass markers, in kDa. Arrows indicate Ig (p) heavy and light chains. This experiment shows no evidence for a 6 heavy chain.
e
jwc5
IVPMBO
Log fluorescence (CD5)
Figure 4. Known bovine B cell antigens are expressed on CD5+ B cells at levels identical to classical CD5- B cells.
1716
J. Naessens and D. J. L. Williams
Eur. J. Immunol. 1992. 22: 1713-1718
Table 1. Percentage of B cells expressing CD5 and C D l l b in lymphoid organs of four cattle Animal
1
3
2
4
PBMC 19.8, 30.YaJ 26.2, 41.0 26.0, 39.2 22.1, 30.5 Spleen 2.9. 8.3 6.2, 6.0 4.4, 9.2 9.1, 10.7 Lymph nodes Prescapular 2.6. 2.5 1.3, 7.9 5.1, 5.2 Prefernural 1.2. 5.0 3.7, 0.4 2.9, 3.9 Peyer's patch 0 , 0 0 , 0.7 2.5, 1.9
a) The left number is the percentage of CD5+ B, cells, the right number is the percentage of C D l l b + B cells.
numbers are expressed as the percentage of the total B cells in that organ, since the numbers of B cells can vary between individual cattle [41]. We found that the percentage of CDS+ B cells in the blood of different animals varied between 5 % and 3 5 % . No correlation was observed between percentages of total B cells in the blood and percentages of CDS+ B cells. Very few CDS+ B cells were detected in the spleen, while none were detected in the lymph nodes or Peyer's patches. In all instances, the percentage of C D l l b + B cells was slightly higher than that of the CDS+ B cells, probably due to the small fraction of CDS- CDllb+ B cells.
3 Levels during T.congoZense infections The kinetics of the increase in CD5+ and CDllb+ B cells during a 7: congolense infection are shown in Fig. S for a representative animal (a) and the mean of four animals (b). B cells were counted weekly using two-color FACS analysis with one mAb against either CD5 or C D l l b and the other against IgM or WC3. At the same time, total and differential white blood cell counts were performed and the percentages of B, T, y6 T cells and macrophages were measured so that the absolute number of B cells per ml of blood could be calculated.The percentages of CDllb+ and CDS+ B cells in the total B cell population did not change for the first 2 weeks after infection. However, by the third week, the percentages had almost doubled (Fig. 5 ) . After that time, percentages increased slowly until, in some animals, up to 80% of the B cells in the blood were CD5+/CDllb+.Thisincrease in CDS+ and C D l l b + B cells has been observed in all the cattle infected with Tcongolense analyzed to date. In another experiment, animals were killed at weekly intervals after infection, to monitor the percentages of B cell subpopulations in various lymphoid organs. In spleen, the magnitude and kinetics of the increase in CDS+ B cells was similar to that found in the peripheral blood. Lymph nodes, tonsils and Peyer's patches remained negative for this subpopulation.
1
1
8
15
22
29
36
43
Days after infection
50
57
1
6
15
22
29
36
43
50
57
Day8 after infection
Figure 5. Changes in cell numbers during an infection with Tcongolense in one animal (a) or in the means of cell numbers in four animals @).The parasitemias are shown in the top panels.The middle panels represent the percentages of CD5+ (solid) and C D l l b + (hatched) B cells in the blood that were measured weekly. In the bottom panel, the left bars represent the absolute number of B cells in the blood (the fraction positivc for CD5+ or C D l l b + is dotted), while the hatched bars represent the serum concentration of IgM.
Eur. J. Immunol. 1992. 22: 1713-1718
The sudden increase in CD5+ B cells betwecn weeks 2 and 3 post infection correlated with a rise in the absolute number of B cells in the blood and an increase in the concentration of scrum IgM (Fig. 5). These correlations were found in four N’Dama cattle (Fig. Sa) and in four Boran cattle (not shown) which were part of the same experiment.The Boran cattle had to be treated as they were more susceptible t o the infection [35], but they expressed similar numbers of CD5+ and CD 1l b + B cells to the N’Damas: by week 3 a mean of 41 % CDS+ and 41% C D l l b + B cells were observed and by weck 5 , SO% CDS+ and 65 % C D l l b + B cells. In Fig. 5, the absolute number of CDS+/CDllb+ B cells is presented in the histogram which depicts the rise in the total B cell count (bottom). These data show that the rise in total B cell numbers was due primarily to an amplification of the CDS+/CDIlb+ subpopulation. In all the cattle which have been analyzed to date, the rise in CD5+/CD1lbt B cells occurred 7-10 days after parasites were first observed in the blood.
4 Discussion CDS+ B cells are present in bovine peripheral blood and spleen, but not in lymph nodes, tonsil or Pcyer‘s patches. They express B cell antigens in normal quantities on their surface, but have higher levels of IgM and in this resemble the human CDS+ and mouse Ly-I+ B cells [21, 221. Bovine B cells do not express IgD, irrespective of whether they expresy the CD5 antigen. All our attempts to raise mAb to IgD have failed and precipitation studies with anti-light chain mAb did not reveal a putative S chain. Triple-color flow cytometry showed that CD5+ B cells also express CDl l b (Mac-1). This is true for cells of both normal and trypanosome-infected animals. A small number of C D l l b + CD5- B cells could be distinguished in thc triple stains, which explains why we always observed more C D l l b + B cells than CDS+ B cells. During an infection with 7: congolense, the number of CDS+/CDllb+ B cells rose sharply about 7-10 days after parasites were first detected in the blood. This increase in the percentage of CDS+ B cells correlates with an increase in total number of B cells, which normally occurs around that time in trypansome-infected cattle [45, 461, and with an increase in the concentration of total scrum IgM. The correlation in timing of these three changes -percentage of CDS+ B cells, total number of B cells and concentration of TgM - suggests they may be linked. Our data show that the major part of the increase in total B cells in bovine trypanosomiasis could be accounted for by an amplification of the CDS+/CDllb+B cell subpopulation. Between week 2 (day 15) and week 3 (day 22) post infection, the absolute number of CD5- B cells increased from around lo5 per ml of blood to about 1.4 x lo5, while the number of B cells expressing CDS and/or C D l l b increased more than fourfold from 0.3 x lo5 to 1.3 x lo5. IgM levels more than tripled during that time. Lyl+ and CDS+ B cells have been linked with IgM secretion [24, 25, 30, 471, production of autoantibodies [26, 271 and production of polyspecific antibodies [31, 321. Therefore, it seems likely that the increase in total IgM concentrations in the blood is a direct consequence of the increased level of CDS+/CDllb+ B cells, which occurs at
CD5+ B cells in trypanosome-infectedcattle
1717
the same time. The appearance of autoantibodies and nonspecific antibodies during trypanosome infections described in other reports (see introduction) has previously been attributed to polyclonal B cell activation [3] or to cross-reactions with trypansosome antigens [48]. However, the large numbers of CD5+ B cells that we observed during infections in cattle are a likely source of such antibodies. Our data raise a number of questions. Why are the CDS+ B cells preferentially amplified? What signal(s) contribute to their increase in blood and spleen? Was their increase a direct result of contact with the parasite or a parasite product, or was it induced by a host molecule (e.g. a cytokine) that is produced in response to infection? Several hypotheses can be considered. First, the non-parasitespecific antibodies observed in infected mice are produced by a T cell-independent process [ 11, 161. If they were secreted specifically by CD5+ B cells, as we suggest, then we would expect that the expansion of the CD5+ B cells was also due to a T cell-independent mechanism. Ly-l+ B cells have been shown to respond to a T cell-independent antigen [30], but were not involved in secondary responses [34]. Thcrcfore. the CD5+ B cells that we observe in infected cattle could result from a high level of Tcellindependent activation that is associated with trypanosome infections. From studies in T.h. uhodesiense-infected mice, it is clear that the antibody response to VSG, which is the immunodominant surface antigen, is a combination of T cell-independent and T cell-dependent processes [49]. The T cell-independent response was generated by the live parasites, whereas purified VSG or fixed parasites only induced a T cell-dependent antibody response [49]. Unfortunately, it is difficult to compare infections in different hosts and with different trypansosome strains. Normal bovine B cells differ from human and mouse B cells in a number of ways: they have an inverted x / h Ig light chain ratio [50] and do not express surface IgD. These are two characteristics of mouse Ly-l+ B cells [25, 471. It is possible that cattle have a different lineage of B cells from that of humans and mice, and respond to chronic infections with a high number of B cells which express CD5.The observation of high numbers of IgM+CDS+ cells in bovine leukemia virus-infected cattle and sheep [51, 521, support such a hypothesis. Ruminants have high numbers of naturally occuring y/6 T cells (53),which are thought to represent a “first front” of immunity, different from that mediated by T cells. Likewise, CD5+ B cells could also represent a more “primitive” level of immunity, and might be preferentially activated during infections in ruminants. The authors acknowledge the excellent technical help of Mr. .loseph Nthale, and the expertise of Mr. J. G. Magondu and I? E Mucheru with the FACStar-PLUS.
Received January 29, 1992.
5 References 1 Jenkins, G . C. and Facer, C. A., inTizardI (Ed.), Immunology and Pathogenesis of Trypansosomiasis, CRC Press Inc., Boca Raton 1985, p. 13. 2 Sileghem, M., in Kierszenbaum F. (Ed.), Immunology and Immunosuppression in Infectious Diseases, Marcel Dekker Tnc., New York, in press.
1718
J. Naessens and D. J. L. Williams
3 Hudson, K. M., Byner, C., Freeman, J. andTerry, R. J., Nature 1976. 264: 256. 4 Baltz. T., Baltz, D., Giroud, C. and Pautrizel, R., Infect. Immun. 1981. 32: 979. 5 Houba, V.; Brown, K. N. and Allison, A. C., Clin. Exp. Immunol. 1969. 4: 113 6 Greenwood, B. M. and Whittle, H. C., Trans. R. SOC.Trop. Med. Hyg. 1980. 74: 716. 7 Frommcl, D., Percy, D. Y. E., Masseycff, R . and Good, R. A , , Nature 1970. 228: 1208. 8 Luckins, A. G. and Mehlitz, D., Ann. Trop. Med. Parasitol. 1976. 70: 479. 9 Nielsen, K., Sheppard. J., Holmes,W. and Tizard, I . , Immunology 1078. 35: 817. 10 Williams, D. J. L., Newson, J. and Naessens, J., Immunobiology 1989. Suppl. 4: 192. 11 Kobayakawa, T., Louis, J., Izui, S. and Lambert, F! H., J. Immunol. 1979. 122: 296. 12 Parratt, D. and Herbert, W. J.. in Lumsden, W. H. R. and Evans, D. A. (Eds.), Biology of the Kinetoplastida, vol. 2., Academic Press, London 1979, p. 523. 13 Rose, L. M., Goldman, M. and Lambert, l? H., J. Immunol. 1982. 128: 79. 14 Corsini, A . C., Clayton, C., Askonas, B. A. and Ogilvie, B. M., Clin. Exp. Irnmunol. 1977. 29: 122. 15 Campbell, G. H., Esser, K . M. and Phillips, S. M., Infect. Irnmun. 1978. 20: 714. 16 Askonas, B. A , , Corsini, A. C., Clayton, C. E. and Ogilvie, B. M., Immunology 1979. 36: 313. 17 Mansficld, J. K. and Kreier, J. P.. Infect. Immun. 1972. 5: 648. 18 MacKenzie, A. R. and Boreham, l? L. F., Immunology 1974. 26: 1225. 19 Klein, F.. Mattern, l?. Kornman. H. J. and Bosch,V. D., Clin. Exp. Imrnunol. 1970. 7: 851. 20 Assoku, R. K. G. and Gardiner, l? R., Vet. Parasitol. 1989.'31: 199. 21 Herzenberg. L. A., Stall, A. M., Lalor. l? A., Sidman, C., Moore, W. A,. Parks, D. R. and Herzenberg, L. A., Irnmunol. Rev. 1086. 93: 81. 22 Kipps,'T. J.. Adv. Immunol. 1990. 47: 117. 23 Dc la Hera, A , , Alvarez-Mon, M., Sanchez-Madrid, E , Martinez-A, C. and Durantez, A., Eur. J. Imrnunol. 1988. 18: 1131. 24 Hayakawa. K., Hardy, R . R., Parks, D. R. and Herzenberg, L. A., .I. Exp. Med. 1983. 157: 202. 25 Hayakawa. K., Hardy, R. and Herzenberg, L. A., Eur. J. Immun. 1986. 16: 450. 26 Casali, l?, Burastero, S.E., Nakamura, M., Inghirami, G. and Notkins, A. L.. Science 1987. 236: 77. 27 Hardy. R. R., Hayakawa, K., Shimizu, M.,Yamasaki, K. and Kishimoto, T., Science 1987. 236: 81.
Eur. J. Immunol. 1992. 22: 1713-1718 28 Hayakawa, K., Hardy, R., Herzenberg, L. A. and Herzenberg, L. A., J. Exp. Med. 1985. 161: 1554. 29 Freedman, A. S., Freeman, G.,Whitman, J., Segi, J., Daley, J. and Nadler, L. M., Blood 1989. 73: 202. 30 Forster, I. and Rajewsky, K., Eur. J. Immunol. 1987.17: 521. 31 Casali, l? and Notkins, A. L., Annu. Rev. Immunol. 1989. 7: 513. 32 Casali, l?, Prabhakar, B. S. and Notkins, A. L., Int. Rev. Immunol. 1988. 3: 17. 33 Nakamura, M., Burastero, S. E., Ueki, Y , Larrick, J. W., Notkins, A. L. and Casali, l?, J. Immunol. 1988. 141: 4165. 34 Raveche, E. S., Clin. Immunol. Imrnunopathol. 1990.56: 135. 35 Paling, R. W., Moloo, S. K., Scott, J. R., McOdimba, F. A.. Logan-Henfrey, L. L., Murray, M. and Williams, D. J. L., Parasite Imrnunol. 1991. 13: 413. 36 Howard, C. J., Morrison,W. I., Brown,W. C., Naessens, J. and Sopp, l?, Anim. Genet. 1989. 20: 351. 37 Howard, C. J., Parsons, K. R., Jones, B. V., Sopp, l? and Pocock, D. H., Vet. Immunol. Immunopathol. 1988. 19: 127. 38 Splitter, G. and Morrison,W. I. ,Vet. Immunol. Immunopathol. 1991. 27: 87. 39 Naessens, J. and Howard, C. J., Vet. Immunol. Immunopathol. 1991. 27: 77. 40 Naessens, J., Newson, J., McHugh, N., Howard, C. J., Parsons, K. and Jones, B., Immunology 1990. 69: 525. 41 Naessens, J., Newson, J., Williams, D. J. L. and Lutje, V., Immunology 1988. 63: 569. 42 Williams, D. J. L., Newson, J. and Naessens, J., Vet. Immunol. Immunopathol. 1990. 24: 267. 43 Clevers, H., MacHugh, N. D., Bensaid, A., Dunlap, S., Baldwin, C. L., Kaushal, A , , Iams, K., Howard, C. J. and Morrison, W. I., Eur. J. Immunol. 1990. 20: 809. 44 Naessens, J., Newson, J., Bensaid, A., Teale, A. J., Magondu, J. G. and Black, S. J., J. Immunol. 1985. 135: 4183. 45 Ellis, J. A., Scott, J. R., McHugh, N. D., Gettingy, G. and Davies, W. C., Parasite Immunol. 1987. 9: 363. 46 Williams, D. J. L., Naessens, J . , Scott, J. R. and McOdimba, F, A., Parasite Immunol. 1991. 13: 171. 47 Sidman, C. L., Shultz, L. D., Hardy, R. R . , Hayakawa, K. and Herzenberg, L. A . , Science 1986. 232: 1423. 48 Musoke, A. J., Nantulya,V. M., Barbet, A. F., Kironde, F. and McGuire, T. C., Parmite Immunol. 1981. 3: 97. 49 Reinitz, D. M. and Mansfield, J. M., Infect. Immun. 1990. 58: 2337. 50 Hood, L., Gray, W. R., Sanders, G. and Dreyer, W. J., Cold Spring Harbor Symp. Quant. Biol. 1967. 32: 133. 51 Depelchin, A., Letesson, J. J., Lostrie-trussart, N., Mammerickx, M . , Portetelle, D. and Burny, A., Immunol. Lett. 1989. 20: 69. 52 Letesson, J. J., Mager, A., Mammerickx, M., Burny, A. and Depelchin, A , , Leukemia 1990. 4: 377. 53 Hein,W. R. and Mackay, C. R., Immunol. Today 1991.12: 30.
Jan NaessensV and Diana J. L. Williams International Laboratory for Research on Animal Diseases, Nairobi
CD5+ B cells in trypanosome-infected cattle
1713
Characterization and measurement of CD5+ B cells in normal and Trypanosoma congolense-infected cattle CD5+ B cells in cattle are present in peripheral blood and spleen, but not in lymph nodes, tonsils or Peyer's patches. Compared to classical B cells, they express similar levels of B cell surface markers, but have higher levels of surface IgM. We failed to find evidence for IgD on bovine B lymphocytes. The CD5+ B cells expressed CDllb (Mac-1). Another small subpopulation of B cells carried C D l l b but not CD5. In cattle infected with Trypanosornu congolense, a dramatic increase in the percentage of CD5+ B cells in blood and spleen was observed.This increase occurred 7-10 days after parasites were first detected in the blood and correlated with the increase in serum IgM and the increase in the absolute number of B cells that is typical to trypanosome-infected animals. The increase in B cells was found to be due mainly to the expansion of the CD5+ B cell subpopulation. The cause of the amplification of the CD5+B cells and their possible involvement in the production of autoantibodies and non-parasite-specific antibodies which have been described in trypanosome-infected animals are discussed.
1 Introduction Infections with African trypanosomes lead to hematological changes such as anemia and thrombocytopenia [l] and to immune defects, including a reduction in the capacity of T cells to become activated and proliferate [2] and an excessive activation of the humoral component of the immune system. High concentrations of IgM were found in the serum of mice [3, 41, man [5, 61, rabbits [7] and cattle [X-9, 101 infected with various species of trypanosomes. An enormous increase in the number of non-parasite specific IgM-secreting cells was observed in spleens from Trypanosomu brucei brucei-infected mice. High levels of spleen cells secreting antibodies to SRBC, TNP, pneumococcal polysaccharide (SIII), chicken y-globulin and FITC were detected, suggesting the occurence of a polyclonal B cell activation during the infection [3, 11, 121. Increases of T15 idiotype-bearing antibodies, which have activity for phosphoryl choline, were also recorded [13]. Less than 10 % of Ig secreted by spleen cells from 7: b. brucei-infected mice could be absorbed by homologous parasites [14]. A similar increase in the number of nonspecific antibodysecreting cells was observed in athymic nulnu [15, 111 and thymectomized mice [16], suggesting that these cells were produced by a T cell-independent mechanism. Some antibodies in trypansome-infected hosts have been demonstrated to have activity for autoantigens. Antibodies to tissue antigens were described in infected rabbits [17, 181. Rheumatoid factor-like substances [19], and antibodies against thymocytes, single-stranded (ss) DNA and
[I 103011 Supported by the Belgian government (ABOS). Correspondence: Jan Naessens, ILRAD, PO Box 30709, Nairobi, Kenya
0 VCH Verlagsgesellschaft mbH, D-6940 Wcinheim, 1992
bromelain-treated mouse RBC were found in mice [ll]. Antibodies against RBC and platelets were described in cattle [20]. These properties, a preference for antibodies of the IgM class and activity for autologous antigens, have been associated with antibodies from the Ly-l+ or CD5+ subpopulation of B cells (reviewed in [21, 221). This subset of B cells differs from conventional CD5- B cells by a number of criteria including their surface phenotype, tissue distribution and antibody specificity. They possess the usual B cell markers, but are IgMhlghIgDIOW.Human leukemic CD5+cells have been shown to express C D l l b [23]. Inmice they occur mainly in the peritoneal cavity, are rare in the spleen and undetectable in blood and lymph nodes [24,25]. In man they constitute 10%-30% of the B cells in the peripheral blood [26, 271. It has not yet been established whether they constitute a different lineage with different precursors [28], or whether they are a maturational stage in B cell differentiation, because evidence exists that CD5 can be acquired on the cell surface of normal B cells [29]. CD5+ B cells are responsible for most of the normal serum IgM [24,25,30] and for the response to the thymusindependent antigen a( 1-3) dextran [30]. They do not appear to be involved in a secondary antigen-specific response [31]. In addition, the CD5' B cells produce polyreactive antibodies and antibodies with specificity for autologous antigens [31, 321. Correlations were found between increased numbers of CD5+ B cells and certain autoimmune diseases [33] and cell sorting experiments proved that the CD5+ B cells were responsible for the production of auto-antibodies [26, 271. It is thought that their primary role is to act in homeostasis of the humoral immune system and as a first line defense mechanism when an antigen-specific system is not functioning [21, 22, 341. We have characterized the CD5+ B cell population in cattle and monitored their numbers during 7: congolense infections, to determine whether they could be involved in production of high IgM titers and autoantibodies,which are characteristic of these infections.
+
OO14-298O/92/0707-1713$3.50 .25/O
1714
J. Naessens and D. J. L. Williams
Eur. J. Immunol. 1992. 22: 1713-1718
2 Materials and methods
2.4 Immunofluorescence
2.1 Infection of cattle
Fluorescence staining of cells was done as described previously [44]. About lo6 cells were incubated for 20 min with diluted ascites fluid (SOO- to 5000-fold, depending on the concentration of antibody) at 4°C in 0.05% sodium azide and were then washed twice. FITC-labeled antimouse-Ig (Southern Biotechnology Inc., Birmingham, AL) was added and after a 20-min incubation at 4”C, the cells were washed and fixed in 2 % formaldehyde. For two-color and three-color analysis, mAb were used of different Ig isotypes, and FITC-, PE- (Southern Biotechnology) and biotin-labeled (Amersham Int., Amersham, GB) mouse isotype-specific antibodies were used in a second step. Streptavidin-allophycocyanin(Becton Dickinson) was used in a third step. mAb used in the triple stains for CDS/CDllb/Ig were CC17 or IL-A67 (IgGl), IL-A6 (IgM) and IL-AS8 (IgG2,). Flow cytometry was performed on a FACStar Plus dual laser system. An Argon-ion laser (model 2025, Spectra-Physics, Montafn View, CA) produces the 488-nm beam to excite fluorescein and PE. A dye laser (model 37SB, Spectra-Physics), pumped by an argonion laser (model 164, Spectra-Physics), produced the 630-nm beam which excited the APC. The emission was measured between 630-660 nm for FITC, 575-600 nm for PE and 660-680 nm for APC.
Two-year-old Boran (Bos indicus) and N’Dama (Bos taurus) cattle were infected by bites from five infected Glossina-morsitans centralis, which were infected by feeding on a goat with 7: congolense clone ILNat 3.1 as described [3S]. When their packed cell volume dropped below 1.5 YO,the animals were treated with 7 mg/kg of the trypanocidal drug diminanzene aceturate (Berenil; Hoechst, Frankfurt, FRG). In another experiment, 6month-old Boran calves from the ILRAD breeding herd were infected with Tcongolense clone ILNat 3.1 and were killed sequentially at weekly intervals after infection. Lymphoid organs were removed to analyze the percentages of differrent lymphocyte populations. Blood samples were collected at weekly intervals for routine parasitological and haematological analysis as described before [3S]. 2.2 Preparation of cells
To prepare spleen, lymph node and Peyer’s patch cells, a slice from each organ was minced and the cells were freed by gentle grinding with a glass pestle. The cell suspension was transferred to a 10-ml tube where debris was allowed to sediment. The cell suspension was then transferred to a clean tube and washed twice in Alsever’s solution. To prepare peripheral blood mononuclear cells (PBMC). Whole blood from the jugular vein was drawn into an equal volume of Alsever’s solution and was layered onto FicollPaque (Pharmacia LKB Biotechnology, Uppsala, Sweden). After centrifuging at 1200 x g for 20min, the interface containing the PBMC was lifted off and washed twice in Alsever’s solution. In the later stages of trypanosome infection, contaminating RBC were removed using NH4Cl lysis buffer. 2.3 mAb
All mAb that were used in these studies, except for IL-A6, were tested and discussed at the first Workshop on Bovine and Ovine Leukocyte Antigens, held in Hannover, July 1989 ( Vet. Immunol. Immunopathol., 1991,vol27). CDS is present in cattle in at least two allelic forms [36]. One allele is prevalent in Bos taurus cattle and is detected by mAb CC17 (IgG1) [37]. A second allele, which is expressed in Bos indicus cattle, is detected by mAb IL-A67 (IgGI) [36]. Two mAb were produced against bovine CDllb: mAb IL-A 15 (IgC1) [38]and mAb IL-A6 (IgM),which detects an antigen with the same cell distribution and MW, and which can be inhibited by mAb IL-A1S. mAb to bovine B cell antigens have been described (391and include mAb to WC3 (IL-A6S, IgGzn or CC21, IgGI), which is an antigen that resembles CD21[40]. mAb to bovine Ig light chain (ILAS8, IgGz,) and IgM (IL-A30, IgGl, and IL-ASO, IgG2,) have been described [41,42]. mAb to CD2 (IL-A42, ILA-43), CD4 (IL-All, IL-A12), CD8 (IL-ASl), WC1 (an antigen specific for y b T cells and recognized by IL-A29, IgGl or CClS, IgG2,) [43], and myeloid antigens (IL-A24, IL-A16) have all been described previously (Vet. Irnmunol. Imrnunopathol., 1991, vol. 27).
3 Results 3.1 Cell surface antigens 3.1.1 Fluorometry
Typical examples of two- and three-color stains of cattle peripheral blood lymphocytes are shown in Fig. 1. All T cells, defined by CD2, expressed the CDS antigen strongly (Fig. 1, top left), while the y6 T cells (Fig. 1, top middle), recognized by a mAb to WC1 [43], and some B cells (Fig. 1,top right) expressed CDS weakly.The CDS+ B cells had high amounts of surface IgM. A triple stain for CDS, IgM and C D l l b on lymphocytes taken from a trypanosome-infected cow is represented in Fig. 1 (bottom). The triple stain reveals that some B cells express C D l l b on their surface (Fig. 1, bottom left) and most of the C D l l b + cells are also CD5+ (Fig. 1, bottom middle). By selecting only the IgM+ cells from total PBMC (Fig. 1, bottom right), three different B cell subsets could be distinguished: one which is CDS- and CDllb- and corresponds to the classical B cells, a second which is CDS+ and C D l l b + and corresponds to the Lyl+ or CDS+ B cells described in mouse and man, and a small third fraction which is CDS- but C D l l b + . The same patterns and subpopulations were observed in non-infected animals, although the proportions of CDS+ and C D l l b + B lymphocytes were smaller.The forward scatter of CDS+ and CDSB cells overlapped but was higher for the CDS+ B cells (Fig. 2). 3.1.2 Immunoglobulins
The amount of IgM expressed on the surface of classical CDS- B cells ranged from low to high (Fig. 1, top right). In
Eur. J. Immunol. 1992. 22: 1713-1718
CDS+ B cells in trypanosome-infected cattle
1715
3.1.3 Other B cell antigens Four B cell-specific antigens were analyzed: WC3 [40], WC5, a B cell antigen of 220 kDa and the B cell antigen recognized by mAb VPM30 [39]. No differences in the expression of these B cell antigens were observed between the CD5+ and CD5- B cells (Fig. 4).
3.2 Organ distribution Table 1shows the percentage of CD5' and CDllb+ B cells in the different lymphoid organs of healthy animals. The
M Figure f. Two-color stains (top) and a three-color stain (bottom) of bovine PBMC. In the two-color stains, the expression of CDS is measured on CD2+ T cells (top left), y/6 T cells (top middle) and Ig+ B cells (top right) from peripheral blood of a normal animal. In the three-color analysis, cells from an animal with a high percentage of CDS+ B cells were stained for Ig, C D l l b and CDS and are represented in the three bottom panels. C D l l b is co-expressed on some Ig+ (bottom left panel) and CD5+ cells (bottom middle). When the same cells are gated for B cells (Ig+) only, one observes a subpopulation of B cells which co-express CDS and C D l l b and a another smaller subpopulation which is C D l l b + but CD5(bottom right). All CDS+ B cells expressed C D l l b .
A
B
C
D
20092-
68-
4630-
n
14-
i CD5*B
(I)
200
1000
200
1000
Forward scatter (Channel number)
Figure2. Comparison of the forward and side (90") scatter distribution of CD5- (left) and CDS+ (right) B cells. The same window was drawn in the two panels to aid comparison.
contrast, the expression of IgM was high on all CD5+ B cells. Neither polyspecific Ab nor mAb against bovine IgD have been produced, despite at least 10 fusions in our laboratory alone using purified B cells (resting and activated) as an antigen source, which yielded several mAb t o Ig and B cell markers. To determine whether IgD was present on the surface of bovine B cells, we performed several immune precipitations using mAb to IgM (IL-A5O) and Ig light chains (IL-A58). Two bands corresponding to the p heavy and the Ig light chain were precipitated with both antibodies. and no second heavy chain could be seen with IL-A58 (Fig. 3). To rule out the possibility that a bovine b chain exists with the same molecular weight as the p chain, we preabsorbed the lymphocyte lysate with anti-IgM and used the anti-Ig light chain antibody (IL-A58) to precipitate any remaining Ig. mAb IL-A58 did not precipitate an additional molecule (Fig. 3), proving that only IgM is detectable on the surface of bovine B cells.This analysis was performed using lymphocytes from spleen, lymph node and blood.
Figure 3. Immune precipitation of radioiodinated surface immunoglobulins from bovine B lymphocytes with an antibody to p heavy chain (IL-A5O) and an antibody to light chain (IL-A58).The precipitated Ig were analyzed by SDS-PAGE and exposed for a long time to reveal the weakly labeled light chains. Precipitations: lane A: IL-ASO; lane B: IL-ASS; lane C: IL-ASO, after preabsorption with IL-ASO; lane D: IL-ASS, after preabsorption with IL-ASO. Lane M contains molecular mass markers, in kDa. Arrows indicate Ig (p) heavy and light chains. This experiment shows no evidence for a 6 heavy chain.
e
jwc5
IVPMBO
Log fluorescence (CD5)
Figure 4. Known bovine B cell antigens are expressed on CD5+ B cells at levels identical to classical CD5- B cells.
1716
J. Naessens and D. J. L. Williams
Eur. J. Immunol. 1992. 22: 1713-1718
Table 1. Percentage of B cells expressing CD5 and C D l l b in lymphoid organs of four cattle Animal
1
3
2
4
PBMC 19.8, 30.YaJ 26.2, 41.0 26.0, 39.2 22.1, 30.5 Spleen 2.9. 8.3 6.2, 6.0 4.4, 9.2 9.1, 10.7 Lymph nodes Prescapular 2.6. 2.5 1.3, 7.9 5.1, 5.2 Prefernural 1.2. 5.0 3.7, 0.4 2.9, 3.9 Peyer's patch 0 , 0 0 , 0.7 2.5, 1.9
a) The left number is the percentage of CD5+ B, cells, the right number is the percentage of C D l l b + B cells.
numbers are expressed as the percentage of the total B cells in that organ, since the numbers of B cells can vary between individual cattle [41]. We found that the percentage of CDS+ B cells in the blood of different animals varied between 5 % and 3 5 % . No correlation was observed between percentages of total B cells in the blood and percentages of CDS+ B cells. Very few CDS+ B cells were detected in the spleen, while none were detected in the lymph nodes or Peyer's patches. In all instances, the percentage of C D l l b + B cells was slightly higher than that of the CDS+ B cells, probably due to the small fraction of CDS- CDllb+ B cells.
3 Levels during T.congoZense infections The kinetics of the increase in CD5+ and CDllb+ B cells during a 7: congolense infection are shown in Fig. S for a representative animal (a) and the mean of four animals (b). B cells were counted weekly using two-color FACS analysis with one mAb against either CD5 or C D l l b and the other against IgM or WC3. At the same time, total and differential white blood cell counts were performed and the percentages of B, T, y6 T cells and macrophages were measured so that the absolute number of B cells per ml of blood could be calculated.The percentages of CDllb+ and CDS+ B cells in the total B cell population did not change for the first 2 weeks after infection. However, by the third week, the percentages had almost doubled (Fig. 5 ) . After that time, percentages increased slowly until, in some animals, up to 80% of the B cells in the blood were CD5+/CDllb+.Thisincrease in CDS+ and C D l l b + B cells has been observed in all the cattle infected with Tcongolense analyzed to date. In another experiment, animals were killed at weekly intervals after infection, to monitor the percentages of B cell subpopulations in various lymphoid organs. In spleen, the magnitude and kinetics of the increase in CDS+ B cells was similar to that found in the peripheral blood. Lymph nodes, tonsils and Peyer's patches remained negative for this subpopulation.
1
1
8
15
22
29
36
43
Days after infection
50
57
1
6
15
22
29
36
43
50
57
Day8 after infection
Figure 5. Changes in cell numbers during an infection with Tcongolense in one animal (a) or in the means of cell numbers in four animals @).The parasitemias are shown in the top panels.The middle panels represent the percentages of CD5+ (solid) and C D l l b + (hatched) B cells in the blood that were measured weekly. In the bottom panel, the left bars represent the absolute number of B cells in the blood (the fraction positivc for CD5+ or C D l l b + is dotted), while the hatched bars represent the serum concentration of IgM.
Eur. J. Immunol. 1992. 22: 1713-1718
The sudden increase in CD5+ B cells betwecn weeks 2 and 3 post infection correlated with a rise in the absolute number of B cells in the blood and an increase in the concentration of scrum IgM (Fig. 5). These correlations were found in four N’Dama cattle (Fig. Sa) and in four Boran cattle (not shown) which were part of the same experiment.The Boran cattle had to be treated as they were more susceptible t o the infection [35], but they expressed similar numbers of CD5+ and CD 1l b + B cells to the N’Damas: by week 3 a mean of 41 % CDS+ and 41% C D l l b + B cells were observed and by weck 5 , SO% CDS+ and 65 % C D l l b + B cells. In Fig. 5, the absolute number of CDS+/CDllb+ B cells is presented in the histogram which depicts the rise in the total B cell count (bottom). These data show that the rise in total B cell numbers was due primarily to an amplification of the CDS+/CDIlb+ subpopulation. In all the cattle which have been analyzed to date, the rise in CD5+/CD1lbt B cells occurred 7-10 days after parasites were first observed in the blood.
4 Discussion CDS+ B cells are present in bovine peripheral blood and spleen, but not in lymph nodes, tonsil or Pcyer‘s patches. They express B cell antigens in normal quantities on their surface, but have higher levels of IgM and in this resemble the human CDS+ and mouse Ly-I+ B cells [21, 221. Bovine B cells do not express IgD, irrespective of whether they expresy the CD5 antigen. All our attempts to raise mAb to IgD have failed and precipitation studies with anti-light chain mAb did not reveal a putative S chain. Triple-color flow cytometry showed that CD5+ B cells also express CDl l b (Mac-1). This is true for cells of both normal and trypanosome-infected animals. A small number of C D l l b + CD5- B cells could be distinguished in thc triple stains, which explains why we always observed more C D l l b + B cells than CDS+ B cells. During an infection with 7: congolense, the number of CDS+/CDllb+ B cells rose sharply about 7-10 days after parasites were first detected in the blood. This increase in the percentage of CDS+ B cells correlates with an increase in total number of B cells, which normally occurs around that time in trypansome-infected cattle [45, 461, and with an increase in the concentration of total scrum IgM. The correlation in timing of these three changes -percentage of CDS+ B cells, total number of B cells and concentration of TgM - suggests they may be linked. Our data show that the major part of the increase in total B cells in bovine trypanosomiasis could be accounted for by an amplification of the CDS+/CDllb+B cell subpopulation. Between week 2 (day 15) and week 3 (day 22) post infection, the absolute number of CD5- B cells increased from around lo5 per ml of blood to about 1.4 x lo5, while the number of B cells expressing CDS and/or C D l l b increased more than fourfold from 0.3 x lo5 to 1.3 x lo5. IgM levels more than tripled during that time. Lyl+ and CDS+ B cells have been linked with IgM secretion [24, 25, 30, 471, production of autoantibodies [26, 271 and production of polyspecific antibodies [31, 321. Therefore, it seems likely that the increase in total IgM concentrations in the blood is a direct consequence of the increased level of CDS+/CDllb+ B cells, which occurs at
CD5+ B cells in trypanosome-infectedcattle
1717
the same time. The appearance of autoantibodies and nonspecific antibodies during trypanosome infections described in other reports (see introduction) has previously been attributed to polyclonal B cell activation [3] or to cross-reactions with trypansosome antigens [48]. However, the large numbers of CD5+ B cells that we observed during infections in cattle are a likely source of such antibodies. Our data raise a number of questions. Why are the CDS+ B cells preferentially amplified? What signal(s) contribute to their increase in blood and spleen? Was their increase a direct result of contact with the parasite or a parasite product, or was it induced by a host molecule (e.g. a cytokine) that is produced in response to infection? Several hypotheses can be considered. First, the non-parasitespecific antibodies observed in infected mice are produced by a T cell-independent process [ 11, 161. If they were secreted specifically by CD5+ B cells, as we suggest, then we would expect that the expansion of the CD5+ B cells was also due to a T cell-independent mechanism. Ly-l+ B cells have been shown to respond to a T cell-independent antigen [30], but were not involved in secondary responses [34]. Thcrcfore. the CD5+ B cells that we observe in infected cattle could result from a high level of Tcellindependent activation that is associated with trypanosome infections. From studies in T.h. uhodesiense-infected mice, it is clear that the antibody response to VSG, which is the immunodominant surface antigen, is a combination of T cell-independent and T cell-dependent processes [49]. The T cell-independent response was generated by the live parasites, whereas purified VSG or fixed parasites only induced a T cell-dependent antibody response [49]. Unfortunately, it is difficult to compare infections in different hosts and with different trypansosome strains. Normal bovine B cells differ from human and mouse B cells in a number of ways: they have an inverted x / h Ig light chain ratio [50] and do not express surface IgD. These are two characteristics of mouse Ly-l+ B cells [25, 471. It is possible that cattle have a different lineage of B cells from that of humans and mice, and respond to chronic infections with a high number of B cells which express CD5.The observation of high numbers of IgM+CDS+ cells in bovine leukemia virus-infected cattle and sheep [51, 521, support such a hypothesis. Ruminants have high numbers of naturally occuring y/6 T cells (53),which are thought to represent a “first front” of immunity, different from that mediated by T cells. Likewise, CD5+ B cells could also represent a more “primitive” level of immunity, and might be preferentially activated during infections in ruminants. The authors acknowledge the excellent technical help of Mr. .loseph Nthale, and the expertise of Mr. J. G. Magondu and I? E Mucheru with the FACStar-PLUS.
Received January 29, 1992.
5 References 1 Jenkins, G . C. and Facer, C. A., inTizardI (Ed.), Immunology and Pathogenesis of Trypansosomiasis, CRC Press Inc., Boca Raton 1985, p. 13. 2 Sileghem, M., in Kierszenbaum F. (Ed.), Immunology and Immunosuppression in Infectious Diseases, Marcel Dekker Tnc., New York, in press.
1718
J. Naessens and D. J. L. Williams
3 Hudson, K. M., Byner, C., Freeman, J. andTerry, R. J., Nature 1976. 264: 256. 4 Baltz. T., Baltz, D., Giroud, C. and Pautrizel, R., Infect. Immun. 1981. 32: 979. 5 Houba, V.; Brown, K. N. and Allison, A. C., Clin. Exp. Immunol. 1969. 4: 113 6 Greenwood, B. M. and Whittle, H. C., Trans. R. SOC.Trop. Med. Hyg. 1980. 74: 716. 7 Frommcl, D., Percy, D. Y. E., Masseycff, R . and Good, R. A , , Nature 1970. 228: 1208. 8 Luckins, A. G. and Mehlitz, D., Ann. Trop. Med. Parasitol. 1976. 70: 479. 9 Nielsen, K., Sheppard. J., Holmes,W. and Tizard, I . , Immunology 1078. 35: 817. 10 Williams, D. J. L., Newson, J. and Naessens, J., Immunobiology 1989. Suppl. 4: 192. 11 Kobayakawa, T., Louis, J., Izui, S. and Lambert, F! H., J. Immunol. 1979. 122: 296. 12 Parratt, D. and Herbert, W. J.. in Lumsden, W. H. R. and Evans, D. A. (Eds.), Biology of the Kinetoplastida, vol. 2., Academic Press, London 1979, p. 523. 13 Rose, L. M., Goldman, M. and Lambert, l? H., J. Immunol. 1982. 128: 79. 14 Corsini, A . C., Clayton, C., Askonas, B. A. and Ogilvie, B. M., Clin. Exp. Irnmunol. 1977. 29: 122. 15 Campbell, G. H., Esser, K . M. and Phillips, S. M., Infect. Irnmun. 1978. 20: 714. 16 Askonas, B. A , , Corsini, A. C., Clayton, C. E. and Ogilvie, B. M., Immunology 1979. 36: 313. 17 Mansficld, J. K. and Kreier, J. P.. Infect. Immun. 1972. 5: 648. 18 MacKenzie, A. R. and Boreham, l? L. F., Immunology 1974. 26: 1225. 19 Klein, F.. Mattern, l?. Kornman. H. J. and Bosch,V. D., Clin. Exp. Imrnunol. 1970. 7: 851. 20 Assoku, R. K. G. and Gardiner, l? R., Vet. Parasitol. 1989.'31: 199. 21 Herzenberg. L. A., Stall, A. M., Lalor. l? A., Sidman, C., Moore, W. A,. Parks, D. R. and Herzenberg, L. A., Irnmunol. Rev. 1086. 93: 81. 22 Kipps,'T. J.. Adv. Immunol. 1990. 47: 117. 23 Dc la Hera, A , , Alvarez-Mon, M., Sanchez-Madrid, E , Martinez-A, C. and Durantez, A., Eur. J. Imrnunol. 1988. 18: 1131. 24 Hayakawa. K., Hardy, R . R., Parks, D. R. and Herzenberg, L. A., .I. Exp. Med. 1983. 157: 202. 25 Hayakawa. K., Hardy, R. and Herzenberg, L. A., Eur. J. Immun. 1986. 16: 450. 26 Casali, l?, Burastero, S.E., Nakamura, M., Inghirami, G. and Notkins, A. L.. Science 1987. 236: 77. 27 Hardy. R. R., Hayakawa, K., Shimizu, M.,Yamasaki, K. and Kishimoto, T., Science 1987. 236: 81.
Eur. J. Immunol. 1992. 22: 1713-1718 28 Hayakawa, K., Hardy, R., Herzenberg, L. A. and Herzenberg, L. A., J. Exp. Med. 1985. 161: 1554. 29 Freedman, A. S., Freeman, G.,Whitman, J., Segi, J., Daley, J. and Nadler, L. M., Blood 1989. 73: 202. 30 Forster, I. and Rajewsky, K., Eur. J. Immunol. 1987.17: 521. 31 Casali, l? and Notkins, A. L., Annu. Rev. Immunol. 1989. 7: 513. 32 Casali, l?, Prabhakar, B. S. and Notkins, A. L., Int. Rev. Immunol. 1988. 3: 17. 33 Nakamura, M., Burastero, S. E., Ueki, Y , Larrick, J. W., Notkins, A. L. and Casali, l?, J. Immunol. 1988. 141: 4165. 34 Raveche, E. S., Clin. Immunol. Imrnunopathol. 1990.56: 135. 35 Paling, R. W., Moloo, S. K., Scott, J. R., McOdimba, F. A.. Logan-Henfrey, L. L., Murray, M. and Williams, D. J. L., Parasite Imrnunol. 1991. 13: 413. 36 Howard, C. J., Morrison,W. I., Brown,W. C., Naessens, J. and Sopp, l?, Anim. Genet. 1989. 20: 351. 37 Howard, C. J., Parsons, K. R., Jones, B. V., Sopp, l? and Pocock, D. H., Vet. Immunol. Immunopathol. 1988. 19: 127. 38 Splitter, G. and Morrison,W. I. ,Vet. Immunol. Immunopathol. 1991. 27: 87. 39 Naessens, J. and Howard, C. J., Vet. Immunol. Immunopathol. 1991. 27: 77. 40 Naessens, J., Newson, J., McHugh, N., Howard, C. J., Parsons, K. and Jones, B., Immunology 1990. 69: 525. 41 Naessens, J., Newson, J., Williams, D. J. L. and Lutje, V., Immunology 1988. 63: 569. 42 Williams, D. J. L., Newson, J. and Naessens, J., Vet. Immunol. Immunopathol. 1990. 24: 267. 43 Clevers, H., MacHugh, N. D., Bensaid, A., Dunlap, S., Baldwin, C. L., Kaushal, A , , Iams, K., Howard, C. J. and Morrison, W. I., Eur. J. Immunol. 1990. 20: 809. 44 Naessens, J., Newson, J., Bensaid, A., Teale, A. J., Magondu, J. G. and Black, S. J., J. Immunol. 1985. 135: 4183. 45 Ellis, J. A., Scott, J. R., McHugh, N. D., Gettingy, G. and Davies, W. C., Parasite Immunol. 1987. 9: 363. 46 Williams, D. J. L., Naessens, J . , Scott, J. R. and McOdimba, F, A., Parasite Immunol. 1991. 13: 171. 47 Sidman, C. L., Shultz, L. D., Hardy, R. R . , Hayakawa, K. and Herzenberg, L. A . , Science 1986. 232: 1423. 48 Musoke, A. J., Nantulya,V. M., Barbet, A. F., Kironde, F. and McGuire, T. C., Parmite Immunol. 1981. 3: 97. 49 Reinitz, D. M. and Mansfield, J. M., Infect. Immun. 1990. 58: 2337. 50 Hood, L., Gray, W. R., Sanders, G. and Dreyer, W. J., Cold Spring Harbor Symp. Quant. Biol. 1967. 32: 133. 51 Depelchin, A., Letesson, J. J., Lostrie-trussart, N., Mammerickx, M . , Portetelle, D. and Burny, A., Immunol. Lett. 1989. 20: 69. 52 Letesson, J. J., Mager, A., Mammerickx, M., Burny, A. and Depelchin, A , , Leukemia 1990. 4: 377. 53 Hein,W. R. and Mackay, C. R., Immunol. Today 1991.12: 30.
