T Cell–Mediated Elimination of B7.2 Transgenic B Cells

Immunity, Vol. 6, 327–339, March, 1997, Copyright 1997 by Cell Press

T Cell–Mediated Elimination of B7.2 Transgenic B Cells Sylvie Fournier,* Jeffrey C. Rathmell,† Christopher C. Goodnow,† and James P. Allison* *Department of Molecular and Cellular Biology University of California at Berkeley Berkeley, California, 94720 † Howard Hughes Medical Institute and Department of Microbiology and Immunology Beckman Center Stanford University Stanford, California, 94305

Summary Transgenic mice were generated to explore the effects on lymphoid development and immune function of constitutive expression of murine B7.2 on B and T cells. The number of B lymphocytes in primary and secondary lymphoid tissues is normal in B7.2 transgenic lines expressing low levels of B7.2 on B cells, but markedly reduced in transgenic lines expressing moderate to high levels of the transgene on B cells. This reduction is not due to an intrinsic abnormality of the transgenic B cells, but is rather the consequence of an elimination by an immune mechanism requiring the engagement of CD28 on T cells. Interestingly, during cognate antigen-specific interaction with T cells in vivo, B7.2 transgenic B cells are not eliminated, but proliferate and differentiate normally. Our findings suggest that, in the absence of high affinity ligand for the TCR, the CD28–B7.2 system participates in the regulation of B cell homeostasis.

Introduction T lymphocytes require two signals for optimal proliferation and induction of effector functions (Schwartz, 1990). Signal one is delivered by the engagement of the T cell antigen receptor (TCR) with the peptide–major histocompatibility complex (MHC) complex. The second signal is antigen independent and provided by the ligation of accessory molecules. Activation of the TCR in the absence of costimulatory signals results in T cell unresponsiveness or anergy (Mueller et al., 1989). The occurrence of costimulation in the absence of TCR stimulation is considered to be a neutral event. It is now accepted that the cell surface molecule CD28 is the major costimulatory receptor for T cells (reviewed by Linsley and Ledbetter, 1993; June et al., 1994; Allison, 1994). CD28 is highly homologous to CTLA-4, an inducible protein expressed on T cells (Brunet et al., 1987; Harper et al., 1991). In contrast with CD28, CTLA-4 appears to be a negative regulator of T cell activation (reviewed by Chambers et al., 1996). Although these two receptors have opposing functions, they share common ligands, the members of the B7 family (Linsley et al., 1990, 1991a, 1991b; Azuma et al., 1993a; Lenschow et

al., 1993; Freeman et al., 1993a; Hathcock et al., 1993; Razi-Wolf et al., 1993; Wu et al., 1993; Caux et al., 1994). Two B7 molecules have been identified: B7.1 (CD80) (Freeman et al., 1989, 1991) and B7.2 (CD86) (Azuma et al., 1993a; Freeman et al., 1993a, 1993b; Chen et al., 1994). Transfection studies have demonstrated that both B7.1 and B7.2 can serve as ligands for costimulation in vitro (Linsley et al., 1991b; Gimmi et al., 1991; Freeman et al., 1991, 1993b; Chen et al., 1994). However, it is not known whether these molecules mediate distinct or overlapping functions in vivo. Anti-B7.1 monoclonal antibodies (MAbs) have only minimal effect on primary mixed lymphocyte reactions or T cell proliferation to Mlsa stimulation (Azuma et al., 1993a; Lenschow et al., 1993; Hathcock et al., 1994), and immune responses are only partially impaired in mice in which the B7.1 gene is inactivated by homologous recombination (Sharpe, 1995). In vivo, anti-B7.2 MAb was shown to be more efficient than anti-B7.1 MAb at blocking the expansion of Ag-specific T cells (Kearney et al., 1995) and prolonging allogenic islet graft survival (Lenschow et al., 1995a). B7.2 is expressed at much higher levels than B7.1 on professional antigen-presenting cells (APC) such as dendritic cells, macrophages, and activated B cells (Lenschow et al., 1993, 1994; Caux et al., 1994; Hathcock et al., 1994; Inaba et al., 1994; Stack et al., 1994; Constant et al., 1995). Based on these observations, it has been proposed that B7.2 is the primary costimulatory molecule for the initiation of T cell immune responses. Recent studies have suggested that B7 molecules do not always provide positive costimulatory signals in vivo for T cell–dependent immune responses. Transgenic mice constitutively expressing B7.1 on B lymphocytes are deficient in T-dependent antibody production (Sethna et al., 1994). Moreover, treatment of nonobese diabetic (NOD) mice with anti-B7.1 MAb has been reported to accelerate the disease course (Lenschow et al., 1995b). It has been suggested that these effects may reflect the preferential engagement of CTLA-4 by B7.1. Nevertheless, it has been shown that B7.1 becomes the functionally dominant costimulatory molecule involved in epitope spreading and clinical relapses of murine EAE (Miller et al., 1995). Moreover, B7.1 transgenic models of autoimmunity (Harlan et al., 1994; Guerder et al., 1994a, 1994b) and studies demonstrating enhanced immunogenicity of B7.1-transfected tumors (Chen et al., 1992; Townsend and Allison, 1993; Baskar et al., 1993) indicate that B7.1 has a positive costimulatory function in vivo. Experiments using antibodies to block B7.1 or B7.2 during immune responses in vivo have yielded mixed results. While anti-B7.1 MAb treatment reduced the incidence of disease of murine EAE, anti-B7.2 MAb made the disease worse (Kuchroo et al., 1995). It has been suggested that this differential effect of the B7 MAbs reflected the ability of B7.1 and B7.2 to direct T cell differentiation along the T helper type 1 (Th1) or Th2 pathway, respectively. However, anti-B7.2 MAb treatment blocked the development of diabetes in NOD mice, a Th1-dependent autoimmune disease (Lenschow et al., 1995b). The observation that interruption of B7.1- or

Immunity 328

B7.2-mediated interactions can have opposite outcomes in different models of T cell–mediated autoimmune diseases emphasizes that the interplay between the B7 molecules and their receptors is highly complex. An additional level of complexity in the B7/CD28/ CTLA-4 system is that the expression of the B7 molecules is not exclusively restricted to “conventional” APC. For example, B7.2 is found on unstimulated T cells and B7.1 is expressed after activation (Azuma et al., 1993b; Hathcock et al., 1994; Prabhu Das et al., 1995; Krummel and Allison, 1995). These molecules appear to be reciprocally regulated on the surface of T cell clones (Prabhu Das et al., 1995), but the physiological significance of T cell expression of B7 molecules has yet to be established. To gain a better understanding of the influence and importance of B7.2 on immune responses in vivo, we have generated transgenic mice that constitutively express this molecule in lymphocytes. The transgene we used contains a MHC class I promoter and the immunoglobulin m (Igm) enhancer to allow expression in T and B cells. Analysis of three different transgenic lines reveals that constitutive expression of B7.2 at moderate to high levels on B lymphocytes leads to their elimination by a T cell–mediated mechanism. Our experiments indicate that the elimination of the B7.2 transgenic B cells does not occur when they present a high affinity ligand for the TCR. Together, these results suggest that the B7.2/ CD28 system may play a dual role in the regulation of B cell homeostasis. Results Generation of Transgenic Mice Overexpressing B7-2 To generate transgenic mice overexpressing the mouse B7-2 protein in lymphoid cells, a construct containing the coding region of the B7-2 cDNA, the H-2Kb promoter, and the immunoglobulin heavy chain enhancer (Pircher et al., 1989) was injected into oocytes. Eight founder mice were obtained carrying 5–80 copies of the transgene. Seven founders were crossed successfully to C57Bl/6 mice to establish lines (lines 7, 16, 23, 26, 27,

31, and 33). Heterozygous transgenic mice from all lines were healthy and maintained in a pathogen-free environment according to the Animal Care guidelines (University of California).

B7-2 Transgenic Mice Express the Transgene on B, T, and NK1.11 Cells Cell suspensions from lymphoid tissues were stained with an anti-B7.2 MAb (GL-1) in combination with antibodies directed against various cell surface markers and analyzed by flow cytometry to determine transgene expression in the cell populations. As previously reported (Hathcock et al., 1994; Prabhu Das et al., 1995; Krummel and Allison, 1995), B7.2 was expressed minimally on B lymphocytes and at more significant levels on T cells in nontransgenic animals (Figure 1A). Increased levels of B7.2 were found only on 20%–50% of the T cells in the transgenic line 16. No expression of the transgene was observed in lines 23 and 33. The other transgenic lines showed increased expression of B7.2 on both B and T cells. These lines could be divided into three groups based on the intensity of B7.2 expression on B cells: low (lines 26 and 31), moderate (line 27), and moderate to high (line 7) (Figure 1B). In all lines, T cells expressed very high levels of B7.2 (Figure 1B). Low levels of B7.2 were also detected on NK1.11 cells (data not shown). The levels of B7-2 on macrophages and dendritic cells were identical to those of nontransgenic littermates, and B7.2 expression was not detectable on granulocytes (data not shown). The magnitude of the B7.2 cell surface expression on peripheral B cells in lines 26, 31, and 27 was lower than that of in vitro activated wild-type B cells (Figure 1C). In line 7, the majority of B cells expressed B7.2 levels that were comparable with those of anti-m activated wild-type B cells. In this report we describe the analysis of the heterozygous offspring from the transgenic lines expressing low (line 31), moderate (lines 27), and high levels (line 7) of B7-2 on B lymphocytes. These lines will subsequently be referred as B7.2low (line 31), B7.2 mod (line 27), and B7.2hi (line 7). Figure 1. B7.2 Expression on B and T Cells in Wild-Type and Transgenic Mice Splenocytes from wild-type (A) and B7.2 transgenic mice (B) were stained with MAbs specific for B7.2 (thick lines) and B220 or Thy1.2 or with an isotype-matched irrelevant MAb (thin lines). Data were electronically gated for B2202 (left panel) or Thy-1.21 (right panel) cells. Shown in (C) is B7.2 expression on wild-type B cells stimulated with LPS (thick line) or anti-IgM MAb (broken line). Control staining was carried out with an isotypematched irrelevant MAb (thin line). Data presented are gated on B2201 cells.

Effects of a Constitutive Expression of B7.2 on Lymphocytes 329

Table 1. B and T Cell Populations in Lymphoid Tissues of B7.2low and B7.2 mod Transgenic Mice Spleen

Mice B7.2low Wild type Transgenic B7.2mod Wild type Transgenic

Lymph Nodes

Bone Marrow

CD41a

CD81a

B2201 CD41 CD81 CD44hi (3 1026 ) (3 10 26) (3 1026 ) (%)

CD44hi (%)

B2201 (3 10 26)

75 6 22 24 6 7 12 6 2 20 6 2 69 6 9 27 6 11 44 6 28 39 6 3

25 6 6 74 6 4

7.6 6 2.5 7.9 6 2.4 4.6 6 1.5 8 6 1 18 6 3 45 6 6 32 6 7 26 6 6 9.0 6 2.8 6.6 6 3.3 4.6 6 1.7 12 6 3 31 6 3 41 6 6 26 6 9 20 6 7

74 6 14 18 6 5 11 6 7 13 6 6

CD41a CD41 (3 10 26)

CD81 (3 1026)

CD44hi (%)

CD81a

Blood CD44hi B2201 (%) (%)

B2201 IgM2 (%)

B2201 IgM1 (%)

10 6 4 23 6 4 30 6 14 6.2 6 1.7 7.5 6 1.7 4.3 6 1.4 11 6 3 16 6 5 45 6 5 35 6 5 22 6 7 11 6 10 46 6 16 56 6 4 0.8 6 0.6 9.0 6 2.0 5.0 6 2.0 19 6 4 32 6 1 10 6 1 5 6 2 2 6 1

Values represent the arithmetic mean 6 standard deviation obtained from the analysis of four transgenic mice and four wild-type littermates. Mice were 4–5 weeks old. a Values are expressed as the percentage of CD41 or CD81 cells expressing high levels of CD44.

B7.2mod and B7.2hi Transgenic Mice Have Reduced Numbers of B Cells B7-2low Transgenic Mice Transgenic mice expressing low levels of B7.2 on B cells (line 31) had similar numbers of splenic and lymph node B2201 cells to that found in wild-type littermates. A normal frequency of B cells was observed in the bone marrow (Table 1). No significant differences in the frequency of B cells in blood were found between transgenic and control mice (wild-type, 50 6 5%, n 5 30; transgenic, 43 6 6%, n 5 40). The phenotype of the transgenic B cells, as assessed by the light scatter properties and expression of MHC class II, MEL-14 (Figure 2), and other cell surface markers (B220, IgM, CD40, CD25, CD44, and HSA), was not significantly different from that of wild-type B cells. These results indicate that transgene-induced expression of B7.2 at low levels did not cause significant alterations in B cell maturation. High levels of B7.2 expression on T cells had no major

Figure 2. Phenotype of B7.2 Transgenic B Cells Splenocytes from wild-type (thin lines) and B7.2low (A), B.72mod (B), and B7.2hi (C) transgenic mice (thick lines) were stained with antibodies specific for B220 and I-Ab or MEL-14. Data presented are gated on B2201 splenocytes.

effect on thymocyte subset composition, and the thymus housed normal numbers of cells (data not shown). However, peripheral T cells consistently differed in some respects from their normal counterparts (Table 1). The CD81 T cell population was increased in the spleen but not in the lymph nodes of the transgenic animals. This augmentation was observed in all transgenic mice analyzed, although the magnitude of the increase was variable. The absolute number of CD41 cells was normal in the peripheral lymphoid tissues. In both the spleen and lymph nodes of transgenic mice, an increased fraction of CD81 and CD41 T cells were CD44 hi (Table 1). However, the level and frequency of expression of early activation markers like CD69 and CD25 were not significantly different (data not shown). B7-2 levels on purified B7.2 low transgenic B cells were insufficient to support an allogenic response by CD41 T cells (data not shown), suggesting that the increased frequency of T cells with a partial activated phenotype is not due to the ability of B cells to provide costimulation. However, in contrast with wild-type T cells, highly purified B7.2 transgenic T cells proliferated in response to submitogenic concentrations of anti-CD3. This proliferative response was completely blocked by an anti-B7.2 MAb (Figure 3), indicating that B7.2 transgenic T cells are capable of autocrine costimulation via the CD28 activation pathway. These results suggest that the increased percentage of CD44 hi T cells in B7.2low transgenic mice might result from the interaction between CD28 and B7.2 expressed by the T cells. Intriguingly, despite the increased frequency of activated CD81 and CD41 T cells, histological examination of solid organs revealed no signs of autoimmunity. Moreover, the transgenic mice showed no alterations in serum immunoglobulin isotypes (Figure 4A). B7.2mod Transgenic Mice In contrast with the absence of alterations in the B cell compartment in the B7.2low mice, transgenic mice expressing moderate levels of B7.2 on B lymphocytes (line 27) exhibited a dramatic reduction of B cells in all peripheral lymphoid tissues (Table 1). This reduction was consistently observed when large numbers of transgenic mice were compared with their wild-type littermates for the frequency of B2201 cells in blood (wild-type, 43 6 5%, n 5 29; transgenic, 17 6 10%, n 5 26). The few B cells found in the periphery had a mature phenotype and expressed B220 and surface IgM at levels comparable with wild-type B cells. However, these B cells were

Immunity 330

Figure 3. Transgenic T Cells Proliferate in Response to Submitogenic Amounts of Anti-CD3 We cultured 10 5 purified wild-type (closed bars) or B7.2low transgenic (hatched bars) lymph node T cells on 96-well plates previously coated with 0.1 mg/ml anti-CD3 MAb. Anti-CD28 MAb (1:1000 dilution of ascites), anti-B7.2 MAb (GL-1; 20 mg/ml), or control rat MAb (2.4G2; 20 mg/ml) was added in indicated wells. Cultures were incubated for 56 hr, pulsed with 1 mCi of [3 H]thymidine, and harvested after an additional 16 hr.

slightly larger and expressed higher levels of MHC class II than nontransgenic B cells (Figure 2), suggesting a state of partial activation. Analysis of the bone marrow cells revealed that the proportion of pro/pre-B cells (B2201IgM2) and immature/mature B cells (B2201IgM1) was drastically reduced in bone marrow of transgenic mice (Table 1), raising the possibility that the peripheral B cell deficiency was due to an impairment in B cell development. The T cell composition of primary and secondary lymphoid tissues was not altered in B7.2 mod transgenic mice. As in the B7.2 low transgenic mice, there was an increased frequency of CD81 and CD41 T cells expressing high levels of CD44 (Table 1). B7.2hi Transgenic Mice The phenotype of the transgenic mice expressing the highest levels of the transgene on B cells (line 7) was quite variable (Table 2). Some mice had the same phenotype as the B7.2mod transgenic line: a sharp reduction in the absolute numbers of B220 1 cells in all peripheral lymphoid tissues (Table 2, transgenic mice 2, 5, 7, and 8). Interestingly, some transgenic mice exhibited a less severe reduction of B cells in the spleen with an increased absolute number in the lymph nodes (Table 2, transgenic mice 1, 4, and 6). In these transgenic mice the frequency of B2201 cells in blood was either normal or slightly increased. However, all transgenic mice, regardless of the absolute number of B cells in their peripheral lymphoid tissues, exhibited a very low frequency of B2201 cells in the bone marrow. One possible explanation for the high variation of the phenotype between mice and between tissues is that while constitutive expression of B7.2 has a deleterious effect on B cell development as observed in the B7.2mod transgenic line, this

Figure 4. Serum Immunoglobulin Levels in B7.2 Transgenic Mice Immunoglobulin levels in sera from wild-type littermates and B7.2low (A) or B7.2 hi (B) transgenic mice were measured by ELISA. Each circle represents an individual mouse.

effect may be masked in some B7.2hi animals by an expansion of the peripheral B cell pool as a result of an ongoing immune response. We examined this possibility by comparing the levels of serum immunoglobulins in transgenic and wild-type littermates. Elevated concentrations of serum IgG1 were found in some unimmunized B7.2hi transgenic mice, whereas the titers of other isotypes were not significantly altered (Figure 4B). The presence of an expanded population of activated B cells in certain B7.2 hi transgenic mice is consistent with the increased size of the transgenic B cells, the higher expression of MHC class II, and the lower expression of MEL-14, which was quite dramatic in some animals as shown in Figure 2. Thus, signs of B cell activation were apparent in some (but not all) B7.2hi transgenic mice. However, a reduced number of B cells in the bone marrow was found in all transgenic mice examined, suggesting that B cell maturation in B7.2hi transgenic mice, as in B7.2mod mice, may be highly perturbed. T Cells Are Responsible for the Paucity of B Cells in B7.2 Transgenic Mice The reduced frequency of B cells in the bone marrow of the B7.2mod and B7.2hi transgenic mice was quite unexpected. This phenotype could be the result of an indirect

Effects of a Constitutive Expression of B7.2 on Lymphocytes 331

Table 2. B and T Cell Populations in Lymphoid Tissues of B7.2hi Transgenic Mice Spleen

Lymph Nodes CD41a

CD81a

Bone Marrow CD41a

CD81a

Mice

Age (weeks)

B2201 (3 10 26)

CD41 (3 1026)

CD81 (3 10 26 )

CD44hi (%)

CD44hi (%)

B2201 (3 1026 )

CD41 (3 10 26)

CD81 (3 1026 )

CD44hi (%)

CD44hi (%)

Blood B2201 B2201 IgM2 (%) (%)

B2201 IgM 1 (%)

WT TG1 TG2 WT TG3 WT WT TG4 TG5 TG6 WT TG7 TG8

4 4 4 5 5 7 7 7 7 7 9 9 9

52 45 26 70 84 105 101 68 10 62 102 48 76

15 10 13 14 31 31 27 25 14 27 44 54 67

8 11 9 9 11 16 10 41 13 17 27 86 97

ND — — — — 19 21 32 34 40 20 35 50

— — — — — 30 25 62 55 70 20 60 75

3.6 10.1 1.0 4.3 7.6 5.3 3.8 8.9 1.1 7.4 4.6 1.2 2.6

4.2 7.7 2.9 5.3 5.2 6.8 6.4 8.6 5.3 5.6 6.5 6.6 4.8

2.5 5.5 2.3 3.9 4.7 4.1 3.4 7.8 4.1 5.0 4.9 6.2 5.0

— — — — — 7 6 30 23 37 17 25 34

— — — — — 17 11 48 39 50 21 37 53

40 53 11 36 53 40 53 57 20 74 49 13 8

24 13 0 20 8 14 19 3 1 1 26 0 1

30 0 7 32 6 18 22 3 20 7 25 4 4

WT, wild type; TG, transgenic; ND, not determined. a Values are expressed as the percentage of CD41 or CD81 cells expressing high levels of CD44.

effect on B cell development caused by an inappropriate insertion of the transgene into the genome. However, the effect was observed in multiple founders, making this possibility unlikely. Alternatively, the reduced B cell numbers may be due to a direct effect caused by the expression of the transgene-derived B7.2 itself, or as a consequence of an immune mechanism mediated by the B7.2 transgenic T cells. To assess the latter possibility, transgenic mice were crossed with mice carrying a genomic deletion in the TCRb locus (TCRb2/ 2 mice) and consequently lacking mature ab T cells (Mombaerts et al., 1991). Thus, if an intrinsic defect in B cell precursors was responsible for the effect observed in the transgenic mice, the perturbations in the B cell compartment should be reproduced in the absence of T cells. As shown in Table 3 for the B7.2 mod transgenic line and Figure 5A for the B7.2 hi transgenic line, we found that TCRb 2/ 2 mice carrying the B7.2 transgene had the same number of splenic and lymph node B cells as their nontransgenic littermates. No significant differences in cellular size or expression of MHC class II (Figure 5B) or other cell surface markers (IgM, B220, CD40, CD44, Fas, LFA-1, ICAM-1, HSA, and MEL-14) were observed on B cells from the B7.2 transgenic TCRb2/2 mice versus the nontransgenic B cells. Most importantly, no defect in B lymphopoiesis was apparent in the transgenic animals

in the absence of ab T cells. The absolute numbers of bone marrow B2201IgM2 and B2201IgM1 cells were similar in the transgenic and nontransgenic littermates (Table 3; Figure 5A). These data indicate that the presence of the B7.2 transgene did not perturb B cell differentiation and that T cells are directly responsible for the reduced number of B cells in both B7.2 mod and B7.2 hi transgenic mice. T Cell Expression of the B7.2 Transgene Is Not Necessary for the Elimination of B7.2 Transgenic B Cells Although transgenic T cells proliferated normally when stimulated with anti-CD3 and anti-CD28 MAbs in vitro (Figure 3), it was possible that the elimination of the B7.2 transgenic B cells was the result of a faulty T–B cell contact or an altered functional status of the T cells due to their very high levels of B7.2. To determine whether B7.2 transgenic expression on T cells was necessary for the elimination of the B7.2 transgenic B cells, we examined the fate of B7.2 transgenic B cells exposed in vivo to syngeneic nontransgenic T cells. Syngeneic wild-type T cells were adoptively transferred into B7.2 hi transgenic TCRb 2/ 2 mice and nontransgenic littermates and the number of B cells in the lymphoid organs was determined by flow cytometry 3 weeks after transfer.

Table 3. Recovery of B7.2mod Transgenic B Cells into TCRb2/2, CD282/2, and lpr/lpr Backgrounds

Mice TCRb2/2 TG2 TG1 CD282/2 TG2 TG1 lpr/lpr TG2 TG1

Blood

Spleen

Lymph Nodes

Bone Marrow

n

B2201 (%)

B2201 (3 10 26)

B2201 (3 10 26)

B2201IgM2 (3 1026)

B2201IgM 1 (3 10 26)

4 4

76 6 8 76 6 5

65 6 4 66 6 10

9.2 6 0.4 11.6 6 3.9

10.9 6 3.8 11.5 6 3.0

8.9 6 4.0 8.8 6 4.6

3 3

43 6 9 38 6 8

83 6 16 77 6 21

19.8 6 2.8 18.1 6 2.0

10.4 6 1.4 10.2 6 1.2

5.0 6 1.7 4.9 6 2.4

3 4

52 6 5 666

70 6 22 665

14.0 6 4.0 3.0 6 2.0

12.0 6 3.0 2.0 6 1.0

7.0 6 1.0 0.0 6 1.0

Values represent the arithmetic mean 6 standard deviation. Mice were 5–6 weeks old. n, the number of mice analyzed. TG, transgenic.

Immunity 332

cell surface markers (CD44, CD45RB, MEL-14, CD25, and CD69), was equivalent in the B7.2 transgenic and control mice (data not shown). Identical results were obtained upon transfer of syngeneic nontransgenic T cells into nu/nu mice reconstituted previously with B7.2hi transgenic bone marrow cells (data not shown). Taken together, these studies demonstrate that the elimination of the B7.2 transgenic B cells does not require the expression of the B7.2 transgene on T cells since it also occurs with normal T cells.

Figure 5. B7.2 Transgenic B Cells Are Eliminated In Vivo in the Presence of Syngeneic Wild-Type T Cells Nontransgenic (TG minus) or B7.2 transgenic (TG plus) C57Bl/6 TCRb2/2 littermates were either untreated (minus) or injected (plus) with 20 3 106 lymph nodes T cells from wild-type C57Bl/6 mice. Mice were sacrificed 3 weeks after transfer. Cells from spleen, lymph nodes (axillary, brachial, and inguinal), and bone marrow were counted, stained with anti-B220, and analyzed by flow cytometry to measure the B cells recovery (A). Spleen cells from untreated (B) or T cell–reconstituted mice (C) were also stained with anti-B220 and anti-I-Ab MAbs to measure cell enlargement by forward light scatter (FSC) and cell surface levels of class II antigen. Data were gated electronically for B220-positive cells from nontransgenic (thin lines) and B7.2 transgenic (dark lines) spleen cells.

Figure 5 shows the results of these adoptive transfer experiments. In the absence of T cells, B cells occur in lymphoid organs in similar numbers in B7.2hi transgenic and nontransgenic TCRb 2/ 2 mice. Similarly, the number of splenic, lymph node, and bone marrow B cells in nontransgenic animals was not significantly different whether or not they had been injected with wild-type T cells. By contrast, in four out of five B7.2 transgenic TCRb 2/ 2 mice given nontransgenic syngeneic T cells, the number of B cells in peripheral lymphoid organs was decreased by z50% compared with B7.2 transgenic TCRb 2/ 2 mice not injected with T cells. A drastic reduction of B2201 cells was also apparent in the bone marrow. In the absence of T cells, B7.2 transgenic and nontransgenic B cells had a similar phenotype (Figure 5B). Interestingly, upon transfer of nontransgenic T cells, B7.2 transgenic B cells increased in size and up-regulated MHC class II (Figure 5C), a phenotype similar to that originally observed in the transgenic mice (Figure 3). In contrast, the degree of T cell activation, as assessed by the cellular size and the expression of various

The Elimination of B7.2 Transgenic B Cells Requires the Expression of CD28 by T Cells To determine whether the T cell–mediated elimination of B cells constitutively expressing B7.2 required the engagement of CD28, the B7.2 mod transgenic line was crossed with mutant mice deficient for the CD28 gene (Lucas et al., 1995). In contrast with the consistent, drastic B cell deficiency observed in the intact transgenic animals (Table 1), in the absence of expression of CD28 mice expressing the B7.2 transgene had normal numbers of peripheral B lymphocytes (Table 3). Moreover, the expression of various cell surface antigens (B220, IgM, I-Ab, CD40, CD25, CD44, LFA-1, ICAM-1, MEL-14, and HSA) on B7.2 transgenic B cells was similar to that found on nontransgenic B cells (data not shown). The frequency of B cells in the bone marrow of the B7.2 transgenic CD282/2 mice was also normal (Table 3). This, together with the observation that B cell development was normal in B7.2 transgenic TCRb2/2 mice, indicates that B cells constitutively expressing B7.2 are not impaired in their development. In addition, these results show that the elimination of the B7.2 transgenic B cells is dependent on CD28–B7.2 interactions.

B7.2 Transgenic B Cells Are Not Eliminated, but Proliferate and Differentiate In Vivo in Presence of Antigen and T Cell Help The T cell–mediated elimination of B cells constitutively expressing moderate to high levels of B7.2 was quite surprising and may have been due to a functional abnormality of the transgenic B cells. The proliferative response of B7.2 transgenic B cells induced by crosslinking of surface IgM with anti-IgM antibody, in the presence or absence of interleukin-4, was found to be normal (data not shown). Similarly, the proliferative response and IgM production of wild-type and B7.2 transgenic B cells to the antigen receptor–independent stimulus lipopolysaccharide (LPS) were not different (data not shown). These results demonstrate that the B7.2 transgenic B cells are not defective in their capacity to proliferate and differentiate into antibody-producing cells upon stimulation with polyclonal activators in vitro. To examine whether B7.2 transgenic B cells respond normally to cognate interactions with Th cells, we crossed the B7.2 hi transgenic mice with mice carrying HEL-specific rearranged IgH and L chain transgenes (Goodnow et al., 1988) and assessed the capacity of the B cells to respond to HEL-specific TCR transgenic T cells (Ho et al., 1994) when transferred together into irradiated histocompatible recipients expressing HEL as

Effects of a Constitutive Expression of B7.2 on Lymphocytes 333

Figure 6. B7.2 Transgenic B Cells Proliferate and Differentiate into Antibody-Secreting Cells In Vivo in the Presence of Antigen and T Cell Help Nontransgenic (TG minus) and B7.2hi transgenic (TG plus) HEL-specific B cells (1.5 3 10 6 cells) were adoptively transferred with (plus) or without (minus) 2.5 3 106 CD41 HEL-specific T cells into irradiated HEL-expressing recipients. The recipient mice were sacrificed 5 days after transfer. Spleen cells were enumerated, stained with antiB220 and anti-IgMa MAbs, and analyzed by flow cytometry to determine the number of cells expressing the transgene-encoded IgM a (A). Spleen cells were also analyzed by spot ELISA to determine frequencies of anti-HEL antibody–producing cells (B).

an autoantigen (Rathmell et al., 1995). As shown in Figure 6, in the absence of HEL-specific T cells, equivalent numbers of B7.2 transgenic and nontransgenic HELspecific B cells were recovered in the spleen of the irradiated recipients, indicating that the same number of B cells of each type established themselves after transfer. In the presence of HEL-specific T cells, B7.2 transgenic HEL-specific B cells expanded and differentiated into antibody-secreting cells as efficiently as nontransgenic HEL-specific B cells. These results demonstrate that the presence of the B7.2 transgene did not impair the B cell capacity to process and present antigen to the T cells, to provide appropriate signals to induce T cell effector functions, and to respond to T cell help. Moreover, these data, together with the adoptive transfer experiments into B7.2 transgenic TCRb2/2 mice, indicate that B cells expressing high to moderate levels of B7.2 expand and differentiate during a high affinity antigen-specific interaction with T cells. They are eliminated only when they encounter T cells that lack TCR capable of high affinity recognition of antigens expressed by the B cells. Taken together, these results suggest that CD28–B7.2 interaction provides different signals depending on the TCR–ligand affinity. The Elimination of the B7.2 Transgenic B Cells Is Fas Independent and Not Mediated by CD81 T Cells As shown in Figure 5C, upon transfer of T cells into B7.2 transgenic TCRb2/2 mice, B cells increased in size and up-regulated MHC class II, suggesting that they may be activated prior to their elimination. Fas expression was also induced on B7.2 transgenic TCRb 2/ 2 B cells exposed to normal T cells (data not shown). It has been reported recently that CD41 Th1 cells induce apoptosis of activated B cells by triggering a Fas-dependent pathway (reviewed by Krammer et al., 1994; Nagata and

Figure 7. The Absence of CD81 T Cells Does Not Prevent the Elimination of B7.2 Transgenic B Cells Peripheral blood from 5-week-old nontransgenic (TG minus) and B7.2 hi transgenic (TG plus) b2-microglobulin 2/2 littermates was stained with anti-CD8 and anti-B220 MAbs and analyzed by flow cytometry.

Golstein, 1995). Moreover, Rathmell et al. (1995) have shown that self-reactive B cells chronically binding autoantigen do not proliferate, but instead are eliminated by a Fas-dependent mechanism upon interaction with Ag-specific CD41 T cells. To determine whether the T cell–mediated elimination of B7.2 transgenic B cells was occurring through a Fas–Fas ligand–mediated killing, the B7.2mod transgenic mice were crossed with Fas-deficient mice. As shown in Table 3, the frequency and absolute number of B lymphocytes in the primary and secondary lymphoid organs were very low in the B7.2 transgenic mice homozygous for the lpr mutation, indicating that the T cell–mediated elimination of B cells in the B7.2 transgenic mice does not occur through a Fasdependent mechanism. It has been shown in vitro that CD28–B7 interactions are necessary and sufficient for the proliferation and induction of CTL activity of human and murine CD81 T cells in the absence of exogenous help (Azuma et al., 1992; Harding and Allison, 1993; Lanier et al., 1995; Guerder et al., 1995). Moreover, several studies have demonstrated that transfection of B7 molecules into tumor cells leads to a potent anti-tumor response in vivo mediated by CD81 T cells (Chen et al., 1992; Townsend and Allison, 1993; Baskar et al., 1993; Yang et al., 1995). To examine the possibility that this subset of T cells was responsible for the elimination of the B7.2 transgenic B cells, B7.2hi transgenic mice were crossed to mice lacking CD81 lymphocytes owing to the disruption of the b2-microglobulin gene (Zijlstra et al., 1990). No B cells were detected by flow cytometry in the blood of the b2-microglobulin2/ 2 mice carrying the B7.2 transgene (Figure 7). These results suggest that CD81 T cells are not the primary mediators of the disappearance of the B7.2 transgenic B cells. Discussion Our studies of B7.2 transgenic mice suggest a previously unreported function for the B7.2–CD28 system in the regulation of B cell homeostasis. In two B7.2 transgenic lines with moderate to high levels of expression of the transgene on B cells, the numbers of peripheral and

Immunity 334

bone marrow B cells were significantly reduced (Tables 1 and 2). However, B cells occurred in normal numbers in B7.2 transgenic animals lacking ab mature T cells (Table 3; Figure 5A) or deficient in CD28 expression (Table 3). These results indicate that the paucity of B cells in the B7.2 transgenic mice is not due to a block in B cell development, but is rather the consequence of an elimination by an immune mechanism mediated by the engagement of CD28 on T cells. Two observations indicate that the T cell–mediated elimination of the B cells in the B7.2mod and B7.2 hi transgenic mice is not due to the high expression of B7.2 on T cells. First, B7.2low transgenic mice, which express functionally relevant levels of B7.2 only on T cells, have normal number of B cells in the periphery and in the bone marrow (Table 1). Second, our adoptive transfer experiments in B7.2 transgenic TCRb2/2 mice demonstrate that B cells constitutively expressing B7.2 are eliminated even when exposed to nontransgenic T cells (Figure 5). The disappearance of B cells does not appear to be due to a functional abnormality of the B7.2 transgenic B cells. The transgenic B cells are not impaired in their capacity to proliferate in vitro in response to polyclonal activators (data not shown). Moreover, peptide-pulsed B7.2hi transgenic B cells were able to induce in vitro the proliferation of TCR transgenic T cells specific for either MHC class II or class I as efficiently as LPS-activated wildtype B cells, indicating that the transgene-derived B7.2 is functional (S. F., unpublished data). Most importantly, HEL-specific B7.2 transgenic B cells are not eliminated, but rather proliferate and differentiate in the presence of HEL and HEL-specific T cells in vivo (Figure 6), demonstrating that the expression of the B7.2 transgene does not impair the ability of B cells to respond normally to T cell help. The adoptive transfer experiments in TCRb2/2 mice demonstrate that B cells expressing moderate to high levels of B7.2 are eliminated when exposed to T cells that lack TCR capable of high affinity interaction with self-petide–MHC complexes expressed by the B cells. By contrast, full T–B cell collaboration between HELspecific B7.2 transgenic B cells and HEL-specific T cells is observed in vivo in presence of HEL. From these studies, we suggest that B7.2–CD28 interaction has two important functions in the regulation of B cell responses. During high affinity TCR–ligand interaction, the T cell receives signals from the TCR that induce the expression of CD40 ligand (CD40L) (Roy et al., 1995; Jaiswal et al., 1996). CD40L engagement of CD40 on B cells upregulates the expression of B7 costimulatory molecules (reviewed by Grewal and Flavell, 1996). Under these conditions, CD28 engagement by B7 results in cytokines release by the T cell. These cytokines, together with CD40–CD40L interaction, promote B cell mitogenesis (Banchereau et al., 1994; Clark and Ledbetter, 1994). When a T cell encounters an activated B cell with high level expression of B7.2 but presenting an antigen either ignored by or with low affinity for the TCR, binding of B7.2 to CD28 is not a neutral event but instead promotes B cell death. The physiological relevance of this process may be to limit or terminate clonal expansion of B cells activated by T-independent antigens such as LPS of Gram-negative bacteria or double-stranded RNA. Alternatively, it may represent a mechanism to eliminate newly generated self-reactive B cells that have become

activated by antigen receptor engagement and for which T cell help is not available owing to the self-tolerance in the T cell compartment. The elimination of B cells expressing high levels of B7.2 in the absence of T cell help closely resembles the fate of naive HEL-binding B cells transferred into recipients expressing soluble HEL (Cyster et al., 1994; Cyster and Goodnow, 1995; Fulcher et al., 1996) and that of mature B cells upon cross-linking of murine IgD in vivo (Finkelman et al., 1995). In these studies, B cells that have received a signal via their immunoglobulin receptor, either by antigen binding or MAbs cross-linking, disappear unless T cell help is provided. In these situations, it is very likely that the B cells are induced to express B7.2. Indeed, engagement of surface immunoglobulin with anti-m or anti-IgD MAbs delivers the signals needed for up-regulation of B7.2 (Lenschow et al., 1994). Moreover, it has been shown that B7.2 is rapidly up-regulated on HEL-binding B cells pulsed in vivo with HEL or transferred into HEL-expressing recipients (Constant et al., 1995; Cyster and Goodnow, 1995). It is tempting to postulate that the mechanism of B cell death in the absence of T cell help observed in these studies is similar to the T cell–mediated elimination of B7.2 transgenic B cells in the absence of high affinity antigen-specific TCR engagement observed in our studies. In both cases, the elimination of the B cells was not affected by B cell Fas deficiency (J. C. R., C. C. G., and J. Cyster, unpublished data) and was accompanied by the activation of the B cells before their disappearance (Cyster and Goodnow, 1995; Fulcher et al., 1996). Furthermore, it does not depend on the stage of maturation of the B cells (Fulcher et al., 1996). Blocking experiments with anti-B7.2 antibody will be necessary to determine whether B7.2–CD28 interaction might play a role in the deletion of the B cells in these situations. The mechanism by which B7.2–CD28 interaction in the absence of high affinity antigen-specific TCR engagement results in the elimination of the B cell is not yet clear. One attractive possibility is that B7.2 engagement on the cell surface directly delivers a death signal to the B cell. During a high affinity antigen-specific interaction with T cells this signal might be antagonized by the survival signals transduced by CD40–CD40L interaction or by cytokines secreted by the activated T cells (or both) (Banchereau et al., 1994; Clark and Ledbetter, 1994). B7.2 contains a cytoplasmic tail with multiple sites for potential phosphorylation by protein kinase C and therefore has the potential to be a signaling molecule. Although we cannot formally exclude this possibility, preliminary experiments indicate that the viability of B7.2 transgenic B cells is not affected by B7.2 crosslinking (S. F., unpublished data). It appears unlikely that CTLA-4, which has been shown to deliver an inhibitory signal for T cell responses (reviewed by Chambers et al., 1996), is implicated in the T cell–mediated elimination of B7.2 transgenic B cells, since CTLA-4 is expressed at very low levels if at all in the absence of TCR engagement (Freeman et al., 1992; Lindsten et al., 1993; Linsley et al., 1992; Krummel and Allison, 1995; Perkins et al., 1996). Thus, despite the fact that B7.2 has a higher affinity for CTLA-4 than for CD28 (Linsley et al., 1994), constitutive high level expression of B7.2 on the APC would favor CD28 engagement over CTLA-4 engagement. Furthermore, if B7.2–CTLA-4 interactions were

Effects of a Constitutive Expression of B7.2 on Lymphocytes 335

responsible, the elimination of B7.2 transgenic B cells should also occur in CD282/2 mice, since T cells from CD28-deficient mice are able to express functional levels of CTLA-4 (Green et al., 1994). Therefore, it appears that the death signal is delivered by the T cell to the B cell upon CD28 engagement in the absence of appreciable TCR signaling. This notion suggests that signal two (costimulatory signal) in the absence of signal one (TCR signal) may not be a neutral event for the T cell. T cell responses upon CD28 engagement in the absence of TCR occupancy are not unprecedented (Baroja et al., 1988; Verwilghen et al., 1993; Brinkmann et al., 1996). Alternatively, the death signal may be delivered upon CD28–B7.2 interaction when the TCR engages self-peptide–MHC complexes with partial agonist activity (Jameson and Bevan, 1995). Whether the elimination of B cells constitutively expressing B7.2 is due to the occurrence of costimulation in the absence of TCR stimulation can be assessed by examining the capacity of T cells having a single specificity (TCR transgenic RAG2/ 2 T cells) to induce the elimination of B cells when adoptively transferred into B7.2 transgenic TCRb2/2 mice. These experiments are currently in progress. The T cell–mediated elimination of B7.2 transgenic B cells seems to be more efficient in B7.2mod than in B7.2hi transgenic mice. This effect is not related to the difference in the levels of expression of B7.2 by the peripheral B cells. In both lines, the immature B cell population (B220lowIgM2), which expressed equally high levels of the transgene (data not shown), was similarly reduced in the bone marrow in the presence of ab mature T cells (Tables 1 and 2; Figure 5A). A reduction in the absolute number of B cells was observed in secondary lymphoid tissues of B7.2hi transgenic mice, although a considerable variation among mice and tissues was noted (Table 2). The more probable explanation for this phenomenon is that the T cell–mediated elimination of the B cells is masked by an expansion of the peripheral B cell pool as a result of an ongoing immune response in some animals. This explanation is implied from the observation that some (but not all) B7.2 hi transgenic mice showed elevated levels of serum IgG1 (Figure 4B). Moreover, it is consistent with our data showing that B7.2hi transgenic TCRb2/2 B cells are eliminated when exposed to T cells in the absence of high affinity antigen, but that B7.2 hi transgenic HEL-specific B cells proliferate and differentiate in the presence of HEL and HEL-specific T cells. No expansion of the peripheral B cell compartment was apparent in B7.2 mod transgenic mice, possibly because the levels of B7.2 on B cells are insufficient to drive T cell activation. We do not know whether the immune responses in B7.2 hi transgenic mice result from the presentation of environmental antigens or self-antigens. We did not notice a significant decline in the viability or the health status of B7.2hi transgenic mice compared with control mice with increasing age, arguing against the development of autoimmune diseases in these animals. It is of particular interest that constitutive high levels of B7.2 on B cells resulted in the specific increase of the Th2-induced IgG1 isotype. It has been suggested that B7.2 may preferentially direct the development of Th cells along the Th2 pathway (Kuchroo et al., 1995; Freeman et al., 1995). However, no intrinsic differences in lymphokine production elicited by B7.1

and B7.2 have been detected in other studies (Lanier et al., 1995; Levine et al., 1995). Analysis of the response of B7.2 transgenic TCRb2/2 mice injected with syngeneic T cells and immunized with different antigens may help to resolve this controversial issue. It is noteworthy that the phenotype of transgenic mice constitutively expressing B7.1 on B cells is different from our B7.2 transgenic mice. B7.1 transgenic mice were impaired in their ability to make antibodies in response to T-dependent antigens (Sethna et al., 1994). More recently, it has been shown that spleen cells from double transgenic mice expressing both human m chain and B7.1 transferred into a nontransgenic recipient failed to induce an antibody response or prime for a secondary antibody response following challenge with human m chain in adjuvant (Yuschenkoff et al., 1996). In contrast, we found that constitutive expression of B7.2 on B cells had no effect on their ability to produce antibodies in response to T cell help (Figure 6B). Furthermore, B7.1 transgenic mice had normal numbers of B2201 cells in their lymphoid organs (Sethna et al., 1994), indicating that, in contrast with B7.2, constitutive expression of B7.1 does not trigger the elimination of the B cells by the T cells. The differences in B cell abnormalities associated with a constitutive expression of B7.1 or B7.2 may indicate that the functions of these molecules are not completely overlapping. Indeed, it has been suggested that B7.1 and B7.2 have distinct regulatory functions during the development of diabetes in NOD mice (Lenschow et al., 1995b). Some studies have also suggested that B7.1 and B7.2 might have differential effects on Th cell differentiation (Kuchroo et al., 1995; Freeman et al., 1995). It is not yet clear how the ligation of CD28/ CTLA-4 by B7.1 or B7.2 may result in a different outcome. No differences in second messengers initiated by the two molecules have been reported to date (Ueda et al., 1995; Ghiotto-Ragueneau et al., 1996). However, it has been shown that CD28 engagement with antiCD28 MAb, but not B7.1, activates p21ras (Nune`s et al., 1994). Further, the CDR-3 region of CD28 and CTLA-4 carries determinants for selective recognition of B7.1 and B7.2 (Truneh et al., 1996; Kariv et al., 1996; Morton et al., 1996), raising the possibility of differential signaling in T cell by the two ligands. One aspect of the phenotype of the B7.2 transgenic mice is particularly intriguing. We found that highly purified B7.2 transgenic T cells can proliferate in vitro to limiting amounts of anti-CD3 antibody applied to plastic wells (Figure 3), indicating that these T cells are capable of autocrine stimulation via the CD28 pathway. However, in vivo, peripheral B7.2 transgenic T cells showed only a partial activated phenotype characterized by an upregulation of CD44 but not of early activation markers such as CD69 or CD25. Despite the increased frequency of partially activated CD41 and CD81 T cells, B7.2low transgenic mice showed no signs of autoimmunity and had normal levels of serum immunoglobulin isotypes (Figure 4A). An explanation for this paradoxical finding might be that the regulation of T cell clonal expansion may be altered in B7.2 transgenic mice. As already proposed (Janeway and Bottomly, 1994; Ronchese et al., 1994; Prabhu Das et al., 1995), T cells expressing costimulatory molecules may be able to prime anti-idiotypic regulatory CD81 T cells, which control the expansion

Immunity 336

and function of activated T cells. This mechanism would be in force in B7.2 transgenic mice, where high levels of B7.2 are found on T cells. In this regard, we observed an expansion of the CD81 T cell population in the spleen of many B7.2 transgenic mice (Tables 1 and 2). Alternatively, it is possible that abnormally high level expression of B7.2 on T cells somehow leads to desensitization of the CD28 signal transduction pathway (Linsley et al., 1993). It is also possible that the activation signals generated by the engagement of CD28 by B7.2 expressed on the T cells are overwhelmed by the concomitant engagement of CTLA-4, which transduces signals that down-regulate T cell activation. Further studies will be necessary to distinguish among these possibilities. In summary, our results clearly show that B cells expressing moderate to high levels of B7.2 are eliminated by T cells via a mechanism that requires CD28. This may be a result of a low affinity interaction of the TCR with self-peptides expressed by the transgenic B cells. Alternatively, it may suggest that the B7.2–CD28 pathway, initially shown to be necessary for the initiation of antigen receptor–mediated T cell responses, transduces signals in the absence of appreciable TCR stimulation. Our studies of the B7.2 transgenic mice reveal that the B7.2–CD28 system, in addition to its critical role as a mediator of T cell proliferation, may play a dual role in the regulation of B cell homeostasis. When B cells exposed a high affinity ligand for the TCR, binding of B7.2 to CD28 induces T cell–derived signals, which promotes B cell expansion and differentiation. In the absence of high affinity TCR engagement, ligation of B7.2 to CD28 on the T cell results in the elimination of B cells. This process could be important for the elimination of irrelevant or potentialy dangerous activated B cells (or both) for which T cell help is not available. Experimental Procedures Generation of B7.2 Transgenic Mice A 1.1 kb EcoRI fragment encompassing the complete open reading frame of the murine B7.2 cDNA (provided by Dr. M. Azuma, Juntendo University School of Medicine, Tokyo, Japan) was placed under the transcriptional control of the H2-Kb promoter and immunoglobulin enhancer (Pircher et al., 1989). The transgene was injected into the pronucleus of fertilized eggs from (C57Bl/6J 3 CBA)F1 hybrid donors, which were transferred to pseudopregnant CD-1 mice. Transgenic progeny were identified by Southern blot analysis of tail DNA and propagated by crossing with C57Bl/6J mice. Progeny were tested for the presence of the transgene by staining peripheral blood lymphocytes with anti-B7.2 MAb. Mice used in the described experiments had been backcrossed to C57Bl/6J at least five times. Mice C57Bl/6J, C57Bl/6-TCRb2/2, and B6.MRL-lpr mice were purchased from the Jackson Laboratory (Bar Harbor, ME). CD28- and b2-microglobulin-deficient mice were provided by Dr. D. Y. Loh (Washington School of Medicine, St. Louis, MO) and Dr. D. H. Raulet (University of California at Berkeley), respectively. The derivation of mice carrying the MD4 anti-HEL immunoglobulin H plus L transgenes, the ML5 soluble HEL transgene, and the 3A9 HEL 46-61 peptide-specific TCRa and b transgenes has been described previously (Goodnow et al., 1988; Ho et al., 1994). Identification of mutant or wild-type Fas and CD28 genes in mice tail DNA was performed by PCR analysis as described (Rubio et al., 1996; Lucas et al., 1995). Preparation and Staining of Cell Suspensions Single cell suspensions of femur and tibia bone marrow were prepared by flushing the cavity with RPMI 1640 medium containing 2%

FCS, followed by gentle disaggregation through a 21-gauge needle. Cells were released from spleens and lymph nodes (axillary, brachial, and inguinal) by mincing the tissues and filtrating the resulting suspension through nylon sieves. Peripheral blood lymphocytes were prepared by centrifugation of tail vein blood (z0.3 ml into 1 ml of Alsevier’s solution) over Hystopaque, density 1.0844 (Sigma, St. Louis, MO). Cell suspensions from bone marrow and spleen were treated with 0.165 M NH4 Cl to lyse erythrocytes and washed twice. Cells were counted by hemocytometer, suspended to appropriate concentrations in Hank’s HEPES-buffered salt solution containing 2% FCS and 0.1% sodium azide (FACS medium) and stained with antibodies on ice for 25 min in the presence of 20 mg/ml of purified anti-mouse CD32/CD16 to block antibody binding to murine Fc receptors. Cells were washed twice in FACS medium before flow cytometric analysis on a FACScan (Becton Dickinson, Mountain View, CA). Listmode data were collected with live gating on 10,000 relevant cellular events and analyzed using LYSIS II software (Becton Dickinson). Antibodies used in this study included the following: anti-B220-FITC or biotin (clone RA3-6B2; Caltag Laboratories, South San Francisco, CA), anti-Thy-1.2-PE (clone 5a-8; Caltag), anti-I-Abbiotin (clone KH74; Pharmingen, San Diego, CA), anti-CD62L-PE (clone MEL-14; Pharmingen), anti-CD4-FITC or biotin (clone H129.19; Pharmingen), anti-CD8-PE or FITC (clone 53-5.8; Pharmingen), anti-CD44-FITC or PE (clone IM7; Pharmingen), anti–ab TCR-biotin (clone H57-597; Pharmingen), anti-IgMa-biotin (clone DS-1; Pharmingen), anti-IgM-PE (Caltag), anti-HSA-biotin (clone M1/ 69; Pharmingen), anti-CD40-PE (clone 3123; Pharmingen), anti-LFA1-PE (clone 2D7; Pharmingen), anti-ICAM-1-FITC (clone 3E2; Pharmingen), anti-CD25-FITC or biotin (clone 3C7; Pharmingen), antiCD69-bitotin (clone H1.2F3; Pharmingen), anti-CD45RB-FITC (clone 16A; Pharmingen), anti-B7-2-FITC or PE (clone GL-1; prepared in our laboratory [FITC] or Pharmingen) (PE), anti-Va11-FITC (clone RR8-1; Pharmingen), anti-Gr1 (clone RB6-8C5; Pharmingen), antiNK1.1-biotin (clone PK136; Pharmingen), anti-CD11b-FITC (clone M1/70; Caltag), anti-Fas-FITC (clone Jo2; Pharmingen). TRI-COLORconjugated streptavidin (Caltag) was used to reveal biotin-coupled antibody staining.

T Cell Proliferation Assays Lymph nodes cells were isolated from 6-week-old B7.2low and nontransgenic littermates. Cell suspensions were obtained by gently grinding tissue between frosted end of glass slides and filtrated through nylon sieves. Cells were treated with anti–class II antibodies (28.16.8s and BP107) and a mixture of rabbit and guinea pig complement (GIBCO BRL, Gaithersberg, MA). Viable cells were isolated over Hystopaque 1.119 (Sigma) and washed extensively with complete RPMI 1640 medium (RPMI 1640 supplemented with 2 mM L-glutamine, gentamycin, 5 3 10 25 2-ME, and 10% heat-inactivated FCS). Residual immunoglobulin-positive cells were removed by repetitive panning on rabbit anti–mouse IgG-coated tissue culture plates (100 mg/ml; Jackson ImmunoResearch, West Grove, PA). Preparations analyzed by flow cytometry were found to be .95% Thy-1.21. Round bottom 96-well plates were coated with anti-CD3 (clone 500A2) at 0.1 mg/ml in 50 ml overnight at 48C, washed extensively with PBS, and blocked for 1 hr at 378C with complete RPMI 1640. T cells were added at 105 per well in 200 ml of complete RPMI 1640. Where indicated, anti-CD28 was added at 1:1000 dilution of ascites, anti-B7.2, or isotype control antibody (anti-CD16; clone 2.4G2) at 20 mg/ml. Cultures were incubated at 378C for 56 hr and then pulsed with 1 mCi of [3 H]thymidine (Amersham, Arlington Heights, IL) for an additional 16 hr before harvesting.

Preparation and Activation of Wild-Type B Cells To obtain B cells, splenic suspensions were incubated for 2 hr at 378C on Falcon 3025 tissue culture plates (Becton Dickinson, Cockeysville, MD). Nonadherent cells were removed and depleted of T cells by treatment with anti-CD4 (clone G. K. 1.5) and anti-CD8 (clone clone 3.155) MAbs and complement. Cells were cultured in complete RPMI 1640 for 72 hr at 2.5 3 10 6 per milliliter with 25 mg/ ml LPS (Calbiochem-Novabiochem Corp., La Jolla, CA) or with 10 mg/ml F(ab9)2 anti-IgM (Cappel Research Products, Durham, NC).

Effects of a Constitutive Expression of B7.2 on Lymphocytes 337

Mouse Immunoglobulin Isotype-Specific ELISA Enzyme-linked immunoabsorbent assay (ELISA) plates (Costar, Cambridge, MA) were coated with 100 ml of goat anti–mouse immunoglobulin antibodies (5 mg/ml in 0.1 M NaHCO3 [pH 8.2]; Southern Biotechnology Associates, Birmingham, AL) at 48C overnight. The plates were washed twice with PBS containing 0.05% Tween 20 (PBS/Tween), incubated 1 hr at room temperature with PBS containing 1% of bovine serum albumin (PBS/BSA), and washed three times with PBS/Tween. Test sera were diluted in PBS/BSA, and 100 ml was incubated for 3 hr at room temperature in triplicate ELISA wells. Plates were washed three times with PBS/Tween and incubated for 1 hr with horseradish peroxidase–conjugated goat anti– mouse isotype-specific antibodies (5 mg/ml in PBS containing 1% BSA and 0.05% Tween 20; Southern Biotechnology Associates). Finally, after three washes with PBS/Tween, peroxidase substrate (ABTS, Sigma) at 0.3 mg/ml in 0.1 M citric acid (pH 4.3), 0.03% H2 02, was added. The OD of the reaction mixture was read at 405 nm using an ELISA reader. Concentrations of various immunoglobulin isotypes were calculated from linear standard curves generated with affinity-purified mouse IgM, IgG3, IgG1, IgG2b, and IgG2a proteins (Southern Biotechnology Associates). Adoptive Transfers into TCRb2/2 Mice Lymph nodes T cells from 6-week-old C57Bl/6 mice were prepared as described in T cell proliferation assays, washed extensively with PBS, and injected via the lateral tail vein (20 3 106 cells per recipient). Recipient animals were sex-matched 6-week-old C57Bl/6 TCRb2/2 nontransgenic or B7.2 hi transgenic littermates. Recipient mice were sacrificed 3 weeks after transfer and single cell suspensions from bone marrow, spleen, and lymph nodes (axillary, brachial, and inguinal) were made as described above. Cells were counted by hemocytometer and analyzed by flow cytometry to determine the frequency of B220-positive cells.

Adoptive Transfers into HEL-Expressing Mice C57Bl/6 B7.2hi transgenic mice were crossed to B10.BR/SgSn mice carrying the MD4 anti-HEL immunoglobulin H plus L transgenes to yield (B6 3 B10.Br)F1 offspring. Offspring were screened by flow cytometry analysis of PBL for the presence of the B7.2 transgene and anti-HEL transgene-specific IgM a allotype on B2201 cells. Spleen cell suspensions were incubated with magnetic anti-B220 microbeads (Miltenyi Biotec, Federal Republic of Germany), as described by the manufacturer, followed by a single pass over a MiniMACS column. CD41 T cells from (C57Bl/6 3 B10.BR/SgSn)F1 3A9 TCR transgenic mice were isolated by incubating pooled spleen, mesenteric, and inguinal lymph nodes cell suspensions with magnetic anti-CD4 microbeads (Miltenyi Biotec), followed by two passes over MiniMACS columns. Purity of the CD41 T cells populations was z93%, and the purity of the B2201 cells populations was routinely 90%–95%. Purified B cells (1.5 3 10 6) and CD41 T cells (2.5 3 106 ) were mixed on ice and injected via the lateral tail vein of sublethally irradiated (750 rads) (C57Bl/6 3 B10.BR/SgSn)F1 carrying the ML5 soluble HEL transgene. After 5 days, spleen cell suspensions from the recipient mice were analyzed by immunofluorescent staining and flow cytometry, and antibody-secreting cells were enumerated by spot ELISA assay as described previously (Rathmell et al., 1995).

Acknowledgments We would like to thank D. Cado for performing oocyte injections, S. Grell and P. Schow for technical assistance, and Dr. D. Cassell for useful discussions. We gratefully acknowledge the thoughtful review of the manuscript and helpful discussions of Dr. C. A. Chambers. This work was supported by National Institutes of Health grants to C. C. G. (PO1 AI19512 and PO60 AR20610) and J. P. A. (CA40041). S. F. is a recipient of Fonds de la Recherche en Sante´ du Que´ bec fellowship. C. C. G. is an Investigator of the Howard Hughes Medical Institute.

Received January 27, 1997.

References Allison, J.P. (1994). CD28-B7 interactions in T-cell activation. Curr. Opin. Immunol. 6, 414–419. Azuma, M., Cayabyab, M., Buck, D., Phillips, J.H., and Lanier, L.L. (1992). CD28 interaction with B7 costimulates primary allogeneic proliferative responses and cytotoxicity mediated by small, resting T lymphocytes. J. Exp. Med. 175, 353–358. Azuma, M., Ito, D., Yagita, H., Okumura, K., Phillips, J.H., Lanier, L.L., and Somoza, C. (1993a). B70 antigen is a second ligand for CTLA-4 and CD28. Nature 366, 76–79. Azuma, M., Yssel, H., Phillips, J.H., Spits, H., and Lanier, L.L. (1993b). Functional expression of B7/BB1 on activated T lymphocytes. J. Exp. Med. 177, 845–850. Banchereau, J., Bazan, F., Blanchard, D., Briere, F., Galizzi, J.P., van Kooten, C., Liu, Y.J., Rousset, F., and Saeland, S. (1994). The CD40 antigen and its ligand. Annu. Rev. Immunol. 12, 881–922. Baroja, M.L., Ceuppens, J.L., van Damme, J., and Billiau, A. (1988). Cooperation between an anti-T cell (anti-CD28) monoclonal antibody and monocyte-produced IL-6 in the induction of T cell responsiveness to IL-2. J. Immunol. 141, 1502–1512. Baskar, S., Ostrand-Rosenberg, S., Nabavi, N., Nadler, L.M., Freeman, G.J., and Glimcher, L.H. (1993). Constitutive expression of B7 restores immunogenicity of tumor cells expressing truncated major histocompatibility complex class II molecules. Proc. Natl. Acad. Sci. USA 90, 5687–5690. Brinkmann, V., Kinzel, B., and Kristofic, C. (1996). TCR-independent activation of human CD4145RO2 T cells by anti-CD28 plus IL-2: induction of clonal expansion and priming for a Th2 phenotype. J. Immunol. 156, 4100–4106. Brunet, J.F., Denizot, F., Luciani, M.F., Roux-Dosseto, M., Suzan, M., Mattei, M.G., and Golstein, P. (1987). A new member of the immunoglobulin superfamily: CTLA-4. Nature 328, 267–270. Caux, C., Vanbervliet, B., Massacrier, C., Azuma, M., Okumura, K., Lanier, L.L., and Banchereau, J. (1994). B70/B7-2 is identical to CD86 and is the major functional ligand for CD28 expressed on human dendritic cells. J. Exp. Med. 180, 1841–1847. Chambers, C.A., Krummel, M.F., Boittel, B., Hurwitz, A., Sullivan, T.J., Fournier, S., Cassell, D., Brunner, M., and Allison, J.P. (1996). The role of CTLA-4 in the regulation and initiation of immune T cell responses. Immunol. Rev. 153, 27–46. Chen, C., Gault, A., Shen, L., and Nabavi, N. (1994). Molecular cloning and expression of early T cell costimulatory molecule-1 and its characterization as B7-2 molecule. J. Immunol. 152, 4929–4936. Chen, L., Ashe, S., Brady, W.A., Hellstrom, I., Hellstrom, K.E., Ledbetter, J.A., McGowan, P., and Linsley, P.S. (1992). Costimulation of antitumor immunity by the B7 counterreceptor for the T lymphocyte molecules CD28 and CTLA-4. Cell 71, 1093–1102. Clark, E.A., and Ledbetter, J.A. (1994). How B and T cells talk to each other. Nature 367, 425–428. Constant, S., Schweitzer, N., West, J., Ranney, P., and Bottomly, K. (1995). B lymphocytes can be competent antigen-presenting cells for priming CD41 T cells to protein antigens in vivo. J. Immunol. 155, 3734–3741. Cyster, J.G., and Goodnow, C.C. (1995). Antigen-induced exclusion from follicles and anergy are separate and complementary processes that influence peripheral B cell fate. Immunity 3, 691–701. Cyster, J.G., Hartley, S.B., and Goodnow, C.C. (1994). Competition for follicular niches excludes self-reactive cells from the recirculating B-cell repertoire. Nature 371, 389–395. Finkelman, F.D., Holmes, J.M., Dukhanina, O.I., and Morris, S.C. (1995). Cross-linking of membrane immunoglobulin D, in the absence of T cell help, kills mature B cells in vivo. J. Exp. Med. 181, 515–525. Freeman, G.J., Freedman, A.S., Segil, J.M., Lee, G., Whitman, J.F., and Nadler, L.M. (1989). B7, a new member of the Ig superfamily with unique expression on activated and neoplastic B cells. J. Immunol. 143, 2714–2722. Freeman, G.J., Gray, G.S., Gimmi, C.D., Lombard, D.B., Zhou, L.J., White, M., Fingeroth, J.D., Gribben, J.G., and Nadler, L.M. (1991).

Immunity 338

Structure, expression, and T cell costimulatory activity of the murine homologue of the human B lymphocyte activation antigen B7. J. Exp. Med. 174, 625–631. Freeman, G.J., Lombard, D.B., Gimmi, C.D., Brod, S.A., Lee, K., Laning, J.C., Hafler, D.A., Dorf, M.E., Gray, G.S., Reiser, H., June, C.H., Thompson, C.B., and Nadler, L.M. (1992). CTLA-4 and CD28 mRNA are coexpressed in most T cells after activation: expression of CTLA-4 and CD28 mRNA does not correlate with the pattern of lymphokine production. J. Immunol. 149, 3795–3801. Freeman, G.J., Borriello, F., Hodes, R.J., Reiser, H., Gribben, J.G., Ng, J.W., Kim, J., Goldberg, J.M., Hathcock, K., Laszlo, G., Lombard, L.A., Wang, S., Gray, G.S., Nadler, L.M., and Sharpe, A.H. (1993a). Murine B7-2, an alternative CTLA4 counter-receptor that costimulates T cell proliferation and interleukin 2 production. J. Exp. Med. 178, 2185–2192. Freeman, G.J., Gribben, J.G., Boussiotis, V.A., Ng, J.W., Restivo, V.A., Jr., Lombard, L.A., Gray, G.S., and Nadler, L.M. (1993b). Cloning of B7-2: a CTLA-4 counter-receptor that costimulates human T cell proliferation. Science 262, 909–911. Freeman, G.J., Boussiotis, V.A., Anumanthan, A., Bernstein, G.M., Ke, X.Y., Rennert, P.D., Gray, G.S., Gribben, J.G., and Nadler, L.M. (1995). B7-1 and B7-2 do not deliver identical costimulatory signals, since B7-2 but not B7-1 preferentially costimulates the initial production of IL-4. Immunity 2, 523–532. Fulcher, D.A., Lyons, A.B., Korn, S.L., Cook, M.C., Koleda, C., Parish, C., Fazekas de St. Groth, B., and Basten, A. (1996). The fate of selfreactive B cells depends primarily on the degree of antigen receptor engagement and availability of T cell help. J. Exp. Med. 183, 2313– 2328. Ghiotto-Ragueneau, M., Battifora, M., Truneh, A., Waterfield, M.D., and Olive, D. (1996). Comparison of CD28-B7.1 and B7.2 functional interaction in resting human T cells: phosphatidylinositol 3-kinase association to CD28 and cytokine production. Eur. J. Immunol. 26, 34–41. Gimmi, C.D., Freeman, G.J., Gribben, J.G., Sugita, K., Freedman, A.S., Morimoto, C., and Nadler, L.M. (1991). B-cell surface antigen B7 provides a costimulatory signal that induces T cells to proliferate and secrete interleukin 2. Proc. Natl. Acad. Sci. USA 88, 6575–6579. Goodnow, C.C., Crosbie, J., Adelstein, S., Lavoie, T.B., Smith-Gill, S.J., Brink, R.A., Pritchard-Briscoe, H., Wotherspoon, J.S., Loblay, R.H., Raphael, K., Trent, R.J., and Basten, A. (1988). Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice. Nature 334, 676–682. Green, J.M., Noel, P.J., Sperling, A.I., Walunas, T.L., Gray, G.S., Bluestone, J.A., and Thompson, C.B. (1994). Absence of B7-dependent responses in CD28-deficient mice. Immunity 1, 501–508. Grewal, I.S., and Flavell, R.A. (1996). A central role of CD40 ligand in the regulation of CD41 T-cell responses. Immunol. Today 17, 410–417. Guerder, S., Meyerhoff, J., and Flavell, R.A. (1994a). The role of the T cell costimulator B7-1 in autoimmunity and the induction and maintenance of tolerance to peripheral antigen. Immunity 1, 155–166. Guerder, S., Picarella, D.E., Linsley, P., and Flavell, R.A. (1994b). Costimulator B7-1 confers antigen-presenting-cell function to parenchymal tissue and in conjunction with tumor necrosis factor a leads to autoimmunity in transgenic mice. Proc. Natl. Acad. Sci. USA 91, 5138–5142. Guerder, S., Carding, S.R., and Flavell, R.A. (1995). B7 costimulation is necessary for the activation of the lytic function in cytotoxic T lymphocyte precursors. J. Immunol. 155, 5167–5174. Harding, F.A., and Allison, J.P. (1993). CD28-B7 interactions allow the induction of CD81 cytotoxic T lymphocytes in the absence of exogenous help. J. Exp. Med. 177, 1791–1796. Harlan, D.M., Hengartner, H., Huang, M.L., Kang, Y.H., Abe, R., Moreadith, R.W., Pircher, H., Gray, G.S., Ohashi, P.S., Freeman, G.J., Nadler, L.M., June, C.H., and Aichele, P. (1994). Transgenic mice expressing both B7 and viral glycoprotein on pancreatic b cells along with glycoprotein-specific transgenic T cells develop diabetes due to a breakdown of T lymphocyte unresponsiveness. Proc. Natl. Acad. Sci. USA 91, 3137–3141.

Harper, K., Balzano, C., Rouvier, E., Mattei, M.G., Luciani, M.F., and Golstein, P. (1991). CTLA-4 and CD28 activated lymphocyte molecules are closely related in both mouse and human as to sequence, message expression, gene structure, and chromosomal location. J. Immunol. 147, 1037–1044. Hathcock, K.S., Laszlo, G., Dickler, H.B., Bradshaw, J., Linsley, P., and Hodes, R.J. (1993). Identification of an alternative CTLA-4 ligand costimulatory for T cell activation. Science 262, 905–907. Hathcock, K.S., Laszlo, G., Pucillo, C., Linsley, P., and Hodes, R.J. (1994). Comparative analysis of B7-1 and B7-2 costimulatory ligands: expression and function. J. Exp. Med. 180, 631–640. Ho, W.Y., Cooke, M.P., Goodnow, C.C., and Davis, M.M. (1994). Resting and anergic B cells are defective in CD28-dependent costimulation of naive CD41 T cells. J. Exp. Med. 179, 1539–1549. Inaba, K., Witmer-Pack, M., Inaba, M., Hathcock, K.S., Sakuta, H., Azuma, M., Yagita, H., Okumura, K., Linsley, P.S., Ikehara, S., Muramatsu, S., Hodes, R.J., and Steinman, R.M. (1994). The tissue distribution of the B7-2 costimulator in mice: abundant expression on dendritic cells in situ and during maturation in vitro. J. Exp. Med. 180, 1849–1860. Jaiswal, A.I., Dubey, C., Swain, S.L., and Croft, M. (1996). Regulation of CD40 ligand expression on naive CD4 T cells: a role for TCR but not co-stimulatory signals. Int. Immunol. 8, 275–285. Jameson, S.C., and Bevan, M.J. (1995). T cell receptor antagonists and partial agonists. Immunity 2, 1–11. Janeway, C.A., and Bottomly, K. (1994). Signals and signs for lymphocytes responses. Cell 76, 275–285. June, C.H., Bluestone, J.A., Nadler, L.M., and Thompson, C.B. (1994). The B7 and CD28 receptor families. Immunol. Today 15, 321–331. Kariv, I., Truneh, A., and Sweet, R.W. (1996). Analysis of the site of interaction of CD28 with its counter-receptors CD80 and CD86 and correlation with function. J. Immunol. 157, 29–38. Kearney, E.R., Walunas, T.L., Karr, R.W., Morton, P.A., Loh, D.Y., Bluestone, J.A., and Jenkins, M.K. (1995). Antigen-dependent clonal expansion of a trace population of antigen-specific CD41 T cells in vivo is dependent on CD28 costimulation and inhibited by CTLA-4. J. Immunol. 155, 1032–1036. Krammer, P.H., Dhein, J., Walczak, H., Behrmann, I., Mariani, S., Matiba, B., Fath, M., Daniel, P.T., Knipping, E., Westendorp, M.O., Stricker, K., Ba¨ umler, C., Hellbardt, S., Germer, M., Peter, M.E., and Debatin, K. (1994). The role of APO-1-mediated apoptosis in the immune system. Immunol. Rev. 142, 175–191. Krummel, M.F., and Allison, J.P. (1995). CD28 and CTLA-4 have opposing effects on the response of T cells to stimulation. J. Exp. Med. 182, 459–465. Kuchroo, V.K., Das, M.P., Brown, J.A., Ranger, A.M., Zamvil, S.S., Sobel, R.A., Weiner, H.L., Nabavi, N., and Glimcher, L.H. (1995). B7-1 and B7-2 costimulatory molecules activate differentially the Th1/ Th2 developmental pathways: application to autoimmune disease therapy. Cell 80, 707–718. Lanier, L.L., O’Fallon, S., Somoza, C., Phillips, J.H., Linsley, P.S., Okumura, K., Ito, D., and Azuma, M. (1995). CD80 (B7) and CD86 (B70) provide similar costimulatory signals for T cell proliferation, cytokine production, and generation of CTL. J. Immunol. 154, 97–105. Lenschow, D.J., Su, G.H., Zuckerman, L.A., Nabavi, N., Jellis, C.L., Gray, G.S., Miller, J., and Bluestone, J.A. (1993). Expression and functional significance of an additional ligand for CTLA-4. Proc. Natl. Acad. Sci. USA 90, 11054–11058. Lenschow, D.J., Sperling, A.I., Cooke, M.P., Freeman, G., Rhee, L., Decker, D.C., Gray, G., Nadler, L.M., Goodnow, C.C., and Bluestone, J.A. (1994). Differential up-regulation of the B7-1 and B7-2 costimulatory molecules after Ig receptor engagement by antigen. J. Immunol. 153, 1990–1997. Lenschow, D.J., Zeng, Y., Hathcock, K.S., Zuckerman, L.A., Freeman, G., Thistlethwaite, J.R., Gray, G.S., Hodes, R.J., and Bluestone, J.A. (1995a). Inhibition of transplant rejection following treatment with anti-B7-2 and anti-B7-1 antibodies. Transplantation 60, 1171– 1178.

Effects of a Constitutive Expression of B7.2 on Lymphocytes 339

Lenschow, D.J., Ho, S.C., Sattar, H., Rhee, L., Gray, G., Nabavi, N., Herold, K.C., and Bluestone, J.A. (1995b). Differential effects of antiB7-1 and anti-B7-2 monoclonal antibody treatment on the development of diabetes in the nonobese diabetic mouse. J. Exp. Med. 181, 1145–1155. Levine, B.L., Ueda, Y., Craighead, N., Huang, M.L., and June, C.H. (1995). CD28 ligands CD80 (B7-1) and CD86 (B7-2) induce longterm autocrine growth of CD41 T cells and induce similar patterns of cytokine secretion in vitro. Int. Immunol. 7, 891–904. Lindsten, T., Lee, K.P., Harris, E.S., Petryniak, B., Craighead, N., Reynolds, P.J., Lombard, D.B., Freeman, G.J., Nadler, L.M., Gray, G.S., Thompson, C.B., and June, C.H. (1993). Characterization of CTLA-4 structure and expression on human T cells. J. Immunol. 151, 3489–3496. Linsley, P.S., and Ledbetter, J.A. (1993). The role of the CD28 receptor during T cell responses to antigen. Annu. Rev. Immunol. 11, 191–212. Linsley, P.S., Clark, E.A., and Ledbetter, J.A. (1990). T-cell antigen CD28 mediates adhesion with B cells by interacting with activation antigen B7/BB-1. Proc. Natl. Acad. Sci. USA 87, 5031–5035. Linsley, P.S., Brady, W., Urnes, M., Grosmaire, L.S., Damle, N.K., and Ledbetter, J.A. (1991a). CTLA-4 is a second receptor for the B cell activation antigen B7. J. Exp. Med. 174, 561–569. Linsley, P.S., Brady, W., Grosmaire, L., Aruffo, A., Damle, N.K., and Ledbetter, J.A. (1991b). Binding of the B cell activation antigen B7 to CD28 costimulates T cell proliferation and interleukin 2 mRNA accumulation. J. Exp. Med. 173, 721–730. Linsley, P.S., Greene, J.L., Tan, P., Bradshaw, J., Ledbetter, J.A., Anasetti, C., and Damle, N.K. (1992). Coexpression and functional cooperation of CTLA-4 and CD28 on activated T lymphocytes. J. Exp. Med. 176, 1595–1604. Linsley, P.S., Bradshaw, J., Urnes, M., Grosmaire, L., and Ledbetter, J.A. (1993). CD28 engagement by B7/BB-1 induces transient downregulation of CD28 synthesis and prolonged unresponsiveness to CD28 signaling. J. Immunol. 150, 3161–3169. Linsley, P.S., Greene, J.L., Brady, W., Bajorath, J., Ledbetter, J.A., and Peach, R. (1994). Human B7-1 (CD80) and B7-2 (CD86) bind with similar avidities but distinct kinetics to CD28 and CTLA-4 receptors. Immunity 1, 793–801. Lucas, P.J., Negishi, I., Nakayama, K., Fields, L.E., and Loh, D.Y. (1995). Naive CD28-deficient T cells can initiate but not sustain an in vitro antigen-specific immune response. J. Immunol. 154, 5757– 5768. Miller, S.D., Vanderlugt, C.L., Lenschow, D.J., Pope, J.G., Karandikar, N.J., Dal Canto, M.C., and Bluestone, J.A. (1995). Blockade of CD28/B7-1 interaction prevents epitope spreading and clinical relapses of murine EAE. Immunity 3, 739–745. Mombaerts, P., Clarke, A.R., Hooper, M.L., and Tonegawa, S. (1991). Creation of a large genomic deletion at the T-cell antigen receptor b-subunit locus in mouse embryonic stem cells by gene targeting. Proc. Natl. Acad. Sci. USA 88, 3084–3087. Morton, P.A., Fu, X.T., Stewart, J.A., Giacoletto, K.S., White, S.L., Leysath, C.E., Evans, R.J., Shieh, J.J., and Karr, R.W. (1996). Differential effects of CTLA-4 substitutions on the binding of human CD80 (B7-1) and CD86 (B7-2). J. Immunol. 156, 1047–1054. Mueller, D.L., Jenkins, M.K., and Schwartz, R.H. (1989). Clonal expansion vs. functional clonal activation: a costimulatory pathway determines the outcome of T cell receptor occupancy. Annu. Rev. Immunol. 7, 445–480. Nagata, S., and Golstein, P. (1995). The Fas death factor. Science 267, 1449–1456. Nune`s, J.A., Collette, Y., Truneh, A., Olive, D., and Cantrell, D.A. (1994). The role of p21ras in CD28 signal transduction: triggering of CD28 with antibodies, but not the ligand B7-1, activates p21ras. J. Exp. Med. 180, 1067–1076. Perkins, D., Wang, Z., Donovan, C., He, H., Mark, D., Guan, G., Wang, Y., Walunas, T., Bluestone, J., Listman, J., and Finn, P.W. (1996). Regulation of CTLA-4 expression during T cell activation. J. Immunol. 156, 4154–4159.

Pircher, H., Mak, T.W., Lang, R., Ballhausen, W., Ruedi, E., Hengartner, H., Zinkernagel, R.M., and Burki, K. (1989). T cell tolerance to Mlsa encoded antigens in T cell receptor Vb8.1 chain transgenic mice. EMBO J. 8, 719–727. Prabhu Das, M.R., Zamvil, S.S., Borriello, F., Weiner, H.L., Sharpe, A.H., and Kuchroo, V.K. (1995). Reciprocal expression of co-stimulatory molecules, B7-1 and B7-2, on murine T cells following activation. Eur. J. Immunol. 25, 207–211. Rathmell, J.C., Cooke, M.P., Ho, W.Y., Grein, J., Townsend, S.E., Davis, M.M., and Goodnow, C.C. (1995). CD95 (Fas)-dependent elimination of self-reactive B cells upon interaction with CD41 T cells. Nature 376, 181–184. Razi-Wolf, Z., Galvin, F., Gray, G., and Reiser, H. (1993). Evidence for an additional ligand, distinct from B7, for the CTLA-4 receptor. Proc. Natl. Acad. Sci. USA 90, 11182–11186. Ronchese F., Hausmann, B., Hubele, S., and Lane, P. (1994). Mice transgenic for a soluble form of murine CTLA-4 show enhanced expansion of antigen-specific CD41 T cells and defective antibody production in vivo. J. Exp. Med. 179, 809–817. Roy, M., Aruffo, A., Ledbetter, J., Linsley, P., Kehry, M., and Noelle, R. (1995). Studies on the interdependence of gp39 and B7 expression and function during antigen-specific immune responses. Eur. J. Immunol. 25, 596–603. Rubio, C.F., Kench, J., Russell, D.M., Yawger, R., and Nemazee, D. (1996). Analysis of central B cell tolerance in autoimmune-prone MRL/lpr mice bearing autoantibody transgenes. J. Immunol. 157, 65–71. Schwartz, R.H. (1990). A cell culture model for T lymphocyte clonal anergy. Science 248, 1349–1356. Sethna, M.P., van Parijs, L., Sharpe, A.H., Abbas, A.K., and Freeman, G.J. (1994). A negative regulatory function of B7 revealed in B7-1 transgenic mice. Immunity 1, 415–421. Sharpe, A.H. (1995). Analysis of lymphocyte costimulation in vivo using transgenic and ‘knockout’ mice. Curr. Opin. Immunol. 7, 389–395. Stack, R.M., Lenschow, D.J., Gray, G.S., Bluestone, J.A., and Fitch, F.W. (1994). IL-4 treatment of small splenic B cells induces costimulatory molecules B7-1 and B7-2. J. Immunol. 152, 5723–5733. Townsend, S.E., and Allison, J.P. (1993). Tumor rejection after direct costimulation of CD81 T cells by B7-transfected melanoma cells. Science 259, 368–370. Truneh, A., Reddy, M., Ryan, P., Lyn, S.D., Eichman, C., Couez, D., Hurle, M.R., Sekaly, R.P., Olive, D., and Sweet, R. (1996). Differential recognition by CD28 of its cognate counter receptors CD80 (B7.1) and B70 (B7.2): analysis by site directed mutagenesis. Mol. Immunol. 33, 321–334. Ueda, Y., Levine, B.L., Huang, M.L., Freeman, G.J., Nadler, L.M., June, C.H., and Ward, S.G. (1995). Both CD28 ligands CD80 (B7-1) and CD86 (B7-2) activate phosphatidylinositol 3-kinase, and wortmannin reveals heterogeneity in the regulation of T cell IL-2 secretion. Int. Immunol. 7, 957–966. Verwilghen, J., Vandenberghe, P., Wallays, G., de Boer, M., Anthony, N., Panayi, G.S., and Ceuppens, J.L. (1993). Simultaneous ligation of CD5 and CD28 on resting T lymphocytes induces T cell activation in the absence of T cell receptor/CD3 occupancy. J. Immunol. 150, 835–846. Wu, Y., Guo, Y., and Liu, Y. (1993). A major costimulatory molecule on antigen-presenting cells, CTLA4 ligand A, is distinct from B7. J. Exp. Med. 178, 1789–1793. Yang, G., Hellstrom, K.E., Hellstrom, I., and Chen, L. (1995). Antitumor immunity elicited by tumor cells transfected with B7-2, a second ligand for CD28/CTLA-4 costimulatory molecules. J. Immunol. 154, 2794–2800. Yuschenkoff, V.N., Sethna, M.P., Freeman, G.J., and Parker, D.C. (1996). Coexpression of B7-1 and antigen blocks tolerance induction to antigen presented by resting B cells. J. Immunol. 157, 1987–1995. Zijlstra, M., Bix, M., Simister, N.E., Loring, J.M., Raulet, D.H., and Jaenisch, R. (1990). b2-microglobulin deficient mice lack CD4281 cytolytic T cells. Nature 344, 742–746.