Unexpected Role of B and T Lymphocyte Attenuator in Sustaini

【摘要】 b and t lymphocyte attenuator (btla; cd272) can deliver inhibitory signals to b and t cells upon binding its ligand herpesvirus entry mediator. because cd28, ctla-4, programmed death-1, and icos regulate the development of acute graft-vs-host disease (gvhd), we wished to assess if btla also played a role in this t cell-mediated response. in the nonirradiated parental-into-f1 model of acute gvhd, btla+/+ and btlac/c donor lymphocytes showed equivalent engraftment and expansion during the first week of the alloresponse. unexpectedly, btlac/c donor t cells failed to sustain gvhd, showing a decline in surviving donor cell numbers beginning at day 9 and greatly reduced by day 11. similarly, inhibition of btla-herpesvirus entry mediator engagement by in vivo administration of a blocking anti-btla ab also caused reduced survival of donor cells. microarray analysis revealed several genes that were differentially expressed by btlac/c and btla+/+ donor cd4+ t cells preceding the decline in btlac/c donor t cells. several genes influencing th cell polarization were differentially expressed by btla+/+ and btlac/c donor cells. additionally, the re-expression of the il-7r subunit that occurs in btla+/+ donor cells after 1 wk of in vivo allostimulation was not observed in btlac/c donor cd4+ cells. the striking loss of btlac/c t cells in this model indicates a role for btla activity in sustaining cd4+ t cell survival under the conditions of chronic stimulation in the nonirradiated parental-into-f1 gvhd.

【关键词】 unexpected lymphocyte attenuator sustaining survival allostimulation

introduction

b and t lymphocyte attenuator (btla)2 is a lymphoid-specific cell surface receptor that is expressed by b cells, t cells, dendritic cells, macrophages, and nk cells in c57bl/6 mice (1, 2, 3, 4). btla shows both structural and expression polymorphisms in mice, with balb/c and c57bl/6 alleles differing at 10 aa within the extracellular region, and the balb/c allele lacking expression by macrophages or nk cells (2). functional analysis has suggested that btla exerts inhibitory actions, indicating a role more similar to ctla-4 and programmed death-1 (pd-1) than to activating receptors of the cd28/b7 family (1, 4).

btla represses ag-driven t cell proliferation upon binding its ligand, herpesvirus entry mediator (hvem) (5), a member of the tnfr superfamily, expressed mainly by t lymphocytes and immature dendritic cells (6). an agonist ab to murine btla inhibits il-2 secretion and t cell proliferation in vitro, further suggesting an inhibitory role for btla (4). phosphorylation of btla leads to recruitment of src homology domain 2-containing protein tyrosine phosphatases (shp-1 and shp-2) to two tyrosine motifs in the cytoplasmic domain of mouse and human btla (3, 5). human btla possesses an additional tyrosine residue that may also participate in inhibitory signaling (7). btla expression is increased on in vivo-anergized t cells, but not on cd25+ t regulatory cells, in contrast to ctla-4 and pd-1, suggesting it may have a distinct role in modulating immune responses (2). notably, mouse and human btla contain a conserved intracellular tyrosine motif suggestive of a grb-2 recruitment site (1, 3), which can interact with grb-2 and the p85 subunit of pi3k in vitro (8). although the function of this region has not been established, it suggests that btla may have the capability to function in a costimulatory or prosurvival manner.

the effects of btla-hvem interactions may also be influenced by the interactions of hvem with its additional ligands, tumor necrosis family members light (tnfsf14) and lymphotoxin (lt) (9). in contrast to the inhibitory nature of btla-hvem interactions, hvem-light binding exerts a costimulatory effect on t cell activation (10, 11, 12). this regulatory network is increasingly complex because hvem has the potential to bind btla and light simultaneously (13, 14, 15). consequently, the balance between costimulatory and inhibitory signals may be regulated by a complex including btla, hvem, and light.

graft-vs-host disease (gvhd) is an immune response against alloantigens, such as foreign mhc molecules (reviewed in refs. 16 and 17). in the parental-into-f1 model of gvhd, parental t cells react against alloantigens of the f1 host, whereas f1 host t cells are tolerant to parental cells (18, 19). the type of gvhd that develops in nonirradiated, immunocompetent f1 recipients is dependent on the strain of parental donor cells transferred (18). in the most widely reported version of this model, transfer of c57bl/6 (h-2b) splenocytes into a nonirradiated, immunocompetent c57bl/6 x dba/2 f1 (b6d2f1; h-2b/d) host induces acute gvhd (agvhd), characterized by a th1 cytokine driven, cell-mediated response against host tissues (18, 20). donor t cells exert cytotoxic effects through fas-fas ligand interactions, perforin, and inflammatory cytokines (21, 22, 23). in contrast, transfer of dba/2 (h-2d) splenocytes into b6d2f1 hosts results in a chronic form of gvhd (cgvhd), characterized by production of th2 cytokines and activation of host b cells, resulting in autoantibody development and a systemic lupus erythematosus-like syndrome (24, 25).

in this study, a related parental-into-f1 model of gvhd was examined. in this model, transfer of c57bl/6 donor cells into c57bl/6 x balb/c f1 (cb6f1; h-2b/d) yields agvhd, whereas balb/c donor cells produce chronic gvhd (26). although gvhd induced in cb6f1 mice is generally similar to that in b6d2f1, as described above, strain-dependent differences have been identified. balb/c-into-cb6f1 resulted in chronic gvhd that was comparable to that of dba/2-into-b6d2f1 (26). the agvhd resulting from c57bl/6-into-cb6f1, characterized by donor cell expansion and antihost ctl activity, was similar to that in c57bl/6-into-b6d2f1 during the first 3 wk following transfer (26). after 3 wk, while c57bl/6 donor cells continued to expand in b6d2f1 hosts, the percentage of parental cells progressively decreased in cb6f1 hosts (26). by 12 wk, the c57bl/6-into-cb6f1 mice had transitioned to cgvhd, with an absence of anti-host ctl activity and the development of autoantibodies (26). in this study, we analyze the 2 wk following transfer of c57bl/6-into-cb6f1, a point at which only agvhd parameters are observed.

costimulatory signals have been demonstrated to be essential to the development of disease in a variety of gvhd models. blocking b7/cd28 costimulatory interactions with mabs to b7-1 and b7-2 (27, 28), ctla-4ig, a soluble competitive inhibitor of cd28 (29, 30, 31), or by using cd28c/c splenocytes (32) inhibited donor t cell activation and disease development. blockade of icos signaling strongly suppressed th2-driven cgvhd disease yet augmented donor t cell expansion in th1-driven agvhd (33). several studies have established that inhibitory receptors slow the development of gvhd. blocking abs against ctla-4 (34) or pd-1 (35) accelerated gvhd following bone marrow transplant. ctla-4 and pd-1 do not appear fully redundant in their regulation of gvhd because an additive increase in disease severity occurs when both pathways are inhibited (35).

in this study, we use the nonirradiated parental-into-f1 model of gvhd to examine the function of btla in regulating alloresponses. we anticipated that btlac/c parental cells would have augmented alloreactivity, similar to that seen in the absence of inhibitory signals such as ctla-4 or pd-1. unexpectedly, we observed that btlac/c donor splenocytes are unable to sustain this gvhd response. although btlac/c lymphocytes react to host alloantigens and expand in an early effector response, this response is not sustained. in comparison to btla+/+ donor cells, btlac/c donor t cells undergo a rapid contraction in vivo beginning on day 10, which is accompanied by resolution of gvhd pathology. dna microarray analysis indicates that cd4+ btlac/c donor cells have altered expression of several important genes that may influence the effector response or survival of t cells, including il-7r (36, 37, 38), il-18r (39), and il-4 (reviewed in ref. 40). these results suggest that btla may act not only as an inhibitor as shown previously (1, 2, 4, 5) but, under some conditions, may participate in providing signals that promote the immune responses.

materials and methods

reagents

the following abs used for facs analysis were from bd pharmingen: cd4-pe-cy7 (rm4-5), cd8-pe or allophycocyanin (53-6.7), b220-allophycocyanin (ra3-6b2), h-2kd-fitc or pe (sf1-1.1), h-2kb-fitc or pe (af6-88.5), il-15r-pe (tm-1), il-6r-pe (d7715a7), ifn-r-biotin (gr20), brdu-fitc, annexin v-allophycocyanin, and streptavidin-allophycocyanin. il-7r-fitc (a7r34) and btla-pe (6f7, pan-allele specific) were obtained from ebioscience. il-18r-biotin (polyclonal) and goat igg-biotin were obtained from r&d systems. four-color analysis was performed on a facscalibur (bd biosciences) and analyzed using flowjo (tree star). for in vivo blockade, endotoxin-free anti-btla (6a6, c57bl/6-allele specific) was purified from hybridoma supernatants according to standard procedures (2). isotype control hamster igg was obtained from jackson immunoresearch laboratories.

mice

c57bl/6, balb/c, and c57bl/6 x balb/c f1 (cb6f1) mice were bred in our facility. btlac/c mice were backcrossed to c57bl/6 or balb/c for at least nine generations (1). in some experiments, btla+/c x btla+/c (generation 9) littermate mice were used as donors, and the btla genotypes of mice were determined by surface staining of b220+ peripheral blood cells with 6f7-pe (anti-btla). all donor and host mice were 8c12 wk of age and sex matched. syngeneic cb6f1 mice were used as control donors.

induction of gvhd

unless otherwise stated, single-cell suspensions of pooled donor splenocytes were subjected to rbc lysis and suspended in sterile, endotoxin-free hbss. agvhd was induced as previously described in ref. 33 , except that 5 x 107 c57bl/6 donor cells (which we found to yield similar results as the reported 6 x 107 cells) were injected via the lateral tail vein into normal, nonirradiated cb6f1 host mice. control mice received donor cells from syngeneic cb6f1 mice. to induce cgvhd, 8 x 107 pooled balb/c donor splenocyte were transferred into cb6f1 mice. at indicated time points, splenocytes from each gvhd-induced mouse were analyzed separately.

elisa for anti-dna abs

serum autoantibodies specific for ssdna and dsdna were detected by elisa. after precoating with 0.01% poly-l-lysine, elisa plates were coated with 3 g/ml single-stranded calf thymus dna (sigma-aldrich) or 2.5 g/ml double-stranded calf thymus dna (sigma-aldrich) in pbs. nonspecific binding was blocked by incubating with 10% fcs. serial dilutions of each experimental serum were plated and developed with anti-mouse-igg-hrp (southern biotechnology associates). serum from an mrl/lpr mouse was used as a positive control for autoantibodies. the od reading for mrl/lpr serum was set to an arbitrary unit of 1, and the level of autoantibody in serum from experimental mice was expressed relative to this value.

in vivo treatment with anti-btla mab

experimental mice were injected i.p. with 100 g of purified 6a6 (monoclonal anti-bl/6 btla) (2) or control hamster-igg (jackson immunoresearch laboratories) at the time of gvhd induction (day 0) or again on day 3 and 6 following transfer, as indicated.

assessment of proliferation by cfse dilution and brdu incorporation

to assess cell division during the initiation of gvhd, donor cells were labeled with 1 m cfse (molecular probes) before transfer, as in ref. 5 . to assess proliferation at later stages of the response, in vivo 5-brdu (bd pharmingen) labeling and facs staining were conducted according to the manufacturer?s suggested protocol (brdu flow kit; bd pharmingen). briefly, 18 h before harvest, mice were injected i.p. with 1 mg of brdu in pbs. brdu incorporation was assessed by facs staining of fixed and permeabilized cells with anti-brdu-fitc.

cfse-based cytotoxicity assay

this facs-based ctl assay was conducted as reported by jedema et al. (41), with alterations noted below. briefly, allogeneic p815 (h-2kd) or syngeneic el4 (h-2kb) target cells were labeled with 1 m cfse. target cells were suspended in medium at 1 x 105 cells/ml and 100 l/well plated in round-bottom 96-well plates. whole splenocytes from gvhd-induced mice, subjected only to rbc lysis, served as effector cells and were plated at the indicated e:t ratios. plates were incubated for 4 h at 37?c. wells were harvested, and 7-aminoactinomycin (7-aad) was added just before facs analysis. lysed targets were determined by the percentage of cfse+ cells that were 7-aad+ (molecular probes). nonspecific lysis was assessed in wells without effector cells added. splenocytes from naive cb6f1 served as a syngeneic control.

microarray analysis

on days 9 and 10 following gvhd induction, the cd4+ donor-derived population was sorted from pooled splenocytes of several experimental mice that had received btla+/+ or btlac/c and stained with anti-h-2kd-fitc, anti-btla-pe, and anti-cd4-allophycocyanin. postsort analysis showed >95% purity of the sorted populations. rna was extracted with the rneasy kit (qiagen). biotinylated crna target was generated using the two-cycle cdna synthesis kit (affymetrix). each crna was hybridized to the affymetrix mouse genome 430 2.0 array. data were analyzed using dchip ().

results

btla is required for maintenance of donor lymphocytes in the parental-into-f1 gvhd model

to investigate the functions of btla during an in vivo allogeneic immune response stimulated by mhc mismatch, we compared the actions of btla+/+ and btlac/c cells in a model of parental-into-f1-induced agvhd. transfer of c57bl/6 (h-2kb) btla+/+ splenocytes into unirradiated cb6f1 (h-2b/d) hosts elicited a progressive expansion of donor cd4+ and cd8+ lymphocytes characteristic of agvhd, as expected (fig. 1, a and b). however, transfer of btlac/c cells was unable to sustain donor cell expansion over the course of a 14-day response (fig. 1, a and b). to characterize the kinetics of this response, the absolute number (fig. 1a) and percentage (fig. 1b) of donor t cells in the spleens of gvhd-induced mice were assessed by facs on days 3, 7, 9, 11, and 14 after transfer. mice receiving either btla+/+ and btlac/c donor cells showed expansion of donor t cells during the initial 9 days following gvhd induction (fig. 1, a and b). however, btlac/c donor cells began to decline in numbers after day 10, in contrast to the progressive increase in btla+/+ donor cd4+ and cd8+ t cells (fig. 1, a and b). to investigate whether btla expressed by host cells was important in controlling gvhd, donor cells were also transferred into btlac/c cb6f1 hosts. again, btla+/+ t cells continually expanded in btlac/c recipients, whereas btlac/c t cells showed a rapid contraction by days 11 and 14, suggesting that btla expressed by the host is not required for maintenance of gvhd by donor cells (data not shown).

figure 1. c57bl/6 btlac/c donor t lymphocytes fail to persist long term in cb6f1 hosts. a total of 5 x 107 donor splenocytes from c57bl/6 btla+/+ (h-2b, ) or c57bl/6 btlac/c (h-2b, ) mice 5 were transferred into cb6f1 (h-2b/d) hosts. spleens were harvested at days 3, 7, 9, 11, and 14 after transfer and analyzed by facs to establish the kinetics of donor-host chimerism. donor-derived cells were distinguished by the absence of h-2kd. a, the absolute number of live donor cd4+ (top panel) and cd8+ (bottom panel) splenocytes were calculated at each time point. b, the percentage of splenic cd4+ (top panel) or cd8+ (bottom panel) compartments that were donor derived is shown at each time point. c, btlac/c donor cells fail to elicit a strong antihost response, as demonstrated by maintenance of the host b cell compartment. the absolute number of live b220+ host splenocytes (top panel) and the total percentage of b220+ splenocytes (bottom panel) are shown at each time point. average values were calculated from three mice per group at each time point with error bars indicating sem. asterisks indicate time points at which student?s t tests indicate significant differences of p < 0.05 between btla+/+ and btlac/c donors. data are representative of five independent experiments.

the inability of btlac/c donor lymphocytes to sustain agvhd was also demonstrated by the absence of an antihost cytotoxic response, apparent from the lack of depletion of the host b220+ population (fig. 1c). on day 14, spleens from mice receiving btla+/+ cells were <10% b220+, indicating cytolysis of host b cells, as expected. in contrast, mice receiving btlac/c cells retained a normal number of b cells, 60% of splenocytes.

to determine whether btlac/c donor lymphocytes were engrafting and expanding, we transferred and analyzed cfse-labeled donor cells (fig. 2). cd4+ and cd8+ t cells from btla+/+ and btlac/c donors showed similar percentages and numbers of cells that had undergone division after 3 days (fig. 2). these results indicate that later disappearance of btlac/c donor cells is not due to a failure in initial activation by host alloantigens. we also used btla+/+, btla+/c, and btlac/c littermates generated from an intercross of c57bl/6 btla+/c mice as donor cells to test for any minor histocompatibility mismatch with cb6f1 host mice. again, only the btlac/c donor cells failed to survive at day 14, whereas both btla+/+ and btla+/c donor cells expanded to occupy 60c70% of the cd4+ and cd8+ compartments (fig. 3), suggesting that the findings were due to the absence of donor cell btla expression rather than a result of transplant incompatibility.

figure 2. btlac/c donor cells engraft and divide during the initiation stage of gvhd. a total of 5 x 107 cfse-labeled cb6f1, c57bl/6 btla+/+, or btlac/c splenocytes were transferred to cb6f1 host mice. splenocytes were harvested 3 days posttransfer and gated on live, cfse+ donor lymphocytes by facs. a, donor cells were divided into cd4+ and cd8+ populations (first column), and the percentage of the donor cd4+ (second column) and cd8+ (third column) populations that had undergone one or more divisions was assessed by cfse dilution. histogram gates indicate the percentage of each population undergoing one or more divisions. b, following the gating scheme above, the absolute number of donor cells that had undergone one or more divisions was calculated. bar graphs show an average of three mice, with error bars indicating sem (btla+/+ donors, ; btlac/c donors, ). student?s t tests did not show significant differences between btla+/+ and btlac/c donor cells (cd4+, p = 0.143; cd8+, p = 0.084). data are representative of two independent experiments.

figure 3. failed survival of btlac/c donor cells cannot be attributed to a minor histocompatibility mismatch. btla+/+, btla+/c, or btlac/c littermates from a btla+/c x btla+/c breeding (generation 9 backcross to c57bl/6) were used as donor splenocytes. donor mice were tail bled and btla genotyped confirmed by facs staining of b220+ cells before transfer (data not shown). donor splenocytes (5 x 107) from a single mouse were transferred into a cb6f1 host and harvested on day 14. the percentage of donor-derived cells (h-2kd negative) in the cd4+ (second column) and cd8+ (third column) populations are indicated by the left gate on each histogram. the h-2kd-positive population (right gates) indicates host cells.

we also evaluated a model of cgvhd characterized by gradual development of a th2 cytokine-induced, autoantibody-mediated disease (24, 25). transfer of balb/c btla+/+ donor splenocytes into cb6f1 hosts induced expansion of cd4+ and cd8+ donor t cells (although to a lower extent than seen in agvhd, as expected (26)), the development of anti-dsdna and anti-ssdna abs and activation of host-derived b cells as measured by cd69 induction (fig. 4), as expected for balb/c donors (25, 42). kinetic analysis revealed that balb/c btlac/c donor t lymphocytes expand equivalently to btla+/+ cells during the initial 10 days following transfer (fig. 4a). however, similar to the transfer of c57bl/6 donor cells, balb/c btlac/c donor cd4+ and cd8+ cells rapidly contract by day 12 of the response (fig. 4a). transfer of balb/c btlac/c donor splenocytes also failed to show the induction of anti-dna autoantibodies characteristic of cgvhd (fig. 4b). additionally, host b cells from mice receiving btlac/c donor cells were not activated, as measured by a failure to up-regulate cd69 (fig. 4c). thus, balb/c btlac/c donor cells show a similar defect in sustaining cgvhd as was found for c57bl/6 btlac/c donor cells in sustaining agvhd.

figure 4. transfer of balb/c btlac/c donor cells fails to elicit a chronic gvhd response. chronic gvhd was induced by adoptively transferring 8 x 107 balb/c btla+/+ or btlac/c donor splenocytes into cb6f1 hosts. a, spleens from mice receiving btla+/+ (, h-2hd) or btlac/c (, h-2kd) donor cells were harvested at days 5, 8, 10, 12, and 21 after transfer and analyzed by facs as in fig. 1, except that donor cells were h-2kb negative. the absolute number of live donor cd4+ (left panel) and cd8+ (right panel) splenocytes was calculated at each time point. average values were calculated from three mice per group at each time point with error bars indicating sem. asterisks indicate time points at which student?s t tests indicate significant differences of p < 0.05 between btla+/+ and btlac/c donor cells. b, serum was obtained from gvhd-induced mice at the indicated time points and assessed by elisa for development of anti-dsdna (left graph) or anti-ssdna (right graph) abs. naive mrl/lpr mouse serum () was used as a positive control for autoantibodies, whereas naive cb6f1 serum () was used as a negative control. the amount of autoantibody in mrl/lpr serum was set to an arbitrary value of 1, and the amount of anti-dna ab in serum from mice receiving btla+/+ cells () or btlac/c cells () was calculated relative to mrl/lpr serum. the average of four mice per group is shown, with error bars indicating sem. asterisks indicate points at which student?s t tests indicate significant differences of p < 0.05. c, transfer of btlac/c donor cells fails to induce activation of host b cells. splenocytes were harvested on day 11 after gvhd induction. the expression level of the activation marker cd69 on host b cells (b220+, h-2kb positive) is shown. data are representative of two independent experiments.

btla blockade by ab treatment prevents maintenance of btla+/+ donor lymphocyte in agvhd

the 6a6 mab is specific for the c57bl/6 allele of btla and blocks binding to its ligand hvem (2). administration of 6a6 at the time of transfer of btla+/+ donor cells resulted in a loss of donor lymphocytes by day 14 of the response, similar to the transfer of btlac/c cells (fig. 5). to ensure that the abrogation of donor cell expansion was not due to opsonization and rapid host clearance of btla-expressing donor cells by the ab, we examined the kinetics the gvh response in the presence of 6a6 administration. (fig. 5). expansion of c57bl/6 btla+/+ donor cells in recipients treated with isotype control or anti-btla ab was equivalent at days 3 and 8 following transfer (fig. 5). by day 14, few donor cd4+ and cd8+ lymphocytes were present in mice that had received anti-btla abs (fig. 5). a single ab treatment at the time of transfer was sufficient to cause the loss of donor cells, as additional ab administration at days 3 and 6 had no additional effect (data not shown). the observation that blocking btla/hvem interactions through ab treatment yields the loss of donor cells similar to transfer of btlac/c splenocytes suggests that the interaction of btla with its ligand hvem is necessary to support the survival of donor lymphocytes in agvhd.

figure 5. anti-btla ab treatment at the initiation of gvhd inhibits the long-term survival of btla+/+ donor lymphocytes. c57bl/6 btla+/+ splenocytes (5 x 107) were transferred into cb6f1 hosts. a total of 100 g of hamster igg (isotype control, ) or anti-btla ab (6a6, ) was administered i.p. to host mice concurrent with donor cell transfer (day 0). splenocytes were harvested after 3, 8, and 14 days and analyzed for donor/host chimerism as in fig. 1. absolute numbers of donor cd4+ (left graph) and cd8+ (right graph) were calculated at each time point. average values are shown for three mice per group at each time point with error bars indicating sem. asterisks indicate points at which student?s t tests indicate significant differences of p < 0.05 between isotype- and 6a6-treated mice. data are representative of three independent experiments.

btlac/c donor cells demonstrate reduced brdu incorporation and antihost ctl activity before contraction

contraction of btlac/c donor cells could be a result either of decreased proliferation or increased cell death. therefore, we examined these parameters at days 9 and 10, a time just before the onset of rapid contraction, using brdu labeling, markers of apoptosis, and assays for cytolytic activity (fig. 6). at day 9, the percentage of btlac/c donor t cells that had incorporated brdu is approximately two-thirds of the percentage of brdu+btla+/+ donors (fig. 6a). on day 10, brdu incorporation by btlac/c donor cells is further reduced to approximately one-third the level of btla+/+ donors (fig. 6a). btlac/c donor cd4+ and cd8+ t cells showed a similar degree of reduced brdu incorporation. however, we did not observe a detectable difference in the degree of annexin v staining, a marker of apoptotic cells, of cd4+ or cd8+ btlac/c and btla+/+ donor cells at day 9 (fig. 6b). analysis of cytolytic activity at day 10 showed that btlac/c donor cells had a greatly reduced capacity to kill allogeneic p815 (h-2kd) cells in vitro compared with btla+/+ donor cells (fig. 6c). this result agrees with our finding of normal levels of host b220+ cells persisting in mice receiving btlac/c donor cells (fig. 1c). overall, these results suggest that the failure of btlac/c donor cells to sustain gvhd is due to a progressive decrease in donor cell proliferation, survival, and insufficient antihost ctl activity.

figure 6. btlac/c donor cells incorporate less brdu but do not exhibit increased apoptosis before contraction. gvhd was induced by transferring 5 x 107 btla+/+ c57bl/6 () or btlac/c () splenocytes into cb6f1 hosts. a, splenocytes were harvested at days 9, 10, and 11 following transfer. eighteen hours before harvest, mice were injected i.p. with 1 mg of brdu to observe cellular proliferation. splenocytes were subjected to facs staining to identify the percentage of donor cd4+ (left graph) and cd8+ (right graph) cells incorporating brdu. data at each time point is representative of the average of two mice per group. b, the percentage of annexin v-positive donor cd4+ (left panels) or cd8+ (right panels) splenocytes from mice receiving btla+/+ (upper panels) or btlac/c (lower panels) splenocytes was assessed at day 9 by facs. gates indicate the percentage of annexin v-positive cells. data are representative of two mice per group. c, at day 10, unfractionated splenocytes from gvhd-induced mice were analyzed for ctl activity against allogeneic (h-2kd) p815 cells. effector splenocytes were incubated with cfse-labeled targets at the indicated ratios for 4 h. a vitality dye (7-aad) was added, and the percentage of target cells lysed (cfse+7-aad+) was determined by facs. ctl activity was not observed toward syngeneic (h-2kb) el4 cell targets (data not shown). the mean ? sem of three mice per group is shown.

differential gene expression of btla+/+ and btlac/c donor cd4+ cells during gvhd

we performed dna microarray studies to characterize the potential molecular mechanism accounting for the contraction of btlac/c donor populations during gvhd. gvhd was induced by transferring either c57bl/6 btla+/+ or btlac/c donor cells into cb6f1 recipients, and donor cd4+ lymphocytes were harvested 9 and 10 days after transfer. we selected these times for analysis because they directly precede the decline in btlac/c donor cells and may reveal cellular changes leading to their loss. consistent with comparable annexin v staining on btla+/+ and btlac/c donor cells (fig. 6b), no differential expression of either pro- or antiapoptotic genes (including bcl-2, bcl-xl, and bad) was seen by dna microarray (tables i and ii). we identified several immune specific genes that were differentially expressed by btla+/+ and btlac/c donor cd4+ populations at day 9 (table i) and day 10 (table ii). these included il-7r, il-18r, il-6r, and ifn-r1 (fig. 7).

table i. global gene analysis reveals genes induced and inhibited in btla+/+ and btlac/ccd4+ donor cells 9 days after acute gvhd inductiona

table ii. global gene analysis reveals genes induced and inhibited in btla+/+ and btlac/c cd4+ donor cells 10 days after acute gvhd inductiona

figure 7. differential gene expression of btla+/+ and btlac/c donor cd4+ cells during gvhd. c57bl/6 btla+/+ splenocytes (5 x 107) were transferred into cb6f1 hosts. splenocytes were harvested on days 9 and 10. the surface levels of various receptors are compared with btla+/+ (solid lines) and btlac/c (dashed lines) cd4+ donor cells. isotype control staining of a mixture of btla+/+ and btlac/c splenocytes is shown (shaded regions). the first column indicates the receptor level on naive btla+/+ c57bl/6 cells. expression levels of all receptors are similar between btla+/+ and btlac/c cells at day 7. il-7r, il-18r, and ifn-r1 levels are increased on btla+/+ compared with btlac/c donor cells at days 9 and 11. il-6r expression is higher on btlac/c donor cells at day 11, whereas il-15r levels remain equivalent between btla+/+ and btlac/c cells throughout. expression of il-15r, but not il-7r, was lower on cd8+btlac/c donor cells than btla+/+ donor cells at day 9 (right column).

il-7r (cd127) showed 3.2- and 1.8-fold higher expression in btla+/+ compared with btlac/ccd4+ t cells on days 9 and 10, respectively (table i). on day 10, t1/st2, an il-1r family member (43), was expressed 11.1-fold higher in btla+/+cd4+ donor cells compared with btlac/c cells (table ii). il-18r showed a 1.7-fold higher expression at day 9 and a 7.3-fold higher expression at day 10 in btla+/+ donor cells. ifn-r1 (cd119) expression was 3.3-fold higher in the btla+/+ donor cd4+ population than the btlac/c donors at day 10 (table ii). some genes were expressed more highly by btlac/ccd4+ donor cells compared with btla+/+cd4+ donor cells. il-4 expression was 3.4-fold higher in btlac/c cells at day 9 (table i), which increased to a 10.4-fold higher expression on day 10 (table ii). il-6r had 1.6-fold higher expression in btlac/c cells at day 9 and 4-fold higher expression at day 10 (table i).

we verified some of these differences by facs analysis (fig. 7). il-7r surface expression on btla+/+cd4+ donor cells increased progressively over time, ultimately returning to levels similar to that of naive cd4+ cells, whereas btlac/c donor cells failed to re-express the receptor (fig. 7). in contrast, the expression of il-7r on cd8+btlac/c donor cells was equivalent to that of btla+/+ donors at day 9. although the il-15r-chain (cd122) remained equivalently expressed by both btla+/+ and btlac/ccd4+ donor cells populations at all time points, cd8+btlac/c donor cells expressed decreased levels of il-15r at day 9 (fig. 7). the il-18r was expressed equivalently at day 7 but later increased on btla+/+ donor cd4+ cells and decreased on btlac/c donor cells (fig. 7). ifn-r1 expression remained unchanged on btlac/ccd4+ donor cells but was increased on btla+/+cd4+ donor cells at days 9 and 11 (fig. 7). il-6r was expressed equivalently between btla+/+ and btlac/c cells at day 7 but was up-regulated on only btlac/c donors at day 11 (fig. 7).

discussion

the major finding of this study is the unexpected stimulatory, rather than inhibitory, role of btla in an in vivo model of chronic allostimulation. in the parental-into-f1 nonirradiated gvhd model, we found that btlac/c donor cells were unable to sustain gvhd pathology, as opposed to the expected augmentation of disease. these data suggest that btla, acting either as a receptor or as a ligand for hvem, has prosurvival activity in this in vivo model of sustained alloresponse. although the initial activation and expansion of btlac/c and btla+/+ donor cells was similar, btlac/c donor cell numbers began to decrease at day 10, eventually disappearing (fig. 1). thus, under this model of chronic allostimulation, btla is acting in a prosurvival manner and is clearly necessary for the maintenance, rather than the initiation, of the effector t cell response.

btla was initially identified as having inhibitory actions on t cell proliferation and cytokine production in vitro (1, 44). inhibitory actions observed in vivo in models of experimental allergic encephalomyelitis (1) and allergic airway inflammation (45) support these observations. engagement of btla on murine cd4+ t cells by the btla ligand hvem expressed on apcs (5) or by a hvem-fc fusion protein (13) also caused a reduction in early t cell proliferation. a chimeric receptor containing the extracellular domain of cd28 fused to the human btla cytoplasmic tail was able to potently inhibit il-2 production by primary human cd4 t cells, again indicating inhibitory actions for btla (7). finally, a recent report describing increased homeostatic expansion of btlac/c t cells in lymphopenic hosts and increased numbers of cd8+ central memory cells in btlac/c and hvemc/c mice further demonstrates the inhibitory activity of btla (46).

however, not all reported actions of btla have been inhibitory. two different systems of cardiac transplantation (47) were used to evaluate the role of btla in allograft rejection (48). in the first system, partially mhc mismatched allografts, differing at a single class ii mhc locus, are tolerated for long periods by wild-type recipient mice. these allografts were rapidly rejected, rather than tolerated, by btlac/c or hvemc/c recipients, consistent with an inhibitory action of btla in the recipient immune response (48). in the second system, fully mhc mismatched, with differences at both class i and class ii mhc loci, cardiac allografts are rapidly rejected by wild-type mice. these allografts showed slightly prolonged survival in btlac/c recipients (48). although the basis for this difference was not established, these results suggested a positive action of btla in the recipient immune response.

in addition to itim and immunoreceptor tyrosine-based switch motif inhibitory motifs (1), the cytoplasmic domains of human and mouse btla contain other conserved tyrosine-containing motifs that may be relevant to signaling, such as sites that may recruit grb-2 (7, 8). in one study, mutation of all cytoplasmic tyrosine residues was required to abolish the inhibitory actions of a chimeric signaling protein containing the human btla cytoplasmic domain (7). another study indicated that the one grb-2 recruitment motif in murine btla may interact with both grb-2 and the p85 subunit of pi3k (8). while indirect evidence, a peptide containing the conserved phosphorylated grb-2 recruitment region of the btla cytoplasmic domain had the ability to interact with pi3k, suggesting that btla may sometimes have a positive, rather than inhibitory, action (8). however, neither of these studies examined the in vivo role of these conserved signaling motifs.

to analyze the role of btla in immune regulation in vivo, we have surveyed various models of pathogen infections, induced and spontaneous autoimmunity, and various transplantation systems. as part of this survey, we examined the nonirradiated parental-into-f1 gvhd system (18), expecting to observe augmented immune responses against host cells in the absence of btla?s inhibitory actions. this nonirradiated gvhd model is distinct from other models of gvhd models that use host irradiation and bone marrow transplantation to more closely mimic gvhd that arises in clinical settings using myeloablative conditioning (reviewed in ref. 17). indeed, each model system contributes unique factors that complicate the analysis of the allogeneic response. irradiation causes epithelial damage and extensive inflammatory cytokine production that exacerbate gvhd (16), whereas the nonirradiated model is complicated by an intact host immune system that may respond to inflammatory stimuli and contribute an antigraft response (49). consequently, different mechanisms may account for the pathology in the irradiated and nonirradiated systems, and distinct roles for perforin (50, 51) and ifn- (52, 53) have been revealed in each model.

global gene expression analysis of btla+/+ and btlac/c donor cells harvested from hosts on days 9 and 10 revealed several genes with differential expression (table i). many of the differentially expressed genes are known to function during allogeneic responses. polarization of donor cd4+ t cells toward the th1 phenotype is important for development of agvhd (54, 55, 56). interestingly, genes associated with th2 cells were increased in btlac/c cd4+ donors cells, whereas genes associated with th1 cells were decreased in btlac/c cd4+ donors. for example, il-4 was increased in btlac/c cd4+ donors (table i) and is th2 specific (40), whereas il-18r and cxcr6 were decreased in btlac/c cd4+ donors cells and are th1 specific (39, 57, 58). thus, the inability of btlac/c donor cells to mount a sustained effector response to allogeneic stimulation may be due to altered cd4+ t cell polarization.

we also found that il-7r failed to become highly expressed on btlac/c cells to the same extent as btla+/+ donor cd4+ cells. il-7 is critical for homeostatic survival of naive t cells (59) and regulates the transition of cd4+ effector t cells to long-lived and memory t cells (36, 37, 38). it was also recently reported that persistence of gvhd in an irradiated model system involved the development of alloreactive cd4+ or cd8+ memory t cells whose survival required the cytokines il-2, il-7, and il-15 despite chronic allostimulation (60, 61). conceivably, failure of btlac/c cd4+ cells to re-express il-7r in the nonirradiated gvhd model could prevent necessary survival signaling such as bcl-2 (62) or pi3k (63), causing the contraction observed after day 10 in vivo.

hvem activity was shown recently to be required for manifestation of disease in a similar parental-into-f1 nonirradiated model of gvhd (64). hvemc/c and lightc/c donor t cells showed decreased survival and reduced antihost ctl activity compared with wild type donor cells 7c10 days following transfer (64). previously, blockade of light using soluble ltr-ig or anti-light ab was found to reduced severity of gvhd in this system (12). notably, cotransfer of wt and lightc/c t cells showed that wild-type donor cells had a significant survival advantage after 11 days (64).

that study (64), taken together with the results reported here, indicate that hvem, light, and btla all have similar actions on promoting donor cell survival in the parental-into-f1 model of gvhd. we have independently confirmed that hvemc/c donor cells show reduced survival in this model (data not shown), as reported previously (64). if hvem signaling, activated by light or lt, delivers a prosurvival signal required for donor cell maintenance, then our results suggest that btla may be playing some role in augmenting this signaling pathway. thus, sustaining a chronic alloresponse may require a complex involving btla, light, and hvem. in the absence of btla, the light-hvem complex may either induce an apoptotic signal or be insufficient to maintain cell expansion and survival. in summary, our data identify a condition in which btla unexpectedly promotes, rather than inhibits, an immune response. we do not know whether this action is mediated by btla signaling or by btla acting as a ligand for hvem, with hvem mediating the prosurvival signals. distinguishing between these and other possibilities will likely require extensive additional analysis and involve mutations of both btla and hvem.

disclosures

the authors have no financial conflict of interest.

【参考文献】
watanabe, n., m. gavrieli, j. r. sedy, j. yang, f. fallarino, s. k. loftin, m. a. hurchla, n. zimmerman, j. sim, x. zang, et al 2003. btla is a lymphocyte inhibitory receptor with similarities to ctla-4 and pd-1. nat. immunol. 4: 670-679.

hurchla, m. a., j. r. sedy, m. gavrielli, c. g. drake, t. l. murphy, k. m. murphy. 2005. b and t lymphocyte attenuator exhibits structural and expression polymorphisms and is highly induced in anergic cd4+ t cells. j. immunol. 174: 3377-3385.

gavrieli, m., n. watanabe, s. k. loftin, t. l. murphy, k. m. murphy. 2003. characterization of phosphotyrosine binding motifs in the cytoplasmic domain of b and t lymphocyte attenuator required for association with protein tyrosine phosphatases shp-1 and shp-2. biochem. biophys. res. commun. 312: 1236-1243.

krieg, c., p. han, r. stone, o. d. goularte, j. kaye. 2005. functional analysis of b and t lymphocyte attenuator engagement on cd4+ and cd8+ t cells. j. immunol. 175: 6420-6427.

sedy, j. r., m. gavrieli, k. g. potter, m. a. hurchla, r. c. lindsley, k. hildner, s. scheu, k. pfeffer, c. f. ware, t. l. murphy, k. m. murphy. 2005. b and t lymphocyte attenuator regulates t cell activation through interaction with herpesvirus entry mediator. nat. immunol. 6: 90-98.

croft, m.. 2005. the evolving crosstalk between co-stimulatory and co-inhibitory receptors: hvem-btla. trends immunol. 26: 292-294.

chemnitz, j. m., a. r. lanfranco, i. braunstein, j. l. riley. 2006. b and t lymphocyte attenuator-mediated signal transduction provides a potent inhibitory signal to primary human cd4 t cells that can be initiated by multiple phosphotyrosine motifs. j. immunol. 176: 6603-6614.

gavrieli, m., k. m. murphy. 2006. association of grb-2 and pi3k p85 with phosphotyrosile peptides derived from btla. biochem. biophys. res. commun. 345: 1440-1445.

mauri, d. n., r. ebner, r. i. montgomery, k. d. kochel, t. c. cheung, g. l. yu, s. ruben, m. murphy, r. j. eisenberg, g. h. cohen, et al 1998. light, a new member of the tnf superfamily, and lymphotoxin are ligands for herpesvirus entry mediator. immunity 8: 21-30.

scheu, s., j. alferink, t. potzel, w. barchet, u. kalinke, k. pfeffer. 2002. targeted disruption of light causes defects in costimulatory t cell activation and reveals cooperation with lymphotoxin in mesenteric lymph node genesis. j. exp. med. 195: 1613-1624.

tamada, k., k. shimozaki, a. i. chapoval, y. zhai, j. su, s. f. chen, s. l. hsieh, s. nagata, j. ni, l. chen. 2000. light, a tnf-like molecule, costimulates t cell proliferation and is required for dendritic cell-mediated allogeneic t cell response. j. immunol. 164: 4105-4110.

tamada, k., k. shimozaki, a. i. chapoval, g. zhu, g. sica, d. flies, t. boone, h. hsu, y. x. fu, s. nagata, et al 2000. modulation of t cell-mediated immunity in tumor and graft-versus-host disease models through the light co-stimulatory pathway. nat. med. 6: 283-289.

gonzalez, l. c., k. m. loyet, j. calemine-fenaux, v. chauhan, b. wranik, w. ouyang, d. l. eaton. 2005. a coreceptor interaction between the cd28 and tnf receptor family members b and t lymphocyte attenuator and herpesvirus entry mediator. proc. natl. acad. sci. usa 102: 1116-1121.

compaan, d. m., l. c. gonzalez, i. tom, k. m. loyet, d. eaton, s. g. hymowitz. 2005. attenuating lymphocyte activity: the crystal structure of the btla-hvem complex. j. biol. chem. 280: 39553-39561.

cheung, t. c., i. r. humphreys, k. g. potter, p. s. norris, h. m. shumway, b. r. tran, g. patterson, r. jean-jacques, m. yoon, p. g. spear, et al 2005. evolutionarily divergent herpesviruses modulate t cell activation by targeting the herpesvirus entry mediator cosignaling pathway. proc. natl. acad. sci. usa 102: 13218-13223.

reddy, p., j. l. ferrara. 2003. immunobiology of acute graft-versus-host disease. blood rev. 17: 187-194.

welniak, l. a., b. r. blazar, and w. j. murphy. immunobiology of allogeneic hematopoietic stem cell transplantation. annu. rev. immunol. in press.

shearer, g. m., r. p. polisson. 1980. mutual recognition of parental and f1 lymphocytes: selective abrogation of cytotoxic potential of f1 lymphocytes by parental lymphocytes. j. exp. med. 151: 20-31.

hakim, f. t., s. o. sharrow, s. payne, g. m. shearer. 1991. repopulation of host lymphohematopoietic systems by donor cells during graft-versus-host reaction in unirradiated adult f1 mice injected with parental lymphocytes. j. immunol. 146: 2108-2115.

via, c. s.. 1991. kinetics of t cell activation in acute and chronic forms of murine graft-versus-host disease. j. immunol. 146: 2603-2609.

via, c. s., s. o. sharrow, g. m. shearer. 1987. role of cytotoxic t lymphocytes in the prevention of lupus-like disease occurring in a murine model of graft-vs-host disease. j. immunol. 139: 1840-1849.

ferrara, j. l.. 1993. cytokine dysregulation as a mechanism of graft versus host disease. curr. opin. immunol. 5: 794-799.

iwasaki, t., t. hamano, k. saheki, t. kuroiwa, y. kataoka, y. takemoto, a. ogata, j. fujimoto, e. kakishita. 2000. graft-versus-host-disease-associated donor cell engraftment in an f-1 hybrid model is dependent upon the fas pathway. immunology 99: 94-100.

gleichmann, e., e. h. van elven, j. p. van der veen. 1982. a systemic lupus erythematosus (sle)-like disease in mice induced by abnormal t-b cell cooperation: preferential formation of autoantibodies characteristic of sle. eur. j. immunol. 12: 152-159.

dewit, d., m. vanmechelen, c. zanin, j. m. doutrelepont, t. velu, c. gerard, d. abramowicz, j. p. scheerlinck, p. debaetselier, j. urbain, et al 1993. preferential activation of th2 cells in chronic graft-versus-host reaction. j. immunol. 150: 361-366.

tschetter, j. r., e. mozes, g. m. shearer. 2000. progression from acute to chronic disease in a murine parent-into-f1 model of graft-versus-host disease. j. immunol. 165: 5987-5994.

blazar, b. r., a. h. sharpe, p. a. taylor, a. panoskaltsismortari, g. s. gray, r. korngold, d. a. vallera. 1996. infusion of anti-b7.1 (cd80) and anti-b7.2 (cd86) monoclonal antibodies inhibits murine graft-versus-host disease lethality in part via direct effects on cd4+ and cd8+ t cells. j. immunol. 157: 3250-3259.

lang, t. j., p. nguyen, r. peach, w. c. gause, c. s. via. 2002. in vivo cd86 blockade inhibits cd4+ t cell activation, whereas cd80 blockade potentiates cd8+ t cell activation and ctl effector function. j. immunol. 168: 3786-3792.

blazar, b. r., p. a. taylor, p. s. linsley, d. a. vallera. 1994. in vivo blockade of cd28/ctla4: b7/bb1 interaction with ctla4-ig reduces lethal murine graft-versus-host disease across the major histocompatibility complex barrier in mice. blood 83: 3815-3825.

hakim, f. t., r. cepeda, g. s. gray, c. h. june, r. abe. 1995. acute graft-versus-host reaction can be aborted by blockade of costimulatory molecules. j. immunol. 155: 1757-1766.

via, c. s., v. rus, p. nguyen, p. linsley, w. c. gause. 1996. differential effect of ctla4ig on murine graft-versus-host disease (gvhd) development?ctla4ig prevents both acute and chronic gvhd development but reverses only chronic gvhd. j. immunol. 157: 4258-4267.

yu, x. z., p. j. martin, c. anasetti. 1998. role of cd28 in acute graft-versus-host disease. blood 92: 2963-2970.

ogawa, s., g. nagamatsu, m. watanabe, s. watanabe, t. hayashi, s. horita, k. nitta, h. nihei, k. tezuka, r. abe. 2001. opposing effects of anti-activation-inducible lymphocyte-immunomodulatory molecule/inducible costimulator antibody on the development of acute versus chronic graft-versus-host disease. j. immunol. 167: 5741-5748.

blazar, b. r., p. a. taylor, a. panoskaltsis-mortari, a. h. sharpe, d. a. vallera. 1999. opposing roles of cd28:b7 and ctla-4:b7 pathways in regulating in vivo alloresponses in murine recipients of mhc disparate t cells. j. immunol. 162: 6368-6377.

blazar, b. r., b. m. carreno, a. panoskaltsis-mortari, l. carter, y. iwai, h. yagita, h. nishimura, p. a. taylor. 2003. blockade of programmed death-1 engagement accelerates graft-versus-host disease lethality by an ifn--dependent mechanism. j. immunol. 171: 1272-1277.

li, j., g. huston, s. l. swain. 2003. il-7 promotes the transition of cd4 effectors to persistent memory cells. j. exp. med. 198: 1807-1815.

kondrack, r. m., j. harbertson, j. t. tan, m. e. mcbreen, c. d. surh, l. m. bradley. 2003. interleukin-7 regulates the survival and generation of memory cd4 cells. j. exp. med. 198: 1797-1806.

seddon, b., p. tomlinson, r. zamoyska. 2003. interleukin-7 and t cell receptor signals regulate homeostasis of cd4 memory cells. nat. 4: 680-686.

hoshino, k., h. tsutsui, t. kawai, k. takeda, k. nakanishi, y. takeda, s. akira. 1999. cutting edge: generation of il-18 receptor-deficient mice?evidence for il-1 receptor-related protein as an essential il-18 binding receptor. j. immunol. 162: 5041-5044.

murphy, k. m., s. l. reiner. 2002. the lineage decisions of helper t cells. nat. rev. immunol. 2: 933-944.

jedema, i., n. m. van der werff, r. m. barge, r. willemze, j. h. falkenburg. 2004. new cfse-based assay to determine susceptibility to lysis by cytotoxic t cells of leukemic precursor cells within a heterogeneous target cell population. blood 103: 2677-2682.

morris, s. c., r. l. cheek, p. l. cohen, r. a. eisenberg. 1990. autoantibodies in chronic graft versus host result from cognate t-b interactions. j. exp. med. 171: 503-517.

mitcham, j. l., p. parnet, t. p. bonnert, k. e. garka, m. j. gerhart, j. l. slack, m. a. gayle, s. k. dower, j. e. sims. 1996. t1/st2 signaling establishes it as a member of an expanding interleukin-1 receptor family. j. biol. chem. 271: 5777-5783.

han, p., o. d. goularte, k. rufner, b. wilkinson, j. kaye. 2004. an inhibitory ig superfamily protein expressed by lymphocytes and apcs is also an early marker of thymocyte positive selection. j. immunol. 172: 5931-5939.

deppong, c., t. i. juehne, m. hurchla, l. d. friend, d. d. shah, c. m. rose, t. l. bricker, l. p. shornick, e. c. crouch, t. l. murphy, et al 2006. cutting edge: b and t lymphocyte attenuator and programmed death receptor-1 inhibitory receptors are required for termination of acute allergic airway inflammation. j. immunol. 176: 3909-3913.

krieg, c., o. boyman, y. x. fu, j. kaye. 2007. b and t lymphocyte attenuator regulates cd8+ t cell-intrinsic homeostasis and memory cell generation. nat. immunol. 8: 162-171.

corry, r. j., h. j. winn, p. s. russell. 1973. primarily vascularized allografts of hearts in mice. the role of h-2d, h-2k, and non-h-2 antigens in rejection. transplantation 16: 343-350.

tao, r., l. wang, r. han, t. wang, q. ye, t. honjo, t. l. murphy, k. m. murphy, w. w. hancock. 2005. differential effects of b and t lymphocyte attenuator and programmed death-1 on acceptance of partially versus fully mhc-mismatched cardiac allografts. j. immunol. 175: 5774-5782.

gleichmann, e., h. gleichmann. 1985. pathogenesis of graft-versus-host reactions (gvhr) and gvh-like diseases. j. invest. dermatol. 85: 115s-120s.

shustov, a., i. luzina, p. nguyen, j. c. papadimitriou, b. handwerger, k. b. elkon, c. s. via. 2000. role of perforin in controlling b cell hyperactivity and humoral autoimmunity. j. clin. invest. 106: r39-r47.

baker, m. b., r. l. riley, e. r. podack, r. b. levy. 1997. graft-versus-host-disease-associated lymphoid hypoplasia and b cell dysfunction is dependent upon donor t cell-mediated fas-ligand function, but not perforin function. proc. natl. acad. sci. usa 94: 1366-1371.

ellison, c. a., j. m. fischer, k. t. hayglass, j. g. gartner. 1998. murine graft-versus-host disease in an f1-hybrid model using ifn- gene knockout donors. j. immunol. 161: 631-640.