Repertoire Analysis of CD8+ T Cell Responses to Minor Histocompatibility Antigens Involved in Graft-Versus-Host Disease
This information is current as Thea M. Friedman, Michael Gilbert, Constance Briggs and Robert of October 2, 2021. Korngold J Immunol 1998; 161:41-48; ; http://www.jimmunol.org/content/161/1/41 Downloaded from References This article cites 36 articles, 15 of which you can access for free at: http://www.jimmunol.org/content/161/1/41.full#ref-list-1
Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision
• No Triage! Every submission reviewed by practicing scientists
• Fast Publication! 4 weeks from acceptance to publication
*average by guest on October 2, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Repertoire Analysis of CD8؉ T Cell Responses to Minor Histocompatibility Antigens Involved in Graft-Versus-Host Disease1
Thea M. Friedman, Michael Gilbert, Constance Briggs, and Robert Korngold2
Graft-vs-host disease (GVHD) is a major complication of allogeneic bone marrow transplantation. Experimentally, lethal GVHD can be induced in MHC-matched strain combinations differing in expression of multiple minor histocompatibility Ags (miHA). Recently, the GVHD potential of C57BL/6By (B6) T cells in irradiated BALB.B (both H2b) and related CXB recombinant inbred strains of mice has been studied to determine the scope of the response to miHA in vivo and how it compared with CTL responses to immunodominant miHA in vitro. The GVHD response in these strain combinations appeared to be limited to a few Ags, yet there was no correlation of these miHA with that of in vitro CTL responses. To further investigate the role of CD8؉ T cells in Downloaded from GVHD, we analyzed positively selected miHA-specific donor CD8؉ thoracic duct lymphocytes (TDL) collected from irradiated BALB.B and CXBE mice, 5 to 6 days after transplantation of B6 T cells. Flow cytometric analysis of B63BALB.B TDL did not indicate expansion of any particular TCR V family, whereas V10 and V14 families were significantly expanded in the B6->CXBE TDL. However, PCR-based complementarity-determining region 3 size spectratyping revealed overlapping involve- ment of donor V1, 6, 8, 9, 10, and 14 families in both BALB.B and CXBE recipients and unique utilization of the V4 family in BALB.B mice, suggesting oligoclonal T cell responses to a limited number of miHA. In addition, the injection of CD8؉V14؉ http://www.jimmunol.org/ B6 T cells into irradiated BALB.B and CXBE mice induced lethal GVHD, confirming the involvement of miHA-specific T cells within an individual V family. The Journal of Immunology, 1998, 161: 41–48.
llogeneic bone marrow transplantation (BMT)3 is an im- Multiple minor histocompatibility Ags (miHA) play an impor- portant therapy for the treatment of hematologic disor- tant role in the induction of GVHD following MHC-matched al- A ders such as leukemia and aplastic anemia (1). The prin- logeneic BMT (8–11). MiHA are presented by MHC class I and cipal complication of BMT is the development of T cell-mediated class II molecules (9, 12), the former of which are recognized by graft-vs-host disease (GVHD) (2–5). Although the approach of CTL that can be generated in vitro by secondary stimulation of depleting T cells from the donor marrow inoculum has been suc- primed effector T cells (13). As an experimental system, the T cell by guest on October 2, 2021 cessful in minimizing the development of GVHD, most clinical response of C57BL/6By (B6) mice directed against miHA ex- studies have also reported an increased incidence of graft failure, pressed by H2b-matched BALB.B mice has been thoroughly in- opportunistic infection, and leukemic relapse (6). One possible vestigated. Despite estimated differences of more than 29 miHA means of reducing all of these risk factors would be to selectively loci between these two strains, it was found that the in vitro CTL deplete those T cells from the donor marrow that are responsible response was directed to only a few immunodominant Ags (14). for inducing GVHD and to allow for residual T cells to support These BALB.B miHA were detected because of their differential engraftment and/or decrease the occurrence of infection and leu- expression in a panel of target cells from the CXBE, G, I, J and K kemic relapse. Alternatively, donor T cells can be infused into recombinant inbred (RI) strain, generated from F1 crosses from an BMT recipients at later times to lessen GVHD severity (7), but this original B6 and BALB/c mating. The hierarchy of the in vitro B6 process could also be optimized by selective depletion of T cells CTL reactivity indicated that CXBG and CXBK strains expressed reactive to host alloantigens. With either of these approaches, se- first-order immunodominant miHA, whereas CXBE expressed sec- lective depletion would be practical only if there were a limited ond-order miHA which could stimulate a response only in the repertoire of GVHD-inducing T cells. absence of the first-order Ags (14). However, GVHD studies in- volving the transplantation of B6 T cells and marrow into irradi- ated BALB.B and CXB RI strains indicated that the in vitro im- Kimmel Cancer Institute, Jefferson Medical College, Philadelphia, PA 19107 munodominant miHA hierarchy did not correlate with GVHD Received for publication September 18, 1997. Accepted for publication February potential (15). In contrast to strong GVHD responses observed in 26, 1998. the BALB.B and CXBE strains, GVHD was not evident in the The costs of publication of this article were defrayed in part by the payment of page CXBG and CXBK recipients. An interstrain GVHD response anal- charges. This article must therefore be hereby marked advertisement in accordance ysis with the BALB.B and CXB RI strains suggested that a min- with 18 U.S.C. Section 1734 solely to indicate this fact. imum of two distinct MHC class I-restricted miHA (or groups of 1 This research was supported by U.S. Public Health Service Research Grants HL55593 and CA60630. miHA) could account for induction of disease in the parental 3 2 Address correspondence and reprint requests to Dr. Robert Korngold, Kimmel Can- B6 BALB.B combination, designated GVH-1 and GVH-2 (16). cer Institute, Jefferson Medical College, 233 S. 10th Street, Philadelphia, PA 19107. Consistent with the observed importance of MHC class I-restricted 3 Abbreviations used in this paper: BMT, bone marrow transplantation; ATBM, anti- miHA, all positive B6 GVHD responses in the BALB.B and CXB Thy-1 mAb plus C-treated bone marrow; B6, C57BL/6By; GVHD, graft-vs-host dis- RI strains were found to involve mediation by CD8ϩ T cells, most ease; MST, median survival time; miHA, minor histocompatibility antigen/s; RI, re- of which appeared to be dependent on CD4ϩ T cell help for their combinant inbred; TDL, thoracic duct lymphocytes; CDR3, complementarity- ϩ determining region 3, DEPC, diethyl pyrocarbonate. generation; in BALB.B mice, CD4 T cells also independently
Copyright © 1998 by The American Association of Immunologists 0022-1767/98/$02.00 42 TCR V REPERTOIRES IN GVHD caused a high level of lethal GVHD (17). Furthermore, a recent Irradiation ϩ  phenotypic study of the B6 CD4 TCR V repertoire during the Recipient mice received a lethal dose of whole body irradiation (825 cGy) early development of GVHD in BALB.B and CXBE mice has from a Gammacell (Atomic Energy of Canada, Kanata, Ontario, Canada) given the first indication of involvement of a limited anti-host 137Ce source at a dose rate of 116 cGy/min. miHA response (18). Monoclonal antibodies In the current investigation, we approached the more critical ϩ Ascites fluid for anti-Thy-1.2 (J1j, rat IgM (22), anti-CD4 (RL172, rat IgM question of the nature and extent of the B6 CD8 T cell GVHD (23), and GK1.5, rat IgG (24)) mAb were used for cell preparations. Af- response to miHA in the BALB.B and CXBE recipients. Positively finity-purified goat anti-mouse IgG (whole molecule) Ab was purchased selected miHA-specific T cell blasts were collected from the tho- from Cappel-Organon Teknika (Westchester, PA). For magnetic cell sort- racic duct lymphocyte (TDL) pool of irradiated BALB.B and ing and flow cytometric analyses, all mAb were purchased from PharM- ingen (San Diego, CA) and included FITC-conjugated anti-CD4 (clone CXBE mice, 5 days after transplantation of host-primed B6 T ϩ RM4-5; no. 01064D) and anti-CD8 (clone 53-6.7; no. 01044D) mAb and cells. Initial flow cytometric phenotype analysis of the CD8 T biotinylated mAb specific for V2, 3, 4, 5.1/5.2, 6, 7, 8.1/8.2, 9, 10, 11, 12, cells in the B63BALB.B TDL did not suggest significant expan- 13, and 14.  3 sions in any TCR V family, whereas B6 CXBE TDL displayed Preparation of donor cells significant expansions of the V10 and V14 families. TCR V repertoire complexity was further examined by the Bone marrow cells were flushed from the femurs and tibias of B6 donor mice with PBS ϩ 0.1% BSA and washed. To prepare anti-T cell-depleted highly sensitive PCR-based complementarity-determining region 3 bone marrow (ATBM), cells were incubated with J1j mAb (1:100 dilution (CDR3) size spectratyping analysis, in which the bulk TCR sizes of ascites fluid) and guinea pig C (1:30 dilution) in 6 ml of PBS ϩ 0.1% for any given V family of a control population exhibit a Gaussian BSA at 37°C for 50 min, and washed three times. ATBM was adjusted to Downloaded from ϫ 7 distribution ladder of in-frame expressible bands separated by 1.6 10 cells/ml in PBS for i.v. injection (0.1 ml) into recipients. T cell-enriched donor cells were prepared from pooled spleen and lymph three bases (19, 20). A skewing of the normal size distribution, node cell suspensions from appropriate BALB.B or CXBE primed and reflected by increased band intensity, is indicative of an expanded boosted B6 mice (i.p. injection of 1.5 ϫ 107 spleen cells 2.5 wk apart). The CDR3 size expression and suggests an oligoclonal T cell response cells were washed and resuspended in Gey’s balanced salt solution con- taining 0.7% NH Cl to remove RBC. The cells were washed twice and (21). Characterization of the TCR V repertoire of the responding 4 filtered through a cell strainer to remove dead cells. B cells were removed
ϩ http://www.jimmunol.org/ donor CD8 T cells is the first step toward ultimate identification by panning the cell suspension on goat anti-mouse IgG precoated plastic of the specific T cells involved in the GVHD response. By this petri dishes for1hat4°C, as previously described (16). The nonadherent approach, we have found that B63BALB.B TDL T cells exhib- (whole T cell enriched) cells were harvested. Cell purity was Ͼ85% pos- itive for CD3 expression, as determined by flow cytometry. For induction ited biased CDR3 size usage in the V1, 4, 6, 8, 9, 10, and 14 ϩ 3 of GVHD, whole T cells were depleted of CD4 T cells by incubation with families, whereas B6 CXBE TDL T cells exhibited overlapping RL172 mAb (1:100 dilution of ascites fluid) and guinea pig C (1:30 dilu- usage of the V1, 6, 8, 9, 10, and 14 families. The V4 response tion) in 6 ml of PBS ϩ 0.1% BSA at 37°C for 50 min and washed three ϩ appeared to be unique in recognition of miHA expressed only by times. CD8 enriched cells were incubated with either biotinylated or  BALB.B recipients. The implications of these findings are that FITC-conjugated mAb specific for V 14 in PBS for 30 min at 4°C. Cells were washed twice in PBS and incubated with MACS Magnetic strepta-
GVHD in this model system seems to involve an oligoclonal re- vidin microbeads (Miltenyi Biotec, Bergisch-Gladbach, Germany) for 30 by guest on October 2, 2021 sponse to a limited number of immunodominant miHA. Further- min at 4°C. The cells were washed twice, resuspended in PBS ϩ 0.1% more, these miHA appear capable of inducing T cell responses in BSA, and positively selected by either fluorescent cell sorting using an both the CXBE RI and BALB.B parental strains, with no evidence EPICS ELITE flow cytometer (Coulter Electronics, Hialeah, FL) or the Vario MACS Column System (Miltenyi). CD8ϩV14ϩ T cells were of competitive inhibition as previously observed in vitro (14). In ϩ ϩ washed twice and resuspended in PBS for i.v. injection. For the GVHD addition, CD8 V14 B6 T cells displayed high GVHD potential experiments with CXBE recipients, CD4ϩ T cells were not depleted from upon transplantation into irradiated BALB.B and CXBE mice, the donor cell populations. supporting the involvement of T cells from an individual V fam- GVHD assay ily in the development of GVHD. BALB.B or CXBE mice were lethally irradiated (825 cGy) and ϳ6 h later injected i.v. (in a volume of 0.5 ml of PBS) with the appropriate prepara- tion of ATBM (2 ϫ 106) alone or in conjunction with either TDL or the Materials and Methods experimental cell preparation to be tested. Mice were checked daily for morbidity and mortality and weighed twice per week until the experiments Mice were terminated. Statistical comparisons between groups were performed C.B10-H2b/LiMcdJ (BALB.B), CXB-2/By (CXBE), and C57BL/6By (B6) by the Wilcoxon nonparametric test. mice were purchased from The Jackson Laboratory (Bar Harbor, ME) Thoracic duct cannulation and/or raised in our breeding colony from breeder pairs provided by The Jackson Laboratory. For all experiments, sex-matched mice were used as Thoracic duct cannulation was used to obtain positively selected host donors and recipients between the ages of 7 and 16 wk. Mice were kept in miHA-specific T cells, as previously described (18, 25). BALB.B or CXBE a pathogen-free environment in autoclaved microisolator cages and were mice (3–4 mice/group) were lethally irradiated (825 cGy) and injected i.v. provided with acidified (pH 2.5) water and autoclaved food ad libitum. 6 h later with unseparated T cells (1–1.5 ϫ 107 cells in 0.25 ml of PBS) from host-primed donor B6 mice and were cannulated 5 days later. Briefly, mice were anesthetized by an i.p. injection of avertin (1:50 dilution in PBS Media of a stock solution of 10 g of tribromoethanol in 10 ml of tert-amyl alcohol; Aldrich Chemical, Milwaukee, WI), an i.v. saline line inserted in the tail, PBS (BioWhittaker, Walkerville, MD) supplemented with 0.1% BSA (Sig- and a cannula (PE-50 intramedic tubing; Clay Adams, Parsipanny, NJ) ma Chemical, St. Louis, MO) was used for all in vitro manipulations of the surgically implanted into the thoracic duct. TDL were collected over a 20-h donor bone marrow cells and lymphocytes. RPMI 1640 (Mediatech, Hern- period and the cells pooled from individual mice of each group. don, VA) supplemented with 10% FCS (Sigma), 2 mM L-glutamine (Me- Flow cytometric analysis of TCR V expression diatech), 50 IU/ml penicillin (Mediatech), 50 g of streptomycin (Medi- atech), and 5 IU/ml heparin sulfate (Schein Pharmaceutical, Florham Park, For fluorescent staining, cells were washed, resuspended in FACS medium, NJ) was used for TDL collection. PBS supplemented with 1% BSA and and incubated for 5 min at 4°C with culture supernatant containing anti- 0.1% NaN3 was used as a medium for staining cells for flow cytometry FcR␥ll mAb (clone 2.4G2, HB197; American Type Culture Collection, (FACS media). PBS alone and supplemented with 1% BSA was used dur- Rockville, MD; (26)) to prevent nonspecific staining by mAb. An isotype- ing magnetic cell sorting. matched FITC-conjugated mAb was used as a negative control. TDL T The Journal of Immunology 43
cells were assayed for the percentage of V expressing CD8ϩ T cells by two-color flow cytometry. Cells were first incubated with each biotinylated anti-V mAb in combination with FITC-anti-CD8 mAb for 30 min at 4°C. The cells were then washed twice and incubated with PE-streptavidin (Caltag, San Francisco, CA) for 30 min at 4°C, washed twice, and fixed in 1% paraformaldehyde. Cells were analyzed on a EPICS C flow cytometer (Coulter Electronics). Data from individual replicate experiments were pooled, and statistical significance was determined by the Mann-Whitney Wilcoxon rank test. Preparation of RNA and cDNA CD8ϩ T cells were first prepared by depletion of CD4ϩ T cells via panning of TDL cell suspensions on GK1.5 mAb (1:20 dilution of ascites fluid in PBS ϩ 0.1% BSA) precoated plastic petri dishes for1hat4°C. This procedure yielded a restricted population with Ն92% CD8ϩ cells. Total cellular RNA was then generated from 106 to 107 nonadherent CD8ϩ cells by homogenization in 1 ml of Ultraspec (Biotecx Laboratories, Houston, TX), separating cellular DNA and protein by the addition of a 1:5 volume  of chloroform, vortexing for 5 s, and centrifuging at 12,500 rpm for 15 min. FIGURE 1. TCR V repertoire analysis of positively selected 3 3 ϩ The aqueous phase was transferred to a clean Eppendorf tube, and RNA B6 BALB.B and B6 CXBE CD8 TDL. BALB.B and CXBE mice 7 was precipitated at 4°C by adding an equal volume of isopropanol and were lethally irradiated (825 cGy) and transplanted with 3 ϫ 10 host- centrifuging at 12,500 rpm for 15 min. The supernatant was removed, and presensitized B6 donor T cells. Five days posttransplant, the animals were the pellet was washed with 75% ethanol in diethyl pyrocarbonate (DEPC) cannulated and TDL were collected and pooled for V analysis by two- Downloaded from treated water and centrifuged, as above. The RNA pellet was resuspended color flow cytometry. The values for the BALB.B and CXBE recipients are ϩ in 25 l of DEPC water, heated to 55–65°C for 10 min and stored at the mean percentage Ϯ SEM of V expression on CD8 TDL T cells Ϫ 20°C. Recovery of RNA was determined by spectrophotometry. The pooled from five and six replicate experiments, respectively. Naive splenic poly(A)ϩ portion of the total RNA was converted into cDNA using oli- B6 T cells were analyzed to serve as baseline controls and are expressed as go(dT) as a primer for reverse transcription. Two micrograms of total RNA in a volume of 9.5 l were heated to 70–80°C, centrifuged briefly, and the mean values from the pool of four replicate experiments. Significant increases in V expression ( p Ͻ 0.05) are indicated (*). placed on ice. Master mix (17.5 l), containing 1 l of RNasin (40 U/l), http://www.jimmunol.org/ 6 lof5ϫ Maloney murine leukemia virus reverse transcriptase reaction buffer, 3 l of oligo(dT) primer (20 mM), 1.5 l of deoxynucleotide triphosphates A, G, C, T (25 mM each), and 3 l of Maloney murine the appropriate recipient type to generate any significant TCR V leukemia virus reverse transcriptase (300 U/ l), was incubated at 37°C for family expansion of the CD8ϩ T cell population within the first 1.5 h to synthesize cDNA. All reagents were purchased from Promega (Madison, WI). The reaction mixture was then heated to 95°C for 3 min, week posttransplantation. The recipients underwent thoracic duct and the cDNA was stored at Ϫ20°C. PCR was performed using murine cannulation on day 5, and TDL was collected over a 20-h period -actin primers to establish the quality of the cDNA. and pooled from three to four mice per group. The positively se- lected TDL were Ͼ95% CD3ϩ and consisted of 55 to 67% CD4ϩ PCR amplification of cDNA and CDR3 size spectratyping ϩ
and 33 to 45% CD8 T cells. A significant percentage (20–25%) by guest on October 2, 2021 PCR was performed using a labeled constant primer (Cb) and a V of the TDL were blast-like in size, in contrast to TDL collected   primer specific for each V family to be analyzed. C b was labeled using from B6 mice transplanted with T cells (1–1.5 ϫ 107) from syn- polynucleotide kinase (Promega) and [␥-32P]ATP (Dupont-NEN, Boston, geneic-presensitized B6 mice, which yielded few blast-like cells MA). All the primers used have been previously described (27) with the ϩ exception of the V20 primer, which was designed with a sequence of and were Ͼ98% CD4 . Therefore, the TCR V repertoire analy- GGTCAAGGAGAGATTCTCAGCTGT. Five microliters of 10ϫ Taq ses of TDL CD8ϩ T cells from GVHD recipients were compared polymerase buffer B (Promega) were added to 5 l of cDNA plus 2 lof with the TCR V repertoire of normal B6 splenic CD8ϩ T cells. MgCl2 (25 mM), 4 l deoxynucleotide triphosphates A,G, C, T (25 mM   each), 2.5 lofV sense primer (20 M), 4 lC b labeled antisense TCR V repertoire analysis by flow cytometry primer (12 M), 28 l DEPC-treated water, and 1 lofTaq polymerase (2–5 U/l, Promega). Thirty cycles of amplification were conducted and on TCR V repertoire analysis was conducted by two-color flow cy- completion of PCR, 50 lof2ϫ loading buffer containing 95% form- tometry of the TDL using a panel of anti-V mAb along with amide, 20 mM EDTA, 0.05% bromphenol blue,, and 0.05% Xylene Cyanol  FF was added to each reaction. PCR reactions were either stored at Ϫ20°C anti-CD8 mAb. Of the 13 V families for which specific mAb or electrophoresed. Each reaction tube was heated to 70°C for 3 min, and were available, positively selected B63BALB.B TDL did not dis- 6 l of each were applied to a prewarmed 6% acrylamide sequencing gel, play any significant V family expansion (Fig. 1). On the other as previously described (20). The gels were run at 55 mA for 1.5 h to hand, the B63CXBE TDL exhibited significant expansions of the maximize band resolution. Sequencing gels were dried, and autoradiogra- V10 (6.1% to 9.9%; p Ͻ 0.05) and V14 (3.3% to 6.8%; p Ͻ phy was generally performed for 15 h at room temperature without inten- 0.05) families; and marginal expansion of the V11 family (8.7% sifying screens. Densitometric scanning of autoradiographs was performed ϩ on a Personal Densitometer SI using ImageQuant software (Molecular Dy- to 11.8%; p Ͼ 0.05), relative to naive B6 splenic CD8 T cells namics, Sunnyvale, CA). The interpretations of the spectratypes were (Fig. 1). based on the criteria previously established by Gorski et al. (19, 20). The relative band intensities within a given spectratype were examined and Induction of GVHD by B63BALB.B TDL T cells compared with the intensity patterns in the spectratype of the naive B6 7 control group. The other bands within a given spectratype act as the in- B63BALB.B TDL T cells (1.4 ϫ 10 ) were transplanted along ternal controls for any variability due to the PCR expansion or to different with 2 ϫ 106 ATBM into lethally irradiated (825 cGy) BALB.B or sample sizes. CXBE recipients to demonstrate that these cells were capable of inducing GVHD. Control recipients received an injection of 2 ϫ Results 106 ATBM alone. By day 25, both the BALB.B and CXBE recip- General approach ients exhibited the clinical symptoms of GVHD, including diar- B6 T cells (1–1.5 ϫ 107) were injected into irradiated (825 cGy) rhea, ruffled fur, and weight loss (Ͼ20%), compared with control BALB.B and CXBE mice for induction of GVHD. It was neces- ATBM mice (Fig. 2A). All of the BALB.B recipients of TDL sary to use cells from donor mice that had been presensitized succumbed by day 50 (MST of 30 days), more rapidly than the (primed and boosted by i.p. injection of 1.5 ϫ 107 spleen cells) to CXBE recipients which died by day 72 (MST of 50 days; Fig. 2B). 44 TCR V REPERTOIRES IN GVHD
family, additional bands were enhanced in the B63 BALB.B CD8ϩ TDL spectratype which were not skewed in the B63CXBE CD8ϩ. Most notably, biased CDR3 usage in the V4 family was unique to the B63BALB.B CD8ϩ TDL, suggesting a response to a BALB.B-specific miHA.
Induction of GVHD by T cells from a single TCR V family Fluorescent cell sorting was used to positively select CD8ϩV14ϩ T cells from CXBE-presensitized B6 mice which were then trans- planted into irradiated (825 cGy) BALB.B recipients to demon- strate that cells from a single V family were capable of inducing GVHD. Recipients were injected with either the combination of 3 ϫ 106 naive B6 CD4ϩ T cells (to provide Th function) and 2 ϫ 106 ATBM alone or with either 4 ϫ 106 CXBE-presensitized un- separated B6 CD8ϩT cells or 1 ϫ 105 CD8ϩV14ϩ-enriched T cells (Ͼ99% V14ϩ). Recipients of either the unseparated or V14ϩCD8ϩ T cells exhibited the clinical symptoms of GVHD between days 45 and 50, including weight loss (22 and 30%, re- spectively; Fig. 4A). By day 76, 60% (MST of 49 days) of the Downloaded from BALB.B mice that received the unseparated CD8ϩ T cells and 80% (MST of 67 days) of those given the V14ϩCD8ϩ T cells had died, compared with 20% mortality in the ATBM control group (Fig. 4B). In a similar manner, lethally irradiated (825 cGy) CXBE mice were injected with 2 ϫ 106 ATBM either alone or with either 7 6 ϩ 3 ϫ 10 CXBE-presensitized unseparated or 3.5 ϫ 10 V14 http://www.jimmunol.org/ enriched (Ͼ84% V14ϩ selected by magnetic cell sorting) B6 T FIGURE 2. GVHD induction following the adoptive transfer of posi- cells. Both groups of CXBE recipients that received T cells ex- tively selected B63BALB.B TDL. Lethally irradiated (825 cGy) BALB.B hibited the clinical symptoms of GVHD by day 40, including and CXBE mice were adoptively transferred with 2 ϫ 106 B6 ATBM alone weight loss (17 and 25%, respectively; Fig. 5A). By day 45, mice or in a mixture with 1.4 ϫ 107 B63BALB.B TDL T cells (unseparated) that received injections of V14ϩ-enriched CD8ϩ T cells had 80% 7 collected on day 5 after injection of 3 ϫ 10 BALB.B-presensitized B6 fatality (MST of 48 days), and by day 70, the mice that had been donor T cells. A, At each time indicated, the mean body weight for each injected with unseparated CD8ϩ T cells had 40% fatality, as com- ϭ ϩ 3 recipient group (n 5 for all groups, except TDL ATBM BALB.B; pared with 100% survival in the control ATBM group (Fig. 5B). n ϭ 4) was calculated and normalized as a percentage of the initial starting by guest on October 2, 2021 weight. B, The survival of transplanted recipients as described above. The results are from a representative of two similar experiments. Discussion The TCR V repertoires of positively selected TDL collected from presensitized B63BALB.B and B63CXBE transplanted mice Molecular analysis of TCR were analyzed to gain insight into the CD8ϩ T cell populations that The CDR3 sequence encoding the variable -chain V-D-J junc- mediate GVHD directed against multiple miHA disparities. The tional region defines the unique TCR clonotype(s) specific for a TDL pool is an excellent site for monitoring T cell responses since miHA and acts as a fingerprint for the T cell lineage(s) bearing it. recently activated T cells expand and enter the thoracic duct after PCR-based analysis, known as CDR3 size spectratyping (19, 20), initially encountering miHA in the spleen and lymph nodes. was used to determine the nature of the V family repertoires Highly viable (Ն99%) cells can be retrieved from the TDL pool, observed in the B63BALB.B and B63CXBE CD8ϩ TDL, as a containing significant levels of blast-like T cells, in contrast to reflection of the initial GVHD response. The V bands of a spec- retrieval of T cells from the peripheral lymphoid organs (28). Due tratype from normal B6 splenic CD8ϩ T cells followed a Gaussian to the overall low frequency responses involved in anti-miHA re- distribution (Fig. 3A), with the center bands of median CDR3 activity and the slower development of primary CD8ϩ T cell re- length being most intense and the outer bands of longer and shorter sponses in vivo relative to CD4ϩ T cells, significant phenotypic lengths being less intense. In contrast, the V band patterns of skewing of V families was not detectable over a period of 3 to 8 skewed spectratypes from TDL did not exhibit Gaussian distribu- days posttransplantation of naive B6 T cells, although significant tions, with some bands being more intense relative to the same skewing occurred in the CD4ϩ T cell population (18). The can- band distribution in the B6 control sample (exemplified in Fig. 3). nulation of mice at later time points was impractical, since the The CDR3 size spectratypes of 17 V families expressed in B6 mice began developing GVHD-related symptoms and could not mice (V1-16 and V20) were examined for band skewing sug- survive the procedure. As an alternative, anti-host miHA T cells gestive of an oligoclonal expansion of particular T cell lineages. were first expanded by presensitization and boosting of donor B6 The results (summarized in Table I) supported the findings of the mice in vivo with host-type splenocytes. We then investigated the phenotypic analysis; i.e., V6 exhibited biased CDR3 usage in the response patterns of these donor cells when placed in a GVHD- B63 BALB.B CD8ϩ TDL and V6, V10, and V14 were inducing environment in the irradiated BALB.B or CXBE recipi- skewed in the B63CXBE CD8ϩ TDL (Fig. 3). In addition, the ents. To confirm that the TDL contained alloreactive anti-miHA- spectratype analysis indicated that both strain combinations exhib- specific T cells, TDL collected from the B63 BALB.B ited skewing of bands in the V1, 6, 8, 9, 10, and 14 families. In combination were injected into lethally irradiated BALB.B and all cases, the same CDR3 size band was enhanced in the same V CXBE recipients. The ability of these cells to induce GVHD in family between the two groups (Fig. 3). However, in the V14 both hosts suggested that the TDL contained alloreactive CD8ϩ T The Journal of Immunology 45
Table I. Summary of TCR V utilization of CD8ϩ TDL from BALB.B and CXBE recipients as analyzed by CDR3 size spectratyping
Skewed Pattern ina
V Family CXBE BALB.B
1 ϩϩ 2 ϪϪ 3 ϪϪ 4 Ϫϩ 5.1/.2 ϪϪ 6 ϩϩ 7 ϪϪ 8.1/.2/.3 ϩϩ 9 ϩϩ 10 ϩϩ 11 ϮϮ 12 ϪϪ 13 ϪϪ 14 ϩϩ 15 ϪϪ
16 ϪϪDownloaded from 20 ϪϪ
a Irradiated CXBE and BALB.B mice were injected with 1 to 1.5 ϫ 107 host- presensitized B6 donor T cells and cannulated 5 days later. TDL were collected from three to four mice per group and CD8ϩ T cells were isolated. CDR3 size spectratype analysis was then performed as described in Materials and Methods. Results were consistent from two separate experiments involving different sets of mice, except for V11, where Ϯ indicates that weak skewing was observed in one experiment but not the other. http://www.jimmunol.org/
cells. Lethal GVHD in the B63BALB.B model is mediated by either CD4ϩ or CD4-dependent CD8ϩ T cells, while in the B63CXBE model only CD4-dependent CD8ϩ T cells are respon- sible for pathogenesis (17). The phenotypic analysis was not sen- sitive enough to detect any major shifts in the B63BALB.B TDL but did indicate expansions of at least the V10 and V14 families by guest on October 2, 2021 in the B63CXBE TDL (Fig. 1). This observation was consistent with the involvement of a limited number of distinct miHA epitopes in the development of the B6 CD8ϩ anti-CXBE GVHD response. It was not overly surprising that phenotypic analysis of the B63BALB.B TDL failed to detect a specific V family response. This approach is capable of detecting only major shifts in V utilization, and even if successful, it is unable to further charac- terize the response and determine whether it is oligoclonal. A phe- notypic increase in V utilization could also be due to polyclonal expansion which would represent increased expression of many TCR V-D-J sequences within that V family. For example, poly- clonal expansion was previously observed for the CD4ϩ V3 fam- ily in the B63BALB.B strain combination and was likely due to MTV-6 superantigen stimulation (18). T cell repertoire analysis based on the size heterogeneity of the CDR3 region is, therefore, a much more sensitive and powerful tool for the study of GVHD and other types of immune responses (21, 29, 30). It also provides a mechanism for differentiating the specific TCR V-D-J sequences expressed within the same V family, and can reveal much more subtle skewing within a V family than can be detected by FACS analysis.
are the spectratypes of V10, 14, 4, and 5. B, Densitometric scanning of the gel films in A generated histograms to compare the size distribution of CDR3 bands between the B6 naive splenic T cells (histogram 1), and FIGURE 3. V specific primers were used to generate PCR products B63CXBE (histogram 2)orB63BALB.B TDL (histogram 3). Arrows from purified (lane 1) B6 naive splenic T cells, B63CXBE (lane 2), and indicate CDR3 bands that display increased relative intensities within the B63BALB.B TDL RNA (lane 3). A, CDR3 size spectratyping analysis spectratype. Starred peaks in the V4 histograms are contaminants and not was performed as described in Materials and Methods. The results shown part of the spectratype. 46 TCR V REPERTOIRES IN GVHD Downloaded from http://www.jimmunol.org/
FIGURE 4. GVHD induction following the injection of CXBE-presen- sitized B6 CD8ϩV14ϩ T cells into BALB.B mice. Lethally irradiated FIGURE 5. GVHD induction following the injection of CXBE-presen- ϩ (825 cGy) BALB.B mice were injected with 3 ϫ 106 naive B6 CD4ϩ T sitized B6 V14 T cells into CXBE mice. Lethally irradiated (825 cGy) 6 7 cells, 2 ϫ 106 ATBM, and either 4 ϫ 106 CXBE-presensitized B6 CD8ϩ CXBE mice were injected with 2 ϫ 10 ATBM, and either 3 ϫ 10 CXBE- 6 ϩ T cells or 1 ϫ 105 CD8ϩV14ϩ T cells. Control recipients received in- presensitized unseparated B6 T cells or 3.5 ϫ 10 V14 T cells. Control jections of B6 CD4ϩ and ATBM only. A, At each time point indicated, the recipients received injections of ATBM only. A, At each time point indi- mean body weight for each recipient group (n ϭ 5) was calculated and cated, the mean body weight for each recipient group (n ϭ 5) was calcu-
normalized, as a percentage of the initial starting weight. B, The survival lated and normalized, as a percentage of the initial starting weight. B, The by guest on October 2, 2021 of transplanted recipients as described above. survival of transplanted recipients as described above.
The increased V heterogeneity observed in the CDR3 size archy was also found to operate for the generation of miHA-spe- spectratyping analysis would suggest that GVHD is mediated by a cific CTL in vitro (14) and was possibly due to the comparative heterogeneous population of alloreactive CD8ϩ T cells that rec- abilities of miHA to compete successfully for Ag-binding residues ognize either the same miHA, several miHA, or multiple epitopes in the appropriate MHC molecules presented by APC. The affinity/ of the same Ag in the host. Yet, the overall responses to miHA in avidity of the interaction of the specific TCR for miHA/self-MHC the B63BALB.B and B63CXBE strain combinations were lim- may also be an important factor of miHA immunodominance. Al- ited to only a few V families, supporting the notion that a limited ternatively, T cells that respond more vigorously to strong miHA number of miHA or epitopes were being recognized. The involve- may down-regulate weaker responses to other miHA via cytokine ment of a limited number of immunodominant miHA in GVHD production. Each of these potential mechanisms or combinations has been previously found by Perreault et al. in another model thereof could account for the phenomenon of competitive immu- system (31). In addition, expansion of a limited set of CD8ϩ V nodominance. Recent studies comparing CTL immunodominant families in local liver GVHD pathogenesis has been observed by specificities and skin graft rejection in the B6 anti-BALB.B and Howell et al. (32). It has also been shown that Ag-specific CD8ϩ CXB strain combinations have found a distinct lack of correlation T cells exhibit limited heterogeneity at the level of TCR V gene between the two responses, with CTL recognizing only a limited usage during a primary response to HIV infection (33). The notion number of miHA operative in vivo (13). of heterogeneous V responses developing against a single or lim- A similar lack of correlation with CTL immunodominant spec- ited number of miHA is highly possible since T cell recognition of ificities had previously been found in GVHD (15, 17). For GVHD Ag involves the interaction of both V and V␣ chains (34). Each then, the relevant question becomes how extensive are the number of the responding V chains could associate with a different V␣ of miHA that are actually involved in GVHD development? There chain to recognize the same epitope. This has been observed in the must be certain limitations and qualifications for miHA to serve in diversity of CD4ϩ T cell responses to influenza antigenic this capacity. At least two of these criteria would seem to be im- determinants (35). munogenicity and tissue distribution. In terms of the latter, it is It is estimated that more than 29 miHA locus differences exist known that some miHA have unique tissue expression (38, 39), between the B6 and BALB/c (and therefore the BALB.B) inbred and it would be expected that for optimum GVHD pathogenesis, mouse strains (36). Yet only a few immunodominant miHA appear Ag should be available not only in cells of hemopoietic origin but to be recognized in vivo by MHC-matched allogeneic T cells re- also in the primary target organs, including the intestine, liver, and sponsible for skin graft rejection (37). An immunodominant hier- skin. The lack of GVHD in irradiation chimeric models that have The Journal of Immunology 47 miHA expressed only in the host hemopoietic compartment sup- References ports the notion that the presence of miHA in both locations is 1. Thomas, E. D., R. A. Clift, and R. Storb. 1984. Indications for marrow trans- required (40). plantation. Annu. Rev. Med. 35:1. Thus, there may only be a limited number of miHA that fit the 2. Gale, R. P. 1985. Graft-versus-host disease. Immunol. Rev. 88:193. criteria for GVHD. In this regard a recent study of the CXB in- 3. Santos, G. W., A. D. Hess, and G. B. Vogelsang. 1985. Graft-versus-host reac- terstrain lethal GVHD responses suggested the minimal involve- tions and disease. Immunol. Rev. 88:169. 4. Deeg, H. J., and R. Storb. 1984. Graft-versus-host disease: pathophysiological ment of two distinct immunodominant miHA (or associated groups and clinical aspects. Annu. Rev. Med. 35:11. of miHA) in the B63 BALB.B GVHD response (16). One class 5. Ferrara, J., and H. J. Deeg. 1991. Graft-versus-host disease. N. Engl. J. Med. I-restricted miHA (GVH-1) appeared to be shared by the CXBE 324:667. 6. Kernan, N. A. 1994. T-cell depletion for prevention of graft-versus-host disease. strain, while the second Ag (GVH-2) was uniquely expressed by In Bone Marrow Transplantation. S. J. Forman, K. G. Blume, and E. D. Thomas, the BALB.B strain. The biased CDR3-size skewing of V1, 6, 8, eds. Blackwell Scientific Publications, Cambridge, MA, p. 124. 9, 10, and 14 families in both the BALB.B and CXBE recipients 7. Kolb, H. J., J. Mittermuller, C. H. Clemm, E. Holler, G. Ledderose, G. Brehm, M. Heim, and W. Wilmanns. 1990. Donor leukocyte transfusions for treatment of may represent a response to common miHAs shared by BALB.B recurrent chronic myelogenous leukemia in marrow transplant patients. Blood and CXBE mice, i.e., GVH-1. Although the same CDR3 size 76:2462. bands were skewed in each corresponding V family (except for 8. Korngold, R., and J. Sprent. 1978. Lethal graft-versus-host disease following V14), the final identity of these specificities will depend on se- bone marrow transplantation across minor histocompatibility barriers in mice: prevention by removing mature T cells from marrow. J. Exp. Med. 148:1687. quence analysis of the CDR3 segments involved in each response. 9. Perreault, C., F. Decary, S. Brochu, M. Gyger, R. Belanger, and D. Roy. 1990. Furthermore, the biased CDR3 usage of V4 in the BALB.B re- Minor histocompatibility antigens. Blood 76:1269.
cipients, not present in the spectratype of the CXBE recipients, 10. Goulmy, E., P. Voogt, C. van Els, M. de Bueger, and J. van Rood. 1991. The role Downloaded from of minor histocompatibility antigens in GVHD and rejection: a mini-review. might represent a response to the unique BALB.B miHA, i.e., Bone Marrow Transplant. 1:49. GVH-2. Experiments are under way to determine whether posi- 11. Martin, P. J. 1991. Increased disparity for minor histocompatibility antigens as a tively selected CD8ϩV4ϩ T cells will induce GVHD in BALB.B potential cause of increased GVHD risk in marrow transplantation from unrelated but not CXBE recipients. donors compared with related donors. Bone Marrow Transplant 8:217. 12. Rammensee, H. G., K. Falk, and O. Rotzschke. 1993. Peptides naturally pre- The biased CDR3 size skewing of V14 displayed the only sented by MHC class I molecules. Annu. Rev. Immunol. 11:213. 3 13. Nevala, W. K., C. Paul, and P. J. Wettstein. 1997. Reduced diversity of CTLs nonidentical skewing pattern between the B6 BALB.B and http://www.jimmunol.org/ B63CXBE T cell responses. The BALB.B spectratype exhibited specific for multiple minor histocompatibility antigens relative to allograft rejec- tion in vivo. J. Immunol. 158:1102. two additional skewed bands. The additional bands present in the 14. Wettstein, P. J. 1986. Immunodominance in the T-cell response to multiple non- V14 spectratype of the BALB.B recipients might also represent H-2 histocompatibility antigens. II. Observation of a hierarchy among dominant a response to the unique BALB.B miHA (GVH-2). Although B6 antigens. Immunogenetics 24:24.  ϩ 15. Korngold, R., and P. J. Wettstein. 1990. Immunodominance in the graft-vs-host V 14 T cells injected into either BALB.B or CXBE recipients disease T cell response to minor histocompatibility antigens. J. Immunol. 145: induced lethal GVHD (80% mortality by day 80 for both groups of 4079. recipients; Figs. 4B and 5B), the number of cells injected into the 16. Korngold, R., C. Leighton, L. E. Mobraaten, and M. A. Berger. 1997. Inter-strain ϫ 5 ϫ 6 graft-vs.-host disease T-cell responses to immunodominant minor histocompati- BALB.B recipients was 30-fold less (1 10 vs 3.3 10 ) than bility antigens. Biol. Blood Marrow Transplant. 3:57. into the CXBE recipients. This difference in the response sug- 17. Berger, M., P. J. Wettstein, and R. Korngold. 1994. T cell subsets involved in by guest on October 2, 2021 gested a possible additive effect of different anti-miHA clonotypes lethal graft-versus-host disease directed to immunodominant minor histocompat- ibility antigens. Transplantation 57:1095. in the BALB.B, consistent with GVH-1 and GVH-2. The capacity ϩ  ϩ 18. Berger, M. A., and R. Korngold. 1997. Immunodominant CD4 T cell receptor of presensitized V 14 donor T cells to cause lethal GVHD is V repertoire involved in graft-versus-host disease responses to minor histocom- noteworthy on its own accord. It has been very difficult in the past patibility antigens. J. Immunol. 159:77. to demonstrate GVHD induction with normal donor T cells di- 19. Gorski, J., M. Yassai, X. Zhu, B. Kissella, C. Keever, and N. Flomenberg. 1994. Circulating T cell repertoire complexity in normal individuals and bone marrow rected to limited miHA specificities. The use of miHA congenic recipients analyzed by CDR3 size spectratyping. J. Immunol. 152:5109. strain combinations has failed to generate detectable GVHD, even 20. Gorski, J., T. Piatek, M. Yassai, and K. Maslanka. 1995. Improvements in rep- when donor cells were presensitized to host Ags (16, 41). Single ertoire analysis by CDR3 size spectratyping: bifamily PCR. Ann. NY Acad. Sci. cloned T cells have been used successfully before to induce 76:99. ϩ 21. DePalma, R., and J. Gorski. 1995. Restricted and conserved T-cell repertoires GVHD (42), but these were CD4 T cells. Furthermore, any clone involved in allorecognition of class II major histocompatibility complex. Proc. (CD4ϩ or CD8ϩ), due to their extensive expansion in culture, Natl. Acad. Sci. USA 92:8836. could potentially differ from normal T cells in mediation of disease 22. Bruce, J., F. W. Symington, T. J. McKearn, and J. Sprent. 1981. A monoclonal antibody discriminating between subsets of T and B cells. J. Immunol. 127:2496. pathogenesis (e.g., by abnormal acquisition or loss of adhesion 23. Ceredig, R., J. W. Lowenthal, M. Nabholz, and H. R. MacDonald. 1985. Ex- molecules responsible for migration into target tissue). The use of pression of interleukin-2 receptors as a differentiation marker on intrathymic stem normal donor T cells with limited heterogeneity to generate cells. Nature 314:98.  ϩ 24. Dialynas, D. P., D. B. Wilde, P. Marrack, A. Pierres, K. A. Wall, W. Havran, GVHD, as with the V 14 cells, will ultimately enable the inves- G. Otten, M. R. Loken, M. Pierres, J. Kappler, and F. W. Fitch. 1983. Charac- tigation of associations between specific anti-miHA responses and terization of the murine antigenic determinant, designated L3T4a, recognized by tissue distribution of lesions. monoclonal antibody GK1.5: expression of L3T4a by functional T cell clones appears to correlate primarily with class II MHC antigen-reactivity. Immunol. Characterization of the TCR usage by T cells responding to Rev. 74:29. miHA and mediating GVHD will help us to understand the scope 25. Korngold, R., and J. R. Bennink. 1984. Collection of mouse thoracic duct lym- of these complex responses and ultimately the nature of the miHA phocytes. Methods Enzymol. 108:270. responsible for their induction. With future developments in diag- 26. Unkeless, J. C. 1979. Characterization of a monoclonal antibody directed against mouse macrophage and lymphocyte Fc receptors. J. Exp. Med. 150:580. nostic capabilities, our appreciation for the GVHD response rep- 27. Casanova, J. L., P. Romero, C. Widmann, P. Kourilsky, and J. L. Maryanski. ertoires may also eventually allow for new targeted strategies for 1991. T cell receptor genes in a series of class I major histocompatibility com- prevention of GVHD. plex-restricted cytotoxic T lymphocyte clones specific for a Plasmodium berghei nonapeptide: implications for T cell allelic exclusion and antigen-specific reper- toire. J. Exp. Med. 174:1371. Acknowledgments 28. Sprent, J. 1977. Migration and lifespan of lymphocytes, In: B and T Lymphocytes in Immune Recognition, F. Loor and G. Roelants, eds. Wiley & Sons, Chichester, We thank David Dicker for his expert technical assistance in performing U.K., pp. 59–82. the flow cytometric analysis and fluorescent cell sorting, as part of the 29. Monteiro, J., R. Hingorani, R. Peroglizzi, B. Apatoff, and P. K. Gregersen. 1997. Kimmel Cancer Institute Flow Cytometry Facility. Oligoclonality of CD8ϩ T cells in multiple sclerosis. Autoimmunity 23:127. 48 TCR V REPERTOIRES IN GVHD
30. Hingorani, R., J. Monteiro, R. Furie, E. Chartash, C. Navarrete, R. Pergolizzi, and 37. Wettstein, P. J., and D. W. Bailey. 1982. Immunodominance in the immune P. K. Gregersen. 1996. Oligoclonality of V 3 TCR chains in the CD8ϩ T cell response to “multiple” histocompatibility antigens. Immunogenetics 16:47. population of rheumatoid arthritis patients. J. Immunol. 156:852. 38. Johnson, L. L., D. W. Bailey, and L. E. Mobraaten. 1981. Genetics of histocom- 31. Perreault, C., J. Jutras, D. C. Roy, J. G. Filep, and S. Brochu. 1996. Identification patibility in mice. IV. Detection of certain minor (non-H-2) H antigens in selected of an immunodominant mouse minor histocompatibility antigen (MiHA): T cell organs by the popliteal node test. Immunogenetics 14:63. response to a single dominant MiHA causes graft-versus-host disease. J. Clin. Invest. 98:622. 39. Griem, P., H.-J. Wallny, K. Falk, G. Schonrich, G. Hammerling, and H.-G. Ram- 32. Howell, C. D., J. Li, E. Roper, and B. L. Kotzin. 1995. Biased liver T cell receptor mensee. 1991. Uneven tissue distribution of minor histocompatibility protein V repertoire in a murine graft-versus-host disease model. J. Immunol. 155:2350. versus peptides is caused by MHC expression. Cell 65:633. 33. Pantaleo, G., J. F. Demarest, H. Soudeyns, C. Graziosi, F. Denis, 40. Korngold, R., and J. Sprent. 1982. Features of T cells causing H-2-restricted J. W. Adelsberger, P. Borrow, M. S. Saag, G. M. Shaw, R. P Sekaly, and lethal graft-vs.-host disease across minor histocompatibility barriers. J. Exp. Med. ϩ A. S. Fauci. 1994. Major expansion of CD8 T cells with a predominant V 155:872. usage during the primary immune response to HIV. Nature 370:463. 41. Blazar, B. R., D. C. Roopenian, P. A. Taylor, G. J. Christianson, 34. Schwartz, R. H. 1985. T-lymphocyte recognition of antigen in association with A. Panoskaltsis-Mortari, and D. Vallera. 1996. Lack of GVHD across classical, gene products of the major histocompatibility complex. Annu. Rev. Immunol. single minor histocompatibility (miH) locus barriers in mice. Transplantation 3:237. 35. Caton, A. J., and W. Gerhard. 1992. The diversity of the CD4ϩ T cell response 61:619. in influenza. Immunology 4:85. 42. Miconnet, I., R. Huchet, D. Bonardelle, R. Motta, C. Canon, E. Garay-Rojas, 36. Bailey, D. W., and L. E. Mobraaten. 1969. Estimates of the number of loci M. Kress, M. Reynes, O. Halle-Pannenko, and M. Bruley-Rosset. 1990. Graft- contributing to the histoincompatibility between C57BL-6 and BALB-c strains of versus-host mortality induced by noncytolytic CD4ϩ T cell clones specific for mice. Transplantation 7:394. non-H-2 antigens. J. Immunol. 145:2123. Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021