TCRVα usage by effector CD4⁺Vβ11⁺ T cells mediating graft-versus-host disease directed to minor histocompatibility antigens

Abstract

T cell receptor (TCR) Vα (TRAV) and Vβ (TRBV) chains provide the T cell specificity for recognition of major histocompatibility complex (MHC)-bound antigens. However, there is limited information on the diversity of TRAV usage within an antigen response. Previous investigation of CD4⁺ T cell-mediated graft-versus-host disease (GVHD) in the minor histocompatibility antigen (miHA)-mismatched C57BL/6 (B6) -> BALB.B irradiated murine model determined that Vβ11⁺ T cells were involved in the severity of disease. Polymerase chain reaction (PCR)-based complementarity determining region-3 (CDR3)-size spectratype analysis of B6 Vβ11⁺ T cells from the spleens of recipient BALB.B mice undergoing GVHD indicated biased usage within the Vα6, 9, 13, 14, 18, and 22 families. In order to probe deeper into this limited Vα response, the current study was undertaken to further define TRAV-Jα (TRAJ) nucleotide sequences found in host-presensitized B6 Vβ11⁺ T cells proliferating in response to in vitro stimulation with BALB.B splenocytes. Using the nonpalindromic adaptor-PCR method, we found dominant usage of the TRAV13-TRAJ16 transcript combination. Then, using laser capture microdissection (LMD), we found use of the identical TRAV-TRAJ nucleotide sequence in areas dominated by infiltrating Vβ11⁺ CD4⁺ T cells during development of GVHD in both the rete-like prominences of the dorsal lingual epithelium and the ileal crypts of the small intestine.

INTRODUCTION

The relative contribution of the α- and β-chain, and particularly the TRAV and TRBV regions, to the diversity and specificity of the TCR has been controversial. Although several studies have indicated that the α-chain dictates recognition of self-MHC molecules during MHC-restricted responses [1–5], other investigations have also found a significant role of the α-chain in peptide recognition [3,6,7]. In addition, crystal structure solution of a TCR interacting with a peptide-MHC II complex revealed that TRAV actually made more peptide contacts than the TRBV [8], and studies using single α- or β-chain transgenic mice have indicated that most responding T cells to specific antigens could use various β-chains, but only one α-chain [9–14]. Thus, it is of interest to also determine the role that the TCR α-chain plays in alloantigenic recognition by donor T cells in bone marrow transplantation models, particularly as it relates to the development of GVHD.

We have previously investigated TCR diversity during CD4⁺ T cell alloresponses in murine acute GVHD directed to multiple miHA differences using the B6->BALB.B H2^b-matched strain combination [15–17]. In this lethally irradiated model, donor B6 CD4⁺ T cells alone can cause severe and lethal GVHD characterized by weight loss, diarrhea, and tissue injury in the skin, intestine, and liver [17–19]. The T cell response was analyzed by PCR-based TCR Vβ CDR3-size spectratyping and found to be marked by a limited, yet heterogeneous TRBV usage [20,21]. Among the eleven utilized Vβ families, Vβ2 and Vβ11 were subsequently determined to be most important for the severity of GVHD, in that combined transfer of only these two Vβ families was able to cause a high level of mortality in BALB.B recipients, and their selective deletion from a donor T cell inoculum allowed for significantly improved survival [22]. Further experiments have also indicated that the B6 CD4⁺Vβ11⁺ T cells alone are capable of mediating severe GVHD [23], and sequence analysis of the skewed CDR3-size band in the CD4⁺Vβ11⁺ response identified the dominant clonal specificity involved as utilizing the Jβ2.5 sequence [22]. To focus on the Vα usage, B6 CD4⁺Vβ11⁺ T cells were isolated from the spleens of recipient BALB.B mice undergoing GVHD, and analyzed by the Vα CDR3-size spectratype approach. We found skewing indicative of restricted responses in the Vα6,9,13,14,18, and 22 families [23].

The current study was undertaken to further define TRAV usage in the B6 CD4⁺Vβ11⁺ T cell response against BALB.B miHA. Using the nonpalindromic adaptor-PCR method, we found dominant usage of TRAV13 in combination with the TRAJ16 chain transcript. Then, using laser microdissection (LMD), we found use of the identical TRAV-TRAJ nucleotide sequences in areas dominated by infiltrating CD4⁺Vβ11⁺ T cells during development of GVHD in both the rete-like prominences of the dorsal lingual epithelium and the ileal crypts of the small intestine. These data indicate that there is a very restricted TRAV usage by CD4⁺Vβ11⁺ T cells involved in GVHD in this model.

MATERIALS AND METHODS

Mice

C.B10/LiMcdJ (BALB.B; H2^b) and C57BL/6By (B6; H2^b) mice were purchased from the Jackson laboratory (Bar Harbor, ME). In addition, BALB.B mice were also bred in our own colony. Male mice, 6–12 weeks of age, were always used as donors, whereas both female and male mice, 8–12 weeks of age, were used as recipients. Mice were housed in a pathogen-free environment, housed in autoclaved microisolator cages, and were provided with autoclaved food and water, ad libitum. For lethal irradiation exposure, mice were administered 10 Gy total body irradiation from a Mark I model 68A ¹³⁷Cs source (J.L Shepherd, San Fernando, CA). Use of animals was approved by the Institutional Animal Care and Use Committees of both Thomas Jefferson University and Hackensack University Medical Center.

Media

Lymphocytes, isolated by homogenization of spleens and lymph node (LN), and bone marrow (BM) cells, extracted from femurs, were suspended in phosphate buffered saline (PBS; BioWhittaker, Walkersville, MD) containing 0.1% bovine serum albumin (BSA; Sigma, St. Louis, MO): PBS-BSA. For culturing, lymphocytes were suspended in RPMI 1640 media containing 50U/mL penicillin, 50μg streptomycin sulfate, 2mmol/L L-glutamine (all from Mediatech, Herndon, VA), 10% fetal bovine serum (FBS; Sigma), 50μM 2-mercaptoethanol (Life Technologies, Gaithersburg, MD), and sodium pyruvate (1mM; Fisher Scientific, Fair Lawn, NJ), collectively termed ‘complete media’. PBS-BSA and 0.01% NaN₃ (FC buffer) was used for analysis of cells by flow cytometry and for cell separation. All Ab used for cell subset depletions were resuspended in PBS-BSA.

Preparation of Donor Cells and GVHD

B6 mice were immunized with 2x10⁷ BALB.B splenocytes intraperitoneally (i.p.) and spleens and LN harvested 17–21 days later. The tissues were homogenized and the cells resuspended in Gey’s balanced salt solution containing 0.7% NH₄Cl to lyse erythrocytes. B cells were removed by panning technique on a Petri dish coated with goat anti-mouse IgG antibody (Ab; 1:200 dilution; Cappel-Organon Teknika, West Chester, PA; 1h at 4^oC). The non-adherent enriched T cell population was then incubated with rat IgM anti-CD8 monoclonal Ab (mAb; 3.168; [24]) at 1:100 dilution along with guinea pig complement (1:6 dilution; Rockland, Boyertown, PA) for 45 minutes at 37^oC to enrich for CD4⁺ T cells. BM was extracted from femurs of primed B6 mice, suspended in PBS-BSA, and depleted of T cells by treatment with anti Thy-1 mAb (J1j; [25]) and complement for 45 minutes at 37^oC. The remaining anti-Thy-1-treated BM (ATBM) cells were confirmed to have 0% T cells by flow cytometry. Donor ATBM cells (2x10⁶), along with enriched (97–99%) CD4⁺ T cells (1.7–2x10⁷) were injected in PBS intravenously (i.v.) via the tail vein into lethally irradiated (10 Gy) BALB.B recipients.

Carboxyfluorescein Diacetate Succinimidyl Ester (CFSE) Labeling and Cell Culture

B6 mice were immunized i.p. and boosted with 2x10⁷ BALB.B splenocytes each on days 0, 21, and 42, and spleens and LN harvested on day 63. The cells were enriched for CD4⁺ T cells, as described above, suspended in PBS (2.5x10⁷/mL), and labeled with 1.5umol/L CFSE (Molecular Probes, Eugene, OR) for 10 minutes at 37^oC, as previously described [26]. The CFSE-labeled B6 CD4⁺ T cells were washed in complete media 4x at 4^oC and cultured in 96-well round-bottom plates (2.5x10⁵ cells/well) along with irradiated (30 Gy) T cell-depleted BALB.B splenocytes at a 1:2 ratio, respectively (3 days; 7% CO₂; 37oC). Mimosin (300μM; Sigma) was added to wells of a control group to inhibit cell proliferation.

Preparation of RNA and cDNA

The in vitro BALB.B-stimulated CFSE-labeled B6 CD4⁺ T cells were pooled from the plates, washed 4x with FC buffer, and sorted on a MoFlo high-speed cell sorter (Dako Cytomation, Carpinteria, CA), based on CFSE intensity, CD4, and Vβ11 expression. CFSE⁺CD4⁺Vβ11⁺ cells were collected directly into cell lysis buffer and RNA was extracted according to the manufacturer’s instructions (RNA Easy kit, Qiagen, Valencia, CA). Murine primers for cDNA synthesis and the protocol for the amplification of cDNA were as previously described with some minor variations [27,28]. Briefly, Superscript III 1^st strand synthesis super mix (Invitrogen, Carlsbad, CA) was used with a NotI-Cα3 gene-specific primer for 1^st strand synthesis. Following 2^nd strand synthesis, the cDNA had a nonpalindromic double-stranded adaptor ligated onto both blunt ends by overnight incubation with T4 DNA ligase (1.4U; Invitrogen). NotI enzyme (7.5U; Promega; Madison, WI) was added for restriction digestion for 3 hours at 37^oC. The digested product was placed into a Min Elute PCR Purification column (Qiagen) to eliminate residual traces of enzyme. PCR was performed using a high fidelity Platinum® Pfx DNA Polymerase enzyme (Invitrogen), with the first round using mCα2 and EcoRI-XmnI primers, and followed by 35 rounds of amplification. The PCR products were placed into a Min Elute PCR purification column to remove traces of excess PCR buffers, and then subjected to a second round of PCR using mCα1 and EcoRI-XmnI primers in 35 rounds of amplification. PCR products were run on a 1.5% agarose gel, and bands ~150–700 base pairs (bp) in size were gel-purified using the Gel Purification Kit (Qiagen). Purified products were directly ligated into the pGem T-Easy vector system (Promega), and transformed into DH5α E. coli. White colonies were selected, DNA extracted using a Miniprep Kit (Qiagen), and sequenced on an ABI Prism 377 DNA Sequencer (Applied Biosystems, Foster City, CA).

Immunohistochemistry

Tongue and intestinal tissues were harvested from irradiated BALB.B recipients 15d after transfer of B6 ATBM and host-primed B6 CD4⁺ T cells. Half of the organ samples were collected in 10% phosphate-buffered formalin (Fisher) and embedded in paraffin for routine H&E staining, while the remaining tissues were embedded in optimal cutting temperature (OCT) medium (Miles Laboratories, Elkhart, IN) and snap-frozen in liquid nitrogen. Staining was performed as previously described [29]. Briefly, frozen sections were cut into 5μm thick sections at −20^oC, air-dried overnight, fixed in cold acetone for 5 minutes, and air-dried. Sections were then incubated with PBS containing 5% BSA and 0.1% Tween-20 (PBS-BSA/Tween), followed by goat blocking serum, and stained with fluorescein isothiocyanate (FITC)-labeled anti-murine Vβ11 or Vβ5 mAb, or with FITC-IgG control Ab (all from Pharmingen and diluted 1:50 in PBS-BSA/Tween) for 1 hour at room temperature. Secondary Ab were: biotin-conjugated rat anti-mouse IgG and rat anti-FITC (Vector Labs, Burlingame, CA), all diluted at 1:200 in PBS-BSA/Tween. Slides were developed using a biotin-avidin-horse radish peroxidase (HRP) ABC kit along with peroxidase NovaVector Red (Vector Labs). Sections were counterstained with Mayer’s hematoxylin solution.

LMD and DNA Extraction

Frozen lingual and intestinal sections were cut, Ab-stained, and counterstained as described above, but were placed uncoverslipped on vinyl acetate-coated slides (Leica Microsystems, Bannockburn, IL). The Leica AS LMD instrument was used to target and capture specific areas in the tissues, which were then placed into the caps of 0.2mL tubes containing 10uL 1X PCR buffer and 0.5mg/mL proteinase K (Fisher). Samples were heated for 1 hour at 50^oC, and proteinase K was inactivated at 95^oC for 10 minutes.

PCR Amplification

Primers used for multiplex PCR were designed using the Primo Multiplex 3.4 program (Chang Bioscience, Castro Valley, CA). This program designed primers that did not form primer-primer dimers, and each primer was tested individually before use. DNA extracted from each LMD section was added to a multiplex PCR mixture containing 1X PCR buffer, Platinum® Pfx DNA Polymerase enzyme (Invitrogen), 10mM MgCl₂, 20μM of each primer, and a total of 800μM dNTPs (Invitrogen; 200μM of each dATP, dTTP, dCTP, dGTP) in a 30uL volume. DNA was amplified for 40 cycles using the program: 94^oC for 5 minutes, 60^oC for 1 minute, 72^oC for 1 minute, and 72^oC for 4 minutes. Multiplexed PCR products were run on a 3% agarose gel, gel extracted, and then cloned into the pGem T-Easy vector (Promega). DNA was extracted from white colonies and sequenced, as described above. Sequences were compared to sequences found in the IMGT database.

Primers, 5’ to 3’:

TRAV13-TRAJ16 (69bp); Forward: ATCACAGACTCAGGCACTTATC, Reverse: CCCTGGCCAAAAACCAGCTTC

TRAV13-TRAJ17 (79bp), Forward: GGCACTTATCTCTGTGCCCTC, Reverse: AGCACCCTGGTTCCGATTCC

CD4 (285bp): Forward: AGGGGCCACCACTTGAACTAC, Reverse: AGCAGATTGTCCAGGGACACC

IMGT Database and Sequence Alignments

The IMGT database and tools (IMGT/V-QUEST) (http://imgt.cines.fr/) were used to align TRAV and TRAJ sequences found in the CFSE and LMD experiments [30–33]. CLUSTALW multiple sequence alignments were done using the online Biology Workbench (http://workbench.sdsc.edu) [34,35].

RESULTS

Identification of Vβ11⁺ T Cells in GVHD Target Organs

In order to examine the TRAV utilization within a GVHD-relevant TRBV CD4⁺ T cell family, a focus was placed on the Vβ11 family which was previously found to be associated with severe GVHD development in the miHA-disparate B6->BALB.B irradiated strain combination [22]. Pathological analyses using this model had indicated that the BALB.B intestine, particularly the distal ileum, was targeted by donor B6 CD4⁺ T cells [20]. In addition, in the dorsal lingual epithelium of the murine tongue, the primary targets of T cell-mediated injury during GVHD were found to be cytokeratin 15⁺ epithelial cells residing in structures called rete-like prominences (RLP; [36,37]) that resemble human skin rete ridges [38]. Initial experiments were conducted to determine if CD4⁺Vβ11⁺ T cells infiltrated the gut and lingual target tissues of GVHD, and whether their level of involvement varied. GVHD was induced by i.v. transfer of 2x10⁷ B6 CD4⁺ T cells and 2x10⁶ ATBM cells into lethally irradiated (10 Gy) BALB.B recipients. At days 10 or 15 post-transplant, recipients that received B6 ATBM cells did not exhibit any histological signs of GVHD (e.g., mononuclear infiltration in either the lingual or intestinal tissues [data not shown]). In contrast, recipients of B6 CD4⁺ T cells exhibited cellular infiltration in dorsal lingual epithelium by day 10 (Figure 1A) and expanded infiltration on day 15 (Figure 1B). The ileal crypts of intestinal tissues exhibited a low level of infiltration on day 10 (Figure 2A), but the infiltration was more evident by day 15 (Figure 2B). Staining with Vβ11-specific mAb resulted in detection of numerous infiltrating cells within RLP in the lingual tissues at day 10 (>50% of the mononuclear cell component; Figure 1C) and increasing numbers at day 15 (Figure 1D). In a similar pattern, the intestinal tissues exhibited some Vβ11⁺ T cell infiltration on day 10 (Figure 2C) and an increased amount on day 15 (Figure 2D). Since the Vβ5 family was never found to be involved in GVHD development in this model at any time point, Vβ5 mAb staining was used as the control for non-specific T cell infiltration [22]. Vβ5⁺ cells were found neither in the lingual (inset, Figure 1D) or intestinal (inset, Figure 2D) sections, suggesting that there was selective T cell infiltration into these tissues. Furthermore, sections from tongue (Figures 1E,F) and intestine (Figures 2E,F) of irradiated syngeneic B6 recipients of B6 CD4⁺ T cells failed to demonstrate Vβ11⁺ T cell infiltration at either day 10 or day 15. Taken together, these data indicate that Vβ11⁺ T cells are selectively infiltrating both lingual and intestinal tissues during the GVHD response.

Figure 1.

Evolution of B6->BALB.B GVHD in lingual epithelium, days 10 and 15. On day 10, H&E histology reveals lymphocytes within the superficial submucosa and epithelium (A); one can easily differentiate the small, somewhat hyperchromatic lymphocyte nuclei from the larger, more open nuclei of the squamous mucosal or columnar epithelial target cells they are infiltrating. The lymphocytes increase in number by day 15 (B), when they become more concentrated at the rete-like ridges and associated with target cell apoptosis (arrow; inset is B6->B6 d15 control). Immunohistochemistry demonstrates that >90% of these lymphocytes express Vβ11 (C&D); note the tendency of these cells to associate with target cells in the basal layer by day 15 (arrow in D; inset represents negative staining for Vβ5 in adjacent section). B6->B6 control tongue failed to show Vβ11 infiltration at day 10 (E) and day 15 (F).

Figure 2.

Evolution of B6->BALB.B GVHD in intestinal epithelium, days 10 and 15. Pathological changes are not apparent at day 10 (A) but are noted by day 15 (B), consisting of lymphoid infiltration and apoptosis of crypt epithelium (arrows). B, inset depicts day 15 B6 -> B6 control. Rare Vβ11⁺ cells are present at d10 (C, arrow), whereas numerous positive cells are present at day 15 (D; inset is staining for Vβ5 in adjacent section). B6->B6 control gut fails to show significant infiltration by Vβ11-positive cells at either day 10 or 15 (E&F, respectively).

Sequencing Sorted Divided CD4⁺Vβ11⁺ T Cells in the B6->BALB.B Response

The identification of Vβ11⁺ T cells within the lingual and intestinal infiltrates of BALB.B recipients allowed us to investigate the corresponding TRAV diversity in this physiologically relevant TRBV population. The Vβ11⁺ T cell family constituted less than 2.8% of the naïve B6 CD4⁺ T cell population [39]. To specifically expand the numbers of anti-BALB.B miHA-specific B6 T cells, B6 mice were serially-sensitized with BALB.B splenocytes. At the end of the sensitization period (63 days), B6 CD4⁺ T cells were isolated from pooled spleen and LN of the sensitiz ed mice, labeled with CFSE to allow detection of dividing cells [20,26], and cultured with irradiated (15 Gy) T cell-depleted BALB.B splenocytes. After 3 days of incubation, CFSE^low/negative CD4⁺Vβ11⁺ T cells were sorted by flow cytometry from the culture to isolate the specific cells that had divided in response to the miHA stimulation (Figure 3). A significant number of the primed Vβ11⁺ T cells appeared to have proliferated rapidly, as might be expected for memory T cell reactivation.

Figure 3.

Isolation of responding B6 CD4⁺Vβ11⁺ T cells. B6 mice were serially-immunized with BALB.B splenocytes (2x10⁷, i.p.) on days 0, 17, and 42. On day 63, spleen and LN CD4⁺ T cells from these mice were labeled with CFSE and cultured for 3 days with either L-mimosin to distinguish non dividing cells (upper panel) or irradiated BALB.B splenocytes that were depleted of all T cells (lower panel). Histograms, gated on CFSE⁺CD4⁺Vβ11⁺ T cells, indicated that most of the allostimulated cells underwent division, marked by a decrease in CFSE intensity.

To determine the TRAV nucleotide sequence diversity of the sorted CFSE^low/negative CD4⁺Vβ11⁺ T cells, nonpalindromic adaptor ligation and PCR (NPA-PCR) amplification was performed [27,28]. This technique allowed for amplification of TRAV mRNA with unknown 5’ ends. Since there is so much variability with TRAV sequences between and within haplotypes, subtle subfamily differences could be seen at the nucleotide level [4,40]. NPA-PCR allowed for a determination of TRAV nucleotide sequences in the responding population without skewing results by selecting for particular α-chain sequences or subfamilies. Using semi-nested PCR, products were amplified ranging from ~150–700bp and those bands were cloned into a vector. Many of the bands were faint and could not be visually resolved, but it is important to note that the brightest bands were found to be ~200bp in size, suggesting that most of the TRAV sequences were in this band size. Once the cloned sequences were analyzed and aligned using a CLUSTALW program, it was evident that there was TRAV nucleotide conservation between the 12 sequences (Figure 4A). When the sequences were separated and aligned with those that were similar and in-frame, it was found that they utilized TRAV13 along with two dominant TRAJ sequences. Five in-frame sequences contained TRAJ16 (Figure 4B) and five other out-of-frame sequences contained TRAJ17. Sequences containing TRAJ17 had 13 N-region nucleotide additions, resulting in an out-of-frame sequence. Sequences containing TRAJ16 were in-frame and contained a N-region that contained a tryptophan (W) surrounded by two alanines (A). The other two sequences contained TRAJ43 and TRAJ16 regions, but their TRAV regions could not be aligned due to the sequence being out-of-frame or there being no apparent TRAV region, respectively. Since neither TRAV-specific mAb nor strain-specific primers had been used for amplification, the TRAV sequence results were entirely representative of the anti-BALB.B miHA-specific B6 CFSE^{low/ negative} CD4⁺Vβ11⁺ T cells.

Figure 4.

Alignment of TRAV sequences isolated from CFSE-labeled B6 CD4⁺Vβ11⁺ T cells. Following RNA extraction from CFSE^low-negative CD4⁺Vβ11⁺ T cells, cDNA was enriched for TCRα chain transcripts using a NotI-Cα3 primer, and subsequently amplified using NPA-PCR. Products were visualized on an agarose gel, gel purified, then cloned into a pGEM T-Easy vector. DNA sequences isolated from bacteria colonies and sequences were aligned for similarity using the CLUSTALW program. All sequences were compared to those found in the IMGT database using the IMGT/V-QUEST program. (A) Sequence alignment of all sequences; in-frame and out-of-frame. All sequences were compared against CD-D16 which was used as a reference sequence that the other sequences were compared to for similarity or differences. Hyphens represent nucleotide identity and dots represent nucleotide gaps. Nucleotides that differed from the reference sequence have the letters written in the areas that differed. CD-D16, CD-D14, CD-D2, CD-D6, and CD-D10 contained the sequences for TRAV13-TRAJ16. CD-D12, CD-D37 CD-D1, CD-D3, and CD-D4 contained the sequences with TRAJ17. CD-D9 contained an out-of-frame sequence for TRAV13-TRAJ43. CD-D13 was an incomplete sequence and only contained a TRAJ and a TRAC region. (B) TRAV13-TRAJ16. Note that the amino acids are included above the nucleotide sequences in the alignment. TRAV, N (N-region additions), TRAJ, and TRAC regions are indicated above nucleotides.

Extraction of Tissue-Infiltrating B6 CD4⁺Vβ11⁺ T Cells by LMD

The restricted use of TRAV13 sequence by the in vitro anti-BALB.B responding B6 Vβ11⁺ CD4⁺ T cells was consistent with recognition of a limited number of miHA. In order to establish biological significance of our analysis, we performed LMD [1] to excise restricted areas within BALB.B target tissues to determine whether the restricted TRAV sequences were found at sites of CD4 infiltration and concomitant GVHD-associated tissue damage. Taking the day 15 post-transplantation frozen tissue sections, previously described in Figures 1 & 2, ~1mm linear sections were excised from areas that exhibited significant staining with the Vβ11-specific mAb, particularly within the RLP of the dorsal lingual epithelium (Figures 5A,B) and surrounding the ileal crypts of the small intestine (Figures 5C,D). We also performed LMD on similar tissue samples from irradiated B6 mice that had only received B6 ATBM, to serve as a negative control for any potential non-specific infiltration of cells associated with the irradiation conditioning of the recipients (Figures 5E-H).

Figure 5.

LMD targeting of B6 Vβ11⁺ T cells in BALB.B organs. B6 CD4⁺ T cells (2x10⁷) and ATBM (2x10⁶) were transferred into lethally irradiated (10 Gy) BALB.B or control B6 recipients. Immunohistochemistry of uncoverslipped tissue sections stained with Vβ11 antibody and counterstained with hematoxylin. (A) LMD-targeted areas of lingual epithelium RLP before and (B) after LMD. (C) Targeted areas of small intestines before and (D) after LMD. Control mice were lethally irradiated (10 Gy) and injected with 2x10⁶ B6 ATBM. (E) RLP of control mice before and (F) after LMD. (G) Targeted intestinal areas of control mice before and (H) after LMD. Note the areas that have been extracted by LMD. The pictures are from one mouse, and are representative of two separate experiments, each with n=3.

PCR Amplification of Restricted B6 TRAV Sequences at the Sites of GVHD Tissue Damage

DNA was extracted from the LMD samples, and multiplex PCR performed with primers that were specifically designed to amplify genomic CD4 along with either the TRAV13-TRAJ16 or TRAV13-TRAV17 segments that were previously found in the in vitro experiment. This approach was taken since TRAV-TRAJ rearrangement would have already taken place during T cell development in the thymus, and due to the distance between the TRAJ and constant regions, which would have made direct cloning of such large fragments difficult. In addition, CD4 primers were used as an internal PCR control. Since the primers for CD4 were meant to amplify genomic DNA, CD4 was expected to be present in all the samples analyzed. Each of the LMD samples from individual mice showed the presence of the control CD4 band along with either TRAV13-TRAJ16 or TRAV13-TRAJ17 (data not shown) present in both the lingual (Figures 6A) and intestinal (Figures 6B) samples. The ATBM control samples from these tissues only amplified CD4 (Figures 6C,D), indicating that the three identified TRAV sequences were specifically involved in the CD4⁺ cellular infiltration of the GVHD target tissues. In all, these tissue-specific sequence results for TRAV-TRAJ utilization were completely consistent with those found in the isolated divided CD4⁺Vβ11⁺ T cells in the in vitro CFSE experiment. Of lesser note, although the TRAV13-TRAJ17 sequence was always out-of-frame and could never be expressed as a protein, it was still specifically found in B6 T cells that were infiltrating BALB.B tissues, alongside the in-frame TRAV13-TRAJ16 sequence.

Figure 6.

PCR amplification of TRAVs isolated by LMD. Amplified products were visualized on 3% agarose gels. CD4 (285bp) and TRAV13-TRAJ16 (69bp)-specific primers were used to amplify DNA from LMD-extracted lingual (A,C) and intestinal epithelium (B,D) 15 days following transplant of either 2x10⁷ B6 CD4⁺ T cells and 2x10⁶ ATBM into BALB.B recipients (10 Gy) (A,B) or 2x10⁶ B6 ATBM into B6 recipients (10 Gy) (C,D). For each gel, lane 1 is a 100bp ladder and lanes 2–4 are each amplified products from a single mouse, and are representative of two separate experiments (n=6).

DISCUSSION

Despite the availability of numerous miHA expressed in the host, the donor B6 CD4⁺ T cell alloresponse is limited in its TCR Vβ diversity following transplantation into BALB.B recipients [39]. Subsequent investigation revealed that there was a correlation between the Vβ11 and/or Vβ2 T cell family responses and the severity of acute GVHD [20–22]. The findings presented in the current study confirm the ability of B6 CD4⁺Vβ11⁺ T cells to infiltrate both lingual and intestinal target tissue during the course of GVHD and the restricted use of TRAV13-TRAJ16 during the in vitro proliferative response of these specific T cells against BALB.B miHA. LMD techniques were then employed for in situ analysis of infiltrating B6 CD4⁺Vβ11⁺ T cells to establish that the in vitro identified TRAV sequence was relevant to the GVHD-associated immunopathological response.

Infiltrating Vβ11⁺ T cells were found in both the lingual and intestinal tissues at days 10 and 15 post-transplantation. The latter observation is of interest because it is consistent with the known association of severe GVHD induction by Vβ11⁺ and/or Vβ2⁺ T cells in BALB.B, and not CXB-2 recombinant inbred mice, with more detectable GVHD-related intestinal pathology [20]. The presence of CD4⁺Vβ11⁺ T cells with the identical TRAV sequence in both the RLP regions of the tongue and the ileal crypts of the intestine in the BALB.B recipients strongly suggests that the miHA being recognized by these specific effector cells are expressed in both target tissues. However, previous studies showed differential tissue damage mediated by B6 donor CD4⁺ T cells was only seen in the intestine of BALB.B and not CXB-2 mice [20]. One possibility to explain this dichotomy is that BALB.B mice uniquely express the relevant Vβ11-associated miHA in the intestine, either qualitatively or quantitatively, and this presentation induces a more severe inflammatory response that directly impacts on the health status of the recipient. In contrast, inflammation in the squamous mucosa of the tongue, and by extension to the general skin surface, probably has less impact on the overall health and survival of the recipient. To extend our previous repertoire analysis of B6 CD4⁺ T cells associated with GVHD responses in BALB.B mice to the TCRα-chain, we sequenced responding TRAV in the CD4⁺Vβ11⁺ T cell response. NPA-PCR allowed us to enrich for α-chain transcripts and to sequence TRAV found within the proliferating CFSE-diminished CD4⁺Vβ11⁺ T cells without creating primers for all the known TRAV. We saw a limited TRAV presence within this population of responding cells. Sequence alignments of these TRAV showed a restricted use of TRAV13 combined with TRAJ16. The ability to clone different sequences, including some of which were out of a reading frame (TRAJ17), indicated that our analysis was not being done on one TRAV that was present in the PCR mixture and continuously amplified. The restricted use of TRAV corroborates the findings in other antigenic systems [8–14,41]. It is also worth noting that unlike the TCRβ-chain, there is no allelic exclusion for the α-chain, so both loci can rearrange in a developing T cell in the thymus, and thus two TCRα-chain proteins can be expressed on the surface of a T cell with a single TCRβ-chain [42,43]. This possibility serves to emphasize the uniqueness of the TRAV13-TRAJ16 rearrangement as representing the only α-chain that appears to interact with the as yet unknown BALB.B MHC class II-restricted miHA.

What appears to be a single uncharacterized miHA, able to elicit severe GVHD mediated by the CD4⁺Vβ11⁺ T cell response, may be the first to be observed as such in murine models. Its existence had been previously hypothesized after a series of inter-strain GVHD studies in the B6->BALB.B model and between members of the closely related recombinant inbred CXB strains [18]. In one other instance, in an investigation by Miconnet et al., the CD4⁺ Vβ response of donor T cells was identified in a strain combination that involved both miHA and strong Mtv superantigen stimulation, but the contribution of miHA themselves was not clear [44]. Although previous studies that analyzed TRAV diversity, using transgenic mice that expressed a single TCRα-chain, had the benefit of knowing the antigenic peptide along with both the α- and β-chains, our CFSE studies circumvented a need to know the specific miHA being recognized by the CD4⁺Vβ11⁺ T cells. It is commonly held that alterations in MHC type in the host can have profound effects on the grade, target organs, and phenotype of GVHD [45]. Since MHC can select a TCR repertoire by positive and negative selection in the thymus and present different types of peptides in the periphery, it seems quite reasonable to consider that the severity of GVHD in MHC-matched situations is due to the types and expression levels of miHA peptides that specific MHC can display [46]. Therefore, an immune response’s dependence upon TRBV versus TRAV restriction would seem to depend upon the nature of their target antigens. In studies where target antigens were known, varying degrees of TCRα and TCRβ dependence were observed. Some systems relied upon TCRα conservation while others equally relied upon TCRα and TCRβ [9,11,12,41,47–49]. In comparison, in our B6->BALB.B CD4⁺-mediated GVHD model, with as yet unidentified miHA, we focused on the seemingly important CD4⁺Vβ11⁺ T cell responses to examine TRAV usage. The finding of a single restricted TRAV association suggests that these specific T cells are recognizing only a single miHA and that TRAV interaction is important. This observation of a restricted TRAV usage is also consistent with a study in a CD8⁺ T cell model of GVHD that suggested miHA recognition was dependant upon peptide contact with the TCRα-chain [50].

Without having a full repertoire of murine TRAV antibodies available, we could not definitively determine whether the in vitro identified and sequenced TRAVs from the CFSE-labeled CD4⁺Vβ11⁺ T cells were actually expressed on the cell surface. In order to provide evidence that the observed TRAV restriction was relevant to the in vivo GVHD situation, at sites of tissue destruction, we isolated tissue sections containing the infiltrating Vβ11⁺ T cells by use of LMD technology. The exact TRAV13-TRAJ16 sequence was found within the lingual and intestinal GVHD-afflicted BALB.B tissues. The sequence for the out-of-frame TRAV13-TRAJ17 rearrangement was also found in equal proportion, as in the CFSE experiment. This could imply that there may be a preference for TRAV13 to combine with TRAJ16 or 17. The two TRAJ gene segments are situated adjacent to one another on their locus and there have been reports showing that TRAV have a preference for certain TRAJ, depending on their distance or accessibility to one another [51–54]. It should also be noted that although NPA-PCR allowed us to sequence α-chains without knowing the variable region, this method did not allow us to sequence the entire TRAV region, so we could not entirely distinguish between TRAV subfamilies. It is certain that the one, prevalent, in-frame TRAV sequence was TRAV13, but since subfamilies have over 75% homology, further identification was difficult. The apparent absence of the other five Vα families in the CFSE-reactive population, previously observed to be skewed in the spleens of BALB.B mice undergoing GVHD, could be due to the absence of those specificities in the target tissue infiltrate and/or the result of multiple presensitization of the donor B6 mice with BALB.B splenocytes, focusing the T cell response towards a more dominant miHA.

Considering the immense diversity found within the repertoire of TCRα-chains, the restricted response that was identified during the B6 CD4⁺Vβ11⁺ T cell response implies that a strong selective pressure may be activating a limited number of T cells expressing particular α and β-chains. We postulate that this selective pressure may be in the form of a limited number of immunodominant miHA that elicit the B6-mediated GVHD reaction and are responsible for the pathological severity in the BALB.B recipients [17,45,55–57]. This present study extends our understanding of this limited T cell repertoire response to the TCR α-chain, as well.