Skewed T cell receptor repertoire of Vδ1⁺ γδ T lymphocytes after human allogeneic haematopoietic stem cell transplantation and the potential role for Epstein–Barr virus-infected B cells in clonal restriction

Abstract

The proliferation of Vδ1⁺ γδ T lymphocytes has been described in various infections including human immunodeficiency virus (HIV), cytomegalovirus (CMV) and malaria. However, the antigen specificity and functions of the human Vδ1⁺ T cells remain obscure. We sought to explore the biological role for this T cell subset by investigating the reconstitution of T cell receptor (TCR) repertoires of Vδ1⁺ γδ T lymphocytes after human allogeneic haematopoietic stem cell transplantation (HSCT). We observed skewed TCR repertoires of the Vδ1⁺ T cells in 27 of 44 post-transplant patients. Only one patient developed EBV-associated post-transplant lymphoproliferative disorder in the present patient cohort. The -WGI- amino acid motif was observed in CDR3 of clonally expanded Vδ1⁺ T cells in half the patients. A skew was also detected in certain healthy donors, and the Vδ1⁺ T cell clone derived from the donor mature T cell pool persisted in the recipient's blood even 10 years after transplant. This T cell clone expanded in vitro against stimulation with autologous EBV–lymphoblastoid cell lines (LCL), and the Vδ1⁺ T cell line expanded in vitro from the same patient showed cytotoxicity against autologous EBV–LCL. EBV-infected cells could also induce in vitro oligoclonal expansions of autologous Vδ1⁺ T cells from healthy EBV-seropositive individuals. These results suggest that human Vδ1⁺ T cells have a TCR repertoire against EBV-infected B cells and may play a role in protecting recipients of allogeneic HSCT from EBV-associated disease.

Introduction

Immune reconstitution is an essential component of allogeneic haematopoietic stem cell transplantation (HSCT) for the treatment of haematological malignancies, severe aplastic anaemia, congenital immunodeficiency and metabolic inborn errors. Post-transplant reconstitution of the αβ T cell receptor (TCR) repertoire correlates with the control of viral infections [1]. We have demonstrated previously that recipients of allogeneic HSCT have a skewing of the αβ TCR repertoire after transplant and this is a result of the proliferation of donor-derived mature T lymphocytes, which can survive in recipients for several years [2–5]. Expansion of the population of donor-derived TCRαβ⁺ T clones from a cytomegalovirus (CMV)-seropositive bone marrow donor was observed in a CMV-seropositive recipient, who developed CMV-associated disease after transplantation [5]. These findings suggest that the clonal expansion of TCRαβ⁺ T cells and resulting skewed αβ TCR repertoire in peripheral blood following allogeneic HSCT reflect primarily the response to microorganisms.

Reconstitution of the diversity of human γδ T cells depends upon the expression of Vδ segments. Vγ2/Vδ2 (formerly Vγ9/Vδ2) T lymphocytes account for the majority of circulating blood γδ T cells [6], and reconstitution of the repertoire diversity of this subset is achieved during the early phase of transplantation [7]. In contrast, our previous study using a limited patient cohort demonstrated that some had a skew of the Vδ1⁺ TCR γδ⁺ T cell repertoire after transplantation, that clonal predominance of Vδ1⁺ TCR γδ⁺ T cells was already present in blood of some donors and that those clones were transferred and still existing 1·5 years after transplantation [7].

The proliferation of Vδ1⁺ γδ T lymphocytes has been described in certain human diseases including rheumatoid arthritis, multiple sclerosis and various infections such as HIV [8–10], CMV [11–13] and malaria [14]. Although the antigen specificity and functions of the Vδ1⁺ γδ T lymphocyte population expanded in vivo remain obscure, the findings described above have raised the hypothesis that Vδ1⁺ T lymphocytes may have a TCR repertoire and play a role in protection against microorganisms which can cause latent infection observed commonly in the human being. In the present study, we have sought to explore the antigen specificity and biological functions of the Vδ1⁺ γδ T lymphocytes in humans by extending our observation on the reconstitution of Vδ1⁺ TCR γδ⁺ T cell repertoire after human allogeneic HSCT and the in vitro responses of Vδ1⁺ γδ T lymphocytes against autologous Epstein–Barr virus (EBV)-infected B cells.

Materials and methods

Patients

Forty-four recipients of allogeneic haematopoietic stem cell grafts were included in this study (Table 1). Informed consent was obtained from patients and donors before blood samples were taken. The ethical committee of our institution approved the experimental protocol. All but two patients were conditioned with myeloablative chemoradiotherapy, consisting mainly of fractionated total body irradiation (12 Gy in six fractions) and cyclophosphamide (60 mg/kg/day for 2 days), followed by infusion of allogeneic marrow or blood stem cell grafts from serologically human leucocyte antigen (HLA)-matched sibling or unrelated donors. All patients received cyclosporin A and short-term methotrexate for the prophylaxis of acute graft-versus-host disease (GVHD). Engraftment was achieved in all patients. Donor chimerism of the myeloid lineage was examined after engraftment by identifying the donor-derived sex-chromosomes in bone marrow cells or typing the red blood cells if there was a sex mismatch or ABO incompatibility between the donor and recipient pair, and all testable recipients in the present patient cohort showed full donor chimerism. Donor chimerism of the lymphoid lineage was not determined. Patients were given oral acyclovir for the prophylaxis of herpes simplex virus infection during the first month of transplantation, and were monitored for CMV infection by weekly CMV antigenaemia assays from the time when the granulocyte count reached 500/µl until day 100 after transplantation. Patients who tested positive in the CMV antigenaemia assay received prophylactic ganciclovir (5 mg/kg/day, 3 days a week) from the time when the granulocyte count was greater than 1000/µl.

Table 1.

Patient characteristics (n = 44).

	No. of patients
Age (years)
Median	30
Range	16–58
Sex
Male	22
Female	22
Disease
AML	15
ALL/LBL	16
CML	8
MDS	3
Atypical CML	1
Aplastic anaemia	1
Graft
Bone marrow	35
Blood stem cell	7
Cord blood	2
Outcome
Alive	29
Dead	15
Observation period, months
Alive	9–112
Dead	2–43

AML, acute myelogenous leukaemia; ALL, acute lymphoblastic leukaemia; LBL, lymphoblastic lymphoma; CML, chronic myelogenous leukaemia; MDS, myelodysplastic syndrome.

CDR3 size spectratyping of TCRDV

The size of CDR3 of TCR δ-chains was determined as described elsewhere [15]. Total RNA was extracted from peripheral blood mononuclear cells (PBMCs) of recipients using an RNeasy Total RNA Kit (Qiagen, Hilden, Germany) and was used for first-strand cDNA synthesis with an oligo-dT primer (First-Strand cDNA Synthesis Kit; Amersham Pharmacia Biotech, Uppsala, Sweden). In some experiments, mononuclear cells were isolated from the grafts. Aliquots of the cDNA were amplified with a Vδ1-specific oligonucleotide (designated VD1) and a Cδ primer (CD1). Polymerase chain reaction (PCR) amplification was performed for 40 cycles in a 20-µl reaction mixture containing 0·2 µM of each primer and 0·5 units of Taq polymerase (TaKaRa, Osaka, Japan). Conditions for the PCR on the thermal cycler were as follows: denaturation at 94°C for 1 min, annealing at 55°C for 1 min and extension at 72°C for 1·5 min. Following the 40 cycles of PCR, an additional extension at 72°C for 15 min was performed.

Aliquots (4 µl) of the unlabelled Vδ–Cδ PCR products were subjected to one cycle of elongation (run-off) with a FAM-labelled Cδ primer (FAM-CD3). The run-off reaction conditions were as follows: denaturation at 94°C for 2 min, annealing at 55°C for 1 min and extension at 72°C for 15 min. The oligonucleotide sequences of the VD1, CD1 and FAM-CD3 primers were as follows: VD1, GTGGTCGCTATTCTGTCAACT; CD1, AACAGCATTCGTAGCCCAAGCAC; FAM-CD3, FAM-GTTTATGGCAGCTCTTTGAAGGT [16]. The labelled PCR products were electrophoresed on acrylamide sequencing gels for the determination of size and fluorescence intensity in an automated DNA sequencer (ABI 377; Perkin-Elmer Applied Biosystems, Foster, CA, USA), followed by analysis using GeneScan software (Perkin-Elmer Applied Biosystems). The results were depicted as peaks corresponding to the intensity of the fluorescence. CDR3 size patterns that failed to exhibit a bell-shaped distribution due to the appearance of predominant peaks with twice higher intensity compared to other peaks with a reduced peak number (< 10 peaks) were judged to be skewed. This judgement was made by three different investigators to minimize interindividual differences.

Sequencing of CDR3 in the TCR-δ chain

PCR products of the TCR-δ chain were cloned into the PCR2·1 TA cloning vector (Invitrogen, Carlsbad, CA, USA) and were sequenced using a Big-Dye Terminator Cycle Sequencing Kit (Perkin-Elmer Applied Biosystems). Sequence analysis was performed using an Applied Biosystems 377 A automated DNA sequencer.

Flow cytometry

Cells were harvested after the indicated time periods and were analysed using flow cytometry (FACSCalibur; Becton Dickinson, San Jose, CA, USA). The monoclonal antibodies used in this study were as follows: anti-CD3 (SK7, IgG1; Becton Dickinson); anti-TCR-γ/δ-1 (Becton Dickinson); anti-Vδ1 (R9·12, IgG1; Immunotech, Marseille, France); anti-Vδ2 (Immu389, IgG1; Immunotech); anti-Vδ3 (P11·5B; Immunotech) and control mouse IgG (×40, IgG1; Dako, Glostrup, Denmurk).

Cell culture

PBMCs were suspended in RPMI-1640 medium containing 10% heat-inactivated autologous serum and were stimulated with irradiated autologous EBV-transformed B cells or allogeneic Burkitt lymphoma Raji cells. Human recombinant interleukin (IL)-2 (Roche, Mannheim, Germany) was added to the culture at the final concentration of 10 U/ml every 3 days after 14 days of culture. In some experiments, PBMCs were stimulated with various antigens of microorganisms such as the extracts from heat-killed bacteria [17], including Escherichia coli, Pseudomonas aeruginosa, Haemophilus influenzae, Streptococcus pneumoniae and Staphyllococcus. aureus, the extract of cytomegalovirus AD-169-infected cells (Chemicon, Temecula, CA, USA), cell wall-derived mannan of Candida albicans (Takara, Otsu, Japan) and lipopolysaccharide O55:B5 (List Biological Laboratories, Campbell, CA, USA).

Establishment of Vδ1 γδ T cell lines and clones

PBMCs were cultured with irradiated autologous EBV–lymphoblastoid cell lines (LCL) in the presence of IL-2 (10 U/ml) for 14 days, and then γδ T lymphocytes were isolated by positive selection with a magnetic affinity cell sorter (MACS) TCRγ/δ MicroBead kit (Miltenyi Biotec, Auburn, CA, USA). Positively selected γδ T lymphocytes were cultured with irradiated autologous EBV–LCL and allogeneic PBMCs as feeder cells, phytohaemagglutinin (PHA) (1 µg/ml) and IL-2 (10 U/ml) for 7 days. Vδ1 T cells were enriched by staining with fluorescein isothiocyanate (FITC)-conjugated anti-TCRαβ, TCRVδ2 and TCRVδ3 monoclonal antibodies (MoAbs), followed by depletion of αβ T cells, Vδ2⁺ and Vδ⁺ T cells with an anti-FITC MicroBead kit and LD columns (Miltenyi Biotec). In some experiments, Vδ1 T cell clones were established by limiting dilution.

Cytotoxicity assay

Cytotoxicity was measured by standard 4-h ⁵¹Cr release assay. Then 5 × 10^{3 51}Cr-labelled target cells were incubated with indicated numbers of effector cells in triplicate wells in 96-well round-bottomed microtitre plates. Maximum release was obtained by replacing the effectors with 1% Triton X. The results were expressed as the percentage of maximum chromium release after subtraction of the spontaneous release. Spontaneous release was always below 15% of maximum release.

Results

Restriction of Vδ1-expressing T lymphocyte repertoires in recipients of allogeneic haematopoietic stem cell grafts

Previously, we observed that recipients of allogeneic haematopoietic stem cell grafts had a skewed Vδ1⁺ TCR repertoire during the early phase of transplantation based on CDR3 size spectratyping in a limited number of patients [7]. We extended our study to learn how often restriction of Vδ1⁺ T cell repertoires would occur after allogeneic HSCT. CDR3 size spectratyping demonstrated that 27 of 44 patients had a highly skewed TCR repertoire of Vδ1⁺ TCR (Fig. 1). There was no significant association between Vδ1 TCR clonality and any of the clinical parameters including the source of grafts, GVHD, virus-associated pneumonia, serological status for CMV and EBV, CMV antigenaemia and outcome (Table 2).

Fig. 1.

CDR3 size distribution patterns of Vδ1–Cδ transcripts of blood γδ T lymphocytes from recipients of allogeneic haematopoietic stem cell grafts. Results of 12 representative patients are shown. Patients UPN-BM20, BM22, BM24, BM25, BM26, BM29, PB06 and CB01 showed skewed Vδ1 T cell receptor repertoires. BM, PB and CB indicate the source of grafts. BM: bone marrow; PB: peripheral blood stem cell; CB: cord blood.

Table 2.

Clonality of Vδ1 T cell receptor (TCR) repertoire and clinical parameters.

	Clonality of Vδ1 TCR
	Oligoclonal	Polyclonal	P-value*
Age (years)
16–29	11	11
30 or over	16	6	0·216
Graft
Bone marrow	23	12
Blood stem cell	2	5	0·160
Cord blood	2	0	0·817
Acute GVHD
Grade 0	18	10
Grades I–IV	9	7	0·838
Chronic GVHD
No	16	8
Yes	5	6	0·414
Virus-associated pneumonia
No	25	14
Yes	2	3	0·579
CMV serology
Negative	2	3
Positive	24	14	0·611
CMV antigenaemia assay
Negative	7	8
Positive	16	8	0·368
EBV serology
Negative	0	0
Positive	14	10	n.d.
Disease progression
No	20	10
Yes	7	7	0·468
Outcome
Alive	19	10
Dead	8	7	0·645

^*P-values were calculated by the χ² test (Yates). Blood stem cell and cord blood grafts were compared to bone marrow. Thirty-five patients were evaluable for chronic graft-versus-host disease (GVHD). Forty-three and 24 patients were evaluable for serological status for cytomegalovirus (CMV) and Epstein–Barr virus (EBV) before transplantation, respectively. The data on CMV antigenaemia assay were available for 39 patients.

Oligoclonal expansions of Vδ1⁺ T cells were confirmed by DNA sequencing of CDR3 in 17 patients with highly skewed repertoires of Vδ1⁺ T lymphocytes (Table 3). There was no obvious sequence homology in CDR3 among recipients. However, it is intriguing that the -WGI- amino acid sequence was observed in this region of TCR in clonally predominant Vδ1⁺ T cells in nine of 17 patients. The -GGY- and -YWG- sequences were recurrent in four and five patients, respectively. The -WGI- and -GGY- sequences are encoded by the TCRDD3 segment (http://imgt.cines.fr). The selective growth of these -WGI- containing Vδ1 clones in different individuals suggests that a common antigenic pressure may shape the Vδ1 TCR repertoire, although the -WGI- motif was not borne by the most prominently expanded clones, with the exception of BM08, BM09, BM24 and CBT01 patients (Table 3).

Table 3.

Clonal restriction of Vδ1-expressing T lymphocytes in recipients of allogeneic haematopoietic stem cell grafts.

Patient	Time of analysis*	Vd1	N-D-N	Jd1	Colony frequency†
BM01	8 M	CALG	KSLPKYWGRPF	YTDKLIFGKG	5/13
		CALGE	GIPYYAGGYPF	TDKLIFGKG	2/13
BM04	11 M	CALGE	ETRGPLDVES	DKLIFGKG	9/15
		CALGE	PIFPTGTGGYDP	YTDKLIFGKG	6/15
BM08	3 M	CALGE	LPLFPHLYWGISRA	DKLIFGKG	8/11
BM09	2 M	CALGE	TAFLWGIQ	YTDKLIFGKG	3/8
		CALGE	ISYDESPFLLNWAPSC	TDKLIFGKG	3/8
BM10	2 M	CALG	VPTGGY	DKLIFGKG	14/17
BM16	3 M	CAL	SPWGR	DKLIFGKG	4/11
		CALGE	LVHRYWGTLS	TDKLIFGKG	4/11
		CALG	DPMFSYVSSYWGIA	DKLIFGKG	2/11
BM20	2 M	CALG	DRPRYVLGD	TDKLIFGKG	13/13
BM22	2 M	CALGE	QRAFQKRSPRY	TDKLIFGKG	21/29
		CALG	PTGGYVPDE	KLIFGKG	4/29
		CALGE	SLWGIRY	TDKLIFGKG	3/29
BM24	2 M	CALGE	SPSPSGWGILKSSY	TDKLIFGKG	6/25
		CALGE	PPLLGDAG	TDKLIFGKG	5/25
		CALGE	LSHLRRLYWGHA	DKLIFGKG	3/25
		CALG	DPIPAFWGFY	TDKLIFGKG	3/25
		CALG	APWPPNEGVRGIAY	TDKLIFGKG	2/25
		CALGE	RVGFLNWGIGNPIDPS	TAQLFFGKG	2/25
BM25	2 M	CALGE	LSSSRGYAGY	TDKLIFGKG	13/29
		CALG	GLPSYPPWGHL	TDKLIFGKG	7/29
		CALGE	LSSTGSRY	TDKLIFGKG	2/29
		CALGE	LSFPWTS	TDKLIFGKG	2/29
BM26	2 M	CALG	GGKISYLRIHGYS	DKLIFGKG	21/25
		CALGE	LARNRVWQY	TDKLIFGKG	2/25
BM27	2 M	CALGE	PHWGMKA	DKLIFGKG	12/15
BM30	2 M	CALGE	RLRHSPTGDTRG	TDKLIFGKG	8/9
BM32	3 M	CALGE	ILPYWTY	TDKLIFGKG	5/20
		CALGE	FHYEGWGIV	TDKLIFGKG	4/20
PB02	2 M	CALGE	PRFWGINS	DKLIFGKG	6/10
CBT01	20 M	CALGE	LENFLWGIPES	DKLIFGKG	14/17
CBT02	19 M	CALGE	LPTSSYWGLY	YTDKLIFGKG	6/14
		CALGE	LYPLPFWGIL	TDKLIFGKG	2/14

^*Blood samples were taken at the indicated time (months) after transplantation.

^†White colonies of transformed Escherichia coli containing plasmids with Vδ1–Cδinserts were picked up randomly and Vδ1–Cδtranscripts were sequenced. T cell clones appearing more than once are presented. The -WGI- amino acid sequence is underlined.

Peripheral expansion of donor-derived Vδ1 T cells and their longevity

Next, we addressed the questions of whether clonally predominant Vδ1⁺ T cells in blood were derived from the donor by peripheral expansion, and how long they could survive in the host if this was the case. Longitudinal analysis of CDR3 size distribution patterns of Vδ1⁺ T cells in blood from the patients BM01 and BM09 demonstrated that post-transplant skewing of the Vδ1⁺ TCR repertoire persisted for at least 7 years and 20 months, respectively (Fig. 2). However, some Vδ1⁺ T cell clones could be expanded selectively after transplantation, as in the case with BM09.

Fig. 2.

CDR3 size spectratyping of Vδ1–Cδ transcripts of blood γδ T lymphocytes from patients UPN-BM01 and BM09. Note that an oligoclonal peak of Vδ1⁺ T cells with a CDR3 of 532 base pairs was detectable in patient UPN-BM01 8 months after transplant, and was still present after 7 years. In patient UPN-BM09, the donor-derived clone was detected until 20 months after transplantation.

Expansion of the long-term surviving donor-derived Vδ1⁺ T cell clone against autologous EBV-infected B cells

EBV causes latent infection in human adults and post-transplant reactivation of EBV is also a well-known complication in recipients of allogeneic haematopoietic stem cell grafts [18]. Moreover, it has been reported previously that EBV-transformed B lymphocyte cell line (EBV–LCL) and Burkitt lymphoma cells stimulate the Vδ1⁺ T cells to proliferate in vitro [19]. In patient BM01, the donor-derived Vδ1 T cell clone with the -KSLPKYWGRPF- CDR3 sequence was still detected in the recipient's blood 7 years after transplant (Fig. 2). We wondered whether this clone could expand against autologous EBV-infected B lymphocytes (EBV–LCL). After obtaining informed consent, 10 ml of peripheral blood was drawn from the BM01 patient 10 years after transplant, and EBV–LCL was first established, and then PBMCs were co-cultured with irradiated autologous EBV–LCL. As shown in Fig. 3, γδ T lymphocytes using Vδ1, Vδ2 and Vδ3 segments all proliferated in response to autologous EBV–LCL. CDR3 size distribution patterns of the Vδ1⁺ TCR became skewed and the Vδ1 T cell clone with the -KSLPKYWGRPF- CDR3 sequence predominated in vitro after stimulation with EBV–LCL (Fig. 3).

Fig. 3.

In vitro expansion of γδ T cells from a long-term survivor of allogeneic haematopoietic stem cell transplantation (UPN-BM01) post-transplant recipient against autologous EBV–LCL. Peripheral blood mononuclear cells (1 × 10⁶/ml) were cultured with irradiated autologous Epstein–Barr virus–lymphoblastoid cell lines (1 × 10⁵/ml) in RPMI-1640 supplemented with heat-inactivated human antibody serum and interleukin-2. At the indicated interval after starting culture, cells were harvested for cell counting, flow cytometry and T cell receptor repertoire analysis.

Cytotoxic activity of the Vδ1⁺ T cell line and clones against autologous EBV–LCL

To examine the functions of the Vδ1 T cell clones expanded in vivo, we propagated Vδ1 T cells in vitro using PHA and IL-2 in trying not to affect the repertoire after cultivating PBMCs with autologous EBV–LCL from the recipient BM01 [20]. The in vitro-propagated Vδ1 T cell line was able to kill autologous EBV–LCL, and also efficiently lysed allogeneic EBV–LCL (Fig. 4). We then isolated Vδ1 T cell clones from the same patient in order to exclude the possibility that cytotoxic activity of the Vδ1 T cell line might have been attributable to contaminated TCRαβ T cells. Two Vδ1 T cell clones were isolated successfully from the patient BM01 in two separate experiments, although they were distinct from the clone with the -KSLPKYWGRPF- junctional sequence. Both clones clearly showed cytotoxicity against autologous EBV–LCL (data not shown).

Fig. 4.

Cytotoxic activity of the Vδ1⁺ T cell line against autologous Epstein–Barr virus–lymphoblastoid cell lines established from a long-term survivor after allogeneic haematopoietic stem cell transplantation. Peripheral blood mononuclear cells were isolated from the recipient BM01 and the Vδ1 T cell line was established according to the procedure described in Materials and methods. Purity of the Vδ1 T cell line was 91% and contained 1·7% Vδ2⁺ and 1·4% Vδ3⁺ T cells.

Oligoclonal expansion of Vδ1⁺ γδ T cells by EBV-infected B cells in healthy individuals

We then addressed the question of whether EBV–LCL could induce clonal expansion of Vδ1⁺ T cells in vitro from healthy donors with positive serology for EBV. The CDR3 size distribution patterns of Vδ1⁺ TCR became skewed 28 days after stimulation with autologous EBV–LCL or allogeneic EBV-infected Burkitt lymphoma Raji cells (data not shown). The T cell clone with the -GLPHALIMWGDLAY- CDR3 sequence was detected recurrently in blood from donor no. 1 and predominated in the culture with autologous EBV–LCL (Table 4). This clone was not observed in the day 28 culture containing allogeneic Raji cells. In donor no. 2, the T cell clone with the -DPSLPTAPDWGYGRP- CDR3 sequence was detected twice in blood and predominated in the culture with autologous EBV–LCL (Table 4). These results suggest the possibility that CDR3 structure may contribute to the recognition of EBV-infected B cells by Vδ1⁺ T cells. Neither heat-killed bacteria, lipopolysaccharide, CMV-infected cell lysates nor fungal components induced the in vitro oligoclonal expansion of Vδ1⁺ T cells from healthy individuals in our culture system (data not shown).

Table 4.

Induction of a skew of the Vδ1⁺ T cell repertoire in vitro in response to autologous Epstein–Barr virus–lymphoblastoid cell lines (EBV–LCL).

Blood donor	Time of analysis	Vd1	N-D-N	Jd	Colony frequency
No. 1	Day 0	CALGE	GLPHALIMWGDLAY	TDKLIFGKG	3/20
	Day 28	CALGE	LEEYWGLPH	TDKLIFGKG	10/27
	With EBV–LCL	CALGE	GLPHALIMWGDLAY	TDKLIFGKG	7/27
		CALG	GVLYWGIRR	TDKLIFGKG	2/27
		CALGE	SLWGIRY	TDKLIFGKG	2/27
		CALGE	LGETTPLLGGYSFA	LTAQLFFGKG	2/27
	Day 28	CALG	VSGLARGGSL	KLIFGKG	6/25
	With Raji	CALGE	ADWGIRARILY	TDKLIFGKG	4/25
		CALGE	PRAILGDTRIKRMY	TDKLIFGKG	4/25
		CALGE	LEEYWGLPH	TDKLIFGKG	3/25
		CALGE	DPGLPFLWY	TDKLIFGKG	2/25
		CALG	DLNLLWGIRSILPG	TDKLIFGKG	2/25
No. 2	Day 0	CALG	DPSLPTAPDWGYGRP	LIFGKG	2/12
		CALGE	DEGSLWCDT	YTDKLIFGKG	2/12
	Day 28	CALG	DPSLPTAPDWGYGRP	LIFGKG	4/8
	with EBV–LCL

Peripheral blood mononuclear cells (PBMC) from adult healthy donors positive for EBV VCA-IgG were co-cultured with either irradiated (30Gy) autologous EBV–LCL or allogeneic Raji cells in RPMI-1640 supplemented with 10% autologous serum. After 7 days of culture, interleukin-2 (10 U/ml) was added to the medium twice a week, and the medium was exchanged depending upon the confluence of cultures. Cells were harvested at day 28 of culture for analysis of T cell receptor repertoires. T cell clones appearing more than once are presented. In a separate experiment using PBMC derived from donor 1, the absolute numbers of Vδ1 T cells were enumerated after 7 days of culture. The numbers of Vδ1 T cells in the absence or presence of autologous EBV–LCL were 0·63 × 10⁵ cells/ml and 1·87 × 10⁵ cells/ml, respectively.

Discussion

We have demonstrated in the present study that skewed TCR repertoires of the Vδ1⁺ T cells are found in half the recipients of allogeneic grafts. The Vδ1⁺ T cell clone derived from the donor mature T cell pool was still detected in the recipient's blood even 10 years after transplant, and autologous EBV–LCL induced an expansion of the identical T cell clone in vitro. We have also shown that EBV-transformed B cells induce in vitro oligoclonal expansions of Vδ1⁺ T cells from healthy individuals. These results support the hypothesis that Vδ1⁺ T lymphocytes have a TCR repertoire against microorganisms which can cause latent infection observed commonly in human being.

The Vδ1⁺ T cell clones with the -WGI- motif were not always expanded prominently in allogeneic HSCT recipients. Moreover, the -WGI- motif containing T cell clones were not dominant in EBV–LCL-stimulated blood T lymphocytes from healthy individuals (Table 4). Because several distinct viruses can be reactivated simultaneously and various types of opportunistic infections can occur in recipients of allogeneic HSCT, we cannot conclude that this CDR3 motif is crucial for the recognition of EBV-infected B cells.

One can raise the question as to why EBV seropositive recipients did not always show a skew of TCR repertoires of the Vδ1⁺ T cells (Table 2). Although we are not ready to demonstrate the evidence to answer that question, it is conceivable that the magnitude and chronological profiles of clonal T cell expansions in response to any given antigen could be different among individuals, and our methodology may not be sensitive enough to detect the small clonal expansions.

As Vδ1⁺ T lymphocyte counts were not enumerated in the present study, restricted repertoires of Vδ1⁺ γδ T cells might not reflect the true expansion of this particular subset. In fact, our previous study did not show any increase in the absolute numbers of total γδ T lymphocytes in recipients of allo-HSCT [7]. However, the crucial point that we would like to make in this study is that certain Vδ1⁺ γδ T cell clones with particular CDR3 structures regenerate preferentially in vivo, and those T cell clones derive presumably from the donor mature T cell pool.

Expansion of Vδ1⁺ T cells in blood has also been reported in patients infected with HIV [8–10, 21], CMV [11–13] or malarial parasites [14]. It has been reported that B cells are involved in the proliferation of Vδ1⁺ T lymphocytes in HIV patients [22], and it is well known that reactivation of EBV can occur during the progression of HIV infection. EBV reactivation is also common among children infected withmalaria [23, 24]. Moreover, it has been suggested that Vδ1⁺ human γδ T cells respond to ligands expressed by EBV-infected B cell lines [19, 25]. Collectively, these findings suggest that EBV-infected B cells may cause expansion of Vδ1⁺ T cells in humans, although natural ligands for Vδ1⁺ TCR remain to be elucidated.

The role for TCR in antigen recognition by Vδ1⁺ γδ T cells is not understood fully. However, there is evidence supporting the hypothesis that the TCR is involved in antigen recognition by Vδ1⁺ γδ T cells. In healthy individuals, the complexity of the Vδ1⁺ γδ T cell repertoire is diverse during the early phase of life, but decreases with ageing as a consequence of the expansion of a few T cell clones [26–28]. This age-related skewing of the Vδ1⁺ TCR repertoire appears to have features similar to those of the TCRαβ⁺ T cell repertoire [29, 30]. CMV seropositivity has been shown to drive the TCRαβ⁺ T cell repertoire toward clonality in healthy aged individuals [31]. In addition, recurrent observation of the -WGI- amino acid sequence in the CDR3 regions of TCR on clonally predominant Vδ1⁺ T cells after allogeneic HSCT suggests that a common antigenic pressure might have generated skewing of the Vδ1 T cell repertoire, and that junctional sequences might contribute to antigen recognition by Vδ1⁺ T cells. Spada et al. have demonstrated that Vδ1⁺ T cells recognize non-polymorphic CD1c molecules and CD1c-specific Vδ1⁺ T cell clones are cytotoxic using perforin- and Fas-mediated cytotoxicity [32]. Because CD1c molecules are expressed on some circulating B cells in healthy individuals and mantle zone B cells in lymph nodes [33], Vδ1⁺ T cells may play a protective role in EBV infection via the TCR/CD1c interaction.

One crucial question can be raised as to whether there is any association between skewing of the Vδ1 T lymphocyte repertoire and post-transplant EBV-reactivation. We experienced one case with EBV-associated post-transplant lymphoproliferative disorder in the present patient cohort and the patient showed an oligoclonal expansion of Vδ1⁺ γδ T lymphocytes (data not shown). We would like to address this crucial issue in a future prospective study. Finally, the Vδ2⁺ T cells also proliferated vigorously in vitro in response to autologous EBV-infected B cells (Fig. 3). Although we have focused on the proliferation and function of the Vδ1⁺ T cell subset in trying to explore the biological relevance of skewed TCR repertoires of this T cell subset after allogeneic human HSCT, the role for Vδ1⁺ γδ T lymphocytes in the control of EBV infection should be also elucidated in the near future.