Identification of TCR Vβ11-2-Dβ1-Jβ1-1 T cell clone specific for WT1 peptides using high-throughput TCRβ gene sequencing

We previously identified a TCR Vβ21 T cell clone which was specific to CML patients, and demonstrated that TCR Vα13/β21 gene-modified CD3+ T cells had specific cytotoxicity for HLA-A11+ K562 cells. However, it remains unclear which antigen is specifically recognized by the TCR Vβ21 T cell clone. In this study, CD3+ T cells from healthy donor peripheral blood were stimulated with the WT1 peptide or mixed BCR-ABL peptides in the presence or absence of IL-2 and IL-7. The distribution of the TCR Vβ repertoire was analyzed after different stimulations. We found that the mixed BCR-ABL peptides induced clonally expanded Vβ7–9-Dβ2-Jβ2–7 T cells while the Wilms Tumor 1 peptide induced clonally expanded Vβ11–2-Dβ1-Jβ1–1 T cells by high-throughput TCRβ sequencing and GeneScan. Interestingly, the sequence and CDR3 motif of Vβ11–2 T cell clone are similar to the TCR Vβ21 (a different TCR V region naming system) T cell clone that we previously found in CML patients. Thus, our findings suggest that the TCR Vβ21 T cell clone found in CML patients might be a T cell clone that specifically recognizes WT1. Electronic supplementary material The online version of this article (10.1186/s40364-019-0163-1) contains supplementary material, which is available to authorized users.


Background
Chronic myelogenous leukemia (CML) is a common hematological malignancy in adults and has the molecular characteristic of BCR-ABL fusion proteins, which exhibit abnormal kinase activity [1]. Although many CML patients benefited from the development and application of tyrosine kinase inhibitors (TKIs) [2], a part of patients still suffer from primary and acquired resistance to TKIs [3,4]. Another therapeutic approach for CML is hematopoietic stem cell transplantation (HSCT), including allogenic-HSCT and haploidentical-HSCT; however, their use is limited for older patients [5,6]. There is evidence demonstrating that CML patients who have undergone recurrent allo-HSCT could be aided by donor lymphocyte infusion (DLI) [7][8][9]. These findings suggest that adoptive T cell immunotherapy may be a potentially effective strategy for CML patients.
Mechanistically, infusing donor-derived cytotoxic T lymphocytes (CTLs) induces CTL-mediated leukemia cell death through the recognition of leukemia-associated antigens. However, DLI also causes graft-versus-host disease (GVHD), mainly because CTLs are multi-clonal T cells that also recognize allo-antigens expressed in host-normal tissues [10]. Therefore, infusing leukemic antigen-specific CTLs is a better strategy for overcoming GVHD for adoptive T cell immunotherapy.
Wilms Tumor 1 (WT1) is a tumor suppressor gene involved in the etiology of Wilms' tumor. It is overexpressed mainly in myeloid leukemias, such as acute myeloid leukemia (AML) and CML, myelodysplastic syndrome (MDS), and several solid tumors [11][12][13][14]. There are evidences demonstrating that WT1 overexpression is closely associated with CML progression, and the poor therapeutic effect of TKIs [15][16][17][18][19]. These results highlight WT1 as a common therapeutic target for leukemia. For example, a clinic trail showed that WT1 peptide vaccination in WT1-expressing AML and MDS patients without curative treatment option had clinical benefit including complete remission or stable diseases (SDs) with more than 50% blast reduction. The effect is accompanied by the emergence of a predominant TCR Vβ + T cell clone both in blood and bone marrow [20][21][22][23]. However, the types of leukemia associated-antigens in AML patients are relatively complex, thus not allowing a clear definition of the types of antigens recognized by clonally expanded TCR Vβ T cells after injection with a WT1 vaccine.
We previously identified a TCR Vβ21 monoclone in blood from patients with CML and demonstrated that TCR Vα13/β21 gene-modified T cells could induce cell death in HLA-A11 + K562 cells [24,25]. However, it remains unclear whether TCR Vβ21 T cell clones specifically recognize BCR-ABL or other CML-associated antigens. Therefore, in this study, we analyzed the distribution of the TCR Vβ repertoire in CD3 + T cells from healthy donor peripheral blood after different stimulations with a WT1 peptide or mixed BCR-ABL peptides in the presence or absence of interleukin (IL)-2 and IL-7.

CD3 + T cell sorting
Peripheral blood mononuclear cells (PBMCs) were isolated from the peripheral blood of a healthy HLA-0201 + donor with informed consent by Ficoll-Hypaque gradient centrifugation. CD3 + T cells were then sorted by immunomagnetic beads from the PBMCs. The immunnomagnetic beads were purchased from MACS, and the sorting operation was performed according to the manufacturer's instructions (Miltenyi Biotec, Germany). This study was approved in writing by the Ethics Committee of the first affiliated hospital of Jinan University.
Cell culture and treatment CD3 + T cells (2.5 X 10 6 cells/mL) were cultured in RPMI 1640 without fetal bovine serum overnight. Fresh media containing 100 UI/mL IL-7, 100 UI/mL IL-2, and antigen peptides was then added to the cells. Untreated cells served as the control group, and the cytokines group comprised cells treated only with IL-2 and IL-7. The cells in the WT1 group were treated with a WT1-specific antigen peptide (RMFPNAPYL HLA A0201), while the cells in the BCR-ABL (B3A2) group were treated with six mixed antigen BCR-ABL peptides (Additional file 1: Table S1). The cells were cultured for 3 weeks. IL-2 was added into the media twice a week, and IL-7 was added into the media once a week. Finally, cultured T cells from different groups were collected for RNA isolation.

RNA extraction and TCRβ sequencing
Total RNA was extracted from samples with TRIzol (Invitrogen, 15,596) according to the manufacturer's instructions. The RNA was dissolved by ddH 2 O after drying out. Then, the immune library sequences were amplified by 5′rapid amplification of cDNA ends (RACE). After amplification, the concentration and integrity of the fragments were determined by Qubit, Agilent, and Q-PCR. Qualified libraries were sequenced by HiSeq or MiSeq. The mixcr (v1.8.2) program was used to identify the sequences in each sample. Sequences containing the complementarity-determining region 3 (CDR3) that had greater than four amino acids and a nucleic acid length that was a multiple of three without stop codon were retained as qualified clones. Bioinformatics analysis was performed after obtaining qualified clones. The amplification and sequencing of TCR Vβ and primary analysis were performed by the Huayin Health Company.

RT-PCR, sanger sequencing and GeneScan analysis for TCR Vβ subfamily clonality
Twenty-four TCR Vβ primers and a TCR Cβ primer were used in unlabeled PCR to amplify the TCR Vβ subfamily members. PCR was performed as described in our previous study. A portion of the PCR product was used for direct sequencing, which was performed by Invitrogen Biotechnology Company. The sequences of the different samples were analyzed with BLAST (https://blast.ncbi. nlm.nih.gov/Blast.cgi). The remaining PCR product was used to perform runoff PCR with the addition of fluorescent primers labeled at the 5′ end with a FAM (5-Carboxyfluorescein) fluorophore (Cβ-FAM) (TIB MOLBIOL GmbH, Germany). Then, the labeled runoff PCR products (2.0 μL) were mixed with 9.5 μL formamide (Hi-Di Formamide, ABI, USA) and 0.5 μL Size Standards (GENESCAN™-500-LIZ™, Perkin Elmer, ABI) and heatdenatured at 94°C for 4 min. The samples were resolved by electrophoresis using a 310 DNA sequenator (Perkin Elmer, ABI) in a 310 POP-4™ gel (Performance Optimized Polymer-4, ABI). The size and fluorescence intensity were determined by GeneScan software. The PCR protocol was performed as described in our previous study [26,27].

Distribution of the TCR Vβ repertoire in T cells after stimulation with WT1 and BCR-ABL peptides
We treated four groups of CD3 + T cells from the same healthy donor's PBMCs under different conditions and then analyzed the distribution of the TCR Vβ repertoire by TCRβ gene sequencing. Approximately 16.1 million effective reads were generated from the CD3-positive T cell populations (Table 1). There were approximately 60 types of the Vβ gene and 14 types of the Jβ gene detected in each group. The number of unique VDJ rearrangements was approximately 1.4 thousand in each group. The number of unique CDR3 amino acid sequences (corresponding to an in-frame effective rearrangement of the CDR3 nucleotide sequence) in the control group was 17,789, while it was 13,828 in the cytokines group. For the WT1 and B3A2 groups, the numbers were 14,472 and 11,747, respectively. From these data, we could also determine that there was no significant difference in the numbers of Vβ genes, Jβ genes, VDJ gene rearrangements, and unique CDR3 amino acid sequences in the four groups.
However, there were some differences in the usage of the Vβ genes and Jβ segments among the four groups (Fig. 1a, b). To obtain more detail regarding differences in the four groups, we further analyzed the data by the frequency of the Vβ subfamilies. We found that the top three Vβ subfamilies in the control group were Vβ20-1 (20.3%), Vβ5-1 (11.1%), and Vβ29-1 (9.2%), which was   Fig. 1c); however, in the WT1 group, the top three Vβ subfamilies changed to Vβ29-1 (15.4%), Vβ12-3 (13.5%), and Vβ20-1 (10.8%), while in the B3A2 group, Vβ7-9 (30.3%), Vβ20-1 (24.8%), and Vβ5-1 (15.4%) were the most frequent. These data demonstrate that stimulation with cytokines did not change the type of Vβ subfamilies used in normal T cells as expected, while stimulation with different peptides could induce the selective proliferation of different Vβ subfamily T cell clones.

Identification of a WT1-or BCR-ABL-specific TCR Vβ clone
We employed RT-PCR and GeneScan for further analysis. Based on GeneScan analysis (Fig. 3a), we found that the Vβ21 clone in the WT1 group was biclonal (bi-peaks) and the size of the fragment was approximately 150 bp. In the B3A2 group (Fig. 3b), we found that the Vβ6 clone was biclonal, and the control and cytokines groups were polyclonal (multi-peaks). The sequencing results (Fig. 3c) indicated that the CDR3 sequence of TCR Vβ6 in the B3A2 group was from the TCR Vβ7-9-Dβ2-Jβ2-7 clone, and the CDR3 sequence of TCR Vβ 21 in the WT1 group was from the TCR Vβ11-2-Dβ1-Jβ1-1 clone.

Discussion
Adoptive T cell transfusion using cancer antigen-specific T cells is the most effective immunotherapy [28]. However, there are issues that limit the application of this approach including difficulties in generating a sufficient number of cancer antigen-specific T cells for each patient in vitro in a short period of time, and patient antigen-specific T cells demonstrating low activation [29]. With the exception of chimeric antigene receptor (CAR)-T cells, TCR-engineered T cells have emerged from pre-clinical research to clinical trials and can overcome the low numbers of patient-derived CTLs [30][31][32][33][34][35]. Thus, identification of tumor antigen-specific TCRs is a key issue for such TCR-T cell generation and application.
To design specific T cell immunotherapies for CML, the identification of common CML-specific TCRs might be focused on BCR-ABL or WT1 antigens [36][37][38][39]. Based on our previous finding that the TCR Vα13/β21 gene derived from CML patient-modified CD3 + T cells can specifically target HLA-A11 + K562 cells [25], it would be interesting to further characterize the target of this TCR.
In this study, we first compared the frequent usage of the TCR Vβ repertoire in CD3 + T cells treated with WT1 or BCR-ABL peptides by high-throughput TCRβ gene sequencing. The predominant TCR Vβ clone in WT1 peptide-induced T cells was TCR Vβ11-2, while TCR Vβ7-9 was predominant in BCR-ABL mixed peptide-induced T cells. As expected, the clonal response of the TCRVβ subfamily cells appeared to vary with different leukemiaassociated antigen epitopes. To confirm this finding, we detected the expression of both Vβ11-2 and Vβ7-9 in treated CD3 + T cell samples by RT-PCR, and the PCR products were further analyzed by GeneScan to confirm the clonality of the T cells and direct Sanger sequencing to confirm the CDR3 rearrangement [40]. Significantly, clonal expansion of T cells expressing TCR Vβ11-2 or TCR Vβ7-9 was identified, and the CDR3 sequences were also confirmed as Vβ11-2-Dβ1-Jβ1-1 and Vβ7-9-Dβ2-Jβ2-7, respectively. Both are in-frame rearrangements, and their predicted CDR3 amino acid sequences were CASSSSGTGPNTEAFF (WT1 related) and CASSLAER-EYPLEQYF (BCR-ABL related). Therefore, both novel TCR clones might be responsible for BCR-ABL or WT1 epitopes, and whether they could be used to produce TCR-modified T cells requires further investigation. Interestingly, we found that the TCR Vβ11-2-Dβ1-Jβ1-1 sequence is similar to the sequence in the Vβ21 T cell clone (a different TCR V region naming system) in CML patients that we previously identified, which could mediate specific cytotoxicity against CML with the TCR-modified T cell technique [25]. Whether these Vβ21 T cell clones from CML patients specifically recognize the WT1 peptide or cross respond based on different individuals requires more investigation. Previous studies have been showed that WT1 specific T cell clones were induced in AML patients by the same WT1 126-134 peptide vaccines which was used in this study, such T cell clone also expressed TCR Vβ11, however, the CDR3 sequence (ASS-DYNEQF) is different from the TCR Vβ11-2-Dβ1-Jβ1-1 (CASSSSGTGPNTEAFF) which we found in the study and the CML patients [20,[41][42][43]. The reason that different TCR clone amplification induced by the same peptide, is due to the individual T cell response from different donors and patients. There were also studies showing different TCR clone (TCR Vβ5-1-Dβ2-Jβ2-5) inducted by WT1 (CMTWNQMNL) peptides [44]. Overall, the new identified TCR Vβ11-2-Dβ1-Jβ1-1 clone in this study may provide new data of WT1 specific TCR clone bank. Moreover, on the other hand, all of this identified TCR clone may be thought as one of the immune biomarker of WT1 specific T cell clone in WT1 + malignancies.

Conclusion
In summary, we characterized the different usage patterns of the TCRVβ repertoire in T cells after WT1 and BCR-ABL peptide stimulation and identified two novel TCR clones (Vβ11-2-Dβ1-Jβ1-1 and Vβ7-9-Dβ2-Jβ2-7) related to both antigens. Functional studies will be performed to confirm their anti-CML cytotoxicity by producing TCR gene-modified T cells, it may be possible to provide a new TCR-T cell clone for WT1 + leukemia and maybe for WT1 + solid tumors immunotherapy.

Additional file
Additional file 1: Table S1.