Abstract
INTRODUCTION
γδ T-cells expressing T-cell receptor (TCR) type Vγ9Vδ2 are well-established mediators of anti-tumor immunity and have demonstrable HLA-independent cytotoxic activity against both B- and T-cell non-Hodgkin lymphoma cells. Initial clinical trials using ex vivo expanded, adoptively transferred γδ T-cells to enhance tumor immunity have shown some promise but overall results have been lackluster. This indicates that wholesale polyclonal expansion prior to transfer is not likely to be effective. Thus more refined approaches are necessary. This requires a better understanding of their TCR repertoire. We recently developed a next generation sequencing (NGS)-based method for evaluating the spectrum T-cell receptor gamma (TRG) gene rearrangements present in clinical samples. In an effort to better understand the TCR repertoire of γδ T-cells we used this strategy to evaluate a series of samples from normal individuals and from individuals with B and T-cell lymphoproliferative disorders (LPDs).
METHODS
DNA was isolated from samples from 11 normal individuals (all peripheral blood, PB), 11 patients with a T-cell LPD (6 PB, 3 bone marrow; BM, 2 FFPE tissues), and 5 patients with a B-cell LPD (Hairy cell leukemia; HCL, 4 PB, 1 BM). γδ T-cells were sorted from the PB of 4 of the healthy donors by FACS. TRG rearrangements were PCR amplified using consensus primers and NGS libraries were prepared and sequenced on the Ion Torrent PGM platform. The data was analyzed as follows. NGS typically yielded up to 400,000 sequencing reads which were grouped by identical V, J, and CDR3 sequences into unique rearrangements (typically 15-30,000). The prevalence of a particular TRG rearrangement (or CDR3 sequence) was determined by the number of individual NGS reads with this unique sequence per the entire data set (percent of total reads). All rearrangements were then ranked by their prevalence.
RESULTS
We sequenced the TRG repertoire of isolated γδ T-cells from the peripheral blood of normal individuals (n=4). We found that a recurrent Vγ9-JγP rearrangement with the CDR3 sequence CALWEVQELGKKIKVF was always (4 of 4 samples) the most prevalent rearrangement in normal γδ T-cells (3.2-11.7-fold more prevalent than the second most common rearrangement; representing 4-11.9% of total reads). Similarly, analysis of a larger set of unsorted normal peripheral blood samples demonstrated high prevalence of the same canonical CDR3 in 5 out of 7 samples, confirming that it is very common in most individuals relative to all TRG rearrangements (among the top 10 most prevalent CDR3s in 4 of 5 samples). We also sequenced the TRG repertoire in 11 samples from patients demonstrating evidence of involvement by a T-cell LPD. Unexpectedly, all Vγ9-JγP type rearrangements were strongly suppressed in these samples including the one with the canonical CDR3. The canonical rearrangement (present in 6 /11 cases), or the most abundant Vγ9-JγP rearrangement when the canonical rearrangement was absent (absent in 5/11 cases), represented on average 0.036 ±0.024 % of all NGS reads in the samples from patients with T-cell LPDs compared to 0.72 ±0.72 % of total NGS reads in the normal controls. This represents on average a 19.9 fold reduction in the T-cell LPD samples. The median rank by abundance of the top Vγ9-JγP rearrangement dropped from 9th (normals) to 291st (T-cell LPD cases) indicating that the suppression was not simply a consequence of the presence of an abundant malignant T-cell clone in the data. The overall distribution of TRG V-segment usage in the normal and neoplastic samples was comparable (Vγ9: 9.9 ±0.55 % and 7.2 ±2.79 %, respectively). In 2 samples from patients with HCL, Vγ9-JγP rearrangements were reduced to a similar extent to that seen in the T-cell LPD cases (9.5-fold reduced; average 0.06% of NGS reads; average rank order 451st). In the other 3 cases of HCL the findings were very similar to those seen in the normal samples.
CONCLUSIONS
We identified a recurrent Vγ9-JγP rearrangement by NGS representing the most abundant CDR3 in sorted γδ T-cells from normal individuals. This population, along with other clones with Vγ9 rearrangements, appeared specifically suppressed in all samples from patients with T-cell LPDs and in 2 or 5 samples from patients with B-cell LPDs (HCL), perhaps indicating a role in the disease process. Additional samples from a wider range of T- and B-cell LPDs are being analyzed.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.