Purpose: T-cell receptor (TCR) variable Vα and Vβ gene diversity is a surrogate biomarker for the therapeutic potential of adoptive immunotherapy and cellular immunity. Therefore, creating a straightforward, rapid, sensitive, and reliable method to view the global changes of both TCRVα and Vβ transcripts in heterogeneous populations of T cells is appealing.

Experimental Design: We designed a “direct TCR expression assay” (DTEA) using a panel of customized bar-coded probes that simultaneously detects and quantifies 45 Vα and 46 Vβ transcripts in a nonenzymatic digital multiplexed assay from a small number of cells (104 cells) or as little as 100 ng of total RNA.

Results: We evaluated DTEA on total RNA samples of tumor-infiltrating lymphocytes and peripheral blood obtained from patients with melanoma after adoptive T-cell therapy. DTEA detected a similar spectrum of the dominant patterns of TCRVβ gene usage as sequencing cloned TCRVβ CDR3 regions. However, DTEA was rapid, achieved a level of sensitivity to identify rare T-cell populations, and simultaneously tracked the full array of Vα and Vβ transcripts.

Conclusions: DTEA can rapidly and sensitively track changes in TCRVα and gene usages in T-cell pools following immune interventions, such as adoptive T-cell transfer, and may also be used to assess impact of vaccination or reconstitution of T-cell compartment after hematopoietic stem cell transplantation. Clin Cancer Res; 18(17); 4733–42. ©2012 AACR.

Translational Relevance

Changes in TCR transcriptome are linked with both the efficacy of immunotherapy and the pathogenesis of various human diseases, such as DiGeorge syndrome, severe combined immunodeficiency syndromes, autoimmune disease, infection, chronic inflammation, aging, and cancer. Therefore, the development of a direct approach to track and measure the TCRVα and gene usages in a heterogeneous mixture of T cells is valuable for assessing the therapeutic potential of adoptive immunotherapy and understanding pathology associated with skewed TCR expression. To this end, we developed the “Direct TCR Expression Assay” (DTEA) as a sensitive, rapid, reliable, and high-throughput digital technique to directly quantify both TCRVα and Vβ transcripts in a single reaction using a small number of T cells or total RNA.

The T-cell receptor (TCR) expressed on a mature α/β T cell is composed of a heterodimer of genomically rearranged α- and β-chains generated through V(D)J somatic recombination with the addition/subtraction of nontemplated bases at recombination junctions that are fused in frame to constant (C) regions (1–4). The variable (V) regions contain 3 hyper variable complementarity-determining regions (CDR1, CDR2, and CDR3) that confer T-cell specificity via the recognition of small peptide antigens (8–12 mer) in the context of MHC proteins (1). The CDR3 sequences of V regions are highly polymorphic and largely determine ability of T cells to recognize peptide antigen. CDR3 is unique to each rearranged TCR on a T-cell clone and thus TCRVα- and Vβ-chain genes can be identified within families based on the shared sequences in hinge regions flanking the unique CDR3 sequences in the α- and β-chains (1, 2). Currently, there are 45 TCRVα- and 48 TCRVβ-chains isolated and sequenced from the human genome (www.imgt.org/) which pair to form a mature and functional αβ TCR (2, 5–7). Methods to assess TCR diversity/usage within a T-cell population include (i) sequencing of CDR3 regions from TCRVβ genes (5, 8–11), (ii) spectratyping to analyze polymorphisms in length of CDR3 within Vβ family (5, 12–14), and (iii) flow cytometry using monoclonal antibodies to identify cell surface expression of TCRVβ-chains (15). Up until now, analysis of TCR clonotype has generally relied on PCR-based amplification of the DNA sequences that incorporate the CDR3 region using primers recognizing conserved Vβ sequences flanking the CDR3 (2, 5, 6, 8, 9). This has been referred to as Vβ clonotyping. Methods to amplify and sequence CDR3 regions from genes have been less commonly used.

Changes in the TCR usage, including a skewing towards monoclonality or oligoclonality can be desired, such as the emergence of T cells with therapeutic potential after vaccination and adoptive transfer (9, 10, 13, 16, 17), or unwanted as associated with inborn errors such as DiGeorge syndrome and severe combined immunodeficiency syndromes, autoimmune disease, infection, chronic inflammation, aging, and cancer (8–10, 12, 13, 18–22). In particular, investigators have described adoptive transfer of tumor-infiltrating lymphocytes (TILs) into patients with metastatic melanoma that result in emergence of T cells in peripheral blood identified as expressing a subset of TCRVβ-chains found in the original infusion product and this skewing is linked to superior antitumor response rates (9, 10, 17, 23–25). Thus, the serial assessment of TCRVβ diversity has been successfully used to evaluate the persistence and inform on the therapeutic potential of transferred TILs (9, 10).

To reveal the complexity of TCR diversity, investigators have sequenced CDR3 rather than measuring their length. Massively parallel sampling of TCR usage by sequencing unique Vβ CDR3 enzymatically amplified amplicons has identified clonotypes within T-cell pools (6, 7, 26). This approach to profiling TCR sequences reveals the nucleotide sequences that compose the diversity and length of CDR3, but has been primarily undertaken within TCRVβ family due to higher sequence variation at the Vα loci (11), as well as the downstream expense and time needed for acquisition and bioinformatics to analyze data. As an alternative to profiling T-cell Vβ metagenomes by high-throughput sequencing, cloning and sequencing of TCRVβ genes can be undertaken, despite that this method is laborious and time-consuming (8–10, 20). However, low frequency, yet potentially clinically important T-cell populations may not be detected by this approach, as only a limited number of bacterial clones carrying TCRVβ gene inserts can be isolated and sequenced. Furthermore, this approach has been limited to analyzing the TCRVβ diversity and not patterns of Vα usage. Therefore, a practical and sensitive alternative method that can rapidly measure or screen for the expression frequencies of all TCRVα and Vβ transcripts in heterogeneous populations of T cells is warranted.

Here, we report a new nonenzymatic approach to directly analyze TCR diversity using a method that directly identifies and measures the frequency of both TCRVα and gene expression in RNA isolates or cell lysates without the need for PCR (including 5′ RACE) and gene cloning. The DTEA uses the nCounter assay system (NanoString Technologies, Seattle, WA). This system uses panels of multiple fluorescently tagged (bar-coded) probes that specifically bind mRNA transcripts (without the enzymatic amplification of RNA) in a small number of cells or RNA, making possible the digitalization of gene expression (27). In the current study, we designed custom probes to identify 45 Vα and 46 Vβ TCR genes present in the human genome. We evaluated the applicability of DTEA to serially measure changes in TCR gene expression profiles before and after adoptive transfer of ex vivo-propagated TILs by serially sampling the peripheral blood of patients with metastatic melanoma. We validated DTEA by comparing it with TCRVβ gene cloning/sequencing and found that DTEA yielded reproducible results that reliably identified changes in the TCR usage in a highly sensitive manner over time, and observed a high degree of concordance between DTEA and TCRVβ cloning/sequencing. However, low frequency clonotypes not detected using Vβ cloning/sequencing could be identified and tracked by DTEA. In aggregate, DTEA provides a rapid, accurate, and sensitive global view to quantify the 45 TCRVα and 46 TCRVβ transcripts in a T-cell population.

Samples for TCR analyses

TILs and peripheral blood mononuclear cells (PBMC) samples were obtained from 14 patients after informed consent at The University of Texas MD Anderson Cancer Center who had stage IIIc-IV metastatic melanoma and were participating in a U.S. Food and Drug Administration (FDA)-approved phase II T-cell therapy clinical trial (IND# BB-IND 12192, Institutional Review Board protocol 2004-0069) evaluating the efficacy of numerically expanded autologous TILs. The detailed information of this clinical trial was described in http://www.cancer.gov/clinicaltrials NCT00338377. TILs harvested from surgically resected visceral metastatic lesions were initially expanded with high-dose interleukin-2 (6,000 IU/mL), and then subjected to a 2-week rapid expansion protocol to generate the final autologous TIL product used for infusion. Before TIL infusions, patients were treated with cyclophosphamide (60 mg/kg) and fludarabine (25 mg/m2) to render recipients lymphopenic. After TIL infusions, peripheral blood samples were obtained from patients at 2 weeks, 1 month, 5 to 6 months, and 11 to 12 months (24, 25). The PBMCs were then isolated using Ficoll-Hypaque, and 5 × 106 PBMCs or TILs were lysed in RNeasy buffer (Qiagen, catalog #74104) for RNA isolation. Altogether, 18 TIL samples and 25 PBMC samples were serially collected after TIL infusions from 14 patients participating in the clinical trial and all were analyzed by both DTEA and sequencing (Supplementary Table S1).

TCRVβ gene cloning and sequencing

Total RNA was isolated from TILs and PBMCs using an RNeasy Kit and treated with DNase to remove genomic DNA contamination (Qiagen), and the quality of the RNA preparation was validated by agarose gel electrophoresis. To clone TCRVβ transcripts, the SMARTer RACE cDNA amplification kit (5′RACE, Clontech, catalog #634924) was used as described previously (9, 10, 17). Briefly, 1 μg total RNA was used to generate full-length cDNA using 5′-RACE CDS primer (provided in the kit), and the 5′ adaptor (SMARTer II A primer) was then incorporated into each cDNA according to manufacturer's instructions. The TCRVβ-specific cDNA was then amplified by PCR using a 3′ primer (5′-CGA GGT AAA GCC ACA GTC-3′) that binds to the constant region of both C1 and C2 of the TCRVβ genes and a 5′ primer (provided in the kit) that is complement to the adaptor sequence. The resulting TCRVβ-specific PCR products were purified from the agarose gel and subcloned into the TA cloning vector TOPO4 (Invitrogen). Plasmid DNA was prepared from 96 bacterial colonies using Direct Prep96 BioRobot kit (Qiagen, catalog #96234). These 96 plasmid DNA isolates were further confirmed by PCR using primers that nested to TCR constant region (5′-TCTGATGGCTCAAACACAG-3′) and nested to the 5′ adaptor (Nested Universal Primer A, NUP; 5′-AAGCAGTGGTATCAACGCAGAGT-3′) provided by manufacturer. Finally, the TCR-positive clones (usually 60–96 DNA plasmid isolates) were subjected to DNA sequencing at the core facility of MD Anderson Cancer Center (Houston, TX). These sequencing data were analyzed using the international IMGT (ImMunoGeneTics) information system (http://imgt.org/) to characterize the individual TCRVβ clones.

DTEA to quantify TCRVα and Vβ gene expression

The IMGT database (http://imgt.org/) was also used to identify TCRVα and Vβ sequences and customized probes for 45 Vα and 46 Vβ families, which were designed and manufactured as an nCounter Gene Expression Assay Kit (NanoString Technologies). Because of 100% cross-hybridization between TCRVα8-2 and TCRVα8-4, the latter was omitted from the panel. Similarly, TCRVβ6-3 and Vβ12-4 were also omitted from the panel because of 100% cross-hybridization with Vβ6-2 and Vβ12-3. Thus, only 45 of 46 TCRVα and 46 out of 48 TCRVβ family members could be selected for the analysis. These TCRVα and Vβ code sets and the housekeeping control genes are listed in Supplementary Tables S2, S3, and S4. Target molecules were detected according to hybridization to capture reporter sequences, each approximately 50 nt in length, that targeted a contiguous 100-bp sequence in each of the Vα and Vβ mRNA species. Target regions were screened against the RefSeq database to exclude direct and inverted repeat elements and evaluated to eliminate cross-hybridization. Each color-coded barcode is attached to a single target-specific reporter probe corresponding to a gene of interest as described previously (27). Bar-coded probes, each 50 bp in length, were selected for melting temperatures between 78°C to 83°C. Hybridizations for each sample were set up as follows: 5 μL (100 ng) of total RNA, 10 μL of hybridization buffer, 10 μL of reporter probes, and 5 μL of capture probes that were mixed in PCR tubes and incubated at 65°C for 12 to 18 hours in a thermal cycler with a heated lid (Pelletier, BIO-RAD DNA Engine). After hybridization, the samples were processed in an nCounter PrepStation and counted using an nCounter Digital Analyzer (model no. NCT-SYST-120; NanoString Technologies). The processing and counting steps were fully automated and required no user interaction

Detection limit of TCRVα and Vβ by DTEA

To determine the sensitivity of DTEA, DMF5 cells were serially diluted in Jurkat cells to assess the minimal cell numbers that were required for detecting both TCRVα and TCRVβ expression. DMF5 is a primary human CD8 T-cell clone generated from melanoma TILs and coexpresses TCRVα12-2 and TCRVβ6-4 (28). The DMF5 T-cell line was sorted with a fluorescence-activated cell sorter based after CD8 and Mart-1 tetramer (Beckman Coulter) staining to obtain more than 95% CD8 expression and Mart-1 tetramer–positive expression. Jurkat is a human acute lymphoblastic leukemia T-cell line that expresses TCRVβ7-2 and apparently several TCRVα genes, including TCRVα18, 19, and 8-2 (all measured by DTEA from this work). Serial dilutions of DMF5 were spiked into Jurkat cells, resulting in a final DMF5 cell concentration ranging from 0.001% (1 DMF5 cell in 105 Jurkat cells) to 10% (104 DMF5 cells in 9 × 104 Jurkat cells). A total of 105 cells were then lysed in 5 μL RNeasy lysis buffer (RLT) buffer from RNeasy Mini Kit (Cat # 74104, Qiagen) and used for DTEA assay with the same conditions used to determine TCRVα and TCRVβ expression in total RNA of TILs and PBMCs.

Statistical analysis

To determine the reproducibility of DTEA, TIL RNA samples from 3 patients were analyzed and each RNA sample was run in triplicates. Gene expression data was analyzed according to nCounter data analysis guidelines. To obtain “specific hybridization counts” for the positive spike controls, housekeeping genes, and test genes, the average of the negative spike control counts in each lane was subtracted from that of the raw positive control counts, housekeeping genes counts, and test genes counts, respectively. Specific test gene hybridization counts in each lane were then normalized to both the positive control and housekeeping genes by the following formula: (test gene–specific hybridization counts × positive spike control normalization factor × housekeeping genes normalization factor). Normalization factors for the positive control and housekeeping genes in each lane were derived using the following formula: (sum of average counts/lane-specific sum of counts). Following normalization, triplicate measurements for each gene in each sample were averaged and SD and coefficient of variation (% CV) was determined. To calculate % CV, SD of the set of replicate measurements was divided by average value for those replicates (29, 30). Coefficient variation (CV) and median CV for these samples were analyzed in Microsoft Excel version 2007.

Reproducibility and sensitivity of DTEA to detect patterns of TCRVα and Vβ usage

DTEA should be capable of simultaneously assessing the expression of both TCRVα- and Vβ-chains in a mixed T-cell population and this was tested in triplicate using TILs from 3 patients (253, 303, and 232) participating in a T-cell therapy clinical trial (Fig. 1A and B and Supplementary Fig. S1). Variations in the measured transcript numbers of individual TCRVα and genes in these TIL samples ranged from 12% to 21% (median CV for TIL253 was 12%, TIL303 was 20%, and TIL232 was 21%). To address reproducibility between probe batches, the DTEA assay was conducted on a single RNA sample (TIL152) using 2 different lots of Vβ probes. These experiments yielded similar results in terms of the spectra and quantities of detected TCRVβ genes (Fig. 1C). These data show that the variation associated with detecting TCR transcripts is minimal (29).

Figure 1.

Reproducibility and sensitivity of the DTEA. TCRVα (A) and TCRVβ (B) gene expression in total RNA from TILs of patient 253 was measured. Data shown are average of triplicates with SD. C, RNA was isolated from TILs of patient 152 and subjected to 2 separate analyses. The relative expression level for each TCRVβ gene is presented as the percentage of total T-cell clones, and the results of the 2 analyses are compared. Filled bar and open bar represent experiments 1 and 2, respectively (D and E). DMF5 cells (coexpressing TCRVα12-2 and TCRVβ6-4) were spiked at 10%, 1%, 0.1%, 0.01%, and 0% with Jurkat cells (coexpressing TCRVβ7-2 and TCRVα18, 19, and 8-2) resulting in a final DMF5 cells at 105, 104, 103, 102, 101, and 10−1. Five microliters of cellular lysates containing 105 total cells was used for DTEA. The experiment was carried out twice with duplicate measurements at each time. Background TCRVα12-2 and TCRVβ6-4 expression counts in control Jurkat cells was subtracted from the spiked sample before data analysis. One DMF5 cell in 1,000 Jurkat cells can be detected showing the sensitivity and accuracy of the assay. Normalized gene expression counts were Log10 transformed and plotted in y-axis. The x-axis represents the 10-fold serial dilutions of DMF5 in Jurkat cells; “0” = 100% (105 DMF5 cells); “1” = 10% (104 DMF5 cells and 9 × 104Jurkat cells); “2” = 1% (103 DMF5 and 9.9 × 104 Jurkat cells); “3” = 0.1% (102 DMF5 cells and 9.99 × 104 Jurkat cells) and “4” = 0.01% (10 DMF5 cells and 9.999 × 104 Jurkat cells). The solid circles represent data from the first experiment and the solid squares represent data from the second experiment.

Figure 1.

Reproducibility and sensitivity of the DTEA. TCRVα (A) and TCRVβ (B) gene expression in total RNA from TILs of patient 253 was measured. Data shown are average of triplicates with SD. C, RNA was isolated from TILs of patient 152 and subjected to 2 separate analyses. The relative expression level for each TCRVβ gene is presented as the percentage of total T-cell clones, and the results of the 2 analyses are compared. Filled bar and open bar represent experiments 1 and 2, respectively (D and E). DMF5 cells (coexpressing TCRVα12-2 and TCRVβ6-4) were spiked at 10%, 1%, 0.1%, 0.01%, and 0% with Jurkat cells (coexpressing TCRVβ7-2 and TCRVα18, 19, and 8-2) resulting in a final DMF5 cells at 105, 104, 103, 102, 101, and 10−1. Five microliters of cellular lysates containing 105 total cells was used for DTEA. The experiment was carried out twice with duplicate measurements at each time. Background TCRVα12-2 and TCRVβ6-4 expression counts in control Jurkat cells was subtracted from the spiked sample before data analysis. One DMF5 cell in 1,000 Jurkat cells can be detected showing the sensitivity and accuracy of the assay. Normalized gene expression counts were Log10 transformed and plotted in y-axis. The x-axis represents the 10-fold serial dilutions of DMF5 in Jurkat cells; “0” = 100% (105 DMF5 cells); “1” = 10% (104 DMF5 cells and 9 × 104Jurkat cells); “2” = 1% (103 DMF5 and 9.9 × 104 Jurkat cells); “3” = 0.1% (102 DMF5 cells and 9.99 × 104 Jurkat cells) and “4” = 0.01% (10 DMF5 cells and 9.999 × 104 Jurkat cells). The solid circles represent data from the first experiment and the solid squares represent data from the second experiment.

Close modal

To assess the minimal number of T cells needed to detect TCRVα and Vβ usages, the T-cell clone DMF5, which expresses different TCRVα and Vβ than Jurkat cells, was spiked with Jurkat cells at different ratios and subjected to the DTEA assay. A total of 105 cells were used in each DTEA reaction. The assay was conducted in 2 independent experiments. As shown in Fig. 1D and E, the mRNA levels coding for TCRVα12-2 and TCRVβ6-4 varied in a linear fashion and proportional to the dilution of the DMF5 clone (P < 0.0001 in both cases). The limit for detecting the DMF5-specific TCRVα12-2 and TCRVβ6-4 expression from the diluted samples was 0.1% (approximately 100 DMF5 in ∼105 Jurkat cells).

DTEA can detect the same pattern of TCRVβ usage as cloning and sequencing of CDR3s

A side-by-side analysis of TCRVβ gene expression comparing DTEA versus classical TCR gene cloning/sequencing in TILs from patient 228 revealed that DTEA detected all of the TCRVβ clones that were identified by TCR gene cloning/sequencing, as well as additional TCRVβ genes undetectable by cloning/sequencing. Both DTEA and TCR gene cloning/sequencing identified the major TCR clonotypes, TCRVβ11-2 and Vβ4-3. The calculated frequencies for TCRVβ11-2 and Vβ4-3 were similar for DTEA (54% and 18%; Fig. 2A), and TCR gene cloning/sequencing (57% and 24%; Fig. 2C). In contrast, the minor clonotypes TCRVβ20-1 and Vβ28 were detected at a frequency of 7% and 5.3%, respectively, using DTEA (Fig. 2A), but at a frequency of only 1.2% and 0% using the cloning/sequencing (Fig. 2C) methodology. Both TCRVβ12-3 and Vβ 6-2 were detected at a frequency of 1.15% by gene cloning, but only 0.47% and 0.35% by DTEA (Fig. 2B and D). This difference may be due to sampling error stemming from the limited number of TCR clones analyzed by cloning (usually 60–96 clones for each sample), which results, in this case, in an overestimation of the TCRVβ frequencies, especially of less dominant clonotypes. Expression analysis of TCRVβ genes from 5 additional TIL samples confirmed that DTEA and gene cloning could detect the same TCRVβ transcripts, although DTEA detected a larger array of clonotypes (Supplementary Table S5). These data show that DTEA not only detects the same TCR clonotypes as a TCR cloning/sequencing technique but also seems to be more sensitive and less prone to missing TCRVβ transcripts that are present at low frequencies.

Figure 2.

DTEA and classical gene cloning yield similar TCRVβ patterns of usage. A, RNA (100 ng) from TIL228 was subjected to TCRVβ gene expression analysis by DTEA, and (C) 1μg of same RNA was used for gene cloning/sequencing analysis. The TCRVβ gene expression level (percentage of total T-cell clones) detected using the 2 methods yielded similar TCRVβ expression patterns. The low expression clone, such as TCRVβ12-3 and TCRVβ6-2, was detected to be 1.15% by cloning (D) but only 0.47 and 0.35%, respectively, by DTEA (B).

Figure 2.

DTEA and classical gene cloning yield similar TCRVβ patterns of usage. A, RNA (100 ng) from TIL228 was subjected to TCRVβ gene expression analysis by DTEA, and (C) 1μg of same RNA was used for gene cloning/sequencing analysis. The TCRVβ gene expression level (percentage of total T-cell clones) detected using the 2 methods yielded similar TCRVβ expression patterns. The low expression clone, such as TCRVβ12-3 and TCRVβ6-2, was detected to be 1.15% by cloning (D) but only 0.47 and 0.35%, respectively, by DTEA (B).

Close modal

Tracking changes in levels of TCRVα and Vβ transcripts after TIL adoptive transfer

To determine whether DTEA could be used to track T-cell persistence and variations in the TCR usage (e.g., numeric expansion or contraction of T-cell clones) in patients after TIL infusions, we compared results obtained using TCR cloning/sequencing and DTEA assays from PBMCs obtained at different times following TIL transfer (Fig. 3). Both methods revealed a similar pattern of TCRVβ gene expression in patients' PBMCs over time. The same dominant TCRVβ transcripts were identified by both methods with some minor differences in the frequencies of TCRVβ gene expression at certain time points (Fig. 3C). To further assess whether DTEA can monitor changes in TCR gene usage, associated with infusion of TIL, we serially followed both TCRVα and TCRVβ gene expression in 6 patients following adoptive immunotherapy. As shown in Supplementary Figs. S2–S7, DTEA could track the persistence and changes of the original pattern of TCR usage exhibited by the infused TIL when measured over time in patients' peripheral blood.

Figure 3.

DTEA and TCR gene cloning/sequencing to longitudinally track TCRVβ usage in peripheral blood. RNA samples were prepared from TIL and PBMC samples obtained from 4 different patients at 1, 2, 5, 11, or 12 months after TIL infusion. The RNA samples were analyzed for T-cell clone persistence using both DTEA and cloning with different amounts of RNAs (100 ng for DTEA and 1 μg for cloning). The persistence of one dominant T-cell clone in each patient is shown: (A) TCRVβ13 in patient 231, (B) TCRVβ4-3 in patient 188, (C) TCRVβ28 in patient 172, and (D) TCRVβ9 in patient 106. Overall, the 2 methods yielded similar TCRVβ gene expression frequency patterns, although substantial differences in frequencies (up to 36%) between the 2 methods was observed at some points, as seen with TCRVβ28 in peripheral blood obtained from patient 172 at 2 months (C).

Figure 3.

DTEA and TCR gene cloning/sequencing to longitudinally track TCRVβ usage in peripheral blood. RNA samples were prepared from TIL and PBMC samples obtained from 4 different patients at 1, 2, 5, 11, or 12 months after TIL infusion. The RNA samples were analyzed for T-cell clone persistence using both DTEA and cloning with different amounts of RNAs (100 ng for DTEA and 1 μg for cloning). The persistence of one dominant T-cell clone in each patient is shown: (A) TCRVβ13 in patient 231, (B) TCRVβ4-3 in patient 188, (C) TCRVβ28 in patient 172, and (D) TCRVβ9 in patient 106. Overall, the 2 methods yielded similar TCRVβ gene expression frequency patterns, although substantial differences in frequencies (up to 36%) between the 2 methods was observed at some points, as seen with TCRVβ28 in peripheral blood obtained from patient 172 at 2 months (C).

Close modal

Tracking low frequency TCRVα and Vβ usage in T cells by DTEA after TIL-adoptive transfer

To quantity the magnitude of TCR clonotypes not detected by TCR cloning/sequencing, but identified using DTEA, we tested samples obtained from TILs and peripheral blood collected before and after infusion for TCRVβ genes expression using both methodologies. The results are summarized in Fig. 4 and Supplementary Table S5. TCR gene cloning/sequencing detected only 6 and 12 different clonotypes in TIL228 and in TIL231, respectively. In contrast, DTEA of the same samples revealed the expression of 30 additional TCR clonotypes in TIL228 and 22 additional TCR clonotypes in TIL231, with frequencies ranging from 0.005% to 6.7% (Supplementary Table S5). TCRVβ gene cloning/sequencing could not detect a number of low expression Vβ transcripts, such as Vβ14 (patient 152), and Vβ19, or Vβ24 (patient 172) in the initial TIL inocula or peripheral blood samples shortly after the infusion (Fig. 4). Tracking of TCRVβ gene usage for longer time intervals after TIL infusion found that while TCR gene cloning/sequencing detected TCRVβ14 (3.2% frequency, patient 152) and TCRVβ19 (1.36% frequency, patient 172), at only 11 and 5 months, DTEA detected these Vβ transcripts already in the initial TIL infusion sample and at earlier time points following administration (Fig. 4A and B). Similar results were seen with TCRVβ24 (Fig. 4C). This suggests that some of the expanding or emerging T-cell clones in the peripheral blood of patients who have undergone TIL transfer may arise from a low frequency pool of T-cell clones that were undetectable by the classical TCR gene cloning/sequencing method. In aggregate, these results indicate that DTEA is more sensitive compared with cloning/sequencing in detecting rare TCR gene usage.

Figure 4.

Serial tracking of low frequency TCR usage in peripheral blood by DTEA. Low frequency T-cell clones that cannot be detected by classical cloning/sequencing can be detected by DTEA. TCRVβ14 in patient 152 (A), TCRVβ19 (B), and TCRVβ24 in patient 172 (C). These T cells can only be detected by cloning/sequencing at a later time point as emerging in peripheral blood at 5 months (Vβ19 and Vβ24) and 11 months (Vβ14) after TIL infusion.

Figure 4.

Serial tracking of low frequency TCR usage in peripheral blood by DTEA. Low frequency T-cell clones that cannot be detected by classical cloning/sequencing can be detected by DTEA. TCRVβ14 in patient 152 (A), TCRVβ19 (B), and TCRVβ24 in patient 172 (C). These T cells can only be detected by cloning/sequencing at a later time point as emerging in peripheral blood at 5 months (Vβ19 and Vβ24) and 11 months (Vβ14) after TIL infusion.

Close modal

Simultaneous tracking of TCRVα and Vβ gene usage before and after TIL infusion

To further test the use of DTEA, we used DTEA to simultaneously track TCRVα and Vβ expression in TIL and PBMC samples at different times following TIL infusion. As shown in Fig. 5, we were able to concurrently determine TCRVα22 (45.15% frequency) and Vβ4-3 (44.17% frequency) representing 2 of the highest expressing TCRVα and Vβ clonotypes in TIL172 (Fig. 5A and B). A similar result was obtained using PBMCs 2 weeks after infusion from patient 117, such that the most frequent TCR clonotypes were TCRVα13-2 (85.12%) and Vβ7-2 (82.35%; Fig. 5D and E). For both patient samples, the frequency of other TCR clonotypes was at least 30% and as much as 7-fold lower. These results suggest that TCRVα22 and TCRVβ4-3 (in patient 172) or TCRVα13-2 and TCRVβ7-2 (in patient 117) may be paired in the same T cells. Further supporting our hypothesis was the observation of overlapping expression patterns for TCRVα22/Vβ4-3 and TCRVα13-2/Vβ7-2 in all PBMC samples obtained from patients 172 and 117, respectively (Fig. 5C and F). Moreover, CDR3 sequence analysis obtained using TCR cloning/sequencing identified only a single T-cell clone expressing TCRVβ4-3 or TCR Vβ7-2 in TIL172 and in PBMC117, respectively. Given our results with TIL172 and PBMC117, we extended our DTEA multiplex analysis to samples from patients 231, 228, and 106 (Supplementary Table S6). For TIL 231 TCRVα38-2 (17.15%), TCRVα14 (25.31%), TCRVβ7-8 (18.76%), and TCRVβ-13 (27.89%) were most highly represented, which on the basis of similar frequencies, suggests pairing of TCRVα38-2 with Vβ7-8, and TCRVα14 with Vβ-13 (Supplementary Table S6). In some samples, we observed a larger variation in major TCRVα and Vβ expression, such as TCRVα29 (13.76%) and TCRVβ9 (42.82%) in TIL 106, suggesting that the sample contained more than one TCRVβ9 clonotypes, each presumably pairing with a different Vα-chains (Supplementary Table S6). Overall, these results suggest that DTEA may be able to predict TCR αβ-chain pairing for a particular T-cell clone within an oligoclonal population when there is significant enrichment of this clone and when the measured TCRVα- and Vβ-chains are detected at similar frequencies.

Figure 5.

DTEA simultaneously tracks TCRVα and Vβ usage after adoptive T-cell therapy. TCRVα (A) and TCRVβ (B) gene usages in TIL172 were compared. TCRVα22 and Vβ4-3 were highly expressed in this sample, with frequencies of 45.15% and 44.17%, respectively. C, the expression patterns of TCRVα22 and Vβ4-3 in PBMCs at different times after adoptive transfer were similar. TCRVα (D) and (E) gene usage in peripheral blood of patient 117 at 2 weeks after TIL infusion were compared. TCRVα13-2 and Vβ7-2 were highly expressed in this sample, with frequencies of 85.12% and 82.35%, respectively. F, similar usage patterns for TCRVα13-2 and Vβ7-2 in TIL and PBMC at different times after infusion.

Figure 5.

DTEA simultaneously tracks TCRVα and Vβ usage after adoptive T-cell therapy. TCRVα (A) and TCRVβ (B) gene usages in TIL172 were compared. TCRVα22 and Vβ4-3 were highly expressed in this sample, with frequencies of 45.15% and 44.17%, respectively. C, the expression patterns of TCRVα22 and Vβ4-3 in PBMCs at different times after adoptive transfer were similar. TCRVα (D) and (E) gene usage in peripheral blood of patient 117 at 2 weeks after TIL infusion were compared. TCRVα13-2 and Vβ7-2 were highly expressed in this sample, with frequencies of 85.12% and 82.35%, respectively. F, similar usage patterns for TCRVα13-2 and Vβ7-2 in TIL and PBMC at different times after infusion.

Close modal

DTEA seems to be a sensitive, fast, and reliable high-throughput digital technique to directly quantify both TCRVα and Vβ usage in a single nonenzymatic reaction using a small number of cells (100 target T cells) or as little as 100 ng total RNA. Currently, the cost runs about $1 to $1.25 per gene per sample or about $91 to $114 to survey the direct level of expression of all 45 TCRVα and 46 TCRVβ genes in a sample. Using T-cell samples recovered from patients with melanoma that had received autologous ex vivo-propagated TILs, we observed that DTEA could be successfully used to (i) measure TCR Vα and Vβ usage and diversity and (ii) track low frequency T-cells in TIL and emerging in peripheral blood following adoptive immunotherapy. There was a high degree of concordance between DTEA and classical TCRVβ gene cloning/sequencing for measuring the frequency of TCRVβ transcripts when present at high abundance. However, DTEA was superior at identifying low frequency TCRVβ (and presumably ) transcripts, which is a major advantage over the TCR cloning/sequencing method.

Measuring a broad complement of TCR chains can inform on the therapeutic potential of infusions of heterogeneous mixtures of tumor-specific T cells propagated ex vivo (23, 24, 28, 31–34). The antitumor effect of adoptive immunotherapy can be gauged by longitudinally tracking the infused T cells identified by measuring TCRVα and genes expression profiles. Currently available methods for measuring TCR diversity are primarily limited to TCRVβ usage and are based on (i) detection by flow-cytometry using antibody panel against TCRVβ, (ii) CDR3 length spectratyping by semiquantitative reverse transcriptase PCR (RT-PCR), (iii) cDNA sequencing of the rearranged CDR3 regions, and (iv) deep sequencing of PCR-amplified rearranged genomic TCRVβ genes. Although these approaches can detect oligoclonal expansions of T cells, in aggregate, they are labor-intensive, can be costly, require relatively long preparation as well as assay and analysis times, depend on availability of detection antibodies, and can require substantial amount of samples (9–11, 14, 15, 35). In contrast, DTEA uses a panel of bar-coded probes to directly quantify 45 Vα and 46 Vβ mRNA species in a multiplexed assay so that changes in the diversity of both Vα and Vβ TCR chains can be simultaneously assessed in a highly sensitive, reproducible, and rapid fashion. DTEA is a direct measurement of the number of RNA transcripts found in the starting material (RNA or cells) by direct hybridization with bar-coded probes, without prior amplification required for cloning and sequencing and any unknown errors this may lead to. Moreover, because TCR sequencing involves random clone selection, of which a limited number may be feasibly sequenced, it is prone to both over- and underrepresenting clonotype frequencies in a given population of cells. Finally, the readout of the DTEA technique is the number of transcripts for each probe, therefore depicting the true frequency of the different TCR genes or chains found in an unmanipulated sample.

Identifying specific TCR chain usage from tumor-specific T-cell populations may reveal new TCR sequences that can be cloned and expressed in recombinant form, producing a new tumor-specific usage for adoptive cell transfer (28, 34). DTEA can rapidly assess TCRVα and Vβ usage of even low frequency clonotypes and potentially predict TCR pairing. It may therefore be a valuable tool for the determination of specific antitumor TCRs suitable for cloning and subsequent genetic modification of T cells. Available approaches to identifying paired TCRαβ-chains rely on T-cell staining with specific TCRVα or Vβ antibodies and subsequent analysis using flow cytometry. By sorting T cells expressing specific TCR Vα/Vβ pairs, the hypervariable CDR regions of both chains can be sequenced after gene cloning. However, a major limitation of this approach is the lack of antibodies recognizing many specific TCRVα or Vβ proteins. In contrast, DTEA uses probes specific for all currently known TCRVα and Vβ mRNA species enabling detection of all TCRVα and genes simultaneously. Using DTEA, we found that certain TCRVα- and Vβ-chains were expressed at similar frequencies in a given TIL population, suggesting possible pairings of these chains. This observation was made from a wide range of TILs from different patients and different Vα/Vβ combinations could be predicted in the samples.

Different T-cell clones may share the same TCRVβ gene but have a different CDR3 sequence, a region that interacts with antigenic peptides bound to MHC molecules. As of now, human T cells are known to express 48 TCRVβ genes, but investigators have identified or predicted a much larger number of CDR3 sequences (6). TCR cloning not only identifies TCRVβ usage but can also determine CDR3 diversity within the same Vβ-chain (9, 10, 36). In contrast, DTEA cannot distinguish T-cell clones that have the same TCRVβ, but different CDR3 sequences because the probes are specific for a given V region of each TCRVα and Vβ. This distinction may also explain the differences in the calculated frequencies of the same TCRVβ detected using DTEA versus the cloning methodology. Classical TCR gene cloning/sequencing of CDR3 could thus be considered as a complementary approach to DTEA. We do not intend for DTEA to be a substitute for DNA deep sequencing of genomic TCRVβ loci that can identify thousands of unique Vβ and CDR3 sequences in populations of T-cells. However, DTEA can readily aid investigators seeking a rapid, inexpensive, and nonlaborious method to measure TCRVα and Vβ usages. DTEA also complements deep sequencing and cloning/sequencing by first identifying the families of TCRVα and genes that are expressed enabling a focused and in-depth analysis by deep sequencing efforts to identify TCR usage in individual T-cell clones.

We speculate that DTEA will also prove valuable in other situations in which rapid, reliable, and highly sensitive monitoring of the TCRVα and gene usage is warranted, such as in patients who receive vaccinations, elderly patients, patients with autoimmune diseases and infections, and recipients undergoing hematopoietic stem cell transplantation. The single copy number sensitivity of DTEA will enable identification of a wide range of possible TCR clone frequencies, including those with frequencies as low as 1 in 1,000 when sufficient T cells are analyzed. Such low frequency clones associated with these TCR genes can later become dominant clones in different pathophysiologic states.

DTEA may thus provide insight into whether T-cell clones emerge from endogenous T-cell reconstitution or expand from a rare TIL present in the transfused cell populations following lymphodepletion and infusion of TIL. DTEA may be able to detect and monitor epitope spreading whereby the immunogenicity of a given immune target expands because of tissue damage resulting in the expansion of T cells (and the associated TCR usage) not initially engaged. In this case, the appearance of new highly expressed TCRVα and genes in blood samples may be a first indication of the expansion of new T-cell clones due to epitope or antigen spreading.

No potential conflicts of interest were disclosed.

Conception and design: M. Zhang, S. Maiti, B. Rabinovich, P. Hwu, L. Radvanyi, L.J.N. Cooper

Development of methodology: M. Zhang, S. Maiti, H. Huls, B. Rabinovich, L.M. Vence, P. Hwu, L. Radvanyi

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Zhang, S. Maiti, C. Bernatchez, P. Hwu, L. Radvanyi

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Zhang, S. Maiti, C. Bernatchez, L. Radvanyi

Writing, review, and/or revision of the manuscript: M. Zhang, S. Maiti, C. Bernatchez, B. Rabinovich, R.E. Champlin, L.M. Vence, P. Hwu, L. Radvanyi, L.J.N. Cooper

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Zhang, S. Maiti

Study supervision: M. Zhang, S. Maiti, R.E. Champlin, R.E. Champlin, L. Radvanyi

Developed TCR panels: S. Maiti

Supervised the experimental work: L. Radvanyi

The authors thank The University of Texas MD Anderson Cancer Center's DNA Analysis Facility for sequencing.

This work was supported in part by NIH grants CA111999, CA16672, CA124782, CA120956, CA141303, CA116127, 1S10RR026916; Burroughs Wellcome Fund; Cancer Prevention Research Institute of Texas; CLL Global Research Foundation; Department of Defense; Dr. Miriam and Sheldon Adelson Medical Research Foundation; Estate of Noelan L. Bibler; Harry T. Mangurian, Jr., Fund for Leukemia Immunotherapy; Gillson Longenbaugh Foundation; the Institute of Personalized Cancer Therapy; Melanoma Research Alliance; Miller Foundation; Mr. and Mrs. Joe H. Scales; National Foundation for Cancer Research; Pediatric Cancer Research Foundation; Sister Institution Network Fund; and William Lawrence and Blanche Hughes Children's Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Mak
TW
. 
The T cell antigen receptor: “The hunting of the Snark”
.
Eur J Immunol
2007
;
37
Suppl 1
:
S83
93
.
2.
Yassai
MB
,
Naumov
YN
,
Naumova
EN
,
Gorski
J
. 
A clonotype nomenclature for T cell receptors
.
Immunogenetics
2009
;
61
:
493
502
.
3.
Sims
JE
,
Tunnacliffe
A
,
Smith
WJ
,
Rabbitts
TH
. 
Complexity of human T-cell antigen receptor beta-chain constant- and variable-region genes
.
Nature
1984
;
312
:
541
5
.
4.
Hayday
AC
,
Diamond
DJ
,
Tanigawa
G
,
Heilig
JS
,
Folsom
V
,
Saito
H
, et al
Unusual organization and diversity of T-cell receptor alpha-chain genes
.
Nature
1985
;
316
:
828
32
.
5.
Akatsuka
Y
,
Martin
EG
,
Madonik
A
,
Barsoukov
AA
,
Hansen
JA
. 
Rapid screening of T-cell receptor (TCR) variable gene usage by multiplex PCR: application for assessment of clonal composition
.
Tissue Antigens
1999
;
53
:
122
34
.
6.
Freeman
JD
,
Warren
RL
,
Webb
JR
,
Nelson
BH
,
Holt
RA
. 
Profiling the T-cell receptor beta-chain repertoire by massively parallel sequencing
.
Genome Res
2009
;
19
:
1817
24
.
7.
Robins
HS
,
Campregher
PV
,
Srivastava
SK
,
Wacher
A
,
Turtle
CJ
,
Kahsai
O
, et al
Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta T cells
.
Blood
2009
;
114
:
4099
107
.
8.
Packer
AN
,
Muraro
PA
. 
Optimized clonotypic analysis of T-cell receptor repertoire in immune reconstitution
.
Exp Hematol
2007
;
35
:
516
21
.
9.
Robbins
PF
,
Dudley
ME
,
Wunderlich
J
,
El-Gamil
M
,
Li
YF
,
Zhou
J
, et al
Cutting edge: persistence of transferred lymphocyte clonotypes correlates with cancer regression in patients receiving cell transfer therapy
.
J Immunol
2004
;
173
:
7125
30
.
10.
Zhou
J
,
Dudley
ME
,
Rosenberg
SA
,
Robbins
PF
. 
Persistence of multiple tumor-specific T-cell clones is associated with complete tumor regression in a melanoma patient receiving adoptive cell transfer therapy
.
J Immunother
2005
;
28
:
53
62
.
11.
Jouvin-Marche
E
,
Fuschiotti
P
,
Marche
PN
. 
Dynamic aspects of TCRalpha gene recombination: qualitative and quantitative assessments of the TCRalpha chain repertoire in man and mouse
.
Adv Exp Med Biol
2009
;
650
:
82
92
.
12.
Gorski
J
,
Yassai
M
,
Zhu
X
,
Kissela
B
,
Kissella
B
,
Keever
C
, et al
Circulating T cell repertoire complexity in normal individuals and bone marrow recipients analyzed by CDR3 size spectratyping. Correlation with immune status
.
J Immunol
1994
;
152
:
5109
19
.
13.
Echchakir
H
,
Asselin-Paturel
C
,
Dorothee
G
,
Vergnon
I
,
Grunenwald
D
,
Chouaib
S
, et al
Analysis of T-cell-receptor beta-chain-gene usage in peripheral-blood and tumor-infiltrating lymphocytes from human non-small-cell lung carcinomas
.
Int J Cancer
1999
;
81
:
205
13
.
14.
Dokai
H
,
Nomura
Y
,
Fujikawa
Y
,
Nihei
K
,
Segawa
M
,
Shinomiya
N
. 
A study of the factors inducing the development of childhood-onset myasthenia gravis using CDR3 spectratyping analysis of the TCR repertoire
.
J Neuroimmunol
2007
;
187
:
192
200
.
15.
Beck
RC
,
Stahl
S
,
O'Keefe
CL
,
Maciejewski
JP
,
Theil
KS
,
Hsi
ED
. 
Detection of mature T-cell leukemias by flow cytometry using anti-T-cell receptor V beta antibodies
.
Am J Clin Pathol
2003
;
120
:
785
94
.
16.
Ikeda
H
,
Sato
N
,
Matsuura
A
,
Sasaki
A
,
Takahashi
S
,
Kozutsumi
D
, et al
Clonal dominance of human autologous cytotoxic T lymphocytes against gastric carcinoma: molecular stability of the CDR3 structure of the TCR alphabeta gene
.
Int Immunol
1996
;
8
:
75
82
.
17.
Zhou
J
,
Dudley
ME
,
Rosenberg
SA
,
Robbins
PF
. 
Selective growth, in vitro and in vivo, of individual T cell clones from tumor-infiltrating lymphocytes obtained from patients with melanoma
.
J Immunol
2004
;
173
:
7622
9
.
18.
Buchholz
VR
,
Neuenhahn
M
,
Busch
DH
. 
CD8+ T cell differentiation in the aging immune system: until the last clone standing
.
Curr Opin Immunol
2011
;
23
:
549
54
.
19.
Vallejo
AN
. 
Age-dependent alterations of the T cell repertoire and functional diversity of T cells of the aged
.
Immunol Res
2006
;
36
:
221
8
.
20.
Pappas
J
,
Jung
WJ
,
Barda
AK
,
Lin
WL
,
Fincke
JE
,
Purev
E
, et al
Substantial proportions of identical beta-chain T-cell receptor transcripts are present in epithelial ovarian carcinoma tumors
.
Cell Immunol
2005
;
234
:
81
101
.
21.
Plasilova
M
,
Risitano
A
,
Maciejewski
JP
. 
Application of the molecular analysis of the T-cell receptor repertoire in the study of immune-mediated hematologic diseases
.
Hematology
2003
;
8
:
173
81
.
22.
Codina-Busqueta
E
,
Scholz
E
,
Munoz-Torres
PM
,
Roura-Mir
C
,
Costa
M
,
Xufre
C
, et al
TCR bias of in vivo expanded T cells in pancreatic islets and spleen at the onset in human type 1 diabetes
.
J Immunol
2011
;
186
:
3787
97
.
23.
Rosenberg
SA
,
Yang
JC
,
Sherry
RM
,
Kammula
US
,
Hughes
MS
,
Phan
GQ
, et al
Durable complete responses in heavily pretreated patients with metastatic melanoma using T-cell transfer immunotherapy
.
Clin Cancer Res
2011
;
17
:
4550
7
.
24.
Dudley
ME
,
Wunderlich
JR
,
Yang
JC
,
Sherry
RM
,
Topalian
SL
,
Restifo
NP
, et al
Adoptive cell transfer therapy following non-myeloablative but lymphodepleting chemotherapy for the treatment of patients with refractory metastatic melanoma
.
J Clin Oncol
2005
;
23
:
2346
57
.
25.
Dudley
ME
,
Wunderlich
JR
,
Robbins
PF
,
Yang
JC
,
Hwu
P
,
Schwartzentruber
DJ
, et al
Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes
.
Science
2002
;
298
:
850
4
.
26.
Robins
HS
,
Srivastava
SK
,
Campregher
PV
,
Turtle
CJ
,
Andriesen
J
,
Riddell
SR
, et al
Overlap and effective size of the human CD8+ T cell receptor repertoire
.
Sci Transl Med
2010
;
2
:
47ra64
.
27.
Geiss
GK
,
Bumgarner
RE
,
Birditt
B
,
Dahl
T
,
Dowidar
N
,
Dunaway
DL
, et al
Direct multiplexed measurement of gene expression with color-coded probe pairs
.
Nat Biotechnol
2008
;
26
:
317
25
.
28.
Johnson
LA
,
Heemskerk
B
,
Powell
DJ
 Jr
,
Cohen
CJ
,
Morgan
RA
,
Dudley
ME
, et al
Gene transfer of tumor-reactive TCR confers both high avidity and tumor reactivity to nonreactive peripheral blood mononuclear cells and tumor-infiltrating lymphocytes
.
J Immunol
2006
;
177
:
6548
59
.
29.
Malkov
VA
,
Serikawa
KA
,
Balantac
N
,
Watters
J
,
Geiss
G
,
Mashadi-Hossein
A
, et al
Multiplexed measurements of gene signatures in different analytes using the Nanostring nCounter Assay System
.
BMC Res Notes
2009
;
2
:
80
.
30.
de Jonge
HJ
,
Fehrmann
RS
,
de Bont
ES
,
Hofstra
RM
,
Gerbens
F
,
Kamps
WA
, et al
Evidence based selection of housekeeping genes
.
PLoS One
2007
;
2
:
e898
.
31.
Besser
MJ
,
Shapira-Frommer
R
,
Treves
AJ
,
Zippel
D
,
Itzhaki
O
,
Hershkovitz
L
, et al
Clinical responses in a phase II study using adoptive transfer of short-term cultured tumor infiltration lymphocytes in metastatic melanoma patients
.
Clin Cancer Res
2010
;
16
:
2646
55
.
32.
Brentjens
RJ
,
Riviere
I
,
Park
JH
,
Davila
ML
,
Wang
X
,
Stefanski
J
, et al
Safety and persistence of adoptively transferred autologous CD19-targeted T cells in patients with relapsed or chemotherapy refractory B-cell leukemias
.
Blood
2011
;
118
:
4817
28
.
33.
Hunder
NN
,
Wallen
H
,
Cao
J
,
Hendricks
DW
,
Reilly
JZ
,
Rodmyre
R
, et al
Treatment of metastatic melanoma with autologous CD4+ T cells against NY-ESO-1
.
N Engl J Med
2008
;
358
:
2698
703
.
34.
Morgan
RA
,
Dudley
ME
,
Yu
YY
,
Zheng
Z
,
Robbins
PF
,
Theoret
MR
, et al
High efficiency TCR gene transfer into primary human lymphocytes affords avid recognition of melanoma tumor antigen glycoprotein 100 and does not alter the recognition of autologous melanoma antigens
.
J Immunol
2003
;
171
:
3287
95
.
35.
Attuil
V
,
Bucher
P
,
Rossi
M
,
Mutin
M
,
Maryanski
JL
. 
Comparative T cell receptor repertoire selection by antigen after adoptive transfer: a glimpse at an antigen-specific preimmune repertoire
.
Proc Natl Acad Sci U S A
2000
;
97
:
8473
8
.
36.
Maryanski
JL
,
Attuil
V
,
Bucher
P
,
Walker
PR
. 
A quantitative, single-cell PCR analysis of an antigen-specific TCR repertoire selected during an in vivo CD8 response: direct evidence for a wide range of clone sizes with uniform tissue distribution
.
Mol Immunol
1999
;
36
:
745
53
.

Supplementary data