TCRs mediate CTL specificity, but TCRs recognizing the same epitope often differ between persons due to their stochastic derivation. The role of this variability in the pathogenesis of virus infections and malignancies has been technically difficult to study. We apply an adaptation of TCR spectratyping to study HIV-specific CTLs, defining the clonal breadth and sequences of epitope-specific TCRs from PBMCs without cellular sorting or molecular cloning. Examining 48 CTL responses in 12 persons reveals a mean of 4.5 ± 2.7 clones per response, of both public and private clonotypes. The number of identified epitope-specific TCRs correlates with CTL frequency across epitopes, suggesting that clonal breadth limits the magnitude of the CTL response against HIV-1 in vivo. HLA A- and B-restricted CTLs are similar in their TCR breadth in this small cohort, preliminarily suggesting that qualitative differences may account for their disparate impacts on pathogenesis. Overall, these findings demonstrate that the magnitude of the CTL response in chronic HIV-1 infection is constrained by TCR clonal breadth, suggesting maximal expansion of CTLs in response to chronic antigenic stimulation.

CD8+ T lymphocytes play a key role in malignancies and viral infections through recognition of epitopes presented on MHC class I, resulting in clearance of target cells. Their specificity is mediated through the TCR, an Ig-like molecule that is generated through random recombination and splicing of TCR variable genes to produce α- and β-chains that pair as heterodimers. This stochastic process can generate multiple distinct TCRs against any given epitope/MHC complex (1).

The influence of the number of TCRs and/or their sequences in the control of malignancies or viral infections remains poorly understood. For situations where epitope variability is high, such as HIV-1 infection, it has been proposed that a greater breadth of TCRs against a given epitope may be beneficial for immune containment (2), because each TCR can have a distinct pattern of promiscuity in recognizing epitope variants (3). However, the number and sequences of TCRs generated against a specific epitope can vary greatly between persons; there can be “public clonotypes” defined as commonly shared variable gene usages for TCRs between individuals, as well as “private clonotypes” that are specific to certain individuals (4). In some situations, such as inbred mice infected with cloned virus strains, the CTL response is consistently and predictably dominated by public TCR clonotypes (5). In contrast, TCR clonotypes and numbers against the same epitope may be highly variable between individuals, such as in human HIV-1 infection (1).

CTLs play a key protective role in HIV-1 infection, and it has been proposed that TCR breadth may have an important influence on HIV-1 immunopathogenesis. TCR numbers and sequences can vary greatly between infected persons, even in the case of identical twins infected with the same strain of virus (1, 6), although for some epitopes there are clear patterns of public clonotype usage (69). The best currently available approaches for quantitating epitope-specific TCR breadth and sequences, however, are highly labor-intensive, using either CTL cloning followed by TCR sequencing (9) or construction of peptide–MHC tetrameric complexes for flow cytometric cell sorting followed by limiting dilution molecular cloning of TCR sequences (2, 8, 10).

In this study we use a modification of a molecular tool previously termed spectratyping or immunoscoping, initially developed by Pannetier et al. (11) to analyze size distributions qualitatively within TCR variable gene families as a global analysis of TCR diversity, and modified by Uko et al. (12) and Killian et al. (13) to examine Ag-specific expansions after in vitro stimulation with proteins or peptides. Improving this assay to examine TCRs in a quantitative manner, it is possible to identify objectively epitope-specific TCRs after epitope stimulation in vitro and isolate their sequences often without any cellular or molecular cloning. The ability of this approach to quantitate epitope-specific TCRs and isolate their sequences within the peripheral blood cells of HIV-1–infected persons is explored. Finally, it is applied to study a central question in HIV-1 pathogenesis regarding the limitation of CTL frequency during infection.

Blood samples were collected from healthy control and chronically HIV-1–infected volunteers under University of California Los Angeles Institutional Review Board-approved protocols. PBMCs were isolated by Ficoll gradient and washed twice with Hanks’ buffered saline, then viably cryopreserved.

CTL responses against previously defined HIV-1 optimal epitopes (14) were assessed using standard IFN-γ ELISPOT assay screening of polyclonal CD8+ T cells as previously described (15, 16). Spot-forming cells were enumerated with an automated ELISPOT reader (Autoimmun Diagnostika, Strassberg, Germany) and expressed as cells per million CD8+ T lymphocytes.

PBMCs (106) were cultured in 2 ml RPMI 1640 medium (Sigma-Aldrich, St. Louis, MO) supplemented with 10% heat-inactivated FCS, 2 mM l-glutamine, 50 U/ml penicillin, 50 μg/ml streptomycin, and 12.5 U/ml recombinant human IL-2 (National Institutes of Health AIDS Research and Reference Reagent Repository), with 1 μg/ml of the indicated epitope peptides (Sigma). After 3 or 4 d, the cultures were fed by replacing 1 ml medium. After 7 d, the CD8+ T lymphocytes were isolated using positive immunomagnetic bead selection (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s protocol.

The total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA), and reverse-transcribed to cDNA using random primers (high-capacity reverse transcription kit; Applied Biosystems, Carlsbad, CA). Using these cDNA, real-time PCR (IQ5; Bio-Rad, Hercules, CA) was performed for each BV family. The nomenclature for TCR BV families and genes follows that of the International ImMunoGeneTics database (17, 18). Each BV-specific amplification was performed separately using one of the forward primers given in Supplemental Table II, with the C region-specific 5′-labeled probe Cy5-TGTTCCCACCCGAGGTCGC-BHQ2 and the C region-specific reverse primer 5′-CTTCTGATGGCTCAAACAC-3′ that was fluorescence tagged as described in Supplemental Table III. These PCR reactions were performed using IQ-Supermix (Bio-Rad) for 35 cycles under the following conditions: denaturation at 95°C (30 s), annealing at 55°C (30 s), 72°C (1 min). A standard control plasmid was prepared by cloning BV20 gene product into TOPO TA cloning vector (Invitrogen); this standard was run in parallel. Based on the relationship of Ct (threshold cycles to detection) to control copy number, the number of starting copies within each BV PCR reaction was estimated. The normalized relative concentration of each family was then calculated as the ratio of copies to the median number of copies across all families.

The PCR products from each BV family were resolved by capillary electrophoresis and detection of the dye-labeled Vβ-specific primers (3130 Genetic Analyzer; Applied Biosystems), after being combined in run groups as listed in Supplemental Table III. The resulting histograms for each family were analyzed with GeneMapper v3.7 (Applied Biosystems) to determine the area under the curve of each size peak. The relative concentration of each size peak within the family was determined first by calculating the fraction of each size within its family (ratio of its peak area to the sum of peak areas for that family), then multiplying the relative concentration of the total family by this fraction to determine the relative concentration for the peak.

The PCR-amplified product of BV families, which showed epitope-specific TCR–BV size-peak expansions, were selected and PCR purified (Invitrogen). Using this purified BV family, bulk PCR product dye-terminator sequencing PCR was performed with a TCR β-chain constant region reverse primer (5′-CTTCTGATGGCTCAAACAC-3′), and the data were analyzed by an ABI-3130 capillary genetic analyzer (Applied Biosystems).

Simpson’s diversity index (19, 20) was applied to measure TCR clonal diversity for each epitope-specific CTL response. The diversity index (DS) was calculated as

DS=1i=1cni(ni1)N(N1)
,

where c is the number of identified BV peak expansions, and for each expansion (ith expansion) ni is the relative concentration of that expansion (increase in concentration units as defined by the quantitative spectratype measurement, as shown in Fig. 2B), and N is the sum of changes in relative concentrations across all BV expansions.

FIGURE 2.

Enumeration of epitope-specific TCR by quantitative spectratyping after epitope-specific CTL expansion. PBMCs from subject 00048 were stimulated with the HLA B*15-restricted HIV-1 Gag epitope GLNKIVRMY or passaged without peptide stimulation (subject 00048 had been found to have a CTL response against this epitope; Supplemental Table I). Quantitative spectratyping was performed on purified CD8+ T lymphocytes, comparing the cells without peptide (Unstimulated) or with peptide (Stimulated). A, Relative concentrations of TCR size peaks for 13 BV families are plotted. B, The change in each of these peaks in the absence and presence of epitope stimulation is plotted for the same 13 families. Size peaks no. 4 in BV11 and no. 2 in BV29 were statistically significant expansions, based on control experiments defining 4 SDs as 1.04 U (Supplemental Fig. 1). C, Sequencing histograms from the bulk BV family PCR product for two families with expansions (BV11 and BV29) are shown, indicating clonal overgrowth of the population with a single BV sequence. D, Sequencing histograms from the bulk BV family PCR product for two families without expansions (BV19 and BV20) are shown, demonstrating mixed sequences.

FIGURE 2.

Enumeration of epitope-specific TCR by quantitative spectratyping after epitope-specific CTL expansion. PBMCs from subject 00048 were stimulated with the HLA B*15-restricted HIV-1 Gag epitope GLNKIVRMY or passaged without peptide stimulation (subject 00048 had been found to have a CTL response against this epitope; Supplemental Table I). Quantitative spectratyping was performed on purified CD8+ T lymphocytes, comparing the cells without peptide (Unstimulated) or with peptide (Stimulated). A, Relative concentrations of TCR size peaks for 13 BV families are plotted. B, The change in each of these peaks in the absence and presence of epitope stimulation is plotted for the same 13 families. Size peaks no. 4 in BV11 and no. 2 in BV29 were statistically significant expansions, based on control experiments defining 4 SDs as 1.04 U (Supplemental Fig. 1). C, Sequencing histograms from the bulk BV family PCR product for two families with expansions (BV11 and BV29) are shown, indicating clonal overgrowth of the population with a single BV sequence. D, Sequencing histograms from the bulk BV family PCR product for two families without expansions (BV19 and BV20) are shown, demonstrating mixed sequences.

Close modal

Calculations of means, medians, standard deviations, and linear correlations were performed using Excel software (Microsoft, Redmond, WA), and plots were produced using Excel or GraphPad Prism (version 3) statistical software (GraphPad Software, La Jolla, CA).

Spectratyping of TCRs is an approach that has allowed gross estimations of alterations in T cell populations based on subjective “perturbations” of size distribution profiles (21, 22) but not finer analysis of the Ag-specific T cell expansions driving the changes. We developed a spectratyping-based method that allows rapid and objective identification of epitope-specific TCRs and isolation of TCR sequences (Fig. 1A). Prior work had shown that the spectratype-identified BV distribution of TCRs that were enriched in this manner closely reflected BV usage as demonstrated by direct BV staining of peptide–MHC tetramer–labeled CTL (13), although identification of such expansions was subjective. Now, real-time RT-PCR was used to calculate the relative concentration of each variable gene family (Fig. 1B), followed by capillary electrophoretic size resolution of TCRs within each family and determination of the relative concentration of each size-subset within the entire TCR population (Fig. 1C).

FIGURE 1.

Scheme for quantitative TCR spectratyping. A, RNA is extracted from purified CD8+ T lymphocytes, followed by reverse transcription. Real-time PCR is then performed to quantitate the absolute number of isolated RNA transcripts for each of the 24 BV families. These values are then used to calculate the relative concentration of each family, normalized as the ratio of the absolute concentration of each family to the median absolute concentration of all families. Each family is resolved to reveal size distributions of TCRs within the family. Based on the contribution of each size peak within each family, the relative concentration of each peak is calculated from its percentage of the total family relative concentration. B, Mean relative concentrations of each family from the PBMCs of seven healthy control subjects are plotted, demonstrating the reproducibility of these measurements across normal individuals. Error bars represent SEs. C, Mean relative concentrations of each size peak within the first 12 BV families for the seven control subjects are plotted. Error bars represent SEs.

FIGURE 1.

Scheme for quantitative TCR spectratyping. A, RNA is extracted from purified CD8+ T lymphocytes, followed by reverse transcription. Real-time PCR is then performed to quantitate the absolute number of isolated RNA transcripts for each of the 24 BV families. These values are then used to calculate the relative concentration of each family, normalized as the ratio of the absolute concentration of each family to the median absolute concentration of all families. Each family is resolved to reveal size distributions of TCRs within the family. Based on the contribution of each size peak within each family, the relative concentration of each peak is calculated from its percentage of the total family relative concentration. B, Mean relative concentrations of each family from the PBMCs of seven healthy control subjects are plotted, demonstrating the reproducibility of these measurements across normal individuals. Error bars represent SEs. C, Mean relative concentrations of each size peak within the first 12 BV families for the seven control subjects are plotted. Error bars represent SEs.

Close modal

This method was applied to CD8+ T cells from PBMCs cultured without or with HIV-1 epitopes to elicit epitope-specific CTL enrichment in vitro. Comparison of TCR size distributions in the control unstimulated versus stimulated cells (Fig. 2A, 2B) demonstrated clear epitope-specific expansions of TCR size-subsets in CD8+ T cells from persons who had been defined previously to have CTL responses by IFN-γ ELISPOT and/or chromium release cytolysis assay (Supplemental Table I). Control experiments with cells from HIV-1–uninfected persons revealed a mean change in TCR size-subsets across all BV families of 0.00 (SD, 0.26; range, −4.18–3.24; Supplemental Fig. 1); based on these data, a statistical cutoff of 4 SDs (1.04 relative concentration units) increase was chosen as a criterion for significant expansions of size-subset peaks within a BV family after epitope stimulation. Control spiking experiments indicated that an increase of one relative concentration unit was equivalent to ∼0.21% added clone (data not shown). Consistent with clonal TCR expansions, sequence analysis of bulk BV families (the raw RT-PCR product used for spectratype analysis of a whole family) containing epitope-associated expansions revealed a clearly dominant sequence compared with families where there were no epitope-associated expansions (Fig. 2C, 2D). Of 92 BV families containing peak expansions of ≥3 relative concentration units after epitope stimulation, analysis of the bulk PCR product revealed clear sequences for 75 BV families (Supplemental Table IV). In some cases, sequencing of the bulk BV family PCR product did not reveal a single sequence due to multiple expansions in the same family, but dominant sequences could be identified by cloning (data not shown). As expected, families with two expansions yielded no readable sequence from bulk sequencing when the expansions were similar in magnitude (e.g., subject 00011, epitope TSTLQEQIGW, BV05a; Supplemental Table IV) but yielded a single sequence when one expansion was much larger (e.g., subject 00016, epitope KELYPLASL, BV04; Supplemental Table IV). Overall, these findings suggested that each peak expansion after stimulation was due to a single dominant clonal CTL expansion. As a whole, these data indicate that epitope-specific TCRs can be identified rapidly with this quantitative spectratyping approach, and additionally that TCR sequences can often be obtained without the need for molecular or cellular cloning.

This method was applied to the analysis of HIV-1–specific CTL responses in persons with chronic untreated infection. Examining 48 CTL responses against HIV-1 minimal epitopes from 12 persons, which had been identified with IFN-γ ELISPOT assays (Supplemental Table I), TCR BV usage was diverse both within and between individuals (Fig. 3). An average of 4.5 ± 2.7 (mean ± SD) peak expansions in 3.6 ± 2.0 BV families was noted for each CTL response. For two responses, expansions were not identified, because of either the stringency of our criteria for positivity or CTL usage of a BV gene that was not amplified by the primers in our spectratyping assay. BV gene usage was comprehensive, spanning 23 of the 24 assayed families. CTL responses against different epitopes within the same person varied in their BV usage, indicating the specificity of detection. These data demonstrated the capability of the approach to identify epitope-specific TCR across a broad repertoire of BV families.

FIGURE 3.

Diversity of TCR BV usage for HIV-1–specific CTL responses. Epitope-specific quantitative spectratyping was performed for 48 CTL responses in 12 persons with chronic untreated HIV-1 infection. BV usage for each of these responses is indicated (top panel), as well as the frequency for which each BV family was used across all responses (bottom panel).

FIGURE 3.

Diversity of TCR BV usage for HIV-1–specific CTL responses. Epitope-specific quantitative spectratyping was performed for 48 CTL responses in 12 persons with chronic untreated HIV-1 infection. BV usage for each of these responses is indicated (top panel), as well as the frequency for which each BV family was used across all responses (bottom panel).

Close modal

Three prior reports have described Vβ usage and sequences for CTL responses against an immunodominant HLA B*57-restricted CTL response against an HIV-1 Gag p24 epitope (KF11); for comparison, we examined our results for two persons (subjects 00036 and 00052) with responses against this epitope (Fig. 4). Subject 00036 PBMCs demonstrated three expansions in families BV11, BV18, and BV19; subject 00052 PBMCs demonstrated four expansions in families BV6, BV19 (two different peaks), and BV25. This shared BV19 usage was similar to data from Gillespie et al. (9) and Yu et al. (7), who found predominance of BV19 (Vβ17 in serologic nomenclature) usage for this response, although in contrast Simons et al. (8) found BV19 usage in only one of eight persons and predominant BV07 usage. Direct sequence analysis of the bulk BV family PCR products used for spectratyping revealed sequences for six of the seven expansions (only the dominant sequence of the two expansions in BV19 from subject 00052 was revealed by bulk sequencing; Supplemental Table IV). Of these six identified CDR3 sequences, the two BV19 CDR3 sequences from subjects 00036 and 00052 differed by a single amino acid insertion/deletion and were very similar or identical to CDR3 sequences reported in the three prior studies of KF11–specific TCR sequences (Fig. 4A). Furthermore, we observed other BV gene usages (Fig. 4B), including BV 6.1, 11.2, 18, and BV 25.1, the latter two being usages that were not reported previously. Finally, using a highly immunodominant A*0201-restricted influenza epitope, this method also identified public clonotype TCR usage matching the known pattern for that epitope (data not shown). Overall, these results indicate that our approach yields TCR sequence data that are consistent with prior reports, as well as additional TCR sequences that are unique to our study subjects, validating this method to assess TCRs.

FIGURE 4.

Public and private BV usages and CDR3 motifs of TCR recognizing an immunodominant HLA B*57-restricted CTL epitope. TCR BV usages and CDR sequences were determined for CTL responses against the immunodominant B*57-restricted Gag epitope KAFSPEVIPMF. A, CDR3 sequences and β-chain joining region (BJ) usages of two BV19 TCRs are aligned against similar TCR sequences that were previously reported by Gillespie et al. (9), Yu et al. (7), and Simons et al. (8). B, BV usages, CDR3 sequences, and BJ usages are given for four TCRs that are dissimilar to previously reported TCRs recognizing this epitope.

FIGURE 4.

Public and private BV usages and CDR3 motifs of TCR recognizing an immunodominant HLA B*57-restricted CTL epitope. TCR BV usages and CDR sequences were determined for CTL responses against the immunodominant B*57-restricted Gag epitope KAFSPEVIPMF. A, CDR3 sequences and β-chain joining region (BJ) usages of two BV19 TCRs are aligned against similar TCR sequences that were previously reported by Gillespie et al. (9), Yu et al. (7), and Simons et al. (8). B, BV usages, CDR3 sequences, and BJ usages are given for four TCRs that are dissimilar to previously reported TCRs recognizing this epitope.

Close modal

CTL responses are subject to ongoing antigenic stimulation throughout untreated chronic HIV-1 infection, and the typically exhausted phenotype of HIV-1–specific CTLs suggests that they are driven to maximal expansion and proliferation. Given this scenario, we reasoned that the number of clones recognizing an epitope thus would be the limiting factor in the magnitude of the response against that epitope. Comparing the number of TCR expansions identified by quantitative spectratyping to the magnitude of epitope-specific CTL responses identified by standard IFN-γ ELISPOT assay, there was a positive correlation (Fig. 5A; r2 = 0.20, p = 0.001). On average, there were 168 ± 178 (±SD) detected CTLs per 106 CD8+ T cells for each TCR-Vβ size-peak expansion. Moreover, the diversity of TCR usage, as calculated by Simpson’s diversity index (19, 20), was directly related to the number of identified TCR expansions (Fig. 5B; r2 = 0.35, p < 0.0001), indicating that the increase in CTL magnitude generally was not dominated by biased usages of TCR subsets (because this diversity index takes into account not only the number of different TCR but also their distribution). These data demonstrate that greater epitope-specific TCR breadth is associated with greater magnitude and diversity of the CTL response, consistent with clonal expansion being a limiting factor for the response.

FIGURE 5.

Correlations of TCR genetic breadth with CTL frequency and TCR diversity. The numbers of TCR BV expansions for each CTL response were assessed for roles in CTL magnitude and distribution of TCR usage. A, Numbers of expansions are plotted against CTL frequency for each epitope. p = 0.001. B, Numbers of expansions are plotted against the diversity of TCR usage for each epitope as defined by Simpson’s diversity index, taking into account the distribution of individual TCR expansion magnitudes. p < 0.0001.

FIGURE 5.

Correlations of TCR genetic breadth with CTL frequency and TCR diversity. The numbers of TCR BV expansions for each CTL response were assessed for roles in CTL magnitude and distribution of TCR usage. A, Numbers of expansions are plotted against CTL frequency for each epitope. p = 0.001. B, Numbers of expansions are plotted against the diversity of TCR usage for each epitope as defined by Simpson’s diversity index, taking into account the distribution of individual TCR expansion magnitudes. p < 0.0001.

Close modal

Because HLA class I A-restricted and B-restricted CTL responses appear to have differential influence on the pathogenesis of HIV-1 infection (23), TCRs of A-restricted and B-restricted responses were compared. The mean numbers of detected TCR expansions for A-restricted versus B-restricted CTL responses were similar (Fig. 6A). The mean ratios of CTL frequencies to detected TCR expansions also were similar, suggesting a comparable degree of CTL expansion per clone (Fig. 6B). Furthermore, the mean diversities of TCRs were similar between A-restricted and B-restricted CTL responses, although there was a trend for more diversity in the latter (Fig. 6C). Overall, these findings suggested that B-restricted CTL responses are not markedly different from A-restricted responses in terms of TCR breadth or clonal expansion of CTLs within this small group of HIV-1–infected persons, although further quantitative measurements of TCRs in larger cohorts will be required to confirm similarity or difference in TCR breadth of HLA A-restricted versus HLA B-restricted HIV-1–specific CTL responses.

FIGURE 6.

HLA A-restricted and HLA B-restricted CTL responses have similar TCR breadth and relationships of TCR to CTL frequency. Data from the HLA A- and B-restricted CTL responses are compared. A, The numbers of TCR BV expansions associated with each epitope are compared. Means are indicated (4.4 versus 4.5 for A-restricted versus B-restricted CTL responses, respectively). B, The ratios of CTL frequency (determined by IFN-γ ELISPOT assays) to TCR BV expansions for each epitope are compared. Means are indicated (163 versus 169 cells per million CD8+ T lymphocytes for A-restricted versus B-restricted CTL responses, respectively). C, The diversities of TCR usage (Simpson’s diversity index) for each epitope are compared. Means are indicated (0.39 versus 0.52 for A-restricted versus B-restricted CTL responses, respectively).

FIGURE 6.

HLA A-restricted and HLA B-restricted CTL responses have similar TCR breadth and relationships of TCR to CTL frequency. Data from the HLA A- and B-restricted CTL responses are compared. A, The numbers of TCR BV expansions associated with each epitope are compared. Means are indicated (4.4 versus 4.5 for A-restricted versus B-restricted CTL responses, respectively). B, The ratios of CTL frequency (determined by IFN-γ ELISPOT assays) to TCR BV expansions for each epitope are compared. Means are indicated (163 versus 169 cells per million CD8+ T lymphocytes for A-restricted versus B-restricted CTL responses, respectively). C, The diversities of TCR usage (Simpson’s diversity index) for each epitope are compared. Means are indicated (0.39 versus 0.52 for A-restricted versus B-restricted CTL responses, respectively).

Close modal

Defining the genetic composition of T cell responses has important implications for pathogenesis studies of viral infections and malignancies, and isolating virus- or tumor-specific TCR sequences for gene transfer holds promise for immunotherapeutics. To date, however, approaches have been limited to either 1) deriving T cell clones for TCR sequencing or 2) sorting epitope-specific T cells using custom peptide–HLA tetrameric complexes, followed by limiting dilution cloning of TCRs. The first approach clearly is limited by the technical challenges, laborious nature, and bias of deriving T cell clones; the second approach requires synthesis of peptide–HLA tetrameric complexes for each epitope of interest, followed by labor-intensive molecular cloning and screening of TCRs. In the current study, we adapt the TCR spectratyping assay, which has been used qualitatively to identify alterations in TCR populations (demonstrated as “perturbations” from the expected native Gaussian distribution of TCR sizes within each family), and which has been used subjectively to examine epitope-specific TCR expansions (1, 13). Taking advantage of epitope-specific expansion of T cell clones and developing a formally quantitative spectratyping measurement of TCR expansions, we find that epitope-specific TCR genes can be sequenced directly from the BV family PCR products used for spectratyping.

The use of quantitative spectratyping to analyze epitope-specific expansions within PBMCs has the potential to allow rapid examination of TCR clonal breadth and TCR sequence determination for any epitope-specific CTL response, requiring only synthetic peptide as a reagent (rather than peptide–MHC construction). This information can be applied to pathogenesis studies examining the role of TCR breadth, or TCR sequence. Moreover, combining TCR α-chain (data not shown) with β-chain spectratyping allows reconstruction of the complete TCR sequences, and comparisons of degrees of α-chain and β-chain expansion may help determine the proper pairings of individual chains. Thus, this method promises to facilitate rapid isolation of epitope-specific TCRs such as those being developed for gene therapeutic approaches to viral infections and cancer.

Our methodology is a significant technical advance over similar technologies that have been reported previously. For example, Naumov et al. (24, 25) previously examined the diversity of TCR for CTL responses against immunodominant A*02-restricted influenza epitope using a similar approach. Those studies also used spectratyping after epitope stimulations of PBMCs to identify clonal TCR expansions but required a subjective assessment for increases in size populations within variable gene families. Furthermore, this group’s approach to obtaining TCR sequences involved generation of CTL clones or highly enriched epitope-specific cell lines after weeks of peptide stimulations, followed by variable gene-specific RT-PCR and molecular cloning before sequencing. Our current study is also a marked advance over our prior approach to identifying epitope-specific spectratype peaks, which was not quantitative (using lack of expansion in presence of 5-fluorouracil as a control for subjective assessments) and did not pursue TCR sequencing (13).

Our data demonstrate the application of this tool to clarifying a key aspect of HIV-1 immunopathogenesis. The factor(s) determining the magnitude of the HIV-1–specific CTL response during chronic infection are poorly understood; despite persistent viremia, this response is often lower in frequency than that of CMV-specific CTLs in chronic human CMV infection, which averages ∼10% of all blood CD8+ T cells (26). We find that the clonal breadth of epitope-specific CTL is correlated with the overall frequency of CTL against that epitope. This suggests that the number of clones raised against an HIV-1 epitope is the limiting factor in the magnitude of the response against that epitope. This hypothesis is consistent with observations that HIV-1–specific CTLs have an exhausted and ineffectual phenotype (27), are dominated by effector memory and terminal effector differentiation (28, 29), have limited proliferative capacity (30), and face ongoing vigorous antigenic stimulation (as reflected by persistent viremia).

Our observation that the magnitude of a CTL response generally reflects the TCR breadth and diversity comprising that response appears to contradict the conclusions of prior work suggesting that HIV-1–specific CTL expansions are biased. Wilson et al. (31) found that some expansions of whole BV families are composed of clonal expansions within these families, indicating that some clonal expansions are large enough to skew entire families. However, our results can be reconciled with these findings. Our data demonstrate a general trend across numerous CTL responses (Fig. 5A); there clearly is biologic variability for any given clone in regard to expansion capacity, with outliers that expand to higher levels. Wilson et al. took the approach of analyzing skewed families (rather than defining all BV usages for individual CTL responses) and therefore preferentially focused only on the greatest expansions. Our conclusions also differ somewhat from those of a study by Meyer-Olson et al. (32), which suggested that the clonal breadth of the TCR repertoire against an epitope is flexible. That study examined a single Nef epitope response, finding that it was composed of three to eight different BV families in eight persons not on antiretroviral therapy. In longitudinal assessment of three persons, two of whom had partially controlled viremia after structured treatment interruptions during acute infection, they found that spontaneously increased viremia correlated with increased clonal breadth of CTLs against that epitope and concluded that there is flexibility of TCR clonal breadth. However, these increases in clonal breadth appeared to be single TCR expansions in all three persons, and the majority of the epitope-specific repertoire was stable. Thus, although their data suggested that there is some flexibility in the TCR repertoire against that epitope, they still are consistent with a general limitation in clonal breadth. Finally, Yang et al. (33) showed that epitopic TCR breadth could be broadened by therapeutic vaccination of persons with chronic HIV-1 infection, seemingly contradicting our hypothesis that clonal breadth of TCRs is limited chronic infection. The likely reason for the difference from our findings is that the subjects in the vaccination study were receiving successful antiretroviral therapy, which restored CD4+ T cell help to better maintain the proliferative capacity of CTLs including previously exhausted or subdominant clones.

It should be noted that there are potential limitations and caveats to our approach. Our inability to detect TCR for two CTL responses confirmed by ELISPOT suggests that sensitivity is not 100%, indicating limitations to the methodology. First, there is potential bias in the detection of epitope-specific TCRs. Detection is limited to CTL clones that can expand in response to epitope stimulation, but chronic antigenic stimulation (such as occurs during HIV-1 infection) can lead to senescent CTLs that fail to proliferate (34). Also, our statistical criteria for defining a positive expansion (increase of 1.04 expansion units = 4 SDs defined in negative controls) are stringent to minimize false positives across the 205 size-peaks per spectratype. We have detected clonal sequences from bulk BV family PCR products when they have contained smaller expansions and therefore likely are underestimating TCR usage. A further issue may be the epitope sequence and/or peptide concentration used for stimulations. The in vivo HIV-1 epitope sequence may differ from that used for in vitro stimulations and fail to elicit CTL proliferation. Conversely, the high concentrations of peptide could allow cross-reactive detection of resting memory CTLs against epitope variants that previously escaped in vivo. Moreover, the supraphysiologic concentrations used for in vitro stimulation also could give aberrant TCR signaling that fails to stimulate proliferation or even elicits apoptosis. These issues all could contribute to bias in detection of TCR variable gene usage, although others have found that in vitro epitope stimulations appear to yield minimal bias in CTL expansions (8, 35). Finally, it should be noted that the delineation of TCR sequences from bulk BV PCR products may sometimes require cloning due to multiple expansions in the same family and that even a single expanded size-peak could contain more than one sequence, because there can be multiple TCR against a single epitope that vary only by amino acid substitutions (as shown in Fig. 4).

In summary, we introduce a rapid method that allows convenient and rapid quantification and sequencing of epitope-specific TCRs. This approach demonstrates that the overall magnitude of the HIV-1–specific CTL response in chronic infection is likely limited by the clonal breadth of TCR raised against viral epitopes. Our method may be useful for further pathogenesis studies and the isolation of TCRs for future immune gene therapies.

We thank the study participants for generously providing sample donations. Recombinant human IL-2 was provided by the National Institutes of Health AIDS Reagent Repository.

Disclosures The authors have no financial conflicts of interest.

This work was supported by Public Health Service Grant AI043203. The University of California AIDS Institute CFAR grant provided supportive infrastructure for this project.

The online version of this article contains supplemental material.

1
Yang
O. O.
,
Church
J.
,
Kitchen
C. M.
,
Kilpatrick
R.
,
Ali
A.
,
Geng
Y.
,
Killian
M. S.
,
Sabado
R. L.
,
Ng
H.
,
Suen
J.
, et al
.
2005
.
Genetic and stochastic influences on the interaction of human immunodeficiency virus type 1 and cytotoxic T lymphocytes in identical twins.
J. Virol.
79
:
15368
15375
.
2
Douek
D. C.
,
Betts
M. R.
,
Brenchley
J. M.
,
Hill
B. J.
,
Ambrozak
D. R.
,
Ngai
K. L.
,
Karandikar
N. J.
,
Casazza
J. P.
,
Koup
R. A.
.
2002
.
A novel approach to the analysis of specificity, clonality, and frequency of HIV-specific T cell responses reveals a potential mechanism for control of viral escape.
J. Immunol.
168
:
3099
3104
.
3
Brander
C.
,
Hartman
K. E.
,
Trocha
A. K.
,
Jones
N. G.
,
Johnson
R. P.
,
Korber
B.
,
Wentworth
P.
,
Buchbinder
S. P.
,
Wolinsky
S.
,
Walker
B. D.
,
Kalams
S. A.
.
1998
.
Lack of strong immune selection pressure by the immunodominant, HLA-A*0201-restricted cytotoxic T lymphocyte response in chronic human immunodeficiency virus-1 infection.
J. Clin. Invest.
101
:
2559
2566
.
4
Kedzierska
K.
,
La Gruta
N. L.
,
Stambas
J.
,
Turner
S. J.
,
Doherty
P. C.
.
2008
.
Tracking phenotypically and functionally distinct T cell subsets via T cell repertoire diversity.
Mol. Immunol.
45
:
607
618
.
5
Deckhut
A. M.
,
Allan
W.
,
McMickle
A.
,
Eichelberger
M.
,
Blackman
M. A.
,
Doherty
P. C.
,
Woodland
D. L.
.
1993
.
Prominent usage of V beta 8.3 T cells in the H-2Db-restricted response to an influenza A virus nucleoprotein epitope.
J. Immunol.
151
:
2658
2666
.
6
Yu
X. G.
,
Lichterfeld
M.
,
Williams
K. L.
,
Martinez-Picado
J.
,
Walker
B. D.
.
2007
.
Random T-cell receptor recruitment in human immunodeficiency virus type 1 (HIV-1)-specific CD8+ T cells from genetically identical twins infected with the same HIV-1 strain.
J. Virol.
81
:
12666
12669
.
7
Yu
X. G.
,
Lichterfeld
M.
,
Chetty
S.
,
Williams
K. L.
,
Mui
S. K.
,
Miura
T.
,
Frahm
N.
,
Feeney
M. E.
,
Tang
Y.
,
Pereyra
F.
, et al
.
2007
.
Mutually exclusive T-cell receptor induction and differential susceptibility to human immunodeficiency virus type 1 mutational escape associated with a two-amino-acid difference between HLA class I subtypes.
J. Virol.
81
:
1619
1631
.
8
Simons
B. C.
,
Vancompernolle
S. E.
,
Smith
R. M.
,
Wei
J.
,
Barnett
L.
,
Lorey
S. L.
,
Meyer-Olson
D.
,
Kalams
S. A.
.
2008
.
Despite biased TRBV gene usage against a dominant HLA B57-restricted epitope, TCR diversity can provide recognition of circulating epitope variants.
J. Immunol.
181
:
5137
5146
.
9
Gillespie
G. M.
,
Stewart-Jones
G.
,
Rengasamy
J.
,
Beattie
T.
,
Bwayo
J. J.
,
Plummer
F. A.
,
Kaul
R.
,
McMichael
A. J.
,
Easterbrook
P.
,
Dong
T.
, et al
.
2006
.
Strong TCR conservation and altered T cell cross-reactivity characterize a B*57-restricted immune response in HIV-1 infection.
J. Immunol.
177
:
3893
3902
.
10
Yu
X. G.
,
Addo
M. M.
,
Perkins
B. A.
,
Wej
F.
,
Rathod
A.
,
Geer
S. C.
,
Parta
M.
,
Cohen
D.
,
Stone
D. R.
,
Russell
C. J.
, et al
.
2004
.
Differences in the expressed HLA class I alleles effect the differential clustering of HIV type 1-specific T cell responses in infected Chinese and caucasians.
AIDS Res. Hum. Retroviruses
20
:
557
564
.
11
Pannetier
C.
,
Cochet
M.
,
Darche
S.
,
Casrouge
A.
,
Zöller
M.
,
Kourilsky
P.
.
1993
.
The sizes of the CDR3 hypervariable regions of the murine T-cell receptor beta chains vary as a function of the recombined germ-line segments.
Proc. Natl. Acad. Sci. USA
90
:
4319
4323
.
12
Uko
G. P.
,
Fraser
P. A.
,
Awdeh
Z. L.
,
Fici
D. A.
,
Crawford
K. D.
,
Larsen
C. E.
,
Alper
C. A.
.
2001
.
Hepatitis B surface antigen- and tetanus toxoid-specific clonal expansion of CD4+ cells in vitro determined by TCRBV CDR3 length and nucleotide sequence.
Genes Immun.
2
:
11
19
.
13
Killian
M. S.
,
Sabado
R. L.
,
Kilpatrick
S.
,
Hausner
M. A.
,
Jamieson
B. D.
,
Yang
O. O.
.
2005
.
Clonal breadth of the HIV-1-specific T-cell receptor repertoire in vivo as determined by subtractive analysis.
AIDS
19
:
887
896
.
14
Frahm
N.
,
Baker
B.
,
Brander
C.
.
2008
.
Identification and optimal definition of HIV-derived cytotoxic T lymphocyte (CTL) epitopes for the study of CTL escape, functional avidity, and viral evolution
. In
HIV Molecular Immunology 2008.
Korber
B. T. M.
,
Brander
C.
,
Haynes
B.
,
Koup
R.
,
Moore
J. P.
,
Walker
B. D.
,
Watkins
D. I.
, eds.
Los Alamos National Laboratory, Theoretical Biology and Biophysics
,
Los Alamos, NM
, p.
3
24
.
15
Ibarrondo
F. J.
,
Anton
P. A.
,
Fuerst
M.
,
Ng
H. L.
,
Wong
J. T.
,
Matud
J.
,
Elliott
J.
,
Shih
R.
,
Hausner
M. A.
,
Price
C.
, et al
.
2005
.
Parallel human immunodeficiency virus type 1-specific CD8+ T-lymphocyte responses in blood and mucosa during chronic infection.
J. Virol.
79
:
4289
4297
.
16
Sabado
R. L.
,
Kilpatrick
S.
,
Ali
A.
,
Dagarag
M.
,
Ng
H. L.
,
Cao
H.
,
Yang
O. O.
.
2005
.
Detection of HIV-1-specific CTL responses in clade B infection with clade C peptides and not clade B consensus peptides.
J. Immunol. Methods
296
:
1
10
.
17
Folch
G.
,
Lefranc
M. P.
.
2000
.
The human T cell receptor beta variable (TRBV) genes.
Exp. Clin. Immunogenet.
17
:
42
54
.
18
Rowen
L.
,
Koop
B. F.
,
Hood
L.
.
1996
.
The complete 685-kilobase DNA sequence of the human beta T cell receptor locus.
Science
272
:
1755
1762
.
19
Mahajan
V. S.
,
Leskov
I. B.
,
Chen
J. Z.
.
2005
.
Homeostasis of T cell diversity.
Cell. Mol. Immunol.
2
:
1
10
.
20
Venturi
V.
,
Kedzierska
K.
,
Turner
S. J.
,
Doherty
P. C.
,
Davenport
M. P.
.
2007
.
Methods for comparing the diversity of samples of the T cell receptor repertoire.
J. Immunol. Methods
321
:
182
195
.
21
Gorochov
G.
,
Neumann
A. U.
,
Kereveur
A.
,
Parizot
C.
,
Li
T.
,
Katlama
C.
,
Karmochkine
M.
,
Raguin
G.
,
Autran
B.
,
Debré
P.
.
1998
.
Perturbation of CD4+ and CD8+ T-cell repertoires during progression to AIDS and regulation of the CD4+ repertoire during antiviral therapy.
Nat. Med.
4
:
215
221
.
22
Killian
M. S.
,
Monteiro
J.
,
Matud
J.
,
Hultin
L. E.
,
Hausner
M. A.
,
Yang
O. O.
,
Gregersen
P. K.
,
Detels
R.
,
Giorgi
J. V.
,
Jamieson
B. D.
.
2004
.
Persistent alterations in the T-cell repertoires of HIV-1-infected and at-risk uninfected men.
AIDS
18
:
161
170
.
23
Kiepiela
P.
,
Leslie
A. J.
,
Honeyborne
I.
,
Ramduth
D.
,
Thobakgale
C.
,
Chetty
S.
,
Rathnavalu
P.
,
Moore
C.
,
Pfafferott
K. J.
,
Hilton
L.
, et al
.
2004
.
Dominant influence of HLA-B in mediating the potential co-evolution of HIV and HLA.
Nature
432
:
769
775
.
24
Naumov
Y. N.
,
Hogan
K. T.
,
Naumova
E. N.
,
Pagel
J. T.
,
Gorski
J.
.
1998
.
A class I MHC-restricted recall response to a viral peptide is highly polyclonal despite stringent CDR3 selection: implications for establishing memory T cell repertoires in “real-world” conditions.
J. Immunol.
160
:
2842
2852
.
25
Naumov
Y. N.
,
Naumova
E. N.
,
Hogan
K. T.
,
Selin
L. K.
,
Gorski
J.
.
2003
.
A fractal clonotype distribution in the CD8+ memory T cell repertoire could optimize potential for immune responses.
J. Immunol.
170
:
3994
4001
.
26
Sylwester
A. W.
,
Mitchell
B. L.
,
Edgar
J. B.
,
Taormina
C.
,
Pelte
C.
,
Ruchti
F.
,
Sleath
P. R.
,
Grabstein
K. H.
,
Hosken
N. A.
,
Kern
F.
, et al
.
2005
.
Broadly targeted human cytomegalovirus-specific CD4+ and CD8+ T cells dominate the memory compartments of exposed subjects.
J. Exp. Med.
202
:
673
685
.
27
Zhang
D.
,
Shankar
P.
,
Xu
Z.
,
Harnisch
B.
,
Chen
G.
,
Lange
C.
,
Lee
S. J.
,
Valdez
H.
,
Lederman
M. M.
,
Lieberman
J.
.
2003
.
Most antiviral CD8 T cells during chronic viral infection do not express high levels of perforin and are not directly cytotoxic.
Blood
101
:
226
235
.
28
Appay
V.
,
Dunbar
P. R.
,
Callan
M.
,
Klenerman
P.
,
Gillespie
G. M.
,
Papagno
L.
,
Ogg
G. S.
,
King
A.
,
Lechner
F.
,
Spina
C. A.
, et al
.
2002
.
Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections.
Nat. Med.
8
:
379
385
.
29
Champagne
P.
,
Ogg
G. S.
,
King
A. S.
,
Knabenhans
C.
,
Ellefsen
K.
,
Nobile
M.
,
Appay
V.
,
Rizzardi
G. P.
,
Fleury
S.
,
Lipp
M.
, et al
.
2001
.
Skewed maturation of memory HIV-specific CD8 T lymphocytes.
Nature
410
:
106
111
.
30
Migueles
S. A.
,
Laborico
A. C.
,
Shupert
W. L.
,
Sabbaghian
M. S.
,
Rabin
R.
,
Hallahan
C. W.
,
Van Baarle
D.
,
Kostense
S.
,
Miedema
F.
,
McLaughlin
M.
, et al
.
2002
.
HIV-specific CD8+ T cell proliferation is coupled to perforin expression and is maintained in nonprogressors.
Nat. Immunol.
3
:
1061
1068
.
31
Wilson
J. D.
,
Ogg
G. S.
,
Allen
R. L.
,
Goulder
P. J.
,
Kelleher
A.
,
Sewell
A. K.
,
O’Callaghan
C. A.
,
Rowland-Jones
S. L.
,
Callan
M. F.
,
McMichael
A. J.
.
1998
.
Oligoclonal expansions of CD8(+) T cells in chronic HIV infection are antigen specific.
J. Exp. Med.
188
:
785
790
.
32
Meyer-Olson
D.
,
Brady
K. W.
,
Bartman
M. T.
,
O’Sullivan
K. M.
,
Simons
B. C.
,
Conrad
J. A.
,
Duncan
C. B.
,
Lorey
S.
,
Siddique
A.
,
Draenert
R.
, et al
.
2006
.
Fluctuations of functionally distinct CD8+ T-cell clonotypes demonstrate flexibility of the HIV-specific TCR repertoire.
Blood
107
:
2373
2383
.
33
Yang
H.
,
Dong
T.
,
Turnbull
E.
,
Ranasinghe
S.
,
Ondondo
B.
,
Goonetilleke
N.
,
Winstone
N.
,
di Gleria
K.
,
Bowness
P.
,
Conlon
C.
, et al
.
2007
.
Broad TCR usage in functional HIV-1-specific CD8+ T cell expansions driven by vaccination during highly active antiretroviral therapy.
J. Immunol.
179
:
597
606
.
34
Dagarag
M.
,
Ng
H.
,
Lubong
R.
,
Effros
R. B.
,
Yang
O. O.
.
2003
.
Differential impairment of lytic and cytokine functions in senescent human immunodeficiency virus type 1-specific cytotoxic T lymphocytes.
J. Virol.
77
:
3077
3083
.
35
Manuel
E. R.
,
Charini
W. A.
,
Sen
P.
,
Peyerl
F. W.
,
Kuroda
M. J.
,
Schmitz
J. E.
,
Autissier
P.
,
Sheeter
D. A.
,
Torbett
B. E.
,
Letvin
N. L.
.
2006
.
Contribution of TCR repertoire breadth to the dominance of epitope-specific CD8+ T lymphocyte responses.
J. Virol.
80
:
12032
12040
.