Monitoring of NY-ESO-1 specific CD4⁺ T cells using molecularly defined MHC class II/His-tag-peptide tetramers

Abstract

MHC-peptide tetramers have become essential tools for T-cell analysis, but few MHC class II tetramers incorporating peptides from human tumor and self-antigens have been developed. Among limiting factors are the high polymorphism of class II molecules and the low binding capacity of the peptides. Here, we report the generation of molecularly defined tetramers using His-tagged peptides and isolation of folded MHC/peptide monomers by affinity purification. Using this strategy we generated tetramers of DR52b (DRB3*0202), an allele expressed by approximately half of Caucasians, incorporating an epitope from the tumor antigen NY-ESO-1. Molecularly defined tetramers avidly and stably bound to specific CD4⁺ T cells with negligible background on nonspecific cells. Using molecularly defined DR52b/NY-ESO-1 tetramers, we could demonstrate that in DR52b⁺ cancer patients immunized with a recombinant NY-ESO-1 vaccine, vaccine-induced tetramer-positive cells represent ex vivo in average 1:5,000 circulating CD4⁺ T cells, include central and transitional memory polyfunctional populations, and do not include CD4⁺CD25⁺CD127⁻ regulatory T cells. This approach may significantly accelerate the development of reliable MHC class II tetramers to monitor immune responses to tumor and self-antigens.

Soluble MHC-peptide tetramers, allowing the direct visualization, characterization, and isolation of antigen-specific T cells, have become essential tools for T-cell analysis. MHC class I tetramers incorporating short CTL peptide epitopes, originally developed by Altman and Davis (1) have been generated for a large number of murine and human alleles, incorporating a variety of peptides of microbial, tumor, and self-antigen origin (2). The development of MHC class II tetramers, however, and particularly of those incorporating peptides from tumor and self-antigens, has been far less successful (3–5). One limiting factor is the high polymorphism of the human MHC class II molecules, especially those encoded by the DRB1 locus, the most frequently studied. Another limiting factor is the binding affinity of antigenic peptides derived from tumor and self-antigens, which is generally lower than that of peptides from pathogens.

NY-ESO-1 (ESO), a tumor-specific antigen of the cancer/testis group frequently expressed in tumors of different histological types (6), is an important candidate for the development of generic cancer vaccines (7). In a recent vaccination trial using a recombinant ESO protein (rESO) administered with Montanide ISA 51 and CpG ODN 7909, we have observed induction of CD4⁺ T cell responses in all vaccinated patients (8). By assessing vaccine-induced CD4⁺ T cells, we have identified an immunodominant epitope (ESO_119–143, core region ESO_123–137) restricted by HLA-DR52b (DRB3*0202), an allele expressed by half of Caucasians (9). DRB3-, DRB4-, and DRB5-encoded molecules are less polymorphic than those encoded by DRB1, and are therefore attractive candidates for the development of generic MHC class II tetramers.

Our initial attempts to construct DR52b/ESO tetramers using an approach previously described by Kwok and colleagues, by peptide-loading of class II molecules incorporating “leucine zipper” motifs (10), failed to generate efficient tetramers. We therefore designed a strategy using His-tagged peptides that allows isolation of folded MHC/peptide monomers by affinity purification before tetramerization. Tetramers generated according to this procedure avidly and stably bound to ESO-specific CD4⁺ T cells, allowing their direct ex vivo enumeration, phenotyping, and isolation from circulating lymphocytes of vaccinated patients. The application of this strategy to other tumor- and self-antigen-derived peptides may significantly accelerate the development of reliable MHC class II tetramers to monitor antigen-specific CD4⁺ T cells.

Results

Generation of Molecularly Defined MHC Class II Tetramers Using His-tag-Peptides and Their Validation.

We initially attempted to generate DR52b tetramers incorporating peptide ESO_123–137 using a strategy previously described by Kwok and colleagues (10). The tetramers synthesized according to this procedure, however, failed to significantly stain ESO-specific DR52b-restricted CD4⁺ T-cell clones (Fig. S1). We reasoned that this failure might be because of a suboptimal formation of DR52b/ESO complexes, resulting in the presence of low proportions of folded monomers in the tetramer preparation. To overcome this limitation, we synthesized an ESO peptide containing an N-terminal His-tag added via a short linker. After loading DR52b molecules with the His-tag peptide, monomers were purified by affinity chromatography on Ni²⁺-NTA columns, followed by gel filtration chromatography (Fig. S2). The purity of the isolated biotinylated monomers was assessed in a shift assay with avidin (Fig. S3). Tetramerization was carried on using phycoerythrin-labeled streptavidin. The tetramers prepared using this procedure efficiently stained ESO-specific DR52b-restricted CD4⁺ clonal T cells at low concentrations, similar to those generally used for MHC class I/peptide tetramers (11, 12), with low background on control populations (Fig. 1 A and B). To further assess the staining obtained with molecularly defined tetramers, we tested the influence of temperature and incubation time. No significant staining was detectable upon incubation at 4 °C, even after prolonged incubation (Fig. 1C). We observed low but significant staining at 23 °C, particularly after long incubation. Staining at 37 °C, however, was much more efficient and displayed a more rapid kinetic. To assess the persistence of tetramer staining, we incubated specific clones with tetramers for 1 h at 37 °C, removed the excess tetramers by washing, and incubated the cells at various temperatures for different times. No decrease in the staining intensity was detected upon incubation at 4 °C or 23 °C, up to 24 h (Fig. 1D). Even at 37 °C, the staining was maintained up to 4 h and gradually decreased afterward, remaining detectable at 24 h. Together, these results show that molecularly defined DR52b/ESO tetramers avidly and stably bind specific CD4⁺ T cells with negligible background staining on nonspecific CD4⁺ T cells.

Fig. 1.

Molecularly defined DR52b/ESO_123–137 tetramers stain specific CD4⁺ T cell clones. (A and B) ESO-specific DR52b-restricted and control clonal populations were stained with DR52b/ESO_123–137 tetramers for 1 h at 37 °C followed by staining with anti-CD4 mAb and flow cytometry analysis. Examples of dot plots for both populations are shown in A and the mean fluorescence intensity (MFI) of tetramer staining for all concentrations is summarized in B. (C) ESO-specific and control clonal populations were stained with DR52b/ESO_123–137 tetramers (3 μg/mL) at 4 °C, 23 °C, or 37 °C for the indicated periods and analyzed as in A. (D) ESO-specific clonal cells were stained with DR52b/ESO_123–137 tetramers (3 μg/mL) for 1 h at 37 °C, extensively washed, and further incubated at 4 °C, 23 °C, or 37 °C for the indicated periods before flow cytometry analysis.

To assess the capacity of the tetramers to identify specific CD4⁺ T cells within polyclonal populations, we stained peptide-stimulated cultures from DR52b⁺ and DR52b⁻ cancer patients immunized with the rESO vaccine (8). After 1 h incubation at 37 °C, DR52b/ESO tetramers stained a significant proportion of CD4⁺ T cells in the cultures from DR52b⁺ but not from DR52b⁻ patients (Fig. 2A). On selected cultures, we compared the staining obtained after incubation for different time periods. Similar proportions of tetramer-positive cells were detected after incubation for different times, with the mean fluorescence intensity of the tetramer-positive populations being higher after longer incubation periods (Fig. 2B).

Fig. 2.

DR52b/ESO_123–137 tetramers stain peptide-stimulated postvaccine CD4⁺ T cell cultures from DR52b⁺ patients. (A) Postvaccine CD4⁺ T cells from DR52b⁺ and DR52b⁻ patients stimulated in vitro with a pool of overlapping long ESO peptides were stained with DR52b/ESO_123–137 tetramers (3 μg/mL) for 1 h at 37 °C and anti-CD4 mAb and analyzed by flow cytometry. Dot plots for two patients and data for all patients tested are shown. (B) Peptide stimulated CD4⁺ T cells were stained with tetramers for the indicated time periods and analyzed by flow cytometry. Numbers in dot plots correspond to the percentage of tetramer-positive cells and numbers between brackets indicate the MFI of tetramer staining of the tetramer-positive population. Results are shown for one patient (C05) representative of three tested.

Whereas peptide ESO_123–137 was the minimal peptide optimally recognized by clonal CD4⁺ T cells, the ESO peptide originally used to identify the DR52b-restricted epitope was significantly longer, corresponding to the 25mer ESO_119–143 (9). We therefore synthesized DR52b/ESO_119–143 tetramers and assessed them on specific and control clonal populations. Similar to ESO_123–137 tetramers, tetramers prepared with untagged ESO_119–143 failed to stain ESO-specific clones. In contrast, DR52b/ESO_119–143 tetramers prepared using a His-tag ESO_119–143 peptide and purified monomeric complexes stained specific clonal populations with a slightly increased efficiency as compared to DR52b/ESO_123–137 tetramers and displayed similar low background on control clones (Fig. 3A). Both DR52b/ESO_123–137 and DR52b/ESO_119–143 tetramers identified similar proportions of specific CD4⁺ T cells in peptide-stimulated cultures from postvaccination samples and failed to detect specific cells in peptide-stimulated cultures from samples obtained before vaccination (Fig. 3 B and C).

Fig. 3.

DR52b/ESO_119–143 and DR52b/ESO_123–137 tetramers stain specific clones as well as peptide-stimulated CD4⁺ T cells from postvaccine but not from prevaccine samples. (A) ESO-specific DR52b restricted and control clonal populations were stained with DR52b/ESO_123–137 or DR52b/ESO_119–143 tetramers and anti-CD4 mAb and analyzed by flow cytometry. Dot plots of an ESO-specific clone stained with both tetramers at 3 μg/mL and MFI of tetramer staining at all concentrations tested are shown. (B) Postvaccine peptide-stimulated CD4⁺ T cells from DR52b⁺ patients were stained with DR52b/ESO_119–143 or DR52b/ESO_123–137 tetramers and analyzed by flow cytometry. Dot plots for patient C06 and data for all patients tested are shown. (C) Peptide-stimulated CD4⁺ T cells from prevaccine and postvaccine samples from DR52b⁺ patients C02 and N10 were stained and analyzed as in B.

Because our approach involves the addition of a His-tag to the ESO peptides, it was important to address the effect of this modification on peptide binding to class II molecules and recognition by specific CD4⁺ T cells. With this aim, we assessed the relative efficiency of untagged and His-tagged ESO peptides to bind to DR52b using a previously described competition assay (13). As shown in Fig. S4A, addition of the His-tag did not significantly modify peptide binding to DR52b for both ESO_123–137 and ESO_119–143. In addition, whereas the His-tagged ESO_123–137 was recognized by specific CD4⁺ T cells with moderately improved efficiency, untagged and His-tagged ESO_119–143 peptides were recognized with similar efficiency (Fig. S4B). Together, these data demonstrate that the success of our approach in generating efficient tetramers is not because of the effect of the His-tag itself on peptide binding or T-cell recognition, but to its use to purify folded DR52b/ESO peptide complexes.

Molecularly Defined DR52b/ESO Tetramers Allow Direct ex Vivo Enumeration and Phenotyping of CD4⁺ T Cells Induced by the rESO Vaccine.

The high efficiency and specificity of staining obtained with the molecularly defined DR52b/ESO tetramers prompted us to assess their capacity to detect vaccine-induced CD4⁺ T cells ex vivo. With this aim, we isolated CD4⁺ T cells from peripheral blood mononuclear cells (PBMCs) of DR52b⁺ healthy donors and patients using magnetic cell sorting, and stained them with the tetramers for 2 h at 37 °C in combination with CD45RA-specific antibodies. As shown in Fig. 4, the frequency of DR52b/ESO tetramer-positive cells among CD4⁺CD45RA⁻ T cells from healthy donors was below 1:100,000. We obtained similar results when assessing CD4⁺ T cells from patients before vaccination. In contrast, in postvaccine samples, DR52b/ESO tetramer-positive cells were clearly detectable among CD4⁺CD45RA⁻ T cells at a frequency ranging between 1:2,500 and 1:7,000 (average 1:5,000). The quality of memory CD4⁺ T-cell responses elicited by pathogens or vaccines has been correlated with their phenotype. Specifically, it has been inferred that a protective memory response should include not only effector cells (CCR7⁻) but also significant proportions of “reservoir” memory cells, including central memory (CCR7⁺) and transitional memory (CCR7⁻CD27⁺) populations (14, 15). To more extensively characterize vaccine-induced CD4⁺ T cells, we costained them with tetramers and antibodies directed against markers that distinguish distinct differentiation stages of memory cells. This analysis revealed that vaccine-induced DR52b/ESO tetramer-positive populations included significant proportions of CCR7⁺ central memory cells (Fig. 5 A and B). In addition, among tetramer-positive CCR7⁻ cells, the majority were CD27⁺ transitional memory T cells. Another important criterion to select candidate anticancer vaccines is their ability to elicit helper CD4⁺ T-cell responses, but not suppressive CD25⁺CD127⁻ Treg (16, 17). To address this point, we costained postvaccine CD4⁺ T cells with DR52b/ESO tetramers in combination to antibodies to CD45RA, CD25, and CD127. As shown in Fig. 5 C and D, whereas CD25⁺CD127⁻ Treg populations were clearly detected among CD4⁺ T cells of vaccinated patients, the large majority of vaccine-induced tetramer-positive cells were CD25⁻CD127⁺. These results clearly demonstrate that the rESO/Montanide/CpG vaccine mainly induces central and transitional memory CD4⁺ T cells and does not induce ESO-specific Treg.

Fig. 4.

DR52b/ESO tetramers allow direct ex vivo quantification of specific vaccine-induced CD4⁺ T cells. CD4⁺ T cells purified from PBMC from DR52b⁺ healthy donors (HD) and from pre- and postvaccine samples from DR52b⁺ patients were stained ex vivo with DR52b/ESO_119–143 tetramers (3 μg/mL) for 2 h at 37 °C and were then stained with anti-CD45RA mAb and analyzed by flow cytometry. Dot plots for one HD, one prevaccine sample, and all postvaccine samples are shown in A and data for all samples tested are summarized in B. Numbers in dot plots correspond to the percentage of tetramer-positive cells among memory CD45RA⁻ CD4⁺ T cells.

Fig. 5.

DR52b/ESO tetramers allow ex vivo phenotyping of specific vaccine-induced CD4⁺ T cells. (A and B) Postvaccine CD4⁺ T cells from DR52b⁺ patients were stained ex vivo with DR52b/ESO_119–143 tetramers, as in Fig. 4, as well as with CD45RA-, CCR7-, CD27-, and CD28-specific mAb. Dot plots for patient N13 are shown gated on tetramer-negative (Upper) and tetramer-positive (Lower) cells in A and data corresponding to the percentage of central memory (CM, CD45RA⁻CCR7⁺), transitional-memory (TM, CD45RA⁻CCR7⁻CD27⁺) and effector-memory (EM, CD45RA⁻CCR7⁻CD27⁻) cells among tetramer-positive cells for all patients are summarized in B. (C and D) Samples were stained with DR52b/ESO_119–143 tetramers as in A, as well as with CD45RA-, CD25-, and CD127-specific mAb. Dot plots for patient C02 are shown gated on memory tetramer-negative and tetramer-positive cells in C and data obtained for all patients are summarized in D.

MHC Class II Tetramer-Guided Isolation and Functional Characterization of ESO-Specific CD4⁺ T Cells.

To assess vaccine-induced CD4⁺ T cells functionally, we isolated them by tetramer-guided flow cytometry cell sorting and expanded them in vitro (Fig. 6A). The resulting populations contained > 90% tetramer-positive cells. To address the type of CD4⁺ T-cell response induced by the vaccine, we used antibodies directed against different cytokines that characterize different T_H subsets. Vaccine-induced tetramer-positive cells displayed a clear T_H1 profile, as they mainly produced IFN-γ and contained only minor proportions of IL-4- and IL-17-secreting cells and no detectable IL-10-secreting cells (Fig. 6B). CD4⁺ T-cell populations able to produce TNF-α and IL-2 in addition to IFN-γ (called polyfunctional) are associated with enhanced cellular-mediated protection (18). As illustrated in Fig. 6C, the large majority of DR52b/ESO tetramer-positive cells were polyfunctional as they cosecreted IFN-γ, TNF-α, and IL-2. To get insight into the functional avidity of antigen recognition of the tetramer-positive cell populations, we assessed them using DR52b⁺ EBV-B as APC incubated with serial dilutions of ESO peptide. All populations specifically recognized the peptide, displaying 50% maximal recognition at a concentration comprised between 0.1 and 1 μM (Fig. 6D). Tetramer-positive cell populations were also able to recognize rESO processed and presented by DR52b⁺ monocyte-derived dendritic cells, with 50% maximal recognition of the protein in the same range of concentrations as that of the peptide. Efficient recognition of the untagged ESO peptide by tetramer-positive cells isolated from vaccinated patients further ruled out the possibility that the tetramers may detect T cells directed against the His-tag or against a His-tag/ESO fusion sequence.

Fig. 6.

DR52b/ESO tetramers allow the isolation and functional characterization of specific CD4⁺ T cells. (A) Peptide-stimulated postvaccine samples were stained with tetramers (Left) and tetramer-positive and tetramer-negative cells were isolated by flow cytometry cell sorting. Aliquots of sorted cells were directly reanalyzed by flow cytometry (Middle). Tetramer-positive cells were expanded in vitro and the purity of the resulting polyclonal populations was assessed by flow cytometry analysis following tetramer staining (Right). Results are shown for one patient and are representative of data obtained for four patients. (B and C) ESO-specific polyclonal cultures were stimulated with PMA and ionomycin and cytokine production was assessed in a 4-h intracellular cytokine staining assay. Dot plots are shown for one patient and are representative of data obtained for four patients. (D) ESO-specific polyclonal cultures were incubated either with DR52b⁺ EBV-B cells and ESO_119–143 or control peptide (Left) or with DR52b⁺ monocyte-derived dendritic cells preincubated with rESO or control protein (Right) and IFN-γ was measured by ELISA in 24-h culture supernatants.

We have previously reported that more than 50% of ESO-specific DR52b-restricted CD4⁺ T-cell clones isolated from vaccinated patients use Vβ2, suggesting a highly restricted TCR repertoire for T cells recognizing this epitope (9). To directly assess Vβ2 usage by tetramer-positive cells, we costained the cultures with tetramers and anti-Vβ2 specific antibodies. To minimize inhibition of tetramer binding by anti-Vβ antibodies, we first incubated the cultures with tetramers for 1 h at 37 °C, washed them, and then incubated them with anti-Vβ2 antibodies. Under these conditions, we detected a proportion of tetramer-positive Vβ2⁺ T cells in the cultures comprised between 25 and 65% (Fig. 7). To address if other Vβ frequently used by tetramer-positive CD4⁺ T cells could be identified by this approach, we costained some of the cultures with the tetramers and a panel of anti-Vβ antibodies covering together about 50% of the TCR repertoire. This approach, however, failed to identify other relevant Vβ.

Fig. 7.

DR52b/ESO tetramer staining allows the direct assessment of TCR Vβ usage by specific CD4⁺ T cells. Peptide-stimulated postvaccine CD4⁺ T cells were first stained with DR52b/ESO tetramers and then with a panel of anti-TCR Vβ mAb and analyzed by flow cytometry. Dot plots obtained with anti-Vβ2 and anti-Vβ5.1 mAb are shown for one patient in A and data showing the percentage of Vβ2⁺ cells among tetramer-positive cells for all patients are summarized in B.

Discussion

Because of the popularity of MHC class I/peptide tetramers, originally described in 1996 and used since in thousands of studies, attempts to generate efficient MHC class II/peptide tetramers have been pursued during the last decade, yet have met only modest success. Fundamental structural differences between MHC class I and class II molecules have required significantly different approaches for their design. For class I molecules, refolding of the single heavy chain in the presence of peptides and β₂-microblobulin yields folded stable monomeric complexes (19). Class II molecules, however, are noncovalent dimers of α and β chains that display variable stability in solution (20). To reliably generate stable class II molecules in soluble form, Kwok et al. have constructed class II molecules incorporating leucine zipper motifs that replace the transmembrane and cytoplasmic portions of the molecules (10, 21). The advantage of this approach, with respect to others involving the synthesis of covalent single-chain class II/peptide molecules (22), is that empty class II molecules can be loaded with any selected peptide, increasing tremendously the number of epitopes that can be studied. The disadvantage, however, is that, as the α- and β-chain complex is formed irrespective of the antigenic peptide, the proportion of folded class II/peptide complexes in the preparation can be highly variable for different peptides. Thus, whereas this approach has been successfully used in some cases, it has not been generally applicable for the study of a large variety of antigenic peptides, particularly those derived from tumor and self-antigens, which often bind class II molecules with lower affinity than those derived from pathogens (2–5).

Another problem in generating class II tetramers for generic use is the extensive polymorphism in humans, particularly in the case of the DRB1 gene encoding the prevalent β-chain of the DR isotype. At variance with the β-chain of the mouse I-E molecule (homolog to HLA-DR) that is encoded by a single gene, in humans, several additional genes encode other β-chains, namely DRB3 (DR52), DRB4 (DR53), and DRB5 (DR51). Whereas DRB1 is present in all individuals, DRB3, DRB4, and DRB5 are only present in some of them, and are in strong linkage disequilibrium with defined DRB1 alleles. These alternate DR molecules are generally expressed at lower levels when compared with those encoded by DRB1, but are fully functional with respect to antigen presentation (13, 23, 24). These molecules are less polymorphic than DRB1-encoded molecules and therefore represent ideal candidates for the generation of generic class II tetramers (25). In particular, DR52b, encoded by one of the main DRB3 alleles, is expressed by half of Caucasians. In recent years, an increasing number of studies have concentrated on alternate DR molecules, describing their structure and binding characteristics and peptide binding motifs have been defined for several of them (13, 26).

Because of the failure of our initial attempts to generate efficient DR52b/ESO tetramers by peptide loading of DR52b molecules incorporating leucine zipper motifs, as previously described by Kwok et al. (10, 21), we designed a strategy that uses His-tagged peptides, enabling the isolation of folded class II/peptide monomers by affinity purification before tetramer formation. Together, the data reported in this study clearly show that molecularly defined DR52b/ESO tetramers are reliable reagents for the detection, characterization, and isolation of ESO-specific CD4⁺ T cells. We obtained efficient staining of clonal and polyclonal ESO-specific DR52b-restricted CD4⁺ T-cell populations using concentrations of molecularly defined class II tetramers similar to those generally used for class I/peptide tetramers (1–10 μg/mL) (11, 12). Consistent with other reports (27, 28), and at variance with most class I/peptide tetramers that efficiently stain specific CD8⁺ T cells at 4 °C or at 23 °C, efficient staining with class II tetramers was optimally achieved upon incubation at 37 °C. The molecular basis for this difference, which might be in relation with a lower functional avidity of CD4⁺ T cells or with a higher need for TCR clustering, remains to be fully elucidated. Importantly, we obtained efficient staining not only using tetramers incorporating the previously defined minimal peptide required for optimal T-cell recognition (15 amino acids long, ESO_123–137), but also using a His-tagged 25 amino acid-long peptide, ESO_119–143, extended at both the N- and C-terminal ends of the core region. This finding indicates that there are no major limitations in the length of peptides that can be incorporated into DR52b molecules and implies that the use of molecularly defined DR52b tetramers incorporating long peptides from defined protein regions, possibly preselected on the basis of the presence of binding motifs or through functional binding assays, may be an efficient strategy, allowing the rapid identification of immunodominant DR52b epitopes from a large number of antigens. These findings are also compatible with the fact that, in contrast to the strict length requirement of class I-bound peptides (8–10 mers) that need to perfectly fit a groove that is closed at both ends, often by adopting a kinked conformation, the class II binding groove, open at both ends, can easily accommodate long peptides (15–25 mers) that bind in an extended form (29, 30).

It is noteworthy that, although we failed to detect a major effect of the His-tag on the binding of ESO peptides to DR52b, a minor effect of the His-tag on peptide binding or on the stability of the MHC/peptide complex cannot be excluded. In addition, beside this particular system, an effect of the His-tag on the binding of these or of other peptides to other DR molecules cannot be excluded. Furthermore, although the addition of the His-tag to the amino terminus of a long peptide, well outside the region critical for TCR recognition, would unlikely affect the latter, such an effect cannot be definitely excluded and should be assessed in each particular case.

Because of the high quality of the molecularly defined tetramers, we could identify, enumerate, and phenotype ex vivo ESO-specific CD4⁺ T cells induced by immunization of cancer patients with a rESO vaccine that is presently under trial for cancer immunotherapy. This ability allowed us to address some important issues regarding the nature of CD4⁺ T-cell responses elicited by the vaccine. We could unambiguously show that, whereas DR52b/ESO tetramer-positive T cells were below detection limits in healthy donors and patients before vaccination, they were clearly induced in remarkably similar proportions among different individuals following vaccination. Combination of staining with tetramers and antibodies directed against activation/differentiation markers allowed us to demonstrate that vaccine-induced CD4⁺ T cells were mostly composed of central and transitional memory cells, a phenotype that has been associated with protective memory responses to viruses (14, 15). In addition and importantly, we could demonstrate that most vaccine-induced CD4⁺ T cells were T-helper polyfunctional cells of type I (18) and not suppressive Treg. Together, these results illustrate the usefulness of molecularly defined class II/peptide tetramers to monitor tumor antigen-specific CD4 T cells, a crucial aspect in the development of anticancer vaccines. It is noteworthy that, whereas DR52b/ESO tetramers allowed us to monitor vaccine-induced CD4⁺ T cells in 50% of vaccinated patients, the remaining 50% express another alternate DR molecule, DRB4*0101–0103, that has also been reported to present ESO-derived peptides (31), suggesting that the use of only two tetramers might be sufficient to monitor ESO-specific CD4⁺ T cells in the large majority of individuals.

In summary, the combination of a technical advance in the synthesis of class II/peptide tetramers (use of His-tagged peptides and affinity purification of peptide-loaded monomers before tetramerization), together with the use of frequently expressed alternate DR molecules, has the potential to significantly accelerate the development of reliable MHC class II/peptide tetramers, allowing the monitoring of CD4⁺ T cells specific for many other antigens in a variety of pathological conditions as well as in the course of immune interventions.

Materials and Methods

Generation of HLA-DR52b/ESO Peptide Tetramers.

Soluble DR52b molecules were produced in D. mel-2 cells, as detailed in SI Materials and Methods, and purified by anti-DR (clone L243) affinity chromatography. The DR52b eluate was brought to the optimal peptide loading pH of 6.0 with 100 mM citric acid, loaded at a peptide to protein molar ratio of 50:1, at 28 °C for 24 h in the presence of protease inhibitor mixture (Roche) and 0.2% octyl β-D-glucopyranoside (Sigma), and then biotinylated using the BirA enzyme (Avidity). When DR52b molecules were loaded with untagged ESO peptides, pMHC complexes were directly purified by gel filtration in PBS pH 7.4, 100 mM NaCl on a Superdex S200 column (GE Healthcare Life Sciences) and the fractions corresponding to the monomeric pMHC complexes were pooled and concentrated. Alternatively, ESO peptides were extended at the N terminus by a sequence containing 6-His residues and a linker (Ser-Gly-Ser-Gly). DR52b/His-tag-ESO peptide complexes were purified using the HisTrap HP 1 mL column (GE Healthcare Life Sciences) before purification by gel filtration (Fig. S2). Finally, biotinylation and purity, as assessed by SDS/PAGE in an avidin shift assay, were both > 90% (Fig. S3). Biotinyated DR52b/peptide complexes were multimerized by mixing with small aliquots of streptavidin-PE (Invitrogen) up to the calculated 4:1 stoichiometrical amount.

Patients’ Samples, Cells, Tissue Culture, Tetramer Staining, and Flow Cytometry Analysis and Sorting.

Peripheral blood samples were collected from cancer patients enrolled in a clinical trial of vaccination with rESO, Montanide ISA 51, and CpG 7909 (8) upon informed consent and approval by the Institutional Review Boards. Peripheral blood samples from healthy donors were obtained from the Etablissement Français du Sang Pays de la Loire (Nantes, France). MHC class II alleles were determined by high resolution molecular typing (9). ESO_119–143-specific DR52b-restricted CD4⁺ T cell clonal populations were obtained from postvaccine samples as previously described (9). For assessment of specific CD4⁺ T-cell responses following in vitro stimulation, CD4⁺ cells were enriched from PBMC by magnetic cell sorting (Miltenyi Biotec Inc.), stimulated with irradiated autologous APC in the presence of a pool of overlapping long peptides spanning the ESO sequence, rhIL-2 and rhIL-7, as previously described (9), and maintained in culture for 10 to 15 days before tetramer staining. Peptide stimulated cultures and clonal populations were incubated with tetramers at a final concentration of 3 μg/mL for 1 h at 37 °C, unless otherwise indicated, in complete IMDM medium, washed, and then stained with CD4- (BD Biosciences) or TCR Vβ- (Beckman Coulter) specific mAb in PBS, 5% FCS for 15 min at 4 °C and analyzed by flow cytometry (FACSAria, BD Biosciences). To generate specific polyclonal T cell populations, tetramer-positive cells within peptide-stimulated cultures were sorted by flow cytometry (FACSAria, BD Biosciences) and expanded by stimulation with PHA and irradiated allogeneic PBMC in the presence of rhIL-2 (32). For ex vivo enumeration and phenotyping of specific cells, CD4⁺ cells enriched from PBMC were rested overnight, incubated with tetramers (3 μg/mL) for 2 h at 37 °C, and then stained with CD45RA-, CCR7-, CD25-, CD27-, CD28-, and CD127-specific mAb, as indicated, and analyzed by flow cytometry.

Antigen Recognition Assays.

DR52b⁺ ESO-specific CD4⁺ T cell clones or polyclonal cultures were stimulated in the absence or presence of ESO peptides (2 μM) or PMA (100 ng/mL) and ionomycin (1 μg/mL), as indicated, and cytokine production was assessed in a standard 4-h intracellular cytokine staining assay using mAb specific for IFN-γ, TNF-α, IL-2, IL-4, IL-10 (BD Biosciences), and IL-17 (eBiosciences), as previously described (9, 33). In other experiments, specific polyclonal cultures were incubated for 24 h with either DR52b⁺ EBV-B cells and serial dilutions of ESO peptides or monocyte-derived dendritic cells preincubated overnight with serial dilutions of rESO. IFN-γ was measured by ELISA (Invitrogen) in 24-h culture supernatants, as previously described (8, 9).