Journal of Immunological Methods

Transcription

1 Journal of Immunological Methods 340 (2009) Contents lists available at ScienceDirect Journal of Immunological Methods journal homepage: Technical note Combining MHC tetramer and intracellular cytokine staining for CD8 + T cells to reveal antigenic epitopes naturally presented on tumor cells Nektaria Dimopoulos a, Heather M. Jackson a, Lisa Ebert a, Philippe Guillaume b, Immanuel F. Luescher b, Gerd Ritter c, Weisan Chen a, a Ludwig Institute for Cancer Research, Melbourne Branch, Austin Health, Heidelberg, VIC 3084, Australia b Ludwig Institute for Cancer Research, Lausanne, Switzerland c Ludwig Institute for Cancer Research, NY, USA article info abstract Article history: Received 5 July 2008 Received in revised form 26 September 2008 Accepted 30 September 2008 Available online 26 October 2008 Keywords: MHC Tetramer Intracellular cytokine staining Tumor antigenic epitope Natural presentation As more tumor antigens are discovered and as computer-guided T cell epitope prediction programs become more sophisticated, many potential T cell epitopes are synthesized and demonstrated to be antigenic in vitro. However, it is estimated that about 50% of such tumor antigen-specific T cells have not been demonstrated to recognize the naturally presented epitopes due to either technical difficulties, such as T cell cloning which is still challenging for many laboratories; or the predicted T cell epitopes are not generated or not generated in sufficient amounts by the antigen processing machinery. However, to potentially identify clinically relevant vaccine candidate epitopes, it is essential to demonstrate natural antigen presentation. Here we combine the advantages of MHC tetramer and intracellular cytokine staining to sensitively detect natural antigen presentation by tumor cells for epitopes of interest. The novel method does not require T cell cloning or long-term T cell culture. Because the antigen-specific T cells are positively identified, this method is much less influenced by IFNγ producing cells with unknown specificities and should be widely applicable Elsevier B.V. All rights reserved. 1. Introduction CD8 + T cells (T CD8+ ) can eliminate tumors by recognizing tumor antigenic peptides presented by the class I major histocompatibility molecules (MHC) on tumor cells. Different MHC molecules select different peptides according to their binding motifs. The understanding of such binding motifs has led to the development of various computer algorithms which enhanced our ability to identify potential tumor antigenderived T cell epitopes. However, due to limited prediction capacity on antigen processing, it has been estimated that nearly 50% of identified epitopes are not naturally presented by tumor Abbreviations: APC, antigen presenting cells; BFA, Brefeldin A; DC, Dendritic cell; FCS, fetal calf serum; ICS, Intracellular cytokine staining; PBMC, peripheral blood mononuclear cells; ELISpot, enzyme-linked immunosorbent spot. Corresponding author. Tel.: ; fax: address: weisan.chen@ludwig.edu.au (W. Chen). cells (Clark and Vonderheide, 2005). Amongst these epitopes many are simply not generated by the antigen processing machinery. Obtaining information on natural presentation is vital, because it is easily understood that an epitope not presented by tumor cells cannot be a useful vaccine candidate. There are several different ways to demonstrate and potentially to quantitate MHC/peptide presentation in vitro. The most quantitative method for MHC/peptide complexes is to either use mabs that mimic the specific Tcell receptor (TCR) or multivalent TCRs obtained from specific T CD8+. These reagents enable a rapid and precise quantitation of MHC/ peptide complex on the surface of individual antigen presenting cell (APC). However, they are extremely difficult to produce and are far less sensitive than antigen-specific T CD8+. Another approach is to acid extract MHC-bound peptides from APC surface followed by HPLC fractionation and peptide identification using mass spectrometry or sequencing. However, this method requires special skills and expensive equipments, which limit its usage to just a few laboratories worldwide /$ see front matter 2008 Elsevier B.V. All rights reserved. doi: /j.jim

2 N. Dimopoulos et al. / Journal of Immunological Methods 340 (2009) Other methods used for demonstrating peptide presentation by class I or class II molecules all involve antigen-specific Tcells. Tcells remain the most sensitive tool to read out natural antigen presentation. The most commonly used functional T cell assays to date are based on cytokine production after T cell activation including ELISpot and intracellular cytokine staining (ICS) (Jung et al., 1993). If an antigen-specific T cell population is either very pure or clonal, activation by tumor cells directly reflects natural presentation (Echchakir et al., 2001). However, to derive high purity T cell lines with specificity for a single peptide, multiple rounds of antigen stimulation are typically required, which may lead to lower T cell avidity. On the other hand, to clone antigen-specific T cells when the line is of low purity can be a daunting task. When low purity, bulk T cell lines are used for demonstrating natural antigen presentation, it is often difficult to differentiate small populations of peptide-specific T cells from, sometimes, a larger population of T cells activated by other unknown antigens expressed on the same tumor cells. For this reason, in classical cytotoxicity assays, it is a common practice to include peptide-pulsed cold tumor targets to suppress the specific killing to radio-labeled hot targets (same tumor cells) to demonstrate that the killing is peptidespecific. Furthermore, the potential difference detected between hot targets with or without the inhibition of killing by the cold targets is then subject to statistical analysis to increase the confidence of data interpretation (Matsueda et al., 2005). Tetrameric MHC molecules complexed with antigenic peptides have been used to identify antigen-specific T cells (Altman et al., 1996). Such tetramers are now generated routinely by many laboratories and are also made available commercially by leading biotech companies. They are by far the most common reagents used to identify antigen-specific CD8 + and CD4 + T cells these days. In the present study we describe a simple yet robust method for assessing natural tumor antigen presentation by combining tetramer staining with ICS. After 4 h activation by co-culture with tumor cells, antigen-specific T cells are directly identified as tetramer positive and cytokine producing cells. The advantages of this method are: firstly, it does not require high purity Tcell lines or clones. A low purity, yet still high avidity T cell culture is perfectly suited; secondly, it is much less influenced by IFNγ producing cells with unknown specificities because the antigen-specific T cells are positively identified both phenotypically and functionally; and finally, it provides semi-quantitative assessment for the quantity of the presented epitopes on APC, as reflected by the activation-induced T cell receptor down modulation. 2. Materials and methods 2.1. Synthetic peptides, antibodies and tetramers All peptides were synthesized and purified (N95% purity) by Chiron Mimotopes (Clayton, Victoria, Australia) and were dissolved in DMSO (Sigma-Aldrich, St Louis, Mo, USA) at 10 mg/ml as stocks. Anti-CD4 and anti-cd3 (APC), anti-cd8 (Cy-Chrome) and anti-ifnγ (FITC) were purchased from Becton Dickinson (BD, North Ryde, NSW, Australia). All tetramers were made in our Ludwig Institute Tetramer facility (Lausanne, Switzerland) and labeled with PE Cell culture The human melanoma cell lines LM-Mel-26 and LM-Mel-3 were established from melanoma biopsy samples as described elsewhere (Jackson et al., 2006) while the SK-Mel-14 melanoma line was obtained from the Memorial Sloan-Kettering Cancer Center. All lines were maintained in RPMI-1640 supplemented with 10% fetal calf serum (FCS, RP-10 ) (Thermo Trace, Melbourne, Australia), L-glutamine (2 mm), 2-ME ( M) and antibiotics (penicillin 100 U/mL, streptomycin 100 µg/ml) (Invitrogen) Peptide-specific T cell bulk culture peripheral blood mononuclear cells (PBMC) from melanoma patients were pulsed with 1 µg/ml (10 6 M) of Melan A (EAAGIGILTV), or NY-ESO (LAMPFATPM), or NY- ESO (APRGPHGGAASGL) peptide in RP-10 at 37 C for 60 min. Peptide-pulsed PBMC were washed and cultured in RP- 10 containing human recombinant IL-2 at U/ml (Roche, CA, USA) for normally days (Jackson et al., 2006) Generating monocyte derived dendritic cells (MoDCs) and antigen cross presentation CD14 + monocytes were isolated from PBMCs from an HLA- Cw3 + volunteer using magnetic beads (Miltenyi Biotec) and cultured with GM-CSF and IL-4 for 6 7 days as previously described (Schnurr et al., 2004). MoDCs at /ml were incubated with NY-ESO-1 ISCOMATRIX vaccine for 16 h at 37 C. After washing, MoDCs and NY-ESO-1 peptide specific T cell lines were co-cultured in 96 well round-bottom plates in the presence of 10 µg/ml BFA for 4 h. Cells were then washed and stained with the specific tetramers and also for intracellular IFNγ (see below) Intracellular cytokine staining (ICS) and tetramer co-staining The ICS method was described earlier (Jackson et al., 2006). In brief, peptides at 1 µg/ml, or as indicated, were added to T cells, or T cells were incubated with the same number of tumor cells. 10 µg/ml BFA was added for 4 h. The cells were then harvested and stained with anti-cd4 and anti-cd8 in 50 µl of PBS at 4 C for 30 min for single ICS staining; or the cells were stained with tetramers at 0.3 µg/ml for 20 min before addition of anti-cd4 and anti-cd8 for a tetramer and ICS double staining. The cells were washed and fixed with 1% paraformaldehyde in PBS at room temperature for 20 min and further stained with anti-ifnγ-fitc in the presence of 0.2% saponin. 100,000 cells were acquired on a FACScalibur or FACSCanto II (BD) and analyzed with Flowjo software. 3. Results and discussion It has been reported that upon T cell activation, the TCR is rapidly down regulated (Liu et al., 2000) which could potentially affect tetramer staining. To appreciate such a potential influence on tetramer staining, we first analyzed TCR expression on T cells newly activated by its cognate peptide at various concentrations. 4 h after peptide activation,

3 92 N. Dimopoulos et al. / Journal of Immunological Methods 340 (2009) these T cells were stained for their TCR expression using anti- CD3 (also anti-cd8) and specific tetramer, and their IFNγ production was measured as mean channel fluorescence intensity (MCF) as shown in Fig. 1A. It was clear that the antigen-specific T cells significantly down regulated their TCR when they were activated with large amount of peptides. For instance, at 10 5 M the TCR expression became minimal (Fig. 1B). As the peptide dose decreased, the TCR down regulation became more modest. At ~10 8 M, the TCR down regulation was about 50%. Most importantly, judging from IFNγ and tetramer staining patterns, the antigen-specific T cells were easily and positively identified. It was also shown in Fig. 1B that the amount of IFNγ produced in these cells inversely correlated to TCR down regulation. Having demonstrated that tetramer could stain newly activated antigen-specific T cells, we sought to use this method to detect natural antigen presentation by tumor cells. It is well observed that some T cells, as well as CD56 + NK cells, in a shortterm T cell culture often produce IFNγ with unknown specificities when co-cultured with various tumor cell lines (unpublished data). Thus, when an antigen-specific T cell population in any given line is small, it becomes difficult to differentiate the specific signal from noise of unknown specificities. More importantly, when different tumor cell lines that do or do not express the tumor antigen or that do or do not present the restricting MHC molecule of interest are used to show natural presentation, it is difficult to be sure that the low number of T cells activated by the tumor cells expressing the right antigen and the right restricting MHC molecule are actually the same cells that are activated by the synthetic peptide. For this reason, such experiments are often performed with either high purity CTL lines, e.g. N60% (Jackson et al., 2006), or clones to avoid ambiguity (Echchakir et al., 2001). We generated a T cell line specificformelana /HLA-A2 with high purity (52.7% pure amongst all CD8 + T cells, data not shown), and co-cultured these T cells with tumor cell lines expressing the tumor antigen Melan A, but with (SK-Mel-14) or without (LM-Mel-3) the presenting molecule HLA-A2. 62% of all antigen specific T cells were activated by the tumor line that expressed both HLA-A2 and Melan A. Some antigen-specific T cells not only produced large amounts of IFNγ but also lost tetramer staining completely, most likely due to dramatic TCR down regulation resulting from strong activation (Fig. 2 top right panel). Such activation was antigen-specific and HLA-A2- restricted because there was no activation when the Tcells were co-cultured with an HLA-A2 negative tumor line (Fig. 2 top left panel). Because many T cells within this line were activated by tumor cells, even if there was no tetramer to identify them, it Fig. 1. Most peptide-activated T cells can be stained simultaneously by tetramer and ICS. Melan A specific T cell line was established as described in the Materials and methods. The T cells were activated by direct addition of varied concentrations of Melan A peptide in the presence of BFA. 4 h later, the cells were stained with anti-cd3, anti-cd8, tetramer and then ICS for IFNγ. In A, T cell activation was shown for three representative peptide concentrations. Percentages shown indicate tetramer + and IFNγ + cells amongst all antigen-specific T cells (including single tetramer + and single IFNγ + cells). In B, the full peptide titration data were converted and plotted as fluorescence intensity of tetramer + cells (right Y-axis) and CD8 + and IFNγ + cell percentages amongst all CD8 + T cells (left Y-axis).

4 N. Dimopoulos et al. / Journal of Immunological Methods 340 (2009) Fig. 2. Tetramer and intracellular IFNγ double staining T cells reveal natural presentation by melanoma cell lines. Melan-A specific T cell line was generated the same way as described in the Materials and methods. The T cell line (purity 52.7% amongst total CD8 + cells, data not shown) was then incubated with Melan A- expressing tumor lines with (SK-MEL-14) or without (LM-MEL-3) expressing HLA-A2 for 4 h in the presence of BFA. T cell activation was then detected by tetramer and ICS (Fig. 2A). In the lower panels (Fig. 2B), a low purity NY-ESO specific T cell line (purity 1.0% amongst total CD8 + cells, data not shown) was generated the same way and co-cultured with HLA-B7-expressing melanoma lines with (SK-MEL-14) or without (LM-MEL-26) expressing NY-ESO-1. Due to the low HLA expression, both lines were treated with IFNγ at 100 ng/ml for 48 h before the co-culture. Percentages in the upper left, upper right and lower right quadrants represent the antigen-specific T cells amongst total tetramer + and IFNγ + (including single positive cells) populations. would have been a clear demonstration of natural presentation of Melan A However, we also generated a low purity T cell line to a novel NY-ESO-1 epitope (Ebert et al., submitted) and tested natural tumor antigen presentation as shown in Fig. 2 lower panels. Within the antigen-specific T cell population identified by the NY-ESO /HLA-B7 tetramer (at ~1% of total CD8 + cells, data not shown) none of them produced IFNγ when co-cultured with a tumor cell line, LM-Mel-3, that did not express the NY-ESO-1 antigen. When the T cell line was cocultured with a melanoma line, SK-Mel-14, that is both NY-ESO- 1 expressing and HLA-B7 positive, 28% of tetramer positive cells also produced IFNγ, a population not much bigger in size than the tetramer negative, IFNγ-producing population (of unknown specificities) that responded to the same tumor line (Fig. 2 lower panels). In this case, if the antigen specific T cells were not positively identified by the specific NY-ESO / HLA-B7 tetramer, based upon such a small signal, it would have been difficult to know whether the natural antigen presentation or extra cells of unknown specificities were detected. Subsequently, we have shown natural presentation for other tumor or viral antigenic epitopes, such as the BMLF /HLA-A2 (GLCTLVAML) epitope from Epstein Bar Virus (EBV) presented on EBV transformed B cell lines (data not shown). As anti-tumor immune responses are believed to be initiated by DCs through cross-presentation, we asked whether the novel method might allow us to measure cross-presentation in vitro. T cell lines specific tony-eso with high (99.4%) and low purity (5.78%) were co-cultured with monocyte-derived DC (MoDC) which were loaded with NY-ESO-1 ISCOMATRIX vaccine followed by tetramer staining and ICS. As shown in Fig. 3, T cells within both the high- and low-purity lines were activated by the cognate peptide (1 µg/ml) directly added into the culture, which was reflected by slightly reduced tetramer staining of the IFNγ-producing cells due to TCR down regulation. Importantly, the MoDCs pulsed with the NY-ESO-1 ISCOMATRIX vaccine efficiently activated these T CD8+ through cross-presentation (the tetramer negative population seen in Fig. 3 top right panel are the antigen presenting MoDCs). So, our method is also suitable for detecting antigen cross-presentation, which is generally considered a less efficient process compared to direct antigen presentation. As more tumor epitope specific T cells are identified through the reverse immunological method with the aid from a computer algorithm (Matsueda et al., 2005) and as it has been demonstrated that peptide exposure in vitro could lead to T cell priming in vitro (Zippelius et al., 2002), it is critical to

5 94 N. Dimopoulos et al. / Journal of Immunological Methods 340 (2009) Fig. 3. Tetramer and intracellular IFNγ double staining T cells detect antigen cross-presentation by MoDCs. High purity (99.4%) and low purity (5.78%, data not shown) T cell lines specific to NY-ESO /HLA-Cw3 were established as described in the Materials and methods. MoDCs were generated from purified monocytes as described in the Materials and methods. MoDCs were then loaded with NY-ESO-1 ISCOMATRIX vaccine at 20 µg/ml overnight and their presentation of NY-ESO was detected by antigen-specific T cell lines. NY-ESO peptide (1 µg/ml) directly added to T cells served as a positive control. The percentages shown in the quadrants are the same as those described in Fig. 2. assess the relevance of such T cells for their potential antitumor activity. Ideally, such demonstration should be achieved by using tumor cells directly isolated from patients, which can be very challenging. Alternatively, natural presentation of such epitopes should be shown using in vitro established tumor lines, or at least using gene-transfected cell lines if a tumor line expressing both the tumor antigen and the presenting HLA is not available. Combining our method with strategies for antigen delivery into cross-presentation pathway, one might be able to show that for some predicted epitopes although these were not able to be generated from direct presentation, might be efficiently cross-presented by DCs. Such results should allow one to consider that the T cell line used is of high enough avidity to demonstrate natural antigen presentation. Thus, the reason that tumor cell lines do not present such an epitope then is likely due to the inability of the antigen processing machinery to generate such an epitope. The most important aspect of our novel method is that it is able to be used with relatively low purity T cell population. Many people stimulate T cells with high peptide dosage, which has been shown to induce lower T cell avidity (Kroger and Alexander-Miller, 2007). That is also why many laboratories emphasize early tetramer-guided sorting of antigen-specific T cells followed with T cell cloning. In the absence of cloning capacity, yet with the availability of tetramers, our method described here will provide equal certainty for showing natural tumor antigen presentation to that assessed by pure Tcell clones. Acknowledgements References Altman, J.D., Moss, P.A., Goulder, P.J., Barouch, D.H., McHeyzer-Williams, M.G., Bell, J.I., McMichael, A.J., Davis, M.M.,1996. Phenotypic analysis of antigenspecific T lymphocytes. Science 274, 94. Clark, C.E., Vonderheide, R.H., Gettingtothesurface: a linkbetween tumor antigen discovery and natural presentation of peptide-mhc complexes. Clin. Cancer Res. 11, Echchakir, H., Mami-Chouaib, F., Vergnon, I., Baurain, J.F., Karanikas, V., Chouaib, S., Coulie, P.G., A point mutation in the alpha-actinin-4 gene generates an antigenic peptide recognized by autologous cytolytic T lymphocytes on a human lung carcinoma. Cancer Res. 61, Jackson, H., Dimopoulos, N., Mifsud, N., Tai, T., Chen, Q., Svobodova, S., Browning, J., Luescher, I., Stockert, L., Old, L., Davis, I., Cebon, J., Chen, W., Striking immunodominance hierarchy of naturally-occurring CD8+ and CD4+ T cell responses to tumor antigen NY-ESO-1. J. Immunol. 176, Jung, T., Schauer, U., Heusser, C., Neumann, C., Rieger, C., Detection of intracellular cytokines by flow cytometry. J. Immunol. Methods 159, 197. Kroger, C.J., Alexander-Miller, M.A., Cutting edge: CD8+ T cell clones possess the potential to differentiate into both high- and low-avidity effector cells. J. Immunol. 179, 748. Liu, H., Rhodes, M., Wiest, D.L., Vignali, D.A., On the dynamics of TCR: CD3 complex cell surface expression and downmodulation. Immunity 13, 665. Matsueda, S., Takedatsu, H., Yao, A., Tanaka, M., Noguchi, M., Itoh, K., Harada, M., Identification of peptide vaccine candidates for prostate cancer patients with HLA-A3 supertype alleles. Clin. Cancer Res. 11, Schnurr, M., Chen, Q., Shin, A., Chen, W., Toy, T., Jenderek, C., Green, S., Miloradovic, L., Drane, D., Davis, I.D., Villadangos, J., Shortman, K., Maraskovsky, E., Cebon, J., Tumor antigen processing and presentation depends critically on dendritic cell type and the mode of antigen delivery. Blood 105, Zippelius, A., Pittet, M.J., Batard, P., Rufer, N., de Smedt, M., Guillaume, P., Ellefsen, K., Valmori, D., Lienard, D., Plum, J., MacDonald, H.R., Speiser, D.E., Cerottini, J.C., Romero, P., Thymic selection generates a large T cell pool recognizing a self- peptide in humans. J. Exp. Med. 195, 485. This project was partly supported by the NHMRC project grant to WC (ID No: ) and grants from Cancer Council Victoria to WC (ID No: ) and LE (ID No: ).