Human Immunology 75 (2014) 514–519

Contents lists available at ScienceDirect

www.ashi-hla.org

journal homepage: www.elsevier.com/locate/humimm

Review

T-cell epitope discovery technologies Govinda Sharma a,b, Robert A. Holt a,b,c,⇑ a

Michael Smith Genome Sciences Centre, BC Cancer Agency, 675 W 10th Avenue, Vancouver, British Columbia V5Z 1L3, Canada Department of Medical Genetics, University of British Columbia, C201 – 4500 Oak Street, Vancouver, British Columbia V6H 3N1, Canada c Department of Molecular Biology and Biochemistry, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia V5A 1S6, Canada b

a r t i c l e

i n f o

Article history: Received 12 November 2013 Accepted 27 March 2014 Available online 19 April 2014 Keywords: T-cell receptor Major histocompatibility complex Epitopes High-throughput screening

a b s t r a c t Despite tremendous potential utility in clinical medicine and research, the discovery and characterization of T-cell antigens has lagged behind most other areas of health research in joining the high-throughput ‘-omics’ revolution. Partially responsible for this is the complex nature of the interactions between effector T cells and antigen-presenting cells. Further contributing to the challenge is the vastness of both the T-cell repertoire and the large number of potential T-cell epitopes. In this review, we trace the development of various discovery strategies, the technical platforms used to carry them out, and we assess the level of success achieved in the field today. Ó 2014 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.

Contents 1. 2. 3. 4.

5.

6. 7.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early genomic/cDNA library screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proteomic approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MHC tetramer approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Flow cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Tetramer microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell-based methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Functional assays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Antigen presentation strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Computational prediction strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions and future directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction The elucidation of T-cell antigens is crucial to the understanding of the molecular etiology of immune related disorders and the development of novel therapeutic strategies. Reliable identification of T-cell antigens would, in particular, address an unmet need in the fields of cancer immunology [1], autoimmunity [2] and ⇑ Corresponding author at: Michael Smith Genome Sciences Centre, BC Cancer Agency, 675 W 10th Avenue, Vancouver, British Columbia V5Z 1L3, Canada. E-mail addresses: [email protected] (G. Sharma), [email protected] (R.A. Holt).

514 515 516 516 516 517 517 517 517 518 518 518 518

infectious disease [3]. T-cell epitopes are short peptides displayed at the surface of antigen-presenting cells by the membrane-bound major histocompatibility complex (MHC) proteins, which are categorized as either class I or II. Class I molecules are expressed on the surface of nearly every cell of the body and present a sampling of short (8–14 residue [4]) peptides derived from proteolytic turnover of proteins of both endogenous and exogenous origin. These MHC class I antigens are targets of direct attack from cytotoxic T-lymphocytes. MHC class II exists on the surface of professional antigen-presenting cells (pAPCs) and is responsible for priming naïve T-lymphocytes in peripheral lymphoid tissue. Notably,

http://dx.doi.org/10.1016/j.humimm.2014.03.003 0198-8859/Ó 2014 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.

G. Sharma, R.A. Holt / Human Immunology 75 (2014) 514–519

peptide determinants presented on the surface of class II molecules are not subject to strict length constraints like their MHC class I counterparts. T-cell recognition of peptide-MHC (pMHC) is mediated by the ab-T-cell receptor (TCR), a heterodimeric integral T-cell membrane protein composed of an a and b subunit, each encoded at a separate genomic locus. Each of these TCR subunit genes has a hypervariable region that encodes complementarity-determining region 3 (CDR3), which is the primary region of direct engagement with MHC-presented peptide epitopes. This hypervariability is derived from stochastic somatic rearrangement of gene segments present within the germline locus of each subunit [5]. Post-recombination, nascent T cells undergo positive and negative selection against self pMHC to yield a diverse repertoire optimized for tolerance of self antigens but poised for recognition of any foreign antigen that may be encountered. The discovery of T-cell epitopes has proven, historically, to be a difficult endeavor given numerous characteristics of T-cell antigen recognition that must be accounted for. Firstly, the extreme diversity of the T-cell repertoire generally ensures that clonotypes of interest are present in very low numbers. Moreover, T-cell epitope recognition is a notoriously low affinity interaction that must occur in the context of polygenic and highly polyallelic MHC molecules. Meanwhile, processing and presentation of both exogenous and endogenous peptides on MHC molecules makes for an enormous T-cell epitope space to be screened. Contributing to the complexity associated with T-cell antigen discovery is the substantial level of cross-reactivity present in the T-cell repertoire. Theoretical calculations have estimated the number of pMHC antigens recognized by a single TCR to be on the order of 106 [6] and such estimates have since received experimental support [7]. Further, it has been noted that these cross-reactive epitopes need not share significant sequence similarity and that TCRs are capable of binding different pMHC via numerous different mechanisms [8]. With respect to antigen-discovery applications, these observations indicate the importance of shifting towards large-scale approaches such as combinatorial library screening to effectively probe pMHC/TCR reactivity. Importantly, from a practical standpoint, high-dimensional screening for cross-reactive epitopes can afford the opportu-

515

nity for therapeutic intervention by revealing antigens with a higher capacity for priming a T-cell response than the natural epitopes restricted to the target pathology [9–11]. Extensive research and development efforts in the field of T-cell antigen discovery have been ongoing over the previous three decades (summarized in Fig. 1) with many of these technologies poised for success in the new era of high-dimensional biological research. The subsequent sections of this review outline the evolution of T-cell antigen discovery and assesses some of the main challenges remaining.

2. Early genomic/cDNA library screening Initial TCR antigen discovery efforts were focused on melanoma since these tumor cells are generally more amenable to the creation of stable cell-lines [12] and are highly mutated, thus providing an abundance of mutational epitopes to characterize. In the late 1980s and early 1990s, pioneering work was done in which cytotoxic T-lymphocyte (CTL)-sensitive cells isolated from tumor were selected by co-culture with reactive CTLs until stable antigen-loss variants were isolated. A cosmid library could then be constructed from the genomic DNA of the original cells, transfected into the antigen-loss variant line, and co-cultured with T-cell clones of interest derived from patient peripheral blood. Subsequent chromium-51 release assays would reveal transfectants in which CTL sensitivity was restored and cosmid vectors could be recovered for characterization [13]. These experiments led to the discovery of the now well-known melanoma associated antigens, MAGE-1, -2, and -3. In another study, a similar approach was used whereby melanoma derived cDNA libraries instead of genomic DNA libraries were transfected into a non-melanoma cell line for screening against patient derived tumor-infiltrating lymphocytes (TIL) [14]. These investigations led to the identification of another classic melanoma antigen, MART-1. A major impediment of the above methodologies was the requirement to create stable target cell lines expressing genes from the tissue under interrogation. Many other important diseases,

Fig. 1. A summary of the various T-cell antigen discovery approaches with respect to both antigen-presentation strategies and assay methods.

516

G. Sharma, R.A. Holt / Human Immunology 75 (2014) 514–519

including many other tumors, cannot be investigated in this way due to the inherent difficulties in establishing permanent cell-lines from most types of primary tissue and from the potential for high background to be observed when using surrogate cell lines. An alternative cDNA screening method provided a partial solution to these problems by utilizing an indirect route to T-cell antigen discovery; serological analysis of recombinant cDNA expression libraries (SEREX). Fundamentally, SEREX is an antibody-epitope screening technique involving the display of a prokaryoticallyexpressed, surface-presented cDNA library screened against patient sera [15]. Given that humoral immune responses require the assistance of CD4 T-cells, antigens discovered using this approach often provide an indirect indication of T-cell specificity, including CD8 T-cells. Indeed, the well-known CTL antigen, NYESO-1, was first identified in a SEREX screen [12] although further experiments were required to confirm its activity in cell-mediated immunity [16].

defective translation products depends on the quality of the reference proteomes containing known PTMs. 4. MHC tetramer approaches A major development in the progression of epitope screening technology was the advent of MHC tetramers [25]. The first tetramer system employed engineered molecules consisting of a specific pMHC complex fused to a biotinylation signaling domain that, when conjugated to biotin, could be bound to phycoerythrin-conjugated avidin. The natural tetramerization properties of avidin resulted in the combination of multiple identical pMHC proteins to form a single molecule while the fluorescence characteristics of phycoerythrin enabled detection. Circumventing the problems associated with the weak and transient interactions characteristic of pMHC/TCR binding in cell-free experiments, the higher avidity tetramers provided a means of directly staining cognate TCRs with candidate antigen complexes.

3. Proteomic approaches 4.1. Flow cytometry Acid elution of MHC-bound peptide followed by analysis by bulk Edman sequencing was an early technique used in the characterization of T-cell epitopes. It was this peptide based approach that first provided evidence for the notion that individual MHC alleles bound specific and restricted peptide motifs [17]. Subsequently, tandem mass spectrometry (MS) for peptide sequencing enabled higher-resolution identification of epitopes isolated by peptide elution [18]. Using this method, investigators were able to rapidly characterize individual peptides present in high-performance liquid chromatography (HPLC) fractions that were capable of eliciting in vitro T-cell responses. From these experiments came another well known melanoma epitope, Pmel-1 or gp100 (280– 288). However, it should be noted that this protocol also identified epitopes that were not able to reconstitute T-cell recognition in validation experiments, indicating a high level of noise in this system. An important feature of these peptide-elution strategies is that in many cases, complex antigen transfection schema are not necessary for the study of pathology-specific epitopes; cancers often can be modeled with existing cell lines that constitutively express epitopes of interest while other epitopes can be introduced by infecting cells with pathogen in vitro [19] or pulsing target cells with protein [20]. Contemporary peptide-elution techniques maintain an important role in the general sphere of T-cell antigen discovery, particularly with respect to MHC class II antigens [19,20]. The lack of stringent size restriction on class II antigens makes many of the short-peptide encoding approaches outlined below generally intractable. While cDNA libraries remain an option for uncovering antigenic proteins, peptide sequencing by mass spectrometry offers a more direct way to discover sequences of specific class II epitope regions of an antigenic protein. Also, developments in modern MS systems have resulted in significantly enhanced sensitivity, mass accuracy, and acquisition speed [21] resulting in an enhanced ability to detect and resolve peptide epitopes. Additionally, a fraction of peptides presented by MHC I and II molecules maintain post-translational modifications (PTMs) [22] and in some cases elevated T-cell immunogenicity of post-translationally modified peptides relative to their unmodified counterparts is observed, indicating that the PTMs themselves actually form the basis of antigenic recognition [23]. Modern MS provides a powerful tool for the analysis of post-translationally modified peptides and has been used extensively for this purpose in other contexts [24]. However, for T-cell epitope analysis, the ability to precisely determine specific PTM sites or to discriminate conserved PTMs from

Since the original description of MHC tetramer staining and subsequent flow cytometric analysis, this type of protocol has become a robust and widely used assay for the detection of T-cell antigen recognition when candidate antigens are known. However, limitation in the number fluorophores available to conduct multiparameter tetramer-based flow cytometry has typically constrained screening to one or two antigens per experiment [26]. Recent work in transitioning this approach into a high-throughput format has resulted in the development of combinatorial encoding schemes. In these studies, tetramer stains are generated by conjugating each unique pMHC to multiple different fluorophores such that labeled T-cells are endowed with a multi-colored surface code. This combinatorial staining can then be analyzed by flow cytometry to successfully enable the detection of 25 antigens using 2dimensional staining of 8 different fluorochromes [27]. Alternatively, higher dimensional staining has been employed to yield detection of 15 antigens using relatively fewer fluorophores [28]. These advances, coupled with advances in fluorophore [29] and cytometer [30] technology, place MHC tetramer flow cytometry in a position to become a high-throughput assay capable of probing hundreds of T-cell/APC interactions simultaneously. However, the utility of this approach as an unbiased discovery platform is limited as the presence of epitope cross-reactivity will introduce ambiguity into epitope identification. While the probability of cross-reactivity being observed within libraries of naturally occurring epitopes is low [6], screening of libraries containing fully random synthetic peptide libraries would be impractical by combinatorial tetramer staining. Recent work has expanded the capability of conventional tetramer based flow cytometry by leveraging the highly multi-parametric nature of mass cytometry. While conceptually similar to flow cytometry, mass cytometry obtains measurements in a manner similar to mass spectrometry. Single cells are labeled with antibodies or tetramers conjugated to stable heavy metal isotopes rather than organic fluorophores and focused into an inductively coupled plasma torch, ionizing the entire contents of individual cells. Masses of the isotopes tagging each cell can then be detected and the corresponding parameters can be monitored with minimal signal spillover. Using the combinatorial approach typical for fluorescence flow cytometry, 3-dimensional staining using 10 different metal tags identified 109 different antigens [31]. Theoretically, the dynamic range of the mass cytometry would enable researchers to probe cells with approximately 100 different metal tags. However,

G. Sharma, R.A. Holt / Human Immunology 75 (2014) 514–519

as of this writing, the number of isotope tags currently available is far less and metal-conjugated tetramers suitable for mass cytometry are not commercially available. Regardless, this type of multiparametric monitoring has the potential for very large numbers of antigens to be monitored with combinatorial screening or for the non-combinatorial screening of cross-reactive antigens by using many more labels than would be available in fluorescence flow cytometry. It is important to note that unlike most other antigen-discovery methods under discussion, positive interactors discovered by mass spectrometry cannot be recovered, as cells are completely ablated during detection.

4.2. Tetramer microarrays An alternative stream of MHC tetramer research is the use of tetramer microarrays. Arrays provide an attractive prospect due to spatial addressability of epitope probes potentially enabling much higher-throughput studies than flow cytometry. The first microarrays consisted of pMHC tetramer complexes immobilized to glass slides coated with polyacrylamide, washed with buffer containing a labeled T-cell clone of interest, and visualized by microscopy [32]. In addition to direct quantitation of T-cell binding, MHC tetramer arrays offer an advantage over flow cytometric analysis by being able to monitor functional activation produced by TCR/pMHC interaction. This can be accomplished by co-spotting arrays with capture antibodies for specific cytokines of interest [33]. However, significant technical challenges have thus far prevented this platform from becoming a true high-throughput antigen discovery technology. Among these challenges are the sensitivity and reproducibility of tetramer microarray experiments. The issue of reproducibility and spot homogeneity has been improved by spotting microarray slides with DNA oligonucleotides and conjugating complementary DNA molecules to pMHC tetramers, such that pMHC tetramers are immobilized via highly specific nucleic acid interactions rather than by non-specific polystyrene–protein interactions [34]. With respect to sensitivity, strategies such as the addition of adhesion molecules and co-stimulatory molecules to the array surface [33] or applying non-shearing flow rates across the array surface [35] have significantly improved the detection limit to within the range of the endogenous T-cell repertoire. Despite these improvements, microarray methods have not achieved the level of sensitivity possible using flow cytometry [26]. Perhaps the most daunting challenge facing high-throughput MHC tetramer microarray platforms, and indeed flow cytometry platforms, is the onerous manufacturing process necessary for the construction of each individual pMHC tetramer molecule to be spotted on the array. In large part, this is due to the instability of empty MHC molecules. Both the peptide and MHC components of the complex must be present in the folding reaction to result in viable pMHC for tetramerization. While collections of commercially available manufactured tetramers continue to grow, this characteristic makes the possibility of tractably constructing large, unbiased antigen screening libraries using conventional methods very remote. However, recent development of conditional pMHC tetramers provides a solution to this problem [36,37]. These specialized molecules are pre-folded MHC containing a placeholder peptide ligand modified to contain a photocleavable moiety. Upon cleavage of the photolabile analog, a peptide of interest can then be added to the reaction to bind and rescue MHC molecules from denaturation. Although conditional pMHC technology has the potential to allow for the construction of epitope libraries more rapidly than previously possible, tetramer technology is still somewhat limited as it requires the manufacture of individual MHC alleles to be done separately.

517

5. Cell-based methods Despite the impact of MHC tetramer technology on the field of T-cell epitope discovery, its persistent shortcomings in highthroughput assay design have ensured the continued relevance of cell-based epitope screening. Since the description of the original genomic and cDNA library screening protocols and bulk chromium-51 release measurements, cell-based assays have evolved both in terms of antigen encoding and functional readout. 5.1. Functional assays The contemporary cell-based functional method of T-cell-antigen recognition is the enzyme-linked immunospot (ELISPOT) assay [38]. This assay relies on the detection of cytokines, usually IFNc, secreted by antigen-stimulated T-cells upon recognition of antigen-presenting cells pulsed or transfected with synthetic minimal peptides, whole proteins, or DNA-coded peptides or proteins. Similar in principle to the more widespread enzyme-linked immunosorbent assay (ELISA), ELISPOT is considerably more sensitive than ELISA and provides data regarding the frequency of reactive T-cells as spot counts [39]. An alternative functional readout of TCR-antigen interaction involves the use of a reporter T-cell line constructed with a b-galactosidase [40,41] or fluorescent protein [42] gene placed under the control of the NFAT promoter. Expression of this transcription factor upon T-cell activation enables high sensitivity detection of positive TCR-antigen interactions. In spite of the requirement for the construction of a CTL hybridoma, this method is advantageous as it develops low background signal unlike ELISPOT, which can encounter problems involving high relative noise. 5.2. Antigen presentation strategies Previous observations have indicated that antigen-coated beads are efficiently ingested by pAPCs via endocytosis and presented on both the MHC class I and II molecules [41]. Hence, conjugation of recombinant peptides to beads followed by incubation with pAPCs provides a unique means of constructing epitope display libraries. Moreover, co-culture of these beads with pAPCs for antigen processing does not require properly folded protein to be expressed as is necessary for tetramers. This methodology, referred to as TCAD (T-Cell Antigen Discovery) screening, has been validated in higher-complexity screens involving the epitope mapping of pathogenic genomes as well as the characterization of mimotope libraries of known antigens [43,44]. Baculoviral expression systems have been described as a means of T-cell antigen display on the surface of insect cells [45]. A notable feature of this method is the unbiased means by which antigens are presented. Short epitope sequences are encoded into baculoviral vectors as fusions with b2-microglobulin or b-chain cDNA (in the case of MHC I or II, respectively). Meanwhile a-chain sequences are placed under a separate promoter and are fused to a baculoviral transmembrane domain to ensure anchoring to the insect cell or viral particle surface. This approach enables the presentation of T-cell antigens in the absence of the mammalian antigen-processing pathway. Beneficially, this strategy allows for precise control of the epitopes presented to TCR for screening while minimizing background from endogenously presented peptides, but is not immediately informative as to whether the epitope would be naturally processed and presented. To date, antigens expressed on insect cell surfaces have only been assayed using TCR tetramers. Relying on TCR tetramers for screening presents its own challenges, as they are difficult to manufacture due to poor solubility of recombinant TCRs.

518

G. Sharma, R.A. Holt / Human Immunology 75 (2014) 514–519

Another approach is the construction and screening of a synthetic minigene library [46]. By this method, antigen-presenting cells are transfected with synthetic mRNA generated from in vitro transcription of an artificial cDNA sequence containing barcoded short peptide-coding sequences. Positive interactions can then be assayed functionally. This method has been shown to be effective in screening massive peptide libraries, however, limited numbers of barcode sequences require some antigens to be pooled under the same barcode and experimentally deconvoluted post-screening. A recently developed and very promising epitope discovery platform involves the use of plasmid-encoded short peptides expressed in an APC and screened by NFAT-GFP reporter T-cell hybridomas [42]. These plasmid encoded peptide libraries provide another unbiased means of delivering epitopes of interest to target cells but with the added benefit of providing a single-cell functional readout that can enable the recovery and resurrection of reactive T-cell clones. What is particularly unique about this method is that recovery of T-cell clones interacting with response-eliciting antigen presenting cells affords the opportunity to recover sequences of both the epitope and the TCR a- and bchains from a single positive interaction. Unfortunately, the throughput of this type of screening is, at the moment, very low as positively interacting T-cell/target cell pairs are assayed by microscopy and recovered by microcapillary pipet. 6. Computational prediction strategies As T-cell antigen screening methods shift towards the use of precisely defined DNA encoded short peptides, efficient methods of refining the list of candidate sequences to screen to a manageable number become very important. Much of this refinement can be facilitated using in silico models of peptide-MHC binding [47]. At present, numerous tools are available to predict the level of binding of a candidate sequences to a given MHC allele and are reviewed in depth in [48,49]. Broadly speaking, these tools are generally either sequence based or structure based. The construction of sequence-based models requires the availability of experimental training data and knowledge of MHC restriction. Sequence-based MHC-peptide binding prediction can be accomplished either by using heuristic methods such as motif matrices, probabilistic methods such as hidden Markov models, or machine learning algorithms such as artificial neural networks, support vector machines, and decision trees. The choice of algorithm to use is typically dependent on the type of data available to train the model. In the absence of empirical data, 3D structure modeling has been employed to provide an estimate of MHC-peptide affinity. Most algorithms of this type are based on the estimation of physicochemical parameters, therefore enabling them to forego the need for large test datasets. The flexibility of structure-based methods is further enhanced by their ability to model rarer MHC alleles of which the structure is not known. Though it is generally accepted that sequence-based epitope prediction algorithms outperform structure-based algorithms, there is no consensus regarding the usage of any one particular tool or class of tool in making high quality epitope-MHC binding predictions. Indeed, quality of prediction can be improved by combining multiple approaches. For example, a recent study indicated that the use of sequence based models in conjunction with a molecular docking algorithm improved predictive performance by as much as 6-fold relative to the sequence-based model alone [50]. 7. Conclusions and future directions To date, advances in molecular techniques, unique chemistries, instrument capabilities, and computational algorithms have

resulted in numerous success stories of T-cell antigen discovery as evidenced, for example, by the over 80,000 positively interacting T-cell epitope sequences currently deposited in the Immune Epitope Database (IEDB) [51]. However, as observations regarding the prevalence of cell-mediated immunity in infectious disease, autoimmune disease, cancer, and transplantation medicine accumulate, it is clear that there remains a need for a reliable and scalable assay for positive interactions between MHC–epitope and Tcell receptor complexes. This impetus is further underscored by the current intractability of defining the network of interactions among all cross-reactive pMHCs and TCRs in a given system. The ability to map these interactions comprehensively would provide a greater understanding of the etiology of immune disorders and inform the rational design of immune interventions. At present, there is a paucity of knowledge regarding the interactions between the known TCR sequences and known T-cell epitope sequences. Existing technologies reviewed herein do, to various degrees, enable the screening of monoclonal T-cell populations against epitope libraries for the purpose of delineating specific TCR/pMHC pairs. However, none have yet provided the means to perform epitope library screening against multiple T-cell clones to map the full complexity of immunoreactivity in a particular physiological or pathological state. Potential solutions to the challenges remaining in T-cell antigen discovery may lie in any number of largely unexplored tools that have not yet been applied to the problem. Among them is the rapidly growing field of microfluidics, which has already been applied to other areas of T-cell biology such as the study of T-cell subsets [52]. With respect to the challenge of delineating specific-receptor interactions, the ability to perform clone-by-clone analysis of thousands of cells in parallel is a powerful tool that should find a prevalent role in T-cell antigen discovery studies. Another potentially valuable technique is next-generation sequencing (NGS). The extraordinary depth of sequencing provided by this method enables the identification of rare sequences and has, indeed, been used to characterize the diversity of the T-cell receptor repertoire [53–56]. NGS is a powerful approach for defining candidate antigens which, when coupled with the T-cell epitope discovery methodologies described herein, has the potential to facilitate the interrogation of T-cell antigen diversity [57]. Acknowledgment The authors would like to acknowledge Canadian Institutes of Health Research grant MOP-102679. References [1] Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science 2011;331: 1565–70. [2] Dornmair K, Meinl E, Hohlfeld R. Novel approaches for identifying target antigens of autoreactive human B and T cells. Sem Immunopathol 2009;31:467–77. [3] Appay V, Doueck DC, Price DA. CD8+ T cell efficacy in vaccination and disease. Nat Med 2008;14:623–8. [4] Davis MM, Bjorkman PJ. T-cell antigen receptor genes and T-cell recognition. Nature 1988;334:395–402. [5] Ekeruche-Makinde J, Miles JJ, van den Berg HA, Skowera A, Cole DK, Dolton G, et al. Peptide length determines the outcome of TCR/peptide-MHCI engagement. Blood 2013;121:1112–23. [6] Mason D. A very high level of crossreactivity is an essential feature of the T-cell receptor. Immunol Today 1998;19:395–404. [7] Woolridge L, Ekereche-Makinde J, van den Berg HA, Skowera A, Miles JJ, Tan MP, et al. A single autoimmune T cell receptor recognizes more than a million different peptides. J Biol Chem 2012;287:1168–77. [8] Sewell AK. Why must T cells be cross-reactive? Nat Rev Immunol 2012;12:669–77. [9] Pentier JM, Sewell AK, Miles JJ. Advances in T-cell epitope engineering. Front Immunol 2013;4:133. http://dx.doi.org/10.3389/fimmu.2013.00133.

G. Sharma, R.A. Holt / Human Immunology 75 (2014) 514–519 [10] Buhrman JD, Jordan KR, Munson DJ, Moore BL, Kappler JW, Slansky JE. Improving antigenic peptide vaccines for cancer immunotherapy using a dominant tumor-specific T cell receptor. J Biol Chem 2013;288:33213–25. [11] Ekeruche-Makinde J, Clement M, Cole DK, Edwards ES, Ladell K, Miles JJ, et al. T-cell receptor-optimized peptide skewing of the T-cell repertoire can enhance antigen targeting. J Biol Chem 2012;287:37269–81. [12] Chen Y-T, Scanlan MJ, Sahin U, Tureci O, Gure AO, Tsang S, et al. A testicular antigen aberrantly expressed in human cancers detected by autologous antibody screening. Proc Natl Acad Sci 1997;94:1914–8. [13] van der Bruggen P, Traversari C, Chomez P, Lurquin C, De Plaen E, Van den Eynde BJ, et al. A gene encoding an antigen recognized by cytolytic Tlymphocytes on a human melanoma. J Immunol 1991;254:1643–7. [14] Kawakami Y, Eliyahu S, Delgado CH, Robbins PF, Rivoltini L, Topaliani SL, et al. Cloning of the gene coding for a shared human melanoma antigen recognized by autologous T cells infiltrating into tumor. Proc Natl Acad Sci 1994;91:3515–9. [15] Sahin U, Tureci O, Schmitt H, Cochlovius B, Johannes T, Schmits R, et al. Human neoplasms elicit multiple specific immune responses in the autologous host. Proc Natl Acad Sci 1995;92:11810–3. [16] Jager E, Chen Y-T, Drijfhout JW, Karbach J, Ringhoffer M, Jager D, et al. Simultaneous humoral and cellular immune response against cancer-testis antigen NY-ESO-1 – definition of human histocompatibility leukocyte antigen (HLA)-A2-binding peptide epitopes. J Exp Med 1988;187:265–70. [17] Falk K, Rotzschke O, Stevanovic S, Jung G, Rammensee H-G. Allele-specific motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature 1991;351:290–6. [18] Cox AL, Skipper JS, Chen Y, Henderson RA, Darrow TL, Shabanowitz J, et al. Identification of a peptide recognized by five melanoma-specific human cytotoxic T cell lines. Science 1994;5159:716–9. [19] Karunakaran KP, Rey-Ladino J, Stoynov N, Berg K, Shen C, Jiang X, et al. Immunoproteomic discovery of novel T cell antigens from the obligate intracellular pathogen Chlamydia. J Immunol 2008;180:2459–65. [20] van Haren SD, Herczenik E, ten Brinke A, Mertens K, Voorberg J, Meijer AB. HLA-DR-presented peptide repertoires derived from human monocytederived dendritic cells pulsed with blood coagulation factor VIII. Mol Cell Proteomics 2011;10:6. http://dx.doi.org/10.1074/mcp.M110.002246. [21] Holcapek M, Jirasko R, Lisa M. Recent developments in liquid chromatography–mass spectrometry and related techniques. J Chromatogr A 2012;1259:3–15. [22] Peterson J, Purcell AW, Rossjohn J. Post-translationally modified T cell epitopes: immune recognition and immunotherapy. J Mol Med 2009;87: 1045–51. [23] Mohammed F, Cobbold M, Zarling AL, Salim M, Barrett-Wilt GA, Shabanowitx J, et al. Phosphorylation-dependent interaction between antigenic peptides and MHC class I: a molecular basis for the presentation of transformed self. Nat Immunol 2008;9:1236–43. [24] Silva AMN, Vitorino R, Domingues MRM, Spickett CM, Domingues P. Posttranslational modifications and mass spectrometry detection. Free Radic Biol Med 2013;65:925–41. [25] Altman JD, Moss PAH, Goulder PJR, Barouch DH, McHeyzer-Williams MG, Bell JI, et al. Phenotypic analysis of antigen-specific T lymphocytes. Science 1996;5284:94–6. [26] Hadrup SR, Schumacher TN. MHC-based detection of antigen-specific CD8+ T cell responses. Cancer Immunol Immunother 2010;59:1425–33. [27] Hadrup SR, Bakker AH, Shu CJ, Andersen RS, van Veluw J, Hombrink P, et al. Parallel detection of antigen-specific T-cell responses by multidimensional encoding of MHC multimers. Nat Meth 2009;6:520–6. [28] Newell EV, Klein LO, Yu W, Davis MM. Simultaneous detection of many T-cell specificities using combinatorial tetramer staining. Nat Meth 2009;6:497–9. [29] Chattopadhyay PK, Price DA, Harper TF, Betts MR, Yu J, Gostick E, et al. Quantum dot semiconductor nanocrystals for immunophenotyping by polychromatic flow cytometry. Nat Med 2006;12:927–77. [30] Perfetto SP, Chattopadhyay PK, Roederer M. Seventeen-colour flow cytometry: unravelling the immune system. Nat Rev Immunol 2004;4:648–55. [31] Newell EW, Sigal N, Nair N, Kidd BA, Greenberg HB, Davis MM. Combinatorial tetramer staining and mass cytometry analysis facilitate T-cell epitope mapping and characterization. Nat Biotechnol 2013;31:623–9. [32] Soen Y, Chen DS, Kraft DL, Davis MM, Brown PO. Detection and characterization of cellular immune responses using peptide–MHC microarrays. PLoS Biol 2003;1:429–38. [33] Stone JD, Demkowicz WE, Stern LJ. HLA-restricted epitope identification and detection of functional T cell responses by using MHC–peptide and costimulatory microarrays. Proc Natl Acad Sci 2005;102:3744–9.

519

[34] Kwong GA, Radu CG, Hwang K, Shu CJ, Ma C, Koya RC, et al. Modular nucleic acid assembled p/MHC microarrays for multiplexed sorting of antigen-specific T cells. J Am Chem Soc 2009;131:9695–703. [35] Deviren G, Gupta K, Paulaitis ME, Schneck JP. Detection of antigen-specific T cells on p/MHC microarrays. J Mol Recognit 2007;20:32–8. [36] Toebes M, Coccoris M, Bins A, Rodenko B, Gomez R, Nieuwkoop NJ, et al. Design and use of conditional MHC class I ligands. Nat Med 2006;12:246–51. [37] Bakker AH, Hoppes R, Linnemann C, Toebes M, Rodenko B, Berkers CR, et al. Conditional MHC class I ligands and peptide exchange technology for the human MHC gene products HLA-A1, -A3, -A11, and -B7. Proc Natl Acad Sci 2008;105:3825–30. [38] Czerkinsky CC, Nilsson L-A, Nygren H, Ouchterlony O, Tarkowski A. A solidphase enzyme-linked immunospot (ELISPOT) assay for enumeration of specific antibody-secreting cells. J Immunol Meth 1983;65:109–21. [39] Carvalho LH, Hafalla JCR, Zavala F. ELISPOT assay to measure antigen-specific murine CD8(+) T cell responses. J Immunol Meth 2001;252:207–18. [40] Karttunen J, Shastri N. Measurement of ligand-induced activation in single viable T cells using the lacZ reporter gene. Proc Natl Acad Sci 1991;88:3972–6. [41] Turner MJ, Abdul-Alim S, Willis RA, Fisher TL, Lord EM, Frelinger JG. T-cell antigen discovery (T-CAD) assay: a novel technique for identifying T cell epitopes. J Immunol Meth 2001;256:107–19. [42] Siewart K, Malotka J, Kawakami N, Wekerle H, Hohlfeld R, Dornmair K. Unbiased identification of target antigens of CD8+ T cells with combinatorial libraries coding for short peptides. Nat Med 2012;18:824–7. [43] Valentino M, Frelinger J. An approach to the identification of T cell epitopes in the genomic era: application to Francisella tularensis. Immunol Res 2009;45:218–28. [44] Valentino MD, Abdul-Alim CS, Maben ZJ, Skrombolas D, Hensley LL, Kawula TH, et al. A broadly applicable approach to T cell epitope identification: application to improving tumor associated epitopes and identifying epitopes in complex pathogens. J Immunol Meth 2011;373:111–26. [45] Crawford F, Jordan KR, Stadinski B, Wang Y, Huseby E, Marrack P, et al. Use of baculovirus MHC–peptide display libraries to characterize T-cell receptor ligands. Immunol Rev 2006;210:156–70. [46] Hondowicz BD, Schwedhelm KV, Kas A, Tasch MA, Rawlings C, Ramchurren N, et al. Discovery of T cell antigens by high-throughput screening of synthetic minigene libraries. PLoS One 2012;7(1):e29949. http://dx.doi.org/10.1371/ journal.pone.0029949. [47] Lundegaard C, Lund O, Buus S, Nielsen M. Major histocompatibility complex class I binding predictions as a tool in epitope discovery. Immunology 2010;130:309–18. [48] Lafuente EM, Reche PA. Prediction of MHC–peptide binding – a systematic and comprehensive overview. Curr Pharm Des 2009;15:3209–20. [49] Patronov A, Doytchinova I. T-cell epitope vaccine design by immunoinformatics. Open Biol 2013;3:120139. http://dx.doi.org/10.1098/ rsob.120139. [50] Zhang XW. A combination of epitope prediction and molecular docking allows for good identification of MHC class I restricted T-cell epitopes. Comput Biol Chem 2013;45:30–5. [51] Vita R, Zarebski L, Greenbaum JA, Emami H, Hoof I, Salimi N, et al. The immune epitope database 2.0. Nucleic Acids Res 2010;38:D854–62. [52] Ma C, Fan R, Ahmad H, Shi Q, Comin-Anduix B, Chodon T, et al. A clinical microchip for evaluation of single immune cells reveals high functional heterogeneity in phenotypically similar T cells. Nat Med 2011;17:738–43. [53] Freeman JD, Warren RL, Webb JR, Nelson BH, Holt RA. Profiling the T-cell receptor beta-chain repertoire by massively parallel sequencing. Genome Res 2009;19:1817–24. [54] Warren RL, Freeman JD, Zeng T, Choe G, Munro S, Moore R, et al. Exhaustive Tcell repertoire sequencing of human peripheral blood samples reveals signatures of antigen selection and a directly measured repertoire size of at least 1 million clonotypes. Genome Res 2011;21:790–7. [55] Robins HS, Srivastava SK, Campregher PV, Turtle CJ, Andriesen J, Riddell SR, et al. Overlap and effective size of the human CD8+ T cell receptor repertoire. Sci Transl Med 2010;2:47–64. http://dx.doi.org/10.1126/scitranslmed. 3001442. [56] Robins HS, Campregher PV, Srivastava SK, Wacher A, Turtle CJ, Kahsai O, et al. Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta T cells. Blood 2009;114:4099–107. [57] Heemskerk B, Kvistborg P, Schumacher TNM. The cancer antigenome. EMBO J. 2013;32:194–203.