In mammals, thymus-derived T lymphocytes and bone-marrow-derived B lymphocytes can generate wide repertoires of antigen receptors [T-cell receptors (TCR) and B-cell receptors (BCR), respectively]. These receptors enable the recognition of a myriad of antigens in mammals through primary participation in cell-mediated immunity (T cells) and humoral immunity (B cells). T cells are divided into the following two subsets: αβ and γδ T cells, according to the heterodimeric, disulphide bond-linked TCR structure (1). The domain structures of TCRs are strikingly similar to the structure of immunoglobulin Fab fragments, therefore, TCRs are also classified as members of the immunoglobulin superfamily. However, unlike immunoglobulins, which recognize intact protein antigens, the complementarity-determining region (CDR) loops of TCRs only recognize processed fragments of antigens that are presented by major histocompatibility complex (MHC) molecules on antigen-presenting cells (APCs), such as B cells, dendritic cells and macrophages (2,3). These antigens that are recognized by TCR family members include peptides, glycoproteins, lipids and small molecule metabolites. Thus, the resulting TCR-peptide-MHC complex is of great importance in cellular immunity.

αβ TCR are mainly expressed on the surfaces of αβ T and NKT cells, where they play a significant role in adaptive immune response by interacting with foreign antigens that are presented by MHC class I/II molecules on APCs. However, recent studies have described unconventional subsets of αβ T cells bearing TCRs that can recognize ligand antigens that are presented by molecules of the MHC-like CD1 family or vitamin B based metabolites that are bound to MHC-related protein 1 (MR1). These subsets include: i) mucosal-associated invariant T cells which exhibited restricted diversity and are involved in antibacterial immunity (4,5); and ii) invariant natural killer T (NKT) cells and germ-line-encoded mycolyl-reactive (GEM) T cells, which recognize glycolipids with respect to CD1d and CD1b, respectively (6,7). Other less well-characterized molecules are also known to recognize lipids that are presented by CD1a and CD1c molecules. In contrast, γδ TCRs are mainly expressed on the surfaces of γδ T cells, which comprise minority of the T cell pool but may represent more than half of the total T cells in tissues such as the gut and skin. In contrast to αβ TCRs, γδ TCRs recognize lipid-based antigens that are presented by CD1 molecules.

Regarding the functions of these two T-cell subsets, αβ T cells play an important role in protective immunity against various antigens, whereas γδ T cells were previously considered to be a necessary component of innate immunity, the term ‘T cell’ generally refers to αβ T cells. However, in recent years, they have been considered to share important characteristics of both the innate and adaptive immune responses and have garnered increasing attention in the field of immunotherapy because of their prominent functions and direct or indirect participation in disease recovery (8).

Abundant TCR diversity is a prerequisite for broad immunological recognition, and antigen-specific TCRs are an important step in TCR gene transduction-based immunotherapy. Therefore, a clear understanding of TCR structure and diversity generation represents an elementary step in determining the best incorporation of αβ TCRs in disease treatment strategies. Moreover, knowledge regarding TCR bias in different diseases could promote a good prognosis in patients treated via gene transfer. Accordingly, in the present review, we briefly introduce background information regarding TCR structure and diversity. In addition, we summarize recent data concerning TCR selection bias in some diseases, which might serve as a theoretical basis for clinical applications, and thus, further illuminate potential functions in disease treatment.

The highly variable TCR comprises a number of segments that are encoded by genes at discrete chromosomal locations: the α and γ glycoprotein chains, which comprise V (variable), J (joining) and C (constant) regions and the β and δ chains, which feature an additional D (diversity) region (9). The re-arrangement of V (D) and J through recombination signal sequences (RSS) and the formation of a functional antigen receptor via the activity of the lymphoid-specific protein recombination activating gene (RAG) 1 and RAG 2 exclusively occurs during thymocyte development (Fig. 1). This process involves the random re-arrangement of various V and J genes at the TCR-α locus and V, D and J genes at the TCR-β locus during T-cell development, additional diversity is produced by the random insertion or deletion of >20 non-germline nucleotides at region junctions (V-(N)-J, V-(N)-D and/or D-(N)-J; N represents nucleotide), thus, yielding a TCR recognition spectrum with an estimated theoretical diversity of ~10¹⁸ in human (10) and 10¹⁵ in mouse (2) although most of those specificities will never be used during an individual's life, as peripheral repertoire in human are composed of 25×10⁶ clonotypes (11) and 2×10⁶ in mouse (1).

Three defined TCR hypervariable regions (CDR1, CDR2 and CDR3) combine to form the TCR antigen-binding site. Crystallographic analysis has revealed that CDR1 and CDR2 interact with a particular MHC molecule, whereas CDR3 interacts with the antigenic peptide bound to the MHC molecule (12,13). Most variation in each chain lies within CDR3, which is responsible for the specific recognition of and interaction with peptide antigens that are presented by MHC molecules. As this region determines the antigen specificity of a T cell, each CDR3 represents a single T-cell clone (14). CDR1α, CDR1β, CDR2α and CDR2β are entirely encoded by the V region in germline DNA segments, whereas CDR3 loops are encoded by the V (D) J junction and the N additions and deletions during the recombination process; relative to CDR1 and CDR2 loops, CDR3 loops are significantly more diverse, and thus, mediate contact between TCR and the antigenic peptide-MHC complex. In addition, the D segment insertion occurs in CDR3 of the TRB/TRD locus, thus, yielding a broader TRB repertoire for antigen recognition compared with TRA and TRG. Whereas the CDR3 loop size distributions of the IgH and L chains and TCR γ and δ chains are relatively broad and dissimilar, in contrast, the CDR3 loop size distribution of the α and β chain are narrow and closely matched, suggesting that a pairing of TCR α and β chains with similar CDR3 loop sizes might be generally required to generate a functional αβ TCR repertoire (14,15). Furthermore, the greater diversity of the CDR3β loop (relative to the CDR3α loop) might indicate that the TCR Vβ chain is the main factor in determining TCR usage bias (16).

As each CDR3 sequence represents one T cell clone, the polyclonal or oligoclonal expansion of T cells can be determined through the detection of CDR3 spectratype by various methods such as PCR DNA blotting, PCR GeneScan sequence analysis, and high throughput TCR sequencing (TCR-seq) (17). The biased TCR AV and BV gene families are considered to be antigen-specific and can be used in immunotherapy.

TCR diversity largely remains constant throughout the life span, except in infants and the elderly, because most specificity is generated in the thymus. However, under conditions of antigen exposure, the expansion of specific clonotypes is a common feature of immunity (Table I). Antigen-specific TCRs will demonstrate a preferential usage of particular TCR variable region gene segments, resulting in a TCR spectrum with a ‘skewed’, ‘immunodominant’, ‘restricted’ or ‘limited’ distribution, known as a ‘biased’ repertoire (16). TCR bias can be mainly classified into three types. Type 1 bias is characterised by selecting a single TCR gene family but with obvious diversity in the CDR3 sequences of different individual clones. Type 2 bias refers to the selection of the same amino acid motif in the CDR3 region of an antigen-specific TCR. Type 3 bias indicates a similar spectrum sequence, with differences resulting from different α and β chain pairings. Several factors can influence TCR bias, including thymic selection, the initial immune response, persistent infection and selection of a ‘private’ versus ‘public’ TCR (16).

In addition to the various disease treatments that have already been applied in the clinic, including tyrosine inhibitors and monoclonal antibody targeting of tumor-associated surface antigens, and potential suppression of antitumor immune responses, cytokines represent another form of immunotherapy that can modify the outcomes of innate or adaptive immunity, antitumor vaccines and adaptive gene therapies (68–78). However, the increased toxicity resulting from cytokine therapy can induce lymphocyte degeneration, reduce TCR repertoire diversity and increase susceptibility to opportunistic pathogens (79). To date, there are three kind of therapies: tumor-infiltrating lymphocytes (CILs), chimeric antigen receptors (CARs) and T cell receptor engineered T cells (80). This kind of transfer of lymphocytes to mediate an effector function is known as adaptive cell transfer (ACT). TILs have been developed with slow but continuing progress (80). Antigen-associated monoclonal TCR Vα and Vβ gene families can be defined according to the length and sequence of the CDR3 region. In vitro experiments have confirmed that following efficient separation and amplification, these specific TCR genes can be incorporated into antitumor or antiviral disease therapy. However, it is difficult to gain a sufficient number of CTLs from the patients' T cell populations, thus, presenting a challenge to the application of this technique. In addition to the limited number of CTLs, immunosuppressive factors might help to downregulate effector T-cell functions. Accordingly, it would be meaningful to study the features and amplification methods that are associated with tumor-specific CTLs because these will play pivotal roles in antitumor adoptive immunotherapy.

All of these factors increase the attractiveness of T cell transduction-based immunotherapy, particularly with regard to long-term survival. The endowment of T cells with high-affinity antigen specificity via TCR transduction has highlighted the growing importance of this effective method with respect to the control and elimination of malignant cells and infectious agents.

Mainstream research has focused on the transduction of tumor or viral antigen-related TCR genes into normal T cells, including mandatory expression of the idiotypic TCR; as a result, the recognition patterns of endogenous TCRs are changed, leading to an increase in cytokine secretion and lethality after a large amplification, an immune response and the generation of the adequate numbers of specific CTLs during treatment. During this process, the TCR repertoires are dynamically monitored before, and for several cycles after therapy, by analysing the CDR3 length distributions within CD4⁺ and CD8⁺ T cell subsets. Recently, an in vitro expansion of tumor-specific CTLs was applied with clinical results. These TCR gene-engineered T cells were first used in the clinical treatment of melanoma. The introduction of a MART-1 antigen specific TCR gene into the peripheral blood lymphocytes of patients with melanoma may restore the capacity to resist positive tumor-surface antigens and might completely alter the disease prognosis.

Two methods can facilitate the transfer of exogenous genes into cells: viral vectors or chemical/physical methods (81). Viral vectors like gamma retrovirus and lentivirus enable efficient gene transfer with the least amount of damage; however, viral vectors are expensive, and the process represents a waste of time and resources. In addition, viral vectors cannot deliver large DNA fragments because of package capacity limitations. Furthermore, the potential for widespread viral contamination limits the application of this technique. In comparison, although physical methods, such as electroporation, can induce more cell death, these methods are faster, easier and safer methods of gene transfection, which explains their prominence. In recent years, a more advanced electroporation-based gene transfer method involving nucleofection has been introduced; this method transfers the gene of interest into the nuclei of non-dividing cells, with an high transfection efficiency.

In 1999, CD8⁺ T cells that were genetically engineered to express the melanoma MART-1-restricted TCRα/β gene exhibited cytotoxicity against HLA-A2-restricted melanoma cells (82). In 2006, the first clinical trial was conducted (83); condition of 2 patients was completely restrained in 15 patients, win 18 months life time without this disease. Many examples of immunotherapy have been reported. The transduction of an HIV-specific TCR gene into patients with HIV-1⁺ can enhance immunity. In addition, Tang et al (84) recently reviewed the mechanism by which the tumor microenvironment (TME) affects the efficiency of immunotherapy and how to best use TME as a contributor to immunotherapy. However, this kind of introduction of exogenous TCR gene may lead to mispairing of the endogenous TCR subunits with introduced TCR chain, which may alter the TCR specificity of antigen recognition, not only lead to the loss of antitumor, but also the formation of autoreactive T cells (85–87). Hu et al (91) have shown the humanized mouse (hu-mouse) model which transplantation of human fetal thymus tissue and CD34⁺ fetal liver cells into immunodeficient mice, leading to the development of human lymphohematopoietic cells and the formation of secondary lymphoid organs (88–90). They used this hu-mouse model to generate melanoma antigen (MART-1) specific human T cells for studies of human cancer immunotherapy. They found that MART-1 specific T cells can be generated after transduction of lentiviruses containing MART-1 specific TCR gene, implying that it was possible to produce large quantity of MART-1 specific T cells with antitumor activity.

The third and a powerful new immunotherapeutic approach intended to target TAAs comprises chimeric antigen receptors (CARs), which comprise an extracellular region - the antigen-binding domain of a monoclonal antibody (containing the variable domains of the light and heavy chains), a stem-like region, a transmembrane region and the intracellular signalling portion of a T-cell receptor, can combine the specificity and effectiveness of a monoclonal antibody with the cytotoxic effects and long-term persistence of CTLs (92), thus, it can make it possible to endow immune system with reactivity and add benefits which can be seen with cytotoxic chemotherapy or targeted therapy of the rapid onset of action. This method has been proven safe and effective for the induction of tumor remission in patients with neuroblastoma (93) and hematological malignancies (94). To date, the most successful CARs were those specific for CD19 on B cell malignances (92). Several research centres have confirmed the compelling efficacy of CARs against acute and chronic B-cell malignancies (95).

Compared with TCRs, there are several different parameters. First of all, these two format receptors target different antigens, TCRs react to natural ligands with MHC restriction, and it is necessary to enhance the affinity of the TCR with specific peptide-MHC in order to mediate the activities of T cells against the peptide-MHC antigen during the process of adoptive T cell therapy. In contrast, CARs are not MHC-dependent, and can function both in CD4⁺ and CD8⁺ T cells. Furthermore, CARs can induce activity even against low-density antigens. The CAR density, antigen-binding domain affinity still need to be optimized to match the antigen density of tumor and tissue (92). However, risk also exits because both on-target and off-target recognition of normal tissue occur in engineered T cells. For instance, in patients treated with carcino-embryonic antigen specific TCR modified T cells showed on-target toxicity, resulting in inflammatory colitis in normal colon (96). Apart from toxicity, both TCR and CARs will give rise to autoimmunity and inflammation from the infusion of ex vivo activated autologous lymphocytes. However, the incidence can be calculated to be less than one in 1,000 patients, which is lower than cytotoxic chemotherapy (97). A common challenge to all therapies, including antigen specific TCR gene based therapy, is to develop cost-effective, and efficient manufacturing and delivery capabilities (80).

Gene re-arrangement during the thymic development of T cells introduces sufficient TCR diversity to meet the varied requirements for antigen recognition. Over the last two decades, considerable research has demonstrated a close relationship between TCR diversity and immune responses against various antigens. Under natural conditions, the TCR repertoire is diverse and exhibits a Gaussian distribution, and the CDR3 of each gene family is represented by at least eight different lengths. As each CDR3 sequence represents a specific T-cell clone, CDR3 region spectratyping can reflect TCR clonality, and antigen-specific TCRs can be selected for immunotherapeutic use. A basic understanding of the T-cell receptor structure and recombination mechanisms can help us to understand the aetiology of TCR diversity and thus identify antigens.

An analysis of TCR bias in tumor, pathogenic infection, autoimmune disease and other disease settings is a basic prerequisite to a deep understanding of the immune mechanisms underlying antigen-specific responses and specific immunotherapy. Many studies have described how TCR abundance might be affected by certain infections, malignancies and immune dysfunctions. A broader specific peptide-MHC complex repertoire can help to reduce the emergence of a persistent viral infection. Three factors, thymic selection, initial immune response and persistent infection, are all known to influence specific TCR selection. In addition, some pathologies can induce the appearance of specific TCR clonotypes, which might be useful as immunological markers in clinical settings.

An understanding of T-cell responses against specific antigens is particularly important with respect to adaptive immunity. The immunotherapeutic role of αβ T cells has been strongly focused on their function in recovery from disease. In contrast, how would one capture the ideal TCR? It is necessary to monitor CDR3 spectratypes and conserved amino acid motifs. In addition to several above mentioned common methods, nascent high-throughput sequencing (TCR-seq) using next-generation sequencing also provides new insights to the sequence analysis of TCR repertoires under conditions of health and disease. The transduction of specific TCR genes with conserved CDR3 motifs into normal T cells can alter the original identification model and form antigen-specific T cells; this approach to disease treatment has yielded results that could further enhance the likelihood of survival. However, new strategies and methods still need be developed to enhance specific recognition between TCRs and antigens, with the aim of improving immunotherapeutic approaches to persistent infection and cancer.