Immunodominance and Immunodomination: Critical Factors in Developing Effective CD8+ T‐Cell–Based Cancer Vaccines
Introduction
The terms “immunodominance” and “immunodominant” were originally used to describe prominent humoral responses to various antigens in the late 1960s and early 1970s (Benacerraf 1972, Curtiss 1975, Johnston 1968). This was not long before the discovery of major histocompatibility complex (MHC) restriction (Zinkernagel and Doherty, 1974) but a decade prior to the discovery that antigenic peptides are presented by MHC molecules (Allen 1984, Bixler 1983, Townsend 1986). It was originally noticed that the cytotoxic T‐lymphocyte (CTL) responses to antigens, such as the minor histocompatibility antigen (MiHA) H‐Y and lymphocytic choriomeningitis virus (LCMV), were restricted by a single or just a few H‐2 haplotypes (Bevan 1975, Zinkernagel 1975). More detailed studies of CTL responses to mouse sarcoma virus (MSV) (Gomard et al., 1977) and influenza A viruses (IAV) (Doherty et al., 1978) revealed that responses were strongly linked to mouse H‐2 alleles. Using recombinant inbred mouse strains expressing different H‐2 genes and then transplanting lymphomas as targets, Gomard et al. (1977) identified H‐2Kd as the major restriction allele for anti‐MSV (mouse sarcoma virus) CTL responses whereas the H‐2Dd allele was not involved in these responses. Doherty et al. (1978), using similar mouse strains and fibroblast cell lines as CTL targets, defined H‐2Kb as a “nonresponder”MHC gene locus in anti‐IAV responses despite this being a responder in antirecombinant vaccinia virus (rVV) responses. These observations were made as part of the demonstration of MHC‐restriction of CTL responses and constituted the first indication of the H‐2 allele‐associated immunodominance. It is now well established that antigen‐specific immunodominant responses are linked to particular MHC alleles (Belz 2000a, Schirmbeck 2002, Tourdot 2002, Tussey 1995), known initially as “immune response (Ir)” genes (Solinger 1979, Zinkernagel 1978).
From such beginnings, it was soon found that even within the strains that responded to a given virus, the CTL responses were highly focused on just a few proteins and often on a single polypeptide (Bennink 1988, Bennink 1987). These “immunodominant responses” were traced to a subregion of the protein antigen (Lamb and Green, 1983) around the time when the peptide nature of antigenic determinants was being elucidated (Allen 1984, Bixler 1983, Townsend 1986).
Nearly simultaneously, Sercarz and colleagues studying CD4+ T‐cell responses to an artificial antigen, hen egg lysozyme (HEL), observed that the CD4+ T‐cell responses to some determinants were easily detectable whereas responses to other determinants were much smaller and consequently harder to demonstrate. There were yet other determinants that were not detected under normal circumstances unless very high levels of antigen were used for priming. Determinants involved in detectable responses were defined as either immunodominant determinants (IDDs) or subdominant determinants (SDDs) depending upon their reproducibility and magnitude; the third category of undetectable determinants were called cryptic determinants (Sercarz et al., 1993), a term derived from their earlier studies on anti‐HEL B‐cell responses and autoimmune responses (Furman 1981, Wicker 1984a, Wicker 1984b).
Similar phenomena were observed in mouse models of IAV, LCMV, herpes simplex virus (HSV), and Listeria monocytogenes (LM) infection. In a given mouse strain, the major responses to these pathogens were often directed toward a single IDD (Table I). However, progress in unraveling the factors that controlled immunodominance was limited by the lack of tools for immune monitoring. The field accelerated when novel methods for enumerating arrived in the late 1990s. These technologies included MHC‐peptide tetramers (Altman et al., 1996) and intracellular cytokine staining (ICS) of antigen‐specific T cells (Jung et al., 1993). These new techniques allowed accurate enumeration of specific T cells in combination with their surface and functional markers in the absence of in vitro T‐cell expansion (Butz 1998, Flynn 1998, Murali‐Krishna 1998). Moreover, the newer methods were up to 100‐fold more sensitive in detecting Ag‐specific than the established limiting dilution analysis (LDA) that measured CTL precursor frequencies indirectly through target killing (Lalvani 1997, McMichael 1998, Murali‐Krishna 1998). This observation suggested that historical estimates of Ag‐specific numbers might have been drastically underestimated. Since then, many experiments have reassessed these systems and revised the estimates of Ag‐specific numbers (Flynn 1998, Murali‐Krishna 1998). The newer technologies were not only more sensitive in detecting the immunodominant but in particularly the subdominant were more readily appreciated. It was then possible to quantitate responses reproducibly and define “immunodominance hierarchies” (Belz 2000a, Belz 2000b, Chen 2000).
While tetramers, or multimers of MHC/peptide complexes, allowed assessment of specific T cells at various stages of their development and differentiation following antigen‐specific activation, another key technological advance was the development of T‐cell receptor (TCR) transgenic (Tg) mice expressing a “monoclonal” T‐cell repertoire (Berg 1988, Bluthmann 1988). Transfer experiments involving TCR Tg‐T cells enabled these T cells to be tracked directly ex vivo at early stages of the immune response and also permitted the in vivo study of T cell–T cell (T‐T) and/or T cell–antigen‐presenting cell (T‐APC) interactions.
A large body of work has since shown how immunodominance might be controlled or influenced by the steps involved in the creation of antigenic peptides (Ag‐processing) and their presentation by the MHC molecules on the APC (Ag‐presentation). Therefore, in this chapter, we begin by addressing the potential contribution of antigen processing and presentation to the establishment of immunodominance hierarchies; we then focus on T‐T and T‐APC interactions; we next discuss the positive and negative roles that immunodominant play during viral and tumor escape of immune surveillance. Finally, we explore the possibility of better vaccine development, utilizing the knowledge accumulated from studying immunodominance. We confine our discussions to MHC class I (MHC‐I)–restricted responses as MHC class II‐restricted immunodominance has been reviewed elsewhere (Latek 1999, Sercarz 2003).
Section snippets
The Phenomenon: Immunodominance
Immunodominance has been mostly observed in antiviral and antibacterial immune responses, both in mouse models and in human diseases. Table I shows a list of the best‐studied antigen systems in mouse models illustrating the generality of well‐focused immune responses in antiviral and antibacterial immunity. Compared to laboratory mouse models, human populations express many different human leukocyte antigen (HLA) haplotypes and are repeatedly exposed to various pathogens (some
Theoretical Contributions of Antigen Processing and Presentation to Immunodominance
The demonstration that antigen presentation occurs in the form of short peptides bound to MHC‐I molecules (Allen 1984, Bixler 1983, Bjorkman 1987, Townsend 1986) led to the elucidation of the detailed biochemistry and pathway of this process, which has been reviewed extensively (Kloetzel 2004b, Trombetta 2005, Van Kaer 2002, Yewdell 1999, Yewdell 1999) and will only be briefly addressed here.
The steps involved in antigen processing and presentation predict a number of factors that might lead to
Immunodomination and its Possible Mechanisms
Immunodomination, also described as T‐T competition (Kedl et al., 2003), refers to circumstances in which the T‐cell response to a given antigenic determinant is inhibited or suppressed either directly or indirectly by T cells specific to other antigenic determinant(s). Immunodomination usually involves T cells specific for determinants from the same antigen or same pathogen. The suppressing are usually specific to IDD, although this might not be necessarily true in the context of
Immunodominance is Susceptible to Viral Escape
Viruses have coevolved with humans and other species and developed many clever mechanisms to escape the host immune surveillance. There are many general types of escape strategies. For instance, HSV encodes the protein ICP47 to block TAP function; CMV encodes multiple proteins, including US2 and US11, which cause host MHC molecules to exit the ER thus targeting them for cytosolic degradation. In both instances, the result is to limit antigen presentation by infected host cells. HIV and
Immunodominance in Antitumor Responses and Tumor Escape
An important role for in suppressing some tumors has been demonstrated experimentally and is termed immunosurveillance (Smyth and Trapani, 2001). also play a critical role in shaping tumor development, and this is referred to as immunoediting (Dunn et al., 2002). In many human tumors, a common feature is the loss of MHC‐I expression. In at least 80% of such cases, a single HLA molecule is lost (Marincola 2000, Ruiz‐Cabello 1991). Haplotype loss occurs much less frequently (Romero et
Immunodominance and Cancer Vaccines
Over the past decade or so, many tumor antigens and their antigenic determinants for both CD4+ and CD8+ T cells have been identified. These serve as the primary focus for many different cancer vaccine strategies both as therapeutic targets and for immune monitoring. Therapeutic cancer vaccines are different from prophylactic childhood immunizations, which are nearly always conducted in the absence of disease. Most cancer vaccines are used in patients with either in situ or resected tumors. In
Conclusions
The development of effective, cellular immunity‐based viral or cancer vaccines will depend upon our general understanding of the mechanisms that determine immunodominance and immunodomination. Taken together, lessons learned from anti‐viral immunity and other model CTL systems are highly likely to be relevant in anti‐tumor immunity. Optimizing the balance of dominant and subdominant responses will require a detailed understanding of key target antigens and host HLA types. In addition, the
Acknowledgments
The authors thank Dr. Ken Pang for critical reading of this chapter.
WC is supported by an International Senior Research Fellow Fellowship from the Wellcome Trust (066646/Z/01/Z).
References (223)
Fusions of breast cancer and dendritic cells as a novel cancer vaccine
Clin. Breast Cancer
(2003)- et al.
Epitope specificity of H‐2Kb‐restricted, HSV‐1‐, and HSV‐2‐cross‐reactive cytotoxic T lymphocyte clones
Virology
(1993) - et al.
Massive expansion of antigen‐specific CD8+ T cells during an acute virus infection
Immunity
(1998) - et al.
The role of dendritic cell subsets in immunity to viruses
Curr. Opin. Immunol.
(2003) - et al.
Dissecting the multifactorial causes of immunodominance in class I‐restricted T cell responses to viruses
Immunity
(2000) - et al.
Immunodominance of H60 is caused by an abnormally high precursor T cell pool directed against its unique minor histocompatibility antigen peptide
Immunity
(2002) - et al.
The relative immunodominance of haptenic determinants on a complex hapten phage conjugate
Immunochemistry
(1975) - et al.
Virus‐specific CD8+ T cells in primary and secondary influenza pneumonia
Immunity
(1998) - et al.
MHC antigens and tumor escape from immune surveillance
Adv. Cancer Res.
(2001) - et al.
In vivo depletion of CD11c(+) dendritic cells abrogates priming of CD8(+) T cells by exogenous cell‐associated antigens
Immunity
(2002)