Introduction: the CTL response to Epstein–Barr virus

Epstein–Barr virus (EBV) is a γ-herpesvirus that is carried by more than 90% of adults worldwide as an asymptomatic persistent infection. This virus is the causative agent of infectious mononucleosis, a benign lymphoproliferative disease and is also implicated in the pathogenesis of an increasing number of malignant diseases, including: Burkitt's lymphoma (BL), nasopharyngeal carcinoma and Hodgkin's disease (Miller, 1990). The importance of cellular immunity, and especially of CTLs, in controlling the potentially harmful consequences of EBV infection is underscored by the increased incidence of EBV-driven lymphoproliferations in organ transplant recipients receiving immunosuppressive therapy and in AIDS patients (Thomas, Allday & Crawford, 1991).

The pathogenic potential of EBV is well illustrated by the ease with which the virus can infect resting B lymphocytes in vitro, causing growth-transformation and the establishment of lymphoblastoid cell lines (LCLs) with infinite doubling capacity. This in vitro infection model provided the first clue as to how the virus is controlled in vivo. If circulating lymphocytes from EBV-seropositive donors are infected with EBV in vitro without first removing or inactivating the T lymphocytes, then the initial proliferation of EBV-infected B lymphocytes is followed by a regression at two to three weeks post-infection because of reactivation of EBV-specific memory T lymphocytes that eliminate the virus-infected cells (Rickinson et al, 1981). Activated T cells isolated from such cultures contain CTL populations that are HLA class I-restricted and specific for EBV-encoded antigens.

Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory disease with autoimmune features chiefly affecting the synovial joints. In common with many other autoimmune diseases, there is emerging evidence to suggest that susceptibility to RA is determined by several genes (Wordsworth & Bell, 1991). However, despite the evidence of a genetic predisposition (in particular HLA-DR4 or HLA-DR1), it is clear that the genetic component is probably relatively small since concordance rates in monozygotic twins are no more than 30%, suggesting that other factors are involved (Wordsworth & Bell, 1991; Rigby et al., 1991; Silman, 1991).

The target tissue for the autoimmune and inflammatory response in RA is the synovial joint. In healthy individuals, the synovial membrane is virtually acellular, but in RA it is infiltrated extensively by large numbers of cells from the blood, including activated T cells, macrophages and plasma cells (Janossy et al., 1981; Fϕrre, Dobloug & Natvig, 1982; Burmester et al., 1982). In addition, there is proliferation of fibroblasts in the lining layer of the synovial membrane. Many of these infiltrating and resident cells including endothelial cells, fibroblasts and T cells normally express little, if any, HLA class II antigens but are activated in the RA synovium and express abundant HLA class II antigens (Førre et al., 1982; Klareskog et al., 1982; Burmester et al., 1987). These cells may act as antigen-presenting cells and be involved in the stimulation of autoreactive T cells.

Introduction

Many animal viruses have the property of being able to modulate the expression of the MHC antigens of infected host cells and sometimes of uninfected cells of the host animal. Since the MHC is central to the immune system, this may be beneficial to the virus in evading an immune response and, therefore, have consequences for disease. The significance of such modulation can be seen from how widespread it is, its effects on pathogenicity and its prevalence in common and important human (and animal) infections. It will be shown here and below (Chapters 7 to 12) that modulation of MHC by viruses (and, briefly, other pathogens) is indeed widespread, does affect pathogenicity and is present in common and important infections. It will, therefore, be possible to conclude that modulation of MHC antigens is a highly important characteristic of viruses in general.

This chapter describes the different viruses for which modulation of MHC has been observed. Different mechanisms involved in modulation, both potential and actual, are summarized with suitable examples where known. This information is related to the pathogenicity of the virus where data are available. In the overall context of infectious disease, a brief note is made of other infectious organisms (i.e. microorganisms other than viruses) that are known to affect expression of functional MHC antigens with immunological consequences. Conclusions are drawn from these data on the importance of the modulation of MHC expression and on suitable directions for further research.

Introduction: immune recognition of tumour cells

In the late 1980s it became clear that cancer cells develop by multiple genetic alterations (reviewed in Weinberg, 1989; Bishop, 1991). These alterations include activation of proto-oncogenes as well as inactivation of tumour suppressor genes. Proto-oncogenes, more commonly called oncogenes, exert important functions in cell proliferation and differentiation and their activity is usually tightly controlled to ensure a minimal risk of inappropriate activity. Activation of oncogenes in animal and human tumours may occur through several mechanisms including amplification, elevated expression and point mutations. These genetic alterations usually result in an altered activity of the oncogene-encoded protein and this contributes to uncontrolled proliferation.

In the light of the crucial role that HLA class I antigens play in the interaction of altered self antigens or viral antigens with CTLs (see Chapter 1), one would expect that antigens specifically present in tumours are presented by HLA class I molecules. Such antigens might be viral antigens in the case of virally induced tumours or altered self proteins in the case of tumours induced by xenobiotics like carcinogens or radiation. In numerous animal and human tumours, mutations in oncogenes have been shown to be responsible for the tumorigenic properties of the tumour cell. These mutations are potential targets for recognition by the immune system of the host. Candidate oncogene and tumour suppressor proteins are ras and p53 proteins, respectively; these can be activated by various mutations and are involved in many forms of human cancer (reviewed in Bos, 1989; Levine, Momand & Finlay, 1991).

Introduction

Demyelinating disorders of the CNS, while bound together by a common process involving damage to or improper laying-down of the myelin sheath, do not share a common aetiology. This is demonstrated by recent advances, which have enabled the classification of these conditions into three groups based on disease aetiology: genetic, viral and autoimmune.

Pelizaeus–Merzbacher disease exemplifies the genetic component of certain demyelination conditions. This X-linked recessive disorder affecting CNS myelination in children is attributed to deficient biosynthesis of the proteolipid protein (PLP) (Koeppen et al., 1987). Lack of PLP, one of the most abundant components of the myelin sheath, contributes to physical symptoms such as spastic extremities and lack of purposeful movement and head control.

The viral aetiology of demyelinating diseases is seen in progressive multifocal leukoencephalopathy (PML). Originally described as a complication of leukaemia and Hodgkin's disease (Aström, Mancall & Richardson, 1958), PML develops mostly in immunocompromised patients with chronic diseases and, more recently, in individuals with AIDS (Levy, Bredsen & Rosenblum, 1985; Niedt & Schinelle, 1985). Symptoms in PML reflect the widespread destruction of the CNS white matter. Neurological deficits include dementia, confusion, aphasia, hemiparesis and ataxia. The viral aetiology of PML has been demonstrated by isolation of the papovavirus JC from the brains of diseased patients (Padgett et al., 1971) and by the demonstration that transgenic mice with part of the JC viral genome develop a neurological disorder that involves dysmyelination of the CNS (Small et al., 1986; Feigenbaum, Hinrichs & Jay, 1992).

Introduction

T lymphocytes play a central role in antigen-specific immune responses through their interactions with other cell types. These interactions are guided by specific T cell recognition of antigen associated with cell surface molecules of the MHC. The two main types of MHC molecule (class I and class II) are both involved in antigen presentation, but to different types of T cell (expressing CD8 or CD4 antigens, respectively). In addition, they normally show very different patterns of expression within tissues: most cell types of the body express MHC class I antigens, whereas the expression of MHC class II antigens is restricted primarily to cells of the immune system – macrophages, dendritic cells, B lymphocytes and (in humans) activated T lymphocytes. Teleologically, this accords with the usual functions of the T cells, which interact with antigen associated with the different types of MHC protein. Thus, class I-restricted, CD8 cytotoxic T cells may be required to interact with antigen presented by almost any cell of the body (for example, following viral infection). By contrast, a major role of class II-restricted, CD4+ helper T cells is to interact with, and help, other cells of the immune system.

Tissue distribution of MHC antigens

The distribution of MHC antigens is now known to be complex. The expression of class I antigens varies in intensity with, for example, a gradation of expression from strong to weak or negative through the following cell types: cells of the immune system, epithelial cells of the gastrointestinal and respiratory tracts, endocrine cells, striated muscle cells, hepatocytes and neurons (reviewed by Pujol-Borrell & Todd, 1987).

Introduction

The MHC is localized on the short arm of chromosome 6 in band 21.3. It extends over 4 Mb of DNA, which have been intensively studied for several decades. Detailed physical maps have been derived and the entire complex has been cloned in cosmid and yeast artificial chromosome (YAC) vectors. About 80 genes have been identified so far, which include the genes encoding the class I or classical transplantation antigens, the class II immune response genes, the class III genes originally defined by complement components and a number of novel genes involved in antigen processing or of unknown function (Trowsdale, Ragoussis & Campbell, 1991). It is intriguing that so many genes involved in immune responses are closely linked to each other in the genome. The class I and class II sequences have been maintained together through evolution and can be found on the same chromosome in species like chicken, mouse and humans. The organization and functional relationships between MHC genes will be presented in this chapter along with a description of MHC transcripts.

The class I region

The human class I region

The class I region is 2 Mb in length and contains the class I multigene family comprising about 20 non-allelic DNA sequences (Fig. 2.1).

Introduction

Human adenoviruses (Ads) possess two well-studied mechanisms for modulation of MHC class I expression. The mechanism that is predominant in infected cells involves down-regulation of surface class I antigens caused by a block in the transport of class I heavy chains to the cell surface. This block is effected by a protein product of a viral early gene (termed E3) and seems to operate in cells infected with most virus serotypes. In adenovirus-transformed cells, the level of surface class I antigens can be either elevated or decreased depending on the serotype of the transforming adenovirus and is controlled mainly at the step of transcription of class I genes. Modulation of the rate of initiation of class I gene transcription in these cells is mediated by the product(s) of a different viral early gene (E1A). Interestingly, it is the expression of the E1A gene of highly oncogenic adenoviruses (such as Adl2) that results in down-regulation of class I transcription, providing a possible mechanism whereby oncogenic Adl2-transformed cells can escape host immune surveillance and form tumours. Adenovirus-infected and adenovirus-transformed cells are, therefore, interesting experimental systems for the study of class I modulation.

Adenovirus transformation and oncogenicity have been extensively reviewed (Knippers & Levine, 1989; Boulanger & Blair, 1991; Chinnadurai, 1992; Moran, 1993) and comprehensive reviews on virus structure, biology and pathogenicity are also available (Ginsberg, 1984; Doerfler, 1986; Horwitz, 1990a,b). The wider aspects of the immunobiology of adenoviruses have also been described (Wold & Gooding, 1991; Braithwaite et al, 1993).

Introduction

Human cytomegalovirus (CMV) is a ubiquitous agent that rarely causes disease in healthy individuals but is an important pathogen in the immunocompromised or the immunologically immature foetus. Individuals at high risk are those with deficiencies in cell-mediated immunity and it is, therefore, believed that in the immunocompetent host the cell-mediated immune response is the most important form of defence against CMV (Grundy, 1991). In the immunocompromised, CMV infection can be associated with a wide range of symptoms. The most serious complication in allogeneic transplant recipients is CMV interstitial pneumonitis, which is associated with a high mortality (Meyers, Flournoy & Thomas, 1982), and the virus remains the most important infectious cause of death following bone marrow transplantation. In AIDS patients, CMV can cause sight-threatening retinitis (Collaborative DHPG Treatment Study Group, 1986) and serious disease throughout the gastrointestinal tract. Congenital CMV infection can be associated with disseminated disease, including in the CNS, and the virus is an important cause of deafness and mental retardation (Alford et al., 1990). Thus, CMV causes disease in many organ systems in various patient populations. However, even in the immunocompetent host, the virus is not eradicated from the body following primary infection and can persist in a latent form throughout life. Hence, whilst the normal host immune response is usually able to prevent disease associated with primary CMV infection, the virus ultimately evades the host response and becomes latent.

Introduction

This chapter deals with transcriptional regulation of MHC class I genes in normal physiology and development. The main focus is transcription factors involved in regulating class I gene expression and the mechanism of action of these factors.

Class I genes are widely expressed in many adult tissues. However, levels of class I expression vary among different tissues; some tissues do not express class I genes at a significant level (Klein, 1975). While lymphoid tissues express class I genes at high levels, there is virtually no expression in the CNS. The lack of class I gene expression in the CNS constitutes a unique immunological environment, which accounts for persistent infection by some viruses (Joly, Mucke & Oldstone, 1991). Other tissues, such as pancreas and muscle, appear to be low in class I gene expression (David-Watine, Israel & Kourilsky, 1990a). Class I gene expression is developmentally controlled: class I mRNAs are not found until mouse embryos reach mid-somite stage (Ozato, Wan & Orrison, 1985). Levels of MHC class I expression in embryonic tissues vary but are higher in haemato/lymphopoietic tissues than in other tissues (Hedley et al, 1989). Class I mRNA levels rise sharply during the first week of neonatal development in the mouse (Kasik, Wan & Ozato, 1987). So far, the molecule responsible for the primary onset of class I expression during embryogenesis, or that which triggers the second, post-natal upsurge of class I expression, has not been identified.

Introduction

The MHC class II (also known as immune response-associated antigens, or Ia) genes are a multigene family encoded on chromosome 6 in humans (Fig. 5.1) and chromosome 17 in mice (see Chapter 2). The functional role of the MHC class II antigens in the immune response is described extensively in Chapters 1 and 3. This includes their function in antigen presentation, thymic education and T cell recognition/activation, their contribution to transplant acceptance or rejection, their genetic association with diseases, their aberrant up-regulation in a number of autoimmune and inflammatory diseases as well as their absence in specific immunodeficiencies. The involvement of MHC class II gene products in a large number of clinical conditions confers practical importance to the study of their regulation. Furthermore, the tissue-, cellspecific, differentiation-dependent and cytokine-regulated expression of MHC class II genes represent features that have made these genes a model system to study gene regulation. In this chapter, the expression and regulation of MHC class II genes in normal and disease states will be discussed, as well as the molecular basis for the constitutive and inducible modes of class II gene regulation.

Role of the MHC class II antigens in the immune response

The specificity of the immune system is conferred by the binding of foreign antigens through receptors on the surfaces of B and T lymphocytes, the immunoglobulin (Ig) molecule and the T cell receptor (TCR), respectively.

Introduction

Cervical cancer and pre-cancer form a disease continuum ranging from cervical intraepithelial neoplasia (CIN) through microinvasion to invasive carcinoma; about 70% of the tumours are squamous and 30% are adeno- and adenosquamous carcinomas (Buckley & Fox, 1989). Most tumours are thought to develop from an area of intraepithelial neoplasia within the transformation zone (Coppelston & Reid, 1967). Cervical cancer is estimated to be the second most common female cancer with approximately 500 000 new cases per annum worldwide (Parkin, Laara & Muir, 1980). Sexually transmitted infections are recognized as one of the major risk factors and the active agents are thought to be specific types of human papillomavirus (HPV) (Munoz et al., 1992).

Papillomaviruses are small DNA viruses associated with benign and malignant proliferative lesions of cutaneous epithelium. Over 60 different types of papillomavirus have been described and they can be segregated into groups distinguished by DNA sequence homology and the specific lesions with which they are associated (de Villiers, 1989). HPV 6 and 11 are found most commonly in cervical condyloma, benign lesions that tend to regress spontaneously, and low-grade CIN. HPV 16 and 18 are the types most commonly associated with high-grade CIN lesions and invasive carcinoma of the cervix. The viruses will replicate only in specific differentiation stages of epithelia, which limits the use of in vitro culture methods for producing HPV. To circumvent this, molecular biological techniques have been utilized extensively to characterize HPV.

Introduction

The expression of MHC products by pancreatic islet cells has been extensively studied because these cells are the targets of the destructive autoimmune response that leads to insulin-dependent diabetes mellitus (IDDM) and also because they constitute a set of distinct cell populations ideal for the study of peripheral tolerance by gene targeting in transgenic (tg) mice. Many of the studies described in this chapter were first prompted by concepts previously proposed in the aberrant class II expression hypothesis of endocrine autoimmunity (Bottazzo et al., 1983).

The aberrant class II expression hypothesis of endocrine autoimmunity

Aberrant (actually ectopic) HLA class II expression, i.e. the expression of HLA class II by cells of lineages that are normally class II negative, was first detected in the thyroid follicular cells of glands resected from patients suffering from autoimmune thyrotoxicosis, or Graves' disease (Hanafusa et al., 1983). This finding, together with the demonstration that thyroid follicular cells can be induced to express HLA class II antigens in vitro (Pujol-Borrell et al., 1983), led to the formulation of the aberrant class II expression hypothesis. This is based on the assumption that the lack of active immune response to endocrine cells and other scarce and differentiated cells present in peripheral tissues is the result of their low expression of HLA proteins, which makes them ‘invisible’ to T cells (immunological silence).

Among the various known ecological factors that determine host–parasite relationships is the specific binding of microorganisms to cells and tissues. This important process is called adherence; the microbe-borne molecule that connects with a host receptor is an ‘adhesin’. Like those tenacious microorganisms inhabiting streams and other marine environments, the flora indigenous to skin and mucosa has the selective advantage of being able to stick to substrates, resisting the abrasive forces of air and fluid currents that would otherwise wash it away. Furthermore, like the attachment of viruses to their target cells, microbial adherence is a significant, if not crucial, step in infectivity and in subsequent infectious disease. Indeed, the molecular principles previously established for specific viral attachment have been found appropriate for bacteria and fungi as well.

The macroscopic perspective

Although recognized as early as 1908 when G. Guyot observed attachment of bacteria to erythrocytes, medical interest in adherence stems from the research of Gibbons and colleagues on the microbial ecology of the oral cavity. Their analyses of the general phenomenon of adherence were soon verified with intestinal, vaginal, and nasal mucosal cells. Reflecting the relative proportions of flora observed in the mouth, Streptococcus salivarius adhered poorly to teeth, moderately to buccal cells, but in high numbers to epithelial cells of the tongue, while Strep. mitis attached well to teeth and cheek but only moderately so to tongue. In contrast, Enterococcus faecalis and Escherichia coli, named for their normal intestinal habitat, were infrequently detected on oral surfaces and had a correspondingly low adherence.

Apart from bacteriophage, which can readily be demonstrated in sebaceous material expressed from glands on the face, viruses cannot be considered a component of the normal flora of the skin. They are rarely, if ever, detected on the surface of the skin in the absence of clinically apparent lesions. Virus may be present in the deeper layers of the skin during the later incubation period after infection, or there may be latent infection with viruses such as papillomavirus. This is unlike the mucous membranes of the oropharynx, genitalia and gastrointestinal tract, where infective viruses may often be detected in the absence of clinical disease. For instance, herpes simplex virus (HSV) cannot be isolated from normal skin, although it may be found in the oropharynx of about 1 per cent of normal healthy adults; early studies suggesting its presence in uncomplicated eczemas have been contradicted.

The skin is a frequent site of manifestation of virus infection. Lesions may be localized or widespread as part of a systemic infection.

Microbiology

All viruses are obligate intracellular parasites and are characterized by containing only one type of nucleic acid, either DNA or RNA, which may be double or single stranded. These provide the main criteria for the classification of viruses. The virion comprises the nucleocapsid, which may be contained within an envelope derived from the cell membrane of the cell in which it replicated.

Normal fungal flora of the skin

As with bacteria, the fungal flora on the surface of the stratum corneum of otherwise healthy persons can be regarded as consisting of two types of populations – transient and resident commensals. The resident members are the Pityrosporum (Malassezia) or lipophilic yeasts, which are found on the skin surface of all adults and cluster around the openings of sebaceous glands. The existence of a transient fungal skin flora has not been as well documented as that of the bacteria, although temporary colonization of nails and hair by dermatophytes, and of other sites by Candida spp. is recognized. Transient colonization may also reflect contamination from another site where carriage rates are comparatively higher. For example, the isolation of yeasts from the perineum may well reflect their presence in the gastrointestinal tract and, in women, in the vagina.

Malassezia (Pityrosporum) yeasts

Members of the genus Pityrosporum are thick-walled yeast fungi that inhabit the superficial layers of the stratum corneum. They comprise three main species or varieties – Pityrosporum ovale, P. orbiculare and P. pachydermatis. The first two of these are found chiefly on human hosts and P. pachydermatis on a variety of other hosts. Pityrosporum ovale and P. orbiculare are associated with certain human diseases including pityriasis versicolor, where yeasts and short, stubby hyphae can be demonstrated in skin scrapings. These hyphae were originally named, without cultural confirmation, Malassezia furfur, but are now known to represent a parastic phase of pityrosporum.