Superantigen implicated in dependence of HIV-1 replication in T cells on TCR V beta expression.

In the pathogenesis of AIDS it is not yet understood whether the small fraction of CD4+ T cells (approximately 1%) infected with the human immunodeficiency virus (HIV) are randomly targeted or not. Here we present evidence that human CD4 T-cell lines expressing selected T-cell antigen receptor V beta gene products can all be infected in vitro with HIV-1, but give markedly different titres of HIV-1 virion production. For example, V beta 12 T-cell lines from several unrelated donors reproducibly yielded up to 100-fold more gag gene product (p24gag antigen) than V beta 6.7a lines. This is consistent with a superantigen effect, because the V beta selectivity was observed with several divergent HIV-1 isolates, was dependent on antigen-presenting cells and on major histocompatibility complex (MHC) class II but was not MHC class II-restricted. The in vivo significance of these findings is supported by the preferential stimulation of V beta 12+ T cells by freshly obtained irradiated antigen-presenting cells from some HIV-1-seropositive but not HIV-1-negative donors. Moreover, cells from patients positive for viral antigen (gp120) were enriched in the V beta 12 subpopulation. V beta 12+ T cells were not deleted in AIDS patients, however, raising the possibility that a variety of mechanisms contribute to T-cell depletion. Our results indicate that a superantigen targets a subpopulation of CD4+ cells for viral replication.     

Homology of human T-cell leukaemia virus envelope gene with class I HLA gene

Human T-cell leukaemia virus (HTLV), first isolated in the United States from a patient with cutaneous T-cell lymphoma, is a unique horizontally transmitted retrovirus which is highly associated with certain adult T-cell malignancies. Also, HTLV can be transmitted in vitro to cord blood T-lymphocytes. In the accompanying paper it was shown that all T cells producing HTLV, whether cultured from infected persons or infected in vitro, bind a monoclonal antibody (4D12) which recognizes an epitope shared by certain cross-reactive class I major histocompatibility antigens. This antigen may account for the extra HLA-A and -B specificities detected in HTLV-infected cells using alloantisera. Because of the unusual findings of apparently inappropriate HLA antigens in HTLV infected cells, we had previously looked for rearrangement of class I-related genes in HTLV infected cells but failed to find any. Here, using molecular clones of HTLV and human major histocompatibility antigen DNA, we have shown homology between the envelope gene region of HTLV and the region of an HLA-B locus gene which codes for the extracellular portion of a class I histocompatibility antigen.     

Functional interaction between human T-cell protein CD4 and the major histocompatibility complex HLA-DR antigen

Mature T cells segregate phenotypically into one of two classes: those that express the surface glycoprotein CD4, and those that express the glycoprotein CD8. The CD4 molecule is expressed primarily on helper T cells whereas CD8 is found on cytotoxic and suppressor cells. A more stringent association exists, however, between these T-cell subsets and the major histocompatibility complex (MHC) gene products recognized by their T-cell receptors (TCRs). CD8+ lymphocytes interact with targets expressing class I MHC gene products, whereas CD4+ cells interact with class II MHC-bearing targets. To explain this association, it has been proposed that these 'accessory' molecules bind to monomorphic regions of the MHC proteins on the target cell, CD4 to class II and CD8 to class I products. This binding could hold the T cell and its target together, thus improving the probability of the formation of the trimolecular antigen: MHC: TCR complex. Because the TCR on CD4+ cells binds antigen in association with class II MHC, it has been difficult to design experiments to detect the association of CD4 with a class II molecule. To address this issue, we devised a xenogeneic system in which human CD4 complementary DNA was transfected into the murine CD4-, CD8- T-cell hybridoma 3DT-52.5.8, the TCR of which recognizes the murine class I molecule H-2Dd. The murine H-2Dd-bearing target cell line, P815, was cotransfected with human class II HLA-DR alpha, beta and invariant chain cDNAs. Co-culture of the parental T-cell and P815 lines, or of one parental and one transfected line resulted in a low baseline response. In contrast, a substantial increase in response was observed when CD4+ 3DT-52.5.8 cells were co-cultured with HLA-DR+ P815 cells. This result strongly indicates that CD4:HLA-DR binding occurs in this system and that this interaction augments T-cell activation.     

Localization of type 1 diabetes susceptibility to the MHC class I genes HLA-B and HLA-A

The major histocompatibility complex (MHC) on chromosome 6 is associated with susceptibility to more common diseases than any other region of the human genome, including almost all disorders classified as autoimmune. In type 1 diabetes the major genetic susceptibility determinants have been mapped to the MHC class II genes HLA-DQB1 and HLA-DRB1 (refs 1-3), but these genes cannot completely explain the association between type 1 diabetes and the MHC region. Owing to the region's extreme gene density, the multiplicity of disease-associated alleles, strong associations between alleles, limited genotyping capability, and inadequate statistical approaches and sample sizes, which, and how many, loci within the MHC determine susceptibility remains unclear. Here, in several large type 1 diabetes data sets, we analyse a combined total of 1,729 polymorphisms, and apply statistical methods-recursive partitioning and regression-to pinpoint disease susceptibility to the MHC class I genes HLA-B and HLA-A (risk ratios >1.5; P(combined) = 2.01 x 10(-19) and 2.35 x 10(-13), respectively) in addition to the established associations of the MHC class II genes. Other loci with smaller and/or rarer effects might also be involved, but to find these, future searches must take into account both the HLA class II and class I genes and use even larger samples. Taken together with previous studies, we conclude that MHC-class-I-mediated events, principally involving HLA-B*39, contribute to the aetiology of type 1 diabetes.     

A binding site for the T-cell co-receptor CD8 on the alpha 3 domain of HLA-A2

Adhesion measurements between CD8 and 48 point mutants of HLA-A2.1 show that the CD8 alpha-chain binds to the alpha 3 domain of HLA-A2.1. Three clusters of alpha 3 residues contribute to the binding, with an exposed, negatively charged loop (residues 223-229) playing a dominant role. CD8 binding correlates with cytotoxic T-cell recognition and sensitivity to inhibition by anti-CD8 antibodies. Impaired alloreactive T-cell recognition of an HLA-A2.1 mutant with reduced affinity for CD8 is not restored by functional CD8 binding sites on an antigenically irrelevant class I molecule. Therefore, complexes of CD8 and the T-cell receptor bound to the same class I major histocompatibility complex molecule seem to be necessary for T-cell activation.     

Polymorphism in the alpha 3 domain of HLA-A molecules affects binding to CD8

Cytotoxic T lymphocytes (CTL) expressing the CD8 glycoprotein recognize peptide antigens presented by class I major histocompatibility complex (MHC) molecules. This correlation and the absence of CD8 polymorphism led to the hypothesis that CD8 binds to a conserved site of class I MHC molecules. Using a cell-cell binding assay we previously demonstrated specific interaction between human class I MHC (HLA-A,B,C) molecules and CD8. Subsequent analysis of the products of 17 HLA-A,B alleles revealed a natural polymorphism for CD8 binding in the human population. Two molecules, HLA-Aw68.1 and HLA-Aw68.2, which do not bind CD8, have a valine residue at position 245 whereas all other HLA-A,B,C molecules have alanine. Site-directed mutagenesis shows that this single substitution in the alpha 3 domain is responsible for the CD8 binding phenotype and also affects recognition by alloreactive and influenza-specific CTL. Our results indicate that CD8 binds to the alpha 3 domain of class I MHC molecules.     

Direct binding of influenza peptides to class I HLA molecules

Activation of T lymphocytes requires the intracellular fragmentation of foreign antigens and their presentation by class I or class II major histocompatibility complex (MHC) glycoproteins. The direct binding of peptides to class II molecules has been demonstrated using equilibrium dialysis, gel filtration and fluorescence energy transfer at planar membranes, and its specificity compared to that of T-cell activation. In contrast, direct binding of peptides to class I molecules has been difficult to detect; although peptide sensitization experiments and the crystallographic structure of HLA-A2 (ref. 9) persuasively argue for its occurrence and importance. Here we describe a gel filtration assay from which we derive direct evidence for selective binding of an influenza matrix peptide to HLA-A2 and for binding of an influenza nucleoprotein peptide to HLA-B37. These two peptides have previously been shown to act respectively as targets for certain HLA-A2 or HLA-B37 restricted influenza-specific cytotoxic T lymphocytes (CTL). In addition we demonstrate binding to some, but not all, HLA allospecificities that cannot present these peptides to CTL. We estimate that less than 0.3% of the HLA molecules present in any given purified preparation were able to bind the added peptides.     

Skewed maturation of memory HIV-specific CD8 T lymphocytes

Understanding the lineage differentiation of memory T cells is a central question in immunology. We investigated this issue by analysing the expression of the chemokine receptor CCR7, which defines distinct subsets of naive and memory T lymphocytes with different homing and effector capacities and antiviral immune responses to HIV and cytomegalovirus. Ex vivo analysis of the expression of CD45RA and CCR7 antigens, together with in vitro analysis of the cell-division capacity of different memory CD8+ T-cell populations, identified four subsets of HIV- and CMV-specific CD8+ T lymphocytes, and indicated the following lineage differentiation pattern: CD45RA+ CCR7+ --> CD45RA- CCR7+ --> CD45RA- CCR7- --> CD45RA+ CCR7-. Here we demonstrate through analysis of cell division (predominantly restricted to the CCR7+ CD8+ T-cell subsets) that the differentiation of antigen-specific CD8+ T cells is a two-step process characterized initially by a phase of proliferation largely restricted to the CCR7+ CD8+ cell subsets, followed by a phase of functional maturation encompassing the CCR7- CD8+ cell subsets. The distribution of these populations in HIV- and CMV-specific CD8+ T cells showed that the HIV-specific cell pool was predominantly (70%) composed of pre-terminally differentiated CD45RA- CCR7- cells, whereas the CMV-specific cell pool consisted mainly (50%) of the terminally differentiated CD45RA+ CCR7- cells. These results demonstrate a skewed maturation of HIV-specific memory CD8+ T cells during HIV infection.     

Peptide binding to empty HLA-B27 molecules of viable human cells

Intracellular binding of antigenic peptides by polymorphic class I major histocompatibility complex molecules creates the ligands recognized by receptors of CD8+ T cells. Previously described in vitro assays of peptide binding to class I molecules have been limited by either the low proportion of accessible binding sites or the lack of allelic specificity. Here we describe a system in which the human class I molecule HLA-B27 binds considerable amounts of an influenza peptide with precise allelic discrimination. Binding requires viable cells, is stimulated by gamma-interferon and is inhibited by brefeldin A. Our results are consistent with the presence of fairly stable 'empty' HLA-B27 molecules at the cell surface. By contrast, analysis of the binding of a second influenza peptide indicates that empty HLA-Aw68 molecules are relatively short-lived. We speculate that HLA-B27 might bind extracellular peptides in vivo and that this property could underlie its association with autoimmune disease.     

Strong xenogeneic HLA response in transgenic mice after introducing an alpha 3 domain into HLA B27.

The pronounced response by mouse T cells to the major histocompatibility complex (MHC) class I antigens of the same species is characterized by a relatively large fraction of responding cells. Responses to MHC class I allelles of other species are, however, generally much weaker. T lymphocytes are positively selected on thymic MHC antigens, resulting in a T-cell repertoire with strong alloreactivity. This has been explained in terms of a mouse T-cell repertoire that is not efficiently selected for recognition of HLA molecules owing to the absence of HLA in mice. Here we show that mice transgenic for HLA mount a T-cell response against allogeneic HLA that is no better than in normal mice. We decided instead to test whether the mouse accessory molecule Lyt-2 on cytotoxic T lymphocytes could interact efficiently with the alpha 3 domain of HLA. To do this, we replaced the alpha 3 domain of HLA-B27 by a murine alpha 3 domain in a gene construct used to produce transgenic mice, and then used the spleen cells from these mice to stimulate normal mouse T cells. Under these conditions cytotoxic T lymphocytes were generated with the same frequency against xenogeneic HLA-B27 determinants as against allogeneic mouse class I antigens. These findings indicate that the normally weak xeno-MHC response is due to the inefficient interaction of the murine Lyt-2 accessory molecule with HLA class I, and not to limitations of the mouse T-cell repertoire.     

Cell lines producing human T-cell lymphoma virus show altered HLA expression

Human T-cell leukaemia/lymphoma virus (HTLV) can be identified in fresh and cultured T-lymphocytes from patients with adult T-cell malignancies. HLA typing of the peripheral blood lymphocytes and cultured cell lines from the patient from which the virus was originally isolated suggested the expression of additional HLA-A and -B locus antigens on the HTLV positive cultured T-cells that were not present on the EBV transformed B-cell line or on the peripheral blood lymphocytes. Peripheral blood lymphocytes (PBLs) and T-cell lines established from patients and cord blood lymphocytes, infected with virus by co-culture with T-cell lines, were typed for HLA antigens with alloantisera and in addition tested for reactivity with a monoclonal antibody (4D12) which recognizes a polymorphic HLA class-I antigen. In all HTLV positive cells, with demonstrable provirus replication, altered HLA alloantigen expression was observed. This may be explained by the observations reported in the accompanying paper which shows homology between the envelope gene region of HTLV and the region of an HLA-B locus gene which codes for the extracellular portion of a class I histocompatibility antigen.     

MHC class II interaction with CD4 mediated by a region analogous to the MHC class I binding site for CD8.

Interactions between major histocompatibility complex (MHC) molecules and the CD4 or CD8 coreceptors have a major role in intrathymic T-cell selection. On mature T cells, each of these two glycoproteins is associated with a class-specific bias in MHC molecule recognition by the T-cell receptor. CD4+ T cells respond to antigen in association with MHC class II molecules and CD8+ T cells respond to antigen in association with MHC class I molecules. Physical interaction between the CD4/MHC class II molecules and CD8/MHC class I molecules has been demonstrated by cell adhesion assay, and a binding site for CD8 on class I has been identified. Here we demonstrate that a region of the MHC class II beta-chain beta 2 domain, structurally analogous to the CD8-binding loop in the MHC class I alpha 3 domain, is critical for function with both mouse and human CD4.     

PD-1 expression on HIV-specific T cells is associated with T-cell exhaustion and disease progression

Functional impairment of T cells is characteristic of many chronic mouse and human viral infections. The inhibitory receptor programmed death 1 (PD-1; also known as PDCD1), a negative regulator of activated T cells, is markedly upregulated on the surface of exhausted virus-specific CD8 T cells in mice. Blockade of this pathway using antibodies against the PD ligand 1 (PD-L1, also known as CD274) restores CD8 T-cell function and reduces viral load. To investigate the role of PD-1 in a chronic human viral infection, we examined PD-1 expression on human immunodeficiency virus (HIV)-specific CD8 T cells in 71 clade-C-infected people who were naive to anti-HIV treatments, using ten major histocompatibility complex (MHC) class I tetramers specific for frequently targeted epitopes. Here we report that PD-1 is significantly upregulated on these cells, and expression correlates with impaired HIV-specific CD8 T-cell function as well as predictors of disease progression: positively with plasma viral load and inversely with CD4 T-cell count. PD-1 expression on CD4 T cells likewise showed a positive correlation with viral load and an inverse correlation with CD4 T-cell count, and blockade of the pathway augmented HIV-specific CD4 and CD8 T-cell function. These data indicate that the immunoregulatory PD-1/PD-L1 pathway is operative during a persistent viral infection in humans, and define a reversible defect in HIV-specific T-cell function. Moreover, this pathway of reversible T-cell impairment provides a potential target for enhancing the function of exhausted T cells in chronic HIV infection.     

HLA-A and B polymorphisms predate the divergence of humans and chimpanzees

Major histocompatibility complex (MHC) glycoproteins bind processed fragments of proteins and present them to the receptors of T lymphocytes. The extraordinary polymorphism of class I MHC molecules in man (HLA-A, B and C) and mouse (H-2 K, D and L) poses many questions concerning their diversification and evolution. Comparison of allelic sequences within a species suggests diversity is generated by the assortment of point mutations into varied combinations by mechanisms of recombination and gene conversion. We have now compared class I MHC alleles in two closely related species: humans (Homo sapiens) and chimpanzees (Pan troglodytes). Chimpanzee homologues of HLA-A, HLA-B and a non-classical gene have been identified. No features distinguishing human and chimpanzee alleles could be found. Individual HLA-A or B alleles are more closely related to individual chimpanzee alleles than to other HLA-A or B alleles. These results show that a considerable proportion of contemporary HLA-A and B polymorphism existed before divergence of the chimpanzee and human lines. The stability of the polymorphism indicates that hyper-mutational mechanisms are not necessary to account for HLA-A, B and C diversity.     

Serine phosphorylation-independent downregulation of cell-surface CD4 by nef

A decline in the T-cell population usually marks the onset of progressive immunological disease in individuals infected with the human immunodeficiency virus (HIV). Because CD4+ cells help to coordinate efficient immune responses, some of the defects in the immune function in advanced cases of AIDS may be explained by the disappearance of these cells. Therefore, an understanding of the mechanisms used by HIV to induce the reduction of CD4+ cells is important. Here we use a Moloney murine leukaemia virus-based retroviral vector in order to express the nef gene of HIV-1 in three lymphocytic cell lines expressing CD4. In all cases we find that cell-surface CD4 expression is inversely related to nef expression. However, nef does not alter steady-state levels of CD4 RNA or CD4 protein. Also, nef can downregulate a CD4 triple mutant (Ser----Ala) that is neither phosphorylated nor down-regulated by phorbol esters, indicating that nef is acting by a different mechanism.     

No T-cell tyrosine protein kinase signalling or calcium mobilization after CD4 association with HIV-1 or HIV-1 gp120.

The T lymphocyte surface protein CD4 is an integral membrane glycoprotein noncovalently associated with the tyrosine protein kinase p56lck. In normal T cells, surface association of CD4 molecules with other CD4 molecules or other T-cell surface proteins, such as the T-cell antigen receptor, stimulates the activity of the p56lck tyrosine kinase, resulting in the phosphorylation of various cellular proteins at tyrosine residues. Thus, the signal transduction in T cells generated through the surface engagement of CD4 is similar to that observed for the class of growth factor receptors possessing endogenous tyrosine kinase activity. As CD4 is also the cellular receptor for the human immunodeficiency virus (HIV), binding of the virus or gp120 (the virus surface protein responsible for specific CD4+ T-cell association) could mimic the types of immunological interactions that have previously been found to stimulate p56lck and trigger T-cell activation pathways. We have evaluated this possibility and report here that binding of HIV-1 or the virus glycoprotein gp120 to CD4+ human T cells fails to elicit detectable p56lck-dependent tyrosine kinase activation and signalling, alterations in the composition of cellular phosphotyrosine-containing proteins, or changes in intracellular Ca2+ concentration.     

Positive selection of antigen-specific T cells in thymus by restricting MHC molecules

Thymus-derived lymphocytes (T cells) recognize antigen in the context of class I or class II molecules encoded by the major histocompatibility complex (MHC) by virtue of the heterodimeric alpha beta T-cell receptor (TCR). CD4 and CD8 molecules expressed on the surface of T cells bind to nonpolymorphic portions of class II and class I MHC molecules and assist the TCR in binding and possibly in signalling. The analysis of T-cell development in TCR transgenic mice has shown that the CD4/CD8 phenotype of T cells is determined by the interaction of the alpha beta TCR expressed on immature CD4+8+ thymocytes with polymorphic domains of thymic MHC molecules in the absence of nominal antigen. Here we provide direct evidence that positive selection of antigen-specific, class I MHC-restricted CD4-8+ T cells in the thymus requires the specific interaction of the alpha beta TCR with the restricting class I MHC molecule.     

Phenotypic heterogeneity of cerebrospinal fluid-derived HIV-specific and HLA-restricted cytotoxic T-cell clones

A variety of clinical syndromes, including AIDS and neurological disorders, may follow as a consequence of infection with the human immunodeficiency virus type 1 (HIV-1). It is not yet clear, however, to what extent the destruction of lymphocytes and neural cells associated with these conditions is caused by adverse immune responses to HIV-1 or how much is due to cytopathic effects of the virus itself. Here we document the existence of HLA-restricted, HIV-1-specific cytotoxic T lymphocytes in the cerebrospinal fluid of two AIDS patients manifesting neurologic disorders. These cytotoxic T lymphocytes showed dual specificity, recognizing target cells coated with purified HIV-1 envelope glycoprotein (gp 120) or inactivated HIV-1 in the context of HLA antigens. Cytotoxic T-cell clones derived from one of the AIDS patients revealed restriction specificities representing both HLA class I and HLA class II antigens. Considerable phenotypic heterogeneity was observed amongst these clones, some expressing conventional combinations of cytotoxic T-cell surface markers, and others displaying unusual phenotypes. The presence of HIV-specific cytotoxic T lymphocytes in AIDS patients, and in particular in their cerebrospinal fluid, suggests that these cytotoxic effectors may participate in the lymphoid cell and/or neurologic damage observed in such patients.