HLA B37 determines an influenza A virus nucleoprotein epitope recognized by cytotoxic T lymphocytes

Loss of HLA-A,B,C allele products and lymphocyte function-associated antigen 3 in colorectal neoplasia

The expression of HLA-A,B,C antigens and lymphocyte function-associated antigen 3 in human colorectal adenomas and adenocarcinomas was studied by immunohistochemistry. None of 10 adenomas and only 1 of 30 carcinomas had lost expression of all HLA-A,B,C molecules. On the other hand, focal loss of an HLA-B product was seen in 2 of the adenomas, and complete losses of tumor cell HLA-A2 (in 7 of 13 cases), HLA-Bw4 (in 4 of 13 cases), and HLA-A3 (in 1 of 6 cases) were seen in the carcinomas. No complete losses of HLA-A1 (in 6 cases) or HLA-Bw6 (in 22 cases) occurred in the carcinomas. In addition, 1 of 20 adenocarcinomas totally lacked lymphocyte function-associated antigen 3. Because a loss of tumor cell HLA-A,B,C antigen or lymphocyte function-associated antigen 3 could be selected for through an advantage in escape from cytotoxic T-lymphocyte attack, our results suggest that immunoselection may be a more important mechanism in tumor progression than has previously been assumed.

Immunization against multiple viruses by using solid-matrix-antibody-antigen complexes

We have previously shown that mice immunized with solid-matrix-antibody-antigen (SMAA) complexes in the absence of adjuvants show vigorous humoral and cell-mediated immune responses to the immunizing antigen. Here we report that various proteins involved in inducing protective immune responses to different viruses can easily and simply be incorporated into SMAA complexes and that such complexes act as powerful multivalent immunogens. Construction of such SMAA complexes may be one of the most practical and effective ways of producing multivalent subunit vaccines for use in humans and animals.

Induction of ovalbumin-specific cytotoxic T cells by in vivo peptide immunization

CTL recognize peptide forms of processed, foreign antigens in association with class I molecules encoded by the MHC and are usually directed against endogenously synthesized "cellular antigens," such as those expressed by virus-infected cells. In vitro studies have shown that small exogenous peptides can directly associate with class I molecules on the cell surface and mimic the target complex derived by intracellular processing and presentation. We have recently generated OVA-specific, H-2Kb-restricted CTL by immunizing C57BL/6 mice with a syngeneic tumor line transfected with the OVA cDNA. The CTL recognize the OVA transfectant E.G7-OVA and the synthetic peptide OVA258-276, but fail to recognize the native protein. We reasoned that given the potential for direct peptide/class I association observed in vitro, OVA258-276 may induce CTL after in vivo priming. However, we found that this is not the case. OVA258-276 and peptides of increasing lengths up to OVA242-276 and OVA242-285, which are all able to form the target complex in vitro, are inefficient at priming E.G7-OVA-specific CTL responses after intravenous injection. This is also true for both native and denatured OVA. In contrast to these results the synthetic peptide OVA229-276 corresponding to a peptide in a partial tryptic digestion of OVA can efficiently prime C57BL/6 mice in vivo after intravenous injection. This peptide elicits CTL that appear identical to those derived from animals immunized with syngeneic cells producing OVA endogenously. These results are discussed in terms of separate class I and class II antigen presentation pathways and the ability of only certain, exogenous antigens to enter the cytoplasmic, class I pathway.

Specificity of proliferative response of human CD8 clones to mycobacterial antigens

Human CD8 T lymphocyte clones (TLC) were generated from the pleural effusion of patients with tuberculosis using a protocol that required, in addition to antigen, coculture of purified CD8+ T cells, accessory cells, interleukin 2 (IL2) and anti-CD3-Sepharose. The TLC obtained were stimulated by mycobacterial soluble extracts in an IL2-dependent and MHC class I-restricted manner. When antigen-responsive TLC were screened with extracts from the recombinant mycobacterial library they were found to respond to either the Y3125 (100-kDa) or the Y3111 (71-kDa) lambda gt11 clones. Polyacrylamide gel immunoblot analysis demonstrated that the CD8 TLC responded to fractions with the molecular mass range 27-45 kDa in the Y3125 lysogen and 60-90 kDa in the mycobacterial soluble extract. The specificity of TLC reactive with the Y3111 clone was confirmed using the 71-kDa antigen purified from the same lysogen. These TLC recognized sequences common to the 71-kDa protein derived from mycobacteria, E. coli or a human cell line. Studies of three TLC using antigen-presenting cells of known genetic haplotype indicated that stimulation with both the Y3125 and the 71-kDa antigens were restricted by determinants encoded by HLA-B8.

HIV-1 gag-specific cytotoxic T lymphocytes defined with recombinant vaccinia virus and synthetic peptides

Current candidate vaccines fail to protect primates against challenge with human immunodeficiency virus (HIV) in the presence of antibody responses; this underlines the importance of studying cell-mediated immunity to HIV and identifying specific epitopes that stimulate cytotoxic T lymphocytes (CTL). Using a recombinant vaccinia virus to express the gag protein of HIV-1 we found HLA class-I-restricted gag-specific CTL in thirteen out of fifteen healthy HIV seropositive patients. We then used short synthetic peptides in the lysis assay to screen for gag CTL epitopes. In one patient we have identified a peptide in p24 that is recognized by CTL in association with HLA-B27. This peptide, and further peptide sequences defined by these methods, could be incorporated in vaccines designed to induce cell-mediated immunity against HIV.

Effect of mutations and variations of HLA-A2 on recognition of a virus peptide epitope by cytotoxic T lymphocytes

Cytotoxic T lymphocytes (CTL) specific for influenza A virus were prepared from 15 donors. Those with HLA-A2 recognized autologous or HLA-A2-matched B-lymphoblastoid cells in the presence of synthetic peptide representing residues 55-73 or 56-68 of the virus matrix protein sequence. Influenza A virus-specific CTL from donors without HLA-A2 or with an HLA-A2 variant type failed to respond to this peptide. CTL lines specific for HLA-A2 plus peptide did not lyse peptide-treated target cells from HLA-A2 variant donors. They also failed to lyse peptide-treated cells with point mutations that had been inserted into HLA-A2 at positions 62-63, 66, 152, and 156 and, in some instances, mutations at positions 9 and 70. CTL lysed peptide-treated target cells with mutations in HLA-A2 at positions 43, 74, and 107. The results imply that this defined peptide epitope therefore interacts with HLA-A2 in the binding groove so that the long alpha-helices of HLA-A2 make important contact with the peptide at positions 66, 152, and 156. Different amino acids at position 9, which is in the floor of the peptide binding groove of HLA-A2 and the closely related position 70, modulate the peptide interaction so that some T-cell clones react and some do not.

T-immunogenic peptides are constituted of rare sequence patterns. Use in the identification of T epitopes in the human immunodeficiency virus gag protein

The sequences of a set of 63 peptides of demonstrated T immunogenicity have been analyzed and compared with two different randomly generated sets of sequences. This study indicates a statistically significant tendency of T immunogenic peptides to be constituted of clusters of rare tetrapeptides, as evaluated from the available sequence data banks. This result has been used to locate potential T epitopes in the human immunodeficiency virus (HIV) gag protein. Four peptides corresponding to the best candidate T epitopes (chosen in regions of conserved sequence among different virus isolates) have been synthesized and found to be recognized by a HIV-1-specific, HLA-A2-restricted human cytotoxic T cell line.

Selection of class I MHC-restricted peptides with the strip-of-helix hydrophobicity algorithm

A strip-of-helix hydrophobicity algorithm to predict class II MHC-restricted peptides, on the basis of their structural similarity to an amphipathic, alpha-helix in Ii, also predicted peptides which were presented to cytotoxic T-cells by class I MHC molecules. This algorithm ranked peptides according to mean Kyte-Doolittle hydrophobicity values of amino acids at positions n, n + 4, n + 7, n + 11, n + 14 and n + 18 in a sequence which when coiled as a putative alpha-helix, had the indicated residues in an axial strip along one side of the helix. Sequences selected for highly scoring, hydrophobic strips were required to have at least 1 of the 4 adjacent strips scoring more negatively than -1 in the strip-of-helix hydrophobicity index and the entire sequence could contain no prolines. This algorithm predicted the class I MHC-restricted, T-cell-presented peptides in sequences of 4 proteins from which some class I MHC-restricted, T-cell-presented sequences had been experimentally determined. Since both class I and class II MHC-restricted peptides could be identified with this algorithm, one can propose that: (1) foreign peptide-binding sites (desetopes) of the class I and class II MHC molecules are structurally similar; and (2) any one T-cell-presented peptide can be presented by some specific allele of both a class I and a class II MHC antigen.

An immunodominant epitope present in multiple class I MHC molecules and recognized by cytotoxic T lymphocytes

CTL derived from (C3H x B6.K1)F1 animals were sensitized against L cells that express the transfected gene product Q10d/Ld. These CTL were highly crossreactive against three other class I molecules, H-2Kbm1, H-2Ld, and H-2Kd. In an attempt to define this crossreactive epitope it was noted that between 25 and 39% of amino acids in the alpha helices and central beta strands of these three molecules vary from Q10d. These amino acids represent residues that have been proposed to potentially interact with a peptide antigen or TCR (21). However, all four molecules share the amino acid tyrosine at positions 155 and 156. Additionally, Q10d, H-2Kbm1, and H-2Ld share alanine at position 152, while H-2Kd has an aspartic acid. We showed that these residues were important in controlling this epitope by the finding that anti-Q10d CTL did not recognize H-2Kbm1 revertant molecules that had either the position 152 alanine changed back to the wild-type H-2Kb residue (glutamic acid) or position 155 and 156 tyrosines changed back to wild-type residues arginine and leucine. Further evidence that these molecules share a crossreactive epitope was noted by the failure of (C3H x H-2Kbm1)F1 animals to generate CTL that recognized H-2Ld or H-2Kd, and the inability of (C3H x BALB/c)F1 animals to generate CTL reactive against H-2Kbm1. CTL from these mice were still able to recognize Q10d/Ld indicating that other epitopes could be detected if natural tolerance prevented recognition of the crossreactive epitope. To further define the epitope, CTL clones were generated against Q10d/Ld and maintained on either H-2Kbm1 or BALB/c feeder cells. In addition to testing these clones on the target cells described above, mutant molecules derived from H-2Ld, which have amino acid substitutions in their alpha 1 domain, were analyzed. It was noted that some anti-Q10 clones that did not crossreact on H-2Ld did react against H-2Ld mutant antigens that had H-2Dd amino acid substitutions in the alpha 1 domain at positions 63, 65, 66, and 70. Other clones had differential reactivities on these H-2Ld mutants further substantiating that alpha 1 domain amino acids play a role in controlling the expression of the crossreactive epitope. Thus, four class I molecules with multiple amino acid differences in their alpha 1 and alpha 2 domains share a crossreactive epitope readily recognized by alloreactive CTL.(ABSTRACT TRUNCATED AT 250 WORDS)

Induction of cytotoxic T lymphocytes by primary in vitro stimulation with peptides

Antigen-specific cytotoxic T cells can be generated by primary in vitro stimulation of spleen cells from C57BL/6 mice with appropriate peptide fragments. This response can be elicited without prior in vivo immunization. Chicken OVA fragmented with either cyanogen bromide (CN OVA) or trypsin (T OVA) was used as a source of mixed peptides. A synthetic peptide, NP365-380, representing the sequence 365-380 from influenza virus A/PR/8 nucleoprotein, was also used, since this contains the main determinants recognized by CTL generated from H-2b mice infected with A/PR/8 virus. The primary in vitro cytotoxic T cell response was peptide specific, since targets were lysed only in the presence of appropriate peptide antigens. Native OVA could not elicit primary effectors in vitro nor could it sensitize targets for lysis by OVA digest-specific CTL. A synthetic peptide corresponding to residues 111-122 within the OVA sequence could sensitize targets for lysis by effectors induced against T OVA. Effectors generated by in vitro stimulation were CD8+, CD4-, and H-2Db-restricted for NP365-380 and T OVA recognition. CN OVA-specific effectors were also CD8+, CD4-, but surprisingly, were able to lyse a range of H-2-different targets in an antigen-specific manner. These effectors failed to lyse a tumor line that does not express class I MHC molecules. This broad MHC restriction pattern was also apparent at the clonal level. In all cases, the antipeptide CTL generated by primary in vitro stimulation were inefficient in lysing target cells expressing endogenous forms of antigens, such as influenza virus-infected cells or cells transfected with the OVA cDNA. However, cytotoxic T cell lines generated in vitro against the NP365-380 peptide did contain a minor population of virus-reactive cells that could be selectively expanded by stimulation with A/PR/8-infected spleen cells. These results are discussed in terms of class I-restricted T cell stimulation in the absence of antigen processing by high surface densities of peptide/MHC complexes.

Structural model of HLA-DR1 restricted T cell antigen recognition

Two human helper T cell determinants in influenza have been identified, one in the hemagglutinin and the other in the matrix protein (M1). Both were shown to be DR1 restricted by using transfected L cells to present antigen. Comparison of the sequences of the two peptides revealed a similar pattern that could account for their DR1 specificity if the peptides adopt a helical conformation. The model was supported by the demonstration that hybrid peptides, composed of the amino acids that interact with DR1 from one determinant and the residues that interact with the T cell receptor from the other, were recognized by each clone. The generality of the motif was confirmed by the finding that DR1 individuals respond to a ragweed peptide containing the defined pattern.

Recognition of influenza virus-infected B-cell lines by human influenza virus-specific CTL

The cytotoxic activity on influenza virus-infected Epstein-Barr virus (EBV)-transformed B-lymphoblastoid cell lines (LCL-Flu) and influenza virus-infected phytohemagglutinin lymphoblasts (PHA-Flu) was compared with the use of influenza-A virus-specific cytotoxic T lymphocytes (CTL), generated in short-term bulk cultures. Cold-target inhibition experiments showed that the lysis of PHA-Flu was completely blocked by both cold LCL-Flu and cold PHA-Flu whereas the lysis of LCL-Flu was completely inhibited by cold LCL-Flu, but only partially by cold PHA-Flu, indicating that structures can be recognized on LCL-Flu which are absent from PHA-Flu. Monoclonal antibody (McAb) directed against a monomorphic determinant of major histocompatibility complex (MHC) class I molecules inhibited the lysis of PHA-Flu more strongly than the lysis of LCL-Flu. Since LCL have a high expression of MHC class II molecules compared to PHA lymphoblasts, we examined whether class II-restricted CTL activity was responsible for the (anti)class I McAb-resistant lysis of LCL-Flu. Neither anti-CD4 McAb nor anti-class II McAb inhibited the lysis of LCL-Flu which argues against a contribution of MHC class II-restricted CTL. Depletion of CD16+ cells, containing the majority of the nonspecific cytotoxic cells, did not affect the lysis of LCL-Flu, indicating that the remaining lysis on LCL-Flu was also not due to a nonspecific component. We suggest that cell-type-dependent variations exist in the nature of the immunogenic determinants to which CTL respond.

Analyses of the cytotoxic T lymphocyte responses to glycoprotein and nucleoprotein components of lymphocytic choriomeningitis virus

The outcome of infection by lymphocytic choriomeningitis virus (LCMV) in the natural murine host is determined in large part by the cytotoxic T lymphocyte response (CTL) mounted by the host. In order to define the specificities of CTL induced by LCMV infection, we have cloned and expressed the full-length nucleoprotein (NP) gene and 75% of the glycoprotein (GP) gene of LCMV in vaccinia virus vectors and have used these recombinant viruses to sensitize syngeneic target cells to lysis by anti-LCMV CTL. We have studied the anti-LCMV CTL responses induced on three different mouse H2 (major histocompatibility complex) backgrounds. First, we find that the relative recognition of the two LCMV proteins differs markedly on different H2 haplotypes; both proteins are seen on the H2bb background, while only NP is recognized on two other haplotypes (H2dd and H2qq). Second, we show that on the H2bb background the anti-GP CTL response comprises a major component of the overall CTL response, in marked contrast to several other viruses, e.g., influenza virus, vesicular stomatitis virus, and respiratory syncytial virus where anti-GP responses, if present, comprise only a minor portion of the whole. Third, LCMV GP can be a major target antigen for CTL induced by a serotypically distinct strain of LCMV, again in contrast to the above virus systems in which CTL cross-reactivity among different serotypes is dependent largely on the recognition of "internal" proteins.

T cell mediated immunity in virus infection

Subpopulations of T cells have many different functions which are important for the regulation of immune responses, and as effector cells to rid the host of intracellular virus infections. In this report, we describe a few features of T cells: their viral recognition patterns, the genetics of antiviral T cell responses, and the variation in the function of T cell subpopulations in different virus infections. Our particular emphasis relates to influenza, which is one of the best analysed model systems to date.

On the role of the transmembrane anchor sequence of influenza hemagglutinin in target cell recognition by class I MHC-restricted, hemagglutinin-specific cytolytic T lymphocytes

We have examined the requirement for the transmembrane hydrophobic anchor sequence of the influenza hemagglutinin (HA) in the formation of the antigenic moiety on the surface of target cells recognized by class I MHC-restricted murine CTL. For this analysis we have used a line of CV-1 monkey epithelial cells that express the transfected murine H-2Kd gene product as target cells and have used recombinant SV40-based late replacement vectors to achieve expression of genes encoding wild-type and mutant forms of HA. We have found that the majority of Kd-restricted HA-specific CTL clones recognize target cells that express a secreted HA molecule that lacks the transmembrane and cytoplasmic domains of the parent glycoprotein. Several Kd-restricted CTL clones that recognize subtype-specific and crossreactive epitopes on HA fail to recognize the anchor-negative, secreted HA or chimeric HA molecules containing the transmembrane and cytoplasmic domains of unrelated glycoproteins. These CTL clones appear to be directed to antigenic epitopes located within the transmembrane domain of HA, as defined by their capacity to recognize target cells sensitized with a synthetic 23-amino-acid peptide corresponding to sequences within this domain. The implications of these results for class I MHC-restricted CTL recognition are discussed.

Class I MHC molecules rather than other mouse genes dictate influenza epitope recognition by cytotoxic T cells

Influenza nucleoprotein (NP) is an important target antigen for influenza A virus cross-reactive cytotoxic T cells (Tc). Here we examine the NP epitope recognized by cloned and polyclonal BALB/c Tc and the genetics of this recognition pattern. We can define NP residues 147-161 as the epitope seen in conjunction with Kd, the only H-2d class I responder allele for NP restriction. H-2d/H-2b F1 mice (C57BL X DBA/2) primed by influenza infection lyse only H-2d target cells treated with peptide 147-161 while H-2b targets are recognized only after treatment with NP residues 365-379 (previously found to be recognized by Db restricted Tc cells). Tc cell recognition of NP peptide 147-161 is entirely dictated by expression of Kd and not by other B10 or C3H background genes of congenic mice. Restriction of a unique NP sequence by each responder class I major histocompatibility complex (MHC) allele suggests that antigen and class I MHC interact for Tc recognition.

Use of synthetic peptides of influenza nucleoprotein to define epitopes recognized by class I-restricted cytotoxic T lymphocytes

The conserved epitopes of influenza nucleoprotein (NP) recognized by class I MHC-restricted CTL from CBA (H-2k) and C57BL/10 (H-2b) mice have been defined in vitro with synthetic peptides 50-63 and 365-379, respectively. Two Db-restricted clones were described that recognize different epitopes on peptide 365-379. Finally, the recognition of complete NP was shown to be approximately 200-fold less efficient than peptide in the cytotoxicity assay. These phenomena are closely related to results with class II-restricted T cells and they strengthen the hypothesis that influenza proteins are degraded in the infected cell before recognition by class I-restricted CTL.

Mapping of T cell epitopes using recombinant antigens and synthetic peptides

Two complementary approaches were used to determine the epitope specificity of clonal and polyclonal human T lymphocytes reactive with the 65-kd antigen of Mycobacterium leprae. A recombinant DNA sublibrary constructed from portions of the 65-kd gene was used to map T cell determinants within amino acid sequences 101-146 and 409-526. Independently, potential T cell epitopes within the protein were predicted based on an empirical analysis of specific patterns in the amino acid sequence. Of six peptides that were predicted and subsequently synthesised, two (112-132 and 437-459) were shown to contain human T cell epitopes. This corroborated and refined the results obtained using the recombinant DNA sublibrary. Both of these regions are identical in M. leprae and M. tuberculosis and are distinct from the known B cell epitopes of the 65-kd protein. This combination of recombinant DNA technology and peptide chemistry may prove valuable in analysis of the cellular immune response to infectious agents.

Enhancement of human T-lymphocyte growth by human transferrin in the presence of fetal bovine serum

All dividing cells require transferrin as a growth factor. During in vitro culture of human lymphocytes, transferrin is usually supplied in the form of serum, either synergic or xenogenic (usually fetal bovine serum (FBS)). In the present work the growth of certain human T-cell lines was examined; these lines were derived from the synovium of rheumatoid arthritis patients and maintained in 10% FBS and 1% synovial fluid. Their growth especially at limiting dilutions was found to be strongly dependent on the presence of synovial fluid at low concentration (0.05-0.1%) in culture medium containing 10% FBS. Further studies indicated that this effect of synovial fluid was duplicated by human serum or plasma, and was due to the presence of human transferrin. A significant effect on T-cell growth was observed using 2 micrograms/ml human transferrin with optimal growth at 10-20 micrograms/ml. This requirement for human transferrin was not a peculiarity of the synovium-derived T-cell lines, but was observed with all T-cell lines tested irrespective of phenotype or function. These observations suggest that bovine transferrin is inadequate for T-cell growth, and that the growth enhancing properties of FBS do not primarily reflect the provision of transferrin. Since some T cells have recently been shown to be capable of secreting transferrin upon activation, endogenous synthesis of transferrin may be an important factor in the in vitro growth of T cells so that such cells would be selected when FBS is the source of serum used to grow human T-cell lines or clones.

Cytotoxic T lymphocytes recognize a fragment of influenza virus matrix protein in association with HLA-A2

Both human and murine cytotoxic T cells (CTL) elicited in response to infection with influenza A viruses have been shown to be specific for internal viral proteins, such as the matrix and nucleoprotein. Individual CTL epitopes have been identified in the nucleoprotein by successfully substituting short synthetic peptides for the intact virus in the preparation of target cells in cytotoxicity assays. The defined peptide epitopes have each been recognized by CTL in association with individual class I major histocompatibility complex (MHC) proteins, H-2Db, H-2Kk, H-2Kd (Taylor, P. et al., unpublished data) and HLA-B37. A logical strategy to investigate the molecular details of the interaction between antigen and MHC class I proteins would be to define an epitope recognized by the MHC class I molecule HLA-A2. This is because the amino-acid sequence is known, several variants of A2 have been characterized and the protein has been purified and crystallized. Here we describe a peptide derived from the influenza matrix protein that is recognized by human CTL in association with the HLA-A2 molecule.

Identification of viral molecules recognized by influenza-specific human cytotoxic T lymphocytes

Human cytotoxic T cells specific for influenza A virus were tested for recognition of each of the ten influenza A virus proteins expressed in target cells using recombinant vaccinia viruses. They recognized the matrix M1, polymerase PB2, and nucleoproteins of influenza virus in association with MHC class I antigens. These internal viral proteins were seen by CTL in conjunction with one or more of the available dependent HLA gene products. There was no detectable recognition of influenza virus surface glycoproteins in target cells.