The amino acid sequences recognized by five I-E(d)-restricted and one E alpha A beta d-restricted murine T cell clones were determined. The clones had been raised to a synthetic peptide representing amino acids 305-328 of influenza virus hemagglutinin. It was found that although all of the T cell clones recognized a single 10-residue region of the peptide, 307KYVKQNTLKL316, different clones could recognize minimal ("core") determinants spanning 8, 9, or 10 of these amino acids. To see whether particular amino acids within the sequence 307-316 were universally important for T cell recognition, the six clones were assayed for their ability to tolerate single amino acid substitutions of the 10 residue peptide. In all, 190 analogues of the peptide in which each amino acid in the sequence was replaced, in turn, by each of the other 19 naturally occurring amino acids were tested. It was shown that 1) the six T cell clones had very different requirements for recognition of the peptide, 2) substitutions at every single position within the peptide could be shown to affect recognition in a T cell-specific manner, and 3) every single position within the peptide could be replaced by a large number of amino acids and still be recognized by at least one T cell clone. These results demonstrate the great diversity exhibited by the T cell repertoire in recognizing a 10-amino acid determinant, as well as the degeneracy of peptide binding to I-E(d).

The determinants recognized by helper T cells specific for the site E region of H3 subtype influenza virus hemagglutinin (HA) have been defined by examining the reactivity of T-cell clones with sets of overlapping peptides of various lengths covering the site. Two overlapping sequences, TLIDALLG and LIDALLGDP, were identified as the minimal determinants for four of five representative clones. These sequences are located within a loop of the molecule closed by a disulfide bond and presumably require cleavage of this bond for interaction with the class II major histocompatibility molecule. In contrast, the determinant recognized by the fifth clone was dependent on the presence of an intact disulfide bond for its expression and could not be represented by a synthetic peptide homolog of the linear sequence. Both TLIDALLG and LIDALLGDP are conserved within all field strains of the H3 subtype. Nevertheless, recognition of these sequences by the T-cell clones is affected by the glycosylation pattern of the hemagglutinin and by residues lying outside the minimal determinant. Three distinct clones directed towards the sequence LIDALLGDP were remarkably similar in their pattern of response to a set of synthetic analogs of the determinant, suggesting that residues of the T-cell receptor other than those contacting the minimal determinant may be responsible for the different specificities observed for these clones with different field strains of virus.

The determinant recognized by a class II restricted helper T cell clone raised against a peptide corresponding to the C-terminal 24 residues of the heavy chain of influenza virus hemagglutinin (HA) was examined in detail. The sequence 309VKQNTLKL316 was identified as the minimal determinant for T cell activation but its stimulatory capacity was augmented by extension at either end. Sets of peptide analogs, in which each residue within the minimal determinant was replaced in turn by every one of the other naturally occurring amino acids, revealed either an absolute requirement for the native residue or a very limited degree of replaceability, at seven of the eight positions. Only the N-terminal residue 309V could be replaced with almost any other amino acid without loss of reactivity; in fact, substitution at this position with residues containing bulky side groups enhanced the response. The reactivity of the clone with analogs of the longer peptide 307KYVKQNTLKL316, which induces maximal levels of stimulation, revealed a very different pattern of replaceability for certain residues; in particular, the requirement for a lysine at position 310 was no longer apparent. This study presents a complete analysis of the importance of each individual residue to the integrity of a T cell determinant and provides evidence that the critical requirement for a particular amino acid at a given location may be overridden by N-terminal extension of the minimal determinant. These findings indicate that, within different homologs of the native sequence, particular residues may assume quite different roles.

Anti-idiotypic (anti-Id) antibodies were raised against two murine monoclonal antibodies (mAb11and mAb21) which recognise two distinct and well-characterised epitopes on a 24-residue synthetic peptide representing part of the haemagglutinin (HA) of influenza virus. A monoclonal anti-Id antibody, specific for mAb21, could bind to mAb21 when the paratope of the latter was occupied with peptide, indicating that this anti-Id antibody is directed to a framework idiotope. In contrast, an anti-Id mAb derived from mAb11-immunised mice was inhibited in its binding to Id by the parent peptide and also by the heptapeptide NVPEKQT which constitutes the epitope recognised by mAb11. The small size of this synthetic peptide eliminates the possibility of significant steric inhibition in the system, and establishes that this mAb is a true paratope-directed anti-Id antibody. The interaction of this anti-Id mAb with the paratope of mAb11 in the presence of a set of peptide homologues of the epitope was also examined. A peptide as short as 5 residues, which contains two of the three irreplaceable residues of the epitope, could inhibit binding between the two mAbs.

The number of neutralizing monoclonal IgG molecules that can bind to a single trimeric molecule of influenza viral hemagglutinin (HA) was calculated by estimating the molecular weight of the immune complexes formed under conditions of antibody excess and was found to be dependent upon the antigenic site to which the MAb is directed. Whereas three antibody molecules directed to site A or site E are able to bind simultaneously to a single trimer of HA, generally only one molecule directed to site B (the "tip") or site B/D ("tip/interface") can be accommodated. Using mixtures of MAbs, more IgG molecules can be accommodated, but steric hindrance limits simultaneous binding of different MAbs directed to the same antigenic site or even to neighboring sites. At limiting antibody concentration, some MAbs can form much larger aggregates in which several HA molecules are crosslinked by antibody. However, the fact that certain MAbs do not crosslink HA molecules in this way indicates that MAbs directed to different epitopes within the same general antigenic site differ significantly in their geometry of binding.

Fifteen T-cell clones were derived from BALB/c or DBA/2 mice immunized with a synthetic peptide corresponding to the C-terminal 24 residues (residues 305 to 328) of the HA1 chain of H3 subtype influenza virus hemagglutinin. All of the clones proliferated when the peptide was presented in association with I-Ed. By using shorter homologs, it was shown that the T-cell response was focused predominantly on the region at the N-terminal end of the peptide encompassed by residues 306 to 319. Individual clones recognizing this region differed in their absolute requirements for residues at the extremities of the site and also in their patterns of efficiency of recognition of shorter homologs. One particular clone defined another site of T-cell recognition within residues 314 to 328. The response of the clones to peptide analogs identified certain residues within the sites that were critical for recognition, with the substitution Gln-311----Ser having a differential effect on clones responding to the N-terminal site. Only one of the clones responded well to influenza virus itself. This clone also required relatively low concentrations of the parent peptide for optimum stimulation and was suppressed by higher concentrations. The data demonstrate striking heterogeneity in the T-cell response even to a short synthetic peptide, with different T-cell clones recognizing slightly different but overlapping areas of the molecule.

Anti-idiotypic (anti-Id) antibodies were raised in rabbits against five monoclonal antibodies (MAbs) specific for different antigenic sites on the hemagglutinin (HA) of influenza virus Mem71H-BelN (H3N1) [A/Memphis/1/71 (H3N2) x A/Bel/42 (H1N1)]. Each of the anti-Id sera was directed predominantly towards a unique (private) idiotype of the immunizing MAb, none of the five idiotypes being detectable in pooled BALB/c antisera against Mem71H-BelN virus or on most other anti-HA MAbs tested. Partial idiotypic sharing was observed, however, between certain MAbs, from different mice, having the same or similar epitope specificity for HA. When used as immunogens in BALB/c mice, two of the five anti-Id preparations induced antibodies that reacted with Mem71H-BelN virus and displayed neutralizing activity. Mice of other inbred strains responded similarly, indicating that the response was not genetically restricted by the Igh locus. From their pattern of reactivity with mutants of Mem71H-BelN virus with known single amino acid substitutions in the HA molecule, the antiviral antibodies elicited by anti-Id antibodies were shown to be directed to the same antigenic site on A/Memphis/1/71 HA as the original immunizing MAb (site A or site E, respectively). However, several of these antisera were shown to contain additional distinct subpopulations of antibodies specific for heterologous influenza A virus strains, either of the H3 subtype or of a different HA subtype (H1 or H2). Since the induction of antibodies to HA of different subtypes is not a feature of the antibody response to influenza virus itself, their induction by anti-Id antibodies merits further investigation.

The immune response to a synthetic peptide, H3 HA1(305-328), representing the C'-terminal 24 amino acid residues of the HA1 chain of the hemagglutinin of the H3 subtype of influenza virus is controlled by genes in the I region of the major histocompatibility complex. Mice of the H-2d haplotype are high responders and produce antibody for several months after a single injection of peptide without carrier. Mice of the H-2b, H-2k, and H-2q haplotypes are low antibody responders. Investigation of recombinant and congenic mouse strains revealed that high responsiveness requires the genes that encode the I-Ed molecule. Immunoassays, involving direct binding to analogs of this peptide and inhibition by both these analogs and synthetic epitopes, were used to analyze the specificity of the polyclonal response. In BALB/c mice, the primary antibody response is directed principally against the antigenic site 314-LKLAT-318, whereas the secondary response after a boost is predominantly directed to a distinct site, 320-MRNVPEKQT-328. The T-cell response to the peptide H3 HA1(305-328), as measured by antigen-induced proliferation of primed T cells in vitro, is also I-Ed restricted in high-responder H-2d mice and is directed against an antigenic site that does not require the four C-terminal residues unique to the H3 influenza subtype. A different epitope appears to be recognized by T cells from CBA (H-2k) mice, which proliferate to a moderate extent on exposure to the peptide but, nevertheless, do not provide help for an antibody response.

Antibodies raised against the synthetic peptide corresponding to the carboxy-terminal 24 amino acids (305-328) of the heavy chain of the hemagglutinin molecule of influenza virus A/X-31 (H3) bind this peptide at three antigenic sites. These sites were identified by assaying binding of polyclonal BALB/c mouse antipeptide sera to the complete set of all possible di-, tri, tetra-, penta-, hexa-, hepta-, and octapeptides homologous with the 24-residue sequence. Individual epitopes were defined and essential residues identified by testing the binding of monoclonal antibodies to sets of peptide analogues in which every one of the homologous residues was replaced in turn by each of the 19 alternative genetically coded amino acids. The immunodominant epitope was shown to be a linear sequence of five amino acids, 314LKLAT318. Replacement of any one of these residues with any other amino acid resulted in loss of antibody binding, indicating that all five are essential to the interaction and that they are probably contact residues. Another antigenic site contains at least two overlapping epitopes: polyclonal sera recognize predominantly an epitope or epitopes encompassed by the linear sequence 320MRNVPEKQT328, whereas the epitope defined by a particular monoclonal antibody comprises the seven amino acids 322NVPEKQT328, of which N322, E325, and Q327 were implicated as contact residues.

Two monoclonal antibodies recognizing distinct epitopes the outer boundaries of which are separated by only three amino acid residues, a maximum of 10A, were demonstrated to bind simultaneously to a short synthetic peptide. The affinity of binding of the two monoclonal antibodies and of Fab' fragments derived from them was determined. The stoichiometry of the interaction was analysed by velocity sedimentation and by gel permeation chromatography experiments. The results indicate that the immune complexes formed are composed of two antibody molecules in association with one or two peptide molecules.

Synthetic peptides of increasing length and corresponding in sequence to the C-terminal end of the HA1 molecule of influenza virus were constructed and examined for their immunogenic and antigenic properties. Peptides containing at least the four C-terminal amino acids, when coupled to keyhole limpet hemocyanin, were capable of eliciting antibody in BALB/c mice that bound to the 24-residue parent peptide H3 HA1 (305 to 328). In the absence of a carrier, the C-terminal decapeptide was the shortest peptide capable of eliciting antibody. The specificity of this antibody was indistinguishable from that of a monoclonal antibody to the parent peptide which recognizes an epitope encompassed by the C-terminal seven residues. All peptides containing at least the C-terminal four residues were able to inhibit completely the binding of this monoclonal antibody to the parent peptide. Taken together, these results indicate that (i) the tetrapeptide is capable of eliciting specific antibody when coupled to a carrier, (ii) this tetrapeptide possesses all of the antigenic information necessary to occupy the paratope of a monoclonal antibody elicited by the longer parent peptide, and (iii) the decapeptide contains all of the information necessary to elicit a specific immune response and therefore carries an epitope recognized by T cells as well as one recognized by B cells.

The mitogenic activity of influenza virus is a function of the hemagglutinin (HA) molecule. Purified HA is mitogenic for murine B lymphocytes but not T lymphocytes. Furthermore, like the intact virus, HA of the H2 (but not H3) subtype is mitogenic only for B cells expressing the class II major histocompatibility complex glycoprotein I-E. Since virus bearing uncleaved HA is as mitogenic as virus bearing cleaved HA, the membrane fusion activity of the HA molecule is not involved.

The specificity and function of two T-cell clones derived from A/Memphis/1/71 (H3) influenza virus (Mem 71)-immune BALB/c spleen cells have been compared. One clone, X-31 clone 1, was subtype specific, proliferating in response to influenza strains of the H3 subtype only. The other, Jap clone 3, cross-reacted in proliferation assays with heterologous subtypes of influenza A, but not type B. Both clones recognized the HA1 chain of the hemagglutinin (HA) molecule and their proliferation in response to detergent-disrupted virus could be specifically inhibited by monoclonal antibodies to the HA. The T-cell clones were of the L3T4+ phenotype. Both recognized antigen in association with I-Ed, as indicated by studies with H-2 recombinant strains of mice and by blocking with monoclonal anti-I-E antibody. In vivo, both clones elicited a delayed-type hypersensitivity (DTH) reaction when inoculated into mouse footpads together with virus, X-31 clone 1 again displaying subtype specificity and Jap clone 3 being cross-reactive. The clones were also able to provide factor-mediated help in vitro to virus-primed B cells in an anti-HA antibody response. The cross-reactive T-cell clone provided help not only for B cells primed with influenza A subtype H3 and responding to H3 virus in culture, but also for H2 virus-primed B cells making anti-H2 antibody.

The relationship between the mitogenic activity of influenza type A viruses for murine B lymphocytes and the receptor-binding specificity of their hemagglutinin was examined. Receptor-binding specificity was determined by the ability of the virus to agglutinate erythrocytes that had been sialidase treated and then enzymatically resialylated to contain sialyloligosaccharides with defined sequences. Distinct differences in receptor-binding specificity were observed between strongly and weakly mitogenic viruses of the H3 subtype, with strong mitogenic activity correlating with the ability of the virus to recognize the sequence N-glycolylneuraminic acid alpha 2,6 galactose (NeuGc alpha 2,6Gal). Viruses isolated early in the evolution of the H3 subtype (from 1968 to 1971) are relatively weak mitogens and recognize the sequence N-acetylneuraminic acid alpha 2,6 galactose (NeuAc alpha 2,6Gal) but not NeuGc alpha 2,6Gal. H3 viruses isolated since 1972 are strongly mitogenic, and these viruses recognize both NeuGc alpha 2,6Gal and NeuAc alpha 2,6Gal. The amino acid substitution of Tyr for Thr at residue 155 of HA1 may be critical to this change in receptor-binding specificity and mitogenic activity of the later H3 viruses. Horse serum-resistant variants of H3 viruses, which bind preferentially to the sequence NeuAc alpha 2,3Gal, are poorly mitogenic. Differences were also observed between the receptor-binding specificity of the strongly mitogenic H3 viruses and viruses of the H2 and H6 subtypes, the mitogenic activity of which is limited to strains of mice that express the class II major histocompatibility complex glycoprotein I-E. The results indicate that the receptor-binding specificity of the hemagglutinin plays a critical role in determining the mitogenic activity of influenza viruses.

A synthetic peptide comprising the C-terminal 24 amino acids of the heavy chain (HA1) of influenza virus hemagglutinin was constructed and examined for antigenic and immunogenic activity. Monoclonal antibodies as well as polyclonal antisera raised against the synthetic peptide were able to bind to intact virus. This binding was greatly enhanced if the virus was first subjected to pH 5, suggesting that this treatment exposes the C-terminus of HA1. Using synthetic analogs of the native sequence it was shown that the epitope recognized by one of the monoclonal antibodies encompasses one or more of the C-terminal four amino acids of HA1 (residues 325–328), which are conserved within subtypes but differ between subtypes, while the other monoclonal antibody recognizes a different epitope which involves at least one of the five variable residues at positions 311–315.

Influenza A viruses behave as T cell-independent B cell mitogens in vitro. The magnitude of the proliferation induced varies with the haemagglutinin subtype of the virus, the order being H2 greater than H6 greater than H3 greater than H1 for Balb/c mice. H3 viruses are equally mitogenic for all strains of mice tested. In contrast, the mitogenic response to H2 and H6 viruses is controlled by the I-E subregion of the major histocompatibility complex. These viruses are mitogenic only for strains of mice that express surface I-E antigen (haplotypes a, d, k, p, r), and not for haplotypes b, f, q, s, which fail to synthesize a normal E alpha chain and do not express surface I-E antigen. Mitogenesis by H2 and H6 viruses may involve direct interaction of virus with I-E molecules on the B lymphocyte or an accessory cell, whereas mitogenesis by H3 viruses does not appear to involve I-E.

Exposure of influenza virus haemagglutinin to pH 5 results in conformational changes occurring in the molecule which are accompanied by antigenic modifications. Furthermore, isolated haemagglutinin (HA) at a concentration of 0.1 nM undergoes dissociation from the trimeric to a monomeric form when exposed to pH 5. Whether present on intact virus or as the isolated monomer, each form of haemagglutinin from pH 5 exhibits similar alterations in antigenic characteristics. These forms of HA show modifications in the antigenic sites located in the hinge (site C), tip (site B) and subunit interface (site D) regions. Whereas binding of monoclonal antibodies recognizing the tip and interface is abrogated or diminished, binding of antibodies to the hinge region is greatly enhanced following exposure of virus or the monomeric form of HA to pH 5.

The recognition of influenza virus hemagglutinin (HA) by T lymphocytes was examined by assaying the T cell proliferative response of influenza virus-primed T cells to purified HA of different influenza A subtypes or to isolated heavy (HA1) or light (HA2) polypeptide chains of the HA molecule. The proliferative response to HA was dependent on the activation of an Ly-1+2- subset of T cells and required the presence of nylon wool-adherent, radiation-resistant accessory cells. T cells from mice primed by infection with one strain of type A influenza virus cross-reacted with other purified HA not only of the same subtype as the priming virus but also of serologically distinct subtypes of influenza A (but not B) virus. The response of virus-primed T cells to the homologous HA or to HA of the same subtype was shown to involve recognition of determinants on both the HA1 and the HA2 chains. The recognition of HA of different subtype by cross-reactive T cells appeared to be directed predominantly to determinants on HA2. Because the antibody response to influenza virus HA is not cross-reactive between subtypes and is directed predominantly to determinants on HA1, the present results indicate that at least some of the determinants on HA recognized by T cells are different from those recognized by B cells and that the HA2 chain may be involved primarily in stimulation of T cell rather than B cell immunity.

A number of peptides of the hemagglutinin (HA) of X-31 influenza virus have been synthesised. The amino acid sequences of some of these peptides represent regions of HA which have been postulated [Wiley et al., Nature, Lond. 289, 373–378 (1981)] to form the antigenic sites of this molecule. Animals were immunized with free peptide or peptide conjugated to a carrier and the resulting antisera examined for their capacities to bind to homologous peptide, whole HA, reduced and alkylated HA, and intact virus. Not all peptides examined in this way were immunogenic. Only antibodies raised against the C-terminus of HA1 peptide displayed binding to virus. This antiserum bound to the intact HA but not to the reduced and alkylated form of the molecule. These results raise questions as to the feasibility of using synthetic peptides of the influenza HA in short linear sequences to elicit neutralising antibody.

UV-inactivated influenza virus A strains of subtypes H1, H2, H3, and H6 were shown to be mitogenic for unprimed splenic lymphocytes from BALB/c mice. Representative viruses of these four subtypes all behaved as T cell-independent B cell mitogens. The magnitude of the proliferative response was determined by the subtype of the hemagglutinin molecule: H2 and H6 viruses were the most potent mitogens, and H3 viruses were moderately mitogenic, whereas H1 viruses induced only low, but significant, levels of proliferation. Mitogenesis was inhibited by antiviral sera and by monoclonal antibodies directed against hemagglutinin.