Increased immunoglobulin G, but not M, binding to endogenous retroviral antigens in HIV-1 infected persons.

The modes of interaction between products of human endogenous retroviral (HERV) sequences and the immune system are largely unknown. In HIV infected persons, an exogenous retrovirus adds further complexity to the situation. Therefore, 14 synthetic peptides with sequences derived from conserved regions of various endogenous retroviruses (ERVs) and from related exogenous retroviruses were used to search for IgG and IgM antibodies that bind to such antigens in 15 HIV-1 seropositive and 17 seronegative immunosuppressed patients. IgG binding to three peptides, namely, the C-terminal half of murine leukemia virus (MLV) capsid protein, the conserved portion of HERV-H transmembrane protein, and the Pol region of human mouse mammary tumor virus (MMTV)-like (HML3) sequence, was observed in both groups. Binding was, however, more frequent and more firm in HIV-1 positive samples (P<0.0001, Wilcoxon rank sum test). IgM binding to the same peptides showed no significant differentiation between the two groups of patients. Binding to both immunoglobulin isotypes was sometimes variable over time in both groups. No correlation of either IgG or IgM peptide binding with progression to AIDS in HIV-1 infected individuals was observed. Inhibition studies using analogous endogenous and exogenous retroviral peptides, including HIV-1, demonstrated specificity of the IgG antibodies for a narrow range of MLV- and MMTV-like retroviral antigens, and excluded cross-reactivity of antibodies to HIV-1 as a cause of these observations. Thus, unlike IgG, IgM binding to retroviral antigens was ubiquitous. It is suggested that anti-HERV IgM belong to a class of natural antibodies and might serve as primers in the mediation of humoral immune responses to more or less related exogenous retroviruses. Increased IgG binding in HIV-1 infected individuals could result from such priming, or reflect higher HERV antigen expression.

Structural prerequisites for intersubtype B and D antigenicity of the third variable envelope region (V3) of human immunodeficiency virus type 1.

To elucidate the structural requirements for intersubtype antigenicity of human immunodeficiency virus type 1 (HIV-1) third variable envelope region (V3), synthetic peptides were used in enzyme immunoassays (EIAs) with serum samples from persons with proven or probable subtype B and D infections. Mathematical analyses of results from EIAs with singly substituted V3 peptides revealed important residues determining overall N-terminal V3 peptide antigenicity. This information was used to design V3 immunogens, rabbit antiserum to which were tested in EIA and for in vitro neutralization of molecular clones of HIV-1(MN) and HIV-1(MAL). Intersubtype-reactive epitopes were distributed toward the N-terminal half of the V3 loop. Lysine at position 310, arginine at position 311, and isoleucine at position 314, all derived from the MN primary sequence, were major determinants of intersubtype V3 antigenicity. Combinations of residues that enhanced antigenicity often contained lysine at position 310. Threonine at position 308 was common in the least advantageous combinations. V3 immunogens modified to achieve optimal antigenicity induced antiserum with augmented cross-neutralization of virus from MAL and MN molecular clones, suggesting one approach to subunit vaccine development.

Comparative studies on neutralisation of primary HIV-1 isolates by human sera and rabbit anti-V3 peptide sera.

IgG binding to V3 peptides and serum neutralising responses were studied in four HIV-1 infected individuals with progressive disease over a period of 31-70 months. The 18-20 mer peptides comprised residues 299-317 (numbering of HIV1 MN) in the N-terminal half of the V3 loop of the envelope glycoprotein gp120 and were derived from the sequences of autologous, as well as heterologous isolates. All four individuals studied lacked anti-V3 IgG binding to at least one autologous V3 sequence. V3 peptides to which autologous sera lacked binding IgG were all immunogenic in rabbits and induced antisera that were broadly cross-reactive by EIA and broadly cross-neutralising to primary HIV-1 isolates. This indicates that the peptides are immunogenic per se and that the respective human hosts have selective defects in recognising the corresponding V3 sequences. Despite the absence of antibody binding to autologous V3 peptides, the human sera had neutralising antibodies to autologous (three out of four cases), as well as heterologous isolates (all cases). Moreover, in vitro exposure of the patients' isolates to autologous neutralising serum or the homologous rabbit antiserum selected for variants with amino acid substitutions close to the crown of the V3 loop or in regions outside the sequence corresponding to peptides used for immunisation. The amino acid exchanges affected V3 positions known to be antigenic and which are also prone to change successively in infected persons. It is likely that neutralising antibodies recognise both linear and conformational epitopes in the V3 loop. Apparently, there are several, but restricted, numbers of ways for this structure to change its conformation and thereby give rise to neutralisation resistant viruses.

Continuity and discontinuity in the anti-V3 IgG response of human immunodeficiency virus type 1-infected persons in a cross-sectional and longitudinal study using synthetic peptides.

The principal neutralization domain (PND) of the V3 region of human immunodeficiency virus type 1 (HIV-1) gp120 is central to HIV pathogenesis. The IgG antibody response to PND was followed in 15 HIV-1-infected persons from southern Sweden over 2-5 years using 32 synthetic V3 peptides. Five peptides had amino acid sequences derived from isolates from each of 5 patients. Sera obtained simultaneously with isolate almost always reacted strongly with these cognate peptides; however, reactivity was undetectable in 1 patient's serum and short lived in the sera of another, indicating inducible holes in the antibody repertoire, which would facilitate dissemination of the corresponding virus strains. Reactivity to other V3 peptides correlated with sequence similarity to the cognate peptide. Strong, stable reactivity to peptides with sequences similar to a south Swedish V3-consensus was accompanied by transient activity to less similar ones. The latter may reflect viral variation, B lymphocyte clonal depletion, or both. Certain IgG responses appeared to preclude others, suggesting clonal dominance.

A survey of synthetic HIV-1 peptides with natural and chimeric sequences for differential reactivity with Zimbabwean, Tanzanian and Swedish HIV-1-positive sera.

OBJECTIVE: To determine whether the known sequence differences between African and non-African HIV-1 strains are reflected in the serological response. DESIGN AND METHODS: We investigated the antibody reactivity of 34 Swedish, 30 Tanzanian and 42 Zimbabwean HIV-1-positive sera to 67 synthetic peptides with sequences from North American and African HIV-1 isolates, mostly derived from regions of gag and env known to be antigenic. Not all sera were tested against all peptides. RESULTS: Differences in frequency of reactivity were noted with peptides covering the entire third variable domain (V3), which is a primary neutralization determinant, and the carboxyl terminus of gp120, in two regions of gp41, and the carboxyl terminus of p24. In env Tanzanian sera reacted preferentially with a V3 peptide from the strain JY1 (Zaire). Gradual substitutions in the central motif in V3 of ELI from GLGQ to GPGR, typical of many non-African strains, led to a gradual increase in reactivity of many Swedish sera, but did not affect Tanzanian and Zimbabwean sera, suggesting that the major epitopes recognized by these African sera are outside GPGR. V3 peptides from the MN and Z3 strains reacted with most sera, but missed 30% of those of Tanzanian origin. In the carboxyl terminus of gp120 both sets of African sera reacted preferentially with peptides from strains JY1 and MAL. Swedish sera reacted strongest with analogues from strains Z321 and HXB2. In gp41, Swedish sera showed a weak preference for reactivity with HXB2-derived peptides in the immunodominant region (amino acids 590-620), and further towards the carboxyl terminus (amino acids 620-665). CONCLUSION: The differences in serological reactivity were as great between Zimbabwe and Tanzania as between the two African sets and the Swedish. The geographical differences in the pattern of reactivity with HIV peptides probably depend on both host and viral variation and may be developed into a seroepidemiological tool, useful for optimization of future HIV vaccines.

PIP: The objective of this study was to determine whether the known sequence differences between African and non-African HIV-1 strains are reflected in the serological response. The authors investigated the antibody reactivity of 34 Swedish, 30 Tanzanian, and 42 Zimbabwean HIV-1 positive sera to 67 synthetic peptides with sequences from North American and African HIV-1 isolates, mostly derived from regions of gag and env known to be antigenic. Not all sera were tested against all peptides. The authors noted several results. Differences in frequency of reactivity were noted with peptides covering the entire third variable domain (V3), which is a primary neutralization determinant, and the carboxyl terminus of gp120, in 2 regions of gp41, and the carboxyl terminus of p24. In env, Tanzanian sera reacted preferentially with a V3 peptide from the strain JY1 (Zaire). Gradual substitutions in the central motif in V3 of ELI from GLGQ to GPGR, typical of many non-African strains, led to a gradual increase in reactivity of many Swedish sera, but did not affect Tanzanian and Zimbabwean sera, suggesting that the major epitopes recognized by these African sera are outside GPGR. V3 peptides from the MN and Z3 strains reacted with most sera, but missed 30% of those of Tanzanian origin. In the carboxyl terminus of gp120, both sets of African sera reacted preferentially with peptides from strains JY1 and MAL. Swedish sera reacted strongest with analogues from strains Z321 and HXB2. In gp41, Swedish sera showed a weak preference for reactivity with HXB2-derived peptides in the immunodominant region (amino acids 620-665). The differences in serological reactivity were as great between Zimbabwe and Tanzania as between the 2 African sets and the Swedish. The geographical differences in the pattern of reactivity with HIV peptides probably depend on both host and viral variation and may be developed into a seroepidemiological tool, useful for optimization of future HIV vaccines.

Seroepidemiology of human retroviruses in Ogun State of Nigeria.

We analyzed sera collected during 1987 and 1988 from 385 healthy business employees of both sexes, of Ogun state in Nigeria, for antibodies to the 3 human retroviruses HIV-1, HIV-2 and HTLV-I. No serum was HIV-1 positive, 1 was HIV-2 positive and 2 were HTLV-I positive. A few sera were false-positive in the antibody screening tests which preceded the confirmatory antibody tests. In the confirmatory tests, we found that in the HIV-1 Western blot test 1 serum reacted only with the HIV-1 gag protein p17, and 2 sera reacted only with the HIV-1 pol proteins p64, p53 and p31. None of these reactivities fulfill internationally accepted criteria for HIV-1 seropositivity. We conclude that HIV-1 was rare in the study population and that HIV-2 and HTLV-I are present at a low frequency. The false positive serological reactions observed are similar to those described previously from Africa and elsewhere. The findings emphasize the importance of routinely testing blood donations for antibodies to these retroviruses in Nigeria.