Search

Several South, Central and North American Indian populations have been characterized over the past 9 years. A summary is given here. Very few major antigens have been found, namely A2, A9, A28 and Aw31 and B5, B15, Bw16, Bw35 and B40. However, several known splits of these antigens have been identified, and new variants are also likely. Immunogenetic analyses of the Amerindians should be able to provide clues both to the origins of the Amerindian cultures and migrations of early man, and also provide new insights into the antigens and the loci of the MHC.

A computer program initially written by the Milwaukee Blood Bank has been modified to use a new algorithm for the assignment of HLA specificities to antisera. The assignment is based on the reactions of cells with known specificities. Specificities which are present only on cells which do not react are first ruled out. This step is followed by one or more steps in which the 'least reactive' specificities are ruled out. The rationale for the algorithm is discussed and an example is presented.

A cytotoxic monoclonal antibody, SFR8-B6, produced by immunizing KGH rats with partially purified glycoprotein from JY, a human B lymphoblastoid cell line, detects HLA-Bw6, a unique public antigen on the HLA-B molecule. By cytotoxicity, SFR8-B6 has complete concordance with the presence of Bw6 as defined by alloantisera, and its reactivity segregates with HLA in families. Because sera detecting Bw4 and Bw6 are rare, this monoclonal antibody has value as a perpetual standard. In addition, SFR8-B6 offers potential as a probe for investigating the molecular relationships of HLA antigens.

Polynesians living on the island of Mauke in the Cook Island group were typed for HLA-A and -B locus antigens. The Mauke population has restricted HLA polymorphism, with five A-locus antigens and four B-locus antigens accounting for a majority of the HLA phenotypes. Although some differences in antigen frequency were found when Mauke Islanders were compared with Polynesians from Easter Island and Samoa, the Mauke Islanders were closer in their HLA antigenic profile to polynesians than to Melanesians.

Varying the sensitization time of the complement-dependent antibody mediated cytotoxicity test, while carefully controlling other test conditions, has enabled us to characterize the subtle reactivities of several broadly reactive sera. In particular, we can discriminate between classes of cells, e.g., HLA-B7 from B27; subclasses cells, e.g., Bw51 from Bw52; and cells carrying new antigens, e.g., Ao79. This manipulation of the cytotoxicity test is discussed in view of its usefulness in separating primary from cross-reactive specificities and in detecting new specificities.

Fourty-four unrelated North American Blacks and one Black family were tested for B-cell specific antigens with 7th International Workshop antisera. DR specificities were clearly defined in this group, but were generally less frequent than reported for Black Americans in the 7th Workshop report and were most similar in frequency to those reported for African Blacks. Five new B-cell specificities (DuB40-43, 45) were identified. In contradistinction to Caucasians, Black Americans type for HLA-D with homozygous typing cells failed to exhibit strong linkage disequilibrium between D and DR types.

A positive association of Ra3 skin test responses with HLA-A2 has previously been reported to be evident in individuals with low IgE levels and to a greater extent in these individuals than those with high IgE levels. We give evidence based on an analysis of data from 133 individuals that Ra3 response is positively correlated with HLA-A2 among individuals with low Antigen E response and negatively associated with HLA-A2 among individuals with high Antigen E response. Furthermore, we have evidence that any observed interaction between Ra3, IgE, and HLA-A2 can be explained by the correlation between IgE and Antigen E response, and that it is Antigen E response which interacts in the relationship between HLA-A2 and Ra3 skin test response.

Sera, previously found to react specifically with B lymphoid cultured cells, were tested on isolated T and B peripheral blood lymphocytes in a microcytotoxicity assay. Studies were performed on lymphocytes obtained from several large Amish families. The sera used in these studies were cytotoxic to peripheral blood, B lymphocytes, but not cytotoxic to T lymphocytes. The antigens detected followed the inheritance pattern of HL-A haplotypes. The strong linkage disequilibrium with HL-A antigens suggests that genes controlling the expression of B lymphocyte antigens are linked to genes controlling HL-A alloantigens.

The cytotoxicity activity of an anti-HL-A2 antiserum cross-reactive with W28 appeared in three distinct peaks after the serum was subjected to isoelectric focusing. The pH at which the peaks formed was reproducible using pH 5-8 or pH 3.5-10 gradients. Anti-HL-A2 activity occurred in each peak, but the cross-reactive anti-W28 cytotoxicity appeared to be largely confined to the second and third peaks. The first peak was exclusively IgG, the second contained IgG and IgM, and the third contained IgG, IgA, and IgM.

An anti-HL-A 3 antiserum with cross-reacting activity for HL-A 1 and HL-A 11 was subjected to isoelectric focusing over a pH 5-8 gradient. The cytotoxic activity of the serum focused into three distinct peaks, one at the basic end of the gradient, one between pH 6 and pH 7 and one at the acidic end. The first and second peaks reacted with HL-A 3 positive cells and HL-A 3 negative cells positive for cross-reacting antigens. The third peak reacted only with HL-A 3 positive cells.

The role of HLA antigens in the generation of cytotoxic cells in CML has been investigated. Cytotoxic effector cells were generated in MLC among HLA-A or HLA-A and HLA-B disparate, HLA-D identical siblings, and among HLA-A and HLA-B disparate, MLC identical (%RR less than or equal to 2 3.6) unrelated individual. The data indicate that HLA-D differences and poliferative MLC responses as measured by 3H-thymidine incorporation are not requisite for the in vitro generation of cytotoxic cells and suggest the existence of a CML-S locus (loci) distinct from HLA-A, HLA-B and HLA-D. The degree of cytotoxicity generated in a proliferative versus a "nonproliferative" MLC was comparable. In addition, these studies demonstrate that antigens other than the currently definable HLA-A, HLA-B, HLA-C, and HLA-D can serve as target determinants in cell-mediated lympholysis.

The HLA antigen profile of the Bari and Yupa Indians who live in the Motilones Valley on the border between Venezuela and Colombia was studied. Both groups showed very limited polymorphism. HLA--A1, A3, A11, Aw23, A25, A26, A29, Aw30, Aw32, and Aw33, and HLA--B7, B8, B12, B13, B14, B17, B18, Bw22, and B27 were not observed in either population.

Human sera were tested for cytotoxicity to pairs of long-term tissue-cultured cell lines. Each pair had been derived from the same individual and one of the pairs possessed the characteristics of either "T" or "B" cells. The alloantisera used were HL-A-typing reagents or sera obtained from Amish multiparas. Selected cytotoxicity was found against the B-cell lines by direct testing. Cytotoxicity was abolished by absorption with B-cell line but not by absorption with the T-cell lines. The results suggest that a group of allotypic antigens may be expressed exculsively on human B cells.

The HL-A antigens of a Chinese population now living on the island of Taiwan have been investigated with reagents obtained from and characterized in Caucasian populations and with antisera obtained from Chinese multiparous women. Both HL-A2, the most frequent specificity found, and HL-A9 appear highly heterogeneous in this Chinese population since groups of antisera used to define HL-A2 or 9 in Caucasians gave discordant reactions. Moreover, new associations were found using chi-square and cluster analyses, further demonstrating the complexity of HL-A in this Taiwanese population. Fourteen % of 1183 Chinese multipara sera and 237 sera from multiple tranfused individuals were lymphocytotoxic and could be placed by cluster analysis into three major groups.

Shared histocompatibility antigens between spouses may affect reproductive outcome adversely as a result of prenatal selection against compatible fetuses. Evidence from both animal and human studies suggest that histocompatible fetuses may not initiate a maternal immunologic response that prevents rejection of the embryo. Therefore, parents sharing HLA antigens may produce compatible fetuses and consequently experience a greater frequency of early fetal losses and show poorer reproductive outcome than couples not sharing antigens. In the Hutterites, an inbred human isolate that proscribes contraception, we tested the hypothesis that couples sharing HLA antigens have poorer reproductive outcomes than couples who do not. The Hutterites are characterized by high fertility and large family sizes. Couples that share zero (no. = 21), one (no. = 15), and more than one (no. = 10) HLA-A or HLA-B antigens were compared for reproductive performance. Median intervals between births were larger among couples that share more than one antigen in eight of 11 intervals examined. In addition, the median intervals from marriage to first, fifth, and tenth birth were consistently larger among couples that share more than one antigen. Differences among the groups appear to become larger with increasing parity, suggesting that the effect of histocompatibility on reproductive performance becomes more evident in later pregnancies. These differences in reproductive performance between couples that share zero, one, or more than one HLA-A or HLA-B antigens may have significant evolutionary consequences. However, our results demonstrate that sharing HLA antigens does not preclude normal pregnancy and caution should be exercised before concluding that shared HLA antigens are solely responsible for repeated fetal losses.

Sera detecting part or all of the 4c group of HL-A antigenic specificities were compared in two American Indian populations, the Ixils and the Pimas, and in one Spanish-Indian hybrid population, the Mestizos of Peru. Several new variants appeared. These were not identical to HL-A5, W5 or W18. Two of the variants appeared to represent specificities Z57 and Z51 as described during the 1972 Histocompatibility Testing Workshop. In addition, three new specificities could be detected. These were Z42, Ao88 and Ao93. Z42, Ao88 appeared to belong to the 4c complex and Ao93, a new specificity of the second locus, may belong to this same cross-reactive group.

The HLA profile of three New Guinean populations, two Highland (Asaro, Watut), and one Coastal is presented. The Highland populations are characterized by a low average number of alleles segregating at the HLA loci and also by a low mean value of heterozygosity at these loci. The genetic affinities of the two Highland groups with other Melanesian populations in the Pacific are remote. The Coastal group, on the other hand, shows strong similarities in its antigenic diversity and haplotypic combinations with other Melanesian populations. Nonetheless, the two Highland groups show significant divergence from each other in terms of allelic and haplotypic frequencies. Two different waves of migration settled in the Highlands of New Guinea between 10,000 and 15,000 years ago, and it is possible that the Watut, an Angan speaking group, represents the remnants of the first migration into the interior, whereas the Asaro, members of the Eastern Central family of the Trans-New Guinea phylum, arrived at a later date.

HLA antigens and Gm, A2m, and Km allotypes were examined in Caucasian patients with malignant melanoma. No significant associations were found for any of the HLA antigens tested. Significant association was found with Gm(2), and the relative risk for individuals with this marker was calculated at 1.9. The data indicate that Caucasians positive for Gm(2) are almost twice as likely to develop malignant melanoma as those without this marker.

Anti-HLA alloantisera inhibit mixed lymphocyte responses in which normal lymphocytes are used as stimulator cells. These same antisera are unable to inhibit lymphocyte proliferative responses stimulated by lymphoblastoid cells from cultured lymphoid cell lines. They also fail to inhibit either the generation of cytotoxic effector cells by lymphoblastoid cells or lymphocyte-mediated cytotoxicity against the lymphoblasts. Although the number of HLA antigens on the surface of lymphoblasts is reported to be greater than on normal lymphocytes, the failure of alloantisera to inhibit lymphoblast-induced responses in vitro does not appear to be due to insufficient amounts of antiserum to react with the antigenic sites. Rather, the data are interpreted to suggest that antigens which are not HLA and are not closely associated with HLA on the lymphocyte membrane are responsible for the stimulation of allogeneic lymphocytes by lymphoblastoid cells. Although lymphoid cell lines are known to contain the genome of the Epstein-Barr virus, antisera against products of the viral genome fail to inhibit proliferative responses to lymphoblastoid cells, suggesting that these antigens do not directly participate in lymphocyte activation.

Two hundred and twenty-five Polynesians were selected from a larger study population for the evaluation of potential genetic influences on the susceptibility to bancroftian filariasis. Analysis showed that there was significantly familial clustering of patients with filariasis and that this clustering was most compatible with genetic transmission of disease susceptibility. The data best fit a model in which the hypothetical gene for filariasis was recessive with a frequency of 0.82 +/- 0.15 in the population and a penetrance of 0.62 +/- 0.14. The alternative hypothesis that susceptibility was environmentally (i.e., not genetically) determined was also compatible with the data but was estimated to be 1.9 times less likely to account for the observed findings than the genetic hypothesis. Extensive evaluation of HLA-A and -B locus specificities failed to detect significant linkage either between particular antigen specificities and the clinical manifestations of filariasis or between individual haplotypes (indicated by HLA markers in studies of large families) and the predisposition to filarial infection or disease.

The antibodies occasionally produced in gravid females usually cause no adverse effects and the mother may become hyporesponsive. Exceptions may occur in some cases of abortion where frequency of antibodies appears to be unduly high. Sperm antigens in the mouse include T locus alleles as well as H-2 and H-Y. The T locus is involved in differentiation. T locus alleles are expressed on early but not late embryos. A search for similar factors in man is in progress. The antigenic differences between sperm and mother and between foetus and mother may be important in natural selection. It is possible that the union of sperm with ovum is non-random and that the ovum can select "compatible" sperm through recognition of cell surface markers present on sperm. Immunologic differences between mother and foetus may cause local "graft-versus-host" reactions in the placenta. This leads to increased placental size and may be a factor in hybrid vigour as well as a protective device.

The evolution of HLA specificities has been toward ever-increasing refinement; one example is a subdivision of HLA-B5, the supertype specificity originally defined as 4a. HLA-B5 can now be further subdivided into Bw51 amd Bw52 by serologic means. Whereas the specificity Bw51 can be detected by specific sera, the identification of Bw52 must frequently be deduced from knowledge of B5 and Bw51, although serology has progressed rapidly. There has been no comparable development in identifying fine specificities by cellular cytotoxicity in populations. We have now found that cytotoxic effectors of exquisite specificity can be generated against Bw52 by sensitization of cells from a Bw51 donor and vice versa; Bw51 and Bw52 can in this way be recognized with equal ease. This may set a precedent for recognizing fine specificities of other HLA antigens that cannot yet be identified serologically or can be identified only imprecisely. These fine distinctions may have great relevance in allotransplantation and in understanding disease susceptibility.

Five of seven patients with idiopathic refractory sideroblastic anemia carried an HLA-A3 alloantigen (relative risk, 7.3; P = 0.02). The significance of this association was strengthened by study of two pedigrees. An abnormality in iron metabolism was found in two siblings who had an HLA-A3,B14 haplotype in common with the first proband. A second proband with idiopathic refractory sideroblastic anemia had clinically manifest hemochromatosis. His brother had clinically manifest hemochromatosis but not sideroblastic anemia. This proband and his brother shared only the HLA-A3,B12 haplotype. Our findings infer that patients with idiopathic refractory sideroblastic anemia carry a single allele for hemochromatosis, that this allele accounts for the increased iron loading in this form of anemia, and that clinically manifest hemochromatosis may develop in an occasional patient with only one allele for hemochromatosis in the presence of the sideroblastic factor.

Linkage analyses between 21 genetic markers including HLA-A, B, and the postulated locus for determining total serum IgE levels were done to try to clarify the inheritance of total IgE levels and to map the locus. A total of 316 individuals from five Mormon kindreds were studied, and data from an additional 204 Amish individuals from 11 families were analyzed for possible HLA linkage. Segregation analyses of both data sets did not give clear definition of the mode of inheritance of total IgE levels, but purely environmental models were rejected. Linkage analyses gave significant evidence against HLA linkage with the codominant, recessive, or dominant model of inheritance for total IgE levels. No significant evidence for linkage with any of the genetic markers was obtained. Since total serum IgE levels are correlated with allergies, understanding the genetics of total IgE levels is important to understanding the genetics of allergic disease in man.

In summary, two large families were subjected to genetic analysis of the distribution of the immune response. S linkage analysis indicated that this locus for positive skin test reaction to antigen E and/or personal history or seasonal allergy is linked to HLA-B. The maximum likelihood estimate of the recombination frequency is 0.13 in males and 0.53 in females, with a lod score of 2.53.