HLA disease association and protection in HIV infection among African Americans and Caucasians

In a previous investigation, we demonstrated an increased progression of overt AIDS in the African American population compared to the Caucasian population as reflected by the significantly lower absolute number of CD4+ lymphocytes detected in the African American population in an earlier study. The present study elucidates some of the possible genetic factors which may contribute to disease association or protection against HIV infection. The HLA phenotypes expressed as A, B, C, DR and DQw antigens were revealed by the Amos-modified typing procedure. NIH scoring was utilized to designate positive cells taking up trypan blue. A test of proportion equivalent to the chi 2 approximation was used to compare the disease population (n = 62; 38 African Americans, 24 Caucasians) to race-matched normal heterosexual local controls (323 African Americans, 412 Caucasians). Significant p values were corrected for the number of HLA antigens tested. HLA markers associated with possible protection from infection for African Americans were Cw4 and DRw6, whereas Caucasians expressed none. Disease association markers present in the African American population were A31, B35, Cw6, Cw7, DR5, DR6, DRw11, DRw12, DQw6 and DQw7, whereas in the Caucasian population A28, Aw66, Aw48, Bw65, Bw70, Cw7, DRw10, DRw12, DQw6 and DQw7 were demonstrated. The highest phenotypic frequency for a disease association marker in the study was for HLA-DR5 (62.9%) in the HIV-infected African American population without Kaposi's sarcoma compared to a frequency of 28.9% for the regional control group (p = 0.0012). We conclude that genetic factors do have a role in HIV infection since only 50-60% of those exposed to the AIDS virus will become infected.     

Monosomy 6 in a human lymphoma line induced by selection with a monoclonal antibody

The human Epstein Barr Virus-superinfected B lymphoma cell line BJAB-B95.8.6 was mutagenized by gamma irradiation, and HLA mutants were selected with the HLA-Bw6-specific monoclonal antibody SFR8-B6. One of the mutants obtained, BM19, had lost one of the chromosomes 6 present in the wild type cells. Electrophoretic analysis of phosphoglucomutase isozyme PGM3 and erythrocyte glyoxalase 1 from both cells supports this conclusion. The HLA antigens expressed on BM19 were HLA-A2, B13, Bw4, C-, DR2 (questionable), DRw52 (weak) and DQw1. This constitutes one of the haplotypes of the wild type cells, the other (lost from BM19 cells) being HLA-A1, B35, Bw6, Cw4, DR5, DRw52 (strong) and DQw3. Possibilities to employ BM19 cells for the analysis of the major histocompatibility complex and other chromosome 6-encoded genes as well as their products are discussed.     

Unique HLA-DR/DQ associations revealed by family studies in Warao Amerindians. Haplotype and homozygosity frequencies

HLA-A, Cw, B and A, Cw, B, DR genotypes have been assigned, respectively, to 318 and 175 Warao Amerindians belonging to 73 sibships, who were tested with International Histocompatibility Workshop reagents. Only 33% of the theoretically possible three-loci and 7% of the possible four-loci haplotypes were found, with 10 and 6 of them accounting, respectively, for 75% of the total observed. This limited haplotype variability, expected in an inbred population, was not accompanied by either an increased or a decreased frequency of homozygous individuals, as demonstrated by population and family analysis. Inheritance of five HLA loci haplotypes in 20 families showed the expected distribution of parental haplotypes in sibling pairs. The study revealed DR2sh (DRw16), DR4 and DRw6 in association with DQw7, and DRw8 with DQw4, and significant positive linkage disequilibria between Bw62 CW1, B51 DRw16, B39 DR4, Bw62 DRw6, and DRw8 DQw4. The DR2-DQw7 and DRw6-DQw7 associations and the first three paired loci disequilibria mentioned are described for the first time in Amerindians and have not yet been found among Japanese, Negroid, or Caucasoid populations.     

Autoreactive T cells in rheumatic disease. II. Function and specificity of an autoreactive T helper cell clone established from a HLA-B27+ reactive arthritis

T cell lines were established by limiting dilution of peripheral blood (PBL) and synovial fluid lymphocytes (SFL) of a patient with HLA-B27+ reactive arthritis. Among these cell lines, the CD4 phenotype was dominant. Functionally, the majority of these cell lines exhibited helper activity for the immunoglobulin production by autologous B cells and proliferated in response to autologous mononuclear cells. In most cases, this autoreactive response was associated with alloreactivity. Only one cell line, the autoreactive CD4+ T cell clone, UA-S2, which was derived from the synovial fluid, proliferated in a highly specific manner in response to a determinant associated with MHC class II products present on autologous mononuclear cells. The restriction element was shown to be associated with DR molecules by inhibition experiments with monoclonal antibodies. Within the patient's family, the capacity of mononuclear cells to stimulate a proliferative response of UA-S2 segregated together with the HLA haplotype A2 or 32, B27, Cw1, DRw11 which was contributed by the patient's mother. UA-S2 proved to be a functional helper T cell clone. In the absence of additional antigen or mitogen, it induced IgG and IgM synthesis of autologous and family members' B cells. This helper activity of UA-S2 showed the same MHC restriction as the proliferative response. Although the patient's father also typed DRw11, this haplotype was not recognized by UA-S2. It is suggested that this autoreactive T cell clone detects a microheterogeneity of the serologically defined DRw11 haplotype. Indeed, typing of the patient's family members with cellular reagents established a difference between the two DRw11 haplotypes.     

A T-cell clone recognizing an MLC stimulating epitope located on the DRw11β1 chain

An alloreactive proliferative T-cell clone, 6065 WS, was obtained in an intrafamilial cell combination where the stimulator was a homozygous DRw11, DRw52, DQw3 son and the responder his haploidentical mother. Proliferation assays on eight local DRw11 families, 75 homozygous B-cell lines (Tenth Histocompatibility Workshop panel), and blocking assays with monoclonal antibodies showed that clone 6065 WS recognizes an epitope on the DRw11β1 chain. Comparison of the reactivity of clone 6065 WS with cells expressing the three known DRw11β1 amino acid sequences identified two unique amino acids at positions 71 and 86 which contribute to determining the specific recognition by the T-cell clone 6065 WS. Our data suggest that one or both of these amino acids can either be directly involved in the recognition by the T-cell receptor or responsible for critical conformation of the determinants on the DR molecule. Alternatively, they could affect recognition of a self peptide bound to the major histocompatibility complex class II molecule.     

HLA typing with monoclonal antibodies: evaluation of 356 HLA monoclonal antibodies including 181 studied during the 10th International Histocompatibility Workshop

During the 10th International Histocompatibility Workshop (10th WS), 181 HLA MoAbs were studied using lymphocytotoxicity micro-technique (LCT) and/or enzyme immuno-assay (EIA), and their capacity to serve as typing reagents was evaluated. 129 MoAbs were tested by both techniques. Results obtained with 92 class I and 86 class II polymorphic MoAbs (10th WS) were compared to published data concerning 180 class I and 176 class II polymorphic MoAbs, listed in an HLA-MoAbs Register maintained in our laboratory. The following conclusions can be proposed: 1/HLA-A, B typing by LCT with MoAbs is possible for about 14 specificities. Some specificities are clearly recognized (HLA-A3, B8, B13, Bw4, Bw6), others are recognized as cross-reacting groups (B7+27+w22+40), others are not currently recognized by any MoAb with restricted specificity (B5, B15). Several MoAbs confirmed the existence of shared epitopes between products from a single locus (A2-A28, A25-A32), or from A and B loci (A2-B17, Bw4-A9-A32). A single HLA-Cw MoAb has been described. 2/HLA class II typing by LCT with MoAbs is more difficult than class I typing. DR2, DR3, DR4, DR5 and DR7 as well as DRw52 and DRw53 are well defined; other DR specificities are poorly or not at all defined. Particular associations (DR1+DR4, DR3+DRw6, all DR except DR7) are recognized by several MoAbs. All DQw specificities are well recognized, including new specificities defined only by MoAbs: WA (DQw4), TA10 (DQw7), 2B3 (DQw6+w8+w9). Only two HLA-DP MoAbs have been described. 3/Satisfactory results, similar to those of LCT, were obtained with EIA using lymphoid cell lines as targets. 4/Human MoAbs (12 in the Register) are satisfactory typing reagents. They could represent in the future a significant contribution to HLA typing with MoAbs.     

HLA polymorphism in a Mataco South American Indian tribe: Serology of class I and II antigens: Molecular analysis of class II polymorphic variants

In the present study, HLA-A, B, C, DR, DQ, and DP loci were analyzed in a group of Mataco Amerindians of Argentina. Using reagents from the 11th International Histocompatibility Workshop (11th IHW), class I specifities such as Bw70, Bw75, and Bw48 were found in this population, other than the HLA determinants commonly described in South American Indians. The class II antigens found were DR4, DRw14, and DRw8 at the DR locus, and DQw4 and DQw7 at the DQ locus. The analysis of DRB1-DR4 related alleles, performed by PCR amplification and oligonucleotide probe hybridization, showed the presence of DRB1*0403, *0404, *0405, and *0411 in individuals from this ethnic group. By the analysis of DRB1-DRw14 related alleles, two variants were found: DRB1*1402 and DRB1*1406, the latter provisionally called DRB1 14.6 in 11th IHW. The DRw8-related allele present was DRB1*0802. The analysis of DRB3 gene revealed only the presence of DRB3*0101 allele in DRw14 individuals. DPB1 locus was also analyzed in unrelated individuals of the same population. Only five DPB1 alleles were found: DPB1*0201, *0301, *0402, *0501, and *1301 over the 19 previously described in the literature. These findings emphasize the restricted HLA class I and II variation observed in this ethnic group as it has been previously shown in other American groups. Some particular haplotypes in this Mataco tribe are described in this work.     

Association of HLA-Bw65 with two major complotypes

The association between the HLA-B14 subtypes Bw64 and Bw65 and complement allotypes (C2, Bf and C4) was investigated in both population and family studies. Bf, C4A and C4B allotyping was performed on 37 Bw64 and 35 Bw65 positive unrelated Welsh/English subjects. Sixteen HLA-Bw65 bearing haplotypes were characterized for HLA-ABC, DR and DQ antigens and complement allotypes, including C2. The findings of the population study suggested that the complement haplotype associated with Bw64 is BfS, C4A2, C4B2. The population and family studies revealed two major complement haplotypes associated with HLA-Bw65: (i) C2C, BfF, C4A3, C4A1 - often associated with HLA-A3, Cw8 and DRw13, and (ii) C2C, BfS, C4A2, C4B2 - often associated with HLA-Aw33, Cw8 and DR1 or with A28, Cw8 and DRw13. The HLA-Bw65 bearing haplotypes of three families carried a C4B2B1 duplication of the C4B locus. In these families three C4B gene products were identified in the Bw65 positive members using an anti-C4B monoclonal antibody. It is suggested that most, if not all, HLA-Bw65 bearing haplotypes may possess a C4B locus duplication.     

Anti-HLA-B7,B27,Bw42,Bw54,Bw55,Bw56, Bw67,Bw73 monoclonal antibodies: specificity, idiotypes, and application for a double determinant immunoassay

The monoclonal antibodies (MoAbs) KS3 and KS4 are secreted by hybridomas constructed with splenocytes from a BALB/c mouse sequentially immunized with the cultured lymphoid cells JKu and LG-2 which share only the HLA-B27 specificity. Serologic and immunochemical assays have shown that the two MoAbs recognize the same (or spatially close) determinant expressed by HLA-B7,B27,Bw42,Bw54,Bw55,Bw56,Bw67, and Bw73 alloantigens. This determinant is spatially close but distinct from those defined by the anti HLA-B27 monoclonal antibodies described in the literature. The syngeneic antiidiotypic MoAb T12-105 and T12-211 elicited with MoAb KS4 were shown to recognize idiotopes within the antigen combining site of MoAb KS3 and KS4. Neither idiotope was detected on the anti HLA class I and anti HLA class II monoclonal antibodies tested. The MoAb KS4 in combination with the anti human beta 2-microglobulin MoAb NAMB-1 was utilized to develop a double determinant immunoassay (DDIA). The latter represents a sensitive method to detect and quantitate HLA-B27 antigens in spent culture medium of lymphoid cell lines and in serum. Typing for HLA-B27 antigens with the DDIA of sera from HLA typed donors yielded results highly correlated with those of the conventional lymphocytotoxicity assay.     

CD4+T cells restricted by DQ not by DR support CD8+T cells to grow

Much attention has been paid whether there are any differences in regulating the human immune response between HLA-DR and -DQ molecules encoded by the genes within the HLA class II multigene family. Previous studies have suggested that HLA DQ molecules control low responsiveness through activating CD4 T cells which generate CD8 positive T cells, whereas HLA -DR molecules control high responsiveness through activating CD4 helper T cells. To examine this model we investigated the streptococcal cell wall antigen (SCW) specific T cell lines restricted by either DR or DQ molecule. To identify the restricting molecules, L cell transfectants expressing DQw1, DR2AB1 or DR2AB5 from Dw12 haplotype or DQw4, DR4 or DRw53 from DW15 haplotype were used. 1. From individuals with Dw12 which is a low responder haplotype to SCW, T cell clones specific to SCW and restricted by HLA-DQw1 or DR2 were identified, whereas from individuals with Dw15 which is a high responder haplotype, only DR4 or DRw53 restricted T cell clones were identified and DQw4 restricted T cells were never observed. 2. SCW specific CD4 T cells restricted by DQw1 were able to support the growth of CD8 positive cells, whereas those restricted by DR4 could not do so. 3. The CD8 T cells also required autologous antigen presenting cells and SCW to grow, and they completely blocked the immune response to SCW in vitro. These observations clearly demonstrated the distinct function of HLA-DQ and -DR molecules in regulating the human immune response to SCW.     

HLA association of idiopathic Peyronie's disease: An indication of autoimmune phenomena in etiopathogenesis?

Abstract: HLA typing for class I and class II antigens was done in 52 unrelated patients suffering from idiopathic Peyronie's disease. The controversially discussed association with the HLA-B7 cross-reacting group could not be confirmed. Marked deviations of antigen frequencies were observed for HLA-A1, B8, Cw7, DR3 and DQw2 compared to healthy local controls. After correction of p-values, Al (p_c < 0.05) and DQw2 (p_c < 0.01) remained significant. A possible association of Peyronie's disease with markers of the HLA-A1, B8, Cw7, DR3, DQw2 haplotype, as first described here, would suggest autoimmunological factors in this disorder of otherwise unknown etiopathogenesis.     

A T cell clone defining a common DP/DR determinant in strong linkage disequilibrium with HLA-A9

Clone F-7 was generated by limiting dilution of lymphocytes stimulated by allogeneic PBL on MLC. Priming against the A23, Cw6, B45, DR7, DRw53, DQw2 haplotype was performed between two HLA haploidentical first degree relatives. The clone was tested for its ability to proliferate in response to a panel of 38 homozygous B lymphoblastoid cell lines plus three local T cell lines. It showed a pattern of reactivity corresponding to HLA-A9 specificity (r = 1) and presented a concomitant cytotoxic activity. Phenotypically, this clone consisted entirely of CD4 cells, as determined by indirect immunofluorescence. Its reactivity was completely blocked by anti-DR (GSP4.1, PL8, L243) and anti-DP (B7/21, PL15) Mo-Abs, whereas anti-DQ (1A3, TU22) and anti-class I (w6/32, BB7.7) Mo-Abs and anti-A9 antibodies did not inhibit its reactivity. These results may suggest that clone F-7 could recognize a DP specificity sharing common determinants with DR, which occurs in linkage disequilibrium with HLA-A9.     

A monoclonal antibody detecting an HLA-DQwl-related determinant

A complement fixing monoclonal antibody (moab) was prepared which reacts with a polymorphic determinant on HLA class II molecules. The moab IIB3 recognises all DQwl (DC1, MB1, LB-E12) positive cells as well as some DR4, DR7, DRw8 and DRw9 positive cells. The moab reacts mainly with B-cells and not or with only a minority of the monocytes. Segregation of the determinant with HLA-DR could be shown. The determinant is strongly expressed on DR2, DR4 and DRw6 positive cell lines but only weakly on DR1 lines. In contrast to a monoclonal antibody against a monomorphic determinant on class II molecules IIB3 did not give a distinct inhibition of the MLC nor did it inhibit the generation of CTLs in MLC as has been described for the DQwl like moab BT 3/4 (Corte et al. 1982). Immunoprecipitation indicates that IIB3 reacts with DQ-like molecules.     

HLA association of idiopathic Peyronie's disease: an indication of autoimmune phenomena in etiopathogenesis?

HLA typing for class I and class II antigens was done in 52 unrelated patients suffering from idiopathic Peyronie's disease. The controversially discussed association with the HLA-B7 cross-reacting group could not be confirmed. Marked deviations of antigen frequencies were observed for HLA-A1, B8, Cw7, DR3 and DQw2 compared to healthy local controls. After correction of p-values, A1 (pc less than 0.05) and DQw2 (pc less than 0.01) remained significant. A possible association of Peyronie's disease with markers of the HLA-A1, B8, Cw7, DR3, DQw2 haplotype, as first described here, would suggest autoimmunological factors in this disorder of otherwise unknown etiopathogenesis.     

T-cell recognition of class II products that result from the combined presence of two different HLA haplotypes.

To analyze DR2 haplotypes as recognized by alloreactive T cells, lymphocytes from a DR7; DQw2 homozygous donor were cocultured with irradiated lymphocytes that were DRw15, DR7; DQw6, DQw2 heterozygous. In this report, we focus on two HLA-DQ-specific T-cell clones obtained from this priming. These two clones (c3518 and c3523) responded to the positive control (original stimulator) and five of 66 panel donors. Three of these donors typed DRw15, DR7; DQw6, DQw2, as did the positive control. One stimulatory donor typed DRw15, DR7; DQw6, DQw9 and one stimulatory donor typed DRw14, DR7; DQw5, DQw2. Oligonucleotide typing revealed that recognition by the clones depended on the simultaneous presence of the DQB1*0602 gene on one haplotype and DRB1*0701 or DQA*0201 on the other. The hypothesis that c3518 and c3523 recognize an HLA class II product that results from the combination of two different HLA haplotypes was further confirmed in family studies. In three families, it was shown that the DRw15, DR7; DQw6, (DQw2 or DQw9)-positive individuals were recognized, whereas the cells carrying either DRw15; DQw6, DR7; DQw2, or DR7; DQw9 were nonstimulatory. Our results can be explained in two ways: (a) the T cells recognize a class II dimer that results from trans-complementation of DQA1*0101 and DQB1*0602, and (2) the T cells recognize a DR7-derived peptide that is presented by DQw6.     

The HLA system in the Korean population

The frequencies of HLA-A, B, C, DR, and DQ antigens, HLA-D (HTC-defined) haplotypes, and the HLA-linked genetic markers glyoxalase I (GLO), factor B (Bf), C2 and C4 were studied in 162 healthy unrelated Koreans. Antigens A2, A24, A26, B44, B51, Bw62, B35, Cw1, Cw3, DR2, DR4, DRw6, DR7, and DRw8 were observed at frequencies of 15% or greater, and GLO-2, BfS, C4A*3, C2C, C4A*4, C4B*1, and C4B*2 were also frequently observed. The antigens A23, A25, B18, Bw42, Bw47, and B21 were not observed at all. HLA-DR4 was the most common class II antigen and was associated with a series of HLA-D-defined haplotypes including Dw4, Dw10, Dw13, and Dw15. The HLA-DRw6, DR2,Dw8, and DRw8 haplotypes were also found frequently. DR2 haplotypes were either Dw2 or Dw12, while all DRw8 haplotypes tested corresponded to the DB7 or Dw "8.3" specificity that has been described in other Oriental populations. Significant linkage disequilibrium was found between the alleles A2,Cw1; A30,B13; A30,Cw6; A30,DR7; Cw1,Bw22; Cw5,B12; Cw6,B13; Cw6,DR7; B7,DR1; B12,Dw6; B12,DR7; B12,Dw7; B13,DR7, B17,DR3; Bw22,C4B*6; DRw6,BfF; and C4A*4,C4B*2. A comparison of gene frequencies and commonly observed haplotypes between Koreans, Chinese, Japanese, and Caucasians showed that while Koreans share several characteristics in common with other Oriental populations, there are allelic frequencies and haplotypes in Koreans that are distinct.     

The effects of HLA class II genes on the susceptibility to autoimmune thyroid diseases

Immunogenetic factors such as HLA, C4, T cell receptor and immunoglobulin allotypes were investigated in 115 Japanese patients with Graves' disease. The patients showed strong positive association with HLA-A2 (R.R. = 3.45, chi 2 = 14.93, Pc less than 0.002), Bw46 (R.R. = 6.47, chi 2 = 16.25, Pc less than 0.002), Cw11 (R.R. = 4.47, chi 2 = 9.19, Pc less than 0.04) and DRw8 (R.R. = 2.22, chi 2 = 5.62, P less than 0.03, Pc: n.s.) which form one of the typical HLA haplotypes in Japanese population due to the strong linkage disequilibria. On the other hand, the patients showed negative association with HLA-B7 (R.R. = 0.15, chi 2 = 7.15), Bw52 (R.R. = 0.24, chi 2 = 7.86), DR1 (R.R. = 0.07, chi 2 = 9.71) and DQw1 (R.R. = 0.45, chi 2 = 5.62), which form HLA-B7-DR1-DQw1 and Bw52-DR2-DQw1 haplotypes. Because HLA-A2 -Bw46-Cw11-DR9 haplotype was reported to be associated with Chinese Graves' patients, and because Bw46 showed the strongest association with the Japanese patients, it was suggested that HLA class 1 antigen, Bw46, might be the primary immunogenetic factor involved in the pathogenesis of Graves' disease. Since HLA-DQw6 was reported to be associated negatively with Hashimoto's thyroiditis as same as the current observation in Graves' disease, it was suggested that HLA-DQw6 may determine the resistance to autoimmune thyroiditis. The effect of HLA-DQw6 gene, therefore, on the experimental autoimmune murine thyroiditis (EAMT) was examined, using DQw6-transgenic mouse. F1 with C3H mouse, and backcross progeny between the F1 and C3H mouse which is a susceptible strain to EAMT. The measurement of anti-thyroglobulin antibody indicated that C3H mouse, (C3H x DQ-B6) F1 and backcross progeny between the F1 and C3H were high responders to the thyroglobulin, but that B6 mouse and DQ-B6 mouse were low responders. The histological examination of the thyroid gland of these mice failed to demonstrate the significant difference in susceptibility to EAMT among these mice. These results suggested that the immune response to the thyroglobulin was controlled by H-2k haplotype and that the effect of HLA-DQw6 gene on the immune response to thyroglobulin and on the autoimmune thyroiditis was marginal.     

A new duplication at the C4B locus associated with the HLA-Aw68, Cw8, Bw65 haplotype

A family with a cross-over between HLA-B and HLA-DR was analysed for its complement alleles. This allowed location of the cross-over between HLA-B and C4. Furthermore, the same family showed a previously undescribed duplication at the C4B locus (C4B* 2,2) that was associated with the HLA-Aw68, Cw8, Bw65, C2*1, Bf*S, C4A*2, DR7, DQw2 haplotype.     

Analysis of human class II antigens by cloned cytolytic T cell reagents: a study using HLA loss mutant lymphoblastoid cell lines and monoclonal antibodies detecting the HLA-DP product(s)

We have utilized cloned T cell reagents and ionizing radiation-induced mutants of an HLA heterozygous lymphoblastoid cell line (LCL) to investigate the determinants detected by the cell-mediated lympholysis (CML) assay. Cells of an LCL clone, 721.501, an HLA haplotype loss mutant expressing the HLA-A2-Cw1-Bw51-DR1-Dw1-DQw1-DPw2-GLO haplotype were used as sensitizing cells for responder cells in vitro. "Cloned" reagents were generated by single-cell deposition of cells of a bulk reagent primed against 721.501 cells. Those clones were screened for cytolytic activity against HLA loss mutant targets (derived from LCL 721) of four different categories; HLA-A2 loss only, A2-Cw1-Bw51 loss, A2-Cw1-Bw51-DR1-DQw1 loss, and the entire HLA haplotype loss. Of 196 clones tested, 36 were cytolytic, including three anti-A2, five anti-Bw51/Cw1, 12 anti-DR1/DQw1, 13 anti-DP region associated with DPw2, and three of undetermined specificity, based on cytolytic patterns against the HLA loss mutant targets. Of 25 anti-HLA class II lytic clones, 23 (92%) fitted the characteristics of helper cell-independent cytolytic T cells (HITc), whereas only two of eight (25%) anti-class I clones were HITc. The 13 anti-DP region clones were divided into three subgroups defined by blocking by anti-FA and not Tü39 monoclonal antibodies (MoAb), by Tü39 and not anti-FA, and by both MoAbs.     

The HLA genetic constitution of the Bushmen (San)

The HLA-A, -B, and -C antigens of 290 and the DR antigens of 212 !Kung San individuals were characterized. The most frequent antigens were HLA-A30 gene frequency (gf) = 0.193, Bw58 (gf = 0.303), Cw6 (gf = 0.327), DR4 (gf = 0.273), and DQw3 (gf = 0.553). An unexpected finding was the low frequency of the classic African black antigen Bw42 (gf = 0.004). Marked differences as well as similarities in HLA gene frequencies were observed between the San and the South African Negroes, supporting the view that they had a common origin and were then separated for a very long time. During this period differences developed as a result of selective advantage in the Negroes following the pastoralist-agriculturalist way of life as opposed to the hunter-gatherer way of life. The picture is further complicated by the fact that gene flow, mostly from the San to the southern African Negroes, took place when they met again a few hundred years ago. The data also illustrate HLA haplotypes, linkage disequilibria, and four-locus haplotypes not previously seen in other human populations. The most frequent four-locus haplotype in the San, HLA-Aw43, Cw7, B7, DRw6 was also different from A30, Cw2, Bw42, DR3, the most common among the South African Negroes.