Autoreactive T cell clones specific for class I and class II HLA antigens isolated from a human chimera.

T cell clones of donor origin that specifically react with recipient cells were obtained from a SCID patient successfully reconstituted by allogeneic fetal liver and thymus transplantation performed 10 yr ago. The majority of these clones displayed both cytotoxic and proliferative responses towards PBL and an EBV-transformed B cell line derived from the patient. In addition, these T cell clones had proliferative and cytotoxic responses towards the parental PBL, EBV cell lines, and PHA blasts. Blocking studies with anti-class I and anti-class II HLA mAbs indicated that the activity of the CD4+ T cell clones was specifically directed against class II HLA antigens of the recipient. On the other hand, the cytotoxic and proliferative responses of the CD8+ T cell clones were specific for class I HLA antigens which are ubiquitously expressed on the recipient cells. Thus, the establishment of transplantation tolerance observed in this stable human chimera is not due to the elimination of host-reactive T cells from the repertoire and suggests the presence of a peripheral autoregulatory suppressor mechanism.

Rearrangements and point mutations of P450c21 genes are distinguished by five restriction endonuclease haplotypes identified by a new probing strategy in 57 families with congenital adrenal hyperplasia.

Congenital adrenal hyperplasia (CAH) is caused by disorders of the P450c21B gene, which, with the P450c21A pseudogene, lies in the HLA locus on chromosome 6. The near identity of nucleotide sequences and endonuclease cleavage sites in these A and B loci makes genetic analysis of this disease difficult. We used a genomic DNA probe that detects the P450c21 genes (A pseudogene, 3.2 kb; B gene, 3.7 kb in Taq I digests) and the 3' flanking DNA not detected with cDNA probes (A pseudogene, 2.4 kb; B gene, 2.5 kb) to examine Southern blots of genomic DNA from 68 patients and 165 unaffected family members in 57 families with CAH. Of 116 CAH-bearing chromosomes, 114 could be sorted into five easily distinguished haplotypes based on blots of DNA digested with Taq I and Bgl II. Haplotype I (76 of 116, 65.6%) was indistinguishable from normal and therefore bore very small lesions, presumably point mutations. Haplotype II (4 of 116, 3.4%) and haplotype III (8 of 116, 6.9%) had deletions and duplications of the P450c21A pseudogene but had structurally intact P450c21B genes presumably bearing point mutations; point mutation thus was the genetic defect in 88 of 116 chromosomes (75.9%). Haplotypes IV and V lack the 3.7-kb Taq I band normally associated with the P450c21B gene. Haplotype IV (13 of 116, 11.2%) retains all other bands, indicating that the P450c21B gene has undergone a gene conversion event, so that it is now also associated with a 3.2-kb band. Haplotype V (13 of 116, 11.2%) lacks the 2.4-kb Taq I fragment and the 12-kb Bgl II fragments normally associated with the P450c21A pseudogene, as well as lacking the 3.7-kb Taq I fragment, indicating deletion of approximately 30 kb of DNA, resulting in a single hybrid P450c21A/B gene. Most (114 of 116, 98%) CAH alleles thus can easily be classified with this new probing strategy, eliminating many ambiguities resulting from probing with cDNA.

Kidney-graft survival in simultaneous kidney-pancreas transplantation.

Patient and kidney survival rates were compared between 69 diabetic patients undergoing simultaneous kidney-pancreas transplantation (group 1) and 723 nondiabetic patients undergoing kidney transplantation (group 2). The patients were treated with different immunosuppressive regimens over the years: steroids plus antilymphocyte globulin (ALG) plus azathioprine (Aza); cyclosporin A (CsA) plus ALG; steroids plus ALG plus Aza, replacing Aza 1 mo posttransplantation; or low doses of steroids plus CsA plus Aza. One-year kidney survival rates with the different regimens were 50, 42, 54, and 76%, respectively, in group 1 and 71, 74, 78, and 84%, respectively, in group 2. Patient survival was 60, 57, 71, and 86%, respectively, in group 1 and 93, 95, 94, and 96%, respectively, in group 2. Differences between the two groups were statistically significant for the first three protocols but not for the one used in this study. In group 1, 38 patients (55%) had a functioning kidney graft, whereas 15 (21%) lost their kidney to rejection. Between these two patient categories, there was no significant difference in age, sex, duration of diabetes, time on dialysis, blood transfusion number, HLA immunization, or HLA matching. Thus, since 1984, kidney-graft survival has not been inferior in diabetic patients. This improvement is mainly due to a decreased mortality related to better patient preparation and improvement in immunosuppression.

Gene conversions and rearrangements cause discordance between inheritance of forms of 21-hydroxylase deficiency and HLA types.

Congenital adrenal hyperplasia (CAH) can be caused by a variety of defects in the functional gene encoding 21-hydroxylase (P450c21), which lies in the midst of the human leukocyte antigen (HLA) locus on chromosome 6. As a result, Mendelian genetics permit clinically distinct forms of CAH to be traced genetically by HLA and complement typing of family members. The recent cloning of probes for P450c21 now permits tracing of the affected gene directly. A consanguineous family had three members affected with three clinically distinct forms of CAH. Two of these individuals had identical extended haplotypes, including nine HLA and complement loci. Despite this extensive identity, the patterns of genomic DNA fragments digested with endonuclease EcoRI and detected by a P450c21 cDNA probe differed greatly in these two individuals. Thus, the DNA diagnosis of allelic variation was much more sensitive than the HLA diagnosis. Genomic DNA digested with endonuclease TaqI and probed with P450c21 cDNA revealed the 3.2-kilobase (kb) band, which is generally associated with the nonfunctional P450c21 A pseudogene, in all family members, and also revealed the 3.7-kb band associated with the functional P450c21 B gene in all family members except the severely affected index case. Probing of the same blots with a genomic probe also permitted examination of the adjacent downstream TaqI fragments, showing retention of both the 2.4-kb (A pseudogene) and 2.5-kb (B gene) fragments. Similarly, BglII-digested genomic DNA from all individuals contained both the 12-kb (A pseudogene) and 11-kb (B gene) bands. These data indicate that the basis of 21-hydroxylase deficiency in the index case was due to a homozygous gene conversion event and not to gene deletion. These results show that the DNA in and around the 21-hydroxylase gene is genetically very active, so that the usual generalization concerning linkage and inheritance may yield incorrect conclusions and diagnoses.

Purification of HLA antigens from urine.

HLA antigens were purified from urines of kidney transplanted patients, HLA was recovered as a single peak of 45,000 mol wt that was dissociated into beta2-microglobulin and a 33,000 mol wt fraction bearing the allospecificity. The purified fractions contain carbohydrates but no lipids. Electrophoretical mobility and the relative salt concentration of eluting buffers in DEAE-Sephadex chromatography were determined for six antigens of the A locus and seven antigens of the B locus. Isolectric points of antigens A1, A2, A9, and B12 were measured. Physiochemical characteristics of HLA purified from urine appear to be similar to those of papain-solubilized cell membrane HLA. Urinary HLA was shown to originate from serum and not from renal or ureteric tissue.

The value of platelet transfusions as preparation for kidney transplantation.

The role of platelet transfusion as a preparative method for kidney transplantation is still a matter of debate. Two groups of 28 male patients transplanted between 1983 and 1988, paired for age, date of transplant, absence of anti-HLA antibody and immunosuppressive therapy have been compared. Group I was given 5 purified platelet transfusions at 1-week intervals before transplantation. Each transfusion contained 7.6 x 10(6) platelets contaminated by less than 1 leukocyte in 10(5) platelets. Group II received from 3 to 5 whole blood transfusions. In all cases it was a first transplant from cadaveric donors and previously untransfused patients before entering the protocol. No patient in group I developed cytotoxic antibodies. Acute tubular necrosis occurred with the same incidence in group I and in group II but was more severe and longer in group I, requiring hemodialysis in 62.5% and only 22% in group II. ATN was significantly associated with graft loss in group I (P less than 0.05). The total number of rejections and the number of patients undergoing rejection were not significantly different in both groups. However, the intensity of rejection was significantly higher in group I with 41% (21/51) of severe or irreversible rejections versus 9/46 (19.5%) in group II (P less than 0.05). The first rejection occurred significantly earlier in group I than in group II since 75% of the first rejection episodes occurred in the first 10 days versus 38% in group II (P less than 0.02) with a mean delay of 12.8 +/- 3.2 and 19.10 +/- 3.3 days, respectively. Although platelet transfusions are devoid of leukocytes the incidence of CMV infection was not significantly different in both groups: 57% in group I and 68% in group II. Purified platelet transfusions did not induce humoral immunization but lack of sensitization does not imply indefinite graft prolongation. Because platelets do not carry class II antigens, purified platelets transfusions represent a useful model to analyze the role of class I antigens alone in the induction of unresponsiveness in organ transplantation.

Influence of HLA-A, B, and DR matching on the outcome of kidney transplant survival in preimmunized patients.

The outcome of 893 prospectively typed (HLA-A, B, and DR) and matched cadaveric kidney transplants--all first grafts, with patients being transfused before transplantation--was studied using actuarial survival methods. The effect of HLA-A, B and DR matching was only found to be significantly beneficial to graft survival in the group of 289 presensitized recipients: 70% and 43% graft survival at two years in the case of best-matched (4-6 HLA-A, B, and DR) identities versus mismatched (0 and 1 HLA-A, B, and DR) identities, respectively (P = 0.05). Although a cumulative effect of matching for antigens belonging to the 3 HLA-A, B, and DR series was observed among the group of preimmunized recipients, a trend arose in favor of the prominent role of the HLA-B alleles. No significant difference related to HLA matching was observed in the group of nonsensitized recipients. These results confirm previous observations and support efforts to give priority for matched kidneys to preimmunized patients.

Cytomegalovirus infection--an etiological factor for rejection? A prospective study in 242 renal transplant patients.

The study aimed at analyzing the role of CMV infection as a risk factor for rejection occurring after CMV infection because of the clinical consequences of the prevention of CMV infection that might lead to the decrease in rejection episodes. Two hundred forty-two consecutive renal transplant patients were prospectively checked for the occurrence of CMV infection. CMV infection was defined virologically by a positive viremia or/and a positive viruria or/and a seroconversion or/and a significant rise of the anti-CMV antibody titers. Viremia, viruria, and serology were performed weekly for the first month and then at day 90, day 180, and every 6 months, and moreover if clinical symptoms related to a viral infection occurred. Rejection episode was defined by a creatininemia rise of 25%, after cyclosporine nephrotoxicity and urological complications had been discarded, and by the response to the antirejection therapy, steroids, or OKT3 in case of steroid-resistant rejection. The outcome factor was rejection episode occurring from day 4 after the diagnosis of CMV infection. A patient undergoing "a rejection episode after CMV infection" could also be exposed to other potential confounding factors that can be considered as risk factors of rejection among our patients. Rejection occurring before CMV infection was the main factor because it was linked both to CMV infection itself and to "rejection after." Thus infected and noninfected patients were randomly paired off. To the noninfected patient of the pair was attributed the date of a fictitious CMV infection that was the date of the CMV infection of the infected member of the pair. Therefore, "rejection after" and "rejection before" were defined in infected and noninfected patients of the pair according to the time of onset of CMV infection of the infected member of the pair. The incidence of CMV infection was 65%, 157 of the 242 patients were infected, and 85 not infected. Thus 85 pairs of infected-noninfected patients were studied. The incidence of "rejection after" the diagnosis of CMV infection was significantly higher in the group of patients with CMV infection: 45% among infected (38/85) versus 10.60% among noninfected (9/85) (P < 0.0001). Among the 85 pairs, 48 pairs were concordant in which patient of the pair evinced the same outcome factor: 43 showed no rejection after, and 5 showed one.(ABSTRACT TRUNCATED AT 400 WORDS)

Immunodeficiency diseases and expression of HLA antigens.

The role of HLA antigens in lymphocyte differentiation is strongly suggested by the existence of a recently identified immunodeficiency associated with, and probably resulting from, the lack of expression of HLA-A, B, and C antigens as well as beta2 microglobulin on various cells of hematopoietic origin. This "bare lymphocyte syndrome" has been described in a family where the transmission appeared to be autosomal recessive, and the responsible gene was not borne by the sixth chromosome. Another infant with a severe combined immunodeficiency disease has been treated with fetal liver and thymus transplants (FLTT). A persisting chimerism has been documented: T cells derived from the donor and B cells from the host. Despite complete HLA-A, B, and DR mismatch, T and B cells did cooperate resulting in significant antibody production, and defense against viral infection has been normal. Such an observation may suggest that "allogeneic restriction" of T-cell effector functions can be circumvented.

A division of HLA-DQw1 associated with DR1, DR1x-DQw1, DR2 short, DRw10, and DRw14. I. Definition by alloantiserum LY 1327.

We have identified an alloantiserum, LY 1327, directed against part of DQw1-positive cells. This split of DQw1 includes DR1, DR1x, DR2sh, DRw10, and DRw14; the other DQ-associated specificities, -DR2 long and DRw13, are unreactive. Segregation was ascertained in 11 informative Caucasoid families and in 33 genotyped individuals. DR1x refers to a specificity typing as DR blank DQw1, detected by certain anti-DR1 sera and recognized cellularly by HTC DwBON DR blank DQw1 (A. Cambon, Toulouse). Biochemical analysis by two-dimensional gel electrophoresis and DNA analysis by restriction fragment length polymorphism will be discussed since they support the existence of this division of DQw1.

A possible new HLA-DR allele.

In routine screening of anti-HLA DR reagents, serum 901 was obtained from a woman of negroid origin ten days after delivery of a second child. The spouse was also of black origin. This serum contained polyspecific HLA A and B antibodies. After platelet absorption it reacted with the B cells of 10 out of 119 European Caucasoid panel donors (8.4%) typed for DR1 to DRW10 and for MT1, MT2. Each of these ten donors had only one recognized DR antigen, the other was "blank." Serum 901 gave negative reactions with the homozygous typing cells (HTC) DW1 to DW8. In 18 normal informative Caucasoid families, serum 901 recognized one HLA haplotype in one (or both) parents and segregated with this haplotype in one or more than one child. In one family in which both parents reacted with serum 901, two children were homozygous for this marker and reacted as HTCs. One of these HTCs typed as DW9 and was found to be identical to a DW9.HTC (8W207).

HLA-DRw11, DQw7 has four cellular subtypes revealed by homozygous typing cells and undetected by restriction fragment length polymorphism.

HLA-DRw11 is in strong linkage disequilibrium with DQw7. We have investigated the heterogeneity of DRw11, DQw7 by cellular typing and by HLA cDNA probes. The results obtained by local and other homozygous typing cells in four informative Caucasoid families and 20 genotyped individuals demonstrate the existence of at least four different cellular subtypes. The frequency of DRw11, DQw7 is 25.7% with the following distribution of subtypes: Dw5 (17.3%), JAC (4.2%), JVM (2.8%), TIS (0.7%), and blank (0.7%). JVM (10W 9039) has been previously described; TIS (10W 9042) and JAC are postulated new specificities from our laboratory. None of these subtypes typed for DRw11, DQw1 cells. The cellular heterogeneity contrasts with the absence of polymorphism observed by serology and by restriction fragment length polymorphism after hybridization with DRB, DQA, and DQB probes. Variation in amino acids of the DRB1 chain has been reported for at least three variants: Dw5, JVM, and TIS (more recently).

Diversity among DRw13 DQw7 haplotypes as revealed by serology Dw typing and restriction fragment length polymorphism.

HLA-DRw13 is in linkage disequilibrium with DQw6; an unusual association, DRw13 DQw7, is found in 2% of our Caucasoid population. Investigation of genotyped individuals and of families by two allosera and by Dw typing revealed two subtypes: one recognized by homozygous typing cell HAG and by two allosera, the other subtype remained unreactive with both reagents. Analysis of DNA fragments with DNA probes indicated that the HAG-negative subset had DNA fragments in common with DRw6 while the HAG-positive subset shared DNA fragments with DR5. However, all DRw13 DQw7 cells are Dw24 as seen by hybridization with DRB probe.

Additional complexity within the HLA-D region: sequence analysis of two new DRw13-DQw7 haplotypes.

HLA-DRB1 is by far the most polymorphic locus within the HLA-D region with now well over 40 alleles. Nearly one fourth of these alleles are subtypes of DRw6, and these are in most cases undetectable by routine typing procedures. In this paper we present the molecular characterization of two new Caucasian DRw13-DQw7 haplotypes by DNA sequencing of the polymorphic first domain exons of DRB1 and DRB3 loci. The first haplotype, DRB1*1301-DRB3*0101-DQB1*0301, has arisen by a recombination between locus DRB1 from a DRw13-DQw6 haplotype and DQA1 from a DR4-DQw7 haplotype, as determined by DNA sequencing, DQ oligotyping, and restriction fragment length polymorphism typing. The second haplotype, DRB1*1305-DQB1*0301, is characterized by the novel DRB1*1305 allele differing from DRB1*1301 by three amino acids. It probably arose by a gene conversion event between a DRw13-DQw6 allele and DRB1*1101. This allele represents a DRw11/DRw13 hybrid DR molecule with a DRw13 serological epitope in the second hypervariable region and a Dw5 cellular epitope in the third hypervariable region. As determined by sequencing of locus DRB3, this allele is associated with DRw52b. Our molecular analysis of the complex HLA-DRw13 group now allows unambiguous DNA typing of all five DRw13 alleles with seven oligonucleotides, a significant improvement in the context of organ transplantation.

DNA typing of DRw6 subtypes: correlation with DRB1 and DRB3 allelic sequences by hybridization with oligonucleotide probes.

Among MHC class II antigens, the DRw6/Dw6 complex represents a special situation where typing on a routine basis is often troublesome, mainly because monospecific alloantisera are rare and individual subtypes numerous. We demonstrate here that the use of oligonucleotide DNA typing permits an analysis of the polymorphism within DRw6 haplotypes and provides a molecular basis for correlations with functional data. Synthetic oligonucleotide probes, most of them locus- and allele-specific, were derived from the DNA sequences of three alleles of locus DRB1 and three alleles of locus DRB3. These probes allow the positive identification of distinct DRw6 subtypes. As analyzed on a panel of 26 well-defined DRw6 cell lines, oligotyping allows a direct and absolute correlation with the DRw13 serologic specificity and with the cellularly defined Dw9,Dw16,Dw18, and Dw19 specificities. Correlations of the polymorphism at the DRB1 locus with the polymorphism at the DRB3 locus (DRw52 alleles) allow us to identify preferential allelic associations such as DRw13-Dw18-DRw52a/52b, DRw13-Dw19-DRw52c, and DRw13/Dw19 haplotype, the Dw19 cellular reactivity might involve, at least DRw14-Dw9-DRw52b. In view of the absolute segregation of the DRw52c allele with the DRw13/Dw19 haplotype, the Dw19 cellular reactivity might involve, at least in part, epitopes on the DRw52c allele. The identification of DRw6 subtypes, as well as of other HLA class II subspecificities, by oligotyping can now complement and possibly replace serologic and cellular typing. It represents a particularly useful contribution to the optimization of class II matching in the case of bone marrow transplantation with unrelated donors.

Both HLA-DR and HLA-DQ determinants contribute to HLA-Dw typing.

The respective contribution of HLA-DR and HLA-DQ gene products in the induction of allogeneic proliferative responses in primary mixed lymphocyte reaction and, therefore, in HLA-Dw typing, is still unclear or controversial. This is in part due to a strong linkage disequilibrium between HLA-DR and -DQ genes. We used DR- or DQ-restricted influenza-specific T-cell clones to define DR and DQ products on a large panel of allogeneic antigen presenting cells. With this functional screening assay, we identified two haplotypes with unusual DR/DQ associations. Cells of these haplotypes were then used as responder cells in mixed lymphocyte culture and stimulated by homozygous typing cells displaying DR or DQ incompatibilities. Our results indicate that DR or DQ incompatibilities alone can give rise, in both cases, to strong T-cell proliferation in a mixed lymphocyte reaction. This was further verified by blocking experiments of secondary mixed lymphocyte reactions by HLA-specific monoclonal antibodies. Anti-DQ, but not anti-DR, antibodies inhibited DQ-incompatible responses. Conversely, anti-DR, but not anti-DQ, antibodies could block DR-incompatible mixed lymphocyte reactions. Together, the results suggest that both HLA-DR and DQ gene products can be involved in HLA-Dw typing. Finally, in dual DR- and DQ-incompatible mixed lymphocyte reaction combinations, HLA-DR molecules seem to have an immunodominant effect, because the response is mostly inhibited by anti-DR antibodies. Immunodominance of HLA-DR allodeterminants may, at least in part, explain some of the controversial conclusions reported by others concerning the role of HLA-DQ molecules in HLA-Dw typing.

A study of HLA-DR2 associated HLA-Dw/LD specificities.

HLA-Dw2 and Dw12 are both associated with HLA-DR2; however, these specificities accounts for only 86% (161/188) of the DR2+ haplotypes in our North American Caucasian panel. In an attempt to identify new DR2 associated antigenic clusters, we have generated four primed lymphocyte (LD) typing (PLT) reagents in haploidentical familial combination against DR2+ Dw blank haplotypes. These reagents were positively restimulated by 11 of 16 DR2+ Dw blank cells tested, with good discrimination from Dw2 and Dw12+ cells, thus identifying a new antigenic cluster provisionally termed LD-MN2. We have compared the LD-MN2 specificity with the specificity LD-5a defined by two DR2+ HTCs, BAS and REM, (Layrisse, Caracas) which have been included in the pre-1984 Workshop Cluster DB9. Although none of our DR2+ cells gave typing responses to these two HTCs defining LD-5a, PLT studies did indicate an interrelationship between these specificities and with the specificity tb24 defined with the HTC, FJO (Betuel). The LD-5a HTCs, four LD-5a heterozygous cells, and two additional HTCs (WJR-Hansen, Seattle and FJO/tb24--Betuel, Lyon) significantly restimulated the anti-MN2 PLT reagents, though usually not as strongly as the MN2+ cells. MN2+ cells primed against the LD-5a HTCs were restimulated by only the LD-5a+ cells. Dw2+ cells primed against FJO were restimulated by some, but not all MN2+ cells. These results suggest that MN2, tb24, and LD-5a share some determinants, not shared with most cells which type as Dw2 and Dw12, though differing by other stimulatory determinants. These studies emphasize the necessity of studying new antigenic clusters by both PLT and HTC methodologies as well as testing different ethnic groups.

Sequence of a new HLA-DR4 allele with an unusual residue at position 88 that does not seem to affect T-cell allo recognition.

New HLA alleles can be identified by unorthodox patterns observed during low-resolution typing performed with sequence specific oligonucleotide probes (SSOP). One of the best examples is locus DRB1, where allelic subtypes are characterized by a combination of a limited number of residues located in three hypervariable regions of exon 2. HLA-DR oligotyping analysis of a female caucasoid bone marrow donor led to the identification of an individual that typed as DRB1*11, DRB3*02, DRB4*01, DQB1*0301-0302. This donor was, however, typed by serology as DR11 DR4, DR52, DR53, DQ7 DQ8. PCR-SSP typing for DR4 subtype revealed an amplification pattern typical for DRB1*0404. After sequencing the entire exon 2, a new DRB1 allele was identified: DRB1*04var that is identical to DRB1*0404, except for one nucleotide at codon 88 resulting in a Ser-->Arg exchange. This mutation had prevented amplification with the DR generic primers. Cellular typing by three HTCs-DRB1*0404/DW14 from the 9th Workshop showed that this DRB1*04var typed exactly like a DW14 cell. This suggests that residue 88 does not affect T cell recognition.

Generation of stable monocyte-derived dendritic cells in the presence of high concentrations of homologous or autologous serum: influence of extra-cellular pH.

Recent studies have highlighted the high degree of differentiation of monocytes. Indeed, dendritic cells (DC) can be generated from monocytes, in the presence of appropriate cytokines. However, human serum is usually avoid in such cultures. Here, we report that human serum does not inhibit generation of mature DC from blood monocytes, but rather that extra-cellular pH may play an important role in the regulation of monocyte differentiation. Indeed, monocytes cultured at pH 7.4 in the presence of high concentrations of human serum developed efficiently into mature DC, as opposed with monocytes cultured at pH 7. These pH 7.4 cultured DC presented features characteristic of mature DC, at the phenotypical, functional and morphological levels. In addition, these DC were stable, with respect to their sustained expression of CD83 and CD86, upon withdrawal of cytokines. Finally, when autologous plasma was used instead of homologous serum, differentiation of monocytes into mature DC was efficient, as well. Thus, altogether, our data show the importance of extra-cellular pH on differentiation of monocyte-derived DC in the presence of human serum, which should be maintained at plasma levels.

Apparent HLA DR triplet due to the coexpression of a DRB5-encoded molecule on a DR1 haplotype.

HLA DR1 molecules are coded by a single polymorphic DRB1 gene. We have observed rare DR1 cells in one Caucasoid family and three unrelated individuals that also reacted with some anti-DR2 sera. Since the second DR antigen was normally expressed, these cells appeared as triplets. Contrary to serology, the cells were not typed by HTCs defining Dw2, Dw12, and Dw21. Further investigations on these unusual DR1+2* haplotypes were conducted by DNA oligotyping and by sequencing of the DRB first-domain exon. The results showed that these DR1 haplotypes, besides their DRB1*0101 allele, carried also a DRB5*0101 allele.