Search

We describe here the production of a human monoclonal antibody to the HLA-DR5 antigen. A human B-cell line secreting cytotoxic antibody that reacted preferentially with DR5-positive targets was fused to the mouse myeloma P3X63Ag8.653 and the resulting heterohybridomas cloned twice. The clones secreted human IgM (lambda light chain), which showed specificity for the DR5 antigen in cytotoxicity assays and reacted with DRw11-positive but not DRw12-positive targets. These results demonstrate the potential of this approach to the production of human monoclonal antibodies to transplantation antigens.

Eleven leukaemic patients and 43 potential unrelated marrow donors were typed serologically, tested in mixed lymphocyte reaction (MLR) and typed by restriction fragment length polymorphism (RFLP) analysis. RFLP typing data were compared with DR serology and MLR results. Eleven of the 54 individuals showed discrepancies between DR serology and RFLP DR assignment. RFLP DR/DQ mismatch always correlated with positive MLR, but RFLP identity was present in both MLR negative and MLR positive pairs. RFLP typing cannot reliably predict a negative MLR response, but we suggest that it should be used in the selection of unrelated marrow donors to exclude mismatched donors from testing in the MLR. This will facilitate donor searches by reducing the number of MLR performed.

Allogeneic lymphocytes can stimulate cell-mediated cytotoxicity (CMC) in lymphocytes from leukaemia patients against autologous leukaemia target cells. We have compared the capacity of different allogeneic lymphoid cells to stimulate CMC to fresh (i.e., patient) and cultured (MOLT 4, K562) leukaemic target cells in lymphocytes from an acute leukaemic patient and his HLA-identical siblings. Allogeneic lymphoid cells, and particularly a lymphoblastoid cell line, were effective in stimulating CMC to leukaemia targets. In some instances, however, leukaemia cells derived from the patient, mixed with allogeneic lymphoid cells stimulated synergistic CMC to the patient's leukaemia. We also found that the patient's leukaemia cells alone were able to stimulate CMC in HLA-identical sib lymphocytes to fresh and cultured leukaemia targets. Extra specificities on fresh leukaemia cells were revealed when these cells induced unpredicted CMC on normal lymphocyte targets when added to mixed lymphocyte cultures (MLC) between related and unrelated lymphocytes. Cytotoxic lymphocytes generated in MLC against the patient's HLA antigens were absorbed by monolayers of lymphocytes and leukaemia cells of the same HLA type as the patient, leaving residual CMC to fresh (patient) and cultured (K562) leukaemia target cells. In addition, CMC to the patient's leukaemia cells, stimulated in lymphocytes from the patient's HLA-identical sib by allogeneic cells, was absorbed by a monolayer of these allogeneic cells. This suggests cross reactivity between determinants on the leukaemia and allogeneic lymphocytes. The results of this study are consistent with expression of 'leukaemia antigen', which are not restricted to leukaemia cells but may also be expressed on lymphocytes.

The aim of our study was to define target determinants other than those coded for by the classical HLA-A, -B, -C or -D loci which were responsible for killing in CML. In one of the families studied, strong evidence was found for the existence of a determinant coded for within the HLA region. CML was restricted to targets carrying the classical HLA-Bw35 and Cw4 determinants but the targets were neither HLA-Bw35 nor Cw4 themselves. We therefore concluded that this new HLA determinant was either the product of a new locus closely associated with HLA-B or that it was a product of the classical HLA-B locus which has not been recognized by serology.

Interleukin 2-dependent alloreactive cytotoxic T cell lines, with activity predominantly directed against the HLA-A2 antigen, have been generated in vitro by stimulating blood mononuclear cells from donors nonimmune to the Epstein-Barr (EB) virus with appropriate numbers of EB virus-transformed B cells from A2-homozygous individuals. Such effector cells were tested against a panel of EB virus-transformed target cell lines all expressing the serologically defined A2 antigen but typed into "common A2" and "variant A2" subgroups on the basis of their recognition by A2-restricted EB virus-specific cytotoxic T cells. "Variant A2" responder cells cocultivated with "common A2"-bearing stimulators gave rise to effector T cell lines which recognized only the "common A2"-bearing subgroup of targets. By contrast, responder cells from A2-negative donors stimulated with "common A2"-bearing cells produced effector T cell lines in which the strong lysis of "common A2"-bearing targets was accompanied by a lower, but still significant, lysis directed against all targets within the "variant A2" subgroup. In both cases, lysis of the target cells was blocked equally well by the anti-A2-specific monoclonal antibody MA2.1 as by the monoclonal antibody W6/32 specific for HLA-A, -B, and -C determinants. This suggests that HLA-A2 molecules possess at least two distinct sets of epitopes capable of inducing alloreactive T cell cytotoxicity: first, epitopes probably associated with T cell-restricting sites, which generate subgroup-specific responses, and second, epitopes shared by all A2 molecules, and perhaps associated with serologically defined sites, which generate "pan A2" group-specific responses.

The potential advantages of automated microscopy for reading tissue typing results are accuracy and speed. A technique which depends on fluorescence emission of both Carboxy fluorescein (CF) in live cells and Propidium Iodine (PI) in dead cells has been developed. Factors which influence the data have been investigated and appropriate mathematical corrections proposed. A practical guide is given in an appendix. In comparison with other methods double fluorescence was shown to be more accurate than single fluorescence when compared with results obtained by visual reading.

The purpose of this study was to perform a rigorous statistical analysis of the benefits of HLA-A,B and DR matching in renal transplantation. Graft survival in 2282 first cadaver kidney transplants, recorded and followed up by the United Kingdom Transplant Service (UKTS), was analyzed using the piecewise proportional hazards regression method. The results show that substantial improvements in graft survival are obtained when there is DR compatibility and at most one A or B mismatch, but that there is little advantage in tissue matching unless this degree of matching can be attained. So far, few graft recipients have benefited substantially through tissue matching (24% of kidneys exchanged through UKTS in 1984). This is partly attributable to unresolved technical problems in DR typing. However simulations show that under ideal conditions, with a pool of 3000 patients awaiting transplantation, considerable improvements in graft survival can be obtained in over 60% of recipients.

The conditions for cryopreservation and reconstitution after thawing of cytotoxic effector cells for cell mediated lympholysis (CML) tests are studied. Percentage recovery of functional activity is analysed not by the commonly used method of comparisons at a single concentration of effector cells but by the movement of the curve of functional activity on the axis of cell concentration in culture. By this method the position of the ascending slope of the dose-response curve, at several different effector to target cell ratios, is compared between fresh and frozen-thawed cells. Cooling rates giving optimal recovery were found to vary between experiments. Using simple techniques, optimal cooling rates were found to range from 0.3 to 1.0 degrees C/min, when using dimethyl sulphoxide (10% v/v). Dead cell debris, but not intact dead cells (which take up eosin), were shown to inhibit the lytic ability of functionally active frozen-thawed effector cells. This could be removed by centrifuging the thawed cells over Ficoll-Hypaque. In the recovered population the proportion of cells which excluded the vital dye, eosin, was greater than the proportion of functioning effector cells. This suggested that the cells which excluded eosin contained both functional and nonfunctional, partially damaged effector cells. Thus dye exclusion methods generally over-estimated the functional activity of thawed effector cells. When cells to be used as targets were preserved prior to treatment with phytohaemagglutinin (PHA) no abnormalities were detected in their behaviour to fresh cells.

A single enzyme/multiple probe system of HLA-DR and DQ typing using restriction fragment-length polymorphism (RFLP) analysis is presented. TaqI-digested genomic DNAs are hybridized sequentially with short DR beta, DQ beta, and DQ alpha cDNA probes. The DR beta probe discriminates between the DR allelic specificities DR1 to DRw14, with the two exceptions of some DR3/DRw13 and some DR7/DRw9 combinations. We describe the positive identification of a DRw10-specific RFLP and demonstrate its segregation in families. The DQ beta probe defines an allelic system that identifies the alleles DQw1, DQw2, and DQw3. This permits the resolution of DR3/DRw13 and DR7/DRw9 alleles by defining the DR/DQ association caused by linkage disequilibrium. The DQ alpha probe defines another allelic series interrelated with, but independent from, the DQ beta series. Specific DQ beta/DQ alpha RFLP combinations correlate with known Dw splits of DR2, DRw6, and DR7. Combined use of the three probes permits the identification of HLA-DR, DQ, and certain Dw specificities and provides an effective and easily interpretable system for major histocompatibility complex class II allogenotyping.

We have demonstrated that in more than half of the mixed lymphocyte reactions, proliferative reactions, in which responders were reacted against HLA-D-compatible homozygous typing cells, were significantly greater than zero. These reactions often masked the true HLA-D identity of the cells under test, as was subsequently shown by primed lymphocyte testing.

Eighty-three patients with rheumatoid arthritis (RA) were investigated for HLA-DQ and DR locus restriction fragment length polymorphisms (RFLP). Of the 83 patients, 61 (73%) possessed the DR4 allele and within this group we have investigated the relative frequencies of two DQ beta gene variants of DQw3, DQw7 and DQw8, one of which we had previously found to be raised in Felty's syndrome. This analysis revealed a significantly higher frequency of DQw7 containing haplotypes in DR4 positive rheumatoid patients (64%) than in DR-matched healthy controls (42%). Furthermore, the distribution of DQw7 was biased towards those patients with greater disability indicated by the HAQ score, more systemic disease and higher titres of rheumatoid factor, suggesting that DQw7 may contribute to disease expression.

The effects of 0-30% methanol (vol/vol) on the Km an Vm values for both the forward and reverse directions of the L-glutamate dehydrogenase reaction were determined at 0 degrees C. The decrease in temperature alone had very little effect on these parameters. However, in the forward reaction, 30% methanol resulted in a 14-fold decrease in the Km value for glutamate, a slight decrease in the Km value for NADP, and a thirty-fold decrease in Vm. Substrate inhibition by glutamate was observed at concentrations greater than 4 mM. In the reverse reaction, 30% methanol caused a decrease in the Km values for alpha-ketoglutarate and ammonia and a 10-fold decrease in Vm. Substrate inhibition by both alpha-ketoglutarate and NADPH was observed at concentrations of either substrate above 0.03 mM. The dependence of Km for glutamate and Vm values for the forward reaction on methanol concentration suggests that they are similarly affected by methanol, in direct contrast to results obtained for NADP. Methanol appeared to cause a general tightening of complexes, which may arise from an effect on the "activities" of species in solution. The use of methanol not only allows for the study of reaction intermediates by slowing the reaction with the cryogenic method, but may also serve as a mechanistic probe by affecting several polarity as well as Km, Vm, and K1 values.

We have demonstrated that there is a similarity in the distribution of restimulation responses of primed lymphocyte typing (PLT) cells raised between related and unrelated individuals, thus leading to the suggestion that HLA-D is a homogeneous determinant. It is proposed that within HLA a hitherto unrecognized heterogeneous structure may exist that is closely associated with HLA-D and which may give rise to intermediate reactions in primary mixed lymphocyte reactions and PLT.

1. New initiatives are required to investigate late allograft failure and its causes. 2. In the UK from 1978 to 1988 major reductions in graft loss rates occurred during the posttransplant epoch from 16-365 days but beyond 1-year failure rates show no overall changes. 3. Cyclosporine has its greatest impact during the interval from 16-365 days posttransplant and there is no evidence from these data of its impact in reducing graft losses beyond 1 year, but switching to other drugs was not taken into account. 4. The relative risk of graft loss associated with beneficial HLA matching apparently dwindles and is negligible beyond 1 year. 5. Major risk factors affecting late kidney allograft failure have yet to be modeled and evaluated.

Soluble FcR-like material can be demonstrated in the culture supernatants of activated lymphocytes. We modified as assay for its detection which was based upon inhibition of complement dependent hemolysis, an unreliable phenomenon in our hands. The micromethod presented was developed using the culture supernatants of FcR producing and FcR non-producing lymphoblastoid cell lines. Suppression of hemolysis was consistently observed providing the assay was performed below 22 degrees C. It is suggested that the classical pathway of complement activation is inhibited by this material but not the alternative pathway. By this method FcR material can be detected in small cultures where cells are limited in number.

In 53 patients with aplastic anemia (AA) a number of parameters concerning immune responsiveness were analyzed. Severe monocytopenia and B-lymphocytopenia was detected in most patients, especially in those with diffuse lymphocyte infiltration in the bone marrow. T-inducer (OKT4)/T-suppressor (OKT8) ratios were normal. The mean IgG level was significantly decreased. The frequency of specific antibodies to common viruses was comparable to that of healthy donors except for cytomegalovirus to which antibodies were less frequently found. The responding capacity in MLC was normal but patients' lymphocytes were often less stimulating than control lymphocytes. Radioresistant suppressor cells were found in 1 patient. In vitro lymphocyte reactivity to mitogens and antigens was severely impaired, related to the numbers of monocytes present. When allogeneic monocytes were added, mitogen responses could be restored.

The application of standard tissue typing techniques to cells other than peripheral blood lymphocytes has been accompanied by the problem of extra reactions. This applies as well to Epstein-Barr virus transformed lymphoblastoid cell lines (LCL) as to leukemic cells and human spleen cells. These extra reactions are attributable to additional antibodies in the typing sera which are not apparent under standard conditions with PBLs. Two types are described: Type 1 extras, which becomes apparent after longer incubation times and are attributed to weak antibodies and type 2 extras which are apparent after shorter incubation times and are attributed to subpopulation specific or differentiation antigens. Technical modifications are proposed by which these extras can be circumvented. They include: Only start typing when cells have been cultured for 2 to 3 days. Remove dead cells by spinning over standard ficoll-hypaque or 11% triosil. Use shorter incubation times. Avoid using sera that give too many type 2 extras. In this way phenotypes can be accurately identified on LCL's obtained from kidney transplant donors and recipients. When LCL's were compared with their matching PBL, HLA phenotypes were concordant in 87% of cases for HLA-A, 90% for HLA-B, 81% for HLA-C and 70% for HLA-DR.

Epstein-Barr (EB) virus-specific cytotoxic T-cell preparations, produced by stimulation in vitro of peripheral blood lymphocytes with the autologous virus-transformed cell line, are HLA-A and B antigen-restricted and, with some donors, show preferential restriction through one or two of the four relevant antigens of the donor's HLA type. It has now been demonstrated that such EB virus-specific cytotoxic T cells may also be reactivated by stimulation with allogeneic virus-transformed cells provided that there is no mismatch of the HLA-A and B antigens between the responder and stimulator cell donors. In particular, virus-transformed cell lines from HLA-homozygous donors HLA-A and B antigen-matched to one of the haplotypes of an HLA-heterozygous responder were shown to reactivate selectively only those EB virus-specific cytotoxic T cells restricted through the HLA-A and B antigens present on the allogeneic stimulating cells. In addition to confirming the polyclonal nature of the HLA-restricted EB virus-specific cytotoxic T-cell response, this new experimental procedure has allowed the production, and subsequent expansion as cell lines dependent upon T-cell growth factor, of those effector cells restricted through the "nonpreferred" HLA antigens that are poorly represented in the response induced by stimulation with autologous virus-transformed cells.

Thirty-four IgG anti-D human monoclonal antibodies (mAb) derived from 18 donor were assessed for their ability to mediate lysis of D+ red cells by lymphocytes in antibody-dependent cell-mediated cytotoxicity assays. Cell-bound antibody was quantified and the mAb were compared at similar levels of sensitization. The majority (23/31) of IgG1 and all (3/3) IgG3 mAb were ineffective; two donors produced both lytic and non-lytic anti-D mAb. Greater sensitivity was achieved using fluid-phase antibody (as culture supernatants) in the assay than was obtained with pre-sensitized red cells. Minimum levels of 2000 anti-D molecules per cell were required for lysis using pre-sensitized cells. Partial D red cells (DIVa, DVa and DVI) were lysed by three mAb that were lytic with normal D+ cells. There was no relationship between lytic ability and Gm allotype or D epitope specificity of the antibodies. Four mAb to other blood group specificities were tested: two (anti-E and anti-G) were lytic and two (anti-c and anti-Kell) were not lytic. Possible reasons for the heterogeneity of the lytic activity by the mAb are discussed.

Stratified proportional hazards regression is described as a method of estimating multifactorially preoperative factor effects on graft survival--and, at the same time, making due allowances for unknown transplant-center-specific influences. The multifactorial aspect of the method overcomes the biases inherent in analyzing transplant survival data one factor at a time, and stratification allows the data from many centers to be used simultaneously without the dangers associated with simple pooling of data from many sources. The proportionality and regression assumptions implicit in the method enable the data to be used efficiently, but must be validated on the data. Methods by which these assumptions may be relaxed are described--in particular, the stratified piece-wise proportional hazards regression method.

In a population of 282 Caucasoid patients in end-stage renal failure (ESRF) Gm and Km typing has confirmed a significant increase in the frequency of the Gm1,2,17;21 haplotype. This was confined to an increase in the number of heterozygous Gm(1,2,3,17;21,5,10,13,14) individuals. A concomitant decrease in the frequency of presumptive Gm(3,5,10,13,14) heterozygotes was also confirmed. Analysis of individual renal diseases revealed significant immunoglobulin allotype, phenotype and haplotype frequency deviations in those patients presenting with hypertension, glomerulonephritis or pyelonephritis. The allotype distribution in patients with physical abnormalities of the urinary-genital tract, whether congenital or acquired, was normal.

1. The risks associated with nonbeneficial matching constitute a persistent hazard to the transplant over a long period. 2. The benefits of HLA matching and CsA are separate and additive. 3. There is no evidence that beneficial matching is not relevant to all (UK) centers. 4. Contemporary typing methods are expected to improve the accuracy of typing and reduce kidney cold-ischemia times. 5. Organ sharing increases the numbers of beneficial matches. 6. Patients with easily matchable ABO-HLA phenotypes should wait for beneficially matched transplants. 7. Patients with phenotypes that are difficult to match should be intelligently mismatched after due consideration of responder status, acceptable mismatches, and cross-reactive groups. 8. Organ sharing does not prejudice graft survival. 9. Wastage is minimized with a central clearing house. 10. Organ sharing reduces the incidence of high sensitization. 11. Organ sharing is cost effective. For these reasons we repudiate the view that organ sharing is now superfluous.

The exon 2 nucleotide sequences of HLA-DQwl-associated and DQw3-associated HLA-DR"Br" alleles were determined from genomic DNA amplified by the Taq polymerase chain reaction technique. Both alleles reveal identical exon 2 nucleotide sequences. Comparison with other DR alleles suggests that DR"Br" may have originated from DR1 by gene conversion with DR4-Dw10, DR5, or DRw6-Dw18 third hypervariable region sequences.

A monolayer absorption technique was used to test the hypothesis that killer cells directed to self HLA-associated minor histocompatibility antigens (H-Y) were divisible into subsets. The results showed that sensitized killer cells, which recognized two combined antigens HLA-A2; H-Y and HLA-B7; H-Y could indeed be divided into two populations. One was directed to HLA-A2; H-Y and the other to HLA-B7; H-Y. These results can be interpreted in the context of the altered self hypothesis. However, when interpreted in the context of the dual recognition hypothesis, they strongly suggest that independant clones of killer T cells exist which are committed to the recognition of self HLA.

A PLT cell was raised between a responder cell which carried the HLA-D and DR specificities Dw3, 8; DRw3, 8 and a stimulator cell which was most likely homozygous for HLA-Dw3, DRw3. The PLT cell appeared to recognize a determinant (PL3A) which was (1) different from the officially recognized HLA-D and DR determinants, (2) was associated with HLA-A1, B8, Dw3 and DRw3, and (3) segregated in three informative families with HLA. Another responder-stimulator combination could be selected to raise PLT cells which recognized the same determinant.

At least two different methods using cellular responses have been described for defining the determinants of the HLA-D region: typing with HLA-D homozygous cells and primed LD typing. Primed LD typing cells were generated in one-haplotype-different combinations and grouped on the basis of two or more cells appearing to define the same HLA-D-region-determined PL antigen. Such cells were restimulated with homozygous typing cells for several of the presently known HLA DW clusters. A very strong correlation was noted: PLT cells defining the antigen PL1 were restimulated with homozygous typing cells for DW3, those PLT cells defining the antigen PL2 were restimulated by homozygous typing cells for DW2, and those defining PL5 were restimulated by homozygous typing cells for DW1.

Our next quinquennial plan for transplantation studies extends MPI to corneal and unrelated marrow transplantation. It applies Bayesian hierarchical modelling to regional variation in donor procurement and continues a program of special studies to augment national databases with respect to kidney, corneal, heart and liver transplantation. It promotes research collaboration among European and other organ exchange organizations as pioneered in the Council of Europe 1986 Study on High Sensitization, which showed the effectiveness of the network of European organ exchange organizations in liaising with transplant units.

HLA-DR1 and DR 'Br' allogenotype patterns, generated using several restriction endonucleases and visualised using four HLA-DR beta cDNA probes in Southern analysis, are indistinguishable. We suggest that HLA-DR 'Br' may be a variant of the HLA-DR1 allele.

An analysis of pooled data from transplants performed between 1977 and 1981 in 29 centers throughout the United Kingdom and Ireland revealed that the pattern of loss varied according to cause and postoperative time. Loss from rejection was characterized by a bimodal pattern in which early (0-25 days) and late (26-100 days) peaks of rejection were distinguishable. Rejected second transplants exhibited this phenomenon more than first transplants, and the loss was proportionately greater during the early period, suggesting that prior sensitization played an important role. Graft loss from technical causes and recipient death showed distinctly different patterns of loss. These findings suggested that, when possible, transplant survival statistics should be analyzed separately according to postoperative time and cause of loss. In applying these preliminary observations of the pooled data to a comparative study of the results in the different centers it was noted that such comparisons could be substantially affected by random variability in estimates of actuarial survival rates. Therefore, a simple method of ranking was developed in which centers were allocated to high or low survivorship categories, or to an indeterminate category when the standard error in estimated actuarial survival was relatively large. Whereas the variation in loss rate from death with a functioning transplant (DWFT) was found to be indistinguishable from random variability, both nonimmunological failure (NIF) and immunological failure (IF) of the graft were found to be legitimate bases for ranking. Furthermore, center ranking based on IF at 0-25 days failed to exhibit a significant relationship with IF at 26-100 days, which could indicate important center differences associated with antirejection treatments during these two periods. These results showed that, ideally, time-cause parameters should be analyzed separately when comparing transplant survival statistics in different centers.

Eight monoclonal antibodies (mabs) to the Rh antigen D produced by Epstein-Barr virus transformed B-lymphoblastoid cell lines from two individuals have been compared for their behaviour in in vitro cell-mediated assays. Three IgG1 Glm(1,17) and two IgG3 G3m(21) mabs from one donor and three IgG1 Glm(3) mabs from another were used. IgG3 anti-D mabs induced greater adherence and phagocytosis of sensitized red cells by U937 monocytes than IgG1 anti-D mabs or the polyclonal anti-D. Minimum sensitization levels for rosetting and phagocytosis by U937 monocytes were 2,000 molecules IgG/cell for IgG3 and 5,000 molecules/cell for IgG1 mabs; maximum rosetting mediated by both IgG1 and IgG3 mabs was obtained at 15,000-20,000 molecules/cell. The IgG3 anti-D mabs were comparable to polyclonal anti-D in mediating binding of sensitized red cells to gamma-interferon stimulated monocyte-derived cultured macrophages and were markedly more effective than the IgG1 anti-D mabs. However, in lymphocyte ADCC assays, only anti-D mabs which were IgG1 Glm(3) were effective in mediating high levels of lysis of sensitized red cells, unlike the IgG1 Glm (1,17) or IgG3 G3m(21) mabs. Minimum sensitization levels required for this lymphocyte-mediated red cell lysis were found to be approximately 5,000 molecules/cell with one IgG1 Glm(3) mab; maximum lysis with this mab was obtained at 10,000 molecules/cell. Polyclonal anti-D containing both IgG1 and IgG3 was effective in all three assays. These observations suggest that different isotypes and allotypes of anti-D antibodies mediate red cell removed or destruction by monocyte or lymphocyte effector cell through functionally dissimilar Fc receptor interactions.

A group of 43 Australian Aborigines was typed for HLA--DR determinants and other antigens on B cells, using pregnancy sera from Caucasoid women. The group could not be considered as random, since a selection was made for an equal number of Au antigen positive and Au antigen negative individuals The typing results showed that the specificities DRw2, 4 and 8 were as well defined in this population as in Caucasoids, whereas the specificities DRw1, 3, 5 and 7 could not be defined. Even more interesting was the finding that a cluster of six sera, defining the local B cell specificity LB12, gave positive reactions in 41 out of 43 Aborigines. Although LB12 is found to be highly associated with DRw1, 2 and 6 in the Caucasoid population, in 36 cases the LB12 positive Aborigines could not be demonstrated to carry the specificities DRw1 or 2. The recognition of DRw6 is difficult at the moment and for the reason its presence or absence in this material is uncertain. However, three individuals typed positive for DRw4 and 8 and for LB12 as well. These findings strongly support the hypothesis that the genes coding for LB12 and its postulated allele LB13 are separate from the genes coding for the known DR determinants. LB12 is very likely identical to the recently described determinants DC1 (Tosi & Tanigaki 1979) and MB1 (Duquesnoy et al. 1979a).

HLA-DR and -DQ serotyped cell lines and peripheral blood leucocytes were analysed by Southern blot allogenotyping. Using a short DQ beta cDNA probe, a DQ beta allelic series was defined by restriction fragment length polymorphism (RFLP) with the restriction endonuclease TaqI. This DQ beta allelic series correlates with, and defines splits of, the HLA-DQ serological specificities DQw1 (DQ beta 1a and DQ beta 1b RFLPs), DQw2 (DQ beta 2a and DQ beta 2b RFLPs) and DQw3 (DQ beta 3a and DQ beta 3b RFLPs). By sequential use of a short DQ alpha cDNA probe a second, DQ alpha allelic series is defined by RFLP. This series correlates to a lesser extent than DQ beta RFLPs with the HLA-DQ serological specificities. Thus, two DQ alpha RFLPs correlate with a single DQ serotype (DQ alpha 1a and DQ alpha 1c with DQw1), but three DQ alpha RFLPs correlate with more than one DQ serotype (DQ alpha 1b with DQw1 and DQw3; DQ alpha 2 with DQw2 and DQw3; DQ alpha 3 with DQw2 and DQw3). Individual DQ beta and DQ alpha RFLP subtypes appear to correlate with single, or associated HLA-DR specificities. Specific combinations of DQ beta with DQ alpha RFLPs also correlate with HLA-Dw splits of DR2 and DRw6. A system for HLA-DNA typing is described, which uses RFLP patterns generated by sequential hybridization of TaqI-digested DNAs with short DR beta, DQ beta and DQ alpha cDNA probes. The DQ beta and DQ alpha probes not only identify the DQ allele, but because of linkage disequilibrium with DR, help to assign the DR allele, which may not always be identified with a DR beta probe alone.