Nomenclature 1991 Foreword Julia G. Bodmer
THE RATIONALE A generally accepted and universally used system of nomenclature is an essential tool in any area of study. This is particularly true in an area such as that of the human histocompatibility system (HLA), which consists of a cluster of genes, 35 so far officially named, on c h r o m o s o m e 6, which displays a level of polymorphism, 250 officially named alleles, unmatched so far by any other known human system. T h e World Health Organization ( W H O ) N o m e n c l a t u r e C o m m i t t e e for HLA, the only body charged with assigning names, has met since 1967 after every International Histocompatibility Workshop, and recently in between Workshops, to decide on nomenclature for new genes, alleles, and antigens. These assignments have been based on widely verifiable standards and agreed criteria since the reagents used to define the alleles have been available to the large n u m b e r of laboratories participating in the international workshops. In the early days, antigen names were assigned on the basis of patterns of reaction with widely available antisera. The loci were defined and the alleles were assigned to them based on segregation in families. Incidentally, the reason that alleles at the H L A - A and B loci share a single set of numbers, rather than each having their own as for other loci, is that when the first specificities were named it was not clear that there were two loci rather than only one. Thus, HLA-A1, A2, A3, A9, B5, B7, and so on, were named as antigens, products of alleles at two linked loci. The first problems arose when it was clear that an antigen such as A9 was not a single specificity, but included at least two different specificities with slightly different serological patterns and different ethnic distributions. The decision was made that the next available numbers would be assigned to these "splits" of A9 and they were accordingly assigned Aw23 and Aw24 (now A23 and A24), indicating a finer or more informative typing. The term A9 continues to be used as before 2 0198-8859/92/$5.00
when it has not been possible to identify the subtype. Another p r o b l e m was the question of the nature of Bw4 and Bw6. Were all B-locus antigens "splits" of Bw4 and Bw6? Some people decided to ignore them completely as they could now give a m o r e precise typing. It is only now that we have sequence data that we see that Bw4 and Bw6 are epitopes present on some B locus and, in the case of Bw4, both some B-locus and some A-locus alleles. The HLA-A, B, and C loci are what are also called class I genes and are found on most nucleated cells. However, products of another gene or series of genes, called class II, were found by mixed lymphocyte culture on a m o r e limited set of cells and these were given names starting with D, which was the next available letter. In practice, all H L A - D specificities have retained the W o r k s h o p or provisional suffix "w" as there was no unambiguous individual locus to which the "Dw" types identified by mixed lymphocyte stimulation could be assigned. It was then shown that these products could be identified serologically on cell subsets and the first antigens so defined were called DR, meaning "D related." As other genes were found in this region, they were called D Q , DP, and so on, DP having previously been identified by primed lymphocyte typing. It became clear that D Q has polymorphism on both the c~ and the/~ chains, so both have to be taken into account. With the advent of molecular typing, even m o r e complexity has to be dealt with for the definition of an H L A type. There are not one, but at least five expressed D R B chains giving different specificities. N o t all D R B chains are expressed in every haplotype and combinations of two will vary with each D R type. Furthermore, for each antigen or type defined serologically, it became apparent that there were two or sometimes several alternative D N A sequences that, in the cases tested, would stimulate a Tcell response and should be noted for the purposes of a complete H L A typing. Human Immunology 34, 2-3 (1992) © American Society for Histocompatibility and Immunogenetics, 1992
Nomenclature 1991 Foreword
It is, therefore, impossible to define simply a system that by its nature is complicated and whose complexity is important in defining, for example, a good match for transplantation or a person at increased risk of a particular disease. Molecularly defined alleles were named first in the 1987 nomenclature report, and in this report a very large n u m b e r of these alleles have been named. The four digits reserved for the allele names are needed to avoid ambiguity, the first two digits correlating with the previously assigned serological name and the last two for any n u m b e r of variants that are found. For example, D R B l * 0 4 1 2 indicates the allele sequence for the 12th variant of the D R 4 specificity. Since it has been shown that a single base-pair difference in sequence can generate a T-cell response, and in rodents to cause p r o m p t graft rejection, it is important to be able to identify these minor differences. Occasionally a fifth digit is added to indicate the presence of a noncoding substitution, e.g., DRB1*11042. This fifth digit indicates a mutation that is unlikely to cause any immunological response and can be practically ignored. T h e * was inserted to separate the locus number, e.g., D R B 1 , from the allele number. In this report, for the first time, serological specificities have been named on the basis of serological identification of the product of a known D N A sequence. Thus, a "short" HLA-A2 serological specificity, identified on a cell with an A-locus sequence of HLA-A*0203, was named H L A - A 2 0 3 in order to maintain the correspondence between serological and molecular definition. In one or two cases, there were anomalies when the sequence similarity indicated naming it with one allele and the commonly found serological specificity would align it with another, for example, in the case of B N 2 1 , named B ' 4 0 0 5 . In this case, the sequence similarity has preference since the B ' 4 0 0 5 has only two amino acid differences from B ' 4 0 0 2 . It has 15 differences from B ' 5 0 0 1
3
and 20 differences from B ' 4 9 0 1 , which are the two sequences associated with B21. THE ARGUMENT
Why does the Nomenclature Committee keep changing the nomenclature? This is a question that is often asked; but as can be seen from the above, from the report published in this issue, and if you wish to look, from all the previous reports of the N o m e n c l a t u r e Committee, the names have not changed. The original HLA-A1, A2, and so on, are still there. It is just that we have been able to define splits and subsets of these, often with very different ethnic frequencies, and it is to these that we have had to assign names. We have identified at the most basic sequence level the nucleotide differences between these variants and named them in a way to keep their connection with the serologically defined names. W h e r e there are new serological specificities, we have named them to match their defined D N A sequence. The two sets of names go hand in hand. It is also asked, "Which are the strong immunogens and against what background?" It is only by identifying matches and mismatches as specifically as possible, as can best be done by molecular typing, that we will be able to analyze the data to discover which are the important and which are the negligible mismatches. H o w can the busy clinician deal with this? The first question is, should she need to? Does she expect the tissue-typing laboratory to understand the finer points of the surgery planned? The grade of match, on an agreed scale of goodness of match, provided by the tissue-typing laboratory experts, based on their knowledge of differences of sequence and, increasingly, the immunogenicity of the observed differences, should provide the guidance needed by the clinician. Matching is a complex science and perhaps should be left to the specialists.
THE RATIONALE A generally accepted and universally used system of nomenclature is an essential tool in any area of study. This is particularly true in an area such as that of the human histocompatibility system (HLA), which consists of a cluster of genes, 35 so far officially named, on c h r o m o s o m e 6, which displays a level of polymorphism, 250 officially named alleles, unmatched so far by any other known human system. T h e World Health Organization ( W H O ) N o m e n c l a t u r e C o m m i t t e e for HLA, the only body charged with assigning names, has met since 1967 after every International Histocompatibility Workshop, and recently in between Workshops, to decide on nomenclature for new genes, alleles, and antigens. These assignments have been based on widely verifiable standards and agreed criteria since the reagents used to define the alleles have been available to the large n u m b e r of laboratories participating in the international workshops. In the early days, antigen names were assigned on the basis of patterns of reaction with widely available antisera. The loci were defined and the alleles were assigned to them based on segregation in families. Incidentally, the reason that alleles at the H L A - A and B loci share a single set of numbers, rather than each having their own as for other loci, is that when the first specificities were named it was not clear that there were two loci rather than only one. Thus, HLA-A1, A2, A3, A9, B5, B7, and so on, were named as antigens, products of alleles at two linked loci. The first problems arose when it was clear that an antigen such as A9 was not a single specificity, but included at least two different specificities with slightly different serological patterns and different ethnic distributions. The decision was made that the next available numbers would be assigned to these "splits" of A9 and they were accordingly assigned Aw23 and Aw24 (now A23 and A24), indicating a finer or more informative typing. The term A9 continues to be used as before 2 0198-8859/92/$5.00
when it has not been possible to identify the subtype. Another p r o b l e m was the question of the nature of Bw4 and Bw6. Were all B-locus antigens "splits" of Bw4 and Bw6? Some people decided to ignore them completely as they could now give a m o r e precise typing. It is only now that we have sequence data that we see that Bw4 and Bw6 are epitopes present on some B locus and, in the case of Bw4, both some B-locus and some A-locus alleles. The HLA-A, B, and C loci are what are also called class I genes and are found on most nucleated cells. However, products of another gene or series of genes, called class II, were found by mixed lymphocyte culture on a m o r e limited set of cells and these were given names starting with D, which was the next available letter. In practice, all H L A - D specificities have retained the W o r k s h o p or provisional suffix "w" as there was no unambiguous individual locus to which the "Dw" types identified by mixed lymphocyte stimulation could be assigned. It was then shown that these products could be identified serologically on cell subsets and the first antigens so defined were called DR, meaning "D related." As other genes were found in this region, they were called D Q , DP, and so on, DP having previously been identified by primed lymphocyte typing. It became clear that D Q has polymorphism on both the c~ and the/~ chains, so both have to be taken into account. With the advent of molecular typing, even m o r e complexity has to be dealt with for the definition of an H L A type. There are not one, but at least five expressed D R B chains giving different specificities. N o t all D R B chains are expressed in every haplotype and combinations of two will vary with each D R type. Furthermore, for each antigen or type defined serologically, it became apparent that there were two or sometimes several alternative D N A sequences that, in the cases tested, would stimulate a Tcell response and should be noted for the purposes of a complete H L A typing. Human Immunology 34, 2-3 (1992) © American Society for Histocompatibility and Immunogenetics, 1992
Nomenclature 1991 Foreword
It is, therefore, impossible to define simply a system that by its nature is complicated and whose complexity is important in defining, for example, a good match for transplantation or a person at increased risk of a particular disease. Molecularly defined alleles were named first in the 1987 nomenclature report, and in this report a very large n u m b e r of these alleles have been named. The four digits reserved for the allele names are needed to avoid ambiguity, the first two digits correlating with the previously assigned serological name and the last two for any n u m b e r of variants that are found. For example, D R B l * 0 4 1 2 indicates the allele sequence for the 12th variant of the D R 4 specificity. Since it has been shown that a single base-pair difference in sequence can generate a T-cell response, and in rodents to cause p r o m p t graft rejection, it is important to be able to identify these minor differences. Occasionally a fifth digit is added to indicate the presence of a noncoding substitution, e.g., DRB1*11042. This fifth digit indicates a mutation that is unlikely to cause any immunological response and can be practically ignored. T h e * was inserted to separate the locus number, e.g., D R B 1 , from the allele number. In this report, for the first time, serological specificities have been named on the basis of serological identification of the product of a known D N A sequence. Thus, a "short" HLA-A2 serological specificity, identified on a cell with an A-locus sequence of HLA-A*0203, was named H L A - A 2 0 3 in order to maintain the correspondence between serological and molecular definition. In one or two cases, there were anomalies when the sequence similarity indicated naming it with one allele and the commonly found serological specificity would align it with another, for example, in the case of B N 2 1 , named B ' 4 0 0 5 . In this case, the sequence similarity has preference since the B ' 4 0 0 5 has only two amino acid differences from B ' 4 0 0 2 . It has 15 differences from B ' 5 0 0 1
3
and 20 differences from B ' 4 9 0 1 , which are the two sequences associated with B21. THE ARGUMENT
Why does the Nomenclature Committee keep changing the nomenclature? This is a question that is often asked; but as can be seen from the above, from the report published in this issue, and if you wish to look, from all the previous reports of the N o m e n c l a t u r e Committee, the names have not changed. The original HLA-A1, A2, and so on, are still there. It is just that we have been able to define splits and subsets of these, often with very different ethnic frequencies, and it is to these that we have had to assign names. We have identified at the most basic sequence level the nucleotide differences between these variants and named them in a way to keep their connection with the serologically defined names. W h e r e there are new serological specificities, we have named them to match their defined D N A sequence. The two sets of names go hand in hand. It is also asked, "Which are the strong immunogens and against what background?" It is only by identifying matches and mismatches as specifically as possible, as can best be done by molecular typing, that we will be able to analyze the data to discover which are the important and which are the negligible mismatches. H o w can the busy clinician deal with this? The first question is, should she need to? Does she expect the tissue-typing laboratory to understand the finer points of the surgery planned? The grade of match, on an agreed scale of goodness of match, provided by the tissue-typing laboratory experts, based on their knowledge of differences of sequence and, increasingly, the immunogenicity of the observed differences, should provide the guidance needed by the clinician. Matching is a complex science and perhaps should be left to the specialists.
