FIELD OF THE INVENTION

The present invention relates to the field of human leukocyte antigen (HLA) typing, and to its application to tissue matching for transplantation.

BACKGROUND TO THE INVENTION

The human leukocyte antigen (HLA) genes, and the alleles encoded by these genes, are a major focus of tissue typing for organ and tissue transplantation. The HLA antigen complex, better known as the human Major Histocompatibility Complex (MHC) is divisible into 3 separate regions, identified as the Class I, Class II and Class III genes, and spans approximately 3.5 million base pairs on the short arm of chromosome 6 (6p21.3). The MHC is a highly polymorphic region containing about 200 genes and contains a set of genes that serve as the backbone of antigen presentation. The HLA genes encompass the most diverse antigenic system in the human genome, encoding literally hundreds of alleles that fall into several distinct subgroups or subfamilies.

The Class I proteins, classically involved in presenting endogenous antigens to CD 8+ T- cells, are expressed by genes located in the HLA-A, -B and C loci. In contrast, the Class II proteins, which associate with and present exogenous antigens to CD4+ T-cells, are encoded in the HLA-D region at the HLA— DR, -DQ and -DP loci. Each locus is highly polymorphic; for example, there are more than 500 alleles in the HLA-A, -B and -DRB 1 loci. A clear understanding of the differences between HLA polymorphisms has provided an insight into why and how foreign tissue is rejected by the host, and as such been a critical enabler of the field of transplantation. Today, a variety of techniques are applied for HLA tissue typing, providing an important tool to increase the success rate of human transplants.

In addition, within the Class I region there are nonclassical Class I genes including, but not limited to, HLA-E, HLA-F, HLA-G, HLA-H, HLA-J and HLA-X. HLA-A and HLA-C comprise eight exons and seven introns, whereas HLA-B comprises seven exons and six introns. The DNA sequences of the HLA-A, -B and -C are, generally speaking. highly conserved. Allelic variation, however, occurs predominantly in exons 2 and 3, which encode the functional domains of the molecules and are flanked by introns 1, 2 and 3.

The allocation of organs in tissue and organ transplantation (e.g. renal, pancreas or heart transplantation) is influenced by two main factors: a) the requirement to human leukocyte antigen (HLA) match as closely as possible, taking into account the constraints of a small donor pool, and the need to minimise cold ischaemic time; b) the requirement to avoid transplanting patients with donor organs which feature an HLA antigen to which the patient has raised an antibody.

For many years, HLA polymorphisms have been typed by serological responses to HLA antigens. Sera containing antibodies to Class I and II proteins were collected from multiparous women, or individuals who had received multiple blood transfusions. In addition, polymorphisms in Class II proteins were analyzed by T-cell responses in the Mixed Lymphocyte Reaction (MLR). However, the advent of recombinant DNA technology, which paved the way to identifying genetic differences among the HLA loci directly, has led some laboratories to abandon classic serological typing methods. Today, several different types of HLA DNA typing methods exist and are commonly used by clinical laboratories.

In one method, e.g. as described in WO 92/1571 1, HLA genotypes are determined at the molecular level by direct sequencing of DRB, DQB, DQA, DPB, DPA, HLA- A, HLA-B and HLA-C transcripts enzymatically amplified using a limited number of non-allele- specific oligonucleotides. DNA samples are enzymatically amplified using different combinations of oligonucleotides for each locus and directly sequenced with Taq polymerase using an internal oligonucleotide. Many other variations of such DNA- sequencing based HLA typing methods exist, for example see WO 97/23650.

In other methods, HLA DNA can be typed either by hybridizing labeled, sequence- specific oligonucleotide probes to HLA loci amplified by the polymerase chain reaction (PGR), or by using PGR to amplify the HLA DNA directly through differential primer extension.

Renal transplantation is regarded as the treatment of choice for patients with end-stage renal failure. Donor HLA specific antibodies (DIISA) are regarded as a contra-indication to successful renal transplantation, causing hyperacute rejection (Patel & Terasaki (1969) N. Engl. J. Med 280:735-9). Depending on their breadth of specificity these antibodies seriously impede the selection of suitable donors leading to increased waiting time on dialysis with associated morbidity and quality of life issues. Although the liver can survive quite high levels of DHSA, HLA antibodies to donor antigens are also a contraindication to most other solid organ transplants.

Sensitive and specific methods are available for the detection of antibodies to the extreme diversity of HLA antigens. The most sensitive methods are based on purified or recombinant HLA molecules attached to microbeads suitable for flow cytometric analysis, with Luminex ¹ ™ based methods being most popular. This has taken antibody detection to the molecular level. In some cases, the epitopes to which these antibodies bind can be derived from these studies. For instance, El-Awar et al. (2007), Transplantation 84:532-540, recognise 103 of these epitopes for HLA Class 1.

Unfortunately reporting methods and the commercial software for defining antibody specificity requires a list of HLA alleles to which the antibodies bind. This is both cumbersome and inaccurate - to take two typical examples:

1) An antibody detected to the serological antigens HLA-B7, 8, 45, 64, 65, 62, 75, 76, 39, 50, 54, 55, 56, 35, 60, 61, 41, 42, 46, 48, 67, 71, 72, 73 is binding the discrete public epitope Bw6 which is defined by the amino acid asparagine at position 80 of the alpha chain of Class I (80 ). However the above list is by no means accurate as there are 34 amino-acid variants of HLA-B8 alone (HLA-B*0801-34) and HLA-B*0802 and 0803 do not have asparagine at position 80 and are therefore Bw6 negative.

2) A further illustration of the complexity of antibody analysis is that epitopes are shared between alleles at more than one locus. For example, an antibody found to bind HLA-A2, 25, 26, 29, 31, 32, 33, 34, 43, 66, 68, 69, 74, B73, Cw7 and 17 (this list would be much longer using the standard amino-acid level four digit HLA nomenclature) is simply binding glutamine at position 253 of the alpha chain of HLA Classl (253Q).

Thus although HLA antibodies bind discrete clusters of amino-acids (i.e. epitopes) currently definition of these antibodies relies on a list of alleles to which the antibody binds, with no reference to the epitope(s). However, as HLA typing for solid organ transplantation is provided at only medium resolution, this means the list is often inaccurate. In 2007 in the UK 7% of deceased donor ^■ kidneys were transported to a recipient who had anti-donor antibodies, resulting in increased cold ischaemic damage to the donor organ due to prolonged storage on ice while a suitable recipient was identified.

The problem is two-fold: we are using outdated nomenclature to define antibodies, and we are not defining HLA antigens to an adequate level. One possible alternative would be to type HLA alleles to the four digit amino-acid level which would require nucleotide sequencing of all HLA loci. This would be very expensive, generate unnecessary data and take about 10 hours per sample. This is not compatible with acute deceased donor transplantation where a medium resolution HLA type takes under four hours from specimen receipt to result.

HLAMatchmaker is one approach which has been described for identifying compatible HLA types for highly allosensitized patients without the need for extensive serum screening (see e.g. Duquesnoy and Marrari (2002), Human Immunology 63:353-363 and Duquesnoy and Marrari (2007), Human Immunology 68: 12-25. This method is based on a computer algorithm which breaks each HLA antigen down into a string of triplet- defined epitopes or more recently spacially arranged amino-acid groups called eplets that have the potential of inducing humoral immune responses. The method relies on the premise that allosensitized patients cannot produce antibodies to eplets on mismatched HLA antigens if such eplets are present in the same sequence location of any of the patient's own HLA molecules. Thus HLAMatchmaker can determine that certain mismatches are acceptable and are unlikely to be immunogenic. A problem with the HLAMatchmaker method is that it still relies on the standard HLA classification system. Thus the input information is a set of serologically defined HLA types, and eplets are assigned to these HLA types based on amino acid sequence information of what is estimated to be the most common molecularly defined allele of each serologically defined HLA antigen. Therefore the eplet assignments are not always accurate. The HLAMatchmaker method could be improved by inputting the HLA allele to the four digit amino-acid level, based on DNA-typing data, but as mentioned above this is not practical in most situations. Moreover, the HLAMatchmaker approach relies on providing a complete amino acid sequence for each HLA antigen in both donor and recipient, whether this is estimated from a serological determination or based more accurately on DNA sequencing, and dividing this complete sequence into triplets. The method fails to focus on the specific regions of HLA alleles in a donor which may actually interact with DHSA in a recipient, and does not allow for matching based on defining the true specificity of the DHSA which are present in the recipient.

Accordingly, there is still a need for an HLA typing method which overcomes the problems associated with the above methods. In particular, there is a need for a method which is accurate and can discriminate between the molecularly defined alleles within a serologically-defined antigen type, whilst being quick and reducing the amount of DNA sequencing required. Moreover, there is a need for a method which can accurately determine whether a recipient has pre-formed antibodies to an HLA type found in the recipient.

SUMMARY OF THE INVENTIO

The present invention is based on characterizing HLA types in terms of epitopes, which is referred to herein as "epityping".

In one aspect the present invention provides a method for HLA typing, comprising detecting a plurality of epitopes present in one or more HLA molecules in a sample from a subject, wherein each epitope comprises a region of the HLA molecule which can be recognised by an anti-HLA antibody, and thereby determining a set of HLA epitopes characteristic of the HLA type of the sub ject. In one embodiment, each epitope comprises a sequence of one, two, three or four amino acid residues in the HLA molecule. Preferably at least 10, at least 30, at least 50, at least 100, at least 150, at least 200 or at least 250 different epitopes are detected. For instance, in one embodiment the method comprises detecting at least 50, at least 100 or about 150 different HLA Class I epitopes. In another embodiment, the method comprises detecting at least 50 or about 100 different HLA Class II epitopes. In a further embodiment, the method comprises detecting at least 100, at least 150, at least 200 or about 250 different epitopes, wherein the set of epitopes comprises HLA Class I and Class II epitopes.

In one embodiment, the HLA molecule or locus comprises HLA-A, HLA-B, HLA-C, HLA-DRB 1, HLA-DRB3, HLA-BRB4, HLA-DRB5, HLA-DQA l, HLA-DQB1, HLA- DPA1 or HLA-DPB 1. For example, in an embodiment wherein the HLA molecule is HLA-A the method comprises detecting one or more HLA-A epitopes selected from, but not limited to, 44K, 150V, 107W, 65G, 62L, 161D, 163R, 144K, 166D, 167G, 76A, 90D, 62G, 142T, 145H, 127K, 801, 82L, K3R, 149T, 62E, 56E, 19K, 56R, 167W, 109L, 17S or 253Q; 43Q+62G, 62G+79G, 65G+151R, 127K+144Q, 127 +151R, 156Q+166D, 156Q+167G, 109L + 163T, 109L + 131R, 158V+167W, 142I+144K, 144K+145R, 158A+163R, 76V + 144K, 152E + 156W, 801 + 149A, 80I+90A, 138M + 144Q, 43Q + 62R, 62R+109F, 62R + 163T, 90A + 171H, 163E + 166E, 163E+167W, 79R + 127N, 56G + 65R, 43Q+90D, 19E + 79G or 62R + 65R; 149A+150A+163R, 127K+144K+145R, 9F + 142T + 149A or 9F + 145H + 149A; 9Y + 41A + 63E + 951 or 32Q + 62R + 77N + 80T.

In another embodiment the HLA molecule is HLA-C and the method comprises detecting one or more HLA-C epitopes selected from, but not limited to, 167W, 194L, 253Q, 21H, 177K or 267Q; 109L+131R, 103L+163T, 163E+166E, 163E+167W, 77N+80K, 163L+167W or 76V+80N; or 73T+76V+80N+90A.

In one embodiment, each epitope is detected by detecting a nucleic acid sequence in the sample which encodes the epitope. Preferably the epitope is detected by a nucleic acid which selectively hybridizes to the epitope-encoding nucleic acid sequence. For example, the nucleic acid may comprise a nucleic acid primer and the method comprises selectively amplifying the epitope-encoding nucleic acid sequence.

In one embodiment, the 3' terminus of the primer is complementary to a nucleotide encoding an amino acid present in the HLA epitope. When the epitope is a discontinuous epitope comprising two or more amino acid residues, the method may comprise amplifying the epitope-encoding nucleic acid sequence using two or more primers, wherein the 3 ⁵ terminus of each primer is complementary to a nucleotide encoding a different amino acid present in the HLA epitope.

In one embodiment, the nucleic acid primer comprises a primer sequence as defined in Table 1 or Table 2.

In another embodiment, the nucleic acid comprises an oligonucleotide probe. For example, the method may comprise contacting the sample with a plurality of oligonucleotide probes bound to a solid support, wherein each oligonucleotide probe selectively hybridises to a different epitope-encoding nucleic acid sequence. In an alternative embodiment, the method comprises contacting the sample with a plurality of types of bead, wherein an oligonucleotide probe which selectively hybridises to a different epitope-encoding nucleic acid sequence is attached to each bead type.

In one embodiment, the method comprises selectively amplifying an HLA locus, and detecting the epitope-encoding sequence in the amplified products. Preferably the amplified products from the HLA locus are purified before detecting the epitope- encoding sequence.

In a further aspect, the present invention provides a method for detecting a set of HLA epitopes which are recognised by anti-HLA antibodies in a subject, comprising detecting binding of antibodies in a sample from the subject to each of a set of HLA epitope- containing polypeptides, wherein each polypeptide comprises a fragment of an HLA molecule comprising an HLA epitope known to be recognised by an anti-HLA antibody.

In one embodiment of this method, each HLA epitope-containing fragment comprises a single HLA epitope. In a further aspect, the present invention provides a method for tissue matching, comprising determining a first set of HLA epitopes for a potential donor subject according to a method as defined above, and a second set of HLA epitopes for a potential recipient subject according to a method as defined above, and comparing the first set of HLA epitopes with the second set of HLA epitopes to determine matching between the two sets.

In a further aspect, the present invention provides a method for tissue matching, comprising determining a first set of HLA epitopes for a potential donor subject according to a method as defined above, and determining a second set of HLA epitopes which are recognised by anti-HLA antibodies in a potential recipient subject, and comparing the first set of HLA epitopes with the second set of HLA epitopes to determine whether any HLA epitopes are present in both sets.

In one preferred embodiment, the tissue matching is performed for solid organ transplantation, e.g. kidney, liver or heart transplantation. In general however, the method may be used for tissue matching for any type of cell, tissue or solid organ allografts of e.g. pancreatic islets, stem cells, bone marrow, skin, muscle, corneal tissue, neuronal tissue, heart, lung, combined heart-lung, kidney, liver, bowel, pancreas, trachea or oesophagus.

In a further aspect, the present invention provides a set of nucleic acid primers, wherein each primer or primer pair in the set is capable of selectively amplifying an HLA epitope- encoding sequence, wherein each HLA epitope-encoding sequence encodes an HLA epitope known to be recognised by an anti-HLA antibody.

In one embodiment, the set comprises 2 or more primers as defined in Table 1 or Table 2.

In a further aspect, the present invention provides a set of oligonucleotide probes, wherein each probe in the set is capable of selectively hybridising to an HLA epitope- encoding sequence, wherein each HLA epitope-encoding sequence encodes an HLA epitope known to be recognised by an anti-HLA antibody. In a further aspect, the present invention provides a set of HLA epitope-containing polypeptides, wherein each polypeptide comprises a fragment of an HLA molecule comprising an HLA epitope known to be recognised by an anti-HLA antibody.

In one embodiment, the set comprises one or more HLA-A epitopes selected from, but not limited to, 44K, 150V, 107W, 65G, 62L, 161D, 163R, 144K, 166D, 167G, 76A, 90D, 62G, 142T, 145H, 127K, 801, 82L, 83R, 149T, 62E, 56E, 19K, 56R, 167W, 109L, 17S or 253Q; 43Q+62G, 62G+79G, 65G+151R, 127K+144Q, 127 +151R, 156Q+166D, 156Q+167G, 109L + 163T, 109L + 131R, 158V+167W, 142I+144K, 144K+145R, 158A+163R, 76V + 144K, 152E + 156W, 801 + 149 A, 801+90A, 138M + 144Q, 43Q + 62R, 62R+109F, 62R + 163T, 90A + 171H, 163E + 166E, 163E+167W, 79R + 127N, 56G + 65R, 43Q+90D, 19E + 79G or 62R + 65R; 149A+150A+163R, 127K+144K+145R, 9F + 142T + 149A or 9F + 145H + 149A; 9Y + 41A + 63E + 951 or 32Q + 62R + 77N + 80T, and/or one or more HLA-C epitopes selected from, but not limited to, 167W, 194L, 253Q, 21H, 177 or 267Q; 109L+131R, 103L+163T, 163E+166E, 163E+167W, 77N+80K, 163L+167W or 76V+80N; or 73T+76V+80N+90A.

Preferably the set comprises primers, probes or polypeptides specific for at least 10, at least 30 or at least 50 HLA epitopes.

In a further aspect, the present invention provides a kit comprising a set of nucleic acid primers or probes or polypeptides as defined above, and optionally one or more reagents suitable for performing a method as defined above.

In one embodiment, the kit further comprises a pair of nucleic acid primers specific for one or more HLA loci.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows the results of PGR amplification of DNA samples using HLA-A epitope- specific primers as listed in Table 1. The DNA samples were known to be positive for the epitope for which the primers were specific, as they had previously undergone Sequence Based Typing. The well numbers correspond to those shown in Table 1. Figure.2 shows the results of PGR amplification of DNA samples using HLA-A epitope- specific primers as listed in Table 1. The DNA samples were known to be negative for the epitope for which the primers were specific, as they had previously undergone Sequence Based Typing. The well numbers correspond to those shown in Table 1.

Figure 3 shows the results of PGR amplification of DNA samples using HLA-A locus- specific primers. The DNA samples used were patient samples chosen to cover as many different alleles as possible.

Figure 4 shows the results of PGR amplification of a random DNA sample using HLA-A epitope-specific primers as listed in Table 1, without prior amplification of the HLA-A locus. The well numbers correspond to those shown in Table 1.

Figure 5 shows the results of PGR amplification using HLA-A epitope-specific primers as listed in Table 1 , wherein the amplification was performed on the products of a prior amplification using HLA-A locus-specific primers on a random DNA sample. The well numbers correspond to those shown in Table 1.

Figure 6 shows the results of PGR amplification using HLA-A epitope-specific primers as listed in Table 1, wherein the amplification was performed on purified products of a prior amplification using HLA-A locus-specific primers on a random DNA sample. The well numbers correspond to those shown in Table 1.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is based on the radical proposal to characterise a subject's HLA type in terms of a set of antibody-detectable HLA epitopes, rather than as a collection HLA antigens per se. This method is referred to herein as epityping.

Although information concerning potential HLA epitopes could be derived from complete sequence information for a set of HLA alleles, for instance by resolution to the four-digit amino acid level, this is time-consuming and expensive. Moreover, this type of classification does not take into account the fact that there are a limited number of actual HLA epitopes, i.e. the number of short sequence motifs which provoke anti-HLA immune responses is relatively small compared to the total number of such motifs present in all HLA alleles.

The epityping method of the present invention obviates the need for such extensive sequencing by focusing specifically on determining which immunogenic HLA epitopes are present in an individual, for example by using nucleic acid primers specific for each individual epitope. A subject's HLA type can then be characterised in terms of the set of HLA epitopes which is present. Thus even without deriving the complete sequence of each HLA allele in the subject, this method provides an HLA classification which is highly accurate and specific in terms of its value in predicting tissue compatibility. This is because it is the HLA epitopes themselves which are determinative of anti-donor immune responses and consequently acute and chronic rejection.

The epityping method is much more accurate in this respect, and therefore of greater predictive value, than methods employing the standard HLA classification, especially where resolution is only at the two-digit serological level. In such methods, a donor may be classified as having the same HLA type as a recipient even though at the molecular level, the HLA alleles are polymorphic in respect of an immunogenic HLA epitope. Conversely a potential donor and recipient may be considered to be mismatched in respect of an HLA locus, and therefore incompatible for transplant, even though the polymorphism does not produce an immunogenic HLA epitope in the donor. By characterising HLA epitopes according to the present invention, the inaccuracies of low or medium resolution HLA typing methods are avoided.

Moreover, the epityping method of the present invention allows the HLA type of a potential donor to be directly compared to the profile of any anti-HLA antibodies present in the recipient. Known methods for antibody determination rely on testing serum from the recipient against a panel of HLA antigens. Due to the highly polymorphic nature of HLA molecules, this is typically done only on a limited number of common HLA antigens. In any case, there would be no use in having a higher resolution description of HLA alleles to which a recipient has antibodies if an HLA type for the donor was available only at low resolution. The output of such known testing methods may be a set of HLA antigens at the two-digit (serological) level to which the recipient is considered to have antibodies. This HLA profile may be matched to a potential donor on the basis that none of the HLA antigens to which the recipient has antibodies are present in the donor. However, due to polymorphisms within individual alleles in a single serological grouping, the recipient may in fact have DHSA directed against an allele present in the donor, leading to antibody-mediated acute rejection of the transplanted organ. According to the prior art methods, the only way to resolve these ambiguities would be to test the recipient's serum against a complete set of all HLA alleles, and to compare this with a description of the donor's HLA type at high resolution, e.g. from DNA sequence-level typing.

Epityping according to the present invention avoids this problem by providing information on a set of HLA epitopes, which are capable of being recognised by anti- HLA antibodies, and which are present in the potential donor. This can be compared directly to information on a set of HLA epitopes whic are recognised by anti-HLA antibodies actually present in the potential recipient. The epitopes recognised by anti- HLA antibodies in the recipient may be derived using much less onerous methods than testing all known HLA alleles against the recipient's serum, for instance by directly detecting peptides containing the epitopes or deriving them from testing a smaller number of HLA antigens to which the serum reacts.

HLA typing

The present invention relates in one aspect to a method of HLA typing, which is referred to as "epityping". By reference to HLA typing it is meant that the invention provides a characteristic profile of the nature of one or more HLA molecules present in a subject. However, epityping according to the present invention typically does not provide a standard "HLA type" defined in terms of a list of HLA antigens at the two- or four-digit resolution level. By "HLA typing" according to the present invention, it is meant that a set of HLA epitopes characteristic of the subject is obtained. The HLA type provided may include a list of epitopes present at a single HLA locus (e.g. HLA-A), or may include epitopes which are derived from two or more loci (e.g. HLA-A, HLA-B and HLA-C). In some embodiments, it may be possible to convert data on the set of HLA epitopes present into a standard HLA type, i.e. to derive the HLA antigens from the epitopes which they contain. However this is not a requirement of the present method.

HLA epitopes

According to embodiments of the present invention, HLA epitopes present in a subject are detected. By "HLA epitope" it is meant a portion of an HLA molecule which can be recognised by an anti-HLA antibody, e.g. one or more amino acid residues, or a sequence of amino acid residues, within an HLA molecule to which an anti-HLA antibody is known to bind. Preferably the epitope comprises one, two, three or four amino acid residues, in some embodiments, the epitope may comprise one or more residues which are not directly contacted by an anti-HLA antibody but which are essential for determining the conformation of the epitope, i.e. which are indirect determinants of antibody binding. In embodiments of the present invention, HLA epitopes in the subject are detected directly, i.e. individual epitopes within HLA molecules are detected without identifying the entire sequence of the HLA molecule.

HLA epitopes according to the present invention are typically sequences to which known anti-HLA antibodies bind. Thus "HLA epitopes" does not typically refer to any region of an HLA molecule which could potentially bind an antibody, but rather to epitopes which have been characterised as the target of anti-donor HLA antibodies (i.e. DHSA). Embodiments of the present invention focus specifically on these epitopes in view of their importance in determining tissue compatibility.

HLA epitopes according to the present invention may be continuous or discontinuous. For example, an epitope may comprise one, two, three, four or more (preferably one to four) amino acid residues which are contiguous in the amino acid sequence of an HLA molecule. Alternatively, a discontinuous epitope may comprise two, three, four or more (preferably two to four) amino acid residues which are separated by one or more (sometimes many) intervening amino acid residues in the sequence of the HL A molecule.

Identifying HLA epitopes Many epitopes in both class I and class II HLA antigens are known in the art, and are described for example in El-Awar et ah, Transplantation (2007), 84:532-540 and Cai et ah, Curr. Op in. Immunol. (2008), 20(5):602-6. These epitopes were identified using alloanti bodies from multiparous women, or patients undergoing transfusions or transplants. The autoantibodies were first adsorbed onto cell lines recombinantly expressing a single HLA antigen. The adbsorbed antibodies were eluted from the cell lines then tested with a panel of single HLA , antigens bound to beads. The individual epitopes recognised by such antibodies can be derived by. defining the amino acids shared exclusively by positive antigens for each antibody.

Epitopes described by El-Awar or in other publications may be used in embodiments of the present invention. For instance, Duquesnoy et al. Tissue Antigens (2009) 74: 1 17-133 discusses the epitopes described by El-Awar and their relevance to the HLAMatchmaker method, and describes further HLA epitopes which may be used in the present invention. Alternatively the methods described in the above documents may be used to identify further epitopes. Thus the HLA epitopes which are detected according to embodiments of the present invention are necessarily immunogenic.

For each antibody, a set of HLA antigens to which the antibody binds (from a single antigen bead assay) can be converted to an epitope specificity by using an epitope search program to identify distinguishing amino acids that are exclusively shared by positive antigens at particular sequence positions. A number of possible 1 to 4 amino acid epitopes may be generated by such an analysis. In order to reduce the number of possibilities, the method may involve a step of only selecting positions which are exposed to the surface of the HLA molecule and which are within the antibody binding span (estimated at 494 A ² (19 x 26 A) or 750 A). Approximate distances in angstroms between two amino acids may be calculated using the Cn3D Viewer software (available at: http://wvvvv.ncbi.nlm.nih.gov/Structure/CN3D/cn3d.shtml.) and based on the published three-dimensional structures of HLA molecules. Amino acids that are exclusively unique to a group of antigens that are reactive with an antibody, preferably exposed to the surface of the molecule, and within the binding span of the antibody may be considered a di stinguishing characteristic of the epitope. The antibodies used by El-Awar recognised epitopes comprising one to four unique amino acids at specific positions. The individual amino acids forming an epitope were not all contiguous but were within a conformational distance formed by protein folding allowing antibody binding. Such methods have allowed the identification of at least 103 HLA class I epitopes and 83 HLA class II epitopes. However, the number of functional HLA epitopes is probably much lower than the total number of permutations of amino acid combinations which are present in HLA molecules. It is also much lower than the total number of HLA alleles. This is an advantageous feature of the present invention, since by focusing .on the epitopes it is possible to reduce unnecessary sequence analysis in determining HLA compatibility.

Detecting a plurality of epitopes

The epityping method of the present invention preferably involves detecting multiple HLA epitopes, such that a characteristic definition of the subject's HLA type can be provided. For instance, the method may comprise detecting at least 2, 3, 4, 5, 10, 20, 30, 50, 100, 200, 300 or 500 epitopes. In one embodiment, a plurality of epitopes (e.g. at least 2, 5, 10, 30 or 50 epitopes, typically 10 to 20 epitopes) within a single HLA molecule is detected. By "detecting epitopes" it is meant that the present method involves performing an assay to determine whether or not such epitopes are present in the sample. Thus when referring to the detection of a particular number of epitopes, it is meant that this number of epitopes is assayed for in the sample, and not that any one individual sample is actually positive for this number of epitopes.

In another embodiment, at least one epitope within each of a plurality of different HLA molecules is detected. For example, epitopes within two or more of HLA- A, HLA-B, HLA-C, HLA-DR, HLA-DQ and HLA-DP molecules may be detected. In one embodiment, two or more epitopes (e.g. at least 2, 5, 10, 30 or 50 epitopes) are detected within each of a plurality of different HLA molecules. In an embodiment which involves detecting HLA Class I epitopes, typically 100 to 200 or about 150 epitopes (in e.g. HLA- A, HLA-B and HLA-C) are detected. In an embodiment which involves detecting HLA Class II epitopes, typically 50 to 150 or about 100 epitopes (in e.g. HLA-DR, HLA-DQ and HLA-DP) are detected. In an embodiment which involves detecting both HLA Class

I and Class II epitopes, typically 200 to 300 or about 250 epitopes are detected.

HLA molecules

By "HLA molecule" it is meant any HLA polypeptide, protein, or glycoprotein including both class I and class II antigens. In one embodiment, the HLA molecule is a class I antigen, e.g. HLA-A, HLA-B or HLA-C. In another embodiment, the HLA molecule is a class II antigen, e.g. HLA-D , HLA-DQ or HLA-DP. HLA epitopes present in any variant polypeptide chain of the HLA molecule may be detected, e.g. in the case of class

II antigens the epitopes may be in either the a or β chain.

In one embodiment the HLA molecule is HLA-A. Preferably one or more HLA-A epitopes such as defined in Table 1 below, or in El-Awar et ah, Transplantation (2007), 84:532-540, are detected. For example, the method may comprise detecting at least 1, 2, 3, 5, 10, 20, 30, 50 or more- of the following HLA-A epitopes:

(a) single amino acid epitopes - 44K, 150V, 107W, 65G, 62L, 161D, 163R, 144K, 166D, 167G, 76A, 90D, 62G, 142T, 145H, 127K, 801, 82L, 83R, 149T, 62E, 56E, 19K, 56R, 167W, 109L, 17S or 253Q;

(b) two amino acid epitopes: 43Q+62G, 62G+79G, 65G+151R, 127K+144Q, 127K+151R, 156Q+166D, 156Q+167G, 109L + 163T, 109L + 131R, 158V+167W, 142I+144K, 144K+145R, J 58A+163R, 76V + 144 , 152E + 156W, 801 + 149A,

80I+90A, 138M + 144Q, 43Q + 62R, 62R+109F, 62R + 163T, 90A + 171H, 163E + 166E, 163E+167W, 79R + 127N, 56G + 65R, 43Q+90D, 19E + 79G or 62R + 65R;

(c) three amino acid epitopes - 149A+150A+163R, 127 +144K+145R, 9F + 142T + 149 A or 9F + 145H + 149A;

(d) four amino acid epitopes - 9Y + 41A + 63E + 951 or 32Q + 62R + 77N + 80T.

In another embodiment, the HLA molecule is HLA-C. Preferably one or more HLA-C epitopes such as defined in Table 2 below, or in El-Awar et ah, Transplantation (2007), 84:532-540, are detected. For example, the method may comprise detecting at least 1, 2, 3, 5, 10 or more of the following HLA-C epitopes:

(a) single amino acid epitopes - 167W, 194L, 253Q, 21H, 177 or 267Q;

(b) two amino acid epitopes - 109L+131R, 103L+163T, 163E+166E, 163E+ 167W, 77N+80K, 163L+167W or 76V+80N;

(c) four amino acid epitope - 73T+76V+80N+90A. Sample

The sample may be any type of sample derived from the subject, provided that it permits the identification of HLA epitopes in the subject. Typically the sample comprises DNA derived from the subject, e.g. genomic or cDNA from the subject. For example, the sample may comprise a tissue or cellular sample or a purified DNA preparation from the subject. Preferably the sample comprises purified genomic DNA from a blood sample from the subject. Methods for the isolation of genomic DNA or preparation of cDNA from tissue samples are well-known in the art.

Detection of epitope-encoding nucleic acids

According to embodiments of the present invention, epitopes in HLA molecules may be detected by any suitable method. Preferably epitopes are detected at the nucleic acid level, e.g. each epitope is detected by detecting a nucleic acid sequence which encodes the epitope. Typically epitope-encoding nucleic acid sequences are detected in genomic DNA, although in alternative embodiments it is also possible to detect epitopes present in mRNA, e.g. by the use of RT-PCR methods to produce cDNA , For example, total RNA may be obtained from an appropriate cellular source such as peripheral blood mononuclear cells. The RNA may be reverse transcribed using antisense primers, which can be specific for specific HLA locus or group of loci to produce cDNA.

HLA molecules are encoded by various loci within the short arm of chromosome 6. Each of the class I antigens HLA-A, HLA-B and HLA-C is encoded by a single locus, whereas each of the class II antigens is encoded by two loci. Thus the a and β subunits of HLA- DQ are encoded by the HLA-DQA 1 and HLA-DQB1 loci respectively. In a similar way the loci HLA-DP A 1 or HLA-DPB 1 encode the a and β subunits of HLA-DP. HLA-DR is encoded by the loci HLA-DRA1, which encodes the a subunit, and HLA-DRB 1 , HLA- DRB3, HLA-DRB4 and HLA-DRB 5, each of which encode a β subunit which can pair with the a subunit. In embodiments of the present invention, HLA epitopes encoded by a gene at any one or more of the above loci may be detected. HLA-DRA is less preferred since in contrast to other loci this locus shows a very low degree of polymorphism between individuals.

Preferably the method involves detecting HLA epitopes encoded by a gene at one or more of the HLA-A, HLA-B, HLA-C, HLA-DPA1, HLA-DPB 1 , HLA-DQA 1 , HLA- DQB1 , HLA-DRB 1, HLA-DRB 3, HLA-DRB4 or HLA-DRB 5 loci. Preferably the method involves detecting epitopes from at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 of these loci, in one embodiment epitopes encoded by at least the HLA-A, HLA-B and HLA-DRB 1 loci are detected. More preferably, the method comprises detecting HLA epitopes encoded by each of the HLA-A, HLA-B, HLA-C, HLA-DPA1, HLA-DPB 1, HLA- DQA 1, HLA-DQBl, HLA-DRB 1, HLA-DRB 3, HLA-DRB4 and HLA-DRB 5 loci.

The nucleotide sequences of HLA genes, and the polypeptide sequences which they encode, are well known and are published in publicly-available databases. Moreover the nucleotide and amino acid sequences of the wide diversity of individual HLA alleles are well characterized and available to the public. For example, such sequences may be accessed via the Anthony Nolan website at http ://www.anthony n o I an . org. u k, at http://hla.alleles.org/data/index.html, or the IMGT/HLA Database at http://www.ebi.ac.uk/imgt/hla/, as accessed 13 August 2009.

Methods for detecting epitope-encoding nucleic acids

Accordingly, any method suitable for detecting nucleic acid sequences may be employed in embodiments of the present invention. However, since the present invention focuses on detecting known HLA epitopes, it is preferable to use methods which are based on determining whether a predetermined nucleic acid sequence is present in an HLA molecule in a subject. In other words, rather than using DNA sequencing methods which simply provide a complete nucleotide sequence for a gene at an HLA locus, according to the present invention a method may be used which determines whether or not a specific nucleic sequence encoding a defined epitope within an HLA molecule is present in a subject. Thus the present methods provide a single positive or negative result for whether a particular HLA epitope is present in the sample.

These embodiments of the invention involve detecting an "epitope-encoding nucleic acid sequence". Typically this involves detecting a short sequence of nucleotides which is specific for the epitope of interest. Since each amino acid in the epitope is encoded by a single codon, i.e. a triplet of bases, the length of the epitope-encoding nucleic acid sequence may be determined according to the number of amino acids in the epitope. For example, in the case of a 3 -amino acid epitope, the epitope-encoding nucleic acid sequence may comprise a specific sequence of 9 nucleotides (i.e. 3 codons). Thus epitope-encoding nucleic acid sequences correspond to particular polymorphisms in HLA genes, wherein these polymorphisms encode an HLA epitope which is known to be recognised by an anti-HLA antibody. Accordingly, epitope-encoding nucleic acid sequences may be detected by any known methods for detecting polymorphisms in DNA.

Typically, in order to identify a particular epitope as being present in the correct context, i.e. at a specific position in an HLA molecule, it is necessary to detect in addition a sequence to one or both sides of the individual codons corresponding to the epitope itself. Moreover, in some embodiments where an adjacent sequence is also detected, it may be sufficient that only a single nucleotide within the codon corresponding to the epitope itself is detected. Thus "epitope-encoding nucleic acid sequence" is used herein to mean a sequence which is specific for a particular epitope, i.e. a sequence which comprises at least one epitope-speci ic nucleotide and which may or may not include nucleotides adjacent to the codons corresponding to the epitope itself.

Preferably each epitope-encoding nucleic acid sequence is 1 to 50 nucleotides in length, e.g. 5 to 50 nucleotides, 10 to 30 nucleotides or 15 to 25 nucleotides in length. In some embodiments each epitope-encoding nucleic acid sequence may comprise 1 to 12, e.g. 1 , 2, 3, 4, 5, 6, 7, 8 or 9 nucleotides which encode the amino acids in the epitope itself, and which are specific for that epitope.

For instance, the method may involve using a nucleic acid which selectively hybridizes to a defined nucleotide sequence which encodes the HLA epitope of interest. The nucleic acid may be an oligonucleotide primer or probe which comprises a region which is complementary to the epitope-encoding sequence. A skilled person can easily design appropriate primers or probes based on a knowledge of the nucleotide sequence encoding an epitope of interest.

In some embodiments, epitope-encoding nucleic acid sequences may be detected directly in a DNA sample derived from a subject. In other embodiments, the method involves first enriching the sample for an HLA-encoding region. For example, the method may involve a step of selectively amplifying a sequence at an HLA locus, e.g. one or more of the HLA-A, HLA-B, HLA-C, HLA-DPAl , HLA-DPB 1 , HLA-DQAl, HLA-DQB1, HLA-DRB 1 , IILA-DRB3, HLA-DRB4 or HLA-DRB5 loci. Primers suitable for PGR amplification at such HLA loci are well known in the art. Detection of HLA-epitopes encoded within one or. more HLA genes may then be performed by applying the epitope- selective primers or probes to the amplified product. Preferably the products of the locus-specific amplification step are purified in a DNA purification step before the epitope-specific detection step.

For example, in one embodiment PGR amplification of DNA is first used as the means to enrich for a selected DNA region encoding one or more HLA antigens. In a subsequent step, the specific epitopes present in these HLA antigens are detected by detecting the epitope-encoding nucleotide sequences. Various methods for the detection of epitope- encoding sequences may be employed, involving the use of, for example, sequence- specific primers (SSP, e.g. as described in Olerup, O., et al. (1992) Tissue Antigens 39:225); direct sequence-specific oligonucleotide probes (SSOP, e.g. as described in Saiki, RK., et al. (1986) Nature 324: 163); restriction fragment length polymorphisms (RFLP, e.g. as described in Maeda, M., et al. (1989) Tissue Antigens 34:290); and reverse SSOP dot blot technologies (see for example Bugawan, TL., et al. (1990) Immunogenctics 32: 231). The use of an initial locus-specific PGR amplification step is particularly preferred when epitope-encoding sequences are detected using oligonucleotide probes (e.g. SSOP), but is also preferable in some methods involving epitope-specific primers. In another embodiment, epitope-encoding sequences may be detected by conformation analysis methods, e.g. using reference strand mediated conformation analysis as described in Turner et al. (2001), Human Immunology 62:414- 418 or Argiiello et al. (1998), Nature Genetics 18:192-194.

Although similar methods involving SSP, SSOP and the like may be used in some standard HLA typing procedures, it is important to note that according to the present invention, these techniques are used to detect specific epitopes rather than HLA alleles as such. For example, according to the present method, the SSP or SSOP which are used are specific only for sequences in HLA loci which encode epitopes which are known to be recognised by an anti-HLA antibody. In contrast, according to standard DNA-based HLA typing methods, SSP or SSOP are used in order to detect the presence of a specific HLA allele, and therefore in order to be truly accurate need to uniquely distinguish that allele from all other known alleles. In some known methods, the SSP or SSOP are specific for a sequence found in a group of related alleles, and therefore do not provide a sufficiently detailed description of the subject's HLA type. Because they do not focus on the HLA epitopes as such, these known methods need to use primers or probes which are capable of characterising HLA alleles to the level of their complete sequence in order to be accurate, whereas this is not required according to the present invention.

Primers

In epityping methods involving primers, the primer is typically extended (e.g. using known methods such as PGR) only when annealed to a specific nucleotide sequence (template). For example, a sequence-specific primer can be used to selectively produce an amplification product only when the specific HLA epitope-encoding sequence is present in the sample.

For example, in one embodiment the primer comprises a sequence complementary to an HLA epitope-encoding sequence. If the primer binds the complementary epitope- encoding sequence and produces an amplified product, then the PGR product can be detected by standard techniques. If the primer is not complementary to the polymorphism, then it cannot bind and facilitate amplification, so no PGR product is detected. By constructing an array of PGR primers complementary to the range of HLA epitopes, it is possible to detect a plurality of HLA epitopes directly by PGR.

The nature of the primer is not particularly limited, provided that it is capable of specifically hybridising to a HLA epitope-encoding sequence. The length of the primer is preferably 5 to 50 nucleotides, more preferably 10 to 50 nucleotides, more preferably 15 to 30 nucleotides, e.g. 17 to 23 nucleotides. Suitable primers may be designed according to standard techniques known to those skilled in the art for selecting primers for polymerase reactions, such as for amplification of DNA by the polymerase chain reaction (PGR).

For example, in one embodiment an aqueous solution of the primer is added to a DNA sample. Hybridisation conditions are then selected so that the primer hybridises selectively to the epitope-encoding sequence, according to criteria well known to those skilled in the art. An appropriate temperature and salt content for hybridisation needs to be selected according to the ^' length of the oligonucleotide primer and its G-C content, amongst other things (see Old & Primrose (1994), Principles of Gene Manipulation, Blackwell Science and Maniatis et al. (1992), Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, New York). Typically the hybridisation temperature should be close to the melting temperature (T _m ) of the primer. Tm is defined as the temperature at which the primer and its target are 50% dissociated, and for oligonucleotide primers may be approximated according to the "Wallace rule" by the following formula:

T _m = 4(number of G:C base-pairs) + 2(number of A:T base-pairs)

Preferably the hybridisation temperature should be within 2°C of T _ra . Accordingly, for a 20-mer oligonucleotide primer with 50% G-C content, the T _m is about 60°C and a suitable hybridisation temperature would be 58°C. Various kits are available commercially for the detection and identification of HLA alleles using SSP HLA typing methods. Examples include SSP UniTray® from Dynal Biotech, Olerup SSP™ HLA Kits from Qiagen and HLA SSP Typing Kits from R.O.S.E. Europe GmbH. These kits detect HLA alleles rather than HLA epitopes as such, and are therefore not directly applicable to the methods of the present invention. However the principles underlying such detection methods are in many ways similar to those used in the present invention. Thus these commercially available kits can be modified for use in the present invention in some cases by redesigning the sequence-specific primers such that they recognise HLA epitope-encoding sequences as defined herein.

For example, in one embodiment the method utilizes a set of primer pairs designed for specifically detecting individual epitopes in one or more HLA genes. The PGR reactions accomplished with these primers produce well defined DNA fragments of different length if their respective epitope-encoding sequences are present in the sample. Control primers for use as an internal standard (e.g. for amplifying the human globin or human growth hormone gene) may also be included in the reaction. The PGR reaction products may be detected on an electrophoretic gel, e.g. an agarose gel by dyeing the double stranded DNA with ethidium bromide and exposure to ultraviolet light. The gel may be documented by photography and interpreted.

According to some embodiments of the present invention, only one primer in a pair is specific for an HLA epitope-encoding sequence. The other primer may, for example, anneal to a known sequence on the other DNA strand which is downstream of the selected epitope-encoding sequence in the HLA gene. Amplification using this primer pair will produce a double-stranded fragment of a particular length only if the epitope- encoding sequence (to which the epitope-specific primer anneals) is present in the sample. Alternatively, each primer in a pair may be specific for an HLA epitope- encoding sequence, e.g. where discontinuous epitopes are detected.

Preferably the 3' terminus of the primer hybridizes to an HLA epitope-specific nucleotide sequence, i.e. one or more nucleotides which encode one or more amino acids in the epitope itself. In other words, the primer is designed such that one or more bases specific for the HLA epitope of interest are complementary to one or more nucleotides at the 3' terminus of the primer. As is well known in the art, mismatches at the 3' terminus of a primer are more likely to disfavour primer elongation than at any other location. Thus primers in which the 3 ⁵ terminus is complementary to one or more epitope-specific nucleotides are most likely to discriminate between samples which either contain or do not contain the HLA epitope of interest, and are preferred in methods of the present invention.

For instance, if a specific HLA epitope is encoded by a codon in which the third base is a G which is characteristically present at position 159, a forward primer might be designed comprising the sequence of residues 140 to 159 (i.e. with a G at the 3' terminus). The reverse primer need not be specific for the epitope encoded by G159, e.g. could be complementary to a downstream region present in all HLA alleles. In an alternative embodiment, the reverse primer could comprise the complement of residues 159 to 178, i.e. with a C at the 3' terminus, and the forward primer may comprise an upstream sequence not specific to the epitope.

In another embodiment, the HLA epitope may be discontinuous, i.e. it is specifically encoded by two or more nucleotide residues (or sequences of nucleotide residues, e.g. three nucleotides comprising a codon) separated by one or more intervening nucleotides. In such embodiments, it may be preferable for one (e.g. the forward) primer to be specific for a first epitope-specific nucleotide or nucleotide sequence (e.g. the 3' terminus of the forward primer is complementary to the first epitope-specific nucleotide or nucleotide sequence), whereas the other (e.g. the reverse) primer is specific for a second epitope- specific nucleotide or nucleotide sequence (e.g. the 3' terminus of the reverse primer is complementary to a second epitope-specific nucleotide or nucleotide sequence). The first and second epitope-specific nucleotides or nucleotide sequences are typically separated by one or more intervening nucleotides in the HLA gene sequence and are on different strands.

Analogous methods may be employed for detecting discontinuous epitopes encoded by three or more specific nucleotides or codons separated by one or more intervening nucleotides. For example, an epitope may be encoded by 3 codons, in each of which there is an epitope-specific nucleotide present as the third base. These epitope-specific nucleotides may be referred to as Nl, N2 and N3 and are separated by intervening nucleotides. Pairs of epitope-specific nucleotides may be confirmed as being present by amplifying the .region between them using forward and reverse primers specific for each. Thus a PGR product of a specific size is produced by a primer pair in which the forward primer is specific for Nl and the reverse primer is specific for N2, whereas if the forward primer is specific for Nl and the reverse primer is specific for N3 a different fragment size is amplified. A reaction between a forward primer specific for N2 and a reverse primer specific for N3 will also produce a specific sized amplified product. If all three expected fragment sizes can be amplified from the sample, the discontinuous epitope comprising Nl, N2 and N3 is present. The same method can be repeated for discontinuous epitopes encoded by four or more characteristic nucleotides. in an alternative embodiment useful for detecting discontinuous HLA epitopes, particularly where the first and second epitope-specific nucleotides or codons are separated by a small number of intervening nucleotides, a single primer which is complementary to both epitope-specific nucleotides on the same strand may be used. For instance, the primer may anneal to a sequence spanning both epitope-specific nucleotides, provided that the primer will only be extended when both epitope-specific nucleotides are present in the template. A single primer may be complementary to a region spanning three, four or more epitope-specific nucleotides.

The methods described above for detecting discontinuous epitopes may be applied in some embodiments for haplotyping, i.e. for detecting a combination of epitopes encoded by the same chromosome. In other words, such methods may be used for discriminating between epitopes present on each of the two (maternal and paternal) alleles for each HLA locus in an individual. The present methods may also be combined with well-known methods for haplotyping, e.g. allele-specific primers as described in DNA Sequence (1996), 6(2):87-94 and coupling of primers to magnetic beads (e.g. as described in Biotechniques (1991), 10(l):30-34). In some embodiments, a particular epitope may be characterised by the presence of single nucleotide at a particular position in an HLA sequence. In such embodiments, well- known methods for the detection of single nucleotide polymorphisms (SNPs) may be employed.

Oligonucleotide probes

Oligonucleotide probes may be used in combination with well-known hybridisation assays to detect an epitope-encoding sequence. For example, the- oligonucleotide probe may be attached to a solid support. In one embodiment, a plurality of oligonucleotide probes may be attached to a solid support in the form of an array, e.g. a DNA micro- array. In an alternative embodiment, different probes may be attached to individual beads or microspheres, e.g. Luminex™ microspheres.

In one embodiment, sequence-specific oligonucleotides (SSOs) are used to identify which HLA epitopes are present in a PGR amplified sample. A set of SSOs is employed which recognises a plurality of HLA epitopes present at a particular locus. Each of the different probes is homologous to a sequence within the amplified DNA which encodes a unique epitope. In other words, these probes are designed so that each probe preferentially hybridizes to a complementary region of an HLA locus which may or may not be present in the amplified DNA.

In some embodiments, the amplified DNA may also be hybridized to one or more consensus probes homologous to sequences present in all HLA alleles of a locus. The signal obtained for the consensus probe(s) can serve as an internal standard to control for variations in the success of the amplification and hybridization procedures (e.g. differences in the type of biological material, method of purification, amount and integrity of genomic DNA and amplified product). Thus the results generated from such a method can be used to determine the presence or absence of particular DNA sequences in amplified DNA and to identify the HLA epitopes present in the sample.

In one embodiment, the probes are attached to microspheres, e.g. to Luminex™ microspheres designed for use with a Luminex™ Instrument. A large number (e.g. up to 100) different types of microspheres can be mixed and analyzed together. In one embodiment the method is carried out in a single reaction vessel. Each population of microspheres can be distinguished by its unique fluorescence signature or colour.

In one embodiment, a different SSO probe is attached to each colour microsphere. A mixture of several probes can be distinguished from each other by virtue of their association with particular colour microspheres. In some embodiments, a label is introduced into the PGR product during the initial amplification reaction, e.g. by using biotin-labelled primers. The relative amounts of labeled PGR product hybridizing to each microsphere can then be quantified. The relative signal obtained can be used to assign the probes as having positive or negative reactivity with the amplified DNA sample. This in turn provides the information needed to determine whether an HLA epitope is present in the sample.

Various SSO HLA-typing kits are commercially available, for instance Dynal RELI™ SSO kits from Dynal Biotech and Lifecodes Class I and II Typing Kits from Tepnel Life Sciences. The methods employed in such kits may be modified for use in the present invention by using oligonucleotide probes which are specific for an HLA epitope.

In alternative embodiments, the method may involve a reverse dot blot or SSOP probe, for example in which an enzyme label and colourimetric substrate is used to detect specific hybridization. In one embodiment, a particular HLA locus is first amplified using PGR. The amplified product is then hybridised with a labeled, epitope-specific oligonucleotide probe. The PCR-amplified DNA may, for example, be blotted onto a nitrocellulose or nylon filter for hybridization. The dot blot may then be incubated with a radiolabelled epitope-specific oligonucleotide probe. The epitope-specific probe only binds to the filter producing a radiolabelled spot if the complementary epitope-encoding HLA nucleic acid sequence is present in the DNA amplified by PGR. A skilled person can easily determine appropriate conditions to produce the required sensitivity and specificity. In some embodiments, a plurality of probes is used to identify an array of HL A epitopes. More preferably, the epitope-specific specific oligonucleotides are labeled with a label such as a fluorescent dye, biotin, digoxigenin, or directly with an enzyme such as horseradish peroxidase. The appropriate substrate can be added to produce a fluorometric (or colorimetric) readout without the problems associated with radioactivity. Epitope- specific oligonucleotide probes labelled with such reagents can be applied to an amplified PGR product from an HLA locus and visualized using known techniques.

In some embodiments, the PGR primers used to amplify the HLA locus are pre-labelled, e.g. with biotin. These PGR products are then incubated with unlabeled, epitope-specific oligonucleotide probes immobilized on a solid support such as a membrane. After incubation and wash steps, the hybridization products can be detected, e.g. by streptavidin-horseradi sh peroxidase and a chromogenic substrate.

Oligonucleotide arrays

In some embodiments, oligonucleotide probes immobilized on an array may be used to detect HLA epitope-encoding sequences. Oligonucleotide arrays are generally known in the art and arrays for HLA typing are also known, see for example WO 00/79006. Arrays comprising probes specific for HLA epitope-encoding sequences may be prepared by appropriately modifying such known methods.

Oligonucleotide arrays may be prepared, for example, by in situ combinatorial oligonucleotide synthesis or by conventional synthesis followed by on-chip immobilization of the oligonucleotide onto the solid support. The solid support may be, for example, a glass slide.

In some embodiments, sample DNA may be amplified by PGR, labeled with a fluorescent tag and hybridized to the oligonucleotide array. The hybridization pattern may be measured, for example, by fluorescence scanning and the intensity of each hybridization signal is quantified using appropriate software. Oligonucleotide arrays provide the ability to assay many different HLA epitopes simultaneously.

Oligonucleotide probes suitable for immobilization on an array, as well as those used in other SSO embodiments, may be of any suitable length provided that they are specific for an HLA pitope-encoding sequence. Typically oligonucleotide probes may be 10 to 50,

15 to 30, 17 to 23 or about 20 nucleotides in length.

Preferably an array comprises at least 10, 30, 50, 100, 200, 500, 1000, 10,000 or 100,000 different epitope-specific oligonucleotide probes. A single array may detect a plurality of epitopes at a single HLA locus or epitopes from a plurality of loci.

Detecting anti-HLA antibodies in a subject

In some embodiments of the present invention, anti-HLA antibodies are detected in the serum of a subject (e.g. a potential transplant recipient) in order to determine whether any epitopes which they recognise are present in a potential transplant donor.

In one embodiment, anti-HLA antibodies may be detected using standard methods. For instance, many such assays are based on detection of binding of an antibody to a single HLA antigen immobilized on a solid substrate, e.g. to a bead or an array.

In one embodiment, affinity, purified Class I or Class II HLA glycoproteins are conjugated to beads (e.g. Luminex beads). Different bead types may comprise different HLA antigens. An aliquot of beads is allowed to incubate with a small volume of test serum sample. The sensitized beads are then washed to remove unbound antibody. An anti-human immunoglobulin (e.g. anti-IgG) antibody conjugated to a detectable marker (e.g. phycoeryihrin) is then added. The sample is then analyzed, e.g. in a flow analyser, by separation of the different bead types and detecting for presence of the marker. The signal intensity from each bead is compared to the signal intensity of a negative control bead included in the bead preparation to determine if the bead is positi ve or negative for bound alloantibody.

A method involving a panel of single HLA antigen beads such as that described in El- Awar et al. (2007), Transplantation 84:532-540 may be used. Commercially available kits for detecting anti-HLA antibodies include Lifecodes ID and LSA kits from Tepnel Molecular Diagnostics, Stamford CT, and L AB Screen beads from One Lambda Inc. Canoga Park, CA). By following the above methods, it is possible to derive a set of HLA antigens which are recognised by anti-HLA antibodies in the recipient. According to the present invention, it is necessary to determine the epitopes to which these anti-HLA antibodies bind. This may be done using the computational methods described in El-Awar et al. (2007), Transplantation 84:532-540 as discussed herein. By selecting an appropriate set .of HLA antigens for use in the assay, it is possible to screen for specific epitopes. Since there are a limited number of HLA epitopes, it is typically not necessary to screen for binding to every known HLA antigen in order to screen for each significant epitope. In fact, it is an advantage of the present method that a significantly lower number of HLA antigens needs to be screened for antibody binding in order to adequately describe the antibody repertoire in a subject. The result is a set of HLA epitopes recognised by the anti-HLA antibodies in the subject (e.g. potential transplant recipient).

In an alternative embodiment, HLA epitopes are detected ^" directly, rather than by derivation from binding of antibodies to single HLA antigens. For example, HLA epitopes may be included in antigenic polypeptide fragments derived from HLA antigens. Preferably each polypeptide fragment comprises a single HLA epitope, i.e. the fragment is short enough in length such that antibody binding to different HLA epitopes in the same HLA molecule can be discriminated. In some embodiments, such antigenic (epitope-containing) fragments comprise 3 to 100, 3 to 50, 5 to 30 or 5 to 15 amino acids. The length of the fragment may be varied depending on the length of the polypeptide sequence spanned by the epitope in an HLA antigen, e.g. including intervening amino acids in a discontinuous epitope. Preferably the fragment is selected such that the epitope assumes its native conformation in the HLA antigen, and can be recognised by the anti- HLA antibody. Anti-HLA antibodies may be detected directly by binding to HLA epitope-containing fragments using analogous methods to those discussed above using full-length HLA antigens.

Matching

In one embodiment, the present invention provides a method for matching a potential tissue or organ donor to a potential recipient. Matching may be performed based on a set of HLA epitopes determined to be present in the donor, versus a set of HLA epitopes determined to be present i a potential recipient. A mismatch in one or more HLA epitopes, e.g. the presence of an epitope in the donor not found in the recipient, may be used as an indicator that the donor and recipient are incompatible. In some cases, one or more mismatches may. be considered to be acceptable, depending on the required level of compatibility and immunogenicity of individual epitopes. Preferably there are no epitope mismatches between donor and recipient.

In another embodiment, the set of HLA epitopes in a potential donor is ^' compared to a set of HLA epitopes recognised by pre-formed anti-HLA antibodies in a potential recipient. This method may be performed, for example, where there are mismatches between the set of HLA epitopes determined to be present in donor and recipient and it is desired to screen for DHSA in the recipient. Preferably there is no overlap between the specificity of recipient anti-HLA antibodies and donor HLA epitopes.

The method is suitable for any type of tissue or organ transplantation, for example for cell, tissue or solid organ allografts of e.g. pancreatic islets, stem cells, bone marrow, skin, muscle, corneal tissue, neuronal tissue, heart, lung, combined heart-lung, kidney, liver, bowel, pancreas, trachea or oesophagus. Preferably the tissue matching is performed for solid organ transplantation, e.g. kidney, liver or heart transplantation.

Kits

In a further aspect, the present invention provides a kit comprising one or components for performing the epityping methods of the present invention as described herein. For example, the kit may comprise a set of nucleic acids which selectively hybridize to particular HLA epitope-encoding nucleic acid sequences.

Preferably the kit comprises a set of nucleic acid primers, e.g. suitable for PGR amplification of HLA epitope-encoding sequences. In one embodiment each primer or primer pair is capable of selectively amplifying a different HLA epitope-encoding sequence. In a more specific embodiment, the set comprises primers suitable for amplifying nucleic acid sequences encoding at least 2, 3, 5, 10, 20, 30, 50 or 100 different HLA epitopes. For example, the set may comprise at least 2, 3, 5, 10, 20, 30, 50 or 100 different primers from those listed in Table 1 and/or Table 2. Typically the kit may comprise about 2, 3 or 4 primer pairs for each epitope detected, e.g. in order to detect ^' .each amino acid of a discontinuous epitope. Therefore a set of primers for detecting HLA Class I epitopes may comprise, for example, 300 to 600 or about 450 primer pairs. A set of primers for detecting HLA Class II epitopes may comprise, for example, 150 to 450 or about 300 primer pairs. A set of primers for detecting both HLA Class I and II epitopes may comprise 600 to 900 or about 750 primer pairs.

In another embodiment, the kit comprises a set of oligonucleotide probes, preferably wherein each probe in the set is capable of selectively hybridismg to a different HLA epitope-encoding sequence. In one embodiment, each oligonucleotide probe may be bound to a solid phase such as a bead or microsphere, e.g. wherein different probes are bound to different bead types (e.g. Luminex™ beads). Preferably the set comprises probes which hybridise to nucleic acid sequences encoding at least 2, 3, 5, 10, 20, 30, 50 or 100 different HLA epitopes.

In some embodiments, the kit may further comprise a pair of primers specific for one or more HLA loci, e.g. HLA-A, HLA-B, HLA-C, HLA-DQA1, HLA-DQB1 , HLA-DPA1, HLA-DPBl, HLA-DRB 1 , HLA-DRB 3, HLA-DRB4 and/or HLA-DRB5.

The above kits may comprise one or more further reagents for performing the methods of the invention. For example, kits comprising PGR primers may contain reagents suitable for performing PGR reactions, such as dNTPs, a thermostable polymerase (e.g. Taq polymerase) and a suitable PGR buffer. Kits comprising oligonucleotide probes may further comprise one or more reagents for detecting specific hybridisation, depending on the visualization technique which is used. For example, in some embodiments the kit may further comprise an HLA locus-specific primer pair which comprises a label or a first member of a binding pair (e.g. biotin) which becomes incorporated into the products of an initial PGR step. The kit may further comprise the second member of the binding pair (e.g. streptavidin) conjugated to a detection reagent (e.g. phycoerythrin). In another aspect, the kit may comprise a set of polypeptides comprising HLA epitopes for detecting ant i -HLA antibodies to those epitopes. For instance, each polypeptide in the set may comprise a single HLA epitope, or a limited number of epitopes (e.g. many fewer than are present in a full length HLA antigen). Preferably each polypeptide comprises a different HLA. epitope. In one embodiment, the set comprises polypeptides comprising at least 2, - 3, 5, 10, 20, 30, 50 or 100 different HLA epitopes. In one embodiment, each polypeptide may be bound to a solid phase, e.g. a bead or microsphere, and different epitopes may be conjugated to different bead types (e.g. Luminex™ beads). The kit may further comprise one or more reagents suitable for detecting serum antibody binding to the polypeptides, e.g. an anti-human immunoglobulin such as anti-IgG. In one embodiment, the reagent for detecting antibody bindin (e.g. anti-IgG) may further comprise a label or visualization reagent such .as phycoerythrin. The kit may further comprise one or more buffers for performing the assay.

Any of the above kits may further comprise suitable controls (e.g. as discussed herein in relation to methods of the invention) and may be provided in suitable packaging with instructions for performing the methods of the invention.

The invention will now be described, by way of example only, with reference to the following non-limiting specific embodiments.

EXAMPLES

Example 1: Epityping at HLA- A Locus !ising SSP

Sequence-specific PGR primers were designed to detect the HLA-A epitopes identified by El-Awar et al 2007 (Human Leukocyte Antigen Class I Epitopes: Update to 103 Total Epitopes, Including the C Locus. Transplantation 2007;84:532-540). Primer specificity was confirmed by using the Probe and Primer Search Tool available at www.ebi.ac.uk/imgt/hla/probe.html. The Probe and Primer Search Tool searches the known coding sequence of any HLA allele, for a particular nucleotide motif.

Primers specific for each HLA-A epitope are shown in Table 1 below. Epitopes may be comprised of one, two, three or four specific amino acids as defined in the left-hand column. Amino acids in brackets are present in an epitope, and essential to antibody binding, but not directly contacted by the antibody. The numbers in brackets following each primer sequence are SEQ ID NOs.

Table 1

PGR reactions were performed on genomic DNA samples using the primers listed above. Epitope specific primers were pre-aliquoted and dried overnight. The final volume of reaction mix per well was ΙΟμΙ, which contained approximately lx Reddy Buffer IV (Abgene), l x Buffer IV (Abgene), 0.2mM dNTPs (Abgene), 134nM 63/64 control primer, l .l75mM MgCl ₂ (Abgene), 0.4 units Taq polymerase (Abgene) and 43ng DNA. Thermal cycling conditions were (a) 94°C for 1 minute, followed by (b) 5 cycles of [94°C for 25 sec, 70°C for 45 sec and 72°C for 30 sec], followed by (c) 20 cycles of [94°C for 25 sec, 63°C for 45 sec and 72°C for 30 sec], followed by (d) 5 cycles of [94°C for 25 sec, 55°C for 1 min and 72°C for 2 mins], followed by (e) hold at 20°C. There was no purification of the PGR products at this stage.

Each pair of primers was tested in a separate PGR reaction. Primers were tested at 4, 2, 1 and 0.5μΜ concentrations. Each of the primers was tested on a DNA sample already known to be positive or negative for a particular HLA epitope, e.g. by resolution of HLA alleles in the sample to the four-digit level. The PGR products were separated by gel electrophoresis and visualized using ethidium bromide and ultraviolet light. The results are shown in Figures 1 and 2, in which the well numbers correspond to the primers defined in Table 1 above.

Amplification using each primer pair resulted in a PGR product of the expected size when the sample was known to be positive for the HLA epitope specific for the primer (see Figure 1). In the majority of cases when a primer pair was added to a sample which was known to be negative for the epitope recognized by the primers, no band at the expected size was seen. However, in a small number of cases an unexpected band was seen to be produced by a known negative sample, i.e. a false positive was produced (see Figure 2).

A possible explanation for the false positive results could be the presence of the many pseudogenes and gene fragments which are homologous to the expressed HLA genes. When considering the HLA-A locus, HLA-H is likely to be responsible for false positive reactions, as this pseudogene is thought to be due to a gene duplication event involving HLA-A. Therefore a very high level of sequence similarity is found between HLA-A and HLA-H. HLA-G is also similar to HLA-A and may lead to further false positive results.

Therefore in a modification of the method, HLA-A locus specific primers were used to first amplify the HLA-A region, and then this amplified product was used to test the epitope specific primers. To amplify the HLA-A locus, the following cycling conditions were used: (a) 96°C for 3 minute, followed by (b) 26 cycles of [96°C for 25 sec, 65°C for 45 sec and 72°C for 30 sec], followed by (c) 4 cycles of [96°C for 25 sec, 55°C for 1 min and 72°C for 2 mins], followed by (d) 72°C for 5 mins, followed by (e) hold at 4°C. The final volume of reaction mix per well was 50μ1, which contained Ix Reddy Buffer IV (Abgene), Ix Buffer IV (Abgene), 0.2mM dNTPs (Abgene), 0.2μΜ 63/64 control primer, 2.5mM MgCl ₂ (Abgene), 2 units Taq polymerase (Abgene) and 50ng DNA.

The HLA-A specific primer sequences used were taken from Hurley et al 13 ^th IHWS Technology Joint Report - Typing for Class I HLA-A alleles with sequence specific oligonucleotide reagents: Primers and probes. The HLA-A primer sequences used were as follows: A Locus Forward Primer 1 - GGCCTCTGTGGGGAGAAGCAA (SEQ ID NO: 155); A Locus Forward Primer 2 - GGCCTCTGCGGGGAGAAGCAA (SEQ ID NO: 156); A Locus Reverse Primer - GTCCCAATTGTCTCCCCTCCTT (SEQ ID NO: 157). These were used to amplify the HLA-A locus of a selection of random samples. A band of approximately 1 OOObp showed the amplification had been successful (see Figure 3). These samples were then diluted (1 in 100,000, decided after testing several dilutions) and the amplified products tested using the epitope specific primers as described above. 10 random genomic DNA samples were selected to test the epitope specific primers. In specific embodiments, each sample was tested with and without amplifying the HLA-A locus. The results are shown in Figures 4 and 5, in which the well numbers correspond to the primers defined in Table I. When the primers were tested with the non-HLA-A amplified DNA, additional bands can be seen in some of the wells (see Figure 4, e.g. lanes 17-19, 50-53) when compared with the wells corresponding to the HLA-A amplified DNA. These reactions are due to the same epitopes being present on alleles other than those of the HLA-A locus. These false-positive bands are removed by the HLA-A locus-specific amplification step.

In some cases, when the primers are tested with the HLA-A amplified DNA, additional bands representing an unexpected fragment size were also seen in some wells (see Figure 5, e.g. lanes 1, 2, 9, 10), when compared with the wells corresponding to non-HLA-A amplified DNA. However these bands could be removed by performing a DNA purification step on the product of the first PGR amplification using HLA-A locus- specific primers, before amplification using the epitope-specific primers. For example, the product of the initial PGR step was purified using the QIAquick Spin Purification procedure. This removes excess primers and nucleotides which may influence the second PGR reaction. Following HLA-A locus-specific amplification and DNA purification, amplification with each of the epitope-specific primer pairs produced a band of the predicted size with no unexpected bands (see Figure 6).

10 random genomic DNA samples known to encode particular HLA-A epitopes were tested using this method. 75 separate reactions were performed on each sample in order to determine the reproducibility of the method. Below shows the results when the epitope specific primers were tested using genomic (non-HLA-A locus amplified) DNA and HLA-A locus amplified and purified DNA. The table below shows the number of times that the expected results were obtained. The results were read blind and of the 75 separate reactions the result column shows the number of times the results were correct (i.e. as expected). R is the correlation coefficient, and the results show that this is above 0.9 for the HLA-A amplified and DNA purified samples. Genomic DNA

DNA number Result R =

28950 70/75 0.93

28951 62/75 0.83

28952 68/75 0.91

28953 61/75 0.81

28954 65/75 0.87

28955 65/75 0.87

29611 66/75 0.88

29612 67/75 0.89

29613 67/75 0.89

29614 63/75 0.84

Average R = 0.872

A amplified 'cleaned' DNA

DNA number Result R =

28950 72/75 0.96

28951 72/75 0.96

28952 73/75 0.97

28953 71/75 0.95

28954 69/75 0.92

28955 70/75 0.93

29611 70/75 0.93

29612 73/75 0.97

29613 70/75 0.93

29614 70/75 0.93

Average R = 0.945 Example 2: Epityping at HLA-C Locus using SSP

In this example, sequence-specific PGR primers were designed to detect the HLA-C epitopes identified by El-Awar et al 2007 as described in Example 1. The primers are shown in Table 2 below. The numbers in brackets following each primer sequence are SEQ ID NOs.

Table 2

73Tf CAAGCGCCAGGCACAGACT 90Arl CGGGGTCACTCACCGGC (209) 80 18

(73T)+76V+ (208) 90Ar2 CGGGGTCACTCACTGGC (210)

80N+90A (73T)+76V+80N+90A

76Vf GGCACAGACTGACCGAGTG 90Arl CGGGGTCACTCACCGGC (212) 71 19

(73TH-76V+ (211) 90Ar2 CGGGGTCACTCACTGGC (213)

80N+90A (73T)+76V+80N+90A

These primers are used to amplify DNA samples in PGR reactions using similar methods to those as described in Example 1. HLA-C locus specific primers may be used in an initial amplification step on the sample, and optionally a DNA purification step may be performed on the PGR product before amplification using the epitope-specific primers.

Each of the applications and patents mentioned in this document, and each document cited or referenced in each of the above applications and patents, including during the prosecution of each of the applications and patents ("application cited documents") 'and any manufacturer's instructions or catalogues for any products cited or mentioned in each of the applications and patents and in any of the application cited documents, are hereby incorporated herein by reference. Furthermore, all documents cited in this text, and all documents cited or referenced in documents cited in this text, and any manufacturer's instructions or catalogues for any products cited or mentioned in this text, are hereby incorporated herein by reference.

Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments and that many modifications and additions thereto may be made within the scope of the invention. Indeed, various modifications of the described modes for carrying out the invention which are obvious to those skilled in molecular biology or related fields are intended to be within the scope of the claims. Furthermore, various combinations of the features of the following dependent claims can be made with the features of the independent claims without departing from the scope of the present invention.