11111llBO-

hnmunogenetics 36: 277-282, 1992

genetics

© Springer-Verlag 1992

Original articles Molecular typing of HLA-B27 alleles O. Dominguez 1, E. Coto 1, E. Martinez-Naves ], S.Y. Choo 2, C. L6pez-Larrea ~ Inmunologfa, Hospital Central, 33006 Oviedo, Spain 2 Fred Hutchinson Cancer Research Center, Seattle, WA, USA Received November 14, 1991; revised version received January 16, 1992

Abstract. HLA-B27 represents a family of closely related antigens. Six alleles which differ in a limited number of nucleotide substitutions have been described (B*2701-B'2706). These changes are clustered in c~l and c~2 domains. Polymerase chain reaction strategies were designed to amplify specific regions of class I exons 2 and 3. Amplified sequences were tested with eight sequencespecific oligonucleotides to distinguish all B27 subtypes. We also subtyped B27 in 50 healthy Spanish individuals using this procedure. The B'2705 subtype is overrepresented in our population (96 %). The remaining 4 % carried the B'2702 allele. This finding is in agreement with the frequencies described by other techniques (cytotoxic T lymphocytes and isoelectric-focusing) for Caucasian populations. Class I oligotyping is a poorly developed field with significant potential applications. This procedure of genotyping B27 alleles is a reliable method which can be used in transplantation and B27-associated disease studies.

Introduction Cloning and sequencing of the polymorphic exons of HLA class II has made possible the use of the polymerase chain reaction (PCR) and sequence-specific oligonucleotides (SSOs) for the genotyping of HLA alleles (Saiki et al. 1986; Morel et al. 1990). With increasing knowledge of class I sequences it should be possible to design a PCR/SSO typing system as has already been done for class II alleles. The polymorphism of class I genes has been classically defined by serological techniques. Additional techniques such as isoelectric-focusing (IEF), cytotoxic T lymphocyte (CTL) typing, and nucleotide or amino acid sequencing have revealed further diversity within the serologically defined alleles (Neefjes et al. 1986; Yang

1989; Beatty et al. 1989). Analysis of sequenced alleles reveal a high degree of polymorphism located mainly in exons 2 and 3 (Bjorkman and Parham 1990; Zemmour and Parham 1991). HLA-B27, initially identified by serologically, has been shown to be polymorphic using different techniques including CTL, IEF, and restriction fragment length polymorphism (RFLP) analysis (Breuning et al. 1982; Choo et al. 1986; Ness and Grumet 1987; Calvo et al. 1990). Six B27 alleles (B'2701-06) were described at the Tenth International Histocompatibility Workshop (Van Seventer et al. 1989). The six B27alleles were studied by protein (Ezquerra et al. 1985a; Vega et al. 1985a, b, 1986; Rojo et al. 1987a, b) and nucleotide sequencing (Weiss et al. 1985; Seemann et al. 1986; Choo et al. 1988). The polymorphic differences are concentrated into four regions which comprise both exons 2 and 3 (Fig. 1). In this study we describe specific DNA amplification by PCR suitable for B27oligotyping. Standardization of this method makes it a simple, rapid, and accurate procedure for DNA typing of HLA-B2 7 alleles which can be used in a variety of applications.

Material and methods Samples. Previously characterized B-lymphoblastoid cell lines were used as positive reference controls for the six different B27 subtypes described at the Tenth International Histocompatibility Workshop (Van Seventer et al. 1989; Table 2). Serologically typed 20 HLA-B27 negative samples containing 31 different HLA-B alleles were used to determine

EXON 2

B*'2702 B*2 703 B ~ 704 B*2 706 B*2706

EXON 3

DRENLRIA DREDLRTL DRESLRTL DREDLRTL DRESLRTL

Address correspendence and offprint requests to: C. L6pez-Larrea, Ser-

Fig. l. Correlation between the B2 7subtypes and the sequence variations

vicio de Inmunologia, Hospital Central, 33006 Oviedo, Spain

of exon 2 and 3.

O. Dominguez et al. : HLA-B27subtyping by PCR and SSO probes

278

Table 1. Allele-specific oligonucleotide probes used for analysis of B27 subtypes PCR primers

Exon

Sequence (5'--3')

Residues

FA0s E90as E91s E 136as Or-1 Or-2

exon 2

GCCGCGAGTCCGAGAGA GGCCTCGCTCTGGTTGTA GGGTCTCACACCCTCCAGAAT CGGCGGTCCAGGAGCT AGGCACTCTTCCAGCGTTCC TCTTTGCGGATGTCCACGTCA

AASPRE YNQSEA GSHTLQN SSWTAA

exon 3 /3-actin

SSOs

B'27 subtype

CL-1 CL-2 CL-3 CL-4 CL-5 CL-6 CL-7 PAN-B27

2701 AAGGCACAGACTTACCGAGAGAAC 2702 AAGGCACAGACTGACCGAGAGAAC 2703 CCGGAGCATTGGGACCGG 2704,06 GACCGAGAGAGCCTGCGGACCCTG 2705,03 GACCGAGAGGACCTGCGGACC 2706 TATGACCAGTACGCCTA 2701,02,03,04,05 TACCACCAGGACGCCTAC 2701,02,03,04,05,06 CACAGATCTGCAAGGCCAAGGCAC

KAQTYREN KAQTDREN PEHWDR DRESLRTL DREDLRT YDQYAY YHQDAY TQICKAKA

Th

Tw

50 o 50 o 50 o 60° 60 o 40 o 50 o 60 o

67 o 67 o 61 o 67 o 66 o 51 o 55 o 67 o

(40-45) (85-90) (91-97) (131- 6) (1349-68) (1523-43)

(70-77) (70-77) (57-62) (74-81) (74-80) (113- 8) (113- 8) (65-72)

The specifity of SSO probes is based on nucleotide and amino acid sequences reported previously (see Material and methods). Tw: Stringent washing temperature; Th: Hybridization temperature.

the specificity of the PCR/SSO method (Table 3). In addition, 50 HLAB27-positive healthy donors were chosen to ascertain the distribution of B27 subtypes in the Spanish population.

PCR. Primers E40s and E90as were chosen in order to amplify a region from codons 40-90 (exon 2) of HLA-B-related alleles (Table 1). Primers E91s and E136as were designed to amplify codons 91 to 136 of exon 3 from HLA-B27alleles. Briefly, 10 ng of genomic DNA per gl containing 0.8 pM of each primer were amplified by PCR procedures for 30 cycles using DNA polymerase (Saiki et al. 1986); 0.1 gM of/3-actin was included as an internal control in the exon 3 amplifications. Genomic DNA was mixed with PCR buffer (17 mM (NH4)2SO4, 67 mM Tris HCI, pH 8.8, 1.5 mM MgSO4, 0.025% NP-40). The cycling reaction was 96 °C for 10 s, 65 °C for 20 s and 70 °C for 20 s, with the initial denaturation (97 °C, 5 min) and the last extension step (70 °C, 5 min).

Oligonucleotidetyping. Table 1 summarizes the sequences of the (SSOs) used in this study. The SSOs CL-2, CL-3, CL-5, CL-7, and PAN-B27 SSOs were derived from sequences reported previously (Weiss et al. 1985; Seeman et al. 1986; Choo et al. 1988). CL-1, CL-4, and CL-6 were deduced from overlapping of HLA-class I nucleotide sequences that share the same amino acid region (Bjorkman and Parham 1990; Zemmour and Parham 1991). The position and sequence of the SSOs are given in Figure 2. One Ixl of PCR products was denatured (NaOH) and immobilized to Zeta Probe membranes (BioRad, Richmond, CA). Filters were prehybridized at 65 °C for 30 min in 2 x SET 8.5 (0.3 M NaCI, 1 mM ethylenediaminetetraacetate (EDTA), 20 mM Tris HC1 pH 8.5), 2% sodium dodecyl sulfate (SDS), 5 x Denhardt's solution, and hybridized with 32p-end labeled SSOs in the same solution at 106 cpm/ml for 1 h. The hybridization temperatures are shown in Table 1. Twenty four-mer SSOs were washed in TMAC solution (3 M TMAC, 5 mM EDTA, 20 mM Tris HC1, pH 8.5, 1% SDS), whereas the other oligos were washed in 2 x SET 8.5, 1% SDS.

Results

HLA-B2 7-specific amplification. T y p i n g for the different B27 subtypes was p e r f o r m e d by amplification o f e x o n 2

and e x o n 3 sequences. In o r d e r to a v o i d S S O crosshybridizations, attempts w e r e m a d e to specifically amplify B27 subtypes. E x o n 2 was amplified by using specific p r i m e r s E40s and E90as flanking codons 4 0 - 9 0 . A c c o r d i n g to the sequences o f HLA class I genes recently s u m m a r i z e d ( Z e m m o u r and P a r h a m 1991), p r i m e r FA0s was designed to allow extension o f only s o m e o f the HLAB alleles (-B7,-B14,-B8,-B38,-B39,-B42) and all B27subtypes. T h e 3'-end o f E40s w o u l d avoid the amplification of HLA-A,-C, and non-classical class I genes (E,F, G). The a m p l i f i e d products [150 base pair (bp), Fig. 3] containing the h y p e r v a r i a b l e regions o f e x o n 2 w e r e used to e x a m i n e the HLA-B2 7 subtypes (B'2701,02,03,04/06,05). The typing specificity was obtained with specific probes (CL-1, - 2 , - 3 , - 4 , - 5 , P A N - B 2 7 ) . To r e s o l v e B'2704 f r o m B'2706, it was necessary to analyze the e x o n 3. Because p o l y m o r p h i c sequences o f e x o n 3 distinguishing b e t w e e n the different B27 alleles are present on HLA-C alleles, P C R p r i m e r s that only amplify the B2 7 sequences o f exon 3 w e r e used. A r e g i o n o f e x o n 3 limited by p r i m e r s E91s and E136as ( 9 1 - 1 3 6 codons), containing the changes that differentiate the B2 7 subtypes, was amplified (135 bp). M i s m a t c h at the 3' end o f E91s with non-B27 alleles assured that only B27 alleles w e r e amplified. All B2 7 p o s i t i v e samples (n = 60) w e r e amplified. N o n e o f the B2 7 n e g a t i v e samples (n = 20) y i e l d e d P C R products. T h e reaction efficiency was d e m o n s t r a t e d in n e g a t i v e samples by simultaneous amplification o f ¢3-actin g e n e fragments. Thus, B2 7 p o s i t i v e samples g a v e both P C R products, that is, B27exon 3 (135 bp) and/3-actin (100 and 195 bp) g e n e fragments. O n the other hand, B2 7 n e g a t i v e samples g a v e only the/3-actin f r a g m e n t (Fig. 3). Subsequently, B*2 704 and B ' 2 7 0 6 subtypes w e r e defined with specific probes (CL-6, CL-7).

O. Dominguezet al. : HLA-B27 subtyping by PCR and SSO probes

Consensus

279

190 GCCGCGAGTCCGAGAGAGGAGCCGCGGGCGCCGTGGATAGAGCAGGAGGGGCCGGAGTATTGGGACCGGGAGA sense

B-2701 B-2702 B'2703 B'2704 B'2705 B'2706

262

.... ******************************************************** .....

C--

263

342

CACAGAACTTCAAGGCCCACGCACAGAcTGAcCGAGAGAACCTGCGGAAC•TGCGCGGCTACTACAACCAGAGCGAGGcC antisense B'2701 ]3*2702 ]3*2703 13"2704 13"2705 ]3*2706

EXON

......

T--G

.......

~-G

...... ......

T--G T--G

....... .......

~-G A-G

......

T--G

.......

A-G .... *****

[- . . . . . ......

T ..........

T--G .......

A-G---+---q

T--G

A - G . . . . * * * * * ~ "I

.......

********************* t ...... G

T-GC--T-C C ..... T-C

G

C .... *********

G

C _ L ~ _ _I .~ * *7* * *.* *

$

.......

Cq---=T-C ..................

3

Consensus

343 GGGTCTCACACCCTCCAGAATATGTATGGCTGCGACGTG~CGGACGGGCGCCTCCTCCGCCG sense

407

B'2701 13"2702 B'2703 B'2704 B'2705 B'2706 408 TATGACCAGTACGCCTACGACGGCAAGGATTACATCGCCCTGAACGAGGACCTGAGCTCCTGGACCGCCG

476 antisense

B-2701 13-2702 B-2703 B-2704 B-2705

--CO --CC --CC ~-CC --CC

B-2706

[

..... ..... ..... ..... .....

G ........ G G G G

************************************

~ ************************************

................

Fig. 2. Nucleotide sequence comparisons of the exon 2 and 3 of HLA-B27 alleles. Partial sequences of B'2701, B'2704, and B'2706 have been predicted (see Material and methods). The boxes indicate the sequence recognized by the eight SSOs. Numbering refers to codon position. The bars show the position of the primers used for the amplifications.

Fig. 3. Exon 2 and B27-specific exon 3 amplifications. Samples 1-4 (B27 + ) were amplified by using Edu40s and Edu90as primers (exon 2), giving an expected 150 bp band. Samples 5-12 were amplified by primers E91s and E136as (exon 3.) The expected bands (135 bp) were only detected in the B27+ samples (lanes 5-8); this band was not seen in all B27- samples (lanes 9-12). Simultaneous amplifications of/3-actin gene, included as PCR controls, gave 100 and 195 bp fragments (lanes 5-12).

Oligotypingfor B2 7subtypes. SSOs were chosen to define six subtypes of B27 based on polymorphism in codons 59,74,77,80,81 (exon 2), and 114,116 (exon 3, Fig. 2). Figure 4 shows the hybridization reactions for the six B2 7 alleles from the control samples. The patterns of hybridization with the selected oligonucleotides are shown in Table 2. All HLA-B2 7 subtypes were identified by the oligo PAN-B27 (codons 65 to 72; Table 1 and Figure 4). The B*2701,B*2702, and B'2705 can be unambiguously defined with regard to the exon 2 by oligotyping analysis (codons 74-81) with the probes CL-1 ,CL-2, and CL-5, respectively. The B'2703 differs from B'2705 subtype by a single change at position 59 (His). Probe CL-3 uniquely recognizes and defines B'2703 (codons 57-62). B'2704 and B'2706 differ from each other at position 114 and 116. The identification of B'2704 and B'2706 alleles requires the combination of probes CL-4 (exon 2) and CL-6/CL-7 both hybridizing

280

O. D o m i n g u e z et al.:

with exon 3. B'2704 was defined by hybridization with oligo CL-4 and CL-7. B'2706 was identified with CL-4 and CL-6 (codons 113-118). Genotypes could be assigned unambiguously to each cell line. It is of interest that the pattern of SSOs hybridization of the exon 2 (CL1,CL-2,CL-4,CL-5) and exon 3 (CL-6,CL-7) are mutually exclusive, being positive with only one probe for each exon (Fig. 4). This oligotyping method has been used to test a panel of 20 B2 7negative controls containing

subtyping b y P C R a n d SSO probes

HLA-B27

most of the HLA class I alleles defined previously (Table 3). The results correlated with the typing data, being negative for all SSOs ofexon 2. All samples were negative for exon 3 amplification. The overall data confirm the specificity of both SSO probes (exon 2) and PCR amplification (exon 3).

Allelicfrequencies of B27. This method was used to type a panel of 50 B27 positive cells. Figure 5 shows the hybridization pattern of the B27 subtypes. The B27 Spanish population examined shows a distribution similar to other Caucasian populations. Only B*2 702 and B*2 705 subtypes were represented in our population. B'2701, B'2703, B'2704, and B'2706 were not found in this panel, as was expected from the frequencies of B2 7 subtypes described in Caucasoids by CTL and IEF analysis (Breur-Vriesendorp et al. 1985; Choo et al. 1986). Nevertheless, B'2702 variant (4%) was found slightly underrepresented in comparison with other Caucasian populations. In addition we observed higher frequencies for B'2705 subtype (96%).

Discussion The great number of closely related HLA class I genes and the high degree of polymorphism has limited the use of oligonucleotide typing. With the increasing knowledge of class I sequences it should be possible to use SSOs for genotyping HLA class I. Here we describe the first procedure of genotyping class I antigens of the B27 family that could be used in other studies of related alleles. This method allows the unambiguous assignment of the six B27 allelic variants. Heterozygote B2 7 genotypes can also be detected. Only the infrequent genotype combination B'2703/2705 remains ambiguous (B'2703/-). This

F i g . 4. H L A - B 2 7 o l i g o t y p i n g analysis. P C R - a m p l i f i e d D N A f r o m six positive cell lines containing the six B 2 7 subtypes was hybridized with the probes s h o w n (Table 2). P a n B 2 7 , C L - 1 , C L - 2 , C L - 3 , C L - 4 , a n d C L - 5 hybridize to the exon 2 a n d C L - 6 a n d C L - 7 probes hybridize specifically to exon 3. H o m o z y g o u s typing cell controls are as follows: L H ( B * 2 7 0 1 ) , R56(B*2702), CHI(B*2703), JSL(B*2704),

B27

LG15(B*2705),

PAR(B*2706).

T a b l e 2. S u b t y p i n g o f

HLA-B2Z

SSOs Cell line LH* R56 R34 * CHI* LAR WEWAK-I* JSL KNE LG15* R69 PAR LIE*

CL-1

CL-2

CL-3 .

CL-4 .

.

CL-5

CL-6

.

CL-7

PAN-B27

B27

alleles

"2701

+

.

+

+

-

+

.

.

.

.

+

+

*2 702

-

+

.

.

.

.

+

+

*2 702

-

-

+

-

+

-

+

+

*2703

-

-

+ -

+

+ -

-

+ +

+ +

*2 7 0 3 *2 7 0 4

--

-

-

+

-

-

+

+

*2704

-

-

-

+

-

-

+

+

*2 704

+ +

-

+ +

+ +

*2705

. .

. .

. .

. .

*2705

-

-

--

+

-

+

-

+

*2 706

-

-

-

+

-

+

-

+

*2706

+ symbols indicate positive hybridization, and - negative hybridization. * D o n o r s used for sequence analysis (see Introduction).

O. Dominguez et al.: HLA-B27 subtyping by PCR and SSO probes

281

Table 3. HLA types of the B27 negative controls used in the study.

1986). HLA evolution studies can be performed by this approach on class I-related genes that have recently diversified in different ethnic groups (i.e., A2). Phylogenetic analysis of the HLA-B27 relationship has suggested gene conversion, point mutation, and recombination events (L6pez de Castro 1989). The B'2706 could have arisen from B'2704 with sequences at position 114-116 (DQY) of Cwl as a potential donor. This correlates with the HLAB27/Cwl linkage disequilibrium described previously (Breur-Vriesendorp et al. 1988). This observation can indicate that codon exchange between linked alleles of different HLA loci can also contribute to HLA diversification in humans. Typing for B2 7 alleles with SSOs can also be applied to the study of HLA-B2 7-associated disease. This gene has been shown to be strongly associated with different spondyloarthropathies (Schlosstein et al. 1973; Woodrow, 1977). Conserved structural motif shared by B'2701-06 has been suggested to be involved in ankylosing spondylitis (AS; Benjamin and Parham 1990). Interestingly, absence of AS disease has been observed in West African populations where the B'2703 subtype was found overrepresented (Hill et al. 1991). The recently reported crystalographic data of B27 (Madden et al. 1991) can contribute to identifying the specific sequence motif involved in antigen presentation (Benjamin et al. 1991). Finally, matching for HLA class I and II antigens is essential for case bone marrow transplantation. Very precise HLA-matching appears to be critical in minimizing the risk of rejection and graft-vs-host disease. Bone marrow transplant rejection associated with a single amino acid change in HLA-class I antigens has recently been observed (Fleischhauer et al. 1990; Parham 1991). These molecular variations can be analyzed by the PCR/SSO system and HLA matching in unrelated bone marrow transplantation could be optimized.

Awl Aw3 Aw4 Aw7 Aw9 Awl0 Awl4 Awl5 Awl8 Aw20 Aw23 Aw25 Aw26 Aw28 Aw34 Aw36 Aw37 Aw47 Aw49 Aw50

A

B

C

1, 11 2, 23 1, 33 9, 29 2, 24 1, 2 2, 33 32, 68 1, 9 26, 32 34, 36 24, 30 2, 36 2, 11 1, 24 1, 26 2, 24 25, x 29, x 2, 29

8, 44 15, 35 22, 62 7, 39 44, 50 37, 52 50, 53 47, 51 8, 18 44, 61 14, 42 13, 40 52, 60 50, 54 44, 55 14, 38 51, 63 57, 76 7, 45 44, 58

w7 w4, w9 w7 w4, w6 w4, w6, w7 w2, w8 w2, w7, w4, w7,

w7

w6 w6 w7 w4 w6 w9 w7 w9

w4 w2 w6 w7

Serologically types HLA-B2 7 negative samples (n = 20) containing most different HLA-B alleles (n=31) were used to determine the specificity of the PCR/SSO method. Samples do not hibridize to either SSOs of exon 2. All samples were negative for exon 3 amplification. These data show the specificity of both, the SSO probes detection (exon 2) and PCR amplification (exon 3).

Fig. 5. HLA-B27 typing of a panel ofB27samples. DNA was amplified with primers as described in Material and methods. The filters containing amplified DNA were hybridized subsequently with the SSO probes (Table 1). Only B'2702 and B'2705 subtypes were represented in the Spanish population and gave positive hybridization with oligos CL-2 (Samples A2,C2) and CL-5 (samples A1 ,B1 ,C1 ,D1 ,El,E2), respectively. PAN-B27 only hybridized B27 sequences and gave negative signals in the B27negafive samples (B2,D2). Twenty negativeB27samples were included as controls, being negative for all of the exon 2 probes.

system can also be utilized to identify new B2 7 alleles which may show the a "patchwork" pattern. The PCR/SSO B27 system can be applied in anthropological studies testing large populations that previously could only be determined by laborious techniques (IEF,CTLs). The B2 7 Spanish population examined showed a distribution similar to other Caucasian populations (Breur-Vriesendorp et al. 1985, 1988; Choo et al.

Acknowledgments. O. Dominguez and E. Martinez-Naves are recipients of FISS and FPI grants, respectively. This work was supported by Fondo de Investigaciones Sanitarias, Ministerio de Sanidad, Spain (FISS).

References Beatty, P. G., Layrisse, Z., Van Seventer, G.A., Fan, L., Magli, C., Ivanyi, P., Jenkin, J., and Hansen, J. A.: T-cell recognition of HLA class I molecules. In Bo Dupont (ed.): Immunobiology of HLA: pp. 359-362, Springer, New York, 1989 Benjamin, R. and Parham, P.: Guilt by association: HLA-B27 and ankylosing spondilitys. Immunol Today 11: 137-142, 1990 Benjamin, R.J., Madrigal, J. A., and Parham, P.: Peptide binding to empty HLA-B27 molecules of viable human cells. Nature 351: 74-77, 1991 Bjorkman, P. and Parham, P.: Structure, function, and diversity of HLA-class I Major Histocompatibility Complex molecules. Annu Rev Biochem 59: 253-288, 1990

282

O. Dominguez et al.: HLA-B27 subtyping by PCR and SSO probes

Breuning, M.H., Lucas, C.J., Breur, B.S., Engelsma, M.Y., de Lange, G. G., Dekker, A. J., Biddison, W. E., and Ivanyi, P.: Subtypes of HLA-B27 detected by cytotoxic T lymphocytes and their role in self recognition. Hum Immunol 5: 259-268, 1982 Breur-Vriesendorp, B. S., Dekker, A. J., Breuning, M., and Ivanyi, P.: Subtypes of antigen HLA-B27 (B27W and B27K) defined by cytotoxic T lymphocytes. In M. Ziff and S.B Cohen (eds.): Advances in Inflammation Research Vol 9. The Espondyloarthropaties, pp. 55-65, Raven Press, New York, 1985

Breur-Vriesendorp, B.S., De Waal, L.P., and Ivanyi, P.: Different linkage disequilibria of HLA-B27 subtypes and HLA-C locus alleles. Tissue Antigens 32: 74-77, 1988 Calvo, V., Rojo, S., Ldpez, D., Galocha, B., and Ldpez de Castro: Structure and diversity of HLA-B27 specific T cell epitopes. J Immunol 144: 4038-4045, 1990 Choo, S. Y., Antonelli, P., Nisperos, B., Nepom, G. T., and Hansen, J. A.: Six variants of HLA-B27 identified by isoelectric-focusing. Immunogenetics 23: 24-29, 1986 Choo, S.Y., John, T.S., Orr, H.T., and Hansen, J.A.: Molecular analysis of the variant alloantigen HLA-B27d (HLA-B*2703) identifies a unique single amino acid substitution. Hum Immunol 21: 209-219, 1988 Ezquerra, A., Bragado, R., Vega, M.A., Strominger, J. L., Woody, J.M., and Ldpez de Castro, J.A.: Primary structure of papainsolubilized human histocompatibility antigens HLA-B27. Biochem 24: 1733-1744, 1985 Fleischhauer, K., Kernan, N. A., O'Reilly, R. J. B., and Yang, S. Y. : Bone marrow-allografl rejection by lymphocytes recognizing a single aminoacid defference in HLA-B44. N Engl J Med 323: 1818-1822, 1990 Hill, A. V. S., Allsopp, C. E. M., Kwiatkowski, D., Anstey, N.M., Greenwood, B. M., and McMichael, A. J.: HLA class I typing by PCR: HLA-B27 and an African B 27 subtype. Lancet337: 640-642, 1991 L6pez de Castro, J.A.: HLA-B27 and HLA-A2 subtypes: structure, evolution and function, lmmunol Today 10: 239-246, 1989 Madden, D. R., Gorga, J. C., Strominger, J. L., and Wiley, D. C.: The structure of HLA-B27 reveals nonamer self-peptides bound in an extended conformation. Nature 353: 321-325, 1991 Morel, C., Zwahlen, F., Jeannet, M., Mach, B., and Tiercy, J.M.: Complete analysis of HLA-DQB1 polymorphism and DR-DQ linkage disequilibrium by oligonucleotide typing. Hum Immunol 229: 64-67, 1990 Neefjes J.J., Breur-Vriesendorp, B. S., Van Seventer, G.A., Ivanyi, P., and Ploegh, H.L.: An improved biochemical method for the analysis of polymorphism of HLA class I antigens. Definition of new HLA class-I subtypes. Hum lmmunol 16: 169-181, 1986 Ness, D.B. and Grumet, F.C: New polymorphism of HLA-B27 and other B-locus antigens detected by RFLP using a locus-specific probe. Hum lmmunol 18: 65-73, 1987

Parham P.: Making just the right match. Nature 350: 111-112, 1991 Rojo, S., Aparicio, P., Choo, S. Y., Hansen, J. A., L6pez de Castro., J. A.: Structural analysis of an HLA-B27 population variant, B27f. Multiple patterns of aminoacid changes within a single polypeptide segment generate polymorphism in HLA-B27. J lmmunol 139: 831-836, 1987a Rojo, S., Aparicio, P., Hansen, J.A., Choo, S.Y., and L6pez de Castro, J. A. : Structural analysis o fan HLA-B27 functional variant, B27d, detected in American Blacks. J Immunol 139: 3396-3401, 1987b Saiki, R. H., Bugawan, T. L., Horn, G. T., Mullis, K. B., and Erlich, H.A.: Analysis of enzymatically amplified /3-globin and HLADQc~ DNA with allele-specific oligonucleotide probes. Nature 324: 163-166, 1986 Schlosstein, L., Terasaki, P.J., Bluestone, R., and Pearson, C.M.: High association of HLA antigen W27 with ankylosing spondilitys. N Engl J Med 288: 704-706, 1973 Seemann, G. H. A., Rein, R. S., Brown, C. S., and Ploehg, H. L. : Gene conversion generates polymorphism in human class-I genes. EMBO J 5: 547-552, 1986 Van Seventer, G.A., van der Horst, A. R., Waal, L. P., Reekers, P., and Ivanyi, P.: Antigen Society 13 Report (B7, B27, Bw47, Bw73): Public Determinants of the HLA-B7 CREG defined by Antibodies and T-Cell Clones. In Bo Dupont (ed.): Immunobiology of HLA, pp. 204-206, Springer, New York, 1989 Vega, M. A., Walace, L., Rojo, S., Aparicio P., and L6pez de Castro, J.A.: Delineation of functional sites in HLA-B27 antigens. Molecular analysis of HLA-B27 variant WewaK I defined by cytolytic T lymphocytes. J Immunol 135: 3323-3332, 1985a Vega, M. A., Ezquerra, A., Rojo, S., Aparicio, P., Bragado, R., and L6pez de Castro J.A.: Structural analysis of an HLA-B27 functional variant: identification of residues that contribute to the specificity of recognition by cytotytic T lymphocytes. Proc Natl Acad USA 82: 7394-7398, 1985b Vega, M.A., Bragado, R., Ivanyi, P., Pelaez, J.L., and L6pez de Castro, J. A.: Molecular analysis of a functional subtype of HLAB27. A possible evolutionary pathway for HLA--B27 polymorphism. J Immunol 137." 3557-3565, 1986 Weiss, E. H., Kuon, W., Dorner, C., Lang, M., and Riethmuller, G.: Organization, sequence and expression of the HLA-B27 gene. A molecular approach to analyze HLA and disease associations. Immunobiology 170: 376-380, 1985 Woodrow J.C.: Histocompatibility antigens and rheumatic diseases. Sem Arth Rheum 6: 257-276, 1977 Yang S. Y.: Population analysis of class I HLA antigens by one-dimensional isoelectric focusing gel electrophoresis: workshop summary report. In Bo Dupont (ed.): Immunobiology of HLA, pp. 30%331, Springer, New York, 1989 Zemmour, J. and Parham, P. : HLA class I nucleotide sequences. Tissue Antigens 37: 174-180, 1991