Restriction fragment length polymorphisms and an HLA-DRw52-associated split

Restriction Fragment Length Polymorphisms and an HLA-DRw52-Associated Split May Chatila, Lorie Luyrink, James McEleney, Thomas Spies, Jack L. Strominger, and David N. Glass

ABSTRACT: The restriction fragment length polymorphism technique, employing seven restriction endonucleases, and a DR-specific repeatedelementprobe were used to study a large pand of horaozygous typing cells in order to delineate haplotypic differences within the HLA-DRw52 supertype. Most of the restriction endonucleases revealedthe presence of two allelic restrictionfragment length polymorphisms correlatipg with and splitting the HLA-DRw52 supertypic specificity. One designated A, correlated with the HLA-DRw52 typing in HLA-DR5, HLA-DR3, (non-A1, B8), and some HLA-DRw6 haplotypes. The other, designated a, correlated with the HLA-DRw52 typing in HLA-DR5, HLA-DRw8, HLA-DR3 (A1, B8), and the remaining HLA-DRw6 haplotypes. The general apph'cability of thesefindings was validated in 47 HLA-typed laboratory controls. ABBREVIATIONS RFLP restriction fragment length polymorphism HTC homozygous typing cells SSC standardsaline citrate SDS sodiumdodecyl sulfate INTRODUCTION The restriction fragment length polymorphism (RFLP) technique has been applied to the study of the HLA system at a genomic level [1-6] and has made possible the splitting of many of the HLA types currently defined by alloantisera [7-12]. These subdivisions not only advance our understanding of the HLA genes but also are useful in disease association studies, an example of which is the use of RFLPs to split the HLA-DQ alleles [ 13-16]. The HLA-DR region is perceived to play a crucial role in the pathogenesis of many autoimmune diseases [19-20]. Hence, the detection of RFLPs that split the HLA-DR types and supertypes is particularly worthwhile, especially if they split the HLA-DRw52/53 supertypes encoded by the DR beta III gene, which is

From the Departments of Rheumatology and Immunology, Brigham and Women's Hospital (M.C., L.L., J.McE., D.N.G.), Dana Farber Cancer Institute (J.L.S.), Department of Biochemistry and Molecular Biology (T.S., J.L.S.), Harvard University, Boston, Massachusetts. Address reprint requeststo Dr. David N. Glass, Division of Rheumatology, Children's Hospital Medical Center, Blland and Bethesda Avenues, Pavilion 1-29, Cincinnati, OH 45229-2899. Received May 25, 1987; acceptedNovember 11, 1987.

Human Immunology 21, 89-97 (1988) © American Society for Histocompatibili~ and Immunogenetics. 1988

89 0198-8859/88153.50

90

M. Chatila et al. less polymorphic than the DR beta I gene that encodes the HLA-DR types [ 19]. Splits of HLA-DRw52 and of HLA-DRw53 were not defined during the last International HLA Workshop [21], although some evidence of additional polymorphism had been obtained serologically. The DNA sequence basis for a split of the HLA-DRw52 supertype has recently been described and supported by oligonucleotide probe [22-23]. To facilitate the identification of these RFLPs, we have used an HLA-DR repeated element probe [24]. This probe is specific for the DR region, and unlike most currently available class II probes it lacks crosshybridization to the products of other class II genes, thus ensuring specificity to the detected RFLPs.

MATERIALS AND METHODS Homozygous Typing Cells Thirty-six Epstein-Barr virus-transformed homozygous typing cells (HTC) were maintained and expanded in RPMI 1640 (Gibco Lab, Grand Island, NY) plus 1096 heat-inactivated fetal calf serum (Gibco Lab, Grand Island, NY), penicillin, and streptomycin. Most of the HTC were the gifts from originating laboratories; the rest were purchased from Coriel Institute for Medical Research (Camden, NJ). Updated information on consanguinity and HLA typing of the HTCs used (Table 1) was generously provided by Dr. J. Bodmer. Normal Population Forty-seven unrelated, healthy, HLA-typed volunteers were used as controls. Cells Peripheral blood mononudear leukocytes from the 47 normal individuals were obtained by density gradient centrifugation through Ficoll-Paque (Pharmacia, Piscataway, NJ). HLAoA, -13, -C Typing HI~-A, -B, and -C typing for 39 specificities was carried out by microdroplet complement-dependent cytotoxicity (CDC).

HI~-DR Typing HI~-DR specificities were typed by complement-dependent cytotoxicity in a Bcell-enriched mononuclear lymphocyte population on nylon wool. Restriction Endonudeases Seven restriction endonucleases were used: BamH I, Taq I, Pst I, Pvu lI, Bgl II, Hind III, and EcoR V (New England Biolabs, Beverley, MA). DNA Extraction and Hybridization Total genomic DNA from HTC and from lymphocyte nuclei of normal individuals was extracted following Bell's modification [25] of the original Kunkel's procedure [26]. An additional step in the extraction from HTC was the addition of RNase (to a final concentration of 100/~g/ml) and incubation at 37°C for 2 h, following the first phenol extraction.

H L A - D R w 5 2 - A s s o c i a t e d RFLPs

91

TABLE 1 HTCS used and their HLA-A, -B, -DR, -Dw, and -DRw typing ~

Name

N

A

B

DR

HOM-2 MAJA METI"E IBw 4 MST PGF BRISTOL-8 WT49 LKT HAR QBL AVL WT51 SJ-AH KT~ PRIESS AS HA JHA SWEIG MICH IDF FPF JVM APD WVB HHK ARNT u t ,A •uT~T ' ~lT JEFF JMF MANN HERLUFF MADURA-T 23.1 DKB

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 ,-*m'x z~ 30 31 32 33 34 35 36

3 2 3 2 3 3 3 3 12 2 1 1 26 1 9 24 24 2 2 24

27 35 5 18 35 7 7 8 13 17 8 8 18 8 14 40 54 15 27

29 2 32 26 69 1 2 1 2 3 2

40 15 27 38 18 35 18 60 16 7 38 39

23 29 2 2 2 24

44 44 44 35 40 27 60

1 1 1 1 2 2 2 14 3 3 3 3 3 4 4 4 ?X 4 4 4 4 5 5 5 5 5 13a 13a 13a 13a i3b w6 7 7 w12 w8 X w8 w8

Dw

DRw52/53

1 1 1 1 2 2 3 3 3 3 3 4 15 15 4 4

5 5 ?X BLANK 18 18 18 6 (18)

7 7 DB6 8 8 DB5

52 52 52 52 52 52 53 53 53 53 53

52 52 52 52 52 52 52 52 52 52 52

C 0 N

H 0 M

N N N Y N Y N Y N Y Y Y Y

Y Y

N N

y Y y Y Y Y Y Y Y Y Y

Y N N Y Y Y Y Y N

53 53 52 52 52 53

Y Y N Y N Y

Y Y Y Y Y

Y Y Y Y Y

a Abbreviations: CON, consanguineous; HOM, homozygous; 13a, local split of D R w l 3 associated with D w l 8 ; 13b, local split of DRwl3 associated with Dw19; N, code number HTC; Y/N, yes/no.

Ten micrograms of D N A were digested with restriction endonuclease (10 units//zg of D N A , but not exceeding 10% of the total volume) in a total volume of 100 Izl for 18 h. The digested D N A was electrophoresed on 0.7% agarose horizontal gels for 1 8 - 2 0 h, then transferred to nitrocellulose filters (Schleicher & Schuell, Keene, N H ) in 2 0 x standard saline citrate (SSC) [27]. The probes were nick-translated to a specific activity of approximately 109 cpm/0zg of D N A . Filters were prehybridized and hybridized at 42°C in 50% formamide, 5 x Denhardt's solution, 6 x SSC, 0.5% sodium dodecyl sulfate (SDS), 10 mM EDTA, 10 mM Tris CI at pH 7.5, and 100 Izg/ml of denatured salmon sperm D N A . Following three low-stringency washes, the filters were washed in 0.2 x SSC-0.5% SDS at 55°C for I h, and then in 0.1 x SSC-0.5% SDS at 55°C for an

92

M. Chatila et al. additional hour. Filters were then exposed with Kodak XAR-5 films at -70°C. Films were developed in 24 h and again after 3 days.

Probe The two-kilobase (kb) HLA-DR-specific repeated element probe, which represented the signal sequence and promoter region of the HLA-DR beta genes, was genomicaUy derived from the Priess HLA-DR4 cell line. In addition, three similar repeats not related to any beta gene are found in the DR region [24]. Restriction Fragment Length Polymorphism Analysis Lambda Hind III digests were used as size markers for each RFLP. Faint and questionable bands were not considered for analysis, and bands of similar intensifies with questionable size difference were considered as one. Unique or specific bands refer to bands present exclusively in haplotypes with a given HLA-DR type. RESULTS Restriction Fragment Length Polymorphism Correlations with the HLA-DRw52 Supertype The RFLPs correlating w~th the HLA-DRw52 serologic type were detected (Tables 2 and 3) using EcoR V, we were able to identify an 18.5- and 16.2-kb duplex in HLA-DRw52 positive cell lines only (Table 2). In addition, a pair of BamH I RFLPs, designated A and a, typed for and split serologicaUy defined HLADRw52 (Figure 1); their sizes were 5.8 and 2.9 kb, respectively. All restriction endonucleases used subsequently, except Hind III, revealed equivalent pairs of allelic bands, with minor differences (Table 3). The RFLP A correlated with nonHLA-A1, B8, DR3 (2/2), with most HLA-DR5 haplotypes (5/6), and with some FIGURE 1 BamH I-digested DNA from HTC demonstrating: the pair of aUelic HLA DRw52-associated RFLPs of 5.8 kb (A) and 2.9 kb (*). The numbers correspond to HTCs listed in Table 1.

HLA-DRw52-Associated RFLPs TABLE 2

93

Specific HTC-RFLP correlations with the HLA DRw52+ supertypic family

HLA HLA - DR

2

3

3

3

3

3

5

5

5

Ntt

7

8

9

10

11

12

20

21

22

Size Kb -

Enzyme"

18.5- 1 18`2-

RV

~.PA

8`5.8- B 29- B

11.8- B } 10.0- P 7.7- P 10.7 - B

1.8- P 9.3-H

8,2- RV 5.0o 4.6-

4.8 - H 2.8- H 1.9- H 2.0- P

** N: Code number of HTC (see Table 1). * Enzymes used: BamH I (B), Pst I (P), EcoR V (RV), Pvu II (PV), and Hind III (H).

HLA-DRw6 (2/5) haplotypes; RFLP a correlated with HLA-A1, B8, DR3 (3/3), with HLA-DR5 (1/6), with HLA-DRw8 (2/2), and with the remaining HLADRw6 haplotypes (3/5) (Table 4). (One HLA-DR2, HTC, Bristol -8, typed as both DRw52A and a.) The only HLA-DR5 HTC associated with allele a typed as HLA-DRw12, and the A/a division of HLA-DRw6 haplotypes was not related to HLA-DRwl 3a/13b splits. No HLA-DRw6 (DRw14) HTC were included in this

TABLE 3

Sizes of HLA-DRw52-associated pairs of allelic RFLPs

Enzyme

Akb

akb

BamH I EcoR V Taq Ia Pvu II" Pst I"

5.8 10.0 3.3 6.8 2.3

2.9 >33.0 3.5/3.1 2.5 6.0

"With PVU II, Pst I, and Taq i, not all the bands were specific to the DRw52 supertypic family.

94

M . C h a t i l a e t al.

TABLE 4

Distribution of the HLA-DRw52-associated RFLPs A and a in HTCs

N"

Name of HTC

DR/Dw

A

a

7 8 9 10 11

Bristob8 WT49 LKT HAR QBL

2-14/3/3 3/3 3/3 3/3

+ +

+ + + -

12

AVL

3/3

-

+

20 21 22 23 24 33 25 26 27 28 29 34 35

SWEIG MICH IDF FPF JVM HERLUFF APD WVB HHK ARNT DAUDI MADURA-T 23.1

5/5 5/5/5-?X 5/Blank 5/12/4 13a/18 13a/18 13a/18 13d18 13b/w8/8

+ + + + + + + + -

+ + + +

-

+

w8/8

" N , code n u m b e r o f H T C (see Table 1).

study. One HLA-DRw6 HTC (DAUDI) had a 1.9-kb band with Hind IIl instead of the 2.8-kb band present in all other HLA-DRw52 cell lines (Table 2). Many other bands were shared by several haplotypes belonging to the HLADRw52 supertypic family; they may represent additional splits of HLA-DRw52 (Table 2). The Applicability of HLA-Typed Normal Controls The general applicability of the findings, particularly the HLA-DRw52-associated allelic RFLPs A and a, was tested in 47 HLA-typed controls using BamH I-digested D N A (Table 5). The results were in agreement with the findings in HTCs, in that the A RFLP was present in HLA-DR3, non-A1, B8 (4/5), in HLADR5 (8/9), and in HLA-DRw6 (2/5); and the a RFLP was present in HLA-A1, B8, DR3 (10/10), HLA-DR3, non-Al, B8 (1/5), HLA-DR5 (1/9), HLA-DRw8 TABLE 5

Distribution of the HLA-DRw 52-associated RFLPs A and a in 47 normal laboratory controls H L A - D R haplotypes

RFLP

DRw8

DR3 (A 1, B8)

DR3 ( N o n - A 1, B8)

DR5

DRw6

DR (Blank)

A

0

0

4

8

2

0

a Neither Totals

2 0 2

10 0 10

1 0 5

1 0 9

3 0 5

1 0 1

HLA-DRw52-Associated RFLPs

95

(2/2), and HLA-DRw6 (3/5). In addition, the a allele was present in one control subject who had typed for the HLA-DR4 and HLA DRw52/53 specificities but not for any other DR allele (Table 5). Some difficulty was encountered in assigning the A RFLP because the HLADR2 haplotypes had a 5.5- and 5.7-kb duplex difficult to distinguish in certain cases from the 5.8-kb A RFLP. However, an additional 5.0- and 5.2-kb HLADR2-associated, specific, duplex enabled us to eliminate false-positive findings. DISCUSSION With the help of a DR-specific repeated element probe that eliminated the problem of cross-hybridization with HLA-DQ and DP loci [23], we detected RFLPs that correlate with the HLA-DRw52 supertypic specificity. The aUelic pair of HLA-DRw52-associated RFLPs, A and a, is of particular interest; they not only split the HLA-DRw52 specificity, but also did so in a manner generally consistent with the linkage equilibria between other HLA antigens and HLA-DRw52. In D N A from both the HTCs and the laboratory controls, the A allele was associated with HLA-DR5, HLA-DR3 (non-A1, B8), and some HLA-DRw6 haplotypes. The other allele was associated with HLA-DR3 (A1, B8), HLA-DR5 (w12), HLA-DRw8, and some HLA-DRw6 haplotypes. The HLA-DRw6 haplotypes were least consistent in terms of the A/a associations. This finding may reflect limitation of our sample base, because HLA-DRw6 (Dwl4) HTCs were not available to us. The finding in one control subject of HLA-DRw52 that was defined both genomically and serologically but occurred in the apparent absence of an appropriate HLA-DR beta I gene product is consistent with previous reports [28,29]. Bands equivalent to the A/a RFLPs were readily detected with most of the restriction endonucleases used, suggesting a coding region polymorphism [30]. We suspect that RFLPs equivalent to A and a have been described by other investigators who found that they split the HLA-DR3 and w6 haplotypes, respectively [8,10]. These observations, however, were limited in that it was not possible to detect an overall association with H I ~ - D R w 5 2 because of the exclusion of HLA-DK) and I-ILA-DKw8 hapiotypes. Of more direct relevance are the findings of Gorsld and Mach, who recently described two HI,A-DR beta III haplotypes, a and b, on the basis of D N A sequence analysis and oligonucleotide genotyping of HTC [22,23]. The RFLP designated as a here correlates with Gorski and Mach's HLA-DRw52 a type, and RFLP A correlates with their H I ~ DRw52 b type. These findings provide a means of HLA-DRw52 typing and splitting at a genomic level that is applicable to the study of HLA, including its disease associations. This is of practical importance, because the available HLA-DRw52 specific alloantisera do not permit associated splits to be established. ACKNOWLEDGMENTS

This work was supported by grants from The Lions Clubs of Massachusetts and the New England Peabody Home for Crippled Children, and by NIH Grants AM030486 and CRU-AM-05577, and Multipurpose Arthritis Center Grant AM20580. The authors are grateful to Agnes Maier and Louise Kittredge for editorial assistance. REFERENCES 1. Wake CT, Long EO, Mach B: Allelic polymorphism and complexh~ of the genes for HLA-DR beta-chains--Direct analysis by DNA-DNA hybridization. Nature 300:372, 1982.

96

M. Chatila et al. 2. Anderson M, Bohme J, Anderson G, Moiler E, Thorsby E, Rask L, Peterson P: Genomic hybridization with class II transplantation antigen cDNA probes as a complementary technique in tissue typing. Hum Immunol 11:57, 1984. 3. Inoki H, Ando A, ho M, Tsuji K: Southern hybridization analysis of DNA polymorphism in the HLA-D region. Hum Immunol 16:304, 1986. 4. Hui K, Festenstein H, de Klein A, Grosveld G: HLA-DR genotyping by restriction fragment length polymorphism analysis. Immunogenetics 22:231, 1985. 5. Kohonen-Corish MRJ, Serjeantson SW: HLA-DR beta gene DNA polymorphisms revealed by Taq I correlate with HLA-DR specificities. Hum Immunol 15:263, 1986. 6. Tilanus MG, van Eggermond MC, van der Bijl M, Morolli B, Schreuder GM, De Vries RR, Giphart MJ: HLA class II DNA analysis by RFLP reveals novel class II polymorphism. Hum Immunol 18:265, 1987. 7. Angelini G, Tanigaki N, Tosi R, Ferrara G: Southern blot and microfinger printing analysis of two DR7 haplotypes. Immunogenetics 24:63, 1986. 8. Bontrop R, Tilanus M, Mikulsi M, Van Eggermond M, Termijtelen A, Giphart M: Polymorphisms within the HLA-DR3 haplotypes. Immunogenetics 23:401, 1986. 9. Holbeck S, Kim S, Silver J, Hansen J, Nepom G: HLA-DR4-associated haplotypes are genotypically diverse within HLA. J Immunol 135:637, 1985. 10. Angelini G, De Preval C, Gorski J, Mach B: High resolution analysis of the human HLA-DR polymorphism by hybridization with sequence-specific oligonucleotid probes. Proc Nail Acad Sci USA 83:4489, 1986. 11. Gregersen P, Shen M, LiSong Q, Merryman P, Degar S, Seki T, Maccari J, Goldberg D, Murphy H, Schwenzer J, Yi Wang C, Winchester R, Nepom J, Silver J: Molecular diversity of HLA-DR4 haplotypes. Proc Nail Acad Sci USA 83:2642, 1986. 12. Tilanus M, Morolli B, Van Eggermond M, Schreuder G, de Vries R, Giphart M: Dissection of HLA class II haplotypes in HLA-DR4 homozygous individuals. Immunogenetics 23:333, 1986. 13. Schreuder G, Tilanus M, Bontrop R, Bruining G, Giphart M, Van Rood J, de Vries 1 ) . . ~ A - ~ ~lymorphism ass~'.iat~ with ,-o~: x,.. ..... t......... t,, type .~ diabetes . . . . ..a-.,--.o'o,a . . with monoclonal antibodies, isoelectric point differences, and restriction fragment length polymorphism. J Exp Med 164:938, 1986. 14. Festenstein H, Awad J, Hitman GA, Cutbush S, Groves AV, Cassell P, Oilier W, Sachs JA: New HLA-DNA polymorphism associated with autoimmune diseases. Nature 322:64, 1986. 15. Nepom B, Palmer J, Kim S, Hansen J, Holbeck S, Nepom G: Specific genomic markers for the HLA-DQ subregion discriminate between DR4 + insulin dependent diabetes mellitus and DR4+ seropositive juvenile rheumatoid arthritis. J Exp Med 164:345, 1986. 16. Bohme J, Carlsson B, Wallin J, Moiler E, Persson B, Peterson P, Rask L: Only one DQ-beta restriction fragment pattern of each DR specificity is associated with insulindependent diabetes. J Immunol 137:941, 1986. 17. Inoko H, Ando A, Tsuji K, Matsuki K, Juji T, Honda Y: HLA-DQ beta chain DNA restriction fragments can differentiate between healthy and narcoleptic individuals witn HLA-DR2. Immunogenetics 23:126, 1986. 18. Howell M, Austin R, Kelleher D, Nepom G, Kagnoff M: An HLA-D region restriction fragment length polymorphism associated with celiac disease. J Exp Med 164:333, 1986. 19. Korman A, Boss J, Spies T, Sorrentino R, Okada K, StromingerJL: Genetic complex-

HLA-DRw52-Associated RFLPs

20.

21.

22. 23.

24.

25. 26.

27. 28

29. 30.

97

ity and expression of human class II histocompatibility antigens. Immunol Rev 85:45, 1985. Strominger JL: Biology of the human histocompatibility leukocyte antigen (HLA) system and a hypothesis regarding the generation of autoimmune diseases. J Clin Invest 77:1411, 1986. Albert E, Barbalho T, Baur M, Christiansen F, Deppe H, Keller E, Marshall W, McNicholas A, Luten T, Raffoux C, Schiessl B, Scholz S, Sierp G, Sigmund M: The serological analysis of the data of the Ninth Histocompatibility Workshop. In Albert et al (eds): Histocompatibility Testing 1984. Heidelberg, Springer-Verlag, 1984, p. 72. Gorski J, Mach B: Polymorphism of human la antigens: Gene conversion between two DR beta loci results in a new HI,A-D/DR specificity. Nature 322:67, 1986. Gorski J, Tilanus M, Giphart M, Mach B: Oligonucleotide genotyping shows that alleles at the HI,A-DR beta Ill locus of the DRw52 supertypic group segregate independently of known DR or Dw specificities. Immunogenetics 25:79, 1987. Spies T, Sorrentino R, Boss J, Okada K, StromingerJL: Structural organization of the DR subregion of the human major histocompatibility complex. Proc Natl Acad Sci USA 82:5165, 1985. Bell G, Karam J, Rutter W: Polymorphic DNA region adjacent to the 5' end of the human insulin gene. Proc Nad Acad Sci USA 78:5759, 1981. Kunkel L, Smith K, Boyers S, Borgaonkar D, Wachtel S, Miller O, Breg W, Jones H, Jr, Rary J: Analysis of human Y-chromosome specific reiterated DNA in chromosome variants. Proc Natl Acad Sci USA 74:1245, 1977. Southern EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503, 1975. Schre~der G, Kennedy L, Gebuhrer L, AwadJ, Betuel H, Degos L, Jeannet M: HLADRw6 and its subgroups HLA-DRw13 and HLA-DRw14. In Albert et al (eds): Histocompatibility Testing 1984. Heidelberg, Springer-Verlag, 1984, p. 192. Herziot A, Lepage V, Freidel A, Betuel H, Degos L, Charron D: Biochemistry of HLA-DRw6: Evidence of seven distinct haplotypes. J Immunol 136:3767, 1986. Bell J, Denney D, Jr, MacMurray A, Foster L, Wading D, McDevitt H: Molecular mapping of Class II polymorphisms in the human major histocompau'bility complex. I. DR beta I. J lmmunol 139:562-573, 1987.