Tissue Antigens (1978), 11, 39-44 Published by Munksgaard, Copenhagen, Denmark No part may be reproduced by any process without written permission from the author(s)

Bw2 1 Partition and Crossreactivity of the Components F. C. Grumet(’), B. W. Colombe(2) and P. Goodfellow(2) Department of Pathology“) and Department of Medicine(2) Stanford University School of Medicine, Stanford, California, U.S.A. Available lymphocytotoxic antisera permitted the clear partition of the Bw21 antigen into two distinct components, Bw21.1 and Bw21.2. Bw21.1 is associated with W4 and is approximately twofold more frequent than the W6-associated Bw21.2. Cells of either Bw21 subtype were capable of absorbing specific anti-Bw21.1, anti-Bw21.2, and anti-Bw21 (21.1 21.2) antibodies. Further, an F(ab ’)* fragment prepared from an anti-Bw21.1 serum blocked cytotoxicity of antiBw21.1, anti-Bw21.2 and anti-Bw21 (21.1 21.2) sera. Based on the crossreactivity (by absorption) and blocking data, a model is proposed relating the Bw21 subtypes and the W4 and W6 antigens.

+

+

Received for publication 6 June, revised, accepted 4 July 1977

The HLA-Bw21 antigen appears t o consist of two components originally designated SL-ET and ET* (Thorsby et al, 1970, Thorsby et al. 1971, Legrand & Dausset 1972). These splits are consistent with cluster analysis performed during the 1975 International Histocompatibility Workshop (Bodmer 1975) and the data and analysis of the First HLA Workshop of the Americas (Fuller 1976) in which the designations Bw21.1 and Bw21.2 were proposed for SL-ET and ET* respectively. In this report, population and family data are shown to confirm this partition of Bw21, and CYNAP and F(ab’)g blocking studies are presented to eiucidate the relation between the splits of Bw2 1. Supported in Part by NCI Contract NO1 CB53935

Materials and Methods Cell donors consisted of normal healthy residents of the San Francisco Bay area. Cytotoxic HLA antisera were obtained from the authors’ laboratory or by exchange with other investigators (sera Haynes & Stitz from Dr. R. Payne, Stanford; Ket from Dr. T. Fuller, Boston; and Heiser from Dr. R. Duquesnoy, Milwaukee). Routine lymphocyte microcytoxicity testing was performed by the fluorochromasia method (Bodmer et al. 1967). Antibody absorptions with platelets (Legrand & Dausset 1974) used 6 X lo9 HLA typespecific platelets to absorb each 0.5 cc of serum. Lymphocyte microabsorption studies were also performed in typing trays as previously described (Bodmer et al. 1967).

40

Serum # Simp Chamb Kend Heiser Haynes Stitz Ket

GRUMET ET AL.

NIH # 2-91 -4-12-17-01

-

A195

-

A1 90

-

2-66-4-05

Table 1 1st North Amer. Workshop #

-05 -01

A187

Specificities Bw21.1 Bw21.1 Bw21 .I, B12 Bw21.2 Bw21.2, A10, Aw34, 912, Bw40 Bw21.2, Bw21.2, B12, B13, Bw40 Bw21.1. Bw21.2

microabsorption studies were also per- ethnic groups were identified as Bw21formed in typing trays as previously de- positive using the sera listed in Table 1. As shown in Fig. 1 , 28 of these individuals scribed (Bodmer e t al. 1967). Specific blocking of cytotoxicity was were designated Bw21.1 and 14 were deperformed with F(ab')2 fragments pre- signated Bw21.2. The remaining five pared from selected HLA antisera (Madsen persons, all Japanese, reacted with only & Rodkey 1976). In blocking experiments two of the three anti-Bw21.1 sera and with the target lymphocytes were incubated at only one, or with neither of the two broad 2OoC with F(ab')2 fragments for 15-30 anti-Bw21 sera. The phenotypes of each of min immediately prior t o their transfer these individuals were (a) Aw24, A(-); B5, (without washing) into standard typing Bw21.1; (b) Aw24, A(-); B5, Bw21.1; (c) trays preloaded with antisera. The addition Aw24, Aw26; B5, Bw21.1; (d) A2, Aw24; of complement and incubations were then Aw21.1, Aw35; and (e)A2,Aw24;Bw21.1, performed as for routine fluorochromasia Bw35. The two reactive sera Simp and microcytotoxicity testing. Chamb fail t o show any association in our local data with B5, Bw35, or Bw24 in Results Caucasians or Mexican-Americans, and in the 1975 Workshop of the Americas Simp 1 ) Population and Family Data: was not associated with B5 in Japanese. Forty-seven individuals in three different

Figure 1 . Reactivity of antiBw21 antisera with Bw21(+) cells from four ethnic groups; Caucasians, Mexican-Americans (MA), American Negroes and Japanese. Shaded blocks or bars are (+) reactions, N D = n o t tested, and f = equivocal reactions.

41

BW21 PARTITION

Because of the reported crossreactivity between B5, Bw35, and Bw21 (Legrand & Dausset 1972, Bodmer 1975), it is possible that these antisera may be recognizing a variant of the B5-Bw35 complex present in some Japanese. In the absence of family studies for these five Japanese it was not possible to determine whether or not the antigens reacting with the Simp and Chamb sera would segregate independently of the B5-Bw35 complex. Based on these data, the five Japanese in this study have been designated as only questionably Bw21 positive. In the Caucasian population eight informative backcross families were found (Table 2). In four, the Bw21.1 specifically segregated as a clearcut B locus antigen. Among the 1 3 offspring of these (+/-) X (-/-) matings, eight were positive for the Bw21.1 antigen. In the remaining four families Bw21.2 segregated as a clearcut

2 ) Absorption and Blocking Studies:

Serum absorptions were carried out with platelets of different specificities as shown in Table 3 and the absorbed sera were tested for residual cytotoxicity. (In this part of the study, only Caucasian cells were used.) For each serum, control absorption with Bw21 (-) cells failed t o re-

Table 2 Genotypes of Caucasian Backmoss Families Parents (Generation I) Generation 11

Family -

ah -

X -

cld -

X

B1, Bw21.1, C-; A3, B8, CAw24, B27; A28, B12 A3, B7;A2, Bw21.1

2.

A3, B7, C W ~A28, ; B12, CA l , B8;A2, Bw21.1

X

3.

A3, B7; A3, B7

X

1.

B locus specificity and that antigen was present in 1 3 of 21 (+/-)X (-1-1 offspring. Four of the Bw21.1 and four of the Bw21.2 unrelated Caucasians were also typed for study in the 1977 International HistocompatibiIity Workshop. Based on typings for the W 4 and W 6 antigens with those sera, it was noted that three Bw21.1 (+) individuals had only a W4 antigen and one had both W4 and W6; and that two Bw21.2 (+) individuals had only the W6 antigen and two had both W 4 and W6.

4.

A1 ,B8; A3, B7

X

5.

A2, B7; A3, B14

X

6. 7.

A l , Bw17; A2, Bw21.2 A2, Bw21.2, C-; A l l , Bw40, Cw3

X

8.

A 2 , Bw21.1, C - ; A l l , Bw37, C-

* e = A3, Bw15

X

X

A l l , Bw21.1; Aw32, Bw35 A 3 , Bw21.2; A l l , Bw22 A l , Bw17; A2, B14 A 2 , B12, C-; Aw23, Bw40, C-

A3, B18, C-; Aw25, B7, C-

Generation I11

blc, bld

-

aid, blc, blc

-

a( or b)ld,

-

a(or bYd a(or b ) I d , a(or b)/c ald, ald, blc

-

blc

e */C

bld, ald alc, alc, alc, blc, blc blc, blc, b Id alc, ale, blc, bld

42

GRUMET E T AL. Table 3 Absorption Crossreactivity of Anti-Bw2l Sera Absorbing Platelets Cytotoxicity* vs. Test Cell Possessing

Serum

a) Bw21.1 b ) Bw21.1 c) Bw21.2d) Bw21.2

+ 1:4

+ 1:4

1) None

Simp (anti-Bw21.1)

Hayn (anti-Bw21.2)

2) A2, Bw21.1, C(-); A3, B7, Cc-) 3) A2, Bw21.1, (3-1; A l , B8, C(-) 4) A2, Aw24; B ~ 2 1 . 2 Bw35 , 5) A3, B ~ 2 1 . 2 C(-); , A3, B14, C(-)

-

-

-

-

-

-

-

-

-

-

-

-

4) A ~ , B W ~ I . ~ , C ( - - ) ; A ~ ~ , BW~~, 5 ) A2; Bw21.1,C(-);

-

A3, B7,C(-)

-

+ 1:2 -

+ 1:2 f

-

C(-) 5 ) A3, Bw21.2, C(-);A3, B14, C(-)

1 ) None Ket 2 ) A2, Bw21.1, C(-); (anti-Bw21.1 3) A2, Bw21.1, C(-); anti-Bw21.2) 4) AZ,Bw21.2, C(-); C(-) 5 ) A3,Bw21:2, C(-);

+

-

-

-

-

+ 1:2

+1:2

+1:2

+l:2

-

-

-

+

-

-

-

-

-

-

-

A3,B7, C ( - ) A3,B7, C(-) A l l , Bw37,

-

A3, B14, C(-)

-

* Expressed as maximal dilution at which serum was reactive with the test cell. If no dilution is stated the test results given represent reactions for undiluted sera. Prozone phenomena were n o t observed f o r any sera. move the original specific activity for Bw21, Bw21.1, or Bw21.2. Bw21 (-) cells were not lysed with either absorbed or unabsorbed sera. Additional absorption studies with the other sera listed in Table 1 produced similar results; e.g., with a single absorption both the specific anti-Bw21.1 antibodies (sera Chamb and Kend) and the specific anti-Bw21.2 antibodies (serum Heis) were removed (usually completely) by either Bw21.1 or Bw21.2 (+) platelets, and either subtype was capable of completely removing or significantly decreasing both anti-Bw21.1 and anti-Bw21.2 antibody activity from the broad anti-Bw21 serum (Stit). Lymphocyte microabsorption studies (data not shown) confirmed the platelet absorption data, with cells of

either Bw21 subtype capable of absorbing either and-Bw21.1 or anti-Bw21.2 antibodies. Studies t o demonstrate cytotoxicity blocking were undertaken with an F(ab ’)2 fragment of serum Simp, and representative data are shown in Fig. 2. It can be seen that this fragment of a specific anti-Bw21.1 antibody effectively blocked cytotoxicity of anti-Bw21.1, anti-Bw21.2 and broad anti-Bw21 (Bw21.1 4-Bw21.2) antisera. In control tests, the Simp F(ab’)Z fragment had no effect on the cytotoxicity of nonBw21 antisera; e%., serum Gillespie, an anti-HLA-Al. Neither did a Gillespie F(ab’)2 fragment block serum Simp cytotoxicity against Bw21.1 (+) cells.

BW21 PARTITION

’“r

I, : u

100-

3

80-

0

IBlocked 0 Not Blocbd

J

w

--

60-

+

-

40-

N

3

m 20-

Figure 2. Blocking of cytotoxicitybythe addition of F(ab‘), fragments of anti-Bw21.1 antibody (serum Simp). Open bars show cytotoxicity (percentage of target cells killed) of test antisera in the absence of F(ab‘), fragments. Solid bars show cytotoxicity of test antisera when cells had been pretreated with Simp F(ab’), fragment.

Dkcussion In addition t o confirming the previous reports partitioning Bw21 into two components, the present study demonstrates the presence of strong CYNAP crossreactivity between the two components. Following the nomenclature agreed upon at the last Workshop of the Americas (1975), and using several of the specific and readily available reference sera used there, the Bw21.1 antigen (equivalent t o SL-ET) and the Bw21.2 antigen (equivalent to ET*) are clearly identified by the sera Simp and Heis. In Caucasians, American-Negroes and Mexican-Americans, each of these components of Bw21 is

43

readily recognized; however, their identification among Japanese is not yet clear. Among a small group of individuals it was also confirmed that Bw21.1 is W4-associated and Bw21.2 is W6-associated (Bodmer 1975, Bright 1976), demonstrating the usefulness of the “public” antigens W 4 and W6 in splitting B locus antigens (van Rood e t al. 1970). The Bw21.1 and the Bw21.2 antisera are entirely non-crossreactive by cytotoxicity, yet each can be completely absorbed by cells bearing the alternate Bw21 subtype. Broad anti-Bw21 (Bw21.1 -I- Bw21.2) antisera are similarly completely absorbed by cells of either specificity, and the F(ab‘)z fragment of an anti-Bw21.1 serum can block the anti-Bw21.1, anti-Bw21.2 and broad anti-Bw21 sera, These data are consistent with a model in which every Bw21 antigen would be a complex structure possessing both a Bw21.1 and a separate, but adjacent, Bw21.2 haptenic determinant. Specific anti-Bw21.1 antibodies would bind only to the Bw21.1 determinant, and specific anti-Bw2 1.2 antibodies would bind only to the adjacent determinant; however, broad anti-Bw21 antibodies could bind t o either determinant because of crossreactivity. As a result of the proximity of the two haptens on each Bw21 molecule, when antibody [or F(ab’)z] fragments specific for one of the Bw21 splits is bound only to its particular hapten, binding t o the other hapten will be blocked for antibody that is specific for the other Bw21 split, as well as for any broad anti-Bw21 antibodies, A particular spatial orientation of the Bw21 complex that is associated with the W4 antigen could then permit bound anti-Bw21.1 to activate complement closely enough to the cell surface to be Iytic; but for that particular cell, bound anti-Bw21.2 would activate

GRUMET ET AL.

44

complement too far from the cell surface for effective lysis. Such cells would be classified as Bw21.1. Associated with W6 there would be an alternate spatial orientation of the Bw21 complex which would reverse the positions of the haptens (and their antibodies) and such cells would be classified as Bw21.2. This proposed model would thus consider W 4 and W6 t o charactize two possible “backbone” structures providing alternate spatial orientation of the B locus antigens that appear to reside on the same molecular complex as these “public” antigens (Bernoco et al. 1972, Bright 1974, Ayres & Cresswell 1976). Acknowledgements The authors wish to thank Drs. Duquesnoy, Fuller and Payne for their generous gifts of antisera, Ms. Dolly Ness for her technical assistance, and Ms. Mary Zimmerman for her assistance in preparation of the manuscript. References Ayres, J . & Cresswell, P. (1976) HLA-B specificities and w4, w6 specificities are on the same polypeptide. Eur. J. Immunol. 6, 794799. Bernoco, D., Cullen, S., Scudeller, G., Trinchieri, G., & Cepellini, R. (1972) HLA molecules at the cell surface. Histocompatibility Testing 1972, pp. 527-537. Munksgaard, Copenhagen. Bodmer, J. (1975) Joint Report (Workshop Antisera). Histocompatibility Testing 1975, pp. 21-99. Munksgaard, Copenhagen. Bodmer, W., Tripp, M., & Bodmer, J. (1967) Application of a fluorochromatic cytotoxcity assay to human leukocyte typing. Histocom-

patibility Testing 1967, pp. 341-350. Munksgaard, Copenhagen. Bright, S. (1974) Comparative studies o n the rates of re-expression of histocompatibility antigens after papain treatment. Tissue Antigens 4, 306-312. Bright, S. (1976) Second locus HL-A antigens and 4a or 4b. Tissue Antigens 7,23-34. Fuller, T. (1976) The First HLA Workshop of the Americas: Joint Summary Report. Tissue Antigens 8,55-60. Legrand, L. & Dausset, J. (1972) Serological evidence of the existence of several antigenic determinants (or factors) on the HL-A gene products. Histocompatibility Testing 1972, pp. 441-453. Munksgaard, Copenhagen. Legrand, L. & Dausset, J. (1974) The complexity of the HL-A gene product I: Study of a serum produced against HL-A5 in an HL-A semi-identical situation. Tissue Antigens 4, 329-345. Madsen, L. H. & Rodkey, L. S. (1976) A method for preparing IgG F(ab’), fragments using small amounts of serum. J. Immunol. Methods 9,355-361. Thorsby, E., Bratlie, A., Mayr, W., Sparck, J., Svejgaard, A., Kjerbye, K. E., & KissmeyerNielsen, F. (1971) The HL-A system: Antigenic heterogeneity and four new antigens. Tissue Antigens 1, 32-38. Thorsby, E., Kissmeyer-Nielsen, F. & Svejgaard, A. (1970) New Alleles of the HL-A system: Serological and genetic studies. Histocompatibility Testing 1970, pp. 137-151. Munksgaard, Copenhagen. van Rood, J. J., van Leeuwen, A. & Zweerus, R. (1970) The 4a and 4b antigens: Do they or don’t they? Histocompatibility Testing 1970, pp. 93-103. Munksgaard, Copenhagen. Address : F. C. Grumet, M.D.

Dept. of Pathology (L-235) Stanford University School of Medicine Stanford California 94305 USA.