Tissue (1972), 2, 262-266 Published by Munksgaard, Copenhagen, Denmark No part may be reproduced by any process without written permission from the author(s)

Is the H-2K Locus of the Mouse Stronger Than the H-2D Locus?

JAN KLEIN Department of Oral Biology, School of Dentistry, Department of Human , Medical School, The , Ann Arbor, Michigan 48104

Rrceived for publication 23 February, ncceptcd 4 Apiil 1972

Recently, evidence has been gathered dence for the “superiority of the K-end in- showing that the H-2 complex of tlic compatibilities over the D-end incompati- mouse consists of only two histocompati- bilities” (RychlikovA et al. 1971). In this bility (regions), H-2K and H-2D communication, evidence is presented that : (Klein & Shreffler 1972). The rest of the 1) skin graft rejection does occur across H-2 regions described in the literature the H-2D barrier, and 2) H-2K, at least (C,V,E,A) are probably serological arti- in the combinations studied, presents a facts. Kychlikovi and co-workers ( 197 1) stronger histocompatibility barrier than reported that significant transformation of H-2D. allogeneic in mixed cultures Four intra-H-2 crossovers (Table I ) , (MLC) occurred when congenic strain two proven (i.e. H-2h-3Sg,H-2i-2Sg) and combinations differed at the H-2I< locus two suspected 1i.e. H-2a,H-2m), and four and did not occur when the combinations H-2 chromosomes from which the cross- differed at the H-2D locus. According to overs were derived ( H-2b,H-2d,H-2k,H- Dkmant ( 1970), a similar phenomenon 24), were arranged in eight different do- can be observed in the graft-versus-host nor-recipient combinations (Table 11). (GVH) reaction. Here again, the 13-2K Four of these combinations differed at the incompatibilities lead to a strong reaction H-2K locus and the other four at the while H-2D incompatibilities cause only H-2D locus. All the H-2 chromosomes insignificant or very mild splenomegaly. were on the same genetic background of The authors interpret these results as evi- the C57RL/lOSn (=BlO) strain, thus ex-

This work was supported by USPHS Rewarch Grants DE-02731 and GRI-18314. STRENGTH5 OF TlIE R4OU\E II-ZK AND H-2D LOCI 263

Table I Origin of four H-2 crussovers

€3-2(1 H-2Kd I H-2D - ____ - I------+ H-2" = H-2KkH-2D" (13lO.Aj" I1-2k I-I-2Kk I H-2Dk

H-2k H-2Kk €I 2Dk _- - I- - I-- -__- + H-2" :H-2KkH-2Dq (ElO.AKM)* H-2'1 H-2KQ I H-2DQ

I 11-2" H-2Kk I H-2Dd = __--- ____ H-2h-Zsg = H-2KkH-2D" (BlO.A[2R]) H-2b €1-2Kb i H-2Db

I---- 13-28 H-2K'' 1 H-2Dd = i __ ~- i H-2'-2% = I-I-2KbH-2D" (BlO.A[5R]) H-2b €F2Kb I H-2Db

* Suspected but not proven H-2 crossovers.

Table II Survival times of skin allografts transplanted in donor-recipient combinations zuhich difler tit the H-2K OY the H-21) locus

Donor (H-2 chromosome) Recipient H-2 difference MST 2S.D." (H-2 chromosome)

1310.A (H-2") B1O.BR (H-2k) H-2Dd 17.0 * 1.3 H10.D2 (H-2d) H-2Kk 11.4 A- 0.8 131O.AKM (H-2m) BlO.I3R (H-2k) H-2D4 14.0 -C 1.5 BIO.G (H-24) H-2Kk 11.7 i 0.7 1llO.A (2R) (H-2h-3%) B10.BR (H-2k) H-2Db 15.0 st 1.3 B10 (H-2b) H-2Kk 11.8 = 0.9 R1O.A. (SR) (H-2'-"sr) I310 (H-2b) H-2D" 15.3 i 1.1 RlO.L)2 (H-Zd) H-2Kb 10.1 + 0.8

# Mean survival time = standard deviation. cluding any non-H-2 incompatibilities. Bailey 1971) was used. The bandages were Both the donors and the recipients were 2- removed seven days after transplantation, to 3-month-old males. The tail skin grafts, and the grafts were inspected daily until about 1 cm2 in size, were obtained by the end of the experiment. In each com- simply cutting off the tail, stripping the bination seven recipients were used. The skin and cutting it into five pieces. The results (Table 11) indicate that in all corn- grafts were placed on the left flank of the binations which differ at the H-2D locus, recipients. Otherwise, the same grafting the skin grafts do not survive longer than technique described elsewhere (Klein & an average of 17 days after grafting. This 264 KLEIN

mean survival time is shorter than MST 1967). It was suggested that the stimula- of grafts exchanged across any of the tion could be due to a third locus which known non-H-2 barriers. Thus, therc is no had no serologically detectable antigens as- question that the H-2D locus alone is a sociated with it. In the mouse, several cir- histocompatibiiity locus. In accordance cumstantial findings were interpreted by with this conclusion, we have recently Amos as evidence for a third locus (Amos shown that the H-2D difference can cause 1971, Amos & Yunis 1971). On the basis a definite stimulation in MLC (Klein et of the third locus hypothesis, the difference al. 1972). between the H-2K and H-2D incompati- However, the results of skin grafting also bilities could be explained by assuming show that in all the combinations tested, that the third locus is stronger than either the grafts transplanted across the H-2D H-2K or H-2D. Alternatively a cumula- barrier survive, on the average, longer than tive effect between the third locus and those transplanted across the H-2K bar- H-2K could result in a stronger incompati- rier. These results can be interpreted in bility than the single H-2D difference. Ac- three different ways. First, the difference cording to the third locus hypothesis, all between the H-2K and H-2D loci could the H-2K differences in Table I1 would be fortuitous. Some 13-2 combinations also have to include third-locus differences. could be more immunogenic than others. This is compatible with the origin of the In the limited number of combinations four H-2 crossovers. The advantage of the tmted, perhaps by chance, the weaker ones third locus hypothesis is that it could ex- coincide with the H-2D incompatibilities. plain not only the differential behavior T'esting additional combinations, as they of the H-21i and H-2D loci in the experi- become available, should either prove or ments of the Prague group and experi- disprove this possibility. Second, the H-2K ments reported in this communication, but locus for some reason may be stronger than also some inconsistencies associated with the H-2D locus. (The H-2D product the H-2 complex. A striking example of could be more deeply embedded in the this is the asymmetrical behavior of the membrane, or it could be slightly different parental variants selected from H-2 hete- chemically, etc.) . This interpretation seem5 rozygous tumors as described by George to be favored by the Prague group. Third, Klein's group in Stockholm. It has been perhaps the H-2K and H-2D loci are of shown by E. Klein ( 1961) that H-2./H-2S cqual strength, and the observed difference heterozygous tumors when transplanted in graft survival is attributable to a third onto an H-2d/H-2s recipient can give rise lOCU5 which is closely linked to H-2 and to a variant which loses the H-2Kk allele located between the H-2K locus and the but retains the H-2Dd allele of the original centromere. The products of this hypo- H-2a ( = H-2KkH-2Dd) chromosome. thetical third locus may not be detectable However, transplantation onto an H-2k/ serologically, but play a significant role in H-2s recipient always leads to a concurrent cellular . Such a locus has been loss of both H-2Kk and H-2Dd alleles, al- postulated both for the mouse and man. though the selection is directed against the In man, such a possibility was first con- H-2Dd allele only. A variant which retains sidered in connection with the finding that the H-2Kk allele but loses the H-2Ddallele leukocytes of some HL-A-identical sib- has never been found. The explanation lings were mutually stimulatory in MLC offered for this asymmetry in variant for- tests (Amos & Bach 1969, Bach & Amos mation is that the variant arises by mitotic STRENGTHS OF THE hlOUSE 11-2K AND 1%-2D LOCI 265

Tumor producing F1 hybrid: --H -Za - XaH-2KkH-2Dd H - 2’ Xs H -2K’H -2DS 1 Tumor .) t Selective F1 hybrids: --H-Zd - XdH-2KdH-2Dd --H-2k - XkH-2KkH-2Dk H - 2’ XSH - 2KSH - 2DS H - 2’ Xs H - 2KSH- 2DS

Selection against Xa and Selection against Xa and ka k da d H-2K (X and H-2K are H-2D (X andH-2D are k lost), no selection against lost 1, H - 2K is also lost d d H-2D (H-2D is retained) due to its central position

- ‘dH-2D H - 2a chromosome of the tumor variant: Figure, 1. An explanation of the asymmetry in tumor variant formation on the basis of the third- locus hypothesis. (X = hypothetical third histocompatibility locus).

crossing-over within the H-2 complex. H-2k/H-2s recipient not only at the H-D2 Such an explanation rests on the assump- locus but also at the third locus (X), ap- tion that the centromere in the IXth link- parently cannot lose the H-2Dd allele with- age group of the mouse is proximal to the out losing also the centrally located H-2K H-2D locus and distal to the H-2K locus. locus. This explanation requires the allcle Recent evidence shows, however, that the at the X-locus in H-2k to be distinct from order in the IXth linkage group is ccntro- the allele in H-2a, an assumption which is mere . . . . H-2K . . . . H-2D . . . , (Lyon et not quite in line with the cross-over hypo- al. 1968, Klein 1970, Miller et al. 1971 and thesis of the origin of H-2a from H-2d and J. Klein, unpublished results). This con- H--2k. However, this discrepancy can be tradicts the order on which the mi- explained by an assumption of mutation at totic crossing-over hypothesis was proposed. the X-locus of H-2”, or additional recom- An explanation of the asymmetry of tumor bination between Xa and H-2Kk during variant formation on the basis of the hypo- the production of strains carrying the H-2” thetical third locus (locus “X”) is shown chromosome. in Figure 1. A third histocompatibility lo- It should be stressed that at this moment cus at the centromeric side of H-2 would it is not possible to decide between the put the H-2K locus in the middle. An alternative explanations offered for the ob- H-2a/H-25 tumor which differs from an served difference in strength of the anti- 266 E;I.EIN gens controlled by the 13-2K and I-I-2D Klein, J. (1970) Order of loci in the 2nd lin- loci. kage group of the mouse with respect to the centromere. Genetics 64, 35. Klein, J. & Bailey, D. W. ( 1971) Histocompati- Acknowledgments bility differences in wild mice. Further evi- I thank Mr. J. M. Park for his outstanding dence for the existence of deme structure in technical assistance and Mr. D. B. Murphy natural populations of the house mouse. Genc- for critical reading of the manuscript. tics 68, 287-297. Klein, J. & Shreffler, D. C. (1972) Evidence supporting a two-gene model for the H-2 histocompatibility system of the mouse. J. exp. Referelices Med. 135, 924-937. Amos, D. B. (1971) An alternative explanation Klein, J., Widmer, M. B., Segall, M. & Bach, of a paradoxical skin grafting result. Trans- F. H. (1972) Mixed culture re- plantation 12, 411-412. active and H-2 histocompatibility loci differ- Amos, D. B. & Bach, F. H. (1968) Phenotypic ences. Immunol., in press. expression of the major histocompatibility Lyon, M. F., Butler, J. M. & Kemp, K. (1968) locus in nian (HL--A): Leukocyte antigens The positions of the centromeres in linkage and mixed leukocyte culture reactivity. J. exp. groups I1 and IX of the mouse. Genet. Res. Med. 128, 623-637. 11, 193-199. Amos, D. B. & Yunis, E. J. (1971) A new Miller, D. A., Kouri, R. E., Dev, V. G., Grewal, interpretation of the major histocompatibility M. S., Hutton, J. J. & Miller, 0. J. (1971) gene complexes of man and mouse. Cell. Im- Assignment of four linkage groups to chro- munol. 2, 517-520. rnosorncs in Mas musculus and a cytogenetic Bach, F. H. & Amos, D. B. (1967) Hu-1: Major method for locating their centrorneric ends. histocompatibility locus in man. Science 156, Proc. nat. Acad. Sci. (Wash.) 68, 2699-2702. 1506-1 508. RychlikovL, M., Dhmant, P. & IvLnyi, P. (1971) DCmant, P. (1970) Genetic requirements for Histocompatibility gene organization and graft-versus-host reaction in the mouse. Dif- mixed lymphocyte reaction. Nature Nerv Biol. ferent efficacy of incompatibility at D- and 230, 271-272. K-ends of the H-2 locus. Folia Biol. (Praha) 16, 273-275. Address : Klein, E. (1961) Studies on the mechanism of Dr. Jan Klcin isoantigcnic variant formation in heterozygous Department of Oral Biology mouse tumors. I. Behavior of H-2 antigens D University of Michigan School of Dentistry and K: Quantitative absorption tests on mouse 101 1 North University Avenue sarcomas. J. nat. Cancer Inst. 27, 1069-1093. Ann Arbor, Michigan 48104