A genetic analysis of human minor Histocompatibility antigens demonstrates Mendelian segregation independent from HLA.

\2t\

Immunogenetics 38: 98-105, 1993

Immunp-genetics

© Springer-Verlag 1993

Α genetic analysis of human minor histocompatibility antigens

demonstrates Mendelian segregation independent of HLA

Geziena M. T. Schreuder

, Jos Pool

, Eis Blokland

, Cedle van Eis

, Astrid Bakkeri, Jon J. van Rood3>

Eis Goulmyi

1 Department of Immunohaematology and Blood Bank, Umversity Hospital Leiden, P. O. Box 9600, 2300 RC Leiden, The Netherlands 2 Department of Immunobiology, RIVM, P. O. Box 1, 3720 BABilthoven, The Netherlands

3 Europdonor Foundation, University Hospital Leiden, P. O. Box 9600, 2300 RC Leiden, The Netherlands Received November 16, 1992

Abstract. An analysis of the genetic traits of human

minor histocompatibility (mH) antigens is, unlike with

inbred mice, rather complicated. Moreover, the fact that

mH antigens are recognized in the context of MHC

molecules creates an additional complication for

relia-ble segregation analysis. To gain insight into the mode

of inheritance of the mH antigens, we relied upon a

series of HLA-A2-restricted cytotoxic T-cell (CTL)

clones specific for four mH antigens. To perform

segre-gation analysis independent of HLA-A2, we

trans-fected HLA-A2-negative cells with the HLA-A2 gene:

this results in the cell surface expression of the

HLA-A2 gene product and, if present, mH antigen

recogni-tion. The mode of inheritance of the HLA-A2-restricted

mH antigens HA-1, -2, -4, and -5 was analyzed in

25 families whose members either naturally expressed

HLA-A2 or were experimentally rendered

HLA-A2-positive. Analysis of distribution of the mH antigens in

the parent population among the mating types, together

with their inheritance pattems in the families,

demon-strated that the four mH antigens behaved as Mendelian

traits, whereby each can be considered a product of a

gene with two alleles, one expressing and one not

ex-pressing the detected specificity. We also showed that

the loci encoding the HA-1 and HA-2 antigens are not

closely linked to HLA (lod scores Ζ (0 = 0.05) <-4.0).

Some indication was obtained that the 4- and

HA-5-encoding loci may be losely linked to HLA. While we

are aware of the limited results of this nonetheless

comprehepsive study, we feel the similarity in

immuno-genetic traits between human and mouse mH antigens

is at least striking.

Correspondence to: E. Goulmy.

Introduction

An argument that leucocyte groups do exist was found

in family studies which showed that leucocyte

iso-an-tigens could be inherited (Van Rood 1962). Since the

mode of inheritance of human minor histocompatibility

(mH) antigens is sofar unknown, their existence is

ap-parently questionable. The classical definition of mH

loci is, however, also based on their immunogenicity in

skin transplant experiments in mice (Snell 1948). The

mH antigenic immunogenicity in humans is reflected

by the occurrence of graft-vs-host-disease (GVHD) as

well as the rejection of grafts in HLA genotypically

identical bone marrow donor/recipient combinations

(Thomas et al. 1975; Deeg and Storb 1984).

Until recently, mH antigen identiflcation was

ap-proached solely by in vitro cellular means. In that way,

a limited number of Τ helper (Th) cells and cytotoxic

Τ lymphocyte (CTL) mH antigenic determinants was

identified (see review Goulmy 1988). Characteristic for

mH antigens is that they are presented by MHC class I

and II molecules. By using selective depleted T-cell

subsets, we showed that the mH antigen Th cell

re-sponses are mediated by CD4 + ve class II (HLA-DR

or -DP) restricted Τ cells; the CTL responses have the

CD8 phenotype and recognize the mH antigens in the

context of class I (HLA-A or -B) molecules (Goulmy

1988; Van Elsetal. 1990).

G M T Schreuder et al Segregation of human minorhistocompatibility antigens 99

similar immuno and biochemical procedures, we and

others have isolated human mH antigenic peptides

(Sekimata et al 1992, De Bueger et al 1993) Despite

the latter advances, the exact amino acid sequence and

identity of the protein from which the classical munne

or human mH antigens onginate remam to be

deter-mined

We recently also obtained Information on the

poly-morphism of the human mH antigens We used a senes

of distinct CTL clones specific for five HLA class

I-restncted mH antigens to analyze the phenotype

frequencies of mH antigens HA-1, -2, -4, and -5 in the

healthy population These immunogenetic studies

re-vealed that some appeared with high frequencies

(HA-2 95%), whereas other mH antigens occurred with low

frequencies (HA-5 7%, Van Elsetal 1992) Usingthe

same CTL clones as genetic probes, we analyzed the

mode of inhentance of four HLA-A2- restncted mH

antigens in twenty-five randomly chosen farmhes

Segregation of human mH antigens was proposed

earh-er (Goulmy et al 1982b, 1988, Ziearh-er et al 1983) Hearh-ere

we report that mH antigens can be considered as

mde-pendent dominant Mendehan traits Moreover, lod

score analyses were carned out on the present family

matenal to analyze the hnkage between the mH antigen

loci themselves and between the loci encoding these

mH antigens and HLA

Materials and methods

mH antigen HA-1 -2, 4, and 5 specific CTL clones Five cytotoxic Τ cell hnes (CTLs), designated HA 1 through -5, were isolated from penpheral blood lymphocytes (PBL) obtained from five patients after in vivo sensitization by a bone marrow allograft from their HLA genotypically identical siblings (Goulmy 1988) The CTL hnes were suspended at 1 5 cells/ml in a feeder cell mixture and plated at 0 2 ml/well (ι e, 0 3 cells/well) of 96-well round bottom microtiter plates Clones with antihost cytotoxic activity were selected and large scale expansion (30 to 60-fold) of these clones for extensive panel typing (N = 100) was performed Four distinct clusters of mH antigen-specific clones, restncted by HLA-A2 (HA-1, 2, 4, and 5), were identified in this way (Van Eis et al 1992)

Cell mediated lympholysis (CML) assay Cell-mediated lympholysis was measured in vitro by usmg a Standard chromium release assay (Goulmy 1982 a)

J Dausset In fourteen famihes both parents were HLA A2 positive From mne of these famihes lymphocytes from the HLA A2 nega tive members (N = 14) were rendered A2 positive by electropora tion In addition lymphocytes from all the members of five HLA A2 negative famihes were rendered A2 positive by electroporation Those five famihes were only tested foi HA 1 and HA 2

Test for segregation in famihes If the genetics of mH antigens are comparable to blood groups they should be inherited in a simple way (Race and Sanger 1954) They should be controlled by sets of allelic genes that follow the Mendehan laws We tested the hypothe sis that each of the HLA A2 restncted mH antigens is inherited as an independent dominant Mendehan character

The matenal presented here enabled us to perform the following analyses

1) Using the parents as a random population to compare their respec tive phenotype (pf) and gene (gf) frequencies, with those previously pubhshed (Van Eis et al 1992),

2) To test whether Mendehan traits are randomly distnbuted among the mating types of the parents,

3) To test whether Mendehan traits are inhented by the progeny in a predictive fashion However, as pointed out by Smith (1956), a shght over representation of individuals expressing the dominant trait may influence the X2 unfavorably We followed the method suggested by Srnith (1956), which compares the distnbution of recessive offspnng within and among the famihes tested, with the expected distnbution The X2 was calculated as X2 = (observed - expected)2/vanance

Recessive children can be a result of + x - mating if the parental genes were + - x — or from a + x+ mating if the parental genes were Η—χ Η—

The test can then be divided into two parts 1) matings with at least one recessive child should results in a reasonable number of recessive children in all, 2) The distnbution of matings resulting in recessive children can be checked against those without any reces sive children

The gf of HA-1 HA 4, and HA 5 were pubhshed previously as 0 44322 0 08348, and 0 03564, respecti vely, whereas the pf of HA-2 was estimated as 95% (see Table 1) The mH antigen phenotypes are designated HA-1, HA 2, and so on, and the non expressing pheno type ha 1, ha-2, and so on The frequency of the allele encoding the expressed mH antigen can be called ρ and of the nonexpressed allele q, where p+q = 1 The pf and gf are shown in Table 1

When the number of observations is very close to the expected number, X2 values are low and consequently, ρ values are high, ι e , >0 05, indicating a good fit to the expectations

Lod score analyses for hnkage of each of the mH antigens with HLA was performed in double backcross famihes using the forrnula ofMorton(1956)

Ζ (θ) = log 2S' + log [6S(1 Θ5' + Θ5r(l 0)s], where θ = the recombination fraction (theta),

s = the total number of children,

and r = the number of children having one type

Class I gene transfectwn We used electroporation (Potter et al 1984) to mtroduce cloned HLA genes into the Epstem-Barr Virus (EBV) transformed B-cell hnes We transfected the HLA A2 gene cloned in the pHEBO vector (Sugden et al 1985, Shimizu et al

1986) into HLA-A2 negative cells Fluorescence activated cell sorter (FACS) analyses using the class I and HLA A2 specific monoclonal antibodies were carned out and demonstrated the surface expression of the HLA-A2 gene product on all transfected cells Subsequently, the transfected cell hnes were subjected to mH typing (Goulmy et al

1991)

Famihes Twenty five famihes, previously HLA typed for vanous reasons, were available Four of these were kindly provided by

Table 1. Phenotype (pf) and gene (gf) frequencies of the HLA-A2 restncted mH antigens (Van Eis et al 1992)

Spec Phenotype

100 G M T Schreuder et al Segregation of human minorhistocompatibihty antigens Table 2. Antigen expression in families

Α Expression of HA-1 Fam No Name children

and HA-2 antigens HA-1 Pl P2 Children pos neg HA-2 P l * P 2 * Children pos neg Nach Nijs* Hart Dirw+ Pool+ Otte+ Creg Hoge Kop deRy* vdBr+ Wild* WdVl WdV2 Goek* Back+ Dyeu Viss Koud* Gijs* Rooy Pasc Capt* Aang* Robi* 3 3 4 4 4 5 6 6 8 9 10 3 3 4 4 5 5 5 5 10 13 4 4 5 7 Nys* Nach Dyeu Robi* Kop Gijs* WdVl Hart WdV2 Koud* Creg Hoge deRy* Rooy Pasc Viss 3 3 5 7 8 10 3 4 4 5 6 6 9 13 4 5 ntnt nt nt 3 3 2 4 3 5 6 6 8 9 7 1 3 1 3 5 3 4 5 8 13 0 0 0 0 0 0 2 0 1 0 0 0 0 0 3 2 0 3 1 0 2 1 0 Ζ. 0 4 4 5 7 2 3 3 6 3 1 0 0 0 0 0 0 0 0 nt nt 1 0 2 0 5 7 3 4 4 4 6 6 9 13 nt nt 3 3 1 4 4 5 6 6 8 9 10 3 3 4 4 5 5 5 5 10 5 4 4 5 7 0 0 2 0 0 0 0 2 1 0 0 0 2 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0 0 0

Β Expression of HA-4 and Fam No Name children HA-5 antigens HA-4 Pl P2 Children pos neg HA-5 Pl P2 Children pos neg 3 3 3 6 8 8 3 2 3 4 6 6 7 13 4 5 * Families with at least two //Z«4-j42-positive haplotypes, cells from A2-negative family members were rendered A2-positive by

electro-poration

G M T Schreuder et al Segregation of human minor histocompatibüity antigens Table 3. Segregation of HA 1

101

Α Number of ha-1 children in Mating type + X + + x -Farn s u e 4 10 3 4 5 10

B Number of famihes with at Mating type + X + + x -Fam size 3 4 5 6 8 9 10 3 4 5 10 13 Total X2 (a+b) farmlies with No fams 2 1 1 2 2 1

least one ha-1

No fams 2 3 1 2 1 1 _i 11 2 2 4 1 _1 10

at least one ha-1 child No of ha-1 children obs 3 _2 6 2 4 3 _2 11

child vs all farmhes

exp 2 926 2 649 5 575 1 714 4 266 5 162 5 005 16 147

No of fams with ha-1 children

obs 0 2 0 0 0 0 _1 3 1 2 2 1 J } 6 exp 0 595 1058 0 393 0 847 0 464 0 477 0 486 4 320 1256 1347 2 782 0718 0717 6 820 vanance 0 840 1 592 2 432 0 490 1 564 2 164 2 478 6 696 vanance 0 418 0 685 0 239 0 489 0 249 0 249 0 250 2 579 0.467 0 880 0 849 0 203 0.202 2 601 4d f X2 0 074 (n s) 3 956 (p = 0 047) X2 0 675(n s) 0.258 4 963 (n s )

Results and discussion

mH antigen HA-1. In Table 2A the distribution of HA-1

in the 25 families is shown.

1) The HA-1 phenotype distribution among the

50 parents was compared with the expected distribution

as shown below:

Phen Genotype Ν exp Ν obs HA-1 ++ = p2 = 0 442 = 0 194

+ - = 2pq = 2 x 0 . 4 4 x 0 5 6 = 0 493

0 687 34 3 32 0 15 ha-1 — = q2 =0562 = O3 1 4 157 18 0 34

50 0 50 0 49 = X2

We can conclude that the HA-1 distribution in the

parent population is not significantly different from the

expected values (X

= 0.49, ρ >0.05).

2) The distribution of mating types over the families

was compared with the expected distribution, which

can be calculated as follows:

Mating types HA-1 χ HA-1 HA-Ix ha-1 ha-1 xha-1 = 0 6862 = 0 470 = 2 x 686χ 313 = 0 432 = 03132 = 0 097 1000 exp 1175 10 80 _245 25 00 obs 11 0.048 10 0 059 _4 0 980 25 1087 =

We can conclude that the distribution of the mating

types of the 25 families is according to expectation

(X2= 1.087,p>0.05).

102 G M T Schreuderetal Segregaüon of human rmnorhistocompaübility antigen Table 4. HA-2 Phenotype distribution

Α In parent population

HA-2 ha-2

Β Among mating types Mating + X + + x X -obs 48 _2 50 obs 23 2 _Q 25 C Segregation of HA-2 companng 1) Number of ha-2 children

Mating type + X + + x

-2) Number of famihes with Matmg

type + X + + x

-Total X2 (c 1+2)

in famihes with at least ha-2 children obs exp 47 36 _2J4 50 00 exp 22 56 2 37 006 24 99 X2 0 04 0 01 0 05 (n s ) X2 0 008 0 057 0060 0125(n s)

one ha-2 child (methods analogous to those given in Table 3)

exp

11 3 520 0

at least one ha-2 child vs all families Fam with ha-2 cmld

obs 2 0 exp 2 387 0 745 vanance 1435 vanance 2 138 0 466 3 d f X2 38 98 X2 0 07 _I_12 40 24

children was small, it was according to expectation (X2 = 0.675, Table 3 B). Ten families showed the + χ -mating type. The number of ha-1 children in these families was slightly different from the expected num-ber (X2 = 3.956, ρ = 0.047), but the fact that they were observed in six of ten families was according to expec-tation (X2 = 0.258, Table 3). The total X2 (a+b)

indi-cates that the HA-1 segregation agrees satisfactorily with our hypothesis of a genetic trait that follows Men-delian laws (Table 3). The same methods as described above were used to analyze the segregation of HA-2, HA-4, and HA-5.

mH antigen HA-2. As can be seen in Table 2A, only

two parents had the ha-2 phenotype. However, this was according to our expectations as shown in Table 4A. The observation of only two families with the + χ -mating type was not significantly different from that expected (Table 4B).

The number of ha-2 children detected in the two + χ + families was much larger than expected

(X2 = 38.98), but this may be due to having had so few

families. However, the fact that only two such families were found among all the families was according to expectation (see Table 4 c). The very high frequency of HA-2 greatly restricts the number of analyzable famihes.

mH antigen HA-4. Six of 28 parents were HA-4

posi-tive, and they all belonged to + χ - matings (Table 2B). Both observations were according to expectation (Table 5 A, B). Since the + χ + mating type was not represented, we could only test the number of recessive children in the + χ - matings. Their number, as well as the number of famihes in which they were observed, fulfilled the criteria for Mendelian segregation (Table 5 C).

mH antigen HA-5. The HA-5 specificity is even less

G M T Schreuder et al Segregation of human minorhistocompatibihty antigens Table 5. HA-4 phenotype distnbution

103

Α In parent population

HA-4 ha 4

Β Among mating types Mating + X + + x X -obs 6 22 28 obs 0 6 _8 14

C Segregation of HA-4 companng

exp 4 81 23 19 28 00 exp 0 42 3 92 9 66 14 00

1) Number of ha-4 children in famihes with at least one ha-4 child Mating

type ha-4 children

obs + x + 0 + x - 15

2) Number of famihes with at least one ha-4 child vs ; Mating

type

+ X + + X +

Total X2 (c 1+2)

Farn with ha-4 child obs 0 4 exp 12 327 all famihes exp 5 16 X2 0 29 0 06 0 35 X2 0 42 1 13 0.27 1 82 vanance 5 462 vanance 0 645 2 d f X2 1 308 X2 2.086 3 394 (n s )

observed HA-5-positive parents giving rise to four

+ X- matings all agreed with the expected values.

Thus, HA-5 behaves as a Mendelian trait (Table 6).

mH antigens and HLA. We have shown that each of the

four mH antigens can be considered a product of a locus

with two alleles. Are these loci 1) independent of HLA

and 2) independent of each other?

For each of the mH antigens a number of double

back-cross families were analyzed for linkage with

HLA. In families with the + χ + HA mating types, only

the HA-negative children were counted. The results are

shown in Table 7. HA-1 and HA-2 did not appear to be

closely linked to HLA, since their lod scores are below

-2.0 at recombination fractions 0 = 0.05, and 0.10.

HA-4 showed a positive lod score which was highest at

0 = 0.1. However, this score was not sufficiently high

to ensure linkage between both loci. Weak but

insignif-icant positive linkage between HLA and HA-5 was

found. In one family an HLA-ArB recombinant

hap-pened to be informative for HA-5. If HA-5 is linked to

HLA its location is most likely centromeric to HLA-A.

Linkage studies between the mH antigens

them-selves revealed hardly any informative double

back-cross families. This is due to the very high frequency of

HA-2 and the very low frequencies of HA-4 and HA-5.

Two families were informative for HA-1: HA-5 linkage

and one family for HA-2: HA-5, both resulting in weak

negative lodscores (Table 7). There were no

informa-tive families for the other combinations.

Although these data are really very few and

insignif-icant, they at least do not contradict the above

observa-tion that HA-1 and HA-2 loci are definitely not linked to

HLA, whereas HA-5 may be linked. The linkage

be-tween HA-1 and HA-2 could not be established with

only one informative family.

104 G M T Schreuder et al Segregation of human minorhistocompatibility antigens Table 6. HA 5 phenotype distnbution

Α In parent population

HA 5 ha 5

Β Among matmg types Mating obs 4

m

12

16

01

23

116

160

1) Number of ha 5 children in famihes with at least one ha 5 child Mating

type

ha 5 children

obs exp _vanance

+ X +

+ x

-0

15 11 356

2) Number of famihes with at least one ha 5 child vs all fanuhes

4 861 2 73 Mating type + X + + x -TotalX2(cl+2)

Farn with ha-6 child obs 0 4 exp 3 765 vanance 0213 2 d f X* 0.26 2 99(n s)

Table 7. Lod scores for hnkage between HLA and HA 1, HA-2, HA-4, and HA 5

θ= 0 05 0 10 0 20 0 30

HLA HA 1 (counts)

(2 1,4 0,3 2, 6 4,2 0,1 1, 3 0, 3 0) HLA HA 2(3 0,1 2, 5 3,5 3) HLA HA-4 (4 i,4 0,0 3,4 2,8 0) HLA Α HA 5 (1 4, 0 4, 0 4, 6 3) HLA BDR HA 5(1 4, 0 4, 0 4, 7 2) HA 1 HA 5(4 1,1 1) HA 1 HA > (2 0) HA 2 HA 5(2 1) -4 815 -4 000 1 905 -0 187 1 181 -0 907 0 258 -0 721 -2 357 -2 211 2 275 0 597 1 549 -0 466 0215 -0 444 -0519 -0 772 2 037 1 101 1 490 -0 070 0 134 -0 194 0010 -0 232 1 344 0 207 0 975 0 019 0 064 -0 076

expressing and one not expressing the detected

speci-iicity Due to the high frequency of HA-2 (pf = 95%,

gf = 0 77), only two famihes were detected with

HA-2-negative children The distnbution of HA-2-negative children

in these two famihes was not accordmg to Mendehan

segregation, which might be due to the small number of

observations But all other tests were in favor of the

hypothesis that HA-2 IS also encoded by an

mdepen-dent gene

G M T Sthreuder et al Segregation of human minor histocompatibihty antigens 105

between the other mH-encoding loci and HLA or

be-tween the mH loci themselves

The mode of inhentance of the respective HA

an-tigens has considerable implications for bone marrow

transplantation Since the HA antigens are not tightly

linked to HLA and also probably not to each other, the

selection of HLA-identical family members does not

implicate matching for the mH antigens However,

among family members the chance of being HLA and

HA-identicdl IS higher than among HLA-identical

unre-lated individuals (Martin 1991) It is clear that the

differences in frequency distnbution of the mH

an-tigens is also important The chance of matching two

unrelated HLA-identical individuals for mH antigens

with very high (HA-2) or very low frequencies (HA-5)

will be higher than for those minors with an equal

allehc distnbution, such as HA-1 However, our

obser-vations only concern HLA-A2-restncted mH antigens,

and many more as yet unknown antigens may play a

role in GVHD Although multiple mH antigen

dispan-ties between HLA-matched individuals may exist,

T-cell responses against immunodominant mH

an-tigens will prevail, as was clearly demonstrated earher

in mouse modeis (Wettstein and Bailey 1982) and more

recently supposed by our own studies (Van Eis et al

1992)

Our contnbution concermng the mode of inhentance

of minor histocompatibihty antigens and their

distnbu-tion in the populadistnbu-tion may help to elucidate further

causes of GVHD in the future

Acknowledgments We would hke to thank Professor J Dausset for raaking available to us EBV-transformed cell hnes from four HLA A2-positive famihes We are grateful to Dr L Cavalh-Sforza, Dr J Mountain, Dr C Falk, and Dr J D'Amaro for helpful discussion and valuable advice, and to Ms Ν Warmerdam and Mrs I Cunel for editing the manuscript This work was supported in part by grants from the Dutch Foundation for Medical and Health Research (NWO) and Radiopathology and Radiation Protection (IRS)

References

De Bueger, Μ , Verreck, F , Blokland, Ε , Dnjfhout, J W , Amons, R , Konmg, F , and Goulmy, Ε Isolation of an HLA A2 1 ex-tracted human minor histocompatibihty peptide Eur J Immunol 3 614-618,1993

Deeg, Η J and Storb, R Graft-versus-host disease pathophysio-logical and chnical aspects Annu Rev Med 35 11 - 24, 1984 Goulmy, Ε HLA-Α, -Β restnction of cytoxic Τ cells In S Ferrone

and Β G Solheim (eds) HLA Typing Methodology and Chm cal Aspects, vol 2, pp 105-122, CRC press, New York, 1982 a Goulmy, Ε Minor Histocompatibihty antigens in man and their role

in transplantation In J Morris and Ν L Tilney (eds ) Trans

plantation Rev vol 2 pp 29-53, W Β Saunders, Harcourt Brace Jovanovich Philadelphia 1988

Goulmy Ε Gratama J W Blokland Ε Zwaan F Ε and Van Rood J J Recognition of an - as yet unknown - minor trans plantation antigen by post-transpiant lymphocytes from an Α Μ L patient ExpHematol 10(Suppl 10) 127-129 1982b Goulmy Ε , Pool J Blokland Ε and Geraghty D Transfected human class I gene product adequately assembles minor His tocompatibility antigen Immunogenetws 34 270-272 1991 Martin Ρ J Increased disparity for minor Histocompatibihty an

tigens as a potential cause of increased GvHD nsk in marrow transplantation from unrelated donors compared with related donors Bone Marrow Transplant 8 271-273 1991

Morton, Ν Ε Sequential tests for the detection of hnkage Am J Hum Genet 7 277 - 318 1956

Potter, Η Weir L , and Leder Ρ Enhancer dependent expression of human k immunoglobuhn genes introduced into mouse pre Β lymphocytes by electroporation Proc Natl Acad Sei USA 81 7161-7165,1984

Race R R and Sanger, R Blood Groups in Man 2nd edn Oxford, Blackwell, 1954

Rotzshke, Ο , Falk Κ , Wallny, Η J , Faath, S and Rammensee Η -G Charactenzation of naturally occurring minor Histocom patibihty peptides including Η 4 and Η Υ Science 249 283-287, 1990

Sekimata, Μ , Gnem, Ρ , Egawa, Κ , Rammensee Η -G , and Takigu chi, Μ Isolation of human minor Histocompatibihty peptides Int Immunol 4 301 - 304, 1992

Shimizu, Υ , Koller, Β , Geraghty, D , Orr, Η , Shaw, S , Kavathas, Ρ , and De Mars, R Transfer of cloned human class I major histocompatibihty complex genes into HLA mutant lympho blastoid cells MolCellBwlo 1074-1987,1986

Smith, C Α Β Α test for segregation ratlos in family data Ann Hum Genet20 257-265, 1956

Snell, G D Methods for the study of histocompatibihty genes JGenet49 87-103, 1948

Sugden, Β , Marsh, Κ, and Yates, J Α vector that repheates as a plasmid and can be efficiently selected in B-lymphoblasts trans-formedby Epstein Barr virus CellBiol5 410-413,1985 Thomas, Ε D , Storb, R , Chft, R Α , Fefer, Α , Johnson, F L ,

Neiman, Ρ Ε, Lerner, Κ G, Glucksburg, Η, and Buckner, C D Bone marrow transplantation New Engl J Med 292 832-902, 1975

Van Eis, C , D'Amaro, J , Pool, J , Blokland, Ε , Bakker, Α , Van den Elsen, Ρ J , Van Rood, J J , and Goulmy, Ε Immunogenetics of human minor histocompatibihty antigens their polymorphism and immunodominance Immunogenetics 35 161-165,1992 Van Eis, C Α C Μ , Zantvoort, Ε , Jacobs, Ν , Bakker, Α , van

Rood, J J , and Goulmy, Ε Graft versus-host disease associated Τ helper cell responses specific for minor Histocompatibihty antigens are mainly restneted by HLA-DR Bone Marrow Transplant 5 365-372,1990

Van Rood, J J Leucocyte grouping Α Method and its Application, ρ 2, Thesis, Umversity of Leiden, Leiden, 1962

Wettstein, Ρ J and Bailey, D W Immunodominance in the immune response to 'multiple' histocompatibihty antigens Immuno geneticsl6 47-58,1982