Ann. Hum. Genet., Lond.(1975), 38, 383 Printed in Great Britain
383
Data on the HL-A linkage group BY L. U. LAMM,* INGE-LIS THORSEN,* G. BRUUN PETERSEN,? J. JORGENSEN,* K. HEWINGSEN,$ BENTE BECHS AND F. KISSMEYER-NIELSEN* The HL-A system is without competition the most polymorphic chromosomal region yet identified in man. It contains, among others, both loci coding for _SerologicallyDefined (SD) antigens and loci determining surface factors responsible for the activation in mixed lymphocyte cultures Defined) (Bach, 1973). It is well established that there are strong gametic (LD = hmphocyte associations between genes within this region which is contained in only a few centimorgans of the genome. Up to the present only serological determinants have been used routinely for determining the segregation of HL-A. These antigens are coded for by two well-known loci, LA and FOUR, separated by approximately 0.8 centimorgans (Svejgaard et al. 1971). The number of alleles at these loci are at least 14 and 16, respectively (Dausset & Colombani, 1972), making a total of 224 different haplotypes. The extreme polymorphism is further illustrated by the rareness of the most frequent haplotype, HL-Al,S,which has a frequency in the order of 10 % in Caucasians. The system is, therefore, exceptionally well suited for linkage studies as nearly every random family will be segregating for HL-A both in the father and the mother. The rareness of good test sera has, however, restricted the use of HL-A as a linkage marker to a small number of laboratories. Therefore, the literature on this subject is surprisingly meagre. The linkage between HL-A and phosphoglucomutase-3 (PGM,) is well established by family studies (Lamm, Svejgaard & Kissmeyer-Nielsen, 1971) as well as by hybridization work (van Someren et al. 1973),and it is likely that PGM, is on the FOUR side of the HL-A region (Lamm et al. 1972). The assignment of this linkage group on chromosome 6 is also documented both by hybridization (Jongsma et al. 1973) and family studies (Lamm et al. 1974). Strong evidence for close linkage between Chido and FOUR has been published (Middleton et al. 1974). It has been claimed that the loci for adenosine deaminase (ADA) and the P antigen also belong to this linkage group, but these data are controversial and will be discussed later. The purpose of this paper is to report our data on the HL-A linkages from a computer analysis of all families which have been HL-A typed at the tissue typing laboratory. These families were typed for up to 19 other marker loci, and include 14 families with cystic fibrosis. The data presented herewith include and update previous published data : Lamm, KissmeyerNielsen & Henningsen (1970); Lamm et al. (1971) and Edwards et al. (1972).
* t
Blood Bank and Tissue Typing Laboratory, The University Hospital, 8000 Arhus C, Denmark. Institute of Human Genetics, University of ifrhus, 8000 Arhus C, Denmark. $ Serological Department, University Institute of Forensic Medicine, Copenhagen, Denmark. 8 Paediatric Department TG, The University Hospital, Copenhagen, Denmark.
25
H C E 38
384
L.U.LAMMAND
OTHERS
METHODS
Linkage analy& was performed with a slightly modified version of the program written
by Professor J. H. Edwards, Birmingham, implemented on a CDC 6400 computer system. It is based on factor union algebra, each test-factor or antigen being represented by one bit (Edwards, 1972). It is therefore completely general and will accommodate any kind of genetic system. This program calculates lod scores for preselected recombination values (8) using the population frequencies of genes and haplotypes t o account for parents or grandparents with equivocal genotypes. The program h t breaks down large kindreds into nuclear families which are analysed separately. A nuclear family consists of children, parents and grandparents, and it will be subjected to analysis only if it fulfils two conditions: (1) both parents are tested for both markers; (2) there is either more than one child or a t least one tested grandparent. The next step is to determine whether one of the parents may be heterozygous a t both loci; if not, the particular nuclear family is skipped as uninformative. These measures were introduced to save computer time and do not involve any selection bias. Lod scores are calculated for recombination values of 0.06, 0.10, 0.20, 0.30 and 0.40 and labelled to show whether the father or the mother is the double heterozygous parent. Families in which both parents are double heterozygous (double intercrosses) are not scored. I n two-generation families the HL-A sero-results were transformed and coded in the conventional form (parents: AB x CD; children: AC, AD, BC and BD) and analysed according to this simple scheme. If, however, the genotypes of the parents could not be determined, or if the families were of three generation, it is appropriate to utilize the phenotypes in terms of all tested antigens. This involves some complications as the program, due to space limitations, only allows for three HL-A antigens. Thus the information on approximately 30 antigens must be trimmed down to the three most informative ones. First, non-segregating antigens are deleted. Then, for each antigen, the profile of presence and absence among the children is compared with all other tests and those antigens with the same or complementary profiles deleted. This will discard antigens determined by the same haplotype as well as allelic genes that are not shared by the mate. While the first type of deletion never involves any loss of information, the second may do SO. This trimming will in the majority of cases suffice, But if not, it should be continued, first, by deleting antigens possessed by both parents (intercrosses), secondly, by random deletions. This last step has only been necessary in a few families. The program produces for each nuclear family one punch card per linkage relations containing lod scores and appropriate identifications. Finally, lod scores from these cards are summed by a very simple program (LODSUM). . HL-A antigens were identified by lymphocytotoxic antisera and for most of the specificities often also checked by complement-fixation on thrombocytes. The reliability of the sera as well as of the technique has been controlled by participation in the last four work-shops in histocompatibility testing. Most of the typings for H p and Gc and a11 typing for Gm and Inv was performed at the Institute of Forensic Medicine, University of Copenhagen. Blood grouping and enzyme typing was performed with conventional technique. References to techniques were given in Lamm et al. (1970). Complement-types (C’3) and Glutamic-Pyruvic-transaminase types (GPT) were
Data on the HL-A linkage group
385
Table 1. Family material Purpose of investigation
Number of nuclear families
Normal family with more than one child Kidney transplantation Linkage study of ADA Linkage study of cystic fibrosis Linkage study of Kell Proband with leukaemia Choriocarcinoma Miscellaneous Total
229
* t
Patients not considered for linkage analysis. 13 three generation families. $ 7 three generation families.
determined as described by Azen, Smithies & Hiller (1969) and Chen & Giblett (1971), respectively. Most of the typings were performed in duplicate and always without knowledge of the family relatives. Intrafamilial inconsistencies were extremely rare except for the MN system. I n part of the material segregation of most markers was checked against current genetic hypothesis while the phenotype frequencies were checked for internal consistencies and for agreement with other Danish or Scandinavian population studies. None of these provisions revealed any serious deviations except for the MN system, which, accordingly, was not considered for linkage analysis. MATERIAL AND RESULTS
The family material consists of 365 families, the great majority of which are of only two generations. Among these, 229 nuclear families informative for at least one linkage relation were found. These families were selected for various reasons as summarized in Table 1. The only groups with a considerable number of three generation families are those ascertained because of segregation for either the Kell or ADA loci. I n some cases HL-A typing was performed because of known segregation at PGM,. The test of the families was unselected with respect to this linkage study. Two intra HL-A recombinant children were not considered for linkage analysis. The lod scores for HL-A and PGM, are given in Tables 2 and 3, respectively. In table 4 lod scores for other relevant relations are summarized. Approximately 5300 central processor seconds were used for calculating these lod scores, DISCUSSION
computer programs must be assumed to contain errors, and this may be especially true for linkage programs which will be confronted with data of varying structure and content. The present program was constantly monitored by checking the result whenever a new type of marker data or family structure was encountered. Of course, corrections of some errors have had to be made, but in these cases the program clearly signalled the error by stopping the analysis. We therefore believe that our results are reliable, especially considering the wide application of the program. Every genetic marker system is handled according to the same rules. 25-2
L. U. LAMMAND OTHERS
386
Table 2. Lod scores on HL-A System
0
ABO
P M
APH*
P M
ADA AK
C.F.7 C'3 Duffy
GC Gm GPTt
*P Inv. Kell Lutheran
P
P M P M P M P M P M
P M P M
P M P
M P M
P M P M P
M PGM,
P
M PGBI,
Rhesus Secretor Tf.
P
0.05
0'10
0'20
0.30
0.40
No. of families
- 20' 179 - 12.620 - 18.591
- 12.092 - 6.852
- 4'93 I - 2.204
- 1.766
-0.355 - 0'079
38 39
- 10.667
-4119 - 0.446
- 0.426
27 I8
- 5'727
- 8.484 - 4'3 I 0 - 4.884 - 5 '499 - 1.556 -
-4.502 - 2.346
- 7'245 - 5.338 - 9'979
-4'354 - 3'039
-
- 5.560
-7276
- 12.638
- 7'024
- 5'967 - 6.322
- 5.005 - 2.954
-4.028 -7.538 - 7.604
-2.163 - 1.349 -6.016 - 2.932
- 4.346 - 8.337 - I 1.208 -4'777 - 16.379 - 10.826 6.032
- 9.728
P M
- 12'942 - 18.117 - 4'345 - 6.737
P M
- 3'173 - 3'330 - 0.894
- 13.952
M
P M
- 2.695
- 0.721 0'000
*
- 3.046 - 3'40 I
- 2.884 - 1.590 - 2.260 - 4402 - 4'43 I - 1'332 -0.611 - 2.775
- 0.565 - 1.499
0.158
- 1.374 - 0.75 I - 1.572 - 1.415 - 0.352 - 1'801 - 1.131 - 1'939
-0.313 -0.173
- 0'049
- 0.747 - 0.542
- 0.265
- 0.676 - 0.362 - 0.599
-0.180 - 0.056
- 2.053
-0.353
-0'015
- 2.406
-0.654
- 0.099
-04318
-0.119
0.015
- 0.970 - 1.026 - 0.486
- 0.206
-0.080 -
-0'012
19 8
-0.125
4 5
-0.015
I2
-
-0.123
0
9
I4 34 44 24 I7
- 0.825
-0.27 I
- 0.068 - 0.059 - 0'039 - 0.056
- 1'733 - 1.741 - 0.582 - 0.065 - 0.452
- 0.607
-0.129
I1
-0.610
-0.131
I0
- 0.228
- 0'054
3
0.07 I
0.04 I
2
-0.318
-0.117
0.108
0.081
13 8
8 I2
21
- 0.582
-0.173
- 2.378
- 0.83 I
- 0.26 I
-4'590 - 6.702 - 2.336
- 1.576
- 0.46 I
- 2'737
- 1.004
- 3'942
0.199 - 1.400 - 0.496
- 0.270
20
- 0.063
28
- 1.386
4'015 -0.198
I -460 0.029
33 I8
-3.112 - 3'255
- 1'25 I - 0.768
- 0'433 - 0.061
30 36
- 0.422 - 1.892
0.045
- 4.258
- 0'743
0.057 -0.172
28 26
- 0'444
-0.194
-0.076
-0.018 0'000
0
- 1.648
- 9'727 - 5.900 7.288
- 5.038 - 7.756
- 10.124
-2.124
0'000
- 0.389
- 1.925 6.379
0'000
0'000
APH, red cell acid phosphatase. C.F., cystic fibrosis. $ GPT, glutamic-pyruvic-transaminase.
t
0.132
- 0.03 I - 0.052 - 0.079 - 0.225 0.196
8 I1
16 35 25
I
Data on the HL-A linkage group
387
Table 3. Lod scores on PGM, System
0
0.05
0'10
0'20
ABO
P M
- 6.770
- 3.601
- 0'942
P M
- 13'996
APH* ADA AK C.F.1. C'3
Duffy CC
Gm GPTS
P R.1 P M
Inv.
Kell
P PGM, Rhesus Secretor
0'201
- 0.023
- 8.142
- 3'140
- 0.499
- 1'012 - 0.05I
-0'133
- 1'972 - I.474
- 2.976
-0.857 - 1.561 - 1.116
- 2'020 - 2.627
- 1.906
- 0.3 10
- 0.066 - 0.642
- 0.448
0.045 0.092 - 0'239 -0.162
23
16 21
0.019 0.058 0.043
9
7 5
- 0.055 - 0.036
4 2
P
0.350
0.298
0'093
0.024
4
- 0.46I
- 0.287
-0.120
- 0.04
-00'010
I
P M P M P M
- 2.884
- 1'775
- 0'776
- 0'303
-0.071 0.059
2
1' M P P nf P hi P 31
Lutheran
0.004 - 0.272
No. of families
A1
M HP
- 3'925
0.40
- 1.149
- 3.608
- 6.671
0.30
P RI P 1cI
- 0.628
0.190
-0.167 - 4.690 - 3'533 - 6.810 - 3'592 -6.214
- 1.882 - 1'390 - 2.380 - 1'330
- 9'304 - 3.348
- 5'532
- 2.262
- 2.005
- 0.836
-0.315
1.070 - 8.509 - 1.721
0.930 - 44353
- 4.584
- 2.852
0.636 - 1.812 0.953 - 1.260 -0'194
0.340 -0'592 0.798 - 0'494 - 0.076
- 7.919 - 6.043 - 12.227
- 6.386
- 10.897
0.005
- 0.444
- 0.72I
-0'315 - 1.183
0.265
- 0.672 -3'103
- 5'044 - 3 '000
- 1.517 - 0'33 I - 1.820
- I '045 - 3 '404 - 9.600
- 5'250 - 3.965 - 6978
0.I 29
- 2.236
- 0.825
-0.150
-0.104 - 0.08I - 0.094 -0*125
-0.183 - 0.073 0.098
- 1'742 - 1'423
- 0.459 - 0.440 - 1'012
R.1
- 7.784
- 3.821
- 2.789 - 0'752
P M
- 5'753 -2.162
- 3'539 - 1.328
- 1.526 -0.581
2
- 0.005
5 5 34 I8
- 0.060
I8
0,093
- 0.081 - 0'255
23
0.I 26
20
0'110
-0.587 - 0.226
5 I
- 0.049
- 7.068 - 11.809
5 2
6
- 0.393
0.228
I8 9
-0.018
-0.019 -0.124 0.008
-0.102
I2
I2
0.341
0.193
23
I4
- 0.087 - 0'530
- 0.523
28 I9
0.287 -0.116
- 0.254 - 1.358
0'495
3
-0.118
0.I 09
P M P
* t
0.147 - 0.677 - 0.489 - 0.679 - 0.447 - 0.694
-0'133 0.052
-
I1
38 I8
APH, red cell acid phosphatase.
C.F., cystic fibrosis. $ GPT, glutamic-pyruvic-transaminase.
Table 4. Lod Scores on Rh-PGM, and ABO-AK Relation
0
Rh-PGM,
P M I' hi
ABO-AK
0.05
-4.278 - 7'973
0.304 1.218
*
0'10
- 1.138 -4'493 0'454 1.077
0'20
0.804 - 1'531
0'434 0'770
P, parental; M, maternal.
0.30
0.865 -0'432 0.275 0.438
No. of 0.40 nuclear families
0.324 -0.068 0.090 0.135
22 20
3 3
L. U. LAMMAND
388
OTHERS
Table 5. Lod scores on HL-A-P Source
N.Y.
G B A (35H
N.Y. G B A (25) Total
e P
M
-
0.05
- 2,48 -
- 11'21 - 2'43 -
- 4'77 -
0'10
1-77
- 1'45
0'20
-
- 0.56
0-72 - 6.70
- 2.73
- 5.66
-
- 2'00
- 1.32 0.72
- 2'34 - 4'94 - 10.60
0.40
1'20
-
- 0'20
-
- 0.44 - 0.39
0.30
-
0.30 1'00 0.30
I '04
-0.13 - 0.30
- 0.04 - 0.23 -
- 0.05 -
0'20
0.19
-
0.81
-
-
1'11
-
* Source: N.Y., New York; G, Galton Laboratory; B, Birmingham; W, Weitkamp (Rochester, U.S.A.); WO, workshop data all from Edwards et al. (1972); A, Aarhus, present study; T, Tariwerdian et al. (1969). $ No. of informative matings. t P, paternal; M, maternal. It is inevitable that there will be some loss of information especially by not considering distant relationship. But this loss is independent of phenotypes and therefore unbiased. I n this context it was encouraging to see the results obtained for PGM,-Rh, the weakest linkage yet discovered in man (see Table 4). Our (male-dependent) lod scores amount to approximately one tenth of those obtained by Robson et al. (1973) in a set of data nearly ten times larger. The data relevant to the linkage between HL-A and PGM, now consist of 51 nuclear families. From these it is estimated - by graphical interpolation - that the paternal recombination , ratio of 3.6 for this linkage is fraction (8,) is 0.11 while the maternal (Of) is 0.40. The 8, to 8 the highest yet reported in man (Renwick, 1969). The rather extensive data on HL-A-P do not indicate linkage but a positive lod score of 0.199 is found at 8, = 0.30. When combined with data from other sources a slightly positive trend for 8, persists (Table 5). In a recent critical report on manlmouse hybrids, Fellous et al. (1973) presented very strong evidence for synteny of HL-A and P . They found only one discordant hybrid clone among 53 investigated, and, furthermore, with a double staining technique they nearly (but not entirely) ruled out the possibility that the antigenic specificities of HL-A and P are in the same cell surface molecule. Our rather extensive data on PGM,-P could indicate a weak linkage with a 8, of 0.30. These data are all consistent with the sequence HL-A:PGM,:P,as the male distance of HL-A-P, estimated from HL-A-PGM, (0.11) and PGM,-P (0.30), of 0.41 would make this linkage almost undetectable from family studies. The relation between HL-A and ADA has for some time presented a dilemma with positive lod scores in several family series (Edwards et al. 1972), in face of a rather confident assignment by cell hybridization work of ADA to chromosome no. 20 and not to no. 6. This was found by two different systems in two laboratories (Creagan et al. 1972; Jongsma et al. 1973). Our data represent a strong argument against the HL-A-ADA linkage, not only by negative lod scores, but especially by the finding in nine phase-known matings of 15 recombinants and 12 nonrecombinants between ADA and HL-A (Table 6 ) . Thus, both available family and hybridization data now agree that ADA is not on the same chromosome as HL-A. It is disturbing that the lod scores for P-ADA from four series (Table 7 ) are very suggestive
Data on the HL-A linkage group
389
Table 6 . Linkage data on HL-A-ADA Source*
ot
0.05
0.10
c wo
P
-3.73
- 1.58
-
A (19)$ M
-1.10
0.01
-
-8.48
-4'50
- 1'37
-
-7.18
-
-2.62
-0.37 0.59
-
-
Total
0.20
1.01
-4.31
-2.35
-0.75
-
-2.13
-
-
-9.31
-
*, t,
0.40
0.30
0.39 -0.07
Nonrecombinant Recombinant
0.27
-
-0.31
-0.05
-
0'01
0.48
1.02
-
0.49 -0.17 1'34
-0.01
-
1'35
$ See footnotes of Table 5.
Table 7. Lod scores on ADA-P Source*
ot
w
P
G A (21)*
w
M
0.05
0'10
0'20
- 0.46
0.3 I
- 0.26 - 3'33
0.26 - 1'99
- 4'05
- 1.42
0'22
0.39
0.16
0.3 I - 2'35
0.34 0.24 0.34
0'22
0.06 0.27 0.06
0.92
I '07
0'39
0.03
0.09
0.03
1-17
1'55
0.58
0'I I
-6.13 0.27
T
- 5'75 - 1.36
Total
0.40
0.61 0.42 -0.81
G A (9)
P+M
0.30
- I 1.16 *, t, $
0.39
- 1.65 - 0'49 - 3'56
0.41 0.27 - 0.29
0.14 0.09 - 0.07
0.66 0.19
See footnotes of Table 5.
of linkage with a peak value of + 1.55 at 6 = 0.30. I n fact, this value is the only positive score, in the data discussed in this paper (except for established linkages), which could not easily be explained as fortuitous. It is therefore still not established whether P belongs to the HL-A linkage group or not. Edwards et al. (personal communication) found in 48 families evidence for linkage between the locus for cystic fibrosis (C.F.) and HL-A. Our small data are positive for linkage of C.F.to PBM, but not for linkage to HL-A, and thus inconclusive. I n studies of metachromasia of fibroblasts from patients when stained with toluidine blue 0 three different patterns have been identified (Danes & Bearn, 1969; Danes & Winge-Flensborg, 1971). This may reflect genetic heterogeneity of the disease which could be proven by linkage studies. The data on linkage between HL-A-PGM, and other marker loci are mostly negative and, therefore, add to the list of non-linkages, which for some marker-loci, are still very meagre. It is remarkable that neither the gammaglobulin markers Gm and Inw nor the complement marker C'3 shows linkage to HL-A-PGM,. SUMMARY
Lod scores from a study in 229 families of the linkage relations of HL-A-PGM, to 19 marker loci and cystic fibrosis are reported. The data exclude that ADA belongs to this linkage group
L. U. LAMM AND OTHERS while they give weak support for the inclusion of P.There is weak evidence for linkage of cystic 390
fibrosis t o PGM,, but none for linkage to HL-A. No new suggestive linkages appeared. We wish to express our gratitude to Mm Elly Andersen and Miss Hanne Pedersen for enthusiastic secrotarial assistance. Lars U. Lamm is indebted t o Professor J. H. Edwards for the linkage program and for many discussions, especially of the trimming procedure, and I also wish to thank Karen Glenn, Birmingham, for help with the implementation of the program. I am further indebted to the staff a t RECAU, University of Arhus, for excellent computer facilities. I was supported by grants Nos. 512-1513 and 512-2509 from Statens Laegevidenskabelige Fomkningsrhd. REFERENCES
b ~ E. A., ,S ~ T H I E S0. , & HILLER, 0. (1969). High-voltage starch-gel electrophoresis in the study of postalbumin proteins and C'3 globulin). Bwchem. Genet. 3, 215. BACH,F. H. (1973). Genetic control of histocompatibility antigens. Am. J . Hum. Genet. 25, 208. CHEN,S.-H. BE GIBLETT,E. R. (1971). Polymorphism of soluble glutamic-pyruvic-transaminme : a new genetic marker in man. Science, N . Y . 173, 148. CREAOAN, R. P., TISCHEBIELD, J. A., NICHOLS,ELIZABETH A. & RUDDLE,F. H. (1973). Autosomal assignment of the gene for the form of adenosine deaminase which is deficient in patients with combined immunodeficiency syndrome. Lancet p. 1449. DANES,B. S. & BEARN, A. G. (1969). Cystic fibrosis of the pancreas. A study in cell culture. J. Exp. Med. 129, 775. DANES, B. S. & WINOEFLENSBORQ, E. (1971). Cystic fibrosis: cell culture studies on a Danish population. Am. J . Hum. Genet. 23, 297. DAUSSET, J. & COLOMBANI, J. (Eds). (1972). Histocompatibility Testing. Copenhagen: Munksgaard. EDWARDS, J. H. (1972). A marker algebra. Clin. Genet. 3, 371. EDWARDS, J. H., ALLEN, F. H., GLENN, KARENP., LAMM,L. U. & ROBSON,ELIZABETH B. (1972). Tho linkage relationships of HL-A. Histocompatibility Teating, p. 745. Copenhagen : Munksgaard. FELLOUS, M., COUILLIN,P., NEAUPORT-SAUTES, C., FREZAL, J., BILLARDSON, c. & DAUSSET,J. (1973). Studies of human allomtigens on man-mouse hybrids: possible syntheny between HL-A and 1' Systems. Eur. J. Immunol. 3, 543. JONGSMA, A., VAN SOMEREN, H., WESTERVELD, A., HAOEMEIJER, ANN& PEARSON, P. (1973). Localizatiori of genes on human chromosomes by studies of human-Chinese hamster somatic cell hybrids. Assignment of PGM, to chromosome C6 and regional mapping of the PGD, PGM, and Pep-C genes on chromosome A1 . Humangenetik 20, 195. LAMM,L. U., FRIEDRICH, URSULA,PETERSEN, G. BRUUN,JQROENSEN, J., NIELSEN,J., THERKELSEN, A. J. & KISSMEYER-NIELSEN, F. (1974). Assignment of the major histocompatibility complex to chromosomo no. 6 in a family with a pericentric inversion. Human H e r d . 24, 273. LAMM, L. U., KISSMEYER-NIELSEN, F. & HENNINGSEN, K. (1970). Linkage and association studies of two phosphoglucomutase loci (PGM, and PGM,) to eighteen other markers. Human H e r d . 20, 305. LAMM,L. U., KISSMEYER-NIELSEN, F., SVEJOAARD, A., PETERSEN, G. B., THORSBY,E., MAYR,W. & HOGMAN, C. (1972). On the orientation of the HL-A region and the PGM, locus in the chromosome. Tieme Antigens 2, 205. LAMM, L. U., SVEJOAARD,A. & KISSMEYER-NIELSEN, F. (1971). PGM,: HL-A is another linkage in man. Nature New Biol. 231, 109. MAYNARD-SMITH, SHEILA,PENROSE, L. S. & SMITH,C. A. B. (1961). Mathematical Tables for Research FVorkers in Human Genetiw. London: J. and A. Churchill Ltd. MIDDLETON, JANICE,CROOKSTON,MARIE C., FALK,JUDITH A., ROBSON, ELIZABETH B., COOK,P. J. L., BATCHELOR, J. R., BODMER, JULIA,FERRARA, G. B., FESTENSTEIN, H., HARRIS,R.,KISSMEYER-NIELSEN, F., LAWLER,SYLVIAD., SACHS, J. A. & WOLF,EVA(1974). Linkage of Chido and HL-A. Tissue Antigens 4, 366. RENWICK, J. H. (1969). Progress in mapping human autosomes. Brit. Med. Bull. 25, 65. ROBSON,E. B., COOK,P. J. L., CORNEY,G., HOPKINSON, D. A., NOADES,J. & CLEGHORN, T. E. (1973). Linkage data on Rh, PGM,, PGD, peptidase C and Fy from family studies. Ann. Hum. Genet., Lond. 36, 393. SOMEREN, H. VAN,WESTERVELD, A., HAGEMEIJER, A., MEES,J. R.,KHAN,P. M. & ZAALBERC, 0. B. (1973). Human antigen and enzyme markers in man/Chinese hamster somatic cell hybrids. Genetics (in the Press). F., LINDBLOM, SVEJOAARD, A., BRATLIE,A., HEDIN,P. J., HOOMAN,C., JERSILD,C., KISSMEYER-NIELSEN, B., Low, B., MESSETER,L., MOLLER,ERNA,SANDBERG, LENA,STAUB-NIELSEN, LILIAN& THORSBY, E. (1971). The recombination fraction of the HL-A system. Tissue Antigens 1, 81.
383
Data on the HL-A linkage group BY L. U. LAMM,* INGE-LIS THORSEN,* G. BRUUN PETERSEN,? J. JORGENSEN,* K. HEWINGSEN,$ BENTE BECHS AND F. KISSMEYER-NIELSEN* The HL-A system is without competition the most polymorphic chromosomal region yet identified in man. It contains, among others, both loci coding for _SerologicallyDefined (SD) antigens and loci determining surface factors responsible for the activation in mixed lymphocyte cultures Defined) (Bach, 1973). It is well established that there are strong gametic (LD = hmphocyte associations between genes within this region which is contained in only a few centimorgans of the genome. Up to the present only serological determinants have been used routinely for determining the segregation of HL-A. These antigens are coded for by two well-known loci, LA and FOUR, separated by approximately 0.8 centimorgans (Svejgaard et al. 1971). The number of alleles at these loci are at least 14 and 16, respectively (Dausset & Colombani, 1972), making a total of 224 different haplotypes. The extreme polymorphism is further illustrated by the rareness of the most frequent haplotype, HL-Al,S,which has a frequency in the order of 10 % in Caucasians. The system is, therefore, exceptionally well suited for linkage studies as nearly every random family will be segregating for HL-A both in the father and the mother. The rareness of good test sera has, however, restricted the use of HL-A as a linkage marker to a small number of laboratories. Therefore, the literature on this subject is surprisingly meagre. The linkage between HL-A and phosphoglucomutase-3 (PGM,) is well established by family studies (Lamm, Svejgaard & Kissmeyer-Nielsen, 1971) as well as by hybridization work (van Someren et al. 1973),and it is likely that PGM, is on the FOUR side of the HL-A region (Lamm et al. 1972). The assignment of this linkage group on chromosome 6 is also documented both by hybridization (Jongsma et al. 1973) and family studies (Lamm et al. 1974). Strong evidence for close linkage between Chido and FOUR has been published (Middleton et al. 1974). It has been claimed that the loci for adenosine deaminase (ADA) and the P antigen also belong to this linkage group, but these data are controversial and will be discussed later. The purpose of this paper is to report our data on the HL-A linkages from a computer analysis of all families which have been HL-A typed at the tissue typing laboratory. These families were typed for up to 19 other marker loci, and include 14 families with cystic fibrosis. The data presented herewith include and update previous published data : Lamm, KissmeyerNielsen & Henningsen (1970); Lamm et al. (1971) and Edwards et al. (1972).
* t
Blood Bank and Tissue Typing Laboratory, The University Hospital, 8000 Arhus C, Denmark. Institute of Human Genetics, University of ifrhus, 8000 Arhus C, Denmark. $ Serological Department, University Institute of Forensic Medicine, Copenhagen, Denmark. 8 Paediatric Department TG, The University Hospital, Copenhagen, Denmark.
25
H C E 38
384
L.U.LAMMAND
OTHERS
METHODS
Linkage analy& was performed with a slightly modified version of the program written
by Professor J. H. Edwards, Birmingham, implemented on a CDC 6400 computer system. It is based on factor union algebra, each test-factor or antigen being represented by one bit (Edwards, 1972). It is therefore completely general and will accommodate any kind of genetic system. This program calculates lod scores for preselected recombination values (8) using the population frequencies of genes and haplotypes t o account for parents or grandparents with equivocal genotypes. The program h t breaks down large kindreds into nuclear families which are analysed separately. A nuclear family consists of children, parents and grandparents, and it will be subjected to analysis only if it fulfils two conditions: (1) both parents are tested for both markers; (2) there is either more than one child or a t least one tested grandparent. The next step is to determine whether one of the parents may be heterozygous a t both loci; if not, the particular nuclear family is skipped as uninformative. These measures were introduced to save computer time and do not involve any selection bias. Lod scores are calculated for recombination values of 0.06, 0.10, 0.20, 0.30 and 0.40 and labelled to show whether the father or the mother is the double heterozygous parent. Families in which both parents are double heterozygous (double intercrosses) are not scored. I n two-generation families the HL-A sero-results were transformed and coded in the conventional form (parents: AB x CD; children: AC, AD, BC and BD) and analysed according to this simple scheme. If, however, the genotypes of the parents could not be determined, or if the families were of three generation, it is appropriate to utilize the phenotypes in terms of all tested antigens. This involves some complications as the program, due to space limitations, only allows for three HL-A antigens. Thus the information on approximately 30 antigens must be trimmed down to the three most informative ones. First, non-segregating antigens are deleted. Then, for each antigen, the profile of presence and absence among the children is compared with all other tests and those antigens with the same or complementary profiles deleted. This will discard antigens determined by the same haplotype as well as allelic genes that are not shared by the mate. While the first type of deletion never involves any loss of information, the second may do SO. This trimming will in the majority of cases suffice, But if not, it should be continued, first, by deleting antigens possessed by both parents (intercrosses), secondly, by random deletions. This last step has only been necessary in a few families. The program produces for each nuclear family one punch card per linkage relations containing lod scores and appropriate identifications. Finally, lod scores from these cards are summed by a very simple program (LODSUM). . HL-A antigens were identified by lymphocytotoxic antisera and for most of the specificities often also checked by complement-fixation on thrombocytes. The reliability of the sera as well as of the technique has been controlled by participation in the last four work-shops in histocompatibility testing. Most of the typings for H p and Gc and a11 typing for Gm and Inv was performed at the Institute of Forensic Medicine, University of Copenhagen. Blood grouping and enzyme typing was performed with conventional technique. References to techniques were given in Lamm et al. (1970). Complement-types (C’3) and Glutamic-Pyruvic-transaminase types (GPT) were
Data on the HL-A linkage group
385
Table 1. Family material Purpose of investigation
Number of nuclear families
Normal family with more than one child Kidney transplantation Linkage study of ADA Linkage study of cystic fibrosis Linkage study of Kell Proband with leukaemia Choriocarcinoma Miscellaneous Total
229
* t
Patients not considered for linkage analysis. 13 three generation families. $ 7 three generation families.
determined as described by Azen, Smithies & Hiller (1969) and Chen & Giblett (1971), respectively. Most of the typings were performed in duplicate and always without knowledge of the family relatives. Intrafamilial inconsistencies were extremely rare except for the MN system. I n part of the material segregation of most markers was checked against current genetic hypothesis while the phenotype frequencies were checked for internal consistencies and for agreement with other Danish or Scandinavian population studies. None of these provisions revealed any serious deviations except for the MN system, which, accordingly, was not considered for linkage analysis. MATERIAL AND RESULTS
The family material consists of 365 families, the great majority of which are of only two generations. Among these, 229 nuclear families informative for at least one linkage relation were found. These families were selected for various reasons as summarized in Table 1. The only groups with a considerable number of three generation families are those ascertained because of segregation for either the Kell or ADA loci. I n some cases HL-A typing was performed because of known segregation at PGM,. The test of the families was unselected with respect to this linkage study. Two intra HL-A recombinant children were not considered for linkage analysis. The lod scores for HL-A and PGM, are given in Tables 2 and 3, respectively. In table 4 lod scores for other relevant relations are summarized. Approximately 5300 central processor seconds were used for calculating these lod scores, DISCUSSION
computer programs must be assumed to contain errors, and this may be especially true for linkage programs which will be confronted with data of varying structure and content. The present program was constantly monitored by checking the result whenever a new type of marker data or family structure was encountered. Of course, corrections of some errors have had to be made, but in these cases the program clearly signalled the error by stopping the analysis. We therefore believe that our results are reliable, especially considering the wide application of the program. Every genetic marker system is handled according to the same rules. 25-2
L. U. LAMMAND OTHERS
386
Table 2. Lod scores on HL-A System
0
ABO
P M
APH*
P M
ADA AK
C.F.7 C'3 Duffy
GC Gm GPTt
*P Inv. Kell Lutheran
P
P M P M P M P M P M
P M P M
P M P
M P M
P M P M P
M PGM,
P
M PGBI,
Rhesus Secretor Tf.
P
0.05
0'10
0'20
0.30
0.40
No. of families
- 20' 179 - 12.620 - 18.591
- 12.092 - 6.852
- 4'93 I - 2.204
- 1.766
-0.355 - 0'079
38 39
- 10.667
-4119 - 0.446
- 0.426
27 I8
- 5'727
- 8.484 - 4'3 I 0 - 4.884 - 5 '499 - 1.556 -
-4.502 - 2.346
- 7'245 - 5.338 - 9'979
-4'354 - 3'039
-
- 5.560
-7276
- 12.638
- 7'024
- 5'967 - 6.322
- 5.005 - 2.954
-4.028 -7.538 - 7.604
-2.163 - 1.349 -6.016 - 2.932
- 4.346 - 8.337 - I 1.208 -4'777 - 16.379 - 10.826 6.032
- 9.728
P M
- 12'942 - 18.117 - 4'345 - 6.737
P M
- 3'173 - 3'330 - 0.894
- 13.952
M
P M
- 2.695
- 0.721 0'000
*
- 3.046 - 3'40 I
- 2.884 - 1.590 - 2.260 - 4402 - 4'43 I - 1'332 -0.611 - 2.775
- 0.565 - 1.499
0.158
- 1.374 - 0.75 I - 1.572 - 1.415 - 0.352 - 1'801 - 1.131 - 1'939
-0.313 -0.173
- 0'049
- 0.747 - 0.542
- 0.265
- 0.676 - 0.362 - 0.599
-0.180 - 0.056
- 2.053
-0.353
-0'015
- 2.406
-0.654
- 0.099
-04318
-0.119
0.015
- 0.970 - 1.026 - 0.486
- 0.206
-0.080 -
-0'012
19 8
-0.125
4 5
-0.015
I2
-
-0.123
0
9
I4 34 44 24 I7
- 0.825
-0.27 I
- 0.068 - 0.059 - 0'039 - 0.056
- 1'733 - 1.741 - 0.582 - 0.065 - 0.452
- 0.607
-0.129
I1
-0.610
-0.131
I0
- 0.228
- 0'054
3
0.07 I
0.04 I
2
-0.318
-0.117
0.108
0.081
13 8
8 I2
21
- 0.582
-0.173
- 2.378
- 0.83 I
- 0.26 I
-4'590 - 6.702 - 2.336
- 1.576
- 0.46 I
- 2'737
- 1.004
- 3'942
0.199 - 1.400 - 0.496
- 0.270
20
- 0.063
28
- 1.386
4'015 -0.198
I -460 0.029
33 I8
-3.112 - 3'255
- 1'25 I - 0.768
- 0'433 - 0.061
30 36
- 0.422 - 1.892
0.045
- 4.258
- 0'743
0.057 -0.172
28 26
- 0'444
-0.194
-0.076
-0.018 0'000
0
- 1.648
- 9'727 - 5.900 7.288
- 5.038 - 7.756
- 10.124
-2.124
0'000
- 0.389
- 1.925 6.379
0'000
0'000
APH, red cell acid phosphatase. C.F., cystic fibrosis. $ GPT, glutamic-pyruvic-transaminase.
t
0.132
- 0.03 I - 0.052 - 0.079 - 0.225 0.196
8 I1
16 35 25
I
Data on the HL-A linkage group
387
Table 3. Lod scores on PGM, System
0
0.05
0'10
0'20
ABO
P M
- 6.770
- 3.601
- 0'942
P M
- 13'996
APH* ADA AK C.F.1. C'3
Duffy CC
Gm GPTS
P R.1 P M
Inv.
Kell
P PGM, Rhesus Secretor
0'201
- 0.023
- 8.142
- 3'140
- 0.499
- 1'012 - 0.05I
-0'133
- 1'972 - I.474
- 2.976
-0.857 - 1.561 - 1.116
- 2'020 - 2.627
- 1.906
- 0.3 10
- 0.066 - 0.642
- 0.448
0.045 0.092 - 0'239 -0.162
23
16 21
0.019 0.058 0.043
9
7 5
- 0.055 - 0.036
4 2
P
0.350
0.298
0'093
0.024
4
- 0.46I
- 0.287
-0.120
- 0.04
-00'010
I
P M P M P M
- 2.884
- 1'775
- 0'776
- 0'303
-0.071 0.059
2
1' M P P nf P hi P 31
Lutheran
0.004 - 0.272
No. of families
A1
M HP
- 3'925
0.40
- 1.149
- 3.608
- 6.671
0.30
P RI P 1cI
- 0.628
0.190
-0.167 - 4.690 - 3'533 - 6.810 - 3'592 -6.214
- 1.882 - 1'390 - 2.380 - 1'330
- 9'304 - 3.348
- 5'532
- 2.262
- 2.005
- 0.836
-0.315
1.070 - 8.509 - 1.721
0.930 - 44353
- 4.584
- 2.852
0.636 - 1.812 0.953 - 1.260 -0'194
0.340 -0'592 0.798 - 0'494 - 0.076
- 7.919 - 6.043 - 12.227
- 6.386
- 10.897
0.005
- 0.444
- 0.72I
-0'315 - 1.183
0.265
- 0.672 -3'103
- 5'044 - 3 '000
- 1.517 - 0'33 I - 1.820
- I '045 - 3 '404 - 9.600
- 5'250 - 3.965 - 6978
0.I 29
- 2.236
- 0.825
-0.150
-0.104 - 0.08I - 0.094 -0*125
-0.183 - 0.073 0.098
- 1'742 - 1'423
- 0.459 - 0.440 - 1'012
R.1
- 7.784
- 3.821
- 2.789 - 0'752
P M
- 5'753 -2.162
- 3'539 - 1.328
- 1.526 -0.581
2
- 0.005
5 5 34 I8
- 0.060
I8
0,093
- 0.081 - 0'255
23
0.I 26
20
0'110
-0.587 - 0.226
5 I
- 0.049
- 7.068 - 11.809
5 2
6
- 0.393
0.228
I8 9
-0.018
-0.019 -0.124 0.008
-0.102
I2
I2
0.341
0.193
23
I4
- 0.087 - 0'530
- 0.523
28 I9
0.287 -0.116
- 0.254 - 1.358
0'495
3
-0.118
0.I 09
P M P
* t
0.147 - 0.677 - 0.489 - 0.679 - 0.447 - 0.694
-0'133 0.052
-
I1
38 I8
APH, red cell acid phosphatase.
C.F., cystic fibrosis. $ GPT, glutamic-pyruvic-transaminase.
Table 4. Lod Scores on Rh-PGM, and ABO-AK Relation
0
Rh-PGM,
P M I' hi
ABO-AK
0.05
-4.278 - 7'973
0.304 1.218
*
0'10
- 1.138 -4'493 0'454 1.077
0'20
0.804 - 1'531
0'434 0'770
P, parental; M, maternal.
0.30
0.865 -0'432 0.275 0.438
No. of 0.40 nuclear families
0.324 -0.068 0.090 0.135
22 20
3 3
L. U. LAMMAND
388
OTHERS
Table 5. Lod scores on HL-A-P Source
N.Y.
G B A (35H
N.Y. G B A (25) Total
e P
M
-
0.05
- 2,48 -
- 11'21 - 2'43 -
- 4'77 -
0'10
1-77
- 1'45
0'20
-
- 0.56
0-72 - 6.70
- 2.73
- 5.66
-
- 2'00
- 1.32 0.72
- 2'34 - 4'94 - 10.60
0.40
1'20
-
- 0'20
-
- 0.44 - 0.39
0.30
-
0.30 1'00 0.30
I '04
-0.13 - 0.30
- 0.04 - 0.23 -
- 0.05 -
0'20
0.19
-
0.81
-
-
1'11
-
* Source: N.Y., New York; G, Galton Laboratory; B, Birmingham; W, Weitkamp (Rochester, U.S.A.); WO, workshop data all from Edwards et al. (1972); A, Aarhus, present study; T, Tariwerdian et al. (1969). $ No. of informative matings. t P, paternal; M, maternal. It is inevitable that there will be some loss of information especially by not considering distant relationship. But this loss is independent of phenotypes and therefore unbiased. I n this context it was encouraging to see the results obtained for PGM,-Rh, the weakest linkage yet discovered in man (see Table 4). Our (male-dependent) lod scores amount to approximately one tenth of those obtained by Robson et al. (1973) in a set of data nearly ten times larger. The data relevant to the linkage between HL-A and PGM, now consist of 51 nuclear families. From these it is estimated - by graphical interpolation - that the paternal recombination , ratio of 3.6 for this linkage is fraction (8,) is 0.11 while the maternal (Of) is 0.40. The 8, to 8 the highest yet reported in man (Renwick, 1969). The rather extensive data on HL-A-P do not indicate linkage but a positive lod score of 0.199 is found at 8, = 0.30. When combined with data from other sources a slightly positive trend for 8, persists (Table 5). In a recent critical report on manlmouse hybrids, Fellous et al. (1973) presented very strong evidence for synteny of HL-A and P . They found only one discordant hybrid clone among 53 investigated, and, furthermore, with a double staining technique they nearly (but not entirely) ruled out the possibility that the antigenic specificities of HL-A and P are in the same cell surface molecule. Our rather extensive data on PGM,-P could indicate a weak linkage with a 8, of 0.30. These data are all consistent with the sequence HL-A:PGM,:P,as the male distance of HL-A-P, estimated from HL-A-PGM, (0.11) and PGM,-P (0.30), of 0.41 would make this linkage almost undetectable from family studies. The relation between HL-A and ADA has for some time presented a dilemma with positive lod scores in several family series (Edwards et al. 1972), in face of a rather confident assignment by cell hybridization work of ADA to chromosome no. 20 and not to no. 6. This was found by two different systems in two laboratories (Creagan et al. 1972; Jongsma et al. 1973). Our data represent a strong argument against the HL-A-ADA linkage, not only by negative lod scores, but especially by the finding in nine phase-known matings of 15 recombinants and 12 nonrecombinants between ADA and HL-A (Table 6 ) . Thus, both available family and hybridization data now agree that ADA is not on the same chromosome as HL-A. It is disturbing that the lod scores for P-ADA from four series (Table 7 ) are very suggestive
Data on the HL-A linkage group
389
Table 6 . Linkage data on HL-A-ADA Source*
ot
0.05
0.10
c wo
P
-3.73
- 1.58
-
A (19)$ M
-1.10
0.01
-
-8.48
-4'50
- 1'37
-
-7.18
-
-2.62
-0.37 0.59
-
-
Total
0.20
1.01
-4.31
-2.35
-0.75
-
-2.13
-
-
-9.31
-
*, t,
0.40
0.30
0.39 -0.07
Nonrecombinant Recombinant
0.27
-
-0.31
-0.05
-
0'01
0.48
1.02
-
0.49 -0.17 1'34
-0.01
-
1'35
$ See footnotes of Table 5.
Table 7. Lod scores on ADA-P Source*
ot
w
P
G A (21)*
w
M
0.05
0'10
0'20
- 0.46
0.3 I
- 0.26 - 3'33
0.26 - 1'99
- 4'05
- 1.42
0'22
0.39
0.16
0.3 I - 2'35
0.34 0.24 0.34
0'22
0.06 0.27 0.06
0.92
I '07
0'39
0.03
0.09
0.03
1-17
1'55
0.58
0'I I
-6.13 0.27
T
- 5'75 - 1.36
Total
0.40
0.61 0.42 -0.81
G A (9)
P+M
0.30
- I 1.16 *, t, $
0.39
- 1.65 - 0'49 - 3'56
0.41 0.27 - 0.29
0.14 0.09 - 0.07
0.66 0.19
See footnotes of Table 5.
of linkage with a peak value of + 1.55 at 6 = 0.30. I n fact, this value is the only positive score, in the data discussed in this paper (except for established linkages), which could not easily be explained as fortuitous. It is therefore still not established whether P belongs to the HL-A linkage group or not. Edwards et al. (personal communication) found in 48 families evidence for linkage between the locus for cystic fibrosis (C.F.) and HL-A. Our small data are positive for linkage of C.F.to PBM, but not for linkage to HL-A, and thus inconclusive. I n studies of metachromasia of fibroblasts from patients when stained with toluidine blue 0 three different patterns have been identified (Danes & Bearn, 1969; Danes & Winge-Flensborg, 1971). This may reflect genetic heterogeneity of the disease which could be proven by linkage studies. The data on linkage between HL-A-PGM, and other marker loci are mostly negative and, therefore, add to the list of non-linkages, which for some marker-loci, are still very meagre. It is remarkable that neither the gammaglobulin markers Gm and Inw nor the complement marker C'3 shows linkage to HL-A-PGM,. SUMMARY
Lod scores from a study in 229 families of the linkage relations of HL-A-PGM, to 19 marker loci and cystic fibrosis are reported. The data exclude that ADA belongs to this linkage group
L. U. LAMM AND OTHERS while they give weak support for the inclusion of P.There is weak evidence for linkage of cystic 390
fibrosis t o PGM,, but none for linkage to HL-A. No new suggestive linkages appeared. We wish to express our gratitude to Mm Elly Andersen and Miss Hanne Pedersen for enthusiastic secrotarial assistance. Lars U. Lamm is indebted t o Professor J. H. Edwards for the linkage program and for many discussions, especially of the trimming procedure, and I also wish to thank Karen Glenn, Birmingham, for help with the implementation of the program. I am further indebted to the staff a t RECAU, University of Arhus, for excellent computer facilities. I was supported by grants Nos. 512-1513 and 512-2509 from Statens Laegevidenskabelige Fomkningsrhd. REFERENCES
b ~ E. A., ,S ~ T H I E S0. , & HILLER, 0. (1969). High-voltage starch-gel electrophoresis in the study of postalbumin proteins and C'3 globulin). Bwchem. Genet. 3, 215. BACH,F. H. (1973). Genetic control of histocompatibility antigens. Am. J . Hum. Genet. 25, 208. CHEN,S.-H. BE GIBLETT,E. R. (1971). Polymorphism of soluble glutamic-pyruvic-transaminme : a new genetic marker in man. Science, N . Y . 173, 148. CREAOAN, R. P., TISCHEBIELD, J. A., NICHOLS,ELIZABETH A. & RUDDLE,F. H. (1973). Autosomal assignment of the gene for the form of adenosine deaminase which is deficient in patients with combined immunodeficiency syndrome. Lancet p. 1449. DANES,B. S. & BEARN, A. G. (1969). Cystic fibrosis of the pancreas. A study in cell culture. J. Exp. Med. 129, 775. DANES, B. S. & WINOEFLENSBORQ, E. (1971). Cystic fibrosis: cell culture studies on a Danish population. Am. J . Hum. Genet. 23, 297. DAUSSET, J. & COLOMBANI, J. (Eds). (1972). Histocompatibility Testing. Copenhagen: Munksgaard. EDWARDS, J. H. (1972). A marker algebra. Clin. Genet. 3, 371. EDWARDS, J. H., ALLEN, F. H., GLENN, KARENP., LAMM,L. U. & ROBSON,ELIZABETH B. (1972). Tho linkage relationships of HL-A. Histocompatibility Teating, p. 745. Copenhagen : Munksgaard. FELLOUS, M., COUILLIN,P., NEAUPORT-SAUTES, C., FREZAL, J., BILLARDSON, c. & DAUSSET,J. (1973). Studies of human allomtigens on man-mouse hybrids: possible syntheny between HL-A and 1' Systems. Eur. J. Immunol. 3, 543. JONGSMA, A., VAN SOMEREN, H., WESTERVELD, A., HAOEMEIJER, ANN& PEARSON, P. (1973). Localizatiori of genes on human chromosomes by studies of human-Chinese hamster somatic cell hybrids. Assignment of PGM, to chromosome C6 and regional mapping of the PGD, PGM, and Pep-C genes on chromosome A1 . Humangenetik 20, 195. LAMM,L. U., FRIEDRICH, URSULA,PETERSEN, G. BRUUN,JQROENSEN, J., NIELSEN,J., THERKELSEN, A. J. & KISSMEYER-NIELSEN, F. (1974). Assignment of the major histocompatibility complex to chromosomo no. 6 in a family with a pericentric inversion. Human H e r d . 24, 273. LAMM, L. U., KISSMEYER-NIELSEN, F. & HENNINGSEN, K. (1970). Linkage and association studies of two phosphoglucomutase loci (PGM, and PGM,) to eighteen other markers. Human H e r d . 20, 305. LAMM,L. U., KISSMEYER-NIELSEN, F., SVEJOAARD, A., PETERSEN, G. B., THORSBY,E., MAYR,W. & HOGMAN, C. (1972). On the orientation of the HL-A region and the PGM, locus in the chromosome. Tieme Antigens 2, 205. LAMM, L. U., SVEJOAARD,A. & KISSMEYER-NIELSEN, F. (1971). PGM,: HL-A is another linkage in man. Nature New Biol. 231, 109. MAYNARD-SMITH, SHEILA,PENROSE, L. S. & SMITH,C. A. B. (1961). Mathematical Tables for Research FVorkers in Human Genetiw. London: J. and A. Churchill Ltd. MIDDLETON, JANICE,CROOKSTON,MARIE C., FALK,JUDITH A., ROBSON, ELIZABETH B., COOK,P. J. L., BATCHELOR, J. R., BODMER, JULIA,FERRARA, G. B., FESTENSTEIN, H., HARRIS,R.,KISSMEYER-NIELSEN, F., LAWLER,SYLVIAD., SACHS, J. A. & WOLF,EVA(1974). Linkage of Chido and HL-A. Tissue Antigens 4, 366. RENWICK, J. H. (1969). Progress in mapping human autosomes. Brit. Med. Bull. 25, 65. ROBSON,E. B., COOK,P. J. L., CORNEY,G., HOPKINSON, D. A., NOADES,J. & CLEGHORN, T. E. (1973). Linkage data on Rh, PGM,, PGD, peptidase C and Fy from family studies. Ann. Hum. Genet., Lond. 36, 393. SOMEREN, H. VAN,WESTERVELD, A., HAGEMEIJER, A., MEES,J. R.,KHAN,P. M. & ZAALBERC, 0. B. (1973). Human antigen and enzyme markers in man/Chinese hamster somatic cell hybrids. Genetics (in the Press). F., LINDBLOM, SVEJOAARD, A., BRATLIE,A., HEDIN,P. J., HOOMAN,C., JERSILD,C., KISSMEYER-NIELSEN, B., Low, B., MESSETER,L., MOLLER,ERNA,SANDBERG, LENA,STAUB-NIELSEN, LILIAN& THORSBY, E. (1971). The recombination fraction of the HL-A system. Tissue Antigens 1, 81.
