Journal of Autoimmunity (2000) 14, 169–178doi: 10.1006/jaut.1999.0357, available online at http://www.idealibrary.com on
A Susceptibility Locus for Human Systemic LupusErythematosus (hSLE1) on Chromosome 2q
Anna-Karin B. Lindqvist1, Kristjan Steinsson2, Bo Johanneson1, Helga Kristjansdottir2,Alfred Arnasson2,7, Gerdur Grondal2, Inger Jonasson1, Veronica Magnusson1,Gunnar Sturfelt3, Lennart Truedsson4, Elisabet Svenungsson5, Ingrid Lundberg5,Joseph D. Terwilliger6, Ulf B. Gyllensten1 and Marta E. Alarcon-Riquelme1
1Department of Genetics and Pathology,Section for Medical Genetics,Uppsala University, 751 85 Uppsala,Sweden2Department of Rheumatology andCenter for Rheumatology Research,Landspitalinn, University Hospital,101 Reykjavik, Iceland3Department of Rheumatology, UniversityHospital, Lund, 221 85, Sweden4Department of Clinical Microbiology,University Hospital, Lund, 221 85,Sweden5Department of Rheumatology,Karolinska Hospital, 171 76, Stockholm,Sweden6Department of Psychiatry and ColumbiaGenome Center, Columbia University,New York, USA7The Blood Bank, Landspıtalinn,University Hospital, 101, Reykjavik,Iceland
Correspondence to: Marta E. Alarcon-Riquelme, Department ofGenetics and Pathology, Section of Medical Genetics, RudbeckLaboratory, Dag Hammarskjolds vag 20, Uppsala University, S-75185, Uppsala, Sweden. Fax: +46 18 471 48 08. E-mail:[email protected]
Introduction
Systemic lupus erythematosus (SLE) is an auto-immune disease of unknown aetiology, characterizedby the presence of autoantibodies to various cellularcomponents and chronic inflammation of differentorgan systems. It affects primarily women from 20 to55 years of age, although children and the elderly arealso frequently affected. Epidemiological and geneticdata suggest that SLE is the result of a complexinterplay between genetic and environmental factors
1690896–8411/00/020169+10 $35.00/0
[1, 2]. A genetic contribution to the aetiology is indi-cated both by the concordance among monozygotictwins (25–69%) as compared to dizygotic twins (2–3%)[3–6] and the high risk to siblings of SLE patientsrelative to the population prevalence (�s=47–58) [7–9].
The genetics of SLE have been studied using bothanimal models and human families. Genome scansperformed on the New Zealand mouse strains, one ofseveral mouse models for SLE and lupus nephritis,have revealed multiple susceptibility loci involved inthe development of nephritis, and production of highlevels of polyclonal antibodies and autoantibodies[10–13] [reviewed by 14, 15]. Analysis of other murinemodels, such as the MRL-lpr/lpr mouse model, hasrevealed additional putative SLE loci [16, 17]. Thesestudies show that lupus-like disease in the mouseis genetically heterogeneous and results from the
interaction of multiple genes. In humans, little isknown at present about the loci involved in thedevelopment of SLE. Inherited deficiencies of compo-nents of the classical pathway of complement C1q, C2and C4 are known to confer a high risk to develop SLE[18], and various candidate genes have been proposedto be involved in the development of the disease,among them cytokine genes such as IL10, apoptosisgenes, and the genes encoding the Fcg receptors [19].A recent report using the method of affected sibpairanalysis (ASP) described increased allele sharing in aregion at 1q41–42 [20] denoted SLEB1. This regionmay possibly be homologous to the sle1 mouse locus[11] but this has not been formally shown. Candidategenes within the region have been studied and associ-ation was shown with a polymorphism in the PARP(poly(ADP-ribose) polymerase [21]. Recently, threegenome scans for human SLE were published [22–24],supporting the observations in murine lupus ofextended genetic heterogeneity.
In order to identify loci for susceptibility to SLE, weperformed genome scans on extended pedigrees withtwo or more cases of SLE from Iceland, a geographicaland genetic isolate [25], and from Sweden, a relativelyhomogeneous population. We have chosen to use alikelihood-based linkage analysis [26] in order toextract all possible linkage information from the pedi-grees under study. Our results show the feasibility ofstudying the genetics of SLE and the importance ofusing well-defined and homogeneous populations.
Materials and Methods
Family material
Two sets of families were used. The set of Icelandicmaterial consisted initially of eight extended pedi-grees with 147 individuals (Figure 1). Ninety-six indi-viduals were genotyped including patients andinformative healthy relatives. The Icelandic familiesoriginate from various parts of the country. Prelimi-nary analyses revealed genetic heterogeneity at lociwithin the HLA region among the Icelandic families asshown by lod score differences when heterogeneitywas assumed using the program HOMOG (seebelow). As partial C4A deficiency, defined as C4A nullalleles (C4AQ0) in the heterozygous or homozygousstate has been shown to confer risk to develop SLE[18], and the C4A gene is located within the HLA, weexcluded families without C4A deficiency and ana-lysed only those having C4AQ0. Families with C4AQ0were families 1,2,4,5,7 and 8 (Figure 1), and thosewithout C4AQ0 were families 3 and 6. These familiesoriginate from North western Iceland, and particu-larly family 6 originates from a highly isolated area.C4AQ0 was observed by serologic, high-voltage aga-rose electrophoretic analysis as described [27]. Theexclusion of families 3 and 6 reduced the heterogen-eity substantially (�=1.0 for all loci in chromosome 6).Noticing this effect, we decided to consider onlyfamilies with C4AQ0 alleles for the rest of the genome
analysis. Only results from families with C4AQ0 arepresented. The Swedish material consisted of 11 fam-ilies, 10 families from Southern Sweden (Skane), sevenof which have been described previously [28] and onefamily from Middle Sweden. In total, 54 individualswere analysed. Of the Swedish families, five have aconfirmed C4A gene deletion [28] and two havepatients with homozygous C4AQ0 alleles definedserologically. One family has no C4AQ0 [28], and onehad a C2 deficiency [28]. Two families with serologicalanalysis could not be unambiguously defined as nothaving C4AQ0. No heterogeneity was observed in thepreliminary analyses of the Swedish families and forthis reason they were not stratified.
All the multiplex families were selected for havingat least two or more patients with definite SLE andfulfilling four or more of the 1982 ARA classificationcriteria for SLE [29]. When considering only thoseindividuals fulfilling four or more of the ARA classi-fication criteria for SLE and having a definite SLEdiagnosis, there were 16 SLE patients in the Icelandicfamilies and 28 patients in the Swedish families. Twoof the Icelandic patients with SLE were deceased.The female to male ratio of the Icelandic and Swedishpatients were 7:1 and 6:1, respectively. The meanage of onset in the Icelandic cohort was 28.8 years andin the Swedish material 38.8 years. Age of onset wasdefined as the age of the first appearance of an ARAcriteria.
In the primary statistical analysis, this definition ofaffected individuals was used. However, other indi-viduals within these families had manifestations com-patible with SLE and titers of antinuclear antibodies(ANA) ≥ 1:100 with homogeneous fluorescence (cor-responding to 3.5 IU/ml, WHO standard 66/233),although not fulfilling four ARA criteria. Althoughthese individuals were not diagnosed as having defi-nite SLE, they may share some genetic factors withtheir affected relatives. Therefore we included theseindividuals as affected in a second analysis, in whichwe used a broader definition of SLE. Applying thebroad definition of SLE, a further seven individualswere considered affected in the Icelandic families(Figure 1, half-shaded individuals) and a further threeindividuals in the Swedish families. Finally a numberof individuals were observed who had 1–2 clinicalmanifestations compatible with SLE, although unspe-cific for the disease, and low titers of ANA (<1:100) oreven negative ANA. These individuals were consid-ered unknown in the statistical analyses (individualswith a ‘?’ symbol in Figure 1).
The same rheumatologists (K.S. and G.G. in Iceland,G.S. in Lund, and E.S. and I.L. in Stockholm) exam-ined all the patients included. A summary of theclinical data according to the ARA criteria for the SLEgroup is shown in Table 1 and the distribution ofindividuals with different numbers of criteria isshown in Table 2. In the Icelandic and Skanefamilies, all individuals within the pedigrees wereexamined. The study was approved by the localethics committees and all individuals had beeninformed of the purpose of the project before bloodwas drawn.
A new susceptibility locus for human SLE on 2q 171
Figu
re1.
Ped
igre
esof
the
Icel
and
icfa
mili
es.I
ndiv
idua
lsw
ith
mor
eth
anfo
urA
RA
crit
eria
and
dia
gnos
edas
defi
nite
SLE
are
show
nsh
aded
.Ind
ivid
uals
wit
hSL
Em
anif
esta
tion
san
dhi
ghti
ters
ofA
NA
(≥1:
100)
but
not
fulfi
lling
four
crit
eria
are
half
shad
ed.F
amili
es3
and
6w
ere
C4A
Q0-
nega
tive
asd
eter
min
edby
sero
logi
cal
allo
typi
ng.I
ndiv
idua
lsw
ith
1–2
man
ifes
tati
ons
com
pati
ble
wit
hSL
Ean
dne
gati
veor
low
AN
Ati
ters
(<1:
100)
are
show
nw
ith
a‘?
’.G
enot
yped
ind
ivid
uals
(n=
96)
have
ad
oton
the
uppe
rri
ght
corn
er.
172 A. B. Lindqvist et al.
Genotyping
The genome-wide scan was carried out using fluor-escent labeled primers from the CHLC (Weber set 6),employing the panels previously described for maxi-mum multiplexing [30, 31]. The marker set consists of86% tri- and tetranucleotide markers. A total of 336markers with an average heterozygosity of 0.76 wasused, allowing for a 10 cM resolution with some gapsof 15–20 cM. Additional markers were included forchromosomes 1, 2 and 6, including D1S225, D1S213,D1S103, D2S117, D2S118, D2S150, TNFa, TAP, D6S273and D6S276, giving a total of 346 markers. MarkersD1S225, D1S213 and D1S103 are located within theregion previously described at 1q41–42 (20). Inaddition, markers D2S140 and D2S345 (Genethon)were analysed to confirm the results on 2q37. PCRswere performed in a volume of 15 µ containing 20 nggenomic DNA, 1.5 mM MgCl2, 1×PCR buffer (Perkin-Elmer, Inc., Foster City, CA, USA), 200 nM of eachdNTP, 2.5 pmol of each primer and 0.2 units of TaqPolymerase (Perkin-Elmer, Inc.). The PCR amplifi-cation was performed using ABI 877 robots (AppliedBiosystems, Inc., Foster City, CA, USA) and thereactions were pooled automatically.
Alternatively, the amplifications were performed inthe Tetrad thermal cycler (PTC-225, Peltier thermalcycler, MJ research) and the markers pooled manually.
Linkage analysis
Two-point linkage analysis was performed usingMLINK [32, 33] in the FASTLINK (version 4.0) pack-age [34, 35] as implemented in the ANALYSIS pro-gram [26, 36]. The genetic heterogeneity betweenfamilies was tested by the admixture test of theHOMOG program [33], included as a routine in theANALYSIS program. The proportion of linked fam-ilies at each locus is expressed as � [37]. SLE affectswomen more often than men (female:male ratio 9:1)for reasons that are unclear. This difference was incor-porated in the models by reducing the penetrance formales substantially, resulting in practice in an ‘affectedonly’ analysis for males, which circumvents the effectsof incomplete penetrance in the males.
Since the mode of inheritance of the susceptibilityloci for SLE is unknown and not necessarily the samefor each locus, both dominant and recessive modelswere considered in the likelihood analysis. Threeanalysis models were used. First, it was assumed thata rare, dominantly inherited factor was segregatingwith affected individuals (male or female): pD=0.002,f(female)=0.50, fmales=0.0005, where pD is the diseasegene frequency and f the penetrance. Second, it wasassumed that a more common, dominantly inheritedfactor was segregating with affected individuals shar-ing this allele: pD=0.02, f(females)=0.50, f(males)=0.0005.Third, the analysis assumes a recessively inheritedfactor shared by all affected individuals: pD=0.03,f(female)=0.70, f(male)=0.0005. For each marker, allelefrequencies were estimated by allele counting usingthe family material.
Results
Table 1. Frequencies of individual ARA criteria amongindividuals diagnosed as definite SLE*
We performed genome scans in Icelandic and Swedishfamilies with multiple cases of SLE. The two popu-lations were primarily analysed separately and whenthere were indications of overlapping linkage results,the whole family cohort was analysed in combination.For the Icelandic families, when applying the stricterdefinition of SLE (>four ARA criteria) (see Materialsand Methods), four regions were identified with lodscores above 2.0 (Table 3 and Figure 2, bold lines).
After pooling, the markers were mixed with theTAMRA GS-350 or 500 size standard, denatured for2 min at 95°C, and loaded on ABI 377 instruments.The fragments were separated in 4% PA gels, on 36 cmwell-to-read plates. The data was analysed using theGeneScan software (Applied Biosystems, Inc) and theallele calling was performed using the Genotypersoftware (Applied Biosystems, Inc). The non-Mendelian inheritance of microsatellites was analysedusing the GAS package version 2.0 (Alan Young,Oxford University, 1993–1995) on a SunSparc UNIXsystem.
A new susceptibility locus for human SLE on 2q 173
These were 2q37 (D2S125, Z=2.06, dominant model,pD=0.002), 4p15–13 (D4S1627, Z=3.20, recessivemodel, pD=0.03), 19p13 (D19S247, Z=2.58, dominantmodel, pD=0.002) and 19q13 (D19S246, Z=2.06, re-cessive model, pD=0.03). Of interest, the region at4p15–13 is homologous with the mouse locus lmb2[17] and the region at 19q13 is syntenic with themouse locus sle3 [11]. No linkage was observed to theHLA region. Only the marker D6S1280, locatedapproximately 10 cM from the HLA region, providedresults indicative of linkage (Z=1.24, dominant model,pD=0.002, Table 2). No evidence for linkage wasobserved for any region on chromosome 1, whichhas previously been shown to be linked to human SLE[20, 21].
When applying the broader definition of the disease(see Materials and Methods), in which the individualsexpressing SLE manifestation but not fulfilling fourARA criteria were included as well (Figure 1, halfshaded symbols), only one region showed a lodscore over 2.0: 9p22 (gata62f03, Z=2.27, dominantmodel, pD=0.02). Another region of interest wasobserved at 15q26 (D15S657, Z=1.06, recessivemodel, pD=0.03) (Table 4). We note that these regionsare homologous to the mouse loci sle2 and sle3,respectively [11–13].
Table 3. Summary of linked chromosomal regions in thegenome scan of Icelandic and Swedish families, applying thestrict definition of SLE (>four ARA criteria)
† Chromosomal locations are available from CHLC(http://lpg.nci.nih.gov/CHLC).* Results are expressed as Maximum lod scores (Z). All MLSabove 1.0 are presented. The result for 2q37 is in bold.§ neg means MLS Z=0.0.
Table 4. Linked regions in the analysis of the broad definition ofSLE
* Individuals considered affected in the linkage analysis includepatients with definite SLE as well as individuals with less thanfour ARA criteria and high titers of ANA (see Materials andMethods).
† Marker order available from the genetic location database(LDB) (cedar.genetics.soton.ac.uk) and Marshfield genetic map(www.marshmed.org/genetics).* Two-point lod scores using the strict disease definition anddominant inheritance model (pD=0.002).
Genome scan of the Swedish families
In the genome scan of the Swedish families whenapplying the strict definition of SLE (>four ARAcriteria) we observed one region with a lod score >2.0
(D2S125, Z=2.18, dominant model, pD=0.002), whichwas the 2q37 region, also identified in the Icelandicfamilies. When combining the Icelandic and Swedishfamilies for D2S125, a maximum lod score of Z=4.24was obtained (Table 3, Figure 2). Two additionalmarkers were analysed in the region in both Icelandicand Swedish families: D2S345, approximately 10 cMcentromeric of D2S125, and D2S140, 3 cM telomericof D2S125 (according to the maps of LDB andMarshfield) (Table 5), both supporting the finding inchromosome 2q37. Based on the lod score for D2S125and D2S140 in the combined family material, abovethe significance level of 3.30 [38], and the indication oflinkage of the D2S345 markers (Table 5), we considerthis to be a new susceptibility locus for SLE.
The HLA region showed slight evidence of linkagewith three markers (Z=1.87, TNF�, dominant model,pD=0.02; Z=1.54, D6S276, dominant model, pD=0.02;Z=1.24, D6S273, dominant model, pD=0.02) in theSwedish families (Table 3 and Figure 2). No evidenceof linkage was observed at 1q41–42, the region pre-viously described by Tsao et al. [20], while a lod scoreof Z=1.61 (D1S1660, recessive model, pD=0.03) wasobtained in the Swedish families for 1q31, at a markerthat lies approximately 32 cM from the 1q41–42
0
2Chromosome 1
103210321
a
b
c
0
2Chromosome 2
103210321
a
b
c
0
2Chromosome 3
103210321
a
b
c
Chromosome 4
0
4Chromosome 5
2
0
2
0
2
a
b
c
0
2Chromosome 6
103210321
0
4
2
0
2
0
2
a
b
c
a
b
c
Chromosome 7
0
210
a
b
c
210
3
21
3
Chromosome 8
0
103
a
b
c
103
2
12
Chromosome 9
0
103
a
b
c
103
2
12
Chromosome 10
0
2
03
a
b
c
103
2
12
1
Chromosome 11
0
10a
b
c
210
3
21
3
Chromosome 12
0
10a
b
c
210
3
21
3
Chromosome 13
0
10a
b
c
210
3
21
3
Chromosome 142
0a
b
c
210
3
21
3
0
1
Chromosome 15
0
2
0a
b
c
210
3
21
3
1
Chromosome 16
0a
b
c
210
3
21
3
1
0
Chromosome 17
0a
b
c
210
3
21
3
1
0
Chromosome 18
0a
b
c
210
3
21
3
1
0
Chromosome 19
0
4
0
2
0
2
2a
b
c
Chromosome 20
0
1
0
2
0a
b
c
3
1
23
1
Chromosome 21
0
1
0
2
0a
b
c
3
1
23
1
2Chromosome 22
0
1
0
2
0a
b
c
3
1
23
1
2Chromosome X
0
1
0
2
0a
b
c
3
1
23
1
D1S
468
gata
29a0
5D
1S55
2
D1S
1622
D1S
2130
D1S
1588
D1S
1631
D1S
1677
D1S
1679
D1S
1595
D1S
534
D1S
1675
D1S
1669
D1S
1665
D1S
1728
D1S
551
D1S
1678
D1S
1660
D1S
518
D1S
549
D1S
2141
D1S
1663
D1S
225
D1S
547
D1S
103
D1S
213
ata2
9c07
D1S
1656
D2S
1400
D2S
1352
D2S
1360
D2S
405
D2S
1394
D2S
442
D2S
1328
D2S
118
D2S
117
D2S
1776
D2S
1353
D2S
150
D2S
441
D2S
1777
D2S
1790
D2S
1384
D2S
1649
D2S
125
D2S
338
D2S
427
D2S
1363
D2S
434
D2S
1326
D2S
1399
D2S
1391
D2S
436
D3S
1050
D3S
2403
D3S
3038
D3S
2432
D3S
1768
D3S
2409
D3S
1766
D3S
2406
D3S
2465
D3S
2459
D3S
3045
D3S
2460
ata2
8h11
D3S
1763
D3S
3053
D3S
2427
D3S
2398
D3S
2418
D3S
1311
D5s
1505
D5S
820
D5S
816
ata2
3a10
D5S
1456
tap
D6S
1006
D6S
1021
D6S
474
gata
29c0
9D
6S12
80D
6S10
17
D5S
408
tnf
D6S
1040
D6S
1009
D6S
1027
D6S
503
D6S
1277
D6S
305
D6S
1003
D6S
1281
D6S
273
D6S
1053
D6S
276
D5S
1471
gata
68a0
3D
5S14
53
D5S
1719
gata
67d0
3
D5S
1473
D5S
817
gata
84e1
1D
5S14
92
D4S
1652
D4S
2417
D4S
2431
D4S
2368
D4S
1629
D4S
1625
D4S
1644
D4S
2394
D4S
2623
D4S
2367
gata
28f0
3D
4S16
27
D4S
2408
D4S
2397
D4S
2639
D4S
403
D4S
2366
D6S
1056
D7S
821
D7S
1804
D7S
1799
gata
44f0
9
D7S
2195
D7S
1805
D7S
1824
D7S
2212
D7S
1818
gata
31a1
0D
7S81
7
D7S
1802
D7S
513
D7S
1808
D7S
2201
D7S
1826
D7S
550
D8S
1136
gaat
1a4
D8S
1113
D8S
1119
D8S
592
D8S
1179
D8S
1132
gata
12b0
6
D8S
1110
D8S
1477
D8S
136
D8S
1100
D8S
1130
D8S
1099
D8S
373
D8S
1128
D8S
1106
D8S
1145
D10
S67
7
D9S
930
D9S
910
D10
S14
35
D9S
934
D10
S12
39
D9S
302
D9S
938
D9S
257
D9S
1122
D9S
118
D10
S12
30
D9S
921
gata
62f0
3
D10
S21
2
D10
S12
37
D9S
741
D9S
925
ggat
1a4
D10
S14
23D
10S
1225
D10
S12
20D
10S
1426
D10
S14
23D
10S
674
gaat
5f06
D10
S14
12D
10S
189
D11
S23
71
D11
S19
98
D11
S20
02D
11S
2000
D11
S23
59
D11
S19
86
D11
S13
92
ata3
4e08
D11
S19
81D
11S
1999
D11
S19
84
D12
S37
2
D12
S37
5
ata2
5f09
D12
S12
94
D12
S10
64
gata
35f0
5D
12S
1045
D12
S39
5
D12
S39
8
D12
S10
90
D13
S78
7
D12
S37
3
D12
S37
4
ggaa
29h
03
D12
S39
2
D12
S39
1
D12
S10
42
D14
S74
2
D13
S17
3
D14
S30
6
gata
31b0
9D
14S
297
D13
S79
6
D13
S77
9
D13
S80
0D
13S
788
D13
S79
3
D13
S31
7
D14
S61
1
D14
S61
7D
14S
606
D14
S53
D14
S58
8
D14
S59
2
D15
S65
3
D15
S65
7
D15
S65
2D
15S
816
D17
S13
03
D16
S40
3
D15
S64
2
D15
S15
3
actc
D15
S65
9
D15
S16
5D
15S
817
D16
S74
8
D16
S26
22
D17
S13
08
D17
S96
9
D17
S97
4D
17S
1298
D17
S12
94D
17S
1293
D17
S12
2
D16
S53
9
D17
S80
9
D16
S26
24ga
ta22
f09
D17
S12
90
D17
S12
99
D17
S92
1
D16
S51
8
D18
S54
2D
18S
843
gata
26c0
3
D18
S87
7D
18S
535
D18
S97
6D
18S
59
D17
S13
01ga
ta31
b11
D17
S92
8
D17
S78
4
D18
S85
1D
18S
858
D18
S64
D18
S84
4D
18S
541
D20
S95
D20
S47
7
D20
S60
4D
20S
470
D20
S48
1D
20S
109
D20
S47
8
D19
S25
4D
19S
589
D19
S71
4
gata
21g0
5D
19S
247
gata
46c0
1D
20S
173
D21
S14
35
D21
S14
40
D21
S14
46
D22
S44
6D
22S
689
D22
S42
0
D21
S12
70
D21
S14
37
D22
S68
3D
22S
445
D22
S68
5
D21
S14
32
DX
S10
47D
XS
6804
DX
S71
32ga
ta31
e8
DX
S67
89
DX
S68
00
DX
S67
99
DX
S10
01
DX
S68
10
DX
S68
07
D12
S10
52
D13
S28
5
D14
S74
9
DX
S67
97D
XS
1193
D20
S47
3
D19
S60
1
D19
S24
6
D19
S43
3
D19
S58
6
D16
S77
1
Figure 2. Graph showing the two-point linkage results of the genome scan of the strict definition of SLE in Icelandic (boldlines) and Swedish (thin lines) families for a) recessive model pD=0.03, female penetrance 0.7, male penetrance 0.0005, b)dominant model pD=0.02, female penetrance 0.5, male penetrance 0.0005, and c) dominant model pD=0.002, femalepenetrance 0.5, male penetrance 0.0005. Points are joined only for the sake of clarity. In addition, maximum lod scores of 0.0are equal to the baseline.
A new susceptibility locus for human SLE on 2q 175
region. In addition, the marker D2S436 at 2q11showed a lod score of Z=2.13 (recessive model,pD=0.03) in the Swedish families (Table 3).
The analysis of the broader definition of SLE (seeMaterials and Methods) for the Swedish familiesrevealed a suggestion of linkage for the 15q26 region(D15S652, Z=1.68, dominant model, pD=0.002 andD15S657, Z=0.89, recessive model), a region alsoobserved in the Icelandic families. Analysis of thecombined material yielded a lod score for both sets ofZ=1.95 for marker D15S657 (Table 4). This region ishomologous to the genomic fragment includingmouse sle3 and linkage to this region has beenobserved in studies of IDDM (IDDM3) and coeliacdisease, both diseases with autoimmune pathogenesis[39, 40].
obtained, indicated by the � value presented byHOMOG.
Although the linkage for the HLA region was notincreased, we knew that the C4A deficiency was seg-regating in the families, inferring a risk factor for SLEdevelopment. Therefore we applied the stratificationto the rest of the genome in order to find other suscep-tibility factors in these families. We noted that the lodscores for a number of regions were substantiallyincreased. Particularly noteworthy was the increase inthe lod score for D2S125 at 2q37 (before stratification:Z=0.16 assuming homogeneity; Z=0.49 assuming het-erogeneity, �=0.38; after stratification: Z=2.06, �=1.0).This observation may suggest a possible epistatic in-teraction between the presence of C4AQ0 alleles andthe locus on 2q. We can not rule out, however, that analternative explanation is that we have only made ourIcelandic family set more homogeneous, for two rea-sons: the same markers showed linkage at 2q37 in allSwedish families regardless of the presence of C4AQ0(D2S125, Z=2.18, �=1.0; D2S140, Z=1.23, �=1.0), andthe Icelandic family 6 in particular is originated from ahighly isolated area in Northern Iceland with a differ-ent distribution of alleles of the MHC (Arnasson, A.,personal communication) showing that the geneticbackground of this family may be quite different fromthat of the rest of the Icelandic population. As thedetermination of C4AQ0 in the Icelandic set and infour of the Swedish families was only performed sero-logically, we cannot at present say if these familiescarry a C4A gene deletion, but the analysis is underway. In future studies it will be important to determinethe presence of C4AQ0 in families by both serologyand gene analysis and to define whether linkage to2q37 is indeed correlated with the presence ofC4AQ0 in SLE patients from various populations. Weare presently analysing this possibility in more sets offamilies. Nevertheless, the high genetic complexityof diseases such as SLE underscores the need forreducing the genetic heterogeneity potential by strati-fying the family material based on biologically relevantphenotypes in order to detect susceptibility loci, par-ticularly when populations of mixed origin are used.
In the present study, no evidence was found forlinkage to the region 1q41–42 in either family set. Weincluded a number of markers used in the previousstudies [20], but no marker showed any sign oflinkage. Thus, we are unable to confirm linkage to thischromosome 1 region. A possible reason for thisdiscrepancy could be ethnic differences between thefamilies used in the different studies.
Three genome scans have recently been publishedfor human SLE. Interestingly, we describe here regionssuggesting linkage also detected by other groups[22–24, 50]. It will be important to confirm all theseregions in other cohorts of SLE families. An associ-ation of SLE to the IL10 gene located at 1q31–q32 haspreviously been suggested [51]. We observed a lodscore indicative of linkage to 1q31 in the Swedishfamilies. However, we have been unable to showlinkage or association of the IL10 gene to SLE inpatients and families from three different populationsusing an intragene dinucleotide repeat marker [52].
Discussion
In order to identify chromosomal regions containingsusceptibility loci for SLE, two independent genomescans were performed in Icelandic and Swedish fam-ilies. We have chosen to study these well-definedpopulations in order to deal with the challenge ofgenetic heterogeneity suggested by previous studiesof lupus models [reviewed by 14, 15] and by recentgenome scans of human SLE [22–24]. Presumably, thegenetic heterogeneity for a complex disease in a popu-lation like that of Iceland is lower than in the admixedpopulation of the USA for example, increasing thechances of locating genetic factors involved in SLE.Linkage was found to a region on chromosome 2q37(D2S125, Z=4.24, �=1.0; D2S140, Z=3.53, �=1.0)obtained when both family sets were combined(dominant model, pD=0.002) (Table 5). This locus isnot syntenic to any hitherto described mouse SLElocus, and represents a new locus for SLE.
The increased risk for SLE in the presence of com-plete inherited C4 deficiency (homozygous lack of thesubcomponents C4A and C4B) is well established [18],and almost invariably such homozygous individ-uals develop an SLE-like disease. However, suchdeficiency is extremely rare [41]. Partial C4 deficiencyexpressed as C4AQ0 has been associated with SLEin several populations [40–48] and a large deletioninvolving most of the C4A gene and part of the21OHA gene within the HLA-B8, DR3 haplotype hasbeen reported in nearly two thirds of Caucasianswith SLE. During the initial analyses using the eightIcelandic families we observed evidence of heterogen-eity within the HLA region, demonstrated by theprogram HOMOG. We suspected that the hetero-geneity could be due to the lack of C4AQ0 in two ofthe families. Based on the known association betweenC4AQ0 and SLE [40–49], we hypothesized thatexcluding the two families without C4AQ0 from theanalyses would increase the linkage for the HLAregion where the C4 genes are located. The lod scoresat the HLA region were not increased, except forD6S1280, which lies approximately 10 cM from theHLA region, but homogeneity for the region was
176 A. B. Lindqvist et al.
A number of genes located in the 2q37 region couldbe considered as candidate genes for SLE. Among thegenes of interest is the INPP5D gene [53, 54], whichencodes for a SH2-containing phosphatase which hasbeen shown to be recruited through tyrosine phos-phorylation of the Fc�RIIB, CD19 or CD22 [55, 56]after cross-ligation of the B cell receptor. It appearsto be important in Fc�RIIB-mediated inhibitionof B cell activation. Studies have been initiatedto analyze the involvement of this gene in thedevelopment of SLE.
In summary, our results indicate that SLE is amen-able to genetic analysis and the susceptibility loci forthe disease appear to be possible to define. Our resultstogether with the previous studies of human SLE[22–24] support the suggestion of extensive geneticheterogeneity based on mouse lupus genome scanresults [14, 15]. A key issue for identification of sus-ceptibility loci for human SLE is to reduce the geneticheterogeneity and possibly also the phenotypic com-plexity. We believe that using well-characterizedhuman family resources from relatively homogeneouspopulations will be a necessity.
Acknowledgements
The authors would like to thank first of all the patientsand their relatives for their participation in this study.This work has been supported by grants from theTore Nilssons Foundation, Ake Wibergs Foundation,the Marcus Borgstroms Foundation and the SwedishMedical Research Council (Grant no. 12673) toM.E.A-R. The Swedish Medical Research Council toG.S. (Grant no. 9528) and L.T., the Landspitalınn andUniversity of Iceland Research Funds to K.S., theSwedish National Association against Rheumatism toM.E.A-R., G.S., L.T. and I.L., the Gustaf V 80thBirthday Fond to G.S., L.T. and I.L., the MargaretaRheumatology Research Foundation to I.L., and theBeijer Foundation and the Borje Dahlins Fond toU.B.G.. The work was also supported by the EuropeanCommission (Grant no. BMH4-CT98–3489). J.D.T.has been partly supported by a Hitchings-ElionFellowship from the Burroughs Wellcome Fund.
References1. Alarcon-Segovia D., Diaz-Jouanen E. 1980. Lupus
subsets: relationship to genetic and environmentalfactors. Semin Arthritis Rheum 10: 18–24
2. Winchester R.J., Nunez-Roldan A. 1982. Some geneticaspects of systemic lupus erythematosus. ArthritisRheum 25: 833–837
3. Reichlin M., Harley J.B., Lockshin M.D. 1992. Serologicstudies of monozygotic twins with systemic lupuserythematosus. Arthritis Rheum 35: 457–464
4. Deapen D., Escalante A., Weinrib L., Horwitz D.,Bachman B., Roy-Burman P., Walker A., Mack T.M.1992. A revised estimate of twin concordance insystemic lupus erythematosus. Arthritis Rheum 35:311–318
6. Jarvinen P., Aho K. 1994. Twin studies in rheumaticdiseases. Semin Arthritis Rheum 24: 19–28
7. Hochberg M.C. 1987. The application of geneticepidemiology to systemic lupus erythematosus.J Rheumatol 14: 867–869
8. Gudmundsson S., Steinsson K. 1990. Systemic lupuserythematosus in Iceland 1975 through 1984. Anationwide epidemiological study in an unselectedpopulation. J Rheumatol 17: 1162–1167
9. Jonsson H., Nived O., Sturfelt G., Silman A. 1990.Estimating the incidence of systemic lupuserythematosus in a defined population using multiplesources of retrieval. Br J Rheumatol 29: 185–188
12. Drake C.G., Babcock S.K., Palmer E., Kotzin B.L. 1994.Genetic analysis of the NZB contribution to lupus-likeautoimmune disease in (NZB ×NZW) F1 mice. ProcNatl Acad Sci USA 91: 4062–4066
13. Kono D.H., Burlingame R.W., Owens D.G., KuramochiA., Balderas R.S., Balomenos D., Theofilopoulos A.N.1994. Lupus susceptibility loci in New Zealand mice.Proc Natl Acad Sci USA 91: 10168–10172
14. Morel L., Wakeland E.K. 1998. Susceptibility to lupusnephritis in the NZB/W model system. Curr OpinImmunol 10: 718–725
18. Walport M.J. 1993. The Roche Rheumatology PrizeLecture. Complement deficiency and disease. Br JRheumatol 32: 269–73
19. Alarcon-Segovia D., Alarcon-Riquelme M.E. 1998.Etiopathogenesis of systemic lupus erythematosus: Atale of three troikas. In: Systemic Lupus Erythematosus.R. Lahita, ed, Academic Press; pp. 55–65
20. Tsao B.P., Cantor R.M., Kalunian K.C., Chen C.J.,Badsha H., Singh R., Wallace D.J., Kitridou R.C., ChenS.L., Shen N., Song Y.W., Isenberg D.A., Yu C.L., HahnB.H., Rotter J.I. 1997. Evidence for linkage of acandidate chromosome 1 region to human systemiclupus erythematosus. J Clin Invest 99: 725–731
21. Tsao B.P., Cantor R.M., Grossman J.M., Shen N.,Teophilov N.T., Wallace D.J., Arnett F.C., Hartung K.,Goldstein R., Kalunian K.C., Hahn B.H., Rotter J.I.1999. PARP alleles within the linked chromosomalregion are associated with systemic lupuserythematosus. J Clin Invest 103: 1135–1140
A new susceptibility locus for human SLE on 2q 177
23. Moser K.L., Neas B.R., Salmon J.E., Yu H.,Gray-McGuire C., Asundi N., Bruner G.R., Fox J.,Kelly J., Henshall S., Bacino D., Dietz M., Hogue R.,Koelsch G., Nightingale L., Shaver T., Abdou N.I.,Albert D.A., Carson C., Petri M., Treadwell E.L., JamesJ.A., Harley J.B. 1998. Genome scan of human systemiclupus erythematosus: evidence for linkage onchromosome 1q in African-American pedigrees. ProcNatl Acad Sci USA 95: 14869–14874
24. Shai R., Quismorio Jr F.P., Li L., Kwon O.J., MorrisonJ., Wallace D.J., Neuwelt C.M., Brautbar C.,Gauderman W.J., Jacob C.O. 1999. Genome-widescreen for systemic lupus erythematosus susceptibilitygenes in multiplex families [In Process Citation]. HumMol Genet 8: 639–644
25. Cavalli-Sforza L.L., Menzzi P., Piazza A. 1993. Thehistory and geography of human genes. 2nd edn.Princeton University Press, Princeton
26. Terwilliger J.D. 1995. A powerful likelihood methodfor the analysis of linkage disequilibrium between traitloci and one or more polymorphic marker loci. Am JHum Genet 56: 777–787
27. Steinsson K., Jonsdottir S., Arason G.J., KristjansdottirH., Fossdal R., Skaftadottir I., Arnason A. 1998. Astudy of the association of HLA DR, DQ, andcomplement C4 alleles with systemic lupuserythematosus in Iceland. Ann Rheum Dis 57:503–505
28. Truedsson L., Sturfelt G., Johansen P., Nived O.,Thuresson B. 1995. Sharing of MHC haplotypes amongpatients with systemic lupus erythematosus fromunrelated Caucasian multicase families: diseaseassociation with the extended haplotype [HLA-B8,SC01, DR17]. J Rheumatol 22: 1852–1861
29. Tan E.M., Cohen A.S., Fries J.F., Masi A.T., McShaneD.J., Rothfield N.F., Schaller J.G., Talal N., WinchesterR.J. 1982. The 1982 revised criteria for the classificationof systemic lupus erythematosus. Arthritis Rheum 25:1271–1277
30. Sheffield V.C., Weber J.L., Buetow K.H., Murray J.C.,Even D.A., Wiles K., Gastier J.M., Pulido J.C., YandavaC., Sunden S.L., et al. 1995. A collection of tri- andtetranucleotide repeat markers used to generate highquality high resolution human genome-wide linkagemaps. Hum Mol Genet 4: 1837–1844
31. Lindqvist A.K., Magnusson P.K., Balciuniene J.,Wadelius C., Lindholm E., Alarcon-Riquelme M.E.,Gyllensten U.B. 1996. Chromosome-specific panels oftri- and tetranucleotide microsatellite markers formultiplex fluorescent detection and automatedgenotyping: evaluation of their utility in pathologyand forensics. Genome Res 6: 1170–1176
32. Morton N.E. 1955. Sequential tests for the detection oflinkage. Am J Hum Genet 7: 277–318
33. Ott J. 1991. Analysis of human linkage. The JohnsHopkins University Press, Baltimore, USA, pp. 325
34. Cottingham Jr R.W., Idury R.M., Schaffer A.A. 1993.Fast Sequential Genetic Linkage computations. Am JHum Genet 53: 252–263
35. Scaffer A.A., Gupta S.K., Shiram K., Cottingham JrR.W. 1994. Avoiding recomputation in Genetic LinkageAnalysis. Hum Hereditary 44: 225–37
36. Kuokkanen S., Sundvall M., Terwilliger J.D., TienariP.J., Wikstrom J., Holmdahl R., Pettersson U., PeltonenL. 1996. A putative vulnerability locus to multiplesclerosis maps to 5p14–p12 in a region syntenic to themurine locus Eae2. Nat Genet 13: 477–480
37. Smith C.A.B. 1963. Testing for heterogeneity ofrecombination fraction values in human genetics. AnnHum Genet 27: 175–182
38. Lander E., Kruglyak L. 1995. Genetic dissection ofcomplex traits: guidelines for interpreting andreporting linkage results. Nat Genet 11: 241–247
39. Field L.L., Tobias R., Magnus T. 1994. A locus onchromosome 15q26 (IDDM3) produces susceptibility toinsulin-dependent diabetes mellitus. Nat Genet 8:189–194
40. Houlston R.S., Tomlinson I.P., Ford D., Seal S.,Marossy A.M., Ferguson A., Holmes G.K., Hosie K.B.,Howdle P.D., Jewell D.P., Godkin A., Kerr G.D.,Kumar P., Logan R.F., Love A.H., Johnston S., MarshM.N., Mitton S., O’Donoghue D., Roberts A.,Walker-Smith J.A., Stratton M.F. 1997. Linkage analysisof candidate regions for coeliac disease genes. HumMol Genet 6: 1335–1339
41. Arnett F.C., Moulds J.M. 1991. HLA class III moleculesand autoimmune rheumatic diseases. Clin ExpRheumatol 9: 289–296
42. Fielder A.H., Walport M.J., Batchelor J.R., Rynes R.I.,Black C.M., Dodi I.A., Hughes G.R. 1983. Family studyof the major histocompatibility complex in patientswith systemic lupus erythematosus: importance of nullalleles of C4A and C4B in determining diseasesusceptibility. Br Med J (Clin Res Ed) 286: 425–428
43. Sturfelt G., Truedsson L., Johansen P., Jonsson H.,Nived O., Sjoholm A.G. 1990. Homozygous C4Adeficiency in systemic lupus erythematosus: analysisof patients from a defined population. Clin Genet 38:427–433
44. Hong G.H., Kim H.Y., Takeuchi F., Nakano K., YamadaH., Matsuta K., Han H., Tokunaga K., Ito K., Park K.S.1994. Association of complement C4 and HLA-DRalleles with systemic lupus erythematosus in Koreans.J Rheumatol 21: 442–447
45. Davies E.J., Steers G., Ollier W.E., Grennan D.M.,Cooper R.G., Hay E.M., Hillarby M.C. 1995. Relativecontributions of HLA-DQA and complement C4A lociin determining susceptibility to systemic lupuserythematosus. Br J Rheumatol 34: 221–225
46. Skarsvag S. 1995. The importance of C4A null genesin Norwegian patients with systemic lupuserythematosus. Scand J Immunol 42: 572–576
47. Steinsson K., Arnason A., Erlendsson K., Fossdal R.,Skaftadottir I., Jonsdottir S., Fjalarson M.,Thorsteinsson J. 1995. A study of the majorhistocompatibility complex in a Caucasian family withmultiple cases of systemic lupus erythematosus:association with the C4AQ0 phenotype. J Rheumatol 22:1862–1866
48. Granados J., Vargas-Alarcon G., Andrade F.,Melin-Aldana H., Alcocer-Varela J., Alarcon-Segovia D.1996. The role of HLA-DR alleles and complotypesthrough the ethnic barrier in systemic lupuserythematosus in Mexicans. Lupus 5: 184–189
178 A. B. Lindqvist et al.
49. Sim E., Cross S.J. 1986. Phenotyping of humancomplement component C4, a class-III HLA antigen.Biochem J 239: 763–767
50. Johannesson B., Steinsson K., Lindqvist A.-K.,Kristjansdottir H., Grondal G., Sandino S., TjernstromF., Sturfelt G., Granados-Arriola J., Alcocer-Varela J.,Lundberg I., Jonasson I., Truedsson L., SvenungssonE., Klareskog L., Alarcon-Segovia D., Gyllensten U.B.,Alarcon-Riquelme M.E. 1999. A comparative analysisof genome-scans performed in multicase families withSystemic Lupus Erythematosus from differentpopulation groups. J Autoimm 13: 137–141
51. Eskdale J., Wordsworth P., Bowman S., Field M.,Gallagher G. 1997. Association betweenpolymorphisms at the human IL-10 locus and systemiclupus erythematosus. Tissue Antigens 49: 635–9
52. Alarcon-Riquelme M.E., Lindqvist A.K., Jonasson I.,Johannesson B., Sandino S., Alcocer-Varela J.,Granados J., Kristjansdottir H., Grondal G.,Svenungsson E., Lundberg I., Steinsson K., KlareskogL., Sturfelt G., Truedsson L., Alarcon-Segovia D.,
Gyllensten U.B. 1999. Genetic analysis of thecontribution of IL-10 to Systemic LupusErythematosus. J Rheumatol 26: 2148–2152
53. Ware M.D., Rosten P., Damen J.E., Liu L., HumphriesR.K., Krystal G. 1996. Cloning and characterization ofhuman SHIP, the 145-kD inositol 5-phosphatase thatassociates with SHC after cytokine stimulation. Blood88: 2833–2840
54. Liu Q., Dumont D.J. 1997. Molecular cloning andchromosomal localization in human and mouse of theSH2-containing inositol phosphatase, INPP5D (SHIP).Amgen EST Program. Genomics 39: 109–112
55. Ono M., Bolland S., Tempst P., Ravetch J.V. 1996. Roleof the inositol phosphatase SHIP in negativeregulation of the immune system by the receptor Fc(gamma)RIIB. Nature 383: 263–266
56. Sato S., Tuscano J.M., Inaoki M., Tedder T.F. 1998.CD22 negatively and positively regulates signaltransduction through the B lymphocyte antigenreceptor. Semin Immunol 10: 287–297
