Genes and Immunity (2001) 2, 205–210
 2001 Nature Publishing Group All rights reserved 1466-4879/01 $15.00
www.nature.com/gene
A genome screen for multiple sclerosis in Italian
families
S Broadley1, S Sawcer1, S D’Alfonso2, A Hensiek1, F Coraddu1, J Gray1, R Roxburgh1, D Clayton3,
C Buttinelli4, A Quattrone5, M Trojano6, L Massacesi7 and A Compston1
1
University of Cambridge Neurology Unit, Addenbrooke’s Hospital, Hills Road, Cambridge, CB2 2QQ, UK; 2Chair of Human
Genetics, Department of Medical Sciences, University of Piemonte Orientale ‘A. Avogadro’, Novara, Italy; 3MRC Biostatistics Unit,
Institute of Public Health, University Forvie Site, Robinson Way, Cambridge, CB2 2SR, UK; 4Department of Neurological Sciences,
University “La Sapienza”, Rome, Italy; 5Department of Medical Sciences, University of Catanzaro and Institute of Experimental
Medicine and Biotechnology, CNR, Italy; 6Department of Neurological and Psychiatric Sciences, University of Bari, Italy;
7
Department of Neurological and Psychiatric Sciences, University of Florence, Italy
We have screened the whole genome for linkage in 40 Italian multiplex families with multiple sclerosis using 322 markers.
The GENEHUNTER-PLUS program was used to analyse these data and revealed eight regions of potential linkage where
the lod score exceeds the nominal 5% significance level (0.7). No region of linkage with genome-wide significance was
identified and none of the markers showed evidence of statistically significant transmission disequilibrium. Overall these
results have refined the linkage data relating to this disease in Italians modestly supporting some previously identified
areas of interest and helping to exclude others. Genes and Immunity (2001) 2, 205–210.
Keywords: multiple sclerosis; linkage; genome screen; genetic susceptibility
Introduction
It is well established that genetic factors influence susceptibility to multiple sclerosis1 but, the responsible genes
remain largely undefined, association with class II alleles
of the MHC being the only established genetic feature of
the disease. In recent years, developments in the human
genome project2–6 have enabled researchers to perform
systematic searches for linkage in complex diseases
where the genetic effects attributable to involved genes
are modest and therefore very large genotyping efforts
are required. To date, four such whole genome linkage
screens have been completed in multiple sclerosis.7–10
Although none of these screens succeeded in identifying
statistically unequivocal linkage, each suggested several
regions of potential linkage and together they provide the
best available summary of the likely location of the relevant genes.11 In considering these screens, it is important
to remember that the power of this approach is critically
dependant upon the frequency of susceptibility alleles in
the population studied.12 Despite the failure of previous
linkage screens to solve the basis for genetic susceptibility
Correspondence: L Massacesi, Dip. Scienze Neurologiche e Psichiatriche,
Univ. di Firenze, V.le Morgagni, 85 50134 Firenze, Italy.
E-mail: luca.massacesi얀unifi.it
This work was supported by the Multiple Sclerosis Society of Great
Britain and Ireland, the Medical Research Council, the Wellcome
Trust, the Italian Multiple Sclerosis Foundation (FISM), the
National Project on Multiple Sclerosis of the Istituto Superiore di
Sanità and by the Italian Ministry of University Scientific Research
and Technology.
Received 24 November 2000; revised and accepted 9 March 2001
in multiple sclerosis, the approach is still worth pursuing
in populations where the frequency of susceptibility
alleles might be more favourable.
The prevalence of multiple sclerosis in Italy is approximately 56 per 100 000 and this is intermediate between
the rates seen in northern Europe and those seen in north
Africa.13,14 While this intermediate rate of disease might
be the result of environmental factors, variation in the
frequency of susceptibility alleles is a more likely explanation.15 These observations suggest that, for some loci,
linkage studies based on Italian families might have
greater power than those based on families from populations with higher rates of disease where the normal frequency of susceptibility alleles may be disadvantageous.
Working on this principle we have previously searched
for genetic linkage to multiple sclerosis in Italian families,
concentrating on those regions of the genome suggested
to be the most promising in existing linkage screens.16
Since the results of this analysis revealed a degree of concordance with previous screens but left most of the genome unexplored, we have now extended the analysis of
Italian families to include the remainder of the genome.
Results and discussion
The 40 families considered in this screen included 37 with
two affected siblings, one with three affected siblings and
two with two affected first cousins. Our previous study16
involved 36 of these families (34 of the sibling pairs, one
of the cousin pairs and the sibling trio) together with
three additional sibling pair families, an additional
cousin pair and an uncle/nephew pair; unfortunately
Italian MS genome screen
S Broadley et al
206
Figure 1 The lod score calculated by GENEHUNTER-PLUS18,19 at each point along the genome is indicated, one graph for each chromosome.
The length of each x-axis is proportional to the genetic length of the corresponding chromosome and the map position of the markers is
indicated by the tick marks. Marker names are placed as near as possible to their appropriate tick mark within the limits of clarity; tick
marks for markers with less than 2 cM separation have in some instances been superimposed. The y-axis is scaled from 0 to 1.5 in each
case. The families and markers considered in this screen overlap with those included in our previous study.16 The mean observed heterozygosity for the 322 markers was 79% and their mean PIC was 76%. The genetic position of each marker was established from the location
database (LDB, http://cedar.genetics.soton.ac.uk/publicFhtml/)24 except for markers not included in this database which were placed
midway between their flanking markers as derived from the map of Davies et al.25 Because only small quantities of genomic DNA were
available on each individual, a polymerase extension PCR (PEP) protocol was used non-specifically to amplify individual DNA samples
prior to genotyping.26 Genotyping was then achieved by PCR amplification of each microsatellite followed by electrophoresis on Applied
Biosystems 373A sequencing machines, with allele sizing and calling being performed using GENESCAN and GENOTYPER software
(Perkin-Elmer, Foster city, CA, USA). Allele frequencies used in the analysis were derived from the data using the SPLINK program
version 1.
insufficient DNA was available from these additional five
families to allow their inclusion in this extended study.
The four new families included three with two affected
siblings and one with two affected cousins. A total 322
markers (321 microsatellites and HLA-DRB1) were used.
Genes and Immunity
In our previous study16 we typed just 67 microsatellite
markers (together with HLA-DRB1 and APOE). Only 42
of these 67 microsatellite markers are re-included in the
current analysis; the remaining 25 were omitted in order
to avoid distorting the marker map, a process which can
Italian MS genome screen
S Broadley et al
207
Figure 1 Continued
Genes and Immunity
Italian MS genome screen
S Broadley et al
208
Figure 1 Continued
exaggerate lod scores.17 Analysis of these overlapping 42
markers was extended to include the four families not
previously typed. All 40 families were typed for the 279
new microsatellite markers.
The results of multipoint nonparametric linkage analysis on each chromosome are shown in Figure 1. The
analysis was performed using GENEHUNTER-PLUS.18,19
Although there are no regions suggestive of linkage with
genome-wide statistical significance,20 eight regions were
identified where the lod score exceeds 0.7 (the nominal
5% significance level), chromosomes 1q42, 1q44, 2q36,
5q33, 6pter, 6q22, 10cen and 15q21. The conservative
interpretation is that the number and size of the linkage
peaks seen in this screen do not exceed chance expectation.
The extent to which the markers have extracted the
available genetic information from these 40 families was
Genes and Immunity
also tested by GENEHUNTER-PLUS. Average information content across the genome was 51%, ranging from
89% where informative markers were typed to 8% where
no markers were typed such as the short arm of chromosome 22.
Each of the microsatellite markers was tested for evidence of transmission disequilibrium using the TRANSMIT program. Five markers (D2S169, D4S406, D8S257,
D22S21 and DXS991) showed evidence of transmission
disequilibrium at the nominal 5% significance level.
However after correcting for multiple testing none of
these potential associations was statistically significant.
In summary, we have completed a genome-wide
screen for linkage in Italian families with multiple sclerosis, several regions of interest have been identified but
none of these have genome-wide statistical significance.
The previous linkage genome screens in multiple scler-
Italian MS genome screen
S Broadley et al
209
Figure 1 Continued
osis7–10 used populations of northern European descent
and clearly demonstrated that, in these populations, the
effect of individual genes determining susceptibility to
multiple sclerosis is modest.21,22 It remained possible that
in a southern European population the frequency of susceptibility alleles might be more favourable12 and thereby
enable linkage to be identified with greater ease.
That this is not the case has several possible explanations. The modest number of families considered
means that the power of our study is low, making it possible for important effects easily to be missed and suggesting that our statistically negative results could represent
a type II error. We had hypothesised that the lower frequency of multiple sclerosis in southern Europeans might
reflect a lower frequency of susceptibility alleles in this
population. We further hypothesised that this might
result in southern European populations being more
favourable for linkage analysis. Tacit in this hypothesis
is the suggestion that the susceptibility alleles for some
genes have frequencies in northern Europeans that are
prohibitively high (⬎90%). However it is possible that
susceptibility allele frequencies in northern Europeans
are more modest (20–80%) and by moving to a southern
European population we have simply reduced them to
an even less favourable position (⬍10%). If this is the
case, then the southern European population offers no
special advantage in terms of power. Unfortunately the
power of our present study is too limited to know which
of these possibilities is correct.
Despite the limitations, the peaks of linkage we have
observed show some concordance with results from other
studies. Perhaps the most interesting result is that on
chromosome 10cen, which is close to a region previously
identified in the multiple sclerosis genome screen of
Haines et al8 and broadly corresponds to the location of
the iddm10 locus,23 suggesting that a gene conferring risk
to autoimmunity in general may be encoded at this locus.
The result in the region of HLA is as expected. The
association of class II HLA alleles with multiple sclerosis
is well known but the effect of this locus on susceptibility
is modest and therefore evidence for linkage also
expected to be small, as in the previous genome
screens.7–10
In considering how this new screen contributes to our
understanding of previously suggested regions of interest it is important to remember that, in such a small
study, any positive lod score can be considered to be supportive, even if it is not statistically significant. Modest
but genuine effects which have previously generated
results that are only just significant would be expected
to be even less obvious in smaller replicate analyses,
whereas false positive results would be more likely to
show zero lod scores. Furthermore, linkage peaks can be
quite ectopic relative to the location of the causative gene.
Similarly the limited power of this study means that we
cannot confidently rule out the possibility that projects
based on modest risk populations may be better able to
detect susceptibility genes than those based on high
risk populations.
Although additional strategies are being developed,
family-based linkage genome screening can still make
important contributions to the genetic analysis of complex traits such as multiple sclerosis and the continued
expansion of such linkage data should be encouraged.
Comparing and contrasting the growing body of linkage
screens will undoubtedly aid in the reliable identification
of genomic regions containing susceptibility genes.
Genes and Immunity
Italian MS genome screen
S Broadley et al
210
References
1 Hogancamp WE, Rodriguez M, Weinshenker BG. The epidemiology of multiple sclerosis. Mayo Clin Proc 1997; 72: 871–878.
2 Weber JL, May PE. Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction.
Am J Hum Genet 1989; 44: 388–396.
3 Diehl SR, Ziegle J, Buck GA, Reynolds TR, Weber JL. Automated
genotyping of human DNA polymorphisms. Am J Hum Genet
1990; 47: A177.
4 Ziegle JS, Su Y, Corcoran KP et al. Application of automated
DNA sizing technology for genotyping microsatellite loci.
Genomics 1992; 14: 1026–1031.
5 Gyapay G, Morissette J, Vignal A et al. The 1993–94 Genethon
human genetic linkage map. Nat Genet 1994; 7: 246–339.
6 Dib C, Faure S, Fizames C et al. A comprehensive genetic map
of the human genome based on 5,264 microsatellites. Nature
1996; 380: 152–154.
7 Sawcer SJ, Jones HB, Feakes R et al. A genome screen in multiple
sclerosis reveals susceptibility loci on chromosomes 6p21 and
17q22. Nat Genet 1996; 13: 464–468.
8 Haines JL, Ter-Minassian M, Bazyk A et al. A complete genomic
screen for multiple sclerosis underscores a role for the major
histocompatibility complex. Nat Genet 1996; 13: 469–471.
9 Ebers GC, Kukay K, Bulman DE et al. A full genome search in
multiple sclerosis. Nat Genet 1996; 13: 472–476.
10 Kuokkanen S, Gschwend M, Rioux JD et al. Genomewide scan
of multiple sclerosis in Finnish multiplex families. Am J Hum
Genet 1997; 61: 1379–1387.
11 Wise LH, Lanchbury JS, Lewis CM. Meta-analysis of genome
searches. Ann Hum Genet 1999; 63: 263–272.
12 Risch N, Merikangas K. The future of genetic studies of complex
human diseases. Science 1996; 273: 1516–1517.
13 Compston A, Ebers G, Lassmann H et al. McAlpine’s Multiple
Sclerosis. Churchill Livingstone: London, 1998.
14 Totaro R, Marini C, Cialfi A, Giunta M, Carolei A. Prevalence of
Genes and Immunity
15
16
17
18
19
20
21
22
23
24
25
26
27
multiple sclerosis in the L’Aquila district, central Italy. J Neurol
Neurosurg Psych 2000; 68: 349–352.
Compston DAS. The dissemination of multiple sclerosis. The
Langdon-Brown lecture 1989. J R Coll Phys Lond 1990; 24: 207–
218.
D’Alfonso S, Nistico L, Zavattari P et al. Linkage analysis of multiple sclerosis with candidate region markers in Sardinian and
continental Italian families. Eur J Hum Genet 1999; 7: 377–385.
Kruglyak L. The use of a genetic map of biallelic markers in
linkage studies. Nat Genet 1997; 17: 21–24.
Kruglyak L, Daly MJ, Reeve-Daly MP, Lander ES. Parametric
and nonparametric linkage analysis: a unified multipoint
approach. Am J Hum Genet 1996; 58: 1347–1363.
Kong A, Cox NJ. Allele-sharing models: LOD scores and accurate linkage tests. Am J Hum Genet 1997; 61: 11790–1188.
Lander E, Kruglyak L. Genetic dissection of complex traits:
guidelines for interpreting and reporting linkage results. Nat
Genet 1995; 11: 241–247.
Sawcer SJ, Goodfellow PN, Compston DAS. The genetic analysis
of multiple sclerosis. TIGs 1997; 13: 234–239.
Dyment DA, Sadovnick AD, Ebers GC. Genetics of multiple
sclerosis. Hum Mol Genet 1997; 6: 1693–1698.
Reed P, Cucca F, Jenkins S et al. Evidence for a type 1 diabetes
susceptibility locus (IDDM10) on human chromosome 10p11q11. Hum Mol Genet 1997; 6: 1011–1016.
Collins A, Frezal J, Teague J, Morton NE. A metric map of
humans: 23,500 loci in 850 bands. Proc Natl Acad Sci USA. 1996;
93: 14771–14775.
Davies JL, Kawaguchi Y, Bennett ST et al. A genome-wide search
for human type 1 diabetes susceptibility genes. Nature 1994; 371:
130–136.
Zhang L, Cui X, Schmitt K et al. Whole genome amplification
from a single cell: implications for genetic analysis. Proc Natl
Acad Sci USA 1992; 89: 5847–5851.
Holmans P, Clayton D. Efficiency of typing unaffected relatives
in an affected sib-pair linkage study. Am J Hum Genet 1995; 57:
1221–1232.