Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy, genetic homogeneity, and mapping of the locus within a 2-cM interval

(1)

Article

Reference

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy, genetic homogeneity, and mapping of the

locus within a 2-cM interval

DUCROS, A, et al.

Abstract

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is a recently identified autosomal dominant cerebral arteriopathy characterized by the recurrence of subcortical infarcts leading to dementia. A genetic linkage analysis conducted in two large families recently allowed us to map the affected gene on chromosome 19 in a 12-cM interval bracketed by D19S221 and D19S215. In the present study, these first 2 families and 13 additional ones, including a total of 199 potentially informative meiosis, have been genotyped with eight polymorphic markers located between D19S221 and D19S215. All families were linked to chromosome 19. The highest combined lod score (Zmax = 37.24 at theta = .01) was obtained with marker D19S841, a new CAn microsatellite marker that we isolated from chromosome 19 cosmids. The recombinant events observed within these families were used to refine the genetic mapping of CADASIL within a 2-cM interval that is now bracketed by D19S226 and D19S199 on 19p13.1. These data strongly suggest the genetic homogeneity of this recently identified condition and establish [...]

DUCROS, A, et al. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy, genetic homogeneity, and mapping of the locus within a 2-cM interval.

American Journal of Human Genetics, 1996, vol. 58, no. 1, p. 171-81

PMID : 8554054

Available at:

http://archive-ouverte.unige.ch/unige:46776

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

Am. J. Hum. Genet. 58:171-181, 1996

Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy, Genetic Homogeneity, and Mapping of

the Locus within

a

2-cM Interval

A.

Ducros,'

T. Nagy,1 S.

Alamowitch,'

A.

Nibbio,'

A.

Joutel,'

K. Vahedi," ² H. Chabriat,2 M. T. Iba-Zizen,3 J. Julien,9 P. Davousi J. Y. Goas,10 0. Lyon-Caen,5 B. Dubois,5

X.

Ducrocq,"1

F. Salsa,12 M. Ragno,13 P.

Burkhard,"4

^C.

Bassetti,"5

M. Hutchinson,16 M.

Verin,"7

F. Viader,18 F. Chapon,18 M. Levasseur,6 J. L. Mas,7 ^0. Delrieu,' J. Maciazek,1 M. Prieur,8 H. Mohrenweiser,'9 J. F. Bach,1 M. G. Bousser,2 and E. Tournier-Lasserve1

'INSERM U25,2Service deneurologie,CHU St.Antoine, 3IRMH6pitaldesQuinze-Vingt, 4H6pital d'Argenteuil, 'CHUPiti6-Salpetriere, 6H6pital d'Orsay, 7CH St.Anne,and 8Laboratoire decytogenetique, Paris; 9H6pital HautL'Eveque, Pessac, France; '"CHU deBrest, Brest;

"CHU de Nancy, Nancy; '2Ospedale Civile, Bassanodel Grappa, Italy;13C&G MazzoniHospital,AscoliPisceno, Italy; 1'H6pitalcantonal deGeneve, Geneva; `1H6pitaldeBern, Bern; "6St.Vincent's Hospital, Dublin; "7CHU deRennes, Rennes; "8CHU deCaen, Caen;and '9Lawrence Livermore, CANational Laboratories, Livermore,CA

Summary

Cerebral autosomal dominantarteriopathywith subcortical infarcts and leukoencephalopathy (CADASIL)is arecently identified autosomal dominant cerebralarteriopathy characterized by the recurrence of subcortical infarcts leading to dementia. A genetic linkage analysis conducted in two large families recently allowed us to map the affected gene on chromosome 19 in a 12-cM interval bracketed by D19S221 and D19S215. In the present study, these first 2 families and 13 additional ones, including a total of 199 potentially informative meiosis, have been genotyped with eight polymorphic markers located between D19S221 and D19S215. All families were linked tochromosome19. The highest combined lod score

(Z,,i.

⁼37.24 at0 = .01) was obtained with marker D19S841, a new

CA&

microsatellite marker that we isolated from chromosome 19 cosmids.

Therecombinant events observed within these families were used to refine the genetic mapping ofCADASIL within a 2-cM interval that is now bracketed by D19S226 and D19S199 on 19pI3.1. These data strongly suggest the genetic homogeneity of this recently identified condition and establish the value of its clinical andneuroimaging diagnostic criteria. Besides their importance for the ongoing posi- tional cloning of the CADASIL gene, these data help to refine the genetic mapping ofCADASILrelative tofamilial hemiplegic migraine and hereditaryparoxysmal cerebellar ataxia, conditions that we both mapped within the same chromosome 19region.

Received August7, 1995;accepted forpublication October 5, 1995.

Address for correspondence andreprints: Dr. E.Tournier-Lasserve, INSERM U25, Facult6 de Medecine Necker, 156 rue de Vaugirard, Paris75015, France.

0002-9297/96/5801-0020$02.00

Introduction

Since1977, severalfamiliessufferingfromanautosomal dominant stroke condition of unknown etiology have been reported undervarious names such as hereditary multi-infarct dementia or familial sclerosingvasculopa- thy (Souranderand Walinder 1977; Stevens etal. 1977;

Colmant 1980; Sonninen andSavontaus1987).In1991, the clinical, neuroimaging, and genetic parameters of this hereditaryconditionwere precisely defined, on the basisoftheanalysis ofaverylargepedigreeoriginating from France (Tournier-Lasserve et al. 1991). This autosomal dominant disorder, now designated under the acronym CADASIL (cerebralautosomal dominantarte- riopathywithsubcortical infarcts andleukoencephalop- athy), is characterized, in the absence of any vascular risk factor, by the recurrence of subcortical infarcts startinginearly/midadulthood andleadingtodementia.

Cerebral magnetic resonance imaging (MRI) shows, in allclinically affected individuals,well delineatedimages of abnormal signal highly suggestive of small deep infarcts as well as a diffuse hypersignal on T2-weighted images of the cerebral white matter (Tournier-Lasserve et al. 1991). It is interesting that these white matter abnormalities (WMAs) are also observed in "at risk,"

butyetasymptomatic,individuals. It represents anearly stage of the disease, and MRI scanning is an absolute requirement to establish accurately the status of an individual for linkage analysis (Tournier-Lasserve et al.

1993). This condition is underlaid by a nonatheroscle- rotic, nonamyloid angiopathy affecting mainly the small arteriesof the white matter and the basal ganglia (Bau- drimont et al. 1993). Linkage analysis of this first large Frenchpedigree allowedus to map the affected gene on chromosome 19 in a 14-cM intervalbetween D19S221 and D19S222. These results were confirmed in a second unrelated Frenchfamily(Tournier-Lasserve et al. 1993).

171

(3)

Fl France

F3France

1 2

4 5 6

F6 France

3 4 5 6 7 8 9 10 11

Fl0 France

2::i~nL:I-

3 U 4 5 6 7 8 9 10 11 12

Fll France

6 9 10 11 12 13 14 U

F13 France F14Ireland F16 Switzerland

2 3 4 5 6 7 8 9 10

F18 Switzerland F19France F21 Italy

2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 4 5 6 7 8 9

Figure 1 Pedigrees. Black-filled symbolsrepresentclinically affectedindividuals;half-blackenedsymbolsrepresentasymptomaticsubjects

with abnormal MRI; opensymbols representhealthy subjects; cross-hatched symbolsrepresent subjectsof unknown status; symbolswitha black central point represent ancestors whowere mostlikelycarriersof the affected gene; andquestionmarkinthe middleofanemptysymbol

represents individual for whom no history was available. The letter U represents individuals of unknown status that haplotype analysis contributed tothe high-resolutiongeneticmapping of the D19S221-D19S215 interval. The originofeachpedigree^isindicated.

F2France

(4)

Ducros etal.: Mapping of CADASIL

Table 1

CADASIL Families

No. OFSAMPLED INDIVIDUALS Affected

Unknown

FAMILY Clinical MRIOnly Healthy Status Spouses Total

F1 11 9 8 0 10 38

F2 2 4 3 1 2 12

F3 2 2 0 2 2 8

F5 6 5 12 6 4 33

F6 3 1 7 0 3 14

F8 9 8 10 3 2 32

F10 7 0 4 1 1 13

F11 4 3 7 0 3 17

F12 4 4 14 3 4 29

F13 4 1 2 0 2 9

F14 8 2 1 0 2 13

F16 2 1 7 9 7 26

F18 4 0 3 2 2 11

F19 3 0 4 1 0 8

F21 2 3 3 1 1 10

Total 71 43 85 29 45 273

NOTE.-Detailedclinical descriptions are available formostof these15families: F1(Tournier-Lasserve etal. 1991);F2 (Davous etal. 1991); F6, F8, F10, F11,and F13(Chabriatetal. 1995);F12(Ragnoetal.

1995); F18 (Jung etal. 1995); and F21 (Gray et al. 1994).

Subsequently, the size of the mapping interval was re- ducedto 12 cM on19pbetweenD19S221andD19S215 (Joutel et al. 1993).

Since 1991,anincreasingnumber of families originat- ingfromdifferent countries and sharing clinical andneuroimaging featuresstrikingly similar to thosepreviously described have been reported (Davouset al. 1991; Mas etal. 1992;Salvietal. 1992; Grayetal. 1994; Gutierrez- Molina etal. 1994; Jungetal. 1995; Ragno et al. 1995;

Sabbadini etal. 1995; Chabriat etal. 1995). Fifty such families have beenreferredtoourlabsince 1993,which suggests that the prevalence of this disorder is much higher than anticipated. Detailed clinical analysis of thesefamilies showed that the "stroke phenotype" present in >80% of the patients was not the only clinical phenotype of CADASIL patients, since >30% of them also suffer from migraine with aura (as compared to 1%-6% in controlpopulations), and 15% have mood disorders (Chabriatetal. 1995).Incidentally, this obser- vationled us to suspect the implication of the CADASIL gene in the pathophysiology of migraine and to map a generesponsible for familial hemiplegic migraine (FHM) within thesameregion onchromosome 19 (Joutel et al.

1993). Thepresence of a cerebellar ataxia in -20% of FHMfamilies led us to suspect and establish the genetic mapping of another neurological autosomal dominant disease, hereditary paroxysmal cerebellar ataxia (HPCA), on chromosome 19 in close vicinity to the FHMand CADASIL loci (Vahedi et al. 1995).

Chr.19

I p-

I

D196221

D19S226 D19S841 DD19S253-D199244

D19S411- I .01

D04

D19S215

I qtr

Figure 2 Schematic regional map of chromosome 19p. Pub- lished markers used for linkage analysis are indicated with their respec- tivedistancesontheright(0).Markerspositioned by genetic analysis of the CADASILpedigreesare ontheleft. According to our haplotyp- ingresults, D19S411, which was previously mapped at 0 = 0 from D19S226(Gyapayetal. 1994), is nowpositioned between D19S841 and D19S199. The positionof D19S841, a new marker that we isolated from a chromosome 19 cosmid, is indicated according to multipointandhaplotypingresults.

173

(5)

~~~~~~~~~~~~~~~~~~~~Am.J.Hum. Genet. 58:171-181, 1996

Table 2

PairwiseLinkageDataforCADASILandChromosome 19 Markers ZATO0=

LocusAND ^"

FAMILY .00 .01 .05 .10 .20 .30 .40 TOTAL 0

D19S221:

F1

F2 F3 F5 F6 F8

F11

F12 F13 F14 F16 F18 F19 F21

Total D19S226:

F1

F2 F3 FS F6 F8

F10

Fit F12 F13 F14 F16 F18 F19 F21

Total D19S841:

F1

F2 F3

FS

F6 F8

F10

Fit F12 F13 F14 F16 F18 F19 F21

Total D19S253:

F1

F2 F3 FS F6

-99.00 2.64 -3.09 -99.00 2.90 -99.00 -99.00 3.14 1.77 1.08 3.18 1.72

1.80 -99.00

-99.00 1.71 -3.40 -99.00 2.77 -99.00 2.33 -99.00 3.88

3.93 2.05 -.07 1.51 .55 -99.00

7.32 2.75 .66 -99.00 2.73 3.72 2.66 2.91 2.96 1.39 2.39 2.54 1.78 -.01 1.81 -99.00

5.37 .56 .70 2.36 .35

1.12 2.59 -1.10 1.99 2.85 6.08 .07 3.15 1.73 1.05 3.12 1.68

1.77 26.10

6.42 1.68 -1.40 3.11 2.72 4.25 2.28 -.19 4.42

3.86 2.01 -.06 1.48 .54 31.12

7.12 2.70 .65 2.27 2.69 3.64 2.61 2.86 2.95 1.36 2.33 2.49 1.74 -.01 1.77 37.24

5.82 .55 .68 2.33 .34

2.80 2.40 -.44 3.00 2.64 6.16 .63 3.11 1.59 .90 2.88 1.54

1.69 28.90

6.58 1.52 -.72 3.43 2.52 4.46 2.09 .40 4.63

3.56 1.85 -.56 1.35 .50 32.11

6.71 2.50 .59 2.71 2.49 3.39 2.41 2.65 2.82 1.24 2.09 2.27 1.60 -.01 1.63 35.08

4.91 .52 .62 2.18 .29

3.16 2.14 -.19 3.09 2.37 5.68 .74 2.93 1.41 .73 2.56 1.36

27.42

6.19 1.32 -.44 3.25 2.26 4.14 1.85 .55 4.38

3.17 1.65 -.04 1.18 .43 29.89

6.07 2.24 .52 2.65 2.23 3.10 2.15 2.37 2.59 1.10 1.79 1.98 1.42 -.01 1.44 31.61

4.34 .49 .53 1.94 .24

2.90 1.61 .01 2.61 1.78 4.36 .64 2.33 1.04 .44 1.91 .98

1.04 21.65

5.00 .92 -.19 2.51 1.70 3.18 1.33 .50 3.55

2.36 1.21 -.02 .83 .30 23.18

4.71 1.67 .37 2.14 1.68 2.38 1.58

1.76

1.97 0.79 1.19 1.40 1.01 -.01 1.04 23.69

3.41 .26 .39 1.55 .15

2.14 1.03 .07 1.75 1.14 2.86 .37 1.53 .63 .21 1.24 .56

.60 14.13

3.50 .52 -.08 1.54 1.09 2.07 .78 .29 2.48

1.50 .74 -.01 .47 .16 15.05

3.20 1.07 .23 1.39 1.07 1.54 .97 1.08 1.22 .46 .64 .85 .58 .00 0.60 14.89

2.29 .10 .24 1.07 .07

1.08 .44 .06 .70 .45 1.30 .10 .62 .22 .06 .58 .18

.19 6.00

1.74 .16 -.02 .51 .44 .93 .25 .08 1.22

.64 .28 -.00 .14 .04 6.41

1.53 .43 .10 .52 .43 .67 .35 .36 .42 .15 .20 .33 .18 .00 0.19 5.88

28.90

32.24

37.24

.05

.03

.01

1.06 .01 .10 .56 .02

(continued) 174

(6)

175 Ducrosetal.: Mappingof CADASIL

Table2 (continued)

ZAT0=

LOCUSAND Zmax

FAMILY .00 .01 .05 .10 .20 .30 .40 TOTAL 0

F8 F10 F11 F12 F13 F14 F16 F18 F19 F21

Total D19S244:

F1 F2 F3 F5 F6 F8 F10 F11 F12 F13 F14 F16 F18 F19 F21

Total D19S411:

F1 F2 F3 F5 F6 F8 F10 F11 F12 F13 F14 F16 F18 F19 F21

Total D19S199:

F1 F2 F3 F5 F6 F8 F10 F11 F12 F13 F14

5.35 .69 1.81 -99.00 1.22

1.20 1.20 -99.00 -99.00 1.34 .70 -99.00 .35 6.16 .51 .36 -99.00 1.80 3.07 1.51 .85 -.01 1.51 -99.00 3.44 .38 .70 -99.00 2.25 2.46 .81 1.72 3.17 1.35

1.20 .90 -99.00 -99.00 2.62 .39 -99.00 .47 7.58 2.10 -99.00 -99.00 1.79 2.54

5.24 .67 1.77 .34 1.19

.18

1.18

1.19 19.37 1.10 1.35 .68 .97 .34 6.05 .50 .35 .56 1.76 3.01 1.47 .84 -.01 1.47 20.46 3.37 .37 .68 .06 2.20 2.43 .79 1.69 3.15 1.32

1.18 .88 18.13 4.80 2.57 .38 1.83 .46 7.43 2.05 1.16 -2.24 1.76 2.48

4.80 .60 1.63 1.50

...

1.08

1.09 1.07 19.58 1.62 1.34 .62 1.42 .29 5.59 .45 .28 1.71 1.62 2.75 1.34 0.80 -.01 1.35 21.18 3.10 .33 .62 .65 2.02 2.26 .71 1.57 3.03

...

1.20 1.07 .62 17.18 5.04 2.38 .34 2.28 .41 6.84 1.85 1.64 -.85 1.62 2.24

4.24 .48 1.40 1.80 0.91

...

.95 .90 18.07 1.68 1.27 .55 1.42 .25 5.00 .39 .23 1.97 1.44 2.41 1.18 .74 -.01 1.18 19.69 2.75 .29 .55 .79 1.79 2.00 .61 1.42 2.81

...

1.05

...

.94 .55 15.56 4.75 2.13 .29 2.24 .35 6.08 1.59 1.66 -.33 1.43 1.93

3.07 .32 1.03 1.66 .64

...

.72 .63 13.66 1.47 1.00 .39 1.14

.15

3.75 .26 .16 1.83

1.05

1.72 .90 .56 -.01 .83 15.22 2.00 .19 .39 .76 1.30 1.39 .41

1.05

2.24 .73

*..

.65 .39 11.50 3.78 1.59 .18 1.81 .23 4.51 1.04 1.36 .03 1.05 1.33

1.85 .16 .58 1.22 .35

...

.44 .33 8.63 1.09 .65 .24 .78 .08 2.46 .13 .11 1.35 .64 1.02 .62 .33 -.00 .46 9.95 1.22 .10 .24 .57 .80 .78 .21 .61 1.50 .39

...

.36 .24 7.02 2.55 1.02 .09 1.16 .12 2.89 .49 .86 .10 .64 .75

.66 .04 .16 .68 .10

...

.15

.08 3.20 .58 .29 .10 .37 .02 1.16 .04 .03 .66 .23 .38 .31 .11 -.00 .13 4.66

19.79 .03

21.43 .04

.48 .03 .10 .29 .30 .56 .06 .18 .68

...

.11

...

.11 .10

2.67 18.13 .01

1.14 .44 .02 .40 .03 1.30 .12 .29 .07 .23 .25

(continued)

(7)

Am. J. Hum. Genet.58:171-181, 1996

Table 2 (continued)

ZATO =

LOCUSAND Zmax

FAMILY .00 .01 .05 .10 .20 .30 .40 TOTAL 0

F16 2.52 2.45 2.20 1.87 1.19 .49 -.03

F18 1.13 1.10 1.01 .89 .62 .33 .09

F19 -.33 -.11 .22 .34 .33 .21 .07

F21 1.51 1.47 1.35 1.18 .19 .09 .02

Total -99.00 27.59 28.57 26.40 19.24 11.79 4.44 28.57 .05

D19S215:

F1 -99.00 .24 2.57 3.17 3.07 2.32 1.20

F2 -99.00 -1.44 -.20 .19 .36 .24 .07

F3 ... ... ...

F5 -99.00 -2.97 -.47 .47 .97 .80 .35

F6 2.36 2.31 2.14 1.90 1.41 .87 .32

F8 1.25 1.55 1.81 1.74 1.33 .81 .27

F11 -99.00 -1.63 -.37 .04 .25 .19 .06

F12 -99.00 -.56 1.81 2.46 2.48 1.88 .96

F13 -99.00 -.53 .07 .23 .25 .16 .05

F14 -99.00 -1.66 -.45 -.08 .09 .07 .02

F16 -4.64 -2.28 -1.26 -.69 -.18 .02 .05

F18 .47 .45 .40 .33 .19 .08 .02

F19 ^... ^... ^... ^... ^... ^... ^...

F21 ... ... ... ... ... ...

Total -99.00 -6.51 6.02 9.76 10.22 7.44 3.38 10.45 .16

Herein, the analysis of13 additionalCADASIL families allowed us to demonstrate the genetic homogeneity ofthis condition. The responsible gene is now mapped within a 2-cM interval bracketed by D19S226 and D19S199.

Subjects and Methods Families andStatus Definition

A totalof 15 unrelated families originating from five different European countries were genotyped (fig. 1).

Twoofthem, F1 andF2, havealready been genotyped with severalchromosome 19 markers and were further investigatedwith new markers (Tournier-Lasserve et al.

1993). Thirteenadditional familieswereanalyzed. They have all been selected on the basis of thefollowingcrite- ria: (1) ahistory ofrecurrentstrokes withoutanyvascu- lar risk factors for the propositus and at least one rela- tive, (2) a leukoaraiosis with or without small, deep infarcts on brain MRI in allclinically affected patients, and (3) a pattern of inheritance consistent with ^{an au-} tosomal dominant mode. Fifty such families were re- ferred; 13 included at least six potentially informative meiosis; they were therefore considered as potentially informative andgenotyped. Two hundredtwenty-eight subjects (>18 yearsofage) from these15 familiesgave their informed consentfor thisstudy. Theywere examined by a board-certified neurologist, underwent cere-

bral MRI, and were blood drawn. As described elsewhere, MRIwasused toestablish the status for genetic linkageanalysis(Tournier-Lasserveetal. 1993).Inbrief, individuals with an abnormal MRI were considered as affected, whetheror notclinicallysymptomatic;asymp- tomaticindividuals withanormalMRIwereconsidered as healthy when :35 years of age and as having an unknown status when <35 years of age (table 1). One hundred fourteen individuals had an abnormal MRI;

among them, 71 members experienced neurological symptoms, and 43 were totally asymptomatic. Neuro- logicalsymptoms included transientischemic attacksor completed strokes(51 patients), migraine with aura(26 patients, including 15 patients suffering from both strokesand migraine withauraand11patientssuffering only from migraine with aura), dementia (22 patients with preceding strokes and 6 patients with an isolated progressive dementia), and mood disorders (8 patients with associated strokes and 3 patients with isolated mooddisorders).Onehundred fourteen individuals had anMRIshowingneitherWMA norsubcortical infarcts:

85 of them Ba35 years of age wereclassifiedashealthy, 29 of them were <35 years of age and classified as

havinganunknownstatus.Amongthe 85patientsclas- sified as healthy, 2 individuals had an MRI

showing

sequelae from largevessel cerebral infarcts due to ath- erosclerosis, and2patients sufferedfrommigrainewith aura.

176

(8)

Ducros etal.:Mappingof CADASIL

Pathological data were available insix of these families (Fl-11,F10-PM, F12-PM, F16-PM, F18-1, andF21- 3) and showedinall cases anonatherosclerotic,nonamyloid angiopathy affecting the small arteries of the cerebral whitematterand the basalganglia strikinglysimilar to previously reported lesions (Souranderand Walinder 1977; Baudrimont et al. 1993; Gray et al. 1994; Jung etal. 1995). Karyotype analysis performedinone index casefromeight families (Fl, F2, F5, F8, F10, F11, F13, and F14) wasunremarkable.

Markers

All 15 familieswereanalyzedwitheight polymorphic microsatellite markers chosen from the Genethon linkage map (D19S221, D19S226, D19S411, and D19S215) (Gyapay et al. 1994) and other published maps (D19S253, D19S244, and D19S199) (Hudson et al.

1992; Weber and May 1989). The last one, D19S841, was recently isolated in ourgroup from a cosmid previously mapped by FISH to 19pl3.1 (De Jong et al.

1989; Human Genome Center, Lawrence Livermore National Laboratory, URL http://www.bio.llnl.gov/

bbrp/genome/chrommap.html). D19S841 is a (CA)24 repeat contained within cosmid 12909 (Ducros 1995;

Genome Data Base [GDB]identification GOO-593-305).

All oligonucleotide sequences are avaible through the GDB (John Hopkins University, Baltimore). D19S253 and D19S244 are (GATA)n repeats, and the others are (CA)nrepeats.Ageneticpartial mapof chromosome 19 ispresented infigure 2.

Genotypingand Linkage Analysis

We extracted DNA from peripheral blood from all consenting membersincluding 199potentiallyinforma- tive meiosis, 45 spouses, and 29 subjects of unknown status (Miller etal. 1988). For subjects F18-1 and F21- 3,DNAwasextracted from autopsy material. Polymor- phic genomic sequences were amplified by PCR as de- scribedelsewhere (Tournier-Lasserve et al. 1993). Link- age analysis was performed using version 5.1 of the LINKAGE program package (Lathrop et al. 1984) using publishedallelesfrequencies from CEPH pedigrees. The frequency of the D19S841 alleles was determined by genotyping 70 unrelated subjects, including 28 CEPH families founders (Ducros 1995; GDB polymorphism identification G00-593-357). CADASIL gene frequency was set at .0001. Lod scores obtained with markers D19S221 and D19S226 for families F1 and F2 have beenreported elsewhere (Tournier-Lasserve et al. 1993) and were incorporated here in the combined lod score calculation. Haplotype studies were performed on all families, and the most likely haplotypes were inferred byminimizing the number of crossover events in each sibship.

To facilitate the multilocus linkage analysis, geno-

types were processed to produce a maximum offive alleles at each marker locus while preserving linkage information. The most likely position of D19S841 with respect to the others markers was established within a subset of informative CADASIL families by using the LINKMAP program. Then, the same program was run with all pedigreesto establish the best estimate of the CADASIL gene location. Because of computational limitations,twosubintervals were ana- lyzed successively. Homogeneity was examined using the admixture testfrom the HOMOGprogrampack- age (Ott 1991).

Results

GeneticHomogeneity

Two-point linkage analysis data are shown in table 2. Significant lodscores (>3) were obtained infamilies F1, F5, F8, F12, F14,andF16with severalmarkers,and positivelodscores wereobtainedinallother families. A maximum combined lod score of 37.24 was reached with marker D19S841 at 0 =.01. Admixtureanalysis of the two-pointdata with the HOMOGpackage showed evidence for linkage with homogeneity for all eight markers, with a significance level (P) of the X2 test

<.0001 for all of them. Admixture analysis of the multipoint data also supported a strong evidence for genetichomogeneity (P< .0001)and estimated thepro- portion of linked families to be 1.0 (95% confidence interval .81-1.00).

Refining of the CADASIL Interval

High-resolution genetic map of the D1 9S221- D19S215interval.-The previous orderof the polymorphic markers within the D19S221-D19S215 interval was the following: tel-D19S221-(D19S226/

D19S411)-(D19S253/D19S244)-D19S199-D19S215- cen.Theanalysis ofinheritedhaplotypes,bothhigh-and low-risk haplotypes inherited from the affected parent and haplotypes inherited from the unaffected parent, re- vealedsixcrossoverevents,which ledustorefine the position of marker D19S411 as well as to map marker D19S841 precisely (fig. 2). These six crossover events, shown in figure 3, strongly suggest that the most likely order of the markers is the following: tel-D19S221- D19S226-D19S841-(D19S253/D19S244/D19S411)- D19S199-cen.Themultipointlinkage analysis conducted in order to estimate the most likely position of the new marker D19S841 throughout the fixed map tel- D19S226-(.01)-D19S253-(.0l)-D19S199-cen showed the following maximum-likelihood order: tel-D19S226- (.009)-D19S841-(.001)-D19S253-(.01)-D19S199- cen,odds againstallalternatives being at least 10,000:1.

177

(9)

Am. J. Hum. Genet. 58:171-181,1996 F5-U F10-U Fll-U F3-1 F8-7 F11-2

_ _ _ _ _ _

_ _ _ ____ _ _

_ _ _ _ _ _

_ _ _ _ _ _ _ _ _ _ _ _

_ _ _ _ _ _

C-I

_

___I,

_ _ _ _

_ _ _ _ _

_ _ _ _ , _

___,,

^P_

_

_ . _ _ _

_ _ _ _

_,,,,_

_ _ _ _

_ _ _ _ _ _

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

_ _ _ _ _ _

hapotpesransmittedbya ulninfornativein this givenpamtandcarriedbyhis region

homologouschromosomes

Figure 3 Crossover events observed within the D19S221- D19S215 interval. Haplotype analysis ofthe 15 families revealed six

crossover events thatled torefine the position of marker D19S411, previously mappedat0 =0fromD19S226,andtoprecisely locatethe

newmarkerD19S841. Aschematic diagram of the relevantregionon

19p isshownonthe left(nottoscale).Becausenorecombinationevent wasobserved between markersD19S253, D19S244,andD19S41 1,they

arerepresentedasasingle locus.Foreachindividualinwhomhaplotype analysis revealeda crossoverevent,the involved chromosomeisrepre-

sented by a bar. Above the bar, the family and the position in the pedigree of this individual areindicated. Threecrossover eventswere

observedin individualsof unknown status(F5-U, F10-U, andF11-U)

onthe chromosomeinherited from their unaffectedparent.Thesethree subjectsareindicatedinfig. 1 byaletterU.

Genetic mapping of the CADASIL locus between D I9S226andD I9S199.-Haplotype analysisinrecom-

binant individuals established that the CADASIL locus isflankedproximally bymarker D19S199 (threerecom-

binationevents: F1-24, F11-5, andF12-17)and distally by marker D19S226 (three recombination events: F3- 1, F8-7, and F11-2). These recombination events were

observed in three affected individuals (two clinicallyaf- fected individuals having an abnormal MRI [F3-1 and F11-2] and one asymptomatic individual with anobvi- ouslyabnormal MRI[F1-24])andthreehealthy subjects with anormal MRI, aged >40years (F8-7, F11-5, and F12-17) (fig. 4).

Twoclinically asymptomatic subjects,F1-28 (27years

old) and F5-19 (40 years old), diagnosed as affected

on the basis of the MRI, have inherited the low-risk haplotype from theiraffectedparent. F1-28 is uninformative at three markers from the interval (D19S841, D19S253, and D19S411), and F5-19 is uninformative for D19S253. These two individuals could be double recombinants but most likely are in fact misclassified, since their MRI showed only one small hypersignal, whichmaybe duetoanothercause.Anotherasymptom-

aticsubject, F12-25, classifiedas unaffected,had inher-

ited the high-risk haplotype from her affected mother;

she is uninformative at D19S841 and D19S411. This individual, age42years, mayeither be^a double^recombinant or representa rare caseof late onsetofthis dis-

ease. Exceptfor thesethreeindividuals,norecombinant

event was observed between D19S841, D19S253, D19S244, D19S411,and the CADASIL locus.

Multipoint analysiswasusedtobestestimatetheposi- tion of the CADASIL gene within two intervals: tel- D19S221-(.06)-D19S226-(.02)-D19S199-cen and tel-D19S226^-(.009)-D19S841^-(.011)-D19S199-cen.

Asshowninfigure5,the bestestimate of theCADASIL

gene location is between D19S226 and D19S199. The odds against the placement beyond D19S226 and D19S199 are, respectively, 209:1 and 3,715:1.

Discussion

We previously mapped the CADASIL locus in a 12- cM interval on chromosome 19 by a linkage analysis conducted on two pedigrees. Herein, wereportgenetic linkage data obtained in 13 additional families. These data strongly suggest the genetic homogeneity of this condition and allow us to refine the mapping of the affected locusto a2-cM intervalbracketedby D19S226 and D19S199.

Recently, strong linkage to 19p was also reported in

an Italian CADASIL pedigree (Sabbadini et al. 1995).

By contrast, negative linkage results were reported on a Scottish pedigree affected by hereditary multi-infarct dementia, and the authors suggested that hereditary multi-infarct dementia,anothernameused todesignate CADASIL, may begenetically heterogeneous (St. Clair

tic

D19S221

D19S226 D19S841

D19S253/5244/S411 D19S199 _ D19S215 I-

cal

F3-1 F8-7 F11-2 F1-24 F11-5 F12-17 F1-28 F5-19 F12-25

NH

A H A ^A

H H

0

A

0

H

Figure4 Recombinationeventsleadingtotheprecise mapping oftheCADASIL locus. For each individualin whomhaplotype analysis revealedarecombination event,the chromosome inherited from theaffectedparentisrepresented byabar. The code above each bar indicatesthefamilyandthepositioninthepedigree (F3-1:individual 1 from family 3). A schematic mapof 19p (notto scale) isshown

on the left. Because no crossover event occurred between markers D19S253, D19S244, and D19S411,the latter are representedas a

singlelocus. A=affected;H=healthy; plus sign (+)=recombinant;

minus sign (-) = nonrecombinant; 0 = uninformative; and ND

= notdone.

178

tel D19S221 D19S226 D19S841 D19S253/S244/S411 D19S199 D19S215

cen

(10)

Ducrosetal.: Mapping of CADASIL 41A

40."

D19S221 I 'D19S19 cen tel D19S226 ' I 'D19S199 cen

D19S226 D19S841

Figure 5 Multipoint analysis. Multipoint lod scores for different positions of the CADASIL locus are shown with respect to two

subsets of threemarkers,onthe leftpanel:D19221-(.06)-D19S226-(.02)-D19Sl99;ontherightpanel: D19S226-(.009)-D19S841-(.011)- D19S199.The solid lines indicate the 10:1 odds (1- lodunit) intervalfortheplacementof thegene.Oddsagainstalternativesareshown for themostlikely placement of the CADASIL locusineach interval.

etal. 1995).Unfortunately, inthe latterstudy, thestatus

data used for genetic linkage analysis have not been established onthe basis ofMRIformany familymem-

bers, leadingtoahigh risk of misclassification andthere- fore false recombinants. Until this majorrequirement is fulfilled, itwill bedifficult todraw anyfirm conclusion infavororagainstlinkage^totheCADASILlocusinthis family.

These data establish the value of combined clinical and neuroimagingdiagnosiscriteriaforCADASIL. The follow-up of subjects carrying the high-risk haplotype already helpedto precise all clinical features as well as

the natural clinical and neuroimaging history of this disorderinhomogenous families (Chabriat^etal. 1995).

This is of particular importance because of the ^recent identification of this condition. In brief, most patients

>40^yearsofagesuffer ofrecurrentstrokes (84%) leading,within 10-20years,toasubcortical dementia.Mi- graine isalsoafrequentsymptomof thisdisorder(23 %) and oftenprecedes theoccurrence ofstrokes by several

years. Insome patients,the previous^symptoms maybe associated with mood disorders (21%). Aconstantfea-

ture is the presence of a leukoencephalopathy on cere-

bralMRI,anabnormality that ismostlikely the first ^to be detectable. The diagnostic marker provided by the indirectgenotypicdiagnosis will also allowthe selection ofhomogenous^groupsofpatientsfor therapeutic trials.

Indirect genetic testing is now possible in families and

may be used as a presymptomatic diagnosis test, with

respect of allethical rules inthis severecondition.

It is interesting that two other autosomal dominant diseases, FHM and HPCA, have been mapped recently

on 19pinclosevicinityto CADASIL,raisingthequestion of the allelismofthese three disorders (Jouteletal. 1993;

Kramer et al. 1995; Tean Teh et al. 1995; Vahedi ^et al

1995, Von Brederlow et al. 1995). The data reported hereinstronglysuggest thatCADASILandHPCA are not allelic disorders, since the HPCA gene resides distal to D19S226, whereas the CADASIL locusiscentromeric to this marker. With regardto FHM, it iswell established that 50% of the FHM families map ^on 19p(Joutel etal.

1994; Ophoff et al. 1994). The most likely interval is bracketedbyD19S413and D19S199 overlapping, therefore, the CADASIL interval reported here (Joutel et al.

1994). HoweverOphoffet al.(1994)reportedanaffected recombinant at D19S226 that favored the locationof the FHMgene distal to this marker. We suggest that the analysis of additional families isrequiredbefore any firmcon- clusion can be drawn for FHM.

In addition, these data will facilitate the genetic screening of otherfamilial cerebral arteriopathies sharing clinical and neuroimaging features with CADASIL and suspected to be due to the alteration ofthe same gene, such as cerebrovascular disease with thin skin, alopecia, and diskdisease,anautosomal recessive disorder initially described by Yamamura et al. (1987) or retinal cerebral arteriopathy, an autosomal dominant vasculopathyfor whichlinkage analysis with 19p markers is underway (Jen et al. 1995).

Acknowledgments

We thank all families for their participation inthisstudy.

HelpfromDr.G. Lutz (CHUSt.Antoine,Paris), Dr. S.Mail- lard (CH deSt.Dizier, St. Dizier),andDr. F. Robert (Hopital Notre-Dame, Montreal) was greatly appreciated as was the secretarial help ofR. Communal and the excellenttechnical help ofM. Cecillion, V. Domenga, and P. Lego. This work wassupported by bioMerieux,INSERM, InstitutNecker,Uni- versite Paris V, Assistance Publique des Hopitaux de Paris

179