Review 1608178

Eur J Hum Genet 1995;3:273-284

Erik Fransen3' CRASH Syndrome: Clinical Spectrum Vance Lemmon0 Guy Van Campa of Hypoplasia, Lieve Vitsa Retardation, Adducted Thumbs, Paul Couckea Patrick J. Willemsa Spastic Paraparesis and a Department of Medical Genetics, Due to Mutations in University of Antwerp, Belgium; One Single Gene, b Case Western Reserve University, Cleveland, Ohio, USA

Keywords Abstract X-linked disorder LI is a neuronal cell adhesion molecule with important func­ Mental retardation tions in the development of the nervous system. The gene Hydrocephalus encoding LI is located near the telomere of the long arm of the MASA syndrome in Xq28. We review here the evidence that Adducted thumbs several X-linked mental retardation syndromes including X- Corpus callosum agenesis linked hydrocephalus (HSAS), MASA syndrome, X-linked Spastic paraplegia complicated spastic paraparesis (SPI) and X-linked corpus LI callosum agenesis (ACC) are all due to mutations in the LI CRASH gene. The inter- and intrafamilial variability in families with Mutation analysis an LI mutation is very wide, and patients with HSAS, MASA, SPI and ACC can be present within the same family. There­ fore, we propose here to refer to this clinical syndrome with the acronym CRASH, for Corpus callosum hypoplasia, Retar­ dation, Adducted thumbs, Spastic paraplegia and Hydroceph­ alus.

Clinical Aspects for Hydrocephalus due to Stenosis of the Aqueduct of Sylvius. This designation was X-Linked Hydrocephalus based upon the presence of aqueductal steno­ X-linked hydrocephalus (MIM No. sis in many HSAS patients [1]. However, later 307000) was originally described by Bickers studies reported several HSAS patients with and Adams in 1949. It is referred to as HSAS hydrocephalus but without aqueductal steno-

Received: April 21,1995 Patrick Willems © 1995 Revision received: July 28,1995 Department of Medical Genetics S. Karger AG, Basel Accepted: August 4, 1995 University of Antwerp B-2610 Antwerp (Belgium) sis [2]. Therefore, the occurrence of aqueduc- reported, and some families have been rede­ tal stenosis in X-linked hydrocephalus might fined as representative of the MASA syn­ not be the cause of hydrocephalus. Neverthe­ drome [9-17], The clinical features show in­ less, the designation of HSAS is still used in ter- and intrafamilial variation, with mental the literature. HSAS is the most common retardation being the only obligatory symp­ genetic cause of hydrocephalus with an inci­ tom. Although MASA syndrome is an X- dence of approximately 1/30,000 male births, linked recessive condition, some expressing and it is responsible for about one quarter of carrier females have been reported, possibly male patients with hydrocephalus not associ­ due to unfavorable lyonization of the wild- ated with spina bifida [3]. The clinical spec­ type gene [9,14], Corpus callosum agenesis or trum of HSAS is very broad and shows great dysgenesis is also a frequent symptom in intra- and interfamilial variation. The only MASA patients [15]. Even though earlier obligate feature is mental retardation, with studies [9] noticed the clinical overlap be­ IQs usually ranging between 20 and 50. Hy­ tween MASA and HSAS, a clear distinction drocephalus presents from impressive macro- was made between those two conditions, cephaly, resulting in pre- or perinatal death, to based upon the presence or absence of hydro­ minimal enlargement of the ventricles. More­ cephalus in HSAS and MASA, respectively. over, HSAS families have been described in However, in later clinical studies, the sharp which some mentally retarded members did boundaries between HSAS and MASA be­ not have hydrocephalus [4, 5]. Adducted gan to fade away as (1) HSAS families were thumbs are seen in approximately 90% of the reported in which some affected members cases. Other features are spastic paraplegia, did not have hydrocephalus [4, 5], (2) some nystagmus and structural brain defects such MASA families contained HSAS patients as hypoplasia of the corpus callosum and sep­ with hydrocephalus [16], and (3) linkage stud­ tum pellucidum [5], HSAS is a recessive X- ies assigned the MASA disease locus to Xq28, linked disorder with no symptoms in carrier the same region as HSAS [16, 17]. This led us females. In 1990, we assigned the HSAS dis­ to suggest that HSAS and MASA were allelic ease locus to the distal part of the long arm of disorders due to mutations in the same gene the X chromosome in Xq28 by linkage analy­ [6]. sis [6], Further mapping studies refined the disease locus to a region of 2 Mb, distal to Spastic Paraplegia DXS52 and proximal to F8C [7, 8]. As the Spastic Paraplegia (SP) is a clinically and neural cell adhesion molecule LI (LICAM) is genetically heterogeneous condition. Some in­ located in this interval, we suggested that vestigators divide SP into pure and compli­ HSAS might be due to mutations in LI [7], cated forms [18], In pure SP, spasticity of the lower extremities is the only clinical sign. In MASA Syndrome complicated SP, additional symptoms such as In 1974, Bianchine and Lewis described an mental retardation and/or congenital brain X-linked disorder characterized by Mental re­ malformations are present. Most hereditary tardation, , .Shuffling gait and Ad­ forms of SP are inherited in an autosomal ducted thumbs [9], Referring to these four dominant or recessive way, whereas X-linked clinical features, the name ‘MASA syndrome’ SP is rare. A locus for X-linked SP (SP2) has was assigned to this disease (MIM No. been localized on the long arm of the X chro­ 303350). Several MASA families have been mosome in Xq21-q22, both in families with

274 Fransen/Lemmon/V an Camp/Vits/ CRASH Syndrome Coucke/WiUems pure [19] and complicated [20] SP. This lat­ linked recessive condition with ACC and a ter form of SP results from mutations in the variety of symptoms including macrocephaly, proteolipid protein (PLP) gene and is allelic anal anomalies, Hirschsprung’s disease, con­ to Pelizaeus-Merzbacher disease (MIM No. stipation, hypotonia and mental retardation 312080), a severe X-linked disorder in which [22], The gene responsible for FG syndrome progressive demyelination of the central ner­ has been mapped to Xql 3 [Briault, pers. com­ vous system occurs [21], Kenwrick et al. [12] mun.]. Several ACC/DCC syndromes with have described a form of X-linked compli­ unknown sublocalization on the X chromo­ cated SP (SPI, MIM No. 312900) in a family some have been described. Menkes et al. [23] with SP, mental retardation and adducted have reported a family with X-linked ACC/ thumbs. Linkage analysis assigned the SPI DCC, seizures, mental retardation and disease locus to the distal part of the long arm Hirschsprung’s disease, which might repre­ of the X chromosome in Xq28. Later clinical sent an example of FG syndrome. An X- studies of this SPI family revealed that the linked DCC family with hydrocephalus, mi­ clinical spectrum was compatible with MASA crocephaly, mental retardation and brachy- syndrome [17], suggesting that SPI and dactyly has been reported by Kang et al. [24], MASA syndrome are coallelic. Kaplan [25] reported two males with com­ plete absence of the corpus callosum and Agenesis of the Corpus Callosum severe psychomotor retardation, one of the The corpus callosum is the largest fiber patients having adducted thumbs and slightly tract in the brain, and plays a key role in the enlarged ventricles. This spectrum of anoma­ communication between the two brain hemi­ lies is reminiscent of MASA syndrome. The spheres. Corpus callosum agenesis (ACC) or fact that ACC/DCC in some families might dysgenesis (DCC) is a rather aspecific abnor­ represent MASA, is also corroborated by the mality. Many cases of DCC, and to a lesser finding that ACC or DCC have been reported extent ACC, are found incidentally at necrop­ in patients with MASA syndrome [15] and sy in individuals who had normal cerebral HSAS [5]. This suggests that there might be function. The frequency of ACC/DCC in the just one gene for HSAS, MASA, SPI and general population has been estimated at 0.3- ACC/DCC. 0.7%, whereas it is found in 2-3% of the men­ tally retarded. In most cases, ACC/DCC is an isolated finding. However, ACC and DCC Molecular Aspects can also occur as part of a syndrome, when accompanied by a number of other malforma­ Identification of LI as the tions. Many chromosomal and monogenic Disease-Causing Gene in HSAS, MASA, disorders have ACC/DCC in their spectrum SPI and ACC/DCC of anomalies. X-linked forms of ACC/DCC, The localization of the HSAS gene in Xq28 however, are very rare. , an in a 2-Mb interval containing the LI X-linked disorder that occurs only in females, (LICAM) gene suggested that LI was a candi­ has a clinical spectrum including ACC, cho­ date gene for HSAS [7], Therefore, extensive rioretinal lacunae, infantile spasms and psy­ mutation analysis was performed in the LI chomotor retardation. The gene involved in gene from a large number of HSAS families. Aicardi syndrome has provisionally been In 1992, the first LI mutation in an HSAS mapped to Xp22. FG syndrome is an X- family was reported by Rosenthal et al. [26].

275 Table 1. LI mutations in HSAS, MASA, SPI and ACC families

No. Mutation Exon Domain Disease Ref­ cDNA level protein level erence

1 321 Del 80 FS 107 frameshift 4 Ig2 HSAS 29 2 G551A R184Q missense 6 Ig2 HSAS 30 3 C630G H210Q missense 6 Ig2 M ASA/ACC 32 4 G791T C264Y missense 7 Ig3 HSAS 28 5 G1354A G452R missense 11 Ig5 HSAS 30 6 C1453T R485STOP nonsense 12 Ig5 HSAS 30 7 G1792A D598N missense 14 Ig6 MASA 32 8 2137 Ins 39 8121ns 13AA insertion intron 17 Fn2 HSAS 26 9 2884 Del G FS 961 frameshift 22 Fn4 HSAS 30 10 3489 Del TG FS 1163 frameshift 26 cytoplasmic SPI 30 11 C3581T S1194L missense 28 cytoplasmic MASA 31 12 3541 Dpi 125 FS 1223 frameshift 28 cytoplasmic HSAS 27 13 Del 3541-end FS1181 frameshift 26-28 cytoplasmic MASA 32

This was a novel mutation in a branch point LI Protein and Function signal of an intron, leading to aberrant splic­ LI, also referred to as LI CAM, is a neuro­ ing. Soon thereafter, we discovered a second nal cell adhesion molecule belonging to the LI mutation [27], a duplication in the original superfamily of the immunoglobulins [33, 34], HSAS family, which was used to localize the It is a transmembrane protein, expressed on HSAS gene [5]. This confirmed that muta­ the outgrowing axon of postmitotic nerve cells tions in the LI gene are responsible for HSAS. both in the central and peripheral nervous Since then, we and others have found many system and on Schwann cells. LI is implicated LI mutations in HSAS families [28-30; Ken- in the pathfinding of an outgrowing axon and wrick and Willems, unpubl. results]. in neural cell migration during the develop­ As MASA syndrome was thought to be alle­ ment of the nerve system, as it plays a role in lic with HSAS, the LI gene of MASA patients the adhesion between neurons, stimulates was then analyzed, and several LI mutations neurite outgrowth and fasciculation, and me­ were found [30-32], LI mutations were also diates the interaction between neurons and found in the original complicated SPI family Schwann cells during meylination [35-38], LI [30] which was later redefined as a MASA is also important in the regeneration of dam­ family [12], and in an ACC/DCC family with aged nerve tissue [39], MASA features [15, 32], This proves that The protein has a molecular mass of HSAS, MASA, SPI and ACC/DCC are allelic 200 kD with 138 kD due to polypeptide and disorders due to mutations in a single gene, LI the rest to carbohydrate [40], It is composed (table 1). of various domains including (1) a cytoplas­ mic domain which is highly conserved among

276 Fransen/Lemmon/Van Camp/Vits/ CRASH Syndrome Coucke/Willems E ■o

_Q E E

6 Immunoglobulin domains 5 Fibronectin domains £ Cytoplasmic domain

NH3 COOH

Fig. 1. Schematic representation of the LI molecule. The LI protein is represented with its different domains: six immunoglobulin domains (loops), five fibronectin domains (filled ovals), the transmembrane and the cytoplasmic domain.

various species, (2) a transmembrane do­ binding [42] (fig. 2a). The activated FGFR main, composed of 23 hydrophobic residues, then phosphorylâtes phospholipase C, which (3) five fibronectin-III-like domains, and (4) triggers a cascade of events producing arachi- six immunoglobulin-type C2-like domains, donic acid. The elevated concentration of ara- each containing one disulfide bridge (fig. 1, chidonic acid causes a Ca2+ influx into the cell 2). The extracellular part of LI interacts with through N- and L-type Ca2+ channels. The its surroundings through homophilic binding same cascade is activated by soluble forms of with other LI molecules [41] and through the LI molecule, suggesting that cell adhesion heterophilic interactions with other neural as such is not required to activate this signal cell adhesion molecules such as axo- transduction pathway [43], Four different nin-l/TAG-1, F3/F11 and phosphocan. The FGFR genes have been identified. All have a cytoplasmic tail of LI interacts with ankyrin, motif of five consecutive amino acids homol­ a protein involved in the binding of the cyto- ogous to a protein sequence located between skeleton to the cell membrane. It is unclear the third and fourth immunoglobulin do­ whether the cytoplasmic domain of LI only mains of LI. In the same FGFR domain ami­ offers an anchor to the cytoskeleton, or is also no acid stretches are found which show ho­ part of a secondary signal transduction sys­ mology to N-CAM and members of the tem (fig. 2a). The different effects of LI are cadherin family. This domain is now referred not only mediated by cell adhesion, but are to as the FGFR CAM-homology domain also the consequence of intracellular changes (FGFR-CHD). brought about by binding of LI to the fibro­ blast growth factor receptor (FGFR). Recent­ The LI Gene ly, evidence has been provided that LI acti­ The LI gene spans 16 kb [Rosenthal, pers. vates the FGFR tyrosine kinase upon ligand commun.] and is composed of 28 exons. The

277 1 immunoglobulin domain

2 fibronectin domain

3 transmembrane domain

4 cytoplasmatic domain 5 putative signal transduction system 6 CAM homology domain (CHD) 7 tyrosine kinase domain 1 8 tyrosine kinase domain 2 9 cell membrane

10 intercellular space 11 intracellular space

FGFR a

cTa pp pq

if

b c

278 Fransen/Lemmon/V an Camp/Vits/ CRASH Syndrome Coucke/Willems Fig. 2. Hypothetical models for the disturbance of open reading frame has 3,825 bp, and encodes normal LI function by different mutations, indicated a protein of 1,275 amino acids. The first 19 by an asterisk, a Normal functioning of LI. Homo- amino acids form a signal peptide that is philic interaction between LI molecules on neighbor­ ing neurons. The fibroblast growth factor receptor cleaved from the native protein, which is (FGFR) interacts with LI via its CAM homology 1,256 amino acids long. The exons 3-14 en­ domain and dimerizes upon activation. This leads to code the 6 immunoglobulin domains with two activation of the two tyrosine kinase domains, b A mis- exons for each domain (fig. 1). Exons 15-24 sense mutation in the extracellular part of LI. The encode the 5 fibronectin domains, exon 25 the homophilic interaction between LI molecules is weak­ ened or absent, leading to absent or abnormal intracel­ transmembrane domain and exons 26-28 lular transduction, c A mutation in the cytoplasmic form the cytoplasmatic domain. An alterna­ domain. Adhesion between neurons, and between LI tive splicing variant lacking four amino acids and the FGFR remains unaffected as the extracellular in the cytoplasmatic domain has been ob­ domains are intact. However, the cytoplasmic domain served in the peripheral nerve system [44] but is truncated or absent. This might lead to absent or abnormal intracellular transduction, or impaired inter­ its functional importance is unknown. Struc­ action with the cytoskeleton. d A nonsense mutation in tural and functional homologues of LI have one of the immunoglobulin domains leading to a pre­ been found in mouse (LI), rat (NILE), chick mature extracellular stop codon. As the truncated pro­ (G4/8D9) and Drosophila (neuroglian). The tein lacks some immunoglobulin domains, the fibro- cytoplasmic domain is by far the most con­ nectin domains, the transmembrane segment and the intracellular domain, it is secreted into the extracellu­ served, sharing 100% homology at the amino lar space. There it might be rapidly degraded, it might acid level between human and mouse [34]. be stable but not functional, or it might block normal interaction of other cell adhesion molecules with LI Mutations immunoglobulin domains. All mutations found to date are private mutations that have never been observed in more than one family. The mutations are dispersed throughout the whole LI gene, and include deletions, a duplication, missense, nonsense, splice site mutations, and a novel type of mutation in a branch point signal leading to aberrant splicing. An overview of lr 4 all LI mutations reported so far in HSAS, d MASA, SPI and ACC/DCC is given in ta­ ble 1. a The mutations can be classified into three d d broad categories. The first class consists of mis­ Q P tr sense point mutations in the extracellular do­ main (mutations 2-5, 7). In these cases, the ■Q sP mutant protein is still an integral membrane protein with a normal transmembrane and cytoplasmic domain. However, the binding ca­ pacities of the LI protein towards its ligands or its signaling pathways might be distorted, as the structure of the extracellular domains has d been altered at one amino acid position (fig.

279 2b). The second class of mutations has a dis­ tions with the cytoskeleton, has been observed rupted cytoplasmic domain, either by missense in a family with a mild MASA phenotype [14, mutation (mutation 11), frameshift (mutation 32]. In contrast, missense mutations can lead 10) or large rearrangements such as a duplica­ to a very severe clinical picture [28, 31]. It is tion (mutation 12) or a deletion (mutation 13). also unlikely that a good genotype-phenotype In all of these cases, the extracellular and trans­ correlation exists, because of the large intra­ membrane domains remain intact (fig. 2c). familial clinical variability with patients with The influence of these mutations on the func­ HSAS, MASA, SPI and ACC/DCC being tioning of LI is unclear. It is possible that present within the same family. Schrander- mutations in the cytoplasmic domain disturb Stumpel et al. [pers. commun.] have reviewed the transduction of extracellular signals into the clinical characteristics of many families the cell via an unknown signal pathway. Alter­ with HSAS, MASA, SPI and ACC/DCC, and natively, it is possible that these mutations lead found that most families have a mixed pheno­ to impaired binding with the intracellular cy- type with large intrafamilial variability. Most toskeleton. The third class of mutant LI pro­ families have been referred to as HSAS, teins has a premature stop codon in the extra­ MASA, SPI or ACC/DCC, depending on the cellular part, due to a nonsense mutation (mu­ clinical phenotype of the proband. In many tation 6) or a frameshift mutation (mutations 1 families, however, affected relatives had dif­ and 9). The mutant protein could be secreted ferent phenotypes. into the extracellular space as no transmem­ LI is not the only gene showing pleiotropic brane domain is present (fig. 2d). Obviously, manifestations. More than 100 genes are such a protein loses its normal function as con­ known in which different mutations cause tact with the nerve cell is lost. It is intriguing, more than one clinical disorder. One of the however, that truncated soluble forms of LI most interesting examples is the RET protoon­ still seem to be able to activate FGFRs and cogene. RET is a transmembrane protein, with stimulate neurite outgrowth in vitro, although a cadherin-like extracellular domain, and a cys­ they miss the transmembrane and cytoplasmic teine-rich extracellular region next to the trans­ domain [43]. This might suggest that LI has membrane region. Although RET mutations essential functions not mediated through the cause four different clinical disorders, there is a FGFR signal transduction pathway. It is, how­ good genotype-phenotype correlation. Inacti­ ever, also possible that the naturally occurring vating mutations in several different domains truncated LI proteins in these patients are of RET cause Hirschsprung’s disease, a devel­ unstable and degraded quickly. opmental disorder of the enteric nervous sys­ tem. Missense mutations of specific cysteine residues in the cysteine-rich region cause either Genotype-Phenotype Correlation multiple endocrine neoplasia type 2A and 2B, or medullary thyroid carcinoma through a Mutations in LI cause a complex and vari­ dominant-negative oncogenic mechanism able spectrum of clinical abnormalities. How­ [45]. In contrast, the different LI mutations do ever, no obvious relationship has been found not cause a constant phenotype. The compli­ between the disease phenotype and the type cated signal transduction pathway, of which L1 or position of the LI mutation. Deletion of a is one of the many components, might be large part of the cytoplasmatic domain, which redundant and allow for variation in the ex­ would be likely to disrupt signaling or interac­ pression of a loss-of-function mutation in LI

280 Fransen/Lemmon/Van Camp/Vits/ CRASH Syndrome Coucke/Willems by the varying influence of other components. drocephalus is present and the diagnosis On the other hand, gain-of-function mutations CRASH is suspected, it is highly probable that might also have a variable clinical picture. an LI mutation will be present. However, Understanding the clinical variation encoun­ Strain et al. [48] have postulated the existence tered in CRASH syndrome might increase of a second locus for CRASH, approximately when we know more about the functioning of 10 cM proximal from the LI gene. The link­ LI. The interactions of LI with FGFR open age analysis upon which the authors based another exciting field to be investigated. Re­ their findings, however, was remarkable, in cently, several mutations in the genes encod­ that it showed an impressive number of inter­ ing FGFR 1 and 2 have been discovered, each marker recombinants over a short distance. of them associated with different craniosynos- Subsequently, a LI mutation was identified in tosis syndromes including Apert, Crouzon, this family, excluding the existence of the Jackson-Weiss and Pfeiffer syndromes. Muta­ hypothesized second locus [Brock, pers. com­ tions in FGFR3, on the other hand, give rise to mun.]. In our own experience of more than 50 achondroplasia [46]. None of the latter syn­ HSAS or MASA families with multiple af­ dromes show clinical overlap with CRASH fected cases, we have not found evidence for syndrome, and clinical features such as cra- linkage heterogeneity. Only one small Ger­ niosynostosis or dwarfism encountered in man family was found in which a presumably cases of an FGFR mutation have not been healthy normal male had the same Xq28 hap­ observed in families with an LI mutation. lotype as his brother and a cousin who were This is not surprising, as until now LI muta­ both affected with HSAS [7], However, the tion analysis was only performed in families normal male refused brain imaging or IQ test­ with CRASH features. It is possible that LI is ing, so that it could not be excluded that he a candidate gene for Xq28-linked diseases was mildly affected. In some cases of the with bone malformations. In this respect, it CRASH syndrome, the typical facial features would be interesting to look for mutations in of including elongated patients with oto-palato-digital syndrome, a face and large everted cars are present [5], bone disorder gene located in Xq28. X-linked spasticity and mental retardation (SPI) due to mutations in LI have to be dif­ ferentiated from SP2 [19,20]. This can be dif­ Differential Diagnosis ficult when SPI is not associated with other symptoms from the CRASH clinical spec­ In view of the large clinical spectrum of the trum (table 2). However, mental retardation CRASH syndrome, the clinical diagnosis can is always moderate to severe in the CRASH be difficult, especially when hydrocephalus syndrome and mild in SP2. The gene respon­ and adducted thumbs are absent. Nearly all sible for SP2 is PLP, located in Xq22. The patients have moderate to severe mental re­ PLP genes has two gene products, produced tardation, but it is still unclear how frequently by alternative splicing: DM20, involved in LI mutations will be found in nonspecific oligodendrocyte maturation, and PLP itself, Xq28-linked mental retardation, but without involved in myelin sheet compaction. Muta­ hydrocephalus, spastic paraparesis or ad­ tions affecting both gene products lead to Pe­ ducted thumbs [47]. It is possible, therefore, lizaeus-Merzbacher disease, whereas muta­ that patients with LI defects are not recog­ tions in PLP but spliced out in DM20 lead to nised as having CRASH syndrome. When hy­ SP2 [21],

281 Table 2. Clinical spectrum of diseases associated with LI and LI mutations PLP mutations PLP mutations HSAS MAS A SPI PMD SP2

Mental retardation ++ ++ ++ ± ± Hydrocephalus ++ + Motor deterioration ++ + Adducted thumbs + + ± Chorea ++ ± Cerebellar ataxia ++ ± Nystagmus ++ ± Hypotonia ++ Demyelination on MRI ++ ±

MRI = Magnetic resonance imaging.

Prenatal Diagnosis itself to be able to perform prenatal diagnosis and carrier detection. In large families with HSAS, MASA, SPI or ACC/DCC linked to the Xq28 region, DNA diagnosis based upon DNA analysis is Conclusion possible as the Xq28 region contains many highly polymorphic markers. RFLP analysis Mutations in the neural cell adhesion mol­ can be performed using probe F814, which is ecule LI, located in Xq28, can lead to a wide a hybrid probe recognizing both a VNTR clinical spectrum of neurologic anomalies in­ marker DXS52 and a two-allellic RFLP in cluding corpus callosum hypoplasia, mental F8C. DXS52 is approximately 1-2 cM proxi­ retardation, adducted thumbs, spastic para­ mal to LI, whereas F8C is 1-2 cM distal to paresis and hydrocephalus. We therefore pro­ LI. LI itself is not very polymorphic as only pose to refer to this clinical spectrum as the one RFLP with a very low polymorphism CRASH syndrome. information content has been found in LI [49]. Nowadays, polymorphic microsatellites, amplified by PCR, are preferred, because this Acknowledgements speeds up the analysis and requires a smaller amount of DNA, whereas microsatellites are The research reviewed here was supported in part by the NFWO (National Fund for Scientific Research), much more informative than RFLP markers. the AFM (Association Française contre les Myopa­ In somes cases, a reliable diagnosis with flank­ thies) and a concerted action from the University of ing markers is not possible, as (1) the distal Antwerp. E.F. holds a scholarship from the IWT (Insti­ Xq28 markers are not always informative, tute for the Support of Research in Science and Tech­ (2) not enough DNA samples are available nology). Y.L. received support from the NEI (National Eye Institute) and NINDS (National Institute of Neu­ and (3) there is a de novo mutation. In such rological Disorders and Stroke). cases, it is necessary to identify the mutation

282 Fransen/Lemmon/Van Camp/Vits/ CRASH Syndrome Coucke/Willems References

1 Bickers DS, Adams RD: Hereditary 12 Kenwrick S, Ionanescu V, Ionanes­ 22 Opitz JM, Richieri-da Costa A, Aase stenosis of the aqueduct of Sylvius cu G, Searby C, King A, Dubowitz JM, Benke PJ: FG syndrome update as a cause of congenital hydrocepha­ M, Davies KE: Linkage studies of X- 1988: Note of 5 new patients and lus. Brain 1949;72:246-262. linked recessive spastic paraplegia bibliography. Am J Hum Genet 2 Landrieu P, Ninane J, Fernere G, using DNA probes. Hum Genet 1988;30:309-328. Lyon G: Aqueductal stenosis in X- 1986;73:264-266. 23 Menkes JH, Philippart M, Clark linked hydrocephalus: A secondary 13 Straussberg R, Blatt I, Brand N, DB: Hereditary partial agenesis of phenomenon? Dev Med Child Neu­ Kessler D, Goodman RM: X-linked the corpus callosum. Arch Neurol rol 1979;21:637-652. mental retardation with bilateral 1964;11:198-208. 3 Halliday J, Chow CW, Wallace D, clasped thumbs: Report of another 24 Kang WM, Huang CC, Lin SJ: X- Dansk DM: X-linked hydrocepha­ affected family. Clin Genet 1991,40: linked inheritance of dysgenesis of lus: A survey of a 20 year period in 337-341. corpus callosum in a Chinese family. Victoria, Australia. J Med Genet 14 Macias VR, Day DW, King TE, Wil­ Am J Med Genet 1992;44:619-623. 1986;23:23-31. son GN: Clasped thumb-mental re­ 25 Kaplan P: X-linked recessive inheri­ 4 Fried K: X-linked mental retarda­ tardation (MASA) syndrome: Con­ tance of agenesis of the corpus callo­ tion and/or hydrocephalus. Clin firmation of linkage to Xq28. Ge­ sum. J Med Genet 1983;20:122- Genet 1972;3:258-263. netics 1992;43:408-414. 124. 5 Willems PJ, Brouwer OF, Dijkstra I, 15 Boyd E, Schwartz CE, Schroer RJ, 26 Rosenthal A, Jouet M, Kenwrick S: Wilmink J : X-linked hydrocephalus. May MM, Shapiro SD, Arena F, Aberrant splicing of neural cell ad­ Am J Med Genet 1987;27:921-928. Lubs HA, Stevenson RE: Agenesis hesion molecule LI mRNA in a 6 Willems PJ, Dijkstra I, Van der Au- of the corpus callosum associated family with X-linked hydrocepha­ wera BJ, Vits L, Coucke P, Raey- with MASA syndrome. Clin Dys- lus. Nat Genet 1992;2:107-112. makers P, Van Broeckhoven C, morphol 1993;2:332-341. 27 Van Camp G, Vits L, Coucke P, Consalez GG, Freeman SB, Warren 16 Schrander-Stumpel C, Legius E, Lyonnet S, Schrander-Stumpel C, ST, Brouwer OF, Brunner HG, Re­ Fryns JP, Cassiman JJ: MASA syn­ Darby J, Holden J, Munnich A, Wil­ nier WO, Van Elsen AF, Dumon JE: drome: New clinical features and lems P: A duplication in the L1CAM Assignment of X-linked hydroceph­ linkage analysis using DNA probes. gene associated with X-linked hy­ alus to Xq28 by linkage analysis. J Med Genet 1990;27:688-692. drocephalus. Nat Genet 1993;4: Genomics 1990;8:367-370. 17 Winter RM, Davies KE, Bell MV, 421-425. 7 Willems PJ, Vits L, Raymakers P, Huson SM, Patterson MN: MASA 28 Jouet M, Rosenthal A, Macfarlane J, Beuten J, Coucke C, Holden JJ, Van syndrome: Further clinical delinea­ Kenwrick S, Donnai D: A missense Broeckhoven C, Warren ST, Sagi M, tion and chromosomal localisation. mutation confirms the LI defect in Robinson D, Dennis N, Friedman Hum Genet 1989;83:367-370. X-linked hydrocephalus. Nat Genet KS, Magnay D, Lyonnet S, White 18 Harding AE: Classification of the 1993;4:331. BN, Wittwer BH, Aylsworth AS, hereditary ataxias and paraplegias. 29 Coucke P, Vits L, Van Camp G, Ser­ Reicke S: Further localization of X- Lancet 1983;i:l 151-1155. vine F, Lyonnet S, Kenwrick S, Ro­ linked hydrocephalus in the chro­ 19 Keppen LD, Leppert MF, O’Connel senthal A, Wehnert M, Munnich A, mosomal region Xq28. Am J Hum PO, Nakamura D, Stauffer M, La- Willems PJ: Identification of a Genet 1992;51:307-315. throp M, Lalouel JM, White R: Etio­ 5'splice site mutation in intron 4 of 8 Lyonnet S, Pelet A, Briard ML, Ser­ logical heterogeneity in X-linked the LI CAM gene in an X-linked hy­ vine F, Le Maree B, Pfeiffer RA, spastic paraplegia. Am J Hum Ge­ drocephalus family. Hum Mol Ge­ Hors MC: Further linkage data at net 1987;41:933-943. net 1994;3:671-673. the X-linked hydrocephalus locus 20 Bonneau D, Rozet JM, Bulteau C, 30 Jouet M, Rosenthal A, Armstrong with seven polymorphic DNA Berhier M, Metthey R, Munnich A, G, Macfarlane J, Stevenson R, Pa­ markers of Xq28. Am J Med Genet Le Mener M: X linked spastic para­ terson J, Matzenberg A, Ionanescu 1991 ;49(suppl):Al 959. plegia (SPG2): Clinical heterogene­ V, Temple K, Kenwrick S: X-linked 9 Bianchine JW, Lewis RC: The ity at a single gene locus. J Med Ge­ spastic paraplegia (SPG1), MASA MASA syndrome: A new hereditary net 1993;30:381-384. syndrome and X-linked hydroceph­ mental retardation syndrome. Clin 21 Saugier-Veber P, Munnich A, Bon­ alus result from mutations in the LI Genet 1974;5:298-306. neau D, Rozet JM, Le Mener M, Gil gene. Nat Genet 1994;7:402-407. 10 Gareis FJ, Mason MD: X-linked R, Boesplug-Tanguy O: X-linked 31 Fransen E, Schrander-Stumpel C, mental retardation associated with spastic paraplegia and Pelizaeus- Vits L, Coucke P, Van Camp G, bilateral clasp thumb anomaly. Am Merzbacher disease are allelic disor­ Willems PJ: X-linked hydrocepha­ J Med Genet 1984;17:333-338. ders at the proteolipid locus. Nat lus and MASA syndrome present in 11 Yeatman GW: Mental retardation- Genet 1994;6:257-262. one family are due to a single mis­ clasped thumb syndrome. Am J sense mutation in exon 28 of the Med Genet 1984;17:339-344. LI CAM gene. Hum Mol Genet 1994;3:2255-2256.

283 32 Vits L, Van Camp G, Coucke P, 38 Williams EJ, Doherty P, Turner G, 43 Doherty P, Williams E, Walsh FS: A Wilson G, Schrander-Stumpel C, Reid RA, Hemperly JJ, Walsh FS: soluble chimeric form of LI glyco­ Schwartz C, Willems P: MASA syn­ Calcium influx into neurons can protein stimulates neurite out­ drome is due to mutations in the solely account for cell contact-de­ growth. Neuron 1995;14:57-66. L1CAM gene. Nat Genet 1994;7: pendent neurite outgrowth stimu­ 44 Reid RA, Hemperly JJ: Variants of 408-413. lated by transfected LI. J Cell Biol human LI-cell adhesion molecule 33 Moos M, Tacke R, Scherer H, Te- 1992;119:883-892. arise through alternate splicing of plow D, Früh K, Schachner M: Neu­ 39 Persohn E, Schachner M: Immu- RNA. J Mol Neurosci 1992;3:127— ral adhesion molecule LI as a mem­ noelectron microscopic localization 135. ber of the immunoglobulin super­ of the neural cell adhesion molecules 45 Romeo G, McKusick VA: Pheno­ family with binding domain similar LI and N-CAM during postnatal de­ typic diversity, allelic series and to fibronectin. Nature 1988;334: velopment of the mouse cerebellum. modified genes. Nat Genet 1994,7: 701-702. J Cell Biol 1987;105:569-576. 451-453. 34 Hlavin ML, Lemmon V: Molecular 40 Rathjen FG, Schachner M: Immu- 46 Mulvihill JJ: Craniofacial syn­ structure and functional testing of nocytological and biochemical char­ dromes: No such thing as a single human L1CAM: an interspecies acterization of a new neuronal cell gene disease. Nature Genet 1995;9: comparison. Genomics 1991 ; 11 : surface component (LI antigen) 101-103. 416-423. which is involved in cell adhesion. 47 Nordstrom AM, Penttinen M, von 35 Lagenaur C, Lemmon V: An Ll-like EMBOJ 1984;3:1-10. Koskull H: Linkage to Xq28 in a molecule, the 8D9 antigen, is a po­ 41 Lemmon V, Farr KL, Lagenaur C: family with nonspecific X-linked tent substrate for neurite extension. LI-mediated axon outgrowth occurs mental retardation. Hum Genet Proc Natl Acad Sci USA 1987;84: via a homophilic binding mecha­ 1992;90:263-266. 7753-7757. nism. Neuron 1989;2:1597-1603. 48 Strain L, Gosden CM, Brock DJH, 36 Wood PM, Schachner M, Bunge RP: 42 Williams EJ, Furness J, Walsh FS, Bonthron DT: Genetic heterogene­ Inhibition of Schwann cell myelina- Doherty P: Activation of the FGF ity in X-linked hydrocephalus: Link­ tion in vitro by antibody to the LI receptor underlies neurite out­ age to markers within Xq27.3. Am J adhesion molecule. J Neurosci growth stimulated by LI, N-CAM Hum Genet 1994;54:236-243. 1990;10:3636-3645. and n-cadherin. Neuron 1994; 13: 49 Willems PJ, Vits L, De Boulle K: 37 Asou H, Miura M, Kobayashi M, 583-594. Mspi RFLP in the LI CAM gene in Uyemura K, Itoh K: Cell adhesion Xq28. Nucleic Acids Res 1991;19: molecule LI guides cell migration in 5448. primary reaggregation culture of mouse cerebellar cells. Neurosci Lett 1992;144:221-224.

284 Fransen/Lemmon/Van Camp/Vits/ CRASH Syndrome Coucke/Willems