|Home | About | Journals | Submit | Contact Us | Français|
Cohen syndrome (CS) is an autosomal recessive disorder with variability in the clinical manifestations, characterised by mental retardation, postnatal microcephaly, facial dysmorphism, pigmentary retinopathy, myopia, and intermittent neutropenia. Mutations in the gene COH1 have been found in an ethnically diverse series of patients. Brief clinical descriptions of 24 patients with CS are provided. The patients were from 16 families of different ethnic backgrounds and between 2.5 and 60 years of age at assessment. DNA samples from all patients were analysed for mutations in COH1 by direct sequencing. Splice site mutations were characterised using reverse transcriptase PCR analysis from total RNA samples. In this series, we detected 25 different COH1 mutations; 19 of these were novel, including 9 nonsense mutations, 8 frameshift mutations, 4 verified splice site mutations, 3 larger in frame deletions, and 1 missense mutation. We observed marked variability of developmental and growth parameters. The typical facial gestalt was seen in 23/24 patients. Early onset progressive myopia was present in all the patients older than 5 years. Widespread pigmentary retinopathy was found in 12/14 patients assessed over 5 years of age. We present evidence for extended allelic heterogeneity of CS, with the vast majority of mutations leading to premature termination codons in COH1. Our data confirm the broad clinical spectrum of CS with some patients lacking even the characteristic facial gestalt and pigmentary retinopathy at school age.
Cohen syndrome (CS; MIM #216550) is an uncommon autosomal recessive disorder with variability in the clinical manifestations including psychomotor retardation, microcephaly, typical facial dysmorphism, progressive pigmentary retinopathy appearing at mid‐childhood, early onset and severe myopia, and intermittent neutropenia.1,2,3,4 The characteristic craniofacial appearance includes downwards slanting and wave shaped palpebral fissures, short philtrum, heavy eyebrows, thick hair, and prominent nasal base. Finnish patients present with a homogeneous phenotype as result of a specific allele. However, a broad clinical spectrum in non‐Finnish cases and the age dependent appearance of some clinical signs may contribute to a delay in making the diagnosis.
Recently, the CS phenotype was found to be associated with mutations in the gene COH1 in different populations.5,6,7,8COH1 maps to chromosome 8q22 and codes for various splice forms. It comprises up to 62 exons and its longest transcript is 14093 nucleotides in length with an open reading frame of 4022 codons.5COH1 encodes a potential transmembrane protein presumably involved in vesicle mediated sorting and intracellular protein transport.5,9,10 The mouse homologue of COH1 is widely expressed in neurones of the postnatal and adult brain suggesting a role in neuronal differentiation.8 This suggests that COH1 primarily has a function in postmitotic cells, which may be the reason for the postnatal microcephaly seen in CS.8
Overall, more than 50 COH1 mutations have been reported in association with CS. Most are termination mutations and predicted to result in a null allele, while missense mutations and larger deletions are less common.
To determine the variability of the clinical manifestations and the spectrum of mutations associated with this condition we investigated 24 patients with CS from 16 families, with a wide range of geographical and ethnic origins.
This study included 24 patients from 16 families (17 male and 7 female patients, ranging in age from 2.5 to 60 years). The diagnosis of CS was based on clinical manifestations including developmental delay/mental retardation and at least two of three criteria: typical facial gestalt (fig 11),), pigmentary retinopathy or myopia, and neutropenia (table 11).). There were 12 non‐consanguineous families, comprising three French families (one with two affected children and two with one affected child each), three German families (two with one affected child and one with five affected male members), one Turkish family (one affected child), one Belgian family (one affected child), one family of English and German origin (one affected child), one Canadian family from English/Scottish descent (one affected child), and two Polish families (with one and two affected children, respectively). The four consanguineous families reported were all Turkish (two with affected sibling pairs and two with one affected child each).
Clinical assessments were carried out on all patients by at least one of us. Blood samples were collected from the parents and from affected and unaffected siblings. Signed informed consents were obtained from the parents of the affected and unaffected children.
Genomic DNA was isolated from peripheral blood using standard techniques. For mutation screening, 62 exons and five potential alternative exons including their flanking splice sites, and the predicted promoter region of COH1 were analysed. Primer pairs for genomic amplification were generated on the basis of the sequence of a chromosome 8 genomic contig (GenBank accession number NT_008046.15).7 All primer sequences are available on request. PCR products were purified using exonuclease I and shrimp alkaline phosphatase, and then sequenced directly with a commercial kit (BigDye Terminator cycle sequencing kit, version 1.1, Applied Biosystems) and analysed on an automated DNA analyser (model 3730; Applied Biosystems).
Total RNA was extracted from peripheral blood, and reverse transcription (RT) perfomed using commercial kits (PAXgene blood RNA validation kit, PreAnalytiX and Omniscript RT kit, Qiagen). Primer pairs for cDNA amplification were designed from the sequence of the full length COH1 transcript (GenBank accession number NM_017890.3). Analysis of PCR products was performed by purifying single bands from agarose gel using a gel extraction kit (QIAquick; Qiagen), and sequenced directly as described above.
In our series of 24 patients with CS from 16 families, we observed marked variability of developmental and growth parameters (table 11).). At the time of assessment, short stature (height at or below the third centile) was present in 16 of 24 cases. Postnatal microcephaly was found in 20 of 24 patients. Truncal obesity was documented in the majority of patients analysed (17/24). Mental retardation and intellectual impairment varied from moderate to severe. All but one of the patients (patient 2‐2) learned to walk independently. The majority of patients achieved verbal communication skills and most used short sentences at mid‐childhood. However, two adult patients (2‐2 and 2‐3) and two 5 year old patients (14‐1 and 15‐2) were non‐verbal. Stereotypic movements were seen in patients 5 and 15‐2.
Typical craniofacial features were seen in 23/24 patients (fig 11).). Patient 6, who presented with everted lips, bulbous nasal tip, and normal appearing philtrum, lacked the facial gestalt characteristic of CS (fig 22).). The affected siblings of family 2 were 44 and 60 years old at assessment and both had the characteristic facial features of CS including short philtrum and downwards slanting palpebral fissure.
Progressive visual disability was a consistent finding. An early onset of progressive myopia with refractive errors ranging from −0.3 to −8 dioptres was evident in all the patients older than 5 years. Patient 8‐2 was too young to assess this clinical sign. Retinal examination had been performed on all study patients. All but two of the patients older than 5 years had widespread pigmentary retinopathy with vessel attenuation and optic pallor. Two siblings (13‐1 and 13‐2) had normal retinal examinations at the ages of 16.5 and 14.5 years. Electrodiagnostic tests have not been carried out on these two patients but functional impairment such as nyctalopia was not recorded. Six patients between 3.5 and 5 years of age had normal retinas on ophthalmological examination. The adult patients assessed by us showed progressive ophthalmological changes. Severe retinochoroidal atrophy, generalised bone spiculae‐like pigment accumulations, cataracts, and optic nerve pallor were found in two affected brothers aged 53 and 44 years (2‐4 and 2‐5, respectively). The ophthalmological abnormalities led to total blindness in mid‐adulthood in patients 2‐1 and 2‐5.
Narrow hands and feet were found in all patients. Progressive kyphoscoliosis developed after the age of 40 years in patients 2‐4 and 2‐5 but was also present in an affected teenager (patient 6). Neutropenia defined as a neutrophil count <1.5×109/l was shown in the majority of patients analysed for that clinical sign (14/16 patients).
Hypogonadotropic hypogonadism, severe and complete growth hormone deficiency, insulin resistance, and compensatory hyperinsulinemia were found in patient 10.
We analysed all 24 patients by direct genomic sequencing of the whole coding region including the exon splice sites and the promoter region. In the full length transcript of COH1, we identified 25 different mutations; 19 of these were novel, including 9 nonsense mutations, 8 frameshift mutations, 4 verified splice site mutations, 3 larger in frame deletions, and 1 missense mutation identified in two unrelated families. The mutations identified are summarised in table 22.. Each mutation cosegregated in the respective family in all cases and was not seen in 150 chromosomes from control subjects.
Heterozygous mutation c.1563G→A identified in family 1 affected the last base pair of exon 11 (table 22).). To analyse whether the mutation caused an altered splicing of COH1, we performed RT‐PCR on a total RNA sample from peripheral blood of patient 1. Amplified cDNA with primers located in exons 8 and 16 showed the regularly processed product in a very low level. Amplification with primers placed in exon 8 and intron 11 resulted in a mutant cDNA fragment. Sequencing demonstrated an abnormal transcript including more than 177 bp of the intron 11, resulting in a new reading frame of 19 codons followed by a stop codon in intron 11.
In family 9, we identified a heterozygous mutation, c.7322_7322+1delGGinsATGGAGC (table 22),), at the splice donor site. RNA analysis with primers located in exons 39 and 45 and in exon 39 and intron 40, respectively, showed lack of splicing with an altered codon 2441, followed by 35 novel codons and a stop codon. A heterozygous splice donor site mutation c.2934+1_2934+2delGT was identified in family 11 (table 22).). We identified complete skipping of exon 20 with primers located in exons 19 and 24. Consequently, the altered transcript lacked 110 nucleotides. Mutation c.6732+1G→A, first reported by Kohlemainen et al,6 changed the splice donor site of intron 37 (family 3; table 22).). Total RNA from the affected individual analysed with primers located in exons 35 and 39 revealed a cryptic splice site in intron 37. The defective splicing leads to an addition of four bases into the COH1 transcript. The sequence of the cryptic splice site (GTGGGT) is consistent with the consensus motif (GURAGU) at five of six positions, including the highly conserved dinucleotide GT at the beginning.
In this paper, we report the clinical and molecular data on 24 individuals with CS, from 16 families, caused by mutations in COH1. Overall, 11 compound heterozygous cases, 4 homozygous cases with known consanguinity, and 1 homozygous case with unknown consanguinity were observed. In one patient (family 10), only one heterozygous mutation was detected. This patient was also assumed to be compound heterozygous. The lack of a second pathogenic mutation points to the existence of further alternative exons and/or other transcripts of COH1.5,7,11 Alternatively, further mutations could reside within intronic sequences, which were analysed only at the conserved splice sites, or could represent larger deletions.
In total, we identified 25 different mutations, of which 19 were novel (table 22).). Most of the patients described here carried mutations resulting in premature termination codons. This is consistent with previously reported cases. In total, 61 of the 73 mutations in COH1 published to date resulted in premature termination through frameshift or nonsense mutations,5,6,7,8,12,13 and six mutations are larger in frame deletions. Thus, the pathogenic implications of missense mutations associated with CS (6 of 73), identified only in two of our cases, have still to be clarified. Several coding SNPs found indicate that the COH1 product appears to have some tolerance to such variations (table 33).
Nearly all mutations except for those found in Finnish patients5,6 were seen only once or in a few families. Four novel splice site mutations (families 1, 3, 9, and 11) were confirmed on RNA analysis. The splice site mutation c.1563G→A (family 1) resulted in intron retention and an aberrant stop codon. The second splice site mutation, c.7322_7322+1delGGinsATGGAGC (family 9), led to intron retention with an aberrant stop after 2477 residues. The splice site mutation found in family 11 (c.2934+1_2934+2delGT) resulted in exon skipping. The fourth splice mutation, c.6732+1G→A (family 3), resulted in activation of a cryptic 5′ splice site within the intronic sequence and a subsequent frameshift.
Our data help in defining the phenotypic spectrum of CS. Consistent with previously reported cases of CS,7,8,12,13 the patients reported by us demonstrate an obviously broad phenotypic variability, in contrast to the patients with apparent founder mutations, who have a fairly homogeneous phenotype. Thus, COH1 mutations may be present in patients with CS lacking either microcephaly or truncal obesity. The majority of patients in our study showed short stature. Most patients reported by us (21/24) had moderate mental retardation. Three patients were classified as having profound mental retardation with complete lack of speech, and one adult patient could not walk.
In previous studies early onset of high myopia and pigmentary retinopathy have been a hallmark of CS.14,15 In all our patients older than 5 years, the severity of myopia was comparable to that reported earlier.14,15 Pigmentary retinopathy was not confirmed by retinal examination in two patients who were teenagers at the time of assessments although these patients did not undergo electrodiagnostic testing. Thus, retinopathy present at school age may not to be obligatory for the diagnosis of CS.
Despite some variability in the facial appearance in CS, craniofacial abnormalities including wave shaped palpebral fissures and short philtrum were seen in almost all patients and in different ethnic backgrounds.5,6,7,12 However, the facial dysmorphism may be subtle, which can lead to delay in diagnosis.12 In our opinion, despite mild hypertelorism, the facial appearance of one patient reported by Falk et al12 (given as sketch drawing in that report) is consistent with the typical facial features of CS. One patient in this study (patient 6) had facial features not typical of CS, including everted lips, bulbous nasal tip, and normal philtrum (fig 22).). Therefore, the diagnosis of CS cannot be ruled out in the absence of characteristic facial features.
Although no significant endocrinal abnormalities were found in Finnish patients on the examination of pituitary function,4 one of our patients, of German origin, showed growth hormone deficiency and hypogonadotropic hypogonadism.
Long term follow up and clinical information on patients older than 40 years are rare in the literature.3 Our patients between 44 and 60 years of age show that the features characteristic of CS remain in older patients, and can be used in making the diagnosis even at that age. Total blindness found in two adult patients reported here contrasts with the Finnish group of CS patients, in whom marked deterioration of visual function occurred over the age of 50 years, but no patient was reported to be completely blind.3 However, one third of CS patients from Britain were registered with severe visual impairment or blindness.16
Kyphoscoliosis was seen in three of our patients as teenagers or adults and this tends to be progressive through adult life. As CS patients do not have life threatening disorders and are generally in good health, their life span does not seem to be markedly reduced.
The detailed clinical information and the results of the molecular analysis of COH1 on patients with CS presented here enabled us to provide evidence of extended allelic heterogeneity as well as to obtain further information regarding the clinical findings, particularly the facial features, in patients of different ethnic backgrounds.
We thank all patients and their families for their collaboration. This work was supported by a grant from the Deutsche Forschungsgemeinschaft to H C Hennies and D Horn.
CS - Cohen syndrome
Competing interests: there are no competing interests.
Signed informed consents were obtained from the parents of the affected and unaffected children and guardians of affected adults. The study was approved by the local ethics committee.