Article Text

PDF

Broader geographical spectrum of Cohen syndrome due to COH1 mutations
  1. G H Mochida1,2,
  2. A Rajab3,*,
  3. W Eyaid4,*,
  4. A Lu1,5,
  5. D Al-Nouri4,
  6. K Kosaki6,
  7. M Noruzinia7,
  8. P Sarda7,
  9. J Ishihara6,8,
  10. A Bodell1,
  11. K Apse1,
  12. C A Walsh1,5
  1. 1Howard Hughes Medical Institute, Beth Israel Deaconess Medical Center, and Department of Neurology, Harvard Medical School, Boston, MA, USA
  2. 2Pediatric Neurology Unit, Department of Neurology, Massachusetts General Hospital, Boston, MA, USA
  3. 3Genetic Unit, DGHA, Ministry of Health, Muscat, Sultanate of Oman
  4. 4Department of Genetics and Paediatric Endocrinology, King Fahad National Guard Hospital, Riyadh, Saudi Arabia
  5. 5Program in Neuroscience, Harvard Medical School, Boston, MA, USA
  6. 6Department of Pediatrics, Keio University School of Medicine, Tokyo, Japan
  7. 7Department of Human Genetics, Arnaud de Villeneuve Hospital, Montpellier, France
  8. 8Department of Pediatrics, Yokohama Citizens Hospital, Yokohama, Japan
  1. Correspondence to:
 C A Walsh MD PhD
 Howard Hughes Medical Institute, Beth Israel Deaconess Medical Center, and Department of Neurology, Harvard Medical School, 4 Blackfan Circle, HIM 816, Boston, MA 02115, USA; cwalshbidmc.harvard.edu

Statistics from Altmetric.com

Cohen syndrome (COH1: MIM 216550) is an autosomal recessive disorder, first described in 1973.1 Cardinal clinical features of Cohen syndrome include microcephaly, non-progressive mental retardation, characteristic facial features, neutropenia, and ophthalmologic findings.2 It is overrepresented in Finland, though cases have been reported worldwide.

The genetic locus for Cohen syndrome was mapped to chromosome 8q in several Finnish pedigrees.3,4 Recently, a novel gene, COH1, in this locus was shown to carry mutations in many patients with Cohen syndrome.5COH1 is a large gene, consisting of 62 exons and encoding a protein of 4022 amino acids, whose biological function is not known. So far all the patients with reported mutation in the COH1 gene are of Finnish or other northern European origin. Therefore it has not been known if the COH1 gene is responsible for those cases of Cohen syndrome outside northern Europe.

Here we report novel mutations in the COH1 gene in four non-Finnish (Omani, Saudi Arabian, Japanese, and French) pedigrees, demonstrating that COH1 mutations are responsible for Cohen syndrome in non-Finnish populations. Variable phenotypes among these patients supports the idea that non-Finnish Cohen syndrome due to COH1 mutations has a broader clinical spectrum than the Finnish subtype. We also perform the first expression analysis of the mouse COH1 homologue, in order to investigate the pathogenesis of microcephaly in Cohen syndrome.

METHODS

Patients

Four pedigrees with a total of 11 affected children from a wide geographical distribution (Omani, Saudi Arabian, Japanese, and French) were enrolled into the study of COH1 mutations. The pedigree structures are shown in figure 1. The patients were identified as having features of Cohen syndrome according to the proposed diagnostic criteria.2,6 Clinical information of the four pedigrees is presented in table 1. We studied these pedigrees according to a study protocol approved by the institutional review board of Beth Israel Deaconess Medical Center and by the participating institutions. Formal ophthalmologic examinations were performed by ophthalmologists, though electroretinography was not performed. MRI images of the brain were obtained as a part of diagnostic work-up for mental retardation in some patients. Blood samples from affected children, as well as unaffected siblings and parents, when available, were collected after informed consent was obtained. DNA was extracted using standard protocols.

Table 1

Summary of clinical findings in four non-Finnish Cohen syndrome pedigrees

Figure 1

Pedigree structure and novel mutations in the four families with Cohen syndrome. (A) The Omani family carried a homozygous 7934 G>A transition in exon 43, which results in a G2645D missense mutation. (B) The Saudi Arabian family carried a homozygous 1219 C>T transition in exon 9, resulting in Q407X nonsense mutation. (C) The Japanese family was a putative compound heterozygote with a heterozygous one base pair deletion (7221delG) in exon 40, which results in frameshift. No other mutant allele was found in this family in the coding region of the COH1 gene. (D) The French family was a compound heterozygote with a one base pair deletion (11598delA) in exon 61, which results in frameshift, and a previously reported 7051 C>T transition in exon 39, resulting in R2351X nonsense mutation (not shown). Nucleotides mutated in the patients are indicated by arrowheads. Deletion followed by frameshift is indicated by “ΔFS......”. In the heterozygous patients (probands of C and D), protein translation of the mutant allele is shown.

Mutation analysis

Primers were designed to amplify each individual exon of the full length COH1 transcript (GenBank accession no. AY223814). Some larger exons required more than two primer pairs, and a total of 88 primer pairs were used. Primer sequences used for mutation analysis are available upon request. PCR products from genomic DNA were purified using PSI Clone kit (Princeton Separations), and direct sequencing was performed using BigDye Terminators kit (Applied Biosystems) and ABI 3700 capillary sequencers (Applied Biosystems).

Key points

  • Our work aimed at identifying new mutations in the COH1 gene in non-Finnish Cohen syndrome patients, and to perform the first expression study of the COH1 gene.

  • We found four novel mutations of the COH1 gene in four non-Finnish pedigrees, providing evidence that COH1 is responsible for Cohen syndrome over a wide geographical distribution. Some patients deviate from the typical Finnish Cohen syndrome phenotype, for instance in lacking microcephaly, showing that the clinical spectrum of COH1 mutations is wider in non-Finnish populations. This may be consistent with the absence of the Finnish founder mutation in our study population.

  • We performed the first expression analysis of the COH1 homologue (Coh1) in the mouse brain. Coh1 is expressed widely in neurons of the postnatal brain, but has a low level of expression embryonically, suggesting that the cardinal role of the COH1 gene may be in neuronal differentiation, but not in proliferation.

Identification of the mouse homologue of the COH1 gene and in situ hybridisation

A BLAST analysis of the human COH1 gene was performed against the public database (http://www.ncbi.nih.gov/BLAST/) and the Cerela mouse genome sequence database (http://www.celera.com/). Predicted mouse transcripts for the RIKEN cDNA c330002D13 gene (GenBank accession no. XM_283276 and Celera ID no. mCT5410) were identified and appeared to represent a partial sequence of the mouse COH1 homologue, and a PCR primer pair was designed to amplify a 907 base pair segment of this transcript, which corresponds to the amino acid residues 82 through 385 of the human COH1 protein. The primers were synthesised as chimeras, with T3 or T7 promoter sequence added to the 5′-end. PCR reaction was performed, using first-strand cDNA made from mouse 15 day embryo total RNA (BD Biosciences) as a template. Non-radioactive in situ hybridisation was performed as described,7 using a digoxigenin (DIG)-labelled cRNA probe generated from the PCR product and frozen mouse brain sections.

RESULTS AND DISCUSSION

Mutation detection

We have identified four novel COH1 mutations and one previously reported mutation in the four pedigrees studied (fig 1). The identified mutations are summarised in table 2. Patients in the Omani pedigree were homozygous for a novel mutation, 7934 G>A in exon 43. This mutation is predicted to substitute aspartic acid for glycine at residue 2645 (G2645D), which changes charge and so is very likely harmful. Both parents were confirmed to be heterozygous for this mutation. The clinical features of the patients in the Omani family are not milder than others with nonsense or frameshift mutations, suggesting that this missense mutation is also probably a null mutation.

Table 2

Summary of the mutations in four non-Finnish Cohen syndrome pedigrees

Affected individuals in the Saudi Arabian pedigree were homozygous for a second novel mutation, 1219 C>T in exon 9. This mutation changes a glutamine residue at position 407 of 4022 to a termination codon (Q407X), creating a severely truncated predicted protein. Parents were heterozygous for the mutation, and the unaffected sibling tested was homozygous for the wild-type allele.

Patients in the Japanese pedigree were heterozygous for a novel COH1 mutation, 7221delG in exon 40. This mutation creates a translational frameshift after residue 2407, with predicted premature termination at residue 2413 (Q2407H...V2414X). Their mother also carried this base pair change in the heterozygous state, while their father was not available for testing. These siblings are presumably compound heterozygotes for two different COH1 gene mutations, but a second mutation was not found despite complete sequencing of all known coding exons of the full-length COH1 transcript.

The patients in the French pedigree were all compound heterozygotes for COH1 mutations. One mutation, 7051 C>T in exon 39, has already been reported in a Belgian patient,5 and is predicted to change an arginine residue at position 2351 to a termination codon (R2351X). The other mutation, 11598delA in exon 61, has not been reported before, and causes a translational frameshift, leading to premature termination after the addition of 10 abnormal amino acids (E3867K...V3877X; fig 1). None of the four novel mutations we found were observed in at least 56 Caucasian (112 chromosomes) and 26 Middle Eastern (Omani) unrelated normal control individuals (52 chromosomes).

It is interesting to note that we did not find in any of the pedigrees we studied the mutation that seems to be most prevalent in Finland, 3348_3349delCT. Out of 27 Finnish patients who were studied previously, 26 had at least one allele of this mutation.5 Furthermore, it appears likely that different mutations prevail among different ethnic groups, although further mutation screening is necessary to confirm this.

Genotype–phenotype correlation

No clear genotype–phenotypic correlation was apparent in the patients we studied. However, the patients reported here seem to show more phenotypic variability compared to the Finnish Cohen syndrome patients, who have a highly homogeneous clinical phenotype. For example, patients 7 and 10 in our study lack microcephaly. In addition, patients 1, 2, 3, and 7 were not found to have neutropenia, though neutropenia may be intermittent, and this cannot be ruled out as these patients had blood counts tested on one occasion. Both microcephaly and neutropenia appear to be invariant features in reported Finnish patients,2 and reported patients from other populations may lack these features.6 The variability in clinical features that we found may be consistent with the absence of a Finnish founder mutation in our study population and, possibly, in other reported non-Finnish Cohen syndrome patients.

One of the mutations found in the French family was a novel frameshift mutation that truncates only the C-terminal region of the predicted protein. Perhaps this mutation results in partial loss of function of the COH1 protein, which may account for the relatively preserved head circumference in this pedigree, though otherwise it shows a classic phenotype similar to the Finnish Cohen syndrome. Though the biological function of the COH1 protein is not known, there is a potential endoplasmic reticulum retention signal detected by the PSORT II program at the C-terminal region5,8; therefore, the French mutation may partially disrupt the function of the protein by interfering with subcellular localisation.

Identification of the mouse homologue of the COH1 gene and in situ hybridisation

A partial sequence for the mouse homologue of the COH1 gene was identified through BLAST searches. Predicted mouse transcripts for the RIKEN cDNA c330002D13 gene (GenBank accession no. XM_283276 and Celera ID no. mCT5410), which appear to represent a partial sequence of the mouse COH1 homologue (Coh1), were identified. The predicted amino acids translated from XM_283276, for example, are 90% identical to the first 415 amino acids of the human COH1 protein. In situ hybridisation was performed using a probe designed from this sequence. In the mouse brain at postnatal day 21, roughly equivalent to early childhood to juvenile age in humans, the Coh1 gene was expressed widely in the central nervous system. In the cerebellum, a high level of expression was seen in the internal granule layer and Purkinje cells, but not in the white matter (fig 2A). In the cerebral cortex, a high level of expression was seen in all cortical layers except for layer I, but expression was absent in the subcortical white matter (fig 2B). A high level of expression was also seen in the dentate gyrus of hippocampus and throughout Ammon’s horn. The adult mouse brain showed a similar pattern of expression (data not shown). These data suggested that all or most neurons in the adult brain express Coh1, with no obvious regional differences.

Figure 2

In situ hybridisation of the mouse homologue of the COH1 gene. (A) Cerebellum of the mouse at postnatal day 21. A high level of expression is seen in the internal granule layer (g), and little expression is seen in the molecular layer (m) and white matter (wm). Purkinje cells (inset) also show expression. (B) Cerebral cortex of the mouse at postnatal day 21. Many cells in layers II through VI of the cerebral cortex (II–VI) show expression, but only a few cells are positive in layer I of the cortex (I) and white matter (wm). (C) Cerebral wall of the mouse at embryonic day 14. A low level of expression is seen in both the cortical plate (cp) and ventricular zone (vz). Scale bars represent 100 μm in A and C, 250 μm in B.

In contrast to the adult brain, the embryonic mouse brain showed low or undetectable levels of Coh1 expression using the same labelled probe (fig 2C). Repeated attempts gave a low signal not greatly different from background and less intense than hybridisation seen with a sense control probe. Thus, Coh1 appears to be widely expressed in neurons of the postnatal and adult brain, but is not highly expressed in the embryonic brain.

Our results indicate that the gene is expressed widely in the neurons of the central nervous system. However, there was little expression in the white matter tracts of cerebrum and cerebellum, suggesting that the gene is not expressed in glial cells at a high level. In addition, the level of expression appeared to be higher in the postnatal brain than in the embryonic brain. This may suggest that the COH1 gene primarily functions in postmitotic neurons, and not in neuronal progenitor cells. These findings are consistent with the fact that patients with Cohen syndrome are generally born with normal head circumference and develop microcephaly postnatally.2COH1 mutations may cause microcephaly by disrupting, for example, dendritic or axonal outgrowth, which continues postnatally and is important for normal postnatal brain growth.

Further molecular genetic studies, in combination with biochemical studies of the COH1 protein, will likely help us understand the pathogenetic mechanism of Cohen syndrome. In addition, studying additional patients with atypical phenotypes (for example, less severe mental retardation or without neutropenia) may provide additional insights into the function of the COH1 gene. Furthermore, defining the clinical-genetic spectrum of Cohen syndrome will potentially benefit patients and their families by allowing accurate diagnosis of the syndrome.

Acknowledgments

We are deeply grateful to the patients and their family members for their participation in this study. We thank Drs A James Barkovich and P Ellen Grant for help with MRI analysis, Sarah Teillon for help with DNA sequencing, Dr Tao Sun for providing embryonic mouse brain sections, and Dr Urs Berger for help with in situ hybridisation. We are grateful to the Dana-Farber/Harvard Cancer Center High-Throughput DNA Sequencing Facility and the Harvard Medical School Biopolymers Facility for help with DNA sequencing. CAW is an Investigator of the Howard Hughes Medical Institute.

REFERENCES

View Abstract

Footnotes

  • GHM is supported by the Charles H. Hood Foundation as a Medical Foundation Research Fellow, and by the William Randolph Hearst Fund. CAW is supported by grants from the National Institute of Neurological Disease and Stroke (2R37NS35129 and P01NS39404), the March of Dimes, and the McKnight Foundation.

  • Conflict of interest: none declared.

  • * These authors contributed equally to this work.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.