Cohen syndrome is caused by mutations in a novel gene, COH1, encoding a transmembrane protein with a presumed role in vesicle-mediated sorting and intracellular protein transport

Abstract

Cohen syndrome is an uncommon autosomal recessive disorder whose diagnosis is based on the clinical picture of nonprogressive psychomotor retardation and microcephaly, characteristic facial features, retinal dystrophy, and intermittent neutropenia. We have refined the critical region on chromosome 8q22 by haplotype analysis, and we report the characterization of a novel gene, COH1, that is mutated in patients with Cohen syndrome. The longest transcript (14,093 bp) is widely expressed and is transcribed from 62 exons that span a genomic region of approximately 864 kb. COH1 encodes a putative transmembrane protein of 4,022 amino acids, with a complex domain structure. Homology to the Saccharomyces cerevisiae VPS13 protein suggests a role for COH1 in vesicle-mediated sorting and transport of proteins within the cell.

Figure 1

Characteristic facial features in patients with Cohen syndrome. A, A 12-year-old girl. B, A 21-year-old man. Both patients are Finnish and are homozygous for the c.3348_3349delCT mutation.

Figure 2

Positional identification of the COH1 gene. A, Haplotypes in 40 COH1 chromosomes from one patient in each of 20 Finnish families with Cohen syndrome. Twelve markers covering the 2.2-Mb COH1 critical region were used. Italic type indicates the five microsatellite markers published elsewhere. Marker CA-CEN1 is within intron 33 of COH1. Haplotypes associated with the c.3348_3349delCT mutation are boxed. Historical recombinations refine the COH1 region to between D8S257 and D8S559. Nondefined alleles are marked with “N.” B, The COH1 critical interval. Shown are the positions of three polymorphic markers, known genes analyzed for mutations in patients with Cohen syndrome (represented as arrows in the direction of transcription), and the COH1 transcript (assembled using in silico homology searches, exon predictions, and RT-PCR); the direction of transcription is indicated (arrowhead).

Figure 3

Genomic structure of COH1 and schematic representation of COH1 alternative splicing. A, Genomic structure of COH1. The COH1 transcript, comprising 14,093 nt, covers an 864-kb genomic region and is transcribed from 62 exons. The alternatively spliced exons present in ESTs are shown as colored boxes. The exons are shown in scale relative to each other. The lengths of the intronic sequences are in scale relative to each other but not relative to the exons. B, Schematic representation of alternative splicing of COH1. The longest predicted transcript (GenBank accession number AY223814) uses exons 8, 17, 28, and 31 and has an ORF of 12,066 bp. Use of exon 28b, instead of 28, results in a transcript (GenBank accession number AY223815) that is 75 bp shorter. Additional variation is predicted from in-frame splicing of the 204-bp exon 31 present in ESTs but not experimentally verified. The presence of exons 28 and 28b predicts a shorter transcript, with an ORF of 4,281 bp (GenBank accession number AY223816). Alternative exons 17b and 8b result in ORFs of 2,589 bp (Genbank accession number AY223817) and 1,236 bp (GenBank accession number AY223818), respectively.

Figure 4

Expression pattern of the COH1 gene. COH1 is widely expressed. A 489-bp probe from the coding region of COH1 was used in human MTN blot analysis (fetal MTN [A], MTN I [B], and MTN II [C]; Clontech). Two isoforms (∼2 and ∼5 kb) were present in all tissues. Expression of the longest ∼12–14-kb transcript was detectable in prostate, testis, ovary, and colon. A β-actin probe (bottom) was used as a control for RNA loading.

Figure 5

Mutations in the COH1 gene. A–I, sequencing chromatograms showing the nine identified mutations from seven patients with Cohen syndrome and controls. Two mutations (A and C) are homozygous, whereas the rest are in heterozygous form. Nucleotide changes (arrows) and codons with a single aa change (box) are shown. Out-of-frame sequences are underlined.

Figure 6

Predicted membrane topology and domains of COH1. The predicted transmembrane domains are numbered 1–10. The locations of the mutations identified in patients with Cohen syndrome are shown (blackened arrows). The positions of the predicted functional motifs (blue arrows) and vacuolar sorting–associated VPS13 domains (gray boxes) are indicated. Conserved patterns observed in Prosite database searches are also indicated (brown arrows).