In a consanguineous family of three affected siblings who were suffering from a complex congenital disorder, homozygosity mapping had revealed a linkage interval on chromosome 4, and mutation screening of the coding regions in this interval by exon enrichment and Illumina/Solexa sequencing revealed a frameshift mutation in SRD5A3.
Mutations in SRD5A3
have been identified independent of this study in other patients and families with partially overlapping clinical features and congenital defects of glycosylation.3 SRD5A3
encodes steroid 5α
-reductase type 3, an enzyme which is required for the conversion of polyprenol to dolichol, the lipid anchor for N
-glycosylation in the endoplasmic reticulum.3 SRD5A3
is highly expressed in fetal brain.3
Cantagrel and colleagues3
and Morava et al
(personal communication) identified a total of seven different SRD5A3
mutations in nine families. Most of these cases were ascertained by abnormal transferrin isoelectric focussing patterns suggesting a type I congenital defect of glycosylation (‘CDG type Ix', OMIM 212067).4
In the family reported here, no clear abnormal transferrin mobility had been detected in two independent routine tests for CDG screening. Interestingly, in the family reported by Al-Gazali et al
now also known to carry an SRD5A3
repeated CDG testing had failed to detect any abnormality, too. Thus, in patients with clinical features suggestive of Kahrizi syndrome, SRD5A3
mutations should be ruled out even if CDG testing is negative.
Apart from severe mental retardation which was observed in all patients with SRD5A3
mutations, the presence of additional clinical features seems to vary considerably, not only between different families, but also within families. Cerebellar ataxia was present in most patients, but malformation of the cerebellum or the cerebellar vermis were absent in the original family with Kahrizi syndrome and only detectable in some patients by MRI investigation. Ophthalmic problems (cataract, iris coloboma, optic nerve hypoplasia) were frequently observed. Cataracts were only present in two of the patients of the Cantagrel/Morava cohort.3
In the family reported here, all three siblings developed cataracts at the age of about 17 years. As the oldest patient in the Cantagrel/Morava cohort was only 12-years old, it may be expected that other patients will also develop cataracts as they grow older.
The syndrome reported here has some phenotypic overlap with Peter's plus syndrome (OMIM 261540), an autosomal recessive disorder which is caused by mutations in the gene encoding B3GALTL, a putative glycosyltransferase.6
Also, in patients with Peter's plus syndrome, routine CDG testing by isoelectric focussing of transferrin failed to reveal any abnormalities. This example, and the family reported here, may serve as a reminder that a normal isoelectric focussing of transferrin result does not always exclude glycosylation defects, and – based on the clinical phenotype – additional molecular genetic tests may be necessary.
In summary, the detection of an SRD5A3 mutation in this family shows that the recently described Kahrizi syndrome (OMIM 612713) and a specific CDG type Ix subtype (OMIM 212067) are allelic disorders. This study also demonstrates the effectiveness of combining exon enrichment and next generation sequencing technology for detecting causative gene defects.