We report our experience to date in mutation identification in the Nor
rie disease (ND) gene. We carried out mutational analysis in 26 kindre
ds in an attempt to identify regions presumed critical to protein func
tion and potentially correlated with generation of the disease phenoty
pe. All coding exons, as well as noncoding regions of exons 1 and 2, 6
36 nucleotides in the noncoding region of exon 3, and 197 nucleotides
of 5' flanking sequence, were analyzed for single-strand conformation
polymorphisms (SSCP) by polymerase chain reaction (PCR) amplification
of genomic DNA. DNA fragments that showed altered SSCP band mobilities
were sequenced to locate the specific mutations. In addition to three
previously described submicroscopic deletions encompassing the entire
ND gene, we have now identified 6 intragenic deletions, 8 missense (s
even point mutations, one 9-bp deletion), 6 nonsense (three point muta
tions, three single bp deletions/frameshift) and one 10-bp insertion,
creating an expanded repeat in the 5' noncoding region of exon 1. Thus
, mutations have been identified in a total of 24 of 26 (92%) of the k
indreds we have studied to date. With the exception of two different m
utations, each found in two apparently unrelated kindreds, these mutat
ions are unique and expand the genotype database. Localization of the
majority of point mutations at or near cysteine residues, potentially
critical in protein tertiary structure, supports a previous protein mo
del for norrin as member of a cystine knot growth factor family (Meiti
nger et al., 1993). Genotype-phenotype correlations were not evident w
ith the limited clinical data available, except in the cases of larger
submicroscopic deletions associated with a more severe neurologic syn
drome. This lack of correlation suggests that complex epigenetic facto
rs may play a significant role in the physiological and neurodevelopme
ntal expression of the disease phenotype. Given the remarkable intra-
and interfamilial variability in hearing and brain dysfunction in this
disease, it may be expected that other phenotypic expressions of the
disease gene, which do not match the ''classical'' Norrie phenotype, m
ay be identified in the future. (C) 1995 Wiley-Liss, Inc.