RNA analysis reveals splicing mutations and loss of expression defects in MLH1 and BRCA1

Authors

  • Andrew Sharp,

    Corresponding author
    1. Wessex Regional Genetics Laboratory, Salisbury District Hospital, Salisbury, Wiltshire SP2 8BJ, UK
    Current affiliation:
    1. Department of Genome Sciences K360, University of Washington, 1705 Pacific St., Seattle, WA 98195-7730
    • Department of Genome Sciences K360, University of Washington, 1705 Pacific St., Seattle, WA 98195-7730
    Search for more papers by this author
  • Gabriella Pichert,

    1. Wessex Clinical Genetics Service, Princess Anne Hospital, Southampton, Hampshire, SO16 5YA, UK
    Search for more papers by this author
  • Anneke Lucassen,

    1. Wessex Clinical Genetics Service, Princess Anne Hospital, Southampton, Hampshire, SO16 5YA, UK
    Search for more papers by this author
  • Diana Eccles

    1. Wessex Clinical Genetics Service, Princess Anne Hospital, Southampton, Hampshire, SO16 5YA, UK
    Search for more papers by this author

Abstract

The classical paradigm of mutation screening seeks to relate alterations in DNA sequence to their effect at the protein level. However, the majority of missense mutations are problematic as their pathological significance is often unclear. In order to test the hypothesis that many missense mutations primarily cause defects at the RNA rather than the protein level, we have performed retrospective RNA analysis of 12 individuals carrying missense mutations in the cancer predisposition genes APC, BRCA1, BRCA2, MLH1, and MSH2. RNA was extracted from peripheral blood samples and RT-PCR performed in order to assess the splicing and expression of the mutant allele in each case. Four of the 12 missense mutations analysed were associated with RNA defects. We detected two cases of exon skipping and one case of partial intron inclusion with activation of a cryptic intronic splice site in MLH1. A fourth case was associated with monoallelic expression of BRCA1. In addition, allele-specific analysis of common coding polymorphisms identified a further case of monoallelic BRCA1 expression in one of two individuals who had previously screened as mutation-negative. Although we were unable to identify the underlying cause of this loss of expression, it strongly suggests the presence of a pathogenic defect in BRCA1 in this case, highlighting the use of allelic expression studies as a method of mutation scanning. Finally, we used our dataset to test the ability of several in silico sequence analysis tools to identify splicing defects. Our results suggest that a significant number of missense mutations in cancer predisposition genes are associated with defects of RNA splicing, and that the use of gene- and splice site prediction software can aid in identifying such mutations. © 2004 Wiley-Liss, Inc.

Ancillary