POSTER NO: 460

Mutation Screening of Mitotic Spindle Checkpoint Genes in Sporadic Colorectal Cancer

¹Lise Wallis, ¹Ming Chao, ¹Justin Harvey, ²Margaret McDonald, ²Christine Morris, ¹David Markie
¹Dunedin School of Medicine, University of Otago, P.O. Box 913, Dunedin, New Zealand, ²Christchurch School of Medicine, University of Otago, P.O. Box 4385, Christchurch, New Zealand

The mitotic spindle checkpoint monitors spindle attachment to kinetochores during mitosis and delays anaphase until all chromosomes have the appropriate bipolar attachment. Failure of the spindle checkpoint results in premature anaphase with chromosome non-disjunction. Defects in this pathway have been implicated in the development of cancers that are characterised by frequent aneuploidy, such as colorectal cancer, and it has been suggested that chromosomal instability could be a mutator phenotype capable of driving carcinogenesis. If this is the case, then loss of function of this pathway should be an early event in the development of many colorectal tumours, and likely to be detectable as common somatic mutations in the genes encoding pathway components. Although occasional mutations in the spindle checkpoint genes of colorectal cancers have been reported, these have been limited to studies based on cDNA in cultured cell lines, and may not represent the mutation spectrum in primary tumours.

We have selected eight human genes (BUB1, BUB3, MAD1L1, MAD2L1, MAD2L2, BUB1B, CDC20 and TTK) that are presumed to be involved in the spindle checkpoint by virtue of their homology to yeast genes. DNA sequences flanking the exons were identified by database searching, and augmented by limited sequencing of genomic clones where necessary. Primer pairs were designed for PCR amplification of fragments that contained all coding regions of these genes, with the exception of the last 347 coding nucleotides of the MAD1L1 gene, for which no genomic sequence or clone could be identified. Each exon was amplified from the DNA of between 40 and 48 primary colorectal cancers, and screened for variation using single strand conformation analysis followed by DNA sequencing. Matched non-tumour DNA was also examined to determine if variation was inherited or resulted from somatic mutation.

Overall sixty-one DNA sequence variants were found in these eight genes, but only one of these was a somatic mutation, producing the amino acid substitution Q163N in the BUB1B gene. In the absence of a functional assay it is difficult to assess the significance of this single mutation. It seems, however, that although occasional somatic mutations can be detected in these eight putative spindle checkpoint genes, they do not make a major contribution to the development of sporadic colorectal cancers. The explanation for aneuploidy in most colorectal cancers may therefore be found in epigenetic silencing of one or more of these genes, in novel genes in this pathway, or even outside the mitotic spindle checkpoint.

Although the majority of inherited variations that were observed are likely to be neutral polymorphisms, at least one is of interest. Out of the sixty inherited variants, forty were in non-coding regions and did not affect consensus splice sites. The majority of these were single nucleotide changes, but included in this group are five short insertions or deletions, and one trinucleotide repeat polymorphism. The remaining twenty variants were in coding regions, but thirteen of these were silent, producing no amino acid change. Six variants produced amino acid substitutions, three of them were seen multiple times and three were present only once in the sample of tumours examined. It remains possible that at least some of these variants could have altered function. The remaining and most remarkable variant observed was an inherited frameshift in MAD1L1 (1248delG in codon 416 out of a total of 719 codons) resulting in premature termination after a further six codons. Although it can not be certain that inherited heterozygosity for this variant contributed to cancer in this individual, it does demonstrate the presence of a significantly variant MAD1L1 allele in the population. Further studies will be required to determine if such alleles are more common in colorectal cancer patients and may therefore be responsible for an inherited predisposition to this malignancy.