Activating K-Ras mutations outwith 'hotspot' codons in sporadic colorectal tumours - implications for personalised cancer medicine

Abstract

Background: Response to EGFR-targeted therapies in colorectal cancer patients has been convincingly associated with Kirsten-Ras (K-Ras) mutation status. Current mandatory mutation testing for patient selection is limited to the K-Ras 'hotspot' codons 12 and 13.

Methods: Colorectal tumours (n=106) were screened for additional K-Ras mutations, phenotypes compared in transformation and Ras GTPase activating assays and gene and pathway changes induced by individual K-Ras mutants identified by microarray analysis. Taqman-based gene copy number and FISH analyses were used to investigate K-Ras gene amplification.

Results: Four additional K-Ras mutations (Leu(19)Phe (1 out of 106 tumours), Lys(117)Asn (1 out of 106), Ala(146)Thr (7 out of 106) and Arg(164)Gln (1 out of 106)) were identified. Lys(117)Asn and Ala(146)Thr had phenotypes similar to the hotspot mutations, whereas Leu(19)Phe had an attenuated phenotype and the Arg(164)Gln mutation was phenotypically equivalent to wt K-Ras. We additionally identified a new K-Ras gene amplification event, present in approximately 2% of tumours.

Conclusions: The identification of mutations outwith previously described hotspot codons increases the K-Ras mutation burden in colorectal tumours by one-third. Future mutation screening to facilitate optimal patient selection for treatment with EGFR-targeted therapies should therefore be extended to codon 146, and in addition should consider the unique molecular signatures associated with individual K-Ras mutations.

Figure 1

The Ras cycle. Ras proteins are key components of signal transduction pathways leading from cell-surface receptors to the control of cell proliferation, differentiation or death. Active Ras, where tumour-specific mutations lock Ras in the GTP-bound conformation, stimulates the RAS–RAF–MEK–ERK–MAP kinase signalling pathway.

Figure 2

Location of novel and hotspot Kirsten-Ras (K-Ras) mutations. The location of new (red) and hotspot (blue) K-Ras mutations are illustrated on a representation of the K-Ras protein sequence, together with the position of the novel K-Ras single-nucleotide polymorphism (SNP) (black). Putative GTP and effector binding sites and GAP and GEF interaction domains are highlighted.

Figure 3

Focus formation assays. NIH3T3 cells were transfected with pEF.6 or plasmids containing wt or mutant K-Ras and foci visualised after crystal violet staining. (A) Cells transfected with empty vector or wt K-Ras were compared with previously described hotspot mutations. (B) The transforming potential of L19F, K117N, A146T and R164Q mutations were compared with the hotspot mutation G12V and wt K-Ras. All experiments were performed in duplicate and each set of transfections was repeated three times. (C) The combined results of all transfections, wherein total foci counts are presented ± SD are illustrated.

Figure 4

Ras GTPase activating assays. Ras GTPase activating assays were carried out for each of the novel Kirsten-Ras (K-Ras) mutations and western blotting used to assess the expression of (A) active GTP-bound K-Ras (B) total K-Ras and (C) β-actin (loading control). The K-Ras G12V construct was included as a positive control for K-Ras protein in the active GTP-bound conformation.

Figure 5

Hierarchical clustering analysis. RNA transcription profiling analysis was carried out as described in Materials and Methods. Lists of differentially expressed genes were generated comparing each of the Kirsten-Ras (K-Ras) mutants with empty vector control using probes that exhibited an adjusted P-value (false discovery rate (FDR)) ⩾0.05. Following hierarchical clustering analysis, gene clusters were represented on a heat map, where upregulated genes are highlighted in green and downregulated genes highlighted in red.

Figure 6

Kirsten-Ras (K-Ras) gene copy number assay. A novel K-Ras gene copy number assay was designed as described in Materials and Methods and used to screen genomic DNA samples extracted from 96 paired normal (N) and colorectal tumour (T) tissues. Representative results from six normal/tumour pairs are illustrated, highlighting increased copy number in samples 1233 and 1264.

Figure 7

FISH analysis. Chromosome metaphase spreads were hybridised with FISH probes for (A) chromosome 12p (red) and (B) chromosome 12q (green) to confirm probe specificity and FISH analysis performed as described in Materials and Methods. Representative colorectal tumour sections (C) without and (D) with a K-Ras gene amplification are illustrated.