PIK3CA mutation in colorectal cancer: relationship with genetic and epigenetic alterations

Abstract

Somatic PIK3CA mutations are often present in colorectal cancer. Mutant PIK3CA activates AKT signaling, which up-regulates fatty acid synthase (FASN). Microsatellite instability (MSI) and CpG island methylator phenotype (CIMP) are important molecular classifiers in colorectal cancer. However, the relationship between PIK3CA mutation, MSI and CIMP remains uncertain. Using Pyrosequencing technology, we detected PIK3CA mutations in 91 (15%) of 590 population-based colorectal cancers. To determine CIMP status, we quantified DNA methylation in eight CIMP-specific promoters [CACNA1G, CDKN2A (p16), CRABP1, IGF2, MLH1, NEUROG1, RUNX3, and SOCS1] by real-time polymerase chain reaction (MethyLight). PIK3CA mutation was significantly associated with mucinous tumors [P = .0002; odds ratio (OR) = 2.44], KRAS mutation (P < .0001; OR = 2.68), CIMP-high (P = .03; OR = 2.08), phospho-ribosomal protein S6 expression (P = .002; OR = 2.19), and FASN expression (P = .02; OR = 1.85) and inversely with p53 expression (P = .01; OR = 0.54) and beta-catenin (CTNNB1) alteration (P = .004; OR = 0.43). In addition, PIK3CA G-to-A mutations were associated with MGMT loss (P = .001; OR = 3.24) but not with MGMT promoter methylation. In conclusion, PIK3CA mutation is significantly associated with other key molecular events in colorectal cancer, and MGMT loss likely contributes to the development of PIK3CA G>A mutation. In addition, Pyrosequencing is useful in detecting PIK3CA mutation in archival paraffin tumor tissue. PIK3CA mutational data further emphasize heterogeneity of colorectal cancer at the molecular level.

Figure 1

PIK3CA Pyrosequencing assay design. (A) The exon 9 PCR products were sequenced by the three different Pyrosequencing primers (9-RS1, 9-RS2, and 9-RS3). Nucleotide positions of the hot spot mutations are underlined and in bold capitals. (B) The exon 20 PCR products were sequenced by the Pyrosequencing primer (20-RS). Nucleotide positions of the hot spot mutations are underlined and in bold capitals.

Figure 2

PIK3CA exon 9 Pyrograms (antisense strand). (A) Wild type exon 9 by the 9-RS1 primer. (B) The c.1634A>G mutation (arrow) causes a shift in reading frame and results in a new peak at A (arrowhead), which serves as quality assurance. (C) The c.1636C>A mutation (arrow) causes a shift in reading frame and results in a new peak at C (arrowhead), which serves as quality assurance. (D) Wild type exon 9 by the 9-RS2 primer. (E) The c.1633G>A mutation (arrow) causes a shift in reading frame and results in new peaks (arrowheads), which serves as quality assurance. (F) Wild type exon 9 by the 9-RS3 primer. (G) The c.1624G>A mutation (arrow) causes a shift in reading frame and results in a new peak at A (arrowhead), which serves as quality assurance. Mut indicates mutant; WT, wild type.

Figure 3

PIK3CA exon 20 Pyrograms (antisense strand). (A) Wild type exon 20. (B) The c.3129G>T mutation (arrow) causes a shift in reading frame and results in new peaks at T and G (arrowheads), which serves as quality assurance. (C) The c.3139C>T mutation (arrow) causes a shift in reading frame and results in a new peak at T (arrowhead), which serves as quality assurance. (D) The c.3140A>T mutation (arrow) causes a shift in reading frame and results in a new peak at G (arrowhead), which serves as quality assurance. (E) The c.3140A>G mutation (arrow) causes a shift in reading frame and results in a new peak at G (arrowhead), which serves as quality assurance. Mut indicates mutant; WT, wild type.

Figure 4

Frequencies of PIK3CA G>A and non-G>A mutations according to MGMT status. PIK3CA G>A mutation was significantly associated with MGMT loss (P = .001). No significant association was found between PIK3CA non-G>A mutation and MGMT loss. CI indicates confidence interval; OR, odds ratio; WT, wild type.