Although the roles of the PI3K catalytic subunit in tumorigenesis is well established, current understanding of the function of PI3K regulatory subunits such as PIK3R3 is still limited. A previous study showed that PIK3R3 gene had copy number gain in high-grade glioma, and the Comparative Genomic Hybridization (CGH) ratios for this locus were positively-correlated with proliferation signatures
]. Furthermore, the IGF2-PIK3R3 signaling axis was involved in promoting the growth of a subclass of highly aggressive human glioblastomas
]. In this study, we compared PIK3R3 expression in 126 paired GC with non-neoplastic gastric mucosa control tissues from the same patients, and found that approximately 9.5% to 15% of GC showed increased PIK3R3 expression. IPA pathway analysis suggested a key role of PIK3R3 in cell proliferation. Knockdown of PIK3R3 by siRNA decreased HGC27 GC cell proliferation and induced cell cycle blockade at G0
. Taken together, these results suggest that PIK3R3 stimulates cell proliferation in GC and may be a potential therapeutic target for a subgroup of GC.
PIK3R3 shares significant sequence identity with PIK3R1 and PIK3R2 regulatory subunit in a proline-rich motif and two Src homology 2 (SH2) domains; however, it has an unique 24-residue NH2 terminus
]. PIK3R3 binds to Rb via this unique region as the N-terminal 24 amino acids of PIK3R3 (N24) are sufficient for binding to Rb
]. Ectopic expression of N24 also inhibited cyclin D1 and E promoter activity and caused G0
phase arrest in several cancer cell lines
]. Likewise, in the current study, knockdown of PIK3R3 caused G0
cell cycle arrest in HGC27 and TMK1 GC cells. Consistent with this observation, Rb phosphorylation, cyclin D1, and PCNA protein levels also were decreased after PIK3R3 knockdown. Interestingly, the pAkt levels did not change after PIK3R3 knockdown. Additionally, when PIK3R3 was overexpressed in AGS cells, a low PIK3R3-expressing gastric cancer cell line, we did not observe any significant change in phosphorylation of Akt (data not shown). Taken together, these findings show that both increased and decreased expression of PIK3R3 do not affect pAkt levels in these cell lines under our experimental conditions. Of note, we have observed similar effects in several colon cancer cell lines and MCF7 breast cancer cells (
]and unpublished data). This lack of change in pAkt levels may be due to either preferential nuclear localization of PIK3R3
] or low PIK3R3 expression relative to p85 regulatory subunit isoforms. In contrast to our findings, Soroceau et al.
found decreased pAkt in G96 glioblastoma cells after PIK3R3 knockdown which may be due to different experimental conditions or different tissue origins of cell lines as PIK3R3 expression is high in the brain
]. Nevertheless, our current and previous studies suggest that PIK3R3 can promote cell growth through a novel signaling pathway by regulating Rb phosphorylation and cyclin D1.
Finally, we investigated the mechanism of PIK3R3 over-expression in GC. Zhang et al.
found PIK3R3 was over-expressed in ovarian cancer due to gene copy number gain
]. However, no gene amplification or DNA de-methylation of PIK3R3 gene was found in our set of GC samples. Interestingly, Demoulin et al.
found that PDGF activated transcription factor SREBP-1 in a PI3K-dependent manner
], and ectopic expression of SREBP-1 induces the expression of PIK3R3 in AG01518 human foreskin fibroblasts
]. Therefore, these results suggested that activated PI3K signaling could induce the expression of PIK3R3. Indeed, we observed a positive correlation between PIK3CA and PIK3R3 mRNA expression in our 126 GC samples. Furthermore, the PIK3CA inhibitor, LY294002 decreased PIK3R3 expression in HGC27 and IM95 GC cell lines. Our results thus provide direct experimental evidence that PIK3R3 expression can be regulated by PIK3CA.