PTEN-deficient cancers depend on PIK3CB

Abstract

Deregulation of the PI3K signaling pathway is observed in many human cancers and occurs most frequently through loss of PTEN phosphatase tumor suppressor function or through somatic activating mutations in the Class IA PI3K, PIK3CA. Tumors harboring activated p110alpha, the protein product of PIK3CA, require p110alpha activity for growth and survival and hence are expected to be responsive to inhibitors of its lipid kinase activity. Whether PTEN-deficient cancers similarly depend on p110alpha activity to sustain activation of the PI3K pathway has been unclear. In this study, we used a single-vector lentiviral inducible shRNA system to selectively inactivate the three Class IA PI3Ks, PIK3CA, PIK3CB, and PIK3CD, to determine which PI3K isoforms are responsible for driving the abnormal proliferation of PTEN-deficient cancers. Down-regulation of PIK3CA in colorectal cancer cells harboring mutations in PIK3CA inhibited downstream PI3K signaling and cell growth. Surprisingly, PIK3CA depletion affected neither PI3K signaling nor cell growth in 3 PTEN-deficient cancer cell lines. In contrast, down-regulation of the PIK3CB isoform, which encodes p110beta, resulted in pathway inactivation and subsequent inhibition of growth in both cell-based and in vivo settings. This essential function of PIK3CB in PTEN-deficient cancer cells required its lipid kinase activity. Our findings demonstrate that although p110alpha activation is required to sustain the proliferation of established PIK3CA-mutant tumors, PTEN-deficient tumors are dependent instead on p110beta signaling. This unexpected finding demonstrates the need to tailor therapeutic approaches to the genetic basis of PI3K pathway activation to achieve optimal treatment response.

Fig. 1.

PIK3CA is required for PI3K signaling and growth in p110α-mutant cell lines. Dox = doxycycline. (A) HCT116 and DLD1 colorectal cancer cell lines transduced with scramble control or PIK3CA-inducible shRNA were cultured in the presence or absence of doxycycline at 10 ng/ml for 72 h and harvested for Taqman analysis. PIK3CA mRNA levels in the control samples (− DOX) were set to 100%. (B) Similarly treated cells were analyzed by Western blot to monitor changes in PI3K pathway signaling. The phosphorylation states of PRAS40 at Thr-246, Foxo1 at Thr-24, Foxo3a at Thr-32, and p70/p85 S6K at Thr 389 were assessed. Endogenous GAPDH is shown as a loading control. (C) Proliferation of stable HCT116 and DLD1 shRNA cell lines under low (0.5% FBS) serum conditions in the presence or absence of doxycycline (10 ng/ml) were monitored using CellTiterGlo over a 10-day period. Results are shown as mean ± SE of 3 replicates. (D) Cells cultured in a 6-well dish for 14 days in the presence (10 ng/ml) or absence of doxycyline were stained with crystal violet to visualize colony growth. All experiments were done in triplicates. (E) HCT116 cells were grown in semisolid medium for 14 days in the presence (100 ng/ml) or absence of doxycycline. Colonies were visualized by Hoechst 33342 staining and photographed using a Nikon fluorescence microscope. Colonies were counted (mean ± SE, in triplicate) using ImagePro software.

Fig. 2.

PIK3CA activity is not required for PI3K signaling or growth in PTEN-deficient cancer cell lines. DOX = doxycycline. (A) BT549, PC3, and U87MG cell lines transduced with scramble control or PIK3CA inducible shRNA were cultured in the presence or absence of doxycycline (10 ng/ml) for 72 h and harvested for Taqman analysis. PIK3CA mRNA levels in the control samples (− DOX) were set to 100%. (B) Similarly treated cells were analyzed by Western blot to monitor changes in p110α levels upon PIK3CA knockdown and the effect on p473AKT signaling. Constitutively expressed TetR protein was used as a loading control. (C) Proliferation of PC3 cells was monitored over 10 days using CellTiterGlo. BT549 and U87MG cell proliferation was monitored by MTS assay over 10 days. (D) shRNA stable cell lines were seeded in triplicates onto 6-well dishes (1500 cells per well) and allowed to attach for 24 h. Doxycycline was added to the indicated wells at a final concentration of 10 ng/ml. Following 10-day incubation, dishes were fixed and stained with crystal violet. (E) Control or PIK3CA shRNA PC3 cells were grown in semisolid medium for 14 days in the presence or absence of doxycycline (100 ng/ml). The resulting colonies were visualized by Hoechst 33342 staining and photographed using a Nikon fluorescence microscope. Colonies were counted using ImagePro software (mean ± SE, in triplicate).

Fig. 3.

PIK3CA is required for growth of p110α mutant, but not PTEN-deficient, tumors in vivo. Dox = doxycycline. (A and B) HCT116 and PC3 cells stably transduced with PIK3CA inducible shRNA were implanted into nude mice as described in Materials and Methods. Mice containing tumors of at least 100 mm³ were administered vehicle control or doxycycline either freely in drinking water (ad libitum) or once a day (qd) by oral gavage (p.o.). Tumor volume was monitored by calipering (mean ± SEM). (C) Changes in PI3K signaling in tumor xenografts were assessed by Western blot using indicated antibodies. Phosphorylation states of PRAS40 at Thr 246 and S6 at Ser-235/S236 were assessed.

Fig. 4.

PTEN-deficient cells require PIK3CB for growth. Dox = doxycycline. (A) Stable control or PIK3CA shRNA–containing HCT116 cells were transfected with either control or PTEN siRNA. Cells grown in the absence or presence of doxycycline were harvested for Western blot analysis after 72 h of treatment. (B) BT549, PC3, and U87MG cells stably transduced with either scramble control or PIK3CB were grown in the presence or absence of doxycycline for 72 h. PIK3CB transcript levels were assessed at the end of treatment by Taqman RT-PCR. (C) Described cells were cultured in a 6-well dish in the presence or absence of doxycycline (10 ng/ml) for 14 days. The effects of PIK3CB silencing on foci formation were visualized by crystal violet staining.

Fig. 5.

p110β-Lipid kinase activity is required for PI3K pathway signaling and growth in PTEN-deficient cells. Stable PC3 cells containing inducible PIK3CB targeting shRNA were transduced with either wild-type or kinase-dead (D937A) shRNA-resistant p110β cDNA. Dox = doxycycline. (A) PIK3CB transcript levels were assessed by Taqman analysis in the absence or presence of 10 ng/ml of doxycycline treatment for 72 h. (B) The effect on PI3K pathway signaling was assessed by Western blot analysis upon expression of the respective rescue construct for 96 h. The effect on PRAS40 phosphorylation at Thr 246 was determined. (C) Stable PC3 cells containing the respective rescue constructs were cultured in a 6-well dish for 14 days in the presence (10 ng/ml) or absence of doxycycline. Cells were stained with crystal violet to visualize colony growth. All experiments were done in triplicate.

Fig. 6.

PIK3CB is required for growth of PTEN-deficient tumors in vivo. Dox = doxycycline. (A) Stable PC3 cells containing PIK3CB inducible shRNA were implanted into nude mice and administered vehicle control or doxycycline by oral gavage (p.o.) once per day (qd) upon tumors reaching 100 mm³. Tumor volume was measured using calipers (mean ± SEM). (B) Tumors harvested from either untreated or doxycycline-treated mice were analyzed by Western blot with the respective antibodies. The phosphorylation states of PRAS40 at Thr 246, of S6 at Ser-235/S236, and of GSK3β at Ser-9 were determined.