PIK3CA mutations in androgen receptor-positive triple negative breast cancer confer sensitivity to the combination of PI3K and androgen receptor inhibitors

Abstract

Introduction: Triple negative breast cancer (TNBC) is a heterogeneous collection of biologically diverse cancers, which contributes to variable clinical outcomes. Previously, we identified a TNBC subtype that has a luminal phenotype and expresses the androgen receptor (AR+). TNBC cells derived from these luminal AR + tumors have high frequency phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) mutations. The purpose of this study was to determine if targeting phosphoinositide 3-kinase (PI3K) alone or in combination with an AR antagonist is effective in AR + TNBC.

Methods: We determined the frequency of activating PIK3CA mutations in AR + and AR- TNBC clinical cases. Using AR + TNBC cell line and xenograft models we evaluated the effectiveness of PI3K inhibitors, used alone or in combination with an AR antagonist, on tumor cell growth and viability.

Results: PIK3CA kinase mutations were highly clonal, more frequent in AR + vs. AR- TNBC (40% vs. 4%), and often associated with concurrent amplification of the PIK3CA locus. PI3K/mTOR inhibitors had an additive growth inhibitory effect when combined with genetic or pharmacological AR targeting in AR + TNBC cells. We also analyzed the combination of bicalutamide +/- the pan-PI3K inhibitor GDC-0941 or the dual PI3K/mTOR inhibitor GDC-0980 in xenograft tumor studies and observed additive effects.

Conclusions: While approximately one third of TNBC patients respond to neoadjuvant/adjuvant chemotherapy, recent studies have shown that patients with AR + TNBC are far less likely to benefit from the current standard of care chemotherapy regimens and novel targeted approaches need to be investigated. In this study, we show that activating PIK3CA mutations are enriched in AR + TNBC; and, we show that the growth and viability of AR + TNBC cell line models is significantly reduced after treatment with PI3K inhibitors used in combination with an AR antagonist. These results provide rationale for pre-selection of TNBC patients with a biomarker (AR expression) to investigate the use of AR antagonists in combination with PI3K/mTOR inhibitors.

Figure 1

Identification of PIK3CA mutations in androgen receptor (AR) + triple-negative breast cancer (TNBC) cell lines and tumors. (A) Table displays the frequency of PIK3CA mutations (H1047R) in 25 TNBC AR + and 25 AR- TNBC human tumors. (B) Panel displays the TNBC molecular subtypes of The Cancer Genome Atlas (TCGA) breast cancer tumors with corresponding heatmaps showing the relative levels of AR RNA (RNA-seq) and AR protein (reverse-phase protein array (RPPA)). Those cases lacking protein or RNA evaluation are colored in black. The bottom row is a color bar indicating PIK3CA mutations (red) or wild-type (black) within the TCGA TNBC cases.

Figure 2

Androgen receptor (AR) + triple-negative breast cancer (TNBC) cell lines have activated PI3K signaling and respond to dihydrotestosterone (DHT) stimulation (A) Graphs display relative viability of the prostate cancer cell line (LNCAP, red line) compared to each of the AR + TNBC cell lines (blue line) 5 days after addition of increasing doses of DHT in charcoal-stripped media. (B) Heatmap displays relative protein levels (reverse-phase protein array, RPPA) of AR, p-AKT (S473 and T308), p-GSK3β (S9 and S21) and PTEN across indicated TNBC cell lines. Unsupervised hierarchical clustering was performed on PI3K pathway proteins (PTEN, p-AKT and p-GSK3β). Known PI3K pathway aberrations are indicated in the colorbar as PTEN loss (blue) or PIK3CA mutations (red). (C) Immunoblot displays relative protein levels of AR, p-AKT, p-S6 in AR-expressing prostate cancer (LNCaP), primary cultures of human mammary epithelial cells (HMECs) and the indicated LAR TNBC cell lines. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) serves as a loading control. (D) Immunohistochemistry of indicated proteins was performed on cell lines to assess levels of AR and p-AKT (S473). Results are representative of three independent experiments.

Figure 3

Androgen receptor (AR) + triple-negative breast cancer (TNBC) cell lines are sensitive to PI3K inhibitors. Bar graphs show the 50% inhibitory concentration (EC50) for TNBC cell lines treated for 72 h with single-agent (A) pan PI3K inhibitors (GDC-0941 and (B) NVP-BKM-120) or dual PI3K/mTOR inhibitors (GDC-0980 and NVP-BEZ235). Black horizontal bars above graphs indicate cell lines in which the EC50 was not reached at maximal concentration. Colorbar below indicates primary human mammary epithelial cell (HMECs) (black), AR + cell lines (green) and cell lines with PIK3CA mutations (red) or PTEN loss (blue). Error bars represent SD for three independent infections and experiments.

Figure 4

Genetic targeting of androgen receptor (AR) is additive in combination with PI3K inhibition. (A) Immunoblot shows AR protein levels at 72 h following transduction with two shRNAs targeting AR (shAR-1 and shAR-2) compared to transduction with nontargeting shRNA (shNT) in the indicated AR + triple-negative breast cancer (TNBC) cell lines. Actin levels are shown for the loading control. (B) Line graphs display relative viability of AR + cell lines transduced with nontargeting (shNT) or shRNAs targeting AR (shAR-1 and shAR-2) after 72 h treatment with increasing concentrations of the pan-PI3K inhibitor GDC-0941 (top) or the dual PI3K/mTOR inhibitor GDC-0980. Error bars represent SD for three independent infections.

Figure 5

Pharmacological targeting of androgen receptor (AR) with bicalutamide (CDX) is additive in combination with of GDC0941 and GDC0980 in AR + triple-negative breast cancer (TNBC) cell lines. (A) Line graphs show viability of AR + cell lines treated with increasing concentrations of GDC-0941 (top) or GDC-0980 (bottom) alone (blue) or in combination (red) with 25 μM CDX. Dashed black line depicts the theoretical line of additivity of both drugs determined from the effect of CDX alone and either GDC-0941 or GDC-0980 alone. Error bars represent SD for three independent experiments. (B) Immunoblots from AR + TNBC cell lines treated with either CDX (25 μM), GDC-0941 (300 nM) or GDC0980 (100 nM) as single agents or CDX in combination with either GDC-0941 or GDC-0980 for 48 h analyzed for AR, p-AKT, AKT, p-S6, S6 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) protein.

Figure 6

Simultaneous pharmacological targeting of androgen receptor (AR) and PI3K decreases viability of AR + triple-negative breast cancer (TNBC) cell lines grown in 3-dimensional suspension culture or in vivo as xenograft tumors. (A) Bright field images display 3-dimensional cell aggregates of MDA-MB-453 cells treated with increasing doses of GDC-0941 or GDC-0980 in the absence or presence of CDX (25 μM). (B) Line graphs display relative viability of 3-dimensional cell aggregates treated with GDC-0941 or GDC-0980 alone (blue) or in combination (red) with 25 μM bicalutamide (CDX). Dashed black line depicts the theoretical line of additivity determined from the effect of CDX alone and either GDC-0941 or GDC-0980 alone. (C) Athymic nude mice bearing established xenograft tumors from AR + TNBC cell lines (MDA-MB-453 and CAL-148) were divided into equal cohorts (n = 8) treated with either CDX (100 mg/kg/d, black hashed line), GDC-0941 (100 mg/kg/d, blue line), GDC-0980 (7.5 mg/kg/d red line) or with the combination of CDX and either GDC0-0941 (blue hashed line) or GDC-0980 (red hashed line). Serial-tumor volumes (mm³) were measured at the indicated days. Each data point represents mean tumor volume of 16 tumors; error bars represent standard error of the mean.