Proline rich 11 (PRR11) overexpression amplifies PI3K signaling and promotes antiestrogen resistance in breast cancer

Abstract

The 17q23 amplicon is associated with poor outcome in ER⁺ breast cancers, but the causal genes to endocrine resistance in this amplicon are unclear. Here, we interrogate transcriptome data from primary breast tumors and find that among genes in 17q23, PRR11 is a key gene associated with a poor response to therapeutic estrogen suppression. PRR11 promotes estrogen-independent proliferation and confers endocrine resistance in ER⁺ breast cancers. Mechanistically, the proline-rich motif-mediated interaction of PRR11 with the p85α regulatory subunit of PI3K suppresses p85 homodimerization, thus enhancing insulin-stimulated binding of p110-p85α heterodimers to IRS1 and activation of PI3K. PRR11-amplified breast cancer cells rely on PIK3CA and are highly sensitive to PI3K inhibitors, suggesting that PRR11 amplification confers PI3K dependence. Finally, genetic and pharmacological inhibition of PI3K suppresses PRR11-mediated, estrogen-independent growth. These data suggest ER⁺/PRR11-amplified breast cancers as a novel subgroup of tumors that may benefit from treatment with PI3K inhibitors and antiestrogens.

Fig. 1. PRR11 is associated with poor clinical outcome of ER⁺ breast cancers treated with antiestrogens.

a Volcano plot of genes differentially expressed in non-responding tumors compared to responding tumors. Log₂ fold change (FC) and false discovery rates (FDR) were calculated using DeSeq2 package. b Log₂ FC of 17q23 locus genes in ER⁺/HER2⁻ breast cancers treated with long-term letrozole (n = 51). Genes in 17q23 locus were selected based on the Atlas of Genetics and Cytogenetics. Red bars indicate genes with FDR < 0.05. c Recurrence-free survival in ER⁺ breast cancers, treated with long-term neoadjuvant letrozole, with low or high PRR11 mRNA levels. A PRR11 FPKM cut-off (3.93) obtained from the human protein atlas was used to divide PRR11-high (n = 43) and -low (n = 15) tumors (https://www.proteinatlas.org/ENSG00000068489-PRR11/pathology/breast+cancer). The Mantel–Cox model was used to calculate the hazard ratio (HR) and p value. d Relapse-free survival of ER⁺/HER2⁻ breast cancers, treated with endocrine therapy, with low (n = 104) or high (n = 98) PRR11 mRNA levels by the auto select best cutoff in Kaplan–Meier Plotter. HR and p were adopted from the Kaplan–Meier Plotter (http://kmplot.com/analysis/). e, f Correlation between the on-treatment percent of Ki67⁺ tumor cells and PRR11 mRNA level in breast tumors from the cohort of Giltnane et al. and Miller et al. (Pearson correlation). g Representative PRR11 immunohistochemistry (IHC) images of primary ER⁺ breast tumors. h PRR11 levels were plotted as a function of response to estrogen suppression with letrozole in trial NCT00651976. Data represent the mean ± SD (n = 91, 25, 39 for drug sensitive, intermediate and resistant group, respectively; two-tailed unpaired t-tests). Source data are provided as a Source data file.

Fig. 2. PRR11 is a key gene in 17q23 associated with endocrine resistance.

a Frequency of PRR11 amplification in MBC project, TCGA or METABRIC ER⁺ breast cancers. b Oncoprint of putative endocrine-resistant drivers in metastatic ER⁺ tumors from MBC project dataset. ER⁺ tumors classified as ‘METASTATIC DISEASE PRESENT’ were interrogated (n = 44). c Plot of disease-free survival of METABRIC ER⁺/HER2⁻ breast cancer patients treated with anti-hormone therapy as a function of PRR11 copy number [gain/amplification (n = 163) vs. diploid/deletion (n = 834)]. d Venn diagram of genes in 17q23 that significantly correlated with on-treatment Ki67 levels (Pearson r > 0.4; p < 0.05). Source data are provided as a Source data file.

Fig. 3. PRR11 overexpression confers resistance to antiestrogens.

a Lysates of MDA-MB-134VI and MDA-MB-175VII cells that had been transduced with pLX304-PRR11, -BRIP1, -TACO1, and -SMARCD2 were subjected to immunoblot analysis. b Low density monolayers of MDA-MB-134VI and MDA-MB-175VII pLX304-PRR11, -BRIP1, -TACO1, and -SMARCD2 cells were grown in estrogen-deprived condition. After 2 weeks, cell monolayers were stained with crystal violet and cell viability quantified as described in Methods. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). c MCF7 LTED and HCC1428 LTED cells were transfected with PRR11 siRNAs. Low density monolayers of cells were treated ± 1 nM estrogen (E2) for 10 days. Cell monolayers were stained with crystal violet. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). d, e MCF7 LTED and HCC1428 LTED cells were transduced with shRNA targeting the 3′ UTR of PRR11 and then, re-transduced with pLX304-GFP or pLX304-PRR11. Cell lysates were subjected to immunoblot analysis (d). Upper and lower arrows indicate exogenous and endogenous PRR11, respectively. Low density monolayers of cells shown in d were grown in absence of E2 for 10 days (e). Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). f MCF7 cells stably expressing doxycycline-inducible-PRR11 shRNA and control shRNA were injected s.c. in the dorsum of athymic ovariectomized mice supplemented with a 14-day release 17β-estradiol pellet. After 4 weeks, mice were randomized to treatment with 10 mg/kg doxycycline (i.p) for 4 weeks. Each data point represents the mean tumor volume in mm³ ± SD; n of mice per group are shown in parenthesis (two-tailed unpaired t-tests). g Fulvestrant sensitivity of ER⁺ breast cancer cell lines (n = 11, PRISM Repurposing 19Q3 dataset). Y-axis, drug sensitivity, represents relative barcode abundance following fulvestrant treatment (Pearson correlation). h, i Fulvestrant GR₅₀ were calculated using the GR metrics calculator (http://www.grcalculator.org/grcalculator/). Cell numbers on days 0 and 6 were used as the input data. MDA-MB-134VI and MDA-MB-175VII cells were stably transduced with pLX302-LacZ (control) or pLX302-PRR11 (h). MCF7 FulvR cells were transfected with control or PRR11 siRNA (i). Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). Source data are provided as a Source data file.

Fig. 4. PRR11 overexpression enhances cancer cell proliferation.

a Single sample gene set analysis was performed using a set of 125 previously reported breast cancer-related signatures. Gene sets that were differentially enriched between PRR11 high vs. low tumors (FDR < 0.01). b Complementary DNA (cDNA) of MCF7 LTED cells transfected with control or PRR11 siRNA was tested in a 84-cell cycle gene PCR array. Expression of 6 genes in the array was reduced by PRR11 siRNA transfection (FC < 0.5). Each data point represents the average of duplicate experiments. c MCF7 LTED and HCC1428 LTED cells transfected with control or PRR11 siRNA for 48 h were stained with propidium iodide and analyzed by flow cytometry. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). d MCF7 LTED and HCC1428 LTED cells transduced with shRNA targeting the 3′ UTR of PRR11 and either pLX304-GFP or pLX304-PRR11 were stained with propidium iodide and analyzed by flow cytometry. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). e MDA-MB-134VI and MDA-MB-175VII cells transduced with pLX302-LacZ or pLX302-PRR11 were grown in estrogen (E2)-deprived condition for 4 days. Cells were stained with propidium iodide and analyzed by flow cytometry. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). f HCC1428 LTED, MCF7 LTED, MCF7 TamR, and MCF7 FulvR cells were transfected with control or PRR11 siRNA for 48 h. Cell lysates were subjected to immunoblot analysis. g MDA-MB-134VI and MDA-MB-175VII cells transduced with pLX302-LacZ or pLX302-PRR11 were grown ± 1 nM E2 for 4 days. Cell lysates were subjected to immunoblot analysis. Source data are provided as a Source data file.

Fig. 5. PRR11 overexpression reduces p85 homodimers and enhances ligand-induced PI3K activation.

a Lysates of MCF7 LTED, FulvR, TamR, and HCC1428 LTED cells transfected with PRR11 or control siRNA for 48 h were subjected to immunoblot analysis. b Lysates of MDA-MB-175VII and MDA-MB-134VI cells transduced with pLX302-LacZ or pLX302-PRR11 were subjected to immunoblot analysis. c MCF7 parental, MCF7 LTED, HCC1428 parental, and HCC1428 LTED cell lysates were immunoprecipitated with PRR11 or IgG antibodies. Immune complexes were then subjected to immunoblot analysis. d MCF7 LTED cells and HEK293 cells transduced with pLenti7.3-PIK3R1-Flag and pLX302-PRR11-V5 were subjected to proximity ligation assay (PLA) with PRR11, p85α, V5 and Flag antibodies. e Schema of p85 monomers associating via a SH3-PR domain intermolecular interaction, potentially disrupted by PRR11 overexpression. f, g HEK293 cells were co-transduced with pLenti7.3-PIK3R1-Flag, -PIK3R1-HA, and pLX302-PRR11-V5 (0, 0.25, 0.5, 1 µg; f). MCF7 LTED cells that had been stably transduced with pIND-PIK3R1-HA and pLenti6.3-PIK3R1-Flag were transfected with PRR11 siRNA (0, 1, 5, 25 pM) for 48 h in presence of 2 µg/mL doxycycline (g). Cell lysates were precipitated with HA or Flag antibodies and then subjected to immunoblot analysis. h, i MCF7 LTED cells transfected with control or PRR11 siRNA (h) and MDA-MB-175VII cells transduced with pLX302-LacZ or pLX302-PRR11 (i) were serum-starved for 24 h and then treated with 100 nM insulin for 10 min. Cell lysates were prepared and immunoprecipitated with a p110α antibody or IgG. Antibody pulldowns were then subjected to immunoblot analysis. j MDA-MB-175VII cells transduced with pLX302-LacZ or pLX302-PRR11 were serum-starved for 24 h and then treated with 100 nM insulin for 10 min. MCF7 LTED and HCC1428 LTED cells transfected with control or PRR11 siRNA were treated with insulin in the same way. Cell lysates were subjected to immunoblot analysis. k HEK293 cells were co-transduced with pLenti7.3-PIK3R1-Flag and either pLX302-PRR11 wild type (WT) or pLX302-PRR11 ∆PR, a mutant lacking the PR motif. Lysates were prepared and immunoprecipitated with a Flag antibody; immune complexes were then subjected to immunoblot analysis. Source data are provided as a Source data file.

Fig. 6. PRR11 amplification is associated with hyperactivation of the PI3K pathway in ER⁺ breast cancers.

a, b Signature score of the PI3K gene set in ER⁺/HER2⁻/PIK3CA wild type breast cancers in METABRIC plotted as a function of PRR11 copy number (a: n = 503 and 135 for deletion/diploid and gain/amplification group, respectively). Signature score of IGF1 gene set in ER⁺/HER2^- breast cancers in METABRIC plotted as a function of PRR11 copy number (b: n = 1058 and 199 for deletion/diploid and gain/amplification group, respectively). Data represent the mean ± SD (two-tailed unpaired t-tests). c GSEA of mRNA expression data from ER⁺/HER2^- tumors in METABRIC (PRR11 gain/amplification vs deletion/diploid); analyses show enrichment in PI3K_AKT_MTOR_SIG and MTORC1_SIG signatures. d Connectivity scores (tau) were computed using the connectivity map (CMap) with genes significantly upregulated in ER⁺ (TCGA) breast cancers harboring PRR11 amplification vs. no amplification and ER⁺/HER2⁻ (METABRIC) breast cancers with PRR11 gain/amplification vs. PRR11 deletion/diploid. Connectivity score of 44 perturbation classes out of 171 are highlighted (tau < –95). e PIK3CA mutation frequency in ER⁺/HER2⁻ (METABRIC^: n = 1398) and ER⁺ (TCGA: n = 594) breast cancers plotted as a function of PRR11 copy number alterations vs. no alterations (two-tailed Fisher’s exact test). Source data are provided as a Source data file.

Fig. 7. PRR11-amplified breast cancer cells are dependent on the PI3K pathway.

a PIK3CA dependency scores of 57 breast cancer cell lines were plotted against PRR11 copy number (DEMETER2 V5 dataset; Pearson correlation). b Sensitivity of 27 breast cancer cell lines to pictilisib and taselisib was plotted against PRR11 copy number (PRISM Repurposing 19Q3 dataset; Pearson correlation). Y-axis shows the log 2 cell fraction as per the relative barcode abundance following drug treatment. c Pictilisib sensitivity score of 26 breast cancer cell lines in the LINCS MGH/Sanger dataset (n = 22 and 4 for non-amplification and amplification group, respectively). Data represent the mean ± SD (two-tailed unpaired t-tests). d, e Alpelisib (d) and taselisib (e) GR₅₀ of MDA-MB-134VI and MDA-MB-175VII cells transduced with pLX302-LacZ or pLX302-PRR11 were calculated using the GR metrics calculator. Cell numbers counted at day 0 and day 6 were used as the input data. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). f Low density monolayers of MDA-MB-134VI pLX302-LacZ and -PRR11 cells were grown in estrogen (E2)-free medium. After 14 days, cell monolayers were stained with crystal violet. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). g, h Low density monolayers of MDA-MB-175VII and MDA-MB-134VI cells transduced with pLX302-PRR11 or pLX302-LacZ were treated ± 1 µM alpelisib (g) or ± 1 µM taselisib (h) in estrogen-free medium. After 14 days, cell monolayers were stained with crystal violet. Data represent the mean ± SD of three replicates (two-tailed unpaired t-tests). Source data are provided as a Source data file.