PRR15 deficiency facilitates malignant progression by mediating PI3K/Akt signaling and predicts clinical prognosis in triple-negative rather than non-triple-negative breast cancer

Abstract

Triple-negative breast cancer (TNBC) is the most aggressive subtype of breast neoplasms with a higher risk of recurrence and metastasis than non-TNBC. Nevertheless, the factors responsible for the differences in the malignant behavior between TNBC and non-TNBC are not fully explored. Proline rich 15 (PRR15) is a protein involved in the progression of several tumor types, but its mechanisms are still controversial. Therefore, this study aimed to investigate the biological role and clinical applications of PRR15 on TNBC. PRR15 gene was differentially expressed between TNBC and non-TNBC patients, previously described as an oncogenic factor in breast cancer. However, our results showed a decreased expression of PRR15 that portended a favorable prognosis in TNBC rather than non-TNBC. PRR15 knockdown facilitated the proliferation, migration, and invasive ability of TNBC cells in vitro and in vivo, which was abolished by PRR15 restoration, without remarkable effects on non-TNBC. High-throughput drug sensitivity revealed that PI3K/Akt signaling was involved in the aggressive properties of PRR15 silencing, which was confirmed by the PI3K/Akt signaling activation in the tumors of PRR15^Low patients, and PI3K inhibitor reversed the metastatic capacity of TNBC in mice. The reduced PRR15 expression in TNBC patients was positively correlated with more aggressive clinicopathological characteristics, enhanced metastasis, and poor disease-free survival. Collectively, PRR15 down-regulation promotes malignant progression through the PI3K/Akt signaling in TNBC rather than in non-TNBC, affects the response of TNBC cells to antitumor agents, and is a promising indicator of disease outcomes in TNBC.

Fig. 1. PRR15 expression is decreased in TNBC but increased in non-TNBC, and predicts a more favorable prognosis in the former.

a DEGs in TNBC (left panel) and non-TNBC (right panel) relative to normal breast tissues. Data analyzed from TCGA and GTEx database, P value <0.05, |Log₂-FC | > 2. b The overlap of the up-regulated genes (left panel) and down-regulated genes (right panel) of TNBC and non-TNBC obtained in Fig. 1a. c Transcriptomic PRR15 expression in the overall breast cancer and normal samples from TCGA database. d PRR15 expression in distinct molecular subtypes of breast carcinoma versus normal mammary epithelium. e PRR15 expression in different subnets of breast neoplasia identified by single-cell transcriptomics archived in the Single Cell Portal project. Left panel, quantity and distribution of subpopulations of different cell types; right panel, expression of PRR15. f RT-qPCR of PRR15 in a panel of cell lines. g PRR15 expression in luminal (n = 26) and TNBC/basal (n = 27) breast neoplasm cell lines available in CCLE. h PRR15 expression examined by IHC in breast cancer tissue microarrays, which included 128 cases of primary breast cancer lesions, including 24 cases of TNBC and 100 cases of non-TNBC (35 cases of luminal A, 54 cases of luminal B, 7 cases of HER2amp, and 4 cases unknown). i, j Statistical analysis of the differences in PRR15 expression between TNBC and non-TNBC (i) and among four molecular subsets of breast cancer (j) in tissue microarrays from Fig. 1h. k Effect of PRR15 expression on OS in TNBC as shown by Kaplan–Meier survival curve using the TCGA data. High and low PRR15 expression was discriminated using a 50% (median) cut-off value. l Differential PRR15 mRNA expression between multiple tumors and corresponding normal samples determined using the RNA-seq data from TCGA in combination with GTEx consortium. Data are presented as mean ± SEM (c, d, f, g) or median and interquartile range (i, j, l), and analyzed by unpaired t-test (c, d, f, g) or Wilcoxon rank-sum test (i, j, l). ^*P < 0.05, ^**P < 0.01,^***P < 0.001. Scale bars: h left 500 μm, right 200 μm. DEGs differentially expressed genes, TCGA The Cancer Genome Atlas, GTEx Genotype-Tissue Expression, TNBC triple-negative breast cancer, CCLE Cancer Cell Line Encyclopedia, RT-qPCR quantitative reverse transcription polymerase chain reaction, IHC immunohistochemistry, OS overall survival.

Fig. 2. PRR15 silencing is required for TNBC initiation and progression in vitro and in vivo.

a Proliferation of MDA-MB-231 with stable knockdown of PRR15 and scramble control monitored by the IncuCyte system. b Migration ability of PRR15-silenced MDA-MB-231 compared to control cells assessed by wound-healing scratch assays. c Effect of PRR15 knockdown on the invasive ability of MDA-MB-231 investigated using the transwell invasion assays. d Comparison of the proliferative ability of MDA-MB-231 stably overexpressing PRR15 with control cells by IncuCyte platform. e Cell motility of MDA-MB-231 with ectopic PRR15 and counterparts determined through wound-healing scratch assays. f Effect of PRR15 overexpression on the invasive ability of MDA-MB-231 examined using the transwell invasion assays. g Cell proliferation of PRR15-silenced MDA-MB-231 with or without PRR15 restoration, as well as that of their control cells assessed using Incucyte technology. h Comparison of the scratch healing rate in control, PRR15-knockdown, and PRR15-restored MDA-MB-231. i Analysis of the invasive potential in the control, PRR15-knockdown, and PRR15-restored MDA-MB-231 using the transwell invasion assay. j–q Representative gross images (j, n), volume (k, o), and weight (l, p) of xenograft tumors formed by PRR15-silenced MDA-MB-231 and CAL51 and their controls subcutaneously implanted into Balb/c nude mice (n = 4–5), along with the weight of tumor-bearing mice (m, q) described in Figs. 2j, n. r Representative bright-field (left panel) and H&E-stained (right panel) images of lungs from Balb/c-nude mice after the intravenous injection of PRR15-silenced MDA-MB-231 and counterparts (n = 7). s–u Number of metastatic nodules (s), lung weight (t), and body weight (u) of mice carrying lung metastatic lesions. Scale bars: b 200 μm, c 100 μm, e 200 μm, f 100 μm, h 200 μm, i 200 μm; j, n, 10 mm, r left 5 mm, right 2 mm. Data are presented as mean ± SEM, and analyzed by unpaired t-test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. H&E hematoxylin and eosin.

Fig. 3. Screening of more potent antineoplastic drugs effective on TNBC cells with decreased PRR15 expression using high-throughput drug sensitivity testing.

a Schematic diagram of a high-throughput drug screen using a library of 397 marketed anti-cancer drugs. MDA-MB-231 cells with or without PRR15 impairment treated with the indicated compounds for 72 h and cell viability measured by CCK8 assay. b Composition of the pathways targeted by the agents included in the library of antineoplastics. c Changes in IC50 values of the tested drugs on MDA-MB-231 cells after PRR15 knockdown. d Distribution of the molecular pathways targeted by the drugs with altered IC50 values in the Fig. 3c. e, f Representative antineoplastic agents with enhanced (i.e., reduced IC50, e) and attenuated (i.e., increased IC50, f) inhibitory effects on PRR15-silenced MDA-MB-231 cells. IC50 half-maximal inhibitory concentration.

Fig. 4. PI3K/Akt pathway is a pivotal mechanism by which PRR15 deficiency promotes the progression of TNBC.

a, b Significant GO clusters for the CC categories (a) and KEGG pathways (b) enriched by the DEGs between PRR15-knockdown MDA-MB-231 cells and control cells. c Western blot analysis of the expression of the key factors of the PI3K/Akt signaling in MDA-MB-231 (top panel) and CAL51 (bottom panel) cells with PRR15 knockdown and overexpression. d Schematic representation of the study design and timeline of LY194002 treatment in mice receiving an injection of PRR15-silenced MDA-MB-231 and control cells into the tail vein. e–g Representative gross (e, left panel) and histopathological (e, right panel) images of the lungs of the mice mentioned in Fig. 4d, as well as the number of metastatic foci (f) and lung weight (g). h–j Representative images of the expression of PRR15, p-PI3K, and p-Akt in the tumor mass of TNBC patients with high and low expression of PRR15 detected by IHC (h), and statistics of IHC staining scores of p-PI3K (i) and p-Akt (j). k Co-expression heat map of PRR15 and critical molecules of EMT signaling for BRCA data analysis archived by TCGA. l Scatter plots of the correlation between PRR15 expression and the expression of the mesenchymal markers vimentin (VIM, left panel) and SNAI1 (right panel). m Transwell migration analysis of the indicated MDA-MB-231 cells with different expression levels of PRR15. n, o Relative expression of vimentin (VIM) and SNAI1 in CAL51 cells with PRR15 knockdown (n) and PRR15 overexpression (o) along with the corresponding control cells evaluated by RT-qPCR. p Western blot analysis of EMT markers in MDA-MB-231 cell with silencing or overexpression of PRR15 and their controls. Data are presented as mean ± SEM (f, g, m–o) or median and interquartile range (i, j), and analyzed by unpaired t-test (f, g, m–o), Wilcoxon rank-sum test (i, j), or Spearman correlation coefficients (k, l). ^*P < 0.05, ^***P < 0.001. Scale bars: e, left 2.5 mm, right 2 mm; h, 200 μm; m, 100 μm. GO Gene Ontology, CC Cellular Component, KEGG Kyoto Encyclopedia of Genes and Genomes, DEGs differentially expressed genes, IHC immunohistochemistry, EMT epithelial-mesenchymal transition, BRCA breast invasive carcinoma, TCGA The Cancer Genome Atlas, RT-qPCR quantitative reverse transcription polymerase chain reaction.

Fig. 5. PRR15 expression acts as a robust predictor of the disease outcome in TNBC patients.

a Comparison of clinical staging in breast cancer patients with high and low PRR15 expression using the TCGA-BRCA dataset. b IHC staining score of PRR15 in TNBC patients with or without metastases in the breast cancer cohort (cases corresponding to tissue microarray). c, d Statistics of the number of metastatic sites in the PRR15^High and PRR15^Low patients in our TNBC cohort (c) together with the profile of metastatic sites (d). e–h Kaplan–Meier survival curves for DFS in the overall TNBC cohort (e), for DFS (f) and OS (g) in the subgroup of TNBC patients receiving EC-P treatment, and for DFS in the subgroup of TNBC patients with high Ki-67 index (>30, h). Data are presented as percentage of the total (a, c, d) or median and interquartile range (b), and analyzed by χ2 test (a), rank-sum test (b, c), or log-rank test (e–h). TCGA The Cancer Genome Atlas, BRCA breast invasive carcinoma, IHC immunohistochemistry, TNBC triple-negative breast cancer, DFS disease-free survival, OS overall survival, EC-P epirubicin and cyclophosphamide followed by paclitaxel.