Arsenic Alters Exosome Quantity and Cargo to Mediate Stem Cell Recruitment Into a Cancer Stem Cell-Like Phenotype

Abstract

Inorganic arsenic is a human carcinogen that can target the prostate. Accumulating evidence suggests arsenic can disrupt stem cell (SC) dynamics during the carcinogenic process. Previous work demonstrated arsenic-transformed prostate epithelial (CAsE-PE) cells can recruit prostate SCs into rapidly acquiring a cancer SC (CSC) phenotype via the secretion of soluble factors. Exosomes are small, membrane-derived vesicles that contain lipids, RNA, and proteins, and actively contribute to cancer initiation and progression when taken up by target cells. Here we hypothesized that CAsE-PE cells are recruiting SCs to a CSC-like phenotype via exosomal signaling. CAsE-PE cells secreted 700% more exosomes than parental RWPE-1 cells. CAsE-PE exosomes were enriched with oncogenic factors, including oncogenes (KRAS, NRAS, VEFGA, MYB, and EGFR), inflammation-related (cyclooxygenase-2, interleukin 1B (IL1B), IL6, transforming growth factor-β, and tumor necrosis factor-A), and apoptosis-related (CASP7, CASP9, and BCL2) transcripts, and oncogenesis-associated microRNAs. When compared with SCs cultured in exosome-depleted conditioned medium (CM), SCs cultured in CM containing CAsE-PE-derived exosomes showed increased (198%) matrix metalloproteinase activity and underwent an epithelial-to-mesenchymal transition in morphology, suggesting an exosome-mediated transformation. KRAS plays an important role in arsenic carcinogenesis. Although KRAS transcript (>24 000%) and protein (866%) levels were elevated in CAsE-PE exosomes, knock-down of KRAS in these cells only partially mitigated the CSC-like phenotype in cocultured SCs. Collectively, these results suggest arsenic impacts both exosomal quantity and cargo. Exosomal KRAS is only minimally involved in this recruitment, and additional factors (eg, cancer-associated miRNAs) likely also play a role. This work furthers our mechanistic understanding of how arsenic disrupts SC dynamics and influences the tumor microenvironment during carcinogenesis.

Figure 1.

EVs play a role in recruiting SCs to an oncogenic phenotype. A, Secreted MMP-2 activity of SCs grown in CAsE-PE CM containing EVs or devoid of EVs for 3 weeks. Asterisk denotes statistical significance (one factor ANOVA, p = .02, n = 3). Post-hoc comparisons were made using Tukey’s HSD test. Bars ± SEM. B, Morphology of SCs grown in CAsE-PE CM for 3 weeks (scale bars = 200 µm).

Figure 2.

Arsenite transformed CAsE-PE cells secrete more exosomes than RWPE-1 cells. A, Nanoparticle tracking analysis of isolated exosomes. B, Western blot analysis of exosomal markers CD9 and CD81 and nonexosomal markers GRP94, CANX, and HSP70. C, CAsE-PE cells secrete 702% more exosomes than RWPE-1 cells (1 factor ANOVA, p = .0002, n = 3, Bars ± SEM).

Figure 3.

Gene and protein expression in exosomes. mRNA expression of (A) inflammation-related genes, (B) apoptosis-related genes, and (C) oncogenes in isolated exosomes. (D) Western blot analysis of KRAS protein levels in exosomes and cell lysate. Asterisk denotes statistical significance (1 factor ANOVA, p < .05, n = 3). Bars ± SEM.

Figure 4.

A, Western blot analysis of KRAS protein levels in cell lysate and exosomes following KRAS KD. B, Western blot analysis of KRAS, BCL-XL, and BCL2 protein levels in SCs following 3 weeks of coculture with KRAS-KD CAsE-PE cells. C, Secreted MMP-2 (2 factor ANOVA, time (p = .34), KD (p = .23), interaction (p = .34)), and (D) MMP-9 activity (2 factor ANOVA, time (p = .38), KD (p = .013), interaction (p = .38)) in cocultured SCs. E, Invasive capacity (one factor ANOVA, p = .02) and (F) colony formation in soft agar (1 factor ANOVA, p = .8) of SCs after 3 weeks of coculture. Asterisk denotes statistical significance (p < .05, n = 3–4). Results are mean ± SEM.