Elevated AKR1C3 expression promotes prostate cancer cell survival and prostate cell-mediated endothelial cell tube formation: implications for prostate cancer progression

Abstract

Background: Aldo-keto reductase (AKR) 1C family member 3 (AKR1C3), one of four identified human AKR1C enzymes, catalyzes steroid, prostaglandin, and xenobiotic metabolism. In the prostate, AKR1C3 is up-regulated in localized and advanced prostate adenocarcinoma, and is associated with prostate cancer (PCa) aggressiveness. Here we propose a novel pathological function of AKR1C3 in tumor angiogenesis and its potential role in promoting PCa progression.

Methods: To recapitulate elevated AKR1C3 expression in cancerous prostate, the human PCa PC-3 cell line was stably transfected with an AKR1C3 expression construct to establish PC3-AKR1C3 transfectants. Microarray and bioinformatics analysis were performed to identify AKR1C3-mediated pathways of activation and their potential biological consequences in PC-3 cells. Western blot analysis, reverse transcription-polymerase chain reaction (RT-PCR), enzyme-linked immunosorbent assay (ELISA), and an in vitro Matrigel angiogenesis assays were applied to validate the pro-angiogenic activity of PC3-AKR1C3 transfectants identified by bioinformatics analysis.

Results: Microarray and bioinformatics analysis suggested that overexpression of AKR1C3 in PC-3 cells modulates estrogen and androgen metabolism, activates insulin-like growth factor (IGF)-1 and Akt signaling pathways, as well as promotes tumor angiogenesis and aggressiveness. Levels of IGF-1 receptor (IGF-1R) and Akt activation as well as vascular endothelial growth factor (VEGF) expression and secretion were significantly elevated in PC3-AKR1C3 transfectants in comparison to PC3-mock transfectants. PC3-AKR1C3 transfectants also promoted endothelial cell (EC) tube formation on Matrigel as compared to the AKR1C3-negative parental PC-3 cells and PC3-mock transfectants. Pre-treatment of PC3-AKR1C3 transfectants with a selective IGF-1R kinase inhibitor (AG1024) or a non-selective phosphoinositide 3-kinases (PI3K) inhibitor (LY294002) abolished ability of the cells to promote EC tube formation.

Conclusions: Bioinformatics analysis followed by functional genomics demonstrated that AKR1C3 overexpression promotes angiogenesis and aggressiveness of PC-3 cells. These results also suggest that AKR1C3-mediated tumor angiogenesis is regulated by estrogen and androgen metabolism with subsequent IGF-1R and Akt activation followed by VEGF expression in PCa cells.

Figure 1

Influence of steroid hormone signaling in PC3-AKR1C3 transfectants. (A) Genes whose expression is altered in PC3-AKR1C3 transfectants were identified to be regulated by 17β-estradiol, androstenediol, or 5α-DHT. Genes that are labeled in red and green represent up- and down-regulated genes, respectively, in PC3-AKR1C3 transfectants as compared to PC3-mock transfectants. The fold changes are reflected by corresponding numbers under the gene names. (B) FST expression can be regulated by AKR1C3-mediated testosterone and/or progesterone metabolism in PC3-AKR1C3 transfectants. The legend for the gene shapes is at lower right corner.

Figure 2

Activation of the IGF-1 pathway in PC3-AKR1C3 transfectants. (A) Ingenuity analysis identified that activation of the IGF-1 signaling pathway is statistically significant in PC3-AKR1C3 transfectants. Up-regulated genes, or focus genes, are marked by in filled gray; and uncolored genes indicate they are expressed in the microarray datasets under any conditions and may participate in signal transduction. Branches of the IGF-1 pathway that did not have at least one focus gene were not included in the diagram. (B) Western blot analysis of total and phosphorylated IGF-1R β and Akt in PC3-AKR1C3 stable transfectants. The analysis was performed 3 times; and all experiments showed consistent elevated phosphorylation of IGF-1R β (Tyr 1131) and Akt (Ser 473) in PC3-AKR1C3 transfectants. Image analysis confirmed that levels of phosphorylated IGF-1R β and Akt are statistically higher in PC3-AKR1C3 transfectants as compared to PC3-mock transfectants.

Figure 3

Phenotypic presentation of PC3-AKR1C3 transfectants from literature analysis. Relationships between differentially expressed genes in PC3-AKR1C3 transfectants and the most significantly overrepresented keywords published with those genes in MEDLINE abstracts and titles are identified. Genes are circled, keywords are not circled. Thickness of the line correlates with the mutual information between the terms (thicker lines mean more mutual information). (A) Up-regulated genes that have been documented as related to aggressive carcinomas, PSA, β-catenin, and ER. (B) Down-regulated genes that are related to angiogenesis, tumor suppression and, more specifically, tumor suppression via p53.

Figure 4

Identification of TREs for differentially regulated genes identified in PC3-AKR1C3 transfectants. (A) TREs shared by genes that are up-regulated in PC3-AKR1C3 transfectants. (B) TREs shared by genes that are down-regulated in PC3-AKR1C3 transfectants. Red color indicates significant TREs with P < 0.05 and FDR < 0.3.

Figure 5

Levels of VEGF mRNA expression in PC3-AKR1C3 transfectants. Total RNA was isolated from PC3-mock and PC3-AKR1C3 C1 transfectants as well as from PC3-AKR1C3 C1 clone treated with either 20 µM AG1024 or 10 µM LY294002. Target VEGF A, B, C, and D mRNA species were PCR amplified using isofrom-specific primer pairs. PCR products were separated by 1% agarose gel electrophoresis; and images of ethidium bromide-stained gels were acquired.

Figure 6

Angiogenic properties of PC-3 cells overexpressing AKR1C3 on Matrigel. (A) PC-3 cell-mediated HMEC-1 tube formation was performed in a co-culture system on Matrigel basement membrane matrix. Formation of HMEC-1 tubes was imaged at 24 hours after HMEC-1 co-culture with either PC3-mock or PC3-AKR1C3 transfectants. (B) Quantification of HMEC-1 tube formation. The number of EC tubes were counted; and results were compared between PC3-mock and PC3-AKR1C3 transfectants and presented as mean ± SEM from 12 experiments for the 4 independent clones. The number of HMEC-1 tubes formed were also determined by pre-treating PC-3 transfectants with 20 µM AG1042 or 10 µM LY294002; and results were presented as mean ± SEM from at least 3 independent assays. * indicates P < 0.05 between PC3-mock and PC3-AKR1C3 transfectants.