The retinoblastoma protein regulates hypoxia-inducible genetic programs, tumor cell invasiveness and neuroendocrine differentiation in prostate cancer cells

Abstract

Loss of tumor suppressor proteins, such as the retinoblastoma protein (Rb), results in tumor progression and metastasis. Metastasis is facilitated by low oxygen availability within the tumor that is detected by hypoxia inducible factors (HIFs). The HIF1 complex, HIF1α and dimerization partner the aryl hydrocarbon receptor nuclear translocator (ARNT), is the master regulator of the hypoxic response. Previously, we demonstrated that Rb represses the transcriptional response to hypoxia by virtue of its association with HIF1. In this report, we further characterized the role Rb plays in mediating hypoxia-regulated genetic programs by stably ablating Rb expression with retrovirally-introduced short hairpin RNA in LNCaP and 22Rv1 human prostate cancer cells. DNA microarray analysis revealed that loss of Rb in conjunction with hypoxia leads to aberrant expression of hypoxia-regulated genetic programs that increase cell invasion and promote neuroendocrine differentiation. For the first time, we have established a direct link between hypoxic tumor environments, Rb inactivation and progression to late stage metastatic neuroendocrine prostate cancer. Understanding the molecular pathways responsible for progression of benign prostate tumors to metastasized and lethal forms will aid in the development of more effective prostate cancer therapies.

Keywords: hypoxia; invasion; neuroendocrine differentiation; prostate cancer; retinoblastoma protein.

Figure 1. Ablation of Rb leads to transcriptional dysregulation of HIF1-target genes involved in metastasis and angiogenesis

(A) LNCaP cells stably infected with either a shRNA to Rb (shRb) or a scrambled negative control shRNA (shSCX) were maintained in normoxic conditions or at 1% O₂ for 24 h then gene expression was determined by quantitative real-time PCR after isolation and reverse transcription of total RNA. Rb expression was normalized to the constitutively active 36B4 gene expression. (B) Immunoblot analysis of whole cell extracts from shSCX and shRb LNCaP cells using anti-Rb or anti-α-tubulin primary antibodies. α-tubulin is the loading control. (C) CXCR4 and (D) VEGF mRNA accumulation was determined by RT-PCR after shSCX and shRb LNCaP cells were treated as described in (A). Open bars represent normoxia (20% O₂) and closed black bars represent hypoxia (1% O₂). Error bars represent ± S.D. and statistical significance was determined using a one-way ANOVA (*p < 0.05).

Figure 2. Hypoxia-inducible increase in invasion but not cell cycle or proliferation in LNCaP prostate cancer cells lacking Rb

(A) shRNA LNCaP cells (1 × 10⁴) were seeded in Matrigel invasion chambers and then maintained in normoxic conditions or at 1% O₂ for 36 h. Chambers were then prepared according to manufacturers protocols and cells were counted under a microscope. Assays were performed in triplicate. Error bars represent ± S.D. and statistical significance was determined using a one-way ANOVA (*p < 0.05). (B) Knockdown of Rb in LNCaP cells does not alter cell proliferation in response to hypoxia. Cells were either left at normoxia or treated with 1% O₂ and cells were counted at 0, 12, 24, 36, 48, and 72 h later. Error bars represent ± S.E.M. and statistical significance was determined using a one-way ANOVA (*p < 0.05). (C) Knock-down of Rb in LNCaP cells does not alter cell cycle in response to hypoxia. Cell cycle status was determined by propidium iodide (PI) staining and flow cytometry. LNCaP cells with a scrambled negative control or with Rb ablated, were treated with hypoxia or left at normoxic conditions for 36-hours. The percentage of cells in each stage of the cell cycle was determined using FlowJo analysis software based on the PI staining profile of FSC/SSC-gated population. Assay was performed three times and each sample was read in triplicate. Error bars represent ± S.E.M.

Figure 3. The role of Rb in hypoxia-regulated transcription

LNCaP cells infected with either a short-hairpin control RNA (shSCX) or a short-hairpin to Rb (shRb) were maintained under hypoxic (1% O₂) or normoxic (20% O₂) conditions for 24 h. Extracted RNA was subjected to microarray analysis. Experiments were performed in triplicate and only genes that were up- or down-regulated at least 2-fold under hypoxic conditions with a p-value < 0.05 were considered significant. (A) Venn diagram of up-regulated genes, showing the overlap between hypoxia inducible genes, genes sensitive to loss of Rb and up-regulated genes sensitive to Rb-loss and hypoxia in combination. Olive shaded area represents genes that are from the shRb-hypoxia-treated data set that were up-regulated significantly (p < 0.05) at least 2.0 fold when compared to the other treatments. Blue shaded area represents genes that are hypoxia-inducible. Red shaded area represents genes that are up-regulated by Rb-loss. (B) A pie chart was used to illustrate the percentage of genes on each chromosome that were up- or down-regulated in Rb knockdown LNCaP cells. (C) Genes regulated by the Rb-HIF1 complex are in close proximity to one another at certain loci on select chromosomes. Chromosome maps of all the up- and down-regulated genes on chromosomes 1 and 21 according to their annotated start and stop sites are represented to scale.

Figure 4. Rb-loss leads to transcriptional dysregulation of hypoxia-regulated genes involved in metastasis, angiogenesis and neuroendocrine differentiation

(A) Heat map of the 50 genes displaying the largest increase in fold gene expression in response to hypoxia after knock-down of Rb. (B) The top systems and diseases identified by Ingenuity Pathway Assist (IPA) analysis. The top 12 systems and diseases associated with the 50 most up-regulated and 50 most down-regulated hypoxia sensitive genes after knock-down of Rb. The –log of the p-value is represented on the ordinate. (C) Nervous system development and function network identified by IPA analysis. Nodes identified are indicated with red circles. Up-regulated genes identified in our micro-array screen are filled with varying shades of pink and red with the most highly expressing genes shaded red and the lower expressing genes shaded pink.

Figure 5. Confirmation of Rb-sensitive and hypoxia-inducible neuroendocrine targets identified through microarray analysis

(A) LNCaP or (B) 22Rv1 cells infected with either a short-hairpin control RNA (shSCX) or a short-hairpin to Rb (shRb) were maintained under hypoxic (1% O₂) or normoxic (20% O₂) conditions for 24 h. Extracted RNA was subjected to qRT-PCR. Transcriptional responses of ENO2, KISS1R, HTR5A and PLOD2 are displayed and expression was normalized to the constitutively active 36B4 gene expression. Open bars represent normoxia (20% O₂) and closed black bars represent hypoxia (1% O₂). Error bars represent ± S.D. and statistical significance was determined using a one-way ANOVA (*p < 0.05).

Figure 6. Loss of Rb results in increased expression of proteins involved in metastasis and neuroendocrine differentiation in prostate cancer cells in a hypoxia-dependent fashion

(A) LNCaP and (B) 22Rv1 shRNA cells were exposed to normal O₂ levels or 1% O₂ for various times up to 7 days as indicated. Whole cell lysates were fractionated by SDS-PAGE and examined by immunoblotting using affinity purified antibodies to α–tubulin, Rb, AR, ENO2, KISS1R, RORα, PLOD2, NDRG1 and HTR5A.

Figure 7. Kisspeptin-10 activates calcium signaling in 22Rv1 cells

Application of 1 uM kisspeptin-10 peptide in (A) LNCaP or (B) 22rV1 cells lacking Rb (shRb) and subjected to 96 hours of hypoxia led to a significant increase in intracellular calcium levels compared to control scrambled (shSCX) hypoxia cells and to normoxia cells (both shRb and shSCX) in 22Rv1 cells only. After hypoxia or normoxia treatment, cells were loaded with 2 uM indo-1AM dye (30 min at 37°C) and then imaged on an inverted fluorescent microscope. Application of 1 uM kisspeptin-10 (indicated by red arrows) led to a significant increase intracellular calcium levels in 22Rv1 cells as indexed by ratiometric fluorescence measurements taken at 405 and 485 nm. A one-way ANOVA was performed followed by a post-hoc Tukey test. *p < 0.05, **p < 0.01 vs 1:20 time point, n = 5.