Angiocrine factors modulate tumor proliferation and motility through EphA2 repression of Slit2 tumor suppressor function in endothelium

Abstract

It is well known that tumor-derived proangiogenic factors induce neovascularization to facilitate tumor growth and malignant progression. However, the concept of "angiocrine" signaling, in which signals produced by endothelial cells elicit tumor cell responses distinct from vessel function, has been proposed, yet remains underinvestigated. Here, we report that angiocrine factors secreted from endothelium regulate tumor growth and motility. We found that Slit2, which is negatively regulated by endothelial EphA2 receptor, is one such tumor suppressive angiocrine factor. Slit2 activity is elevated in EphA2-deficient endothelium. Blocking Slit activity restored angiocrine-induced tumor growth/motility, whereas elevated Slit2 impaired growth/motility. To translate our findings to human cancer, we analyzed EphA2 and Slit2 expression in human cancer. EphA2 expression inversely correlated with Slit2 in the vasculature of invasive human ductal carcinoma samples. Moreover, analysis of large breast tumor data sets revealed that Slit2 correlated positively with overall and recurrence-free survival, providing clinical validation for the tumor suppressor function for Slit2 in human breast cancer. Together, these data support a novel, clinically relevant mechanism through which EphA2 represses Slit2 expression in endothelium to facilitate angiocrine-mediated tumor growth and motility by blocking a tumor suppressive signal.

Figure 1. `Angiocrine' signals from endothelial cells enhance tumor cell growth and motility, which is diminished in the absence of endothelial EphA2 receptor function

(A) Upper panels show photomicrographs of 4T1 tumor cell spheroids cultured in control base medium (Optimem/2% FCS) versus conditioned medium (CM) from wild-type (WT) or EphA2-deficient (KO) endothelial cells (EC). Lower panels show confocal images of spheroids stained with E-cadherin (green) and Topro3 (blue) nuclear counterstain. Arrowheads indicate invasive protrusions. Colony size was quantified based on pixel area of 4 independent colonies/photomicrograph in replicate cultures from 3 to 5 independent experiments. Scale bar = 25 μm upper panels, 10 μm lower panels. (B) 4T1 tumor cell proliferation in two-dimensional culture was scored by BrdU incorporation, and 4T1 tumor cell migration was measured by transwell assay. (C) Schematic for microfluidic chamber device. Tumor cells (4T1-GFP, green) and endothelial cells (CellTracker Orange dye labeled, red) were seeded into two adjacent cell culture chambers separated by a narrow PDMS barrier. Release of the valve barrier permits cross-migration of both cell types through the central chamber. Arrows indicate direction of cross-migration through the open chamber upon barrier removal. (D) Cross-migration of tumor cells and endothelial cells was quantified simultaneously based on cell morphology and differential fluorescent labeling. Arrowheads indicate cells that migrated into the central chamber (boundaries between cell chambers and central chamber marked with dashed lines). Scale bar = 10 μm upper panels, 50 μm lower panels. Data are a representation of 3 to 5 independent experiments using conditioned medium/cells from at least three independent WT versus KO EC isolates for all cell culture experiments. Graphs display average ± standard deviation.

Figure 2. Effect of `Angiocrine' signals from endothelial cells on tumor growth in vivo

(A) Schematic for tumor cell-endothelial cell co-transplantation experimental design. 4T1 tumor cells were ad-mixed with WT or KO EC (labeled with adenovirus encoding β-galactosidase), resuspended in growth factor-reduced Matrigel, and injected subcutaneously into the dorsal flank of recipient mice. The resulting tumors were harvested 2 to 7 days post-injection for analysis. (B) Proliferation of tumor cells in proximity to exogenous EC (blue*) within the interior of the tumors for WT versus KO EC was quantified based on nuclear PCNA staining (arrowheads), 4 days post-transplantation (p <0.05), as well as proliferation in the tumor periphery, where tumor cells are proximal to host blood vessels (arrows). Right hand panels show lower magification photomicrographs, with boxed areas indicating regions shown in higher magnification panels on left. Scale bar = 10 μm (left hand panels) or 50 μm (right hand panels). (C) Tumor volume was scored over time at day 2, 4, and 7 post-transplantation. (D) Schematic for tumor cell-endothelial cell CM co-transplantation experimental design. 4T1 tumor cells were ad-mixed with 5× concentrated CM from WT or KO EC versus control base medium, resuspended in growth factor-reduced Matrigel, and injected subcutaneously into the dorsal flank of recipient mice. The resulting tumors were harvested 7 days post-injection for analysis. Tumor volume was measured at day 7. Data were consolidated from 6 to 10 independent animals/condition in at least two experiments, using conditioned medium/cells from at least three independent WT versus KO EC isolates for all in vivo studies. Graphs display average ± standard deviation.

Figure 3. Slit2 expression and activity are elevated in EphA2-deficient endothelial cells

(A) Slit2 expression in WT versus KO EC was measured by Real-Time qRT-PCR analysis. Expression was analyzed using RNA harvested from 5 independent EC isolates/genotype. Graph displays average ± standard deviation. (B) Slit2 protein was detected in EC by immunofluorescence staining. Staining for the endothelial cell marker CD31 was performed to confirm that greater than 95% of the cells in isolates were of endothelial origin. Scale bar = 20 μm. (C) To correlate expression data with Slit function, phosphorylation of Robo4 receptor in WT versus KO EC was used as a surrogate for Slit activity. The absence of EphA2 protein expression in KO EC lysates versus WT EC was confirmed by immunoblot analysis. Blots were stripped and re-probed for actin to validate uniform loading. (D) To determine if Slit activity present in EC CM affected tumor cells, phosphorylation of Robo1 receptor in 4T1 tumor cells was assessed upon stimulation with 10× CM from WT versus KO EC (plus or minus soluble Robo1-Fc) versus base medium (negative control) or recombinant Slit2 (positive control) after Robo1 immunoprecipitation. Expression of myc-tagged, recombinant Slit2 in HEK293 producer cell lysates was confirmed by immunoblot analysis.

Figure 4. Modulating Slit2 activity in endothelium affects angiocrine-mediated tumor growth and motility in the context of endothelial EphA2 receptor function

(A) Upper panels show photomicrographs of 4T1 tumor cell spheroids cultured in WT EC CM ± recombinant Slit2/control IgG or KO EC CM ± soluble Robo1-Fc/control IgG for 5 days. Lower panels show confocal images of spheroids stained with E-cadherin (green) and Topro3 (blue) nuclear counterstain. Arrowheads indicate invasive protrusions. Scale bar = 25 μm upper panels, 10 μm lower panels. Colony size was quantified based on pixel area of 4 independent colonies/photomicrograph in replicate cultures from 3 to 5 independent experiments. Graph displays average ± standard deviation. (B) 4T1 tumor cell proliferation was quantified in two-dimensional culture by BrdU incorporation, and 4T1 tumor cell migration was measured by transwell assay. (C) Cross-migration of tumor cells and endothelial cells was quantified simultaneously based on cell morphology and differential fluorescent labeling in microfluidic chamber devices. Data are a representation of 3 to 5 independent experiments using conditioned medium from at least three independent WT versus KO EC isolates for all cell culture experiments. Graphs display average ± standard deviation. (D) 4T1 tumor cells were admixed with 5× concentrated WT EC CM ± Slit2/control IgG or KO EC CM ± Robo1-Fc/control IgG, resuspended in growth factor-reduced Matrigel, and injected subcutaneously into the dorsal flank of recipient mice. Tumor dimensions were measured at 1, 3, and 7 days post-injection for comparison of tumor volume over time, and harvested on day 7 for analysis. Data were consolidated from 6 to 10 independent animals/condition in at least two experiments for in vivo studies. Graph displays average ± standard deviation.

Figure 5. EphA2 and Slit2 expression profiles in human breast cancer

(A) EphA2 and Slit2 protein expression in human invasive ductal carcinoma samples was analyzed in breast cancer tissue microarrays (TMA; Cybrdi, Inc.). Representative photographs of EphA2-negative /Slit2-positive tumor blood vessels (left), and of EphA2-positive/Slit2-negative (right) tumor blood vessels, are shown. Arrowheads indicate tumor blood vessels. 17 out of 53 samples harbored EphA2-positive/Slit2-positive tumor vessels, compared with 36 out of 53 samples with tumor vessels that were EphA2-postive/Slit2-negative (p = 0.009 Chi Square Test). Scale bar = 50 μm upper panels, 10 μm lower panels. (B) Controls for EphA2 antibody specificity. MMTV-Neu tumor tissue sections from EphA2 WT and KO animals were stained with anti-EphA2 antibodies. Slit2 protein was detected in tumor parenchyma of normal/hyperplastic human breast tissue epithelium, but not in tumor epithelium from Stage II invasive breast cancer, in TMA samples. Scale bar = 50 μm. (C) Kaplan-Meier kinetic analyses of the van der Vijver dataset, with microarray profiles of 295 human breast tumors and associated clinical data. The impact of elevated slit2 expression on overall survival and recurrence-free survival was analyzed by Log-Rank tests.