Endothelial TRPV4 channels prevent tumor growth and metastasis via modulation of tumor angiogenesis and vascular integrity

Abstract

Transient receptor potential vanilloid 4 (TRPV4) is a ubiquitously expressed polymodally activated ion channel. TRPV4 has been implicated in tumor progression; however, the cell-specific role of TRPV4 in tumor growth, angiogenesis and metastasis is unknown. Here, we generated endothelial-specific TRPV4 knockout (TRPV4^ECKO) mice by crossing TRPV4 ^lox/lox mice with Tie2-Cre mice. Tumor growth and metastasis were significantly increased in a syngeneic Lewis lung carcinoma tumor model of TRPV4^ECKO mice compared to TRPV4l^ox/lox mice. Multiphoton microscopy, dextran leakage and immunohistochemical analysis revealed increased tumor angiogenesis and metastasis that were correlated with aberrant leaky vessels (increased width and reduced pericyte and VE-cadherin coverage). Mechanistically, increases in VEGFR2, p-ERK and MMP-9 expression and DQ gelatinase activity were observed in the TRPV4^ECKO mouse tumors. Our results demonstrated that endothelial TRPV4 is a critical modulator of vascular integrity and tumor angiogenesis and that deletion of TRPV4 promotes tumor angiogenesis, growth, and metastasis.

1. Introduction

TRPV4 is a polymodal cation channel activated by hypotonicity, temperature, shear stress, arachidonic acid, epoxyeicosatrienoic acid (EET), matrix stiffness and mechanical stretch in various cell types[1–9]. Recently, TRPV4 has been implicated in the growth and progression of breast, gastric, colon, and melanoma cancers[10–15]. In contrast to these findings, Peters et al. demonstrated that activation of TRPV4 induced oncosis and apoptosis of breast cancer cells in vitro and reduced tumor growth in vivo[16]. These findings suggest that either activation or inhibition of TRPV4 in tumor cells can differentially regulate tumor growth and metastasis of different tumors.

In addition to tumor cells, TRPV4 is expressed in cells of the tumor microenvironment. Notably, the mechanosensitive role of TRPV4 in endothelial cells (ECs) has attracted increased attention. In breast tumor-derived ECs but not in normal human ECs (HMVECs), TRPV4 was shown to regulate arachidonic acid-induced migration[17], implicating it in tumor angiogenesis. However, this study did not measure angiogenesis in vitro or in vivo. We recently demonstrated enhanced tumor growth in vivo in global TRPV4 knockout (KO) mice[1]. Nevertheless, it is not clear whether the enhanced tumor growth in the global TRPV4KO mice is due to the effects of TRPV4 on tumor cells themselves, tumor-associated ECs or both types of cells. To demonstrate the cell-specific role of TRPV4, we generated endothelial-specific TRPV4KO mice and investigated its role in tumor angiogenesis, growth, and metastasis.

2. Materials and Methods

2.1. Generation of endothelial-specific TRPV4 knockout mice

All animal (C57BL/6J background) experiments were approved by the Institutional Animal Care and Use Committees of Northeast Ohio Medical University (NEOMED). EC-specific TRPV4 knockout mice (TRPV4^f/f-Tie2-Cre; TRPV4^ECKO) were generated by a two-step process: in the first generation, TRPV4^lox/lox mice[18] were crossed with Tie2-Cre mice, and the resulting heterozygous (Cre/Loxp^het) mice were again crossed to generate homozygous TRPV4^{f/f Tie2-Cre} mice. Littermates without Tie2–Cre were employed as controls. Genotyping was performed by PCR using ear punch/tail snip biopsies. Endothelial deletion of TRPV4 was further confirmed by measuring TRPV4 expression (RT-PCR) and calcium influx (calcium imaging) in primary ECs isolated from aortic explants.

2.2. Cell culture

Lewis lung carcinoma cells (LLCs) were cultured as described previously [2, 5]. Isolated ECs were characterized and cultured as previously described[19,1,20,21]. Briefly, ECs were isolated from the aorta in a two-step positive isolation process that utilized anti-CD31 antibody conjugated to magnetic beads and then anti-ICAM2 conjugated beads[22]. The EC phenotype was characterized by measuring a) the endothelial markers CD31 and VEGFR2, b) uptake of acetylated LDL and c) tube formation (2D angiogenesis) on Matrigel (Suppl. Fig. 1). Cells from passages 2–4 were employed to carry out all the experiments.

2.3. Syngeneic tumor model

LLCs injection and tumor growth assessment were performed as described previously[1,3].

2.4. RT-PCR and qPCR

RNA isolation, cDNA synthesis, RT-PCR, and qPCR were performed as previously described[23]. The following real-time mouse-specific primers were obtained from IDT Technologies (USA):

TRPV4 (F-5’-CAAGTACCCCGTGGTCTTCAT −3’; R: 5’-GTTCAACAGGAGCACGAAGG −3’),

VEGFR2 (F-5’-CTACAAGTGCTCGTACCGGG −3’; R-5’-TCGGTGATGTACACGATGCC-3’),

VEGF (F-5’-CAAACCTCACCAAAGCCAGC −3’; R-5’-GCGCTTTCGTTTTTGACCCT −3’),

MMP-9 (F −5’-GGTCTTCCCCAAAGACCTGAAA −3’; R-5’-AGAGACTGCTTCTCTCCCATCA −3’),

MMP-2 (F-5’-AACGGTCGGGAATACAGCAG −3’; R-5’-TGGTAAACAAGGCTTCATGGGG −3’),

GAPDH (F-5’-GGGTCCCAGCTTAGGTTCATC −3’, R-5’-ATCCGTTCACACCGACCTTC −3’) and

TIMP-2 (F-5’-AGAAGGAGGTGGATTCCGGG-3’, R-5’-AGAGGGGGCCGTGTAGATAAA-3’).

2.5. Immunohistochemistry for OCT-embedded tissues

Aorta, lungs, liver, and retina were extracted from the TRPV4^lox/lox and TRPV4^ECKO mice and frozen in OCT (Tissue Tek, USA,) or fixed in paraformaldehyde. Tissue sections (10 μm) were incubated with the following primary antibodies: mouse anti-CD31 (1:50; Invitrogen, Carlsbad, CA, USA) and rabbit anti-TRPV4 (1:50; Biorbyt, UK). After the samples were washed with PBS, the tissue sections were incubated with appropriate secondary antibodies coupled to Alexa Fluor-488 or Alexa Fluor-594 (Invitrogen) and mounted with DAPI Fluoromount (Vector Labs, USA). Retinal flat mounts were stained with Alexa Fluor-594-conjugated isolectin IB4. Images (x10, x20) were captured using either an Olympus IX81 or FV1000 confocal microscope and processed using NIH ImageJ software.

2.6. Western blot

Tumor tissues were lysed, and the protein content was measured by Western blotting. The membranes were incubated with primary antibodies against total VEGFR2 (1:1000; Cell Signaling, Danvers, MA, USA), total and phosphor-ERK1/2 (1:1000; Cell Signaling), HIF-1α (1:1000; Cell Signaling Technologies, USA) and GAPDH (1:5000; Cell Signaling Technologies, USA) followed by goat anti-rabbit (1:20,000) antibodies conjugated with horseradish peroxidase (Cell Signaling Technologies, USA). Clarity Western ECL Substrate (Bio-Rad, Hercules, CA, USA) was employed to detect the signals using a FluorChem M Simple Imager (Protein Simple, USA).

2.7. Calcium imaging

Calcium imaging was performed as previously described[19,1,2,23,20]. Cells were stimulated with the TRPV4 selective agonist GSK1016790A (GSK101; 100 nM, Sigma, USA).

2.8. Tumor vascular leakage/permeability

Tumor vascular leakage/permeability was assessed as previously described[1,3]. Briefly, dextran, Texas Red^™, (3000 MW, Thermo Fisher Scientific, USA) was injected into tumor-bearing mice via tail vein injections. Tumors were extracted and fixed, and sections were imaged. Image processing and quantification were performed by using ImageJ software (National Institutes of Health). The results are shown as the mean fluorescence intensity.

2.9. Multiphoton microscopy

Mice bearing LLC tumors were anesthetized with isoflurane (2–3%) and injected with heparin (100 units/100 μl) via the inferior vena cava to prevent blood clotting. One hundred microliters of Alexa Fluor^™ 594 isolectin GS-IB4 conjugate (1 μg/μl) (Molecular Probes, Thermo Fisher Scientific, USA) was injected via the inferior vena cava, and after 15 min, the mice were euthanized by Fatal-Plus solution. The mice were perfused with 10 ml of 10% formalin followed by 10 ml of PBS. Tumors were harvested, and deep vascular imaging (20x) was performed using either multiphoton or confocal microscopy. Image processing and quantification were performed by Fluoview Viewer software. The tortuosity index was measured by dividing the segmented line of the vessel by the distance between the start and end of the vessel using NIH ImageJ software.

2.10. Lung metastasis

The metastasis of LLCs to the lung tissue was monitored as previously described[3], and the total number of metastatic foci was calculated per lung. The data are reported as the total number of metastatic foci.

2.11. Immunohistochemistry of OCT-embedded tumors

Tumor immunohistochemistry was performed as previously described[1,3,21]. The following antibodies were employed: mouse-anti-CD31 (1:50; Invitrogen,USA), anti-CD45 (1:50; Cell Signaling Technologies, USA), rabbit anti- NG2 (1:200) (Millipore, USA), rabbit and mouse anti- VE-cadherin (1:100) (BD Biosciences, USA), rabbit anti-VEGFR2 (1:50; Cell Signaling Technologies, USA) and appropriate secondary antibodies coupled to Alexa Fluor-488 or Alexa Fluor-594 (Invitrogen, USA). The total number of CD31⁺ stained vessels and the number of vessels that colocalized with NG2/VE-cadherin/VEGFR2 were counted and are expressed as a percentage of the VEGFR2⁺ vessels (CD31⁺ + VEGFR2⁺/CD31⁺) and mural cell coverage of NG2 and vascular integrity with VE-cadherin. Images were captured using an Olympus IX81 microscope and processed using NIH ImageJ software.

2.12. In situ zymography

Frozen tumor sections of the TRPV4^lox/lox and TRPV4^ECKO mice were brought to room temperature and rinsed with distilled water. Tumor tissue sections were incubated with DQ-gelatin fluorescein conjugate (1 mg/ml; 1: 50 dilution; Thermo Fisher) in reaction buffer (50 mM Tris-HCl, 150 mM NaCl, 5 mM CaCl₂; pH 7.6) at 37°C for 2 hrs. After incubation, slides were washed with Milli-Q water and then fixed in 4% paraformaldehyde solution for 10 min at room temperature. The slides were then washed with PBS three times and mounted with Fluoro-G containing DAPI. Confocal images were acquired at 40X, and the DQ-gelatin fluorescence intensity was quantified using ImageJ software.

2.13. Statistical analyses

Statistical analyses were carried out using either GraphPad Prism 5 or Microsoft Excel. Analyses were performed using Student’s t-test or Mann-Whitney tests. All values reported are the mean ± SEM. A value of *p < 0.05, **p < 0.01, or ***p < 0.001 was considered significant.

3. Results

3.1. Endothelial TRPV4 deletion enhances tumor growth and results in leaky and tortuous tumor vasculature

To determine the role of endothelial TRPV4 in tumor growth, we generated TRPV4^ECKO mice by crossing TRPV4^lox/lox mice with Tie2-Cre transgenic mice. First, we confirmed the TRPV4 deletion by genotyping using specific primers for Cre and loxp[18] and found deletion of TRPV4 in the TRPV4^ECKO animals but not the TRPV4^lox/lox animals (not shown). Immunohistochemical analysis of aortic cross sections showed TRPV4 staining in the endothelium (red; arrowheads in Fig. 1A) and smooth muscle layer (red) of the TRPV4^lox/lox mice; this staining was colocalized with CD31 in the endothelium (yellow; arrows; Fig. 1A). Importantly, TRPV4 staining was absent in the endothelium of the TRPV4^ECKO mouse aorta, showing no colocalization with CD31 (green; Fig. 1A), but was present in the smooth muscle layer (red), indicating specific deletion of TRPV4 in the endothelium. Further, we confirmed the deletion of TRPV4 in ECs isolated from the TRPV4^lox/lox mice but not from the TRPV4^ECKO mice using a two-step isolation protocol employing anti-CD31 and anti-ICAM2 magnetic beads. Next, we found that these isolated ECs expressed the cell surface receptor CD31 (Suppl. Fig. 1A) and VEGFR2 (Fig. 1 B), exhibited uptake of acetylated-LDL (Suppl. Fig. 1 B) and formed tubes on Matrigel (Suppl. Fig. 1C), confirming the typical EC functional phenotype. We next confirmed TRPV4 deletion in these ECs. First, RT-PCR analysis revealed that TRPV4 was expressed in ECs from the TRPV4^lox/lox mice but not in ECs from the TRPV4^ECKO mice (Fig. 1B). Finally, calcium imaging revealed that ECs from the TRPV4^ECKO mice failed to induce Ca²⁺ influx when stimulated with the TRPV4 selective agonist GSK1016790A (Fig. 1C, D), confirming the functional deletion of TRPV4 in the endothelium. Importantly, immunohistochemical analysis of various organs revealed that there was no apparent vascular phenotype in the TRPV4^ECKO mice compared to the TRPV4^lox/lox mice (Suppl. Fig. 2).

Figure 1. Characterization of endothelial TRPV4 knockout (TRPV4^ECKO) deletion in mice.

A) Immunohistochemistry shows the colocalization (yellow; arrows) of the endothelial marker CD31 (green) with TRPV4 (red; arrowheads) in the TRPV4^lox/lox mouse aortae; this finding is absent in TRPV4^ECKO mouse aortae (only green). Scale bar: 10 μm. Note that TRPV4 (red) is present only in the smooth muscle layer of the TRPV4^ECKO mouse aortae. B) RT-PCR analysis of TRPV4, VEGFR2, and GAPDH expression in isolated mouse aortic endothelial cells from the TRPV4^lox/lox and TRPV4^ECKO mice. C) Representative average traces showing relative changes in cytosolic calcium in response to GSK1016790A (GSK101) in the Fluo-4/AM-loaded TRPV4^lox/lox and TRPV4^ECKO endothelial cells (n=64). The arrow represents the time when the cells were stimulated with GSK101 (F/F0 = ratio of normalized Fluo-4 fluorescence intensity relative to time). D) Quantitative analysis of cytosolic calcium influx induced by GSK101 in the TRPV4^lox/lox and TRPV4^ECKO endothelial cells. All data represented are the mean ± SEM from at least three independent experiments. Statistical significance is indicated by *** p≤0.001.

To confirm a direct role for endothelial TRPV4 in tumor angiogenesis in vivo, we induced syngeneic tumors in the TRPV4^ECKO and TRPV4^lox/lox mice by subcutaneously injecting mouse LLCs and monitored tumor growth for 21 days. We found that the tumors were palpable by 7 days and that there was a significant increase in tumor growth by 14 days. Tumor growth was 1.5-fold higher in the TRPV4^ECKO mice than in the TRPV4^lox/lox mice at 21 days (Fig. 2A, B). Multiphoton and confocal microscopy revealed that the tumors from the TRPV4^ECKO mice exhibited abnormal angiogenesis, as shown by increased vessel density characterized by tortuous and larger vessels (Fig. 2C–F). Quantification of vessel tortuosity revealed a significant increase in the tumors from the TRPV4^ECKO mice compared to the TRPV4^lox/lox mice (Suppl. Fig. 3). Further, the TRPV4^ECKO mouse tumors displayed vessels with reduced pericyte coverage (NG2/CD31) and vascular integrity (CD31/VE-cadherin) compared to the tumors from the TRPV4^lox/lox mice (Fig. 3A, B). Finally, we found increased leakage of intravenously injected fluorescent dextran in the TRPV4^ECKO mouse tumors compared to the TRPV4^lox/lox mouse tumors (Fig. 3C). Tie-2 Cre may impact cells other than ECs, such as CD45-positive hemopoietic cells[24]. Therefore, we measured the recruitment of CD45-positive cells in the tumors and found no difference between the tumors in the TRPV4^lox/lox and TRPV4^ECKO groups (Suppl. Fig. 4). suggesting that the observed effects on tumor growth are mainly due to deletion of TRPV4 in ECs.

Figure 2. Deletion of endothelial TRPV4 channels promotes tumor growth and abnormal vasculature.

A) Tumor growth curves of the TRPV4^lox/lox (n =9) and TRPV4^ECKO (n =11) mice bearing Lewis lung carcinoma (LLC) cells. B) Tumor volumes of the TRPV4^lox/lox and TRPV4^ECKO mice at day 21. C-D) Multiphoton microscopy images of tumor vasculature from the TRPV4^lox/lox and TRPV4^ECKO mice (n=3) showing the blood vessel diameter. Mice were infused with Alexa Fluor 594-conjugated isolectin B4 prior to tumor extraction. Scale bar: 10 μm. E) Representative immunofluorescence images of tumor sections of the TRPV4^lox/lox and TRPV4^ECKO mice stained with an endothelial marker (CD31, red) (n=3). Scale bar: 10 μm. F) Quantification of vessel density from sections of the TRPV4^lox/lox and TRPV4^ECKO mouse tumors. All data are presented as the mean ± SEM, and statistical significance is indicated by * p≤0.05, **p≤0.01, and *** p≤0.001.

Figure 3. Endothelial TRPV4 channel deletion disrupts vascular integrity, leading to leaky vasculature.

A) Representative merged immunofluorescence images from tumor sections of the TRPV4^lox/lox and TRPV4^ECKO mice stained with an endothelial marker (CD31, red) and pericyte marker (NG2, green) (n=3). Scale bar: 10 μm. Quantification of pericyte coverage (NG2, pericytes in green/CD31, EC in red). B) Representative merged immunofluorescence images from the TRPV4^lox/lox and TRPV4^ECKO mouse tumor sections stained with an endothelial marker (CD31, red) and endothelial adherens junctional protein (VE-cadherin, green) (n=3). Scale bar: 10 μm. Quantification of VE-cadherin coverage (VE-cadherin, adherens junctional protein in green/CD31, ECs in red). C) Representative immunofluorescence images showing the leakage of Texas red dextran from the vasculature of the TRPV4^lox/lox and TRPV4^ECKO mouse tumors at day 21. Quantification of mean fluorescence intensity (MFI) (n=3). All data are presented as the mean ± SEM, and statistical significance is indicated by **p≤0.01 and *** p≤0.001.

3.2. Loss of endothelial TRPV4 increases the expression of VEGFR2 in tumors

To delineate the molecular mechanism underlying abnormal tumor angiogenesis in the TRPV4^ECKO mice, we focused on the VEGFR2 pathway, which is downstream of TRPV4 in ECs[23]. Western blot analysis and immunohistochemistry revealed a significant increase in VEGFR2 expression and concomitant colocalization with CD31 in the tumors from the TRPV4^ECKO group compared to the TRPV4^lox/lox group (Fig. 4A, B). Further, we found significant upregulation of the p-ERK level in the tumors of the ECKO mice compared to the TRPV4^lox/lox mice (Suppl Fig. 5). However, we found no difference in the expression of VEGF or HIF-1α between the tumors in the TRPV4^ECKO group and the TRPV4^lox/lox group (Suppl. Figs. 6 and 7).

Figure 4. Loss of endothelial TRPV4 increases VEGFR2 expression in LLC tumors.

Western blot analysis (A) and quantification (B) of VEGFR2 expression in LLC tumors from the TRPV4^lox/lox and TRPV4^ECKO mice (n=3). C) Representative merged immunofluorescence images of tumor sections stained with endothelial markers (CD31, red) and VEGFR2 (green) from the TRPV4^lox/lox and TRPV4^ECKO mice. Scale bar: 10 μm. D) Quantification of VEGFR2-positive vessels in tumor sections from the TRPV4^lox/lox and TRPV4^ECKO mice (n=3). All data are presented as the mean ± SEM, and statistical significance is indicated by **p≤0.01 and ***p≤0.001.

3.3. TRPV4 deletion in the endothelium promotes lung metastasis

To explore the functional consequence of abnormal tumor angiogenesis (leaky vessels) in the TRPV4^ECKO mice, we measured the metastasis of LLCs to the lung tissue. Consistent with our previous findings with the global TRPV4KO mice[3], lung metastasis was significantly higher in the TRPV4^ECKO mice at 21 days than in the TRPV4^lox/lox mice (Fig. 5A). Quantitative analysis revealed a significant increase (TRPV4^lox/lox vs TRPV4^ECKO; 8 vs 18; p<0.05) in the number of metastatic foci in the lungs of the TRPV4^ECKO mice compared to those of the TRPV4^lox/lox mice (Fig. 5B). Further, we found increased expression of matrix metalloproteinase-9 (Fig. 5C) but not matrix metalloproteinase-2 (Fig. 5D) in the TRPV4^ECKO mouse tumors compared to the TRPV4^lox/lox mouse tumors. Importantly, in situ zymography revealed increased DQ gelatinase activity in the TRPV4^ECKO mouse tumors compared to the TRPV4^lox/lox mouse tumors (Suppl. Fig. 8) with no difference in TIMP-2 expression (Suppl. Fig. 9).

Figure 5. TRPV4 deletion in the endothelium promotes lung metastasis via increased MMP-9 expression in tumors.

Representative H&E-stained images of lung sections (A) and quantification of metastatic foci (B) from the LLC tumor-bearing TRPV4^lox/lox (n=9) and TRPV4^ECKO mice (n=11). Scale bar: 200 μm. qPCR analysis of relative MMP-9 (C) and MMP-2 (D) expression in the TRPV4^lox/lox and TRPV4^ECKO mouse tumors (n=3). All data are presented as the mean ± SEM, and statistical significance is indicated by *p≤0.05 and ** p≤0.01, NS: nonsignificant.

4. Discussion

In the present study, we demonstrated that deletion of endothelial TRPV4 enhanced tumor progression compared to that of the TRPV4^lox/lox mice. Tumors from the TRPV4^ECKO mice exhibited increased vessel density characterized by large and tortuous vessels. Further, endothelial TRPV4 deletion produced leaky tumor vessels with a concomitant increase in metastasis to the lung compared with that of the TRPV4^lox/lox mice. Ion channels have recently been identified as promising VEGF alternative therapeutic targets for tumor growth and metastasis. Among these channels, TRPV4 has attracted increased attention in the regulation of tumor endothelial function and angiogenesis[1,3,17,23]. However, TRPV4 in tumor cells has been shown to exhibit opposing effects on tumor growth and metastasis in different cancers[10–16]. These contrasting reports indicate that TRPV4 may have a tumor cell-independent role in inducing tumor growth and metastasis.

Tumors perfectly exploit the endothelium to induce tumor angiogenesis and metastasis by triggering the formation of abnormal, leaky vasculature. We previously demonstrated that TRPV4 expression is reduced in tumor ECs in vitro and that tumor growth and angiogenesis are increased in a subcutaneously implanted tumor in global TRPV4KO mice[1]. Although our studies demonstrated a role for TRPV4 in the regulation of endothelial function in vitro and vasodilation ex vivo[19,1,23,9,20], no studies have directly investigated the specific role of endothelial TRPV4 in tumor growth and metastasis in vivo. To confirm the specific role of endothelial TRPV4 in tumor angiogenesis and progression, we generated endothelial-specific knockout of TRPV4. To our knowledge, this is the first report to describe the role of endothelial TRPV4 in angiogenesis using endothelial-specific knockout of TRPV4. We did not observe any obvious vascular phenotype in the TRPV4^ECKO mice because there was no change in the retinal vasculature compared to that of the TRPV4^lox/lox mice. However, syngeneic tumor experiments revealed enhanced tumor growth and increased expression of p-ERK in the TRPV4^ECKO mice compared to the TRPV4^lox/lox mice. These findings agree with our previous findings from the global TRPV4KO mice[1,3] and tumor ECs[20] and confirm the endothelial-specific role of TRPV4 in tumor angiogenesis and growth. Multiphoton microscopy revealed large and convoluted vessels in the TRPV4^ECKO tumors compared to the TRPV4^lox/lox tumors, indicating abnormal vessel formation. Both NG2 and VE-cadherin staining, and Texas red dextran injections confirmed that loss of endothelial TRPV4 triggered vascular destabilization, leading to leaky vessels. In fact, these leaky vessels resulted in enhanced metastasis to the lung. VE-cadherin is critical for endothelial barrier integrity, which was shown to be compromised in the tumor vasculature[25,26]. A small molecule that directly potentiates VE-cadherin-mediated EC junctions induced tumor vascular normalization and reduced metastasis in mice[27].

Mechanistically, TRPV4 deletion in the endothelium appears to induce activation of the VEGF/VEGFR2 pathway. Both VEGF and VEGFR2 expression and activation have been shown to be enhanced during tumor angiogenesis, which has become the major target for cancer therapy[28]. We have recently demonstrated that downregulation or absence of TRPV4 in ECs in vitro induces translocation of VEGFR2 from the Golgi to the plasma membrane through YAP. Further, we found increased tyrosine phosphorylation of VEGFR2 at Y-1175[23]. Consistent with this observation, we found that the tumors from the TRPV4^ECKO mice displayed increased expression of VEGFR2 and its downstream effector ERK1/2, indicating that deletion of endothelial TRPV4 augments the VEGF/VEGFR2 pathway. However, we did not find a significant difference in VEGF and HIF-1α expression between the tumors in the TRPV4^ECKO and TRPV4^lox/lox groups. We measured the VEGF and HIF-1α levels only at the end of the experiment (21 days) and cannot rule out the possibility that VEGF/HIF-1a expression may peak at earlier stages (7–14 days). Interestingly, we observed increased expression of matrix metalloproteinase-9 and DQ gelatinase activity in the TRPV4^ECKO mouse tumors compared to the TRPV4^lox/lox mouse tumors, suggesting that increased MMP-9 may be responsible for increased metastasis in the absence of endothelial TRPV4.

Our study demonstrates that endothelial TRPV4 suppresses tumor angiogenesis, tumor growth, vascular disruption, and metastasis, most likely by controlling VEGF/VEGFR2 localization/activation. These results indicate that angiogenesis is regulated by a tight coordination between mechanical and soluble (growth factor) signaling. Although the endothelium is continuously exposed to mechanical forces such as shear stress and cyclic strain, most studies on the endothelial role in tumor angiogenesis have focused on soluble factors such as VEGF. Previous work by Mammoto et al. demonstrated that ECM stiffness governs VEGFR2 gene promoter activity and expression via a mechanotranscriptional mechanism to stimulate angiogenesis[29]. However, the identity of the proximal mechanosensor(s) involved remains unknown. We recently demonstrated that the mechanosensitive ion channel TRPV4 regulates angiogenesis in an ECM stiffness-dependent manner in vivo[3]. Further, we showed that the absence or downregulation of TRPV4 in ECs induces membrane translocation and phosphorylation of VEGFR2 through the Rho/YAP pathway[23]. Taken together, these findings suggest that mechanical stiffness regulates angiogenesis through a mechanotranscriptional mechanism through both VEGFR2 expression and/or its activation, which is regulated by TRPV4. VEGF and its receptor VEGFR2 are critical for angiogenesis, yet direct targeting of the VEGF/VEGFR2 pathway yielded modest results in the clinic[30–33]. Nevertheless, most vascular normalization therapies still target growth factors such as VEGF and PDGF[34,28]. In conclusion, we propose that endothelial TRPV4 can be an alternate therapy for cancer instead of current strategies that inhibit angiogenesis by targeting VEGF and its receptors. Importantly, targeting endothelial TRPV4 could also offer a potential novel strategy for vascular normalization and cancer therapy.