TRPV2 is upregulated in ESCC cells, and overactivation of TRPV2 promotes cell proliferation
Transient receptor potential vanilloid receptor 2 (TRPV2) is functionally thermosensitive and can be activated by heat at high temperature (> 52°C) [25]. To explore the expression of TRPV2 in ESCC cells (Eca-109 and TE-1) and non-tumor esophageal squamous cells (NE2) (the anchorage-independent growth of the cell lines was tested by a soft agar colony formation assay, Additional file 1: Figure S1a), we conducted western blotting. TRPV2 protein (86 kDa) was detectable among all three cell lines, and the expression of TRPV2 was found to be upregulated in ESCC cells compared with NE2 cells (Fig. 1a-b). To further examine the expression and localization of TRPV2 among these cell lines, immunocytofluorescence was carried out. TRPV2 was found to be expressed and localized predominantly in the plasma membrane of ESCC cells and NE2 cells (Fig. 1c and Additional file 1: Figure S1b).
Cellular proliferation viability was evaluated via a single-cell culturing assay, based on previous studies that tumors could originate from a single cell and further grow uncontrollably, eventually leading to a malignant state [26, 27] and our observation that tumor cells are more tolerant of ‘unfavorable’ conditions than non-tumor cells, therefore they could survive an extremely scarce nutritious environment.
Cellular proliferation of Eca-109 was promoted significantly upon exposure to recurrent brief heat stimuli (54°C) or frequent application of O1821 (20 μM), which could activate TRPV2, as confirmed by a calcium imaging assay (Additional file 2: Figure S2a-l). In this study, the term ‘overactivation’ was used to describe the abovementioned patterns of heat treatment (54°C) or the application of O1821 (20 μM), which could cause frequent activations of TRPV2. The pro-proliferative effects on Eca-109 cells were abrogated by either simultaneous application of tranilast (120 μM), a TRPV2 antagonist, or by TRPV2 knockout using CRISPR-Cas9 (Fig. 1d–f). Similar effects could be observed in another ESCC cell line, TE-1 (Additional file 3: Figure S3a-b). It is worth noting that cellular proliferation of Eca-109 was decreased substantially by the knock-out of TRPV2. The proliferation of wild-type NE2 (NE2-WT, which was with low expression level of TRPV2, Fig. 1a-b) cells was affected by neither exposure to recurrent brief heat stimuli (54°C) nor
application of O1821 (20 μM) (Fig. 1g and Additional file 3: Figure S3c). Conversely, when the proliferation assay was conducted on NE2 cells with ectopically expressed TRPV2 (NE2-VR2), the cellular proliferation of NE2 cells was enhanced markedly upon exposure to recurrent brief heat stimuli (54°C) or O1821 (20 μM), and these effects were attenuated by tranilast (120 μM) (Fig. 1h and Additional file 3: Figure S3d). Together, these data indicated that overactivation of TRPV2 could promote the proliferation of esophageal squamous cells and upregulation of TRPV2 might play a role in the pathology of ESCC.
Overactivation of TRPV2 enhances the migration and invasion of ESCC cells in vitro
To assess the impact of TRPV2 activation on ESCC cell migration, a wound healing assay was applied. Cellular migration of ESCC cells (Eca-109 and TE-1) was considerably accelerated upon the activation of TRPV2 by recurrent brief heat (54°C) stimuli or administration of O1821 (20 μM), and these effects were abolished by either the TRPV2 antagonist, tranilast (120 μM), or by TRPV2 knockout using CRISPR-Cas9 (Fig. 2a-b and Additional file 4: Figure S4a-b). These data suggested that overactivation of TRPV2 could promote the migratory ability of ESCC cells. On the other hand, the migratory ability of wild-type NE2 cells (NE2-WT, which had a low expression level of TRPV2, Fig. 1a-b) remained unaffected upon the overactivation of TRPV2 by heat (54°C) stimuli or O1821 (20 μM) treatment (Fig. 2c and Additional file 4: Figure S4c). Conversely, cellular migration was enhanced profoundly upon the overactivation of TRPV2 by both cues in the NE2 cells with ectopically overexpressed TRPV2 (NE2-VR2) (Fig. 2d and Additional file 4: Figure S4d), hence further verifying the promigratory role of TRPV2 in the cells.
To evaluate the cellular invasive process, we used a new platform adopting AIM 3D chips [28]. The trajectory of invasive cells can be monitored over time during the experiments, and the invasive process (including invasive cell numbers and furthest invaded distance by the most front cell) can be easily visualized and measured using these chips in the assay, compared with the traditional way of using Boyden chambers [29-31]. The number of invasive ESCC cells (Eca-109 and TE-1) and the furthest invaded distance of these cells were enhanced markedly by recurrent brief heat (54°C) stimuli or exposure to O1821 (20 μM), and these effects were attenuated significantly by tranilast (120 μM) or by TRPV2 knockout using CRISPR-Cas9 (Fig. 2e-f and Additional file 5: Figure S5a-c). It is noteworthy that both cellular migration and invasion of Eca-109 cells was significantly inhibited by the knockout of TRPV2 ((Fig. 2a-b, e-f and Additional file 5: Figure S5a, c). Taken together, these findings suggested that overactivation of TRPV2 could promote the migratory and invasive ability of ESCC cells.
For the non-tumor line, the invasive ability (including invasive cell numbers and furthest invaded distance) of wild-type NE2 (NE2-WT) cells were affected neither by the overactivation of TRPV2 by recurrent brief heat stimuli (54°C) nor by O1821 (20 μM) treatment, whereas the invasive ability (including invasive cell numbers and furthest invaded distance) of ectopically expressed TRPV2 NE2 (NE2-VR2) cells was elevated substantially upon the overactivation of TRPV2 by recurrent brief heat stimuli (54°C) or O1821 (20 μM) administration. Again, these effects were abolished by tranilast (120 μM) (Fig. 2g and Additional file 5: Figure S5c), which further corroborates the pro-invasive role of TRPV2.
Overactivation of TRPV2 in ESCC cells promotes tumor-related angiogenesis
The capability to promote angiogenesis within or surrounding the tumor tissue is a hallmark of many oncogenic factors [32]. To examine the impact of TRPV2 overactivation in ESCC cells on angiogenesis, a tube formation assay was applied (Fig. 3a). Human umbilical vein endothelial cells (HUVECs) were used as angiogenesis progenitor cells. The recruitment of HUVECs and the total length of newly formed microvessels were all significantly promoted by conditioned medium derived from Eca-109 cells following the overactivation of TRPV2 by recurrent brief heat stimuli (54°C) or the treatment with O1821 (20 μM), and these effects were attenuated remarkably by tranilast (120 μM). Meanwhile, the pro-angiogenic effect (on the total length of newly formed microvessels) of the conditioned medium derived from Eca-109-VR2-/- (TRPV2 knocked-out Eca-109) cells following recurrent brief heat stimuli (54°C) was arrested (Fig. 3b-c). Furthermore, the number of junctions and branches of the newly formed microvessels was also significantly increased by the conditioned medium derived from Eca-109 cells following the overactivation of TRPV2 by recurrent brief heat stimuli (54°C) or by the administration of O1821 (20 μM), and these effects were markedly inhibited by tranilast (120 μM). Conditioned medium derived from Eca-109-VR2-/- cells following recurrent brief heat stimuli (54°C) induced much less microvessel formation (with much fewer junctions and branches of the newly formed microvessels) versus the control in the assay (Fig. 3b, d-e). Collectively, these findings indicated that overactivation of TRPV2 could promote tumor-related angiogenesis in ESCC cells and thus might promote the tumorigenesis of ESCC.
Overactivation of TRPV2 promotes ESCC growth and invasion in xenograft models
The biological role of TRPV2 in ESCC progression in vivo was investigated using BALB/c nude mice to generate a tumor xenograft model. In the ESCC formation assay, the tumors originating from Eca-109 cells followed recurrent brief heat (54°C) challenge or O1821 (20 μM) application were significantly larger, in both size and weight, than the tumors from control cells, and these effects were attenuated markedly by tranilast (120 μM) (Fig. 4a–c). Notably, the tumors formed by TRPV2 knockout cells (Eca-109-VR2-/-) followed recurrent brief heat (54°C) treatment were clearly smaller and had substantially lower tumor weights than the tumors formed by control cells (Fig. 4a–c). By contrast, wild-type NE2 cells were subcutaneously injected into the BALB/c nude mice, but no tumor formation was found even up to 30 days post inoculation. Conversely, when we used ectopically expressed TRPV2 NE2 (NE2-VR2) cells (Additional file 6: Figure S6a-b) to perform the similar experiments, on day 12 after inoculation, tumors were palpable in the groups which were subcutaneously injected with NE2-VR2 cells followed recurrent brief heat (54°C) challenge or O1821 (20 μM) application. To the end of the assay, it was observed that overactivation of TRPV2 by recurrent brief heat (54°C) challenge or O1821 (20 μM) treatment could significantly promote tumor formation, which was displayed with larger tumor sizes and greater tumor weights compared to the control group, and these effects were compromised markedly by tranilast (120 μM) (Additional file 6: Figure S6c-e). These findings further verified that overactivation of TRPV2 could significantly promote ESCC tumor formation in nude mouse models.
Moreover, an experimental tail vein metastasis model was established using BALB/c nude mice. The photon flux of Eca-109-gl cells (GFP and luciferase dual-labeled Eca-109 cells) followed recurrent acute heat (54°C) challenge or O1821 (20 μM) administration to the lungs was profoundly enhanced compared with that in the control group, and these effects were attenuated significantly by tranilast (120 μM) application or by TRPV2 knockout using CRISPR Cas9 (Fig. 4d and Additional file 6: Figure S6f), suggesting that overactivation of TRPV2 markedly promotes the ability of Eca-109-gl cells to metastasize to the lungs following cell injection through the tail vein. Lungs derived from the group followed overactivation of TRPV2 in ESCC cells by recurrent acute heat (54°C) challenge showed more metastatic tumor nodules than lungs from the control group (the rightmost panel of Fig. 4d). Consistently, the H&E staining of the lungs showed clear tumor lesions in the group followed over-activation of TRPV2 in the ESCC cells by recurrent acute heat (54°C) challenge, whereas tumor lesions were attenuated markedly by tranilast (120 μM). Meanwhile, tumor lesion was nearly absent from the TRPV2 knockout (Eca-109-gl-VR2-/-) group (Fig. 4e). Immunohistocytochemical analysis revealed that the group followed overactivation of TRPV2 in ESCC cells by recurrent acute heat (54°C) challenge showed markedly increased percentages of Ki-67-positive cells and of CD31-positive cells (which represents the proliferative cells and the newly formed microvessels respectively) compared to those of the control group (Fig. 4e-g). When combined with tranilast (120 μM), the group followed over-activation of TRPV2 in ESCC cells by recurrent acute heat (54°C) challenge demonstrated much smaller Ki-67 proliferation indices and weaker newly formed microvessels signals (Fig. 4e-g), implying that overactivation of TRPV2 could promote tumor proliferation and tumor-related angiogenesis in vivo. Moreover, ESCC tumor formation under the skin, tumor metastasis to the lungs, the proliferation index and newly formed microvessels signal of the mouse lungs were all reduced by the knockout of TRPV2 (Fig. 4), further supported the notion that the driving role of TRPV2 in the progress of ESCC.
To sum up, these data suggest that the overactivation of TRPV2 contributes to the augmented proliferative, angiogenic and metastatic capacity of ESCC cells and hence drives ESCC progression in vivo.
Overactivation of TRPV2 activates HSP and PI3K/Akt/mTOR signaling pathways
To explore the mechanism(s) underlying the role that overactivation of TRPV2 plays in the progression of ESCC, western blotting was carried out. Eca-109 cells were treated as described in Drug administration and thermal stimulation protocol (see Methods) prior to western blotting assay.
It is well known that HSF1 (heat shock factor 1) mediates heat shock stress within organisms; therefore, we tested whether HSF1 is involved in the heat treatment of the ESCC cells. As expected, HSF1 was upregulated by overactivation of TRPV2 by heat stimuli (54°C), and this effect was inhibited by tranilast (120 μM) or TRPV2 knockout using CRISPR-Cas9 (Fig. 5e-f), indicating that HSF1 was modulated during the overactivation of TRPV2 by heat.
Next, five members of the HSP family, including HSP27, HSP40, HSP60, HSP70, HSP90, were examined during the assay, and only HSP70 and HSP27 were found to be upregulated by overactivation of TRPV2 upon exposure to heat stimuli (54°C) or O1821 (20 μM) (Fig. 5a-c, e-f). The expression levels of both proteins were returned to near baseline when the activation was antagonized by tranilast (120 μM), suggesting that the expression of these HSP proteins was mediated by TRPV2 (Fig. 5a, c, e-f). Previous works had reported that these two HSP proteins are involved in the progression of multiple types of cancers [33-35]. HSP70 and HSP27 may also play a role in the tumorigenesis of ESCC following the overactivation of TRPV2. The expression levels of calmodulin were inversely proportional to the overactivation of TRPV2 (Fig. 5b, d), possibly because more calmodulin protein was employed to modulate the increased influx of Ca2+ following overactivation of TRPV2, which is consistent with the results of Ca2+ imaging assays (Additional file 2: Figure S2a-b, e-f).
Given that noxious heat stimulation may induce inflammatory reactions, we then measured two important inflammation-related pathways, PI3K and NFkB. The expression of PI3K was enhanced with the overactivation of TRPV2 by either heat stimuli (54°C) or O1821 (20 μM), and these effects were inhibited by the TRPV2 inhibitor tranilast (120 μM) or by TRPV2 knockout using CRISPR Cas9, whereas the expression of NFkB remained unchanged during the overactivation of TRPV2 (Fig. 5e-f), suggesting that PI3K may be involved in the overactivation of TRPV2, while NFkB is not.
To further investigate the PI3K pathway playing in the progression of ESCC driven by overactivation of TRPV2, we measured the expression of up- and downstream signaling proteins of this pathway. PDK1, which can be activated by PI3K, was found to be upregulated upon overactivation of TRPV2, and as expected, AKT1 and mTORC1, the target proteins of PDK1, were accordingly upregulated by overactivation of TRPV2 (Fig. 5g-h). In contrast, PTEN, the negative regulatory protein of PI3K, was conversely regulated during the process of overactivation of TRPV2 (Fig. 5g-h), suggesting that the PI3K signal was activated during the overactivation of TRPV2 and may be amplified by PTEN, thus significantly promoting the aggressiveness of ESCC upon overactivation of TRPV2.
In addition, it is worth noting that TNFα was upregulated by heat stimuli, while this effect may not be regulated by TRPV2 because its expression remained unchanged when tranilast (120 μM) was simultaneously applied or TRPV2 was knocked-out using CRISPR-Cas9 (Fig. 5e-f). The activation of TNFα results in cell death [36]. Indeed, we did observe a small portion of cell death during the heat stimulation process, while the overall cell numbers were not decreased but increased in response to overactivation of TRPV2 (Fig. 1d, f), suggesting that both pro-cell death and pro-cell proliferation signals could be simultaneously activated during the process of overactivation of TRPV2 by heat stimuli, while the latter signal exceeded the former, thus leading to the substantial increase in ESCC cell numbers upon overactivation of TRPV2.
Last, to further confirm the role of TRPV2-PI3K/Akt/mTOR playing in the progression of ESCC, VS5584 (a pan-PI3K/mTOR kinase inhibitor) and oroxin B (a PTEN protein activator and a pan-PI3K/mTOR kinase inhibitor) were applied in the Eca-109 cellular proliferation assay. As expected, both compounds could significantly attenuated Eca-109 proliferation following overactivation of TRPV2 by frequent acute heat (54°C) challenge or O1821 application (Additional file 7: Figure S7a-b). Thus, we further verified the driving role of the TRPV2-PI3K/Akt/mTOR axis in the progression of ESCC. Taken together, these data suggest that the HSP and PI3K/Akt/mTOR signaling pathways participate in the overactivation of TRPV2 and may play an important role in the tumorigenesis of ESCC upon the overactivation of TRPV2 (Additional file 7: Figure S7c).
TRPV2 is upregulated in ESCC tumor tissues
To ask the question of whether tumor tissues from patients have similar TRPV2 expression patterns to the ESCC cells used in the present study, we performed western blotting on 31 fresh samples derived from ESCC patients to detect the expression of TRPV2 protein and conducted IHC on 193 pathological slides of ESCC patients obtained from multiple hospitals (patient information is shown in Additional file 9: Table S2).
As expected, upregulation of TRPV2 was found in over 87% (27 out of 31 cases) of the fresh ESCC tissues compared with their matched adjacent non-tumor tissues (Fig 6a-c). The staining of TRPV2 in the IHC assay showed that, in comparison to the non-tumor tissue, upregulation of TRPV2 was found in 84.1% of ESCC tumor tissues compared with adjacent non-tumor tissues (Fig 6d, f). Intriguingly, the expression of TRPV2 was found to be upregulated among 77.5% of the squama of the para-tumor in the tumor slides compared with that of the non-tumor slides (Fig 6e, g). Further efforts are needed to identify whether TRPV2-positive cells in the squama of those para-tumor tissues are transformed or untransformed cells.