Activation of TRPA1 Channel by Antibacterial Agent Triclosan Induces VEGF Secretion in Human Prostate Cancer Stromal Cells

Prostate cancer is among the most frequent malignancies in industrialized nations, and continuous research efforts are undertaken in order to better understand its development and progression. Androgen dependence of this tumors is long known (1–3), but epidemiologic studies suggest that factors from Western life style also contribute to its development (4, 5).

Epithelial stromal interactions are considered important for prostate cancer development and progression (6, 7). Carcinomas in general are composed of two interdependent components: the neoplastic epithelial cells and the supporting tumors stroma, which plays decisive roles in pivotal processes such as tumors proliferation, vascularization, and invasion (8–10) via the secretion of growth factors and cytokines. Following epithelial changes in carcinogenesis, the surrounding stroma is modified by cancer cell–derived factors. These modifications drive the emergence of the characteristic reactive stroma: the modified stromal cells secreting extracellular matrix proteins and soluble factors (cytokines, growth factors), which in turn play important roles in initiation and/or progression of certain carcinomas (11–14), including breast and prostate cancers (9, 15–17). Modulation of these stromal factors secretion might affect the initiation or progression of prostate cancers. In addition to the epithelial-derived and inflammatory factors, environmental molecules might target the stromal cells to modify their physiology, including their secretion capacity. One of the mechanisms involved in the effects of the environmental compounds is the modification of the cellular calcium homeostasis (18). It is well established that Ca²⁺ responses to extracellular stimuli can lead to rapid secretion of growth factors and hormones through exocytosis (19).

Accumulating data indicate that exposure to environmental compounds may adversely affect human health. Triclosan (TCS; 2,4,4′-trichloro-2′-hydroxydiphenyl ether; a chlorophenol) is an antimicrobial agent widely used as preservative in toothpastes, soaps, shampoos, and cosmetics at concentrations up to 0.3% or 10 mmol/L (20). TCS is known to be a highly toxic chemical for aquatic flora and fauna (21) and thus has been included in the probable list of endocrine disruptors (ED). Triclosan is also detected in human blood plasma (22), breast milk (at a concentration of 1–10 μmol/L; ref. 23), and urine (24), and daily exposure to TCS of breast tissue was estimated to be of 5.5 μmol/L (25, 26). At such levels, TCS has been shown to modulate the growth of mammary cancer cells (27), but its potential effects are not known on human prostate cancer epithelial and stromal cells. The mode of action of TCS as an ED is controversial, and it was recently showed that TCS could exert its effects via several mechanisms, including the modulation of androgen receptor (28), the modulation of the activity of an adenylyl cyclase enzyme (29) and, recently, the activation of calcium signaling (25). The latter effect of TCS is of importance because it is well established that in most cell types, increase in cell calcium (Ca²⁺) concentrations is associated with cell growth and/or secretory responses (30).

As epithelial-stromal interactions play a pivotal role in the progression of prostate cancers, environmental factors like TCS may affect these interactions to favor this progression. The present work was designed to determine the effects of TCS on both epithelial and stromal cells derived from prostate cancer tissues. Our work focused on the effects of this environmental factor on calcium signaling, an important parameter controlling the secretion of the mitogenic and/or angiogenic factors by stromal cells known to trigger epithelial and endothelial cells' development.

Using a combination of immunostaining, calcium imaging and electrophysiology techniques, we show here for the first time that the widely used antimicrobial TCS induces the activation of TRPA1 calcium permeable channels and the subsequent release of vascular endothelial growth factor (VEGF) in human prostate cancer stromal cells. We also show the exclusive expression of the TRPA1 protein in stromal cells of human prostate cancer tissue. TCS could thus have a significant impact on prostate carcinogenesis through the activation of TRPA1 channels and subsequent release of VEGF.

All chemicals were from Sigma-Aldrich. Triclosan (Irgasan) and triclocarban were dissolved in DMSO, conserved at −20°C and diluted at the desired concentration on the day of the experiments.

LNCaP, PC-3, and DU145 prostate cancer cell lines and the HEK293 (human embryonic kidney-derived 293) cell line were obtained from the American Type Culture Collection (ATCC) and cultured as described by Gackiere and colleagues (31).

Human prostate cancer biopsies were obtained from consenting patients following the local ethical considerations. All experiments involving patient tissues were carried out under approval number CP 01/33, issued by the Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale de Lille. A portion of prostate tissue suspicious for carcinoma was incised, and one-half of the sliced tissue was submitted for immediate microscopic examination on cryostat sections. After establishment of the diagnosis of adenocarcinoma, the remaining half of the tissue was used for primary culture. The tissue was cut in multiple minute cubicles, placed on a plastic surface, and grown in Phenol red-free RPMI 1640 containing charcoal-stripped fetal calf serum (FCS; CS-RPMI) containing 1 nmol/L DHT for the growth of prostate cancer stromal cells (PrSC) and in keratinocyte medium supplemented with 50 mg/mL bovine pituitary extract, 5 ng/mL epidermal growth factor (EGF), 100 mg/mL streptomycin, and 100 U/mL penicillin (Life Technologies, Inc.) and 2% FCS for the epithelial cells growth. PrSC cultures were validated by immunofluorescence using anti–α-actin and mouse anti-Vimentin (Dako) antibodies.

PrSC were transfected overnight with 25 nmol/L of control siRNA (targeting Luciferase mRNA; Eurogentec) or a siRNA raised against TRPA1 mRNA (siTRPA1-1, target sequence 5′-GGUGGGAUGUUAUUCCAUA(dTdT)-3′ or siTRPA1-2: 5′-GAAGGACGCUCUCCACUUA(dTdT)-3′) as previously described (32) and used 48-hour posttransfection. The efficiency of the siRNAs was validated by RT-PCR and by Western blot studies (shown for siTRPA1-1 in Fig. 2D). RT-PCR experiments and HEK293 stably transfected by either hTRPA1 coding sequence (HEK-hTRPA1) or by pcDNA3 vector (HEK-vector) were realized as described earlier (32). The PCR primers (hTRPA1: 5′-AGTGGCAATGTGGAGCAA-3′ and 5′-TCTGATCCACTTTGCGTA-3′; β-actin: 5′-CAGAGCAAGAGAGGCATCCT-3′ and 5′-GTTGAAGGTCTCAAACATGATC-3′ used in this study were designed on the basis of established GenBank sequences and synthesized by Invitrogen. The amplified PCR products were of 510 and 212 bp, respectively.

The protein expression studies of the ion channels in prostate cancer cells and tissues were determined by indirect immunofluorescence analysis performed on acetone-fixed cells and formalin-fixed paraffin embedded (FFPE) tissues as previously described (32) and analyzed by confocal microscopy (Zeiss LSM 780; acquisition parameters: objective 40×/1.3; thickness of confocal slide, 1 μm). Three prostate non-cancer and four prostate cancer tissues were used in the present work to study the expression of the TRPA1 channel. The prostate cancer tissues were from 54-, 60-, 61-, 76-year-old patients all presenting the same TNM (T2N0M0) with Gleason grades of 3, 3, 4, and 4 and Gleason scores of 6, 5, 7, and 7, respectively. For the expression of the TRPA1 channel in breast cancer, two grade 2 breast carcinoma (infiltrating ductal carcinoma) from 52- and 61-year-old patients were used, and the expression of the channel was examined in cancer and in the adjacent non-cancer regions.

PrSC and HEK293 cells were cultured at 80% of confluence and total proteins extracted. Ten to μm of each sample were analyzed by Western blotting using hTRPA1 (Alomone Labs, 1/500e) β-actin (Sigma-Aldrich, 1/2000e) and Calnexin (Santa Cruz Biotechnology, 1/200e) antibodies and ECL technique as described previously (32).

Cells were grown on glass coverslips, and [Ca²⁺]_i imaging experiments were performed in Hanks Balanced Salt Solution (HBSS) in mmol/L; 142 NaCl, 5.6 KCl, 1 MgCl₂, 2 M CaCl₂, 0,34 Na₂PO₄, 0,44 KH₂PO₄, 10 Hepes, 5.6 glucose, and buffered to pH 7.4), as previously described (32) using Fura-2 AM as calcium dye. To represent the variation of the [Ca²⁺]_i, the fluorescence intensity ratio represented by F340/F380 was used as an indicator of changes in cytosolic Ca²⁺ concentration. Each experiment was repeated at least 4 times in duplicate on different cell cultures on a field of 35 to 45 cells and representative experiments performed on 60 to 90 cells as mean ± SE are presented.

Whole-cell patch-clamp recordings were performed using a RK-300 patch-clamp amplifier (Biologic) as previously described (31). Bath medium used for whole-cell experiments consisted in HBSS supplemented with 10 mmol/L TEA to block potassium currents. Recording pipettes were filled with a solution containing (in mmol/L) 140 KGluconate, 10 NaCl, 10 HEPES, 2 MgCl₂, with 1 EGTA. Osmolarity and pH were adjusted to 290 mOsm.L⁻¹ and 7.2, respectively.

Primary cultured stromal cells (5 × 10⁵ cells per condition) were cultured in CS-RPMI supplemented with 0.1% FCS with or without TCS for 3 days. Culture supernatants were collected, centrifuged, and then incubated with PC-3 cells for the cell growth studies. PC-3 cells seeded in 96-well plates (3,000 cells/well) were cultured in RPMI/10% FCS for 24 hours. The medium was then replaced by 150 μL of PrSC-conditioned medium and incubated for 3 days. The control PC3 cells were cultured in CS-RPMI supplemented with 0.1% FCS for 3 days with or without TCS. Cell growth was estimated using the commercial assay MTS/PMS kit as previously described (32). PC-3 cell proliferation assays were also evaluated by Malassez manual cell counting. Briefly, PC-3 cells were cultured in 12-well plates (4-well per condition) and after incubation periods, cells were trypsinized and suspended in 500 μL PBS and 10 μL of each sample were used for manual cell counting. Each sample was counted at least twice, and the results are expressed in number of cells. Assays were performed in triplicate. Differences between samples and the corresponding control (CTL) were determined by the one-way ANOVA analysis, where P < 0.05 was considered statistically significant.

PrSC cells were incubated in CS-RPMI supplemented with 0.1% FCS with or without TCS for 3 days into 12-well plates. The supernatants of the cell cultures were then withdrawn, centrifuged, and a VEGF immunoassay kit (Abcam) was used for VEGF detection according to the manufacturer's instructions.

Plots were produced using Origin 8.0 (Microcal Software). Results are expressed as mean ± SE. Statistical analysis was performed using unpaired t tests or ANOVA tests followed by either Dunnett (for multiple control vs. test comparisons) or Student–Newman–Keuls posttests (for multiple comparisons). The Student t test was used for statistical comparison of the differences, and P < 0.05 was considered statistically significant.

We first assessed the effects of Triclosan (TCS) and its structural analog, Triclocarban (TCC) on prostate cancer stromal cells. The PrSC primary culture were first characterized for the expression of stromal cells markers by immunofluorescence studies using antibodies raised against α-actin and vimentin. Seven different cultures originated from 7 patients were included in this study. We observed that the proportion of cells expressing α-actin and/or vimentin varied between each primary cell culture. Globally, the PrSC cultures were divided in two categories, the first one in which 90% of cells expressed vimentin (fibroblasts) and the remaining 10% expressed α-actin (smooth muscle cells; Fig. 1C, inset), and the other one in which almost 100% of cells expressed both vimentin and α-actin (myofibroblasts; Fig. 1D, inset), suggesting the presence of only stromal cells in our cultures.

On these characterized stromal cells, by calcium imaging, we studied the effects of TCS and its structural analogue, TCC (see chemical structures in insets in Fig. 1A and B), on cytosolic free calcium concentration ([Ca²⁺]_i). The [Ca²⁺]_i was continuously monitored as a fluorescence ratio F340/F380 using the calcium probe Fura-2/AM in a normal HBSS medium containing 2 mmol/L CaCl₂ and then, at the time indicated, TCS or TCC was applied to the cells. In these experiments, TCS induced a rapid and dose-dependent increase in [Ca²⁺]_i in PrSC (Fig. 1A). These observations suggest that TCS is able to induce either a calcium entry and/or a Ca²⁺ release from intracellular stores. In the same experimental conditions, TCC failed to elicit any significant modulation of basal [Ca²⁺]_i in PrSC (Fig. 1B). Two types of TCS-induced calcium increase were observed with regard to the kinetics and amplitudes (Fig. 1C and D). TCS induced an initial rapid and transient first phase followed by a sustained plateau phase characterized by either a low (Fig. 1C) or a high amplitude (Fig. 1D). The combined calcium imaging and immunofluorescence studies using α-actin and vimentin antibodies suggest that these two different types of calcium response are correlated with the preponderance of myofibroblasts for the high-amplitude plateau phase (Fig. 1D) and with the preponderance of fibroblasts for the low-amplitude plateau phase (Fig. 1C) in stromal cells populations. This robust effect of TCS on calcium entry was observed on PrSC from 6 out of 7 cancer tissues.

In calcium imaging experiments, cells were challenged with TCS in a Ca²⁺-free HBSS medium (Fig. 1E) followed by the addition of 2 mmol/L calcium still in the presence of TCS. In the absence of external calcium, TCS induced a slight calcium response due to a mobilization of calcium from intracellular pools, through the activation of Type I Ryanodine Receptors (RyRs) as previously described in mouse skeletal myotubes (25). Interestingly, as shown in Fig. 1E, when calcium was added in the external buffer, TCS induced a rapid and high-amplitude calcium entry in human PrSC via the plasma membrane calcium channels. The same experiments performed on human prostate cancer epithelial cells (hPEC) showed that TCS only induced a calcium mobilization without promoting a calcium entry in these cells (Fig. 1F). We therefore undertook experiments in order to identify the calcium permeable membrane channels involved in the effects of TCS in primary cultured PrSC.

To investigate the membrane ion channels involved in TCS-induced calcium entry, we used an extensive pharmacology comprising blockers of voltage-dependent, non–voltage-dependent calcium channels and potassium channels (Supplementary Fig. S1A). Surprisingly, many tested blockers [inhibitors of L-Type voltage-dependent calcium channels and non–voltage-dependent calcium channels (TRP)] were able to elicit an increase in [Ca²⁺]_i somehow similar to the one induced by TCS (Supplementary Fig. S1B and S1C). According to the published data, all these compounds were activators of a unique ion channel, the non–voltage-dependent TRPA1 calcium channel (33–37). We thus assessed whether TRPA1 could be the ion channel activated by TCS in human PrSC.

We first used ruthenium red (RR) as a non-selective inhibitor of TRPA1 channel. As shown in Fig. 2A, the TCS-induced calcium entry was dose dependently blocked by RR. Moreover, the calcium increase observed by application of flufenamic acid was also inhibited by ruthenium red (Fig. 2B), suggesting the involvement of TRPA1 in flufenamic acid effects in these cells. We next used a selective inhibitor of TRPA1 channel, the compound HC-030031, which showed a dose-dependent inhibitory effect on the TCS-induced calcium response with a maximal inhibition for a concentration of 200 μmol/L (Fig. 2C and Supplementary Fig. S2A–S2F). The involvement of the TRPA1 channel in the TCS-induced calcium response was confirmed by the use of AP-18 and A967079, two other inhibitors of the TRPA1 channel (Supplementary Fig. S3A–S2F). In these experiments, we also observed a significant decrease in basal calcium concentration after the addition of HC-030031 (200 μmol/L) and RR (50 μmol/L; Fig. 2B and C). These observations suggest that TRPA1 is involved in setting the basal calcium concentration in PrSC. To confirm these results, we also used siRNAs targeting TRPA1 (siTRPA1) to downregulate the TRPA1 protein. Using RT-PCR and Western blot techniques, we first showed that the siTRPA1 transfection (25 nmol/L) of stromal cells induced a significant downregulation of the TRPA1 mRNA and protein in these cells (Fig. 2D). We then observed in calcium experiments that the TCS-induced calcium entry was almost completely blocked in siTRPA1-transfected cells (Fig. 2E), leading to the conclusion that TRPA1 is the main ion channel involved in TCS-induced calcium entry. To confirm the functionality of TRPA1 channel in prostate cancer stromal cells, the effects of a known TRPA1 agonist (Allyl Isothiocyanate, AITC) on cytosolic free calcium concentration were studied in PrSC. As shown in Fig. 2F, AITC (30 μmol/L) induced a calcium increase which was inhibited by HC-030031 (200 μmol/L).

We further investigated the potential activation of TRPA1 by TCS in PrSC by electrophysiological approaches using whole-cell patch clamp techniques. A voltage ramp of −80 to +80 mV was applied, and the resulting currents were recorded in the absence (CTL) or in the presence of TCS in the extracellular medium. As shown in Fig. 3A–D, TCS induced within few seconds after the onset of perfusion a large increase in membrane conductance at all membrane potentials. TCS-induced membrane current displayed a slight outward rectification and reversed around 0 mV. This TCS-induced current was reversible when cells were washed with TCS-free HBSS and was inhibited by the addition of HC-030031 (50 μmol/L). The action of TCS was similar to that of AITC, which was able to activate TRPA1 current in PrSC, an effect also inhibited by HC-030031 (50 μmol/L; Fig. 3E and F). To confirm the implication of TRPA1 in TCS-induced current, the effects of TCS were studied in HEK-hTRPA1 cells stably expressing the human form of TRPA1. In these cells, both TCS (10 μmol/L) and AITC (50 μmol/L) reversibly increased membrane conductance, as in stromal cells, whereas they were ineffective in vector-transfected cells (HEK-vector cells; Fig. 4A–D). Moreover, TCS- and AITC-induced currents were also inhibited by HC-030031 (50 μmol/L). These data clearly show that the antibacterial TCS induces the activation of hTRPA1.

We further studied the expression of the TRPA1 channel in PrSC and human prostate cancer epithelial cells. First, by RT-PCR analysis, we observed an absence of TRPA1 expression in epithelial cells while stromal cells obtained from different prostate cancer tissues all expressed the mRNA for TRPA1 (Fig. 5A). TRPA1 expression was also confirmed in HEK-hTRPA1 cells by Western blot experiments. As shown in Fig. 5B, the antibody revealed a band of expected size (∼130 kDa) and a band of higher molecular weight, which could represent a glycosylated form of TRPA1 protein. The same antibody was then used in Western blot experiments performed on different prostate cancer epithelial cell lines (LNCaP and PC-3), primary cultured epithelial cells (hPEC) and stromal cells of different patients (PrSC) used in the present study. As shown in Fig. 5C, the TRPA1 proteins were present only in protein extracts from stromal cells. This is in agreement with our results showing a stimulation of an HC-030031-sensitive calcium entry by TCS in stromal but not in epithelial cells. We further studied by immunofluorescence staining the expression of the TRPA1 channel protein in 7 different non-cancer (BPH) and cancer (prostate cancer) tissues using formalin-fixed, paraffin-embedded tissues. As α-actin (smooth muscle actin) expression was restricted to the stroma, the positive staining for α-actin was used to discriminate between the epithelium and the stroma. Coimmunolabeling for TRPA1 (in green) and α-actin (in red) showed overlapping staining in prostate cancer regions showing a preferential expression of TRPA1 protein in stromal cells of the cancer regions (Fig. 5D and E) and in the primary culture stromal cells (Fig. 5F). The same coimmunolabeling experiments performed in non-cancer (BPH) regions did not reveal any staining for the TRPA1 protein neither in epithelial cells nor in stromal cells (Fig. 5H), whereas α-actin (in red) was clearly expressed in stromal cells (Fig. 5G and H). A control immunofluorescence experiment is presented (Fig. 5I), where the primary antibody (anti-TRPA1) was omitted.

These data clearly show the preferential expression of TRPA1 in prostate cancer stromal cells. To extend these data to other cancers, we studied the expression of TRPA1 in 3 breast non-cancer and cancer tissues. As shown in Fig. 5J–L, TRPA1 protein is increased in cancer regions of breast tissues. These data suggest a common feature of TRPA1 expression in cancer and that the effects of TCS on TRPA1 channels could be of importance in cancer progression by inducing calcium entry and a possible secretion of factors necessary for the epithelium–stroma interactions.

To assess the influence of TCS on epithelial–stromal interactions, we examined the impact of TCS pretreatments of PrSC on the proliferation of PC-3 prostate cancer epithelial cells. In this context, using the MTS assay, we observed that when PrSC cells were cultured in CS-RPMI supplemented with 10% FCS, the PrSC-conditioned media stimulated an increase in proliferation of PC-3 cells. When PrSCs were cultured in CS-RPMI with 0.1% FCS, the PrSC-conditioned media failed to stimulate the proliferation of PC-3 cells. Interestingly, in this low serum condition, the conditioned media from TCS-pretreated PrSC induced an increase in PC-3 cells growth (Fig. 6A), whereas in high serum condition (10% FCS), the conditioned media of TCS-treated PrSC induced only a faint increase of the PC-3 cell proliferation (data not shown). These observations were confirmed when the experiments were performed by manual cell counting technique (Fig. 6B). We then carried out ELISA measurements of growth factors secretion in CTL and TCS-treated–conditioned media of PrSC. Interestingly, the incubation of stromal cells with TCS (1–10 μmol/L) clearly induced an increase in VEGF secretion (Fig. 6C). In addition, TCS-induced VEGF secretion was inhibited by the TRPA1 inhibitor HC-030031, suggesting the involvement of the TRPA1 channel in the effects of TCS on VEGF secretion by prostate stromal cells. In these experiments, the TRPA1 inhibitor HC-030031 showed a dose-dependent inhibitory effect on the TCS-induced VEGF secretion reaching a maximum of inhibition for a concentration of 200 μmol/L (Fig. 6C and Supplementary Fig. S4). The involvement of TRPA1 channel in the TCS-induced VEGF secretion was confirmed by the use of AP-18 and A967079, two other inhibitors of TRPA1 channel (Supplementary Fig. S4). These data suggest that TCS, by stimulating the secretion of factors by stromal cells, might promote prostate tumorigenesis (Fig. 6D).

The present study reports for the first time the potential impact of TCS, a widely used antibacterial in cosmetics and hygienic products, on human prostate cancer–associated stromal cells. Indeed, we have shown that TCS induced an important calcium entry in primary cultured stromal cells of human prostate cancer tissues by activating the TRPA1 channel.

The TRPA1 channel has been studied for its role in the (patho)physiology of several organs and its potential as a drug target for the management of various pathologic conditions has previously been suggested (38, 39). Initially reported to detect noxious cold (40), TRPA1 receptors have subsequently been shown to act as sensors of many environmental irritants, including acrolein, formaldehyde, zinc, pungent plant ingredients such as mustard oil, cinnamaldehyde, and allicin, as well as endogenously produced substances such as hydrogen peroxide and 4-hydroxynonenal (38). Our data suggest that the antimicrobial agent TCS has to be included in the long list of environmental factors inducing TRPA1 channel activation. However, the fixation site of TCS on the TRPA1 channel needs further investigations, using channel mutants in different intracellular and extracellular ligand binding domains of the channel.

We have also shown here for the first time the exclusive expression of TRPA1 channels in human PrSC. According to the published data, the stroma is transformed into “reactive stroma” during early prostate cancer development and it co-evolves with the cancer progression. Previous studies have established that reactive stroma biology influences prostate tumorigenesis and progression. In the present work, we show that TCS, via activation of TRPA1, induces VEGF secretion. The VEGF receptors have been shown to be present in epithelial, stromal, and endothelial cells (41), suggesting a complex interplay between the cell components of the prostate cancer microenvironment. VEGF can potentiate different functions in prostate cancer, many of which may be responsible for its growth and survival in the androgen-refractory stage, the aggressive form of the disease. The main function of VEGF is its involvement in lymphangiogenesis where the growth and sprouting of lymphatic endothelium occurs from a preexisting lymphatic vessel (42). Increased lymphatic vessel formation surrounding a tumor may provide potential routes for tumor cells to migrate to distant organs. As a first line of treatment, androgen-deprivation therapy triggers tumor growth arrest and tumor apoptosis in the primary site, and therefore an increase in the metastatic potential that helps the tumor cells to migrate to a different location and survive. Interestingly, several works have shown the potential VEGF-mediated function that may be important for the growth and survival of prostate cancer from its early phases to transition to the androgen-refractory stage (41). Thus, TCS-induced VEGF secretion might favor the progression of prostate cancer by modulating the growth and transformation of prostate epithelial cells, angiogenesis, and finally metastasis. These aspects need to be investigated in in vivo studies to address the impact of TCS on the prostate cancer epithelial–stromal and stromal–endothelial cell interactions. Secondly, the impact of TCS on the cancer cell tumor development (in nude mice) and the potential variations of vessel density in these prostate cancer cell xenografts require further investigations. In this context, our preliminary work showed that the TCS and AITC treatment of nude mice coinjected with human prostate cancer epithelial and stromal cells promoted tumor development. These data suggest that targeting the TRPA1 channel could be an alternative to the actual therapies (AR ablation) leading to relapse in the majority of cases.

In addition to environmental factors, several endogenous TRPA1 activators have also been detected. These may modulate the secretion of paracrine factors by stromal cells. TRPA1 is an attractive candidate, because it can be activated by molecules that are produced during oxidative phosphorylation, such as hydrogen peroxide (H2O2), 4-Hydroxynonenal (4-HNE), cyclopentenone prostaglandins (PGJ2; ref. 43), and Methylglyoxal, formed from triose phosphates during secondary glucose metabolism in hyperglycemic condition (39). Recently, TRPA1 has also been shown to be activated by formaldehyde (44).

Interestingly, clinical data have shown that formaldehyde concentration is elevated (2.8-fold) in the urine of patients with prostate and bladder cancer (45) and in the expired air from tumor-bearing mice and breast cancer patients (46). Formaldehyde is considered to be a cancer development risk factor. A recent work showed that in tissues from breast and lung cancer patients, the formaldehyde concentration was 0.75 mmol/L with the highest concentration of 2.35 mmol/L, concentrations being strong enough to activate TRPA1, thereby inducing the secretion of tumor-promoting factors (44). These data suggest that cancer-derived substances (formaldehyde) could have pro-cancer effects by activating TRPA1 in stromal cells, leading to the secretion of the paracrine factors themselves promoting evolution of cancers toward more aggressive forms.

Overall, our data suggest that targeting TRPA1 channel might be an advantage to limit tumor progression. Highlighting the direct effects of Triclosan on the TRPA1 channel expressed in cancer could allow considering new and/or complementary therapies targeting this channel, and/or preventive measures in the treatment of prostate cancers, which take into account the impact of these environmental factors.

No potential conflicts of interest were disclosed.

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conception and design: M. Roudbaraki

Development of methodology: S. Derouiche, M. Roudbaraki, P. Mariot, E. Vancauwenberghe, M. Warnier, G. Bidaux, P. Gosset, E. Dewailly

Acquisition of data (acquired and managed patients, provided facilities, etc.): S. Derouiche, P. Mariot, E. Vancauwenberghe, M. Warnier, P. Gosset, B. Mauroy, J.-L. Bonnal, C. Slomianny

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Roudbaraki, N. Prevarskaya, S. Derouiche, P. Mariot, G. Bidaux, P. Gosset, C. Slomianny,

Writing, review, and/or revision of the manuscript: M. Roudbaraki, S. Derouiche, E. Vancauwenberghe, G. Bidaux, P. Gosset,

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): E. Vancauwenberghe, P. Delcourt

Study supervision: M. Roudbaraki

We would like to thank Dr. L.R. Sadofsky (Cardiovascular and Respiratory Studies, The University of Hull, Castle Hill Hospital, Cottingham, Hull HU16 5JQ, UK) for the hTRPA1-pCDNA3 expression vector construction and E. Richard for the technical assistance in images analysis by confocal microscopy.

This work was supported by grants from Région Nord Pas-de-Calais (M. Roudbaraki), Institut National de la Santé et de la Recherche Médicale (INSERM; all authors received), the Ministère de l'Education Nationale de l'Enseignement Supérieur et de la Recherche (all authors received), and La Ligue Nationale Contre le Cancer (N. Prevarskaya). S. Derouiche was supported by the Région Nord Pas-de-Calais and Association pour la Recherche sur les Tumeurs de la Prostate (ARTP; S. Derouiche and M. Roudbaraki).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.