Inhibition of Diacylglycerol–Sensitive TRPC Channels by Synthetic and Natural Steroids

TRPC channels are a family of nonselective cation channels that regulate ion homeostasis and intracellular Ca2+ signaling in numerous cell types. Important physiological functions such as vasoregulation, neuronal growth, and pheromone recognition have been assigned to this class of ion channels. Despite their physiological relevance, few selective pharmacological tools are available to study TRPC channel function. We, therefore, screened a selection of pharmacologically active compounds for TRPC modulating activity. We found that the synthetic gestagen norgestimate inhibited diacylglycerol-sensitive TRPC3 and TRPC6 with IC50s of 3–5 µM, while half-maximal inhibition of TRPC5 required significantly higher compound concentrations (>10 µM). Norgestimate blocked TRPC-mediated vasopressin-induced cation currents in A7r5 smooth muscle cells and caused vasorelaxation of isolated rat aorta, indicating that norgestimate could be an interesting tool for the investigation of TRP channel function in native cells and tissues. The steroid hormone progesterone, which is structurally related to norgestimate, also inhibited TRPC channel activity with IC50s ranging from 6 to 18 µM but showed little subtype selectivity. Thus, TRPC channel inhibition by high gestational levels of progesterone may contribute to the physiological decrease of uterine contractility and immunosuppression during pregnancy.


Introduction
Transient receptor potential canonical (TRPC) channels belong to the TRP cation channel superfamily. Seven TRPC channels, TRPC1-7, have been found in rodents. They all conduct Ca 2+ in addition to monovalent cations and can be activated via membrane receptors linked to phospholipase C signaling. Downstream of phospholipase C the activation mechanisms of TRPC channels are not well defined but one subclass, namely TRPC3, 26 and 27, as well as the structurally more distinct TRPC2 respond to the phospholipase C hydrolysis product diacylglycerol. The remaining family members, however, are diacylglycerol-insensitive (reviewed in [1][2][3]).
Despite the substantial progress made, important questions regarding TRPC channel function and regulation remain, and it would be highly desirable to verify and extent studies in genetically modified mice by pharmacological means in non-engineered animals. Unfortunately, such experiments have proved difficult due to the lack of specific compounds that modulate TRPC channels. Tools such as SKF96365 [13,14], KB-R7943 [15] or BTP-2 [16] do not discriminate between TRPC homologs and also alter the activity of other ion channels or transporters [14,17]. Only recently the first subtype -specific TRPC inhibitor was discovered. This compound, the pyrazole Pyr3, inhibited TRPC3 with a half maximal inhibitory concentration (IC 50 ) of 0.7 mM without having effects on other TRPC channels. Using Pyr3 an involvement of TRPC3 in the development of cardiac hypertrophy could be demonstrated in vivo [18].
To enrich the repertoire of pharmacological tool compounds for TRPC channels and to identify new natural channel modulators, we screened a library of pharmacologically and biologically active compounds for their effects on TRPC6-mediated Ca 2+ entry. We found that norgestimate, a synthetic steroid and active ingredient of certain contraceptives, preferentially inhibited TRPC3 and 26 channels at low micromolar concentrations. The endogenous gestagen progesterone also inhibited TRPC channels but showed little subtype selectivity.
Interestingly, a recent study by Majeed et al. [19] demonstrated inhibition of TRPC5 by neurosteroids including pregnenolone sulphate and progesterone. Our data now show that steroid modulation is a common feature of TRPC family members and reveal structural determinants of selective TRPC channel inhibition by progestins.
Hence, TRPC channels may contribute to diverse steroid actions ranging from progesterone-induced vascular remodelling and decrease in uterine contractility during pregnancy to cardiovascular side-effects of oral contraceptives.
For measurement of norgestimate effects on adrenoceptor signaling a CHO cell line stably expressing human a 1A -receptors (GenBank accession number NM_000680.2) under control of a CMV promoter was used. CHO cells were kept in HAM's F-12 medium (with glutaMAX I) supplemented with 10% (v/v) fetal bovine serum (Biochrom), 1 mM glutamine, and 0.6 mg/ml geneticin.

Measurement of intracellular calcium concentration ([Ca 2+ ] i )
Fluo-4 measurements. Cells grown to an almost confluent monolayer on black poly-D-lysine coated 96-well plates (Greiner, Frickenhausen, Germany) were washed with standard extracellular solution (140 mM NaCl, 1 mM MgCl 2 , 5.4 mM KCl, 2 mM CaCl 2 , 10 mM HEPES, 10 mM glucose, pH 7.35) and stained (30 min, room temperature) with dye solution (2 mM fluo-4 AM, 0.02% pluronic F127, 0.1% bovine serum albumin in standard extracellular solution). Cells were washed and either incubated with standard extracellular solution only or with different concentrations of test compounds for 10 min. Fluo-4 fluorescence was excited at 488 nm with an argon laser and measured using a fluorometric imaging plate reader (Molecular Devices, Sunnyvale, USA). All fluorometric measurements were performed at room temperature.
Ca 2+ entry into TRPC3 CHO and TRPC6 HEK-FITR cells was elicited by application of the diacylglycerol analog 1-oleoyl-2acetyl-sn-glycerol (OAG). For calculation of norgestimate-and progesterone-induced inhibition of TRPC3-and TRPC6-mediated Ca 2+ entry fluorescence values were plotted versus time and the area under the curve was considered as a measure of Ca 2+ influx. Baseline fluorescence in non-OAG-stimulated cells and OAGstimulated Ca 2+ influx into control cells not treated with inhibitors were defined as 0% and 100% influx, respectively.
TRPC4-and TRPC5-mediated Ca 2+ influx was stimulated by application of trypsin that activates endogenous G q -coupled protease activated receptors in HEK cells [20]. As trypsin, via phospholipase C activation, causes Ca 2+ release from internal stores (PI response) and subsequent Ca 2+ entry through native store-operated channels, these components had to be subtracted from the total fluorescence signal in order to determine TRPCmediated Ca 2+ entry. For this purpose trypsin-induced Ca 2+ responses in non-induced TRPC4-and TRPC5 HEK-FITR cells lacking functional expression of the TRPC channels were measured in parallel to responses in induced cells and were subtracted from the fluorescence traces. Compound effects were then calculated from the area under the curve and normalized to the Ca 2+ influx into trypsin-stimulated, non-compound-treated cells.
[Ca 2+ ] i measurements in CHO cells stably expressing human a 1A -receptors were performed essentially as described for HEK-FITR cells. The a 1 -agonist phenylephrine was used for receptor stimulation.

Electrophysiological techniques
The whole-cell patch clamp technique [21] was employed to measure ion currents from single cells. Heat-polished patch pipettes with resistances of 2-4 MV were pulled from borosilicate glass capillaries (Hilgenberg, Malsfeld, Germany) using a DMZ-Universal puller (Zeitz-Instruments, Munich, Germany) and filled with standard intracellular solution containing: 120 mM CsOH, 120 mM gluconic acid, 2 mM MgCl 2 , 3 mM CaCl 2 , 5 mM Cs 4 -BAPTA, 10 mM HEPES (pH 7.4 adjusted with gluconic acid). For agonist-independent stimulation of TRPC currents the intracellular solution was supplemented with 30 mM AlCl 3 and 10 mM NaF and cells were infused with AlF 4 2 through the patch pipette. AlF 4 2 is a non-specific activator of heterotrimeric G proteins that consequently leads to increased phospholipase C activity and TRPC channel activation.
For measurement of [Arg 8 ]-vasopressin (AVP) -induced currents in A7r5 cells the Ca 2+ concentration of the extracellular solution was lowered to 0.2 mM.
Cells grown on poly-L-lysine-coated coverslips were continuously superfused with extracellular solution and substances were applied using an ALA BPS-8 perfusion system (ALA Scientific Instruments, Westbury, USA). Whole-cell recordings were performed with an EPC-10 amplifier and Pulse software (HEKA, Lambrecht, Germany). Cells were held at 270 mV, and currentvoltage (I-V) relationships were routinely measured every 2 or 3 s by applying voltage ramps (180 ms) from 2100 mV to +80 mV. Data was acquired at 6.67 kHz and filtered with 2.22 kHz. After establishing the whole cell configuration the series resistance was usually ,10 MV and was compensated 50-70%. All experiments were performed at room temperature.

Ethics statement
Animals were used for organ retrieval only and were sacrificed according to Sanofi-Aventis Ethical Committee guidelines and to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Sanofi-Aventis is an authorized institution to house and handle laboratory animals according 111 German Animal Welfare Act, and has a nominated animal welfare officer. Sanofi-Aventis is also AAALAC (Association for Assessment and Accreditation of Laboratory Animal Care International) accredited. The competent authority (Regierungsprä sidium Darmstadt) has been notified of the procedure under protocol number T04-01/102.

In vitro vascular function
Adult male Wistar-Unilever rats (8-11 weeks old; Harlan Winkelmann, Borchen, Germany) were sacrificed by decapitation. Thoracic aortas were excised quickly, transferred to cold (4uC), carbogen (95% O 2 and 5% CO 2 ) bubbled physiological salt solution (119 mM NaCl, 4.7 mM KCl, 1.2 KH 2 PO 4 , 1.2 mM MgSO 4 , 1.6 mM CaCl 2 , 25 mM NaHCO 3 , 11 mM glucose) and rinsed. After removal of connective tissue and perivascular fat aortas were dissected in 5 mm rings, connected to force transducers and equilibrated in carbogen bubbled physiological salt solution at 37uC for 15 min. Aortic rings were set at 1000 mg passive tension and contractile forces were then measured isometrically using standard bath procedures as described earlier [22]. Vessels strongly contracting after application of 60 mM KCl were defined as intact and used for further experiments. To minimize the contribution of endothelium-derived nitric oxide to compound-induced force changes experiments were performed in the presence of the nitric oxide synthase inhibitor nitro-L-arginine methyl ester (L-NAME).

Statistics
For statistical analysis, analysis of variance was performed with Origin 6.0 software (Microcal Software Inc., Northampton, USA). P values less than 0.05 were considered as statistically significant. IC 50 were calculated with SigmaPlot (Systat software, San Jose, USA) using the sigmoidal Hill-model: f = 100 (C n /(IC 50 n +C n )); where f is the inhibition, C is the applied drug concentration and n is the Hill coefficient. The half maximal effective concentration (EC 50 ) of norgestimate-induced vasorelaxation was calculated analogously with R = y 0 +y 1 C n /(EC 50 n +C n ); where R is relaxation, y o and y 1 are constants, C is the applied drug concentration and n is the Hill coefficient. Averaged data is expressed as means 6 SEM.

Results
In order to discover compounds that modulate TRPC activity we developed fluorescence-based assays that allowed to measure TRPC-mediated Ca 2+ entry using fluorometric imaging plate reader technology. Screening a selection of pharmacologically and biologically active compounds using these assays identified norgestimate as a potent TRPC3 and 26 channel inhibitor. Norgestimate blocked OAG-induced Ca 2+ -entry in TRPC3 and TRPC6 expressing cells with similar potency. IC 50 s amounted to 3.0 and 5.2 mM, respectively (Figure 1). To directly investigate the effect of norgestimate on TRPC currents we performed patch clamp experiments using the same TRPC expressing cell lines used for Ca 2+ measurements. These experiments confirmed that 10 mM norgestimate almost completely inhibited TRPC3 and 26 (Figure 2A-C). Further evaluation of the dose dependence of TRPC6 current inhibition by norgestimate yielded an IC 50 of 3.06 mM ( Figure 2D). Thus, both direct current measurements and indirect fluorescence experiments showed a similar inhibitory activity of norgestimate towards TRPC3 and 26.
Steroid-sensitivity had previously been established for TRPC5 [19] which together with its close homolog TRPC4 defines a distinct subclass of TRPC channels that is not activated by diacylglycerol. We wanted to test the subtype selectivity of norgestimate and, therefore, studied the effect of the compound on TRPC5.
As illustrated in Figure 3 norgestimate suppressed TRPC5mediated currents with an IC 50 of 13.6 mM, indicating a considerably lower potency of the synthetic gestagen towards this diacylglycerol-insensitive TRPC channel. Unfortunately, we were unable to determine the effects of norgestimate on TRPC4 currents since compound effects at 10 mM were apparently small, irreversible and could not be separated from the significant rundown of TRPC4-mediated currents.
Having established the prominent inhibitory effect of norgestimate on recombinant TRPC6 we next asked whether the compound similarly affects native TRPC6-mediated currents. In the aortic smooth muscle cell line A7r5 TRPC6 and/or TRPC6/7 channel complexes had previously been demonstrated to underlie vasopressin V 1a -receptor-activated cation currents [23][24][25]. Hence, we chose this model and measured inhibition of AVPinduced receptor-operated currents by norgestimate ( Figure 4A). When 10 mM norgestimate was applied to AVP-stimulated A7r5 cells, receptor-operated cation currents measured at 270 mV were reversibly reduced by 86.566.0% (n = 8), in good agreement with the effect on recombinant TRPC6 channels. Since AVPinduced Ca 2+ release was not sensitive to norgestimate ( Figure 4B), receptor-operated channel inhibition is likely due to a direct effect on the channels proteins rather than interaction of the compound with upstream signaling elements.
We extended the study of functional effects of norgestimate to intact vessel segments derived from thoracic rat aorta. Several investigators have reported expression of TRPC3/6 in this preparation [5,26]. Isometric force measurement showed that norgestimate concentration-dependently relaxed aortic segments pre-contracted with the a 1 -receptor agonist phenylephrine ( Figure 5). To exclude the possibility that norgestimate-induced relaxation was due to direct receptor blockage we tested the effect of the compound on recombinant a 1 -receptors expressed in CHO cells. Surprisingly, we observed that norgestimate significantly stimulated phenylephrine-induced Ca 2+ release in this model ( Figure S1). This stimulatory action on a 1 -receptor signaling may contribute to the relatively low vasorelaxing activity of norgestimate in isolated aortic rings. Nonetheless, our results indicate that norgestimate can be used to modulate TRPC-dependent functions in situ.
Recent experiments revealed that TRPC5 is inhibited by certain steroids including progesterone in a stereo-selective manner [19]. Because progesterone is known to induce vasorelaxation of diverse vessels, we wondered if part of this nongenomic progesterone effect may be due to inhibition of vascular TRPC channels. Therefore, we investigated the effects of progesterone on recombinant TRPC3, 24, 25 and 26. Figure 6 illustrates that progesterone indeed attenuated Ca 2+ entry through all TRPCs tested. 12.4 mM and 18 mM progesterone were required for half maximal inhibition of OAG-induced [Ca 2+ ] i transients in TRPC3-and TRPC6 expressing cells. The hormone also suppressed trypsin-induced Ca 2+ entry via TRPC4 and 25 with IC 50 s of 6.2 and 11.8 mM, respectively, demonstrating that diacylglycerol-insensitive TRPC family members were at least as sensitive to progesterone treatment as TRPC3 and 26.
These experiments confirm TRPC channels as novel targets of steroids and raise the possibility that TRPCs contribute to the vasoregulatory properties of these substances.

Discussion
Our search for new pharmacological TRPC channel modulators led to the discovery of two steroids, namely norgestimate and progesterone, which differentially inhibited TRPC-mediated currents and Ca 2+ influx. Norgestimate was found to be a potent inhibitor of diacylglycerol-sensitive TRPC channels. The compound blocked TRPC6 in patch-clamp experiments with an IC 50 of ,3 mM and showed a similar or slightly higher inhibitory potency on TRPC3 currents and TRPC3-mediated Ca 2+ entry, respectively. Diacylglycerolinsensitive TRPC5 currents were also reduced by norgestimate but required a considerably higher concentration (,14 mM) for half maximal inhibition.
The effects of norgestimate on TRPC currents occurred rapidly, suggesting a direct modulation of channel proteins rather than genomic effects via steroid receptors. We could not discern a significant interference of 10 mM norgestimate with cytosolic calcium responses elicited by vasopressin receptors in A7r5 cells. However, the compound up-regulated a 1 -receptor induced Ca 2+ release in CHO cells. Thus, further evaluation of norgestimate specificity is needed to fully estimate the potential applications and limitations of the compound as a blocker of diacylglycerol-sensitive TRPC channels. Nevertheless, the rather stimulatory effects of norgestimate on intracellular Ca 2+ release exclude inositol 1,4,5trisphosphate receptor antagonism or inhibition of the G q/11 / phospholipase C signaling cascade as mechanisms of TRPC channel inhibition.  Notably, norgestimate inhibited recombinant TRPC6 channels and endogenous vasopressin receptor-activated cation currents in A7R5 cells with almost the same potency. There is evidence suggesting that the AVP-induced cation current in A7R5 is carried by TRPC6/7 heteromeric channels [25]. Based on this assumption it seems plausible, that other possible heteromers formed by diacylglycerol-sensitive TRPC proteins share a similar high sensitivity to norgestimate. Therefore, the steroid will be a useful tool for characterisation of native TRPC3/6/7 containing channel complexes. In line with this view, our experiments on isolated rat aorta indicate that norgestimate can be used to investigate TRPCmediated functions in tissue preparations.
Norgestimate is a progestin (a synthetic gestagen). Combined with ethinyl estradiol it is a component of oral contraceptives widely used in humans. It would be of great interest to investigate whether other synthetic gestagens share the TRPC-blocking activity of norgestimate and whether any of the numerous cardiovascular effects of norgestimate-or other progestin-containing oral contraceptives can be assigned to TRP channel inhibition.
Prior to this study, Maheed et al. [19] showed that TRPC5 is inhibited by the pregnancy-maintaining hormone progesterone and other neurosteroids. Our Ca 2+ measurements extent these findings and demonstrate that in addition to TRPC5 progesterone also inhibited TRPC3, 24, and 26. However, in contrast to norgestimate no preferential effects of progesterone on diacylglycerol-sensitive channel subtypes were observed. This differential sensitivity of TRPC family members to progesterone and norgestimate provides further support to the notion that channel-steroid interaction is based on direct, structure-driven binding and not caused by unspecific steroid-induced perturbation of membrane lipid bilayers.
The physiological significance of TRPC channel regulation by progesterone remains to be established. Our data suggest that in addition to neurological effects of neurosteroids [19] TRPCs may contribute to vascular effects of steroid hormones. Several studies have shown that progesterone rapidly relaxed vessels, e.g. pig coronary arteries [27] and rat aorta [28,29]. This vasorelaxation is endothelium-independent and mediated at least partly through inhibition of L-type Ca 2+ channels [30,31]. Nevertheless, an involvement of other ion channels, including receptor-operated cation channels, has also been proposed [28,29] and would be compatible with the widespread expression of TRPC channels in smooth muscle (reviewed in [32,33]).
It is of note, that several TRPC channel proteins, in particular TRPC1, 24, and 26 have been detected in human myometrium [34][35][36]. Pregnancy is characterized by elevated progesterone plasma levels that can reach up to 1 mM [31,37]. Given that progesterone is highly lipophilic and has a large volume of distribution [38], tissue concentrations are likely even higher and may achieve the effective concentrations for TRPC channel inhibition. Therefore, it is conceivable that TRPC channel blockade by high gestational progesterone concentrations is part of the adaptive process that limits uterine contractility during pregnancy [39].
With estimated 15 mM [40], progesterone concentrations within the placenta are even higher than in other tissues. Such levels of progesterone have demonstrated immuno-suppressive effects important for prevention of fetal-maternal rejection in utero [41]. As many ion channels including TRPCs are involved in immunomodulation [42] attenuation of TRPC-dependent Ca 2+ signaling could play a role in progesterone-induced immunosuppression.
To further corroborate these hypotheses studies of TRPC channel regulation by steroid hormones in native tissues are highly desirable.