Selective inhibition of TRPM2 channel by two novel synthesized ADPR analogues

Transient receptor potential melastatin‐2 (TRPM2) channel critical for monitoring internal body temperature is implicated in the pathological processes such as neurodegeneration. However, lacking selective and potent TRPM2 inhibitors impedes investigation and validation of the channel as a drug target. To discover novel and selective TRPM2 inhibitors, a series of adenosine 5′‐diphosphoribose analogues were synthesized, and their activities and selectivity were evaluated. Whole‐cell patch‐clamp recordings were employed for screen and evaluation of synthesized compounds. Two compounds, 7i and 8a, were identified as TRPM2 inhibitors with IC 50 of 5.7 and 5.4 μm, respectively. Both 7i and 8a inhibited TRPM2 current without affecting TRPM7, TRPM8, TRPV1 and TRPV3. These two TRPM2 inhibitors can serve as new pharmacological tools for further investigation and validation of TRPM2 channel as a drug target, and the summarized structure–activity relationship (SAR) may also provide insights into further improving existing inhibitors as potential lead compounds.

Substitution of pyrophosphate linkage oxygen of ADPR has rarely been investigated. Xu et al. [29] introduced a methylenebisphosphonate linkage to cADPR and found that the activity of cADPR is very sensitive to the modifications in the pyrophosphate moiety, indicating that pyrophosphate moiety contributes to cADPR binding on its receptor. Considering the homology of cADPR and ADPR, we, in this study, synthesized a new series of ADPR analogues ( Figure 2) by integrating the modifications of pyrophosphate linkage and adenosine, in which the pyrophosphate linkage was replaced by methylenebisphosphate or difluoromethylenebisphosphate. The activity of all synthesized individual ADPR analogues was assessed. Their selectivity against other TRP channels including TRPM7, TRPM8, TRPV1 and TRPV3 was also evaluated. Two novel TRPM2 inhibitors, 7i and 8a, were identified in this study, and both compounds can be primarily used to probe physiological and pharmacological functions of TRPM2 channel, and their potency can also be further improved.

| Chemistry
All commercial chemicals and solvents were purchased from commercial suppliers of analytical grade and used without further purification. 1 H and 13 C NMR spectra were recorded on a Bruker Avance III 400 spectrometer. Chemical shifts were reported as values from an internal tetramethylsilane standard. High-resolution mass spectra (HRMS) were recorded on a Bruker Apex IV FTMS. LC/MS analyses were performed on an Agilent 1200-6110 instruments. Silica gel (200~300 mesh) was used for chromatography. HPLC (Gilson, France) was used to purify the protected and the final deprotected ADPR analogues, whereby a C18 reversedphase column (Venusil XBP C18-2, 21.5 mm×250 mm, 10 μm, Agela Technologies, Beijing, China) was employed.
Generally, the synthesis of ADPR analogues could be divided into two routes ( Figure 3). One was adenosine diphosphate (ADP) coupled with 5-tosyl ribose (Figure 3), in which ADP analogues were originally synthesized from C-2 position substituted adenosines and 5-tosyl ribose was originally synthesized from D-ribose. The other one was 5′-tosyl nucleoside coupled with ribose diphosphate (RDP) (Figure 3), in which 5-tosyl adenosines were originally synthesized from C-2 position substituted adenosine and RDP analogues were originally synthesized from Dribose. Protected ADPR analogues were synthesized employing above synthetic route and were further deprotected with aqueous HCl and purified by HPLC to give the final ADPR analogues. Detailed characterization data by 1 H, 13 [30,31] Compound 2a (0.65 g, 1.4 mmol) and tris(tetran-butylammonium) hydrogen methylenediphosphonate (1.60 g, 1.8 mmol) were mixed with 3 ml dry CH 3

| Western blot
Western blot was conducted as previously described. [34] In brief, after transfection for 24-36 hr, HEK293 cells were rinsed with ice-cold PBS and lysed in ice-cold protein lysis buffer (RIPA buffer, Beyotime Institute of Biotechnology, China) for 30 min. After centrifuging the lysates at 20,817 g at 4°C for 10 min, the supernatants were collected and stored at −80°C until use. The protein concentrations were determined using a BCA protein assay kit (Beyotime Institute of Biotechnology, China). Thirty to fifty micrograms of protein/lane was diluted in standard SDS sample buffer and subjected to electrophoresis on 12% SDSpolyacrylamide gels. Proteins were then transferred to polyvinylidene difluoride membranes (Millipore, USA), blocked with 5% BSA (Sangon Biotech, China) in Tris-buffered saline containing 0.05% Tween-20 (TBST) for 2 hr at room temperature and incubated with the primary antibody (TRPM2: Ab11168, Abcam, UK) overnight at 4°C. The membranes were then washed with TBST and incubated with the secondary antibody (goat anti-rabbit IgG-HRP: 31420, Thermo Fisher Scientific, USA). Protein bands were visualized using an Odyssey Infrared Imaging System (Li-Cor Biosciences, USA). Quantity One software (BioRad, USA) was used for densitometric scanning.
Test compound was added to either ECS or ICS with a concentration of 0.1 mm, to determine the extracellular or intracellular effect of test compound on TRPM2. The ADPR concentration was fixed at 0.1 mm in the ICS that also contains test compound when tested for intracellular effect. ECS with compound was perfused for at least 60 s before switching to acidic ECS (pH 5.0) that blocks TRPM2 current.
Change of extracellular solution was achieved using an RSC-160 system (Biologic Science Instruments, France) in which the solution changing time was about 300 ms. Cell was held at 0 mV. Voltage ramps with 500 ms duration from −100 to 100 mV were applied every 5 s. Data were acquired at 10 kHz and filtered offline at 50 Hz. Capacitive currents and series resistance were determined and corrected before each voltage ramp. For analysis, the mean of the first three ramps before channel activation was used for leak-subtraction of all subsequent current recordings.
For selectivity evaluation, test compound (0.1 mm) was added to the intracellular solution (ICS) to determine intracellular effect of compound on individual TRPM7, TRPM8, TRPV1 and TRPV3, respectively.
For TRPV3 recordings, both ECS and ICS contained 130 mm NaCl, 0.2 mm EDTA. 0.1 mm 2-APB in the ECS was used to activate TRPV3 and was washed out with standard ECS.

| Data analysis
All results from patch-clamp recordings were expressed as mean ± SEM, with n indicating the number 558 | LUO et aL. T A B L E 1 (Continued) of individual cells from at least three independent experiments. Statistical analysis was performed using twotailed paired Student's t test for comparison between groups, with p < .05 considered to be statistically significant. Prism 5 software was used for all statistical analyses.

| Chemistry
All the ADPR analogues (Table 1; Figure 2) were synthesized in two simple routes (Figure 3; Schemes 1-3). ADP analogues and RDP analogues were key intermediates for the synthesis of protected ADPR analogues. The protected ADPR analogues were deprotected with aqueous HCl to give final ADPR analogues.

| Intracellular, but not extracellular, inhibition of TRPM2 by two novel ADPR analogues 7i and 8a
To screen for TRPM2 inhibitors, the effects of all synthesized ADPR analogues (Table 1) were evaluated on TRPM2 channels stably expressed in HEK293 cells by whole-cell patch-clamp recordings. In HEK293 cells, the expression of TRPM2 channel proteins was detected by Western blot analysis (Figure 4a). It is known that TRPM2 is only activated by intracellular ADPR that binds directly to TRPM2 channel's enzymatic NUDT9-H domain in the C-terminal tail. [6] We tested whether extracellular administration of individual ADPR analogues had any effect on TRPM2 currents (Table  S1). When administered extracellularly (100 μm), none of the synthesized ADPR analogues showed any inhibition on TRPM2 currents induced by intracellular ADPR, in which compounds 7i and 8a serving as representatives (Figure 4b).
In contrast, intracellular administration of 7i or 8a at 3 μm resulted in a reduction of TRPM2 current by 39% and 36%, respectively, as compared with 100 μm ADPR as a control (Figure 4c-e, p < .005). Applying different concentrations of 7i or 8a gave rise to a dose-dependent inhibition of TRPM2 current with IC 50 s for 7i at 5.7 μm and 8a at 5.4 μm (Figure 4f). These results suggest compounds 7i and 8a inhibited TRPM2 intracellularly, possibly through competing with ADPR by binding to the intracellular NUDT9-H domain.

| Lack of inhibitory effect on other TRP channels by compound 7i or 8a
To evaluate the compounds selectivity, the effects of both 7i and 8a on other TRP channels, including the closely related channels TRPM7 and TRPM8, and the more distantly related channels TRPV1 and TRPV3, were tested. TRPM7 currents overexpressed in HEK293T cells were activated by low concentrations of Ca 2+ and Mg 2+ (both in 0.1 mm) in the extracellular solution (ECS). Compared with the control without 7i (Figure 5k), the intracellular administration of 7i (100 μm) showed no effect on TRPM7 currents (Figure 5a). Similarly, adding 100 μm 7i compound in the pipette did not affect TRPM8 pharmacology in response to either activator, menthol (1 mm) or blocker 2-APB (100 μm) (Figure 5b,l). Adding compound 7i (100 μm) in the pipette had no effect on activating TRPV1 induced by capsaicin, although appeared to delay the channel activation (Figure 5c,m). The intracellular effect of compound 7i on TRPV3 was also tested. As shown in Figure 5d,n, application of compound 7i (100 μm) in pipette had no effect on TRPV3 currents activated by 100 μm  (Figure 5a-e).
Similarly, the effects of compound 8a on TRPM7, TRPM8, TRPV1 and TRPV3 were also tested. Intracellular administrations of compound 8a (100 μm) did not show any inhibitory effect on these TRP channels (Figure 5f-j,k-n). Taken together, these results indicate that compounds 7i and 8a are specific in inhibiting TRPM2 channel without affecting other TRP channels.
To begin with adenosine modification, ADPR analogues (7h, 7g, 10h, 10g) with 6-methylamino or 6-dimethylamino substitution showed no inhibitory activity, indicating the necessity of 6-amino for inhibitory activity. This is in accordance with Moreau's report, in which 6-oxygen substituted ADPR (IDPR) shows no inhibitory activity. [28] Substitutions at C-2 position of purine base showed different influences. Compounds 7i (2-chloro) and 8a (2-hydrogen) exhibited moderate activity, while other substitutions such as 2-bromo, 2-iodo, 2-amino and 2-methoxy (7a~7e, 7h, 7j~7o, 10a~10o) did not. Thus, more kinds of substitutions should be investigated to further reveal the effect of C-2 position modification on inhibitory activity, which may lead to novel potent inhibitors. As to adenosine ribose part, neither compound 13 nor compound 14 with 2′-methoxy substitution showed any inhibitory activity indicating the importance of C-2′ position for analogues binding. With regard to pyrophosphate, both two active compounds 7i and 8a possess a substituted pyrophosphate (O → CH 2 or CF 2 ). In contrast, inactive compound 7o reserves oxygen-linked pyrophosphate. Previous report considered pyrophosphate as an activity-keeping group and none of the sulphonamide, squarate or triazole substituted ADPR analogues showed any activity. [28] Therefore, appropriate modifications on the pyrophosphate can be tolerable.
As to the terminal ribose, it has been shown that a fivemembered ring but the hydroxyl groups of the terminal ribose plays a critical role in filling the binding site. [28] Here, compounds 7i and 8a (both terminal ribose hydroxyl protected) showed inhibitory activity, while their corresponding deprotected products 10i and 10a lost inhibitory activity. Taken together, we suppose, applying saturated ring mimicking essential terminal ribose is feasible, which may improve the inhibitory activity.

| DISCUSSION
Searching for selective TRPM2 inhibitors as tool molecules is critical for probing pharmacological functions of the channel.
Up to date, there are non-specific TRPM2 inhibitors such as 2-APB, clotrimazole and FFA. These inhibitors exhibit low potency on TRPM2 channel and also are lack of selectivity on other TRP channels. Moreau et al. previously reported several ADPR analogues as specific TRPM2 inhibitors without interfering Ca 2+ release induced by cADPR, NAADP or IP 3 . [28] However, the selectivity of these analogues against other TRP channels remains unclear.
To discover novel TRPM2 selective inhibitors, we systematically synthesized a new series of ADPR analogues (Schemes 1-3) in this study. Modifications were mainly focused on pyrophosphate linkage and adenosine, to improve the potency and membrane permeability. Two major routes were employed for ADPR analogue synthesis ( Figure 3). Both synthetic routes are relative simple, of which ADP analogues and RDP analogues were the key intermediates (Schemes 1-3). These two key intermediates were obtained in moderate yield (50%~70%) and can be employed for ADPR analogues derivatization easily. Compared with previous report of which ADPR analogues are synthesized through quite a number of individual synthetic routes, [28] a more efficient way for synthesis of ADPR analogues was developed in this F I G U R E 5 Selectivity evaluations of intracellular 100 μm 7i (a-d, n = 3) or 100 μm 8a (f-i, n = 3) in the pipette on other TRP channels (TRPM7, TRPM8, TRPV1, TRPV3). (k-n) The effect of controls with their corresponding agonist on the other TRP channels (TRPM7, TRPM8, TRPV1, TRPV3, n = 4), without 7i or 8a in pipette. (e) Summary of the effect of compound 7i on TRP channels (n = 3), compared with their corresponding inhibitor (100 μm 2-APB for TRPM2 and TRPM8; 2 mm of Ca 2+ and 1 mm Mg 2+ for TRPM7; 100 μm ruthenium red for TRPV1; and washout with bath solution for TRPV3). (j) Summary for the effect of compound 8a on TRP channels (n = 3), compared with their corresponding inhibitor (100 μm 2-APB for TRPM2 and TRPM8; 2 mm of Ca 2+ and 1 mm Mg 2+ for TRPM7; 100 μm ruthenium red for TRPV1; and washout with bath solution for TRPV3) [Colour figure can be viewed at wileyonlinelibrary.com] F I G U R E 6 Structure-activity relationship of ADPR-analogue inhibitor. Purine C-2 position (R1, red plaque): appropriate substitutions may contribute to inhibitory activity; purine C-6 position (R2, pink plaque): amino is important for inhibitory activity; ribose 2′-position (blue plaque): hydroxyl is important for inhibitory activity; pyrophosphate (yellow plaque): pyrophosphate is important, but appropriate modifications are tolerable; terminal ribose (purple plaque): ribose is necessary, and appropriate ring substitution may increase activity [Colour figure can be viewed at wileyonlinelibrary.com] study, though esterification and deprotection steps give the desired products (protected ADPR analogues and ADPR analogues) in low yield.
Totally 39 new ADPR analogues were synthesized. Compounds 7i and 8a were confirmed as TRPM2 inhibitors with moderate activity, exerting intracellular inhibition of TRPM2 current, and demonstrating selectivity over other TRP members including TRPM7, TRPM8, TRPV1 and TRPV3. The characters of these two inhibitors probably attribute to the nature of ADPR analogues. It is well known that ADPR gates TRPM2 through binding to the cytosolic C-terminal NUDT9-H domain which is a unique structure domain among TRP channels. [5] The two inhibitors 7i and 8a likely compete with ADPR for binding to the NUDT9-H domain, serving as intracellular inhibitors. These two new inhibitors are firstly identified TRP-subtype selective TRPM2 inhibitors, which are selective over TRPM7, TRPM8, TRPV1 and TRPV3, providing new tool molecules for probing pharmacological functions of TRPM2 channel. Previous reported nonselective inhibitors such as FFA and 2-APB also show significant effects on other ion channels. FFA can also inhibit TRPM3, TRPM4 and TRPM5, [35,36] while 2-APB also activates TRPV1, TRPV2 and TRPV3. [25] When employing selective inhibitors 7i or 8a on probing TRPM2 function, these unwanted side-effects will be eliminated, so that the results will be more relevant to TRPM2 channel. Though the two inhibitors still exhibit poor membrane permeability limiting their extracellular application in probing TRPM2 function, applying liposomes or nucleotide prodrug technique will allow the transport of 7i or 8a across the membrane. These newly identified TRP-subtype selective inhibitors can serve as new pharmacological tools for probing TRPM2 channel.
What is more, the identification of ADPR-analoguebased TRPM2 inhibitors provides further insights into optimizing potency and membrane permeability. With regarding to improving membrane permeability, none of the ADPR analogues show any inhibitory activity when administered extracellularly; thus, it is still insufficient though substituting pyrophosphate oxygen with methylene or difluoromethylene. Recently, a neutral triazole substituted cADPR (P 2 O 6 − → triazole) that retains its ability to activate Ca 2+ release was reported by Swarbrick et al. [37] This encourages employing more kinds of neutral substitutes, though the pyrophosphate group seems critical for keeping the activity. For potency optimizing, more kinds of substitutions at C-2 position should be investigated, which may lead to novel potent inhibitors. Other successful cases that took purine C-2 position substitution support this hypothesis. Fischer et al. reported 2-chloro-adenosine-5′-O-α-boranodiphosphate (2-Cl-α-BH 3 -ADP) as a potent agonist (EC 50 = 7 nm) of P2Y1 receptor, whose endogenous substrate is adenosine diphosphate (ADP). [38] Similarly, 2-methylthio-AMP was reported as an inhibitor of P2Y12 receptor, for which ADP serves as the endogenous substrate as well. [39,40] As far as to terminal ribose, neither ribose nor isopropylidene-protected ribose determines the inhibitory activity, thus employing more kinds of rings is an alternative way.

| CONCLUSIONS
Searching for selective TRPM2 inhibitors as tool molecules is critical for probing pharmacological functions of the channel. In this study, we synthesized 39 new ADPR analogues, of which the modifications were mainly focused on pyrophosphate linkage and adenosine. Their effects on TRPM2 channel were evaluated. Compounds 7i and 8a were confirmed as TRPM2 inhibitors with moderate activity, exerting intracellular inhibition of TRPM2 current, and demonstrating selectivity over other TRP members including TRPM7, TRPM8, TRPV1 and TRPV3. Besides, a more sufficient SAR of ADPRanalogue TRPM2 inhibitors was summarized in this study.
In summary, we synthesized and obtained two novel subtype selective TRPM2 inhibitors. The newly identified TRPM2 inhibitors can serve as new pharmacological tools for further investigation and validation of TRPM2 channel as a drug target. The summarized SAR may also provide insights into further improving existing inhibitors as potential lead compounds.