Functional expression of TRPM7 as a Ca2+ influx pathway in adipocytes

Abstract In adipocytes, intracellular Ca2+ and Mg2+ modulates physiological functions, such as insulin action and the secretion of adipokines. TRPM7 is a Ca2+/Mg2+‐permeable non‐selective cation channel. TRPM7 mRNA is highly expressed in adipose tissue, however, its functional expression in adipocytes remains to be elucidated. In this study, we demonstrated for the first time that TRPM7 was functionally expressed in both freshly isolated white adipocytes and in 3T3‐L1 adipocytes differentiated from a 3T3‐L1 pre‐adipocyte cell line by whole‐cell patch‐clamp recordings. Consistent with known properties of TRPM7 current, the current in adipocytes was activated by the elimination of extracellular divalent cations and the reduction of intracellular free Mg2+ concentrations, and was inhibited by the TRPM7 inhibitors, 2‐aminoethyl diphenylborinate (2‐APB), hydrogen peroxide (H2O2), N‐methyl maleimide (NMM), NS8593, and 2‐amino‐2‐[2‐(4‐octylphenyl)ethyl]‐1,3‐propanediol (FTY720). Treatment with small‐interfering (si) RNA targeting TRPM7 resulted in a reduction in the current to 23 ± 7% of nontargeting siRNA‐treated adipocytes. Moreover a TRPM7 activator, naltriben, increased the TRPM7‐like current and [Ca2+]i in 3T3‐L1 adipocytes but not in TRPM7‐knockdown adipocytes. These findings indicate that TRPM7 is functionally expressed, and plays a role as a Ca2+ influx pathway in adipocytes.


Introduction
Transient receptor potential (TRP) proteins are a superfamily of cation channels, most of which are Ca 2+ -permeable and thereby involved in Ca 2+ -dependent signaling pathways. TRP melastatin (TRPM) 7 channel permeates Ca 2+ and Mg 2+ , and is expressed in most cell types (Nadler et al. 2001;Runnels et al. 2001;Fonfria et al. 2006;Kunert-Keil et al. 2006). TRPM7 is a very unique membrane protein in that it is a channel harboring an alpha-kinase at the carboxyl-terminal. Down-regulation of TRPM7 in cells has been reported to affect diverse physiological functions, including cell proliferation, survival, and migration (reviewed in Bates-Withers et al. 2011;Fleig and Chubanov 2014).
In adult humans, TRPM7 mRNA is most highly expressed in adipose tissue (Fonfria et al. 2006). In pre-adipocytes, it has been demonstrated that the pharmacological blockade or siRNA of TRPM7 inhibited proliferation and adipogenesis via reduction in the expression of cyclin D1 and cyclin E kinases, and phosphorylated ERK1/2 and Akt (Chen et al. 2014). In contrast, it remains unclear whether TRPM7 is functionally expressed and plays physiological roles in mature adipocytes. It has been reported in mature adipocytes that [Ca 2+ ] i affects insulin signaling (Draznin et al. 1987;Kelly et al. 1989;Whitehead et al. 2001;Worrall and Olefsky 2002), lipolysis (Tebar et al. 1996;Xue et al. 1998), and the secretion of adipokines, including leptin (Levy et al. 2000;Cammisotto and Bukowiecki 2004), adiponectin (Sukumar et al. 2012;Komai et al. 2014;El Hachmane et al. 2015), and fatty acid binding protein 4 (FABP4) (Schlottmann et al. 2014). Calcium plays permissive roles in these physiological processes rather than triggers them; however, the regulation in [Ca 2+ ] i levels appears to be important for proper metabolic signaling in mature adipocytes. Moreover it has been reported that intracellular Mg 2+ deficiency is related to metabolic disorders, such as diabetes mellitus (Resnick et al. 1993;Takaya et al. 2004; Barbagallo and Dominguez 2007). Thus, if it is expressed in adipocytes, TRPM7 might affect the physiological functions through the regulation of [Ca 2+ ] i and/or [Mg 2+ ] i .
Here we demonstrated that TRPM7 was functionally expressed in both freshly isolated mature adipocytes and differentiated 3T3-L1 adipocytes by patch-clamp experiments. Consistently, ratiometric Ca 2+ imaging with Fura-2 revealed that the augmentation of TRPM7 current by a TRPM7 activator, naltriben (Hofmann et al. 2014), induced an increase in [Ca 2+ ] i in differentiated 3T3-L1 adipocytes.
Portions of this work have appeared previously in abstract form (Inoue et al. 2017).

Animals
All experimental procedures involving animals were approved in advance by the Institutional Animal Care and Use Committee of Tokyo Medical University and by the Ethics Review Committee for Animal Care and Experimentation of the National Institute for Physiological Sciences. C57BL/6 mice aged 10 weeks were purchased from Japan SLC (Shizuoka, Japan).

Preparation of white adipocytes and differentiated 3T3-L1 cells
Freshly isolated murine white adipocytes were prepared as described previously (Inoue et al. 2010). 3T3-L1 cells were grown in Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich) supplemented with 10% calf serum, Gluta-MAX TM -I (Thermo Fisher Scientific), and 1% penicillinstreptomycin (Thermo Fisher Scientific) at 37°C in 5% CO 2 under humidified conditions. Three days after confluence, cells were differentiated into adipocytes with a differentiation medium, consisting of DMEM supplemented with 10% fetal bovine serum, 2 mmol/L glutamine, penicillin-streptomycin, and a differentiation cocktail (0.25 lmol/L DEX, 10 lg/mL insulin, and 0.5 mmol/L IBMX). After 42 h, the medium was changed to a fresh differentiation medium without the differentiation cocktail, and then replaced with fresh differentiation medium every 2 days. Differentiated 3T3-L1 cells were harvested by trypsin digestion and plated on Matrigel â (Corning, Corning, NY)-coated glass coverslips for patch-clamp experiments and Ca 2+ imaging experiments, or for siRNA transfection.

siRNA transfection
Transfection of TRPM7-targeted siRNA into 3T3-L1 adipocytes was performed as described previously (Kilroy et al. 2009). In a collagen-coated 48-well plate (BD, Franklin Lakes, NJ), 10 lL Opti-MEM (Thermo Fisher Scientific) and 10 lL of 2 lmol/L siRNA (final concentration: 100 nmol/L) were mixed and incubated for 5 min at room temperature. TRPM7-targeted siRNA and nontargeted siRNA were used (Silencer â Select s81666 and negative control 1 from Thermo Fisher Scientific). Subsequently, 18.6 lL Opti-MEM and 1.4 lL of a lipidbased transfection reagent, DharmaFECT 4 (GE Healthcare Life Science, Buckinghamshire, UK), were added and incubated for 20 min at room temperature. A suspension of differentiated 3T3-L1 adipocytes was then added to the plate containing siRNA complex at a density of 1.02-1.16 9 10 5 cells/cm 2 and cultured for 24 h. To visualize cells transfected with siRNA, siGLO Green Transfection Indicator (Thermo Fisher Scientific) was co-transfected. Twenty-four hours after transfection, cells were re-plated onto Matrigel-coated glass cover slips and then cultured an additional 18 to 24 h before patch-clamp experiments or Ca 2+ imaging experiments.

Quantitative real-time polymerase chain reaction (PCR)
First-strand cDNA of 3T3-L1 adipocytes was prepared using the SuperScript TM III CellsDirect cDNA Synthesis System (Thermo Fisher Scientific) following manufacturer's instructions. The expression level of TRPM7 was determined by quantitative real-time polymerase chain reaction (PCR) in each sample using the ABI 7500 Real-Time PCR system (Thermo Fisher Scientific). The expression level of the house-keeping gene TATA box binding protein was used as an internal control. Primers and probes used for target genes were TaqMan Gene Expression Assays (Mm00446973_m1 for TATA box-binding protein; Mm00457998_m1 for TRPM7) purchased from Thermo Fisher Scientific.

Patch-clamp experiments
All experiments were conducted at room temperature (23-25°C). A salt bridge containing 3 mol/L KCl in 2% agarose was used to connect a reference Ag-AgCl electrode. The patch electrodes, prepared from borosilicate glass capillaries, had a resistance of 1.5-2.5 MΩ when filled with a pipette solution (see below). Series resistance (<5 MΩ) was compensated to 80% to minimize voltage errors. Currents were recorded using an Axopatch 200B amplifier (Molecular Devices, San Jose, CA) coupled to a DigiData 1321A A/D and D/A converter (Molecular Devices). Current signals were filtered at 1 kHz using a four-pole Bessel filter and were digitized at 5 kHz. pClamp 10.4 software was used for the command pulse protocol, data acquisition, and analysis. The time courses of current activation and recovery were monitored by repetitively applying (every 10 sec) ramp pulses from À100 to +100 mV (1 sec duration) from a holding potential of 0 mV. The control bath solution consisted of (mmol/L): 135 NaCl,

Statistical analysis
Data are shown as the mean AE standard error of the mean (SEM) of observations. Statistical analysis was performed between the two groups using a paired t-test (Figs. 1F-H, J and K and 2E-G) or an unpaired t-test ( Fig. 3A and B). A one-way analysis of variance (ANOVA) followed by post hoc Bonferroni test was used to analyze the data of Figures 1C, 2D, and 4C. Data were considered to be significant at P < 0.05.

TRPM7-like current in acutely isolated mouse white adipocytes
Whole-cell patch-clamp recordings were performed to investigate whether native mouse white adipocytes functionally express TRPM7 current. It has been reported that TRPM7 is (1) inhibited by intracellular Mg 2+ (Nadler et al. 2001;Schmitz et al. 2003;Chokshi et al. 2012b,c);  (Chubanov et al. 2012;Tashiro et al. 2014Tashiro et al. , 2019, and FTY720 (Qin et al. 2013;Takahashi et al. 2018). Whole-cell recordings in native white adipocytes revealed that outwardly rectifying currents were gradually activated after the establishment of whole-cell configurations with the pipette solution that contained 29 lmol/L [Mg 2+ ] in control bath solution (Fig. 1A). As TRPM7 current is inhibited by intracellular free Mg 2+ (normallỹ 0.9 mmol/L in cytosol), these currents may contain TRPM7 current, which is activated due to a reduction in intracellular Mg 2+ concentration to 29 lmol/L by intracellular perfusion with the pipette solution . The current was further activated by perfusion with DVF in native white adipocytes (3.5 AE 0.5 and À0.7 AE 0.2 pA/pF in the control solution vs. 11.1 AE 3.6 and À15.5 AE 5.9 pA/pF in DVF at +80 and À80 mV [n = 4], respectively) ( Fig. 1A  (IC 50 ) of 187 lmol/L (Fig. 1D). At 500 lmol/L, H 2 O 2 maximally inhibited the current (Fig. 1A-D 1D). Application of 2-APB (200 lmol/L), NMM (50 lmol/L), or FTY720 (10 lmol/ L) inhibited the DVF-activated current in white adipocytes ( Fig. 1F-H). Moreover the outward current in the presence of extracellular divalent cations was also inhibited by H 2 O 2 (500 lmol/L) as well as NS8593 (20 lmol/ L), which has been reported to inhibit TRPM7 current at low concentrations with an IC 50 of 1.6 lmol/L (Chubanov et al. 2012), though the inward current was too small to detect the significant inhibitory effects of these reagents ( Fig. 1I-K). These electrophysiology results suggest the existence of TRPM7-like current in white adipocytes.

Discussion
Adipocytes secrete various hormone-like factors, the socalled adipokines. It has been suggested that Ca 2+ modulates intracellular signaling, which mediates the secretion of adipokines (Levy et al. 2000;Cammisotto and Bukowiecki 2004;Sukumar et al. 2012;Komai et al. 2014;Schlottmann et al. 2014;El Hachmane et al. 2015) and the insulin response (Draznin et al. 1987;Kelly et al. 1989;Whitehead et al. 2001;Worrall and Olefsky 2002). It has been reported that there are several functional Ca 2+ influx pathways, including VDCC (Draznin et al. 1987;Pershadsingh et al. 1989;Levy et al. 2000),  found that TRPM7, a Ca 2+ /Mg 2+ -permeable cation channel, is functionally expressed in both native white adipocytes and differentiated 3T3-L1 adipocytes, and plays a role as a Ca 2+ influx pathway. The characteristics of the DVF-activated, TRPM7-like current in native white adipocytes and differentiated 3T3-L1 adipocytes were quite consistent with those reported for heterologously overexpressed TRPM7 current. TRPM7-like currents in both cell types were activated by a reduction of [Mg 2+ ] i , an elimination of extracellular divalent cations, and were inhibited by 2-APB, NS8593, and FTY720 ( Figs. 1 and 2), which were consistent with previous reports (Nadler et al. 2001;Schmitz et al. 2003;Chubanov et al. 2012;Chokshi et al. 2012b,c;Qin et al. 2013;Tashiro et al. 2014Tashiro et al. , 2019. We previously reported that TRPM7 current was inhibited by H 2 O 2 in a [Mg 2+ ] iand [ATP] i -dependent manner in HEK293 cells overexpressing murine TRPM7 ). In the presence of 5 mmol/L [ATP] i , H 2 O 2 failed to inhibit overexpressed TRPM7 current ). In the present study, the endogenous TRPM7-like current was inhibited by H 2 O 2 in a [Mg 2+ ] i -dependent manner; however, the inclusion of 5 mmol/L [ATP] i did not affect the current inhibition by H 2 O 2 (Figs. 1A-D, I and J, 2B-D, 3C). Although it remains unclear why the effect of ATP differs between endogenous TRPM7-like current and overexpressed TRPM7 current, based on overall similarities in their characteristics and the observation that siTRPM7 effectively reduced the current, our data suggest that TRPM7 is functionally expressed in native adipocytes and in differentiated 3T3-L1 adipocytes.
Naltriben augmented Ca 2+ influx as well as the current (Fig. 4), indicating that TRPM7 is one of the Ca 2+ influx pathways in adipocytes. Although the current recording showed the growing current during a 4 min application of naltriben (Fig. 4A), the increase in [Ca 2+ ] i reached a plateau within 2 min (Fig. 4D). This might result from the inhibition of the channel by intracellular Ca 2+ Matsushita et al. 2005) that did not occur during the whole-cell path-clamp recordings due to intracellular perfusion by the pipette solution, and/or the activity of Ca 2+ extrusion by calmodulin-sensitive Ca 2+ pumps (Pershadsingh et al. 1980) and Na + /Ca 2+ exchanger (Pershadsingh et al. 1989) in adipocytes. Since [Ca 2+ ] i is finely tuned by multiple molecules, TRPM7-knockdown did not significantly affect resting [Ca 2+ ] i in 3T3-L1 adipocytes (Fig. 4E). In contrast, the naltriben-stimulated [Ca 2+ ] i increase was clearly observed in siControl-adipocytes but not in siTRPM7-adipocytes. Thus, Ca 2+ influx via TRPM7 might exert physiological effects, especially when the TRPM7 channel is activated.
It is well established that TRPM7 channel activity is tonically regulated by intracellular free Mg 2+ as well as Mg-nucleotides (Penner and Fleig 2007); however, little is known about the physiological stimuli that change the TRPM7 channel activity. Several reports indicate that activation of phospholipase C (PLC) induces Ca 2+ influx via TRPM7 in intact cells (Langeslag et al. 2007;Qian et al. 2019). There are contradicting reports showing that TRPM7 is inactivated by G aq -coupled receptor stimulation via the PLC-mediating depletion of phosphatidylinositol 4,5-bisphosphates (PIP 2 ) which is necessary for TRPM7 channel activity (Runnels et al. 2002;Kozak et al. 2005;Gwanyanya et al. 2006;Bates-Withers et al. 2011;Zhelay et al. 2018). Therefore, the physiological stimuli that activate PLC in adipocytes might increase or decrease [Ca 2+ ] i by activation or inactivation of TRPM7. Insulin has been reported to activate PLC in adipocytes; however it remains controversial whether insulin induces Ca 2+ flux (Draznin et al. 1987;Cammisotto and Bukowiecki 2004 In pre-adipocytes, it has been reported that TRPM7 is functionally expressed and involved in proliferation and adipogenesis (Chen et al. 2014). Consistently, we examined TRPM7 mRNA expression in 3T3-L1 pre-adipocytes and found that the expression levels of TRPM7 were comparable between pre-adipocytes and differentiated 3T3-L1 adipocytes (data not shown). Thus, unlike Cav3.1, which is downregulated in mature adipocytes (Oguri et al. 2010), or TRPC1/C5, which are upregulated by differentiation (Sukumar et al. 2012), TRPM7 was expressed constantly irrespective of differentiation in adipose cells. The physiological roles of TRPM7 in mature adipocytes remain unknown. Since TRPM7 is a Ca 2+ /Mg 2+ -permeable channel that contains a protein kinase domain, it likely exerts functions via Ca 2+ /Mg 2+ influx and/or its kinase activity. In adipocytes, Ca 2+ modulates lipolysis (Tebar et al. 1996;Xue et al. 1998), glucose uptake (Draznin et al. 1987;Kelly et al. 1989;Whitehead et al. 2001;Worrall and Olefsky 2002), and the secretion of adipokines (Levy et al. 2000;Cammisotto and Bukowiecki 2004;Sukumar et al. 2012;Komai et al. 2014;Schlottmann et al. 2014;El Hachmane et al. 2015). It has been reported that the reduction in intracellular Mg 2+ content is related to impaired glucose metabolism and insulin resistance (Kandeel et al. 1996;Takaya et al. 2004;Barbagallo and Dominguez 2007). One of the reported substrates of TRPM7 kinase, myosin IIA, is involved in glucose transporter 4 translocation upon insulin stimulation (Steimle et al. 2005;Fulcher et al. 2008;Chung et al. 2010;Stall et al. 2014). Phosphorylation of myosin IIA by TRPM7 inhibits myosin II-based contractility (Clark et al.  2006, 2008). Thus, it is likely that downregulation of TRPM7 affects these functions. In our pilot studies investigating the effects of TRPM7-knockdown on lipolysis and insulin-stimulated glucose uptake, there was no significant difference between siControl and siTRPM7 (data not shown). Furthermore, released adipokines in culture supernatants were roughly compared using a Profiler Mouse Adipokine Array (R&D Systems, Minneapolis, MN), and the results revealed that there was no marked difference between siControl and siTRPM7 in our current preparations (~65% knockdown of TRPM7 in 3T3-L1 adipocytes differentiated for 9 days in culture). Thus, further study is needed to understand the physiological roles of TRPM7 in adipocytes under diverse conditions (e.g., using TRPM7-deficient or overexpressing matured adipocytes, and adipocyte-specific TRPM7-deficient mice).
In conclusion, we determined the functional expression of TRPM7 by whole-cell path-clamp and Ca 2+ imaging experiments in adipocytes. TRPM7 may modulate physiological functions of adipocytes by permeating Ca 2+ /Mg 2+ and/or phosphorylating the substrates.