The Zinc Transporter, Slc39a7 (Zip7) Is Implicated in Glycaemic Control in Skeletal Muscle Cells

Dysfunctional zinc signaling is implicated in disease processes including cardiovascular disease, Alzheimer's disease and diabetes. Of the twenty-four mammalian zinc transporters, ZIP7 has been identified as an important mediator of the ‘zinc wave’ and in cellular signaling. Utilizing siRNA targeting Zip7 mRNA we have identified that Zip7 regulates glucose metabolism in skeletal muscle cells. An siRNA targeting Zip7 mRNA down regulated Zip7 mRNA 4.6-fold (p = 0.0006) when compared to a scramble control. This was concomitant with a reduction in the expression of genes involved in glucose metabolism including Agl, Dlst, Galm, Gbe1, Idh3g, Pck2, Pgam2, Pgm2, Phkb, Pygm, Tpi1, Gusb and Glut4. Glut4 protein expression was also reduced and insulin-stimulated glycogen synthesis was decreased. This was associated with a reduction in the mRNA expression of Insr, Irs1 and Irs2, and the phosphorylation of Akt. These studies provide a novel role for Zip7 in glucose metabolism in skeletal muscle and highlight the importance of this transporter in contributing to glycaemic control in this tissue.


Introduction
Cellular zinc storage, release and distribution are controlled by a family of zinc transporters and metallothioneins. In mammals two families of zinc transporters exist: the zinc efflux (Slc30/ZnT) and the zinc influx (Slc39/ZIP) proteins [1]. ZnT proteins transport zinc out of the cell or into subcellular compartments in the presence of high cytoplasmic zinc. In contrast, ZIP proteins transport zinc into the cell or out of subcellular compartments when cytosolic zinc is low or depleted [2].
There is increasing interest in the importance of zinc transporters in diseases associated with dysfunctional cellular signaling. In particular, a significant role for these transporters in maintaining essential glucose and lipid metabolism has been identified. For example, in myocytes isolated from the femoral muscle of ZnT7 knockout mice, a reduction in insulin signaling pathway activity was observed [3]. The ZnT7 null mice were susceptible to diet-induced glucose intolerance and insulin resistance and this was associated with a decrease in the expression of the insulin receptor, insulin receptor substrate 2 and Akt1 [3]. ZnT3, ZnT5 and ZnT8 gene expression are differentially regulated by glucose in INS-IE cells, and streptozotocin-treated ZnT3 null mice have decreased insulin gene expression and insulin secretion that resulted in hyperglycemia [4]. Moreover, ZnT8 plays a critical role in the synthesis and secretion of insulin and therefore represents a pharmacological target for treating disorders of insulin secretion including diabetes [5].
Zinc mediates its effects through two mechanisms; early zinc signaling (EZS) and late zinc signaling (LZS) [6]. LZS occurs several hours after an extracellular signaling event and depends on changes in the expression of zinc-related molecules such as zinc transporters and metallothioneins [6,7]. In contrast, EZS occurs minutes after an extracellular stimulus and does not involve transcriptional-dependent changes [6,7]. Zinc signaling mechanisms are involved in eliciting an increase in intracellular zinc concentrations 2 the 'zinc wave' phenomenon [8]. Thus, in this situation zinc acts as a second messenger that activates pathways associated with cellular signaling. In fact, zinc has been categorized as an insulin-mimetic with several groups examining the role of its mimetic activity on glucose [9][10][11][12][13] and lipid [13,14] metabolism. In this context ZIP7 has been identified as a key zinc transporter implicated in the ''zinc wave'' and is suggested to be a ''gatekeeper'' of cytosolic zinc release from the ER [8]. Endogenous ZIP7 is predominately localized to the Golgi apparatus [15], the ER [16], or both [17] and has been implicated in breast cancer progression [8,17,18]. Studies in tamoxifen-resistant MCF-7 breast cancer cells identified that ZIP7 was responsible for activation of multiple tyrosine kinases that are implicated in the aggressive phenotype of tamoxifen-resistant breast cancer [8,19,20]. Recent evidence in MCF7 cells suggests that ZIP7 is phosphorylated by CK2 and is associated with the regulated release of zinc from intracellular stores to phosphorylate kinases implicated in cell proliferation and migration [8].
Given the role of ZIP7 in modulating zinc flux, and the role of zinc as an insulin mimetic in cellular processes, we propose that ZIP7 may also be implicated in metabolic processes associated with glycaemic control. Here we report evidence for a novel role for Zip7 in modulating glycaemic control in skeletal muscle cells. We find that the attenuation of Zip7mRNA in C2C12 skeletal muscle cells modulates genes involved in carbohydrate metabolism and glycogen synthesis. These studies demonstrate a previously unprecedented role for Zip7 in regulating glycaemic control in skeletal muscle and provide a platform to further explore the potential of this transporter in skeletal muscle insulin resistance.

Cell culture
Proliferating mouse C2C12 myoblasts in all experiments were cultured and maintained in DMEM supplemented with 10% Fetal Bovine Serum and physiological zinc concentrations (20 mM ZnSO 4 ), (Life Technologies, Mulgrave, Victoria, Australia). Differentiation of myoblasts into post-mitotic, multi-nucleated myotubes was induced by mitogen withdrawal (i.e. DMEM supplemented with 20 mM Zn SO 4 and 2% horse serum for three days). Assessment of the muscle-specific, contractile and metabolic C2C12 muscle phenotype was assessed by measuring the expression of markers of differentiation and metabolic processes as previously described [21]. The time course experiments on differentiated C2C12 skeletal muscle cells were performed over 60 minutes in the presence of 10 nM insulin, 20 mM ZnSO 4 and 10 mM pyrithione (see Figures S1 and S2).

RNA Extraction and cDNA Synthesis
Mouse quadriceps muscle was a kind gift from Dr. Paul Lewandowski, Deakin University, Australia with approval from the Deakin University Animal Welfare Committee (A37/2007). Total RNA was extracted from C12C2 cells and C57Bl/6J mouse quadriceps using TRI-Reagent (Sigma-Aldrich, Castle Hill, NSW, Australia) according to the manufacturer's protocol. Total RNA was then treated with 2 U of DNase1 for 30 min at 37uC followed by purification of the RNA through an RNeasy purification column system (Qiagen, Chadstone, Victoria, Australia). RNA quantity and quality was measured using a Nanodrop spectrophotometer (Thermo Scientific, Scoresby, Victoria, Australia). A High Capacity cDNA Synthesis kit was used to synthesize cDNA from 2 mg of total RNA using random hexamers according to the manufacturer's instructions (Life Technologies). The cDNA was diluted to 400 ml in nuclease-free water and stored at 220uC.

Mouse glucose metabolism and zinc transporter arrays
The Mouse Glucose Metabolism RT 2 Profiler PCR Array was purchased from SA Biosciences, Qiagen; Catalogue No. 330321 PAMM-006ZA. This array profiles the expression of 84 genes involved in the regulation and enzymatic pathways of glucose and glycogen metabolism (Table S1). The Zinc Transporter RT 2 Profiler Custom PCR Array (Qiagen) contained the genes for the two zinc transporter families, Slc30a1-10 and Slc39a1-14.
cDNA synthesis of RT 2 mouse glucose metabolism and zinc transporter PCR array cDNA synthesis using the RT 2 First Strand Kit was performed as described by the manufacturer (Qiagen). Briefly, potential genomic DNA was eliminated from 500 ng of total RNA using buffer GE at 42uC for 5 min. The RNA was then reverse transcribed in a total of 20 ml reaction volume for 15 mins at 42uC then the reaction was stopped by heating the sample at 95uC for 5 min.

Quantitative Real-time PCR
Quantitative PCR (qPCR) was performed on a RealPlex PCR detection system (Eppendorf, North Ryde, New South Wales, Australia) in triplicate on at least three independent RNA preparations. Target cDNA levels were analyzed in 10 ml reactions with SensiMix SYBR No-ROX (Bioline, Alexandria, New South Wales, Australia). Primers (GeneWorks, South Australia, Australia) for markers of skeletal muscle cell differentiation and metabolism, Myogenin, Tnni1, Tnni2, Abca1, Fabp3 and Srebp-1c (Table S2) have been previously described [21][22][23]. Other primers (Table S2) for the amplification of target gene sequences were designed using the NCBI Primer Blast Tool http://www.ncbi.nlm.nih.gov/tools/ primer-blast/index.cgi, with the exception of Irs1 (PrimerBank ID: 29825829a1), Irs2 (PrimerBank ID: 3661525a1) and Insr (Primer-Bank ID: 6754360a1) which were obtained from the PrimerBank Database http://pga.mgh.harvard.edu/primerbank/index.html [24][25][26]. Note: all primers were rigorously analyzed by BLAST for target gene specificity and designed to be genomic resistant (i.e. at least one primer crossed an exon-exon boundary). qPCR was performed using 4 ml of cDNA template (20 ng) and 40 cycles of 95uC for 15 seconds, 57uC for 15 seconds and 72uC for 20 seconds. The relative level of target gene expression was normalized to Gapdh, or eukaryotic elongation factor 2 (Eef2) as described in the results section and associated errors were calculated using the guidelines described by Bookout and Mangelsdorf [27].

Primer design to detect endogenous and exogenous pCMV-Zip7
Primers were designed to specifically target endogenous and exogenous Zip7 mRNA. For the pCMV-Zip7 overexpression plasmid, we placed the forward primer on the Zip7 mRNA sequence and the reverse primer on the plasmid C-myc tag (see Table S2). For the specific amplification of endogenous Zip7 we designed primers on the 59UTR. This region is omitted on the pCMV-Zip7 expression plasmid.
Quantitative Real-time PCR of RT 2 mouse glucose metabolism and zinc transporter PCR array The qPCR for the RT 2 glucose metabolism and zinc transporter arrays were performed as outlined in the guidelines supplied by the manufacturers (Qiagen). Briefly, cDNA (102 ml) from the RT 2 First Strand Kit was added to 1350 ml of 2 x RT 2 SYBR Green and 1248 ml of H 2 O. To each well of the glucose metabolism array 96-well plate, 25 ml of sample was added. qPCR was performed on a RealPlex PCR detection system (Eppendorf) using 45 cycles of amplification consisting 95uC for 15 seconds and 60uC for 1 minute.

Transient transfections of siRNA molecules
The transient transfection of siRNA molecules Zip7 (Catalogue No: AM16708), Zip1 (Catalogue No: AM16708), Gapdh (Catalogue No. 4390771) and the scramble control (Catalogue No: AM4635) (Life Technologies) were performed using RNAiMAX reagent as instructed by the manufacturer (Life Technologies). Briefly, C2C12 cells were transfected in 6-well dishes with 10 nM of siRNA molecule or the scramble control in RNAiMAX reagent. The cells were subsequently maintained in 2% horse serum and differentiated over three days and collected in 1 ml of TRI-Reagent per three wells for RNA extraction or 1 ml of RIPA Buffer (Thermo Scientific) (containing Halt Protease and Phosphatase Inhibitor Cocktail; Thermo Scientific) for protein analysis.

Zip7 overexpression plasmid and transient transfection in C2C12 skeletal muscle cells
A full-length mouse cDNA Zip7 expression plasmid was obtained from Origene Technologies, Inc (Clone ID MR216531; Rockville, MD). The pCMV control plasmid was created by excision of the full-length Zip7 gene by restriction digest at enzyme sites Sgf1 and Mlu1 followed by end-filling and blunt-end ligation. Briefly, 1 mg of pCMV-Zip7 plasmid was digested in the presence of 10X Fast Digest Green Buffer and 1 unit of Sgf1 and Mlu1 restriction enzymes (Fermentas, Thermo Scientific) for 5 min at 37uC. For the end-filling, approximately 1 mg of pCMV plasmid DNA was incubated with 0.5 mM dNTPs and 1 unit of Klenow fragment and incubated at 30uC for 15 minutes. The pCMV control plasmid was circularized in the presence of 2 ml of 10X T4 DNA ligase, 100 ng of plasmid vector and 1 ml of T4 DNA ligase and incubated at 4uC for approximately 16 h.
The pCMV-Zip7 and pCMV control plasmid were transiently transfected into C2C12 skeletal muscle cells using Lipofectamine 2000 (Life Technologies) as instructed by the supplier. Briefly, C2C12 skeletal muscle cells were grown to 80% confluence and 4 mg of pCMV-Zip7 and the pCMV control vector were mixed with 5 ml of Lipofectamine 2000 and 500 ml of optimen. Following a 20 min incubation at RT, the Lipofectamine-DNA reagent was pipetted onto the C2C12 cells contained in a 6-well plate and supplemented with 1.5 ml of differentiation media. The cells were differentiated for 72 hr and subsequent RNA and protein was extracted as described below.

Protein Extraction and Western Blot
Total cellular protein from the scramble control and the siRNA-Zip7 transfected C2C12 cells was isolated by scraping cells with RIPA buffer (contains protease and phosphatase inhibitor cocktail) then the samples were place on ice for 1 hr with constant vortexing every 10 mins. The cells were then sonicated for 5 second pulses for 30 seconds at 50% duty followed by boiling for 10 min. The protein samples were centrifuged at 13,000 rpm for 5 min and the supernatant collected. Total protein concentration was measured using a BCA kit (BIORAD, Gladesville, New South Wales) as outlined by the manufacturer's instructions.
Total soluble protein (100 mg) from the scramble control and siRNA-Zip7 transfected C2C12 cell lines was resolved on a 4-15% SDS-PAGE gradient gel (BIORAD) and transferred to a nitrocellulose membrane. The membranes were blocked overnight in 5% skim milk in TBS-Tween 20 followed by an overnight incubation with either Glut4 ( for 1 hr at RT. Immunoreactive signals were detected using enhanced chemiluminescence SuperSignal West Pico Substrate (Pierce) and visualized by autoradiography or a UVITEC Alliance digital imaging system (Thermo Fisher Scientific, Victoria Australia). Note: to assess protein loading consistency, the membranes were stripped with Restore Plus Western Blot Stripping Buffer (Thermo Fisher) by incubating the membrane in the buffer for 15 min at RT. Membranes were subsequently washes in TBS-Tween and blocked with 5% skim milk before adding the primary antibody.

Glycogen synthesis assay
The glycogen synthesis assay was performed as described by the manufacturer's (BioVision, Life Research, Scoresby Victoria, Australia). Briefly, C2C12 cells were transfected with siRNA-Zip7 and the scramble control as described above. Following differentiation, C2C12 cells were treated with 10 nM insulin for 60 minutes. Cell lysates were collected in 200 ml of dH 2 O on ice and homogenates were boiled for 5 min to inactivate enzymes. Samples were then centrifuged at 13000 rpm for 5 min and the supernatant was collected. Samples were then prepared by performing hydrolysis of glycogen to glucose and then mixed with OxiRed probe to generate colour (l max = 570 nm). Note: a glucose control was also performed in the absence of glucoamylase to determine background glucose levels. These were subsequently subtracted from the glycogen readings. The glycogen concentration in the samples was calculated by C = Ay/Sv where Ay is the amount of glycogen (mg) in the sample as determined from a standard curve and Sv is the sample volume (ml).

Statistics
Data obtained from individual qPCR was assessed by a Student's unpaired t-test on at least three independent biological replicates. Statistical significance was denoted as the average 6 standard deviation of the mean. Data was considered statistically significance when the P value was #0.05. *P,0.05; **P,0.01 and ***P,0.001. The analysis of the gene arrays was performed with the RT 2 Profiler PCR Array Data Analysis Software v3.5 (SA Biosciences).

The Slc39a (Zip) zinc transporters are differentially expressed in C2C12 skeletal muscle cells and mouse quadriceps
To determine the expression levels of the Slc39a zinc transporter family in mouse C2C12 skeletal muscle cells and mouse quadriceps we utilized a custom zinc gene array (SABiosciences, Qiagen) with primer sequences that are specific for the Slc39a (Zip) mouse zinc transporter genes (i.e. Slc39a1-14). Quantitative realtime PCR (qPCR) was performed and the expression of each zinc transporter transcript was measured relative to the 'housekeeping gene', Gapdh.
The zinc transporters Slc39a1 and Slc39a7 were highly expressed in C2C12 skeletal muscle cells ( Figure 1A). Lower levels of expression were observed for Slc39a3, 6, 9, 10, 11, 13 and 14. Minimal or no expression was observed in Slc39a2, 4, 5, 8, and 12 ( Figure 1A). In mouse quadriceps we observed high levels of expression for all of the Slc39a transporters with the exception of Slc39a5 ( Figure 1B).

Slc39a7 (Zip7) is expressed during C2C12 skeletal muscle differentiation
We were most interested in Slc39a7 (Zip7) as this transporter is predominately localized to the ER and the Golgi apparatus and is suggested to be involved in the 'zinc wave' and associated cellular signaling [8]. Accordingly, to elucidate the role of Zip7 in skeletal muscle we initially investigated the expression profile of this zinc transporter relative to Eef2 in the mouse C2C12 myoblast cell line. Proliferating myoblasts can be induced to biochemically and morphologically differentiate into post-mitotic multinucleated myotubes by mitogen withdrawal. This transition from a nonmuscle phenotype to a contractile phenotype is associated with the activation and repression of a structurally diverse group of genes responsible for contraction and the extreme metabolic demands placed on this tissue [21]. During this period of differentiation, we observed that Zip7 mRNA is highly expressed in proliferating myoblasts and was constitutively expressed during skeletal muscle cell differentiation when normalized to Eef2 (Figure 2A).
In order to assess the differentiation status of the C2C12 cells and to demonstrate that they had acquired a differentiated, contractile and metabolic phenotype, qPCR was performed on the marker genes myogenin (MyoG), a gene that encodes the hierarchical basic helix loop regulator and is specifically required for differentiation [28], the slow twitch (type I) and the fast twitch (type II) isoforms of the contractile protein troponin I (Tnni1 and TnniII), and the metabolic genes Abca1 (ATP-binding cassette proteins), Fabp3 (fatty acid binding protein 3) and Srebp1c (sterol regulatory binding element protein). Expression of both MyoG and the contractile protein genes (type I and II, Tnni1 and Tnni2, respectively) were dramatically increased and confirmed the differentiation of the myoblast C2C12 skeletal cell line to the myotube phenotype ( Figure 2B-D). Additionally, genes involved in lipid metabolism (Abca1 and Srebp1c) ( Figure 2E and 2G) were also induced while Fabp3 was downregulated during muscle differentiation ( Figure 2F) which is consistent with previous studies [22,[29][30][31] and confirms that the muscle cells had acquired the appropriate contractile and metabolic phenotype.

siRNA-Zip7 Expression Represses Endogenous Zip7 mRNA in Skeletal Muscle Cells
To elucidate the biological role of Zip7 in the context of glucose metabolism we selectively ablated the expression of this transporter in C2C12 skeletal muscle cells utilizing a siRNA-Zip7 molecule. An siRNA targeting mouse Gapdh and a scramble sequence that contains no known homology to the mouse, rat or human genome were utilized as controls. The siRNA-Gapdh was used to determine the robustness of the transfection and the ability to successfully attenuate a specific target gene that is constitutively expressed. Accordingly, C2C12 cells were transfected with the scramble control, siRNA-Gapdh or the siRNA-Zip7 and subsequently differentiated for three days.
Initially, we aimed to validate the specificity and robustness of the siRNA transfection in C2C12 cells by transfecting an siRNA-Gapdh to determine transfection efficacy and siRNA specificity. We identified a significant reduction in Gapdh mRNA (4-fold, p = 0.0023) in the siRNA-Gapdh transfected cells compared to the scramble control ( Figure 3A). We then transfected C2C12 skeletal muscle cells with an siRNA targeting Zip7 mRNA. Quantitative PCR was then performed to measure the expression levels of endogenous Zip7 relative to Eef2 in RNA isolated from the scramble control and Zip7 transfected cell lines. We observed a significant reduction in the mRNA levels of Zip7 (4.6-fold, p = 0.0006) when compared to the scramble control ( Figure 3B).
To determine that the attenuation of Zip7 was not due to differential Eef2 mRNA expression, qPCR was also performed on Eef2 normalized to Gapdh. No change in the level of Eef2 in the Zip7-siRNA cell lines were observed when normalized to Gapdh mRNA ( Figure 3C). We also tested the relative expression of Zip7 in siRNA-Gapdh C2C12 cells. There was no change in Zip7 mRNA expression in the Gapdh reduced C2C12 cell lines ( Figure 4D).
Since Zip1 was also highly expressed in C2C12 skeletal muscle cells ( Figure 1) we decided to selectively reduce the expression of this transporter with an siRNA-Zip1 to determine if there were any compensatory changes in Zip7 expression. C2C12 cells were transfected with the scramble control and siRNA-Zip1 and endogenous Zip1 and Zip7 mRNA was measured. We successfully attenuated endogenous levels of Zip1 mRNA (approximately 3fold, p = 0.0025) in the C2C12 cell lines ( Figure 3E). No change in endogenous expression of Zip7 mRNA (p = 0.1040) was observed in the siRNA-Zip1 cell lines ( Figure 3F).

The attenuation of Zip7 resulted in no change in other zinc transporters
To determine the expression status of the other zinc transporter family members in the presence of the Zip7 reduced C2C12 cell lines we utilized a custom gene array that contains the primer sequences for the Slc30a/ZnT (1-10) and Slc39a/Zip (1-14) family members. cDNA from the scramble control and the siRNA-Zip7 C2C12 cells were assayed to assess for compensatory changes in the other family members due to reduced Zip7 mRNA. We identified that the reduction of Zip7 had no effect on the expression of the Slc30a/ZnT family members ( Figure 4A). In the Slc39a/Zip arrays, reduced expression of Zip7 resulted in a significant attenuation of Zip7 mRNA as expected. We also  . Relative expression of Slc39a7 (Zip7) and markers of skeletal muscle differentiation in C2C12 cell lines. A). Slc39a7 (Zip7) expression relative to Eef2, B-D). Markers of skeletal muscle differentiation: myogenin (MyoG) and the troponins 1 and 2 (Tnni1 and Tnni2), respectively. E-G). Markers of metabolism: ATP-binding cassette transporter protein 1 (Abca1), fatty-acid binding protein 3 (Fabp3) and sterol regulatory element binding protein 1c (Srebp-1c), respectively. PMB = proliferating myoblasts; D1-3 = day 1 to day 3 of differentiation of myotubes, respectively. Error bars indicated the 6 SD from three independent biological samples. doi:10.1371/journal.pone.0079316.g002 Figure 3. Zip7 mRNA is attenuated by si-RNA-Zip7. Relative expression of Gapdh, Zip7, Zip1 and Eef2 in the scramble control and corresponding siRNA cells, respectively. A). Gapdh relative to Eef2 in siRNA-Gapdh cells B). Zip7 relative to Eef2 in siRNA-Zip7 cells C). Eef2 relative to Gapdh in siRNA-Zip7 cells D). Zip7 relative to Eef2 in siRNA-Gapdh cells E). Zip1 relative to Eef2 in siRNA-Zip1 cells, and F). Zip7 relative to Eef2 in siRNA-Zip1 cells. Error bars indicated the 6 SD from three independent biological samples. **P,0.01, ***P,0.001. doi:10.1371/journal.pone.0079316.g003 observed a small, but significant reduction in the expression of Zip13 and Zip14 mRNA ( Figure 4B).
To further assess the reduced expression of Zip13 and Zip14 in the Zip7 reduced C2C12 cells we designed primer pairs specific for Zip13 and Zip14 to independently test the validity of this observation. We performed qPCR on Zip13 and Zip14 expression in the scramble control and siRNA-Zip7 C2C12 cells. No significant changes in the level of expression for these zinc transporters were observed ( Figure 4C).
We further validated several of these genes with a focus on glycogen metabolism (Pgm2, Phkb, Pygm and Gbe1) by designing new primer pairs and performing qPCR on the scramble control versus the siRNA-Zip7 cDNA. We observed significant downregulation in these genes in concordance with the PCR array data ( Figure 5A-D).
We speculated that given genes implicated in glycogen metabolism were affected by reduced Zip7 mRNA levels, that perhaps the glucose transporter, Glut4 might be downregulated in the siRNA-Zip7 cells. Glut4 predominately transports glucose across the plasma membrane which is further processed by oxidative (glycolysis) or non-oxidative (glycogenesis) pathways [32]. Accordingly, qPCR was performed for Glut4 mRNA expression in the scramble control and the siRNA-Zip7 C2C12 cells. We observed a significant downregulation of Glut4 in the siRNA-Zip7 cells (p = 0.0096) ( Figure 6A). We also tested for Glut4 immunoreactive protein in the scramble control and siRNA-Zip7 C2C12 cells. Accordingly we observed a significant reduction in immunoreactive Glut4 in the siRNA-Zip7 C2C12 cells compared to the scramble control ( Figure 6B). Gapdh was used as a protein loading control and showed that similar amounts of total soluble protein were resolved ( Figure 6B).

Reduced Zip7 compromises insulin-induced glycogen synthesis and phosphorylation of AKT in C2C12 skeletal muscle cells
Cellular glucose utilization by Glut4 is responsible for glycogenesis in muscle [33] and with increasing plasma insulin concentration, glycogen synthase is activated by insulin and glycogen synthesis predominates [34]. Moreover, a core component of glycogen synthesis is the insulin-induced phosphorylation of AKT in a process that leads to the activation of glycogen synthase [34]. To test the efficacy of insulin to induce phosphorylation of Akt and thus confirm the robustness of the C2C12 skeletal muscle cell line to respond to insulin, skeletal muscle cells were treated with 10 nM insulin over 60 mins mins and subsequent protein was extracted as described in Material and Methods. We observed that 10 nM of insulin activated pAkt after 5 min followed by a robust phosphorylation of Akt over the 60 min time course (Figure S1), and thus confirmed the validity of skeletal muscle cell line to respond to insulin treatment ( Figure S1). Similarly, to test whether maintaining the C2C12 line in the presence of 20 mM ZnSO 4 (see Materials and Methods, Cell culture) affected the phosphorylation status of AKT we treated cells with ZnSO 4 alone and ZnSO 4 in the presence of an ionphore, pyrithione ( Figure S2). Accordingly, 10 mM of pyrithione in the presence of 20 mM ZnSO 4 induced a rapid phosphorylation of AKT within 15 minutes and this increased further over 30 and 60 minutes of treatment ( Figure S2). We did not observe an increase in AKT phosphorylation in the presence of ZnSO 4 alone and thus confirmed that maintaining our cell culture system in the presence of 20 mM ZnSO 4 had no effect on AKT phosphorylation.
Given that Zip7 modulates core genes implicated in glucose metabolism, we tested whether glycogen synthesis was compromised in the attenuated Zip7 skeletal muscle cells. We treated the scramble and siRNA-Zip7 C2C12 cells with 10nM insulin over 60 minutes and performed glycogen synthesis. We observed a significant reduction in glycogen synthesis in the siRNA-Zip7 when compared to the scramble control ( Figure 6C). As expected, we observed a significant induction of glycogen synthesis on exposure to insulin in the scramble control cells, however this effect was blunted in the Zip7-siRNA C2C12 ( Figure 6C).
To determine a potential mechanism of action for the reduced glycogen synthesis in the presence of reduced Zip7 mRNA we performed qPCR on the insulin receptor (Insr) and the most predominant isoforms of the insulin receptor substrate molecules that are expressed in skeletal muscle, insulin receptor substrate 1 (Irs1), and insulin receptor substrate 2 (Irs2) [32]. These substrates serve as docking molecules for several SH2-containing proteins and the subsequent activation of downstream signaling molecules that result in the activation of AKT, which mediates many of insulin's metabolic effects by modulating gluconeogenesis, protein synthesis and glycogen synthesis [33]. Accordingly, the reduced expression of Zip7 in the C2C12 skeletal muscle cells resulted in a significant reduction in the expression of the Insr, Irs1 and Irs2 ( Figure 7A-C). In order to confirm that the reduction of these key genes was associated with a reduction in signaling we performed immunoblot analysis on phosphorylated Akt (pAkt). We observed a significant reduction in pAkt in the Zip7-siRNA compared to the scramble control ( Figure 7D). Overexpression of Zip7 in C2C12 cells induces genes associated with glucose metabolism We observed that reduced Zip7 mRNA in C2C12 skeletal muscle cells was associated with changes in genes implicated in glucose metabolism. For example, a significant reduction in the expression of Pgm2, Phkb, Pygm, Gbe1, Glut4, Insr, Irs1 and Irs2 was observed in the Zip7-siRNA C2C12 cells compared to the scramble control (see Figures 5, 6, 7). Accordingly, to determine if by overexpressing Zip7 we could observe the converse effect on gene expression, we transiently transfected an overexpression Zip7 plasmid (pCMV-Zip7) into C2C12 skeletal muscle cells and after 72 hours collected RNA for subsequent qPCR analysis. We observed a significant induction in the expression of exogenous Zip7 in the pCMV-Zip7 expressing C2C12 cells compared to the pCMV control ( Figure 8A and B). To confirm that the major Zip7 mRNA transcript observed was from the overexpression of the pCMV-Zip7 plasmid we performed PCR using primers that were specific for the endogenous form of Zip7 mRNA. We observed that Zip7 mRNA was expressed at relatively much lower levels in both the pCMV and pCMV-Zip7 transfected cells ( Figure 8C) confirming that the major Zip7 transcript resulted from the overexpression system. Moreover, we found that the overexpression of Zip7 mRNA induced the expression of the insulin receptor (Insr); insulin receptor substrate 1 (Isr1) and insulin receptor substrate 2 (Isr2) (see Figure 8D-F). This was in contrast to Figure 7 where a reduction in the expression of Zip7 mRNA resulted in reduced expression of Insr, Isr1 and Isr2. We also observed an increase in Glut4 mRNA in the pCMV-Zip7 overexpression system, however this result did not attain significance (p = 0.0590). Similarly, Glut4 protein levels were not significantly changed in the pCMV-Zip7 overexpression system when compared to the pCMV control (data not shown).

Discussion
Intracellular zinc homeostasis is largely regulated by two families of zinc transporters (ZnTs and ZIPs) that traffic zinc across biological membranes [35,36]. Dysregulation of zinc signaling leads to a number of disease states including cancer [19,37], autoimmune disease [38,39], cardiovascular disease [40,41] and diabetes [42][43][44][45]. Of this family, ZIP7 is important in maintaining physiological and cellular zinc homeostasis through its ability to initiate the 'zinc wave' and provide cytosolic zinc ions that are involved in cellular signaling processes. Although many zinc transporters respond to fluctuating zinc levels and alter their subcellular localization, ZIP7 is an exception and is restricted constitutively to the membrane of the Golgi apparatus and/or the endoplasmic reticulum [15][16][17]. Furthermore, Zip7 gene expression and intracellular location are not altered in response to changes in intracellular zinc status [3]. Studies in breast cancer cells have elucidated a role for this transporter in cell signaling events [8,20]; however, the role of ZIP7 with respect to the control of the genetic programs associated with carbohydrate metabolism in skeletal muscle has not been addressed. Here we provide the first evidence for a metabolic role for Zip7 in modulating glycaemic control in skeletal muscle and provide support for further studies in processes associated with insulin resistance in this tissue.
Zip7 mRNA is highly expressed in differentiated C2C12 cells and mouse quadriceps. Although Slc39a1 was also highly expressed in C2C12 skeletal muscle cells, homozygous knockout of Slc39a1 in mice produces no phenotype when dietary zinc intake is normal [46] suggesting compensatory actions from other family members. To explore compensatory mechanisms from other zinc transporters we performed qPCR on all of the family members in the scramble control and the siRNA-Zip7 C2C12 skeletal muscle cells. We did not observe major changes in expression of the other members of the zinc transporters which suggest that the attenuation of Zip7 has no other effect on these genes. Of the zinc transporters, it should be emphasized that, in addition to ZIP13, [47] ZIP7 is the only other zinc transporter localized to the Golgi apparatus and not the plasma membrane [15] and compensation is therefore unlikely. Given that ZIP7 is localized exclusively on the Golgi apparatus or the ER [17], and the fact that no compensatory changes in the other transporters were observed in the reduced Zip7 C2C12 cells suggests that this transporter may be unique in its specialized function in transporting zinc from the ER or Golgi into the cytosol [19].
In contrast to the Zip expression profile in C2C12 cells, we also observed moderate levels of expression for all of the Zip transporters (except for Zip5) in mouse quadriceps. Zip7 mRNA was more highly expressed in C2C12 cells (approximately 15-fold) when compared to the expression found in quadriceps. It should be noted that quadriceps contain a mix of muscle fibre-types (oxidative type 1 and glycolytic type II) [48]. Similar studies on  Zip7 Regulates Glucose Metabolism in Muscle other muscle fibre types; soleus (type I), plantaris (type II) and anterior tibialis (type II) also demonstrated differences in the level of expression for the orphan nuclear receptor, Coup-tfII in these tissues in comparison to C2C12 cells [49]. Moreover, studies on protein arginine methyltransferase 3, 4 and 5 (PRMT3-5) in mouse skeletal muscle tissue and C2C12 cells found high expression of PRMT3-5 in gastrocnemius in comparison to only high expression of PRMT4 (with no or minimal expression of PRMT3 and 5, respectively) [50]. Although these relative expression discrepancies exist between in vitro and in vivo model systems, the C2C12 cell culture model is a well-established and validated system to study the effects of metabolic processes [21,51,52]. For example, data derived from this in vitro model with liver X receptor (LXR) and peroxisome proliferating activated receptor (PPAR) agonists and their role in metabolism (e.g. energy expenditure, running endurance, lipid metabolism and cholesterol efflux) has been validated and reproduced in mice [51][52][53][54][55].
Our study revealed that subsets of genes involved in glucose metabolism (Agl, Dlst, Galm, Gbe1, Idh3g, Pck2, Pgam2, Pgm2, Phkb, Pygm, Tpi1, Gusb and Glut4) are altered when Zip7 expression was reduced. This is further highlighted by the fact that related genes in similar and other pathways (see Table S1) were refractory to the attenuation of Zip7 expression. These data are noteworthy for several reasons. For example, in skeletal muscle, GLUT4 predominately transports glucose across the plasma membrane which is further processed by oxidative (glycolysis) or nonoxidative (glycogenesis) pathways [32]. Thus, the decline in Glut4 protein in the Zip7-reduced C2C12 cells would suggest a reduction in glucose transport and subsequent genes associated with oxidative and non-oxidative pathways. Similarly, this is supported by the observation that several genes implicated in glycolysis (Galm, Gusb, Pgam2, Pgm2 and Tpi1) and glycogen synthesis (Gbe1, Agl, Pgm2, Pygm and Phkg2) were reduced in the Zip7 attenuated cells. In skeletal muscle, these genes play critical roles in the oxidative and non-oxidative pathways, respectively. This is further supported by the reduction in the mRNA of the insulin receptor (Insr), and the insulin receptor substrates 1 and 2 (Isr1 and Isr2) and the subsequent reduction of basal and insulin-mediated glycogen storage in C2C12 myotubes when Zip7 expression was attenuated (see Figure 6).
Skeletal muscle is particularly important in maintaining glucose homeostasis because approximately 70-90% of whole body insulin-mediated induction of glucose uptake occurs in muscle where it is incorporated into glycogen for storage [56]. Moreover, in insulin-resistant states, insulin-induced glucose uptake and glycogen synthesis is markedly reduced in skeletal muscle [32,57]. Accordingly, in association with a reduction in genes involved in glycogen metabolism and the fact that there was a reduction in glycogen synthesis, we also observed a significant decrease in the phosphorylation status of AKT in the reduced Zip7 expressing C2C12 cells. This is consistent with recent studies where a siRNA targeting ZIP7 significantly decreased zinc-induced pAKT after 5 minutes of 20 mM zinc treatment in MCF-7 tamoxifen-resistant breast cancer cells [8]. Moreover, in a recent study in a Zip9 gene knockout chicken DT40 cell model, the levels of phosphorylation of Akt and Erk were significantly reduced [58]. Given the role of Zip7 in facilitating zinc flux into the cytosol [8], and the fact that previous studies have shown that zinc can activate pAKT [8,59], it will be important to determine whether Zip7 in skeletal muscle plays a similar role in mediating zinc flux and signaling events that lead to phosphorylation of AKT and the mobilization of glucose transporters.
Zinc is a well-known inhibitor of protein tyrosine phosphatases (PTPs) [60] with a reported inhibition constant in the nanomolar range [61]. Zinc inhibits PTP1B, a cytoplasmic phosphatase that interacts with the insulin receptor and catalyzes its dephosphorylation resulting in the attenuation of insulin signaling [62]. Based on these results, and the fact that the insulin signaling pathway depends on the status of tyrosine phosphatases and the release of zinc into the cytosol, we hypothesize that reduced expression of Zip7 could lead to a reduction in the cytosolic zinc pool that is available for cellular signaling. For example, in the testes of diabetic mice treated with the zinc chelator, TPEN, a significant down-regulation of Akt-mediated glucose metabolism signaling was observed that was reflected by reduced phosphorylation of Akt and Gsk-3b [63]. Moreover, treatment of 3T3-L1 adipocytes with ZnCl 2 increased tyrosine phosphorylation of the insulin receptor beta subunit and enhanced the transport of glucose in the absence of insulin through the PI3-kinase/Akt pathway [53]. Furthermore, in myocytes isolated from the femoral muscle of mice with a ZnT7 knock-out (these mice display low zinc status) there was reduced insulin signaling pathway activity and these mice were insulin resistant. This was also congruent with a reduction in the mRNA expression of Insr, Irs2 and Akt [3].
Based on the observations that Zip7 plays a crucial role in facilitating cytosolic zinc flux [8], and given the role of zinc as a second messenger that activates pathways associated with cellular signaling, these studies now show a new role for Zip7 in regulating the critical gene programs involved in glucose uptake and glycogen storage in skeletal muscle. In particular, the mRNA downregulation of Insr, Irs1 and Irs2, in association with reduced phosphorylation of Akt and reduced Glut4 expression, suggests that Zip7 activity may be amenable to manipulation as a novel approach for the treatment of insulin resistance in skeletal muscle. Figure S1 A. Western blot analysis for insulin-induced phosphorylation of AKT in C2C12 skeletal muscle cells. C2C12 skeletal muscle cells were differentiated in 2% horse serum for 3 days and then treated in the absence or presence of 10 nM of insulin for 60 minutes. Total cellular protein was collected and the presence for immunoreactive pAkt and Akt was assessed. This immunoblot is a representation of three independent biologically insulin-treated C2C12 cell preparations. B. Average densitometry quantification of pAkt/Akt. pAkt quantified by densitometry on immunoblots from three independent experiments normalized to total Akt and displayed as the mean 6 SD with significant (**P#0.01,*** P#0.001) changes over time 0. (TIF) Figure S2 Western blot analysis for zinc induced phosphorylation of AKT in the absence and presence of 10 mM pyrithione in C2C12 skeletal muscle cells. C2C12 skeletal muscle cells were differentiated in 2% horse serum for 3 days and then treated in the presence (+) or absence (-) of 10 mM of pyrithione over 60 minutes. Total cellular protein was extracted and the presence for immunoreactive pAKT and AKT was performed by western blot analysis. This immunoblot represents at least three independent biological replicates. (TIF)

Supporting Information
Table S1 Fold changes in expression of glucose metabolic genes in the siRNA-Zip7 compared to the scramble control. (DOC)