Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You and Your Team.

Learn More →

The ERG1a potassium channel increases basal intracellular calcium concentration and calpain activity in skeletal muscle cells

The ERG1a potassium channel increases basal intracellular calcium concentration and calpain... Background: Skeletal muscle atrophy is the net loss of muscle mass that results from an imbalance in protein synthesis and protein degradation. It occurs in response to several stimuli including disease, injury, starvation, and normal aging. Currently, there is no truly effective pharmacological therapy for atrophy; therefore, exploration of the mechanisms contributing to atrophy is essential because it will eventually lead to discovery of an effective therapeutic target. The ether-a-go-go related gene (ERG1A)K channel has been shown to contribute to atrophy by upregulating ubiquitin proteasome proteolysis in cachectic and unweighted mice and has also been implicated in calcium modulation in cancer cells. Methods: We transduced C C myotubes with either a human ERG1A encoded adenovirus or an appropriate control 2 12 virus. We used fura-2 calcium indicator to measure intracellular calcium concentration and Calpain-Glo assay kits (ProMega) to measure calpain activity. Quantitative PCR was used to monitor gene expression and immunoblot evaluated protein abundances in cell lysates. Data were analyzed using either a Student’s t test or two-way ANOVAs and SAS software as indicated. Results: Expression of human ERG1A in C C myotubes increased basal intracellular calcium concentration 51.7% (p < 2 12 0.0001; n = 177). Further, it increased the combined activity of the calcium-activated cysteine proteases, calpain 1 and 2, by 31.9% (p <0.08; n = 24); these are known to contribute to degradation of myofilaments. The increased calcium levels are likely a contributor to the increased calpain activity; however, the change in calpain activity may also be attributable to increased calpain protein abundance and/or a decrease in levels of the native calpain inhibitor, calpastatin. To explore the enhanced calpain activity further, we evaluated expression of calpain and calpastatin genes and observed no significant differences. There was no change in calpain 1 protein abundance; however, calpain 2 protein abundance decreased 40.7% (p <0.05; n = 6). These changes do not contribute to an increase in calpain activity; however, we detected a 31.7% decrease (p <0.05; n = 6) in calpastatin which could contribute to enhanced calpain activity. Conclusions: Human ERG1A expression increases both intracellular calcium concentration and combined calpain 1 and 2 activity. The increased calpain activity is likely a result of the increased calcium levels and decreased calpastatin abundance. Keywords: ERG1A, Skeletal muscle atrophy, Calpains, Calpastatin, Intracellular calcium * Correspondence: apond@siumed.edu Anatomy Department, Southern Illinois University School of Medicine, Carbondale, IL 62902, USA Southern Illinois University, 1135 Lincoln Drive, Carbondale, IL 62902, USA Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Whitmore et al. Skeletal Muscle (2020) 10:1 Page 2 of 15 Background until we demonstrated that ERG1a protein abundance Skeletal muscle comprises approximately 40% of total hu- increases in the skeletal muscle of mice in response to man body weight and contains 50–75% of all bodily pro- hind limb suspension and tumor expression [18]. We teins. Skeletal muscle is needed for the production of further showed that, when ectopically expressed in the mechanical energy, body posture, modulation of body skeletal muscle of weight bearing mice, ERG1a increases temperature, and for generating force and movement. the abundance of the UPP E3 ligase, MuRF1, and overall Thus, a certain amount of skeletal muscle tissue is neces- UPP activity [18]. These data suggest that ERG1a partici- sary for well-being and a reduction in this tissue could pates in the process of skeletal muscle atrophy at least compromise health [1]. Skeletal muscle mass is maintained partially through modulation of the UPP [15]. We hy- by a continuous, fluctuating balance between protein deg- pothesized that ERG1a could affect other proteolytic radation and protein synthesis; however, when the rate of pathways. Indeed, human ERG1A (HERG1A) has been degradation increases or the rate of protein synthesis de- shown to increase the basal intracellular calcium con- 2+ creases, muscle mass can be lost in a process known as at- centration ([Ca ]i) of SKBr3 breast cancer cells [19] and rophy. Skeletal muscle atrophy is defined as a 5% or greater is detected in the t-tubules of cardiac tissue [17, 20] decrease in muscle mass and strength and can be induced where it has the potential to affect the calcium release by certain stimuli: muscle disuse, denervation, starvation, mechanism. Thus, we hypothesized that HERG1A would disease (e.g., diabetes and cancer), loss of neural input, and increase intracellular concentration in C C myotubes 2 12 even normal aging [2, 3]. Treatments for skeletal muscle at- and consequently enhance calpain activity. Here, we de- rophy currently under study include administration of scribe studies designed to explore this hypothesis and pharmaceuticals such as growth factors [4], beta-agonists demonstrate that indeed, ERG1A enhances both intra- [5], inhibitors of proteolysis [6, 7], stimulators of protein cellular calcium concentration and calpain activity. synthesis [8], and myostatin inhibitors [9–11]; however, these are not adequately effective. Thus, further investiga- Methods and materials tion into the mechanisms resulting in atrophy is needed to Antibodies reveal new and improved targets for therapy. The following antibodies were used: Calpain-1 polyclonal The protein degradation that contributes to atrophy antibody 3189-30 T (BioVision, Milpitas, CA); Calpain-2 occurs mainly through four proteolytic pathways: the polyclonal antibody 3372-30 T (BioVision, Milpitas, CA); ubiquitin proteasome pathway (UPP), cathepsins (the Calpain-3 polyclonal antibody A11995 (ABclonal, Woburn, autophagy-lysosome system), caspases (the apoptosis MA); Calpastatin polyclonal antibody A7634 (ABclonal, protease system), and calpain enzymes. Calpains are a Woburn, MA); MF-20 myosin antibody (Developmental family of calcium activated cysteine proteases that cleave Studies Hybridoma Bank, Iowa City, IA); laminin antibody specific proteins to release large fragments [7]. In skel- NBP2–44751 from rat (Novus, Centennial, CO); erg1 anti- etal muscle, calpain activity disassembles the sarcomere, body P9497 (Sigma, St. Louis, MO); and GAPDH poly- releasing actin and myosin to become accessible for ubi- clonal antibody ABS16 (Sigma, St. Louis, MO). quination and subsequent degradation by the prote- asome (i.e., the UPP) [12–14]. Indeed, calpains have Cell culture been shown in vitro to act upon anchoring proteins (e.g., C C myoblasts were grown in Dulbecco’smodification 2 12 titin, nebulin, and desmin) which attach the sarcomere’s of Eagle’s medium (DMEM) supplemented with 10% fetal myofilaments to the sarcomeric Z-disc [13]. The cleav- bovine serum (FBS) and maintained in a humidified incu- age of these proteins subsequently releases α-actinin and bator with 10% CO at 37 °C. To differentiate myoblasts thus results in the release of the actin thin filament from into myotubes, cells were grown in DMEM supplemented the myofibril [13, 14]. Calpains have also been shown to with 10% FBS to ~ 85% confluence. The FBS medium was degrade tropomyosin and troponin proteins [13] and, then replaced with DMEM medium supplemented with combined with the cleavage of titin, this degradation al- 2% heat-inactivated horse serum. Cells were incubated for lows for the removal of the thick filaments from the 4 days to allow for terminal differentiation. myofibrils. Calpain activity has also been shown to affect the Akt pathway which modulates the balance of protein Viral transduction synthesis and degradation [14]. Terminally differentiated C C myotubes were treated 2 12 The ERG1a (ether-a-go-go related gene) gene encodes with 200 MOI virus to produce HERG1A protein after a potassium channel known to conduct cardiac I 48 h. Specifically, for experimentation one set of cells was Kr current and be partially responsible for the repolariza- treated with control GFP encoded adeno-virus (VQAd tion of the heart action potential [15–17]. ERG1 is de- EMPTY-eGFP; ViraQuest, New Liberty, IA) while the tected in numerous mammalian tissues including brain other received the same GFP encoded adeno-viral parti- and heart, but had not been reported in skeletal muscle cles also encoding the human ERG1A K channel (VQAd Whitmore et al. Skeletal Muscle (2020) 10:1 Page 3 of 15 CMV Herg-GFP; ViraQuest). The cells were then incu- 340FX camera. The nuclei of myosin-positive cells were bated for 48 h and monitored via fluorescence to verify counted in three fields from ten slides (five treated with that the transduction was successful. HERG-encoded virus and five treated with control virus). Animals 2+ All procedures were approved by the Southern Illinois Resting intracellular Ca assay University Carbondale (SIUC) Animal Care and Use C2C12 myoblasts were cultured in DMEM supple- Committee. A total of 80 ND4-Swiss Webster 7–8- mented with 10% FBS and 1% P/S and plated at a dens- week-old male mice (Harlan-Sprague; Indianapolis, IN) ity of 5 × 10 cells/well in a black-walled 96-well plates were used. Animals were housed in SIUC vivarium facil- (Corning Life Sciences). Once myoblasts reached 80– ities on a 12 h light/dark cycle, monitored by lab animal 90% confluency, culturing media was exchanged for dif- veterinarians, and provided food and water ad libitum. ferentiation media (DMEM supplemented with 2% horse serum and 1% P/S) to promote differentiation and fusion Western blot of myoblasts into myotubes. Myoblasts were differenti- Membrane proteins were extracted from C C myo- ated for 3–4 days (2–3 days prior to a decrease in myo- 2 12 blasts and myotubes for Fig. 1a and from C C myo- tube viability within a 96-well plate), and the 2 12 tubes at 48 h after viral transduction for Figs. 1, 5, 6, 7, differentiation media was exchanged daily. Using a and 8c, b, b. Membrane proteins were extracted from multiplicity of infection of 100 (based on the initial C C cells using Tris buffer (10 mM, pH 7.4) containing number of myoblasts plated), myotubes were transduced 2 12 1 mM EDTA, 2% Triton X-100, and protease inhibitors with adenovirus encoding EGFP control or HERG. Myo- (0.5 mM pefabloc, 0.5 mM PMSF, 1 mM benzamidine, tubes were grown for two additional days, and the differ- 2+ 1 mM pepstatin, and 1 mM 1,10-phenanthroline). Sam- entiation media was refreshed daily. Prior to Ca ples were triturated using a tuberculin syringe and 23G measurements, the media was removed and myotubes needle and allowed to incubate on ice at 4 °C for 30 min were washed twice with 200 μL PBS. Then, 5 μM Fura2- and then centrifuged for 2 min at 15,000 rpm. Cellular pro- AM (Molecular Probes, Eugene, OR) was diluted in teins for Fig. 2b were extracted from C2C12 myotubes at Krebs-Ringer HEPES buffer (KRBH), and each well of 24, 48, and 72 h after transduction using Tris buffer myotubes was incubated in 100 μL of this solution for (10 mM, pH 7.4) containing 1 mM EDTA, and protease in- 1 h at RT. KRBH contained 134 mM NaCl, 3.5 mM KCl, hibitors (0.5 mM pefabloc, 0.5 mM PMSF, 1 mM benzami- 1.2 mM KH PO , 0.5 mM MgSO , 1.5 mM CaCl , 2 4 4 2 dine, 1 mM pepstatin, and 1 mM 1,10-phenanthroline). 5 mM NaHCO , and 10 mM HEPES and was supple- The samples were then centrifuged for 2 min at 15, mented with 0.05% fatty-acid free BSA (pH 7.4). After 000 rpm. All supernatants were collected and the protein this period, the Fura2-AM was removed, and myotubes content was determined using a DC protein assay kit were washed twice with KRBH. Lastly, myotubes were (BioRad, Hercules, CA) and manufacturer’s instructions. equilibrated in KRBH for 30 min at RT. Fura2 fluores- Samples were electrophoresed through a 4% poly- cence was monitored every 0.7 s for a total of 15 s using acrylamide stacking gel followed by a 7.5% poly-acrylamide a Synergy 4 Multimode Microplate Reader (BioTek In- separating gel and finally transferred to PVDF membrane struments, Winooski, VT). Fura2 was excited using a (BioRad, Hercules, CA). Membranes were immunoblotted 340/20 nm band-pass excitation filter or 380/20 nm using one or more of the antibodies listed above and devel- band-pass excitation filter, and emission was collected in oped with Immun-Star AP chemiluminescent substrate both cases using a 508/20 nm band-pass emission filter. (BioRad, Hercules, CA). Optical densities of the protein The 340/380 nm ratio at each time point was calculated bands were determined using ImageJ software (NIH). by dividing the Fura2 signal collected at 340 nm by 380 m, and these data points were averaged to yield a 2+ Fusion index resting 340/380 nm ratio, or resting Ca level, for each Myoblasts were grown on glass coverslips coated with well of myotubes. Seven independent calcium measure- rat tail collagen and then treated with either the HERG- ments were performed, with each experiment containing encoded or the control virus and allowed to terminally between six and 16 replicates, and the average 340/ differentiate. These were then immunostained for my- 380 nm ratio ± SE was calculated among all wells for osin using the DSHB antibody recognizing myosin and a GFP- and HERG-transduced myotubes. mouse on mouse (M.O.M.) Kit (Vector Labs, Inc.; Bur- lingame, CA) per manufacturer’s instructions. The cov- Quantitative real time PCR erslips were then mounted to slides with a mounting Total RNA was extracted from C C myotubes using Tri- 2 12 substance containing DAPI, and images were acquired zol reagent (Life Technologies; Carlsbad, CA) according using a Leica DM4500 microscope with a Leica DFC to manufacturer’s instructions followed by chloroform Whitmore et al. Skeletal Muscle (2020) 10:1 Page 4 of 15 Fig. 1 Transduction of C C myotubes with a HERG1A-encoded adenovirus results in elevated HERG1A protein. a Immunoblot of equal protein 2 12 content (50 μg) from lysates of non-transduced cells reveals that native ERG1 protein is 40.7% (p < 0.01; n = 6; Student’s t test) more abundant in myotubes than in myoblasts. Coomassie stained membrane confirms that equal amounts of cell lysate protein were loaded into each lane. b Immunohistochemistry labeling ERG1 protein with Alexfluor 488 (green) secondary antibody confirms that native ERG1 protein is more abundant in myotubes than in myoblasts. Representative images of immune-stained cells: (1) myoblasts immunostained with ERG1 primary antibody; (2) myoblasts immunostained without ERG1 primary antibody as control; (3) myotubes immunostained with ERG1 primary antibody; (4) myotubes immunostained without ERG1 primary antibody as control. Scale bar = 50 μm. c Transduction of C C myotubes with a HERG1A-encoded 2 12 adenovirus results in synthesis of HERG1A protein as demonstrated by immunoblot (p < 0.05; n = 6; two-way ANOVA). Coomassie stained membrane (blue) reveals that equal amounts of cell lysate protein were loaded into each lane Whitmore et al. Skeletal Muscle (2020) 10:1 Page 5 of 15 Fig. 2 Transduction of myotubes with HERG1A-encoded adenovirus is a valid in vitro skeletal muscle atrophy model. a The area of myotubes treated with HERG1A-encoded adenovirus is a significant 26.4% smaller (p < 0.01; n = 3 experimental sets) than that of control myotubes at 48 h after transduction and a significant 19.3% smaller (p < 0.01; n = 3 experimental sets) at 72 h after transduction. Scale bar = 100 μm. Bars of the graph represent the mean myotube area (μm ) while the error bars represent the standard error of the mean. b Immunoblot shows that transduction of C C myotubes with a HERG1A-encoded adenovirus yields an early increase in MuRF1 E3 ligase protein abundance while it does 2 12 not increase abundance of ATROGIN1 protein. Immunoblots are representative of three experiments solubilization and ethanol precipitation. Contaminating measure of amplicon. Changes in gene expression were DNA was degraded via DNase (RQ1 RNase-Free DNase; determined using the Livak method to normalize the gene ProMega, Madison WI). The total RNA was then reverse of interest to the “housekeeping gene.” transcribed using a GOScript™ Reverse Transcription Sys- tem Kit (Promega) per manufacturer’s instructions. Quan- Tissue sections and immunohistochemistry titative PCR was then performed using PowerUp SYBR For Fig. 4, mouse Gastrocnemius muscles were embed- green master mix (Applied Biosystems, Foster City, CA) ded in OCT, cryo-sectioned (20 μm), and stained for β- and primers for the gene of interest along with primers galactosidase (lacZ) activity as described earlier [18]. for the 18S ribosomal subunit “housekeeping gene” Sections for immunohistochemistry were fixed in cold (Table 1). An Applied Biosystems 7300 real-time PCR sys- methanol at − 20 °C for 10 min. These were then rinsed tem was used to detect SYBR green fluorescence as a with PBS at room temperature (RT) and incubated in Whitmore et al. Skeletal Muscle (2020) 10:1 Page 6 of 15 Table 1 Sequences of primers used for quantitative PCR a a Primer name (mouse) Primer sequence 5′–3′ Size (bp) Tm (°C) GC (%) Amplicon size (bp) Merg1a forward cctcgacaccatcatccgca 20 59.6 55.0 145 Merg1a reverse aggaaatcgcaggtgcaggg 20 60.3 60.0 18S subunit forward cgccgctagaggtgaaattct 21 57.2 52.4 101 18S subunit reverse agaacgaaagtcggaggttc 20 57.0 52.4 Calpain 1 forward gctaccgtttgtctagcgtc 20 58.73 55.0 98 Calpain 1 reverse taactcctctgtcatcctctggt 23 59.99 47.83 Calpain 2 forward ttttgtgcggtgtttggtcc 20 59.83 50.0 107 Calpain 2 reverse aactcagccacgaagcaagg 20 60.89 55.0 Calpain 3 forward ttcacaggaggggtgacaga 20 60.11 55.0 122 Calpain 3 reverse ttcgtgccatcgtcaatggag 21 61.01 52.38 Calpastatin forward gccttggatgacctgataga 20 53.8 50.0 115 Calpastatin reverse gtgcctcaaggtaggtagaa 20 53.7 50.0 bp base pair 3% H O for 1 h. These were then rinsed thoroughly in and methods adapted from those published previously 2 2 PBS and incubated with blocking reagent I (10% normal [22]. Brightness values were recorded as integers ranging goat serum [NGS], 0.1% bovine serum albumin [BSA; from 0 (no signal) to 256 (white). The average brightness Sigma, St. Louis, MO], and 0.1% Tween-20 in PBS) for value (± standard error of the mean, SEM) for each section 1 h at RT. The slides were then incubated for one hour was determined and analyzed by two-way ANOVA using with the laminin antibody (2 μg/mL in blocking reagent the General Linear Model Procedure of SAS 9.4 (SAS In- II–5% NGS and 0.2% TritonX100 in PBS) or in blocking stitute Inc., Cary, NC). reagent II only as a control for primary antibody bind- ing. After a thorough rinsing with PBS, the slides were Plasmids incubated overnight in the erg1 antibody (1:10 in block- The mouse Erg1a (Merg1a) clone in pBK/CMV plasmid ing reagent 2) or in blocking reagent 2 alone on the con- [23] was a generous gift from Dr. Barry London (Cardio- trol sections. The next day, the sections were rinsed vascular Institute, University of Pittsburgh, PA). The thoroughly in PBS containing 0.1% Tween-20. All sec- phRL synthetic Renilla luciferase reporter vector was tions were then incubated for 1 h at RT in Alexafluor purchased from ProMega (Madison, WI). 568 goat anti-rat IgG (1:1000 in blocking reagent II) to bind the laminin primary antibody from rat. The slides Electro-transfer were then again rinsed with PBS and incubated for one Mouse anesthesia was induced with 4% isoflurane in a hour at RT in the goat anti-rabbit secondary antibody vented chamber and maintained by administration of from the Alexafluor 488 Tyramide Super Boost Kit (Invi- 2.5% isoflurane in oxygen using a properly ventilated trogen, Carlsbad, CA). The tyramide reaction was carried nose cone with anesthesia machine and scrubber. Once out per manufacturer’sinstructionsto identify ERG1 pro- the animals were well anesthetized, the hind limbs were tein with green fluorescence. Finally, the sections were shaved and the Gastrocnemius muscles were injected rinsed thoroughly with PBS and mounted with Fluoro- with expression plasmids in 50 μL sterile saline and then mount G with DAPI (EMS; Hatfield, PA). Two sections stimulated with 8 pulses at 200 V/cm for 20 ms at 1 Hz from each muscle mid-section were analyzed. with an ECM 830 ElectroSquare Porator (BTX; Haw- thorne, NY). This method has been shown to result in Imaging ERG1a protein synthesis in skeletal muscle [15, 18]. Images were acquired using a Leica DM4500 microscope with a Leica DFC 340FX camera. Acquisition parameters Animal study design were maintained identically across samples to allow for Study 1 comparison of immunofluorescence levels when these The Merg1a plasmid (30 μg) and a plasmid encoding comparisons were made. For assay of laminin protein Renilla reporter (5 μg) were injected into the left Gastro- fluorescence, two fields were imaged per slide (one slide cnemius muscles of mice (n = 40). An empty control per mouse) and the single point brightness was measured plasmid (30 μg) and the Renilla reporter plasmid (5 μg) for 50 random consecutive points within the sarcolemma were injected into the Gastrocnemius muscles of the of each complete fiber within each field using ImageJ [21] right legs. All legs were electro-transferred to improve Whitmore et al. Skeletal Muscle (2020) 10:1 Page 7 of 15 plasmid uptake and expression. Each day, at days 0–7, measured again 10 min later to ensure that the reaction five mice were humanely killed and the Gastrocnemius had reached an end point after the first 10 min. The data muscles were harvested and frozen immediately in liquid are reported in relative light units (RLU). nitrogen. These were then stored at − 80 °C. All muscles were later thawed, homogenized, and assayed for (1) Calpain assay protein content, (2) Renilla activity to determine trans- A Calpain-Glo Kit (ProMega; Madison, WI) was used to fection efficiency, and (3) calpain activity. determine calpain activity in both myotubes and mouse muscle. Study 2 The Gastrocnemius muscles of a second set of animals, Myotubes consisting of five animals per day for days 0–5and 7 Myotubes were terminally differentiated and then trans- (n = 35), were injected and electro-transferred as de- duced with either a HERG1A-encoded adeno-virus or scribed above. After the appropriate amount of time, the the same (but non-HERG1A-encoded) virus as control animals were humanely sacrificed, the muscles were har- (12 wells each). At 48 h post-transduction, wells were vested, and total RNA was extracted for rtPCR assay. washed with two changes of 37 °C PBS and then PBS (200 μL) containing 0.2% Triton X-100 and 200 nM Study 3 epoxomicin (BostonBiochem, Cambridge, MA, Cat. #I- The Merg1a plasmid (30 μg) and a plasmid encoding a 110) was added to permeabilize the cells and to inhibit β-galactosidase (LacZ) reporter (5 μg) were injected into the proteasome, respectively. Six wells per viral treat- the left Gastrocnemius muscles of mice (n = 5). An ap- ment (HERG1A or control) received the buffer described propriate empty control plasmid (30 μg) and the LacZ (i.e., native activity); however, six wells per viral treat- reporter plasmid (5 μg) were injected into the Gastrocne- ment received buffer supplemented with the calpain in- mius muscles of the right legs. All legs were electro- hibitor MDL28170 (50 μM). These were allowed to sit at transferred to improve plasmid uptake and expression. room temperature for 5 min to ensure the myotubes At day 5, the five mice were humanely killed and the were permeabilized and the inhibitors had taken effect. Gastrocnemius muscles were harvested and frozen im- Then 200 μL of Calpain-Glo reagent was added to all mediately in liquid nitrogen. These were then stored at wells, mixed gently, and allowed to sit at room − 80 °C. All muscles were later thawed and painstakingly temperature. After 15 min, a 200 μL aliquot of the liquid serially sectioned. Serial sections were then stained for was removed from each well and placed in a white- either lacZ or dually immunostained for MERG1 and walled 96-well plate and luminescence was read using a laminin proteins as described above. Synergy H1 Hybrid Reader (BioTek Instruments, Wi- nooski, VT). The remaining well contents were scraped Protein assay from the back of the plate, triturated using a syringe and The BCA D/C Protein Assay Reagents (BioRad; Carls 26 gauge needle, and then centrifuged (13,000×g; 3 min) Bad, CA) were used to assay both samples and standards to remove any solid material. The supernatant was (0, 0.25, 0.5, 1.0, 1.25, 1.5, 2.0 mg/mL bovine serum al- assayed for protein content using the BioRad DC Protein bumin in Passive Lysis Buffer [ProMega; Madison, WI]) Assay kit. The protein data were used to normalize the for protein content, using a Synergy H1 Hybrid Reader calpain RLU activity. (BioTek; Winooski, VT) to measure absorbance at 605 nm light wavelength. Sample absorbances were in- Mouse muscle samples terpolated against the standard curve to determine the The Gastrocnemius muscles were thawed, weighed, and protein concentration of each sample. homogenized in Passive Lysis Buffer (PLB; ProMega) at a concentration of 2.5 μL buffer/μg tissue. The sample Renilla activity homogenates were aliquoted and frozen at − 80 °C. Prior To control for differences in transfection efficiency in to assay, the homogenates were thawed and sample ali- the animal muscle, a plasmid encoding the Renilla lucif- quots (40 μL) and positive control (purified porcine cal- erase enzyme was electro-transferred into muscle along pain) were added to wells of 96-well plates with assay with the Merg1a plasmid (as described above). The buffer (40 μL) having either 2 mM calcium (to activate Renilla-Glo™ Luciferase Assay System (ProMega) was calcium dependent enzymes) or 2 mM calcium plus used, according to manufacturer’s instructions, to assay 50 mM MDL28170 (to inhibit calpain specifically while homogenates for Renilla enzyme activity. The reaction allowing other calcium activated enzymes to function). was allowed to proceed for the recommended 10 min Each 96-well plate was read with a Synergy H1 Hybrid and luminescence was measured using a Synergy H1 Hy- Reader (BioTek; Winooski, VT) and activity was mea- brid Reader (BioTek; Winooski, VT). Luminescence was sured in RLU. Calpain activity was determined by Whitmore et al. Skeletal Muscle (2020) 10:1 Page 8 of 15 subtracting the RLU of the wells treated with 2 mM cal- test)and by 19.3% at72hpost transfection (p <0.01; n =6; cium and MDL28170 from the RLU of the wells treated Student’s t test). Within each time point, the difference be- with 2 mM calcium only and normalizing this RLU to tween the HERG1A-treated and control myotubes is statis- the RLU assayed with the Renilla kit to control for dif- tically significant (p < 0.01); however, there is no significant ferences in transfection efficiencies. The result was then difference in size between the myotubes treated with normalized to protein content (RLU/mg protein). HERG1A-encoding virus at the two different time points (Fig. 2a). Also similarly to mouse skeletal muscle expressing Statistics Merg1a [23], myotubes transduced with HERG1A exhibit In general, statistics were done using either a simple Student increased levels of the UPP E3 ligase, MuRF1, but not the t test or an ANOVA (as indicated in results section and fig- E3 ligase ATROGIN1 (Fig. 2b). However, when we treated ure legends) and SAS (SAS Inc.; Carey, NC). Results were myoblasts with either the HERG-encoded or the control considered significant when p < 0.05 unless otherwise noted. virus and allowed them to differentiate, we found that the HERG-expressing samples did not haveasignificantlydif- Results ferent number of myotubes containing two or more nuclei Transduction of C C myotubes with a HERG1A-encoded than the cells treated with the control virus. That is, the fu- 2 12 adenovirus results in elevated HERG1A protein sion index (myosin-positive multi-nucleated cells:total Immunoblot of equal protein aliquots from both non- myosin-positive cells evaluated) was 33.5 ± 5.0% (mean ± virus treated C C myoblast and myotube lysates detects SEM) for the cells treated with the HERG-encoded virus 2 12 a40.7% (p <0.01; n =6; Student’s t test) greater abundance while it was 31.6 ± 2.3% for the control-treated myoblasts of the ERG1 protein in myotubes than in myoblasts (p <0.74; n =14; Student’s t test). Thedatademonstrate (Fig. 1a). Immunohistochemistry work also demonstrates that HERG1A treatment of myotubes results in atrophy that there is more ERG1 protein in the C C myotubes (i.e., reduced myotube area) as it does in mouse skeletal 2 12 than in the myoblasts, revealing a stronger signal in myo- muscle; however, it does not affect the degree to which the tubes that is dispersed over the surface of the cell, while in myoblasts fuse. We conclude that we have developed a myoblasts it reveals only a very faint fluorescent signal valid in vitro model of skeletal muscle atrophy. which appears to be mainly nuclear (Fig. 1b). We trans- fected myotubes with either virus-encoding HERG1A Transduction of myotubes with a HERG1A-encoded (and GFP) or with the same, but not HERG1A-encoded, adenovirus yields a basal increase in both intracellular virus as control. Immunoblot of the lysates shows that calcium levels and calpain activity C C myotubes transfected with virus encoding HERG1A We transduced C C myotubes with either a GFP- and 2 12 2 12 do synthesize the HERG1A protein, which appears as a HERG1A-encoded adenovirus or an appropriate control single band of higher mass (likely a result of differential GFP-only encoded adenovirus. At 48 h after viral treat- glycosylation) than the native mouse ERG1 and is absent ment, we used a fura-2 calcium indicator assay and ob- from the myotubes treated with the control virus (Fig. 1c; served a significant 51.7% increase (p < 0.0001; n =90 GFP p < 0.05; two-way ANVOA). Coomassie stained mem- and n = 87 HERG1A transduced wells; Student’s t test) in brane confirms that equal amounts of protein were loaded basal intracellular calcium levels in HERG1A transduced into each well of the gel for immunoblot. myotubes relative to control (Fig. 3a). This demonstrates that HERG1A must either increase calcium influx and/or Transduction of C C myotubes with a HERG1a-encoded intracellular calcium release and/or decrease intracellular 2 12 adenovirus results in decreased myotube area and calcium re-uptake. Because HERG1A transduction results increased MuRF1 E3 ligase abundance, but no change in in increased basal intracellular calcium levels, we investi- myoblast fusion index gated the downstream effects of this increase. Specifically, We transfected myotubes with either virus-encoding using a Calpain-Glo assay kit (ProMega), we measured the HERG1A (and GFP) or with the same, but not HERG1A- combined activity of the calpain 1 and 2 enzymes in myo- encoded, virus as control. Fluorescent imaging demon- tubes treated with either the control or the HERG1A- strates that both viral particles infect myotubes (Fig. 2a). encoded virus. Some myotubes from both viral treatments Further, when the average area (μm ) of fluorescent myo- were treated with either 50 μM MDL28170 to inhibit cal- tubes from both sets is determined at both 48 and 72 h pains or an equal volume of buffer vehicle. We observed after transfection, we discover that, similarly to mouse skel- that basically the same amount of enzyme activity (control etal muscle fibers electro-transferred with Merg1a plasmid myotubes = 160.8 ± 7.3 and HERG1A-expressing myotubes [23], the myotubes transfected with HERG1A are signifi- = 167.5 ± 5.34 RLU/mg protein; n = 24) was not blocked in cantly smaller than control myotubes. Specifically, the area each well treated with the MDL28170, indicating that a ra- of the HERG1A-expressing myotubes is decreased by ther high level of non-calpain activity was assayed. None- 26.4% at 48 h post transfection (p < 0.01; n =6; Student’s t theless, we find that in control cells, the calpain activity is Whitmore et al. Skeletal Muscle (2020) 10:1 Page 9 of 15 control plasmid (n = 68 mice). We then assayed total RNA extracted from the muscles for Merg1a expression (n =28) and the muscle homogenates for calpain activity (n = 40). Quantitative PCR reveals that the electro-transfer did pro- duce Merg1a expression which was significantly higher than day 0 at days 3–5(p <0.05; Student’s t test was used to compare each day to day 0; Fig. 4a). Merg1a expression also yielded an increase in calpain activity, increasing nearly 4-fold (over day 0) by day 3 and 7.5-fold by day 4 (p <0.05; Student’s t test was used to compare each day to day 0; Fig. 4b). It returns to day 0 control levels by day 5. Thus, we show that MERG1a overexpression increases calpain ac- tivity and thus protein degradation. It is possible that the increase in intracellular calcium could lead to myofiber de- generation. Thus, we electro-transferred left mouse Gastro- cnemius muscle with a Merg1a-encoded plasmid and a Lac-Z-encoded plasmid while expressing lacZ-encoded plasmid and a an appropriate control plasmid in the right Gastrocnemius muscle and performed studies to determine if over-production of this protein would bring about changes indicative of degeneration, specifically changes in the number of centrally located nuclei or in the abundance of basal laminin. Thus, we painstakingly stained muscle ser- ial sections for lacZ (Fig. 4c) as a marker for MERG1 and dually immunostained matching serial sections for both Fig. 3 Transduction of myotubes with HERG1A-encoded adenovirus MERG1 (green fluorescence, Fig. 4d) and laminin (red increases basal intracellular calcium levels and basal calpain activity. a Fura-2 dye experiments reveal that expression of HERG1A in C C fluorescence, not shown) and used a DAPI containing 2 12 myotubes yields a 51.9% increase (p < 0.0001; n = 90 GFP and n =87 immunomount to identify nuclei (Fig. 4d). There was no re- HERG1A transduced wells) in basal intracellular calcium levels sponse in sections not stained with primary antibody relative to myotubes transduced with a control virus. b Calpain assay (Fig. 4e). The lacZ stain (blue fibers in Fig. 4c) identifies reveals that transduction of C C myotubes with a HERG1A- 2 12 where the MERG1 overexpression occurs. We find no evi- encoded adenovirus increases combined native calpain 1 and 2 activity a significant 31.9% (p < 0.08; n = 24; two-way ANOVA) over dence of any changes in the number of centrally located control myotubes. All bars represent the mean while error bars nuclei (Fig. 4d) nor in the amount of laminin fluorescence represent the standard error of the mean (Fig. 4f) in the fibers overexpressing MERG1 in any of the five mice examined nor have we seen any evidence of these 22.1% of the total native activity while it is 38.5% of the total occurrences in any of our past studies. in HERG1A-treated cells, demonstrating an increase in cal- pain activity in the HERG1A-treated cells. Because a two- HERG1A expression in myotubes does not affect way ANVOA reveals there is no real difference in the level expression of calpains 1–3 or calpastatin although it does of MDL28170 inhibited activity, we can compare the differ- affect certain protein abundances ences in assayed native activity (control versus HERG1A Calpain activity will augment with increased intracellular treated) and find that there is a 31.9% increase (p < 0.08) in calcium; however, we cannot assume that the increased activity in the HERG1A-expressing myotubes over the con- calcium is the only explanation for the increased calpain trols (Fig. 3b). Although the 0.08 probability is greater than activity. Thus, we asked if expression and/or protein the generally accepted statistical significance level of 0.05, abundances of either calpains 1, 2, or 3 or calpastatin we believe that the difference is nonetheless real. were affected by HERG1A expression. We used quanti- tative real-time PCR to discover that HERG1A expres- Merg1a expression in mouse Gastrocnemius muscle sion does not produce a statistically significant change in increases calpain activity, but did not change the number calpain 1 mRNA levels for up to 84 h after viral treat- of centrally located nuclei or laminin abundance ment (Fig. 5a). As well, no change in gene expression To test the effect of Merg1a expression on calpain activity was detected for calpains 2 or 3 (data not shown). Fur- in animals, we electro-transferred the left Gastrocnemius ther, our results indicate that there is no significant muscle of mice with an expression plasmid encoding change in calpain 1 protein abundance (Fig. 5b; n =6; Merg1a and the right leg muscle with an appropriate Student’s t test). Calpain 2, when autolyzed and hence Whitmore et al. Skeletal Muscle (2020) 10:1 Page 10 of 15 Fig. 4 Expression of mouse erg1a in mouse Gastrocnemius muscle increases Merg1a transcription and native calpain activity, but does not increase the number of centrally located nuclei or the abundance of laminin protein. a Quantitative PCR shows that electro-transfer of an expression plasmid encoding mouse erg1a (Merg1a) into mouse skeletal muscle produces Merg1a expression which is significantly higher than day 0 at days 3–5(p < 0.05; n = 28). The enclosed circles of the line graph represent the mean while the error bars represent the standard error of the mean. b Merg1a transfection in mouse skeletal muscle increases calpain activity nearly 4-fold (over day 0) by day 3 and nearly 7.5-fold by day 4(p < 0.05; n = 40). It returns to day 0 control levels by day 5 post transfection. Bars represent the mean calpain activity while error bars represent the standard error of the mean. c Positive assay for the β-galactosidase reporter (as an indicator of electro-transfer of plasmid encoding the Merg1a gene) produces a blue color. d Immunostain for MERG1 (green) of a serial section matched to the section in c demonstrates that there is indeed a greater amount of MERG1 in the fibers colored blue in c. There were no greater number of centrally located nuclei in the green fibers of any sections (n = 5 mice). e Representative of sections immunostained without primary antibody. f Over-expression of Merg1a does not produce a change in laminin abundance (p = 0.3; n = 5). Bars represent the mean single point laminin fluorescence while error bars represent the standard error of the mean. All scale bars = 50 μm activated, appears as a doublet found at ~ 75 kD [24]. abundance declined by a statistically significant 31.7% Interestingly, our results show that there is a 40.7% de- (Fig. 7b; p < 0.05; n = 6; Student’s t test). Additionally, crease (p < 0.05; n = 6; Student’s t test) in total calpain 2 there is a decrease in two of the three noted calpain 3 protein abundance in response to 48 h of HERG1A autocatalytic products (25; Fig. 8): the 114 kD isoform is treatment (Fig. 6). Calpastatin expression was not signifi- down 29.6% and the 60 kD isoform is down 29.2%, al- cantly affected by the HERG1A channel for up to 84 h though the 30 kD isoform is not affected (p < 0.03; n =6; post-transduction (Fig. 7a); however, calpastatin protein Student’s t test within protein isomer). When the optical Whitmore et al. Skeletal Muscle (2020) 10:1 Page 11 of 15 Fig. 5 Neither calpain 1 expression nor protein abundance changes after transduction of myotubes with HERG1A-encoded adenovirus. a Quantitative PCR reveals that there is no change in expression of calpain 1 for up to 84 h after transduction (n = 15). bImmunoblot demonstrates that there is no significant change in calpain 1 protein abundance at 48 h after viral transduction (n = 6). Bars represent the mean and the error bars represent the standard error of the mean. Coomassie staining of the blotted membrane shows that equal amounts of protein were loaded into each well of the gel densities for all protein bands are summed, there is a either control or HERG1A-encoded adenovirus, we show 2+ total 21.0% decrease in calpain 3 protein abundance. that HERG1A expression also increases basal [Ca ]i and calpain activity. There are numerous potential sources of 2+ Discussion the calcium that contributes to the increased [Ca ]i. For The ERG1a voltage-gated K channel is responsible for example, it is possible that ERG1A is modulating Cav1.1 late phase repolarization of the cardiac action potential channels in the skeletal muscle sarcolemmal membrane, and was reported to be absent from skeletal muscle [23, resulting in an influx potentially from both the external 25]; however, the Pond and Hannon labs demonstrated milieu and internal stores. Further, because ERG1A is lo- that this protein is detectable in the atrophying skeletal cated in the t-tubules of cardiac tissue [17, 20], it is pos- muscle of mice and in very low abundance in healthy ro- sible that it is located in the t-tubules of skeletal muscle, dent muscle with careful use of protease inhibitors and where it could contribute to the release of calcium from concentration of solubilized membrane proteins [18]. Sub- internal stores by modulation of ryanodine receptors and/ sequent studies showed that ERG1a expression leads to an or IP3 receptors. Indeed, changes in regulation of sarco- increase in abundance of the MURF1 E3 ubiquitin ligase lemmal permeability could have severe consequences for protein and enhances ubiquitin proteasome proteolysis, a skeletal muscle tissue, potentially producing diseases such pathway known to contribute to skeletal muscle atrophy as muscular dystrophies and Niemann-Pick disease [26, [15, 18]. Here, using C C myotubes transduced with 27]. The source of the increased calcium is currently 2 12 Whitmore et al. Skeletal Muscle (2020) 10:1 Page 12 of 15 blocking calpain activity reduced the activation of cal- pain 1 gene expression and attenuated skeletal muscle atrophy [29]. Here, we report that there is no detectable change in calpain 1 protein abundance in myotubes transduced with HERG1A while surprisingly we detect a decrease in calpain 2 protein abundance. These data demonstrate that the increased calpain activity is not a result of increased enzyme protein abundance. We sug- gest that the decreased calpain 2 protein abundance could result from either decreased calpain 2 synthesis and/or increased calpain 2 protein degradation. Quanti- tative PCR data demonstrate that there is no significant change in transcription of calpain 1 or 2 genes for up to 84 h post transduction. Interestingly, we observe a de- crease in calpain 2 protein abundance without detecting a change in transcription of that gene. Thus, although mRNA production is not always directly correlated with protein abundance, we can speculate that the calpain 2 Fig. 6 Calpain 2 protein abundance decreases (p < 0.05; n =6) 48 h protein may be undergoing an increased level of degrad- after myotube transduction with HERG1A-encoded adenovirus. Bars ation. Indeed, these proteins may be undergoing autoly- represent the mean and error bars represent the standard error of sis or it is possible that ubiquitin proteasome proteolysis the mean. Coomassie staining of the blotted membrane confirms that equal amounts of protein were loaded into each well of the gel of calpain 2 is enhanced. Indeed, we have shown that in- creased ERG1 expression increases UPP activity. Calpastatin is a native calpain inhibitor which inhibits under investigation in our laboratories. However, because conventional calpains 1 and 2, but not calpain 3. Calpas- we find no change in the fusion index or an increase in ei- tatin requires calcium to bind calpains so that when the ther the number of centrally located nuclei or in the abun- calcium concentrations rise, calpain activity is increased, dance of laminin fluorescence in the fibers over- but so is calpastatin binding [13, 30]. Indeed, a decrease expressing Merg1a, we believe that our data suggest that in calpastatin protein would lower the inhibition of cal- the channel (which we find to be in very low abundance pains and allow for increased calpain-mediated proteoly- in muscle normally) is simply upregulating protein deg- sis. Certainly, the increased level of calpain activity radation in our myotubes. It is also possible that the low assayed in the mouse muscle homogenates, in which the 2+] levels of increased calcium are affecting signaling path- [Ca i is disrupted, suggests that something other than 2+ ways, but that remains to be investigated. [Ca ]i must contribute to enhanced calpain activity. The explanation for the increased calpain activity may Calpain 3 is a non-classical calpain which is detected 2+ seem obvious—the increase in [Ca ]i. However, we ec- mainly in skeletal muscle. It undergoes calcium- topically expressed mouse erg1a (Merg1a) in mouse mediated autolysis that has been reported to be en- Gastrocnemius muscle and homogenized the muscle, hanced by ATP at lower calcium concentrations [31, 2+ thereby disrupting the [Ca ]i pool and equalizing the 32]. Evidence has shown that the absence of calpain 3 calcium concentration throughout the sample. We then leads to a reduction in protein turnover and results in assayed for calpain activity and discovered that even in accumulation of damaged and/or misfolded proteins the homogenate it is still higher in the Merg1a-express- which can lead to cellular stress and eventual muscle 2+ ing tissue. This study is evidence that increased [Ca ]i pathology [33, 34]. Indeed, the absence or reduction of may not be the only factor that contributes to the this protein has been shown to lead to limb-girdle mus- ERG1A-induced increase in calpain activity. Other pos- cular dystrophy type 2A (LGMD2A) in humans [30–32, sible contributors include increased calpain 1 and/or 2 34–37]. Studies suggest that calpain 3 takes part in re- protein and/or decreased calpastatin protein. modeling of the sarcomere in response to cellular dam- Calpains 1 (μ-calpain) and 2 (m-calpain) are both clas- age such as atrophy [34, 36, 37]. Interestingly, studies sical calpains and are detected throughout the body, in- with calpain 3 knockout mice suggest that calpain 3 acts cluding skeletal muscle [28]. Indeed, calpain activity has upstream of the UPP, although it is uncertain if calpain been demonstrated to contribute to muscle atrophy [28]. 3 directly cleaves proteins to make them accessible for For example, Shenkman and colleagues inhibited calpain ubiquitination [34]. Thus, calpain 3 appears to be pro- activity in hind limb suspended mice by treatment with tective against muscle loss and its protein abundance the calpain inhibitor PD150606 and demonstrated that might be expected to be lower in an atrophic situation. Whitmore et al. Skeletal Muscle (2020) 10:1 Page 13 of 15 Fig. 7 Calpastatin expression does not change after transduction with HERG1A-encoded adenovirus although protein abundance decreases. a Quantitative PCR reveals that levels of calpastatin mRNA do not significantly change for up to 84 h after viral transduction with HERG1A encoded adenovirus. b, c. Immunoblot detects a significant 31.7% decrease in protein abundance (p < 0.05; n = 6) at 48 h after transduction. All bars represent the mean ± the standard error of the mean. Coomassie staining of the blotted membrane confirms that equal amounts of protein were loaded into each well of the gel Indeed, we report that calpain 3 protein abundance de- calpain activity. This is not surprising considering that creases in response to HERG1A expression. The de- calpastatin binding is also enhanced by intracellular cal- crease may be related to a decreased ability to remodel cium. Calpain 3 activity was not measured here; how- the sarcomere during/after atrophy; however, this possi- ever, the decline in calpain 3 protein is consistent with bility would require much additional investigation. an atrophic environment. Interestingly, classical calpain In summary, we show that HERG1A increases calpain activity has been shown to degrade sarcomeric anchor activity in myotubes, likely resulting from the increase in proteins (e.g., titin, nebulin) and this allows for release of 2+ [Ca ]i. We detect no increases in abundances of cal- contractile proteins (e.g., myosin and actin) into the pains 1 or 2 proteins which would otherwise contribute cytosol where they can be accessed and degraded by the to enhanced calpain activity. In fact, we report a decline UPP [30, 38]. Here, we show that HERG1A modulates in the abundance of calpain 2 protein. Thus, it would intracellular calcium and calpain activity. Because its 2+ appear that the increased [Ca ]i could be the main con- interaction with calcium and calpains is upstream of the tributor to the enhanced calpain activity; however, there UPP, and it also modulates UPP activity [18], we is a significant decline in calpastatin protein abundance hypothesize that ERG1A may indeed contribute to co- which likely also contributes to the measured increase in ordination of proteolytic systems which produce skeletal Whitmore et al. Skeletal Muscle (2020) 10:1 Page 14 of 15 Fig. 8 Calpain 3 protein abundance decreased 21.0% in response to transduction of myotubes with HERG1A-encoded adenovirus. Immunoblot shows that calpain 3 degraded into numerous fragments as expected, including three notable autocatalytic products: 114 kD (down 29.6%), 60 kD (down 29.2%), and 30 kD which was not affected. Bars represent the mean ± the standard error of the mean. Coomassie staining of the blotted membrane shows that equal amounts of protein were loaded into each well of the gel muscle atrophy, specifically calpain and UPP activities. fura-2 assays to determine intracellular calcium concentrations. LA cultured and transduced myotubes and then imaged myotubes and determined their Further study is needed to learn how ERG1A functions area. KB, SML, and MNH performed the electro-transfer on mice hind limbs. in skeletal muscle. Indeed, because of the role of the KB and SML performed the calpain assays on the electro-transferred muscles. ERG1A/ERG1B heteromultimeric channel in cardiac ac- AKU imaged myotubes and consulted on content and writing of manuscript. RW provided direction on calpain assays and consulted on content and writ- tion potential repolarization, ERG1A will likely never be ing of manuscript. JKD cultured, imaged, and evaluated myoblasts and myo- a target for pharmacological treatment of atrophy; how- tubes and consulted on content and writing of manuscript. WHW cloned the ever, continuing study of this protein may reveal other HERG1A construct into the viral cassette, provided guidance for primer devel- opment, and consulted on content of manuscript. GHH provided over all possible targets to combat atrophy. guidance to EP for measurement of calcium concentration, acted as co-PI on the grant which funded the bulk of this work, and consulted on content and Abbreviations writing of manuscript. ALP worked in the laboratory to produce some of the DMEM: Dulbecco’s modification of Eagle’s medium; ERG1A: Ether-a-gogo- data, provided over all guidance to the project, acted as co-PI on the grant related gene; FBS: Fetal bovine serum; HERG1A: Human ether-a-gogo-related which funded the bulk of this work, and consulted on content and writing gene; Merg1a: Mouse ether-a-gogo-related gene; RLU: Relative light units; of manuscript. All authors read and approved the final manuscript. UPP: Ubiquitin proteasome pathway Authors’ information Acknowledgements Not desired. Not applicable. Funding Authors’ contributions This work was funded in part by the Southern Illinois University to ALP and CW cultured and transduced myotubes, performed calpain assays, in part by the Department of Defense Office of the Congressionally Directed completed the PCR and immunoblotting work, and wrote the original draft Medical Research Programs in the form of a Discovery Award (PR170326) to of the manuscript. EP cultured and transduced myotubes and performed ALP and GHH. Whitmore et al. Skeletal Muscle (2020) 10:1 Page 15 of 15 Availability of data and materials 17. Jones EM, Roti Roti EC, Wang J, Delfosse SA, Robertson GA. Cardiac IKr The datasets used and/or analyzed during the current study are available channels minimally comprise hERG 1a and 1b subunits. J Biol Chem. 2004; from the corresponding author on reasonable request. 279:44690–4. 18. Wang X, Hockerman GH, Green HW 3rd, Babbs CF, Mohammad SI, Gerrard D, et al. Merg1a K+ channel induces skeletal muscle atrophy by activating Ethics approval and consent to participate the ubiquitin proteasome pathway. FASEB J. 2006;20:1531–3. All animal work and studies were approved the SIU IACUC. 19. Perez-Neut M, Shum A, Cuevas BD, Miller R, Gentile S. Stimulation of hERG1 channel activity promotes a calcium dependent degradation of cyclin E2, Consent for publication but not cyclin E1, in breast cancer cells. Oncotarget. 2015;6:1631–9. Not applicable. 20. Rasmussen HB, Moller M, Knaus H, Jensen BS, Olesen S, Jorgensen NK. Subcellular localization of the delayed rectifier K channels KCNQ1 and ERG1 Competing interests in the rat heart. Am J Physiol Heart Circ Physiol. 2003;286:H1300–H9. The authors declare that they have no competing interests. 21. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671–5. PMID 22930834 Author details 22. Cheatwood JL, Emerick AJ, Schwab ME, Kartje GL. Nogo-a expression after Anatomy Department, Southern Illinois University School of Medicine, focal ischemic stroke in the adult rat. Stroke. 2008;39:2091–8. https://doi.org/ Carbondale, IL 62902, USA. Medicinal Chemistry and Molecular 10.1161/STROKEAHA.107.507426. Pharmacology, Purdue University, West Lafayette, IN 47906, USA. Doisey 23. London B, Trudeau MC, Newton KP, Beyer AK, Copeland NG, Gilbert DJ, College of Health Sciences, Saint Louis University, St. Louis, MO 63103, USA. et al. Two isoforms of the mouse ether-a-go-go-related gene coassemble to School of Liberal Arts, Sciences and Education, Ivy Tech State college, form channels with properties similar to the rapidly activating component Lafayette, IN 47905, USA. Biochemistry Department, Southern Illinois of the cardiac delayed rectifier K+ current. Circ Res. 1997;81:870–8. University School of Medicine, Carbondale, IL 62902, USA. Gene Editing Core 24. Hongqui L, Thompson VF, Goll DE. Effects of autolysis on properties of μ- Facility, Purdue University, West Lafayette, IN 47906, USA. Southern Illinois and m-calpain. Biochim Biophys Acta. 1691;2004:91–103. https://doi.org/10. University, 1135 Lincoln Drive, Carbondale, IL 62902, USA. 1016/j.bbamcr.2003.12.006. 25. Curran M, Splawski I, Timothy KW, Vincent GM, Green ED, Keating MT. A Received: 4 June 2019 Accepted: 18 December 2019 molecular basis for cardiac arrhythmia: HERG mutations cause long QT syndrome. Cell. 1995;80:795–803. 26. Michailowsky V, Li H, Mittra B, Iyer SR, Mazala DAG, Corrotte M, Wang Y, References Chin ER, Lovering RM, Andrews NW. Defects in sarcolemma repair and 1. Frontera WR, Ochala J. Skeletal muscle: a brief review of structure and skeletal muscle function after injury in a mouse model of Niemann-pick function. Calcif Tissue Int. 2015;96:183–95. type a/B disease. Skelet Muscle. 2019;9:1–5. 2. Bodine SC, Baehr LM. Skeletal muscle atrophy and the E3 ubiquitin ligases 27. Gumerson JD, Michele DE. The dystrophin-glycoprotein complex ini the MuRF1 and MAFbx/atrogin-1. Am J Physiol Endocrinol Metab. 2014;307: prevention of muscle damage. J Biomed Biotechnol. 2011;210797:1–13. E469–84. 28. Huang J, Zhu X. The molecular mechanism of calpains action on skeletal 3. Bonaldo P, Sandri M. Cellular and molecular mechanisms of muscle atrophy. muscle atrophy. Physiol Res. 2016;65:547–60. Dis Model Mech. 2013;6:25–39. 29. Shenkman BS, Belova SP, Lomonosova YN, Kostrominova TY, Nemirovskaya 4. Baoge L, Van Den Steen E, Rimbaut S, Philips N, Witvrouw E, Almqvist KF, TL. Calpain-dependent regulation of the skeletal muscle atrophy following et al. Treatment of skeletal muscle injury: a review. ISRN Orthop. 2012;2012: unloading. Arch Biochem Biophys. 2015;584:36–41. 30. Sorimachi H, Ono Y. Regulation and physiological roles of the calpain 5. Lynch GS, Ryall JG. Role of beta-adrenoceptor signaling in skeletal muscle: system in muscular disorders. Cardiovasc Res. 2012;96:11–22. implications for muscle wasting and disease. Physiol Rev. 2008;88:729–67. 31. Murphy RM, Lamb GD. Endogenous calpain-3 activation is primarily 6. Derbre F, Ferrando B, Gomez-Cabrera MC, Sanchis-Gomar F, Martinez-Bello 2+ governed by small increases in resting cytoplasmic [Ca ] and is not VE, Olaso-Gonzalez G, et al. Inhibition of xanthine oxidase by allopurinol dependent on stretch. J Biol Chem. 2009;284:7811–9. prevents skeletal muscle atrophy: role of p38 MAPKinase and E3 ubiquitin 32. Murphy RM, Vissing K, Latchman H, Lamboley C, McKenna MJ, Overgaard K, ligases. PLoS One. 2012;7:e46668. Lamb GD. Activation of skeletal muscle calpain-3 by eccentric exercise in 7. Fareed MU, Evenson AR, Wei W, Menconi M, Poylin V, Petkova V, et al. humans does not result in its translocation to the nucleus or cytosol. J Appl Treatment of rats with calpain inhibitors prevents sepsis-induced muscle Physiol. 2011;111(5):1448–58. proteolysis independent of atrogin-1/MAFbx and MuRF1 expression. Am J 33. Duguez S, Bartoli M, Richard I. Calpain 3: a key regulator of the sarcomere? Physiol Regul Integr Comp Physiol. 2006;290:R1589–97. FEBS J. 2006;273:3427–36. 8. Guasconi V, Puri PL. Epigenetic drugs in the treatment of skeletal muscle 34. Kramerova I, Kudryashova E, Venkatraman G, Spencer MJ. Calpain 3 atrophy. Curr Opin Clin Nutr Metab Care. 2008;11:233–41. participates in sarcomere remodeling by acting upstream of the ubiquitin- 9. Han HQ, Mitch WE. Targeting the myostatin signaling pathway to treat proteasome pathway. Hum Mol Genet. 2005;14:2125–34. muscle wasting diseases. Curr Opin Support Palliat Care. 2011;5:334–41. 35. Meznaric M, Writzl K. Limb-girdle muscular dystrophies: different types and 10. Hemmati-Brivanlou A, Kelly O, Melton D. Follistatin, an antagonist of activin, diagnosis. In: Willem M, editor. Skeletal Muscle. London: Nova Science is expressed in the Spemann organizer and displays direct neuralizing Publishers, Inc.; 2013. p. 105–27. activity. Cell. 1994;77:283–95. 36. Murphy RM, Goodman CA, McKenna MJ, Bennie J, Leikis M, Lamb GD. Calpain- 11. Smith RC, Lin BK. Myostatin inhibitors as therapies for muscle wasting 3 is autolyzed and hence activated in human skeletal muscle 24 h following a associated with cancer and other disorders. Curr Opin Support Palliat Care. single bout of eccentric exercise. J Appl Physiol. 2007;103:926–31. 2013;7:352–60. 37. Ono Y, Ojima K, Shinkai-Ouchi F, Hata S, Sorimachi H. An eccentric calpain, 12. Ma L, Chu W, Chai J, Shen C, Li D, Wang X. ER stress and subsequent CAPN3/p94/calpain-3. Biochimie. 2016;122:169–87. activated calpain play a pivotal role in skeletal muscle wasting after severe 38. Campbell RL, Davies PL. Structure-function relationships in calpains. burn injury. PLoS One. 2017;12:e0186128. Biochem J. 2012;447:335–51. 13. Goll DE, Neti G, Mares SW, Thompson VF. Myofibrillar protein turnover: the proteasome and the calpains. J Anim Sci. 2008;86:E19–35. 14. Smith IJ, Lecker SH, Hasselgren PO. Calpain activity and muscle wasting in Publisher’sNote sepsis. Am J Physiol Endocrinol Metab. 2008;295:E762–71. Springer Nature remains neutral with regard to jurisdictional claims in 15. Pond AL, Nedele C, Wang WH, Wang X, Walther C, Jaeger C, et al. The published maps and institutional affiliations. mERG1a channel modulates skeletal muscle MuRF1, but not MAFbx, expression. Muscle Nerve. 2014;49:378–88. 16. Vandenberg JI, Perry MD, Perrin MJ, Mann SA, Ke Y, Hill AP. hERG K(+) channels: structure, function, and clinical significance. Physiol Rev. 2012;92: 1393–478. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Skeletal Muscle Springer Journals

The ERG1a potassium channel increases basal intracellular calcium concentration and calpain activity in skeletal muscle cells

Loading next page...
 
/lp/springer-journals/the-erg1a-potassium-channel-increases-basal-intracellular-calcium-c067ZE40Cs
Publisher
Springer Journals
Copyright
Copyright © The Author(s). 2020
eISSN
2044-5040
DOI
10.1186/s13395-019-0220-3
Publisher site
See Article on Publisher Site

Abstract

Background: Skeletal muscle atrophy is the net loss of muscle mass that results from an imbalance in protein synthesis and protein degradation. It occurs in response to several stimuli including disease, injury, starvation, and normal aging. Currently, there is no truly effective pharmacological therapy for atrophy; therefore, exploration of the mechanisms contributing to atrophy is essential because it will eventually lead to discovery of an effective therapeutic target. The ether-a-go-go related gene (ERG1A)K channel has been shown to contribute to atrophy by upregulating ubiquitin proteasome proteolysis in cachectic and unweighted mice and has also been implicated in calcium modulation in cancer cells. Methods: We transduced C C myotubes with either a human ERG1A encoded adenovirus or an appropriate control 2 12 virus. We used fura-2 calcium indicator to measure intracellular calcium concentration and Calpain-Glo assay kits (ProMega) to measure calpain activity. Quantitative PCR was used to monitor gene expression and immunoblot evaluated protein abundances in cell lysates. Data were analyzed using either a Student’s t test or two-way ANOVAs and SAS software as indicated. Results: Expression of human ERG1A in C C myotubes increased basal intracellular calcium concentration 51.7% (p < 2 12 0.0001; n = 177). Further, it increased the combined activity of the calcium-activated cysteine proteases, calpain 1 and 2, by 31.9% (p <0.08; n = 24); these are known to contribute to degradation of myofilaments. The increased calcium levels are likely a contributor to the increased calpain activity; however, the change in calpain activity may also be attributable to increased calpain protein abundance and/or a decrease in levels of the native calpain inhibitor, calpastatin. To explore the enhanced calpain activity further, we evaluated expression of calpain and calpastatin genes and observed no significant differences. There was no change in calpain 1 protein abundance; however, calpain 2 protein abundance decreased 40.7% (p <0.05; n = 6). These changes do not contribute to an increase in calpain activity; however, we detected a 31.7% decrease (p <0.05; n = 6) in calpastatin which could contribute to enhanced calpain activity. Conclusions: Human ERG1A expression increases both intracellular calcium concentration and combined calpain 1 and 2 activity. The increased calpain activity is likely a result of the increased calcium levels and decreased calpastatin abundance. Keywords: ERG1A, Skeletal muscle atrophy, Calpains, Calpastatin, Intracellular calcium * Correspondence: apond@siumed.edu Anatomy Department, Southern Illinois University School of Medicine, Carbondale, IL 62902, USA Southern Illinois University, 1135 Lincoln Drive, Carbondale, IL 62902, USA Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Whitmore et al. Skeletal Muscle (2020) 10:1 Page 2 of 15 Background until we demonstrated that ERG1a protein abundance Skeletal muscle comprises approximately 40% of total hu- increases in the skeletal muscle of mice in response to man body weight and contains 50–75% of all bodily pro- hind limb suspension and tumor expression [18]. We teins. Skeletal muscle is needed for the production of further showed that, when ectopically expressed in the mechanical energy, body posture, modulation of body skeletal muscle of weight bearing mice, ERG1a increases temperature, and for generating force and movement. the abundance of the UPP E3 ligase, MuRF1, and overall Thus, a certain amount of skeletal muscle tissue is neces- UPP activity [18]. These data suggest that ERG1a partici- sary for well-being and a reduction in this tissue could pates in the process of skeletal muscle atrophy at least compromise health [1]. Skeletal muscle mass is maintained partially through modulation of the UPP [15]. We hy- by a continuous, fluctuating balance between protein deg- pothesized that ERG1a could affect other proteolytic radation and protein synthesis; however, when the rate of pathways. Indeed, human ERG1A (HERG1A) has been degradation increases or the rate of protein synthesis de- shown to increase the basal intracellular calcium con- 2+ creases, muscle mass can be lost in a process known as at- centration ([Ca ]i) of SKBr3 breast cancer cells [19] and rophy. Skeletal muscle atrophy is defined as a 5% or greater is detected in the t-tubules of cardiac tissue [17, 20] decrease in muscle mass and strength and can be induced where it has the potential to affect the calcium release by certain stimuli: muscle disuse, denervation, starvation, mechanism. Thus, we hypothesized that HERG1A would disease (e.g., diabetes and cancer), loss of neural input, and increase intracellular concentration in C C myotubes 2 12 even normal aging [2, 3]. Treatments for skeletal muscle at- and consequently enhance calpain activity. Here, we de- rophy currently under study include administration of scribe studies designed to explore this hypothesis and pharmaceuticals such as growth factors [4], beta-agonists demonstrate that indeed, ERG1A enhances both intra- [5], inhibitors of proteolysis [6, 7], stimulators of protein cellular calcium concentration and calpain activity. synthesis [8], and myostatin inhibitors [9–11]; however, these are not adequately effective. Thus, further investiga- Methods and materials tion into the mechanisms resulting in atrophy is needed to Antibodies reveal new and improved targets for therapy. The following antibodies were used: Calpain-1 polyclonal The protein degradation that contributes to atrophy antibody 3189-30 T (BioVision, Milpitas, CA); Calpain-2 occurs mainly through four proteolytic pathways: the polyclonal antibody 3372-30 T (BioVision, Milpitas, CA); ubiquitin proteasome pathway (UPP), cathepsins (the Calpain-3 polyclonal antibody A11995 (ABclonal, Woburn, autophagy-lysosome system), caspases (the apoptosis MA); Calpastatin polyclonal antibody A7634 (ABclonal, protease system), and calpain enzymes. Calpains are a Woburn, MA); MF-20 myosin antibody (Developmental family of calcium activated cysteine proteases that cleave Studies Hybridoma Bank, Iowa City, IA); laminin antibody specific proteins to release large fragments [7]. In skel- NBP2–44751 from rat (Novus, Centennial, CO); erg1 anti- etal muscle, calpain activity disassembles the sarcomere, body P9497 (Sigma, St. Louis, MO); and GAPDH poly- releasing actin and myosin to become accessible for ubi- clonal antibody ABS16 (Sigma, St. Louis, MO). quination and subsequent degradation by the prote- asome (i.e., the UPP) [12–14]. Indeed, calpains have Cell culture been shown in vitro to act upon anchoring proteins (e.g., C C myoblasts were grown in Dulbecco’smodification 2 12 titin, nebulin, and desmin) which attach the sarcomere’s of Eagle’s medium (DMEM) supplemented with 10% fetal myofilaments to the sarcomeric Z-disc [13]. The cleav- bovine serum (FBS) and maintained in a humidified incu- age of these proteins subsequently releases α-actinin and bator with 10% CO at 37 °C. To differentiate myoblasts thus results in the release of the actin thin filament from into myotubes, cells were grown in DMEM supplemented the myofibril [13, 14]. Calpains have also been shown to with 10% FBS to ~ 85% confluence. The FBS medium was degrade tropomyosin and troponin proteins [13] and, then replaced with DMEM medium supplemented with combined with the cleavage of titin, this degradation al- 2% heat-inactivated horse serum. Cells were incubated for lows for the removal of the thick filaments from the 4 days to allow for terminal differentiation. myofibrils. Calpain activity has also been shown to affect the Akt pathway which modulates the balance of protein Viral transduction synthesis and degradation [14]. Terminally differentiated C C myotubes were treated 2 12 The ERG1a (ether-a-go-go related gene) gene encodes with 200 MOI virus to produce HERG1A protein after a potassium channel known to conduct cardiac I 48 h. Specifically, for experimentation one set of cells was Kr current and be partially responsible for the repolariza- treated with control GFP encoded adeno-virus (VQAd tion of the heart action potential [15–17]. ERG1 is de- EMPTY-eGFP; ViraQuest, New Liberty, IA) while the tected in numerous mammalian tissues including brain other received the same GFP encoded adeno-viral parti- and heart, but had not been reported in skeletal muscle cles also encoding the human ERG1A K channel (VQAd Whitmore et al. Skeletal Muscle (2020) 10:1 Page 3 of 15 CMV Herg-GFP; ViraQuest). The cells were then incu- 340FX camera. The nuclei of myosin-positive cells were bated for 48 h and monitored via fluorescence to verify counted in three fields from ten slides (five treated with that the transduction was successful. HERG-encoded virus and five treated with control virus). Animals 2+ All procedures were approved by the Southern Illinois Resting intracellular Ca assay University Carbondale (SIUC) Animal Care and Use C2C12 myoblasts were cultured in DMEM supple- Committee. A total of 80 ND4-Swiss Webster 7–8- mented with 10% FBS and 1% P/S and plated at a dens- week-old male mice (Harlan-Sprague; Indianapolis, IN) ity of 5 × 10 cells/well in a black-walled 96-well plates were used. Animals were housed in SIUC vivarium facil- (Corning Life Sciences). Once myoblasts reached 80– ities on a 12 h light/dark cycle, monitored by lab animal 90% confluency, culturing media was exchanged for dif- veterinarians, and provided food and water ad libitum. ferentiation media (DMEM supplemented with 2% horse serum and 1% P/S) to promote differentiation and fusion Western blot of myoblasts into myotubes. Myoblasts were differenti- Membrane proteins were extracted from C C myo- ated for 3–4 days (2–3 days prior to a decrease in myo- 2 12 blasts and myotubes for Fig. 1a and from C C myo- tube viability within a 96-well plate), and the 2 12 tubes at 48 h after viral transduction for Figs. 1, 5, 6, 7, differentiation media was exchanged daily. Using a and 8c, b, b. Membrane proteins were extracted from multiplicity of infection of 100 (based on the initial C C cells using Tris buffer (10 mM, pH 7.4) containing number of myoblasts plated), myotubes were transduced 2 12 1 mM EDTA, 2% Triton X-100, and protease inhibitors with adenovirus encoding EGFP control or HERG. Myo- (0.5 mM pefabloc, 0.5 mM PMSF, 1 mM benzamidine, tubes were grown for two additional days, and the differ- 2+ 1 mM pepstatin, and 1 mM 1,10-phenanthroline). Sam- entiation media was refreshed daily. Prior to Ca ples were triturated using a tuberculin syringe and 23G measurements, the media was removed and myotubes needle and allowed to incubate on ice at 4 °C for 30 min were washed twice with 200 μL PBS. Then, 5 μM Fura2- and then centrifuged for 2 min at 15,000 rpm. Cellular pro- AM (Molecular Probes, Eugene, OR) was diluted in teins for Fig. 2b were extracted from C2C12 myotubes at Krebs-Ringer HEPES buffer (KRBH), and each well of 24, 48, and 72 h after transduction using Tris buffer myotubes was incubated in 100 μL of this solution for (10 mM, pH 7.4) containing 1 mM EDTA, and protease in- 1 h at RT. KRBH contained 134 mM NaCl, 3.5 mM KCl, hibitors (0.5 mM pefabloc, 0.5 mM PMSF, 1 mM benzami- 1.2 mM KH PO , 0.5 mM MgSO , 1.5 mM CaCl , 2 4 4 2 dine, 1 mM pepstatin, and 1 mM 1,10-phenanthroline). 5 mM NaHCO , and 10 mM HEPES and was supple- The samples were then centrifuged for 2 min at 15, mented with 0.05% fatty-acid free BSA (pH 7.4). After 000 rpm. All supernatants were collected and the protein this period, the Fura2-AM was removed, and myotubes content was determined using a DC protein assay kit were washed twice with KRBH. Lastly, myotubes were (BioRad, Hercules, CA) and manufacturer’s instructions. equilibrated in KRBH for 30 min at RT. Fura2 fluores- Samples were electrophoresed through a 4% poly- cence was monitored every 0.7 s for a total of 15 s using acrylamide stacking gel followed by a 7.5% poly-acrylamide a Synergy 4 Multimode Microplate Reader (BioTek In- separating gel and finally transferred to PVDF membrane struments, Winooski, VT). Fura2 was excited using a (BioRad, Hercules, CA). Membranes were immunoblotted 340/20 nm band-pass excitation filter or 380/20 nm using one or more of the antibodies listed above and devel- band-pass excitation filter, and emission was collected in oped with Immun-Star AP chemiluminescent substrate both cases using a 508/20 nm band-pass emission filter. (BioRad, Hercules, CA). Optical densities of the protein The 340/380 nm ratio at each time point was calculated bands were determined using ImageJ software (NIH). by dividing the Fura2 signal collected at 340 nm by 380 m, and these data points were averaged to yield a 2+ Fusion index resting 340/380 nm ratio, or resting Ca level, for each Myoblasts were grown on glass coverslips coated with well of myotubes. Seven independent calcium measure- rat tail collagen and then treated with either the HERG- ments were performed, with each experiment containing encoded or the control virus and allowed to terminally between six and 16 replicates, and the average 340/ differentiate. These were then immunostained for my- 380 nm ratio ± SE was calculated among all wells for osin using the DSHB antibody recognizing myosin and a GFP- and HERG-transduced myotubes. mouse on mouse (M.O.M.) Kit (Vector Labs, Inc.; Bur- lingame, CA) per manufacturer’s instructions. The cov- Quantitative real time PCR erslips were then mounted to slides with a mounting Total RNA was extracted from C C myotubes using Tri- 2 12 substance containing DAPI, and images were acquired zol reagent (Life Technologies; Carlsbad, CA) according using a Leica DM4500 microscope with a Leica DFC to manufacturer’s instructions followed by chloroform Whitmore et al. Skeletal Muscle (2020) 10:1 Page 4 of 15 Fig. 1 Transduction of C C myotubes with a HERG1A-encoded adenovirus results in elevated HERG1A protein. a Immunoblot of equal protein 2 12 content (50 μg) from lysates of non-transduced cells reveals that native ERG1 protein is 40.7% (p < 0.01; n = 6; Student’s t test) more abundant in myotubes than in myoblasts. Coomassie stained membrane confirms that equal amounts of cell lysate protein were loaded into each lane. b Immunohistochemistry labeling ERG1 protein with Alexfluor 488 (green) secondary antibody confirms that native ERG1 protein is more abundant in myotubes than in myoblasts. Representative images of immune-stained cells: (1) myoblasts immunostained with ERG1 primary antibody; (2) myoblasts immunostained without ERG1 primary antibody as control; (3) myotubes immunostained with ERG1 primary antibody; (4) myotubes immunostained without ERG1 primary antibody as control. Scale bar = 50 μm. c Transduction of C C myotubes with a HERG1A-encoded 2 12 adenovirus results in synthesis of HERG1A protein as demonstrated by immunoblot (p < 0.05; n = 6; two-way ANOVA). Coomassie stained membrane (blue) reveals that equal amounts of cell lysate protein were loaded into each lane Whitmore et al. Skeletal Muscle (2020) 10:1 Page 5 of 15 Fig. 2 Transduction of myotubes with HERG1A-encoded adenovirus is a valid in vitro skeletal muscle atrophy model. a The area of myotubes treated with HERG1A-encoded adenovirus is a significant 26.4% smaller (p < 0.01; n = 3 experimental sets) than that of control myotubes at 48 h after transduction and a significant 19.3% smaller (p < 0.01; n = 3 experimental sets) at 72 h after transduction. Scale bar = 100 μm. Bars of the graph represent the mean myotube area (μm ) while the error bars represent the standard error of the mean. b Immunoblot shows that transduction of C C myotubes with a HERG1A-encoded adenovirus yields an early increase in MuRF1 E3 ligase protein abundance while it does 2 12 not increase abundance of ATROGIN1 protein. Immunoblots are representative of three experiments solubilization and ethanol precipitation. Contaminating measure of amplicon. Changes in gene expression were DNA was degraded via DNase (RQ1 RNase-Free DNase; determined using the Livak method to normalize the gene ProMega, Madison WI). The total RNA was then reverse of interest to the “housekeeping gene.” transcribed using a GOScript™ Reverse Transcription Sys- tem Kit (Promega) per manufacturer’s instructions. Quan- Tissue sections and immunohistochemistry titative PCR was then performed using PowerUp SYBR For Fig. 4, mouse Gastrocnemius muscles were embed- green master mix (Applied Biosystems, Foster City, CA) ded in OCT, cryo-sectioned (20 μm), and stained for β- and primers for the gene of interest along with primers galactosidase (lacZ) activity as described earlier [18]. for the 18S ribosomal subunit “housekeeping gene” Sections for immunohistochemistry were fixed in cold (Table 1). An Applied Biosystems 7300 real-time PCR sys- methanol at − 20 °C for 10 min. These were then rinsed tem was used to detect SYBR green fluorescence as a with PBS at room temperature (RT) and incubated in Whitmore et al. Skeletal Muscle (2020) 10:1 Page 6 of 15 Table 1 Sequences of primers used for quantitative PCR a a Primer name (mouse) Primer sequence 5′–3′ Size (bp) Tm (°C) GC (%) Amplicon size (bp) Merg1a forward cctcgacaccatcatccgca 20 59.6 55.0 145 Merg1a reverse aggaaatcgcaggtgcaggg 20 60.3 60.0 18S subunit forward cgccgctagaggtgaaattct 21 57.2 52.4 101 18S subunit reverse agaacgaaagtcggaggttc 20 57.0 52.4 Calpain 1 forward gctaccgtttgtctagcgtc 20 58.73 55.0 98 Calpain 1 reverse taactcctctgtcatcctctggt 23 59.99 47.83 Calpain 2 forward ttttgtgcggtgtttggtcc 20 59.83 50.0 107 Calpain 2 reverse aactcagccacgaagcaagg 20 60.89 55.0 Calpain 3 forward ttcacaggaggggtgacaga 20 60.11 55.0 122 Calpain 3 reverse ttcgtgccatcgtcaatggag 21 61.01 52.38 Calpastatin forward gccttggatgacctgataga 20 53.8 50.0 115 Calpastatin reverse gtgcctcaaggtaggtagaa 20 53.7 50.0 bp base pair 3% H O for 1 h. These were then rinsed thoroughly in and methods adapted from those published previously 2 2 PBS and incubated with blocking reagent I (10% normal [22]. Brightness values were recorded as integers ranging goat serum [NGS], 0.1% bovine serum albumin [BSA; from 0 (no signal) to 256 (white). The average brightness Sigma, St. Louis, MO], and 0.1% Tween-20 in PBS) for value (± standard error of the mean, SEM) for each section 1 h at RT. The slides were then incubated for one hour was determined and analyzed by two-way ANOVA using with the laminin antibody (2 μg/mL in blocking reagent the General Linear Model Procedure of SAS 9.4 (SAS In- II–5% NGS and 0.2% TritonX100 in PBS) or in blocking stitute Inc., Cary, NC). reagent II only as a control for primary antibody bind- ing. After a thorough rinsing with PBS, the slides were Plasmids incubated overnight in the erg1 antibody (1:10 in block- The mouse Erg1a (Merg1a) clone in pBK/CMV plasmid ing reagent 2) or in blocking reagent 2 alone on the con- [23] was a generous gift from Dr. Barry London (Cardio- trol sections. The next day, the sections were rinsed vascular Institute, University of Pittsburgh, PA). The thoroughly in PBS containing 0.1% Tween-20. All sec- phRL synthetic Renilla luciferase reporter vector was tions were then incubated for 1 h at RT in Alexafluor purchased from ProMega (Madison, WI). 568 goat anti-rat IgG (1:1000 in blocking reagent II) to bind the laminin primary antibody from rat. The slides Electro-transfer were then again rinsed with PBS and incubated for one Mouse anesthesia was induced with 4% isoflurane in a hour at RT in the goat anti-rabbit secondary antibody vented chamber and maintained by administration of from the Alexafluor 488 Tyramide Super Boost Kit (Invi- 2.5% isoflurane in oxygen using a properly ventilated trogen, Carlsbad, CA). The tyramide reaction was carried nose cone with anesthesia machine and scrubber. Once out per manufacturer’sinstructionsto identify ERG1 pro- the animals were well anesthetized, the hind limbs were tein with green fluorescence. Finally, the sections were shaved and the Gastrocnemius muscles were injected rinsed thoroughly with PBS and mounted with Fluoro- with expression plasmids in 50 μL sterile saline and then mount G with DAPI (EMS; Hatfield, PA). Two sections stimulated with 8 pulses at 200 V/cm for 20 ms at 1 Hz from each muscle mid-section were analyzed. with an ECM 830 ElectroSquare Porator (BTX; Haw- thorne, NY). This method has been shown to result in Imaging ERG1a protein synthesis in skeletal muscle [15, 18]. Images were acquired using a Leica DM4500 microscope with a Leica DFC 340FX camera. Acquisition parameters Animal study design were maintained identically across samples to allow for Study 1 comparison of immunofluorescence levels when these The Merg1a plasmid (30 μg) and a plasmid encoding comparisons were made. For assay of laminin protein Renilla reporter (5 μg) were injected into the left Gastro- fluorescence, two fields were imaged per slide (one slide cnemius muscles of mice (n = 40). An empty control per mouse) and the single point brightness was measured plasmid (30 μg) and the Renilla reporter plasmid (5 μg) for 50 random consecutive points within the sarcolemma were injected into the Gastrocnemius muscles of the of each complete fiber within each field using ImageJ [21] right legs. All legs were electro-transferred to improve Whitmore et al. Skeletal Muscle (2020) 10:1 Page 7 of 15 plasmid uptake and expression. Each day, at days 0–7, measured again 10 min later to ensure that the reaction five mice were humanely killed and the Gastrocnemius had reached an end point after the first 10 min. The data muscles were harvested and frozen immediately in liquid are reported in relative light units (RLU). nitrogen. These were then stored at − 80 °C. All muscles were later thawed, homogenized, and assayed for (1) Calpain assay protein content, (2) Renilla activity to determine trans- A Calpain-Glo Kit (ProMega; Madison, WI) was used to fection efficiency, and (3) calpain activity. determine calpain activity in both myotubes and mouse muscle. Study 2 The Gastrocnemius muscles of a second set of animals, Myotubes consisting of five animals per day for days 0–5and 7 Myotubes were terminally differentiated and then trans- (n = 35), were injected and electro-transferred as de- duced with either a HERG1A-encoded adeno-virus or scribed above. After the appropriate amount of time, the the same (but non-HERG1A-encoded) virus as control animals were humanely sacrificed, the muscles were har- (12 wells each). At 48 h post-transduction, wells were vested, and total RNA was extracted for rtPCR assay. washed with two changes of 37 °C PBS and then PBS (200 μL) containing 0.2% Triton X-100 and 200 nM Study 3 epoxomicin (BostonBiochem, Cambridge, MA, Cat. #I- The Merg1a plasmid (30 μg) and a plasmid encoding a 110) was added to permeabilize the cells and to inhibit β-galactosidase (LacZ) reporter (5 μg) were injected into the proteasome, respectively. Six wells per viral treat- the left Gastrocnemius muscles of mice (n = 5). An ap- ment (HERG1A or control) received the buffer described propriate empty control plasmid (30 μg) and the LacZ (i.e., native activity); however, six wells per viral treat- reporter plasmid (5 μg) were injected into the Gastrocne- ment received buffer supplemented with the calpain in- mius muscles of the right legs. All legs were electro- hibitor MDL28170 (50 μM). These were allowed to sit at transferred to improve plasmid uptake and expression. room temperature for 5 min to ensure the myotubes At day 5, the five mice were humanely killed and the were permeabilized and the inhibitors had taken effect. Gastrocnemius muscles were harvested and frozen im- Then 200 μL of Calpain-Glo reagent was added to all mediately in liquid nitrogen. These were then stored at wells, mixed gently, and allowed to sit at room − 80 °C. All muscles were later thawed and painstakingly temperature. After 15 min, a 200 μL aliquot of the liquid serially sectioned. Serial sections were then stained for was removed from each well and placed in a white- either lacZ or dually immunostained for MERG1 and walled 96-well plate and luminescence was read using a laminin proteins as described above. Synergy H1 Hybrid Reader (BioTek Instruments, Wi- nooski, VT). The remaining well contents were scraped Protein assay from the back of the plate, triturated using a syringe and The BCA D/C Protein Assay Reagents (BioRad; Carls 26 gauge needle, and then centrifuged (13,000×g; 3 min) Bad, CA) were used to assay both samples and standards to remove any solid material. The supernatant was (0, 0.25, 0.5, 1.0, 1.25, 1.5, 2.0 mg/mL bovine serum al- assayed for protein content using the BioRad DC Protein bumin in Passive Lysis Buffer [ProMega; Madison, WI]) Assay kit. The protein data were used to normalize the for protein content, using a Synergy H1 Hybrid Reader calpain RLU activity. (BioTek; Winooski, VT) to measure absorbance at 605 nm light wavelength. Sample absorbances were in- Mouse muscle samples terpolated against the standard curve to determine the The Gastrocnemius muscles were thawed, weighed, and protein concentration of each sample. homogenized in Passive Lysis Buffer (PLB; ProMega) at a concentration of 2.5 μL buffer/μg tissue. The sample Renilla activity homogenates were aliquoted and frozen at − 80 °C. Prior To control for differences in transfection efficiency in to assay, the homogenates were thawed and sample ali- the animal muscle, a plasmid encoding the Renilla lucif- quots (40 μL) and positive control (purified porcine cal- erase enzyme was electro-transferred into muscle along pain) were added to wells of 96-well plates with assay with the Merg1a plasmid (as described above). The buffer (40 μL) having either 2 mM calcium (to activate Renilla-Glo™ Luciferase Assay System (ProMega) was calcium dependent enzymes) or 2 mM calcium plus used, according to manufacturer’s instructions, to assay 50 mM MDL28170 (to inhibit calpain specifically while homogenates for Renilla enzyme activity. The reaction allowing other calcium activated enzymes to function). was allowed to proceed for the recommended 10 min Each 96-well plate was read with a Synergy H1 Hybrid and luminescence was measured using a Synergy H1 Hy- Reader (BioTek; Winooski, VT) and activity was mea- brid Reader (BioTek; Winooski, VT). Luminescence was sured in RLU. Calpain activity was determined by Whitmore et al. Skeletal Muscle (2020) 10:1 Page 8 of 15 subtracting the RLU of the wells treated with 2 mM cal- test)and by 19.3% at72hpost transfection (p <0.01; n =6; cium and MDL28170 from the RLU of the wells treated Student’s t test). Within each time point, the difference be- with 2 mM calcium only and normalizing this RLU to tween the HERG1A-treated and control myotubes is statis- the RLU assayed with the Renilla kit to control for dif- tically significant (p < 0.01); however, there is no significant ferences in transfection efficiencies. The result was then difference in size between the myotubes treated with normalized to protein content (RLU/mg protein). HERG1A-encoding virus at the two different time points (Fig. 2a). Also similarly to mouse skeletal muscle expressing Statistics Merg1a [23], myotubes transduced with HERG1A exhibit In general, statistics were done using either a simple Student increased levels of the UPP E3 ligase, MuRF1, but not the t test or an ANOVA (as indicated in results section and fig- E3 ligase ATROGIN1 (Fig. 2b). However, when we treated ure legends) and SAS (SAS Inc.; Carey, NC). Results were myoblasts with either the HERG-encoded or the control considered significant when p < 0.05 unless otherwise noted. virus and allowed them to differentiate, we found that the HERG-expressing samples did not haveasignificantlydif- Results ferent number of myotubes containing two or more nuclei Transduction of C C myotubes with a HERG1A-encoded than the cells treated with the control virus. That is, the fu- 2 12 adenovirus results in elevated HERG1A protein sion index (myosin-positive multi-nucleated cells:total Immunoblot of equal protein aliquots from both non- myosin-positive cells evaluated) was 33.5 ± 5.0% (mean ± virus treated C C myoblast and myotube lysates detects SEM) for the cells treated with the HERG-encoded virus 2 12 a40.7% (p <0.01; n =6; Student’s t test) greater abundance while it was 31.6 ± 2.3% for the control-treated myoblasts of the ERG1 protein in myotubes than in myoblasts (p <0.74; n =14; Student’s t test). Thedatademonstrate (Fig. 1a). Immunohistochemistry work also demonstrates that HERG1A treatment of myotubes results in atrophy that there is more ERG1 protein in the C C myotubes (i.e., reduced myotube area) as it does in mouse skeletal 2 12 than in the myoblasts, revealing a stronger signal in myo- muscle; however, it does not affect the degree to which the tubes that is dispersed over the surface of the cell, while in myoblasts fuse. We conclude that we have developed a myoblasts it reveals only a very faint fluorescent signal valid in vitro model of skeletal muscle atrophy. which appears to be mainly nuclear (Fig. 1b). We trans- fected myotubes with either virus-encoding HERG1A Transduction of myotubes with a HERG1A-encoded (and GFP) or with the same, but not HERG1A-encoded, adenovirus yields a basal increase in both intracellular virus as control. Immunoblot of the lysates shows that calcium levels and calpain activity C C myotubes transfected with virus encoding HERG1A We transduced C C myotubes with either a GFP- and 2 12 2 12 do synthesize the HERG1A protein, which appears as a HERG1A-encoded adenovirus or an appropriate control single band of higher mass (likely a result of differential GFP-only encoded adenovirus. At 48 h after viral treat- glycosylation) than the native mouse ERG1 and is absent ment, we used a fura-2 calcium indicator assay and ob- from the myotubes treated with the control virus (Fig. 1c; served a significant 51.7% increase (p < 0.0001; n =90 GFP p < 0.05; two-way ANVOA). Coomassie stained mem- and n = 87 HERG1A transduced wells; Student’s t test) in brane confirms that equal amounts of protein were loaded basal intracellular calcium levels in HERG1A transduced into each well of the gel for immunoblot. myotubes relative to control (Fig. 3a). This demonstrates that HERG1A must either increase calcium influx and/or Transduction of C C myotubes with a HERG1a-encoded intracellular calcium release and/or decrease intracellular 2 12 adenovirus results in decreased myotube area and calcium re-uptake. Because HERG1A transduction results increased MuRF1 E3 ligase abundance, but no change in in increased basal intracellular calcium levels, we investi- myoblast fusion index gated the downstream effects of this increase. Specifically, We transfected myotubes with either virus-encoding using a Calpain-Glo assay kit (ProMega), we measured the HERG1A (and GFP) or with the same, but not HERG1A- combined activity of the calpain 1 and 2 enzymes in myo- encoded, virus as control. Fluorescent imaging demon- tubes treated with either the control or the HERG1A- strates that both viral particles infect myotubes (Fig. 2a). encoded virus. Some myotubes from both viral treatments Further, when the average area (μm ) of fluorescent myo- were treated with either 50 μM MDL28170 to inhibit cal- tubes from both sets is determined at both 48 and 72 h pains or an equal volume of buffer vehicle. We observed after transfection, we discover that, similarly to mouse skel- that basically the same amount of enzyme activity (control etal muscle fibers electro-transferred with Merg1a plasmid myotubes = 160.8 ± 7.3 and HERG1A-expressing myotubes [23], the myotubes transfected with HERG1A are signifi- = 167.5 ± 5.34 RLU/mg protein; n = 24) was not blocked in cantly smaller than control myotubes. Specifically, the area each well treated with the MDL28170, indicating that a ra- of the HERG1A-expressing myotubes is decreased by ther high level of non-calpain activity was assayed. None- 26.4% at 48 h post transfection (p < 0.01; n =6; Student’s t theless, we find that in control cells, the calpain activity is Whitmore et al. Skeletal Muscle (2020) 10:1 Page 9 of 15 control plasmid (n = 68 mice). We then assayed total RNA extracted from the muscles for Merg1a expression (n =28) and the muscle homogenates for calpain activity (n = 40). Quantitative PCR reveals that the electro-transfer did pro- duce Merg1a expression which was significantly higher than day 0 at days 3–5(p <0.05; Student’s t test was used to compare each day to day 0; Fig. 4a). Merg1a expression also yielded an increase in calpain activity, increasing nearly 4-fold (over day 0) by day 3 and 7.5-fold by day 4 (p <0.05; Student’s t test was used to compare each day to day 0; Fig. 4b). It returns to day 0 control levels by day 5. Thus, we show that MERG1a overexpression increases calpain ac- tivity and thus protein degradation. It is possible that the increase in intracellular calcium could lead to myofiber de- generation. Thus, we electro-transferred left mouse Gastro- cnemius muscle with a Merg1a-encoded plasmid and a Lac-Z-encoded plasmid while expressing lacZ-encoded plasmid and a an appropriate control plasmid in the right Gastrocnemius muscle and performed studies to determine if over-production of this protein would bring about changes indicative of degeneration, specifically changes in the number of centrally located nuclei or in the abundance of basal laminin. Thus, we painstakingly stained muscle ser- ial sections for lacZ (Fig. 4c) as a marker for MERG1 and dually immunostained matching serial sections for both Fig. 3 Transduction of myotubes with HERG1A-encoded adenovirus MERG1 (green fluorescence, Fig. 4d) and laminin (red increases basal intracellular calcium levels and basal calpain activity. a Fura-2 dye experiments reveal that expression of HERG1A in C C fluorescence, not shown) and used a DAPI containing 2 12 myotubes yields a 51.9% increase (p < 0.0001; n = 90 GFP and n =87 immunomount to identify nuclei (Fig. 4d). There was no re- HERG1A transduced wells) in basal intracellular calcium levels sponse in sections not stained with primary antibody relative to myotubes transduced with a control virus. b Calpain assay (Fig. 4e). The lacZ stain (blue fibers in Fig. 4c) identifies reveals that transduction of C C myotubes with a HERG1A- 2 12 where the MERG1 overexpression occurs. We find no evi- encoded adenovirus increases combined native calpain 1 and 2 activity a significant 31.9% (p < 0.08; n = 24; two-way ANOVA) over dence of any changes in the number of centrally located control myotubes. All bars represent the mean while error bars nuclei (Fig. 4d) nor in the amount of laminin fluorescence represent the standard error of the mean (Fig. 4f) in the fibers overexpressing MERG1 in any of the five mice examined nor have we seen any evidence of these 22.1% of the total native activity while it is 38.5% of the total occurrences in any of our past studies. in HERG1A-treated cells, demonstrating an increase in cal- pain activity in the HERG1A-treated cells. Because a two- HERG1A expression in myotubes does not affect way ANVOA reveals there is no real difference in the level expression of calpains 1–3 or calpastatin although it does of MDL28170 inhibited activity, we can compare the differ- affect certain protein abundances ences in assayed native activity (control versus HERG1A Calpain activity will augment with increased intracellular treated) and find that there is a 31.9% increase (p < 0.08) in calcium; however, we cannot assume that the increased activity in the HERG1A-expressing myotubes over the con- calcium is the only explanation for the increased calpain trols (Fig. 3b). Although the 0.08 probability is greater than activity. Thus, we asked if expression and/or protein the generally accepted statistical significance level of 0.05, abundances of either calpains 1, 2, or 3 or calpastatin we believe that the difference is nonetheless real. were affected by HERG1A expression. We used quanti- tative real-time PCR to discover that HERG1A expres- Merg1a expression in mouse Gastrocnemius muscle sion does not produce a statistically significant change in increases calpain activity, but did not change the number calpain 1 mRNA levels for up to 84 h after viral treat- of centrally located nuclei or laminin abundance ment (Fig. 5a). As well, no change in gene expression To test the effect of Merg1a expression on calpain activity was detected for calpains 2 or 3 (data not shown). Fur- in animals, we electro-transferred the left Gastrocnemius ther, our results indicate that there is no significant muscle of mice with an expression plasmid encoding change in calpain 1 protein abundance (Fig. 5b; n =6; Merg1a and the right leg muscle with an appropriate Student’s t test). Calpain 2, when autolyzed and hence Whitmore et al. Skeletal Muscle (2020) 10:1 Page 10 of 15 Fig. 4 Expression of mouse erg1a in mouse Gastrocnemius muscle increases Merg1a transcription and native calpain activity, but does not increase the number of centrally located nuclei or the abundance of laminin protein. a Quantitative PCR shows that electro-transfer of an expression plasmid encoding mouse erg1a (Merg1a) into mouse skeletal muscle produces Merg1a expression which is significantly higher than day 0 at days 3–5(p < 0.05; n = 28). The enclosed circles of the line graph represent the mean while the error bars represent the standard error of the mean. b Merg1a transfection in mouse skeletal muscle increases calpain activity nearly 4-fold (over day 0) by day 3 and nearly 7.5-fold by day 4(p < 0.05; n = 40). It returns to day 0 control levels by day 5 post transfection. Bars represent the mean calpain activity while error bars represent the standard error of the mean. c Positive assay for the β-galactosidase reporter (as an indicator of electro-transfer of plasmid encoding the Merg1a gene) produces a blue color. d Immunostain for MERG1 (green) of a serial section matched to the section in c demonstrates that there is indeed a greater amount of MERG1 in the fibers colored blue in c. There were no greater number of centrally located nuclei in the green fibers of any sections (n = 5 mice). e Representative of sections immunostained without primary antibody. f Over-expression of Merg1a does not produce a change in laminin abundance (p = 0.3; n = 5). Bars represent the mean single point laminin fluorescence while error bars represent the standard error of the mean. All scale bars = 50 μm activated, appears as a doublet found at ~ 75 kD [24]. abundance declined by a statistically significant 31.7% Interestingly, our results show that there is a 40.7% de- (Fig. 7b; p < 0.05; n = 6; Student’s t test). Additionally, crease (p < 0.05; n = 6; Student’s t test) in total calpain 2 there is a decrease in two of the three noted calpain 3 protein abundance in response to 48 h of HERG1A autocatalytic products (25; Fig. 8): the 114 kD isoform is treatment (Fig. 6). Calpastatin expression was not signifi- down 29.6% and the 60 kD isoform is down 29.2%, al- cantly affected by the HERG1A channel for up to 84 h though the 30 kD isoform is not affected (p < 0.03; n =6; post-transduction (Fig. 7a); however, calpastatin protein Student’s t test within protein isomer). When the optical Whitmore et al. Skeletal Muscle (2020) 10:1 Page 11 of 15 Fig. 5 Neither calpain 1 expression nor protein abundance changes after transduction of myotubes with HERG1A-encoded adenovirus. a Quantitative PCR reveals that there is no change in expression of calpain 1 for up to 84 h after transduction (n = 15). bImmunoblot demonstrates that there is no significant change in calpain 1 protein abundance at 48 h after viral transduction (n = 6). Bars represent the mean and the error bars represent the standard error of the mean. Coomassie staining of the blotted membrane shows that equal amounts of protein were loaded into each well of the gel densities for all protein bands are summed, there is a either control or HERG1A-encoded adenovirus, we show 2+ total 21.0% decrease in calpain 3 protein abundance. that HERG1A expression also increases basal [Ca ]i and calpain activity. There are numerous potential sources of 2+ Discussion the calcium that contributes to the increased [Ca ]i. For The ERG1a voltage-gated K channel is responsible for example, it is possible that ERG1A is modulating Cav1.1 late phase repolarization of the cardiac action potential channels in the skeletal muscle sarcolemmal membrane, and was reported to be absent from skeletal muscle [23, resulting in an influx potentially from both the external 25]; however, the Pond and Hannon labs demonstrated milieu and internal stores. Further, because ERG1A is lo- that this protein is detectable in the atrophying skeletal cated in the t-tubules of cardiac tissue [17, 20], it is pos- muscle of mice and in very low abundance in healthy ro- sible that it is located in the t-tubules of skeletal muscle, dent muscle with careful use of protease inhibitors and where it could contribute to the release of calcium from concentration of solubilized membrane proteins [18]. Sub- internal stores by modulation of ryanodine receptors and/ sequent studies showed that ERG1a expression leads to an or IP3 receptors. Indeed, changes in regulation of sarco- increase in abundance of the MURF1 E3 ubiquitin ligase lemmal permeability could have severe consequences for protein and enhances ubiquitin proteasome proteolysis, a skeletal muscle tissue, potentially producing diseases such pathway known to contribute to skeletal muscle atrophy as muscular dystrophies and Niemann-Pick disease [26, [15, 18]. Here, using C C myotubes transduced with 27]. The source of the increased calcium is currently 2 12 Whitmore et al. Skeletal Muscle (2020) 10:1 Page 12 of 15 blocking calpain activity reduced the activation of cal- pain 1 gene expression and attenuated skeletal muscle atrophy [29]. Here, we report that there is no detectable change in calpain 1 protein abundance in myotubes transduced with HERG1A while surprisingly we detect a decrease in calpain 2 protein abundance. These data demonstrate that the increased calpain activity is not a result of increased enzyme protein abundance. We sug- gest that the decreased calpain 2 protein abundance could result from either decreased calpain 2 synthesis and/or increased calpain 2 protein degradation. Quanti- tative PCR data demonstrate that there is no significant change in transcription of calpain 1 or 2 genes for up to 84 h post transduction. Interestingly, we observe a de- crease in calpain 2 protein abundance without detecting a change in transcription of that gene. Thus, although mRNA production is not always directly correlated with protein abundance, we can speculate that the calpain 2 Fig. 6 Calpain 2 protein abundance decreases (p < 0.05; n =6) 48 h protein may be undergoing an increased level of degrad- after myotube transduction with HERG1A-encoded adenovirus. Bars ation. Indeed, these proteins may be undergoing autoly- represent the mean and error bars represent the standard error of sis or it is possible that ubiquitin proteasome proteolysis the mean. Coomassie staining of the blotted membrane confirms that equal amounts of protein were loaded into each well of the gel of calpain 2 is enhanced. Indeed, we have shown that in- creased ERG1 expression increases UPP activity. Calpastatin is a native calpain inhibitor which inhibits under investigation in our laboratories. However, because conventional calpains 1 and 2, but not calpain 3. Calpas- we find no change in the fusion index or an increase in ei- tatin requires calcium to bind calpains so that when the ther the number of centrally located nuclei or in the abun- calcium concentrations rise, calpain activity is increased, dance of laminin fluorescence in the fibers over- but so is calpastatin binding [13, 30]. Indeed, a decrease expressing Merg1a, we believe that our data suggest that in calpastatin protein would lower the inhibition of cal- the channel (which we find to be in very low abundance pains and allow for increased calpain-mediated proteoly- in muscle normally) is simply upregulating protein deg- sis. Certainly, the increased level of calpain activity radation in our myotubes. It is also possible that the low assayed in the mouse muscle homogenates, in which the 2+] levels of increased calcium are affecting signaling path- [Ca i is disrupted, suggests that something other than 2+ ways, but that remains to be investigated. [Ca ]i must contribute to enhanced calpain activity. The explanation for the increased calpain activity may Calpain 3 is a non-classical calpain which is detected 2+ seem obvious—the increase in [Ca ]i. However, we ec- mainly in skeletal muscle. It undergoes calcium- topically expressed mouse erg1a (Merg1a) in mouse mediated autolysis that has been reported to be en- Gastrocnemius muscle and homogenized the muscle, hanced by ATP at lower calcium concentrations [31, 2+ thereby disrupting the [Ca ]i pool and equalizing the 32]. Evidence has shown that the absence of calpain 3 calcium concentration throughout the sample. We then leads to a reduction in protein turnover and results in assayed for calpain activity and discovered that even in accumulation of damaged and/or misfolded proteins the homogenate it is still higher in the Merg1a-express- which can lead to cellular stress and eventual muscle 2+ ing tissue. This study is evidence that increased [Ca ]i pathology [33, 34]. Indeed, the absence or reduction of may not be the only factor that contributes to the this protein has been shown to lead to limb-girdle mus- ERG1A-induced increase in calpain activity. Other pos- cular dystrophy type 2A (LGMD2A) in humans [30–32, sible contributors include increased calpain 1 and/or 2 34–37]. Studies suggest that calpain 3 takes part in re- protein and/or decreased calpastatin protein. modeling of the sarcomere in response to cellular dam- Calpains 1 (μ-calpain) and 2 (m-calpain) are both clas- age such as atrophy [34, 36, 37]. Interestingly, studies sical calpains and are detected throughout the body, in- with calpain 3 knockout mice suggest that calpain 3 acts cluding skeletal muscle [28]. Indeed, calpain activity has upstream of the UPP, although it is uncertain if calpain been demonstrated to contribute to muscle atrophy [28]. 3 directly cleaves proteins to make them accessible for For example, Shenkman and colleagues inhibited calpain ubiquitination [34]. Thus, calpain 3 appears to be pro- activity in hind limb suspended mice by treatment with tective against muscle loss and its protein abundance the calpain inhibitor PD150606 and demonstrated that might be expected to be lower in an atrophic situation. Whitmore et al. Skeletal Muscle (2020) 10:1 Page 13 of 15 Fig. 7 Calpastatin expression does not change after transduction with HERG1A-encoded adenovirus although protein abundance decreases. a Quantitative PCR reveals that levels of calpastatin mRNA do not significantly change for up to 84 h after viral transduction with HERG1A encoded adenovirus. b, c. Immunoblot detects a significant 31.7% decrease in protein abundance (p < 0.05; n = 6) at 48 h after transduction. All bars represent the mean ± the standard error of the mean. Coomassie staining of the blotted membrane confirms that equal amounts of protein were loaded into each well of the gel Indeed, we report that calpain 3 protein abundance de- calpain activity. This is not surprising considering that creases in response to HERG1A expression. The de- calpastatin binding is also enhanced by intracellular cal- crease may be related to a decreased ability to remodel cium. Calpain 3 activity was not measured here; how- the sarcomere during/after atrophy; however, this possi- ever, the decline in calpain 3 protein is consistent with bility would require much additional investigation. an atrophic environment. Interestingly, classical calpain In summary, we show that HERG1A increases calpain activity has been shown to degrade sarcomeric anchor activity in myotubes, likely resulting from the increase in proteins (e.g., titin, nebulin) and this allows for release of 2+ [Ca ]i. We detect no increases in abundances of cal- contractile proteins (e.g., myosin and actin) into the pains 1 or 2 proteins which would otherwise contribute cytosol where they can be accessed and degraded by the to enhanced calpain activity. In fact, we report a decline UPP [30, 38]. Here, we show that HERG1A modulates in the abundance of calpain 2 protein. Thus, it would intracellular calcium and calpain activity. Because its 2+ appear that the increased [Ca ]i could be the main con- interaction with calcium and calpains is upstream of the tributor to the enhanced calpain activity; however, there UPP, and it also modulates UPP activity [18], we is a significant decline in calpastatin protein abundance hypothesize that ERG1A may indeed contribute to co- which likely also contributes to the measured increase in ordination of proteolytic systems which produce skeletal Whitmore et al. Skeletal Muscle (2020) 10:1 Page 14 of 15 Fig. 8 Calpain 3 protein abundance decreased 21.0% in response to transduction of myotubes with HERG1A-encoded adenovirus. Immunoblot shows that calpain 3 degraded into numerous fragments as expected, including three notable autocatalytic products: 114 kD (down 29.6%), 60 kD (down 29.2%), and 30 kD which was not affected. Bars represent the mean ± the standard error of the mean. Coomassie staining of the blotted membrane shows that equal amounts of protein were loaded into each well of the gel muscle atrophy, specifically calpain and UPP activities. fura-2 assays to determine intracellular calcium concentrations. LA cultured and transduced myotubes and then imaged myotubes and determined their Further study is needed to learn how ERG1A functions area. KB, SML, and MNH performed the electro-transfer on mice hind limbs. in skeletal muscle. Indeed, because of the role of the KB and SML performed the calpain assays on the electro-transferred muscles. ERG1A/ERG1B heteromultimeric channel in cardiac ac- AKU imaged myotubes and consulted on content and writing of manuscript. RW provided direction on calpain assays and consulted on content and writ- tion potential repolarization, ERG1A will likely never be ing of manuscript. JKD cultured, imaged, and evaluated myoblasts and myo- a target for pharmacological treatment of atrophy; how- tubes and consulted on content and writing of manuscript. WHW cloned the ever, continuing study of this protein may reveal other HERG1A construct into the viral cassette, provided guidance for primer devel- opment, and consulted on content of manuscript. GHH provided over all possible targets to combat atrophy. guidance to EP for measurement of calcium concentration, acted as co-PI on the grant which funded the bulk of this work, and consulted on content and Abbreviations writing of manuscript. ALP worked in the laboratory to produce some of the DMEM: Dulbecco’s modification of Eagle’s medium; ERG1A: Ether-a-gogo- data, provided over all guidance to the project, acted as co-PI on the grant related gene; FBS: Fetal bovine serum; HERG1A: Human ether-a-gogo-related which funded the bulk of this work, and consulted on content and writing gene; Merg1a: Mouse ether-a-gogo-related gene; RLU: Relative light units; of manuscript. All authors read and approved the final manuscript. UPP: Ubiquitin proteasome pathway Authors’ information Acknowledgements Not desired. Not applicable. Funding Authors’ contributions This work was funded in part by the Southern Illinois University to ALP and CW cultured and transduced myotubes, performed calpain assays, in part by the Department of Defense Office of the Congressionally Directed completed the PCR and immunoblotting work, and wrote the original draft Medical Research Programs in the form of a Discovery Award (PR170326) to of the manuscript. EP cultured and transduced myotubes and performed ALP and GHH. Whitmore et al. Skeletal Muscle (2020) 10:1 Page 15 of 15 Availability of data and materials 17. Jones EM, Roti Roti EC, Wang J, Delfosse SA, Robertson GA. Cardiac IKr The datasets used and/or analyzed during the current study are available channels minimally comprise hERG 1a and 1b subunits. J Biol Chem. 2004; from the corresponding author on reasonable request. 279:44690–4. 18. Wang X, Hockerman GH, Green HW 3rd, Babbs CF, Mohammad SI, Gerrard D, et al. Merg1a K+ channel induces skeletal muscle atrophy by activating Ethics approval and consent to participate the ubiquitin proteasome pathway. FASEB J. 2006;20:1531–3. All animal work and studies were approved the SIU IACUC. 19. Perez-Neut M, Shum A, Cuevas BD, Miller R, Gentile S. Stimulation of hERG1 channel activity promotes a calcium dependent degradation of cyclin E2, Consent for publication but not cyclin E1, in breast cancer cells. Oncotarget. 2015;6:1631–9. Not applicable. 20. Rasmussen HB, Moller M, Knaus H, Jensen BS, Olesen S, Jorgensen NK. Subcellular localization of the delayed rectifier K channels KCNQ1 and ERG1 Competing interests in the rat heart. Am J Physiol Heart Circ Physiol. 2003;286:H1300–H9. The authors declare that they have no competing interests. 21. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671–5. PMID 22930834 Author details 22. Cheatwood JL, Emerick AJ, Schwab ME, Kartje GL. Nogo-a expression after Anatomy Department, Southern Illinois University School of Medicine, focal ischemic stroke in the adult rat. Stroke. 2008;39:2091–8. https://doi.org/ Carbondale, IL 62902, USA. Medicinal Chemistry and Molecular 10.1161/STROKEAHA.107.507426. Pharmacology, Purdue University, West Lafayette, IN 47906, USA. Doisey 23. London B, Trudeau MC, Newton KP, Beyer AK, Copeland NG, Gilbert DJ, College of Health Sciences, Saint Louis University, St. Louis, MO 63103, USA. et al. Two isoforms of the mouse ether-a-go-go-related gene coassemble to School of Liberal Arts, Sciences and Education, Ivy Tech State college, form channels with properties similar to the rapidly activating component Lafayette, IN 47905, USA. Biochemistry Department, Southern Illinois of the cardiac delayed rectifier K+ current. Circ Res. 1997;81:870–8. University School of Medicine, Carbondale, IL 62902, USA. Gene Editing Core 24. Hongqui L, Thompson VF, Goll DE. Effects of autolysis on properties of μ- Facility, Purdue University, West Lafayette, IN 47906, USA. Southern Illinois and m-calpain. Biochim Biophys Acta. 1691;2004:91–103. https://doi.org/10. University, 1135 Lincoln Drive, Carbondale, IL 62902, USA. 1016/j.bbamcr.2003.12.006. 25. Curran M, Splawski I, Timothy KW, Vincent GM, Green ED, Keating MT. A Received: 4 June 2019 Accepted: 18 December 2019 molecular basis for cardiac arrhythmia: HERG mutations cause long QT syndrome. Cell. 1995;80:795–803. 26. Michailowsky V, Li H, Mittra B, Iyer SR, Mazala DAG, Corrotte M, Wang Y, References Chin ER, Lovering RM, Andrews NW. Defects in sarcolemma repair and 1. Frontera WR, Ochala J. Skeletal muscle: a brief review of structure and skeletal muscle function after injury in a mouse model of Niemann-pick function. Calcif Tissue Int. 2015;96:183–95. type a/B disease. Skelet Muscle. 2019;9:1–5. 2. Bodine SC, Baehr LM. Skeletal muscle atrophy and the E3 ubiquitin ligases 27. Gumerson JD, Michele DE. The dystrophin-glycoprotein complex ini the MuRF1 and MAFbx/atrogin-1. Am J Physiol Endocrinol Metab. 2014;307: prevention of muscle damage. J Biomed Biotechnol. 2011;210797:1–13. E469–84. 28. Huang J, Zhu X. The molecular mechanism of calpains action on skeletal 3. Bonaldo P, Sandri M. Cellular and molecular mechanisms of muscle atrophy. muscle atrophy. Physiol Res. 2016;65:547–60. Dis Model Mech. 2013;6:25–39. 29. Shenkman BS, Belova SP, Lomonosova YN, Kostrominova TY, Nemirovskaya 4. Baoge L, Van Den Steen E, Rimbaut S, Philips N, Witvrouw E, Almqvist KF, TL. Calpain-dependent regulation of the skeletal muscle atrophy following et al. Treatment of skeletal muscle injury: a review. ISRN Orthop. 2012;2012: unloading. Arch Biochem Biophys. 2015;584:36–41. 30. Sorimachi H, Ono Y. Regulation and physiological roles of the calpain 5. Lynch GS, Ryall JG. Role of beta-adrenoceptor signaling in skeletal muscle: system in muscular disorders. Cardiovasc Res. 2012;96:11–22. implications for muscle wasting and disease. Physiol Rev. 2008;88:729–67. 31. Murphy RM, Lamb GD. Endogenous calpain-3 activation is primarily 6. Derbre F, Ferrando B, Gomez-Cabrera MC, Sanchis-Gomar F, Martinez-Bello 2+ governed by small increases in resting cytoplasmic [Ca ] and is not VE, Olaso-Gonzalez G, et al. Inhibition of xanthine oxidase by allopurinol dependent on stretch. J Biol Chem. 2009;284:7811–9. prevents skeletal muscle atrophy: role of p38 MAPKinase and E3 ubiquitin 32. Murphy RM, Vissing K, Latchman H, Lamboley C, McKenna MJ, Overgaard K, ligases. PLoS One. 2012;7:e46668. Lamb GD. Activation of skeletal muscle calpain-3 by eccentric exercise in 7. Fareed MU, Evenson AR, Wei W, Menconi M, Poylin V, Petkova V, et al. humans does not result in its translocation to the nucleus or cytosol. J Appl Treatment of rats with calpain inhibitors prevents sepsis-induced muscle Physiol. 2011;111(5):1448–58. proteolysis independent of atrogin-1/MAFbx and MuRF1 expression. Am J 33. Duguez S, Bartoli M, Richard I. Calpain 3: a key regulator of the sarcomere? Physiol Regul Integr Comp Physiol. 2006;290:R1589–97. FEBS J. 2006;273:3427–36. 8. Guasconi V, Puri PL. Epigenetic drugs in the treatment of skeletal muscle 34. Kramerova I, Kudryashova E, Venkatraman G, Spencer MJ. Calpain 3 atrophy. Curr Opin Clin Nutr Metab Care. 2008;11:233–41. participates in sarcomere remodeling by acting upstream of the ubiquitin- 9. Han HQ, Mitch WE. Targeting the myostatin signaling pathway to treat proteasome pathway. Hum Mol Genet. 2005;14:2125–34. muscle wasting diseases. Curr Opin Support Palliat Care. 2011;5:334–41. 35. Meznaric M, Writzl K. Limb-girdle muscular dystrophies: different types and 10. Hemmati-Brivanlou A, Kelly O, Melton D. Follistatin, an antagonist of activin, diagnosis. In: Willem M, editor. Skeletal Muscle. London: Nova Science is expressed in the Spemann organizer and displays direct neuralizing Publishers, Inc.; 2013. p. 105–27. activity. Cell. 1994;77:283–95. 36. Murphy RM, Goodman CA, McKenna MJ, Bennie J, Leikis M, Lamb GD. Calpain- 11. Smith RC, Lin BK. Myostatin inhibitors as therapies for muscle wasting 3 is autolyzed and hence activated in human skeletal muscle 24 h following a associated with cancer and other disorders. Curr Opin Support Palliat Care. single bout of eccentric exercise. J Appl Physiol. 2007;103:926–31. 2013;7:352–60. 37. Ono Y, Ojima K, Shinkai-Ouchi F, Hata S, Sorimachi H. An eccentric calpain, 12. Ma L, Chu W, Chai J, Shen C, Li D, Wang X. ER stress and subsequent CAPN3/p94/calpain-3. Biochimie. 2016;122:169–87. activated calpain play a pivotal role in skeletal muscle wasting after severe 38. Campbell RL, Davies PL. Structure-function relationships in calpains. burn injury. PLoS One. 2017;12:e0186128. Biochem J. 2012;447:335–51. 13. Goll DE, Neti G, Mares SW, Thompson VF. Myofibrillar protein turnover: the proteasome and the calpains. J Anim Sci. 2008;86:E19–35. 14. Smith IJ, Lecker SH, Hasselgren PO. Calpain activity and muscle wasting in Publisher’sNote sepsis. Am J Physiol Endocrinol Metab. 2008;295:E762–71. Springer Nature remains neutral with regard to jurisdictional claims in 15. Pond AL, Nedele C, Wang WH, Wang X, Walther C, Jaeger C, et al. The published maps and institutional affiliations. mERG1a channel modulates skeletal muscle MuRF1, but not MAFbx, expression. Muscle Nerve. 2014;49:378–88. 16. Vandenberg JI, Perry MD, Perrin MJ, Mann SA, Ke Y, Hill AP. hERG K(+) channels: structure, function, and clinical significance. Physiol Rev. 2012;92: 1393–478.

Journal

Skeletal MuscleSpringer Journals

Published: Jan 16, 2020

There are no references for this article.