Novel γ-sarcoglycan interactors in murine muscle membranes

Affinity-purified rabbit polyclonal anti-γ-sarcoglycan and anti-archvillin

Rabbit polyclonal antibodies specific for extracellular residues 72–290 in murine γ-sarcoglycan (Sgcg, NCBI reference sequence NP_036022.1) and for murine archvillin (mAV, Swiss Protein Q8K463.1) amino acids 121–568 were generated against bacterially expressed proteins at Cocalico Biologicals, Inc. (Stevens, PA). Affinity purification employed columns containing the corresponding GST-tagged proteins and protocols detailed previously for high-avidity antibodies against human supervillin [35]. After a series of stringent washes, high-avidity antibodies were eluted from a column with covalently bound immunogen using 4.5 M MgCl2, 72.5 mM Tris-HCl pH 6.0–7.0. PCR templates were a murine Sgcg plasmid from Dr. Elizabeth McNally [36] and an EGFP-tagged mAV plasmid [27], respectively. PCR primers included the underlined sites for directed restriction cloning, as shown in Additional file 1, Supplementary Table S1. PCR products were generated with Pfu Turbo DNA polymerase (Agilent Technologies, Santa Clara, CA) according to the manufacturer’s directions, gel purified, cloned into TOPO-pCR2.1 vector (ThermoFisher Scientific, Waltham, MA), and verified by DNA sequencing. Coding sequences for Sgcg and mAV were recovered from doubly digested vectors and ligated into identically digested pGEX-6P-1 (Sigma-Aldrich, St. Louis, MO). Soluble GST-mAV was isolated from Rosetta 2(DE3)pLysS chemically competent bacteria (EMD-Millipore-Sigma, Burlington, MA) induced overnight with 0.2 mM isopropyl b-D-thiogalactopyranoside and purified on glutathione-Sepharose™ (Sigma-Aldrich). The ~ 26-kDa mAV immunogen was generated by cleavage with PreScission Protease (Sigma-Aldrich), dialyzed to remove residual glutathione, and purified by removing GST with a second glutathione-Sepharose column, followed by electrophoresis on a 15% acrylamide SDS-gel [37]. Because GST-Sgcg was incorporated into inclusion bodies and was poorly cleaved after urea renaturation, His-tagged Sgcg was generated by cloning the BamHI and EcoRI DNA fragment into doubly cut pET-30a vector (Sigma-Aldrich). Urea-solubilized His-Sgcg was purified on Ni-NTA agarose columns (Qiagen, Germantown, MD), as described [38], eluting with 50-100 mM histidine in a step gradient of 10–250 mM histidine in column buffer (7 M urea, 5 mM glycine, 50 mM Tris-HCl, pH 8.0, 0.3 M NaCl, 0.5 mM DTT, 1 mM PMSF). The ~ 29-kDa His-Sgcg remained soluble after dialysis against 1 mM DTT, PBS, pH 7.4, and was used as immunogen without further purification. Rabbits were boosted until high titers were observed on immunoblots, and each antibody was affinity purified with the corresponding GST-tagged protein covalently bound to CNBr-activated Sepharose (Sigma-Aldrich, C9142) [35]. After affinity purification, the rabbit anti-mAV antibody was used at 1:5000–1:10000 for immunoblots (IB), and the rabbit anti-Sgcg was used at 1:5000 for IB and at 1:200 for immunofluorescence microscopy (IF).

Commercial antibodies

Dilutions refer to those used for IB, IF, and immunoprecipitations (IP). Murine monoclonal anti-HA-Tag (6E2, #2367; IB: 1:1500, IF: 1:200) and anti-DYKDDDDK/Flag epitope (9A3, #8146; IB: 1:1000, IF: 1:200) antibodies were obtained from Cell Signaling Technologies (Beverly, MA). Murine monoclonal hybridoma supernatants were obtained from the Developmental Studies Hybridoma Bank (Iowa City, IA) to stain dystrophin (MANDRA1, 7A10, IB: 1:1000; MANDYS1, 3B7, IF 1:10) [39], α-sarcoglycan (IVD3(1)A9; IB 1:50), β-dystroglycan (MANDAG1 (7A11); IB 1:200), contractile myosins II (A4.1025; IB 1:100) [40, 41], troponin T (JLT12; IB 1:300) and SERCA1 (CaF2-5D2; IB 1:70). Rabbit monoclonal anti-HA (C29F4, #3724; IB: 1:1000, IP: 18 μl/60 μl of Dynabeads) and affinity-purified rabbit polyclonal anti-DYKDDDDK (#2368; IB: 1:1000, IP: 18 μl per 60 μl of Dynabeads) antibodies were from Cell Signaling Technologies. Rabbit monoclonal anti-PP1β was from Abcam (#ab53315, Cambridge, MA, USA; IB: 1:1000, IF: 1:100). Affinity-purified rabbit polyclonal antibodies against MYPT2 (#13366-1-AP; IB: 1:1000), NKCC1 (#13884-1-AP; IB: 1:1000, IF: 1:500) and Sgcg (#18102-1-AP; IB: 1:1000, IF: 1:100) were purchased from Proteintech Group Inc. (Rosemont, IL). Other rabbit polyclonal antibodies were anti-PP1β (#LS-C482256, Life Span BioSciences Inc., Seattle, WA; IB: 1:500), anti-NKCC1 (#ANT-071, Alomone Labs, Radassah Einkerem, Jerusalem, Israel; IB: 1:500, IF: 1:50), and anti-ERK1/2 (#ABS44, EMD Millipore Sigma; IB: 1:1000). Isotype-specific control antibodies were rabbit IgG (EMD-Millipore #12-370) and mouse IgG2a (#401502, BioLegend, San Diego, CA). For chemiluminescence detection in IB, we used horseradish peroxidase-conjugated donkey anti-mouse (#715-035-150) and donkey anti-rabbit (#711-035-152) secondary antibodies from Jackson ImmunoResearch (West Grove, PA) at 1:20,000 dilution. IB with near-infrared fluorescent detection used primary antibodies from Cell Signaling Technology against phospho-ERK1/2 (#9101; 1:2000) and total-ERK1/2 (#9102; 1:2000). Primary antibodies against NKCC1 were rabbit T-NKCC (1:2000, #13884-1-AP; Proteintech) and sheep polyclonal antibodies against P-NKCC1: phospho Thr 203, Thr 207, Thr 212 (IB: 1:1000, #S763B) and phospho Thr 212, Thr 217 (IB: 1:1000, #S603D), both from the MRC Protein Phosphorylation and Ubiquitylation Unit, University of Dundee, (United Kingdom). Secondary antibodies (#925-68071, anti-rabbit and #925-32214, anti-mouse, both 1:15,000) were from Li-Cor Biosciences (Lincoln, NE). IF signals from transfected RH30 cells were visualized with goat-anti-mouse Alexafluor 350 (#A 21049) and goat anti-rabbit Alexafluor 568 (#A11036) conjugated antibodies, and the actin cytoskeleton was stained with Alexafluor 350- or 488-conjugated phalloidin (all from ThermoFisher Scientific, diluted 1:500). Nuclei were stained using 4,6-diamidine-2-phenylindole dihydrochloride (DAPI, Sigma-Aldrich).

Animal models

All animals used in this study were bred, housed, and treated in accordance with standards set by the Animal Care and Use Committees at the University of Florida and the University of Massachusetts Medical School. Mice with a null mutation of the Sgcg gene were generated, bred, and phenotyped, as described previously [16]. Mice lacking the long muscle-specific archvillin (SV2) splice-form encoded by the Svil gene were generated by Ingenious Targeting Laboratory (Ronkonkoma, NY). The biochemical and functional consequences of this mutation are under investigation and will be described in detail elsewhere. In brief, the guide RNAs 5′-TGTAGGGCGATCCAAAGAAGAAG-3′ and 5′-TCTTCAATGCTTACCTGGCTCGG-3′ and CRISPR/Cas9 were used to generate a targeting vector in which the 609-nt exon 3 in the Svil gene, called exon 2 in the initial cDNA cloning paper [27], was replaced by an ivNeo selection cassette flanked by 759 bp and 838 bp of genomic sequence [42]. The targeting construct was verified by DNA sequencing, linearized with NotI, and electroporated into C57BL/6 embryonic stem (ES) cells. Targeted C57BL/6 FLP ES cells were microinjected into Balb/c blastocysts, and the resulting chimeras with a high percentage of black coat color were mated to C57BL/6 (WT) mice to generate germline Neo-deleted mice, which were identified by PCR screening of tail DNA. Neo-deleted mice were back-crossed four times with C57BL/6NCrl mice (Charles River Laboratories, Wilmington, MA) and screened by PCR for the absence of the FLP transgene and for proper integration at the Svil locus. Immunoblotting with rabbit polyclonal antibodies against archvillin N-terminal sequences confirmed the absence of full-length archvillin in skeletal muscle (Additional file 2, Fig. S1A).

Mouse muscle plasma membranes

Mouse gastrocnemius muscles were dissected and trimmed of extraneous tissue, cut into two ~ 100 mg pieces (wet tissue mass), and flash frozen in liquid nitrogen. To prepare for co-IP, one piece of frozen gastrocnemius muscle was very finely minced with a razor blade at room temperature on weighing paper while thawing, transferred to a 1.5-ml plastic mortar tube (#749520-0090, Kimble-Kontes, DWK Life Sciences, Millville, NJ), and processed according to a novel method adapted from those used to detect glucose transporter 4 [43], and other muscle membrane proteins [44,45,46,47,48,49] (Fig. 1). Minced muscles were covered with 900 μl of Buffer 1 (50 mM Tris, pH 8.0, 0.5 mM DTT, 0.1% NP-40, 10% glycerol, and protease inhibitor cocktail (Sigma-Aldrich #P8340) and lysed by manual grinding for 5–10 s with a blue plastic Kimble-Kontes pestle. Up to 4 samples were processed at one time, and all were stored on ice between steps. When all samples in a group were ground, they were immediately sonicated for 30 pulses with a microtip probe on a Branson Sonifier Cell Disrupter (Emerson Electric, Danbury, CT) with settings of 25% duty cycle, output ~ 1.5. A 100-μl portion of each sample was taken for gel analysis (Muscle Lysate, WCL). The remaining volumes were centrifuged at 250×g for 1 min at 4 °C. The cloudy supernatant was removed to a fresh 1.5-ml polypropylene microcentrifuge tube, and 450 μl of Buffer 1 was added to the remaining pellet, which was again ground manually for 5–10 s and sonicated. After another 1 min centrifugation at 250×g, this also-cloudy supernatant was combined with the first and centrifuged together at 4 °C for 10 min at 750×g. The remaining 250×g pellet was homogenized in an equivalent volume of 2x RIPA1 buffer (100 mM MOPS, pH 7.5, 0.3 M NaCl, 2% IGEPAL CA-630, 1% deoxycholic acid, 0.2% SDS, plus protease inhibitors) and incubated on ice while the other fractions were further processed; this incubation period was ≥ 25 min before sonication (see below). The supernatant from the 750×g spin was removed (Discard supernatant 1), and the remaining (white and somewhat fluffy) membrane pellet was resuspended by vortexing in 450 μl of Buffer 1. The membranes were again sedimented at 750×g for 10 min. The supernatant was removed (Discard supernatant 2), and the pelleted membranes were resuspended by briefly vortexing in 600 μl of cold Rx Buffer (25 mM MOPS, pH 7.5, 0.1 M NaCl, 0.5 M KCL, 10 mM MgCl2, 5 mM ATP (freshly made), 60 mM octylglucopyranoside, 0.5% IGEPAL CA-630, 10% glycerol, and protease inhibitors). The membranes were then sonicated three times, as above, with an incubation on ice of 1 min in between, until the suspension was slightly pearly. The pellets from the RIPA1-solubilized membranes also were sonicated once at this time, and both fractions were incubated on ice for 10 min after the last sonication. All tubes were centrifuged for 10 min at 10,000×g, and the supernatants removed to fresh tubes. The Rx membrane pellet was resuspended by briefly vortexing in 400 μl of Rx buffer and sonicated once before centrifugation at 10,000×g for 10 min. This final Rx pellet was resuspended in an equal volume of 2x Laemmli Sample Buffer [37]. The two Rx supernatants were combined and subjected to a final centrifugation at 10,000×g for 10 min. The RIPA1 pellet was resuspended in a Laemmli sample buffer volume equivalent to that of the corresponding RIPA supernatant.

Fig. 1figure 1

Flow chart for rapid membrane enrichment from mouse muscle. Skeletal muscle was fractionated as shown and described in detail in “Methods.” Lane numbers refer to analyses for total protein and immunoblotted sarcolemmal proteins, as shown in Fig. 2. Immunoprecipitations (IP, green ovals) with anti-Sgcg antibody were performed on extracts in either Rx buffer (high salt, 0.5% IGEPAL CA-630 in Rx buffer; lane 6) or with RIPA buffer (1% IGEPAL CA-630, 0.5% deoxycholate, 0.1% SDS; lane 8). WCL, whole cell lysate

Light and confocal microscopy

Phase images of extracted muscle membranes were obtained with a Leica DMI 6000B inverted fluorescence microscope with a Leica DFC 365 FX camera, a Leica HCX PL Fluotar 10x/0.30 PH1 lens, and Leica Application Suite 3.2.0.9625 software (Leica Microsystems, Exton, PA). Immunofluorescent images were taken on the same system using a Leica HC PL APO 63x/1.40-0.60 oil lens. Images of fluorescently stained muscle sections were obtained on a Leica SP5 (II) AOBS laser scanning confocal microscope with a HCX PL APO CS 40.0×/1.30 oil UV lens and using Leica Application Suite Advanced Fluorescence (LAS-AF) 2.7.3.9723 software (Leica Microsystems CMS GmbH, Mannheim, Germany). Optical z-sections of 0.29 μm were obtained sequentially for each color channel through the muscle sample. Selected sequential sections comprising no more than 1.2 μm of z-thickness were processed using the Maximum Intensity Projection function in the software. All images were exported as TIF files, then uniformly adjusted and assembled with Adobe Photoshop CS3 software (Adobe Systems, Inc., San Jose, CA).

Immunoblotting and analyses

Proteins were electrotransferred overnight onto Amersham™ Protran™ 0.45-μm nitrocellulose membranes (GE Healthcare Life Science, Marlborough, MA; #10600002). Chemiluminescent signals were visualized by SuperSignal West Pico or Femto reagents (ThermoFisher Scientific) on a Chemidoc MP Imaging System, using ImageLab™ software, version 4.1 (Bio-Rad Life Science Research, Hercules, CA). Densitometry of protein bands was determined using GelQuant.Net (version 1.7.8, BiochemLabSolutions.com; University of California, San Francisco, CA) software. Ratios of Unbound to Input signals were calculated using Microsoft Excel (Microsoft, Redmond, WA) and analyzed using GraphPad Prism 8.4.0 software (GraphPad Software, L.L.C., La Jolla, CA). Comparisons of multiple datasets were carried out using standard or non-parametric one-way ANOVAs, as indicated in the figure legends. Immunoblots were uniformly adjusted and assembled with Adobe Photoshop CS6 software.

For P/T-ERK1/2 and P/T-NKCC1 immunoblotting, extensor digitorum longus (EDL) muscles and tibialis anterior (TA) muscles were powdered using a mortar and pestle in dry ice and homogenized in RIPA2 buffer supplemented with PMSF and inhibitors of proteases (#P8340; Sigma-Aldrich) and phosphatases (#P5726; Sigma-Aldrich). Homogenates were incubated on ice for 60 min and centrifuged at 15,000×g for 15 min. Protein quantifications of muscle lysates were determined by the Bradford assay (#1863028; Thermo Fisher Scientific). For P/T-ERK1/2 analyses, EDL lysates (20 μg protein) were loaded in 4–12% Bis-Tris Midi Protein Gels (#WG1402A, Thermo Fisher Scientific) and electrotransferred to nitrocellulose. Each membrane was blocked for 90 min at room temperature with 5% BSA (BP9703100, Thermo Fisher Scientific) in TBS. Blots were incubated with anti-phospho-ERK1/2 and anti-total-ERK1/2 primary antibodies overnight at 4 °C, washed and incubated for 90 min at room temperature with secondary antibodies (Li-Cor Biosciences).

For P/T-NKCC1 analysis, TA lysates (20 μg protein) were loaded in 3–8% Bis-Tris Gel 1.0 mm (WG1002BOX; Thermo Fisher Scientific) and transferred to Immobilon-FL PVDF, 0.45-μm porosity (IPFL00010, EMD-Millipore) for 16 h at 40 mA in 10% methanol, 0.01% SDS, 0.1% NuPage Antioxidant (NP0005, Thermo Fisher Scientific). Each membrane was blocked for 120 min at room temperature with 5% BSA in TBS. Both P-NKCC antibodies were pre-incubated for 30 min with 10 μg/ml non-phospho peptide. After washing, blots were incubated for 90 min at room temperature with secondary antibodies (#925-68071, anti-rabbit; #925-32214, anti-goat/sheep; both at 1:15,000) from Li-Cor Biosciences. After washing, all blots were scanned with the Odyssey CLx Imaging System (Li-Cor Biosciences). The band intensity was automatically determined by the accompanying software Image Studio v.5.2 (Li-Cor Biosciences).

Co-IPs from muscle membranes

For each IP from muscle lysates, 15 μg of antibody was used per 50 μl of Dynabeads Protein A (ThermoFisher Scientific; #10001D). Protein A-bound IgG was cross-linked using 20 mM dimethylpimelimidate (DMP, Pierce/ThermoFisher Scientific), as previously described [35]. The Rx membrane lysate was diluted with an equal volume of Rx buffer without KCL or ATP (25 mM MOPS, pH 7.5, 0.1 M NaCl, 10 mM MgCl2, 60 mM octylglucopyranoside, 0.5% IGEPAL CA-630, 10% glycerol), and the RIPA1 extract was diluted to 750 μl using RIPA1 buffer. Both the diluted Rx and RIPA1 lysates were pre-cleared with 20 μl of Dynabeads Protein A (pre-rinsed in Rx or RIPA1 buffer) for 30 min at 4 °C. A 100-μl aliquot was removed as “Input” before proceeding with each IP. The lysates were divided evenly among the IP antibodies (200 μl each for RIPA1, 300 μl each for Rx), and incubated at 4 °C for 2 h with gentle rotation. Dynabeads were collected magnetically, and the Unbound supernatants were transferred to a fresh tube. The beads were rinsed 5 times with 500 μl of ice-cold 0.5 × TBST (5 mM Tris, pH 7.5, 83 mM NaCl, 0.05% Tween-20); beads were transferred to a fresh tube at the second rinse. Immunoprecipitated proteins were eluted under non-reducing conditions in Laemmli sample buffer lacking dithiothreitol (DTT) [37], and the eluate was transferred to a fresh tube.

In-gel protein digestion and LC-MS/MS analysis

IPs and analyses were performed using triplicate biological samples for each antibody and muscle genotype. Gastrocnemius muscles containing the SC were from C57BL/6 (WT) mice and a new mouse strain homozygous for the genetic ablation of a large 5′ coding exon in the Svil gene (Svil−/−). Negative controls were gastrocnemius muscles from Sgcg−/− mice [16] and IPs with nonspecific rabbit IgG. DTT (final concentration of 2 mM) was added to the immunoprecipitated samples before they were electrophoresed into a 10% Laemmli resolving gel [37], without a stacking gel, until the dye front was ~1 cm below the bottom of the loading wells; at least two blank wells separated samples. Gel proteins were then visualized with the Invitrogen Novex Colloidal Blue Staining Kit, according to the manufacturer's protocol (ThermoFisher Scientific; #LC6025). Each IP sample was excised as a single gel piece. Gel pieces were subjected to in-gel trypsin digestion after reduction with DTT and alkylation with iodoacetamide. Peptides eluted from the gel were lyophilized and re-suspended in 25 μl of 5% acetonitrile with 0.1% (v/v) trifluoroacetic acid (TFA). Using an injection volume of 3 μl, peptides were loaded by a Waters nanoAcquity UPLC (Waters Corp., Milford, MA) in 5% acetonitrile (0.1% formic acid (v/v)) at 4.0 μl/min for 4.0 min to a 100 μm I.D. fused-silica pre-column (Kasil frit) packed with 2 cm of 5 μm (200Å) Magic C18AQ (Bruker-Michrom, Billerica, MA). Peptides were then eluted at 300 nl/min via a 75-μm I.D. gravity-pulled analytical column packed with 25 cm of 3 μm (100Å) Magic C18AQ using a linear gradient from 5-35%B (mobile phase A, water + 0.1% (v/v) formic acid; mobile phase B, acetonitrile + 0.1% (v/v) formic acid) over 90 min. Ions were introduced by positive electrospray ionization via liquid junction electrode into a Thermo Scientific Q Exactive hybrid mass spectrometer. Mass spectra were acquired over m/z 300-1750 at 70,000 resolution (m/z 200) using an AGC (automatic gain control) target ion population of 1e6. Data-dependent acquisition selected the top 10 most abundant precursor ions for tandem mass spectrometry using higher-energy C-trap dissociation (HCD) using an isolation width of 1.6 Da, a normalized collision energy of 27, a maximum ion fill time of 110 ms, and an AGC target ion population of 1e5. Tandem mass spectra were acquired at 17,500 (m/z 200) resolution.

Proteomics data analysis

Raw data files were peak processed with Proteome Discoverer (version 1.4, Thermo Scientific) followed by identification using Mascot (version 2.5, Matrix Science, Boston, MA) against the SwissProt Mus musculus database (download 04/2018). Search parameters included full trypsin specificity with up to 2 missed cleavages, fixed modification of cysteine carbamidomethylation, and variable modifications of methionine oxidation, glutamine to pyroglutamic acid conversion, and protein N-terminal acetylation. Assignments were made using a 10 ppm mass tolerance for the precursor and 0.05 Da mass tolerance for the fragments. All non-filtered search results were processed by Scaffold software (version 4.4.4, Proteome Software, Inc., Portland. OR), utilizing the Trans-Proteomic Pipeline (Institute for Systems Biology, Seattle, WA). Rx extracts and RIPA1 extracts were analyzed separately. For both extracts, threshold values were set at 80% for peptides (false-discovery rates (FDR) of 0.45% for Rx extract and 0.53% for RIPA extracts) and 99% for proteins (2 peptide minimum; Rx extracts, FDR 20.0%; RIPA extracts, FDR 12.0%). Quantitative comparisons were made in the Scaffold software, using normalized weighted spectra, with an ANOVA significance level P < 0.05 without a post-test. Application of a Benjamini-Hochberg multiple test correction eliminated as significant all of the Rx buffer interactors and β-sarcoglycan and all proteins with higher P values in RIPA1 buffer, but the overall P values were unchanged. Interactors were listed in order of increasing P value from both Rx (Additional file 3, Table S2) and RIPA (Additional file 4, Table S3) extracts and edited manually for candidate Sgcg interactors, as described below.

Proteins specifically co-immunoprecipitating with Sgcg were defined as those that had significantly higher total normalized weighted spectral counts in IPs from C57BL/6 wild-type muscles and/or from Svil-targeted (Svil−/−) muscles, as compared to spectral counts from Sgcg−/− muscles and spectral counts in IPs with control IgG from any muscle type. Spectral counts were normalized based on each protein’s predicted molecular mass. “Top candidate interactors” were represented by ≥ 3 total normalized spectral counts from WT or Svil−/− muscles and were selected based on P values < 0.05 and quantitative profiles that showed increases of total normalized spectral counts that were ≥ 2-fold over those in each of the two types of negative controls. In most cases, total normalized spectral counts were greatly reduced in Sgcg−/− muscles and very few counts were observed with the IgG control antibody. “Other candidate interactors” also had P values < 0.05, but were below the 2-fold spectral count threshold or were elevated only in Svil−/− muscles. Two proteins (myotilin, titin) made this cut-off when the high background counts observed with control IgG were subtracted first. Because titin and myotilin mutations are responsible for LGMD, type 2J and type 1A, respectively [50], we drew the cut-off for statistical significance at P < 0.05 under these conditions. Exceptions were made for α- and β-sarcoglycans, known to be part of the SC [51], and for candidate PP1β-binding interactors close to the arbitrary 2-fold cut-off. A group of F-actin-binding proteins, e.g., spectrin, dystrophin, and filamin C, were recovered at approximately equal abundance from all three muscle types with anti-Sgcg, but not with control IgG, and were not considered further (Additional files 3 and 4; Tables S2 and S3).

This experimental approach may be inherently limited by binding of residual actin filaments to IgG [52], leading to high backgrounds of F-actin-binding proteins. Co-IPs also were performed with affinity-purified antibodies against the actin-binding protein archvillin (anti-mAV) and Rx and RIPA1 extracts of WT muscle and muscles deficient in archvillin. No archvillin-associated candidate interactors were identified because of high background binding with rabbit IgG.

Direct and indirect structural and signaling relationships among the candidate Sgcg interactors were identified using the Connect command in Ingenuity Pathway Analysis (IPA) Path Designer software, version 01-14 (Qiagen Bioinformatics, Redwood City, CA) Sgcg interactions with Svil [26] and NKCC1 were added manually and positioned with the IPA Auto-Layout command. The B-Raf, MEK1/2, and ERK1/2 signaling cascade was added manually, along with their connections to sarcolemmal proteins. Interactor colors and designations were manipulated for clarity using Adobe Photoshop CS6 (Adobe Systems, Inc.).

Plasmids

Elizabeth McNally (Northwestern University Feinberg School of Medicine, Chicago, IL) generously provided the plasmid encoding untagged murine Sgcg [36]. Myc-DDK (Flag)-tagged murine β- (#MR204617), δ- (#MR221060) and γ- (#MR223013) sarcoglycan cDNAs were from OriGene Technologies (Rockville, MD), as were Myc-DDK-tagged cDNAs encoding mouse Ppp1r12b/MYPT2 (#MR226968) and mouse tensin 2 (tensin like C1 domain-containing phosphatase, isoform 1/Tenc1; #MR211954). The plasmid encoding EGFP-PP1β/δ (EGFP-C1 vector) was obtained from Mathieu Bollen, (KU Leuven, Leuven, Belgium), via A. J. Baucum (Indiana University-Purdue University, Indianapolis, IN) [53]. Plasmid encoding human PP1β-EGFP (EGFP-N3 vector) was a gift from Angus Lamond and Laura Trinkle-Mulcahy (Addgene plasmid #44223; http://n2t.net/addgene:44223; RRID:Addgene_44223) [54]. EGFP-tensin2 (human) was a gift from David Critchley and Kenneth Yamada (Addgene plasmid #105298; http://n2t.net/addgene:105298; RRID:Addgene_105298) [55]. HA-tagged human NKCC1 (pdDNA3.1-HA-CFP-hNKCC1 WT (NT15-H)) was a gift from Biff Forbush (Addgene plasmid #49077; http://n2t.net/addgene:49077; RRID:Addgene_49077 [56]).

GST-bSV1398-1792 was described previously [57]. A pIRES2-EGFP vector containing full-length human Sgcg cloned between the EcoRI and BamHI sites [58] was used as a template to create plasmids encoding only the 35 N-terminal residues fused with C-terminal EGFP. The vector was first modified to contain an internal BamHI site by converting Leu-36 to glycine and Tyr-37 to serine using the QuikChange Site-Directed Mutagenesis kit (Agilent Technologies; #200519) and the L36Y37-BamHI-sense and L36Y37-BamHI-antisense primers in Additional file 1, Table S1. The modified vector was restriction digested with BamHI and NotI, and the similarly digested EGFP cassette from pEGFP-N3 was directionally ligated to be in-frame with the Sgcg cytoplasmic domain, i.e., residues 1–35. The Sgcg1-35-EGFP vector was then used as a template to generate hSgcg-1-35-EGFP fragments to clone into either the pET30a vector for His-tagging using the Sgcg-BglII-start-For and Sgcg-EcoRI-end-Rev primer pair or into the pMALc5x vector (New England BioLabs, Beverly, MA; #N8108S) for maltose-binding protein tagging using the Sgcg-For-EcoRV-1Met and Scgc-EcoRI-end-Rev primer pair. EGFP only controls were cloned into the same vectors using a BamHI-NotI double digest from pEGFPN3 into pET30a, and an XmnI-EcoRI double digest from pET30a-EGFP into pMALc5x. Vectors containing wild-type hSgcg-1-35-EGFP were then mutated at Tyr-6 to Ala using the QuikChange mutagenesis kit and the Scgc-Y6A-sense and antisense primer pair. All restriction enzymes were from New England BioLabs (Beverly, MA).

The plasmid encoding 3X-HA-tagged human NKCC1 synthetic cDNA with convenient restriction sites [56] was modified in two ways. First, we eliminated the sequence encoding all of the transmembrane domains and much of the cytoplasmic domains with a double restriction digest using HpaI and EcoRV. We then re-ligated the blunt ends to generate 3xHA-CFP-hNKCC1-cyto, which encodes the 3xHA-CFP tag, followed by the cytosolic NKCC1 N-terminal amino acids Glu-2 through Val-141 and the 91 residues of the cytosolic NKCC1 C-terminus (I-1118 through S-1196) [56]. Second, we generated a plasmid encoding a control 3XHA-CFP protein, which lacks the entire NKCC1 sequence, by inserting a TAA stop codon immediately downstream of the CFP-coding sequence using PCR with the QuikChange® II XL site-directed mutagenesis kit (Agilent Technologies; #200521) and primers HA-CFP-Stp For and HA-CFP-Stp Rev (Additional file 1, Table S1). All plasmids were verified by end sequencing.

Cell culture, screens with exogenously expressed proteins, and proximity ligation assay

Human RH30 rhabdomyosarcoma cells (SJC-RH30, American Type Culture Collection #CRL-2061, Manassas, VA) were maintained at 37 °C and 5% CO2 in RPMI-1640 medium modified to contain 2 mM L-glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 4500 mg/L glucose, 1.5 g/L sodium bicarbonate (#A1049101, Thermo Fisher Scientific), supplemented with 10% heat-inactivated fetal bovine serum, an additional 2 mM L-glutamine, and Pen-Strep. Transfections were done in 6-cm dishes that had been plated the day before so that the cells would be ≤ 60% confluent, using Lipofectamine 2000 (Thermo Fisher Scientific) and a scaled-down version of the manufacturer’s instructions. First, an equimolar mixture of C-terminally myc-DDK-tagged β-- δ-- and γ-sarcoglycan cDNAs was made in water at a total concentration of 0.3 μg/μl. For co-transfections in a 6-cm dish, with or without coverslips, 5 μl of this mix (1.5 μg) or 1.5 μg of an empty Flag vector control were then mixed with 1.5 μg of the other mammalian expression plasmid before dilution to 250 μl in OptiMEM reduced-serum medium (#31985, Thermo Fisher Scientific). Concurrently, 10 μl of Lipofectamine 2000 per transfection were diluted to 250 μl in OptiMEM. Both solutions were incubated for 5 min at room temperature and then combined (500 μl per transfection dish) and incubated at room temperature for an additional 10 min. The culture medium in each plate was replaced with 4.5 ml of fresh medium immediately before the plasmid DNA complexes in OptiMEM were added drop-wise across each plate. Plates were then incubated for 3 to 4 h at 37 °C and 5% CO2 before the medium was replaced with 5 ml of fresh growth medium and the incubation was continued overnight. Cells on coverslips in 6-well plates were cultured and transfected the same way, except using half the amounts of plasmid DNA and Lipofectamine 2000 in 100 μl of OptiMEM and a total volume of 2 ml medium per well. Cells were harvested 23 to 25 h post-transfection for either co-IP or fixation for immunofluorescence or proximity ligation assay (PLA). Fixation was performed as previously described [59], using either ice-cold methanol for 15 min or 4% paraformaldehyde in CSK buffer (10 mM PIPES, pH 6.8, 300 mM sucrose, 100 mM sodium chloride, 3 mM magnesium chloride, 1 mM EGTA [60]) for 30 min on ice. Pre-extraction experiments involved treating cells on coverslips with 0.1% Triton X-100 in CSK buffer for 4 min on ice and rinsing briefly with PBS before fixation and immunofluorescence imaging, as described above.

PLA was performed using the Duolink In Situ Orange kit (Sigma #DUO92102), according to the manufacturer’s instructions, with mouse anti-HA (Cell Signaling #2367, 1:200) and rabbit anti-Sgcg (Proteintech #18102-AP, 1:200) as primary antibodies. PLA speckles were counted by hand after signal inversion and ~3-fold enlargement using Adobe Photoshop. Nuclei associated with zero speckles were assumed to be untransfected cells. Graphical and statistical analyses were performed with GraphPad Prism.

Co-IPs and supernatant-depletion assays

Standard co-IP experiments and supernatant-depletion assays were carried out after extraction of RH30 cells with modified RIPA1 (no SDS). Cells between passage 5 and 28 in 6-cm dishes were transfected for these experiments as follows: (1) with a mixture of Myc-DDK (Flag)-tagged murine β-, δ-, γ-sarcoglycans and either HA-CFP, HA-CFP-Cyto, or HA-CFP-NKCC1; or (2) with Myc-Flag-tagged sarcoglycans, or an empty Flag vector, and either PP1β-EGFP, EGFP- PP1β or EGFP. Transfected cells were grown for 24 h and were 50–60% confluent when extracted on ice for 15 min in 350 μl of pre-chilled modified RIPA1 buffer (no SDS), plus protease and phosphatase inhibitors (# P8340, P2850, P5726, Sigma-Aldrich). Extracts were collected by scraping and transferred to 1.5-ml polypropylene centrifuge tubes, sonicated for 25 pulses with a Branson sonifier, as above, and centrifuged at 18,000×g for 10 min at 4 °C. The resulting supernatants were removed to a fresh 1.5-ml tube and used for either co-IPs or supernatant-depletion assays.

For experiments with HA-tagged proteins in RH30 cell lysates, Protein A Dynabeads were incubated for 1 h at room temperature either with monoclonal rabbit anti-HA antibody, clone C29F4, (for HA-CFP-NKCC1 experiments), or with rabbit polyclonal anti-Flag. Beads were rinsed once with RIPA1 buffer and split into equal fractions. The rinse was removed from each tube, and the beads were resuspended in 300 μl of each cell extract and incubated with rotation for 1 h at 4 °C. For experiments involving EGFP, 25 μl of GFP-Trap Dynabeads (product code gtd, Chromotek Inc., Islandia, NY, USA) were used per IP and washed once with ice-cold RIPA1 buffer before the addition of cell extract.

Co-IPs and supernatant-depletion assays were carried out similarly, except co-IPs focused on the amounts of proteins in pellets (“Bound”) and supernatant-depletion assays focused on the differences in protein concentration before (“Input”) and after (“Unbound”) centrifugation. The latter approach better detects low-avidity interactions [61]. In supernatant-depletion assays, either 20 μl of beads with bound anti-HA antibody or 25 μl of GFP-Trap Dynabeads were mixed with 300 μl of each cell extract for 1 h at 4 °C. A 100-μl “Input” sample was taken, and the remainder of each mixture was quickly sedimented through a cushion of 20% sucrose in RIPA1 buffer using narrow-bore BioRad tubes (#223-9502) and a swinging-bucket rotor at 800×g for 5 min. The top 100 μl was taken as the “Unbound” sample. Input and Unbound samples were resolved on 12% SDS-polyacrylamide gels and electroblotted.

Supernatant-depletion assays [61] also were carried out with purified, recombinant N-terminally tagged archvillin C-terminus and C-terminally tagged Sgcg cytoplasmic domain. The GST-bSV1398-1792 protein is respectively 94.43% and 96.96% identical to the corresponding sequences in murine and human archvillins (from Clustal2.1, Conway Institute UCD, Dublin, Ireland), and was made as described previously [57]. All recombinant EGFP proteins were made in Rosetta 2 pLyS bacteria (#71400, Novagen/Sigma-Aldrich), as described for antigen production. His-EGFP-tagged proteins were purified via NiNTA agarose (Qiagen), and MBP-tagged proteins were purified via amylose (New England BioLabs), according to the manufacturers’ instructions. We removed the tags from both sets of proteins (thrombin digestion for His tags, Factor Xa proteolysis for MBP), and re-purified them with NiNTA agarose or by ion-exchange chromatography, respectively. GST-bSV-1398-1792 was mixed with the EGFP proteins in a total volume of 250 μl and incubated for 30 min at 4 °C with rotation before the addition of 25 μl of glutathione-Sepharose beads (#17-0756-01, GE Healthcare). After an additional 30-min incubation, 100 μl of the total slurry was removed as the “Input” sample; the rest was loaded atop a 25% sucrose cushion in a narrow diameter tube and centrifuged, as described above. Input and Unbound samples were resolved on 12% SDS-polyacrylamide gels and electroblotted.

Muscle immunostaining

To determine the localization of NKCC1 in EDL muscles, 10-μm cryosections were obtained from EDL muscles of C57 and Sgcg−/− mice. Both longitudinal and cross sections were utilized. Sections were washed thrice in PBS, 10 min per wash. Sections were fixed for 5 min in 4% paraformaldehyde in PBS, followed by permeabilization with 0.5% Triton-X, PBS. Following blocking in 5% bovine serum albumin (BSA) in PBS for 1 h at room temperature, sections were incubated overnight at 4 °C with primary antibodies diluted in 5% BSA against dystrophin (MANDYS1, 3B7) [39] and NKCC1 (1:500 NKCC1, rabbit pAb 13884-1-AP; Proteintech) or with 5% BSA alone as a negative control. Sections were then washed thrice in PBS, 10 min/wash, and incubated 1 h at room temperature in the dark with secondary antibodies diluted in 5% BSA (1:10,000, Alexafluor 488 IgG anti-mouse (#A11029); 1:1000 Alexafluor 568 IgG anti-rabbit (#A11036) (Invitrogen). Sections were washed again in the dark (PBS 3 times, 10 min each), air-dried, and covered with a mounting agent (ProLongTM Diamond Antifade with Dapi, Cat#P36962, Thermo Fisher Scientific) and coverslip.

NKCC1 inhibition in vivo

EDL muscles from C57 or Sgcg−/− mice (9-14 wks old) were isolated and prepared for muscle mechanics as previously described [62]. EDLs with suture loops were mounted to the 800A in vitro Muscle Apparatus (Aurora Scientific, Ontario, CAN) via a rigid hook and hook to a servomotor arm. The muscles were positioned between two platinum plate electrodes, and in a Radnoti vessel filled with Ringer’s solution equilibrated with 95% O2/5% CO2 and maintained at 22 °C. EDL stimulation was controlled and carried out using Aurora Scientific (Ontario, CAN) hardware and software (i.e., High Power Bi-Phase Current Stimulator, 300C Dual-Motor Level System, Digital Controller Interface, and DMC v5.420). Following determination of optimal length (Lo), muscles were subjected to three bouts of supramaximal stimulation at 150 Hz for 500 ms, separated by 3 min to obtain maximum isometric force (Max Force). After completion of initial stimuli, EDLs were treated with bumetanide dissolved in 100% ethanol (final concentration 50 μM bumetanide, 1% ethanol) or ethanol vehicle alone (final concentration 1% ethanol) for 20–30 min. A second series of Max Force measurements were conducted to determine potential changes in twitch and tetanic forces after treatment. After a 5-min rest period, EDLs underwent five eccentric contractions, consisting of a 10% Lo lengthening during the final 200 ms of stimulation, with a 5-min rest interval between stimuli. A 30-min wait period followed muscle mechanics. EDLs were then lightly dried, weighed, snap frozen, and stored at − 80 °C. For those EDLs remaining at rest, they were pinned loosely in a petri dish with oxygenated Ringers solution containing either 50 μM bumetanide, 1% ethanol, or neither substance for 30 min. Similarly, non-mechanics EDLs were snap frozen afterwards and stored at − 80 °C for future biochemical and histological analyses.

P/T-ERK1/2 and P/T-NKCC1 activation assays

EDL and TA muscles were powdered using a mortar and pestle in dry ice and homogenized in RIPA2 buffer supplemented with PMSF and inhibitors of proteases (P8340; Sigma-Aldrich) and phosphatases (P5726; Sigma-Aldrich). Homogenates were incubated on ice for 60 min and centrifuged at 15,000×g for 15 min. Protein quantifications of muscle lysates were determined by the Bradford assay (#1863028; Thermo Fisher Scientific). For P/T-ERK1/2 analyses, EDL lysates (20 μg protein) were loaded in 4–12% Bis-Tris Midi Protein Gels (#WG1402A, Thermo Fisher Scientific) and electrotransferred to nitrocellulose. Each membrane was blocked for 90 min at room temperature with 5% BSA (BP9703100, Thermo Fisher Scientific) in TBS. Blots were incubated with anti-phospho-ERK1/2 and anti-total-ERK1/2 primary antibodies overnight at 4 °C, washed and incubated for 90 min at room temperature with secondary antibodies (Li-Cor Biosciences).

For P/T-NKCC1 assays, TA lysates (20 μg protein) were loaded in 3–8% Bis-Tris Gel 1.0 mm (WG1002BOX; Thermo Fisher Scientific) and transferred to Immobilon-FL PVDF, 0.45-μm porosity (IPFL00010, EMD-Millipore) for 16 h at 40 mA in 10% methanol, 0.01% SDS, 0.1% NuPage Antioxidant (NP0005, Thermo Fisher Scientific). Each membrane was blocked for 120 min at room temperature with 5% BSA in TBS. Both P-NKCC antibodies were pre-incubated for 30 min with 10 μg/ml non-phospho peptide. After washing, blots were incubated for 90 min at room temperature with secondary antibodies (#925-68071, anti-rabbit; #925-32214, anti-goat/sheep; both at 1:15,000) from Li-Cor Biosciences. After washing, all blots were scanned with the Odyssey CLx Imaging System (Li-Cor Biosciences). The band intensity was automatically determined by the accompanying software Image Studio v.5.2 (Li-Cor Biosciences).

留言 (0)

沒有登入
gif