Skeletal muscle mitochondrial volume and myozenin-1 protein ... - PeerJ

Skeletal muscle mitochondrial volume and myozenin-1 protein differences exist between high versus low anabolic responders to resistance training Michael Roberts Corresp., 1 , Matthew Romero 1 , Christopher Mobley 1 , Petey Mumford 1 , Paul Roberson 1 , Cody Haun 1 , Christopher Vann 1 , Shelby Osburn 1 , Hudson Holmes 1 , Rory Greer 1 , Christopher Lockwood 2 , Hailey Parry 1 , Andreas Kavazis 1 1 2

School of Kinesiology, Auburn University, Auburn, Alabama, United States Research, Lockwood, LLC, Draper, Utah, United States

Corresponding Author: Michael Roberts Email address: [email protected]

Background. We sought to examine how 12 weeks of resistance exercise training (RET) affected skeletal muscle myofibrillar and sarcoplasmic protein levels along with markers of mitochondrial physiology in high versus low anabolic responders. Methods. Untrained college-aged males were classified as anabolic responders in the top 25th percentile [HI; n=13, dual x-ray absorptiometry total body muscle mass change (Δ) =+3.1±0.3 kg, Δ vastus lateralis (VL) thickness =+0.59±0.05 cm, Δ muscle fiber CSA =+1426±253 μm2) and bottom 25th percentile (LO; n=12, +1.1±0.2 kg, +0.24±0.07 cm, +5±209 μm2; pLO, p=0.018, Cohen’s d=0.737) and time (PRE>POST, p=0.037, Cohen’s d=-0.589) were observed for citrate synthase activity, although no significant interaction existed. POST myofibrillar myozenin-1 protein levels were up-regulated in the LO cluster (+25%, p=0.025, Cohen’s d = 0.691). No interactions or main effects existed for other assayed markers. Our data suggest myofibrillar or sarcoplasmic protein concentrations do not differ between HI versus LO anabolic responders prior to or following a 12-week RET program. Discussion. Greater mitochondrial volume in HI responders may have facilitated greater anabolism, and myofibril myozenin-1 protein levels may represent a biomarker that differentiates anabolic responses to RET. However, mechanistic research validating these hypotheses is needed.

PeerJ Preprints | https://doi.org/10.7287/peerj.preprints.26933v1 | CC BY 4.0 Open Access | rec: 16 May 2018, publ: 16 May 2018

1 2

Skeletal muscle mitochondrial volume and myozenin-1 protein differences exist between high versus low anabolic responders to resistance training

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Michael D. Roberts1,2*, Matthew A. Romero1, Christopher B. Mobley1, Petey W. Mumford1, Paul A. Roberson1, Cody T. Haun1, Christopher G. Vann1, Shelby C. Osburn1, Hudson M. Holmes1, Rory A. Greer1, Christopher M. Lockwood3, Hailey A. Parry1, Andreas N. Kavazis1 Affiliations: 1School of Kinesiology, Auburn University, Auburn, AL USA; 2Department of Cell Biology and Physiology, Edward Via College of Osteopathic Medicine, Auburn, AL USA 3Lockwood, LLC, Draper, UT USA E-mail addresses: M.A.R.: [email protected] C.B.M.: [email protected] P.W.M.: [email protected] P.A.R.: [email protected] C.T.H: [email protected] C.G.V: [email protected] S.C.O.: [email protected] H.M.H.: [email protected] R.A.G.: [email protected] C.M.L.: [email protected] H.A.P.: [email protected] A.N.K.: [email protected] *Address

correspondence to: Michael D. Roberts, Ph.D. Associate Professor, School of Kinesiology, Auburn University Director, Molecular and Applied Sciences Laboratory 301 Wire Road, Office 286 Auburn, AL 36849 Phone: 334 - 844 -1925 Fax: 334 - 844 -1467 E-mail: [email protected] Short title: High and low responders to exercise


38

ABSTRACT

39

Background. We sought to examine how 12 weeks of resistance exercise training (RET)

40

affected skeletal muscle myofibrillar and sarcoplasmic protein levels along with markers of

41

mitochondrial physiology in high versus low anabolic responders. Methods. Untrained college-

42

aged males were classified as anabolic responders in the top 25th percentile [HI; n=13, dual x-ray

43

absorptiometry total body muscle mass change (Δ) =+3.1±0.3 kg, Δ vastus lateralis (VL)

44

thickness =+0.59±0.05 cm, Δ muscle fiber CSA =+1426±253 μm2) and bottom 25th percentile

45

(LO; n=12, +1.1±0.2 kg, +0.24±0.07 cm, +5±209 μm2; pLO, p=0.018, Cohen’s d=0.737) and time (PRE>POST, p=0.037, Cohen’s d=-0.589)

51

were observed for citrate synthase activity, although no significant interaction existed. POST

52

myofibrillar myozenin-1 protein levels were up-regulated in the LO cluster (+25%, p=0.025,

53

Cohen’s d = 0.691). No interactions or main effects existed for other assayed markers. Our data

54

suggest myofibrillar or sarcoplasmic protein concentrations do not differ between HI versus LO

55

anabolic responders prior to or following a 12-week RET program. Discussion. Greater

56

mitochondrial volume in HI responders may have facilitated greater anabolism, and myofibril

57

myozenin-1 protein levels may represent a biomarker that differentiates anabolic responses to

58

RET. However, mechanistic research validating these hypotheses is needed.

59 60

Keywords: MYOZ1, citrate synthase, muscle hypertrophy


62

INTRODUCTION

63

Numerous studies have reported resistance exercise increases both muscle protein synthesis

64

(MPS) and myofibrillar protein synthesis (MyoPS) rates several days following a single exercise

65

bout (Damas et al. 2016; Mitchell et al. 2012; Phillips et al. 1997; Phillips et al. 1999; Wilkinson

66

et al. 2008), and numerous studies have also reported weeks to months of resistance exercise

67

training (RET) increases muscle fiber cross sectional area (fCSA) (Mitchell et al. 2013; Mobley

68

et al. 2017; Petrella et al. 2008; Reidy et al. 2016; Staron et al. 1994). These parallel findings

69

have led to a general consensus that RET-induced increases in fCSA likely coincide with

70

increased myofibrillar protein content. The addition of myofibrils to pre-existing myofibrillar

71

structures involves proteins such as alpha-actinin 2 (ACTN2), myozenin 1 (MYOZ1), myotilin

72

(MYOT), and Sorbin And SH3 Domain Containing 2 (SORBS2) (Sanger et al. 2002). This

73

process has been observed in rapidly growing cardiomyocytes (LoRusso et al. 1997), skeletal

74

muscle myotubes (White et al. 2014), developing zebrafish (Sanger et al. 2009), and embryonic

75

chicken heart rudiments (Ehler et al. 1999). Given that resistance exercise acutely upregulates

76

MyoPS, it seems logical RET would upregulate these genes in order to increase myofibril protein

77

content in hypertrophying muscle fibers.

78

From a bioenergetics perspective RET-induced increases in MPS and MyoPS are

79

energetically costly given upwards of four ATP molecules are required per peptide bond

80

synthesized (Stouthamer 1973). Thus, increases in mitochondrial function or volume are likely

81

needed to sustain muscle growth during RET due to the increased energy demand required for

82

intracellular protein accretion. However, a recent review by Groennebaek and Vissing

83

(Groennebaek & Vissing 2017), which included 16 studies examining how chronic “high load”

84

RET affected markers of mitochondrial volume in whole-tissue lysates, cited 14 of these studies

85

observed no change or a decrease in these biomarkers. While this report suggests RET likely


86

does not increase markers of mitochondrial volume, it remains possible that high anabolic

87

responders to RET may experience greater increases in mitochondrial volume in order to better

88

support anabolism.

89

We recently published a study examining skeletal muscle biomarkers related to ribosome

90

biogenesis, inflammation, and androgen signaling that were differentially expressed between

91

high versus modest and low anabolic responders following a 12-week full body RET program

92

(Mobley et al. 2018); notably, vastus lateralis (VL) thickness changes was the clustering

93

variable. Herein, we adopted a refined approach similar to Davidsen et al. (Davidsen et al. 2011)

94

in order to define high versus low anabolic responders in these subjects based upon three

95

hypertrophic indices including total muscle fCSA, VL thickness, and total body skeletal muscle

96

mass (TBMM) assessed using dual x-ray absorptiometry (DEXA). Next, we sought to examine

97

if total myofibrillar and sarcoplasmic protein concentrations, myosin and actin protein content,

98

myofibrillar protein levels of genes involved with myofibril formation, or markers of

99

mitochondrial physiology differed between clusters (HI = anabolic responders in the top 25th

100

percentile and LO = anabolic responders in the bottom 25th percentile). We hypothesized HI

101

responders would exhibit greater changes in myofibrillar and sarcoplasmic protein

102

concentrations relative to LO responders following RET. Additionally, we hypothesized HI

103

responders would exhibit greater indices of mitochondrial volume or biogenesis relative to LO

104

responders prior to and/or following RET.

105 106

MATERIALS & METHODS

107

Ethical approval and study design


108

This study was approved by the Institutional Review Board at Auburn University and conformed

109

to the standards set by the latest revision of the Declaration of Helsinki. In the current study we

110

analyzed muscle specimens from select participants that participated in a study we previously

111

published (Mobley et al. 2017) and registered on ClinicalTrials.gov (Identifier: NCT03501628,

112

date registered: April 18, 2018). However, the current study is not a clinical trial per the

113

definition of the World Health Organization or National Institutes of Health given that health-

114

related biomedical outcomes were not assessed. Apparently healthy, untrained college-aged

115

male subjects provided written consent to participate in this study and performed a testing battery

116

prior to (PRE) and 72 hours after the last training bout (POST) following a 12-week full body

117

RET program. The testing battery consisted of a VL muscle biopsy, full-body dual DEXA scan,

118

and VL thickness assessment using ultrasound. More in-depth descriptions regarding the RET

119

protocol as well as assessments of body composition, VL thickness, and fCSA can be found in

120

past publications by our group (Mobley et al. 2018; Mobley et al. 2017).

121 122

Muscle tissue processing

123

Muscle biopsies from PRE and POST testing sessions were collected using a 5 gauge needle

124

under local anesthesia as previously described (Mobley et al. 2017). Immediately following

125

tissue procurement, ~20-40 mg of tissue was embedded in cryomolds containing optimal cutting

126

temperature (OCT) media (Tissue-Tek®, Sakura Finetek Inc; Torrence, CA, USA) for fCSA

127

assessment. The remaining tissue was teased of blood and connective tissue, wrapped in pre-

128

labelled foils, flash frozen in liquid nitrogen (LN2), and subsequently stored at -80ºC until

129

protein and citrate synthase activity analyses described below.

130


131

Western blotting of cell lysates

132

For whole tissue lysate protein analysis, ~30 mg tissue was powdered on a LN2-cooled ceramic

133

mortar and placed in 1.7 mL microcentrifuge tubes on ice containing 500 µL of general cell lysis

134

buffer [20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM Na2EDTA, 1 mM EGTA, 1% Triton;

135

Cell Signaling, Danvers, MA, USA] pre-stocked with protease and Tyr/Ser/Thr phosphatase

136

inhibitors (2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4, 1 µg/mL

137

leupeptin). Samples were then homogenized on ice by hand using tight micropestles, insoluble

138

proteins were removed with centrifugation at 500 g for 5 minutes, and obtained sample lysates

139

were stored at -80ºC prior to Western blotting.

140

Upon first thaw total protein content was determined in duplicate using a BCA Protein

141

Assay Kit (Thermo Scientific; Waltham, MA, USA). Lysates were immediately prepared

142

thereafter for Western blotting using 4x Laemmli buffer at 1 µg/µL. Following sample

143

preparation, 15 µL samples were loaded onto 4-15% SDS-polyacrylamide gels (Bio-Rad;

144

Hercules, CA, USA) and subjected to electrophoresis (180 V for 45-60 minutes) using pre-made

145

1x SDS-PAGE running buffer (Ameresco; Framingham, MA, USA). Proteins were subsequently

146

transferred (200 mA for 2 hours) to polyvinylidene difluoride membranes (PVDF) (Bio-Rad

147

Laboratories), Ponceau stained, and imaged to ensure equal protein loading between lanes.

148

Membranes were then blocked for 1 hour at room temperature with 5% nonfat milk powder in

149

Tris-buffered saline with 0.1% Tween-20 (TBST; Ameresco). Rabbit anti-human Peroxisome

150

Proliferator-Activated Receptor Gamma Coactivator 1-Alpha (PGC1-α, 1:1000; catalog #:

151

GTX37356; GeneTex; Irvine, CA, USA), and mouse anti-human total OXPHOS antibody

152

cocktail (1:250; catalog #:ab110413; Abcam; Cambridge, UK) were incubated with membranes

153

overnight at 4º C in TBST with 5% bovine serum albumin (BSA). The following day,


154

membranes were incubated with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG

155

(catalog #: 7074; Cell Signaling; Danvers, MA, USA) or HRP-conjugated anti-mouse IgG

156

(catalog #: 7072; Cell Signaling) in TBST with 5% BSA at room temperature for 1 hour

157

(secondary antibodies diluted 1:2000). Membrane development was performed using an

158

enhanced chemiluminescent reagent (Luminata Forte HRP substrate; EMD Millipore, Billerica,

159

MA, USA), and band densitometry was performed using a gel documentation system and

160

associated software (UVP; Upland, CA, USA). Densitometry values for each target were

161

divided by whole-lane Ponceau densities. Regarding data presentation, values for a given

162

protein target were normalized to the HI PRE group mean values whereby the HI PRE group

163

average was 1.00, and data were expressed as relative expression units (REUs) as reported in a

164

recent publication by our laboratory (Mobley et al. 2018).

165 166

Total myofibrillar and sarcoplasmic protein assessment

167

Myofibrillar and sarcoplasmic protein isolations were performed based on the methods of

168

Goldberg’s laboratory (Cohen et al. 2009). Briefly, frozen powdered muscle (8-13 mg) was

169

weighed using an analytical scale sensitive to 0.0001 g (Mettler-Toledo; Columbus, OH, USA),

170

and immediately placed in 1.7 mL polypropylene tubes containing 190 μL of ice cold

171

homogenizing buffer (20 mM Tris-HCl, pH 7.2, 5 mM EGTA, 100 mM KCl, 1% Triton-X100).

172

Samples were homogenized on ice using tight-fitting pestles, and centrifuged at 3000 g for 30

173

min at 4ºC. Supernatants (sarcoplasmic fraction) were collected, placed in 1.7 mL

174

polypropylene tubes, and stored at -80ºC until concentration determination. The resultant pellet

175

was resuspended in homogenizing buffer, and samples were centrifuged at 3000 g for 10 min at

176

4ºC. Resultant supernatants from this step were discarded, resultant pellets were resuspended in


177

190 μL of ice cold wash buffer (20 mM Tris-HCl, pH 7.2, 100 mM KCl, 1 mM DTT), and

178

samples were centrifuged at 3000 g for 10 min at 4ºC; this specific process was performed twice.

179

Final myofibril pellets were resuspended in 200 μL of ice cold storage buffer (20 mM Tris-HCl,

180

pH 7.2, 100 mM KCl, 20% glycerol, 1 mM DTT) and frozen at -80ºC until concentration

181

determination.

182

Sarcoplasmic protein concentrations were determined in triplicate using the microplate

183

BCA assay according to manufacturer’s instructions (20 μL of 5x diluted sample + 180 μL

184

Reagent A + B, absorbance reading at 580 nm) (Thermo Scientific) and normalized to input

185

muscle weights. The average coefficient of variation (CV) values for all triplicate readings were

186

1.6%. Myofibril protein concentrations were initially determined in triplicate using the

187

microplate BCA assay (Thermo Scientific). However, some wells (~10%) contained noticeable

188

myofibril aggregates yielding a relatively high average CV (9.2%). Hence, we adapted the BCA

189

protocol to a cuvette-based assay whereby a larger volume of myofibril resuspensions were

190

sampled (100 μL of 5x diluted sample + 900 μL Reagent A + B), and this visually yielded

191

uniform absorbances in all samples. Samples were run in duplicate (not triplicate) using this

192

method due to resource constraints, and the average CV proved to be lower for duplicate

193

readings relative to the microplate method (5.0%). Myofibrillar protein concentrations from the

194

cuvette-based assay were normalized to input muscle weights.

195

Proteins of select genes associated with new myofibril formation were assayed using

196

aforementioned Western blotting techniques, but the myofibril fraction was assayed rather than

197

the whole tissue lysate. For these assays, myofibril suspensions were prepared for and subjected

198

to SDS-PAGE, proteins were transferred to PVDF membranes, membranes were Ponceau

199

imaged, and membranes were blocked as described above. Rabbit anti-human ACTN2 (1:1000;


200

catalog #: GTX103219; GeneTex), rabbit anti-human SORBS2 (1:1000; catalog #: GTX81600;

201

GeneTex), rabbit anti-human MYOZ1 (1:1000; catalog #: GTX107334; GeneTex), and rabbit

202

anti-human MYOT (1:1000; catalog #: GTX109905; GeneTex) were incubated with membranes

203

overnight at 4º C in TBST with 5% BSA. Thereafter, secondary antibody incubations,

204

membrane development, and data procurement occurred similar to PGC1-α and OXPHOS

205

described above.

206 207

Determination of myosin heavy chain and actin content

208

SDS-PAGE preps from resuspended myofibrils were performed using: a) 10 μL resuspended

209

myofibrils, b) 65 μL distilled water (diH2O), and c) 25 μL 4x Laemmli buffer. Samples were

210

then loaded (15 μL) on pre-casted 4-15% SDS-polyacrylamide gels (Bio-Rad Laboratories) and

211

subjected to electrophoresis (200 V for 40 minutes) using pre-made 1x SDS-PAGE running

212

buffer (Ameresco). Following electrophoresis gels were rinsed in diH2O for 15 min, and

213

immersed in Coomassie stain (LabSafe GEL Blue; G-Biosciences; St. Louis, MO, USA) for 2 h.

214

Thereafter, gels were destained in diH2O for 60 min, bright field imaged using a gel

215

documentation system (UVP), and band densities were determined using associated software.

216

Myosin and actin content were expressed as arbitrary units (AU)/mg muscle.

217 218

Citrate synthase activity assays

219

Tissue lysates obtained through cell lysis buffer processing (described above) were batch

220

processed for citrate synthase activity as previously described by our laboratory (Kephart et al.

221

2015). This metric was used as a surrogate for mitochondrial content per the findings of Larsen

222

et al. (Larsen et al. 2012) suggesting citrate synthase activity highly correlates with transmission


223

electron micrograph (TEM) images of mitochondrial content (r=0.84, p0.500 and d≤0.800) or large (d>0.800).

255

Bivariate correlations were also performed on select variables, and significant correlations were

256

established at p≤0.050. Assumptions tests, ANOVAs, and correlations were performed using

257

SPSS v22.0 (IBM Corp; Armonk, NY, USA), and effect size calculations were performed using

258

Microsoft Excel v2013 (Redmond, WA, USA).

259 260

RESULTS

261

Cluster differences in anabolic indices and other variables prior to and following training

262

Figure 1 diagrams DEXA TBMM, VL thickness, and fCSA values between clusters prior to and

263

following training. HI responders (n=13) presented the following change scores (mean ±

264

standard error values): ΔTBMM=+3.1±0.3 kg, ΔVL thickness=+0.59±0.05 cm,

265

ΔfCSA=+1426±253 μm2. LO responders (n=12) presented the following change scores:

266

ΔTBMM=+1.1±0.2, ΔVL thickness=+0.24±0.07, ΔfCSA=+5±209 μm2. Notably, all Δ scores

267

were significantly different when comparing HI versus LO responders (p