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