RESEARCH ARTICLE
Fast-to-Slow Transition of Skeletal Muscle
Contractile Function and Corresponding
Changes in Myosin Heavy and Light Chain
Formation in the R6/2 Mouse Model of
Huntington’s Disease
Tanja Hering
1,2
, Peter Braubach
1,3
, G. Bernhard Landwehrmeyer
2
, Katrin S. Lindenberg
2
,
Werner Melzer
1
*
1 Institute of Applied Physiology, Ulm University, Ulm, Germany, 2 Department of Neurology, Ulm
University, Ulm, Germany, 3 Institute of Pathology, Hannover Medical School, Hannover, Germany
*
werner.melzer@uni-ulm.de
Abstract
Huntington´s disease (HD) is a hereditary neurodegenerative disease resulting from an
expanded polyglutamine sequence (poly-Q) in the protein huntingtin (HTT). Various studies
report atrophy and metabolic pathology of skeletal muscle in HD and suggest as part of the
process a fast-to-slow fiber type transition that may be caused by the pathological changes
in central motor control or/and by mutant HTT in the muscle tissue itself. To investigate
muscle pathology in HD, we used R6/2 mice, a common animal model for a rapidly pro-
gressing variant of the disease expressing exon 1 of the mutant human gene. We investi-
gated alterations in the extensor digitorum longus (EDL), a typical fast-twitch muscle, and
the soleus (SOL), a slow-twitch muscle. We focussed on mechanographic measurements
of excised muscles using single and repetitive electrical stimulation and on the expression
of the various myosin isoforms (heavy and light chains) using dodecyl sulfate polyacryl-
amide gel electrophoresis (SDS-PAGE) of whole muscle and single fiber preparations. In
EDL of R6/2, the functional tests showed a left shift of the force-frequency relation and
decrease in specific force. Moreover, the estimated relative contribution of the fastest myo-
sin isoform MyHC IIb decreased, whereas the contribution of the slower MyHC IIx isoform
increased. An additional change occurred in the alkali MyLC forms showing a decrease in
3f and an increase in 1f level. In SOL, a shift from fast MyHC IIa to the slow isoform I was
detectable in male R6/2 mice only, and there was no evidence of isoform interconversion in
the MyLC pattern. These alterations point to a partial remodeling of the contractile appara-
tus of R6/2 mice towards a slower contractile phenotype, predominantly in fast glycolytic
fibers.
PLOS ONE | DOI:10.1371/journal.pone.0166106 November 7, 2016 1 / 18
a11111
OPEN ACCESS
Citation: Hering T, Braubach P, Landwehrmeyer
GB, Lindenberg KS, Melzer W (2016) Fast-to-Slow
Transition of Skeletal Muscle Contractile Function
and Corresponding Changes in Myosin Heavy and
Light Chain Formation in the R6/2 Mouse Model of
Huntington’s Disease. PLoS ONE 11(11):
e0166106. doi:10.1371/journal.pone.016610 6
Editor: Laszlo Csernoch, University of Debrecen,
HUNGARY
Received: June 30, 2016
Accepted: October 24, 2016
Published: November 7, 2016
Copyright: © 2016 Hering et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: The authors received no specific funding
for this work.
Competing Interests: The authors have declared
that no competing interests exist.
Introduction
Huntington’s disease (HD) is a genetic neuropathy associated with severe motoric dysfunction
[
1, 2]. HD is of autosomal dominant inheritance and results from an expanded cytosine-ade-
nine-guanine triplet repeat (CAG)
n
in exon 1 of the IT15 gene (HTT) leading to an elongated
polyglutamine (polyQ) stretch in the N-terminal region of the protein huntingtin (HTT). The
pathology in HD is thought to result mainly from a gain of function caused by toxic polygluta-
mine-containing fragments that are capable of forming protein aggregates [
3]. In the brain,
neurons of the striatum and cortex are predominantly affected [
4, 5]. The pathomechanism
likely involves a combination of effects, for instance on vesicular transport, endocytosis, the
ubiquitin-proteasome system and transcription [
6, 7]. A number of studies indicate mitochon-
drial dysfunction as an important contributor to the pathology [
8–14]. In addition several
results point to altered Ca
2+
homeostasis and excitotoxicity in affected neurons [15, 16].
HTT is ubiquitously expressed both in the central nervous system and in peripheral organs
[
17–20] leading to tissue-specific changes. The most obvious clinical alteration in peripheral
tissues in HD are muscle weakness and wasting [
21, 22], probably caused by both neuronal and
muscle-intrinsic mHTT-induced changes [
23–27] [28][29, 30].
Muscle as an extensive and easily accessible excitable tissue may, therefore, provide clues
regarding mechanisms of the disease development and has been considered as a potential bio-
marker for its progression and for monitoring therapeutic interventions [
31]. Among the animal
models, the R6/2 mouse used in the present investigation has been one of the most frequently
studied. It originated from an insertion of the N-terminal fragment (exon 1) of the mutant
human huntingtin gene (mHTT, with 144 CAG repeats) [32]. Widespread mHTT-positive inclu-
sion formation in several neuronal and non-neuronal tissues which include muscle have been
observed in R6/2 mice [
19, 24, 25, 27, 29]. Several studies have addressed skeletal muscle function
in these mice and identified higher sensitivity of mitochondria for Ca
2+
-induced permeability
transition [
33], alterations in synaptic transmission and in Na
+
, K
+
and Cl
-
channels [34, 35] and
compromised depolarization-triggered Ca
2+
release and removal activity and contraction [30,
36].
RNA microarray analysis of the pattern of gene expression in skeletal muscle of human HD
patients and R6/2 mice suggested a “progressive loss of fast glycolytic fibers and concomitant
gain in slow fibers” [
31]. The four main types of muscle fibers found in mouse skeletal muscle,
I, IIA, IIX (or IID) and IIB, are characterized by their expression of the four myosin heavy
chain (MyHC) isoforms I, IIa, IIx (or IId) and IIb, respectively. I and IIA are oxidative and IIX/
D and IIB glycolytic fibers with I representing the slowest and IIB the fastest contracting fiber
type [
37, 38]. Each myosin heavy chain is associated with two light chains (MyLC), termed
essential (or alkali) and regulatory (or phosphorylatable) light chain, respectively. Light chains
are present in mouse muscle in 5 different isoforms (1f, 2f, 3f, 1s and 2s). Here 1 and 3 indicate
essential and 2 regulatory isoforms and the letters indicate their predominant presence in fast
glycolytic (f) and slow oxidative (s) muscle fiber types, respectively. The hypothesis of a fast to
slow transition would imply that the expression pattern of myosin isoforms changes [
37, 39].
Mielcarek et al. [
30] report significant transcriptional upregulation of Myh7, the gene encoding
the slow isoform MyHCI, in several fast twitch muscles. In contrast, determining protein alter-
ations in R6/2 interosseus muscle (containing predominantly fast type IIA fibers), our group
found several alterations in the MyLC protein pattern but none in MyHC isoforms that would
have pointed to a significant fast to slow fiber type transition [36]. Therefore, our aim in this
study was to specifically investigate the fast glycolytic extensor digitorum longus muscle (EDL)
in comparison to the slow oxidative soleus (SOL) muscle to determine contractile properties
Skeletal Muscle and Huntington’s Disease
PLOS ONE | DOI:10.1371/journal.pone.0166106 November 7, 2016 2 / 18
and to quantify changes in muscle fiber type by determining relative alterations of the various
myosin isoforms using electrophoretic protein separation and immunostaining.
Materials and Methods
Ethics statement
All experimental procedures performed on mice were in accordance with German animal pro-
tection laws and conducted under the project licence (C/0.113) of the Institutional Animal
Care and Use Committee of Ulm University (Tierforschungszentrum, Universität Ulm and the
Regierungspräsidium Tübingen) that specifically approved this study.
Experimental animals
R6/2 mice [
32] (B6CBA-Tg (HDexon1)62Gpb/1J x C57BL6J/CBA/caF1; The Jackson Labora-
tory) with 190 ± 10 CAG repeats and wild-type (WT) littermates were kept at a 12 hour light-
dark cycle with unrestricted access to food and water. Animal handling was in agreement with
the regulations of the local animal welfare committee. Genotype and CAG length were deter-
mined from tail biopsies according to Mangiarini et al [32]. Mice, showing weight loss (on
average 20–25% compared to WT), muscle atrophy and clasping, were sacrificed at the age of
12 to 13 weeks by CO
2
application and rapid cervical dislocation.
Contraction measurements
Extensor digitorum longus (EDL) and soleus (SOL) muscles were dissected from the hind
limbs of mice and transferred to carbogen-bubbled Ringer’s solution [
40]. Muscles were verti-
cally mounted in a temperature-controlled test chamber and connected to a force transducer
(FT03, Grass Instruments, Quincy USA). After stretching the muscle to optimal length, single
or repetitive contractions were elicited with supra-maximal electric current pulses of 1 ms
duration passed through platinum electrodes positioned near the muscle and recorded using a
bridge amplifier and data acquisition system (Digidata 1200, Axon Instruments) controlled by
custom-made software. Cross sectional area was calculated from muscle diameter under the
assumption of a circular cross-section. In force-frequency experiments contractile forces were
normalized to the force observed during the 125 Hz (EDL) and 100 Hz (SOL) tetani.
Myosin heavy chain analysis
Muscle specimens were immersed in a solution consisting of 50% glycerol and 50% Krebs-
Ringer’s solution, then rapidly frozen in liquid nitrogen and stored at -80°C. For the following
MyHC determination always one muscle (SOL and EDL) was used, while the second SOL and
EDL muscle was used for MyLC determination. After re-thawing a protein extraction accord-
ing to Singh et al. [
41] was performed. Briefly, one SOL and EDL per mouse was subjected to a
protein extraction buffer containing (in mM) 300 NaCl, 100 NaH
2
PO
4
, 50 Na
2
HPO
4
, 10
Na
4
P
2
O
7
, 1 MgCl
2
, 10 EDTA, 1.4 2-mercapto-ethanol. They were cut into small pieces and vor-
texed for 10 s. A repetitive freezing (liquid nitrogen) and re-thawing procedure improved pro-
tein extraction and was followed by 10 min centrifugation at 10,000 rpm. For the MyHC
determination in single fibers, freshly dissected muscles were incubated under permanent
shaking for 45 minutes at 35°C in 3 ml Ringer’s solution containing 2 mg/ml collagenase. The
digestion was stopped with PBS (phosphate-buffered saline; 20012; Invitrogen GmbH; Darm-
stadt) containing 10% FCS. After thorough washing with Krebs-Ringer’s solution and transfer
to a depolarizing buffer (58 mM TES, 7.8 mM MgCl
2
, 50 mM EGTA, 1 mM KH
2
PO
4
, 6.2 mM
Na
2
ATP, pH 7.1) intact single fibers were randomly selected and used for protein extraction as
Skeletal Muscle and Huntington’s Disease
PLOS ONE | DOI:10.1371/journal.pone.0166106 November 7, 2016 3 / 18
described by Tikunov et al. [42]. Total protein in the supernatant was estimated using Brad-
ford’s method. A Mini-Gel system (Bio-Rad, Munich, Germany) was used for SDS-PAGE. The
gels (8% polyacrylamide) were loaded with 5 μg extracted protein from whole muscle or total
single fiber lysate [
38] and stained with Roti1-Blue (Carl Roth; Karlsruhe, Germany) and the
protein bands in the scanned gels were integrated using the software ImageJ (National Insti-
tutes of Health). Quantitative Western Blot analysis to identify the stained MyHC isoforms
was performed as described by Braubach et al. [
36].
Myosin light chain analysis
To avoid a contamination by other proteins of similar molecular weight, a myosin extraction
method [
43] was used when studying MyLC isoforms. For this, the second SOL and EDL of
each mouse was used. Separation of 25 μg myosin extraction was performed using 12% poly-
acrylamide gels.
Statistics
Statistical calculations were carried out using R 3.1 [
44] running under Ubuntu Linux 15.04
and Microsoft Windows 7. Data are presented as mean value ± SEM (n = number of values),
count data as relative frequencies and binomial proportion confidence intervals (95% CI) using
Wilson’s method. Group means were compared by Student’s two-sided t-test, distributions by
the chi-squared test. In figures significant differences are marked as follows:
: p < 0.05,
:
p < 0.01,
: p < 0.001.
Results
Contraction measurements
EDL and SOL differ strongly in their fiber type pattern. EDL is composed of mainly glycolytic
fibers, whereas SOL contains large amounts of fast (type IIa) and slow (type I) oxidative fibers [
45]
and exhibits a slower twitch contraction [40]. On average, isolated muscles of R6/2 mice showed a
reduced weight compared to their age-matched WT counterparts (6.67 vs. 12.1 mg for EDL and
6.27 vs. 8.00 mg for SOL) and smaller estimated cross-sectional area (2.25 vs. 2.84 mm
2
for EDL
and 2.49 vs. 2.72 mm
2
for SOL). To characterize the contractile properties, isometric force of the
muscles was recorded in response to single or repetitive electrical shocks of 1 ms duration. Similar
to Mielcarek and coworkers [
30], it was found that contractile force in R6/2 EDL was reduced.
Peak twitch and tetanic force of the EDL were significantly smaller both in absolute size and when
normalized for cross-sectional area (Twitch: 1.88 vs. 7.67 mN/mm
2
; Fig 1A); tetanus: 24.2 vs. 59.6
mN/mm
2
). Therefore, reduced force is not a mere consequence of smaller muscle size. In the SOL,
which produced lower specific force than EDL the changes were considerably smaller and not sig-
nificant (Twitch: 0.92 vs. 1.35 mN/mm
2
; tetanus: 39.4 vs. 41.4 mN/mm
2
).
To determine force-frequency relations, incomplete and fused tetani (lasting 350 ms) were
measured in EDL at frequencies of 25, 50, 75, 100 and 125 Hz. In SOL, frequencies of 10, 25,
50, 75, 90, and 100 Hz were applied.
Fig 1C shows mean maximum force as a function of stim-
ulation frequency normalized to the value at the highest frequency, respectively. The normal-
ized force-frequency relations of R6/2 and WT were almost identical in SOL whereas in EDL a
clear shift to the lower frequencies can be noticed for R6/2.
Consistent with these observations, kinetic parameters of single twitch responses showed mus-
cle type specific changes: A significant increase in the time to the peak of the twitch (t
peak
) and the
half-time of relaxation (t
1/2
) was observed in EDL (by 128% and 190%, respectively), but no corre-
sponding alteration could be noticed in SOL (
Fig 1B). Data of the contraction experiments are
Skeletal Muscle and Huntington’s Disease
PLOS ONE | DOI:10.1371/journal.pone.0166106 November 7, 2016 4 / 18
Fig 1. Lower force and slowed kinetics of contraction in R6/2. (A) Mean twitch force (left panel) and specific
force (i.e. normalized by cross sectional area; right panel) compared in EDL and SOL muscles of WT (n = 15 and
n = 15, respectively) and R6/2 (n = 10 and n = 10, respectively). (B) Comparison of half time of relaxation (t
1/2
, left
panel) and time to peak (t
peak
, right panel). (C) Comparison of force frequency relations. Error bars indicate SEM.
doi:10.1371/journal.pone.0166106.g001
Skeletal Muscle and Huntington’s Disease
PLOS ONE | DOI:10.1371/journal.pone.0166106 November 7, 2016 5 / 18