MCB Accepts, published online ahead of print on 11 November 2013 Mol. Cell. Biol. doi:10.1128/MCB.01232-13 Copyright © 2013, American Society for Microbiology. All Rights Reserved.
1
Pknox1/Prep1 regulates mitochondrial oxidative
2
phosphorylation components in skeletal muscle
3
Timo Kanzleiter #1, Michaela Rath1, Dmitry Penkov2,3, Dmytro Puchkov4, Nadja Schulz1,
4
Francesco Blasi2, Annette Schürmann1
5
1: Department of Experimental Diabetology, German Institute of Human Nutrition, Potsdam-
6
Rehbruecke; 2: IFOM (Foundation FIRC Institute of Molecular Oncology), Milano; 3: Moscow
7
State University, Moscow, 4: Leibniz Institut für Molekulare Pharmakologie (FMP), Berlin
8
9
10
Running title: Prep1 regulates OXPHOS components #Address correspondence to Timo Kanzleiter, [email protected]
11
1
12
Abstract
13
The homeodomain transcription factor Prep1 was previously shown to regulate insulin
14
sensitivity. Our aim was to study the specific role of Prep1 for the regulation of energy
15
metabolism in skeletal muscle. Muscle specific ablation of Prep1 resulted in increased
16
expression of respiratory chain subunits. This finding was consistent with an increase in
17
mitochondrial enzyme activity without affecting mitochondrial volume fraction as assessed
18
by electron microscopy. Metabolic phenotyping revealed no differences in daily energy
19
expenditure or body composition. However, during treadmill exercise challenge Prep1
20
ablation resulted in a higher maximal oxidative capacity and better endurance. Elevated
21
PGC-1 expression was identified as a cause for increased mitochondrial capacity in Prep1
22
ablated mice. Prep1 stabilizes p160 Mybbp1a, a known inhibitor of PGC-1 activity. Thereby,
23
P160 protein levels were significantly lower in muscle of Prep1 ablated mice. By a ChIPseq
24
approach PREP1 binding sites in genes encoding mitochondrial components (e.g. Ndufs2)
25
were identified that might be responsible for elevated OXPHOS proteins in the muscle of
26
Prep1 nullmutants. These results suggest that Prep1 exhibits additional direct effects on
27
regulation of mitochondrial proteins. We therefore conclude that Prep1 is a regulator of
28
oxidative phosphorylation components via direct and indirect mechanisms.
29
Introduction
30
Obesity can give rise to a multitude of pathological conditions collectively referred to as the
31
metabolic syndrome. The underlying key metabolic defect is insulin resistance, which can be
32
caused by ectopic fat storage mainly in muscle and liver (1). Skeletal muscle is the major site
33
of oxidative glucose and lipid metabolism and dysregulation of either of these metabolic
2
34
pathways can contribute to the development of metabolic diseases such as type 2 diabetes
35
and cardiovascular complications (2).
36
The Pknox1 gene encodes the homeodomain transcription factor Prep1 that belongs to the
37
family of TALE (three-amino acid loop extension) proteins (3, 4). It dimerizes with Pbx
38
proteins to increase its target specificity (5–7). Prep1, as most homeodomain factors, is also
39
involved in the regulation of development and consequently deletion of Prep1 leads to
40
embryonic lethality. However, Prep1 hypomorphic mice (Prep1i/i), which express about 2% of
41
Prep1 mRNA have a survival rate of 25% whereas the remaining 75% suffers intra-uterine
42
death with developmental defects in hematopoesis, oculogenesis and angiogenesis (8).
43
Using the Prep1 hypomorphic mouse model, it was recently reported that Prep1 is involved
44
in the regulation of glucose metabolism (9). Prep1-hypomorphic mice were shown to be
45
more insulin sensitive than wildtype animals and it was concluded that this was due to
46
increased GLUT4-mediated glucose uptake in skeletal muscle (9). Prep1-hypomorphic have
47
other phenotypes relevant to systemic glucose metabolism such as changes in beta-cell
48
proliferation (9), enhanced hepatic insulin responsiveness and reduced hepatic glucose
49
output (10). Therefore we ablated Prep1 specifically in skeletal muscle in order to test the
50
hypothesis that Prep1 is involved in the regulation of energy metabolism in skeletal muscle.
51
Material and Methods
52
Animal studies
53
Animals were kept in a temperature-controlled room (22±1°C) on a 12 h light/dark cycle with
54
free access to food and water. All animal studies were conducted in accordance with the NIH
55
guidelines for the care and use of laboratory animals, and all experiments were approved by
3
56
the ethics committee of the State Agency of Environment, Health and Consumer Protection
57
(State of Brandenburg, Germany).
58
Generation of Prep1-hypomorphic and Prep1-loxP mice was described elsewhere (8, 11).
59
Briefly, Prep1-hypomorphic mice have a retroviral vector (VICTR45) inserted into intron1 and
60
Prep1-loxP mice have loxP sites flanking exon 6 and 7 of the Prep1 gene. Genotyping of the
61
mice was performed by genomic PCR on DNA isolated from tail biopsies (Fig1A; Primers A-E).
62
Primer sequences used were A: GGCACATCGTGAAGTTGGG, B: GCAGGTTAGAAAGGGAGGAC,
63
C: CCAAGGGCAGTAAGAGAAGCTCTGCAG, D: CAAAATGGCGTTACTTAAGCTAGCTTGCC, E:
64
GGAGTGCCAACCATGTTAAGAAGAAGTCCC . All three mouse lines (Prep1f/f, Prep1i/i and MCK-
65
Cre) were backcrossed to C57BL/6 mice for at least ten generations to minimize variation
66
due to differences in genetic background.
67
Body composition was weekly analyzed by NMR (nuclear magnetic resonance; Minispec
68
LF50, Bruker Biospin Corporation; Billerica, USA). Energy expenditure and respiratory
69
quotient was measured by indirect calorimetry as described elsewhere (12).
70
For assessment of glucose tolerance animals were fasted for 6h and then received
71
intraperitoneal injection of 2mg/g body weight glucose. For insulin tolerance tests (ITT) food
72
was withdrawn for 1 hour and then mice received 0.75mU/g body weight insulin ip
73
(Actrapid, Novo Nordisk, Mainz, Germany). Blood sugar was measured in blood sampled
74
from the tail 15, 30, 60 and 120 (glucose tolerance only) minutes after injection. For random
75
blood glucose measurements animals were measured in the morning without food
76
withdrawal.
77
The maximal oxidative capacity and endurance of the mice was measured on an airtight
78
metabolic treadmill (Columbus Instruments, Columbus, USA) connected to an Oxymax 4
79
oxygen and carbon dioxide analyzer (Columbus Instruments, Columbus, USA). The animals
80
were running on a 20° incline and the treadmill speed was increased by 2m/min every
81
2 minutes. The treadmill was stopped either when (a) the mice refused to run any further
82
and remained on the shock grid for more than 5 seconds, (b) if the oxygen consumption did
83
not increase any further after increase of the treadmill speed or (c) the respiratory quotient
84
rose rapidly over 1 indicating anaerobic conditions.
85
Mitochondrial properties
86
Mitochondrial density was assessed by qPCR amplification of genomic DNA (glucagon) and
87
mitochondrial DNA (cytochrome b) in isolated DNA from tibialis anterior muscle. The ratio of
88
relative abundance values between mtDNA and genomic DNA was used as an indicator of
89
mitochondrial density.
90
For measurement of citrase synthase activity frozen muscle samples were homogenized in
91
Tris-EDTA buffer (pH=7.4) and cleared supernatant after centrifugation was used for
92
spectrophotometric assay. The conversion rate of acetyl-CoA and oxaloacetate to citrate and
93
CoA-SH by citrate synthase is proportional to the coupled reaction of CoA-SH and DTNB
94
(5,5 -Dithiobis(2-nitrobenzoic acid); Sigma Aldrich) to NTB (2-nitro-5-thiobenzoate), which
95
was measured at 412 nm.
96
RNA and protein analysis
97
For cells and tissue samples RNA was isolated according to the manufacturer's instructions
98
using TRI-reagent (Sigma Aldrich). RNA integrity was measured on LabChips in an Agilent
99
Bioanalyzer and concentration was determined spectrophotometrically using a Nanodrop
100
spectrophotometer (Thermo Scientific). For conversion into cDNA random primers and the
101
cDNA synthesis kit from Qiagen were used. Quantitative real time PCR was performed on 5
102
Roche Lightcycler 480 using gene specific primers and corresponding UPL Probes (Roche)
103
(Table S1). For muscle samples the eukaryotic translation elongation factor 2 (Eef2) was used
104
as endogenous control. Transcriptional profilling of muscle samples were performed on
105
Agilent 4x44k whole genome microarrays by Source (Berlin). Array results were analyzed by
106
gene set enrichment using the MetaCore database (Thomson Reuters).
107
For immunoblot analysis frozen muscle samples were powdered and solubilized in
108
radioimmunoprecipitation (RIPA) buffer supplemented with protease and phosphatase
109
inhibitors (Roche). Protein concentration was measured in cleared supernatants using the
110
BCA assay (Pierce). Protein, 30 µg, was diluted in Laemmli SDS buffer and separated on an
111
SDS gel. After transfer to a polyvinylidene fluoride (PVDF) membrane, antibodies against
112
Prep1 (sc-6245; Santa Cruz), PGC-1
113
(Ambion), alpha-tubulin (Sigma Aldrich) and MitoProfile Total OXPHOS Rodent WB Antibody
114
Cocktail (MitoSciences) were used for the immunodetection of respective proteins.
115
Electron microscopy
116
Mice were transcardially perfused with 2% paraformaldfehyde/2,5% glutaraldehyde. Tibialis
117
anterior (TA) muscles were isolated and postfixed in the same fixative for 1 day. TA muscles
118
were washed and cut longitudinally to stripes of 0.5 mm in thickness. Following osmification
119
with 2% osmium tetroxide and 1% uranyl acetate en bloc staining tissue was routinely
120
dehydrated in methanol gradient solutions and embedded in Epoxy resin. After
121
polymerization median regions of the muscle were trimmed for subsequent ultramicroscopic
122
analysis. Images were taken at Technai transmission electron microscope at x3500
123
magnification using Gatan CCD camera. Multiple images were combined in order to
124
reconstruct an area of
(Cell signalling), p160 Mybp1 (Invitrogen), Gapdh
16 by 16 µm per each muscle fibre analysed. Images from 25 6
125
individual muscle fibers and 3 animals per genotype were analysed. Mitochondrial volume
126
fraction was estimated by superimposing a grid over an image of muscle fibre and getting a
127
ratio of crossings targeting mitochondria over total number of crossing targeting a muscle
128
fibre.
129
ChIP assay
130
ChIPs were performed using standard methods as described elsewhere (7) on differentiated
131
C2C12 myotubes. We used anti-Prep1 antibody for immunoprecipitation of sheared
132
chromatin fragments. Enrichment for Prep1 binding in immunoprecipitated chromatin was
133
tested by standard endpoint PCR (30cycles) as well as qPCR using SYBRgreen.
134
Microarray Dataset
135
Microarray data was deposited at Gene Expression Omnibus was assigned the accession
136
number XXXX.
137
Results
138
Generation of double heterozygous Prep1 MCK-Cre-mice
139
In a first approach, mice with loxP sites flanking exon 6 and 7 of the Prep1 gene (Prep1f/f;
140
Fig1A; described in (11)) were crossed with MCK-Cre mice (13). The Cre expressing offspring
141
had Prep1 expression in skeletal muscle at a level of about 50% as compared to the non-Cre
142
expressing mice (data not shown). In order to increase the efficiency of Prep1 deletion we
143
combined the Prep1 flox MCK-Cre mice with Prep1 hypomorphic mice (Prep1i/i; Fig1A (8))
144
and crossed Prep1i/+ with heterozygous MCK-Cre mice, and the resulting Prep1i/+ Cre/+ with
145
Prep1f/f mice. Double heterozygous Prep1f/i
146
Prep1
SM
Cre/+
offspring (from here on referred to as
) was used as a model for Prep1 ablation and compared to Prep1f/+ +/+ mice (referred 7
147
to as Prep1f/+; Fig1B). Prep1
148
of Prep1 mRNA expression in skeletal muscle as compared with Prep1f/+ control mice (Fig1C).
149
PREP1 protein levels were significantly reduced to about 60% of Prep1f/+ control mice
150
(Fig1D). Since the MCK promoter is also active in heart we studied the expression of Prep1 in
151
this organ and detected a 45% reduction in Prep1
152
Prep1 is a negative regulator of respiratory chain subunits
153
In order to identify target genes that are affected by Prep1 ablation transcriptional profiling
154
was performed. Therefore, isolated RNA from tibialis muscle of eight week old male
155
Prep1
156
differentially expressed genes (p
1
Pknox1/Prep1 regulates mitochondrial oxidative
2
phosphorylation components in skeletal muscle
3
Timo Kanzleiter #1, Michaela Rath1, Dmitry Penkov2,3, Dmytro Puchkov4, Nadja Schulz1,
4
Francesco Blasi2, Annette Schürmann1
5
1: Department of Experimental Diabetology, German Institute of Human Nutrition, Potsdam-
6
Rehbruecke; 2: IFOM (Foundation FIRC Institute of Molecular Oncology), Milano; 3: Moscow
7
State University, Moscow, 4: Leibniz Institut für Molekulare Pharmakologie (FMP), Berlin
8
9
10
Running title: Prep1 regulates OXPHOS components #Address correspondence to Timo Kanzleiter, [email protected]
11
1
12
Abstract
13
The homeodomain transcription factor Prep1 was previously shown to regulate insulin
14
sensitivity. Our aim was to study the specific role of Prep1 for the regulation of energy
15
metabolism in skeletal muscle. Muscle specific ablation of Prep1 resulted in increased
16
expression of respiratory chain subunits. This finding was consistent with an increase in
17
mitochondrial enzyme activity without affecting mitochondrial volume fraction as assessed
18
by electron microscopy. Metabolic phenotyping revealed no differences in daily energy
19
expenditure or body composition. However, during treadmill exercise challenge Prep1
20
ablation resulted in a higher maximal oxidative capacity and better endurance. Elevated
21
PGC-1 expression was identified as a cause for increased mitochondrial capacity in Prep1
22
ablated mice. Prep1 stabilizes p160 Mybbp1a, a known inhibitor of PGC-1 activity. Thereby,
23
P160 protein levels were significantly lower in muscle of Prep1 ablated mice. By a ChIPseq
24
approach PREP1 binding sites in genes encoding mitochondrial components (e.g. Ndufs2)
25
were identified that might be responsible for elevated OXPHOS proteins in the muscle of
26
Prep1 nullmutants. These results suggest that Prep1 exhibits additional direct effects on
27
regulation of mitochondrial proteins. We therefore conclude that Prep1 is a regulator of
28
oxidative phosphorylation components via direct and indirect mechanisms.
29
Introduction
30
Obesity can give rise to a multitude of pathological conditions collectively referred to as the
31
metabolic syndrome. The underlying key metabolic defect is insulin resistance, which can be
32
caused by ectopic fat storage mainly in muscle and liver (1). Skeletal muscle is the major site
33
of oxidative glucose and lipid metabolism and dysregulation of either of these metabolic
2
34
pathways can contribute to the development of metabolic diseases such as type 2 diabetes
35
and cardiovascular complications (2).
36
The Pknox1 gene encodes the homeodomain transcription factor Prep1 that belongs to the
37
family of TALE (three-amino acid loop extension) proteins (3, 4). It dimerizes with Pbx
38
proteins to increase its target specificity (5–7). Prep1, as most homeodomain factors, is also
39
involved in the regulation of development and consequently deletion of Prep1 leads to
40
embryonic lethality. However, Prep1 hypomorphic mice (Prep1i/i), which express about 2% of
41
Prep1 mRNA have a survival rate of 25% whereas the remaining 75% suffers intra-uterine
42
death with developmental defects in hematopoesis, oculogenesis and angiogenesis (8).
43
Using the Prep1 hypomorphic mouse model, it was recently reported that Prep1 is involved
44
in the regulation of glucose metabolism (9). Prep1-hypomorphic mice were shown to be
45
more insulin sensitive than wildtype animals and it was concluded that this was due to
46
increased GLUT4-mediated glucose uptake in skeletal muscle (9). Prep1-hypomorphic have
47
other phenotypes relevant to systemic glucose metabolism such as changes in beta-cell
48
proliferation (9), enhanced hepatic insulin responsiveness and reduced hepatic glucose
49
output (10). Therefore we ablated Prep1 specifically in skeletal muscle in order to test the
50
hypothesis that Prep1 is involved in the regulation of energy metabolism in skeletal muscle.
51
Material and Methods
52
Animal studies
53
Animals were kept in a temperature-controlled room (22±1°C) on a 12 h light/dark cycle with
54
free access to food and water. All animal studies were conducted in accordance with the NIH
55
guidelines for the care and use of laboratory animals, and all experiments were approved by
3
56
the ethics committee of the State Agency of Environment, Health and Consumer Protection
57
(State of Brandenburg, Germany).
58
Generation of Prep1-hypomorphic and Prep1-loxP mice was described elsewhere (8, 11).
59
Briefly, Prep1-hypomorphic mice have a retroviral vector (VICTR45) inserted into intron1 and
60
Prep1-loxP mice have loxP sites flanking exon 6 and 7 of the Prep1 gene. Genotyping of the
61
mice was performed by genomic PCR on DNA isolated from tail biopsies (Fig1A; Primers A-E).
62
Primer sequences used were A: GGCACATCGTGAAGTTGGG, B: GCAGGTTAGAAAGGGAGGAC,
63
C: CCAAGGGCAGTAAGAGAAGCTCTGCAG, D: CAAAATGGCGTTACTTAAGCTAGCTTGCC, E:
64
GGAGTGCCAACCATGTTAAGAAGAAGTCCC . All three mouse lines (Prep1f/f, Prep1i/i and MCK-
65
Cre) were backcrossed to C57BL/6 mice for at least ten generations to minimize variation
66
due to differences in genetic background.
67
Body composition was weekly analyzed by NMR (nuclear magnetic resonance; Minispec
68
LF50, Bruker Biospin Corporation; Billerica, USA). Energy expenditure and respiratory
69
quotient was measured by indirect calorimetry as described elsewhere (12).
70
For assessment of glucose tolerance animals were fasted for 6h and then received
71
intraperitoneal injection of 2mg/g body weight glucose. For insulin tolerance tests (ITT) food
72
was withdrawn for 1 hour and then mice received 0.75mU/g body weight insulin ip
73
(Actrapid, Novo Nordisk, Mainz, Germany). Blood sugar was measured in blood sampled
74
from the tail 15, 30, 60 and 120 (glucose tolerance only) minutes after injection. For random
75
blood glucose measurements animals were measured in the morning without food
76
withdrawal.
77
The maximal oxidative capacity and endurance of the mice was measured on an airtight
78
metabolic treadmill (Columbus Instruments, Columbus, USA) connected to an Oxymax 4
79
oxygen and carbon dioxide analyzer (Columbus Instruments, Columbus, USA). The animals
80
were running on a 20° incline and the treadmill speed was increased by 2m/min every
81
2 minutes. The treadmill was stopped either when (a) the mice refused to run any further
82
and remained on the shock grid for more than 5 seconds, (b) if the oxygen consumption did
83
not increase any further after increase of the treadmill speed or (c) the respiratory quotient
84
rose rapidly over 1 indicating anaerobic conditions.
85
Mitochondrial properties
86
Mitochondrial density was assessed by qPCR amplification of genomic DNA (glucagon) and
87
mitochondrial DNA (cytochrome b) in isolated DNA from tibialis anterior muscle. The ratio of
88
relative abundance values between mtDNA and genomic DNA was used as an indicator of
89
mitochondrial density.
90
For measurement of citrase synthase activity frozen muscle samples were homogenized in
91
Tris-EDTA buffer (pH=7.4) and cleared supernatant after centrifugation was used for
92
spectrophotometric assay. The conversion rate of acetyl-CoA and oxaloacetate to citrate and
93
CoA-SH by citrate synthase is proportional to the coupled reaction of CoA-SH and DTNB
94
(5,5 -Dithiobis(2-nitrobenzoic acid); Sigma Aldrich) to NTB (2-nitro-5-thiobenzoate), which
95
was measured at 412 nm.
96
RNA and protein analysis
97
For cells and tissue samples RNA was isolated according to the manufacturer's instructions
98
using TRI-reagent (Sigma Aldrich). RNA integrity was measured on LabChips in an Agilent
99
Bioanalyzer and concentration was determined spectrophotometrically using a Nanodrop
100
spectrophotometer (Thermo Scientific). For conversion into cDNA random primers and the
101
cDNA synthesis kit from Qiagen were used. Quantitative real time PCR was performed on 5
102
Roche Lightcycler 480 using gene specific primers and corresponding UPL Probes (Roche)
103
(Table S1). For muscle samples the eukaryotic translation elongation factor 2 (Eef2) was used
104
as endogenous control. Transcriptional profilling of muscle samples were performed on
105
Agilent 4x44k whole genome microarrays by Source (Berlin). Array results were analyzed by
106
gene set enrichment using the MetaCore database (Thomson Reuters).
107
For immunoblot analysis frozen muscle samples were powdered and solubilized in
108
radioimmunoprecipitation (RIPA) buffer supplemented with protease and phosphatase
109
inhibitors (Roche). Protein concentration was measured in cleared supernatants using the
110
BCA assay (Pierce). Protein, 30 µg, was diluted in Laemmli SDS buffer and separated on an
111
SDS gel. After transfer to a polyvinylidene fluoride (PVDF) membrane, antibodies against
112
Prep1 (sc-6245; Santa Cruz), PGC-1
113
(Ambion), alpha-tubulin (Sigma Aldrich) and MitoProfile Total OXPHOS Rodent WB Antibody
114
Cocktail (MitoSciences) were used for the immunodetection of respective proteins.
115
Electron microscopy
116
Mice were transcardially perfused with 2% paraformaldfehyde/2,5% glutaraldehyde. Tibialis
117
anterior (TA) muscles were isolated and postfixed in the same fixative for 1 day. TA muscles
118
were washed and cut longitudinally to stripes of 0.5 mm in thickness. Following osmification
119
with 2% osmium tetroxide and 1% uranyl acetate en bloc staining tissue was routinely
120
dehydrated in methanol gradient solutions and embedded in Epoxy resin. After
121
polymerization median regions of the muscle were trimmed for subsequent ultramicroscopic
122
analysis. Images were taken at Technai transmission electron microscope at x3500
123
magnification using Gatan CCD camera. Multiple images were combined in order to
124
reconstruct an area of
(Cell signalling), p160 Mybp1 (Invitrogen), Gapdh
16 by 16 µm per each muscle fibre analysed. Images from 25 6
125
individual muscle fibers and 3 animals per genotype were analysed. Mitochondrial volume
126
fraction was estimated by superimposing a grid over an image of muscle fibre and getting a
127
ratio of crossings targeting mitochondria over total number of crossing targeting a muscle
128
fibre.
129
ChIP assay
130
ChIPs were performed using standard methods as described elsewhere (7) on differentiated
131
C2C12 myotubes. We used anti-Prep1 antibody for immunoprecipitation of sheared
132
chromatin fragments. Enrichment for Prep1 binding in immunoprecipitated chromatin was
133
tested by standard endpoint PCR (30cycles) as well as qPCR using SYBRgreen.
134
Microarray Dataset
135
Microarray data was deposited at Gene Expression Omnibus was assigned the accession
136
number XXXX.
137
Results
138
Generation of double heterozygous Prep1 MCK-Cre-mice
139
In a first approach, mice with loxP sites flanking exon 6 and 7 of the Prep1 gene (Prep1f/f;
140
Fig1A; described in (11)) were crossed with MCK-Cre mice (13). The Cre expressing offspring
141
had Prep1 expression in skeletal muscle at a level of about 50% as compared to the non-Cre
142
expressing mice (data not shown). In order to increase the efficiency of Prep1 deletion we
143
combined the Prep1 flox MCK-Cre mice with Prep1 hypomorphic mice (Prep1i/i; Fig1A (8))
144
and crossed Prep1i/+ with heterozygous MCK-Cre mice, and the resulting Prep1i/+ Cre/+ with
145
Prep1f/f mice. Double heterozygous Prep1f/i
146
Prep1
SM
Cre/+
offspring (from here on referred to as
) was used as a model for Prep1 ablation and compared to Prep1f/+ +/+ mice (referred 7
147
to as Prep1f/+; Fig1B). Prep1
148
of Prep1 mRNA expression in skeletal muscle as compared with Prep1f/+ control mice (Fig1C).
149
PREP1 protein levels were significantly reduced to about 60% of Prep1f/+ control mice
150
(Fig1D). Since the MCK promoter is also active in heart we studied the expression of Prep1 in
151
this organ and detected a 45% reduction in Prep1
152
Prep1 is a negative regulator of respiratory chain subunits
153
In order to identify target genes that are affected by Prep1 ablation transcriptional profiling
154
was performed. Therefore, isolated RNA from tibialis muscle of eight week old male
155
Prep1
156
differentially expressed genes (p
