Induced in vivo knockdown of the Brca1 gene in skeletal muscle results in skeletal muscle weakness

Abstract

Key points

• Breast cancer 1 early onset gene codes for the DNA repair enzyme, breast cancer type 1 susceptibility protein (BRCA1). The gene is prone to mutations that cause a loss of protein function.

• BRCA1/Brca1 has recently been found to regulate several cellular pathways beyond DNA repair and is expressed in skeletal muscle.

• Skeletal muscle specific knockout of Brca1 in mice caused a loss of muscle quality, identifiable by reductions in muscle force production and mitochondrial respiratory capacity.

• Loss of muscle quality was associated with a shift in muscle phenotype and an accumulation of mitochondrial DNA mutations.

• These results demonstrate that BRCA1 is necessary for skeletal muscle function and that increased mitochondrial DNA mutations may represent a potential underlying mechanism.

Abstract

Recent evidence suggests that the breast cancer 1 early onset gene (BRCA1) influences numerous peripheral tissues, including skeletal muscle. The present study aimed to determine whether induced‐loss of the breast cancer type 1 susceptibility protein (Brca1) alters skeletal muscle function. We induced genetic ablation of exon 11 in the Brca1 gene specifically in the skeletal muscle of adult mice to generate skeletal muscle‐specific Brca1 homozygote knockout (Brca1KO^smi) mice. Brca1KO^smi exhibited kyphosis and decreased maximal isometric force in limb muscles compared to age‐matched wild‐type mice. Brca1KO^smi skeletal muscle shifted toward an oxidative muscle fibre type and, in parallel, increased myofibre size and reduced capillary numbers. Unexpectedly, myofibre bundle mitochondrial respiration was reduced, whereas contraction‐induced lactate production was elevated in Brca1KO^smi muscle. Brca1KO^smi mice accumulated mitochondrial DNA mutations and exhibited an altered mitochondrial morphology characterized by distorted and enlarged mitochondria, and these were more susceptible to swelling. In summary, skeletal muscle‐specific loss of Brca1 leads to a myopathy and mitochondriopathy characterized by reductions in skeletal muscle quality and a consequent kyphosis. Given the substantial impact of BRCA1 mutations on cancer development risk in humans, a parallel loss of BRCA1 function in patient skeletal muscle cells would potentially result in implications for human health.

Key points

• Breast cancer 1 early onset gene codes for the DNA repair enzyme, breast cancer type 1 susceptibility protein (BRCA1). The gene is prone to mutations that cause a loss of protein function.

• BRCA1/Brca1 has recently been found to regulate several cellular pathways beyond DNA repair and is expressed in skeletal muscle.

• Skeletal muscle specific knockout of Brca1 in mice caused a loss of muscle quality, identifiable by reductions in muscle force production and mitochondrial respiratory capacity.

• Loss of muscle quality was associated with a shift in muscle phenotype and an accumulation of mitochondrial DNA mutations.

• These results demonstrate that BRCA1 is necessary for skeletal muscle function and that increased mitochondrial DNA mutations may represent a potential underlying mechanism.

Introduction

Breast cancer 1 early onset gene (BRCA1, human) codes for the large, multidomain breast cancer type 1 susceptibility protein (BRCA1) and is highly susceptible to genetic mutation in humans, with over 1600 mutations having been identified (Godet & Gilkes, 2017). Although some of these mutations are somatic (Hennessy et al. 2010; Bell et al. 2011), most BRCA1 mutations are germ line (Satagopan et al. 2001) and result in a non‐functional BRCA1 protein or complete loss of BRCA1 protein content as a result of protein instability (Yoshikawa et al. 1999; Godet & Gilkes, 2017). As a tumor suppressing protein, a central physiological function of BRCA1 is the repair of DNA (Wang et al. 2000; Liu & West, 2002; Kim et al. 2004; Yoshida & Miki, 2004; Saha et al. 2010; Roy et al. 2012); thus, the loss of BRCA1 function or protein content results in genomic instability and a greater risk of developing breast cancer and, to a lesser extent, ovarian cancer (Silver & Livingston, 2012). Historically, most studies have therefore focused on understanding the link between BRCA1 and cancer development in reproductive tissues. However, BRCA1 is expressed in all cell types and with varying levels of expression dependent on cell type (Anon, 2018). Recent studies have thus expanded the physiological role and necessity of BRCA1/Brca1 beyond DNA repair and reproductive tissues, describing a polyfunctional protein involved in the regulation of adipose, brain, smooth vascular muscle, cardiac muscle and skeletal muscle (Shukla et al. 2011; Ortega et al. 2012; Jackson et al. 2014; Lovren et al. 2014; Suberbielle et al. 2015).

BRCA1 is found on chromosome 17q21 and contains 24 exons, 22 of which code for the protein (Miki et al. 1994). Exon 11 is the largest and accounts for more than 60% of the 1863 amino acid polypeptide (Miki et al. 1994; Mehrgou & Akouchekian, 2016). At the nucleotide level, the murine form of Brca1 contains 72% sequence identity with the human form, with an amino acid similarity greater than 80% (Szabo et al. 1996). The RING domain of BRCA1 facilitates E3‐ubquitin ligase activity (Freemont, 2000; Ruffner et al. 2001), whereas the tandem BRCT (C‐terminus) domains allow for complex formation with phosphorylated proteins (Dulic et al. 2001; Manke et al. 2003; Yu et al. 2003; Williams et al. 2005). Together, these domains allow BRCA1 to regulate multiple distinct cellular functions (Venkitaraman, 2002; Deng, 2006). The regions of BRCA1 coding for exons 11–13 and the RING and tandem BRCT domains are not only the most functionally relevant, but also the most frequently mutated (Clark et al. 2012). These features of BRCA1/BRCA1 necessitate further exploration to fully understand the scope and scale of BRCA1‐regulated cellular processes in tissues throughout the body, as well as the potential consequences on systemic health in individuals with a loss of BRCA1 function or protein content.

As in the reproductive tissue, BRCA1 shares a pivotal role in DNA repair in non‐reproductive tissues. For example, Brca1 supports neural integrity and cognitive function in mice via protection of the neural genome, which experiences frequent double‐strand breaks (DSBs) following heightened neural activity (Suberbielle et al. 2015).

DSBs are similarly elevated in human atrial biopsies following ischaemic stress, and are associated with significantly increased BRCA1 expression (Shukla et al. 2011). Cardiomyocyte‐specific Brca1 knockout (KO) mice suggest that this upregulation is necessary for Brca1‐mediated DSB repair because loss of BRCA1 results in reduced DSB repair and impaired cardiac muscle remodelling, leading to increased mortality following cardiac ischaemia (Shukla et al. 2011). Brca1 has also been implicated in the protection of smooth muscle against oxidative stress (Lovren et al. 2014) and the attenuation of high‐fat diet‐induced aortic plaque lesions, as well as improved capillary density and blood flow following hind‐limb ischaemia in mice (Singh et al. 2013). The range of cellular functions influenced by BRCA1/Brca1 also extends to metabolic pathways, regulating lipid metabolism in human breast cancer cells (MCF7) (Magnard et al. 2002; Moreau et al. 2006) and lipid biosynthesis in adipose tissue via direct interactions between the BRCT domain and phosphorylated acetyl CoA carboxylase (p‐ACC) (Ortega et al. 2012). Despite the growing significance of BRCA1 in the regulation of numerous cellular functions across multiple tissues, as well as the involvement of skeletal muscle in the progression and development of a host of diseases and co‐morbidities, the literature concerning the role of BRCA1 within skeletal muscle remains sparse.

To date, our laboratory has shown that BRCA1 and Brca1 are expressed in human and murine skeletal muscle, respectively, and also that Brca1 localizes within skeletal muscle to the nucleus and mitochondria (Jackson et al. 2014, 2018). Furthermore, we have identified BRCA1/Brca1 as a regulator of skeletal muscle metabolism via BRCA1–p‐ACC interactions (Jackson et al. 2014) and shown that skeletal muscle‐specific loss of Brca1 protects mice against high‐fat diet‐induced perturbations in glucose and insulin tolerance (Jackson et al. 2018). The extent of BRCA1/Brca1 regulation on skeletal muscle function remains largely unknown. For example, the central function of skeletal muscle is locomotion, and loss of skeletal muscle strength and quality is associated with increased mortality (Metter et al. 2002), although the effect of Brca1 loss on skeletal muscle strength and quality has not been investigated. Additionally, although Brca1 is a known modulator of skeletal muscle metabolism and localizes to mitochondria, the relationship between Brca1 and mitochondrial function is not well understood. In the present study, we utilized an inducible skeletal muscle specific KO mouse model (Brca1KO^smi) coupled with an integrative experimental approach aiming to determine whether the loss of Brca1 expression is critical for skeletal muscle function in adult mice.

Methods

Ethical approval and animal experiments

The authors understand the ethical standards under which The Journal of Physiology operates and confirm that the animal procedures adhered to in the present study conform with these standards. All animal procedures and their usage were approved by the Institutional Review Committee at East Carolina University (Animal Usage Protocol #Q332a). Animal care complied with the Guide for the Care and Use of Laboratory Animals by the Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council (https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf). The present study utilized a total of 70 animals, 54 HSA‐mER‐Cre‐mER (+)‐Brca1(fl⁺/fl⁺) mice and 16 HSA‐mER‐Cre‐mER(+) Brca1(fl⁺/fl⁻) mice. Mice were bred on a C57BL/6NJ background, and were a mix of male and female mice aged 30–34 weeks (weighing 30–50 g) at the time of death. Mice were housed under a 12:12 h light/dark cycle at 22°C with access to food and water available ad libitum. Animals were killed via isoflurane‐induced anaesthesia, followed by cervical dislocation.

Animal generation and genotyping

We used the previously described HSA‐Cre(+)‐Brca1(fl/fl) mice (Jackson et al. 2018). All animals were genotyped based on the presence of HSA‐Cre, and hetero‐ and homozygote Brca1 flox expression using genomic DNA. Cre‐mediated recombination was confirmed via PCR analysis of genomic DNA using a DNA isolation kit (Qiagen, Valencia, CA, USA). Briefly, a short bout of tamoxifen delivery via i.p. injections results in the deletion of exon 11 in the Brca1 gene. We have previously demonstrated that this model results in the deletion of Brca1 protein content specifically in the skeletal muscle with no off‐target effects (Jackson et al. 2018). We confirmed the deletion of exon 11 in the current cohort of animals at 30–34 weeks of age via PCR‐amplified verification of DNA recombination (data not shown), as described previously (Jackson et al. 2018). Successful knockdown of Brca1 mRNA was also confirmed in the plantaris muscle for each group (data not shown).

Experimental approach

Ten‐ to 12‐week‐old HSA‐ mER‐Cre‐mER(+) Brca1(fl⁺/fl⁺) and HSA‐ mER‐Cre‐mER(+) Brca1(fl⁺/fl⁻) male and female mice were injected for 5 consecutive days with 2 mg day^–1 tamoxifen or vehicle solution. Mice were allowed to age for an additional 20 weeks prior to experimentation. We chose 20 weeks because we have previously demonstrated that Brca1 protein was ablated from skeletal muscle 8 weeks post‐tamoxifen treatment. Mice were placed under isoflurane‐induced anaesthesia for the collection of muscle tissue, before being humanely killed. Animals of the genotype HSA‐ mER‐Cre‐mER(+) Brca1(fl⁺/fl⁺) administered tamoxifen or sunflower seed oil and ethanol are referred to as Brca1KO^smi and wild‐type (WT) mice, respectively. Animals of genotype HSA‐ mER‐Cre‐mER(+) Brca1(fl⁺/fl⁻) administered tamoxifen are referred to as Het mice. Those mice administered vehicle are referred to as WT mice. In addition, a cohort of HSA‐ mER‐Cre‐mER(+) was assessed to ensure no off target effects of tamoxifen delivery. To date, we have found no effect of the tamoxifen delivery in any of our outcome measurements, including treadmill running, mitochondrial respiration, organ mass and/or body weight (data not shown).

Muscle fibre size and type

Muscle fibre cross‐sectional area (CSA) and muscle fibre type were measured in the tibialis anterior (TA) muscle as a result of mixed muscle fibre type, as described previously (Schmidt et al. 2017). Images of muscle cross‐sections were taken by a blinded investigator from six different locations across the belly of the muscle section and were probed with primary antibodies against myosin heavy chain type I (BA‐F8), IIa (SC‐71) and IIb (BF‐F3) (Development Studies Hybridoma Bank, University of Iowa, Iowa City, IA, USA) and anti‐dystrophin (Rb‐9024; ThermoFisher, Waltham, MA, USA). Myosin heavy chain types I, IIa and IIb and dystrophin were stained with secondary fluorescence probes 350, 488, 546 and 647, respectively. Stained cross‐sections were imaged using an EVOS FL automicroscope and the accompanying software (Life Technologies, Bothell, WA, USA). Fibre type and CSA were assessed using Image J (NIH, Bethesda, MD, USA), as described previously (Schmidt et al. 2017). All images were coded and randomized to allow for blinded analyses.

Histochemical staining of muscle fibres

TA tissue sections were stained with haematoxylin and eosin (H&E) in accordance with standard protocols. TA muscle collagen content was visualized using Sirius Red staining. Tissue sections stained for H&E and Sirius Red were imaged using standard light microscopy and analysed using ImageJ, version 1.6.0 (NIH). TA muscle was used for consistency across microscopy methods. The content of muscle fibres positive for the mitochondrial complexes cytochrome c oxidase (COX) and succinate dehydrogenase (SDH) was visualized using previously described methods (Otis et al. 2004). All images were coded and randomized to allow for blinded analyses.

Vascular density assessment

Vascular density was assessed in TA tissue by immunohistochemical staining, and is reported as the mean percentage CD31⁺ (endothelial cell marker) per 20× field of view as described previously (Ryan et al. 2016). TA muscle was used for consistency across microscopy methods. All images were coded and randomized to allow for blinded analyses.

Isometric force production

Isometric force production was assessed in vivo for the plantar flexor muscles of WT and KO mice, as well as in vitro for extensor digitorum longus (EDL) muscles of WT, Het and KO mice, as described previously (Spangenburg et al. 2008; Iyer et al. 2016). For the in vivo experiments, mice were exposed to 4–5% isoflurane in an induction chamber and then placed on a nose cone for maintenance of anaesthesia using a precision isoflurane vaporizer. Toe pinch reflex was used to confirm proper anaesthetic depth. Optimal voltage was determined for each protocol. For the in vivo protocol, force was normalized to muscle mass as described previously as a result of the multiple muscle groups involved (Iyer et al. 2016).

In vitro fatigue resistance was assessed in EDL muscles for all three groups using a 10 min fatigue protocol consisting of 300 contractions at 30 Hz. Optimal voltage was determined for each protocol. Plantar flexor muscles were used for their capacity to measure isometric force in vivo. EDL muscles were used for in vitro measurements because it is easy to excise and tie are both proximal and distal tendons without damaging the muscle. For the in vitro protocol, muscle length and mass were recorded for the calculation of physiological CSA and specific force, expressed as N cm^–2 (Barton et al. 2008).

Muscle lactate assay

Muscle lactate production was assessed by measuring the change in lactate present in the contractile bath media before and after the 10 min fatigue protocol in EDL muscles. Muscle lactate was measured using a Lactate Assay Kit (MAK064‐1KT; Sigma‐Aldrich, St Louis, MO, USA) in accordance with the manufacturer's instructions and normalized to the EDL muscle mass.

Preparation of permeabilized muscle fibre bundles

Permeabilized muscle fibre bundles were prepared from portions of the medial gastrocnemius muscle as a result of its oxidative phenotype and ease of mechanical fibre separation, as described previously (O'Rourke et al. 2018). Medial gastrocnemius muscle was used because of its predominantly red phenotype. Medial gastrocnemius muscle was excised and placed in ice‐cold buffer X [in mm: 7.23 K₂EGTA, 2.77 CaK₂EGTA, 20 imidazole, 20 taurine, 5.7 ATP, 14.3 phosphocreatine, 6.56 MgCl₂·6H₂O and 50 Mes, pH 7.1, 295 mosmol (kgH₂O)^–1]. Connective tissue and fat were removed using a dissecting microscope, and fibre bundles of ∼1 mg wet weight were separated using microdissecting tweezers. Muscle fibre bundles were permeabilized in buffer X containing 30 μg mL^–1 saponin with continuous rotation at 4°C for 30 min. Muscle bundles were promptly transferred to ice‐cold buffer Z [in mm: 110 K‐Mes, 35 KCl, 1 EGTA, 5K₂HPO₄, 3 MgCl₂·6H₂O and 5 mg mL^–1 BSA, pH 7.4, 295 mosmol (kgH₂O)^–1] and washed with continuous rotation at 4°C for 15 min.

High‐resolution respirometry

High‐resolution respirometry measurements were made on permeabilized muscle fibre bundles using the OROBOROS Oxygraph‐2K (Oroboros Instruments, Innsbruck, Austria) as described previously (O'Rourke et al. 2018). Experiments were conducted at 37°C with a starting oxygen concentration of ∼300–350 μm in buffer Z, containing 20 mm creatine monohydrate and 25 μm blebbistatin. Mitochondrial respiration was assessed by the sequential addition of substrates at a final concentration of 4 mm pyruvate, 0.5 mm malate, 5 mm glutamate, 2.5 mm ADP, 5 mm succinate, 5 μm cytochrome c, 10 μm rotenone, 5 μm anti‐mycin A, 2 mm ascorbic acid and 0.5 mm TMPD (N,N,N′,N′‐tetramethyl‐p‐phenylenediamine dihydrochloride). Mitochondrial membrane integrity was confirmed by the addition of exogenous cytochrome c. Muscle fibre bundles that produced a >10% increase in respiration were excluded. Muscle fibre bundles were rinsed in distilled H₂O, freeze‐dried (Labconco, Kansas City, MO, USA) and weighed (Orion Cahn C‐35; Thermo Electron, Beverly, MA, USA). O₂ consumption rates were normalized to dry weight and converted to pmol s^–1 mg^–1 dry weight O₂. All chemicals and reagents were purchased from Sigma‐Aldrich.

Mitochondrial H₂O₂ production

Mitochondrial H₂O₂ emission was measured fluorometrically in permeabilized muscle fibre bundles, as described previously (O'Rourke et al. 2018). Briefly, muscle fibre bundles were placed in 1 mL cuvettes containing Amplex Ultra Red (25 μm), horseradish peroxidase (1 U mL^–1) and blebbistatin (25 μm) at 37°C, and mixed using a magnetic stirrer. Resorufin fluorescence was detected at Ex:EM 565:600 via a spectrofluorometer (SPEX Fluoromax 3; HORIBA Jobin Yvon, Kyoto, Japan). Mitochondrial H₂O₂ emission was stimulated through the addition of succinate (10 mm) and mitochondrial H₂O₂ buffering capacity was assessed via the addition of thioredoxin reductase and glutathione reductase inhibitors, auranofin (AF) (1 μm) and carmustine (BCNU) (100 μm), respectively. Fibre bundles were rinsed in distilled H₂O, freeze‐dried and weighed after each experiment. Standard curves were generated and fluorescence was converted to nm mg^–1 dry weight H₂O₂. Scavenger index, a measure of H₂O₂ buffering capacity, was calculated as the percentage change in resorufin fluorescence between succinate and the succinate + AF/BCNU stages.

Skeletal muscle mitochondrial isolation

Skeletal muscle mitochondria were isolated as described previously (Hepple et al. 2015). Freshly isolated mitochondria were used for the osmotic swelling assay and the remaining isolate was stored at –80°C for assessment of mitochondrial DNA (mtDNA) mutations.

Mitochondrial osmotic swelling assay

Mitochondrial osmotic swelling was measured using a protocol adapted from one described previously (Lee et al. 2005). Briefly, 250 μg of isolated mitochondria was incubated in 1 mL of KCl buffer (in mm: 140 KCl and 3 Hepes, pH 7.4) at room temperature and mixed via a magnetic stir bar. Following a brief background measure, CaCl₂ was added (final concentration of 200 μm), after which 3.125 μg of alamethicin was added following a further measuring period. Mitochondrial swelling as a result of solute flux in and out of the mitochondrial matrix was recorded by the change in absorbance at λ_540 nm using a spectrofluorometer. Rates of swelling were calculated based on the Δabsorbance_540 nm min^–1 between the initial linear background rate and the rate following the addition of CaCl₂. ΔAbsorbance_540 nm was normalized to maximal mitochondrial swelling, as determined via the addition of alamethicin.

Mitochondrial DNA mutations assay

The mutation frequency in mtDNA was measured using an adaptation of a protocol described previously (Vermulst et al. 2008). Isolated mitochondria were processed using a Qiagen minispin kit (model 27104) in accordance with the manufacturer's recommendations for the extraction of mtDNA. Mutations were detected via a quantitative PCR assay using primer sets mControl forward: TCGGCGTAAAACGTGTCAAC; mControl reverse: CCGCCAAGTCCTTTGAGTTT; mTaq634 forward: ACTCAAAGGACTTGGCGGTA; and mTag634 reverse: AGCCCATTTCTTCCCATTTC. Mitochondrial DNA isolated from 1‐year‐old C57BL/6 and old Polg mutator mice served as negative and positive controls, respectively. Polg mutator mice are known to accumulate significant mtDNA mutations by 1 year of age as a result of the expression of a proofreading‐deficient version of polymerase gamma (Trifunovic et al. 2004; Kujoth et al. 2005)

Electron microscopy

Soleus muscle was used because of its high mitochondrial content. The soleus was cut into 1–2 mm strips and fixed for 1–3 hours at 4°C in a solution of 2% glutaraldehyde in 0.1 m sodium cacodylate buffer (pH 7.4). After rinsing, tissues were post‐fixed in a solution of 1% osmium tetroxide (Stevens Metallurgical, New Yor, NY, USA) in 0.1 m sodium phosphate buffer (pH 7.4). Rinsed tissues were then dehydrated by passage through a graded ethanol series and embedded in Spurr's media (Electron Microscopy Sciences, Hatfield, PA, USA). Ultra‐thin sections (70 nm) were placed on 200‐mesh copper grids coated with Coat Quick ‘G’ (Daido Sangyo Co., Kawasaki, Japan) and then stained sequentially by immersion in 2% aqueous uranyl‐acetate followed by Reynold's lead citrate. Grids were examined in a Jeol 1200 EX electron microscope (Jeol Inc., Peabody, MA, USA) at an accelerating voltage of 80 kV, and images were recorded using a MegaView III CCD camera (Olympus, Tokyo, Japan) and iTEM software (Soft Imaging Systems, Münster, Germany).

Muscle homogenization and western blotting

TA muscles were excised and immediately flash frozen. Samples were stored at –80°C. Muscle samples were placed in ice‐cold cell lysis buffer (50 mm Tris‐HCl, EDTA 1 mm, NaCl 150 mm, SDS 0.1%, sodium deoxycholate 0.5%, igepel Ca 630 1%, pH 7.5) containing protease and phosphatase inhibitors. Samples were homogenized and centrifuged. Protein concentrations were determined via a bicinchoninic acid assay (ThermoFisher). Proteins (20–30 μg) were separated on SDS‐PAGE gels (Criterion TGX Stain‐Free Gels; Bio‐Rad, Hercules, CA, USA), which were photo‐activated prior to transfer to polyvinylidene difluoride (PVDF) membranes (Trans‐Blot Turbo Transfer System; Bio‐Rad). PVDF membranes were imaged to record total lane protein before being blocked at room temperature for 1 h in 5% non‐fat dry milk or 5% BSA, and then incubated overnight with primary antibodies (Table 1) at 4°C. Following primary antibody incubation, membranes were incubated at room temperature for 1 h with goat anti‐rabbit HRP‐conjugated secondary antibodies (ab97051; Abcam, Cambridge, MA, USA) at concentrations of 1:2000 to 1:20,000. Proteins were visualized via chemiluminescence with care being taken to ensure signal detection was in a linear range and not saturated (Clarity Western ECL Substrate, Bio‐Rad; SuperSignal West Femto, ThermoFisher). Captured ECL signals were quantified using Image Lab Software, version 5.2.1 (Bio‐Rad) and normalized to total lane protein content. See Table 2 for primary antibodies used, dilation factors and exposure times.

Table 1.

Immunohistochemical antibodies source information

Antibody	Vendor	Catalogue number and reference code	Dilution factor
Myosin heavy chain type I	Development Studies Hybridoma Bank, University of Iowa (Iowa City, IA, USA)	#BA‐F8 RRID:AB_10572253	1:50
Myosin heavy chain type IIa		#SC‐71 RRID:AB_2147165	1:50
Myosin heavy chain type IIb		#BF‐F3 RRID:AB_2266724	1:50
Anti‐dystrophin	ThermoFisher (Waltham, MA, USA)	#Rb‐9024‐P RRID:AB_149805	1:50
Alexa Fluor 350	ThermoFisher	#A‐21140 RRID:AB_2535777	1:250
Alexa Fluor 488	ThermoFisher	#A‐21121 RRID:AB_2535764	1:250
Alexa Fluor 546	ThermoFisher	#A‐21045 RRID:AB_2535714	1:250
Alexa Fluor 647	ThermoFisher	#A‐21245 RRID:AB_141775	1:250

Table 2.

Western blot primary and secondary antibodies source information

Antibody	Vendor	Catalogue number and reference code	Dilution factor and exposure time
4‐HNE (4‐hydroxynonenal)	Abcam (Cambridge, MA, USA)	#Ab46545 RRID:AB_722490	1:3000 75 s
APE1 (apurinic/apyrimidinic endonuclease)	Abcam	#Ab137708 RRID:AB_2732912	1:1000 5 s
TTC11/Fis1 (mitochondrial fission 1 protein)	Abcam	#Ab71498 RRID: AB_1271360	1:1000 30 s
Glutathione reductase	Abcam	#Ab137513 RRID:AB_2732913	1:1000 20 s
NTH1 (endonuclease III‐like protein 1)	Abcam	#Ab191413 RRID:AB_2732911	1:1000 5 s
Ogg1 (8‐oxoguanine glycosylase)	Abcam	#Ab135940 RRID:AB_2732910	1 ug/ml 15 s
Parkin	Abcam	#Ab15954 RRID:AB_443270	1:1000 15 s
p‐Parkin Ser65	Abcam	#Ab154995 RRID:AB_2732915	1:1000 6 s
TXNRD1 (thioredoxin reductase 1)	Abcam	#Ab124954 RRID:AB_10975643	1:1000 5 s
TXNRD2 (thioredoxine reductase 2)	Abcam	#Ab180493 RRID:AB_2732914	1:1000 10 s
Goat anti‐rabbit HRP‐conjugated secondary antibody	Abcam	#ab97051 RRID:AB_10679369	1:2000 – 1:20,000
Drp1 (dynamin‐related protein 1)	Cell Signaling (Beverly, MA, USA)	#8570s RRID:AB_10950498	1:500 60 s
p‐Drp1 Ser616	Cell Signaling	#3455 RRID:AB_2085352	1:1000 6 s
MFN2 (mitofusin‐2)	Cell Signaling	#9482s RRID:AB_2747394	1:1000 30 s
OPA1 (optic atrophy type 1)	Cell Signaling	#80471 RRID:AB_2734117	1:1000 10 s

Statistical analysis

D'Agostino–Pearson omnibus tests were conducted to confirm that data followed a normal distribution. Student's t tests were used to analyse differences between WT and KO groups when data were normally distributed. One‐way analysis of variance tests were used to analyse differences in contractile function and mitochondrial respiration between the WT, Het and KO groups when data were normally distributed. Multiple comparisons were made using Sidak post hoc tests. Data that did not follow a normal distribution were analysed using a Mann–Whitney U test when comparing WT and KO groups only, and a Kruskal–Wallis test with Dunn's multiple comparisons when comparing WT, Het and KO groups. Statistical analyses were conducted using Prism, version 7.03 (GraphPad Software Inc., San Diego, CA, USA). The α level was set at 0.05. All data are reported as the mean ± SEM.

Results

Brca1KO^smi mice develop kyphosis and altered skeletal muscle size

Twenty weeks post‐tamoxifen injection, Brca1KO^smi mice developed kyphosis, which was not seen in WT mice 20 weeks post‐vehicle injection (Fig. 1 A), nor was it identified in Cre(+) animals injected with tamoxifen (data not shown). The development of kyphosis occurs for a variety of reasons, although commonly as a result of muscle weakness. Thus, we aimed to determine whether loss of Brca1 altered the overall phenotype of the skeletal muscle. The absolute mass of the gastrocnemius, quadriceps and TA muscles was similar between WT and Brca1KO^smi groups (Fig. 1 B and C). However, both overall TA muscle diameter and TA muscle fibre CSA were significantly greater in Brca1KO^smi compared to WT skeletal muscle (Fig. 1 D and E). In agreement, Brca1KO^smi skeletal muscle exhibited a rightward shift in the frequency distribution of individual muscle fibre size compared to the WT mice (Fig. 1F and G). Because the muscle in the Brca1KO^smi exhibited significant increases in muscle fibre size without changes in muscle mass, we assessed differences in collagen content in each group. We found no quantitative differences in the total amount of Sirius Red staining between groups, suggesting no differences in collagen content (Fig. 1 H and I).

Figure 1. Brca1KO^smi develop kyphosis and altered skeletal muscle characteristics compared to aged‐matched WT mice.

Representative photographs and X‐ray images show the development of kyphosis in Brca1KO^smi mice (A). Muscle mass of gastrocnemius (n = 8 per group), quadriceps (n = 8 per group) and TA (n = 5 per group) muscles was similar between WT and Brca1KO^smi mice (B and C). Despite a similar muscle mass, we found a significantly greater muscle diameter and muscle fibre CSA in TA muscles of Brca1KO^smi compared to WT mice (D–F). Representative images of H&E stained TA muscle cross‐sections used for muscle CSA measurements are included (G). Collagen content was similar between WT and Brca1KO^smi mice (H). Representative images of TA muscle cross‐sections (n = 5 per group) stained for collagen (I). Black bars refer to WT mice and grey bars refer to Brca1KO^smi mice. All images were coded and randomized to allow for blinded analyses. *Statistically different from WT (P < 0.05). Data are the mean ± SEM.

Brca1KO^smi mice display an altered skeletal muscle fibre phenotype

Next, we assessed whether there were differences in skeletal muscle fibre type between WT and Brca1KO^smi mice in TA muscle. The skeletal muscle of Brca1KO^smi had significantly fewer type IIb fibres relative to WT skeletal muscle with no significant differences in the other fibre type populations (Fig. 2 A and B). In addition, COX and SDH staining showed a pattern of expression, also supporting a shift in muscle fibre type. Skeletal muscle from Brca1KO^smi possessed significantly fewer negative COX and SDH fibres, at the same time as also possessing a significantly greater percentage of strong staining COX and SDH fibres (++) (Fig. 2 C and D). Because capillary density often follows muscle fibre type, we assessed the number of capillaries (CD31+ cells) in the muscle cross‐sections. We found a trend toward reduced capillaries content in TA muscle cross‐sections (P = 0.07); however, when corrected to the fibre number, we found a significantly reduced capillary density in Brca1KO^smi compared to WT skeletal muscle (Fig. 2 E–G).

Figure 2. Brca1KO^smi mice display an altered skeletal muscle fibre phenotype compared to aged matched WT mice.

Myosin heavy chain (MHC) labelling of TA muscle cross‐sections shows significant muscle fibre type shifts in Brca1KO^smi mice (A). Representative immunofluorescence MHC images (B) are labelled as: blue = type I, green = type IIa, no stain = IIx and red = IIb. SDH and COX staining of WT and Brca1KOsmi muscle cross‐sections supports a shift toward an oxidative phenotype in Brca1KO^smi skeletal muscle (C). Representative images of SDH and COX staining (D) are labelled as: blue = SDH and brown = COX. Capillary number per mm², based on CD31⁺ positive staining, trended lower in Brca1KO^smi muscle (E) and was significantly lower when CD31⁺ was corrected to muscle fibre number (F). Representative images (purple/red = dystrophin and green/yellow = CD31⁺) demonstrate the significant reductions in capillary density following Brca1 ablation (G). Black bars refer to WT mice and grey bars refer to Brca1KO^smi mice. All images were coded and randomized to allow for blinded analyses. ^*Statistically different from WT (P < 0.05). Data are the mean ± SEM.

Isometric force production is reduced and fatigue resistance is increased in Brca1KO^smi skeletal muscle

To assess whether the altered Brca1KO^smi phenotype impacted muscle force production, we measured isometric force of the plantar flexor group and EDL muscle using an in situ stimulation platform and an in vitro approach, respectively. Normalized muscle force was significantly lower in Brca1KO^smi compared to WT mice using both the in vivo (g/muscle) and in vitro (N/cm²) approaches (Fig. 3 A and B). Reductions in specific force were not significant between WT and Het EDL muscle. Fatigue resistance, measured in the same pool of EDL muscles, demonstrated significantly greater fatigue resistance in EDL muscles of the Brca1KO^smi mice (Fig. 3 C), whereas no differences were observed between WT and Het mice. Lactate produced during the isometric contraction fatigue protocol (Fig. 3 D) was increased in the EDL of Brca1KO^smi mice, suggesting that Brca1KO^smi may be more reliant on glycolytic metabolism compared to WT mice.

Figure 3. Altered contractile responses and lactate production in Brca1KO^smi skeletal muscle compared to skeletal muscle from aged matched WT mice.

In vivo gastrocnemius/plantaris specific force production was significantly reduced in Brca1KO^smi (n = 5) compared to WT (n = 4) mice (A). In vitro isometric force production was similarly reduced in Brca1KO^smi (n = 9) EDL muscles (B); however, Het (n = 11) and WT (n = 13) muscles were not different. In vitro EDL fatigue resistance was significantly greater Brca1KO^smi mice compared to age‐matched Het and WT mice (C). Significantly increased lactate production within Brca1KO^smi EDL muscles during the in vitro fatigue protocol suggests a greater reliance on glycolytic metabolism (D). Black bars refer to WT mice, light grey bars refer to Het and dark grey bars refer to Brca1KO^smi mice. ^*Statistically different from WT (P < 0.05). Data are the mean ± SEM.

Brca1KO^smi mice exhibit a unique mitochondrial phenotype showing reduced respiratory function

Our previous data in human primary myotubes (Jackson et al. 2014) and isolated mitochondria (Jackson et al. 2018) suggest that oxidative capacity might be uniquely affected in the Brca1KO^smi mice; thus, we aimed to directly assess mitochondrial respiration. Accordingly, we employed a permeabilized muscle fibre bundle approach to assess mitochondrial respiration (because it maintains the mitochondria in their native state at the same time as still maintaining the natural architecture of the muscle cell). Using fibre bundles prepared from the medial portion of the gastrocnemius muscle, we found that O₂ consumption from muscle fibre bundles of Brca1KO^smi mice was lower compared to the WT and Het animals (Fig. 4 A–E). Specifically, mitochondrial respiration was significantly lower during state 3 respiration at complexes I compared to WT mice, and complexes I+II, II and IV relative to WT and Het mice. Respiration was comparable between WT and Het mice. We also assessed mitochondrial H₂O₂ emission potential and found no significant differences between the WT, HET or KO mice following the addition of either succinate (Fig. 4 F) or thioredoxin reductase and glutathione reductase inhibitors, AF and BCNU, respectively (Fig. 4 G). However, scavenger index, a representation of H₂O₂ buffering capacity, was significantly elevated in KO mice compared to WT and Het mice (Fig. 4 H).

Figure 4. Skeletal muscle mitochondrial respiration and H₂O₂ emission in Brca1KO^smi mice.

Medical gastrocnemius muscle fibre bundles from WT (n = 15), Het (n = 16) and KO (n = 14) mice were separated and permeabilized in preparation for high‐resolution respirometry and H₂O₂ emission. Substrates and inhibitors were added sequentially to measure complex I leak and ADP‐stimulated complexes I, I+II, II and IV mitochondrial respiration, as well as H₂O₂ emission. Data were corrected to muscle dry weight. Complex I leak (state 2) respiration was similar between WT, HET and Brca1KO^smi fibre bundles (A); however, complex I (state 3) respiration (B), complex I+II (state 3) respiration (C), complex II (state 3) respiration (D) and complex IV (state 3) respiration (E) were significantly lower in Brca1KO^smi fibre bundles. Representative tracing of high‐resolution respirometry measurement with regions of interest labelled (A) to (E) to identify the corresponding image (F). H₂O₂ emission was not different between groups under succinate‐only (G) or succinate + AF/BCNU (thioredoxin and glutathione reductase inhibitors, respectively) H, scavenger index, calculated as the percentage difference in H₂O₂ emission following the addition of AF/BCNU inhibitors, was significantly increased in BCRA1KO^smi compared to Het and WT fibre bundles (I). Black bars refer to WT mice, light grey bars refer to Het and grey bars refer to Brca1KO^smi mice. ^*Statistically different from Het (P < 0.05). Data are the mean ± SEM. [Color figure can be viewed at wileyonlinelibrary.com]

Brca1KO^smi mitochondria show abnormal morphology, a greater accumulation of mtDNA mutations and increased Ca²⁺‐induced swelling

Using electron microscopy of soleus muscle, we observed that Brca1KO^smi muscle was characterized by morphologically abnormal mitochondria, displaying a fused and swollen reticulum compared to that of WT and Het animals. We observed similar mitochondrial morphology between WT and Het skeletal muscle, which was consistent with mitochondrial respiratory data (Fig. 5 A). Because BRCA1/Brca1 is largely recognized as a DNA repair protein (Wang et al. 2000; Liu & West, 2002; Kim et al. 2004; Yoshida & Miki, 2004; Saha et al. 2010; Roy et al. 2012) and co‐localizes with mitochondria and probably mtDNA (Coene et al. 2005; Maniccia et al. 2009; Jackson et al. 2018), we measured the frequency of mtDNA mutations in skeletal muscle. mtDNA was derived from isolated mitochondria sourced from red, white and mixed skeletal muscle from the hind‐ and forelimbs. Skeletal muscle from mBrca1KO^smi mtDNA contained a significantly greater mutational burden compared to the WT (Fig. 5 B). Isolated mitochondria were suspended in KCl buffer and absorbance at λ_540nm was monitored to determine changes in mitochondrial volume via solute flux before and after the addition of Ca²⁺ (CaCl₂). Mitochondria isolated from Brca1KO^smi skeletal muscle were significantly more susceptible to Ca²⁺‐induced increases in mitochondrial volume (Fig. 5 C and D).

Figure 5. Mitochondrial morphology, DNA mutations and Ca²⁺‐induced swelling.

Sections of soleus muscle were cut and fixed for imaging of mitochondrial morphology via transmission electron microscopy (TEM). Mitochondria were isolated from the skeletal muscle of age‐matched WT (n = 14) and Brca1KO^smi (n = 11) mice for the assessment of mtDNAmutations. Isolated mitochondria were equilibrated in a KCl buffer before the addition of CaCl₂. Changes to solute flux in and out of the mitochondrial matrix were recorded by the change in absorbance at λ_540 nm. ΔAbsorbance_540 nm was normalized to maximal mitochondrial swelling, determined via the addition of alamethicin. Lower absorbance equates to greater mitochondrion volume. Distinctive morphological shifts in the mitochondrial network were demonstrated via TEM images of soleus muscle cross‐sections; images are presented for WT, Het and Brca1KO^smi mice at magnifications of 20,000×, 40,000× and 100,000×. Black boxes indicate magnified regions (A). Morphological shifts were associated with significant increases in mitochondrial DNA mutations in Brca1KO^smi mice (B). Mitochondria isolated from Brca1KO^smi muscle further showed an increased susceptibility to Ca²⁺‐induced swelling (C and D). Black bars refer to WT mice and grey bars refer to Brca1KO^smi mice. ^*Statistically different from WT (P < 0.05). Data are the mean ± SEM.

Protein markers of DNA repair, anti‐oxidant defence and oxidative stress are similar between Brca1KO^smi and WT mice

BRCA1 has previously been shown to have regulatory functions in pathways of genomic stability, including base excision repair (BER) (Saha et al. 2010), anti‐oxidant defence (Bae et al. 2004; Gorrini et al. 2013) and cellular ROS levels (Saha et al. 2009). To further investigate the accumulation of mtDNA mutations in Brca1KO^smi mice, we measured key protein markers within these pathways from TA muscle homogenate (Fig. 6 A). We examined the skeletal muscle protein content of the BER enzymes Ogg1, NTH1 and APE1. Ogg1, NTH1 and APE1 were all similar between Brca1KO^smi and WT mice (Fig. 6 B–D). Protein levels of glutathione reductase (Fig. 6 E), cytoplasmic thioredoxin reductase (Fig. 6 F) and mitochondrial‐localized thioredoxin reductase (Fig. 6 G) were similar between Brca1KO^smi and WT mice. Finally, as a biomarker of oxidative stress, we measured the lipid peroxidation by‐product, 4‐HNE, which was similar between Brca1KO^smi and WT mice (Fig. 6 H).

Figure 6. Protein markers of HER, anti‐oxidant defence and oxidative stress.

Whole muscle homogenate from TAmuscles of age‐matched WT (n = 5) and Brca1Ko^smi (n = 4) mice was assessed for protein content of the BER enzymes, Oggl, NTH1 and APE1, anti‐oxidant defence enzymes, glutathione reductase (GR), TXNRDI and TXNRD2, and a biomarker of oxidative stress, 4‐HNE (A). The glycosylase, Oggl, was similar between WT and Brca1Ko^smi mice (B), in contrast to the glycosylase, NTH1 (C) and the endonuclease, APE1 (D), which were unchanged followed Brca1 KO. Protein content of the anti‐oxidant enzymes, glutathione reductase (GR), cytoplasm localized thioredoxin reductase (TXNRD1 and mitochondria localized thioredoxin reductase (TXNRD2) was similar between WT and Brca1KO^smi mice (E–G). The lipid peroxidationmarker, 4‐HNE, was comparable between WT and Brca1KO^smi muscle, which may indicate similar states of oxidative stress (H). Black bars refer to WT mice and grey bars refer to Brca1KO^smi mice. ^*Statistically different from WT (P < 0.05). Data are the mean ± SEM.

Protein markers of mitochondrial dynamics and mitophagy are altered in Brca1KO^smi skeletal muscle

In light of the alteration in mitochondrial morphology, we examined mitochondrial dynamics and mitophagy regulatory proteins by western blotting using TA muscle homogenate (Fig. 7 A). OPA1, an inner mitochondrial membrane fusion protein, showed a strong trend toward greater protein content in Brca1KO^smi muscle (P = 0.052) (Fig. 7 B). The outer mitochondrial membrane fusion regulator, MFN2, was similar between Brca1KO^smi and WT muscles (Fig. 7 C). Drp1, a mitochondrial fission protein, and its phosphorylated form (Drp1^S616) were both significantly reduced in Brca1KO^smi muscles (Fig. 7 D and E). Additionally, the mitochondrial fission protein, Fis1, which promotes Drp1 recruitment to the outer mitochondrial membrane (Palmer et al. 2013), was significantly reduced (P = 0.016) in muscle of Brca1KO^smi mice (Fig. 7 F). Muscle protein content of the pro‐mitophagy protein, Parkin, was not different between Brca1KO^smi and WT muscles. The phosphorylated form, Parkin^S65, was significantly elevated in Brca1KO^smi skeletal muscle (Fig. 7 G and H).

Figure 7. Mitochondrialdynamic and mitophagy protein markers in Brca1KO^smi skeletal muscle.

Tibialis anterior muscles from age‐matched WT (n = 5) and Brca1KO^smi (n = 4) mice were assessed via western blotting for specific protein markers of mitochondrial fusion (OPAl and MFN2), mitochondrial fission (Drp1 and Fis) and mitophagy (Parkin) (A). OPAl, an inner mitochondrial membrane fusion protein, showed a strong trend toward greater content in Brca1KO^smi muscle (B). MFN2, a regulator of outer mitochondrial membrane fusion, was similar between groups (C). The mitochondrial fission protein, Drp1, and its covalently active form, Drp1^S616, were both significantly lower in Brca1KO^smi muscle (D and E). Fis1, a mitochondrial fission promotor and Drp1 recruitment protein, was significantly lower in Brca1KO^smi muscle (F). The respective shifts in fusion and fission proteins imply a profusion state. Total protein content of the mitophagy marker, Parkin, was similar between groups (G), although a significant elevation in phospho‐Parkin^S65 (H) suggests that there may be greater initiation of mitophagy in Brca1KO^smi skeletal muscle. Black bars refer to WT mice and grey bars refer to Brca 1KO^smi mice. ^*Statistically different from WT (P < 0.05). Data are the mean ± SEM.

Discussion

Using our recently described Brca1KO^smi mouse model (Jackson et al. 2018), we demonstrate that the loss of skeletal muscle BRCA1 protein leads to kyphosis associated with a decrease in force‐producing capacity. This functional myopathy was accompanied by significant alterations in the overall phenotype of the skeletal muscle. Specifically, there was transition towards a more oxidative fibre type profile that was not accompanied by an increase in capillary density of the muscle. We discovered a consistent reduction across all of the muscle mitochondrial respiratory states assessed, suggesting the development of a mitochondriopathy. We found that Brca1KO^smi muscle mitochondria were visually swollen, contained a higher burden of mtDNA mutations and possessed a mitochondrial membrane more susceptible to osmotic swelling (as indicated by greater Ca²⁺‐induced membrane permeability). Together, our data demonstrate that KO of the Brca1 gene leads to a skeletal muscle phenotype characterized by weaker muscles, reduced capillary density, and dysfunctional mitochondria with reduced oxidative capacities and mutated mitochondrial DNA.

Mutations in BRCA1 have traditionally been considered in the context of breast and ovarian cancer development. With the exception of our previous studies (Jackson et al. 2014, 2018), little is known about Brca1 in skeletal muscle. The loss of skeletal muscle quality (i.e. force produced per mass of muscle) following the induced loss of Brca1 provides critical proof that Brca1 is critical for the maintenance of skeletal muscle function. The deterioration of muscle quality in the musculature of Brca1KO^smi mice probbaly led to the development of kyphosis. Indeed, muscle from Brca1KO^smi mice produced less force across most stimulation frequencies, regardless of whether muscle force was measured with direct or indirect electrical stimulation. It is noteworthy that the reduction in muscle force was largely the result of an increase in CSA, which produced a significantly lower specific force. The reduction in muscle force therefore is not a result of muscle atrophy but, instead, a loss of muscle quality. In a pre‐clinical setting, these data provide a strong rationale for assessment of skeletal muscle function in individuals carrying BRCA1 mutations that influence protein function. Het mice demonstrated a lower mean specific force compared to WT mice, and these differences were not significantly lower or higher than WT and Brca1KO^smi mice, respectively.

Brca1KO^smi developed an oxidative muscle fibre type profile characterized by greater COX/SDH staining and a reduction in IIb muscle fibres. However, greater muscle fibre CSA and reduced capillary density also developed, which is inconsistent with a more oxidative muscle phenotype (Schiaffino & Reggiani, 2011). The noted reduction in capillary density may have been an indirect result of Brca1 ablation because of a possible loss of cross‐talk between endothelial and skeletal muscle cells. Unexpectedly, Brca1KO^smi muscle demonstrated greater fatigue resistance, which is expected with the oxidative fibre type, although this is at odds with the reduced capillary density. An increase in fatigue resistance is somewhat reconciled by the elevated lactate emitted by Brca1KO^smi muscle during fatiguing contractions, implying a greater reliance on glycolytic metabolism. Similar metabolic shifts have previously been reported following the ablation of BRCA1 in human smooth muscle cells (Lovren et al. 2014) and Brca1KO^smi mice (Jackson et al. 2018), whereas the overexpression of BRCA1 in cultured breast cancer cells downregulated glycolytic gene expression and upregulated oxidative gene expression (Privat et al. 2014). Increased lactate production/glycolytic metabolism following Brca1 loss may represent an adaptation to compromised mitochondrial function in Brca1KO^smi mice.

Impaired mitochondrial respiratory capacity is a further factor potentially contributing to the overall loss of muscle quality in Brca1KO^smi mice. Using intact permeabilized muscle fibre bundles, we demonstrated respiratory impairments at complexes I, I+II, II and IV in Brca1KO^smi mice, supporting previous BRCA1/Brca1‐loss studies in human myotubes (Jackson et al. 2014) and isolated mitochondria from Brca1KO^smi mice (Jackson et al. 2018). However, these data substantially extend previous results by assessing respiration using an approach that ensures the native reticular structure of the mitochondria is maintained without disrupting the architecture of the muscle fibre. In addition, the present study is the first to compare respiration rates in WT and Brca1KO^smi mice with those of heterozygote Brca1 mice. The comparable respiration rates between WT and Het mice indicate that Brca1 is not haploinsufficient in skeletal muscle. The difference in respiratory capacity between the WT and Brca1KO^smi is associated with a lower muscle quality. Similar findings have been reported in other animal models where mitochondrial respiration is compromised and associated with reduced specific force capacity that is not a result of muscle atrophy (Gouspillou et al. 2018). However, it should be noted that other models demonstrate muscle atrophy accompanied by reduced absolute force with reduced respiration kinetics, although a frequent key characteristic of these models is increased mortality (Diaz et al. 2005; Morrow et al. 2017). Mitochondrial respiratory dysfunction reported in Brca1KO^smi mice was associated with increased mtDNA mutations, altered mitochondrial morphology and increased Ca²⁺‐induced membrane permeability.

Mitochondrial DNA is particularly susceptible to oxidative damage (Saha et al. 2010; Alexeyev et al. 2013), partly because of its proximity to the major site of cellular ROS generation: the mitochondrion (Shokolenko et al. 2009). Using a permeabilized muscle fibre approach, we found no direct evidence of increased mitochondrial ROS emission in Brca1KO^smi skeletal muscle, although we did note a significant increase in the scavenger index, indicating greater superoxide quenching potential in Brca1KO^smi mice. Increased mtDNA mutations are probably not the result of increased ROS but, instead, may be a result of reduced repair. BER is the most prevalent mtDNA repair pathway (Alexeyev et al. 2013) and is a target of BRCA1 regulation. BRCA1 has previously been reported to stimulate the expression and activity of the BER enzymes, OGG1, NTH1 and APE1 in MCF‐7 cells, whereas knockdown of BRCA1 reduced enzyme expression (Saha et al. 2010). These data appear to contradict those obtained for Brca1KO^smi mice, which show similar OGG1, NTH1 and APE1 protein levels. Protein levels do not necessarily reflect enzyme activity and thus impaired BER cannot be ruled out as a mechanism of mtDNA mutation in Brca1KO^smi mice. Additionally, contradictions between past and current data may be the result of model and tissue‐specific differences.

BRCA1 has also been reported to positively regulate anti‐oxidant pathways via interactions with Nrf2 (Gorrini et al. 2013). Protein content for glutathione and thioredoxin reductases, which are transcription targets of Nrf2 (Ma, 2013), were unchanged between Brca1KO^smi and WT mice, implying that loss of BRCA1‐Nrf2 anti‐oxidant regulation may not be a key factor in the accumulation mtDNA mutations. These data are consistent with our H₂O₂ data in permeabilized muscle fibre bundles, which would have been expected to elicit increased H₂O₂ emission and reduced scavenger index relative to WT animals. Likewise, 4‐HNE content between Brca1KO^smi and WT mice suggests that mtDNA mutations were not the result of oxidative stress from increased non‐mitochondrial ROS or deficiencies in anti‐oxidant defence pathways.

Analysis of mitochondrial dynamics proteins demonstrates altered regulation of the mitochondrial reticulum, which, in addition to greater membrane permeability, supports the enlarged/swollen morphological adaptations found in Brca1KO^smi skeletal muscle mitochondria. OPA1 and MFN2 are important regulators of inner mitochondrial membrane fusion (Neuspiel et al. 2005; Ishihara et al. 2006; Liu et al. 2009), whereas phosphorylation of Drp1 at serine 616 activates Drp1 fission activity (Taguchi et al. 2007; Kashatus et al. 2015), which is aided by Fis1‐mediated Drp1 recruitment (Palmer et al. 2013). Respective shifts in the skeletal muscle content of these proteins support the fused appearance of the mitochondrial reticulum. The inferred increase in the activity of the mitophagy regulator, Parkin, as a result of increased S65 phosphorylation also suggests that Brca1KO^smi mitochondria may be in a pro‐mitophagy state. Similarly, the knockdown of BRCA1 in fibroblasts increased the expression of autophagy markers, LC3 and LAMP1, and the mitophagy marker, BNIP3 (Salem et al. 2012), whereas LC3II was expressed following the knockdown of BRCA1 in breast cancer cells (Arun et al. 2015). Mitophagy induction is anticipated in response to mtDNA mutations (de Vries et al. 2012); however, mtDNA mutations continued to accumulate in Brca1KO^smi mice, which may be a result of incomplete autophagy/mitophagy activation (de Vries et al. 2012) or could reflect the severity of Brca1 ablation on mtDNA quality.

The Brca1KO^smi model used in the present study is specific to skeletal muscle, although muscle groups are not homogenous and, instead, vary considerably in key characteristics, such as fibre type, contractile strength and oxidative capacity. Adaptations to Brca1 ablation cannot therefore be assumed to occur equally across all muscle groups. Some conclusions, however, do presume a degree of extrapolation to provide potential mechanistic insights. This represents a limitation of the present study. Adaptations in Brca1KO^smi mice were found in TA, plantar flexor, EDL, gastrocnemius and soleus muscles. Furthermore, the development of kyphosis indicated that deleterious effects were not limited to the hind limbs. Together, these results suggest that shifts in muscle phenotype following the deletion of Brca1 are not confined to any particular muscle group, and thus it is expected that some uniformity in adaptations occurs across the muscles examined, although a very detailed analysis is necessary to confirm this idea.

In summary, skeletal muscle‐specific ablation of Brca1 leads to a deterioration in skeletal muscle quality, which manifests physiologically as kyphosis and a reduction in isometric force output. Mitochondrial respiratory capacity is also reduced in Brca1KO^smi skeletal muscle, which is associated with an accumulation of mtDNA mutations and abnormal mitochondrial morphology. Muscle force production, mitochondrial respiration and H₂O₂ emission data, as well as mitochondrial morphology, were similar between WT and Het mice. These data suggest that, in skeletal muscle, Brca1 may not be haploinsufficient, as has previously been reported in MCF‐7 and human breast epithelial cells (Cousineau & Belmaaza, 2007; Konishi et al. 2011), thereby implying that BRCA1/Brca1 haploinsufficiency may be cell or tissue‐specific. Loss of BRCA1/Brca1 has previously been implicated in cognitive, cardiac and vascular health (Shukla et al. 2011; Singh et al. 2013; Suberbielle et al. 2015), as well as comprising a modulator of substrate metabolism and metabolic phenotype in adipose tissue and skeletal muscle (Ortega et al. 2012; Jackson et al. 2014, 2018). This would imply that Brca1 deletion data in mice models are translatable to the loss of BRCA1 in humans. The detrimental impact of skeletal muscle‐specific Brca1 deletion in the present study may therefore be relevant to the maintenance of skeletal muscle quality in humans and, by extension, to the risk of developing disease, particularly those related to muscle contractile and metabolic function. Additionally, disease development and progression is a multifactorial process, such that loss of BRCA1 function in muscle may have further indirect consequences on human health that were not investigated in the current murine study.

The noted similarities between Brca1KO^smi mice and models of accelerated ageing are also of interest because skeletal muscle function is necessary for the preservation of functional independence and quality of life during the natural process of age‐related decline. This strengthens the need to better understand the effects of BRCA1 mutations in skeletal muscle across the lifespan and how such mutations interact with and progress age‐related muscle decline and disease. Advancements in our understanding of Brca1 functional mutations in mouse skeletal muscle can provide mechanistic insights into the role of Brca1 in skeletal muscle, and thereby direct efforts aiming to better understand the role of BRCA1 in human skeletal muscle and health.

Additional information

Competing interests

The authors declare that they have no competing interests.

Author contributions

The experiments were initiated at the University of Maryland (past) with the large majority completed at East Carolina University (present). Transmission electron microscopy was conducted by RHR in the ECU EM imaging core laboratory. MDT, APV, JMM, RML and EES were responsible for the conception of the study. MDT, APV, TER, JMM, RML and EES were responsible for the study design. MDT, APV, KCJ, AJA, NPB, RHR, SJPP and EES were responsible for the acquisition of data. MDT, APV, KCJ, AJA and EES were responsible for the analysis of data. MDT, TER, JMM, RML and EES were responsible for the interpretation of data. MDT and EES were responsible for drafting the manuscript. MDT, TER, JMM, RML and EES were responsible for revising the manuscript. All authors have approved the final version of the manuscript submitted for publication, agree to be accountable for all aspects of the work, and meet the standards required for authorship. All persons meeting the standards of authorship have been included as authors.

Funding

This work was supported by the National Institutes of Health: R01AR06660 and R21AR059913 (EES); R01HL25695 (JMM); R01AR059179 and R21AR067872 (RML); and F32HL129632 (TER).

Biography

Michael Tarpey completed his Master's degree in Exercise Physiology at the University of Hertfordshire, UK, before pursuing a PhD in Molecular and Cellular Science at Virginia Tech. Following the completion of his PhD in 2015, Michael took a postdoctoral position in the laboratory of Dr Espen Spangenburg at the East Carolina Diabetes and Obesity Institute (ECDOI), East Carolina University. Beyond investigating the role of BRCA1 in skeletal muscle, Michael's research is focused on understanding the mechanistic relationship between skeletal muscle function and mitochondrial bioenergetics.