BRCA1 is a novel regulator of metabolic function in skeletal muscle

Abstract

Breast cancer type 1 (BRCA1) susceptibility protein is expressed across multiple tissues including skeletal muscle. The overall objective of this investigation was to define a functional role for BRCA1 in skeletal muscle using a translational approach. For the first time in both mice and humans, we identified the presence of multiple isoforms of BRCA1 in skeletal muscle. In response to an acute bout of exercise, we found increases in the interaction between the native forms of BRCA1 and the phosphorylated form of acetyl-CoA carboxylase. Decreasing BRCA1 content using a shRNA approach in cultured primary human myotubes resulted in decreased oxygen consumption by the mitochondria and increased reactive oxygen species production. The decreased BRCA1 content also resulted in increased storage of intracellular lipid and reduced insulin signaling. These results indicate that BRCA1 plays a critical role in the regulation of metabolic function in skeletal muscle. Collectively, these data reveal BRCA1 as a novel target to consider in our understanding of metabolic function and risk for development of metabolic-based diseases.

Breast Cancer 1, early onset (BRCA1) is an established DNA repair gene and is recognized as an estrogen-sensitive tumor suppressor gene (1, 2). Genetic variation in the BRCA1 gene is associated with increased risk for the development of breast cancer and/or tumorigenesis in reproductive tissues (3). The BRCA1 gene produces either a full-length breast cancer type 1 susceptibility protein (BRCA1) or through alternatively splicing two documented variants, BRCA1Δ11 or BRCA1Δ11b, both of which lack a nuclear localization signal (4). Recently, BRCA1 was identified as a regulator of lipid metabolism in human breast cancer cells (MCF7) as a result of direct interaction with the phosphorylated form of acetyl-CoA carboxylase (ACC-p) at the BRCA1 C-terminal (BRCT) domains (5, 6). The interaction encourages the maintenance of the phosphorylated state of ACC thereby altering lipid metabolism in the cancer cell line (5, 6).

ACC has two isoforms, ACC1 or ACC2, with ACC2 containing an extra 146 amino acids in the NH₂-terminal region. ACC activity is negatively regulated by phosphorylation of residue Ser⁷⁹ on ACC1 and Ser²²¹ on ACC2 (7, 8). In the active form (i.e., dephosphorylated), ACC catalyzes the carboxylation of acetyl-CoA into malonyl-CoA (MaCoA). Changes in cellular MaCoA content alter intracellular lipid dynamics in two specific manners (9, 10). MaCoA directly contributes to de novo synthesis of palmitate via FAS and MaCoA also allosterically inhibits carnitine palmitoyltransferase-1 (CPT-1), a mitochondrial long chain fatty acid transporter (11). Thus, in mammary tissue the ability of BRCA1 to affect ACC activity alters cellular lipid concentrations by indirectly regulating rates of fatty acid synthesis and/or the flux of fatty acids into the mitochondria.

In response to increased energetic demand, such as during exercise, skeletal muscle increases fatty acid entry into the mitochondria through CPT-1-mediated transport (12). During acute exercise, this rate is enhanced in muscle by a reduction in MaCoA content due to reduced ACC activity (10), with the inhibition of ACC mediated by AMP-activated protein kinase (AMPK) (9). Because regular exercise is known to promote metabolic health, multiple investigations have targeted this metabolic mechanism in an effort to treat metabolic disease. ACC function is critical to the regulation of lipid metabolism in skeletal muscle because genetic ablation of ACC results in higher basal and insulin-stimulated palmitate oxidation rates (7). Collectively, these results indicate that ACC plays an important role in skeletal muscle metabolism; however, key regulators of ACC remain unidentified.

A previous investigation has shown BRCA1 mRNA is expressed in C2C12 myoblasts, which would suggest a possible role in skeletal muscle (13). However, to our knowledge, no studies have mechanistically examined the role of BRCA1 in skeletal muscle. Thus, the purpose of this investigation was to establish whether BRCA1 was expressed in skeletal muscle and determine whether it acts as a regulator of metabolic function. Using an integrative and translational approach, we have identified that BRCA1 is expressed in both mouse and human skeletal muscle. Our data indicate that BRCA1 influences mitochondrial function, lipid storage, and insulin responses in the skeletal muscle cell. Overall, the results indicate that BRCA1 function extends beyond reproductive tissues, suggesting that broader roles across multiple tissues should be evaluated.

MATERIALS AND METHODS

Animals

The University of Maryland Institutional Animal Care and Use Committee Review Board approved all aspects involving animal research. Male and female C57Bl/6 mice ranging from 8 to 10 weeks of age were utilized in this study. C57Bl/6 mice were divided into two groups, an exercise group (male, n = 6; female, n = 7) and a sedentary group (male, n = 6; female, n = 7). All animals were treadmill acclimated and then only the exercise group was subjected to an acute bout of treadmill exercise (male: 21.92 ± 0.57 m/min, 40.1 ± 2.75 min, 5% incline; female: 26.57 ± 0.30 m/min, 36.5 ± 4.3 min, 5% incline) while the sedentary animals were placed on the treadmill in a stationary position for an equivalent time. The males were run at a lower speed to maintain similar relative intensities between the males and females. At the end of the exercise bout, animals were euthanized and skeletal muscle was harvested, snap-frozen in liquid nitrogen, and stored at −80°C. For the high-fat diet (HFD) study, 12-week-old C57Bl/6 male mice (n = 4/group) were placed on a normal chow diet (NCD) (10% kcal fat; D12450K Research Diets) or a HFD (45% kcal fat; D12451 Research Diets) for 10 weeks. At the conclusion of 10 weeks, animals were euthanized and skeletal muscle was harvested, snap-frozen in liquid nitrogen, and stored at −80°C.

Human subjects

The Ethics Committee of the Karolinska Institutet approved all aspects of this study. Twenty-six healthy subjects were included in the study, 13 males and 13 females, who all gave their informed consent to participate. The mean (range) age, height, and weight were 26 (21–30) years, 177 (158–190) cm, and 75 (58–90) kg, respectively. The mean (range) maximal oxygen consumption (VO₂max) was 48 (43–64) ml·kg⁻¹·min⁻¹. Well-trained subjects (VO₂max > 65 ml·kg⁻¹·min⁻¹) were excluded to maximize the subjects’ exercise responses (14). Skeletal muscle biopsies from the vastus lateralis (VL) muscle were obtained using the percutaneous needle biopsy technique at rest (prior to the exercise bout) and at 30 min after the exercise bout, alternating between the legs. All biopsy samples were snap-frozen in liquid nitrogen and stored at −80°C until further analysis.

Mouse mRNA analysis

Isolation of RNA and subsequent cDNA synthesis from the gastrocnemius or plantaris muscles and mouse testes (positive control) was performed according to the previously described techniques (15, 16). Specific primers for mouse Brca1 were as follows: forward 5′-CAC AGC GTA TGC CAG AGA AA-3′ and reverse 5′-ATC CTG GGA GTT TGC ATT TG-3′.

Human mRNA analysis

Total RNA from the skeletal muscle biopsies was isolated using standard methods (17) and real-time RT-PCR was used to measure two short BRCA1 splice variants (BRCA1Δ11, BRCA1Δ11b) and total BRCA1 (BRCA1 total). Amplicons were synthesized using previously described sequences (4). GAPDH was used as an internal control (4352934E, Applied Biosystems Inc.). Primer efficiency was tested with standard titration curves and did not differ between the primer pairs. The expression of each target was determined by the DeltaDelta Ct method (18).

Homogenization

Mouse skeletal muscle was mechanically homogenized according to previously described techniques (15, 19). Human skeletal muscle biopsies were homogenized on ice using glass-on-glass homogenizers in RIPA buffer [150 mM NaCl, 10 mM Tris-HCl, 5 mM EDTA, 0.5% Na deoxycholate, 0.1% SDS, 1% Triton X-100, and protein inhibitor cocktail complete mini (Roche Diagnostics)]. Total protein was determined in each sample using the Pierce BCA protein assay (mouse) or Bradford protein assay (human) as previously described (20, 21).

BRCA1 immunoprecipitation

Immunoprecipitation of endogenous BRCA1 protein in mouse (500 μg total protein) or human (150 μg total protein) skeletal muscle homogenate was performed with 2 μg BRCA1 antibody (I-20, sc-646; Santa Cruz Biotechnology, Santa Cruz, CA) and incubated, rocking, overnight at 4°C. The antigen-antibody complex was combined with protein A affinity and then washed through repeated centrifugation steps. After the final wash the pellet was suspended in sample buffer and heated to 100°C for 5 min. The sample was then cooled and the eluted protein was loaded onto an SDS-PAGE gel for Western blot analysis.

Immunoblotting analysis

Immunoblotting for either BRCA1 immunoprecipitation solution or whole muscle homogenate protein was performed as previously described (15, 20, 22). Membranes were probed with an antibody specific for BRCA1 (I-20 or D-20, 1:200; Santa Cruz Biotechnology), ACC-p or ACC total (1:1,000; Cell Signal, Boston, MA), Akt-p or Akt total (1:1,000; Cell Signal), α-actinin (1:5,000; Sigma-Aldrich, Saint Louis, MO), red fluorescent protein (RFP) (1:2,000; Thermo Scientific, Waltham, MA). MCF7 cell lysates or mammary gland tissue from BRCA1 KO mice (BRCA1 MG KO) were used as a positive or negative control, respectively. BRCA1 MG KO lysates also served as a positive control to detect the BRCA1 protein short splice variants (BRCA1Δ11, BRCA1Δ11b) as previously described (23–26).

Nuclear and cytoplasmic fractions and BRCA1 immunoblotting

Nuclear and cytoplasmic fractions were prepared using a nuclear and cytoplasmic extraction kit (NE-PER® nuclear and cytoplasmic extraction reagents, 78833; Thermo Scientific) as previously described (27). Purity and enrichment of nuclear fractions were demonstrated by measuring nuclear protein lamin A/C and the cytosolic protein β-tubulin (T5201; Sigma-Aldrich) as previously described (28).

Mouse MaCoA measures

MaCoA measures were performed on the gastrocnemius muscles from sedentary and acute exercise female and male mice as previously described (20).

Human myoblast cell culture

Human skeletal muscle myoblasts and media were purchased from Zen-Bio (Research Triangle Park, NC). Human skeletal myoblasts were derived from the VL muscle biopsy from a healthy, lean (normal BMI), 24-year-old female and only used at low passage number (<7). Myoblasts were induced to differentiate to myotubes upon reaching ∼90% confluence. All plates were visually examined to ensure that myotubes covered ∼90% of the well prior to any experimental utilization.

Adenovirus shRNA-hBRCA1 myotube infection

To reduce BRCA1 content in the human myotubes, the cells were transduced with either scrambled shRNA adenovirus (scrambled-shRNA) or adenovirus containing a shRNA sequence targeting the coding region of BRCA1 (nt.530-550_NM_007294) (shRNA-hBRCA1) and containing a RFP tag overnight (Vector Biolabs, Philadelphia, PA). Myotubes were then returned to regular growth media for 48 h and equivalent adenovirus infection was confirmed via imaging detection of RFP. BRCA1 mRNA was isolated from adenovirus-infected human myotubes as previously described (29). Reduction in BRCA1 mRNA expression was confirmed in human myotubes 72 h post adenovirus infection using the same primers as above for BRCA1 total, BRCA1Δ11, and BRCA1Δ11b.

Mitochondrial oxygen consumption

Mitochondrial oxygen consumption rates (OCRs) were measured in shRNA-hBRCA1- or scrambled-shRNA-treated human myotubes similar to our previously described technique (20, 30). In parallel plates, ACC phosphorylation was assessed after 5-amino-1-β-D-ribofuranosyl-imidazole-4-carboxamide (AICAR) treatment in the same myotubes with either normal or reduced BRCA1 content.

Human myotube 2-NBDG insulin stimulated glucose uptake

Ad-shRNA-hBRCA1-RFP- or AD-shRNA-RFP-treated human myotubes were serum starved for 4–5 h as previously described (31). Myotubes were then washed three times and incubated with or without 50 nM insulin in warm Krebs-Ringer solution (135 mM NaCl, 10 mM NaHCO₃, 5 mM KCl, 3 mM CaCl₂, 2 mM MgSO₄, 1.2 mM NaH₂PO₄) excluding glucose for ∼30 min at 37°C. At the conclusion of the insulin incubation, myotubes were exposed to 50 μM 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino]-2-deoxy-D-glucoseR (2-NBDG) with or without 50 nM insulin for ∼30 min at room temperature as previously described (31). Myotubes were then washed three times and placed in room temperature Krebs-Ringer solution and 2-NBDG fluorescence (438 nm excitation; 535 nm emission) and RFP fluorescence (556 nm excitation; 586 nm emission) measures were recorded using a BioTek Synergy plate reader (BioTek, Winooski, VT). All 2-NBDG measures were normalized to RFP values.

Human myotube insulin signaling

Scrambled-shRNA- or shRNA-hBRCA1-treated human myotubes were serum starved for 4–5 h in DMEM. Myotubes were then either control treated or treated with 50 nM insulin for 30 min as previously described (32). Protein was then isolated from control-treated or insulin-treated scrambled-shRNA or shRNA-hBRCA1 human myotubes as previously described (33).

Human myotube palmitate/oleate incubation and BODIPY imaging

shRNA-hBRCA1- or scrambled-shRNA-treated human myotubes were incubated in 30 μM BSA-conjugated palmitate/oleate mixture in DMEM for 4 h as previously described (34). Myotubes were then stained with BODIPY (Molecular Probes, Carlsbad CA) as previously described (35). Myotubes were imaged with a Zeiss Axiovision 4 (Zeiss, Oberkochen, Germany) as previously described (34, 36).

Human myotube reactive oxygen species (ROS) measures and imaging

Scrambled-shRNA- or shRNA-hBRCA1-treated human myotubes were placed in 2′,7′-dichlorofluorescein (DCF)-supplemented Krebs Ringer buffer for 30 min at 37°C and then washed three times. H2-DCF was quantified using a fluorescent plate reader (H2; Biotek, Burlington, VT). The myotubes were also visually imaged using an inverted epifluorescence microscope (Zeiss).

Statistical analysis

All data are represented as the mean ± SEM. Statistical analysis was conducted using ANOVA or t-test (paired and unpaired) approaches with Tukey post hoc tests employed when appropriate. P ≤ 0.05 was considered significant.

RESULTS

Brca1 expression in mouse skeletal muscle

In mouse skeletal muscle, Brca1 mRNA and Brca1 protein were detectable (Fig. 1A, B). In addition, using standard immunoblotting or immunoprecipitation approaches with different antibodies specific to either the N- or C-terminal region, we consistently detected Brca1 (220 kDa) in whole muscle homogenates (Fig. 1B, C). The resulting protein mass is consistent with previously published data concerning BRCA1 in other tissues (2, 6, 37).

Fig. 1.

A–F: Identification of Brca1 in male and female mouse skeletal muscle. A: Brca1 mRNA detection identified in female mouse gastrocnemius muscle (SM) with mouse testes serving as a positive control. B: Protein identification and verification of Brca1 (220 kDa) using an antibody comparison, Ab:1 (SC-I:20) Brca1 C-terminal (C-term)-specific antibody, Ab:2 (SC-D:20) Brca1 N-terminal (N-term)-specific antibody. C: Immunoprecipitation Brca1 in female mouse skeletal muscle using Ab1 and Ab2. D: Brca1 protein was greater in male compared with female mouse gastrocnemius muscle (P < 0.05). No significant differences were detected between male and female mice for Brca1Δ11 (78 kDa). Mammary gland from Brca1 KO mouse (MG KO) served as a negative control for full-length Brca1 and a positive control for the Brca1 splice variant (Brca1Δ11). E: No differences in Brca1 and Brca1Δ11 protein content in soleus, gastrocnemius (Gastroc), and TA muscles from adult female mice were detected. F: In skeletal muscle from adult male mice, no differences across muscle groups were detected in Brca1 protein content, while Brca1Δ11 content was significantly higher in the TA and soleus compared with the gastrocnemius muscle. Total protein staining gels were used to ensure equal loading of protein across samples. Data are presented as mean ± SEM (n = 3 mice per group for Western blotting measures). *P < 0.05 female versus male; #P < 0.05 TA versus gastrocnemius; $P < 0.05 TA versus soleus.

In reproductive tissue, Brca1 is an estrogen-sensitive gene (13, 38), however, in the gastrocnemius muscle of mice we detected greater levels of both the full-length and short [Brca1Δ11, 85 kDa as previously defined (26)] Brca1 splice variants in males compared with females (Fig. 1D). Full-length Brca1 protein levels did not differ between gastrocnemius and tibialis anterior (TA) muscles within female or male mice, respectively (Fig. 1E, F). Further, Brca1 protein was undetectable in the soleus muscle (Fig. 1E, F). In female mice, no differences were detected in Brca1Δ11 protein content in the soleus, TA, or gastrocnemius muscle (Fig. 1E). In contrast, in male mice, Brca1Δ11 was significantly higher in TA and soleus muscle compared with the gastrocnemius muscle (P < 0.05) (Fig. 1F).

Brca1 and Acc-p complex after an acute exercise bout in mice

In order to assess the potential interaction between the endogenous forms of Acc-p and Brca1, mice were subjected to an acute bout of treadmill exercise. Exercise induced significant elevations in the ratio of Acc-p/Acc total in both male and female gastrocnemius muscle compared with sedentary animals (P < 0.05) (Fig. 2A, B). As expected, MaCoA content significantly decreased in response to the exercise bout in the female animals, but surprisingly the increase in Acc-p levels did not correspond to a decline in MaCoA levels in the male mice (Fig. 2C). Further, MaCoA levels in sedentary female skeletal muscle were significantly higher compared with sedentary males (P < 0.05) (Fig. 2C). In response to the exercise bout there was a significant increase in Brca1-Acc-p protein-protein interaction in both female and male gastrocnemius muscle compared with sedentary animals (P < 0.05) (Fig. 2D). However, in females, the exercise-induced Brca1-Acc-p interaction tended to be greater than males (P = 0.09) (Fig. 2D).

Fig. 2.

A–E: An acute bout of exercise increases endogenous Acc-p and Brca1 interaction in skeletal muscle in adult male and female mice. A, B: Acc-p levels were significantly higher in response to an acute bout of exercise (Ex) in the gastrocnemius muscle from adult female and male mice compared with sedentary (Sed) mice. C: MaCoA levels were significantly higher in gastrocnemius muscles from sedentary female mice compared with sedentary male mice. Adult female mice exposed to an acute bout of exercise had significantly lower MaCoA levels compared with the female sedentary mice and no significant differences were apparent in gastrocnemius muscle from male mice. D: Interaction between Brca1 and Acc-p was greater in gastrocnemius muscle from sedentary females compared with sedentary males. Brca1 and Acc-p interaction was significantly higher after an acute bout of exercise in gastrocnemius muscles from both male and female mice when compared with their sedentary counterparts. Data are presented as mean ± SEM (n = 6–7 per group). *P < 0.05 sedentary versus exercise; #P < 0.05 male sedentary versus female sedentary; $P = 0.09 male exercise versus female exercise. E: Changes in Brca1 mRNA in the plantaris muscle in age-matched C57Bl/6 female mice after 12 weeks of a NCD or a HFD. Data are presented as mean ± SEM (n = 4 per group). *P < 0.05 NCD versus HFD. IB, immunoblot; IP, immunoprecipitation.

Brca1 mRNA expression is increased in skeletal muscle from mice on a HFD

Adult mice exposed to a HFD (45% kcal fat) exhibited an increase in Brca1 mRNA in the plantaris muscle compared with age-matched mice on a NCD (10% kcal fat) (Fig. 2E).

BRCA1 mRNA and BRCA1 protein expression in human skeletal muscle

In an effort to translate the observations found in our murine model, similar experiments were performed in humans. BRCA1 mRNA (full-length and splice variants) was detected in human skeletal muscle, with no differences between sedentary women and men (Fig. 3A). The two well-recognized short variants in human tissue, BRCA1Δ11 and BRCA1Δ11b mRNA, were readily expressed in human skeletal muscle with no sex differences (Fig. 3B, C). Using antibodies specific to the C- and N-terminal regions, the presence of BRCA1 protein in human skeletal muscle was verified (Fig. 3D), with no detectable sex differences (Fig. 3E). Previous research has suggested that BRCA1 isoforms exhibit specific intracellular localization (4, 37). Thus, BRCA1 protein content was measured in nuclear and cytosolic fractions from male and female skeletal muscle biopsies. The short splice variants (lacking nuclear localization signal, BRCA1Δ11, and BRCA1Δ11b) of BRCA1 were almost exclusively detected in the cytosolic fraction, whereas the full-length BRCA1 protein was predominantly detected in the nuclear fraction (Fig. 3F). Although mRNA for BRCA1Δ11 and BRCA1Δ11b was individually detected, it was not possible to distinguish the protein products due to similar molecular masses.

Fig. 3.

A–F: BRCA1 expression is detectable in biopsies taken from human VL. A–C: No significant differences were detected in mRNA expression of all recognized human BRCA1 variants in skeletal muscle biopsies taken from males and females. All data were normalized to GAPDH. D: The presence of BRCA1 protein in human skeletal muscle was verified using antibodies specific to the C terminus (C-term) (Ab:1) or N terminus (N-term) (Ab:2). E: No significant differences in BRCA1 protein were detected between men and women in biopsies from the VL. F: Full length BRCA1 was detected predominantly in the nuclear fraction isolated from the VL from men and women, whereas the BRCA1Δ11/Δ11 splice variant was detected in the cytoplasmic fraction. Cell lysates isolated from the mammary gland of BRCA1 KO animals were used as a positive control for the BRCA1Δ11/Δ11b splice variants. Laminin was used as a nuclear control protein and β-tubulin as the cytoplasmic control protein. Data are presented as mean ± SEM (n = 13 per group for mRNA analysis; n = 5–6 per group for BRCA1 Western blot analysis).

BRCA1 and Acc-p complex after an acute exercise bout in humans

To assess the role of BRCA1 in skeletal muscle, women and men performed an acute bout of exercise. The exercise bout resulted in a significant increase in ACC-p/ACC total compared with pre-exercise values (P < 0.05) (Fig. 4A, B). There was considerable individual variability in the magnitude of interaction between ACC-p and BRCA1 in both males and females prior to and after the completion of the exercise bout (Fig. 4C). There was no significant difference across sex in the ACC-p and BRCA1 interaction, thus the data were collapsed across sex. In response to an acute bout of exercise, there was an increase in BRCA1 and ACC-p protein interaction when compared with the resting state (Fig. 4D, E).

Fig. 4.

A–E: An acute bout of exercise in men and women resulted in a greater interaction between ACC-p and BRCA1 in muscle biopsies taken from the VL. A: Significant increases were found in ACC-p content compared with pre-exercise (pre) in both male and female subjects. B: Representative results from the ACC immunoblot generated from the human muscle biopsies. C: In response to an acute bout of exercise in men and women, the majority of subjects had increased BRCA1-ACC-p interaction compared with pre-exercise. Each line represents an individual subject and is marked with the sex and magnitude of response. D: Group average of BRCA1-ACC-p interaction in response to the acute bout of exercise. Data from male and female subjects were collapsed. E: Representative immunoprecipitation-immunoblot of the ACC-p-BRCA1 complex in response to the acute exercise. Data are presented as mean ± SEM (n = 5–6 per group for BRCA1 Western blot analysis). *P ≤ 0.05 sedentary versus exercise within female or male subjects.

Reductions in BRCA1 expression results in accumulation of intramyocellular lipid and reduced insulin-induced glucose uptake

Infection of human myotubes with shRNA-hBRCA1 resulted in nearly undetectable BRCA1 mRNA expression compared with infection with scrambled-shRNA (Fig. 5A). Loss of BRCA1 expression (shRNA-hBRCA1) in human myotubes resulted in accumulation of intracellular neutral lipid storage compared with control myotubes (scrambled-shRNA) (Fig. 5B). When myotubes were treated with conjugated-FFA, we observed an additional increase in lipid storage in myotubes with reduced BRCA1 expression compared with treated control myotubes (Fig. 5B). Myotubes with reduced BRCA1 expression exhibited reduced Akt phosphorylation in response to insulin treatment (50 nM for 30 min) compared with the insulin response in the control myotubes (Fig. 5C). Myotubes with reduced BRCA1 expression exhibited reduced insulin-induced glucose uptake compared with the control myotubes (Fig. 5D). There were no differences in basal glucose uptake responses between groups.

Fig. 5.

A–D: Reduction in BRCA1 enhances neutral lipid storage and decreases insulin-induced glucose uptake in primary human myotubes. A: Human myotubes transduced with shRNA for human BRCA1 (shRNA-hBRCA1) presented with reduced BRCA1 total, BRCA1Δ11, and BRCA1Δ11b compared with cells infected with scrambled-shRNA. B: Myotubes with reduced BRCA1 expression exposed to either BSA or 30 μM palmitate/oleate-conjugated BSA exhibited increased neutral lipid accumulation in myotubes compared with myotubes infected with scrambled-shRNA. C: Insulin-induced phosphorylation of Akt was reduced in myotubes with reduced BRCA1 expression compared with control myotubes. Myotubes were treated with 50 nM insulin for 30 min. D: Insulin-induced glucose uptake was reduced in myotubes with reduced BRCA1 expression compared with control myotubes. Insulin-induced uptake values are normalized to basal glucose uptake values. No differences in basal uptake were detected between groups. Data are presented as mean ± SEM (n = 3–5 per group for all analyses). *P < 0.05 scrambled-shRNA versus shRNA-hBRCA1 myotubes.

Reductions in BRCA1 expression results in decreased mitochondrial oxygen consumption in human myotubes

Human myotubes with reduced BRCA1 expression exhibited lower basal OCRs compared with scrambled-shRNA myotubes (P < 0.05) (Fig. 6A, B). Treatment of the myotubes with oligomycin resulted in reduced OCR with no differences detected in OCR values between groups (data not shown). Maximal OCR induced by uncoupling was also lower in shRNA-hBRCA1-treated compared with scrambled-shRNA myotubes (Fig. 6C). Subsequent addition of palmitate also resulted in lower OCR values in shRNA-hBRCA1 myotubes compared with scrambled-shRNA myotubes (Fig. 6D). No differences in mitochondrial ATP synthase content were detected between conditions (Fig. 6E).

Fig. 6.

A–E: Reduction in human myotube BRCA1 content decreases mitochondrial oxygen consumption. A: A representative respiration experiment in human myotubes transduced with scrambled-shRNA (Scramb-shRNA) (black line) or shRNA specific to human BRCA1 (gray line). These data are shown to provide a reader with a visual example of how the respiration experiments were conducted. B: Basal OCR was reduced in shRNA-hBRCA1 human myotubes compared with scrambled-shRNA myotubes. C: Uncoupling the mitochondria with FCCP (400 nM) resulted in a reduced OCR in shRNA-hBRCA1 myotubes compared with scrambled-shRNA. D: Palmitate (PA) (100 μM) stimulated OCR was significantly reduced in shRNA-hBRCA1 myotubes compared with scrambled-shRNA myotubes. E: No difference in the mitochondrial protein ATP 5A was detected in scrambled-shRNA and shRNA-hBRCA1 myotubes. Equal loading was confirmed through RFP measures. Data are presented as mean ± SEM (n = 3–5 per group for all analyses). *P < 0.05 scrambled-shRNA versus shRNA-hBRCA1 myotubes.

Reductions in BRCA1 expression results in reduced AICAR-induced phosphorylation of ACC in human myotubes

AICAR induces phosphorylation of ACC via activation of AMPK; thus we determined the ability of AICAR to induce phosphorylation of ACC under reduced BRCA1 conditions. ACC-p content did not increase after AICAR treatment of the human myotubes infected with shRNA-hBRCA1 compared with scrambled-shRNA-treated myotubes (Fig. 7A, B). However, reducing BRCA1 content resulted in a significant increase in basal phosphorylation of AMPK (∼20% increase) and increased basal phosphorylation of ACC (∼70% increase) suggesting that BRCA1 is not required for phosphorylation of ACC to occur (P < 0.05). In addition, it is possible that the failure of AICAR to further increase ACC phosphorylation under the reduced BRCA1 conditions is due to hyper-phosphorylation of ACC under basal conditions (Fig. 7A, B).

Fig. 7.

A, B: Reduced BRCA1 expression results in increased basal phosphorylation of ACC and failure of AICAR to induce further phosphorylation of ACC. A, B: ACC-p response was attenuated in AICAR-treated shRNA-hBRCA1 compared with scrambled-shRNA. Data are presented as mean ± SEM (n = 3–5 per group for all analyses). *P < 0.05 scrambled-shRNA versus shRNA-hBRCA1 myotubes.

Loss of BRCA1 expression increases ROS production leading to AMPK phosphorylation

We assessed global ROS content in myotubes with or without BRCA1. Using a global ROS indicator (H2-DCF) (Fig. 8A–F), we found higher ROS content in myotubes transduced with shRNA-hBRCA1 compared with scrambled-shRNA-treated myotubes (Fig. 8G).

Fig. 8.

A–G: Reduced BRCA1 expression in primary human myotubes increases ROS accumulation. A–F: Human myotubes transduced with shRNA-BRCA1 exhibit visual increases in ROS signal (green signal marked with white arrows) compared with cells infected with scrambled-shRNA (Scramb-shRNA). Both shRNA-BRCA1 and scrambled-shRNA plasmids contained a RFP tag to ensure appropriate transfection (red signal). A, D: Human myotubes infected with scrambled-shRNA exhibit little to no DCF signal. B, E: Human myotubes infected with shRNA-BRCA1 exhibit localized DCF signal at 10× magnification. C, F: Higher magnification (20×) imaging demonstrates that human myotubes infected with shRNA-BRCA1 exhibit DCF signal localized to unknown vacuoles. G: Quantification of basal ROS accumulation in human myotubes with shRNA-BRCA1 compared with myotubes infected with scrambled-shRNA. Data are presented as mean ± SEM (n = 3–5 per group for all analyses). *P < 0.05 scrambled-shRNA versus shRNA-hBRCA1 myotubes.

DISCUSSION

BRCA1 is a large polyfunctional protein (220 kDa) that regulates a number of different intracellular functions through a variety of mechanisms including protein-protein interactions (39). To date, BRCA1 has been predominantly recognized as a cell cycle regulator and a DNA damage repair protein in reproductive tissues (4, 39). However, the data presented here provide evidence for a novel role for BRCA1 as a regulator of metabolic function in skeletal muscle of both mice and humans. Specifically, in response to an energetic stress (i.e., acute exercise) there is a significant increase in the interaction of BRCA1 with ACC-p. In addition, adult mice consuming a HFD exhibited significant increases in Brca1 mRNA expression compared with mice on a NCD. Reducing BRCA1 content in human myotubes resulted in increased lipid storage, decreased insulin signaling, reduced mitochondrial function, and enhanced ROS production. Overall, our observations provide evidence for a previously undescribed role for BRCA1 and lend support for more detailed examinations of the role of BRCA1 as a metabolic regulator in skeletal muscle.

Genetic alteration in the BRCA1 sequence and/or alterations in BRCA1 expression are strongly correlated with risk of tumorigenesis in reproductive tissues (40). However, a functional role involving BRCA1 in skeletal muscle has gone unrecognized even though Brca1 mRNA content is known to increase in differentiating C2C12 myoblasts (13). Our data significantly extend this initial observation by demonstrating that both the long and short BRCA1 isoforms are present at the mRNA and protein level in adult mouse skeletal muscle and in human skeletal muscle. In addition, we found that mice exposed to a HFD exhibited increased full-length Brca1 mRNA expression in the plantaris muscle compared with mice fed a NCD, suggesting that Brca1 expression is modulated by changes in nutrient exposure. Consistent with previous literature, the full-length BRCA1 protein, which has an intact nuclear localization sequence (4, 24), was detected predominantly in the nuclear fraction of human skeletal muscle while the short splice variants, which lack a nuclear localization sequence (4), were detected in the cytosolic fraction. It is well-established that BRCA1 is an estrogen-sensitive gene (41); thus it was surprising to find higher Brca1 and Brca1Δ11 content in male than in female mouse gastrocnemius muscle. However, this sex difference did not translate to human skeletal muscle, as neither BRCA1 mRNA nor BRCA1 protein content differed across sexes. Finally, few differences were detected across skeletal muscle groups in mouse Brca1 expression and protein content. In fact, only the male mice presented with significant differences in BRCA1Δ11 across muscle groups with the TA and the soleus muscles having greater levels of BRCA1Δ11 compared with the gastrocnemius muscle. In male mice, most muscles, including the soleus muscle, are composed of a mixed phenotype (42). Thus, it is unclear why TA and/or soleus muscle would have significant differences in only BRCA1Δ11 (42). Our data demonstrate that multiple recognized forms of BRCA1 are detectable at both the mRNA and protein level in skeletal muscle from both mice and humans.

In a series of elegant experiments, BRCA1 was previously shown to directly bind to ACC-p in mammary tissue (5, 6). ACC is a critical regulator of lipid metabolism, as it catalyzes the production of MaCoA in skeletal muscle (7, 9, 10). MaCoA can act as a potent allosteric inhibitor of CPT-1 or can serve as a precursor for lipid synthesis (9, 43), with the former being the likely mechanism for ACC regulation of lipid metabolism in skeletal muscle (10, 44). In both mice and humans, we used an acute bout of exercise to energetically challenge the muscle to increase ACC-p content in skeletal muscle. The increase in ACC-p also resulted in increased interaction between BRCA1 and ACC-p. Further, our data in the mouse indicate differential Brca1/Acc-p interactions in females and males both at rest and in response to an acute bout of exercise, which was also associated with differing content of MaCoA in the same muscle. To the best of our knowledge, this is the first time a sex-based difference in skeletal muscle MaCoA content has been demonstrated. Reanalysis of data from a previous publication (45) where male and female mice were treated with AICAR demonstrated a similar finding. Specifically, female mice presented higher levels of resting MaCoA and a greater response to AICAR treatment than age-matched male mice. Due to limited sample amounts, we were unable to directly measure MaCoA content in the human samples to compare the BRCA1/ACC-p interaction in a quantitative fashion. Thus, at this point it is unclear if there was also a differential MaCoA response in our human samples, although acute bouts of exercise similar to the one performed here are known to reduce MaCoA content in human muscle (46). However, it was very clear that the magnitude of increase in interaction between BRCA1 and ACC-p in response to the exercise bout in the human samples was highly variable. Such variability may be, in part, due to the large degree of genetic variation in BRCA1, which can result in altered protein function and/or expression (47). ACC-p interacts with BRCA1 in the C-terminal region, specifically in the BRCT domains (48). Although the BRCT domains in BRCA1 are conserved across species, the BRCT domains in the human BRCA1 gene contain the highest degree of genetic variation (49). Because ACC-p binds to the BRCT domain within BRCA1, it is possible that specific sequence variation in this domain may contribute to variations in the degree of ACC function in humans. Further examination of this hypothesis will be necessary to determine if documented SNPs within the BRCT domains of BRCA1 are associated with the risk of development of metabolic disease in humans. However, to our knowledge no single nucleotide polymorphisms in the BRCA1 gene have been associated with metabolic disease in a genome-wide association study.

To determine whether BRCA1 plays a critical role in regulation of skeletal muscle metabolic function, we reduced BRCA1 expression through shRNA technology. Reducing BRCA1 content in human myotubes resulted in increased neutral lipid storage and resulted in a concurrent reduction in mitochondrial function in the myotubes. Thus, our results in skeletal muscle are in agreement with previous findings in MCF7 cells that reductions in BRCA1 expression results in increased lipid storage (5). BRCA1 localization within the cell remains controversial, however it is generally accepted that BRCA1 is found within the mitochondria (50, 51). Further, recent data have demonstrated that deletion of BRCA1 in cardiac muscle results in lower expression of multiple genes associated with the mitochondria and reductions in fatty acid oxidation rates (52). In addition, because ACC2 localizes to the mitochondria and can be bound by BRCA1, there is strong evidence to suggest that BRCA1 is associated with the mitochondria (5). Collectively, evidence suggests a role for BRCA1 in regulating mitochondrial function that is likely independent of ACC. BRCA1 is considered a poly-functional protein due to its ability to regulate diverse cellular mechanisms including DNA repair. The ability of BRCA1 to influence DNA repair may explain its role in the mitochondria because mitochondrial DNA is highly susceptible to DNA damage (53, 54). Thus, it is possible that loss of BRCA1 leads to loss of genomic integrity in the mitochondria. However, this hypothesis remains untested to date.

Our results show that reduced BRCA1 expression resulted in increased oxidative stress, as evidenced by increased DCF signal accumulation. At this time it is unclear if this is due to increased production of ROS or due to reduced ability to buffer ROS accumulation. Our data are in agreement with previous work demonstrating that the loss of BRCA1 in nonmuscle cells results in the accumulation of ROS (52, 55). BRCA1 appears to regulate antioxidant gene expression, thus protecting macromolecules against oxidative damage through a complex interaction with NRF2 (55, 56). Exposure to elevations in ROS levels in skeletal muscle contributes to dysfunction of a number of mechanisms, including dysregulation of insulin-induced glucose uptake and mitochondrial function (57, 58).

Reduced BRCA1 expression in the human myotubes resulted in reduced insulin-induced glucose uptake and lower Akt phosphorylation compared with the control myotube response. Our data suggest that the metabolic dysfunction induced by acute BRCA1 reduction induces poor insulin responses in human skeletal muscle. At this time, it is unclear if loss of BRCA1 directly impacts insulin signaling or if the reduced insulin response is a consequence of other cellular defects. For example, decreased BRCA1 expression resulted in reduction of mitochondrial function, excess ROS accumulation, and elevated neutral lipid storage, all of which have been associated with the development of insulin resistance. Future experiments will be crucial to identify the role of BRCA1 in the skeletal muscle insulin-signaling pathway. At this time, reduced BRCA1 expression in myotubes leads to a measurable decline in metabolic function of cultured skeletal muscle. Our data suggest that reductions in BRCA1 expression could contribute to metabolic disease susceptibility, thus arguing for more detailed in vivo explorations of the function of BRCA1 in skeletal muscle.

This initial study of BRCA1 in skeletal muscle has raised additional questions regarding the regulation and function of BRCA1 in skeletal muscle that will be addressed in future studies by our laboratory. Specifically, loss of BRCA1 leads to accumulation of neutral lipid, however it is currently unclear if increased lipid droplet content can be completely explained by reduced mitochondrial function or if this is in response to increased rates of triacylglycerol esterification. In addition, due to the number of intracellular changes that developed with the reduction in BRCA1 expression in the primary human myotubes, it was impossible to accurately determine if a reduction in the insulin response is due to a primary effect of BRCA1 reduction or a secondary effect as a consequence of lipid accumulation or ROS production. Finally, there were limitations in the number of measures that could be made with the human skeletal muscle biopsies due to the significant amount of tissue required to perform the immunoprecipitation and MaCoA measures.

In summary, we have identified that BRCA1 expression is an important regulator of multiple intracellular processes that may influence metabolic function in both mouse and human skeletal muscle. Future experiments are critical to more clearly delineate the mechanisms by which BRCA1 influences the intracellular functions of skeletal muscle. These findings significantly extend our understanding of BRCA1 physiology by considering mechanistic aspects outside of its classically known role in mammary tissue. Future studies will be necessary to examine the therapeutic potential of manipulating BRCA1 expression and/or function in skeletal muscle as a means to prevent the development of metabolic diseases.