Defective endoplasmic reticulum-mitochondria contacts and bioenergetics in SEPN1-related myopathy

Abstract

SEPN1-related myopathy (SEPN1-RM) is a muscle disorder due to mutations of the SEPN1 gene, which is characterized by muscle weakness and fatigue leading to scoliosis and life-threatening respiratory failure. Core lesions, focal areas of mitochondria depletion in skeletal muscle fibers, are the most common histopathological lesion. SEPN1-RM underlying mechanisms and the precise role of SEPN1 in muscle remained incompletely understood, hindering the development of biomarkers and therapies for this untreatable disease. To investigate the pathophysiological pathways in SEPN1-RM, we performed metabolic studies, calcium and ATP measurements, super-resolution and electron microscopy on in vivo and in vitro models of SEPN1 deficiency as well as muscle biopsies from SEPN1-RM patients. Mouse models of SEPN1 deficiency showed marked alterations in mitochondrial physiology and energy metabolism, suggesting that SEPN1 controls mitochondrial bioenergetics. Moreover, we found that SEPN1 was enriched at the mitochondria-associated membranes (MAM), and was needed for calcium transients between ER and mitochondria, as well as for the integrity of ER-mitochondria contacts. Consistently, loss of SEPN1 in patients was associated with alterations in body composition which correlated with the severity of muscle weakness, and with impaired ER-mitochondria contacts and low ATP levels. Our results indicate a role of SEPN1 as a novel MAM protein involved in mitochondrial bioenergetics. They also identify a systemic bioenergetic component in SEPN1-RM and establish mitochondria as a novel therapeutic target. This role of SEPN1 contributes to explain the fatigue and core lesions in skeletal muscle as well as the body composition abnormalities identified as part of the SEPN1-RM phenotype. Finally, these results point out to an unrecognized interplay between mitochondrial bioenergetics and ER homeostasis in skeletal muscle. They could therefore pave the way to the identification of biomarkers and therapeutic drugs for SEPN1-RM and for other disorders in which muscle ER-mitochondria cross-talk are impaired.

Fig. 1. Sepn1 KO mice present a metabolic phenotype in vivo with modified body mass composition, altered energy balance and defective fatty acid oxidation.

Follow-up of body mass composition and calorimetric parameters between ages five (N = 12 mice per group) and 47 weeks (N = 6 mice per group). a Body weight; b percentage of fat mass and c lean mass over total body weight. Metabolic cage parameters during resting (day), activity (night), and average are shown: d food intake, e energy expenditure per kg of lean body mass, f energy balance corresponding to the difference between food intake and energy expenditure, g respiratory exchange ratio RER = 0.7 corresponding theoretically to 100% fatty acid oxidation, RER = 1 corresponding to 100% carbohydrate metabolism and h percentage of fatty acid oxidation in 5- and 32-week-old (w.o.) WT and SEPN1 KO.

Fig. 2. Sepn1 KO mice show defective exercise endurance and increased muscle glucose metabolism during exercise.

a Maximal treadmill running distance for 2-month-old mice of indicated genotype (N = 12). b Plasma glucose in mice of indicated genotype at rest and 20 min after exhaustive running was tested (N = 12). c Representative periodic acid-Schiff (PAS) staining in liver and gastrocnemius muscle sections from mice of indicated genotype, at rest and after running. Insets represent samples digested with amylase.

Fig. 3. SEPN1 controls mitochondria bioenergetics.

a Percentage of the ratio of the metabolic activity (MTS assay) of WT and SEPN1 KO HeLa cells between 2 days of culture in low glucose (5.5 mM) and high glucose (25 mM) (N = 9). b Measurement of cellular ATP content in WT and SEPN1 KO HeLa cells. Results from three independent experiments are summarized as a graph (N = 6). c Analysis of mitochondrial membrane potential (Ψm) as measured by TMRM intensity in WT and SEPN1 KO HeLa cells (N = 34). d Mitochondrial respiratory chain specific activity in WT and SEPN1 KO C2C12 (N = 4). e Oxygraphy studies of mitochondrial complexes in tibialis anterior and diaphragm muscles from Sepn1 KO mice (red bars) compared with WT animals (black bars). f Complex I activity in tibialis anterior (TA), diaphragm (DF), quadriceps (QD), and extensor digitorum longus (EDL).

Fig. 4. SEPN1 is a MAM-localized protein in nonmuscular and muscular cells.

a Localization of a cMyc-SEPN1 in murine primary myotubes. Skeletal muscle cells were transfected with a plasmid expressing a cMyc-SEPN1 protein, which co-localized with calnexin (an ER marker), with the outer mitochondrial membrane protein TOM40 and with VDAC2 (a MAM marker). Scale bar is 10 µm. b Protein components of subcellular fractions prepared from muscle (gastrocnemius) and cells (HeLa) revealed by western blot analysis. SEPN1 presence was shown using a specific antibody. IP3R3 and PDI were used as ER markers, β-tubulin as a cytosolic marker, Sigma 1-R as a MAM marker, cytochrome C (Cyt c) as a mitochondrial marker. H homogenate, Mp pure mitochondria, ER endoplasmic reticulum, MAM mitochondria-associated membranes, C cytosol. c Co-localization of SEPN1-Flag (red) and Sigma 1-R-EGFP (a MAM marker, green) in C2C12 (right panel) and HeLa cells (left panel). The lower-right panel displays the merged image of the two stains. The lower-left panels display the SEPN1 signal overlaid with MAMs (MAM boundaries are highlighted in gray) in a rainbow lookup table (LUT). MAMs-SEPN1: Manders coefficient for SEPN1 staining was calculated as the proportion of SEPN1 signal overlapping with the Sigma 1-R marker.

Fig. 5. SEPN1-devoid cells display less ER-mitochondria apposition (MAMs).

a Representative images of ER (green) and Mito (red) trackers in WT and SEPN1 KO HeLa cells following 16 h of tunicamycin (1 μg/mL) or vehicle. The outline of the ER tracker signal was used to create a mask within the Mito stained area that was quantified. The signal of Mito is shown as a color-coded fire image overlaid by the ER mask. b Dot plots representing the Mito area fraction in the ER mask decreased in SEPN1 KO HeLa cells (ER_(mito) area fraction). c The size of ER (ER area fraction) in HeLa cells. d Mitochondrial skeletonized signal was used to evaluate the mitochondria network. The mean branch length was calculated in WT and SEPN1 KO HeLa cells. e Representative images of ER (green) and Mito (red) trackers in WT and SEPN1 KO C2C12 cells. f The Mito area fraction in the ER mask decreased in SEPN1 KO C2C12 cells (ER_(mito) area fraction). g The size of ER (ER area fraction) in C2C12 cells. h Mitochondrial skeletonized signal was used to evaluate the mitochondria network. The mean branch length was calculated in WT and SEPN1 KO C2C12 cells. Each dot represents one cell.

Fig. 6. SEPN1 regulates ER-mitochondria Ca²⁺ homeostasis.

a WT and SEPN1 KO HeLa cells were co-transfected with ER aequorin-based Ca²⁺ sensors and SEPN1 plasmids and calcium refilling of the ER was recorded (Ca²⁺ 1 mM) (N = 10). The traces are representative of ten independent experiments that led to similar results. b Measurements of [Ca²⁺] using recombinant aequorin-based Ca²⁺ sensors in the mitochondria upon agonist stimulation (100 μM Histamine) (N = 10). The traces are representative of ten independent experiments that led to similar results. c Flag and beta-Actin representative immunoblot of proteins extracted from SEPN1 KO HeLa cells transfected with SEPN1 wt-Flag, SEPN1 CC-Flag, and SEPN1 SS-Flag.

Fig. 7. Abnormalities in body composition, ATP levels, and SR-mitochondrial contacts in patients confirm defective bioenergetics as a novel pathomechanism in SEPN1-RM.

A Two typical SEPN1-RM patients showing loss of subcutaneous fat between the ages of 14 and 17 years (Patient a) and extremely reduced body mass with preserved ambulation in adulthood (Patient b). A small group of patients was in the overweight/obesity range, showed abdominal fat accumulation and had severe muscle weakness leading to loss of ambulation and reduced upper limb antigravity movements in their early teens (Patient c, wheelchair-bound from age 13 years). Anthropometric analysis (d) confirmed a positive correlation (p = 0.021) between BMI and disease severity (e). Most underweight patients had moderate disease, only a few of them having mild or more severe forms, while all the overweight cases were severe. B Reduced ATP content in primary fibroblasts from three patients homozygous or compound heterozygous for nonsense mutations leading to a severe reduction of SEPN1 protein (Western blot lanes 1–3; the last lane serves as a control to show SEPN1 expression), compared with four different age- and passage-paired controls. ATP was reduced in cells from both overweight/severe Patient 1 and underweight/moderated Patients 2–3. C Representative electron microscopy images from control (a–b), and SEPN1-RM patient (c–f) muscle biopsies. In control muscle fibers, mitochondria are positioned at the I band and often form pairs (a, black arrows) on both sides of Z lines; CRUs have the classic triad structure and are frequently associated with mitochondria (inset in a and white arrows in b). In muscle fibers from patients, mitochondria are usually misplaced from their correct triadic position, forming clusters and/or longitudinal columns between the myofibrils (empty arrows in c). Relocation of mitochondria leaves areas completely free of mitochondria (lower-right corner in c). In these areas, triads may still be present (white arrows in d). Intermyofibrillar white spots lacking mitochondria and triads are often present within apparently normal myofibrils (e, f, arrowheads). Scale Bars: a, c, e: 1 µm; b, d and f: 0.5 µm; inset: 0.2 µm.