Abstract Muscle degeneration is the most prevalent cause for frailty and dependency in inherited diseases and ageing. Elucidation of pathophysiological mechanisms, as well as effective treatments for muscle diseases, represents an important goal in improving human health. Here, we show that the lipid synthesis enzyme phosphatidylethanolamine cytidyltransferase (PCYT2/ECT) is critical to muscle health. Human deficiency in PCYT2 causes a severe disease with failure to thrive and progressive weakness. Pcyt2 mutant zebrafish and muscle-specific Pcyt2 knockout mice recapitulate the patient phenotypes, with failure to thrive, progressive muscle weakness and accelerated ageing. Mechanistically, muscle Pcyt2 deficiency affects cellular bioenergetics and membrane lipid bilayer structure and stability. PCYT2 activity declines in ageing muscles of mice and humans, and AAV-based delivery of PCYT2 ameliorates muscle weakness in Pcyt2 knockout and old mice, offering a therapy for rare disease patients and muscle ageing. Thus, PCYT2 plays a fundamental and conserved role in vertebrate muscle health, linking PCYT2 and PCYT2 synthesized lipids to severe muscle dystrophy and ageing. Keywords: Lipids, Genetic Disease, Ageing, Progressive Muscle Weakness, Phosphatidylethanolamine, Membrane Physicochemical Properties, Mitochondria, Gene Therapy, Muscle rejuvenation __________________________________________________________________ Skeletal muscle is the biggest organ in the human body, with essential roles in support, mobility and metabolism. Muscle degeneration, either as a result of inherited diseases^[99]1, chronic diseases, or ageing^[100]2 severely impairs the life quality, and health of millions of people. Complete understanding of pathophysiological mechanisms driving this pathology represents an important objective in medicine. Eukaryotic lipidome is complex with a potential of generating up to 100 000 different lipids^[101]3. Differences among specific lipids occur at subcellular compartments, cell and tissue types^[102]4. Tissue-specific differences suggest that certain organs use specific lipid pathways for organ health and longevity. In humans, genetic deficiency in phosphatidylethanolamine cytidyltransferase (PCYT2/ECT), the bottle neck enzyme in PE synthesis through the Kennedy pathway^[103]5, leads to complex and severe hereditary spastic paraplegia (HSP)^[104]6,[105]7. Here, we discover a conserved, essential and specific role for PCYT2 synthesized phosphatidylethanolamine (PE) in muscle health. Pcyt2 mutant zebrafish and muscle specific Pcyt2 knockout mice recapitulate several patient phenotypes, particularly failure to thrive, short stature, impaired muscle development, and progressive weakness, with accelerated ageing and shortened life span. In contrast, mice lacking Pcyt2 in other tissues appeared unaffected. Loss of PCYT2 in muscle results in alterations in the mitochondrial and cellular lipidome, affecting mitochondrial function and physicochemical properties of the lipid bilayer, compromising sarcolemmal stability and exercise tolerance. We further show that PCYT2 activity declines in aging muscles of humans and mice and that Pcyt2 gene-therapy in Pcyt2 knock-out and aged mice improved muscle strength. Thus, PCYT2 and PE synthesized via PCYT2, are essential to muscle health, linking mitochondrial and sarcolemmal lipid bilayer perturbations to muscle degeneration, exercise tolerance and aging. Results Patients with disease-causing PCYT2 variants fail to thrive PCYT2 mutations were recently discovered in patients who manifest a complex disorder that involves developmental gross motor delay and progressive overall muscle weakness^[106]6. Observing these patients, we found that those with a homozygous nonsense variant [107]NM_001184917.2:c.1129C>T (p.Arg377Ter) in PCYT2 exhibited an apparently lower weight and shorter body length from birth, throughout childhood, and adulthood ([108]Figure 1A,[109]B). We also assessed patients with mutations in EPT1, which encodes the final enzyme in PE synthesis via the Kennedy pathway ([110]Extended Figure 1A,[111]B). These patients manifest similar clinical features as those with PCYT2 mutations^[112]7,[113]8. Indeed, patients with a homozygous variant [114]NM_033505.4:c.335G>C (p.Arg112Pro) in EPT1 also exhibited growth defects, further confirming a role for the Kennedy pathway in postnatal growth ([115]Extended Data Figure 1B). Thus, in addition to previously described symptoms, mutations in two critical enzymes that generate PE in the Kennedy pathway are associated with stunted growth. Figure 1. Phenotypes of human PCYT2 rare disease mutations and pcyt2 mutant zebrafish. [116]Figure 1. [117]Open in a new tab (A) Body weight and (B) height gain of patient (male) carrying the homozygous nonsense variant 3c.1129C>T (p.Arg377Ter) in the PCYT2 gene. Controls indicate WHO standards of median weights and heights at the respective ages +/− 2 standard deviations (SD). (C) Representative appearance and quantifications of body length of control and hypomorphic pcyt2 mutant zebrafish at 14 months post fertilization. n=4 for each group. (D) Representative muscle sections and muscle myofiber sizes of control (n=4 animals and 233 myofibers in total) and hypomorphic pcyt2 (n=4 animals and 233 myofibers in total) zebrafish. Scale bar 50μm. Myofibers of the same anatomical region were analyzed with ≥ 50 myofibers per animal (E) Body weight gains of control (n=15) and Myf5Cre-Pcyt2 (n=11) littermates on standard chow diet. Two-Way ANOVA with multiple comparison followed by Bonferroni correction was used. ****p^(genotype) < 0.0001 (F) Appearance of 4 days old (P4) and 56 days old (P56) control and Myf5Cre-Pcyt2 littermates. Scale bars are 1 cm for P4 and 2 cm for P56. (G) Skeletal muscle appearance (quadriceps) isolated from 10 days control and Myf5Cre-Pcyt2 littermate mice. (H) Representative cross sections and (I) skeletal muscle myofiber diameter sizes from 6 months old control (n=4 mice and 570 myofibers) and Myf5Cre-Pcyt2 mice (n=4 mice and 640 myofibers). Myofibers were imaged using 10X magnification with ≥ 100myofibers analyzed per mouse. Scale bar 100μm. (I) Lipidomics analyses from quadriceps muscles isolated from 10 days old Myf5Cre-Pcyt2 and littermate control mice. Data are shown relative to control values. CE-cholesterol ester; Cer-Ceramides; DAG-diacylglycerols; LPC-lysophosphatidylcholines; LPE-lysophosphatidylethanolamines; PC-phosphatidylcholines; PE-phosphatidylethanolamines; PG-phosphatidylglycerols; PI-phosphatidylinositols; PS-phosphatidylserines; SM-sphingomyelins; TAG-triacylglycerols. n=4 per group. (J) Detailed analysis of PE species with different chain lengths from quadriceps muscles of Myf5Cre-Pcyt2 as compared to control mice. Data are shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. Pcyt2 deficiency in zebrafish affects muscle and whole-body growth Since Ept1 loss can be partially compensated by Cept1^[118]9, we focused on the bottleneck enzyme PCYT2 ([119]Extended Figure 1A). Given the ubiquitous tissue expression of PCYT2^[120]10 ([121]Supplementary Figure 3), its loss of function could potentially affect several organs, thus contributing to the complexity and severity of the disease. To gain insight into pathophysiological mechanisms, we first examined hypomorphic mutant pcyt2 zebrafish^[122]6. Similar to the human rare disease patients, pcyt2 mutant zebrafish were significantly smaller ([123]Figure 1C). Zebrafish and mouse models of hereditary spastic paraplegia rarely exhibit whole body growth phenotypes^[124]11 [125]12,[126]13. However, muscle development is essential for whole-body growth, and failure to thrive is a well-known feature of muscular dystrophies^[127]14. Therefore, we examined muscle morphology in pcyt2 mutant zebrafish. We observed significantly smaller skeletal muscles, with smaller fibers without total number change in pcyt2 mutant zebrafish ([128]Figure 1D, [129]Supplementary Figure 4A,[130]B). These results suggest that impaired muscle development could explain the stunted growth associated with pcyt2 loss-of-function mutations in zebrafish and patients. Pcyt2 muscle deficiency in mice impairs muscle and whole-body growth In mice, global disruption of Pcyt2 results in embryonic lethality^[131]15. Therefore, to study the role of Pcyt2 in muscle health, we generated mice with muscle-specific Pcyt2 deletion early during muscle development to recapitulate the human condition. We crossed Pcyt2^flox/flox with Myf5 promoter driven Cre mice to generate Myf5Cre-Pcyt2 offspring, given that Myf5 is the first broadly expressed myogenic regulator in the developing myotome^[132]16. Pcyt2 deletion was validated by RNA sequencing ([133]Extended Data Figure 2A, Myf5Cre-Pcyt2 mice were born at normal Mendelian ratios, but were significantly smaller at birth (postnatal day P1) and early postnatal days (P4), gained less weight, and grew less during the postnatal period compared to controls, as observed for both genders ([134]Figure 1E,[135]F; [136]Extended Data Figure 3A-[137]D). Neither Myf5Cre nor Pcyt2^flox/flox littermate controls displayed a phenotype, therefore we used Pcyt2^flox/flox littermates as controls for all subsequent experiments. We noticed that limb muscles were smaller in Myf5Cre-Pcyt2 mice compared to controls already at P10 and at 2 months old ([138]Figure 1G; [139]Extended Data Figure 3E-[140]H). Myofiber size was reduced in skeletal muscle ([141]Figure 1G-[142]I). Lipidomic analysis of quadriceps muscle isolated from 10-day old Myf5Cre-Pcyt2 mice showed a marked reduction in the levels of PE, particularly of long-chain fatty acid PE species ([143]Figure 1J, [144]K). In addition, our data revealed an increase of PC lipids from the PC branch of the Kennedy pathway and upregulation of several enzymes from PE and PC branch of the Kennedy pathway ([145]Figure 1J, [146]Extended Data Figure 3I). Muscle growth is mediated first by muscle satellite cell proliferation and an increase in myofibers until ~P7, and subsequently via myofiber hypertrophy^[147]17. The number of proliferating cells as well as the rate of myoblast proliferation was unaffected in the developing muscles and in myoblasts of Myf5Cre-Pcyt2 mice ([148]Extended Data Figure 4A, [149]B). Although the number and distribution of Pax7^+ muscle progenitor cells was similar in adult Myf5Cre-Pcyt2 and control mice ([150]Extended Data Figure 4C), Myf5Cre-Pcyt2 mice showed a mild but significant reduction in myoblast fusion, with thinner myofibers ([151]Figure 2A-[152]C). Several pro-fusion and late differentiation markers were reduced in the myoblasts from Myf5Cre-Pcyt2 mice ([153]Figure 2D). There were no apparent changes in the levels of transcriptional regulators of myofiber differentiation, such as MyoD and MyoG ([154]Figure 2D). Membrane lipid composition is essential for myoblast fusion^[155]18. Given that PE lipids are abundant membrane lipids, we wanted to address if unavailable PE in the myoblast membrane would affect fusion into myofibers. We used the non-toxic, PE-specific SA-Ro binding probe^[156]19 to mask the externally exposed membrane PE during myoblast fusion, mimicking PE deficiency. Indeed, myoblast fusion was severely affected ([157]Figure 2E-[158]G), showing that membrane PE is required for efficient myoblast fusion. Figure 2. Pcyt2 deficiency affects muscle stem cell fusion and muscle hypertrophic growth. [159]Figure 2. [160]Open in a new tab (A) Representative images of Mf20 stained myofibers and (B) primary myoblast fusion index quantification of Control and Myf5Cre-Pcyt2 primary myoblasts after differentiation in vitro. Nine biological replicate myoblast cultures from three independent isolations were used. Each dot represents a calculated fusion index from in total n=9 cultures for each group. ≥ 300 nuclei were counted per one culture. Myofibers were imaged using 10X magnification. Scale bar 50μm. (C) Representative images and myofiber diameter quantification of Control (n=145 myofibers) and Myf5Cre-Pcyt2 (n=158 myofibers) after differentiation from primary myoblasts in vitro. Myofibers were imaged using 10X magnification. Scale bar 50μm (D) RT-PCR analysis of fusion and differentiation markers of Control and Myf5Cre-Pcyt2 myoblasts after 48h in differentiation media. N=5 cell cultures from 5 different animals per group. (E) Representative images and myoblast fusion index quantification of primary myoblasts with addition of vehicle (DMSO) and SA-Ro phosphatidylethanolamine binding peptide in differentiation media. Nine biological replicate myoblast cultures from three independent isolations were used. Each dot represents a calculated fusion index from in total n=9 cultures for each group. ≥ 300 nuclei were counted per one culture. Myofibers were imaged using 10X magnification. Scale bar 50μm. (F) Hypertrophic muscle growth in control and Myf5Cre-Pcyt2 mice. Following synergic ablation or sham surgery, M. plantaris weights were determined on the compensating limb. Each dot represents individual mice. Data are shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. To directly examine hypertrophic growth independently of the myoblast fusion process, we performed synergic muscle ablation^[161]20. Muscle overloading resulted in a significant enlargement of the plantaris muscle on the un-operated limb in control mice but not in Myf5Cre-Pcyt2 mice ([162]Figure 2H). Phosphorylation of P-S6K1 and 4E-BP1, downstream markers and effectors of global protein synthesis and translation, appeared to be effective ([163]Extended Data Figure 5A), indicating that impaired protein synthesis does not underlie the observed defect in hypertrophic growth of muscles from Myf5Cre-Pcyt2 mice. The observed impaired myoblast fusion and hypertrophic muscle growth in Myf5Cre-Pcyt2 mice did not affect the fiber type distribution nor fiber number in muscles from adult Myf5Cre-Pcyt2 mice ([164]Extended Data Figure 5B-[165]D). Together these data show that loss of Pcyt2 impairs long-chain fatty acid PE production and compromises both progenitor fusion and hypertrophic growth, leading to smaller myofibers and skeletal muscles. Muscles lacking Pcyt2 exhibit progressive wasting We noticed that adult Myf5Cre-Pcyt2 mice exhibited hindlimb clasping upon tail suspension ([166]Figure 3A,[167]B), indicative of muscle weakness. Indeed, muscle strength was reduced and progressively declined as Myf5Cre-Pcyt2 mice aged ([168]Figure 3C, [169]Extended Data Figure 6A). At 8 months of age all Myf5Cre-Pcyt2 mice developed kyphosis ([170]Figure 3D), which was also seen in PCYT2 disease patients ([171]Figure 3E) and has been reported in mouse models of muscular dystrophy ^[172]21. Atrophy of muscle tissue was evident in Myf5Cre-Pcyt2 mice ([173]Figure 3F) with a high incidence of central nuclei ([174]Figure 3G). Furthermore, we observed tubular aggregates and inflammation in the muscles of 12–15 months old Myf5Cre-Pcyt2 mice ([175]Figure 3H, [176]Extended Data Figure 6B-[177]D). Consequent to the observed muscle weakness that was apparent for both genders, Myf5Cre-Pcyt2 mice developed secondary osteopenia contributing to overall frailty ([178]Figure 3I) and reduced lifespan ([179]Figure 3J). Figure 3. Inactivation of Pcyt2 in mice leads to progressive weakness, muscle atrophy, inflammation and accelerated ageing. [180]Figure 3. [181]Open in a new tab (A) Representative images of 6 months old control and Myf5Cre-Pcyt2 male mice and (B) quantification of progressive worsening of hind limb clasping (B). Each dot represents one mouse, values are average of three measurements per mouse; scale bar 1 cm. (C) Age-dependent decline in grip strength in male control and Myf5Cre-Pcyt2 littermates. Each dot represents one mouse, values are average of three measurements per mouse. (D) Typical kyphosis appearance and kyphosis severity in 8 months old control (n=4) and Myf5Cre-Pcyt2 male mice (n=7). (E) Evident scoliosis (arrows) in a patient carrying the homozygous nonsense variant [182]NM_001184917.2:3c.1129C>T (p.Arg377Ter) in PCYT2. (F) Representative image and quantification of relative muscle mass changes of 12 months old versus 6 months old control and Myf5Cre-Pcyt2 male littermates. QA, quadriceps; GC, gastrocnemius; TA, tibialis anterior muscles. Scale bar = 1 cm; n=7 per group. (G) Quantification of myofibers with central nuclei in quadriceps muscles from 8 months old control and Myf5Cre-Pcyt2 male mice. n=3 mice per group. Scale bar 100μm. (H) Muscle inflammation as determined by H&E staining. Data are from 12 months old male mice. Data are representative for n=4 mice per group. Scale bar 100μm. (I) Representative cross section of tibial bone in 12 months old control and Myf5Cre-Pcyt2 male mice with quantification of tibial bone cortical thickness. Randomly assigned 5 areas from n=4 mice per group were quantified. Scale bar 100μm. (J) Survival curves for control and Myf5Cre-Pcyt2 male mice. n=22 mice per group. For statistical analysis Mantel Cox test). Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. As expected, progressive atrophy had a marked effect on whole-body metabolism. At 6 months of age, glucose clearance was increased with progressive decline in blood glucose levels and food consumption by 8 months of age ([183]Extended Data Figure 6E-[184]F, [185]Supplementary Figure 5A,[186]B) which could explain the increased mortality starting at this age. There were no apparent structural defects in respiratory muscles ([187]Supplementary Figure 6). Although Myf5Cre-Pcyt2 mice were less active ([188]Extended Data Figure 6G), energy expenditure was significantly increased in both light and dark periods ([189]Extended Data Figure 6H). Under thermoneutrality, during the light phase energy expenditure in Myf5Cre-Pcyt2 mice was comparable to Control mice, while during the dark phase the higher energy expenditure was again evident ([190]Extended Data Figure 6I). The lower activity and the muscle weakness of Myf5Cre-Pcyt2 mice were also evident under thermoneutrality ([191]Extended Data Figure 6J,[192]K). Pcyt2 is specifically required in muscle Myf5Cre is active in precursors of both skeletal muscle and brown adipose tissue (BAT)^[193]22. PE species were reduced in the BAT of Myf5Cre-Pcyt2 mice, but to a markedly lower extent than in the skeletal muscle, with an increase of PC lipids ([194]Extended Data Figure 7A,[195]B). Importantly, loss of Pcyt2 and the observed lipid remodeling did not affect in vitro differentiation of adipocyte progenitors into brown fat, thermoregulation through BAT actvity, nor levels of Ucp1 in BAT ([196]Extended Data Figure 7C-[197]F). In addition, mitochondrial ultra-structures, content and respiration appeared normal in BAT of Myf5Cre-Pcyt2 mice ([198]Extended Data Figure 7G-[199]J). Moreover, we crossed Pcyt2^flox/flox mice to the AdipoQCre line to remove Pcyt2 specifically in white and brown adipose tissue^[200]23 and did not observe any differences in growth, blood glucose, nor any apparent pathologies ([201]Extended Data Figure 8A-[202]C). Tissue-specific deletion of Pcyt2 in motor neurons (Mnx1Cre; [203]Extended Data Figure 8D-[204]F), gut epithelium (Villin1Cre; [205]Extended Data Figure 8G-[206]I), and mammary and skin epithelial cells (K14Cre, [207]Extended Data Figure 8J-[208]L), neither resulted in apparent developmental nor degenerative deffects up to 12 months of age. This lack of apparent phenotypes in the above tested Cre-deleter mouse lines suggests that Pcyt2 does not play an essential developmental role in these tissues^[209]24–[210]30. To further explore muscle-specific deletion of Pcyt2, we crossed Pcyt2^flox/flox with Mck-Cre mice. MckCre-Pcyt2 mice have been reported with a beneficial effect at young age^[211]31, which is in contrast to our findings in Myf5Cre-Pcyt2 mice and PCYT2 mutant patients. Critical membrane myotome development is established early in utero, facilitated by addition and fusion of muscle satellite cells (MSC) and adult myoblasts. Perturbations in these early events manifest as rapid onset and severe dystrophies^[212]32,[213]33. Given that McKCre is active late in muscle formation (peak Cre activity at P10), and in mature muscle without activity in early myotome, MSCs and myoblasts^[214]34,[215]35 [216]32,[217]33, this coupled with slow membrane PE turnover^[218]36 might explain observed phenotypical differences. Importantly, whereas young MckCre-Pcyt2 mice up to 4 months of age did not display degenerative phenotypes^[219]31, older MckCre-Pcyt2 mice displayed muscle weakness with a late onset at 18 months of age ([220]Extended Data Figure 8M,[221]N). Pcyt2 muscle deficiency alters mitochondrial function Given that phosphatidylethanolamines are abundant in cell but also mitochondrial membranes^[222]37, we analyzed if Pcyt2 deficiency affects mitochondrial homeostasis. Parallel to whole tissue lipidome changes, there was also a significant reduction of PE species in muscle mitochondria from Myf5Cre-Pcyt2 mice ([223]Figure 4A,[224]B). Global RNA transcriptome analyses of muscle from 10-day old control and Myf5Cre-Pcyt2 mice revealed enrichment of genes associated with mitochondrial dysfunction ([225]Figure 4C). Interestingly, when we analyzed skeletal muscle mitochondrial activity from 2 months old Myf5Cre-Pcyt2 mice, we observed an increase in activity, followed by a drop in the activity from 6 months old Myf5Cre-Pcyt2 mice ([226]Figure 4D; [227]Extended Data Figure 9A-[228]C). Further, we observed an increase in mitochondrial ROS in the isolated myofibers, as well as increased levels of antioxidant catalase activity and protein oxidative damage in the skeletal muscle of 6 months old Myf5Cre-Pcyt2 mice ([229]Figure 4E-[230]G). As expected, markers of cellular stress (pJNK, Foxo1) and muscle wasting (Atrogin, MuRF1, Fbx031) were increased in the muscles of 6 months old Myf5Cre-Pcyt2 mice ([231]Figure 4H,[232]I). The ultrastructural morphology and contents of mitochondria appeared unchanged in muscles even in adult mutant mice with apparent phenotype ([233]Extended Data Figure 9D,[234]E). Thus, increased cellular stress including energy stress, ROS mediated protein damage, and unbalanced proteostasis accelerate muscle degeneration in Myf5Cre-Pcyt2 mice. Figure 4. Pcyt2 deficiency severely affects muscle mitochondrial homeostasis as opposed to brown fat mitochondria. [235]Figure 4. [236]Open in a new tab (A) Total PE levels and (B) global lipidomics analyses of purified mitochondria isolated from 2 months old Control and Myf5Cre-Pcyt2 male mice. N=6 mice per group (C) Pathway enrichment analysis of differentially expressed genes in Control and Myf5Cre-Pcyt2 quadriceps isolated from 10 days old male pups. Evident enrichment of mitochondrial dysfunction linked genes specifically in the muscle of Myf5Cre-Pcyt2 mice. N=4 mice per group. (D) Muscle mitochondrial function assessed by measurements of complex I linked activity on isolated mitochondria from 2 months and 6 months old control and Myf5Cre-Pcyt2 male mice respectively. Paired Student t-test. (E) Measurement and quantification of myofiber mitochondrial reactive oxygen species (mtROS) in isolated myofibers (EDL muscle) from 6 months old male control (n=3 mice and 61 myofiber) and Myf5Cre-Pcyt2 mice (n=3 mice and 59 myofibers) as detected by MitoSox staining. Each dot represents relative amount of mtROS from a single myofiber. Scale bar 25μm. (F) Evidence of increased protein oxidative damage in quadriceps muscles isolated from 6 months old male Myf5Cre-Pcyt2 mice. Representative blots are shown for n=3 mice per group. (G) Catalase anti-oxidant activity in quadriceps muscles from 6 months old control and Myf5Cre-Pcyt2 male mice. (H) Increased levels of phospho-JNK (pJNK) and FoxO1 in quadriceps muscles from 6 months old male Myf5Cre-Pcyt2 mice (n=3) as compared to controls (n=3). 4 animals per group were analyzed in total, representative blot from 3 animals per group is shown (I) Increased levels of myofiber wasting markers in muscles of 8 months old male Myf5Cre-Pcyt2 mice. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. Next, we tested if the accumulation of the Pcyt2 substrate phosphoethanolamine, was affecting the mitochondria. Compared to control mitochondria, there was no additive, inhibitory effect of phosphoethenolamine on the activity of skeletal muscle mitochondria isolated from Myf5Cre-Pcyt2 mice ([237]Extended Data Figure 9F), suggesting that intra-mitochondrial overaccumulation of phosphoethanolamine is not responsible for the observed reduction in activity of skeletal muscle mitochondria from Myf5Cre-Pcyt2 mice. Given that we observed impaired skeletal muscle mitochondrial activity, we attempted to improve the pathological features of Myf5Cre-Pcyt2 mice by administering mitochondria-targeted antioxidant SS-31 daily for two months, starting from 4 months of age. Although there was a mild improvement in the grip strength and tissue weight, the effect was not significant ([238]Extended Data Figure 9G,[239]H). We also tested a potential role of Pcyt2 and Pcyt2-derived PE in sarcolemmal Ca^2+ handling and autophagy. We failed to observe any structural changes or alterations in Ca^2+ release and Ca^2+ uptake in isolated myofibers from 6-month-old Myf5Cre-Pcyt2 mice manifesting gross phenotypes ([240]Extended Data Figure 10 A-[241]C), suggesting that sarcoplasmic reticulum Ca^2+ handling was preserved. PE conjugation to ATG8 is also necessary for autophagy^[242]38. However, PE-ATG8 conjugation was comparable in both quadriceps and diaphragm muscles of 6-month-old control and Myf5Cre-Pcyt2 mice under both fed and fasted conditions, without evident accumulation of p62/SQSTM1 under fed or fasting conditions ([243]Extended Data Figure 10 D-[244]I). Of note, it has been previously shown that genetic inactivation of Atg7 and LC3 PE lipidation in Myf5-derived lineages induces brown fat over-activation, but does not lead to muscle weakness, degeneration, not muscle dystrophy^[245]39, contrasting with our Myf5Cre-Pcyt2 mice. Taken together, loss of Pcyt2 does not affect PE-modifications in autophagy or sarcoplasmic reticulum but affects mitochondrial function. However, given that mitochondria targeted therapy didn’t rescue the muscle weakness of Myf5Cre-Pcyt2 mice, this suggested that there are additional cellular defects driving the observed pathology due to Pcyt2 deficiency. Pcyt2 deficiency affects sarcolemmal lipid bilayer physicochemical properties Impaired sarcolemmal stability causes myofiber degeneration in muscular dystrophies^[246]40. Our muscle whole tissue lipidomics data showed a significant decrease in PEs containing long chain fatty acids (FAs)([247]Figure 1J,[248]K), which are abundant membrane lipids^[249]37. Therefore, we hypothesized that the reduced abundance of PEs containing long chain FAs in Myf5Cre-Pcyt2 mice might also affect formation and stability of the sarcolemmal lipid bilayer, driving the muscular pathology. To test this hypothesis, we first evaluated whether the organization of the sarcolemmal lipid bilayer was altered in Myf5Cre-Pcyt2 mice. Spectral imaging of NR12S-stained giant plasma membrane vesicles (GPMVs) provides structural information of lipid bilayers in their native compositional complexity and structural organization^[250]41. We derived the parameter of general polarization (GP), with higher values corresponding to tightly packed, rigid lipid bilayer and lower values corresponding to loosely packed, soft bilayer. Strikingly, polarization microscopy of GPMVs derived from Myf5Cre-Pcyt2 myoblasts displayed loosely packed and softer lipid bilayer as compared to control myoblasts-derived GMPVs ([251]Figure 5A,[252]B). To further address if the membrane bilayer changes persist in the myofibers, we assessed GPMVs from myofibers immediately after isolation from the tissue ([253]Figure 5C). GPMVs isolated from myofibers of Myf5Cre-Pcyt2 mice again showed significant reduction of lipid packing and soft lipid bilayer ([254]Figure 5D). Figure 5. Loss of Pcyt2 results in altered muscle membrane architectures. [255]Figure 5. [256]Open in a new tab (A) Scheme of GPMV isolation from primary myoblasts. (B) Polarization microscopy analysis of NR12S dye-stained myoblast-derived GPMVs from male control (n=71) and Myf5Cre-Pcyt2 (n=71) myoblasts. Each dot represents GP values of a single GPMV. GPMVs were derived from two independent isolations Scale bar 10μm. (C) Scheme and representative example of GPMVs (arrows) immediately after isolation from skeletal myofibers. Images are taken at 0 and 30 minutes under GPMV conditions. Scale bar 50μm. (D) Polarization microscopy of NR12S-stained GPMVs from control (n=58) and Myf5Cre-Pcyt2 (n=99) from primary myofibers (as shown in C). from two independent isolations. Representative images and quantifications are shown. Each dot represents GP values of a single GPMV. Scale bar 10μm. (E) Scheme of Brillouin light scattering microscopy for freshly isolated myofibers. (F) Surface stiffness analysis measured by Brillouin frequency shift (BFS) from isolated myofibers for male control (n=3 mice and 60 myofibers in total) and Myf5Cre-Pcyt2 (n=3 mice and 60 myofibers in total). Left panels indicate representative Brillouin images. Each data point in the right panel represents a BFS peak value of the individual myofiber surface. (G) Representative qualitative membrane stiffness data of male control and Myf5Cre-Pcyt2 myofibers assessed by atomic force microscopy. Displayed by curve angles in the approach (0 to −1000nm) and retraction phase (−1000 to 0nm), the cantilever bends less for Myf5Cre-Pcyt2 myofibers, indicating lower surface stiffness. In the prolonged part of retraction phase (0 to 400nm) the cantilever remains deeper within the Myf5Cre-Pcyt2 myofibers, indicating higher degree of surface deformity upon pressure. (H) Quantitative myofiber membrane stiffness as assessed by atomic force microscopy (Young’s modulus scale in kilopascal, kPa). For each myofiber we collected ≥4000 measurements (5μm X 5μm area). Matlab’s Randsample function was used to uniformly sample each myofiber measurements. Each dot represents 500 data points per each myofiber, from control (n=20) and Myf5Cre-Pcyt2 (n=26) myofibers. Data are shown as means ± SEM. Data are shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. To directly address how these structural-chemical changes of the membrane lipid bilayer affect mechanical properties, we employed high-resolution Brillouin Light Scattering Microscopy (BLSM) on isolated myofibers^[257]42. Scans using BLSM revealed a significant reduction in the surface stiffness of myofibers isolated from Myf5Cre-Pcyt2 mice compared to controls ([258]Figure 5E,[259]F). Atomic force microscopy on single myofibers further confirmed that Myf5Cre-Pcyt2 myofibers have a higher degree of membrane deformity after applying pressure at a nanoscale level and reduced membrane stiffness compared to Pcyt2-expressing muscle cells ([260]Figure 5G,[261]H). Thus, loss of Pcyt2 in myoblasts and myofibers results in an altered architecture of membrane lipid bilayers, directly perturbing sarcolemmal lipid bilayer mechanical properties of rigidity and stiffness. Pcyt2 deficiency impairs sarcolemmal stability An intact cell membrane architecture is critical for membrane barrier function. The sarcolemma undergoes recurrent injury via contraction mediated-mechanical strain and needs structural stability for proficient skeletal muscle function^[262]43. To determine if the perturbed architecture of the sarcolemma in Myf5Cre-Pcyt2 mice leads to altered permeability, we injected 6-month-old control and Myf5Cre-Pcyt2 mice intraperitoneally with Evans blue (<1kDA)^[263]44. We observed an extensive accumulation of Evans blue in the quadricep muscles of Myf5Cre-Pcyt2 mice relative to controls ([264]Figure 6A,[265]B). To further explore sarcolemmal stability, we induced laser mediated membrane microinjury on freshly isolated myofibers and quantified the extent of damage in real time, by measuring intracellular influx of the fluorescent fm1–43 dye^[266]45. Following laser microinjury, Control myofibers displayed minimal influx of fm1–43, whereas Myf5Cre-Pcyt2 myofibers showed increased permeability to the dye ([267]Figure 6C; [268]Video S1-[269]S4). Figure 6. Pcyt2 is essential for muscle membrane integrity and strain tolerance. [270]Figure 6. [271]Open in a new tab (A) Penetrance of Evans blue into the quadriceps muscle of 6 months old male control and Myf5Cre-Pcyt2 mice after i.p. injection. Gross morphologies and histological sections are shown. Scale bars are 1cm and 100μm. (B) Quantification of Evans blue after extraction from the muscle. n=3 per group. (C) Laser induced damage of isolated myofibers from male 6 months old control (n=9) and Myf5Cre-Pcyt2 (n=12) myofibers. The injured membrane areas of the myofibers are indicated by arrows. Right panel shows quantification of fm43 influx over the indicated time n=9–12 myofibers per group from two independent isolations. Scale bar 50μm (D) Running distance during eccentric exercise regime of 6 months old male control (n=6) and Myf5Cre-Pcyt2 (n=4) mice. (E) Representative histological analysis (H&E staining) of quadriceps muscles isolated from untrained (no training) 6 months old male control (n=4) or Myf5Cre-Pcyt2 mice (n=4) and from 6 months old control (n=6) or Myf5Cre-Pcyt2 mice (n=4) after eccentric exercise (training). Black and blue arrows show inflammation and ectopic fat deposits. Scale bars 100μm (F) Myopathy scores in 6 months old male control (n=6) and Myf5Cre-Pcyt2 (n=4) mice following eccentric exercise. The following parameters were used: inflammation, myofiber necrosis, atrophy, interstitial fibrosis, loss of membrane integrity, regenerating myofibers. Each was scored with 1–4 depending of the severity, and summed. (G) Blood muscle creatine kinase levels inferred from muscle creatine kinase activity from 6 months old sedentary and immediately after eccentric exercise of male control and Myf5Cre-Pcyt2 mice. (H) F-actin staining of skeletal muscle tissue isolated from 6 months old male control and Myf5Cre-Pcyt2 mice after eccentric exercise. Images of quadriceps cross-sections were taken using 20x magnifications. ≥100 myofibers were counted. n=3 mice per group. Scale bar 15μm. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. To directly address sarcolemmal durability to strain in vivo, we subjected control and Myf5Cre-Pcyt2 mice to eccentric exercise regime. Eccentric exercise is a potent inducer of sarcolemma strain, while having very low energy requirement compared to concentric exercise of the same load^[272]46. During the early acclimatization phase (low speed downhill running; 4 meters min^−1 for 40 min) and the intermediate phase (4 meters min^−1 for 40 min plus 9 meters min^−1 for 20 minutes) of the exercise, Myf5Cre-Pcyt2 mice performed similarly to their littermates ([273]Figure 6D). However, during the late stress phase (20 meters min^−1 for 20 minutes), Myf5Cre-Pcyt2 mice failed to complete the exercise ([274]Figure 6D; [275]Video S5). We analyzed the skeletal muscle after the last phase of training. The weights of the quadricep and gastrocnemius muscles were higher after training in control mice while the muscles of Myf5Cre-Pcyt2 mice failed to undergo hypertrophy ([276]Supplementary Figure 7A); instead, their skeletal muscles exhibited apparent muscle damage with foci of inflammation, fat cell deposits and fibrosis ([277]Figure 6E-[278]F, [279]Supplementary Figure 7B). In parallel, we observed a significant increase in blood muscle creatine levels after exercise ([280]Figure 6G). Dysferlin, a sarcolemma associated membrane repair, was aberrantly localized in Myf5Cre-Pcyt2 quadricep muscles after training ([281]Supplementary Figure 7D). In addition, we observed disorganized F-actin networks in Myf5Cre-Pcyt2 mice after training ([282]Figure 6H). Of note, the number of Pax7^+ progenitors in the skeletal muscle appeared similar between trained control and Myf5Cre-Pcyt2 mice ([283]Supplementary Figure 7E). Thus, Pcyt2 PE synthesis is required for sarcolemmal stability, preventing muscle damage and hypertrophy during eccentric exercise. Muscle-specific Pcyt2 gene therapy ameliorates muscle weakness in mutant mice Currently, there is no treatment for the disease caused by PCYT2 deficiency. Since gene therapies made significant advances in the treatment of rare diseases, we sought to therapeutically ameliorate the muscle weakness Myf5Cre-Pcyt2 mice. We cloned Pcyt2 under the control of the muscle creatine kinase 8 (CK8) promoter/enhancer into AAV6 vector cassete^[284]47,[285]48. The CK8:Pcyt2-HA vector-carrying AAV6 viral particles were injected into 4 day old Myf5Cre-Pcyt2 mice and the muscles of these mice were assessed 6 months after the treatment ([286]Figure 7A). This approach resulted in Pcyt2-HA protein overexpression and increased PE levels in the skeletal muscles of Myf5Cre-Pcyt2 AAV6:CK8:Pcyt2-HA injected mice compared to untreated Myf5Cre-Pcyt2 controls ([287]Figure 7B,[288]C, [289]Supplementary Figure 8A). Figure 7. Adenovirus based Pcyt2 gene therapy in mice is efficient for treatment of Pcyt2 deficiency-induced muscle pathology. [290]Figure 7. [291]Open in a new tab (A) Scheme of Pcyt2 muscle specific gene therapy. (B) Grip strength of male control (saline) Myf5Cre-Pcyt2 (saline) and Myf5Cre-Pcyt2 (AAV6-CK8-Pcyt2HA) mice. (C) Muscle weight isolated from 6 months old male control, Myf5Cre-Pcyt2 saline treated and Myf5Cre-Pcyt2 AAV6-Pcyt2 treated mice. Each dot represents single mice. QA, quadriceps; GC, gastrocnemius; TA, tibialis anterior muscles. Scale bars 1 cm. (D) Assessment of Pcyt2HA expression in Myf5Cre-Pcyt2 AAV6-Pcyt2 treated male mice as determined by quadriceps lysate anti-HA immunoprecipitation, followed by anti-Pcyt2 blot, 6 months after the gene delivery. (E) Total phosphatidylethanolamine levels from quadriceps of control, Myf5Cre-Pcyt2 saline treated and Myf5Cre-Pcyt2 AAV6-Pcyt2 treated male mice, 6 months after the treatment. (F) Myofiber diameter sizes from 6 months old control (n=5 mice and ≥ 60 myofibers analyzed per mouse), Myf5Cre-Pcyt2 saline treated (n=5 mice and ≥ 60 myofibers analyzed per mouse), and Myf5Cre-Pcyt2 AAV6-Pcyt2 mice (n=5 mice and ≥ 60 myofibers analyzed per mouse). Myofibers were imaged using 10X magnification. Scale bar 100μm. (G) Polarization microscopy of NR12S-stained muscle-derived GPMVs from control (n=84), Myf5Cre-Pcyt2 saline (n=93) and Myf5Cre-Pcyt2 AAV6-Pcyt2 treated (n=88) 6 months old male mice. GPMVs were derived from three independent isolations. Each dot represents values of a single GPMV. Scale bar 5μm. (H) Surface stiffness analysis by Brillouin frequency shift (BFS) from isolated myofibers of control, Myf5Cre-Pcyt2 saline and Myf5Cre-Pcyt2 AAV6-Pcyt2 treated 6 months old male mice. Each data point in the right panel represents a BFS peak value of the myofiber surface. 20 myofibers from n=4 male mice per group. (I) Muscle mitochondrial respiration of control, Myf5Cre-Pcyt2 saline and Myf5Cre-Pcyt2 AAV6-Pcyt2 treated mice as assessed by complex I linked activity on muscle lysates isolated from 6 months old male mice. N=5 mice per group. Paired Student t-test was used for statistical analysis. Dashed line indicates the average value of mitochondrial activities measured from 5 individual control male mice. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unless otherwise indicated, Multiple comparison One-Way ANOVA with Dunnett correction was used for statistical analysis unless stated otherwise. Strikingly, the AAV6:CK8:Pcyt2-HA treated mutant mice displayed a significant increase in grip strength, increased skeletal muscle mass and myofiber diameter ([292]Figure 7D-[293]F; [294]Supplementary Figure 8B). Moreover, skeletal muscles from Myf5Cre-Pcyt2 AAV6:CK8:Pcyt2-HA injected mice, displayed improved muscle membrane parameters ([295]Figure 7G,[296]H) and improved mitochondrial respiration ([297]Figure 7I). Muscle-specific Pcyt2 gene therapy improves muscle health in aging Progressive muscle atrophy is a critical determinant of frailty in aging. Muscle aging is commonly associated with diminished membrane integrity, increased susceptibility to damage, and diminished repair after exercise^[298]49–[299]51. As Myf5Cre-Pcyt2 mice displayed degenerative features that are also found in aging muscles, we assessed a potential role of Pcyt2 in muscle aging. Indeed, Pcyt2 mRNA expression and enzymatic activity in quadricep muscle were reduced in aged, pre-sarcopenic mice compared to young mice ([300]Figure 8A, [301]Supplementary Figure 8C). Importantly, PCYT2 activity and levels were substantially decreased in quadricep muscle biopsies of otherwise healthy 45–62 year-old compared to 20–30 year-old humans ([302]Figure 8B, [303]Supplementary Figure 8D,[304]E). Thus, in mice and humans, Pcyt2 expression and even more activity, decline with aging. Accoupling this, we found that in the muscles of aged, pre-sarcopenic mice, some of the most significantly affected lipids were PE species ([305]Figure 8C). To test whether increasing the levels of Pcyt2 can improve muscle function in aged mice, we aimed to rejuvenate aged muscles via overexpression of Pcyt2. AAV6:CK8:Pcyt2-HA or saline (as control) were injected retro-orbitally into 24 month-old male C57B6/J mice ([306]Figure 8D). and the expression of Pcyt2-HA and increase of total muscle PE levels in the quadricep muscle was confirmed after 2 months ([307]Figure 8E,[308]F; [309]Supplementary Figure 8F). Figure 8. Pcyt2 activity is reduced in aged muscles from humans and mice and Pcyt2 gene delivery ameliorates age-related atrophy in sarcopenic mice. [310]Figure 8. [311]Open in a new tab (A) PCYT2 activity in quadriceps from young (20–30yr) and middle aged (45–62yr) healthy human volunteers. Each dot represents individual human. (B) Pcyt2 activity in quadriceps from young (6 month) and pre-sarcopenic (24 months old) C57B6/J male mice. (C) Lipidomics analyses from quadriceps isolated from young (6 months old) and pre-sarcopenic (24 months old) C57B6/J male mice. mice. Data are shown relative to control values. CE-cholesterol ester; Cer-Ceramides; DAG-diacylglycerols; LPC-lysophosphatidylcholines; LPE-lysophosphatidylethanolamines; PC-phosphatidylcholines; PE-phosphatidylethanolamines; PG-phosphatidylglycerols; PI-phosphatidylinositols; PS-phosphatidylserines; SM-sphingomyelins; TAG-triacylglycerols. n=5 mice per group. (D) Scheme of adenovirus based, muscle-specific delivery of Pcyt2 to pre-sarcopenic 24 months old C57B6/J male mice. (E) Assessment of Pcyt2HA expression as determined by anti-HA immunoprecipitation, followed by an anti-Pcyt2 blot, from quadriceps isolated 2 months after the gene delivery. (F) Total phosphatidylethanolamine levels from quadriceps of young (6 months old), and aged (26 months old) control (saline) and AAV6-CK8-Pcyt2HA transduced C57B6/J male mice two months after AAV6 injection. (G) Polarization microscopy of NR12S-stained GPMVs isolated from aged (26 months old) control (n=63) (saline) and AAV6-CK8-Pcyt2HA transduced (n=79) C57B6/J male mice two months after AAV6 injection. GPMVs were derived from three independent isolations. Each dot represents values of a single GPMV. Scale bar 10μm. (H) Surface stiffness analysis as measured by Brillouin frequency shift (BFS) from isolated myofibers of aged (26 months old) control (saline) and AAV6-CK8-Pcyt2HA transduced C57B6/J male mice two months after AAV6 injection. Dashed line indicates the average value of Brillouin frequency shift (BFS) measured separately from five 6 month old male mice. Each data point in the right panel represents a BFS peak value of the myofiber surface. 7 myofibers from n=7 mice per group were analyzed. (I) Muscle mitochondrial function of aged (26 months old) control (saline) and AAV6-CK8-Pcyt2HA transduced C57B6/J male mice two months after the treatment as measured by complex I linked activity. Paired Student t-test was used for statistical analysis. Dashed line indicates the average value of mitochondrial activities measured separately from five 6 month old male mice. (J) Myofiber diameter sizes from aged (26 months old) control (saline) (n=5 mice and ≥ 180 myofibers analyzed per mouse.) and AAV6-CK8-Pcyt2HA transduced C57B6/J male mice (n=5 mice and ≥ 180 myofibers analyzed per mouse) two months after the AAV6 injection. Dashed line indicates the average value of myofiber diameter sizes measured separately from five 6 months old male mice (≥ 60 myofibers analyzed per mouse). Scale bar 100μm. (K) Grip strength measurements on aged (26 months old) control (saline; n=15) and AAV6-CK8-Pcyt2HA transduced (n=11) C57B6/J male mice one and two months after AAV6 injection. Repeated Measures Two-Way ANOVA with Bonferroni correction was used for statistical analysis. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. Using our newly established microscopy-based analysis pipeline, we addressed whether the observed AAV6:CK8:Pcyt2-HA dependent increase of PE lipids in the muscles of aged mice would also reflect on myofiber membrane physical parameters. Indeed, there was a significant increase in muscle membrane stiffness as measured by both polarization microscopy and Brillouin spectroscopy ([312]Figure 8G,[313]H). Moreover, the bioenergetics of the muscle showed beneficial changes, with improved mitochondrial capacity and respiratory control ratio, a measure of ATP production efficiency ([314]Figure 8I, [315]Supplementary Figure 8G). Accompanying this, there was a significant increase in the myofiber diameter of AAV6:CK8:Pcyt2-HA mice compared to control mice 2 months after gene delivery ([316]Figure 8J). Remarkably, we observed significantly improved grip strength at 1 and 2 months after gene delivery in AAV6:CK8:Pcyt2-HA mice compared to control mice ([317]Figure 8K, [318]Supplementary Figure 8H). Discussion In summary, our results uncover a critical and conserved role for Pcyt2 and Pcyt2-regulated lipid biosynthesis. We show that loss of Pcyt2-dependent lipid biosynthesis causes a previously unrealized form of muscular dystrophy, characterized by an aberrant muscle development, progressive muscle weakness and wasting, failure to thrive and shortened lifespan. Our work reveals that with Pcyt2 deficiency, reduction of long chain PE synthesis compromises the sarcolemma lipid bilayer stability, as well as myofiber mitochondria homeostasis. Given that PEs are predominantly membrane lipids, we infer that Pcyt2-dependent PE synthesis is essential for the lipid bilayer of sarcolemma and PE enriched membranes of mitochondria. This form of muscular dystrophy is very rare, in that a lipid species provides mechanical support to cellular membranes, as opposed to other forms of dystrophies that are caused by aberrations of cytoskeletal, mitochondrial or extracellular proteins^[319]52, and may thus also have distinct therapeutic implications. Whereas muscular dystrophy is typically caused by the disruption of proteins that support the sarcolemma, we show that loss of Pcyt2 leads to intrinsic changes of the membrane lipid bilayer, thus representing a unique disease mechanism. Besides being essential to sarcolemmal lipid bilayer composition and stability, PE lipids are also abundant in the inner mitochondrial membrane. Although majority of mitochondrial PEs are derived via decarboxylation of phosphatidylserine inside the mitochondria^[320]53, the observed mitochondrial dysfunction with increased generation of mtROS in myofibers of Myf5Cre-Pcyt2 mice clearly demonstrate that Pcyt2-dependent PE synthesis is additionally important in generating mitochondrial PE lipids. Decreased mitochondrial membrane viscosity coupled with impaired mitochondrial activity was already observed in degenerative diseases ^[321]54, while decreased membrane viscosity increases ROS diffusion across the bilayer ^[322]55,[323]56. This phenomenon could explain alterations of mitochondrial respiration as well as increase in ROS levels in myofibers which coupled with sarcolemmal instability, triggers muscle mitochondrial functional decline, myofiber stress, increased ROS levels, oxidative protein damage and activation of the JNK-FoxO1 axis of muscle atrophy in Myf5Cre-Pcyt2 mice. Consequently, these perturbations result in dramatic muscle degeneration, progressing to a severe dystrophy with inflammation and shortening of lifespan. Intriguingly, we and others^[324]31 have observed an initial increase of mitochondrial activity in Pcyt2 deficient muscles, without an increase in total mitochondrial numbers. This initial response most likely represents a compensatory response typical of mitochondrial diseases^[325]57. Mouse models that initiate gene loss very early in myotome development, faithfully recapitulate human muscular dystrophies in many pathological features including stunted growth and progressive muscle degeneration with shortened lifespan^[326]32,[327]33,[328]52. MckCre-Pcyt2 deficient mice did not display muscular weakness at an early adulthood ^[329]31, and we observed muscle weakness in old (20 months old) MckCre-Pcyt2 mice. Mck promoter drives Cre expression late in the muscle development (peak activity P10)^[330]34. However, critical muscle membrane organization occurs very early in the developing myotome^[331]52. Thus, relatively later Cre activation driven by the Mck promoter would miss this critical developmental stage, which is not the case for inherited, disease-causing mutations in humans. Moreover, given that MckCre activity is restricted to mature muscles^[332]34,[333]35, this bypasses the muscle stem cell pool rendering them “wild type”, which can then actively repopulate and repair the Pcyt2 deficient muscle, and delay muscle degeneration. This phenomena of dependence of disease severity on the timing of the gene disruption was already observed for several animal models of membrane-related human dystrophies, where only mouse mutants with early gene deletion faithfully recapitulated patient phenotypes^[334]32,[335]33. Our findings in zebrafish model, mouse mutant and aged models, as well as in rare disease patients with weakness and muscle wasting ^[336]58 support that the Pcyt2-dependent Kennedy pathway is essential for muscle health. As seen in rare disease patients, failure to thrive preceded the apparent progressive weakness and degeneration in Myf5Cre-Pcyt2 mice. Muscle growth by myoblast fusion and hypertrophy is essential for both muscle and whole-body size ^[337]59. Both mechanisms of muscle growth were affected in Myf5Cre-Pcyt2 mice. Apart from being important membrane building blocks, phosphatidylethanolamines are important in modulating membrane physicochemical properties^[338]4. Due to their relatively small polar head group and consequently a conical shape^[339]4, phosphatidylethanolamines form a negative membrane curvature required for membrane bending during the cell fusion process^[340]60. Therefore, any genetic insufficiency of phosphatidylethanolamines would negatively affect the efficiency of both membrane neogenesis during tissue growth and cellular fusion, affecting tissue growth. However, solely stunted growth or impaired myoblast fusion wouldn’t necessarily result in degenerative phenotypes later in adulthood ^[341]61–[342]64. Our findings indicate that the muscle tissue is especially vulnerable to loss of Pcyt2 and Pcyt2 synthesized PE. It is well established that distinct tissues have a diverse membrane lipid composition^[343]4 and may be differentially dependent on Pcyt2. Indeed, mining the Achilles Depmap data portal, which contains gene essentiality scores from 769 cell lines^[344]65, we found that Pcyt2 is not essential for a large majority of the tested cell lines (4.8% dependent cell lines). For comparison, many more cell lines (54% dependent cell lines) are dependent on choline-phosphate cytidylyltransferase (Pcyt1a), a bottleneck enzyme for synthesis of phosphatidylcholines in the parallel branch of the Kennedy pathway. The muscle dependency on Pcyt2 derived PE might be explained by the general chemical properties of PE lipids. Increasing PE concentrations increase the viscosity of the liposomes^[345]66, therefore we hypothesize that the constant mechanical strain and contraction of the myofibers render muscle membranes dependent on PE for mechanical support. The essential dependency of myofibers on Pcyt2 derived PE compared to other cell types, is supported by our findings from various tissue-specific mutants, as well on BAT functionality in Myf5Cre-Pcyt2 mutants. Moreover, current pathophysiological symptoms in the rare disease patients are mainly restricted to growth and neuromuscular parameters, with no other reported physiological defects such as core body temperature maintenance. This indicates that other cell types are able to engage alternative molecular mechanisms to compensate for the deficiency in Pcyt2-dependent PE synthesis. Future research should illuminate which synthesis or lipid uptake mechanisms from periphery are responsible for this. Interestingly, inherited mutation in choline kinase beta (Chkb) results in loss of synthesis of another membrane lipid, phosphocholine and consequently phosphatydilcholine, causing neuronal pathology as well as muscular dystrophy without affecting other tissues^[346]67,[347]68, thus resembling certain pathological features of inherited PCYT2 mutations. It is intriguing that nervous and muscle tissue are particularly vulnerable to deficiency of certain membrane lipids as opposed to other cell types stimulating a wider field of research on cell type-specific changes in cell membrane lipid composition and lipid bilayer physicochemical parameters in various biological processes and pathological conditions. It is important to note that humans the effect of the mutations in either EPT1 or PCYT2 vary between affected individuals ([348]Supplementary Figure 1, [349]Supplementary Figure 2). In some patients where the muscle parameters were addressed the effect on muscle strength was apparent with weakness and muscle wasting while other patients showed no signs of reduced muscle strength^[350]58. Muscle biopsy on one clinical case confirmed the reduction in muscle fiber size^[351]69. Gross muscle strength evaluation revealed that all of our reported EPT1 and PCYT2 mutation patients have muscle weakness. This variation in muscle strength parameters might be explained by the effects of the different mutations on the EPT1 and PCYT2 enzymatic activity resulting in pathogenic thresholds of PE concentrations. A future comparative study of all reported individuals, including detailed characterization of muscle biopsies and determination of enzymatic activities of EPT1 and PCYT2, should provide further insights on the severity of the muscle phenotypes in these patients in regard to the corresponding enzymatic efficiency. Muscle atrophy is a hallmark of aging, and a leading cause of frailty and dependency. We found that Pcyt2 levels and activity markedly declined in muscles from aged rodents and humans. Decreased expression of Pcyt2 mRNA was recently observed in aged rat muscles^[352]70. It is possible that this reduction occurs as a consequence of a metabolic switch in aged muscle, which in aging appears to be more directed towards triglyceride and cholesterol synthesis^[353]71. Indeed, low density lipoprotein, cholesterol oxysterols or LXR (liver X receptor, a transcriptional regulator of cholesterol, fatty acid, and glucose homeostasis) inhibit Pcyt2 ^[354]72. Importantly, our data show that PCYT2 enzymatic activity are significantly reduced in aged muscles, with long chain PE membrane lipids being one of the most significantly reduced lipid species in aged muscles. Moreover, we found that increasing Pcyt2 expression in aged mice improves several parameters of aged myofibers through increasing tissue PE lipid levels, ameliorating muscle strength decline. Taken all this together, Pcyt2 upregulation could be considered as a potential treatment to improve muscle frailty. Methods Studies in Humans Patients Patients with PCYT2 and EPT1 deficiency were identified previously^[355]6,[356]8. Their height and weight were recorded at visits to the hospital. The spinal MRI was performed at the age of 19 years. All data presented are being shared with parental and patient consent under the ethical approval according to institutional and international guidelines for studies with human subjects. Human biopsies All human experiments were approved by the regional ethical review board in Stockholm (2014/516–31/2 and 2010/786–31/3) and complied with the Declaration of Helsinki. Oral and written informed consent were obtained from all subjects prior to participation in the study. 8 healthy young adults (age 21–29) and 8 middle-aged (age 45–62) subjects were recruited. The subjects did not use any medications and were nonsmokers. Biopsies of the quadriceps vastus lateralis muscle were obtained under local anesthesia using the Bergström percutaneous needle biopsy technique and immediately frozen in isopentane, cooled in liquid nitrogen, and stored at −80°C until further analysis. Studies in zebrafish Generation of mutant zebrafish and analysis Zebrafish were raised and maintained at the biological services facility at the University of Manchester under standard conditions. Pcyt2 mutant zebrafish at F0 have been described previously ^[357]6. For histological examination, animals were sacrificed by lethal anesthesia with buffered tricaine methanesulfonate. After gross examinations, whole body was fixed in 4% paraformaldehyde (PFA) for 72 hr. After fixation, 0.5 mm pieces were cut and embedded in paraffin blocks. 3μm sections were further processed for routine hematoxylin and eosin staining. Back muscle cross sectional areas of the same anatomical region were imaged under 10X magnification, followed by analysis with ImageJ software. All animal experiments were approved by the Animal Care and Use Committee at the University of Manchester. Studies in mice Mouse lines and diets All mice were housed in the IMBA mouse colony with a 12 h light/dark cycle in a temperature-controlled environment and fed a standard chow diet. Pcyt2 conditional mice have been described previously^[358]73. In all cases, all mice described in our experiments were littermates, matched for age and sex. Tissue specific Pcyt2 mutant mice were generated by crossing of Pcyt2^flox/flox mice with Cre transgenic mice. Villin Cre mice originate from Institut Curie (Sylvie Robine Lab). The following mouse lines were obtained from the Jackson Laboratory (Jackson Lab, Bar Harbor, US): Adipoq Cre (B6;FVB-Tg(Adipoq-cre)1Evdr/J); Alb Cre (B6.Cg-Speer6-ps1Tg(Alb-cre)21Mgn/J); Mck Cre (B6.FVB(129S4)-Tg(Ckmm-cre)5Khn/J. All animal experiments were approved by the Animal Care and Use Committee at IMBA, University of Physical Education in Budapest, Institut NeuroMyoGène and Institut de Recherche en Santé Digestive in Toulouse. Functional in vivo muscle tests Grip strength: Two-, 4- and 6-month-old mice (control and Pcyt2 Myf5 KO mice) were subjected to grip strength tests using a grip strength meter (Bioseb, USA). Prior to tests, mice were single caged for two weeks. Clasping index was evaluated as described previously^[359]74. Each mouse was scored three times, and an average of scores was calculated. Eccentric exercise: Single caged mice were acclimatized for treadmill exercise for three days on low speed (4m min^−1 for 40 min per day), followed by a 7-day training on a medium speed (4m min^−1 for 40 min plus 9m min^−1 for 20 min per day), and a 2-day stress training (20m min-1 for 20min per day). Immediately after the completion of the exercise, mice were sacrificed, and muscles were collected for histological analysis. Synergic muscle ablation: For synergic ablation experiments, all surgical procedures were performed under aseptic conditions with the animals deeply anaesthetized with pentobarbital sodium (60 mg/kg i.p.). Compensatory overload of the plantaris muscle was performed unilaterally via removal of the major synergistic muscles (gastrocnemius-medialis, -lateralis and soleus). A sham operation was systematically performed on the control hindlimb, which consisted of separating tendons of the soleus and gastrocnemius muscles from the plantaris muscle. Analgesic was administered to the animals for two days following the operation. The overload lasted for 14 days. For maintaining the activity of the animals during the overload period, moderate speed walking training was used on a treadmill (10 degrees ascents, 4–5 m/min, 30 min, 6 times/week). At the end of this period, after animal sacrifice, the plantaris muscle was removed bilaterally, trimmed of excess fat and connective tissue, weighed and processed for further analysis. AAV-based vector delivery: For the AAV6 treatment, Pcyt2 ([360]NM_024229.3) was C-terminally tagged with an HA-tag and cloned into the AAV6-CK8 muscle specific expression vector ^[361]47 using the Sal1-Kpn1 restriction sites. AAV6 viral particles were prepared as previously described ^[362]75. For gene therapy of Myf5Cre-Pcyt2 mutant mice, 4 days old pups were injected i.p. with 2×10^11 vector genomes per mouse. For ageing studies 24-month-old C57B6/J mice (Jackson Labs, Bar Harbor) were single caged for two weeks for acclimatization. Grip strength was measured before AAV6 delivery using the grip strength meter. On the day of the AAV injection, mice were anesthetized with isoflurane, and injected retro-orbitally either with AAV6-Pcyt2HA (5×10^12 vector genomes per mouse) or as a control saline. Expression of the Pcyt2-HA was determined by Western blotting and grip strength was measured one and two months after the injection. Metabolic studies: Animals were fed standard chow diet and blood glucose was measured at fed and fasted state (16h fasting). Standard chow diet was purchased from SSNIFF (V 1184–300; 10mm pellets autoclavable; 49% kcal carbohydrates, 36% kcal protein and 15% kcal fat). Measurements were done at the same time of the day by using Onetouch Verio strips (LifeScan, GmbH) after tail snipping. Food consumption was measured on single cage housed animals over a period of two weeks, after one week acclimatization. For calorimetry, measurements were performed at room temperature (21C-23C) on a 12/12 h light/dark cycle in a PhenoMaster System (TSEsystems, Bad Homburg, Germany) using an open circuit calorimetry system. Mice (4–7 months old) were housed individually, trained on drinking nozzles for 7 days and allowed to adapt to the PhenoMaster chamber for 2 days. Food and water were provided ad libitum. Parameters of indirect calorimetry and activity were determined for 5 consecutive days. Body weights were recorded at the beginning and end of the experiments and average values were plotted against energy expenditure and activity. To address brown fat activity, 6-month-old mice were housed at 4°C and body temperature was measured during the fed period using a thermometer (Thermometer TK 98802; Bioseb). Temperature was recorded every hour over a 4 h period. After a 2 days recovery period at room temperature, the same mice were fasted and the body temperatures determined at 6 and 16 hours of fasting. For autophagy induction, animals were fasted for 24 hours. Isolation and imaging of Giant Plasma Membrane Vesicles (GPMVs) GPMVs were prepared as previously described ^[363]76. Briefly, myoblasts were seeded on a 60 mm petri dish until ~ 70% confluency. Before GPMV formation, they were washed twice with GPMV buffer (150 mM NaCl, 10 mM Hepes, 2 mM CaCl2, pH 7.4) and finally 2 ml of GPMV buffer with 25 mM PFA and 10 mM DTT (final concentrations) was added. After incubation for 2hr at 37^oC, GPMVs were collected from the supernatant. For GMPV preparation from myofibers, the extensor digitorum longus (EDL) muscle was digested in collagenase supplemented medium (type1, 2mg/mL) for 2.5hr at 37^oC, followed by single myofiber isolation. Myofibers were then gently washed twice with FCS free DMEM, followed by a brief 1min wash with GPMV buffer containing 25 mM PFA, to prevent myofiber hypercontraction. Finally, GPMV buffer with 25 mM PFA and 10 mM DTT was added to the myofibers. After incubation for 2hr at 37^oC, GPMVs were collected from the supernatant. GMPVs were labelled with the polarity sensitive membrane probe NR12S (a kind gift from A. Klymchenko) at 0.1μg/ml final probe concentration in phosphate buffered saline (PBS) for 5 minutes and then imaged on a Zeiss LSM 780 confocal microscope equipped with a 32-channel GaAsP detector array. Laser light at 488 nm was used for fluorescence excitation of NR12S. The lambda detection range was set between 490 – 691 nm. Images were analyzed by using a custom plug-in compatible with Fiji/ImageJ3 to measure generalized polarization which reflects membrane lipid packing/order using the following formula where I[560] and I[650] are the fluorescence intensities at 560 nm and 650 nm respectively: [MATH: GP=I560−I650I560+I650 :MATH] Brillouin Microscopy of myofibers Brillouin Light Scattering Microscopy (BLSM) was performed using an inverted confocal sample-scanning microscope with a Brillouin imaging-spectrometer as described previously ^[364]42,[365]77. Briefly the setup employed a 532nm single-frequency probing laser and is based on a 2-stage cross dispersion Virtual Imaged Phase Array (VIPA) with intermediate Fourier and image plane filtering, a cooled EM-CCD camera (Hamamatsu ImagEMII) for detection, with a spectral finesse >85. For more scattering samples, we also employed a heated Iodine absorption cell in the detection path tuned to the laser wavelength to reduce the elastic scattering signal. To generate spatial maps, samples were scanned with a 3-axis piezo electric stage (Physik Instrumente). Imaging was performed through 1.3 Numerical Aperture (NA) Si immersion-oil objective (Olympus) and confocality was assured via a physical pinhole of ~1 Airy Unit (AU) before coupling the light into the spectrometer. Widefield transmitted light images were used to determine the scanning area for each sample. Several cross-sectional scans were performed for each myofiber at positions separated by a 1μm. Acquisition was controlled by a custom Labview based script developed by the company THATec. The acquisition (dwell) time per voxel was 100ms and the power measured at the sample was 2–3mW. Each measured spectrum was de-convolved with the complete system spectral response as determined for the attenuated elastic scattering peak measured prior to each scan in the same sample. Prior and subsequent to each imaging session the spectra of water and ethanol were measured on a separate imaging arm and used for calibrating the spectral projection based on paraxial theory ^[366]78, adapted to a dual VIPA setup. Each spectrum was then analyzed in Matlab (Mathworks) using custom developed scripts employing Spectral Phasor Analysis. Results were confirmed to be in agreement with ones obtained from conventional non-linear least-squares fitting of deconvolved spectra consisting of the Stokes and Anti-Stokes Brillouin peaks using Lorentzian functions. The extracted Brillouin peak frequency, which scales with the local elastic storage modulus, is taken for local stiffness. All measurements were performed at 37°C and 5% CO2. Atomic force microscopy Force spectroscopy by means of Atomic force microscopy was done on individual myofibers after isolation from EDL muscles. Fibers were either cultured on matrigel coated dishes (QI mode) or probed immediately after isolation (Force spectroscopy). A JPK (Bruker) Nanowizard4 AFM atomic force microscope was used for the AFM experiments. The QP-BioAC cantilevers from Nanosensors (0.06N/m, less than 10 nm nominal tip radius) were used because of their ability to work in QI mode with biological samples. The approach and retract speeds were kept constant at 52microns/s. The model used to obtain the Young’s modulus from the acquired data was Hertz/Sneddon. The paraboloid model was chosen as the most suitable for the sharp tips of the QP-BioAC cantilevers. For the fits forces up to 60pN were considered, where indentation depths remained below 500nm. Each of the fibers retrieved ≥ 4000 values (5μm X 5μm sampled in 64×64 data pixels). Matlab’s Randsample function was used to uniformly sample 500 data points. For the force curves acquired for the qualitative comparison, forces up to 600pN were applied to a single location each, using same cantilevers and approach speeds. Samples were kept at a 37°C during the experiment. Laser-induced damage of myofiber cell membranes Myofibers were freshly isolated from EDL muscle of 4-month old mice by digestion in collagenase supplemented media (type1, 2mg/ml, Sigma) for 2 hr at 37°C. After digestion, individual fibers were transferred onto a 4-well chamber slide (Nunc Lab-Tek, Merck, GmbH) containing HBSS with 2.5μM FM1–43 dye (Molecular Probes, Invitrogen, ThermoFisher Scientific, GmbH). Laser-induced cell membrane damage was performed as previously described ^[367]45. Briefly, a 5 × 5 pixel area of the plasma membrane was exposed to a Laser at 20% of maximum power (Enterprise, 80 mW, 351/364 nm) for 5 s using a Zeiss-LSM 510 confocal microscope equipped with a ×63 water immersion lens (N.A. 1.3). Following the laser damage, distribution of the FM1–43 dye was imaged using high speed video captures. Throughout the experiment, cells were kept at a 37°C in a 5% CO[2] chamber. Ca^2+ dynamics under voltage-clamp protocol The flexor digitorum brevis (FDB) and interosseous (IO) muscles were freshly isolated from mice and incubated with collagenase (Sigma type 1) for 1 hour at 37 °C in a Tyrode solution. Single myofibers were isolated by gentle mechanical trituration of the collagenase-treated muscles within the experimental chamber. Fibers were voltage-clamped using the silicone-voltage-clamp technique as described^[368]79 with the voltage-clamp pipette filled with a solution containing (in mM) 140 K-glutamate, 5 Na2-ATP, 5 Na2-phosphocreatine, 5.5 MgCl2, 5 glucose, 5 HEPES and either 15 EGTA, 6 CaCl2, and 0.1 rhod-2 or 0.1 fluo-4FF. The extracellular solution contained (in mM) 140 TEA-methane-sulfonate, 2.5 CaCl2, 2 MgCl2, 1 4-aminopyridine, 10 HEPES and 0.002 tetrodotoxin. For fluo-4FF fluorescence measurements, the extracellular solution also contained 0.05 mM N-benzyl-p-toluene sulfonamide (BTS) to block contraction. All solutions were at pH 7.2. Voltage-clamp and membrane current measurements were done with an RK-400 patch-clamp amplifier (Bio-Logic, Claix, France) in whole-cell voltage-clamp configuration. Command voltage pulse generation was achieved with an analog-digital converter (Digidata 1440A, Axon Instruments, Foster City, CA) controlled by pClamp 9 software (Axon Instruments). Holding voltage was always set to −80 mV. Following insertion of the micropipette extremity into the muscle fiber, a 30 min-long period of equilibration was allowed before taking measurements. Rhod-2 and fluo-4-FF fluorescence were detected with a Zeiss LSM 5 Exciter confocal microscope equipped with a 63× oil immersion objective (numerical aperture 1.4). For detection of rhod-2 and fluo-4 FF fluorescence, excitation was from the 543 nm line of a HeNe laser and from the 488 nm line of an Argon laser, respectively, and fluorescence was collected above 560 nm and above 505 nm, respectively. Both rhod-2 and fluo-4 FF voltage-activated fluorescence changes were imaged using the line-scan mode (x,t) of the system and expressed as F/F0 where F0 is the baseline fluorescence. Rhod-2 Ca2+ transients were triggered by 0.5 s-long depolarizing pulses of increasing amplitude. From these, the rate of SR calcium release (d[CaTot]/dt) was calculated as described^[369]80 Fluo-4-FF was used under non-EGTA buffering intracellular conditions to assess the resistance of the fibers to a fatigue protocol. Fibers were stimulated by consecutive trains of thirty 5 ms-long pulses from −80 mV to +60 mV delivered at 100 Hz: 40 trains were applied, separated by a 0.7 s interval. In vivo muscle permeability measurements To determine in vivo muscle tissue permeability, 1% Evans blue dye (10 ml kg^−1 body weight, Sigma, GmbH) was injected into the intraperitoneal cavity of 8 month old animals. 16hours later mice were sacrificed via cervical dislocation and muscles collected. Quadriceps muscles were weighed and then soaked in formamide (GibcoBRL, UK) for 48 hr at 55°C with gentle shaking. The optical density of Evans blue in the resulting supernatant was measured at 610 nm with a Spectronic 610 spectrophotometer (Milton Roy). To image Evans blue via microscopy, muscles were immediately embedded in OCT and frozen. 10μm longitudinal sections were cut and visualized after counterstaining with DAPI. Pcyt2 enzyme activity Frozen muscle tissue (50mg) was homogenized in a cold lysis buffer (10 mM Tris-HCl [pH 7.4], 1 mM EDTA, and 10 mM NaF) and briefly centrifuged to remove cell debris. Fifty ug of protein was assayed with 0.2 μCi of [14C]-phosphoethanolamine (P-Etn) (American Radiolabeled Chemical) in 50μl of reaction mixture of 50 mM MgCl2, 50 mM DTT, 10 mM unlabeled P-Etn, 20 mM CTP and 100 mM Tris-HCl (pH 7.8). The reaction was incubated at 37°C for 15 min and terminated by boiling (2 min). The reaction product [14C]CDP-Etn was separated from [14-C]PEtn on Silica gel G plates (Analtech) with a solvent system of methanol:0.5%NaCl:ammonia (10:10:1). CDP-Etn and P-Etn in standards and samples were visualized with a 0.5% ninhydrin in ethanol and the [14C]CDP-Etn collected and quantified by liquid-scintillation counting. Pcyt2 activity was expressed in nmol/min/mg protein. Protein content was determined with bicinchronic acid assay from Pierce. Further methods can be found in [370]Supplementary Information. Statistical analysis of mouse studies All mouse data are expressed as mean +/− standard error of the mean (SEM). All figures and mouse statistical analyses were generated using Prism 8 (GraphPad) or R. Details of the statistical tests used are stated in the figure legends. In all figures, statistical significance is represented as *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001. Extended Data Extended Data Fig. 1. [371]Extended Data Fig. 1 [372]Open in a new tab PE synthesis pathways and EPT1 rare disease mutation carriers (A) Schematic diagram of phosphatidylcholines (PC), phosphatidylethanolamines (PE) and phosphatidylserine (PS) phospholipids synthesis. EK-Ethanolamine kinase; PCYT2-CTP:phosphoethanolamine cytidylyltransferase; EPT1-ethanolaminephosphotransferase 1; PSS2-Phosphatidylserine Synthase 2; PSD-Phosphatidylserine decarboxylase; CK-Choline kinase; PCYT1-Choline-phosphate cytidylyltransferase; CEPT1-Choline/ethanolaminephosphotransferase 1; PSS1-Phosphatidylserine Synthase 1. (B) Height and weight gains of three patients (#1 male, #2 female, #3 male) carrying the homozygous missense variant c.335 G>C (p.Arg112Pro) in the EPT1 gene. Controls indicate WHO standards of median weights and heights at the respective ages +/− 2 standard deviations (SD). Extended Data Fig. 2. [373]Extended Data Fig. 2 [374]Open in a new tab Analysis of Pcyt2 deletion in mice. (A) Schematic diagram of exon 2 deletion in Myf5Cre-Pcyt2 male mice and confirmation by RNA sequencing. Exon and introns structures as well as LoxP sites targeted to exon 2 and loss of exon 2 upon Cre-mediated recombination are shown for the murine Pcyt2 locus. n=3 animals per group. Extended Data Fig. 3. [375]Extended Data Fig. 3 [376]Open in a new tab Characterization of Myf5Cre-Pcyt2 mice. (A) Body weights of control and Myf5Cre-Pcyt2 male mice at P1 and P4. (B) Body length gains of control and Myf5Cre-Pcyt2 male mice. n=6 per group for body length analysis. (C) Body weights of 2 months old control and Myf5Cre-Pcyt2 female mice. (D) Body lengths of 2 months old control and Myf5Cre-Pcyt2 female mice. (E) Skeletal muscle and tissue weight isolated from (E) 10 day old control (n=6) and Myf5Cre-Pcyt2 (n=8) and (F) 2 months old (P56) control (n=8) and Myf5Cre-Pcyt2 (n=7) littermate male mice. QA, quadriceps; GC, gastrocnemius; TA, tibialis anterior muscles. Liver and spleen weights are shown as controls. Scale bars 1 cm. (G-H) Gross skeletal muscle appearance of 56 days old control and Myf5Cre-Pcyt2 male littermates. (I) mRNA TPM levels of enzymes from the PE and PC branch of the phospholipid synthesis Kennedy pathway. n=3 mice per group. Data are shown as means ± SEM. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant (unpaired Student t-test). Extended Data Fig. 4. [377]Extended Data Fig. 4 [378]Open in a new tab Myoblast proliferation assessment in Myf5Cre-Pcyt2 mice. (A) Representative images and quantification of BrdU labeled quadriceps from 2 days old control and Myf5Cre-Pcyt2 male mice. Images were taken under 5x magnification, and ≥2000nuclei were counted and analyzed. N=4 animals per group. Scale bar 60μm. (B) Representative images and quantification of EdU labeled primary myoblasts in cell culture isolated from control and Myf5Cre-Pcyt2 male mice. 18 biological replicate cultures from 3 independent isolations were analyzed and images were taken under 5x magnification. ≥100 nuclei counted per each culture. Each dot represents the number of EdU positive (B) cells per each culture. Scale bar 50μm. (C) Number of Pax7 positive nuclei in quadriceps from 6 months old male control (n=5) and Myf5Cre-Pcyt2 mice (n=4). ≥100 nuclei per each individual section from each mouse were counted. Scale bar 50μm. Data are shown as means ± SEM. Unpaired Student t-test with Welch correction was used for statistical analysis. Each dot represents the number of EdU positive cells per each culture. Scale bar 50μm. (C) Number of Pax7 positive nuclei in quadriceps from 6 months old control (n=5) and Myf5Cre-Pcyt2 mice (n=4). ≥100 nuclei per each individual section from each mouse were counted. Scale bar 50μm. Data are shown as means ± SEM. Unpaired Student t-test with Welch correction was used for statistical analysis. Extended Data Fig. 5. [379]Extended Data Fig. 5 [380]Open in a new tab Myofiber type distribution in skeletal muscle of Myf5Cre-Pcyt2 mice (A) Western blot analysis of critical regulators of protein synthesis and translation S6K1 and 4E-BP1 in overloaded M. plantaris from Control and Myf5Cre-Pcyt2 male mice. Each lane represents individual mice. Two-Way ANOVA with multiple comparison followed by Bonferroni correction was used for statistical analysis. (B) Representative images and quantification of MyHC!, MyhCIIA and MyHCIIB fibers in skeletal muscle (quadriceps) from 6 months old control and Myf5Cre-Pcyt2 male mice. Images were taken under 5x magnification, and ≥100 myofibers were counted at 3 different matching histological areas. N=4 animals per group. Scale bar 500μm. (C) Representative images and quantification of oxidative and glycolytic fibers in skeletal muscle (quadriceps) from 6 months old control and Myf5Cre-Pcyt2 male mice. Images were taken under 10x magnification, and ≥1000 myofibers were counted at matching histological areas. N=5 animals per group. Scale bar 500μm. (D) Total number of fibers in skeletal muscle (quadriceps) from 6 months old control and Myf5Cre-Pcyt2 male mice. Images were taken under 2.5x magnification. N=5 animals per group. Scale bar 500μm. Data are shown as means ± SEM. Unless otherwise stated, unpaired Student t-test with Welch correction was used for statistical analysis. Extended Data Fig. 6. [381]Extended Data Fig. 6 [382]Open in a new tab Muscle inflammation and metabolic assessment of Myf5Cre-Pcyt2 mice. (A) Grip strength of 6 months old control and Myf5Cre-Pcyt2 females. Each dot represents one mouse, values are average of three measurements per mouse. (B) Representative electron microscopy images of quadriceps of 15 months old control and Myf5Cre-Pcyt2 male mice. Note accumulation of tubular aggregates in the mutant animals (red arrows). Representative images of 3 animals per group are shown. Scale bar 2μm. (C) Characterization of muscle inflammation in 12 months old Myf5Cre-Pcyt2 male mice. Helper T cells (CD4+) and cytotoxic T cells (CD8+) are shown. Scale bar 100μm for H&E stained and 50 μm for immune cell staining. Representative staining of 3 animals per group are shown. (D) Inflammatory cytokine levels in the quadriceps of 12 months old Myf5Cre-Pcyt2 male mice. (E) Fed blood glucose levels on normal chow diet of 8 months old control and Myf5Cre-Pcyt2 male mice. (F) Food consumption analysis of 6 month and 8 months old control and Myf5Cre-Pcyt2 mice (G) Cage activity of 6 months old male control and Myf5Cre-Pcyt2 mice (n=12 per group). Multiple ANOVA was used to analyze the data. (H) Energy expenditure of 6 months old male control and Myf5Cre-Pcyt2 mice during the resting (light) and active (dark) phases. (I) Cage activity under thermoneutrality of 6 months old male control and Myf5Cre-Pcyt2 mice (n=6 per group). Multiple ANOVA was used to analyze the data. (J) Energy expenditure of 6 months old male control and Myf5Cre-Pcyt2 mice during the resting (light) and active (dark) phases under thermoneutrality. (K) Grip strength assessment of 6 months old male control and Myf5Cre-Pcyt2 mice under thermoneutrality. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant. Unpaired Student t-test with Welch correction was used for statistical analysis unless stated otherwise. Extended Data Fig. 7. [383]Extended Data Fig. 7 [384]Open in a new tab Characterization of the brown adipose tissue from Myf5Cre-Pcyt2 mice. (A-B) Lipidomics analyses from brown fat isolated from 10-day old Myf5Cre -Pcyt2 and control male mice. n=4 per group. (C) Brown fat differentiation in lipid free conditions from 2-day old primary pre-adipocytes isolated from control and Myf5Cre -Pcyt2 male mice. Scale bar 50μm. (D-E) Brown fat activity as addressed by exposure of 6-month-old control and Myf5Cre-Pcyt2 male mice to cold (4C) or during fasting. (F) Ucp1 mRNA levels in brown fat of 6-month-old control and Myf5Cre-Pcyt2 male mice. (G) Mitochondrial content in brown adipose tissue. (H) BAT mitochondrial structure of 6-month-old Myf5Cre-Pcyt2 male mice. Representative images of 3 animals per group are shown. Scale bar 1μm. (I-J) Complex I and II activities of brown fat mitochondria. Paired Student t-test was used to analyze the data. Data are shown as means ± SEM. Each dot represents data point from individual mice unless stated otherwise. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant (unpaired Student t-test, unless otherwise stated) Extended Data Fig. 8. [385]Extended Data Fig. 8 [386]Open in a new tab Specific inactivation of Pcyt2 in multiple mouse tissues. (A) Schematic diagram to generate adipose tissue specific Pcyt2 deficient male mice (AdipoQCre-Pcyt2). (B) Body weights and appearances of 6 months old control and AdipoQCre-Pcyt2 male mice. (C) Fasting blood glucose of 6 months old control and AdipoQCre-Pcyt2 male littermates fed a chow diet. (D) Schematic diagram of motor neuron specific Pcyt2 deficient male mice (Mnx1Cre-Pcyt2). (E) Body weights of 8 months old control and Mnx1Cre-Pcyt2 male mice. (F) Absence of any overt clasping behavior and appearance in 8 months old Mnx1Cre-Pcyt2 male mice. (G) Schematic diagram of intestine epithelium specific Pcyt2 deficient male mice (VilinCre-Pcyt2). (H) Body weights of 6 months old control and VilinCre-Pcyt2 littermates. (I) Histological sections of intestine isolated from 12 months old control and VilinCre-Pcyt2 male mice. Scale bar 100 m. (J) Schematic diagram of skin epithelium Pcyt2 deficient male mice (K14Cre-Pcyt2). Representative images of 3 animals per group are shown. (K) Body weights and appearances of 6 months old control and K14Cre-Pcyt2 male mice. (L) Histological sections of skin isolated from 12 months old control and K14Cre-Pcyt2 male littermates. Representative images of 3 animals per group are shown. Scale bar 100 m. (M) Schematic diagram of mature muscle specific Pcyt2 deficient male mice (MCKCre-Pcyt2). (N) Grip strength of 18 months old control and muscle specific MckCre-Pcyt2 male mice. Data are shown as means ± SEM. Each dot represents individual mice, each mouse was tested in triplicates. Mean values ± SEM are displayed. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, n.s. not significant (unpaired Student t-test). Extended Data Fig. 9. [387]Extended Data Fig. 9 [388]Open in a new tab Assessment of mitochondrial homeostasis and SS-31 treatment. (A-B) Muscle mitochondrial function assessed by measurements of complex II linked activity on isolated mitochondria from 2 months and from (C) 6 months old control and Myf5Cre-Pcyt2 male mice respectively. Paired Student t-test was used for statistical analysis. (D-E) Ultrastructure and total numbers of muscle mitochondria from 8 months -old control (n=6) and Myf5Cre-Pcyt2 male mice (n=6). Scale bar 200nm. Unpaired Student t-test with Welch correction was used for statistical analysis (F) Function of muscle mitochondria under increasing concentrations of phosphoethanolamine, as assessed by measurements of complex I linked activity on isolated mitochondria from 2 months old control and Myf5Cre-Pcyt2 male mice. N=3 mice per group. Two-Way ANOVA with multiple comparison followed by Bonferroni correction was used for statistical analysis. (G-H) Grip strength and organ weight measurements of 6 months old control (vehicle) and Myf5Cre-Pcyt2 male mice that have been treated with either vehicle or ss-31 compound for two months. Data are shown as means ± SEM. Multiple comparison One-Way ANOVA with Dunnett correction was used for statistical analysis. Extended Data Fig. 10. [389]Extended Data Fig. 10 [390]Open in a new tab Assessment of calcium handling, and autophagy markers in skeletal muscle. (A) Representative images of SR ultrastructure in skeletal muscles from 6 months old male control and Myf5Cre-Pcyt2 mice. Scale bar 200 nm. (B) Voltage-dependence of the peak rate of sarcoplasmatic reticulum (SR) Ca2+ release (d[CaTot]/dt) measured from rhod-2 Ca2+ transients in fibers from male control (n=6 mice and 23 myofibers) and Myf5Cre-Pcyt2 (n=5 mice and 21 myofibers). (C) Decline of voltage-activated fluo-4FF Ca2+ transients in muscle fibers from control (n=2 mice and 6 myofibers) and Myf5Cre-Pcyt2 (n=3 mice and 8 myofibers) in response to an exhausting voltage stimulation protocol. (D) LC3 I/II and p62 levels in quadriceps from 8 months old control and Myf5Cre-Pcyt2 male mice under fed and fasting (24h) conditions. N=3 mice per group. (E) LC3 I/II and p62 levels in diaphragm from 8 months old control and Myf5Cre-Pcyt2 male mice under fed and fasting (24h) conditions. N=3 mice per group. (F-G) Quantification of p62 levels under fed and fasting conditions from quadriceps and diaphragm muscle respectively. Each dot represents individual mice. Data are shown as means ± SEM. Unpaired Student t-test with Welch correction was used for statistical analysis. Supplementary Material Supplementary methods and Supp. Figs.1-8 [391]NIHMS2010970-supplement-Supplementary_methods_and_Supp__Figs_1-8.p df^ (3.9MB, pdf) Supp data 1 [392]NIHMS2010970-supplement-Supp_data_1.xlsx^ (10KB, xlsx) Supp data 3 [393]NIHMS2010970-supplement-Supp_data_3.xlsx^ (12.4KB, xlsx) Supp data 2 [394]NIHMS2010970-supplement-Supp_data_2.xlsx^ (10.5KB, xlsx) Supp data 5 Supplementary Data 5 Statistical Source Data [395]NIHMS2010970-supplement-Supp_data_5.xlsx^ (15.4KB, xlsx) Supp data 4 [396]NIHMS2010970-supplement-Supp_data_4.xlsx^ (36KB, xlsx) Supp data 6 Supplementary Data 6 Statistical Source Data [397]NIHMS2010970-supplement-Supp_data_6.xlsx^ (28.6KB, xlsx) Supp data 7 Supplementary Data 7 Statistical Source Data [398]NIHMS2010970-supplement-Supp_data_7.xlsx^ (28.2KB, xlsx) Supp data 8 Supplementary Data 8 Statistical Source Data [399]NIHMS2010970-supplement-Supp_data_8.xlsx^ (26.6KB, xlsx) Source data Fig. 1.xlxs Source Data Fig. 1 Statistical Source Data [400]NIHMS2010970-supplement-Source_data_Fig__1_xlxs.xlsx^ (38.6KB, xlsx) Source data Fig. 3.xlxs Source Data Fig. 3 Statistical Source Data [401]NIHMS2010970-supplement-Source_data_Fig__3_xlxs.xlsx^ (17.7KB, xlsx) Source data Fig. 2.xlxs Source Data Fig. 2 Statistical Source Data [402]NIHMS2010970-supplement-Source_data_Fig__2_xlxs.xlsx^ (19.2KB, xlsx) Source data Fig. 4.xlxs Source Data Fig. 4 Statistical Source Data [403]NIHMS2010970-supplement-Source_data_Fig__4_xlxs.xlsx^ (15.6KB, xlsx) Source data Fig. 4.pdf Source Data Fig. 4 Unprocessed western Blots [404]NIHMS2010970-supplement-Source_data_Fig__4_pdf.pdf^ (214.1KB, pdf) Source data Fig. 5.xlxs Source Data Fig. 5 Statistical Source Data [405]NIHMS2010970-supplement-Source_data_Fig__5_xlxs.xlsx^ (208.3KB, xlsx) Source data Fig. 6.xlxs Source Data Fig. 6 Statistical Source Data [406]NIHMS2010970-supplement-Source_data_Fig__6_xlxs.xlsx^ (27.9KB, xlsx) Source data Fig. 7.xlxs Source Data Fig. 7 Statistical Source DataUnprocessed western Blots [407]NIHMS2010970-supplement-Source_data_Fig__7_xlxs.xlsx^ (38.6KB, xlsx) Source data Fig. 8.pdf Source Data Fig. 8 Unprocessed western BlotsUnprocessed western Blots [408]NIHMS2010970-supplement-Source_data_Fig__8_pdf.pdf^ (88.5KB, pdf) Source data Fig. 7.pdf [409]NIHMS2010970-supplement-Source_data_Fig__7_pdf.pdf^ (120.2KB, pdf) Source data Fig. 8.xlxs [410]NIHMS2010970-supplement-Source_data_Fig__8_xlxs.xlsx^ (50.5KB, xlsx) Source data extended Data Fig. 4.xlsx Source Data Extended Data Fig./Table 4 Statistical Source Data [411]NIHMS2010970-supplement-Source_data_extended_Data_Fig__4_xlsx.xlsx ^(12.7KB, xlsx) Source data extended Data Fig. 3.xlsx Source Data Extended Data Fig./Table 3 Statistical Source Data [412]NIHMS2010970-supplement-Source_data_extended_Data_Fig__3_xlsx.xlsx ^(17.4KB, xlsx) Source data extended Data Fig. 1.xlsx Source Data Extended Data Fig./Table 1 Statistical Source Data [413]NIHMS2010970-supplement-Source_data_extended_Data_Fig__1_xlsx.xlsx ^(93.5KB, xlsx) Source data extended Data Fig. 5.xlsx Source Data Extended Data Fig./Table 5 Statistical Source DataUnprocessed western Blots [414]NIHMS2010970-supplement-Source_data_extended_Data_Fig__5_xlsx.xlsx ^(15.6KB, xlsx) Source data extended Data Fig. 5.pdf [415]NIHMS2010970-supplement-Source_data_extended_Data_Fig__5_pdf.pdf^ (112.5KB, pdf) Source data extended Data Fig. 6.xlsx Source Data Extended Data Fig./Table 6 Statistical Source Data [416]NIHMS2010970-supplement-Source_data_extended_Data_Fig__6_xlsx.xlsx ^(62KB, xlsx) Source data extended Data Fig. 8.xlsx Source Data Extended Data Fig./Table 8 Statistical Source Data [417]NIHMS2010970-supplement-Source_data_extended_Data_Fig__8_xlsx.xlxs ^(13.5KB, xlxs) Source data extended Data Fig. 9.xlsx Source Data Extended Data Fig./Table 9 Statistical Source Data [418]NIHMS2010970-supplement-Source_data_extended_Data_Fig__9_xlsx.xlxs ^(16.9KB, xlxs) Source data extended Data Fig. 7.xlsx Source Data Extended Data Fig./Table 7 Statistical Source Data [419]NIHMS2010970-supplement-Source_data_extended_Data_Fig__7_xlsx.xlxs ^(14.7KB, xlxs) Source data extended Data Fig. 10.xlsx Source Data Extended Data Fig./Table 10 Statistical Source Data [420]NIHMS2010970-supplement-Source_data_extended_Data_Fig__10_xlsx.xlx s^ (14.4KB, xlxs) Source data extended Data Fig. 10.pdf Source Data Extended Data Fig./Table 10 Unprocessed western Blots [421]NIHMS2010970-supplement-Source_data_extended_Data_Fig__10_pdf.pdf^ (142.6KB, pdf) Video S1: Laser-induced myofiber damage assessment on myofibres isolated 6-month-old control mice Supplementary Video 1 Laser induced myofiber damage assessment on myofibers isolated 6-month-old control mice. [422]Download video file^ (369.2KB, mp4) Video S3: Laser-induced myofibre damage assessment on myofibres isolated 6-month-old Myf5Cre Pcyt2 mice Supplementary Video 3 Laser induced myofiber damage assessment on myofibers isolated 6-month-old Myf5Cre Pcyt2 mice. [423]Download video file^ (430.3KB, mp4) Video S4: Laser-induced myofibre damage assessment on myofibres isolated 6-month-old Myf5Cre Pcyt2 mice#2 Supplementary Video 4 Laser induced myofiber damage assessment on myofibers isolated 6-month-old Myf5Cre Pcyt2 mice.Video of the last bout of eccentric exercise (day 13) of 6-month-old control and Myf5Cre-Pcyt2 mice. First, mice were adjusted to a lower speed (9m min-1; t=0–10s), followed by a higher speed bout (20 m min-1; t=10–40s). [424]Download video file^ (407.5KB, mp4) Video S5: Video of the last bout of eccentric exercise (day 13) of 6-month-old control and Myf5Cre-Pcyt2 mice. First, mice were adjusted to a lower speed (9 m min−1; t = 0–10 s), followed by a higher speed bout (20 m min−1; t = 10–40 s) [425]Download video file^ (21.8MB, mp4) Video S2: Laser-induced myofibre damage assessment on myofibres isolated 6-month-old control mice2 Supplementary Video 2 Laser induced myofiber damage assessment on myofibers isolated 6-month-old control mice. [426]Download video file^ (379KB, mp4) Acknowledgements