Suppression of myopathic lamin mutations by muscle-specific activation of AMPK and modulation of downstream signaling

Abstract

Laminopathies are diseases caused by dominant mutations in the human LMNA gene encoding A-type lamins. Lamins are intermediate filaments that line the inner nuclear membrane, provide structural support for the nucleus and regulate gene expression. Drosophila melanogaster models of skeletal muscle laminopathies were developed to investigate the pathological defects caused by mutant lamins and identify potential therapeutic targets. Human disease-causing LMNA mutations were modeled in Drosophila Lamin C (LamC) and expressed in indirect flight muscle (IFM). IFM-specific expression of mutant, but not wild-type LamC, caused held-up wings indicative of myofibrillar defects. Analyses of the muscles revealed cytoplasmic aggregates of nuclear envelope (NE) proteins, nuclear and mitochondrial dysmorphology, myofibrillar disorganization and up-regulation of the autophagy cargo receptor p62. We hypothesized that the cytoplasmic aggregates of NE proteins trigger signaling pathways that alter cellular homeostasis, causing muscle dysfunction. In support of this hypothesis, transcriptomics data from human muscle biopsy tissue revealed misregulation of the AMP-activated protein kinase (AMPK)/4E-binding protein 1 (4E-BP1)/autophagy/proteostatic pathways. Ribosomal protein S6K (S6K) messenger RNA (mRNA) levels were increased and AMPKα and mRNAs encoding downstream targets were decreased in muscles expressing mutant LMNA relative controls. The Drosophila laminopathy models were used to determine if altering the levels of these factors modulated muscle pathology. Muscle-specific over-expression of AMPKα and down-stream targets 4E-BP, Forkhead box transcription factors O (Foxo) and Peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α), as well as inhibition of S6K, suppressed the held-up wing phenotype, myofibrillar defects and LamC aggregation. These findings provide novel insights on mutant LMNA-based disease mechanisms and identify potential targets for drug therapy.

Introduction

Laminopathies are a collection of diseases caused by dominant mutations in the human LMNA gene encoding A-type lamins (1,2). Lamins are intermediate filaments that line the inner nuclear membrane where they provide structural support for the nucleus and regulate gene expression (1,2). Laminopathies include autosomal dominant Emery–Dreifuss muscular dystrophy (EDMD2, OMIM #181350), Limb-Girdle muscular dystrophy type 1B (LGMD1B, 159001), congenital muscular dystrophy (MDC, OMIM #613205), dilated cardiomyopathy type 1A (CMD1A, OMIM #115200), familial partial lipodystrophy type 2 (FPLD2, OMIM #151660) and early on-set aging syndromes such as Hutchinson–Gilford progeria syndrome (HGPS; OMIM #176670) (1–6). It is unclear how LMNA mutations result in tissue-specific defects when mutant lamins are expressed in nearly all tissues (7,8). The pathogenic mechanisms of laminopathies are not well defined; hence, a greater understanding is needed to support the development of therapeutic interventions.

Over 400 distinct mutations have been identified in the LMNA gene, among the highest number of mutations discovered in a single human gene (1–3,7,9). The majority of these are point mutations throughout the gene that give rise to single amino acid substitutions in lamins A and C, two isoforms derived from alternatively spliced LMNA messenger RNA (mRNA). Amino acid substitutions that give rise to skeletal muscular dystrophy are often accompanied by congenital muscular dystrophy (CMD) (4,9–13). EDMD2 in particular is characterized by progressive muscle weakness, joint contractures and CMD with conduction defects (9,10). While much is known about the functions of lamins in the nucleus where they play a role in maintaining nuclear envelope (NE) integrity and organizing the genome (11–14), their functions in signaling pathways are becoming equally important with respect to disease mechanisms (1,2,15,16). For example, mutant lamins cause perturbations of the mammalian target of rapamycin (mTOR) pathway, which can be partially reversed with mTOR inhibitors such as rapamycin and temsiromilus (17,18). Genetic ablation of S6K1 (encoding ribosomal protein S6 protein kinase 1), a downstream substrate of mTOR, improved muscle function and extended lifespan of Lmna^-/- mice (19). mTOR activity inversely correlates with the rate of autophagy, which plays a role in regulating nuclear-to-cytoplasmic transport and degradation of Lamin B1 (20). Consistent with these findings, activation of autophagy suppressed cardiac laminopathy in a Drosophila model (21). Thus, regulation of the mTOR pathway is critical for muscle health in the context of laminopathies, however, which factors upstream and downstream of mTOR play key roles needed further investigation.

To evaluate the role of TOR signaling and autophagy in lamin-associated muscle disease, we established Drosophila melanogaster (fruit fly) models of laminopathies. Drosophila models have proved to be powerful in defining the mechanistic basis of human disease, including muscle disorders associated with cytoskeletal defects (2,22–29). In addition, Drosophila models have been used to identify potential therapeutic targets for human aging disorders (2,29–33). Relevant for this study, Drosophila indirect flight muscle (IFM) models have been successfully used to define the molecular basis for muscle organization and disorganization (29,32,34–38). Importantly, expression of dominant negative (DN) mutants and knock-down (KD) of IFM-specific genes does not cause lethality in flies (29,32,36), allowing evaluation of pathophysiological aspects of progressive muscle degeneration without effects on the remainder of the organism. A dominant flightless phenotype with abnormal wing position provides powerful visual markers of defective IFM function (36,39). D. melanogaster, with its high degree of genome conservation to humans and manipulability through versatile genetic techniques, is an excellent model for understanding the molecular mechanisms of mutant lamin-induced skeletal muscle defects (29,36).

The expression of D. melanogaster Lamin C (LamC) gene is developmentally regulated and nearly ubiquitously expressed, similar to the human LMNA gene (6). LamC shares amino acid sequence identity with human lamins A and C. Lamins have a conserved protein domain structure with a globular head, coiled-coil rod and a tail domain possessing an immunoglobulin-fold (Ig-fold) (40). In addition, LamC localizes to the NE in all Drosophila tissues investigated including cardiac and larval body wall muscle tissue, supporting Drosophila as a useful model (16,25,41). Furthermore, the pathogenic genes and pathways described in this study are highly conserved between Drosophila and humans, offering the possibilities for the identification of conserved drug targets. The genetic and pharmacological manipulation of these pathways will provide mechanistic tests for potential skeletal muscle laminopathy therapies (4,6,7,16,42).

To address the molecular basis of skeletal muscle laminopathies, mutations were made in Drosophila LamC analogous to those that cause muscle disease in humans. Muscle-specific expression of mutant LamC resulted in muscle functional defects that were accompanied by a plethora of cellular abnormalities including cytoplasmic aggregation of NE proteins. We hypothesize that these cytoplasmic aggregates trigger signaling pathways and alter cellular and metabolic homeostasis, which results in muscle dysfunction. In support of our hypothesis and to reveal relevance to human pathology, transcriptomics data obtained from human muscle biopsy tissue showed misregulation of genes in the AMP-activated protein kinase (AMPK)/TOR/autophagy signaling pathways. Genetic manipulation of these pathways in Drosophila IFM suppressed the muscle defects, suggesting that misregulation of these pathways was causal to the muscle pathology. Overall, our analysis identified potential new therapeutic targets for lamin-associated skeletal myopathies and possibly other laminopathies.

Results

Expression of mutant LamC resulted in a held-up wing phenotype due to myofibril defects

To understand the tissue-specific functions of lamins and to address the molecular basis of muscle laminopathies, we developed Drosophila models. Point mutations were generated in D. melanogaster LamC analogous to dominant muscle disease-causing mutations in the human LMNA gene (16,24,25,28). LamC A177P and R205W affect the rod domain and correspond to human lamin A/C L162P and R190W, respectively; LamC G489V and V528P affect the Ig-fold domain and correspond to human lamin A/C G449V and L489P, respectively (Fig. 1A). These mutant LamC transgenes were expressed in the IFM using the Gal4/UAS system (43) with three IFM-specific Gal4 `drivers’: Act88F-Gal4 (44), Fln-Gal4 (45) and DJ-694-Gal4 (44). While exclusively expressed in IFM, these drivers have different temporal expression profiles. IFM performance was assayed by flight tests (29,39,46). Expression of the mutant LamC transgenes with Act88F-Gal4 and Fln-Gal4 that drive expression prior to sarcomere maturation resulted in severe muscle dysfunctions (measured as flight capability defects) (Fig. 1B and Supplementary Material, Fig. S1A and B). In contrast, expression of wild-type LamC and Gal4 alone using these drivers had no obvious effects on muscle function (Fig. 1B and Supplementary Material, Fig. S1A and B). Expression of A177P produced a moderate flight defect (Fig. 1B and Supplementary Material, Fig. S1A and B) that worsened with age, suggesting progressive muscle weakening, which is characteristic of human laminopathies (25,28,47). Expression of R205W, G489V and V528P resulted in a flightless (non-flight, NF) phenotype (Fig. 1B and Supplementary Material, Fig. S1A), suggesting severe muscle defects. Expression of mutant LamC with DJ694-Gal4, which drives expression after sarcomere assembly (44), resulted in less severe flight defects compared to the other Gal4 drivers tested and also worsened with age. In contrast, no defects were observed upon expression of wild-type LamC (Fig. 1C versus B and Supplementary Material, Fig. S1A versus C). Taken together, these data indicate that expression of mutant LamC causes IFM defects and might interfere with sarcomere maturation.

Figure 1.

IFM expression of mutant LamC caused loss of muscle function and altered muscle architecture. (A) Diagram of LamC protein domains showing amino acid substitutions used in this study (black, Drosophila numbering; red, corresponding human numbering). (B) Flight indexes (FIs) of females expressing mutant LamC via the Fln Gal4 driver (expressed during muscle maturation) display severe (R205W, G489V and V528P) and subtle (A177P) progressive flight defects, compared with age-matched controls (n = 100 female flies per genotype). (NF, no flight i.e. flightless phenotype). (C) FIs of female flies expressing mutant LamC via the DJ694 Gal4 driver (expressed post muscle maturation) show mild, yet progressive flight defects compared to controls (n = 100 females flies for each). FIs of 3-day- and 1-week-old adults expressing R205W, G489V and V528P were not statistically different from age-matched wild-type LamC-expressing adults or from the driver only (DJ694-Gal4/+) controls. FIs for adults expressing the Gal4 drivers alone and wild-type LamC were not statistically different from that of the non-transgenic host control w¹¹¹⁸ (represented as +/+ in panels B and C). (^*= P < 0.05, ^**= P < 0.01 and ^***= P < 0.001; NS = not significant for (B) and (C), using one-way analysis of variance (ANOVA) with post hoc Tukey test). (D) Images of the held-up wing phenotype in the male flies upon expression of mutant LamC (R205W, G489V and V528P) in the IFM. Females showed a similar phenotype (data not shown). (E) Ultrastructural analysis of IFM from pupae expressing mutant LamC showed sarcomere assembly/maturation defects. Longitudinal sections of IFM myofibrils from stage-15 control pupae (∼90 h after puparium formation) show normal sarcomeric organization. Longitudinal sections of IFM myofibrils from age-matched stage-15 pupae IFM expressing R205W and G489V revealed disrupted Z-discs (Z) and M-lines (M). (F) Quantification of sarcomere length, Z-disc and M-line integrity showed that R205W and G489V caused a reduction in sarcomere length and disorganization of Z-discs and M-lines. Analyses were carried out from 50–300 sarcomeres for each genotype using ImageJ as described in the methods section. For panel (F) statistical analysis using one-way ANOVA with post hoc Tukey test (^***= P < 0.001 and NS = not significant) was performed.

To determine if the muscle dysfunction was caused by mutant LamC proteins or overall increased levels of LamC, western analyses were performed. Protein extracts from 1-week-old adults expressing wild-type and mutant LamC via the Fln-Gal4 and DJ-694-Gal4 drivers showed similar levels of expression. Expression of the wild-type LamC transgene in an otherwise wild-type genetic background produced ∼1.5- to 2.0-fold higher levels than endogenously produced LamC (Supplementary Material, Fig. S1D). However, this slight increase did not affect the physiological parameters examined in muscle, as they were indistinguishable from age-matched controls expressing the Gal4 driver alone (Fln-Gal4/+) (Fig. 1B and E). Furthermore, despite similar expression levels for R205W, G489V and V528P with Fln-Gal4 and DJ-694-Gal4, the flies showed a severe and mild phenotype, respectively. These results suggested that the severe defects obtained with Fln-Gal4 were due to interference of LamC aggregates with sarcomere maturation. Additionally, Fln-Gal4 driven expression of wild-type and mutant LamC in a heterozygous LamC (null/+) background (homozygous null mutations in LamC are lethal) (48) produced similar phenotypes as observed in a wild-type genetic background (Supplementary Material, Fig. S2A). Overall, these data demonstrate that the muscle dysfunction was caused by mutant LamC, not increased levels of total LamC.

The severe flight defect caused by R205W, G489V and V528P when expressed via Fln-Gal4 resulted in a wing posture defect (held-up wings) (Fig. 1D, almost 100% of the flies showed this wing phenotype after 1 week of age). This phenotype has been observed with mutations in genes encoding contractile proteins that give rise to muscle defects (29,49). Held-up wings can also be caused by mitochondrial dysfunction (50). To more closely examine sarcomere assembly/maturation, transmission electron microscopy (TEM) was performed on pupal IFM expressing R205W and G489V and wild-type LamC. Muscles expressing the mutant LamC transgenes showed cytoskeletal assembly defects at stage P15 of pupal IFM development, with abnormalities of Z-discs and M-lines (Fig. 1E). In contrast, muscle-specific expression of wild-type LamC showed no apparent defects. Furthermore, controls expressing the Gal4 driver alone (Act88F/+) showed normal sarcomere assembly/maturation (Fig. 1E). Sarcomere length was reduced in muscles expressing R205W and G489V, relative to muscles expressing wild-type LamC and the Gal4 driver alone (Fig. 1F). The M-lines and Z-discs were disrupted in muscles expressing mutant LamC, whereas they appeared normal in muscle expressing wild-type LamC and the Gal4 driver alone (Fig. 1F). Thus, the held-up wing phenotype caused by IFM-specific expression of mutant LamC (Fig. 1D) can be at least partially attributed to myofibrillar disorganization and represents an excellent marker for genetic/pharmacological screens for suppression of muscle defects.

Mutant LamC caused cytoplasmic aggregation of NE proteins, nuclear blebbing and disorganization of cytoskeletal proteins

To determine the cellular consequences of expressing mutant LamC, IFMs were cryosectioned and viewed by confocal microscopy as previously described (51,52). Briefly, adult thoraces were flash frozen, and cryosections of the IFM (30 μm thick) were stained with antibodies to nuclear and cytoplasmic proteins and phalloidin, which recognizes F-actin. Expression of wild-type LamC with Fln-Gal4 showed LamC localization to the NE as anticipated with no apparent defects in nuclear shape and myofilament organization (Fig. 2A–A”). In contrast, expression of R205W, G489V and V528P with these same drivers showed LamC localization at the NE, but also abnormal aggregation in the cytoplasm (arrowheads). In addition, there was abnormal nuclear morphology, including blebbing of the NE (arrows) and disrupted actin-containing fibers (asterisks) (Fig. 2A”). Nuclear blebbing has been reported for fibroblasts from individuals with HGPS and might be a marker and/or contribute to accelerated aging (53). These muscle defects were apparent in 3-day-old flies (Fig. 2A–A”) and correlated with severe loss of muscle function (Fig. 1B). Furthermore, these abnormalities increased in severity with age, as shown by comparison of 3-day- and 3-week-old adults (Fig. 2B–B”). In contrast to these findings, expression of A177P did not show LamC aggregation in 3-day-old adults (Fig. 2A–A”), which was consistent with the normal flight index (FI) (Fig. 1B and Supplementary Material, Fig. S1A). However, 3-week-old adults expressing A177P showed cytoplasmic aggregates of LamC (Fig. 2B and B’) that correlated with loss of muscle function (Fig. 2B–B” and Supplementary Material, Fig. S1A). Quantitation of the LamC aggregates using ImageJ showed that the relative area occupied by the aggregates per total area surveyed was greater in IFMs expressing mutant LamC versus those expressing wild-type LamC (Supplementary Material, Fig. S2B). Furthermore, the amount of cytoplasmic aggregates increased with age (Supplementary Material, Fig. S2B, red asterisks). The number of nuclei showing dysmorphology (enlarged size and/or atypical shape), the relative percentage of nuclei showing blebs and the relative percentage of nuclei showing misalignment within muscle fibers (Supplementary Material, Fig. S2C) were increased for muscles expressing R205W, G489V and V528P (and in some cases A177P), compared to age-matched controls expressing wild-type LamC. Moreover, the differences between muscles expressing mutant LamC and wild-type LamC increased with age (Supplementary Material, Fig. S2C, red asterisks).

Figure 2.

Mutant LamC caused myofibrillar disorganization, nuclear blebbing and cytoplasmic aggregation of NE proteins. (A and B) Confocal images of cryosectioned IFMs from 3-day- and 3-week old adults were stained with antibody against LamC (orange) and DAPI (blue, DNA). (A’ and B’) Sections of the stained IFM (from above) showing enlarged images stained with DAPI and anti-LamC antibodies. (A” and B”) merged images of (A) and (A’) and (B) and (B’), respectively, with F-actin labeled by phalloidin (gray). Expression of mutant LamC resulted in NE morphological defects (arrows), disorganization of actin-containing myofibrils (asterisks) and cytoplasmic aggregates of LamC (arrowheads). (orange, LamC; blue, DAPI) (C) Confocal images of cryosectioned IFMs of 3-week-old adults were stained with antibodies against FG-containing nuclear pore proteins (Nups) (white) and DAPI (blue). Expression of mutant LamC resulted in nuclear enlargement (arrows) and cytoplasmic aggregation of Nups (arrowheads). These defects are not observed in IFM from similar aged adults expressing wild-type LamC. (C’) Sections of the stained IFM (from directly above) showing only the staining with anti-Nup antibodies (red) and DAPI (blue). (C”) Merged images of (C) and (C’) stained with antibodies to muscle myosin II (green) showing disorganization of myosin-containing myofibrils (asterisks).

Expression of A177P, R205W, G489V and V528P via the DJ-694-Gal4 driver caused slight, but significant, functional defects in 3-week-old adults (Fig. 1C and Supplementary Material, Fig. S1C), which correlated with cytoplasmic aggregation of LamC (Supplementary Material, Fig. S3B). In contrast, expression of wild-type LamC did not cause these abnormalities (Supplementary Material, Fig. S3A and B). Quantification showed that expression of mutant LamC increased the number of nuclei with dysmorphology (enlarged size and/or atypical shape), the relative percentage of nuclei with blebs and the relative percentage of nuclei misaligned within muscle fibers, compared with age-matched controls expressing wild-type LamC (Supplementary Material, Fig. S3C). Thus, cytoplasmic LamC aggregates correlated with altered nuclear morphology and physiological dysfunction.

The nuclear lamina provides a structural lattice for nuclear pore proteins (Nups) (16,25). We examined the localization of FG-repeat containing Nups. In IFMs of 3-week-old adults expressing wild-type LamC, FG-repeat containing Nups localized to the NE as expected. In contrast, IFMs of 3-week-old adults expressing mutant LamC showed cytoplasmic aggregation of FG-repeat containing Nups (Fig. 2C–C”). The relative area occupied by the Nup aggregates was quantitated using ImageJ (Supplementary Material, Fig. S2D). The presence of cytoplasmic Nup aggregates correlated with disorganization of the myosin-containing myofibrillar structure with expression of R205W, G489V and V528P causing severe disruptions, A177P less severe and wild-type LamC no apparent disruptions (Fig. 2C–C”). Cytoplasmic aggregation of Nups has been reported in Drosophila larval body wall muscles expressing mutant LamC G489V and human muscle biopsy tissue possessing LMNA mutations, including a mutant allele that produces G449V, which is analogous to Drosophila G489V (16). The observation that both LamC and Nups aggregate in the cytoplasm provides insights on the etiology of the skeletal muscle dysfunction and suggests that reducing abnormal cytoplasmic protein aggregation might be an avenue for therapy.

Ultrastructural and cytological analyses revealed that mutant LamC caused abnormal myofibrillar organization, autophagic defects, mitochondrial dysmorphology and nuclear defects

To examine the mutant LamC-induced cytological defects at the ultrastructural level, TEM was performed on IFMs as described (29,54,55). Expression of wild-type LamC via Fln-Gal4 showed no ultrastructural IFM defects in 3-day-old adults, similar to that of 3-day-old adults expressing Gal4 alone (Fig. 3A, top row). In contrast, expression of R205W and G489V showed partial loss of Z-discs and M-lines (Fig. 3A, middle panel) in 3-day-old adults. The ultrastructural analysis also revealed that LamC caused enlarged nuclei that contained blebs for R205W and G489V (Fig. 3A, middle row), which was consistent with images obtained from confocal microscopy and has been observed in electron micrographs of biopsied muscle tissue from individuals with muscle laminopathies (47). This phenotype was not seen in IFM expressing wild-type LamC and Gal4 alone. Quantification of the TEM images revealed that expression of R205W and G489V resulted in shortening of sarcomere length and disruption of the M-lines and Z-discs, which were not observed in the controls (Fig. 3C). The TEM images also revealed that R205W and G489V caused severe mitochondrial dysmorphology including small and fragmented mitochondria (Fig. 3A, lower row, Fig. 3C and Supplementary Material, Fig. S5), possibly by interfering with mitochondrial mRNA import as reported in muscle for a Drosophila model of progeria (2). Mitochondrial defects have also been reported for a mouse model of HGPS cause by mutations in Lmna (55). Using a recombinant stock possessing Fln-Gal4 and UAS-mito-GFP, which labels mitochondria with Green fluorescent protein (GFP) (45), we discovered that expression of mutant LamC caused mitochondrial dysmorphology (Fig. 3B and Supplementary Material, Fig. S4). Expression of mutant LamC resulted in fragmented and deformed mitochondria, whereas expression of wild-type LamC and the Gal4 driver alone yielded mitochondria and actin-containing myofibrillar organization that appeared normal (Fig. 3B and Supplementary Material, Fig. S4). Furthermore, expression of the mito-GFP transgene did not affect the flight performance of flies expressing wild-type LamC, the non-transgenic control host stock (w¹¹¹⁸) and the Gal4 driver alone, all of which gave nearly identical values, demonstrating the mitochondrial defects were caused by mutant LamC.

Figure 3.

Expression of mutant LamC caused nuclear enlargement, mitochondrial abnormalities and autophagic defects. (A) Ultrastructural analysis using TEM revealed that IFM-specific expression of mutant LamC caused myofibril disarray, nuclear enlargement and abnormal mitochondrial shape compared to IFM expressing wild-type LamC. Occasional myofibrils with partially intact Z-discs (Z) and M-lines (M) were observed (arrows and arrowheads, upper panels). However, the majority of the myofibrils were completely disorganized (arrows) and possessed nuclei with abnormal morphology (middle panels). Abnormal mitochondria (m) and completely disorganized myofibrils (arrows) were evident (lower panels). (B) Confocal images of cryosectioned IFMs from 3-day-old adults expressing mito-GFP (green) and phalloidin (red). Compared with age-matched adults expressing the Gal4 drivers alone and wild-type LamC, expression of R205W and G489V resulted in fragmented and diffuse mitochondria (see Supplementary Material, Fig. S4 for quantification). (C) Quantification of sarcomere length, Z-disc and M-line integrity and mitochondrial dysmorphology in TEM images. Expression of R205W and G489V resulted in a reduction of sarcomere length, disorganization of Z-discs and M-lines and mitochondrial dysmorphology. Analyses were carried out from 10–400 sarcomeres from three independent flies for each genotype using ImageJ as described in the methods section. For (C) statistical analysis using one-way ANOVA with post hoc Tukey test (^***= P < 0.001 and NS = not significant) was performed. (D and E) Representative western analysis (top) and antibody dot blot analysis of Ref(2)P protein levels (bottom), normalized to histone H2B, in IFMs from 3-week-old female flies expressing R205W and G489V compared to IFMs from the same age control flies expressing wild-type LamC. Statistical analysis using one-way ANOVA with post hoc Tukey test (^*= P < 0.5 and ^**= P < 0.01, n = 3) was performed. (F) Up-regulation of Ref(2)P was detected by immunofluorescence in IFM of 3-week-old flies (green foci) expressing mutant LamC (R205W and G489V) and compared with that of age-matched controls expressing wild-type LamC [green, Ref(2)P; blue, DAPI; red, phalloidin]. (F’) Immunofluorescence images of IFM stained with antibodies to Ref(2)P (green, cytoplasmic foci) and LamC (orange, cytoplasmic aggregates) upon expression of mutant LamC -R205W and G489V (arrowheads).

The TEM analysis also revealed that expression of R205W and G489V yielded double membrane structures containing mitochondria and vacuoles, suggesting autophagic defects (Supplementary Material, Fig. S5). Both perinuclear and intranuclear vacuoles containing a variety of materials have been observed in electron micrographs of human muscle biopsy tissue (47). Taken together, these ultrastructural studies show that mutant LamC caused myofibrillar, autophagic, mitochondrial and nuclear defects, all of which likely contribute to the poor health of the IFM.

Mutant LamC caused increased levels of Ref(2)P/p62 in IFM

Increased cytoplasmic protein aggregation often leads to up-regulation of the autophagy cargo receptor p62 (16,21,56,57). This was the case when expressing mutant LamC in larval body wall muscle and cardiac tissue (16,21). Ref(2)P is the Drosophila orthologue of mammalian p62 (56,58,59). We examined levels of Ref(2)P in IFM of adults expressing wild-type and mutant LamC. Expression of R205W and G489V via Fln-Gal4 showed up-regulation of Ref(2)P relative to that of adults expressing wild-type LamC as shown by immuno- and dot blot analyses and immunohistochemical staining (Fig. 3D–F). A subset of the LamC aggregates co-localize with Ref(2)P puncta (Fig. 3F’, arrowheads). Thus, expression of mutant LamC caused increased levels of Ref(2)P, which is indicative of the loss of proteostasis and metabolic homeostasis (56,57,59). Up-regulation of p62 has also been observed in the Drosophila heart and human muscle biopsy tissue from individuals with muscle laminopathies (16,21), suggesting that up-regulation of p62 might be a useful marker for laminopathy-induced skeletal muscle dysfunction.

RNA-seq analysis of human muscle biopsy tissue revealed dysregulation of proteostasis, autophagy and signaling kinases

To determine if our findings in Drosophila were relevant to human disease, we performed RNA-seq analysis on human muscle biopsy tissue from an individual possessing a LMNA point mutation (1346T > C) that produces Lamin A/C G449V (analogous to Drosophila LamC G489V) and muscle tissue from an age-matched control. RNA was extracted, high thoroughput sequencing was performed and expression of the 36 144 genes represented in the Ensemble database (https://www.ensembl.org/biomart) was analyzed (One Array, Phalanx Biotech Group). The results showed that expression of 3047 genes was altered in the diseased versus control muscle (1.5-fold or greater change). The molecular functions of the protein products encoded by these genes were analyzed using DAVID, BIND and STRING online tools (60–62). Pathways and genes involved in biological processes misregulated by mutation of LMNA are listed in Supplementary Material, Tables S1 and 2, respectively. Genes associated with metabolic and cellular processes were enriched among those misregulated (Supplementary Material, Table S1). The function of these genes and their direction of change (up- or down-regulated) were in good agreement with our findings in Drosophila (Figs 1–3, Supplementary Material, Table S2). For example, several kinases in signaling pathways such as MAPK and AKT/mTOR were up-regulated in the diseased muscle relative to the control. Several autophagy related genes were up-regulated [ULK1 (orthologue of Drosophila Atg1), ULK3 (orthologue of Drosophila CG8866), ATG4, ATG7, ATG10, ATG101, BECN1 and epsin 3]. AMPK transcripts for alpha and three gamma subunits were significantly down-regulated; transcripts for both beta subunits remained unchanged. Transcripts of genes downstream of AMPK and genes associated with proteostasis were down-regulated in the diseased muscle compared to the age-matched control [examples: 4E-BP, Foxo, PGC1α, Adipose triglyceride lipase (ATGL), sirtuin and heat shock transcription factor 1 (HSF1)]. Thus, the transcriptomic data from an individual with LMNA muscular dystrophy showed changes that were consistent with the findings in Drosophila, strengthening the fly models.

AMPKα over-expression suppressed mutant LamC-induced muscle dysfunction and cytological defects

The transcript analyses of human diseased muscle suggested alterations in several signaling pathways that regulate proteostasis, a process of interconnected cellular networks that includes protein folding and metabolism. We tested for this using the Drosophila models, which allowed for manipulation of specific pathways exclusively in the IFM without affecting the rest of the organism. Over-expression (OE), RNA interference (RNAi) KD, and expression of DN proteins in these pathways was used to modulate their activity in IFM and then muscle function assays and cytology analyses were performed (Supplementary Material, Table S2).

The human muscle gene expression data revealed that AMPKα kinase transcripts were down-regulated 1.6-fold in diseased muscle. AMPK is a sensor of cellular energy and a regulator of cellular and organismal metabolism. AMPK activity has been linked to autophagy, proteostasis and mitochondrial function (63–66). In both mammals and flies, AMPK regulates proteostasis via autophagy, HSF1 and its downstream targets that include 4E-BP, Foxo and PGC1α (67). Our data from Drosophila and human muscle suggested that autophagy might be impaired by expression of mutant LamC. Therefore, we hypothesized that OE of AMPKα would suppress the muscle defects through down-regulation of TOR activity, which increases the rate of autophagy (68). IFM-specific OE of AMPKα via Fln-Gal4 suppressed the flight defect caused by IFM-specific expression of mutant LamC (Fig. 4) and had no effect on the FI of adults expressing wild-type LamC (Fig. 4A). In contrast, OE of a truncated form of AMPKα (lacking kinase activity) and other mutated forms (69) did not show suppression (Supplementary Material, Table S3). Consistent with this finding, IFM-specific RNAi KD of AMPKα did not suppress the mutant LamC-induced muscle dysfunction. In addition, RNAi KD of AMPKα did not appear to affect muscles expressing mutant and wild-type LamC (Fig. 4). IFM-specific expression of an RNAi against GFP and OE of GFP (non-specific target) showed no suppression of muscle defects caused by mutant LamC (Fig. 4A). In addition, AMPKα OE, but not KD, suppressed the mutant LamC-induced held-up wings phenotype (Fig. 4B). Consistent with these data, IFM-specific OE of AMPKα, but not KD of AMPKα and GFP (non-target control) reduced levels of Ref(2)P (Fig. 4C). Taken together, these findings suggest that AMPK kinase activity is required for suppression of muscle defects caused by mutant LamC.

Figure 4.

Suppression of mutant LamC-induced IFM muscle dysfunction and cytoplasmic LamC aggregation by AMPKa over-expression (OE). (A) IFM-specific OE of AMPKα suppressed functional defects (flightless phenotype) caused by G489V in 3-week-old flies (n = 100). KD of AMPKa and GFP (non-target control) did not suppress the flightless phenotype. (B) The mutant LamC-induced held-up wing phenotype was suppressed upon OE of AMPKα (female flies shown). KD of AMPKα and GFP did not suppress the held-up wing phenotype. (C) Quantitative dot blot analysis of Ref(2)P protein levels normalized to levels of histone H2B, in IFMs from 3-week-old female flies expressing either wild-type or mutant LamC in combination with AMPKα OE, AMPKα KD or GFP KD. G489V caused up-regulation of Ref(2)P, which was reduced upon AMPKα OE, but not AMPKα KD and GFP KD (D and D’). Confocal images of cryosectioned IFMs from 3-week-old adults stained with antibodies against Drosophila LamC (orange), DAPI (blue) and phalloidin (red). IFM-specific OE of AMPKα suppressed the cytoplasmic aggregation of LamC and myofibril disorganization caused by G489V. No suppression was observed upon AMPKa KD and GFP KD. Arrows represent mutant LamC-induced enlarged nuclei, asterisks represent disorganized actin-containing myofibrils, and arrowheads represent cytoplasmic aggregation of LamC. (E and E’), Confocal images of cryosectioned IFMs from 3-week-old adults expressing wild-type LamC stained as in panel D. (F–I), Quantification of mutant LamC-induced LamC aggregates, nuclear dysmorphology (blebs) and percentage of enlarged and misaligned nuclei. Quantification of LamC aggregates and nuclear dysmorphology was performed as previously reported (21,54) and described in the methods section. Briefly, the area represented by the aggregates/nuclei was divided by the total area surveyed and compared among genotypes for a minimum of six images/genotype from three independent files' IFM. As indicated in the methods section, for each genetic modifier, the area occupied by the aggregates/nuclei was reported relative to the values obtained with the genetic modifier alone. Statistical indicators in red text are results obtained from comparisons to wild-type LamC; indicators in black text are results obtained from comparisons to the specific mutant LamC. Statistical analysis using one-way ANOVA with post hoc Tukey test (^***= P < 0.001, ns = non-significant) was performed.

To determine if suppression of the flight defect by AMPKα OE was due to cytological changes in the IFMs, immunohistochemistry was performed. AMPKα OE caused no observable cytological changes in IFMs expressing wild-type LamC (Fig. 4D and E). In contrast, AMPKα OE showed suppression of cytoplasmic LamC aggregation, nuclear morphology defects and myofibril disorganization in muscle expressing mutant LamC (Fig. 4D, D’, and F-1). IFM-specific RNAi KD of AMPKα exacerbated mutant LamC-induced nuclear morphology and myofibril defects, whereas IFM-specific expression of an RNAi against AMPKα and GFP had no observable effects on controls (Fig. 4D, E and F-I). Overall, these findings identify the AMPK pathway as a potential target for treating skeletal muscle laminopathies.

Muscle-specific activation of AMPK downstream targets, PGC1α (spargel) and Foxo, suppressed muscle defects caused by mutant LamC

The transcriptomics data from human muscle biopsy tissue revealed that in addition to AMPKα down-regulation, transcripts of AMPKα targets were also down-regulated (Supplementary Material, Table S2). We hypothesized that OE of a conserved downstream target gene, such as peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α) (called dPGC-1 and spargel in Drosophila), a master regulator of mitochondrial biogenesis (70), would suppress the muscle defects. IFM-specific OE of dPGC-1 in muscle expressing wild-type LamC produced no noticeable defects (Fig. 5A). In contrast, IFM-specific OE of dPGC-1 in muscle expressing mutant LamC suppressed the held-up wing phenotype and flight defect (Fig. 5A and B). This suppression was accompanied by a reduction in cytoplasmic LamC aggregation, nuclear defects and myofibril disorganization (Fig. 5C). Consistent with these findings, IFM-specific KD of dPGC-1 enhanced these cytological defects (Fig. 5C). Quantification revealed that OE of dPGC-1 suppressed G489V-induced LamC aggregates, nuclear blebs and misaligned nuclei (Fig. 5E–H), In contrast, suppression was not observed with an RNAi against dPGC-1 (Fig. 5E–H). In fact, some of the abnormal phenotypes induced by G489V were enhanced upon dPGC-1 KD. Thus, it appeared that at least a portion of the suppressive effects of AMPKα OE was likely through dPGC-1, potentially by maintaining mitochondrial and redox homeostasis (70).

Figure 5.

Suppression of mutant LamC-induced IFM dysfunction and cytoplasmic LamC aggregation by OE of dPGC-1- and Foxo. (A) FIs of female flies (n = 100 per genotype) showed suppression of G489V-induced functional defect (flightless phenotype) upon over expressing dPGC-1. No suppression of the flight defect was observed with Foxo OE. In addition, no suppression was observed upon dPGC-1 and Foxo KD. (B) The mutant LamC-induced held-up wing phenotype was suppressed upon dPGC-1 OE. Suppression was not observed upon dPGC-1 OE, Foxo OE and Foxo KD. (C) Confocal images of cryosectioned IFMs from 3-week-old adults stained with an antibody to LamC (orange), DAPI (blue) and phalloidin (red). IFM-specific OE of dPGC-1 and Foxo partially suppressed the cytoplasmic aggregation of LamC, the disorganization of myofibrils and nuclear blebbing caused by G489V. Suppression was not observed upon dPGC-1 KD or Foxo KD. Arrows represent mutant LamC-induced enlarged nuclei, asterisks represent disorganized actin-containing myofibrils and arrowheads represent cytoplasmic aggregates of LamC. (D) Confocal images of cryosectioned IFMs from 3-week-old adults expressing wild-type LamC as in (D and D’). (E–H) Quantification of mutant LamC-induced aggregates, nuclear dysmorphology (blebs) and percentage of enlarged and misaligned nuclei. Quantification of LamC aggregates and nuclear dysmorphology was performed as reported for Fig. 4F–I. Statistical indicators in red text are results obtained from comparisons to wild-type LamC; indicators in black text are results obtained from comparisons to the specific mutant LamC. Statistical analysis using one-way ANOVA with post hoc Tukey test (^*= P < 0.05, 0.01 = P < 0.01, ^***= P < 0.001, ns = non-significant) was performed.

Transcripts encoding Foxo, a phosphorylation target of AMPKα, were reduced in the human muscle biopsy tissue (Supplementary Material, Table S2). Foxo transcription factor activity reduces age-associated protein toxicity by eliminating protein aggregates via autophagy (71). IFM-specific OE of Foxo in muscle expressing wild-type LamC had no noticeable effect (Fig. 5A). IFM-specific OE of Foxo did not suppress the flight defect caused by mutant LamC (Fig. 5A); however, cytological analysis revealed a mild reduction in LamC aggregation, nuclear morphological defects and myofibril disorganization (Fig. 5C and Supplementary Material, Fig. S5). Moreover, IFM-specific KD of Foxo enhanced cytoplasmic LamC aggregates and nuclear deformation (Fig. 5C). Quantitative analyses agreed with the qualitative assessment of the effects of Foxo OE, showing partial suppression of cytological defects, but no restoration of muscle function (Fig. 5E–H). These results suggest that much of the suppressive effects of AMPKα OE are unlikely due to down-regulation of Foxo, which might play a contributory role in muscle pathology.

Muscle-specific inhibition of S6K and OE of 4E-BP (Thor) suppressed muscle defects caused by mutant LamC

The transcriptomic data from the human muscle biopsy tissue revealed that mRNAs encoding ribosomal protein S6K (S6 kinase) were up-regulated (more than 3-fold) relative to the control (Supplementary Material, Table S2). S6K is a downstream effector of TOR that phosphorylates a number of substrates including ribosomal protein S6 for translation of capped mRNAs (72). The TOR-S6K axis is part of a conserved signaling cascade that regulates many aspects of cellular physiology including glucose homeostasis, insulin sensitivity and natural aging, processes that are altered in individuals with laminopathies (17,18). In mouse laminopathy models, rapamycin and temsirolimus, both inhibitors of AKT/mTOR signaling and, therefore, S6K phosphorylation, had beneficial effects on muscle function (17,18). In Drosophila, extension of lifespan by rapamycin was blocked by ubiquitous OE of a constitutively active S6K (72). It is unclear which step(s) of TOR signaling are misregulated by mutant lamins and the overall consequences on autophagy flux, which increases upon rapamycin treatment. To address this issue, we tested for effects of factors downstream of TOR on the suppression of pathologies caused by mutant LamC. IFM-specific expression of a DN version of S6K suppressed the held-up wing phenotype and flight defect caused by mutant LamC (Fig. 6A and B). Furthermore, the DN S6K suppressed cytoplasmic LamC aggregates, nuclear morphological defects and disorganization of actin-containing myofibrils (Fig. 6C and E–H ). OE of wild-type S6K in muscle expressing wild-type LamC had no observable effects and did not suppress defects of muscles expressing mutant LamC (Fig. 6A–C). Collectively, these data suggest that inhibition of the TOR-S6K axis, particularly at the level of S6K, might have therapeutic value for individuals with muscle laminopathies.

Figure 6.

Suppression of mutant LamC-induced muscle dysfunction and cytoplasmic LamC aggregation by downstream targets of mTOR (S6K and 4E-BP). (A) FIs of female flies showing suppression of G489V-induced functional defects (flightless phenotype) upon OE of a DN S6K and OE of Thor (4E-BP). OE of wild-type S6K and KD of Thor did not suppress the flightless phenotype. (n = 100 flies per genotype) (B) G489V-induced wing phenotype was suppressed by OE of DN S6K and OE of Thor. OE of wild-type S6K and KD of Thor showed no suppression. (C) Confocal images of cryosectioned IFMs from 3-week-old adults expressing G489V stained with an antibody against Drosophila LamC (orange), DAPI (blue) and phalloidin (red). IFM-specific expression of DN S6K and OE of Thor suppressed the cytoplasmic aggregation of LamC, nuclear blebbing and disorganization of myofibrils. No suppression was observed with OE of wild-type S6K and KD of Thor. Arrows represent mutant LamC-induced enlarged nuclei, asterisks represent disorganized actin-containing myofibrils and arrowheads represent cytoplasmic aggregates of LamC. (D) Confocal images of cryosectioned IFMs from 3-week-old adults expressing wild-type LamC stained as in (D and D’). (E–H) Quantification of mutant LamC-induced aggregates, nuclear dysmorphology (blebs) and the percentage of enlarged and misaligned nuclei. Quantification of LamC aggregates and nuclear dysmorphology was performed as reported for Fig. 4F–I. Statistical indicators in red text are results obtained from comparisons to wild-type LamC; indicators in black text are results obtained from comparisons to the specific mutant LamC. Statistical analysis using one-way ANOVA with post hoc Tukey test (^*= P < 0.05, ^***= P < 0.001, ns = non-significant) was performed.

In addition to S6K phosphorylation, TOR phosphorylates 4E-binding protein 1 (4E-BP1) (orthologue of Drosophila Thor) (71,73), which triggers its release from the pre-translation complex formed on capped mRNAs and promotes translation (73). Previous studies revealed that Thor OE reduced age-related cytoplasmic protein aggregates in muscle (71,73). Here, Thor OE had no discernible effects on IFMs expressing wild-type LamC; however, it suppressed the held-up wing phenotype and flight defect in IFMs expressing mutant LamC (Fig. 6B). In agreement with the aging study, Thor OE reduced the cytoplasmic LamC aggregates, nuclear morphological defects and myofibril disorganization (Fig. 6C and E–H). RNAi KD of Thor enhanced the cellular phenotypes and did not suppress the held-up wing phenotype and flight defect (Fig. 6A and B). Taken together, these data suggest that down-regulation of the TOR pathway suppressed muscle dysfunction by reducing protein synthesis through S6K and Thor phosphorylation, promoting muscle health. Additionally, we have that the UAS-GFP RNAi and UAS-GFP OE did not affect muscle function in flies with a wild-type genetic background and flies expressing wild-type and mutant LamC (Supplementary Material, Fig. S6 D and E). Over-expression of GFP did not suppress G489V-mutant induced aggregation (Supplementary Material, Fig. S6 B).

To further understand the impact of these suppressors and the mechanistic basis of the mutant LamC-induced muscle disease, we evaluated the influence of each suppressor (AMPKα, dPGC-1, Foxo, S6K and 4E-BP) on protein quality control using reduction in the overall levels of LamC as an indicator. OE of AMPKα and other modulators in the presence of G489V reduced overall levels of LamC to varying extents (Supplementary Material, Fig. S7A). The greatest reduction was observed with AMPKα OE and the least with Foxo OE. These modulations did not impact levels of LamC in the wild-type background suggesting that these genetic modulators maintain protein quality control by removing LamC aggregates, possibly by activating autophagy.

The suppressors were also evaluated for effects on mitochondria using mito-GFP driven by Fln-Gal4. Muscle-specific expression of mito-GFP caused no apparent mitochondrial defects in flies expressing wild-type LamC and the Gal4 driver alone (Figs 4–7 and Supplementary Material, Fig. S4), allowing for analysis of mitochondria in the presence of genetic modifiers. OE of AMPKα, dPGC-1, 4E-BP and inhibition of S6K suppressed G489V-induced mitochondrial fragmentation and dysmorphology (Fig. 7A–A” and Supplementary Material, Fig. S7B) to nearly that observed in wild-type controls (Fig. 7B–B” and D–D” and Supplementary Material, Fig. S7B). OE of Foxo only partially rescued G489V-induced mitochondrial dysfunction (Fig. 7A–A” and Supplementary Material, Fig. S7B). However, OE of Foxo suppressed G489V-induced LamC aggregation and myofibrillar disarray, but not the held-up wing phenotype and flight defect (Fig. 5). Mitochondrial dysfunction can produce the held-up wing and flightless phenotype; therefore, the lack of suppression of these defects might be due to the residual mitochondrial defects (50).

Figure 7.

Suppression of mutant LamC-induced mitochondrial dysmorphology in the IFM. (A, A' and A”), Confocal images of cryosectioned IFMs from 3-day-old adults expressing G489V and mito-GFP (green), DAPI (blue) and phalloidin (red) in the absence or presence of each modulator. Quantification of G489V-induced muscle dysfunction and mitochondrial dysmorphology in the absence and presence of each modulator is shown in Supplementary Material, Fig. S7B. Expression of G489V resulted in muscle dysfunction and mitochondrial dysmorphology, which were suppressed by each genetic modulator to a varying extent. The least amount of suppression was obtained with Foxo OE, which resulted in small fragmented mitochondria (see Supplementary Material, Fig. S7B). (B, B' and B”) Representative confocal images showing mito-GFP localization in the absence and presence of each genetic modulator. (C, C' and C”) and (D, D' and D”) Representative confocal images of IFM from non-transgenic (w¹¹¹⁸) and Fln-Gal4; mito-GFP adults, respectively.

To determine how generalizable these findings were among mutant versions of LamC, we performed similar genetic tests with flies with IFM-specific expression of R205W, which in an amino acid substitution in the rod domain (Fig. 1A). Effects of IFM-specific OE of AMPKα, Foxo, dPGC-1 and Thor were investigated (Figs 4–6, Supplementary Material, Table S3). OE of AMPKα, Foxo and dPGC-1 suppressed the defects caused by R205W(Supplementary Material, Table S3). Surprisingly, OE of Thor caused only minor improvements in the cytological defects and did not suppress the held-up wing phenotype and flight defect (Supplementary Material, Fig. S8). Collectively, these data suggest that both common and personalized therapies might be successful for individuals with laminopathies.

Overall, using gain- and loss-of-function genetic manipulations in Drosophila, we showed that regulation of the AMPK and TOR signaling pathways had beneficial effects in the context of muscle laminopathy. The transcriptomic data from human diseased skeletal muscle showed misregulation of genes in these pathways. Subsequent tests of genetic suppression in Drosophila supported their involvement. For the first time, we have evaluated members of TOR (i.e. S6K, 4E-BP) and AMPK (i.e. Foxo, PGC1α) signaling and their association with cytoplasmic protein aggregation in the context of muscle laminopathy. Taking these data into consideration, we have developed a model that links cytoplasmic aggregates to cellular signaling pathways that impact muscle health (Fig. 8). Our findings imply that modulation of the AMPK, TOR and autophagy pathways serve as new potential targets for therapy.

Figure 8.

Model for the molecular basis of the muscle pathology. Expression of mutant lamins in muscle causes cytoplasmic aggregation of NE proteins (cloud-shaped images). This leads to the up-regulation (up arrow) of the autophagy chaperone Ref(2)P/p62, which binds to the aggregates. Accumulation of Ref(2)P/p62 causes down-regulation of autophagy (down arrow) and up-regulation of mTOR. Down-regulation of autophagy leads to inactivation of AMPKα and downstream effectors, thereby causing mitochondrial defects and loss of energy homeostasis. Abundant p62 also causes up-regulation of RPTOR (called raptor in Drosophila), which binds to TOR and inhibits autophagy. Up-regulation of TOR causes increased S6K activity contributing to the loss of energy homeostasis and muscle dysfunction. Asterisks represent protein products of genes tested with both gain and loss of function alleles.

Discussion

Although several hundred mutations in the LMNA gene have been identified and many studies have been performed on lamins, the pathogenic mechanisms of laminopathies remain not well understood. Greater insights are needed for therapeutic interventions. To address the molecular pathology of laminopathies and to understand the functions of lamins, we developed Drosophila models of skeletal myopathies. Mutations in Drosophila LamC were generated that are analogous to human LMNA mutations (Fig. 1A) and expressed exclusively in the IFM, a muscle that produces a readily visible held-up wing phenotype (Fig. 1D) upon muscle dysfunction.

Three of the four LamC mutants studied here (R205W, G489V and V528P) caused severe muscle defects upon expression with Act88F and Fln Gal4 drivers (expressed before sarcomere assembly/maturation). In contrast, A177P caused only moderate functional defects when expressed with the same drivers, despite similar levels of LamC protein (Fig. 1B, Supplementary Material, Fig. S1A, B and D). These data demonstrate that the severity of the abnormal phenotypes is mutation-specific. This is similar to the human disease condition in which individuals with different LMNA mutations exhibit a wide range of disease severity depending on the location of the amino acid substitution (47). Expression of the mutant lamins after sarcomere assembly/maturation via the DJ-694 Gal4 driver resulted in only moderate functional defects (Fig. 1C). Taken together, these data suggested that mutant LamC interfered with sarcomere assembly/maturation. This idea was supported by TEM images showing disruption of sarcomere organization when using the Act88F and Fln Gal4 drivers (Fig. 1E). During sarcomere assembly, several proteins are produced de novo and presence of cytoplasmic LamC aggregates might interfere with the formation of multi-protein complexes that are associated with the contractile apparatus. It is also possible that cytoplasmic LamC aggregates interfere with proteostasis by sequestering chaperone proteins that facilitate protein folding following de novo synthesis (54). Loss of sarcomere structure and mitochondrial defects are known to cause the held-up wing phenotype and loss of flight (Fig. 1D). Both of these phenotypes are useful phenotypes for drug screens (29,49,50). The fact that we identified mutation-specific variation in muscle disease severity further suggest that the Drosophila models will be useful for identifying modifier genes, which provide another level of complexity with regard to the range of disease severity observed in individuals, including family members with the same LMNA mutation.

To better understand the molecular and cellular basis of the muscle pathology, we performed an in-depth analysis of the Drosophila models. Cytological analysis revealed cytoplasmic aggregates of LamC and nuclear pore proteins, nuclear blebbing, disruption of the cytoskeletal organization and mitochondrial morphology (Fig. 2 and Supplementary Material, Fig. S4). During the natural aging process, accumulation of aggregates often results from defective proteostasis (54). Interestingly, protein aggregation in Huntington disease leads to amyloids that cause sarcomeric assembly defects due to loss of proteostasis (54). We propose that the abnormal accumulation of cytoplasmic NE protein aggregates leads to an impairment of proteostasis, causing loss of muscle function (Figs 2 and 3). Our findings are consistent with cytoplasmic aggregation of NE proteins in muscle biopsies from individuals with skeletal muscle laminopathy (16,25). Thus, our results show that the IFM defects in the Drosophila models share characteristics with the human diseased muscle.

To further define the pathological mechanisms of mutant LamC in skeletal muscle, we have examined the effects of mutant LamC on autophagy and metabolic signaling. We found that Ref(2)P, the Drosophila homologue of mammalian polyubiquitin binding protein p62, is up-regulated in IFM expressing mutant LamC relative to controls (Fig. 3B and C). Misregulation of nuclear factor (erythroid-derived 2)-like 2 (Nrf2)/Keap1 redox signaling mediated by p62 has also been associated with muscle atrophy and cardiomyopathy, and this pathway is predicted to influence autophagy (16,21,74,75). p62 is an adaptor protein that binds protein aggregates and targets them for autophagy and proteasome-based destruction (58,59). However, it is unknown how p62 influences laminopathy-mediated autophagic defects (21). Studies in mice show that loss of A-type lamins leads to cardiac and muscle defects due to alterations in mTOR signaling, which influences the rate of authophagy (17,18). Autophagy is responsible for the regulation of lamin B1 levels (20). Thus, it is possible that autophagy flux is impaired by NE protein aggregates. This might lead to the persistence of defective mitochondria (Fig. 3A), resulting in the up-regulation of the TOR pathway (2,21), which in turn contributes to the down-regulation of autophagy.

Transcriptomic data from human laminopathy muscle allowed us to (1) obtain unbiased insights into the gene expression profile of human diseased muscle, (2) compare the data obtained with Drosophila to human diseased muscle to validate the use of our models and (3) establish the translational potential of the Drosophila models. Based upon the knowledge gained from the RNA sequencing of human muscle biopsy tissue, we identified pathogenic pathways and then modulated those pathways using the rapid genetics offered by Drosophila. Through these genetic manipulations, we were able to reduced (and eliminated in some cases) NE protein aggregation and alter intracellular signaling to ameliorate muscle dysfunction. Through modulation of AMPK, PGC1α, Foxo, S6K and 4E-BP, we identified key players that regulate autophagy in suppressing laminopathy-induced skeletal myopathy and mitochondrial dysmorphology (Figs 4–7). The pathway components identified might serve as valuable disease markers and provide new targets for the development of rational therapeutic strategies.

Based on our human muscle transcriptomics and genetic manipulations in Drosophila, we showed that activation of AMPK suppressed muscle laminopathy (Fig. 4). AMPK is a sensor of cellular energy and metabolism that is linked to regulating autophagy, proteostasis and mitochondrial function (63–66). AMPK has conserved functions in many species, including Drosophila, and occurs universally as heterotrimeric complexes containing catalytic α-subunits and regulatory β-and γ-subunits (67). Increased expression of AMPK prevented age-related phenotypes in old mice, such as weight gain and decline of mitochondrial function (47). Activation of the AMPK pathway improved lamin-induced myopathy by removing abnormal aggregates (Supplementary Material, Fig. S7A), achieving autophagic and mitochondrial homeostasis (68) (Figs. 5C and 7). Consistent with these findings, the AMPK activator metformin lowered progerin (a specific mutant form of lamin A/C) levels and suppressed defects in the HGPS-induced pluripotent stem cell model (76).

Our data extend these findings by showing that the positive effects of AMPK activation are mainly through PGC1α, with contributions from Foxo, both of which maintain metabolic and cellular homeostasis. Previously, it was shown that Foxo/4E-BP signaling regulates age-induced proteostasis, including suppression of age-associated aggregation in skeletal muscle (71). As observed with rapamycin treatment in mouse models (17,18), activation of 4E-BP, a key downstream effector of the mTOR complex (73), is thought to reduce TOR activity. Muscle-specific expression of 4E-BP suppressed age-related protein aggregates and metabolic defects in Drosophila and mouse models (71,77). However, whole-body OE of 4E-BP1 shortened the lifespan of Lmna^-/- mice possibly by enhancing lipolysis (19). In the Drosophila IFM models, activation of S6K enhanced muscle deterioration and a DN version of S6K suppressed muscle dysfunction, presumably by activating autophagy as evidenced by the reduction of cytoplasmic aggregation of NE proteins (Fig. 6). Overall, we identified specific downstream targets of AMPK that suppress muscle laminopathy.

Based upon our findings and those in the literature, we propose a model that describes how cytoplasmic aggregates of NE proteins impact autophagy and signaling pathways and contribute to muscle pathology (Fig. 8). According to this model, cytoplasmic aggregation of NE proteins lead to increased levels of Ref(2)P/p62, which bind to the protein aggregates (16,25). Accumulation of p62 causes up-regulation of the TOR pathway that leads to inhibition of autophagy in the skeletal muscle (56,78). Accumulation of Ref(2)P/p62 also causes up-regulation of the regulatory associated protein of MTOR complex 1 (RPTOR), which binds mTOR and inhibits autophagy (78). Thus, autophagy is down-regulated by two mechanisms, causing a disruption in proteostasis. Moreover, up-regulation of the mTOR pathway causes increased S6K activity, which leads to imbalance in energy homeostasis (79,80). Consistent with this model, the transcriptomic data from the laminopathy muscle biopsy tissue showed up-regulation of RPTOR and S6K, implying that autophagy is down-regulated. Also, in support of this model, KD of S6K in Drosophila IFM suppressed the muscle defects (Fig. 6). Inhibition of autophagy is predicted to cause a reduction in AMPK activity (67). Consistent with this idea, all three AMPKα transcripts were down-regulated in the laminopathy muscle biopsy tissue. OE of AMPKα in Drosophila IFM suppressed the muscle defects (Fig. 4). AMPK inactivation leads to the activation of PI3K/AKT/mTOR pathway, which was also up-regulated in the human muscle biopsy. Another important function of AMPK is to control the expression of genes involved in energy metabolism and aging by enhancing the activity of sirtuin 1 (SIRT1) (81). SIRT1 controls the activity of downstream targets such as PGC-1α, the master regulator of mitochondrial biogenesis, and Foxo, which is involved in delaying the aging process, by reducing protein aggregation through controlling its target 4E-BP (82–85). The transcriptomic data showed that SIRT1 and downstream targets, PGC-1α, Foxo and 4E-BP, were down-regulated, which would cause an imbalance in cellular energy metabolism leading to cellular stress and compromising skeletal muscle function. In agreement, OE of dPGC-1, Foxo and 4E-BP in the Drosophila models suppressed the abnormal muscle phenotypes (Figs 5 and 6). Several kinases were up-regulated in the human muscle biopsy tissue (Supplementary Material, Table S2). Up-regulation of these kinases has been observed in lamin-associated cardiomyopathy (86). Genetic modulation of these kinases (Supplementary Material, Table S2) is needed to test their effectiveness in suppressing muscle laminopathy and other lamin-based disorders.

Overall, our data provide new insights on potential targets for small molecular screens. As proof-of-principle, dietary supplementation of rapamycin (TOR inhibitor) or 5-Aminoimidazole-4-carboxamide ribonucleotide (AICAR), activator of AMPK, in Drosophila media suppressed the mutant LamC-induced muscle defects (Supplementary Material, Table S3). Drosophila allows evaluation of the effects of these compounds on functional and cellular defects caused by mutant LamC in the context of a whole organism. Promising compounds can then be tested in pre-clinical mouse laminopathy models (20,29). Given that lamin-associated muscular dystrophies have pathophysiological features shared with other laminopathies and diseases such as diabetes, our findings have the potential for broad impact.

Materials and Methods

Drosophila stocks, gene expression system and tests of genetic modifiers

Drosophila transgenic stocks expressing wild-type UAS-LamC and mutant versions of LamC (R205W, G489V and V528P) were generated as previously described (16,21,25,28). Briefly, the desired mutation was introduced into full length Lamin C complementary DNA (cDNA) that was amplified by polymerase chain reaction (PCR) from RNA isolated from 18- to 21-hour Drosophila embryos (Clonetech, Palo Alto, CA). The cDNA was cloned into pCR2.1-TOPO (Invitrogen, San Diego, CA) and used as a template for site-directed mutagenesis (Quick Change Kit, Qiagen, Hilden, Germany) to generate the A177P construct; primer sets: 5′ GAAGTTCGAGAGGATCAGCCCAAGGAGCTCGTCTG3′ (forward) and 5′ CAGAGCGAGCTCCTTGGGCTGATCCTCGAACTT (reverse) were used. The cDNA possessing the desired mutations was cloned into the pUAST P element transformation vector and transformants were generated using standard procedures (16,25).

Wild-type and mutant LamC were expressed in IFM using the Act88F (Act88F), flightin (fln) and DJ-694 Gal4 drivers (44,45). Act88F-Gal4 and Flightin-Gal4 drivers express the yeast Gal4 transcription factor in the IFM during sarcomere assembly/maturation (44,45) and DJ694-Gal4 expresses Gal4 in the IFM after sarcomere assembly/maturation (44). Flies possessing each of these Gal4 drivers were independently crossed to flies possessing each of the UAS-LamC transgenes. F1 progeny of each transgenic cross were collected and maintained at 22°C in standard sucrose and corn meal medium (54). Male and female flies were aged separately for the examination of age- and mutation-dependent functional, cytological and ultrastructural and biochemical defects linked with IFM-specific expression of the LamC transgenes.

Flies possessing the Fln-Gal4 driver and the recombinant Fln-Gal4:UAS-mitoGFP were used for genetic tests of enhancement and suppression of the muscle phenotypes and muscle mitochondria health, respectively. Recombinant stocks were made by first placing the Gal4 driver over a marked balancer chromosome (possessing multiply inverted chromosome segments to prevent recombinants) possessing a dominant marker by crossing the Fln-Gal4 (w; Fln/CyO; +/+) with the A-12 line (w; CyO/Bl; TM2/TM6B). In the F1 generation, progeny with the genotype w; Fln/CyO; +/TM6B and w; Fln/CyO; +/TM2 were collected. These two genotypes were self-crossed to obtain w; Fln/CyO; TM2/TM6B. These flies were then crossed to each of the LamC transgenic stocks that were also maintained over a marked balancer chromosome (example: w; Fln/CyO; TM2/TM6B X w; CyO/Bl; G489V/G489V). In the F-1 generation, flies (both males and females) having the genotype w; Fln/CyO; G489V/TM6B were collected and maintained as a w; Fln/CyO; G489V/G489V stock for tests of enhancement and suppression.

For genetic tests of potential enhancement and suppression, flies expressing the LamC transgenes were crossed with each of the candidate modifier genes. For example, flies from the AMPK OE stock (w; +/+; AMPK OE/AMPK OE) were crossed with w; Fln/CyO; G489V/G489V and the progeny with the genotype w; Fln/+; G489V/ AMPK OE were collected and used for functional and cytological studies. Similar crosses were performed for potential modifier genes on the third chromosome. For candidate modifiers on the second chromosome, such as w; PGC-1α OE/PGC-1α OE; +/+, flies were crossed with w; Fln/CyO; G489V/G489V and progeny having the genotype w; Fln/PGC-1α OE; G489V/+ were collected. In addition, flies from the stock w; Fln/CyO; G489V/G489V were crossed to y, w; LamC^Ex296/CyO; +/+; resulting progeny were analyzed to determine the effects of mutant LamC in a background heterozygous for a null allele of LamC. Genetic crosses for the recombinant Fln-Gal4:UAS-mitoGFP/CyO were performed as described above for Fln-Gal4 driver.

Functional analysis of muscle

Muscle function was assessed by performing flight tests on adults with IFM-specific expression of wild-type and mutant LamC as described (29,39). Briefly, flies were released into the center of a Plexiglas box (43 cm high × 27.5 cm wide × 43 cm long) with a light source positioned at the top. Points were assigned based the ability of the adult to fly upwards (6.0), horizontal (4.0), down (2.0) or no flight at all (0.0). The average FI was calculated by dividing the sum of the individual FI values by the number of individuals for each group. The FIs were calculated for flies of different ages and genotypes.

Cytological analysis of muscle

Immunostaining was performed on IFM from at least ten thoraces from 3-day- and 3-week-old adults (51,52,54). The thoraces were dissected and fixed in a 4% PFA solution for 20 min. The fixed thoraces were washed thrice in PBS with 10 min incubation time between each wash. The thoraces were then aligned longitudinally in a cryomold filled with optimal cutting temperature (OCT) compound and flash frozen in dry ice. The fixed thoraces were cryosectioned at 30 nm thickness and the samples were washed in 1X PBST (phosphate buffered saline with Tween-20) containing 1% Triton X-100 thrice with 15 min incubation between each wash. Antibody LC 28.26 (anti-mouse, 1:100 dilution; Developmental Studies Hybridoma Bank, University of Iowa) was used to detect Drosophila LamC. Antibody Mab414 (anti-mouse, 1:100 dilution, Abcam, Cambridge, United Kingdom) was used to recognize nuclear pore proteins with FG-repeats. An antibody directed against muscle myosin (1:400 dilution, a generous gift from Dr Kiehart, Duke University) was used to detect the myosin containing myofibrillar organization in the IFM. Dylight 657 goat anti-mouse was used as the secondary antibody to detect LC 28.26 and Mab414 and goat anti-rabbit CFL 488 (Santa Cruz Biotechnology, Dallas, TX) was used to detect the myosin antibody. A primary antibody against Ref(2)P (1:100 dilution, a generous gift from Dr Kim Finley, San Diego State University) and a goat anti-rabbit CFL 488 secondary antibody were used to detect Ref(2)P. The muscle samples were also stained with 0.1 μM phalloidin (TRITC) and mounted with Vectashield Mounting Medium containing DAPI (Vector Laboratories Inc., Burlingame, CA). For analyses using mito-GFP, muscle samples were fixed as described above, stained with 0.1 μM phalloidin Tetramethylrhodamine (TRITC) and mounted with Vectashield Mounting Medium containing DAPI. Imaging was performed using a Zeiss 710 Confocal Microscope.

Quantification of cytoplasmic aggregates, mitochondrial dysmorphology/fragmentation, nuclear dysmorphology and blebs and misaligned nuclei were performed using ImageJ as previously described (21,54). Statistics were performed for all quantitative measurements using one-way analysis of variance (ANOVA) with post hoc Tukey test (^*= P < 0.05, ^**= P < 0.01, ^***= P < 0.001 and NS = not significant) was performed. Briefly, the total area (μm²) occupied by the aggregates (or mitochondria) per micrograph was determined from a minimum of six confocal images representing three independent adults for each genotype using ImageJ. The relative area (or percentage) was calculated by dividing the average area of the aggregates (or mitochondria) in muscle expressing G489V by the average area of aggregates (or mitochondria) in muscle expressing the wild-type LamC control. A ratio of these averages (average for G489V: average for control) was calculated. The ratios were normalized, with the control sample set to one. A similar procedure was performed for determining the relative area of the aggregates (or mitochondria) in the presence of genetic modifiers and compared to values obtained in the absence of the genetic modifier.

For the quantification associated with nuclear phenotypes, the average area of the nuclei (μm²) in myonuclei expressing wild-type and mutant LamC was determined using ImageJ for a minimum of six confocal images (total nuclei > 50 per genotype) representing three independent adults. Expression of mutant LamC resulted in myonuclei that were significantly larger than the control myonuclei. A ratio of the average nuclear area of myonuclei expressing mutant LamC over the average area of myonuclei expressing wild-type LamC was determined. The ratios were normalized, setting the control to 1. In addition, the impact of each modulator (ratio of modular with G489V/modulator with wild-type LamC) on nuclear area was calculated in a similar manner. For quantification of nuclear blebbing, a ratio of major/minor myonuclear radius was measured for nuclei expressing wild-type LamC and mutant LamC. Statistical differences in this ratio between the control and experimental samples were considered as a `blebbed nuclei’. The percentages were reported as relative values compared to the control, which was set at one. Similar measurements and calculations were made in the presence and absence of genetic modifiers. For measurements of nuclear alignment, the distance between two neighboring myonuclei measured in six different images (total nuclei > 50 per genotype) from three independent flies for each genotype using ImageJ. The average relative distance between myonuclei was reported, with the average distance between myonuclei in the control sample given a value of one. The resulting values were reported as a percentage of the control. Statistical differences in the percentages calculated for muscles expressing mutant versus wild-type LamC were considered misaligned. In a similar manner, the impact of each modulator on nuclear misalignment was determined.

Ultrastructure analysis of muscle

TEM was performed on stage 16 pupae and 3-day-old adults with IFMs expressing mutant LamC and compared with those expressing wild-type LamC and Fln-Gal4/+ controls as described (29,54,87). TEM analysis was used to determine the impact of each LamC mutant on myofibrillar organization, cytoskeletal integrity, architecture of the NE and mitochondria as previously described (29,54,55). Quantification of sarcomere length, intact M-lines and Z-discs and mitochondrial defects was carried out using ImageJ. Specifically, for the evaluation of intact M-lines and Z-discs in EM figures (Figs 1E–F, 3A and C and Supplementary Material, Fig. S5), two criteria were evaluated. (1) The M-line and Z-disc were scored as intact if they were continuous along the sarcomere width. For example, in the control sample, the M-line and Z-disc lengths were identical to the width of sarcomere. In contrast, in muscle expressing mutant LamC the M-lines and Z-discs are completely or partially missing (Figs 1E, 3A and Supplementary Material, Fig. S5, see arrows). (2) The alignment of the M-lines and Z-discs within the sarcomere was determined by drawing a straight line on each using ImageJ. While almost all the M-lines and Z-discs showed up as straight lines in the Gal4 driver only and wild-type LamC controls, muscle expressing mutant LamC showed misalignment (Figs 1E and 3A, see arrows). The threshold for a normal M-line or Z-disc was that it was both straight and intact. Any tracings of the M-lines and Z-discs that did not yield a straight and intact line were counted as abnormal.

Western and immunodot blot analysis

LamC protein levels were determined by performing western analysis on protein extracts from dissected IFM from 3-day-old adults. Briefly, 10 thoraxes were added to 125 μl of lysis buffer (62 mm Tris pH 7.5, 2% SDS (sodium dodecyl sulfate), protease inhibitors), and the samples were sonicated for 30 s using an E220 Covaris Sonicator. The samples were boiled for 5 min at 95^oC and centrifuged to collect the supernatant. Sample loading buffer was added and then the mixture was boiled for 1 min at 95^oC. The samples were then electrophoresed on AnykD Mini-PROTEAN TGX Precast Gels (Bio-Rad, Hercules, CA). The protein bands were transferred to polyvinylidene difluoride (PVDF) membrane (Bio-Rad, Hercules, CA) and incubated with LC 28.26 anti-mouse antibody (1:300; Developmental Biology Hybridoma Bank, University of Iowa) and beta-actin antibody (1:1000, Cell Signaling, Danvers, MA, USA). For the secondary antibody, horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgG were used at 1:1000 dilutions. SuperSignal West Pico chemiluminescent substrate was used to detect the signal. Signal from the membranes was collected and the data quantified using ImageJ software. Western analysis was performed on three independent biological replicates for each genotype. Immunodot blots were performed using dissected thoraxes (10 IFMs/genotype) as described (41,54) using antibodies that recognize Ref(2)P (1:1000 dilution; gift from Kim Finley) and LC 28.26 anti-mouse antibody (Development Studies of Hybridoma Bank, University of Iowa). Antibodies that recognize histone H2B (Cell Signaling, Danvers, MA, USA) were used as a control.

RNA-seq of human muscle biopsy tissue and bioinformatics

A de-identified flash frozen human quadriceps muscle biopsy tissue from an individual with skeletal muscle laminopathy (LMNA 1346G > T) and an age-matched control were obtained from the Wellstone Muscular Dystrophy Cooperative Research Center, University of Iowa (16,25,28). This mutation produces lamin A/C G449V, which is analogous to Drosophila lamin C G489V. Total RNA was extracted from the muscle tissue and RNA-seq was performed to generate an average of 75 million reads per sample in the Illumina 2 × 100 base read format on the Illumina HiSeq 2500 platform (ribo-zero) (30,88). Sequenced reads were mapped to the genome, normalized and quality control measures were performed; only reads aligning uniquely to the genome were considered (30,88). The reference sequence from Ensembl GRCh38 and annotation files from Ensembl BioMart were used for data analysis. Gene products of the transcripts were functionally annotated using DAVID, BIND and STRING online tools (60–62) to identify pathways and functional clusters that are affected by the LMNA mutation. The Student t-test was used to identify differentially expressed transcripts as described (30,88).

Statistical analysis

Quantitative functional, structural and biochemical parameters collected from the control, experimental and rescued groups (i.e. both sexes, at least three replicates from independent adults) were analyzed using one-way ANOVA with post hoc Tukey test. Multiple comparisons were used to assess the statistical significance of differences among groups with Prism (Graph Pad) software as previously described (54). Significant differences were assumed for P-values < 0.05.