Autophagy increase in Merosin-Deficient Congenital Muscular Dystrophy type 1A

Abstract

The autophagy process recycles dysfunctional cellular components and protein aggregates by sequestering them in autophagosomes directed to lysosomes for enzymatic degradation. A basal level of autophagy is essential for skeletal muscle maintenance. Increased autophagy occurs in several forms of muscular dystrophy and in the merosin-deficient congenital muscular dystrophy 1A mouse model (dy3k/dy3k) lacking the laminin-α2 chain. This pilot study aimed to compare autophagy marker expression and autophagosomes presence using light and electron microscopes and western blotting in diagnostic muscle biopsies from newborns affected by different congenital muscular myopathies and dystrophies. Morphological examination showed dystrophic muscle features, predominance of type 2A myofibers, accumulation of autophagosomes in the subsarcolemmal areas, increased number of autophagosomes overexpressing LC3b, Beclin-1 and ATG5, in the merosin-deficient newborn suggesting an increased autophagy. In Duchenne muscular dystrophy, nemaline myopathy, and spinal muscular atrophy the predominant accumulation of p62+ puncta rather suggests an autophagy impairment.

Ethical Publication Statement

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Congenital myopathies and muscular dystrophies are hereditary myopathies featuring weakness, hypotonia, and characteristic features on muscle biopsy.¹ In Italy, congenital merosin-deficient muscular dystrophy (MDC1A; OMIM #607855) is a rare disease with a prevalence of 0.136 per 100,000 in the all-age population accounting for 24.11% of congenital muscular dystrophies (CMDs) ^2,3 in which laminin-α2 (LAMA2), a large glycoprotein of the basal lamina, is reduced or absent.^4,5 Nemaline myopathies (NMs) are a heterogeneous group of CMDs caused by de novo dominant or recessive mutations in at least 12 genes being similarly rare muscle diseases with an overall prevalence of 0.2 per 100.000.⁶ Another congenital disease presenting with profound muscle weakness derives from irreversible loss of motor-neurons of spinal cord and brain stem nuclei (spinal muscular atrophy type 0, SMA 0) being one of the diseases for primary differential diagnosis with CMDs.

Autophagy is a catabolic pathway in which altered intracellular components are sequestered in autophagolysosomes for degradation. In the LAMA2 deficient mouse (dy3k/dy3k) model of MDC1A, there is increased expression and function of autophagy molecules; systemic administration of 3-methyladenine, a PI3K inhibitor of autophagosome formation, improves the dystrophic phenotype, increasing lifespan and motility.⁷ To determine whether an enhanced expression of autophagy-related molecules is a general, aspecific reaction of myofibers to different noxae or limited to specific muscle dystrophies, we analyzed by immunohistochemistry (IHC), immuno-electron microscopy (IEM) and western blotting (WB), the expression of LC3b, a late phase marker of autophagy and p62/SQSTM1, a shuttle receptor of LC3. Beclin-1, an early phase marker of phagophore formation and ATG5, an intermediate phase marker, were also morphologically evaluated. We compared the densities of IHC expression of LC3 and p62 in diagnostic muscle biopsies of newborns affected by MDC1A, NM10, SMA 0 with biopsies of Duchenne muscular dystrophy (DMD)-affected boys, and control muscle tissues.

Table 1.

Clinical, laboratory and morphometric findings of patients.

Diagnosis	Age at biopsy	Sex	Onset clinical presentation	Summary of MRI findings	Creatine kinase at biopsy time (IU/L)2
MDC1A (congenital merosin-deficient muscular dystrophy)	10 days	M	Significant postnatal hypotonia, weak feeding and crying	Normal signal intensities on T1--and T2-weighted images of brain GM and WM	43765
NEM10 (Nemaline myopathy 10)	8 days	F	Low foetal movements, significant postnatal hypotonia, no spontaneous movements and no elicitable reflexes, weak feeding	Normal signal intensities on T1--and T2-weighted images of brain GM and WM	ND
SMA 0 (Spinal muscular atrophy 0)	13 days	F	Respiratory distress at birth and significant postnatal hypotonia	Normal signal intensities on T1--and T2-weighted images of brain GM and WM	ND
DMD (Duchenne-type muscular dystrophy)	5 years	M	Frequent falls, kyphoscoliosis, waddling gait, respiratory distress	ND	792
DMD (Duchenne-type muscular dystrophy)	7 years	M	Weakness, short stature, difficult running, Gowers sign,	Neck flexors, biceps, triceps, quadriceps hypotrophy and fibro-fatty replacement	2189
IMNM (Immune-mediated necrotizing myopathy)	81	F	Limb proximal weakness, myalgia	Myofascial quadriceps, harmstring, and gluteal oedema.	2416
CTRL (control muscle tissue)	24 years	M	Asymptomatic, occasional elevated serum creatine kinase	ND	1543

Materials and Methods

Single open biopsies of muscle tissue obtained for diagnostic purposes from 3 newborns, 2 children, and 2 adult patients were retrospectively reviewed. The patients had clinical, pathological, and genetic confirmation of a specific myopathy (Table 1). Three newborns are affected by MDC1A, NM10, and SMA 0, the other two by DMD. As positive control of autophagy marker accumulation, we used muscle tissues from an adult patient diagnosed with immune-mediated necrotizing myopathy (IMNM).⁸ A normal control specimen was derived from an apparently healthy subject with mildly elevated serum levels of CK but without any histological abnormality, including the absence of myonecrosis, dystrophic features, and inflammatory infiltrates (Tables 1, 2 and 3). All procedures performed in this study were in accordance with the ethical standards of the 1964 Helsinki Declaration and later amendments. All patients or their relatives signed informed consent for diagnostic and research analyses and specimen inclusion in a muscle biobank.

Table 2.

Myopathological, genetic, and morphometric findings of patients.

Diagnosis	Mean area fraction of p62+ puncta	Mean number of p62+ puncta per myofiber	Mean p62+ puncta diameter	% of fibers with p62+ puncta	% of necrotic myofibers	% of type 2A myofibers
MDC1A	1.82	2.75±5.73	0.35	6.42	7.04	61.1
NEM10	1.25	4.04±4.24	0.36	14.27	1.28	6.8
SMA 0	1.32	3.33±3.21	0.34	13.07	0.14	7.5
DMD	2.64	6.5±1.97	0.58	19.98	1.15	29.7
DMD	2.72	5.07±1.88	0.61	18.92	0.96	35.8
IMNM	5.21	12.85±7.58	0.75	2.45	1.97	39.2
CTRL	0.58	0.6±0.89	0.14	0.91	0.0	22.8

Muscle Biopsy Analysis

The muscle biopsy specimens were rapidly frozen in isopentane, cooled in liquid nitrogen, and kept at −80°C until use. Serial, transverse, unfixed sections underwent standard histochemical and IHC stainings, as previously detailed.^9,10 Primary antibodies diluted in blocking buffer were used to recognize Beclin-1 (1:10; Abcam, code ab51031; Cambridge, United Kingdom), ATG5 (1:100; Abcam, code ab78073), LC3b (1:300; Abcam, code ab48394), p62/SQSTM1 (1:50; Abcam, code ab56416), laminin-α2-β1-γ1 (1:200; Sigma-Aldrich, code L9393; Darmstadt, Germany), laminin-α2 (merosin; 1:20; Monosan, clone Mer3/22B2; Uden, The Netherlands), laminin-α1 (1:200; Novus Biologicals, code CL3183; Minneapolis, MN, USA) and laminin-α4 (Novus Biologicals, code NBP2-42393). Negative controls were prepared by omitting the primary antibodies, pre-adsorbing the primary antibodies with an excess of Tableantigen when available, and mismatching the secondary antibodies. Immunolabelled sections were photographed under a Leica TCS SP5 confocal laser scanning microscope (Leica Microsystems, Mannheim, Germany) using a sequential scan procedure. Confocal images were taken with 20x, 40x, 63x oil lenses applying digital zoom. Z-stacks of serial optical planes (projection images) and single optical planes were analysed by ImageJ software (NIH, Bethesda, MD, USA). Quantitative evaluations were performed by F.G., blinded for patient characteristics and diagnosis in all biopsies, by means of computer-aided morphometry applied to microscopic images. Positive puncta were counted on 30 myofibers at least using the default ‘analyze particles’ plugin in ImageJ software. Puncta with an area of 0.1–1.767 μm² were quantitated for LC3b and p62. Results were expressed as average values ± SD. No statistical analysis was performed.

Table 3.

Other morphometric findings of patients.

Diagnosis	Mean area fraction of p62+ puncta	Mean number of p62+ puncta per myofiber	Mean p62+ puncta diameter	% of fibers with p62+ puncta	% of necrotic myofibers	% of type 2A myofibers
MDC1A	1.82	2.75±5.73	0.35	6.42	7.04	61.1
NEM10	1.25	4.04±4.24	0.36	14.27	1.28	6.8
SMA 0	1.32	3.33±3.21	0.34	13.07	0.14	7.5
DMD	2.64	6.5±1.97	0.58	19.98	1.15	29.7
DMD	2.72	5.07±1.88	0.61	18.92	0.96	35.8
IMNM	5.21	12.85±7.58	0.75	2.45	1.97	39.2
CTRL	0.58	0.6±0.89	0.14	0.91	0.0	22.8

IEM Analysis

Frozen muscle specimens from the MDC1A and NM10 were fixed in 2.5% glutaraldehyde and processed as previously detailed.¹¹ Ultra-thin sections were observed using a transmission electron microscopy Morgagni 268 (FEI Company, Hillsboro, OR, USA). Immunoelectron microscopy was performed as previously detailed.¹² Briefly, ultrathin sections were mounted on formwar-coated gold grids and incubated overnight with the LC3b antibody (Abcam) diluted 1:300 in blocking buffer (TBS 0.1 M, pH 7.4 + BSA1%) at 4 °C. After rinsing with TBS (0.1 M, pH 7.4), the grids were incubated at a dilution of 1:20 of 20-nm gold-conjugated anti rabbit IgG (Sigma) in TBS for 1 hr at RT. After several rinses in TBS, the grids were lightly stained with uranyl acetate and lead citrate. The grids were observed under a Morgagni 268 electron microscope (FEI).

WB Analysis

Frozen muscle specimens (20 mg) from the MDC1A, NM10, anti-HMGCR⁺ IMNM, DMD, and control biopsy were transferred into precooled tubes and immediately solubilized by using an Ultra Turrax T8 durable homogenizer (IKA, Wilmington, North Carolina) in at least 10 volumes of extraction buffer (radio-immuno-precipitation assay buffer; Thermo Fisher Scientific, Waltham, Massachusetts) added with a cocktail of protease inhibitors (Roche Diagnostics, Mannheim Germany). The immunoblotting method was previously detailed.⁹

The membranes containing the blotted proteins were incubated with primary antibodies LC3b (1:750; Abcam), p62 (1:1000; Abcam), and glyceraldehyde 3- phosphate dehydrogenase (GAPDH; 1:3,000; Sigma, St Louis, Missouri) overnight at 4°C. After having been washed with 0.1% Tween 20 PBS, the membranes were incubated with near-infrared fluorescent secondary antibodies IRDye 800CW anti-rabbit (LI-COR, Lincoln, Nebraska) or IRDye 800CW anti-mouse (1:7,000; LI-COR) secondary antibodies for 1 h at room temperature. For immunoblotting analysis, the LI-COR Odyssey infrared imaging system was used.

Results

The newborns were admitted to the neonatal intensive care unit of Bari University Hospital for severe hypotonia, respiratory distress at birth and weak feeding (Table 1). On admission, complete blood count, serum electrolytes, bilirubin, lactic acid, and ammonium levels were normal. Muscle enzymes were increased, especially CK levels (Table 1). Needle electromyography showed abnormal findings, whereas brain MRI was normal (Table 2). Muscle biopsy was performed, demonstrating the presence of necrotic myofibers and other specific findings (Table 3). The impact of muscle biopsy on the diagnostic workflow was evident in the MDC1A patient, whose muscle tissue showed a complete lack of LAMA2 immunohistochemical staining around myofibers (Suppl. Fig. 1A, C).

Compared to the control muscle (Suppl. Fig. 1B), a compensatory hyper-expression of laminin-α1 and -α4 (Suppl. Fig. 1D, E) was observed, together with hyper-expression of the other laminin subunits, detected using a pan-laminin antibody (Suppl. Fig. 1A). In addition, in the MDC1A muscle biopsy, immunofluorescence labelling showed that myofibers accumulated compartmentalized puncta positive for Beclin-1 (Fig. 1A), ATG5 (Fig. 1B), LC3b (Fig. 1C, F), and p62 (the shuttle receptor between LC3 and ubiquitinated proteins, Fig. 1E, L). Indeed, the comparison of different morphological parameters related to immunolabelling for two significant autophagy markers, LC3b and p62, among the different neuromyopathies revealed the highest density of compartmentalized puncta and % of myofibers accumulating LC3b in MDC1A compared to the other infantile neuromyopathies, whereas p62 appeared less expressed than LC3b in MDC1A (Fig. 1; Table 1). Adult control muscle was unstained with these autophagy markers (Fig. 1D, K, Q). A notable invasion of MHC-II⁺ inflammatory cells was noted in the analyzed neuromyopathies with no preferential association to autophagy puncta (Fig. 1F, H-J, L, N-P), an interesting feature instead noted in IMNM myofibers (Fig. 1G, M).

Fig 1.

Comparison of autophagy markers immunolabelling among different neuromyopathies. (A-C, E, F, L) Sections of MDC1A muscle biopsy double immunolabeled with anti-pan-laminin (red channel in A-C, F, M) and 4 different autophagy markers (green channel) such as anti-Beclin-1 (A; Abcam, code ab51031), an early phase phagophore formation marker, anti-ATG5 (B; Abcam, code ab78073), intermediate phase marker, anti-LC3b (C, F; Abcam, code ab48394), late phase marker, and anti-p62/SQSTM1 (E, L; Abcam, code ab56416), a shuttle receptor of LC3 from ubiquitin pathway. These different autophagy markers localize as green stained puncta in sarcoplasms of several myofibers (arrows). (C, F-J) LC3b+ stained puncta are more frequently seen in MDC1A and IMNM than DMD, SMA, and NM. (E, L-P) p62+ stained puncta are more frequently seen in DMD and IMNM than the other neuromyopathies. (D, K, Q) No LC3 and p62 immunostainings are normally detectable in sarcoplasmic adult healthy control muscle (CM). (F-Q) Both LC3b and p62 appeared preferentially localizing in proximity to MHC-II+ endomysial cells in IMNM (G, M), but separated in the other neuromyopathies. Scale bars: A-E: 50 μm; insets horizontal side in A (27 μm), C (47 μm), E (67 μm); F, L: 2 μm; G-K, M-Q: 10 μm.

Fig 2.

Transmission electron microscopy (TEM) images of autophagy. (A) The TEM image obtained from frozen sample of the MDC1A-affected muscle biopsy shows an autophagosome (arrowhead) appearing as a multilayered vesicle containing electron-dense material. In an earlier step, a three-layered membrane phagophore (white arrows) appears as an open bag (black arrow) enclosing a mitochondrion. (B) Immunogold electron microscopy image from the same MDC1A patient shows LC3B immune-conjugated 20 nm nanoparticles (arrows) localized on a membrane enclosing two mitochondria (m). (C) Multilayered membrane autophagosomes (arrowheads) are also localized in the perinuclear region (n) of myofibers from NM-affected newborn, where an early endosome (arrow) and dilated cisterns of endoplasmic reticulum (er) appeared in close proximity to the nuclear membrane. Scale bars: A, B: 0.5 μm; C: 2 μm.

Ultrastructural signs of autophagy activation, such as sarcoplasmic multilayered membrane vesicles (the autophagosomes) engulfed with amorphous dense materials or mitochondria, were observed in the MDC1A muscle tissue (Fig. 2A). At IEM, LC3 was localized in large phagophores containing mitochondrial remnants (Fig. 2B). Autophagosomes engulfed with whorls of endoplasmic reticulum (ER) membranes and damaged mitochondria were also observed in the NM perinuclear areas where dilated cisterns of ER appeared to expand to include sarcoplasmic material (Fig. 2C). Vacuolar degeneration typical of autophagic cell death and nuclear condensation typical of apoptosis and autosis were not observed at light and electron microscope in the biopsies, while necrotic (pale, swollen, or myophagocytosed) myofibers were more frequently seen in MDC1A muscle tissue than the other neuromyopathies (Table 1). WB analysis was carried out and revealed relatively higher LC3 and p62 expression in MDC1A than NM, IMNM, and DMD patients (Suppl. Fig. 2). Altogether, the morphometric results suggest an active autophagy pathway in which p62 and cargo are degraded and do not accumulate in MDC1A, whereas in the other infantile neuromyopathies both LC3 and p62 seem to accumulate. Only in the MDC1A patient there was an excessive prevalence of type 2A myofibers (61.1%) calculated on ATPase stainings (pH 4.3, 4.6, 9.4; Suppl. Fig. 3; Table 1), even if myofiber sizes of all types appeared within the normal range, as compared with previously published data.¹³

Discussion

In this pilot study, by IHC, IEM, and WB, we found hyperexpression of LC3, a marker of active autophagy, in MDC1A-affected muscle tissue. We compared this MDC1A-affected muscle tissue with those obtained for diagnostic purposes from other children affected by different neuromyopathies, NM10, SMA 0, DMD, from an adult subject affected by IMNM used as positive control,^l,8 and from an apparently healthy subject. The molecular activation of autophagy in MDC1A depends on intracellular signaling of LAMA2 via integrin binding involving the Akt and FOXO3 transcription factors, which inhibits autophagy.⁷ In the MDC1A mouse model, the lack of LAMA2 is associated with excessive autophagy, Akt/PKB inhibition, and muscle degeneration.⁷ Administration of 3-methyladenine (3-MA), an inhibitor of the autophagosome organizer Vps34, was sufficient to ameliorate the muscle morphology, locomotion and lifespan of the affected mice.⁷

In the MDC1A-affected muscle tissue, other autophagy markers were also expressed, namely Beclin-1 and ATG5, together with p62/SQSTM1, indicating an activated degrading flux in myofibers. Abundant p62/SQSTM1 accumulation seems prominent in DMD-affected myofibers than the other infantile neuromyopathies, at a level similar to that of adult IMNM. p62 accumulation could reveal impaired autophagy or lysosomal dysfunction, rather than abnormal activation.¹⁴ In fact, p62 accumulation has also been reported in the prototypical example of autophagy flux impairment in a muscle disorder, namely glycogen storage disease type II due to a defect in the lysosomal enzyme acid α-glucosidase.¹⁵ The shuttle molecule p62 has many roles in cellular processes, including selective autophagy of ubiquitinated cargos, cell survival, cell death, oxidative stress, DNA repair and inflammation.¹⁶

This acts as substrate of caspase 6 and 8,¹⁷ promotes fully activated cullin-3-modified caspase 8,¹⁶ and plays an important role in childhood-onset neurodegeneration,¹⁸ through damaged mitochondria clustering in the perinuclear region of cultured HeLa cells.¹⁹ At light and electron microscopy, morphological analysis of MDC1A and NEM10 demonstrates the presence of necrosis but the absence of autophagic cell death characterized by the presence of many vacuoles and autolysis with nuclear membrane shrinkage.²⁰ The increase of autophagosome density observed in the analyzed infantile neuromyopathies can be interpreted as either activation or impairment of autophagy. In fact, autophagy impairment causes accumulation of damaged organelles and autophagosomes as already found in DMD and mdx mice, a model of DMD disease, with persistent phosphorylation of Akt and mammalian target of rapamycin (mTOR) and corresponding downregulation of the autophagy-inducing genes LC3, Atg12, Gabarapl1 and Bnip3.²¹ Also in SMA mouse gastrocnemius, lower levels of LC3II, Beclin-1, and p62 proteins were observed in the pre-symptomatic stage.²² NMs have not been previously associated with altered autophagy. This study also provides preliminary evidence of autophagosome accumulation in myofibers of a NEM10-affected patient. In addition, autophagy flux impairment could increase unwanted antigen presentation of muscular neoepitopes and/or secretion of phagocytic “eat me” signals to macrophages.^8,9,23 This hypothesis is not corroborated by our observation of scattered MHC-II+ cells not preferentially associated with LC3+ or p62+ puncta. In addition, fast-twitching type II myofibers are mostly affected by impaired autophagy,²⁴ but could better survive increased activation of autophagy. The primary limitation of our study is the heterogeneous single cases of exceptionally rare neuromyophaties analyzed and, hence static morphologic assessment of autophagy instead of appropriate dynamic flux analysis. Nevertheless, our study focuses on two main autophagy markers, LC3 and p62, providing evidence that increased LC3 protein correlates with autophagy activity whereas p62 correlates with autophagy impairment.

In conclusion, our pilot study shows, in an MDC1A-affected newborn, upregulated LC3b, myofiber expression of other autophagy markers (Beclin-1, ATG5), and the presence of autophagosomes frequently internalizing concentric whorls of ER membranes and damaged mitochondria in the perinuclear areas. These results suggest that increased autophagy characterizes MDC1A pathogenesis, confirming the observations made in the dy3k/dy3k animal model of MDC1A.7 In DMD, NEM10, and SMA the accumulation of p62+ puncta suggests an autophagy impairment.

List of acronyms

Akt

serine/threonine kinase

ATG5

Autophagy-related 5

ATPase

adenosine triphosphatase

Bnip3

Bcl-2/adenovirus E1B 19kDa interacting protein 3

CK

creatine kinase

CM

control muscle

CMDs

congenital muscular dystrophies

CTRL

control

DMD

Duchenne muscular dystrophy

DNA

deoxyribonucleic acid

dy3k/dy3k

merosin-deficient congenital muscular dystrophy 1A mouse model

ER

endoplasmic reticulum

Fig.

figure

FOXO3

Forkhead box O3 transcription factor

Gabarapl1

GABA type A receptor associated protein like 1

HB

healthy babies

HMGCR

3-hydroxy-3-methyl-glutaryl-coenzyme A reductase

LAMA2

laminin-α2

IEM

immuno-electron microscopy

IHC

immunohistochemistry

IMNM

immune-mediated necrotizing myopathy

LC3

microtubule-associated protein 1A/1B-light chain 3

MDC1A

merosin-deficient congenital muscular dystrophy 1A

Mdx

mouse strain with a point mutation in its DMD gene

MHC-II

major histocompatibility complex type II

MRI

magnetic resonance imaging

mTOR

mammalian target of rapamycin

NM

Nemaline myopathy

p62/SQSTM1

polyubiquitin-binding protein p62/ Sequestosome 1

PI3K

Phosphoinositide 3-kinases

PKB

Protein kinase B

SMA 0

Spinal muscular atrophy type 0

TEM

Transmission electron microscopy;

Vps34

vacuolar protein sorting 34

WB

western blotting

Funding Statement

Funding: ork was supported by the Italian Ministry of Education, University and Research (MIUR, Finanziamento delle attività base di ricerca 2017 to FG).