Diverse myopathological features in the congenital myasthenia syndrome with GFPT1 mutation

Abstract Introduction Mutations in the GFPT1 gene are associated with a particular subtype of congenital myasthenia syndrome (CMS) called limb‐girdle myasthenia with tubular aggregates. However, not all patients show tubular aggregates in muscle biopsy, suggesting the diversity of myopathology should be further investigated. Methods In this study, we reported two unrelated patients clinically characterized by easy fatigability, limb‐girdle muscle weakness, positive decrements of repetitive stimulation, and response to pyridostigmine. The routine examinations of myopathology were conducted. The causative gene was explored by whole‐exome screening. In addition, we summarized all GFPT1‐related CMS patients with muscle biopsy in the literature. Results Pathogenic biallelic GFPT1 mutations were identified in the two patients. In patient one, muscle biopsy indicated vacuolar myopathic changes and atypical pathological changes of myofibrillar myopathy characterized by desmin deposits, Z‐disc disorganization, and electronic dense granulofilamentous aggregation. In patient two, muscle biopsy showed typical myopathy with tubular aggregates. Among the 51 reported GFPT1‐related CMS patients with muscle biopsy, most of them showed tubular aggregates myopathy, while rimmed vacuolar myopathy, autophagic vacuolar myopathy, mitochondria‐like myopathy, neurogenic myopathy, and unspecific myopathic changes were also observed in some patients. These extra‐synaptic pathological changes might be associated with GFPT1‐deficiency hypoglycosylation and altered function of muscle‐specific glycoproteins, as well as partly responsible for the permanent muscle weakness and resistance to acetylcholinesterase inhibitor therapy. Conclusions Most patients with GFPT1‐related CMS had tubular aggregates in the muscle biopsy, but some patients could show great diversities of the pathological change. The myopathological findings might be a biomarker to predict the prognosis of the disease.


INTRODUCTION
Congenital myasthenia syndrome (CMS) includes a large group of rare inherited endplate myopathies characterized by dysfunctions of neuromuscular junction transmission due to genetic defects (Engel et al., 2015;Finsterer, 2019). CMS shows great clinical and genetic heterogeneities characterized by abnormal fatigability, transient or permanent muscle weakness with varied age of onset. The main inheritance pattern of this disease is autosomal recessive, but a small part is inherited in autosomal dominant mode. There are at least 32 kinds of genes that have been identified in CMSs, while the number is still being updated (Iyadurai, 2020). Mutations in CHRNA1, CHRNB1, CHRND, or CHRNE are the most causative genes accounting for more than 30% of the cases, while mutations in RAPSN, COLQ, and DOK7 involve about 10% to 15% of the cases, and GFPT1 is accountable to approximately 3% of the cases (Engel et al., 2015;Finsterer, 2019).
Among the various types of CMS, the limb-girdle form is characterized by a muscle weakness and fatigability predominant in proximal muscles with minor or no involvement of ocular, facial, and bulbar muscles (Belaya et al., 2012). Mutations in the glutamine-fructose-  (Huh et al., 2012;Senderek et al., 2011).
Although genetic screening may be conveniently available for these patients through next-generation sequence (NGS), muscle biopsy is typically the first assessment conducted in these CMS patients who predominantly present with limb-girdle muscle weakness. Considering that some subtypes of CMS may be treatable genetic diseases, it is very important to make a timely diagnosis as early as possible (Farmakidis et al., 2018). Therefore, accurate identification of the various myopathological changes is very important to the diagnosis of CMS.
However, not all patients show tubular aggregates in muscle biopsy (Guergueltcheva et al., 2012), suggesting a need to re-recognize and summarize the diversity of muscle pathology in patients with CMS associated with GFPT1 mutations.
In this study, we described two CMS patients with GFPT1 mutations: one presented with vacuolar myopathy with myofibrillar destruction, and the other showed typical myopathy with tubular aggregates.
To further explore the pathological characteristics of CMS caused by GFPT1 mutation, we summarized the muscle pathological features in all reported GFPT1-related CMS cases with muscle biopsy.

Subjects
Patients with GFPT1 mutations were recruited from our in-home

Ethical statement
All patients' tissue samples were obtained after a written consent signed by each individual in compliance with the bioethics laws of China as well as the Declaration of Helsinki. The research was approved by ethics committee of the first affiliated hospital of Nanchang University.

Genetic screening
Genomic DNA was extracted from peripheral blood samples.

Muscle pathological examination
Muscle biopsies were performed from the right bicep or left gastrocnemius of the two cases, respectively. The muscle tissue was frozen and then cut at 8 µm sections. These sections were stained according to standard histological and enzyme histochemical procedures For electron microscopy, muscle specimens were fixed in 2.5% glutaraldehyde in phosphate buffer and post-fixed in 1% osmium tetroxide in the same buffer. Specimens were then dehydrated and embedded in Epon 812. The ultrathin sections of muscle tissue were double stained with uranyl acetate and lead citrate, and then examined with an electron microscope (JEM-1230 JEOL Inc. Tokyo, Japan).

Literature review
We searched the literature in multiple databases including PubMed, EMBASE, Scopus, Web of Science, EBSCO, and Google Scholar database using the keywords "congenital myasthenia syndrome" and "GFPT1 gene." All included cases were required to have muscle biopsy, then the clinical characteristics, laboratory results, treatments, complications and outcomes of all patients were summarized and reanalyzed.

Patient one
The patient was a 37-year-old man who had limb weakness for more than 30 years. At age 5, the patient has noticed poor ability in walking and running compared to his peers. At age 15, he had difficulty in standing up after squatting, and frequently falling down. The symptom of muscle weakness was better in the morning, but worse in the evening. He was diagnosed with myasthenia gravis. Corticosteroid was initially administered, but no efficacy was observed. Afterward, he showed some responses to pyridostigmine, while the muscle weakness gradually progressed to walking difficulty and bath inability. His family history was unremarkable.
Physical examination on admission revealed symmetrical limb weakness without facial, bulbar, neck, and respiratory muscle involvement.
Muscle strength (Medical Research Council, MRC) was 4 grade in the proximal upper limbs, 5-grade in the distal upper limbs, 3 grade in the proximal lower limbs, and 4 grade in the distal lower limbs. Deep tendon reflexes were decreased. Pathological reflexes were negative. No muscle atrophy or fasciculation could be observed. There was no evidence of sensory disturbance, ataxia, or autonomic dysfunction.
The blood count, blood biochemistry, thyroid function, parathyroid hormone, blood acylcarnitines and urine organic acid profiles, paraneoplastic antibody spectrum, and antibodies of myasthenia gravis were all negative. Muscle MRI of thigh revealed a slightly diffused hyperintensity on T1WI except adductor magnus and semimembranosus muscles; and muscle MRI of leg also showed a slightly diffused hyperintensity except medial gastrocnemius ( Figure 1a). Nerve conduction velocity (NCV) and electromyography (EMG) were not obvious abnormal. Nevertheless, repetitive nerve stimulation (RNS) at 3 Hz F I G U R E 1 Muscle MRI changes of lower limb in patient one (a) and patient two (b) with GFPT1-related CMS. Thigh level showed a slightly diffused hyperintensity on T1WI except adductor magnus and semimembranosus muscles; leg level showed a slightly diffused hyperintensity except medial gastrocnemius that simultaneously had mild high-signal on STIR revealed positive decrements of compound muscular action potential (CMAP) in the deltoid.

Patient two
The patient was a 21-year-old man who had limb weakness for 14 years. At age 7, he showed a poor performance in physical education class, and had a little difficulty in running and stairs climbing. Since then, he had complained of muscle fatigue and fluctuating weakness. He was diagnosed with lipid storage disease at age 12, and was given riboflavin and coenzyme Q10, but no benefits were observed. On this admission, he showed permanent weakness characterized by difficulties in climbing stairs, standing up after squatting, and combing hair.
Physical examination showed waddling gait and symmetrical proximal limb weakness without facial, bulbar, neck, and respiratory mus- EMG showed rapid recruitment of motor units suggestive of a myopathic pattern. In addition, RNS at 3 Hz revealed positive decrements in the deltoid and abductor digiti minimi muscle.

Genetic findings
Genetic sequencing disclosed compound heterozygous mutations in  Table S1). A homology search in different species demonstrated that the amino acids at residues 111 and 512 were evolutionally highly conserved, respectively ( Figure 2c). The variants were predicted to be damaging by several in silico tools. The significance of variants was evaluated as pathogenic according to the American College Medical Genetics and Genomics (ACMG) criteria (Li et al., 2017). No causative mutations associated with other CMS or myopathies were found in the genetic screening.

Muscle pathological changes
The myopathological changes of patient one showed an appearance of multiple small vacuoles ( Figure 3a) and a few rimmed vacuoles ( Figure 3b) in some fibers, accompanied with variation of fiber size, central nuclei, fiber splitting and mild interstitial proliferation.
Some fibers with small vacuoles had dark aggregations on MGT stain

Response to therapy
The patient one has been taking pyridostigmine (180 mg/day) since the age of 15. The medicine worked well at first 15 years while the response became less pronounced gradually. After a definite diagnosis, he was prescribed salbutamol (6 mg/day) and fluoxetine (20 mg/day), but his symptoms showed no significant alleviation. After joint prescription to patient two of pyridostigmine (180 mg/day) and albuterol (6 mg/day), his symptoms of muscular weakness improved considerably.

Muscle pathological review
We summarized all reported cases of GFPT1-related CMS in the past 10 years from 2011 to the present. A total of 77 patients with clinical details were reviewed (Table S2) crisis in a few patients at birth, most of them started with muscle weakness, fatigue or frequent falls due to the involvement of proximal limbs.
Eighteen of 51 patients with muscle biopsy were also examined by electron microscopy. Nine patients showed tubular aggregates on muscle ultrastructure. Extensive autophagic vacuoles were found in two patients. Among the 18 patients, endplate analysis was performed in 12 patients, of which 11 patients revealed significantly reduced and poorly developed junctional fold membrane compared to the normal neuromuscular junction.

DISCUSSION
In this study, we described psychomotor delay were also found in our patients, while no more extra-muscular symptoms were observed. Accordingly, clinical physicians should carefully make differential diagnosis between the GFPT1related CMS and myasthenia gravis, metabolic myopathies, or limbgirdle muscle dystrophy (Witherick & Brady, 2018).
There is a lack of an inherent association between the severity of muscle weakness and the abnormal extent of muscle MRI. The finding of a relatively normal muscle MRI in a patient who showed marked weakness possibly suggested a disorder of neuromuscular junction (Finlayson et al., 2016). This study showed that muscle MRI of GFPT1-related CMS had a tendency of selective distribution of mild fat infiltration characterized by diffusely involving in thigh muscles but sparing of adductor magnus and semimembranosus muscles, as well as diffusely involving in the leg muscles but sparing of medial gastrocnemius. Additionally, the mild hyperintensity in muscles without fat infil- Although about 70% of GFPT1-related CMS patients showed tubular aggregates that were believed to represent aggregations of misfolded proteins (Schiaffino, 2012), our studies indicated that the pathological changes simultaneously had great diversities. The impairment of neuromuscular junction is a main target due to heavy glycosylation of many important proteins in the neuromuscular junction, while it is possible that GFPT1 defect could have additional direct pathological effects on extra-synaptic regions (Hugo & Schlegel, 2017;Niimi et al., 2001). Muscle specimens of patients with hypoglycosylated myasthenia have shown prominent myopathic features including fiber-type disproportion, degenerating mitochondria, and destruction of the muscle fiber organelles associated with autophagy (Bauché et al., 2017;Guergueltcheva et al., 2012;Helman et al., 2019;Huh et al., 2012;Luo et al., 2019;Ma et al., 2021;Maselli et al., 2014;Matsumoto et al., 2019;O'grady et al., 2016;Selcen et al., 2013;Senderek et al., 2011;Szelinger et al., 2020;Yiş et al., 2017;Zhao et al., 2021). Zebrafish model with GFPT1 knock down also showed abnormalities of both muscle structure and neuromuscular junction (Hugo & Schlegel, 2017). Therefore, it is reasonable that some vacuolar or nonspecific myopathic changes could appear in CMS specimens attributed to GFPT1-related hypoglycosylation of multiple muscle proteins.
It was puzzling that GFPT1 defect in patient one was associated with atypical pathological changes of myofibrillar myopathy (MFM) characterized by desmin deposits, Z-disc disorganization, and electronic dense granulofilamentous aggregation. More than 200 known glycosyltransferases are responsible for the glycosylation of thousands of proteins in muscle (Zoltowska et al., 2013), of which many MFMrelated proteins, such as desmin, plectin, myotilin, LDB3, and FLNC, should be glycosylated to accomplish physiological functions (Hong et al., 2011). Among these MFM-related proteins, the plectin crosslinks intermediate filaments to their targets in different tissues, and has been associated with MFM, CMS, and limb-girdle muscle dystrophy (Winter et al., 2014). In this sense, the underlying hypoglycosylation of plectin that will cause the dysfunction of the protein might be partly in charge of the MFM-like pathological changes.
The dysfunction of neuromuscular junction is the essence of CMS.
The endplates morphology showed that the folds of postsynaptic membrane usually were reduced and simplified, but unspecific abnormalities and even normal endplates could also be observed in GFPT1related CMS (Zoltowska et al., 2013). Intriguingly, besides the poorly developed endplates, some ring-like or block-like materials with electronic dense were observed beneath endplates in our patient. These materials might originate from the disturbance of Z lines or myofibrillar structures. The pathological basis of endplate changes likely stems from hypoglycosylation and altered function of endplate-specific glycoproteins, such as MUSK, agrin, and dystroglycans (Willems et al., 2016).
In summary, besides the common tubular aggregates, the muscle pathological changes of GFPT1-related CMS also can show rimmed vacuolar myopathy, autophagic vacuolar myopathy, mitochondrialike myopathy, MFM-like myopathy, neurogenic features, and unspecific myopathy changes. This extra-synaptic pathology might be in part responsible for the permanent muscle weakness and resistance to acetylcholinesterase inhibitor therapy. To some extent, the pathological findings might be one of the predictors of the disease outcome.

ACKNOWLEDGMENTS
We thank the patients and their families for cooperation. We thank Ms. Yaqing Yu for the work in preparations for pathological sections.

CONFLICT OF INTEREST
The authors declare that they have no competing interests.

AUTHOR CONTRIBUTIONS
Kaiyan Jiang and Yilei Zheng contributed to analysis, interpretation and drafting. Jing Lin contributed to genetic analysis. Yanyan Yu, Xiaobing Li, Xiaorong Wu, and Xin Fang contributed to the acquisition and analysis of data. Meihong Zhou performed the pathological study, Meihong Zhou performed the electrophysiological analysis. Daojun Hong contributed to the study design and revising the manuscript, as well as funding acquisition.

DATA AVAILABILITY STATEMENT
All relevant data are within the paper and its Supporting Information files.

SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publisher's website.