GFPT1 deficiency in muscle leads to myasthenia and myopathy in mice

Abstract

Glutamine-fructose-6-phosphate transaminase 1 (GFPT1) is the rate-limiting enzyme in the hexosamine biosynthetic pathway which yields precursors required for protein and lipid glycosylation. Mutations in GFPT1 and other genes downstream of this pathway cause congenital myasthenic syndrome (CMS) characterized by fatigable muscle weakness owing to impaired neurotransmission. The precise pathomechanisms at the neuromuscular junction (NMJ) owing to a deficiency in GFPT1 is yet to be discovered. One of the challenges we face is the viability of Gfpt1-/- knockout mice. In this study, we use Cre/LoxP technology to generate a muscle-specific GFPT1 knockout mouse model, Gfpt1tm1d/tm1d, characteristic of the human CMS phenotype. Our data suggest a critical role for muscle derived GFPT1 in the development of the NMJ, neurotransmission, skeletal muscle integrity and highlight that a deficiency in skeletal muscle alone is sufficient to cause morphological postsynaptic NMJ changes that are accompanied by presynaptic alterations despite the conservation of neuronal GFPT1 expression. In addition to the conventional morphological NMJ changes and fatigable muscle weakness, Gfpt1tm1d/tm1d mice display a progressive myopathic phenotype with the presence of tubular aggregates in muscle, characteristic of the GFPT1-CMS phenotype. We further identify an upregulation of skeletal muscle proteins glypican-1, farnesyltransferase/geranylgeranyltransferase type-1 subunit α and muscle-specific kinase, which are known to be involved in the differentiation and maintenance of the NMJ. The Gfpt1tm1d/tm1d model allows for further investigation of pathophysiological consequences on genes and pathways downstream of GFPT1 likely to involve misglycosylation or hypoglycosylation of NMJs and muscle targets.

Figure 1.

Generation of muscle-specific Gfpt1 knockout mice. (A) Schematic demonstrating the genomic region of the Gfpt1 gene where Exon 7 is floxed by loxP sites. Conversion of the Gfpt1^tm1c allele to the Gfpt1^tm1d allele in skeletal and cardiac muscle of Gfpt1 mutant mice carrying a Cre transgene under the control of the Ckm promoter. (B) PCR demonstrating recombination events that take place in muscle and non-muscle tissues of Gfpt1^tm1d/tm1d mice. (C) Protein lysates from brain, kidney, skeletal muscles and the heart of control and mutant mice were analysed by western blotting against a GFPT1 antibody. An antibody against GAPDH was used as a loading control (n = 6).

Figure 2.

A comparison of body weight and muscle strength between control and Gfpt1^tm1d/tm1d mice. (A) Growth curves demonstrating changes in body weight over a 6 months period (n = 8). (B) Quantitative analysis of latency to fall from a wire grid at various time points up to the age of 6 months (n = 8). Gfpt1^tm1d/tm1dmice perform worse than control mice at all time points (P < 0.01). (C) Quantitative analysis of force generated by the TA muscle after every 10 stimulations of the CPN in 3 months old control and Gfpt1^tm1d/tm1d mice (n = 5). Data are expressed as a percentage of baseline force. (D) Quantification of fatigability of the TA muscle after 80, 90 and 100 stimulations of the CPN in 3 months old control and Gfpt1^tm1d/tm1dmice (n = 5). Data are expressed as the percentage decrease of baseline force. (E) Isometric tetanic maximal force in diaphragm muscle from control and Gfpt1^tm1d/tm1d mice (n = 4). (F) Quantitative analysis of force generated by the diaphragm muscle after every 10 stimulations in 3 months old control and Gfpt1^tm1d/tm1d mice (n = 4). Data are expressed as a percentage of baseline force. (G) Quantification of the fatigability of the diaphragm muscle between 50 and 100 stimulations in 3 months old control and Gfpt1^tm1d/tm1dmice. Data are expressed as a percentage reduction in force. Data are mean±SEM. NC, no change; *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 3.

Characteristics and myopathic changes in muscle from Gfpt1^tm1d/tm1d mice. (A) Representative TA, intercostal, soleus, EDL and diaphragm muscles stained with haematoxylin and eosin from control and Gfpt1^tm1d/tm1d mice. Rounded myofibres (asterix), centrally located nuclei in myofibres (black arrows), tubular aggregates (black arrow head), necrotic fibres (white arrow head), adipose tissue (white arrows). Hypertrophic and atrophic myofibres are also present. (B) Quantitative analyses demonstrating the distribution of myofibre size according to CSA. Data are median, 25th percentile, 75th percentile, minimum and maximum values (including outliers). **P < 0.01; ***P < 0.001; ns, not significant (Mann–Whitney U test) (n = 4).

Figure 4.

Fibre type labelling in control and Gfpt1^tmld/tmld mouse muscle. Soleus (A) and TA (C) muscles were labelled with antibodies against MHC: Type 1 (blue) I, Type IIa (red), Type IIb (green) and laminin (green). Type IIx fibres were labelled on a second section (not shown). There were no differences in the fibre type proportion in both the soleus (n = 4) (B) and TA muscles (n = 4) (D).

Figure 5.

Aberrant NMJs in 3 months old Gfpt1^tm1d/tm1d mice. (A) Whole-mount TA, intercostal, soleus and EDL muscles were labelled with anti-neurofilament, anti-synaptophysin and Alexa fluor 594 α-bungarotoxin. (B) Whole-mount TA muscles labelled with anti-synaptophysin and Alexa fluor 594 α-bungarotoxin demonstrating the degree of co-localization of pre- and postsynaptic components. Quantitative analysis demonstrating AChR cluster area (C), and mean number of fragments/AChR cluster (D) in control and Gfpt1^tm1d/tm1d mice (n = 6). AChR fragments (blue arrows), discontinuous and disorganized axons (white arrow), presynaptic and postsynaptic overlap (magenta arrows). Data are mean±SEM. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.

Figure 6.

Altered morphology at the ultrastructural level in Gfpt1^tm1d/tm1d mouse muscle. Representative electron micrographs from 3 months old control and Gfpt1^tm1d/tm1d intercostal muscles. Examples of control (A), and Gfpt1^tm1d/tm1d(B–D) NMJs. Tubular aggregates (E) and subsarcolemmal vesicular structures (F–H) in Gfpt1^tm1d/tm1d mouse muscle. Synaptic vesicles (*), junctional folds (red arrow), mitochondria (M), tubular aggregates (TA), subsarcolemmal vesicular structures (white arrow), dense filamentous-like material (blue arrowhead) (n = 4).

Figure 7.

Caveolin-3 expression in muscles from control (n = 3) and Gfpt1^tm1d/tm1d (n = 4) mice. Transverse sections of the soleus (A) and TA muscles were labelled with an antibody against caveolin-3. No significant differences in the percentage of myofibres expressing caveolin-3 were observed in the soleus (B) and TA (C) muscles from Gfpt1^tm1d/tm1d mice compared with controls. Caveolin-3 positive punctate were significantly larger in Gfpt1^tm1d/tm1d soleus (D) and TA (E) muscles compared with controls. Data are mean±SEM. **P < 0.01, ****P < 0.0001.

Figure 8.

Proteins regulated as a consequence of GFPT1 depletion in 3 months old intercostal muscles. (A) STRING analyses demonstrating the spectrum of affected proteins. Proteins that show upregulation (green) and downregulation (red) are shown. (B) Immunoblot studies show upregulation of glypican-1 and MuSK proteins in Gfpt1^tm1d/tm1d mice. Quantitative analysis showing the relative expression levels of glypican-1 (C) (n = 3), and MuSK (D) (n = 2) normalized to their corresponding α-actinin loading controls.