Family 2
Subject II2 is the only known affected member of a Portuguese non-consanguineous family (Figure 1). She had no history of physical development delay. She reported cramps in the lower limbs and a tremor of the upper extremities in early adulthood. At 30 years, she displayed muscular weakness with walking impairment and difficulties climbing stairs. Her father also experienced cramps but there was no familial history otherwise and no consanguinity. The disease slowly progressed, and at age 44, her clinical examination revealed distal weakness in the upper limbs (wrist extensor and interosseous scored as 4/5 on MRC scale) and a proximal deficit in the lower limbs (quadriceps, psoas, and hamstring muscles as 4/5 on MRC scale), with pes cavus . Jerk reflexes were all present except at the ankle. No pyramidal signs or cranial nerve impairments were observed. Electroneuromyography showed neurogenic features, in the forearms, and both proximal and distal in the lower limbs, with normal motor and sensory conduction (Table 1). CK level was mildly elevated: 258 UI/L (Normal < 200 UI/L). A whole-body muscle MRI was performed at 46 years of age, showing a severe, bilateral, and symmetric fibro-fatty substitution more apparent in muscles of the calf. In the thighs and pelvis, there was a mild fatty replacement of posterior thigh musculature and glutei (Fig1 B). In the upper limbs, there was also a mild wasting of biceps brachialis, and deltoids.

Molecular findings and bioinformatics analysis

Whole-exome sequencing was performed for both families. The two affected individuals of the first family bear the same heterozygous mutation c.1330G>A (p.Gly444Arg) in the MORC2 gene (NM_001303256.2). Patient II2 from family 2 bears the heterozygous mutation c.1338C>A (p.His446Gln) in the MORC2 gene. Sequencing of the MORC2 gene in her mother and sister’s DNA (both asymptomatic) revealed no mutation. Her father was deceased hence no DNA was available.
MORC2 protein (NP_001290185) contains several predicted domains (Figure 2A). Interestingly, all known mutations affect the ATPase domain or its transducer S5 domain (Douse et al., 2018). Both mutations described here affect conserved amino acids in the S5 domain at positions p444 and p446 (Figure 2B). Neighboring mutations in the S5 domain are p.C407Y (Ando et al., 2017), p.T424R (Schottmann et al., 2016; Zanni et al., 2017), p.A431V (Ando et al., 2017), p.D466N (Semplicini et al., 2017). Interestingly, mutation p.G444R has been reported by Albulym, et al (Albulym et al., 2016) as a likely pathogenic variant without familial or functional evidence.
Bioinformatic prediction by HumVAr (PolyPhen-2) identified mutations p.G444R and p.H446Q as probably pathogenic with a score of 0.945 and 0.878 on a 0 to 1 scale, with a sensitivity of 0.65 or 0.71 and a specificity of 0.91 or 0.89, respectively. Likewise, the mutation taster software found that both mutations were pathogenic and were neither found in gnomAD (The Genome Aggregation Database) nor 1000G. Interestingly, the crystal structure of MORC2 was solved (Douse et al., 2018) allowing to map the mutations (Figure 2C) and predicting their impact on the 3D structure of the protein. This allowed the HOPE software to predict that the mutations were structurally damaging. Indeed, the arginine mutant residue in position 444 is bigger and introduces a charge in a buried residue which probably perturbs protein folding. The glutamine mutant residue at position 446 breaks a hydrogen bond between Histidine 444 and Alanine 442 residue and therefore destabilizes the alpha helix.

Protein construct and subcellular localization

A flag tag (also called DYK) (Figure sup1A) was fused to the C-terminal extremity of MORC2 and an internal ribosome entry site (IRES) and eGFP were added downstream (pCAGIG plasmid, addgene #11159). To check the subcellular localization of the MORC2-flag fusion protein, SH-EP neuroblastoma cells were transfected (Reddy et al., 1991) and murine primary cortical neurons were electroporated. Two days after transfection, the cells were fixed and MORC2 was stained with the anti-Flag antibody. Nuclei were counterstained with DAPI (Figure sup1B). In both cell types, MORC2 was detected in the nucleus.
Site-directed mutagenesis was done on pCAGIG MORC2-Flag to reproduce the mutations identified in the patients. As positive pathogenic controls mutations were introduced at the following positions of the MORC2 coding sequence (NM_001303256): c.754 C>T leading to MORC2 p.R252W related to CMT2 (Sevilla et al., 2016), c.1271 C>G leading to MORC2 p.T424R related to SMA-like (Schottmann et al., 2016; Zanni et al., 2017). Similarly, both variants of unknown significance were generated by mutagenesis at positions c.1330G>A leading to MORC2 p.G444R and c.1338C>A leading to p.H446Q. By convention, these variants will be designated in the text by their position (ie: p.G444R as p444). To evaluate the expression of the protein, the same amount of each plasmid was transfected in SH-EP and total proteins were solubilized in Laemmli buffer 24 hours later and analyzed by western Western Blot (Figure sup1B). Ponceau Red staining revealed the total protein loaded whereas anti-MORC2 or anti-Flag antibodies were used to detect MORC2. Altogerther, WT and mutant MORC2 proteins were expressed at the same level. Furthermore, immunofluorescence and confocal microscopy detected no difference in the subcellular localization of WT or MORC2 mutants (data not shown).

MORC2 mutants alter SH-EP survival and trigger apoptosis

To determine the impact of MORC2 variants on SH-EP survival over time, transfected cells (GFP positive) were analysed after 1, and 3 days in culture (Figure 3A). The cell number was attributed a value of 100 % at day 1. To evaluate survival, eGFP positive cells were counted at days 3 and normalized to the number of cells at day 1. At day 3, all mutants caused a decrease in survival compared to control, including the variants p444 (84%) and p446 (82%) of unknown significance (Figure 3B).
Activation of the caspase 3 by proteolytic cleavage is a landmark of apoptosis-mediated cell death. To evaluate if reduced survival induced by MORC2 mutants involved the induction of apoptosis, activated caspase 3 was detected by immunofluorescence at days 3 of culture after transfection. Quantification of activated caspase 3 positive cells at day 3 revealed a significant increase in all MORC2 mutants: p.252 (110%), p424 (127%) or MORC2 p444 (124%), p446 (130%) compared to WT MORC2 (100%). Altogether, these results suggest that MORC2variants decrease cell survival through apoptosis.

Mutant MORC2 affect cortical neuron survival

Since mutations in MORC2 are associated with neurodegenerative disorders, we tested their effect in primary cells more relevant for to the pathophysiology of MORC2-related disorders. Cortical neurons of the motor cortex, also called upper motor neurons, are easy to purify in large quantities and can be transfected by electroporation. To determine the impact of MORC2 mutants on the survival, cortical neurons were electroporated with WT, p424, p444 or p446 MORC2 constructs immediately after dissociation and analysed at days 1, 2 and 5 (Figure 4A). Cortical neuron counting overtime showed that a decrease in survival was already observed at day 2 with the pathogenic mutation (p424, 87%) and with the p444 (84%) and p446 (77%) variants compared to WT MORC2. Similar results were observed at days 5 after electroporation with the pathogenic mutant p424 (60.34%) and the two variants of unknown significance p444 (67.19%) and p446 (47.97%) compared to the WT (100%), suggesting that these variant are pathogenic.
Since we observed increased apoptosis in SH-EP cell expressing the MORC2 mutants, we investigated whether it was also the case in motor neurons. Activated-caspase 3 staining was performed 2 days after electroporation (Figure 4C). The percentage of activated-caspase3 positive neurons in electroporated neurons (i.e. GFP positive) indicated that after 2 days, apoptosis was increased in MORC2 p424 (199 %), p444 (239 %), p446 (235 %) compared to WT MORC2 (100%) electroporated neurons.

MORC2 affect cortical neuron neurite outgrowth

Axonal forms of CMT and SMA affect axonal growth or maintenance in patients. To investigate the impact of MORC2 mutations on the axonal compartment, we evaluated neurite outgrowth in neurons electroporated with the eGFP reporter gene (Figure 4C). Box plot of the results showed a significant reduction of neurite outgrowth in neurons expressing p424 (mean: 87.96), p444 (53.37) and p446 (66.43) compared to neurons expressing WT MORC2 (107.85). Altogether, these results show that MORC2 mutants interfere with neurite outgrowth, suggesting that they cause axonal stress. This supports the conclusion drawn from survival and apoptosis experiments performed in SH-EP and motor neurons that indicated that the p444 and p446 variants of unknown significance are probably pathogenic.