Reply: A novel MCM3AP mutation in a Lebanese family with recessive Charcot-Marie-Tooth neuropathy

Le résumé fourni par la source

Sir, We have read with great interest the letter of Kennerson et al. (2018) as well as the letter of Karakaya et al. (2017), which confirm our finding of recessive mutations in minichromosome maintenance complex component 3 associated protein (MCM3AP) underlying Charcot-Marie-Tooth neuropathy (CMT4). We initially reported patients in five families (Ylikallio et al., 2017). Kennerson et al. (2018) describe three siblings with homozygous MCM3AP variant p.Leu870Ser, which is located in the Sac3 (suppressor of actin) domain of the protein encoded by MCM3AP, germinal centre associated nuclear protein (GANP). Karakaya et al. (2017) studied patients in three independent families. Earlier Schuurs-Hoeijmakers et al. (2013) had also described a putative pathogenic MCM3AP variant in one family. Thus the total number of reported patients with neuropathy linked to MCM3AP is now 18, from 10 distinct families. Based on the so far reported studies, we can conclude that MCM3AP mutations cause early-onset CMT4. At worst the disease is progressive and leads to loss of ambulation in childhood. At best, progression is slow and ambulation is preserved in early adulthood. The MCM3AP-related neuropathy is axonal and predominantly motor. Sensory findings are less pronounced; however, at least a mild degree of sensory impairment, i.e. elevated pain or sensory thresholds or reduced sensory nerve action potentials, have been reported in patients where sensory testing data is available. The most consistent additional finding is intellectual disability, which was nevertheless unreported in the patients of Kennerson et al. (2018). Thus, intellectual disability or suspicion thereof is found in 10/18 patients (Schuurs-Hoeijmakers et al., 2013; Karakaya et al., 2017; Ylikallio et al., 2017). Of the other potentially associated symptoms, obesity merits mention as the patients described by Kennerson et al. (2018) had developed obesity in adulthood, and similarly the two Canadian patients reported in our study who had reached adulthood were also obese. The latter two also had ovarian failure (Ylikallio et al., 2017), which was not found in the adult patients of Kennerson et al. (2018). The reported MCM3AP mutations fit broadly into two categories: (i) homozygous point mutations in the Sac3 domain (five mutations, 10 patients); and (ii) compound heterozygous nonsense and missense variants or a small duplication outside of the Sac3 domain (four mutations, seven patients). Additionally the Australian patient reported in our study (Ylikallio et al., 2017) has a compound heterozygous missense mutation p.Ala867Asp in the Sac3 domain together with a nonsense mutation p.Tyr889* predicted to cause mRNA instability, making the patient essentially hemizygous for the Sac3 domain mutation (Ylikallio et al., 2017). As originally pointed out by Karakaya et al. (2017) there is a trend towards early onset and severe disease course for the compound heterozygous mutations outside of the Sac3 domain, and a more benign disease course for the homozygous Sac3 mutations. The findings of Kennerson et al. (2018) appear to support this hypothesis, as their homozygous patients had disease onset at ages 6–10 and have maintained ability to manage daily activities with help from their family. In contrast, compound heterozygous patients with non-Sac3 domain mutations have had disease-onset between birth and age 4 (Karakaya et al., 2017; Ylikallio et al., 2017). Among the most severely affected were the Finnish patients who were compound heterozygous for p.Pro148Leufs* and p.Val1272Met, which we showed to lead to GANP depletion in skin fibroblasts. The oldest of the Finnish patients lost independent ambulation at age 10–11, and was placed on ventilator at age 14 (Ylikallio et al., 2017). The Dutch patient who carried nonsense and splice donor mutations (p.Asn1317fs*18 and p.Gln619=), which are likely to affect GANP protein amount (Ylikallio et al., 2017), was unable to sit or walk at age 3.5. However, one of the Turkish patients with the homozygous Sac3 mutation p.Met762Thr also showed a relatively aggressive disease course and lost ambulation at age 10 (Ylikallio et al., 2017). Thus, we agree with Karakaya et al. (2017) and Kennerson et al. (2018) that no definite genotype–phenotype correlations can be established. Investigation of the consequences of the different mutation types on GANP function could clarify what determines the severity of disease in each case. Based on the crystal structure of the yeast homologue of GANP Kennerson et al. (2018) show that Sac3 domain mutations could inhibit the interaction of GANP with PCID2, another subunit of the transcription export complex 2 (TREX-2) for which GANP serves as a scaffold (Williams et al., 2018). This finding supports the hypothesis that axonal degeneration caused by Sac3 domain mutations results from impaired nucleocytoplasmic mRNA export. As we have shown for some of the non-Sac3 domain mutations, defective mRNA export could also arise from depletion of GANP, preventing its scaffold role for other nuclear export proteins (Ylikallio et al., 2017). As Kennerson et al. (2018) point out, further functional studies are needed to investigate whether defective mRNA transport is the disease mechanism in MCM3AP patients. In particular, it remains to be explained why the phenotype appears neuron-specific although mRNA export is universal to all cell types. Hypothetical explanations are that defective GANP impairs the export of one or more mRNAs that are locally translated in axons (Zappulo et al., 2017) or that the proposed role of GANP in selective mRNA export is particularly important for neurons (Wickramasinghe et al., 2014).

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.