Extensive heterogeneity in patients with ALS with mutations in SOD1 in France

Abstract

Muratet and colleagues present an update on mutations in the cytosolic free radical scavenging enzyme SOD1 in 470 French patients with familial ALS. With the development of gene and antibody therapies targeting SOD1 and with over 220 variants found in SOD1 globally, key questions asked are, if all mutations are pathogenic, do they cause neurodegeneration by the same mechanism? which mutations are to be studied in clinical trials? Obviously, it is advantageous in trials to include patients with the same mutation to get a homogenous study population as possible. The paper is therefore timely and interesting for several reasons. From an epidemiological point of view, 38 of the families are heterozygous for 1 of 27 mutations, but only 5 of the mutations were found in more than one family. The p.Gly94Cys is the most frequent SOD1 mutation in France and was found in five families. It is also the most prevalent mutation in neighbouring Belgium, making this a particular suitable mutation to study in drug trials. However, four of the SOD1 mutations reported here are unique to France and found only in single families. This makes it difficult to ascertain if these four are pathogenic for ALS or not. Such private mutations are therefore usually not suitable for inclusion in drug trials. To get a clear readout in an ALS drug trial, studying patients with aggressive forms of ALS is often advantageous. There are very aggressive mutations (e.g., p.Ala5Thr, p.Ala5Val, p.Gly42Ser, p.His44Arg and p.Leu85Phe) with survival times around 1 year in France. These have also been found in other countries and are well documented pathogenic for ALS. Hence, patients in these families should be suitable for participation in drug trials. A surprise finding is that no oligogenetic families were found (the patients have mutations in more than one pathogenic gene, eg, SOD1+C9orf72HRE). Such families are difficult in genetic counselling and ‘most unwanted’ in clinical drug trials. Lastly, there is genetic heterogeneity: while 25 of the mutations are missense mutation switching one amino acid for another, 2 mutations change the length of the SOD1 protein: the c.358–304C>G creates a new pseudoexon deep inside intron 4 and will not be detected by standard DNA diagnostic analysis (which does not sequence the introns). This mutation adds seven novel amino acids after p.valine 119 in exon 4 but loses all of exon 5 truncating the mutant protein at 126 amino acids. The c.358–304C>G has previously convincingly been demonstrated to be pathogenic in a large French Canadian family and should be suitable for inclusion in drug trials. While the c.358–304C>G produces a much shorter mutant protein, the c.358–10T>G mutation does precisely the opposite: the c.358–10T>G was originally found by Peter Sapp and colleagues in 1994 in a single patient in a US pedigree with familial ALS, but pathogenicity was not proven. This mutation squeezes in three amino acids between exons 4 and 5 (residues p.Val119 and p.Val120), expanding the overall peptide to 156 residues. The addition of three residues in a confined part of the molecule will disrupt the tertiary structure and ligand binding of p.His121 to the catalytic copper atom. The mutant peptide will be unable to form a 3D structure resembling the native structure and will be prone for aggregate formation, a hallmark of ALScausing SOD1 mutations. As has been found for other mutations that change the length of the mutant SOD1 protein, patients with the c.358–10T>G develop a typical ALS disease with limb onset, dominant lowermotor neuron signs, no atypical features and a survival time of about 4 years. The disease penetrance is complete but with a surprising high mean age at onset at 60 years contrasting with 47 years for most mutations with complete penetrance. Muratet and colleagues present four families with the c.358–10T>G and convincingly demonstrate that the c.358–10T>G is a cause of ALS and suitable for participation in antiSOD1 trials.