Further phenotypic heterogeneity of CoQ10 deficiency associated with steroid resistant nephrotic syndrome and novel COQ2 and COQ6 variants

To the Editor: Genes involved in the coenzyme Q10 biosynthetic pathway (COQ2, COQ6, PDSS1, PDSS2 and ADCK4) are mutated in about 1% of steroid resistant nephrotic syndrome (SRNS) cases and are often associated with neurological symptoms (1, 2). Here, we describe novel phenotypes associated with pathogenic mutations in two genes of the CoQ10 pathway, COQ2 and COQ6, in three patients with SRNS. Written informed consent was obtained from each patient and/or their parents. The study was performed according to the guidelines of the Declaration of Helsinki and the local Ethics Committees of University Hospitals of Foggia and Bari (Italy). One novel homozygous COQ2 variant, p.Gly390Ala (p.Gly340Ala, according to KU877220 GenBank sequence (2)) was identified by Next Generation Sequencing (NGS) (Appendix S1, Supporting information) in two cousins (patients P1 and P2; Fig. 1a; Tables S1 and S2) with SRNS associated with focal segmental glomerulosclerosis (FSGS) lesions and podocyte foot process effacement on renal biopsy (Fig. 1b, panels 1–4; Table S1). The pathogenicity of this variant is supported by its absence in public SNP (Single Nucleotide Polymorphism) database; the segregation with the disease (Fig. S1, panel 1); in silico predictions (Fig. 1c, Table S2); the reduced rate of respiratory growth and CoQ levels of yeast expressing this allele (Fig. 1d); the presence of numerous dysmorphic mitochondria on renal biopsy (Fig. 1b, panels 5 and 6). Remarkably, both patients harboring COQ2 change developed SRNS in adolescence with rapid progression to end-stage renal disease (ESRD) and had only mild neurological symptoms (Table S1). Of note, both cousins received a successful kidney transplant without recurrence of proteinuria and started CoQ10 treatment immediately after the genetic diagnosis. They are currently asymptomatic without any neurological symptoms after a 2-year follow up. Other authors reported patients with inherited COQ2 changes presented with isolated renal symptoms (2), however, to date, the reported age of onset of SRNS in patients carrying COQ2 variants was before the age of 2.5 years (2). To our knowledge, this is the first report describing COQ2 variants in patients with adolescent-onset SRNS and with a clinical spectrum resembling another CoQ10-glomerulopathy caused by mutations in ADCK4 (3). This finding recapitulates what has been observed in the Pdss2 (kd/kd) mice (2) and can be explained by the relatively mild effect of the p.Gly390Ala (p.Gly340Ala) allele, as documented by yeast studies (Fig. 1d). Mutational screening of patient P3 (Fig. 1e; Table S1) revealed a novel homozygous missense change in COQ6 gene: p.Pro261Leu (Fig. 1f; Table S2). The functional effect of this variant is supported by its low MAF in the European population (1:16,683, Table S2); the segregation with the disease (Fig. S1, panel 2); in silico predictions (Fig. 1g; Table S2) and the inability of COQ6 p.Pro261Leu allele to rescue the growth defect of the deleted yeast (Fig. 1h). Few patients with COQ6 changes (12) have been reported so far, they showed SRNS with onset at a median age of 1.2 years and sensorineural deafness (7/12) but some had also encephalopathic features (4). Our patient developed SRNS at 8 months with progressive deterioration of renal function and ESRD at 20 months. Of note, he did not present deafness or encephalopathic features. He is now on peritoneal dialysis treatment, however, CoQ10 treatment was started to prevent neurological symptoms. Interestingly, there was no history of schwannomatosis in the patient’s family, and unless further evidence supporting the link between heterozygous COQ6 mutations and schwannomatosis becomes available, we do not recommend mutational screening of COQ6 gene for schwannomas carriers (5). The most frequent glomerular lesion associated with COQ6 changes is FSGS (4). Remarkably, the analysis of renal biopsy of our patient revealed membranoproliferative glomerulonephritis and C3 deposits (Fig. 1f) usually associated with variants and risk haplotypes of complement-pathway genes, whose analysis did not reveal any variants with MAF (Minor Allele Frequency) <0.01 (data not shown). In summary, this study shows that we should analyze COQ2 and COQ6 genes also in patients with adolescent-onset of SRNS