LAUSR

We thank Dr. Finsterer for the comments (Finsterer, 2021) regarding our manuscript (Torraco et al., 2021) and below you will find our responses to his concerns. First of all we agree that the sentence reported in the first lines of the discussion line 3 since we have described seven patients carrying four novel homozygous mutations, as specified in the results section and in table 1. Second, he argued that, despite the virtually undetectable NDUFA12 protein in all samples we analyzed, phenotypic variability is observed. He sustains that this condition might be due to a several factors, such as mtDNA variants and/or mtDNA amount, to different complex I (CI; NADH‐ ubiquinone oxidoreductase) activity in our patients and to different treatments they have received. Indeed phenotype variability is the common baseline of the mitochondrial disease and several reports have been published with high phenotype variability even in the context of the same genetic defect. Clinical symptoms and progression can be different even among siblings who share the same genetic variation, but whether this variability is due to genetic background or environmental contribution is still under debate. Mitochondrial DNA is a potential modifier in mitochondrial disorders caused by a specific nuclear gene defects but to link a mtDNA variant or mtDNA amount to disease phenotype onset, variability and progression would require specific disease models (Hämäläinen et al., 2013; Li et al., 2018) and any attempt to make a correlation between mtDNA‐nDNA genotype interaction in the context of our group of patients would have been just speculative. Vice versa, no qualitative (single/multiple deletions) or quantitative (depletion) alterations of mtDNA have ever been associated with nuclear genes encoding for protein been part of complex I. In addition, a limited availability of tissues in our cohort of patients prevented us to carry extensive functional analysis that could at least in part shed light on the different phenotypes observed. We do agree that CSF lactate levels in some cases might correlate with a more severe disease course, even though reports in the literature are controversial (Bakare et al., 2021). On the other hand, we would like to highlight that some patients in our cohort, who seem to exhibit a milder disease course (patients 4 and 5) present high CSF lactate values. It is well known that antioxidant therapy unfortunately does not modify the natural history of Leigh syndrome. Moreover, patients 4 and 5 in our cohort are described having a milder course in absence of any antioxidant therapy. Regarding the reason why the patient‐1 showed bilateral T2‐hyperintense brainstem lesions in the first MRI (5 years) that disappeared on follow‐up imaging at age 7 years, we do not have a clear explanation but can hypothesize that this is due to the resolution of cytotoxic edema caused by a metabolic decompensation. In conclusion, disease variability is a hallmark of the mitochondrial disease and the purpose of our study was to make mitochondrial disorders more recognizable through a good clinical description and to link this phenotype to a gene, which is rarely found in association with mitochondrial defect.