A PRDX1 mutant allele causes a MMACHC secondary epimutation in cblC patients

To date, epimutations reported in man have been somatic and erased in germlines. Here, we identify a cause of the autosomal recessive cblC class of inborn errors of vitamin B12 metabolism that we name “epi-cblC”. The subjects are compound heterozygotes for a genetic mutation and for a promoter epimutation, detected in blood, fibroblasts, and sperm, at the MMACHC locus; 5-azacytidine restores the expression of MMACHC in fibroblasts. MMACHC is flanked by CCDC163P and PRDX1, which are in the opposite orientation. The epimutation is present in three generations and results from PRDX1 mutations that force antisense transcription of MMACHC thereby possibly generating a H3K36me3 mark. The silencing of PRDX1 transcription leads to partial hypomethylation of the epiallele and restores the expression of MMACHC. This example of epi-cblC demonstrates the need to search for compound epigenetic-genetic heterozygosity in patients with typical disease manifestation and genetic heterozygosity in disease-causing genes located in other gene trios.

associated CpG island. There is also an increase in H3K36me3 around this site in individuals with this mutation. H3K36me3 is a mark that is associated with active transcription of PolII and is involved in splicing, supporting the other evidence of antisense read-through of the promoter of the neighbouring MMACHC (i.e. aberrant transcripts). They provide some functional evidence for this hypothesis by showing that knock down of the aberrant PRDX1 restores transcription of MMACHC. This mutation was discovered as when inherited as a compound heterozygote with loss of function mutations within the MMACHC coding region, it is sufficient to cause cblC metabolic disorder due to loss of MMACHC expression. They argue that this is an example of a constitutive epimutation that is transgenerationally inherited as the hypermethylation phenotype is present in sperm in cases with paternal transmission.
Novelty: These findings are novel in that this is a rare and novel mutation that is associated with an inherited disorder and has diagnostic value for cases of suspected cblC in which homozygous MMACHC mutations are absent. With regards to the claim of a transgenerationally inherited epimutation; I would apply caution. This is clearly a genetic mutation that links to phenotype through altering the epigenetic state, as they show. The presence of promoter hypermethylation in sperm does not demonstrate transgenerational epigenetic inheritance, (i.e. failure to be erased either in primordial germ cells and/or the post-fertilisation embryo). The evidence they present suggests that hypermethylation will be established in any tissue in which PRDX1 is expressed (including the male germ-line during spermatogenesis) as an obligatory consequence of the genetic mutation. They provide no direct evidence to suggest that there is failure to erase this during the developmental periods of genome-wide DNA methylation erasure. Furthermore, this is not the first example of read-through antisense transcription altering the epigenetic state of a promoter. Therefore, with regards to mechanistic insight, this is not novel (e.g. Tufarelli, C. et al. Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nature Genet. 34, 157-165 (2003).) I think the conclusions and work presented in the manuscript are reasonable. I have some minor comments with regards to specific analyses and presentation, provided below.
I believe the greatest relevance and interest in this work will be for the diagnostic benefit of this specific disorder, rather than an insight into novel biological mechanisms.
Specific comments: I found the manuscript difficult to read and perhaps some effort could be used to simplify language where possible, e.g. lines 279-281 "rare disease produced by compound epigenetic/genetic heterozygosity in a reverse1 (R1)/forward 2(F2)/reverse3(R3) trio of genes)," -with the WGS, did you consider trans variants as well, if not, what do you consider a cis variant?
-the layout of the figures is confusing. E.g. Figure 1 relates to one pedigree. Figure 2a also relates to this pedigree as does the top part of 2b, whereas the middle of 2b is the next pedigree and the bottom the other patient, before 2c goes on to relate to the second pedigree. Having figures correspond to the order they are mentioned in the text AND grouped into related information would make it easier for the reader.
- Figure 6b; they use the wrong control Reviewer #3 (Remarks to the Author): Methylmalonic aciduria and homocystinuria, cblC type, is rare recessive disease caused by biallelic loss of function of the MMACHC gene on chromosome 1. The authors initially identified three patients in which one MMACHC allele has a DNA sequence mutation and the other one an epimutation (promoter hypermethylation). In two families the epimutation is also present in the patient's fathers, and in at least one of them, it also appears to be present in sperm. Interestingly, the epimutation is in phase with a DNA sequence mutation in a neighbouring gene (PRDX1), which is transcribed towards the MMACHC gene from the other strand (convergent promoters). The authors show that the mutation leads to skipping of the last PRDX1 exon and the transcription termination signal, so that the polymerase reads -in antisense -through the MMACH1 gene into the neighbouring gene on the other side (CCDC163P). It is likely that PRDX1 read through transcription induces methylation of the MMACHC promoter. This notion is supported by the presence of H3K36me3 at the MMACHC promoter and by hypomethylation and reactivation of MMACHC expression after silencing of PRDX1. In a consecutive study, the authors identified five more cases with an PRDX1 mutation and MMACHC epimutation.
In summary, Gueant et al. have identified an epimutation that is caused by a cis-acting DNA mutation. Similar cases involving read through transcription induced DNA methylation have been described before (e.g. HBA2 and MSH2), although in these cases aberrant methylation was not present in sperm. The authors claim that their study presents the first evidence for transgenerational transmission of an epimutation in humans.
For transparency I would like to state that I reviewed the manuscript previously for another journal. While the findings are very interesting and the paper has improved since the first submission, I have several points of criticism.
1. Aberrant HBA2 and MSH2 methylation in the afore-mentioned cases is restricted to somatic tissues, because the neighbouring genes LUC7L and EPCAM, respectively, are expressed in a tissuespecific manner. In contrast, PRDX1 is expressed in all tissues including germ cells and stem cells. Germ cell expression of PRDX1 most probably explains why MMACHC methylation also occurs in sperm. Thus, we are dealing with the transgenerational transmission of a DNA (PRDX1) mutation that causes MMACH1 methylation in each cell, and not with the transgenerational transmission of an epimutation. The MMACHC epimutation per se is not stable, because it easily lost after blocking PRDX1 read through ( Fig. 6; this also needs to be discussed more thoroughly). The term "transgenerational inheritance of an epimutation" is reserved for the inheritance of a (stable) epimutation that has occurred spontaneously in the absence of any DNA sequence change during the life of a parent. It is important to distinguish between these two different modes of inheritance, especially with respect to the discussion of the inheritance of acquired traits. Thus, the authors should not use the term "transgenerational inheritance of an epimutation" for their cases.
2. The authors claim that the epimuation is also present in sperm from the fathers of CHU-12122 (CHU-14061) and WG-3838 (CDH-867). However, MMACHC methylation and SNRPN methylation (contamination control) is only shown for CHU-14061. The MMACHC and SNPRPN data for CDH-867 need to be shown, also.
3. I am not convinced that the patients need to be reclassified as having a new type of Cbl inherited disorder, epi-cblC. I assume that MMACHC methylation mimics an MMACHC mutation and that the clinical features are the same. 5. Terminology is sometimes wrong or imprecise. Examples for wrong terminology: Line 113 and ff: "c.-302G genotype of the rs3748643 …2". This should read "allele" rather than "genotype".Lines 212-213"The level of anti-sense transcripts was high in control and case fibroblasts". In the controls, there is no MMACHC antisense-transcription, just the sense PRDX1 transcription. Lines 217-219: "The activation of cryptic acceptor sites and resulting skipping of exons and introns of MMACHC …". The exons and introns are not skipped; they are not recognized, because they are on the other strand. Line 235: "compound heterozygosity for a constitutional epimutation … in cis and MMACC coding mutation in trans." This is an inappropriate use of the words "cis" and "trans".
General comments: «The authors present several cases with a severe clinical diagnosis of an inborn error or cobalamin metabolism, which would usually be caused by recessive mutations of the MMACHC gene, who instead have a compound genetic coding mutation of one allele of MMACHC and an epimutation of the other allele. The data presented are compelling and, in my opinion, their conclusions are consistent with their data. Firstly, they demonstrate through bisulphite sequencing flanking an (uninvolved) promoter SNP in at least two of their cases that the MMACHC promoter methylation was monoallelic. They also show through allelic expression analyses of cDNA that this resulted in allelic loss of expression of one allele, and the allele affected by the epimutation was the wild-type allele (hence in trans with the germline coding mutation). They demonstrated these molecular characteristics in samples from other members of the respective families form prior generations, including the blood and, interestingly, also in the sperm of the carrier fathers. Hence they provide strong evidence that the MMACHC epimutations were transmitted from one generation to the next, intact via the gametes. This is the first time, to my knowledge, that this has bee demonstrated in humans. In other examples of epimutations, the epigenetic manifestations are erased in the spermatozoa, even if the epimutation is subsequently reinstated in the somatic cells of the offspring. These findings are also consistent with what, in the field, we term a "secondary" epimutation (follows Mendelian laws of intergenerational transmission because it is caused by an underlying genetic mutation in cis). This paper is also comprehensive in that, not only does it demonstrate the epigenetic manifestation provides a new mechanism of causation of this disease, but they also unravel the underlying cis-genetic cause, and furthermore, provide data from functional studies that corroborate their genetic findings. They identify the genetic basis of the epimutation as a cisacting splice mutation of the neighboring PRDX1 gene, which is expressed in the antisense direction on the opposite strand. This genetic mutation, they demonstrate, causes exon-skipping resulting in loss of the polyadenylation (transcription termination) signal, with consequent aberrant continuation of expression of PRDX1 into the MMACHC. Their finding of excess H3K36 trimethylation through this region is consistent with this extended antisense transcription. (No other gene mutations were identified by WGS or panel exome sequencing of other metabolic genes segregating with the phenotype in all families). The authors then proceed to demonstrate via siRNA transfection that reduced PRDX1 expression partially ameliorates the methylation in appropriate cell lines. The finding of this PDRX1 genetic mutation and it's functional impact are consistent with the inheritance pattern through 3 generations, the sequence of events leading to the epigenetic manifestations including elevated H3K36 trimethylation and MMACHC promoter methylation, and that the epimutation is isolated at MMACHC (which they provide convincing evidence of through their Human Methylation 450k array data analyses with relevant controls). All in all, the findings presented are novel, in the following respects: They show epimutation can serve as a cause for recessive disease by affecting one allele, while a genetic mutation accounts for the loss of function of the other allele. They show the epimutation is transmitted via the gamete, likely due to the high levels of expression of the causative mechanisms via PRDX1 antisense transcription in this cell type.»: We thank the reviewer for these appreciations of our findings.  We have revised the language throughout the manuscript and we have reworded the headings of the paragraphs of results section (limited to 60 characters) to address the suggestion of the reviewer. We have simplified the mechanistic scheme of Fig 6e as suggested.
Point 5: « It has been a pleasure reading this manuscript and I must congratulate the authors on a very impressive body of work. » We warmly thank Reviewer 1 for this very positive feedback and constructive suggestions.

Reviewer #2:
General comments: "The authors identify a rare genetic variant located in the PRDX1 gene that is recessive for cblC class of disorders of vitamin B12 metabolism. They demonstrate that this mutation affects a splice acceptor site, the consequence of which is read-through transcription of PRDX1 into the MMACHC gene in an antisense orientation. Antisense transcription through the promoter of MMACHC is associated with transcriptional silencing of this promoter and hypermethylation of the promoter associated CpG island. There is also an increase in H3K36me3 around this site in individuals with this mutation. H3K36me3 is a mark that is associated with active transcription of PolII and is involved in splicing, supporting the other evidence of antisense read-through of the promoter of the neighbouring MMACHC (i.e. aberrant transcripts). They provide some functional evidence for this hypothesis by showing that knock down of the aberrant PRDX1 restores transcription of MMACHC. This mutation was discovered as when inherited as a compound heterozygote with loss of function mutations within the MMACHC coding region, it is sufficient to cause cblC metabolic disorder due to loss of MMACHC expression. They argue that this is an example of a constitutive epimutation that is transgenerationally inherited as the hypermethylation phenotype is present in sperm in cases with paternal transmission.
Novelty: These findings are novel in that this is a rare and novel mutation that is associated with an inherited disorder and has diagnostic value for cases of suspected cblC in which homozygous MMACHC mutations are absent. With regards to the claim of a transgenerationally inherited epimutation; I would apply caution. This is clearly a genetic mutation that links to phenotype through altering the epigenetic state, as they show. The presence of promoter hypermethylation in sperm does not demonstrate transgenerational epigenetic inheritance, (i.e. failure to be erased either in primordial germ cells and/or the post-fertilisation embryo)." The evidence they present suggests that hypermethylation will be established in any tissue in which PRDX1 is expressed (including the male germ-line during spermatogenesis) as an obligatory consequence of the genetic mutation. They provide no direct evidence to suggest that there is failure to erase this during the developmental periods of genome-wide DNA methylation erasure. Furthermore, this is not the first example of read-through antisense transcription altering the epigenetic state of a promoter. Therefore, with regards to mechanistic insight, this is not novel (e.g. Tufarelli, C. et al. Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nature Genet. 34, 157-165 (2003).) I think the conclusions and work presented in the manuscript are reasonable. I have some minor comments with regards to specific analyses and presentation, provided below. I believe the greatest relevance and interest in this work will be for the diagnostic benefit of this specific disorder, rather than an insight into novel biological mechanisms." We thank the reviewer for these general appreciations of our work on its limitations, novelty and relevance.
Concerning the following comment: "With regards to the claim of a transgenerationally inherited epimutation, I would apply caution. This is clearly a genetic mutation that links to phenotype through altering the epigenetic state, as they show. The presence of promoter hypermethylation in sperm does not demonstrate transgenerational epigenetic inheritance, (i.e. failure to be erased either in primordial germ cells and/or the post-fertilisation embryo). ", we would like to point out that we do not claim that the transmission was directly "inherited". We have written that there is a transgenerational transmission of the epimutation, with the presence of the epimutation in three generations and its presence in sperm. In addition, we have written that the transmission is triggered by the antisense transcription of PRDX1 produced by the PRDX1 mutations, in one heading of the result section "The epimutation was generated by aberrant extension of anti-sense transcription of PRDX1 through the MMACHC exon1 and the MMACHC/CCDC163P bidirectional promoter." and in the second paragraph of the discussion: "The forced antisense transcription of MMACHC resulted from the skipping of the last exon of PRDX1. This skipping was the causative defect that produced the epimutation since the silencing of PRDX1 decreased the methylation of exon 1 and the promoter and restored the transcription of MMACHC, in WG-3838 and MeWo-LC1 cells." In regard to the comment of the reviewer, it seems that we have to explain this point more clearly. To address this comment, we have revised the second sentence of the discussion as follows: "The epimutation is directly involved in the mechanism of the disease and is transmitted in 3 generations through the forced antisense transcription of the adjacent mutated PRDX1." We have also revised the abstract with the following sentence "The epimutation was transmitted in three generations through PRDX1 mutations that forced antisense transcription of MMACHC, resulting in a H3K36me3 mark in the promoter." Concerning the following comment: "The evidence they present suggests that hypermethylation will be established in any tissue in which PRDX1 is expressed (including the male germ-line during spermatogenesis) as an obligatory consequence of the genetic mutation." we would like to indicate that we agree with the reviewer. This is what we have tried to express in the discussion: "The transgenerational transmission and the presence of the MMACHC epimutation in sperm may be explained by the ubiquitous high expression of PRDX1 in germ cells, stem cells, and somatic cells." We have revised the sentence as follows, to make this point clearer: "The transgenerational transmission and the presence of the MMACHC epimutation in DNA from sperm, fibroblasts and blood may be explained by the ubiquitous high expression of PRDX1 in germ cells, stem cells, and somatic cells." To address the following comment: "They provide no direct evidence to suggest that there is failure to erase this during the developmental periods of genome-wide DNA methylation erasure.", we have added the following sentence in the third paragraph of discussion "We demonstrated that the epimutation escaped spermatozoa erasure, in contrast to previous reports of other diseases 5,14 , but we could not assess the possibility of a hypothetical failure to erase this epimutation during early embryonic development." Regarding the following comments: "Furthermore, this is not the first example of read-through antisense transcription altering the epigenetic state of a promoter. Therefore, with regards to mechanistic insight, this is not novel (e.g. Tufarelli, C. et al. Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nature Genet. 34, 157-165 (2003).)" , we would like to point out that we agree with the reviewer. The article by Tufarelli et al. in Nat. Genet. 2003 is one of the previous examples of gene silencing by methylation through a gene disruption that caused antisense transcription across the CpG island of the promoter. However, these authors reported the absence of methylation in sperm, and studied only two generations. We have added a sentence of the third paragraph of discussion section to address this comment: "Gene silencing by methylation through a gene disruption that caused antisense transcription across the CpG island of the promoter was previously reported for HBA2 (alpha-thalassemia) and MLH1 (familial cancer syndrome)." More precisely, the initial genetic data of the cases of alpha-thalassemia were reported by Barbour, Tufarelli et al in a former article published in Blood, 2000. The Nat Genet paper cited by the referee corresponds to subsequent molecular studies in ES cells and transgenic mouse models. The report of these cases in the article of Blood indicates that the promoter was not methylated in sperm, as pointed out in our discussion. This lack of methylation is clearly indicated in page 803 of the article published in Blood: "the CpG island (H) associated with the remaining α2 gene on the α2-ZF chromosome was unmethylated in spermatocytes ( Figure 4C), but it was methylated in peripheral blood, EBV lymphocytes, and the interspecific hybrid". In addition, the methylation of α2 gene was studied only in two generations, e.g. the proband, his affected mother and his unaffected father (table 1).

Specific remarks
Point 1: " I found the manuscript difficult to read and perhaps some effort could be used to simplify language where possible, e.g. lines 279-281 "rare disease produced by compound epigenetic/genetic heterozygosity in a reverse1 (R1)/forward 2(F2)/reverse3(R3) trio of genes)," We have tried to tighten up the language throughout the manuscript and we have simplified headings of paragraphs in the Results section. We have revised the sentence: "rare disease produced by compound epigenetic/genetic heterozygosity in a reverse/forward/reverse trio of genes)"

Point 2: "-with the WGS, did you consider trans variants as well, if not, what do you consider a cis variant?"
As explained result section, we used an informative variant, rs3748643 to identify the position in cis of the PRDX1 mutation and the epimutation. This is explained in the Results section for CHU-12122: "We detected a c.-302G genotype of the rs3748643 c.-302 G>T polymorphism in the allele bearing the epimutation and a c.-302T genotype in the non-methylated allele. The epimutation and c.-302G genotype of rs3748643 were absent in DNA from the mother and maternal grandmother (Fig. 1 a,c)." and also in the same section for WG-4152: "The c.-302G genotype of rs3748643 was detected in the allele bearing the epimutation and the c.-302T genotype in the non-methylated allele (Fig. 2b,c)." These positions according to rs3748643 were confirmed in WGS. We have reworded the corresponding sentence of the Results section to address the comment of the reviewer: "The PRDX1 variants and the polymorphism rs3748643 associated with the epimutation were present in the same allele as evidenced by DNA sequencing and transmission in the heterozygous relatives." Point 3: "the layout of the figures is confusing. E.g. Figure 1 relates to one pedigree. Figure 2a also relates to this pedigree as does the top part of 2b, whereas the middle of 2b is the next pedigree and the bottom the other patient, before 2c goes on to relate to the second pedigree. Having figures correspond to the order they are mentioned in the text AND grouped into related information would make it easier for the reader." We have revised the layout of the figures as suggested. Figure 1 relates only to data from CHU-1222 and her relatives. Figure 2a is now Figure 1d as it relates to CHU-12122. The middle of 2b and 2c are now Fig 2a and Fig 2b, respectively. They relate to WG-3838 and relatives . Fig 2 c, d, and e present the data of HM250 methylome profiling in the cases and their relatives.
Point 4: "- Figure 6b; they use the wrong control" We used two types of controls in Fig 6b, control fibroblasts and absence of RNA extract in the RT-PCR. In lane 6, RT-PCR shows no detectable antisense RNA in RNA extracted from HDF control fibroblasts. When no RNA is added in the RT-PCR reaction mixture, no artifactual amplification is observed. We have revised the legend as follows: "Lanes 2, 4 and 6 correspond to RT-PCR of RNA from fibroblasts CHU-12122, WG-3838 and HDF (control fibroblasts). Lanes 3 and 5 correspond to control experiments without fibroblast RNA in the reaction mixture of RT-PCR. They show no artifactual amplification."

Reviewer #3:
General comments: "Methylmalonic aciduria and homocystinuria, cblC type, is rare recessive disease caused by biallelic loss of function of the MMACHC gene on chromosome 1. The authors initially identified three patients in which one MMACHC allele has a DNA sequence mutation and the other one an epimutation (promoter hypermethylation). In two families the epimutation is also present in the patient's fathers, and in at least one of them, it also appears to be present in sperm. Interestingly, the epimutation is in phase with a DNA sequence mutation in a neighbouring gene (PRDX1), which is transcribed towards the MMACHC gene from the other strand (convergent promoters). The authors show that the mutation leads to skipping of the last PRDX1 exon and the transcription termination signal, so that the polymerase reads -in antisense -through the MMACH1 gene into the neighbouring gene on the other side (CCDC163P). It is likely that PRDX1 read through transcription induces methylation of the MMACHC promoter. This notion is supported by the presence of H3K36me3 at the MMACHC promoter and by hypomethylation and reactivation of MMACHC expression after silencing of PRDX1. In a consecutive study, the authors identified five more cases with an PRDX1 mutation and MMACHC epimutation.

In summary, Gueant et al. have identified an epimutation that is caused by a cis-acting DNA mutation. Similar cases involving read through transcription induced DNA methylation have been described before (e.g. HBA2 and MSH2), although in these cases aberrant methylation was not present in sperm. The authors claim that their study presents the first evidence for transgenerational transmission of an epimutation in humans.
For transparency I would like to state that I reviewed the manuscript previously for another journal. While the findings are very interesting and the paper has improved since the first submission, I have several points of criticism." We thank the reviewer for these general appreciations of our work and we address the "several points of criticism" in the following answers: Point 1: "Aberrant HBA2 and MSH2 methylation in the afore-mentioned cases is restricted to somatic tissues, because the neighbouring genes LUC7L and EPCAM, respectively, are expressed in a tissue-specific manner. In contrast, PRDX1 is expressed in all tissues including germ cells and stem cells. Germ cell expression of PRDX1 most probably explains why MMACHC methylation also occurs in sperm. Thus, we are dealing with the transgenerational transmission of a DNA (PRDX1) mutation that causes MMACH1 methylation in each cell, and not with the transgenerational transmission of an epimutation. The MMACHC epimutation per se is not stable, because it easily lost after blocking PRDX1 read through ( Fig. 6; this also needs to be discussed more thoroughly). The term "transgenerational inheritance of an epimutation" is reserved for the inheritance of a (stable) epimutation that has occurred spontaneously in the absence of any DNA sequence change during the life of a parent. It is important to distinguish between these two different modes of inheritance, especially with respect to the discussion of the inheritance of acquired traits. Thus, the authors should not use the term "transgenerational inheritance of an epimutation" for their cases." We agree with the reviewer that PRDX1 is expressed in all tissues including germ cells and stem cells and that germ cell expression of PRDX1 most probably explains why MMACHC methylation also occurs in sperm. This is indicated in the third paragraph of discussion, "The transgenerational transmission and the presence of the MMACHC epimutation in sperm may be explained by the ubiquitous high expression of PRDX1 in germ cells, stem cells, and somatic cells." and "PRDX1 was also ubiquitously expressed in E7-E10 mouse embryos (http://www.informatics.jax.org; http://dbtmee.hgc.jp/) and adult humans (http://www.proteinatlas.org)." In contrast to the aberrant HBA2 and MSH2 methylation, we observed the aberrant methylation of MMACHC in somatic tissues and in sperm. We have revised the sentence as follows, to make this point clearer: "The presence of the MMACHC epimutation in DNA from sperm, fibroblasts and blood may be explained by the ubiquitous high expression of PRDX1 in germ cells, stem cells, and somatic cells." Concerning the following comments: "Thus, we are dealing with the transgenerational transmission of a DNA (PRDX1) mutation that causes MMACH1 methylation in each cell, and not with the transgenerational transmission of an epimutation. The epimutation is directly involved in the mechanism of the disease and is transmitted in 3 generations through the forced antisense transcription of the adjacent PRDX1 mutated gene." and "The term "transgenerational inheritance of an epimutation" is reserved for the inheritance of a (stable) epimutation that has occurred spontaneously in the absence of any DNA sequence change during the life of a parent. It is important to distinguish between these two different modes of inheritance, especially with respect to the discussion of the inheritance of acquired traits. Thus, the authors should not use the term "transgenerational inheritance of an epimutation" for their cases.", we would like to point out that we do not claim that this transmission is directly inherited. This is the reason why we use the term "transgenerational transmission" instead of "transgenerational inheritance". We have clearly written that the "transgenerational transmission" of the epimutation is triggered by the antisense transcription of PRDX1 produced by the PRDX1 mutations, in the second paragraph of the discussion: "The forced antisense transcription of MMACHC resulted from the skipping of the last exon of PRDX1. This skipping was the causative defect that produced the epimutation since the silencing of PRDX1 decreased the methylation of exon 1 and the promoter and restored the transcription of MMACHC, in WG-3838 and MeWo-LC1 cells." To make this point clearer and to address the comments of the reviewer, we have revised the second sentence of the discussion as follows: "The epimutation is directly involved in the mechanism of the disease and is transmitted in 3 generations through the forced antisense transcription of the adjacent mutated PRDX1." We have also revised the abstract with the following sentence "The epimutation was transmitted in three generations through PRDX1 mutations that forced antisense transcription of MMACHC, resulting in a H3K36me3 mark in the promoter." Regarding the comment on the MMACHC epimutation stability and its loss after blocking PRDX1, we would like to point out that our experimental data do not support that this epimutation is easily lost and unstable, considering that the silencing of PRDX1 produced only 10-15% hypomethylation of the allele initially fully methylated. We have discussed this point in the second paragraph of discussion: "However, the silencing of PRDX1 produced only 10-15% hypomethylation of the allele initially fully methylated, suggesting a limited reversibility of the epimutation (Fig. 6g).
Point 2: " The authors claim that the epimutation is also present in sperm from the fathers of CHU-12122 (CHU-14061) and . However, MMACHC methylation and SNRPN methylation (contamination control) is only shown for CHU-14061. The MMACHC and SNPRPN data for CDH-867 need to be shown, also." We have now added these data in the revised figure 2b (Sanger of bisulfated DNA), Fig 2e (HM 450 methylome profiling) and in the supplementary figure S1 (methylation of SNRP imprinted gene in sperm from proband's fathers and two controls). Point 3. "I am not convinced that the patients need to be reclassified as having a new type of Cbl inherited disorder, epi-cblC. I assume that MMACHC methylation mimics an MMACHC mutation and that the clinical features are the same." We consider "epi-cblC" as a new cause for the cblC inherited disorder, but not as a new group of Cbl inherited disorders. This is now more clearly indicated to address the comment of the reviewer, in the summary and in the introduction section: "We report the transgenerational transmission of a new cause of the autosomal recessive cblC class of inborn errors of vitamin B12 metabolism that we named "epi-cblC", the first heading of the Results section: "Identification of a new cause of the cblC disorder named epi-cblC…" and the first sentence of the discussion : "Our "epi-CblC| cases represent a new cause for the autosomal recessive cblC disorder…". However, we cannot consider our epi-cblC cases as "classical" cblC patients as we do not know whether the PRDX1 heterozygous mutations could have an influence on the clinical symptoms throughout life. In addition, it is important to distinguish the epi-CblC cases from the classical cblC patients for preimplantation genetic diagnosis. This was the case for the parents of case CHU-12122. Point 4. "Bisulfite sequencing of the MMACHC promoter in sperm DNA of 14061 should be shown in Fig 1, not Fig. 2

."
We have modified the Figure 1 and Figure 2 as suggested, in addition to the changes indicated by reviewer 2.
Point 5: « Terminology is sometimes wrong or imprecise. Examples for wrong terminology: Line 113 and ff: "c.-302G genotype of the rs3748643 …2". This should read "allele" rather than "genotype".Lines 212-213"The level of anti-sense transcripts was high in control and case fibroblasts". In the controls, there is no MMACHC antisense-transcription, just the sense PRDX1 transcription. Lines 217-219: "The activation of cryptic acceptor sites and resulting skipping of exons and introns of MMACHC …". The exons and introns are not skipped; they are not recognized, because they are on the other strand. Line 235: "compound heterozygosity for a constitutional epimutation … in cis and MMACC coding mutation in trans." This is an inappropriate use of the words "cis" and "trans"." We have changed wording from "genotype" to "allele" We thank the reviewer for pointing out the error in Lines 212-213. As shown in Fig 6b, the MMACHC antisense-transcription was undetectable in control fibroblasts. We have corrected the sentence as follows: "The level of anti-sense transcripts was high in case fibroblasts and undetectable in control fibroblasts." Regarding the comment on Lines 217-219, the RNA seq experiments showed that the forced antisense transcription produced the activation of cryptic acceptor sites that generated several antisense MMACHC transcripts. We also identified some of these transcripts by RT-PCR.
We deleted "exons and introns" and we reworded the related sentences from the revised first paragraph of page 7, to address the comment: "The forced antisense transcription produced the activation of cryptic acceptor sites that generated several antisense MMACHC transcripts" and "This transcript resulted from the activation of a cryptic antisense splicing site located in the middle of MMACHC exon 1 (Fig. 6d,e)." To answer the comment on Line 235, we have deleted trans and cis and indicated that "Our epi-Cbl cases represent a new type of cblC inherited disorder, resulting from hypermethylation of the CpG island in the bidirectional promoter of MMACHC on one allele and a coding mutation on the other. " We thank the reviewers for their help in improving the quality of our manuscript and we hope that it is now acceptable for publication.