We would like to comment on Dixon et al., 2012 who questioned the pathogenicity of private NLRP7 variants found in a heterozygous state in singleton cases with one to three hydatidiform moles. The interesting part of this work is in the analysis of nine patients with recurrent androgenetic moles despite that only reports demonstrating that three of these patients had had recurrent androgenetic moles are provided either in the supplementary materials (for case 6)[1] or referenced in the Materials and Methods and had been previously reported by the same group (for case 2 in Buyukurt et al.[2] and case 1 in van der Smagt et al.[3]). Among patients with recurrent androgenetic moles, the authors found only one patient with a non-synonymous variant, R413Q that is not reported in the 1000 Genomes database. Curiously, we previously reported another novel DNA substitution affecting the same amino acid, R413W, in a patient with one mole and one live birth[5].

We agree with the authors that finding a rare or private variant in a patient but not in controls does not imply its pathogenicity. However, when several patients, each with a single rare variant, are found, and if these variants are not found in controls [4-6], such variants cannot be ignored, and the question of their pathogenicity should be raised and experimental methods addressing this question should be performed. Our group has reported ten different variants in heterozygous state in patients with one to three moles, but not in our controls. Determining the pathogenicity of private missense variants is very challenging in human genetics. Below, we discuss the arguments that are in favour of, or against, the pathogenicity of the private missense we reported in heterozygous state.

First, a good and simple genetic argument to determine if a variant is pathogenic is its absence in controls, which was done for every variant. The fact that some of these variants were later reported in some populations in the 1000 Genomes database is not against their involvement in the disease. Because in these populations, these variants were found at low frequencies that are very close to the incidences of sporadic moles in these populations. Basically, the definition of a susceptibility variant implies its presence in the general population and only a subset of the subjects carrying it will develop the disease.

Second, another good genetic argument is to show segregation of the variant in question in families with other mutations but on different haplotypes. This was done for one missense, A719V, that we reported in two patients in a heterozygous state[4, 5,6]. This variant was recently reported in a fourth patient and again in a heterozygous state and not in controls [7]. Definitely, more data are needed for the other variants.

Third, as mentioned by Dixon et al., the predicted pathogenicity scores by Polyphen-2 for seven of the ten variants we reported as single defective alleles are equal or higher than the predicted pathogenicity scores for missenses seen in patients with two defective alleles.

Fourth, to date, the only known functional role of NLRP7 is its requirement for normal IL1? secretion by peripheral blood mononuclear cells that we demonstrated in cells carrying different NLRP7 mutations [8] and was confirmed using independent in vitro studies [9]. To investigate whether the identified single rare variants in our patients affect IL1? secretion, we had performed the cytokine assay on patients cells carrying each of the four variants, C399Y, G380R, A719V (Fig. 1 in8), and R413W (unpublished data), in a heterozygous state, and again these cells displayed reduced IL1? and TNF secretion.

Fifth, the reproductive outcomes of patients with one defective allele are milder (approximately 81.6% reproductive wastage, 18.4% live birth) than those of patients with two defective alleles (97.5% reproductive wastage, 2.5% live births) [5] and more severe than those of patients with sporadic moles (10-22% reproductive wastage, 90-78% live births) [10].

One of the arguments used by Dixon et al., [1] is that androgenetic moles are not caused by inherited variants is the fact that androgenetic moles have not been seen in families. While we agree with this good argument, we believe that this may not be the only explanation. Another possibility is that a genetic defect leading to an androgenetic mole implies having an oocyte whose pronucleus, by as yet unknown mechanisms, is not able to meet and fuse with the male pronucleus and form a diploid biparental zygote. Such defect may be very severe, lethal, and not compatible with the life of the patient when caused only by a genetic defect in a homozygous state. However, in a heterozygous state, such defect would be expected to be milder, compatible with the life of the patient, and will confer genetic susceptibility to androgenetic moles only in few of her pregnancies when combined with other environmental and/or genetic factors.

Taking into consideration the emerging views about the mechanisms leading to androgenetic moles and the presence of complex postzygotic aneuploidies at their origin, it is therefore unlikely to see the same resulting parental contribution to products of conception occurring in other family members and consequently such conceptions will have other histopathological features and lead to other forms of reproductive wastage (such as spontaneous abortions, blighted ovum, ectopic pregnancies, etc.). This would make it difficult to health care professionals to connect these different clinical entities under the same unifying common defect segregating in the family. We did notice familial aggregation of reproductive wastage in the relatives of some patients with moles but these are difficult to detect and require well-prepared questionnaire and pedigrees. In few instances, we saw patients with sporadic moles who reported the occurrence of moles in their mothers. Although, it is impossible to review the histopathology of the mother's moles to validate the patients' statements, which is a real limitation in this research field, however, based on family history, there are rare cases with dominant transmission of a mild genetic susceptibility. Because males do not manifest moles, theoretically, only 50% of families of patients who carry predisposing genetic variants are at risk to manifest reproductive wastage and of these families only those with females who have tried to conceive may manifest the disease. For all these reasons and the genetic complexity of the mechanisms leading to androgenetic moles, we believe that a very small fraction of patients with a single missense "mutation" would be expected to have another member who had had androgenetic moles.

In conclusion, we agree with Dixon et al., that recurrent androgenetic moles do not seem to be caused by a strong causal genetic defect. However, we believe that there are several arguments, at least for the time being, indicating that some of the single rare variants in NLRP7 predispose the patients to moles and reproductive wastage. With the advances in this field, the accumulation of more data from various groups, and the development of additional functional tests to determine easily the pathogenicity of rare variants of unknown significance, patients with one mutation or one private variant can be assured that their chances of having live births are much higher than patients with two defective alleles5.

References

1 Dixon PH, Trongwongsa P, Abu-Hayyah S, Ng SH, Akbar SA, Khawaja NP, Seckl MJ, Savage PM, Fisher RA. Mutations in NLRP7 are associated with diploid biparental hydatidiform moles, but not androgenetic complete moles. J Med Genet 2012;49(3):206-11.

2 Buyukkurt S, Fisher RA, Vardar MA, Evruke C. Heterogeneity in recurrent complete hydatidiform mole: presentation of two new Turkish families with different genetic characteristics. Placenta 2010;31(11):1023 -5.

3 van der Smagt JJ, Scheenjes E, Kremer JA, Hennekam FA, Fisher RA. Heterogeneity in the origin of recurrent complete hydatidiform moles: not all women with multiple molar pregnancies have biparental moles. Bjog 2006;113(6):725-8.

4 Deveault C, Qian JH, Chebaro W, Ao A, Gilbert L, Mehio A, Khan R, Tan SL, Wischmeijer A, Coullin P, Xie X, Slim R. NLRP7 mutations in women with diploid androgenetic and triploid moles: a proposed mechanism for mole formation. Hum Mol Genet 2009;18(5):888-97.

5 Messaed C, Chebaro W, Roberto RB, Rittore C, Cheung A, Arseneau J, Schneider A, Chen MF, Bernishke K, Surti U, Hoffner L, Sauthier P, Buckett W, Qian J, Lau NM, Bagga R, Engert JC, Coullin P, Touitou I, Slim R. NLRP7 in the spectrum of reproductive wastage: rare non-synonymous variants confer genetic susceptibility to recurrent reproductive wastage. J Med Genet 2011;48(8):540-8.

6 Qian J, Cheng Q, Murdoch S, Xu C, Jin F, Chebaro W, Zhang X, Gao H, Zhu Y, Slim R, Xie X. The Genetics of Recurrent Hydatidiform Moles in China: Correlations between NLRP7 Mutations, Molar Genotypes, and Reproductive Outcomes. Mol Hum Reprod 2011;17(10):612-9.

7 Landolsi H, Rittore C, Philibert L, Hmissa S, Gribaa M, Touitou I, Yacoubi MT. NLRP7 mutation analysis in sporadic hydatidiform moles in Tunisian patients: NLRP7 and sporadic mole. Arch Pathol Lab Med 2012;136(6):646-51.

8 Messaed C, Akoury E, Djuric U, Zeng J, Saleh M, Gilbert L, Seoud M, Qureshi S, Slim R. NLRP7, a NOD-like receptor protein, is required for normal cytokine secretion and co-localizes with the Golgi and the microtubule organizing center. J Biol Chem 2011;286(50):43313-433123.

9 Khare S, Dorfleutner A, Bryan NB, Yun C, Radian AD, de Almeida L, Rojanasakul Y, Stehlik C. An NLRP7-containing inflammasome mediates recognition of microbial lipopeptides in human macrophages. Immunity 2012;36(3):464-76.

10 Lan Z, Hongzhao S, Xiuyu Y, Yang X. Pregnancy outcomes of patients who conceived within 1 year after chemotherapy for gestational trophoblastic tumor: a clinical report of 22 patients. Gynecol Oncol 2001;83(1):146-8.

Conflict of Interest:

None declared

We have read the entire paper with great care. We totally agree with the findings of the respected authors in this respective article. The article seems to be used by many researchers as a reference article so we need to update the status of SMA diagnosis for non deleted SMA patients. Many new approaches have been employed to diagnose non deleted SMA. One of such methods is long range PCR method (LR-PCR) (Clemont et al., 2004). Secondly, we can not just focus on the mutations described in this article. We need to focus on all known mutations in the SMN1 gene in non deleted SMA patients. Although Clermont and colleagues claimed that 90% of the mutations in the SMN1 gene can be found by their method (Clermont et al.,2004), yet many other groups have reported individual mutations. We would like to enumerate the list of all these mutations. All the mutations are divided into four main groups; missence mutation, Nonsense mutation, Frame-shift mutation and Splice site mutations.

The missence mutations (total 25) have been reported in Ex.1 in type II and III SMA patients (Parsons et al., 1998), Ex. 2a in type II and III SMA patients (Sun et al., 2005), Ex.3 in type I (Cusco et al., 2004, Clermont et al., 2004, Prior et al., 2007, Sun et al., 2005, Koatni et al., 2007), Ex.3 in type II (Clermont et al., 2004, Sun et al., 2005), Ex.3 in type III SMA patient (Sun et al., 2005, Ex.4 in type I SMA patient (Zapletalova et al., 2007), Ex. 6 in type III SMA patients (Rochette et al., 1997, Clermont et al., 2004, Sun et al., 2005, Hahnen et al., 1997, Parson et al., 1998, Wirth et al., 1999), Ex.6 in type I SMA patient (Clermont et al., 2004), Ex.6 in type II SMA patient (Alias et al., 2009, Prior et al., 2007, Lefebvre et al., 1995, Wirth et al., 1999, Rochette et al., 1997, Clermont et al., 2004, Zapletalova et al., 2007, Hahnen et al., 1997, Parson et al., 1998, Sun et al., 2005, Ex.6 in type III SMA patients (Hahnen et al., 1997, Parsons et al., 1998, Wirth et al., 1999, Sun et al., 2005, Zapletalova et al., 2007, Alias et al., 2009, Burglen et al., 1996, Skordiset et al., 2001), Ex.7 in type I SMA patient (Talbot et al., 1997) and Ex.7 in type II and III SMA patients (Wang et al., 1998). The nonsense mutations (total 5) were curated to be included in Ex.1 in type I SMA patient (Wirth et al., 1999), Ex.1 in type III SMA patient (Sun et al., 2005), Ex.3 in type II SMA patient (Sossi et al., 2001, Prior et al., 2007), Ex.3 in type III SMA patient (Wirth et al., 1999, Sun et al., 2005, Sossi et al., 2001, Prior et al., 2007), Ex.3 in type I SMA patient (Bricchta et al., 2008), Ex.4 in type I SMA patient (Alias et al., 2009), Ex.5 in type I SMA patient (Tsai et al., 2001). Considering the fram-shift mutations (total 18), it included mutation in Ex.1 of type I SMA patient (Wirth et al., 1999), Ex.1 in type III SMA patient (Wirth et al., 1999), Ex.2a in type I SMA patient (Prior et al., 2007, Zapletalova et al., 2007), Ex.2a in type I SMA patient (Wirth et al., 1999, Alias et al., 2008), Ex.2b in type I SMA patient (Clermont et al., 2004), Ex. 2b in type III SMA patient (Wirth et al., 1999), Ex.3 in type I SMA patient (Brahe et al., 1996, Sossi et al., 2001, Alias et al., 2008), Ex.3 in type II SMA patient (Bussaglia et al., 1995, Martin et al., 2002, Cusco et al., 2003, Alias et al., 2008), Ex.3 in type III SMA patient (Alias et al., 2008, Clermont et al., 2004, Bussaglia et al., 1995, Martin et al., 2002, Cusco et al., 2003), Ex.4 in type I SMA patient (Clermont et al., 1997, Clermont et al., 2004), Ex.4 in type II SMA patient (Parons et al., 1998, Wirth et al., 1999), Ex.4 in type III SMA patient (Parson et al., 1998), Ex.5 in type I, II and III SMA patients (Wirth et al., 1999, Clermont et al., 2004), Ex.6 in type I, II and III SMA patients (Wirth et al., 1999, Clermont et al., 2004, Parsons et al., 1996, Parsons et al., 1998, Martin et al 2002, Clermont et al., 2004, Alias et al., 2008, Martin et al., 2002). The splice site mutations in the SMN1 gene in non deleted SMA patients include 40Int.4 in type I SMA patient (Brichta et al., 2008), 41Int.6 in type I SMA patient (Martin et al., 2002), 42Int.6 in type I SMA patient (Eggermann et al., 2008), 43Int.6 in type I SMA patient (Lefebvre et al., 1995), 44Int.7 in type I SMA patient (Wirth et al., 1999) and 45Int.7 in type II SMA patient (Lefebvre et al., 1995). The copy number analysis of the SMN1 gene is not enough as the SMA modified genes; SMN2 and NAIP are reported to modify the disease severity (Watihayati et al., 2009). We will emphasize that the title of the article is making the readers confusing as considering the article to be the diagnostic protocol for the spinal muscular atrophy in non deleted SMA patients. The severity of the disease could be estimated by many parameters even in the non deleted SMA patients and recently published report has prooven that (Watihayati et al., 2009) beside the report of Teguh and colleagues which states that only PCR detection of exon7 of the SMN1 is enough to diagnose SMA (Teguh et al., 2011) but yet the authors did not explain the diagnosis of non deleted SMA patients. Very humbly we would like to add this response with a request to the respected authors to consider the findings of SMN2 copy number and the NAIP gene deletion with the available data of this patient.

Conflict of Interest:

None declared

It was with great interest that I read the study by Tang and co- authors [1], in which they discovered twenty-five novel mutations in DNA polymerase gamma. Based on the presence of p.G268A substitution in heterozygosis in 19 subjects from a cohort of 2697 unrelated patients, they proposed to reclassify this mutation as a neutral polymorphism or a polymorphic modifier rather than a pathological mutation. Smith and co- authors [2] confirmed that POLG G268A is not pathogenic but represent neutral polymorphism on the basis of the presence of this mutations in three healthy subjects and on the basis of the frequencies of the allele p.G268A in their patient cohort, which were similar to those found in a control population of European origin by the NHLBI exome sequencing project [3].

I and co-authors previously demonstrated that in yeast the mip1G224A allele, corresponding to the human p.G268A substitution, was associated with a 2.2-fold increase in petite frequency, which is the frequency of mutants with large deletions and/or loss of mtDNA [4]. A further analysis made in our laboratory on yeast mip1 mutations corresponding to human substitutions classified as neutral SNPs [5] conserved from yeasts to mammals (E. Baruffini and T. Lodi, unpublished results), suggested the existence of a class of substitutions, that also includes G224A, which are not neutral. These mutations caused a 1.5 to 2.5-fold increase in petite frequency, which was lower compared to the petite frequency caused by mutations recognized as pathological, for which at least a 10-fold increase in petite frequency or the total loss of mtDNA were observed in our laboratory [4, 6-9]. This class includes also the mip1E900G and mip1Q766C alleles, corresponding to the human p.E1143G and p.R964C substitutions, respectively. The former (2-fold increase in petite frequency) is a substitution classified in humans as a phenotypic modifier that can modulate disease mutations [10, 11]. The latter (2-fold increase) seems to be pathological in humans when in heterozygosis with p.A862T [8,12] or with p.A962T [1] but not pathological in homozygosis [13]; however it is associated with mitochondrial toxicity susceptibility to stavudine both in humans [13] and in yeast [14] and the mutant DNA polymerase gamma showed an impaired polymerase activity [15].

Conclusion: analysis in yeast suggests that p.G268A is neither a pathological mutation nor a neutral SNP. It can be classified, together with R964C and E1143G, as a "phenotypic modifier", as suggested by Copeland [10], as an "unclassified variant", as suggested by Tang and co- authors [1] or as an "Ecogenetic Single Nucleotide Variant (ESNV)", as suggested by Saneto and Naviaux for genetic variations whose phenotypic expression is determined by interaction with genetic, epigenetic and environmental factors [16]. In general, these observations suggest that the validation of the pathological significance of novel mutations, whose discovery is exponentially increasing thanks to the next generation sequencing, can take advantage, when possible, of experimental analysis in model systems to confirm or disavow their role in the disease.

References 1. Tang S, Wang J, Lee NC, Milone M, Halberg MC, Schmitt ES, Craigen WJ, Zhang W, Wong LJ. Mitochondrial DNA polymerase gamma mutations: an ever expanding molecular and clinical spectrum. J Med Genet 2011;48:669-681. 2. Smith C. POLG p.G268A and p.G517V are not pathogenic mutations. J MedGenet 2011; eletter 3. NHLBI exome sequencing project, https://esp.gs.washington.edu/drupal/ 4. Baruffini E, Lodi T, Dallabona C, Puglisi A, Zeviani M, Ferrero I. Genetic and chemical rescue of the Saccharomyces cerevisiae phenotype induced by mitochondrial DNA polymerase mutations associated with progressive external ophthalmoplegia in humans. Hum Mol Genet 2006;15:2846 -2855. 5. dbSNP, http://www.ncbi.nlm.nih.gov/snp/ 6. Baruffini E, Ferrero I, Foury F. Mitochondrial DNA defects in Saccharomyces cerevisiae caused by functional interactions between DNA polymerase gamma mutations associated with disease in human. Biochim Biophys Acta 2007;1772:1225-1235. 7. Spinazzola A, Invernizzi F, Carrara F, Lamantea E, Donati A, Dirocco M, Giordano I, Meznaric-Petrusa M, Baruffini E, Ferrero I, Zeviani M. Clinical and molecular features of mitochondrial DNA depletion syndromes. J Inherit Metab Dis 2009;32:143-158. 8. Stricker S, Pr?ss H, Horvath R, Baruffini E, Lodi T, Siebert E, Endres M, Zschenderlein R, Meisel A.J. A variable neurodegenerative phenotype with polymerase gamma mutation. J Neurol Neurosurg Psychiatry 2009;80:1181 -1182. 9. Baruffini E, Horvath R, Dallabona C, Czermin B, Lamantea E, Bindoff L, Invernizzi F, Ferrero I, Zeviani M, Lodi T. Predicting the contribution of novel POLG mutations to human disease through analysis in yeast model. Mitochondrion 2011;11:182-190. 10. Human DNA Polymerase Gamma Mutation Database, http://tools.niehs.nih.gov/polg/ 11. Chan SS, Longley MJ, Copeland WC. Modulation of the W748S mutation in DNA polymerase gamma by the E1143G polymorphismin mitochondrial disorders. Hum Mol Genet 2006;15:3473-3483. 12. Wong LJ, Naviaux RK, Brunetti-Pierri N, Zhang Q, Schmitt ES, Truong C, Milone M, Cohen BH, Wical B, Ganesh J, Basinger AA, Burton BK, Swoboda K, Gilbert DL, Vanderver A, Saneto RP, Maranda B, Arnold G, Abdenur JE, Waters PJ, Copeland WC. Molecular and clinical genetics of mitochondrial diseases due to POLG mutations. Hum Mutat 2008;29:E150-E172. 13. Yamanaka H, Gatanaga H, Kosalaraksa P, Matsuoka-Aizawa S, Takahashi T, Kimura S, Oka S. Novel mutation of human DNA polymerase gamma associated with mitochondrial toxicity induced by anti-HIV treatment. J Infect Dis 2007;195:1419-1425. 14. Baruffini E, Lodi T. Construction and validation of a yeast model system for studying in vivo the susceptibility to nucleoside analogues of DNA polymerase gamma allelic variants. Mitochondrion 2010;10:183-187. 15. Bailey CM, Kasiviswanathan R, Copeland WC, Anderson KS. R964C mutation of DNA polymerase gamma imparts increased stavudine toxicity by decreasing nucleoside analog discrimination and impairing polymerase activity. Antimicrob Agents Chemother 2009;53:2610-2612. 16. Saneto RP, Naviaux RK. Polymerase gamma disease through the ages. Dev Disabil Res Rev 2010;16:163-174.

Conflict of Interest:

None declared

In their recently published paper describing mutations in mitochondrial DNA polymerase gamma, Tang et al.(1) propose that the POLG p.G268A (c.803G>C) and p.G517V (c.1550G>T) variants which have previously been reported as pathogenic mutations should be considered as unclassified variants that may represent rare neutral polymorphisms or polymorphic modifiers.

We have also identified these variants in our own cohort of 627 patients that were referred with a suspected disorder of mitochondrial DNA maintenance. We detected p.G268A in 9 individuals (allele frequency 0.72%) of ages from 3 months to 63 years. Symptoms were variable ranging from severe early onset mitochondrial DNA depletion syndrome to milder late onset neuropathy/ataxia/ophthalmoplegia. In all 9 cases the variant occurred as a heterozygous change with no other pathogenic POLG mutation, suggesting that p.G268A is not a recessive mutation. Parental samples were available in 3 cases, and for each of these we found the variant in an unaffected parent, suggesting that p.G268A is very unlikely to be a dominant mutation. Furthermore, one of the 9 index cases was later found to be compound heterozygous for pathogenic mutations in another gene associated with mtDNA maintenance disorders, DGUOK. In another case the proband had a similarly affected sibling who did not have p.G268A.

Similarly, we have identified p.G517V 8 times (allele frequency 0.64%), always as a heterozygous change with no other pathogenic POLG mutation. The age of the individuals ranged from 1-81 yrs and the phenotype varied from infantile epilepsy/failure to thrive to adult onset ophthalmoplegia/ataxia/myopathy. The unaffected parent of 1 of these individuals was also heterozygous for p.G517V. A second affected individual was later found to have pathogenic mutations in RRM2B.

Furthermore, the allele frequencies of p.G268A and p.G517V in our patient cohort are remarkably similar to those found in a large control population of European origin (2700 alleles) by the NHLBI exome sequencing project (2); 0.59% and 0.86% respectively.

Therefore, our data, taken together with that for Tang et al. and the NHLBI exome sequencing project, confirm that POLG p.G268A and p.G517V are not pathogenic and are highly likely to represent neutral polymorphisms.

References. (1) Tang S, Wang J, Lee NC, Milone M, Halberg MC, Schmitt ES, Craigen WJ, Zhang W, Wong LJ. Mitochondrial DNA polymerase gamma mutations: an ever expanding molecular and clinical spectrum. J Med Genet 2011;48:669-681. (2) https://esp.gs.washington.edu/drupal/

Conflict of Interest:

None declared