Frank L. Barnes

Mitochondrial DNA quantitation—making sense of contradictory reports

Sir, We read the article by Ravichandran et al., which appeared in the June 2017 issue of Human Reproduction, with interest (Ravichandran et al., 2017). The manuscript reports the findings of a retrospective analysis of mitochondrial DNA (mtDNA) levels in blastocyst transfers, and concludes that ‘mtDNA quantity can serve as an independent biomarker for the prediction of euploid blastocyst implantation potential’. In the article, the authors mention our study, which in contrast to theirs did not show a statistically significant correlation between mtDNA levels and implantation (Victor et al., 2017). Ravichandran et al. question the conclusions of our study asserting that it was ‘hampered by a variety of technical issues’. We would like to take the opportunity to comment on the points listed as technical concerns. First, the authors claim that the Next Generation Sequencing (NGS) technique used in our study is incapable of accurately quantifying mtDNA levels as it is ‘known to provide insufficient coverage of the mitochondrial genome’. This assertion is not followed by a reference. To the best of our knowledge, there is no published report that leads to such a conclusion. To be able to make this claim, it is necessary to perform a proper analysis such as by comparing different detection platforms in samples with known mtDNA quantities. We carried out such a cross-platform validation in our study (Victor et al., 2017). Second, the authors point out that the absolute quantitation we performed fails to account for sample-to-sample variation due to technical batch effect and biopsy sample size. Here, the authors clearly had a misconception about the method employed in our study. We carried out absolute quantitation of a locus in the mtDNA ‘as well’ as a locus in the nuclear DNA, and calculated the ratio between the two in order to normalize for the variables stated above. Third, the use of a single copy locus in the nuclear DNA for normalization is declared to lead to ‘almost meaningless results’ due to alleledrop-out during whole genome amplification, something the authors test experimentally in the manuscript. It is unclear how well the qRTPCR assays used in their experiment were validated, as no standard curves or serial dilutions are shown testing the efficiencies of the various assays. We would like to point to a more recent study that makes use of a nuclear multi-copy locus and still concludes that mtDNA quantitation has no predictive power in regards to implantation (Treff et al., 2017). On a different note, we are generally puzzled by the authors’ focus on our technical methods, when there is a clear explanation how our studies could be reconciled and both be ‘right’, as explained below. All data in our paper stemmed from transfers at a single IVF center, thereby controlling for possible inter-clinic variables such as culture conditions, biopsy techniques, or equipment. In their manuscript, Ravichandran et al. indicate that multiple clinics provided data to their study, and they show stark differences in percentage of embryos that contain elevated mtDNA levels at each site (Table I). In fact, approximately half (17 out of 35) of the clinics that contribute data to their study did not show any blastocysts with elevated mtDNA amounts. The most logical conclusion is that our center is one of those that do not produce embryos with high mtDNA levels. From this, we gather that the phenomenon of elevated mtDNA is not a universal biological occurrence that happens stochastically in a percentage of embryos in any given setting, but is rather a center-induced event. We feel that a much more intriguing and fundamental question is what could be eliciting these large deviations in mtDNA levels across clinics.

Production of embryos from in vitro-matured primary human oocytes**Supported by a grant from IVF America Corporation, Purchase, New York.

OBJECTIVE To determine the factors that influence the number and quality of embryos produced from primary oocytes collected from untreated regularly ovulating and irregular or anovulatory polycystic women. DESIGN A direct comparison between two patient groups whose oocytes were matured in vitro and a comparison of the embryo development of in vitro-matured oocytes from untreated patients with in vivo-matured oocytes of superovulated IVF-ET patients obtained during the same period. SETTING The Monash IVF Clinic, involving patients who expressed the desire to avoid super-ovulation with fertility drugs. MAIN OUTCOME MEASURES The completion of nuclear maturation of oocytes after 36 or 48 hours culture, fertilization in vitro, and embryo development ratio. RESULTS Oocytes from regular cycling patients matured and fertilized at significantly higher rates than irregular cycling and anovulatory women and their embryos had significantly higher mean embryo development ratio. The mean embryo development ratio of embryos of regular cycling patients was similar to superovulated IVF patients but irregular cycling and anovulatory patients had a significantly lower embryo development ratio. Culture of oocytes for 48 hours increased maturation of oocytes from 57% to 82% but did not affect fertilization or cleavage rates. Embryo development was not affected significantly by the grade of follicular cell cover of oocytes. CONCLUSIONS The developmental capability of primary oocytes is higher in regular cycling women than in irregular cycling and anovulatory women with polycystic ovary disease.