Search

Humans present remarkable diversity in their mitochondrial DNA (mtDNA) in terms of variants across individuals as well as across tissues and even cells within one person. We have investigated the timing of the first appearance of this variant-driven mosaicism. For this, we deep-sequenced the mtDNA of 254 oocytes from 85 donors, 158 single blastomeres of 25 day-3 embryos, 17 inner cell mass and trophectoderm samples of 7 day-5 blastocysts, 142 bulk DNA and 68 single cells of different adult tissues. We found that day-3 embryos present blastomeres that carry variants only detected in that cell, showing that mtDNA mosaicism arises very early in human development. We classified the mtDNA variants based on their recurrence or uniqueness across different samples. Recurring variants had higher heteroplasmic loads and more frequently resulted in synonymous changes or were located in non-coding regions than variants unique to one oocyte or single embryonic cell. These differences were maintained through development, suggesting that the mtDNA mosaicism arising in the embryo is maintained into adulthood. We observed a decline in potentially pathogenic variants between day 3 and day 5 of development, suggesting early selection. We propose a model in which closely clustered mitochondria carrying specific mtDNA variants in the ooplasm are asymmetrically distributed throughout the cell divisions of the preimplantation embryo, resulting in the earliest form of mtDNA mosaicism in human development.

Children born using assisted reproductive technologies (ART) have an increased risk of a lower birth weight, the cause of which remains unclear. As a causative factor, we hypothesized that variants in the mitochondrial DNA (mtDNA) that are not associated with disease, may explain changes in birth weight. We deep-sequenced the mtDNA of 451 ART and spontaneously conceived (SC) individuals, 157 mother-child pairs and 113 individual oocytes from either natural menstrual cycles or cycles with ovarian stimulation (OS). The mtDNA genotypes were compared across groups and logistic regression and discriminant analysis were used to study the impact of the different factors on birth weight percentile. ART individuals more frequently carried variants with higher heteroplasmic loads in protein and rRNA-coding regions. These differences in the mitochondrial genome were also predictive of the risk of a lower birth weight percentile, irrespective of the mode of conception but with a sex-dependent culture medium effect. The higher incidence of these variants in ART individuals results both from maternal transmission and de novomutagenesis, which we found not to be caused by OS but to be associated to maternal ageing. MtDNA variants in protein and rRNA coding regions are associated with a lower birth weight and are more frequently observed in ART children. We propose that these non-disease associated variants can result in a suboptimal mitochondrial function that impacts birth weight. Future research will establish the long-term health consequences of these changes and how these findings will impact the clinical practice and patient counselling in the future.

BackgroundChildren born using assisted reproductive technologies (ART) have an increased risk of a lower birth weight, the cause of which remains unclear. As a causative factor, we hypothesized that variants in the mitochondrial DNA (mtDNA) that are not associated with disease, may explain changes in birth weight.MethodsWe deep-sequenced the mtDNA of 451 ART and spontaneously conceived (SC) individuals, 157 mother-child pairs and 113 individual oocytes from either natural menstrual cycles or cycles with ovarian stimulation (OS). The mtDNA genotypes were compared across groups and logistic regression and discriminant analysis were used to study the impact of the different factors on birth weight percentile.ResultsART individuals more frequently carried variants with higher heteroplasmic loads in protein and rRNA-coding regions. These differences in the mitochondrial genome were also predictive of the risk of a lower birth weight percentile, irrespective of the mode of conception but with a sex-dependent culture medium effect. The higher incidence of these variants in ART individuals results both from maternal transmission and de novo mutagenesis, which we found not to be caused by OS but to be associated to maternal ageing.ConclusionsMtDNA variants in protein and rRNA coding regions are associated with a lower birth weight and are more frequently observed in ART children. We propose that these non-disease associated variants can result in a suboptimal mitochondrial function that impacts birth weight. Future research will establish the long-term health consequences of these changes and how these findings will impact the clinical practice and patient counselling in the future.

Humans present remarkable mitochondrial DNA (mtDNA) variant mosaicism, not only across tissues but even across individual cells within one person. The timing of the first appearance of this mosaicism has not yet been established. In this study, we hypothesized it occurs during preimplantation development. To investigate this, we deep-sequenced the mtDNA of 254 oocytes from 85 donors, 158 single blastomeres of 25 day-3 embryos, 17 inner cell mass and trophectoderm samples of 7 day-5 blastocysts, 142 bulk DNA and 68 single cells of different adult tissues. We found that day-3 preimplantation embryos already present blastomeres that carry variants unique to that cell, showing that the first events of mtDNA mosaicism happen very early in human development. We classified the mtDNA variants based on their recurrence or uniqueness across sibling oocytes and embryos, and between single cells and samples from the same embryos or adult individuals. Variants that recurred across samples had higher heteroplasmic loads and more frequently resulted in synonymous changes or were located in non-coding regions than variants that were unique to one oocyte or single embryonic cell. These differences were maintained through developmental stages, suggesting that the mtDNA mosaicism arising in preimplantation development is maintained into adulthood. Further, the results support a model in which close clustering of mitochondria carrying specific mtDNA variants in the ooplasm leads to asymmetric distribution of these mitochondria throughout the cell divisions of the preimplantation embryo, resulting in the appearance of the first form of mtDNA mosaicism in human development.

Study question Can mitochondrial DNA (mtDNA) variants explain the differences in birthweight between ART and spontaneously conceived (SC) individuals and how do they originate? Summary answer Children born after ART carry more frequently a different mtDNA variant composition, both maternally inherited and de novo, which are predictive of their birthweight percentile. What is known already Children born after ART show an increased risk of lower birthweight and of developing a mild abnormal cardio-metabolic profile later in life. Variation in the mtDNA associates with overall health in the general population, including cardio-metabolic fitness, and can result in changes in mitochondrial function. We hypothesized that mitochondrial DNA variants could explain the differences in birthweight between ART and SC individuals and that these differences may result from maternal transmission and/or from the ovarian stimulation (OS) used in ART. Study design, size, duration We deep-sequenced the mtDNA of 472 individuals of who 283 ART and 189 SC, 182 mother-child pairs and 113 single oocytes from both natural menstrual cycles and OS cycles. The mtDNA was compared between groups and Fisher linear discriminant analysis was used as predictive model for the birthweight percentile. Participants/materials, setting, methods Mitochondrial DNA was enriched by long-range PCR and subsequently sequenced on an Illumina platform. mtDNA server and MuTect were used for variant calling for variants with a load higher than 1.5%, versus the reference NC_012920.1. An orthogonally rotated factor analysis was used to reduce the dimensionality of the studied dependent variables in the complex data of the heteroplasmic variants. Main results and the role of chance ART individuals carried more frequently haplogroup U4 (p = 0.004) and component analysis indicated that they carry a different mtDNA heteroplasmic variant composition than SC individuals (p = 0.01), driven by non-synonymous protein-coding and rRNA-coding variants. These differences were also predictive of the risk of a lower birthweight percentile, especially for the SC children, together with the absence of haplogroup T, the presence of homoplasmic tRNA-variants, pregnancy-induced hypertension and the embryo culture medium used. The differences in heteroplasmic variation observed in the ART children resulted from both maternal transmission (p = 0.03) and de novo mutagenesis (p = 0.02). Mothers of ART children showed a similar mtDNA genotype as their children and differed in the same variant composition when compared to the mothers of SC children (p = 0.03). Furthermore, the comparison of oocytes from the same donors retrieved in natural menstrual cycles and after one OS cycle showed that OS does not increase de novo mutagenesis. Additionally, clinical parameters such as the total dosage of FSH units, the number of oocytes retrieved, and maternal age did not show any correlation with the differences observed in ART individuals. Limitations, reasons for caution This study is observational with no functional tests being performed. Wider implications of the findings We demonstrate an association between a lower birthweight percentile and a mtDNA variant composition which is more frequently carried by ART children. These non-disease associated mtDNA variants could cause a suboptimal mitochondrial function affecting the birthweight. Long-term health consequences of these differences remain to be further elucidated. Trial registration number Not applicable

STUDY QUESTION Does the presence of single nucleotide polymorphisms (SNPs) in the FSH receptor gene (FSHR) and/or FSH beta subunit-encoding gene (FSHB) influence ovarian response in predicted normal responders treated with rFSH? SUMMARY ANSWER The presence of FSHR SNPs (rs6165, rs6166, rs1394205) has a statistically significant impact in ovarian response, although this effect is of minimal clinical relevance in predicted normal responders treated with a fixed dose of 150 IU rFSH. WHAT IS KNOWN ALREADY Ovarian reserve markers have been a breakthrough in response prediction following ovarian stimulation. However, a significant percentage of patients show a disproportionate lower ovarian response, as compared with their actual ovarian reserve. Studies on pharmacogenetics have demonstrated a relationship between FSHR or FSHB genotyping and drug response, suggesting a potential effect of individual genetic variability on ovarian stimulation. However, evidence from these studies is inconsistent, due to the inclusion of patients with variable ovarian reserve, use of different starting gonadotropin doses, and allowance for dose adjustments during treatment. This highlights the necessity of a well-controlled prospective study in a homogenous population treated with the same fixed protocol. STUDY DESIGN, SIZE, DURATION We conducted a multicenter multinational prospective study, including 368 patients from Vietnam, Belgium, and Spain (168 from Europe and 200 from Asia), from November 2016 until June 2019. All patients underwent ovarian stimulation followed by oocyte retrieval in an antagonist protocol with a fixed daily dose of 150 IU rFSH until triggering. Blood sampling and DNA extraction was performed prior to oocyte retrieval, followed by genotyping of four SNPs from FSHR (rs6165, rs6166, rs1394205) and FSHB (rs10835638). PARTICIPANTS/MATERIALS, SETTING, METHODS Eligible were predicted normal responder women MAIN RESULTS AND THE ROLE OF CHANCE The prevalence of derived allele homozygous SNPs in the FSHR was rs6166 (genotype G/G) 15.8%, rs6165 (genotype G/G) 34.8%, and rs1394205 (genotype A/A) 14.1%, with significant differences between Caucasian and Asian women (P T) was very rare (0.5%). Genetic model analysis revealed that the presence of the G allele in FSHR variant rs6166 resulted in less oocytes retrieved when compared to the AA genotype (13.54 ± 0.46 vs 14.81 ± 0.61, estimated mean difference (EMD) −1.47 (95% CI −2.82 to −0.11)). In FSHR variant rs1394205, a significantly lower number of oocytes was retrieved in patients with an A allele when compared to G/G (13.33 ± 0.41 vs 15.06 ± 0.68, EMD −1.69 (95% CI −3.06 to −0.31)). A significantly higher prevalence of hypo-responders was found in patients with the genotype A/G for FSHR variant rs6166 (55.9%, n = 57) when compared to A/A (28.4%, n = 29), ORadj 1.87 (95% CI 1.08–3.24). No significant differences were found regarding the FORT across the genotypes for FSHR variants rs6166, rs6165, or rs1394205. Regarding the FOI, the presence of the G allele for FSHR variant rs6166 resulted in a lower FOI when compared to the A/A genotype, EMD −13.47 (95% CI −22.69 to −4.24). Regarding FSHR variant rs6165, a lower FOI was reported for genotype A/G (79.75 ± 3.35) when compared to genotype A/A (92.08 ± 6.23), EMD −13.81 (95% CI −25.41 to −2.21). LIMITATIONS, REASONS FOR CAUTION The study was performed in relatively young women with normal ovarian reserve to eliminate biases related to age-related fertility decline; thus, caution is needed when extrapolating results to older populations. In addition, no analysis was performed for FSHB variant rs10835638 due to the very low prevalence of the genotype T/T (n = 2). WIDER IMPLICATIONS OF THE FINDINGS Based on our results, genotyping FSHR SNPs rs6165, rs6166, rs1394205, and FSHB SNP rs10835638 prior to initiating an ovarian stimulation with rFSH in predicted normal responders should not be recommended, taking into account the minimal clinical impact of such information in this population. Future research may focus on other populations and other genes related to folliculogenesis or steroidogenesis. STUDY FUNDING/COMPETING INTEREST(S) This study was supported by an unrestricted grant by Merck Sharp & Dohme (MSD). N.P.P. reports grants and/or personal fees from MSD, Merck Serono, Roche Diagnostics, Ferring International, Besins Healthcare, Gedeon Richter, Theramex, and Institut Biochimique SA (IBSA). N.L.V. and M.T.H. report consultancy and conference fees from Merck, Ferring, and MSD, outside the submitted work. P.D. has received honoraria for lecturing and/or research grants from MSD, Ferring International, and Merck. D.S. reports grants and/or personal fees from MSD, Ferring International, Merck Serono, Cook, and Gedeon Richter. A.R.N., B.A.M., C.S., J.M., L.H.L., P.Q.M.M., H.T., and S.G. report no conflict of interests. TRIAL REGISTRATION NUMBER NCT03007043

Summary In this study, we deep-sequenced the mtDNA of human embryonic and induced pluripotent stem cells (hESCs and hiPSCs) and their source cells and found that the majority of variants pre-existed in the cells used to establish the lines. Early-passage hESCs carried few and low-load heteroplasmic variants, similar to those identified in oocytes and inner cell masses. The number and heteroplasmic loads of these variants increased with prolonged cell culture. The study of 120 individual cells of early- and late-passage hESCs revealed a significant diversity in mtDNA heteroplasmic variants at the single-cell level and that the variants that increase during time in culture are always passenger to the appearance of chromosomal abnormalities. We found that early-passage hiPSCs carry much higher loads of mtDNA variants than hESCs, which single-fibroblast sequencing proved pre-existed in the source cells. Finally, we show that these variants are stably transmitted during short-term differentiation., Graphical Abstract, Highlights • Early-passage hESCs carry mainly mtDNA variants inherited from the embryos • Clonal culture takeover results in increased variant load and pathogenic potential • hESCs and somatic cell cultures are mosaic in terms of mtDNA genotypes • hiPSCs carry high levels of variants due to clonal reprogramming of a mosaic source, In this work, Spits and colleagues present an in-depth study of mtDNA variants in hPSCs. Early-passage hESCs carry few low-frequency variants, while higher variant loads are observed in late-passage hESCs and hiPSCs. The presence of variants is determined by the somatic or embryonic origin of the line and modulated by random drift, spontaneous mutagenesis, and clonal culture takeover.

Massive parallel sequencing (MPS) can accurately quantify mitochondrial DNA (mtDNA) single nucleotide variants (SNVs), but no MPS methods are currently validated to simultaneously and accurately establish the breakpoints and frequency of large deletions at low heteroplasmic loads. Here we present the thorough validation of an MPS protocol to quantify the load of very low frequency, large mtDNA deletions in bulk DNA and single cells, along with SNV calling by standard methods. We used a set of well-characterized DNA samples, DNA mixes and single cells to thoroughly control the study. We developed a custom script for the detection of mtDNA rearrangements that proved to be more accurate in detecting and quantifying deletions than pre-existing tools. We also show that PCR conditions and primersets must be carefully chosen to avoid biases in the retrieved variants and an increase in background noise, and established a lower detection limit of 0.5% heteroplasmic load for large deletions, and 1.5 and 2% for SNVs, for bulk DNA and single cells, respectively. Finally, the analysis of different single cells provided novel insights into mtDNA cellular mosaicism.European Journal of Human Genetics advance online publication, 23 August 2017; doi:10.1038/ejhg.2017.129.