Background Mitochondrial DNA (mtDNA) disorders have a high clinical variability, mainly explained by variation of the mutant load across tissues. The high recurrence risk of these serious diseases commonly results in requests from at-risk couples for prenatal diagnosis (PND), based on determination of the mutant load on a chorionic villous sample (CVS). Such procedures are hampered by the lack of data regarding mtDNA segregation in the placenta.

The objectives of this report were to determine whether mutant loads (1) are homogeneously distributed across the whole placentas, (2) correlate with those in amniocytes and cord blood cells and (3) correlate with the mtDNA copy number.

Methods We collected 11 whole placentas carrying various mtDNA mutations (m.3243A>G, m.8344A>G, m.8993T>G, m.9185T>C and m.10197G>A) and, when possible, corresponding amniotic fluid samples (AFSs) and cord blood samples. We measured mutant loads in multiple samples from each placenta (n= 6–37), amniocytes and cord blood cells, as well as total mtDNA content in placenta samples.

Results Load distribution was homogeneous at the sample level when average mutant load was low (<20%) or high (>80%) at the whole placenta level. By contrast, a marked heterogeneity was observed (up to 43%) in the intermediate range (20%–80%), the closer it was to 40%–50% the mutant load, the wider the distribution. Mutant loads were found to be similar in amniocytes and cord blood cells, at variance with placenta samples. mtDNA content correlated to mutant load in m.3243A>G placentas only.

Conclusion These data indicate that (1) mutant load determined from CVS has to be interpreted with caution for PND of some mtDNA disorders and should be associated with/substituted by a mutant load measurement on amniocytes; (2) the m.3243A>G mutation behaves differently from other mtDNA mutations with respect to the impact on mtDNA copy number, as previously shown in human preimplantation embryogenesis.