King Tutankhamun’s Family and Demise

Abstract

To the Editor: In their study, Dr Hawass and colleagues reported ancient DNA data from 11 royal Egyptian mummies and used microsatellites to ascertain kinship among specimens. We question the reliability of the genetic data presented in this study and therefore the validity of the authors’ conclusions. Furthermore, we urge a more critical assessment of the ancient DNA data in the context of DNA degradation and contamination. The long-term survival of DNA is determined by the environmental history of the samples, and Gilbert et al argued in reference to mitochondrial DNA that “in most, if not all, ancient Egyptian remains, [ancient DNA] does not survive to a level that is currently retrievable.” The age of the mummies (more than 3300 years before the present) coupled with their preservation history suggests that DNA survival is highly unlikely. Longterm survival of nuclear DNA sequences, as accessed by Hawass et al, is even less likely than mitochondrial DNA, given lower copy numbers per cell. Success in the retrieval of putative nuclear DNA sequences is also surprising given the use of traditional polymerase chain reaction techniques rather than newly developed capture approaches coupled with second-generation sequencing that allow for successful capture of degraded (shorter) DNA sequences. Contamination is a major obstacle in human ancient DNA research. Although laboratory members involved in the study were genotyped, no persons handling the specimens prior to the study were included, raising a question of the reliability of the microsatellite profiles. Precautions such as genotyping of associated nonhuman remains and including information on the microsatellite allele frequencies in presentday Egypt would have clarified the issue of modern contamination. Another cause for concern is the lack of reported quality control measures in the genotyping of microsatellites. Potential genotyping errors include allelic stutters, allelic dropout, short allele dominance, and null alleles, all of which can result in the incorrect identification of alleles. Even small error rates (0.01 per allele) can lead to high error rates in downstream applications, such as false paternity exclusion in kinship testing.