Ancient DNA Methodology: Thoughts from Brian M. Kemp and David Glenn Smith on "Mitochondrial DNA of Protohistoric Remains of an Arikara Population from South Dakota"

Our paper "Mitochondrial DNA of Protohistoric Remains of an Arikara Population from South Dakota: Implications for the Macro-Siouan Language Hypothesis" [Lawrence et al. 2010 (this issue)] has been reviewed by two scholars, and although we agree with the comments of the second reviewer, we object to the first reviewer's statement that

the Authors do not use currently accepted methods in ancient DNA studies. the Authors have not followed the now classical stringent standards for the authentications of ancient DNA. Unfortunately, Authors did not do [the following]: (i) test to quantify the preservations of macromolecules (amino acid or collagen), (ii) quantifications of macromolecules, (iii) detections of long amplificates, (iv) amplifications of associated remains (if any), (v) cloning, (vi) independent replications. Further, in this study there is another fundamental question: the impossibility to track down modern DNA human contaminations (see Sampietro et al. 2006) so the results, even though interesting, are not supported by a strong scientific methodology.

We believe that the review was unfair and reflects viewpoints that are potentially stifling to future discoveries and progression of the ancient DNA (aDNA) field as a whole.

Reviewer 1 refers to six "classical stringent standards" to which we did not adhere. Our rebuttal is not to argue that, in fact, we did adhere to any fixed set of "standards" but rather to argue that (1) following these standards cannot alone authenticate aDNA results. Conversely, (2) failing to follow all these standards does not necessarily lead to aberrant and/or unreliable results, and (3) site-specific and/or sample-specific circumstances are suitable for dictating the analytical standards to be followed.

We recognize that the "classical stringent standards" to which the reviewer refers originated from an opinion piece published in Science: Cooper and Poinar's (2000) "Ancient DNA: Do It Right or Not at All." [This was not the first or only list of recommendations; however, according to Google Scholar, this opinion piece has been cited 305 times, making it a classic for the aDNA field (website accessed December 22, 2009). A few years later, Pääbo et al. (2004) expanded on these recommendations and explored them in more detail. Pääbo's review paper was co-authored by Hendrik Poinar, but did not include Alan Cooper. In this letter, we will contrast, where appropriate, the recommendations of Cooper and Poinar (2000) with those put forth by Pääbo et al. (2004).]

The intention of Cooper and Poinar's (2000) paper was to provide some guidelines for the field of aDNA, which had taken a major detour in the early 1990s into what now appears to be a realm of impossibility. During this period, for example, attempts were made to extract and analyze DNA from dinosaurs (Woodward et al. 1994), Miocene chloroplasts (Golenberg et al. 1990; Soltis et al. 1992), and 25-135-million-year-old amber-embedded insects and their gut flora (Cano et al. 1992a, 1992b, 1993, 1994), reports of which were published in leading scientific journals. (Hendrik Poinar was part of the research team that reported on the DNA retrieved from the amber-embedded insects specimens.) Upon reanalysis, the dinosaur DNA was found to closely resemble that of mammals, especially humans (Allard et al. 1995; Hedges and Schweitzer 1995; Henikoff 1995; Zischler et al. 1995) and the chloroplast and insect DNA sequences, collected and analyzed in many cases under substandard conditions, failed to meet evolutionary expectations (Hebsgaard et al. 2005) and/or were not reproducible (Austin et al. 1997). Although it was clearly an opportune time to make some recommendations to ground those practicing in the aDNA field, we object to anyone dictating any methodology that needs to be followed in all circumstances. We dispute that the foremost experts in the study of aDNA, a group in which several of us, in addition to Cooper and Poinar and presumably Reviewer 1, also hold membership, regard these "classical standards" as unequivocally valid requirements or broadly accept them as standards. Although stringent standards are always desirable...