Re: Ancient DNA, facts and assumptions

From
Date
Cc
Reply-To
Tom Loy presents good ideas here for advancing the aDNA field.  In fact,
many of these are already in use in closely related fields.  For example,
in the collection of deep sea cores, an archival half is kept so some will
remain available, and the working half is seldom given away entirely.  This
poses some problem for museums, but as a museum representative, I would
endorse a strong effort to maintain "archive and working pieces" of fossil
materials used in DNA analysis.  These should be kept  in appropriate
storage facilities that ensure the preservation of the DNA, if any.   It is
one of the advances that paleo museums must be willing to accept.  Like
type specimens, DNA specimens should be deposited before or at publication,
with museum identifiers, and be available to any legitimate scientific
borrower after the original publication.  The Museum of Paleontology here
will do that.  Editors of journals might insist or at least suggest that
this proceedure be adopted to authors submitting mss on ancient DNA.

>The recent few contributions are heartening. It seems that there may be a
>shift from "I've got the oldest.../ most novel....aDNA" approach to the
>analysis of ancient DNA to problems that have also bugged me for some time.
>Shortly after I saw Woodward's paper I offered(via email) the services of
>my lab to duplicate his findings and he replied that the entire bone had
>been destroyed during the process of analysis and replications. As an
>archaeologist and analyst of aDNA this is a troublesome situation. One of
>my basic tenets is never to destroy the entire unit under study, but to
>save some parts for either independant analysis, or for some future
>analysis using more advanced techniques. I have complained when 'residue
>analysts' (by in large using immunological methods) use the entire residue
>from a tool surface in attempts to determine species of origin using
>methods that are less than definitive, and generally insensitive (World
>Archaeology V25 #1 1994).
>
>In a communication to the aDNA list more recently I advanced the view that
>at this juncture in the development of the (sub?) discipline should adopt
>some guidelines that may better offer, if not proof, at least some higher
>level of certainty about our results
>
>>... about proof of authenticity: that there must be some way other than the
>>Woodward approach >(used by lots of others as well) to the effect that "it
>>must be real 'cause I don't find the sequence on >the database anywhere" There
>>are so few organisms for which sequences are known, and such a >huge place to
>>explore at the moment, that it seems to me absolutely imperative to adopt
>>procedures >that incorporate standard scientific logic: proof by weight of
>>evidence, and proof by >non-falsifiability. To do so means that  either  (1)
>>the target sequence is really well known (eg >D-loop stuff FOR HUMANS; or (2)
>>that the sequences we examine to "prove the authenticity " of >our ancient DNA
>>and (for my research the validity of identifications between control bone DNA
>>and >blood residues on tools) must have limiting or unique features that are
>>diagnostic at some level.
>
>As I mentioned then, the primers we have been using (Naito's A &B primers
>for the v3 region of 28s ribosomal RNA gene will not amplify from lower
>eukaryot nor plant organisms, eliminating a source of great uncertainty
>when dealing with tool residues or bones that have been exposed to fungal,
>bacterial and algal contamination. Now, having examined about 40 sequences
>including eutherian, metaeutherian (including 2 species of extinct
>Diprotodon species from bone ca 100,000 years old), monotreme, amphibian,
>bird, fish and crustacean sources using this region, it is apparent that
>within the v3 region there are "signature" sequences that enable both
>species-level identity, but there is additional sequence variation that
>reflects higher orders of taxonomic affiliation within the amplified
>region. This approach gives a simple  taxonomic double-check that the match
>of the sequence with a control species, whether ancient or modern, "fits"
>the profile at higher taxonomic levels; it may be an approach that will not
>work in many regions of the genome, but certainly gives added weight of
>evidence in our cases.
>
>Finally, much of the variation in whether or not people can extract aDNA
>from a variety of sources, (and this consideration has raised the question
>of its survival at all) I believe stems from the plethora of extraction
>methods currently in use in various laboratories for different sources of
>aDNA, some capable of extraction and purification of extremely small
>amounts of aDNA, others not. Eventually these methods need to be subjected
>to interlaboratory trials, using the same starting material and the same
>target region. This is made difficult by the usually small sample size and
>uniqueness of the sample, or in the case of human bone the politically and
>ethically sensitive nature of repeated destructive analysis. There are
>hower a few large animals of not too great antiquity (eg. woolly mammoths)
>that have served similar purposes in  radiocarbon dating laboratories (see
>Stafford et al. "Radiocarbon" 1987, 29(1). The analogy with radiocarbon
>dating analysis is apt; it is in the best interest of radiocarbon
>laboratories to conduct regular checks on their accuracy  sensitivity and
>contamination, both internally and inter-laboratory using standard and well
>characterized sources of radiocarbon.
>It may well be that as the analysis of aDNA proceeds that some standard
>ancient source of DNA must be adopted as the analogue of the "ANU sucrose
>standard" in radiocarbon dating. This does not preclude different
>laboratories using different methods, and targets of analysis, but would at
>least provide a benchmark for detection of contamination, sensitivity,
>specificity, and uniformity that lie at the heart of any methodology.
>
>In addition, a few ground rules should be applied to our research: (1) we
>should agree that no sample should be completely destroyed by our analysis,
>but a portion kept for future (and possibly, independant) analysis and that
>samples be kept as small as possible in order to acheive this aim; (2) that
>the source of original reference aDNA should be clearly taxonomically
>identified, it is not good enough to assert as Woodward did that because
>the coal seam contained dinosaur bone that the bone he analyzed was
>therefore dinosaurian; (3) that we should publish and openly discuss the
>problems of 'contamination' we encounter, and not always assume that the
>contaminant is 'in house';  (4) that the concept of "proof by weight of
>evidence" must not rest solely upon the number of replicates that were in
>agreement with each other, or, that a sequence must be "xxxx" because it is
>not on the international data base;   Finally (5) that some large beast
>like a single woolly mammoth skeleton be designated as a standard against
>which to test our own and other's methods and constitute an interlaboratory
>standard. Following from Ken Reed's proposal, choosing an extinct animal
>which is most unlikely to have shed DNA around the lab, or be in commercial
>preparations is highly desireable as a control species.  Most analysis it
>seems deals with bone as the primary source, although for tissue there are
>also large animals with copious naturally freeze-dried tissue (bison, horse
>etc from Alaska) that can serve as a source of aDNA for those who work with
>tissue; &ct.
>
>Thank you Jery for the apt quote from Darwin, it is now on the wall of my
>student lab.
>
> Martin Evison's comments prompt me to suggest that perhaps (speaking for
>myself as one for whom just getting to the conference is a major-- and this
>year, insurmountable-- cost) might  the organizers of aDNA III arrange for
>electronic transmission of presented papers via the aDNA list? It seems a
>shame that information sharing is limited to the (always) small number of
>people who, by way of proximity or access to travel funds are able to
>actually learn what is taking place elsewhere by being there in person ?
>The aDNA list is a grapevine, not a publication venue so, take a chance and
>say something, pass on papers and other information resulting from the aDNA
>III.
>
>Any response to this?
>cheers
>t.loy
>
>
>
>
>-------------------------------------------
>Thomas H. Loy, Principal Research Fellow
>Department of Anthropology and Sociology
>                                AND
>Centre for Molecular and Cellular Biology
>University of Queensland,  Brisbane, 4702
>Email: t.loy@mailbox.uq.oz.au
>CMCB=Phone (+61) 7 365-4483 // Fax:(+61) 7 365-4388