Woodward et al.

[rowe@lepomis.psych.upenn.edu (Mickey Rowe)]

As promised, I'm writing about the recent claims of the dinosaur DNA
find from BYU.  First off, the two references are:

Gibbons, A. (1994). "Possible Dino DNA Find Is Greeted With
     Skepticism", _Science_, 266:1159.

and:

Woodward, S. R., Weyand, N. J. and Bunnell, M. (1994). "DNA Sequence
     from Cretaceous Period Bone Fragments", _Science_, 266:1229-1232.

To clear up a bit of past confusion--no real attempt was made at
determining from what sort of cells the DNA might have been extracted.
Fragments from the interior of the bones were ground up for the
extraction experiments.  Other bone samples were sectioned and stained
for viewing under a light microscope.  These sections indicated the
possible presence of osteocytes (mature bone cells), which were
strongly stained by eosin and hematoxylin.  I believe that those
stains indicate the presence of nuclear DNA.  In any case, for the
current report, Woodward et al. cast their net specifically for a
section of mitochondrial DNA, more specifically DNA which codes for
cytochrome b (I suspect that they searched for other sequences as well
and reported only on the one in which they achieved a success, but I'm
basing that more on a comment in the Gibbons article than on anything
in the Woodward article).

In passing, they mention that the bone appeared to be woven rather
than lamellar.  If I recall correctly, woven bone is an indicator of
rapid growth, like that typically seen (amongst modern animals) only
in birds and mammals.

As for the bones: bone 1 is thought to be a fragment from a long bone.
It's dimensions are 20 cm x 6 cm x 3 cm.  It's not possible to
determine what bone it is let alone what sort of animal it might have
supported.  Bone 2 appears to have been from a rib, and not
necessarily from the same animal as bone 1.  It was 20 cm x 2 cm x 1
cm.  Analysis of the constituents of the bones indicate that very
little mineralization took place--the composition is similar to that
of bones extracted from recently living animals.

Now for the sequences: 9 sequences were obtained with the help of PCR,
7 from bone 1, and 2 from bone 2.  A consensus sequence was generated
from the seven sequences from bone 1.  (That is, each of the seven
sequences was given "one vote" on what each base should be at each
location along the strand, and the base indicated by the majority of
the seven strands was assigned to the corresponding position along the
consensus strand).  All 9 of the strands were then compared to the
consensus strand, and all 10 strands (9 experimentally determined and
the one consensus) were compared with sequences reported from other
animals.

I'm not a molecular biologist, so I don't know what sort of variation
you would expect to find between strands obtained and amplified from
modern organisms (anybody else want to chime in?).  However, what was
found here was differences (arguments between the sequences if you
will) at around 9% of the sites.  I'm having a bit of difficulty
determining the length of the strands they found and would appreciate
any available help.  The text of the article claims that the sequence
found was 174 base pairs long.  In the figure showing all of the
sequences, it looks to me like the *longest* sequence is 134 base
pairs long.  In the Gibbons article it is claimed that the *shortest*
sequence was 134 base pairs long.  In any case, the sequences were all
pretty short so you can't draw strong conclusions from them,
particularly in light of the discrepancies between the individual
sequences.  As for the latter, a very plausible explanation for these
differences is that each of the samples were damaged at some locations
(since most people would have told you that all 80 million year old
DNA would be fubar, finding evidence of damage at some sites doesn't
exactly call for a suspension of disbelief :-) However, in most cases
the disagreements are between the consensus sequence and only one or
two of the actual sequences, so it's a reasonable assumption that the
the consensus sequence is the "correct" sequence for the animal that
gave us bone 1.  There are 9 sites at which the sequences from bone 2
disagree with the consensus sequence, but agree with each other.  The
authors suggest this to mean that the bones were from different
animals, and that at these 9 sites, the two sequences from bone 2 are
"correct". 

With regards comparisons between these sequences and others from
contemporary animals, the authors say very little other than that the
data don't allow for the conclusion that the sequences are more
closely related to either mammals, or birds, or reptiles (their word,
not mine--I can only assume that they include box turtle, garter
snake, and alligator in the last category since they mention those
three animals by name in the list of additional sequences obtained for
comparison in their own lab).

Naturally there's more to the paper than what I've written, but I
don't want to go on forever unless people ask for more information.
The take home message is (from the last sentence of Woodward et al.):

     ... this result demonstrates that the recovery of DNA from
     well-preserved Cretaceous period bone may be possible.

Any stronger conclusion would be unwarranted embellishment IMHO.

-- 
Mickey Rowe     (rowe@lepomis.psych.upenn.edu)