March 18, 1953
Dr. Linus C. Pauling
California Inst. of Technology
Pasadena 4, California
Dear Dr. Pauling:
I have just finished reading your paper in the February "Proc. Nat. Acad. Sci." on
a new type structure which you have proposed for nucleic acids. I have been interested
in the biochemistry of the nucleic acids for some time and have recently constructed
an hypothetical duplicating mechanism for peptide chain and nucleic acid synthesis
which seems to be logically consistant and which makes use of reaction types that
are familiar to the biochemist. However this mechanism requires functionally available
phosphate groups as you will see if you care to glance through the enclosed reprint.
Therefore a single nucleotide chain or helix would be required as postulated by Astbury
and by Furberg, and if the three-chain helix postulated by you is correct, my mechanism
must go into the waste basket.
Now I am sure that my mechanism, which after all is only a hypothesis, cannot be entirely
correct, and of course it may be entirely incorrect; but it does suggest certain biochemical
approaches to the problem of peptide chain synthesis, and hence might be of some use
even if it is only partially correct. Therefore, before I do throw it in the waste
basket I should like to take the liberty of asking a few questions about your proposed
nucleic acid structure which you may well have answers for, but for which answers
are not apparent in your paper. These questions are as follows:
1. Do the positions of the hydrogen atoms associated with the oxygens of the phosphate
tetrahedra indicate that these hydrogens would be difficult to replace by metals such
as sodium, potassium, calcium, barium, etc.; by dye molecules (methyl green); and
by basic proteins (histones or protamines); or for that matter by ordinary proteins?
It would seem to me that in reactions involving replacement of the phosphate hydrogens
by any of the above-mentioned substances (present in solution as cations), there might
be difficulties if the cation replacing the hydrogen could not because of steric hinderance
get close to the oxygens of the phosphate group bearing the negative charges. If
such steric hinderance should be appreciably-present, then any hydrogen ions which
were present should still be attracted into positions close to the negatively charged
oxygen atoms, and it would seem to me that the net result would be to render the hydrogens
very non-dissociable from the phosphate groups. In view of these considerations,
can you make your model work properly in exchange reactions involving the hydrogens
of the phosphate groups with other cations, and would it work as well as the Astbury
or Furburg models?
Dr. Linus C. Pauling
- 2 -
March 18, 1953
2. How good is the evidence for the cross sectional area of the DNA
molecule? Astbury makes a rough calculation of cross-sectional area based
on estimates of the dimensions of a nucleotide molecule (15 X 7.5 Ǻ) = 112.5 Ǻ2.
As far as I can see this assumes a ribbon-like molecule with or without
coiling. If this value of Astbury for area is divided into 338 Ǻ3, your figure for the volume per nucleotide, obtained from the average residue molecular
weight and the density, we obtain a figure of about 3.01 which is somewhere near the
distance along the fiber length occupied by a repeating unit (3.4 Ǻ)
as calculated from the x-ray measurements. This would indicate as I under-
stand it that the observed spacing probably corresponds to the distance
between nucleotides in a single chain. On the other hand if we use your
figure for area, obtained I presume from an assumed circular cross section
with radius of about 10 Ǻ ,we get for the distance occupied by a nucleotide
residue along the fiber axis a value of about 1.12 Ǻ which is only a third
of the observed spacing (3.4 Ǻ). For this reason, I take it, you feel that there
have to be three nucleotide residues per distance of 3.4 Ǻ along the chain, and therefore
construct a model with three intertwined chains. Hence the crux of the matter seems
to be to get an accurate estimate of the cross sectional area, as free as possible from assumptions, and that is why I am asking question number
3. Regardless of the mechanism proposed for nucleic acid synthesis,
it seems likely that enzymes must be involved. Although this question
obviously cannot be answered with any certainty, I wonder whether difficulties
will not arise in allowing access of synthetic enzymes to a close packed
structure of three interwoven helices such as you propose. The same question
might be asked in regard to enzymes of depolymerization which cause rupture
of inter-nucleotide linkages and hence must presumably form complexes with groups
very close to the phosphate groups. Incidentally, would your structure predict the
products of DNA-ase action as well as a single chain or helix?
4. How much change is likely to occur in structure on liberating nucleic acids from
the proteins to which they are bound, presumably through purines and pyrimidines,
in the natural state?
In closing, I might state that I am not trained in advanced physical chemistry or
physics, and might therefore not appreciate all of the basis for your model, but I
would enjoy hearing from you regarding the questions even if only very briefly.
I would also appreciate having any comments which you might have regarding my hypothetical
duplicating mechanism, if you feel that this is worthy of comment.
Alexander L. Dounce
P.S. I should appreciate a reprint of your article (referred to in the letter),
and any reprints which you might be able to spare on peptide chain structure
and hemoglobin chemistry. A.L.D.