Dear Dr. Pauling:
As an old-timer of the Collagentsia-quite solo in these parts- I am inconsiderate
enough to detract the strained Nobel-laureate from the extraverts of the day to the
intraverts of collagen. A few lines reading on some recent developments in the chemistry
of collagen might then be a relaxation.
According to the findings given in the enclosed notes, the degree of the hydrothermal
stability of collagens is directly related to their content of hypro (from 70° to
35° versus hypro % from 13 to 6). Investigations of exhaustively acetylated bovine
collagens indicate 3 out of 4 hypro-OH groups to form the link - OH ••••• OC. The
OH group resisting acetyla-tion appears to be the mainstay of collagens structure
HN, generally. By O-acetylation of bovine collagen, i.e., the breaking of the hydrogen
bonds given above, its shrinkage temperature is lowered from 65 to 40°, or to the
point of teleostian collagen. There are some indications that the link responsible
for the ultimate stability of collagen is an ester link of the type -0 • CO. Thus,
Grassmann (the Aug. issue of Z. Naturforschung) cites exp. evidence suggesting the
presence of an ester link in procollagen. Collagen is insoluble in sat. solution of
LiBr, whereas silkfibroin and synthetic polypeptides, mainly H-bonded, are easily
solubilized. The alkoli binding capacity of limed collagens is about 0.5 meq. base
per g. collagen, while gelatin binds 0.8 - 0«,9 meq. The free caxboxyl groups amount
to about 1 meq. per g. protein in both instances. The presence of the masked OH groups
(0.3 – 0.4 meq.) in ester linking with carboxyl groups would satisfactorily account
for the difference.
As to the interchain nature of the OH •••• CO link, deduced from the trend of the
TS, this canlusion would be invalidated if an intrachain link of this type would suffice
for the hydrothermal stabilization of collagen. Thus, according to Huggins model in
his recent JACS-note, an intrachain hydrogen bond of the type suggested by me is assigned
a distance of 2.9 Å. If you might get a chance looking into this, it would be fine
to have a chat with you at the social tea after your Tuesday-night lecture. No writing
or acknowledging please.
Finally, a recent finding of Zahn published in a leather chemist journal, which likely
is not available in Pasadena, will undoubtedly interest you. By making a sulfone out
of Sanger's reagent, an excellent crosslinking agent for collagen is obtained. By
hydrolysis of the sulfone-treated collagen and the dinitrodiphenylsulfone-bis-lysine
was isolated. I am enclosing a reprint, by the way. Evidently, the two lysine residues
are close enough to be within the reach of the S-connected benzene rings. The may
be conceived as supporting Bear's idea of the bulky side chains being located in the
bands and the matching of the bands in fibrils. In my Cincinnati lecture, as you
perhaps remember, I stressed the fact that by crosslinking of collagen its hydrothermal
stability is improved, while the cross-linking has practically no effect of the mechanical
strength of the fibers (cohesion), explaining some findings of our mutual friend John
Highberger. This issue takes on a new aspect if the hydrothermal stabilization should
be due to intrahelical crosslinking.
These simple demonstrations may be of helpful in attempts of promulgating structures
for collagen. Your helical concept has stirrd the thinking members of my scribe.
We thank you for your impulses and your inspiration and we wish you health and solitude
for the final arrival.
With kind personal regards,
Cordially yours,
