Biochtmica et Biophysics Acta, 1000 (1989) 197-199
Elsevier

Commentary by

Christian B. Anfinsen

Department of Biology, Johns Hopkins University, Baltimore, MD (U.S.A.)

     on `Studies on the structural basis of ribonuclease activity'
by C.B. Anfinsen, W.F. Harrington, Aa. Hvidt, K. Linderstrum-Lang, M. Ottesen
                   and J. Schellman

Biochim. Biophys. Acta 17 (1955) 141-142

The short communication selected for inclusion in   involving six of us working at the Carlsberg Laboratory
the 1000th volume of Biochimica et Biophysics Acta   in 1955 under the guidance of Professor K.
was the result of experimental efforts and discussions   Linderstrram-Lang. This communication is a beautiful

11971


example of how an entirely acceptable conclusion can
be reached that is entirely wrong because of the paucity
of knowledge at that particular time. Indeed, I spent the
following 15 years or so completely disproving the
conclusions reached in this communication, namely,
that an ordered secondary structure in a protein is quite
unnecessary for its properties as a catalyst. These were
the days when, except for Sanger's work on insulin, no
sequence information of any significance was available
for protein molecules, and three-dimensional structure
derived from crystallographic work was still a product
of the future.

In the 3 or 4 years that followed, I was fortunate to
be joined in my laboratory at the National Institutes of
Health by Michael Sela, William Harrington, Fred
White, and a number of others. In a relatively short
time we discovered that the ribonuclease molecule, which
is indeed highly disoriented in strong urea solutions, is
held together and stabilized in its native conformation
by its substrate and, indeed, by a number of other
polyvalent cations such as polymetaphosphate,
poly(aspartic acid), and even orthophosphate itself. We
have frequently referred to this kind of stabilization of
structure in denaturing solvents as `rigidification', and
it seems to be a common phenomenon with many
enzymes whose activities are preserved by substrate
molecules or substrate analogs. When I now look at the
original BBA communication, I find myself thinking in
terms of the current witch-hunting that is now so popu-
lar in which published material based on inadequate or
incomplete data is occassionally referred to as "fraud".
In think this term is frequently probably correctly ap-
plied, although I do believe that in many instances such
premature incorrect conclusions may simply reflect the
fact that the advance of science and the deeper under-
standing of nature are under continual modification.
Ongoing refinement of data frequently requires consid-
erable reinterpretation of formerly held ` truths'.

As the result of a number of studies with many
colleagues on the refolding of denatured molecules,
accumulated even after conversion to extended poly-
peptide chains by disulfide bond reduction, it became
clear that a generality could be proposed that seemed to
be applicable to essentially all proteins that were ex-
amined. This generality stated that the details of the
three-dimensional structure of a protein molecule were
determined entirely by the amino-acid sequence of the
molecule and that no other outside information was
required. The air oxidation of a reduced protein fre-
quently led to `scrambled' molecules with incorrect
pairing of half-cystine residues, particularly when the
concentration was too high during the reoxidation. In
dilute solutions, however, such mispairing, or inter-
molecular bonding, was generally avoided. I remember,
as one of my more exciting moments, an experiment
which Dr. Edgar Haber and I carried out on ribonuclease

refolding from the reduced form. After stirring in air
overnight, almost no activity had been regenerated, but
the addition of a small amount of mercaptoethanol led
to a rapid reshuffling of the scrambled SS bridges with
full recovery of activity in a relatively short time. The
path of our research on formation of three-dimensional
structure thereafter was fairly easy sailing.

In recent years, with the acceptance of the generality
of the process of spontaneous refolding based on se-
quence alone, a very large effort has been progressing in
many laboratories to elucidate the nature of the inter-
acting forces that lead to the correct structure and to
deduce the three-dimensional structure of proteins from
the sequence alone, employing a large number of differ-
ent thermodynamic and stereochemical parameters. One
of these days, our thermodynamic and computer experts
will solve this problem of prediction, and the so-called
`folding problem' can be put on the shelf along with
other solutions to Nature's secrets.

Correspondence: C.B. Anfinsen, Department of Biology, The Johns
Hopkins Uniuersity, 34th & Charles Streets, Baltimore, MD 21218,
U.S.A.

11981