chapter 6

THE BIOLOGICAL ACTIVITY
OF PROTEINS
IN RELATION TO STRUCTURE ~

W. may almost define the life sciences as those
concerned with the elucidation of the mechanisms by which molecules
exert their specific actions on living cells. In the case of many simple
inorganic ions and organic molecules it has been possible to arrive
at an approximate understanding of their mechanism of action. We
have some understanding, for example, of the physiological sequelae
of raising or lowering the tonicity of body fluids by the administration
or withdrawal of sodium chloride. Similarly, the aberrations in the
synaptic transmission of nerve impulses following the administration
of physostigmine may be partially explained by the action of this
drug on the enzyme acetylcholinesterase. It is a tribute to cellular
complexity that even such well-studied systems as these continue to
be frontiers of research and speculation among those individuals who
are fully conversant with them.

Protein chemists naturally feel that the most likely approach to the

126

understanding of cellular behavior lies in the study of structure and
function of protein molecules. This is, perhaps, not an entirely un-
reasonable point of view. Except for those rare phenomena in biol-
ogy which are purely physical, the “aliveness” of cells is basically the
summation of enzymic catalysis and its regulation.

The field of protein chemistry has now reached a stage of relative
sophistication allowing us to think of proteins as organic chemicals
rather than as conglomerates of amino acids. In spite of their
enormous complexity we now can measure the extent of such phe-
nomena as “denaturation” in terms of rather well-defined changes in
specific types of chemical bonds, some of which have been discussed
in Chapter 5. Because of this happy situation, we may approach,
in a rational way, the relation of specific aspects of covalent and non-
covalent protein chemistry to biological activity. Proteins, as mol-
ecules, must be thought of as consisting of one or more polypeptide
chains, cross-linked and coiled through a variety of chemical bonds
of varying strengths. Modifications of any of these lead to an entity
which is not identical to the original native molecule and which,
in a purist sense, may be considered “denatured.” From the stand-
point of function, however, we may take a somewhat narrower view.
The nativeness of an enzyme, as estimated by its ability to catalyze a
reaction, need not involve the whole of its structure.

Studies of the effect of specific degradation on biologically active
proteins are fairly recent. It was observed, however, more than
twenty years ago, that various reactive groupings on proteins could
be modified by substitution, or conversion to other forms, without loss
of function. Perhaps the best-known example of such work is the
series of investigations by Herriott and Northrop of the activity of
pepsin during progressive acetylation. Pepsin was treated with
ketene which converted free amino groups and tyrosine hydroxy]
groups to their acetyl derivatives. By this method Herriott was able
to prepare a crystalline derivative which contained seven acetyl
groups per mole of pepsin and had approximately 60 per cent of the
enzyme activity of the original pepsin. He demonstrated that the
ultraviolet absorption spectrum of this 60 per cent active material
was modified to the extent to be expected for the masking of three
tyrosine hydroxyl groups. Further, by cautious hydrolysis at pH
0.0 or at pH 10.0, three acetyl groups could be removed with a simul-
taneous increase in enzyme activity. These, and other, studies indi-
cated that tyrosine residues were somehow involved in activity and
also that acetylation of a number of the free amino groups in the pro-
tein had no effect on function.

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 127
Experiments of this sort have now become relatively commonplace,
and there is no doubt that the structures of a large number of en-
zymes and hormones may be tampered with without inactivation.
In spite of such information, the concept persisted, until very re-
cently, that the structure of biologically active proteins was more or
less “inviolable” and that the complete and integrated three-dimen-
sional architecture of these proteins was required for function. This
concept was nurtured by several theoretical considerations in which
the protein molecule was pictured as a network of resonating path-
ways through which electrons might surge and redistribute during
the process of catalysis. The observations of immunology added fur-
ther support to this concept, since it was well known that relatively
minor changes in, for example, the structure of a hapten could
lead to a profound modification in the effectiveness of reaction with
a specific antibody.

The idea of the “inviolability” of protein structure is now slowly
being replaced with the idea of the “functional adequacy of less than
the whole.” Shortly after Sanger and his colleagues had completed
their monumental studies on beef insulin, it was shown by Lens? that
a known, degradative change in the hormone, namely, the removal
of the C-terminal alanine residue of the B chain, did not lead to loss
of biological activity. (The evolutionary implications of this observa-
tion were not particularly obvious at the time since the experiment
was the first of its kind and had to be thought of as an isolated ex-
ample. Now, however, as the result of many similar observations,
we must concern ourselves (see Chapter 11) with the question of
why the C-terminal alanine residue is preserved in insulin as a con-
stant structural feature if it serves no function in the biological activity
of the hormone. )

Insulin has been subjected to other, more extensive, studies of this
sort. However, for the purpose of illustration of the extent to which
proteins may be degraded without inactivation, we shall consider
instead three other examples about which we have somewhat more
information: the pituitary hormone, ACTH, the pancreatic enzyme,
ribonuclease, and the plant enzyme, papain. In the following dis-
cussion of these examples, two quite different approaches to the struc-
tural basis of biological activity will be considered more or less simul-
taneously. One aim is to show that active polypeptides can be de-
graded without loss of function, and is concerned mainly with indi-
cating the magnitude of unessential parts of structure. The other is
directed deliberately toward the delineation of the essential parts,
that is, the active centers.

128 THE MOLECULAR BASIS OF EVOLUTION

Adrenocorticotropic Hormone

Complete sequences have been elucidated for adrenocorticotropic
hormone (ACTH) isolated from three species, hog, beef, and sheep.
The problem of species variation will be discussed elsewhere. For
our present purposes we shall consider only the material isolated
from hog pituitaries since the relation between structure and function
appears to be the same for the hormones of all three species.
Adrenocorticotropic hormone is one of a large number of polypep-
tides which can be demonstrated in extracts of the anterior pituitary
gland. The hormone has been prepared in highly purified form and
has been shown to cause the selective stimulation of the adrenal cor-
tex with the release of adrenal steroid hormones. It also causes de-
pletion of the ascorbic acid and cholesterol stores of the adrenal gland
and is capable of stimulating the repair of histological changes owing
to hypophysectomy.

The sequence of pig ACTH is shown in Figure 67. The polypep-
tide is composed of 39 amino acid residues and contains several
intriguing groups of amino acids, notably the sequence of three hy-
droxyamino acids, Ser.Tyr.Ser, at the N-terminus and the highly
basic grouping, Lys.Lys.Arg.Arg, in positions 15 to 18. The action
of carboxypeptidase removes three residues from the C-terminal end
of the chain without loss in hormonal activity. A more severe deg-
radation results during limited pepsin digestion with the removal of
the eleven C-terminal amino acids, still without change in activity.
These experiments establish that, as a maximum, only the first 28
residues of ACTH are required for hormonal function. Paul Bell
and his co-workers have recently reduced this estimate, reporting that
by mild acid hydrolysis of the chain the last 15 residues may be re-
moved without inactivation.’

It has also been possible to characterize the activity of ACTH in
terms of specific aspects of structure which are essential. Thus it has
been found that degradation of the N-terminal end of the chain with
leucine aminopeptidase results in complete loss of activity after only
one or two residues have been removed. Recent studies on the
action of fibrinolysin by White and Gross‘ have further underlined
the importance of the first half of the hormone. As is shown in
Figure 67, this proteolytic enzyme cleaves the chain at two points,
following residues 8 and 15, with complete inactivation.

These various degradative studies establish that at least 16, and
perhaps as many as 24, of the amino acids of ACTH are essential
ones. The ultimate elucidation of the exact structural requirements

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 129
So

QBOSOOSEHOOOO]
T aC)
G2OSOOGOGOSO” “
Ge
SQQQQQQ_Q@S

Figure 67. The structure of porcine «corticotropin. Points of attack by

chymotrypsin (C), trypsin (T), pepsin (P), and fibrinolysin (F) are indicated
by the arrows,

—]

for ACTH activity must probably await the step-by-step synthesis
of the sequence, and such studies are in progress in Klaus Hofmann’s
laboratory,* where the unequivocal synthesis of a peptide correspond-
ing to at least the first 14 amino acids in the polypeptide has been
achieved. At last report, the synthetic material had not yet reached
sufficient length to exhibit hormonal activity. In the meantime,
Boissonnas and his colleagues* have reported the synthesis of a biologi-
cally active peptide consisting of the first 20 N-terminal amino acids
of ACTH. Although the synthesis is not entirely unequivocal, the
results give strong support to the conclusions drawn from the
degradative work.

The studies on ACTH are of particular interest since they have to
do with a substance which is assayed by in vivo methods. Investiga-
tions of the changes in activity during the degradation of enzymes
must generally be controlled by in vitro tests. The possibility that
functional adequacy in the test tube does not give a true picture of
activity in an organized cellular environment adds an aspect of un-
certainty to the interpretations. Discounting the rather slim possi-
bility that degraded ACTH is reconverted to native ACTH during
activity assays, we have, for this hormone, clear-cut evidence for
the “functional adequacy of less than the whole.”

Papain

Papain is a powerful proteolytic enzyme present in papaya latex
which has been crystallized, both as the native protein and as the

130 THE MOLECULAR BASIS OF EVOLUTION

mercuri-derivative, by Emil Smith and his colleagues.’ The enzyme
is of considerable interest as a model for the study of the relation-
ships between structure and function because it appears to contain
a great deal of structure which is “superfluous” for activity. The
molecule contains six sulfhydryl groups, five of which are not reactive
with the usual sulfhydryl reagents in the native protein but appear
from within the inner regions of the structure only after suitable
denaturation. The single reactive sulfhydryl is essential for activity,
and complete inactivation occurs when two molecules of the protein
react with a single molecule of mercuric ion to form the dimer mer-
captide. No other sulfur is present beyond the six cysteine residues,
and cross-linkage of the 180 amino acid residue chain through cystine
bridges is therefore excluded.

The most exciting development in the studies on papain has been
the observation by Hill and Smith® that approximately 80 residues
may be removed from the N-terminal end of the chain by leucine
aminopeptidase without loss in catalytic activity. The remaining
C-terminal stretch of amino acids is relatively poor in basic amino
acids but rich in tyrosine and valine. The noncritical nature of
lysine side chains is indicated by the fact that the «amino groups
may be converted to the guanido derivatives, by treatment with
O-methylisourea, without activation.

Papain is reversibly inactivated by strong urea solutions and, to
some extent, irreversibly by guanidinium ions. Under such condi-
tions of denaturation the shape of the molecule is considerably altered
as evidenced by changes in the optical rotatory, viscosity, and ultra-
centrifugal properties. Therefore, although only a portion of the
papain chain is essential for catalysis and is apparently not dependent
on covalent cross-linkages, there are obviously important secondary
and tertiary structural features involved in the determination of its
catalytically active center.

Ribonuclease

The present status of our knowledge of the sequence of ribonuclease
has been reviewed in an earlier chapter together with a considera-
tion of some of the physical properties of this enzyme, both in its
native state and under denaturing conditions. With this background
we can now examine some of the covalent and noncovalent aspects
of its structure in connection with the problem of catalytic activity.
As we have seen, the protein contains two rather longish “tails,” one
at the N-terminal end consisting of 25 residues and one at the

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 131
C-terminal end consisting of 14 (Figure 62, Chapter 5). Both of
these parts of the chain have been subjected to controlled hydrolysis
with proteolytic enzymes. It was first shown by F. M. Richards?
that native ribonuclease, when digested with the bacterial enzyme
subtilisin, was rapidly and specifically hydrolyzed at the Ala.Ser bond
(residues 20-21) in the N-terminal “tail.” The resulting derivative,
in which the 20 residues of the “tail” portion were still attached to the
body of the enzyme through some particularly strong noncovalent
interaction, could be isolated on ion exchange columns and was
shown to retain the full activity of the original protein. This active,
degraded ribonuclease molecule now also contains, in addition to the
N-terminal lysine residue characteristic of the native enzyme, a new
N-terminal serine residue.

The C-terminal end of the chain can also be degraded without
inactivation. It has been possible to remove the C-terminal valine
residue by carboxypeptidase treatment, together with a fraction of
the serine and alanine residues which immediately precede the valine.
In such samples of degraded protein no evidence for inactivation has
been observed.

The site of these noninactivating cleavages with proteolytic enzymes
may be assigned to specific portions of the linear polypeptide chain.
Some active products of more extensive digestion have also been de-
scribed for which, unfortunately, the exact sites of cleavage cannot
be pinpointed at present. Uziel, Stein, and Moore,’ for example,
have reported the isolation of active derivatives of ribonuclease which
has been subjected to trypsin digestion under conditions of partial
unfolding of the protein in dilute urea solutions. Similarly, Kalnit-
sky and Rogers" describe experiments using carboxypeptidase diges-
tion in which the digesting enzyme apparently contained other pro-
teases of unknown types which caused extensive degradation of the
molecule without loss of much activity. As the analysis of these
latter studies becomes more complete and it becomes possible to as-
sign the changes that have occurred to definite parts of the polypep-
tide chain, we shall be able to make much more useful statements
about the location and nature of the “active center” of this enzyme.

We may also derive useful information from experiments designed
to determine the minimum changes necessary to cause inactivation of
the enzyme. Two particular experiments are of special interest here.
The first has to do with the inactivation of ribonuclease activity dur-
ing photooxidation. Weil and Seibles'? have shown that the enzyme,
when exposed to light in the presence of methylene blue, undergoes
slow loss of activity and that this loss parallels, at least during the

132 THE MOLECULAR BASIS OF EVOLUTION

early stages of photooxidation, the disappearance of one (possibly
two) of the four histidine residues that occur in this protein. Al-
though the exact location of the histidine residue (or residues) in-
volved has not been determined, it should be noted that two of the
four are found in the C-terminal twenty amino acids of the chain.

A second sort of inactivating treatment which may ultimately help
to elucidate the chemical basis of ribonuclease activity involves pep-
sin digestion under carefully controlled conditions. When the en-
zyme is exposed to pepsin at pH values in the neighborhood of 1.8,
there occurs an extremely rapid hydrolysis of what appears to be a
single, particularly labile, peptide bond.'* The macromolecular por-
tion of the reaction mixture may be isolated on ion exchange columns
and is completely devoid of activity. The only other detectable
fragment is a tetrapeptide, Asp.Ala.Ser.Val, the sequence of which
indicates that it was derived from the C-terminal end of the chain
by the cleavage of the peptide bond following the phenylalanine
residue at position 120 in the polypeptide chain (see Figure 62).
It has not been possible to detect the presence of cleavages elsewhere
in the inactive macromolecular derivative, and both physical studies
and C-terminal amino acid analysis support the conclusion that
nothing else has happened beyond this unique hydrolysis.

Since the earlier studies on the carboxypeptidase treatment indicate
the nonessentiality of the three C-terminally located amino acid resi-
dues, these results with pepsin suggest that the aspartic acid residue
located at position 121 has some special function in the maintenance,
or mechanics, of the catalytically active center. We shall discuss this
point further in connection with observations on the spectral prop-
erties of ribonuclease and its various active and inactive derivatives.

Although the combined consideration of the degradative experi-
ments indicates that some portions of the sequence of ribonuclease
have greater importance in activity than others, we cannot conclude
that catalysis by this enzyme requires only a simple linear sequence
of amino acids. Both oxidized ribonuclease and ribonuclease that
has been converted to a random, open chain by reduction of the
disulfide bridges are completely inactive. It seems very likely that
one or more portions of the sequence, perhaps quite widely separated
in the sense of linear distance along the chain, are essential, and that
their relation to one another in space is fixed by restrictions intro-
duced by disulfide bridges and by other bridges of a noncovalent
nature.

Some recent experiments on the relation between controlled, step-
wise reduction of the disulfide bridges in relation to activity are of

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 133
 

 

 

 

100
a
75 om
Vv e
2 °
= 50h ° ° 7
2 veNX
° A oe
25 . 7
° Oe o
o e
oe
i | |
°5 2 4 6 8
SH groups

Figure 68. Activity of ribonuclease at various stages of reduction (expressed as
percentage of the specific activity of native ribonuclease) as a function of the
number of moles of sulfhydryl per mole of enzyme. Solid triangle, reduction in
absence of urea; solid circle, reduction in 8 M urea; open square, reoxidation of
fully reduced, inactive ribonuclease; open circle, reoxidation of samples contain-
ing more than six sulfhydryl groups per average molecule; open triangle, reoxida-
tion of samples containing about four sulfhydryl groups per average molecule.
From M. Sela, F. H. White, and C. B. Anfinsen, Science, 125, 691 (1957).

interest in this regard.1* The disulfide bonds may be reduced by
thioglycollic acid and by various other reducing agents as discussed
in Chapter 5. It has been found that the reduction of one of the
four bridges in ribonuclease goes fairly rapidly in the absence of urea,
but that in 8 M urea, in which the three-dimensional structure of the
protein has been disoriented and “loosened,” all four are easily
cleaved. During reduction the activity falls off in a manner suggest-
ing that one, and perhaps two, of the bridges may be dispensed with
from the standpoint of enzymatic activity (Figure 68). Thus when
an average of one SS link has been reduced, we can still demonstrate
the presence of 80 to 90 per cent of the initial activity in spite of the
absence of all but traces of native enzyme. Even at an average level
of four SH groups per mole of protein (equivalent to the rupture of
two SS bridges), considerably more than half the activity remains.
It has recently been possible to separate some of the intermediates
of early stages of reduction on ion exchange columns, and the iden-
tification of the bridges that are most susceptible to reductive cleavage
is now under investigation. Preliminary results indicate that nearly

134 THE MOLECULAR BASIS OF EVOLUTION

full activity is retained after the disulfide bridge between the first
and sixth half-cystine residue is cleaved.

The reduction experiments, and the proteolytic studies discussed
earlier, permit us to perform a certain amount of paper surgery on
the enzyme and to construct a two-dimensional picture of the max-
imum structural requirement for activity (Figure 69). It seems likely
that amputations of other parts of this structure may be possible
without inactivation. We can already safely conclude, however, that
considerable portions of the covalent structure of ribonuclease must
either be unimportant evolutionary vestige or involve aspects of
ribonuclease function and intracellular behavior about which we are
still in complete ignorance.

  
   
 

photooxidation here
| Removed, after subtilisin
cleavage, by carboxypeptidase;

“tail” can still reactivate
“core”

Active after reduction
of 1-6 disulfide

 
 
   

L
Inactivated by cleavage here
with pepsin at pH 18

C-terminal residue with
carboxypeptidase

Figure 69. A schematic drawing of the ribonuclease molecule in two dimensions
(see also Figure 62). Various experimental modifications of the native molecule
are indicated, together with the consequences of such modifications in terms of
enzyme activity. Also included are some results of studies on the comparative
structures of bovine and ovine pancreatic ribonucleases (see Chapter 7). The
data summarized in this figure make it tempting to suggest that much of or all
the N-terminal “tail” of the molecule, together with only a portion of the “core,”
might eventually be shown to be sufficient for catalytic activity. ‘The “active
center” must obviously be a complicated three-dimensional structure but may not
necessarily involve more than a fraction of the entire protein. This drawing is
extremely speculative and is included here only to indicate some of the directions
that research on the biological activity of this enzyme may take in the future.

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 135
Before leaving the subject of enzymatic structure in relation to
function (and the reader must excuse my preoccupation with a spe-
cific enzyme about whose properties I happen to be able to write
with less labor than certain others of equal illustrative value), let us
consider a few aspects of noncovalent structure in terms of catalytic
activity." We have referred, in Chapter 5, to some of the physical
characteristics of ribonuclease under conditions of reversible dena-
turation by urea. The rather large changes that occurred in the
structure of the enzyme under such conditions (e.g., the intrinsic vis-
cosity of ribonuclease in 8 M urea shifts to about 0.085 as compared
with a value of 0.036 in dilute salt solution) made it of interest to ex-
amine whether or not activity was retained in 8M urea solutions.
It was quite surprising to find that hydrolysis of RNA was unimpaired
and, indeed, even somewhat stimulated under these conditions of hy-
drogen bond disorientation. The conclusion was drawn that activity
was independent of organized, three-dimensional structure. That
this conclusion might be premature, however, was suggested when
it was subsequently observed that low concentrations of phosphate
ions or other polyvalent anions (e.g, arsenate) could reverse the un-
folding effect of urea. The natural substrate for ribonuclease, RNA,
is itself a polyanion and might also cause refolding of the enzyme
under the conditions of assay. Various technical difficulties having
to do with the spectral and optical properties of ribonucleic acid
made the direct test of this hypothesis difficult. However, model
substances such as polymetaphosphate and uridine-3’-phosphate
(which is the product of ribonuclease attack on the synthetic sub-
strate, uridylic-2’,3’-cyclic-phosphate) could also cause a refolding of
the molecule as based on measurements of its spectrum, optical ro-
tation, and intrinsic viscosity (Table 6). The refolding was not com-
plete, as evidenced by the somewhat lower optical rotation of the
enzyme in 8 M urea in the presence of these agents as compared with
that of the native enzyme in water, and by an incomplete regenera-
tion of viscosity behavior. However, spectral differences were com-
pletely abolished.

Although these experiments suggested that complete “nativeness”
might not be essential, as indicated by the partial but significant ir-
reversibility of the optical rotatory and viscosity properties of the en-
zyme to anions, it appeared that the portion of the three-dimensional
structure of the enzyme which is responsible for the peculiar spectral
properties of several of the six tyrosine residues (page 120, Chap-
ter 5) must be in the proper configuration.

In an effort to support this hypothesis, the spectra of various ac-

136 THE MOLECULAR BASIS OF EVOLUTION

 

TABLE 6
Some Physical Properties of Ribonuclease and Modified Ribonucleases
Unfolding
Ribonuclease Agent Anion Ae.1073" lal, Nsp/e”
Native - = 1120 —71.7° 0.036
Native 8 M urea - 0 —103.7° 0.085
Native 8Murea 0.4M 110 —98.8°
chloride
Native 8 Murea 0.15 M 1120 —81.0° 0.050
phosphate
Native 8 Murea 0.15 M 1120 —80.7°
arsenate
Native 8Murea 0.15 M 0.052
uridylate
Oxidized - - 0 —91.6° 0.116
Reduced and car-
boxymethylated - - 0 0.149
Pepsin inactivated - - 0 —83.8° 0.039
Subtilisin digested - - 1120

 

The concentration of the proteins was 2.5-2.7 grams/liter for spectro-
photometric readings, 25-30 grams/liter for polarimetric, and 14—15 grams/
liter for viscosimetric measurements. All measurements were carried out
at pH 6-7. Anions were added as their sodium salts.

The active derivative of ribonuclease produced by limited subtilisin diges-
tion was purified according to Richards (see page 132).

*Change in extinction relative to the absorption of ribonuclease in 8 M
urea in the absence of salts, at 285 mu.

> Reduced viscosity, in (grams/100 ml)~4. All viscosity measurements
were carried out in 0.1 M KCI.

tive and inactive derivatives of ribonuclease were examined. It was
found that all the derivatives of ribonuclease that possessed activity
showed a normal “shifted’ *spectrum, whereas all those that were inac-
tive possessed the spectral characteristics of tyrosine residues with
unmodified resonance properties. We might tentatively conclude
from these studies that the presence of a shifted, “native,” spectrum
leads to a “diagnosis” of catalytic activity and that factors which
cause changes in the spectrum may be expected to inactivate.

To summarize our impressions at this point, it appears that at least
one bond in the N-terminal and part of the C-terminal “tail” are dis-
pensable, and that one, and perhaps two, of the SS bridges may be
opened with impunity. Various findings suggest that a part of the

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 137
structure in the vicinity of the C-terminal “tail” has special impor-
tance, specifically the aspartic acid residue at position 12]. Finally,
there appears to be good correlation between the presence of a
“shifted” spectrum, as occurs in the native enzyme, and activity.
Since the evidence is now rather convincing that such shifted spectra
may involve hydrogen bonding between carboxylate groups in the
protein and hydroxyl groups on tyrosine residues, it may be sug-
gested, as one alternative, that the free carboxylate group of the
aspartic acid residue at position 121 is involved in such a linkage
and functions as one of the determinants of the three-dimensional
form of the active center of the enzyme.

The importance of considering the enzymatic activity of a protein
in terms of three-dimensional structure is strongly emphasized by an
especially intriguing aspect of ribonuclease chemistry, recently re-
ported by F. M. Richards.® In a continuation of his studies on the
enzymatically active derivative prepared by limited subtilisin diges-
tion, he has investigated the nature of the noncovalent bond or
bonds (see page 132) which hold the N-terminal “tail” peptide to the
macromolecular portion of the protein. The attachment may be
broken by treatment of the derivative with trichloroacetic acid, and
the N-terminal peptide fragment may be separated by dialysis and
purified by electrophoretic methods. Richards has shown that both
the peptide and the macromolecular component are inactive, but that
upon mixing the two activity is completely regenerated as shown in
Figure 70.

Richards has also examined the spectral properties of the separate
fragments and of the reconstituted, active mixture. The 20-amino
acid peptide, which contains no tyrosine, of course shows no absorp-
tion at 280 mp. The large piece, however, has the expected absorp-
tion in this region of the ultraviolet, but shows none of the spectral
shift observed with the native protein or with the subtilisin-digested
enzyme before separation of the two fragments. Its spectrum is ex-
tremely similar to that observed for native ribonuclease in urea or
after treatment with inactivating reagents such as pepsin. Upon
mixing the two components, the shifted tyrosine spectrum reappears
almost completely, indicating that the peptide “tail” plays an impor-
tant part in the determination of that part of the three-dimensional
structure of the protein which is responsible for the anomalous tyro-
sine absorption.

We have, in Richards’ experiments, an instance of biological func-
tion in a relatively small peptide which is strongly reminiscent of the
action of some of the peptide hormones of the anterior and posterior

138 THE MOLECULAR BASIS OF EVOLUTION

 

 

 

 

150 | V7
/
7
9’ ©
2100}. —---4- 5 4
>
3 °
2
8
5
* 50 9° -
| | _t |
99 5 10 4 25 50

Volume of N-terminal peptide solution added, ul.

Figure 70. The regeneration of ribonuclease activity by the addition of the N-
terminal peptide (which is split off from the native molecule by subtilisin) to the
inactive macromolecular “core.” As described in the text, the “core” was pre-
cipitated from the subtilisin digest with trichloroacetic acid. The N-terminal
polypeptide, containing twenty amino acid residues, causes complete regenera-
tion of activity when it has been added in amounts approximately equimolar with
the concentration of the macromolecular component. From F. M. Richards,
Proc. Soc. Natl, Acad. Sct. U.S., 44, 162 (1958).

parts of the pituitary gland. The complete regeneration of ribonu-
clease activity occurs at concentrations of the order of 10~¢ M, a level
within quite reasonable limits for hormonal action. On the basis of
these studies alone, the N-terminal peptide fragment does not neces-
sarily have to be an intrinsic part of the active center. It may con-
ceivably function only as a determinant of proper folding in the rest
of the molecule. Whatever its role, the forces that bind this small
polypeptide to the body of ribonuclease must be multiple and highly
specific.

In addition to being an essential partner to the catalytic “center” of
ribonuclease, Richards’ peptide fragment also appears to possess the
ability to stabilize the rest of the molecule. The active subtilisin
derivative is stable in aqueous solution, reflecting the considerable
strength of the association between the two components of the com-
plex. In urea, however, where these components dissociate, as they
do in trichloroacetic acid, the macromolecular fraction is rapidly de-
natured and cannot be reactivated by the addition of the peptide,
even following removal of the dissociating agent. The specific, cata-

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 139
lytically active configuration in the intact subtilisin derivative is thus
dependent on the participation of certain bonds and amino acid resi-
dues contributed by the macromolecular body of the derivative, sta-
bilized by the peptide “tail” fragment.

It is much too early to attempt to synthesize the various observa-
tions on the chemical basis of ribonuclease into a coherent whole.
It is a tribute to the progress of modern protein chemistry, however,
that we can even hope to do so in the near future. The pattern al-
ready emerging for ribonuclease, in which certain widely separated
areas of the protein are strongly implicated as components of an
active center, and other portions (at least one disulfide bond and an
assortment of amino acid residues at various points in the chain) are
not, is an object lesson in the necessity for reorienting our thinking
about proteins from planar to solid geometry.

REFERENCES

1. Described in Crystalline Enzymes, John H. Northrop, Moses Kunitz, and
Roger M. Herriott, second edition, Columbia University Press, New York,
1948,

. J. Lens, Biochim. et Biophys. Acta, 3, 367 (1949).

. P. H. Bell, K. S. Howard, R. G. Shepherd, B. M. Finn, and J. H. Meisen-
helder, J. Am. Chem. Soc. 78, 5059 (1956).

. W. F. White and A. M. Gross, J. Am. Chem. Soc., 79, 1141 (1957).

- K. Hofmann, T. A. Thompson, and E. T. Schwartz, J. Am, Chem. Soc., 79,
6087 (1957).

6. R. A. Boissonnas, S, Guttmann, J. P. Waller, and P. A. Jaquenoud, Ex-

perientia, 12, 446 (1956).

7. E. L. Smith, Federation Proc., 16, 801 (1957).

8. R. L. Hill and E. L. Smith, Biochim. et Biophys. Acta, 19, 376 (1956).

9. F. M. Richards, Proc. Natl. Acad. Sci. U.S., 44, 162 (1958).

10. M. Uziel, W. H. Stein, and S. Moore, Federation Proc., 16, 263 (1957).

11. G, Kalnitsky and W. I. Rogers, Biochim. et Biophys. Acta, 20, 378 (1956).

12. L. Weil and T. S. Seibles, Arch. Biochem. Biophys., 54, 368 (1955).

13. C. B. Anfinsen, Federation Proc., 16, 783 (1957); M. Sela and C. B. Anfin-
sen, Biochim, et Biophys. Acta, 24, 229 (1957); M. Sela, C. B. Anfinsen,
and W. F. Harrington, Biochim. et Biophys, Acta. 26, 502 (1957); M. Sela,
F. H. White, Jr., and C. B. Anfinsen, Science, 125, 691 (1957).

on

Cr be

SUGGESTIONS FOR FURTHER READING

Anfinsen, C, B., and R. R. Redfield in Advances in Protein Chemistry, volume
11, Academic Press, New York, 1956.

140 THE MOLECULAR BASIS OF EVOLUTION

Enzymes: Units of Biological Structure and Function (O. H. Gaebler, editor),
Academic Press, New York, 1956.

Molecular Structure and Biological Specificity, edited by L. Pauling and H. A.
Itano, Publication 2, American Institute of Biological Sciences, Washington,
D. C., 1957.

Symposium on Protein Structure (A. Neuberger, editor), John Wiley & Sons,
New York, 1958,

THE BIOLOGICAL ACTIVITY OF PROTEINS IN RELATION TO STRUCTURE 141