Kinetics of Folding of Staphylococcal Nuclease

Abstract. Staphylococcal nuclease undergoes a reversible structural transition
between pH 3 and 4 which can be measured by changes in tryptophan fluores-
cence. A stopped-flow spectrofluorometer was used to study the kinetics of
renaturation of nuclease from the acidified form on neutralization. The refolding
is fast, and the data can be described as a sequence of two first-order processes,
with half times of about 55 and 350 milliseconds, respectively,

The folding of polypeptide chains is
an intermediate step in making available
the information of the genetic code
for the biological function of the cor-
responding protein. Although many
studies have been made of the kinetics
of denaturation (J), few direct mea-
surements exist of the folding of pro-
teins from a disordered state to their
native conformation. We present here
measurements by stopped-flow spectro-
fluorometry of the renaturation of staph-
ylococcal nuclease from acid solution.

Nuclease is a single polypeptide chain
of 149 amino acids (molecular weight
16,800 daltons) of known sequence
(2) which has been partially synthesized
(3). There are no cysteine or cystine
residues. There is a single tryptophan
residue (position 140), located in the
2-A Fourier map just on the COOH-
terminal side of the last length of a-
helix and largely shielded from water
(4); the fluorescence of the tryptophan
is sensitive to conformational changes
in the protein (5). We have used the

70;

 

 

6014 549
—
£ ro"
.
3 50~— 391
2
s
s
5 40
s& 3
3 30
2 336

3.06
@ 20
2
oO
ond
oa oe
Buffer 1 l
° 300 350 400 450
Wavelength (nm)

Fig. 1. Fluorescence emission spectra of
staphylococcal nuclease at several pH

values from 7.02 to 3.06 at 26°C. Protein
(8 x 10°°A4) in 0.01M sodium phosphate
and 0.084 sodium chloride was adjusted
to the specified pH values with 1N HCl,
without significant change in volume.
Excitation light, at 285 nm, was passed
threugh a horizontal polarizing filter to
minimize light scatter (15); emission was
measured through a monochromator
grating cut for maximum diffraction effi-
ciency at 300 nm. The detector response
curve of this instrument is linear between
300 and 400 nm, therefore these are true
emission spectra (1/5).

changes in the quantum yield of
tryptophan fluorescence as an index
of the renaturation of nuclease when
the pH is changed from 3.2 to 6.7.

The stopped-flow instrument has been
described (6). With compressed air
(2.72 atm) as the driving force on the
syringes containing the reagents, we
obtained a mixing time of 10 to 20
msec. The mixing chamber and the
observation tube of quartz glass were
similar to that described except that a
four-jet ball mixer was employed in-
stead of a “T” type mixer (7).

The observation tube was placed in
an Aminco-Bowman spectrophotofiuo-
rometer fitted with an emission grating
cut for maximum diffraction efficiency
at 750 nm, which had good transmis-
sion in the second order. The R136
phototube was operated with a 900-volt
battery pack, and the anode current
was amplified and displayed on a vari-
able persistence oscilloscope. An elec-
tronic time constant of 10 msec was
employed. The oscilloscope tracings
were photographed on Polaroid film and
read on a Nikon profile projector at
tenfold magnification.

The fluorescence emission spectra of
staphylococcal nuclease (8), with exci-
tation at 285 nm, at several values of
PH are shown in Fig. 1. The major
emission, centered between 330 and 350
nm, results from the tryptophan ex-
cited state; there is some tyrosine emis-
sion near 300 nm. There is little change
in the tryptophan emission spectrum
between pH 7 and 4, but between pH 4
and 3 there is a marked decrease in the
amplitude of the fluorescence and a
slight shift in the emission maximum
toward longer wavelength. When fluor-
escence emission at 335 nm was mea-
sured as a function of pH by detailed
titrations, the acid-induced spectral
transition was fully reversible. The
abruptness of the transition suggested
that the protonation of one or several
groups with pK’s of about 3.8 caused
the transition.

The intrinsic viscosity of the nuclease
at pH 3 was about threefold greater
than that of the native nuclease (9).
This indicates that the spectral transi-

tion is accompanied by a general dis-
ruption of the native compact, globular
structure. Studies by difference spec-
troscopy, optical rotation, and circular
dichroism also denoted generalized con-
formational changes during the acid
transition (9). A nuclear magnetic reso-
nance study of the four histidine resi-
dues of nuclease has shown that their
chemical shifts become equivalent to
that of exposed histidine at pH 3.4
(10).

Figure 2A shows fluorescence emis-
sion during the mixing of nuclease at
PH 3.2 with an acid buffer, with no sig-
nificant change in pH. The dead time
in this experiment is 25 to 30 msec.

The oscilloscope tracing of the re-
naturation from acid solution of nu-
clease is shown in Fig. 2B. The acidi-
fied nuclease was mixed with an equal
volume of phosphate buffer to yield a
final pH of 6.70. The initial, rapid rise
in fluorescence intensity, as the neutral-
ized protein filled the cuvette, is fol-
lowed by a slower increase in fiuores-
cence as the nuclease molecules change
their conformation to that which is
stable at neutral pH. This change is
clearly measurable over the first 400
msec of the reaction, The value of
the amplitude of fluorescence after
a minute of reaction is the upper part
of the heavy white line, at about four

 

Fig. 2. (A) Photograph of the oscilloscope
tracing of the mixing of nuclease (initial
concentration, 1.2 x 10*M) in 0.002M
citric acid and 0.09M sodium chloride (to
keep ionic strength constant), pH 3.20,
with an equal volume of the same buffer
at 26°C. Time, increasing from right to
left, is measured in 50 msec per division
along the abscissa; fluorescence intensity
(ordinate) at 335 nm is monitored with
the second order transmission of the grat-
ing. At the arrow the mixed nuclease
solution began to fill the observation cu-
vette and the fluorescence of acidified
nuclease was registered. (B) Photograph
of oscilloscope tracing of mixing experi-
ment identical to the previous one except
that the buffer syringe contained 0.05M
sodium phosphate and 0.01M sodium
chloride, pH 6.90. At the arrow there was
a rapid rise in fluorescence as the mixed
nuclease solution at pH 6.70 filled the
cuvette; the following gradual! increase
in fluorescence is a measure of the re-
naturation of the protein.
 

Log (a—x}

 

 

2 1 1 } I 1 I |
oO 50 100 180 200 250 300 350 400

Time (msec)

Fig. 3. First-order kinetic plot of the re-
naturation of acidified nuclease from pH
3.20 to pH 6.70. The value of log (a — x)
is plotted on the ordinate where a is the
total increase in fluorescence from the be-
ginning to the observed end of the reac-
tion and x is the increase in fluorescence
at any time.

and a half divisions on the ordinate
above the base line.

The kinetics of this reaction have
been analyzed by first-order plots, with
the fluorescence of the acidified nu-
clease as the value for zero time and
the final recorded value as that for
infinite time. The first-order plots of
three experiments at a final concentra-
tion of protein of 6 x 10—5M fit one
line for about the first 150 msec of
the reaction (Fig. 3). Subsequently,
when the spectral change is more than
75 percent completed, they follow a
different line. After we subtract the
contribution of the second slope to the
first, these lines are equivalent to proc-
esses with first-order rate constants
(k,) of 12.1 and 1.9 sec—1, respec-
tively.

This reaction has been studied at pro-
tein concentrations from 1.5 to 6.0 x
10-5M, There was no significant
change in the initial first-order rate
constant.

The folding of nuclease from the
extensively disrupted acidified state to
the native form apparently can occur
very rapidly. The kinetic analysis, by
the simplest although not the only pos-
sible method, is compatible with an
initial, very rapid folding, with a half

Reprinted from SCIENCE

6 February 1970, Volume 167, pp. 886-887

time of about 55 msec (k, of 12.1
sec—1), which is accompanied by a
large change in tryptophan emission.
This is followed by at least one slower
conformational change, with a half time
of about 350 msec (k, of 1.9 sec—!),
which causes further change in the
environment of the tryptophan residue.

Most estimates of renaturation rates
have been established only indirectly
from the assumption of a two-state
transition whose apparent equilibrium
constant is the quotient of the renatura-
tion and denaturation rate constants
(11). However, Hauschka and Harring-
ton (J2) have recently directly mea-
sured the renaturation of Ascaris col-
lagen, at pH 6,25, by optical rotation
and have interpreted their data at 25°C
in terms of three first-order processes,
with half times of about 0.20, 2.5, and
11 minutes. Pohl (73) has used temper-
ature-jump relaxation methods in the
study of the reversible denaturation of
chymotrypsin and trypsin. The half
time values near 25°C of the renatura-
tion of these enzymes are between
about 7 and 70 seconds. The renatura-
tion rates of staphylococcal nuclease
appear significantly faster than the rates
of conformational change in these
proteins and faster than those in the
several proteins whose _ renaturation
rates have been estimated indirectly
(1). The need for more than one first-
order rate constant to fit our data indi-
cates the existence of kinetic intermedi-
ates in the conformational changes of
nuclease.

The rate of renaturation of nuclease
is slower by several orders of magnitude
than those established by temperature-
jump methods for the conformational
isomerizations of enzymes during their
catalytic function, and many orders
slower than those values established for
the helix-coil transformation (/4).
These results for the renaturation of
nuclease give an approximate idea of
the time required for this polypeptide
chain of 149 amino acids to assume a
large enough subset of all possible back-
bone and side-chain dihedral angles so
that the presumably statistical process
of folding reaches the thermodynamic-
ally stable, native conformation.

The measurement of the renaturation
kinetics of several spectral and hydro-
dynamic properties of nuclease, in ad-
dition to tryptophan fluorescence, is
necessary for an understanding of the
sequence of processes in folding. The
correlation of these results with the
structural information in the 2-A model
of nuclease offers an approach to an
understanding of the mechanism of
folding of this protein.

ALaN N. SCHECHTER
RAYMOND F, CHEN
CHRISTIAN B. ANFINSEN

National Institutes of Health,
Bethesda, Maryland 20014

References and Notes

ray

. For reviews see J. Steinhardt and E. M.
Zaiser, Advan. Protein Chem. 10, 152 (1955);
M. Joly, 4 Physico-Chemical Approach to the
Denaturation of Proteins (Academic Press,
New York, 1965), pp. 191-260; C. Tanford,
Advan. Protein Chem. 23, 122 (1968).

2. H. Taniuchi, C. B. Anfinsen, A. Sodja, J.
Biol. Chem. 242, 4752 (1967).

3. C. B. Anfinsen, Pure Appl. Chem. 17, 461
(1968); D. A. Ontjes and C. B. Anfinsen,
Proc. Nat. Acad. Sci. U.S., in press.

4. A. Amone, C. J. Bier, F. A. Cotton, E. E.
Hazen, D. C. Richardson, J. §. Richardson,
ibid,, in press, and personal communication.

5. P. Cuatrecasas, H. Taniuchi, C. B. Anfinsen,
Brookhaven Symp. Biol. 21, 172 (1968); H.
Edelhoch, R. L. Perlman, M. Wilchek, Ann.
N.Y. Acad. Sci, 158, 391 (1969); H. Taniuchi
and C. B. Anfinsen, J. Biol. Chem. 244, 3964
(1969), For a review of the physical basis of
such fluorescence changes in proteins see R.
F. Chen, H. Edelhoch, R. F. Steiner, in
Physical Principles and Techniques of Pro-
tein Chemistry, S. J. Leach, Ed. (Academic
Press, New York, 1969), p. 171.

. R. Harvey, Anal. Biochem. 29, 58 (1969);
R. F, Chen, A. N. Schechter, R. L. Berger,
ibid., p. 68.

7. R. L. Berger, B. Balko, H. F. Chapman,
Rev, Sci. Instrum. 39, 493 (1968).

. Highly purified nuclease was prepared from
the growth medium of Staphylococcus aureus
(Foggi strain) as described by S. Fuchs, P.
Cuatrecasas, C. B. Anfinsen [J, Biol. Chem.
242, 4768 (1967)} and L. Mordvek, C. B.
Anfinsen, J. L. Cone, H. Taniuchi [ibid.
244, 497 (1969)].

9. D, C. Richardson, A. N. Schechter, C. B.
Anfinsen, unpublished data. The value of the
intrinsic viscosity of the acidified nuclease is
very sensitive to changes in ionic strength.

10, J. §, Cohen, personal communication.

11. See for example, C. Tanford, R. H. Pain,
N. S. Otchin, J. Mol. Biol, 15, 489 (1966).
12. P. Hauschka, thesis, Johns Hopkins Univer-

sity (1969).

13. F. M. Pohl, Eur, J. Biochem, 4, 373 (1968);
ibid. 7, 146 (1969).

14. M. Eigen and G. Hammes, Advan, Enzymol.
25, 1 (1963); G. Hammes, Advan. Protein
Chem, 23, 1 (1968); R. Lumry, R. Legare, W.
G. Miller, Biopolymers 2, 489 (1964); E.
Hamori and H. A. Scheraga, J. Phys. Chem.
71, 4147 (1967).

15. R. F. Chen, Anal. Biochem. 14, 497 (1966);

ibid. 20, 339 (1967).

21 August 1969; revised 3 November 1969

an

oO

Copyright’ 1970 bythe

American Association for the Advancement of Science