Local stability identification and the role of a key aromatic
amino acid residue in staphylococcal nuclease refolding
Zhengding Su
3
, Jiun-Ming Wu
1
, Huey-Jen Fang
1
, Tian-Yow Tsong
2,3
and Hueih-Min Chen
1
1 Institute of BioAgricultural Sciences, Academia Sinica, Taipei, Taiwan, ROC
2 Institute of Physics, Academia Sinica, Taipei, Taiwan, ROC
3 Department of Biochemistry, University of Minnesota College of Biological Sciences, St Paul, MN, USA
Staphylococcal nuclease (SNase) is a single-domain
protein of 149 amino acids. Its 3D structure has been
examined by NMR [1–3] and X-ray crystallography
[4,5]. However, only part of the structure (positions 1–
141) can be confirmed. The segment 142–149 has not
been defined with certainty because of its apparent
flexibility. Structural observations made with NMR or
X-ray have led to the prediction that certain amino
acid(s) in the flexible segment stabilize the rest of the
structure, in particular the key amino acid(s) located
close to this flexible segment. The tryptophan at posi-
tion 140, for example, may play an important role in
maintaining the protein structure formed by amino
acids 1–141. In this study, we used site-directed muta-
genesis to generate point mutations and truncations
around this position to explore the above prediction.

acid(s) which act as nucleation centres before proton
dissociation. Subsequently, these centres trigger the
condensation of random polypeptide chains into the
compact form of the native state.
In this study, the effects of mutating W140 [10]
on SNase protein conformation and stability were
Keywords
aromatic amino acid; refolding; stability;
staphylococcal nuclease
Correspondence
H-M. Chen, Institute of BioAgricultural
Sciences, Academia Sinica, Taipei,
Taiwan 115, R.O.C.
Fax: +886 2 2788 8401
Tel: +886 2 2785 5696 ext. 8030
E-mail:
(Received 3 May 2005, revised 3 June
2005, accepted 13 June 2005)
doi:10.1111/j.1742-4658.2005.04814.x
Staphylococcal nuclease (SNase) is a model protein that contains one
domain and no disulfide bonds. Its stability in the native state may be
maintained mainly by key amino acids. In this study, two point-mutated
proteins each with a single base substitution [alanine for tryptophan
(W140A) and alanine for lysine (K133A)] and two truncated fragment
proteins {positions 1–139 [SNase(1–139) or W140O] and positions 1–141
[SNase(1–141) or E142O]} were generated. Differential scanning micro-
calorimetry in thermal denaturation experiments showed that K133A and
E142O have nearly unchanged DH
cal
relative to the wild-type, whereas

Figure 1A shows the CD spectra of W140A, E142O
and K133A. There are no significant differences from
that of the wild-type protein. In contrast, W140O has a
completely different CD spectrum. There are no peak
minima at 222 nm and 208 nm, but instead a peak mini-
mum at 200 nm, indicating that it has no secondary
structure only random coil. We can therefore postulate
that, when the segment beyond W140 was removed, the
secondary structure of the protein disintegrated. Using
the peak at 222 nm of the wild-type protein as an index
of helical conformational stability, this CD spectrum of
W140O suggests a decrease in helicity on removal of all
amino acids beyond position 140.
In CD spectra in the near-UV region (Fig. 1B), the
wild-type protein and mutants E142O and K133A show
strong intensity at  277 nm (h  )77 degreesÆcm
)2
Æ
dmol
)1
), revealing an intact tertiary conformation. In
contrast, W140A and W140O both lacked tertiary struc-
ture, as their intensities at 277 nm were only h  )22
and )13 degreesÆcm
)2
Ædmol
)1
(Fig. 1B). However,
W140A at 295 nm had a similar spectrum to that of the
wild-type protein, but W140O did not (h  0). This may

Z. Su et al. Staphylococcal nuclease refolding
FEBS Journal 272 (2005) 3960–3966 ª 2005 FEBS 3961
of the wild-type protein and K133A and E142O were
88.45, 84.70, and 83.60 kcalÆmol
)1
, respectively. The
melting point and DSC profile of E142O (52.8 °C) was
almost identical with that of the wild-type protein
(52.32 °C). K133A had a similar DH
cal
but a slightly
different melting point (47.70 °C) compared with the
wild-type protein. These results suggest that lysine at
position 133 and the residues that follow glutamic acid
at position 142 play negligible roles in maintaining the
native structure of SNase.
In contrast, W140A and W140O revealed no distinct
maxima in their DSC profiles. This is very different
from the wild-type protein and K133A and E142O,
which showed a significant enthalpy change. This
implies that either the protein has poor stability or no
secondary ⁄ tertiary structure results from either repla-
cing the tryptophan at position 140 with alanine or the
removal of a fragment from positions 140–149. This
is supported by the melting points of the proteins
(52.32 °C for the wild-type and ‘not detectable’ for
W140A and W140O). Flanagan et al. [15] reported
that multiple mutations can cause large changes in the
average conformation of denatured proteins. Here we
show that a specific single mutation or removal of a

Fig. 2. Steady-state fluorescent spectra of wild-type and mutant
SNase. Spectra of five proteins [wild-type (WT), W140A, E140O,
W140O and K133A]. Protein concentration was 0.4 mgÆmL
)1
.
Fig. 3. Calometric melting curves of wild-type and mutants of
SNase. DSC curves of five proteins [wild-type (WT), W140A,
E140O, W140O and K133A]. The curves of W140A and W140O
are nearly linear in terms of intensity. All protein concentrations
were 2 mgÆmL
)1
.
Staphylococcal nuclease refolding Z. Su et al.
3962 FEBS Journal 272 (2005) 3960–3966 ª 2005 FEBS
cannot maintain its tertiary conformation or even the
secondary motif. However, if residues 142–149 are
removed, the protein largely retains its 3D structure
(Table 1 and Fig. 2). Related experiments have been
performed previously. Parker et al. [18] showed that
the apparent association constant of SNase(1–126)
bound to pdTp in the presence of Ca
2+
is approxi-
mately threefold lower than that of the wild-type pro-
tein. Therefore, some or all of the residues in the 127–
149 segment are critical for maintenance of the native
conformation of the protein. In addition, Griko and
coworkers [19], studying SNase(1–136), showed that
this fragment (residues 1–136) has no secondary struc-
ture but retains the tertiary conformation. They thus

picosecond time scale. Taniuchi et al. [23] used both
SNase(1–126) and SNase(50–149) as typical fragments
to form type I and type II (in equilibrium) complex
proteins. They reported that these two overlapping
fragments could form enzymically active complement-
ary structures, and their folding rates were not related
to any decrease in energy from the unfolded to the
folded state. Feng et al. [24] created the SNase frag-
ments SNase(1–110), SNase(1–121) and SNase(1–135)
and used them in NMR spectroscopy experiments to
study the folding process. They concluded that the
conformation of these fragments could be considered
as native-like partially folded and unfolded states.
Recently, they further used the short N-terminal frag-
ments SNase(1–20), SNase(1–28) and SNase(1–36) to
show that the folding nucleation sites of SNase may
start from the N-terminus [25]. However, we used the
C-terminal fragments SNase(1–140) and SNase(1–142)
to show that SNase(1–140) plays a role in the assembly
of the protein during refolding. Our results also show
that SNase(1–140), without Trp at position 140, does
not have any structure, whereas the fragment including
Trp at position 140, i.e. SNase(1–142), has a similar
structure to the wild-type enzyme.
Our experiments show that tryptophan at position
140 plays an important role in maintaining protein ter-
tiary integrity. Figure 4A shows how tryptophan (loop
1) may interact with loop 2 and loop 3. These inter-
actions seem to form a ‘lower neck’ in the protein as
compared with the ‘upper neck’ which is formed by

intermediate state such as SNase(1–139) exists
Table 1. DSC results of SNase and its mutants. Phosphate buffer (25 mM Na
2
HPO
4
⁄ 50 mM NaH
2
PO
4
⁄ 200 mM NaCl, pH adjusted to
7.0) was used in the experiments. All proteins were used at a concentration of 2 mgÆmL
)1
. Difference from DH of WT (%) calculated by
[(DH
mutant
) DH
WT
) ⁄DH
WT
] · 100. Difference from DC
p
of WT (%) calculated by [(DC
p mutant
) DC
pWT
) ⁄DC
pWT
] · 100. WT, Wild-type; ND,
not detectable.
Average T

FEBS Journal 272 (2005) 3960–3966 ª 2005 FEBS 3963
(Fig. 4B, State 2). When tryptophan occupies position
140, the whole protein is folded into its native confor-
mation (Fig. 4B, State 3).
Summary
Trp140 located in the C-terminal loop of SNase plays
an important role in protein stability and conforma-
tional integrity. During protein unfolding ⁄ refolding,
the addition of 2.5 protons (both E75 and E129 are
the targets of protonation) results in SNase unfolding,
with refolding at an early stage through the formation
of a SNase(1–139) fragment in which the tryptophan
at position 140 is needed for formation of the native
structure.
Experimental procedures
Materials
Luria–Bertani broth and isopropyl thio-b-d-galactoside
were purchased from USB (Cleveland, OH). Salmon testes
DNA and some analytical grade chemicals such as EDTA,
Tris ⁄ HCl, CaCl
2
, NaCl and mineral oil were obtained from
Sigma (St Louis, MO, USA). Salmon testes DNA for the
enzyme activity test was used without further purification.
Guanidine hydrochloride and dNTPs were purchased from
Boehringer (Mannheim, Germany). Ethanol (> 99%) was
obtained from Panreac (Barcelona, Spain). Urea was a
product of Acros (Pittsburgh, PA, USA). The Quick-
change
TM

the mixture was overlaid with 30 lL mineral oil. A PCR
consisting of 16 cycles of 50 °C (1.5 min), 68 °C (14 min),
and 94 °C (1 min) was performed using a PerkinElmer 480
thermal cycler (Wellesley, MA, USA). The wild-type DNA
template was then digested by adding 1 lL DpnI restriction
enzyme to the PCR mixture and incubating at 37 °C for
1 h. Then 10 lL of the reaction mixture (containing undi-
gested mutant plasmid) was used to transform 100 lL
B
A
Fig. 4. Global segment interactions and the folding profiles in wild-
type SNase. (A) W140, in loop 1 interacts with loop 2 and loop 3
which forms a ‘lower neck’ network area in maintaining protein ter-
tiary structure. State 1 denotes the nascent polypeptide fragment.
(B) The fragment folding pathway induced via the SNase(1–139)
fragment and formed to its minimum energy native state via the
addition of either tryptophan at position 140 or fragment 140–149.
Staphylococcal nuclease refolding Z. Su et al.
3964 FEBS Journal 272 (2005) 3960–3966 ª 2005 FEBS
competent JM105 cells. The mixture was incubated on ice
for 1 h, and at 42 °C for 2 min, followed by 2 min incuba-
tion on ice. After transformation, 800 lL Luria–Bertani
medium was added, and the mixture incubated at 37 °C for
1 h. Transformed cells were selected on ampicillin plates,
and mutant DNA was isolated from the resulting colonies.
Mutant plasmids were then identified by BamHI and NcoI
restriction digestion, and their sequences confirmed by
DNA sequencing.
DNA sequencing
Plasmid DNA was isolated with the GFX

wild-type and confirmed to have the correct mutant
sequences.
Protein purification
Escherichia coli JM105 carrying recombinant plasmids were
grown in Luria–Bertani broth containing 100 lg ÆmL
)1
ampicillin at 37 °C. Protein expression was induced by
adding isopropyl thio-b-d-galactoside. The cells were har-
vested after 4 h of incubation and suspended in chilled
buffer A (6 m urea, 0.05 m Tris, 0.2 m NaCl, pH 9.2, fil-
tered through a 0.45 lm membrane). Proteins were collec-
ted after two alcohol precipitations and stored in buffer B
(6 m urea, 0.05 m Tris, pH 9.2, filtered through a 0.45 lm
membrane). The recombinant proteins were purified by
cation-exchange chromatography (washed CM-25 ion-
exchange gel column). The proteins were dialysed after
purification for 2 days at 4 °C against phosphate buffer
(25 mm NaH
2
PO
4
⁄ 50 mm NaHPO
4
⁄ 200 mm NaCl, pH
adjusted to 7.0) and were then lyophilized. The average
yield of recombinant proteins was  25 mgÆL
)1
. SNase
purity was investigated by SDS ⁄ PAGE. The gel was
stained with Coomassie blue and analyzed by densitome-

. Excitation was set at
298 nm, and emissions were observed at 350 nm. The fluor-
escence spectra were measured between 300 and 550 nm
with a scanning speed of 150 nmÆs
)1
and an excitation slit
of 5.0 nm.
Calorimetric measurements
Thermal analysis of protein denaturation was performed
with DSC (a model 6100 Nano II; Calorimetry Sciences
Corp., Provo, UT, USA). Lyophilized wild-type and
mutant SNase were dissolved in phosphate buffer at a
concentration of 2 mg ÆmL
)1
. Samples were first sonicated
for 15 min. Then 1 mL buffer or sample was loaded into
a clean reference or sample cell, respectively, ensuring that
the samples were free of air bubbles. Samples were heated
from 20 °Cto75°C under 3 atm at a heating rate of
1 °CÆmin
)1
. The melting point (T
m
) of protein tested was
directly obtained from the DSC curve. The enthalpy
change (DH
cal
) of each protein was calculated by integra-
tion of the curve covering area (T
m

tions for the origin of protein thermostability. J Mol
Biol 303, 125–130.
6 Chen HM, Chan SC, Leung KW, Wu JM, Fang HJ &
Tsong TY (2005) Local stability identification and the
role of key acidic amino acid residues in staphylococcal
nuclease unfolding. FEBS Journal 272, 3967–3974.
7 Chen HM, Markin VS & Tsong TY (1992) pH induced
folding ⁄ unfolding of staphylococcal nuclease: determin-
ation of kinetic parameters by the sequential jump
method. Biochemistry 31, 1483–1491.
8 Chen HM, Markin VS & Tsong TY (1992) Kinetic evi-
dence of microscopic states in protein folding. Biochem-
istry 31, 12369–12375.
9 Su ZD, Arooz TM, Chen HM, Gross CJ & Tsong TY
(1996) Least activation path for protein folding: inves-
tigation of staphylococcal nuclease folding by stopped-
flow CD. Proc Natl Acad Sci USA 93, 2539–2544.
10 Eftink M & Wasylewski Z (1992) Time-resolved fluor-
escence studies of the thermal and guanidine induced
unfolding of nuclease A and its unstable mutant.
Time-Resolved Laser Spectroscopy Biochemistry III
1640, 579–584.
11 Rodger A & Norde
´
n B (1997) Circular Dichroism and
Linear Dichoism. Oxford University Press Inc., New
York.
12 Nakanishi K, Berova N & Woody RH (1994) Circular
Dichroism Principles and Applications . VCH Publishers,
Weinheim.

21 Parikh I, Corley L & Anfinsen CB (1971) Semisynthetic
analogues of an enzymically active complex formed
between two overlapping fragments of staphylococcal
nuclease. J Biol Chem 246, 7392–7397.
22 Choy W-Y & Kay LE (2003) Probing residual interac-
tions in unfolded protein states using NMR spin relaxa-
tion techniques: an application to D131D. J Am Chem
Soc 125, 11988–11992.
23 Taniuchi H, Parker DS & Bohnert JL (1977) Study of
equilibration of the system involving two alternative,
enzymically active complementing structures simulta-
neously formed from two overlapping fragments of sta-
phylococcal nuclease. J Biol Chem 252, 125–140.
24 Feng Y, Liu D & Wang J (2003) Native-like partially
folded conformations and folding process revealed in
the N-terminal large fragments of staphylococcal nucle-
ase: a study by NMR spectroscopy. J Mol Bol 330,
821–837.
25 Dai J, Wang X, Feng Y, Fan G & Wang J (2004)
Searching for folding initiation sites of staphylococcal
nuclease: a study of N-terminal short fragments. Bio-
polymers 75, 229–241.
26 Hirano S, Kamikubo H, Yamazaki Y & Kataoka M
(2005) Elucidation of information encoded in trypto-
phan 140 of staphylococcal nuclease. Proteins: Struc-
ture, Function and Bioinformatics 56, 271–277.
27 Gill SC & von Hippel H (1989) Calculation of protein
extinction coefficients from amino acid sequence data.
Anal Biochem 181, 319–326.
Staphylococcal nuclease refolding Z. Su et al.