302 Biotechnol. Prog.

2008, 24, 302−310

Preparation, Characterization and Refolding in Vitro of a Recombinant Human

Cyclophilin A Mutant: Effect of a Single Pro/Ser Substitution on Cyclophilin A
Structure and Properties
Li-Ren Xu, Xiao Yan, Man Luo, Yi-Xin Guan,* and Shan-Jing Yao
Department of Chemical and Biochemical Engineering, Zhejiang University, Hangzhou 310027, China

A recombinant cyclophilin A (CypA) mutant, which carries a serine instead of proline at sequence
16, was prepared for structural and functional assessment for human CypA. Soluble expression
of the recombinant CypA mutant in E. coli was obtained under 30 °C, 180 rpm culture condition
after being induced by IPTG. Ion exchange chromatography was used to purify the CypA mutant
in a single step, and a high activity recovery of target protein with a high purity was achieved.
Peptide fragments produced by trypsin proteolysis were applied to MALDI-TOF-MS, and
searching results from the NCBI protein databank confirmed the protein attribution as well as
the mutation sequence. Peptidyl-prolyl cis-trans isomerase activity was assayed for the CypA
mutant using tetrapeptide substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide, and the calculated kcat/
Km value was 1.5 × 106 M-1 s-1 at 10 °C, which was 10-fold lower than the previously reported
constant for wild-type CypA. An Eyring plot was also carried out. Inhibition by cyclosporine A
demonstrated that the IC50 value was 26.5 nM. Meanwhile the expected enhancement of intrinsic
tryptophan fluorescence was quenched by the mutation. The effect of CypA mutant on accelerating
protein refolding in vitro was investigated in ribonuclease A refolding process, and it was found
that 10% slow phase could be catalyzed by CypA. The protein was subject to urea and GdmCl
denaturation, where both activity and fluorescence served as structural probes. Activity recovery
indicated this CypA mutant was extremely sensitive to GdmCl and the susceptibility to urea
was increased. Low pH could also destabilize CypA. Furthermore the refolding of this CypA
mutant itself was studied. Although the activity yield was nearly unchanged, the former proposed
folding/assembly pathway might be altered. Fluorescence chart also demonstrated that the folding
time was extended, and fast-folding and slow-folding analysis indicated the slow-folding rate
constant presented a concentration dependence property denoting the autocatalysis of the foldase.

Introduction diverse proteins, and finally established as a rate-limiting step

in folding process (5).
Among the 20 essential amino acids, imino acid proline is Peptidyl-prolyl cis-tans isomerase (PPIase) is the specific
the only one which lacks the ability to form the classical enzyme catalyzing the Xaa-Pro bond isomerization and promot-
hydrogen bond. Incorporation of the five-membered pyrrolidine ing protein folding (6). It was first isolated from pig kidney in
ring of proline restricts the conformational space available for 1984 (7). In the same year, cyclophilin, the predominant
a polypeptide backbone and further imposes some restrictions immunosuppressant cyclosporin A (CsA)-binding protein, was
on secondary and tertiary structures of folded proteins. Gener- purified to homogeneity from bovine thymocytes (8). Interest-
ally, proline residues are considered as R-helix-breaking and ingly, in 1989, cyclophilin and peptidyl-prolyl cis-tans isomerase
â-sheet-breaking elements, though sometimes they are still found were found to be identical proteins independently by two
in the middle of R-helices where they cause some geometry research groups (9, 10). Later, other peptidyl-prolyl cis-trans
deformation (1, 2). Hence, proline residue plays a significant isomerases such as FK506-binding proteins FKBPs and rapa-
role in protein function; for instance, proline 75 is known to be mycin-binding proteins were also identified.
a key residue in thioredoxin-fold proteins (3). Among cyclophilin family, cyclophilin A (CypA), the 17.7
Except that proline plays an indispensable role in protein kDa cytosolic protein, appears abundantly expressed in all
structure, more intriguing is the proline isomerization coupled human tissues (6). CypA displays a high efficiency in the
in protein folding process. Rates of protein folding reactions isomer-specific proteolysis assay using tetrapeptide derivatives
vary in a wide range, even for a single polypeptide species. (usually Suc-Ala-Xaa-Pro-Phe-p-nitroanilide, with Xaa for any
Brandts (4) proposed the “proline hypothesis” to explain the native amino acids), in some cases, rate constants close to the
fast- and slow-folding phenomena: the two species differed in diffusion limit of 2 × 108 M-1 s-1 were reported (11). As a
conformational states, and cis-trans isomerization of one or more folding catalyst, CypA has also been found to accelerate prolyl
Xaa-Pro peptide bonds was responsible for the structural isomerization in protein substrates such as immunoglobulin light
differentiation. Now, this hypothesis was tested extensively for chain, S-protein fragment of bovine RNase A (12), tendamistat
(13), RNase T1(14), creatine kinase (15) and N-terminal domain
* To whom correspondence should be addressed. Tel: +86-571- of HIV-1 capsid (16). Value of kcat/Km of 73 000 was calculated
87951982, Fax: +86-571-87951015. E-mail: guanyx@zju.edu.cn. for CypA in catalyzing the refolding of reduced and carboxy-
10.1021/bp070259m CCC: $40.75 © 2008 American Chemical Society and American Institute of Chemical Engineers
Published on Web 03/12/2008
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Biotechnol. Prog., 2008, Vol. 24, No. 2 303

methylated form of (S54G/P55N) RNase T1 (11). Meanwhile CypA, and the cultivation condition was adjusted to 30 °C and
chaperone activity in assisting protein folding in the early stage 180 rpm. After incubation for additional 9 h, cells were
was also observed. harvested by centrifugation at 8 000 rpm for 10 min at 4 °C.
Site-directed mutagenesis has already been applied to inves- The wet cells were washed with 50 mM Tris-HCl (pH 8.5),
tigate the catalytic properties of CypA and brought us with deep and resuspended in 10% volume of the fermentation broth of
understanding of the mechanisms of proline isomerization in the same buffer. The cells were lysed by ultrasonication. After
the presence of peptidyl-prolyl cis-trans isomerase (17). Sys- discarding of the cell debris by centrifugation at 14 000 rpm
tematic replacement of all four cysteine residues in CypA with for 40 min at 4 °C, 6 mL of supernatant was loaded onto ion
alanine did not block the catalytic ability and activity in the exchange chromatography where a 25 mL DEAE Sepharose
binding assay of the isomerase, which ruled out cysteine as a Fast Flow resin column was used. The purification process was
participating residue in catalysis (18). Subsequently, it was found carried on an ÄKTA Explorer 100 (GE healthcare, USA) and
that the PPIase activity of the human CypA mutant (Arg55 f the absorbance at 280 nm was monitored. Purified CypA
Gly55) fell below 1% of the wild-type enzyme (19). Together fractions can be collected and protein concentration was
with other results, this finding corroborated the tentative that determined by denaturing SDS-PAGE and Bradford method.
the cis-trans interconversion of a Xaa-Pro bond occurred through Peptide Mass Fingerprint of CypA. Protein identification
binding of a peptide substrate in the hydrophobic cleft of the was confirmed by mass spectrometry. Single CypA band on
enzyme, and the binding was accompanied by out-of-plane coomassie brilliant blue stained SDS-PAGE gel was cut for
deformation of amide bond which was then stabilized by a MALDI-TOF-MS. After washed by Milli-Q H2O and 1:1
hydrogen bonding to the amino acid side chain of the binding acetonitrile(ACN)/50 mM NH4HCO3, the gel was dehydrated
cleft (1). by ACN. Then, dried gel was continually washed by NH4HCO3
In this paper, a recombinant CypA mutant with a substitution and ACN for several times. Trypsin was used to digest the gel
of proline16 to serine16 (which is more flexible in structure) for 4-6 h at 37 °C. The mass spectra were recorded by using
was prepared, and peptide mass fingerprint plot was used to a time-of-flight delayed extraction MALDI mass spectrometer
characterize this mutant. The catalytic efficiency toward tet- (Bruker Autoflex). Monoisotopic peptide masses obtained from
rapeptide Suc-Ala-Ala-Pro-Phe-p-nitroanilide was assayed and MALDI-TOF-MS were queried against entries for protein
inhibition of the PPIase activity by CsA was investigated; databases in NCBI using a protein search program. (Mascot,
meanwhile its ability in catalyzing RNase A refolding in vitro Matrix Science Ltd).
was also tested. It was shown that this mutant had the sensitivity PPIase Activity Assay. An improved spectrophotometric
to denaturant GdmCl and the susceptibility to urea was assay (7, 20) was used in this experiment. A typical 1 mL assay
increased. Finally, the refolding mechanism of this CypA mutant mixture consisted of 44 mM Hepes buffer (pH 8.0), 83 µM
was illustrated, and autocatalyzed folding was observed. Suc-Ala-Xaa-Pro-Phe-p-nitroanilide and 0.5 mg/mL R-chymot-
rypsin. Previously R-chymotrypsin was prepared and stocked
Materials and Methods in 1 mM HCl at the concentration of 10 mg/mL, and Suc-Ala-
Chemicals. Tryptone, yeast extract, ampicillin, and lithium Xaa-Pro-Phe-p-nitroanilide was dissolved in TFE and 450 mM
chloride (LiCl) were purchased from BBI. Hepes free acid and LiCl. All components of the assay mixture except for the peptide
R-chymotrypsin were from Amresco Co. Extra pure trifluoro- substrate were put together and preincubated at 10 °C. After
ethanol (TFE) was provided by J&K Chemical Ltd. Guanidine thermal equilibrium achieving, peptide substrate was added and
hydrochloride was from Sinopharm Chemical Reagent Co. Ltd. quickly mixed to initiate the reaction, and the absorbance at
Tetrapeptide substrate Suc-Ala-Xaa-Pro-Phe-p-nitroanilide and 390 nm was recorded per 2 s by an Ultraspec 3000 spectrometer
bovine pancreatic ribonuclease A were obtained from Sigma (GE Healthcare, USA). The data were fit to an exponential curve
Chemical Co. DEAE Sepharose Fast Flow resin was from GE using origin 7.5 (OriginLab Corporation), and first-order rate
healthcare. All other chemicals were analytical grade. constants could be derived. The mathematical equation for the
Bacterial Strain and Plasmid. E. coli BL21(DE3) strain was kinetic analysis was At ) A∞ - (A∞ - A0) e-kt, i.e., ln(A∞ - At)
used in this work, which was F-, ompT, hsdSB, rB- mB- (λ ) - kt + ln(A∞ - A0), where k denoted first-order rate constant
Ci857, ind1, Sam7, nin5, lacUV5-T7.AL). Plasmid pAVD30 and At was the 390 nm absorbance at time t. One unit of activity
which encodes the human CypA sequence was a gift from Prof. was defined as (kobs - k0)/k0, where kobs was the observed first-
Fersht in Centre for Protein Engineering MRC, Cambridge. The order rate constant of the enzyme-catalyzed reaction and k0 the
CypA sequence was introduced into a pRSET vector. This vector rate constant of the reaction without addition of CypA.
also codes an ampicillin-resistant gene, which could help to Inhibition by CsA. The same isomerase assay (using 10 nM
screen the CypA transformant. There was a mutation of the CypA) was performed as above, but a definite concentration of
amino acid at position 16 of CypA, instead of a proline, it was CsA ranging from 1 nM to 1 µM was included in this part (8,
a serine residue. The expression can be induced by IPTG under 18). To investigate the time course of CsA inhibition, CsA and
the control of strong T7 promotor. CypA were first mixed and preincubated for different time
Preparation of Recombinant CypA. The pAVD30 plasmid periods before launching the assay reaction.
encoding CypA was transformed into E. coli BL21(DE3) strain Refolding of Ribonuclease A. Bovine pancreatic ribonu-
by the calcium chloride method and transformants were screened clease A of 10 mg/mL was first denatured in 8 M urea at low
on 2×YT plates containing 150 µg/mL ampicillin. Cells were pH. Then denatured ribonuclease A was directly diluted to 100
grown aerobically in the medium composed of tryptone 16 g/L, mM sodium acetate buffer (pH 5.6) in the presence and absence
yeast extract 10 g/L, NaCl 5 g/L, and mannitol 10 g/L, which of 1 µM CypA, respectively, to initiate refolding reaction. The
was supplemented with 20 mg/L ampicillin. The transformant final concentration of ribonuclease A was maintained at 200
was precultured overnight under 37 °C, 220 rpm, and inoculated µg/mL and the refolding temperature was controlled at 13 °C.
to 200 mL medium in a 500 mL shake flask with a volume The refolding kinetics was monitored continuously by a
ratio of 10%. When the OD600 reached 1.3, IPTG was added to fluorescence spectrometer (F-4500, Hitachi). The excitation
the final concentration of 1 mM to induce the expression of wavelength was 278 nm (5 nm bandwidth), and the emission
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
304 Biotechnol. Prog., 2008, Vol. 24, No. 2

wavelength was 307 nm (2.5 nm bandwidth). Data points were

taken every a second time span. The observed changes in
fluorescence intensity were normalized by setting the total
fluorescence change equal to 1.
Determination of Urea and GdmCl Denaturation Curve.
CypA was incubated in 100 mM K2HPO4/KH2PO4 (pH 7.0 and
6.0, respectively) and 10 mM DTT in the presence of varying
concentration urea and GdmCl for 22 h at 30 °C (21).
Fluorescence emission of the denatured solutions at 350 nm (5
nm bandwidth) after excitation at 280 nm (2.5 nm bandwidth)
was measured as a probe of the extent of unfolding. The
concentrations of the denaturants in the stock solutions were
determined by the refractive index and urea solution should be
freshly prepared prior to use to avoid contamination (22). For
normalization, total fluorescence change was set to 1 and first-
order rate constant change between native and total denatured
CypA was also set to 1.
Refolding of CypA Protein. Different fractions of 100 µM
CypA denatured in 8 M urea and 10 mM DTT were diluted to
100 mM K2HPO4/KH2PO4 (pH 7.0) buffer to initiate refolding Figure 1. Induced expression of recombinant CypA in E. coli. (Lane
(23). Refolding temperature was 25 °C. The refolding kinetics 1: E. coli no plasmid and inducement. Lane 2: E. coli no plasmid but
was monitored by emission at 350 nm (5 nm bandwidth) after with inducement. Lane 3: E. coli with plasmid but no inducement. Lane
4: E. coli with plasmid and inducement.)
excitation at 280 nm (2.5 nm bandwidth). Refolded CypA was
further tested for the recovery of PPIase activity. The aggregate
in a concentrated initially 10 µM refolding sample was
investigated using native PAGE.
Results
Confirmation for the Mutation. Before transforming the
plasmid and expressing the target protein, DNA sequencing was
applied to know the coding sequence of recombinant CypA and
confirm the mutation. The experiment was performed on ABI
PRISM 3730, and the sequence kit was BigDye terminator v3.1.
Both T7 promoter and SP6 primers have been used for
sequencing from two opposite directions. For T7 promoter, only
about 300 nucleotides had detectable signal; fortunately, the
mutation site was included in the partial result. At the site
corresponding to the 16 position of amino acid, the codons were
TCC (which encode serine), but not the proline encoding codons
CCC in the wild-type. For the second primer SP6, complemen-
tary chain for the CypA encoding sequence was all sequenced,
and the result indicated that all nucleotides were consistent to
the wild-type except for the mutation site AGG, which also Figure 2. SDS-PAGE of the expression of recombinant CypA in E.
demonstrated a proline to serine mutation. coli under different cultivation conditions. (Lane 1: the precipitate at
Overexpression and Purification of CypA Protein. A high- 37 °C, 220 rpm. Lane 2: the supernatant at 37 °C, 220 rpm. Lane 3:
producing strain has been picked out by SDS-PAGE detection the precipitate at 30 °C, 180 rpm. Lane 4: the supernatant at 30 °C,
180 rpm.)
and preserved for CypA preparation. As shown in Figure 1,
only E. coli transformed with pAVD30 plasmid after induced ature and rotating speed were investigated. The results indicated
by IPTG could express a high-level of recombinant CypA that a decreased temperature to 30 °C and rotating speed to
protein. During the fermentation process, while the transformants 180 rpm after the addition of 1 mM IPTG was an efficient
were grown continuously under 37 °C and 220 rpm, large strategy to make target protein soluble. Under these conditions,
amounts of the recombinant proteins unfortunately aggregated the synthesis of recombinant CypA protein was slowed down.
to form inclusion bodies. After the cells were lysed by Therefore, the formation of inclusion body was minimized and
ultrasonication, both the supernatant and precipitate were the expression of soluble CypA was increased, as shown in
investigated by SDS-PAGE, and the results are shown in Figure Figure 2 (Lanes 3 and 4). Finally we achieved a production of
2. From that we know a large part of CypA protein fell into the 66.7 mg soluble CypA per liter fermentation broth at optimized
precipitate band (Lane 1), while the fraction in the supernatant conditions, and the yield was quite higher than that earlier
was not so rich as well (Lane 2). It was speculated that inclusion reported (18).
body might be formed for the heterologous protein. This was A single-step ion exchange chromatography was applied to
validated by following TEM detection (JEM-1230, JEOL, yield purified recombinant protein. The pI value for CypA was
Japan). TEM picture indicated that more than one-half of the quite high as 9.1 (18); therefore, even under pH 8.5 buffer
E. coli cells were filled with stained homogeneous materials, adsorption the protein could not bind to the anion resin. After
which were probably the inclusion bodies. being loaded onto the column, CypA sample was eluted in the
To increase the soluble expression of recombinant CypA in breakthrough peaks while other proteins were adsorbed by the
E. coli cells, the most important parameters including temper- column. Coomassie brilliant blue staining of the collected
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Biotechnol. Prog., 2008, Vol. 24, No. 2 305

Figure 3. Chromatogram of CypA protein purification using ion exchange chromatography and SDS-PAGE of the purified CypA.

Table 1. Purification of CypA Protein Using a Single-Step Ion Exchange Chromatographya

step total protein (mg) total activity (U) specific activity (U/mg) protein recovery (%) activity recovery (%)
lysate supernatant 154.9 6.32 × 105 4.08 × 103 100 100
IEC eluate 54.9 5.20 × 105 9.49 × 103 35.5 82.4
a Samples from lysate supernatant and IEC eluate were assayed for protein concentration and PPIase activity in the same condition. The amount of total

protein and PPIase activity in lysate supernatant was defined as 100%. Protein concentrations were assayed by Bradford method, and PPIase activity was
determined by chymotrypsin-coupled assay.

fractions gave only one visible band, noted in Figure 3. The velocity equation can be simplified to ν ) kcat/Km[E][S] ) k
purified CypA had the molecular mass of 18 kDa. The [S],where k ) kcat/Km[E]. kcat/Km can be obtained from the slope
purification results are summarized in Table 1, where specific of Figure 5B. The calculated kcat/Km for our mutant protein was
activity was calculated as the ratio of activity to the correspond- 1.5 × 106 M-1 s-1 at 10 °C, which was nearly 10-fold less
ing concentration of CypA. It seems that the separation process than several reported values as 1.37 × 107 M-1 s-1 (18) or 1.46
was effective and the activity recovery was quite high. × 107 M-1 s-1 (20). As kcat/Km reflects the inherent catalytic
Peptide Mass Fingerprint for CypA. The MS result was ability of enzymes, it was speculated here that the mutation
then delivered to NCBI for identification of target protein using Pro16 f Ser16 led to the decrease of PPIase activity of CypA
MASCOT software. The search matched with protein IDs of toward the tetrapeptide.
gi|1431788 and gi|1633054, which represented that CypA was The temperature dependence of the CypA mutant catalysis
complex with CsA and dipeptide Gly-Pro, respectively. For the was also carried out. Activation parameters were determined
peptide fragments 1-19 including the mutation site proline f by the isomerization of oligopeptide substrate from Eyring plots
serine, SwissProt protein mass tool predicated theoretical mass (Figure 6) of ln[k(p/k*T)] versus 1/T, where p is Planck’s
was 2067.0215 Da. From Figure 4, a rich peptide fragment in constant, k* is Boltzmann’s constant, and T is expressed in
the mass peaks was observed with mass value 2067.0067 Da, Kelvin. The standard concentration for CypA was 10 nM. In
which was basically in agreement with theoretical value. These these plots, the enthalpy and entropy of activation can be
results confirmed that we had cloned, expressed and purified obtained from the slope ∆Hq/R and y intercept ∆Sq/R, where R
the target protein mutant. is the gas constant(24). For noncatalyzed reaction, ∆Hq and
PPIase Activity Assay. Fischer (7) pioneered this specific -T∆Sq were equal to 23.62 ( 0.92 kcal/mol and -4.33 ( 0.90
isomer chymotrypsin-coupled assay. In this improvement of kcal/mol, respectively, whereas for the catalyzed reaction, ∆Hq
activity assay, TFE/LiCl incubation and low-temperature reac- and -T∆Sq were equal to 6.01 ( 0.51 kcal/mol and 12.74 (
tion were vital to the method (20). After this improvement, the 0.50 kcal/mol, respectively. Furthermore, it also revealed the
cis fraction of the substrate in this assay was increased to about signal-to-noise value in the temperature range from 10 to 25
55%, larger than the 9% value in former incubation system. °C, and at 20 °C, the background of the reaction has unfavorably
Meanwhile reaction at 10 °C significantly increased the signal- outweighed the enzymatic catalysis.
to-noise value, which was acceptable for a quantitative analysis Inhibition of CypA Activity by CsA. CypA forms very tight
of the PPIase activity. The original produced data are shown in stoichiometric complexes with CsA (8). Values in the range of
Figure 5A, indicating that variation of the amount of enzyme 1-1000 nM CsA were estimated for the activity inhibition of
led to a set of curves. The more CypA was added, the sharper isomerase with CypA concentration of 10 nM due to CsA
the curve displayed. The first-order rate constant k generated binding efficiently to CypA. For the time course study, catalysis
from the curves versus CypA added is drawn in Figure 5B. for 2 min was enough to achieve the equilibrium inhibition
Chymotrypsin was added to be excessive so that the isomer- and further incubation of CsA with CypA was unnecessary.
ization process was the rate-limiting step of the whole reaction, Dependence of residual PPIase activity on CsA concentration
which appeared to be first-order reaction kinetics. For the is shown in Figure 7. In the semilogarithmic plot, data were fit
isomerization, since the assay concentration of substrate was to a sigmoidal curve and IC50 was calculated to be 26.5 nM.
substantially lower than Michaelis-Menten constant Km, the Since the assay was not under typical Michaelis-Menten
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
306 Biotechnol. Prog., 2008, Vol. 24, No. 2

Figure 4. Peptide mass fingerprint of CypA mutant after trypsin treatment.

conditions (i.e., [E] , [S] or [I]), only an IC50 rather a Ki value probe. For the irreversible unfolding of CypA, these curves were
can be obtained. The observed IC50 was in agreement with not intrinsic equilibrium unfolding curves. However, the
previously reported 19 nM for the wild-type recombinant CypA sensitivity of CypA to the denaturants can be partially estimated
(18). from these curves. The denaturation process followed a two-
For dissociation constant determination, fluorescence measure state mechanism. GdmCl was more powerful in denaturing this
was also tried. However, no obvious fluorescence enhancement CypA mutant, less than 0.5 M GdmCl yielded half loss of the
was observed after CsA binding. Former research reported that PPIase activity and the activity was totally lost in the presence
a strong enhancement of intrinsic tryptophan fluorescence was of nearly 1 M GdmCl. This finding supposed that in assisting
generated by CypA-CsA complex (8, 18). The results indicated the folding of GdmCl-denatured protein, it might be troublesome
that the hydrophobic environment of tryptophan Trp121 residue to control the concentration of GdmCl. More thermodynamic
in this mutant changed little consequent to CsA binding. In stability was found in the presence of urea denaturant. Nonethe-
addition, the tyrosine residue Tyr48 in CypA structure should less, the susceptibility to urea was increased compared with
also be responsible for that because it was close to the mutation previously reported urea-denaturation curve of wild-type CypA.
site. In this experiment, 0.5-1.0 M urea was less needed to achieve
Refolding of Ribonuclease A. Bovine pancreatic ribonu- the same fraction of activity loss both at pH 6.0 and pH 7.0 at
clease A was previously reported to undergo proline isomer- same denaturation conditions strictly reserved (21). Urea
ization during refolding in vitro (25). Here, bovine pancreatic denaturation curves also showed that this CypA mutant was
ribonuclease A was used to examine the acceleration of protein less stable at pH 6.0 than at pH 7.0.
folding in vitro in the presence of CypA as a refolding aid. Autocatalyzing Refolding of CypA. Spontaneous refolding
From Figure 8, an obvious positive effect was observed in this of wild-type CypA occurred to an activity yield of about 30%
experiment. Because the mixing dead time elapsed several and was finished in less than 1 min (23). However, the Pro16
seconds, any phases with short lifespans cannot be detected in f Ser16 mutant displayed different refolding properties. The
this fluorescence spectrum. Figure 8 indicated that during the autocatalyzing refolding kinetics of recombinant CypA mutant
initial 250 s of the ribonuclease A refolding in the presence is shown in Figure 10, where fluorescence was used as the probe
and absence of CypA, respectively, the signals were nearly of structural information. The fluorescence intensity took about
identical and the amplitude covered nearly 90% of the total 300-400 s to become stable.
fluorescence change, whereas in the later phase, CypA notably The activity recovery of refolded CypA after equilibrium was
accelerated the folding of ribonuclease A. We concluded that consistent with the wild-type, in the range of 30-40%. Previous
the slow 10% fraction of ribonuclease A experienced proline results indicated the low recovery was due to aggregation of
isomerization and was catalyzed by the CypA mutant. CypA in the refolding. To understand the refolding pathway, a
Determination of Urea and GdmCl Denaturation Curve. native-PAGE was carried out to detect whether the aggregation
Both urea and GdmCl denaturation curves are presented in decreased the activity recovery. Finally, in the refolded CypA
Figure 9. The residual activity of CypA in the presence of mutant sample, only a small fraction of molecules fell into an
varying concentration of denaturants was used as the structural aggregation of dimers while most of the molecules presented a
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Biotechnol. Prog., 2008, Vol. 24, No. 2 307

Figure 7. Concentration dependence of PPIase activity of the CypA

mutant on cyclosporine A inhibitor.

Figure 5. (A) Chymotrypsin-coupled PPIase activity assay for different

concentrations of CypA mutant. (B) Variation of rate constant k with Figure 8. Refolding of ribonuclease A in vitro in the presence and
concentration of CypA mutant. absence of CypA monitored by fluorescence intensity change.

Figure 6. Eyring plot for catalyzed and uncatalyzed isomerization

reactions of tetrapeptide Suc-Ala-Xaa-Pro-Phe-p-nitroanilide. Figure 9. Denaturation curve of CypA mutant by urea and GdmCl
using PPIase activity as a probe.
band around 18 kDa. Thus, compared with wild-type CypA, it
might be speculated that the Pro16 f Ser16 mutation changed (FK 506 binding protein) (26) and parvulin (27) were reported
the folding/assembly pathway and refolding intermediates fell to display autocatalytic folding property. It was not easy to
into a kinetic trap in the refolding process under the observed establish the same properties for CypA protein because of the
time course and reduced aggregation. irreversible unfolding and folding. However, kinetic analysis
PPIase are folding enzymes and thus have the potential to concerning the fluorescence change is still presented here for
catalyze their self- folding. Other two prolyl isomerases FKBP12 further research. We fit the fluorescence refolding curves into
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
308 Biotechnol. Prog., 2008, Vol. 24, No. 2

Figure 11. Three-dimensional structure of human CypA complex with

tetrapeptide substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide (plotted with
Molsoft ICM software).

rate of enzyme renaturation was observed (31). Pro30 f Ser30

substitution in a fixed sequence of Drosophila and rat cyto-
chrome c imposed an effect on the local destabilization of the
protein structure rather than on the overall protein stability (32).
Also, the Pro97 f Ser97 mutant of nucleoside diphosphate
kinase of Drosophila melanogaster affected only the stability
but not catalytic efficiency of the enzyme (29). Proline
substituted for other residues also brings about new properties
for native enzymes. E. coli aspartate transcarbamoylase with
the Pro286 replaced for Ala exhibited a 40-fold reduction in
enzyme activity and weakened substrate affinity toward car-
bamoyl phosphate and aspartate (33). For yeast phosphogly
cerate kinase, although the secondary and tertiary structure of
the Pro204 f His204 protein is generally similar to that of the
wild-type enzyme, the enzyme is less stable to heat and
guanidine chloride denaturation (34). There were also the
opposite cases: mutation of proline 207 in mitochondrial
creatine kinase to alanine leads to increased acid stability (35).
Figure 10. (A) Concentration dependence of CypA mutant autocata- A wealth of information on the three-dimensional structure
lyzing refolding monitored by fluorescence intensity change. (B) is available for both the unligated recombinant human CypA
Semilogarithmic plot of concentration dependence of CypA mutant (36) and CypA-substrate complex such as CypA with a proline-
autocatalyzing folding monitored by fluorescence intensity change. containing tetrapeptide and a dipeptide Ala-Pro (37), CypA
binding with CsA or its derivatives (38), such as the most
a two-phase process (At ) A∞ + ∑n1 Ai exp(t/τi), n ) 2) (28), commonly used substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide
and the correlation coefficient can achieve to 0.996. When the (39) and amino-terminal domain of HIV-1 capsid protein (40).
refolding concentration was 1 or 2 µM, the magnitudes of two These materials provide us with insight into the structure of
phases were more or less the same: about 55% and 45% for CypA and mechanism of the catalysis. The ribbon diagram of
the slow and fast phases, respectively. For refolding rate constant wild-type CypA complex with Suc-Ala-Ala-Pro-Phe-p-nitroa-
determination, urea concentration was adjusted to be constant nilide is shown in Figure 11, which was drawn using Molsoft
so that the impact of denaturant can be ignored. The results ICM software with original data from RCSB Protein Data Bank.
indicated that basically the characteristic rate constant for the The sites of substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide, pro-
fast phase at different protein concentrations was around 0.0448 line 16 and tryptophan 121 are highlighted in the diagram using
s-1. Concerning the slow phase, increase of concentration of bat model. Residues of CypA involved in the interaction with
the folding catalyst also resulted in the increase of the folding substrate Suc-Ala-Ala-Pro-Phe-p-nitroanilide are Arg55, Ile57,
rate constant from 0.0080 to 0.0145 s-1 when the CypA Phe60, Met61, Gln63, Ala101, Asn102, Gln111, Phe113,
concentration doubled. Trp121, Leu122 and His126 (39). The crystal structure supported
the mechanism of the “protonation on nitrogen” and pinpointed
Discussion the catalytic group as Arg55. This general acid catalysis was
Several examples of proline to serine substitutions in the not in conflict with the observed pH independence of the cis f
primary structure of proteins are available (29). Generally this trans activity and consistent with the observation that mutation
substitution does not affect the protein free energy change, but of Arg55 to Ala retained less than 1% of the wild-type catalytic
native structure becomes more susceptible toward denaturants activity (19).
considering the stabilization of the denatured form of protein From Figure 11, it can be seen that Pro16 in the structure of
as the result of entropic factors (30). For example, the Pro87 wild-type CypA was far away from the catalytic center of the
f Ser87 mutant of E. coli adenylate kinase had a diminution PPIase. However, Pro16 occupied the crosspoint of two long
in catalytic efficiency but a large increase sensitivity to strands on the opposite of the small globular protein. Substitution
proteolysis by trypsin; however, only a marginal effect on the with Ser16 might substantially increase the flexibility of
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Biotechnol. Prog., 2008, Vol. 24, No. 2 309

microenvironment: reconstitution of the two strands and (2) Eyles, S. J.; Gierasch, L. M. Multiple roles of prolyl residues in
neighboring secondary structure. Moreover, the other ends of structure and folding. J. Mol. Biol. 2000, 301, 737-747.
the two strands were located in the core of the globular protein (3) Su, D.; Berndt, C.; Fomenko, D. E.; Holmgren, A.; Gladyshev, V.
N. A conserved cis-proline precludes metal binding by the active
and indirectly connected to the hydrophobic pocket in the surface site thiolates in members of the thioredoxin family of proteins.
of wild-type CypA. Urea denaturation curves validated the Biochemistry 2007, 46, 6903-6910.
supposition that fluctuation was increased and structural stability (4) Brandts, J. F.; Halvorson, H. R.; Brennan, M. Consideration of
was decreased. Another notable phenomenon was the fluores- the possibility that the slow step in protein denaturation reactions is
cence property of CypA mutant Pro16 f Ser16. The expected due to cis-trans isomerism of proline residues. Biochemistry 1975,
strong enhancement of intrinsic tryptophan fluorescence was 14, 4953-4963.
(5) Dobson, C. M. Protein folding and misfolding. Nature 2003, 426,
not observed and the fluorescence change induced by urea and
884-890.
GdmCl fell into a confused trend. What should be noted was (6) Wang, P.; Heitman, J. The cyclophilins. Genome Biol. 2005, 6,
that Trp121 was located in the binding pocket of Suc-Ala-Ala- 226.
Pro-Phe-p-nitroanilide and can be closely related to the catalytic (7) Fischer, G.; Bang, H.; Mech, C. Determination of enzymatic
efficiency. Taking into account of the results of decrease in catalysis for the cis-trans-isomerization of peptide binding in proline-
catalytic efficiency toward the tetrapeptide, holding CsA-binding containing peptides. Biomed. Biochim. Acta 1984, 43, 1101-1111.
property, increased susceptibility to urea and great sensitivity (8) Handschumacher, R. E.; Harding, M. W.; Rice, J.; Drugge, R. J.;
Speicher, D. W. Cyclophilin: a specific cytosolic binding protein
to GdmCl, together with fluorescence property change and for cyclosporin A. Science 1984, 226, 544-547.
presumably altered folding pathway, it was concluded that Pro16 (9) Fischer, G.; Wittmann-Liebold, B.; Lang, K.; Kiefhaber, T.; Schmid,
might play an important role in maintaining the structural and F. X. Cyclophilin and peptidyl-prolyl cis-trans isomerase are probably
functional integrity. identical proteins. Nature 1989, 337, 476-478.
It is very interesting to investigate the impact of a proline (10) Takahashi, N.; Hayano, T.; Suzuki, M. Peptidyl-prolyl cis-trans
residue in a peptidyl-prolyl cis-trans isomerase on its folding isomerase is the cyclosporin A-binding protein cyclophilin. Nature
1989, 337, 473-475.
properties. Proline isomerization was now firmly established (11) Gothel, S. F.; Marahiel, M. A. Peptidyl-prolyl cis-trans isomerases,
as a rate-limiting step in protein folding, and PPIase was a a superfamily of ubiquitous folding catalysts. Cell. Mol. Life Sci.
catalyst to accelerate proline isomerization in protein folding 1999, 55, 423-436.
(5). Coupling of these two aspects in the commonly studied (12) Lang, K.; Schmid, F. X.; Fischer, G. Catalysis of protein folding
PPIase generated new finding in CypA folding. Although by prolyl isomerase. Nature 1987, 329, 268-270.
refolding yield was still not high compared with the folding of (13) Pappenberger, G.; Bachmann, A.; Muller, R.; Aygun, H.; Engels,
J. W.; Kiefhaber, T. Kinetic mechanism and catalysis of a native-
wild-type CypA, the very problem was not the expected
state prolyl isomerization reaction. J. Mol. Biol. 2003, 326, 235-
aggregation of folding intermediates and lack of chaperone 246.
assistance (21, 23). Native-PAGE revealed that only quite a (14) Kiefhaber, T.; Quaas, R.; Hahn, U.; Schmid, F. X. Folding of
small fraction of the mutants formed dimers, compared with ribonuclease T1. 1. Existence of multiple unfolded states created
the ratio of activity loss. Many monomeric polypeptides fell by proline isomerization. Biochemistry 1990, 29, 3053-3061.
into the folding trap and failed to form active conformation, (15) Huang, G. C.; Zhou, J. M. The two slow refolding processes of
hence the folding/assembly pathway has been changed. Fluo- creatine kinase are catalyzed by cyclophilin. J. Protein Chem. 2000,
19, 285-289.
rescence intensity change indicated that previously reported 1 (16) Bosco, D. A.; Eisenmesser, E. Z.; Pochapsky, S.; Sundquist, W.
min refolding time for wild-type CypA was not enough for this I.; Kern, D. Catalysis of cis/trans isomerization in native HIV-1
Pro16 f Ser16 mutant to be refolded. Two-phase kinetics capsid by human cyclophilin A. Proc. Natl. Acad. Sci. U.S.A. 2002,
applied well to the folding process for this CypA mutant. Fast- 99, 5247-5252.
folding and slow-folding forms of unfolded proteins were both (17) Hur, S.; Bruice, T. C. The mechanism of cis-trans isomerization
observed (28). Interestingly, characteristic time constants for of prolyl peptides by cyclophilin. J. Am. Chem. Soc. 2002, 124,
7303-7313.
the fast phase remained constant in the CypA mutant folding (18) Liu, J.; Albers, M. W.; Chen, C. M.; Schreiber, S. L.; Walsh, C.
process with different protein concentrations, while the slow T. Cloning, expression, and purification of human cyclophilin in
phase was accelerated by increase of the PPIase concentration. Escherichia coli and assessment of the catalytic role of cysteines by
This explanation could be resorted to autocatalysis or self- site-directed mutagenesis. Proc. Natl. Acad. Sci. U.S.A. 1990, 87,
catalysis (26, 27), but a definite elucidation was still unclear 2304-2308.
for the complication of the low yield of activity recovery. Except (19) Zydowsky, L. D.; Etzkorn, F. A.; Chang, H. Y.; Ferguson, S. B.;
Stolz, L. A.; Ho, S. I.; Walsh, C. T. Active site mutants of human
for Pro16 and Pro4 (which was supposed to have less influence
cyclophilin A separate peptidyl-prolyl isomerase activity from
on the whole structure), there were still four proline residues in cyclosporin A binding and calcineurin inhibition. Protein Sci. 1992,
the mutant: Pro30, Pro58, Pro95, Pro105. These proline residues 1, 1092-1099.
were all presented in the conjunction or turn of the secondary (20) Kofron, J. L.; Kuzmic, P.; Kishore, V.; Colon-Bonilla, E.; Rich,
structure, so attribution of the observed fast- and slow-folding D. H. Determination of kinetic constants for peptidyl prolyl cis-
species to one or more proline isomerization required more trans isomerases by an improved spectrophotometric assay. Bio-
detailed investigation: mutagenesis or proteolytic analysis might chemistry 1991, 30, 6127-6134.
(21) Zahn, R.; Spitzfaden, C.; Ottiger, M.; Wuthrich, K.; Pluckthun,
contribute more insight. A. Destabilization of the complete protein secondary structure on
Acknowledgment binding to the chaperone GroEL. Nature 1994, 368, 261-265.
(22) Pace, C. N. Determination and analysis of urea and guanidine
This work was financially supported by the National Natural hydrochloride denaturation curves. Methods Enzymol. 1986, 131,
Science Foundation of China (no. 20476093). We also thank 266-280.
Professor Fersht in Centre for Protein Engineering MRC, (23) Zahn, R.; Buckle, A.; Perrett, S.; Johnson, C.; Corrales, F.; Golbik,
Cambridge very much for the gift of the CypA plasmid R.; Fersht, A. Chaperone activity and structure of monomeric
pAVD30. polypeptide binding domains of GroEL. Proc. Natl. Acad. Sci. U.S.A.
1996, 93, 15024-15029.
References and Notes (24) Harrison, R. K.; Stein, R. L. Mechanistic studies of peptidyl prolyl
(1) Galat, A.; Metcalfe, S. M. Peptidylproline cis/trans isomerases. cis-trans isomerase: evidence for catalysis by distortion. Biochem-
Prog. Biophys. Mol. Biol. 1995, 63, 67-118. istry 1990, 29, 1684-1689.
 15206033, 2008, 2, Downloaded from https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070259m by Indian Institution Of Chem Biology, Wiley Online Library on [01/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
310 Biotechnol. Prog., 2008, Vol. 24, No. 2

(25) Lin, L. N.; Hasumi, H.; Brandts, J. F. Catalysis of proline (34) McHarg, J.; Kelly, S. M.; Price, N. C.; Cooper, A.; Littlechild, J.
isomerization during protein-folding reactions. Biochim. Biophys. A. Site-directed mutagenesis of proline 204 in the ‘hinge’ region of
Acta 1988, 956, 256-266. yeast phosphoglycerate kinase. Eur. J. Biochem. 1999, 259, 939-
(26) Scholz, C.; Zarnt, T.; Kern, G.; Lang, K.; Burtscher, H.; Fischer, 945.
G.; Schmid, F. X. Autocatalytic folding of the folding catalyst (35) Forstner, M.; Muller, A.; Rognan, D.; Kriechbaum, M.; Walli-
FKBP12. J. Biol. Chem. 1996, 271, 12703-12707. mann, T. Mutation of cis-proline 207 in mitochondrial creatine kinase
(27) Scholz, C.; Rahfeld, J.; Fischer, G.; Schmid, F. X. Catalysis of to alanine leads to increased acid stability. Protein Eng. 1998, 11,
protein folding by parvulin. J. Mol. Biol. 1997, 273, 752-762. 563-568.
(28) Schmid, F. X. Fast-folding and slow-folding forms of unfolded
proteins. Methods Enzymol. 1986, 131, 70-82. (36) Ke, H.; Zydowsky, L. D.; Liu, J.; Walsh, C. T. Crystal structure
(29) Lascu, I.; Chaffotte, A.; Limbourg-Bouchon, B.; Veron, M. A of recombinant human T-cell cyclophilin A at 2.5 A resolution. Proc.
Pro/Ser substitution in nucleoside diphosphate kinase of Drosophila Natl. Acad. Sci. U.S.A. 1991, 88, 9483-9487.
melanogaster (mutation killer of prune) affects stability but not (37) Ke, H.; Mayrose, D.; Cao, W. Crystal structure of cyclophilin A
catalytic efficiency of the enzyme. J. Biol. Chem. 1992, 267, 12775- complexed with substrate Ala-Pro suggests a solvent-assisted mech-
12781. anism of cis-trans isomerization. Proc. Natl. Acad. Sci. U.S.A. 1993,
(30) Lin, T. Y.; Kim, P. S. Evaluating the effects of a single amino 90, 3324-3328.
acid substitution on both the native and denatured states of a protein. (38) Kallen, J.; Mikol, V.; Taylor, P.; Walkinshaw, D. M. X-ray
Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 10573-10577. structures and analysis of 11 cyclosporin derivatives complexed with
(31) Gilles, A. M.; Saint-Girons, I.; Monnot, M.; Fermandjian, S.; cyclophilin A. J. Mol. Biol. 1998, 283, 435-449.
Michelson, S.; Barzu, O. Substitution of a serine residue for proline-
87 reduces catalytic activity and increases susceptibility to proteolysis (39) Zhao, Y.; Ke, H. Crystal structure implies that cyclophilin
of Escherichia coli adenylate kinase. Proc. Natl. Acad. Sci. U.S.A. predominantly catalyzes the trans to cis isomerization. Biochemistry
1986, 83, 5798-5802. 1996, 35, 7356-7361.
(32) Koshy, T. I.; Luntz, T. L.; Schejter, A.; Margoliash, E. Changing (40) Gamble, T. R.; Vajdos, F. F.; Yoo, S.; Worthylake, D. K.;
the invariant proline-30 of rat and Drosophila melanogaster cyto- Houseweart, M.; Sundquist, W. I.; Hill, C. P. Crystal structure of
chromes c to alanine or valine destabilizes the heme crevice more human cyclophilin A bound to the amino-terminal domain of HIV-1
than the overall conformation. Proc. Natl. Acad. Sci. U.S.A. 1990, capsid. Cell 1996, 87, 1285-1294.
87, 8697-8701.
(33) Jin, L.; Stec, B.; Kantrowitz, E. R. A cis-proline to alanine mutant Received August 1, 2007. Accepted January 30, 2008.
of E. coli aspartate transcarbamoylase: kinetic studies and three-
dimensional crystal structures. Biochemistry 2000, 39, 8058-8066. BP070259M