Original Paper |
|
|||
|
Acta Biochim Biophys |
||||
|
doi:10.1111/j.1745-7270.2006.00199.x |
Design of a Novel Plasminogen
Activator Based on the Structure of Hirudin
Yu-Gao Zhang1,
Lu Yue2,
Yu-Xiong Wang1,
Xian-Mei Tao1, and Hou-Yan Song1*
1
The Key Laboratory of Molecular Medicine, Ministry of Education, Fudan
University, Shanghai 200032, China;
2 Department
of Oncology, Medical Hospital of Qingdao University, Qingdao 266003, China
Received: March 19,
2006
Accepted: May 23,
2006
*Corresponding author: Tel/Fax, 86-21-64033738; E-mail,
[email protected]
Abstract Using a phage library, seven peptide sequences with high
affinity to human microplasminogen were obtained. Caseinolytic assay indicated
only the synthesized peptide P07 had slight fibrinolytic activity. To enhance
its plasminogen activation ability, peptide P07 was fused into loop 32–35 of hirudin. In vitro assay
demonstrated that this hirudin-like fusion protein can activate human
plasminogen and retain the function of thrombin inhibition. Fusing the sequence
“SPDASRL” into hirudin generated a plasminogen activation activity
100 times higher than peptide P07 in chromogenic and radial caseinolytic assay.
This significant functional improvement might originate from a more specific
active structure due to the hirudin scaffold.
Key words plasminogen activator; hirudin; fusion protein;
thrombolytic agent; anticoagulant
Activation of plasminogen (Plgn; EC 3.4.21.7) is a key event in the
fibrinolytic system that results in the dissolution of blood clots in a
fibrin-dependent manner, and also promotes cell migration and tissue
remodelling. The zymogen activation of Plgn results in the serine protease
plasmin (Plm), which consists of five kringle domains at the amino-terminus and
a serine protease domain in the carboxyl-terminus [1]. Human microplasminogen (mPlgn) is a
single polypeptide of 261 residues from the carboxyl-terminal portion of native
Plgn [2] and still retains the fibrinolytic function of the latter. However,
like Plgn, mPlgn is the inactive precursor of the fibrinolytic enzyme
microplasmin, and it can be activated by tissue-specific plasminogen activator
(t-PA), vampire bat t-PA, urokinase, and two bacterial protein co-factors,
streptokinase (SK) and staphylokinase (SAK) [3–5].
SK and SAK are not enzymes themselves, but form 1:1 stoichiometric
complexes with Plgn and Plm, which acquire a remarkable specificity and
efficiency to activate Plgn. After complex formation the specificity switches,
whereas plasmin shows a preference for “extended” substrates like
fibrin(ogen), the SK-Plgn and SAK-Plm complexes have activities against
“narrower” substrates, such as the activation loop of Plgn [5,6].
This binding-activation mechanism of co-factors SK and SAK provides clues for
new plasminogen activator (PA) design.
Using a phage displayed random peptide library, we screened out a
specific peptide with high affinity to mPlgn. However, amidolytic
assay showed no PA activity for the synthesized peptide, whereas the M13
strain, having the same epitope, does have this ability. To restrict the
conformation of this peptide, we constructed a fusion protein by inserting this
sequence into loop 32–35 of hirudin. In chromogenic assay, the fusion protein (designated
8067) demonstrated PA activity, which was further proved by radial caseinolytic
assay.
Materials and Methods
Materials
Bacterial strains of Escherichia coli DH5a and ER2738,
wild-type M13 phage, yeast strain of Pichia pastoris and pPIC9K were
stored in our laboratory. All the restriction endonucleases, T4 DNA ligase and Pfu
polymerase were purchased from New England Biolabs (Ipswich, USA). IPTG, X-gal,
tetracycline and primers were obtained from Sangon Biotech (Shanghai, China).
Recombinant SAK [7], chromogenic substrate S-2390 and Plgn were prepared in our
laboratory. Sephacryl S-100 and Lys-Sepharose were obtained from Pharmacia
Biotech (Uppsala, Sweden). Peptides were synthesized by GL Biochem (Shanghai,
China). All peptides were purified with HPLC to more than 90% purity. Disulfide
bonds after oxidation, together with the free amino and acid group at the N-
and C-termini, were verified with mass spectrometry. For example, peptide P07
was prepared with a purity of 92.2%. The sequence of P07 is GGSACSPDASRLCGGSAE,
with several flanking residues at both the N- and C-termini to increase its
stability and solubility. These flanking residues were derived from the
Ph.D-C7C peptide library (New England Biolabs). Human Plgn was isolated from
fresh frozen plasma using Lys-Sepharose affinity chromatography following the
established procedures [8,9]. All other reagents were of analytic purity.
Cloning, expression and
purification of mPlgn and fusion protein
The structural gene of mPlgn was retrieved from the genomic DNA of Homo
sapiens by polymerase chain reaction amplification and cloned into yeast
expression vector pPIC9K [10]. The full-length gene of hirudin-like fusion
protein (8067) was synthesized (Sangon Biotech) then cloned into pPIC9K (Fig.
1). Both mPlgn and hirudin-derived fusion protein were expressed in P.
pastoris. The supernatant of yeast fermentation was precipitated using 40%
ammonium sulfate solution and loaded onto a Sephacryl S-100 column using a low
velocity of flow (0.5 ml/min, 0.02 M phosphate buffer, pH 7.8) to avoid pigment
contamination, then purified by ion exchange chromatography (QFF) (Fig. 2).
Mass spectrometry verified the molecular weight of 8067 as 7286 Da. Purified
proteins were lyophilized and stored at –20 ºC.
Panning using epitope mapping
phage display
The 7 mer random peptide library (Ph.D-C7C peptide library kit) was
purchased from New England Biolabs. Using the purified mPlgn as the target protein,
the panning process was carried out according to the manual (Table 1).
Radial caseinolytic assay
Functional activity of fusion protein was also estimated by radial
caseinolytic assay. Petri dishes containing 1.2% agarose, 1% skim milk and 10 mg/ml Plgn were
prepared [11]. On this solidified agarose plate, wells of equal diameter were
bored and an equal quantity of protein was added and kept at 37 ºC for 6–7 h. The
diameter of the halo around the well was measured to check the fibrinolytic
activity.
Amidolytic assay
Amidolytic activity assay was carried out with small modifications.
Considering the relative low PA ability of the screened peptide, we used S2390
(D-Val-Phe-Lys-p-nitroanilide) as the chromogenic substrate instead of
S2251(D-Val-Leu-Lys-p-nitroanilide) [12]. S2390 has a lower Km value than the standard S2251, and it also has a much higher Vmax/Km value than most other substrates of this type
[13]. Plgn (1.25 mM) was incubated with sample moieties (25 mM) at 37 ºC in activation
buffer (10 mM phosphate buffer, pH 7.4, 0.01% Tween 80). S2390 was added to a final
concentration of 1 mM and initial velocity determinations for the enzymatic
activities were performed by monitoring the hydrolysis of the tripeptidyl-p-nitroanilide
at 405 nm. The absorbance change (DA405) was
monitored for up to 12 h using a spectrophotometer (Microplate Reader
Benchmark; Bio-Rad, Hercules, USA).
Determination of specific
activity of hirudin-like fusion protein by thrombin titration
Human a-thrombin was obtained from Huashan Hospital (Shanghai, China). The
recombinant hirudin variant 1 (rHV1) was prepared in our laboratory and
standardized with a specific activity of 10,000 ATU/mg [14]. The specific
activity of hirudin-like fusion protein 8067 was determined using a modified
thrombin titration method [15,16]. Briefly, 200 ml of 0.5% fibrinogen
solution was added to a polystyrene cuvette (path length=1 cm). Assay working
buffer was composed of 50 mM sodium chloride, 50 mM Tris-HCl, 0.1% BSA and 0.1%
PEG-6000, and was adjusted to pH 7.4. Fibrinogen was incubated at 37 ºC with 50
ml
of sample for 5 min. Then 5 ml of thrombin (20 NIH units/ml) was added, mixed and incubated at 37
ºC for 1 min with 8067 or rHV1. The addition of thrombin increased turbidity
because of the fibrin clot formation, and the absorbance was monitored by the
microplate reader at 405 nm.
Results
Panning of phage library
After two cycles of panning, the random library was enriched with
strains representing highly specific epitopes for mPlgn (Fig. 3).
Thirty blue phage plaques were picked out for sequencing, from which seven
different sequences were identified, as shown in Fig. 3. We named the
peptides P02, P07, P08, P16, P17, P21 and P22, with the suffix number
designating the serial number of the clone by which the sequence was
identified. Among the 30 picked blue plaques, three had the same sequence as
P07; six clones had different sequences as P02, P08, P16, P17, P21 and P22;
five clones had sequences partly unreadable; and the remaining 16 plaques were
identified as wild-type M13 phage (vanishingly small levels of contaminating
environmental wild-type phage has a growth advantage over the library phage).
These seven peptides were subsequently synthesized for function assay.
Radial caseinolytic assay
The seven synthesized peptides were assayed. However, at a final
concentration of 5 mM, only P07 showed a small lytic halo (Fig. 4). The
hirudin-like fusion protein 8067 showed little fibrinolytic activity at 10–20 mM, whereas 50 mM of 8067 led to
a clearly visible lytic circle (Fig. 5). The pattern of caseinolytic
assay indicated concentration-dependent activity of protein 8067 in the
fibrinolytic process.
Amidolytic activity assay
To examine the Plgn activation properties of hirudin-like fusion
proteins, the pattern of Plgn activation by protein 8067 was compared with that
of SAK, a specific and robust activator for mPlgn. Protein 8067
activated Plgn into Plm very slowly, to approximately 80% of Plgn activation in
approximately 12 h. Activation of Plgn in the presence of 8067 occurred
progressively with a prolonged lag phase followed by an exponential increase in
Plgn activation. The synthesized peptide P07 showed only marginal (approximately
8%) PA activity. In contrast, the catalytic amount of SAK (5 nM) induced rapid
activation of Plgn to Plm, resulting in more than 95% of Plgn activation within
6 min. Hirudin demonstrated no specific PA ability compared with the control
(PBS buffer) (Fig. 6). PBS control and rHV1 showed an almost identical
noise pattern. These results suggested that the ability of 8067 to activate
Plgn is much lower than SAK but significantly higher than P07.
Inhibition of thrombin by the hirudin-like
fusion protein 8067
To provide evidence that the fusion protein 8067 binds to thrombin
and is able to inhibit the activity of thrombin, we carried out the titration
assay of anti-thrombin activity (Fig. 7). Protein 8067 and rHV1 at
concentrations of 500 ng/ml and 100 ng/ml, respectively, were sufficient to
completely inhibit the clotting activity of thrombin. Our results showed that
the inhibition of thrombin catalytic activity by 8067 (2000 ATU/mg) was
impaired markedly compared to that of rHV1 (10,000 ATU/mg).
Discussion
The fibrinolysis mechanism is mainly responsible for the maintenance
of the potency of blood vessels during the elimination of intravascular clots. Plasminogen
activators play a key role in fibrinolysis. Current clinically approved
thrombolytic agents have significant drawbacks, including re-occlusion and
bleeding complications. Recently, attempts have been made to obtain hybrid
proteins on the basis of known PA such as t-PA, SAK and thrombin inhibitors
like hirudin [18–24]. They consist of hybrid fusion proteins from active compounds,
or fragments which are necessary for hemostasis. However, simple incorporation
of functional domains into one molecule might raise problems such as
difficulties in gene engineering and expression, incorrect protein folding and
sophisticated interaction between different original fragments.
In this work we take advantage of peptide mimics to simplify the
structure of chimera molecules. The goal is to obtain effective thrombolytic
and antithrombotic protein and simplify the way the agent is given, to restore
vessel potency.
To obtain active peptide mimics of PA, we used a random peptide
library that was conformationally restricted by a disulfide bond (formed by two
cysteines at both termini of the random sequence region). We had previously
screened a linear peptide library, which led to more unified sequences, but
had less potential to improve its PA activity. Cyclic peptide has a more steady
structure and possibly favors its interaction with target molecules. As the PA
activity of the primary peptide P07 was quite poor, we selected hirudin as a
scaffold for further molecular design.
Hirudin is a classic anticoagulant and very robust. Both fragments
derived from its C-terminal tail and N-terminal core have inhibition abilities
to thrombin [25,26]. Loop 32–35 of rHV1 is relatively distant from the active region, as
mentioned above, making it an ideal site for engineering. In addition, the
neighboring b-strands are tightly strained, enabling this loop a typical b-turn motif
[27], which is similar to the architecture of the cyclic peptide P07. With
these factors in mind, we introduced the sequence “SPDASRL” into loop
32–35
of rHV1 by replacing mutation. The disulfide bond of P07 was substituted by the
native structure of rHV1 for motif configuration (Fig. 8).
Fusion of “SPDASRL” into rHV1 generated a PA activity
100 times higher than P07 in chromogenic and radial caseinolytic assays. This
significant functional improvement might originate from the more specific
active structure due to the rHV1 scaffold. Although the fusion protein has a
high affinity motif for Plgn displayed at the molecular surface, protein 8067 is
believed to be a non-enzymatic PA, as our research originated from the
binding-activation theory. A mechanism is proposed for this enhancement that
involves the rendezvous and local concentration of Plgn. By complex formation
with trace Plm, the Plm-8067 complex activates Plgn and initiates the
enzyme-catalyzed reactions. This hypothesis is based on the Plgn activation
mechanism of SAK [28,29], which acts as a co-factor in the fibrinolytic
process. More evidence is needed to confirm the mechanism of Plgn activation
by 8067.
Anti-thrombin activity of 8067 was reduced remarkably compared with
rHV1, suggesting insertion of the sequence “SPDASRL” in loop 32–35 does impair
the interaction between hirudin and thrombin. The specific activity of 8067 was
approximately 20% of rHV1. In spite of this drawback, 8067 remains an effective
anticoagulant due to its potent parent form.
In conclusion, a functional epitope screen targeting mPlgn provided
clues for developing a peptide agonist of Plgn. Our finding revealed a new PA
or co-activator, although the mechanism needs further research. By grafting
the peptide in hirudin, we developed a new agent with evident thrombolytic and
anticoagulant properties. Ongoing studies on the fusion protein will provide
further details about its functional mechanism and therapeutic potential.
References
1 Collen D, Van Hoef B, Schlott B, Hartmann M,
Guhrs KH, Lijnen HR. Mechanisms of activation of mammalian plasma fibrinolytic systems
with streptokinase and with recombinant staphylokinase. Eur J Biochem 1993,
216: 307–314
2 Yecies LD, Kaplan AP. Partial
characterization of a low molecular weight fragment derived from human
plasminogen. Thromb Res 1979, 14: 729–738
3 Collen D. Thrombolytic therapy. Thromb
Haemost 1997, 78: 742–746
4 Collen D, Lijnen HR. Thrombolytic agents.
Thromb Haemost 2005, 93: 627–630
5 Collen D. Staphylokinase: A potent, uniquely
fibrin-selective thrombolytic agent. Nat Med 1998, 4: 279–284
6 Gonzalez-Gronow M, Siefring GE, Castellino
FJ. Mechanism of activation of human plasminogen by the activator complex,
streptokinase-plasmin. J Biol Chem 1978, 253: 1090–1094
7 Tang QQ, Zhang XX, Yu M, Song HY. Isolation,
purification and crystallization of recombinant staphylokinase (r-SAK).
Pharmacia Biotech 1997, 4: 1–4
8 Deutsch DG, Mertz ET. Plasminogen:
Purification from human plasma by affinity chromatography. Science 1970, 170:
1095–1096
9 Mannhalter C. Purification of plasma protein.
Haemostasis 1988, 18: 115–119
10 Song G, Guan XQ, Song HY. Expression of
microplasminogen in methylotrophic yeast Pichia pastoris. Chin J Biotech
1999, 15: 211–214
11 Hauptmann J, Guhrs KH, Hartmann M, Schlott B.
The fibrinolytic activity of staphylokinase mutants in the fibrin plate assay.
Haemostasis 1995, 25: 272–276
12 Sharma SK, Castellino FJ. The chemical
synthesis of the chromogenic substrates, H-D-Val-L-Leu-L-Lys-p-nitroanilide
(S2251) and H-D-Ile-L-Pro-L-ARG-p-nitroanilide (S2288). Thromb Res 1990,
57: 127–138
13 Schnyder J, Marti R, Cooper PH, Payne TG.
Spectrophotometric method to quantify and discriminate urokinase and
tissue-type plasminogen activators. Anal Biochem 1992, 200: 156–162
14 Liu X, Mo W, Dai L, Yan X, Song H. Structure
study of recombinant RGD-hirudin by vibrational and circular dichroism
spectroscopy. Protein Pept Lett 2006, 13: 47–51
15 Markwardt F, Sturzebecher J, Walsmann P. The
hirudin standard. Thromb Res 1990, 59: 395–400
16 Markwardt F. The development of hirudin as an
antithrombotic drug. Thromb Res 1994, 74: 1–23
17 Thompson JD, Gibson TJ, Plewniak F, Jeanmougin
F, Higgins DG. The ClustalX windows interface: flexible strategies for multiple
sequence alignment aided by quality analysis tools. Nucleic Acids Res 1997,
24:4876–4882
18 Icke C, Schlott B, Ohlenschlager O, Hartmann
M, Guhrs KH, Glusa E. Fusion proteins with anticoagulant and fibrinolytic
properties: Functional studies and structural considerations. Mol Pharmacol
2002, 62: 203–209
19 Lian Q, Szarka SJ, Ng KK, Wong SL. Engineering
of a staphylokinase-based fibrinolytic agent with antithrombotic activity and
targeting capability toward thrombin-rich fibrin and plasma clots. J Biol Chem
2003, 278: 26677–26686
20 Szemraj J, Walkowiak B, Kawecka I, Janiszewska
G, Buczko W, Bartkowiak J, Chabielska E. A new recombinant thrombolytic and
antithrombotic agent with higher fibrin affinity – a staphylokinase variant. I.
In vitro study. J Thromb Haemost 2005, 3: 2156–2165
21 van Zyl WB, Pretorius GH, Hartmann M, Kotze HF.
Production of a recombinant antithrombotic and fibrinolytic protein, PLATSAK,
in Escherichia coli. Thromb Res 1997, 88: 419–426
22 Lijnen HR, Wnendt S, Schneider J, Janocha E,
Van Hoef B, Collen D, Steffens GJ. Functional properties of a recombinant chimeric
protein with combined thrombin inhibitory and plasminogen-activating potential.
Eur J Biochem 1995, 234: 350–357
23 Wnendt S, Janocha E, Schneider J, Steffens GJ.
Construction and structure-activity relationships of chimeric prourokinase
derivatives with intrinsic thrombin-inhibitory potential. Protein Eng 1996, 9:
213–223
24 Wirsching F, Luge C, Schwienhorst A. Modular
design of a novel chimeric protein with combined thrombin inhibitory activity
and plasminogen-activating potential.
Mol Genet Metab 2002, 75: 250–259
25 Markwardt F. Hirudin as alternative
anticoagulant – a historical review. Semin Thromb Hemost 2002, 28: 405–414
26 Chang JY. Production, properties, and thrombin
inhibitory mechanism of hirudin amino-terminal core fragments. J Biol Chem
1990, 265: 22159–22166
27 Rydel TJ, Ravichandran KG, Tulinsky A, Bode W,
Huber R, Roitsch C, Fenton JW 2nd. The structure of a complex of recombinant
hirudin and human a-thrombin. Science 1990, 249: 277–280
28 Rabijns A, De Bondt HL, De Ranter C. Three-dimensional
structure of staphylokinase, a plasminogen activator with therapeutic
potential. Nat Struct Biol 1997, 4: 357–360
29 Lijnen HR, Van Hoef B, Collen D. Interaction
of staphylokinase with different molecular forms of plasminogen. Eur J Biochem
1993, 211: 91–97