Proc. Nati. Acad. Sci. USA Vol. 87, pp. 3530-3533, May 1990 Biochemistry
Amino acid residues that affect interaction of tissue-type plasminogen activator with plasminogen activator inhibitor 1 (enzyme kinetics/rate constant for inhibition/rate constant for association)
EDWIN L. MADISON*, ELIZABETH J. GOLDSMITH*, ROBERT D. GERARD*t, MARY-JANE H. GETHINGt, JOSEPH F. SAMBROOK*, AND RHONDA S. BASSEL-DUBYt Departments of *Biochemistry and tInternal Medicine, and tHoward Hughes Medical Institute, University of Texas Southwestern Medical Center, Dallas, TX 75235
Communicated by Bernard N. Fields, February 16, 1990
Fibrinolysis is regulated in part by the interABSTRACT action between tissue-type plasminogen activator (t-PA) and plasminogen activator inhibitor 1 (PAMi, a serine protease inhibitor of the serpin family). It is known from our earlier work that deletion of a loop of amino acids (residues 296-302) from the serine protease domain of t-PA suppresses the interaction between the two proteins without altering the reactivity of t-PA towards its substrate, plainogen. To derme more precisely the role of individual residues within this loop, we have used site-directed mutagenesis to replace Lys-296, Arg298, and Arg-299 with negatively charged glutamic residues. Replacement of all three positively charged amino acids generates a variant of t-PA that associates inefficiently with PAMI and is highly resistant to inhibition by the serpin. Two t-PAs with point mutations (Arg-298 -- Glu and Arg-299 -* Glu) are partially resistant to inhibition by PAM- and associate with the serpin at intermediate rates. Other point mutations (Lys-296-> Glu, His-297 -+ Glu, and Pro-301 -* Gly) do not detectably affect the interaction of t-PA with PAMI. None of these substitutions has a significant effect on the rate of catalysis by t-PA or on the affinity of the enzyme for its substrate, plasminogen. On the basis of these results, we propose a model in which positively charged residues located in a surface loop near the active site of t-PA form ionic bonds with complementary negatively charged residues C-terminal to the reactive center of PAMI.
+
K296 t-PA
,
1
TYR
SR39
SER
1.)ILE19
R299/
LS 15
i
.....
BPTI
FIG. 1. A model of the interaction between t-PA and PAM- based on the structure of the complex between bovine pancreatic trypsin and BPTI (Protein Databank file 2 ptc.dat). The region of interaction between trypsin and BPTI is displayed in INSIGHT (11). The a-carbon backbone of trypsin is represented by dotted lines while that of BPTI is shown by a solid line. The side chains of Tyr-39 of trypsin and Ile-19 of BPTI are shown to indicate the van der Waals contact between the residues in the trypsin-BPTI complex. The position of the t-PA seven amino acid insertion, residues 2%-302 (KHRRSPG in the standard single-letter symbols), is indicated schematically by a bold line and the positively charged residues in this insertion are labeled.
Tissue-type plasminogen activator (t-PA) is a serine protease that converts the zymogen plasminogen into the active enzyme plasmin. Plasmin degrades fibrin and consequently plays an important role in the dissolution of blood clots (1-3). In human plasma, circulating t-PA is present predominantly as a complex with a 50-kDa protein inhibitor, plasminogen activator inhibitor 1 (PAI-1), a member of the serpin family of serine protease inhibitors (4). t-PA interacts with PAT-i to form a 1:1 molar complex that is enzymatically inactive and stable to treatment with SDS (5). No crystallographic structure is available for an intact serpin or for a serpin-protease complex. However, the catalytic mechanism of t-PA is fundamentally similar to that of trypsin (6-8), and the interaction between trypsin and bovine pancreatic trypsin inhibitor (BPTI) is known in great detail (9). Because both BPTI and PAT-i bind to the active-site region of their target proteases they may interact directly with similar structural motifs (8). We have therefore used the known structure ofthe trypsin-BPTI complex to identify amino acid residues in the serine protease domain of t-PA that are predicted to make contact with PAl-i (10). By contrast to trypsin, t-PA contains an insertion of seven amino acids (residues 296-302) that are
predicted to form a surface loop adjacent to the active site of the enzyme. The basic and charged nature of the amino acids in this loop (Lys-His-Arg-Arg-Ser-Pro-Gly) suggests that electrostatic interactions could contribute to the interaction between t-PA and PA-IM. Deletion of the entire loop or substitution of Arg-304, which is predicted to be located at the edge of the active site of t-PA, yields variants that are resistant to inhibition by PAT-i and whose ability to catalyze the activation of plasminogen is essentially undiminished (10). The surface loop of t-PA (residues 2%-302) contains several positively charged amino acids, including a lysine and two arginine residues (see t-PA insertion in Fig. 1). The region of PAT-i that is predicted to make contact with the surface loop, residues 350-355, contains several negatively charged amino acids (Glu-Glu-Ile-Ile-Met-Asp). The interaction of t-PA and PA-IM may therefore be mediated by the formation of salt bridges between positively charged amino acids in t-PA and negatively charged residues in PAM-i (10).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviations: t-PA, tissue plasminogen activator; PAT-i, plasminogen activator inhibitor 1; BPTI, bovine pancreatic trypsin inhibitor.
3530
Biochemistry: Madison et al. In this report, we have used site-directed mutagenesis to assess the contribution of individual amino acids in the loop to the binding and recognition of PAI-1 by t-PA. Our results demonstrate that positively charged amino acids in the loop play an essential role in the formation of t-PA-PAI-1 complexes but are not involved in recognition of the substrate, plasminogen.
MATERIALS AND METHODS Site-Specific Mutagenesis and Expression of Mutated Enzymes. Oligonucleotide-directed site-specific mutagenesis was performed by the method of Zoller and Smith (12) as modified by Kunkel (13). Mutations were introduced into the 472-base-pair (bp) EcoRI fragment of wild-type t-PA cDNA that had been subcloned in bacteriophage M13 mpl8. The nomenclature assigned to the mutants and the oligonucleotides used for mutagenesis are: 5'-ATC1TTGCCGAGCACAGGA-3' K2%---E: 5'-1TTGCCAAGTACAGGAGGT-3' H297--Y: 5'-GCCAAGCACGAGAGGTCGCCC-3' R298--E: 5'-AAGCACAGGGAGTCGCCCGG-3' R299--*E: 5'-AGGAGGTCGGGCGGAGAGCG-3' P301--G: K2%,R298,R299 5'-GCCATC1TTGCCGAGCACGAGG>-)E,E,E: AGTCGCCCGGAGA-3' After mutagenesis the entire 472-bp EcoRI fragment was sequenced to verify the presence of the desired mutation. The mutated fragment was then recovered and used to reconstruct a full-length cDNA encoding the desired variant of t-PA. The full-length cDNAs were ligated either into the transient expression vector pSVT7(RI-) (10, 14, 15) or into a derivative of pSVT7 called pSTE. pSTE was constructed by replacement of the 350-bp Cla 1-HindIII promoter/origin fragment of pSVT7 with the 418-bp Hpa IT-HindIlI fragment spanning the promoter/origin region of simian virus 40 cs 1085 (16). One microgram of each of the resulting constructs was used to transfect COS-1 cells by the DEAE-dextran method (15). Twelve hours after the chloroquine boost, the medium (Dulbecco's modified Eagle's medium containing 10o fetal bovine serum) was replaced with medium lacking serum, and incubation was continued for a further 60 hr before the conditioned medium was collected. One-chain forms of wild-type t-PA and each of the variant proteins were purified by immunoaffinity chromatography on separate columns to avoid cross-contamination. Conditioned media from transfected COS-1 cells were loaded onto columns that contained rabbit anti-PAI-i IgG conjugated to Sepharose-4B (Pharmacia) and the flow-through was adsorbed onto mouse IgG1 monoclonal antibody to t-PA (designated 1A5) that also was conjugated to Sepharose-4B. Protein bound to the second column was eluted with 0.2 M glycine hydrochloride (pH 2.2). The eluates were neutralized with 2 M Tris HCl (pH 8), concentrated by using Centricon microconcentrators (Amicon), and equilibrated with Dulbecco's phosphate-buffered saline containing 0.01% Tween-80. The concentration oft-PA was measured by solid-phase radioimmunoassay as previously described (17). Enzyme Assays. Indirect chromogenic assays of t-PA were performed as described previously (10). Desafib (soluble fibrin), human [Lysiplasminogen, and Spectrozyme PL were purchased from American Diagnostica (Greenwich, CT). Inhibition of Activity. The rate constants for inhibition of wild-type or mutant t-PAs by PAT-i were measured as described by Beatty et al. (18) and Holmes et al. (19). Briefly, purified wild-type or mutant t-PAs (3-50 fmol) were incubated at 22°C for periods of time varying from 0 to 120 min
Proc. Natl. Acad. Sci. USA 87 (1990)
3531
with purified recombinant PAI-1 (35-1330 fmol) (20) that had been activated with guanidine hydrochloride (21). After this incubation, the mixtures were diluted, and the residual enzymatic activity was determined in a standard indirect chromogenic assay. Data were analyzed by plotting ln(residual activity/initial activity) versus time and determining the slope of the resulting straight line. Pseudo-first-order rate constants were then derived by dividing the slope by the concentration of the inhibitor in the reaction. Rate Constants for Association. The rate constants for the association of PAI-1 with wild-type and mutant t-PAs were measured by a modification of the immunoradiometric assay described by Hekman and Loskutoff (22). Purified wild-type or mutant t-PAs (100 pM) were mixed with purified PAI-1 (0.12-10 nM) that had been activated with guanidine hydrochloride (21). After incubation for 1-90 min at 37TC in phosphate-buffered saline containing 3% bovine serum albumin and 0.01% Tween-80, the reactions were terminated by the addition of 4-amidinophenylmethanesulfonyl fluoride (pAPMSF, Boehringer Mannheim) to a final concentration of 10 mM. Aliquots of the mixtures were then added to 96-well microtiter plates that had been coated with goat antibody to t-PA (American Diagnostica). After incubation for 60 min at 370C, the plates were washed with phosphate-buffered saline containing 0.1% bovine serum albumin and 0.01% Tween-80. Bound PAI-1 was detected with rabbit anti-PAI-i, followed by 1251-labeled donkey anti-rabbit IgG (Amersham). The amount of PAI-1 bound to the plate was calculated from a standard curve constructed from data obtained with known amounts of t-PA-PAI-1 complexes. Data were analyzed as previously described (22) by plotting log([EO][I]/[E][IT]) versus time and determining the slope of the resulting straight line. Apparent second-order kinetic association constants were then derived by multiplying the slope by 2.303 and dividing by ([EO] - [Io]). [EO] and [Io] are the initial concentrations of t-PA and PAI-1, respectively, and [E] and [I] are the concentrations of uncomplexed t-PA and PAI-1 in samples withdrawn from the reaction mixtures after incubation for different lengths of time. When PAI-1 was used in large molar excess, pseudo-first-order rate constants were derived as described above.
RESULTS Construction of t-PA Mutants. Previous data from our laboratory have suggested that a loop of seven amino acids, located on the surface of the catalytic domain of t-PA, is involved in the formation of enzymatically inactive complexes with the serpin PAI-1 (10). This surface loop (Fig. 1) contains several positively charged residues that could form electrostatic bonds with negatively charged amino acids in the contact region of PAI-1. To assess the importance of these positively charged amino acids, we have used sitedirected mutagenesis to construct variants of t-PA in which Lys-296, Arg-298, and Arg-299 have been replaced singly or collectively by negatively charged residues (Table 1). In addition, we have tested whether other amino acids in the loop interact specifically with residues in PAI-1, by making substitutions in which His-297 and Pro-301 are replaced with Tyr and Gly, respectively (Table 1). cDNAs encoding wildtype t-PA and each of these variants were ligated into transient expression vectors and used to transfect COS-1 cells. Wild-type or mutated t-PA proteins secreted into the medium were purified by immunoaffinity chromatography, quantitated by radioimmunoassay, and assayed for enzymatic activity as described below. Activity Assays. Indirect chromogenic assays of wild-type and variant t-PAs were performed in the presence of satulrating concentrations of the positive effector Desafib and various concentrations of the substrate [Lys]plasminogen. Analysis of the resulting data by the method of Lineweaver
3532
Biochemistry: Madison et al.
Proc. Natl. Acad. Sci. USA 87 (1990)
Table 1. Compairson of the primary sequence of wild-type and mutant t-PAs in theregion of the t-PA surface loop (residues
2%-302) Enzyme
Amino acid sequence K H R R S P G E R EH R R S P G E R K XR R S P G E R K HER S P G E R K H RES P G E R K H R R S fG E R EH E E S P G E R K H R R S P G E a K H R R S P G E E
F L306 t-PA 294F A F L F A K296-+E F L F A H297-+Y F L F A R298--E F L F A R299-*E F L F A P301--G F L K296,R298,R299--E,E,E F A F L F A *K304-S F L F A *R304-E F A - - - - - - - - - F L *A2%-302 Amino acid substitutions in the mutant t-PAs are underlined. *Mutants that have been previously described (10).
Active PAI-1 (tmol)
and Burk yielded the values for Km and kcat shown in Table 2. All mutants are enzymatically active and their specific activities vary less than 20% from the activity of standardized preparations of t-PA. The values of Km and kit for wild-type t-PA, the five substitution mutants, and the deletion mutant (A296-302) are similar to one another and to values published elsewhere by our laboratory (10, 17, 23) and others (24, 25). All of the mutants respond to the positive effector Desafib in a manner similar to the wild-type enzyme. The maximal stimulation was 20- to 40-fold (data not shown), in agreement with our previous results (10, 16, 23) and those of others (26). In each case, half-maximal stimulation occurred when Desafib was present at a concentration of approximately 1-2 ttg/ml. These data show that substitution of residues within the surface loop has not significantly altered the ability of t-PA to activate plasminogen or to be stimulated by soluble fibrin fragments. Inhibition of Enzymatic Activity by PAIl-. To test whether the enzymatic activity of the variant t-PAs was resistant to inhibition by PAIl-, the mutant and wild-type forms of the enzyme were incubated with various concentrations of purified recombinant PAIl-, and the residual enzymatic activity was then measured by using the indirect chromogenic assay. The results, presented in Fig. 2, show that the enzymes can be grouped into three categories: (i) Those that are similar to wild-type t-PA (K296-)E, H297-*Y, and P301-+G). These mutants are inhibited to the same extent as wild-type t-PA. (ii) Those that are partially inhibited by PA-IM (R298-*E and R299-*E). These mutants retain about 60% of their original activity after incubation with an amount of PA-IM that completely inhibits an equivalent amount of wild-type t-PA. (iii) Those that are highly resistant to inhibition by PA-IM
FIG. 2. Effect of PA-IM on the activity of wild-type and mutant t-PA enzymes. Equal amounts of wild-type and mutant t-PAs were incubated with various amounts of PAl-i (0-1250 fmol) for 20 min at room temperature. The residual activity of the enzymes was then determined by using the indirect chromogenic assay in the presence of 0.4 mM Spectrozyme PL, 0.16 ,uM [Lysiplasminogen, and soluble fibrin monomer, Desafib, at 25 ,g/ml. o, Wild-type t-PA; +, K296-* E; o, H297-*Y; A, R298-*E; A, R299-*E; m, K296,R298,R299-* E,E,E; and *, P301-*G. mutant A296-302 is highly resistant to inhibition by PAI-1 and belongs in class iii. Taken together, these data demonstrate that positively charged amino acids located in or immediately C-terminal to the charged loop of t-PA play a critical role in the inhibition of t-PA by PAIT-. Other residues in the same region (Arg-296, His-297, and Pro-301) are not essential for inhibition. However, more extensive mutagenesis will be required to determine whether any amino acid can be tolerated at these positions. Kinetic Analysis of the Interaction Between PAI-i and t-PA. The pseudo-first-order rate constants for the inhibition of the enzymatic activities of wild-type or mutant t-PAs by PAI-i were measured at 220C as previously described (18, 19). The inhibition of wild-type t-PA by PAIT- has a rate constant of 1.4 x 106 M-1 sol at 220C (Table 2). The rate constants obtained with mutant t-PAs belonging to class ii, which contain single amino acid substitutions (R304-*S, R304-*E, R298--E, and R299-*E), are approximately 4-fold, 61-fold, 64-fold, and 60-fold, respectively, less than the rate constant obtained with wild-type t-PA (Table 2). Mutants of class iii are far more severely affected: the rate constants for inhibition of the deletion mutant A296-302 and the triple mutant K296,R298,R299--E,E,E are reduced by factors of approximately 465 and 2800, respectively (Table 2). These values are the average of results obtained from five separate experiments, which showed less than 25% variation. The rates at which wild-type and mutant t-PA proteins enter into physical association with PAI-i were measured at
(K296,R298,R299--E,E,E). We have previously reported (10) that substitution mutants R304--S and R304--E, which alter the positively charged arginine residue that maps on the lip of the active site adjacent to the charged loop, are partially inhibited by PAI-i. These variants can therefore be placed in class ii. The t-PA deletion Table 2. Kinetic analysis of t-PA mutants
kat suMs-1
Km, Enzyme t-PA
R304--S R304-.E R298--E R299- E
A296-302 K296,R298,R299 --E,E,E
0.024 0.019 0.023 0.027 0.033 0.029 0.027
0.22 0.23 0.22 0.24 0.26 0.17 0.24
Inhibition rate constant,
M-1ls-1 (220C) 1.4 x 106 3.3 2.3 2.2 2.4 3.0 5.0
Association rate constant,
M-1ls-1 (370C) 3.5 8.0 3.9 1.4
105 104 104 104
Amino acid residues that affect interaction of tissue-type plasminogen activator with plasminogen activator inhibitor 1 (enzyme kinetics/rate constant for inhibition/rate constant for association)
EDWIN L. MADISON*, ELIZABETH J. GOLDSMITH*, ROBERT D. GERARD*t, MARY-JANE H. GETHINGt, JOSEPH F. SAMBROOK*, AND RHONDA S. BASSEL-DUBYt Departments of *Biochemistry and tInternal Medicine, and tHoward Hughes Medical Institute, University of Texas Southwestern Medical Center, Dallas, TX 75235
Communicated by Bernard N. Fields, February 16, 1990
Fibrinolysis is regulated in part by the interABSTRACT action between tissue-type plasminogen activator (t-PA) and plasminogen activator inhibitor 1 (PAMi, a serine protease inhibitor of the serpin family). It is known from our earlier work that deletion of a loop of amino acids (residues 296-302) from the serine protease domain of t-PA suppresses the interaction between the two proteins without altering the reactivity of t-PA towards its substrate, plainogen. To derme more precisely the role of individual residues within this loop, we have used site-directed mutagenesis to replace Lys-296, Arg298, and Arg-299 with negatively charged glutamic residues. Replacement of all three positively charged amino acids generates a variant of t-PA that associates inefficiently with PAMI and is highly resistant to inhibition by the serpin. Two t-PAs with point mutations (Arg-298 -- Glu and Arg-299 -* Glu) are partially resistant to inhibition by PAM- and associate with the serpin at intermediate rates. Other point mutations (Lys-296-> Glu, His-297 -+ Glu, and Pro-301 -* Gly) do not detectably affect the interaction of t-PA with PAMI. None of these substitutions has a significant effect on the rate of catalysis by t-PA or on the affinity of the enzyme for its substrate, plasminogen. On the basis of these results, we propose a model in which positively charged residues located in a surface loop near the active site of t-PA form ionic bonds with complementary negatively charged residues C-terminal to the reactive center of PAMI.
+
K296 t-PA
,
1
TYR
SR39
SER
1.)ILE19
R299/
LS 15
i
.....
BPTI
FIG. 1. A model of the interaction between t-PA and PAM- based on the structure of the complex between bovine pancreatic trypsin and BPTI (Protein Databank file 2 ptc.dat). The region of interaction between trypsin and BPTI is displayed in INSIGHT (11). The a-carbon backbone of trypsin is represented by dotted lines while that of BPTI is shown by a solid line. The side chains of Tyr-39 of trypsin and Ile-19 of BPTI are shown to indicate the van der Waals contact between the residues in the trypsin-BPTI complex. The position of the t-PA seven amino acid insertion, residues 2%-302 (KHRRSPG in the standard single-letter symbols), is indicated schematically by a bold line and the positively charged residues in this insertion are labeled.
Tissue-type plasminogen activator (t-PA) is a serine protease that converts the zymogen plasminogen into the active enzyme plasmin. Plasmin degrades fibrin and consequently plays an important role in the dissolution of blood clots (1-3). In human plasma, circulating t-PA is present predominantly as a complex with a 50-kDa protein inhibitor, plasminogen activator inhibitor 1 (PAI-1), a member of the serpin family of serine protease inhibitors (4). t-PA interacts with PAT-i to form a 1:1 molar complex that is enzymatically inactive and stable to treatment with SDS (5). No crystallographic structure is available for an intact serpin or for a serpin-protease complex. However, the catalytic mechanism of t-PA is fundamentally similar to that of trypsin (6-8), and the interaction between trypsin and bovine pancreatic trypsin inhibitor (BPTI) is known in great detail (9). Because both BPTI and PAT-i bind to the active-site region of their target proteases they may interact directly with similar structural motifs (8). We have therefore used the known structure ofthe trypsin-BPTI complex to identify amino acid residues in the serine protease domain of t-PA that are predicted to make contact with PAl-i (10). By contrast to trypsin, t-PA contains an insertion of seven amino acids (residues 296-302) that are
predicted to form a surface loop adjacent to the active site of the enzyme. The basic and charged nature of the amino acids in this loop (Lys-His-Arg-Arg-Ser-Pro-Gly) suggests that electrostatic interactions could contribute to the interaction between t-PA and PA-IM. Deletion of the entire loop or substitution of Arg-304, which is predicted to be located at the edge of the active site of t-PA, yields variants that are resistant to inhibition by PAT-i and whose ability to catalyze the activation of plasminogen is essentially undiminished (10). The surface loop of t-PA (residues 2%-302) contains several positively charged amino acids, including a lysine and two arginine residues (see t-PA insertion in Fig. 1). The region of PAT-i that is predicted to make contact with the surface loop, residues 350-355, contains several negatively charged amino acids (Glu-Glu-Ile-Ile-Met-Asp). The interaction of t-PA and PA-IM may therefore be mediated by the formation of salt bridges between positively charged amino acids in t-PA and negatively charged residues in PAM-i (10).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviations: t-PA, tissue plasminogen activator; PAT-i, plasminogen activator inhibitor 1; BPTI, bovine pancreatic trypsin inhibitor.
3530
Biochemistry: Madison et al. In this report, we have used site-directed mutagenesis to assess the contribution of individual amino acids in the loop to the binding and recognition of PAI-1 by t-PA. Our results demonstrate that positively charged amino acids in the loop play an essential role in the formation of t-PA-PAI-1 complexes but are not involved in recognition of the substrate, plasminogen.
MATERIALS AND METHODS Site-Specific Mutagenesis and Expression of Mutated Enzymes. Oligonucleotide-directed site-specific mutagenesis was performed by the method of Zoller and Smith (12) as modified by Kunkel (13). Mutations were introduced into the 472-base-pair (bp) EcoRI fragment of wild-type t-PA cDNA that had been subcloned in bacteriophage M13 mpl8. The nomenclature assigned to the mutants and the oligonucleotides used for mutagenesis are: 5'-ATC1TTGCCGAGCACAGGA-3' K2%---E: 5'-1TTGCCAAGTACAGGAGGT-3' H297--Y: 5'-GCCAAGCACGAGAGGTCGCCC-3' R298--E: 5'-AAGCACAGGGAGTCGCCCGG-3' R299--*E: 5'-AGGAGGTCGGGCGGAGAGCG-3' P301--G: K2%,R298,R299 5'-GCCATC1TTGCCGAGCACGAGG>-)E,E,E: AGTCGCCCGGAGA-3' After mutagenesis the entire 472-bp EcoRI fragment was sequenced to verify the presence of the desired mutation. The mutated fragment was then recovered and used to reconstruct a full-length cDNA encoding the desired variant of t-PA. The full-length cDNAs were ligated either into the transient expression vector pSVT7(RI-) (10, 14, 15) or into a derivative of pSVT7 called pSTE. pSTE was constructed by replacement of the 350-bp Cla 1-HindIII promoter/origin fragment of pSVT7 with the 418-bp Hpa IT-HindIlI fragment spanning the promoter/origin region of simian virus 40 cs 1085 (16). One microgram of each of the resulting constructs was used to transfect COS-1 cells by the DEAE-dextran method (15). Twelve hours after the chloroquine boost, the medium (Dulbecco's modified Eagle's medium containing 10o fetal bovine serum) was replaced with medium lacking serum, and incubation was continued for a further 60 hr before the conditioned medium was collected. One-chain forms of wild-type t-PA and each of the variant proteins were purified by immunoaffinity chromatography on separate columns to avoid cross-contamination. Conditioned media from transfected COS-1 cells were loaded onto columns that contained rabbit anti-PAI-i IgG conjugated to Sepharose-4B (Pharmacia) and the flow-through was adsorbed onto mouse IgG1 monoclonal antibody to t-PA (designated 1A5) that also was conjugated to Sepharose-4B. Protein bound to the second column was eluted with 0.2 M glycine hydrochloride (pH 2.2). The eluates were neutralized with 2 M Tris HCl (pH 8), concentrated by using Centricon microconcentrators (Amicon), and equilibrated with Dulbecco's phosphate-buffered saline containing 0.01% Tween-80. The concentration oft-PA was measured by solid-phase radioimmunoassay as previously described (17). Enzyme Assays. Indirect chromogenic assays of t-PA were performed as described previously (10). Desafib (soluble fibrin), human [Lysiplasminogen, and Spectrozyme PL were purchased from American Diagnostica (Greenwich, CT). Inhibition of Activity. The rate constants for inhibition of wild-type or mutant t-PAs by PAT-i were measured as described by Beatty et al. (18) and Holmes et al. (19). Briefly, purified wild-type or mutant t-PAs (3-50 fmol) were incubated at 22°C for periods of time varying from 0 to 120 min
Proc. Natl. Acad. Sci. USA 87 (1990)
3531
with purified recombinant PAI-1 (35-1330 fmol) (20) that had been activated with guanidine hydrochloride (21). After this incubation, the mixtures were diluted, and the residual enzymatic activity was determined in a standard indirect chromogenic assay. Data were analyzed by plotting ln(residual activity/initial activity) versus time and determining the slope of the resulting straight line. Pseudo-first-order rate constants were then derived by dividing the slope by the concentration of the inhibitor in the reaction. Rate Constants for Association. The rate constants for the association of PAI-1 with wild-type and mutant t-PAs were measured by a modification of the immunoradiometric assay described by Hekman and Loskutoff (22). Purified wild-type or mutant t-PAs (100 pM) were mixed with purified PAI-1 (0.12-10 nM) that had been activated with guanidine hydrochloride (21). After incubation for 1-90 min at 37TC in phosphate-buffered saline containing 3% bovine serum albumin and 0.01% Tween-80, the reactions were terminated by the addition of 4-amidinophenylmethanesulfonyl fluoride (pAPMSF, Boehringer Mannheim) to a final concentration of 10 mM. Aliquots of the mixtures were then added to 96-well microtiter plates that had been coated with goat antibody to t-PA (American Diagnostica). After incubation for 60 min at 370C, the plates were washed with phosphate-buffered saline containing 0.1% bovine serum albumin and 0.01% Tween-80. Bound PAI-1 was detected with rabbit anti-PAI-i, followed by 1251-labeled donkey anti-rabbit IgG (Amersham). The amount of PAI-1 bound to the plate was calculated from a standard curve constructed from data obtained with known amounts of t-PA-PAI-1 complexes. Data were analyzed as previously described (22) by plotting log([EO][I]/[E][IT]) versus time and determining the slope of the resulting straight line. Apparent second-order kinetic association constants were then derived by multiplying the slope by 2.303 and dividing by ([EO] - [Io]). [EO] and [Io] are the initial concentrations of t-PA and PAI-1, respectively, and [E] and [I] are the concentrations of uncomplexed t-PA and PAI-1 in samples withdrawn from the reaction mixtures after incubation for different lengths of time. When PAI-1 was used in large molar excess, pseudo-first-order rate constants were derived as described above.
RESULTS Construction of t-PA Mutants. Previous data from our laboratory have suggested that a loop of seven amino acids, located on the surface of the catalytic domain of t-PA, is involved in the formation of enzymatically inactive complexes with the serpin PAI-1 (10). This surface loop (Fig. 1) contains several positively charged residues that could form electrostatic bonds with negatively charged amino acids in the contact region of PAI-1. To assess the importance of these positively charged amino acids, we have used sitedirected mutagenesis to construct variants of t-PA in which Lys-296, Arg-298, and Arg-299 have been replaced singly or collectively by negatively charged residues (Table 1). In addition, we have tested whether other amino acids in the loop interact specifically with residues in PAI-1, by making substitutions in which His-297 and Pro-301 are replaced with Tyr and Gly, respectively (Table 1). cDNAs encoding wildtype t-PA and each of these variants were ligated into transient expression vectors and used to transfect COS-1 cells. Wild-type or mutated t-PA proteins secreted into the medium were purified by immunoaffinity chromatography, quantitated by radioimmunoassay, and assayed for enzymatic activity as described below. Activity Assays. Indirect chromogenic assays of wild-type and variant t-PAs were performed in the presence of satulrating concentrations of the positive effector Desafib and various concentrations of the substrate [Lys]plasminogen. Analysis of the resulting data by the method of Lineweaver
3532
Biochemistry: Madison et al.
Proc. Natl. Acad. Sci. USA 87 (1990)
Table 1. Compairson of the primary sequence of wild-type and mutant t-PAs in theregion of the t-PA surface loop (residues
2%-302) Enzyme
Amino acid sequence K H R R S P G E R EH R R S P G E R K XR R S P G E R K HER S P G E R K H RES P G E R K H R R S fG E R EH E E S P G E R K H R R S P G E a K H R R S P G E E
F L306 t-PA 294F A F L F A K296-+E F L F A H297-+Y F L F A R298--E F L F A R299-*E F L F A P301--G F L K296,R298,R299--E,E,E F A F L F A *K304-S F L F A *R304-E F A - - - - - - - - - F L *A2%-302 Amino acid substitutions in the mutant t-PAs are underlined. *Mutants that have been previously described (10).
Active PAI-1 (tmol)
and Burk yielded the values for Km and kcat shown in Table 2. All mutants are enzymatically active and their specific activities vary less than 20% from the activity of standardized preparations of t-PA. The values of Km and kit for wild-type t-PA, the five substitution mutants, and the deletion mutant (A296-302) are similar to one another and to values published elsewhere by our laboratory (10, 17, 23) and others (24, 25). All of the mutants respond to the positive effector Desafib in a manner similar to the wild-type enzyme. The maximal stimulation was 20- to 40-fold (data not shown), in agreement with our previous results (10, 16, 23) and those of others (26). In each case, half-maximal stimulation occurred when Desafib was present at a concentration of approximately 1-2 ttg/ml. These data show that substitution of residues within the surface loop has not significantly altered the ability of t-PA to activate plasminogen or to be stimulated by soluble fibrin fragments. Inhibition of Enzymatic Activity by PAIl-. To test whether the enzymatic activity of the variant t-PAs was resistant to inhibition by PAIl-, the mutant and wild-type forms of the enzyme were incubated with various concentrations of purified recombinant PAIl-, and the residual enzymatic activity was then measured by using the indirect chromogenic assay. The results, presented in Fig. 2, show that the enzymes can be grouped into three categories: (i) Those that are similar to wild-type t-PA (K296-)E, H297-*Y, and P301-+G). These mutants are inhibited to the same extent as wild-type t-PA. (ii) Those that are partially inhibited by PA-IM (R298-*E and R299-*E). These mutants retain about 60% of their original activity after incubation with an amount of PA-IM that completely inhibits an equivalent amount of wild-type t-PA. (iii) Those that are highly resistant to inhibition by PA-IM
FIG. 2. Effect of PA-IM on the activity of wild-type and mutant t-PA enzymes. Equal amounts of wild-type and mutant t-PAs were incubated with various amounts of PAl-i (0-1250 fmol) for 20 min at room temperature. The residual activity of the enzymes was then determined by using the indirect chromogenic assay in the presence of 0.4 mM Spectrozyme PL, 0.16 ,uM [Lysiplasminogen, and soluble fibrin monomer, Desafib, at 25 ,g/ml. o, Wild-type t-PA; +, K296-* E; o, H297-*Y; A, R298-*E; A, R299-*E; m, K296,R298,R299-* E,E,E; and *, P301-*G. mutant A296-302 is highly resistant to inhibition by PAI-1 and belongs in class iii. Taken together, these data demonstrate that positively charged amino acids located in or immediately C-terminal to the charged loop of t-PA play a critical role in the inhibition of t-PA by PAIT-. Other residues in the same region (Arg-296, His-297, and Pro-301) are not essential for inhibition. However, more extensive mutagenesis will be required to determine whether any amino acid can be tolerated at these positions. Kinetic Analysis of the Interaction Between PAI-i and t-PA. The pseudo-first-order rate constants for the inhibition of the enzymatic activities of wild-type or mutant t-PAs by PAI-i were measured at 220C as previously described (18, 19). The inhibition of wild-type t-PA by PAIT- has a rate constant of 1.4 x 106 M-1 sol at 220C (Table 2). The rate constants obtained with mutant t-PAs belonging to class ii, which contain single amino acid substitutions (R304-*S, R304-*E, R298--E, and R299-*E), are approximately 4-fold, 61-fold, 64-fold, and 60-fold, respectively, less than the rate constant obtained with wild-type t-PA (Table 2). Mutants of class iii are far more severely affected: the rate constants for inhibition of the deletion mutant A296-302 and the triple mutant K296,R298,R299--E,E,E are reduced by factors of approximately 465 and 2800, respectively (Table 2). These values are the average of results obtained from five separate experiments, which showed less than 25% variation. The rates at which wild-type and mutant t-PA proteins enter into physical association with PAI-i were measured at
(K296,R298,R299--E,E,E). We have previously reported (10) that substitution mutants R304--S and R304--E, which alter the positively charged arginine residue that maps on the lip of the active site adjacent to the charged loop, are partially inhibited by PAI-i. These variants can therefore be placed in class ii. The t-PA deletion Table 2. Kinetic analysis of t-PA mutants
kat suMs-1
Km, Enzyme t-PA
R304--S R304-.E R298--E R299- E
A296-302 K296,R298,R299 --E,E,E
0.024 0.019 0.023 0.027 0.033 0.029 0.027
0.22 0.23 0.22 0.24 0.26 0.17 0.24
Inhibition rate constant,
M-1ls-1 (220C) 1.4 x 106 3.3 2.3 2.2 2.4 3.0 5.0
Association rate constant,
M-1ls-1 (370C) 3.5 8.0 3.9 1.4
105 104 104 104
