I Knockaert

IDENTIFICATION OF POSITIVELY CHARGED RESIDUES CONTRIBUTING TO THE STABILITY OF PLASMINOGEN ACTIVATOR INHIBITOR 1

Plasminogen activator inhibitor 1 (PAI‐1), a member of the serpins, has a unique conformational flexibility. A typical characteristic is its intrinsic lability resulting in the conversion of the active conformation to a latent conformation. In the present study, we have evaluated the effect of substitution of positively charged residues located at the turn connecting strand s4C with strand s3C, either with negatively charged or with neutral residues, on the functional stability of PAI‐1. The following mutants were constructed, purified and characterized in comparison to wild‐type (wt) PAI‐1: PAI‐1‐R186E,R187E (Arg186→Glu and Arg187→Glu), PAI‐1‐H190E,K191E (His190→Glu and Lys191→Glu) and PAI‐1‐H190L,K191L (His190→Leu and Lys191→Leu). In contrast to wtPAI‐1 the mutants exhibited no inhibitory activity. Whereas latent wtPAI‐1 can be reactivated (up to a specific activity of 78±19%) by treatment with guanidinium chloride, a similar treatment applied to these mutants resulted in a significant but relatively small increase of specific activity (i.e. to 14%). Evaluation of the functional stability (at 37°C, pH 5.5, 1 M NaCl) revealed a strongly decreased functional stability compared to wtPAI‐1 (i.e. 3–9 h for the mutants vs. >24 h for wtPAI‐1). Further characterization by heat denaturation studies and plasmin susceptibility confirmed that removal or reversal of the positive charge on the turn connecting s4C with s3C results in PAI‐1 mutants with a highly accelerated conversion of active to latent forms. We can therefore conclude that the pronounced positive charge in the turn connecting s4C with s3C is of the highest importance for the functional stability of PAI‐1.

Evaluation of the mechanism of inactivation of plasminogen activator inhibitor-1 by monoclonal antibodies using a stable variant

Summary A number of studies have shown that plasminogen activator inhibitor-1 (PAI-1) can be inactivated through different mechanisms. In the current study we have carried out a comparative analysis of the effects of various PAI-1 neutralizing antibodies on wild-type PAI-1 (wtPAI-1, t1/2 ≈ 2 h) and a stable PAI-1 mutant (PAI-1-stab, t1/2 ≈ 145 h). MA8H9D4, MA-33H1 and MA-55F4, switching active wtPAI-1 to substrate PAI-1, exerted qualitatively similar effects on PAI-1-stab. Yet, the effects observed with MA-33H1 and MA-55F4 were much less pronounced on PAI-1-stab. MA-33B8 and MA-35A5 appear to exert their neutralizing properties through an acceleration of the conversion of active to latent PAI-1, thereby reducing the half-life of wtPAI-1 400-fold, whereas the half-life of PAI-1-stab was reduced up to 7000-fold. Consequently, in the presence of MA-33B8 or MA-35A5, PAI-1-stab was only slightly more stable (i.e. 4- to 6-fold) than wtPAI-1. MA-56A7C10, converting active wtPAI-1 to the latent form, also inactivates PAI-1-stab even though in the latter no plasmin cleavable site, typically present in the latent conformation of wtPAI-1, was generated. This suggests that the non-reactive conformation induced in PAI-1-stab by MA-56A7C10 is not identical to the latent conformation induced in wtPAI-1. Thus, in spite of the highly increased stability of PAI-1-stab compared to wtPAI-1, PAI-1 neutralizing antibodies may inactivate this stable PAI-1 variant with an unexpected high efficiency. This indicates that the intramolecular interactions responsible for the increased stability of PAI-1-stab do contribute only marginally to its stability after interaction with the monoclonal antibodies.

Characterization of plasminogen activator inhibitor 1 mutants containing the P13 to P10 region of ovalbumin or antithrombin III: evidence that the P13 residue contributes significantly to the active to substrate transition.

The serpin plasminogen activator inhibitor 1 (PAI-1) can occur, in vitro, in both an inhibitory and a non-inhibitory but cleavable substrate form. In the present study, we have evaluated the effect of replacing the P13 to P10 region of PAI-1 (Val-Ala-Ser-Ser), with the P13 to P10 region of either the non-inhibitory serpin ovalbumin (Glu-Val-Val-Gly; PAI-1-ovalbumin) or the inhibitory serpin antithrombin III (Glu-Ala-Ala-Ala; PAI-1-antithrombin III). In addition, we have replaced Val at position P13 with Glu (PAI-1-P13 (Val-->Glu)). Wild-type (wt) PAI-1 revealed specific activities of 80+/-9% (mean+/-S.D., n=4) of the theoretical maximum value towards t-PA. PAI-1-ovalbumin, PAI-1-antithrombin III and PAI-1-P13 (Val-->Glu) revealed specific activities of 86+/-15%, 77+/-11%, and 100+/-30% respectively, towards t-PA and similar inhibitory properties towards u-PA. Surprisingly, upon inactivation at 37 degreesC, the active conformation of the PAI-1 mutants converted partly into a substrate conformation (i.e. 52+/-5.2%, 55+/-8.2% and 46+/-4.6% for PAI-1-ovalbumin, PAI-1-antithrombin III and PAI-1-P13 (Val-->Glu), respectively) and partly into a latent conformation. This is in contrast to active wtPAI-1 which, as expected, is converted to the latent conformation (i.e. 86+/-1.0%). In conclusion, even though replacement of the P13 to P10 region of PAI-1 by the corresponding region of a non-inhibitory serpin or of an inhibitory serpin, does not directly affect its inhibitory properties, the nature of the amino acids in this region and of P13 in particular, contributes to its conformational transitions.

Expression, purification, and characterization of recombinant rat plasminogen activator inhibitor-1

Summary Plasminogen activator inhibitor-1 (PAI-1) is a physiologically important regulatory protein of the fibrinolytic system. To study in vivo its influence on a variety of biological events (thrombosis, atherosclerosis, tumour progression) a number of different experimental models in various animals, including rats, have been established. A major drawback of these models is the lack of the purified endogenous proteins and/or specific reagents for their detection in each particular species. In this study we describe the expression, in Escherichia coli (E. coli) cells, the purification, and the characterization of rat PAI-1. As expected, sodium dodecyl sulphate polyacrylamide gel electrophoresis revealed a protein with an apparent molecular mass of ∼45 kDa. Recombinant rat PAI-1 had a specific inhibitory activity of 434 000±74 000 U/mg (mean±SD, n=11) towards human tissue-type plasminogen activator (t-PA), corresponding to 58±10% of the theoretical maximum value. Evaluation of the reaction products formed upon interaction between recombinant rat PAI-1 and its target proteinases t-PA or urokinase-type plasminogen activator revealed the presence of large amounts of covalent complex, small amounts of cleaved PAI-1 and residual latent PAI-1. Active recombinant rat PAI-1 converted spontaneously to the latent form with a half-life of 2±0.2h (n=6). The second-order rate constant of inhibition of human t-PA by recombinant rat PAI-1 was two-fold lower compared to that by recombinant human PAI-1 (0.6±0.02×107M−1s−1 vs 1.4±0.14×107M−1s−1, respectively, P The current data demonstrate the biochemical equivalence between rat PAI-1 and human PAI-1. The availability of purified recombinant rat PAI-1 will allow the development of research tools required to evaluate the in vivo role of PAI-1 in various rat models.

Plasminogen activator inhibitor 1 (PAI-1) in its active conformation: crystallization and preliminary X-ray diffraction data

Because of its intrinsic lability, wild-type plasminogen activator inhibitor 1 (PAI-1) cannot be crystallized in its active conformation. Therefore, a stable variant of PAI-1 was used to retain the active conformation during crystallization. Four different crystallization conditions were evaluated in detail and two major types of crystals were detected. Whereas solutions consisting of either (i) cacodylate and sodium acetate, (ii) lithium sulfate and polyethylene glycol 4K, or (iii) imidazole, sodium chloride and sodium potassium phosphate buffer revealed thin platelet crystals, a solution (iv) containing ammonium acetate, citrate and polyethylene glycol 4K appeared to enhance the formation of clustered brush-like crystals. Crystals grown under condition (iii) were found to be suitable for X-ray data collection and consequent structural investigation. Data collection was 79.8% complete with a maximum resolution of 2.92 A. Importantly, PAI-1 retained its functional properties under all conditions.

Construction and characterization of plasminogen activator inhibitor-1 mutants in which part of the active site loop is deleted

Summary Plasminogen activator inhibitor-1 (PAI-1) can occur in a labile active inhibitory conformation, a non-inhibitory but cleavable substrate conformation and a non-reactive latent conformation. The active conformation is not stable and converts spontaneously into the latent conformation. To prevent this conversion, we have constructed, purified and characterized two mutants in which part of the reactive site loop is deleted: PAI-1-d(P6-P4) lacking residues P6 to P4 (Val-Ile-Val), and PAI-1-d(P9-P4) lacking residues P9 to P4 (Ser-Thr-Ala-Val-Ile-Val). Wild-type PAI-1 (wtPAI-1) revealed a specific activity of 54 ± 8% (mean ± SD, n = 4) of the theoretical maximum value towards tissue-type plasminogen activator (t-PA). Both deletion mutants revealed no inhibitory activity towards t-PA or towards urokinase-type plasminogen activator (u-PA). Conformational analysis of PAI-1-d(P6-P4) and PAI-1-d(P9-P4) revealed the formation of a cleaved derivative originating from the substrate form (55 ± 3% and 84 ± 8%, respectively) and non-reactive material (45 ± 3% and 16 ± 8%, respectively). Incubation at 37°C resulted in the conversion of the substrate conformation of both PAI-1 mutants into a non-reactive conformation. Reactivation experiments with guanidinium chloride revealed that for both mutants this non-reactive conformation represented a latent form from which the substrate properties could be restored up to 70–90%. In contrast to the latent form of wtPAI-1 and PAI-1-d(P6-P4), the non-reactive conformation of PAI-1-d(P9-P4) is not susceptible to plasmin cleavage outside the reactive site loop. Heat denaturation experiments revealed that the substrate conformation, as well as the latent conformation and the cleaved substrate derivative of PAI-1-d(P6-P4) and PAI-1-d(P9-P4), are less resistant to denaturation compared to wtPAI-1. In conclusion, shortening the reactive site loop of PAI-1 results in non-inhibitory mutants with an increased substrate behaviour. In contrast to the stable substrate conformation of wtPAI-1, the substrate conformation of these deletion mutants reveals an unexpected transition to the latent form.