- Email: [email protected]
Structural basis of the cofactor function of denatured albumin in plasminogen activation by tissue-type plasminogen activator
BBRC Biochemical and Biophysical Research Communications 341 (2006) 736–741 www.elsevier.com/locate/ybbrc
Structural basis of the cofactor function of denatured albumin in plasminogen activation by tissue-type plasminogen activator Rita Gala´ntai a, Ka´roly Mo´dos a, Judit Fidy a, Krasimir Kolev a
b,* ,
Raymund Machovich
b
Department of Biophysics and Radiation Biology, Semmelweis University, Budapest, Hungary b Department of Medical Biochemistry, Semmelweis University, Budapest, Hungary Received 6 January 2006 Available online 19 January 2006
Abstract Certain denatured proteins function as cofactors in the activation of plasminogen by tissue-type plasminogen activator. The present study approached the structural requirements for the cofactor activity of a model protein (human serum albumin). Heat denaturation of 100–230 lM albumin (80 °C and 60–90 min) reproducibly yielded aggregates with radius in the range of 10–150 nm. The major determinant of the cofactor potency was the size of the aggregates. The increase of particle size correlated with the cofactor activity, and there was a minimal requirement for the size of the cofactor (about 10 nm radius). Similar to other proteins, the molecular aggregates with cofactor function contained a significant amount of antiparallel intermolecular b-sheets. Plasmin pre-digestion increased the cofactor efficiency (related to C-terminal lysine exposure) and did not affect profoundly the structure of the aggregates, suggesting a long-lasting and even a self-augmenting cofactor function of the denatured protein. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Tissue-type plasminogen activator; Plasminogen; Protein denaturation
Degradation of fibrin is catalyzed by proteases and among them plasmin is commonly considered to be the main fibrinolytic enzyme. Plasmin is formed from its zymogen, plasminogen by plasminogen activators via hydrolysis of a peptide bond in plasminogen (Arg561-Val562). The main endogenous activators are urokinase-type plasminogen activator (uPA) and tissue-type plasminogen activator (tPA) and the latter requires a cofactor for efficient function. A wide range of molecules and molecular systems have been identified as cofactors, such as fibrin (that is the substrate of plasmin at the same time), actin, myosin, endothelial cells, some extracellular matrix components, and also certain proteins in their denatured state (reviewed in [1]). The variety of cofactors raises the question whether plasmin plays a role in processes different from the intravascular fibrinolysis, e.g., removal of certain damaged, abnormal proteins from blood plasma or other extracellular compartments. The fact that denatured proteins play a *
Corresponding author. Fax: +36 1 2670031. E-mail address: [email protected] (K. Kolev).
0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.01.027
role in plasminogen activation is well documented [2–5], but the structural requirements for the cofactor function of these proteins are largely unknown. Recent studies emphasize the role of intermolecular b-sheets in aggregates of fibrin- and amyloid-derived peptides as well as glycated albumin for the cofactor function of these proteins [6,7]. However, the complex spatial arrangement of the multiple binding sites for tPA and plasminogen in fibrin (reviewed in [8]) suggests that the restriction of the cofactor function to a single structural prerequisite in denatured proteins may be an oversimplification. This prompted us to search for additional structural characteristics contributing to the stimulation of plasminogen activation. In this study, we selected the heat denaturation of human serum albumin (HSA) as a tool to induce cofactor function and to study its structural basis. Earlier studies have shown that the heat denaturation of HSA is a complex process involving unfolding/refolding of the molecules, aggregation of unstable non-native structures by forming intermolecular disulfide bridges or intermolecular b-sheets. The relative contribution of these processes
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
to the final state of denatured HSA may considerably depend on the conditions of heat treatment [9–11]. Further diversity may arise from the fact that prior to denaturation, the HSA sample may contain microaggregates and the monomeric HSA itself is often a mixture of mercapt- and non-mercaptalbumin [12–14]. We demonstrate now that the state/structure of denatured HSA strongly influences the cofactor potency of albumin identifying the aggregate particle size as a structural characteristic, which has a definite impact on the cofactor activity of denatured albumin. Materials and methods Chemicals. Bovine a-crystallin, 8-anilino-1-naphthalene-sulfonic acid (ANS), thioflavin T (ThT), and guanidinium hydrochloride (GdnHCl) were obtained from Sigma (St. Louis, MO) and sodium-dihydrogenphosphate, disodium-hydrogen-phosphate, and ethanol from Merck (Darmstadt, Germany). All chemicals were of spectroscopic grade. The chromogenic plasmin substrate Spectrozyme-PL (H-D-norleucyl-hexahydrotyrosyl-lysine-p-nitroanilide) and tPA were products of American Diagnostica (Hartford, CT) and Genentech Inc. (South San Francisco, CA), respectively. Cyanogen bromide human fibrinogen fragments (FDP) (Chromogenix, Molndal, Sweden) were used as a positive control cofactor of tPA. HSA preparations. Once crystallized and lyophilized HSA (Lot. No.: 127F9320 and 21K7603) was the product of Sigma (St. Louis, MO). HSA was dissolved in 50 mM sodium phosphate buffer at pH 7.4 and was used without further purification. Concentration of HSA was determined using e280nm = 37,400 M1 cm1 in phosphate buffer [15]. Two batches of HSA were studied. They were distinct as received in terms of their aggregate content. Measurements showed that the batches differed in the absolute values of the examined structural properties and cofactor potency in denatured state, but the trends of changes in the experiments were identical. Therefore, for a clear understanding of our results the comparison of the batches is not presented in the main text, only in Supplementary material. Plasminogen preparation. Human plasma was collected from healthy volunteers and published procedures were used for the isolation of plasminogen from citrated human plasma [16], as well as for the generation of plasmin and determination of its active concentration [17]. SDS gel electrophoresis and determination of the monomer content of HSA samples. HSA was dissolved to 0.1 g/l final concentration in 100 mM TRIS, pH 8.2, buffer containing 100 mM NaCl and 1% Na dodecyl sulfate (in certain samples 1% b-mercaptoethanol was used for reduction) and subjected to electrophoresis on 4–15% polyacrylamide gels (Phastsystem, Amersham Biosciences, Uppsala, Sweden). Following silver staining the resolved protein bands were analyzed densitometrically using SigmaGel 1.0 software (Jandel Scientific, Erkrath, Germany). Absorbance spectra for concentration determination of HSA and ThT, and for the correction of self-absorbance were recorded by a Cary 4E UV–Visible Spectrophotometer. The sample and the reference cells were connected to a Peltier system in order to control the temperature with an accuracy of ±1 °C. Fluorescence spectra were measured by using an Edinburgh Analytical Instruments (Edinburgh, UK) CD900 luminometer equipped with a 75 W Xenon lamp. A cooled Hamamatsu R955 PMT was used as photodetector in order to increase the signal/noise ratio. The temperature of the sample container was regulated by a Peltier system with an accuracy of ±1 °C. Dynamic light scattering measurements were performed using an ALV goniometer with a Melles Griot diode-pumped solid-state laser at 457.5 nm wavelength (type: 58 BLD 301). The intensity of the scattered light was measured at 90° and the autocorrelation function was calculated using an IBM PC-based data acquisition system developed in the institute. Particle size distributions were determined by the maximum entropy
737
method [18]. Since the scattered light intensity is strongly dependent on the size of the particles, the relative contribution of the different sized particles to the autocorrelation function does not show only the number of particles of a given radius, but overemphasizes those of larger size. Thus, minor amount of large sized impurities will have significant contribution. In order to eliminate this falsifying effect, the contribution values were divided by the corresponding radii, and these corrected values were plotted versus the radius. The radius of the particles was calculated using sphere approximation. Heat denaturation of HSA was done by incubating the protein sample at the selected temperature in the measuring cuvette of absorbance spectrophotometer or luminometer. Different denaturation conditions (heating temperature, time, and HSA concentration during the heating) were checked. The details of the selection of the denaturation condition (80 °C, 90 min, and 100 lM) applied in most of the experiments discussed below are presented in Supplementary material. Plasminogen activation assay was carried out as published previously [19]. Briefly, plasminogen (3.2 lM) was activated with 74 nM tPA in the presence of 0.1 mg/ml cofactor in 10 mM Hepes–NaOH, pH 7.4, buffer containing 150 mM NaCl at 37 °C. At time intervals during the activation, samples were taken and the generated plasmin activity was measured on 0.1 mM Spectrozyme-PL. The hydrolysis of the plasmin substrate was monitored by measuring the absorbance at 405 nm with a Beckman DU7500 spectrophotometer. ThT binding studies. The fluorescence of ThT is a good indicator for the presence of larger protein aggregates that contain intermolecular b-sheets [20]. Stock solution of ThT (1 mg/ml) was prepared in 50 mM sodium-phosphate buffer at pH 7.4. Since according to the manufacturer, the dye content of the powder was just about 75%, the precise concentration was determined using e416nm = 26,600 M1 cm1. The binding of ThT to HSA was studied at [ThT]/[protein] molar ratios of 0 to 110. The dye was always added to the albumin at room temperature, all spectra were recorded at 25 °C. Emission spectrum of ThT was measured in the range of 460–570 nm by excitation at 450 nm. The effect of self-absorbance and turbidity on the fluorescence spectra was corrected using Aex Aem F corr ¼ F obs 10ð 5 þ 2 Þ
ð1Þ
where Fcorr and Fobs are the corrected and observed fluorescence intensities, Aex and Aem are the absorbances at the excitation and the emission wavelengths, respectively [21].
Results Cofactor efficiency of different HSA samples Native and heat-denatured albumin were tested for their cofactor activity in plasminogen activation with tPA (Fig. 1). HSA had no cofactor activity in native state. However, following heat denaturation it accelerated the tPAcatalyzed plasminogen activation. Since earlier studies showed that pre-digestion of denatured proteins with plasmin increases the efficiency of plasminogen activation [4], we checked this effect for our samples. Plasmin pre-digestion of native HSA did not influence the plasminogen activation with tPA. Pre-digestion of denatured albumin, however, resulted in cofactor potency increased by a factor of about 2. Structural parameters related to the cofactor function of HSA Several methods were used to identify the structural characteristics necessary for the cofactor activity of HSA.
738
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
After heating (80 °C, 90 min, and 100 lM HSA), the HSA sample contained large aggregates that could not enter the gel under non-reducing conditions (Fig. 2). The plasmin treatment itself did not eliminate these aggregates despite the extensive digestion seen under reducing conditions.
Fig. 1. Cofactor effects in the plasminogen activation with tPA. Plasminogen (3.2 lM) was activated with tPA (74 nM) in the presence of the indicated cofactors (0.1 mg/ml) and the generated plasmin activity was monitored as described in Materials and methods. The plasmin activity measured after 4 min activation is presented in relative units (the activity in the absence of cofactors is 1). Abbreviations: nHSA, native HSA; dHSA, HSA denatured in 100 lM concentration at 80 °C for 90 min; +P, digestion of the indicated sample with 0.35 lM plasmin for 2 h at 37 °C prior to the activation assay; Cryst, native a-crystallin.
SDS gel electrophoresis Gel electrophoresis under non-reducing conditions showed the presence of monomeric (67 kDa) and aggregated (mainly dimeric) forms in the SDS treated HSA (Fig. 2). Reduction eliminated the aggregates indicating that the presence of disulfide bridges between HSA molecules was the reason for dimer formation in the native sample. Referring to the lack of cofactor activity of the native albumin (Fig. 1), we can conclude that the monomeric albumin and the aggregated structures present in the initial (‘‘native’’) state are not enough for the cofactor effect.
Fig. 2. Evaluation of aggregates in HSA by SDS gel electrophoresis. SDS gel electrophoresis was carried out as described in Materials and methods under reducing (R) or non-reducing conditions (NR). Abbreviations: nHSA, native HSA; dHSA, HSA denatured in 100 lM concentration at 80 °C for 90 min; +P, digestion of the indicated sample with 0.35 lM plasmin for 2 h at 37 °C.
Particle size distribution by dynamic light scattering Since SDS gel electrophoresis does not provide information about the particle size distribution of heat-denatured HSA, dynamic light scattering was applied for this purpose (Fig. 3). The distribution curve of 5 lM of native HSA had a rather wide peak around 4 nm covering also the size of dimers (around 9 nm) in agreement with the SDS gel electrophoresis results. Identical pattern of distribution was observed with 10–100 lM HSA (not shown). After plasmin treatment, the distribution curve narrowed mainly because of the elimination of the particles that are larger than monomers, implying that the higher sized particles (dimers at least) were favored in plasmin digestion. After incubating 100 lM of HSA at 80 °C for 90 min, a smaller peak appeared at around 9–10 nm and a more intense one around 40 nm, indicating that after heat treatment the sample contained larger aggregates, but a smaller amount of dimers was also present. Plasmin digestion only mildly shifted the peak of larger-sized particles to smaller radii and widened the distribution curve in agreement with the findings for native HSA (the largest aggregates were better substrates of plasmin, but no extensive digestion occurred). Comparing the size and the cofactor activity of HSA samples not pre-digested with plasmin (see Fig. 1), one can see that the increase of particle size correlated with the increase of cofactor activity.
Fig. 3. Size distribution of particles in solutions of native and denatured HSA determined by dynamic light scattering. The relative contribution to the autocorrelation function of the particles with different radii was evaluated for native (nHSA), plasmin pre-digested (nHSA + P), denatured (dHSA), and denatured and plasmin pre-digested (dHSA + P) HSA samples with the dynamic light scattering procedure described in Materials and methods. Denaturation and plasmin pre-digestion conditions are the same as described in Fig. 1.
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
739
Correlation between the particle size and the cofactor activity of denatured HSA Heat-denatured HSA (230 lM, 80 °C, and 60 min) was centrifuged (105g, 60 min), and the upper 50 ll fraction (supernatant containing 13% of the total protein subjected to centrifugation) and the lower 50 ll fraction (sediment, containing 17% of the total protein) were examined in parallel for cofactor potency and particle size distribution. The particle size of different fractions of HSA and their cofactor potency are summarized in Fig. 4. Centrifugation eliminated particles of size larger than 100 nm from both fractions and significantly reduced the contribution of the 10– 100 nm radius size in the supernatant fraction. The sediment contained dimers and aggregates formed by three, four or more monomers. The cofactor potency of the supernatant, containing mainly small monomeric and
Fig. 5. Evaluation of repetitious b-sheet structures in HSA by ThT binding experiments. HSA (2 lM) was mixed with various concentrations of ThT and the fluorescence (excitation at 450 nm and emission at 482 nm) was measured as described in Materials and methods. Symbols: circle, native HSA; square, heat-denatured HSA.
dimeric particles, was significantly lower than the potency of the sediment.
Fig. 4. Effect of particle size on the cofactor properties of heat-denatured HSA in plasminogen activation by tPA. Heat-denatured HSA (230 lM, 80 °C, and 60 min) (dHSA) and its fractions after centrifugation at 100,000g for 60 min (upper 50-ll, supernatant, and lower 50-ll, sediment) were used separately for the determination of particle size distribution (A) and the cofactor potency (B) measurement. In (A), the measurement was carried out as in Fig. 3 and the lines are mean values from two independent fractionations; the data for the unfractionated denatured HSA are shown with solid line; for the supernatant with dashed-anddotted line, for the sediment with dashed line. In (B), plasminogen activation was performed as in Fig. 1 and the plasmin activity measured after 4 min activation is presented in relative units (the activity in the absence of cofactors is 1). The bars represent the mean and standard deviation of five parallel measurements, asterisks indicate difference significant at p < 0.001 level compared to the other two samples.
ThT binding In an earlier study, the analysis of CD spectra showed that after heat denaturation of HSA the oligomers contained intermolecular b-structures [9]. Since ThT binds to ordered b-structures and binding is accompanied by increase in its fluorescence, it can be used to differentiate the repetitious antiparallel b-sheet structures and the irregular aggregates of fibrils [20]. In our study, we examined the binding of ThT to the different HSA samples in order to compare their b-structure content. Negligible increase of the fluorescence intensity was observed when 2 lM of native albumin was mixed with increasing concentration of ThT (Fig. 5). The denaturation of HSA, however, resulted in a dramatic increase of ThT fluorescence, showing that this state of albumin was able to bind the dye, so the albumin aggregates contained repetitious b-sheets. Plasmin predigestion did not change the ThT binding ability either of native or denatured HSA (not shown). The lack of cofactor activity and ThT binding in the case of native albumin raised the possibility for a role of the repetitious b-sheets in the cofactor potency as previously reported [6,7]. We checked this hypothesis using a-crystallin instead of HSA as a cofactor. Native a-crystallin exists in the form of large oligomers of about 800 kDa with significant antiparallel b-sheet content. The size of the oligomer is about 10 nm [22]. The monomers contain b-structure in about 30% [23]. The cofactor activity of a-crystallin in plasminogen activation was higher than that of native HSA (Fig. 1), but did not exceed that of denatured albumin. Discussion In agreement with our earlier report [4], the present results confirm that in the process of plasminogen activation by tPA: (i) HSA does not behave as a cofactor in
740
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
native state; (ii) when HSA is heat denatured, it gains cofactor properties; (iii) pre-digestion by plasmin increases the cofactor activity. Our present study extends these findings providing insights into the structural background of the cofactor potency. Role of particle size in cofactor activity of HSA SDS gel electrophoresis showed that the main components in native HSA were monomers and dimers. Since native HSA did not affect the plasminogen activation by tPA, we suggest that the HSA dimers are too small and the requirement for this effect is a minimal size (at least 10 nm of radius). This conclusion is in good agreement with the particle size distribution and cofactor activity of denatured HSA samples: after heat denaturation, cofactor activity developed in parallel with the formation of aggregates larger than dimers. The dynamic light scattering data for the fractions gained with differential centrifugation (Fig. 4) provided further evidence for the correlation of the particle size of the denatured albumin and its cofactor potency, and for the minimal size requirement in cofactor activity. The supernatant, that contained only a small amount of aggregates with higher than 10 nm radius, had much lower cofactor activity than the sediment or the denatured HSA, in which samples significant amount of larger aggregates were present. These observations raise the question about the structural role of the particle size in the cofactor effect. Based on the results with fibrin as a cofactor, the cofactor has to fulfill, at least, two important functions: (i) to bring plasminogen and tPA near each other and orient them [24] and (ii) to induce the conformational change of plasminogen into an extended, flexible structure in which plasminogen can be more easily activated [25]. In both cases, the cofactor has to bind to tPA and/or plasminogen. The assumption of the direct contact between the cofactor and, at least, plasminogen is in good agreement with the conclusion of the studies where plasminogen was suggested to be the target of all denatured proteins having cofactor activity [3,4]. Our present results suggest that the extensive surface of the aggregates exceeding certain size is required to provide a template for the binding partners (activator and plasminogen). This may be a general mechanism explaining why a wide range of molecules can act as cofactors. Role of intermolecular b-sheets in cofactor activity ThT binding studies on different HSA samples indicated that the cofactor activity correlates with the amount of repetitious b-sheets in the structure. This finding is in agreement with the recent data for the fibrillar aggregates of amyloid and glycated albumin [6,7]. However, not the bsheet structure on its own, but rather its role in the maintenance of large intermolecular aggregates is essential for
the cofactor potency. The moderate cofactor activity of a-crystallin supports this conclusion. a-Crystallin is rich in intra- and intermolecular b-sheets, but its size does not exceed 10 nm, and thus its cofactor activity is low (Fig 1). On the other hand, one of the interactions that connect the monomers to each other in the aggregates is the formation of intermolecular b-sheets. Thus, the increase of ThT binding in samples with higher cofactor activity is in line with the higher amount of b-sheets in the larger aggregates with better cofactor properties. The interaction between bstructures is not the only force that builds up the aggregates. Reduction of disulfide bridges results in disaggregation of the largest particles (Fig. 2) in agreement with the observation that the amount of irreversible denaturation is reduced by blocking the free –SH group of cysteine-34 in HSA [9]. Thus, the presence of cross b-sheets seems to be an essential, but not exclusive prerequisite for the cofactor potency. Role of plasmin pre-digestion in cofactor activity Plasmin pre-digestion eliminated only the largest aggregates slightly shifting the size distribution curve to smaller radii (Fig. 3) and did not affect the ThT binding capability of HSA. Thus, the enhancement of the cofactor potency following plasmin digestion of heat-denatured HSA (Fig. 1) cannot be explained with the structural properties discussed above, additional factors that influence the cofactor efficiency should be considered. Because plasmin cleaves peptide bonds next to lysine residues in the substrate proteins, the newly exposed C-terminal lysine residues provide additional binding sites for plasminogen [26]. Binding of ligands to plasminogen profoundly changes the conformation of the plasminogen molecule; its maximal longitude is increased from 15 to 24 nm [27]. Adopting this open conformation plasminogen is more susceptible to some PAs [28]. Binding of plasminogen to the lysine residues exposed when denatured HSA is digested with plasmin could result in conformational changes similar to those induced by small ligands and in an analogous way to the situation of plasminogen activation on fibrin surface [29–34]. In conclusion, our results support a concept that plasminogen activation by tPA is stimulated by several well-defined structural changes of denatured proteins. Formation of extensive aggregates maintained in part by intermolecular b-sheet interactions and disulfide bridges provide initial template for the assembly of the activation complex, the cofactor properties of which are subsequently enhanced by the generated plasmin. This general mechanism can contribute to the non-fibrinolytic functions of plasminogen (recently reviewed in [35]). Acknowledgments The technical assistance of R. Marka´cs and Gy. Oravecz is highly appreciated. The authors acknowledge the
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
support from Hungarian Grants OTKA T-031891 (R.M.), K60123 (K.K.), M041729 (J.F.), ETT 017/2003 (R.M.), ETT 501/2003 (J.F.), NKFP-1A/0023/2002 (R.M.), and the Wellcome Trust (069520/Z/02/Z) (K.K.).
[16] [17]
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/ j.bbrc.2006.01.027.
[18]
[19]
References
[20]
[1] K. Kolev, R. Machovich, Molecular and cellular modulation of fibrinolysis, Thromb. Haemost. 89 (2003) 610–621. [2] R. Radcliffe, T. Heinze, Stimulation of tissue plasminogen activator by denatured proteins and fibrin clots: a possible role for plasminogen activator? Arch. Biochem. Biophys. 211 (1981) 750–761. [3] R. Radcliffe, A critical role of lysine residues in the stimulation of tissue plasminogen activator by denatured proteins and fibrin clots, Biochim. Biophys. Acta 743 (1983) 422–430. [4] R. Machovich, W.G. Owen, Denatured proteins as cofactors for plasminogen activation, Arch. Biochem. Biophys. 344 (1997) 343–349. [5] R. Machovich, E. Komorowicz, K. Kolev, W.G. Owen, Facilitation of plasminogen activation by denatured prothrombin, Thromb. Res. 94 (1999) 389–394. [6] O. Kranenburg, B. Bouma, L.M.G. Kroon-Batenburg, A. Reijerkerk, Y.P. Wu, E.E. Voest, M.F.B.G. Gebbink, Tissue-type plasminogen activator is a multiligand cross-b structure receptor, Curr. Biol. 12 (2002) 1833–1839. [7] B. Bouma, L.M.G. Kroon-Batenburg, Y.P. Wu, B. Brunjes, G. Posthuma, O. Kranenburg, P.G. de Groot, E.E. Voest, M.F.B.G. Gebbink, Glycation induces cross-b structure in albumin, J. Biol. Chem. 278 (2003) 41810–41819. [8] L. Medved, W. Nieuwenhuizen, Molecular mechanisms of initiation of fibrinolysis by fibrin, Thromb. Haemost. 89 (2003) 409–419. [9] R. Wetzel, M. Becker, J. Behlke, H. Billwitz, S. Bo¨hm, B. Ebert, H. Hamann, J. Krumbiegel, G. Lassmann, Temperature behavior of human serum albumin, Eur. J. Biochem. 104 (1980) 469–478. [10] M. Fukuoka, T. Kobayashi, T. Satoh, A. Tanaka, A. Kubodera, Studies of quality control of 99m Tc-labelled macroaggregated albumin part 1 aggregation of non-mercaptalbumin and its conformation, Nucl. Med. Biol. 20 (1993) 643–648. [11] G.A. Pico, Thermodynamic features of the thermal unfolding of human serum albumin, Int. J. Biol. Macromol. 20 (1997) 63–73. [12] M. Sogami, S. Nagoka, S. Era, M. Honda, K. Noguchi, Resolution of human mercapt- and nonmercaptalbumin by high-performance liquid chromatography, Int. J. Pept. Protein Res. 24 (1984) 96–103. [13] M. Sogami, S. Era, S. Nagoka, K. Kuwata, K. Kida, K. Miura, H. Inouye, E. Suzuki, S. Hayano, S. Sawada, HPLC-studies on nonmercapt–mercapt conversion of human serum albumin, Int. J. Pept. Protein Res. 25 (1985) 398–402. [14] S. Era, T. Hamaguchi, M. Sogami, K. Kuwata, E. Suzuki, K. Miura, K. Kawai, Y. Kitazawa, H. Okabe, A. Noma, S. Miyata, Further studies on the resolution of human mercapt- and nonmercaptalbumin and on human serum albumin in the elderly by high-performance liquid chromatography, Int. J. Pept. Protein Res. 31 (1988) 435–442. [15] E. Reddi, C.R. Lambert, G. Jori, M.A. Rogers, Photokinetic and photophysical measurements of the sensitized photooxidation of the
[21] [22] [23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
741
tryptophyl residue in N-acetyl tryptophanamide and in human serum albumin, Photochem. Photobiol. 45 (1987) 345–351. D.G. Deutsch, E.T. Mertz, Plasminogen: purification from human plasma by affinity chromatography, Science 170 (1970) 1095–1096. K. Kolev, I. Le´ra´nt, K. Tenekejiev, R. Machovich, Regulation of fibrinolytic activity of neutrophil leukocyte elastase, plasmin, and miniplasmin by plasma protease inhibitors, J. Biol. Chem. 269 (1994) 17030–17034. R.K. Bryan, in: P.F. Fougere (Ed.), Maximum Entropy and Bayesian Methods, Kluwer Academic Publishers, The Netherlands, 1990, pp. 221–232. K. Kolev, W.G. Owen, R. Machovich, Dual effect of synthetic plasmin substrates on plasminogen activation, Biochim. Biophys. Acta 1247 (1995) 239–245. H. LeVine III, Quantification of b-sheet amyloid fibril structures with thioflavin T, Methods Enzymol. 309 (1999) 274–284. J.R. Lakowitz, Corrected emission spectra, in: Principles of Fluorescence Spectroscopy, Plenum Press, New York, 1983, pp. 40–43. J. Horwitz, Alpha-Crystallin, Exp. Eye Res. 76 (2003) 145–153. P.N. Farnsworth, H. Frauwirth, B. Groth-Vasselli, S. Kamalendra, Refinement of 3D structure of bovine lens alpha A-crystallin, Int. J. Biol. Macromol. 22 (1998) 175–185. G. Spraggon, S.J. Everse, R.F. Doolittle, Crystal structures of fragment D from human fibrinogen and its crosslinked counterpart from fibrin, Nature 389 (1997) 455–462. R. Machovich, R.D. Litwiller, W.G. Owen, Requirement of zymogen modification for activation of porcine plasminogen, Biochemistry 31 (1992) 11558–11561. V. Fleury, E. Angles-Cano, Characterization of the binding of plasminogen to fibrin surfaces: the role of carboxy-terminal lysines, Biochemistry 30 (1991) 7630–7638. W.F. Mangel, B.H. Lin, V. Ramakrishnan, Characterization of an extremely large, ligand-induced conformational change in plasminogen, Science 248 (1990) 69–73. V. Sinniger, R.E. Merton, P. Fabregas, J. Felez, C. Longstaff, Regulation of tissue plasminogen activator activity by cells. Domains responsible for binding and mechanism of stimulation, J. Biol. Chem. 274 (1999) 12414–12422. M. Vaskuilen, A. Vermond, G.H. Veeneman, J.H. van Boom, E.A. Klasen, N.D. Zegers, W. Nieuwenhuizen, Fibrinogen lysine residue Alpha 157 plays a crucial role in the fibrin-induced acceleration of plasminogen activation, catalyzed by tissue-type plasminogen activator, J. Biol. Chem. 262 (1987) 5944–5946. G. Tsurupa, L. Medved, Identification and characterization of novel tPA- and plasminogen-binding sites within fibrin(ogen) aC-domains, Biochemistry 40 (2001) 801–808. P. Grailhe, W. Nieuwenhuizen, E. Angles-Cano, Study of tissue-type plasminogen activator binding sites on fibrin using distinct fragments of fibrinogen, Eur. J. Biochem. 219 (1994) 961–967. O. Yonekawa, M. Voskuilen, W. Nieuwenhuizen, Localization in the fibrinogen gamma-chain of a new site that is involved in the acceleration of the tissue-type plasminogen activator-catalysed activation of plasminogen, Biochem. J. 283 (1992) 187–191. C. de Vries, H. Veerman, E. Koornneef, H. Pannekoek, Tissue-type plasminogen activator and its substrate Glu-plasminogen share binding sites in limited plasmin-digested fibrin, J. Biol. Chem. 265 (1990) 13547–13552. E. Suenson, L.C. Petersen, Fibrin and plasminogen structures essential to stimulation of plasmin formation by tissue-type plasminogen activator, Biochim. Biophys. Acta 870 (1986) 510–519. E.F. Plow, J. Hoover-Plow, The functions of plasminogen in cardiovascular disease, Trends Cardiovasc. Med. 14 (2004) 180–186.
Structural basis of the cofactor function of denatured albumin in plasminogen activation by tissue-type plasminogen activator Rita Gala´ntai a, Ka´roly Mo´dos a, Judit Fidy a, Krasimir Kolev a
b,* ,
Raymund Machovich
b
Department of Biophysics and Radiation Biology, Semmelweis University, Budapest, Hungary b Department of Medical Biochemistry, Semmelweis University, Budapest, Hungary Received 6 January 2006 Available online 19 January 2006
Abstract Certain denatured proteins function as cofactors in the activation of plasminogen by tissue-type plasminogen activator. The present study approached the structural requirements for the cofactor activity of a model protein (human serum albumin). Heat denaturation of 100–230 lM albumin (80 °C and 60–90 min) reproducibly yielded aggregates with radius in the range of 10–150 nm. The major determinant of the cofactor potency was the size of the aggregates. The increase of particle size correlated with the cofactor activity, and there was a minimal requirement for the size of the cofactor (about 10 nm radius). Similar to other proteins, the molecular aggregates with cofactor function contained a significant amount of antiparallel intermolecular b-sheets. Plasmin pre-digestion increased the cofactor efficiency (related to C-terminal lysine exposure) and did not affect profoundly the structure of the aggregates, suggesting a long-lasting and even a self-augmenting cofactor function of the denatured protein. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Tissue-type plasminogen activator; Plasminogen; Protein denaturation
Degradation of fibrin is catalyzed by proteases and among them plasmin is commonly considered to be the main fibrinolytic enzyme. Plasmin is formed from its zymogen, plasminogen by plasminogen activators via hydrolysis of a peptide bond in plasminogen (Arg561-Val562). The main endogenous activators are urokinase-type plasminogen activator (uPA) and tissue-type plasminogen activator (tPA) and the latter requires a cofactor for efficient function. A wide range of molecules and molecular systems have been identified as cofactors, such as fibrin (that is the substrate of plasmin at the same time), actin, myosin, endothelial cells, some extracellular matrix components, and also certain proteins in their denatured state (reviewed in [1]). The variety of cofactors raises the question whether plasmin plays a role in processes different from the intravascular fibrinolysis, e.g., removal of certain damaged, abnormal proteins from blood plasma or other extracellular compartments. The fact that denatured proteins play a *
Corresponding author. Fax: +36 1 2670031. E-mail address: [email protected] (K. Kolev).
0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.01.027
role in plasminogen activation is well documented [2–5], but the structural requirements for the cofactor function of these proteins are largely unknown. Recent studies emphasize the role of intermolecular b-sheets in aggregates of fibrin- and amyloid-derived peptides as well as glycated albumin for the cofactor function of these proteins [6,7]. However, the complex spatial arrangement of the multiple binding sites for tPA and plasminogen in fibrin (reviewed in [8]) suggests that the restriction of the cofactor function to a single structural prerequisite in denatured proteins may be an oversimplification. This prompted us to search for additional structural characteristics contributing to the stimulation of plasminogen activation. In this study, we selected the heat denaturation of human serum albumin (HSA) as a tool to induce cofactor function and to study its structural basis. Earlier studies have shown that the heat denaturation of HSA is a complex process involving unfolding/refolding of the molecules, aggregation of unstable non-native structures by forming intermolecular disulfide bridges or intermolecular b-sheets. The relative contribution of these processes
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
to the final state of denatured HSA may considerably depend on the conditions of heat treatment [9–11]. Further diversity may arise from the fact that prior to denaturation, the HSA sample may contain microaggregates and the monomeric HSA itself is often a mixture of mercapt- and non-mercaptalbumin [12–14]. We demonstrate now that the state/structure of denatured HSA strongly influences the cofactor potency of albumin identifying the aggregate particle size as a structural characteristic, which has a definite impact on the cofactor activity of denatured albumin. Materials and methods Chemicals. Bovine a-crystallin, 8-anilino-1-naphthalene-sulfonic acid (ANS), thioflavin T (ThT), and guanidinium hydrochloride (GdnHCl) were obtained from Sigma (St. Louis, MO) and sodium-dihydrogenphosphate, disodium-hydrogen-phosphate, and ethanol from Merck (Darmstadt, Germany). All chemicals were of spectroscopic grade. The chromogenic plasmin substrate Spectrozyme-PL (H-D-norleucyl-hexahydrotyrosyl-lysine-p-nitroanilide) and tPA were products of American Diagnostica (Hartford, CT) and Genentech Inc. (South San Francisco, CA), respectively. Cyanogen bromide human fibrinogen fragments (FDP) (Chromogenix, Molndal, Sweden) were used as a positive control cofactor of tPA. HSA preparations. Once crystallized and lyophilized HSA (Lot. No.: 127F9320 and 21K7603) was the product of Sigma (St. Louis, MO). HSA was dissolved in 50 mM sodium phosphate buffer at pH 7.4 and was used without further purification. Concentration of HSA was determined using e280nm = 37,400 M1 cm1 in phosphate buffer [15]. Two batches of HSA were studied. They were distinct as received in terms of their aggregate content. Measurements showed that the batches differed in the absolute values of the examined structural properties and cofactor potency in denatured state, but the trends of changes in the experiments were identical. Therefore, for a clear understanding of our results the comparison of the batches is not presented in the main text, only in Supplementary material. Plasminogen preparation. Human plasma was collected from healthy volunteers and published procedures were used for the isolation of plasminogen from citrated human plasma [16], as well as for the generation of plasmin and determination of its active concentration [17]. SDS gel electrophoresis and determination of the monomer content of HSA samples. HSA was dissolved to 0.1 g/l final concentration in 100 mM TRIS, pH 8.2, buffer containing 100 mM NaCl and 1% Na dodecyl sulfate (in certain samples 1% b-mercaptoethanol was used for reduction) and subjected to electrophoresis on 4–15% polyacrylamide gels (Phastsystem, Amersham Biosciences, Uppsala, Sweden). Following silver staining the resolved protein bands were analyzed densitometrically using SigmaGel 1.0 software (Jandel Scientific, Erkrath, Germany). Absorbance spectra for concentration determination of HSA and ThT, and for the correction of self-absorbance were recorded by a Cary 4E UV–Visible Spectrophotometer. The sample and the reference cells were connected to a Peltier system in order to control the temperature with an accuracy of ±1 °C. Fluorescence spectra were measured by using an Edinburgh Analytical Instruments (Edinburgh, UK) CD900 luminometer equipped with a 75 W Xenon lamp. A cooled Hamamatsu R955 PMT was used as photodetector in order to increase the signal/noise ratio. The temperature of the sample container was regulated by a Peltier system with an accuracy of ±1 °C. Dynamic light scattering measurements were performed using an ALV goniometer with a Melles Griot diode-pumped solid-state laser at 457.5 nm wavelength (type: 58 BLD 301). The intensity of the scattered light was measured at 90° and the autocorrelation function was calculated using an IBM PC-based data acquisition system developed in the institute. Particle size distributions were determined by the maximum entropy
737
method [18]. Since the scattered light intensity is strongly dependent on the size of the particles, the relative contribution of the different sized particles to the autocorrelation function does not show only the number of particles of a given radius, but overemphasizes those of larger size. Thus, minor amount of large sized impurities will have significant contribution. In order to eliminate this falsifying effect, the contribution values were divided by the corresponding radii, and these corrected values were plotted versus the radius. The radius of the particles was calculated using sphere approximation. Heat denaturation of HSA was done by incubating the protein sample at the selected temperature in the measuring cuvette of absorbance spectrophotometer or luminometer. Different denaturation conditions (heating temperature, time, and HSA concentration during the heating) were checked. The details of the selection of the denaturation condition (80 °C, 90 min, and 100 lM) applied in most of the experiments discussed below are presented in Supplementary material. Plasminogen activation assay was carried out as published previously [19]. Briefly, plasminogen (3.2 lM) was activated with 74 nM tPA in the presence of 0.1 mg/ml cofactor in 10 mM Hepes–NaOH, pH 7.4, buffer containing 150 mM NaCl at 37 °C. At time intervals during the activation, samples were taken and the generated plasmin activity was measured on 0.1 mM Spectrozyme-PL. The hydrolysis of the plasmin substrate was monitored by measuring the absorbance at 405 nm with a Beckman DU7500 spectrophotometer. ThT binding studies. The fluorescence of ThT is a good indicator for the presence of larger protein aggregates that contain intermolecular b-sheets [20]. Stock solution of ThT (1 mg/ml) was prepared in 50 mM sodium-phosphate buffer at pH 7.4. Since according to the manufacturer, the dye content of the powder was just about 75%, the precise concentration was determined using e416nm = 26,600 M1 cm1. The binding of ThT to HSA was studied at [ThT]/[protein] molar ratios of 0 to 110. The dye was always added to the albumin at room temperature, all spectra were recorded at 25 °C. Emission spectrum of ThT was measured in the range of 460–570 nm by excitation at 450 nm. The effect of self-absorbance and turbidity on the fluorescence spectra was corrected using Aex Aem F corr ¼ F obs 10ð 5 þ 2 Þ
ð1Þ
where Fcorr and Fobs are the corrected and observed fluorescence intensities, Aex and Aem are the absorbances at the excitation and the emission wavelengths, respectively [21].
Results Cofactor efficiency of different HSA samples Native and heat-denatured albumin were tested for their cofactor activity in plasminogen activation with tPA (Fig. 1). HSA had no cofactor activity in native state. However, following heat denaturation it accelerated the tPAcatalyzed plasminogen activation. Since earlier studies showed that pre-digestion of denatured proteins with plasmin increases the efficiency of plasminogen activation [4], we checked this effect for our samples. Plasmin pre-digestion of native HSA did not influence the plasminogen activation with tPA. Pre-digestion of denatured albumin, however, resulted in cofactor potency increased by a factor of about 2. Structural parameters related to the cofactor function of HSA Several methods were used to identify the structural characteristics necessary for the cofactor activity of HSA.
738
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
After heating (80 °C, 90 min, and 100 lM HSA), the HSA sample contained large aggregates that could not enter the gel under non-reducing conditions (Fig. 2). The plasmin treatment itself did not eliminate these aggregates despite the extensive digestion seen under reducing conditions.
Fig. 1. Cofactor effects in the plasminogen activation with tPA. Plasminogen (3.2 lM) was activated with tPA (74 nM) in the presence of the indicated cofactors (0.1 mg/ml) and the generated plasmin activity was monitored as described in Materials and methods. The plasmin activity measured after 4 min activation is presented in relative units (the activity in the absence of cofactors is 1). Abbreviations: nHSA, native HSA; dHSA, HSA denatured in 100 lM concentration at 80 °C for 90 min; +P, digestion of the indicated sample with 0.35 lM plasmin for 2 h at 37 °C prior to the activation assay; Cryst, native a-crystallin.
SDS gel electrophoresis Gel electrophoresis under non-reducing conditions showed the presence of monomeric (67 kDa) and aggregated (mainly dimeric) forms in the SDS treated HSA (Fig. 2). Reduction eliminated the aggregates indicating that the presence of disulfide bridges between HSA molecules was the reason for dimer formation in the native sample. Referring to the lack of cofactor activity of the native albumin (Fig. 1), we can conclude that the monomeric albumin and the aggregated structures present in the initial (‘‘native’’) state are not enough for the cofactor effect.
Fig. 2. Evaluation of aggregates in HSA by SDS gel electrophoresis. SDS gel electrophoresis was carried out as described in Materials and methods under reducing (R) or non-reducing conditions (NR). Abbreviations: nHSA, native HSA; dHSA, HSA denatured in 100 lM concentration at 80 °C for 90 min; +P, digestion of the indicated sample with 0.35 lM plasmin for 2 h at 37 °C.
Particle size distribution by dynamic light scattering Since SDS gel electrophoresis does not provide information about the particle size distribution of heat-denatured HSA, dynamic light scattering was applied for this purpose (Fig. 3). The distribution curve of 5 lM of native HSA had a rather wide peak around 4 nm covering also the size of dimers (around 9 nm) in agreement with the SDS gel electrophoresis results. Identical pattern of distribution was observed with 10–100 lM HSA (not shown). After plasmin treatment, the distribution curve narrowed mainly because of the elimination of the particles that are larger than monomers, implying that the higher sized particles (dimers at least) were favored in plasmin digestion. After incubating 100 lM of HSA at 80 °C for 90 min, a smaller peak appeared at around 9–10 nm and a more intense one around 40 nm, indicating that after heat treatment the sample contained larger aggregates, but a smaller amount of dimers was also present. Plasmin digestion only mildly shifted the peak of larger-sized particles to smaller radii and widened the distribution curve in agreement with the findings for native HSA (the largest aggregates were better substrates of plasmin, but no extensive digestion occurred). Comparing the size and the cofactor activity of HSA samples not pre-digested with plasmin (see Fig. 1), one can see that the increase of particle size correlated with the increase of cofactor activity.
Fig. 3. Size distribution of particles in solutions of native and denatured HSA determined by dynamic light scattering. The relative contribution to the autocorrelation function of the particles with different radii was evaluated for native (nHSA), plasmin pre-digested (nHSA + P), denatured (dHSA), and denatured and plasmin pre-digested (dHSA + P) HSA samples with the dynamic light scattering procedure described in Materials and methods. Denaturation and plasmin pre-digestion conditions are the same as described in Fig. 1.
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
739
Correlation between the particle size and the cofactor activity of denatured HSA Heat-denatured HSA (230 lM, 80 °C, and 60 min) was centrifuged (105g, 60 min), and the upper 50 ll fraction (supernatant containing 13% of the total protein subjected to centrifugation) and the lower 50 ll fraction (sediment, containing 17% of the total protein) were examined in parallel for cofactor potency and particle size distribution. The particle size of different fractions of HSA and their cofactor potency are summarized in Fig. 4. Centrifugation eliminated particles of size larger than 100 nm from both fractions and significantly reduced the contribution of the 10– 100 nm radius size in the supernatant fraction. The sediment contained dimers and aggregates formed by three, four or more monomers. The cofactor potency of the supernatant, containing mainly small monomeric and
Fig. 5. Evaluation of repetitious b-sheet structures in HSA by ThT binding experiments. HSA (2 lM) was mixed with various concentrations of ThT and the fluorescence (excitation at 450 nm and emission at 482 nm) was measured as described in Materials and methods. Symbols: circle, native HSA; square, heat-denatured HSA.
dimeric particles, was significantly lower than the potency of the sediment.
Fig. 4. Effect of particle size on the cofactor properties of heat-denatured HSA in plasminogen activation by tPA. Heat-denatured HSA (230 lM, 80 °C, and 60 min) (dHSA) and its fractions after centrifugation at 100,000g for 60 min (upper 50-ll, supernatant, and lower 50-ll, sediment) were used separately for the determination of particle size distribution (A) and the cofactor potency (B) measurement. In (A), the measurement was carried out as in Fig. 3 and the lines are mean values from two independent fractionations; the data for the unfractionated denatured HSA are shown with solid line; for the supernatant with dashed-anddotted line, for the sediment with dashed line. In (B), plasminogen activation was performed as in Fig. 1 and the plasmin activity measured after 4 min activation is presented in relative units (the activity in the absence of cofactors is 1). The bars represent the mean and standard deviation of five parallel measurements, asterisks indicate difference significant at p < 0.001 level compared to the other two samples.
ThT binding In an earlier study, the analysis of CD spectra showed that after heat denaturation of HSA the oligomers contained intermolecular b-structures [9]. Since ThT binds to ordered b-structures and binding is accompanied by increase in its fluorescence, it can be used to differentiate the repetitious antiparallel b-sheet structures and the irregular aggregates of fibrils [20]. In our study, we examined the binding of ThT to the different HSA samples in order to compare their b-structure content. Negligible increase of the fluorescence intensity was observed when 2 lM of native albumin was mixed with increasing concentration of ThT (Fig. 5). The denaturation of HSA, however, resulted in a dramatic increase of ThT fluorescence, showing that this state of albumin was able to bind the dye, so the albumin aggregates contained repetitious b-sheets. Plasmin predigestion did not change the ThT binding ability either of native or denatured HSA (not shown). The lack of cofactor activity and ThT binding in the case of native albumin raised the possibility for a role of the repetitious b-sheets in the cofactor potency as previously reported [6,7]. We checked this hypothesis using a-crystallin instead of HSA as a cofactor. Native a-crystallin exists in the form of large oligomers of about 800 kDa with significant antiparallel b-sheet content. The size of the oligomer is about 10 nm [22]. The monomers contain b-structure in about 30% [23]. The cofactor activity of a-crystallin in plasminogen activation was higher than that of native HSA (Fig. 1), but did not exceed that of denatured albumin. Discussion In agreement with our earlier report [4], the present results confirm that in the process of plasminogen activation by tPA: (i) HSA does not behave as a cofactor in
740
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
native state; (ii) when HSA is heat denatured, it gains cofactor properties; (iii) pre-digestion by plasmin increases the cofactor activity. Our present study extends these findings providing insights into the structural background of the cofactor potency. Role of particle size in cofactor activity of HSA SDS gel electrophoresis showed that the main components in native HSA were monomers and dimers. Since native HSA did not affect the plasminogen activation by tPA, we suggest that the HSA dimers are too small and the requirement for this effect is a minimal size (at least 10 nm of radius). This conclusion is in good agreement with the particle size distribution and cofactor activity of denatured HSA samples: after heat denaturation, cofactor activity developed in parallel with the formation of aggregates larger than dimers. The dynamic light scattering data for the fractions gained with differential centrifugation (Fig. 4) provided further evidence for the correlation of the particle size of the denatured albumin and its cofactor potency, and for the minimal size requirement in cofactor activity. The supernatant, that contained only a small amount of aggregates with higher than 10 nm radius, had much lower cofactor activity than the sediment or the denatured HSA, in which samples significant amount of larger aggregates were present. These observations raise the question about the structural role of the particle size in the cofactor effect. Based on the results with fibrin as a cofactor, the cofactor has to fulfill, at least, two important functions: (i) to bring plasminogen and tPA near each other and orient them [24] and (ii) to induce the conformational change of plasminogen into an extended, flexible structure in which plasminogen can be more easily activated [25]. In both cases, the cofactor has to bind to tPA and/or plasminogen. The assumption of the direct contact between the cofactor and, at least, plasminogen is in good agreement with the conclusion of the studies where plasminogen was suggested to be the target of all denatured proteins having cofactor activity [3,4]. Our present results suggest that the extensive surface of the aggregates exceeding certain size is required to provide a template for the binding partners (activator and plasminogen). This may be a general mechanism explaining why a wide range of molecules can act as cofactors. Role of intermolecular b-sheets in cofactor activity ThT binding studies on different HSA samples indicated that the cofactor activity correlates with the amount of repetitious b-sheets in the structure. This finding is in agreement with the recent data for the fibrillar aggregates of amyloid and glycated albumin [6,7]. However, not the bsheet structure on its own, but rather its role in the maintenance of large intermolecular aggregates is essential for
the cofactor potency. The moderate cofactor activity of a-crystallin supports this conclusion. a-Crystallin is rich in intra- and intermolecular b-sheets, but its size does not exceed 10 nm, and thus its cofactor activity is low (Fig 1). On the other hand, one of the interactions that connect the monomers to each other in the aggregates is the formation of intermolecular b-sheets. Thus, the increase of ThT binding in samples with higher cofactor activity is in line with the higher amount of b-sheets in the larger aggregates with better cofactor properties. The interaction between bstructures is not the only force that builds up the aggregates. Reduction of disulfide bridges results in disaggregation of the largest particles (Fig. 2) in agreement with the observation that the amount of irreversible denaturation is reduced by blocking the free –SH group of cysteine-34 in HSA [9]. Thus, the presence of cross b-sheets seems to be an essential, but not exclusive prerequisite for the cofactor potency. Role of plasmin pre-digestion in cofactor activity Plasmin pre-digestion eliminated only the largest aggregates slightly shifting the size distribution curve to smaller radii (Fig. 3) and did not affect the ThT binding capability of HSA. Thus, the enhancement of the cofactor potency following plasmin digestion of heat-denatured HSA (Fig. 1) cannot be explained with the structural properties discussed above, additional factors that influence the cofactor efficiency should be considered. Because plasmin cleaves peptide bonds next to lysine residues in the substrate proteins, the newly exposed C-terminal lysine residues provide additional binding sites for plasminogen [26]. Binding of ligands to plasminogen profoundly changes the conformation of the plasminogen molecule; its maximal longitude is increased from 15 to 24 nm [27]. Adopting this open conformation plasminogen is more susceptible to some PAs [28]. Binding of plasminogen to the lysine residues exposed when denatured HSA is digested with plasmin could result in conformational changes similar to those induced by small ligands and in an analogous way to the situation of plasminogen activation on fibrin surface [29–34]. In conclusion, our results support a concept that plasminogen activation by tPA is stimulated by several well-defined structural changes of denatured proteins. Formation of extensive aggregates maintained in part by intermolecular b-sheet interactions and disulfide bridges provide initial template for the assembly of the activation complex, the cofactor properties of which are subsequently enhanced by the generated plasmin. This general mechanism can contribute to the non-fibrinolytic functions of plasminogen (recently reviewed in [35]). Acknowledgments The technical assistance of R. Marka´cs and Gy. Oravecz is highly appreciated. The authors acknowledge the
R. Gala´ntai et al. / Biochemical and Biophysical Research Communications 341 (2006) 736–741
support from Hungarian Grants OTKA T-031891 (R.M.), K60123 (K.K.), M041729 (J.F.), ETT 017/2003 (R.M.), ETT 501/2003 (J.F.), NKFP-1A/0023/2002 (R.M.), and the Wellcome Trust (069520/Z/02/Z) (K.K.).
[16] [17]
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/ j.bbrc.2006.01.027.
[18]
[19]
References
[20]
[1] K. Kolev, R. Machovich, Molecular and cellular modulation of fibrinolysis, Thromb. Haemost. 89 (2003) 610–621. [2] R. Radcliffe, T. Heinze, Stimulation of tissue plasminogen activator by denatured proteins and fibrin clots: a possible role for plasminogen activator? Arch. Biochem. Biophys. 211 (1981) 750–761. [3] R. Radcliffe, A critical role of lysine residues in the stimulation of tissue plasminogen activator by denatured proteins and fibrin clots, Biochim. Biophys. Acta 743 (1983) 422–430. [4] R. Machovich, W.G. Owen, Denatured proteins as cofactors for plasminogen activation, Arch. Biochem. Biophys. 344 (1997) 343–349. [5] R. Machovich, E. Komorowicz, K. Kolev, W.G. Owen, Facilitation of plasminogen activation by denatured prothrombin, Thromb. Res. 94 (1999) 389–394. [6] O. Kranenburg, B. Bouma, L.M.G. Kroon-Batenburg, A. Reijerkerk, Y.P. Wu, E.E. Voest, M.F.B.G. Gebbink, Tissue-type plasminogen activator is a multiligand cross-b structure receptor, Curr. Biol. 12 (2002) 1833–1839. [7] B. Bouma, L.M.G. Kroon-Batenburg, Y.P. Wu, B. Brunjes, G. Posthuma, O. Kranenburg, P.G. de Groot, E.E. Voest, M.F.B.G. Gebbink, Glycation induces cross-b structure in albumin, J. Biol. Chem. 278 (2003) 41810–41819. [8] L. Medved, W. Nieuwenhuizen, Molecular mechanisms of initiation of fibrinolysis by fibrin, Thromb. Haemost. 89 (2003) 409–419. [9] R. Wetzel, M. Becker, J. Behlke, H. Billwitz, S. Bo¨hm, B. Ebert, H. Hamann, J. Krumbiegel, G. Lassmann, Temperature behavior of human serum albumin, Eur. J. Biochem. 104 (1980) 469–478. [10] M. Fukuoka, T. Kobayashi, T. Satoh, A. Tanaka, A. Kubodera, Studies of quality control of 99m Tc-labelled macroaggregated albumin part 1 aggregation of non-mercaptalbumin and its conformation, Nucl. Med. Biol. 20 (1993) 643–648. [11] G.A. Pico, Thermodynamic features of the thermal unfolding of human serum albumin, Int. J. Biol. Macromol. 20 (1997) 63–73. [12] M. Sogami, S. Nagoka, S. Era, M. Honda, K. Noguchi, Resolution of human mercapt- and nonmercaptalbumin by high-performance liquid chromatography, Int. J. Pept. Protein Res. 24 (1984) 96–103. [13] M. Sogami, S. Era, S. Nagoka, K. Kuwata, K. Kida, K. Miura, H. Inouye, E. Suzuki, S. Hayano, S. Sawada, HPLC-studies on nonmercapt–mercapt conversion of human serum albumin, Int. J. Pept. Protein Res. 25 (1985) 398–402. [14] S. Era, T. Hamaguchi, M. Sogami, K. Kuwata, E. Suzuki, K. Miura, K. Kawai, Y. Kitazawa, H. Okabe, A. Noma, S. Miyata, Further studies on the resolution of human mercapt- and nonmercaptalbumin and on human serum albumin in the elderly by high-performance liquid chromatography, Int. J. Pept. Protein Res. 31 (1988) 435–442. [15] E. Reddi, C.R. Lambert, G. Jori, M.A. Rogers, Photokinetic and photophysical measurements of the sensitized photooxidation of the
[21] [22] [23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
741
tryptophyl residue in N-acetyl tryptophanamide and in human serum albumin, Photochem. Photobiol. 45 (1987) 345–351. D.G. Deutsch, E.T. Mertz, Plasminogen: purification from human plasma by affinity chromatography, Science 170 (1970) 1095–1096. K. Kolev, I. Le´ra´nt, K. Tenekejiev, R. Machovich, Regulation of fibrinolytic activity of neutrophil leukocyte elastase, plasmin, and miniplasmin by plasma protease inhibitors, J. Biol. Chem. 269 (1994) 17030–17034. R.K. Bryan, in: P.F. Fougere (Ed.), Maximum Entropy and Bayesian Methods, Kluwer Academic Publishers, The Netherlands, 1990, pp. 221–232. K. Kolev, W.G. Owen, R. Machovich, Dual effect of synthetic plasmin substrates on plasminogen activation, Biochim. Biophys. Acta 1247 (1995) 239–245. H. LeVine III, Quantification of b-sheet amyloid fibril structures with thioflavin T, Methods Enzymol. 309 (1999) 274–284. J.R. Lakowitz, Corrected emission spectra, in: Principles of Fluorescence Spectroscopy, Plenum Press, New York, 1983, pp. 40–43. J. Horwitz, Alpha-Crystallin, Exp. Eye Res. 76 (2003) 145–153. P.N. Farnsworth, H. Frauwirth, B. Groth-Vasselli, S. Kamalendra, Refinement of 3D structure of bovine lens alpha A-crystallin, Int. J. Biol. Macromol. 22 (1998) 175–185. G. Spraggon, S.J. Everse, R.F. Doolittle, Crystal structures of fragment D from human fibrinogen and its crosslinked counterpart from fibrin, Nature 389 (1997) 455–462. R. Machovich, R.D. Litwiller, W.G. Owen, Requirement of zymogen modification for activation of porcine plasminogen, Biochemistry 31 (1992) 11558–11561. V. Fleury, E. Angles-Cano, Characterization of the binding of plasminogen to fibrin surfaces: the role of carboxy-terminal lysines, Biochemistry 30 (1991) 7630–7638. W.F. Mangel, B.H. Lin, V. Ramakrishnan, Characterization of an extremely large, ligand-induced conformational change in plasminogen, Science 248 (1990) 69–73. V. Sinniger, R.E. Merton, P. Fabregas, J. Felez, C. Longstaff, Regulation of tissue plasminogen activator activity by cells. Domains responsible for binding and mechanism of stimulation, J. Biol. Chem. 274 (1999) 12414–12422. M. Vaskuilen, A. Vermond, G.H. Veeneman, J.H. van Boom, E.A. Klasen, N.D. Zegers, W. Nieuwenhuizen, Fibrinogen lysine residue Alpha 157 plays a crucial role in the fibrin-induced acceleration of plasminogen activation, catalyzed by tissue-type plasminogen activator, J. Biol. Chem. 262 (1987) 5944–5946. G. Tsurupa, L. Medved, Identification and characterization of novel tPA- and plasminogen-binding sites within fibrin(ogen) aC-domains, Biochemistry 40 (2001) 801–808. P. Grailhe, W. Nieuwenhuizen, E. Angles-Cano, Study of tissue-type plasminogen activator binding sites on fibrin using distinct fragments of fibrinogen, Eur. J. Biochem. 219 (1994) 961–967. O. Yonekawa, M. Voskuilen, W. Nieuwenhuizen, Localization in the fibrinogen gamma-chain of a new site that is involved in the acceleration of the tissue-type plasminogen activator-catalysed activation of plasminogen, Biochem. J. 283 (1992) 187–191. C. de Vries, H. Veerman, E. Koornneef, H. Pannekoek, Tissue-type plasminogen activator and its substrate Glu-plasminogen share binding sites in limited plasmin-digested fibrin, J. Biol. Chem. 265 (1990) 13547–13552. E. Suenson, L.C. Petersen, Fibrin and plasminogen structures essential to stimulation of plasmin formation by tissue-type plasminogen activator, Biochim. Biophys. Acta 870 (1986) 510–519. E.F. Plow, J. Hoover-Plow, The functions of plasminogen in cardiovascular disease, Trends Cardiovasc. Med. 14 (2004) 180–186.