8846
Langmuir 2004, 20, 8846-8852
Time-Dependent Conformational Changes in Fibrinogen Measured by Atomic Force Microscopy Aashiish Agnihotri† and Christopher A. Siedlecki*,†,‡ Department of Bioengineering and Department of Surgery, Pennsylvania State University, College of Medicine, Biomedical Engineering Institute, Hershey, Pennsylvania 17033 Received March 23, 2004. In Final Form: June 2, 2004 Tapping-mode atomic force microscopy was used to study the time-dependent changes in the structure of fibrinogen under aqueous conditions following adsorption on two model surfaces: hydrophobic graphite and hydrophilic mica. Fibrinogen was observed in the characteristic trinodular form, and the dimensions of the adsorbed molecules were consistent with previously reported values for these surfaces. On the basis of the differences in the relative heights of the D and the E domains, four orientation states were observed for fibrinogen adsorbed on both the surfaces. On graphite, the initial asymmetric orientation states disappeared with spreading over time. Some small lateral movements of the adsorbed proteins were observed on mica during repeated scanning, whereas no such movement was observed on graphite, indicating strong adhesion of fibrinogen to a hydrophobic surface. Spreading kinetics of fibrinogen on the two surfaces was determined by measuring the heights of the D and E domains over a time period of ∼2 h. On graphite, the heights of both the D and E domains decreased with time to a lower plateau value of 1.0 nm. On mica, the heights of both the D and E domains showed an increase, rising to an upper plateau value of ∼2.1 nm. The spreading of the D and E domains on graphite was analyzed using an ‘exponential-decay-of-height’ model. A spreading rate constant of ∼4.7 × 10-4 s-1 was observed for the whole fibrinogen molecule adsorbed on graphite, corresponding to a free energy of unfolding of ∼37 kT. Extrapolation of the exponential curve in the model to t ) 0 yielded values of 2.3 and 2.2 nm for the heights of the D and the E domains at the time of contact with the hydropbobic graphite substrate, significantly less than their free solution diameters. A two-step spreading model is proposed to explain this observation.
Introduction A fundamental principle guiding much of the current research in biomaterials science is that protein adsorption to a biomaterial substrate is the first event that occurs following contact with the biological environment, and this step is a principal determinant of the ultimate biological response. Certainly, the adsorption and activity of plasma proteins with biomaterial surfaces represents a key variable in the observed blood compatibility of cardiovascular biomaterials. Despite extensive research in this field, there is still lack of a detailed understanding of protein/surface interactions particularly as new techniques begin to address these interactions at the molecular and submolecular levels. Among the multitude of proteins adsorbed to blood-contacting biomaterials, fibrinogen has received considerable interest because it plays a pivotal role in the process of surface-induced thrombosis.1,2 Surface-bound fibrinogen demonstrates a capability for supporting platelet adhesion and initiating platelet aggregation, the first microscopically observable step in thrombosis.2 Platelets interact with adsorbed fibrinogen through adhesion receptors, primarily the integrin receptor RIIbβ3 present on the platelet membrane. Fibrinogen is the third most prevalent protein in plasma with a circulating concentration of approximately 2.6-3 mg/mL. It has a symmetrical dimeric structure with two sets of three intertwined polypeptide chains, designated as AR, Bβ, and γ, linked together by 29 disulfide bonds. In 1959, Hall and Slayter used electron microscopy studies * Corresponding author. Phone: (717) 531-5716. Fax: (717) 5314464. E-mail:
[email protected]. † Department of Bioengineering. ‡ Department of Surgery. (1) Baier, R. E.; Dutton, R. C. J. Biomed. Mater. Res. 1969, 3, 191. (2) Horbett, T. A. Cardiovasc. Pathol. 1993, 2, S137.
to propose a trinodular structure with an overall molecular length of 47.5 nm.3 The fibrinogen molecule was modeled as two outer nearly-spherical domains (termed the D domains) of diameter 6.5 nm, connected by thin linear helical regions to a central E domain of diameter 5 nm. The E domain was proposed to have the N termini of all the six chains while each set of the carboxyl termini folds in the D domains. Subsequent refinement of the structure through a combination of techniques has confirmed the general dimensions proposed in the trinodular model and revealed finer details of the structure.4 According to current models, the carboxyl terminus of the Bβ chain and the γ chain fold independently into separate nodules within the D domain, which are diagonally displaced from the long axis of the molecule. The carboxyl terminal of the AR chain, which is rich in hydrophilic amino acids, extends from the terminal domains of the β and γ chains back toward the central E domain and forms an independent nodule. The N-terminal regions of all six chains fold into an approximately globular dimeric central nodule. The six chains are joined by disulfide bonds in antiparallel fashion, giving the molecule a characteristic S shape. The amino acid residues of the three chains between the domains form a coiled-coil region where the chains are in an R-helical conformation.5 Fibrinogen has six putative sites for the platelet membrane integrin receptor RIIbβ3. The two sites located at the C terminus of each of the γ chains, commonly referred to as the γ-chain dodecapeptides (residues 400-411, HHLG(3) Hall, C. E.; Slayter, H. S. J. Biophys. Biochem. Cytol. 1959, 5, 11. (4) Weisel, J. W.; Stauffacher, C. V.; Bullitt, E.; Cohen, C. Science 1985, 230, 1388. (5) Hantgan, R. R.; Simpson-Haidaris, P. J.; Francis, C. W.; Marder, V. J. In Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 4th ed.; Colman, R. W., Hirsh, J., Marder, V. J., Clowes, A. W., George, J. N., Eds.; Lippincott Williams and Wilkins: Philadelphia, PA, 2001; p 203.
10.1021/la049239+ CCC: $27.50 © 2004 American Chemical Society Published on Web 08/21/2004
Time-Dependent Fibrinogen Conformations
GAKQAGDV), have been shown to be the primary mediators of platelet adhesion.6 Proteins undergo conformational transitions upon adsorption to a surface.2,7 Surface-induced transitions are often viewed as spreading or relaxation of proteins, where protein/surface contact area increases at a rate that varies with surface properties. A number of studies using various techniques including antibody assays,8 circular dichroism,7,9 Fourier transformed infrared/total internal reflection (FTIR/ATR) spectroscopy,10 time-of-flight secondaryion mass spectroscopy (ToF-SIMS),11 and total internal reflection fluorescence (TIRF)12 have reported postadsorption conformational transitions that may modulate the ability of these proteins to interact with other proteins. AFM is a highly versatile microscopy and analysis technique and has been widely used to study proteins. Most of the earlier AFM studies were limited to highresolution imaging of proteins in an aqueous environment with the focus on observing the detailed morphology of adsorbed proteins.13,14 Sit et al. reported quantitative dimensional analysis of fibrinogen adsorbed to three model surfaces, hydrophobic OTS self-assembled monolayer (SAM), positively charged APTES SAM, and negatively charged mica from their AFM study.15 The molecular length and width were observed to increase according to the order mica < APTES < OTS, suggesting that the spreading of fibrinogen increases with the hydrophobicity of the surface. AFM and other related force probe methods have been used to study protein-surface and protein-protein interactions. In these techniques, the probe is modified with the protein of interest and force-distance curves are acquired on model surfaces,16 surface-adsorbed protein layers,17 and membrane proteins.18 While these studies provide valuable information about the binding strength of protein-surface and protein-protein interactions, they provide little or no information about conformation and spreading of surface-bound proteins. AFM in force mode has been also used for conformational analysis of large, multidomain proteins by the so-called force spectroscopy, where a single molecule of a protein is unfolded by applying a mechanical force with the AFM tip.19,20 Along with the extent of spreading (or the final conformation of protein on the surface), another parameter that can be used to determine the strength of surfaceprotein interactions is the rate of spreading. There have been only a limited number of reports concerning the spreading kinetics of fibrinogen on surfaces. Wertz et al. investigated the spreading rate for albumin and fibrinogen (6) Farrell, D. H.; Thiagarajan, P.; Chung, D. W.; Davie, E. W. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 10729. (7) Tanaka, M.; Motomura, T.; Kawada, M.; Anzai, T.; Kasori, Y.; Shiroya, T.; Shimura, K.; Onishi, M.; Mochizuki, A. Biomaterials 2000, 21, 1471. (8) Goldberg, M. E. Trends Biochem. Sci. 1991, 16, 358. (9) Greenfield, N. J. Trac-Trends Anal. Chem. 1999, 18, 236. (10) Chittur, K. K. Biomaterials 1998, 19, 357. (11) Lhoest, J. B.; Wagner, M. S.; Tidwell, C. D.; Castner, D. G. J. Biomed. Mater. Res. 2001, 57, 432. (12) Bos, M. A.; Kleijn, J. M. Biophys. J. 1995, 68, 2566. (13) Marchant, R. E.; Barb, M. D.; Shainoff, J. R.; Eppell, S. J.; Wilson, D. L.; Siedlecki, C. A. Thromb. Haemost. 1997, 77, 1048. (14) Cacciafesta, P.; Humphris, A. D. L.; Jandt, K. D.; Miles, M. J. Langmuir 2000, 16, 8167. (15) Sit, P. S.; Marchant, R. E. Thromb. Haemost. 1999, 82, 1053. (16) Kidoaki, S.; Matsuda, T. Colloid Surf., B: Biointerfaces 2002, 23, 153. (17) Chowdhury, P. B.; Luckham, P. F. Colloid Surf., A: Physicochem. Eng. Asp. 1998, 143, 53. (18) Lee, I.; Marchant, R. E. Ultramicroscopy 2003, 97, 341. (19) Best, R. B.; Fowler, S. B.; Toca-Herrera, J. L.; Clarke, J. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12143. (20) Oberdorfer, Y.; Fuchs, H.; Janshoff, A. Langmuir 2000, 16, 9955.
Langmuir, Vol. 20, No. 20, 2004 8847
using TIRF on hydrophobic and hydrophilic self-assembled monolayer surfaces.21,22 They observed a constant spreading rate with a linear growth in the protein footprint for at least 15 min and a dramatic reduction in spreading after 2 h. The spreading rate was observed to increase with the surface hydrophobicity. As there were a number of other parameters in the experiments in these studies, including changes in bulk solution concentration and wall shear rate, it is difficult to predict the true effect of the surface on spreading kinetics. Also, TIRF and other traditional methods measure ensemble-averaged properties and thereby provide only a general idea of the state of the proteins at the surface. In this article, we report the use of AFM to obtain the time-dependent structural changes in fibrinogen adsorbed to two model surfaces, hydrophobic highly ordered pyrolitic graphite (HOPG) and hydrophilic muscovite mica. The structural changes were observed in the topographical images of the proteins and show direct evidence of the spreading occurring over time. Experimental Section Human fibrinogen, >95% pure as stated by the supplier, was purchased as lyophilized powder from Calbiochem (La Jolla, CA) and used without further characterization. Stock solutions at a concentration of 200 µg/mL were prepared in 1 mM phosphate buffer (pH 7.4) and stored at -70 °C in 100 µL aliquots. The stock solution was thawed at 37 °C for 30 min and diluted to the final concentrations for AFM imaging using 1 mM phosphate buffer immediately prior to adsorption. In this manner, protein was only subjected to one freeze-thaw cycle. The post-adsorptive dimensional changes in fibrinogen structure are on the order of just a few nanometers to a few Angstroms. Thus, to observe these changes by AFM, ultrasmooth model surfaces are necessary. The model surfaces ensure that the proteins can be detected easily and their dimensions measured precisely. Two materials with vastly different surface wettability were used in these studies: HOPG (grade II, Structure Probe Inc., PA) was used as a model hydrophobic surface, while muscovite mica (Ted Pella Inc., CA) was used as a model hydrophilic surface. These two materials are traditional surfaces for AFM imaging. A Nanoscope III Multimode AFM (Digital Instruments, Santa Barbara, CA) was used for all AFM imaging. Imaging was carried out in aqueous tapping mode using a glass fluid cell. In this mode of operation, the probe oscillates at high frequency in the Z-direction as it is raster-scanned through the X-Y plane. The intermittent contact between the probe and the surface significantly reduces the lateral forces applied to the test surface by the probe and allows for submolecular resolution imaging even in the presence of relatively weak protein/substrate adhesion. For optimal imaging, a free vibration root-mean-square (RMS) amplitude of 20 nm was used and the imaging set point was chosen such that the dampening in RMS oscillation amplitude during imaging was
