SCANNING VOL. 37, 158–164 (2015) © Wiley Periodicals, Inc.
Adsorption of Human Serum Albumin Onto Highly Orientated Pyrolytic Graphite Surface Studied by Atomic Force Microscopy XIAO PENG,1,2,3 HAORAN FU,4† RUISI LIU,1,2,3 LIN ZHAO,1,2,3 YUANGANG ZU,1,2,3 FENGJIE XU,1,2,3
AND
ZHIGUO LIU1,2,3,*
1
State Engineering Laboratory of Bio-Resource Eco-Utilization, Harbin, People’s Republic of China Engineering Research Center of Forest Bio-preparation, Ministry of Education, Northeast Forestry University, Harbin, People’s Republic of China 3 Key Laboratory of Forest Plant Ecology of Ministry of Education, Northeast Forestry University, Harbin, People’s Republic of China 4 School of Chemistry and Environment, Beihang University, Beijing 100191, People’s Republic of China 2
Summary: It is important to know the adsorption behavior and assembly structure of human serum albumin (HSA) molecules onto a carbonaceous substrate for further application of carbon nanomaterials in biomedical field. Individual HSA molecules and oligmers (dimer and trimer) adsorbed onto HOPG surface have been imaged by atomic force microscopy (AFM). Individual HSA molecule appeared as an ellipsoid on HOPG surface with average length of 12.6, width of 6.5, and height of 1.9 nm when they were incubated at the physiological condition (pH 7.4). HSA molecules also can form the interconnected chains, uniform network, and monolayer by tuning the initial concentrations and adsorption time. Furthermore, HSA molecules can assemble into quite different network structures and irregular chains at pH of 2, 5, and 10. This study could expand our knowledge of the interactions between protein and carbonaceous surfaces. SCANNING 37:158–164, 2015. © 2015 Wiley Periodicals, Inc. Key words: human serum albumin (HSA), HOPG, atomic force microscopy (AFM)
Contract grant sponsor: Fundamental Research Funds for the Central Universities; Contract grant number: DL12DA02; Contract grant sponsor: National Public-welfare Research Fund for Industry of Forestry; Contract grant number: 201104002-3. Conflicts of interest: None. † These authors contributed equally to this work (co-first author). Address for reprints: Zhiguo Liu, Key Laboratory of Forest Plant Ecology of Ministry of Education, Northeast Forestry University, Harbin 150040, People’s Republic of China E-mail: [email protected] Received 19 November 2014; revised 30 December 2014; Accepted with revision 9 January 2015 DOI: 10.1002/sca.21193 Published online 13 February 2015 in Wiley Online Library (wileyonlinelibrary.com).
Introduction Human serum albumin (HSA) is the most abundant protein in human blood. Many studies have revealed that HSA has a high affinity to a wide range of materials such as metal ions, fatty acids, amino acids, and a lot of drug compounds. Recent studies further proved that HSA is a perfect carrier for many kinds of anticancer drug (Kratz, 2008). Recently, new class of carbon materials such as carbon nanotubes and graphene that is a single-atom thick, two-dimensional sheet of hexagonally arranged carbon atoms isolated from its three-dimensional parent material, graphite, have received considerable attention due to their potential applications in variety of fields. Biological and biomedical applications of these carbon nanomaterials are of great interest. Some in vitro and in vivo biomedical applications such as for drug delivery and in bone regeneration have been explored in the past few years (Liu et al., 2009; Tran et al., 2009). Especially for graphene and its derivatives such as graphene oxide (GO), reduced graphene oxide (RGO), and GO-nanocomposites in biomedical field, significant progresses have been made in recent years (Yang et al., 2013). It has been demonstrated that graphene can be used for anti-cancer drug delivery, cancer therapy, and biological imaging. For further biomedical applications of graphene, the most important issue is still the potential long-term toxicity concerns. Protein adsorption on graphene surfaces is believed to mediate cell uptake and thus toxic responses, and hence is a very important issue to investigate. However, there have been few studies of graphene–protein interactions. More research including microscopic observation is clearly needed on the interaction of proteins with graphene-family nanomaterials. The comprehensive assessment of adsorption behaviors of protein on graphene and its derivatives at the physiological condition is not only important to
Z. Liu et al.: Adsorption of HSA Onto HOPG
investigate their toxicology, but also critical to improve their performance in various biomedical imaging and therapy applications. When carbon nanomaterials enter into the circulatory system, the interactions between the HSA molecules and carbonaceous surfaces are inevitable. It is believed that the interactions are very crucial for the function and destination of carbon nanomaterials as well as the loaded cargo on their surfaces. Thus, it is important to know the adsorption behavior and assembly structures of HSA molecules onto a carbonaceous substrate. However, there are few reports about the behavior and structures of HSA molecules on carbonaceous surfaces. More recently, Bradley et al. explored HSA adsorption and imaging on the graphite surfaces (Orasanu-Gourlay and Bradley, 2006). In this study, we further extend to study molecular structure and adsorption behavior of HSA on a carbonaceous surface at different concentration and pH conditions. The adsorption of HSA at the concentration of 1–1,000 ng/ml and different pH condition onto highly orientated pyrolytic graphite (HOPG) surfaces was investigated using atomic force microscopy (AFM). HOPG, a well defined carbonaceous substrate with large and atomically flat terraces can be used as a surrogate for the various carbonaceous materials to study protein structures on their surfaces. Adsorption from solution at relatively low concentration (1 ng/ml) allowed individual protein molecules to be imaged. When the adsorption was performed at 10 ng/ml or more high concentration, a protein network and monolayer can be obtained.
Materials and Methods HSA was obtained from Sigma–Aldrich (A9731, lyophilized powder, 96%). HOPG of ZYB grade was provided by MikroMasch, Estonia. The water used in all experiments was purified by a MilliporeMilli-Q system, and its resistivity was >18 MV/cm. All other reagents are analytical grade and used as-received without further purification. 1–1000 ng/ml of HSA solutions under different pH condition were prepared from its pure powder and different buffers. The HSA solutions at pH 7.4, 5.0 were prepared from 0.2 M phosphate buffer and acetate acid buffer, respectively; protein solutions at pH 2.0 and 10.0 were obtained by further adjusting of phosphate buffer and acetate acid buffer with HCl and NaOH solution, respectively. For adsorption of HSA samples onto HOPG surfaces, typically 40 ml of the asprepared protein solutions were deposited on the freshly cleaved HOPG surface and incubated for a predefined time. After incubation, the protein droplet was removed and HOPG surfaces washed by pure water at least three times. Finally, the HOPG sample was dried in air or under a flow of pure Ar gas before AFM imaging.
159
All AFM images were obtained by a PicoPlus II AFM system from Molecular Imaging, Inc. (Tempe, AZ,). The scanner was calibrated with a standard reference of 2 mm pitch lines with a depth of 200 nm provided by Molecular Imaging, Inc., NSC35 type silicon tips (radius of curvature is less than 10 nm) with a nominal force constant of 4.5 N/m provided by MikroMasch, Estonia were used. Original AFM images were obtained at 512 512 pixel. The dimensions of HSA topography were measured manually by using the PicoScan5.3.3 software. Three-dimensional representation of the HSA was performed with Scanning Probe Image Processor (Image Metrology ApS, Lyngby, Denmark)
Results and Discussion Figure 1 shows the representative AFM topographic and phase image of HSA on HOPG surface. Some wellseparated ellipsoidal particles can be identified. They were considered to be the individual HSA molecules based on their topography and dimension. In addition, there are some dimer and trimer of HSA as the arrows shown in Figure 1(a), which were further confirmed by their corresponding phase image as shown in Figure 1 (b). The phase image reflects the phase lag of the cantilever oscillation relative to the signal sent to the cantilever’s piezo driver, which provides the adhesion and visco-elasticity information of the surface and can be used for identifying different surfaces. HSA molecules and HOPG surfaces can be differentiated from their contrast in the phase images, in which HSA molecule appears darker, whereas HOPG surfaces appear bright. This result confirms that the observed objects in the AFM images are HSA molecules rather than artifacts on the HOPG surface. Previous electron microscope observation and calculated mode indicated that bovine serum albumin (BSA) tend to form dimer and trimers (Bloomfield, ’66; Slayter, ’65). More recently, self-association of BSA has been further investigated by fluorescence resonance energy transfer (Levi and Flecha, 2002). It is revealed that association between BSA monomers occurs through a reversible path that involves specific interactions between the protein molecules (Levi and Flecha, 2002). The observed HSA oligomers in present study revealed that HSA also tend to self-associate to different oligomers, which is very similar to those of BSA. Figure 2(a) and (b) shows the typical topographic and three-dimensional image of a HSA molecule on the HOPG surface. It can be seen that HSA appeared as an ellipsoid on HOPG surface.The average dimensions of these individual HSA molecules were measured. The average length of long axis is 31.3 4.2, short axis is 22.8 2.8, and average height is 1.91 0.22 nm (50 well separated individual HSA molecules were measured for average). The actual dimension of HSA should
160
SCANNING VOL. 37, 2 (2015)
Fig. 1. The representative AFM topographic and phase image of HSA (1 ng/ml incubated for 72 min) on HOPG surface. (a) Topographic image; (b) phase image.
Fig. 2. The typical AFM topographic and three-dimensional image of a HSA molecule on the HOPG surface. (a) Topographic image; (b) three-dimensional image.
Z. Liu et al.: Adsorption of HSA Onto HOPG
be less than these values due to the well-known broadening effect caused by the finite probe tip radius in AFM imaging. The actual dimension of HSA can be estimated by a previously proposed formula to deconvolute the tip effects (Liu et al., 2008). The calculated length of HSA after deconvolution is 12.6 nm and the width is 6.5 nm when the AFM tip radius assumed to be 10 nm. This result is very close to those from scanning tunneling microscopy (STM) imaging by Feng et al. (Feng et al., ’89). They found that HSA dried on a graphite surface to appear as an elongated (6 12 nm) molecule showing three linear domains. Further investigation of the HSA molecule structure in Figure 2(a) and (b) indicated that the three homologous domains of HSA are aligned linearly like as three tennis balls in a can. The boundary of three domains can be identified as shown in Figure 2(b). The present structural mode of HSA is very similar to that of BSA proposed previously by Bloomfield (Bloomfield, ’66). Albumin structures in solution and their crystal structrues have been widely investigated in the past (Scheider et al., ’76; Carter et al., ’89; He and Carter, ’92; Sugio et al., ’99). There has been general consensus from many physical measurements that in solution, HSA and BSA have the cigar-shape form of an ellipsoid of revolution (Peters, ’95). However, X-Ray diffraction now shows that albumin in a crystal, at least fatty acidfree albumin has a triangular or heart-like shape (an equilateral triangle of 8 * 8 nm dimensions and 3 nm height) (Sugio et al., ’99). This favors the rounded rather than the elongated picture obtained by most physical chemical studies. The different pictures may arise from the flexibility of the albumin molecule, which allows it to assume different configurations when packed tightly in a crystal or floating free in solution, and their reconciliation should be enlightening about its behavior. Our present observation by AFM revealed that HSA at least on HOPG surface has an ellipsoidal shape, which is consistent with the reported structures in solution. Bradley et al. observed both round and oblong shape of HSA on HOPG surface in their recent report (OrasanuGourlay and Bradley, 2006) though the sample preparation processes between their and our work are different. Interestingly, triangular and oblong features of HSA have been observed on mica surface (Quist et al., ’95). Therefore, in despite of the difference of sample preparation and adsorption surfaces, HSA can be transformed into a similar isomer featured with an oblong shape. Protein concentration is an important factor to influence their self-association and assembly structures. When more high concentration of HSA was used for adsorption, we observed quite different HSA structures on HOPG surface. Figure 3(a) shows the representative AFM topographic image of HSA incubated from 10 ng/ ml solution. HSA form a very uniform network structure on the HOPG surface instead of separated individual HSA molecules. The height of the network structure can
161
be measured from the exposed pores on HOPG surface. Interestingly, the height of the network structure is consistent across the scanning area as indicated in the section analysis in Figure 3(a). The average height of the segments of network is 2.15 0.31 nm. In comparison with the dimension of individual HSA molecule as described above, we can conclude that the network structure is made up of the interconnected individual HSA. Furthermore, most pores exposed out of the network structure have a polygonal shape. This phenomenon can be attributed to the aligning ability of carbonaceous surfaces in which the top two or top three graphene layers were considered to be involved in the surface interaction (Liu et al., 2012; Liu et al., 2013). As we reported previously, both circular and linear DNA molecules tend to form hexagonal patterns on HOPG surface (Liu et al., 2013). Further increasing the HSA concentration to 1,000 ng/ml, the obtained HSA structure is shown in Figure 3(b). HSA formed a flat monolayer on which there were some big protein aggregates. The number and size of the exposed pores are dramatically decreased under this relatively high concentration in comparison with that at low concentration (10 ng/ml). This result revealed that the high concentration of HSA facilitates more proteins to be adsorbed on HOPG surface. We further investigated the influences of the adsorption time on HSA structures on HOPG surface. Figure 4 shows the AFM topographic image of HSA incubated from 10 ng/ml solution for 1 and 30 min, respectively. After 1 min incubation, HSA formed into some interconnected chains as shown in Figure 4(a). As increasing adsorption time to 30 min, a monolayer of HSA was formed as indicated in Figure 4(b). From the HSA images in Figure 4(a), (b), and Figure 3(a), it is indicated that the adsorption of HSA on HOPG surface is a time-dependent process. As the adsorption time was increased, HSA evolved from the interconnected chains, network structure to a monolayer. As revealed by previous studies, HSA molecule undergoes several well-recognized changes in conformation under different pH conditions (Peters, ’95). Besides the assembly structures of HSA at pH 7.4 as described above, the adsorption of HSA under other pH conditions has been investigated. Figure 5 shows the representative AFM topographic image of HSA (10 ng/ ml) incubated from the solutions at pH of 2.0, 5.0, and 10.0, respectively. At pH 2.0, HSA formed a network structure as indicated in Figure 5(a). The average height of the segments of network is 2.88 0.40 nm. At pH 5.0, a heterogeneous network was obtained as shown in Figure 5(b). The average height of the segments of network is 3.90 0.52 nm, which is higher than that at pH 2.0. Under pH 11.0, some irregular chains with average height of 2.7 0.48 nm were presented as illustrated in Figure 5(c). All these results indicated that the pH condition can influence HSA assembly structure.
162
SCANNING VOL. 37, 2 (2015)
Fig. 3. The representative AFM topographic image of HSA incubated from 10 ng/ml and 1,000 ng/ml solution for 15 min on HOPG surface, respectively. (a) 10 ng/ml; (b) 1,000 ng/ml.
pH 5 is a very important point for albumin conformation since both the isoionic point and the isoelectric point of albumin are around 5.0 (Peters, ’95). The isoionic point of albumins, the pH of a thoroughly deionized solution, is about pH 5.2. The isoelectric point, in contrast to the isoionic point, is the pH at which
Fig. 4.
the net charge of a molecule, including any bound ions, is zero. For undefatted albumin in 0.15 M NaCl the isoelectric pH is about 4.7 (Peters, ’95). Thus, the electrostatic interactions between the HSA are minimized around pH 5.0 whereas self-association of the HSA through hydrophobic interactions are enhanced.
The AFM topographic image of HSA incubated from 10 ng/ml solution for 1 and 30 min, respectively. (a) 1 min; (b) 30 min.
Z. Liu et al.: Adsorption of HSA Onto HOPG
163
Fig. 5. The representative AFM topographic image of HSA (10 ng/ml) incubated from the solutions at pH of 2, 5, and 10, respectively. (a) pH ¼ 2; (b) pH ¼ 5; (c) pH ¼ 10.
Therefore, the relative higher height of the segment of the network at pH 5.0 can be understandable. They should be composed of two or more HSA molecules rather than the interconnected individual HSA. The
topographic image in Figure 5(b) appears to support this speculation. At extremely pH condition of 2.0, the albumin is considered in its extended form (E form) wherein all
164
SCANNING VOL. 37, 2 (2015)
carboxyl groups have become protonated (Peters, ’95). The positive charges of lysine, arginine, and histidine residues would cause mutual repulsion between the domains and subdomains of the molecule. An unfolding occurs until the molecule appears to be as fully expanded as its disulfide bonding structure allows. A previous electron microscope study reported that BSA at pH 1.9 has a thread-like form with approximately 25 nm long and 2.1 nm in diameter (Slayter, ’65). In present study, HSA formed a network structure rather than a separated individual HSA molecule. A possible reason for network structure formation is that HSA molecules connected to each other by hydrogen bonding between the carboxyl groups of acid residues. At pH 10, HSA is considered in its basic form (B form) wherein imidizole is the main ionizable group (Peters, ’95). It is indicated that B isomerization is a structural fluctuation, a loosening of the molecule with loss of rigidity (Peters, ’95). Furthermore, the irreversible conformational changes may occur in basic conditions caused by disulfide interchange initiated by Cys-34. The observed irregular chains in Figure 5(c) implied that some domain of HSA have irreversible conformational changes, which is not so obvious at other pH conditions.
Conclusions Individual HSA molecules and oligmers (dimer and trimer) adsorbed on HOPG surface have been imaged by AFM. Individual HSA molecule appeared as an ellipsoid on HOPG surface with average length of 12.6, width of 6.5, and height of 1.9 nm when they were incubated at the physiological condition (pH 7.4). HSA molecules also can form the interconnected chains, uniform network, and monolayer by tuning the initial concentration and adsorption time. Furthermore, HSA molecules can assemble into quite different network structures and irregular chains at pH of 2, 5, and 10. This study could expand our knowledge of the interactions between protein and carbonaceous surfaces.
References Bloomfield V. 1966. The structure of bovine serum albumin at low pH. Biochemistry 5: 684–689. Carter DC, He X-M, Munson SH, et al. 1989. Three-dimensional structure of human serum albumin. Science 244: 1195–1198. Feng L, Andrade J, Hu C. 1989. Scanning tunneling microscopy of proteins on graphite surfaces. Scanning Microsc 3: 399–410. He XM, Carter DC. 1992. Atomic structure and chemistry of human serum albumin. Nature 358: 209–215. Kratz F. 2008. Albumin as a drug carrier: design of prodrugs, drug conjugates and nanoparticles. J Control Release 132: 171–183. Levi V, Flecha FLG. 2002. Reversible fast-dimerization of bovine serum albumin detected by fluorescence resonance energy transfer. Biochimica et Biophysica Acta 1599: 141–148. Liu Z, Lin Z, Zhen Z, Sun T, Zu Y. 2012. Assembly of singlestranded DNA onto HOPG surface at different temperature: atomic force microscopy study. Scanning 34: 302–308. Liu Z, Tabakman S, Welsher K, Dai H. 2009. Carbon nanotubes in biology and medicine: in vitro and in vivo detection, imaging and drug delivery. Nano Res 2: 85–120. Liu Z, Zu Y, Fu Y, Zhang Z, Meng R. 2008. Assembling and imaging of his-tag green fluorescent protein on mica surfaces studied by atomic force microscopy and fluorescence microscopy. Microsc Res Tech 71: 802–809. Liu ZG, Zhao L, Zu YG, et al. 2013. Unusual DNA structures formed on bare highly oriented pyrolytic graphite surfaces studied by atomic force microscopy. Microsc Microanal 19: 544–552. Orasanu-Gourlay A, Bradley RH. 2006. Protein adsorption by basal plane graphite surfaces: molecular images and nanostructured films. Adsorp Sci Technol 24: 117–130. Peters T. 1995. All about albumin: biochemistry, genetics, and medical applications. San Diego, CA: Academic Press. pp 9–54. Quist AP, Bjorck LP, Reimann CT, Oscarsson SO, Sundqvist BUR. 1995. A scanning force microscopy study of human serum albumin and porcine pancreas trypsin adsorption on mica surfaces. Surf Sci 325: L406–L412. Scheider W, Dintzis HM, Oncley JL. 1976. Changes in the electric dipole vector of human serum albumin due to complexing with fatty acids. Biophys J 16: 417–431. Slayter EM. 1965. An electron microscope study of the conformational change in bovine serum albumin at low pH. J Mol Biol 14: 443–452. Sugio S, Kashima A, Mochizuki S, Noda M, Kobayashi K. 1999. Crystal structrue of human serum albumin at 2.5A resolution. Protein Eng 12: 439–446. Tran PA, Zhang L, Webster TJ. 2009. Carbon nanofibers and carbon nanotubes in regenerative medicine. Adv Drug Deliv Rev 61: 1097–1114. Yang K, Feng L, Shi X, Liu Z. 2013. Nano-graphene in biomedicine: theranostic applications. Chem Soc Rev 42: 530–547.
Adsorption of Human Serum Albumin Onto Highly Orientated Pyrolytic Graphite Surface Studied by Atomic Force Microscopy XIAO PENG,1,2,3 HAORAN FU,4† RUISI LIU,1,2,3 LIN ZHAO,1,2,3 YUANGANG ZU,1,2,3 FENGJIE XU,1,2,3
AND
ZHIGUO LIU1,2,3,*
1
State Engineering Laboratory of Bio-Resource Eco-Utilization, Harbin, People’s Republic of China Engineering Research Center of Forest Bio-preparation, Ministry of Education, Northeast Forestry University, Harbin, People’s Republic of China 3 Key Laboratory of Forest Plant Ecology of Ministry of Education, Northeast Forestry University, Harbin, People’s Republic of China 4 School of Chemistry and Environment, Beihang University, Beijing 100191, People’s Republic of China 2
Summary: It is important to know the adsorption behavior and assembly structure of human serum albumin (HSA) molecules onto a carbonaceous substrate for further application of carbon nanomaterials in biomedical field. Individual HSA molecules and oligmers (dimer and trimer) adsorbed onto HOPG surface have been imaged by atomic force microscopy (AFM). Individual HSA molecule appeared as an ellipsoid on HOPG surface with average length of 12.6, width of 6.5, and height of 1.9 nm when they were incubated at the physiological condition (pH 7.4). HSA molecules also can form the interconnected chains, uniform network, and monolayer by tuning the initial concentrations and adsorption time. Furthermore, HSA molecules can assemble into quite different network structures and irregular chains at pH of 2, 5, and 10. This study could expand our knowledge of the interactions between protein and carbonaceous surfaces. SCANNING 37:158–164, 2015. © 2015 Wiley Periodicals, Inc. Key words: human serum albumin (HSA), HOPG, atomic force microscopy (AFM)
Contract grant sponsor: Fundamental Research Funds for the Central Universities; Contract grant number: DL12DA02; Contract grant sponsor: National Public-welfare Research Fund for Industry of Forestry; Contract grant number: 201104002-3. Conflicts of interest: None. † These authors contributed equally to this work (co-first author). Address for reprints: Zhiguo Liu, Key Laboratory of Forest Plant Ecology of Ministry of Education, Northeast Forestry University, Harbin 150040, People’s Republic of China E-mail: [email protected] Received 19 November 2014; revised 30 December 2014; Accepted with revision 9 January 2015 DOI: 10.1002/sca.21193 Published online 13 February 2015 in Wiley Online Library (wileyonlinelibrary.com).
Introduction Human serum albumin (HSA) is the most abundant protein in human blood. Many studies have revealed that HSA has a high affinity to a wide range of materials such as metal ions, fatty acids, amino acids, and a lot of drug compounds. Recent studies further proved that HSA is a perfect carrier for many kinds of anticancer drug (Kratz, 2008). Recently, new class of carbon materials such as carbon nanotubes and graphene that is a single-atom thick, two-dimensional sheet of hexagonally arranged carbon atoms isolated from its three-dimensional parent material, graphite, have received considerable attention due to their potential applications in variety of fields. Biological and biomedical applications of these carbon nanomaterials are of great interest. Some in vitro and in vivo biomedical applications such as for drug delivery and in bone regeneration have been explored in the past few years (Liu et al., 2009; Tran et al., 2009). Especially for graphene and its derivatives such as graphene oxide (GO), reduced graphene oxide (RGO), and GO-nanocomposites in biomedical field, significant progresses have been made in recent years (Yang et al., 2013). It has been demonstrated that graphene can be used for anti-cancer drug delivery, cancer therapy, and biological imaging. For further biomedical applications of graphene, the most important issue is still the potential long-term toxicity concerns. Protein adsorption on graphene surfaces is believed to mediate cell uptake and thus toxic responses, and hence is a very important issue to investigate. However, there have been few studies of graphene–protein interactions. More research including microscopic observation is clearly needed on the interaction of proteins with graphene-family nanomaterials. The comprehensive assessment of adsorption behaviors of protein on graphene and its derivatives at the physiological condition is not only important to
Z. Liu et al.: Adsorption of HSA Onto HOPG
investigate their toxicology, but also critical to improve their performance in various biomedical imaging and therapy applications. When carbon nanomaterials enter into the circulatory system, the interactions between the HSA molecules and carbonaceous surfaces are inevitable. It is believed that the interactions are very crucial for the function and destination of carbon nanomaterials as well as the loaded cargo on their surfaces. Thus, it is important to know the adsorption behavior and assembly structures of HSA molecules onto a carbonaceous substrate. However, there are few reports about the behavior and structures of HSA molecules on carbonaceous surfaces. More recently, Bradley et al. explored HSA adsorption and imaging on the graphite surfaces (Orasanu-Gourlay and Bradley, 2006). In this study, we further extend to study molecular structure and adsorption behavior of HSA on a carbonaceous surface at different concentration and pH conditions. The adsorption of HSA at the concentration of 1–1,000 ng/ml and different pH condition onto highly orientated pyrolytic graphite (HOPG) surfaces was investigated using atomic force microscopy (AFM). HOPG, a well defined carbonaceous substrate with large and atomically flat terraces can be used as a surrogate for the various carbonaceous materials to study protein structures on their surfaces. Adsorption from solution at relatively low concentration (1 ng/ml) allowed individual protein molecules to be imaged. When the adsorption was performed at 10 ng/ml or more high concentration, a protein network and monolayer can be obtained.
Materials and Methods HSA was obtained from Sigma–Aldrich (A9731, lyophilized powder, 96%). HOPG of ZYB grade was provided by MikroMasch, Estonia. The water used in all experiments was purified by a MilliporeMilli-Q system, and its resistivity was >18 MV/cm. All other reagents are analytical grade and used as-received without further purification. 1–1000 ng/ml of HSA solutions under different pH condition were prepared from its pure powder and different buffers. The HSA solutions at pH 7.4, 5.0 were prepared from 0.2 M phosphate buffer and acetate acid buffer, respectively; protein solutions at pH 2.0 and 10.0 were obtained by further adjusting of phosphate buffer and acetate acid buffer with HCl and NaOH solution, respectively. For adsorption of HSA samples onto HOPG surfaces, typically 40 ml of the asprepared protein solutions were deposited on the freshly cleaved HOPG surface and incubated for a predefined time. After incubation, the protein droplet was removed and HOPG surfaces washed by pure water at least three times. Finally, the HOPG sample was dried in air or under a flow of pure Ar gas before AFM imaging.
159
All AFM images were obtained by a PicoPlus II AFM system from Molecular Imaging, Inc. (Tempe, AZ,). The scanner was calibrated with a standard reference of 2 mm pitch lines with a depth of 200 nm provided by Molecular Imaging, Inc., NSC35 type silicon tips (radius of curvature is less than 10 nm) with a nominal force constant of 4.5 N/m provided by MikroMasch, Estonia were used. Original AFM images were obtained at 512 512 pixel. The dimensions of HSA topography were measured manually by using the PicoScan5.3.3 software. Three-dimensional representation of the HSA was performed with Scanning Probe Image Processor (Image Metrology ApS, Lyngby, Denmark)
Results and Discussion Figure 1 shows the representative AFM topographic and phase image of HSA on HOPG surface. Some wellseparated ellipsoidal particles can be identified. They were considered to be the individual HSA molecules based on their topography and dimension. In addition, there are some dimer and trimer of HSA as the arrows shown in Figure 1(a), which were further confirmed by their corresponding phase image as shown in Figure 1 (b). The phase image reflects the phase lag of the cantilever oscillation relative to the signal sent to the cantilever’s piezo driver, which provides the adhesion and visco-elasticity information of the surface and can be used for identifying different surfaces. HSA molecules and HOPG surfaces can be differentiated from their contrast in the phase images, in which HSA molecule appears darker, whereas HOPG surfaces appear bright. This result confirms that the observed objects in the AFM images are HSA molecules rather than artifacts on the HOPG surface. Previous electron microscope observation and calculated mode indicated that bovine serum albumin (BSA) tend to form dimer and trimers (Bloomfield, ’66; Slayter, ’65). More recently, self-association of BSA has been further investigated by fluorescence resonance energy transfer (Levi and Flecha, 2002). It is revealed that association between BSA monomers occurs through a reversible path that involves specific interactions between the protein molecules (Levi and Flecha, 2002). The observed HSA oligomers in present study revealed that HSA also tend to self-associate to different oligomers, which is very similar to those of BSA. Figure 2(a) and (b) shows the typical topographic and three-dimensional image of a HSA molecule on the HOPG surface. It can be seen that HSA appeared as an ellipsoid on HOPG surface.The average dimensions of these individual HSA molecules were measured. The average length of long axis is 31.3 4.2, short axis is 22.8 2.8, and average height is 1.91 0.22 nm (50 well separated individual HSA molecules were measured for average). The actual dimension of HSA should
160
SCANNING VOL. 37, 2 (2015)
Fig. 1. The representative AFM topographic and phase image of HSA (1 ng/ml incubated for 72 min) on HOPG surface. (a) Topographic image; (b) phase image.
Fig. 2. The typical AFM topographic and three-dimensional image of a HSA molecule on the HOPG surface. (a) Topographic image; (b) three-dimensional image.
Z. Liu et al.: Adsorption of HSA Onto HOPG
be less than these values due to the well-known broadening effect caused by the finite probe tip radius in AFM imaging. The actual dimension of HSA can be estimated by a previously proposed formula to deconvolute the tip effects (Liu et al., 2008). The calculated length of HSA after deconvolution is 12.6 nm and the width is 6.5 nm when the AFM tip radius assumed to be 10 nm. This result is very close to those from scanning tunneling microscopy (STM) imaging by Feng et al. (Feng et al., ’89). They found that HSA dried on a graphite surface to appear as an elongated (6 12 nm) molecule showing three linear domains. Further investigation of the HSA molecule structure in Figure 2(a) and (b) indicated that the three homologous domains of HSA are aligned linearly like as three tennis balls in a can. The boundary of three domains can be identified as shown in Figure 2(b). The present structural mode of HSA is very similar to that of BSA proposed previously by Bloomfield (Bloomfield, ’66). Albumin structures in solution and their crystal structrues have been widely investigated in the past (Scheider et al., ’76; Carter et al., ’89; He and Carter, ’92; Sugio et al., ’99). There has been general consensus from many physical measurements that in solution, HSA and BSA have the cigar-shape form of an ellipsoid of revolution (Peters, ’95). However, X-Ray diffraction now shows that albumin in a crystal, at least fatty acidfree albumin has a triangular or heart-like shape (an equilateral triangle of 8 * 8 nm dimensions and 3 nm height) (Sugio et al., ’99). This favors the rounded rather than the elongated picture obtained by most physical chemical studies. The different pictures may arise from the flexibility of the albumin molecule, which allows it to assume different configurations when packed tightly in a crystal or floating free in solution, and their reconciliation should be enlightening about its behavior. Our present observation by AFM revealed that HSA at least on HOPG surface has an ellipsoidal shape, which is consistent with the reported structures in solution. Bradley et al. observed both round and oblong shape of HSA on HOPG surface in their recent report (OrasanuGourlay and Bradley, 2006) though the sample preparation processes between their and our work are different. Interestingly, triangular and oblong features of HSA have been observed on mica surface (Quist et al., ’95). Therefore, in despite of the difference of sample preparation and adsorption surfaces, HSA can be transformed into a similar isomer featured with an oblong shape. Protein concentration is an important factor to influence their self-association and assembly structures. When more high concentration of HSA was used for adsorption, we observed quite different HSA structures on HOPG surface. Figure 3(a) shows the representative AFM topographic image of HSA incubated from 10 ng/ ml solution. HSA form a very uniform network structure on the HOPG surface instead of separated individual HSA molecules. The height of the network structure can
161
be measured from the exposed pores on HOPG surface. Interestingly, the height of the network structure is consistent across the scanning area as indicated in the section analysis in Figure 3(a). The average height of the segments of network is 2.15 0.31 nm. In comparison with the dimension of individual HSA molecule as described above, we can conclude that the network structure is made up of the interconnected individual HSA. Furthermore, most pores exposed out of the network structure have a polygonal shape. This phenomenon can be attributed to the aligning ability of carbonaceous surfaces in which the top two or top three graphene layers were considered to be involved in the surface interaction (Liu et al., 2012; Liu et al., 2013). As we reported previously, both circular and linear DNA molecules tend to form hexagonal patterns on HOPG surface (Liu et al., 2013). Further increasing the HSA concentration to 1,000 ng/ml, the obtained HSA structure is shown in Figure 3(b). HSA formed a flat monolayer on which there were some big protein aggregates. The number and size of the exposed pores are dramatically decreased under this relatively high concentration in comparison with that at low concentration (10 ng/ml). This result revealed that the high concentration of HSA facilitates more proteins to be adsorbed on HOPG surface. We further investigated the influences of the adsorption time on HSA structures on HOPG surface. Figure 4 shows the AFM topographic image of HSA incubated from 10 ng/ml solution for 1 and 30 min, respectively. After 1 min incubation, HSA formed into some interconnected chains as shown in Figure 4(a). As increasing adsorption time to 30 min, a monolayer of HSA was formed as indicated in Figure 4(b). From the HSA images in Figure 4(a), (b), and Figure 3(a), it is indicated that the adsorption of HSA on HOPG surface is a time-dependent process. As the adsorption time was increased, HSA evolved from the interconnected chains, network structure to a monolayer. As revealed by previous studies, HSA molecule undergoes several well-recognized changes in conformation under different pH conditions (Peters, ’95). Besides the assembly structures of HSA at pH 7.4 as described above, the adsorption of HSA under other pH conditions has been investigated. Figure 5 shows the representative AFM topographic image of HSA (10 ng/ ml) incubated from the solutions at pH of 2.0, 5.0, and 10.0, respectively. At pH 2.0, HSA formed a network structure as indicated in Figure 5(a). The average height of the segments of network is 2.88 0.40 nm. At pH 5.0, a heterogeneous network was obtained as shown in Figure 5(b). The average height of the segments of network is 3.90 0.52 nm, which is higher than that at pH 2.0. Under pH 11.0, some irregular chains with average height of 2.7 0.48 nm were presented as illustrated in Figure 5(c). All these results indicated that the pH condition can influence HSA assembly structure.
162
SCANNING VOL. 37, 2 (2015)
Fig. 3. The representative AFM topographic image of HSA incubated from 10 ng/ml and 1,000 ng/ml solution for 15 min on HOPG surface, respectively. (a) 10 ng/ml; (b) 1,000 ng/ml.
pH 5 is a very important point for albumin conformation since both the isoionic point and the isoelectric point of albumin are around 5.0 (Peters, ’95). The isoionic point of albumins, the pH of a thoroughly deionized solution, is about pH 5.2. The isoelectric point, in contrast to the isoionic point, is the pH at which
Fig. 4.
the net charge of a molecule, including any bound ions, is zero. For undefatted albumin in 0.15 M NaCl the isoelectric pH is about 4.7 (Peters, ’95). Thus, the electrostatic interactions between the HSA are minimized around pH 5.0 whereas self-association of the HSA through hydrophobic interactions are enhanced.
The AFM topographic image of HSA incubated from 10 ng/ml solution for 1 and 30 min, respectively. (a) 1 min; (b) 30 min.
Z. Liu et al.: Adsorption of HSA Onto HOPG
163
Fig. 5. The representative AFM topographic image of HSA (10 ng/ml) incubated from the solutions at pH of 2, 5, and 10, respectively. (a) pH ¼ 2; (b) pH ¼ 5; (c) pH ¼ 10.
Therefore, the relative higher height of the segment of the network at pH 5.0 can be understandable. They should be composed of two or more HSA molecules rather than the interconnected individual HSA. The
topographic image in Figure 5(b) appears to support this speculation. At extremely pH condition of 2.0, the albumin is considered in its extended form (E form) wherein all
164
SCANNING VOL. 37, 2 (2015)
carboxyl groups have become protonated (Peters, ’95). The positive charges of lysine, arginine, and histidine residues would cause mutual repulsion between the domains and subdomains of the molecule. An unfolding occurs until the molecule appears to be as fully expanded as its disulfide bonding structure allows. A previous electron microscope study reported that BSA at pH 1.9 has a thread-like form with approximately 25 nm long and 2.1 nm in diameter (Slayter, ’65). In present study, HSA formed a network structure rather than a separated individual HSA molecule. A possible reason for network structure formation is that HSA molecules connected to each other by hydrogen bonding between the carboxyl groups of acid residues. At pH 10, HSA is considered in its basic form (B form) wherein imidizole is the main ionizable group (Peters, ’95). It is indicated that B isomerization is a structural fluctuation, a loosening of the molecule with loss of rigidity (Peters, ’95). Furthermore, the irreversible conformational changes may occur in basic conditions caused by disulfide interchange initiated by Cys-34. The observed irregular chains in Figure 5(c) implied that some domain of HSA have irreversible conformational changes, which is not so obvious at other pH conditions.
Conclusions Individual HSA molecules and oligmers (dimer and trimer) adsorbed on HOPG surface have been imaged by AFM. Individual HSA molecule appeared as an ellipsoid on HOPG surface with average length of 12.6, width of 6.5, and height of 1.9 nm when they were incubated at the physiological condition (pH 7.4). HSA molecules also can form the interconnected chains, uniform network, and monolayer by tuning the initial concentration and adsorption time. Furthermore, HSA molecules can assemble into quite different network structures and irregular chains at pH of 2, 5, and 10. This study could expand our knowledge of the interactions between protein and carbonaceous surfaces.
References Bloomfield V. 1966. The structure of bovine serum albumin at low pH. Biochemistry 5: 684–689. Carter DC, He X-M, Munson SH, et al. 1989. Three-dimensional structure of human serum albumin. Science 244: 1195–1198. Feng L, Andrade J, Hu C. 1989. Scanning tunneling microscopy of proteins on graphite surfaces. Scanning Microsc 3: 399–410. He XM, Carter DC. 1992. Atomic structure and chemistry of human serum albumin. Nature 358: 209–215. Kratz F. 2008. Albumin as a drug carrier: design of prodrugs, drug conjugates and nanoparticles. J Control Release 132: 171–183. Levi V, Flecha FLG. 2002. Reversible fast-dimerization of bovine serum albumin detected by fluorescence resonance energy transfer. Biochimica et Biophysica Acta 1599: 141–148. Liu Z, Lin Z, Zhen Z, Sun T, Zu Y. 2012. Assembly of singlestranded DNA onto HOPG surface at different temperature: atomic force microscopy study. Scanning 34: 302–308. Liu Z, Tabakman S, Welsher K, Dai H. 2009. Carbon nanotubes in biology and medicine: in vitro and in vivo detection, imaging and drug delivery. Nano Res 2: 85–120. Liu Z, Zu Y, Fu Y, Zhang Z, Meng R. 2008. Assembling and imaging of his-tag green fluorescent protein on mica surfaces studied by atomic force microscopy and fluorescence microscopy. Microsc Res Tech 71: 802–809. Liu ZG, Zhao L, Zu YG, et al. 2013. Unusual DNA structures formed on bare highly oriented pyrolytic graphite surfaces studied by atomic force microscopy. Microsc Microanal 19: 544–552. Orasanu-Gourlay A, Bradley RH. 2006. Protein adsorption by basal plane graphite surfaces: molecular images and nanostructured films. Adsorp Sci Technol 24: 117–130. Peters T. 1995. All about albumin: biochemistry, genetics, and medical applications. San Diego, CA: Academic Press. pp 9–54. Quist AP, Bjorck LP, Reimann CT, Oscarsson SO, Sundqvist BUR. 1995. A scanning force microscopy study of human serum albumin and porcine pancreas trypsin adsorption on mica surfaces. Surf Sci 325: L406–L412. Scheider W, Dintzis HM, Oncley JL. 1976. Changes in the electric dipole vector of human serum albumin due to complexing with fatty acids. Biophys J 16: 417–431. Slayter EM. 1965. An electron microscope study of the conformational change in bovine serum albumin at low pH. J Mol Biol 14: 443–452. Sugio S, Kashima A, Mochizuki S, Noda M, Kobayashi K. 1999. Crystal structrue of human serum albumin at 2.5A resolution. Protein Eng 12: 439–446. Tran PA, Zhang L, Webster TJ. 2009. Carbon nanofibers and carbon nanotubes in regenerative medicine. Adv Drug Deliv Rev 61: 1097–1114. Yang K, Feng L, Shi X, Liu Z. 2013. Nano-graphene in biomedicine: theranostic applications. Chem Soc Rev 42: 530–547.
