8078
Langmuir 2008, 24, 8078-8089
Immobilization of Functional Light Antenna Structures Derived from the Filamentous Green Bacterium Chloroflexus aurantiacus Arati Sridharan,† Jit Muthuswamy,† Jeffrey T. LaBelle,†,‡ and Vincent B. Pizziconi*,† Harrington Department of Bioengineering, Arizona State UniVersity, Tempe, Arizona 85287, and Biodesign Institute, Arizona State UniVersity, Tempe, Arizona 85281 ReceiVed NoVember 26, 2007. ReVised Manuscript ReceiVed April 15, 2008 The integration of highly efficient, natural photosynthetic light antenna structures into engineered systems while their biophotonic capabilities are maintained has been an elusive goal in the design of biohybrid photonic devices. In this study, we report a novel technique to covalently immobilize nanoscaled bacterial light antenna structures known as chlorosomes from Chloroflexus aurantiacus on both conductive and nonconductive glass while their energy transducing functionality was maintained. Chlorosomes without their reaction centers (RCs) were covalently immobilized on 3-aminoproyltriethoxysilane (APTES) treated surfaces using a glutaraldehyde linker. AFM techniques verified that the chlorosomes maintained their native ellipsoidal ultrastructure upon immobilization. Results from absorbance and fluorescence spectral analysis (where the Stokes shift to 808/810 nm was observed upon 470 nm blue light excitation) in conjunction with confocal microscopy confirm that the functional integrity of immobilized chlorosomes was also preserved. In addition, experiments with electrochemical impedance spectroscopy (EIS) suggested that the presence of chlorosomes in the electrical double layer of the electrode enhanced the electron transfer capacity of the electrochemical cell. Further, chronoamperometric studies suggested that the reduced form of the Bchl-c pigments found within the chlorosome modulate the conduction properties of the electrochemical cell, where the oxidized form of Bchl-c pigments impeded any current transduction at a bias of 0.4 V within the electrochemical cell. The results therefore demonstrate that the intact chlorosomes can be successfully immobilized while their biophotonic transduction capabilities are preserved through the immobilization process. These findings indicate that it is feasible to design biophotonic devices incorporating fully functional light antenna structures, which may offer significant performance enhancements to current silicon-based photonic devices for diverse technological applications ranging from CCD devices used in retinal implants to terrestrial and space fuel cell applications.
Introduction The functional immobilization of light antenna complexes and photosynthetic reaction centers (RCs) has shown to be a promising approach in the design and development of hybrid devices for a wide variety of bioelectronic applications ranging from biosensors to solar fuel cells to artificial visual prosthetics.1–8 The biophotonic components derived from nature possess the ability to photosynthesize over a wide dynamic range of photonic spectra and light intensities ranging from bright sunlight conditions to dim light intensities. Current photodetection systems, such as silicon-based solar cells and charge coupled devices (CCDs) that have high relevance in biomedical and biotechnological applications, lose their efficiency under low light intensity * To whom correspondence should be
[email protected]. † Harrington Department of Bioengineering. ‡ Biodesign Institute.
addressed.
E-mail:
(1) Lam, K. B.; Johnson, E. A.; Chiao, M.; Lin, L. J. Microelectromech. Syst. 2006, 15, 1243–1249. (2) Morishima, K.; Yoshida, M.; Furuya, A.; Moriuchi, T.; Ota, M.; Furukawa, Y. J. Micromech. Microeng. 2007, 17, S274-S279. (3) Trammell, S. A.; Wang, L.; Zullo, J. M.; Shashidhar, R.; Lebedev, N. Biosens. Bioelectron. 2004, 19, 1649–1655. (4) Nakamura, C.; Hasegawa, M.; Nakamura, N.; Miyake, J. Biosens. Bioelectron. 2003, 18, 599–603. (5) Brune, A.; Jeong, G.; Liddell, P. A.; Sotomura, T.; Moore, T. A.; Moore, A. L.; Gust, D. Langmuir 2004, 20, 8366–8371. (6) Chiao, M.; Lam, K. B.; Lin, L. J. Micromech. Microeng. 2006, 16, 2547– 2553. (7) Kuritz, T.; Lee, I.; Owens, E. T.; Humayun, M.; Greenbaum, E. IEEE Trans. Nanobiosci. 2005, 4, 196–200. (8) Lakhanpal, R. R.; Yanai, D.; Weiland, J. D.; Fujii, G. Y.; Caffey, S.; Greenberg, R. J.; Eugene de Juan, J.; Humayun, M. S. Curr. Opin. Ophthalmol. 2003, 14, 122–127.
conditions, especially in the blue range of the visible spectrum.9,10 The effort to study and mimic biological systems may elucidate novel design strategies for constructing high efficiency, low power photodetection systems. A major obstacle in developing biohybrid photodetectors lies at the interface, where maintaining the photonic functionality of the biological component upon immobilization to a substrate is a challenge. One strategy is to chemically modify naturally derived photosynthetic molecules, such as chlorophyll, into synthetically engineered interfaces to preserve the functionality of the molecule.11–14 While this method is highly promising, methods still need to be addressed for controlling photobleaching and also matching the performance and robustness of photosynthetic elements as found in nature. Another promising approach that would be beneficial is to study the immobilization of naturally derived photosynthetic supramolecular complexes and assess their functionality at the interface. Upon immobilization, however, these typically larger and more complex biophotonic components face serious impediments in terms of performance loss due to degradation and denaturation.15 The main challenge, therefore, is to preserve the photonic transducing capabilities of these biological components at the interface defined by the biological complex and the engineered substrate. In this paper, light antenna complexes without RCs derived from the filamentous thermophilic green bacterium Chloroflexus aurantiacus, known as chlo(9) Dandin, M.; Abshire, P.; Smela, E. Lab Chip 2007, 7, 955–977. (10) Yotter, R. A.; Wilson, D. M. IEEE Sens. J. 2003, 3, 288–303. (11) Amao, Y.; Yamada, Y. Langmuir 2005, 21, 3008–3012. (12) Amao, Y. Y. Yuriko Biosens. Bioelectron. 2007, 22, 1561–1565. (13) Furukawa, H.; Inoue, N.; Watanabe, T.; Kuroda, K. Langmuir 2005, 21, 3992–3997. (14) Balaban, T. S. Acc. Chem. Res. 2004, 38, 612–623. (15) Miyake, J.; Hara, M. Mater. Sci. Eng.: C 1997, 4, 213–219.
10.1021/la703691a CCC: $40.75 2008 American Chemical Society Published on Web 07/01/2008
Light Antenna Structures from C. aurantiacus
rosomes, are covalently immobilized and characterized for biophotonic performance on various substrates. Chlorosomes, which are ellipsoidal bodies with average dimensions of ∼100-200 nm length × 20-50 nm width × 10-20 nm height, are known to be one of the most energetically efficient supramolecular nanophotonic structures.16,17 These highly organized, nanoscaled “bags” are mainly composed of bacteriochlorophyll c (BChl-c) oligomers encased in a lipid monolayer. The precise supramolecular organization of BChl-c oligomers within the chlorosomal envelope, which lends the relatively high quantum yields, is still under debate.18,19 Evolutionarily, the photosynthetic apparatus of the Chloroflexus green bacterium is closer to that of the purple photosynthetic bacterium and supposedly derived the chlorosome structure through horizontal gene transfer from the green sulfur bacteria.20,21 In the natural system, light is captured by the self-assembled BChl-c oligomers in the chlorosome and transferred to a BChla/protein complex known as the base plate. This energy, in turn, is funneled to a quinone-based RC (similar to purple bacteria), resulting in an energy cascade that eventually drives a proton gradient in the bacterium. Unlike other photosynthetic apparatus that function with a pigment-protein interaction, chlorosomes derived from C. aurantiacus are unique in that while proteins are present in the chlorosome structure, proteins are not known to be essential for BChl-c oligomer assembly or its light-harvesting function in the bacterium. 22 The same is true for similar, but slightly larger, light antenna structures also known as chlorosomes derived from the Chlorobium (green sulfur bacterium) species, though these require Fenna-Matthews-Olson (FMO) metalloproteins for subsequent energy transfer to a Fe-S type RC.23 Given that the chlorosome of the C. aurantiacus functions in nature under extreme thermal conditions (∼70 °C) and is not reliant on proteins for energy capture, the immobilization of this complex structure may elucidate fundamental methods for efficient energy transfer in future studies. A variety of methods have been attempted to immobilize lightharvesting structures, such as the electrostatic layer-by-layer technique using poly-L-lysine,24 physical adsorption of lightharvesting structures on glass25 and self-assembled thiol-based films,26,27 or immobilization on hydrophobically treated surfaces.28 Groups working with purple photosynthetic bacterial, polypeptide-based light-harvesting structures, which are evolutionarily similar to the RCs from C. aurantiacus, also have had success in immobilization using electrostatic principles on aminated surfaces.29–32 While these physical immobilization methods may be advantageous in certain situations, a stronger (16) Olson, J. M. Photochem. Photobiol. 1998, 67, 61–75. (17) Blankenship, R. E. Spectrum 1996, 9, 2–6. (18) Fetisova, Z.; Freiberg, A.; Novoderezhkina, V.; Taisova, A.; Timpmann, K. FEBS Lett. 1996, 383, 233–236. (19) Psencı´k, J.; Ikonen, T. P.; Laurinma¨ki, P.; Merckel, M. C.; Butcher, S. J.; Serimaa, R. E.; Tuma, R. Biophys. J. 2004, 87, 1165–172. (20) Hofman-Marriot, M. F.; Blankenship, R. E. FEBS Lett. 2007, 581, 800– 803. (21) Olson, J. M.; Blankenship, R. E. Photosynth. Res. 2004, 80, 373–386. (22) Miller, M.; Simpson, D.; Redlinger, T. E. Photosynth. Res. 1993, 35, 275–283. (23) Frigaard, N.-U.; Chew, A. G. M.; Li, H.; Maresca, J. A.; Bryant, D. A. Photosynth. Res. 2003, 78, 93–117. (24) Saga, Y.; Kim, T.-Y.; Hisai, T.; Tamiaki, H. Thin Solid Films 2006, 500, 278–282. (25) Shibata, Y.; Saga, Y.; Tamiaki, H.; Itoh, S. Biophys. J. 2006, 91, 3787– 3796. (26) Kincaid, H. A.; Niedringhaus, T.; Ciobanu, M.; Cliffel, D. E.; Jennings, G. K. Langmuir 2006, 22, 8114–8120. (27) Ko, B. S.; Babcock, B.; Jennings, G. K.; Tilden, S. G.; Peterson, R. R.; Cliffel, D.; Greenbaum, E. Langmuir 2004, 20, 4033–4038. (28) LaBelle, J. T.; Pizziconi, V. B. U.S. Patent 7067293, 2006. (29) Iida, K.; Inagaki, J. I.; Shinohara, M.; Suemori, Y.; Ogawa, M.; Dewa, T.; Nango, M. Langmuir 2005, 21, 3069–3075.
Langmuir, Vol. 24, No. 15, 2008 8079
and more stable interconnection is desired between the chlorosome light antenna structure and the substrate, especially in the presence of components that may facilitate desorption from the surface. In envisioning potential bioelectronic and photoelectrochemical applications, chlorosome immobilization achieved by way of a covalent bond may be required for sufficient and sustained energy transfer. Covalent immobilization techniques that have been attempted with purple photosynthetic bacterial RCs include using genetically modified polyhistidine tags on the polypeptides present in the L and H components of the RCs.3,33 Other groups have immobilized whole spinach thylakoids by directly platinizing the RC component to a metal substrate.34–36 These works show that covalent immobilization of RCs need not be detrimental to the light-harvesting function. However, to our knowledge, covalent immobilization with light antennas derived from the green bacteria has not yet been attempted. In this study, a more generic immobilization scheme is presented where chlorosomes without RCs were immobilized on 3-aminoproyltriethoxysilane (APTES) treated surfaces activated with glutaraldehyde. The utilization of this technique is plausible with chlorosomes, since proteins are known not to be essential to the light-capturing function of the chlorosome. The same technique could in principle be used for chlorosomes derived from the green sulfur bacterium as well. Here the immobilized chlorosomes were characterized for structural and functional integrity using absorbance and emission spectroscopy, selected electron microscopic techniques, and electrochemical impedance spectroscopy (EIS). The study presented herein demonstrates that the covalent immobilization technique maintains the structural and functional integrity of the chlorosome ultrastructure. In order to tap into the tremendous energy-capturing capacity of the light antenna structure, which occurs at high quantum yields (>90%), the immobilization of the chlorosome holds much promise not only in the development of biohybrid devices but also in further elucidating its fundamental mechanism of energy transfer.
Materials and Methods Chlorosome Isolations and Growth of Bacteria. J-10-Fl C. aurantiacus bacteria were grown at 300-500 lx using an incandescent light source in a custom-made incubator at 55 °C under anaerobic conditions. Chlorosomes were isolated using a similar protocol as Feick and Fuller.37 Briefly, green-colored cultures were harvested by centrifuging at 3000g for 60 min after 7 days of growth (late exponential phase). The pellet was resuspended in 2 M sodium thiocyanate buffer at a (w/v) ratio of 1 mg of bacteria to 4 mL of buffer. After homogenizing the resuspended bacteria in the cold 10 times, the bacteria were French-pressed (Thermo IEC, Model OMFA078A, Needham Heights, MA) at 20 000 psi three times to obtain fragments containing chlorosomes. The resulting mixture was then ultracentrifuged at 100 000g for 18 h at 4 °C in a 5%-40% sucrose gradient. The gradient then was aliquoted into 1 mL fractions (30) Suemori, Y.; Fujii, K.; Ogawa, M.; Nakamura, Y.; Shinohara, K.; Nakagawa, K.; Nagata, M.; Iida, K.; Dewa, T.; Yamashita, K.; Nango, M. Colloids Surf., B 2007, 56, 182–187. (31) Suemori, Y.; Nagata, M.; Nakamura, Y.; Nakagawa, K.; Okuda, A.; Inagaki, J.-i.; Shinohara, K.; Ogawa, M.; Iida, K.; Dewa, T.; Yamashita, K.; Gardiner, A.; Cogdell, R., J.; Nango, M. Photosynth. Res. 2006, 90, 17–21. (32) Valiokas, R.; Vaitekonis, S.; Klenkar, G.; Trinkujnas, G.; Liedberg, B. Langmuir 2006, 22, 3456–3460. (33) Das, R.; Kiley, P. J.; Segal, M.; Norville, J.; Yu, A. A.; Wang, L.; Trammell, S. A.; Reddick, L. E.; Kumar, R.; Stellacci, F.; Lebedev, N.; Schnur, J.; Bruce, B. D.; Zhang, S.; Baldo, M. Nano Lett. 2002, 4, 1079–1083. (34) Lee, I.; Lee, J. W.; Greenbaum, E. Phys. ReV. Lett. 1998, 79, 3294–3297. (35) Lee, J. W.; Lee, I.; Greebaum, E. Biosens. Bioelectron. 1996, 11, 375– 387. (36) Ciobanu, M.; Kincaid, H. A.; Jennings, G. K.; Cliffel, D. E. Langmuir 2005, 21, 692–698. (37) Feick, R. G.; Fitzpatrick, M.; Fuller, R. C. J. Bacteriol. 1982, 150, 905– 915.
8080 Langmuir, Vol. 24, No. 15, 2008 and characterized for the chlorosome-containing fraction via absorbance spectroscopy. The chlorosome-containing fractions were pooled and dialyzed five times into 50 mM Tris buffer (pH 8.0) for storage. The number of chlorosomes was determined using absorbance spectroscopy and an empirical relationship based on a modified ASTM standard method derived by LaBelle.38 The average size and hydrodynamic radius of chlorosomes in solution were also modeled and characterized by photocorrelation spectroscopy, also known as dynamic light scattering (DLS) (Wyatt Technologies, Santa Barbara, CA). The measured hydrodynamic radius (Rh) of the spherical model of the chlorosome was then used to estimate the elliptical dimensions of its native structure by equating their volumes using the geometric relationship 4/3π(abc) ) 4/3πRh3, where a, b, and c are the corresponding elliptical radii representing the length, width, and height of the average chlorosome and Rh represents the hydrodynamic radius. Immobilization of Chlorosomes. Microscope coverslips were bought from VWR, and indium tin oxide (ITO) coated conductive glass was purchased from Delta Technologies Limited (Stillwater, MN). Surface treatment protocols using APTES and glutaraldehyde were modified from the work of Cass and Ligler.39 The substrates were cleaned in 50/50 v/v 1 M hydrochloric acid and 30% hydrogen peroxide and then rinsed in distilled water six times and air-dried. The surface was then incubated at room temperature in 5% APTES for 1 h and cured at 80 °C overnight in enclosed boxes. Next the substrates were treated with 2.5% glutaraldehyde for 1 h and rinsed in distilled water three times to remove excess unreacted reagents. Isolated chlorosomes in solution derived from C. aurantiacus, as described in the previous subsection, were adsorbed onto the treated substrates overnight at 4 °C. The resulting substrates were then washed six times with distilled water to remove any unbound chlorosomes, dried, and stored in the dark. The various phases of the substrate surface treatments were characterized by equilibrium contact angle goniometry using the static sessile drop method with a 3 µL drop of distilled water using a Rame´-Hart goniometer (Netcong, NJ). The angles were captured and calculated using a pixel-based imaging analysis by the built-in Dropimage imaging software. The Dropimage software was used to estimate surface energiesbasedonacombinationofYoung-LaplaceandYoung-Dupre´ equations along with other known fits with a built-in library of parameters for known surfaces and liquids. Characterization of Chlorosome-Immobilized Substrates. Imaging of Chlorosome-Immobilized Surfaces. Immobilized chlorosomes were characterized for structural integrity and morphology using field emission scanning electron microscopy (FESEM, Leica Cambridge Stereoscan 360FE). Samples were prepared fresh before each run, sputtered with pure platinum for 1 min in a high vacuum chamber, and imaged. Contact-mode atomic force microscopy (AFM) was performed using a Digital Instruments scanner and subsequent chlorosome size analysis was done using the built-in Nanoscope software. Confocal imaging of chlorosome-immobilized microscope slides was done using the Leica SP2 multiphoton scanning laser microscope housed in the Keck Bioimaging Laboratory at Arizona State University. The samples were stimulated with an excitation wavelength of 488 nm, and fluorescence emission was filtered and collected from 700 to 900 nm. Subsequent image analysis was done using Adobe Photoshop software. Spectroscopic Characterization. Absorbance spectroscopy was done from 400 to 900 nm wavelengths on the chlorosomeimmobilized substrate using the appropriate substrate (ITO or coverslip) as reference. The resulting absorbance data were recorded and analyzed using Excel analysis software. Absorbance spectroscopy of chlorosomes immobilized on ITO in the presence of oxidants was done using 50 µM potassium triiodide/potassium iodide in 50 mM Tris-base buffer. The immobilized chlorosome electrodes were exposed to the oxidant for ∼1-2 min and the absorbance was measured specifically at 740 nm, where Bchl-c aggregates absorb. To reverse the oxidation of the Bchl-c aggregates, the electrodes (38) LaBelle, J. T. Dissertation, Arizona State University, 2001. (39) Cass, T.; Ligler, F. S. (Eds.) Immobilized Biomolecules in Analysis; Oxford University Press: New York, 1998.
Sridharan et al. were rinsed in distilled water six times for ∼1 min per wash. The absorbance spectra were recorded after each rinse, until a maximum value was observed after the fifth rinse, indicating a recovery of the 740 nm peak. The peaks were stable in the absence of oxidants for at least ∼1 month. Emission spectroscopy was done using a custom modified setup of a Shimadzu fluorescence spectrometer fitted with an IR-enhanced PMT (Columbia, MD). Essentially the thin films of immobilized chlorosomes on substrates were stimulated with an excitation wavelength of 470 nm at an angle of 45° and the resulting fluorescent emission of the immobilized chlorosomes was captured by a sensitive photodetector that was subsequently recorded and analyzed using Excel software. Electrochemical Characterization. EIS and chronoamperometry were conducted using a CH-660 (CH Instruments, Austin, TX) electrochemical station in a custom electrochemical cell. A conductive ITO glass, with or without chlorosomes, served as the working electrode, platinum metal was the counter electrode, and silver/ silver chloride (BASi, West Lafayette, IN) in 3 M KCl served as the reference. The working and counter electrodes were placed 0.5 cm apart for the scan. For the EIS experiments, the potentiostat varied frequencies from 1 to 100 000 Hz with a 5 mV amplitude and measured their respective impedances. The resulting data were then modeled using equivalent electrical circuit models in ZSimpwin (EChem Software, Ann Arbor, MI). For chronoamperometric experiments, chlorosomes were immobilized on ITO electrodes similar to the manner described above using APTES and glutaraldehyde. Electrochemical cells were constructed similar to the configuration for EIS experiments. Tris-base (pH 8.0) buffer at 50 mM and 50 µM potassium triiodide/potassium iodide in 50 mM Tris-base buffer were used as electrolyte solutions for the electrochemical cell. Since the iodine-based Tris buffer electrolyte and plain Tris buffer electrolyte inherently shift the base current properties of the entire electrochemical cell, all cells were assessed for voltage bias that gave the maximal current response using cyclic voltammetry. For the Tris buffer cells, it was found to be 0.7 V and for iodinebased electrolyte cells it was 0.4 V. The experimental design was such that the same pair of electrodes, i.e., bare, APTES-coated, and chlorosome-immobilized electrodes, were used to assess the current properties first in Tris base electrolyte and subsequently the iodinebased Tris buffer. The electrodes were rinsed to remove oxidants and were found to recover their amperometric response in Tris buffer (data not shown).
Results and Discussion Structural Characterization. Chlorosomes were immobilized onto glass coverslips using a covalent immobilization technique with APTES and glutaraldehyde as described in more detail in the Materials and Methods section. Contact angle measurements, taken at various stages during the immobilization procedure using a Rame´-Hart goniometer, are shown in Figure 1. It can be seen that sequential treatment of APTES and then glutaraldehyde dramatically changes the contact angle and subsequently the surface energy, based on estimates from the build-in Dropimage software. As previously noted, these were estimates ofsurfaceenergiesderivedfromYoung-LaplaceandYoung-Dupre´ equations along with other known fits with a built-in library of parameters for known surfaces and liquids. The initial low contact angle values obtained for bare glass were thought to be due to the substantial cleaning of the substrate, rendering the surface hydrophilic. Also shown in Figure 1 is the water contact angle of ∼83° obtained after treatment with APTES, which is comparable to values obtained in the literature.40–44 The (40) Howarter, J. A.; Youngblood, J. P. Langmuir 2006, 22, 11142–11147. (41) Krishnan, A.; Liu, Y.-H.; Cha, P.; Woodward, R.; Allara, D.; Vogler, E. A. Colloids Surf., B 2005, 43, 95–98. (42) Wang, Y. P.; Yuan, K.; Li, Q. L.; Wang, L. P.; Gu, S. J.; Pei, X. W. Mater. Lett. 2005, 59, 1736–1740.
Light Antenna Structures from C. aurantiacus
Langmuir, Vol. 24, No. 15, 2008 8081
Figure 1. Contact angle analysis of sequential surface treatments in immobilization of chlorosomes. Water contact angles were measured using a Rame`-Hart goniometer and surface energies were estimated using its built-in software Dropimage. Representative surface treatments on glass coverslips (a) are shown. APTES treatment of glass surface (b) generated the highest contact angles, while subsequent glutaraldehyde (c) and chlorosome immobilization (d) rendered the surface more hydrophilic. Both surface treatments had higher contact angles compared to the bare glass.
corresponding estimated surface energy for the APTES-coated surface was significantly decreased compared to the bare glass surface, indicating the presence of a highly hydrophobic surface. The high contact angle obtained suggested that multiple layers of APTES were deposited onto the substrate, where the overall hydrophobic effect is a cumulative result of the domination of the alkyl and silyl groups. With subsequent treatment of the substrate with glutaraldehyde, the estimated surface energy was slightly increased, possibly due to the polar carbonyl group present on the molecule. Note that the addition of glutaraldehyde does not render the surface totally hydrophilic, since the APTES is still present on the surface. The data presented thus far conform to similar results obtained from standard APTES/glutaraldehyde immobilization techniques typically used for protein immobilization as presented elsewhere.44,45 In the final step of immobilization, in which a layer of covalently bound chlorosomes was self-assembled on the substrate, a moderate increase in the estimated surface energy was observed upon chlorosome introduction relative to estimated surface energies from the previous surface treatments. This differential increase in surface energy could possibly be due to a combination of the inherent surface charges present on the chlorosome body with its proteins and pigments together with an expected increase in surface roughness. It should be noted that the chlorosomeimmobilized substrate still had a lower surface energy compared to cleaned bare glass. To further verify that intact chlorosomes were immobilized on the surface, results obtained from field emission scanning electron microscopy (FESEM) images revealed numerous rounded, nanoscale particles adhered to the surface, as shown (43) Han, Y.; Mayer, D.; Offenhausser, S.; Ingebrandt, S. Thin Solid Films 2006, 510, 175–180. (44) Qin, M.; Houa, S.; Wang, L.; Feng, X.; Wang, R.; Yang, Y.; Wang, C.; Yu, L.; Shao, B.; Qiao, M. Colloids Surf., B 2007, 60, 243–249. (45) Wang, Z.-H.; Jin, G. J. Immunol. Methods 2004, 285, 237–243.
Figure 2. FESEM of chlorosome-immobilized glass substrates. (a) Representative glass substrate with covalently immobilized chlorosomes with an average surface concentration of∼50 chlorosomes/µm2. (b) Representative glass substrate with only APTES as a comparative control for background with the same scale as part a. (c) Higher resolution image of covalently immobilized chlorosomes at above surface concentration. The high degree of background surface roughness was due to the cracking of the APTES surface under high voltage conditions (15 kV).
in Figure 2a. For the sake of clarity and contrast between the background and the chlorosome adhered layer, images were obtained using a surface coverage of ∼50 chlorosomes/µm2. The value of ∼50 chlorosomes/µm2 used in the samples was estimated using an empirical relationship developed by LaBelle between chlorosome absorbance at 650 nm and physical chlorosome-particle counts.38 Independently, we also estimated the average number of chlorosomes present on the surface using Fluorchem software (Alpha Innotech, San Leandro, CA) to analyze the FESEM images, and determined ∼47-52/µm2 chlorosomes present on the surface with an approximate error of 4%-6%. Therefore, the FESEM images suggest that almost all the chlorosomes that were adsorbed on the surface were bound to the surface.
8082 Langmuir, Vol. 24, No. 15, 2008
Figure 3. Contact-mode AFM imaging of immobilized chlorosomes on glass substrates. (Left) Representative image of covalently immobilized chlorosomes taken using contact mode AFM. (Right) A higher resolution image of a typical single immobilized chlorosome. Note that the typical native ellipsoidal shape of chlorosomes is maintained upon immobilization. The average dimensions measured for 12 samples were (123 ( 11.6 nm) length × (43.8 ( 1.2 nm) width × (11.2 ( 1.8 nm) height.
An unexpected result of the FESEM imaging of chlorosomes was the significant background surface roughness, as seen in the higher resolution scan represented in Figure 2c. This could be due to the high voltages (15 kV) required to obtain nanoscale resolution in the image, which probably led to breaking and cracking of the underlying APTES thin film. Background surface roughness was especially prominent at low chlorosomal surface concentration levels (∼5 chlorosomes/µm2, data not shown). Another unexpected result was the rounded geometry of the immobilized chlorosomes, as seen in Figure 2a,c. The estimated range of the chlorosome diameters in the FESEM images depicted in Figure 2b was ∼110-150 nm. This range in chlorosome diameter suggested a much greater chlorosome size based upon its native elliptical geometry compared to that reported by others earlier.16 Although this result still fell within the higher end of reported values for native chlorosome dimensions, it was quite possibly due to an artifact of the platinum coating sputter procedure used during sample preparation, as well as a result of the high-vacuum environment to which the chlorosome samples were subjected during imaging. Similar effects of shape distortion and increases in size were also observed by Saga et al. and Martinez-Planells et al.46,47 To verify if the original chlorosomes in the stock solution were ellipsoidal in shape, another sample of chlorosomes from the same stock solution were immobilized using the same technique and imaged via contact-mode AFM. A representative AFM image of the immobilized chlorosome is shown in Figure 3, where it clearly depicts its more familiar ellipsoidal structure. This result suggests that the rounded shape distortions seen in the FESEM images in Figure 2 were likely an artifact of sample preparation and imaging conditions. Using the AFM-Nanoscope analysis software tools, average sizes of chlorosomes immobilized in the same fashion as those imaged using FESEM were determined to be (123 ( 11.6 nm) × (43.8 ( 1.2 nm) × (11.2 ( 1.8 nm), which represent an ellipsoid geometry with major and minor axis that are well within the range of length scale reported in literature. To ensure that the average size determinations derived for immobilized chlorosomes were representative of the chlorosomes found in solution, DLS was done on a freshly isolated chlorosome (46) Martinez-Planells, A.; Arellano, J. B.; Borrego, C. A.; Lopez-Iglesias, C. F. G.; Garcia-Gil, J. S. Photosynth. Res. 2002, 71, 83–90. (47) Saga, Y.; Wazawa, T.; Nakada, T.; Ishii, Y.; Yanagida, T.; Tamiaki, H. J. Phys. Chem. B 2002, 106, 1430–1433.
Sridharan et al.
Figure 4. Dynamic light scattering of chlorosomes in solution. Shown is the estimated hydrodynamic radius of the chlorosome sample, which was 18.5 ( 2.6 nm.
sample, similar to procedures reported by Montan˜o et al.48 DLS, which estimates particle size based on the diffusive properties of the sample solution, calculated that the average hydrodynamic radius of a chlorosome was 18.5 ( 2.6 nm, assuming a spherical particle model as shown in Figure 4. This result compared well to the calculated hydraulic radius, Rh, of 19.6 nm obtained from direct chlorosome size measurements made from our AFM studies, which is well within the margin of error of the estimated radius using DLS (18.5 ( 2.6 nm). Taken together, these data further support the conclusion that the overall size and structural integrity of the individual chlorosomes were retained when covalently bound to a substrate according to the method reported herein. Spectroscopic Assessment of Chlorosome Functionality. Additional results obtained from absorbance spectroscopy studies confirmed the ability of the BChl-c aggregates (the main pigment in the bacterial light antenna structure) to function in a photosensitive manner that further verifies that the structural and functional integrity of the covalently immobilized chlorosomes was maintained. Covalently immobilized chlorosomes subjected to wavelengths from the visible spectrum (400-900 nm) were assessed for absorbance. The absorbance spectra of immobilized chlorosomes on ITO substrates were compared to that of chlorosomes in solution, as shown in Figure 5a. As seen, the characteristic peaks of the BChl-c aggregate at 740 nm and the Soret at 465 nm were preserved. It is also noted that the RC, which would typically absorb at 808/866 nm, was not observed in the samples and thus confirms that no RCs were present. Additionally, the immobilized chlorosomes in this study did not show any obvious sign of degradation, such as the dissolution of the BChl-c aggregate (characterized by the 740 nm peak) to BChl-c monomer (characterized by 670 nm peak). This was highly suggestive that the immobilization method with glutaraldehyde utilized in this study did not disturb the internal architecture of the chlorosomal pigments. Moreover, the 795 nm peak, characteristic of BChl-a pigments in the chlorosome base plate, was also present, suggesting that the base plate was not denatured. In contrast to the absorbance spectrum of the sample in solution, where the slope in the 550-700 nm region was typically constant, the linear upward trend in the 550-700 nm region for the immobilized chlorosome was possibly due to surface scatter from the substrate. This interference from the scatter may have led to the slight decrease in the Soret peak for the immobilized chlorosome compared to that in solution. Also, the ratio of Qy band to base plate (Abs 740/Abs 795) was 10:1 in both samples, suggesting that nonspecific degeneration of neither BChl-c aggregates nor BChl-a took place. Similar (48) Montano, G. A.; Bowen, B. P.; LaBelle, J. T.; Woodbury, N. W.; Pizziconi, V. B.; Blankenship, R. E. Biophys. J. 2003, 85, 2560–2565.
Light Antenna Structures from C. aurantiacus
Figure 5. (a) Comparative absorbance spectroscopy of immobilized (thick line) and free-floating chlorosomes in solution (thin line). Shown are chlorosomes that were immobilized on conductive glass surfaces (ITO) with 100% coverage. Similar absorbance spectra were obtained with nonconductive glass surfaces (data not shown). Note no degradation of the Bchl-c aggregate within the chlorosome (characterized by a 670 nm absorbance peak) is observed upon immobilization. Also no reaction centers are present, which would be characterized by absorbance peaks at 810 and 866 nm. The characteristic peaks for chlorosomal Bchl-c aggregates at 465 nm (Soret) and at 740 nm and the chlorosomal Bchl-a base plate peak at 795 nm are preserved upon immobilization. (b) Comparative absorbance spectroscopy of immobilized chlorosomes on glutaraldehyde activated substrates (thick, black line), chlorosome-treated nonactivated APTES-treated surfaces (thin, black line), and chlorosometreated plain glass coverslip surfaces (gray line). The characteristic peaks for chlorosomal Bchl-c aggregates at 465 nm (Soret) and at 740 nm and the chlorosomal Bchl-a base plate peak at 795 nm are preserved upon immobilization on silanized surfaces activated with glutaraldehyde. These peaks are not seen on nonactivated silanized or plain glass samples exposed to chlorosomes, indicating that a stable immobilization did not occur on these controls.
absorbance spectra of immobilized chlorosomes were obtained on nonconductive glass (data not shown). In addition, controls of APTES-silanized surfaces and plain glass surfaces exposed to chlorosome-containing solutions showed that chlorosomes do not strongly adsorb to these surfaces, as indicated by absorbance measurements. Both controls showed no stable absorbance at the characteristic peaks as would be expected for the chlorosome at 462, 740, 795 nm if it had been adsorbed (Figure 5b). Previous work testing chlorosome adsorption on plain glass using laser confocal microscopy corroborated this result.38 Chlorosomes bound on APTES-treated surfaces activated with glutaraldehyde did not show a change in absorbance with rinsing, as seen in the reported study. The stable absorbance spectra of chlorosomes on glutaraldehyde-activated silanized surfaces suggested that chlorosomes were strongly bound to these surfaces and did not desorb with rinsing as compared to the controls. To assess the immobilized chlorosome functionality further, the fluorescence activity of immobilized chlorosomes were qualitatively compared using confocal microscopy. Using an excitation wavelength of 488 nm (Ar laser line), the emitted photons from a patterned microscope slide were captured by a PMT with a long pass filter (>700 nm) at room temperature, where the chlorosomes were expected to fluoresce. As seen in Figure 6, the immobilized chlorosomes had ∼7 times higher photon count than the APTES-treated or untreated surface. The
Langmuir, Vol. 24, No. 15, 2008 8083
smaller degree of photon counts evident in the controls was likely due to noise and surface scatter. Therefore, the confocal images indicated that the fluorescence exhibited by chlorosomes in solution was not quenched upon immobilization. To further evaluate the fluorescent response of the immobilized chlorosome, emission spectra of chlorosomes in solution and those that were immobilized on the glutaraldehyde-activated silanized surface were measured at an excitation wavelength of 470 nm using a custom setup of the Shimadzu spectrofluorometer. The characteristic fluorescent peaks at ∼750 nm due to the BChl-c aggregates and at ∼805 nm due to BChl-a pigments in the base plate are seen in Figure 7. The presence of a peak at ∼805 nm was highly suggestive that energy transfer from the BChl-c aggregates to the base plate was maintained in the immobilized state of the chlorosome. Work by other groups showed that BChl-a fluorescence was quenched or significantly diminished where there was no spacer between the substrate and the chlorosome body.25 In this immobilization study, we estimated the minimum distance between the substrate and the chlorosome using the average bond lengths, and determined that the APTES-glutaraldehyde link formed at the minimum a ∼1.4 nm spacer. This length was consistent with intermolecular distance of less than 1.5 nm between the pigments in the base plate and the pigments in the P808/866 protein complex in their native state. This result suggests that a spacer g1.5 nm was needed to observe the fluorescent functionality of the base plate, which may become quenched at lesser distances due to Dexter- or Fo¨rster-type energy transfer mechanisms. It should also be noted that, although not experimentally verified in this study, the majority of chlorosomes covalently bound to the surface are expected to be oriented with their more hydrophobic base plate regions aligned toward the surface due to the hydrophobic nature of the APTES treatment. Additional solution studies of chlorosome suspensions using emission spectroscopy showed a concentration-dependent emission response, as shown in Figure 7a. A similar concentrationdependent photonic response for immobilized chlorosomes was observed in Figure 7b, where the 805 nm peak signal appeared to decrease to a 4-fold proportion from a 10-fold decrease in concentration change in a similar manner to the solution. This result suggested that the response of the immobilized chlorosomes was controllable to some extent by changing surface concentration. The concentration-dependent response was used to establish the upper and lower detection limits for fluorescence for this test system, where the upper limit is controlled by surface concentration with a maximum surface coverage of ∼1010 chlorosomes and the lower emission limit of ∼109 chlorosomes was determined by the detection limits of the PMT within the fluorimeter. Control experiments with just APTES surface treatments did not generate any emission response (data not shown). As seen in the data in Figure 7a,b, upon excitation at 470 nm, the Bchl-c aggregates fluoresce at 750 nm (Bchl-c) and the Bchl-a pigments in the base plate fluoresce at 805 nm, which represents the Fo¨rster energy transfer from the Bchl-c aggregates. This result further suggests that the chlorosomes have maintained their functionality. In further evaluation, comparison of the ratio of emission at 750 nm to that of 805 nm may give insight into the interactive energy transfer dynamics between the Bchl-c aggregates and the Bchl-a pigments. Comparison of emission ratios (750/805) for immobilized samples reveals a fairly consistent result of 0.54 for the upper concentration limit and 0.45 for the lower concentration for the immobilized chlorosome samples. Comparisons of the 750 nm/805 nm ratio of other similarly prepared samples reveal similar values (data not shown). This suggests that, in this set of experiments, the immobilization procedure maintained a similar
8084 Langmuir, Vol. 24, No. 15, 2008
Sridharan et al.
Figure 6. (a)-Confocal imaging of a patterned glass slide with immobilized chlorosomes. In comparison to bare glass and APTES-coated glass surfaces, immobilized chlorosomes visually had a higher photon count. Images were taken at 400× magnification. (b) Images converted to grayscale and analyzed using Photoshop. The percentage of photon counts was measured by the proportion of white pixels in the image. Areas with immobilized chlorosomes were higher in photon counts than the controls.
Figure 7. Comparative emission spectroscopy of (a) free-floating vs (b) immobilized chlorosomes on nonconductive glass. The thick lines represent a 10-fold chlorosome concentration decrease compared to data shown as thin lines. For immobilized chlorosome samples, the emission levels reached the detection limits of the PMT, resulting in unsmoothed data collection. Note that no reaction centers (characterized by an 870 nm emission peak) are present in either of the isolated chlorosome samples in solution or on a surface.
configuration of the chlorosomes on the surface. However, comparison of emission ratios from chlorosomes in solution results in 0.24 for the higher concentration and 0.55 for the lower concentration. This suggests that fluorescence quenching of Bchl-c aggregates from nearby chlorosomes in the solution with the higher concentration of chlorosomes may have occurred, even
though this does not affect the internal energy transfer from Bchl-c aggregates to Bchl-a pigments in the base plate. In contrast, the lower chlorosome concentration yields a higher ratio that was similar to the immobilized samples, suggesting that the distance between individual chlorosomes was such that quenching did not occur. A similar concentration-dependent quenching effect
Light Antenna Structures from C. aurantiacus
is not observed in the immobilized samples, possibly due to the relatively fixed orientation of the Bchl-c aggregates on the planar surface, where nearby chlorosomes are not expected to be within the same emission plane to cause quenching. Given the ability to control the biophotonic response of the immobilized chlorosome, as seen by the large Stokes shift from 470 to 750 and 805 nm, it is envisioned that the photonic device efficiencies, especially in the blue portion of the visible spectrum, could enhance current photodetection systems, as previously demonstrated by LaBelle in a related biohybrid system.38 For instance, the typical efficiency of commercial silicon-based photodetection systems is ∼15% for wavelengths greater than 550 nm. For lower wavelengths, i.e. blue regions, the efficiencies quickly fall to
