912
Langmuir 2009, 25, 912-918
Characterizing Antibody Specificity to Different Protein Morphologies by AFM Min S. Wang,† Andleeb Zameer,†,‡ Sharareh Emadi,† and Michael R. Sierks*,† Department of Chemical Engineering, Arizona State UniVersity, Tempe, Arizona 85287-6006, and Corinne Goldsmith Dickinson Center for Multiple Sclerosis and Department of Neurology, Mount Sinai School of Medicine, New York, New York 10029 ReceiVed August 8, 2008. ReVised Manuscript ReceiVed October 8, 2008 Protein misfolding and aggregation can lead to several neurodegenerative diseases including Alzheimer’s Disease (AD), Parkinson’s Disease (PD) and Huntington’s Disease (HD). While the respective proteins involved in each disease differ in their pathological effects and amino acid sequences, the aggregated forms all share a common cross β-sheet conformation. Substantial controversy exists over the roles of the different aggregate morphologies in disease onset and progression, and analytical tools such as morphology specific antibodies are needed to distinguish between the different protein morphologies in situ. Here we utilize atomic force microscopy (AFM) to characterize the binding of three single chain antibody fragments (scFvs) to different morphologies of R-synuclein (RS). From the topographic images generated using the AFM, we were able to show that one scFv bound all morphologies of RS, a second bound only oligomeric RS, and a third bound only fibrillar RS by comparing the height distribution of the different RS morphologies with and without addition of the different scFvs. These results demonstrate the versatility of the AFMbased technique as an easy tool to characterize specific antigen-antibody binding and the potential applications of scFvs as promising immunodiagnostics for protein misfolding diseases.
Protein folding is a relatively fast and efficient process whereby a protein self-assembles into its native structure and conformation. While there are many different folding pathways that proteins can utilize, thermodynamics require proteins to assume the conformation with the lowest available energy, defined by an energy landscape.1 Some proteins, however, have the tendency to fold into intermediate states where the polypeptide may settle into “kinetic traps” on the energy landscape1,2 and thus increase the risk of protein misfolding and protein aggregation.3 Such intermediate states have been implicated in a number of neurodegenerative diseases involving amyloidogenic proteins, including Alzheimer’s Disease (AD), Parkinson’s Disease (PD), bovine spongiform encephalopathy (mad cow disease), familial amyloid polyneuropathy, and Huntington’s Disease (HD).4-6 The respective misfolded proteins, while differing in their pathogenesis and amino acid sequences,7 share a common “crossβ” structured motif6 and form insoluble amyloid fibrils or plaques.4 In AD, the disorder is characterized by presence of β-amyloid (Aβ) plaques and neurofibrillary tangles (NFTs) in the brain.8 The pathological hallmark Lewy Bodies associated with PD are predominantly comprised of aggregated and insoluble fibrils of * Corresponding author. Address: Department of Chemical Engineering, Arizona State University, Tempe, AZ 85287-6006. Phone: 480-965-2828. Fax: 480-965-0037. E-mail:
[email protected]. † Arizona State University. ‡ Mount Sinai School of Medicine.
(1) Dill, K. A.; Chan, H. S. Nat. Struct. Biol. 1997, 4(1), 10–19. (2) Nevo, R.; Brumfeld, V.; Kapon, R.; Hinterdorfer, P.; Reich, Z. EMBO Rep. 2005, 6(5), 482–486. (3) Dobson, C. M. Protein Pept.Lett 2006, 13(3), 219–227. (4) Agorogiannis, E. I.; Agorogiannis, G. I.; Papadimitriou, A.; Hadjigeorgiou, G. M. Neuropathol. Appl. Neurobiol. 2004, 30(3), 215–224. (5) Cavagnero, S.; Jungbauer, L. M. Trends Biotechnol. 2005, 23(3), 157–162. (6) Soto, C.; Estrada, L.; Castilla, J. Trends Biochem. Sci. 2006, 31(3), 150– 155. (7) Lipfert, J.; Franklin, J.; Wu, F.; Doniach, S. J. Mol. Biol. 2005, 349(3), 648–658. (8) Lee, V. M. Y.; Balin, B. J.; Otvos, L.; Trojanowski, J. Q. Science 1991, 251(4994), 675–678.
R-synuclein (RS).9,10 For HD, the characteristic aggregates contain expanded polyglutamine (polyQ) repeats of the huntingtin protein.11 While numerous studies have implicated insoluble fibrils as the neurotoxic component in these diseases,12,13 an increasing number of recent studies suggest that other morphologies, including soluble oligomeric aggregates and protofibrils, are the primary neurotoxic species.14,15 A simple and convenient method to directly observe and characterize the binding activity of the antigen-antibody complex would be useful to probe the specific intermolecular interactions between a single chain variable domain antibody fragment (scFv) and its antigen. Atomic force microscopy (AFM) has long been used to characterize the aggregation mechanism and to describe the morphological details of various amyloid-like proteins, including RS,14,16-20 Aβ,21-23 insulin,24,25 and huntingtin pro(9) Sulzer, D. Nat. Med. 2001, 7(12), 1280–1282. (10) Volles, M. J.; Lansbury, P. T. Biochemistry 2003, 42(26), 7871–7878. (11) Bhattacharyya, A. M.; Thakur, A. K.; Wetzel, R. Proc. Natl. Acad. Sci. U.S.A. 2005, 102(43), 15400–15405. (12) Bellotti, V.; Mangione, P.; Merlini, G. J. Struct. Biol. 2000, 130(2-3), 280–289. (13) Spillantini, M. G.; Goedert, M.; Jakes, R.; Klug, A. Proc. Natl. Acad. Sci. U.S.A. 1990, 87(10), 3947–51. (14) Ding, T. T.; Lee, S. J.; Rochet, J. C.; Lansbury, P. T. Biochemistry 2002, 41(32), 10209–10217. (15) Lansbury, P. T.; Ding, T.; Lashuel, H.; Rochet, D. C. Abstr. Pap. Am. Chem. Soc. 2002, 223, C41-C42. (16) Apetri, M. M.; Maiti, N. C.; Zagorski, M. G.; Carey, P. R.; Anderson, V. E. J. Mol. Biol. 2006, 355(1), 63–71. (17) Conway, K. A.; Harper, J. D.; Lansbury, P. T. Nat. Med. 1998, 4(11), 1318–1320. (18) Conway, K. A.; Lee, S. J.; Rochet, J. C.; Ding, T. T.; Williamson, R. E.; Lansbury, P. T. Proc. Natl. Acad. Sci. U.S.A. 2000, 97(2), 571–576. (19) Lashuel, H.; Nowak, R.; Lansbury, P.; Petre, B.; Walz, T.; Wall, J.; Simon, M. J. Mol. Biol. 2002, 322(5), 1089–1102. (20) Zhang, F.; Lin, X. J.; Ji, L. N.; Du, H. N.; Tang, L.; He, J. H.; Hu, J.; Hu, H. Y. Biochem. Biophys. Res. Commun. 2008, 368(2), 388–394. (21) Harper, J. D.; Wong, S. S.; Lieber, C. M.; Lansbury, P. T., Jr Biochemistry 1999, 38(28), 8972–80. (22) Legleiter, J.; Czilli, D. L.; Gitter, B.; DeMattos, R. B.; Holtzman, D. M.; Kowalewski, T. J. Mol. Biol. 2004, 335(4), 997–1006. (23) Stine, W. B., Jr.; Snyder, S. W.; Ladror, U. S.; Wade, W. S.; Miller, M. F.; Perun, T. J.; Holzman, T. F.; Krafft, G. A. J. Protein Chem. 1996, 15(2), 193–203.
10.1021/la8025914 CCC: $40.75 2009 American Chemical Society Published on Web 12/17/2008
Characterization of Antibody Specificity
tein.26 Here we report an AFM-based study that characterizes the interactions between different morphologies of RS and single chain antibody fragments (scFvs) with different morphological specificities that were previously isolated in our laboratory including an anti-oligomeric,27 a pan-specific scFv28 and antifibrillar scFv29 by measuring the height changes of each sample. We have taken advantage of the fast and label-free surface characterization capabilities of the AFM for direct detection and quantification of specific antigen-antibody binding30-32 to detect and characterize binding interactions between scFv’s and specific protein morphologies associated with neurodegenerative diseases. Calculation of height distribution data with the AFM enables precise characterization of scFv binding to different protein aggregate morphologies.
Materials and Methods Materials. All chemicals were purchased from Sigma (SigmaAldrich, St. Louis, MO) and used as is unless otherwise specified. Production of rS. RS plasmid was generously donated by Dr. Michael Volles (Brigham and Women’s Hospital, Harvard Medical School), and RS was produced as previously described.33 Briefly, RS plasmid was transformed into competent E. coli cells BL21 (DE3, Invitrogen) using heat shock treatment by placing the cells on ice for 30 min, and then heated at 42 °C for 30 min, and finally placed on ice for another 3 min. An aliquot of 100 µL of the mixture was plated onto LB-amp agar plates (100 µg/ml of ampicillin) and grown overnight in a 37 °C incubator. Single colonies of BL21 (DE3) were prepared as previously described.34 Briefly, 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) was added to the culture and induced overnight at 30 °C. The overnight culture was centrifuged at 15 000 g for 20 min, and the pellet was resuspended in 4 mL TEN buffer (50 mM Tris-HCl (pH 8.0), 10 mM EDTA and 150 mM NaCl) and immediately frozen at -80 °C. Purification of rS. RS was purified as previously described.34 Briefly, the -80 °C frozen cells were boiled in a water bath at 100 °C for 10 min. After boiling, the cells were centrifuged at 10 000g for 15 min. The cell pellet was discarded, and the supernatant was transferred to a new tube where streptomycin sulfate (136 µL of a 10% solution/mL supernatant) and glacial acetic acid (228 µL per mL supernatant) were added and the mixture was gently mixed by inverting the tube a few times. The mixture was then centrifuged, and the pellet was discarded. The supernatant was precipitated with 1.7 M sodium citrate in a 1:1 (v/v) ratio and centrifuged at 10 000g for 10 min. The pellet was resuspended in 100 mM ammonium acetate in a 1:1 (v/v) ratio and vortexed. An equi-volume of 100% ethanol was added to the solution and pelleted. The pellet was washed twice in 100% ethanol, and the final precipitate was resuspended in 100 mM ammonium acetate and dialyzed against water. The concentration of RS was determined using BCA Protein Assay (Pierce, Rockford, IL). Aliquots of 200 µL were transferred into Epperdorf tubes (VWR, West Chester, PA) and stored at 4 °C until use. (24) Grudzielanek, S.; Velkova, A.; Shukla, A.; Smirnovas, V.; Tatarek-Nossol, M.; Rehage, H.; Kapurniotu, A.; Winter, R. J. Mol. Biol. 2007, 370(2), 372–84. (25) Jansen, R.; Dzwolak, W.; Winter, R. Biophys. J. 2005, 88(2), 1344–53. (26) Dahlgren, P. R.; Karymov, M. A.; Bankston, J.; Holden, T.; Thumfort, P.; Ingram, V. M.; Lyubchenko, Y. L. Nanomedicine 2005, 1(1), 52–7. (27) Emadi, S.; Barkhordarian, H.; Wang, M.; Schulz, P.; Sierks, M. J. Mol. Biol. 2007, 368(4), 1132–1144. (28) Zhou, C.; Emadi, S.; Sierks, M. R.; Messer, A. Mol. Ther. 2004, 10(6), 1023–31. (29) Barkhordarian, H.; Emadi, S.; Schulz, P.; Sierks, M. Protein Eng., Des. Sel. 2006, 19(11), 497–502. (30) Browning-Kelley, M.; Wadu-Mesthrige, K.; Hari, V.; Liu, G. Langmuir 1997, 13(2), 343–350. (31) Kaur, J.; Singh, K.; Schmid, A.; Varshney, G.; Suri, C.; Raje, M. Biosens. Bioelectron. 2004, 20(2), 284–293. (32) Willemsen, O.; Snel, M.; van der Werf, K.; de Grooth, B.; Greve, J.; Hinterdorfer, P.; Gruber, H.; Schindler, H.; van Kooyk, Y.; Figdor, C. Biophys. J. 1998, 75(5), 2220–2228. (33) Volles, M.; Lansbury, P. Nucleic Acids Res. 2005, 33(11), 3667–3677. (34) Volles, M.; Lansbury, P. J. Mol. Biol. 2007, 366(5), 1510–1522.
Langmuir, Vol. 25, No. 2, 2009 913 Preparation of Monomeric, Oligomeric and Fibrillar rS. The different morphologies of RS were prepared in a similar fashion by first diluting the RS to a stock concentration of 1 mg/mL (∼70 µM) in Tris-HCl buffer (25 mM Tris, 150 mM NaCl, pH 7.4). This stock RS was further diluted to a final concentration of 30 µM in Tris-HCl buffer. The RS sample obtained immediately after the sample preparation was predominantly monomeric. To obtain predominantly oligomeric RS, the samples were incubated at 37 °C with shaking at 250 rpm for 3-4 days. A predominantly fibrillar RS sample was formed after 2 weeks of incubation under the same conditions. All RS morphologies were confirmed by AFM prior to use. The different RS samples were not purified to separate different morphologies. Rather they were collected at different incubation time points where a single morphology is the most dominant species (monomeric, oligomeric, or fibrillar) as determined by AFM. This ensured that a variety of monomeric, oligomeric, and fibrillar forms were respectively present in each sample. Production and Purification of scFv. The anti-oligomeric RS scFv D5,27 pan-specific RS scFv D10,28 anti-fibillar RS scFv 6E,29 and antiphosphorylase B (RPLB) scFv were produced and purified as previously described.29,35 The final concentrations of the D5, D10, 6E and RPLB scFvs were 0.6 mg/mL (∼20 µM), 0.2 mg/mL (∼7 µM), 0.6 mg/mL (∼20 µM), and 1 mg/mL (∼34 µM), respectively. Protein-Treated Mica Surfaces. A 10 µL aliquot of RS was added to freshly cleaved mica (Spruce Pine, NC) and incubated at room temperature for 10 min. The mica was thoroughly rinsed with copious amounts of deionized (DI) water (18.1 MΩ, Millipore, MA) and dried under a gentle stream of N2 gas. For scFv binding, an additional 10 µL of each scFv was added separately to the RStreated mica and incubated for another 10 min before rinsing with DI water to remove unbound particles. The rinsing was done by applying 1 mL of DI water directly over the mica surface while holding the mica with a pair of tweezers and letting the water drain into a collecting beaker. The mica was rinsed three times before drying with N2. All treated surfaces were stored under vacuum until imaged by AFM. AFM Imaging. Topographic images were carried out in air at room temperature using a tapping mode AFM (TMAFM) with a Nanoscope IIIa controller (Veeco, Santa Barbara, CA). Images were acquired using oxide sharpened Si3N4 AFM tips (k ) 40 N/m, fo ∼ 300 kHz) (model: OTESPA, Veeco, Santa Barbara, CA) at scan rates of 2 Hz and at a scan resolution of 512 samples per line. Images were subjected to second-order polynomial flattening as needed to reduce the effects of image bowing and tilt. Height Distribution. AFM images were obtained using a TMAFM, equipped with a Nanoscope IIIa controller (Veeco, Santa Barbara, CA) and were analyzed automatically with the Scanning Probe Imaging Processor (SPIP) software (Image Metrology, ver. 4.5, Denmark) to generate height distributions for each sample. The height of each image was reset to absolute zero automatically by the SPIP software, and the z-dimension was used to estimate the size of the features on the image. Surface features having heights less than 0.5 nm were disregarded to account for the contribution of background pixels in the image. The size distribution was obtained by measuring all pixels on the AFM image after the z-height was reset. Statistical Analysis of Mean Particle Height. Statistical analysis of the mean particle height for each sample was performed using the Grain Analysis Module on the SPIP software. To determine the mean particle heights, 150-200 particles from the monomeric sample series, 60-100 particles from the oligomeric samples series, and 3-8 individual fibrils from the fibrillar sample series were randomly chosen from three independent AFM images of each sample. The detection limit was set at 0.5, 1, and 2 nm for the monomeric, oligomeric, and fibrillar sample series, respectively, to eliminate smaller particles and background pixels. Each particle is defined by a contour and segment as detected by the SPIP software based on the height values. The contour creates a boundary or outline around (35) Emadi, S.; Liu, R.; Yuan, B.; Schulz, P.; McAllister, C.; Lyubchenko, Y.; Messer, A.; Sierks, M. R. Biochemistry 2004, 43(10), 2871–8.
914 Langmuir, Vol. 25, No. 2, 2009
Wang et al.
Table 1. Particle Height Distribution of D5, D10, 6E and rPLB scFvs, Monomeric, Oligomeric, and Fibrillar rS, and rS-scFv Complexesa % Height Distribution sample
0.5-1 nm 1-2 nm 2-4 nm 4-6 nm >6 nm
D5 D10 6E RPLB
63.2 54.7 0.2 9.8
36.8 40.8 88.3 83.4
0.0 1.6 10.4 5.0
0.0 1.1 0.8 1.6
0.0 1.8 0.3 0.3
Monomeric RS RS RS+D5 RS+D10 RS+6E RS+RPLB
90.8 96.6 76.5 94.1 83.5
7.9 1.8 9.8 1.3 15.2
0.8 1.0 13.5 4.5 1.2
0.4 0.0 0.1 0.1 0.0
0.2 0.6 0.1 0.0 0.0
Oligomeric RS RS RS+D5 RS+D10 RS+6E RS+RPLB
18.3 24.0 11.7 5.7 2.7
77.3 53.5 75.4 87.0 96.1
3.4 22.4 12.3 7.3 1.0
0.6 0.0 0.6 0.0 0.1
0.4 0.0 0.0 0.0 0.0
0.6 0.8 0.2 0.3 1.2
4.0 12.1 1.0 3.4 7.1
87.2 82.1 35.0 58.5 85.1
2.9 2.9 59.4 30.4 2.9
5.4 2.1 4.4 7.5 3.7
Fibrillar RS RS RS+D5 RS+D10 RS+6E RS+RPLB
a Samples were incubated on mica for 10 min, rinsed with DI water, and dried with N2 gas. For binding analysis, scFvs were added to RS-treated mica and incubated for an additional 10 min, rinsed with DI water, and dried with N2 gas.
Table 2. Mean Particle Heights of D5, D10 and 6E scFv’sa scFv
n
mean particle height ( SEM (nm)
D5 D10 6E RPLB
55 28 37 43
0.86 ( 0.01 1.00 ( 0.05 1.20 ( 0.11 1.45 ( 0.03
a Data represent the average of three independent samples ( SEM, n ) number of randomly selected particles per sample. Samples were incubated on freshly cleaved mica for 10 min at room temperature, rinsed with DI water, and dried with N2 gas.
Results
Figure 1. Mean particle heights of (a) monomeric RS series, (b) oligomeric RS series, and (c) fibrillar RS series. RS (black column), RS+D5 (striped column), RS+D10 (dotted column), RS+6E (gray column), RS+RPLB (white). An aliquot of 10 µL of RS was incubated on mica for 10 min, rinsed with DI water, and dried under N2 gas before imaging. For the RS-scFv complex, an additional 10 µL of scFv was added to the RS-treated mica, incubated for 10 min, rinsed with DI water, and dried before imaging. Statistical analysis was determined using a two-tail t-test to determine significance by comparing the mean particle heights from the untreated RS sample with that of the scFvtreated samples: (*) p < 0.05; (**) p < 0.01; (***) p < 0.001. Error bars represents the standard error means from three independent samples. n ) 150-200 (monomeric RS samples); n ) 60-100 (oligomeric RS samples); n ) 3-8 (fibrillar RS samples).
Antigen Binding Characterized by Height Distribution. Height Distribution. In order to determine antibody specificity to different protein morphologies, we anticipated that the height distribution of the different RS samples would shift to larger sizes when an scFv was added that specifically bound the target RS morphology. We utilized three different RS samples containing predominantly monomeric, oligomeric, or fibrillar morphologies, respectively, and studied their interactions with three different scFv’s (D5, D10, and 6E) previously isolated in our laboratory. A non RS specific scFv RPLB was used as a negative control. Height images for each RS and scFv sample were carried out using TMAFM, and the height distribution data were generated using the SPIP software (Table 1). The scFv only samples were mostly small aggregates (
