Protein and Peptide Letters, 2005, 12, 79-83
79
Atomic Force Microscopy Study of Human Amylin (20-29) Fibrils Victoria L. Sedman, Stephanie Allen, Weng C. Chan, Martyn C. Davies, Clive J. Roberts, Saul J.B. Tendler* and Philip M. Williams Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The University of Nottingham, Nottingham. NG7 2RD UK. Abstract: Here we present atomic force microscopy images of the fibrils formed by human amylin(20-29). This peptide is a fragment of the polypeptide amylin, the major proteinaceous component of amyloid deposits found in cases of type-II diabetes mellitus. Our results demonstrate that the amylin(20-29) peptide fragment forms amyloidlike fibrils that display polymorphic structures. Twisting along the axis of fibrils was often observed in fibrils aged for 6 hours but disappeared in mature fibrils aged for longer time periods.
Keywords: protein aggregation, designability, protein folding, protein design, amyloid. INTRODUCTION
MATERIALS AND METHOD
A common pathological feature of human type-II diabetes mellitus is the deposition of amyloid fibrils in pancreatic islet cells [1]; this material is thought to contribute to the progressive dysfunction of the islet betacells. The major proteinaceous component of these fibrillar deposits is amylin, a 37 amino acid polypeptide [2], normally co-secreted with insulin by the islet beta-cells [3]. The abnormal aggregation and deposition of amylin suggests that type-II diabetes mellitus is one of a growing number of degenerative mammalian disorders including Alzheimer’s disease and transmissible spongiform encephalopathies (Scrapie, BSE, CJD). The fibrillar deposits associated with such diseases are comprised of structures termed amyloid. Amyloid fibrils are typically linear unbranched structures with diameters in the range of 4 to 20 nm and of varying lengths [4, 5]. Amyloid fibrils have a generic structural hierarchy irrespective of the comprising protein, in which the proteins aggregate into protofilaments that assemble together forming protofibrils, which then form higher order amyloid fibrils [6]. X-ray diffraction patterns of amyloid demonstrate a common core β-sheet structure in which the individual proteins are stacked perpendicular to the fibril axis [6-8].
The peptide fragment of human amylin corresponding to residues 20 to 29 (sequence H2N– S-N-N-F-G-A-I-L-S-S COOH) was synthesized using solid phase peptide synthesis [16, 17] and purified using high performance liquid chromatography. The chemical identity of the peptide fragment was established using liquid chromatography-mass spectrometry (results not shown).
Secondary structure predictions for the full-length amylin polypeptide highlight three β-strands at residues 8 to 20 [9], 24 to 29 [10, 11] and 30 to 37 [12]. Extensive study of in vivo and synthetic amylin has revealed a stretch of 10 amino acids that are crucial to fibril formation, corresponding to residues 20 to 29 in the human sequence [13, 14]. Here we present atomic force microscope (AFM) images of fibrils formed by a peptide fragment corresponding to this segment of the human amylin over a time course of up to 48 hours. Whilst previously published electron microscopy data [15] have used fibrils stored for greater than 24 hours. Here we have focused in detail on the morphologies of the amylin(2029) fibrils stored for up to 18 hours. *Address correspondence to this author at the Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The University of Nottingham, Nottingham, NG7 2RD,. UK; Tel: +44(0)115-9515063; Fax: +44(0)1159515110; E-mail:
[email protected] 0929-8665/05 $50.00+.00
The lyophilized peptide was solubilised in ultrapure water (pH 6, resistivity 18.2 MΩ cm) at a concentration of 1 mg/ml. Samples were left to form fibrils at room temperature for 6 to 48 hours. In some instances, to prevent molecular overcrowding of the AFM scan area, stock solutions were diluted to concentrations of 100 µg/ml in ultrapure water immediately prior to imaging. Aliquots of 10 µl were dropped onto freshly cleaved mica (1 cm2), dried in a N 2 gas stream and imaged immediately. Ultrapure water stored under identical conditions and for the same time period was used as a control preparation; here imaging showed no fibrillar structures. Images presented were generated using a Nanoscope IIIa AFM (Digital Instruments, Veeco Metrology Group, Santa Barbara CA). All imaging was performed in air in tapping mode using silicon tapping probes (supplied by Veeco Metrology Group) mounted on cantilevers with spring constants of 34.4-37.2 N/m and resonant frequency of 280360 kHz. Topography and phase information are shown as contrast images, with lighter colours indicating higher topographic features or a more positive phase shift. All images obtained have a resolution of 512 x 512 pixels. Dimensions given are the mean ± the standard deviation of a sample (n >10) of randomly selected fibrils within each scan area. RESULTS AND DISCUSSION Fibril Morphology The representative AFM images presented in this paper demonstrate the polymorphic nature of the fibrils formed by the peptide fragment amylin(20-29). The images in figures 1 to 4 are of fibrils stored in solution for 18 hours, dried onto © 2005 Bentham Science Publishers Ltd.
80
Protein and Peptide Letters, 2005, Vol. 12, No. 1
Tendler et al.
result width dimensions may be an overestimate of the true value, however height measurements are accurate. Nonetheless, these measurements suggest that the protofibrils formed by amylin(20-29) are flat ribbon-like structures. This provides further evidence that this fragment of amylin forms fibrillar structures similar in morphology to those formed by full-length amylin [15].
Figure 1. AFM image of individual fibrils of amylin(20-29) stored for 18 hours. Figure 1a is a 1.5 µm x 1.5 µm topography image, the boxed area was zoomed in, using the AFM to give the 500 nm x 500 nm topography image in figure 1b. Arrows indicate two different widths (16.7 (±3.1) nm and 23.9 (±2.6) nm) of the protofibril. Solution diluted to 100 µg/ml prior to imaging.
mica and then imaged in air. Dimensions can be seen to vary from fibril to fibril. However individually, widths appear uniform along each fibril axis. The majority of fibrillar structures were straight and rod-like with heights of 4-5 nm, and widths that fell into two main population distributions with means of 27 nm and 34 nm. In this paper, fibrillar structures with these height and width dimensions are referred to as protofibrils. When taking dimension measurements one must take into account the effects of AFM tip convolution, in which features appear broader. As a
A common feature often observed was the thinning at the end of a fibril. An example of this is presented in figure 1b, the arrows highlight the thicker and the thinner portions of the protofibril; a noticeable height difference of 1.0 nm can be observed. One explanation for this could be that the thicker and taller protofibril with a width of 23.9 (±2.6) nm, is undergoing elongation with the thinner (width 16.7 (±3.1) nm) lower-order protofibril visible. Although this phenomenon could be due to two protofibrils aligning on top of one another, it seems unlikely as this was observed in several protofibrils during imaging of other scan areas. Furthermore, a similar phenomenon has been noted previously in growing fibrils of full-length amylin [5]. Several distinctive morphological features that were encountered during imaging are presented in figure 2. Protofibrils were observed fanning out of a dense central aggregate (figure 2a). Fibrils were sometimes found to branch into two new protofibrils (figure 2b), with the new branches of equivalent height and width as the stem protofibril. In some instances, the branches then remerged to form a single protofibril (figure 2c and d). The phase images
Figure 2. Representative AFM images of amylin(20-29) fibril morphologies at 18 hours. Figure 2a shows a 500 nm x 500 nm topography (left) and phase (right) image showing fibrils fanning out from a central point. The 850 nm x 850 nm topography image shown in figure 2b is of a fibril splitting or branching. The topography image (1 µm x 1 µm) shown in figure 2c is another example of fibril branching. The image in figure 2d is a zoom in (500 nm x 500 nm) of the structure shown in the box in figure 2b indicating a fibril splitting and remerging. Solution diluted to 100 µg/ml prior to imaging.
Atomic Force Microscopy Study of Human Amylin (20-29) Fibrils
Protein and Peptide Letters, 2005, Vol. 12, No. 1 81
structures in the full-length protofibrils may be retained within the shorter peptide fragment of amylin (20-29). Although it is not known whether these higher-order assemblies are due to an artifact of the sample preparation method, this phenomenon has been previously observed both by electron microscopy and atomic force microscopy in which sample preparation in the two techniques is different [5,9,19]. Therefore further study of the aggregation properties of amylin(20-29) would prove beneficial to clarify this point, particularly investigation of aggregation patterns in solutions. Figure 3. AFM image of amylin(20-29) fibrils aged 18 hours assembled into twisted bundles. 2.5 µm x 2.5 µm topography image. The arrows indicate two bundles with different sizes and morphologies. Solution diluted to 100 µg/ml prior to imaging.
in figures 2a, 4a and b, provide information on the surface properties of the fibrillar structure and complementary information to the topography data on the morphology of the amylin (20-29) fibrils. The phase data provides additional information on the chemical nature of the structures observed, primarily that the fibrillar structures are soft and sticky as would be expected from a proteinaceous feature [18]. Individual protofibrils were also observed organised into higher-order structures. The image shown in figure 3 is a representative of protofibrils assembled into twisted bundles. The twisting of the bundles appeared to have no regular periodicity. Some bundles appeared more regular, however there was variation in the number of protofibrils incorporated within each bundle. Intriguingly, protofibrils were also found to laterally align forming sheet-like assemblies as demonstrated in the image in figure 4. In such cases, the protofibrils are more linear, producing a uniform array. Also of note, protofibrils within the twisted bundles could extend out from one bundle to form that of another, generating a network, which was not seen with protofibrils that assembled laterally. These observations are complementary to formations previously observed for protofibrils of full-length amylin [5,19], in which it was noted that two similar populations exist; protofibrils that coil with distinct axial periodicity and a second group in which the protofibrils laterally align [19]. These data are suggestive that the interactions governing the formation of higher-order
FIBRIL VARIATION WITH TIME A study was then undertaken to investigate the time scale of fibrillogenesis, with the AFM images taken in air from aliquot samples removed after successive 2 hour time periods. Fibrillar structures were observed after incubation of the solution for 4 hours: prior to this, no fibrous material was seen. Figure 5a is a typical scan area of the fibrillar structures present in preparations incubated for 6 hours. The majority of protofibrils observed display a twisting along their axis, giving the structures a globular appearance. Figure 5c shows a height cross-sectional profile along the axis of the protofibril highlighted within the box in figure 5b. The twisting has a periodicity of approximately 75 nm; however, this varied between different protofibrils and in some instances was not as regular. This periodicity was not observed in solutions of protofibrils incubated for longer periods (18+ hours) (figures 1 to 4). By comparison, protofibrils of full-length amylin aged for long periods of time have been observed by AFM and electron microscopy to demonstrate regular periodicity with twists of 25 and 50 nm [5, 19]. These data suggest that over the time course studied the nascent protofibrils grow and assemble into mature protofibrils, which do not display any periodicity. This observable change in topography over the time course studied may be the result of loss of resolvable underlying structure. Furthermore, it is possible that the structures termed protofibrils in the preparations aged for 6 hours are more accurately nascent protofibril intermediates of the protofibril and a lower order structure. This suggestion provides further support for the highly ordered structural hierarchy in which the amyloid fibril is formed from the aggregation of protofibrils, which are assemblies of protofilaments comprising of several subprotofilaments [6].
Figure 4. AFM image of the lateral assembly of amylin(20-29) fibrils aged 18 hours. Figure 4a , shows a 2.5 µm x 2.5 µm topographic (left) and phase (right) image. Figure 4b, shows a 500 nm x 500 nm topographic (left) and phase (right) image of the AFM zoomed in area highlighted with the box in figure 4a . Solution diluted to 100 µg/ml prior to imaging.
82
Protein and Peptide Letters, 2005, Vol. 12, No. 1
Tendler et al.
Figure 5. AFM images of amylin(20-29) fibrils aged 6 hours. Figure 5a shows a 1.5µm x 1.5µm topography image, a zoom in 663 nm x 663 nm image is shown in figure 5b. Figure 5c is a line plot of the height variations along the axis of the fibril in the box of figure 5b, showing an example of the periodicity observed in younger fibrils. Concentration 1 mg/ml.
These observations provide evidence that amylin(20-29) forms protofibrils that display this hierarchy and thus that the ability to form the highly organized and structured amyloid fibril is intrinsically retained within the fragment of amylin (20-29). Analysis of protofibril sizes revealed that those incubated for up to 18 hours fell into two main populations, with widths of approximately 27 nm and 34 nm (n=46). Whereas at 48 hours there was some evidence of higher-order fibres, with an additional population of individual protofibrils exhibiting widths of approximately 50 nm. Previous electron microscopy data [15] of amylin(20-29) fibrils show them to be straight and rod like structures; however, currently there is no published information on fibrils from solutions stored for less than 24 hours. Here we have investigated the morphology of fibrils from solutions stored for shorter time. Our findings suggest that protofibrils form after a lag period of 4 hours in solution, and that these early fibrils have a cable-like appearance with varied periodicity, indicating that the morphology of amylin(20-29) protofibrils may vary during growth. Younger protofibrils display a periodicity, which disappears when they mature into older fibrils that are straight rod-like structures.
published electron microscopy data [19]. Furthermore, the data suggest that the morphology of individual protofibrils varies with time, with young protofibrils (aged up to 6 hours) displaying a twisted periodicity that was not observable in older protofibrils (18 and 48 hours), which were predominantly straight and rod like. However, further investigation into these early stages of fibrillization is required. The atomic force microscope in particular provides an opportunity to study dynamically fibrillization events in liquid [4] and hence provide insight into mechanisms of the growth and maturation of fibrils. ACKNOWLEDGEMENTS VLS thanks the UK Biotechnology and Biological Sciences Research Council for funding of her studentship. SA thanks Pfizer Global Research and Development for the funding of her lectureship. REFERENCES [1] [2] [3]
CONCLUSION The fibrils of amylin(20-29) displayed polymorphic features and were of varying lengths. Protofibrils were observed on the mica surface individually, grouped in lateral arrays or into irregular twisted bundles that often formed networks, or were laterally aligned into monolayers. Widths of the fibrils observed fell into three main populations, with means of 27 nm, 34 nm and in preparations incubated for 48 hours an additional group of 50 nm fibrils. These fibril dimensions are in general agreement with previously
[4] [5] [6] [7] [8]
Kahn, S.E., Andrikopoulos, S. and Verchere, C.B. (1999) Diabetes, 48(2):640-5. Cooper, G.J.S., Willis, A.C., Clark, A.C., Turner, R.C., Sim, R.B. and Reid, K.B.M. (1987) Proceedings of the National Academy of Sciences, USA, 84:8628-32. Kahn, S.E., D’Alessio, D.A., Schwartz, M.W., Fujimoto, W.Y., Ensinck, J.W., Taborsky, Jr. G.J. and Porte, Jr. D. (1990) Diabetes, 39:634-8. Blackley, H.K.L., Davies, M.C., Sanders, G.H.W., Roberts, C.J., Tendler, S.J.B., Wilkinson, M.J. and Williams, P.M. (2000) Journal of Molecular Biology, 298:833-840. Goldsbury, C., Kistler, J., Aebi, U., Arvinte, T. and Cooper, G.J.S. (1999) Journal of Molecular Biology, 285:33-39. Sunde, M. and Blake, C.C.F. (1997). Advances in Protein Chemistry, 50:123-159. Blake, C.C.F. and Serpell, L.C. (1996) Structure, 4:989-998. Sunde, M., Serpell, L.C., Bartlam, M., Fraser, P.E., Pepys, M.B. and Blake, C.C.F. (1997) Journal of Molecular Biology, 273:729739.
Atomic Force Microscopy Study of Human Amylin (20-29) Fibrils [9] [10] [11] [12] [13] [14]
Jaikaran, A.S., Higham, C.E., Serpell, L.C., Zurdo, J., Gross, M., Clark, A. and Fraser, P.E. (2001) Journal of Molecular Biology, 308: 515-25. Glenner, G.G., Eanes, E.D. and Wiley, C.A. (1988) Biochemical and Biophysical Research Communication, 155(2): 608-14. Betsholtz, C., Christenson, L., Engstrom, U., Rorsman, F., Jordan, K., O’Brien, T.D., Murtaugh, M., Johnson, K.H. and Westermark, P. (1990) Diabetes, 39(1):118-22. Nillson, M.R. and Raleigh, D.P. (1999) Journal of Molecular Biology, 294:1375-1385. Westermark, P., Engstrom, U., Johnson, K.H., Westermark, G.T. and Betsholtz, C. (1990) Proceedings of the National Academy of Sciences USA, 87:5036-40. Azriel, R. and Gazit, E. (2001) Journal of Biological Chemistry, 276(36):34156-61.
Protein and Peptide Letters, 2005, Vol. 12, No. 1 83 [15] [16] [17] [18] [19]
Goldsbury, C., Goldie, K., Pellaud, J., Seeling, J., Frey, P., Muller, S.A., Kistler, J., Cooper, G.J.S. and Aebi, U. (2000) Journal of Structural Biology, 130:352-62. Nilsson, M.R., Nguyen, L.L. and Raleigh, D.P. (2001) Analytical Biochemistry, 288:76-82. Chan, W.C. and White, P.D. (2000) Fmoc Solid Phase Peptide Synthesis: A Practical Approach, Ed. Chan, W.C. and White, P.D. Oxford University Press. García, R. and Pérez, R. (2002) Surface Science Reports, 47: 197301 Goldsbury, C., Cooper, G.J.S., Müller, S.A., Saafi, E.L., Gruijters, W.T.M., Misur, M.T., Engel, A., Aebi, U. and Kistler, J. (1997) Journal of Structural Biology, 119:17-27.
Received on January 6, 2004, accepted on June 23, 2004.
