Supplementary Information for Nairn, Lyons et al.
A synthetic resilin is unstructured
SAXS details
Experimental
SAXS data were collected on the SAXS instrument located at sector 15, ChemMatCARS, at the Advanced Photon Source, Chicago, IL. This beamline employs a high-brilliance undulator source, which delivered more than 2×1012 photons per second to the sample at a wavelength of 1.1 Å (11.3 keV) (1). The camera length was calibrated with silver behenate at a nominal sample to detector distance of 1900 mm, giving an accessible q-range of 0.0077 Å-1 to 0.30 Å-1 (q = 4πsinθ/λ, where λ is the x-ray wavelength, and 2θ is the scattering angle).
The samples were mounted in an air gap of ~2 cm. Two thin windows (silicon nitride and mica) separated the sample air space from the beam transport vacuum. Background patterns taken with no sample in place indicated that these windows contributed negligible amounts of parasitic scatter. During exposure the solutions were held in a capillary with a nominal 1.5 mm bore. A motorized syringe was connected to the capillary, via a Touhy Borst to Luer lock adapter, and was used to draw and expel the solutions. This system allowed different solutions to be loaded without moving the capillary, ensuring that the path length through the solution container remained identical – essential for good quality background subtractions. The motorized syringe was used to move the protein solution through the x-ray beam, in an attempt to avoid beam damage to the sample. A range of exposure times was used, from 1 to 20 seconds, with 4 different protein concentrations (0.1, 1, 5 and 10 mg/ml). A 2 second exposure was sufficient to collect a scattering pattern for a concentration of 5 mg/ml, however all concentrations were measured with a range of exposure times to check detector linearity over the entire scattered intensity range. All patterns were corrected for detector geometry, response and dark-current. Images were then azimuthally integrated to give one dimensional profiles of intensity versus q. Profiles for different samples were normalized to constant incident count rate and transmission. The transmission measurements were made after the SAXS scans were collected to reduce unnecessary radiation damage prior to the collection of scattering data. Compressibility scatter from pure water (q < 0.3Å-1) in the same capillary used for the specimen provided an absolute intensity calibration.
Results
Figure S1 shows a typical scattering profile for AN16 on a log-log scale. The number of experimental data points shown has been reduced for clarity. Also on the figure are two fitted curves as detailed below. The scattering profile can be divided into four regions, each exhibiting a specific scattering feature attributable to a particular sample characteristic.
Figure S1: Log-Log representation of the scattering profile and fitting for AN16. Exposure time 5 seconds, sample concentration 5 mg/ml. Circles: experimental data; triangles: fit (over region 0.019 Å-1 < q < 0.15 Å-1) using Equation 3 (Rg = 48Å and α = 2.05); grey line:fit (over region 0.019 Å-1 < q < 0.15 Å-1) using Equation 4 (Rg = 48Å).
In the case of a dilute solution of monodisperse particles (such as protein molecules), where the x-ray scatter from one particle is unlikely to interfere with that of its neighbor, the SAXS intensity profile, at low q, has a constant value related to the scattering particle volume. As q increases towards the value 1/r, where r is the average particle radius, there will be an exponential-like decay in scattering intensity with q. On a log-log scale this will be observed as a knee in the scattering profile. The knee is typically fitted using the Guinier approximation (2):
, Eq. 1
where Rg is the radius of gyration of the particle. It should be noted that Equation 1 alone is not valid for qRg1. In Figure S1there is a knee in the region 0.018 Å-1q< 0.045 Å-1; however, the scattered intensity is not constant for q0.018 Å-1. This indicates the presence of particles significantly larger than the expected protein molecule size – most likely resulting from protein aggregation. Fortunately, in this case the low-q modification of the scattering profile by aggregation does not mask the scattering from the primary protein.
For ordered, globular proteins in solution, the scattering profile at q beyond the Guinier knee tends to oscillate. This was not observed for the scattering profile of AN16, which displayed a monotonically decreasing slope with a power law of the form:
, Eq. 2
where -α is the power-law exponent. Both the high and low q regions may be modeled using the Unified Fit of Beaucage (3), thus avoiding the restriction of qRg < 1 encountered when fitting with Equation 1 alone . Here a good fit (shown as a solid grey line in Figure S1) is achieved using the equation
, Eq. 3
where G and B are the prefactors for the Guinier and power-law regions respectively. The parameter k is an empirical constant which is known for α > 3 or α ≈ 2. With this model Rg was found to be 48 ± 3 Å, the power-law gradient α = 2.05 ± 0.05, G = 0.13 cm-1 and B = 9.3×10-5 cm-1.
The α = 2 power-law scattering is consistent with that seen for a Gaussian coil – a random chain structure that neither attracts nor avoids itself. Debye (4) formulated an equation, to describe the scattering from a Gaussian coil:
I(q) = 2I(0) (x-1 + e-x)/x2, Eq. 4
where x = q2Rg2. This equation describes both the Guinier region and the power-law scattering and was also used to compare with the Unified Fit of Beaucage. The best fit of this equation to the data is shown in Figure S1 (white triangles connected by a thin black line). Note that it is in excellent agreement for the region 0.02 < q < 0.45 Å-1. From this fit the Rgwas found to be 51 ± 2 Å.
As discussed by Beaucage (5), at high qthe α = 2 power-law scattering of a random coil will be limited by the persistence length of the chain (i.e., the distance over which correlations in the direction of the tangent to the chain are lost), and hence the profile will depart from a linear form at approximately the inverse of the persistence length. For a polymer coil the persistence length is usually considered a rod, leading to α = 1 power-law scattering for q somewhat greater than the inverse of persistence length. While the scattering profile for AN16 departs from the = 2 power law at q≈ 0.2 Å-1, the data in this region are noisy and only allow the statement that persistence length is of the order of several angstroms – appropriate for a protein chain.
Plotting the SAXS data on a Kratky plot (figure S2) allows it to be compared with literature data for urea denatured proteins (for example lysozyme (6)), acid denatured proteins (such as cytochrome C (7)) and heparin (used as a model random coil(6)). The AN16 data combine features of both the model random coil and the denatured protein curves. Like the model random coil(6), the curve does not have a distinct peak but continues to increase in intensity at higher q. A similar shape is also observed for acid denatured cytochrome C (7), and temperature-and-urea denatured Streptomyces subtilisin inhibitor (8)in contrast to the distinct peak observed for urea denatured lysozyme(6) or guanidinium denatured lysozyme (9). However, the low q region of the AN16 data has a similar strongly curved shape to the urea denatured lysozyme. This comparison implies that AN16 is closer to an overall "random coil" conformation than many denatured proteins.
Figure S2: Kratky plot of experimental data from figure S1
Circular Dichroism details
Experimental
AN16 in PBS was further purified by affinity chromatography on a Ni-column, and then dialysed into a 10 mM sodium phosphate buffer. CD spectra were measured on a solution containing 120 g/ml of protein. A Jasco J-810 spectrometer was used, with a 1mm path-length quartz cell. 10 acquisitions per spectrum were run from 185-260 nm at 50 nm/min. The buffer background was subtracted.
The fitting program interface CDPro (Sreerama, Colorado State University) was used to extract secondary structure information from the CD spectrum. The reference spectra sets used were SDP42, SDP48 and SP37A. SDP42 and SDP 48 contain some denatured proteins, along with soluble model proteins. SP37A includes model proteins for the polyproline II (PPII) structure. CDPro uses three different fitting algorithms (ContinLL (10, 11), Selcon3 (12) and CDSSTR (13)). The secondary structure fractions determined by the three algorithms were very similar.
Figure S3: Far-UV CD spectrum of AN16 in 10mM phosphate buffer. The fitted curve was obtained using ContinLL, with reference dataset SP37A.
Table S1:
Table of circular dichroism secondary structure fraction results for various fits to the experimental data in figure S3.
Program / CDSSTR / ContinLL / Selcon3 / CDSSTR / ContinLL / Selcon3 / CDSSTR / ContinLL / Selcon3 / Average / Standard DeviationReference set / sdp48 / sdp48 / sdp48 / sdp42 / sdp42 / sdp42 / sp37a / sp37a / sp37a
Regular alpha helices / -0.005 / 0.001 / 0.004 / -0.005 / 0.002 / -0.006 / -0.002 / 0.004
Distorted alpha helices / 0.029 / 0.024 / 0.016 / 0.018 / 0.025 / 0.024 / 0.023 / 0.005
Regular beta strands / 0.105 / 0.084 / 0.071 / 0.099 / 0.083 / 0.097 / 0.09 / 0.01
Distorted beta strands / 0.057 / 0.058 / 0.052 / 0.065 / 0.071 / 0.067 / 0.062 / 0.007
Turns / 0.106 / 0.106 / 0.105 / 0.119 / 0.138 / 0.129 / 0.12 / 0.01
Unordered / 0.695 / 0.727 / 0.74 / 0.688 / 0.68 / 0.667 / 0.70 / 0.03
Total alpha helices / 0.041 / 0.052 / 0.058 / 0.050 / 0.009
Total beta sheet / 0.229 / 0.162 / 0.165 / 0.19 / 0.04
Turn / 0.136 / 0.118 / 0.117 / 0.12 / 0.01
PPII / 0.108 / 0.109 / 0.093 / 0.103 / 0.009
Unordered / 0.49 / 0.559 / 0.547 / 0.53 / 0.04
Additional Raman Spectra
Figure S4a: Amide I bands of the Raman spectra for crosslinked gels at 230 mg/ml and 423 mg/ml protein. Spectra have been normalized and baseline corrected.
Figure S4b: Uncrosslinked samples concentrated via exposure to PEO or by drying over calcium chloride
Figure S4b shows the Amide I bands (baseline corrected and normalized in intensity) of the Raman spectra for various uncrosslinked samples of AN16:
(i) AN16 solutions (150 uL aliquots of around 20% protein in PBS) that were slowly concentrated further by placing varying amounts (5 to 80 mg) of polyethylene oxide (MW 600000, Aldrich chemical company) in contact with the solutions (in a similar way to the osmotic stress technique described in (14)); and
(ii) Uncrosslinked solid AN16 formed by allowing a solution to dry over calcium chloride.
Figure S4c: Amide I Raman band of dragonfly tendon compared with concentrated (423 mg/ml) crosslinked AN16 gel. Both bands are baseline corrected and normalized in intensity.
Figure S4c compares the Amide I bands for an example of a natural resilin (the tendon from a dragonfly), and a sample of AN16 crosslinked gel with a similar protein concentration. The natural resilin contains less aromatic residues than AN16, as evidenced by the much smaller peak at 1600-1620 cm-1. The Amide I band of the natural resilin is narrower than that of AN16, and also exhibits two distinct peaks (at 1657 cm-1 and approximately 1668 cm-1). This suggests that natural resilin is more ordered than AN16. This additional order is not indicative of a -spiral structure, since the Amide I band for the model -spiral peptide cyclo(VPGVG)3 occurs at 1676 cm-1, and for the corresponding polypeptide (VPGVG)n at 1673 cm-1(15).
Significant differences in Raman bands between model polypeptides and their natural counterparts have also been observed for (VPGVG)n and bovine elastin(16). In this previous case, the natural protein was found to be more disordered than its model counterpart.
References
1. Cookson, D., N. Kirby, R. Knott, M. Lee, and D. Schultz. 2006. Strategies for data collection and calibration with a pinhole-geometry SAXS instrument on a synchrotron beamline. J. Synchrotron Radiat. 13:440-444.
2. Guinier, A. and G. Fournet. 1955. Small-angle Scattering of X-rays. John Wiley & Sons, New York; Chapman & Hall, London.
3. Beaucage, G. 1995. Approximations l eading to a unified exponential power-law approach to small-angle scattering. J. Appl. Crystallogr. 28:717-728.
4. Debye, P. 1947. Molecular-weight determination by light scattering. J. Phys. Colloid Chem. 51:18-32.
5. Beaucage, G. 1996. Small-angle scattering from polymeric mass fractals of arbitrary m ass-fractal dimension. J. Appl. Crystallogr. 29:134-146.
6. Chen, L., K. O. Hodgson, and S. Doniach. 1996. A lysozyme folding intermediate revealed by solution x-ray scattering. J. Mol. Biol. 261:658-671.
7. Pollack, L., M. W. Tate, N. C. Darnton, J. B. Knight, S. M. Gruner, W. A. Eaton, and R. H. Austin. 1999. Compactness of the denatured state of a fast-folding protein measured by submillisecond small-angle x-ray scattering. Proc. Natl. Acad. Sci. U. S. A. 96:10115-10117.
8. Konno, T., M. Kataoka, Y. Kamatari, K. Kanaori, A. Nosaka, and K. Akasaka. 1995. Solution x-ray scattering analysis of cold- heat-, and urea-denatured states in a protein, Streptomyces Subtilisin Inhibitor. J. Mol. Biol. 251:95-103.
9. Segel, D. J., A. Bachmann, J. Hofrichter, K. O. Hodgson, S. Doniach, and T. Kiefhaber. 1999. Characterization of transient intermediates in lysozyme folding with time-resolved small-angle x-ray scattering. J. Mol. Biol. 288:489-499.
10. Provencher, S. W. and J. Glockner. 1981. Estimation of globular protein secondary structure from circular dichroism. Biochemistry 20:33-37.
11. van Stokkum, I. H. M., H. J. W. Spoelder, M. Bloemendal, R. van Grondelle, and F. C. A. Groen. 1990. Estimation of protein secondary structure and error analysis from circular dichroism spectra. Anal. Biochem. 191:110-118.
12. Sreerama N. and Woody R. W. 1993. A self-consistent method for the analysis of protein secondary structure from circular dichroism. Anal. Biochem. 209:32-44.
13. Johnson, W. C. 1999. Analyzing protein circular dichroism spectra for accurate secondary structures. Proteins: Struct., Funct., Genet. 35:307-312.
14. Sohn, S., H. H. Strey, and S. P. Gido. 2004. Phase behavior and hydration of silk fibroin. Biomacromolecules 5:751-757.
15. Thomas, G. J. Jr, B. Prescott, and D. W. Urry. 1987. Raman amide bands of type-II beta-turns in cyclo-(VPGVG)3 and poly-(VPGVG), and implications for protein secondary-structure analysis. Biopolymers 26:921-34.
16. Prescott, B., V. Renugopalakrishnan, and G. J. Thomas Jr. 1987. Raman spectrum and structure of elastin in relation to type-II beta-turns. Biopolymers 26:934-6.