International
Tables for
Crystallography
Volume H
Powder diffraction
Edited by C. J. Gilmore, J. A. Kaduk and H. Schenk

International Tables for Crystallography (2018). Vol. H, ch. 3.7, pp. 318-319

Section 3.7.7. Protein Data Bank (PDB)

J. A. Kaduka,b,c*

aDepartment of Chemistry, Illinois Institute of Technology, 3101 South Dearborn Street, Chicago, IL 60616, USA,bDepartment of Physics, North Central College, 131 South Loomis Street, Naperville, IL 60540, USA, and cPoly Crystallography Inc., 423 East Chicago Avenue, Naperville, IL 60540, USA
Correspondence e-mail: kaduk@polycrystallography.com

3.7.7. Protein Data Bank (PDB)

| top | pdf |

The Protein Data Bank is described in Chapter 24.1[link] of Inter­national Tables for Crystallography Volume F (Berman et al., 2011[link]). Current information is available on the web at https://www.wwpdb.org/ .

3.7.7.1. Powder diffraction by proteins

| top | pdf |

Although powder-diffraction techniques had been applied to proteins as long ago as 1936 (Wyckoff & Corey, 1936[link]; Corey & Wyckoff, 1936[link]), and proof-of-principle experiments had been carried out (Rotella et al., 1998[link], 2000[link]), real progress in protein powder crystallography began with the work of Von Dreele (Von Dreele, 1998[link], 1999[link], 2003[link]; Von Dreele et al., 2000[link]).

Progress in powder crystallography on macromolecules has been reviewed by Margiolaki & Wright (2008[link]) and is also discussed in Chapter 7.1[link] of this volume. Notable studies include the characterization of the binding of N-acetylglucosamine oligosaccharides to hen egg-white lysozyme (Von Dreele, 2007a[link]) and determination of the second SH3 domain of ponsin (Margiolaki et al., 2007[link]).

As with all powder diffraction, peak overlap ultimately limits the information available. Multi-pattern strategies to overcome the overlap problem have been investigated by Von Dreele (2007b[link]). Multiple-pattern resonant-diffraction experiments have enabled study of the binding of PtBr62− ions to lysozyme (Helliwell et al., 2010[link]). A bootstrap approach has been used to determine the structure of bacteriorhodopsin to 7 Å resolution (Dilanian et al., 2011[link]). Parametric resonant-scattering experiments have been used to determine the secondary structures of lysozyme derivatives (Basso et al., 2010[link]). Powder-diffraction experiments have also been used to gain insight into the general features of a nonstructural protein 3 (nsp3) macro domain (Papageorgiou et al., 2010[link]).

The structure of a five-residue peptide has been determined ab initio using laboratory powder data (Fujii et al., 2011[link]). We can expect further useful results at this interface between small-molecule and protein powder crystallography.

As is typical in other areas of science, powder diffraction has proven to be useful in more practical features of protein processing. It has been used to identify insulin (Norrman et al., 2006[link]) and GB1 (Frericks Schmidt et al., 2007[link]) polymorphs and lot-to-lot variations in lyophilized protein formulations (Hirakura et al., 2007[link]), and has been explored for use in structure-based generic assays (Allaire et al., 2009[link]).

3.7.7.2. Calculation of protein powder patterns (with Kenny Ståhl)

| top | pdf |

The Powder Diffraction File contains a few experimental powder patterns of proteins. These include silk fibroin protein (00-054-1394), tubulin (00-036-1547 and 00-036-1548), insulin (00-060-1360 through 00-060-1368), tomato bushy stunt virus (00-003-0001) and tobacco mosaic virus (00-003-0003 and 00-003-0004). Patterns have not yet been calculated from the structures in the Protein Data Bank because the calculated intensities generally fit poorly to those in experimental patterns.

Protein structures in the PDB do not generally contain H-atom positions, and the contributions from the disordered solvent in the solvent channels (which is the major source of the discrep­ancy) is not described (Hartmann et al., 2010[link]). The conventional Lorentz factor tends to infinity when approaching 2θ = 0°. Differences in data-collection temperatures and solvent content between powder and single-crystal specimens often mean that the lattice parameters differ. The relatively poor scattering from the protein and the large scattering from the mother liquor and sample holder result in significant background contributions to experimental powder patterns.

Optimization of the lattice parameters is generally straightforward and is important because most protein crystal structures are determined at low temperatures, while powder data are collected under ambient conditions. Protein crystals contain 30–80% disordered solvent. The solvent contribution to the diffraction pattern is most important for the low-angle powder data. In conventional protein crystallography several correction models have been developed (Moews & Kretsinger, 1975[link]; Phillips, 1980[link]; Jiang & Brünger, 1994[link]), but the flat bulk-solvent model is the simplest one which yields a realistic correction (Jiang & Brünger, 1994[link]; Hartmann et al., 2010[link]). This model includes two parameters: ksol, which defines the level of electron density in the solvent region, and Bsol, which defines the steepness of the border between the solvent and macromolecular regions. These parameters are typically refined in contemporary software and cluster around ksol = 0.35 e Å−3 and Bsol = 46 Å2 (Fokine & Urzhumtsev, 2002[link]).

The flat bulk-solvent correction can be applied using phenix.pdbtools (Adams et al., 2010[link]), which requires a PDB coordinate file and values of ksol and Bsol as input. Average values can be used, but refined values or values from the Electron Density Server (EDS; Kleywegt et al., 2004[link]) can improve the results. The bulk-solvent correction is highly anisotropic, and both parameters affect the anisotropy.

The ideal H-atom positions can be calculated using phenix.pdbtools. The solvent and hydrogen contributions to the pattern can be significant (Fig. 3.7.13[link]).

[Figure 3.7.13]

Figure 3.7.13 | top | pdf |

Overview of the trends from the different corrections. The effects are shown as the relative intensity difference (Inon-corrIcorr)/Inon-corr plotted as functions of the scattering angle 2θ (using Cu Kα1) and resolution d = λ/(2 sin θ). The curves are based on average corrections of lysozyme and insulin data. Inon-corr is the raw intensity from a calculated pattern which has only been Lorentz corrected. The geometric correction curve was calculated using η = 0.045 Å−1. From Hartmann et al. (2010[link]).

The Lorentz factor L describes the fraction of a reflection that is in the diffracting condition. For Bragg–Brentano and Debye–Scherrer geometries it is given by[L = {1 \over {\sin 2\theta }}{1 \over {\sin \theta }}. \eqno (3.7.3)]This equation assumes ideal crystals, resulting in infinitesimally small reciprocal-lattice points. The true size of the lattice points depends on the crystallite size and imperfections (strain). This smearing needs to be included in the Lorentz factor at low angles. A revised Lorentz factor for protein powder diffraction has been derived (Hartmann et al., 2010[link]),[L_{\rm rev} = {1 \over {\sin 2\theta }} {1 \over {\sin \theta }} {{\sin^2\theta} \over {(\sin^2\theta + \lambda^2\eta^2/12)}}, \eqno (3.7.4)]in which η reflects the distribution of scattering-vector amplitudes. For Guinier geometry these equations become more complex (Hartmann et al., 2010[link]). Fig. 3.7.14[link] shows that the Lorentz factor has a smaller effect than the solvent and H atoms, but that it is still significant. By applying these corrections it should be possible for the ICDD editorial staff to calculate useful powder patterns from PDB entries that could be included in the Powder Diffraction File.

[Figure 3.7.14]

Figure 3.7.14 | top | pdf |

Calculated and experimental powder patterns for (a) lysozyme, (b) trigonal insulin and (c) cubic insulin. The calculated patterns (blue) are corrected for bulk-solvent and geometrical effects using the revised Lorentz factor. From Hartmann et al. (2010[link]).

Separating the background from the diffraction pattern is not straightforward (Frankaer et al., 2011[link]). Estimation of the background is greatly assisted by a correct calculated pattern. The calculated pattern can be scaled to the experimental data using PROTPOW (http://www.kemi.dtu.dk/english/Research/PhysicalChemistry/Protein_og_roentgenkrystallografi/Protpow ).

Ståhl et al. (2013[link]) have demonstrated that existing search/match procedures can be used to identify proteins using their powder patterns, and that powder patterns calculated from Protein Data Bank coordinates with proper care can be added to a database and included in the search/match procedure. Several problems can be foreseen when including large amounts of protein data into the Powder Diffraction File. It may be worthwhile including powder patterns with several levels of solvent correction, rather than just an average value. Asymmetry from instrumental effects and specimen transparency, which can affect the peak positions, needs to be taken into account. The use of an average thermal expansion coefficient may be sufficient to account for the differences in lattice parameters between low-temperature single-crystal structures and powder patterns measured under ambient conditions.

References

Adams, P. D., Afonine, P. V., Bunkóczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J. J., Hung, L.-W., Kapral, G. J., Grosse-Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richardson, D. C., Richardson, J. S., Terwilliger, T. C. & Zwart, P. H. (2010). PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Cryst. D66, 213–221.Google Scholar
Allaire, M., Moiseeva, N., Botez, C. E., Engel, M. A. & Stephens, P. W. (2009). On the possibility of using polycrystalline material in the development of structure-based generic assays. Acta Cryst. D65, 379–382.Google Scholar
Basso, S., Besnard, C., Wright, J. P., Margiolaki, I., Fitch, A., Pattison, P. & Schiltz, M. (2010). Features of the secondary structure of a protein molecule from powder diffraction data. Acta Cryst. D66, 756–761.Google Scholar
Corey, R. B. & Wyckoff, R. W. G. (1936). Long spacings in macromolecular solids. J. Biol. Chem. 114, 407–414.Google Scholar
Dilanian, R. A., Darmanin, C., Varghese, J. N., Wilkins, S. W., Oka, T., Yagi, N., Quiney, H. M. & Nugent, K. A. (2011). A new approach for structure analysis of two-dimensional membrane protein crystals using X-ray powder diffraction data. Protein Sci. 20, 457–464.Google Scholar
Fokine, A. & Urzhumtsev, A. (2002). Flat bulk-solvent model: obtaining optimal parameters. Acta Cryst. D58, 1387–1392.Google Scholar
Frankaer, C. G., Harris, P. & Ståhl, K. (2011). A sample holder for in-house X-ray powder diffraction studies of protein powders. J. Appl. Cryst. 44, 1288–1290.Google Scholar
Frericks Schmidt, H. L., Sperling, L. J., Gao, Y. G., Wylie, B. J., Boettcher, J. M., Wilson, S. R. & Rienstra, C. M. (2007). Crystal polymorphism of protein GB1 examined by solid-state NMR spectroscopy and X-ray diffraction. J. Phys. Chem. B, 111, 14362–14369.Google Scholar
Fujii, K., Young, M. T. & Harris, K. D. M. (2011). Exploiting powder X-­ray diffraction for direct structure determination in structural biology: the P2X4 receptor trafficking motif YEQGL. J. Struct. Biol. 174, 461–467.Google Scholar
Hartmann, C. G., Nielsen, O. F., Ståhl, K. & Harris, P. (2010). In-house characterization of protein powder. J. Appl. Cryst. 43, 876–882.Google Scholar
Helliwell, J. R., Bell, A. M. T., Pryant, P., Fisher, S. J., Habash, G., Helliwell, M., Margiolaki, I., Kaenket, S., Watier, Y., Wright, J. P. & Yalamanchilli, S. (2010). Time-dependent analysis of K2PtBr6 binding to lysozyme studied by protein powder and single crystal X-ray analysis. Z. Kristallogr. 225, 570–575.Google Scholar
Hirakura, Y., Yamaguchi, H., Mizuno, M., Miyanishi, H., Ueda, S. & Kitamura, S. (2007). Detection of lot-to-lot variations in the amorphous microstructure of lyophilized protein formulations. Int. J. Pharm. 340, 34–41.Google Scholar
Jiang, J.-S. & Brünger, A. T. (1994). Protein hydration observed by X-ray diffraction. Solvation properties of penicillopepsin and neuraminidase crystal structures. J. Mol. Biol. 243, 100–115.Google Scholar
Kleywegt, G. J., Harris, M. R., Zou, J., Taylor, T. C., Wählby, A. & Jones, T. A. (2004). The Uppsala Electron-Density Server. Acta Cryst. D60, 2240–2249.Google Scholar
Margiolaki, I. & Wright, J. P. (2008). Powder crystallography on macromolecules. Acta Cryst. A64, 169–180.Google Scholar
Margiolaki, I., Wright, J. P., Wilmanns, M., Fitch, A. N. & Pinotsis, N. (2007). Second SH3 domain of ponsin solved from powder diffraction. J. Am. Chem. Soc. 129, 11865–11871.Google Scholar
Moews, P. C. & Kretsinger, R. H. (1975). Refinement of the structure of carp muscle calcium-binding parvalbumin by model building and difference Fourier analysis. J. Mol. Biol. 91, 201–225.Google Scholar
Norrman, M., Ståhl, K., Schluckebier, G. & Al-Karadaghi, S. (2006). Characterization of insulin microcrystals using powder diffraction and multivariate data analysis. J. Appl. Cryst. 39, 391–400.Google Scholar
Papageorgiou, N., Watier, Y., Saunders, L., Coutard, B., Lantez, V., Gould, E. A., Fitch, A. N., Wright, J. P., Canard, B. & Margiolaki, I. (2010). Preliminary insights into the non structural protein 3 macro domain of the Mayaro virus by powder diffraction. Z. Kristallogr. 225, 576–580.Google Scholar
Phillips, S. E. V. (1980). Structure and refinement of oxymyoglobin at 1.6 Å resolution. J. Mol. Biol. 142, 531–554.Google Scholar
Rigaku (2011). COD for PDXL: Integrated Powder X-ray Diffraction Software. http://www.crystallography.net/archives/2011/Rigaku .Google Scholar
Rotella, F. J., Duke, N. & Kaduk, J. A. (1998). X-ray powder diffraction from biological macromolecules. What do we see? What can we tell? American Crystallographic Association Meeting Abstracts, Washington DC, 18–23 July. Buffalo: American Crystallographic Association.Google Scholar
Rotella, F. J., Duke, N. & Kaduk, J. A. (2000). Powder diffraction from biological macromolecules using synchrotron X-rays. American Crystallographic Association Meeting Abstracts, St Paul MN, 22–27 July. Buffalo: American Crystallographic Association.Google Scholar
Ståhl, K., Frankaer, C. G., Petersen, J. & Harris, P. (2013). Monitoring protein precipitates by in-house X-ray powder diffraction. Powder Diffr. 28, S448–S457.Google Scholar
Von Dreele, R. B. (1998). Protein structures by powder diffraction? Abstr. Pap. Am. Chem. Soc. 215, U759.Google Scholar
Von Dreele, R. B. (1999). Combined Rietveld and stereochemical restraint refinement of a protein crystal structure. J. Appl. Cryst. 32, 1084–1089.Google Scholar
Von Dreele, R. B. (2003). Protein crystal structure analysis from high-resolution X-ray powder-diffraction data. Methods Enzymol. 368, 254–267.Google Scholar
Von Dreele, R. B. (2007a). Binding of N-acetylglucosamine oligosaccharides to hen egg-white lysozyme: a powder diffraction study. Acta Cryst. D61, 22–32.Google Scholar
Von Dreele, R. B. (2007b). Multipattern Rietveld refinement of protein powder data: an approach to higher resolution. J. Appl. Cryst. 40, 133–143.Google Scholar
Von Dreele, R. B., Stephens, P. W., Smith, G. D. & Blessing, R. H. (2000). The first protein crystal structure determined from high-resolution X-­ray powder diffraction data: a variant of T3R3 human insulin–zinc complex produced by grinding. Acta Cryst. D56, 1549–1553.Google Scholar
Wyckoff, R. W. G. & Corey, R. B. (1936). X-ray diffraction patterns of crystalline tobacco mosaic proteins. J. Biol. Chem. 116, 51–55.Google Scholar








































to end of page
to top of page