Regular article
Can anomalous signal of sulfur become a tool for solving protein crystal structures?1

https://doi.org/10.1006/jmbi.1999.2743Get rights and content

Abstract

A general method for solving the phase problem from native crystals of macromolecules has long eluded structural biology. For well diffracting crystals this goal can now be achieved, as is shown here, thanks to modern data collection techniques and new statistical phasing algorithms. Using solely a native crystal of tetragonal hen egg-white lysozyme, a protein of 14 kDa molecular mass, it was possible to detect the positions of the ten sulfur and seven chlorine atoms from their anomalous signal, and proceed from there to obtain an electron-density map of very high quality.

Introduction

If one could measure from a crystallography experiment the amplitude and phase of all X-rays diffracted from the crystal, or at least most of them, to a given resolution, a simple Fourier transform would yield an electron-density map at that resolution. From there, the interpretation of this map in terms of an atomic model would proceed. Because only the intensity of the diffracted rays can be measured, various methods were established to obtain an approximate estimation of phases, that would lead, via the Fourier transform, to an electron-density map where molecular features can be recognized and interpreted.

For most “small” molecules (less than 100 atoms) it is possible to obtain a good approximation of the phases from the experimental knowledge of the diffracted intensities only. Recently direct methods have been applied to macromolecules, but they still require data extending to exceptionally high resolution, at least to 1.2 Å. If the atomic structure of a related molecule is known, the phase problem can sometimes be solved by the method of molecular replacement, by modifying one to fit the diffracted intensities of the other. In all other cases, the phase problem can only be solved by a perturbation method: introducing a heavy atom into the molecule in the crystal (isomorphous replacement), or taking advantage of the resonant diffraction behavior of some atoms at specific X-ray wavelengths (anomalous diffraction). This latter method yields weaker signals, but requires less chemical manipulation of the crystals, and provides more reliable phase information to high resolution.

Atoms differ widely in the strength of their resonant scattering behaviour. In the range of wavelengths used for X-ray diffraction, lanthanides exhibit the largest effects. Standard heavy metals, such as Hg or Pt have δf” of about ten electrons at a wavelength just below their LIII edges and about four electrons above the edge, which is the worst wavelength selection for anomalous signal. With the copper radiation the anomalous contribution of those metals is about seven electrons. Selenium, often used for multiwavelength anomalous dispersion (MAD) phasing through incorporation of selenomethionine residues instead of normal methionine residues into proteins, has an δf” value of about four below its K edge, 0.5 above it and about 1.1 anomalous electrons for the copper radiation. In MAD work with selenium precautions are taken to ensure the highest possible accuracy of the measured intensities. Crystals are often aligned to measure the Bijvoet-related reflections on the same exposure or “inverse beam” technique utilized to record Friedel mates close in time, to make sure that the measurements of both mates, F+ and F, are done with similar paths of the X-ray beams in the crystal, so as to minimize the systematic errors in the estimation of ΔFanom. Local scaling of intensities during merging of symmetry equivalent reflections serves the same purpose of minimizing the errors, therefore enhancing the weak anomalous signal present in the X-ray data. Three or four data sets at different wavelengths are usually collected for MAD work, allowing us to exploit different combinations of δf” and δf” contributions. However, with accurate enough measurements it should be possible to obtain phase estimations from a single wavelength data set, using only the anomalous signal based on δf” contribution.

The K X-ray absorption edges of the elements of the third period lie in the low energy region, below 3 keV or in terms of wavelength, beyond 4 Å. It is therefore not realistically possible to utilize the maximum anomalous scattering effect of such elements as chlorine, sulfur or phosphorus for phasing of X-ray diffraction data. However, those elements retain some anomalous scattering effect even far from their absorption edges, at wavelengths that are commonly used for collecting X-ray data. At the copper Kα wavelength of 1.54 Å, a sulfur atom has a δf” value of 0.56 and chlorine 0.70 anomalous electron as estimated from program CROSSEC (Cromer, 1983). Close to 1 Å, often used at synchrotron beamlines, S and Cl have δf” values of 0.24 and 0.31 electron, respectively.

A seminal paper of Hendrickson & Teeter (1981) describes the structure solution of the small protein crambin from the anomalous diffraction of the sulfur atoms naturally present in the molecule. This points to the possibility of a universal phasing method, since sulfur is present in almost all proteins. Phosphorus can be used for the same purpose in nucleic acids. The main obstacle so far has been the quality of the data required to take advantage of the vanishingly small signal of the sulfur atom. We demonstrate that it is now becoming possible by using currently available data collection and statistical phasing methods.

Section snippets

Data acquisition and phasing

We have collected X-ray data on the tetragonal crystal of hen egg-white lysozyme (HEWL) using synchrotron radiation of 1.54 Å, thus mimicking the home laboratory copper anode source, and the MAR345 imaging plate scanner. The data collection protocol was standard for single wavelength experiments, except for the high multiplicity of observations, resulting mainly from four measurement passes with different resolution limits and exposure times (Table 1). The high symmetry of the crystal also

Anion sites

One chloride site was found in the early structure determination of tetragonal HEWL (Blake et al., 1967) on the basis of its low B factor. In a recent paper (Lim et al., 1998) four bromide ions in tetragonal HEWL were identified and a probable fifth site postulated on the basis of the difference in electron density between bromine and chlorine atoms. In other crystal forms HEWL is known to bind nitrates and acetates (Walsh et al., 1998) or iodides (Steinrauf, 1998). The sites of eight chloride

Conclusions

The small but significant anomalous dispersion signal of 17 sulfur and chlorine atoms is enough to estimate the protein phases of lysozyme, a protein with 129 amino acid residues, by the single wavelength anomalous dispersion (SAD) approach. The data have been collected by standard procedures, except for high multiplicity of intensity measurements. The positions of anomalous scatterers were found by “half-baked” direct methods approach and the protein phases were obtained by a maximum

Methods

The tetragonal crystals of HEW lysozyme (from Sigma) were grown in 2 ml batches of 20 mg ml−1 protein solution in sodium acetate buffer (pH 4.6) containing 10 % (w/v) NaCl. Before using for data collection, the crystal was transferred for five seconds to the solution escribed above, but containing 30 % (v/v) glycerol.

A single specimen of size 0.3 mm × 0.3 mm × 0.4 mm was quickly transferred in the fiber loop to a stream of nitrogen gas and frozen at 100 K at the goniostat of the MAR345 imaging

References (21)

  • V.S Lamzin et al.

    Automated refinement for protein crystallography

    Methods Enzymol

    (1997)
  • Z Otwinowski et al.

    Processing of X-ray diffraction data collected in oscillation mode

    Methods Enzymol

    (1997)
  • J.P Abrahams

    Bias reduction in phase refinement by modified interference functionsintroducing the gamma correction

    Acta Crystallog. sect. D

    (1997)
  • C.C.F Blake et al.

    On the conformation of the hen egg-white lysozyme molecule

    Proc. Roy. Soc. ser. B

    (1967)
  • The CCP4 suiteprograms for protein crystallography

    Acta Crystallog. sect. D

    (1994)
  • K Cowtan

    DMan automated procedure for phase improvement by density modification

  • D.T Cromer

    Calculation of anomalous scattering factors at arbitrary wavelengths

    J. Appl. Crystallog

    (1983)
  • D.W.J Cruickshank

    Remarks about protein structure precision

    Acta Crystallog. sect. D

    (1999)
  • G.S French et al.

    On the treatment of negative intensity observations

    Acta Crystallog. sect. A

    (1978)
  • W.A Hendrickson et al.

    Structure of the hydrophobic protein crambin determined directly from the anomalous scattering of sulfur

    Nature

    (1981)
There are more references available in the full text version of this article.

Cited by (212)

  • De novo crystal structure determination of double stranded RNA binding domain using only the sulfur anomalous diffraction in SAD phasing

    2021, Current Research in Structural Biology
    Citation Excerpt :

    In addition, other atoms such as Ca, Cl, Mg, K, Na, are occasionally - and sometimes fortuitously (Hegde et al., 2017) - found within the protein crystal, as determined by the X-ray fluorescence emission spectrum. Their anomalous signal can be exploited together with the sulfur signal for successful native SAD phasing (Dauter et al., 1999; Goulet et al., 2010; Liu et al., 2012; Doutch et al., 2012; Liu and Hendrickson, 2015; Rose et al., 2015; Weinert et al., 2015; Olieric et al., 2016; Olczak and Cianci, 2018). Crambin was the first protein whose structure was solved by sulfur-SAD (Hendrickson and Teet, 1981).

  • Crystallographic phasing from weak anomalous signals

    2015, Current Opinion in Structural Biology
    Citation Excerpt :

    Native biomolecules contain intrinsic light anomalous scatterers, sulfur (Z = 16) in proteins and phosphorous (Z = 15) in nucleic acids. The resonant edges of these native light elements may not be readily accessible; however, off-resonance anomalous scattering from these elements at lower X-ray energy is weak but measurable, which is very attractive for phasing with no need for heavy-atom derivatization and selenomethionyl substitution, thus promising for automated SAD analysis [5–10,11•,12••]. Comparing to heavy atoms, the anomalous signals from light anomalous scatterers are much weaker, limiting their application for routine structure determination by native-SAD.

  • A solution-free crystal-mounting platform for native SAD

    2020, Acta Crystallographica Section D: Structural Biology
View all citing articles on Scopus
1

Edited by I. A. Wilson

View full text