Anaerobic crystallization of proteins
Crystallization has been a bottleneck in the X-ray crystallography of proteins. Although many techniques have been developed to overcome this obstacle, the impurities caused by chemical reactions during crystallization have not been sufficiently considered. Oxidation of proteins, which can lead to poor reproducibility of the crystallization, is a prominent example. Protein oxidization in the crystallization droplet causes inter-molecular disulfide bridge formation, formation of oxidation film, and precipitation of proteins. These changes by oxidation are typically irreversible. The best approach for preventing protein oxidization during crystallization is anaerobic crystallization. Here we review the anaerobic crystallization of proteins, which was originally developed to trap a reaction intermediate of the enzyme in the crystal. We also summarize representative anaerobic crystallizations from our laboratory and the general setup of anaerobic crystallization.
KeywordsX-ray crystallography Crystallization Anaerobic conditions Oxidation Disulfide bridge Reproducibility
Since the tertiary structure of proteins (or biological macromolecules) is a key to analyzing the molecular mechanisms underlying cellular biological processes, much research effort has been expended in analyzing tertiary structures. As a result, more than 130,000 coordinates have been deposited to date in the Protein Data Bank (Burley et al. 2017). While the use of single particle analysis by cryo-electron microscopy has expanded rapidly in recent years (Fernandez-Leiro and Scheres 2016), X-ray crystallography remains a major tool to determine the tertiary structures of proteins. One of the main advantages of X-ray crystallography is its high-throughput nature. The combination of a synchrotron light source, robotics, a pixel array type X-ray detector, and sophisticated software enable more than 100 diffraction data sets to be collected easily within 1 day. Once all these procedures are optimized, it should be possible to collect more than 500 data sets in that same time period. However, to achieve a high-throughput analysis, the biological macromolecule of interest needs to be purified and crystallized—and this step constitutes a bottleneck. To date, many techniques have been developed to overcome this bottleneck. For example, the sparse matrix method has greatly improved the success rate of the initial crystallization screening (Jancarik and Kim 1991). As a result, many screening kits are now commercially available. In addition, robotics has been introduced as a means to improve the efficiency and accuracy of the initial crystallization screening. Indeed, several crystallization robots are commercially available, including a few large-scale crystallization robots for protein crystallization (Morris et al. 1989). Since robots can handle small amounts of sample with high accuracy, their utilization is effective for realizing experiments with high reproducibility. In our laboratory, a large-scale, fully automated crystallization robot is in operation, which can perform every process of the crystallization as well as observe the crystallization droplets (Hiraki et al. 2006).
In addition to the crystallization methods themselves, new methods for the expression and purification of proteins have also accelerated the process of protein crystallization. Both bacterial expression systems and systems using eukaryotic cells have been developed, thereby expanding our targets for structural analysis. The combination of a high-yield expression system and a tag-purification method has improved the quality of purified recombinant proteins. Moreover, although the reconstitution of large protein complexes was previously considered impossible, it has become possible to reconstitute large protein complexes by using recombinant DNA techniques (Imasaki et al. 2011; Bieniossek et al. 2013; Nozawa et al. 2017).
In addition to these developments, assessment of the physicochemical purity, or physicochemical uniformity, has become a common procedure in recent years because conformational variety and thermodynamic instability are not advantageous for achieving crystallization. The physicochemical purity of samples has been pursued via several methods, of which the most frequently used is the truncation of intrinsically disordered regions of the target protein (Chait 1994). In the case of multi-domain proteins, isolation of a single domain is an effective method for crystallization. Since the relative arrangement of the domains can easily be changed in the multi-domain protein, a reasonable strategy is to isolate a domain for crystallization. In addition, the thermodynamic stability of the protein is often assessed before crystallization using the differential scanning fluorimetry (DSF) method (Ericsson et al. 2006), which can quickly estimate the melting temperature (Tm). One of the striking examples of using thermo-stabilization in protein crystallization is the introduction of a T4-lysozyme in a loop region of the G-protein-coupled receptors prior to their crystallization (Cherezov et al. 2007). Many membrane proteins have also been crystallized via thermo-stabilization (Hattori et al. 2012). On the basis of these results, the DSF method has become a popular approach for analyzing the stability of proteins. Indeed, the optimization of buffer conditions using DSF has proven effective for crystallization. Size exclusion chromatography is also routinely utilized to quickly assess the physicochemical properties of a purified protein sample, such as its aggregation and physicochemical heterogeneity.
While these advanced techniques have significantly improved the success rate of protein crystallization, several problems remain. For example, chemical changes of the protein in the crystallization droplet can lead to crystallization failure. A well-known example of just such a chemical change during crystallization is oxidation of the protein sample, which causes inter-molecular disulfide bridge formation, formation of oxidation film, and precipitation of proteins. In our experience, the success rate of the crystallization can be improved by avoiding oxidation of the crystallization sample. In this review, therefore, we provide a brief history of the anaerobic crystallization method used to trap a reaction intermediate and describe the possible advantages of this method, as well as summarizing our experiences with this technique.
Anaerobic crystallization to isolate reaction intermediates
Anaerobic conditions in protein crystallography have been utilized to isolate reaction intermediates of enzymes with a redox center. The redox center of the enzyme controls the reactivity of the enzyme by changing its redox state. When the reduced state is a reactive form of the enzyme, structural information of the reduced form should be obtained to explain the catalytic reaction mechanism. However, in many cases, the reduced form is unstable under aerobic conditions due to its oxidation by oxygen molecules. Thus, to trap the reduced form of the enzyme in the crystal, the enzyme should be crystallized under anaerobic conditions. It would also be possible to reduce a crystal of the oxidized form of the enzyme by soaking the crystal in a solution supplemented with a reducing reagent under anaerobic conditions, unless the crystal is broken by the reduction of the enzyme.
One of the first examples of anaerobic crystallization was that of isopenicillin N-synthase by the Baldwin group. These researchers crystallized a substrate complex form of the isopenicillin N-synthase under anaerobic conditions, revealing the crystal structure of its reaction intermediate (Roach et al. 1996, 1997). Following this advance, many other crystal structures of reaction intermediates that were prepared under anaerobic conditions were reported, including flavoproteins (Senda et al. 2007a), proteins with an iron-sulfur cluster (Colbert et al. 2000; Hsieh et al. 2005; Senda et al. 2007a), and non-heme iron proteins (Uragami et al. 2001; Vaillancourt et al. 2002; Sato et al. 2002). A trapped reaction intermediate can also be utilized to obtain the following reaction intermediate of the catalytic reaction. In the case of naphthalene dioxygenase, an oxygen molecule was introduced into a frozen crystal of a reaction intermediate, yielding an O2-binding form of this enzyme (Karlsson et al. 2003). Several similar examples have also been reported (Tocheva et al. 2004; Kovaleva and Lipscomb 2007). A redox-dependent complex was also crystallized under anaerobic conditions as follows. Ferredoxin and ferredoxin reductase interact in a redox-dependent manner. Upon the reduction of flavin adenine dinucleotide (FAD), BphA4, an FAD-containing ferredoxin reductase, increases its affinity for ferredoxin BphA3. The BphA3–BphA4 complex can then be isolated under anaerobic conditions (Senda et al. 2007b), as can the crystals of this redox-dependent BphA3–BphA4 complex (Senda et al. 2007c).
Anaerobic crystallization to avoid oxidation
Another advantage of anaerobic crystallization is that chemical purity can be maintained by avoiding oxidation of the target protein and crystallization reagents. When free Cys residues are present in the target protein, these residues can form an intra/inter-molecular disulfide bridge(s), causing heterogeneity of the protein in solution. This obstruction of crystallization by oxidation is a well-established phenomenon. The addition of a reducing reagent is a typical strategy to avoid the random formation of the intra/inter-molecular disulfide bridges. However, since it is rather difficult to completely avoid oxidation of the sample for a long period under aerobic conditions, the reproducibility of the experiment tends to be poor. Furthermore, oxidization of the reducing reagent results in heterogeneity of the crystallization solution, which would cause problems with the crystallization. It is therefore quite difficult to maintain all the conditions the same across many crystallization trials using a reducing reagent. To avoid heterogeneity in the crystallization solution, the use of anaerobic conditions is a potent method. For example, anaerobic crystallization can easily avoid the formation of an inter-molecular disulfide bridge and increase the reproducibility of crystallization, as observed in the crystallization of oxidized and reduced forms of BphA3 (Senda et al. 2007b).
General setup for anaerobic crystallization
For convenient anaerobic crystallization, it is preferable to equip the chamber with an incubator and a microscope. By sealing the crystallization plate completely with vacuum grease, such as Apiezon grease (Apiezon, Manchester UK), it becomes possible to remove the crystallization plate from the anaerobic chamber and keep it in an aerobic incubator. However, it is awkward and inconvenient to handle a crystallization plate sealed with vacuum grease. Therefore, it is preferable to perform the incubation and observation of the crystallization plates in the anaerobic chamber. Since a microscope can easily be equipped with a CCD camera, images of crystallization droplets can be observed through a monitor inside or outside the chamber. If the microscope is in the anaerobic chamber, it is possible to freeze the crystal inside the anaerobic chamber.
Effects of anaerobic crystallization
Several trials of crystallization screening of the tandem SH2 domain with various phosphorylated peptides were performed under anaerobic conditions, and eventually micro-crystals of the tandem SH2 domain in complex with the phosphorylated EPIYA-C peptide were obtained. After optimization of the crystallization conditions, rod-shaped crystals were obtained in a reproducible manner under anaerobic conditions in the sitting drop setting (Fig. 4b). Interestingly, the same crystallization conditions rarely gave crystals under aerobic conditions; precipitation occurred in most of these trials (Fig. 4b), suggesting that oxidation of the protein and/or crystallization reagents has adverse effects on crystallization. In many cases, oxidation of the proteins and/or crystallization reagents seem to cause precipitation, which reduces the concentration of protein in the droplet and hampers protein crystallization. Since oxidized protein molecules can be regarded as impurities of the protein solution, oxidized protein molecules hamper protein crystallization. Moreover, in cases in which an oxidation film would be formed on the surface of the crystallization droplet under aerobic conditions, anaerobic crystallization will still be effective because the oxidation film will not form on the droplet surface under anaerobic conditions. Since formation of the oxidation film is typically irreversible and would reduce the concentration of the protein in the droplet, the oxidation film impairs the reproducibility of the crystallization experiment.
While the oxidation of proteins and crystallization reagents has been recognized for some time, there has been no systematic search for fundamental solutions to the problem. Here, we have reviewed the topic of anaerobic crystallization and described how it frequently improves the results and reproducibility of crystallization experiments. Since the oxidation of the protein in the crystallization droplet causes a reduction in the purity of the protein, avoiding oxidation under the anaerobic conditions is theoretically reasonable.
We thank Drs M. Fukuda, E. Masai, S. Kimura, M. Hatakeyama, and T. Hayashi for their continuous collaborations and providing us with protein samples. This work was partly supported by the Platform for Drug Discovery, Informatics, and Structural Life Science (PDIS) and Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS) from AMED, Japan.
Compliance with ethical standards
Conflict of interest
Miki Senda declares that she has no conflicts of interest. Toshiya Senda declares that he has no conflicts of interest.
This article does not contain any studies with human participants or animals performed by any of the authors.
- Burley SK, Berman HM, Kleywegt GJ, Markley JL, Nakamura H, Velankar S (2017) Protein Data Bank (PDB): the single global macromolecular structure archive. In: Wlodawer A, Daughter Z, Jaskolski M (eds) Methods in molecular biology: protein crystallography methods and protocols, 1st edn. Springer, New York, pp 627–641Google Scholar
- Ohnishi N, Yuasa H, Tanaka S, Sawa H, Miura M, Matsui A, Higashi H, Musashi M, Iwabuchi K, Suzuki M, Yamada M, Azuma T, Hatakeyama M (2008) Transgenic expression of Helicobacter pylori CagA induces gastrointestinal and hematopoietic neoplasms in mouse. Proc Natl Acad Sci USA 105:1003–1008CrossRefPubMedPubMedCentralGoogle Scholar
- Senda M, Kimura S, Kishigami S, Senda T (2006) Crystallization and preliminary X-ray analysis of the Rieske-type [2Fe-2S] ferredoxin component of biphenyl dioxygenase from Pseudomonas sp. strain KKS102. Acta Crystallogr F62:590–592Google Scholar
- Senda M, Kishigami S, Kimura S, Senda T (2007b) Crystallization and preliminary X-ray analysis of the reduced Rieske-type [2Fe-2S] ferredoxin derived from Pseudomonas sp. strain KKS102. Acta Crystallogr F63:311–314Google Scholar
- Senda M, Kishigami S, Kimura S, Senda T (2007c) Crystallization and preliminary X-ray analysis of the electron-transfer complex of Rieske-type [2Fe-2S] ferredoxin and NADH-dependent ferredoxin reductase derived from Acidovorax sp. strain KKS102. Acta Crystallogr F53:520–523Google Scholar
- Tocheva EI, Rosell FI, Mauk AG, Murphy ME (2004) Side-on copper-nitrosyl coordination by nitrite reductase. Science 304:867–870Google Scholar
- Vaillancourt FH, Barbosa CJ, Spiro TG, Bolin JT, Blades MW, Turner RF, Eltis LD (2002) Definitive evidence for monoanionic binding of 2,3-dihydroxybiphenyl to 2,3-dihydroxybiphenyl 1,2-dioxygenase from UV resonance Raman spectroscopy, UV/Vis absorption spectroscopy, and crystallography. J Am Chem Soc 124:2485–2496CrossRefPubMedGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.