research papers
The Azotobacter vinelandii suggests a domain swap
of Shethna protein II (FeSII) fromaDepartment of Life Sciences, Imperial College, London SW7 2AZ, United Kingdom, bTurkish Accelerator and Radiation Laboratory, Gőlbaşı, 06830 Ankara, Türkiye, cCambrium GmbH, Max-Urich-Strasse 3, 13355 Berlin, Germany, and dBioheuris Inc., 1100 Corporate Square Drive, St Louis, MO 63132, USA
*Correspondence e-mail: j.w.murray@imperial.ac.uk
The Azotobacter vinelandii FeSII protein forms an oxygen-resistant complex with the nitrogenase MoFe and Fe proteins. FeSII is an adrenodoxin-type that forms a dimer in solution. Previously, the was solved [Schlesier et al. (2016), J. Am. Chem. Soc. 138, 239–247] with five copies in the One copy is a normal adrenodoxin domain that forms a dimer with its mate. The other four copies are in an `open' conformation with a loop flipped out exposing the 2Fe–2S cluster. The open and closed conformations were interpreted as oxidized and reduced, respectively, and the large conformational change in the open configuration allowed binding to nitrogenase. Here, the structure of FeSII was independently solved in the same crystal form. The positioning of the atoms in the is similar to the earlier report. However, the interpretation of the structure is different. The `open' conformation is interpreted as the product of a crystallization-induced domain swap. The 2Fe–2S cluster is not exposed to solvent, but in the crystal its interacting helix is replaced by the same helix residues from a crystal symmetry mate. The domain swap is complicated, as it is unusual in being in the middle of the protein rather than at a terminus, and it creates arrangements of molecules that can be interpreted in multiple ways. It is also cautioned that crystal structures should be interpreted in terms of the contents of the entire crystal rather than of one asymmetric unit.
Keywords: Shethna protein II; FeSII protein; Azotobacter vinelandii; nitrogen fixation; oxygen protection; structural biology; domain swapping.
PDB reference: FeSII from Azotobacter vinelandii, 6yav
1. Introduction
Nitrogen is an essential component of DNA and proteins. Nitrogen gas makes up 80% of the atmosphere, but it is very unreactive. Fixing this unreactive nitrogen is an essential biological and industrial process. Biological 2 tetramer containing molybdenum. The MoFe protein is in turn reduced by the dinitrogenase reductase, the Fe protein, a dimer of NifH. MoFe has two oxygen-sensitive metalloclusters, the P-cluster and the FeMoCo, which contains the Mo atom. The Fe protein has a low-potential 4Fe–4S cluster at the surface bridging the dimer interface. Both nitrogenase components are irreversibly inhibited by oxygen, which reacts with the low-potential iron–sulfur clusters. MoFe has a half-life of around 10 min in air, and the Fe protein has a half-life of around 45 s (Robson, 1979; Robson & Postgate, 1980). The MoFe–Fe protein complex is partially protected from oxygen relative to each protein individually (Schlesier et al., 2016). As well as direct inactivation of the nitrogenase, reactions of oxygen with nitrogenase proteins probably generate damaging reactive oxygen species (Maier & Moshiri, 2000).
is catalysed by the enzyme nitrogenase, which reduces nitrogen gas to ammonia. The reduction of nitrogen is catalysed by the MoFe protein, an (NifDNifK)Organisms that fix nitrogen, known as diazotrophs, use several strategies to protect themselves from oxygen. Some are obligate anaerobes or micro-aerobes. Some diazotrophic aerobes, such as Azotobacter species, use `respiratory protection', a fast respiratory rate, to consume oxygen. A second mechanism is `conformational protection'. It was noticed that a nitrogen-fixing culture of A. chroococcum would stop fixing when oxygenated by strong agitation, but when the agitation stopped the culture would resume at the same rate as before (Hill et al., 1972). This result implied that the nitrogenase was temporarily being inactivated by a `conformational' mechanism.
In the 1960s, Shethna and coworkers isolated several nonheme iron proteins from A. vinelandii (Shethna et al., 1968). The second of these proteins became known as the Shethna protein or FeSII. Early work on purifying nitrogenase found a pink protein that associated with nitrogenase and provided some oxygen protection (Kelly et al., 1967). This pink protein was then identified as the FeSII protein (Bulen & LeComte, 1972; Haaker & Veeger, 1977). FeSII is the basis of the conformational protection (Robson, 1979). It forms a ternary complex with the MoFe and Fe proteins that is catalytically inactive, but from which active nitrogenase can be recovered, with a half-life in air of about an hour (Robson, 1979). FeSII is not essential for A. vinelandii viability, even under nitrogen-fixing conditions, but the nitrogenase was more sensitive to inactivation in an FeSII-knockout mutant (Moshiri et al., 1994).
FeSII has only been shown to have protective activity in A. vinelandii and A. chroococcum, although FeSII-like functionality has been proposed for other ferredoxins associated with nitrogenase operons, for example FdxD in Rhodobacter capsulatus (Hoffmann et al., 2014) and a Gluconacetobacter (Lery et al., 2010; Ureta & Nordlund, 2002). It is possible that the protective function has appeared more than once, or that similar mechanisms exist in other diazotrophs but have not yet been identified.
FeSII is a 13 kDa adrenodoxin-type et al., 1995). Its crystallization was published in 1995 (Moshiri et al., 1995) in an orthorhombic crystal form, but no structure was solved. Recently, Schlesier et al. (2016) reported a 2.2 Å resolution of FeSII in the same orthorhombic crystal form (PDB entry 5ffi). The observed FeSII crystal form has five copies in the One of them is in a `closed conformation', forming a `closed' dimer with a crystallographic symmetry mate. The closed conformation is similar to other adrenodoxin proteins, such as the monomeric Escherichia coli Fdx (Kakuta et al., 2001). The other four copies in the are in an `open' from. In the open form, the so-called N-loop consisting of residues 59–96 is flipped out from the structure so that the 2Fe–2S cluster is revealed. Specifically, the `lid helix' from residues 68 to 78 is moved away from the cluster. The four open FeSII molecules were interpreted as two dimers of the same type as the closed form, but with the N-loop flipped out. Schlesier and coworkers proposed that this conformational change is triggered by oxidation of the cluster, and the change enables FeSII to form the protective ternary complex with the MoFe and Fe proteins. Small shifts in the elution time from in different oxidation states were also consistent with a conformational change. However, the open form of FeSII has not been observed in crystals of any other adrenodoxin-type domains.
protein (midpoint potential −262 mV) with a 2Fe–2S cluster that forms a dimer in solution (MoshiriProtein domain swapping is when two protein molecules exchange equivalent regions in their structures (Liu & Eisenberg, 2002; Wodak et al., 2015). In the simplest form, two adjacent monomers swap a region to form a domain-swapped dimer. Some proteins have evolved natural domain swaps relative to an unswapped ancestor (Cahyono et al., 2020). In other cases the swap is functional or may be involved in disease (Rousseau et al., 2012). Domain swapping may also be induced by chemical treatment (Cahyono et al., 2020) or crystallization. For example, in a crystallized PAS domain one of the six copies had a domain-swapped C-terminal α-helix (Emami et al., 2009). In cyanobacterial Psb29 (Bečková et al., 2017) only one of three crystal forms had a domain-swapped N-terminal α-helix (Bečková et al., 2017). In these cases, the domain-swapped structure may be regarded as a crystallization artefact and the swapped structure is not biologically relevant. In most domain-swapped structures the swap is of one end of the structure containing the N- and C-terminus. Swaps of internal regions are rarer, but have been observed, such as in the structure of IX/X-binding protein from snake venom (Mizuno et al., 1997), which has a loop-exchange heterodimer relative to the homologous rat mannose-binding protein (Weis et al., 1991).
We have independently solved the structure of the FeSII protein in the same crystal form as previously but to a higher resolution. The atoms are in similar positions to the earlier structure, but our interpretation of the open form of the protein is very different. We believe that the open form is caused by a crystallization-induced domain swap. Here, we present evidence that the `open' conformation of FeSII in the
is best interpreted as a domain-swapped structure.2. Materials and methods
2.1. Overexpression and purification of FeSII
A synthetic gene coding for the FeSII from A. vinelandii (AVIN_39700) was synthesized by DNA 2.0 (California, USA) and subcloned into a pRSET-A vector (Invitrogen, Carlsbad, California, USA) with no tag. The FeSII gene was PCR-amplified with primers FESII_FOR, 5′-TATTACATATGGCGACGATCTATTTCAGCAGC-3′, and FESII_REV, 5′-ATTATCTCGAGTTATCAGGCGCCACCCG-3′ (restriction sites are underlined), and the product was digested with NdeI and XhoI, and then ligated into vector cut with NdeI and XhoI restriction enzymes.
FeSII protein was produced by transformation of the vector into chemically competent E. coli KRX cells (Promega, Wisconsin, USA) and cultivation in 1 l Terrific Broth medium supplemented with 100 µg l−1 ampicillin. Expression was induced at an OD600 of 0.6–0.8 with 1 g l−1 L-rhamnose monohydrate. After overnight induction at 18°C, the cells were harvested by centrifugation and resuspended in 10 mM HEPES pH 7.4, 1 mM MgCl2. Cell disruption was performed by sonication. FeSII was purified using a modification of a published method (Moshiri et al., 1995). 0.4%(w/v) NaCl and 11%(w/v) PEG 8000 were added as powder to the lysate. They were mixed gently for 45 min at 4°C and the cell debris was separated by centrifugation. The red supernatant was filtered through a 0.2 µm filter and purified by anion-exchange on a Q Sepharose column equilibrated with 10 mM HEPES pH 7.4, 1 mM MgCl2. FeSII was eluted using a linear gradient of MgCl2 from 0 to 50 mM over ten column volumes. The fractions corresponding to the FeSII protein were pooled and concentrated, and further purified by (HiLoad 16/600 Superdex 200 pg column) in 10 mM HEPES pH 7.4, 5 mM MgCl2, 100 mM NaCl.
2.2. of FeSII
FeSII was crystallized using hanging-drop vapour diffusion with an equal mixture of 0.1 M HEPES pH 8.0, 0.2 M sodium citrate, 24%(w/v) PEG 3350, 2%(v/v) 1-butanol and 40 mg ml−1 FeSII. The crystals were cryoprotected for 30 s in the mother-liquor solution with 30%(v/v) PEG 400 added and were flash-cooled in liquid nitrogen. X-ray diffraction data were collected at 100 K on beamline I03 at Diamond Light Source, UK. For phase determination, a data set was collected at the Fe K edge at a wavelength of 1.734 Å. The data were processed and scaled with xia2 (Winter, 2010) and XDS (Kabsch, 2010). The structure was phased with Fe-SAD using the intrinsic Fe atoms of the protein using AutoSol (Terwilliger et al., 2009) from the Phenix package (Adams et al., 2010; Liebschner et al., 2019). The phases were improved by density modification in DM (Cowtan, 2010) and a preliminary model was built with Buccaneer (Cowtan, 2006). The model was rebuilt manually in Coot (Emsley et al., 2010), cycling with in phenix.refine (Afonine et al., 2012). This model was then refined and rebuilt against a higher resolution native data set at 1.65 Å. The structure was validated with MolProbity (Williams et al., 2018). The final data-collection and are given in Table 1. Dimer surface areas and interaction energies were predicted with PISA (Krissinel, 2010; Krissinel & Henrick, 2007). Molecular figures were produced in PyMOL (DeLano, 2002).
|
3. Results
3.1. FeSII structure
We independently solved the FeSII structure using Fe-SAD to a final resolution of 1.65 Å in the same orthorhombic P21212 crystal form as observed in 1995 (Moshiri et al., 1995) and 2016 (Schlesier et al., 2016). Data-collection and information is given in Table 1. The contains five copies of FeSII. One is in the closed conformation similar to other adrenodoxins. The other four are in the open conformation. Although the crystal form is the same, we have chosen symmetry positions of the chains so that the contains both of the types of dimer that we describe.
The structures are similar, with r.m.s.d.s of 1.1 Å over the closed chain between this work and Schlesier et al. (2016) and of 0.4 Å between the best-matching open chains. The closed chain is more complete than before, with a gap from residues 82 to 92 in the N-loop. In the open chains the N-loop is ordered, and the entire chain is visible, apart from the N- and C-termini.
3.2. Interpretation of the crystal structure
In this crystal form the `closed' chain E forms a dimer with its mate E′ (−x + 1, −y, z). All four open chains form the `open dimer'. Chains A and B form this dimer. Chain C forms this dimer with chain D from the x, y, z − 1 symmetry mate (chain D′). Chains C and D form another extended interface forming what we call the `domain-swap' dimer. Chain A also forms this domain-swap dimer with the B chain related by x, y, z − 1 (chain B′). All open chains are involved in both the open and domain-swap dimers. Fig. 1 shows the dimers in the crystal and the chains forming them for two unit cells.
The work by Schlesier and coworkers proposed that the `lid helix' moves away from the 2Fe–2S cluster in the open dimer, exposing it to solvent. However, in the crystal the lid helix in each open chain is replaced by a lid helix from a symmetry-related chain. Fig. 2(a) shows the lid helix in place for chain E, the closed chain. For the open chain C, Fig. 2(b) shows the lid helix in an equivalent position, but coming from the domain-swap dimer partner chain D in the crystal. In chains A and B, the lid helix is similarly replaced by symmetry-related chains B′ and A′, respectively.
The predicted interface areas, free energies of dissociation and binding energies for the dimer interfaces that exist in the crystal are shown in Table 2. The calculated area of the closed dimer interface is 490 Å2.
|
In the open conformation, the `lid helix' (residues 68–79), as part of the longer `N-loop' (residues 59–96), is folded away from the central domain. It forms additional contacts with the closed dimer partner around Pro12, expanding the closed dimer interface. The calculated area of the closed dimer interface is 490 Å2, but the mean open dimer interface area is increased by 1120 Å2. In the crystal, the open dimers are also involved in the domain-swap dimer. This dimer interface is primarily made up of residues of the opened-out N-loop, but has a mean area of 970 Å2, and is the most energetically favoured of the interfaces. The open and domain-swapped dimers have slightly positive predicted free energies of dissociation (ΔGdiss) ranging from 6.2 to 6.7 kcal mol−1, suggesting a stable dimer. The closed dimer has a predicted ΔGdiss of −5.8 kcal mol−1, implying a marginally stable dimer within the error of the prediction. These predicted energies suggest that the formation of the open and domain-swap interfaces together is energetically favoured in the crystal.
The dimers in the FeSII crystal are shown in Fig. 3 in cartoon and schematic views. The replacement of the lid helix by the lid helix from a symmetry mate (red in Fig. 3) may be interpreted as a domain swap of the N-loop. This domain swap is unusual in that it involves an internal region of the protein rather than an N- or C-terminus. The swap is also unusual in that it creates an extensive new dimer interface, in addition to the dimer interface that is present in the open and closed dimers. Given that these two interfaces can only occur together in the crystal, we suggest that this open conformation of FeSII is induced by crystallization. The flexibility of the `hinge' regions proposed earlier may allow this protein to change shape under the constraints of formation to a more energetically favoured conformation that may not be biologically relevant.
Schlesier and coworkers proposed that the formation of the open dimers is induced by the oxidation of the 2Fe–2S cluster. Our protein and crystals were prepared aerobically, and our protein has an Supplementary Fig. S1) so even the closed chain should be oxidized, and thus any conformational change is not driven by differential oxidation states of the 2Fe–2S cluster.
corresponding to oxidized protein (4. Discussion
Previous research has claimed that the FeSII protein adopts a radically changed conformation in the oxidized state, based on the Rhodobacter capsulatus (Sainz et al., 2006) solved structures of the oxidized and the reduced protein (Armengaud et al., 2001; Sainz et al., 2006) and only small conformational changes were seen, mainly around the iron–sulfur cluster. Based on the behaviour of all other known ferredoxins that function without a large conformational change, it is probable that FeSII can also regulate its interaction with nitrogenase with only small conformational changes. We await of the full MoFe–Fe–FeSII protein complex for a definitive answer to this question.
and small shifts in in We have independently solved the FeSII at 1.65 Å resolution. Based on examination of the full of the crystal, we conclude that the `open' structure seen in the crystal is most likely due to crystallization-induced domain-swapping. Previous work on the homologous VI fromSupporting information
PDB reference: FeSII from Azotobacter vinelandii, 6yav
Supplementary Figure S1. DOI: https://doi.org/10.1107/S2059798324005928/rr5241sup1.pdf
Acknowledgements
The authors would like to thank Diamond Light Source for beam time (proposal mx12579) and the staff of beamlines I02, I03 and I24 for assistance with crystal testing and data collection. Author contributions were as follows. JWM, LL, CARC, AK, AS, CRM and BVK performed research. JWM drafted the paper in consultation with the other authors.
Conflict of interest
The authors declare that they have no conflicts of interest with the contents of this article.
Funding information
This work was funded in part by the BBSRC/NSF Nitrogen Ideas Lab: Oxygen-Tolerant Nitrogenase (BB/L011468/1). The crystallization facility at Imperial College was funded by BBSRC (BB/D524840/1) and the Wellcome Trust (202926/Z/16/Z). CRM was funded by a BBSRC Doctoral Training Programme (BB/J014575/1), as was CARC (BB/F017324/1).
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). Acta Cryst. D66, 213–221. Web of Science CrossRef CAS IUCr Journals Google Scholar
Afonine, P. V., Grosse-Kunstleve, R. W., Echols, N., Headd, J. J., Moriarty, N. W., Mustyakimov, M., Terwilliger, T. C., Urzhumtsev, A., Zwart, P. H. & Adams, P. D. (2012). Acta Cryst. D68, 352–367. Web of Science CrossRef CAS IUCr Journals Google Scholar
Armengaud, J., Sainz, G., Jouanneau, Y. & Sieker, L. C. (2001). Acta Cryst. D57, 301–303. Web of Science CrossRef CAS IUCr Journals Google Scholar
Bečková, M., Yu, J., Krynická, V., Kozlo, A., Shao, S., Koník, P., Komenda, J., Murray, J. W. & Nixon, P. J. (2017). Philos. Trans. R. Soc. B, 372, 20160394. Google Scholar
Bulen, W. A. & LeComte, J. R. (1972). Methods Enzymol. 24, 456–470. CAS PubMed Google Scholar
Cahyono, R. N., Yamanaka, M., Nagao, S., Shibata, N., Higuchi, Y. & Hirota, S. (2020). Metallomics, 12, 337–345. Web of Science CrossRef CAS PubMed Google Scholar
Cowtan, K. (2006). Acta Cryst. D62, 1002–1011. Web of Science CrossRef CAS IUCr Journals Google Scholar
Cowtan, K. (2010). Acta Cryst. D66, 470–478. Web of Science CrossRef CAS IUCr Journals Google Scholar
DeLano, W. L. (2002). PyMOL. http://www.pymol.org. Google Scholar
Emami, K., Topakas, E., Nagy, T., Henshaw, J., Jackson, K. A., Nelson, K. E., Mongodin, E. F., Murray, J. W., Lewis, R. J. & Gilbert, H. J. (2009). J. Biol. Chem. 284, 1086–1096. Web of Science CrossRef PubMed CAS Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. Web of Science CrossRef CAS IUCr Journals Google Scholar
Haaker, H. & Veeger, C. (1977). Eur. J. Biochem. 77, 1–10. CrossRef CAS PubMed Web of Science Google Scholar
Hill, S., Drozd, J. W. & Postgate, J. R. (1972). J. Appl. Chem. 22, 541–558. CAS Google Scholar
Hoffmann, M.-C. C., Müller, A., Fehringer, M., Pfänder, Y., Narberhaus, F. & Masepohl, B. (2014). J. Bacteriol. 196, 633–640. Web of Science CrossRef PubMed Google Scholar
Kabsch, W. (2010). Acta Cryst. D66, 133–144. Web of Science CrossRef CAS IUCr Journals Google Scholar
Kakuta, Y., Horio, T., Takahashi, Y. & Fukuyama, K. (2001). Biochemistry, 40, 11007–11012. Web of Science CrossRef PubMed CAS Google Scholar
Kelly, M., Klucas, R. V. & Burris, R. H. (1967). Biochem. J. 105, 3C–5C. CrossRef CAS PubMed Web of Science Google Scholar
Krissinel, E. (2010). J. Comput. Chem. 31, 133–143. Web of Science CrossRef PubMed CAS Google Scholar
Krissinel, E. & Henrick, K. (2007). J. Mol. Biol. 372, 774–797. Web of Science CrossRef PubMed CAS Google Scholar
Lery, L. M., Bitar, M., Costa, M. G., Rössle, S. C. & Bisch, P. M. (2010). BMC Genomics, 11, S7. Google Scholar
Liebschner, D., Afonine, P. V., Baker, M. L., Bunkóczi, G., Chen, V. B., Croll, T. I., Hintze, B., Hung, L.-W., Jain, S., McCoy, A. J., Moriarty, N. W., Oeffner, R. D., Poon, B. K., Prisant, M. G., Read, R. J., Richardson, J. S., Richardson, D. C., Sammito, M. D., Sobolev, O. V., Stockwell, D. H., Terwilliger, T. C., Urzhumtsev, A. G., Videau, L. L., Williams, C. J. & Adams, P. D. (2019). Acta Cryst. D75, 861–877. Web of Science CrossRef IUCr Journals Google Scholar
Liu, Y. & Eisenberg, D. (2002). Protein Sci. 11, 1285–1299. Web of Science CrossRef PubMed CAS Google Scholar
Maier, R. J. & Moshiri, F. (2000). J. Bacteriol. 182, 3854–3857. Web of Science CrossRef PubMed CAS Google Scholar
Mizuno, H., Fujimoto, Z., Koizumi, M., Kano, H., Atoda, H. & Morita, T. (1997). Nat. Struct. Mol. Biol. 4, 438–441. CrossRef CAS Web of Science Google Scholar
Moshiri, F., Crouse, B. R., Johnson, M. K. & Maier, R. J. (1995). Biochemistry, 34, 12973–12982. CrossRef CAS PubMed Web of Science Google Scholar
Moshiri, F., Kim, J. W., Fu, C. & Maier, R. J. (1994). Mol. Microbiol. 14, 101–114. CrossRef CAS PubMed Web of Science Google Scholar
Robson, R. L. (1979). Biochem. J. 181, 569–575. CrossRef CAS PubMed Web of Science Google Scholar
Robson, R. L. & Postgate, J. R. (1980). Annu. Rev. Microbiol. 34, 183–207. CrossRef CAS PubMed Web of Science Google Scholar
Rousseau, F., Schymkowitz, J. & Itzhaki, L. S. (2012). Adv. Exp. Med. Biol. 747, 137–152. CrossRef CAS PubMed Google Scholar
Sainz, G., Jakoncic, J., Sieker, L. C., Stojanoff, V., Sanishvili, N., Asso, M., Bertrand, P., Armengaud, J. & Jouanneau, Y. (2006). J. Biol. Inorg. Chem. 11, 235–246. Web of Science CrossRef PubMed CAS Google Scholar
Schlesier, J., Rohde, M., Gerhardt, S. & Einsle, O. (2016). J. Am. Chem. Soc. 138, 239–247. Web of Science CrossRef CAS PubMed Google Scholar
Shethna, Y. I., DerVartanian, D. V. & Beinert, H. (1968). Biochem. Biophys. Res. Commun. 31, 862–868. CrossRef CAS PubMed Web of Science Google Scholar
Terwilliger, T. C., Adams, P. D., Read, R. J., McCoy, A. J., Moriarty, N. W., Grosse-Kunstleve, R. W., Afonine, P. V., Zwart, P. H. & Hung, L.-W. (2009). Acta Cryst. D65, 582–601. Web of Science CrossRef CAS IUCr Journals Google Scholar
Ureta, A. & Nordlund, S. (2002). J. Bacteriol. 184, 5805–5809. Web of Science CrossRef PubMed CAS Google Scholar
Weis, W. I., Kahn, R., Fourme, R., Drickamer, K. & Hendrickson, W. A. (1991). Science, 254, 1608–1615. CrossRef PubMed CAS Web of Science Google Scholar
Williams, C. J., Headd, J. J., Moriarty, N. W., Prisant, M. G., Videau, L. L., Deis, L. N., Verma, V., Keedy, D. A., Hintze, B. J., Chen, V. B., Jain, S., Lewis, S. M., Arendall, W. B., Snoeyink, J., Adams, P. D., Lovell, S. C., Richardson, J. S. & Richardson, D. C. (2018). Protein Sci. 27, 293–315. Web of Science CrossRef CAS PubMed Google Scholar
Winter, G. (2010). J. Appl. Cryst. 43, 186–190. Web of Science CrossRef CAS IUCr Journals Google Scholar
Wodak, S. J., Malevanets, A. & MacKinnon, S. S. (2015). Biophys. J. 109, 1087–1100. Web of Science CrossRef CAS PubMed Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.