research letters
Small-angle neutron scattering studies suggest the mechanism of BinAB protein internalization
aProtein Crystallography Section, Radiation Biology & Health Sciences Division, Bhabha Atomic Research Centre, Mumbai 400 085, India, bSolid State Physics Division, Bhabha Atomic Research Centre, Mumbai 400 085, India, and cHomi Bhabha Professor, Bhabha Atomic Research Centre, Mumbai 400 085, India
*Correspondence e-mail: vinay@barc.gov.in, rc@barc.gov.in
Small-angle neutron scattering (SANS) is one of the most widely used neutron-based approaches to study the solution structure of biological macromolecular systems. The selective deuterium labelling of different protein components of a complex provides a means to probe conformational changes in multiprotein complexes. The Lysinibacillus sphaericus mosquito-larvicidal BinAB proteins exert toxicity through interaction with the receptor Cqm1 protein; however, the nature of the complex is not known. Rationally engineered deuterated BinB (dBinB) protein from the L. sphaericus ISPC-8 species was synthesized using an Escherichia coli-based protein-expression system in M9 medium in D2O for `contrast-matched' SANS experiments. SANS data were independently analysed by ab initio indirect Fourier transform-based modelling and using crystal structures. These studies confirm the dimeric status of Cqm1 in 100% D2O with a longest intramolecular vector (Dmax) of ∼94 Å and a (Rg) of ∼31 Å. Notably, BinB binds to Cqm1, forming a heterodimeric complex (Dmax of ∼129 Å and Rg of ∼40 Å) and alters its oligomeric status from a dimer to a monomer, as confirmed by matched-out Cqm1–dBinB (Dmax of ∼70 Å and Rg of ∼22 Å). The present study thus provides the first insight into the events involved in the internalization of larvicidal proteins, likely by raft-dependent endocytosis.
Keywords: SANS; Cqm1 dimer; protein deuteration; Cqm1–BinB complex; contrast matching; BinAB internalization; deuterated BinB.
1. Introduction
Small-angle neutron scattering (SANS) is one of the most widely used neutron-based approaches for solution structure studies of biomacromolecular complexes where crystallization proves to be intractable. Selective deuterium labelling of the component proteins combined with contrast variation can be very useful to distinguish between and model different regions of multi-protein complexes (Neylon, 2008; Dunne et al., 2017). Protein deuteration can be achieved by three different approaches, including deuteration through expression of the protein in D2O-based medium but using an unlabelled (hydrogenous) carbon source, which may yield 60–70% D incorporation with a scattering length density (SLD) close to ∼99% D2O (Koruza et al., 2018; Dunne et al., 2017).
Binary (BinAB) toxin is responsible for the mosquito-larvicidal properties of Lysinibacillus sphaericus. The two component proteins, BinA (the toxic component) and BinB (the receptor-binding component), work synergistically and exert toxicity through interaction with the receptor, Cqm1, in Culex quinquefasciatus (Darboux et al., 2001). Bound apically to the epithelial membrane of the larval midgut cells via a glycosylphosphatidylinositol (GPI) anchor, Cqm1 mediates toxin internalization. Presentation of the receptor on the cell membrane and interaction of the toxin component proteins with the receptor are essential for the larvicidal cytotoxicity. The intracellular toxicity of the BinA component has been proposed to be associated with its glycan affinity (Sharma et al., 2018a). However, the mode of toxin internalization has long been debated and remains unclear (Oei et al., 1992; Opota et al., 2011; Lekakarn et al., 2015). The crystal structures of the Cqm1 and BinAB proteins are known (Sharma & Kumar, 2019; Colletier et al., 2016); however, the structure of the biological complex is not known. Understanding the details of their interaction may provide clues to the toxin internalization mechanism.
Here, a SANS study of Cqm1–BinB interaction using hydrogenous BinB (hBinB) and deuterated BinB (dBinB) is reported. High-yield expression of dBinB protein using an unlabelled carbon source was achieved and the homogenously purified dBinB protein was characterized for proper folding, thermal stability and its interaction with the receptor protein. The Cqm1 protein was investigated in its solution form and in complex with BinB. The SANS study shows that Cqm1 exists as a dimer in solution and undergoes a change in oligomeric status from a dimer to a monomer upon interaction with BinB. The present study also provides low-resolution details of the complex structure and clues to the mechanism of toxin internalization.
2. Materials and methods
2.1. Materials
Escherichia coli XL-10 Gold and BL21 Star (DE3) bacterial strains were obtained from Stratagene and Novagen, respectively. Ni–IDA matrix was obtained from GE Healthcare. Isopropyl β-D-1-thiogalactopyranoside (IPTG), kanamycin, phenylmethylsulfonyl fluoride (PMSF) and SYPRO Orange dye were obtained from Sigma. All other chemicals were of analytical grade.
and Phusion polymerase were obtained from New England Biolabs. The expression vector pET-28a(+) was obtained from Novagen. LB broth and LB agar were procured from HiMedia Laboratories.2.2. Rational engineering of BinB and purification of BinB and Cqm1
The yield of full-length BinB protein from the recombinant plasmid carrying the binB gene from L. sphaericus ISPC-8 was observed to be low. The use of different solubility tags or expression vectors did not improve the yield of soluble and properly folded protein. Overlapping the L. sphaericus ISPC-8 BinB protein sequence (GenBank accession ID EU375309.1) with other BinB sequences available in GenBank revealed that the L. sphaericus ISPC-8 sequence differs from the other sequences primarily at positions 109 and 274 [His109 and Pro274; Supplementary Fig. S1(a)]. These residues occupy surface positions, as revealed from the available BinB [PDB entry 5foy; Colletier et al., 2016; Supplementary Fig. S1(b)]. Hence, we restored the invariant Pro109 and Ser274 residues by the overlap-extension method (Section S1.1, supporting information). DNA sequencing confirmed the substitutions.
The `rationally' engineered pET-28a-binB construct was transformed into E. coli BL21 Star (DE3) cells for protein expression. BinB protein (with an N-terminal 6×His tag of 18 residues) was purified using immobilized metal ion-affinity (IMAC) using Ni–IDA matrix (Section S1.2, supporting information).
The receptor Cqm1 (functional form; residues 23–560), lacking the 22 N-terminal signal peptide residues and 20 C-terminal GPI anchor residues, was expressed in a soluble form using E. coli BL21 Star (DE3) cells with an N-terminal polyhistidine tag of 22 residues and was purified to using Ni–IDA as described previously (Sharma et al., 2018b).
Partially deuterated forms of the Cqm1 and BinB proteins were obtained by buffer exchange [using ultracentrifugal devices; Amicon Ultra, molecular-weight cutoff 10 kDa) of the hydrogenous Cqm1 and BinB proteins into D2O-based buffer C (25 mM HEPES pH 7.5, 25 mM NaCl in 100% D2O). The concentrations of the purified BinB and Cqm1 proteins in buffer C were estimated from their absorbance at 280 nm using extinction coefficients (57 105 and 121 700 M−1 cm−1, respectively) estimated from the amino-acid sequences by the ProtParam tool (https://web.expasy.org/protparam/; Gasteiger et al., 2005).
2.3. Expression and purification of deuterated BinB
To express dBinB protein, E. coli BL21 Star (DE3) cells were adapted from H2O-based to D2O-based M9+ medium (M9 medium supplemented with a high percentage of hydrogenated glucose as the carbon source; Supplementary Table S1) using a three-step approach (Cai et al., 2016; Supplementary Fig. S2; Section S2, supporting information). The third-stage 25 ml pre-culture was inoculated into 250 ml M9+/D2O medium and allowed to grow at 37°C to an OD600 of 0.7. Protein expression was induced with 1 mM IPTG at 20°C. The cells were grown further for an extended period of 48 h at 20°C before harvesting. This protocol ensured a high yield of deuterated protein even using conventional laboratory cell-culture equipment. The concentration of the purified dBinB protein (in buffer C), purified using Ni–IDA matrix, was estimated from the absorbance at 280 nm using an extinction coefficient (57 105 M−1 cm−1) estimated from the amino-acid sequence.
2.4. Biochemical and biophysical characterization of deuterated BinB
BinB and dBinB proteins (each at 0.5 mg ml−1) were monitored to assess their folding state by intrinsic (tryptophan) fluorescence and from estimation of the melting temperature (Tm). Fluorescence spectra were recorded at 25°C on a JASCO spectrofluorometer (FP-8500) over the wavelength range 290–400 nm after excitation at 280 nm. Each spectrum was obtained by averaging three individual scans. For thermal stability analysis using a Thermofluor shift assay, a protein sample (2 µM) was mixed with 5× SYPRO Orange dye and loaded into 96-well plates (Bio-Rad) sealed with Optical Quality Sealing Tape (Bio-Rad). The plate was heated from 20 to 90°C at a ramp rate of 1.0°C min−1 and data were recorded on a CFX96 Real-Time PCR Detection System (Bio-Rad) in FRET mode. Each experiment was performed in triplicate. The melting curves were analysed using the CFX Manager software (Bio-Rad) and Tm was determined from the first derivative of the melting curve.
2.5. Native PAGE and DLS analysis of proteins
Cqm1 and BinB proteins were mixed in a 1:2 molar ratio and maintained at 25°C for 1 h. The formation of the stable complex was adjudged on a 10% native PAGE gel and by dynamic −1; Cqm1, BinB and a mixture of the Cqm1 and BinB proteins) and data analysis was performed using the Zetasizer software v.7.01 (Malvern Instruments).
(DLS) performed on a Zetasizer Nano ZS instrument (Malvern Instruments). For DLS, three sets of measurements were recorded at 25°C for each of the proteins (0.5 mg ml2.6. SANS data collection and analysis
SANS measurements were carried out at the SANS-I facility at the Dhruva reactor, Bhabha Atomic Research Centre, Mumbai, India (Aswal & Goyal, 2000). Neutrons with a wavelength (λ) of 5.2 Å and a wavelength spread Δλ/λ of 15% were selected using the neutron velocity selector. Scattered neutrons were detected using a 1 m long 3He position-sensitive detector. The data were collected in a q range (q = 4πsinθ/λ, where 2θ is the scattering angle) from 0.015 to 0.26 Å−1. Samples were held in quartz cells of 5 mm thickness and the temperature was maintained at 25°C. All data were corrected for solvent and background, and normalized to cross-sectional units using a standard procedure. Owing to high at high q, the data sets were truncated for q > 0.2 Å−1. The concentrations of the different proteins used in SANS data collection were 3.3 mg ml−1 for Cqm1, 3.3 mg ml−1 for hBinB, 5 mg ml−1 for the Cqm1–BinB complexes and 6 mg ml−1 for dBinB. The reduced SANS intensity was normalized to a protein concentration of 1 mg ml−1 to estimate I(0) (the intensity at q = 0).
The SANS data were analysed and modelled using software tools available in ATSAS 2.8 (Franke et al., 2017) following two independent approaches: ab initio modelling by the indirect Fourier transformation (IFT) method using DATGNOM (Petoukhov et al., 2007) and using available structural information (theoretical) by the CRYSON module of ATSAS 2.8 (Svergun et al., 1995). The (Rg) and absolute intensity I(0) (at q = 0) values were estimated from the pair-distance distribution [P(r)]. The molecular weights (MW) of the proteins were estimated by MW = I(0) × (NAdp2/Δρ2), where NA is Avogadro's number, the average protein density (dp) is 1.36 g cm−3 and the average excess scattering length density of proteins, Δρ, is 2.8 × 1010 cm−2. Ab initio modelling of the shape was achieved with the online DAMMIN module (Svergun, 1999) using P(r) values and the automatic algorithm available on the ATSAS online server (https://www.embl-hamburg.de/biosaxs/atsas-online/; Franke et al., 2017). The experimental and theoretical Kratky plots were also calculated using Origin and are given in Supplementary Fig. S6.
The agreement between observed scattering and transformed data were assessed using the reduced χ2 values defined as
where the summation is over all observed experimental data points i, Iifit and IiSANS are the fitted and observed intensities for the ith observation, σi is the experimental error and N is the number of data points in the analysed q range.
The atomic coordinates of Cqm1 (residues 7–537) and BinB (residues 28–446) monomers were extracted from Protein Data Bank (PDB) entries 6k5p (Sharma & Kumar, 2019) and 5foy (Colletier et al., 2016), respectively. The template-based CA-CA-guided docking method was used to model the Cqm1–BinB complex structure using the HADDOCK web server (Xue et al., 2017) and the interfacial residues of Cqm1 (Ser109, Gly139-Gly140 and Ala292; Ferreira et al., 2014) and of the BinB protein (Phe41-Tyr42-Asn43; Singkhamanan et al., 2013).
The atomic structures were fitted into ab initio shape models generated with DAMMIN using the SUPALM and SUPCOMB modules of ATSAS.
3. Results and discussion
Mosquito-larvicidal binary toxin (BinAB) is highly active against Culex and Anopheles mosquitoes but is refractory to Aedes aegypti. The incompetence of the BinAB toxin against Aedes may be due to the inability of the toxin to be internalized across the cell membrane (Lekakarn et al., 2015). Methods are required to understand the interaction of the BinAB components with the receptor protein in order to understand the mechanism underlying toxin internalization. BinB interacts with Cqm1 in solution with high affinity (Kd of ∼10 nM; Sharma et al., 2018b). In this study, we investigated the oligomeric state of Cqm1 and its complex with the BinB protein using hydrogenated and deuterated BinB by `contrast-matched' SANS. All SANS experiments were carried out in 100% D2O. The SANS data were analysed as suggested in the recommendations of the Small-Angle Scattering Validation Task Force (Trewhella et al., 2017).
3.1. Characteristics of deuterated BinB
A rationally engineered pET-28a(+)-binB construct was used for expression of the hBinB and dBinB proteins. The three-step approach to adapt E. coli cells from H2O-based to D2O-based culture medium proved to be successful for large-scale purification of the dBinB protein (Supplementary Fig. S2). A high yield of dBinB protein (15 mg dBinB compared with 30 mg hBinB) could be achieved with a cell density of ∼2 (OD600) from 1 l E. coli culture medium (Supplementary Fig. S3). Notably, an adaptive protocol and a longer induction time (48 h) seem to be critical factors for a higher protein yield. The success of protein deuteration is reflected by its scattering length density (ρp) reaching close to ρs [that of D2O; Fig. 1(b)]. The extent of deuteration for dBinB was estimated from contrast [(ρp − ρs)2] values to be ∼77% (Supplementary Table S2), compared with Cqm1 and BinB deuterated partially through buffer exchange (20% and 16%, respectively). Purified dBinB exhibits proper folding with a λemmax of 327 nm [Supplementary Fig. S4(a)] and its displays a thermal stability similar to that of the hBinB protein, with a Tm value of ∼80°C [Fig. 1(a) and Supplementary Fig. S4(b)].
3.2. SANS modelling
The SANS curves decrease monotonically over the scattering-vector range 0.016–0.2 Å−1. Two independent approaches were employed to achieve fitting to the experimental SANS data: IFT-based ab initio modelling and fitting theoretical scattering curves calculated directly from the atomic structures available in the PDB or from docking solutions.
The scattering curves obtained by IFT and from a dimeric structure of Cqm1 (Supplementary Fig. S7) match the experimental SANS data, with χ2 values close to 0.3 [Fig. 2(a)]. The Rg (∼31 Å) and Dmax (∼94 Å) obtained by the IFT method with a smooth P(r) function closely match the theoretical estimates obtained from the dimer structure [Table 1; Fig. 3(a)]. The molecular weight estimated from the absolute I(0) (∼120 kDa) is within 10% of the value deduced from the amino-acid sequences (Table 1) and a dimer of Cqm1 fits well into the ab initio shape model generated with DAMMIN [Fig. 3(a)]. In contrast, the theoretical curve for a Cqm1 monomer does not match the experimental SANS data (χ2 = 3.6) [Fig. 2(a)]. The dimeric status of Cqm1 in solution also matches the radius of hydration (Rh = ∼42 Å) observed in dynamic light-scattering experiments [Fig. 1(c)] and the elution profile of the protein from the column (Sharma et al.,, 2018b).
|
Likewise, the BinB monomer matches the SANS and DLS experimental data [Figs. 1(b) and 1(c), Table 1]. The three oscillations observed in the P(r) function and the extra bead density observed in the ab initio dummy DAMMIN model [Fig. 3(b)] can be rationalized owing to the presence of a third domain in BinB constituted of residues 1–45, for which atomic coordinates are not available. In comparison to BinB, the deuterated protein did not result in a significant scattering signal in D2O solvent [Fig. 1(b)]. It was thus taken that scattering owing to dBinB was matched out in 100% D2O with 77% deuteration.
Single peaks corresponding to Rh values of about 80 Å in the DLS profiles of Cqm1 mixed with the hBinB and dBinB proteins suggest the formation of stable Cqm1–hBinB and Cqm1–dBinB complexes [Fig. 1(c)], which is also confirmed by the presence of a single protein band, albeit with retarded mobility, on 10% native PAGE (Supplementary Fig. S5). However, analysis of experimental SANS data for the Cqm1–dBinB complex revealed the presence of a Cqm1 monomer in solution [Fig. 2(b)] with the ab initio and theoretical curves fitting very well, with χ2 values of <0.5 [Figs. 2(b) and 3(c), Table 1]. A near `match-out' of dBinB against 100% D2O used in the SANS experiments explains the data.
Fitting of SANS data for the Cqm1–hBinB complex with ab initio IFT modelling indicated Rg and Dmax values of ∼40 and ∼129 Å, respectively [Table 1; Figs. 2(c) and 3(d)]. The molecular weight of the complex was estimated to be ∼121 kDa (compared with the molecular weight of 118 kDa expected from the sequence). These data suggested formation of the Cqm1–BinB complex with a 1:1 stoichiometry. Two shoulder peaks, in addition to the main peak at ∼47 Å, in the pair-distance distribution function may suggest the domain structure of the complex. Attempts to match structural models generated by docking analysis using given interaction constraints provided clues to the placement of BinB in the ab initio shape model. However, the Cqm1 fit did not seem to be good. Attempts were made to obscure the BinB fitted region and fit the Cqm1 monomer, as observed in the `matched-out' Cqm1–dBinB SANS data, to the remainder of the ab initio shape model [Fig. 3(d)]. This results in a low-resolution structural model of the complex, which matches the experimental data with a χ2 value of 0.96.
The −1 (Sharma & Kumar, 2019; Supplementary Fig. S7). Each of the monomers is localized on the lipid rafts via a GPI anchor. It can be speculated that the separation of weakly held Cqm1 monomers in the presence of BinB might disrupt the lipid rafts, resulting in raft-dependent A similar mechanism has been suggested for some viral particles, in which lipid-raft disruption enhanced the release of viral particles with compromized infectivity owing to the leakage of essential viral proteins (Barman & Nayak, 2007).
of the Cqm1 protein suggests a weak dimer with a solvation free-energy gain of ∼4.4 kcal mol4. Conclusion
In the present study, we report the first solution structures of the BinAB toxin receptor Cqm1 and of its complex with the BinB protein. Cqm1 exists as a dimer in both H2O-based and D2O-based buffers. `Contrast-matched' SANS using deuterated BinB and hydrogenous BinB revealed a change in the oligomeric state of the receptor protein from a dimer to a monomer, and provides the first solution model of the Cqm1–BinB complex.
5. SASBDB accession codes
The SANS data have been submitted to the Small Angle Scattering Biological Data Bank (SASBDB; https://www.sasbdb.org; Valentini et al., 2015) with accession codes SASDF87 (receptor Cqm1 protein), SASDF97 (complex of Cqm1 and dBinB proteins) and SASDFA7 (receptor binding BinB protein).
6. Related literature
The following references are cited in the supporting information for this article: Goujon et al. (2010), Pettersen et al. (2004) and Robert & Gouet (2014).
Supporting information
Supplementary Methods, Supplementary Tables and Supplementary Figures. DOI: https://doi.org/10.1107/S2052252519017159/ti5015sup1.pdf
Acknowledgements
We thank Dr Debes Ray for his help with the SANS data collection. We thank the
Board, Department of Atomic Energy, India for providing We thank Dr Anthony Duff of the National Deuteration Facility, ANSTO for helpful discussions.References
Aswal, V. K. & Goyal, P. S. (2000). Curr. Sci. 79, 947–953. Google Scholar
Barman, S. & Nayak, D. P. (2007). J. Virol. 81, 12169–12178. Web of Science CrossRef PubMed CAS Google Scholar
Cai, M., Huang, Y., Yang, R., Craigie, R. & Clore, G. M. (2016). J. Biomol. NMR, 66, 85–91. Web of Science CrossRef CAS PubMed Google Scholar
Colletier, J.-P., Sawaya, M. R., Gingery, M., Rodriguez, J. A., Cascio, D., Brewster, A. S., Michels-Clark, T., Hice, R. H., Coquelle, N., Boutet, S., Williams, G. J., Messerschmidt, M., DePonte, D. P., Sierra, R. G., Laksmono, H., Koglin, J. E., Hunter, M. S., Park, H.-W., Uervirojnangkoorn, M., Bideshi, D. K., Brunger, A. T., Federici, B. A., Sauter, N. K. & Eisenberg, D. S. (2016). Nature, 539, 43–47. Web of Science CrossRef CAS PubMed Google Scholar
Darboux, I., Nielsen-LeRoux, C., Charles, J.-F. & Pauron, D. (2001). Insect Biochem. Mol. Biol. 31, 981–990. Web of Science CrossRef PubMed CAS Google Scholar
Dunne, O., Weidenhaupt, M., Callow, P., Martel, A., Moulin, M., Perkins, S. J., Haertlein, M. & Forsyth, V. T. (2017). Eur. Biophys. J. 46, 425–432. Web of Science CrossRef CAS PubMed Google Scholar
Ferreira, L. M., Romão, T. P., Nascimento, N. A., Costa, M. C. M. F., Rezende, A. M., de-Melo-Neto, O. P. & Silva-Filha, M. H. (2014). Insect Biochem. Mol. Biol. 50, 34–42. Web of Science CrossRef CAS PubMed Google Scholar
Franke, D., Petoukhov, M. V., Konarev, P. V., Panjkovich, A., Tuukkanen, A., Mertens, H. D. T., Kikhney, A. G., Hajizadeh, N. R., Franklin, J. M., Jeffries, C. M. & Svergun, D. I. (2017). J. Appl. Cryst. 50, 1212–1225. Web of Science CrossRef CAS IUCr Journals Google Scholar
Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., Wilkins, M. R., Appel, R. D. & Bairoch, A. (2005). The Proteomics Protocols Handbook, edited by J. M. Walker, pp. 571–607. Totowa: Humana Press. Google Scholar
Goujon, M., McWilliam, H., Li, W., Valentin, F., Squizzato, S., Paern, J. & Lopez, R. (2010). Nucleic Acids Res. 38, W695–W699. Web of Science CrossRef CAS PubMed Google Scholar
Koruza, K., Lafumat, B., Végvári, Á., Knecht, W. & Fisher, S. Z. (2018). Arch. Biochem. Biophys. 645, 26–33. Web of Science CrossRef CAS PubMed Google Scholar
Lekakarn, H., Promdonkoy, B. & Boonserm, P. (2015). J. Invertebr. Pathol. 132, 125–131. Web of Science CrossRef CAS PubMed Google Scholar
Neylon, C. (2008). Eur. Biophys. J. 37, 531–541. Web of Science CrossRef PubMed CAS Google Scholar
Oei, C., Hindley, J. & Berry, C. (1992). J. Gen. Microbiol. 138, 1515–1526. CrossRef PubMed CAS Web of Science Google Scholar
Opota, O., Gauthier, N. C., Doye, A., Berry, C., Gounon, P., Lemichez, E. & Pauron, D. (2011). PLoS One, 6, e14682. Web of Science CrossRef PubMed Google Scholar
Petoukhov, M. V., Konarev, P. V., Kikhney, A. G. & Svergun, D. I. (2007). J. Appl. Cryst. 40, s223–s228. Web of Science CrossRef CAS IUCr Journals Google Scholar
Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C. & Ferrin, J. (2004). J. Comput. Chem. 25, 1605–1612. Web of Science CrossRef PubMed CAS Google Scholar
Robert, X. & Gouet, P. (2014). Nucleic Acids Res. 42, W320–W324. Web of Science CrossRef CAS PubMed Google Scholar
Sharma, M., Gupta, G. D. & Kumar, V. (2018a). J. Invertebr. Pathol. 156, 29–40. Web of Science CrossRef CAS PubMed Google Scholar
Sharma, M., Gupta, G. D. & Kumar, V. (2018b). Insect Biochem. Mol. Biol. 93, 37–46. Web of Science CrossRef CAS PubMed Google Scholar
Sharma, M. & Kumar, V. (2019). Int. J. Biol. Macromol. 140, 1315–1325. Web of Science CrossRef CAS PubMed Google Scholar
Singkhamanan, K., Promdonkoy, B., Srikhirin, T. & Boonserm, P. (2013). J. Invertebr. Pathol. 114, 65–70. Web of Science CrossRef CAS PubMed Google Scholar
Svergun, D., Barberato, C. & Koch, M. H. J. (1995). J. Appl. Cryst. 28, 768–773. CrossRef CAS Web of Science IUCr Journals Google Scholar
Svergun, D. I. (1999). Biophys. J. 76, 2879–2886. Web of Science CrossRef PubMed CAS Google Scholar
Trewhella, J., Duff, A. P., Durand, D., Gabel, F., Guss, J. M., Hendrickson, W. A., Hura, G. L., Jacques, D. A., Kirby, N. M., Kwan, A. H., Pérez, J., Pollack, L., Ryan, T. M., Sali, A., Schneidman-Duhovny, D., Schwede, T., Svergun, D. I., Sugiyama, M., Tainer, J. A., Vachette, P., Westbrook, J. & Whitten, A. E. (2017). Acta Cryst. D73, 710–728. Web of Science CrossRef IUCr Journals Google Scholar
Valentini, E., Kikhney, A. G., Previtali, G., Jeffries, C. M. & Svergun, D. I. (2015). Nucleic Acids Res. 43, D357–D363. Web of Science CrossRef CAS PubMed Google Scholar
Xue, L. C., Rodrigues, J. P. G. L. M., Dobbs, D., Honavar, V. & Bonvin, A. M. J. J. (2017). Brief. Bioinform. 18, 458–466. Web of Science 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.