journal menu
short communications
 | JOURNAL OF APPLIED CRYSTALLOGRAPHY |
accessDiffraction anisotropy and paired of H33, a protein binder to interleukin 10
aCzech Technical University in Prague, Brehova 7, Prague 115 19, Czech Republic, and bInstitute of Biotechnology of the Czech Academy of Sciences, Biocev, Průmyslová 595, Vestec 25250, Czech Republic
*Correspondence e-mail: petr.kolenko@fjfi.cvut.cz
Binder H33 is a small protein binder engineered by ribosome display to bind human interleukin 10. Crystals of binder H33 display severe diffraction anisotropy. A set of data files with correction for diffraction anisotropy based on different local signal-to-noise ratios was prepared. Paired was used to find the optimal anisotropic high-resolution diffraction limit of the data: 3.13–2.47 Å. The structure of binder H33 belongs to the 2% of crystal structures with the highest solvent content in the Protein Data Bank.
Keywords: anisotropy; paired refinement; binder H33.
PDB reference: 8bdu
1. Introduction
The diffraction quality of a crystal is usually different in various directions. Diffraction anisotropy can be caused by crystal growth, the crystal shape, the modulated volume of the irradiated crystal during the measurement and the arrangement of molecules inside the crystal. This phenomenon is often not a serious issue for a successful Most of the macromolecular programs are able to work with weak diffraction anisotropy. But severe diffraction anisotropy may represent a serious threat. The difficulties may appear in the process of phasing and/or structure However, several computational tools have been developed to analyse or even account for diffraction anisotropy, e.g. AIMLESS (Evans & Murshudov, 2013![[Evans, P. R. & Murshudov, G. N. (2013). Acta Cryst. D69, 1204-1214.]](../../../../../../logos/arrows/j_arr.gif "Evans, P. R. & Murshudov, G. N. (2013). Acta Cryst. D69, 1204-1214."){kind=small}
), STARANISO (Tickle et al., 2018) and Diffraction Anisotropy Server (Strong et al., 2006). These methods perform anisotropic cut-off of the data together with rescaling of intensities or structure factors with scales depending on the analysis and model of anisotropy employed by each program. These modifications are beneficial for a large number of crystal structures and are reported in Section 2 (Rupp, 2018).
Paired is a modern method for determining the high-resolution cut-off of diffraction data (Karplus & Diederichs, 2012). For this method, the reference data are selected on the basis of a conservative cut-off [e.g. 〈I/σ(I)〉 < 2]. More and more reflections are added to the model in a stepwise manner, and their positive or negative contribution is evaluated on a number of criteria, mainly Rfree calculated on the reference data. Recently, the method has been implemented in the program PAIREF (Malý et al., 2020, 2021). However, the current protocol implemented in PAIREF does not consider the anisotropic diffraction qualities of the crystals. Both the reference data and the evaluated reflections are in the form of spherical shells.
We investigated the possibility of combining corrections for diffraction anisotropy with the standard paired approach. The crystals of our target protein H33 showed serious anisotropy in the diffraction qualities. H33 is an artificial protein binder that was selected during a directed evolutionary study (Pham et al., 2021); it is a variant of the protein scaffold derived from the N-terminal domain of the PIH1D1 domain of the R2TP cochaperone complex (PDB entry 4psf; Hořejší et al., 2014). The scaffold was trained using the ribosome display technique to bind human interleukin-10 (IL-10), a cytokine of human innate immunity (El Kasmi et al., 2007). Blocking or potentiating IL-10 signalization by artificially evolved non-antibody binders such as H33 could be an important component of the treatment of inflammatory, malignant and autoimmune diseases in which IL-10 plays a role. However, our understanding of the structural aspects of the binding between the binders and IL-10 is quite limited. So far, we have solved the structure of only one IL-10 binder called J61 (PDB entry 7avc; Pham et al., 2021). Therefore, a newly solved structure of H33 will aid in the design of new potent and selective binders.
In our work, we introduced an approach to perform paired using the anisotropic data. Anisotropic scaling proved to have a positive impact on the quality of the observed electron density.
2. Materials and methods
2.1. Protein production and crystallization
Protein production, purification and basic characterization have been described previously (Pham et al., 2021). Briefly, the synthesized DNA strings were cloned into the pET-26b(+) vector. The plasmid was used for transformation into the Escherichia coli strain BL21(DE3). The bacteria were grown in LB medium, and protein expression was induced by the addition of isopropyl-beta-D-thiogalactopyranoside. After cell disruption, the soluble fraction was separated by centrifugation and the protein was purified from the cell lysate by using Strep-Tactin XT resin. The last purification step was performed using (Superdex 75 16/600 column).
The crystals were prepared using the hanging-drop vapour-diffusion method from a protein solution that contained 20 mM Tris, 100 mM NaCl pH 8.0 and the protein at a concentration of 10 mg ml−1. The protein crystallized in a wide range of crystallization conditions. However, the crystals diffracted poorly. The final crystallization conditions were 1 M (NH4)2SO4, 1%(w/v) PEG 3350, 0.1 M bis-Tris pH 5.5. Cryoprotection with 20%(v/v) glycerol was necessary before flash-freezing in liquid nitrogen.
2.2. Diffraction data collection and processing
The synchrotron data were collected on beamline P13 (Cianci et al., 2017) operated by EMBL Hamburg at the PETRA III storage ring (DESY, Hamburg, Germany). The diffraction images were processed with XDS (Kabsch, 2010) up to 2.3 Å resolution. The data quality metrics [decrease in 〈I/σ(I)〉, decrease in CC1/2] indicate radiation damage that started immediately after 180° of total oscillation and progressed to the end of the measurement. Therefore, only half of the images (3600) were used for further data evaluation. Such data treatment should remove the possible impact of absorbed dose on the resulting diffraction anisotropy. Initial scaling of the data was performed using AIMLESS (Evans & Murshudov, 2013). The data were severely anisotropic according to a number of indicators. For example, the estimates of the diffraction limits reported by AIMLESS [based on criterion for 〈I/σ(I)〉 > 1.5 in the highest-resolution shell] were 3.28 and 2.65 Å along the hk plane and the l axis, respectively.
Due to severe diffraction anisotropy, the unmerged scaled data from XDS (XDS_ASCII.HKL file) were merged and corrected for anisotropy using the STARANISO server (Tickle et al., 2018) with four different local spherical 〈Imean/σ(Imean)〉 cut-offs going down from 1.2 (STARANISO default value, A1.2 data) to 1.0 (A1.0 data), 0.75 (A0.75 data) and the lowest available value 0.5 (A0.5 data). The free flags were generated with the program FREERFLAG (Brünger, 1992). Initially, free flags were generated for the A1.2 data. The free flags for the A1.0 data were generated with the option to copy already existing flags for reflections in the A1.2 data. A similar approach was used for the generation of free flags for the A0.75 and A0.5 data. This approach was necessary to maintain the pairwise consistency of the free flags within the different data. Data quality indicators are shown in Table 1.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The was solved by using PHASER (McCoy et al., 2007) employing the J61 variant of the protein binder from the same directed evolutionary study (PDB entry 7avc; Pham et al., 2021) as a search model. Data with the highest 〈Imean/σ(Imean)〉 cutoff (A1.2) were used. Two molecules were found in the Due to the low resolution of the data and the unstable (unacceptable number of Ramachandran outliers and bad bond angles) in REFMAC5 (Murshudov et al., 2011), the structure was restrained to the original scaffold of the PDB entry 4psf refined at 1.58 Å (Hořejší et al., 2014) with PROSMART (Nicholls et al., 2012). The structure was refined with isotropic atomic displacement parameters (ADPs) and no TLS domains defined. Manual corrections to the model were performed with Coot (Emsley et al., 2010).
For the manually launched paired the data with the highest cut-off (A1.2) were initially chosen. of the structure model restrained to the structure of PDB entry 4psf with REFMAC5 was used. To keep the same scheme as used previously, three cycles were performed in each paired step. We also performed several manual paired refinements with ten cycles of Although the results differ in exact values, this change did not lead to a different decision on data usage. Several criteria were evaluated during the paired Mainly, drops in overall Rwork and Rfree were monitored. In addition to that, Rfree in the highest-resolution shell did not exceed the value of 0.42 (the theoretically perfect model gives an R value of 0.42 against random data with no and no translational Evans & Murshudov, 2013), and the values of CCwork and CCfree did not exceed the value of CC* (see Table 1). The main results of the paired are shown in Table 2. The decrease in Rwork and Rfree values in all three steps indicates that the addition of the progressively weaker reflections improved the model quality against the same (stronger) data. Therefore, A0.5 data were used in further structure refinements. The exact values of the final Rwork and Rfree in the fifth column of Table 2 cannot be directly compared because they were calculated against different data. Although the differences in the Rwork and Rfree values can be considered marginal, they are comparable to values published in previous studies (Karplus & Diederichs, 2012; Malý et al., 2020, 2021).
| ||||||||||||||||||||||||||||||||||||||
The final model using ten cycles was carried out using all reflections (work and free) of the A0.5 dataset. Jelly body protocol was used to release the previously used and necessary restraints. The quality of the structure stereochemistry was checked using the validation tools in Coot (Emsley et al., 2010), CCP4 (Agirre et al., 2023; Winn et al., 2011), MOLPROBITY (Chen et al., 2010) and the Protein Data Bank (Berman et al., 2003). The quality indicators of the final structure are shown in Table 3. Raw diffraction data are available from https://doi.org/10.5281/zenodo.4033811. The structure coordinates were deposited under PDB entry 8bdu.
†The lowest high-resolution diffraction limit after the anisotropic cut-off. | ||||||||||||||||||||||||||||||||||||
For analysis of the additional value of anisotropic scaling along with paired in terms of data quality and observed electron density, data processed in the standard (isotropic) way were used in paired with a 2.9 Å starting resolution. The complete cross-validation procedure implemented in PAIREF extended the resolution to 2.8 Å.
3. Results and discussion
The artificially generated binder H33 was successfully crystallized and the diffraction data were collected. The crystal diffracted anisotropically, and correction of the intensities for diffraction anisotropy was performed. The was solved and refined using the A1.2 data. The high-resolution diffraction limit was extended using the paired procedure to that of the A0.5 data. The A0.5 data were used for final structure refinement.
The structure of binder H33 is highly similar to that of binder J61 (Pham et al., 2021) from the same study. The root mean square deviation calculated on 128 Cα atoms is lower than 1.2 Å. The structure has an unusually high solvent content of 76% (Kantardjieff & Rupp, 2003). This high solvent content is present in <2% of the crystal structures in the PDB. The solvent content is probably responsible for the low diffraction quality of the crystals.
Previous studies have shown that diffraction anisotropy is not strictly dependent on crystal packing (Robert et al., 2017). The molecules in the crystal of binder H33 are arranged in tubules perpendicular to the z axis of the The tubules have large channels of solvent between them. The planes with the normal vector perpendicular to the z axis are the least occupied with molecules [see Fig. 1(b)]. In contrast, no large channels are present in the planes with normal vectors perpendicular to the x or y axis.
![[Figure 1]](ap5049fig1thm.gif) | Figure 1 (a) Structure of the monomeric binder H33 determined using the A0.5 data in a secondary structure representation. (b)–(c) filled with molecules viewed from different perspectives. (d)–(f) Residue Leu75 (chain A) with the 2mFo − DFc electron density (blue mesh) at the level of 0.15 e Å−3 calculated for the model before the paired procedure using data A1.2, the output model from the paired procedure using the A0.5 data and the model from the isotropic paired procedure using the Iso data, respectively. (g)–(i) Residue Leu75 (chain A) with the 2mFo − DFc electron density at the 1σ level for the same combination of model versus data as in the previous triplicate. Graphics were generated with CCP4MG (McNicholas et al., 2011 ). |
Using the data range according to paired may result in an improvement of the observed electron density (Karplus & Diederichs, 2015; Malý et al., 2020). Correction for the anisotropy in the diffraction qualities was also shown to improve the observed electron density (Tickle et al., 2018). Paired using shells reflecting the diffraction anisotropy is not automated in any pipeline. The available software, for example PAIREF (Malý et al., 2020) and PDB-REDO (Joosten et al., 2014), use the addition of reflections in spherical shells by increasing the spherical high-resolution diffraction limit. Here, we propose the addition of reflections with the same expected information content in non-spherical shells.
The quality of the electron density depends on the diffraction data and the structural model. In our analysis, we compared electron density maps of (i) the starting model for the paired refined using the A1.2 data against the A1.2 data, (ii) the resulting model from the paired refined using the A0.5 data against the A0.5 data and (iii) the optimal model from the paired using the Iso data at 2.8 Å resolution. The electron density maps were calculated with fast Fourier transformation using the same grid spacing to avoid possible bias (Urzhumtsev et al., 2014). The from the Iso data is the least detailed [see Fig. 1(f)]. Corrections for diffraction anisotropy using the STARANISO server (Tickle et al., 2018) and using the A1.2 data dramatically improved the quality of the observed electron density maps. No differences were observed with the extension of data from A1.2 to A0.5.
The number of reflections in the A0.5 data is approximately equal to that of the Iso data. Apparently, their information content is different. The anisotropic cut-off in the A0.5 data removed a significant portion of noisy reflections. The Iso dataset contains reflections in the weak directions with a low signal-to-noise ratio. Moreover, it does not contain a portion of the strong reflections in the strong directions that are present in the A0.5 data at a resolution higher than 2.8 Å.
In our case, both approaches to data optimization (correction for diffraction anisotropy and paired refinement) have proved useful. Although improvement in observed electron density did not occur after paired of data corrected for diffraction anisotropy, 2676 unique reflections (14.5% from 18 374 reflections in total) were added to the scheme using the A0.5 data. This addition was validated by the decrease in R values (see Table 2).
The current trend in data quality evaluation (paired refinement) is to investigate the `additional value' of more and more observations involved in model Conventional indicators of the quality of the diffraction data are no longer relevant for the estimation of the high-resolution diffraction limit. The diffraction anisotropy makes the problem even more difficult. Our was determined at a nominal diffraction limit of 2.47 Å. However, closer inspection of the diffraction data statistics shows that the diffraction data become dramatically incomplete at better than 3.13 Å resolution. The highest-resolution shell of reflections has a spherical completeness lower than 15%. This indicator must be considered when comparing structures refined `with the same diffraction limits'.
Acknowledgements
We thank Isabel Bento for her assistance in using the beamline.
Funding information
This work was supported by the Ministry of Education, Youth and Sports CR, project CAAS (grant No. CZ.02.1.01/0.0/0.0/16_019/0000778 awarded to the Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague); the institutional grant to the Institute of Biotechnology of the Czech Academy of Sciences (RVO 86652036). We acknowledge CMS-Biocev (Crystallization and diffraction techniques) supported by MEYS CR (grant No. LM2018127), the Grant Agency of the Czech Technical University in Prague (grant No. SGS22/114/OHK4/2T/14).
References
Agirre, J., Atanasova, M., Bagdonas, H., Ballard, C. B., Baslé, A., Beilsten-Edmands, J., Borges, R. J., Brown, D. G., Burgos-Mármol, J. J., Berrisford, J. M., Bond, P. S., Caballero, I., Catapano, L., Chojnowski, G., Cook, A. G., Cowtan, K. D., Croll, T. I., Debreczeni, J. É., Devenish, N. E., Dodson, E. J., Drevon, T. R., Emsley, P., Evans, G., Evans, P. R., Fando, M., Foadi, J., Fuentes-Montero, L., Garman, E. F., Gerstel, M., Gildea, R. J., Hatti, K., Hekkelman, M. L., Heuser, P., Hoh, S. W., Hough, M. A., Jenkins, H. T., Jiménez, E., Joosten, R. P., Keegan, R. M., Keep, N., Krissinel, E. B., Kolenko, P., Kovalevskiy, O., Lamzin, V. S., Lawson, D. M., Lebedev, A. A., Leslie, A. G. W., Lohkamp, B., Long, F., Malý, M., McCoy, A. J., McNicholas, S. J., Medina, A., Millán, C., Murray, J. W., Murshudov, G. N., Nicholls, R. A., Noble, M. E. M., Oeffner, R., Pannu, N. S., Parkhurst, J. M., Pearce, N., Pereira, J., Perrakis, A., Powell, H. R., Read, R. J., Rigden, D. J., Rochira, W., Sammito, M., Sánchez Rodríguez, F., Sheldrick, G. M., Shelley, K. L., Simkovic, F., Simpkin, A. J., Skubak, P., Sobolev, E., Steiner, R. A., Stevenson, K., Tews, I., Thomas, J. M. H., Thorn, A., Valls, J. T., Uski, V., Usón, I., Vagin, A., Velankar, S., Vollmar, M., Walden, H., Waterman, D., Wilson, K. S., Winn, M. D., Winter, G., Wojdyr, M. & Yamashita, K. (2023). Acta Cryst. D79, 449–461. Google Scholar
Berman, H., Henrick, K. & Nakamura, H. (2003). Nat. Struct. Mol. Biol. 10, 980. Web of Science CrossRef Google Scholar
Brünger, A. T. (1992). Nature, 355, 472–475. PubMed Web of Science Google Scholar
Chen, V. B., Arendall, W. B., Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S. & Richardson, D. C. (2010). Acta Cryst. D66, 12–21. Web of Science CrossRef CAS IUCr Journals Google Scholar
Cianci, M., Bourenkov, G., Pompidor, G., Karpics, I., Kallio, J., Bento, I., Roessle, M., Cipriani, F., Fiedler, S. & Schneider, T. R. (2017). J. Synchrotron Rad. 24, 323–332. Web of Science CrossRef CAS IUCr Journals Google Scholar
El Kasmi, K. C., Smith, A. M., Williams, L., Neale, G., Panopolous, A., Watowich, S. S., Häcker, H., Foxwell, B. M. & Murray, P. J. (2007). J. Immunol. 179, 7215–7219. 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
Evans, P. R. & Murshudov, G. N. (2013). Acta Cryst. D69, 1204–1214. Web of Science CrossRef CAS IUCr Journals Google Scholar
Hořejší, Z., Stach, L., Flower, T. G., Joshi, D., Flynn, H., Skehel, J. M., O'Reilly, N. J., Ogrodowicz, R. W., Smerdon, S. J. & Boulton, S. J. (2014). Cell. Rep. 7, 19–26. Google Scholar
Joosten, R. P., Long, F., Murshudov, G. N. & Perrakis, A. (2014). IUCrJ, 1, 213–220. Web of Science CrossRef CAS PubMed IUCr Journals Google Scholar
Kabsch, W. (2010). Acta Cryst. D66, 125–132. Web of Science CrossRef CAS IUCr Journals Google Scholar
Kantardjieff, K. A. & Rupp, B. (2003). Protein Sci. 12, 1865–1871. Web of Science CrossRef PubMed CAS Google Scholar
Karplus, P. A. & Diederichs, K. (2012). Science, 336, 1030–1033. Web of Science CrossRef CAS PubMed Google Scholar
Karplus, P. A. & Diederichs, K. (2015). Curr. Opin. Struct. Biol. 34, 60–68. Web of Science CrossRef CAS PubMed Google Scholar
Malý, M., Diederichs, K., Dohnálek, J. & Kolenko, P. (2020). IUCrJ, 7, 681–692. Web of Science CrossRef PubMed IUCr Journals Google Scholar
Malý, M., Diederichs, K., Dohnálek, J. & Kolenko, P. (2021). Acta Cryst. F77, 226–229. Web of Science CrossRef IUCr Journals Google Scholar
McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C. & Read, R. J. (2007). J. Appl. Cryst. 40, 658–674. Web of Science CrossRef CAS IUCr Journals Google Scholar
McNicholas, S., Potterton, E., Wilson, K. S. & Noble, M. E. M. (2011). Acta Cryst. D67, 386–394. Web of Science CrossRef CAS IUCr Journals Google Scholar
Murshudov, G. N., Skubák, P., Lebedev, A. A., Pannu, N. S., Steiner, R. A., Nicholls, R. A., Winn, M. D., Long, F. & Vagin, A. A. (2011). Acta Cryst. D67, 355–367. Web of Science CrossRef CAS IUCr Journals Google Scholar
Nicholls, R. A., Long, F. & Murshudov, G. N. (2012). Acta Cryst. D68, 404–417. Web of Science CrossRef CAS IUCr Journals Google Scholar
Pham, P. N., Huličiak, M., Biedermannová, L., Černý, J., Charnavets, T., Fuertes, G., Herynek, Š., Kolářová, L., Kolenko, P., Pavlíček, J., Zahradník, J., Mikulecký, P. & Schneider, B. (2021). Viruses, 13, 190. Google Scholar
Robert, X., Kassis-Sahyoun, J., Ceres, N., Martin, J., Sawaya, M. R., Read, R. J., Gouet, P., Falson, P. & Chaptal, V. (2017). Sci. Rep. 7, e17013. Google Scholar
Rupp, B. (2018). Structure, 26, 919–923. Web of Science CrossRef CAS PubMed Google Scholar
Strong, M., Sawaya, M. R., Wang, S., Phillips, M., Cascio, D. & Eisenberg, D. (2006). Proc. Natl Acad. Sci. USA, 103, 8060–8065. Web of Science CrossRef PubMed CAS Google Scholar
Tickle, I. J., Flensburg, C., Keller, P., Paciorek, W., Sharff, A., Vonrhein, C. & Bricogne, G. (2018). STARANISO. Global Phasing Ltd, Cambridge, UK. https://staraniso.globalphasing.org/cgi-bin/staraniso.cgi. Google Scholar
Urzhumtsev, A., Afonine, P. V., Lunin, V. Y., Terwilliger, T. C. & Adams, P. D. (2014). Acta Cryst. D70, 2593–2606. Web of Science CrossRef IUCr Journals Google Scholar
Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G. W., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A. & Wilson, K. S. (2011). Acta Cryst. D67, 235–242. Web of Science CrossRef CAS IUCr Journals 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.
| JOURNAL OF APPLIED CRYSTALLOGRAPHY |
access
