research communications
1.25 Å resolution structure of an RNA 20-mer that binds to the TREX2 complex
aMRC Laboratory of Molecular Biology, Francis Crick Avenue, Cambridge Biomedical Campus, Cambridge CB2 0QH, England
*Correspondence e-mail: ms@mrc-lmb.cam.ac.uk
The 1.25 Å resolution H32:R of a 20 nt ribonucleotide that binds to the TREX-2 complex with high affinity shows a double-stranded RNA duplex arranged along a crystallographic 31 axis in which the antiparallel chains overlap by 18 and are related by a crystallographic twofold axis. The duplex shows C–A, U–U and C–C noncanonical base pairings together with canonical Watson–Crick A–U and G–C pairs and a G–U wobble.
Keywords: RNA; noncanonical base pair.
PDB reference: RNA 20-mer, 5c5w
1. Introduction
The yeast TREX2 complex functions to integrate the nuclear components of the gene-expression pathway with GAL1 to nuclear pores (NPCs). TREX2 is based on a Sac3 scaffold, to which Thp1, Sem1, Cdc31 and two chains of Sus1 bind. The complex can be divided into three regions, each of which has been characterized structurally and functionally: (i) the CID region, containing Cdc31, both Sus1 chains and Sac3 residues 727–805 (Jani et al., 2009), that interacts with nuclear basket components such as Nup1 and, in addition to facilitating export, functions to localize actively transcribing genes to NPCs (Jani et al., 2014), (ii) the N-terminal region of Sac3 (residues 1–140) that binds to the principal yeast nuclear export factor, Mex67–Mtr2 (Dimitrova et al., 2015), and (iii) the M-region, in which Thp1 and Sem1 bind to Sac3 residues 250–563 and which contains two juxtaposed winged-helix domains that bind RNA (Ellisdon et al., 2012). Previous work indicated a level of specificity in RNA binding by TREX2 in that polyuridine binds more strongly than other polyribonucleotides (Ellisdon et al., 2012), and studies on the human analogue of TREX2 that is based on GANP (the Sac3 homologue) also showed some specificity in binding to a subset of transcripts (Wickramasinghe et al., 2014). Here, we report the 1.25 Å resolution of a 20 nt RNA fragment identified as having a higher affinity for TREX2 than polyuridine.
nuclear export as well as localizing actively expressing genes such as2. Methods and materials
An RNA oligonucleotide with the sequence ACCUGAGUUCAAUUCUAGCG was synthesized and HPLC-purified by Integrated DNA Technologies (Interleuvenlaan, Belgium) and dissolved at a concentration of 1.0 mM in 10 mM Tris–HCl pH 7.4. No further steps involving thermal and cooling were taken to promote duplex formation and additional magnesium chloride was not added to the sample. Crystallization conditions were screened using sparse-matrix formulations at 18°C by sitting-drop vapour diffusion using ∼200 nl drops. Crystals were observed in many conditions after several days. A variety of morphologies including needles, clusters and large individual crystals were observed in PEG-based (PEG 2K MME and PEG ranging from 600 to 8K) conditions and conditions where inorganic salts (ammonium sulfate, lithium sulfate, sodium chloride) were the main precipitant and over a range of different pH values (ranging from 5.5 to 10.5). No further optimization from initial screening was carried out. For X-ray data collection, crystals were supplemented with 15–30%(v/v) glycerol prior to vitrification in liquid nitrogen.
Diffraction data were collected on beamline I04-1 at the Diamond Light Source, Didcot, England. Surprisingly, there was no clear correlation between crystal appearance and diffraction quality. Although almost all of the crystals tested produced powder diffraction patterns, one crystal grown in 2.0 M ammonium sulfate, 0.2 M sodium chloride, 0.1 M sodium cacodylate pH 6.5 (Table 1) diffracted to high resolution and data were collected using a φ increment of 0.5°. These crystals diffracted to 1.25 Å resolution, with unit-cell parameters a = 39.9, b = 39.3, c = 156.9 Å, α = β = 90, γ = 120° (Table 2). Owing to the possibility of and the presence of noncystallographic symmetry, it was not possible to distinguish between H3:R and H32:R symmetry at this stage. Data were processed and reduced using XDS (Kabsch, 2010) and AIMLESS (Evans, 2011). tests for H3:R symmetry were equivocal using phenix.xtriage (Adams et al., 2010) and thus it was thought prudent to initially attempt using H3:R symmetry. Conventional using models comprising helices and duplexes was unsuccessful, and thus a set of shell scripts was developed that used fragments of high-quality RNA models from the PDB together with several ab initio models to derive a series of low-scoring solutions in Phaser (McCoy et al., 2007). All of the solutions obtained, irrespective of their scores, were subjected to 20 cycles of simulated annealing in Cartesian space as implemented in phenix.refine (Afonine et al., 2012) and were then used as starting models for phenix.autobuild (Adams et al., 2010) in a fully automated protocol. Model improvement was monitored using both Rfree and correlation coefficients during successive rounds of density modification and reciprocal-space Several different models of highly truncated duplexes, comprising two and three Watson–Crick base pairs, converged to generate almost identical models with Rfree < 40% and with clear features corresponding to bases and phosphate backbone in 2m|Fo| − D|Fc| electron-density maps together with features in the m|Fo| − D|Fc| difference density maps that were not yet accounted for by the model. The density obtained was sufficiently good to enable purines and pyrimidines to be assigned and the ribonucleotide sequence to be fitted unequivocally. In the crystals, 2–20 of the RNA formed an antiparallel helical duplex (Fig. 1) with its axis coincident with a crystallographic 31 screw axis and one of the twofold axes relating the two chains coincident with the position of a crystallographic twofold axis of the H32:R Consequently, this symmetry was adopted for further processing. The density corresponding to cytosine 10 that was related by this twofold was unusual and appeared to be the result of alternative conformations adopted by this nucleotide that also slightly influenced the adjacent (Fig. 2). The influence of the twofold axis perpendicular to the c axis on the distribution of diffraction intensities was very similar to that expected from (h, k, −l) Iterative cycles of rebuilding using Coot (Emsley et al., 2010) and with phenix.refine (Afonine et al., 2012), introduction of anisotropic temperature factors and insertion of waters produced a model with an R factor of 17.9% and an Rfree of 18.8% (Table 3). Some densities assigned to water probably represented ammonium ions that balanced the charge on the RNA phosphates but could not be identified unequivocally. The stereochemistry of the structure was assessed and validated with MolProbity (Chen et al., 2010), which indicated that there were no bond-length/angle outliers, that all ribose puckers were canonical and that the clashscore of 1.43 was in the top 98% of structures determined at this resolution.
|
|
|
3. Results
The C2–G18, G5–C15, A6–U14, U8–A12 and U9–A11 base pairings in the duplex were all of the canonical Watson–Crick form and that of G7–U13 (Fig. 3a) was of the canonical wobble form supplemented by a water-bridged hydrogen bond involving guanine N2, water 5 and uridine ribose O2′. In the similar C3–A17 wobble pairing the adenine N6 formed a single hydrogen bond to O4 of C3 (Fig. 3b), which differs from the arrangement observed in the 2.5 Å resolution structure of a 16-mer duplex (Pan et al., 1998), whereas the U4–U16 pairing (Fig. 3c) was based on hydrogen bonds between N3 and O4 of each base. Cytosine 10 adopted alternate conformations, but both formed 3.2 Å amino–imino hydrogen bonds between N4 and N3 (Fig. 3d) supplemented by hydrogen bonds between water 24 and N4 of both bases. A similar amino–imino hydrogen bonding was observed in the Hepatitis C virus internal ribosome entry site eIF3-binding site and thymidylate synthase where multiple conformations of this base were also observed (Collier et al., 2002; Tavares et al., 2009). There was no clear electron density for adenine 1, which appeared to be disordered owing to its being displaced to facilitate the canonical C–G pairings between bases 19 and 20 in consecutive duplexes along the crystallographic 31 axis. Consistent with this interpretation, the electron density for the ribose of cytosine 2 was also weak.
Although the RNA formed a duplex in the crystals, it appeared to be monomeric in solution and probably formed a hairpin using et al., 2007).
analogous to that observed in the duplex. The formation of duplexes from hairpins is favoured at the high and RNA concentrations employed for crystallization (NakanoFootnotes
‡Present address: Department of Biochemistry, Max Planck Institute for Developmental Biology, Spemannstrasse 35, 72076 Tübingen, Germany.
Acknowledgements
This work was supported by the Medical Research Council (U105178939 to MS) and a Wellcome Trust Programme grant (MS). No conflicts of interest are declared. We are most grateful to our colleagues in Cambridge, especially Andrew Leslie, Phil Evans, Shintaro Aibara, Jimmy Gordon and Neil Marshall, for their many helpful comments, criticisms and suggestions.
References
Adams, P. D. et al. (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
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
Collier, A. J., Gallego, J., Klinck, R., Cole, P. T., Harris, S. J., Harrison, G. P., Aboul-Ela, F., Varani, G. & Walker, S. (2002). Nature Struct. Biol. 9, 375–380. PubMed CAS Google Scholar
Dimitrova, L., Valkov, E., Aibara, S., Flemming, D., McLaughlin, S. H., Hurt, E. & Stewart, M. (2015). Structure, 23, 1246–1257. CrossRef CAS PubMed Google Scholar
Ellisdon, A. M., Dimitrova, L., Hurt, E. & Stewart, M. (2012). Nature Struct. Mol. Biol. 19, 328–336. CrossRef 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
Evans, P. R. (2011). Acta Cryst. D67, 282–292. Web of Science CrossRef CAS IUCr Journals Google Scholar
Jani, D., Lutz, S., Marshall, N. J., Fischer, T., Köhler, A., Ellisdon, A. M., Hurt, E. & Stewart, M. (2009). Mol. Cell, 33, 727–737. CrossRef PubMed CAS Google Scholar
Jani, D., Valkov, E. & Stewart, M. (2014). Nucleic Acids Res. 42, 6686–6697. CrossRef CAS PubMed Google Scholar
Kabsch, W. (2010). Acta Cryst. D66, 133–144. Web of Science CrossRef CAS 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
Nakano, S., Kirihata, T., Fujii, S., Sakai, H., Kuwahara, M., Sawai, H. & Sugimoto, N. (2007). Nucleic Acids Res. 35, 486–494. Web of Science CrossRef PubMed CAS Google Scholar
Pan, B., Mitra, S. N. & Sundaralingam, M. (1998). J. Mol. Biol. 283, 977–984. Web of Science CrossRef CAS PubMed Google Scholar
Tavares, T. J., Beribisky, A. V. & Johnson, P. E. (2009). RNA, 15, 911–922. CrossRef PubMed CAS Google Scholar
Wickramasinghe, V. O., Andrews, R., Ellis, P., Langford, C., Gurdon, J. B., Stewart, M., Venkitaraman, A. R. & Laskey, R. A. (2014). Nucleic Acids Res. 42, 5059–5071. 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.