|Home | About | Journals | Submit | Contact Us | Français|
Interest has been focused on comparative X-ray structure analyses of different tRNAGly acceptor-stem helices. tRNAGly/glycyl-tRNA synthetase belongs to the so-called class II system, in which the tRNA identity elements consist of simple and unique determinants that are located in the tRNA acceptor stem and the discriminator base. Comparative structure investigations of tRNAGly microhelices provide insight into the role of tRNA identity elements. Predominant differences in the structures of glycyl-tRNA synthetases and in the tRNA identity elements between prokaryotes and eukaryotes point to divergence during the evolutionary process. Here, the crystallization and high-resolution X-ray diffraction analysis of a human tRNAGly acceptor-stem microhelix with sequence 5′-G1C2A3U4U5G6G7-3′, 5′-C66C67A68A69U70G71C72-3′ is reported. The crystals belonged to the monoclinic space group C2, with unit-cell parameters a = 37.32, b = 37.61, c = 30.47 Å, β = 112.60° and one molecule per asymmetric unit. A data set was collected using synchrotron radiation and data were processed within the resolution range 50.0–1.18 Å. The structure was solved by molecular replacement.
tRNA identity elements assure the aminoacylation of tRNA by the cognate aminoacyl-tRNA synthetases. Owing to the redundancy of the genetic code, there are several isoacceptor tRNAs which code for the same amino acid but recognize different codons on the messenger RNA. Identity elements which consist of structure and/or sequence motifs on the tRNA determine the correct aminoacylation of tRNA isoacceptors and avoid the misaminoacylation of noncognate tRNAs.
Two different groups of tRNA/aminoacyl-tRNA synthetase systems exist. Class I tRNA/aminoacyl-tRNA synthetases possess complex identity elements which are mostly spread over different parts of the tRNA including the anticodon (Eriani et al., 1990 ). Within the class II system, tRNA identity elements consist of a small number of simple determinants (Eriani et al., 1990 ). These are located in the aminoacyl stem and sometimes include the discriminator base at position 73 of the tRNA. Such tRNA determinants can be so simple as to consist of only one base pair. This has been shown for the Escherichia coli tRNAAla (Hou & Schimmel, 1988 ).
tRNAGly/glycyl-tRNA synthetase (GlyRS) belongs to the class II system. In E. coli, the tRNAGly identity determinants consist of a conserved base pair C2·G71 and the U73 discriminator base (McClain et al., 1991 ). Nevertheless, the glycine system is a special case within class II as there are large differences in the synthetase structure and the tRNA identity elements between eukaryotic/archaebacterial and eubacterial organisms.
The eubacterial GlyRS has been well investigated. E. coli GlyRS, a tetrameric protein, consists of an α2β2 structure; both subunits are required for activity (Ostrem & Berg, 1974 ; Webster et al., 1983 ). The β-subunit is responsible for tRNA binding (Nagel et al., 1984 ). Amino-acid activation, which involves the binding of ATP and glycine, is governed by the α-subunit (Toth & Schimmel, 1990a ,b ). The tRNA identity elements are located on the aminoacyl stem and include the discriminator base, which has to be uracil.
In contrast, the eukaryotic/archaebacterial GlyRS is a simpler enzyme consisting of an α2 structure (Kern et al., 1981 ; Nada et al., 1993 ; Shiba et al., 1994 ; Shiba, 2005 ). The tRNA identity elements also differ between prokaryotes and eukaryotes. A major difference in tRNAGly identity is the discriminator base, which is strictly adenosine in the eukaryotic/archaebacterial case but is known to be uracil in the prokaryotic system (Shiba et al., 1994 ). In general, GlyRS also differs from most other class II aminoacyl-tRNA synthetases with respect to the three highly conserved binding motifs. The eukaryotic/archaebacterial system possesses three unique class II motifs 1–3, which all consist of a strongly conserved sequence. Motif 1 contains a GΦXXΦXXPΦΦ region, motif 2 a (F/Y/H)RX(E/D) stretch and motif 3 possesses a λXΦGΦGΦERΦΦΦΦΦ sequence (Eriani et al., 1990 ), in which X represents any residue, Φ a hydrophobic residue and λ a short-chain residue. The distance between motif 1 and motif 3 is relatively constant and lies between 40 and 80 residues, whereas the distance between motifs 2 and 3 can consist of between 70 and 300 residues (Eriani et al., 1990 ). In the prokaryotic enzymes only motif 3 is present; motifs 1 and 2 are possibly degenerated and could not be found in the primary sequences of the proteins (Moras, 1992 ; Shiba, 2005 ).
The glycine system becomes even more complicated when studying Thermus thermophilus GlyRS. The crystal structure of this enzyme has been solved to 2.75 Å resolution (Logan et al., 1995 ). The structure of the thermophilic GlyRS somewhat resembles the subunit arrangement of the eukaryotic/archaebacterial-type enzymes, whilst the tRNA identity elements show the features of the eubacterial tRNAs.
Here, we report the crystallization and high-resolution X-ray diffraction analysis of the human tRNAGly acceptor stem at 1.18 Å resolution. The structure was solved by molecular replacement. Comparative structure analysis with other tRNAGly acceptor-stem helices (Förster, Brauer et al., 2007 ; Förster et al., 2008 ) will help in understanding the tRNAGly–GlyRS system at the structural level in the region of the acceptor stem.
The human tRNAGly acceptor-stem microhelix isoacceptor investigated here was derived from the gene sequence with tRNA Database ID DG9990 (Sprinzl & Vassilenko, 2005 ) and is the counterpart of the human tRNAGly microhelix structure (derived from DG9992) that has recently been published (Förster et al., 2008 ). The oligonucleotides 5′-GCAUUGG-3′ and 5′-CCAAUGC-3′ which form the tRNAGly (DG9990) microhelix were purchased from IBA (Göttingen, Germany) at HPLC purification grade. The exact concentration of the RNA single strands was determined by alkaline hydrolysis, considering the molar extinction coefficients of the single nucleotides and using UV spectroscopy as described in Sproat et al. (1995 ), and both strands were annealed in water at a concentration of 0.5 mM. For this procedure, the strands were combined and heated to 363 K for 5 min and then cooled to room temperature for several hours. The resulting tRNAGly duplex was used in crystallization experiments without further purification.
We used two different nucleic acid crystal screening kits from Hampton Research (California, USA) for initial crystallization experiments. The Natrix Nucleic Acid Crystallization Kit (HR2-116; Hampton Research, California, USA) containing 48 different crystallization conditions was used with the sitting-drop vapour-diffusion technique and CrystalQuick Lp plates from Greiner Bio-One (Germany). 1 µl of the aqueous RNA solution at a concentration of 0.5 mM was mixed with 1 µl reservoir solution and equilibrated against 80 µl reservoir solution at 294 K. In parallel, the Nucleic Acid Mini Screen (HR2-118; Hampton Research, California, USA) with 24 different conditions was used with the hanging-drop vapour-diffusion technique and Linbro plates (ICN Biomedicals Inc., Ohio, USA). In this screening procedure, 1 µl 0.5 mM aqueous RNA solution was mixed with 1 µl crystallization solution from the screen and equilibrated against 1 ml reservoir solution consisting of 35%(v/v) MPD (2-methyl-2,4-pentanediol) in water pH 7.4 at a temperature of 294 K. Many tiny crystals appeared after 1–2 d using different crystallization conditions with 35%(v/v) MPD in the reservoirs. Only one condition yielded larger crystals with dimensions of approximately 0.3 × 0.1 × 0.1 mm (Fig. 1 ). These crystals grew in 10%(v/v) MPD, 40 mM sodium cacodylate pH 6.0, 12 mM spermine tetrachloride, 80 mM potassium chloride with equilibration against 35%(v/v) MPD at 294 K and appeared after 12 d. This condition was an initial crystallization solution from the Hampton screen and was not further optimized. Nevertheless, several crystallization setups were performed with this crystallization solution using the hanging-drop vapour-diffusion technique in Linbro plates and the best crystals were used for data collection.
Prior to X-ray diffraction data collection, the crystals were flash-frozen in liquid nitrogen. The crystallization buffer, which contained 10%(v/v) MPD, was directly used for freezing without the addition of any further cryoprotectant solution. The X-ray diffraction data were recorded on Diamond Synchrotron beamline I04 (UK) at a wavelength of 0.9696 Å. A high-resolution data set was recorded from 50 to 1.18 Å at a temperature of 100 K. All data were analyzed and processed using programs from the HKL-2000 suite (Otwinowski & Minor, 1997 ). The data set of the tRNAGly helix was analyzed for merohedral twinning using the Padilla and Yeates algorithm (Padilla & Yeates, 2003 ) as implemented in the web server http://nihserver.mbi.ucla.edu/pystats. The structure was solved by molecular replacement using the program Phaser (McCoy et al., 2005 ) with the human tRNAGly acceptor-stem microhelix isoacceptor G9992 (PDB code 2v7r) as a search model (Förster et al., 2008 ). Structure refinement is in progress using the program REFMAC5 (Murshudov et al., 1997 ). Both programs were used as implemented in the CCP4i package (Collaborative Computational Project, Number 4, 1994 ).
For human cytoplasm, three of the four different sequences for tRNAGly isoacceptor genes are available in the database of Sprinzl & Vassilenko (2005 ), with identification numbers DG9990, DG9991 and DG9992, of which DG9990 and DG9991 have identical sequences in their acceptor stems. The crystallization and the X-ray structure of the human tRNAGly isoacceptor derived from gene sequence DG9992 have recently been published (Förster, Szkaradkiewicz et al., 2007 ; Förster et al., 2008 ). The two tRNAGlys DG9992 and DG9990 possess different anticodons, UCC and GCC, respectively, matching the mRNA triplets coding for glycine. Nevertheless, both tRNAs are aminoacylated with the same amino acid; we therefore focused on comparative structural analyses of the aminoacyl stems of these isoacceptors. We have now successfully crystallized the human cytoplasmic tRNAGly isoacceptor with the sequence 5′-G1C2A3U4U5G6G7-3′ and 5′-C66C67A68A69U70G71C72-3′ (derived from gene sequence DG9990; Figs. 1 and 2 ). A high-resolution data set was recorded and X-ray diffraction data were processed in the 50–1.18 Å resolution range.
Like the recently described human tRNAGly microhelix (DG9992; Förster, Szkaradkiewicz et al., 2007 ), the human tRNAGly acceptor-stem microhelix investigated here (DG9990) crystallized in space group C2. A high-resolution data set was acquired using synchrotron radiation and cryogenic cooling on Diamond Synchrotron beamline I04 (UK).
The crystallographic parameters of the tRNAGly acceptor-stem microhelix (DG9990) are listed in Table 1 . The unit-cell parameters of the DG9990 tRNAGly microhelix reported here are fairly similar to those of the previously published DG9992 (Förster, Szkaradkiewicz et al., 2007 ).
A merohedral twinning analysis was applied to the crystallographic data using the Padilla and Yeates algorithm (Padilla & Yeates, 2003 ), since merohedral twinning can occur in crystals of short RNA duplexes (Rypniewski et al., 2006 ; Mueller, Muller et al., 1999 ; Mueller, Schübel et al., 1999 ). Analysis of the tRNAGly microhelix (DG9990) data set showed a curve which corresponds to a theoretically untwinned crystal; there was no indication of merohedral twinning.
The structure of the human tRNAGly microhelix was solved by molecular replacement using the program Phaser (McCoy et al., 2005 ) with the coordinates of the human tRNAGly microhelix isoacceptor with PDB code 2v7r (Förster et al., 2008 ) as a search model. The initial R value after molecular replacement was 36.0%, with a correlation coefficient of 56.6 and Z scores of 10.1 in the rotation function and 10.9 in the translation function. The structure is presently being refined and the R and R free values were approximately 20% and 23%, respectively, after adding solvent molecules (data not shown).
We will focus on comparative structure analysis between the human cytoplasmatic tRNAGly microhelices reported here and published previously (Förster et al., 2008 ). Additionally, the E. coli tRNAGly microhelix X-ray structure (Förster, Brauer et al., 2007 ) will be used for superposition experiments. The two human tRNAGly microhelices crystallize without magnesium, whereas in the E. coli tRNAGly microhelix a magnesium-binding site could be identified within the region of the tRNA identity elements (Förster, Brauer et al., 2007 ). This magnesium-binding site has also been reported for E. coli tRNAAla (Limmer et al., 1993 ) and for yeast tRNAPhe (Shi & Moore, 2000 ). Soaking experiments with the human tRNAGly microhelix crystals are under way to investigate potential magnesium-binding sites. A detailed comparison of the water molecules bound to the human tRNAGly microhelix structures will provide insight into the hydration patterns. This is of importance for initial protein–RNA contacts, which we believe to be facilitated via bound solvent molecules. Our investigations support the structure–function relationship within the different tRNAGly systems with respect to the evolutionary divergence between eubacteria and eukaryotes/archaebacteria.
This work was supported within the RiNA Network for RNA Technologies by the Federal Ministry of Education and Research, the City of Berlin and the European Regional Development Fund. We thank the Fonds der Chemischen Industrie (Verband der Chemischen Industrie eV), the Deutsches Zentrum für Luft- und Raumfahrt (DLR) and the BMBF/VDI-financed BiGRUDI Network of the Robert-Koch-Institute (Berlin) for additional support. We gratefully acknowledge the Diamond Synchrotron facility, UK for providing us with beamtime.