- Email: [email protected]
Crystal structure of a fructokinase homolog from Halothermothrix orenii
Journal of Structural Biology 171 (2010) 397–401
Contents lists available at ScienceDirect
Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi
Structure Report
Crystal structure of a fructokinase homolog from Halothermothrix orenii Teck Khiang Chua a, J. Seetharaman b, Joanna M. Kasprzak c, Cherlyn Ng a, Bharat K.C. Patel d, Christopher Love d, Janusz M. Bujnicki c,e, J. Sivaraman a,* a
Department of Biological Sciences, 14 Science Drive 4, National University of Singapore, Singapore 117543, Singapore X4 Beamline, Brookhaven National Laboratory, Upton, NY, USA c Bioinformatics Laboratory, Institute of Molecular Biology and Biotechnology, Faculty of Biology, Adam Mickiewicz University, ul. Umultowska 98, 61-614 Poznan, Poland d Microbial Gene Research and Resources Facility, School of Biomolecular and Physical Sciences, Griffith University, Brisbane, Queensland 4111, Australia e Laboratory of Bioinformatics and Protein Engineering, International Institute of Molecular and Cell Biology, Ks. Trojdena 4, PL-02-190 Warsaw, Poland b
a r t i c l e
i n f o
Article history: Received 29 November 2009 Received in revised form 17 May 2010 Accepted 17 May 2010 Available online 21 May 2010 Keywords: Hore_18220 Fructokinase FRK homolog Fructose ATP Structure
a b s t r a c t Fructokinase (FRK; EC 2.7.1.4) catalyzes the phosphorylation of D-fructose to D-fructose 6-phosphate (F6P). This irreversible and near rate-limiting step is a central and regulatory process in plants and bacteria, which channels fructose into a metabolically active state for glycolysis. Towards understanding the mechanism of FRK, here we report the crystal structure of a FRK homolog from a thermohalophilic bacterium Halothermothrix orenii (Hore_18220 in sequence databases). The structure of the Hore_18220 protein reveals a catalytic domain with a Rossmann-like fold and a b-sheet ‘‘lid” for dimerization. Based on comparison of Hore_18220 to structures of related proteins, we propose its mechanism of action, in which the lid serves to regulate access to the substrate binding sites. Close relationship of Hore_18220 and plant FRK enzymes allows us to propose a model for the structure and function of FRKs. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction Sugar kinases that are responsible for the phosphorylation of free monosaccharides such as glucose and fructose, the initial step of metabolic pathways, are broadly classified into three superfamilies: the galactokinases, hexokinases and ribokinases. Members of the galactokinase family are involved in diverse pathways, ranging from cholesterol and amino acid synthesis to galactose phosphorylation. As sucrose is the major saccharide in plants; two enzymes are responsible for the phosphorylation of sucrose cleavage products fructose and glucose. (Medina and Sols, 1956) Hexokinases (Hxk; EC 2.7.1.1) preferentially phosphorylate glucose and (Frankart and Pontis, 1976) Fructokinase (FRK; EC 2.7.1.4) (a member of the ribokinase-like superfamily according to PFAM and SCOP databases), is a ubiquitous and highly specific enzyme primarily catalyzing the transfer of a phosphate group from adenosine triphosphate (ATP) donor to a phosphate acceptor D-fructose to result in the formation of D-fructose 6-phosphate (F6P). FRK activity was first reported in 1956, but the protein was isolated and characterized only 20 years later (Medina and Sols, 1956; Frankart and Pontis, 1976). FRK is an enzyme belonging to the ribokinase-like superfamily of sugar kinases that show high substrate * Corresponding author. Fax: +65 6779 5671. E-mail address: [email protected] (J. Sivaraman). 1047-8477/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2010.05.007
specificity. Evolutionary analyses indicate that divergence of the FRK family from a ribokinase ancestor occurred before species divergence (Zhang et al., 2004). FRK specifically phosphorylates fructose with a km of 41–220 lM, at pH 8.0 and have much higher affinities for fructose than Hxk (Renz and Stitt, 1993). As fructose phosphorylation by FRK is irreversible and near rate-limiting, it regulates the rate and localization of carbon usage by channeling fructose into a metabolically active state for glycolysis in plants and bacteria (Zhang et al., 2003). This reaction is particularly important in sink tissues where sucrose assimilation, degradation and conversion to starch is mediated by invertase and/or sucrose synthase (SS), and the fructose produced must be phosphorylated to maintain the carbon flux to starch or respiration. FRKs are widely reported to have a preference for ATP over other nucleotides and unless in the presence of high GTP or UTP concentrations, ATP will be the principle source of phosphate (Chaubron et al., 1995; Martinez-Barajas et al., 1997). The activity of FRK greatly exceeding glucokinase in many tissues is consistent with the view that SS, rather than invertase, is the major route of sucrose degradation, thus producing a larger amount of fructose than glucose. In the course of a random sequence analysis of the Halothermothrix orenii genome, an open reading frame (ORF) encoding a putative FRK enzyme (EC: 2.7.1.4) was identified (Genbank Accession Number: Hore_18220) (Mijts and Patel, 2001). H. orenii is an anaerobic, thermohalophilic bacterium from the class Clostridia found in
398
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
the sediment of a Tunisian salted lake. The strain isolated, H168, oxidizes glucose, fructose, xylose, ribose, cellobiose, and starch (Sigrell et al., 1998). Here we report the crystal structure of Hore_18220, refined to 2.8 Å resolution. Based on the comparative analysis of this structure to related enzymes’ structures and mechanisms described in the literature, we propose a model for FRK-ligand interactions.
Table 1 Data collection and refinement statistics.
2. Purification, crystallization and structure determination Halothermothrix orenii FRK homolog (hereafter referred as Hore_18220) gene was amplified by PCR and ligated into the pTrcHisA expression vector (Invitrogen) at BamHI and KpnI cut sites (Mijts and Patel, 2001). His-tagged Hore_18220 was expressed heterologously in BL21 (DE3). Selenomethionine-substituted protein was expressed using methionine auxotroph Escherichia coli DL41 in LeMaster medium supplemented with 25 mg/L selenomethionine (SelMet). Hore_18220 was first purified by Ni–NTA affinity chromatography (Qiagen) and then by gel filtration (Hiload 16/60 Superdex200). Purified Hore_18220 was concentrated to 10 mg/ml. Crystallization screen was carried out through hanging-drop vapor-diffusion method using Hampton Research (Aliso Viejo, CA, USA) screens as well as by micro batch under-oil technique using JB crystallization screens (Jena Biosciences, Jena, Germany) at room temperature. Initially crystals were small. Obtaining the diffraction quality crystals was the most challenging part of this project. The present data set is the best of many data sets collected. As an approach to improve the data quality, we have also attempted cocrystallization/soaking with the substrates. The best diffraction quality crystals were obtained from 8% PEG 4000, 0.8 M LiCl2 and 0.1 M Tris–HCl, pH 8.5 by using micro batch under-oil technique with 2 ll of the crystallization solution mixed with 2 ll of protein under 15 ll of paraffin oil. Crystals were directly taken from the drop and the synchrotron data were collected at beam lines X12C and X29, NSLS, Brookhaven National Laboratory for the SelMet/native crystals using Quantum 4-CCD detector (Area Detector Systems Corp., Poway, CA, USA) to 2.8 Å resolution (Table 1). Data were processed and scaled using the program HKL2000 (Otwinowski and Minor, 1997). All the four expected selenium sites in an asymmetric unit were located by the program SOLVE (Terwilliger and Berendzen, 1999). Initial phases were further developed by RESOLVE (Terwilliger, 2000), which improved the overall figure of merit (FOM) to 0.68 and automatically built 50%, of the residues of one asymmetric unit. The remaining parts of the molecules were built manually using the program O (Jones et al., 1991). Several cycles of model building alternating with refinement using the program CNS (Brunger et al., 1998) resulted in the final model, with an R-factor of 0.254 (Rfree = 0.288) at 2.8 Å resolution with reflections I > rI was used in the refinement. The final model comprises 278 residues from each monomer and 142 water molecules. The first 21 N terminal residues, His-tag with the linker residues, loop regions comprising residues 144–145 and 267–271, and the C-terminal 21 residues (307–327) were not fully resolved in the electron density map and therefore were not modeled. Procheck (Laskowski et al., 1993) analysis shows that two residues are in the disallowed regions. These residues are found in the turns and are well defined by the electron density map. 3. Overall structure Hore_18220 crystallized with two molecules in the asymmetric unit related by a 2-fold rotation non-crystallographic symmetry approximately parallel to the c-axis. Interestingly, these two molecules are packed one over the other through b-strands, resulting
Data set
Peak
Inflection
Remotea
Data collection Resolution range (Å) Wavelength (Å) Observed reflections >1 Unique reflections Completeness (%) Overall (I/rI) Rsym (%)b
50.0–2.8 0.9790 117,292 30,770 99.9 19.8 4.6
50.0–2.8 0.9794 117,115 30,424 99.9 18.9 4.5
50.0–2.8 0.9600 116,025 30,193 99.9 16.7 4.9
Refinementc and quality Resolution range (Å) Rwork (No. of reflections)d Rfree (No. of reflections)e R.M.S.D. bond lengths (Å) R.M.S.D. bond angles
50.0–2.8 0.254 (25238) 0.288 (1846) 0.008 1.72
Average B-factors (Å2) Monomer A (all atoms) Monomer B (all atoms)
46.8 45.7
Ramachandran plotf Most favored regions (%) Additional allowed regions (%) Generously allowed regions (%) Disallowed regions (%)
83.9 14.5 0.7 0.9
a
NCS restraint was kept throughout the refinement. Rsym = |Ii |/|Ii| where Ii is the intensity of the ith measurement, and is the mean intensity for that reflection. c For all models, reflections with I > rI was used in the refinement. d Rwork = |FP FP(calc)|/FP. e R-free was calculated with 6% of the reflections in the test set. f Statistics for the Ramachandran plot from an analysis using Procheck (Laskowski et al., 1993). b
in the formation of a continuous b-sheet between two subunits of the dimer (Fig. 2). Gel filtration and dynamic light scattering experiments indicate that Hore_18220 exists as a dimer in solution (data not shown). This is consistent with the dimeric arrangement observed in the crystal structure. Each monomer has a mixed a/b fold and a characteristic nucleotide binding domain that resembles the Rossmann fold (Leu22– Ile30, Ser56–Thr108, Ala128–Ile306) (hereafter referred to as the catalytic domain) with a b-sheet ‘‘lid” (or lid region) (Leu31– Gly55, Thr109–Glu127). The substrate binding cleft is located at the interface between the catalytic domain and the lid region with a dimension of approximately 18 Å width and 22 Å length (Fig. 1). The core catalytic domain consists of a b-sheet with nine mostly parallel b-strands. This b-sheet is flanked on both sides by eight helices; of which three are very short (1–2 turns). In addition, each monomer contributes four b-strands at the dimer interface to form a tilted antiparallel b-sheet lid that runs from one subunit in the dimer to the other. This b-sheet maintains the dimeric architecture of Hore_18220. The observation of a dimeric molecule in solution as well as in crystal structure suggests a functionally important role for dimerization of Hore_18220. 4. Comparisons with other proteins A search for structurally similar proteins was performed using DALI (Holm and Sander, 1993). Structures showing overall similarity belonged only to the ribokinase-like superfamily of proteins. The structurally most common feature of these proteins is the substrate binding cleft region. The highest structural similarity is observed between Hore_18220 and a recently deposited (doi:10.2210/pdb3gbu/pdb; but not yet described in the literature) structure of an uncharacterized (putative) sugar kinase PH1459 from Pyrococcus horikoshii (PDB code 3ewm) with RMSD of 2.1 Å for 256 Ca atoms, 33% sequence identity and Z-score 31.9. This is
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
followed by aminoimidazole riboside (AIR) kinase from Salmonella enterica (PDB code 1TYY and 1TZ6) with RMSD of 2.0 Å for 262 Ca atoms, 24% identity and Z-score 31.2. Moreover, D-2-keto-3deoxygluconate (KDG) kinase (PDB code 1V19; RMSD = 2.3 Å for 261 Ca atoms; 26% identity, Z-score 30.1) and E. coli ribokinase (PDB code 1RKD; RMSD = 2.2 Å for 252 Ca atoms; 23% identity, Z-score 27.6) are also structurally similar. In addition, a recently deposited (but not yet described in the literature) structure of FRK homolog from Bacteroides thetaiotaomicron VPI-5482 (PDB code 2QHP; RMSD = 2.2 Å for 247 Ca atoms; 17% identity, Z-score 28.5) is found to be structurally related. In several cases, the individual domains of Hore_18220 and ribokinase homologs superpose well, however, the relative disposition of the domains often varies (Fig. 2A). The structure-based sequence alignment of Hore_18220 with the abovementioned homologs from ribokinase family showed that conservation of the residues is predominantly in the substrate and nucleotide binding pockets (Fig. 2B). To analyze the relationship of Hore_18220 to other proteins we performed sequence database searches with PSI-BLAST (Altschul et al., 1997) and clustered the obtained sequences with CLANS (Frickey and Lupas, 2004). Hore_18220 groups together with members of the COG0524 family (annotated as ‘sugar kinases, ribokinases’) from the Cluster of Orthologous Groups (COG) database, to the exclusion of various paralogs (members of other COGs) that exhibit other functions. Within this family, Hore_18220 belongs to a small group of prokaryotic proteins with the mutual sequence similarity in the range of 57–60%. This group includes PH1459 from P. horikoshii found by the DALI search. Among proteins with known function, plant FRK enzymes are the most closely related to Hore_18220 and PH1459 (sequence similarity in the range of 25–32%). Thus, even though the substrate specificities of Hore_18220 and PH1459 remain unknown, these proteins appear to be potentially the closest structural templates to model the structure and mechanism of action of plant FRKs. 5. Putative ATP and substrate binding pocket Although no position of ATP and substrate in the Hore_18220 structure was identified through crystallization, the binding sites
399
of ATP and the substrate can be predicted by comparison with structures of three known ribokinase family members complexed with ATP (PDB codes 1TZ6, 1V19 and 1RKD) (Fig. 3). To infer the catalytic mechanism of action for Hore_18220 (and, by extension, for plant FRK enzymes), we constructed a model of Hore_18220 complexed with ATP by remodeling the crystal structure of apoHore_18220 to mimic the conformation of bound substrates observed in the crystal structure of PH1459 complexed with ATP and E. coli ribokinase complexed with ribose and ADP (1RKD). We have also constructed a homology-model of a FRK from tomato (Solanum lycopersicum) and its complex with ATP and fructose (whose position was modeled by analogy to ribose in the 1RKD structure) (Fig 3). The superposition of related structures suggest that the overall architecture of the binding site is conserved among Hore_18220, AIR kinase (Zhang et al., 2004), KDG kinase (Ohshima et al., 2004), PH1459 and ribokinase (Sigrell et al., 1998) and these enzymes may interact with their ligands in a similar way. ATP is predicted to bind to a pocket on the catalytic domain of Hore_18220 lined by residues Asp181–Cys183; Lys210–Asp215; Thr243– Gly248; Ala267–Gly274; Leu297; Asn299; Val301 and Phe304. Similarly, the substrate binding pocket is predicted to be lined by residues Leu31, Asp33, Leu43, Gly54–Ser56, Asn59, Phe153 and Asp275, of which Leu31 and Asp33 are from the lid region. Thr243, Gly274 and Asn299 located in the ATP binding pocket are strictly conserved across all FRK homolog sequences compared (Fig. 2B). The equivalent residues of Thr243 and Gly274 in ribokinase, KDG kinase, PH1459, and AIR kinases were found to interact with the phosphate group of the bound ATP mainly through water mediated hydrogen bonds, while the equivalent residue of Asn299 interacts with the adenosine base. These residues in Hore_18220 would probably play a role in forming the oxyanion hole to stabilize the intermediate during phosphorylation as they bind to both substrate and ATP. Moreover, in comparison with ribokinases, the substrate and cofactor interacting residues show higher conservation among bacterial and plant FRKs. In particular, the residues lining the substrate and ATP binding pockets such as Pro182 (H. orenii numbering), Arg185, Val268, Asn299 are highly conserved in plant and
Fig. 1. Crystal structure of Hore_18220. (A) Shows the ribbon representation of the Hore_18220 monomer. (B) Ribbon diagram showing the dimeric Hore_18220 in the asymmetric unit. The catalytic domains of the monomers (residues Leu22–Ile30; Ser56–Thr108 and Ala128–Ile306) are depicted in blue and magenta and the b-sheet ‘‘lid” regions (residues Leu31–Gly55 and Thr109–Glu127) in red and cyan. The N- and C-terminals are labeled. These figures were prepared using the programs PYMOL (DeLano, 2008).
400
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
Fig. 2. Structural and sequence comparison of Hore_18220. (A) Stereo view of the superposition of Hore_18220 and its homologs. Hore_18220 (cyan) and the structures of the five ribokinase family members (protein–ligand complexes) superimposed on the catalytic domain of Hore_18220. Colored lines represent the Ca trace of AIR kinase (PDB code 1TZ6, green), RDK kinase (PDB code 1V1B, yellow), RK (PDB code 1RKD, magenta), PH1459 (PDB code 3EWM, blue) and putative 5-dehydro-2-deoxygluconokinase (PDB code 2QCV, orange). Substrates in the various structures are represented by sticks. This figure was prepared using the programs PYMOL (DeLano, 2008). (B) Structural and sequence alignment of Hore_18220. Top six rows: Structure-based sequence alignment of Hore_18220 (black), SKK (PDB code 3EWM, blue), DGK (PDB code 2QCV, cyan), ARK (PDB code 1TZ6, green), KDK (PDB code 1V19, brown) and RKK (PDB code 1RKD, magenta). The amino acids are in one-letter codes; the conserved residues are in red. Strictly conserved residues are highlighted red (identity above 48%). Secondary structural elements of Hore_18220 belonging to the a/b domain and the b ‘‘lid” are shown in blue and red, respectively. This figure was created using the program BioEdit (Hall, 1999). Middle 7–10th rows: Sequence alignment of Hore_18220 (top, black) with the closest four bacterial FRK homologs (orange) was carried out using CLUSTAL_X (Thompson et al., 1997) and refined manually. Bottom 11–14th rows: Sequence alignment of Hore_18220 (top, black) with the closest four plant FRK homologs (black). The anion hole motif GAGD is indicated by magenta asterisks. Proposed key substrate binding residues of fructose and ATP are indicated by blue and red asterisks, respectively. Suffix: FRK_Ho: Hore_18220 (FRK, H. orenii homolog); SKK_Ph: Sugar Kinase, Pyrococcus horikoshii (3EWM); DGK_Bh: 5-dehydro-2-deoxygluconokinase, Bacillus halodurans (2QCV); ARK_Se: Aminoimidazole riboside kinase, Salmonella enterica (1TZ6); KDK_Tt: 2Keto-3-Deoxygluconate Kinase, Thermus thermophilus (1V19); RKK_Ec: Ribokinase, Escherichia coli (1RKD); FRK_Pm: FRK, Petrotoga mobilis SJ95; FRK_Pd: FRK, Polaribacter dokdonensis; FRK_Cs: FRK, Cellulophaga sp. MED134; FRK_Pt: FRK, Psychroflexus torquis ATCC 700755; FRK_Sl: FRK, Solanum lycopersicum; FRK_At: FRK, Arabidopsis thaliana; FRK_Zm: FRK, Zea mays; FRK_Bv: FRK, Beta vulgaris.
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
401
Protein Data Bank accession code Coordinates and structure factors for the Hore_18220 have been deposited with RCSB Protein Data Bank (PDB) with code 3HJ6. Acknowledgments
Fig. 3. Computationally modeled FRK structure from Solanum lycopersicum (tomato) with Fructose and ATP. Protein backbone is shown as a ribbon, conserved substrate binding residues, the cofactor and fructose are shown as sticks.
Data for this study were measured at beamlines X12C and X29 of the National Synchrotron Light Source, Brookhaven National Laboratories (BNL). We thank Dr. Anand Saxena (BNL) for assistance in data collection. J.S. acknowledges research support from the Academic Research Fund (ARF) Grant No. R154000245112, National University of Singapore (NUS). J.M.K and J.M.B were supported by the Polish Ministry of Science (Grant 188/N-DFG/2008/0) and by the NIH (Grant 1R01 GM081680-01). BKCP acknowledges the support from the Griffith University Research Grants scheme. We thank Mr. Sun Qingxiang for his assistance in the final refinement of the model. Chua Teck Khiang is a Ph.D. student in receipt of a research scholarship from the National University of Singapore (NUS). References
Hore_18220, respectively. This suggests that these residues are mostly required to create the appropriate binding pockets in both plant and Hore_18220. Our analysis of structure and sequence conservation between Hore_18220 and ribokinases (RK) suggest that the substrate binding pocket in Hore_18220 and its close homologs (including plant FRK enzymes) is lined up with smaller side chains to provide more space for large substrates. For instance in RK residues Gly41 to Gly55 (corresponding to Gly54 to Gly68 in Hore_18220) directly interact with ribose (in RK) and the corresponding region is highly conserved in both enzyme families. However, Lys43 and Gly44 in RK are replaced by Ala and Pro in plant FRK and Ser and Pro in Hore_18220. A large positively charged amino acid (Lys43 in RK) is replaced by a small, uncharged residue (Ala or Ser) in Hore_18220. Similarly Gln47 of RK is replaced by small hydrophobic residues in FRK homologs, i.e. Ile in H. orenii and Val in plants. It is further observed that the lid region (Leu31–Gly55; Thr109–Glu127) is highly conserved in both bacterial and plant FRK homologs, whereas this region is not conserved among the ribokinases. This region in FRK homologs might be responsible for the regulation of the open/ closed state of the binding site. Finally, based on the model of the tomato FRK complexed with ligands, we predict that Glu33, Asp37, Ala60, Asn63, Ile117, Arg192 and Asp285 of the plant enzyme may be particularly important for the binding of the substrate (Fig 3). In summary, we have determined the structure of a FRK homolog from H. orenii. The plant FRK proteins present difficulties for purification and characterization (Medina and Sols, 1956; Frankart and Pontis, 1976). Our analysis presented in this article illustrates similarities and differences between sugar kinases acting on different substrates (e.g. ribose vs. fructose) and provides a working model towards studying the mechanism of plant FRKs. In particular: (1) Hore_18220 belongs to a family of bacterial proteins that are the closest prokaryotic homologs of plant FRKs. (2) The structure of Hore_18220 was used as a template to homology-model the structure of tomato FRK, in complex with its substrates (ATP and fructose), which helped predict fructosebinding residues in the plant enzyme. (3) Residues in the putative substrate binding pocket of Hore_18220 are conserved in plant FRKs; they have smaller side chains than corresponding residues in ribokinases, which makes the Hore_18220 and FRK pocket larger and potentially allows for accommodation of larger substrates.
Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., Read, R.J., Rice, L.M., Simonson, T., Warren, G.L., 1998. Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D 54, 905–921. Chaubron, F., Harris, N., Ross, H.A., Davies, H.V., 1995. Partial purification and characterization of fructokinase from developing taproots of sugar beet (Beta vulgaris). Plant Sci. 110, 181–186. DeLano, W.L., 2008. The PyMOL Molecular Graphics System. DeLano Scientific LLC, Palo Alto, CA, USA. Frankart, W.A., Pontis, H.G., 1976. Fructose metabolism in plants. Isolation and properties of pea seed frucktokinase. Acta Physiol. Latinoam. 26, 319–329. Frickey, T., Lupas, A., 2004. CLANS: a Java application for visualizing protein families based on pairwise similarity. Bioinformatics 20, 3702–3704. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41, 95–98. Holm, L., Sander, C., 1993. Protein structure comparison by alignment of distance matrices. J. Mol. Biol. 233, 123–138. Jones, T.A., Zou, J.Y., Cowan, S.W., Kjeldgaard, M., 1991. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, 110–119. Laskowski, R.A., McArthur, M.W., Moss, D.S., Thornton, J.M., 1993. Procheck: a program to check the stereochemical quality of protein structures. J Appl. Crystallogr. 26, 282–291. Martinez-Barajas, E., Luethy, M.H., Randall, D.D., 1997. Molecular cloning and analysis of fructokinase expression in tomato (Lycopersicon esculentum Mill.). Plant Sci. 125, 13–20. Medina, A., Sols, A., 1956. Specific fructokinase in peas. Biochim. Biophys. Acta 19, 378–379. Mijts, B.N., Patel, B.K., 2001. Random sequence analysis of genomic DNA of an anaerobic, thermophilic, halophilic bacterium, Halothermothrix orenii. Extremophiles 5, 61–69. Ohshima, N., Inagaki, E., Yasuike, K., Takio, K., Tahirov, T.H., 2004. Structure of Thermus thermophilus 2-Keto-3-deoxygluconate kinase: evidence for recognition of an open chain substrate. J. Mol. Biol. 340, 477–489. Otwinowski, Z., Minor, W., 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–326. Renz, A., Stitt, M., 1993. Substrate specificity and product inhibition of different forms of fructokinase and hexokinases in developing potato tubers. Planta 190, 166–175. Sigrell, J.A., Cameron, A.D., Jones, T.A., Mowbray, S.L., 1998. Structure of Escherichia coli ribokinase in complex with ribose and dinucleotide determined to 1.8 Å resolution: insights into a new family of kinase structures. Structure 6, 183–193. Terwilliger, T.C., 2000. Maximum-likelihood density modification. Acta Crystallogr. D 56, 965–972. Terwilliger, T.C., Berendzen, J., 1999. Automated MAD and MIR structure solution. Acta Crystallogr. D 55, 849–861. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882. Zhang, S.R., Nichols, S.E., Dong, J.G., 2003. Cloning and characterization of two fructokinases from maize. Plant Sci. 165, 1051–1058. Zhang, Y., Dougherty, M., Downs, D.M., Ealick, S.E., 2004. Crystal Structure of an aminoimidazole riboside kinase from Salmonella enterica: implications for the evolution of the ribokinase superfamily. Structure 12, 1809–1821.
Contents lists available at ScienceDirect
Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi
Structure Report
Crystal structure of a fructokinase homolog from Halothermothrix orenii Teck Khiang Chua a, J. Seetharaman b, Joanna M. Kasprzak c, Cherlyn Ng a, Bharat K.C. Patel d, Christopher Love d, Janusz M. Bujnicki c,e, J. Sivaraman a,* a
Department of Biological Sciences, 14 Science Drive 4, National University of Singapore, Singapore 117543, Singapore X4 Beamline, Brookhaven National Laboratory, Upton, NY, USA c Bioinformatics Laboratory, Institute of Molecular Biology and Biotechnology, Faculty of Biology, Adam Mickiewicz University, ul. Umultowska 98, 61-614 Poznan, Poland d Microbial Gene Research and Resources Facility, School of Biomolecular and Physical Sciences, Griffith University, Brisbane, Queensland 4111, Australia e Laboratory of Bioinformatics and Protein Engineering, International Institute of Molecular and Cell Biology, Ks. Trojdena 4, PL-02-190 Warsaw, Poland b
a r t i c l e
i n f o
Article history: Received 29 November 2009 Received in revised form 17 May 2010 Accepted 17 May 2010 Available online 21 May 2010 Keywords: Hore_18220 Fructokinase FRK homolog Fructose ATP Structure
a b s t r a c t Fructokinase (FRK; EC 2.7.1.4) catalyzes the phosphorylation of D-fructose to D-fructose 6-phosphate (F6P). This irreversible and near rate-limiting step is a central and regulatory process in plants and bacteria, which channels fructose into a metabolically active state for glycolysis. Towards understanding the mechanism of FRK, here we report the crystal structure of a FRK homolog from a thermohalophilic bacterium Halothermothrix orenii (Hore_18220 in sequence databases). The structure of the Hore_18220 protein reveals a catalytic domain with a Rossmann-like fold and a b-sheet ‘‘lid” for dimerization. Based on comparison of Hore_18220 to structures of related proteins, we propose its mechanism of action, in which the lid serves to regulate access to the substrate binding sites. Close relationship of Hore_18220 and plant FRK enzymes allows us to propose a model for the structure and function of FRKs. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction Sugar kinases that are responsible for the phosphorylation of free monosaccharides such as glucose and fructose, the initial step of metabolic pathways, are broadly classified into three superfamilies: the galactokinases, hexokinases and ribokinases. Members of the galactokinase family are involved in diverse pathways, ranging from cholesterol and amino acid synthesis to galactose phosphorylation. As sucrose is the major saccharide in plants; two enzymes are responsible for the phosphorylation of sucrose cleavage products fructose and glucose. (Medina and Sols, 1956) Hexokinases (Hxk; EC 2.7.1.1) preferentially phosphorylate glucose and (Frankart and Pontis, 1976) Fructokinase (FRK; EC 2.7.1.4) (a member of the ribokinase-like superfamily according to PFAM and SCOP databases), is a ubiquitous and highly specific enzyme primarily catalyzing the transfer of a phosphate group from adenosine triphosphate (ATP) donor to a phosphate acceptor D-fructose to result in the formation of D-fructose 6-phosphate (F6P). FRK activity was first reported in 1956, but the protein was isolated and characterized only 20 years later (Medina and Sols, 1956; Frankart and Pontis, 1976). FRK is an enzyme belonging to the ribokinase-like superfamily of sugar kinases that show high substrate * Corresponding author. Fax: +65 6779 5671. E-mail address: [email protected] (J. Sivaraman). 1047-8477/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2010.05.007
specificity. Evolutionary analyses indicate that divergence of the FRK family from a ribokinase ancestor occurred before species divergence (Zhang et al., 2004). FRK specifically phosphorylates fructose with a km of 41–220 lM, at pH 8.0 and have much higher affinities for fructose than Hxk (Renz and Stitt, 1993). As fructose phosphorylation by FRK is irreversible and near rate-limiting, it regulates the rate and localization of carbon usage by channeling fructose into a metabolically active state for glycolysis in plants and bacteria (Zhang et al., 2003). This reaction is particularly important in sink tissues where sucrose assimilation, degradation and conversion to starch is mediated by invertase and/or sucrose synthase (SS), and the fructose produced must be phosphorylated to maintain the carbon flux to starch or respiration. FRKs are widely reported to have a preference for ATP over other nucleotides and unless in the presence of high GTP or UTP concentrations, ATP will be the principle source of phosphate (Chaubron et al., 1995; Martinez-Barajas et al., 1997). The activity of FRK greatly exceeding glucokinase in many tissues is consistent with the view that SS, rather than invertase, is the major route of sucrose degradation, thus producing a larger amount of fructose than glucose. In the course of a random sequence analysis of the Halothermothrix orenii genome, an open reading frame (ORF) encoding a putative FRK enzyme (EC: 2.7.1.4) was identified (Genbank Accession Number: Hore_18220) (Mijts and Patel, 2001). H. orenii is an anaerobic, thermohalophilic bacterium from the class Clostridia found in
398
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
the sediment of a Tunisian salted lake. The strain isolated, H168, oxidizes glucose, fructose, xylose, ribose, cellobiose, and starch (Sigrell et al., 1998). Here we report the crystal structure of Hore_18220, refined to 2.8 Å resolution. Based on the comparative analysis of this structure to related enzymes’ structures and mechanisms described in the literature, we propose a model for FRK-ligand interactions.
Table 1 Data collection and refinement statistics.
2. Purification, crystallization and structure determination Halothermothrix orenii FRK homolog (hereafter referred as Hore_18220) gene was amplified by PCR and ligated into the pTrcHisA expression vector (Invitrogen) at BamHI and KpnI cut sites (Mijts and Patel, 2001). His-tagged Hore_18220 was expressed heterologously in BL21 (DE3). Selenomethionine-substituted protein was expressed using methionine auxotroph Escherichia coli DL41 in LeMaster medium supplemented with 25 mg/L selenomethionine (SelMet). Hore_18220 was first purified by Ni–NTA affinity chromatography (Qiagen) and then by gel filtration (Hiload 16/60 Superdex200). Purified Hore_18220 was concentrated to 10 mg/ml. Crystallization screen was carried out through hanging-drop vapor-diffusion method using Hampton Research (Aliso Viejo, CA, USA) screens as well as by micro batch under-oil technique using JB crystallization screens (Jena Biosciences, Jena, Germany) at room temperature. Initially crystals were small. Obtaining the diffraction quality crystals was the most challenging part of this project. The present data set is the best of many data sets collected. As an approach to improve the data quality, we have also attempted cocrystallization/soaking with the substrates. The best diffraction quality crystals were obtained from 8% PEG 4000, 0.8 M LiCl2 and 0.1 M Tris–HCl, pH 8.5 by using micro batch under-oil technique with 2 ll of the crystallization solution mixed with 2 ll of protein under 15 ll of paraffin oil. Crystals were directly taken from the drop and the synchrotron data were collected at beam lines X12C and X29, NSLS, Brookhaven National Laboratory for the SelMet/native crystals using Quantum 4-CCD detector (Area Detector Systems Corp., Poway, CA, USA) to 2.8 Å resolution (Table 1). Data were processed and scaled using the program HKL2000 (Otwinowski and Minor, 1997). All the four expected selenium sites in an asymmetric unit were located by the program SOLVE (Terwilliger and Berendzen, 1999). Initial phases were further developed by RESOLVE (Terwilliger, 2000), which improved the overall figure of merit (FOM) to 0.68 and automatically built 50%, of the residues of one asymmetric unit. The remaining parts of the molecules were built manually using the program O (Jones et al., 1991). Several cycles of model building alternating with refinement using the program CNS (Brunger et al., 1998) resulted in the final model, with an R-factor of 0.254 (Rfree = 0.288) at 2.8 Å resolution with reflections I > rI was used in the refinement. The final model comprises 278 residues from each monomer and 142 water molecules. The first 21 N terminal residues, His-tag with the linker residues, loop regions comprising residues 144–145 and 267–271, and the C-terminal 21 residues (307–327) were not fully resolved in the electron density map and therefore were not modeled. Procheck (Laskowski et al., 1993) analysis shows that two residues are in the disallowed regions. These residues are found in the turns and are well defined by the electron density map. 3. Overall structure Hore_18220 crystallized with two molecules in the asymmetric unit related by a 2-fold rotation non-crystallographic symmetry approximately parallel to the c-axis. Interestingly, these two molecules are packed one over the other through b-strands, resulting
Data set
Peak
Inflection
Remotea
Data collection Resolution range (Å) Wavelength (Å) Observed reflections >1 Unique reflections Completeness (%) Overall (I/rI) Rsym (%)b
50.0–2.8 0.9790 117,292 30,770 99.9 19.8 4.6
50.0–2.8 0.9794 117,115 30,424 99.9 18.9 4.5
50.0–2.8 0.9600 116,025 30,193 99.9 16.7 4.9
Refinementc and quality Resolution range (Å) Rwork (No. of reflections)d Rfree (No. of reflections)e R.M.S.D. bond lengths (Å) R.M.S.D. bond angles
50.0–2.8 0.254 (25238) 0.288 (1846) 0.008 1.72
Average B-factors (Å2) Monomer A (all atoms) Monomer B (all atoms)
46.8 45.7
Ramachandran plotf Most favored regions (%) Additional allowed regions (%) Generously allowed regions (%) Disallowed regions (%)
83.9 14.5 0.7 0.9
a
NCS restraint was kept throughout the refinement. Rsym = |Ii |/|Ii| where Ii is the intensity of the ith measurement, and is the mean intensity for that reflection. c For all models, reflections with I > rI was used in the refinement. d Rwork = |FP FP(calc)|/FP. e R-free was calculated with 6% of the reflections in the test set. f Statistics for the Ramachandran plot from an analysis using Procheck (Laskowski et al., 1993). b
in the formation of a continuous b-sheet between two subunits of the dimer (Fig. 2). Gel filtration and dynamic light scattering experiments indicate that Hore_18220 exists as a dimer in solution (data not shown). This is consistent with the dimeric arrangement observed in the crystal structure. Each monomer has a mixed a/b fold and a characteristic nucleotide binding domain that resembles the Rossmann fold (Leu22– Ile30, Ser56–Thr108, Ala128–Ile306) (hereafter referred to as the catalytic domain) with a b-sheet ‘‘lid” (or lid region) (Leu31– Gly55, Thr109–Glu127). The substrate binding cleft is located at the interface between the catalytic domain and the lid region with a dimension of approximately 18 Å width and 22 Å length (Fig. 1). The core catalytic domain consists of a b-sheet with nine mostly parallel b-strands. This b-sheet is flanked on both sides by eight helices; of which three are very short (1–2 turns). In addition, each monomer contributes four b-strands at the dimer interface to form a tilted antiparallel b-sheet lid that runs from one subunit in the dimer to the other. This b-sheet maintains the dimeric architecture of Hore_18220. The observation of a dimeric molecule in solution as well as in crystal structure suggests a functionally important role for dimerization of Hore_18220. 4. Comparisons with other proteins A search for structurally similar proteins was performed using DALI (Holm and Sander, 1993). Structures showing overall similarity belonged only to the ribokinase-like superfamily of proteins. The structurally most common feature of these proteins is the substrate binding cleft region. The highest structural similarity is observed between Hore_18220 and a recently deposited (doi:10.2210/pdb3gbu/pdb; but not yet described in the literature) structure of an uncharacterized (putative) sugar kinase PH1459 from Pyrococcus horikoshii (PDB code 3ewm) with RMSD of 2.1 Å for 256 Ca atoms, 33% sequence identity and Z-score 31.9. This is
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
followed by aminoimidazole riboside (AIR) kinase from Salmonella enterica (PDB code 1TYY and 1TZ6) with RMSD of 2.0 Å for 262 Ca atoms, 24% identity and Z-score 31.2. Moreover, D-2-keto-3deoxygluconate (KDG) kinase (PDB code 1V19; RMSD = 2.3 Å for 261 Ca atoms; 26% identity, Z-score 30.1) and E. coli ribokinase (PDB code 1RKD; RMSD = 2.2 Å for 252 Ca atoms; 23% identity, Z-score 27.6) are also structurally similar. In addition, a recently deposited (but not yet described in the literature) structure of FRK homolog from Bacteroides thetaiotaomicron VPI-5482 (PDB code 2QHP; RMSD = 2.2 Å for 247 Ca atoms; 17% identity, Z-score 28.5) is found to be structurally related. In several cases, the individual domains of Hore_18220 and ribokinase homologs superpose well, however, the relative disposition of the domains often varies (Fig. 2A). The structure-based sequence alignment of Hore_18220 with the abovementioned homologs from ribokinase family showed that conservation of the residues is predominantly in the substrate and nucleotide binding pockets (Fig. 2B). To analyze the relationship of Hore_18220 to other proteins we performed sequence database searches with PSI-BLAST (Altschul et al., 1997) and clustered the obtained sequences with CLANS (Frickey and Lupas, 2004). Hore_18220 groups together with members of the COG0524 family (annotated as ‘sugar kinases, ribokinases’) from the Cluster of Orthologous Groups (COG) database, to the exclusion of various paralogs (members of other COGs) that exhibit other functions. Within this family, Hore_18220 belongs to a small group of prokaryotic proteins with the mutual sequence similarity in the range of 57–60%. This group includes PH1459 from P. horikoshii found by the DALI search. Among proteins with known function, plant FRK enzymes are the most closely related to Hore_18220 and PH1459 (sequence similarity in the range of 25–32%). Thus, even though the substrate specificities of Hore_18220 and PH1459 remain unknown, these proteins appear to be potentially the closest structural templates to model the structure and mechanism of action of plant FRKs. 5. Putative ATP and substrate binding pocket Although no position of ATP and substrate in the Hore_18220 structure was identified through crystallization, the binding sites
399
of ATP and the substrate can be predicted by comparison with structures of three known ribokinase family members complexed with ATP (PDB codes 1TZ6, 1V19 and 1RKD) (Fig. 3). To infer the catalytic mechanism of action for Hore_18220 (and, by extension, for plant FRK enzymes), we constructed a model of Hore_18220 complexed with ATP by remodeling the crystal structure of apoHore_18220 to mimic the conformation of bound substrates observed in the crystal structure of PH1459 complexed with ATP and E. coli ribokinase complexed with ribose and ADP (1RKD). We have also constructed a homology-model of a FRK from tomato (Solanum lycopersicum) and its complex with ATP and fructose (whose position was modeled by analogy to ribose in the 1RKD structure) (Fig 3). The superposition of related structures suggest that the overall architecture of the binding site is conserved among Hore_18220, AIR kinase (Zhang et al., 2004), KDG kinase (Ohshima et al., 2004), PH1459 and ribokinase (Sigrell et al., 1998) and these enzymes may interact with their ligands in a similar way. ATP is predicted to bind to a pocket on the catalytic domain of Hore_18220 lined by residues Asp181–Cys183; Lys210–Asp215; Thr243– Gly248; Ala267–Gly274; Leu297; Asn299; Val301 and Phe304. Similarly, the substrate binding pocket is predicted to be lined by residues Leu31, Asp33, Leu43, Gly54–Ser56, Asn59, Phe153 and Asp275, of which Leu31 and Asp33 are from the lid region. Thr243, Gly274 and Asn299 located in the ATP binding pocket are strictly conserved across all FRK homolog sequences compared (Fig. 2B). The equivalent residues of Thr243 and Gly274 in ribokinase, KDG kinase, PH1459, and AIR kinases were found to interact with the phosphate group of the bound ATP mainly through water mediated hydrogen bonds, while the equivalent residue of Asn299 interacts with the adenosine base. These residues in Hore_18220 would probably play a role in forming the oxyanion hole to stabilize the intermediate during phosphorylation as they bind to both substrate and ATP. Moreover, in comparison with ribokinases, the substrate and cofactor interacting residues show higher conservation among bacterial and plant FRKs. In particular, the residues lining the substrate and ATP binding pockets such as Pro182 (H. orenii numbering), Arg185, Val268, Asn299 are highly conserved in plant and
Fig. 1. Crystal structure of Hore_18220. (A) Shows the ribbon representation of the Hore_18220 monomer. (B) Ribbon diagram showing the dimeric Hore_18220 in the asymmetric unit. The catalytic domains of the monomers (residues Leu22–Ile30; Ser56–Thr108 and Ala128–Ile306) are depicted in blue and magenta and the b-sheet ‘‘lid” regions (residues Leu31–Gly55 and Thr109–Glu127) in red and cyan. The N- and C-terminals are labeled. These figures were prepared using the programs PYMOL (DeLano, 2008).
400
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
Fig. 2. Structural and sequence comparison of Hore_18220. (A) Stereo view of the superposition of Hore_18220 and its homologs. Hore_18220 (cyan) and the structures of the five ribokinase family members (protein–ligand complexes) superimposed on the catalytic domain of Hore_18220. Colored lines represent the Ca trace of AIR kinase (PDB code 1TZ6, green), RDK kinase (PDB code 1V1B, yellow), RK (PDB code 1RKD, magenta), PH1459 (PDB code 3EWM, blue) and putative 5-dehydro-2-deoxygluconokinase (PDB code 2QCV, orange). Substrates in the various structures are represented by sticks. This figure was prepared using the programs PYMOL (DeLano, 2008). (B) Structural and sequence alignment of Hore_18220. Top six rows: Structure-based sequence alignment of Hore_18220 (black), SKK (PDB code 3EWM, blue), DGK (PDB code 2QCV, cyan), ARK (PDB code 1TZ6, green), KDK (PDB code 1V19, brown) and RKK (PDB code 1RKD, magenta). The amino acids are in one-letter codes; the conserved residues are in red. Strictly conserved residues are highlighted red (identity above 48%). Secondary structural elements of Hore_18220 belonging to the a/b domain and the b ‘‘lid” are shown in blue and red, respectively. This figure was created using the program BioEdit (Hall, 1999). Middle 7–10th rows: Sequence alignment of Hore_18220 (top, black) with the closest four bacterial FRK homologs (orange) was carried out using CLUSTAL_X (Thompson et al., 1997) and refined manually. Bottom 11–14th rows: Sequence alignment of Hore_18220 (top, black) with the closest four plant FRK homologs (black). The anion hole motif GAGD is indicated by magenta asterisks. Proposed key substrate binding residues of fructose and ATP are indicated by blue and red asterisks, respectively. Suffix: FRK_Ho: Hore_18220 (FRK, H. orenii homolog); SKK_Ph: Sugar Kinase, Pyrococcus horikoshii (3EWM); DGK_Bh: 5-dehydro-2-deoxygluconokinase, Bacillus halodurans (2QCV); ARK_Se: Aminoimidazole riboside kinase, Salmonella enterica (1TZ6); KDK_Tt: 2Keto-3-Deoxygluconate Kinase, Thermus thermophilus (1V19); RKK_Ec: Ribokinase, Escherichia coli (1RKD); FRK_Pm: FRK, Petrotoga mobilis SJ95; FRK_Pd: FRK, Polaribacter dokdonensis; FRK_Cs: FRK, Cellulophaga sp. MED134; FRK_Pt: FRK, Psychroflexus torquis ATCC 700755; FRK_Sl: FRK, Solanum lycopersicum; FRK_At: FRK, Arabidopsis thaliana; FRK_Zm: FRK, Zea mays; FRK_Bv: FRK, Beta vulgaris.
T.K. Chua et al. / Journal of Structural Biology 171 (2010) 397–401
401
Protein Data Bank accession code Coordinates and structure factors for the Hore_18220 have been deposited with RCSB Protein Data Bank (PDB) with code 3HJ6. Acknowledgments
Fig. 3. Computationally modeled FRK structure from Solanum lycopersicum (tomato) with Fructose and ATP. Protein backbone is shown as a ribbon, conserved substrate binding residues, the cofactor and fructose are shown as sticks.
Data for this study were measured at beamlines X12C and X29 of the National Synchrotron Light Source, Brookhaven National Laboratories (BNL). We thank Dr. Anand Saxena (BNL) for assistance in data collection. J.S. acknowledges research support from the Academic Research Fund (ARF) Grant No. R154000245112, National University of Singapore (NUS). J.M.K and J.M.B were supported by the Polish Ministry of Science (Grant 188/N-DFG/2008/0) and by the NIH (Grant 1R01 GM081680-01). BKCP acknowledges the support from the Griffith University Research Grants scheme. We thank Mr. Sun Qingxiang for his assistance in the final refinement of the model. Chua Teck Khiang is a Ph.D. student in receipt of a research scholarship from the National University of Singapore (NUS). References
Hore_18220, respectively. This suggests that these residues are mostly required to create the appropriate binding pockets in both plant and Hore_18220. Our analysis of structure and sequence conservation between Hore_18220 and ribokinases (RK) suggest that the substrate binding pocket in Hore_18220 and its close homologs (including plant FRK enzymes) is lined up with smaller side chains to provide more space for large substrates. For instance in RK residues Gly41 to Gly55 (corresponding to Gly54 to Gly68 in Hore_18220) directly interact with ribose (in RK) and the corresponding region is highly conserved in both enzyme families. However, Lys43 and Gly44 in RK are replaced by Ala and Pro in plant FRK and Ser and Pro in Hore_18220. A large positively charged amino acid (Lys43 in RK) is replaced by a small, uncharged residue (Ala or Ser) in Hore_18220. Similarly Gln47 of RK is replaced by small hydrophobic residues in FRK homologs, i.e. Ile in H. orenii and Val in plants. It is further observed that the lid region (Leu31–Gly55; Thr109–Glu127) is highly conserved in both bacterial and plant FRK homologs, whereas this region is not conserved among the ribokinases. This region in FRK homologs might be responsible for the regulation of the open/ closed state of the binding site. Finally, based on the model of the tomato FRK complexed with ligands, we predict that Glu33, Asp37, Ala60, Asn63, Ile117, Arg192 and Asp285 of the plant enzyme may be particularly important for the binding of the substrate (Fig 3). In summary, we have determined the structure of a FRK homolog from H. orenii. The plant FRK proteins present difficulties for purification and characterization (Medina and Sols, 1956; Frankart and Pontis, 1976). Our analysis presented in this article illustrates similarities and differences between sugar kinases acting on different substrates (e.g. ribose vs. fructose) and provides a working model towards studying the mechanism of plant FRKs. In particular: (1) Hore_18220 belongs to a family of bacterial proteins that are the closest prokaryotic homologs of plant FRKs. (2) The structure of Hore_18220 was used as a template to homology-model the structure of tomato FRK, in complex with its substrates (ATP and fructose), which helped predict fructosebinding residues in the plant enzyme. (3) Residues in the putative substrate binding pocket of Hore_18220 are conserved in plant FRKs; they have smaller side chains than corresponding residues in ribokinases, which makes the Hore_18220 and FRK pocket larger and potentially allows for accommodation of larger substrates.
Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., Read, R.J., Rice, L.M., Simonson, T., Warren, G.L., 1998. Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D 54, 905–921. Chaubron, F., Harris, N., Ross, H.A., Davies, H.V., 1995. Partial purification and characterization of fructokinase from developing taproots of sugar beet (Beta vulgaris). Plant Sci. 110, 181–186. DeLano, W.L., 2008. The PyMOL Molecular Graphics System. DeLano Scientific LLC, Palo Alto, CA, USA. Frankart, W.A., Pontis, H.G., 1976. Fructose metabolism in plants. Isolation and properties of pea seed frucktokinase. Acta Physiol. Latinoam. 26, 319–329. Frickey, T., Lupas, A., 2004. CLANS: a Java application for visualizing protein families based on pairwise similarity. Bioinformatics 20, 3702–3704. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41, 95–98. Holm, L., Sander, C., 1993. Protein structure comparison by alignment of distance matrices. J. Mol. Biol. 233, 123–138. Jones, T.A., Zou, J.Y., Cowan, S.W., Kjeldgaard, M., 1991. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, 110–119. Laskowski, R.A., McArthur, M.W., Moss, D.S., Thornton, J.M., 1993. Procheck: a program to check the stereochemical quality of protein structures. J Appl. Crystallogr. 26, 282–291. Martinez-Barajas, E., Luethy, M.H., Randall, D.D., 1997. Molecular cloning and analysis of fructokinase expression in tomato (Lycopersicon esculentum Mill.). Plant Sci. 125, 13–20. Medina, A., Sols, A., 1956. Specific fructokinase in peas. Biochim. Biophys. Acta 19, 378–379. Mijts, B.N., Patel, B.K., 2001. Random sequence analysis of genomic DNA of an anaerobic, thermophilic, halophilic bacterium, Halothermothrix orenii. Extremophiles 5, 61–69. Ohshima, N., Inagaki, E., Yasuike, K., Takio, K., Tahirov, T.H., 2004. Structure of Thermus thermophilus 2-Keto-3-deoxygluconate kinase: evidence for recognition of an open chain substrate. J. Mol. Biol. 340, 477–489. Otwinowski, Z., Minor, W., 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–326. Renz, A., Stitt, M., 1993. Substrate specificity and product inhibition of different forms of fructokinase and hexokinases in developing potato tubers. Planta 190, 166–175. Sigrell, J.A., Cameron, A.D., Jones, T.A., Mowbray, S.L., 1998. Structure of Escherichia coli ribokinase in complex with ribose and dinucleotide determined to 1.8 Å resolution: insights into a new family of kinase structures. Structure 6, 183–193. Terwilliger, T.C., 2000. Maximum-likelihood density modification. Acta Crystallogr. D 56, 965–972. Terwilliger, T.C., Berendzen, J., 1999. Automated MAD and MIR structure solution. Acta Crystallogr. D 55, 849–861. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882. Zhang, S.R., Nichols, S.E., Dong, J.G., 2003. Cloning and characterization of two fructokinases from maize. Plant Sci. 165, 1051–1058. Zhang, Y., Dougherty, M., Downs, D.M., Ealick, S.E., 2004. Crystal Structure of an aminoimidazole riboside kinase from Salmonella enterica: implications for the evolution of the ribokinase superfamily. Structure 12, 1809–1821.