X-ray Structure Determination of Trypanosoma brucei Ornithine Decarboxylase Bound to d-Ornithine and to G418
2003; Elsevier BV; Volume: 278; Issue: 24 Linguagem: Inglês
10.1074/jbc.m300188200
ISSN1083-351X
AutoresLaurie K. Jackson, Elizabeth J. Goldsmith, Margaret A. Phillips,
Tópico(s)Enzyme Structure and Function
ResumoOrnithine decarboxylase (ODC) is a pyridoxal 5′-phosphate (PLP)-dependent enzyme that catalyzes the rate-determining step in the biosynthesis of polyamines. ODC is a proven drug target to treat African sleeping sickness. The x-ray crystal structure of Trypanosoma brucei ODC in complex with d-ornithine (d-Orn), a substrate analog, and G418 (Geneticin), a weak non-competitive inhibitor, was determined to 2.5-Å resolution. d-Orn forms a Schiff base with PLP, and the side chain is in a similar position to that observed for putrescine and α-difluoromethylornithine in previous T. brucei ODC structures. The d-Orn carboxylate is positioned on the solvent-exposed side of the active site (si face of PLP), and Gly-199, Gly-362, and His-197 are the only residues within 4.2 Å of this moiety. This structure confirms predictions that the carboxylate of d-Orn binds on the si face of PLP, and it supports a model in which the carboxyl group of the substrate l-Orn would be buried on the re face of the cofactor in a pocket that includes Phe-397, Tyr-389, Lys-69 (methylene carbons), and Asp-361. Electron density for G418 was observed at the boundary between the two domains within each ODC monomer. A ten-amino acid loop region (392–401) near the 2-fold axis of the dimer interface, which contributes several residues that form the active site, is disordered in this structure. The disordering of residues in the active site provides a potential mechanism for inhibition by G418 and suggests that allosteric inhibition from this site is feasible. Ornithine decarboxylase (ODC) is a pyridoxal 5′-phosphate (PLP)-dependent enzyme that catalyzes the rate-determining step in the biosynthesis of polyamines. ODC is a proven drug target to treat African sleeping sickness. The x-ray crystal structure of Trypanosoma brucei ODC in complex with d-ornithine (d-Orn), a substrate analog, and G418 (Geneticin), a weak non-competitive inhibitor, was determined to 2.5-Å resolution. d-Orn forms a Schiff base with PLP, and the side chain is in a similar position to that observed for putrescine and α-difluoromethylornithine in previous T. brucei ODC structures. The d-Orn carboxylate is positioned on the solvent-exposed side of the active site (si face of PLP), and Gly-199, Gly-362, and His-197 are the only residues within 4.2 Å of this moiety. This structure confirms predictions that the carboxylate of d-Orn binds on the si face of PLP, and it supports a model in which the carboxyl group of the substrate l-Orn would be buried on the re face of the cofactor in a pocket that includes Phe-397, Tyr-389, Lys-69 (methylene carbons), and Asp-361. Electron density for G418 was observed at the boundary between the two domains within each ODC monomer. A ten-amino acid loop region (392–401) near the 2-fold axis of the dimer interface, which contributes several residues that form the active site, is disordered in this structure. The disordering of residues in the active site provides a potential mechanism for inhibition by G418 and suggests that allosteric inhibition from this site is feasible. Ornithine decarboxylase (ODC) 1The abbreviations used are: ODC, ornithine decarboxylase; Orn, ornithine; PLP, pyridoxal 5′-phosphate; put, putrescine; AZ, antizyme; DFMO, α-difluoromethylornithine; r.m.s.d., root mean square deviation. is a pyridoxal 5′-phosphate (PLP)-dependent enzyme that catalyzes the rate-limiting step in the biosynthesis of polyamines, the decarboxylation of l-Orn to form the diamine putrescine (1Tabor C.W. Tabor H. Microbiol. Rev. 1985; 49: 81-98Crossref PubMed Google Scholar). Polyamines are required for cell proliferation. α-Difluoromethylornithine (DFMO), a suicide inhibitor of ODC, is an approved drug for the treatment of African sleeping sickness caused by the parasitic protozoa Trypanosoma brucei (2Wang C.C. Annu. Rev. Pharmacol. Toxicol. 1995; 35: 93-127Crossref PubMed Google Scholar, 3Kuzoe F.A.S. Acta Trop. 1993; 54: 153-162Crossref PubMed Scopus (137) Google Scholar). It has also been investigated as a potential drug for the treatment of cancer and other proliferative diseases (reviewed in Refs. 4Pegg A.E. Shantz L.M. Coleman C.S. J. Cell Biol. 1995; 22: 132-138Google Scholar and 5McCann P.P. Pegg A.E. Pharmacol. Ther. 1992; 54: 195-215Crossref PubMed Scopus (229) Google Scholar). X-ray structural analysis of the native enzymes from mouse (6Kern A.D. Oliveira M.A. Coffino P. Hackert M. Structure. 1999; 7: 567-581Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar), T. brucei (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar), and human (9Almrud J.J. Oliveira M.A. Grishin N.V. Phillips M.A. Hackert M.L. J. Mol. Biol. 2000; 295: 7-16Crossref PubMed Scopus (128) Google Scholar) demonstrates that the ODC monomer consists of two domains, an N-terminal β/α-barrel domain and a C-terminal β-sheet domain. The T. brucei structure has also been solved in complex with the product putrescine and with DFMO, identifying important active site residues (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar). ODC is an obligate homodimer, and the two identical active sites are formed at the dimer interface, encompassing the β/α–barrel domain from one subunit and β–sheet domain from the other. Catalytic residues are contributed to the active site from both monomers. Both putrescine and DFMO form a covalent Schiff base complex with the PLP cofactor, and the side-chain amino groups of both compounds bind in a well conserved pocket that bridges between subunits. In addition DFMO has undergone decarboxylation and has formed a covalent bond to Cys-360 as expected based on its mechanism of inhibition (10Poulin R. Lu L. Ackerman B. Bey P. Pegg A.E. J. Biol. Chem. 1992; 267: 150-158Abstract Full Text PDF PubMed Google Scholar). The ODC active site is invariant between the host and parasite enzymes; thus, selective toxicity of DFMO for the parasite is not achieved through differential inhibitor binding to the enzyme active site (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar). Instead, it is thought to arise from metabolic differences between the host and parasite, which include the rapid turnover of host ODC when compared with that of the parasite (2Wang C.C. Annu. Rev. Pharmacol. Toxicol. 1995; 35: 93-127Crossref PubMed Google Scholar, 11Phillips M.A. Coffino P. Wang C.C. J. Biol. Chem. 1987; 262: 8721-8727Abstract Full Text PDF PubMed Google Scholar), and the requirement for the polyamine spermidine for the formation of the novel cofactor trypanothione to maintain reduced thiol pools in the parasite (12Fairlamb A.H. Le Quesne S.A. Hide G. Mottram J.C. Coombs G.H. Holmes P.H. Trypanosomiasis and Leishmaniasis. CAB International, Wallingford, Oxon, United Kingdom1997: 149-161Google Scholar). The discovery that DFMO was curative against T. brucei gambiense infections provided the first alternative to the highly toxic melarsoprol for the treatment of late stage disease (13Bacchi C.J. Nathan H.C. Hunter S.H. Science. 1980; 210: 332-334Crossref PubMed Scopus (309) Google Scholar). However, utilization of DFMO is hampered by the large dose requirement and by poor activity against the T. brucei rhodesiense strain of the parasite (2Wang C.C. Annu. Rev. Pharmacol. Toxicol. 1995; 35: 93-127Crossref PubMed Google Scholar). A number of small molecule inhibitors of ODC have been reported, but like DFMO they target the highly conserved active site (14Danzen C. Casasra P. Claverie N. Metcalf B.W. J. Med. Chem. 1981; 24: 16-20Crossref PubMed Scopus (52) Google Scholar, 15Bey P. Danzin C. Dorsselaer V. Mamont P. Jung M. Tardif C. J. Med. Chem. 1978; 21: 50-55Crossref PubMed Scopus (85) Google Scholar, 16Bitonti A.J. Bacchi C.J. McCann P.P. Sjoerdsma A. Biochem. Pharmacol. 1985; 34: 1773-1777Crossref PubMed Scopus (62) Google Scholar). The discovery of novel ODC inhibitors with chemical properties that differ from substrate may overcome the poor pharmacology of DFMO. We had observed previously (17Myers D.P. Jackson L.K. Ipe V.G. Murphy G.E. Phillips M.A. Biochemistry. 2001; 40: 13230-13236Crossref PubMed Scopus (36) Google Scholar) that mutations in the dimer interface of ODC distant from the active site decreased catalytic activity. Thus we sought to explore whether ODC could be inhibited allosterically by binding inhibitors to sites outside of the active site. We have identified an inhibitor of ODC, G418 sulfate (Geneticin), which is non-competitive with substrate. Although the inhibitor has a high KI (5–8 mm) it was sufficiently inhibitory to allow us to observe d-Orn bound in the active site of ODC. Previous attempts to co-crystallize d-Orn and ODC were unsuccessful, because the substrate analog was unexpectedly decarboxylated to putrescine (8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar). Through crystallographic analysis of T. brucei ODC, G418, and d-Orn, we have determined the binding orientation of a carboxylated ligand. The structure reported here, in combination with the DFMO-bound structure and with mutagenic analysis, supports a model where the substrate carboxylate is buried on the re face of PLP on the same face as Lys-69. The inhibitor G418 binds outside the active site of ODC in the interface between the two domains of the ODC monomer. A loop region (residues 392–401) at the symmetry-related center of the dimer interface, which contributes essential residues to the active site pocket, is disordered in this structure. The disorder in this region of the active site may explain the basis for inhibition of ODC by G418. This conformation, apparently available at low energy, is inactive and could be useful in inhibitor design. G418 sulfate (Geneticin) was purchased from Invitrogen (catalogue numbers 55-0273 and 11811-023), and Cytoscint was purchased from ICN Biomedicals, Inc. ODC Expression and Purification—ODC was expressed from a T7 promoter in the BL21 (DE3) strain of Escherichia coli as described for T. brucei (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar, 18Osterman A.L. Grishin N.V. Kinch L.N. Phillips M.A. Biochemistry. 1994; 33: 13662-13667Crossref PubMed Scopus (73) Google Scholar) and human (9Almrud J.J. Oliveira M.A. Grishin N.V. Phillips M.A. Hackert M.L. J. Mol. Biol. 2000; 295: 7-16Crossref PubMed Scopus (128) Google Scholar) ODC. The protein was produced as an N-terminal His6-tagged fusion that included a tobacco etch virus (TEV) protease cleavage site to allow removal of the tag. Protein was purified by column chromatography over nickel-nitrilotriacetic acid-agarose and Sephadex 200. For crystallography, the His tag was removed from purified ODC by digestion with tobacco etch virus (TEV) protease immobilized on glutathione-agarose beads as described (19Grishin N.V. Osterman A.L. Goldsmith E.J. Phillips M.A. Proteins. 1996; 24: 272-273Crossref PubMed Scopus (14) Google Scholar). The purified protein for crystallization experiments was concentrated to 20–25 mg/ml. Site-directed Mutagenesis—D243A, L339A, and E384A mutants of T. brucei ODC were created using PCR-based mutagenesis of the pODC29 expression plasmid using the QuikChange™ site-directed mutagenesis kit from Stratagene (La Jolla, CA). The primers listed below contain the desired mutation (bold), and D243A and L339A also contain silent mutations (bold italics), which remove a restriction site (BseDI and StuI, respectively) for diagnostic purposes: L339A forward, 5′-GACCACGCAGTCGTCAGACCTGCGCCCCAGAGGGAGCC-3′; D243A forward, 5′-GTGGGTTTCCAGGTACGAGGGCTGCACCACTTAAATTTG3′; E384A forward, 5′-GGCTGCTCTTTGCGGATATGGGTGCC-3′. Inhibitor Identification—G418 and other aminoglycosides were identified as potential ODC inhibitors by computational methods using the Available Chemicals Directory (1998 version). Enzyme Assays—Steady-state spectrophotometric assay of the decarboxylation of l-Orn (0.1 to 16 mm) by ODC (20–80 nm) was followed at 37 °C as described (18Osterman A.L. Grishin N.V. Kinch L.N. Phillips M.A. Biochemistry. 1994; 33: 13662-13667Crossref PubMed Scopus (73) Google Scholar) using the Sigma Diagnostics carbon dioxide detection kit (Sigma) in the presence of 50 μm PLP. In this assay CO2 production is linked to the oxidation of NADH (λmax·NADH = 340 nm) by coupling the ODC reaction to phosphoenolpyruvate carboxylase and malate dehydrogenase. Direct assay of l-Orn decarboxylation by ODC was performed by a radioactive assay of 1-14C-l-Orn. For this assay, a 2× substrate solution, 1-14C-l-Orn (32 μm) mixed with cold l-Orn (0.8 mm) in Buffer A (0.2 m Hepes, pH 7.5, 0.05 mm PLP), was added to a culture tube by injection with a Hamilton syringe through a rubber stopper cap into a 2× ODC enzyme solution (0.25 μm ODC in Buffer A). After mixing, the reaction was allowed to proceed at 37 °C for varied times (0 to 8 min) and then quenched with 20% (final) trichloroacetic acid. Sample chambers were incubated for another hour at 37 °C to evolve liberated 1-14CO2 onto paper strips soaked with a saturated Ba(OH)2 solution that were hung in the chamber. The paper strips were submerged in Cytoscint scintillation mixture and counted for radioactive decay. The KI and IC50 values for the inhibition of ODC by G418 sulfate were determined by spectroscopic assay over a range of inhibitor concentrations (0–16 mm). The inhibitors (at 1×) were preincubated with ODC (at 50×) in assay buffer for 15 min at 37 °C before addition to the assay mixture. The assay mixture was also allowed to pre-incubate for 5 min with G418 before the addition of pre-incubated enzyme. Salt controls (NaCl and KSO4) were run in the same concentration range as inhibitor. IC50 values were determined by fitting to Equation 1.νiνo=11+[I]IC50(Eq. 1) The KI values were determined by varying both l-Orn and inhibitor concentration, and KI values for the non-competitive inhibitors were determined by fitting to Equation 2.ν=Vmax[S](Km+[S])(1+[1]KI)(Eq. 2) Kinetic data were fitted to the appropriate equation with Sigma Plot 2001 version 7.0 (SPSS Inc., Chicago, IL) to determine the kinetic constants. Crystallization—T. brucei ODC was co-crystallized with d-Orn and G418 sulfate under the following conditions: T. brucei ODC (23 mg/ml in 10 mm Hepes, pH 7.2, 50 mm NaCl, 10 mm dithiothreitol, 0.5 mm EDTA, 0.01% Brij-20) was preincubated with 25 mm d-Orn and 100 mm G418 sulfate for 10 min at room temperature before setting crystal drops. Equal volumes of preincubated enzyme and well solution (20% polyethylene glycol 3350, 100 mm Hepes, pH 8.0, 10 mm dithiothreitol, 25 mm d-Orn, pH 7.5, 100 mm G418) were used to form the crystal drops. γ-Butyrolactone, 3% (v/v) final, was added to the drops (final pH = 7.0). The crystals formed overnight at 16 °C under conditions of vapor diffusion. Crystals were cryoprotected in 1.2 × crystal conditions with 20% ethylene glycol and frozen in propane. These conditions differed from those used previously (8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar) to obtain the putrescine bound structure in which the γ-butyrolactone additive was not required, and the well solution contained NaOAc in place of G418 (20% polyethylene glycol 3350, 100 mm Hepes, pH 7.5, 10 mm dithiothreitol, 7.5 mm d-Orn, 200 mm NaOAc). Data Collection and Processing—Diffraction data were collected with an R-Axis IV image plate system (MSC, Houston, TX) mounted on a rotating anode and operated at 100 mA and 50 kV, at the University of Texas Southwestern Medical Center. Data were processed with HKL2000 (20Otwinoski A. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref Scopus (38572) Google Scholar). A summary of the statistics for data processing is given in Table I. There was a slight asymmetry in the resolution of the data and also a telescoping effect in the spot shape. After screening many crystals, data adequate for structure determination were collected. Wilson B-values were determined using CCP4 (21Collaborative Computational Project, Number 4 Acta Cryst. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19769) Google Scholar). 2E. Dodson and P. Wilson. WILSON, a CCP4 supported program. (See Ref. 21Collaborative Computational Project, Number 4 Acta Cryst. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19769) Google Scholar.)Table IData collection, processing and refinement statistics for the determination of the T. brucei ODC structure in complex with G418 and D-OrnData collection and processingData sourceR-axis IV image plateWavelength (Å)1.54Temperature (K)-175 °CResolution (Å)2.45Total reflections253681Unique reflections63372Completeness (%) (last shell 1 > σ)96.7 (93.9)Multiplicity4.0Intensities I/σ (last shell)21.5 (2.1)Chi squared1.206Rmerge0.069Wilson temperature factor52RefinementNo. of reflections (F > 2σ)49323 (35-2.45)No. of non-H protein atoms11020No. of water molecules207Rcrystal (%)/Rfree (%) (F > 2σ)26.0/28.3 (35-2.45)r.m.s.d. bonds0.0087r.m.s.d. bond angles1.559Average B-values (Å2)45B r.m.s.d. for bonded main-chain atoms1.453B r.m.s.d. for bonded side-chain atoms1.720 Open table in a new tab Structure Determination and Refinement—Initial phases were calculated by molecular replacement using the program AMoRe (23Navaza J. Acta Crystallogr. Sect. A. 1994; 50: 157-163Crossref Scopus (5029) Google Scholar) and the coordinates of a dimer from the T. brucei ODC/put structure (8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar) (Protein Data Bank code 1f3t), with putrescine and waters removed, as a search model. The model was built in O version 6.2.2 (24Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Cryst. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13011) Google Scholar) and refined using CNS (25Brunger A.T. Adams P.D. Clre G.M. Delano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges N. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16967) Google Scholar) utilizing data collected between 35 and 2.45 Å. Non-crystallographic symmetry constraints were kept for residues 38–152, 170–291, and 317–390 for initial rounds of refinement, non-crystallographic symmetry constraints were released for one round, and then restraints were added again for residues 38–152, 170–238, 247–291, 317–340, and 350–390 for the remainder of the refinement. The peptide torsion angles for 1011 of 1180 non-glycine and non-proline residues fall within the most favored regions, and there are none found in disallowed regions of the Ramachandran plot as determined by the program PROCHECK (26Laskowski R.A. MacArthur M.W. Moss D.S. Thornton J.M. J. Appl. Crystallogr. 1993; 26: 283-291Crossref Google Scholar). Water molecules were added using CNS and edited in the program O. Refinement statistics are listed in Table I. The coordinates have been placed in the Protein Data Bank (Protein Data Bank code 1NJJ). Superposition of Crystal Structures and Determination of r.m.s.d. between Structure Coordinates—C monomers from each of the following T. brucei ODC structures, native ODC (Protein Data Bank code 1QU4; native_ODC), K69A ODC inactivated with DFMO (Protein Data Bank code 2tod; K69_ODC/DFMO), and the structure reported in this paper (ODC/G418/d-Orn), were superimposed with the structure of ODC in complex with putrescine (Protein Data Bank code 1f3t; ODC/put) using Insight II (Accelrys, San Diego, CA). Monomers from each structure were pared down to include only shared residues 37–68, 70–157, 168–296, 312–344, 348–391, and 402–409, and backbone atoms from residues 320–340 and 350–380 (from the β-sheet domain) were superimposed. Superimposed monomers were written out into the same coordinate system using the Biopolymer module of InsightII, and CNS was used to determine the average r.m.s. coordinate difference between structures for residues 45–282 (from the βα-barrel domain) and for residues in the β-sheet domain (38–44, 283–342, and 349–409), using the CNS rmsd.inp input file. Analysis of the Inhibition of T. brucei ODC Activity by G418 Sulfate—G418 sulfate is a weak non-competitive inhibitor of ODC with respect to l-Orn. G418 has a KI of8mm at pH 7.0 and 4.5 mm at pH 8.0. Pre-incubation of ODC with G418 to establish equilibrium before assaying for activity was required to observe constant inhibition suggesting a slow binding step or a slow conformational change promoted by G418. Inhibition was observed using both the enzyme-linked and the direct, radioactive decarboxylation assays. Because primary amines are able to form a Schiff base with the PLP cofactor, the observed noncompetitive inhibition could result from to the sequestration of PLP by the high concentration of G418 (an amine) used. To test for this possibility, we assayed for inhibition over a range of PLP concentrations (10–200 μm), and the inhibition remained constant, suggesting that PLP sequestration is not the mechanism of inhibition. Controls containing equivalent concentrations of the inhibitor counter ions (NaCl or KSO4) were also performed. No inhibition was seen under these conditions. Similar results for inhibition of human ODC by G418 were obtained. Crystallization, Data Collection, and Structure Determination—T. brucei ODC was co-crystallized with d-Orn and G418. The ODC/G418/d-Orn crystal diffracted to 2.45 Å in the space group P21 (a = 67.8 Å, b = 88.5 Å, c = 150.4 Å, β = 90.03°) with two homodimers in the asymmetric unit. The final model contains four monomers, 1404 residues, four PLP, four d-Orn, three G418, and 207 water molecules. The R-factor is 26.0% with an Rfree of 28.3% for data between 35- and 2.45-Å resolution (Table I). Representative electron density for the PLP/d-Orn and G418 binding sites is displayed in Fig. 1. Density for most of the structure was unambiguous with the exception that residues at the N terminus (amino acid residues 1–19), the C terminus (410–425), and three surface loops (31–36, 158–167, and 297–311) were disordered. Residues 343–348 have poorly defined density as observed previously (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar) for all T. brucei ODC structures. Chain specific differences include the following: chain A has an additional six residues (410–414) at the C terminus, chain C contains density for additional residues at the N-terminal end (14Danzen C. Casasra P. Claverie N. Metcalf B.W. J. Med. Chem. 1981; 24: 16-20Crossref PubMed Scopus (52) Google Scholar, 15Bey P. Danzin C. Dorsselaer V. Mamont P. Jung M. Tardif C. J. Med. Chem. 1978; 21: 50-55Crossref PubMed Scopus (85) Google Scholar, 16Bitonti A.J. Bacchi C.J. McCann P.P. Sjoerdsma A. Biochem. Pharmacol. 1985; 34: 1773-1777Crossref PubMed Scopus (62) Google Scholar, 17Myers D.P. Jackson L.K. Ipe V.G. Murphy G.E. Phillips M.A. Biochemistry. 2001; 40: 13230-13236Crossref PubMed Scopus (36) Google Scholar, 18Osterman A.L. Grishin N.V. Kinch L.N. Phillips M.A. Biochemistry. 1994; 33: 13662-13667Crossref PubMed Scopus (73) Google Scholar, 19Grishin N.V. Osterman A.L. Goldsmith E.J. Phillips M.A. Proteins. 1996; 24: 272-273Crossref PubMed Scopus (14) Google Scholar) and for residue 392, and chain D has additional density observed for residues 309–311 and for the backbone atoms of residues 400 and 401. In addition to those regions typically disordered in T. brucei ODC crystals, residues Val-392 to Gln-401, near the active site, were also disordered in the ODC/G418/d-Orn structure (Fig. 2). These residues were well ordered in the previous crystallographic studies of T. brucei ODC with B-factors in the region being comparable with the average B-factor for the overall structure (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar). Residues 392–401 are in the dimer interface, adjacent to the equivalent loop from the opposing monomer, and contribute residues that form the back of the active site pocket (Fig. 3A). Several of these residues (Phe-397, Asn-398, and Phe-400) have been demonstrated to be important for enzyme activity (17Myers D.P. Jackson L.K. Ipe V.G. Murphy G.E. Phillips M.A. Biochemistry. 2001; 40: 13230-13236Crossref PubMed Scopus (36) Google Scholar, 27Jackson L.K. Brooks H.B. Myers D.P. Phillips M.A. Biochemistry. 2003; 42: 2933-2940Crossref PubMed Scopus (30) Google Scholar). Although the space group is the same as with previous T. brucei ODC structures (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar), the lattice and packing differ. Neither the inhibitor binding site nor the disordered loop appear to be influenced by the lattice.Fig. 3Stereo views of the d-Orn binding site.A, selected active site residues from the ODC/G418/d-Orn structure (blue carbon atoms) are superimposed with corresponding residues from the ODC/put structure (green carbon atoms) (Protein Data Bank code 1f3t). The ribbon and dark green residues denote the loop regions from the ODC/put structure that have become disordered in the ODC/G418/d-Orn structure. PLP phosphate atoms are purple. B, the d-Orn carboxyl binding pocket. Selected active site residues from the ODC/G418/d-Orn structure (blue carbon atoms) are displayed. PLP and d-Orn carbon atoms are gray. The position for Leu-166, which is disordered in the ODC/G418/d-Orn structure, is taken from the ODC/put structure and displayed with green carbon atoms.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The d-Orn Binding Site—This structure allowed for the orientation of the substrate stereoisomer, d-Orn, to be determined. d-Orn binds ODC with a Kd of 0.27 mm, similar to the binding constant for putrescine (8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar, 28Osterman A.L. Brooks H.B. Jackson L.K. Abbott J.J. Phillips M.A. Biochemistry. 1999; 38: 11814-11826Crossref PubMed Scopus (46) Google Scholar). During kinetic analysis, d-Orn is not decarboxylated at a detectable rate (kcat/Km < 1 × 10–6 s–1m–1 at 37 °C). However, during the extensive incubation between d-Orn and wild-type T. brucei ODC required for crystallization, d-Orn is converted to putrescine, as observed in the ODC/put structure (8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar). In contrast, in the ODC/G418/d-Orn structure, density for the carboxylate is observed (Fig. 1A). The d-Orn carboxyl group is positioned on the si face of the PLP cofactor, on the solvent-exposed face of the active site (see Fig. 1A and Fig. 3B). Gly-199, Gly-362, and the His-197 side chains are the only contacts within 4.2 Å of the d-Orn carboxyl group. d-Orn forms an external aldimine with the PLP cofactor, and the δ-amino group interacts with Asp-361 and Asp-332 (Fig. 3A). The position for the d-Orn δ-amino group is similar to that observed for ligands in the putrescine and DFMO complexed ODC structures (7Grishin N.V. Osterman A.L. Brooks H.B. Phillips M.A. Goldsmith E.J. Biochemistry. 1999; 38: 15174-15184Crossref PubMed Scopus (137) Google Scholar, 8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar). Lys-69, which forms a Schiff base with PLP in the unliganded structure, has rotated out of the active site to form interactions with Asp-88 and Glu-94, as observed in the structure of T. brucei ODC bound to putrescine (8Jackson L.K. Brooks H.B. Osterman A.L. Goldsmith E.J. Phillips M.A. Biochemistry. 2000; 39: 11247-11257Crossref PubMed Scopus (77) Google Scholar). Differences in the active site structure are also observed. In previous structures residues 158–165, which are located near the active site entrance, are disordered. In addition to these residues, Leu-166 becomes disordered in this structure, apparently displaced by the carboxylate of d-Orn. Cys-360 is also observed in a different position than for previous ligand-boun
Referência(s)