The Catalytic Domain of Escherichia coli Lon Protease Has a Unique Fold and a Ser-Lys Dyad in the Active Site
2004; Elsevier BV; Volume: 279; Issue: 9 Linguagem: Inglês
10.1074/jbc.m312243200
ISSN1083-351X
AutoresIstvan Botos, Edward E. Melnikov, Scott Cherry, Joseph E. Tropea, Anna G. Khalatova, Fatima Rasulova, Zbigniew Dauter, Michael R. Maurizi, Т. В. Ротанова, Alexander Wlodawer, Alla Gustchina,
Tópico(s)Enzyme Structure and Function
ResumoATP-dependent Lon protease degrades specific short-lived regulatory proteins as well as defective and abnormal proteins in the cell. The crystal structure of the proteolytic domain (P domain) of the Escherichia coli Lon has been solved by single-wavelength anomalous dispersion and refined at 1.75-Å resolution. The P domain was obtained by chymotrypsin digestion of the full-length, proteolytically inactive Lon mutant (S679A) or by expression of a recombinant construct encoding only this domain. The P domain has a unique fold and assembles into hexameric rings that likely mimic the oligomerization state of the holoenzyme. The hexamer is dome-shaped, with the six N termini oriented toward the narrower ring surface, which is thus identified as the interface with the ATPase domain in full-length Lon. The catalytic sites lie in a shallow concavity on the wider distal surface of the hexameric ring and are connected to the proximal surface by a narrow axial channel with a diameter of ∼18 Å. Within the active site, the proximity of Lys722 to the side chain of the mutated Ala679 and the absence of other potential catalytic side chains establish that Lon employs a Ser679-Lys722 dyad for catalysis. Alignment of the P domain catalytic pocket with those of several Ser-Lys dyad peptide hydrolases provides a model of substrate binding, suggesting that polypeptides are oriented in the Lon active site to allow nucleophilic attack by the serine hydroxyl on the si-face of the peptide bond. ATP-dependent Lon protease degrades specific short-lived regulatory proteins as well as defective and abnormal proteins in the cell. The crystal structure of the proteolytic domain (P domain) of the Escherichia coli Lon has been solved by single-wavelength anomalous dispersion and refined at 1.75-Å resolution. The P domain was obtained by chymotrypsin digestion of the full-length, proteolytically inactive Lon mutant (S679A) or by expression of a recombinant construct encoding only this domain. The P domain has a unique fold and assembles into hexameric rings that likely mimic the oligomerization state of the holoenzyme. The hexamer is dome-shaped, with the six N termini oriented toward the narrower ring surface, which is thus identified as the interface with the ATPase domain in full-length Lon. The catalytic sites lie in a shallow concavity on the wider distal surface of the hexameric ring and are connected to the proximal surface by a narrow axial channel with a diameter of ∼18 Å. Within the active site, the proximity of Lys722 to the side chain of the mutated Ala679 and the absence of other potential catalytic side chains establish that Lon employs a Ser679-Lys722 dyad for catalysis. Alignment of the P domain catalytic pocket with those of several Ser-Lys dyad peptide hydrolases provides a model of substrate binding, suggesting that polypeptides are oriented in the Lon active site to allow nucleophilic attack by the serine hydroxyl on the si-face of the peptide bond. Rapid proteolysis plays a major role in post-translational cellular control by the targeted degradation of short-lived regulatory proteins and also serves an important function in protein quality control by eliminating defective and potentially damaging proteins from the cell (1Wickner S. Maurizi M.R. Gottesman S. Science. 1999; 286: 1888-1893Crossref PubMed Scopus (907) Google Scholar, 2Gottesman S. Wickner S. Maurizi M.R. Genes Dev. 1997; 11: 815-823Crossref PubMed Scopus (468) Google Scholar, 3Goldberg A.L. Eur. J. Biochem. 1992; 203: 9-23Crossref PubMed Scopus (415) Google Scholar, 4Gottesman S. Maurizi M.R. Microbiol. Rev. 1992; 56: 592-621Crossref PubMed Google Scholar). In all cells, protein degradation is predominantly carried out by ATP-dependent proteases, which are complex enzymes containing both ATPase and proteolytic activities expressed as separate domains within a single polypeptide chain or as individual subunits in complex assemblies. Five ATP-dependent proteases, Lon, FtsH, ClpAP, ClpXP, and HslUV, have been discovered in Escherichia coli, and homologous proteases have been found in all eubacteria and in many eukaryotes. Alone among them, Lon protease has been found in virtually all living organisms, from Archaea to eubacteria to humans.E. coli Lon protease was the first ATP-dependent protease to be identified (5Swamy K.H. Goldberg A.L. Nature. 1981; 292: 652-654Crossref PubMed Scopus (101) Google Scholar, 6Goldberg A.L. Swamy K.H. Chung C.H. Larimore F.S. Methods Enzymol. 1981; 80: 680-702Crossref PubMed Scopus (74) Google Scholar, 7Charette M.F. Henderson G.W. Markovitz A. Proc. Natl. Acad. Sci. U. S. A. 1981; 78: 4728-4732Crossref PubMed Scopus (134) Google Scholar, 8Goldberg A.L. Moerschell R.P. Chung C.H. Maurizi M.R. Methods Enzymol. 1994; 244: 350-375Crossref PubMed Scopus (181) Google Scholar). It is an oligomeric multidomain enzyme whose single polypeptide chain is composed of 784 amino acids (9Amerik A.Y. Antonov V.K. Ostroumova N.I. Rotanova T.V. Chistyakova L.G. Bioorgan. Khim. 1990; 16: 869-880PubMed Google Scholar). Comparison of the amino acid sequences of various members of the Lon family (9Amerik A.Y. Antonov V.K. Ostroumova N.I. Rotanova T.V. Chistyakova L.G. Bioorgan. Khim. 1990; 16: 869-880PubMed Google Scholar, 10Gottesman S. Wickner S. Jubete Y. Singh S.K. Kessel M. Maurizi M. Cold Spring Harb. Symp. Quant. Biol. 1995; 60: 533-548Crossref PubMed Scopus (33) Google Scholar, 11Rotanova T.V. Bioorgan. Khim. 1999; 25: 883-891PubMed Google Scholar) suggested that Lon consists of three functional domains: a variable N-terminal domain, an ATPase domain, and a C-terminal proteolytic domain. The domain organization has been confirmed by expression of functional domains of the E. coli and yeast mitochondrial Lon (12Rasulova F.S. Dergousova N.I. Starkova N.N. Melnikov E.E. Rumsh L.D. Ginodman L.M. Rotanova T.V. FEBS Lett. 1998; 432: 179-181Crossref PubMed Scopus (27) Google Scholar, 13van Dijl J.M. Kutejova E. Suda K. Perecko D. Schatz G. Suzuki C.K. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10584-10589Crossref PubMed Scopus (62) Google Scholar) and by limited proteolysis of the E. coli and Mycobacterium Lons (14Ovchinnikova T.V. Martynova N. Yu. Potapenko N.A. Vasilyeva O.V. Golikova N. Yu. Ginodman L.M. Rotanova T.V. Biomed. Tekhnol. 1998; 8: 62-65Google Scholar, 15Vasilyeva O.V. Kolygo K.B. Leonova Y.F. Potapenko N.A. Ovchinnikova T.V. FEBS Lett. 2002; 526: 66-70Crossref PubMed Scopus (32) Google Scholar, 16Roudiak S.G. Shrader T.E. Biochemistry. 1998; 37: 11255-11263Crossref PubMed Scopus (39) Google Scholar). Despite extensive studies of this enzyme, many of its structural characteristics remain undetermined, although its function in selective energy-dependent proteolysis has been characterized in considerable detail (3Goldberg A.L. Eur. J. Biochem. 1992; 203: 9-23Crossref PubMed Scopus (415) Google Scholar, 17Gottesman S. Annu. Rev. Genet. 1996; 30: 465-506Crossref PubMed Scopus (600) Google Scholar, 18Frank E.G. Ennis D.G. Gonzalez M. Levine A.S. Woodgate R. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 10291-10296Crossref PubMed Scopus (148) Google Scholar).As in other ATP-dependent proteases, the Lon ATPase domain belongs to the superfamily of AAA+ proteins (ATPases associated with different cellular activities) (19Neuwald A.F. Aravind L. Spouge J.L. Koonin E.V. Genome Res. 1999; 9: 27-43Crossref PubMed Google Scholar). The characteristic AAA+ domain consists of 220–250 amino acids that include hallmark Walker A and B motifs and several other regions of high sequence conservation (19Neuwald A.F. Aravind L. Spouge J.L. Koonin E.V. Genome Res. 1999; 9: 27-43Crossref PubMed Google Scholar). An AAA+ module has two structural domains (20Lupas A.N. Martin J. Curr. Opin. Struct. Biol. 2002; 12: 746-753Crossref PubMed Scopus (296) Google Scholar): a RecA-like α/β domain and an α domain that may interact with protein substrates. In response to ATP hydrolysis, these domains undergo changes in conformation and orientation. Transduction of the mechanical motions within the AAA+ module to bound substrates provides the driving force for various functions, including binding and unfolding of target proteins, translocation of proteins to an associated functional domain (in this case, the protease), and coordinated activation of the functional domain (1Wickner S. Maurizi M.R. Gottesman S. Science. 1999; 286: 1888-1893Crossref PubMed Scopus (907) Google Scholar, 20Lupas A.N. Martin J. Curr. Opin. Struct. Biol. 2002; 12: 746-753Crossref PubMed Scopus (296) Google Scholar, 21Maurizi M.R. Li C.C. EMBO Rep. 2001; 2: 980-985Crossref PubMed Scopus (48) Google Scholar, 22Lupas A. Flanagan J.M. Tamura T. Baumeister W. Trends Biochem. Sci. 1997; 22: 399-404Abstract Full Text PDF PubMed Scopus (204) Google Scholar, 23Zwickl P. Baumeister W. Steven A. Curr. Opin. Struct. Biol. 2000; 10: 242-250Crossref PubMed Scopus (78) Google Scholar).The heterooligomeric ATP-dependent proteases (ClpAP, ClpXP, and HslUV), as well as 26 S proteasomes, display similar overall architecture despite a lack of similarity in sequence and three-dimensional folds of their proteolytic subunits (22Lupas A. Flanagan J.M. Tamura T. Baumeister W. Trends Biochem. Sci. 1997; 22: 399-404Abstract Full Text PDF PubMed Scopus (204) Google Scholar). The AAA+ modules are contained in separate subunits, which assemble into six- or seven-membered rings. The proteolytic subunits also assemble into six- or seven-membered rings that stack upon each other to form barrel-shaped complexes with a central cavity containing the proteolytic active sites accessible by narrow axial channels in the rings. The ATPase rings interact at both ends of the protease barrels.Despite considerable efforts, the spatial arrangement and oligomeric state of intact, homooligomeric Lon and FtsH proteases are still not known with certainty. Analysis by proteolysis and expression of functional domains of Lon proteases from several sources points to self-association of the isologous domains into hexameric or heptameric rings (13van Dijl J.M. Kutejova E. Suda K. Perecko D. Schatz G. Suzuki C.K. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10584-10589Crossref PubMed Scopus (62) Google Scholar, 24Stahlberg H. Kutejova E. Suda K. Wolpensinger B. Lustig A. Schatz G. Engel A. Suzuki C.K. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 6787-6790Crossref PubMed Scopus (110) Google Scholar). 1M. R. Maurizi, F. Rasulova, G. Leffers, R. Leapman, and A. C. Steven, unpublished results. 1M. R. Maurizi, F. Rasulova, G. Leffers, R. Leapman, and A. C. Steven, unpublished results. High resolution structures of individual domains of FtsH and Lon have now been reported. The crystal structure of the AAA+ module of FtsH was used to construct a model of a hexameric ring similar to those formed by other AAA+ modules (25Krzywda S. Brzozowski A.M. Verma C. Karata K. Ogura T. Wilkinson A.J. Structure. 2002; 10: 1073-1083Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 26Niwa H. Tsuchiya D. Makyio H. Yoshida M. Morikawa K. Structure. 2002; 10: 1415-1423Abstract Full Text Full Text PDF PubMed Scopus (89) Google Scholar). The α domain of the E. coli Lon AAA+ module, the only structural fragment of Lon to be crystallized previously, displays a fold typically found in AAA+ proteins (27Botos I. Melnikov E.E. Cherry S. Khalatova A.G. Rasulova F. Tropea J. Maurizi M.R. Rotanova T.V. Gustchina A. Wlodawer A. J. Struct. Biol. 2003; (in press)Google Scholar). The quality of the structure of the isolated α domain confirmed the unique stability of these α domains, which are thought to undergo rigid body motions during the catalytic cycle of AAA+ proteins. However, the oligomerization state of the intact protein could not be deduced from the α domain structure (27Botos I. Melnikov E.E. Cherry S. Khalatova A.G. Rasulova F. Tropea J. Maurizi M.R. Rotanova T.V. Gustchina A. Wlodawer A. J. Struct. Biol. 2003; (in press)Google Scholar).The proteolytic components of ATP-dependent proteases contain several different active site types. ClpP has a classic serine protease triad (28Wang J. Hartling J.A. Flanagan J.M. Cell. 1997; 91: 447-456Abstract Full Text Full Text PDF PubMed Scopus (491) Google Scholar); HslV, like the proteasome, has a catalytic N-terminal threonine residue (29Yoo S.J. Shim Y.K. Seong I.S. Seol J.H. Kang M.S. Chung C.H. FEBS Lett. 1997; 412: 57-60Crossref PubMed Scopus (36) Google Scholar); and FtsH is a zinc-dependent metalloprotease (30Tomoyasu T. Gamer J. Bukau B. Kanemori M. Mori H. Rutman A.J. Oppenheim A.B. Yura T. Yamanaka K. Niki H. EMBO J. 1995; 14: 2551-2560Crossref PubMed Scopus (361) Google Scholar). The catalytic residues in Lon have remained uncertain until quite recently. Mutational studies suggested that Lon had a catalytic serine (31Amerik A.Y. Antonov V.K. Gorbalenya A.E. Kotova S.A. Rotanova T.V. Shimbarevich E.V. FEBS Lett. 1991; 287: 211-214Crossref PubMed Scopus (94) Google Scholar); however, other candidate catalytic residues could not be definitively identified (32Starkova N.N. Koroleva E.P. Rumsh L.D. Ginodman L.M. Rotanova T.V. FEBS Lett. 1998; 422: 218-220Crossref PubMed Scopus (48) Google Scholar). Extensive sequence comparisons suggested that Lon contains a catalytic Ser-Lys dyad (32Starkova N.N. Koroleva E.P. Rumsh L.D. Ginodman L.M. Rotanova T.V. FEBS Lett. 1998; 422: 218-220Crossref PubMed Scopus (48) Google Scholar, 33Birghan C. Mundt E. Gorbalenya A.E. EMBO J. 2000; 19: 114-123Crossref PubMed Scopus (166) Google Scholar, 34Rotanova T.V. Melnikov E.E. Tsirulnikov K.B. Bioorgan. Khim. 2003; 29: 97-99PubMed Google Scholar), and this model was experimentally supported by site-directed mutagenesis of candidate residues in E. coli Lon (34Rotanova T.V. Melnikov E.E. Tsirulnikov K.B. Bioorgan. Khim. 2003; 29: 97-99PubMed Google Scholar). Here, we report the high resolution crystal structure of the proteolytic domain of E. coli Lon. The structure confirms the presence of a Ser-Lys catalytic dyad in the active site and reveals a unique structural fold distinct from both the classical serine proteases containing active site catalytic triads and from other hydrolytic enzymes that are utilizing Ser-Lys catalytic dyads. The catalytic domain of Lon in the crystals assembles into hexameric rings, which provides strong support for a hexameric structure of the holoenzyme.EXPERIMENTAL PROCEDURESExpression and Purification of the Full-length Lon-S679A—The intact, proteolytically inactive mutant, Lon-S679A, was expressed and purified as described previously (27Botos I. Melnikov E.E. Cherry S. Khalatova A.G. Rasulova F. Tropea J. Maurizi M.R. Rotanova T.V. Gustchina A. Wlodawer A. J. Struct. Biol. 2003; (in press)Google Scholar, 35Rotanova T.V. Kotova S.A. Amerik A Yu. Lykov I.P. Ginodman L.M. Antonov V.K. Bioorgan. Khim. 1994; 20: 114-125PubMed Google Scholar). This procedure yielded ∼100 mg of Lon-S679A from 40 g of cell paste, with ∼90% purity. All of the purification procedures were performed at 4 °C and monitored by SDS-PAGE on 4–12% NuPAGE gels (Invitrogen). Protein concentration was estimated with a Bio-Rad protein assay using bovine serum albumin as the standard. All of the chromatography columns were from Amersham Biosciences, and the filter membranes were from Millipore (Bedford, MA).Limited Proteolysis of Intact Lon and Purification of the P Domain— Purified Lon-S679A (100 mg) was cleaved with 250 μg of α-chymotrypsin (Sigma) in 50 ml of 20 mm potassium phosphate, pH 8, containing 0.3 m NaCl. After2hof incubation at 30 °C, the reaction was stopped by adding phenylmethylsulfonyl fluoride to 1 mm. The reaction solution was cooled to 4 °C, diluted 6-fold with 20 mm potassium phosphate, pH 8.0, containing 1 mm DTT, 2The abbreviations used are: DTT, dithiothreitol; Se-Met, selenomethionine; PIPES, 1,4-piperazinediethanesulfonic acid; MES, 4-morpholineethanesulfonic acid; r.m.s., root mean square; SPase, signal peptidase. filtered through a 0.45-μm polysulfone membrane and loaded onto a 5-ml HiTrap Heparin HP column equilibrated with the dilution buffer. The target protein was collected in the flow-through, concentrated using a YM-10 membrane, diluted 10-fold with buffer A (20 mm HEPES, pH 7.0, 1 mm DTT), and applied to a 5-ml HiTrap Q-Sepharose HP column equilibrated in buffer A. The column was washed and the protein was eluted with a 100-ml linear gradient of NaCl from 0 to 1 m in buffer A. Fractions eluted within 0.1–0.2 m NaCl were pooled, concentrated on Centriprep-10, and applied to a HiLoad 26/60 Superdex 75 column equilibrated in 20 mm Tris-HCl, pH 8, 0.15 m NaCl, 1 mm DTT buffer. The purity and homogeneity of the target fragment (residues 585–784) were verified by N-terminal sequencing and electrospray ionization mass spectroscopy (Agilent 1100 series).Cloning, Expression, and Purification of the P Domain—Coding regions for the Lon P domain were amplified from a plasmid carrying the gene for Lon-S679A (31Amerik A.Y. Antonov V.K. Gorbalenya A.E. Kotova S.A. Rotanova T.V. Shimbarevich E.V. FEBS Lett. 1991; 287: 211-214Crossref PubMed Scopus (94) Google Scholar), using PCR with the following oligonucleotide primers: 5′-GAG AAC CTG TAC TTC CAG GAC TAT GGT CGC GCT GAT AAC GAA AAC-3′ and 5′-GGG GAC CAC TTT GTA CAA GAA AGC TGG GTT ATT ACT ATT TTG CAG TCA CAA CCT GCA TG-3′ (primer Lon R). The PCR amplicon was subsequently used as template for a second PCR with the following primers: 5′-GGG GAC AAG TTT GTA CAA AAA AGC AGG CTC GGA GAA CCT GTA CTT CCA G-3′ and primer Lon R. The amplicon from the second PCR was inserted by recombinational cloning into the entry vector pDONR201 (Invitrogen), and the nucleotide sequence was confirmed by DNA sequencing. The open reading frame encoding the P domain of Lon (Asp585-Lys784) containing an N-terminal recognition site (ENLYF(Q/D)) for tobacco etch virus protease was cloned into the destination vector pDEST-His-maltose-binding protein to produce pELP2. pELP2 directs the expression of the inactive, S679A P domain mutant of E. coli Lon as a fusion protein with E. coli maltose-binding protein with an intervening tobacco etch virus protease recognition site. The maltose-binding protein contains an N-terminal His6 tag for affinity purification by immobilized metal affinity chromatography. The fusion protein was expressed in the lon-deficient E. coli strain BL21 (DE3) (Novagen, Madison, WI). The cells were grown to mid-log phase (A600 = ∼0.5) at 37 °C in Luria broth containing 100 μg/ml ampicillin and 0.2% glucose. Overproduction of fusion protein was induced with isopropyl-α-d-thiogalactopyranoside at a final concentration of 1 mm for 4 h at 30 °C. The cells were pelleted by centrifugation and stored at -80 °C. The selenomethionine-substituted P domain of Lon was produced using the saturation of the methionine biosynthetic pathway protocol (36Doublie S. Methods Enzymol. 1997; 276: 523-530Crossref PubMed Scopus (791) Google Scholar).Cell paste was suspended in 25 mm HEPES, pH 8, containing 100 mm NaCl, 25 mm imidazole (buffer B), and Complete EDTA-free protease inhibitor mixture (Roche Applied Science). The cells were disrupted with an APV Gaulin Model G1000 homogenizer at 10,000 p.s.i. The homogenate was centrifuged at 20,000 × g for 30 min, and the supernatant was applied to a 10-ml nickel-nitrilotriacetic acid Superflow column (Qiagen) equilibrated in buffer B. The column was washed and eluted with a 100-ml linear gradient of imidazole from 25 to 200 mm. Fractions containing recombinant fusion protein were pooled and incubated overnight at 4 °C with 4–5 mg of His6-tagged tobacco etch virus protease. The digest was diluted 4-fold with 25 mm HEPES, pH 8, containing 100 mm NaCl and applied to a 25-ml nickel-nitrilotriacetic acid Superflow column equilibrated in buffer B. The column effluent containing the Lon P domain was collected, mixed with an equal volume of 25 mm HEPES, pH 8, 1 mm DTT (buffer C), and applied to a 16/5 Mono Q HR column equilibrated in buffer C. The column was washed, and protein was eluted with a 300-ml linear gradient of NaCl from 0 to 1 m. Desired fractions were pooled, concentrated using YM10 membrane, and applied to a HiPrep 26/60 Sephacryl S-100 HR column equilibrated in 20 mm Tris-HCl, pH 7.5, containing 150 mm NaCl and 1 mm DTT. Fractions with the P domain were pooled, concentrated as above, flash-frozen in liquid nitrogen, and stored at -80 °C until use.In all cases the final product was better than 95% pure on the basis of silver staining after SDS-PAGE. Protein concentration was estimated spectrophotometrically using a calculated molar extinction coefficient (37Gill S.C. von Hippel P.H. Anal. Biochem. 1989; 182: 319-326Crossref PubMed Scopus (5023) Google Scholar) of 9650 m-1 cm-1 at 280 nm. The molecular weights of the recombinant P domains of Lon were confirmed by electrospray mass spectrometry; selenomethionine substitution exceeded 99%.Protein Crystallization—Only the inactive Lon-S679A P domain was successfully crystallized, either in its native form or as a Se-Met derivative. For crystallization, the native form was concentrated to 18 mg/ml, and the Se-Met form was concentrated to 23.7 mg/ml. Initial screening of crystallization conditions (38Jancarik J. Kim S.H. J. Appl. Crystallogr. 1991; 21: 916-924Google Scholar) was carried out by the hanging drop, vapor diffusion method (39McPherson A. Preparation and Analysis of Protein Crystals. John Wiley & Sons, Inc., New York1982Google Scholar), using the Hampton (Hampton Research, Laguna Niguel, CA) and Wizard (Emerald Biostructures, Bain-bridge Island, WA) screens. Native P domain, whether proteolytically obtained or recombinant, crystallized under very similar conditions (2 m ammonium sulfate, 0.1 m PIPES, pH 6.5). Typical crystals grew to 0.2 × 0.1 × 0.1 mm in 5–10 days at room temperature. The crystals of the Se-Met derivative grew in 2 m ammonium sulfate, 0.1 m MES, pH 6.0, with the largest crystal growing to the size of 0.5 × 0.45 × 0.2 mm in 14 days at room temperature. Before flash freezing, the crystals were transferred into a cryoprotectant solution, consisting of 80% mother liquor and 20% ethylene glycol for the native and 100% paratone-N for the Se-Met derivative.Crystallographic Procedures—X-ray data for the native Lon P domain were collected on a Mar345 detector, using a Rigaku rotating anode x-ray source with CuKα radiation focused by an MSC/Osmic mirror system. A series of putative heavy atom derivatives produced by soaking were tested without success. Se-Met derivative data were collected on beamline X9B at Brookhaven National Laboratory, National Synchrotron Light Source, on a Quantum4 ADSC detector. Native data were processed using program DENZO and scaled using program SCALEPACK, whereas derivative data collected at the selenium peak wavelength was processed using the HKL2000 package (40Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref Scopus (38368) Google Scholar) (Table I). Single-wavelength anomalous dispersion phasing was carried out using programs SOLVE and RESOLVE (41Terwilliger T.C. Acta Crystallogr. Sect. D Biol. Crystallogr. 2001; 57: 1763-1775Crossref PubMed Scopus (78) Google Scholar). RESOLVE used density averaging based on the 6-fold noncrystallographic symmetry and yielded an excellent map and a partial model with 750 residues. The remainder of the model was built manually into the initial map using program O (42Jones T.A. Kjeldgaard M. Methods Enzymol. 1997; 277: 173-208Crossref PubMed Scopus (504) Google Scholar). Initial rigid body refinement with CNS (43Brünger 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. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16930) Google Scholar), using a maximum-likelihood target, was followed by simulated annealing (44Brünger A.T. Krukowski A. Erickson J.W. Acta Crystallogr. Sect. A. 1990; 46: 585-593Crossref PubMed Scopus (599) Google Scholar) with Engh and Huber parameters (45Engh R. Huber R. Acta Crystallogr. Sect. A. 1991; 47: 392-400Crossref Scopus (2537) Google Scholar). The model was rebuilt into density using program O. Subsequent refinement of the Se-Met structure was carried out with SHELXL (46Sheldrick G.M. Schneider T.R. Methods Enzymol. 1997; 277: 319-343Crossref PubMed Scopus (1874) Google Scholar) modeling in water and alternate conformations for side chains. TLS refinement was carried out with REFMAC5 (CCP4 suite) to an R factor of 19.7% and Rfree (47Brünger A.T. Nature. 1992; 355: 472-474Crossref PubMed Scopus (3849) Google Scholar) of 26.2%. The TLS parameters describe the thermal motion of a rigid body in terms of translation and libration of the group. In simple terms, the T tensor describes the mean square translation, the L tensor describes the mean square libration, and S describes the cross-correlation, respectively (48Schomaker V. Trueblood K.N. Acta Crystallogr. Sect. B Struct. Crystallogr. Cryst. Chem. 1968; 24: 63-76Crossref Google Scholar). The refinement of the native structure was completed with CNS to an R factor of 24.3% and Rfree of 29.6%. The coordinates and structure factors have been submitted to the Protein Data Bank with the accession codes 1rr9 for the native and 1rre for the Se-Met derivative, respectively.Table IStatistics of data collection and structure refinement FOM, figure of merit; FC, calculated structure factors; FP, experimental structure factors; CC, correlation coefficient.NativeSe-MetData collectionSpace groupP31Molecules/a.u.6Wavelength (Å)1.540.97915Unit cell parameters (Å)a = b = 86.21, c = 122.68a = b = 86.37, c = 124.16Resolution (Å)30-2.120-1.75Total reflections127,371460,655Unique reflections58,314105,429Completeness (%)aThe values in parentheses relate to the highest resolution shell98.0 (99.5)100.0 (100.0)Average I/σ14.1 (1.1)19.2 (2.1)Rmerge (%)bRmerge = Σ|I - 〈I〉/ΣI, where I is the observed intensity, and 〈I〉 is the average intensity obtained from multiple observations of symmetry-related reflections after rejections5.6 (62.4)7.5 (50.5)Phasing statistics (20-1.75 Å)Number of Se sites19Mean FOM of phasing (SOLVE)0.26Overall FOM of phasing (RESOLVE)0.47R for FC vs. FP (%)29.7CC of recovered map with RESOLVE map0.68Correlation of noncrystallographic symmetry regions (%)93Refinement statisticsR (%)cR = Σ||Fo| - |Fc||/Σ|Fo|, where Fo and Fc are the observed and calculated structure factors, respectively24.319.7Rfree (%)dRfree defined in Ref. 4729.626.2Root mean square deviation bond lengths (Å)0.0060.021Root mean square deviation angles (degrees)1.33.4Temperature factor (protein, Å3)47.819.5Temperature factor (solvent, Å3)41.441.9Number of protein atoms84618212Number of solvent molecules259792a The values in parentheses relate to the highest resolution shellb Rmerge = Σ|I - 〈I〉/ΣI, where I is the observed intensity, and 〈I〉 is the average intensity obtained from multiple observations of symmetry-related reflections after rejectionsc R = Σ||Fo| - |Fc||/Σ|Fo|, where Fo and Fc are the observed and calculated structure factors, respectivelyd Rfree defined in Ref. 47Brünger A.T. Nature. 1992; 355: 472-474Crossref PubMed Scopus (3849) Google Scholar Open table in a new tab Fold Assignment—The preliminary coordinates of molecule A of the catalytic domain of Lon were submitted to the DALI server (49Holm L. Sander C. J. Mol. Biol. 1993; 233: 123-138Crossref PubMed Scopus (3552) Google Scholar) to search for other proteins with a similar fold. The highest Z score was 5.3 with r.m.s. deviation of 4.3 Å for 107 Cα atoms of phosphomevalonate kinase (Protein Data Bank code 1k47). This level of similarity does not support any evolutionary relationship, especially with only 14% sequence identity. The similarity to any proteolytic enzymes is even lower, with the Z score for fibroblast collagenase of 2.6 with 2.9 Å r.m.s. deviation for only 76 Cα atoms. Thus, we can conclude that the fold of the catalytic domain of Lon is unique and that this enzyme is the first structurally characterized member of a novel family of proteases.RESULTS AND DISCUSSIONStructure Solution and Refinement—The crystal structure of the Lon-S679A P domain was solved using single-wavelength anomalous data for a Se-Met variant. The experimental map based on the anomalous signal, enhanced by solvent flattening and noncrystallographic symmetry averaging, was quite clear and enabled unambiguous tracing of all but a few terminal residues in each of six molecules in the asymmetric unit. This structure was refined at 1.75 Å resolution in parallel with the native protein data extending to 2.1 Å. The two structures were very similar (r.m.s. deviation 0.43 Å for Cα atoms), and only the former one will be discussed here in detail. Multiple conformations were evident for several side chains even in the original experimental map, in particular for methionines and cysteines. In the course of the refinement, it was crucial to model the alternate conformations for the heavy scatterers (selenium and sulfur atoms) to substantially reduce the phasing errors. The absence of density for 18 terminal residues/molecule (9 from either end) accounts for the ∼10% of the structure missing in our final model and contributes to the slightly elevated R factors. To further alleviate this problem, the 1.75 Å structure was refined with the TLS routine in program REFMAC5, reducing the free R factor by a further ∼2%.Earlier efforts to solve the crystal structure were hampered by twinning of the vast majority of the crystals. Most diffraction data, including, for example, data obtained with a mercury derivative, could be scaled equally well in both a trigonal (P31) and hexagonal (P62) space group, with a clear indication that the latter was due to twinning. The UCLA
Referência(s)