【病毒外文文献】2008 Without Its N-Finger, the Main Protease of Severe Acute Respiratory Syndrome Coronavirus Can Form a Novel Dimer thr

JOURNAL OF VIROLOGY May 2008 p 4227 4234 Vol 82 No 9 0022 538X 08 08 00H110010 doi 10 1128 JVI 02612 07 Copyright 2008 American Society for Microbiology All Rights Reserved Without Its N Finger the Main Protease of Severe Acute Respiratory Syndrome Coronavirus Can Form a Novel Dimer through Its C Terminal Domain H17188 Nan Zhong 1 2 Shengnan Zhang 1 2 Peng Zou 1 2 Jiaxuan Chen 1 3 Xue Kang 1 2 Zhe Li 1 3 Chao Liang 1 Changwen Jin 1 2 3 and Bin Xia 1 2 3 Beijing Nuclear Magnetic Resonance Center 1 College of Chemistry and Molecular Engineering 2 and College of Life Science 3 Peking University Beijing 100871 China Received 7 December 2007 Accepted 18 February 2008 The main protease M pro of severe acute respiratory syndrome coronavirus SARS CoV plays an essential role in the extensive proteolytic processing of the viral polyproteins pp1a and pp1ab and it is an important target for anti SARS drug development It was found that SARS CoV M pro exists in solution as an equilibrium of both monomeric and dimeric forms and the dimeric form is the enzymatically active form However the mechanism of SARS CoV M pro dimerization especially the roles of its N terminal seven residues N finger and its unique C terminal domain in the dimerization remain unclear Here we report that the SARS CoV M pro C terminal domain alone residues 187 to 306 M pro C is produced in Escherichia coli in both monomeric and dimeric forms and no exchange could be observed between them at room temperature The M pro C dimer has a novel dimerization interface Meanwhile the N finger deletion mutant of SARS CoV M pro also exists as both a stable monomer and a stable dimer and the dimer is formed through the same C terminal domain interaction as that in the M pro C dimer However no C terminal domain mediated dimerization form can be detected for wild type SARS CoV M pro Our study results help to clarify previously published controversial claims about the role of the N finger in SARS CoV M pro dimerization Apparently without the N finger SARS CoV M pro can no longer retain the active dimer structure instead it can form a new type of dimer which is inactive Therefore the N finger of SARS CoV M pro is not only critical for its dimerization but also essential for the enzyme to form the enzymatically active dimer A novel coronavirus CoV was identified as the etiological agent of the highly epidemic severe acute respiratory syndrome SARS which has infected more than 8 400 people with a high fatality rate of about 10 3 14 16 25 SARS CoV is a positive sense single stranded RNA virus The genome of the virus encodes two overlapping polyproteins pp1a 486 kDa and pp1ab 790 kDa which mediate viral replication and transcription 17 19 20 The main protease M pro of SARS CoV also named 3C like protease plays an important role in the extensive proteolytic processing of the viral polyproteins pp1a and pp1ab which makes it essential for the viral life cycle and represents an attractive target for antiviral agent develop ment 2 29 30 The first crystal structure of SARS CoV M pro was solved in 2003 and the enzyme is a symmetric homodimer with a fold similar to that of the porcine transmissible gastroenteritis virus TGEV M pro 1 30 The N terminal domain residues 1 to 184 of SARS CoV M pro has a chymotrypsin like fold and the C terminal domain residues 201 to 303 has a globular fold containing five H9251 helices 30 It was reported that SARS CoV M pro exists in solution as an equilibrium between monomeric and dimeric forms 10 and only the dimeric form of SARS CoV M pro is active 9 In the crystal structure of SARS CoV M pro the N terminal residues 1 to 7 N finger of each protomer are squeezed in between two protomers and make contacts with both the N terminal and C terminal domains of the other protomer and these contacts are important for dimerization 30 This dimerization pattern is similar to that of the TGEV M pro in which the role of the N finger in the dimerization has been analyzed in detail 1 However previous studies of the N finger deletion mu tants of M pro gave different views of the role of the N finger in SARS CoV M pro dimerization Hsu et al stated that the N finger especially residue R4 is indispensable for the dimeriza tion and enzymatic activity of SARS CoV M pro and the mo nomeric form becomes the predominant form after deletion of the N terminal four to seven residues 6 11 On the other hand Chen et al concluded that the N finger is not crucial for the dimerization of SARS CoV M pro but is fundamental only to the enzymatic activity and they found that the N finger deletion mutant and wild type WT SARS CoV M pro have similar dissociation constants for the dimerization 4 Re cently Wei et al reported that the N finger deletion mutant of SARS CoV M pro could not dimerize at all 24 These contro versial results turn the role of the N finger in SARS CoV M pro dimerization into a mystery Meanwhile Shi et al proposed that the C terminal domain plays a critical role in SARS CoV M pro dimerization based on the observation that the N terminal domain alone is a mono mer and the C terminal domain alone is only a dimer 22 Corresponding author Mailing address Beijing Nuclear Magnetic Resonance Center Peking University Beijing 100871 P R China Phone 86 10 6275 8127 Fax 86 10 6275 3790 E mail binxia pku H17188 Published ahead of print on 27 February 2008 4227 on March 18 2015 by ST ANDREWS UNIV http jvi asm org Downloaded from Nevertheless in the crystal structure of SARS CoV M pro there is almost no direct contact between the two C terminal domains of the homodimer Therefore it is not obvious how the dimerization of the C terminal domain is related to the dimerization of SARS CoV M pro In order to clarify the controversial issues mentioned above we have reinvestigated the dimerization of SARS CoV M pro We found that the SARS CoV M pro C terminal domain alone M pro C exists as a stable monomer and a stable dimer simul taneously There is no obvious conversion between the two forms The dimerization interface of the M pro C dimer is novel and is unrelated to that of SARS CoV M pro in the crystal structure Without the N finger SARS CoV M pro can also form a stable dimer due to the dimerization of its C terminal domain MATERIALS AND METHODS Construction of expression plasmids For WT SARS CoV M pro the DNA fragment encoding residues 1 to 306 was cloned into the pET21a vector and an NdeI restriction site within the coding sequence was removed by changing the codon for H164 from CAT to CAC A hexahistidine tag sequence LEHHHHHH was engineered at the carboxyl terminus of the protein For the N terminal domain of SARS CoV M pro M pro N the DNA fragment encoding residues 1 to 199 was cloned into the pET21a vector For the C terminal domain M pro C the DNA fragment encoding residues 187 to 306 was cloned into the pET21a vector For the N finger deletion mutant of SARS CoV M pro M pro H90047 the DNA fragment encoding residues 8 to 306 was cloned into pET28a with a hexahistidine tag sequence LEHHHHHH attached at the carboxyl terminus Protein production and purification The proteins were all produced in the Escherichia coli Rosetta DE3 pLysS strain The proteins with without a C terminal hexahistidine tag were purified using nickel nitrilotriacetic acid ion exchange chromatography and followed by gel filtration Superdex 75 column on an A KTA fast protein liquid chromatography system FPLC GE Cross linking experiment Protein 0 1 mM was cross linked with 3 mM ethylene glycolbis succinimidylsuccinate EGS Pierce Rockford IL in the reaction buffer 0 1 M potassium phosphate 0 15 M NaCl pH 7 2 The reaction mixture was incubated at room temperature for 30 min and then the reaction was quenched by adding Tris 1 M pH 7 5 to a final concentration of 50 mM Gel filtration analysis Protein oligomerization was analyzed using a home packed 16 70 Superdex 75 HR gel filtration column on an A KTA FPLC All protein samples were in 50 mM potassium phosphate buffer pH 8 0 with 1 mM 1 4 dithiothreitol DTT To estimate the apparent molecular mass based on the retention volume three proteins myoglobin 17 0 kDa 88 0 ml egg albumin 42 7 kDa 73 4 ml and bovine albumin V 68 0 kDa 67 2 ml were used as the molecular mass standard A standard calibration curve was obtained by plotting the ratio Ve V 0 Vt V 0 against the logarithm of mo lecular mass Ve is elution volume V 0 is the void volume and Vt is the total bed volume Enzymatic activity assay The enzymatic activities of WT SARS CoV M pro and the mutant M pro H90047 were measured using a fluorogenic peptide MCA AVLQSGFR Lys Dnp Lys NH 2 more than 95 purity GL Biochem Shang hai Ltd as the substrate The fluorescence intensity was monitored using a Hitachi Tokyo Japan F 4500 fluorescence spectrophotometer with wave lengths of 320 and 405 nm for excitation and emission respectively The reaction buffer consisted of 50 mM Tris HCl pH 7 3 1 mM EDTA and 1 mM DTT The working concentrations of the protease and the substrate were 1 H9262M and 40 H9262M respectively 29 NMR spectroscopy Nuclear magnetic resonance NMR samples of uniformly 15 N labeled 15 N 13 C labeled and 2 H 15 N 13 C labeled M pro C and the uniformly 2 H 15 N labeled M pro H90047 dimer and monomer were prepared All NMR samples were at a concentration of about 1 mM and were prepared in buffer containing 50 mM potassium phosphate pH 7 0 1 mM EDTA and 0 03 NaN 3 in 90 H 2 O 10 D 2 O plus Complete an EDTA free protease inhibitor cocktail Roche Germany All NMR experiments were performed at 298 K on a Bruker Avance 500 MHz with cryoprobe or 600 MHz NMR spectrometer Backbone chemical shift assignments were based on a two dimensional 2D 1 H 15 N het eronuclear single quantum coherence spectrum and three dimensional HNCA HN CO CA HN CA CB HN COCA CB HNCO and HN CA CO experi ment data 21 All NMR spectra were processed with the NMRPipe software program 7 and analyzed using NMRView software 12 The chemical shift in the 1 H dimension was referenced directly to 2 2 dimethyl 2 silapentanesulfonic acid DSS whereas the chemical shifts in the 13 C and 15 N dimensions were indirectly referenced to DSS 26 Dimer structure modeling The model of the M pro C dimer was calculated using the protein protein docking program HADDOCK 8 The docking was initiated from the C terminal domain part of the SARS CoV M pro crystal structure 30 The active residues were defined based on the chemical shift perturbation data in which residues with the combined NH chemical shift difference between two forms exceeding 0 10 ppm average H11001 0 5 H11003 standard deviation were selected They were residues R217 to T225 A260 D263 C265 to A267 K269 to L271 and L282 The passive residues were defined as all other surface accessible residues residues with more than 55 solvent accessible surface area determined using the MOLMOL software program 13 The ambiguous interaction restraints were defined between the active residues of one protomer and all the active and passive residues of the other protomer The active residues were set as flexible segments and the passive residues H11006 2 sequential residues were set as semiflexible segments during the calculation RESULTS Two oligomerization states of M pro C We have produced the N terminal domain alone residues 1 to 199 M pro N and the C terminal domain alone residues 187 to 306 M pro C of SARS CoV M pro in E coli In agreement with the previous report M pro N behaved as a monomer on the gel filtration column retention volume 86 3 ml apparent molecular mass 17 7 kDa and the retention volume did not show significant concentration dependence data not shown 22 Interestingly we found that M pro C was produced in E coli in two forms which could be separated by gel filtration and the retention volumes of the two forms were 74 6 ml and 86 4 ml respectively Fig 1A The apparent molecular masses calcu lated based on their retention volumes are 38 2 kDa for the 74 6 ml fraction and 17 5 kDa for the 86 4 ml fraction On a sodium dodecyl sulfate SDS polyacrylamide gel both frac tions appeared at the same position under either reducing or nonreducing conditions with an apparent molecular mass of H1101113 kDa Fig 1B lanes M2 M3 D2 and D3 Mass spec trometry analysis also confirmed that the molecular masses of both fractions were the same as the theoretical value 13 4 kDa for the M pro C monomer After treatment with the cross linking agent EGS the 86 4 ml fraction appeared at the same position as the untreated sample in SDS polyacrylamide gel electrophoresis PAGE while the 74 6 ml fraction showed two bands One band was at the same position as the untreated sample and the other was at the position of H1101128 kDa Fig 1B lane M1 and D1 These data suggest that M pro C is produced in E coli not only in the dimeric form 74 6 ml fraction as described by Shi et al 22 but also in a monomeric form 86 4 ml fraction Meanwhile the dimerization of M pro C is noncovalent and no disulfide bond is involved even though M pro C has two free cysteine residues Surprisingly we found that the two forms of M pro C are stable and there is no obvi ous conversion between the monomeric and dimeric forms for days at room temperature which was monitored by gel filtra tion analysis Fig 1A Novel interface for the M pro C dimer Carefully examining all the crystal structures of SARS CoV M pro 15 23 27 28 30 we found that there is almost no direct contact between the C terminal domains of the two protomers Only the side 4228 ZHONG ET AL J VIROL on March 18 2015 by ST ANDREWS UNIV http jvi asm org Downloaded from chains of the C terminal domain residues T285 and I286 from each protomer are closer than5 inallofthese structures In the structure of TGEV M pro which has tertiary and quaternary structures similar to those of SARS CoV M pro two hydrogen bonds are found between the C terminal domains of the two protomers However it was suggested that the interactions between the C terminal domains of TGEV M pro appear to be a consequence rather than the cause of the dimerization 1 Thus the current available structural information cannot elu cidate why the SARS CoV M pro C terminal domain alone can form a stable dimer To determine the dimerization interface of the M pro C dimer we carried out the backbone NMR res onance assignments for both the monomeric and dimeric forms of M pro C For the M pro C monomer nearly all back bone NH chemical shift assignments were obtained with the exception of residues F219 and E288 whose NH signals were missing For the M pro C dimer backbone NH signals for res idues F219 R222 F223 and E288 were missing in the 2D 1 H 15 N HSQC spectrum while all the other NH signals have been assigned The missing NH signals probably resulted from intermedi ate time scale conformational exchange which causes broad ening of the NMR signals beyond detection Since the chem ical shifts of NH signals are sensitive to the local chemical environment change for individual NH group the dimerization would cause the chemical environment change for the residues at the dimer interface thus resulting in the NH chemical shift changes for these residues Therefore the dimerization inter face of the M pro C dimer can be identified from the compar ison of the NH chemical shift differences between the mono meric and dimeric forms of M pro C Most of the NH peaks overlap well between the 2D 1 H 15 N HSQC spectra of the monomeric and dimeric forms of M pro C Fig 2A As expected some NH signals exhibit significant chemical shift differences between the two forms Residues with a large combined NH chemical shift difference H9004H9254 comb of H110220 1 ppm include V212 R217 T225 F219 R222 and F223 are missing A260 Q273 except V261 L262 M264 L268 and L272 and M276 Residues with a H9004H9254 comb value of less than 0 1 ppm but more than 0 05 ppm are the following I200 A210 N228 F230 Y239 V261 L268 L272 T280 G283 and E290 In addition the side chain NH 2 signals of residues N214 N221 N274 and N277 also show significant chemical shift differences between the monomeric and dimeric forms Fig 2B Notably all of the residues mentioned above are located at the loop consisting of residues R217 to T225 D loop and a helix formed by residues A260 to Q273 D helix which is right underneath the D loop Thus the D loop and D helix should represent the dimerization interface for the M pro C dimer Mapping this dimer interface onto the crystal structure of the SARS CoV M pro dimer colored blue and red for dif ferent protomers in Fig 3A it is obvious that the dimer ization interface of the M pro C dimer is not related to the dimerization interface in the crystal structure of SARS CoV M pro All residues on the D loop show relatively larger H9004H9254 comb values than those on the D helix Fig 2B NH signals from residues R222 and F223 are observed for the monomeric form but are not detected for the dimeric form of M pro C presum ably due to a difference in the conformational exchange rates between the two forms Also the D helix residues which are facing the D loop have relatively larger H9004H9254 comb values than those on the opposite side Fig 2B Furthermore the 13 C H9251 chemical shift differences of most residues on the D helix are less than 0 2 ppm within the 13 C chemical shift resolution Fig 2C Since the 13 C H9251 chemical shift is sensitive to the secondary structure change this should indicate that this helix does not undergo much conformational change after the dimerization On the contrary some of the residues on the D loop show quite large 13 C H9251 chemical shift differences over 2 ppm between the monomeric and dimeric forms suggesting that this loop probably changes its conformation upon dimer ization Fig 2C This should imply that the NH and 13 C H9251 chemical shift differences observed for residues on the D helix are from a secondary effect of the D loop conformation change due to the dimerization Based on the chemical shift perturbation data a structure model of the M pro C dimer was calculated using the docking program Haddock Comparing the final refined models of the dimer with the C terminal domain structure of SARS CoV M pro it seems that the dimerization is mainly due to the hy drophobic interaction between the residue F223 of one M pro C molecule and the residues W218 F219 and L271 of the other molecule and also possibly a few hydrogen bonds Fig 3B Dimerization of WT SARS CoV M pro and its N finger dele tion mutant Since M pro C can form a stable dimer we tried to FIG 1 A Elution profile of M pro C from gel filtration analysis The solid line is for the purification of the M pro C protein and the elution peaks for the monomeric M and dimeric D forms are indicated The broken line and dotted line are for the purified M pro C monomeric and dimeric protein samples after 3 days at room temper ature respectively B SDS PAGE analysis of M pro C Lanes M2 and M3 are the M pro C monomer with without 10 mM DTT respectively lanes D2 and D3 are the M pro C dimer with without 10 mM DTT respectively lanes M1 and D1 are the M pro C monomeric and dimeric forms treated with cross linking agent EGS and the center lane is the molecular mass marker VOL 82 2008 NOVEL DIMER OF SARS CoV MAIN PROTEASE C TERMINAL DOMAIN 4229 on March 18 2015 by ST ANDREWS UNIV http jvi asm org Downloaded from find out whether WT SARS CoV M pro has a similar C termi nal domain mediated stable dimeric form which is different from the dimer of the crystal structure Our study showed that WT SARS CoV M pro behaves as an equilibrium between the monomeric and dimeric forms on a gel filtration column and no stable dimeric form could be detected The retention vol umes of WT SARS CoV M pro were concentration dependent on the gel filtration column for samples with concentrations of 25 9 2 and 0 1 mg ml the corresponding retention volumes are 70 2 70 8 72 7 and 75 2 ml respectively Fig 4A At the concentration of 0 1 mg ml the estimated apparent molecular mass is 36 7 kDa v