purification and initial biochemical characterization of ... filepurification and initial...

Purification and Initial Biochemical Characterization of the ATP:Cob(I)alaminAdenosyltransferase(EutT) Enzyme of Salmonella enterica

Nicole R. Buan§ and Jorge Escalante-Semerena*Department of Bacteriology, University of Wisconsin-Madison

Running title: the EutT corrinoid adenosyltyransferase of Salmonella*Corresponding author: 1710 University Av., Madison, WI 53726-4087

Telephone: 608-262-7379; FAX: 608-265-7909; E-mail: [email protected]

The ATP:Co(I)balamin adenosyltransferase(EutT) enzyme of Salmonella enterica wasoverproduced, enriched to ~70% homogeneity,and its basic kinetic parameters weredetermined. Abundant amounts of EutT proteinwere produced, but all of it remained insoluble.Soluble, active EutT protein (~70%homogeneous) was obtained after treatment withdetergent. Under conditions where Cbl wassaturating, the Km ATP = 10 µM, kcat = 0.03 s-1, andVmax = 54.5 nM min-1. Similarly, underconditions where MgATP was saturating, Km

Cbl =4.1 µM, kcat = 0.06 s-1, and Vmax = 105 nM min-1.U n l i k e o t h e r A T P : c o ( I ) r r i n o i dadenosyltransferases in the cell (i.e. CobA,PduO), EutT activity was >50-fold higher withATP vs GTP, retained 80% of its activity whenADP substituted for ATP, and was completelyinactive with AMP as substrate, indicating thatthe enzyme required the β-phosphate group ofthe nucleotide substrate. The data suggested thatthe amino group of adenine might play a role innucleotide recognition and/or binding. Unlikethe housekeeping CobA enzyme, EutT was notinhibited by PPPi. Results from 31P-NMRspectroscopy studies identified PPi and Pi as by-products of the EutT reaction. In the absence ofcobalamin, EutT cleaved ATP into adenosineand PPPi, suggesting that PPPi was broken downinto PPi and Pi. Electron transfer proteinpartners for EutT were not encoded by the eutoperon. EutT-dependent activity was detected incell-free extracts of cobA strains enriched forEutT when FMN and NADH were used toreduce cob(III)alamin to cob(I)balamin.

ATP:Co(I)rrinoid adenosyltransferases (CAs)1 areenzymes responsible for the conversion of B12 (akacobalamin, Cbl) into coenzyme B12 (aka adenosyl-B12, AdoB12; adenosylcobalamin, AdoCbl) (Fig. 1).In the enterobacterium Salmonella enterica serovarTyphimurium LT2 (hereafter S. enterica), there arethree different classes of CA enzymes. The

housekeeping CA of this bacterium is encoded bythe cobA gene, which is constitutively expressed(1). Two other CAs (PduO, EutT) (2-4) are encodedby large operons whose functions are needed for thecatabolism of 1,2-propanediol (the propanediolutilization (pdu) operon) (5), or ethanolamine (theethanolamine utilization (eut) operon) (6,7).

Bioinformatics analysis of the CobA, PduO, andEutT proteins shows that they evolved fromdifferent ancestors. CobA, the best characterized ofthe three enzymes, adenosylates incomplete andcomplete corrinoids, its three-dimesional structurecomplexed with its substrates has been solved, andits physiological electron donor partner (i.e.flavodoxin A) is known (8-13). The PduO enzymehas been studied in some detail, and knowledge ofhow it works is of particular interest because of itshomology to CA enzymes found in mammals,including humans (14,15). Structure studies of thearchaeal PduO homolog of T h e r m o p l a s m aacidophilum have provided insights into the basisfor methylmalonic aciduria in humans (16). TheEutT enzyme is the least understood of all threetypes of CA enzymes. Identification of eutT as thegene encoding the CA enzyme of the operon wasjust recently reported (3,4). The reason for thediversity in the lineage of these proteins is notknown.

There are many open questions regarding PduOand EutT function. Ass mentioned above, we do notknow the identity of the electron donor thatgenerates the cob(I)balamin nucleophile needed forformation of the Co-C bond, nor do we understandhow PduO or EutT bind ATP or their corrinoidsubstrates. To better understand the moleculardetails of how the EutT enzyme works, we purifiedthe protein and began the biochemicalcharacterization of its enzymatic activity.

EXPERIMENTAL PROCEDURESMicrobiological Techniques

Assessment of growth phenotypes. Strains weregrown overnight in 2 ml of lysogenic broth (LB)

http://www.jbc.org/cgi/doi/10.1074/jbc.M603069200The latest version is at JBC Papers in Press. Published on April 24, 2006 as Manuscript M603069200

Copyright 2006 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on January 4, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/cgi/doi/10.1074/jbc.M603069200

http://www.jbc.org/

Buan & Escalante-Semerena the EutT corrinoid adenosyltransferase of Salmonella 2

(19,20) containing the appropriate antibiotic;cultures were incubated at 37°C. Growth behaviorwas analyzed using 96-well plates containing 190µ l of no-carbon E (NCE) medium (21),ethanolamine (50 mM, pH 7), trace minerals elixir(22), MgSO4 (1 mM), NH4Cl (50 mM), L-methionine (0.5 mM), hydroxocobalamin (200 nM,HOCbl), and inoculated with a 10-µ l sample(~2x107 cfu) of an overnight LB culture. Plateswere placed in the chamber of a BioTek platereader whose temperature was held at 37°; shakingwas constant at maximum setting.

Recombinant DNA TechniquesPlasmid construction. Strains, plasmids andprimers used in this study are listed in Tables 1,2.Plasmid pEUT12. The S. enterica eutT+ gene wasamplified by PCR using the appropriate primers.The NdeI/XhoI fragment containing the native stopcodon was cloned into pTyb2 (New EnglandBioLabs) to yield plasmid pEUT12 foroverproduction of EutTWT in E. coli.Plasmid pEUT25-27. Plasmid pEUT25-27 encodeEutT variants which were constructed using theSite-Directed Mutagenesis™ Kit (Stratagene).Plasmid pEUT12 was used as template. Mutationsin plasmids pEUT25-27 were verified by DNAsequencing.Plasmids pEUT35-37. To avoid indirect effects byunknown mutations introduced into the cloningvector, the mutant alleles were moved frompEUT25-27 into a non-mutagenized cloning vectorand were placed under the control of a tunablepromoter. For this purpose, primers shown in Table1s (supplemental online data; see plasmidspEUT35-37) were used to amplify the mutant eutTalleles from their respective pTyb2 plasmids(pEUT25-27). Amplification products were cut withEcoRI and XbaI restriction enzymes and placedunder the control of the arabinose-inducible ParaBADpromoter in plasmid pBAD24 (23) cut with thesame enzymes. The resulting plasmids were namedpEUT36 (eutT1168, EutTC79A), pEUT37 (eutT1169,EutTC80A) and pEUT38 (eutT1170, EutTC83A). Theseplasmids were used in in vivo studies aimed atdetermining the functionality of the proteinsencoded by the mutant eutT alleles. For thispurpose, plasmids were transformed (24) into strainJE7180 [metE205 ara-9 cobA366::Tn10d(cat+)eutE18::MudI1734(kan+)]. Growth of the resultingstrains was assessed as described above.

Plasmid pEUT55-58. Wild-type chromosomalalleles of eutP, eutQ, eutT, and combinations ofthem were constructed by PCR amplification usingprimers shown in Table 1s (supplemental onlinedata). Amplification products were cloned under thecontrol of the ParaBAD promoter in vector pBAD24using the appropriate restriction sites.Plasmid pFDX1. The S. enterica ferredoxin gene(fdx+) was amplified, the PCR fragment was cutwith NdeI and BamHI restriction enzymes andcloned into plasmid pET-16b (Novagen) yieldingplasmid pFDX1, which encoded ferredoxin with ahexahistidine tag fused to its N terminus.DNA sequencing. All plasmids were sequencedusing BigDye® protocols (ABI PRISM) by theDNA sequencing Facility at the Biotech Center atthe University of Wisconsin-Madison.

Biochemical TechniquesEnrichment of EutT protein. A transformant of E.coli BL21(λDE3) carrying the appropriate eutT+

plasmid was used to inoculate two sterile culturetubes (16 mm x 150 mm) containing 5 ml of LBplus ampicillin (100 µg/ml) and grown overnight at37°C. The next morning, the 10 ml starter wasadded to 500 ml LB ampicillin, trace minerals,cysteine (0.5 mM), methionine (0.5 mM) in a 2-literErlenmeyer flask. The culture was shaken at 140rpm at 37°C for 1.5 h. Isopropyl-β -D-thiogalactopyranoside (IPTG) was added to 0.5 mMand the temperature was dropped to 15°C forovernight incubation. Cells were harvested bycentrifugation at 7,354 xg in a Beckman CoulterAvanti J-25I centrifuge fitted with a LA 16.250rotor at 4°C for 10 min. Cell paste was transferredto a 50-ml serum vial, flushed with O2-free N2 onice for 10 min and frozen at -80°C under 100 kPa ofpressure until use.

Cells were broken under anoxic conditions. Forthis purpose, frozen cell paste was transferred to ananoxic chamber and resuspended in 10 ml of anoxic1X Bug Buster™ reagent (Novagen) in 20 mMglycine / 20 mM N - c y c l o h e x y l - 2 -aminoethanesulfonic acid (CHES) buffer pH 9.5containing 1 mM phenylmethylsulphonylfluoride(PMSF). The cell suspension was incubated at roomtemperature for 20 min with intermittent agitation(approx. every 5 min). The lysate was moved into astainless-steel centrifuge tube fitted with anexpandable O-ring cap, centrifuged at 4ºC for 30min at 43,667 x g in a Beckman Coulter Avanti J-


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


25I centrifuge fitted with a JA 25.50 rotor. Thecentrifuged lysate was moved back to the anoxicchamber where the supernatant was decanted.Pelleted cell membranes and insoluble materialwere resuspended in 10 ml of an anoxic solutioncontaining 8 mM CHAPS in 20 mM glycine / 20mM CHES buffer pH 9.5. Repeated pipettingensured homogeneous resuspension of the pellet,which was then incubated at room temperature for 1h with gentle agitation every 10 min. Thesuspension was centrifuged in stainless steel tubesfitted with sealable caps at 4ºC for 1 h at 43,667 xg.Tubes were transferred back into the anoxicchamber where the soluble material was decanted,200-µl samples were dispensed into glass serumvials fitted with gray butyl rubber stoppers, andaluminum seals were crimped in place. Sealed vialswere removed from the chamber, pressurized withN2 (100 kPa), and kept at 4°C until use.Protein purity analysis. Protein purity wasassessed by SDS-PAGE (25) using Fotodyne’sFOTO/Eclipse® Electronic Documentation &Analysis System, including software packagesFOTO/Analyst® PC Image v5.0 and TotalLab™1D gel analysis v2003 from NonLinear Dynamics,Ltd.Overexpression and isolation of Fpr, FldA andFdx. Fpr and FldA proteins were overproduced andisolated as previously reported (12,26). Fresh E.coli BL21 (λDE3) (JE3892) / pFDX1 transformantswere grown in two 10-ml LB plus ampicillin (100µg/ml) cultures and incubated overnight at 37°Cwith shaking. Each 10-ml starter culture was usedto inoculate 2 liters of Terrific Broth (TB) mediumsupplemented with 0.1 mg/ml ferrous ammoniumsulfate and 100 µg/ml ampicillin (27); inoculatedflasks were incubated at 37°C for 15 h. Cells wereharvested by centrifugation at 4ºC at 5,000 xg in aBeckman Coulter Avanti J-20 XPI centrifuge fittedwith a JLA 8.1000 rotor, and the cell paste wasstored at -80°C until used. Cell paste wasresuspended in 20 mM KPi buffer pH 7.4containing 0.5 M NaCl, 1 mM PMSF, passedthrough a French press cell twice at 1250 kPa andtreated with DNAse for 10 min on ice. Cell-freeextracts were centrifuged for 30 min at 4ºC at43,000 xg in a Beckman Coulter Avanti J-25Icentrifuge fitted with a JA 25.50 rotor, and cell-freeextracts were passed through a 0.2 µM syringe filterto remove residual particulate material. H6Fdxprotein was purified on a ÄKTA FPLC Explorer

using a 5-ml HiTrap Chelating HP column(Amersham) charged with NiSO4. Fractionscontaining pure protein were pooled, concentratedin an Amicon YM10 Centricon unit (cut off = 10kDa), dialyzed against 20 mM KPi buffer pH 7.4containing 0.2 M NaCl, and concentrated again inan Amicon YM10 Microcon unit. Total yield wasapproximately 1 mg of EutT protein per 10 g of cellpaste.Preparation of cell-free extracts. S.entericastrains were grown in NCE minimal mediumcontaining trace minerals elixir, MgSO4, NH4Cl,methionine at concentrations described above. Inaddition, 1 µM HOCbl (or AdoCbl as indicated),200 µM L(+)-arabinose as inducer, andethanolamine (50 mM, pH 7) or glycerol (30 mM)were added. Cells in stationary phase wereharvested by centrifugation at 4°C, transferred to a50-ml serum vial and flushed with anoxic N2 gas.Cells were broken at room temperature afterresuspension in anoxic lysis buffer [Bug Busterreagent (Novagen), 20 mM glycine / 20 mM CHESbuffer pH 9.5, containing 1 mM PMSF]. The cellsuspension was kept at room temperature for 30min with intermittent agitation. Proteinconcentration was determined using the BradfordReagent (Bio-Rad Laboratories) according tomanufacturer’s instructions. Cell-free extracts wereused immediately for corrinoid adenosylationassays.Corrinoid adenosylation assays. Quartz cuvettesfitted with removable silicon rubber septa (StarnaCells) were flushed with O2-free N2 and filled withdegassed buffer containing 0.2 M TrisCl buffer (pH7, 37°C). HOCbl was added between 0.5-50 µMMnCl2 was added. Ti(III)citrate (2.2 mM) was usedto reduce the corrin from Co3+ to Co1+, ATP wasadded (0.5 µM-1mM), and 0.5 µM purifiedCHAPS/EutT protein or 50 µg cell-free extract wasused per reaction. CHAPS/EutT protein waswarmed to room temperature prior to its addition tothe reaction mixture. A temperature shift to 37°Cwas used to start the reaction whose rate wasmonitored by the decrease in the abundance ofcob(I)balamin substrate at 388 nm. When FldA orFdx were used as reductant, Ti(III)citrate wasreplaced by 2 µM Fpr protein, 0.5 mM NADPH and1 µM of purified H6Fdx (N-terminal H6 tag) orFldAH6 (C-terminal H6 tag) proteins. In assayswhere cell-free extract was used in lieu of purifiedCHAPS/EutT protein, 50 µM FMN, 1 mM NADH


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


and 1 mM ATP were used in place of Ti(III)citrate.Corrinoid adenosylation was assessed by the changein absorbance at 525 nm of the sample after 30 minof incubation at 37°C and after photolysis for 10min on ice.31P-NMR spectroscopy studies. A 10-ml corrinoidadenosylation reaction mixture containing 3 mMTi(III)citrate, 100 µM ATP, 100 µM HOCbl, 0.5mM MgCl2, 16 µM EutT in 0.2 M TrisCl buffer(pH 7, 37°C) was incubated for 2 h at 37°C. Afterincubation, ethylenediaminetetraacetic acid (EDTA)was added to 20 mM, reactions were concentratedunder vacuum to 0.5 ml in a SPD111V SpeedVac®(Thermo Savant) concentrator overnight at roomtemperature; 100% D20 was added to a finalconcentration of 6% (v/v). Standards were added to100 µM when indicated. 31P-NMR spectra wereacquired at the Nuclear Magnetic ResonanceFacility at Madison (NMRFAM) at the Universityof Wisconsin-Madison.Chemicals. All chemicals were obtained fromSigma unless otherwise indicated, and were of highpurity. Ti(III)citrate was prepared from liquid TiCl3.Briefly, 0.2 M citrate was adjusted to pH 8. Thecitrate solution was degassed with O2-free N2 in asealed glass serum vial for 20 min. TiCl3 wasadded by means of a glass syringe into the serumvial with a pressurized O2-free N2 headspace. Thefinal Ti(III)citrate concentration was 88 mM.

RESULTSInit ial purif ication and biochemicalcharacterization of EutT protein. RecombinantEutT protein was overproduced in E. coli, butremained insoluble unless cells containing EutTprotein were lysed under anoxic conditions usingCHES/glycine buffer pH 9.5, followed byresuspension of EutT from the pelleted materialwith 8 mM CHAPS detergent, and a finalcentrifugation step. After these purifications stepsEutT was substantially enriched (~70%homogeneous) (Fig. 2). However, EutT could notbe resolved from contaminating proteins by columnchromatography. All biochemical assays wereperformed using CHAPS-solubilized EutT.EutT activity as a function of reductant and pH.EutT activity were erratic when KBH4 was used asa reductant to convert cob(III)alamin tocob(I)alamin in vitro. To circumvent this problem,Ti(III)citrate was used as reductant in all cobalaminadenosylation assays. The highest EutT activity was

measured at pH 7 (sp. act. 34 nmol of AdoCbl min-1

mg-1 of protein) and less than 30% active at higherpH (e.g. at pH 7.5, sp. act = 9.7; at pH 8, sp. act =7.3; at pH 8.5, sp. act = 7.3 nmol AdoCbl min-1 mg-

1).Metal requirement and inhibition by Ni2+.Several metals were tested for their ability tostimulate activity of EutT (i.e. Ca2+, Mg2+, Zn2+,Na+, K+). Only Mn2+ ions stimulated EutT enzymeactivity (by 32%) above the no-metal-added control(Fig. 3A). Surprisingly, Ni2+ and Zn2+ ions inhibitedEutT activity (62% and 77% activity, respectively.Fifty percent inhibition was the maximal levelmeasured when the ratio of Ni2+ ion:EutT monomerwas 100:1 [at 100 µM Ni2+] (Fig. 3B). The shape ofthe inhibition curve suggested single-site binding.Cysteinyl residues in in the Cys-rich motif ofEutT are required for activity. EutT has aputative HX7HX3CCX2C motif that resembles thecytochrome oxidase CCX2C dicopper binding siteor a 4Fe/4S cluster that was recently identified inBacillus megaterium CbiX protein (Fig. 4) (28).This cysteine-rich motif of EutT was targeted forsite-directed mutagenesis to determine whethercysteinyl residues Cys79, Cys80 and Cys83 wereimportant for structural integrity, enzyme catalysisor both. Residues were mutated to Ala, and theresulting enzymes were tested for their ability tofunction in vivo and in vitro. None of the allelescoding for EutTC79A, EutTC80A or EutTC83A proteinscomplemented a S. enterica cobA eutT strainJE7180 during growth on ethanolamine (data notshown). Variant proteins were overproduced,purified using the protocol for the EutTWT protein,and their adenosyltransferase activity was assayedin vitro (Table 2). EutTC79A was the most active ofthe three variants tested, with 20% of the wild-typeactivity detected in the detergent-soluble fraction.In contrast, EutTC80A and EutTC83A proteins retainedvery little activity (0.6% and 1.2% activity,respectively). Mutant EutT proteins werepredominantly found in the detergent-insolublepellet, likely misfolded and aggregated (Table 2). We propose that residue Cys79 might be moreimportant for structural integrity than residuesCys80 and Cys83.Substrate specificity.Nucleotide substrate. Under conditions where Cblwas saturating, the K m and kcat for ATP weremeasured at 10 µM and 0.03 s-1, respectively (Fig.5A); Vmax was calculated at 54.5 nM min-1. Order-


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


of-addition experiments revealed a modest increasein the specific activity of the enzyme (17%) when itwas incubated with ATP first (sp. act. = 42 vs. 36nmol AdoCbl min-1 mg-1 of protein). Recognition ofNTPs by EutT was unlike that of CobA or PduO(2,10). Not surprisingly, ATP was the preferrednucleotide substrate (Fig. 6A), but there wereseveral differences between CobA and EutT interms of the nucleotide substrate each enzyme canuse. EutT retained reasonable activity (31%) whendATP was the substrate, while CobA does usedATP as substrate at all. EutT retained 38% activitywhen CTP was used as substrate, while CobA canuse CTP 1.5-fold better than ATP. EutT activitywas poor (<10%) when GTP, UTP and ITP,whereas CobA can use the latter 45%, 125%, and85% as effectively as ATP (10). The sharpestdifference between CobA and EutT was the abilityof the latter to use ADP as substrate, retaining 83%of its activity relative to ATP. CobA does use ADPas substrate.Corrinoid substrate. We also determined the kineticparameters of EutT for Cbl under conditions whereMgATP was saturating. Under such conditions, Km= 4.1 µM, kcat = 0.06 s-1, and Vmax = 105 nM min-1

(Fig. 5B). In vitro, EutT adenosylated cobinamide(Cbi), an incomplete corrinoid lacking the DMBlower ligand. When Cbi was used as substrate, thespecific activity of the enzyme (13 nmol AdoCbimin-1 mg-1 of protein) was 21% of the one measuredwhen Cbl was the substrate (sp. act. = 62 nmolAdoCbl min-1 mg-1 of protein), suggesting a role forthe nucleotide loop in substrate binding. Despite itsin vitro activity, a cobA mutant strain (JE7172,Table 1) expressing the eutT+ allele from a plasmidfailed to grow on ethanolamine or glycerol as solecarbon sources when Cbi was provided in themedium. Clearly, the rate of Cbi adenosylationmeasured in vitro was not enough to satisfy eventhe lowest cellular requirement for Cbl, i.e. the levelneeded for methylation of homocysteine tomethionine by the Cbl-dependent methioninesynthase (MetH) enzyme.EutT cleaves PPPi into PPi and Pi. To date,corrinoid adenosyltransferases (10,29) and SAMsynthase (30) are the only enzymes known torelease tripolyphosphate (PPPi) as a reaction by-product without cleaving it to pyrophosphate (PPi)and orthophosphate (Pi). CobA is strongly inhibitedby PPPi (10), and PduO is not (2). EutT was slightlyinhibited by PPPi (15%) or PPi (20%) when eachcompound was present at 10-fold the concentration

of ATP in the reaction mixture. Pi did not inhibitEutT activity (Fig. 6B). Since PPPi only slightlyinhibited EutT, we investigated whether EutTcleaved PPPi. For this purpose, we used 31P-NMRspectroscopy to determine if PPPi, PPi or Pi werestable reaction by-products. NMR data showed thatwhile ATP and PPPi were present in the completereaction (Fig. 7B), PPi and Pi peaks were enhanced.When Cbl was omitted from the reaction mixture,only PPPi signals were observed (Fig. 7A).Therefore, EutT (unlike CobA and PudO) cleavedPPPi to PPi and Pi, and cleaved ATP to Ado andPPPi in the absence of Cbl.Insights into the identity of the electron transferprotein partner for EutT in vivo. Candidateswithin the e u t operon. We investigated thepossibility that a protein encoded by the eut operonmight be the electron transfer protein for EutT. Wefocused on two genes of unknown function, eutPand eutQ. Chromosomal in-frame deletions of thesegenes were constructed singly or in combination. Ina cobA mutant background, each of these deletionswas tested for their ability to affect EutT-dependentgrowth on ethanolamine. If eutP, eutQ, or theircombination were required for cob(II)balamin cob(I)balamin reduction, mutant strains lackingthese functions would be unable to grow onethanolamine as carbon and energy source, i.e. thephenotype would be identical to that of the cobAeutT strain (JE7180). If EutP and/or EutQ proteinswere not required for other metabolic functions,lack of these functions would not affect growth onethanolamine when wild-type cobA or eutT geneswere present on the chromosome. A eutQ cobAstrain (JE8819) displayed a slightly reduced growthrate on ethanolamine relative to the eutQ+cobAstrain (JE1293) (generation time = 0.26 vs 0.23doublings h-1). However, this minor effect of theeutQ mutation on ethanolamine metabolism wasindependent of cobalamin adenosylation because aeutQ cobA+ strain (JE8816) also had a slightlyreduced growth rate compared to wild-type strain(TR6583) (generation time = 0.24 vs 0.29 doublingsh-1). A eutP cobA strain (JE8818) grew as well ascobA strain JE1293 (data not shown). None of theeutQ, eutP or eutPQ mutant strains were defectivefor growth on ethanolamine as sole carbon andenergy source, as reported for the cobA eutT strain(JE7180) (3). Based on the above data, weconcluded that neither EutP nor EutQ were requiredfor EutT function.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


Candidates outside the eut operon. Given the highlyhydrophobic nature of the EutT protein, wepostulated that a plausible source of electrons forEutT function could be the electron transportsystem (ETS). Various electron donors were testedfor the ability to allow AdoCbl production as afunction of the presence of unresolved lysate,cob(III)balamin, ATP and Mn2+. Lysate from cellsexpressing eutT from an overexpression vectorsupported the conversion of cob(III)balamin toAdoCbl when FMN and NADH were included inthe reaction mixture (Fig. 8). Control experimentsshowed that FMN or NADH alone did not result inEutT-dependent AdoCbl synthesis (data notshown). EutT-dependent activity was detectedregardless of the presence of the eut operon on theS. enterica chromosome, indicating that the EutTfunction did not require interactions with other Eutproteins.

AdoCbl was not synthesized fromcob(III)balamin when purified CHAPS/EutTprotein substituted for EutT-enriched cell-freeextracts not treated with CHAPS under conditionsrequiring FMN and NADH (Fig. 8). AdoCbl wasnot detected when the membrane fraction of thecobA eutT strain JE7180 was added to purifiedCHAPS/EutT protein, suggesting that in thepresence of detergent, factors that allow EutTfunction no longer interacted with EutT.

The electron transfer proteins flavodoxin (FldA)and the 2Fe/2S ferredoxin (Fdx) were tested fortheir ability to reduce cob(II)balamin in the activesite of purified CHAPS/EutT enzyme. The S .e n t e r i c a Ferredoxin(flavodoxin):NADP(H)reductase (Fpr) enzyme was used to reduce FldAand Fdx. While activity was detectable, it was only<1% of the activity measured when Ti(III)citratewas used to reduce EutT (sp. act. = 0.49 vs. 61nmol AdoCbl min-1 mg-1 of protein, respectively).At present, we cannot rule out the participation ofFldA or Fdx in EutT function.

DISCUSSIONThe three classes of CA enzymes known to date,

CobA, PduO and EutT, are not evolutionarilyrelated as shown by their profound differences atthe amino acid sequence level. The biochemicalcharacterization of these enzymes has providedbasic knowledge regarding substrate binding andmechanism of catalysis.ATP binding and by-products of EutT aredifferent than other adenosyltransferases. From

the work reported here we have learned that theEutT enzyme requires the amino group of adenineand the β-phosphate for ATP binding, and that theenzyme cleaves ATP to PPPi and adenosine whenCbl is absent from the active site. Unlike CobA orPduO, EutT cleaves PPPi to PPi and Pi, explainingwhy PPPi is a poor inhibitor of itsadenosyltransferase activity. The elucidation ofthree-dimensional EutT structures could providefurther mechanistic insights on substrate binding.Is EutT located outside or inside themetabolosome? Although the data strongly suggestthat EutT is associated with the cell membrane, thedata available do not distinguish whether EutT isinside or outside the metabolosome. One possibilityis that EutT is part of a complex involving other Eutproteins, that the complex is housed within themetabolosome, and that EutT anchors the latter tothe cell membrane. Alternatively, EutT may beassociated with the cell membrane outside of themetabolosome and may not interact with other Eutproteins at all. The latter possibility is feasiblebecause the metabolosome is not a sealedproteinaceous vessel (31), and AdoCbl coulddiffuse into the metabolosome for use byethanolamine ammonia-lyase. Some bacteria withethanolamine ammonia-lyase genes (eutBC) do nothave a eutT homolog in their chromosome, andpresumably rely solely on the housekeeping CobAenzyme. It is unknown whether these organismsmetabolize ethanolamine as sole carbon and energysource like S. enterica and E. coli, both of whichmaintain eutT within the eut operon.Possible means for the generation of thecob(I)balamin nucleophile for the EutT reaction.The data obtained in these studies shed some lighton the identity of the electron donor of EutT. It isunclear whether ferredoxin (Fdx) or flavodoxin A(FldA) is the physiological electron transfer proteinpartner of EutT. The poor activity of EutT whenFdx or FldA was used to reduce cob(II)balamin tocob(I)balamin may reflect limited productiveinteractions between Fdx or FldA and EutT due tothe detergent needed to solubilize EutT. The factthat some activity was detected with FldA and Fdxwarrants further investigation. On the other had, thephenotypic behavior of strains JE8818 (cobA eutP),JE8819 (cobA eutQ), and JE8820 (cobA eutPQ)strongly suggests that the EutP and EutQ proteinsare not involved in EutT-dependent corrinoidadenosylation, or in cob(II)balamin


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


cob(I)balamin reduction. The function of theEutP,Q proteins remains unknown.

A membrane soluble electron donor may providethe reducing power for the reduction of thecorrinoid substrate of EutT. We note that EutT-dependent adenosylation of cob(III)alamin by cell-free extracts enriched for EutT was achieved in thepresence of NADH and FMN. One plausibleexplanation for these results is that NADH sinks itselectrons into the electron transport system, and asemiquinone from dihydroflavins (FMNH2,FADH2) or dihydroquinone reduces cob(II)balamin

to cob(I)balamin in the active site of EutT. There isa need, however, to explain why NADH and FMNdid not support EutT-dependent corrinoidadenosylation when the reaction mixture containedpurified EutT in lieu of EutT-enriched CFE. Inthinking about this problem, one must consider thatpurified EutT was complexed with a detergent(CHAPS). The fact that EutT adenosylatedchemically-reduced cob(I)balamin, but could notaccept electrons from the NADH/FMN pairsuggests that CHAPS does not affect EutT activityas long as the Co(I)rrinoid substrate is available.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


FOOTNOTES.§ Present address, Department of Microbiology, University of Illinois-Urbana, Urbana, IL 61801.

* This work was supported by grant R01 GM40313 from the NIH to J.C.E.-S. N.R.B. is a Howard HughesPredoctoral Fellow. The authors acknowledge the NMRFAM at the University of Wisconsin-Madison forobtaining 31P-NMR spectra.

1Abbreviations used are: CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]propanesulfonic acid; CHES, 2-(cyclohexylamino)ethanesulfonic acid; TrisCl, tris(hydroxymethyl)aminomethane chloride; ATP, adenosinetriphosphate; ADP, adenosine diphosphate, AMP, adenosine monophosphate; GTP, guanosine triphosphate;CTP, cytosine triphosphate; UTP; uracyl triphosphate; PPPi, triphosphate; PPi, pyrophosphate; Pi,orthophosphate; KPi, potassium phosphate; Cbl, cobalamin; AdoCbl, adenosylcobalamin;HOCbl,hydroxocobalamin; EDTA, ethylenediaminotetraacetic acid; PMSF, phenylmethylsulfonylfluoride;IPTG, isopropylthio-b-D-galactopyranoside; H6Fdx, ferredoxin with a hexahistidine tag fused at its N terminus;FldAH6, flavodoxin A with a hexahistidine tag fused at its C terminus; FPLC, fast protein liquidchromatography.

REFERENCES

1. Suh, S.-J., and Escalante-Semerena, J. C. (1995) J. Bacteriol. 177, 921-9252. Johnson, C. L., Pechonick, E., Park, S. D., Havemann, G. D., Leal, N. A., and Bobik, T. A. (2001) J.

Bacteriol. 183, 1577-15843. Buan, N. R., Suh, S. J., and Escalante-Semerena, J. C. (2004) J Bacteriol 186, 5708-57144. Sheppard, D. E., Penrod, J. T., Bobik, T., Kofoid, E., and Roth, J. R. (2004) J Bacteriol. 186, 7635-76445. Bobik, T. A., Xu, Y., Jeter, R. M., Otto, K. E., and Roth, J. R. (1997) J. Bacteriol. 179, 6633-66396. Stojiljkovic, I., Bäumler, A. J., and Heffron, F. (1995) J. Bacteriol. 177, 1357-13667. Kofoid, E., Rappleye, C., Stojiljkovic, I., and Roth, J. (1999) J Bacteriol. 181, 5317-53298. Escalante-Semerena, J. C., Suh, S.-J., and Roth, J. R. (1990) J. Bacteriol. 172, 273-2809. Stich, T. A., Buan, N. R., Escalante-Semerena, J. C., and Brunold, T. C. (2005) J Am Chem Soc 127,

8710-871910. Fonseca, M. V., Buan, N. R., Horswill, A. R., Rayment, I., and Escalante-Semerena, J. C. (2002) J. Biol.

Chem. 277, 33127-3313111. Bauer, C. B., Fonseca, M. V., Holden, H. M., Thoden, J. B., Thompson, T. B., Escalante-Semerena, J.

C., and Rayment, I. (2001) Biochemistry 40, 361-37412. Fonseca, M. V., and Escalante-Semerena, J. C. (2001) J. Biol. Chem. 276, 32101-3210813. Suh, S.-J., and Escalante-Semerena, J. C. (1993) Gene 129, 93-9714. Leal, N. A., Park, S. D., Kima, P. E., and Bobik, T. A. (2003) J. Biol. Chem. 278, 9227-923415. Leal, N. A., Olteanu, H., Banerjee, R., and Bobik, T. A. (2004) J. Biol. Chem. 279, 47536-4754216. Saridakis, V., Yakunin, A., Xu, X., Anandakumar, P., Pennycooke, M., Gu, J., Cheung, F., Lew, J. M.,

Sanishvili, R., Joachimiak, A., Arrowsmith, C. H., Christendat, D., and Edwards, A. M. (2004) J. Biol.Chem. 279, 23646-23653

17. Havemann, G. D., Sampson, E. M., and Bobik, T. A. (2002) J Bacteriol. 184, 1253-126118. Brinsmade, S. R., Paldon, T., and Escalante-Semerena, J. C. (2005) J. Bacteriol. 187, 8039-804619. Bertani, G. (1951) J. Bacteriol. 62, 293-30020. Bertani, G. (2004) J. Bacteriol. 186, 595-60021. Berkowitz, D., Hushon, J. M., Whitfield, H. J., Roth, J., and Ames, B. N. (1968) J. Bacteriol. 96, 215-

22022. Balch, W. E., and Wolfe, R. S. (1976) Appl. Environ. Microbiol. 32, 781-79123. Guzman, L.-M., Belin, D., Carson, M. J., and Beckwith, J. (1995) J. Bacteriol. 177, 4121-413024. Ryu, J.-I., and Hartin, R. J. (1990) Biotechniques 8, 43-4525. Laemmli, U. K. (1970) Nature 227, 680-685


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


26. Bianchi, V., Reichard, P., Eliasson, R., Pontis, E., Krook, M., Jörnvall, H., and Haggård-Ljungquist, E.(1993) J. Bacteriol. 175, 1590-1595

27. Kakuta, Y., Horio, T., Takahashi, Y., and Fukuyama, K. (2001) Biochemistry 40, 11007-1101228. Brindley, A. A., Raux, E., Leech, H. K., Schubert, H. L., and Warren, M. J. (2003) J. Biol. Chem. 278,

22388-2239529. Johnson, C. L., Buszko, M. L., and Bobik, T. A. (2004) J. Bacteriol. 186, 7881-188730. Markham, G. D., Hafner, E. W., Tabor, C. W., and Tabor, H. (1980) J. Biol. Chem. 255, 9082-909231. Kerfeld, C. A., Sawaya, M. R., Tanaka, S., Nguyen, C. V., Phillips, M., Beeby, M., and Yeates, T. O.

(2005) Science 309, 936-93832. Castilho, B. A., Olfson, P., and Casadaban, M. J. (1984) J. Bacteriol. 158, 488-49533. Datsenko, K. A., and Wanner, B. L. (2000) Proc. Natl. Acad. Sci. USA 97, 6640-6645


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


Table 1. Strains and plasmids used in this study. Unless otherwise indicated, all strains and plasmids were constructedduring the course of this work.

Name Genotype ReferenceE.coliBL21 (λDE3) F- dcm ompT hsdS(rB

- mB-) gal λ(DE3) New England

BiolabsS. entericaTR6583 metE205 ara-9 J. Roth via K.

SandersonDerivatives of TR6583JE1096 cobA343::MudI1734(kan+)a Lab collectionJE1293 cobA366::Tn10d(cat+) Lab collectionJE8292 cobA343::MudI1734(kan+) zfa3648*Tn10*zfa3649

(Δeut)JE7172 cobA366::Tn10d(cat+) / pEUT7JE7179 cobA366::Tn10d(cat+) eutE18::MudI1734(kan+) Lab collectionJE7180 cobA366::Tn10d(cat+) eut1141(ΔeutT)b Lab collectionJE8815 eutP1168 (ΔeutP)b

JE8816 eutQ1169 (ΔeutQ)b

JE8817 eutPQ1170 (ΔeutPQ)b

JE8818 cobA343::MudI1734(kan+) eutP1168 (ΔeutP)JE8819 cobA343::MudI1734(kan+) eutQ1168 (ΔeutQ)JE8820 cobA343::MudI1734(kan+) eutPQ1169 (ΔeutPQ)

Plasmid Genotype Protein ReferencepEUT7 bla+ ParaBAD-eutT+ EutTWT (3)pEUT36 bla+ eutT1171 EutTC79A

pEUT37 bla+ eutT1172 EutTC80A

pEUT38 bla+ eutT1173 EutTC83A

pEUT55 bla+ ParaBAD-eutP+ EutPWT

pEUT56 bla+ ParaBAD-eutQ+ EutQWT

pEUT57 bla+ ParaBAD-eutPQ+ EutPQWT

pEUT58 bla+ ParaBAD-eutPQT+ EutPQTWT

pFPR1 bla+ fpr+ FprWT Lab collectionpFLDA4 bla+ fldA+ FldAH6 (3)pFDX1 bla+ fdx+ H6Fdx

a Referred to in the text as MudJ (32).

b In-frame deletion constructed using described methodology (33).


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


Table 1s. List of plasmids and primers used to construct them.

Plasmid (Genotype, protein) Primers

pEUT12 (PT7-eutT+) (EutTWT) EutTNdeIf (5’GGCCTGCAAACTGGCAATCCATATGAACG3’)

EutTXhoIr2 (5’GGGCGCGTTCAACTCGAGTGGCTTCTCTCCCAACCGTTG3’)

pEUT25 (PT7-eutT1165) (EutTC79A) EutTC79Af (5’GCCGCCTGTGAACTGTGCCGCCAG3’)

EutTC79Ar (5’TTCACAGGCGGCTTGTGGGTGCGTATCGCTACTG3’)

pEUT26 (PT7-eutT1166) (EutTC80A) EutTC80Af (5’GCCTGTGCAGAACTGTGCCGCCAGCCG3’)

EutTC80Ar (5’CAGTTGTGCACAGGCCTGTGGATGCGTATCGCTACTGGA3’)

pEUT27 (PT7-eutT1167) (EutTC83A) EutTC83Af (5’GAACTTGCACGTCAGCCGGTGGTGAAAAAGCG3’) and

EutTC83Ar (5’CTGACGTGCAAGTTCGCAGCAGGCCTGTGGGTG3’)

pEUT35 (ParaBAD- eutT1165) (EutTC79A) EutTEcoRIf (5’GTACGTCGCCTGGAATTCAAACTGGC3’)

EutTXbaIr2 (5’GCGCATCTAGAGAAAGACGACTCTGGC3’)

pEUT36 (ParaBAD- eutT1165) (EutTC80A) Same as pEUT35

pEUT37 (ParaBAD- eutT1165) (EutTC83A) Same as pEUT35

pEUT55 (ParaBAD-eutP+) EutPEcoRIf (5’ ATATAAGAATTCTCGATGGCGGTGATGAAACGTATTGCT3’)

EutPXbaIr (5’ TTTTTTTCTAGATTAGCTGTGATAAGTTTTTTC3’)

pEUT56 (ParaBAD-eutQ+) EutQEcoRIf (5’AACAGGAGGAGGGAATTCATGAAAAAACTTATAACAGCT3’)

EutQSalIr (5’AAAAAAGTCGACTCATACGGATTGCCAGTTTG3’)

pEUT57 (ParaBAD-eutPQ+) EutPEcoRIf (5’ ATATAAGAATTCTCGATGGCGGTGATGAAACGTATTGCT3’)

EutQSalIr (5’AAAAAAGTCGACTCATACGGATTGCCAGTTTG3’)

pEUT58 (ParaBAD-eutPQT+) EutPEcoRIf (5’ ATATAAGAATTCTCGATGGCGGTGATGAAACGTATTGCT3’)

EutTXbaIr2 (5’GCGCATCTAGAGAAAGACGACTCTGGC3’)

pFDX1 (pET-16b fdx+) (H6Fdx) SefdxNdeIf (5’ TTTTTTCATATGCCGAAGATTGTTATTCTGCC3’)

SefdxBamHIr (5’ TTTTTTGGATCCTTAGTGTTCACGCGCATGGTTGATTGT3’)


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


Table 2. ATP:Cob(I)alamin adenosyltransferase assays with mutant EutT proteins

Protein Fraction Specific Activitya %WTWT CHAPS soluble 61 ± 0.36 100

Whole lysate 1.4 ± 0.25 2CHAPS soluble 12 ± 1.5 20

C79A

Insoluble fractionb 14 ± 0.49 23

Whole lysate 0.04 ± 0.06 < 0.1CHAPS soluble 0.37 ± 0.11 0.6

C80A

Insoluble fractionb 0.71 ± 0.61 1

Whole lysate 0.37 ± 0.02 0.6CHAPS soluble 0.75 ± 0.02 1

C83A

Insoluble fractionb 0.28 ± 0.04 0.5

a The average of three determinations using Ti(III)citrate as reductant.

b Measured after solubilization with CHAPS.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


FIGURE LEGENDS

Figure 1. Contributions of the CobA and EutT ATP:Co(I)rrinoid adenosyltransferases to the pool of AdoCbl inS. enterica.

Figure 2. Expression and isolation of EutTWT protein. Proteins from cell fractions were separated by SDS-PAGE (25). Numbers on the left of the gel represent protein mass markers (kDa).

Figure 3. Metal requirements and Ni2+ inhibition of EutT adenosyltransferase activity. Panel A shows theenhancement of cobalamin adenosyltransferase activity as a function of added Mn2+ ions. Panel B shows EutTactivity inhibition by Ni2+ ions; the shape of the curve suggests competition for a single site.

Figure 4. Comparison of the cysteine-rich motif of EutT and the 4Fe/4S cluster motif of B. subtilis CbiX. EutThas three cysteines (bold) and two histidines (outlined) that seemed reminiscent of the CCX2C dicopper bidingmotif of cytochrome oxidase and/or the MXCX2C 4Fe/4S motif of B. subtilis CbiX. Residues C79, C80 and C83of EutT were changed to alanine to assess their importance in EutT function.

Figure 5. Kinetics of ATP and Cbl binding to EutT. Panel A. Cbl was held at 50 µM while ATPconcentration was varied. The Km

ATP = 10 µM, kcatATP = 0.03 s-1 and the Vmax = 54 nmoles min-1 mg-1. In B, ATP

was held at 500 µM while Cbl concentration was varied. The KmCbl = 4.1 µM , kcat

Cbl = 0.06 s-1, and Vmax = 105nmoles min-1. Insets show Lineweaver-Burk plots.

Figure 6. NTP substrate analogs and phosphate product inhibition of EutT. Panel A. NTPs tested in lieu ofATP; all NTPs were at 0.5 mM. Panel B. PPPi, PPi or Pi were added to 10X the amount of ATP present in thereaction to assess product inhibition. For ATP, the sp. act = 61 nmol min-1 mg-1.

Figure 7. 31P-NMR spectra of EutT reaction products. Panel A. Reaction mixture lacking Cbl. Panel B.Complete reaction mixture containing EutT, ATP and Cbl. Panels C, D. PPPi and ATP standards, respectively.Panel E. Mixture of ATP, PPPi, PPi, Pi and HOCbl standards, without EutT.

Figure 8. NADH, FMN-dependent adenosylation of Cbl by EutT. Cell-free extracts of several strains of S.enterica were prepared and 50 µg of total protein was used in adenosylation assays. Product was detected bychange in the absorbance at 525 nm after photolysis. pVOC, vector-only control; pEUT7, peutT+. Severalculture media were tested. Cell-free extracts of strains used in these studies were (from left to right): strainJE1293 (cobA eutT+) grown in NCE minimal medium containing ethanolamine (30 mM), and HOCbl (1mM);strains JE7204 (cobA eutT / pVOC) and JE7205 (cobA eutT / peutT+) grown in NCE minimal mediumcontaining ethanolamine (30 mM) AdoCbl (1 mM) and L(+)-arabinose (200 mM); strains JE8941 (cobA eut /pVOC) and JE8942 (cobA eut / peutT+) grown in NCE minimal medium containing glycerol (30 mM), L(+)-arabinose (200 mM), and methionine (0.5 mM). The high level of HOCbl or AdoCbl was used to allow forfaster growth and higher yield. In all cases, reaction mixtures contained FMN (50 µM), NADH (1 mM) andATP (1 mM).


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Co(I)rrinoid + ATP AdoCo(III)rrin + PPPi AdoCbl

Co(II)Cbl

Co(III)CblCo(I)Cbl

Ado

H2O

e-

e-

ATP

Pi + PPiEutT

ATP:Co(I)rrinoid adenosyltransferase (CobA)

Figure 1, Buan & Escalante-Semerena


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

21

29

34

50

77103

EutT

kDa

+CHAPS



ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

0 50 100 1500

20

40

60

Sp

ecifi

c A

ctiv

ityA

doC

bl n

mol

es m

in-1 m

g-1

[Mn+2] mM

A

B

0.0 0.1 0.2 0.3 0.4 0.5

20

30

Sp

ecifi

c A

ctiv

ityA

doC

bl n

mol

es m

in-1 m

g-1

[Ni+2] mM



ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

EutT PQPV GLTSSDT PQACCELCRQPVVKKP7980 837567

CbiX 258 263

MNCDTC

Figure 4, Buan & Escalante-Semerena by guest on January 4, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

0.0 0.1 0.2 0.3 0.4 0.5

20

40

60

0 50 100 150 200

0.02

0.04

0.06

Sp

ecifi

c A

ctiv

ityA

doC

bl n

mol

es m

in-1 m

g-1

[ATP] mM

1/V

1/[ATP] mM

0 20 40

20

40

60

80

0.0 0.5 1.0

0.02

0.04

0.06

1/[Cbl] mM

Sp

ecifi

c A

ctiv

ityA

doC

bl n

mol

es m

in-1 m

g-1

1/V

[Cbl] mM

B

A



ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

60

80

100

120

% S

pec

ific

Act

ivity

Ad

oCb

l nm

oles

min

-1 m

g-1

None

PPPiPPi Pi

Phosphate (1 mM)

BAS

pec

ific

Act

ivity

NC

bl n

mol

es m

in-1 m

g-1

ATP

ADPAM

PdAT

PGTP CTP UTP IT

P

Substrate (0.5 mM)

0

25

50

75

100

dGTPdCTP

dTTP

Ado

Figure 6, Buan & Escalante-Semerena by guest on January 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/



ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

cobA eu

tT+

0

1

2

3

4

5

6

cobA eu

tT /

peutT

+

cobA eu

tT /

pVOC

cobA ∆eu

t / p

eutT

+

cobA ∆eu

t / p

VOC

Ad

oCb

l (nm

ol)

Figure 8, Buan & Escalante-Semerena by guest on January 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Nicole R. Buan and Jorge C. Escalante-Semerenaadenosyltransferase(EutT) enzyme of salmonella enterica

Purification and initial biochemical characterization of the ATP:cob(I)alamin

published online April 24, 2006J. Biol. Chem.

10.1074/jbc.M603069200Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

http://www.jbc.org/content/suppl/2006/06/15/M603069200.DC1


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M603069200

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;M603069200v1&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2006/04/24/jbc.M603069200.citation

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=early/2006/04/24/jbc&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2006/04/24/jbc.M603069200.citation

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2006/06/15/M603069200.DC1

http://www.jbc.org/

purification and initial biochemical characterization of ... filepurification and initial...

Documents