Portal de Periódicos da CAPES

Propionyl-Coenzyme A Synthase from Chloroflexus aurantiacus, a Key Enzyme of the 3-Hydroxypropionate Cycle for Autotrophic CO2 Fixation

2002; Elsevier BV; Volume: 277; Issue: 14 Linguagem: Inglês

10.1074/jbc.m110802200

1083-351X

Birgit E. Alber, Georg Fuchs,

Enzyme Catalysis and Immobilization

The 3-hydroxypropionate cycle has been proposed as a new autotrophic CO2 fixation pathway for the phototrophic green non-sulfur eubacterium Chloroflexus aurantiacus and for some chemotrophic archaebacteria. The cycle requires the reductive conversion of the characteristic intermediate 3-hydroxypropionate to propionyl-CoA. The specific activity of the 3-hydroxypropionate-, CoA-, K+-, and MgATP-dependent oxidation of NADPH in autotrophically grown cells was 0.09 μmol min−1 mg−1 protein, which was 2-fold down-regulated in heterotrophically grown cells. Unexpectedly, a single enzyme catalyzes the entire reaction sequence: 3-hydroxypropionate + MgATP + CoA + NADPH + H+ → propionyl-CoA + MgAMP + PPi + NADP+ + H2O. The enzyme was purified 30-fold to near homogeneity and has a very large native molecular mass between 500 and 800 kDa, with subunits of about 185 kDa as judged by SDS-PAGE, suggesting a homotrimeric or homotetrameric structure. Upon incubation of this new enzyme, termed propionyl-CoA synthase, with the proteinase trypsin, the NADPH oxidation function of the enzyme was lost, whereas the enzyme still activated 3-hydroxypropionate to its CoA-thioester and dehydrated it to acrylyl-CoA. SDS-PAGE revealed that the subunits of propionyl-CoA synthase had been cleaved once and the N-terminal amino acid sequences of the two trypsin digestion products were determined. Two parts of the gene encoding propionyl-CoA synthase (pcs) were identified on two contigs of an incomplete genome data base of C. aurantiacus, and the sequence of the pcs gene was completed. Propionyl-CoA synthase is a natural fusion protein of 201 kDa consisting of a CoA ligase, an enoyl-CoA hydratase, and an enoyl-CoA reductase, the reductase domain containing the trypsin cleavage site. Similar polyfunctional large enzymes are common in secondary metabolism (e.g. polyketide synthases) but rare in primary metabolism (e.g. eukaryotic type I fatty acid synthase). These results lend strong support to the operation of the proposed pathway in autotrophic CO2 fixation. The 3-hydroxypropionate cycle has been proposed as a new autotrophic CO2 fixation pathway for the phototrophic green non-sulfur eubacterium Chloroflexus aurantiacus and for some chemotrophic archaebacteria. The cycle requires the reductive conversion of the characteristic intermediate 3-hydroxypropionate to propionyl-CoA. The specific activity of the 3-hydroxypropionate-, CoA-, K+-, and MgATP-dependent oxidation of NADPH in autotrophically grown cells was 0.09 μmol min−1 mg−1 protein, which was 2-fold down-regulated in heterotrophically grown cells. Unexpectedly, a single enzyme catalyzes the entire reaction sequence: 3-hydroxypropionate + MgATP + CoA + NADPH + H+ → propionyl-CoA + MgAMP + PPi + NADP+ + H2O. The enzyme was purified 30-fold to near homogeneity and has a very large native molecular mass between 500 and 800 kDa, with subunits of about 185 kDa as judged by SDS-PAGE, suggesting a homotrimeric or homotetrameric structure. Upon incubation of this new enzyme, termed propionyl-CoA synthase, with the proteinase trypsin, the NADPH oxidation function of the enzyme was lost, whereas the enzyme still activated 3-hydroxypropionate to its CoA-thioester and dehydrated it to acrylyl-CoA. SDS-PAGE revealed that the subunits of propionyl-CoA synthase had been cleaved once and the N-terminal amino acid sequences of the two trypsin digestion products were determined. Two parts of the gene encoding propionyl-CoA synthase (pcs) were identified on two contigs of an incomplete genome data base of C. aurantiacus, and the sequence of the pcs gene was completed. Propionyl-CoA synthase is a natural fusion protein of 201 kDa consisting of a CoA ligase, an enoyl-CoA hydratase, and an enoyl-CoA reductase, the reductase domain containing the trypsin cleavage site. Similar polyfunctional large enzymes are common in secondary metabolism (e.g. polyketide synthases) but rare in primary metabolism (e.g. eukaryotic type I fatty acid synthase). These results lend strong support to the operation of the proposed pathway in autotrophic CO2 fixation. A new autotrophic CO2 fixation cycle termed 3-hydroxypropionate cycle has been proposed for the phototrophic green non-sulfur bacterium Chloroflexus aurantiacus (1.Eisenreich W. Strauss G. Werz U. Fuchs G. Bacher A. Eur. J. Biochem. 1993; 215: 619-632Crossref PubMed Scopus (68) Google Scholar, 2.Herter S. Farfsing J. Gad′on N. Rieder C. Eisenreich W. Bacher A. Fuchs G. J. Bacteriol. 2001; 183: 4305-4316Crossref PubMed Scopus (68) Google Scholar, 3.Holo H. Arch. Microbiol. 1989; 151: 252-256Crossref Scopus (106) Google Scholar, 4.Holo H. Sirevåg R. Arch. Mirobiol. 1986; 145: 173-180Crossref Scopus (114) Google Scholar, 5.Strauss G. Eisenreich W. Bacher A. Fuchs G. Eur. J. Biochem. 1992; 205: 853-866Crossref PubMed Scopus (73) Google Scholar, 6.Strauss G. Fuchs G. Eur. J. Biochem. 1993; 215: 633-643Crossref PubMed Scopus (165) Google Scholar). This bacterium belongs to the family Chloroflexaceae, which represents one of five major lineages of phototrophic bacteria (7.Xiong J. Fischer W.M. Inoue K. Nakahara M. Bauer C.E. Science. 2000; 289: 1724-1730Crossref PubMed Scopus (387) Google Scholar). C. aurantiacus grows optimally at 55 °C under heterotrophic conditions but can also grow in mineral salt medium with CO2 as sole carbon source (4.Holo H. Sirevåg R. Arch. Mirobiol. 1986; 145: 173-180Crossref Scopus (114) Google Scholar, 8.Pierson B.K. Castenholz R.W. Arch. Microbiol. 1974; 100: 5-24Crossref PubMed Scopus (291) Google Scholar, 9.Pierson B.K. Castenholz R.W. Arch. Microbiol. 1974; 100: 283-305Crossref Scopus (124) Google Scholar, 10.Sirevåg R. Castenholz R.W. Arch. Microbiol. 1979; 120: 151-153Crossref Scopus (28) Google Scholar). In thermal springs, filamentous Chloroflexus sp. and cyanobacteria form microbial mats that probably thrive photoautotrophically (11.van der Meer M.T.J. Schouten S. de Leeuw J.W. Ward D.M. Environ. Microbiol. 2000; 2: 428-435Crossref PubMed Scopus (87) Google Scholar). Indications for the operation of a similar pathway in autotrophic CO2 fixation have been obtained for acidophilic archaebacteria of the phylum Crenarchaeota, such as Acidianus brierleyi, Metallosphaera sedula, and Sulfolobus metallicus (12.Burton N.P. Williams T.D. Norris P.R. Arch. Microbiol. 1999; 172: 349-353Crossref PubMed Scopus (28) Google Scholar, 13.Ishii M. Miyake T. Satoh T. Sugiyama H. Oshima Y. Kodama T. Igarashi Y. Arch. Microbiol. 1997; 166: 368-371Crossref Scopus (70) Google Scholar, 14.Ménendez C. Bauer Z. Huber H. Gad′on N. Stetter K.O. Fuchs G. J. Bacteriol. 1999; 181: 1088-1098Crossref PubMed Google Scholar). The proposed 3-hydroxypropionate cycle is shown below in Fig. 1. Each turn of the cycle results in the net fixation of two molecules of bicarbonate into one molecule of glyoxylate. Acetyl-CoA is carboxylated to malonyl-CoA by conventional ATP-dependent acetyl-CoA carboxylase. In vitro, a straightforward reductive conversion of malonyl-CoA to propionyl-CoA was observed when NADPH, K+, CoA, and MgATP were added to cell extract (6.Strauss G. Fuchs G. Eur. J. Biochem. 1993; 215: 633-643Crossref PubMed Scopus (165) Google Scholar). This finding was surprising, because this conversion formally requires five enzymatic reactions and 3-hydroxypropionate as a free intermediate (see Fig. 1). 3-Hydroxypropionate formation is characteristic for the cycle; this metabolite is even excreted when cell growth becomes limited (3.Holo H. Arch. Microbiol. 1989; 151: 252-256Crossref Scopus (106) Google Scholar). Recently, we have purified the enzyme reducing malonyl-CoA and have shown that it is a bifunctional enzyme that requires two NADPH and forms 3-hydroxypropionate (15.Hügler, M., Ménendez, C., Schägger, H., and Fuchs, G. (2002) J. Bacteriol., in pressGoogle Scholar); 3-hydroxypropionate semialdehyde is an intermediate in this process. This new enzyme, malonyl-CoA reductase, has both alcohol and aldehyde dehydrogenase (CoA acylating) activity; the two partial reactions catalyzed are shown below in Fig. 1. The subsequent reductive conversion of 3-hydroxypropionate to propionyl-CoA formally requires three enzymatic reactions. The first step, activation to 3-hydroxypropionyl-CoA, requires MgATP; the product of ATP hydrolysis, ADP or AMP, has not been identified. This knowledge is important for estimating the energy need for autotrophic CO2 fixation. So far, a CoA ligase catalyzing this reaction was assumed. The second step is the dehydration of 3-hydroxypropionyl-CoA to acrylyl-CoA by an enoyl-CoA hydratase. The third step is the NADPH-specific reduction of acrylyl-CoA to propionyl-CoA by an enoyl-CoA reductase. This study originally aimed at purifying the first of these enzymes, the presumptive CoA ligase, converting 3-hydroxypropionate to its CoA-thioester. Much to our surprise, an enzyme is present in cell extracts of C. aurantiacus, which catalyzes the overall reductive conversion of 3-hydroxypropionate to propionyl-CoA. The enzyme was purified and characterized, and the gene encoding this new enzyme was identified. C. aurantiacus strain OK-70-fl (DSM 636) was grown at 55 °C anaerobically under photoautotrophic conditions on H2/CO2 (80%/20% (v/v)) or photoheterotrophically as described previously (15.Hügler, M., Ménendez, C., Schägger, H., and Fuchs, G. (2002) J. Bacteriol., in pressGoogle Scholar). The cells were harvested at an optical density (578 nm) of 3.5–4.0 and stored in liquid nitrogen until use. 3-Hydroxypropionate was obtained by hydrolysis of 3-hydroxypropionitrile as described previously (2.Herter S. Farfsing J. Gad′on N. Rieder C. Eisenreich W. Bacher A. Fuchs G. J. Bacteriol. 2001; 183: 4305-4316Crossref PubMed Scopus (68) Google Scholar). Two spectrophotometric assays were used. (a) In the standard assay, the ATP-, CoA-, and NADPH-dependent reduction of 3-hydroxypropionate was measured at 55 °C. The standard reaction mixture (0.5 ml) contained 100 mm Tris/HCl (pH 7.8), 2 mm DTE, 1The abbreviations used are: DTEdithioerythritolTristris(hydroxymethyl)aminomethaneMOPS2-(N-morpholino)propanesulfonic acidTAPSN-tris(hydroxymethyl)methyl-3-aminopropanesulfonic acidHPLChigh performance liquid chromatographyORFopen reading frameACSacyl-CoA synthetaseECHenoyl-CoA hydrataseECRenoyl-CoA reductase 5 mm MgCl2, 10 mm KCl, 3 mm ATP, 0.5 mm CoA, 0.4 mm NADPH, and protein. (b) Formation of ADP or AMP was measured using a coupled assay at 45 °C. The reaction mixture (0.5 ml) contained 100 mm Tris/HCl (pH 8.5), 2 mm DTE, 2 mm MgCl2, 100 mm KCl, 3 mm ATP, 0.5 mm CoA, 0.3 unit of myokinase, 0.1 unit of pyruvate kinase, 0.3 unit of lactate dehydrogenase, 0.5 mm NADH (or 0.25 mm NADH plus 0.25 mm NADPH), and protein. Both reactions were started by the addition of 1 mm 3-hydroxypropionate and followed at 365 nm (εNAD(P)H = 3490 m−1cm−1). Protein concentrations were determined by the method of Bradford (16.Bradford M. Anal. Biochem. 1976; 72: 248-254Crossref PubMed Scopus (216440) Google Scholar) using bovine serum albumin as a standard. dithioerythritol tris(hydroxymethyl)aminomethane 2-(N-morpholino)propanesulfonic acid N-tris(hydroxymethyl)methyl-3-aminopropanesulfonic acid high performance liquid chromatography open reading frame acyl-CoA synthetase enoyl-CoA hydratase enoyl-CoA reductase All procedures were done aerobically at 4 °C. Thawed cell paste (10 g wet weight) of autotrophically grown cells was suspended in 15–20 ml of buffer A (100 mm Tris/HCl (pH 7.8), 5 mmMgCl2) containing 0.25 mg of DNase I and passed twice through a chilled French pressure cell at 138 kPa. The cell lysate was centrifuged at 12,000 × g for 15 min, and the supernatant was recentrifuged at 100,000 × g for 1 h. The cell extract was incubated for 30 min at 63 °C and centrifuged at 20,000 × g for 15 min. Saturated ammonium sulfate solution was added to the supernatant to a final concentration of 10% (NH4)2SO4 and centrifuged at 100,000 × g for 1 h. The supernatant from step ii was loaded onto a 25-ml Phenyl-Sepharose column (Amersham Biosciences, Inc.) equilibrated with buffer A containing 200 mm ammonium sulfate. After an 80-ml wash, the column was developed with a 250-ml decreasing linear gradient of 200–0 mm ammonium sulfate at 1 ml min−1. The peak of activity was eluted between 120 and 80 mm salt, and the pooled fractions were concentrated using an ultrafiltration unit fitted with a 10-ml membrane cell (cutoff, 10 kDa; Filtron, Karlstein, Germany). The concentrated enzyme solution from step iii was applied onto an 8-ml MonoQ HR 10/10 anion-exchange column (Amersham Biosciences, Inc.) equilibrated with buffer B (25 mm MOPS/NaOH (pH 7.2), 5 mm MgCl2). The column was washed with 15 ml of buffer B and developed with a linear gradient of 0–1 m NaCl at 1 ml min−1. Propionyl-CoA synthase activity eluted between 300 and 340 mm salt. Active fractions were pooled and frozen at −20 °C. Aliquots of 1 ml from step iv were applied onto a 120-ml Superdex 200 HiLoad 16/60 column (Amersham Biosciences, Inc.) equilibrated with 20 mmMOPS/NaOH (pH 7.2), 100 mm NaCl. The column was developed at a flow rate of 1 ml min−1. The combined active fractions were concentrated as described above, and glycerol and DTE were added to a final concentration of 20% and 1 mm, respectively, and stored at −20 °C. The Km values of CoA, ATP, 3-hydroxypropionate or acrylate, and NADPH of the ATP-, CoA-, and NADPH-dependent reduction of 3-hydroxypropionate or acrylate were determined using propionyl-CoA synthase, obtained after the MonoQ chromatography step, and the standard reaction assay (a). The concentration of one substrate was varied, while keeping the concentration of the other substrates constant; the concentrations given were saturating. The stoichiometry of the reaction was determined by changing the concentration of either ATP, CoA, or 3-hydroxypropionate of the standard reaction assay to concentrations between 10 μm and 0.1 mm. After the reaction had come to completion, the amount of NADPH consumed was determined by measuring the overall absorption change at 365 nm. The stoichiometry of the 3-hydroyxpropionate-dependent formation of AMP was determined using assay (b) in the presence of NADH and NADPH. The pH optimum at 55 °C of the ATP-, CoA-, and NADPH-dependent reduction of 3-hydroxypropionate was determined by using the standard reaction assay (a) with 100 mm instead of 10 mm KCl and, instead of 100 mm Tris/HCl, the following buffers at 100 mmconcentrations (pH at 55 °C): MOPS/NaOH pH 6.8–7.3, HEPES/NaOH pH 7.2–8.1, TAPS/NaOH pH 7.7–8.9. The thermostability of the enzyme was determined by incubating the enzyme at 30, 40, 50, 60, 65, 72, 77, and 90 °C for 15 min. The enzyme solution was cooled to 4 °C before determining the activities relative to those of samples kept at 4 °C throughout the experiment. K+ and Mg2+dependencies were measured by omitting KCl or MgCl2 from the standard reaction assay and adding them back at defined concentrations. The specificity of the ATP-, CoA-, and NADPH-dependent reduction of 3-hydroxypropionate toward its substrate was determined by substituting 3-hydroxypropionate in the standard reaction assay by 10 mm neutralized 3-hydroxybutyrate, crotonate, acrylate, β-alanine, or by 1 mm glycolate and malonate; ATP by 3 mm GTP and UTP; NADPH by 0.4 mm NADH. An optical absorption spectrum of the purified enzyme (1.3 mg ml−1 in 20 mmMOPS/NaOH, pH 7.5, 5 mm MgCl2, 300m KCl, 10% (w/v) glycerol) was collected at 25 °C using a PerkinElmer Life Sciences Lambda 2S spectrometer and the same buffer as a blank. To an aliquot of 200 μl of purified propionyl-CoA synthase (1.0 mg ml−1) 320 μl of 10 mm ammonium bicarbonate buffer (pH 8.0), containing 20 μg of trypsin (Sigma), was added. To a second aliquot 320 μl of 10 mmammonium bicarbonate buffer (pH 8.0), and no trypsin, was added and used as a control ("undigested"). After incubation at 37 °C for 30 min, 20 μl of 1 mg ml−1 trypsin inhibitor was added. The activity of ATP-, CoA-, and 3-hydroyxpropionate-dependent oxidation of NADH and/or NADPH and the stoichiometry of 3-hydroxypropionate used per NAD(P)H supplied, was measured using assay (b), containing either 0.5 mm NADPH or 0.5 mm NADH or 0.25 mm NADPH plus 0.25 mm NADH. An optimized reaction mixture (0.6 ml), based on the standard reaction assay (a), containing 100 mm Tris/HCl (pH 8.5), 2 mm DTE, 2 mm MgCl2, 100 mm KCl, 2 mm ATP, 0.5 mm CoA, 0.7 mm NADPH, and 25 μg of purified propionyl-CoA synthase was used. The enzymatic reaction was started by the addition of 1 mm 3-hydroxypropionate and stopped at different time points by transferring 100 μl of the reaction mixture to 3 μl of 25% HCl. CoA-thioesters were analyzed by reversed-phase HPLC using an RP-C18 column (LiChrospher 100, endcapped, 5 μm, 125 × 4 mm; Merck, Darmstadt, Germany). A 35-min gradient from 2% to 10% (v/v) acetonitrile, in 50 mm potassium phosphate-buffer, pH 6.7, at a flow rate of 1 ml min−1, was used. All CoA-esters were detected at 260 nm. Propionyl-CoA and acrylyl-CoA were synthesized enzymatically at 37 °C (100 mm Tris/HCl, pH 8.5, 2 mm MgCl2, 20 mm KCl, 3 mm phosphoenolpyruvate, 0.5 mm NADH, 2 mm ATP, 0.5 mm CoA, 0.2 unit of myokinase, 0.1 unit of pyruvate kinase, 0.1 unit of lactate dehydrogenase, 0.1 unit of acetyl-CoA synthetase (Sigma), and 20 mm neutralized propionate or acrylate, reaction time: 3 min for propionate and 5 min for acrylate) and used as standards. SDS-PAGE (7%) was performed as described previously (17.Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (207231) Google Scholar). The native molecular mass was determined on a 125-ml Sephacryl S-300 (Amersham Biosciences, Inc.) gel filtration column calibrated with ovalbumin (45 kDa), catalase (250 kDa), ferritin (450 kDa), and thyroglobulin (669 kDa). To an aliquot of 200 μl of purified propionyl-CoA synthase (0.6 mg ml−1) 320 μl of 10 mm ammonium bicarbonate buffer (pH 8.0), containing 20 μg of trypsin (Sigma), was added. After incubation at 37 °C for 30 min, trichloroacetic acid was added to a final concentration of 6% (w/v). The sample was centrifuged at 10,000 × g for 10 min, and the pellet was resuspended in 50 μl of 0.1 m NaOH. The trypsin-digested protein fragments were separated by 8% SDS-PAGE and transferred to an Immobilon-Psq transfer membrane (Millipore, Bedford, MA). N-terminal sequencing was performed by TopLab (Martinsried, Germany) using an Applied Biosystems Procise 492 sequencer (Weiterstadt, Germany). The phenylthiohydantoin derivatives were identified with an on-line Applied Biosystems Analyzer 140 C. Small scale chromosomal DNA from C. aurantiacus was isolated using a standard technique (18.Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Current Protocols in Molecular Biology. John Wiley and Sons, New York1987: 2.4.1-2.4.2Google Scholar). Two 23-mer oligonucleotides (primer echB NdeI, 5′-TTCATCATATGGCCGAAGAGCAG-3′, partially corresponding to nucleotides 11841–11855 of contig 1024 of the genome data base of C. aurantiacus(www.jgi.doe.gov/JGI_microbial/html/chloroflexus/chloro_homepage.html) and primer echE EcoRI, 5′-ACAGCGAATTCGGGTCAACCACT-3′, partially corresponding to nucleotides 445–459 of contig 799 of the same genome data base), 1 μg of chromosomal C. aurantiacus DNA, 2.5 units of Pwo polymerase (Peqlab, Erlangen, Germany), and the Peqlab DNA amplification kit were used to amplify a 974-bp genomic region, which is part of the pcs gene. The PCR product was purified (19.Vogelstein B. Gillespie D. Proc. Natl. Acad. Sci. U. S. A. 1979; 76: 615-619Crossref PubMed Scopus (1067) Google Scholar) and sequenced at the sequencing facility at the University of Freiburg (Germany). The BLAST program (20.Altschul S.F. Madden T.L. Schäffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Nucleic Acids Res. 1997; 25: 3389-3402Crossref PubMed Scopus (59929) Google Scholar) was used to search protein data bases at the National Center for Biotechnology Information (Bethesda, MD), the local C. aurantiacus server (spider.jgi-psf.org/JGI_microbial/bin/psf_blast?PROJECT_ID = 2351478) at the DOE Joint Genome Institute (University of California, CA), and the Conserved Domain data base (Pfam) at Washington University (St. Louis, MO). Chemicals were obtained from Fluka (Neu-Ulm, Germany), Merck (Darmstadt, Germany), Sigma-Aldrich (Deisenhofen, Germany), or Roth (Karlsruhe, Germany); biochemicals were from Roche Diagnostics (Mannheim, Germany), Applichem (Darmstadt, Germany), or Gerbu (Craiberg, Germany). Extracts of autotrophically grown cells of C. aurantiacus catalyzed the 3-hydroxypropionate-dependent oxidation of NADPH, provided that Mg2+, ATP, K+, and CoA were added. The specific activity at 55 °C was 0.09 μmol min−1mg−1 protein; this activity was 2-fold down-regulated in heterotrophically grown cells. The enzyme catalyzing the reductive transformation of 3-hydroxypropionate was purified from 10 g (wet weight) of autotrophically grown cells (Table I). An initial heat precipitation step was essential to remove large amounts of pigments. Three chromatographic steps yielded a nearly homogeneous protein, as judged by SDS-PAGE (Fig. 2). The enzyme eluted as a symmetrical peak from the gel filtration column. The overall recovery was 16% and the enrichment about 30-fold, which means that this protein amounts to at least 3% of the soluble cell protein. Comparison of the soluble protein profile of cells grown hetero- versus autotrophically (Fig. 2) shows that the amount of enzyme is elevated under autotrophic growth condition, consistent with the higher specific activity measured under these conditions.Table IPurification of propionyl-CoA synthase from 10 g (wet weight) of autotrophically grown C. aurantiacusStepTotal activityProteinSpecific activityRecoveryPurificationμmol min−1mgμmol min−1mg−1%-foldCell extract37.44100.091(100)(1)30 min at 63 °C, 10% (NH4)2SO429.22800.10781.1Phenyl-Sepharose21.3171.35714MonoQ15.37.42.14123Gel filtration6.12.42.51628 Open table in a new tab SDS-PAGE of the purified enzyme showed a single protein band of ∼185 kDa (Fig. 2). Gel filtration chromatography of the native protein gave a molecular mass of 500–800 kDa. This indicates that the enzyme is a homotrimer or homotetramer. The UV-visible spectrum (250–800 nm) of the purified enzyme showed a single peak at 280 nm, with no indication for the presence of a chromophoric cofactor (data not shown). The enzyme was stable when incubated for 15 min at temperatures up to 65 °C and during storage in the presence of 20% glycerol and 1 mmDTE at −20 °C for a few weeks. The stoichiometry of the reaction was determined. Per mole of 3-hydroxypropionate added, 1 mol of NADPH was oxidized. The same stoichiometry was observed for ATP and for CoA. The products of the reaction were determined in different ways. HPLC analysis showed that propionyl-CoA was formed. To determine whether AMP plus pyrophosphate or ADP plus phosphate was formed from ATP, a coupled spectrophotometric assay was used. In the presence of NADH and NADPH the rate of NAD(P)H oxidation was increased by 190% over the rate of NADPH oxidation alone; this effect was strictly dependent on the addition of myokinase. Furthermore, the stoichiometry of NAD(P)H oxidized per mole of 3-hydroxypropionate added was 2.9:1. This shows ATP was hydrolyzed to AMP and pyrophosphate. The overall reaction catalyzed by the enzyme is as follows, 3­hydroxypropionate+MgATP+CoA+NADPH+H+→propionyl­CoA+MgAMP+PPi+H2O+NADP+ REACTION1 The enzyme is referred to as propionyl-CoA synthase. The enzyme showed high affinities for its substrates. The reaction followed Michaelis-Menten kinetics with apparent Km values of 15 μm for 3-hydroxypropionate, 10 μm for NADPH, 50 μm for ATP, and 10 μm for CoA. The enzyme is also very specific; it does not act on glycolate, malonate, β-alanine, 3-hydroxybutyrate, or crotonate. Acrylate was transformed at the same rate as 3-hydroxypropionate, albeit with a much higher apparent Km value (0.5 mm). NADH could not substitute for NADPH, whereas ATP (100%) could be substituted by GTP (24%) and UTP (20%). Activity was dependent on K+, and half-maximal activity was observed at 4 mm KCl. The acrylate-dependent oxidation of NADPH in the presence of ATP and CoA was also dependent on K+, indicating that the CoA ligase step is K+-dependent. Monovalent cations with similar ionic radii (Rb+, NH4+, and Cs+) can partially substitute for K+ (95%, 85%, and 20%). There is a clear correlation of decrease in activity with the increase in ionic radius, relative to that of K+, indicating that this K+ dependence is not a general ionic strength effect. The pH dependence at 55 °C of the propionyl-CoA synthase catalyzed reaction shows a rather narrow pH optimum around pH 8 (Fig. 3). Between pH 6.8 and 7.9 the specific activity of the propionyl-CoA synthase increased 30-fold, indicating that a functional group in the enzyme with an estimated p Ka of 7.3 has to be deprotonated for maximum turnover. The vmax value was 4 μmol min−1 mg−1, corresponding to a turnover of 12 s−1 per subunit of 185 kDa. Product formation from 3-hydroxypropionate in the presence or absence of NADPH was studied to detect possible intermediates. In the absence of NADPH an unknown CoA-thioester (compound X) and acrylyl-CoA were formed (Fig. 4B). After addition of 1 mm NADPH, propionyl-CoA was formed (Fig. 4C), and the amount of compound X and acrylyl-CoA decreased. The consumption of compound X and acrylyl-CoA upon addition of NADPH suggests that both are intermediates of the reaction, and it is, therefore, likely that compound X represents 3-hydroxypropionyl-CoA. The assignment of compound X as 3-hydroxypropionyl-CoA is also consistent with its elution time after CoA and before propionyl-CoA. Native (undenatured) propionyl-CoA synthase was digested with the proteinase trypsin. SDS-PAGE analysis indicated that the subunits of the enzyme were principally cleaved once, each subunit yielding two polypeptides, with estimated molecular masses of 150 and 40 kDa (Fig. 5). The trypsin-digested propionyl-CoA synthase was no longer able to catalyze the ATP-, CoA-, and 3-hydroxypropionate-dependent oxidation of NADPH (Table II). By coupling the formation of 1 mol of AMP to the oxidation of 2 mol of NADH using myokinase, phosphoenolpyruvate, pyruvate kinase, and lactate dehydrogenase, it was shown that the trypsin-digested enzyme, however, still catalyzed the 3-hydroxypropionate- and CoA-dependent hydrolysis of ATP (Table II). The rate of 3-hydroxypropionate-, ATP-, and CoA-dependent formation of AMP for the trypsin-digested enzyme (0.5 μmol min−1 mg−1) was even increased over the rate of the same reaction catalyzed by undigested propionyl-CoA synthase (0.34 μmol min−1mg−1). In the presence of NADPH and trypsin-digested enzyme, only traces of propionyl-CoA were formed, whereas compound X was the major CoA-thioester formed (Fig. 4E). Acrylyl-CoA was also formed (Fig. 4E), indicating that the trypsin-digested propionyl-CoA synthase retained 3-hydroxypropionate-CoA ligase and acrylyl-CoA hydratase activity, whereas the acrylyl-CoA reductase activity was almost completely lost.Table IIRates of the ATP-, CoA-, 3-hydroxypropionate-dependent oxidation of NADPH and/or NADH by propionyl-CoA synthase and stoichiometry of NAD(P)H used per 3-hydroxypropionate suppliedPropionyl-CoA synthaseNADPHNADHSpecific activityStoichiometryaRatio is 3-hydroxypropionate used:NAD(P)H oxidized.mmmmμmol NAD(P)H oxidized min−1 mg−1Undigested0.501.51:1.000.50.671:2.20.250.254.31:2.8Trypsin digest0.50<0.1NAbNot applicable, because no activity was detected under these conditions.00.51.01:2.20.250.251.11:2.0The assay was performed in the presence of myokinase, pyruvate kinase, lactate dehydrogenase, and phosphoenolpyruvate at 45 °C to couple AMP formation to the oxidation of 2 NADH.a Ratio is 3-hydroxypropionate used:NAD(P)H oxidized.b Not applicable, because no activity was detected under these conditions. Open table in a new tab The assay was performed in the presence of myokinase, pyruvate kinase, lactate dehydrogenase, and phosphoenolpyruvate at 45 °C to couple AMP formation to the oxidation of 2 NADH. The N-terminal sequences of both polypeptides of the trypsin-digested propionyl-CoA synthase were determined as MIDTA for the 150-kDa polypeptide and ASTLLAAGAH for the 40 kDa polypeptide. A search of the data base of the almost completed genome of C. aurantiacus at the DOE Joint Genome Institute revealed a perfect match of the N-terminal sequence of the 150-kDa polypeptide to a putative protein encoded by an incomplete open reading frame (ORF) on contig 1024. A potential ribosome binding site (GGAGGA) ends 7 nucleotides upstream of the ATG codon for the N-terminal methionine. It is, therefore, concluded, that the MIDTA sequence represents the true N terminus of the propionyl-CoA synthase. The N-terminal sequence of the second (40 kDa) polypeptide revealed a perfect match to an internal sequence of a putative protein encoded by an incomplete ORF on contig 799. Based on the deduced amino acid sequence of this ORF, the N-terminal alanine of the 40-kDa polypeptide is preceded by an arginine residue in the native protein, which itself is preceded by four consecutive serine residues. This indicates a likely cleavage site for the proteinase trypsin. The missing DNA sequence between the two contigs was completed and deposited in the GenBankTM data base (accession number AF445079, Fig. 6). The gene encoding propionyl-CoA synthase from C. aurantiacus is referred to as pcs. The pcs gene encodes a protein with 1822 amino acids and a calculated molecular weight of 201,404. This value is within the estimated subunit molecular mass of 185 kDa for the purified protein (Fig. 2). The calculated molecular masses of the two polypeptides, after cleavage with trypsin, are 159 and 42 kDa, respectively. We have described a new enzyme, propionyl-CoA synthase, which functions in autotrophic CO2 fixation in C. aurantiacus and possibly other prokaryotes. Propionyl-CoA synthase catalyzes the irreversible reductive conversion of 3-hydroxypropionate to propionyl-CoA. The reaction consists of three partial reactions as shown in Fig. 1. The properties of the enzyme are summarized in Table III.Table IIIMolecular and catalytic properties of the new enzyme propionyl-CoA synthase from C. aurantiacusPropertySubstrates3-Hydroxypropionate, ATP, CoA, NADPHCocatalystsK+, Mg2+ProductsPropionyl-CoA, AMP + PPi, NADP+Specific activity4 μmol min−1 mg protein−1Apparent Km3-Hydroxypropionate 15 μm, CoA 10 μm, ATP 50 μm, NADPH 10 μmpH optimum8.1 (55 °C)Native molecular mass500–800 kDaSubunit molecular mass185 ± 10 kDaSuggested compositionα3 or α4Catalytic number12 s−1 (per subunit)Specificity3-Hydroxypropionate 100%, acrylate 100%, glycolate <1%, malonate <1%, β-alanine <1%, 3-hydroxybutyrate <1%, crotonate <1%, NADPH 100%, NADH <1%, ATP 100%, GTP 24%, UTP 20%Influence of cations (mm added)No divalent ions added 6%, Mg2+ (5) 100%, Mn2+ (5) 80%, Ca2+(5) 43%No monovalent ions added 3%, K+ (20) 100%, Rb+ (20) 95%, NH4+ (20) 85%, Cs+ (20) 30%, Li+ (20) 5% Open table in a new tab Propionyl-CoA synthase is a trifunctional enzyme. It belongs to the following enzyme classes: EC 6.2.1., ligases forming carbon–sulfur bonds (acid-thiol ligases); EC 4.2.1., carbon–oxygen lyases (hydro-lyases); and EC 1.3.1., oxidoreductases acting on the CH–CH groups of donors with NAD(P)+ as acceptor. The first partial reaction corresponds to organic acid-CoA ligase (AMP-forming), the second one to enoyl-CoA hydratase, and the third partial reaction to NADPH-dependent enoyl-CoA reductase. A search of protein data bases indicated that propionyl-CoA synthase consists of three protein domains (Fig. 6). Residues 18–850 of the deduced amino acid sequence of pcs showed significant sequence similarities to several acetyl-CoA and short fatty acid acyl-CoA synthetases from various organisms and was, therefore, named acyl-CoA synthetase (ACS) domain. It showed highest percent sequence identity (35%) to an acetyl-CoA synthetase from Pyrobaculum aerophilum (GenBankTM accession number AAD09253). A multiple alignment of several homologous sequences indicated an untypical insertion of about 110 amino acids in the ACS domain of propionyl-CoA synthase, located within a consensus AMP binding domain (Pfam 00501) identified in all aligned sequences. The acyl-CoA synthetase (3-hydroxypropionate-CoA ligase) activity of propionyl-CoA synthase is K+-dependent, as has been shown for the acetyl-CoA synthetase from animal tissue (21.von Korff R.W. J. Biol. Chem. 1953; 203: 265-271Abstract Full Text PDF PubMed Google Scholar). The enoyl-CoA hydratase (ECH) domain (Pcs amino acid residues 858–1051) showed 43% sequence identity to the 3-hydroxybutyryl-CoA dehydratase (crotonase) of Clostridium acetobutylicum(GenBankTM accession number AAK80658) and significant sequence similarity to other members of the enoyl-CoA hydratase family (Pfam 00378). Both glutamate residues shown to be involved in the enoyl-CoA hydratase-catalyzed reaction (22.Hofstein H.A Feng Y. Anderson V.E. Tonge P.J. Biochemistry. 1999; 38: 9508-9516Crossref PubMed Scopus (60) Google Scholar) are conserved in propionyl-CoA synthase (E975 and E995). The enoyl-CoA reductase (ECR) domain (Pcs amino acid residues 1201–1816) is placed within the zinc-binding dehydrogenase family of proteins (Pfam 00107), however, the zinc-binding site is not conserved. An untypical insertion of about 180 amino acids is present in propionyl-CoA synthase compared with other members of this family, where this insertion is absent. Conserved residues of a NAD(P)H consensus binding motif (G XG X2A X3A) correspond to Gly-1415 to Ala-1424 of propionyl-CoA synthase. The highest sequence similarity (overall 28% sequence identity) was found with an alcohol dehydrogenase of Methylobacterium extorquens (GenBankTM accession number AAB26986) involved in an unknown acetyl-CoA to gloxylate pathway (23.Chistoserdova L.V. Lidstrom M.E. Microbiology. 1996; 142: 1459-1468Crossref PubMed Scopus (40) Google Scholar). The ECR domain is also homologous to NADPH-dependent crotonyl-CoA reductase of Streptomyces collinus (24.Wallace K.K. Bao Z.Y. Dai H. Digate R. Schuler G. Speedie M.K. Reynolds K.A. Eur. J. Biochem. 1995; 233: 954-962Crossref PubMed Scopus (55) Google Scholar). The ECR domain has no significant sequence similarity to FAD-dependent acyl-dehydrogenases involved in fatty acid oxidation nor to enoyl-ACP reductases of the dissociated type II fatty acid synthases of prokaryotes. Significant similarity of the ECR domain of propionyl-CoA synthase was observed with the enoyl-ACP reductase domain of fatty acid type II synthases, however, this similarity only consists of a short region centered around the NAD(P)H consensus binding motif. It is this ECR domain that contains the site responsible for the cleavage of native propionyl-CoA synthase into two polypeptides by the proteinase trypsin (Figs. 5 and 6). Upon incubation of the enzyme with trypsin, propionyl-CoA synthase lost its ability to reduce acrylyl-CoA to propionyl-CoA almost completely (Table II and Fig. 4E). The residual activity is likely due to the incomplete digestion of propionyl-CoA synthase in the presence of trypsin (Fig. 5). The enzyme still showed 3-hydroxypropionate-CoA ligase and acylyl-CoA hydratase activity, consistent with the ACS and ECH domains staying mostly intact upon digestion of propionyl-CoA synthase with trypsin. The rate of the CoA- and 3-hydroxypropionate-dependent formation of AMP for the undigested enzyme in the absence of NADPH is 23% of the same rate in the presence on NADPH (Table II). A possible reason for the decrease in rate in the absence of NADPH is the slow release of 3-hydroxypropionyl-CoA or acrylyl-CoA from the enzyme. This is consistent with the increased rate for AMP formation of the trypsin-digested enzyme, where a portion of the holoenzyme is missing, thus yielding possibly to a more open path for product release. The region between the ECH and ECR domains of propionyl-CoA synthase (residues 1132–1164) has 47% sequence identity to the linker region between the enoyl-CoA hydratase and 3-hydroxyacyl-CoA dehydrogenase of the fatty acid oxidation complex (α subunit) of Vibrio cholerae (GenBankTM accession number AAF94206). There is, however, no entry in the protein data bases that represents a homolog to propionyl-CoA synthase having all three functional domains located on a single polypeptide. From a chemical point of view the reaction catalyzed by the enzyme is irreversible. The activation of the substrate to the CoA thioester and water elimination are reversible processes. However, pyrophosphate, the product of the CoA ligase reaction, is rapidly hydrolyzed in most cells, shifting the equilibrium to the product side. Notably, the NADPH-dependent reduction of enoyl-CoA compounds is irreversible. The standard potential of the NADPH/NADP+couple is −0.32 V; because in growing cells the NADPH/NADP+ ratio is normally ≥1, this value might be even slightly more negative under cellular conditions. The standard redox potential of the acrylyl-CoA/propionyl-CoA couple is +69 mV (25.Sato K. Nishina Y. Setoyama C. Miura R. Shiga K. J. Biochem. (Tokyo). 1999; 26: 668-675Crossref Scopus (38) Google Scholar). Thus, the NADPH-dependent reduction of propionyl-CoA cannot be reversed. Reduction of 3-hydroxypropionate to propionyl-CoA is part of the proposed 3-hydroxy-propionate cycle. The enzyme fulfills the proposed role in all respects. First, the specific activity is high enough in autotrophically growing cells, which requires a specific activity for enzymes involved in the CO2 fixation cycle of at least 0.015 μmol−1 mg−1 protein (15.Hügler, M., Ménendez, C., Schägger, H., and Fuchs, G. (2002) J. Bacteriol., in pressGoogle Scholar). Second, the enzyme activity and protein levels are up-regulated in autotrophically versus heterotrophically growing cells. Third, propionyl-CoA synthase is highly specific for its substrates with Km values in the physiological range. Fourth, as a fusion protein catalyzing an entire reaction sequence of the pathway, leading from the key intermediate 3-hydropropionate to the CO2 acceptor propionyl-CoA, the enzyme is highly efficient. Its rate is only determined by the concentration of its diffusible substrates not in addition by the concentration of its intermediates. Furthermore, acrylyl-CoA, due to its double bond a reactive intermediate, is rapidly converted by the enzyme and, therefore, does not accumulate in the cell. These results show that propionyl-CoA synthase is a key enzyme in the 3-hydroxypropionate cycle of autotrophic CO2 fixation. It remains to be shown whether, using the 3-hydroxypropionate cycle, a similar trifunctional enzyme exists in other autotrophs, or whether a multifunctional enzyme complex or even separately functioning enzymes are operating. Very special thanks to N. Gad'on, Freiburg, for growing cells. Thanks are also due to M. Kies of TopLab, München, for N-terminal sequence analyses and G. Igloi, Freiburg, for DNA sequencing. We also thank the reviewer for fruitful and constructive suggestions.