|Home | About | Journals | Submit | Contact Us | Français|
Pantothenate kinase (CoaA) catalyzes the first step of the coenzyme A (CoA) biosynthetic pathway and controls the intracellular concentrations of CoA through feedback inhibition in bacteria. An alternative enzyme found in archaea, pantoate kinase, is missing in the order Thermoplasmatales. The PTO0232 gene from Picrophilus torridus, a thermoacidophilic euryarchaeon, is shown to be a distant homologue of the prokaryotic type I CoaA. The cloned gene clearly complements the poor growth of the temperature-sensitive Escherichia coli CoaA mutant strain ts9, and the recombinant protein expressed in E. coli cells transfers phosphate to pantothenate at pH 5 and 55°C. In contrast to E. coli CoaA, the P. torridus enzyme is refractory to feedback regulation by CoA, indicating that in P. torridus cells the CoA levels are not regulated by the CoaA step. These data suggest the existence of two subtypes within the class of prokaryotic type I CoaAs.
Coenzyme A (CoA) is an essential cofactor synthesized from pantothenate (vitamin B5), cysteine, and ATP (1, 20, 30). The thiol group derived from the cysteine moiety in a CoA molecule forms a thioester bond, which is a high-energy bond, with carboxylates including fatty acids. The resulting compounds are called acyl-CoAs (CoA thioesters) and function as the major acyl group carriers in numerous metabolic and energy-yielding pathways. Since it is thought that the pantetheine moiety in CoA existed when life first came about on Earth (25) and at present, a CoA, acyl-CoA, or 4′-phosphopantethein moiety that is common to CoA and acyl carrier proteins is utilized by about 4% of all enzymes as a substrate (6), these compounds are thought to play a crucial role in the earliest metabolic system.
Bacteria, fungi, and plants can produce pantothenate, which is the starting material of CoA biosynthesis, although animals must take it from their diet (41). The canonical CoA biosynthetic pathway consists of five enzymatic steps: i.e., pantothenate kinase (CoaA in prokaryotes and PanK in eukaryotes; EC 220.127.116.11), phosphopantothenoylcysteine synthetase (CoaB; EC 18.104.22.168), phosphopantothenoylcysteine decarboxylase (CoaC: EC 22.214.171.124), phosphopantetheine adenylyltransferase (CoaD; EC 126.96.36.199), and dephospho-CoA kinase (CoaE; EC 188.8.131.52). The organisms belonging to the domains Bacteria and Eukarya have this pathway (20, 30). CoaB, CoaC, CoaD, and CoaE are detectable in the complete genome sequences as orthologs of the counterparts from E. coli and humans (15, 16, 32). However, there is diversity among the CoaAs and PanKs, depending on their primary structures, and to date, three types of CoaA in bacteria and one type of PanK in eukaryotes have been identified. CoaAs and PanK catalyze the phosphorylation of pantothenate to produce 4′-phosphopantothenate at the first step of the pathway. First, the Escherichia coli CoaA (CoaAEc) was cloned as a prokaryotic type I CoaA after characterization of the properties enzymatically (42-44, 48). Thereafter, the eukaryotic PanK isoforms were isolated from Aspergillus nidulans (AnPanK), mice (mPanK), and humans (hPanK) (10, 17, 28, 29, 33, 34, 54-56). These enzyme activities were clearly regulated by end products of the biosynthetic pathway such as CoA, acetyl-CoA, and malonyl-CoA, and the pantothenate kinases governed the intracellular concentrations of CoA and acyl-CoAs (10, 17, 28, 29, 33, 34, 43, 44, 48, 54, 55). However, CoaAs insensitive to CoA and acyl-CoAs were recently identified from Staphylococcus aureus (CoaASa), Pseudomonas aeruginosa (CoaAPa), and Helicobacter pylori (CoaAHp) as prokaryotic type II and III CoaAs (9, 11, 18, 27). The structural and functional diversity among pantothenate kinases suggests that they are key indicators of the regulation of the CoA biosynthesis. In archaea neither CoaA nor pantothenate synthetase (PanC; EC 184.108.40.206), which catalyzes the condensation of pantoate and β-alanine to produce pantothenate, had been identified biochemically until very recently. COG1829 and COG1701 were assigned as the respective candidates based on comparative genomic analysis (15). COG1701 was reported to be PanC (36), and later the enzyme was revised to phosphopantothenate synthetase, which catalyzed the condensation of phosphopantoate and β-alanine (52). Together with the identification of COG1701, COG1829 was found to be pantoate kinase, responsible for the phosphorylation of pantoate (52). Homologues of pantoate kinase and phosphopantothenate synthetase are found in most archaeal genomes, thus establishing a noncanonical CoA biosynthetic pathway involving the two novel enzymes. However, homologues of the two novel enzymes are missing in the order Thermoplasmatales.
Hence, we proceeded with a search for the kinase genes of the remaining archaea to elucidate the regulatory mechanism(s) underlying archaeal CoA biosynthesis. The PTO0232 gene in the complete genome sequence of Picrophilus torridus was identified as encoding a distant homologue of CoaAEc by a BLAST search. The recombinant protein phosphorylated pantothenate, but the activity was not inhibited at all by CoA or CoA thioesters despite its classification as prokaryotic type I CoaA. This functional difference between P. torridus CoaA (CoaAPt) and CoaAEc can be accounted for by an amino acid substitution at position 247 which possibly interacts with CoA. Here we describe the existence of a second subtype in the class of prokaryotic type I CoaAs.
Materials were purchased from the following suppliers. Restriction enzymes, PrimeSTAR HS DNA polymerase, the Mighty TA-cloning kit for PrimeSTAR including pMD20-T vector, and pG-Tf2 were from Takara Bio, Inc., Shiga, Japan. pET-28a(+) was from Novagen, Madison, WI. E. coli BL21(DE3) and BL21-CodonPlus(DE3)-RIPL were from Stratagene, La Jolla, CA. Ampicillin (Ap), kanamycin (Km), chloramphenicol (Cm), streptomycin (Sm), tetracycline (Tc), phenylmethylsulfonyl fluoride (PMSF), isopropyl-1-thio-β-d-galactopyranoside (IPTG), 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal), and pyrophosphate were from Wako Pure Chemical Industries, Ltd., Osaka, Japan. Oligonucleotides for PCR and lysozyme from chicken egg white, CoA, acetyl-CoA, malonyl-CoA, ADP, AMP, GTP, CTP, and UTP were from Sigma-Aldrich, Inc., St. Louis, MO. Ni-Sepharose 6 Fast Flow and Sephacryl S-200 HR were from GE Healthcare Bio-Sciences Corp., Piscataway, NJ. d-[1-14C]pantothenate (specific activity, 55 mCi/mmol) was from American Radiolabeled Chemicals, Inc., St. Louis, MO. DE81 filters were from Whatman International, Ltd., Maidstone, Kent, England. ScintiSafe 30% was from Fisher Scientific, Dubuque, IA. Bradford dye-binding protein assay solution was from Bio-Rad Laboratories, Inc., Hercules, CA. DNase I was from bovine pancreas, and ATP was from Roche Diagnostics GmbH, Mannheim, Germany. Other chemicals used were of reagent grade.
Picrophilus torridus NBRC 100828 (the same as strain DSM 9790), of which the genome DNA sequence has already been published from the NCBI database (14), was obtained from the NITE Biological Resource Center (NBRC) for a template of PCR. The E. coli strains used in this work were strain JM109 (recA1 endA1 gyrA96 [thi hsdR17 supE44 relA1 λ− Δ(lac-proAB) F′(traD36 proAB+ lacIq lacZΔM15)]) for a general host of DNA manipulations, strain BL21(DE3) [F− ompT hsdSB(rB− mB−) gal(λcI857 ind1 Sam7 nin5 lacUV5-T7 gene 1) dcm (DE3)] for the expression of the archaeal coaA gene, and strain ts9 (leuB6 hisG1 argG6 metB1 rplL9 rts-1 ilu-1 lacY1 gal-6 xyl-7 mtl-2 malA1 tonA2 tsx-1 λR λ− supE44) (13), which was a temperature-sensitive mutant of pantothenate kinase and showed poor growth at 37°C, for the confirmation of CoaA function biologically. For the expression of the recombinant CoaA from P. torridus, an E. coli strain was constructed by introducing the 4.7-kb pSC101-based plasmid with the argU, ileY, and leuW tRNA genes that was prepared from E. coli BL21-CodonPlus(DE3)-RIPL, together with the plasmid pG-Tf2 with the groES, groEL, and tig chaperone genes under Pzt-1 promoter control, into E. coli BL21(DE3) to obtain E. coli BL21(DE3)-RIL-GT (5). The pSC101-based plasmid and pG-Tf2 conferred Smr and Cmr on the prepared E. coli cells, respectively. E. coli cells carrying pMD20-T vector, pET-28a(+), the 4.7-kb pSC101-based plasmid, and/or pG-Tf2 were grown aerobically in Luria-Bertani (LB) broth (37) with 50 μg/ml Ap, 30 μg/ml Km, 50 μg/ml Sm, and/or 25 μg/ml Cm at 37°C or 30°C, respectively.
DNA manipulation was performed according to the methods described by Sambrook and Russell (37). The nucleotide sequence of the pantothenate kinase gene was determined on both strands by the dideoxy-chain termination method of Sanger et al. (38) using an ABI PRISM 3130xl genetic analyzer (Applied Biosystems Japan, Ltd., Tokyo, Japan). The DNA sequence and the deduced amino acid sequence were examined with the sequence analysis programs of GENETYX software (Software Development Co., Tokyo, Japan) and ClustalW (45). Homology searches in the databases were carried out using the BLAST program (4).
A total of 73 sequences (72 CoaA sequences and a uridine kinase sequence from E. coli [NCBI gi 90111379]) containing 274 amino acids were aligned and used for phylogenetic analyses. Phylogenetic trees were constructed using the neighbor-joining (NJ) method and the maximum likelihood (ML) method with a JTT model. The uridine kinase was used as an outgroup in the phylogenetic tree construction. Bootstrap analysis was performed 1,000 replicates to estimate the confidence of tree topology. The following reference CoaA sequences were used, with the NCBI gi numbers given in parentheses: Escherichia coli K-12 strain MG1655 (gi 16131808), Shigella flexneri 5 strain 8401 (gi 110807829), Salmonella enterica serovar Typhimurium LT2 (gi 16767398), Citrobacter koseri ATCC BAA-895 (gi 157147230), Klebsiella pneumoniae subsp. pneumoniae MGH 78578 (gi 152972836), Enterobacter sakazakii ATCC BAA-894 (gi 156935824), Yersinia frederiksenii ATCC 33641 (gi 238790607), Serratia proteamaculans 568 (gi 157368518), Pectobacterium atroseptica SCRI1043 (gi 50119175), Photorhabdus luminescens subsp. laumondii TTO1 (gi 37528548), Sodalis glossinidius strain “morsitans” (gi 85058104), Aeromonas hydrophila subsp. hydrophila ATCC 7966 (gi 117621012), Actinobacillus pleuropneumoniae L20 (gi 126208977), Vibrio cholerae V52 (gi 121730117), Mannheimia succiniciproducens MBEL55E (gi 52426243), Moritella sp. strain PE36 (gi 149910606), Pasteurella multocida subsp. multocida strain Pm70 (gi 15603612), Vibrionales bacterium SWAT-3 (gi 148982168), Haemophilus influenzae PittAA (gi 145635947), “Candidatus Blochmannia pennsylvanicus” strain BPEN (gi 71891972), Shewanella pealeana ATCC 700345 (gi 157959999), Psychromonas ingrahamii 37 (gi 119947224), Baumannia cicadellinicola strain Hc (gi 94677049), Chloroflexus aggregans DSM 9485 (gi 219849400), Azorhizobium caulinodans ORS 571 (gi 158426120), Stigmatella aurantiaca DW4/3-1 (gi 115373059), Rhodopseudomonas palustris BisA53 (gi 115522392), Parvibaculum lavamentivorans DS-1 (gi 154251682), Janibacter sp. strain HTCC2649 (gi 84497885), Myxococcus xanthus DK 1622 (gi 108763280), Labrenzia aggregata IAM 12614 (gi 118593706), Nitrobacter hamburgensis X14 (gi 92115760), Xanthobacter autotrophicus Py2 (gi 154246229), Rhodococcus jostii RHA1 (gi 111022820), Bradyrhizobium japonicum USDA 110 (gi 27375767), Kineococcus radiotolerans SRS30216 (gi 152964692), Methylobacterium chloromethanicum CM4 (gi 218530728), Nocardia farcinica IFM 10152 (gi 54026791), Coxiella burnetii Dugway 5J108-111 (gi 154705789), Streptomyces coelicolor A3(2) (gi 21223117), Thermobifida fusca YX (gi 72163009), Saccharopolyspora erythraea NRRL 2338 (gi 134098560), Mesorhizobium sp. strain BNC1 (gi 110635827), Corynebacterium glutamicum ATCC 13032 (gi 19552218), Mycobacterium tuberculosis H37Ra (gi 167969225), Rhizobium leguminosarum bv. viciae 3841 (gi 116249806), Fulvimarina pelagi HTCC2506 (gi 114707306), Aurantimonas sp. strain SI85-9A1 (gi 90420981), Agrobacterium tumefaciens strain C58 (gi 159184134), Clavibacter michiganensis subsp. michiganensis NCPPB 382 (gi 148273761), Sinorhizobium medicae WSM419 (gi 150398449), Arthrobacter aurescens TC1 (gi 119964235), Ochrobactrum anthropi ATCC 49188 (gi 153008168), Brucella ovis ATCC 25840 (gi 148558913), Actinomyces odontolyticus ATCC 17982 (gi 154507820), Leifsonia xyli subsp. xyli strain CTCB07 (gi 50955530), Bartonella henselae strain Houston-1 (gi 49475017), Bacillus subtilis subsp. subtilis strain 168 (gi 50812263), Propionibacterium acnes KPA171202 (gi 50843257), Listeria monocytogenes EGD-e (gi 16802963), Ferroplasma acidarmanus fer1 (gi 126008038), Picrophilus torridus DSM 9790 (gi 48477304), Enterococcus faecium DO (gi 69247086), Pediococcus pentosaceus ATCC 25745 (gi 116492210), Lactobacillus plantarum WCFS1 (gi 28377743), Streptococcus pneumoniae R6 (gi 15902785), Lactococcus lactis subsp. lactis Il1403 (gi 15673426), Wigglesworthia glossinidia endosymbiont of Glossina brevipalpis (gi 32491263), Tropheryma whipplei TW08/27 (gi 28572319), Leuconostoc mesenteroides subsp. mesenteroides ATCC 8293 (gi 116618957), Oenococcus oeni PSU-1 (gi 116491576), and Chloroflexus aurantiacus J-10-fl (gi 163847600).
The gene encoding pantothenate kinase from P. torridus (locus tag PTO0232) was amplified by PrimeSTAR HS DNA polymerase using two synthetic primers, 5′-CGCATATGGATTTTTACTTCAGACCTGAT-3′ (positions 263466 to 263489 in GenBank accession no. AE017261) and 5′-CGGTCGACTTATATCCTGAAATAGATCGATTC-3′ (positions 264293 to 264270) containing the NdeI and SalI sites (underlined), respectively. The reaction mixture contained 10 μl of 5× PrimeSTAR buffer with 1 mM MgCl2, four deoxynucleoside triphosphates (dNTPs) at 200 μM each, P. torridus NBRC 100828 cells as a template, the primers at 0.25 μM each, and 1.25 U of PrimeSTAR HS DNA polymerase in a final volume of 50 μl. DNA amplification was performed in a temperature cycler (Thermal Cycler Personal; Takara Bio, Inc.), for 30 cycles consisting of a denaturation step for 10 s at 98°C, an annealing step for 15 s at 55°C, and an elongation step for 1 min at 72°C. The amplified 0.8-kb product was inserted into a pMD20-T vector using the TA-cloning kit, pMD-PTO0232, and the nucleotide sequence of the insert DNA was confirmed on both strands. The insert DNA was digested with NdeI and SalI and then ligated into an expression vector of pET-28a(+) digested with the same restriction enzymes, giving pET-PTO0232. For genetic complementation of the pantothenate kinase temperature-sensitive mutant of E. coli ts9, the KpnI-HindIII fragment of pMD-PTO0232 was ligated under the lac promoter of pUC118. The resulting plasmid, pUC-PTO0232, was used for growth complementation of strain ts9.
E. coli BL21(DE3)-RIL-GT cells carrying pET-PTO0232 were grown aerobically in 2 liters of LB broth containing 30 μg/ml Km, 25 μg/ml Cm, and 10 ng/ml Tc for induction of the chaperones at 30°C. When the turbidity of the culture at 600 nm reached ~0.5, IPTG was added to the culture broth at a concentration of 0.1 mM and the mixture was additionally cultivated for 6 h. The E. coli cells were harvested by centrifugation at 8,000 × g for 15 min and stored at −20°C. The cell pellet was resuspended in 100 ml of 20 mM Tris-HCl (pH 7.9) containing 1.5 M NaCl, 1 mM 2-mercaptoethanol, 1 mM PMSF, 1 mg/ml lysozyme, 10 μg/ml DNase I, and 0.1% Triton X-100. The suspension was frozen at −80°C for 2 h and then thawed on ice, and cell debris was removed by centrifugation at 8,000 × g for 30 min. The supernatant was applied onto a Ni-Sepharose 6 Fast Flow column (1.0 by 13 cm) previously charged with 50 mM NiSO4 (40 ml) and equilibrated in 60 ml of loading buffer, 20 mM Tris-HCl (pH 7.9), 1 mM 2-mercaptoethanol, and 1.5 M NaCl. After the column was first washed with loading buffer and then with loading buffer containing 40 mM imidazole (80 ml), a linear gradient elution was conducted with 40 to 500 mM imidazole in loading buffer, the total volume of the gradient being 100 ml. The 2-ml fractions were taken at a flow rate of 24 ml/h. The fractions with proteins were combined and dialyzed against 2 liters of 50 mM Tris-HCl (pH 7.6) with 1 mM 2-mercaptoethanol. The resulting precipitates were recovered by centrifugation at 8,000 × g for 30 min, and after being washed with the dialysis buffer, the proteins were dissolved in 1.5 ml of 25 mM Tris-HCl (pH 7.6) containing 500 mM ammonium sulfate, 1 mM 2-mercaptoethanol, and 50% glycerol. The enzyme solution was stored at −20°C until use. Proteins were measured by the Bradford method (8), with bovine serum albumin as a standard. In column chromatography, protein elution patterns were usually monitored by absorption at 280 nm. All operations of the purification procedure were carried out at 4°C.
CoaAEc was prepared from E. coli BL21(DE3) cells carrying pET15b/bPanK (10). The cells aerobically grown in 100 ml of LB broth with 100 μg/ml Ap at 30°C were collected by centrifugation at 8,000 × g for 15 min after 6 h of the addition of IPTG at 0.5 mM and then stored at −80°C. The cells were resuspended in 5 ml of 20 mM Tris-HCl (pH 7.9), containing 0.5 M NaCl, 1 mM PMSF, 1 mg/ml lysozyme, 10 μg/ml DNase I, and 0.1% Triton X-100. After being kept at −80°C for 2 h, the suspension was thawed on ice and centrifuged, and the supernatant was subjected to Ni-Sepharose resin previously charged with NiSO4. The matrix was washed with loading buffer composed of 20 mM Tris-HCl (pH 7.9) and 0.5 M NaCl and further with a loading buffer containing 40 mM imidazole. The His-tagged proteins were eluted by the buffer with 200 mM imidazole, and the active fractions were combined. The enzyme solution was loaded onto a Sephacryl S-200 column (2.5 by 115 cm) equilibrated with 50 mM Tris-HCl (pH 7.4), and the 2-ml fractions were taken at a flow rate of 24 ml/h. Fractions with the pantothenate kinase activity were combined, concentrated by ultrafiltration with PLGC membrane (Millipore, Co., Billerica, MA) to 1.65 ml, and stored as 50% glycerol solution at −20°C until use.
Standard assays for CoaAPt contained 91 μM d-[14C]pantothenate (specific activity, 55 mCi/mmol), 2.5 mM ATP, 10 mM MgCl2, 50 mM sodium acetate (pH 5.2), and the enzyme solution in a total volume of 40 μl. The mixture was incubated for 10 min at 55°C, and the reaction was stopped by addition of 4 μl of acetic acid. An aliquot of the mixture was deposited onto a Whatman DE81 ion exchange filter disk, and the filter disk was washed in three changes of 1% acetic acid in 95% ethanol to remove unreacted pantothenate. The produced 4′-phosphopantothenate on the disk was quantified by counting the dried disk in 3 ml ScintiSafe 30%. The recovery of 4′-phosphopantothenate on the filter disks was ~65%, and there is a liner relationship between the amount of 4′-phosphopantothenate formed and the incubation time for at least 15 min under these conditions. In a kinetic analysis of CoaAPt, the substrate solution composed of a 1:9 mixture of radioactive and inactive pantothenate was used. The enzyme activity of CoaAEc was measured by the same method as with the archaeal enzyme, except that 50 mM Tris-HCl (pH 7.5) was used in the mixture and the reaction was conducted at 37°C.
To confirm the function of PTO0232 as pantothenate kinase, pUC-PTO0232 with the gene coding for PTO0232 was electroporated into E. coli ts9 cells (field strength, 12.5 kV/cm; capacitor, 25 μF; resistor, 200 Ω). After incubation in 1 ml of SOC medium (37) at 30°C for 1 h, the transformants were plated onto LB agar plate containing 50 μg/ml Ap at 30°C or 37°C overnight.
Pantothenate kinase genes have been cloned from Escherichia coli (42-44), Staphylococcus aureus (11, 27), Pseudomonas aeruginosa (18), Helicobacter pylori (9), Bacillus subtilis (9), B. anthracis (31), Thermotoga maritima (50, 51), Aspergillus nidulans (10), Arabidopsis thaliana (26, 46), mice (29, 33, 34, 54), and humans (28, 55, 56), and their enzyme activities have been verified in vitro. On the other hand, in archaea, pantoate kinase has been identified as the counterpart of CoaA (52). The homologues are widely distributed in archaea, but not in the order Thermoplasmatales. Based on their amino acid sequences, the pantothenate kinases have been classified into four types: prokaryotic types I, II, and III, and the eukaryotic type. To identify the key enzyme of the CoA biosynthetic pathway in the order Thermoplasmatales, a homology search using the NCBI BLAST program of the archaeal genome sequence data was queried with the amino acid sequences of CoaAEc (gi 16131808) as a prokaryotic type I CoaA, CoaASa (gi 21283783) as a prokaryotic type II, CoaAPa (gi 15599475) as a prokaryotic type III, and AnPanK (gi 4191500) and murine pantothenate kinase 1α (mPanK1α; gi 21744715) as eukaryotic enzymes. Consequently, three candidates for an archaeal pantothenate kinase were detected in the genomes of Ferroplasma acidarmanus fer1 (3), P. torridus DSM 9790 (14), and Thermoplasma volcanium GSS1 (24, 40) as a homologue of CoaAEc. These three proteins of F. acidarmanus Faci_03001718 (gi 126008038), P. torridus PTO0232 (gi 48477304), and T. volcanium TVN0137 (gi 13540968) consist of 320, 275, and 182 amino acids, and show 42%, 28%, and 41% identity to CoaAEc at the amino acid level, respectively (Fig. (Fig.1).1). Faci_03001718 and PTO0232 proteins conserve the amino acid residues constituting the catalytic site and the substrate binding sites for pantothenate and ATP, including the typical ATP-binding motif of GXXXXGKS/T, except that in Faci_03001718, the amino acid corresponding to Tyr240 of CoaAEc is replaced with Phe (19, 35, 53). On the other hand, TVN0137 is 134 amino acids shorter than CoaAEc and largely lacks the amino acids corresponding to the C-terminal region of CoaAEc which contains the pantothenate- and ATP-binding sites. The existence of a stop codon after Ile182 was verified by sequencing the DNA fragment between 242 bp upstream and 430 bp downstream of the TV0137 gene amplified by PCR, since sequence errors are sometimes observed in genome-sequenced data. Thus, in this analysis, it was suggested that at least the former two of the three candidates, Faci_03001718 and PTO0232, were the archaeal pantothenate kinases belonging to the class of prokaryotic type I enzymes.
Phylogenetic relationships between the two candidate archaeal CoaAs and other bacterial type I CoaAs were investigated. Phylogenetic trees were constructed using the deduced amino acid sequences of the biochemically characterized type I CoaA, CoaAEc, and of 72 putative type I CoaAs deposited in public databases (Fig. (Fig.2).2). The CoaA sequences were derived from a representative species showing higher similarities to CoaAEc based on BLAST search. The tree showed that the type I CoaA sequences were moderately grouped at the class or phylum level in both NJ and ML trees. In fact, the CoaA sequences from the Alphaproteobacteria, Deltaproteobacteria, and Chloroflexi formed monophyletic clades with over 50% bootstrap values in the both NJ and ML trees. In addition, the CoaA sequences derived from the members of Gammaproteobacteria and Actinobacteria formed monophyletic clades, except for three bacterial species Coxiella burnetii, Wigglesworthia glossinidia (Gammaproteobacteria), and Tropheryma whipplei (Actinobacteria). In contrast, the sequences from Firmicutes and Archaea were not clearly grouped in the NJ and ML trees. The candidate archaeal CoaAs from P. torridus and F. acidarmanus showed a deep branching position in the trees such as those from W. glossinidia and Tropheryma whipplei. Amino acid sequence identities of archaeal CoaAs from P. torridus and F. acidarmanus to any other bacterial type I CoaAs were <36.6% and <48.8%, respectively, indicating that these two archaeal CoaAs were distantly related to other type I CoaAs.
P. torridus is available from NITE Biological Resource Center under NBRC 100828, although F. acidarmanus fer1 is not available commercially. To confirm whether or not the above PTO0232 protein was pantothenate kinase, we cloned and expressed the PTO0232 gene using P. torridus NBRC 100828. The gene was amplified by PCR using P. torridus cells as a template DNA, and then the pUC-PTO0232 and pET-PTO0232 expression plasmids were constructed. First, the function of PTO0232 as the pantothenate kinase was confirmed by E. coli strain ts9, which carried a conditionally defective rts allele (13) and exhibited poor growth at 37°C. The pUC-PTO0232 that carried the gene under lac promoter control complemented the temperature-sensitive growth defect in strain ts9, and a similar number of colonies formed at 30°C, as were observed at 37°C on LB agar plate with ampicillin (Fig. (Fig.3).3). This data clearly indicated that the PTO0232 gene encoded the functional equivalent of an active pantothenate kinase. Subsequently, the PTO0232 protein was expressed in E. coli BL21(DE3)-RIL-GT cells (see Materials and Methods) carrying pET-PTO0232. For expression and recovery of the expressed protein in a soluble fraction, the extra copies of the tRNA genes on the pSC101-based plasmid and the chaperone genes on pG-Tf2 were needed. Consequently, a portion of the expressed protein was recovered in a soluble fraction, although most of the protein formed inclusion bodies (Fig. (Fig.4A).4A). The expressed protein in the soluble fraction phosphorylated pantothenate at 55°C, at which temperature the E. coli enzyme functions only slightly. Surprisingly, this protein was a globulin requiring a high concentration of salt—1.5 M NaCl or 500 mM (NH4)2SO4—to be solubilized in the buffer solution. Therefore, the recombinant protein was extracted with buffer containing 1.5 M NaCl and dialyzed against buffer without added salt in the final step of the purification to collect the protein as a precipitate. The resulting precipitate was dissolved in 25 mM Tris-HCl (pH 7.6) containing 500 mM (NH4)2SO4, 1 mM 2-mercaptoethanol, and 50% glycerol. The results from the overall purification procedure of the recombinant CoaAPt (PTO0232) are summarized in Table Table1.1. The specific activity of the purified enzyme was 360 nmol/min/mg of protein, with a 27.3-fold increase, and the overall recovery was 22.3% in enzyme activity. The purified enzyme was almost homogeneous, and its apparent molecular mass was estimated to be ~33.0 kDa by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; 12.5% polyacrylamide), which is in good agreement with the 34,221 Da predicted from the DNA sequence containing the His tag (Fig. (Fig.4B).4B). The known pantothenate kinases function as homodimers (17, 18, 27, 44, 50, 51, 53, 54). Gel filtration chromatography on Sephacryl S-200 was used to elucidate the subunit structure of the archaeal CoaA. However, despite many attempts, the molecular mass could not be successfully estimated, since precipitation of the protein occurred in the column.
CoaAPt was most active at pH 5, and the pH values between 4.5 and 6 supported more than 70% of its maximum activity, compared with CoaAEc, whose activity was maximal around pH 8 (Fig. (Fig.5A).5A). α-Glucosidase and α-mannosidase from the archaeon also show an optimum pH around 5 (5). Thus, Picrophilus enzymes appear to function well under acidic conditions compared with those from general bacteria, reflecting the cytoplasmic pH of Picrophilus cells, pH 4.6 (49). The kinase activity was optimal at 55°C and maintained more than 80% of the maximum activity at temperatures between 50°C and 70°C (Fig. (Fig.5B),5B), corresponding to the optimal growth of the organism at 60°C (49). In contrast, CoaAEc showed only 12.6% of the maximum activity at 50°C and 2.3% at 60°C. The pantothenate kinases from prokaryotic and eukaryotic cells generally require Mg2+ to form a complex with ATP. Although Mg2+ is the most effective in AnPanK (10), CoaAPt and CoaAEc seemed to prefer Mn2+ to Mg2+ (Table (Table2).2). In particular, the activity of CoaAPt was clearly accelerated by the addition of Mn2+ instead of Mg2+ by 1.75-fold. In comparison, the archaeal CoaA activity was supported by a wide range of divalent cations, even Fe2+ or Zn2+, which supported more than 80% of the Mg2+-dependent activity but less than 20% of the maximum activity of CoaAEc. The nucleotide specificity of CoaAPt was examined using the compounds listed in Table Table3.3. ATP was the best phosphate donor for both CoaAPt and CoaAEc. CoaAPt seems to recognize a number of NTPs (i.e., UTP, GTP, and CTP) in addition to ATP, as compared with the nucleotide specificity of CoaAEc. The fungal AnPanK and the prokaryotic type III CoaAPa also have broad substrate specificity for the nucleotide triphosphates (10, 18). Thus, it seems that the nucleotide recognition of CoaAPt, AnCoaA, and CoaAPa is loose, or rather that CoaAEc has stringent specificity.
The kinetic parameters of the purified CoaAPt for the two substrates were determined (Table (Table4).4). The apparent Km for pantothenate at saturating ATP concentrations was 621 ± 10 μM, and the Km for ATP at saturating pantothenate concentrations was 99 ± 16 μM. The Km value for ATP was similar to that of CoaAEc, but the value for pantothenate was about 20-fold higher than the Km of CoaAEc. The intracellular pantothenate concentration of E. coli possessing the prokaryotic type I CoaA was estimated to be roughly 20 to 90 μM (21, 22). Thus, the Km value of CoaAEc is reasonably consistent with the pantothenate level in the cells, but the Km of CoaAPt was much higher than expected. Most of the pantotenate kinases, except for the prokaryotic type III CoaAs, which require K+ or NH4+ for their activities, show the Km values for ATP to be about 100 μM. On the other hand, the Km values for pantothenate range from 5.5 to 621 μM.
The pantothenate kinases classified into prokaryotic type I and eukaryotic type enzymes are inhibited by CoA and CoA thioesters (10, 17, 28, 29, 33, 34, 43, 44, 48, 54, 55). The prokaryotic type I enzyme of CoaAEc is inhibited by CoA > acetyl-CoA > malonyl-CoA (10, 27, 48). Eukaryotic enzymes show different responses to CoA and acyl-CoAs: for AnPanK, acetyl-CoA > malonyl-CoA = CoA (10); mPanK1α and mPanK3, acetyl-CoA > malonyl-CoA > CoA (34, 54); mPanK1β, acetyl-CoA > malonyl-CoA and CoA activates mPanK1β (33, 34); and hPanK2, very sensitive to feedback inhibition by CoA species (55). On the other hand, the enzymes categorized into the prokaryotic types II and III CoaAs, such as CoaASa and CoaAPa, are refractory to feedback regulation by CoA species (9, 18, 27). Therefore, CoaAPt, which was classified as a prokaryotic type I enzyme based on its amino acid sequence, was assayed in vitro in the presence of different concentrations of CoA, acetyl-CoA, and malonyl-CoA, and compared with the activity of CoaAEc (Fig. (Fig.6).6). The CoaAEc activity was inhibited by the CoA species, and the addition of CoA, acetyl-CoA, or malonyl-CoA at 1,000 μM suppressed the activity by 83.6%, 64.7%, or 66.8%, respectively. However, the CoaAPt activity was not affected by these compounds, and more than 90% of the activity was maintained even in the presence of 1,000 μM of each of these compounds, indicating that CoA biosynthesis in P. torridus was not controlled at the pantothenate kinase step. The resistance to inhibition by CoA was clearly different from the sensitivity of the CoaAEc which belonged to the same class, and rather was characteristic of the prokaryotic type II and III CoaAs.
In this study, the PTO0232 gene from P. torridus was identified as encoding a pantothenate kinase. The gene product catalyzed the phosphate transfer reaction of ATP to pantothenate, with the best performance at pH 5 and 55°C, corresponding to the cytoplasmic pH and the optimal temperature for growth (49). Generally, the prokaryotic type I pantothenate kinase is governed by feedback inhibition by CoA and its thioesters (20, 30, 48). The archaeal CoaA reported here, however, was not regulated by these metabolites in spite of the sequence similarity to the prokaryotic type I enzyme (Fig. (Fig.6).6). The analysis of the structure of CoaAEc, which is representative of type I CoaAs, showed that the inhibitors bind to the enzyme and then cover Lys101 in the ATP-binding motif (53). Thus, regulation occurs because CoA and the ATP substrate overlap at the same site. Arg106, His177, and Phe247 have been identified as the amino acid residues that interact by a hydrophobic bond and a salt bridge only with CoA and not with ATP. Site-directed mutants, each with one substitution of the three amino acids, were refractory to inhibition by CoA (35). Thus, insensitivity of CoaAPt (PTO0232) to CoA inhibition can be explained by the substitution(s) of Arg106 and/or Phe247, as His177 binds not only to CoA but also to pantothenate and is conserved (Fig. (Fig.1).1). In CoaAPt, the amino acids corresponding to positions 106 and 247 of CoaAEc are replaced with Lys and Leu, respectively (Fig. (Fig.1),1), and consequently, the archaeal CoaA is not sensitive to CoA, acetyl-CoA, or malonyl-CoA (Fig. (Fig.6).6). Focusing on the amino acids at positions 106 and 247 which affect the feedback regulation, position 106 is occupied by Arg in most prokaryotic type I CoaAs, but based on an amino acid of position 247, the type I enzymes can be divided into two subtypes. One is the pantothenate kinase highly regulated by CoA and acyl-CoAs such as CoaAEc, and another subtype consists of enzymes refractory to end products such as CoaAPt. Phe247 forms a hydrophobic pocket together with His177 capable of sandwiching the adenine moiety of CoA between them (35, 53). The former subtype has Phe at position 247, and the latter has Leu, which is less hydrophobic and bulky than Phe, resulting in the insensitivity of CoaA to CoA and its thioesters.
Interestingly, the CoaAs possessing Phe at position 247 are mostly limited to Gammaproteobacteria, including most of the members within the family Enterobacteriaceae, although only one species in the Firmicutes also possessed Phe at this position (Fig. (Fig.2).2). The type I enzymes with Leu are widely distributed in a part of the Alphaproteobacteria, Gammaproteobacteria, Deltaproteobacteria, and Chloroflexi in gram-negative bacteria and Actinobacteria and Firmicutes (lactic acid bacteria) in gram-positive bacteria in the domain Bacteria, as well as Picrophilus torridus and Ferroplasma acidarmanus in domain Archaea. To date, the prokaryotic type I pantothenate kinases have been considered to be key enzymes in the CoA biosynthetic pathway and to be regulated by the end products (20, 22, 23, 30, 47, 48). However, this study indicates the existence of an uncontrolled type I enzyme and its distribution in the domains Bacteria and Archaea. Hence, it seems that organisms possessing the CoA biosynthetic pathway stringently controlled in the step of pantothenate kinase are limited to eukaryotes with the eukaryotic PanK and Gammaproteobacteria and only one species in Firmicutes possessing the subtype of the prokaryotic type I with Phe at position 247, since in addition to prokaryotic type II and III CoaAs, another subtype of the prokaryotic type I CoaA with Leu at position 247 is also refractory to feedback regulation by CoA (9, 18, 27). The following 10 of 72 strains used for the phylogenetic analysis in Fig. Fig.22 hold a type III CoaA in addition to a type I CoaA that is refractory to CoA: M. tuberculosis (type III CoaA; gi 148663464), N. farcinica (gi 54022372), R. jostii (gi 111021389), S. erythraea (gi 134097021), S. coelicolor (gi 21221809), T. fusca (gi 72163281), B. subtilis (gi 255767018), L. monocytogenes (gi 16802267), M. xanthus (gi 108763851), and S. aurantiaca (gi 115369916). Thus, although some organisms possess two CoaA genes in their genomes (i.e., the set of prokaryotic type I CoaAs with Leu at position 247 and type III CoaAs or a set of type II and type III CoaAs), neither set is subject to regulation by the end products of the CoA biosynthetic pathway.
So far, to our own best knowledge, there is no detailed information about the phylogenetic relationships of the prokaryotic type I CoaAs. In this study, the phylogenetic analysis of the type I CoaAs was performed on the basis of the amino acid sequences using the NJ and ML methods (Fig. 2A and B). The CoaA protein phylogenetic trees showed that the archaeal CoaA (CoaAPt) was significantly distinct from any other CoaAs. In fact, CoaAPt showed quite low sequence similarities (<36.6%) compared to any other CoaA sequences, indicating the uniqueness of this enzyme. Our phylogenetic analysis also showed that the two archaeal CoaAs from P. torridus and F. acidarmanus and the two bacterial CoaAs from Wigglesworthia glossinidia and Tropheryma whipplei were positioned at substantial branch points in the phylogenetic trees. W. glossinidia is an obligate primary endosymbiont of an insect (tsetse flies), and T. whipplei is the causative agent of human disease (Whipple's disease) and is often found in human intracellular niches such as macrophages and monocytes. The genome sequences of W. glossinidia and T. whipplei reflect their traits as strictly host-adapted microorganisms, e.g., having small genome sizes (0.7 Mb and 0.9 Mb, respectively) and low GC contents (22% and 46%, respectively) (2, 7). These data suggest that archaeal CoaAs that branch substantially within these symbiotic bacteria have likely evolved through unique ecophysiological adaptations.
Very recently, in the hyperthermophilic archaeon Thermococcus kodakarensis, two novel enzymes have been identified as pantoate kinase and phosphopantothenate synthetase, which are alternatives to CoaA and PanC involved in the canonical CoA biosynthetic pathway (52). As homologues of these enzymes are widely distributed in archaea, except for the order Thermoplasmales, the archaea seem to produce 4′-phosphopantothenate through the condensation of phosphopantoate and β-alanine by phosphopantothenate synthetase after the phosphorylation of pantoate by pantoate kinase. On the other hand, at least two species in the order Thermoplasmatales which are capable of growing even at extremely acidic conditions around pH 1 (12, 39, 49), P. torridus and F. acidarmanus, employ the pantothenate kinases belonging to the prokaryotic type I class for their CoA biosynthetic pathways as reported here. Thus, the key enzymes related to the phosphorylation of pantothenate or pantoate have been assigned to the domain Archaea. However, the PanC genes are still missing in the order Thermoplasmatales. Since archaeal CoA biosynthetic pathways seem not to be regulated by these phosphorylation steps, further biochemical characterization of each enzyme composing the pathway, in addition to a search for missing gene(s), is important for the clarification of the regulatory mechanisms that control the cellular CoA content of these organisms.
This work was supported by a Grant-in-Aid for Scientific Research (C) 19580079 from The Japan Society for the Promotion of Science.
We thank Koji Mori, NITE Biological Resource Center, Chiba, Japan, for sending us archaeal strains. We also thank Y. Sako, Graduate School of Agriculture, Kyoto University, Kyoto, Japan, and Y. Kamagata, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan, for helpful discussion.
Published ahead of print on 23 October 2009.