- Email: [email protected]
The macrophage-induced gene (mig) of Mycobacterium avium encodes a medium-chain acyl-coenzyme A synthetase
Biochimica et Biophysica Acta 1521 (2001) 59^65
www.bba-direct.com
The macrophage-induced gene (mig) of Mycobacterium avium encodes a medium-chain acyl-coenzyme A synthetase Christian Morsczeck, Sven Berger, Georg Plum * Institut fu«r Medizinische Mikrobiologie, Immunologie und Hygiene, Universita«t zu Ko«ln, GoldenfelsstraMe 19^21, 50935 Cologne, Germany Received 29 March 2001; received in revised form 26 July 2001 ; accepted 6 August 2001
Abstract The macrophage-induced gene (mig) of Mycobacterium avium has been associated with virulence, but the functions of the gene product were still unknown. Here we have characterized the Mig protein by biochemical methods. A plasmid with a histidine-tagged fusion protein was constructed for expression in Escherichia coli. Mig was detected as a 60 kDa protein after expression and purification of the recombinant gene product. The sequence of the fusion gene and of the parent gene in M. avium were reexamined. This confirmed that the mig gene encodes a 550 amino acid protein (58 kDa) instead of a 295 amino acid protein (30 kDa) as predicted before. The 550 amino acid Mig exhibits a high degree of homology to bacterial acyl-CoA synthetases. Two artificial 30 kDa derivatives of Mig were expressed and purified as histidine-tagged fusion proteins in E. coli. These proteins and the 58.6 kDa histidine-tagged Mig protein were analysed for activity with an acyl-CoA synthetase assay. Among the three investigated proteins, only the 58.6 kDa Mig exhibited detectable activity as an acyl-CoA synthetase (EC 6.2.1.3) with saturated medium-chain fatty acids, unsaturated long-chain fatty acid and some aromatic carbon acids as substrates. Enzymatic activity could be inhibited by 2-hydroxydodecanoic acid, a typical inhibitor of medium-chain acyl-CoA synthetases. We postulate a novel medium-chain acyl-CoA synthetase motif. We have investigated the biochemical properties of Mig and suggest that this enzyme is involved in the metabolism of fatty acid during mycobacterial survival in macrophages. ß 2001 Elsevier Science B.V. All rights reserved. Keywords : Acyl-coenzyme A synthetase; Mycobacterium ; Macrophage; Fatty acid ; Virulence gene
1. Introduction Mycobacterium avium is a slow growing, facultatively pathogenic mycobacterium that is ubiquitous. In countries with low incidence of tuberculosis, this species is common as an opportunistic pathogen in patients with AIDS where it contributes considerably to mortality. Pathogenic mycobacteria like M. avium are intracellular pathogens that are able to survive and multiply within macrophages. One gene, designated mig for macrophageinduced gene, was identi¢ed and shown to be transcribed when M. avium bacilli were growing within macrophages [1]. In DNA hybridization experiments this gene was shown to be speci¢c for M. avium within the Mycobacterium avium intracellulare Complex (MAC) [2]. Moreover, it was shown that a recombinant Mycobacterium smegma-
* Corresponding author. Fax: +49-221-478-3094. E-mail address : [email protected] (G. Plum).
tis strain expressing the Mig protein had an increased resistance against intracellular killing by primary macrophages in an infection model [3]. To fully elucidate the role of such genes in the pathogenicity it is necessary to characterize their function. This would preferably be done by gene targeting with an inactivated mig gene or by puri¢cation and biochemical characterization of the gene product. Gene inactivation has been successfully employed in other slow growing mycobacterial species, e.g. Mycobacterium tuberculosis, Mycobacterium bovis and Mycobacterium intracellulare [4^6]. We have not been successful in applying these same techniques to M. avium for inactivating the gene mig (unpublished data). Characterization of enzymatic activities were a last resort in our attempt to shed any light on the biological function of the gene. M. smegmatis mc2 155 transformed with shuttle plasmid pGPC200 harbouring the gene expressed Mig protein. On a Western blot, the cell wall fraction of M. smegmatis mc2 155/pGPC200 showed a single band of 60 kDa that reacted strongly with the a¤nity-
0167-4781 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 1 ) 0 0 2 8 7 - 1
BBAEXP 93565 17-10-01
60
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
puri¢ed AA22-36 antibody [3]. This result stands in con£ict with the previously published DNA sequence of mig U43598 that was deduced as ORF1 encoding a 30 kDa protein. This DNA sequence contained an additional putative ORF3 downstream from ORF1 separated by a frameshift of one nucleotide [3]. Here we report the expression and puri¢cation of a 60 kDa His-tagged protein that represents the fusion of ORF1 and ORF3. Moreover we characterize this product of mig as a medium-chain fatty acid acyl-CoA synthetase and discuss its role for the pathogenicity for mycobacteria and especially for M. avium. Acyl-CoA synthetases may be involved in intracellular recruitment of energy and carbon sources by degradation of fatty acids during replication in macrophages. Fatty acids are the only abundant C2 carbon sources for pathogens in mammalian tissues [7,8]. The importance of fatty acid degradation for intracellular persistence has recently been shown for the enzyme isocitrate lyase, an enzyme of the glyoxalate shunt pathway, which is strongly induced in intracellular pathogens after phagocytosis in the phagolysosome [7,8]. 2. Materials and methods 2.1. Bacterial strains and growth conditions M. avium strains were grown in Middlebrook 7H9 medium or 7H11 agar supplemented with ADC and 0.02% Tween 80. Escherichia coli strains used for plasmid selection, cloning and conjugation were grown in Luria^Bertani (LB) medium with 25 Wg ml31 kanamycin and/or 100 Wg ml31 ampicillin concentrations. 2.2. Oligonucleotides, polymerase chain reaction (PCR) and cycle sequencing Two oligonucleotides, MIG1005 5P-GTCCGCGCAGCCACCACC-3P and MIG1549 5P-CTCCGACGACCCGACGCC-3P, were designed based on a M. avium DNA sequence (GenBank accession No. U43598). PCR was carried out using genomic M. avium #5-8 or A5 DNA as template. The ampli¢cation conditions were as follows : initial denaturation for 5 min at 95³C, followed by 38 cycles of 1 min at 95³C, 45 s at 61³C and 1 min at 72³C and ¢nally 10 min incubation at 72³C. The nucleotide sequence of the PCR fragment was determined using the cycle-sequencing kit of Perkin-Elmer and an ABI-Prism Model 310 sequencer. 2.3. Construction of Mig expression system Initially the DNA fragment containing the ORF of mig was cloned into the pQE10 (Qiagen, Hilden, Germany) bacterial expression plasmid. Therefore a sense primer MIGHISFW 5P-AAGGATCCCACCACAACAGCATT-
CAC-3P containing a BamHI site (underlined) and an antisense primer MIGHISRW 5P-AACTGCAGGGGACCGCACGAC-3P containing a PstI site (underlined) were designed to amplify ORF1 and ORF3 (Fig. 1) of the published mig sequence [3]. The ampli¢cation conditions were as follows : initial denaturation for 5 min at 94³C, followed by 35 cycles of 1 min at 94³C, 45 s at 60³C and 1 min at 72³C and ¢nally 10 min incubation at 72³C. For expression vector construction a BamHI-PstI fragment of the ampli¢ed DNA was directionally ligated into BamHIPstI digested pQE10 (Qiagen) and transformed into the E. coli XL-1-blue. The nucleotide sequence of pHis5 was compared to the published sequence U43598. The expression vector pHis9 was constructed by deleting an AscIPstI fragment, coding for the carboxy-terminal part of Mig. 2.4. Expression of His-tagged Mig in E. coli E. coli M15 was transformed with pHis5, pHis9 or pGPC150 for His-tagged Mig protein expression. A 60 ml culture was grown overnight at 37³C in LB medium containing kanamycin and ampicillin and used to inoculate 1000 ml of LB medium containing ampicillin and kanamycin. At an OD600nm of 0.6, isopropyl L-D-thiogalactopyranoside (IPTG) was added to a ¢nal concentration of 1 mM, and growth was continued for another 4.5 h at 37³C. Cells were harvested by centrifugation at 1000Ug for 15 min at 4³C and resuspended in lysis bu¡er (50 mM NaH2 PO4 (pH 8.0), 300 mM NaCl, 10 mM imidazole). The cells were disrupted by glass beads with a cell mill (5 min, high speed). Thereafter insoluble matter was removed by centrifugation at 20 000Ug for 30 min at 4³C. The cleared lysate was added to 1 ml of a suspension of Ni-agarose (Ni-NTA agarose, Qiagen), equilibrated with lysis bu¡er, and mixed gently for 1 h at 4³C by end-overend rotation. The mixture was poured into a column and washed twice with wash bu¡er (50 mM NaH2 PO4 (pH 8.0), 300 mM NaCl, 20 mM imidazole). Finally the elution was done batchwise with elution bu¡er (50 mM NaH2 PO4 (pH 8.0), 300 mM NaCl, 250 mM imidazole). All fractions were analysed by 10% SDS^PAGE and stained with Coomassie. 2.5. Western blot analysis Cultures of E. coli and M. avium in exponential growth phase (OD600 : 0.2^0.6) or stationary growth phase were harvested by centrifugation. After cell disruption by agitation with glass beads in a cell mill (5 min, high speed) cleared lysates and insoluble fractions were prepared as stated above. Samples were separated by sodium dodecyl sulphate^polyacrylamide gel electrophoresis (SDS^PAGE, 10% polyacrylamide) under denaturing conditions and then transferred by blotting to PVDF membrane. Immunodetection was performed with a Mig-speci¢c antibody
BBAEXP 93565 17-10-01
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
61
Fig. 1. Old (U43598) and corrected DNA sequences of the mig gene of M. avium with positions of primers and open reading frames. (A) DNA sequences as obtained after direct sequencing of PCR products, using primers MIG1005 and MIG1549 and M. avium A5 or M. avium #5-8 as template, in comparison to homologous sequences from pHis5 and U43598. The deduced amino acid sequences of ORF1 (U43598) and ORFmig (pHis5) in this region are indicated below. The deletion mutation at codon triplet for amino acid 275 of ORF1 is marked grey and the resulting frameshift in the ORFmig amino acid sequence is underlined. (B) Positions of primers MIGHISFW, MIGHISRW, MIG1005 and MIG1549 and of open reading frames (ORF) are indicated.
and the ECL Western blotting detection method (Amersham Pharmacia Biotech). 2.6. Analytical procedures Acyl-CoA synthase activity was measured as described before [9,10]. The following assay conditions were used unless otherwise speci¢ed: 100 mM Tris^HCl (pH 7.8), 5 mM MgCl2 , 1 mM ATP, 0.4 mM CoA, 0.4 mM NADH, 1 mM PEP, 2 mM DTE, 1 U/ml myokinase (EC 2.7.4.3), 2.8 U/ml D-lactate dehydrogenase (EC 1.1.1.27), 1.5 U/ml pyruvate kinase (EC 2.7.1.40), 1^20 Wg/ml protein (Mig) and various concentrations (Table 2) of fatty acids or aromatic acids in a total volume of 1 ml. Incubation was carried out for various times at room temperature. Kinetic studies were done with one of the substrates varied as indicated and with the other substrate at the concentration
used in the standard assay. Inhibition studies were performed in a standard assay using octanoic acid or dodecanoic acid as substrate after preincubation with 2-hydroxydodecanoic acid. Protein was determined according to Bradford (Bio-Rad) with bovine serum albumin as a standard. SDS^PAGE was done according to Laemmli [11]. 3. Results 3.1. ORFmig encodes a 550 amino acid protein The expression vector pHis5 was constructed as described in Section 2 and the DNA sequence was determined three times for each nucleotide. From the resulting nucleotide sequence an open reading frame (ORFmig) was deduced that encoded a 550 amino acid Mig protein (58
Table 1 Feature
Prediction based on
Amino acid position
Orientation
Leader TM AMP TM
SignalP TMpred Prosite TMpred TopPred2 TMpred TopPred2 TopPred2 alignment alignment
1^18 84^103 186^197 245^267 251^271 309^330 311^331 354^370 422^446 430^464
na o-i na i-o na o-i na na na na
TM TM FAD Domain M
Feature, Mig peptide sequence features; TM, transmembrane region; AMP, AMP-binding domain signature; o-i, outside-inside ; i-o, inside-outside; na, not applicable ; FAD, fatty acyl-coenzyme A synthetase signature motif; domain M, medium-chain fatty acyl speci¢city domain (see Section 3).
BBAEXP 93565 17-10-01
62
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
kDa). The only di¡erence between the pHis5 DNA sequence and U43598 [3] is a single deletion of a cytosine residue at the codon triplet for amino acid 275 of ORF1. This results in a frameshift leading to a fusion between ORF1 and ORF3 (Fig. 1). Direct sequencing of PCR products created with primers MIG1005 and MIG1549 veri¢ed this fusion of ORF1 and ORF3 in the genome of M. avium. Fig. 2. SDS^PAGE and Western blot with puri¢ed His-tagged Mig. (A) Expression of His-tagged Mig protein in E. coli. Lanes: 1, molecular weight markers ; 2, lysate of E. coli M15 transformed with pHis5 after induction (80 Wg protein) ; 3, wash fraction; 4^7, elution fractions 1^4 from Ni-agarose column (20^80 Wg protein). (B) Western blot of Mig protein expressed in E. coli. Western blot with a¤nity-puri¢ed AA22-36 antibody [3] is shown. Lanes: W, wash fraction; F2, elution fraction 2 from Ni-agarose column, see above (1 Wg protein).
3.2. Heterologous expression of Mig proteins in E. coli The overexpressions of Mig were carried out using pHis5 and pHis9 and with pGPC150 [3] bacterial expression systems. Ni-agarose a¤nity chromatography resulted in the appearance of a major band on SDS^PAGE with an apparent mobility of approx. 60 kDa (pHis5, Fig. 2A), 31 kDa (pHis9, data not shown) and 28 kDa (pGPC150, data
Fig. 3. Protein domains of fatty acid acyl-CoA synthetases. The predicted primary amino acid sequence of mig gene product was compared with the AMP-binding domain [17] and the fatty acyl-coenzyme A synthetase signature motif [21] of adenylate-building proteins. Amino acid positions responsible for decanoic acid speci¢city (I and II) and for oleic acid speci¢city (III and IV) are on a grey shaded background [21]. Acyl-CoA synthetases: from M. bovis (AcoAS [24]); from P. oleovorans (AlkK [19]); from Archaeoglobus fulgidus (AlkK-5 [32]; putative); from Saccharomyces cerevisiae (FAA1 [33]); from S. cerevisiae (FAA2 [34]); from S. cerevisiae (FAA3 [35]); from E. coli (FadD [16]); putative from M. tuberculosis (FadD19 [20]). M-domain of medium-chain acyl-CoA synthetase: Mig; AlkK [19]; AlkK-5 [32]; Dein (Deinococcus spec. GenBank B75265, putative); and long-chain acyl-CoA synthetases : FadD [16]; FadD19 [20]; ACoAS [24].
BBAEXP 93565 17-10-01
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
63
Table 2 Mig medium-chain fatty acid acyl-CoA synthetase activity using indicated carbon acids as substrates Substrate
Activity (nmol min31 mg31 )
Concentration (WM)
Ethanoic acid (C2:0 ) Butanoic acid (C4:0 ) Pentanoic acid (C5:0 ) Hexanoic acid (C6:0 ) Heptanoic acid (C7:0 ) Octanoic acid (C8:0 ) Decanoic acid (C10:0 ) Dodecanoic acid (C12:0 ) Tridecanoic acid (C13:0 ) Tetradecanoic acid (C14:0 ) Palmitic acid (C16:0 ) Arachidonic acid (C20:4 ) Linolenic acid (C18:3 ) Oleic acid (C18:1 ) Benzoic acid 2-Aminobenzoic acid m-Hydroxybenzoic acid Phenyl acetate
5000 5000 5000 5000 5000 5000 5000 100 100 100 200 10 30 30 2000 2000 2000 2000
S.D. (nmol min31 mg31 )
0.0 0.0 8.0 175.0 170.0 155.0 497.0 153.0 73.0 19.0 0.0 25.7 41.2 111.8 11.6 7.2 3.9 0.0
not shown). The approx. 60 kDa single band (pHis5, Fig. 2B, lane F2) was recognized in Western blot analysis with the a¤nity-puri¢ed AA22-36 anti-Mig antibody [3]. M. avium cultures grown in 7H9 Middlebrook medium supplemented with ADC were analysed by Western blotting. The antibody [3] recognized a band at approx. 60 kDa in cell lysates that was also present in the insoluble fraction. These bands were of a size comparable with the His-Mig fusion protein expressed in E. coli/pHis5. 3.3. Analysis of deduced 58 kDa Mig peptide sequence The Mig amino acid sequence (encoded by ORFmig) was examined. The hydrophilicity pro¢le of the deduced sequence [12] identi¢es several hydrophobic regions. The computer program `Psort' [13] predicted Mig to be a membrane protein. Most notably, the existence of putative transmembrane-associated K-helices (Table 1) was strongly predicted by the programs `TmPred' and `TopPred2' [14,15]. High similarities (data not shown) with procaryotic fatty acid acyl-CoA synthetase were detected for the 58 kDa Mig protein after investigations with NCBI advanced and basic blast programs. This applies especially to the FadD from E. coli [16,17], the long-chain acyl-CoA synthetase from Bacillus subtilis [18], the medium-chain acyl-CoA synthetase from Pseudomonas oleovorans [19]
0.0 0.0 2.5 4.5 3.0 2.5 8.5 28.0 6.0 0.5 0.0 4.9 11.8 11.8 0.1 0.1 0.1 0.0
and a putative fatty acid acyl-CoA synthetase FadD19 from M. tuberculosis [20]. Other signi¢cant characteristics of the 58 kDa Mig protein are a putative leader peptide and an AMP-binding domain signature as described before [3] and a homology with a fatty acyl-coenzyme A synthetase motif [21] of adenylate-forming proteins (Fig. 3). Moreover, we detected a previously undescribed region of high similarity in medium-chain CoA synthetases which is localized C-terminally of the fatty acyl-coenzyme A synthetase motif (Fig. 3, domain M). 3.4. The properties of Mig as a medium-chain fatty acid acyl-CoA synthetase We tested the three derivatives of Mig for acyl-CoA synthetase activity. Only the 58 kDa Mig exhibited an activity using saturated and unsaturated fatty acids (Table 2). This activity depended on the concentration of Mig, of CoA, ATP and fatty acids. Interestingly Mig exhibited acyl-CoA synthetase activities with saturated fatty acids of medium-chain length between C5:0 and C14:0 and with unsaturated fatty acids like oleic acid (C18:1 ), linolenic acid (C18:3 ) and arachidonic acid (C20:4 ). In addition, we detected low activity using benzoic acid as substrate (Table 2). In contrast to the 58 kDa protein no acyl-CoA synthetase activities were detected with any carbon acid as sub-
Table 3 Biochemical characterization of Mig as medium-chain fatty acid acyl-CoA synthetase Value
ATP/CoA
Km (WM) Vmax (nmol min31 mg31 ) Turnover No. (min31 )
36/45 ^ ^
Octanoic acid (C8:0 ) 285 183 11
BBAEXP 93565 17-10-01
Decanoic acid (C10:0 ) 1775 510 30
Dodecanoic acid (C12:0 ) 20 163 10
64
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
strate for both puri¢ed 30 kDa His-Mig derivatives. In order to characterize the kinetic properties of the acyl-CoA synthetase the Km and Vmax values for octanoic acid, decanoic acid and dodecanoic acid and the Km values for ATP and CoA for the recombinant protein were determined (Table 3). It was possible to inhibit the acyl-CoA synthetase activity by preincubation of Mig with 100 WM 2-hydroxydodecanoic acid. The acyl-CoA synthetase activity decreased to 64% in comparison to the untreated Mig protein when 100 WM octanoic acid or 100 WM dodecanoic acid was used as substrate. 4. Discussion Our results demonstrate that the ORF of mig encodes a 550 amino acid peptide. The corrected sequence was veri¢ed by direct sequencing of PCR products ampli¢ed from two di¡erent M. avium strains and by Western blot analysis. This clari¢ed the con£icting results that had previously been achieved after expression of the gene in M. smegmatis, E. coli and M. avium. We demonstrated that Mig is a medium-chain acyl-CoA synthetase. Our results are a ¢rst step towards the elucidation of the biological function of the gene. Together with our previous ¢ndings [1,3] they underscore the importance of enzymes for lipid metabolism in virulent mycobacteria as discussed before [7,22,23]. We characterized the biochemical properties of a puri¢ed medium-chain acyl-CoA synthetase in Gram-positive bacteria for the ¢rst time. The amino acid sequence of Mig shows similarities to acyl-CoA synthetases of E. coli, B. subtilis, P. oleovorans and with peptide synthetases of various bacterial species (data not shown). Interestingly, Mig exhibits only moderate degrees of homology to the M. bovis long-chain acyl-coenzyme A synthase which was characterized before [24,25]. These polypeptides belong to the superfamily of AMP-binding proteins with a characteristic signature motif [17]. In addition to this motif Mig shares homologies with an additional domain of 25 amino acid residues (Fig. 3) that is typical for fatty acylcoenzyme A synthetases. Alterations of amino acid residues within this region have been shown to modulate fatty acid substrate speci¢city [21]. A homologous domain of this fatty acyl-coenzyme A synthetase signature motif is located at the C-terminal part of Mig. The two approx. 30 kDa His-tagged Mig derivatives do not carry this domain and were found to have no acyl-CoA synthetase activity. The 30 kDa proteins that have previously been described [3] are likely functionally inactive degradation products of the protein. Despite the obvious similarities of the gene to the E. coli FadD homologue there are also some remarkable di¡erences from this enzyme. In fact the putative fatty acyl-coenzyme A synthetase motif of Mig has only ¢ve positions of 25 amino acid residues in common with the corresponding domain of FadD. More-
over, three of four important positions for amino acid residues responsible for decanoic acid and oleic acid speci¢city are not identical for the two acyl-CoA synthetases [21], but Mig was still active with these fatty acids as substrate. Here we postulate a novel motif, called domain M, that could be involved in de¢ning the CoA synthetase speci¢city for medium-chain fatty acids (Fig. 3). We found strong similarities between the domain M of Mig and the corresponding domains of medium-chain acyl-CoA synthetases, but a less conserved pattern with long-chain acyl-CoA synthetases (Fig. 3). The substrate speci¢city of Mig is similar to mediumchain acyl-CoA synthetases of Pseudomonas putida and rat kidney [26,27]. Mig was able to ligate CoA to saturated medium-chain fatty acids (C5:0 -C14:0 ) and benzoic acid derivatives as substrates. In the mammalian kidney, activation of benzoic acid derivatives is used for glycine conjugation. This constitutes one of the most important routes of detoxi¢cation, not only for xenobiotic carboxylic acids [27]. In contrast to the characterized M. bovis acyl-CoA synthase or FadD from E. coli, Mig could not ligate coenzyme A to palmitic acid (C16:0 ) or longer saturated fatty acids [24,28]. The acyl-CoA synthetase activities of Mig as a function of the tested fatty acid concentrations followed the Michaelis^Menten kinetics, whereas the rat intestinal medium-chain acyl-CoA synthetase activity as a function of octanoate concentration has been shown to follow a biphasic reaction curve [29]. Mig showed the highest activity (Vmax : 510 nmol min31 mg31 ) with decanoic acid but had the lowest a¤nity to this molecule (Km : 1775 WM) in comparison to octanoic acid (Km : 183 WM) and dodecanoic acid (Km : 20 WM) (Table 3). The Mig protein has four predicted transmembrane regions and is localized in the cytoplasm membrane with the `Psort' program. This prediction is in accordance with the localization of the acyl-CoA synthase (M. bovis) and FadD (E. coli) [16,24]. We conclude that Mig is involved in activation of free fatty acids that will then either be degraded by L-oxidation or serve as intermediates in the synthesis of very long fatty acids, of lipoproteins or glycolipids. In mycobacteria the activation of fatty acids for L-oxidation can be linked to the biosynthesis of glucose via the glyoxalate shunt pathway, which is important for the survival inside the phagosome of macrophages [7]. On the other hand are free fatty acids together with reactive nitrogen intermediates and reactive oxygen intermediates e¡ector molecules in the expression of antimicrobial activity of the macrophage against intracellular mycobacteria [30,31]. Free fatty acids could be detoxi¢ed by CoA ligation and subsequent Loxidation. Furthermore, the arachidonic acid modulating activity of Mig may also interfere with the leukotriene synthesis by the host, where the fatty acid is a precursor of the e¡ectors in the antimicrobial response. Taken together, these ¢ndings argue for a role of this gene in the virulence of this organism.
BBAEXP 93565 17-10-01
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
References [1] G. Plum, J.E. Clark Curtiss, Infect. Immun. 62 (1994) 476^483. [2] M.L. Beggs, R. Stevanova, K.D. Eisenach, J. Clin. Microbiol. 38 (2000) 508^512. [3] G. Plum, M. Brenden, J.E. Clark-Curtiss, G. Pulverer, Infect. Immun. 65 (1997) 4548^4557. [4] V. Balasubramanian, M.S. Pavelka, S.S. Bardarov, J. Martin, T.R. Weisbrod, R.A. McAdam, B.R. Bloom, W.R. Jacobs, J. Bacteriol. 178 (1996) 273^279. [5] A.K. Azad, T.D. Sirakova, L.M. Rogers, P.E. Kolattukudy, Proc. Natl. Acad. Sci. USA 93 (1996) 4787^4792. [6] E. Mahenthiralingam, B.I. Marklund, L.A. Brooks, D.A. Smith, G.J. Bancroft, R.W. Stokes, Infect. Immun. 66 (1998) 3626^3634. [7] J.D. McKinney, K. Honer zu Bentrup, E.J. Munoz-Elias, A. Miczak, B. Chen, W.T. Chan, D. Swenson, J.C. Sacchettini, W.R. Jacobs Jr., D.G. Russell, Nature 406 (2000) 735^738. [8] M.C. Lorenz, G.R. Fink, Nature 412 (2001) 83^86. [9] T. Tanaka, K. Hosaka, S. Numa, Methods Enzymol. 71 (1981) 334^ 341. [10] K. Ziegler, R. Buder, J. Winter, G. Fuchs, Arch. Microbiol. 151 (1989) 171^176. [11] U.K. Laemmli, Nature 227 (1970) 680^685. [12] J. Kyte, R.F. Doolittle, J. Mol. Biol. 157 (1982) 105^132. [13] K. Nakai, M. Kanehisa, Genomics 14 (1992) 897^911. [14] G. von Heijne, J. Mol. Biol. 225 (1992) 487^494. [15] K. Hofmann, W. Sto¡el, Biol. Chem. Hoppe-Seyler 374 (1993) 166^ 178. [16] P.N. Black, C.C. DiRusso, A.K. Metzger, T.L. Heimert, J. Biol. Chem. 267 (1992) 25513^25520. [17] M. Fulda, E. Heinz, F.P. Wolter, Mol. Gen. Genet. 242 (1994) 241^ 249. [18] A. Wipat, N. Carter, S.C. Brignell, B.J. Guy, K. Piper, J. Sanders, P.T. Emmerson, C.R. Harwood, Microbiology 142 (1996) 3067^3078. [19] J.B. van Beilen, G. Eggink, H. Enequist, R. Bos, B. Witholt, Mol. Microbiol. 6 (1992) 3121^3136.
65
[20] S.T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris, S.V. Gordon, K. Eiglmeier, S. Gas, C.E. Barry III, F. Tekaia, K. Badcock, D. Basham, D. Brown, T. Chillingworth, R. Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S. Holroyd, T. Hornsby, K. Jagels, B.G. Barrell et al., Nature 393 (1998) 537^544. [21] P.N. Black, Q. Zhang, J.D. Weimar, C.C. DiRusso, J. Biol. Chem. 272 (1997) 4896^4903. [22] W. Bishai, Nature 406 (2000) 683^685. [23] M.S. Glickman, W.R. Jacobs Jr., Cell 104 (2001) 477^485. [24] A.M. Fitzmaurice, P.E. Kolattukudy, J. Bacteriol. 179 (1997) 2608^ 2615. [25] A.M. Fitzmaurice, P.E. Kolattukudy, J. Biol. Chem. 273 (1998) 8033^8039. [26] M. Fernandez-Valverde, A. Reglero, H. Martinez-Blanco, J.M. Luengo, Appl. Environ. Microbiol. 59 (1993) 1149^1154. [27] F. Kasuya, K. Igarashi, M. Fukui, Chem.-Biol. Interact. 118 (1999) 233^246. [28] K. Kameda, W.D. Nunn, J. Biol. Chem. 256 (1981) 5702^5707. [29] Y. Ohkubo, S. Mori, Y. Ishikawa, K. Shirai, Y. Saito, S. Yoshida, Digestion 51 (1992) 42^50. [30] T. Akaki, K. Sato, T. Shimizu, C. Sano, H. Kajitani, S. Dekio, H. Tomioka, J. Leukocyte Biol. 62 (1997) 795^804. [31] H. Tomioka, K. Sato, C. Sano, T. Akaki, T. Shimizu, H. Kajitani, H. Saito, Clin. Exp. Immunol. 109 (1997) 248^254. [32] H.P. Klenk, R.A. Clayton, J.F. Tomb, O. White, K.E. Nelson, K.A. Ketchum, R.J. Dodson, M. Gwinn, E.K. Hickey, J.D. Peterson, D.L. Richardson, A.R. Kerlavage, D.E. Graham, N.C. Kyrpides, R.D. Fleischmann, J. Quackenbush, N.H. Lee, G.G. Sutton, S. Gill, E.F. Kirkness, B.A. Dougherty, K. McKenney, M.D. Adams, B. Loftus, J.C. Venter et al., Nature 390 (1997) 364^370. [33] R.J. Duronio, L.J. Knoll, J.I. Gordon, J. Cell Biol. 117 (1992) 515^ 529. [34] D.R. Johnson, L.J. Knoll, N. Rowley, J.I. Gordon, J. Biol. Chem. 269 (1994) 18037^18046. [35] L.J. Knoll, D.R. Johnson, J.I. Gordon, J. Biol. Chem. 269 (1994) 16348^16356.
BBAEXP 93565 17-10-01
www.bba-direct.com
The macrophage-induced gene (mig) of Mycobacterium avium encodes a medium-chain acyl-coenzyme A synthetase Christian Morsczeck, Sven Berger, Georg Plum * Institut fu«r Medizinische Mikrobiologie, Immunologie und Hygiene, Universita«t zu Ko«ln, GoldenfelsstraMe 19^21, 50935 Cologne, Germany Received 29 March 2001; received in revised form 26 July 2001 ; accepted 6 August 2001
Abstract The macrophage-induced gene (mig) of Mycobacterium avium has been associated with virulence, but the functions of the gene product were still unknown. Here we have characterized the Mig protein by biochemical methods. A plasmid with a histidine-tagged fusion protein was constructed for expression in Escherichia coli. Mig was detected as a 60 kDa protein after expression and purification of the recombinant gene product. The sequence of the fusion gene and of the parent gene in M. avium were reexamined. This confirmed that the mig gene encodes a 550 amino acid protein (58 kDa) instead of a 295 amino acid protein (30 kDa) as predicted before. The 550 amino acid Mig exhibits a high degree of homology to bacterial acyl-CoA synthetases. Two artificial 30 kDa derivatives of Mig were expressed and purified as histidine-tagged fusion proteins in E. coli. These proteins and the 58.6 kDa histidine-tagged Mig protein were analysed for activity with an acyl-CoA synthetase assay. Among the three investigated proteins, only the 58.6 kDa Mig exhibited detectable activity as an acyl-CoA synthetase (EC 6.2.1.3) with saturated medium-chain fatty acids, unsaturated long-chain fatty acid and some aromatic carbon acids as substrates. Enzymatic activity could be inhibited by 2-hydroxydodecanoic acid, a typical inhibitor of medium-chain acyl-CoA synthetases. We postulate a novel medium-chain acyl-CoA synthetase motif. We have investigated the biochemical properties of Mig and suggest that this enzyme is involved in the metabolism of fatty acid during mycobacterial survival in macrophages. ß 2001 Elsevier Science B.V. All rights reserved. Keywords : Acyl-coenzyme A synthetase; Mycobacterium ; Macrophage; Fatty acid ; Virulence gene
1. Introduction Mycobacterium avium is a slow growing, facultatively pathogenic mycobacterium that is ubiquitous. In countries with low incidence of tuberculosis, this species is common as an opportunistic pathogen in patients with AIDS where it contributes considerably to mortality. Pathogenic mycobacteria like M. avium are intracellular pathogens that are able to survive and multiply within macrophages. One gene, designated mig for macrophageinduced gene, was identi¢ed and shown to be transcribed when M. avium bacilli were growing within macrophages [1]. In DNA hybridization experiments this gene was shown to be speci¢c for M. avium within the Mycobacterium avium intracellulare Complex (MAC) [2]. Moreover, it was shown that a recombinant Mycobacterium smegma-
* Corresponding author. Fax: +49-221-478-3094. E-mail address : [email protected] (G. Plum).
tis strain expressing the Mig protein had an increased resistance against intracellular killing by primary macrophages in an infection model [3]. To fully elucidate the role of such genes in the pathogenicity it is necessary to characterize their function. This would preferably be done by gene targeting with an inactivated mig gene or by puri¢cation and biochemical characterization of the gene product. Gene inactivation has been successfully employed in other slow growing mycobacterial species, e.g. Mycobacterium tuberculosis, Mycobacterium bovis and Mycobacterium intracellulare [4^6]. We have not been successful in applying these same techniques to M. avium for inactivating the gene mig (unpublished data). Characterization of enzymatic activities were a last resort in our attempt to shed any light on the biological function of the gene. M. smegmatis mc2 155 transformed with shuttle plasmid pGPC200 harbouring the gene expressed Mig protein. On a Western blot, the cell wall fraction of M. smegmatis mc2 155/pGPC200 showed a single band of 60 kDa that reacted strongly with the a¤nity-
0167-4781 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 1 ) 0 0 2 8 7 - 1
BBAEXP 93565 17-10-01
60
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
puri¢ed AA22-36 antibody [3]. This result stands in con£ict with the previously published DNA sequence of mig U43598 that was deduced as ORF1 encoding a 30 kDa protein. This DNA sequence contained an additional putative ORF3 downstream from ORF1 separated by a frameshift of one nucleotide [3]. Here we report the expression and puri¢cation of a 60 kDa His-tagged protein that represents the fusion of ORF1 and ORF3. Moreover we characterize this product of mig as a medium-chain fatty acid acyl-CoA synthetase and discuss its role for the pathogenicity for mycobacteria and especially for M. avium. Acyl-CoA synthetases may be involved in intracellular recruitment of energy and carbon sources by degradation of fatty acids during replication in macrophages. Fatty acids are the only abundant C2 carbon sources for pathogens in mammalian tissues [7,8]. The importance of fatty acid degradation for intracellular persistence has recently been shown for the enzyme isocitrate lyase, an enzyme of the glyoxalate shunt pathway, which is strongly induced in intracellular pathogens after phagocytosis in the phagolysosome [7,8]. 2. Materials and methods 2.1. Bacterial strains and growth conditions M. avium strains were grown in Middlebrook 7H9 medium or 7H11 agar supplemented with ADC and 0.02% Tween 80. Escherichia coli strains used for plasmid selection, cloning and conjugation were grown in Luria^Bertani (LB) medium with 25 Wg ml31 kanamycin and/or 100 Wg ml31 ampicillin concentrations. 2.2. Oligonucleotides, polymerase chain reaction (PCR) and cycle sequencing Two oligonucleotides, MIG1005 5P-GTCCGCGCAGCCACCACC-3P and MIG1549 5P-CTCCGACGACCCGACGCC-3P, were designed based on a M. avium DNA sequence (GenBank accession No. U43598). PCR was carried out using genomic M. avium #5-8 or A5 DNA as template. The ampli¢cation conditions were as follows : initial denaturation for 5 min at 95³C, followed by 38 cycles of 1 min at 95³C, 45 s at 61³C and 1 min at 72³C and ¢nally 10 min incubation at 72³C. The nucleotide sequence of the PCR fragment was determined using the cycle-sequencing kit of Perkin-Elmer and an ABI-Prism Model 310 sequencer. 2.3. Construction of Mig expression system Initially the DNA fragment containing the ORF of mig was cloned into the pQE10 (Qiagen, Hilden, Germany) bacterial expression plasmid. Therefore a sense primer MIGHISFW 5P-AAGGATCCCACCACAACAGCATT-
CAC-3P containing a BamHI site (underlined) and an antisense primer MIGHISRW 5P-AACTGCAGGGGACCGCACGAC-3P containing a PstI site (underlined) were designed to amplify ORF1 and ORF3 (Fig. 1) of the published mig sequence [3]. The ampli¢cation conditions were as follows : initial denaturation for 5 min at 94³C, followed by 35 cycles of 1 min at 94³C, 45 s at 60³C and 1 min at 72³C and ¢nally 10 min incubation at 72³C. For expression vector construction a BamHI-PstI fragment of the ampli¢ed DNA was directionally ligated into BamHIPstI digested pQE10 (Qiagen) and transformed into the E. coli XL-1-blue. The nucleotide sequence of pHis5 was compared to the published sequence U43598. The expression vector pHis9 was constructed by deleting an AscIPstI fragment, coding for the carboxy-terminal part of Mig. 2.4. Expression of His-tagged Mig in E. coli E. coli M15 was transformed with pHis5, pHis9 or pGPC150 for His-tagged Mig protein expression. A 60 ml culture was grown overnight at 37³C in LB medium containing kanamycin and ampicillin and used to inoculate 1000 ml of LB medium containing ampicillin and kanamycin. At an OD600nm of 0.6, isopropyl L-D-thiogalactopyranoside (IPTG) was added to a ¢nal concentration of 1 mM, and growth was continued for another 4.5 h at 37³C. Cells were harvested by centrifugation at 1000Ug for 15 min at 4³C and resuspended in lysis bu¡er (50 mM NaH2 PO4 (pH 8.0), 300 mM NaCl, 10 mM imidazole). The cells were disrupted by glass beads with a cell mill (5 min, high speed). Thereafter insoluble matter was removed by centrifugation at 20 000Ug for 30 min at 4³C. The cleared lysate was added to 1 ml of a suspension of Ni-agarose (Ni-NTA agarose, Qiagen), equilibrated with lysis bu¡er, and mixed gently for 1 h at 4³C by end-overend rotation. The mixture was poured into a column and washed twice with wash bu¡er (50 mM NaH2 PO4 (pH 8.0), 300 mM NaCl, 20 mM imidazole). Finally the elution was done batchwise with elution bu¡er (50 mM NaH2 PO4 (pH 8.0), 300 mM NaCl, 250 mM imidazole). All fractions were analysed by 10% SDS^PAGE and stained with Coomassie. 2.5. Western blot analysis Cultures of E. coli and M. avium in exponential growth phase (OD600 : 0.2^0.6) or stationary growth phase were harvested by centrifugation. After cell disruption by agitation with glass beads in a cell mill (5 min, high speed) cleared lysates and insoluble fractions were prepared as stated above. Samples were separated by sodium dodecyl sulphate^polyacrylamide gel electrophoresis (SDS^PAGE, 10% polyacrylamide) under denaturing conditions and then transferred by blotting to PVDF membrane. Immunodetection was performed with a Mig-speci¢c antibody
BBAEXP 93565 17-10-01
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
61
Fig. 1. Old (U43598) and corrected DNA sequences of the mig gene of M. avium with positions of primers and open reading frames. (A) DNA sequences as obtained after direct sequencing of PCR products, using primers MIG1005 and MIG1549 and M. avium A5 or M. avium #5-8 as template, in comparison to homologous sequences from pHis5 and U43598. The deduced amino acid sequences of ORF1 (U43598) and ORFmig (pHis5) in this region are indicated below. The deletion mutation at codon triplet for amino acid 275 of ORF1 is marked grey and the resulting frameshift in the ORFmig amino acid sequence is underlined. (B) Positions of primers MIGHISFW, MIGHISRW, MIG1005 and MIG1549 and of open reading frames (ORF) are indicated.
and the ECL Western blotting detection method (Amersham Pharmacia Biotech). 2.6. Analytical procedures Acyl-CoA synthase activity was measured as described before [9,10]. The following assay conditions were used unless otherwise speci¢ed: 100 mM Tris^HCl (pH 7.8), 5 mM MgCl2 , 1 mM ATP, 0.4 mM CoA, 0.4 mM NADH, 1 mM PEP, 2 mM DTE, 1 U/ml myokinase (EC 2.7.4.3), 2.8 U/ml D-lactate dehydrogenase (EC 1.1.1.27), 1.5 U/ml pyruvate kinase (EC 2.7.1.40), 1^20 Wg/ml protein (Mig) and various concentrations (Table 2) of fatty acids or aromatic acids in a total volume of 1 ml. Incubation was carried out for various times at room temperature. Kinetic studies were done with one of the substrates varied as indicated and with the other substrate at the concentration
used in the standard assay. Inhibition studies were performed in a standard assay using octanoic acid or dodecanoic acid as substrate after preincubation with 2-hydroxydodecanoic acid. Protein was determined according to Bradford (Bio-Rad) with bovine serum albumin as a standard. SDS^PAGE was done according to Laemmli [11]. 3. Results 3.1. ORFmig encodes a 550 amino acid protein The expression vector pHis5 was constructed as described in Section 2 and the DNA sequence was determined three times for each nucleotide. From the resulting nucleotide sequence an open reading frame (ORFmig) was deduced that encoded a 550 amino acid Mig protein (58
Table 1 Feature
Prediction based on
Amino acid position
Orientation
Leader TM AMP TM
SignalP TMpred Prosite TMpred TopPred2 TMpred TopPred2 TopPred2 alignment alignment
1^18 84^103 186^197 245^267 251^271 309^330 311^331 354^370 422^446 430^464
na o-i na i-o na o-i na na na na
TM TM FAD Domain M
Feature, Mig peptide sequence features; TM, transmembrane region; AMP, AMP-binding domain signature; o-i, outside-inside ; i-o, inside-outside; na, not applicable ; FAD, fatty acyl-coenzyme A synthetase signature motif; domain M, medium-chain fatty acyl speci¢city domain (see Section 3).
BBAEXP 93565 17-10-01
62
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
kDa). The only di¡erence between the pHis5 DNA sequence and U43598 [3] is a single deletion of a cytosine residue at the codon triplet for amino acid 275 of ORF1. This results in a frameshift leading to a fusion between ORF1 and ORF3 (Fig. 1). Direct sequencing of PCR products created with primers MIG1005 and MIG1549 veri¢ed this fusion of ORF1 and ORF3 in the genome of M. avium. Fig. 2. SDS^PAGE and Western blot with puri¢ed His-tagged Mig. (A) Expression of His-tagged Mig protein in E. coli. Lanes: 1, molecular weight markers ; 2, lysate of E. coli M15 transformed with pHis5 after induction (80 Wg protein) ; 3, wash fraction; 4^7, elution fractions 1^4 from Ni-agarose column (20^80 Wg protein). (B) Western blot of Mig protein expressed in E. coli. Western blot with a¤nity-puri¢ed AA22-36 antibody [3] is shown. Lanes: W, wash fraction; F2, elution fraction 2 from Ni-agarose column, see above (1 Wg protein).
3.2. Heterologous expression of Mig proteins in E. coli The overexpressions of Mig were carried out using pHis5 and pHis9 and with pGPC150 [3] bacterial expression systems. Ni-agarose a¤nity chromatography resulted in the appearance of a major band on SDS^PAGE with an apparent mobility of approx. 60 kDa (pHis5, Fig. 2A), 31 kDa (pHis9, data not shown) and 28 kDa (pGPC150, data
Fig. 3. Protein domains of fatty acid acyl-CoA synthetases. The predicted primary amino acid sequence of mig gene product was compared with the AMP-binding domain [17] and the fatty acyl-coenzyme A synthetase signature motif [21] of adenylate-building proteins. Amino acid positions responsible for decanoic acid speci¢city (I and II) and for oleic acid speci¢city (III and IV) are on a grey shaded background [21]. Acyl-CoA synthetases: from M. bovis (AcoAS [24]); from P. oleovorans (AlkK [19]); from Archaeoglobus fulgidus (AlkK-5 [32]; putative); from Saccharomyces cerevisiae (FAA1 [33]); from S. cerevisiae (FAA2 [34]); from S. cerevisiae (FAA3 [35]); from E. coli (FadD [16]); putative from M. tuberculosis (FadD19 [20]). M-domain of medium-chain acyl-CoA synthetase: Mig; AlkK [19]; AlkK-5 [32]; Dein (Deinococcus spec. GenBank B75265, putative); and long-chain acyl-CoA synthetases : FadD [16]; FadD19 [20]; ACoAS [24].
BBAEXP 93565 17-10-01
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
63
Table 2 Mig medium-chain fatty acid acyl-CoA synthetase activity using indicated carbon acids as substrates Substrate
Activity (nmol min31 mg31 )
Concentration (WM)
Ethanoic acid (C2:0 ) Butanoic acid (C4:0 ) Pentanoic acid (C5:0 ) Hexanoic acid (C6:0 ) Heptanoic acid (C7:0 ) Octanoic acid (C8:0 ) Decanoic acid (C10:0 ) Dodecanoic acid (C12:0 ) Tridecanoic acid (C13:0 ) Tetradecanoic acid (C14:0 ) Palmitic acid (C16:0 ) Arachidonic acid (C20:4 ) Linolenic acid (C18:3 ) Oleic acid (C18:1 ) Benzoic acid 2-Aminobenzoic acid m-Hydroxybenzoic acid Phenyl acetate
5000 5000 5000 5000 5000 5000 5000 100 100 100 200 10 30 30 2000 2000 2000 2000
S.D. (nmol min31 mg31 )
0.0 0.0 8.0 175.0 170.0 155.0 497.0 153.0 73.0 19.0 0.0 25.7 41.2 111.8 11.6 7.2 3.9 0.0
not shown). The approx. 60 kDa single band (pHis5, Fig. 2B, lane F2) was recognized in Western blot analysis with the a¤nity-puri¢ed AA22-36 anti-Mig antibody [3]. M. avium cultures grown in 7H9 Middlebrook medium supplemented with ADC were analysed by Western blotting. The antibody [3] recognized a band at approx. 60 kDa in cell lysates that was also present in the insoluble fraction. These bands were of a size comparable with the His-Mig fusion protein expressed in E. coli/pHis5. 3.3. Analysis of deduced 58 kDa Mig peptide sequence The Mig amino acid sequence (encoded by ORFmig) was examined. The hydrophilicity pro¢le of the deduced sequence [12] identi¢es several hydrophobic regions. The computer program `Psort' [13] predicted Mig to be a membrane protein. Most notably, the existence of putative transmembrane-associated K-helices (Table 1) was strongly predicted by the programs `TmPred' and `TopPred2' [14,15]. High similarities (data not shown) with procaryotic fatty acid acyl-CoA synthetase were detected for the 58 kDa Mig protein after investigations with NCBI advanced and basic blast programs. This applies especially to the FadD from E. coli [16,17], the long-chain acyl-CoA synthetase from Bacillus subtilis [18], the medium-chain acyl-CoA synthetase from Pseudomonas oleovorans [19]
0.0 0.0 2.5 4.5 3.0 2.5 8.5 28.0 6.0 0.5 0.0 4.9 11.8 11.8 0.1 0.1 0.1 0.0
and a putative fatty acid acyl-CoA synthetase FadD19 from M. tuberculosis [20]. Other signi¢cant characteristics of the 58 kDa Mig protein are a putative leader peptide and an AMP-binding domain signature as described before [3] and a homology with a fatty acyl-coenzyme A synthetase motif [21] of adenylate-forming proteins (Fig. 3). Moreover, we detected a previously undescribed region of high similarity in medium-chain CoA synthetases which is localized C-terminally of the fatty acyl-coenzyme A synthetase motif (Fig. 3, domain M). 3.4. The properties of Mig as a medium-chain fatty acid acyl-CoA synthetase We tested the three derivatives of Mig for acyl-CoA synthetase activity. Only the 58 kDa Mig exhibited an activity using saturated and unsaturated fatty acids (Table 2). This activity depended on the concentration of Mig, of CoA, ATP and fatty acids. Interestingly Mig exhibited acyl-CoA synthetase activities with saturated fatty acids of medium-chain length between C5:0 and C14:0 and with unsaturated fatty acids like oleic acid (C18:1 ), linolenic acid (C18:3 ) and arachidonic acid (C20:4 ). In addition, we detected low activity using benzoic acid as substrate (Table 2). In contrast to the 58 kDa protein no acyl-CoA synthetase activities were detected with any carbon acid as sub-
Table 3 Biochemical characterization of Mig as medium-chain fatty acid acyl-CoA synthetase Value
ATP/CoA
Km (WM) Vmax (nmol min31 mg31 ) Turnover No. (min31 )
36/45 ^ ^
Octanoic acid (C8:0 ) 285 183 11
BBAEXP 93565 17-10-01
Decanoic acid (C10:0 ) 1775 510 30
Dodecanoic acid (C12:0 ) 20 163 10
64
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
strate for both puri¢ed 30 kDa His-Mig derivatives. In order to characterize the kinetic properties of the acyl-CoA synthetase the Km and Vmax values for octanoic acid, decanoic acid and dodecanoic acid and the Km values for ATP and CoA for the recombinant protein were determined (Table 3). It was possible to inhibit the acyl-CoA synthetase activity by preincubation of Mig with 100 WM 2-hydroxydodecanoic acid. The acyl-CoA synthetase activity decreased to 64% in comparison to the untreated Mig protein when 100 WM octanoic acid or 100 WM dodecanoic acid was used as substrate. 4. Discussion Our results demonstrate that the ORF of mig encodes a 550 amino acid peptide. The corrected sequence was veri¢ed by direct sequencing of PCR products ampli¢ed from two di¡erent M. avium strains and by Western blot analysis. This clari¢ed the con£icting results that had previously been achieved after expression of the gene in M. smegmatis, E. coli and M. avium. We demonstrated that Mig is a medium-chain acyl-CoA synthetase. Our results are a ¢rst step towards the elucidation of the biological function of the gene. Together with our previous ¢ndings [1,3] they underscore the importance of enzymes for lipid metabolism in virulent mycobacteria as discussed before [7,22,23]. We characterized the biochemical properties of a puri¢ed medium-chain acyl-CoA synthetase in Gram-positive bacteria for the ¢rst time. The amino acid sequence of Mig shows similarities to acyl-CoA synthetases of E. coli, B. subtilis, P. oleovorans and with peptide synthetases of various bacterial species (data not shown). Interestingly, Mig exhibits only moderate degrees of homology to the M. bovis long-chain acyl-coenzyme A synthase which was characterized before [24,25]. These polypeptides belong to the superfamily of AMP-binding proteins with a characteristic signature motif [17]. In addition to this motif Mig shares homologies with an additional domain of 25 amino acid residues (Fig. 3) that is typical for fatty acylcoenzyme A synthetases. Alterations of amino acid residues within this region have been shown to modulate fatty acid substrate speci¢city [21]. A homologous domain of this fatty acyl-coenzyme A synthetase signature motif is located at the C-terminal part of Mig. The two approx. 30 kDa His-tagged Mig derivatives do not carry this domain and were found to have no acyl-CoA synthetase activity. The 30 kDa proteins that have previously been described [3] are likely functionally inactive degradation products of the protein. Despite the obvious similarities of the gene to the E. coli FadD homologue there are also some remarkable di¡erences from this enzyme. In fact the putative fatty acyl-coenzyme A synthetase motif of Mig has only ¢ve positions of 25 amino acid residues in common with the corresponding domain of FadD. More-
over, three of four important positions for amino acid residues responsible for decanoic acid and oleic acid speci¢city are not identical for the two acyl-CoA synthetases [21], but Mig was still active with these fatty acids as substrate. Here we postulate a novel motif, called domain M, that could be involved in de¢ning the CoA synthetase speci¢city for medium-chain fatty acids (Fig. 3). We found strong similarities between the domain M of Mig and the corresponding domains of medium-chain acyl-CoA synthetases, but a less conserved pattern with long-chain acyl-CoA synthetases (Fig. 3). The substrate speci¢city of Mig is similar to mediumchain acyl-CoA synthetases of Pseudomonas putida and rat kidney [26,27]. Mig was able to ligate CoA to saturated medium-chain fatty acids (C5:0 -C14:0 ) and benzoic acid derivatives as substrates. In the mammalian kidney, activation of benzoic acid derivatives is used for glycine conjugation. This constitutes one of the most important routes of detoxi¢cation, not only for xenobiotic carboxylic acids [27]. In contrast to the characterized M. bovis acyl-CoA synthase or FadD from E. coli, Mig could not ligate coenzyme A to palmitic acid (C16:0 ) or longer saturated fatty acids [24,28]. The acyl-CoA synthetase activities of Mig as a function of the tested fatty acid concentrations followed the Michaelis^Menten kinetics, whereas the rat intestinal medium-chain acyl-CoA synthetase activity as a function of octanoate concentration has been shown to follow a biphasic reaction curve [29]. Mig showed the highest activity (Vmax : 510 nmol min31 mg31 ) with decanoic acid but had the lowest a¤nity to this molecule (Km : 1775 WM) in comparison to octanoic acid (Km : 183 WM) and dodecanoic acid (Km : 20 WM) (Table 3). The Mig protein has four predicted transmembrane regions and is localized in the cytoplasm membrane with the `Psort' program. This prediction is in accordance with the localization of the acyl-CoA synthase (M. bovis) and FadD (E. coli) [16,24]. We conclude that Mig is involved in activation of free fatty acids that will then either be degraded by L-oxidation or serve as intermediates in the synthesis of very long fatty acids, of lipoproteins or glycolipids. In mycobacteria the activation of fatty acids for L-oxidation can be linked to the biosynthesis of glucose via the glyoxalate shunt pathway, which is important for the survival inside the phagosome of macrophages [7]. On the other hand are free fatty acids together with reactive nitrogen intermediates and reactive oxygen intermediates e¡ector molecules in the expression of antimicrobial activity of the macrophage against intracellular mycobacteria [30,31]. Free fatty acids could be detoxi¢ed by CoA ligation and subsequent Loxidation. Furthermore, the arachidonic acid modulating activity of Mig may also interfere with the leukotriene synthesis by the host, where the fatty acid is a precursor of the e¡ectors in the antimicrobial response. Taken together, these ¢ndings argue for a role of this gene in the virulence of this organism.
BBAEXP 93565 17-10-01
C. Morsczeck et al. / Biochimica et Biophysica Acta 1521 (2001) 59^65
References [1] G. Plum, J.E. Clark Curtiss, Infect. Immun. 62 (1994) 476^483. [2] M.L. Beggs, R. Stevanova, K.D. Eisenach, J. Clin. Microbiol. 38 (2000) 508^512. [3] G. Plum, M. Brenden, J.E. Clark-Curtiss, G. Pulverer, Infect. Immun. 65 (1997) 4548^4557. [4] V. Balasubramanian, M.S. Pavelka, S.S. Bardarov, J. Martin, T.R. Weisbrod, R.A. McAdam, B.R. Bloom, W.R. Jacobs, J. Bacteriol. 178 (1996) 273^279. [5] A.K. Azad, T.D. Sirakova, L.M. Rogers, P.E. Kolattukudy, Proc. Natl. Acad. Sci. USA 93 (1996) 4787^4792. [6] E. Mahenthiralingam, B.I. Marklund, L.A. Brooks, D.A. Smith, G.J. Bancroft, R.W. Stokes, Infect. Immun. 66 (1998) 3626^3634. [7] J.D. McKinney, K. Honer zu Bentrup, E.J. Munoz-Elias, A. Miczak, B. Chen, W.T. Chan, D. Swenson, J.C. Sacchettini, W.R. Jacobs Jr., D.G. Russell, Nature 406 (2000) 735^738. [8] M.C. Lorenz, G.R. Fink, Nature 412 (2001) 83^86. [9] T. Tanaka, K. Hosaka, S. Numa, Methods Enzymol. 71 (1981) 334^ 341. [10] K. Ziegler, R. Buder, J. Winter, G. Fuchs, Arch. Microbiol. 151 (1989) 171^176. [11] U.K. Laemmli, Nature 227 (1970) 680^685. [12] J. Kyte, R.F. Doolittle, J. Mol. Biol. 157 (1982) 105^132. [13] K. Nakai, M. Kanehisa, Genomics 14 (1992) 897^911. [14] G. von Heijne, J. Mol. Biol. 225 (1992) 487^494. [15] K. Hofmann, W. Sto¡el, Biol. Chem. Hoppe-Seyler 374 (1993) 166^ 178. [16] P.N. Black, C.C. DiRusso, A.K. Metzger, T.L. Heimert, J. Biol. Chem. 267 (1992) 25513^25520. [17] M. Fulda, E. Heinz, F.P. Wolter, Mol. Gen. Genet. 242 (1994) 241^ 249. [18] A. Wipat, N. Carter, S.C. Brignell, B.J. Guy, K. Piper, J. Sanders, P.T. Emmerson, C.R. Harwood, Microbiology 142 (1996) 3067^3078. [19] J.B. van Beilen, G. Eggink, H. Enequist, R. Bos, B. Witholt, Mol. Microbiol. 6 (1992) 3121^3136.
65
[20] S.T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris, S.V. Gordon, K. Eiglmeier, S. Gas, C.E. Barry III, F. Tekaia, K. Badcock, D. Basham, D. Brown, T. Chillingworth, R. Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S. Holroyd, T. Hornsby, K. Jagels, B.G. Barrell et al., Nature 393 (1998) 537^544. [21] P.N. Black, Q. Zhang, J.D. Weimar, C.C. DiRusso, J. Biol. Chem. 272 (1997) 4896^4903. [22] W. Bishai, Nature 406 (2000) 683^685. [23] M.S. Glickman, W.R. Jacobs Jr., Cell 104 (2001) 477^485. [24] A.M. Fitzmaurice, P.E. Kolattukudy, J. Bacteriol. 179 (1997) 2608^ 2615. [25] A.M. Fitzmaurice, P.E. Kolattukudy, J. Biol. Chem. 273 (1998) 8033^8039. [26] M. Fernandez-Valverde, A. Reglero, H. Martinez-Blanco, J.M. Luengo, Appl. Environ. Microbiol. 59 (1993) 1149^1154. [27] F. Kasuya, K. Igarashi, M. Fukui, Chem.-Biol. Interact. 118 (1999) 233^246. [28] K. Kameda, W.D. Nunn, J. Biol. Chem. 256 (1981) 5702^5707. [29] Y. Ohkubo, S. Mori, Y. Ishikawa, K. Shirai, Y. Saito, S. Yoshida, Digestion 51 (1992) 42^50. [30] T. Akaki, K. Sato, T. Shimizu, C. Sano, H. Kajitani, S. Dekio, H. Tomioka, J. Leukocyte Biol. 62 (1997) 795^804. [31] H. Tomioka, K. Sato, C. Sano, T. Akaki, T. Shimizu, H. Kajitani, H. Saito, Clin. Exp. Immunol. 109 (1997) 248^254. [32] H.P. Klenk, R.A. Clayton, J.F. Tomb, O. White, K.E. Nelson, K.A. Ketchum, R.J. Dodson, M. Gwinn, E.K. Hickey, J.D. Peterson, D.L. Richardson, A.R. Kerlavage, D.E. Graham, N.C. Kyrpides, R.D. Fleischmann, J. Quackenbush, N.H. Lee, G.G. Sutton, S. Gill, E.F. Kirkness, B.A. Dougherty, K. McKenney, M.D. Adams, B. Loftus, J.C. Venter et al., Nature 390 (1997) 364^370. [33] R.J. Duronio, L.J. Knoll, J.I. Gordon, J. Cell Biol. 117 (1992) 515^ 529. [34] D.R. Johnson, L.J. Knoll, N. Rowley, J.I. Gordon, J. Biol. Chem. 269 (1994) 18037^18046. [35] L.J. Knoll, D.R. Johnson, J.I. Gordon, J. Biol. Chem. 269 (1994) 16348^16356.
BBAEXP 93565 17-10-01