Nonorganellar Acyl Carrier Protein from Oleaginous Yeast Is a Homologue of Ribosomal Protein P2*

Acyl carrier protein (ACP) is responsible for carrying the growing fatty acid chain from one enzyme active site to the next during fatty acid biosynthesis. Here we report the identification, purification, immunocytochemical localization, and cloning of ACP from the oleaginous yeast, Rhodotorula glutinis. The soluble fraction of this organism can synthesize triacylglycerol and is able to accept the acyl group from acyl-ACP for the synthesis. The ACP, cloned from the system, showed a significant similarity with ribosomal protein P2. Expression and characterization of the recombinant protein showed that the ACP was acylated in vitro. The recombinant protein was post-translationally modified, since it was observed in [14C]β-alanine labeling and matrix-assisted laser desorption mass spectroscopic analysis. Site-directed mutants were generated to identify a serine residue responsible for phosphopantetheinylation and found that mutation of serine 59 to alanine abrogated the fatty acylation ability of the protein. These results demonstrate that a novel modification of ribosomal protein P2 allows it to act as an acyl carrier protein and participate in acylation reactions.

Acyl carrier protein (ACP) 1 of Escherichia coli is a small, acidic protein containing a prosthetic group 4Ј-phosphopantetheine to which fatty acyl chains are linked as thioester. ACP acts as a carrier for substrates and intermediates from the active site of one enzyme to the other in the fatty acid synthase (1), nonribosomal polypeptide synthetases (2), and polyketide synthase (3). ACP is also involved in phospholipid synthesis (4,5), membrane-bound oligosaccharide synthesis (6) in Escherichia coli , and host-specific lipochitin oligosaccharide signal molecule synthesis in Rhizobium leguminosarum (7). Several isoforms of ACP have been identified in plant chloroplasts and are shown to be involved in specific fatty acid synthesis during seed development (8). In Neurospora crassa (9) and bovine heart (10), ACP was identified in mitochondria as a subunit of the respiratory NADH-ubiquinone oxidoreductase complex. In Saccharomyces cerevisiae, nuclear encoded and 19-kDa molecular mass ACP has been reported in mitochondria (11,12), but this ACP is not part of the respiratory complex (13). No distinct low molecular mass nonorganellar ACP has been identified in eukaryotes, and this could be due to the presence of fatty acid synthase multifunctional enzyme (type I). Eukaryotic ACP possesses characteristic organellar import sequence in front of the known core sequence (12). The overall three-dimensional structure and the amino acid sequences around the phosphopantetheine-binding site of ACP are well conserved (14,15).
A soluble triacylglycerol biosynthetic complex (TBC) from oleaginous yeast was isolated, and the individual components were characterized (16). The complex consists of ACP, acyl-ACP synthetase, lysophosphatidic acid (LPA) acyltransferase, phosphatidic acid phosphatase, diacylglycerol (DAG) acyltransferase (16), and superoxide dismutase (17). The acyl-ACP synthetase in the multienzyme complex activates free fatty acids to acyl-ACP in an ATP-dependent manner, and the synthesized acyl-ACP is used as a substrate by TBC (18). Mutation in any of the acyltransferases present in TBC or alteration in nutritional content leads to decreased synthesis of triacylglycerol in Rhodotorula glutinis (19).
Here we report the identification, purification, and immunolocalization of a nonorganellar ACP from oleaginous yeast that is an integral part of TBC. Cloning, expression, and subsequent characterization of the protein reveals that it has a sequence homology to ribosomal protein P2. The protein is post-translationally modified to accept free fatty acid, and the acyl-ACP formed can act as acyl group donor in acylation reactions. Alteration of serine 59 to alanine by site-directed mutation abolishes the post-translational modification, and therefore the fatty acid esterification ability is lost in the ACP.
Metabolic Labeling of R. glutinis Cells with [ 14 C]␤-Alanine-R. glutinis cells were grown in complete minimal medium (containing all of the amino acids) until early log phase (15 h) and then transferred to aspartate-less minimal medium to which [ 14 C]␤-alanine (10 Ci/ml culture) was added and incubated further for 12 h. Cells were centrifuged, washed, and resuspended in lysis buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 5 mM MgCl 2 , 5% sucrose, and 1 mM phenylmethylsulfonyl fluoride), and glass bead lysis was performed. Cytosol and TBC were prepared (16) and resolved on 7% native PAGE and 15% SDS-PAGE followed by fluorography.
LPA and DAG acyltransferases and phosphatidic acid phosphatase activities are collectively represented as TAG synthase (16). This assay was performed by monitoring the incorporation of [ 14 C]oleoyl-CoA (100,000 dpm) or [ 14 C]oleoyl ACP (35,000 dpm) into TAG in the presence of the acyl acceptor, LPA. The TAG synthase assay mixture consisted of 5-25 g of enzyme, 20 M acyl acceptor, 50 M LPA in a total volume of 100 l. The reaction was carried out at 30°C for 30 min and stopped by lipid extraction (21). Lipids were separated on silica-TLC plates using petroleum ether/diethyl ether/acetic acid (70:30:1, v/v/v) and visualized by staining with iodine vapor. The spot corresponding to TAG was scraped, and radioactivity was measured in a liquid scintillation counter.
Purification of ACP-Acyl carrier protein was purified from R. glutinis as described (22). Late logarithmic phase cells (35 g, wet weight) were centrifuged and resuspended in 10 mM Tris-HCl (pH 7.5), 0.1 M NaCl, 5 mM MgCl 2 , 1 mM phenylmethylsulfonyl fluoride, and 100 mM leupeptin. The cells were lysed using glass beads (0.45-0.6 mm) in the absence of detergent, and the lysate was precleared by centrifugation at 300 ϫ g for 15 min. Lysate was extracted with isopropyl alcohol followed by DEAE-cellulose chromatography and acid precipitation. The purity of the protein was monitored on 15% SDS-PAGE and isoelectric focusing gel electrophoresis. The purified ACP was used as the substrate for E. coli acyl-ACP synthetase.
Antibody Generation and Purification-Antisera were produced against the purified ACP as described earlier (16). The IgG fraction of the polyclonal antibodies was purified by passing the antisera through a protein A-agarose column after adjusting the pH of the antibody preparation to 8.0 (23). The column was washed with 100 mM Tris-HCl (pH 8.0) followed by 10 mM Tris-HCl (pH 8.0). The bound antibodies were eluted with 100 mM glycine (pH 3.0), and the pH was immediately brought back to neutral by the addition of 1 M Tris (pH 8.0). Immunoglobulin fractions were identified by separating them in 10% SDS-PAGE. The titer and the specificity of antibodies were determined by direct and inhibition enzyme-linked immunosorbent assay (24). This purified fraction was used for further analysis.
The peptide CYVVGAQPFGGARGS, corresponding to DAG acyltransferase, was conjugated to bovine serum albumin using m-maleimidobenzoyl-N-hydroxysuccinimide ester, and antisera were raised as described (18). The immunoglobulin fraction from the antisera was precipitated with a 33% saturation of ammonium sulfate, the precipitate was dialyzed extensively against phosphate-buffered saline, and purity was monitored in 10% SDS-PAGE. This immunoglobulin fraction was kept for further use. Protein was estimated by the method of Bradford (25) using bovine serum albumin as the standard.
Purification of Acyl-ACP Synthetase from E. coli-Acyl-ACP synthetase from E. coli was partially purified (26). Overnight grown E. coli (JM109) cells were lysed by passing through a French pressure cell (Aminco) at 16,000 p.s.i. The lysate was centrifuged at 10,000 ϫ g for 20 min, and MgCl 2 was added to the supernatant to make a final concentration of 10 mM and centrifuged at 80,000 ϫ g for 90 min. The resulting pellets were suspended uniformly in 50 mM Tris-HCl (pH 8.0), and NaCl and MgCl 2 were added to reach a final concentration of 0.5 M and 10 mM, respectively. The solution was stirred for 15 min and centrifuged at 80,000 ϫ g for 90 min. The membrane pellet was suspended in 50 mM Tris-HCl, pH 8.0, and Triton X-100 was added to a final concentration of 2%. The suspension was then centrifuged at 80,000 ϫ g for 90 min. ATP was added to the supernatant to make a final concentration of 5 mM and then placed in a 55°C water bath for 7 min. The solution was cooled and centrifuged at 15,000 ϫ g for 20 min, and the resultant supernatant was used as the source of the enzyme.
Double Immunofluorescence-Logarithmic phase (21-h) R. glutinis cells were fixed with 4% paraformaldehyde for 20 min followed by 4% formaldehyde for 60 min. The fixed cells were washed three times with 0.1 M phosphate buffer (pH 5.9) and resuspended (10 7 cells/ml) in 1.2 M sorbitol, and lyticase was added to a final concentration of 100 g/100 l of cells. Cells were incubated at 30°C for 6 h to obtain spheroplasts. The cells were washed, resuspended, and plated onto poly-L-lysine-pretreated slides. The slides were treated with ice-cold methanol for 6 min and acetone for 30 s.
Slides were first treated with purified anti-ACP antibodies (1:100 dilution) followed by TRITC-conjugated goat anti-rabbit secondary antibodies. Slides were washed thoroughly with PBST and phosphatebuffered saline and treated with goat anti-rabbit immunoglobulin (1: 100 dilution) to block the unoccupied sites of primary antibodies. Anti-DAG acyltransferase antibodies (IgG fraction) were added in 1:100 dilution followed by fluorescein isothiocyanate-conjugated goat antirabbit secondary antibodies. The slides were viewed in a confocal laserscanning microscope (Leica, TCS SP, Heidelberg, Germany) for localization.
Purification of Acyl-ACP-Acyl-ACP formed in the reaction mixture was purified through DEAE-Sepharose column chromatography followed by blue Sepharose column chromatography (27). A DEAE (1-ml) column was made and equilibrated with 20 mM Tris-HCl (pH 8.0). The acylation reaction product was loaded onto the column and washed with 10 ml of 20 mM Tris-HCl (pH 8.0). The column was then washed with 50% isopropyl alcohol in 20 mM Tris-HCl (pH 8.0) to remove the unreacted fatty acid. Bound protein was eluted with 0.5 M NaCl, and 1-ml fractions were collected. Acyl-ACP-containing fractions were pooled and dialyzed against 20 mM Tris-HCl (pH 8.0). Fractions were checked for

S12A T98G
Sense: 5Ј-CCTCCTCGTCGCGCCGGCAACA-3Ј Antisense: 5Ј-TGTTGCCGGCGCGACGAGGAGG-3Ј S17A T113G Sense: 5Ј-CGGCAACACCGCGCCCTCGGCCG-3Ј Antisense: 5Ј-CGGCCGAGGGCGCGGTGTTGCCG-3Ј S19A T119G radioactivity in the case of radioactive assay or run on a 15% SDS-PAGE followed by silver staining. Blue Sepharose column (1 ml) was equilibrated with 20 mM Tris-HCl (pH 8.0) containing 2% Triton X-100. The dialyzed fraction from the DEAE column was loaded onto the blue Sepharose column, and flow-through was collected and checked for acyl-ACP. An acyl-ACP (500 l) sample was taken, 100 l of 0.5 N NaOH was added to it, and the sample was incubated at 100°C for 15 min. Reaction mixture was cooled to room temperature and neutralized. Lipids were extracted and resolved on a silica-TLC, using petroleum ether/diethyl ether/acetic acid (70:30:1, v/v/v) as the solvent system followed by autoradiography.
Modification of Sulfhydryl Group by Iodoacetate or Iodoacetamide-ACP (wild type or mutant) was taken as 2 mg/ml solution in 50 mM Tris (pH 8.5), 0.15 M NaCl, 5 mM EDTA. Iodoacetate or iodoacetamide was added to the reaction mixture to a final concentration of 50 mM and incubated for 2 h at room temperature (28). Excess reagent was removed by dialysis, and protein was used for enzymatic assay and MALDI mass analysis. Western Blotting-Proteins were separated by either native or SDS-PAGE and transferred onto a nitrocellulose membrane for immunoblotting. ACP antibodies were used at a dilution of 1:2000 in Western blot analysis and at 1:1000 dilution for cDNA library screening.
Ribosome Preparation-Exponential phase R. glutinis cells were pelleted down, resuspended in an equal volume of ice-cold 2ϫ buffer A (20 mM Tris (pH 7.4), 16 mM MgCl 2 , 0.1 M KCl, 12 mM 2-mercaptoethanol, FIG. 1. Identification and purification of soluble ACP from R. glutinis. A, early log phase R. glutinis cells, grown in aspartate-less minimal medium, were metabolically labeled with [ 14 C]␤-alanine (10 Ci/ml culture) for 12 h. Incorporation of label was monitored by resolving cytosol (50 g) (lane 2) and purified TBC (10 g) (lane 3) in native and SDS-PAGE followed by fluorography. Lane 1, labeling in Luria-Bertani medium. B, the cytosol was separated on a 7% native PAGE, and the proteins were eluted from the gel pieces and assayed for ACP. The gel-eluted fractions were also assayed for the incorporation of labeled free fatty acids into triacylglycerol in the presence of LPA as acyl acceptor, and these are represented as TAG synthase activities. C, cytosol was electrophoresed on 7% native PAGE and transferred to a nitrocellulose membrane that was probed with anti-ACP antibodies. The antibodies recognized two bands under nondenaturing conditions. D, purification profile of ACP was determined by resolving the samples from each purification step on a 13% SDS-PAGE followed by silver staining. Lane 1, isopropyl alcohol supernatant; lane 2, DEAE-cellulose eluate; lane 3, acid precipitation step; lane 4, molecular weight markers. Approximately 10 g of protein was loaded in each lane. E, purified ACP (2.5 g) from R. glutinis (lane 2) was resolved in a 15% SDS-PAGE, and Western blot was performed with the anti-ACP antibodies. Normal rabbit serum was used as a control (lane 1). F, displacement of binding of ACP antibodies to immobilized ACP by the free ACP as determined by competitive inhibition enzyme-linked immunosorbent assay. ACP was coated in a microtiter plate (1 g/well) followed by blocking the unoccupied sites. Varying concentrations of ACP (0.1-1 M) was incubated with ACP antibodies at a dilution that showed 50% of maximum binding to 1 g of the immobilized ACP. and 0.2 mM EDTA), and lysed with glass beads. Lysate was centrifuged at 10,000 ϫ g for 15 min, and the resulting supernatant was centrifuged at 20,000 ϫ g for 30 min. The supernatant was loaded onto 10 -40% linear sucrose density gradient and centrifuged for 17 h at 20,000 ϫ g. The pellet containing polysome was collected and used for analysis. All of the operations were performed at 4°C (29).
Cloning, Expression in E. coli, and Purification of Recombinant ACP-cDNA library was constructed using the cDNA library construction kit and pBK-CMV ZAP expression vector (17). cDNA library was screened using the purified polyclonal antisera raised against R. glutinis ACP. The positive ZAPII phage plaques were excised using a helper phage (ExAssist) and sequenced in both directions using T7 and T3 primers in a DNA sequencer (Applied Biosystems model 377).
For expression of recombinant ACP in E. coli, a 369-bp DNA fragment containing the coding sequence of ACP was generated by PCR amplification of the cDNA clone using forward primer (5Ј-TCTATTCA-GCTCCATATGAAGCACGTCGCC-3Ј) and reverse primer (5Ј-GGGGA-GAGGCGAGAATTCTCATTAGTCGAAGAG-3Ј). The forward primer contains a NdeI site followed by the beginning of an open reading frame; the reverse primer contains an EcoRI site and a stop codon after the open reading frame. PCR (1 min of denaturation at 94°C, 1 min of annealing at 55°C, and 45 s of elongation at 72°C) was performed using Pfu polymerase for 30 cycles with a 10 mM concentration of each primer in a final volume of 100 l. The purified PCR product was digested with NdeI and EcoRI and ligated directionally in predigested pET 21a (Novagen) vector. The construct was transformed into E. coli BL21 (DE3) cells, and the insert was sequenced. The transformed E. coli cells were induced with 0.5 mM IPTG for 4 h, and the cell lysate was resolved on 15% SDS-PAGE to look for the expression.
Recombinant ACP was purified by Ni 2ϩ -NTA affinity column chromatography. The cell pellet was resuspended in lysis buffer containing 50 mM Tris-HCl (pH 8.0) and 300 mM NaCl (buffer A). Cells were disrupted by passing through a French pressure cell at 16,000 p.s.i. The supernatant (10,000 ϫ g) was allowed to bind to the Ni 2ϩ -NTA matrix. The column was washed with buffer A containing 20 mM imidazole. The bound protein was eluted with NaCl containing 250 mM imidazole. One-milliliter fractions were collected and checked for the presence of protein in 15% SDS-PAGE followed by silver staining. Fractions containing ACP were pooled, dialyzed extensively using benzoylated dialysis tubing, concentrated, and used for further assays.
Site-directed Mutagenesis-Wild type ACP template (80 ng), sense, and antisense primers (25 pmol each) were added to PCR tubes containing 0.2 mM dNTPs, 1 mM MgSO 4 , 2.5 milliunits of Pfu polymerase, and 1ϫ reaction buffer. Amplification was carried out using the following conditions: denaturation of the template at 94°C for 4 min followed by 20 cycles at 94°C for 45 s (denaturation), 52°C for 1 min (annealing), and 72°C for 6 min (extension). The reaction was continued for another 20 min at 72°C to complete the extension. The sequences of sense and antisense primers are described in detail in Table I. The PCR-amplified product was treated with DpnI at 37°C for 1 h to digest the methylated template and transformed to E. coli DH5␣ strain. The presence of mutations was confirmed by sequencing the plasmid.
Matrix-assisted Laser Desorption Ionization Mass Spectroscopic Analysis-Two micrograms of WT or mutant ACP was resolved in 15% SDS-PAGE and visualized by Coomassie Brilliant Blue staining. The band of interest was excised from the gel, placed in an acid-washed tube, and dried under vacuum. Then the gel pieces were washed twice with 50 l of water for 30 min. Then the gel pieces were alternately dehydrated and rehydrated in acetonitrile and 100 mM ammonium bicarbonate, respectively, until they became white. Trypsin (80 ng/ml in 25 mM ammonium bicarbonate, pH 8.5) was added to the dry gel pieces and allowed to stand for 45 min on ice. The excess trypsin was removed, and the gel pieces were soaked in a minimum amount of 25 mM ammonium bicarbonate (pH 8.5) and incubated overnight at 30°C. Tryptic fragments were extracted by 50% acetonitrile plus 0.1% trifluoroacetic acid, dried, and suspended in 2,5-dihydroxybenzoic acid and applied to a MALDI sample plate. MALDI mass spectroscopic analysis was performed on a Kratos PCKompact Seq 1.2.2 mass spectrometer in linear mode. The masses obtained were compared with the predicted mass of the tryptic fragments obtained from PeptideMass (available on the World Wide Web at www.expasy.org/tools/peptide-mass).

RESULTS
Identification of an ACP-Acyl carrier protein contains a 4Ј-phosphopantetheine prosthetic group, attached to a conserved serine residue as a post-translational modification. Phosphopantetheine is synthesized from aspartate in a twostep reaction, and ␤-alanine is an intermediate in this pathway (1). To show the presence of ACP in soluble fraction, wild type R. glutinis cells were metabolically labeled with [ 14 C]␤-alanine in aspartate-less medium, and cytosol was resolved on a 13% SDS-PAGE. A small ϳ14-kDa protein was found to be labeled (Fig. 1A). In another set of experiments, incorporation of [ 14 C]␤-alanine in TBC was monitored by isolating TBC from the labeled cells and resolving the purified TBC on SDS-PAGE (Fig. 1A). These experiments suggested that the ϳ14-kDa polypeptide that was labeled by ␤-alanine could be soluble ACP.
To estimate the amount of cellular ACP associated with other components of TBC, cytosol was resolved on a 7% native PAGE, and 1-cm fractions were cut from the resolving gel. Protein was eluted, and ACP and TAG synthase activities were measured as described in detail under "Experimental Procedures." It was observed that 63% of the total ACP activity was associated with TBC, and the remaining activity was found in free form (Fig. 1B). Western blot analysis of cytosol with anti-ACP antibodies, after separating in 7% native-PAGE, showed two reactive bands, one at the place of TBC and other at around 14 kDa (Fig. 1C).
Purification of ACP-Acyl carrier protein was purified from R. glutinis by treating the cell lysate with isopropyl alcohol where ACP partitioned into the isopropyl alcohol supernatant. The ACP was then adsorbed onto DEAE-cellulose followed by elution with increasing concentrations of salt. The acidic nature of ACP was used to selectively precipitate the protein.
ACP was purified to 8-fold with a recovery of only 0.5% (Table  II), which could be due to the loss of activity during purification. The activity of ACP was determined by estimating the formation of acyl-ACP, and the specific activity of the purified protein was calculated to be 1.7 nmol/min/mg protein. The purified ACP showed an aberrant mobility on SDS-PAGE, which could be due to the highly acidic nature of the protein (Fig. 1D). The pI was determined to be ϳ4 by isoelectric focusing (data not shown). Polyclonal antisera were raised against the purified ACP, and the IgG fraction was purified by a protein A-agarose column. In Western blot analysis with cytosol, the purified antibodies recognized a polypeptide of 14 kDa (Fig.  1E). The antiserum was titrated against 1 g of immobilized ACP. The 50% inhibitory binding was found at a 1:600 dilution of the antibodies. Inhibition enzyme-linked immunosorbent assay was performed using the same dilution of antiserum and different concentrations of the purified ACP. The binding (50%) was achieved with 2.6 M protein (Fig. 1F).
Immunocytochemical Localization of ACP in TBC-To determine whether ACP colocalizes with other components of TBC, double immunostaining experiments were performed. This involves the probing of the spheroplasts of oleaginous yeast cells with anti-ACP and anti-DAG acyltransferase antibodies. Confocal analysis revealed a substantial colocalization of both the proteins in cytosol (yellow color in Fig. 2c). Cloning of Acyl Carrier Protein-A cDNA expression library was screened with antibodies raised against the purified ACP. After screening ϳ1 ϫ 10 5 plaque-forming units, three strong positive signals were obtained. All of the clones were excised and propagated into E. coli XLOLR strain; plasmids were prepared and subjected to nucleotide sequencing. Sequences obtained were identical in all three with a few base pair changes.
The primary structure of the cDNA is depicted in Fig. 3. The isolated clone contained a full-length cDNA of 565 nucleotides with a 64-bp 5Ј-untranslated region. Analysis of the nucleotide sequence revealed an open reading frame of 110 amino acids with translation initiation at the 65 ATG 67 codon. The calculated molecular mass of the protein is 11.3 kDa, and the isoelectric point is 3.9. A hydropathy plot of the predicted protein revealed that this clone does not have any transmembrane domain (data not shown). We used PROSITE domain search for structural motifs and found a potent protein kinase C phosphorylation site ( 59 SKK 61 ), two casein kinase 2-dependent phosphorylation sites ( 19 SAED 22 and 100 SDDD 103 ), and three myristoylation sites ( 69 GAAPAA 74 , 77 GGAAAG 82 , and 78 GAAAGG 83 ). BLAST search analysis for the deduced amino acid sequence in all of the available databases suggested that the protein has more than 60% identity with ribosomal protein P2 from Cladosporium, Alternata, Saccharomyces, maize, rats, and humans. Multiple sequence alignment was carried out using a ClustalW algorithm with the ACP clone using the above mentioned ribosomal protein P2, and it was observed that the isolated ACP has an overall sequence homology with ribosomal protein P2 (Fig. 4A). No significant homology was observed when multiple sequence alignment was carried out with different ACP sequences from Helicobacter pylori, Plasmodium, Pseudomonas, E. coli, S. cerevisiae, and Brassica (Fig.  4B). When the most conserved region of the ACP was used to carry out a local alignment (LALIGN), we could see more than 70% identity with the isolated cDNA in the nucleotide level (data not shown). This sequence was submitted to GenBank TM with the identification as nonorganellar ACP (accession number AF434667).
Expression of ACP-The full-length R. glutinis ACP cDNA in pBK-CMV expression vector containing the lac promoter was used to transform E. coli (JM109) and induced with IPTG to produce recombinant protein. Although the induction was observed in SDS-PAGE as well as in Western blot, the level of expression was very low (data not shown). To obtain a high level of expression, the ACP gene was PCR-amplified from the parent vector (pBK-CMV) with gene-specific primers and subcloned into pET 21a vector. The transformed E. coli BL21 (DE3) cells showed a very high level of expression upon induction with 0.5 mM IPTG as compared with uninduced control (Fig. 5A). E. coli BL21 (DE3) cells expressing ACP were simultaneously induced and metabolically labeled with [ 14 C]␤-alanine. Cell lysate from the [ 14 C]␤-alanine labeling reaction was resolved on a 13% SDS-PAGE, and the recombinant protein was found to be radiolabeled after fluorography (Fig. 5B). Uninduced cell lysate showed the labeling of the endogenous E. coli ACP. Western blot analysis with anti-ACP antibodies (generated against the recombinant protein) confirmed the identity of both of the proteins (Fig. 5C). Western blot analysis FIG. 4. Multiple sequence alignment of isolated ACP. A, the conceptual translation product of the cDNA identified from TBC was used to search the SwissProt data base. Six polypeptides (60 S ribosomal protein P2 from Cladosporium (P42038); Schizosaccharomyces pombe (P08094); major allergen of Alternaria (P42037); rat (P02401); human (P05387); and maize (O24415) ribosomal proteins), which appeared as significant hits in a BLAST search (scoring over 60), were multiply aligned with the cloned cDNA using ClustalW algorithm. B, polypeptide sequences of different ACPs (E. coli, S. cerevisiae, H. pylori, Plasmodium, Pseudomonas, Brassica) were obtained from a public data base (SwissProt) and were used to do a multiple sequence alignment with the isolated ACP clone using the ClustalW algorithm. On the consensus line below each section of the alignment, in both A and B, asterisks indicate identical or conserved residues in all sequences in the alignment, colons indicate conserved substitutions, and dots denote semiconserved substitutions. was performed with ribosome using the same antisera, and it recognized a low molecular weight protein (Fig. 5C).
Purification and Functional Characterization of Recombinant ACP-Wild type ACP, overproduced in E. coli Bl21 (DE3) using a T7 RNA polymerase expression system, was induced by IPTG. Most (ϳ80%) of the overexpressed protein was found to be in insoluble inclusion body. The remaining overexpressed protein in the soluble fraction was purified by Ni 2ϩ -NTA affinity chromatography (Fig. 6A).
To determine the acylation capacity of the recombinant wild type protein, purified rACP was subjected to in vitro acylation assay using E. coli acyl-ACP synthetase and radiolabeled oleic acid. The amount of acyl-ACP formed (pmol/min) was comparable with that of native protein. Acyl-ACP synthetase or ACP alone did not show any significant acylation (Fig. 6B).
The acylation reaction in ACP is a unique phenomenon, since it involves the formation of a thioester linkage between the thiol group of phosphopantetheinylated protein and the acyl  5-8). B, Ni 2ϩ -NTA-purified rACP (100 pM) was used in in vitro acylation assay using enriched E. coli acyl-ACP synthetase and [ 14 C]oleic acid. R. glutinis ACP and E. coli ACP were used as positive controls. Four independent experiments were performed in triplicate, and mean Ϯ S.E. has been represented. C, purified rACP (100 pM) was treated with 10 mM iodoacetamide or 10 mM iodoacetic acid for 2 h and extensively dialyzed, and an acylation assay was performed. The percentage of specific activity was represented, considering no treatment as 100% activity. D, acyl-ACP was purified from a bulk cold acylation reaction and passed through DEAE column chromatography. The fractions (containing around 5 g of protein) were checked in 15% nonreducing SDS-PAGE followed by silver staining. group. To understand the specificity of the acylation, purified rACP was pretreated with thiol group blockers, such as iodoacetamide and iodoacetic acid, and an acylation reaction was performed after complete removal of the excess reagent. Iodoacetamide treatment inhibited the acylation capacity to 80%, whereas iodoacetic acid treatment left back around 35% of acylation activity. This difference could be attributed to the fact that iodoacetamide binds the thiol group irreversibly, whereas iodoacetic acid binds reversibly (Fig. 6C).
The acylated ACP was purified from the reaction mixture by passing it through DEAE-Sepharose followed by blue Sepharose column chromatography. The purified acylated product was resolved onto a 15% nonreducing SDS-PAGE followed by silver staining. The acylated ACP binds SDS better and runs at a faster rate in PAGE as compared with ACP (Fig. 6D).
The purified [ 14 C]oleoyl ACP was subjected to alkaline hydrolysis, and the extracted lipids were separated on a TLC followed by autoradiography. The alkaline hydrolyzed product was identified as fatty acid (data not shown).

Site-directed Mutagenesis and Analysis of Mutant Proteins-
Based on the sequence analysis of isolated clone, all eight serine residues present in the protein (Fig. 7A) were subjected to site-directed mutagenesis and substituted with alanine. All of the mutant proteins were overexpressed in E. coli BL21 (DE3) cells, induced with IPTG, and purified through Ni 2ϩ -NTA affinity chromatography. In Western blot analysis, anti-ACP antibodies recognized all of the mutant proteins, and there was no apparent mobility difference in the mutant proteins as compared with wild type (data not shown).
The acylation abilities of all of the mutant proteins were assessed by the in vitro acylation assay using E. coli acyl-ACP synthetase. Replacement of Ser-59 with alanine reduced the acylation capability of the protein to 10 -15%. Other mutations (S12A, S17A, S19A, S40A, S64A, S67A, and S100A) did not show any significant change in their acylation pattern as compared with wild type ACP (Fig. 7B) protein band corresponding to rACP was detected in S59A mutant, whereas all other mutants showed a profile similar to that of wild type (Fig. 7C).
Wild type and S59A mutant proteins were subjected to an acylation reaction using [ 14 C]oleic acid, and the acylated product was separated in a 15% SDS-PAGE in the presence and the absence of 2-mercaptoethanol followed by fluorography. The acylated wild type protein showed a strong signal in fluorography, which disappeared upon 2-mercaptoethanol treatment, whereas mutant protein did not show any signal (Fig. 7D).
TAG Synthesis with Acyl-ACP-To determine the functional role of rACP, [ 14 C]oleoyl ACP was purified and used as a substrate in the TAG synthase assay. Different enzyme sources were used in the assay. Among them only R. glutinis cytosol and purified TBC produced triacylglycerol from acyl-ACP as acyl donor, whereas neither R. glutinis membranes nor developing peanut cytosol and its membranes could accept acyl-ACP (Fig. 7E).
MALDI-MS Analysis of Wild Type and Mutant ACP-Mass spectroscopic analysis was carried out to identify the posttranslationally modified serine residue in ACP. Recombinant ACP (wild type and/or S59A mutant) was digested with trypsin before and after acylation, and the protein was analyzed by MALDI mass spectroscopy (Fig. 8). Phosphopantetheinylation increased the mass by 340 Da, and acylation with oleic acid increased the mass of the same fragment further by 282 Da. We analyzed the tryptic fragment that showed a shift of 340 Da and that, after acylation, shifted by the mass of the fatty acyl group. One of the major peaks obtained by in silico analysis (1159 Da, peak C) fitted with these criteria and corresponds to the peptide fragment 50 DVNEVIAEGSK 60 . In wild type protein, peak C is not detectable, since it showed a shift to peak A that has a mass of 1511 (352-Da shift). After acylation, peak A shifted to peak B, showing a mass of 1786 Da. This tryptic fragment has a single serine residue (Ser-59). We did not observe any evidence of phosphopantetheinylation in any other tryptic fragment. The MALDI-MS analysis of the S59A mutant showed the presence of peak C, but peaks A and B were not detectable, confirming that this fragment contains the posttranslationally modified serine residue. DISCUSSION Acyl carrier protein is a part of a 10 S soluble multienzyme complex involved in TAG biosynthesis in R. glutinis. ACP accepts free fatty acid and serves as acyl donor in acyltransferase reactions in TBC. Quantitation of ACP and immunocytochemical studies with antibodies raised against DAG acyltransferase and ACP provided evidence for the association of ACP with other components of TBC. Cloning and characterization of this nonorganellar ACP from oleaginous yeast was achieved. The following observations revealed the presence of a functional ACP in R. glutinis. 1) The gene encoding for ACP was isolated and overexpressed in E. coli. A fraction of expressed protein was post-translationally modified and found capable of accepting free fatty acid. 2) MALDI-MS analysis of recombinant ACP confirmed the presence of post-translational modification. 3) Site-directed mutagenesis of all of the serine residues in recombinant ACP was performed, and it was found that Ser-59 is indeed involved in post-translational modification.
Sequence analysis of the isolated clone for ACP showed that this protein shares a good percentage of homology with ribosomal protein P2. P2 is an acidic ribosomal protein involved in FIG. 8. Identification of the serine residue involved in phosphopantetheinylation in rACP using MALDI-MS. WT and S59A mutant ACP (2 g) were resolved on a 15% SDS-PAGE under reducing conditions followed by Coomassie Brilliant Blue staining. WT acylated ACP (2 g) was resolved on 15% SDS-PAGE under nonreducing conditions followed by staining with Coomassie Brilliant Blue. The stained ACP bands were excised from the gel and subjected to in-gel trypsin digestion. Tryptic fragments of WT nonacylated (A); WT acylated (B), and S59A mutant ACP (C) were analyzed in MALDI mass spectroscopy. Fragment A contains the post-translationally modified peptide in wild type, which actually shows a 340-dalton shift from fragment C. Fragment B represents the acylated peptide obtained from WT protein, which shows a 280-dalton shift in the spectrum. Tryptic digested peptides of S59A mutant showed only fragment C. The tryptic peptide masses obtained experimentally were compared with the in silico data and are represented at the bottom. ribosomal stalk formation and is present as a part of the 60 S ribosomal subunit of all eukaryotic cells (30). There are two pools of P2 proteins, cytosolic and ribosomal. Cytosolic protein is not phosphorylated; upon phosphorylation, the protein gets translocated to ribosome (31). Western blot analysis of R. glutinis ribosome with anti-ACP antibodies showed a strong signal around 14 kDa, suggesting the presence of a cross-reactive protein. Possibly, the cytosolic pool of unphosphorylated P2 protein could behave as an acyl carrier protein in oleaginous yeast. It was observed that another ribosomal protein L7 was found to be copurified with the lipid inclusion bodies obtained from oleaginous bacteria like Rhodococcus opacus PD360 and Rhodococcus ruber. The significance of this association is not clear (32).
Overexpression of rACP in E. coli cells is mildly toxic, resulting in a retarded growth rate. Normally, ACP exists in apo and holo forms. Under the inducible expression system, most of the ACP was present as apo form, which was shown to be a potent inhibitor of glycerol-3-phosphate acyltransferase (33). Phosphopantetheine protein transferase is responsible for the transfer of 4Ј-phosphopantetheine from CoA to the hydroxyl group of serine residue of acyl carrier protein (34). During overproduction of ACP in bacterial system, the enzymatic activity of phosphopantetheine protein transferase becomes limiting to convert all of the apo-ACP to the holo-ACP form. This was evident from the [ 14 C]␤-alanine labeling experiment with the recombinant protein, where a moderately intense signal in fluorography was observed, although the expression level was very high. The recombinant protein was overproduced in E. coli system, and most of the protein was found in inclusion bodies. The soluble fraction, which was used to purify the recombinant protein, contained mostly the holo form of the protein, since the post-translationally modified peptide fragment was found predominantly in MALDI-MS analysis. Metabolic labeling of yeast cells with [ 14 C]␤-alanine could possibly incorporate the label both in CoA and ACP. Hence, it is difficult to draw a correlation between the amount of label added and the amount incorporated into recombinant ACP in these experiments.
Recombinant ACP was readily acylated using E. coli acyl-ACP synthetase, and this acylation was inhibited by pretreatment of the protein with alkylating agents. This confirmed that acylation happened via the thiol group and resulted in thioester linkage. The anomalous mobility of ACP in SDS-PAGE can be explained by considering the high acidic charge on the protein that prevents binding of SDS to the protein. Upon acylation, more SDS binds with ACP and acylated ACP runs faster than the nonacylated form in SDS-PAGE. The difference in mobility was very clear when purified acyl-ACP was resolved in nonreducing SDS-PAGE.
All of the known ACP sequences contain a conserved serine that was post-translationally modified by the attachment of phosphopantetheine to it (14). Surprisingly, we could not locate any conserved serine in our clone upon multiple sequence alignment, but the overexpressed ACP is post-translationally modified and capable of undergoing fatty acylation. There is only one ACP (open reading frame 5 of the Nod locus) reported from Azorhizobium caulinodans, which does not have a conserved serine but is involved in acylation reactions (35). Sitedirected mutagenesis of the recombinant protein and MALDI-MS analysis after tryptic digestion clearly showed that Ser-59 could be the site of post-translational modification in R. glutinis. Sequence analysis showed that the flanking resi-dues of Ser-59 are highly charged ( 57 EGSKKL 62 ). The conserved motif of phosphopantetheinylation in any other ACP is DSLD, where the charged residues flank the serine. The phosphopantetheine protein transferase in E. coli is known to recognize the conserved motif for the transfer of phosphopantetheine (34). However, our expression studies indicated that the expressed ACP was post-translationally modified even when the conserved motif is not present in the protein. The identification of nonorganellar ACP has significant implications for the understanding of TAG biosynthesis and regulation in oleaginous yeast.