The Neisseria meningitidis Serogroup A Capsular Polysaccharide O-3 and O-4 Acetyltransferase*

Neisseria meningitidis serogroup A capsular polysaccharide (CPS) is composed of a homopolymer of O-acetylated, α1→6-linked ManNAc 1-phosphate that is distinct from the capsule structures of the other meningococcal disease-causing serogroups, B, C, Y, and W-135. The serogroup A capsule biosynthetic genetic cassette consists of four open reading frames, mynA–D (sacA–D), that are specific to serogroup A, but the functions of these genes have not been well characterized. mynC was found to encode an inner membrane-associated acetyltransferase that is responsible for the O-acetylation of the CPS of serogroup A. The wild-type CPS as revealed by 1H NMR had 60–70% O-acetylated ManNAc residues that contained acetyl groups at O-3, with some species acetylated at O-4 and at both O-3 and O-4. A non-polar mynC mutant generated by introducing an aphA-3 kanamycin resistance cassette produced CPS with no O-acetylation. A serogroup A capsule-specific monoclonal antibody was shown to recognize the wild-type O-acetylated CPS, but not the CPS of the mynC mutant, which lacked O-acetylation. MynC was C-terminally His-tagged and overexpressed in Escherichia coli to obtain the predicted ∼26-kDa protein. The acetyltransferase activity of purified MynC was demonstrated in vitro using [14C]acetyl-CoA. MynC O-acetylated the O-acetylated CPS of the mynC mutant and further acetylated the wild-type CPS of serogroup A meningococci, but not the CPS of serogroup B or C meningococci. Genetic complementation of the mynC mutant confirmed the function of MynC as the serogroup A CPS O-3 and O-4 acetyltransferase. MynC represents a new subclass of O-acetyltransferases that utilize acetyl-CoA to decorate the d-mannosamine capsule of N. meningitidis serogroup A.

lar polysaccharide (CPS) 1 that confers serogroup specificity is composed of repeating units of ␣136-linked ManNAc 1-phosphate that is O-acetylated (1). Although there is evidence of other glycosidic linkages (2), the principal linkage between monomer ManNAc residues in this polysaccharide is the ␣136 phosphodiester bond involving the hemiacetal group of carbon 1 and the hydroxyl group of carbon 6 of the mannosamine residues. Serogroup A CPS is structurally distinct from the other disease-causing meningococcal serogroups, B, C, Y, and W-135, which are composed of or contain sialic acid (1,3,4).
Meningococcal serogroups C, Y, W-135, and H also express O-acetylated capsules. Interestingly, the serogroup B CPS is not O-acetylated. The genes encoding the putative CPS Oacetyltransferases OatC and OatWY, responsible for the Oacetylation of meningococcal serogroup C and serogroups W-135 and Y, respectively, have been recently identified (5). Other pathogens such as pneumococcal serotype 9V, Salmonella enterica serovar typhi Vi, Staphylococcus aureus serotypes 5 and 8, and Escherichia coli K1 (6) also express Oacetylated capsules. The biological importance of O-acetylation of CPS appears to be species-or subspecies-dependent. In some pathogens, O-acetylation of capsules is involved in immune recognition (6,7). For meningococcal serogroup A CPS, there is a dramatic reduction in the immunogenicity of the polysaccharide observed upon removal of the O-acetyl groups by chemical treatment (8).
The general genetic organization of CPS genes of N. meningitidis is similar to that of other bacterial systems such as Haemophilus influenzae, E. coli K1, etc., that are classified (9,10) as group II capsules. It is usually composed of a unique biosynthetic genetic cassette and conserved genes involved in translocation of the CPS. The genetic cassette responsible for the biosynthesis of the serogroup A capsule is composed of an ϳ5-kb nucleotide sequence located ( Fig. 1) between ctrA, the outer membrane capsule transporter, and galE, the UDP-glucose 4-epimerase (11). Four open reading frames (designated mynA-D or sacA-D) are cotranscribed as an operon (11) and are not found in the genomes of other meningococcal serogroups or in Neisseria gonorrhoeae. Separated from ctrA by a 218-bp intergenic region, mynA is predicted to encode a 372amino acid protein that has homology to the E. coli UDP-Nacetyl-D-glucosamine 2-epimerase. MynB has been hypothesized to be the capsule polymerase, linking individual UDP-ManNAc monomers together, whereas MynD has been predicted to be involved either in CPS transport assembly or in cross-linking of the capsule to the meningococcal cell surface (11). In this study, we demonstrate that mynC (744 bp) encodes an O-acetyltransferase (247 amino acids) that transfers acetyl groups to the ManNAc residues of the serogroup A CPS.

EXPERIMENTAL PROCEDURES
Materials and Bacterial Strains-The bacterial strains, plasmids, and primers used in this study are described in Table I. The meningococcal serogroup A strains were originally isolated during an outbreak in Nairobi, Kenya, in 1989 (12) and were provided by the Centers for Disease Control and Prevention (Atlanta, GA). Strain F8229 (CDC1750) is encapsulated and was isolated from the cerebrospinal fluid of a patient with meningitis. Strain F8239 (CDC16N3) is an unencapsulated variant originally isolated as a serogroup A strain from the pharynx of an asymptomatic carrier. These strains belong to clonal group III-I and are closely related to strains that have caused epidemics in Saudi Arabia, Chad, Ethiopia, and other parts of the world. Monoclonal antibody (mAb) 14-1-A (13) against the meningococcal serogroup A CPS was generously provided by Dr. Wendell Zollinger (Walter Reed Army Institute of Research). Restriction enzymes were purchased from New England Biolabs Inc. (Beverly, MA). Nickel-nitrilotriacetic acid (Ni-NTA)-agarose gravity flow matrix and anti-pentahistidine monoclonal antibodies were purchased from QIAGEN Inc. (Valencia, CA). The B-PER 6xHis fusion protein purification kit was purchased from Pierce. Nucleotide primers were synthesized at MWG Biotech (High point, NC). [ 14 C]Acetyl-CoA and 4-nitrophenyl acetate were purchased from Sigma. Automated DNA sequence analysis was performed with the Prism dye deoxy terminator cycle sequencing kit (Applied Biosystems, Foster City, CA), and completed reactions were run on an ABI Model 377 automated DNA sequencer at the Microchemical Facility of Emory University.
Growth Conditions-Meningococcal strains were grown with 3.5% CO 2 at 37°C on gonococcal base agar (Difco) supplemented with 0.4% glucose and 0.68 mM Fe(NO 3 ) 3 or in gonococcal broth containing the same supplements and 0.043% NaHCO 3 . Brain/heart infusion medium (37 g/liter) with 1.25% fetal bovine serum was used when kanamycin selection was required. The antibiotic concentrations used for E. coli strains were 100 g/ml ampicillin, 50 g/ml kanamycin, and 300 g/ml erythromycin, and those for N. meningitidis were 80 g/ml kanamycin, 60 g/ml spectinomycin, and 3 g/ml erythromycin. E. coli strain DH5␣ cultured on LB medium was used for cloning and propagation of plasmids. Meningococci were transformed by the procedure of Janik et al. (14). E. coli strains were transformed with a Gene-Pulser (Bio-Rad) according to the manufacturer's protocol.
Construction of Meningococcal Non-polar mynC Mutant NmA001-An internal 745-bp fragment of mynC, produced by PCR amplification using primers SE57 and SE61 (11) and the chromosomal DNA of strain F8229 as a template, was cloned into pCR2.1 to yield pGS201. The aphA-3 fragment obtained from pUC18K (15) with EcoRI and HincII digestion and filled in with Klenow polymerase was inserted into the unique SspI site of mynC in pGS201 to generate pGS202. The correct orientation of aphA-3 was confirmed by colony PCR and direct sequence analysis of pGS202. An ScaI-linearized pGS202 plasmid was used to transform meningococcal serogroup A strain F8229 to generate FIG. 1. Genetic organization and location of the N. meningitidis serogroup A capsule biosynthetic locus mynA-D (sacA-D) and the sites of polar () and non-polar (ࡗ) mutations in these genes. ctrA is the first gene of the capsule transport operon, and galE encodes the UDP-glucose 4-epimerase.  This study  Primers  SE56  AATCATTTCAATATCTTCACAGCC  11  SE57  TTACCTGAATTTGAGTTGAATGGC  11  SE61  CAAAGGAAGTTACTGTTGTCTGC  11  YT79  CATCATAACGGTTCTGGCAAATATTC  60  YT80  CTGTATCAGGCTGAAAATCTTCTCTC  60  SG005  GAACATATGTTATCTAATTTAAAAAAC  This study  SG006  TTACTCGAGATATATATTTTGGATTATGGT  This study  SG007  GGAGATATACATAAGCTTTCTAATTTAAAA  This study  SG008  AGCGAATTCTCAGTGGTGGTGGTGGTGGTG  This study NmA001. The correct homologous recombination of the aphA-3 cassette into the mynC coding sequence was confirmed by PCR with cassettespecific primers and chromosome-specific primers.
Overexpression and Purification of Meningococcal MynC-The complete coding sequence of mynC was obtained by PCR amplification using primers SG005 (NdeI) and SG006 (XhoI) ( Table I). The PCR product, digested with NdeI and XhoI, was subsequently cloned into pET20b(ϩ) cut with the same enzymes to yield pGS203, resulting in a C-terminal His 6 fusion. Plasmid pGS203 was purified and subjected to DNA sequence analysis to confirm the intact mynC sequence and the C-terminal His tag fusion. pGS203 was then transformed into E. coli expression strain BL21(DE3) pLysS. One liter of LB culture of the MynC-overexpressing strain was induced with 1 mM isopropyl-␤-Dthiogalactopyranoside (IPTG) for 5 h. The harvested cells were resuspended in 15 ml of lysis buffer (50 mM sodium phosphate (pH 8.0), 300 mM NaCl, 10 mM imidazole, 1% (v/v) Tween 20, 1 mM phenylmethylsulfonyl fluoride, and 1 mg/ml lysozyme), left on ice for 30 min, and sonicated 10 times for 30 s with 30-s cooling intervals. The cell debris was removed by centrifugation at 14,000 ϫ g for 15 min at 4°C. The overexpressed protein was purified under native conditions on Ni-NTA matrices following the supplier's protocol with a modification in column washing. Briefly, the crude extract was incubated with 2 ml of 50% suspension of Ni-NTA-agarose for 1 h before packing into a column. The column was washed with 5 ml each of 10, 20, and 40 mM imidazole in lysis buffer without lysozyme and then eluted with 5 ml of buffer containing 250 mM imidazole. The MynC protein was also extracted and purified using a B-PER protein extraction kit (Pierce), containing a lysis reagent with a proprietary mild nonionic detergent in 20 mM Tris-HCl (pH 7.5), following the manufacturer's instructions. The purified MynC fractions obtained by either method were concentrated separately using Centricon YM-3 centrifugal filters (Millipore Corp., Bedford, MA) after SDS-PAGE analysis and dialyzed in storage buffer (50 mM HEPES (pH 7.05), 5 mM MgCl 2 , 100 mM NaCl, and 1 mM EDTA). The protein concentration was determined with a BCA protein assay kit (Pierce) using bovine serum albumin as the standard.
Complementation of the NmA001 Mutant-An intact copy of mynC under the control of the tac promoter was constructed on a meningococcal shuttle vector. Full-length mynC with a C-terminal His tag was amplified from pGS203 using primers SG007 (HindIII) and SG008 (EcoRI) ( Table I). The amplified PCR product was cloned into pCR2.1 to yield pGS204. The mynC insert was subsequently released from pGS204 with HindIII and EcoRV digestion and ligated into the HindIII and SmaI sites of pFLAG-CTC to generate pGS205 with mynC under the control of the lac promoter. The construct was confirmed by PCR using vector-specific primers YT79 and YT80. The pGS205 plasmid was then cut with BglI, filled in with Klenow, and ligated into the EcoRV site of the meningococcal shuttle vector pYT250(Erm R ), yielding pGS206. The pGS206 construct was methylated with HaeIII methylase, and the reaction mixture was used directly to transform wild-type strain F8229 and non-polar mynC mutant NmA001, yielding NmAwtc1 and NmAnpc1, respectively.
Meningococcal Membrane and Cytosolic Preparations-Meningococcal membranes and cytosol were separated by the method of Clark et al. (16) from the complemented meningococcal mynC strain NmAnpc1. Briefly, the pellet from a 500-ml culture of NmAnpc1 carrying pGS206, induced overnight with 1 mM IPTG, was used to prepare the inner and outer membranes and cytosolic fractions. The pellet was suspended in 2 ml of buffer containing 1 mM EDTA, 50 mM Tris, and 20% sucrose (pH 8.0) with 1 mg/ml lysozyme for 30 min at 4°C. The cell suspensions were diluted with 20 ml of Tris buffer to generate spheroplasts and were sonicated three times each for 30 s in an ice bath with 30-s resting intervals. The cell debris was removed by centrifugation at 10,000 ϫ g for 15 min at 4°C. The supernatant was freeze-thawed once at Ϫ70°C before ultracentrifugation at 100,000 ϫ g for 90 min at 4°C. The pellet, containing the meningococcal membrane fraction, was washed with Tris buffer. The level of contamination of the membrane fraction with cytoplasmic components was assessed by determining the activity of a cytoplasmic enzyme, malate dehydrogenase (17), for both fractions. The membrane fractions were 97-98% pure. The cytosolic proteins were precipitated by 5% trichloroacetic acid and suspended in 2 ml of 1 M Tris (pH 6.8). Total membrane was solubilized with 2 ml of 2% N-lauroylsarcosine (Sarkosyl) in 10 mM HEPES (pH 7.4) for 1 h at room temperature using an orbital shaker. Soluble inner membrane components and insoluble outer membrane components were separated by ultracentrifugation at 100,000 ϫ g for 2 h at 4°C. The outer membrane pellet was suspended in 500 l of 1 M Tris (pH 6.8). The diluted inner membrane proteins were precipitated by 5% trichloroacetic acid, and the pellet thus obtained was suspended in 500 l of 1 M Tris (pH 6.8). Subcellular fractions were resolved on 10% SDS-polyacrylamide gels. The loadings were standardized based on the same amount of starting meningococci (500 ml of meningococcal culture pellet at A 550 ϭ 1 contains ϳ2.5 ϫ 10 11 cells) and analyzed by Western blotting. To examine the nature of the membrane association of MynC, membrane solubilization experiments were performed as described (18). Briefly, the membrane pellets were extracted with 5 ml of buffer (100 mM sodium phosphate (pH 7.6), 0.2 mM dithiothreitol, 20% sucrose, and 0.2 M KCl) containing 1% Triton X-100, 1 M NaCl, or 6 M urea for 30 min at room temperature (urea), at 30°C (Triton X-100), or on ice (buffer alone and buffer with NaCl). Samples were centrifuged at 130,000 ϫ g for 1 h at 4°C after the extraction. Proteins in the soluble fractions were precipitated with 5% trichloroacetic acid, and the precipitates obtained were washed two times with acetone, dried, and resuspended in 1 M Tris (pH 6.8) before an equal volume of 2ϫ SDS-PAGE sample buffer was added.
CPS Extraction and Structural Characterization-CPS was extracted from 2 liters of meningococcal cultures using the standard method of Gotschlich et al. (19). Briefly, the overnight cultures were treated with a final concentration of 1% Cetavlon, a polycationic detergent that precipitates the polyanionic polysaccharides. The precipitate was collected by centrifugation and resuspended in water, and CaCl 2 was then added to a final concentration of 1 mM to separate the polysaccharide from the detergent. Nucleic acids were precipitated from the solution by adding 25% (v/v) ethanol, followed by centrifugation. CPS in the supernatant was subsequently precipitated by ethanol at a final concentration of 80% (v/v). Contaminating protein, traces of Cetavlon, and other low molecular mass contaminants were removed with proteinase K digestion and extensive dialysis against a buffer composed of 10% ethanol, 50 mM NaCl, and 5 mM Tris. CPS was further purified by Sephacryl 200 gel filtration column using 50 mM ammonium formate elutions. Column fractions were tested for neutral sugar estimation by phenol sulfuric acid assay (20). Void volume fractions were pooled and concentrated by speed vacuuming and analyzed by deoxycholate-PAGE and Alcian blue staining (21).
Compositional and NMR Analyses of CPSs-Compositional analysis of purified CPS was performed on the alditol acetate derivatives of the sugars after removing the phosphate groups by hydrogen fluoride treatment of the purified serogroup A CPS. The alditol acetate derivatives were analyzed by combined gas chromatography/mass spectrometry using a 30-m SP2330 capillary column (Supelco) (22).
Lyophilized wild-type or mutant CPS powder (5 mg) was dissolved in D 2 O (99.999 atom % D; Sigma) to a uniform concentration of 5 mg/ml. Solutions were agitated with vortexing for 10 min at room temperature, followed by low speed centrifugation at 7200 ϫ g for 10 min to eliminate undissolved material. Aliquots (600 l) of the supernatant were transferred to 5-mm NMR tubes and placed in a sonication bath for 10 min to eliminate air bubbles trapped on the inner wall of the NMR tube.
NMR spectra were acquired on a Varian Unity 500 NMR spectrometer equipped with a 5-mm pulsed field gradient triple resonance probe and a high precision temperature controller (ϩ0.1°C) and under the control of VNMR Version 6.1B or on a Varian Inova 500 spectrometer equipped with a 5-mm pulsed field gradient inverse detection heteronuclear probe running under VNMR Version 6.1C and Solaris Version 2.8. One-dimensional proton NMR spectra were collected at 25°C using a standard one-pulse experiment. The transmitter was set at the HDO frequency (4.78 ppm). Standard spectral acquisition conditions are as follows: collection of 64K data points over a spectral window of 8000 Hz, acquisition time of 4.096 s, and a relaxation delay of 26 s, giving a recycle time of 30 s. Typically, 64 scans were averaged. Spectra were Fourier-transformed after applying a 0.2-Hz line broadening function. Integrations were performed using subroutines built into the VNMR software.
Hydrophobic Interaction Chromatography-The cell-surface hydrophobicity of meningococcal strains was tested using a modified method of Karlyshev et al. (23). Disposable plastic columns packed with octyl-Sepharose CL-4B (Sigma) to a height of 2 cm were washed with 10 ml of buffer A (0.2 M ammonium sulfate in 10 mM sodium phosphate buffer (pH 6.8)). Meningococci collected from overnight plate cultures were suspended in phosphate-buffered saline to an absorbance of 10, and a 100-l aliquot was gently pipetted onto the surface of the column and eluted with 5 ml of buffer A. A 100-l cell suspension diluted directly into 5 ml of buffer A was also prepared as a control. The A 600 of both the column flow-through and control samples was determined. Results were calculated as the A 600 of the flow-through divided by that of the control and are expressed as a percentage of cells adsorbed to the column.
Serum Bactericidal Assay-A serum bactericidal assay was performed as described previously (24) using pooled normal human serum at a final concentration of 10% (v/v) with a 30-min incubation at 37°C. Heat-inactivated normal human serum was used as a control.
Immunoblots-Serogroup A wild-type and mynC mutant NmA001 CPSs were resolved on 15% deoxycholate-polyacrylamide gels and transferred onto polyvinylidene difluoride membrane using a transfer buffer of 25 mM Tris, 192 mM glycine (pH 8.3), and 20% methanol. An identical gel was stained with Alcian blue to visualize capsules. Membranes were blocked with 3% bovine serum albumin in Tris/Tween buffer (0.5 M Tris (pH 7.5), 0.9% NaCl, and 0.05% Tween 20). Serogroup A capsule-specific mAb 14-1-A (13) was used as the primary antibody at 1:1000 dilution, whereas alkaline phosphatase-conjugated goat antimouse IgG ϩ IgM (Organon Teknika Corp., West Chester, PA) was used at 1:5000 dilution. All incubations were done at room temperature for 1 h. Blots were developed in 20 ml of alkaline phosphatase buffer (0.1 M Tris (pH 9.5), 0.1 M NaCl, and 0.5 mM MgCl 2 ) containing 40 l of 10% nitro blue tetrazolium in 70% N,N-dimethylformamide and 30 l of 5-bromo-4-chloro-3-indolyl phosphate (50 mg/ml in N,N-dimethylformamide). Colony immunoblots were processed similarly using nitrocellulose membranes. After the meningococci were lifted, the membranes were allowed to air-dry for 30 min at room temperature and then blocked for 1 h with 5% bovine serum albumin in Tris/Tween buffer. Protein samples for Western blots were resolved by a 10% SDS-polyacrylamide gel and transferred to polyvinylidene difluoride membranes as described above. Anti-pentahistidine monoclonal antibodies were used as primary antibodies at 1:1000 dilutions.
Whole Cell ELISA-ELISAs were performed using the methods reported previously (11) with the following modifications. Aliquots (50 l) at 1:9 dilution of meningococcal suspensions (A 550 ϭ 0.1) were applied to microtiter plates and dried overnight at 37°C. mAb 14-1-A was used at 1:30,000 dilution, and alkaline phosphatase-conjugated goat antimouse secondary antibody (Organon Teknika Corp.) was used at 1:10,000 dilution. All incubations were performed at 37°C.
Colorimetric Estimation of Capsule O-Acetylation-O-Acetylation of purified CPSs was also measured colorimetrically according to the method described by Hestrin (25). Aliquots of CPS samples (500 l) were incubated with an equal volume of 0.035 M hydroxylamine in 0.75 M NaOH for 10 min at 25°C, followed by the addition of 1 M of perchloric acid (500 l) and 70 mM ferric perchlorate in 0.5 M perchloric acid (500 l). The pink color resulting from the presence of O-acetyl groups was quantified at 500 nm with a known amount of ethyl acetate as the standard.
Autoradiography-O-Acetyltransferase enzyme activity was determined by autoradiography using [ 14 C]acetyl-CoA as acetyl donor and purified meningococcal CPSs as substrates. In a typical 50-l reaction volume, 50 g of CPS, 10 g of MynC protein, and 0.5 Ci of [ 14 C]acetyl-CoA (0.05 Ci/l, specific activity of 47 Ci/mol) were incubated in buffer composed of 10 mM Tris (pH 7.4), 20 mM NaCl, 1 mM MgCl 2 , and 25 mM EDTA. The reaction mixtures were concentrated to near dryness after a 1-h incubation at 37°C and then resuspended in 10 l of water and 10 l of 2ϫ sample buffer. The samples were resolved on 15% deoxycholate-polyacrylamide gels. Gels were incubated with intensifying solution (DuPont) for 30 min before drying under vacuum. The dried gels were exposed to x-ray films at Ϫ80°C.

RESULTS
Homology of MynC-A BLAST search performed with the deduced MynC amino acid sequence (247 amino acids) identified five proteins in the GenBank TM /EBI Data Bank with Ն25% sequence identity (Table II). Among these were EpsK of Lactococcus lactis spp. cremoris; acetylesterase/xylosidase (EC 3.1.1.6; 266 amino acids) XynC of Caldicellulosiruptor saccharolyticus (27); and a CPS synthesis protein, Cap8I (464 amino acids), from S. aureus spp. aureus MW2 (28). Interestingly, these proteins shared with MynC a semiconserved motif (GSSKGG) of mostly hydrophobic small amino acids in the N-terminal region. Repeated search and pairwise comparison of known O-acetyltransferases from prokaryotes and eukaryotes revealed no significant homology to MynC.
Motif search of the MynC sequence at the ISREC (Swiss Institute for Experimental Cancer Research) and SIB (Swiss Institute for Bioinformatics) sites revealed no matches in the data banks. Search results using the SIB-PROSITE Database of protein families and domains showed no similarity to MynC. However, EMBL-EBI InterProScan predicted MynC as a member of the ␣/␤-hydrolase superfamily, which includes acetylcholinesterases, carboxylesterases, mycobacterial antigens, and acetylesterases.
Glycosyl Compositional Analysis of the Wild-type and mynC Mutant CPSs-The growth of the non-polar mynC mutant on gonococcal medium was not different compared with the growth of the wild-type parent. However, when the pellets from 1-liter cultures of similar growth (A 600 ϭ 1.0) were compared for CPS yields, the mynC mutant consistently yielded 25-30% less CPS compared with the wild-type parent. CPSs from wildtype strain F8229 and the non-polar mynC mutant NMA001 were then prepared, purified, and subjected to compositional and structural analyses. The gas-liquid chromatography/mass spectrometric analysis of the alditol acetate derivatives, after removal of the phosphate groups by hydrogen fluoride treatment, revealed ManNAc as the sole component of CPSs isolated from both the wild-type strain and the mynC mutant.
O-Acetylation by NMR Spectral and Colorimetric Analyses of NmA CPSs-To investigate the extent of O-acetylation and the location of the O-acetyl groups, the CPSs were subjected to oneand two-dimensional 1 H NMR spectroscopic analyses. Assignments of the various protons could be made from the COSY and total correlation spectroscopy NMR analyses. The wild-type 3-O-acetylated CPS proton assignments (Table III) were com- pared with published values (29,30) and were highly consistent with these values. However, the mynC mutant CPS spectrum was quite distinct.
In the wild-type CPS 1 H NMR spectrum shown in Fig. 2A , the H-3 proton of ManNAc was observed at 5.20 ppm when the moiety was acetylated at O-3 due to the deshielding effect of the acetyl group. The absence of this peak in the spectrum of the mutant CPS (Fig. 2B) indicated the lack of acetylation at O-3 on the ManNAc residue. The H-2 resonance at 4.61 ppm was observed in the wild-type CPS, indicating 3-O-acetylation, whereas in the mynC mutant spectrum, this peak was missing (Fig. 2, compare A and B). In the region between 2.05 and 2.10 ppm where N-and O-acetyl methyl protons were observed ( Fig.  2A (inset) and Table III), three peaks were identified in the wild-type CPS spectrum. Two of these peaks corresponded to O-acetyl methyl protons, whereas the other was due to N-acetyl methyl protons. However, in the spectrum (Fig. 2B (inset) and Table III) of the mynC mutant CPS, only one peak corresponding to the N-acetyl methyl proton resonance at 2.08 ppm was observed, suggesting the absence of O-acetylation. These differences in one-dimensional NMR spectra indicate the absence of O-acetylation in the mynC mutant CPS.

FIG. 2. 1 H NMR spectra of CPSs purified from the meningococcal serogroup A wild-type strain F8229 (A), the mynC::aphA-3 NmA001 mutant (B), and the mynC::aphA-3 mutant complemented with pGS205 (NmAnpc1) upon IPTG induction (C).
Insets are enlargements of the N/O-acetyl (NAc/OAc) methyl proton regions. The relative percentages of the CPS populations (Table IV)  Acetyl-CoA:Poly-ManNAc O-Acetyltransferase Activity of MynC-To confirm the acetyltransferase activity, MynC was C-terminally His-tagged, overexpressed in E. coli, and purified under native conditions by Ni-NTA affinity chromatography (Fig. 3) after extraction by the conventional lysozyme lysis method in the presence of 1% Tween 20 or by the B-PER protein extraction procedure (see "Experimental Procedures"). MynC purified by both methods were active and soluble and, when used separately, gave similar results.
Purified MynC was used in in vitro assays containing the serogroup A wild-type or mynC mutant CPS as the substrate and [1-14 C]acetyl-CoA as the acetyl donor. Autoradiography of the CPSs (Fig. 4) revealed that MynC transferred the [ 14 C]acetyl group from acetyl-CoA to the non-acetylated CPS of the mynC mutant (Fig. 4, lanes 5-8). Interestingly, MynC was also capable of further O-acetylating the wild-type CPS (Fig. 4,  lane 4). MynC recognized the serogroup A CPS, but not serogroup B or C CPS (Fig. 4, lanes 2 and 3). Finally, the acetyltransferase activity was not due to the minor contaminating E. coli proteins in the MynC preparations, as the lysate of the vector construct alone did not exhibit activity (Fig. 4, lanes 9  and 10).
The observed O-acetylation activity was dependent on the enzymatic function of MynC, as the reactions showed a concentrationdependent increase in the 14 C incorporation over a range of 1-6 g of MynC protein in a 25-l reaction volume (Fig. 5A). In addition, MynC demonstrated linear time kinetics, with the maximal activity reached at 2 h under the experimental conditions, and did not decrease its activity until 3 h (Fig. 5B). MynC lost its enzyme activity at pH 4.5, and the activity also decreased at pH Ͼ9.5. The optimal pH for the MynC activity was between 5.8 and 7.0 (Fig. 5C). Mg 2ϩ and Na ϩ ions, present in the in vitro Oacetyltransferase reaction buffer, were not essential for the enzyme activity, as the citrate, phosphate, and borate buffers (without Mg 2ϩ and Na ϩ ions) gave similar results (data not shown).
Complementation of the Meningococcal mynC Mutant-An intact copy of mynC under the control of the lac promoter was constructed and subcloned into the meningococcal shuttle vector pYT250 as described under "Experimental Procedures." The plasmid was transformed into the mynC mutant and the wildtype strain to generate strains NmAnpc1 and NmAwtc1, respectively. The wild-type strain, the wild-type strain overexpressing MynC (NmAwtc1), the mynC mutant, and the complemented strain (NmAnpc1) were grown on gonococcal agar plates with or without IPTG and analyzed by colony immunoblotting (data not shown) and ELISAs (Fig. 6A) using the serogroup A capsule-specific mAb 14-1-A. The unencapsulated strain F8239, a serogroup A strain that contains point mutations and deletions in mynA (11), was used as a negative control. The wild-type and NmAwtc1 meningococci were   strongly recognized by mAb 14-1-A, whereas the uncomplemented non-polar mynC mutant and the capsule-negative control strain F8239 did not react with the antibody. The complemented non-polar mynC mutant strain NmAnpc1 reacted strongly, and the intensity was increased with IPTG induction. These data indicate that an O-acetyl group is a component of the epitope specificity of mAb 14-1-A. The CPS isolated from the complemented strain NmAnpc1 was subjected to 1 H NMR analyses, which revealed the restoration of O-acetylation in the polymer (Fig. 2C). Compared with the relative integration values of H-2 resonances of 3-O-acetylated and non-O-acetylated forms, the level of O-acetylation in the complemented strain, even with IPTG, was less than the wild-type levels, although the relative ratio (5:3) of the acetylated species O-3:O-4 was similar to the wild-type ratio (4:2.7) (Table IV). In addition, the mynC mutant CPS O-acetylated in vitro by MynC was now recognized by mAb 14-1-A (data not shown), again confirming the importance of O-acetylation in defining the epitope recognized by this antibody.
In summary, quantitative ELISA, 1 H NMR, and colorimetric assays with the CPS from the complemented strain NmAnpc1 revealed that O-acetylation was restored by genetic complementation. Western blot analysis (Fig. 6B) of the whole cell lysates from the MynC-overexpressing wild-type strain (NmAwtc1) and the complemented mynC mutant (NmAnpc1) was also performed using the anti-pentahistidine mAb. Histagged MynC (Fig. 6B, lanes 4 and 6) was visualized in the complemented meningococci upon IPTG induction, a result correlating with the restoration of O-acetylation in the mutant NmA001 CPS.
MynC Localization-According to a Markov model for transmembrane domain prediction, TMHMM (Centre for Biological Sequence Analysis, Technical University of Denmark, Lyngby, Denmark), no transmembrane domains were detected in MynC at high stringency. The complemented mynC strain NmAnpc1 was used to assess the cellular location of MynC in serogroup A meningococci because the plasmid-encoded copy of mynC contains a His tag and thus enables detection of the MynC protein.
Western blot analysis of the subcellular fractions, which were loaded on the basis of the same cell numbers of cultured meningococci, revealed that MynC was associated with the inner membrane (Fig. 7A). The total membrane and inner membrane components gave strong reactivity, whereas the cytosolic fraction showed weak reaction, and the outer membrane prepara- tion showed no reaction. Similar results were obtained when the gels were loaded with the equivalent protein concentrations of the subcellular fractions (data not shown). To explore the possibility that MynC is a peripheral or transmembrane protein, various extraction procedures were performed. Treatment of total membranes with 6 M urea partially removed the protein, whereas treatment with 1 M NaCl or 1% Triton X-100 (Fig. 7B) did not extract MynC from the membranes, indicating that MynC is a peripheral protein tightly associated with the inner membrane. O-Acetylation of bacterial surface polysaccharides such as CPSs, exopolysaccharides, peptidoglycans, and lipo-oligosaccharides is common in pathogens and in symbionts and has been shown to have immunogenic and functional importance. N. meningitidis, E. coli K1, Streptococcus pneumoniae, S. enterica, S. aureus, and Pseudomonas aeruginosa can express O-acetylated CPS (32,33). In S. enterica serovar typhi (7) and E. coli K1 (6), the loss of O-acetylation from CPS results in loss of immunogenicity, whereas in meningococcal serogroup C (31) and pneumococcal serotype 9V (34), O-acetylation of capsules is not required for induction of protective antibodies. In the extracellular alginate polysaccharide polymers, produced by isolates of P. aeruginosa from patients with cystic fibrosis, Dmannuronic acid is O-acetylated at O-2 and O-3 by three genes, algI, algJ, and algF (35). Alginate O-acetylation had been shown to contribute to biofilm architecture and microcolony formation (36) and resistance to opsonic phagocytosis (37). O-Acetylation is also important for Rhizobium-legume symbiosis. The rhizobial Nod factors may be O-acetylated at distinct sites to define the host specificity and the formation of the preinfection thread and root nodule (38 -40). In Proteus mirabilis, N. gonorrhoeae, and N. meningitidis (41), the C-6 hydroxyl of N-acetylmuramyl residues in peptidoglycans are O-acetylated to confer both intrinsic and complete resistance to lysozyme hydrolysis. These peptidoglycan motifs are pathogen-associated molecular patterns recognized by the innate immune system (42,43).
A number of acetyltransferases that transfer an acetyl group from acetyl-CoA to O-acetylate dissimilar substrates have been identified in prokaryotic and eukaryotic systems, but these proteins share limited sequence homology. Two families of proteins that O-acetylate exported carbohydrate moieties have been reported. Members of the NodL-LacA-CysE family (44 -48), which includes the lipochitin acetyltransferase (NodL) of Rhizobium leguminosarum, galactoside acetyltransferases such as LacA, and the serine acetyltransferase (CysE) of E. coli, are cytoplasmic proteins that use acetyl-CoA as acetyl donor. Interestingly, the proposed sialic acid O-acetyltransferases of meningococcal serogroups W-135 and Y (OatWY), but not of serogroup C (OatC), show sequence homology to the NodL-LacA-CysE family (5). The second family comprises integral membrane proteins. Members of this family include MdmB of Streptomyces mycarofaciens, which O-acetylates macrolide antibiotics such as midecamycin (49) (54), but the respective gene and the protein have not been identified. MynC does show sequence homology to several proteins (Table II), including the acetylesterase (acetylxylosidase) that degrades xylan from the thermophile C. saccharolyticus and a hypothetical esterase/lipase/thioesterase family protein of Arabidopsis thaliana. These proteins share with MynC a semiconserved motif (GSSKGG) in the N-terminal region. Typically, serine esterases contain an N-terminal conserved GSSSG motif (assumed to be the catalytic motif), where the center serine residue (underlined) is the active-site nucleophile (55). However, purified MynC did not show acetylesterase activity when tested with the synthetic substrate p-nitrophenyl acetate used in other in vitro esterase assays (data not shown) (56).
O-Acetylation is critical for N. meningitidis serogroup A CPS immunogenicity and antibody formation (8). The major protective epitope recognized by antibodies induced following vaccination with serogroup A polysaccharide requires O-acetylation. Berry et al. (8) found that bactericidal anti-serogroup A antibodies in the sera of serogroup A polysaccharide-vaccinated individuals are specific for O-acetylated CPS. The importance of Oacetylation in serogroup A capsule immunogenicity was confirmed using O-acetylated and non-O-acetylated polysaccharides and polysaccharide-protein conjugates in immunogenicity studies with mice (8). However, in this study, we found that the serogroup A capsule with or without O-acetylation protected the meningococcus against killing by a low concentration (10%) of normal human serum, i.e. antibody-independent complementmediated killing. Both the wild-type parent and non-polar mynC mutant were protected from killing with 100% survival, in contrast to the unencapsulated strain F8239 and the mynA, mynB, and mynD mutants, which were completely killed under these conditions (data not shown). O-Acetylation may also have a role in the initial stages of colonization and infection by N. meningitidis serogroup A. In preliminary studies of meningococcal colonization in a mouse model (57), mynC mutants expressing non-O-acetylated CPS showed significantly reduced ability to establish colonization compared with the wild-type parent strain. 2 MynC is specific for meningococcal serogroup A (␣136linked ManNAc 1-phosphate) CPS. MynC did not acetylate the sialic acid CPSs of N. meningitidis serogroup B or C. The in vitro O-acetylation studies indicate that MynC recognized the non-O-acetylated and the partially O-acetylated CPS assembled polymers as substrates. Therefore, O-acetylation appears to be a near final step of decorating the serogroup A capsular polymers. Cell-surface hydrophobicity, a marker of capsule expression, measured by hydrophobic interaction column chromatography (23), showed that the mynC mutant cell surface was hydrophilic (Ͻ5% cells retained on the column) and was similar to the wild-type parent (Ͻ4% cells retained on the column). In contrast, the unencapsulated variant F8239 (Ͼ60% cells retained on the column) and the mynA and mynB mutants (Ͼ90% cells retained on the column) demonstrated high cellsurface hydrophobicity. The hydrophilic cell surface and the resistance to killing by normal human sera of the mynC mutant indicate that the O-acetylated capsular polymer in the nonpolar mynC mutant is surface-expressed and functional. Thus, the serogroup A capsule expression and transport or the prevention of killing by normal human sera does not require Oacetylation. The modest reduction in synthesis of the CPS of the mynC mutant, while not affecting biological function, might suggest some polarity of the insertion or disruption of transcript stability. Alternatively, O-acetylation might influence the level of CPS expression in meningococci.
In summary, MynC is the CPS O-3 and O-4 acetyltransferase of serogroup A of N. meningitidis. This ϳ25-kDa inner membrane-associated enzyme utilizes acetyl-CoA as the acetyl group donor and belongs to a new subclass of Oacetyltransferases. Study of the O-acetylation-deficient mutant confirms the importance of O-acetylation in serogroup A polysaccharide immunogenicity, but O-acetylation is not required for capsule expression or to protect the meningococci from killing by normal human sera. O-Acetylation by MynC may be important for vaccine development against N. meningitidis serogroup A. The enzymatic function of MynC may be used to modulate the degree of O-acetylation of serogroup A polysaccharides used for new and existing meningococcal conjugate and polysaccharide vaccines.