ABSTRACT
We report on the identification of a surface-exposed, highly conserved, immunogenic nontypeable Haemophilus influenzae(NTHi) protein, which elicits cross-reactive bactericidal antibodies against NTHi. The protein was extracted from NTHi strain P860295 with KSCN and purified; it migrated as a single band on a sodium dodecyl sulfate-polyacrylamide gel with an apparent molecular mass of 63 kDa. Mouse antiserum generated against the purified protein was reactive on whole-cell enzyme-linked immunosorbent assay (ELISA) with seven NTHi strains and type b Eagan and Whittier strains and exhibited bactericidal activity to homologous and heterologous NTHi strains. However, the protein is made in small amounts in NTHi as corroborated by immunoelectron microscopy. To further study this protein, we cloned, sequenced, and expressed it recombinantly in Escherichia coli. The recombinant protein is localized in the periplasm ofE. coli and has been purified to homogeneity. Both the recombinant and native proteins possess 5′-nucleotidase activity; hence, the protein has been called NucA. Mouse antiserum directed against the recombinant NucA protein was reactive on Western immunoblots and whole-cell ELISA with all H. influenzaestrains tested including Eagan and was bactericidal for two heterologous strains tested. The antiserum also resulted in a log reduction in bacteremia, in an infant-rat protection study withH. influenzae type b as the challenge strain. These features suggest that NucA is a potential subunit vaccine candidate against NTHi disease.
Nontypeable Haemophilus influenzae (NTHi) isolates are causative agents of a wide range of human diseases including tracheobronchitis, pneumonia, epiglottitis, and otitis media (7, 28, 47). Most vaccine development effort has been directed toward integral outer membrane components of NTHi. Outer membrane proteins (OMPs) P1(30, 31), P2 (18, 22, 32, 33, 46), P4 (18), and P6 (13, 19, 20, 29, 34, 36) have been extensively studied, cloned, sequenced, and recombinantly expressed. Antiserum directed against each of these OMPs is both bactericidal and protective against the homologous NTHi strains in the infant-rat model of H. influenzae type b (Hib) meningitis. However, only P4 and P6 are antigenically conserved among clinical isolates. While the presence of serum bactericidal antibody against a particular NTHi strain has been correlated with protection against this strain in children (15), the bactericidal response to natural infection with NTHi appears to be strain specific. Therefore, most laboratories have continued to search for protein targets which elicit bactericidal antibodies that are broadly cross-reactive among NTHi clinical isolates.
Additional, nonintegral, outer membrane-associated proteins have been investigated for vaccine potential against NTHi. Using the chinchilla model of otitis media, Brinton et al. (8) have reported that active immunization with LKP pili or passive immunization with anti-LKP serum could protect chinchillas against otitis media caused by NTHi organisms expressing homologous pili. Barenkamp more recently reported that immunization with high-molecular-weight adhesin proteins could partially protect against and modify NTHi-caused otitis media in chinchillas (4). These results show that an immune response to surface adhesin molecules can also influence NTHi otitis media in animal models.
At least two publications have shown that a secretory immunoglobulin A (IgA) response to surface structures of NTHi can influence colonization in humans (23, 43), and it is possible that auxiliary vaccine components may be beneficial in addition to OMPs. Therefore, our laboratory has been investigating additional surface-exposed, nonintegral membrane proteins for vaccine potential. Experiments were initiated to determine if surface-exposed, nonintegral proteins which were cross-reactive among NTHi strains could be identified to supplement integral OMPs (e.g., P4 and P6). Antiserum from chinchillas immunized with whole-cell extracts from H. influenzae was previously shown to be protective in the chinchilla model of otitis media (21). This antiserum was used to examine KSCN extracts of various NTHi strains, previously shown to contain numerous proteins while leaving most OMPs intact. Western blot analysis showed a consistent band appearing at approximately 63 kDa. Thus, the decision was made to purify this protein and determine its vaccine potential. This paper reports the identification of a surface-exposed, highly conserved, immunogenic NTHi protein which elicits cross-reactive bactericidal antibodies against NTHi. This protein, called NucA, has been identified as a 5′ nucleotidase and has been cloned, sequenced, and expressed recombinantly. It elicits broadly cross-reactive antibody against NTHi strains and has vaccine potential.
MATERIALS AND METHODS
Bacterial strains and growth.The H. influenzaestrains used were P860295, P861454, TN106, and other NTHi strains from the Pittsburgh, Dallas, and Georgetown collections and type b Eagan and Whittier strains. H. influenzae strains were grown in brain heart infusion broth (Difco Laboratories, Detroit, Mich.) supplemented with 10 μg of hemin stock (0.1 g of hemin, 0.1 g of histidine, and 4 ml of triethanolamine per 100 ml of solution) (Sigma Chemical Co., St. Louis, Mo.) per ml and 40 μg of NAD (Sigma) per ml or with 2% Fildes enrichment (Remel, Lenexa, Kans.) and 20 μg of NAD per ml. The Escherichia coli strain used for construction of the phage library was XL1BlueMRF′ (Stratagene, La Jolla, Calif.) grown in YT medium supplemented with 12.5 μg of tetracycline per ml, 0.2% maltose, and 10 mM MgSO4 as required. InvαF′ and TOP10F′ competent cells (Invitrogen Corp., San Diego, Calif.) were used for transformation of plasmid DNA. Transformants were grown in SOB medium (consisting of, per liter, 20 g of Bacto-tryptone [Difco, Sparks, Md.], 5 g of yeast extract [BBL, Cockeysville, Md.], 0.6 g of sodium chloride, 0.2 g of potassium chloride [pH 7.0 to 7.2] [J. T. Baker, Phillipsburg, N.J.]) supplemented with ampicillin (100 μg ml−1) or kanamycin (40 μg ml−1) as required. For recombinant NucA (rNucA) expression, cells were grown to an optical density of 1.0, measured at 600nm (OD600) and induced with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for 2 h. All cultures were grown at 37°C.
Purification of NucA protein.Approximately 70 g (wet weight) of NTHi strain P860295 was suspended in 0.1 M potassium phosphate–3.0 M potassium thiocyanate (KSCN) (pH 6.0) and stirred at room temperature for 60 min to extract NucA protein. Cells and cell debris were removed by centrifugation at 8,000 rpm using a Sorvall GS-3 rotor for 20 min at 4°C. The supernatant was collected and further clarified by centrifugation at 60,000 rpm using a Beckman 70Ti rotor for 1 h at 4°C. The supernatant was collected and dialyzed overnight at 4°C against 50 mM sodium phosphate (pH 8.0). Dialyzed whole-cell extract was loaded onto a ceramic hydroxyapatite column (80-μm bead size) (Bio-Rad, Hercules, Calif.) equilibrated in 50 mM sodium phosphate (pH 8.0). Bound protein was eluted with 0.2 M sodium phosphate followed by a linear sodium phosphate gradient (0.2 to 0.5 M). Fractions were screened for NucA by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and pooled. Three independent hydroxyapatite chromatography runs were performed to process the entire whole-cell extract. Pooled fractions from each of the three runs were combined and concentrated using an Amicon stirred cell with a PM10 membrane. The concentrated material was dialyzed overnight at 4°C in 50 mM Tris-HCl–1 mM EDTA (pH 9.0). The preparation was subsequently loaded onto a MonoQ HR5/5 column (Pharmacia, Uppsala, Sweden) equilibrated in 50 mM Tris-HCl–1 mM EDTA (pH 9.0). Bound protein was eluted with a linear NaCl gradient (0 to 1.0 M NaCl in 50 mM Tris-HCl–1 mM EDTA [pH 9.0]). Fractions were screened for NucA protein by SDS-PAGE and pooled. A total of 2.5 mg of purified NucA was isolated from about 70 g (wet weight) of cells. The concentrations of protein were determined using the Pierce bicinchoninic acid assay with bovine serum albumin (BSA) as a standard.
Recombinant NucA purification.Approximately 30 g (wet weight) of induced bacterial cells was suspended in TE (50 mM Tris-HCl, 1 mM EDTA [pH 8.0]) and disrupted by three passages through a French pressure cell at 1,000 lb/in2. Cell debris and membranes were removed by centrifugation at 55,000 rpm using a Beckman 70Ti rotor for 20 min at 4°C. The supernatant was collected and dialyzed overnight at 4°C against TE. The dialyzed crude-extract supernatant was loaded onto a trimethylaminoethyl (TMAE) Fractogel column (EM Industries, Inc., Hawthorne, N.Y.) equilibrated with TE at room temperature. The column flowthrough, containing recombinant 63-kDa (r63K) protein, was collected and dialyzed overnight at 4°C against 20 mM HEPES–1 mM EDTA (pH 7.0). The material was then loaded onto a MonoS HR10/10 column (Pharmacia) equilibrated in 20 mM HEPES–1 mM EDTA (pH 7.0). Bound protein was eluted with a linear NaCl gradient (0 to 1.0 M NaCl) in 20 mM HEPES–1 mM EDTA (pH 7.0). Fractions were screened for rNucA protein by SDS-PAGE and pooled. A total of 80 mg of purified NucA was isolated from about 30 g (wet weight) of cells.
Immunogenicity study with NucA protein.Swiss-Webster mice were immunized subcutaneously with 5 μg of NucA protein and 50 μg of MPL (3-O-deacylated monophosphoryl lipid A; RIBI ImmunoChem Research, Inc., Hamilton, Mont.) as adjuvant per dose at weeks 0, 4, and 6. Blood samples were collected at weeks 0 and 10 for analyses of antibody titers to protein, whole cells, and bactericidal activity.
Generation of antibodies.Monoclonal antibodies were generated by immunizing 8-week-old female BALB/c mice intraperitoneally (i.p.) at weeks 0, 2, and 4 with 5 μg of NucA protein and 25 μg of MPL adjuvant per dose. The animals were boosted with 5 μg of protein after a 4-month rest period. During the immunization and rest period, mouse sera were tested for antibody activity by NucA protein enzyme-linked immunosorbent assay (ELISA). About 72 h after the last injection, spleens were recovered and combined with nonsecreting X63Ag8.653 myeloma cells. The cells were fused for 4 min in 50% (wt/wt) polyethylene glycol 1500 and 10% dimethyl sulfoxide in Dulbecco modified Eagle medium and then diluted in Dulbecco modified Eagle medium containing hypoxanthine, aminopterin, thymidine, 10% fetal bovine serum, and 10% NCTC-109 medium supplement. Positive reactors were identified by NucA protein ELISA and P860295 whole-cell ELISA.
Rabbit polyclonal sera was generated in New Zealand White rabbits. Two rabbits, with the lowest preexisting titers to NucA protein, were immunized subcutaneously at weeks 0, 4, 8, and 14 with 25 μg of native NucA protein cut from a SDS-polyacrylamide gel and mixed with 25 μg of MPL. Serum was collected at weeks 0, 10 and 16.
ELISA.ELISA was performed for recombinant and native proteins as well as for whole cells. Briefly, 75 μl of a 2.5-μg/ml concentration of a protein preparation was loaded into each well of a Polysorp 96-well Nunc plate, and incubated for 90 min at 37°C. The plates were sealed and stored at 4°C until use. Primary antiserum was diluted in 1× phosphate-buffered saline (PBS)–0.05% Tween, 50 μl was added per well, and the plate was incubated at 22°C for 120 min. The plates were washed in a Biotek washer (Bio-Tek Instruments, Inc., Winooski, Vt.) with a 30-s soak and five washes with 1× PBS–0.1% Tween. Alkaline phosphatase-conjugated goat anti-mouse IgG (Southern Biotechnology Associates, Inc., Birmingham, Ala.) was added at a 1:2,000 dilution in a 50-μl volume per well and incubated at 22°C for 120 min. The plates were washed as above and developed for 60 min at 22°C with 100 μl of p-nitrophenylphosphate (NPP) at 1 mg/ml in 1 M diethanolamine per well. The reactions were stopped with 50 μl of 3 N NaOH per well. The results were read on an ELISA plate reader (Biotek 312/SLT Spectra apparatus) at 405 nm, ref 690 nm.
For whole-cell ELISA, plates were coated with 80 μl of OD620 ≈ 0.2 formalin-fixed cells per well. The plates were allowed to dry in the hood, sealed, and stored desiccated until use. They were washed and then blocked with 1% BSA for 60 min at 22°C. They were then washed in a Biotek washer using a 30-s soak and five rinses, 50 μl of primary serum diluted in 1× PBS–0.3% Tween was added to each well, and the plates were incubated at 37°C for 120 min. The BAC-T 96-well (Nunc) plates were similarly washed, 80 μl of an alkaline phosphatase-conjugated goat anti-mouse secondary antibody was added at a 1:2,000 dilution in 1× PBS–0.05% Tween, and the plates were incubated at 37°C for 60 min. The plates were washed and developed with 1 mg of NPP per ml in 1 M diethanolamine at 22°C for 60 min. The reactions were stopped by addition of 50 μl of 3N NaOH. The plates were read on a Biotek 312/SLT Spectra apparatus.
Protein gel electrophoresis and immunoblot analysis.The methods and buffers used for SDS-PAGE and Western immunoblot analysis were as described by Sambrook et al. (41). Basically, proteins were separated by SDS-PAGE (12% polyacrylamide), electrotransferred onto a sheet of nitrocellulose (Schleicher & Schuell, Keene, N.H.), and blocked with BLOTTO. The membrane was incubated with the appropriate antisera, washed, and incubated with alkaline phosphatase-conjugated antibody. Reactive bands were visualized using nitroblue tetrazolium (NBT)–5-bromo-4-chloro-3-indolylphosphate (BCIP) as the substrate.
IEM.NucA protein on the surface of NTHi cells was visualized using immunoelectron microscopy (IEM) as described elsewhere (42). Briefly, grids were coated with live NTHi cells, blocked, reacted with mouse polyclonal anti-NucA serum, washed in PBS-BSA, and coated with goat-anti mouse IgG-plus-IgM-containing 12-nm gold spheres (Jackson ImmunoResearch Laboratories, West Grove, Pa.). The grids were rinsed with PBS, fixed with 1% glutaraldehyde, washed with water, and stained with Nano Van, a low-atomic-number vanadium stain (Nanoprobes, Stony Brook, N.Y.). IEM was performed on a Zeiss 10C transmission electron microscope operating at 100 kV.
N-terminal protein sequence.For amino-terminal amino acid sequence analysis of the native protein, 100 μg of purified protein was precipitated with 90% ethanol and centrifuged and the pellet was resuspended in 20% acetic acid. For the recombinant protein, 80 μg of purified protein was added to a ProSpin cartridge (Applied Biosystems, Foster City, Calif.) and centrifuged. The polyvinylidene difluoride membrane containing the sample was removed, washed with 20% methanol, and used for amino-terminal amino acid sequence analysis. Amino-terminal amino acid sequence analysis was carried out using an Applied Biosystems model 477A protein-peptide sequencer equipped with an on-line model 120A PTH analyzer (Applied Biosystems). After the cleavage of each successive amino terminus, the anilinothiazolinone derivative formed was converted to the more stable phenylthiohydantion (PTH) derivative by treatment with 25% trifluoroacetic acid at 64°C for 20 min. The PTH derivatives were separated and identified on the PTH analyzer by reversed-phase high-pressure liquid chromatography using a Brownlee PTH C-18 instrument (Applied Biosystems) with a two-solvent gradient system as specified by the manufacturer.
Cloning vectors and methods.The standard molecular biology cloning methods used were as described by Sambrook et al. (41). The λZapII expression library cloning kit was from Stratagene. The TA cloning kit, containing the PCR cloning vector pCRII, and the pTrcHis expression vector were from Invitrogen. The transformation method, using One Shot competent cells, INVαF′ and TOP10F′, was as recommended by the vendor (Invitrogen).
NTHi λZapII library.NTHi-P860295 chromosomal DNA was partially digested with the Tsp509I restriction enzyme (New England Biolabs, Beverly, Mass.), and DNA fragments in the size range of 4 to 10 kb were gel isolated and ligated toEcoRI-digested and dephosphorylated λZapII arms as specified by the manufacturer. The ligations were packaged and plated on XL1 Blue MRF′ cells.
Immunoblot analysis of library.The λZAP library plaques were blotted onto NC nitrocellulose membranes (Schleicher & Schuell) blocked with BLOTTO and reacted with the appropriate anti-NucA antibody. Polyclonal rabbit anti-NucA serum was absorbed against XL1-Blue MRF′ cells and a λZapII lysate prior to use. Individual monoclonal antibodies or mixtures of monoclonal antibodies to NucA protein were also used to probe the library. The filters were incubated with secondary antibody (goat anti-mouse IgG plus IgM, phosphatase conjugated [Kirkegaard & Perry Laboratories, Gaithersburg, Md.]), and the reactive spots were visualized using NBT and BCIP.
Probe generation.A 420-bp digoxigenin (dig)-labeled PCR probe was generated from either nucA-containing plasmid DNA or NTHi-P860295 genomic DNA, using primers 4101ext and 4102ext and dig-dUTP in a PCR as recommended by the supplier (Boehringer Mannheim, Indianapolis, Ind.).
Southern hybridization.Southern transfer was done by the method of Sambrook et al. (41), and the hybridization was done by following the Genius protocol (Boehringer Mannheim). Hybridization and washes using the 420-bp dig-labeled PCR probe were done at 65 to 70°C.
PCR.DNA amplification was performed using a Perkin-Elmer GeneAmp PCR Systems 2400 unit (Applied Biosystems) with TaqDNA polymerase and deoxynucleoside triphosphates from Boehringer Mannheim. A hot start at 95°C for 2 to 5 min was used prior to the initiation of the PCR amplification. Thirty cycles were performed using the following cycle conditions: 40 to 60 s at 95°C, 40 to 60 s at 40 to 60°C, and 1 to 4 min at 72°C. The primers used are given in Table 1.
Primers used in this study
DNA sequencing.All sequencing reactions were performed with the Applied Biosystems Prism Dye Terminator cycle-sequencing core kit based on the Prism protocol supplied by the vendor. Approximately 1 μg of template DNA and 100 ng of primer were used for each cycling reaction. The reactions were cycled on the GeneAmp PCR Systems 2400 unit, purified using the Prism method, and analyzed on an ABI 373A DNA sequencer (Applied Biosystems).
Mouse anti-rNucA sera.BALB/c mice were immunized subcutaneously at weeks 0, 4, and 8 with 5 μg of purified rNucA protein and 50 μg of MPL per dose. The animals were bled at weeks 0 and 10. Antisera were subjected to titer determination for IgG antibodies and tested in a whole-cell ELISA and for bactericidal activity. A control group immunized with 5 μg of native NucA protein and 50 μg of MPL was included.
Bactericidal assay.Immune sera were incubated at 56°C for 30 min to inactivate complement activity. Bacteria were grown to log phase to about 109 CFU/ml in brain heart infusion broth supplemented with hemin and NAD. The bacteria were then suspended to 105 CFU/ml in sterile PCM buffer (10 mM sodium phosphate [pH 7.4], 150 mM NaCl, 0.5 mM MgCl2, 0.15 mM CaCl2) containing 0.25% BSA. Precolostral calf serum was used as the complement source. Complement was preabsorbed with an overnight culture of bacteria by adding pelleted cells from 1 ml of PCM-washed bacteria with 0.5 ml of complement, incubating them at 4°C for 60 min, pelleting them at maximum speed in a microcentrifuge for 3 min at 4°C, and repeating the process twice. Aliquots of preabsorbed serum were stored at −70°C. The assay was performed using flat-bottom 96-well Costar (Corning, N.Y.) plates. The reagents were added as follows: 15 μl of PCM buffer containing BSA, 15 μl of serially diluted antiserum, 15 μl of premixed bacteria and complement (20 μl of bacteria, 40 μl of preabsorbed complement, and 140 μl of PCM containing BSA). Following incubation at 37°C in a 5% CO2 incubator for 45 min, 15-μl aliquots were plated on chocolate agar plates. After overnight incubation at 37°C, the bactericidal activities of serially diluted antisera were assessed by determining the numbers of CFU. The highest dilutions yielding 50% killing were reported as bactericidal titers. A value fourfold or greater over that for the preimmune sera is considered to show bactericidal activity.
Infant-rat protection study.Four-day-old Sprague-Dawley rats were randomized into 10 groups with a mother for each group of 10 infants. The infants were immunized i.p. with 0.1 ml of the appropriate dilutions of mouse rNucA antiserum from week 10. Preimmune serum and PCM buffer were used as negative controls. The positive control group received a monoclonal antibody raised against Hib capsular polysaccharide (MAbE117.5) (38). All dilutions of sera and cells were done in PCM buffer. Approximately 24 h after immunization, the infant rats were challenged i.p. with approximately 50 CFU (0.1 ml) of virulent Hib Eagan. Approximately 20 to 24 h postchallenge, 10 μl of blood was taken from the tail and viable Hib CFU were determined from duplicate dilutions of blood.
nucA deletion strain.Plasmid DNA containing thenucA gene was digested with SwaI andSnaBI restriction enzymes, which cut within the gene at nucleotides 173 and 781 respectively, resulting in a deletion of 608 bp. This region was replaced with a HincII DNA fragment, containing the HaeII DNA region from pACYC184 (New England Biolabs) carrying the chloramphenicol resistance gene (40). Plasmid DNA containing this deletion insertion was used to transform naturally competent KW20 cells by the standard method of Moxon et al. (27). Chloramphenicol-resistant cells were selected and screened by PCR and by Western and Southern blot analysis.
5′-Nucleotidase assay.Inorganic phosphate released upon hydrolysis of 5′-nucleotides was determined colormetrically using a modification of the method of Chen et al. (11). Reaction mixtures contained 100 mM Tris-HCl (pH 9.0), 6.6 mM MgCl2, 1.65 mM 5′-AMP (or other nucleotide substrate), and enzyme in a final volume of 750 μl. The reaction mixture was incubated at 37°C for 30 min, and the reaction was stopped by the addition of 250 μl of 20% trichloracetic acid. Controls were run for each reaction in which enzyme was added after the addition of 20% trichloroacetic acid. Precipitated protein was removed via centrifugation, and 200 μl of the supernatant was withdrawn into a fresh test tube. A 100-μl volume of 1.0 N HCl was added followed by 750 μl of ammonium molybdate reagent (3.0 ml of 10% ascorbic acid mixed with 18 ml of 3.4 mM ammonium molybdate in 1 N H2SO4), and the mixture was incubated at 37°C for 30 min. After the tubes were allowed to cool to room temperature, the absorbance at 650 nm was determined. One unit of 5′-nucleotidase activity is defined as the amount of enzyme which results in an absorbance change of 1.0 at 650 nm per minute at 37°C.
Phosphatase assay.Phosphatase was assayed using NPP as substrate. Reaction mixtures contained 100 mM Tris-HCl (pH 9.0), 25 mM NPP, and enzyme in a final volume of 750 μl. The reaction mixture was incubated at 37°C for 30 min, and the tubes were placed on ice. The absorbance at 410 nm was determined.
RESULTS
Purification of NucA protein.We have developed a three-step process for extracting and purifying NucA protein from an NTHi cell pellet. Cell proteins, mainly cell surface associated, were extracted from bacteria with 3.0 M KSCN–0.1 M potassium phosphate (pH 6.0) as described in Materials and Methods. As illustrated in Fig.1, NucA purified by this method exhibits a single faint protein band on an SDS-polyacrylamide gel corresponding to a molecular size of ∼63 kDa.
Purification of native NucA from strain P860295 analyzed by SDS-PAGE. Lanes: 1, high-molecular-mass protein standards (Life Technologies) (the corresponding masses are shown to the left in kilodaltons); 2, total-cell extract; 3, KSCN cell extract; 4, purified native NucA. The gel was stained with Coomassie brilliant blue.
N-terminal sequence of NucA.The N terminus of purified NucA protein was sequenced. The first 26 residues were unambiguously identified as Lys-Glu-Ala-Pro-Gln-Ala-His-Lys-Ala-Val-Glu-Leu-Ser-Ile-Leu-His-Ile-Asn-Asp-His-His-Ser-Tyr-Leu-Glu-Pro. Since the protein was eluted from the cell membrane, the N-terminal sequence shown is probably of a processed, mature protein with the signal sequence removed. Because we were able to obtain the peptide sequence of purified NucA protein and show that the N-terminal amino acid residue is not a cysteine, this protein is not a lipoprotein.
Immunogenicity of NucA.Mice were immunized with native purified NucA protein, and their antisera were analyzed. The week 10 antisera had an IgG antibody titer of ∼930,000 and elicited whole-cell ELISA titers against the homologous strain, P860295, and five heterologous NTHi strains, as shown in Table2. In another study, anti-NucA sera showed equivalent whole-cell titers against homologous and two Hib strains, Eagan and Whittier (data not shown). These results indicate that NucA is antigenically conserved among all H. influenzaestrains tested. The same antiserum was then examined in a bactericidal assay. Table 3 shows that the antisera exerted bactericidal activity against both homologous and four heterologous NTHi strains tested.
Antibody titers
Bactericidal activity of NucA antiserum against NTHi strains
Antigenic conservation of NucA.Total-cell lysates of homologous and heterologous H. influenzae strains were probed in a Western immunoblot format with mouse or rabbit polyclonal anti-NucA sera. As shown in Fig. 2, a single major reactive band migrating at ∼63 kDa was detected in allH. influenzae strains tested. When an E. coliextract was probed with these antisera, only a very faint reactive 63-kDa band was observed. This is due to spillover of the neighboring gel lanes, since the predicted molecular mass of the NucA-like protein in E. coli, UshA, is 58 kDa and in other gel experiments, using the same sera, there is no band detected in the E. coli extract lane. Also, using a monoclonal antibody to NucA (Nt63-34-25), only the 63-kDa NucA band was detected in the H. influenzae lanes and no band was observed in the E. coli lane (data not shown).
Western immunoblot probed with mouse polyclonal anti-NucA sera. Various whole-cell extracts were analyzed by SDS-PAGE and probed using anti-NucA mouse polyclonal sera. An E. coliInvaF′ extract was used as negative control (lane 12). The NTHi clinical isolates were from Pittsburgh or Dallas. Two Hib strains, Eagan and Whittier, were included. High-molecular-mass protein standards are shown in lane 1 (the apparent masses are shown to the left in kilodaltons). Lanes: 2, P860295; 3, P861454; 4, P880859; 5, P810384; 6, TN106; 7, N1955; 8, N830161E; 9, H305; 10, Eagan; Whittier; 12, E. coli InvaF′ overnight culture; 13, TN106 overnight culture. All other extracts were from OD600 = 1.0 to 2.0 cultures. All extracts were normalized to OD600 = 1.0 for gel loadings.
Initial cloning of the nucA gene.Initially, a λZapII expression library of NTHi P860295 genomic DNA was constructed and probed with polyclonal and monoclonal anti-NucA antisera. No reactive plaques were detected, suggesting that the library was defective (although when probed with P4 antisera, reactive plaques were detected [data not shown]), that expression of NucA protein inE. coli was toxic, or that the nucA gene was incomplete. Phage DNA was then isolated from the λZapII expression library and probed using PCR amplification.nucA-specific degenerate oligonucleotide primers, P1 and P2, were used in combination with λZapII-specific primers, T7 and T3, in separate PCR amplifications. PCR products of ∼1.8 kb and 600 bp were cloned into pCRII and sequenced. Inspection of the DNA sequence revealed an open reading frame whose inferred amino acid sequence was identical to the N-terminal amino acid sequence of the NucA protein. This confirmed the correct cloning of part of the nucA gene. As expected from the size of the PCR product and the predicted size of the nucA gene, the cloned PCR DNA did not contain the complete carboxy end of the gene.
Cloning of the 3′ end of the nucA gene.A restriction map of NTHi P860295 genomic DNA was constructed from Southern blot experiments using a nucA-specific, 420-bp probe. In a Southern blot reaction with the 420-bp probe,EcoRI-digested chromosomal DNA produced an intense, approximately 4.3-kb reactive band and a faint approximately 1.8-kb, reactive band (data not shown). Based on the partial nucADNA sequence obtained above, there is an EcoRI site within the probe region, 40 bp from the 3′ end of the probe. This, as well as the hybridization stringency conditions used in the Southern blot experiment, suggested that the faint, 1.8-kb reactive band should contain the carboxy end of the nucA gene plus additional downstream sequences. Genomic DNA was then digested withEcoRI, self-ligated, and used in an inverse DNA PCR amplification with primers 4313ext and 4102ext. An ∼1.3-kb PCR product was observed, cloned into pCRII, and sequenced. The DNA sequence revealed a partial open reading frame that overlapped with the partial nucA sequence of the PCR clones above and continued downstream from the nucA gene coding region.
Cloning of the complete nucA gene.Multiple attempts to amplify the 5′ end of the nucA gene using various nucA-specific primers and either total, or gel-isolated (3.5-5 kb fragments) EcoRI-digested NTHi P860295 DNA in an inverse DNA PCR amplification failed to generate a correct DNA PCR fragment that contained any upstream nucAsequence. Based on the partial nucA sequence, there is a single NsiI site near the 3′ end of the gene. Southern blot analysis of NsiI-digested genomic DNA revealed an ∼3-kb reactive band with the 420-bp probe. This reactive fragment should contain upstream nucA sequence. Genomic DNA was digested with NsiI, and ∼3-kb DNA fragments were gel purified and used in an inverse PCR assay with the nucA primers 4102ext and 4313ext. The predicted ∼2-kb PCR product was cloned into pCRII and sequenced. The DNA sequence overlapped the N-terminal region of thenucA gene and contained additional upstream DNA sequence. To verify the proposed composite nucA gene sequence, a PCR amplification was performed using outside flanking PCR primers 63K ORF (5′-GTGGAAGTTGATGATTTCTCG-3′) and 63K rev (5′-GAAGTCTTCAAACCTAGGAC-3′) with NTHi P860295 genomic DNA. The predicted 2.0-kb PCR fragment was detected, cloned into pCRII, and sequenced. The sequence agreed with the proposed composite sequence from the partial clones obtained above. Figure3 shows the complete DNA sequence of thenucA gene and its inferred amino acid sequence. The calculated molecular mass of the full-length NucA preprotein is 66,200 Da, and that of the processed NucA mature protein is 63,571 Da.
Complete nucleotide and amino acid sequences of NucA. The presumed ribosome binding site is labeled RBS. Amino acids are listed starting with the initiation codon, methionine. The signal sequence is underlined, and the start codon, lysine, of the mature sequence is shown by an arrow. The locations of various restriction sites are displayed above the sequence. The boxed region demarcates the region of strong protein homology observed among known 5′-nucleotidases. Arrows below the DNA sequence at the 3′ end of the gene depict an inverted-repeat region. ORF, open reading frame.
Expression of rNucA.From the DNA sequence shown in Fig. 3, the inferred NucA protein sequence has a leader region of 25 amino acids that is presumed to be the signal sequence, required for transport of the protein to the cell membrane. For expression inE. coli, the entire nucA gene, with its signal sequence, was cloned into the pTrcHis expression vector and called pPX691. The cloning strategy introduced two additional amino acids at the 5′ end of the signal sequence. Whole-cell extracts from induced cultures containing pPX691 were analyzed on SDS-polyacrylamide gels (Fig. 4). Coomassie blue staining of the gel proteins showed a major induced band of ∼63 kDa, suggesting that this is the rNucA protein (Fig. 4A). Western immunoblot analysis using anti-NucA serum confirmed that the major induced protein band was indeed NucA protein (Fig. 4B). The presumed lower-molecular-mass bands observed in Fig. 4B, lane 3, are most probably breakdown products of NucA, as evidenced by the observed lower-molecular-mass breakdown product of purified NucA protein (lane 4) and no lower-molecular-mass bands in the uninduced E. coli culture (Fig. 4A and B, lanes 2).
SDS-PAGE analysis of E. coli expressing rNucA protein. An isolate of InvαF′ cells harboring pPX691 was grown in SOB medium containing ampicillin and induced with IPTG. Total-cell lysate was subjected to SDS-PAGE and either stained with Coomassie brilliant blue (A) or probed with mouse polyclonal anti-NucA serum in a Western immunoblot assay (B). Lanes: 1, high-molecular-mass protein standards (shown in kilodaltons on the left); 2, isolate carrying pPX691, uninduced extract; 3, same isolate, induced extract; 4, purified rNucA. The arrow depicts the migration of NucA.
The rNucA protein was purified, and the N terminus was sequenced. The amino acid sequence obtained was in complete agreement with the original N-terminal amino acid sequence of the native NucA protein, demonstrating the correct processing of the native signal sequence in the recombinant E. coli clone.
Immunogenicity of rNucA.Animal studies were done to compare the immunogenicities of rNucA and native NucA protein. BALB/c mice were immunized with 5 μg of purified protein and MPL as adjuvant. Antisera were subjected to titer determination for IgG antibodies, whole-cell ELISA, and bactericidal activities. The IgG antibody titers were comparable for native NucA and rNucA immunogens (∼930,000 and ∼960,000, respectively). The antisera also had similar whole-cell ELISA titers to both homologous (P860295) and five heterologous strains, as shown in Table 2. In another study, antiserum to rNucA exhibited whole-cell ELISA activity to 13 NTHi strains and 2 Hib strains, Eagan and Whittier (data not shown). The rNucA antiserum also had similar bactericidal activity to that of the native NucA antiserum (Table 3). In general, there was no significant difference in titers among all of these groups and between native NucA and rNucA antisera. The sera from all the recombinant immunized groups were pooled and used in a passive-immunization protection study of infant rats (discussed below). The anti-rNucA serum also recognized both native and rNucA proteins in a Western immunoblot assay (data not shown).
Infant-rat protection study.The infant-rat animal model for Hib meningitis demonstrates the ability of antibodies to reduce bacteremia (and subsequent death) due to challenge with Hib (29). Infant rats were passively immunized with the mouse anti-rNucA pooled sera above and, approximately 24 h later, challenged with Hib. About 22 h postchallenge, blood was taken from the rats and viable Hib colonies were determined. Table4 summarizes the results from this study. The group receiving a 1:2 dilution of mouse anti-rNucA pooled sera showed about a 10-fold reduction in the level of bacteremia compared to the group vaccinated with week 0 (preimmune) pooled mouse sera. This demonstrates the effectiveness of anti-rNucA sera in reducing the growth of Hib and therefore the potential use of NucA protein as a vaccine candidate against H. influenzae.
Infant-rat protection study
IEM.NucA on the surface of NTHi cells labeled with anti-NucA antiserum and visualized by IEM showed weak labeling. To eliminate any possibility that the weak labeling was due to non-NucA antisera, we adsorbed the sera with a genetically engineered nucA-deleted strain. As expected from the bactericidal activity data shown in Table3, P860295 cells showed poorer IEM labeling (approximately 75% less) compared to TN106 cells. Figure 5 shows the level of anti-NucA labeling to the outer surface of TN106 by using this adsorbed serum. (Note that normal use of a high-atomic-number stain, such as uranyl acetate, will obscure either the body of the bacterium or the gold beads, but with the vanadium stain used here, both are clearly visible.) As expected, the adsorbed serum did not show any labeling of the nucA deletion strain on IEM (data not shown). Overall, the low-level detection of NucA on the surface of cells is in agreement with the poor yield of purified native NucA protein obtained from NTHi cells.
IEM of negatively stained TN106 showing the binding of antiserum to NucA. Binding was detected with goat anti-mouse antiserum conjugated with 12-nm gold spheres, and the cells were stained with vanadium. Mouse polyclonal anti-NucA serum was preabsorbed with a whole-cell nucA knockout mutant to reduce background. Bar, 100 nm.
Conservation of NucA sequences.The nucA gene was cloned from nine NTHi strains and a Hib strain. The DNA sequence was determined, and the derived protein sequences were aligned. Comparison of the NucA protein sequence from strain P860295 to that from the other strains showed a strikingly high degree of amino acid homology among all the strains sequenced. The NucA sequences from four strains (P880859, TN106, SH1015, and Eagan) are 100% identical, while those of the other five strains differ by 1 to 7 amino acid residues, as shown in Table 5. Analysis of the NucA protein derived from the published H. influenzae type d strain Rd sequence (gene designation, HI0206) (15) showed only 6 amino acid residue differences. These results support the notion that thenucA gene is highly conserved among Haemophilusspp.
Mature NucA amino acid residue differences between strain P860295 and other H. influenzae strains
5′-Nucleotidase activity of the NucA protein.A NucA protein homology search was performed using the National Center for Biotechnology Information BLAST program (2). The highest score matches were to genes that encode 5′-nucleotidase activity. A Lipman-Pearson protein alignment (DNAStar) was performed comparing NucA protein with various nucleotidase protein sequences from rat (25), human (26),Vibrio parahaemolyticus NutA (44), and E. coli UshA (10). Two small regions of striking homology were observed among the 5′ nucleotidase proteins as depicted in Fig. 3. This domain has been suggested to be involved in the catalytic activity and/or binding of substrate (25). 5′-Nucleotidase activity was detected on both purified native NucA and rNucA proteins as well as NTHi whole cells. Using 5′ AMP as the substrate, direct enzymatic assay of 5′-nucleotidase in purified preparations of both native and recombinant NucA confirmed the presence of this activity. The specific activities of the purified enzymes were shown to be similar, 6.36 and 6.06 U/min/mg for native NucA and rNucA, respectively. We have therefore named this genenucA (for “nucleotidase”).
Incubation of purified rNucA with a monoclonal antibody (Nt63-34-25) raised against purified native NucA resulted in the immunoprecipitation of the 5′-nucleotidase enzyme activity (data not shown). This confirms that NucA protein has 5′-nucleotidase activity.
Divalent cations stimulate (and in some cases are an absolute requirement for) 5′-nucleotidase activity in both procaryotic and eucaryotic systems (6). Table6 shows that the addition of Mg2+ resulted in maximal stimulation and that Mn2+ gave the second-strongest stimulation. The addition of Zn2+ inhibited the activity of the rNucA protein by 75%; however, an inhibition of only 20% was observed for the native NucA protein. The reason for this inconsistent observation is unclear. It is possible, however, that the native NucA protein, in a more natural confirmation, may bind any endogenous cation more strongly, thereby outcompeting the Zn2+ at the cation binding site. The fact that the NucA is not stimulated to the same degree by Mg2+as is rNucA supports this theory. A 5′-nucleotidase isolated from V. parahaemolyticus was shown to require Cl− for activity (44). Therefore, the effect of KCl on NucA activity was examined. The addition of KCl at concentrations up to 100 mM had no effect on the 5′-nucleotidase activity (data not shown).
Stimulation of 5′-nucleotidase activity by divalent cationsa
Native NucA and rNucA exhibited similarly broad substrate specificity profiles. Both were shown to hydrolyze a variety of nucleoside mono-, di-, and triphosphates (Table 7). The NucA protein displayed no activity with NPP or 3′AMP, suggesting its specificity for 5′-nucleotide substrates (data not shown).
Substrate specificity of 5′-nucleotidase activitya
DISCUSSION
There have been only a few integral OMPs, such as the lipoproteins D, P4, P6, and PCP and a nonlipoprotein, D15, isolated from NTHi strains that are antigenically conserved, induce bactericidal antibodies, and show partial protection in animal models (1, 12, 18, 20, 24). We undertook a different strategy to search for a nonintegral OMP. Using a chaotropic agent, we extracted a novel nonintegral cell surface protein. This nonintegral OMP, NucA, has an apparent molecular mass of 63 kDa and is antigenically conserved and of similar size among all Haemophilus spp. tested in a Western immunoblot assay. However, the amount of NucA surface protein is small as determined from the yield of protein purified from cells. This was corroborated by the number of 12-nm gold beads displayed on the NTHi surface (an average of 150/surface). Since the yield of NucA protein is low, we cloned and expressed the NucA protein in E. coli.
The entire nucA gene and flanking DNA region was cloned inE. coli and sequenced. Examination of the inferred NucA protein sequence revealed a putative signal sequence, as predicted by the SignalP program (37), whose cleavage site, which is similar to the E. coli periplasmic maltose binding protein (14), agreed with the N-terminal amino acid sequence obtained from the mature native protein. This cleavage site is different from the proposed lipoprotein consensus cleavage site sequence of NutA from V. parahaemolyticus, which is a putative lipoprotein (44). It is worthwhile to note that 54 out of 579 residues of the mature NucA protein are lysines, making this protein highly charged and potentially useful as a vaccine conjugate. Immediately upstream of the nucA gene is an open reading frame. This putative open reading frame is transcribed in the same direction as nucA, and its stop codon (UAA) is only 22 bases away from the AUG start codon of nucA. Examination of this intervening sequence does not reveal any obvious promoter sequence for the nucA gene, suggesting that nucA may be part of an operon, carried within a polycistronic mRNA.
The recombinant nucA gene, when expressed in E. coli, produced approximately 20% of the total cellular protein as NucA protein. The N-terminal amino acid sequence of the purified rNucA protein was identical to the sequence of the purified native NucA protein. This confirmed the correct processing of the native signal sequence in E. coli and the possible localization of the rNucA protein to the cell wall. However, crude protein fractionation studies, IEM analysis, and immunofluorescent flow cytometry results all showed low to negligible levels of rNucA on the surface of E. coli (data not shown).
Both native NucA and rNucA proteins elicited similar immunogenicity antibody titers in mice. Anti-NucA sera exhibited limited but detectable bactericidal activity in vitro against certain strains ofH. influenzae tested. The limited activity is most probably due to the low level of NucA observed on the cell surface. Passive immunization of infant rats with mouse rNucA antiserum showed approximately a 10-fold reduction in bacteremia. Similar passive-immunization results have also been observed with the D15 OMP of Hib and NTHi (24). Although our results show limited functional anti-NucA antibody activities, the exact role of the host defense mechanisms against H. influenzae infection has yet to be determined. It has been suggested that complement-directed bacteriolysis may not play the major role in the clearance of H. influenzae and that opsonic activity may be important (39). Likewise, in Bordetella pertussisinfection, there seems to be no correlation between protection and antibody level in human post-acellular pertussis vaccination sera. Other factors should be taken into account. For example, it has recently been demonstrated that gamma interferon produced by T cells plays an important role in controlling infection in animals (3).
In addition to the cross-reactivity observed with anti-NucA sera, additional support that the nucA gene is highly conserved among H. influenzae strains came from analyzing the DNA sequence of nine NTHi strains, one Hib strain, and one Rd strain. Comparison of the deduced protein sequences revealed only a minor number of amino acid differences among all the NucA proteins analyzed, ranging from 98.8% to 100% homologous. Even at the DNA level, there was only a limited number of base differences.
The published genomic DNA sequence of H. influenzae Rd lists the nucA gene as HI0206 and its product as a putative 5′-nucleotidase (16). We have experimentally determined that the nucA gene encodes a 5′-nucleotidase protein and hence have called it NucA. The NucA protein has limited homology to the E. coliperiplasmic bifunctional UshA protein (for “UDP-sugar hydrolase”), mainly in the region thought to contain the catalytic nucleotidase domain (44). Therefore, NucA probably does not possess sugar hydrolase activity as has also been determined for NutA (44). One possible role for 5′-nucleotidase is the hydrolysis of extracellular nucleotides and the transport of the nucleoside products into the cell or into the periplasmic space for bacterial wall polysaccharide biosynthesis (17). The exact location of NucA within the bacteria is unclear. Our antibody and protection data suggest that NucA is surface exposed, but it may also be located in the periplasm based on the Psort program of Nakai and Kanehisa (35).
It is interesting that the nucA gene, being so highly conserved among all H. influenzae strains tested, is not essential for growth in the laboratory, as has also been observed inE. coli and Salmonella enterica serovar Typhimurium (5, 9). In contrast to these observations, it has been shown that 5′-nucleotidase inhibitors have an inhibitory effect on the in vitro growth of Streptococcus mutans and therefore may be candidates to help prevent dental caries (45).
In conclusion, the NucA protein is highly homologous among allHaemophilus strains tested by both Western immunoblot analysis and DNA sequence analysis. The protein is a good immunogen, producing functional antibodies with cross-reactivity. These properties make NucA a good potential subunit vaccine candidate against NTHi infection.
ACKNOWLEDGMENTS
We thank Terri Mininni for antibody production and assays and Robert Hazelo and Brenda Novitsky for oligonucleotide synthesis, PCR, and DNA sequencing. We also thank Stacey Meredith and Kathryn Mason for the in vitro bactericidal assays, Darla Lane for the protein and whole-cell ELISAs, and MaryJane del Mastro, Deb Chakravarti, and Jeffrey Caplan for the protein sequencing.
Notes
Editor: D. L. Burns
FOOTNOTES
- Received 5 January 2000.
- Returned for modification 2 February 2000.
- Accepted 14 February 2000.
- Copyright © 2000 American Society for Microbiology
