Mutant bacteria for production of generalized modules for membrane antigens

Abstract

Gram-negative bacterial strains are generated by inactivating at least one LytM catalytic domain-containing protein, such as NT013, NT017 and NT022 of non typeable H influenzae. The vesicles from these strains are useful for vaccination.

Claims

1. A pharmaceutical composition comprising (a) outer membrane vesicles from a Gram-negative bacterium in which at least one Metalloproteases of the lysostaphin-type (LytM) catalytic domain-containing protein is inactivated and (b) a pharmaceutically acceptable carrier, wherein the composition does not comprise any living and/or whole bacteria.

2. The pharmaceutical composition of claim 1, wherein the at least one LytM catalytic domain-containing protein is knocked out.

3. The pharmaceutical composition of claim 1, wherein the at least one LytM catalytic domain-containing protein comprises: an amino acid sequence with at least 95% identity to the sequence SEQ ID NO: 1, an amino acid sequence with at least 95% identity to the sequence SEQ ID NO: 3, or an amino acid sequence with at least 95% identity to the sequence SEQ ID NO: 5.

4. The pharmaceutical composition of claim 1, wherein the Gram-negative bacterium is selected from the group consisting of: non-typeable Haemophilus influenza (H. influenza), Neisseria meningitidis (N. meningitidis), and Bordetella pertussis (B. pertussis).

5. The pharmaceutical composition of claim 1, wherein the Gram-negative bacterium is non-typeable H. influenzae.

6. The pharmaceutical composition of claim 1, wherein the Gram-negative bacterium is not non-typeable H. influenzae.

7. A method of making the pharmaceutical composition of claim 1 comprising the step of admixing vesicles from a Gram-negative bacterium in which at least one LytM catalytic domain-containing protein is inactivated with a pharmaceutically acceptable carrier.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) FIG. 1: Expression and subcellular localization of NTHi LytM factors. (A) Western blot analysis on different cell compartments extracts were performed using specific antisera raised against NT013, NT017 and NT022. 1Recombinant protein, 2total extract WT, 3 total extract KO, 4outer membrane proteins WT, 5outer membrane proteins KO, 6periplasmic fraction WT, 7periplasmic fraction KO, 8supernatant, WT 9supernatant KO. Red arrows indicate the specific signals. As expected, no specific reactivity is observed with the mutant strains; however the antisera cross-react with other not specific bands present also in the knockout strains which were not characterized. (B) Immunofluorescence microscopy analysis on Hi176 wild type strain and 176ANT022 mutant confirming the surface localization of protein NT022. Bacteria are red and LytM factors in green. (C) Model of LytM proteins localization in NTHi.

(2) FIG. 2: Phenotypic characterization of NTHi Hi176 wild type and LytM mutants. Aggregation phenotype in liquid static cultures growth for 16 h at 37 C. (A) and CFU per milliliter (B) CFU counts was performed at two different OD. The 176ANT017 strain has a growth rate and a CFU similar to the parent strain, while the 176ANT013 and 176ANT022 mutants showed a reduced growth rate growth and a lower CFU. (C) Confocal imaging showing bacterial aggregation of 176ANT013 and 176ANT022 strains, bacteria are stained in red and Dapi in blue.

(3) FIG. 3: Confocal and electron microscopy on LytM mutants. (A) Confocal imaging of Hil76 wt, 176NT013 and 176NT022, bacteria are stained in red (anti total bacterium) and blue (DAPI). (B) Scanning electron microscopy of 176 wt, 176NT013 and 176NT022. The mutant 176NT017 does not show any difference compared to the wild type strain (data not shown).

(4) FIG. 4: Septum formation in LytM mutants. Transmission electron microscopy on Hi176 wt and LytM mutants. Red arrows indicate impaired septum formation in mutants 176N7013 and 176N7022.

(5) FIG. 5: The mutants 176NT013 and 176NT022 release more OMVs than the wild type strain. Transmission electron microscopy of Hil76WT, 176NT013 and 176NT022 mutants and of their respective OMVs preparations. Red arrows indicate OMVs that are released from bacterial surface.

(6) FIG. 6: Analysis of OMVs. Coomassie stained SDS page gel of OMVs prepared from the wild type, 176NT013 and 176NT022 strains (A). Mass spectrometry identification was performed on selected bands (B). Luciferase assay using HEK293 cells stably expressing NF-iB-luciferase reporter cassette and TLR2 (C) or TLR4/MD2/CD14 (D). The stimulation of TLR receptors is assessed by measuring the NF-xB-induced luciferase activity after 6 hours incubation with serially diluted OMVs. IL-6 and TNF levels were measured in hPBMCs stimulated (O.N.) with different dilutions of OMVs purified from wt and mutant strains (E-F).

(7) FIG. 7: Analysis of proteins found in OMVs

(8) FIG. 8: Lipoproteins in OMVs. OMVs from WT, tolR, 17 have similar amounts of lipoproteins, while 13 and 22 mutants are enriched for lipoproteins, in particular NTHI1957 and NTHI 0353 lipoproteins.

(9) FIG. 9: Periplasmic proteins in OMVs. The major periplasmic protein is the periplasmic serine protease do HhoA. This is particularly true from the vesicle derived from 13. Compared to OMVs generated from the WT, OMVs derived from tolR and 22 are enriched in periplasmic proteins. Compared to OMVs generated from the WT, OMVs derived from 17 contain low amount of periplasmic proteins.

(10) FIG. 10: Outer membrane proteins in OMVs. From each type of OMVs, the most abundant outer membrane proteins are the nt099 (outer membrane protein P2) and nt092 (outer membrane protein P5). The main drop in outer membrane protein amount observed for OMVs derived from 13 and 22 is due to the reduction of P2.

(11) FIG. 11: Relative amounts of proteins in OMVs. Relative amounts of proteins nt067, nt014, nt022 and nt016 found in OMVs produced by mutant strains, relative to the wild type strain.

(12) FIG. 12: DAPI staining of MC58 1483 and wild type bacteria.

(13) FIG. 13: TEM (A) and SEM (B) analysis of MC58 1483 and wild type bacteria.

(14) FIG. 14: Protein analysis of vesicles produced by MC58 1483 and wild type bacteria.

(15) FIG. 15: DAPI staining of BP1721 knockout and wild type bacteria.

(16) FIG. 16: Multiple alignment for the NlpD homologues.

(17) FIG. 17: Multiple alignment for the YebA homologues.

(18) FIG. 18: Multiple alignment for the EnvC homologues.

(19) FIG. 19: Protein analysis of vesicles produced by Bp W28 9G/129K NlpD and wild type bacteria.

MODES FOR CARRYING OUT THE INVENTION

(20) Single knock-out non-typeable Haemophilus influenzae (NTHi) mutants were generated for three LytM metalloproteases: NTHI0532 (NT013), NTHI10915 (NT017) and NTHI0830 (NT022). These mutants displayed cell surface defects which caused an increase in the release of Outer Membrane Vesicles (OMVs). Furthermore, these proteins were shown to be involved in bacterial cell division and pathogenesis. In particular, NT013 and NT022 are fundamental for peptidoglycan cleavage and cell splitting. NT017 has a strong influence on NTHi colonization and host immunity evasion.

(21) Methods

(22) Bacterial Strains and Growth Conditions

(23) NTHi Strain 176 was used for this study. It was part of a Finnish otitis media cohort study, as isolate obtained from the middle ear. NTHi was cultivated on chocolate agar polivitex (BioMerieux) incubated at 37 C. with 5% CO.sub.2. Brain-heart infusion (BHI) broth (Difco Laboratories) supplemented with 10 g/mL each of haemin (Fluka Biochemika) and nicotinamide adenine dinucleotide (NAD, Sigma) was used as fluid growth medium. Escherichia coli strains DH5a, HK100 and BL21 (DE3) (Invitrogen) were used for cloning and expression of LytM proteins. They were cultured at 37 C. in Luria Bertani (LB) medium and, when required, supplemented with 100 g/mL ampicillin.

(24) Cell Cultures

(25) Tissue culture cells used in this study are Chang epithelial cells (Wong-Kilbourne derivative, clone 1-5c-4, human conjunctiva, ATCC CCL-20.2) and HEK293 (human kidney, ATCC CRL1573). Chang cells were maintained in Dulbecco's Modified Eagle's Medium (D-MEM; Gibco) supplemented with 25 mM Hepes, 15 mM L-glutamine, antibiotics and 10% (vol/vol) heat-inactivated fetal calf serum (FCS, Invitrogen Corporation). They were grown at 37 C. with 5% CO.sub.2.

(26) HEK293 cells stably expressing TLR2 or TLR4/MD2/CD14 and the NF-B-luciferase reporter cassette, were cultured in DMEM containing 4.5 g/ml glucose, supplemented with 10% heat inactivated FBS, 100 U/ml pelicillin, 100 g/ml streptomycin, 2 mM glutamine, 5 g/ml puromycin 250 g/ml hygromycin (and plus 10 g/ml Blasticidin for HEK293-TLR4 cells).

(27) Cloning of Genes Coding Fr LyM Proteins

(28) LytM genes were cloned into the pET15b+ vector (Novagen) by the polymerase incomplete primer extension (PIPE) method (119). In brief, sequences coding for each protein were amplified by PCR from the HI176 genomic DNA, removing the signal peptide. PCRs generated mixtures of incomplete extension products, by primer design, short overlapping sequences were introduced at the ends of these incomplete extension mixtures, which allowed complementary strands to anneal and produce hybrid vector-insert combinations. Escherichia coli HK100 cells [120] were then transformed with vector-insert hybrids. Single ampicillin-resistant colonies were selected and checked for the presence of the recombinant plasmid by PCR. Plasmids from positive clones were isolated and subcloned into competent E. coli BL21(DE3) cells.

(29) Expression and Purification of Recombinant Proteins

(30) For protein purification, one single colony of E. coli BL21(DE3) strain expressing NTHI0532, NTHI0915 and NTHI0830 were inoculated in LB+ampicillin and grown overnight at 37 C., diluted in fresh LB medium and grown at 30 C. to an OD of 0.6-0.8. The protein over-expression was induced by the addition of 1 mM isopropyl-1-thio--D-galactopyranoside (IPTG; Sigma) for 4 hours. Recombinant 6His-fusion proteins was purified by affinity chromatography on Ni.sup.2+-conjugated chelating fast-flow Sepharose 4B resin (Pharmacia). The purity was checked by SDS-PAGE electrophoresis staining with Coomassie blue. Protein concentration was determined using the bicinchoninic acid (BCA) assay (Thermo Scientific).

(31) Construction of the Knockout Mutants

(32) Deleted mutants of NTHI0532, NTHI0915 and NTH10830 were constructed by allelic replacement of each whole gene with an erythromycin resistance cassette. Upstream and downstream regions of the three genes were amplified by PCR using the primers listed below and cloned in Stratagene pSC-A TOPO vector. Erythromycin resistance cassette was purified from pIM13 plasmid. The constructs containing upstream regions, resistance cassette and downstream regions were assembled. Plasmids obtained were linearized and used to transform 176 NTHi strain using MIV protocol [121]. Knockout strains obtained were confirmed by PCR western blot and locus sequencing.

(33) TABLE-US-00003 NT013 NT013 TTGCACGCGCCAAAATACC SEQID 5FOR NO:25 NT013 TGCATGCATTTACGTGTTGCACTGGCATC SEQID 5REV NO:26 NT013 TGCATGCATTGTTCGTGTTCGTGAAGCAG SEQID 3FOR NO:26 NT013 AACGCGATTGCGTAATGCAG SEQID 3REV NO:28 NT017 NT017 TGCTGGTGCAATTTGATCTTC SEQID 5FOR NO:29 NT017 TGCATGCATTGATTAACGCCAAAACGCAAC SEQID 5REV NO:30 NT017 TGCATGCATATTAGCCGTAAAGGAACGCC SEQID 3FOR NO:31 NT017 TGGCGATCTAATGAACGCAC SEQID 3REV NO:32 NT022 NT022 AAACATTGTGCAACAATGGGG SEQID 5FOR NO:33 NT022 TGCATGCATACAAGACTCAAAGGGAGTAAG SEQID 5REV NO:34 NT022 TGCATGCATGGATCCAGTACGTTACCTAC SEQID 3FOR NO:35 NT022 GTTTCTTTGTCCGCAGGTTC SEQID 3REV NO:36

(34) Preparation of Polyclonal Antisera

(35) Groups of four CD1 mice were immunized to produce polyclonal antisera; 10 g of purified protein was used for each mouse. The recombinant protein was given intraperitoneally in the presence of aluminum. A second (day 21) and a third (day 35) booster doses were administered. Blood sample was taken on day 49.

(36) The treatments were performed in accordance with internal animal ethical committee and institutional guidelines.

(37) Cell Fractionation and Western Blot Analysis

(38) Haemophilus strains were grown in BHI until mid-log phase at 37 C. with 5% CO.sub.2.

(39) Whole cell lysates and periplasmic fractions were purified using PeriPreps Periplasting kit from Epicentre. Outer membrane proteins (OMPs) were recovered on the basis of Sarkosyl-insolubility following the rapid procedure as described by Carlone et al. [122].

(40) To prepare culture supernatants, bacteria were harvested at 13000 g for 10 min at 4 C. 1 ml of culture supernatant was filtered through a 0.22 mm filter and precipitated with vol of 50% TCA for 1 h at 4 C.

(41) After centrifugation at 13000 g for 30 min, the achieved pellet was washed once with 70% ethanol and resuspended in IX sample loading buffer.

(42) Proteins of each cell fraction were separated by SDS-PAGE electrophoresis using NuPAGE Gel System, according to the manufacturer's instructions (Invitrogen), and revealed by Coomassie-blue staining or transferred onto nitrocellulose membranes for Western blot analysis.

(43) Western blots were performed according to standard procedures. The different LytM proteins were identified with a polyclonal mouse antiserum raised against recombinant NTHI0532, NTHI0915 and NTHI0830 (diluted 1:1000) and an anti-mouse antiserum conjugated to horseradish peroxidase (DAKO), as secondary antibody. Bands were visualized with Super Signal Chemiluminescent Substrate (Pierce) and with Opti 4CN Substrate Kit (Bio-Rad) following the manufacturer's instructions.

(44) Confocal Microscopy

(45) The presence of LytM proteins on NTHi surface was checked using confocal imaging. Knockout mutants were used as negative controls. Bacteria were grown until exponential phase, and fixed in 4% paraformaldehyde (Sigma). After multiple washings, bacteria were spread on polylisine-coated slides and blocked with PBS+3% bovine serum albumin (BSA) (Sigma) for 30 min at room temperature. Samples were washed and incubated with specific antisera (1:1000) for 15 min at room temperature. LytM antisera were preadsorbed with intact KO bacteria to minimize cross-reactivity. Bacteria were washed several time with PBS and incubated with Alexa Fluor 488 goat anti-mouse IgG (1:400) (Molecular Probes). Labelled samples were mounted with ProLong Gold antifade reagent with DAPI (Molecular Probes) and analysed with ZeissLSM710 confocal microscope.

(46) Scanning and Transmission Electron Microscopy

(47) Electron microscopy was performed on 176 wt and knockout strains to observe defects in bacterial morphology. Bacteria were grown until exponential phase, washed with PBS and fixed overnight in cacodylate sucrose buffer containing 2.5% glutaraldehyde and 2.5% paraformaldehyde. Samples were then postfixed in 1% OsO4 and 0.15% ruthenium red in cacodylate buffer, blocked with 1% uranyl acetate and dehydrated with serial dilution of acetone.

(48) For SEM, samples were then dried by the critical point method using CO.sub.2 in a Balzers Union CPD 020, sputter-coated with gold in a Balzers MED 010 unit, and observed with a JEOL JSM 5200 electron microscope. For TEM, samples were fixed and dehydrated as described above then embedded in Epon-based resin. Thin sections were cut with a Reichert Ultracut ultramicrotome by use of a diamond knife, collected on collodion copper grids, stained with uranyl acetate and lead citrate, and observed with a JEOL 1200 EX II electron microscope.

(49) Preparation of Outer Membrane Vesicles

(50) Native Outer membrane vesicles (OMVs) were isolated from WT and mutant strains, growing the bacteria until exponential phase in 200 ml BHI cultures. Bacteria were then centrifuged and supernatant were filtered and left at 4 C. overnight adding proteases inhibitor and EDTA. Supernatant were ultracentrifuged for 3 hours at maximum 200000g and final pellet containing OMVs was resuspended in PBS.

(51) Mass Spectrometry

(52) SDS-PAGE Coomassie stained bands were excised and destained in 50 mM NH.sub.4HCO.sub.3 50% acetonitrile. After a drying step, bands were in-gel digested with 12.5 ng/ml Trypsin in 5 mM NH.sub.4HCO.sub.3 overnight at 37 C. The reaction was stopped by the addition of 0.1% final concentration Trifluoroacetic acid (TFA) and the samples were subjected to MALDI-TOF Mass Spectrometry analysis. 1 l of digestion solution was spotted on a PAC target (Prespotted AnchorChip 96, set for Proteomics, Bruker Daltonics) and air-dried at room temperature. Spots were washed with 0.6 l a solution of 70% (vol/vol) ethanol, 0.1% (vol/vol) TFA. Peptide mass spectra were externally calibrated using the standards pre-spotted on the target. Peptide molecular masses determination was performed using a MALDI-TOF/TOF mass spectrometer UltraFlex (Bruker Daltonics, Bremen, GmbH). Ions generated by laser desorption at 337 nm (N2 laser) were recorded at an acceleration voltage of 25 kV in the reflector mode. In general, approximately 200 single spectra were accumulated for improving the signal/noise ration and analyzed by FlexAnalysis (version 2.4, Bruker Daltonics) Peptide mass fingerprints were performed using MASCOT searches against Haemophilus influenzae 86-028NP database using the following parameters: (i) 1 as number of allowed missed cleavages, (ii) methionine oxidation as variable modification, (iii) 75 ppm as peptide tolerance. Only significant hits were considered, as defined by the MASCOT scoring and probability system.

(53) Reactogenicity Assays

(54) For luciferase assay HEK293-TLR2 and HEK293-TLR4 cells were seeded into microclear 96-well bottom plates in 90 l of complete medium in absence of selection antibiotics. After overnight incubation, cells were stimulated in duplicates with different concentration of OMVs (10 l/well) starting from 1 mg/ml diluted 1:2 in PBS, for 6 h. Then the medium were discarded and cells were lysed with 20 l of Passive Lysis Buffer (Promega) for 20 min at room temperature. Luciferase levels were measured by addition of 100 l/well Luciferase Assay Substrate (Promega) using LMax II384 microplate reader (Molecular Devices). Raw light units (RLU) from each sample were divided by the RLU of the control sample (PBS) and expressed as Fold Induction (FI).

(55) PBMCs (Pheripheral Blood Mononuclear Cells) were isolated from buffy coats of healthy donors using Ficoll (Amersham Biosciences) density gradient centrifugation. Cells were seeded into microclear %-well bottom plates in 180 l of RPMI (GIBCO) supplemented with 10% of heat-inactivated FBS, 100 U/ml pelicillin, 100 g/ml streptomycin, 2 mM glutamine. Cells were stimulated with different concentration of OMVs (20 l/well) starting from 1 mg/ml diluted 1:2 in PBS, for overnight. Mesoscale Assay Human-Proinflammatory 7-spot (MSD Technology) is used for detection of inflammatory cytokines following manufacturer's instructions.

(56) Results

(57) NT013, NT017 and NT022 of NTHi show a significant homology with a number of previously characterized LytM-proteins expressed by other Gram-negative bacteria. In particular, E. coli proteins known to be involved in the cell division process such as YebA, EnvC and NlpD show an amino acid identity of 49% with NT013, 40% with NT017 and 43% with NT022, respectively. LytM catalytic domains are the most conserved regions between NTHi and E. coli proteins, in fact, the homology percentage grows up to 79% for this specific domain.

(58) LytM Proteins are Differently Distributed on NTHi Compartments

(59) In order to verify expression and the subcellular localization of LytM proteins NT013, NT017 and NT022, single deletion mutants of the genes codifying for the three LytM proteins were generated in NTHi176 strain. Immunoblotting with specific antisera raised against each of the LytM recombinant proteins was performed to determine the level of expression in periplasmic, outer membrane and supernatant fractions.

(60) As shown in FIG. 1A, NT013 was detected in the outer membrane protein extracts, NT017 in the periplasmic fraction, while NT022 was found in all fractions. As a control, none of the antisera recognized specific bands at 53 kDa, 46 kDa and 42.5 kDa corresponding to NT013. NT017 and NT022 in cell preparations from the respective knockout mutant strains (FIG. 1A).

(61) Surprisingly, confocal immunofluorescence microscopy of bacteria stained for NT013 revealed no specific signal of the protein on the bacterial surface, indicating that NT013 could be associated to the inner layer of the outer membrane as it was found to be present in the outer membrane fraction by western blot analysis. As expected, NT017, which was found in the periplasmatic fraction, was negative by confocal microscopy analysis and NT022 was confirmed to be exposed on the bacterial surface (FIG. 1B). Of interest, it appears that the antigen is translocated on the bacterial surface at specific foci close to the division septum (FIG. 1B). A model of the protein localisation is illustrated in FIG. 1C.

(62) 176NT013 and 176NT022 Exhibit Aberrant Cell Morphology and Severe Cell Separation Defects

(63) To investigate whether NTHi LytM proteins, NT013 and NT022, played a role in cell separation, single isogenic knockout mutants (176NT013 and 176NT022) were cultured on solid or liquid medium and compared to the wild type strain. These showed no differences in colony morphology as visualized by light microscopy. To evaluate the effect of the mutations, knockout strains were grown in liquid BHI at 37 C. There were no significant differences in the growth rate of WT and mutants, but a phenotype of aggregation was observed in liquid cultures for 176NT013 and 176NT022 (FIG. 2A), and was confirmed by confocal imaging (FIG. 2C). The number of bacterial colonies was also measured at two different OD by plating cultures serial dilution on agar chocolate plates. Colony forming units (CFU) per millilitre of 176NT013 and 176NT022 was much lower than the wild type Hi176 and of 176NT017, indeed CFU derived from 176NT013 and 176NT022 are respectively only about 10% and 1% with respect to the parent strain (FIG. 2B).

(64) To verify whether the bacterial aggregation phenotype was due to a failure in cell separation, we used confocal and scanning electron microscopy which clearly showed that 176NT013 and 176NT022 mutants differ from the wild type in dimension and morphology (FIGS. 3A and 3B).

(65) In particular, 176NT013 cells appeared roughly four times longer than the wild type strain and are bended in the central portion. On the other hand, 176NT022 mutant forms longer chains (up to 0.1 mm), while no evident morphological differences were observed for 176NT017. The same phenotype was observed when LytM mutants were generated in a different strain (Hi162), indicating the ubiquitous functional properties of such determinants.

(66) This result indicates that NT013 and NT022 are involved in bacterial separation, although they are not essential for NTHi cell growth.

(67) 176NT013 and 176NT022 mutants release outer membrane vesicles

(68) Defective septum formation in 176NT013 and 176NT022 mutants was observed in transmission electron microscopy (TEM) of bacterial cell surfaces (FIG. 4). Moreover, an exclusive formation of blebs on the surface of both these mutants was highlighted (FIG. 5).

(69) This membrane blebbing is believed to be due to an overproduction of Outer Membrane Vesicles (OMVs) which are naturally secreted by Gram negative bacteria. Native OMVs were purified from these two mutants and from the Hil76 strain to verify the quality and to quantify OMV release. Isolation of OMVs revealed that both mutant strains produce more vesicles with respect to the wild type strain (FIG. 5). TEM analysis of OMV preparations confirmed the presence of vesicles with an apparent diameter of 20 to 100 nm (FIG. 5). OMV overproduction in LytM mutants was quantified and showed a fourfold increase with respect to WT (Lowry Method for protein quantitation).

(70) Protein Composition of Outer Membrane Vesicles

(71) To compare the protein composition of the OMVs extracted from each strain, samples were run on a SDS-page gel and a Coomassie blue staining was performed (FIG. 6A). Protein patterns were similar between wild type and mutant strains, although a few bands showed a different intensity. As expected, mass spectrometry analysis associated these bands to a number of known surface determinants, including HMW1 and 2, HtrA, P2, P5 and OMP26 (FIG. 6B). Comparisons were made with the known TolR mutant as well as wild type.

(72) OMVs purified from the mutants strains were analysed for their protein content and compared to OMVs purified from the wild type strain on the basis of mass spectroscopy of selected bands.

(73) Haemophilus proteins were grouped into five classes according to their localisation within the cell, and the amount of protein from each class present in OMVs was measured. Differences were observed between wild type and mutants and this is suggestive of distinct mechanisms of OMV formation (see FIG. 7). NT013 and NT022 mutant OMVs displayed increased amounts of lipoproteins and reduced amounts of outer membrane proteins. The amounts of individual proteins were quantified within each class. The results for the lipoprotein, periplasmic protein and outer membrane protein classes are illustrated in FIG. 8, FIG. 9 and FIG. 10 respectively.

(74) One difference is a decrease in the quantity of the outer membrane immunodominant proteins (protein 2 and protein 5) in mutant strains (see FIG. 10 and Table below). In particular in wild type the quantity of these P2 and P5 is about 56% of total, in NT013 it decreases to 31% and in NT022 to 23%. P2 and P5 are very abundant and variable proteins and their presence is one of the key reasons for the lack of cross reactivity in the sera generated against haemophilus OMVs. The mutant OMVs with reduced P2 and P5 concentration may therefore be used to improve cross reactivity.

(75) TABLE-US-00004 176 WT TolR NT013 NT022 Total proteins (ng) 101.546 100.127 101.567 102.663 P2 40.2 17.29 13.0889 11.129 P5 16.69 25.488 18.41 12.169 Total proteins (P2 + P5) 44.62 57.34 70.06 79.36 % P2-P5 56.06 42.73 31.02 22.7 % rest of OMP 43.94 57.27 68.98 77.3

(76) A summary of the proteomic data relating to virulence factors in the NT017 mutant is set out below:

(77) TABLE-US-00005 Decrease >50% Similar Increase <50% wt ko wt ko wt ko HtrA 8.281 1.559 D15 1.429 1.529 Protein 5 16.693 23.523 Omp26 3.441 2.202 Protein 6 0.720 0.864 Iga1 protease 0 9.538 NT067 1.870 0.628 phosphate binding 0.760 0.880 long chain fatty 0.648 1.670 Protein E 0.234 0 periplasmic protein acid ABC HMW2B 1.187 0.853 PstS transporter NT018 0.701 0.575 HxuA 0 0.697 NT022 0.336 0.421 opacity protein 0.469 1.106 SurA 0.490 0.370

(78) Immunological studies were performed on OMVs from wild types and mutants to determine if the differences observed in OMVs protein patterns could influence TLRs activation by LPS or lipoprotein components. HEK293-hTLR2 and HEK293-hTLR4/CD14-MD2 cells were stimulated with different dilutions of OMVs from the wild type and knockout strains. No significant differences were detected (FIG. 6C-6D).

(79) Moreover the same stimulation was extended to human Peripheral Blood Mononuclear Cells (hPBMCs) to measure proinflammatory cytokines production. No significant differences were observed (FIG. 6E-6F).

(80) LytM Protein Knock Out in Neissena meningitidis

(81) Three LytM proteins were identified in Neisseria meningitidis MC58 and are: NMB1483, NMB0315 and NMB1333. NMB1483 contains two LysM domains and a M23 peptidase family domain, and is the NlpD homologue found in Neisseria meningitidis group B. NMB0315 is also a lysostaphin-type zinc-dependent metallopeptidase belonging to the M23 peptidase family and is characterized by a conserved active site containing an HXH motif. NMB1333 also presents a conserved domain typical of the M23 peptidase family. The gene and protein sequences for the LytM proteins were identified in Neisseria meningitidis are set out in SEQ ID NOs: 37-42.

(82) TABLE-US-00006 NMB1483 nucleic acid SEQ ID NO: 37 NMB0315 nucleic acid SEQ ID NO: 38 NMB1333 nucleic acid SEQ ID NO: 39 NMB1483 protein SEQ ID NO: 40 NMB0315 protein SEQ ID NO: 41 NMB1333 protein SEQ ID NO: 42

(83) A knockout mutant was generated in strain MC58 of Neisseria meningitidis for gene NMB1483 (NlpD) and it was named: MC58148. A detailed description of the generation of knock out mutants is described elsewhere [123].

(84) Flanking regions to the coding sequence of the gene were amplified using the following sets of primers:

(85) TABLE-US-00007 Up1483_Fw1 gctctagaCGTTACAGCGGCAATTATTGC XbaI SEQID NO:43 Up1483_Rv tcccccgggCGCAGACAGTACAGATAGTAC SmaI SEQID NO:44 Dn1483_Fw tcccccgggATGTTCCGATATATAGCCTG SmaI SEQID NO:45 Dn1483_Rv2 ccgctcgaCCCCTATTTTGTGGAACATC XhoI SEQID NO:46

(86) The plasmid used for generation of the deletion mutant in MC58 was: pBS-UD1483_Ery. By chromosomal allelic exchange, the gene was substituted by an erythromycin resistance cassette.

(87) The MC581483 strain was analysed by confocal microscopy analysis and showed the presence of multiple aggregates of variable size, with respect to the WT strain. For confocal microscopy analysis bacteria were grown in GC medium until exponential phase (OD.sub.600 0.5) and fixed in 4% paraformaldehyde, before DAPI staining (See FIG. 12).

(88) To further characterize the bacterial cell morphology TEM (transmission electron microscopy) and SEM (scanning electron microscopy) analysis was performed. For TEM analysis, bacteria were also grown in GC medium until exponential phase (OD.sub.600 0.5). TEM analysis confirmed the presence of bacterial aggregates and aberrant cell morphology was shown in MC581483, compared to the diplococcic observed in the WT strain. Moreover, the presence of vesicles was clearly visible in the mutant strain (See FIG. 13A).

(89) For SEM analysis, bacteria were also grown in GC medium until exponential phase (ODM.sub.600 0.5). Also SEM analysis confirmed the presence of three-dimensional bacterial aggregates and of aberrant cell morphology in the MC581483 mutant (See FIG. 13B).

(90) The MC581483 mutant was also tested for its ability to produce OMVs. In a first experiment, strains MC58 and MC58483 were grown till stationary phase (OD 1.3-1.5) in 50 ml MCDM I (Meningitis chemically defined medium 1), in 250 ml shaker flasks incubated overnight at 37 C., 5% CO.sub.2 and 185 rpm.

(91) For OMVs isolation, the cultures were centrifuged at 3500 rpm for 30 min at 4 C. and supernatants were filtered using Stericup filter bottles (0.22 m pore size).

(92) Samples were then centrifuged at 35,000 rpm (96,000g, average) at 4 C. for 2 h, then washed with PBS and centrifuged again at 35,000 rpm (96,000g, average) at 4 C. for 2 h. After removal of supernatant, the pellet was resuspended in 200 to 500 l PBS.

(93) For checking the quality of the preparation and compare the amount of vesicles produced by the MC58 WT and MC581483 mutant strains, the same volume of OMVs was loaded for SDS-PAGE analysis and proteins were stained with Coomassie blue. The results showed a different protein pattern in the MC581483 mutant, compared to the WT strain. Total protein quantification by Lowry assay showed a two-fold increase in the production of OMVs from the mutant, compared to the WT strain (FIG. 14A) In a second experiment, strains MC58 and MC581483 were grown till exponential phase (OD.sub.600 0.5) in 40 ml GC medium, in 250 ml shaker flasks incubated at 37 C., 5% CO2 and 185 rpm. For OMVs isolation, the same protocol was followed.

(94) The results from SDS-PAGE analysis showed OMV proteins only from the MC581483 mutant. No proteins were detected when the same volume of OMVs from the WT strain was loaded. Also quantification by Lowry assay confirmed the absence of OMVs in the preparation from the WT strain (FIG. 14B).

(95) Other OMV preparations are made and a mass-spectrometry proteomic analysis of the sample is performed at exponential and stationary growth phases. The presence of the main 4CMenB vaccine antigens is evaluated and compared.

(96) LytM Protein Knock Out in Bordetella pertussis

(97) Six putative peptidases were identified in Bordetella pertussis Tohama I: BP1721, BP2956, BP0608, BP2919, BP3015 and BP1017. BP1721 and BP2919 have both a Lys-M domain (involved in binding to peptidoglygan) and a Lyt-M domain (Lysostaphin-type metallopeptidases) like NlpD from E. coli and NT022 from NTHi, the other four proteins have only the Lyt-M domain. The locus organization does not help to discriminate the different homologues, with the exception of BP1721 which can be clearly identified as the NlpD homologue in B.pertussis. Multiple alignments showed a high conservation of the peptidase catalytic site between BP2956 and NT013, and between BP0608 and NT017. Gene and protein sequences are shown in SEQ ID NO:s 47-58. BP1721 is a NlpD homologue. BP2956 is a putative NT013/YebA homologue. BP0608 is a putative NT017/EnvC homologue.

(98) TABLE-US-00008 Nucleic acid sequence Protein sequence BP1721 SEQ ID NO: 47 SEQ ID NO: 53 BP2956 SEQ ID NO: 48 SEQ ID NO: 54 BP0608 SEQ ID NO: 49 SEQ ID NO: 55 BP2919 SEQ ID NO: 50 SEQ ID NO: 56 BP3015 SEQ ID NO: 51 SEQ ID NO: 57 BP1017 SEQ ID NO: 52 SEQ ID NO: 58

(99) A knockout mutant was generated in strain W28 9K-129G of Bordetella pertussis for gene BP1721 (NlpD homologue). Flanking regions to the coding sequence of the gene were amplified using the following sets of primers:

(100) TABLE-US-00009 BP1721 ccgGAATTCGCGGTTGCGCGCGCAGGGCAT SEQID 5For NO:59 BP1721 ggaGGATCCACGATTCTCCTGTTTGCTCAA SEQID 5Rev NO:60 BP1721 ggaGGATCCCGCCCACGCTCGTTTTCGACC SEQID 3For NO:61 BP1721 cccAAGCTTCCACGTCGGTCTCGCAGTACG SEQID 3Rev NO:62

(101) Deletion of the gene BP1721 was obtained as follows:

(102) A kanamycin resistance cassette was cloned between the BP1721 flanking regions into the suicide vector pSORTP1. The pSORTP1-BP1721KO construct was introduced into B. pertussis by conjugation. Integration of the plasmid into the chromosome following the first crossing-over event was selected for gentamicin resistance (present on the plasmid backbone) and kanamycin resistance (present in the plasmid insert). Loss of the plasmid following the second crossing-over event and replacement of the BP1721 gene with the kanamycin cassette was selected for streptomycin resistance (the plasmid confers sensitivity to streptomycin) and kanamycin resistance. The replacement of the BP1721 gene with the kanamycin cassette was confirmed by PCR amplification using primers external to the flanking regions of BP1721 and using B. pertussis W29 9K/129G as a control. The expected sizes of the amplification products were 2189 bp for the WT strain and 2574 bp for the KO strain. The primers used were: BP1721 EXT 5 FOR: AACCTGGGCTTGAACTCC (SEQ ID NO:63); BP1721 EXT 3 REV: ACACCAGCCAGGTATTGA (SEQ ID NO:64).

(103) Cell aggregates, which could be due to altered cell division mechanisms, were already visible from the culture when the bacterium was grown in liquid medium. Cell division defects were then subsequently confirmed by confocal microscopy analysis. For the microscopy analysis 50 l of a liquid culture were fixed with 4% paraformaldehyde and stained with DAPI to visualize DNA. The KO strain clearly exhibited a severe cell chaining phenotype, compared to the WT strain. (See FIG. 15).

(104) Strains Bp W28 9G/129K and Bp W28 9G/129K NlpD were grown until exponential phase (OD.sub.600 4.3-4.8) in 40 ml of Stainer-Scholte broth (supplemented with 400 g/ml streptomycin) in 250 ml shaker flasks and incubated at 35 C. and 180 rpm.

(105) For OMV isolation, the cultures were centrifuged at 5000g for 45 min at 4 C. and supernatants were filtered using Stericup filter bottles (0.22 m pore size). Samples were then centrifuged at 96000g at 4 C. for 2 h, then washed with PBS and centrifuged again at 96000g at 4 C. for 2 h. After removal of supernatant, the pellet was resuspended in 200 l PBS.

(106) To check the quality of the preparation and compare the amount of vesicles produced by the Bp W28 9G/129K and Bp W28 9G/129K NlpD strains, the same volume of OMVs (10 l) was loaded for SDS-PAGE analysis and proteins were stained with Simply Blue SafeStain (Life Technologies). The results showed a different protein pattern in the mutant, compared to the WT strain (see FIG. 19).

(107) A knockout mutant is generated in strain W28 9K-129G of Bordetella pertussis for gene BP2956 (putative NT013/YebA homologue). Flanking regions to the coding sequence of the gene (bold) are amplified using the following sets of primers:

(108) TABLE-US-00010 BP2956 ccgGAATTCATCAAGAAGCTGGGACGT SEQIDNO:65 5For BP2956 ggaGGATCCAACTTTGCGTTTGAAGCT SEQIDNO:66 5Rev BP2956 ggaGGATCCCAAGCAGCAGATCAAGCT SEQIDNO:67 3For BP2956 cccAAGCTTGTCGGCGTCGTAAGGCTG SEQIDNO:68 3Rev

(109) Deletion of the gene BP2956 is obtained as follows. A cloramphenicol resistance cassette is cloned between the BP2956 flanking regions into the suicide vector pSORTP1. The pSORTP1-BP2956.sub.KO construct is introduced into B. pertussis by conjugation. Integration of the plasmid into the chromosome following the first crossing-over event is selected for gentamicin resistance (present on the plasmid backbone) and cloramphenicol resistance (present in the plasmid insert). Loss of the plasmid following the second crossing-over event and replacement of the BP2956 gene with the kanamycin cassette is selected for streptomycin resistance (the plasmid confers sensitivity to streptomycin) and cloramphenicol resistance.

(110) These deletion mutant strains are analyzed for their cell morphology, OMVs production and proteomic characterization. In particular, the presence of B. pertussis most immunogenic antigens (PT. FHA and 69K) is evaluated. Finally, these mutants are analyzed in vitro for their function in B. pertussis physiology and/or pathogenesis and the OMVs produced are tested in immunogenicity and cross-reactivity studies.

(111) Knock-out mutants for the other putative LytM protein homologues: flanking regions to the coding sequence of the gene (bold) are amplified using the following sets of primers:

(112) TABLE-US-00011 BP0608 ccgGAATTCTGGAAAACCGTTTCACGG SEQIDNO:69 5For BP0608 ggaGGATCCGAATCAGTCCTTTTTCGC SEQIDNO:70 5Rev BP0608 ggaGGATCCATAAATATCGGGAAGTGT SEQIDNO:71 3For BP0608 cccAAGCTTCCGAGTTCCTTCAGATGG SEQIDNO:72 3Rev BP2919 ccgGAATTCCATGATGCCGACTTGCAT SEQIDNO:73 5For BP2919 ggaGGATCCTGAAAGAGGCAGCAAAAC SEQIDNO:74 5Rev BP2919 ggaGGATCCCCGGCGAAACAGCACGTA SEQIDNO:75 3For BP2919 cccAAGCTTAGTTCGAAGCTGGCATTG SEQIDNO:76 3Rev BP3015 ccgGAATTCTTGCCGATATCGGTTTTC SEQIDNO:77 5For BP3015 ggaGGATCCTTGCATCCTGTTATTTGA SEQIDNO:78 5Rev BP3015 ggaGGATCCGTTAAACTGGATCGTTTC SEQIDNO:79 3For BP3015 cccAAGCTTTCGAAGCCGAATTCGTTA SEQIDNO:80 3Rev BP1017 ccgGAATTCAGCAGATGCGCCAGATCA SEQIDNO:81 5For BP1017 ggaGGATCCTTGCGTCGGTCTTGCCCT SEQIDNO:82 5Rev BP1017 ggaGGATCCTTCGGCGTATTGCAGTTC SEQIDNO:83 3For BP1017 cccAAGCTTTGAGTACCTGCCTATCGT SEQIDNO:84 3Rev

(113) It will be understood that the invention has been described by way of example only and modifications may be made whilst remaining within the scope and spirit of the invention. [1] WO2006/046143. [2] Berlanda Scorza et al. (2008) Mol Cell Proteomics 7:473-85. [3] European patent 0011243. [4] Fredriksen et al. (1991) NIPH Ann. 14(2):67-80. [5] WO2004/019977. [6] Hozbor et al. (1999) Curr Microbiol 38:273-8. [7] Tan et al. (2010) N Engl J Med. 362(16):1511-20. [8] Koeberling (2008) J Infect Dis. 198(2):262-70. [9] Uehara T. et al., The EMBO Journal (2010) 29, 1412-1422 [10] Bernadac A., et al. (1998) Journal of Bacteriology 180 (18): 4872-4878 [11] WO2002/062378 [12] Rawlings et al (2006) Nucleic Acids Res. January 1; 34:D270-2 [13] Kessler et al In A. J. Barrett. N. D. Rawlings, and J. F. Woessner (ed.). Handbook of proteolytic enzymes 1998, p. 1476-1478 1st ed. Academic Press, London, United Kingdom [14] Hogg et al (2007) Genome Biology 8:R103. [15] Needleman & Wunsch (1970) J. Mol. Biol. 48, 443-453. [16] Rice et al. (2000) Trends Genet 16:276-277. [17] WO02/09746. [18] WO01/09350. [19] WO02/062378. [20] WO2004/014417. [21] Heterologous Gene Expression in E. coli: Methods and Protocols (Method in Moleclar Biology) (eds. Evans & Xu) Humana Press 2010 (ISBN 161737%62). [22] WO2010/010983 [23] WO2011/161551 [24] Anogv et al. (2008) PLoS ONE 3(5):e2189 [25] WO02/09643. [26] Katial et al. (2002) Infect. Immun. 70:702-707. [27] WO2011/036562 [28] WO2004/019977. [29] European patent 0011243. [30] Fredriksen et al. (1991) NIPH Ann. 14(2):67-80. [31] WO01/91788. [32] WO2005/004908. [33] Gennaro (2000) Remington: The Science and Practice of Pharmacy. 20th edition, ISBN: 0683306472. [34] Vaccine Design (1995) eds. Powell & Newman. ISBN: 030644867X. Plenum. [35] WO90/14837. [36] Podda & Del Giudice (2003) Expert Rev Vaccines 2:197-203. [37] Podda (2001) Vaccine 19: 2673-2680. [38] Vaccine Adjuvants: Preparation Methods and Research Protocols (Volume 42 of Methods in Molecular Medicine series). ISBN: 1-59259-083-7. Ed. O'Hagan. [39]U.S. Pat. No. 5,057,540. [40] Niikura et al. (2002) Virology 293:273-280. [41] Lenz et al. (2001) J Immunol 166:5346-5355. [42] Pinto et al. (2003) J Infect Dis 188:327-338. [43] Gerber et al. (2001) J Virol 75:4752-4760. [44] WO03/024480. [45] WO03/024481. [46] Gluck et al. (2002) Vaccine 20:B10-B16. [47] Meraldi et al. (2003) Vaccine 21:2485-2491. [48] Pajak et al. (2003) Vaccine 21:836-842. [49] Krieg (2003) Nature Medicine 9:831-835. [50] McCluskie et a. (2002) FEMS Immunology and Medical Microbiology 32:179-185. [51] WO98/40100. [52] U.S. Pat. No. 6,207,646. [53] U.S. Pat. No. 6,239,116. [54] U.S. Pat. No. 6,429,199. [55] Schellack et al. (2006) Vaccine 24:5461-72. [56] Johnson et a. (1999) Bioorg Med Chem Lett 9:2273-2278. [57] Evans et al. (2003) Expert Rev Vaccines 2:219-229. [58] Beignon et al. (2002) Infect Immun 70:3012-3019. [59] Pizza et al. (2001) Vaccine 19:2534-2541. [60] Pizza et al. (2000) Int J Med Microbiol 290:455-461. [61] Scharton-Kersten et al. (2000) Infect Immun 68:5306-5313. [62] Ryan et al. (1999) Infect Immun 67:6270-6280. [63] Partidos et a. (1999) Immunol Lett 67:209-216. [64] Peppoloni et a. (2003) Expert Rev Vaccines 2:285-293. [65] Pine et al. (2002) J Control Release 85:263-270. [66] WO99/40936. [67] WO99/44636. [68] Singh et al. (2001) J Cont Release 70:267-276. [69] WO99/27960. [70] U.S. Pat. No. 6,090,406. [71] U.S. Pat. No. 5,916,588. [72] EP-A-0626169. [73] WO99/52549. [74] Andrianov et al. (1998) Biomaterials 19:109-115. [75] Payne et al. (1998) Adv Drug Delivery Review 31:185-196. [76] Stanley (2002) Clin Exp Dermatol 27:571-577. [77] Jones (2003) Curr Opin Investig Drugs 4:214-218. [78] WO99/11241. [79] WO94/00153. [80] WO98/57659. [81] European patent applications 0835318, 0735898 and 0761231. [82] Ogunniyi et al. (2001) Infect Immun 69:5997-6003. [83] WO2006/110603. [84] Mason et al. (2003) Infect Immun 71:3454-3462. [85] Zwijnenburg et al. (2001) J Infect Dis 183:1143-6. [86] Cheeseman M. T. et al. (2011) PLoS Genetics 7 (10): e1002336. [87] Watson (2000) Pediatr Infect Dis J 19:331-332. [88] Rubin (2000) Pediatr Clin North Am 47:269-285, v. [89] Jedrzejas (2001) Microbiol Mol Biol Rev 65:187-207. [90] Bell (2000) Pediatr Infect Dis J 19:1187-1188. [91] Iwarson (1995) APMIS 103:321-326. [92] Gerlich et al. (1990) Vaccine 8 Suppl:S63-68 & 79-80. [93] Vaccines (1988) eds. Plotkin & Mortimer. ISBN 0-7216-1946-0. [94] Del Guidice et al. (1998) Molecular Aspects of Medicine 19:1-70. [95] Gustafsson et al. (1996) N. Engl. J. Med. 334:349-355. [96] Rappuoli et al. (1991) TIBTECH 9:232-238. [97] Costantino et al. (1999) Vaccine 17:1251-1263. [98] Sutter et al. (2000) Pediatr Clin North Am 47:287-308. [99] Zimmerman & Spann (1999) Am Fam Physician 59:113-118, 125-126. [100] McMichael (2000) Vaccine 19 Suppl 1:5101-107. [101] Schuchat (1999) Lancet 353(9146):51-6. [102] WO02/34771. [103] Dale (1999) Infect Dis Clin North Am 13:227-43, viii. [104] Ferretti et al. (2001) PNAS USA 98: 4658-4663. [105] Kuroda et al. (2001) Lancet 357(9264):1225-1240; see also pages 1218-1219. [106] EP-A-0372501 [107] EP-A-0378881 [108] EP-A-0427347 [109] WO93/17712 [110] WO94/03208 [111] WO98/58668 [112] EP-A-0471177 [113] WO00/56360 [114] WO91/01146 [115] WO00/61761 [116] WO01/72337 [117] Research Disclosure, 453077 (January 2002) [118] Current Protocols in Molecular Biology (F. M. Ausubel et al., eds., 1987) Supplement 30. [119] Smith & Waterman (1981) Adv. Appl. Math. 2:482-489. [120] Klock et al (2005) J. Struct. Funct. Genomics, 6:89-94 [121] Herriott, et al (1970) J. Bacteriol., 101:517-524. [122] Carlone et al (1986) J Clin Microbiol. 24(3): 330-332. [123] Echenique-Rivera H, et al (2011) PLoS Pathog. May; 7(5)