Title of Invention

A GENETICALLY-ENGINEERED BLEB PREPARATION AND A VACCINE COMPRISING THE SAME

Abstract

The present invention relates to a genetically-engineered bleb preparation isolated from a modified Neisseria meningitidis strain characterized in that said preparation is obtainable by employing the following process: a process of reducing irnmunodominant variable or non-protective antigens within the bleb preparation comprising the steps of engineering a bacterial strain to produce less or none of the PorA antigen, and making blebs from said strain.

Full Text

The present invention relates to a genetically-engineered bleb preparation and a vaccine comprising the same. FIELD OF THE INVENTION The present invention relates to the field of Gram-negative bacterial vaccine compositions, their manufacture, and the use of such compositions in medicine, More particularly it relates to the field of novel outer-membrane vesicle (or bleb) vaccines, and advantageous methods of rendering these vaccines more effective and safer. BACKGROUND OF THE INVENTION Gram-negative bacteria are separated from the external medium by wo successive layers of membrane structures. These structures, referred to as the cytoplasmic membrane and the outer membrane (OM), differ both structurally and functionally. The outer membrane plays an important role in the interaction of pathogenic bacteria with their respective hosts. Consequently, the surface exposed bacterial molecules represent important targets for the host immune response, making outer-membrane components attractive candidates in providing vaccine, diagnostic and therapeutics reagents. Whole cell bacterial vaccines (killed or attenuated) have the advantage of supplying multiple antigens in their natural micro-environment. Drawbacks around this approach are the side effects induced by bacterial components such as endotoxin and peptidoglycan fragments. On the other band, acellular subunit vaccines containing purified components from the outer membrane may supply only limited protection and may not present the antigens properly to the immune system of the host. Proteins, phospholipids and iipopolysaccharides are the three major constituents found in the outer-membrane of all Gram-negative bacteria. These molecules are distributed asymmetrically: membrane phospholipids (mostly in the inner leaflet), lipooligosaccharides (exclusively in the outer leaflet) and proteins (inner and outer leaflet lipoproteins, integral or polytopic membrane proteins). For many bacterial pathogens which impact on human health, lipopolysaccharide and outer-membrane proteins have been shown to be immunogenic and amenable to confer protection against the corresponding disease by way of immunization. The OM of Gram-negative bacteria is dynamic and, depending on the environmental conditions, can undergo drastic morphological transformations. Among these manifestations, the formation of outer-membrane vesicles or "blebs" has been studied and documented in many Gram-negative bacteria (Zhou, L et at. 1998. FEMS Microbiol. Lett. 163-. 223-22S). Among these, a non-exhaustive list of bacterial pathogens reported to produce blebs include: Bordetello pertussis, Borrelia burgdorferi. Brucella melitensis, Brucella avis, Chlamydia psittaci. Chlamydia trachomatis, Esherichia coli, Haemophilus influenzae, Legionella pneumophila, Neisseria gonorrhoeae, Neisseria meningitidis, Pseudomonas aeruginosa and Yersinia enterocolixica. Although the biochemical mechanism responsible for the production of OM blebs is not fully understood, these outer membrane vesicles have been extensively studied as they represent a powerful methodology in order to isolate outer-membrane protein preparations in their native conformation. In that context, the use of outer-membrane preparations is of particular interest to develop vaccines against Neisseria. Moraxella catarrhalis, Haemophilus influenzae, Pseudomonas aeruginosa and Chlamydia. Moreover, outer membrane blebs combine multiple protcinaceaous and non-proteinaceous antigens that are likely to confer extended protection against intra-species variants. in comparison with the other, more widely used, types of bacterial vaccine (whole cell bacterial and purified subunii vaccines), the inventors will show that outer membrane bleb vaccines (if modified in certain ways) may represent the ideal compromise. The wide-spread use of bacterial subunit vaccines has been due to the intensive study of bacterial surface proteins that have been found to be useful in vaccine applications [for instance B. pertussis pertactin]. These proteins are loosely associated with the bacterial outer membrane and can be purified from culture supernatant or easily extracted from the bacterial cells. However it has also been shown that structural, integral outer membrane proteins are also protective antigens. Examples are PorA for A', meningitidis serogroup B; D15 for H. influenzae; OMP CD for M. catarrhalis: OMP F for P. Aeruginosa. Such proteins however have rather specific structural features, particularly multiple amphipathic p-sheets, which complicates their straightforward use as purified (recombinant) subunit vaccines. In addition, it has become clear that multiple component vaccines are needed (in terms of bacterial surface proteins and integral membrane proteins) to supply a reasonable level of protection. For instance, in the case of B. pertussis subunit vaccines multicomponent vaccines are superior to mono or bicomponent products. In order to incorporate integral outer-membrane proteins into such a subunit product, native (or near-native) conformational folding of the proteins must be present in trie product in order to have a useful immunological effect. The use of excreted outer membrane vesicles or blebs may be an elegant solution to the problem of including protective integral membrane proteins into a subunit vaccine whilst still ensuring that they fold properly. ,V. meningitidis serogroup E (meriB) excretes outer membrane blebs in sufficient quantities to allow their manufacture on an industrial scale. Such multicomponent outer-membrane protein vaccines from naturally-occurring menB strains h?vc b~?n found to be efficacious in protecting teenagers from menB disease and have become registered in Latin America. An alternative method of preparing outer-membrane vesicles is via the process of detergent extraction of the bacterial cells (EP 11243) Examples of bacteria! species from which bleb vaccines can be made are the following. Neisseria meningitidis: Neisseria meningitidis (meningococcus) is a Gram-negative bacterium frequently isolated from the human upper respiratory tract. It occasionally causes invasive bacterial diseases such as bacteremia and meningitis. The incidence of meningococcal disease shows geographical seasonal and annua! differences (Schwartz. B., Moore, P.S., Broome, C.V.; Clin. Microbiol. Rev. 2 (Supplement), S18-S24, 1989). Most disease in temperate countries is due to strains of serogroup B and varies in incidence from 1-10'100,000/year total population sometimes reaching higher values (Kaczmarski, E.B. (1997). Commun. Dis. Rep. Rev. 7: R55-9, 1995; Scholten, R.J.P.M., Bijlmer, H.A., Poolman, J.T. et al. Clin. Infect. Dis. 16: 237-246, 1993; Cruz, C, Pavez, G., Aguilar. E.. e: al. Epidemiol. Infect. 105: 119-126, 1990). Age-specific incidences in the two high risk-groups, infants and teenagers, reach higher levels-Epidemics dominated by serogroup A meningococci occur, mostly in central Africa, sometimes reaching levels up to 1000/100,000/year (Schwartz, B., Moore, P.S., Broome, C.V. Clin. Microbiol. Rev, 2 (Supplement), SI8-S24, 1989). Nearly all cases of meningococcal disease as a whole are caused by serogroup A, B, C, W-135 and Y meningococci, A tetravalent A, C, W-135, Y capsular polysaccharide vaccine is available (Armand, J., Arminjon, F., Mynard, M.C., Lafaix, C, J. Biol. Stand. 10: 335-339, 1982). The polysaccharide vaccines are currently being improved by way of chemically conjugating them to carrier proteins (Lieberman, J.M., Chiu, S.S., Wong, V.K., et al. JAMA 275 : 1499-1503, 1996). A serogroup B vaccine is not available, since the B capsular polysaccharide is non-immunogenic, most likely because it shares strucniral similarity to host components (Wyle, F.A., Artenstein, M.S., Brandt, M.L. et al. J. Infect. Dis, 126: 514-522, 1972; Finne, J.M., Leioonen, M., Makela, P.M. Lancet ii.: 355-357, 1983). For many years efforts have been focused on developing meningococcal outer membrane based vaccines (de Moraes, J.C., Perkins, B.; Camargo, M.C. et al. Lancet 340: 1074-1078, 1992; Bjune, G., Hoiby, E.A. Gronnesby, J.K. et al. 338: 1093-1096, 1991). Such vaccines have demonstrated efficacies from 57% - 85% in older children (>4 years) and adolescents. Most of these efficacy trials were performed with OMV (outer membrane vesicles, derived by LPS depletion from blebs) vaccines derived from wild-type A', meningitidis B strains. Many bacterial outer membrane components are present in these vaccines, such as PorA, PoiB, Rmp, Ope, Opa, FrpB and the contribution of these components to the observed protection still needs further definition. Other bacterial outer membrane components have been defined (using animal or human antibodies) as potentially being relevant to the induction of protective immunity, such as TbpB, NspA (Martin, D., Cadieux, K., Hamel, J., Brodeux, B.R., J. Exp, Med. 185: 1173-1183, 1997: Lissoio, L., Maitre-Wilmotte, C., Dumas, p. et al., Inf. Immun, 63: 884-890, 1995). The mechanism of protective immunity wil! involve antibody mediated bactericidal activity and opsonophagocytosis. The frequency of Neisseria meningitidis infections has risen dramatically in the past few decades. This has been attributed to the emergence of multiple antibiotic resistant strains, and increased exposure due to an increase in social activities (for instance swimming pools or theatres). It is no longer uncommon to isolate Neisseria meningitidis strains thai are resistant to some or all of the standard antibiotics. This phenomenon has created an unmet medical need and demand for new anti-microbial agents, vaccines, drug screening methods, and diagnostic tests for this organism. Moraxelia catarrhalis Moraxelia catarrhalis (also named Branhamella catarrhalis) is a Gram-negative bacterium frequently isolated from the human upper respiratory tract. It is responsible for several pathologies, the main ones being otitis msdia in infants and children, and pneumonia the elderly, h is also responsible for sinusitis, nosocomial infections and, less frequently, for invasive diseases. Bactericidal antibodies have been identified in most adults tested (Chapman, AJ et al. (1985) J. Infect.Dis. 151:878). Strains of M. catarrhalis present vananons in their capacity to resist serum bactericidal activity: in general, isolates from diseased individuals are more resistant than those who are simply colonized (Hoi, C et al. (1993) Lancet 341:1281, Jordan, KL et al, (1990) AmJ.Med. 88 (suppl. 5A):28S). Serum resistance could therfore be considered as a virulence factor of me bacteria. An opsonizing activity has been observed in the sera of children recovering from otitis media. The antigens targetted by these different immune responses in humans have not been identified, with the exception ofOMPBl, a 84 kDa protein, the expression of which is regulated by iron, and that is recognized by the sera of patients with pneumonia (Sethi, S, et al. (1995) Infect.Immun. 63:1516), and of UspAl and UspA2 (Chen D. et al.(1999), InfecLlmmun. 67:1310). A few other membrane proteins present on the surface of M. catarrhalis have been characterized using biochemical methods for their potential implication in the induction of a protective immunity (for review, see Murphy, TF (1996) Microbiol.Rev, 60:267). In a mouse pneumonia model, the presence of antibodies raised against some of them (UspA, CODB) favors a faster clearance of the pulmonary infection. Another polypeptide (OMP CD) is highly conserved among hi. catarrhalis strains, and presents homologies with a ponn of Pseudomonas aeruginosa, which has been demonstrated to be efficacious against this bacterium in animal models. M. catarrhalis produces outer membrane vesicles (Blebs). These Blebs have been isolated or extracted by using different methods (Murphy T.F.. Loeb M.R. 1989. Microb. Pathog. 6: 159-174; Unhanand M., Maciver, I., Ramillo 0., Arencibia-Mireles 0., Argyle l.C, McCracken G.H. Jr., Hansen E.J. 1992. J. Infect. Dis. 165:644-650). The protective capacity of such Bleb preparations has been tested in a murine model for pulmonary clearance of W. catarrhalis. It has been shown that active immunization with Bleb vaccine or passive transfer of anti-Blebs antibody induces significant protection in this model (Maciver I., Unhanand M.. McCracken G.H. Jr., Hansen, E.J. 1993. J. Infect. Dis. 168: 469-472). Haemophilus influenzae Haemophilus infv.-.zae is H. influenzae type b (Hib) is clearly different from the other types in that it is a major cause of bacterial meningitis and systemic diseases. Nontypeable strains of H. influenzae (NTHi) are only occasionally isolated from the blood of patients with systemic disease. NTHi is a common cause of pneumonia, exacerbation of chronic bronchitis, sinusitis and otitis media. NTHi strains demonstrate a large variability as identified by clonal analysis, whilst Hib strains as a whole are more homogeneous. Various proteins of H. influenzae have been shown to be involved in pathogenesis or have been shown to confer protection upon vaccination in animal models. Adherence of NTHi to human nasopharygeai epithelial cells has been reported (Read RC. et al. 1991. J. Infect. Dis. 163:549). Apart from fimbriae and pili (Briuton CC. et al. 1989. Pediatr. Infect. Dis. J. 8:S54; Kar S. et al. 1990. Infect. Immun. 58:903; Giidorf JR. et al. 1992. Infect. Immun. 60:374; St. Geme JW st al. 1991. Infect. Immun. 59:3366; St. Geme JW et al. 1993. Infect. Immun. 61: 2233), many adhesins have been identified in NTHi. Among them, two surface exposed high-molecular-weight proteins designated HMW1 and HMW2 have been shown to mediate adhesion of NTHi to epithelial cells (St. Geme JW. et al. 1993. Prac. Natl. Acad. Sci. USA 90:2875). Another family of high-molecular-weigh: proteins has been identified in NTHi strains that lack proteins belonging to HMW1/HMW2 family. The NTHi 115-kDa Hia protein (Barenkamp SJ., St Geme S.W. 1996. Mol. Microbiol In press) is highly similar to the Hsf adhesin expressed by H. influenzae type b strains (St- Geme JW. et al. 1996. J. Bac:. 178:6281). Another protein, the Hap protein shows similarity to IgAl serine proteases and has been shown lo be involved in both adhesion and cell entry (St. Geme JW, et al. 1994. Mol. Microbiol. :4:2P). Five major outer membrane proteins (OMP) have been identified and numerically numbered. Original studies using H.influenzae type b strains showed that antibodies specific for PI and P2 OMPs protected infant rats from subsequent challenge (Loeb MR. et at. 198". Infect. Immun. 55:2612; Musson RS. Jr. et al. 1983. J. Clin. Invest. 72:677). P2 was found to be able to induce bactericidal and opsonic antibodies, which are directed against the variable regions present within surface exposed loop structures of this integral OMP (Haase EM. et al. 1994 Infect. Immun. 62:3712; Troelstra A. et al. 1994 Infect. Irnmun. 62:779). The lipoprotein P4 also may induce bactericidal antibodies (Green BA. et al. 1991. lnfect.Immun.59:3191). OMP P6 is a conserved peptidoglycan associated lipoprotein making up 1-5 % of the outer membrane (Nelson MB. et a!. 1991. Infect. Immun. 59:2658), Later a lipoprotein of about the same molecular weight was recognized called PCP (P6 cross-reactive protein) (Deich RM. et al. 1990. Infect. Immun. 58:3388). A mixture of the conserved lipoproteins P4, P6 and PCP did not reveal protection as measured in a chinchilla otitis-msdia model (Green BA. et al. 1993. Infect-immun. 61:1950). P6 alone appears to induce protection in the chinchilla model (Demana TF. et al. 1996. Infect. Immun. 64:5187). A fimbnn protein (Miyamoto N., Bakaletz, LO. 1996. Microb. Pathog. 21:343) has also been described with homology to OMP P5, which itself has sequence homology to the inteeral Escherichia coil OmpA (Mivamoto N., Bakaletz, LO. 1996. Microb. Pathog. 21:343; Munson RS. Jr. et al. 1993. Infect. Immun. 61:1017). NTHi seem to adhere to mucus by way of fimbriae. In line with the observations made with gonococci and meningococci, NTHi expresses a dual human transferrin receptor composed of TbpA and TbpB when grown under iron limitation. Anti-TbpB antibody protected infant rats (Loosmore SM. et al. 1996. Mol. Microbiol. 19:575). Hemoglobin / haptoglobin receptor also have been described for NTHi (Maciver I. et al. 1996. Infect. Immun. 64:3703). A receptor for HaemiHemopexir has also been identified (Cope LD. et al. 1994. Mol.Microbiol. 13:868). A lactoferrin receptor is also present amongst NTHi, but is not yet characterized (Schryvers AB. et a!. 1989. J. Med. Microbiol. 29:121). A protein similar to neissenal FrpB-protein has not been described amongst NTHi. An 80kDa OMP, die D15 surface antigen, provides protection against NTHi in a mouse challenge model. (Flack FS. et al. 1995. Gene 156:97). A 42kDa outer membrane lipoprotein, LPD is conserved amongst Haemophilus influenzae and induces bactericidal antibodies (Akkoyunlu M. et al. 1996. Infect. Immun. 64:4586). A minor 98kDa OMP (Kimura A. et al. 1985. Infect. Immun. 47:253). was found to be a protective anrigen, this OMP may very well be one of the Fe-limitation inducible OMPs or high molecular weight adhesins that have been characterized thereafter. H. Influenzae produces IgA 1-protease activity (Mulks MH-, Shoberg RJ. 1994. Meth. Enzymol. 235:543). IgA 1-proteases of NTHi have a high degree of antigenic variability (LomholtH., van AlphenL., Kilian, M. 1993. Infect. Immun. 61:4575). Another OMP of NTHi, OMP26, a 26-kDa protein has been shown to enhance pulmonary clearance in a rat model (Kyd, J.M., Cripps, A.W. 1998. Infect. Immun. 66:2272). The NTHi HtrA protein has also been shown to be a protective antigen. Indeed, this protein protected Chinchilla against otitis media and protected infant rats against H. influenzae type b bacteremia (Loosmore S.M. et al. 1998. Infect. Immun. 66:899). Outer membrane vesicles (or biebs) have been isolated from H. influenzae (Loeb M.R., Zachary A.L., Smith D.H. 1981. J. Bacteriol. 145:569-604; Stull T.L., Mack K., Haas J.E.. Smi; J.. Smith AX. 1985. Anal. Biochem.'l50: 471-480). The vesicles have been associated with the induction of blood-brain barrier permeability (WiwpeJwey B.. Hansen E.J., Scheld W.M. 1989 Infect. Immun. 57: 2559-2560), the induction of meningeal inflammation (Mustafa M.M.. Ramilo 0., Syragiannopoulos G.A., Olsen K.D., McCraken G.H. Jr., Hansen, E.J. 1989. J. Infect Dis. 159: 917-922) and to DNA uptake (Concino M.F., Goodgal S.H. 1982 J. Bacterid. 152: 441-450). These vesicles are able to bind and be absorbed by the nasal mucosal epithelium (Harada T., Shimuzu T., Nishimoto K., Sakakura Y. 1989. Acta Otorhinolarygol. 246: 218-221) showing that adhesins and/or colonization factors could be present in Blebs. Immune response to proteins present in outer membrane vesicles has been observed in patients with various H. influenzae diseases (Sakakura Y., Harada T., Hamaguchi Y., Jin C.S. 1988. Acta Otorhinolarygol. Suppl. (Stockh.) 454: 222-226; Harada T., Sakakura Y„ Miyoshi Y. 1986. Rhinology 24: 61-66). Pseudomonas aeruginosa; The genus Pseudomonas consists of Gram-negative, polarly flagellated, straight and slightly carved rods-tha;-"grow aerobically and do not forms spores. Because of their limited metaboiic requirements, Pseudomonas spp. are ubiquitous and are widely distributed in the soil, the air, sewage water and in plants. Numerous species of Pseudomonas such as P. aeruginosa, P. pseudomallei, P. mallei, P. maltophilia and P. cepacia have also been shown to be pathogenic for humans. Among this list, P. aeruginosa is considered as an important human pathogen since it is associated with opportunistic infection of immuno-compromised host and is responsible for high morbidity in hospitalized patients. Nosocomial infection with P. aeruginosa afflicts primarily patients submitted for prolonged treatment and receiving immuno-suppressive agents, corticosteroids, antimetabolites antibiotics or radiation. The Pseudomonas, and particularly P. aeruginosa, produces a variety of toxins (such as hemolysins, ftbrinolysins, esterases, coaguhses, phospholipases, endo- and exo-toxins) that contribute to the pathogenicity of these bacteria. Moreover, these organisms have high intrinsic resistance to antibiotics due to the presence of multiple drug efflux pumps. This latter characteristic often complicates the outcome of the disease. Due to the uncontrolled use of antibacterial chemotherapeutics the frequency of nosocomial infection caused by P. aeruginosa has increased considerably over the last 30 years. In the US, for example, the economic burden of P. aeruginosa nosocomial infection is estimated to 4.5 billion USS annually. Therefore, the development of a vaccine for active or passive immunization against P. aeruginosa is actively needed (for review see Stanislavsky et al. 1997. FEMS Microbiol. Lett. 21: 243-277). Various cell-associated and secreted antigens of P. aeruginosa have been the subject of vaccine development, Among Pseudomonas antigens, the mucoid substance, which is an extracellular slime consisting predominantly of alginate, was found to be heterogenous in terms of size and imrnunogenicity. High molecular mass alginate components (30-300 kDa) appear to contain conserved epitopes while lower molecular mass alginare components (10-30 kDa) possess conserved epitopes in addition to unique epitopes. Among surface-associated proteins, PcrV, which is pan of the type III secretion-translocation apparatus, has also been shown to be an interesting target for vaccination (Sawa et al. 1999. Nature Medicine 5:392-398). ■- Surface-exposed antigens- including G-antigens (O-specific polysaccharide of- ~ LPS) or H-antigens (flagellar antigens) have been used for serotyping due to their highly immunogenic nature. Chemical structures of repeating units of O-specific polysaccharides have been elucidated and these data allowed the identification of 31 chemotypes of P. aeruginosa. Conserved epitopes among all serotypes of P. aeruginosa are located in the core oligosaccharide and the lipid A region of LPS and immunogens containing these epitopes induce cross-protective immunity in mice against different P. aeruginosa immunotypes. The outer core of LPS was implicated to be a ligand for binding of P. aeruginosa to airway and ocular epithelial cells of animals. However, heterogeneity exists in this outer core region among different serotypes. Epitopes in the inner core are highly conserved and have been demonstrated to be surface-accessible, and not masked by O-specific polysaccharide. To examine the protective properties of OM proteins, a vaccine containing P. aeruginosa OM proteins of molecular masses ranging from 20 to 100 kDa has been used in pre-cltnical and clinical rrials. This vaccine was efficacious in animal models against P. aeruginosa challenge and induced high levels of specific antibodies in human volunteers. Plasma from human volunteers containing anti-P. aeruginosa antibodies provided passive protection and helped the recovery of S7% of patients with severe forms of P. aeruginosa infection. More recently, a hybrid protein containing parts of the outer membrane proteins OprF (amino acids 190-342) and Oprl (amino acids 21-83)/rem Pseudomonas aeruginosa fused to the glutathione-S-transferase was shown to protect mice agamst a 975-fold 50% lethal dose of P. aeruginosa (Knapp et al. 1999. Vaccine. 17:1663-1669). The present inventors have realised a number of drawbacks associated with the above wild-type bleb vaccines (either naturally occurring or chemically made). Examples of such problems are the following: - the presence of immunodominant but variable proteins on the bleb (PorA; TbpE, Opa [A', meningitidis B]; P2, P5 [non-typeable H. influenzae]) - such blebs being effective only against a restricted selection of bacterial species. Type-specificity of the bactericidal antibody response may preclude the use of such vaccines in infancy. the presence of unprotecuve (non relevant) antigens (Rmp, HS, ...) on the bleb - antigens that are decoys for the immune system the lack of presence of important molecules which are produced conditionally (for instance iron-regulated outer membrane proteins, IROMP's, in vivo regulated expression mechanisms) - such conditions are hard to control in bleb production in order to optimise the amount of antigen on the surface - the low level of expression of protective, (particularly conserved) antigens (NspA, P6) the toxicity of the LPS remaining on the surface of the bleb the potential induction of an autoimmune response because of host-identical structures (for example the capsular polysaccharide in Neisseria meningitidis serogroup B, the lacto-N-neotetraose in Neisseria LPS, saccharide structure within ntHi LPS, saccharide structures within Pili), Such problems may prevent the use of bleb vaccines as human vaccine reagents. This is particularly so for paediatric use ( ineffective at immuno-protecting. Accordingly, the present invention provides methods of alleviating the above problems using genetically engineered bacterial strains, which result in improved bleb vaccines. Such methods will be especially useful in the generation of new vaccines against bacterial pathogens such as Neisseiria meningitidis, Moraxella catarrhaiis, Haemophilus influenzae, Pseudomonas aeruginosa, and others. The bleb vaccines of" the invention are designed to focus the immune response on a few protective (preferably conserved) antigens or epitopes - formulated in a multiple component vaccine. Where such antigens are integral OMPs, the outer membrane vesicles of bleb vaccines will ensure their proper folding. This invention provides methods to optimize the OMP and LPS composition of OMV (bleb) vaccines by deleting immunodominant variable as well as non protective OMPs, by creating conserved OMPs by deletion of variable regions, by upregulating expression of - proiective-GMPsv-and by eliminating control mechanisms for expression (such as iron restriction) of protective OMPs. In addition the invention provides for the reduction in toxicity of lipid A by modification of the lipid portion or by changing the phosphoryl composition whilst retaining its adjuvant activity or by masking it. Each of these new methods of improvement individually improve the bleb vaccine, however a combination of one or more of these methods work in conjunction so as to produce an optimised engineered bleb vaccine which is immuno-protective and non-toxic -particularly suitable for paediatric use. SUMMARY OF THE INVENTION The present invention provides a genetically-engineered bleb preparation from a Gram-negative bacterial strain characterized in that said preparation ;s obtainable by employing one or more processes selected from the following group: a) a process of reducing immunodominant variable or non-protective antigens within the bleb preparation comprising the steps of determining the identity of such antigen, engineering a bacterial strain tc produce less or none of said antigen, and making blebs from said strain; h) a process of upregulating. expression of protective, endogenous (and preferably conserved) OMP antigens within the bleb preparation comprising the steps of identifying such antigen, engineering a bacterial strain so as to introduce a stronger promoter sequence upstream of a gene encoding said antigen such that said gene is expressed at a level higher than in the non-modified bleb, and making blebs from said strain; c) a process of upregulating expression of conditionally-expressed, protective ■ (and preferably conserved) OMP antigens within the bleb preparation composing the steps of identifying such an antigen, engineering a bacterial strain so as to remove the repressive control mechanisms of its expression (such as iron restriction), and making blebs from said strain; d) a process of modifying lipid A portion of bacterial LPS within the bleb preparation, comprising the steps of identifying a gene involved in rendering the lipid A portion of LPS toxic, engineering a bacterial strain so as to reduce or switch off expression of said gene, and making blebs from said strain'. e) a process of modifying lipid A portion of bacterial LPS within the bleb preparation, comprising the steps of identifying a gene involved in rendering the lipid A portion of LPS less toxic, engineering a bacterial strain so as to introduce a stronger promoter sequence upstream of said gene such that said gene is expressed at a level higher than in the non-modified bleb, and making blebs from said strain; f) a process of reducing lipid A toxicity within the bleb preparation and increasing the levels of protective antigens, comprising the steps of engineering the chromosome of a bacterial strain to incorporate a gene encoding a Polymyxin A peptide, or a derivative or analogue thereof, fused to a protective antigen, and making blebs from said strain; g) a process of creating conserved OMP antigens on the bleb preparation comprising the steps of identifying such antigen, engineering a bacteria! strain so as to delete variable regions of a gene encoding said antigen, and making blebs from said strain; h) a process of reducing expression within the bleb preparation of an antigen wiiich shares a structural similarity with a human structure and may be capable of inducing an auto-immune response in humans (such as the capsular polysaccharide of N. meningitidis B), comprising the steps of identifying a gene involved in the biosynthesis of the antigen, engineering a bacterial strain so as to reduce or switch off expression of said gene, and making blebs from said strain; or i) a process of upregulating expression of protective, endogenous (and preferably conserved) OMP antigens within the bleb preparation composing the steps of identifying such antigen, engineering a bacterial strain so as to introduce into the chromosome one or more further copies of a gene encoding said antigen controlled by a heterologous, stronger promoter sequence, and making blebs from said strain. Further aspects of die invention include, preferential processes for obtaining the above bieb preparation, including optimal positioning of strong promoters for the upregulaaon of expression of antigens within biebs, preferential antigens for upregulation and downreguation for various bacterial strains in order to obtain bleb preparations particularly suitable for vaccine use. Preferential formulations comprising the blebs of the invention are aiso provided which are particularly suitable for global vaccines against certain disease states. Vectors for producing the blebs of the invention, and modified bacterial strains from which the blebs of the invention are produced are still further aspects of the invention. The present invention provides for the first time a bleb vaccine which is immune-protective and non-toxic when used with children under 4 years of age. BRIEF DESCRIPTION OF THE DRAWINGS Figure I: Reactivity of the 735 mAb on different colonies. Figure 2: Reactivities of specific monoclonal antibodies by whole cell Elisa, Figure 3: Schematic representation of the pCMK vectors used to-deliver genes, operons and/or expression cassettes in the genome of Neisseria meningitidis. Figure 4: Analysis of PorA expression in total protein extracts of recombinant \'. meningitidis serogroupB (H44/76 derivatives). Total proteins were recovered from cps-(lanes 3 and 4), cps- porA::pCMK+ (lanes 2 and 5) and cps-porA::nspA (lanes 1 and 6) recombinant N. meningitidis serogroupB strains and were analyzed under SDS-PAGE conditions in a 12% polyacrylamide gel. Gels were stained with Coomassie blue (lanes ! to 3) or transferred to a nitrocellulose membrane and immuno-stained with an antUPorA monoclonal antibody. Figure 5: Analysis of NspA expression in protein extracts of recombinant N. meningitidis serogroupB strains (H44/76 derivatives). Proteins were extracted from whole bacteria (lanes 1 to 3) or outer-membrane blebs preparations (lanes 4 to 6) separated by SDS-PAGE on a 12% acrylamide gel and analyzed by immuno-blotting using an anti-NspA polyclonal serum. Samples corresponding to cps- (lanes J and 6), cps- pora::pCMK+ (lanes—> and 4}-and cps- porA::nspA (lanes-2 and 5) were analyzed. Two forms of NspA were detected: a mature form (18kDa) co-migrating with the recombinant purified NspA, and a shorter form (15kDa). Figure 6: Analysis of D15/omp85 expression in protein extracts of recombinant N. meningitidis serogroupB strains (H44/76 derivatives). Proteins were extracted from outer-membrane blebs preparations and were separated by SDS-PAGE on a 12% acrylamide gel and analyzed by immuno-blotting using an anti-omp85 polyclonal serum. Samples corresponding to cps- (lane 2), and cps-, PorA+, pCMK+Omp85/Dl5 (lane 1) recombinant N. meningitidis serogroupB strains were analyzed. Figure 7: General strategy for modulating gene expression by promoter delivery (RS stands for restriction site). Figure 8; Analysis of outer-membrane blebs produced by recombinant A". meningitidis serogroupB cps- strains (H44/76 derivatives). Proteins were extracted from outer-membrane bleb preparations and were separated by SDS-PAGE under reducing conditions on a 4-20% gradient polyacrylamide gel. The gel was stained with Coomassie brilliant blue R250. Lanes 2, 4, 6 corresponded to 5ug of toiai proteins whereas ianes 3, 5 and 7 were loaded with lOug proteins. Figure 9: Construction of a promoter replacement plasmid used to up-regulate the expression/production of Omp85/D15 in Neisseria meningitidis H44/76. Figure 10: .Analysis of OMP85 expression in total protein extracts of recombinant AfmB (H44/76 derivatives). Gels were stained with Coomassie blue (A) or transferred to nitrocellulose membrane and immuno-stained with rabbit anti-OMP85 (N.gono) monoclonal antibody (B). Figure 11: Analysis of OMP85 expression in OMV preparations from recombinant NmB (H44/76 derivatives). Gels were stained with Coomassie blue (A) or transferred to nitrocellulose membrane and- immuno-stained with rabbit anti-OMP85- polyclonal antibody (B). Figure 12: Schematic representation of the recombinant PCR strategy used to delete the iacO in the chimeric porA/lacO promoter. Figure 13: Analysis of Hsf expression in total protein extracts of recombinant N. meningitidis serogroup B (H44/76 derivatives). Total proteins were recovered from Cps-PorA+(lanes I), and Cps-PorA+/Hsf (lanes 2) recombinant N. meningitidis serogroup B strains and were analyzed under SDS-PAGE conditions in a 12% polyacrylamide gel. Gels were stained with Coomassie blue . Figure 14: Analysis of GFP expression in total protein extracts of recombinant A'. meningitidis (H44/76 derivative). Total protein were recovered from Cps-, PorA+ (lanel), Cps-, PorA- GFP- (lane2 & 3) recombinant strains. Proteins were separated by PAGE-SDS in a 12% polyacryl amide gel and then stained with Coomassie blue. Figure 15: Illustration of the pattern of major proteins on the surface of various recomtnant bleb preparations as analysed by SDS-PAGE (Coomassie staining). Figure 16: Specific anti-Hsf response for various bleb and recombinant bleb preparations using purified recombinant Hsf protein. Figure 17: Analysis of NspA expression in total protein extracts of recombinant NmB (serogroup B derivatives). Gels were stained with Coomassie blue (A) or transferred to nitrocellulose membrane and immuno-stained with mouse anti-PorA monoclonal antibody (B) or mouse anti-NspA polyclonal antibody (C). DESCRIPTION OF THE INVENTION The present invention relates to a general set of tools and methods capable of being used for manufacturing improved, genetically engineered blebs from Gram-negative bacterial strains. The- invention includes methods used to make recombinant blebs more immunogenic, less toxic and safer for their use in a human and/or animal vaccine. Moreover, the present invention also describes specific methods necessary for constructing, producing, obtaining and using recombinant, engineered blebs from various Gram-negative bacteria, for vaccine, therapeutic and/or diagnostic purposes. By the methods of the invention, the biochemical composition of bacterial bJebs can be manipulated by acting upon/altering the expression of bacterial genes encoding products present in or associated with bacterial outer-membrane blebs (outer membrane proteins or OMPs). The production of blebs using a method of genetic modification to increase, decrease or render conditional the expression of one or more genes encoding outer-membrane components are also included m the scope of this invention. For clarity, [he term "expression cassette" will refer herein to all the genetic elements necessary to express a gene or an operon and to produce and target the corresponding protem(s) of interest to outer-membrane blebs, derived from a given bacterial host. A non-exhaustive list of these features includes control elements (transcriptional and/or translationai), protein coding regions and targeting signals, with appropriate spacing between them. Reference to the insertion of promoter sequences means, for the purposes of this invention, the insertion of a sequence with at least a promoter function, and preferably one or more other genetic regulatory elements comprised within an expression cassette. Moreover, the term "integrative cassette" will refer herein to all the genetic elements required to integrate a DNA segment in given bacterial host. A non-exhaustive list of these features includes a delivery vehicle (or vector), with recombinogenic regions, and selectable and counter selectable markers. Again for the purpose of clarity, the terms 'engineering a bacterial strain to produce less of said antigen' refers to any means to reduce the expression of an antigen of interest, relative to that of the non-modified (i.e., naturally occurring) bleb such that expression is at Isas: J0% lower than that of the non-modified bJeb. Preferably it is at least 50% lower. "Stronger promoter sequence" refers to a regulatory control element that increases transcription for a gene encoding antigen of interest. "Upregulating expression" refers to any means to enhance the expression of an antigen of interest, relative to thai of the non-modified (i.e., naturally occurring) bleb. It is understood that the amount of 'upregulation' will vary depending on the particular antigen of interest but will not exceed an amount that will disrupt the membrane integrity of the bleb. Upregulation of an antigen refers to expression that is at least 10% higher than that of the non-modified bleb. Preferably it is at least 50% higher. More preferably it is at least 100% (2 fold) higher. Aspects of the invention relate to individual methods for making improved engineered blebs, to a combination of such methods, and to the bleb compositions made as a result of these methods. Another aspect of the invention relates to the genetic tools used in order to genetically modify a chosen bacterial strain in order to extract improved engineered blebs from said strain. The engineering steps of the processes of the invention can be carried out in a variety of ways known to the skilled person. For instance, sequences (e.g. promoters or open reading frames) can be inserted, and promoters/genes can be disrupted by the technique of transposon insertion. For instance, for upregulating a gene's expression, a strong promoter could be inserted via a transposon up to 2 kb upstream of the gene's initiation codon (more preferably 200-600 bp upstream, most preferably approximately 400 bp upstream). Point mutation or deletion may also be used (particularly for down-regulating expression of a gene). Such methods, however, may be quite unstable or uncertain, and therefore it is preferred that the engineering step [panicularly for processes a), b), c), d), e), h) and i) as described below] is performed via a homologous recombination event. Preferably, the event takes place between a sequence (a recombinogenic region) of at least 30 nucleotides on the bacterial chromosome, and a sequence (a second recombinogenic region) of at least 30 nucleotides on a vector transformed within the strain. Preferably the regions are 40-1000 nucleotides, more preferably 100-800 nucleotides, most preferably 500 nucleotides). These recombinogenic regions should be sufficiently similar that they are capable of hybridising to one another under highly stringent conditions (as defined later). Recombination events may take place using a single recombinogenic region on chromosome and vector, or via a double cross-over event (with 2 regions on chromosome and vector). In order to perform a single recombination event, the vector ■should be a circular DNA molecule, in order lo perform a double recombination-event, the vector could be a circular or linear DNA molecule (see Figure 7). It is preferable that a double recombination event is employed and that the vector used is linear, as the modified bacterium so produced will be more stable in terms of reversion events. Preferably the two recombinogenic regions on the chromosome (and on the vector) are of similar (most preferably the same) length so as to promote double cross-overs. The double cross-over functions such that the two recombinogenic regions on the chromosome (separated by nucleotide sequence 'X') and the two recombinogenic regions on the vector (separated by nucleotide sequence 'Y') recombine to leave a chromosome unaltered except that X and Y have interchanged. The position of the recombinogenic regions can both be positioned upstream or down stream of, or may flank, an open reading frame of interest. These regions can consist of coding, non-coding, or a mixture of coding and non-coding sequence. The identity of X and Y will depend on the effect desired. X may be all or part of an open reading frame, and Y no nucleotides at all, which would result in sequence X being deleted from the chromosome. Alternatively Y may be a strong promoter region for insertion upstream of an open reading frame, and therefore X may be no nucleotides at all. Suitable vectors will vary in composition depending what type of recombination event is to be performed, and what the ultimate purpose of the recombination event is. Integrative vectors used to deliver region Y can be conditionally replicative or suicide plasmids, bacteriophages, transposons or linear DNA fragments obtained by restriction hydrolysis or PCR amplification. Selection of the recombination event is selected by means of selectable genetic marker such as genes conferring resistance to antibiotics (for instance kanamycin, erythromycin, chloramphenicol, or gentamycin), genes conferring resistance to heavy metals and/or toxic compounds or genes complementing auxotrophic mutations (for instance pur, leu, met, aro). Process a) and f) - Down regulation/Removal of Variable and non-protective immunodominant antigens Many surface antigens are variable among bacterial strains and as a consequence are- protective only against -a limited set of closely related strains. An aspect of this invention covers the reduction in expression, or, preferably, the deletion of the gene(s) encoding variable surface protein(s) which results in a bacterial strain producing blebs which, when administered in a vaccine, have a stronger potential for cross-reactivity against various strains due to a higher influence exerted by conserved proteins (retained on the outer membranes) on the vaccinee's immune system. Examples of such variable antigens include: for Neisseria - pili (PilC) which undergoes antigenic variations, PorA, Opa, TbpB, FrpB; for H. influenzae • ?2, P5, pilin, IgAl-protease; and for Moraxella - CopB, OMP106. Other types of gene that could be down-regulated or switched off are genes which, in vivo, can easily be switched on (expressed) or off by the bacterium. As outer membrane proteins encoded by such genes are not always present on the bacteria, the presence of such proteins in the bleb preparations can also be detrimental to the effectiveness of the vaccine for the reasons stated above. A preferred example to down-regulate or delete is Neisseria Ope protein. Anti-Opc immunity induced by an Ope containing bleb vaccine would only have limited protective capacity as the infecting organism could easily become Ope'. H. influenzae HgpA and HgpB are other examples of such proteins. In process a), these variable or non-protective genes are down-regulated in expression, or terminally switched off. This has the above-mentioned surprising advantage of concentrating the immune system on better antigens that are present in low amounts on the outer surface of blebs. The strain can be engineered in this Way by a number of strategies including transposon insertion to disrupt the coding region or promoter region of the gene, or point mutations or deletions to achieve a similar result. Homologous recombination may also be used to delete a gene from a chromosome (where sequence X comprises part {preferably all) of the coding sequence of the gene of interest). It may additionally be used to change its strong promoter for a weaker (or no) promoter (where nucleotide sequence X comprises part (preferably all) of the promoter region of the gene, and nucleotide sequence Y comprises either a weaker promoter region [resulting in a decreased expression of the gene(s)/operon(s) of interest], or no promoter region). In this-caschis preferable-tor the recombination event to occur within the region of the chromosome 1000 bp upstream of the gene of interest. Alternatively, Y may confer a conditional transcriptional activity, resulting in a conditional expression of the gene(s)/operon(s) of interest (down-reguJarion), This is useful in the expression of molecules that are toxic to or not well supported by the bacterial host. Most of the above-exemplified proteins arc integral OMPs and their variability may be limited only to one or few of their surface exposed loops. Another aspect of this invention [process g)] covers the deletion of DNA regions coding for these surface exposed loops which leads to the expression of an integral CMP containing conserved surface exposed loops. Surface exposed loops of H. influenzae P2 and P5 are preferred examples of proteins that could be transformed into cross-reactive antigens by using such a method. Again, homologous recombination is a preferred method of performing this engineering process. Process b) - Promoter delivery and modulation: A further aspect of the invention relates to modifying the composition of blebs by altering in situ the regulatory region controlling the expression of gene(s) and'or operon(s) of interest. This alteration may include partial or total replacement of the endogenous promoter controlling the expression of a gene of interest, with one conferring a distinct transcriptional activity. This distinct transcriptional activity may he conferred by variants (point mutations, deletions and/or insertions) of the endogenous control regions, by naturally occurring or modified heterologous promoters or by a combination of both. Such alterations will preferably confer a transcriptional activity stronger than the endogenous one (introduction of a strong promoter), resulting in an enhanced expression of the gene(s)/operon{s) of interest (up-regulation). Such a method is particularly useful for enhancing the production of immunologically relevant Bieb components such as outer-membrane proteins and lipoproteins (preferably conserved OMPs, usually present in blebs at low concentrations). Typical strong promoters that may be integrated in Neisseria are porA [SEQ ID NO: 24], porB [SEQ ID NO:26], IgiF, Opa, pUO, 1st, and hpuAB. PorA and PorB are preferred as constitutive, strong promoters. It -has been established (Example 9) that the PorB promoter activity is contained in a fragment corresponding to nucleotides -1 to -250 upstream of the initation codon of porB. In Moraxella, it is preferred to use the ompH, ompG, ompE, OmpB.1, ompB2, ompA, OMPCD and Omp 106 promoters, and in H. influenzae, it is preferred to integrate the P2, P4, P1, P5 and P6 promoters. Using the preferred double cross-over homologous recombination technology to introduce the promoter in the 1000 bp upstream region, promoters can be placed anywhere from 30-970 bp upstream of the initiation codon of the gene to be up-regulated. Although conventionally it is thought the promoter region should be relatively close to the open reading frame in order to obtain optimal expression of the gene, the present inventors have surprisingly found that placement of the promoter further away from the initiation codon results in large increases in expression levels. Thus it is preferred if the promote: is inserted 200-600 bp from the initiation codon of the gene, more preferably 300-500 bp, and most preferably approximately 400 bp from the initiation ATG. Process c) - Bleb components produced conditionally The expression of some genes coding for certain bleb components is carefully regulated. The production of the components is conditionally modulated and depends upon various metabolic and/or environmental signals. Such signals include, for example, iron-limitation, modulation of the redox potential, pH and temperature variations, nutritional changes. Some examples of bleb components known to be produced conditionally include iron-regulated outer-membrane proteins from Neisseiria and Moraxella (for instance TbpB, LbpB), and substrate-inducible outer-membrane porins. The present invention covers the use of the genetic methods described previously (process a) or b)) to render constitutive the expression of such molecules. In this way, the influence of environmental signal upon the expression of gene(s) of interest can be overcome by modifying/replacing the gene's corresponding control- region- so that it -becomes constitutively active (for instance by deleting part [preferably all] or the repressive control sequence - e.g. the operator region), or inserting a constitutive strong promoter. For iron regulated genes the/ur operator may be removed. Alternatively, process i) may be used to deliver an additional copy of the gene/operon of interest in the chromosome which is placed artificially under the control of a constitutive promoter. Processes d), and ej - Detoxification of LPS The toxicity of bleb vaccines presents one of the largest problems in the use of blebs in vaccines. A further aspect of the invention relates to methods of genetically detoxifying the LPS present in Blebs. Lipid A is the primary component of LPS responsible for cell activation. Many mutations in genes involved in this pathway lead to essential phsnorypes. However, mutations in the genes responsible for the terminal modifications steps lead to temperature-sensitive (h:rB) or permissive (msbB) phenotypes. Mutations resulting in a decreased (or no) expression of these genes result in altered toxic activity of lipid A. Indeed, the non-lauroylated (htrB mutant) or non-myristoylated i,msbB mutant) lipid A are less toxic than tie wild-type lipid A. Mutations in the lipid A ^'-kinase encoding gene {ipxK) also decreases the toxic activity of lipid A. Process d) thus involves either the deletion of pan (or preferably all) of one or more of the above open reading frames or promoters. Alternatively, the promoters could be replaced with weaker promoters. Preferably the homologous recombination techniques described above are used to carry out the process. The sequences of the htrB and msbB genes from Neisseria meningitidis B, Moraxella catarrhalis, and Haemophilus influenzae are additionally provided for this purpose. LPS toxic activity could also be altered by introducing mutations in genes/loci involved in polymyxin B resistance (scch resistance has been correlated with addition of aminoarabinose on the 4' phosphate, of lipid A). These genes/loci could be pmrE that encodes a UDP-glucose dehydrogenase, or a region of antimicrobial peptide-resistance genes common to many enterobacteriaciae which could be involved in aminoarabinose synthesis and transfer. The gene pmrF that is present in this region encodes a dolicol-phosphate manosyl transferase {Gunn J.3.,- Kbeng, B.L., Krusger J., Kim K.,Guo L., Hackett M., Miller S.I. 1998. Mol. Microbiol. 27: 1171-1182). Mutations in the PhoP-PhoQ regulatory system, which is a phospho-relay two component regulatory system (I i. PhoP constitutive phenorype, PboP), or low Mg** environmental or culture conditions (that activate the PhoP-PhoQ regulatory system) lead to the addition of aminoarabinose on the 4'-phosphate and 2-bydroxymyristate replacing myristate (hydroxylation of myristate). This modified lipid A displays reduced ability to stimulate E-selectin expression by human endothelial cells and TNF-a secretion from human monocytes. Process e) involves the upregulation of these genes using a strategy as described above (strong promoters being incorporated, preferably using homologous recombination techniques to carry out the process). Alternatively, rather than performing any such mutation, a polymyxin B resistant strain could be used as a vaccine production strain (in conjunction with one or more of the other processes of the invention), as blebs from such strains also have reduced LPS toxicity (for instance as shown for meningococcus - van der Ley, P, Hamstra, HJ, Kramer, M, Steeghs, L, Petrov, A and Poolman. JT. 1994. In: Proceedings of the ninth international pathogenic Neisseria conference. The Guildhall, Winchester, England). As a further alternative (and further aspect of the invention) the inventors provide a method of detoxifying a Gram-negative bacterial strain comprising the step of cuitunng the strain in a growth medium containing O.lmg-lOOg of aminoarabinose per litre medium. As a further still alternative, synthetic peptides that mimic the binding activity of polymyxin B (described below) may be added to the Bleb preparation in order to reduce LPS toxic activity (Rustici, A, Velucchi, M, Faggioni, R, Sironi, M, Ghezzi, P, Quataert, S, Green, B and Porro M 1993. Science 259: 361-365; Velucchi, M, Rustici, A, Meazza, C, Villa, P, Ghezzi, P and Porro, M. 1997.1 Endotox. Res. 4;). Process f) - Anchoring homologous or heterologous proteins to outer-membrane biebs whilst reducing the toxicity of LPS A further aspect of this invention covers the use of genetic sequences encoding polymyxin B peptides-(or analogues thereof) as a means to target fusion protein? to the outer-membrane. Polymyxin B is a cyclic peptide composed of non tRNA-encoded amino acids (produced by Gram-positive actinomycetal organisms) that binds very strongly to the Lipid A part of LPS present m the outer-membrane. This binding decreases the intrinsic toxicity of LPS (endotoxin activity). Peptides mimicking the structure of Polymyxin B and composed of canonical (tRNA encoded) amino acids have been developed and also bind lipid A with a strong affinity. These peptides have been used for detoxifying LPS. One of these peptides known as SAEP-2 (Nterminus-Lys-Thr-Lys-Cys-Lys-Phe-Leu-Lys-Lys-Cys-Cterminus) was shown to be very promising in that respect (Molecular Mapping and detoxifying of the Lipid A binding site by syntiietic peptides (1993). Rustici, A., Velucchi, M., Faggioni, R., Sironi, M., Ghezzi, P., Quataert, S., Green, B. and M. Porro. Science 259, 361-365). Tne present process f) of the invention provides an improvement of this use. It has been found that the use of DNA sequences coding for the SEAP-2 peptide (or derivatives thereof), fused genetically to a gene of interest (encoding for instance a T cell antigen or a protective antigen that is usually secreted such as a toxin, or a cytosolic or periplasms protein) is a means for targeting the corresponding recombinant protein to the outer-membrane of a preferred bacterial host (whilst a! the same time reducing the toxicity of the LPS). This system is suitable for labile proteins which would not be directly exposed to the outside of the bleb. The bleb would therefore act as a delivery vehicle which would expose the protein to the immune system once the blebs had been engulfed by T-cells. Alternatively, the genetic fusion should also comprise a signal peptide or transmembrane domain such that the recombinant protein may cross the outer membrane for exposure to the host's immune system. This targeting strategy might be of particular interest in the case of genes encoding proteins that are not normally targeted to the outer-membrane. This methodology also allows the isolation of recombinant blebs enriched in the protein of interest. Preferably, such a peptide targeting signal allows the enrichment of outer membrane blebs in one or several proteins of interest, which are naturally not found in that given subcellular localization. A non exhaustive list of bacteria that can be used as a recipient host for such a production of recombinant blebs includes Neisseria meningitidis, Neisseiria gonorrhoeae Moruxella catarrhalis, Haemophilus IJ:.^™;:jp, Pseudomonas aeruginosa, Chlamydia trachomatis, and Chlamydia pneumoniae. Although it is preferred that the gene for the construct is engineered into the chromosome of the bacterium [using process i)], an alternative preferred embodiment is for SAEP-2-tagged recombinant proteins to be made independently, and attached at a later stage to a bieb preparation. A further embodiment is the use of such constructs in a method of protein purification. The system could be used as part of an expression system for producing recombinant proteins in general. The SAEP-2 peptide tag can be used for affinity purification of the protein to which it is attached using a column containing immobilised lipid A molecules. process h) ■ Cross-reactive polysaccharides The isolation of bacterial outer-membrane blebs from encapsulated Gram-oeeative bacteria often results ir. the co-purification of capsular polysaccharide. In Some cases, this '"contaminant" material may prove useful since polysaccharide may enhance the immune response conferred by other bleb components. In other cases however, the presence of contaminating polysaccharide materia! in bacterial bleb preparations may prove detrimental to the use of the blebs in a vaccine. For instance, it has been shown at least in the case of A-, meningitidis that the serogroup B capsular polysaccharide does not confer protective immunity and is susceptible to induce an adverse auto-immune response in humans. Consequently, process h) of the invention is the engineering of the bactenal strain for bleb production such that it is free of capsular polysaccharide. The blebs will then be suitable for use in humans. A particularly preferred example of such a bleb preparation is one from N. meningitidis serogroup B devoid of capsular polysaccharide. This may be achieved by using modified bleb production strains in which the genes necessary for capsular biosynthesis and/or export have been impaired. Inactivatiort of the gene coding for capsular polysaccharide biosynthesis or export can be achieved by mutating (point mutation, deletion or insertion) either the control region, the coding region or both (preferably using the homologous recombination techniques described above). Moreover, inactivation of capsular biosynthesis genes may also be achieved by-anrisenae- o^er-ex^ression or transposon mutagenesis. A preferred method is the deletion of some or all of the Neisseria meningitidis cps genes required for polysaccharide biosynthesis and export. For this purpose, the replacement plasmidpMF12i {described in Frosh et al. 1990, Mol. Microbiol, 4; 1215-1218) can be used to deliver a mutation deleting the cpsCAD (^ galE) gene cluster. Alternatively the siaD gene could be deleted, or down-regulated in expression (the meningococcal siaD gene encodes alpha-2,3-sialyltransferase, an enzyme required for capsular polysaccharide and LOS synthesis). Such mutations may also remove host-similar structures on the saccharide portion of the LPS of the bacteria. Process i) - Delivery of one or more further copies of a Gene and/or operon in a host chromosome, or delivery of a heterlogous gene and/or operon in a host chromosome. .An efficient strategy to modulate the composition of a Bleb preparation is to deliver one or more copies of a DNA segment containing an expression cassette into the genome of a Gram-negative bactenum. A non exhaustive list of preferred bacterial species tha: could be used as a recipient for such a cassette includes Neisseria meningitidis, Neisseiria gonorrhoeae, MoraxeUa caiarrhalis, Haemophilus influenzae, Pseudomonas aeruginosa. Chlamydia trachomatis. Chlamydia pneumoniae. The gene(s) contained in the expression cassette may be homologous (or endogenous) (i.e. exist naturally in the genome of the manipulated bacterium) or heterologous (i.e. do not exist naturally in the genome of the manipulated bacterium). The reintroduced expression cassette may consist of unmodified, "natural" promoter/gene/operon sequences or engineered expression cassenes in which the promoter region and/or the coding region or both have been altered. A non-exhaustive list of preferred promoters that could be used for expression includes the promoters porA,porB, IbpB, tbpB,pII0, 1st, kpuAB from A', meningitidis or N. gonorroheae, the promoters p2, p5, p4, ompF, pi. ompH, p6, bin47 from H. influenzae, the promoters ompH, ompG, ompCD, ompE, ompBl, ompB2, ompA of M. catarrhalis, the promoter X.pL, lac, tac, araB of Escherichia coli or promoters recognized specifically by bacteriophage RNA polymerase such as the E. coli bacteriophage T7. A non-exhaustive list of preferred genes that could be expressed in such a system includes Neisseria NspA, Omp85, PilQ, TbpA/B complex, Hsf, PldA, HasR; Chlamydia MOMP^HMWP; Moraxella OMP106, HasR, PilQ, OMP85, PldA; Bordetella pertussis FHA, PRN, PT. In a preferred embodiment of the invention the expression cassette is delivered and integrated in the bacterial chromosome by means of homologous and/or site specific recombination. Integrative vectors used to deliver such genes and/or operons can be conditionally replicative or suicide piasmids, bacteriophages, transposons or linear DNA fragments obtained by restriction hydrolysis or PCR amplification. Integration is preferably targeted to chromosomal regions dispensable for growth in vitro. A non exhaustive list of preferred loci that can be used to target DNA integration includes the porA, porB, opa, ope, rmp, omp26, lecA, cps, IgtB genes of Neisseiria meningitidis and Neisseria gonorrhoeae, the PI, PS. hmwl/2, IgA-proiease, fimE genes of NTHi: the'tecAl, lecA2, omplQ6, uspAl, uspA2 genes of Moraxella catarrhalis. Alternatively, the expression cassette used to modulate the expression of bleb component(s) can be delivered into a bacterium of choice by means of episomal vectors such as circular/linear replicative piasmids. cosmida, phasmids, iysogenic bacteriophages or bacterial artificial chromosomes. Selection of the i recombination event can be selected by means of selectable genetic marker such as genes conferring resistance to antibiotics (for instance kanamycin. erythromycin, chloramphenicol, or eentamvein), genes confemne resistance to heaw metals and/or toxic compounds or genes complementing auxotroohtc mutations (for instance pur, leu. met. arc). Heterologous Genes ■ Expression of foreign proteins in outer-membrane blebs Outer-membrane bacterial blebs represent a very attractive system to produce, isolate and deliver recombinant proteins for vaccine, therapeutic and/or diagnostic uses. A further aspect of this invention is in respect of the expression, production and [argeting of foreign, heterologous proteins to the outer-membrane, and the use of the bacteria to produce recombinant blebs. A preferred method of achieving this is via a process comprising the steps of: introducing a heterologous gene, optionally controlled by a strong promoter sequence, into the chromosome of a Gram-negative strain by homologous recombination. Biebs may be made from the resulting modified strain. A non-exhaustive list of bacteria that can be used as a recipient host for production of recombinant blebs includes Neisseria meningitidis, Neisseiria gonorrhoeae Moraxella catarrhalis, Haemophilus influenzae, Pseudomonas aeruginosa. Chlamydia trachomatis. Chlamydia pneumoniae. The gene expressed in such a system can be of viral, bacterial, fungal, parasitic or higher eukaryocic origin. A preferred application of the invention includes a process for the expression of Moraxella, Haemophilus and/or Pseudomonas outer-membrane proteins (integral, polytopic and/or lipoproteins) in Neisseria meningitidis recombinant blebs. The preferable integration loci are stated above, and genes that are preferably introduced are those that provide protection against the bacterium from which they were isolated. Preferred protective genes for each bacterium are described below. Further preferred applications are: blebs produced from a modified Haemophilus influenzae strain where the heterologous gene is a protective OMP from Moraxella catarrhalis; and blebs produced from a modified Moraxella catarrhalis strain where the heterologous gene is a protective OMP from Haemophilus influenzae (preferred loci for gene insertion are given above, and preferred protective antigens are desenbed below). A particularly preferred application of this aspec*. is in the field of the prophylaxis or treatment of sexually-transmitted diseaseses (STDs). It is often difficult for practitioners to determine whether the principal cause of a STD is due to gonococcus or Chlamydia trachomatis infection. These two organisms are the main causes of salpingitis - a disease which can lead to sterility in the host. It would therefore be useful if a STD could be vaccinated against or treated with a combined vaccine effective against disease caused by both organisms. The Major Outer Membrane Protein (MOMP) of C. trachomatis has been shown to be the target of protective antibodies. However, the structural integrity of this integral membrane protein is important for inducing such antibodies. In addition, the epitopes recognised by these antibodies are variable and define more than 10 serovars. The previously described aspect of this invention allows the proper folding of one or more membrane proteins within a bleb outer membrane preparation. The engineering of a gonococcal strain expressing multiple C. trachomatis MOMP serovars in the outer membrane, and the production of blebs therefrom, produces a single solution to the multiple problems of correctly folded membrane proteins, the presentation of sufficient MOMP serovars to protect against a wide spectrum of serovars, and the simultaneous prophylaxis/treatment of gonococcal infection (and consequently the non-requirement of practitioners to initially decide which organism is causing particular clinical symptoms - both organisms can be vaccinated against simultaneously thus allowing the treatment of the STD at a very early stage). Preferred loci for gene insertion in the gonoccocal chromosome are give above. Other preferred, protective C. trachomatis genes that could be incorporated are HMWP, PmpG and those OMPs disclosed in WO 99/28475. Targeting of heterologous proteins to outer-membrane blebs: The expression of some heterologous proteins in bacterial blebs may require the addition of outer-membrane targeting signal(s). The preferred method to solve this problem is bv creating a genetic fusion between a heterologous gene and a gene coding for a resident OMP as a specific approach to target recombinant proteins to blebs. Most preferably, the heterologous gene is fused to the signal peptides sequences of such an OMP. Neisserial bleb preparations One or more of the following genes (encoding protective antigens) are preferred for upreguiation via processes b) and/or i) when carried out on a Neisserial strain, including gonococcus, and meningococcus (particularly Ar. meningitidis By. NspA {WO 96/29412), Hsf-like (WO 99/31132), Hap (PCT/EP99/02766), PorA, PorB, OMP85 (WO 00/23595), PilQ (PCT/EP99/03603), PldA (PCT/EP99/06718), FrpB (WO 96/31618), TbpA (US 5,912,336), TbpB, FrpA/FrpC (WO 92/01460), LbpA/LbpB (PCT/EP98/05117), FhaB (WO 98/02547), HasR (PCT/EP99/05989), npo02 (PCT/EP99/08315), Tbp2 (WO 99/57280), MltA (WO 99/57280), and ctrA (PCT/EP0O/O0135). They are also preferred as genes which may be heterologously introduced into other Gram-negative bacteria. One or more of the following genes are preferred for downregulation via process a): PorA, PorB, PilC, TbpA, TbpB, LbpA, LbpB, Opa, and Ope. One or more of the following genes are preferred for downregulation via process d); htrB, msbB and IpxK. One or more of the following genes are preferred for upreguiation via process e): pmrA, pmrB, pmrE, and pmrF. Preferred repressive control sequences for process c) are: the fur operator region (particularly for either or both of the TbpB or LbpB genes); and the DtxR operator region. One or more of the fallowing genes are preferred for downregulation via process h); galE, siaA, siaB, siaC, siaD, ctrA, ctrB, ctrC, and ctrD. Pseudomonas aeruginosa bleb preparations One or more of the following genes (sneoding protective antigens) are preferred for upreguiation via processes b) and/or i): PcrV, OprF, Oprl. They are also preferred as genes which may be heterologously introduced into other Gram-negative bacteria. ) Moraxelia catarrhaiis bleb preparations One or more of the following genes (encoding protective antigens) are preferred for upreguiation via processes b) and/or i): OMP106 (WO 97/4173! & WO 96/34960), HasR (PCT/EP99/03824), PilQ (PCT/EP99/03823), OMP85 (PCT/EPOO/01468), lipo06 (GB 9917977.2), IipolO (GB 9918208.1), iipoll (GB 9918302.2), lipolS (GB 9918038.2), P6 (PCT/EP99/03038), ompCD, CopB (Helminen ME, et al (1993) Infect. Immun. 61:2003-2010), D15 (PCT/EP99/03822), OmplA) (PCT/EP99/06781), Hly3 (PCT/EP99/03257), LbpA and LbpB {WO 98/55606), TbpA and TbpB (WO 97/13785 & WO 97/32980), OmpE, UspAl and UspA2 (WO 93/03761), and Omp21. They are also preferred as genes which may be heterologously introduced into other Gram-negative bacteria. One or more of the following genes are preferred for downregulation via process a); CopB, OMP106, OmpBl, TbpA, TbpB, LbpA, and LbpB. One or more of the following genes are preferred for downregulation via process d): htrB, msbB and IpxK. One or more of the following genes are preferred for upregulation via process •e): pmrA, pmrB pcirE,-and pmrF. Haemophilus influenzae bleb preparations One or more of the following genes (encoding protective antigens) are preferred for upregulation via processes b) and/or i): D15 (WO 94/12641), P6 (EP 281673), TbpA, TbpB, P2, P5 (WO 94/26304), OMP26 (WO 97/01638), HMWl, HMW2, HMW'3, HMW4, Hia, Hsf, Hap, Hin47, and Hif (all genes in this operon should be upregulated in order to upregulate pilin). They are also preferred as genes which may be heterologously introduced into other Gram-negative bacteria. One or more of the following genes are preferred for downregulation via process a)-. P2, PS, Hif, IgAl-protease, HgpA. HgpB, HMWl, HMW2, Hxu. TbpA, and TbpB. One or more of the following genes are preferred for downregulation via process d): htrB, msbB and IpxK. One or more of the following genes are preferred for ^regulation via process e): pmrA, pmrB. pmrE. and pmrF. Vaccine Formulations A preferred embodiment of the invention is the formulation of the bleb preparations of the invention in a vaccine which may also comprise a pharmaceutical I y acceptable excipient. The manufacture of bleb preparations from any of the aforementioned modified strains may be achieved by any of the methods well known to a skilled person. Preferably the methods disclosed in EP 301992, US 5,597,572, EP 11243 or US 4,271,147 are used. Most preferably, the method described in Example 8 is used. Vaccine preparation is generally described in Vaccine Design ("The subunit and adjuvant approach" (eds Powell M.F. & Newman M.J.) (1995) Plenum Press New York). The bleb preparations of the present invention may be adjuvanted in the vaccine formulation of the invention. Suitable adjuvants include an aluminium salt such, ss aluminum hydroxide gel (alum) or aluminium phosphate, b;r. JTT.IV JSO bz z salt of calcium (particularly calcium carbonate), iron or zinc, or may be an insoluble suspension of acylated tyrosine, or acylated sugars, cationically or anionically denvatised polysaccharides, or polyphosphazenes. Suitable Thl adjuvant systems that may be used include, Monopbosphoryl lipid A, particularly 3-de-O-acylated monophosphoryl lipid A, and a combination of monophosphoryl lipid A, preferably 3-de-O-acylated monophosphoryl lipid A (3D-MPL) together with an aluminium salt. An enhanced system involves the combination of a monophosphoryl lipid A and a saponin derivative particularly the combination of QS2! and 3D-MPL as disclosed in WO 94/00153. or a less reactogenic composition where the QS21 is quenched with cholesterol as disclosed in W096/33739. A i particularly potent adjuvant formulation involving QS21 3D-MPL and tocopherol in an oil in water emulsion is described in WO95/17210 and is a preferred formulation. The vaccine may comprise a saponin, more preferably QS21. It may also comprise an oi! in water emuision and tocopherol. Unmethylated CpG containing oligo nucleotides (WO 56/02555) are also preferential inducers of a TH1 response and 3 are suitable for use in th: present invention. The vaccine preparation of the present invention may be used to protect or treat a mammal susceptible to infection, by means of administering said vaccine via systemic or mucosal route. These administrations may include injectiori via the intramuscular, intraperitoneal, intradermal or subcutaneous routes; or via mucosal administration to the oral/alimentary, respiratory, genitourinary tracts. Thus one aspect of the present invention is a method of immunizing a human host against a disease caused by infection of a gram-negative bacteria, which method comprises administering to the host an immunoprotective dose of the bleb preparation of the present invention. The amount of antigen in each vaccine dose is selected as an amount which induces an immunoprotective response without significant, adverse side effects in typical vaccinees. Such amount will vary depending upon which specific immunogen is employed and how it is presented. Generally, it is expected that each dose will comprise l-100|ig of protein antigen, preferably 5-50)ig, and most typically in me range 5 • 25u,g. An optimal amount-for-a. parrisutaf Vaccine can be ascertained by standard studies involving observation of appropriate immune responses in subjects. Following an initial vaccination, subjects may receive one or several booster immunisations adequately spaced. Ghost or Killed Whole cell vaccines The inventors envisage that the above improvements to bleb preparations and vaccines can be easily extended to ghost or killed whole cell preparations and vaccines (with identical advantages). The modified Gram-negative strains of the invention from which the bleb preparations are made can also be used to made ghost and killed whole cell preparations. Methods of making ghost preparations (empty cells with intact envelopes) from Gram-negative strains are well known in the art (see for example WO 92701791). Methods of killing whole cells to make inactivated cell preparations for use in vaccines are also well known. The terms 'bleb preparations' and 'bleb vaccines' as well as the processes described throughout this document are therefore applicable to the terms 'ghost preparation' and 'ghost vaccine', arid 'killed whole cell preparation' and 'killed whole cell vaccine', respectively, for the purposes of this invention. Combinations of methods a) - i) It may be appreciated that one or more of the above processes may be used to produce a modified strain from which to make improved bleb preparations of the invention. Preferably one such process is used, more preferably two or more (2, 3, 4, 5, 6, 7, 8 or 9) of the processes are used in order to manufacture the bleb vaccine. As each additional method is used in the manufacture of the bleb vaccine, each improvement works in conjunction with the other methods used in order ID make an optimised engineered bleb preparation. A preferred meningococcal (particularly N. meningitidis B) bleb preparation comprises the use of processes a), b), d) and/or e), and h). Such bieb preparations are safe (no structures similar to host structures), non-toxic, and structured such that the host immune response will be focused on high levels of protective (and preferably conserved) antigens. All the above elements work together in order to provide an - optimised-bleb vaccine. :.- ,-j . Similarly for M. catarrhalis and non-typeable H. influenzae, preferred bleb preparations comprise the use of processes a), b), and d) and/or e). A further aspect of the invention is thus an immuno-protective and non-toxic Gram-negative bleb, ghost, or killed whole cell vaccine suitable forpaediatric use. By paediatric use it is meant use in infants less than 4 years old. By immunoprotective it is meant that at least 40% (and preferably 50, 60, 70, 80, 90 and 100%) of infants seroconven (4-fold increase in bactericidal activity [the dilution of antisera at which 50% of bacteria die - see for example PCT/EP98/Q5U7]) against a set of heterologous strains to be selected from the major clonal groups, known. For meningococcus B these stains should have a different PorA type from the bleb production strain, and should preferably be 2, 3, 4 or, most preferably, all 5 of strains H44/76, M97/252078, BZ10, NGPI65 and CU385. For non-typeable H. influenzae, the sxains should preferably be 2, 3, 4 or, most preferably, all 5 of strains 3224A, 3219C, 3241A, 640645, and A840177, For M. catarrhalis, the strains should preferably be 2, 3. 4 or, most preferably, all 5 of strains ATCC 43617, 14, 35S, 216 and 2926. By nan-toxic it is meant that there is a significant (2-4 fold, preferably 10 fold) decrease of endotoxin activity as measured by the well-known LAL and pyrogcnicity assays. Vaccine Combinations A further aspect of the invention are vaccine combinations comprising the bleb preparations of the invention with other antigens which are advantageously used against certain disease states. It has been found that blebs are particularly suitable for formulating with other antigens, as they advantageously have an adjuvant effect on the antigens they are mixed with. In one preferred combination, the meningoccocus B bleb preparations of the invention are formulated with 1, 2, 3 or preferably all 4 of the following meningococcal capsular polysaccharides which may be plain or conjugated to s protein carrier: A, C, Y or W. Such a vaccine may be advantageously used as a global meningococcus vaccine. Rather than use the meningoccocus B bleb preparations of the invention, it is also envisaged that the formulation could alternatively contain wild-type meningococcus B bleb preparations from 2 or more (preferably several) strains belonging to several subtype/serotypes (for instance chosen from PI.15, P1.7,16,P1.4,andP1.2). In a further preferred embodiment, the meningoccocus B bleb preparations of the invention [or the aforementioned mix of 2 or more wild-type meningococcus B bleb preparations], preferably formulated with 1, 2, 3 or ail 4 of the plain or conjugated meningococcal capsular polysaccharides A, C, Y or W, are formulated with a conjugated H. influenzae b capsular polysaccharide, and one or more plain or conjugated pneumococcal capsular polysaccharides. Optionally, the vaccine may also comprises one or more protein antigens that can protect a host against Streptococcus pneumoniae infection. Such a vaccine may be advantageously used as a global meningitis vaccine. The pneumococcal capsular polysaccharide antigens are preferably selected from serotype; I, 2, 3, 4, 5, 6B, 7F, S, 9N, 9V, I0A, IIA, I2F, 14, 15B, I7F, I8C, 19A. 19F, 20. 22F, 23F and 33F (most preferably from serotypes I, 3. 4, 5, 6B, 7F, 9V, 14, 18C, 19Fand23Fj. Preferred pneumococcal proteins antigens are those pneumococcal proteins which are exposed on the outer surface of the pneumococcus (capable of being recognised by a host's immune system during at least pan of the life cycle of the pneumococcus). or are proteins which are secreted or released by the pneumococcus. Most preferably, the protein is a toxin, adhesm, 2-component signal tranducer, or lipoprotein of Streptococcus pneumoniae, or fragments thereof. Particularly preferred proteins include, but are not limited to-, pneumolysin (preferably detoxified by chemical treatment or mutation) [Mitchell et a!. Nucleic Acids Res. 1990 Jul 11; 18(13): 4010 "Comparison of pneumolysin genes and proteins from Streptococcus pneumoniae types 1 and 2.", Mitchell et al. Biochim Biophys Acta 1989 Jan 23; 1007(1): 67-72 "Expression of the pneumolysin gene in Escherichia coli: rapid purification and biological properties.", WO 96/05859 (A. Cyanamid), WO 90/06951 (Paton et al), WO 99/03884 (NAVA)]; PspA and transmembrane deletion variants : thereof. (US. 5804123 - Briles et al.); PspC. and transmembrane deletion variants thereof (WO 97/09994 - Briles et al); PsaA and transmembrane deletion variants thereof (Berry & Paton, Infect Immun 1996 Dec;64(12):5255-62 "Sequence heterogeneity of PsaA, a 37-kilodaiton putative adhesin essential for virulence of Streptococcus pneumoniae"); pneumococcal choline binding proteins and transmembrane deletion variants thereof; CbpA and transmembrane deletion variants thereof (WO 97/41151; WO 99/51266); Glyceraldehyde-3-phosphate -dehydrogenase (Infect. Immun. 1996 64:3544); HSP70 (WO 96/40928); PcpA (Sanchez-Beato et al. FEMS Microbiol Lett 1998, 164:207-14); M like protein, SB patent application No. EP OS37130; and adhesin 1S627, SB Patent application No. EP 0834568. Further preferred pneumococcal protein antigens are those disclosed in WO 98/18931, particularly those selected in WO 98/18930 and PCT/US99'30390. In a further preferred embodiment, the Moraxella catarrhalis bleb preparations of the invention are formulated with one or more plain or conjugated pneumococcal capsular polysaccharides, and one or more antigens that can protect a host against non-rypeable H influenzae infection. Optionally, the vaccine may also comprise one or more protein antigens that can protect a host against Streptococcus pneumoniae infection. The -\aiciri£ may also optionally comprise one or more antigens that can protect a host against RSV and/or one or more antigens that can protect a host against influenza vims. Such a vaccine may be advantageously used as a global otitis media vaccine. Preferred non-typeable H. influenzae protein antigens include Fimbrin protein (US 5766608) and fusions comprising peptides therefrom (eg LBl Fusion) (US 5843464 - Ohio State Research Foundation), OMP26, P6, protein D, TbpA, TbpB, Hia, Hmwl, Hmw2, Hap, and D15. Preferred influenza virus antigens include whole, live or inactivated virus, split influenza virus, grown in eggs or MDCK cells, or Vero cells or whole flu virosomes (as described by R. Gluck, Vaccine, 1992, 10, 915-920) or purified or recombinant proteins thereof, such as HA, NP, NA. or M proteins, or combinations thereof. Preferred RSV (Respiratory Syncytia! Virus) antigens include the F glycoprotein, the G glycoprotein, the HN protein, or derivatives thereof. In a still further preferred embodiment, the non-typeable H. influenzae bleb preparations of the invention are formulated with one ->r more plain or corjugated pneumococcal capsular polysaccharides, and one or more antigens that can protect a host against M. catarrhaiis infection. Optionally, the vaccine may also comprise one or more protein antigens that can protect a host against Streptococcus pneumoniae infection. The vaccine may also optionally comprise one or more antigens that can protect a host against RSV and/or one or more antigens that can protect a host against influenza virus. Such a vaccine may be advantageously used as a global otitis media vaccine. Nucleotide sequences of the invention A further aspect of the invention relates to the provision of new nucleotide sequences which may be used in the processes of the invention. Specific upstream regions from various genes from various strains are provided which can be used in, for instance, processes a), b), d) and h). In addition, coding regions are provided for performing process d). General method for the analysis of the non-coding flanking region of a bacteria! gene, and us exploitation for modulated exp'ession of the gene in blebs The non-coding flanking regions of a specific gene contain regulatory elements important in the expression of the gene. This regulation takes place both at the transcriptional and translational level. The sequence of these regions, either upstream or downstream of the open reading frame of the gene, can be obtained by DNA sequencing. This sequence information allows the determination of potential regulatory motifs such as the different promoter elements, terminator sequences, inducible sequence elements, repressors, elements responsible for phase variation, die Shine-Dalgamo sequence, regions with potential secondary structure involved in regulation, as well as other types of regulatory motifs or sequences. This sequence information allows the modulation of the natural expression of the gene in question. The upregulation of the gene expression may be accomplished by altering the promoter, the Shine-Dalgamo sequence, potential repressor or operator elements, or any other elements involved. Likewise, downregularion of expression can be achieved by similar typs* of modifications. Alternatively, by changing phase variation sequences, the expression of the gene can be put under phase variation control, or may be uncoupled from mis regulation. In another approach, the expression of the gene can be put under the control of one or more inducible elements allowing regulated expression. Examples of such regulation includes, but is not limited to, induction by temperature shift, addition of inductor substrates like selected carbohydrates or their derivatives, trace elements, vitamins, co-factors, metal ions, etc. Such modifications as described above can be introduced by several different means. The modification of sequences involved in gene expression can be done in vivo by random mutagenesis followed by selection for the desired phenotype. Another approach consists in isolating the region of interest and modifying it by random mutagenesis, or site-directed replacement, insertion or deletion mutagenesis. The modified region can then be reintroduced into the bacterial genome by homologous recombination, and the effect on gene expression can be assessed. In another approach, the sequence knowledge of the region of interest can be used to replace or delete all or pan of the natural regulatory sequences. In this case, the regulatory region targeted is isolated and modified so as to contain the regulatory elements from another gene, a combination cf regulator/ elements from different genes, a synthetic regulatory region, or any other regulatory region, or to delete selected parts of the wild-type regulatory sequences. These modified sequences can then be reintroduced into the bacterium via homologous recombination into the genome. In process b), for example, the expression of a gene can be modulated by exchanging its promoter with a stronger promoter (through isolating the upstream sequence of the gene, in vitro modification of this sequence, and reintroduction into the genome by homologous recombination). Upregulated expression can be obtained in both the bacterium as well as in the outer membrane vesicles shed (or made) from the bacterium. In other preferred examples, the described approaches can be used to generate recombinant bacterial strains with improved characteristics for vaccine applications, as described above. These can be, but are not limited to, attenuated strains, strains with increased expression of selected antigens, strains with knock-outs (or decreased expression) of genes interfering with the immune response, and strains with modulated expression of immunodominant proteins. SEQ ID NO:2-23, 25, 27-38 are all Neisseria! upstream sequences (upstream of the initiation codon of various preferred genes) comprising approximately 1000 bp each. SEQ ID NO: 39-62 are all M. catarrhalis upstream sequences (upstream of the initiation codon of various preferred genes) comprising approximately 1000 bp each. SEQ ID NO: 63-75 are all H. influenzae upstream sequences (upstream of the initiation codon of various preferred genes) comprising approximately 1000 bp each. All of these can be used in genetic methods (particularly homologous recombination) for up-regulating, or down-regulating the open reading frames to which they are associated (as described before). SEQ ID NO: 76-81 are the coding regions for the HtrB and MsbB genes from Neisseria, M. catarrhalis, and Haemophilus influenzae. These can be used in genetic methods (particularly homologous recombination) for down-regulating (in particular deleting) part (preferably all) of these genes [process d)j. Another aspect of the invention is thus an isolated polynucleotide sequence which hybridises under highly stringent conditions to at least a 30 nucleotide portion of the nucleotides in SEQ ID NO; 2-23. 25, 27-81 or a complementary strand thereof Preferably the isolated sequence should be long enough to perform homologous recombination with the chromosomal sequence if it is pan of a suitable vector -namely at least 30 nucleotides (preferably at least 40, 50, 60, 70, 80, 90, 100, 200, 300, 400, or 500 nucleotides). More preferably the isolated po)ynucleoti As used herein, highly stringent hybridization conditions include, for example, 6X SSC, 5X Denhardt, 0.5% SDS, and 100 p-g/mL fragmented and denatured salmon sperm DNA hybridized overnight at 65 °C and washed in 2X SSC, 0.1% SDS one time at room temperature for about 10 minutes followed by one time at 65 °C for about 15 minutes followed by at least one wash in 0.2X SCC, 0.1% SDS at room temperature for at least 3-5 minutes. A further aspect is the use of the isolated polynucleotide sequences of the invention in performing a genetic engineering event (such as transposon insertion, or site specific mutation or deletion, but preferably a homologous recombination event) within 1000 bp upstream of a Gram-negative bacterial chromosomal gene in order to either increase or decrease expression of the gene. Preferably the strain in which the recombination event is to take place is the same as the strain from which the upstream sequences of the invention were obtained. However, the meningococcus A, B, C, Y and W and gonococcus genomes are sufficiently similar that upstream sequence from any of these strains may be suitable for designing vectors for performing such events in the other strains. This is may also be the case for Haemophilus influenzae and non-typeable Haemophilus influenzae. EXAMPLES The examples below are carried out using standard techniques, which are well known and routine to those of skill in the art, except where otherwise described in detail. The examples are illustrative, but do not limit the invention. Example 1: Construction of a Neisseiria meningitidis serogroup B strain lacking capsular polysaccharides. The plasmid pMF121 (Frosch et al., 1990) has been used to construct a Neisseria meningitidis B strain lacking the capsular polysaccharide. This plasmid contains the flanking regions of the gene locus coding for the biosynthesis pathway of the group B polysaccharide (B PS), and the erythromycin resistance gene. Deletion of the B PS resulted in loss of expression of the group B capsular polysaccharide as well as a deletion in the active copy ofgalE leading to the synthesis of galactose deficient LPS. Strain transfarmation: Neisseria meningitidis B H44/76 strain (B:15:P1.7, 16;Los 3,7,9) was selected for transformation. After an overnight C02 incubation on MH plate (without erythromycin), cells were collected in liquid MH containing 10 mM MgCl; (2 ml were used per MH plate) and diluted up to an OD of 0.1 (550 nra). To this 2 ml solution, 4 ul of the plasmid pMFl21 stock solution (0.5 ug/'ml) were added for a 6 hours incubation period at 37°C (with shaking). A control group was done with the same amount of Neisseria meningitidis B bacteria, but without addition of plasmid. After the incubation period, 100 fi] of culture, as such, at 1/10, 1/100 and 1/1000 dilutions, were put in MH plates containing 5, 10, 20, 40 or BO ug erythromycin /ml before incubation for 48 hours at 3?°C. Colony blotting: After plate incubation, 20 colonies were grown and selected from the 10 and 20 ug erythromycin/ml MH plates, while there was no colony growth in the control group without plasmid transformation. The H44/76 wild type strain was unable to grow in the selected erythromycin plates (10 to 80 jig erythromycin/ml). The day after, all the visible colonies were placed on new MH piates without erythromycin in order to let them grow. Afterwards, they were transferred onto nitrocellulose sheets (colony blotting) for presence of B polysaccharide. Briefly, colonies were blotted onto a nitrocellulose sheet and rinsed directly in PBS-0.05 % Tween 20 before cell inactivation for 1 hour at 56°C in PBS-0.05% Tween 20 (diiuant buffer). Afterwards, the membrane was overlaid for one hour in the diiuant buffer at room temperature (RT). Then, sheets were washed again for three times 5 minutes in the diiuant buffer before incubation with the anti-B PS 735 Mab (Boerhinger) diluted at 1/3000 in the diiuant buffer for 2 hours at RT. After a new washing step (3 times 5 minutes), the monoclonal antibody was detected with a btorinylated anti-mouse Ig from Amersham (RPN 1001) diluted 500 times in the diiuant buffer (one hour at RT) before the next washing step (as described above). Afterwards, sheets were incubated for one hour at RT with a solution of streptavidin-peroxidase complex diluted 1/1000 m the diiuant buffer. After the lasr three washing steps using the same method, nitrocellulose sheets were incubated for 15 min in the dark using the revelation solution (30 mg of 4-chioro-1-naphtol solution in 10 ml methanol plus 40 ml PBS and 30 mcl of H,0: 37% from Merck). The reaction was stopped with a distillated water-washing step. Whole cell Elisas: Whole cell Elisas were also done using the two transformed colonies ("D" and "R") and the wild type strain (H44/76) as coated bacteria (20 fig protein/ml), and a set of different monoclonal antibodies used tc characterize Neisseria meningitidis strains. The following Mabs were tested: anti-B PS (735 from Dr Frosch), and the other Mabs from NIBSC: anti-B PS (Ref 95/750) anti-P1.7 (A-PorA, Ref 4025), anti-Pl.16 (A-PorA, Ref 95"20), anti-Los 3,7,9 (A-LPS, Ref 4047), anti-Los 8 (A-LPS, Ref 4048), and anti-P1.2 (A-PorA Ref 95'696). Micror.ter plates (Maxisorp. Nunc) were coated with 100 jii of the recombinant meningococcal B cells solution overnight (ON) at 37DC at around 20 ug/ml in PBS. Afterwards, piates are washed three times with 300 ul of 150 mM NaCi - 0.05 * o Tween 20. and were overlaid with 100 ul of PBS-0.3 % Casein and incubated for 30 min at room temperature with shaking. Plates were washed again using the same procedure before incubation with antibodies. Monoclonal antibodies (100 ul) were used at different dilutions (as shown in Figure 2) in PBS-0.3 % Casein 0.05 % Tween 20 and put onto the cnicropiates before incubation at room temperature for 30 min with shaking, before the next identical washing step. 100 ul of the anti-mouse Ig (from rabbit, Dakopatts E04I3) conjugated to biocin and diluted at 1/2000 in PBS-0.3 % Casein - 0.05 % Tween 20 were added to the wells to detect bound monoclonal antibodies. After the washing step (as before), plates were incubated with a streptavidin-peroxidase complex solution (100 ul of the Amersham RPN 1051) diluted at 1/4000 in the same working solution for 30 min at room temperature undeT shaking conditions. After this incubation and the last washing step, plates are incubated with 100 ui of the chromogeD solution (4 mg orthophenylenediamine (OPD) in 10 ml 0.1 M citrate buffer pH4.5 with 5 ui HjO^ for 15 min in the dark. Plates are then read at 490/620 nm using a spectrophotometer. Results: Figure 1 shows that from the 20 isolated colonies, which were able to growth on the selected medium with erythromycin, only two (the "D" and the "R") colonies were shown negative for presence of B polysaccharide. Among the others, 16 were clearly positive for B PS and still resistant to erythromycin. This indicated that they integrated the plasmid into their genome, but in the wrong orientation, and keeping intact the B PS and LPS gene (no double crossing-over). Positive and negative controls were also tested on the plates, and showed that the H44/76 wild type NmB strain was clearly positive for the B polysaccharide, while meningococcus A (Al) and meningococcus C (CLl) strains were clearly negative with this anti-B PS 735 Mab. These results indicate that around 10 % of the selected colonies correctly integrated the plasmid in their genome by making a double crossing-over, while the other strains/colonies were obtained after a simple crossing-over, keeping the B PS and LPS genes intact and expressed. Using whole eel! Elisa, results (Figure 2 and the Table below) ciearly indicate that the two "D" and "R" transformants (derived from D and R colonies) can not be recognized anymore by the anti-B PS Mabs (735 and 95/750), nor by the anti-Los 3,7,9 and anti-Los 8 Mabs. However, when using specific anti-PorA Mabs, there is a clear reaction with the anti-P1.7 and anti-P 1.16 Mabs on the cells, as aiso observed in ihe wild-type strain. No reaction was observed with a non-specific anti-PorA Mab (anti-P 1.2 mab). These results confirm that the PorA protein, and specifically P1.7 and PI.16 epitopes are still present after transformation, while B polysaccharide and Los 3.7,9 and Los 8 epitopes (LPS) were not. A plasmid allowing homologous recombination and stable integrator, of foreign DNA in the porA iocus. of Neisseria meningitidis was constructed. This delivery vector (genes, operorcs and'br expression cassettes) is useful for constructing Neisseiria meningitidis strains producing recombinant, improved blebs. Typically, such a vector contains at least: (1) a plasmid backbone replicative in E. coli but not in Neisseria meningitidis (a suicide plasmid), (2) at least one, but preferably two regions 4t of homology for targeting the integration in a chromosomal locus such as porA, (3) Efficient transcriptional (promoter, regulatory region and terminator) and translational {optimised ribosome binding site and initiation codon) signals functional in Neisseria meningitidis, (4) a multiple cloning site and (5) selectable gene(s) allowing the maintenance of the plasmid in E. coli and the selection of integrants in Neisseria meningitidis. Additional elements include, for example, uptake sequences to facilitate the entry of foreign DNA in Neisseiria meningitidis, and counter selectable markers such assacB, rpsl,gltS to enhance the frequency of double cross-over events. A schematic drawing of the vector constructed in this example and designated pCMK is represented in Figure 3. Its corresponding complete nucleotide sequence is shown in SEQ. ID NO: I. pCMK derives from a pSLI180 backbone (PharmaciaBiotech, Sweeden), a high copy-number plasmid replicarive in E. coli, harbouring the bla gene (and thereby conferring resistance to ampicillin). In addition to this, pCMK functionally contains two porA flanking regions (porA5' afld porA3' containing a transcription terminator) necessary for homologous recombination, a selectable marker conferring resistance to kanamycin, two uptake sequences, a porA/lacO chimeric promoter repressed in the E.coli host expressing lacN but transcriptionally active in Neisseria meningitidis, and a multiple cloning site (5 sites present: Ndel , Kpnl, Nhel, Pf.nA 1 and Sphl) necessary for the insertion of foreign DNA in pCMK. pCMK was constructed as follows. The porA5' and par A3' recombinogenic regions, the porA/lacO promoter were PCR amplified using the oligonucleotides listed in the table below, cloned in pTOPO and sequenced. These DNA fragments were successively excised from pTOPO and recloned in pSLI 180. The kanamycin resistance cassette was excised from pUC4K (PharmaciaBiotech, Sweeden) and was introduced between the porA5' flanking region and the porA/lacO promoter region. Example 3: Construction of a Neisseiria meningitidis serogroup B strain lacking both capsular polysaccharides and the major immunodominant antigen PorA. Modulating the antigenic content of outer membrane blebs may be advantageous in improving their safety and efficacy in their use in vaccines, or diagnostic or therapeutic uses. Components such as the Neisseiria meningitidis serogroup B capsular polysaccharides should be removed to exclude the risk of inducing autoimmunity (see example 1). Similarly, it is beneficial to suppress the immunodominanee of major outer-membrane antigens such as PorA, which induce strain-specific bactericidal antibodies but fail to confer cross-protection. To illustrate such an approach, we used the pCMK(+) vector to construct a Neisseiria meningitidis serogroup B strain lacking both capsular polysaccharides and the immunodominant PorA outer membrane protein antigen. For this purpose, a deletion of the porA gene was introduced in the H44/76 cps- strain, described in example 1 by means of homologous recombination. The H44/76 cps- strain was prepared competent and transformed with two 2\ig of supercoiled pCMK.(+) plasmid DNA as described previously. Aliquot fractions of the transformation mixture (lOOjil) were plated on Mueller-Hinton plates supplemented with Kanamycin (200u.g/ml) and incubated at 37°C for 24 to 48 hours. Kanamycin-resistant colonies were selected, restreaked on MH-Kn and grown for an additional 24 hours at 37°C. At that stage half of the bacterial culture was used to prepare glycerol stocks (15 % voI./voI.) and was kept frozen at -70°C. Another fraction (estimated to be 108 bacteria) was resuspended in 15 p.1 of distilled water, boiled ten minutes and used as a template for PCR screening. Two porA internal primers named , PPA1 and PPA2, were synthesized and used to perform PCR amplification on boiled bacterial lysates in the conditions described by the supplier (HiFi DN'A polymerase, Boehringer Mannheim, GmbH). The thermal cycling used was the following: 25 times (94°C Imin., 52°C lmin., 72°C 3mm.) and 1 time (72"C 10mm., 4°C up to recover.'). Since a double crossing-over between pCMK DNA and the chromosomal porA locus deletes the region required for #1 and ~2 annealing, clones lacking a 1170bp PCR amplification fragment were selected as porA deletion mutants. These PCR results were further confirmed by analyzing in parallel, the presence of PorA in the corresponding bacteria! protein extracts. For that purpose, another aliquot of bacteria (estimated to be 5.10s bacteria) was re-suspended in 50 p.1 of PAGE-SDS buffer (SDS 5%, Glycerol 30%, Beta-mercaptoethano] 15%, Bromophenol blue 0.3mg/ml, Tns-HCI 250 mM pH6.8), boiled (100°C)frozen(-20°C} / boiled (100°C) three times and was separated by PAGE-SDS electrophoresis on a 12.5 % gel. Gels were then stained by Cootnassie Brilliant blue R250 or transferred to a nitrocellulose membrane and probed with an anti-PorA monoclonal antibody as described in Maniatis et ai. As represented in Figure 4, both Coomasste and immunoblot staining confirmed that porA PCR negative clones do not produce detectable levels of PorA. This result confirm that the pCMK vector is functional and can be used successfully to target DNA insertion in the porA gene, abolishing concomitantly the production of the PorA outer membrane protein antigen. Example 4: Up-regularion of the NspA outer membrane protein production in blebs derived from a recombinant Neisseria meningitidis serogroup B strain lacking functional porA and cps genes. Enriching bleb vesicles with protective antigens is advantageous for improving die efficiency and the coverage of outer membrane protein-based vaccines. In that context, recombinant Neisseria meningitidis strains lacking functional cps and porA genes were engineered so that the expressions level of the outer-membrane protein NspA was up-regulated. For that purpose, the gene coding for NspA was PCR amplified using the NOV-full-Ndel and JVdeI-3'oligonucleotide primers (see table in example 2). The conditions used for PCR amplification v/ere those described by the supplier (HiFi DNA polymerase, Boehringer Mannheim, GmbH). Thermal cycling done was the following: 25 rimes (94°C lmin.. 52°C Imin., 72°C 3mm.) and 1 rime (72°C 10mm., 4CC up to recovery). The corresponding amplicon was digested with Ndel and inserted in the Ndel restriction site of the pCMK(+) deliver,' vector. Insert orientation was checked and recombinant plasmids, designed pCMK(+)-NspA, were purified at a large scale using the QIAGEN maxiprep kit and 2 jig of this material was used to transform a Neisseria meningitidis serogroup B strain lacking functional cps genes {strain described in example I). Integration resulting from a double crossing-over between the pCMK(+)-NspA vector and the chromosomal porA iocus were selected using a combination of PCR and Western blot screening procedures presented in example 3. Bacteria (corresponding to about 5.10' bacteria) were re-suspended in 50 y\ of PAGE-SDS buffer, frozen(-20°C) / boiled (1006C) three times and then were separated by PAGE-SDS electrophoresis on a 12.5 % gel. Gels were then stained by Coomassie Brilliant blue R250 or transferred to a nitrocellulose membrane and probed with an anti-NspA polyclonal serum. Both Coomassie (data not shown} and immunoblot staining (see figure 4) confirmed that porA PCR negative clones do not produce detectable levels of PorA. The expression of NspA was examined in Whole-cell bacterial lysates (WCBL) or outer-membrane bleb preparations derived from NmB [cps-, porA-] or NmB [cps-, porA~, Nspa+). Although no difference was observable by Coomassie staining, immunoblotting with the anti-NspA polyclonal serum detected a 3-5 fold increased in the expression of NspA (with respect to the endogenous NspA level), both in WCBL and outer-membrane bleb preparations (see figure 5). This result confirm that the pCMK(+)-NspA vector is functional and can be used successfully to up-regulate the expression of outer membrane proteins such as NspA, abolishing concomitantly the production of the PorA outer membrane protein antigen. Example 5: Up-regulation of the D15/Omp85 outer membrane protein antigen in blebs derived from a recombinant Neisseria meningitidis serpgroup B strain lacking functional cps genes but expressing PorA. Certain geographically isolated human populations (such as Cuba) are infected by a limited number of Neisseria meningitidis isolates belonging largely to one or few outer membrane protein serotypes. Since PorA is a major outer-membrane protein antigen inducing protective and strain-specific bactericidal antibodies, it is then possible to confer vaccine protection using a limited number of porA serotypes in a vaccine. In such a context, the presence of PorA in outer membrane vesicles may be advantageous, strengthening the vaccine efficacy of such recombinant improved blebs. Such PorA containing vaccines, however, can be improved still further by increasing the level of other cross-reactive OMPs such as ompS5/D 15. In the following example, the pCMK(+) vector was used to up-regulate the expression of the Omp85/D15 outer membrane protein antigen in a strain lacking functional cps genes but expressing porA. For thai purpose, the gene coding for Omp85/D15 was PCR amplified using the D15-AWeI and D\5-Notl oligonucleotide primers. The conditions used for PCR amplification were those described by the supplier (HiFi DNA polymerase, Boehringer Mannheim, GmbH). Thermal cycling done was the following: 25 times (94°C Imin., 52°C 1mm., 72°C 3min.) and I time (72°C 10mm., 4°C up to recovery). The corresponding amplicon was inserted in the pTOPO cloning vector according to the manufacturer's specifications and confirmatory sequencing was performed. This Omp85/D15 DNA fragment was excised from pTOPO by restriction hydrolysis using NdeVNsil and subsequently elonsd in the corresponding restriction sites of die pCMK(+) aHivsry "ector Recombinant plasmids, designed pCMK(+)-D 15 were purified on a large scale using the QIAGEN maxiprep kit and 2 ug of this material was used to transform a Neisseiria meningitidis serogroup B strain lacking functional cps genes (strain described in example 1). In order to preserve the expression of porA, integration resulting from a single crossing-over (either in Omp85/D15 or in porA) were selected by a combination of PCR and Western blot screening procedures. Kanamycin resistant clones testing positive by/jor^-specific PCR and western blot were stored at -70°C as glycerol stocks and used for further studies. Bacteria (corresponding to about 5.10s bacteria) were re-suspended in 50 ul of PAGE-SDS buffer, frozen(-20°C) / boiled (100°C) three times and then were separated by PAGE-SDS electrophoresis on a 12.5 % gel. Gels were then stained by Coomassie Brilliant blue R250 or transferred to a nitrocellulose membrane and probed with an anti-porA monoclonal antibody. As represented in Figure 6. both Coomassie and immunobbt staining confirmed that/?or.-4 PCR positive clones produce PorA. The expression of D15 was examined using outer-membrane bleb preparations derived from NrnB [cps-, porA-] or NmB [cps-, porA-, D15+]. Coomassie detected a significant increase in the expression of D15 (with respect to the endogenous D15 ievel), preparations (see Figure 6). This result confirmed that the pCMK(+)-DI5 vector is functional and can be used successfully to up-regulate the expression of outer membrane proteins such as DI5, without abolishing die production of the major PorA outer membrane protein antigen-Example 6: Construction of versatile promoter delivery vectors Rational: The rational of this approach is represented in Figure 7 and can be summarized in 7 essential steps. Some of these steps are illustrated below with the construction of Vector for up-regulating the expression of NspA and D15/Omp85. Vector for up-regulating the expression of the NspA gene. Step 1. A DNA region (997bp) located upstream from the NspA coding gene was discovered (SEQ. ID NO'.2) in the private Incyte PathoSeq data base containing unfinished genomic DNA sequences of the Neisseria meningitidis strain ATCC 13090. Two oligonucleotide primers referred to as PNS1 and PNS2 (see table in example 2) were designed using this sequence and synthesized. These primers were used for PCR amplification using genomic DNA extracted from the H44/76 strain. Step 2. The corresponding amplicons were cleaned-up using the Wizard PCR kit (Promega, USA) and submitted to digestion with the EcoKUXbal restriction enzymes for 24 hours using the conditions described by the supplier (Boehringer Mannheim, Germany). The corresponding DNA fragments were gel purified and inserted in the corresponding sites of thepUCl8 cloning vector. Step 3. Recombinant plasmids were prepared on a large scale and an aliquot fraction was used as a template for inverse PCR amplification. Inverse PCR was performed using the PNS4 and PNS5 oligonucleotides using the following thermal cycling conditions: 25 times (94°C Imin., 50°C lmin., "2°C 3mm.) and 1 time (72°C lOmin., 4°C up to recovery). Linearized pUC 18 vectors harbouring a deletion in me NspA upstream region insert were obtained. Vector for up-regulating the expression of theD15/ompS5 gene. Step 1. A DNA region (1000 bp) located upstream from the D15/omp85 coding gene was discovered (SEQ. ID NO:3) in the private Incyte PathoSeq database containing unfinished genomic DNA sequences of the Neisseria meningitidis strain ATCC 13090. Two oligonucleotide primers refererred to as PromDl5-5IX and PromD15-S2 (see table in example 2) were designed using this sequence and synthesized. These primers were used for PCR amplification using genomic DNA extracted from the H44/76 strain. Step 2. The corresponding ampiicons were cleaned-up using the Wizard PCR kit (Promega, USA) and submitted to digestion with the EcdRilXbal restriction enzymes for 24 hours in the conditions described by the supplier (Boehringer Mannheim, Germany). The corresponding DNA fragments were gel purified and inserted in the corresponding sites of the pUCI8 cloning vector. Step 3. Recombinant plasmids were prepared on a large scale and an aliquot fraction was used as a template for inverse PCR amplification. Inverse PCR was performed using the DI5-S4 and D15-S5 oligonucleotides using the following thermal cycling conditions: 25 times {94°C lmin., 50°C lmin., 72°C 3min.) and 1 time (72°C lOmin., 4DC up to ,-=.---,■..- ^ i/.icari-td puc is vectors harbouring a dtletior. in :hs DtS.'rmp^. upstream region insert were obtained. Example 7: Fermentation processes for producing recombinant blebs . The examples listed below describe methods for producing recombinant blebs lacking either capsular polysaccharides or capsular polysaccharides and PorA. Such a procedure may be used for a wide range of Neisseiria meningitidis recombinant strains and may be adapted over an extended scale range. Culture mcd\z:Neisseiria meningitidis serogroup B strains were propagated in solid (FNE 004 AA, FNE 010 AA) or liquid (FNE 008 AA) culture media. These new media for growing meningococcus are advantageiously free of animal products, and are considered a further aspect of the invention. *! Flask cultivation of Neisseiria meningitidis serogroup B cps- recombinant blebs: This was performed in two steps comprising precuiture an solid medium followed by liquid cultivation. Solid pre-culture A via! of seed was removed from freezer (-80°C), thawed to room temperature and 0.1 mL was streaked into a Petri dish containing 15 mL of FNE004AA (see above).Tbe Petri dish was incubated at 37DC for 18 ± 2 hours. The surface growth was resuspended in 8 mL of FNE008AA (see above) supplemented with 15 mg/L of erythromycin. Flask culture. 2 mL of resuspended solid pre-culture were added to a 2 litre flask containing 400 mL of FKE008AA supplemented with 15 mg/L of erythromycin. The flask was placed on a shaking table (200 rpm} and incubated at 37°C for 16 f 2 hours. The cells were separated from the culture broth by centrifugation at 5000g at 4°C for 15 minutes. Batch mode cultivation of Neisseiria meningitidis serogroup B est- recombinan" blebs: This was performed in three steps comprising precuiture on solid medium, liquid cultivation and batch mode cultivation. Solid pre-culture. A vial of seed was removed from freezer (-80"C). tbawed to room temperature and 0.! mL was streaked mto 2 Petri dish containing 15 mL of FNE004AA (see above). The Petri dish was separated from the culture broth by centrifugation at SOOOg at 4°C for 15 minutes. of expression in bacteria. Since promoter regulatory elements are classically encompassed within 200 bp upstream and 50bp dowtream from the -I site (Coliado-Vides J, Magasanik B, Gralla JD, 1991, Microbiol Rev 55(3):371-94), the result of such an experiment allows us to identify DNA fragments of about 250 bp carrying strong promoter activities. Major outer membrane proteins such as Neisseria meningitidis PorA, PorB &. Rmp, Haemophilus influenzae PI, P2, P5 &. P6, Moraxella caiarrhalis OmpCD, OmpE, as well as some cyoplasmic and/or iron regulated proteins of these bacteria possess strong promoter elements. As a validation of this general methodology, we mapped the transcriptional start site of the strong Neisseria meningitidis par A and porB promoters using rapid amplification of cDNA elements (51 RACE). The principles of 5' RACE are the following: 1) Total RNA extraction using QIAGEN "RNeasy" Kit. Genomic DNA removing by DNase treatment followed by QIAGEN purification; 2) mRNA reverse transcription with a porA specific 3' end primer i,nainea pt,rA3); Expected cDNA" size: 307 nt. RNA removing by "alkaline hydrolysis; 3) Ligation of a single-stranded DNA oligo anchor (named DT88) to the 3r end of the cDNA using T4 RNA ligase. Expected product size: 335 nt. Amplification of the anchor-!igated cDNA using a combination of hemi-nested PCR; 4) PCR amplification of the anchor-Iigated cDNA using a complementary-sequence anchor primer as the 5' end primer (named DT89) and a Tend primer (named pl-2) wbich is internal to the 3'end RT primer porA3. Expected product size: 292 bp; 5) PCR amplification of previous PCR products using DT89 as 5'end primer and pi -1 as 3'end primer (internal to pl-2). Expected product size: 21Ibp; and 6) Sequencing witbpl-1 primer (expected products size can be calculated because porA transcription start site is known: 59 ti: before the " ATG" translation start site). Experimental procedure Total RNA was extractsd from approximately 10^ cells of Neisseria meningitidis seroeroup B cos- porA + strain. Extraction of 1 ml of a liquid cuirure ai appropriate optical density' CODgrjo = !) was psrformsd by the QIAGEN "RNAsasy" kit according to the manufacturer's instructions. Chromosomal DNA was removed by addition of 10U of RNase-fre; DNase (Ro;he Diagnostics, Mannheim, Germany) TO the 30 y.\ of eluted RNA and was incubated at 37°C for 15 min. The DNA-free RNA was purified with the same QIAGEN kit according to instructions. Reverse transcription reactions were performed using primer porAJ and 200U of SUPERSCRIPT II reverse transcriptase (Life Technologies). The RT reactions were performed in a SOul volume containing: 5ul of 2mM dNTP, 20 pmol of porA3 pimer, 5ul of 10X SUPERSCRIPT II buffer, 9^1 of 25mM MgC12, 4ul of 0.1M DTT, 40L' of recombinant ribonuclease inhibitor and 1 ug of total RNA. The porA3 primer was annealed stepwise (70°C for 2 min, 65'C for 1 min, 60°C for I min, 55°C for i min, 50°C for 1 min, and 45°C for i min) before the SUPERSCRIPT II was added. The RT reaction was performed at 42°C for 30 min, followed by 5 cycles (50°C for J min, 53"C for 1 min and 56°C for 1 min) ro destabilize RNA secondary structure. Two parallel reactions were performed with the reverse transcriptase omitted from one reaction as negative control. The RNA was removed by alkaline hydrolysis cleavage with the addition of 1 ul of 0.5M EDTA followed by 12.5 ul of 0.2 M NaOH before incubation at 68°C for 5 min. The regions ":r^ neutralized by adding ,2.5 -.! of 1 M Trii-HCl (pH7.4) a?,r, precipitated cy the addition of 20 jig of glycogen (Roche Molecular Biochemicals, Mannheim, Germany), 5 u.1 of 3 M sodium acetate and 60 ul of isopropanal. Both samples were resuspended in 20 ul of 10:lTE(10mMTris-HCl,pH7.4; 1 mM EDTA.pHS). T4 RNA Ugase was used to anchor a 5'-phosphorylated, 3'end ddCTP-blocked anchor oligonucleotide DT88 (see table below). Two parallel ligations were performed overnight at room temperature with each containing'. 1.3 ul of 10X RNA ligase buffer (Roche Molecular Biochemicals), 0.4 uM DT88, 10 ul of either cDNA or RT control sample and 3 U of T4 RNA ligase. As negative controls, a second set of ligations reactions was performed, omitting the T4 RNA Sigase. The resulting ligation-reaction mixtures were used directly without purification in the subsequent PCR. The anchor-ligated cDNA was amplified using a combination of hemi-nested and hot-started PCR approaches to increase specificity and product yieid. Four separate first-round PCR were performed on the RT/ligase reaction and controls in a 30 ul volume, each containing: 3 >il of lOXTacPlannium buffer, 3ul of 25 mtvlMgCl2, 1 ul of :0mM dNTP, 10 pmol of each primers and 1 ul of corresponding RNA ligation reaction. The PCR were hot started by the use of Taq Platmium (Life Technologies) DNA polymerase (2U added). The first ligation-anchored PCR (LA-PCR) was performed using 10 pmol of both the anchor-specific primer DT89 and the transcript-specific primer pl-2 (see table below) which is for 1 min (reducing one degree per cycle), 15 cycles at 95°C for 10 s and 60°C for 1 mm. The second hemi-nested LA-PCR was performed under the same conditions using primer DT89 andtbepI-2 internal primer, together with 10 pmo! of pl-1 (see table below) and I ^1 of first-round PCFL Amplification products were purified using the QIAGEN "QIAquick PCR purification" kit according to manufacturer instructions before submitted to sequencing. The CEQTM Dye Terminator Cycle Sequencing kit (Beckman, France) was used to sequence the RACE PCR products using 10 pmol of primer pi-1. Sequencing reactions were performed according to the provided instructions and sequencing products were analyzed by the Ceq2000 DNA Analysis System (Bee km an-Coulter). codon, this result supports that the porB transcriptional start site is located about 50 bp (+/- 30 bp) upstream of the porB ATG. The exact nucleotide corresponding to transcription initiation is presently being determined by DNA sequencing. The above PCR result suppons that a DNA fragment containing nucleotides -1 to -250 with regard to theporB ATG start codon is suitable for driving strong gene expression in Neisseria meningitidis and possibly in other bacterial species such as Haemophilus, Moraxella, Pseudomonas. Example 10: Up-regulatiori of the N. meningitidis sarogroup B OmpSS gene by promoter replacement. The aim of me experiment was to replace the endogenous promoter region of the DI5/Omp85 gene by the strong porA promoter in order to up-regulate the production of the D15/OmpS5 antigen. For that purpose, a promoter replacement plasmid was constructed using E. coh cloning methodologies. A DNA region (1000 bp-) located upstream from the DI5/ompS5 coding gene was discovered fSEQ ID NO:3) in the private Incyte PathoSeq data base containing unfinished genomic DNA sequences of the Neisseria meningitidis, strain ATCC L3090. The main steps of this procedure are represented in Figure 9. Briefly, a DNA fragment (lOOGbp) covering nucleotides -48 to -983 with respect to tbe D15/OmpS5 gene start cocion (ATG) was PCR amplified using oligonucleotides ProD15-51X (5'-GGG CGA ATT CGC GGC CGC CGT CAA CGG CAC ACC GTT G-3*) and ProDI5-52 fS'-OCT CTA GAG CGG AAT GCG GTT TCA GAC G-3') containing EcoBI and Xbal restriction sites (underlined) respectively, This fragment was submitted to restriction and inserted in pUC18 plasmid restricted with the same enzymes. The construct obtained was submitted to in vitro mutagenesis using the Genome Priming system (using the pGPS2 donor plasmid) commsrcialized by New England Biolabs (MA, USA). Clones having inserted a mini-transposon (derived from Tn7 and harboring a chloramphenicol resistance gene) were selected. One clone containing a mini-transposon insertion located in the Dl5/Omp85 5' flanking region, 401 bp downstream from the EcoRl site was isolated and used for further studies. This plasmid was submitted to circle PCR mutagenesis (Jones & Winistofer (1992), Biotechniquesll: 528-534) in order to (i) delete a repeated DNA sequence (Tn7R) generated by the transposition process, (ii) insert raenmsococcal uotake setmeiw: required for transfonnacion, and (iii) insert suitable restriction sites allowing cloning of foreign DNA material such as promoters. The circle PCR was performed using the TnRD15-KpnI/XbaI + US (5'-CGC CGG TAC CTC TAG AGC CGT CTG AAC CAC TCG TGG ACA ACC C-3') & TnR03Cam(KpnI) (5'-CGC CGG TAC CGC CGC TAA CTA TAA CGG TC-3') oligonucleotides containing uptake sequences and suitable restriction sites {Kpnl and Xbal) underlined. The resulting PCR fragment was gel-purified, digested with Asp7\8 (isoschizomer of Kpnl) and Hgated to a I84bp DNA fragment containing Che porA promoter and generated by PCR. using the PorA-01 (5'-CGC CGG TAC CGA GGT CTG CGC TTG AAT TGT G-3') and PorA02 (5'-CGC CGG TAC CTC TAG ACA TCG GGC AAA CAC CCG-3*) oligonucleotides containing Kpnl restriction sites. Recombinant clones carrying a porA promoter inserted in the correct orientation (transcription proceeding in the £coPJ to Xbal direction) were selected and used to transform a strain of Neisseria meninpitidi? serogroup B lacking capsular polysaccharide lrps-) and one of the major outer membrane proteins - PorA {porA-). Recombinant Neisseria meningitidis clones resulting from a double crossing over event (PCR screening using oligonucleotides Cam-05 (5'-GTA CTG CGA TGA GTG GCA GG-3") & proDlS-52) were selected on GC medium containing 5ug/ml chloramphenicol and analyzed for D15/Omp85 expression. As represented in Figure 10, the production of D15/Omp85 was significantly increased in the total protein extracts of Nm strains resulting from promoter replacement, when compared to parental strain (cps-). This result was also observed when analyzing outer-membrane blebs prepared from the same strains (see Figure 17). These results are attributable to the replacement of the endogenous D!5 promoter by the strong porA promoter. In addition, it was surprisingly found that expression, where the porA promoter was introduced approximately 400 bp upstream of the initiator codon, was approximately 50 times greater than when the promoter was placed approximately 100 bp upstream. Altogether, these experiments support that the promoter replacement strategy works and allows the up-regulation of the synthesis of integral outer-membrane proteins in outer -membrane blebs. Certain geographically isolated human populations (such as Cuba) are infected by a limited number of Neisseiria meningitidis isolates belonging largely to one or few outer membrane protein serotypes. Since PorA is a major outer-membrane protein antigen which can induce protective and strain-specific bactericidal antibodies, it may be possible to confer vaccine protection in such a population using a limited mimbei of porA serotypes. Moreover, PorA may interact with or stabilize some other outer membrane proteins. In this context, the presence of PorA. in outer membrane vesicles may be advantageous, strengthening the vaccine efficacy of such recombinant improved blebs. For such a reason, it may be desirable to up-regulate the expression of D15/Omp85 outer membrane protein in a Neisseria meningitidis serogroup B strain lacking functional cps genes but expressing PorA. Genomic DNA was extracted from the recombinant Neisseria meningitidis serogroup B cps-, porA-, D15/Omp85+ strain using the QIAGEN Genomic Tips 100-G kit. lOugr of this material was linearized and used to transform Neisseria meningitidis serogroup B cps- following a classical _ transformation, pr-orocol. Recombinant Neisseria were obtained on GC agar plates containing 5u.gr/ml chloramphenicol. Integrations resulting from a double crossing-over upstream of the DI5 gene were screened by PCR as described previously. As homologous recombinations can occur everywhere in the chromosome, a second PCR screening was performed to control the integrity of the porA locus in the recombinant strain. For this purpose, internal porA primers PPA1 (5- GCG GCC GTT GCC GAT GTC AGC C -3*) and PpA2 ( 5- GGC ATA GCT GAT GCG TGG AAC TGC -3' ) were used in a PCR screening experiment. The amplification of an II 70bp fragment confirms the presence of the porA gene in the recombinant bacteria. Recombinant bacteria (corresponding to about 5.10* bacteria) can be re-suspended in SO n-1 of PAGE-SDS buffer, frozen(-20°C) / boiled (IO0°C) three times and then separated by PAGE-SDS electrophoresis on a 12.5 % gel. Gels can then he Stained by Cootnassie Brilliant blue R250 or transferred to a nitrocellulose membrane and probed either with an anti-porA monoclonal antibody or with an anti-DI5/OmpS5 rabbit polyclonal antibody. Analysis of outer-membrane blebs prepared from the same strains can also be performed. Example 11: Up-regulation of the Hsf protein antigen in a recombinant Neisseiria meningitidis serogroup B strain lacking functional cps series but expressing PocA. As described above, in certain countries, toe presence of PorA in outer membrane vesicles may be advantageous, and can strengthen the vaccine efficacy of recombinant improved blebs. In the following example, we have used a modified pCMK(+) vector to up-regulate the expression of the Hsf protein antigen in a strain lacking functional cps genes but expressing PorA, The original pCMK(-t-) vector contains a chimeric porA/lacO promoter repressed in E. coH host expressing iac!R but transcriptionally active in Neisseria meningitidis. In the modified pCMK(+), the native porA promoter was used to drive the transcription of the faf gene. The gene coding for Hsf was PCR amplified using the HSF 01-NdeI and HSF 02-MeI oligonucleotide primers, presented in the table below. Because of the sequence of the HSF Ol'Ndel primer the Hsf protein expressed will contain two methionine residues at the 5' end. T1^ conditions ■:szd for PCR. amplification w^r; tho:-.? Scribed by the supplier (HiFi DNA polymerase, Boehringer Mannheim, GmbH). Thermal cycling was the following: 25 times (94°C lmin., 48°C lmin., 72°C 3mui.) and 1 time (72DC lOmin., 4DC up to recovery). The corresponding amplicon was subsequently cloned in the corresponding restriction sites of pCMK(+) delivery vector. In this recombinant plasmid, designed pCMK(+)-Hsf, we deleted the lacO present in the chimeric porMacO promoter by a recombinant PCR strategy (See Figure 12). The pCMK(+> Hsf plasmid was used as a template to PCR amplify 2 separate DNA fragments: -fragment 1 contains the porA 5' recombinogenic region, the Kanamycin resistance gene and the porA promoter. Oligonucleotide primers used, RPl(5adI) and RP2. aie presented in the table below. RPl primer u homologous to the sequence just upstream of the lac operator. -fragment 2 contains the Shme-Dalgamo sequence from -Cat porA gene, the hsf gene and the porA 3 ' recombinogenic region. Oligonucleotide primers used, RP3 and RP4(.-lpflI), are presented in the table below. RP3 primer is homologous to the sequence just downstream of the he operator. The 3' end of fragment 1 and the 5'end of fragment 2 have 48 bases overlapping. 500ng of each PCR (1 and 2) were used for a final PCR reaction using primers RPI and RP4. The final amplicon obtained was subcloned in pSLUSO vector restricted with SacR and Apal. Tne modified plasmid pCMK(+)-Hsf was purified at a large scale using the QIAGEN maxiprep kit and 2 ^g of this materia! was used to transform a Neisseria meningitidis seragroup B strain lacking functional cps genes (the strain described in example 1). In order to preserve the expression afporA, integration resulting from a single crossing-over was selected by a combination of PCR and Western blot screening procedures. Kanamycin resistant clones testing positive by porA -specific PCR and western blot were stored ar -?D°C as glycerol stocks and used for further studies. Bacteria (corresponding to about 5.108 bacteria) were re-suspended in 50 ui of PAGE-SDS buffer, frozen (-20°C) / boiled (100°C) three times and then were separated by PAGE-SDS electrophoresis on a 12.5 % gel. The expression of Hsf was examined in Whole-cell bacterial lysates (WCBL) derived from NmB [Cps-, PorA+] or NmB [Cps-, PorA+, Hsf+]. Coomassie staining detected- •> a-igr. Ificarf:" " increase- in the expression or Hsf (wite respect to the endogenous Hsf level) (See in Figure 13), This result confirms that the modified pCMK(+)-Hsf vector is functional and can be used successfully to up-regulate tie expression of outer membrane proteins, without abolishing the production of the major PorA outer membrane protein antigen. Example 12: Expression of the Green Fluorescent Protein in a recombinant Neisseria meningitidis serogroup B strain lacking functional cps genes but expressing PorA. In the following example, the pCMK vector was used to test the expression of a cytoplasmic heterologous protein m Neisseria meningitidis. The Green Fluorescent Protein was amplified from the pK.en-Gfpmut2 plasmid with the primers GFP-Asn-mut2 and GFP-Spe (see table in Example 11). Asnl gives cohesive ends compatible with Ndel , Spel gives cohesive ends compatible with Nhel. The conditions used for PCR amplification were those described by the supplier (HiFi DNA polymerase, Boehringer Mannheim, GmbH). Thermal cycling was the following: 25 times (94°C Imin., 48°C imin., 72°C 3min.) and 1 time (72°C lOmin., 4°C up to recovery).The corresponding amplicon was subsequently cloned in the pCMK(+) delivery vector digested with Ndel and Nhel restriction enzymes. In this recombinant plasmid, designed pCMK(+)-GFP, we deleted the lacO present in the chimeric porAAacO promote: by a recombinant PCR strategy. The pCMK(+}-GFP plasmid was used as template to PCR. amplify 2 separate DNA fragments1. -fragment 1 contained the porA 5' recombinogenic region, the Kanamycin resistance gene and tbeporA promoter. Oligonucleotide primers used, RPl(SacH) and RP2 (see table in example 11). RP1 pnmer is homologous to the sequence just upstream of the lac operator. -fragment 2 contains the PorA Shme-Dalgamo sequence, the gfp gene and the porA 3' recombinogenic region. Oligonucleotide primers used, RP3 and RP4(,4pai), are presented in the table in example 11. RP3 primer is homologous to the sequence just downstream of the lac operator. The 3'end of fragment 1 and the 5'end of fragment 2 have 48 bases overlapping. 500ng of each PCR (1 and 2) were used for a final PCR reaction using printers RPi and RP4. Twenty jig of this PCR fragment were used to transform a Neisseiria meningitidis serogroup B strain lacking functional cps genes. Transformation with linear DNA is less efficient than with circular plasmid DNA but all the recombinants obtained performed a double crossing-over (confirmed by a combination of PCR and Western blot screening procedures). Kanamyctn resistant clones were stored at -70°C as glycerol stocks and used for further studies. Bacteria (corresponding to about 5.108 bacteria) were re-suspended in 50 ul of PAGE-SDS buffer, fro2en (-20°C) / boiled (100°C) three times and then were separated by PAGE-SDS electrophoresis on a 12.5 % gel. The expression of GFP was examined in Whole-cell bacterial lysates (WCBL) derived from NmB [Cps-, PorA+J or NmB [Cps-, PorA-. GFP+], Coomassie staining detected an expression of GFP absent in the recipient Neisseria meningitidis strain (see figure 14). Example 13: Up-regulation of the N. meningitidis serogroup B NspA gene by promoter replacement The aim of the experiment was to replace the endogenous promoter region of the NspA gene by the strong porA promoter, in order to up-regulate the production of the NspA antigen. For that purpose, a promoter replacement plasmid was constructed using E. coli cloning methodologies. A DNA region (924bp) located upstream from tie NspA coding gene was discovered (SEQ ID NO: 7) in the private Incyte PathoSeq data base containing unfinished genomic DNA sequences of the Neisseria meningitidis strain ATCC 13090. A DNA fragment (675bp) covering nucleotides -115 to -790 with respect to the NspA gene start codon (ATG) was PCR amplified using oligonucleotides PNS1' (5'-CCG CGA ATT CGA CGA AGC CGC CCT CGA C-3') and PNS2 fS'-CGT CTA GAC GTA GCG GTA TCC GGC TGC -V) containing EcoRl and Xbal restriction sites (underlined) respectively. The PCR fragment was submitted to restriction with EcoRl and Xbal and inserted in pUC18. This plasmid was submitted to circle PCR mutagenesis ('Jones & Winistofer (1992), Biotechniquesl2: 528-534) in order to insert meningococcal uptake sequences required for transformation, and suitable restriction sites allowing cloning of a CmRfPaiA promoter cassette. The circle PCR was performed using the BAD01-2 (5'- GGC GCC CGG GCT CGA GCT TAT CGA TGG AAA ACG CAG C-3') & BAD02-2 (5'-GGC GCC CGG GCT CGA GTT CAG ACG GCG CGC TTA TAT AGT GGA TTA AC -3') oligonucleotides containing uptake sequences and suitable restriction sites (Xmal and Xhol) underlined. The resulting PCR fragment was gel-purified and digested with Xhol. The CmR/PorA promoter cassette was amplified from the pUC D15/Omp85 plasmid previously described, using primers BAD 15-2 {5'-GGC GCC CGG GCT CGA GTC TAG ACA TCG GGC AAA CAC CCG-3') 8c BAD 03-2 (5'- GGC GCC CGG GCT CGA GCA CTA GTA TTA CCC TGT TAT CCC-3') oligonucleotides containing suitable restriction sites {Xmal, Xbal, Spel and Xhol} underlined. The PCR fragment obtained was submitted to digestion and inserted in the circle PCR plasmid restricted with the corresponding enzymes. 10 ug of the recombinant plasmid were linearized and used to transform a strain of Neisseria meningitidis serogroup B lacking capsular polysaccharide (cps-) and one of the major outer membrane proteins - PorA (porA-). Recombinant Neisseria meningitidis clones resulting from a double crossing over event EPCR screening using oligonucleotides BAD 25 {5'-GAG CGA AGC CGT CGA ACG C -3') & BAD08 The same recombinant pUC plasmid can be used to up-regulaie the expression of NspA in a Neisseria meningitidis serogroup B strain lacking functional cps gens but still expressing PorA. Example 14: Up-regulation of the N. meningitidis serogroup B pldA The aim of the experiment was to replace the endogenous promoter region of the pldA (pmplA) gene by the strong porA promoter in order to up-regulate the production of the PldA (OmpiA I) antigen. For that purpose, a promoter replacement plasmid was constructed using E. coh cloning methodologies. A DNA region (373bp) located upstream from the pldA coding sequence was discovered (SEQ ID NO: 18) in the private Incyte PathoSeq data base of the Neisseria meningitidis strain ATCC 13090. This DNA contains the sequence coding for a putative rpsT gene. The stop codon of rpsT is located 169bp upstream the pldA ATG. To avoid the disruption of this potentially important gene, we decided to insert the CmRfPoiA promoter cassette nucleotides of pldA gene was PCR amplified from Neisseria meningitidis serogroup B genomic DNA using oligonucleotides PLAl Amo5 (5'-OCC GTC TGA ATT TAA AAT TGC GCG TTT ACA G-3') and PLAl Amo3 (5'-GTA GTC TAG ATT CAG ACG GCG CAA TTT GGT TTC CGC AC -3') containing uptake sequences (underlined). PLAl Amo3 contains also a Xbal restriction site. This PCR fragment was cleaned with a High Pure Kit (Roche, Mannheim, Germany) and directly cloned in a pGemT vector (Promega, USA). This plasmid was submitted to circle PCR mutagenesis (Jones & Winistofer (1992)) in order to insert suitable restriction sites allowine cloning of a CmRfPoiA promoter cassette. The circle PCR was performed using the CIRCl-Bgl (5'CCT AGA TCT CTC CGC CCC CCA TTG TCG -3') & either CIRC1-XH-RBS/2 (5'-CCG CTC GAG TAC AAA AGG AAG CCG ATA TGA ATA TAC GGA ATA TGC G-3*) or CIRC2-XHO/2 (5'-CCG CTC GAG ATG AAT ATA CGG .AAT -3') oligonucleotides containing suitable restriction sites [BglR and Xhol) underlined. The Cm&Tor A promoter cassette was amplified from the pUC D15/Omp85 plasmid previously described, using primers BAD20 (5'- TCC CCC GGG AGA TCT CAC TAG TAT TAC CCT GTT ATC CC-3") and CM-PORA-3 (5'- CCG CTC GAG ATA .AAA ACC TAA .AAA CAT CGG GC-3') containing suitable restriction sites [BgUl and Xhol) underlined. This PCR fragment was cloned in the circle PCR plasmid obtained with primers CIRCl-Bgl and CIRC1-XH-RBS/2. This plasmid can be used to transform Neisseria meningitidis serogroup B Icps-] and Icps- porA-) strains. Integration by double crossing-over in the upstream region of pldA will direct the insertion of the porA promoter directly upstream of the pidA ATG. Another cassette was amplified from the genomic DNA of the recombinant Neisseria meningitidis serogroup B lcps-,porA% D15/Omp85+) over-expressing DI5/Omp85 by promoter replacement. This cassette contains the cmR gene, the porA promoter and 400bp corresponding to the 5' flanking sequence of the D15/Omp85 gene. This sequence has been proven to be efficacious for up-regulation of the expression of D15/Omp85 in Neisseria and will be tesred for the up-reguiaiion of the expression of other Neisseria antigens. Primers used for the amplification were BAD 20 and CM-PORA-D1S/3 (5'- CGG CTC GAG TGT CAG TTC CTT GTG GTG C-3'} containing Xhol restriction sites (underlined). This PCR fragment was cloned in the circle PCR plasrrnd obtained with primers CIRCI-Bg! and CTRC2-XHO/1 Tliis plasmid will be used to transform Neisseria meningitidis serogroup B lcps-1 and icps-, porA-) strains. Integration by double crossing-over in the upstream region of pidA will direct the insertion of fat porA promoter400bp upstream thepldA A7G- Example 15: Up-regulation of the N. meningitidis serogroup B tbpA gene by promoter replacement. The aim of the experiment was to replace the endogenous promoter region of the tbpA gene by the strong porA promoter, in order to up-regulate the production of the TbpA antigen. For thai purpose, a promoter replacement plasmid was constructed using E. coli cloning methodologies. A DNA region (731bp) located upstream from the tbpA coding sequence was discovered (SEQ ID NO: 17) in the private Incyte PathoSeq data base of the Neisseria meningitidis strain ATCC 13090. This DNA contains the sequence coding for TbpB antigen. The genes are organized in an operon. The tbpB gene will be deleted and replaced by the CmR/porA promoter cassette. For that puroose. a DNA fragment of 321 Sbp corresponding to the 509bp 5' flanking region of tbpB gene, the 2l39bp tbpB coding sequence, the 87bp intergenic sequence and the 483 first nucleotides of tbpA coding sequence was PCR amplified from Neisseria meningitidis serogroup B genomic DNA using oligonucleotides BADI6 (5'- GGC CTA GCT AGC CGT CTG AAG CGA TTA GAG TTT CAA AAT TTA TTC-3') and BAD17 (5'-GGC CAA GCT TCA GAC GGC GTT CGA CCG AGT TTG AGC CTT TGC-3') containing uptake sequences and Nhel and HmdlU restriction sites (underlined). This PCR fragment was cleaned with a High Pure Kit ( Boerhinger Mannheim, Germany) and directly cloned in a pGemT vector (Promega, USA). This plasmid was submitted to circle PCR mutagenesis (Jones & Winistofer (1992)) in order to (i) insen suitable restriction sites allowing cloning of a CmK/PorA promoter cassette and (ii) to delete 209bp of the 5' flanking sequence of tbpB and the tbpB coding sequence. The circle PCR was performed using the BAD 18 (5'-TCC CCC GGG AAG ATC TGG ACG AAA AAT CTC -AAG AAA CCG-3*) & the BAD 19 (5'-GGA AGA TCT CCG CTC GAG CAA ATT TAC AAA AGG AAG CCG ATA TGC AAC AGC AAC ATT TGT TCC G -3') oligonucleotides containing suitable restriction sites Xmal, Bgrtl and Xhol (underlined). The Cmi!/PorA promoter cassette was amplified frorata.epUG"-D.15'Omp85 pUsmid previously described, using primers BAD21 (5'- GGA AGA TCT CCG CTC GAG ACA TCG GGC AAA CAC CCG-3') & BAD20 (5'- TCC CCC GGG AGA TCT CAC TAG TAT TAC CCT GTT ATC CC-3') containing suitable restriction sites Xmal, Spel, Bgfil and Xhol (underlined). This PCR fragment was cloned in the circle PCR plasmid. This plasmid will be used to transform Neisseria meningitidis serogroup B leps-i and leps- porA-) strains. Integration by double crossing-over in the upstream region of tbpA will direct the insertion of iheporA promoter directly upstream of the tbpA ATG. Example 16: Up-regulation of the A', meningitidis serogroup B pilQ gene by promoter replacement. The aim of the experiment was to replace the endogenous promoter region of the pilQ gene by the strong porA promoter, in order to up-regulate the production of the PilQ antigen. For that purpose, a promoter replacement plasmid was constructed using E. coli cloning methodologies. A DNA region (772bp) located upstream from the pi!Q coding gene was discovered (S£Q ID NO: 12) in the private Incyte PathoSeq data base of the Neisseria meningitidis stram ATCC 13090. This DNA contains the sequence coding for PilP antigen. The pilQ gene, is part of an operon we do not want to disturb, pilins being essentia! elements of the bacteria. The Cmfi/porA promoter cassette was introduced upstream tixpilQ gene following the same strategy described for the up-regulation of the expression of the pldA gene. For that purpose, a DNA fragment of 866 bp corresponding to the 3' pan of the pilP coding sequence, the 18bp intergenic sequence and the 392 first nucleotides of pilQ gene was PCR amplified from Neisseria serogroup B genomic DNA using PQ-recS-Nhe (S*-CTA GCT AGC GCC GTC TGA ACG ACG CGA AGC CAA AGC-3'} and PQ-rec3-Hin (GCC AAGCTT TTC AGA CGG CAC GGT ATC GTC CGA TTC G-3') oligonucleotides containing uptake sequences and Nhel and Hindlll restriction sites (underlined). This PCR fragment was directly cloned in a pGemT vector (Promega, USA). This plasmid was submitted to circle PCR mutagenesis (Jones &. Winistofer (1992)) in order to insert suitable restriction sites allowing cloning of a Cm#/PorA promoter cassette. The circle PCR was performed using the CIRCI-PQ-Bgl (5'-GGA AGA TCT AAT GGA GTA ATC CTC TTC TTA-3') & either CIRC1-PQ-XHO (5'-CCG CTC GAG TAC A.1A AGO AAQ CCC ATA TGA TTA CCA AAC TGA CAA AAA TC-3') or CIRC2-PQ-X (5'-CCG CTC GAG ATG AAT ACC AAA CTG ACA AAA ATC -3') oligonucleotides containing suitable restriction sites figfll and Xhol (underlined). The CmR/PoiA promoter cassette was amplified from the pUC D15/Omp85 plasmid previously described, using primers BAD20 (5'- TCC CCC GGG AGA TCT CAC TAG TAT TAC CCT GTT ATC CC-3') and CM-PORA-3 (5'- CCG CTC GAG ATA AAA ACC TAA AAA CAT CGG GCA AAC ACC C-3') containing suitable restriction sites Bgtll and Xhol (underlined), This PCR fragment was cloned in the circle PCR plasmid obtained with primers CIRCI-PQ-Bgl and CIRC 1-PQ-XHO. This plasmid can be used to transform Neisseria meningitidis serogroup B \cps-) and Ecps-, porA-1 strains. Integration by double crossing-over in the upstream region ofpiiQ will direct the insertion of the porA promote: directiy upstream of thepilQ ATG. Another cassette was amplified from the genomic DKA of the recombinant Neisseria meningitidis serogroup B fops-, porA-, D!5/Omp85+\ over-expressing D 15'0mpS5 by promote: replacement. This cassette contains toe cmR gene, the porA promoter and 400bp corresponding to the 5' flanking sequence of the D15/OmpS5 gene. This sequence has been proven to be efficacious for up-regulation of the expression of Dli/OmpSS in Neisseria meningitidis and will be tested for the up-regulation of the expression of other Neisseria antigens. Pnmers used for the amplification were BAD 20 and CM-PORA-D153 (5'- GGG CTC GAG TGT CAG TTC CTT GTG GTG C-31) containing Xhdl restriction sites (underlined}. This PCR fragment was cloned in the circle PCR plasmid obtained with primers CIRCI-PQ-Bgl and CIRC2-PQ-X. This plasmid can be used to transform Neisseria meningitidis serogioup B \cps-\ and icps-, porA-) strains, integration by double crossing-over in the upstream region of/?i/£>will direct the insertion of the porA promoter 400bp upstream thepilQATG. Example 17: Construction of a kanR/sacB cassette for introducing "clean", unmarked mutations in the N. meningitidis chromosome. The aim of the experiment is to construct a versatile DNA cassette containing a selectable marker for the positive screening of recombination in the chromosome of Neisseria meningitidis fie: kanR gene), and a counter selectable marker to delete the cassette from_the chromosome after; recombination fig: sacB genej. ■ By .this method, any heterologous DNA introduced during homologous recombination will be removed from the Neisseria chromosome. A DNA fragment containing the neoR gene and the sacB gene expressed under the control of its own promoter was obtained by restriction of the pIB 279 plasmid (Blomfield IC, Vaughn V, Rest RF, Eisenstein Bl (1991), Mol Microbiol 5:1447-57) with BamHl restriction enzyme. The recipient vector was derived from plasmid pCMK, previously described. The kanR gene of the pCMK was deleted by restriction with enzymes Nntl and EcoKV. This plasmid was named pCMKs. The neoR/sacB cassette was inserted in the pCMKs at a Bg!U restriction site compatible with BamHl ends. £. coii harboring the plasmid is unable to grow in the presence of 2% sucrose in the culture medium, confirming the functionality of the sacB promoter. This plasmid contains recombioogenic sequences allowing the insertion of the cassette at the por.1 locus in the chromosome of Neisseria meningitidis serogroup B. Recombinant Neisseria were obtained on GC agar plates containing 200ug/ml of kanamycin. Unfortunately, the sacB promoter was not functional in Neisseria meningitidis: nc growth difference was observed on GC agar plates containing 2% sucrose. A new cassette was constructed containing the sacB gene under the control of the kanR promoter. A circle PCR was performed using the plasmid pUC4K ((Amersham Phamiacia Biotech, USA)) as a template with CIRC-Kao-Nco (5'-CAT GCC ATG GTT AGA AAA ACT CAT CGA GCA TC-3*) &. CIRC-Kan-Xba (5'-CTA GTC TAG ATC AGA ATT GGT TAA TTG GTT G-3') oligonucleotides containing A^d and Xbal restriction sites (underlined). The resulting PCR fragment was gel-purified, digested with Ncol and ligaied to the sacB gene generated by PCR from the pIB279 plasmid with SAC/NCO/NEW5 (5'-CAT GCCATGGGA GGA TGA ACG ATG AAC ATC AAA AAG TTT GCA A-3') oligonucleotide containing a Ncol restriction site (underlined) and a RBS (bold) & SACWCOttJEW3 (S'-GAT CCC ATG GTT ATT TGT TAA CTG TTA ATT GTC-3') oligonucleotide containing a Ncol restriction site (underlined). The recombinant E. coli clones can be tested for their sensitivity on agar plates containing 2% sucrose. Th? nrv kr.r.R !sazB -c^ssett? can be subcloned in '^. pCMKj r,.-,c vss£ to transform a Neisseria meningitidis serogroup B cps- strain. The acquired sucrose sensitivity will be confirmed in Neisseria. The pCMKs plasmid will be used to transform the recombinant kan&'SacB Neisseria to delete the entire cassette inserted in the chromosome at the porA locus. Clean recombinant Neisseria will be obtained on GC agar plates containing 2% sucrose. Example 18: Use of small recombinogenic sequences (43bo) to allow homologous recombination in the chromosome of Neisseria meningitidis. The aim of the experiment is to use small recombinogenic sequences (43bp) to drive insertions, modifications or deletions in the chromosome of Neisseria. The achievement of this experiment will greatly facilitate future work, in terms of avoiding subclomng steps of homologous sequences in £. coli (recombinogenic sequences of 43bp can easily be added in the PCR amplification pnmer). The kanR gene was PCR amplified from plasmid pUC4K with oligonucleotides Kan-PorA-5 (5'-GCC GTC TGA ACC CGT CAT TCC CGC GCA GGC GGG AAT CCA GTC CGT TCA GTT TCG GGA AAG CCA CGT TGT GTC-3') containing 43bp homologous to the 5' flanking sequence of A'mB pcv4 gene (bold) and an uptake sequence (underlined) & Kan-PorA-3 (5'-TTC AGA CGG CGC AGC AGG AAT TTA TCG GAA ATA ACT GAA ACC GAA CAG ACT AGG CTG AGG TCT GCC TCG-3') containing 43bp homologous to the 3' flanking sequence of MnB porA gene (bold) and an uptake sequence (underlined). The 1300bp DNA fragment obtained was cloned in pGemT vector (Promega, USA). This plasmid can be used to transform a Neisseria meningitidis serogioupE cps- strain. Recombinant Neisseria will be obtained on GC plates containing 200ug/ml kanamycin. Integrations resulting from a double crossing-over at the porA locus will be screened by PCR with primers PPA1 & PPA2 as described previously. Example 19: Active protection of mice immunized with WT and recombinant Neisseria meningitidis blebs Animals were immunised three times (IP route) with 5 fig of the different OMVs adsorbed on Al(OH)3 on days 0, 14 and 28. Bleedings were done on days 28 (day 14 Post ID and 35 (day 7 post Til), and they were cbailerged on day .?.5 (P route). The challenge dose was 20 x LD50 (-10' CFU/mouse). Mortality rate was monitored for 7 days after challenge. OMVs injected were: Group 1: Cps-, PorA+ blebs Group2: Cps-, PorA- blebs Group3: Cps-, PorA-, NspA-*- blebs Group4: Cps-, PorA-, Omp85+ blebs Group5: Cps-, PorA-, Hsf+ blebs Figure IS illustrates the pattern of these OMVs by analyzed SDS Page (Coomassie staining) 24 hours after the challenge, there was 100% mortality (8/8) in the negative control group (immunised with Al(OH), alone) while mice immunised with the 5 different OMVs preparations were still alive {1 to 8/8 mice survived). Sickness was also monitored during the 7 days and the mice immunised with the NSPA over-expressed blebs appeared to be less sick than the other groups. PorA present in PorA-blebs is likely to confer extensive protection against infection by the homologous strain. However, protection induced by PorA- up-regulated blebs is likely to be due at least to some extent, to the presence of increased amount of NspA, Omp85 or Hsf. The specific Ab response to available recombinant HSF protein was carried out. Microplates were coated with 1 ug/ml full length HSF molecule. The results illustrated in Figure 16 show that there was a good specific HSF response when HSF over-expressed OIvTVs were used to immunize mice (using purified recombinant HSF on the plates). The HSF over-expressed blebs induce a good level of specific antibodies. WE CLAIM: 1. A genetically-engineered bleb preparation isolated from a modified Neisseria meningitidis strain characterized in that said preparation is obtainable by employing the following process: a process of reducing immunodominant variable or non-protective antigens within the bleb preparation comprising the steps of engineering a bacterial strain to produce less or none of the PorA antigen, and making blebs from said strain. 2. A genetically-engineered bleb preparation isolated from a modified Moraxella catarrhalis strain characterized in that said preparation is obtainable by employing the following process: a process of reducing immunodominant variable or non-protective antigens within the bleb preparation comprising the steps of determining the identity of such antigen, engineering a bacterial strain to produce less or none of said antigen, and making blebs from said strain, wherein one or more of said antigens are selected from a list consisting of: CopB. OMP106, OmpBl, TbpA, TbpB, LbpA, and LbpB. 3. A genetically-engineered bleb preparation isolated from a modified Haemophilus influenzae strain characterized in that said preparation is obtainable by employing the following process: a process of reducing immunodominant variable or non-protective antigens within the bleb preparation comprising the steps of determining the identity of such antigen, engineering a bacterial strain to produce less or none of said antigen, and making blebs from said strain, wherein one or more of said antigens are selected from a list consisting of: P2, P5, Hif, IgAl-protease, HgpA, HgpB, HMW1, HMW2, Hxu, TbpA, and TbpB. 4. The genetically-engineered bleb preparation as claimed in Claims 1-3, wherein said preparation is obtainable by employing one or more further processes selected from the following group: b) a process of upregulating expression of protective OMP antigens within the bleb preparation comprising the steps of identifying such antigen, engineering a bacterial strain so as to introduce a stronger promoter sequence upstream of a gene encoding said antigen such that said gene is expressed at a level higher than in the non-modified bleb, and making blebs from said strain; c) a process of upregulating expression of conditionally-expressed, protective OMP antigens within the bleb preparation comprising the steps of identifying such an antigen, engineering a bacterial strain so as to remove the repressive control mechanisms of its expression, and making blebs from said strain; d) process of modifying lipid A portion of bacterial LPS within the bleb preparation, comprising the steps of identifying a gene involved in rendering the lipid A portion of LPS toxic, engineering a bacterial strain so as to reduce or switch off expression of said gene, and making blebs from said strain; e) a process of modifying lipid A portion of bacterial LPS within the bleb preparation, comprising the steps of identifying a gene involved in rendering the lipid A portion of LPS less toxic, engineering a bacterial strain so as to introduce a stronger promoter sequence upstream of said gene such that said gene is expressed at a level higher than in the non-modified bleb, and making blebs from said strain; f) a process of reducing lipid A toxicity within the bleb preparation and increasing the levels of protective antigens, comprising the steps of engineering the chromosome of a bacterial strain to incorporate a gene encoding a Polymyxin A peptide, or a derivative or analogue thereof, fused to a protective antigen, and making blebs from said strain; g) a process of creating conserved OMP antigens on the bleb preparation comprising the steps of identifying such antigen, engineering a bacterial strain so as to delete variable regions of a gene encoding said antigen, and making blebs from said strain; h) a process of reducing expression within the bleb preparation of an antigen which shares a structural similarity with a human structure and may be capable of inducing an auto-immune response in humans, comprising the steps of identifying a gene involved in the biosynthesis of the antigen, engineering a bacterial strain so as to reduce or switch off expression of said gene, and making blebs from said strain; or i) a process of upregulating expression of protective OMP antigens within the bleb preparation comprising the steps of identifying such antigen, engineering a bacterial strain so as to introduce into the chromosome one or more further copies of a gene encoding said antigen controlled by a heterologous, stronger promoter sequence, and making blebs from said strain. 5. The bleb preparation as claimed in claims 1 to 4, wherein the engineering steps of processes a), b), c), d), e), h), and i) are carried out by homologous recombination events between a sequence of at least 30 nucleotides on the bacterial chromosome, and a sequence of at least 30 nucleotides on a vector transformed within the strain. 6. The bleb preparation as claimed in claim 5, wherein the engineering steps are carried out by double cross-over homologous recombination events between two sequences of at least 30 nucleotides on the bacterial chromosome separated by nucleotide sequence "X\ and two sequences of at least 30 nucleotides on a vector transformed within the strain separated by nucleotide sequence 'Y', wherein during the recombination event X and Y are interchanged. 7. The bleb preparation as claimed in claim 6, wherein the two nucleotide sequences are of approximately the same length, and wherein the vector is a linear DNA molecule. 8. The bleb preparation as claimed in claim 6 or 7, wherein the recombination events of processes a), b), c), d), e) and h) are carried out within the region of the chromosome 1000 bp upstream of the initiation codon of the gene of interest. 9. The bleb preparation as claimed in claim 8, wherein for process a), d) or h) nucleotide sequence X comprises part of the promoter region of the gene, and 79 nucleotide sequence Y comprises either a weak promoter region, or no promoter region. 10. The bleb preparation as claimed in claim 6 or 7. wherein the recombination events of processes a), d) and h) are carried out such that nucleotide sequence X comprises part of the coding sequence of the gene of interest. 11. The bleb preparation as claimed in claim 6 or 7, wherein the recombination events of process i) is carried out such that nucleotide sequence Y comprises the further copy of the gene within an expression cassette. 12. The Neisseria meningitidis bleb preparation as claimed in claim 4, obtainable by employing process b) and/or i), wherein one or more genes are upreguiated from a list consisting of: NspA, Hsf-like, Hap, PorB, OMP85, PilQ, PldA, FrpB, TbpA, TbpB, FrpA, FrpC, LbpA, LbpB, FhaB, HasR, lipo02, Tbp2 (lipo28), MltA (lipo30), and ctrA. 13. The Neisseria meningitidis bleb preparation as claimed in claim 4 or 12, obtainable by employing at least process h), wherein one or more genes are downregulated from a list consisting of: galE, siaA, siaB, siaC, siaD, ctrA, etrB, ctrC, and ctrD. 14. The Moraxeila catarrhalis bleb preparation as claimed in claim 4, obtainable by employing process b) and/or i), wherein one or more genes are upreguiated from a list consisting of: OMP106, HasR, PilQ, OMP85, lipo06, lipolO, lipoll, lipolS, P6, ompCD, CopB, D15, OmplAl, Hly3, LbpA, LbpB, TbpA, TbpB, OmpE, UspAl, UspA2, andOmp21. 15. The Haemophilus influenzae bleb preparation as claimed in claim 4, obtainable by employing process b) and/or i), wherein one or more genes are upreguiated from a 80 list consisting of: D15, P6, TbpA, TbpB, P2, P5, OMP26, HMW1, HMW2, HMW3, HMW4, Hia, Hsf, Hap, Hin47, and Hif. 16. A vaccine comprising the bleb preparation as claimed in claims 1 to 15 and a pharmaceuticaUy acceptable excipient. 17. A meningococcal vaccine comprising the bleb preparation as claimed in claims 1, 4, 12, or 13 and one or more plain or conjugated meningococcal capsular polysaccharides selected from the serotypes A, C, Y or W. 18. A meningitis vaccine comprising the preparation as claimed in claims 1, 4, 12 or 13, a conjugated H. influenzae b capsular polysaccharide, and one or more plain or conjugated pneumococcal capsular polysaccharides. 19. An otitis media vaccine comprising the bleb preparation as claimed in claims 2, 4 or 14 and one or more plain or conjugated pneumococcal capsular polysaccharides, and one or more antigens that can protect a host against non-typeable H. influenzae infection. 20. An otitis media vaccine comprising the bleb preparation as claimed in claims 3, 4 or 15 and one or more plain or conjugated pneumococcal capsular polysaccharides, and one or more antigens that can protect a host against Moraxella catarrhalis infection. 21. The otitis media vaccine as claimed in claim 19 or 20 wherein the vaccine additionally comprises one or more protein antigens that can protect a host against Streptococcus pneumoniae and/or one or more antigens that can protect a host against RSV and/or one or more antigens that can protect a host against influenza virus. 22. A modified Gram-negative bacterial strain from which the bleb preparation as claimed in claims 1-15 is made. 81 23. A method of making a vaccine comprising the step of generating said vaccine from the modified Gram-negative bacterial strain as claimed in claim 22.

Documents:

in-pct-2002-0152-che abstract-duplicate.pdf

in-pct-2002-0152-che abstract.pdf

in-pct-2002-0152-che claims-duplicate.pdf

in-pct-2002-0152-che claims.pdf

in-pct-2002-0152-che correspondence-others.pdf

in-pct-2002-0152-che correspondence-po.pdf

in-pct-2002-0152-che description (complete)-duplicate.pdf

in-pct-2002-0152-che description (complete).pdf

in-pct-2002-0152-che drawings-duplicate.pdf

in-pct-2002-0152-che drawings.pdf

in-pct-2002-0152-che form-1.pdf

in-pct-2002-0152-che form-18.pdf

in-pct-2002-0152-che form-26.pdf

in-pct-2002-0152-che form-3.pdf

in-pct-2002-0152-che form-5.pdf

in-pct-2002-0152-che pct search report.pdf

in-pct-2002-0152-che pct.pdf

in-pct-2002-0152-che petition.pdf