The present invention provides means and products for the stimulation of an immune response against tumors in a subject in need thereof. More specifically the invention provides immunogenic compositions containing bacterial outer membrane vesicles loaded with tumor antigens, fusion proteins comprising a bacterial protein and a tumor antigen, and isolated bacterial outer membrane vesicles (OMVs) containing said fusion proteins. The fusion proteins, OMVs and immunogenic compositions according to the invention are used in the prevention and treatment of tumors.
Bacterial Outer Membrane Vesicles (OMVs)
More than 40 years ago, researchers made the observation that Gram-negative bacteria secrete Outer Membrane Vesicles (OMVs) (Beveridge T J (1999) J Bacteriol. 181:4725-33; Mayrand D & Grenier D. (1989) Can J Microbiol. 35:607-13). However, the last 15 to 20 years have brought a greater understanding of the regulation and function of vesiculation. OMVs are closed spheroid particles of a heterogeneous size, 20-300 nm in diameter, generated through a “budding out” of the bacterial outer membrane. Consistent with that, the majority of OMV components are represented by LPS, glycerophospholipids, outer membrane proteins and periplasmic proteins (A. Kulp and Kuehn M. J. (2010) Annu. Rev. Microbiol. 64, 163-184; T. N. Ellis and Kuehen M. J. (2010) Microbiol. Mol. Biol. Rev. 74, 81-94).
OMVs represent a distinct secretory pathway with a multitude of functions, including inter and intra species cell-to-cell cross-talk, bio film formation, genetic transformation, defense against host immune responses and toxin and virulence factor delivery to host cells (A. Kulp and Kuehn M. J. (2010) Annu. Rev. Microbiol. 64, 163-184). OMVs interaction to host cells can occur by endocytosis after binding to host cell receptors or lipid rafts. Alternatively, OMVs have been reported to fuse to host cell membrane, leading to the direct release of their content into the cytoplasm of the host cells (A. Kulp and Kuehn M. J. (2010) Annu. Rev. Microbiol. 64, 163-184; T. N. Ellis and Kuehen M. J. (2010) Micrbiol. Mol. Biol. Rev. 74, 81-94).
OMVs purified from several pathogens, including Neisseria, Salmonella, Pseudomonas, Vibrio cholerae Burkholderia, and E. coli, induce potent protective immune responses against the pathogens they derive from (B. S. Collins (2011) Discovery Medicine, 12, 7-15), and highly efficacious anti-Neisseria OMV-based vaccines are already available for human use (J. Holst et al. (2009) Vaccine, 27S, B3-B12). Such remarkable protection is attributed to three key features of OMVs.
First, they are readily phagocytosed by professional antigen-presenting cells which get activated and present OMV-derived peptides in the context of MHC molecules.
Second, they carry the proper immunogenic antigens which, in extracellular pathogens, usually reside on the surface and therefore are naturally incorporated in OMVs. Indeed, OMV immunization induces potent antibody responses against the major membrane-associated antigens. However, OMV immunogenicity is not restricted to antibody responses. For instance, mice immunized with Salmonella OMVs develop robust Salmonella-specific B and T cell responses, and OMVs stimulate IFN-γ production by a large proportion of CD4+ T cells from mice previously infected with Salmonella, indicating that OMVs are an abundant source of antigens recognized by Salmonella-specific CD4+ T cells (R. C. Alaniz et al., (2007) J. Immunol. 179, 7692-7701).
Third, and most importantly, OMVs carry most of the bacterial Pathogen-Associated-Molecular Patterns (PAMPs) which, by binding to pathogen recognition receptors (PRRs), play a key role in stimulating innate immunity and promoting adaptive immune responses. OMV-associated PAMPs include LPS which, in concert with MD-2 and CD14, binds TLR-4, lipoproteins whose acylpeptide derivatives interact with TLR-1/2 and 2/6 heterodimers, and peptidoglycan whose degradation products bind to intracellular NOD1/2 (A. Moshiri et al., Hum. Vaccines. Immunother. (2012) 8, 953-955; T. N. Ellis et al., (2010) Inn. Immun. 78, 3822-3831; M. Kaparakis et al., (2010) Cell. Miocrobiol. 12, 372-385). The engagement of this group of PPRs results in the activation of transcription factors (NF-kB) and the consequent expression of specific cytokines. Interestingly, LPS, lipoproteins and peptidoglycan can work synergistically, thus potentiating the built-in adjuvanticity of OMVs (D. J. Chen et al., (2010) PNAS, 107, 3099-3104).
An additional interesting property of OMVs is their capacity to induce protection at the mucosal level. Protection at the mucosal sites is known to be at least partially mediated by the presence of pathogen-specific IgAs and Th17 cells. In particular, a growing body of evidence suggests that Th17 cells have evolved to mediate protective immunity against a variety of pathogens at different mucosal sites. Interestingly, Th17 cells have recently also been shown to play a crucial role in the generation of vaccine-induced protective responses. For instance, it has been reported that in mice whole cell pertussis vaccines (Pw) induce Th17 cells and neutralization of IL-17 after vaccination reduces protection against a pulmonary challenge with B. pertussis. Similarly, in a CD4+ T cell dependent, antibody-independent model of vaccine-induced protection following S. pneumoniae challenge, treatment with IL-17-antibodies resulted in reduced immunity to pneumococcal colonization compared to the control serum treated mice (Malley R, et al. (2006) Infect Immun., 74:2187-95).
Elicitation of IgAs and Th17 cells by OMVs has been well documented and this can explain mechanistically the good protective activities of OMVs against several mucosal pathogens. For instance, immunization with Vibrio cholerae-derived OMVs protects rabbits against Vibrio cholerae oral challenge (Roy N. et al. (2010) Immunol. Clinical Microbiol. 60, 18-27) and Pasteurella multocida-derived and Mannheimia haemolytica-derived OMVs protect mice from oral challenge with P. multocida (Roier S. et al., (2013) Int. J. Med. Microbiol. 303, 247-256). In addition, intranasal immunization with Porphyromonas gingivalis OMVs elicits potent IgA production at both serum and mucosal level and immunization with Escherichia coli-derived OMVs prevent bacteria-induced lethality. Protective effect of Escherichia coli-derived OMVs is primarily mediated by OMV-specific, IFN-γ and IL-17 producing, T cells (Kim O Y et al., (2013) J. Immunol. 190, 4092-4102).
Finally, a key feature of OMVs is the possibility to manipulate their protein content by genetic engineering. This feature was demonstrated for the first time by Kesty and Kuehn who showed that Yersinia enterocolitica outer membrane protein Ail assembled on OMVs surface when expressed in E. coli, and that the GFP fluorescence protein fused to the “twin arginine transport (Tat)” signal sequence was incorporated in the OMV lumen (N. C. Kesty and Kuhen M. J. (2004) J. Biol. Chem. 279, 2069-2076). Following the observation by Kesty and Kuehn, an increasing number of heterologous proteins have been successfully delivered to OMVs using a variety of strategies. For instance, heterologous antigens have been delivered to the surface of OMVs by fusing them to the β-barrel forming autotransporter AIDA and to hemolysin ClyA, two proteins that naturally compartmentalized into E. coli OMVs (J. Schroeder and Aebischer T. (2009) Vaccine, 27, 6748-6754; D. J. Chen et al., (2010) PNAS, 107, 3099-3104). Recently, heterologous antigens from Group A Streptococcus and Group B Streptococcus were delivered to the lumen of E. coli vesicles by fusing their coding sequences to the leader peptide of E. coli OmpA. Interestingly, when the recombinant vesicles were used to immunize mice, they elicited high titers of functional antibodies against the heterologous antigens, despite their luminal location (Fantappiè et al., (2014) Journal of Extracellular Vesicles, 3, 24015).
Despite the many strategies successfully used to deliver heterologous antigens to the vesicle compartment, it has to be pointed out that a universal system working for any protein antigen has not been described yet. A strategy that is effective for one specific antigen in terms of level of expression and elicitation of immune responses can be inefficient with other antigens.
Therefore, the identification of novel strategies to deliver antigens to the OMV compartment is highly needed.
In general, the amount of OMV released by Gram-negative bacteria when grown under laboratory conditions is too low to allow the exploitation of OMVs in biotechnological applications. However, it has been shown that under stress conditions, such as high temperature, OMV release is substantially increased. Furthermore, a number of mutations have been described, many of them affecting the composition of the periplasmic and membrane compartments, that result in abundant vesicle production. For instance, in Neisseria meningitidis, a mutation in the gna33 gene, encoding a glucosyltransferase, has been shown to drive the release of several milligrams of vesicles per liter in the culture supernatant (Ferrari et al., (2006) Proteomics, 6, 1856-1866). Similar quantities of vesicles are obtained from Escherichia coli strains carrying deletions in the genes encoding the Tol/Pal system (a protein complex involved in the connection of the inner membrane with the outer membrane) (Bernadac A. et al., (1998) J. Bacteriol. 180, 4872-4878) and in the ompA gene, encoding one of the major outer membrane proteins of E. coli (Fantappiè et al., (2014) Journal of Extracellular Vesicles, 3, 24015). Such quantities make the production process of OMVs highly efficient and inexpensive thus making natural and engineered vesicles extremely attractive for vaccine purposes. A number of other mutations have been described that enhance the production of OMVs in several Gram negative bacteria, including Salmonella and E. coli (Deatherage B. L. et al. (2009) Mol. Microbiol. 72, 1395-1407; McBroom A. J. and Kuehen M. J. (2007) Mol. Microbiol. 63, 545-558), and such mutations are amenable to be exploited to develop scalable OMV production processes.
As far as the production of OMVs for industrial applications is concerned, a number of methods have been described. The first ones to be developed make use of mild detergents that promote the production of great yield of vesicles from biomass and decrease the level of toxicity by removing a substantial amount of LPS (Fredriksen J. H. et al, (1991) NIPH Ann. 14, 67-79). Although these processes have been proved to produce safe and effective vaccines designed to fight Meningococcal B epidemics (Granoff D. (2010), Clin. Infect. Dis. 50, S54-S65) their main drawback is that the detergent treatment favor bacterial cell lysis with the consequence that the OMV preparations are heavily contaminated with cytoplasmic proteins (Ferrari et al., (2006) Proteomics, 6, 1856-1866). This affect the immune response and can reduce the breath of protection. More recently, detergent-free methods for OMV production have been proposed. Such methods involve the separation of the bacterial culture supernatant from biomass and the purification of vesicles from the supernatant using tangential flow filtration (TFF) (Berlanda Scorza F. et al., (2012) PlosOne 7, e35616).
The yield of OMV production using centrifugation couple to TFF can exceed 100 mg/liter of culture and therefore the process is perfectly compatible with large scale production.
One of the potential issues encountered in using OMVs in vaccine applications is the presence of lipopolysaccharide (LPS), an endotoxin known to be reactogenic both in animals and humans. Possible strategies to reduce reactogenicity is to extract LPS from OMVs using mild detergents (Fredriksen J. H. et al, (1991) NIPH Ann. 14, 67-79) or to formulate OMVs with alum hydroxide which absorbs LPS and keeps it confined at the site of injection (Ferrari et al., (2006) Proteomics, 6, 1856-1866; Snape M. D. et al., (2010) Pediatr. Infect. Dis. J. 29, e71-e79). Another strategy is to genetically alter the LPS synthetic pathway of the strain used for OMV production so that the purified vesicles carry modified versions of LPS with reduced reactogenicity. For instance, in Neisseria meningitidis one promising mutant with attenuated endotoxin activity contains a deletion in the lpxL1 gene (also referred to as the msbB gene) (Fisseha M. et al., (2005) Infect. Immun., 73:4070-4080). This mutation results in a LPS carrying a penta-acylated lipid A, which is poorly recognized by human Toll-like receptor 4 (Steeghs L. et al. (2008) Infect. Immun., 76:3801-3807), instead of the more toxic hexa-acylated lipid A, which is present in the LPS produced by wild-type strains. The inactivation of msbB gene to produce less toxigenic OMVs has also been reported for Shigella, Salmonella and E. coli (Berlanda Scorza F. et al., (2012) PlosOne 7, e35616; Lee S-R et al., (2009) J. Microb. Biotechnol. 19, 1271-1279; Dong H. L. et al., (2011) Vaccine, 29, 8293-8301). In E. coli an additional mutation in the pagP gene has been described that, when combined with msbB mutation, results in the production of LPS with a fully penta-acylated lipid A which has a low reactogenicity property (Dong H. L. et al., (2011) Vaccine, 29, 8293-8301). Finally, by using Synthetic Biology, Needham and co-workers (Needham B. D. et al., (2013) PNAS, 110, 1464-1469) have created a collection of novel LPS synthetic pathways which lead to the synthesis of LPS carrying different modifications, each displaying distinct TLR4 agonist activities, cytokine induction and reactogenicity properties. OMVs purified from E. coli carrying such engineered LPS pathways have high potential for the design of vaccines with ad hoc modulated immunogenicity and adjuvanticity properties.
Neisseria meningitidis Factor H Binding Protein
Factor H binding protein (fHbp) is a 28 kDa surface-exposed lipoprotein of Neisseria meningitidis (F. Cantini et al., (2006), J Biol Chem., 281, 7220-7). fHbp is able to bind factor H (fH), the central regulator of the alternative complement pathway, and in this way it down-regulates complement activation (Lo H. et al., (2009) Lancet Infect. Dis. 9: 418-427) and impairs complement-mediated bacterial lysis by human plasma. fHbp is present on the surface of most meningococcal strains (Fletcher L. D. et al., (2004) Infect. Immun., 72: 2088-2100) and high levels of fHbp expression have been found in hyper-virulent meningococcal strains (Masignani et al., (2003) JEM 6: 789-799). Sequence analysis has classified fHbp in 3 main variant groups (var1, var2 and var3) (Masignani V et al (2003) JEM 6: 789-799). To study the immunogenic and functional properties of the protein, fHbp variant 1 was initially divided for convenience into three regions, named “domain” A, B, and C (Giuliani M M et al (2005) Infect Immun. 2: 1151-1160). Domain A encompasses amino acids 27 to 119 of the lipoprotein unprocessed precursor, domain B starts from amino acid 120 and ends at residue 183, and finally domain C spans from amino acid 184 to the end (amino acid 274) (F. Cantini et al., (2006), J Biol Chem., 281, 7220-7). More recent structural studies revealed that fHbp folds to form two β-barrels, with the amino-terminal barrel consisting of the A and part of the B regions and the carboxy-terminal barrel composed of the rest of the B and the C regions (Schneider M C et al., (2009) Nature, 458: 890-893; Faleri A et al, (2014) FASEB J. 4: 1644-53).
In the E. coli model Gram-negative bacterium, in which the lipoprotein sorting process has been well characterized, all outer membrane lipoproteins face the periplasmic space. They reach their final destination in two major steps (M. P. Bos et al. (2007) Ann. Rev. Microbiol., 61: 191-214). First, the protein is synthesized as a precursor carrying at its N-terminal a leader peptide (LP) including a lipobox. This LP is recognized by Sec pathway and the protein crosses the cytoplasmic membrane. The first amino acid of mature lipoprotein is a cysteine. Processing into mature form takes place on the periplasmic side of the inner membrane, where the thiol group on the side chain of the cysteine residue is modified by the covalent attachment of a diacylglycerol moiety and an amide-linked acyl group is attached to the N-terminus. Both the diacylglyceryl group and the amino-terminal acyl group participate to the anchorage of the lipoprotein to the membrane. The second step consists in the translocation of the acylated protein from the inner membrane to the inner leaflet of the outer membrane. Translocation is mediated by the Lol system (Tokuda H. (2009) Biosci. Biotechnol. Biochem., 73, 465-73). In N. meningitidis and other Gram-negative species, cell-surface exposed lipoproteins have been identified. The molecular mechanism responsible for the flipping out of fHbp and other surface lipoproteins from the inner to the outer leaflet of the outer membrane is not known.
fHbp is efficiently incorporated into OMVs released by Neisseria meningitidis and fHbp-containing OMVs have been shown to induce potent bactericidal antibodies against Neisseria meningitidis. Therefore, such engineered OMVs have been proposed as highly efficacious vaccines (EP 22 55 826 A2). fHbp has been successfully expressed in E. coli and the protein accumulates in the membranes fraction and is efficiently lipidated. Whether or not the protein, when expressed in E. coli, is exposed on the surface as it is the case in Neisseria meningitidis has not been reported. If a specific flippase is involved in fHbp translocation, when expressed in E. coli the protein is expected to be anchored to the inner leaflet of the outer membrane but not surface-exposed. Lipidated fHbp has been shown to induce high titers of bactericidal antibodies and an anti-meningococcus B vaccine based on lipidated fHbp purified from a recombinant E. coli strain is close to registration in USA and Europe (L. D. Fletcher et al., (2004) Inf. Immun. 72, 2008-2100; Richmond P C, et al., (2012), The Lancet Infectious Diseases, 12, 597-607).
Neisseria meningitidis NHBA (Neisseria Heparin Binding Antigen)
Neisseria heparin binding antigen (NHBA) is a Neisseria meningitidis surface-exposed lipoproteins which play important roles in pathogenesis. NHBA carries an arginine-rich region responsible for heparin binding, which correlates with an increased survival of N. meningtidis in human serum. NHBA fold consists of an 8-strand β-barrel that closely resembles the C-terminal domain of Nm-fHbp suggesting that they derive from a common ancestor (Esposito, V. et al. J Biol Chem, 2011. 286(48), 4167-75.
Consistent with its biological role NHBA extend out of the bacterial surface and is accessible to antibodies with bactericidal activity.
Aggregatibacter Actinomycetemcomitans Factor H Binding Protein (Aa-fHbp)
A. actinomycetemcomitans is a human pathogen associated with aggressive periodontitis and endocarditis (van Winkelhoff, A. J. and J. Slots, Periodontol 2000, 1999. 20, 122-35; Henderson, B., J. M. Ward, and D. Ready, Periodontol 2000, 2010. 54(1), 78-105).
The genome sequence of the pathogen revealed the presence of an Open Reading Frame (ORF) encoding a lipoprotein (Aa-fHbp) sharing 41% amino acid identity with Nm-fHbp var.1 (
Escherichia coli Maltose Binding Protein
The Maltose Binding Protein (MBP) is a periplasmic protein that specifically binds maltose and maltodextrins, with highest affinity for maltotriose (Schwartz et al., (1976) Eur. J. Biochem. 71, 167-170; Szmelcman et al., (1976) Eur. J. Biochem., 65, 13-19; Wandersman et aL, (1979) J. Bacteriol, 140, 1-13). Studies on the binding specificity have suggested that the binding site recognizes the glycosidic bond linking the glucose moities of maltose. MBP can be purified in dimeric form from bacteria. Maltose induces the conversion of the protein dimers to monomers and upon binding of the substrate, the monomer undergoes a conformational change (Szmelcman et al., (1976) Eur. J. Biochem., 65, 13-19; Zukin, S. (1979) Biochemistry 18, 2139-2145).
MBP is synthesized in large amounts. A fully induced cell may contain up to 40,000 monomers. Expression of the malE gene is activated by the MalT protein in the presence of maltose or maltodextrim and by the cyclic AMP receptor protein.
Like most outer membrane and periplasmic proteins, MBP is synthesized initially with an amino-terminal signal peptide which interacts with the cell secretory apparatus and is cleaved during the export process. The MBP precursor is active in binding maltose (Ferenci and Randall (1979) J. Biol. Chem. 254, 9979-9981). Interestingly, MBP is essential for the energy-dependent translocation of maltose and maltodextrins through the cytoplasmic membrane.
MBP is extensively used to express proteins in E. coli and to increase their solubility. Proteins of interest are usually expressed as MBP-fusions, and this has been shown to prevent aggregation through a mechanism that is not fully understood. MBP can itself be used as an affinity tag for purification of recombinant proteins. The fusion protein binds to amylose columns while all other proteins flow through. The MBP-protein fusion can be purified by eluting the column with maltose. Once the fusion protein is obtained in purified form, the protein of interest is often cleaved from MBP with a specific protease and separated from MBP by ion exchange chromatography.
Since MBP is a periplasmic protein, chimeric MBP carrying a heterologous protein at its C-terminus also accumulates in this bacterial compartment. However, the delivering of MBP chimera to OMVs and their exploitation as immunogens has never been tested.
OmpF
Escherichia coli expresses three major Outer Membrane Proteins (OMPs) OmpA, OmpC and OmpF.
OmpA is noncovalently anchored to peptidoglycan, and is constituted by eight β-barrel strands connected by four long loops at the outer membrane surface and three short periplasmic turns. The other two OMPs, OmpF and OmpC, are porins which serve as passive diffusion pores across the outer membrane. Expression of these two proteins is reciprocally regulated by medium osmolarity. OmpF is preferentially produced at low osmolarity, whereas OmpC is almost exclusively produced at high osmolarity. It has been proposed that the larger OmpF pore size is important for efficient nutrient uptake from nutritionally poor media, whereas the smaller OmpC pore size is important to exclude the passage of toxic bile salts across the outer membrane (Ferrario, M. et al., (1995) J. Bacteriol. 177, 103-1132).
E. coli has a highly sophisticated regulatory system to modulate the reciprocal expression of ompF and ompC. Both are controlled at the transcriptional level by the histidine kinase EnvZ, a transmembrane osmosensor, and the response regulator OmpR, a transcriptional factor (Hall, M. N., and Silhavy, T. J. (1981) J. Mol. Biol. 151, 1-15), and also at the translational level by antisense RNAs: micF for ompF mRNA (Mizuno, et al. (1984) Proc. Natl. Acad. Sci. U.S.A 81, 1966-1970) and micC for ompC mRNA (Chen, S., Zhang, A., Blyn, L. B., and Storz, G. (2004) J. Bacteriol. 186, 6689-66975).
When grown at 37° C. in rich media such as LB or in minimal media, such as M9 medium, supplemented with glycerol, E. coli BL21(D3) mostly produces OmpA and OmpF. The two proteins accumulate in the outer membrane vesicles (OMVs) released by the strain and when the ompA gene is inactivated OmpF constitutes more than 50% of total OMV proteins (Fantappiè et al., (2014) Journal of Extracellular Vesicles, 3, 24015).
Structurally, OmpF is characterized by the trimeric assembly of monomeric 16-stranded β-barrels, each containing its own hydrophilic pore. Each barrel essentially spans the thickness of the membrane. Except for the third loop that folds inward and constricts the channel opening, seven long loops connect adjacent pairs of β-strands on the extracellular side. Recently each OmpF loop has been systematically deleted with the aim of dissecting their role in pH and voltage sensitivity. Despite the fact that the study did not allow to completely clarify the contribution of the loops in OmpF activity, it nicely demonstrated that the protein can be mutilated in any of its external loop without appreciably effecting the level of expression, membrane integration and overall protein function (Basle et al., (2004) Protein engineering, design & selection, 9, 665-672).
No experimental evidence that OmpF-derived chimeras can be generated have been reported so far.
Cancer Vaccines
The notion that the immune system can recognize and mount a response against tumors was postulated in the late nineteenth century by Coley who demonstrated that attenuated bacteria or bacterial products injected into tumor-bearing patients in some cases resulted in tumor regression (Coley W B (1893) Am. J. Med. Sci. 105: 487-511). Nearly a century later, it was demonstrated that immunization of mice with mutated tumor cells could induce a protective anti-tumor immune response against non-immunogenic tumor (Van and Boon, (1982) PNAS, 79, 4718-4722). Together, these studies set a foundation for cancer immunotherapy research and demonstrated the therapeutic potential of strategies targeting immune modulation for tumor eradication and protection against tumor recurrences. Therefore, the development of cancer vaccines capable of generating an active tumor-specific immune response serves as a promising venue for cancer therapy.
Probably the best example that illustrates the potential of the immune system to fight cancer is given by Sipuleucel-T, the recently approved vaccine for prostate cancer patients. The vaccine is produced by isolating an individual patient's CD54+ white cells via leukapheresis, exposing the isolated cells ex vivo to PA2024, a protein antigen expressed in over 95% of prostate cancers, and infusing the vaccine back into the patient. Sipuleucel-T therefore consists of personalized primed APCs and of a mixed cell suspension containing also monocytes, macrophages, B and T cells, exposed to activated APCs (Lu C. et al. (2011) Exp. Opin. Biol. Ther. 11, 99-108). Although complicated and expensive to produce the vaccine clearly indicates that, if properly stimulated, the immune system can control tumor growth and progression.
Other promising cell-based vaccines are being developed by collecting Tumor Infiltrating Lymphocytes (TILs) from freshly dissected tumors, expanding them upon stimulation with tumor antigens (total tumor extracts or selected tumor antigens) and infusing TILs back into the patients (Restifo et al., (2012) Nature Rev. Immunol. 12, 269-281). Also this approach has shown to reduce tumor growth and to prolong overall survival.
A more practical way to develop cancer vaccines is to stimulate patient's immune system by injecting into patients specific cancer antigens formulated with proper adjuvants/immune potentiators (Berinstein N L (2007) Vaccine 25S, B72-B88). This approach has the great advantage to avoid the complication of collecting immune cells from each patient and of re-injecting them back after activation and/or amplification.
Several trials are ongoing exploiting this strategy. Among the most promising ones are two peptide-based vaccines, Her2-E75 (Nelipepimut-S) (Mittendorff E A et al., (2014) Annals of Oncology 25: 1735-1742) and EGFRvIII (Rindopepimut) (Del Vecchio C A et al. (2012), Expert Rev. Vaccines 11, 133-144) against Her2-positive breast cancer and glioblastoma, respectively. These vaccines, which are formulated with the immune stimulator GM-CSF, appear to have different mechanisms of action. The first primarily induces cytotoxic CD8+ T cells while the other mostly elicits humoral response.
However, despite demonstrated efficacy in various murine models, cancer vaccines have found little success in the clinic. Although several factors may contribute to the failure of therapeutic cancer vaccines in the clinic, the most important ones are i) the weak immunogenicity of Tumor Associated Antigens (TAAs), ii) central and peripheral immune tolerance to self TAAs, and iii) various immune evasion mechanisms employed by the progressing tumor.
Therefore, the success of therapeutic cancer vaccines may require formulations that induce potent immune responses that overcome immune tolerance to TAAs as well as reverse or inhibit tumor-mediated immune evasion mechanisms.
Human EGFRvIII
Glioblastoma multiforme (GBM) accounts for over 50% of primary brain tumors (Porter K R et al., (2010) Neuro. Oncol. 12, 520-527). Although gliomas often respond to radiotherapy, recurrence occurs in most of the patients with a median time to progression of 7 months (Grossman S A et al., (2004) Semin. Oncol. 31, 635-644). Tumor recurrence almost inevitably leads to death, the median overall survival of patients being only 15-16 months with surgical resection and combination chemo/radiation therapy (Vredenburgh J J et al. (2007) J. Clin. Oncol. 25, 4722-4729; Stupp R et al., (2005) N. Engl. J. Med. 352, 987-996).
Abnormal cell signaling by EGF receptor has been implicated in numerous cancers. Physiologically, EGF binds to the monomeric form of its receptor and this leads to receptor dimerization and autophosphorylation, which in turn triggers the downstream signal cascades (Salomon D S et al., (1995) Crit. Rev. Oncol. Hematol. 19, 183-232). In the majority of solid tumors, including breast, brain, head-and-neck, non-small-cell lung, renal, ovarian, prostate and colon cancer EGFR is overexpressed (Wong A J et al., (1992) Proc. Natl Acad. Sci. USA 89, 2965-2969; Gorgoulis V et al. (1992) Anticancer Res. 12, 1183-1187; Irish J C et al. (1993) Laryngoscope 103, 42-52; Korc M et al. (1986) Proc. Natl Acad. Sci. USA 83, 5141-5144; Moorghen M et al. (1990) Anticancer Res. 10, 605-611; Ishikawa J et al., (1990) Int. J. Cancer 45, 1018-1021; Zajchowski D et al., (1988) Cancer Res. 48, 7041-7047). EGFR overexpression leads to the enhancement of downstream signaling pathways stimulating growth and invasiveness of cancer cells. Furthermore, cell motility and VEGF-mediated angiogenesis is increased, while cell adhesion requirements are reduced.
In addition to overexpression, there is a naturally occurring variant of the EGF receptor called EGFRvIII. This variant was originally identified in GBM, and is now known to occur in up to 60% of primary GBM. EGFRvIII expression mainly occurs as a consequence of gene rearrangement and amplification even though alternative splicing event has also been implicated. Both gene rearrangement and alternative splicing result in an in-frame 801 base pair deletion of exons 2-7. This deletion gives rise to a truncated receptor that maintains its signal peptide, transmembrane, intracellular kinase and autophosphorylation domains, but lacks a significant portion of the extracellular ligand-binding domain, thus rendering EGFRvIII ligand independent and constitutively active. Other tumors have also been shown to express this variant, including lung, breast, ovarian and prostate cancer, but EGFRvIII is only rarely expressed in normal tissue (Moscatello D K et al., (1995) Cancer Res. 55, 5536-5539).
The in-frame deletion of the extracellular domain of EGFR creates a novel antigenic epitope which is exquisitely tumor-specific (Humphrey et al., (1990) PNAS, 87, 4207). Therefore, the newly generated epitope can be exploited in active and passive immunization. Indeed, a vaccine has been developed (Rindopepimut) which is based on a 14-amino acid peptide (LEEKKGNYVVTDHC) spanning the new epitope conjugated to keyhole limpet hemocyanin (KLH) and formulated with GM-CSF. A number of Phase II clinical trials in EGRFvIII-positive GBM patients have demonstrated that vaccination with Rindopepimut resulted in significantly higher progression-free and overall survival times (Del Vecchio et al., (2012) Exp. Rev. 11, 133). The vaccine is highly promising, but it has margins for improvement. For instance, the immune responses could be optimized. In one clinical trial GMTs were below 910 ng/ml with 6 responders out of 14 patients, and in a second trial all 22 enrolled patients responded but after as many as eight vaccine injections. Furthermore, the vaccine production process is relatively complex (three components with a chemical conjugation). Therefore, the development of an easy-to-produce vaccine with enhanced immunogenicity properties is highly desirable.
Human FAT-1
Human FAT gene family is a subclass of the cadherin superfamily, composed of four giant proteins (FAT1-4) of 500-600 kDa sharing structural similarities from invertebrates to mammals. Human FAT1 is a type 1 transmembrane protein carrying 34 cadherin repeats, five EGF-like repeats, a laminin A-G domain in the extracellular region and a cytoplasmic tail (Dunne, J. et al., (1995) Genomics 30, 207-23). The protein undergoes a proteolytic cleavage by Furin and is predicted to be further cleaved by γ secretase so that its intracellular domain (ICD) can translocate into the nucleus and directly activate cell signaling. FAT1 ICD also interacts with EnaNAPS and Scribble, promotes actin-mediated cell migration and inhibits YAP1-mediated cell proliferation (Moeller, M. J. et al., (2004) The EMBO journal, 23, 3769-79). In addition, FAT1 ICD also interacts with β-catenin and prevents its translocation to the nucleus (Morris, L. G. T. et al., (2013) Nature Genetics 45, 253-61).
Alteration of FAT1 expression and function has been clearly associated to several human cancers however its role in tumors is controversial.
Several publications provide evidence that in many tumors including oral cancer, astrocytoma, glioblastoma and ductal carcinoma FAT1 acts as a tumor suppressor (Chosdol, K. et al., (2009) BMC Cancer, 9, 5; Nakaya, K. et al., (2007) Oncogene, 26, 5300-8; Settakorn, J. et al., (2005) Journal of Clinical Pathology, 58, 1249-54). Indeed, it has been shown that FAT1 depletion leads to a significant stimulation of cell growth and proliferation, while its expression robustly suppresses tumor growth (Morris, L. G. T. et al., (2013) Cell Cycle (Georgetown, Tex.), 12, 1011-2; Morris, L. G. T. et al., (2013) Nature Genetics 45, 253-61). Furthermore, FAT1 gene falls within the genomic locus 4q35, a highly prevalent region of deletion in many types of human cancers. FAT1 deletion or loss of heterozygosity have been described in oral cancer (Nakaya, K. et al., (2007) Oncogene, 26, 5300-8), astrocytoma and glioblastoma (Chosdol, K. et al., (2009) BMC Cancer, 9, 5). Finally, in cholangiocarcinoma FAT1 shows a reduced plasma membrane localization (Settakorn, J. et al., (2005) Journal of Clinical Pathology, 58, 1249-54) and in invasive breast cancer is preferentially down-regulated (Kwaepila, N. et al., (2006) Pathology, 38, 125-31).
By contrast, in other cancers such as acute myeloid leukemia (AML), pre-B acute lymphoblastic leukemia (ALL) and T-ALL (De Bock, C. E. et al., (2012) Leukemia, 26, 918-26) FAT1 has been described to act as tumor promoter. FAT1 appears to be upregulated in hepatocarcinoma (Valletta, D. et al., (2014) Carcinogenesis, 35, 1407-15) and leukemia (de Bock et al. 2012). Furthermore, FAT1 up-regulation is an unfavorable prognostic factor for precursor B-cell acute lymphoblastic leukemia patients (De Bock, C. E. et al., (2012) Leukemia, 26, 918-26). Finally, recent studies in melanoma and pancreatic cancer have demonstrated that FAT1 undergoes an aberrant processing and an altered localization compared to normal cells (De Bock, C. E. et al., (2012) Leukemia, 26, 918-26).
Most recently, (PCT/EP2014/062419(Wo2014/198919) it was discovered that FAT1 is expressed in a large fraction of early and late stage CRCs. Moreover, a murine monoclonal antibody (mAb198.3) was isolated that selectively binds the surface of different FAT1-positive colon cancer cell lines and, upon binding, it is efficiently internalized. mAb198.3 also showed an intrinsic tumor-inhibiting activity in mouse models of human colon cancer xenografts.
mAb198.3 was shown to recognize an epitope present on cadherin domain 8 (D8) and cadherin domain 12 (D12), and antibody binding was efficiently abrogated in the presence of the synthetic peptide IQVEATDKDLGPNGHVTYSIVTDTD designed on the basis of the amino acid sequence of D8 domain.
Overall, this study provides the proof of concept that mAb198.3 could be exploited as novel tools for the treatment of CRC and paves the way to the use of FAT1 as an anti-CRC vaccine component.
Human MUC-1
Human mucin 1 (MUC1) is a large glycoprotein expressed at relative low level on the apical surface of healthy ductal epithelial cells. A peculiar characteristic of MUC1 is the presence of a region of variable number of tandem repeats (VNTR) in its extracellular domain. In particular, the VNTR region is constituted by a 20-amino acid sequence (PDTRPAPGSTAPPAHGVTSA) that repeats an average 20-150 times. In healthy epithelia, the VNTR is highly glycosylated on serines and threonines with long and branched O-linked carbohydrates.
During tumorigenesis, MUC1 loses polarity and its expression is highly upregulated. Furthermore, in tumor cells MUC1 is aberrantly hypoglycosylated. This is due to the downregulation of glycosyltransferases and the upregulation of sialyltransferases in tumor cells resulting in premature termination of glycosylation and expression of novel carbohydrate structures. This aberrant MUC1 expression is found in diverse kinds of human cancers including colon, lung, pancreas, breast, ovarian, prostate, kidney, stomach and head and neck (Ho S B et al., (1993) Cancer Res 53, 641-51; Baldus S E et al., (2002) Histopathology 40, 440-9; Rabassa M E et al. (2006) BMC Cancer 6, 253).
Abnormal expression and reduced glycosylation make MUC1 immunogenic. Cytotoxic T lymphocytes (CTLs) specific for VNTR epitopes have been found in pancreatic, breast and ovarian cancer patients Barnd, D. L. et al. (1989) Proc. Natl. Acad. Sci. USA 86, 7159-7163; Jerome, K. R. et al. (1991) Cancer Res. 51: 2908-2916; Ioannides, C. G. et al. (1993) J. Immunol. 151: 3693-3703). Interestingly, it has been found that peptide epitopes as well as novel truncated glycopeptide epitopes are presented by MHCI and induce MHC-restricted, MUC1-specific CTLs (Vlad, A. M. et al. (2002) J. Exp. Med. 196: 1435-1446). Furthermore, MUC1-specific antibodies have been found in cancer patients and have been associated to a survival benefit in breast, pancreatic, and ovarian cancer patients (vonMensdorff-Pouilly, S. et al. (2000) J. Clin. Oncol. 18: 574-583; Hamanaka, Y. et al. (2003) Int. J. Cancer. 103: 97-100; Pinheiro, S. P. et al. (2010) Cancer Epidemiol. Biomarkers Prev. 19: 1595-1601).
Since its first characterization as a human tumor-associated antigen recognized by T cells MUC1 has been studied as a highly promising antigen for passive (adoptive transfer of antibodies or T cells) and active (vaccines) immunotherapy of multiple human cancers (Kimura T. and Finn O J (2013) Expert Opin. Biol. Ther. 13: 35-49). However, the numerous clinical trials conducted over the last twenty years to demonstrate the potential of MUC1 in immunotherapy were somehow disappointing. Some protection was clearly observed but below the expectations and a MUC1-based cancer immunotherapy still needs substantial optimization before being considered as a valid therapeutic option for the treatment of cancer.
The inventors have found that bacterial Outer Membrane Vesicles (OMVs) loaded with tumor antigens fused to suitable bacterial proteins, are able to induce potent antigen-specific immune response. In particular it was observed that the following bacterial proteins: Factor H Binding Protein (fHbp), Maltose Binding Protein (MBP) and Outer Membrane Protein-F (Omp-F), are able to deliver the tumor antigen to the outer membrane compartment and correctly expose the antigen on the vesicular surface.
Accordingly, the invention provides a fusion protein comprising a bacterial protein selected from Factor H Binding Protein (fHbp), Maltose Binding Protein (MBP), Outer Membrane Protein-F (Omp-F), Neisseria heparin binding antigen (NHBA) and Aggregatibacter actinomycetemcomitans Factor H binding protein (Aa-fHbp), which is fused to one or more copies of a tumor antigen protein.
Preferably, in the fusion protein, the carboxyl end of the bacterial protein is linked to the amino terminus of the tumor antigen, directly or by interposition of a peptide linker. When the bacterial protein is OmpF, the tumor antigen can be inserted in—or replace—any of its external loops.
The fusion protein can contain one or more copies, preferably up to 10 copies, of tumor antigens, optionally separated by a peptide linker. As herein intended, the term “linker” designates any peptide sequence containing from 1 to 20 aa and preferably 2 to 4 aa.
The Factor H Binding Protein (fHbp) according to the invention includes the full-length protein or its domain A.
In preferred embodiments, the bacterial proteins are selected from the group consisting of: fHbp, SEQ ID NO:1; fHbpDomA, SEQ ID NO:2; OMP, SEQ ID NO:3 and MBP, SEQ ID NO:4; NHBA, SEQ ID NO:109; Aa-fHbp, SEQ ID NO:110.
Any tumor antigen can be potentially used to construct the fusion protein according to the invention and particularly the following:
(a) cancer-testis antigens including NY-ESO-1, SSX2, SCP1 as well as RAGE, BAGE, GAGE and MAGE family polypeptides, for example, GAGE-1, GAGE-2, MAGE-1 MAGE-2, MAGE-3, MAGE-4, MAGE-5, MAGE-6, and MAGE-12, which can be used, for example, to address melanoma, lung, head and neck, NSCLC, breast, gastrointestinal, and bladder tumours; (b) mutated antigens, including p53, associated with various solid tumours, e.g., colorectal, lung, head and neck cancer; p21/Ras associated with, e.g., melanoma, pancreatic cancer and colorectal cancer; CDK4, associated with, e.g., melanoma; MUM1 associated with, e.g., melanoma; caspase-8 associated with, e.g., head and neck cancer; CIA 0205 associated with, e.g., bladder cancer; HLA-A2-R1701, beta catenin associated with, e.g., melanoma; TCR associated with, e.g., T-cell non-Hodgkin lymphoma; BCR-abl associated with, e.g., chronic myelogenous leukemia; triosephosphate isomerase; KIA 0205; CDC-27, and LDLR-FUT; (c) over-expressed antigens, including, Galectin 4 associated with, e.g., colorectal cancer; Galectin 9 associated with, e.g., Hodgkin's disease; proteinase 3 associated with, e.g., chronic myelogenous leukemia; WT 1 associated with, e.g., various leukemias; carbonic anhydrase associated with, e.g., renal cancer; aldolase A associated with, e.g., lung cancer; PRAME associated with, e.g., melanoma; HER-2/neu associated with, e.g., breast, colon, lung and ovarian cancer; mammaglobin, alpha-fetoprotein associated with, e.g., hepatoma; KSA associated with, e.g., colorectal cancer; gastrin associated with, e.g., pancreatic and gastric cancer; telomerase catalytic protein, MUC-1 associated with, e.g., breast and ovarian cancer; G-250 associated with, e.g., renal cell carcinoma; p53 associated with, e.g., breast, colon cancer; and carcinoembryonic antigen associated with, e.g., breast cancer, lung cancer, and cancers of the gastrointestinal tract such as colorectal cancer; (d) shared antigens, including melanoma-melanocyte differentiation antigens such as MART-1/Melan A; gplOO; MC1R; melanocyte-stimulating hormone receptor; tyrosinase; tyrosinase related protein-1/TRP1 and tyrosinase related protein-2/TRP2 associated with, e.g., melanoma; (e) prostate associated antigens including PAP, PSA, PSMA, PSH-P1, PSM-P1, PSM-P2, associated with e.g., prostate cancer; (f) immunoglobulin idiotypes associated with myeloma and B cell lymphomas. In certain embodiments, the one or more TAA can be selected from pi 5, Hom/Mel-40, H-Ras, E2A-PRL, H4-RET, IGH-IGK, MYL-RAR, Epstein Barr virus antigens, EBNA, human papillomavirus (HPV) antigens, including E6 and E7, hepatitis B and C virus antigens, human T-cell lymphotropic virus antigens, TSP-180, pl85erbB2, pl 80erbB-3, c-met, mn-23H1, TAG-72-4, CA 19-9, CA 72-4, CAM 17.1, NuMa, K-ras, pi 6, TAGE, PSCA, CT7, 43-9F, 5T4, 791 Tgp72, beta-HCG, BCA225, BTAA, CA 125, CA 15-3 (CA 27.29\BCAA), CA 195, CA 242, CA-50, CAM43, CD68\KP1, CO-029, FGF-5, Ga733 (EpCAM), HTgp-175, M344, MA-50, MG7-Ag, MOV18, NB/70K, NY-CO-1, RCAS1, SDCCAG16, TA-90 (Mac-2 binding protein/cyclophilin C-associated protein), TAAL6, TAG72, TLP, TPS.
Preferably the tumor antigen is the entire sequence, a portion of it, or specific immunogenic epitopes of one of the following human proteins: TCTN1 (Gene ID: ENSG00000204852), TCTN2 (Gene ID: ENSG00000168778), TCTN3 (Gene ID: ENSG00000119977), HIGD2A (Gene ID: ENSG00000146066), HIGD2B (Gene ID: ENSG00000175202), C4ORF32 (Gene ID: ENSG00000174749), FAM62A (E-SYT1, Gene ID: ENSG00000139641), COLEC11 (Gene ID: ENSG00000118004), FSTL5 (Gene ID: ENSG00000168843), FAM82A2 (Gene ID: ENSG00000137824), SCARA5 (Gene ID: ENSG00000168079), VSTM1 (Gene ID: ENSG00000189068), RNF5 (Gene ID: ENSG00000183574), UNQ6126 (Gene ID: gi|169216088), DPY19L3 (Gene ID: ENSG00000178904), SLC39A10 (gene ID: ENSG00000196950), GPR107 (Gene ID: ENSG00000148358), COL20A1 (Gene ID: ENSG00000101203), GLT25D2 (Gene ID: ENSG00000198756), SYTL3 (Gene ID: ENSG00000164674), DENND1B (Gene ID: ENSG00000162701), C6orf98 (Gene ID: EG: 387079), FAM69B (Gene ID: ENSG00000165716), EMID1 (Gene ID: OTTHUMG00000030824), KLRG2 (GENE ID: ENSG00000188883), ERMP1 (GENE ID: ENSG00000099219), VMO1 (Gene ID: ENSG00000182853), C9orf46 (Gene ID: ENSG00000107020), F1137107 (Gene ID: ENSG00000177990), YIPF2 (Gene ID: ENSG00000130733), TRYX3 (PRSS58, ENSG00000258223.2), C14orf135 (Gene ID: ENSG00000126773), ANGPTL7 (Gene ID: ENSG00000171819), TPCN2 (Gene ID: ENSG00000162341), C18orf19 (Gene ID: ENSG00000177150), OLFML1 (Gene ID: ENSG00000183801), LYPD4 (Gene ID: ENSG00000101203), MEGF8 (Gene ID: ENSG00000105429), F1142986 (Gene ID: ENSG00000196460), SLC46A1 (Gene ID: ENSG00000076351), FAM180A (Gene ID: ENSG00000189320), CRISP-3 (GENE ID: ENSG00000096006)
These tumor antigens are disclosed in WO2010/086162, WO2010/086163, WO2011/051278, WO2011/051276, WO2011/051277, WO2011/051280, WO2011/051271, WO2011/135068, WO2014/198919, all in the applicant's name, the content of which is herein incorporated by reference.
More preferably the tumor antigen is selected from hEGFRvIII, hFAT-1 and hMUC-1, or an immunogenic fragment thereof. In preferred embodiments, the immunogenic fragments are selected from the following peptide sequences: LEEKKGNYVVTDH (EGFRvIII), IQVEATDKDLGPNGHVTYSIVTDTD (hFAT-1), GVTSAPDTRPAPGSTAPPAH (hMUC-1).
In preferred invention embodiments the fusion protein is selected from the group consisting of NOs:8 through 25, SEQ ID NOs:111 and 112.
The invention also provides an isolated bacterial outer membrane vesicle (OMV) loaded with a fusion protein as above defined. The isolated OMV can contain a fusion protein carrying one species of tumor antigen or a plurality of fusion proteins carrying different tumor antigens.
The OMV can be isolated and purified from Gram-negative bacteria, including species from any of genera Escherichia, Shigella, Neisseria, Moraxella, Bordetella, Borrelia, Brucella, Chlamydia, Haemophilus, Legionella, Pseudomonas, Yersinia, Helicobacter, Salmonella, Vibrio, etc. For example, the vesicles may be from Bordetella pertussis, Borrelia burgdorferi, Brucella melitensis, Brucella ovis, Chlamydia psittaci, Chlamydia trachomatis, Moraxella catarrhalis, Escherichia coli (including extraintestinal pathogenic strains), Haemophilus influenzae (including non-typeable stains), Legionella pneumophila, Neisseria gonorrhoeae, Neisseria meningitidis, Neisseria lactamica, Pseudomonas aeruginosa, Yersinia enterocolitica, Helicobacter pylori, Salmonella enterica (including serovar typhi and typhimurium), Vibrio cholerae, Shigella dysenteriae, Shigella flexneri, Shigella boydii or Shigella sonnei.
N. meningitidis OMVs have a proven safety record in humans and so they are a preferred choice. Another useful choice is E. coli vesicles, for example the BL21(DE3) strain (see Methods).
OMVs are prepared artificially from bacteria, and may be prepared using detergent treatment (e.g. with deoxycholate), or by non-detergent means (e.g. see WO2004/019977). Techniques for forming OMVs include treating bacteria with a bile acid salt detergent (e.g. salts of lithocholic acid, chenodeoxycholic acid, ursodeoxycholic acid, deoxycholic acid, cholic acid, ursocholic acid, etc.) at a pH sufficiently high not to precipitate the detergent (WO01/91788). Other techniques may be performed substantially in the absence of detergent (WO2004/019977) using techniques such as sonication, homogenisation, microfluidisation, cavitation, osmotic shock, grinding, French press, blending, etc. Methods using no or low detergent can retain useful antigens (WO2004/019977). Thus a method may use an OMV extraction buffer with about 0.5% deoxycholate or lower e.g. about 0.2%, about 0.1%, <0.05% or zero.
Bacterial vesicles can conveniently be separated from whole bacterial culture by filtration e.g. through a 0.22 μm filter. Bacterial filtrates may be clarified by centrifugation, for example high-speed centrifugation {e.g. 20,000×g for about 2 hours). Another useful process for OMV preparation is described in WO2005/004908 and involves ultrafiltration on crude OMVs, instead of high-speed centrifugation. The process may involve a step of ultracentrifugation after the ultrafiltration takes place. A simple process for purifying bacterial vesicles is described in WO2011/036562, comprising: (i) a first filtration step in which the vesicles are separated from the bacteria based on their different sizes, with the vesicles passing into the filtrate e.g. using a function (Basle et al., (2004) Protein engineering, design & selection, 9, 665-672).
In a further embodiment, the invention provides an immunogenic composition comprising a bacterial outer membrane vesicle as herein disclosed, together with pharmaceutical acceptable vehicles and excipients. The immunogenic composition may contain a mixture of outer membrane vesicles differing from each other for the type of tumor antigen, for the bacterial protein, or both.
The compositions of the invention for administration to subjects are preferably vaccine compositions. Vaccines according to the invention may either be prophylactic {e.g. to prevent cancer) or therapeutic {e.g. to treat cancer). Pharmaceutical compositions used as vaccines comprise an immunologically effective amount of antigen(s), as well as any other components, as needed. By ‘immunologically effective amount’, it is meant that the administration of that amount to an individual, either in a single dose or as part of a series, is effective for treatment or prevention. This amount varies depending upon the health and physical condition of the individual to be treated, age, the taxonomic group of individual to be treated {e.g. non-human primate, primate, etc.), the capacity of the individual's immune system to stimulate antibody production, the degree of protection desired, the formulation of the vaccine, the treating doctor's assessment of the medical situation, and other relevant factors. It is expected that the amount will fall in a relatively broad range that can be determined through routine trials. The antigen content of compositions of the invention will generally be expressed in terms of the amount of protein per dose. The amount of OMVs in compositions of the invention may generally be between 10 and 500 μg, preferably between 25 and 200 μg, and more preferably about 50 μg or about 100 μg.
Compositions of the invention may be prepared in various liquid forms. For example, the compositions may be prepared as injectables, either as solutions or suspensions. The composition may be prepared for pulmonary administration e.g. by an inhaler, using a fine spray. The composition may be prepared for nasal, aural or ocular administration e.g. as spray or drops, and intranasal vesicle vaccines are known in the art. Injectables for intramuscular administration are typical. Injection may be via a needle (e.g. a hypodermic needle), but needle-free injection may alternatively be used.
The OMVs and the immunogenic compositions according to the invention are conveniently used for the stimulation of an immune response against tumor in a subject in need thereof. Particularly they can be used for the prevention or treatment of different types of tumor, including but not limited to
bronchogenic carcinoma, nasopharyngeal carcinoma, laryngeal carcinoma, small cell and non-small cell lung carcinoma, lung adenocarcinoma, hepatocarcinoma, pancreatic carcinoma, bladder carcinoma, colon carcinoma, breast carcinoma, cervical carcinoma, ovarian carcinoma, or lymphocytic leukaemias, prostate cancer.
In a preferred embodiment, the isolated bacterial outer membrane vesicles or the immunogenic composition are used in the prevention or treatment of tumors selected from breast, brain, head-and-neck, non-small cell lung, renal, ovarian, kidney, stomach, prostate and colon cancer, oral cancer, astrocytoma, glioblastoma, ductal carcinoma, cholangiocarcinoma, hepatocarcinoma, acute myeloid leukemia, acute lymphoblastic leukemia, melanoma, pancreatic cancer and prostate cancer.
In one embodiment, the DNA coding sequence of a selected tumor antigen is fused to the 3′ end of the gene coding for the full length Neisseria meningitidis Factor H Binding Protein (fHbp) (SEQ ID NO:85) and such gene fusion is inserted into an appropriate plasmid expression vector in order to drive the expression of a fusion protein constituted by fHbp carrying the foreign antigen at its carboxyl terminus. Because of the presence of the fHbp leader sequence and lipidation site the fusion protein is delivered to the bacterial outer membrane and subsequently incorporated into OMVs. It was surprisingly found that tumor antigens fused to the C-terminus of fHbp are exposed to the surface of E. coli cells and of its derived OMVs.
In another embodiment, the DNA coding sequence of a selected tumor antigen is fused to the 3′ end of the gene coding for the domain A of fHbp (fHbpDomA, SEQ ID NO:86) and such gene fusion is inserted into an appropriate plasmid expression vector in order to drive the expression of a fusion protein constituted by fHbp carrying the foreign antigen at its carboxyl terminus. Because of the presence of the fHbp leader sequence and lipidation site the fusion protein is delivered to the bacterial outer membrane and subsequently incorporated into OMVs. It has been surprisingly found that when expressed in E. coli, fHbpDomA is delivered to the outer membrane and tumor antigens fused to the C-terminus of fHbpDomA are exposed to the surface of E. coli cells and of its derived OMVs.
In another embodiment, the DNA coding sequence of a selected tumor antigen is fused to the 3′ end of the full length gene coding for the Neisseria meningitidis NHBA (SEQ ID NO:113) and such gene fusion is inserted into an appropriate plasmid expression vector in order to drive the expression of a fusion protein constituted by NHBA carrying the foreign antigen at its carboxyl terminus. Because of the presence of the NHBA leader sequence and lipidation site the fusion protein is delivered to the bacterial outer membrane and subsequently incorporated into OMVs. It has been surprisingly found that when expressed in E. coli, NHBA is delivered to the outer membrane and tumor antigens fused to the C-terminus of NHBA are exposed to the surface of E. coli cells and of its derived OMVs.
In another embodiment, the DNA coding sequence of a tumor antigen of interest is fused to the 3′ end of the gene coding for the E. coli Maltose Binding Protein (MBP, SEQ ID NO:100). Such gene fusion is inserted into an appropriate plasmid expression vector in order to drive the expression of a fusion protein constituted by MBP carrying the foreign antigen at its carboxyl terminus. It has surprisingly been found that such fusion protein can be efficiently incorporated into OMVs.
In another embodiment, the DNA coding sequence of a selected tumor antigen is inserted into coding sequence of ompF in correspondence of DNA regions coding for the external loops of the protein. Such gene fusions are inserted into appropriate plasmid expression vectors in order to drive the expression of engineered OmpF in which any of the external loop is replaced with the foreign antigen. It has surprisingly been found that the engineered OmpF are efficiently integrated into the E. coli outer membrane and the tumor antigen is efficiently exposed on the bacterial surface, and that the engineered OmpF is also efficiently incorporated into OMVs.
In a further embodiment, the tumor antigen is the EGFRvIII peptide LEEKKGNYVVTDH. The DNA sequence coding for one or more copies of the EGFRvIII peptide is ligated at the 3′ end of gene encoding the full length fHbp in order to generate gene chimeras encoding fHbp with one or more copies of EGFRvIII peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimera are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying fHbp-EGFRvIII peptide fusions induce potent anti-EGFRvIII peptide immune responses.
In another embodiment, the DNA sequence coding for one or more copies of the EGFRvIII peptide is ligated at the 3′ end of gene encoding fHbpDomA (SEQ ID NO:86) in order to generate gene chimera encoding fHbp with one or more copies of EGFRvIII peptide fused at its C-terminus. A relevant aspect of this invention is that when the gene chimera are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying fHbpDomA-EGFRvIII peptide fusions induce potent anti-EGFRvIII peptide immune responses.
In another embodiment, the DNA sequence coding for one or more copies of the EGFRvIII peptide is ligated at the 3′ end of gene encoding the full length NHBA (SEQ ID NO: 113) in order to generate gene chimeras encoding NHBA with one or more copies of EGFRvIII peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimera are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs.
In yet another embodiment, the DNA sequence coding for one copy of the EGFRvIII peptide is ligated at the 3′ end of gene encoding MBP (SEQ ID NO:100) in order to generate a gene chimera encoding MBP with one copy of EGFRvIII peptide fused at its C-terminus. A relevant aspect of this invention is that when the gene chimera is expressed in E. coli the hybrid protein is efficiently incorporated into OMVs. Remarkably, OMVs carrying MBP-EGFRvIII peptide fusion induce potent anti-EGFRvIII peptide immune responses.
In a further embodiment, the DNA sequence coding for one or more copies of EGFRvIII peptide is inserted into coding sequence of ompF in correspondence of DNA regions coding for the external loops of the proteins. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying OmpF-EGFRvIII peptide fusions expose the EGFRvIII peptide to the surface of OMVs and are accessible to anti-EGFRvIII antibody binding.
In another embodiment, the foreign antigen is the FAT1-derived peptide IQVEATDKDLGPNGHVTYSIVTDTD. The DNA sequence coding for one or more copies of FAT1 peptide is ligated at the 3′ end of gene encoding the full length fHbp in order to generate gene chimeras encoding fHbp with one or more copies of FAT1 peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying fHbp-FAT1 peptide fusions induce potent anti-FAT1 peptide immune responses.
In yet another embodiment, the DNA sequence coding for one or more copies of the FAT1 peptide is ligated at the 3′ end of gene encoding fHbpDomA (SEQ ID NO:86) in order to generate gene chimeras encoding fHbp with one or more copies of FAT1 peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying fHbpDomA-FAT1 peptide fusions induce potent anti-FAT1 peptide immune responses.
In another embodiment, the DNA sequence coding for one or more copies of the FAT1 peptide is ligated at the 3′ end of the gene encoding MBP (SEQ ID NO:100) in order to generate gene chimeras encoding fHbp with one or more copies of FAT1 peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the periplasm and incorporated into OMVs. Remarkably, OMVs carrying MBP-FAT1 peptide fusions induce potent anti-EGFRvIII peptide immune responses.
In a further embodiment, the foreign antigen is the MUC1-derived peptide GVTSAPDTRPAPGSTAPPAH. The DNA sequence coding for one or more copies of MUC1 peptide is ligated at the 3′ end of gene encoding the full length fHbp (SEQ ID NO:85) in order to generate gene chimeras encoding fHbp with one or more copies of MUC1 peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying fHbp-MUC1 peptide fusions induce potent anti-MUC1 peptide immune responses.
In a further embodiment, the DNA sequence coding for one or more copies of the MUC1 peptide is ligated at the 3′ end of gene encoding fHbpDomA (SEQ ID NO:86) in order to generate gene chimeras encoding fHbpDomA with one or more copies of MUC1 peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the outer membrane and incorporated into OMVs. Remarkably, OMVs carrying fHbpDomA-MUC1 peptide fusions induce potent anti-MUC1 peptide immune responses.
In another embodiment, the DNA sequence coding for one or more copies of MUC1 peptide is ligated at the 3′ end of gene encoding MBP (SEQ ID NO:100) in order to generate gene chimeras encoding MBP with one or more copies of MUC1 peptide fused at their C-terminus. A relevant aspect of this invention is that when the gene chimeras are expressed in E. coli the hybrid proteins are efficiently delivered to the periplasm and incorporated into OMVs. Remarkably, OMVs carrying MBP-MUC1 peptide fusions induce potent anti-MUC1 peptide immune responses.
In another embodiment, the DNA sequence encoding for the Aggregatibacter actinomycetemcomitans factor H binding protein (Aa-fHbp, SEQ ID NO:116) gene was cloned in E. coli and used to decorate OMVs. A relevant aspect of this invention is that when the gene is cloned in E. coli the protein is efficiently expressed, delivered to the external side of the outer membrane and incorporated into OMVs.
BL21(DE3)/ΔompA (pET-fHbp-FAT1) and BL21(DE3)/ΔompA (pET-fHbpDomA-FAT1) E. coli strains were grown in 10 ml LB medium at 37° C. and when the cultures reached OD600=0.6, the expression of the fusion proteins was induced by addition of 1 mM IPTG. After 2 hours growth bacteria were collected by centrifugation and incubated with 50 μl of an appropriate dilution of anti-FAT1 mAb198.3 or, as negative controls, with PBS containing 1% BSA or with an unrelated mAb. After 1 hour, bacterial cells were washed with PBS containing 1% BSA and subsequently incubated for 30 minutes on ice with goat anti-mouse antibodies added at a final dilution of 1:200. Finally, after 2 wash steps, pellets were re-suspended in 200 μl of PBS and analyzed with FACS CANTOII. Collected data were analyzed with FlowJo software.
(A) Protein sequence alignment of Nm-fHbp and Aa-fHbp. (B) SDS-PAGE analysis of total cell extracts (TL) and OMVs isolated from BL21□ompA (pET_Aa-fHbp). (C) Western blot analysis of OMVs purified from BL21□ompA (pET_Aa-fHbp). (D) Assessment of Aa-fHbp localization by FACS analysis. Aa-fHbp detection was carried out using anti-His-tag antibodies. (E) Assessment of Aa-fHbp localization by proteinase K surface shaving
Engineered OMVs Expressing EGFRvIII Peptide
The detailed description of the construction of the protein chimeras and the preparation of the engineered OMVs decorated with the different protein fusions are reported below.
OMV Engineering with fHBPvIII
Construction of pET-fHbpvIII and pET-fHbpDomAvIII Plasmids
In an attempt to express and deliver the EGFRvIII peptide to the membrane compartment of E. coli OMVs the Neisseria meningitidis fHbp lipoprotein was used as a carrier. To this purpose one or three copies of the EGFRvIII peptide were fused to the 3′ end of either the full length fHbp or fHbpDomA. The first step to achieve this was to amplify the sequence coding for full length fHbp (fHbpFL) and fHbpDomA and to clone the amplified sequences into pET21 plasmid. fHbpFL and fHbpDomA coding sequences (SEQ ID NOs:85 and 86) were amplified by PCR from N. meningitidis MC58 genomic DNA using primers fHbp-ss-F/fHbp R and fHbp-ss-F/fHbp A rev, respectively (Table 1 and
Expression of the fHbpvIII and fHbpDomAvIII Heterologous Proteins in E. coli BL21(DE3)/ΔompA Strain
The four recombinant plasmids encoding fHbp and fHbpDomA fused to one copy and three copies of the EGFRvIII peptide were used to transform E. coli strain BL21(DE3)/ΔompA. Four recombinant strains were obtained: BL21(DE3)/ΔompA(pETfHbpFL-vIII), BL21(DE3)/ΔompA (pETfHbpFLvIIIx3), BL21(DE3)/ΔompA(pETfHbpDomAvIII) and BL21(DE3)/ΔompA (pETfHbpDomAvIIIx3). Each strain was grown in LB medium and when cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., cells were collected and total protein extracts were analyzed by Western Blot. To this aim, 25 μg of total proteins from each strain were separated by SDS-PAGE and proteins were transferred to nitrocellulose filters. The filters were blocked overnight at 4° C. by agitation in blocking solution (10% skimmed milk and 0.05% Tween in PBS), followed by incubation for 90 minutes at 37° C. with a 1:1,000 dilution of rabbit anti-vIII polyclonal antibodies in 3% skimmed milk and 0.05% Tween in PBS. After 3 washing steps in PBS-Tween, the filters were incubated in a 1:2,000 dilution of peroxidase-conjugated anti-rabbit immunoglobulin (Dako) in 3% skimmed milk and 0.05% Tween in PBS for 1 hour, and after 3 washing steps, bound conjugated IgGs were detected using the Super Signal West Pico chemo-luminescent substrate (Pierce). As shown in
in the cell lysates. No immune reactive bands were detected in total lysates from E. coli cells carrying either pET-fHbp expressing full length fHbp or empty pET21 plasmid.
Analysis of fHbpvIII and fHbpDomAvIII Expression in OMVs
Having demonstrated that the four fusion proteins were expressed in E. coli BL21(DE3)/ΔompA strain, the presence of the antigens in the OMV fraction was analyzed. The four recombinant strains BL21(DE3)/ΔompA(pETfHbpFL-vIII), BL21(DE3)/ΔompA(pETfHbpFLvIIIx3), BL21(DE3)/ΔompA (pETfHbpDomAvIII) and BL21(DE3)/ΔompA(pETfHbpDomAvIIIx3) were grown in LB medium and when the cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., vesicles were purified from culture supernatants by using ultrafiltration coupled to ultracentrifugation. More specifically, OMVs were collected from culture supernatants by filtration through a 0.22 μm pore size filter (Millipore) and by high-speed centrifugation (200.000×g for 2 hours). Pellets containing OMVs were finally suspended in PBS. The presence of the antigens in OMVs preparations was verified by Western Blot analysis as described in the previous section (
Analysis of Cellular Localization of fHbpvIII and fHbpvIIIx3
The localization of recombinant fHbpvIII and fHbpvIIIx3 fusion proteins was evaluated by flow cytometry. To this aim, recombinant E. coli strains BL21(DE3)/ΔompA (pETfHbpFL-vIII), BL21(DE3)/ΔompA (pETfHbpFLvIIIx3) and E. coli BL21(DE3)/ΔompA(pET21), as negative control, were grown at 37° C. under agitation. When cultures reached an OD600 value of 0.6, IPTG was added at a final concentration of 1 mM and bacteria were grown for 2 additional hours. Subsequently, bacteria cells corresponding to those contained in 1 ml culture at OD600=1 were collected by centrifugation at 13,000×g for 5 minutes and pellets were re-suspended in 50 ml of PBS containing 1% BSA. 50 μl of cell suspensions were incubated with 50 μl of an appropriate dilution of anti-vIII rabbit polyclonal antibodies raised against the vIII peptide conjugated to Keyhole Limpet Hemocyanin (KLH) or with 50 μl of PBS containing 1% BSA as negative control. After 1 hour, 100 μl of PBS containing 1% BSA were added and the suspensions were centrifuged at 3,000×g for 10 minutes and supernatants discarded. Pellets were washed with 200 μl of PBS containing 1% BSA and bacteria subsequently incubated for 30 minutes on ice with goat anti-rabbit antibodies (Alexa flour488, Life Technology) added at a final dilution of 1:2,000. Finally, After 2 wash steps, pellets were re-suspended in 200 μl of PBS and analyzed with FACS CANTOII evaluating collected data with FlowJo software. As shown in
Engineered OMVs Carrying Recombinant fHbpvIII and fHbpDomAvIII Fusion Proteins Induce High Antibodies Titers in Immunized Mice
To test whether OMVs purified from fHbpvIIIx3 and fHbpDomAvIIIx3 recombinant strains were capable of inducing vIII-specific antibody responses, CD1 mice were i.p. immunized three times at two-week intervals with 10 μg of OMVs formulated in Alum. Blood samples were collected nine days after second dose (post2) and seven days after third dose (post3) administration and anti-vIII IgGs were detected by using plates coated in each well with 0.5 μg of synthetic vIII peptide conjugated to Keyhole limpet hemocyanin (vIII-KHL). Serum deriving from mice immunized with empty OMVs was used as negative control. More specifically, coating was carried out by incubating plates overnight at 4° C. with 100 μl of conjugated peptide (5 μg/ml). Subsequently, wells were washed three times with PBST (0.05% Tween 20 in PBS, pH 7.4), incubated with 100 μl of 1% BSA in PBS for 1 h at room temperature and washed again three times with PBST. Serial dilutions of serum samples in PBST containing 1% BSA were added to the plates, incubated 2 h at 37° C., and washed three times with PBST. Then 100 μl/well of 1:2.000 diluted, alkaline phosphatase-conjugated goat anti-mouse IgGs were added and left for 2 h at 37° C. After triple PBST wash, bound alkaline phosphatase-conjugated antibodies were detected by adding 100 μl/well of 3 mg/ml para-nitrophenyl-phosphate disodium hexahydrate (Sigma-Aldrich) in 1M diethanolamine buffer (pH 9.8). After 10 minute incubation at room temperature, the reaction was stopped with 100 μl 7% EDTA and substrate hydrolysis was analyzed at 405 nm in a microplate spectrophotometer.
As shown in
OMV Engineering with NHBAvIII
Construction of pET-NHBAvIII Plasmids
In order to express and deliver the EGFRvIII peptide to the membrane compartment of E. coli OMVs the Neisseria meningitidis NHBA lipoprotein was used as a carrier. To this purpose one or three copies of the EGFRvIII peptide were fused to the 3′ end of the full length NHBA. The first step to achieve this was to amplify the sequence coding for full length NHBA and to clone the amplified sequences into pET21 plasmid. NHBA coding sequences (SEQ ID NO:113) were amplified by PCR from N. meningitidis MC58 genomic DNA using primers NHBA-F/NHBA-R (Table 1 and
Expression of the NHBAvIII Heterologous Protein in E. coli BL21(DE3)/ΔompA Strain
The two recombinant plasmids encoding NHBA fused to one copy and three copies of the EGFRvIII peptide were used to transform E. coli strain BL21(DE3)/ΔompA. Two recombinant strains were obtained: BL21(DE3)/ΔompA(pET-NHBA-vIII) and BL21(DE3)/ΔompA(pET-NHBAvIIIx3). Each strain was grown in LB medium and when cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., cells were collected and total protein extracts were analyzed by Western Blot. To this aim, 25 μg of total proteins from each strain were separated by SDS-PAGE and proteins were transferred to nitrocellulose filters. The filters were blocked overnight at 4° C. by agitation in blocking solution (10% skimmed milk and 0.05% Tween in PBS), followed by incubation for 90 minutes at 37° C. with a 1:1,000 dilution of rabbit anti-vIII polyclonal antibodies in 3% skimmed milk and 0.05% Tween in PBS. After 3 washing steps in PBS-Tween, the filters were incubated in a 1:2,000 dilution of peroxidase-conjugated anti-rabbit immunoglobulin (Dako) in 3% skimmed milk and 0.05% Tween in PBS for 1 hour, and after 3 washing steps, bound conjugated IgGs were detected using the Super Signal West Pico chemo-luminescent substrate (Pierce). As shown in
Analysis of Cellular Localization of NHBAvIII and NHBAvIIIx3
The localization of recombinant NHBAvIII and NHBAvIIIx3 fusion proteins was evaluated by flow cytometry. To this aim, recombinant E. coli strains BL21(DE3)/ΔompA (pET-NHBAvIII), BL21(DE3)/ΔompA (pET-NHBAvIIIx3) and E. coli BL21(DE3)/ΔompA(pET21), as negative control, were grown at 37° C. under agitation. When cultures reached an OD600 value of 0.6, IPTG was added at a final concentration of 1 mM and bacteria were grown for 2 additional hours. Subsequently, bacteria cells corresponding to those contained in 1 ml culture at OD600=1 were collected by centrifugation at 13,000×g for 5 minutes and pellets were re-suspended in 50 ml of PBS containing 1% BSA. 50 μl of cell suspensions were incubated with 50 μl of an appropriate dilution of anti-vIII rabbit polyclonal antibodies raised against the vIII peptide conjugated to Keyhole Limpet Hemocyanin (KLH) or with 50 μl of PBS containing 1% BSA as negative control. After 1 hour, 100 μl of PBS containing 1% BSA were added and the suspensions were centrifuged at 3,000×g for 10 minutes and supernatants discarded. Pellets were washed with 200 μl of PBS containing 1% BSA and bacteria subsequently incubated for 30 minutes on ice with goat anti-rabbit antibodies (Alexa flour488, Life Technology) added at a final dilution of 1:2,000. Finally, After 2 wash steps, pellets were re-suspended in 200 μl of PBS and analyzed with FACS CANTOII evaluating collected data with FlowJo software. As shown in
OMV Engineering with OmpFvIII
Construction of pET-OmpFvIII Plasmids
With the aim of expressing EGFRvIII peptide on the OMVs surface, the fusion of vIII peptide to Outer Membrane Protein F (OmpF) was attempted. OmpF is a protein embedded in the outer membrane with eight loops (L) exposed to the extracellular compartment (
To substitute loops L1, L2, L4, L6, L7 and L8 with EGFRvIII, the polymerase incomplete primer extension (PIPE) cloning method was used. pET-OmpF was amplified by PCR using the primer couples Loop1Vf/Loop1Vr, Loop2Vf/Loop2Vr, Loop4Vf/Loop4Vr, Loop6Vf/Loop6Vr, Loop7Vf/Loop7Vr and Loop8Vf/Loop8Vr, respectively (Table 1). Each couple was designed to anneal to the flanking regions of a selected loop and to carry complementary tails that when annealed reconstituted the sequence coding for vIII peptide. PCR-amplification followed by E. coli HK100 transformation resulted in replacement of the loop with the DNA sequence coding for EGFRvIII. Plasmids pET-OmpFvIII_L1, pET-OmpFvIII_L2, pET-OmpFvIII_L4, pET-OmpFvIII_L6, pET-OmpFvIII_L7 and pET-OmpFvIII_L8 were obtained. The correctness of the replacement of each loop coding sequence with the small fragment coding for the vIII peptide was verified by sequence analysis (SEQ ID NOs:92, 94, 95, 97, 98 and 99).
In an attempt to maximize the exposure of the peptide to the extracellular milieu two other constructs were generated, in which L1 or L4 were substituted with three copies of EGFRvIII (
Expression of OmpFvIII and OmpFvIIIx3 Heterologous Proteins in E. coli BL21(DE3)/ΔompA
The eight plasmids: pET-OmpFvIII_L1, pET-OmpFvIII_L2, pET-OmpFvIII_L4, pET-OmpFvIII_L6, pET-OmpFvIII_L7, pET-OmpFvIII_L8, pET-OmpFvIIIx3_L1 and pET-OmpFvIIIx3_L4 were used to transform E. coli strain BL21(DE3)/ΔompA. Expression of engineered OmpF carrying the vIII peptide in each loop was analyzed in bacterial lysates by SDS-PAGE and Western Blot. E. coli BL21(DE3)/ΔompA strains carrying each plasmid encoding engineered OmpF were grown in LB medium and when cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., cells were collected and 25 μg of total protein extracts were analyzed by SDS-PAGE. As shown in
Expression of OmpFvIII and OmpFvIIIx3 Heterologous Proteins into OMVs
Having demonstrated that engineered OmpF carrying one or three copies of vIII peptide in correspondence of one of the OmpF external loops were expressed in E. coli BL21(DE3)/ΔompA, the presence of engineered OmpF in the OMV fraction was analyzed. To this aim, the recombinant strains BL21(DE3)/ΔompA(pET-OmpFvIII_L1), BL21(DE3)/ΔompA (pET-OmpFvIII_L2), BL21(DE3)/ΔompA (pET-OmpFvIII_L4), BL21(DE3)/ΔompA (pET-OmpFvIII_L6), BL21(DE3)/ΔompA (pET-OmpFvIII_L7), and BL21(DE3)/ΔompA (pET-OmpFvIII_L8) were grown in LB medium and when the cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., vesicles were purified from culture supernatants by using ultrafiltration coupled to ultracentrifugation. More specifically, OMVs were collected from culture supernatants by filtration through a 0.22 μm pore size filter (Millipore) and by high-speed centrifugation (200,000×g for 2 hours). Pellets containing OMVs were finally suspended in PBS. The presence of the engineered OmpF proteins in each OMV preparation was verified by Western Blot analysis as already described using anti-vIII polyclonal antibodies. As shown in
Analysis of vIII Expression on the Surface of OmpFvIII and OmpFvIIIx3 Recombinant Strains
Finally, the surface expression of the vIII peptide was analyzed by Flow Cytometry in BL21(DE3)/ΔompA(pET-OmpFvIII_L2), BL21(DE3)/ΔompA (pET-OmpFvIII_L4), BL21(DE3)/ΔompA (pET-OmpFvIII_L6), BL21(DE3)/ΔompA (pET-OmpFvIIIx3_L1), and BL21(DE3)/ΔompA (pET-OmpFvIIIx3_L4), strains. Bacterial cultures were grown at 37° C. under agitation. When cultures reached an OD600 value of 0.6, IPTG was added at a final concentration of 1 mM and bacteria were grown for 3 additional hours. Subsequently, 1 ml bacteria cells were collected by centrifugation at 5,000×g for 5 minutes. After a wash in 1% BSA/PBS bacteria were resuspended in 5 ml 1% BSA/PBS. 50 μl of cell suspensions were incubated with 5 μl of an appropriate dilution of rabbit anti-vIII polyclonal antibodies raised against the vIII peptide conjugated to Keyhole Limpet Hemocyanin (KLH) or, as negative control, with 5 μl of PBS containing 1% BSA. After 1 hour, the suspensions were centrifuged at 5.000×g for 5 minutes and supernatants discarded. Pellets were washed with 100 μl of PBS containing 1% BSA and bacteria subsequently incubated for 1 hour at 4° C. with goat anti-rabbit antibodies (Alexa flour488, Life Technology) added at a final dilution of 1:200. Finally, after a wash step, pellets were fixed with 100 μl 4% formaldehyde/PBS, re-suspended in 100 μl of PBS and analyzed with FACS CANTOII evaluating collected data with FlowJo software. As shown in
OMV Engineering with MBPvIII
pET-MBPvIII Plasmid Construction
In an attempt to deliver EGFRvIII peptide to the lumen of OMVs, the Maltose binding protein (MBP), which is naturally delivered to the periplasm, was used as a carrier (
Expression of MBPvIII Heterologous Protein in E. coli BL21(DE3)/ΔompA
The recombinant plasmid encoding MBP fused to one copy of the EGFRvIII peptide was used to transform E. coli strain BL21(DE3)/ΔompA.
One recombinant strain was obtained: BL21(DE3)/ΔompA(pET-MBPvIII). The strain was grown in LB medium and when culture reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration.
After two additional hours of growth at 37° C., cells were collected and 25 μg of total protein extracts were analyzed by SDS-PAGE. As shown in
Expression of MBPvIII into OMVs
Having demonstrated that MBPvIII was expressed in E. coli BL21(DE3)/ΔompA strain, the presence of the antigen in the OMV fraction was analyzed. Strain BL21(DE3)/ΔompA(pET-MBPvIII) was grown in LB medium and when the culture reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., vesicles were purified from culture supernatants by using ultrafiltration coupled to ultracentrifugation. More specifically, OMVs were collected from culture supernatants by filtration through a 0.22 μm pore size filter (Millipore) and by high-speed centrifugation (200,000×g for 2 hours). Pellets containing OMVs were finally suspended in PBS. The presence of the antigens in OMVs preparations was verified by SDS-PAGE and Western Blot analyses as described in the previous section. Data indicate that MBP carrying one copy of vIII peptide was incorporated into OMVs (
Antibody Titers Elicited in Mice Immunized with MBPvIII in Engineered OMVs
To test whether OMVs purified from BL21(DE3)/ΔompA strain expressing MBPvIII were able to induce EGFRvIII-specific antibody responses, CD1 mice were i.p. immunized three times at two-week interval with 10 μg OMVs in Alum. Blood samples were collected nine days after third dose administration and anti-vIII IgGs were detected by using plates coated in each well with 0.5 μg of synthetic vIII peptide conjugated to Keyhole limpet hemocyanin (vIII-KHL). PBS containing 1% BSA was used as negative control. More specifically, coating was carried out by incubating plates overnight at 4° C. with 100 μl of conjugated peptide (5 μg/ml). Subsequently, wells were washed three times with PBST (0.05% Tween 20 in PBS, pH 7.4), incubated with 100 μl of 1% BSA in PBS for 1 h at room temperature and washed again three times with PBST. Serial dilutions of serum samples in PBST containing 1% BSA were added to the plates, incubated 2 h at 37° C., and washed three times with PBST. Then 100 μl/well of 1:2,000 diluted, alkaline phosphatase-conjugated goat anti-mouse IgGs were added and left for 2 h at 37° C. After triple PBST wash, bound alkaline phosphatase-conjugated antibodies were detected by adding 100 μl/well of 3 mg/ml para-nitrophenyl-phosphate disodium hexahydrate (Sigma-Aldrich) in 1M diethanolamine buffer (pH 9.8). After 10 minute incubation at room temperature, the reaction was stopped with 100 μl 7% EDTA and substrate hydrolysis was analyzed at 405 nm in a microplate spectrophotometer. As shown in
Engineered OMVs Expressing FAT1 Peptide
Two strategies were designed to deliver the FAT1 peptide to the E. coli OMVs. The first strategy envisaged the fusion of FAT1 peptide to the carboxyl terminus of either full length fHbp or fHbpDomA. For this purpose, a synthetic DNA fragment encoding three copies of IQVEATDKDLGPNGHVTYSIVTDTD peptide was ligated to the 3′ end of the full length fHbp gene and to the DNA sequence coding for the Domain A of fhbp (fHbpDomA), thus generating chimeric proteins carrying the FAT1 peptide at their C-terminus.
The second strategy was designed to deliver the FAT1 peptide into the lumen of OMVs. To this aim, the synthetic DNA coding for three copies of the FAT1 peptide was fused to the 3′ end of the MBP gene to create an in frame C-terminal fusion.
Construction of pET21_FL-fHbp-FAT1 and pET21_fHbp-DomA-FAT1 Plasmids
The fusion of three copies of the FAT1 peptide to full length fHbp (FL-fHbp) and to fHbp-DomA was carried out in two main steps. First, the DNA fragments encoding FL-fHbp and fHbp-DomA were cloned into plasmid pET21, thus generating plasmids pET21-fHbp and pET21-fHbp-DomA. Subsequently, the two plasmids were linearized to clone the synthetic DNA encoding three copies of FAT1 peptide (FAT1 Minigene) at the end of fHbp and fHbp-DomA coding sequences. The sequence of FAT1 Minigene was designed taking into consideration BL21 E. coli codon usage (SEQ ID NO:102).
pET21 FL-fHbp and pET21 fHbp-DomA were generated as described in section 5.1.1.
To generate pET21_FL-fHbp-FAT1 and pET21_fHbp-DomA-FAT1 plasmids, pET21-fHbp and pET21-fHbp-DomA were linearized by PCR amplification using the two couples of primers FHBP-F/FHBPFL-R and FHBP-F/FHBPDA-R (Table 2) and the linear fragments were combined with the synthetic DNA coding for FAT1 Minigene (Table 2, Sequence 18). FAT1 Minigene was constructed by assembling six complementary oligonucleotides the sequence of which is reported in Table 2 and the assembled DNA fragment was amplified with primers F-FATFH/R-FATFH F-FATDomA/R-FATFH (Table 2) to make its extremities complementary to linearized pET21 FL-fHbp and pET21_FL-fHbpDomA, respectively. The DNA mixtures were then used to transform E. coli HK100 competent cells and clones carrying pET21_FL-fHbp-FAT1 and pET21_fHbp-DomA-FAT1 plasmids were selected on LB agar plates supplemented with 100 μg/ml Amplicillin. From one clone of each transformation the plasmid was purified and the correctness of the fHbp-FAT1 and fHbpDomA-FAT1 gene fusions was verified by DNA sequencing (SEQ ID NOs:104 and 103).
pET-MBP-FAT1 Plasmid Construction
To express FAT1 peptide in the lumen of OMVs, the Maltose binding protein (MBP) which is naturally delivered to the periplasm was used as a carrier and the FAT1 Minigene was cloned as an in frame fusion to the 3′ end of the MBP gene. For this purpose, plasmid pET-MBPvIII (see Section 5.1.3) has been used as template for a PCR reaction carried out according to the PIPE method (Klock H. E. and Lesley S. A (2009) Methods Mol. Biol. 498, 91-103), using primers pET21-MBPF and pET21-MBPR (see Table 2) to generate a linear fragment missing the vIII coding sequence. Then, the linear fragment was ligated to FAT1 Minigene, which was assembled in vitro using the six synthetic oligonucleotides reported in Table 2 and subsequently amplified with primers MBPFA-F and MBPFA-R. The DNA mixture was used to transform MK-100 competent cells and clones carrying pET-MBP-FAT1 plasmid were selected on LB agar plates supplemented with 100 μg/ml of Ampicillin. The correctness of the MBP and FAT1 Minigene fusion in pET-MBP-FAT1 plasmid purified from one of the Ampicillin resistant clones was verified by DNA sequencing (SEQ ID NO:105).
Expression of fHBP-FAT1 and MBP-FAT1 Peptide in E. coli BL21(DE3)ΔompA
Plasmids pET21 FL-fHbp-FAT1, pET21_fHbp-DomA-FAT1 and pET-MBP-FAT1 were used to transform BL21(DE3) ΔompA strain. Recombinant clones were grown in 200 ml LB medium at 37 C and when the cultures reached OD600=0.6, the expression of the fusion proteins was induced by addition of 1 mM IPTG. After 2 hour growth, the expression of protein fusions was assessed by SDS-PAGE and Western Blot. To this aim, the equivalent in volume of 1 OD600 of each bacterial culture was collected, centrifuged at 13,000×g for 5 minutes and pellets were lysed in 200 μl of BPer Reagent, Lysozime 1 mg/ml, DNAase 10 U/ml and 0.1 mM MgCl2 for 30 minutes. Then the samples were centrifuged at 13.000×g for 20 minutes to separate the supernatants (soluble fraction) from the pellets (insoluble fraction). The soluble fraction was collected (200 μl) and diluted with 100 μl of 4×SDS-PAGE loading buffer while the pellets were re-suspended in 300 μl of 2× loading buffer. 20 μl of each sample were loaded onto an SDS-polyacrylamide gel and the proteins separated by electrophoresis (SDS-PAGE).
Then the gel was stained with Coomassie blue overnight at room temperature. For Western blot analysis 5 μl of the same samples were loaded onto a 4-12% polyacrilamide gel (Invitrogen). After electrophoretic separation proteins were transferred onto nitrocellulose filter by standard methods. The filters were blocked overnight at 4° C. by agitation in blocking solution (10% skimmed milk and 0.1% Tween in PBS), followed by incubation for 90 minutes at 37° C. with anti-FAT1 mAb198.3 at 3 μg/ml in 1% skimmed milk and 0.1% Tween in PBS. After three washing steps in PBS-Tween, the filters were incubated in a 1:5.000 dilution of peroxidase-conjugated anti-mouse immunoglobulin (PerkinElmer) in 1% skimmed milk and 0.1% Tween in PBS for an hour, and after three washing steps, bound conjugated IgGs were detected using the Super Signal West Pico chemo-luminescent substrate (Pierce) and the resulting signal was detected by using the Western lighting plus ECL (PerkinElmer).
As shown in
Expression of fHbp-FAT1, fHbpDomA-FAT1 and MBP-FAT1 into OMVs
Having demonstrated that fHbp-FAT1, fHbpDomA-FAT1 and MBP-FAT1 peptide were expressed in E. coli BL21(DE3)/ΔompA, the presence of the fusions in the OMV fraction was analyzed. 200 ml of the bacterial cultures BL21(DE3)/ΔompA (pET-fHbp-FAT1), BL21(DE3)/ΔompA (pET-fHbpDomA-FAT1) and BL21(DE3)/ΔompA (pET-MBP-FAT1) were grown in LB and when the cultures reached an OD600=0.6 were induced with 1 mM IPTG. Vesicles were purified from the culture supernatants by using ultrafiltration coupled to ultracentrifugation. Briefly, cultures were centrifuged at 4000×g for 15 min, and the supernatants were concentrated using a membrane with a cut off of 100 kDA (Amicon) until a final volume of 30 ml was reached. Then an ultracentrifugation step was performed at 160.000×g for 2 hours. The pellets were re-suspended in 200 μl of PBS. Protein quantization was performed by DC Protein Assay (BioRad). The presence of the specific antigen in OMVs preparation was verified by Western Blot and SDS-PAGE analysis as already described by loading 10 μg of total proteins of each OMV preparation. As shown in
The presence of FAT1 peptide in the three fusion proteins was also confirmed by Western Blot, where the anti FAT1 mAb198.3 recognized the three protein species visible in the OMVs preparations by Coomassie Blue staining (
Analysis of fHBP-FAT1 Expression on the Surface of E. coli BL21(DE3) ΔompA by FACS
In order to confirm the ability of fHbp to deliver the FAT1 peptide on the outer membrane of E. coli ΔompA, FAT1 expression on the bacterial surface was analyzed by FACS.
BL21(DE3)/ΔompA (pET-fHbp-FAT1) and BL21(DE3)/ΔompA (pET-fHbpDomA-FAT1) E. coli strains were grown in 10 ml LB medium at 37 C and when the cultures reached OD600=0.6, the expression of the fusion proteins was induced by addition of 1 mM IPTG. After 2 hour growth bacteria cells corresponding to those contained in 1 ml culture at OD600=1 were collected by centrifugation at 13.000×g for 5 minutes and pellets were re-suspended in 50 ml of PBS containing 1% BSA. 50 μl of cell suspensions were incubated with 50 μl of an appropriate dilution of anti-FAT1 mAb198.3 or, as negative control, with 50 μl of PBS containing 1% BSA. After 1 hour, 100 μl of PBS containing 1% BSA were added and the suspensions were centrifuged at 3.000×g for 10 minutes and supernatants discarded. Pellets were washed with 200 μl of PBS containing 1% BSA and bacteria subsequently incubated for 30 minutes on ice with goat anti-mouse antibodies (Alexa flour488, Life Technology) added at a final dilution of 1:200. Finally, After 2 wash steps, pellets were re-suspended in 200 μl of PBS and analyzed with FACS CANTOII evaluating collected data with FlowJo software. As shown in
Antibody Titers Elicited in Mice Immunized with fHbp-FAT1, fHbpDomA-FAT1 and MBP-FAT1 in Engineered OMVs
To test whether OMVs purified from BL21(DE3)/ΔompA strain expressing fHbpFLFAT1, fHbpDomA-FAT1 and MBP-FAT1 were capable of inducing FAT1 peptide-specific antibody responses, groups of CD1 mice (5 mice per group) were i.p. immunized three times at two-week interval with 20 ug OMVs in Alum. After two weeks from the third immunization, sera were collected and pooled to analyze anti-FAT1 antibody titers by ELISA. ELISA was performed using plates coated with synthetic FAT1 peptide conjugated to the carrier protein Keyhole limpet hemocyanin (KLH). Coating was carried out at room temperature for 14 hours by adding to each well 100 μl of conjugated peptide at a concentration of 5 μg/ml. After three washes with 200 μl/well of PBS supplemented with 0.05% Tween 20 (PBST) the plates were incubated one hour at 37 C with 100 μl/well of PBS containing 1% BSA and subsequently washed three times with PBST. Different dilutions of sera in PBST containing 0.1% BSA were added in duplicate in a final volume of 100 μl/well and plates were stored at 37 C for 2 hours. After three washes in PBST, 100 μl of goat anti-mouse antibodies conjugated to alkaline phosphatase (SouthernBiotech, Cat. 1030-04, 1:2.000 dilution) were added to each well and incubated at 37 C for 1 hour. Finally, after three washes, the phosphatase substrate (4-Nitrophenyl phosphate disodium salt) was added to each well at a concentration of 1 mg/ml (100 ml/well) and after 30 minutes incubation at room temperature in the dark, substrate hydrolysis was measured spectrophotometrically at 405 nm. As shown in
Engineered OMVs Expressing MUC1 Peptide
Two strategies were designed to deliver the MUC1 peptide to E. coli OMVs. The first strategy was designed to deliver MUC1 peptide to the membrane of the vesicles by fusing MUC1 to the C-terminus of fHbpDomA. To achieve this, a synthetic DNA fragment encoding five copies of the MUC1 peptide GVTSAPDTRPAPGSTAPPAH was ligated to the 3′ end of the gene coding for Domain A of fHbp (fHbpDomA).
The second strategy was designed to deliver the MUC1 peptide into the lumen of OMVs. To this aim, the synthetic DNA coding for five copies of the peptide was fused to the 3′ end of the MBP gene to create an in frame C-terminal fusion.
OMV Engineering with fHbpDomA-MUC1 Peptide
The fusion of five copies of the MUC1 peptide to fHbp-DomA was carried out in two main steps. First, the DNA fragment encoding fHbp-DomA were cloned into plasmid pET21 thus generating plasmid pET21 fHbp-DomA as already described. Subsequently, the plasmid was linearized by PCR and ligated to the synthetic DNA fragment encoding five copies of MUC1 peptide (MUC1 Minigene). The sequence coding for MUC1 Minigene was designed taking into consideration the E. coli codon usage. More specifically, pET21 fHbp-DomA was generated as follows. The fHbpDomA gene was amplified by PCR from Neisseria meningitidis serogroup B strain MC58 genome. The primers were designed to amplify the gene with its natural leader sequence for secretion and the lipobox. The polymerase incomplete primer extension (PIPE) cloning method (Klock H. et al., 2009) was used to insert the PCR product into plasmid pET21b which was amplified using primers pet-rev/nohisflag (Table 2). In so doing plasmid pET21 fHbp-DomA was generated (see also above for additional details).
To generate pET21 fHbp-DomA-FAT1 plasmid the PIPE cloning method was used. pET21_fHbp-DomA was linearized by PCR amplification using primers FHBP-F and FHBPDA-R (Table 2) and the linear fragment was combined with the synthetic DNA coding for MUC1 Minigene (Table 2 and Sequence 22). MUC1 Minigene was constructed by assembling ten complementary oligonucleotides the sequence of which is reported in Table 3 and the assembled DNA fragment was amplified with primers RMUCFH and FMUCDomA (Table 3) to make its extremities complementary to the amplified vector. The DNA mixture was then used to transform E. coli HK100 competent cells and clones carrying pET21_fHbp-DomA-MUC1 plasmid were selected on LB agar plates supplemented with 100 μg/ml Amplicillin. From one clone the plasmid was purified and the correctness of the fHbpDomA-MUC1 gene fusion was verified by DNA sequencing (SEQ ID NO:107).
pET-MBP-MUC1 Plasmid Construction
To express MUC1 peptide in the lumen of OMVs, the Maltose binding protein (MBP) which is naturally delivered to the periplasm was used as a carrier and the MUC1 Minigene was cloned as an in frame fusion to the 3′ end of the MBP gene. For this purpose, plasmid pET-MBPvIII (see Section 5.1.3) has been used as template for a PCR reaction carried out according to the PIPE method (Klock H. E. and Lesley S. A (2009) Methods Mol. Biol. 498, 91-103), using primers pET21-MBPF and pET21-MBPR (see Table 2) to generate a linear fragment missing the vIII coding sequence. Then, the linear fragment was ligated to MUC1 Minigene constructed as described above and amplified with primers MBPMU-F and MBPMU-R (Table 3) to make its extremities complementary to the amplified vector. The DNA mixture was then used to transform E. coli HK100 competent cells and clones carrying pET21_MBP-MUC1 plasmid were selected on LB agar plates supplemented with 100 μg/ml Amplicillin. From one clone the plasmid was purified and the correctness of the MBP-MUC1 gene fusion was verified by DNA sequencing (SEQ ID NO:108).
Expression of fHbpDomA-MUC1 and MBP-MUC1 Fusions in E. coli BL21(DE3) ΔompA
Plasmids pET21_fHbp-DomA-MUC1 and pET-MBP-MUC1 were used to transform BL21(DE3) ΔompA strain. Recombinant clones were grown in 200 ml LB medium at 37 C and when the cultures reached OD600=0.6, the expression of the fusion proteins was induced by addition of 1 mM IPTG. After 2 hour growth, the expression of protein fusions was assessed by SDS-PAGE. To this aim, the equivalent in volume of 1 OD600 of bacteria culture was collected, centrifuged at 13.000×g for 5 minutes and pellets were lysed in 200 μl of BPer Reagent, Lysozime 1 mg/ml, DNAase 10 U/ml and 0.1 mM MgCl2 for 30 minutes. Then the samples were centrifuged at 13.000×g for 20 minutes to separate the supernatants (soluble fraction) from the pellets (insoluble fraction). The soluble fractions were collected (200 ul) and diluted with 100 μl of 4×SDS-PAGE loading buffer while the pellets were re-suspended in 300 μl of 2× loading buffer. 20 μl of each sample were loaded onto an SDS-polyacrylamide gel and proteins separated by electrophoresis (SDS-PAGE). Then the gel was stained with Coomassie blue overnight at room temperature. As shown in
Compartmentalization of fHbpDomA-MUC1 and MBP-MUC1 Fusions into OMVs
Having demonstrated that fHbpDomA-MUC1 and MBP-MUC1 were expressed in E. coli BL21(DE3)/ΔompA, the presence of the fusions in the OMV compartment was analyzed. To this aim, 200 ml of the bacterial cultures BL21(DE3)/ΔompA (pET-fHbpDomA-MUC1) and BL21(DE3)/ΔompA (pET-MBP-MUC1) were grown in LB and when the cultures reached an OD600=0.6 were induced with 1 mM IPTG. Vesicles were purified from the culture supernatants by using ultrafiltration coupled to ultracentrifugation. Briefly, cultures were centrifuged at 4.000×g for 15 min, and the supernatants were concentrated using a membrane with a cut off of 100 kDA (Amicon) until a final volume of 30 ml was reached. Then an ultracentrifugation step was performed at 160.000×g for 2 hours. The pellets were re-suspended in 200 μl of PBS. Protein quantization was performed by DC Protein Assay (BioRad). The presence of the fusion proteins in the OMVs preparation was verified by SDS-PAGE analysis as already described by loading 10 μg of total proteins of each OMV preparation. As shown in
Antibody Titers Elicited in Mice Immunized with Engineered OMVs Carrying fHbpDomA-MUC1 and MBP-MUC1
To test whether OMVs purified from BL21(DE3)/ΔompA (pET-fHbpDomA-MUC1) and BL21(DE3)/ΔompA (pET-MBP-MUC1) strains were capable of inducing MUC1-specific antibody responses, CD1 mice (5 mice per group) were i.p. immunized three times at two-week intervals with 20 μg of engineered OMVs in Alum. After two weeks from the third immunization, sera were collected and pooled to analyze anti-MUC1 antibody titers by ELISA. ELISA was performed using plates coated with the synthetic MUC1 peptide GVTSAPDTRPAPGSTAPPAH. Coating was carried out at room temperature for 14 hours by adding to each well 100 μl of a solution of synthetic MUC1 peptide at a concentration of 5 μg/ml. After three washes with 200 μl/well of PBS supplemented with 0.05% Tween 20 (PBST) the plates were incubated one hour at 37 C with 100 μl/well of PBS containing 1% BSA and subsequently washed three times with PBST. Different dilutions of sera in PBST containing 0.1% BSA were added in duplicate in a final volume of 100 μl/well and plates were stored at 37 C for 2 hours. After three washes in PBST, 100 μl of goat anti-mouse antibodies conjugated to alkaline phosphatase (SouthernBiotech, Cat. 1030-04, 1:2.000 dilution) were added to each well and incubated at 37 C for 1 hour. Finally, after three washes, the phosphatase substrate (4-Nitrophenyl phosphate disodium salt) was added to each well at a concentration of 1 mg/ml (100 μl/well) and after 30 minute incubation at room temperature in the dark, substrate hydrolysis was measured spectrophotometrically at 405 nm. As shown in
Engineered OMVs Expressing Aa-fHbp
pET-Aa-fHbp-HIS8 Plasmid Construction
The gene encoding the 828 bp lipoprotein gna1870 like-protein was chemically synthetized (GeneArt™ Gene Synthesis, Thermo Fisher Scientific) using the reported gene sequence (EnsemblBacteria gene ID HMPREF9996_00541) from Aggregatibacter actinomycetemcomitans Y4 with the exception that the natural GTG start codon was replaced by an ATG start codon. The synthetic gene (SEQ ID NO:116) was cloned into pET21b+ fused to a 8-HIS tag at the C-term for subsequent detection of the protein using an anti-HIS polyclonal antibody. For cloning: the synthetic gene was amplified by PCR using the AgfHbp_F and AgfHbp_R primers, and annealed to pET21b+ plasmid backbone amplified with pET HIS-F and pET 2-R primers (Table1).
Expression of the Aa-fHbp Heterologous Protein in E. coli BL21(DE3)/ΔompA Strain
To investigate whether Aa-fHbp was surface-associated when expressed in E. coli, the pET_Aa-fHbp-Hiss recombinant plasmid was used to transform E. coli BL21ΔompA and the expression and localization of Aa-fHbp was analyzed as described previously. Each strain was grown in LB medium and when cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., cells were collected and total protein extracts were analyzed by SDS-PAGE followed by Coomassie staining. No bands were visible in total lysates from E. coli cells carrying empty pET21 plasmid. A band corresponding to the Aa-fHbp protein is detected in the total lysate of the strain carrying the pET_Aa-fHbp-Hiss recombinant plasmid as detected by Coomassie.
Analysis of Aa-fHbp Expression in OMVs
Having demonstrated that the exogenous Aa-fHbp protein was well expressed in E. coli BL21(DE3)/ΔompA strain, we then analysed its cellular localization and compartimentalization to the OMV fraction. The recombinant strain BL21(DE3)/ΔompA(pETAa-fHbp-_HIS8) was grown in LB medium and when the cultures reached an OD600 value of 0.6, IPTG was added at 1 mM final concentration. After two additional hours of growth at 37° C., vesicles were purified from culture supernatants by using ultrafiltration coupled to ultracentrifugation. More specifically, OMVs were collected from culture supernatants by filtration through a 0.22 μm pore size filter (Millipore) and by high-speed centrifugation (200.000×g for 2 hours). Pellets containing OMVs were finally suspended in PBS. The presence of the Aa-fHbp_HIS8 fusion protein in OMVs preparations was verified by Coomassie and Western Blot analysis as described in the previous section (
Analysis of Cellular Localization of Aa-fHbp
The localization of recombinant Aa-fHbp protein was evaluated by flow cytometry. To this aim, recombinant E. coli strains BL21(DE3)/ΔompA(pET-Aa-fHbp-HIS8) and E. coli BL21(DE3)/ΔompA(pET21), as negative control, were grown at 37° C. under agitation. When cultures reached an OD600 value of 0.6, IPTG was added at a final concentration of 1 mM and bacteria were grown for 2 additional hours. Subsequently, bacteria cells corresponding to those contained in 1 ml culture at OD600=1 were collected by centrifugation at 13,000×g for 5 minutes and pellets were re-suspended in 50 ml of PBS containing 1% BSA. 50 μl of cell suspensions were incubated with 50 μl of an appropriate dilution of anti-His-tag antibody with 50 μl of PBS containing 1% BSA as negative control. After 1 hour, 100 μl of PBS containing 1% BSA were added and the suspensions were centrifuged at 3,000×g for 10 minutes and supernatants discarded. Pellets were washed with 200 μl of PBS containing 1% BSA and bacteria subsequently incubated for 30 minutes on ice with goat anti-mouse antibodies (Alexa flour488, Life Technology) added at a final dilution of 1:2,000. Finally, After 2 wash steps, pellets were re-suspended in 200 μl of PBS and analyzed with FACS CANTOII evaluating collected data with FlowJo software. As shown in
Number | Date | Country | Kind |
---|---|---|---|
15168024.6 | May 2015 | EP | regional |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2016/061031 | 5/17/2016 | WO | 00 |