Patent Application
20170145061
The targeted intestinal delivery of therapeutic proteins by genetically engineered live bacteria is an emerging field of research with broad potential in the treatment of human diseases. The conventional delivery of protein drugs like anti-inflammatory cytokines to the intestine is challenging as they are unstable when administered orally, or require high doses with severe side effects if administered systemically.
Engineered microbes overcome drawbacks of conventional protein delivery strategies by releasing protein drugs in close proximity of their site of action. Provided herein is a versatile intestinal protein delivery system deploying engineered human gut commensals of the Bacteroides species to secrete heterologous, therapeutic proteins via outer membrane vesicles (OMVs). Delivery via OMVs prevents cargo from dilution and proteolytic degradation. The stable and abundant intestinal colonization by bacteria of the beneficial genus Bacteroides particularly qualifies them as long-term therapeutics.
As shown in the working examples provided herein, heterologous proteins were secreted via Bacteroides OMVs by genetic fusion to small peptide tags derived from proteins naturally occurring in B. thetaiotaomicron and B. ovatus OMVs. First, three OMV proteins (BT1491, BT3238, BACOVA_04502) that target the reporter protein NanoLuc to B. thetaiotaomicron OMVs when fused to its N-terminus were identified. A bioinformatics analysis predicted that the proteins were lipoproteins containing N-terminal signal peptides with signal peptidase II cleavage sites. Second, truncations of the OMV proteins were generated to determine the minimal and optimal length for secretion. Peptide tags of 25-35 amino acids efficiently translocated functional NanoLuc to OMVs in B. thetaiotaomicron, B. fragilis, and B. vulgatus. Further, the anti-inflammatory molecule IL-10 was successfully secreted into the supernatants of B. thetaiotaomicron and B. vulgatus cultures at concentrations of up to 50 ng/mL. IL-10 was identified in purified OMVs at concentrations of 0.3 ng/mL. Additionally, stable intestinal colonization of mice was shown in an experimental model of inflammatory bowel disease with B. thetaiotaomicron, B. fragilis, and B. vulgatus, which had no detrimental impact on gut inflammation.
IL-10 has previously been shown to ameliorate inflammation in experimental animal models of inflammatory bowel disease. Thus, the production of this mammalian immunomodulatory protein and its secretion via outer membrane vesicles of the Bacteroides spp. is a candidate for the development of novel long-term therapies for inflammatory gut diseases. The treatment generally requires low doses and primarily avoids systemic side effects due to its targeted local delivery. Further, the treatment is applicable to a wide range of therapeutic protein drugs for treatment of various intestinal disorders.
These and other embodiments of the present disclosure will be described in greater detail herein.
Microorganisms are essential for human health. The human body accommodates at least as many microbes—collectively referred to as the microbiota—as human cells (1-3). The intestinal mucosa harbors the largest population of microbial cells composed of an impressive variety of 500 to 1,000 different species adding up to an aggregate biomass of about 1.5 kg. Concentrations can exist up to 1011 organisms per milliliter proximal colonic contents (2, 4). Although most members belong to the domain Bacteria, there are also viruses as well as Archaea and Fungi (5-7). Approximately 99.9% of all cultivatable bacteria are obligate anaerobes (8) with Bacteroides, Clostridium, Eubacterium, and Bifidobacterium (2, 9) as common genera. During the first years of live, every human develops a specific composition of microbes, which is continuously shaped by factors like diet, age, and antibiotics (10, 11).
The complex relationship between the gut flora and host can be commensal (i.e. benefitting from the host without affecting it), mutualistic, meaning that both, host and microbiota, benefit from each other, or pathogenic by harming the host (9, 12). Intestinal bacteria contribute to the hosts' health in many ways and are essential for its well-being. Indeed, the microbiota is sometimes referred to as the ‘forgotten organ’ to emphasize its crucial role in human health and disease (13). It endows us with functional features and metabolic pathways we have not evolved ourselves and are unable to perform. These include the fermentation of complex dietary carbohydrates supplying us with 10-15% of our daily calories (14, 15), the biotransformation of conjugated bile acids (16), and the synthesis of certain vitamins, particularly those of the B group and vitamin K, that are essential for human health (17). Moreover, the gut microbiota prevents intestinal colonization of potential pathogens and their translocation through the mucosa (18-20), and educates our immune system to induce tolerance to certain microbial epitopes, which contributes to reduce allergic responses to food and other environmental antigens (21, 22).
Recent efforts focused on the development of microbiota-derived therapeutics including genetically engineered probiotics that augment their intrinsic benefits by expressing recombinant therapeutic molecules (23, 24). As shown in the working examples, abundant species of the mutualistic Bacteroides genus were engineered to secrete therapeutic proteins associated with outer membrane vesicles (OMVs).
Bacteroides are the predominant genus of the healthy human microbiota residing in the distal small intestine and colon (8, 25). They are a pleomorphic group of Gram-negative, obligate anaerobic, rod-shaped, non-spore-forming bacteria, which are essential for the mutualism between gut microbiota and human host. The Bacteroides spp. most abundant in the colon are B. vulgatus, B. thetaiotaomicron, and B. distasonis (ca. 1010 per g dry weight of feces) followed by B. fragilis, B. ovatus, and B. uniformis (ca. 109 per g dry weight of feces) (26). B. fragilis and B. thetaiotaomicron are the most intensively studied species.
Accounting for ˜30% of the total bacterial population in the adult human gut (26), Bacteroides spp. live in an inextricable partnership with their host. They sense and adapt to environmental changes and stressors like altered nutrient availability or low oxygen concentrations, enabling them to thrive in the extremely harsh conditions in the gut (27). For instance, Bacteroides are nutritionally versatile in that they are able to use a wide range of carbon sources, including dietary fibers that are indigestible by the host (26). Lee et al. discovered that commensal colonization factors (ccf) important for glycan utilization are required for the persistent and resilient colonization of the mammalian gut by Bacteroides (28). Additionally, Bacteroides have multiple efflux pump systems to remove toxic substances (27, 29). As a consequence, they are significantly more stable members of the microbiota than the population average (30).
Protection from Disease.
Bacteroides spp. live in a mutually beneficial relationship with their host as long as they are retained within the intestine (4). However, if they escape the gut, usually as a consequence of intestinal surgery or ruptures, Bacteroides cause serious infections and abscesses at multiple body sites including the abdomen, liver, lungs, and brain, as well as severe bacteremia in rare cases (31). Toxigenic variants of B. fragilis are the most commonly encountered anaerobic pathogen and the most virulent Bacteroides species, even though B. fragilis accounts for only 0.5% of the human colonic microbiota (32).
Retained within the gut, Bacteroides substantially contribute to restrict pathogens from colonizing the intestine and invading host tissues by occupying ecological niches, nutrients, and other resources. In addition, B. fragilis makes a major contribution towards the development of the intestinal immune system (18, 33) to further limit the access and proliferation of potential pathogens into the gut. It prevents experimental intestinal inflammatory diseases in mice through a mechanism involving its capsular polysaccharide A (PSA) (34, 35). Further, B. thetaiotaomicron stimulates Paneth cells to produce antimicrobial peptides such as defensins and lectins (36). The secretion of angiogenin-4 in mouse Paneth cells, stimulated by B. thetaiotaomicron, has also been found to have bactericidal activity against certain intestinal Gram-positive pathogens like Listeria monocytogenes (37). Intriguingly, Bacteroides species were recently found to induce a distinct population of regulatory T cells (Tregs) to confine immune-inflammatory responses (38). Monocolonization of germ-free mice with various Bacteroides species revealed increased frequency of these anti-inflammatory, beneficial RORγ+Tregs, especially for B. ovatus, B. vulgatus, and B. thetaiotaomicron.
Nutrient Provision.
Bacteroides species have the remarkable ability to utilize a tremendous variability of nutrients. They ferment a large variety of indigestible dietary plant polysaccharides like amylose and amylopectin as well as host- or microbiota-derived polysaccharides that are not processed by human enzymes (14). They are responsible for the major fraction of polysaccharide digestion in the colon (26, 39), a task that is almost exclusively executed by members of the genus Bacteroides (40, 41). In particular, B. thetaiotaomicron plays an exceptional role in polysaccharide breakdown. The majority of its genome is devoted to an extensive polysaccharide utilization system that comprises 20 sugar-specific transporters, 163 homologs of polysaccharide-binding proteins (SusC and SusD homologs), and 172 glycosylhydrolases (e.g., glucosidases, galactosidases, mannosidases, amylases) (27). Notably, the number of glycosylhydrolases in the proteome of B. thetaiotaomicron is higher than in any other sequenced prokaryote (27). In addition, Bacteroides harbor multiple nutrient sensing mechanisms, including σ-factors and two-component regulatory systems that coordinate gene expression according to nutrient availability in the vicinity (27, 42). Consequently, Bacteroides are capable to adapt to changes and stresses in their environment.
By providing additional nutrients, Bacteroides spp. benefit the host as well as the whole bacterial community (43). Other organisms in the intestine that do not harbor such an array of sugar utilization enzymes can harvest sugars generated by Bacteroides. For example, B. ovatus ferments the fructose polymer inulin to cross-feed other gut species like B. vulgatus, which provide benefits for B. ovatus in return (44).
For these beneficial interactions, protein secretion plays a pivotal role to e.g. disseminate enzymes that make nutrients accessible for bystanders. Traditionally, six major classes of protein secretion machineries are known that translocate soluble molecules across the inner and outer membranes of Gram-negative bacteria (45). However, recent attention was drawn to a seventh, independent secretion system for the transfer of a diverse group of molecules via blebs formed by the outer membrane (46-48). These so called outer membrane vesicles (OMVs) are ubiquitously present in all Gram-negative bacteria (48-50) and possess many important advantages over other secretion systems, as elaborated on in the following two sections.
OMVs are spherical, bilayered proteoliposomes with a diameter ranging from 20 to 250 nm, consisting of the bacterial outer membrane and periplasmic content (
OMV Biogenesis.
OMVs originate from the cell envelope of Gram-negative bacteria, which is composed of two membranes, the outer membrane (OM) and the inner membrane (IM). The membranes are linked by a thin, mesh-like peptidoglycan (PG) network in the periplasmic space between the two and stitched together by protein crosslinks reaching from the IM through the PG network to the OM (
The biogenesis of OMVs is an elaborate, energy-consuming mechanism that takes place during active growth and is not a by-product of cell lysis or a product of simple membrane shearing or blebbing (46). Vesiculation levels are induced by stress, for instance temperature increase (55, 56), amino acid deprivation (57), and antibiotics (58). However, the exact pathway of OMV formation remains unknown. Instead of a universal mechanism, the current literature proposes several mechanistic scenarios and key features that are likely to be involved (reviewed in detail in (59)).
In the first scenario, the OM bulges out in areas where it is dissociated from the underlying PG since protein crosslinks between the two are locally absent or decimated. Evidence for this model comes from hypervesiculating mutants of Escherichia coli that exhibit lower rates of OM-PG crosslinks than wild type E. coli (60). Further, proteins responsible for OM-PG crosslinks are sparse in OMVs; e. g. Braun's lipoprotein Lpp that covalently bridges the OM with the PG layer is excluded from E. coli OMVs (61). A second model of OMV biogenesis assumes vesiculation being a general stress response of bacteria to misfolded proteins or aberrant envelope components like overexpressed periplasmic proteins or excess peptidoglycan fragments (56, 62, 63). This material accumulates in so called nanoterritories at the inner surface of the OM, exerts turgor pressure on the OM and—after Lpp-PG crosslinks are locally removed—causes the OM to bulge outwards and bud off. Consequently, these undesired components are effectively removed from the cell and were found to be enriched in OMVs (56, 62, 64). The third theory assumes that altered biophysical characteristics of the OM change the membrane fluidity and flexibility, resulting in curvature and budding off (46, 48). For instance, charge-to-charge repulsion in microdomains of highly charged B-band LPS in Pseudomonas aeruginosa forces the membrane to curve outward and is enriched in OMVs (58, 65). Importantly, the suggested mechanisms are not mutually exclusive but rather may collectively contribute to the formation of OMVs (59, 66). In all cases, OMV biogenesis does not compromise envelope integrity (67).
Cargo Selection.
The composition of OMVs has been thoroughly analyzed by several groups in a multitude of bacterial strains. Mass spectrometry-based high-throughput profiling of OMVs has provided massive amounts of data about their protein content, which also elucidates their biogenesis and function (reviewed in (68)). Derived from the cell envelope, OMVs contain a similar outer membrane consisting of LPS and phospholipids as well as lipoproteins and membrane proteins like porins, ion channels, adhesins, and enzymes. Apart from periplasmic proteins and peptidoglycans, OMVs also carry specific cargos in their lumina, e. g. proteases, nucleic acids, and toxins such as the cholera toxin in Vibrio cholerae (69) and Cytolysin A in enterotoxic E. coli (70).
Although most proteins detected in OMVs were also found in the outer membrane or periplasm, certain proteins were specifically enriched or excluded in OMVs (61, 71-73). For instance, a proteomic study of the B. fragilis outer membrane and OMVs identified 40 proteins unique to OMVs, mostly glycosidases and proteases (74). This clustering of functionally related proteins supports the presence of a specific mechanism for OMV biogenesis, as discussed above, and additionally suggests a specialized machinery to sort certain proteins into OMVs. In line with this finding, also cytoplasmic and inner membrane proteins were identified in OMVs in low amounts, even after stringent purification steps (73). As these proteins are normally not present in the cell envelope, they must be actively sorted and exported into OMVs. Knowledge about how OMV cargo is selected would facilitate engineering of Gram-negative bacteria to specifically package heterologous proteins into their OMVs. However, the exact mechanism by which proteins are sorted into OMVs is currently not known, but several hypotheses exist (75).
In order to be secreted in OMVs, the cargo must be exported from the cytosol to the periplasm or OM first. Proteins in these two compartments are synthesized in the cytosol as precursors with N-terminal signal peptides. Typical amino acid motifs in the signal peptides target the proteins for translocation across the IM either by the Sec translocon (76-78) or the twin-arginine translocation (Tat) pathway (79, 80). After cleavage of the signal peptide at the periplasmic face of the IM, proteins destined for the OM cross the periplasm in complex with guiding chaperones (81, 82). To date, however, no signal or machinery has been identified to target incorporation of specific proteins into OMVs. One possibility is that OM components like OM proteins or OM lipoproteins prone to budding by increasing membrane curvature or fluidity might be inherently clustered in certain areas of the cell envelope during its biogenesis. Prior to OMV budding, specific proteins might directly or indirectly interact with the periplasmic face of these OM components and become enriched in OMVs (48). However, this model does not explain how cytoplasmic proteins that do not possess signal peptides are transported across the IM (83, 84) to interact with OM proteins and translocate into OMVs.
Cargo Delivery to Host Cells.
After disseminating, OMVs deliver their cargo not only to other bacterial cells (85, 86) but also eukaryotic host cells, for which three mechanisms were proposed (reviewed in (87)) (
As vesiculation is a ubiquitous mechanism in Gram-negative bacteria, it is obvious that OMVs play an integral role in cell physiology and the pathogenesis of infections (97). Depending on the species of origin and their environment, OMVs have diverse functions. In general, they act as long distance delivery vehicles of proteins, lipids, and genetic material from bacteria to bacteria or host cells while protecting their cargo form dilution and proteolytic degradation. OMVs induce changes in the bacterial environment and benefit the survival of the parent bacteria, as illustrated in the following section.
Initially, OMVs were thought to primarily mediate pathogenic processes by supporting the shuttle of virulence factors such as proteases, toxins, or pro-inflammatory molecules like flagellin, LPS, and peptidoglycan, to host cells and competing bacteria (58, 89). However, recent attention has been drawn towards non-pathogenic, commensal bacteria utilizing OMVs to mediate beneficial effects on the host (97-99).
OMVs Benefit the Bacterial Community.
A major role of OMVs in bacterial physiology lies in the response to environmental stress. Here, OMVs are an effective mechanism to quickly relieve the cell of damaging agents such as toxic or misfolded material, antibiotics, and bacteriophages and are particularly crucial for aggregates that are too big for OM pores. For instance, heat stress in E. coli results in the accumulation of misfolded proteins, which are packed into OMVs and thereby removed (56). OMVs may even be essential in the survival of stress situations. McBroom and Kuehn (56) found that when two vesiculation mutant E. coli strains were challenged with lethal envelope stressors, e.g. ethanol or OM-damaging antimicrobial peptides, hypovesiculating mutants succumbed, whereas hypervesiculating mutants survived better than wild type E. coli. Further, bacterial cells exposed to antibiotics produce OMVs to sequester (100) or degrade (101) the antibiotics outside of the cells. Additionally, increased OMV production was shown to protect bacteria from lytic bacteriophages by acting as ‘decoy’ targets for the phages (100, 102). Hence, Manning & Kuehn called OMVs an ‘innate bacterial defense’ (100).
Moreover, OMVs are essential in nutrient acquisition. They can carry and disseminate enzymes that degrade complex macromolecules to make nutrients accessible for bacterial and host cells. For instance, proteomic data revealed that B. fragilis and B. thetaiotaomicron preferentially target acidic hydrolytic enzymes, primarily proteases and glycosidases, to OMVs to help secure nutrients (74). Besides, OMVs can contain iron and zinc acquisition systems to collect these scarce metal ions from the environment and enrich them for the subsequent consumption by bacteria (73).
Importantly, OMVs act as a common resource that benefits whole bacterial populations: They not only provide nutrients for the OMV producing bacterium but also for bystanders. Also, OMVs were found to protect both producing and bystander bacteria from antibiotic stress by sequestration of antibiotics. Consequently, OMVs have an indispensable role for the survival and fitness of whole bacterial communities present in the gut microbiota (103).
OMVs Contribute to Host Health.
Apart from benefitting bacterial populations, OMVs also directly improve the human gastrointestinal physiology. For instance, Stentz et al. found that BtMinpp, a homolog of the mammalian Inositol hexakisphosphate (InsP6) phosphatase (MINPP), is secreted in OMVs by B. thetaiotaomicron and delivered to intestinal epithelial cells (99). BtMinpp-packed OMVs thereby not only contribute to the essential InsP6 homeostasis and free up the vital nutrients phosphate and inositol but also interact with the inositol polyphosphate signaling pathway in host cells. In addition, OMVs released by B. fragilis deliver immunomodulatory molecules to host cells (98). The capsular polysaccharide A (PSA) is selectively associated with B. fragilis OMVs, which are then internalized by dendritic cells (DCs) to program them for an enhanced production of Tregs that secrete the anti-inflammatory cytokine IL-10. This leads to mucosal tolerance and protects mice from experimental colitis. These examples show two mechanisms of a beneficial inter-kingdom communication between microbiota and host mediated by OMVs.
OMVs act as a secretion and delivery system to disseminate bacterial products to distant locations. As compared to whole cells, OMVs are smaller and more mobile, which enables them to reach remote sites and sites inaccessible to bacteria without consuming energy to move themselves (46). In contrast to traditional soluble secretion machineries, OMVs exhibit several advantages. Recently, Hickey et al. found that B. thetaiotaomicron OMVs can access host immune cells in the murine intestinal mucosa (104). Sulfatases contained in these OMVs enabled them to break through the sulfate-containing, net-like mucus layer, cross the epithelial barrier, and deliver their cargo after being engulfed by macrophages.
Unlike soluble secretory pathways, the secretion via OMVs protects cargo from degradation and dilution. Luminal OMV proteins resist proteolytic degradation, e.g. in the GI tract (105), allowing even less stable proteins to reach their destination. Further, OMVs are robust as they show no signs of spontaneous lysis and increased thermal stability (106). Sequestration in the enclosed OMV prevents cargo dilution and enables its delivery and release at high local concentrations over long distances. Due to the co-transport of multiple molecules, for instance the various enzymes required for the degradation of a complex molecule, they reach distant targets simultaneously, which increases their efficacy. Another advantage of OMV-based secretion is to efficiently shed insoluble hydrophobic molecules like lipids, membrane proteins, and certain signaling molecules.
Inflammatory Bowel Diseases and IL-10.
Inflammatory bowel disease (IBD) is a chronic, relapsing, intestinal disorder, frequently manifesting as Crohn's disease (CD) or ulcerative colitis (UC). The diseases are characterized by chronic inflammation, severe diarrhea with rectal bleeding, and malabsorption as a consequence of a dysregulated intestinal immune homeostasis (107). Although the causes are not fully elucidated, disproportionate mucosal immune responses against resident bacteria are thought to be crucially involved and might be fostered by both genetic and environmental factors (108). Innate and adaptive immune cells accumulate in the intestinal mucosa leading to increased levels of pro-inflammatory cytokines like IFN-γ, interleukin (IL)-17, and IL-22 produced by the T helper (Th)1 response in CD, and tumor necrosis factor (TNF)-α, IL-1β, and IL-6 mediated by the Th2-like response in UC (109). The chronic intestinal inflammation results in continuous epithelial damage and destruction of the epithelial barrier, which allows more intestinal microbes to invade and evoke further immune responses (110). CD and UC differ in the localization of the inflammation. While CD can affect any part of the GI tract, it is predominantly found in the terminal ileum with transmural inflammation across the entire intestinal wall. In contrast, UC is restricted to mucosal inflammation of the colon and rectum (111).
Genome-wide association studies have identified IL10 as a susceptibility locus for the development of IBD (112-114). Polymorphisms in the IL10 promoter that reduce serum levels of the anti-inflammatory cytokine IL-10 have been linked to certain forms of IBD (115, 116). Thus, IL-10 supplementation has been regarded as an alternative IBD treatment to the current available options like surgery, aminosalicylates, immunosuppressants, and biologics, which often have low response rates (117). The dose-limiting side effects for these long-term drug treatments range from nausea and headache to severe, long-lasting complications like osteoporosis or bone marrow toxicity resulting in leucopenia and sepsis (118, 119). Therefore, improved treatment options are urgently needed.
The multifunctional anti-inflammatory cytokine IL-10 counteracts excessive inflammatory immune responses and prevents tremendous intestinal damage. It is produced by many cell types of the innate (e.g. dendritic cells (DCs), macrophages, and natural killer (NK) cells) and adaptive immune system (e.g. Th1, Th2, Tregs, CD8+ T cells, and B cells) (120). IL-10 binds its receptor as a homodimer, which is present on most hematopoietic cells and induces a downstream signaling cascade leading to signal transducer and activator of transcription 3 (STAT3) mediated gene expression (121, 122). As a consequence, IL-10 exerts a wide range of immunomodulatory effects. In macrophages, IL-10 inhibits antigen presentation by MHC class II, co-stimulatory molecule expression, and pro-inflammatory cytokine (e.g. TNF-α, IFN-γ) and chemokine production (reviewed in (120)). Further, Th1, Th2, and NK cell responses are inhibited and the differentiation of IL-10 producing Tregs enhanced.
IL-10 supplementation alleviated symptoms in IBD animal models. Intestinal inflammation in several models of experimental colitis were substantially improved by IL-10 treatment in various animals including mice, rats, and rabbits (123-125). However, clinical studies indicate no significantly reduced remission rates or clinical improvements of systemic IL-10 therapy compared to placebo (118, 126). A hypothesis explaining this setback is that local IL-10 concentrations in the intestine were too low to elicit an ameliorating effect. Unfortunately, concentrations of systemically administered IL-10 are limited due to side effects like anemia and headache. To increase the mucosal bioavailability of IL-10 and circumvent systemic side effects, Steidler et al. engineered Lactococcus lactis to secrete IL-10 after intragastric administration (127). Dextran sulfate sodium (DSS) induced colitis was reduced by 50% in mice as determined by histological scores. Further, a small phase I human trial revealed that application of IL-10 producing L. lactis is safe and well-tolerated in humans while systemic side effects are avoided (128). These results show that IL-10 should have therapeutic potential in the treatment of intestinal inflammation if delivered at high concentrations to the diseased mucosa.
Genetically modified bacteria residing in the intestine and secreting therapeutic proteins in situ have crucial advantages over currently available systemic treatments. First, the drug is exclusively produced and released at the desired site. Hence, vital organs are protected from elevated drug doses that lead to toxicity and side effects. Further, it obviates the need to protect unstable drugs like therapeutic proteins from degradation on a long route to the effector site. Second, due to the targeted action of in vivo production and delivery, the required dosage for a comparable therapeutic effect is reduced by several orders of magnitudes (127, 129, 130), which additionally prevents side effects. Third, engineered bacteria can be orally administered and reach the intestine where they release the therapeutic proteins. This increases patient compliance compared to systemic treatment by invasive and inconvenient intravenous or subcutaneous injections. Finally, the treatment costs are drastically reduced as the need for expensive drug purification and formulation is eliminated.
Commensal Bacteroides colonizes the intestine naturally in high abundance providing a high capacity and continuity of drug production. As has been elaborated on in section 2.1., Bacteroides additionally have beneficial features for the host. The group of Simon Carding engineered B. ovatus to produce promising candidates for the long-term treatment of inflammatory gut diseases like transforming growth factor(TGF)-β, keratinocyte growth factor(KGF)-2, and IL-2 (129, 130, 134). The growth factors/cytokines were secreted into the extracellular milieu by B. ovatus and needed to diffuse to their target site in the mucosa. Provided herein is a model where therapeutic proteins are secreted in association with outer membrane vesicles produced by Bacteroides species and hence reach their target cells in a more directed and concentrated form.
Deploying OMVs as drug delivery vehicles for therapeutic proteins produced by engineered bacteria will have several advantages compared to soluble secreted proteins. OMVs can migrate to the inflammatory site in the mucosa and deliver anti-inflammatory molecules directly to the target side. In contrast to whole bacteria migrating to the inflammatory site, OMVs are less immunogenic and therefore less prone to exacerbate the inflammation. Bacteroides OMVs are particularly suited as it was shown that B. fragilis LPS is 10 to 1,000 times less toxic than that of E. coli (31). Another benefit of OMV-based delivery is that multiple different therapeutic proteins can be targeted to the same OMVs and simultaneously be delivered to the same target cell. For instance, IBD could be treated as a combination therapy of various anti-inflammatory proteins that complement or reinforce each other, such as IL-10, IL-22, IL-4, and TGF-β (171).
Since OMV tag-mediated secretion is compatible with at least three different Bacteroides spp., therapeutic protein concentrations could be increased if necessary by employing several species.
As alternative to administration of engineered bacteria that secrete therapeutic proteins via OMVs, purified OMVs packed with protein drugs could be administered without the associated bacteria. Remarkably, Shen et al. showed that dosing of purified B. fragilis OMVs alone were sufficient to protect mice from chemically-induced colitis (172). By this approach, possible environmental and safety concerns raised by applying genetically modified organisms in patients will be overcome while still exploiting the benefits of targeted OMV drug delivery.
The OMV-based secretion system in Bacteroides spp. should substantially improve the quality of life of IBD patients. After a single administration of the engineered bacteria, they will reside in the patient's intestine and secrete therapeutic proteins. By coupling the production of therapeutic proteins to a sensing system for IBD flare-ups, the therapy is adjusted to the disease state in the patient. For instance, the infiltration of neutrophils during intestinal inflammation leads to a release of reactive oxygen species and increased oxygen levels in the gut (173). Additionally, nitric oxide (NO) concentrations were found to correlate with disease activity in ulcerative colitis with 100 times higher levels in the patients than control levels (174, 175). Therefore, oxidative stress pathways or NO-induced gene expression systems could be deployed for that purpose. Eventually, IBD patients will be colonized with engineered Bacteroides spp. that sense and react to disease flare-ups with the secretion of anti-inflammatory proteins directly at the inflamed site, reducing patient exposure to the drug to a minimum. Ideally, the inflammation will be alleviated before the patient suffers from the symptoms. As Bacteroides colonizes the intestine stably and in high abundance, it is better suited for these long-term application than L. lactis or E. coli. Additionally, the anaerobic nature of Bacteroides spp. provides an inherent biosafety feature.
Considering the 2.2 million Europeans and 1.4 million Americans suffering from IBD (176) and the dramatic increase of IBD occurrence in western countries as well as newly industrialized countries in Asia, South America, and the Middle East (177), there is an urgent need for novel effective treatments. Current therapy options like aminosalicylates, immunosuppressants, and biologics either show severe dose-limiting side effects including long-lasting complications like osteoporosis or bone marrow toxicity (118, 119), or have low response rates. Although the highly cost-intensive therapy with anti-TNF antibodies is currently seen as the gold-standard for IBD treatment, ca. 30% of patients to not respond and another 40% lose response over time (117).
Engineered commensal bacteria of the Bacteroides spp. secreting immunomodulatory molecules, such as IL-10, via OMVs, as described herein, should provide a specific and controlled long-term immune therapy for intestinal disorders such as IBD.
Some aspects of the present disclosure relate to engineered Bacteroides. Examples of species of Bacteroides that may be used in accordance with the present disclosure include, without limitation, B. acidifaciens, B. caccae, B. distasonis, B. gracilis, B. fragilis, B. dorei, B. oris, B. ovatus, B. putredinis, B. pyogenes, B. stercoris, B. suis, B. tectus, B. thetaiotaomicron, B. vulgatus, B. eggerthii, B. merdae, B. stercori, and B. uniformis.
In some embodiments, an engineered Bacteroides comprises a nucleic acid encoding a fusion protein comprising a Bacteroides membrane-associated protein linked to a heterologous protein (e.g., a therapeutic protein). In some embodiments, a promoter is an inducible promoter. In some embodiments, an engineered Bacteroides comprises a fusion protein comprising a Bacteroides membrane-associated protein linked to a heterologous protein (e.g., a therapeutic protein). A “fusion protein” is a hybrid polypeptide that comprises protein domains (e.g., at least one peptide) from at least two different proteins (e.g., obtained from two different types of proteins). In some embodiments, the Bacteroides membrane-associated protein may be fused to the N-terminus of the heterologous protein. In other embodiments, the Bacteroides membrane-associated protein may be fused to the C-terminus of the heterologous protein.
A “membrane-associated protein” is a protein, truncated protein or peptide that interacts with, or is part of, a cell membrane (e.g., a lipid bilayer). Non-limiting examples of membrane-associated proteins include integral membrane proteins (e.g., that are permanently anchored or part of the membrane) and peripheral membrane proteins (e.g., that are only temporarily attached to the lipid bilayer or to other integral proteins). In some embodiments, a Bacteroides membrane-associated protein is a Bacteroides lipoprotein. Bacteroides lipoproteins include membrane proteins that play key roles in Bacteroides physiology and pathogenesis, e.g., in host cell adhesion, modulation of inflammatory processes, and translocation of proteins, e.g., virulence factors, into host cells. A lipoprotein may be or comprise a signal peptidase I (SPI) or a signal peptidase II (SPII) lipoprotein. In some embodiments, a lipoprotein does not include a signal peptidase.
In some embodiments, a Bacteroides membrane-associated protein is selected from the group consisting of: BT1491 proteins, BT3238 proteins, and BACOVA_04502 proteins. In some embodiments, a Bacteroides membrane-associated protein is a truncated variant of a BT1491 protein, a BT3238 protein, or a BACOVA_04502 protein. For example, a Bacteroides membrane-associated protein may be a N-terminal peptide of a BT1491 protein, a BT3238 protein, or a BACOVA_04502 protein. In some embodiments, a N-terminal peptide of a BT1491 protein, a BT3238 protein, or a BACOVA_04502 protein has a length of 18-100 amino acids. For example, a N-terminal peptide of a BT1491 protein, a BT3238 protein, or a BACOVA_04502 protein may have a length of 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, or 100 amino acids. In some embodiments, a N-terminal peptide of a BT1491 protein, a BT3238 protein, or a BACOVA_04502 protein has a length of 25-55 amino acids. In some embodiments, a N-terminal peptide of a BT1491 protein, a BT3238 protein, or a BACOVA_04502 protein has a length of 18-50 amino acids. In some embodiments, a N-terminal peptide of a BT1491 protein has a length of 25 amino acids. In some embodiments, a N-terminal peptide of a BT3238 protein has a length of 35 amino acids. In some embodiments, a N-terminal peptide of a BACOVA_04502 protein has a length of 28 amino acids.
In some embodiments, a Bacteroides lipoprotein of the present disclosure comprises a N-terminal signal peptide. A “signal peptide” is a peptide located within the N-terminal region (e.g., 15-60 amino acids) of a protein. A signal peptide, in some instances, is needed for translocation across a cell membrane and thus universally controls in eukaryotes and prokaryotes entry of most proteins into the secretory pathway. A signal peptides generally includes three regions: an N-terminal region of differing length, which usually comprises positively charged amino acids; a hydrophobic region; and a short carboxy-terminal peptide region. Bacteroides lipoproteins of the present disclosure, in some embodiments, comprises a N-terminal signal peptide that is rich in aspartic acids (D). For example, the N-terminal signal peptide may comprise at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, or more aspartic acids. In some embodiments, a N-terminal signal peptide comprises 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, or more aspartic acids.
In some embodiments, a N-terminal signal peptide comprises a cleavage site for a signal peptidase, e.g., a signal peptidase I (SPI) or a signal peptidase II (SPIT). In some embodiments, a signal peptidase cleavage site is within the first 15-50 amino acids of the Bacteroides lipoprotein. For example, a signal peptidase cleavage site may be within the first 15, 20, 25, 30, 35, 40, 45, or 50 amino acids of a Bacteroides lipoprotein.
A fusion protein of the disclosure comprises Bacteroides membrane-associate protein linked to a heterologous protein. A “heterologous protein,” generally, is any non-Bacteroides protein. Non-limiting examples of heterologous protein include recombinant therapeutic proteins, diagnostic proteins and prophylactic proteins. In some embodiments, however, a Bacteroides membrane-associate protein may be linked to a Bacteroides protein. Specific examples of heterologous proteins include, without limitation, biomarkers, transcriptional regulators, epigenetic modifiers, nucleic acid editing enzymes, nucleases, proteases, or any other enzymes of interest.
In some embodiments, a heterologous protein is a therapeutic protein. Therapeutic proteins that may be used in accordance with the present disclosure include, without limitation, antibodies, cytokines, and growth factors. In some embodiments, the therapeutic protein is a growth factor, e.g., a transforming growth factor beta 1 (TGF-β1) or a keratinocyte growth factor (KGF). In some embodiments, a therapeutic protein is a cytokine. Cytokines include small cell-signaling protein molecules secreted by cells. Non-limiting examples of cytokines that may be used in accordance with the present disclosure include Acrp30, AgRP, amphiregulin, angiopoietin-1, AXL, BDNF, bFGF, BLC, BMP-4, BMP-6, b-NGF, BTC, CCL28, Ck beta 8-1, CNTF, CTACK CTAC, Skinkine, Dtk, EGF, EGF-R, ENA-78, eotaxin, eotaxin-2, MPIF-2, eotaxin-3, MIP-4-alpha, Fas, Fas/TNFRSF6/Apo-1/CD95, FGF-4, FGF-6, FGF-7, FGF-9, Flt-3 Ligand fms-like tyrosine kinase-3, FKN or FK, GCP-2, GCSF, GDNF Glial, GITR, GITR, GM-CSF, GRO, GRO-α, HCC-4, hematopoietic growth factor, hepatocyte growth factor, 1-309, ICAM-1, ICAM-3, IFN-γ, IGFBP-1, IGFBP-2, IGFBP-3, IGFBP-4, IGFBP-6, IGF-I, IGF-I SR, IL-1α, IL-1β, IL-1, IL-1 R4, ST2, IL-3, IL-4, IL-5, IL-6, IL-8, IL-10, IL-11, IL-12 p40, IL-12p70, IL-13, IL-16, IL-17, I-TAC, alpha chemoattractant, lymphotactin, MCP-1, MCP-2, MCP-3, MCP-4, M-CSF, MDC, MIF, MIG, MIP-1α, MIP-1β, MIP-1δ, MIP-3α, MIP-3β, MSP-a, NAP-2, NT-3, NT-4, osteoprotegerin, oncostatin M, PARC, PDGF, P1GF, RANTES, SCF, SDF-1, soluble glycoprotein 130, soluble TNF receptor I, soluble TNF receptor II, TARC, TECK, TGF-beta 1, TGF-beta 3, TIMP-1, TIMP-2, TNF-α, TNF-β, thrombopoietin, TRAIL R3, TRAIL R4, uPAR, VEGF and VEGF-D. In some embodiments, a cytokine is interleukin 2 (IL-2), interleukin 10 (IL-10), or interleukin 22 (IL-22). In some embodiments, a therapeutic protein is interleukin 10 (IL-10) or a functional fragment thereof.
In some embodiments, a Bacteroides membrane-associated protein facilitates the secretion of a fusion protein into the periplasm of the engineered Bacteroides. In some embodiments, a Bacteroides membrane-associated protein facilitates the display of the fusion protein on the outer membrane of the engineered Bacteroides. In some embodiments, a fusion protein is incorporated into a Bacteroides outer membrane vesicle (OMV).
Accordingly, some aspects of the present disclosure provides engineered Bacteroides outer membrane vesicles (OMVs) comprising a fusion protein, as described herein. A Bacteroides OMV refers to a spherical bud of the outer membrane filled with outer membrane and periplasmic contents. OMVs are commonly produced by Gram-negative bacteria. The production of OMVs allows bacteria to interact with their environment, and OMVs have been found to mediate diverse functions, including promoting pathogenesis, enabling bacterial survival during stress conditions and regulating microbial interactions within bacterial communities. Derived from the cell envelope, OMVs contain a similar outer membrane consisting of LPS and phospholipids as well as lipoproteins and membrane proteins like porins, ion channels, adhesins, and enzymes. Apart from periplasmic proteins and peptidoglycans, OMVs also carry specific cargos in their lumina, e.g., proteases, nucleic acids, and toxins such as the cholera toxin in Vibrio cholerae and Cytolysin A in enterotoxic E. coli. In some embodiments, a fusion protein of the present disclosure, when incorporated into the engineered Bacteroides OMV, is in the lumen of the OMV, or is displayed on the surface of the OMV.
In some embodiments, an engineered Bacteroides OMV of the present disclosure maybe used to deliver a fusion protein to another cell, e.g., a eukaryotic cell. In some embodiments, a fusion protein is delivered to an immune cell. For example, a fusion protein may be delivered to a B cell, a dendritic cell, a granulocyte, a megakaryocyte, a monocytes/macrophage, a natural killer cell, a platelet, a red blood cell, or a T cell or a thymocyte. In some embodiments, an immune cell is an intestinal mucosal immune cell. An intestinal mucosal immune cell is a component of the mucosal immune system at the gastrointestinal barrier, which contains small foci of lymphocytes and plasma cells are scattered widely throughout the lamina propria of the gut wall. One skilled in the art is familiar with different types of immune cells and the gastrointestinal mucosal immune system.
To deliver the fusion protein to another cell, an engineered Bacteroides OMV interacts with the cell, e.g., the immune cell. Fusion proteins may be delivered in a number of different ways. For example, in some embodiments, a fusion protein is displayed on the surface of an engineered Bacteroides OMV and is recognized by a receptor on the surface of a cell, e.g., an immune cell, receiving the fusion protein. In some embodiments, the engineered Bacteroides OMV undergoes lysis and releases the fusion protein to the vicinity of the cell receiving the fusion protein. In some embodiments, an engineered Bacteroides OMV undergoes membrane fusion with the cell receiving the fusion protein. In some embodiments, an engineered Bacteroides OMV is internalized as a whole entity by the cell receiving the fusion protein via endocytosis.
An engineered Bacteroides or engineered Bacteroides OMV of the present disclosure, may be administered to a subject. In some embodiments, a subject has a disorder that may be treated with a heterologous protein delivered by an OMV of an engineered Bacteroides. In some embodiments, the disorder is an intestinal disorder. For example, an intestinal disorder may be inflammatory bowel disease (IBD) or Crohn's disease. In some embodiments, an engineered Bacteroides or engineered Bacteroides OMV is administered orally or intrarectally.
In some embodiments, a subject is a mammal. In some embodiments, a subject is human.
A “nucleic acid” refers to at least two nucleotides covalently linked together, and in some instances, may contain phosphodiester bonds (e.g., a phosphodiester “backbone”). In some embodiments, a nucleic acid (e.g., an engineered nucleic acid) of the present disclosure may be considered a nucleic acid analog, which may contain other backbones comprising, for example, phosphoramide, phosphorothioate, phosphorodithioate, O-methylphophoroamidite linkages, and/or peptide nucleic acids. Nucleic acids (e.g., components, or portions, of the nucleic acids) of the present disclosure may be naturally occurring or engineered. Nucleic acids of the present disclosure may be single-stranded (ss) or double-stranded (ds), as specified, or may contain portions of both single-stranded and double-stranded sequence (e.g., a single-stranded nucleic acid with stem-loop structures may be considered to contain both single-stranded and double-stranded sequence). It should be understood that a double-stranded nucleic acid is formed by hybridization of two single-stranded nucleic acids to each other. Nucleic acids may be DNA, including genomic DNA and cDNA, RNA or a hybrid/chimeric of any two or more of the foregoing, where the nucleic acid contains any combination of deoxyribo- and ribonucleotides, and any combination of bases, including uracil, adenine, thymine, cytosine, guanine, inosine, xanthine, hypoxanthine, isocytosine, and isoguanine.
An “engineered nucleic acid” is a nucleic acid that does not occur in nature. It should be understood, however, that while an engineered nucleic acid as a whole is not naturally-occurring, it may include nucleotide sequences that occur in nature. In some embodiments, an engineered nucleic acid comprises nucleotide sequences from different organisms (e.g., from different species). For example, in some embodiments, an engineered nucleic acid includes a murine nucleotide sequence, a bacterial nucleotide sequence, a human nucleotide sequence, and/or a viral nucleotide sequence. The term “engineered nucleic acids” includes recombinant nucleic acids and synthetic nucleic acids. A “recombinant nucleic acid” refers to a molecule that is constructed by joining nucleic acid molecules and, in some embodiments, can replicate in a live cell. A “synthetic nucleic acid” refers to a molecule that is amplified or chemically, or by other means, synthesized. Synthetic nucleic acids include those that are chemically modified, or otherwise modified, but can base pair with naturally-occurring nucleic acid molecules. Recombinant nucleic acids and synthetic nucleic acids also include those molecules that result from the replication of either of the foregoing. Engineered nucleic acids of the present disclosure may be encoded by a single molecule (e.g., included in the same plasmid or other vector) or by multiple different molecules (e.g., multiple different independently-replicating molecules).
Engineered nucleic acids of the present disclosure may be produced using standard molecular biology methods (see, e.g., Green and Sambrook, Molecular Cloning, A Laboratory Manual, 2012, Cold Spring Harbor Press). In some embodiments, engineered nucleic acids are produced using GIBSON ASSEMBLY® Cloning (see, e.g., Gibson, D. G. et al. Nature Methods, 343-345, 2009; and Gibson, D. G. et al. Nature Methods, 901-903, 2010, each of which is incorporated by reference herein). GIBSON ASSEMBLY® typically uses three enzymatic activities in a single-tube reaction: 5′ exonuclease, the Y extension activity of a DNA polymerase and DNA ligase activity. The 5′ exonuclease activity chews back the 5′ end sequences and exposes the complementary sequence for annealing. The polymerase activity then fills in the gaps on the annealed regions. A DNA ligase then seals the nick and covalently links the DNA fragments together. The overlapping sequence of adjoining fragments is much longer than those used in Golden Gate Assembly, and therefore results in a higher percentage of correct assemblies.
Engineered nucleic acids of the present disclosure may be included within a vector, for example, for delivery to a cell. A “vector” refers to a nucleic acid (e.g., DNA) used as a vehicle to artificially carry genetic material (e.g., an engineered nucleic acid construct) into a cell where, for example, it can be replicated and/or expressed. In some embodiments, a vector is an episomal vector (see, e.g., Van Craenenbroeck K. et al. Eur. J. Biochem. 261, 5665, 2000, incorporated by reference herein). A non-limiting example of a vector is a plasmid. Plasmids are double-stranded generally circular DNA sequences that are capable of automatically replicating in a host cell. Plasmid vectors typically contain an origin of replication that allows for semi-independent replication of the plasmid in the host and also the transgene insert. Plasmids may have more features, including, for example, a “multiple cloning site,” which includes nucleotide overhangs for insertion of a nucleic acid insert, and multiple restriction enzyme consensus sites to either side of the insert. Another non-limiting example of a vector is a viral vector. Any of the engineered nucleic acids of the present disclosure, for example, a nucleic acid encoding a fusion protein, may be present on a vector (e.g., and delivered to a Bacteroides cell).
A “promoter” refers to a control region of a nucleic acid sequence at which initiation and rate of transcription of the remainder of a nucleic acid sequence are controlled. A promoter may also contain sub-regions at which regulatory proteins and molecules may bind, such as RNA polymerase and other transcription factors. Promoters may be constitutive, inducible, activatable, repressible, tissue-specific or any combination thereof. A promoter drives expression or drives transcription of the nucleic acid sequence that it regulates. A promoter is considered to be “operably linked” when it is in a correct functional location and orientation in relation to a nucleic acid sequence it regulates to control (“drive”) transcriptional initiation and/or expression of that sequence.
Promoters of an engineered nucleic acid construct may be “inducible promoters,” which refer to promoters that are characterized by regulating (e.g., initiating or activating) transcriptional activity when in the presence of, influenced by or contacted by an inducer signal.
Effective treatment of intestinal disorders using protein drugs such as anti-inflammatory cytokines is hampered by too low therapeutic levels at the required site—the intestinal mucosa; oral administration is impeded by protein degradation in the acid-rich and protease-rich upper gastro-intestinal tract; and systemic medication leads to side effects and requires frequent injections of high doses due to the short in vivo half-lives of proteins. One approach to overcome these difficulties is targeted enteric protein delivery by genetically engineered bacteria, as provided herein. Abundant, commensal Bacteroides species naturally and stably colonize the gut in concentrations of up to 30% of the intestinal flora.
In the present study, Bacteroides spp. was engineered to produce therapeutic proteins for intestinal delivery. The proteins were packaged into outer membrane vesicles (OMVs), which are constitutively produced by all Gram-negative bacteria. The circumventing membrane of OMVs protects their cargo from dilution and proteolytic degradation by intestinal proteases even when transported over long distances. Surprisingly, Bacteroides OMVs were found to cross the epithelial layer and deliver their cargo to host immune cells (104). This was harnessed by packing anti-inflammatory molecules such as interleukin-10 into Bacteroides OMVs, which deliver the molecules to mucosal immune cells, leading to amelioration of intestinal inflammation.
Heterologous proteins were targeted to OMVs by genetic fusion to peptide tags derived from Bacteroides OMV proteins. Three proteins enriched in OMVs were first identified by literature research and whether fusion to a reporter protein leads to its integration and activity in OMVs was tested. To prohibit disturbing effects due to the bulky OMV proteins, the minimal sequence tags crucial for the export to OMVs were determined. The anti-inflammatory cytokine IL-10 was then targeted to OMVs for future usage in the treatment of inflammatory bowel diseases (IBD). Additionally, the colonization behavior and tolerance of different Bacteroides spp. in the mouse gut were analyzed to prepare for subsequent in vivo tests of the designed fusion proteins delivered by the optimal chassis.
Thus, provided herein are treatment systems used for intestinal protein delivery that require low dosage due to targeted high concentrations, are cost-effective, are well-tolerated, and are potentially applicable to many diseases. The methods of the present disclosure are suitable for a wide range of therapeutic proteins that currently encounter difficulties with regard to intestinal delivery.
To develop a versatile and effective intestinal protein delivery system by harnessing a human gut commensal, Bacteroides were engineered to produce outer membrane vesicles (OMVs) that carry heterologous proteins which can be delivered to host cells in the gut—once applied to humans. Bacteroides spp. are one of the numerically dominant beneficial genera of the human intestinal microbiota with stable and robust colonization (28, 30), which particularly qualifies them for long-term therapeutics. Recently, the genetic parts for the precise modulation of gene expression in B. thetaiotaomicron were designed by our group (140). The isopropyl β-D-1-thiogalactopyranoside (IPTG)-inducible expression system, which integrates into the Bacteroides genome after conjugation from E. coli, was deployed. The present study to generate therapeutic Bacteroides spp. was conducted in a two-part process. In the first part, for suitable fusion tags derived from proteins in B. thetaiotaomicron and B. ovatus OMVs, which are able to target heterologous proteins to OMVs of Bacteroides spp., were examined. In the second part, the in vivo tolerance and colonization behavior of different Bacteroides species were tested in a mouse model, identifying exemplary species for use in intestinal delivery of the fusion proteins.
While mass spectrometry and biochemical analyses have shown that some proteins are enriched or exclusively found in OMVs, including those generated by Bacteroides spp. (74), the mechanism of their selection remains elusive. At present, it is unknown if a particular part of an OMV protein is responsible for its secretion via Bacteroides OMVs. This question was addressed through the investigation of whether the fusion of heterologous proteins to native OMV proteins promotes their targeting to OMVs and if truncated peptide tags derived from the OMV proteins' N-termini suffice in doing so.
The OMV Proteins BT1491, BT3238, and BACOVA_04502 Promote Enrichment of Heterologous Proteins in OMV-Containing Culture Supernatants.
First, the literature for Bacteroides proteins found enriched in OMVs was searched in order to identify possible candidates for OMV targeting. In the only large-scale proteomic analysis of B. thetaiotaomicron OMVs, Elhenawy et al. reported the association of 60 proteins with OMVs that were not detectable in the bacterial outer membrane, indicating specific targeting to OMVs (74). Of these proteins, two were selected: the hypothetical protein BT1491 and the SusD (starch binding protein) homolog BT3238. Both were detected not only by sensitive MS approaches but also as strong bands on a Coomassie gel of purified OMVs. This indicates high concentrations rather than contaminations due to poor purification. Further, the inulinase BACOVA_04502, a B. ovatus OMV protein (141), was chosen and found to be substantially enriched in OMVs after expression in B. fragilis, suggesting its universal sorting mechanism among Bacteroides spp. (74). BT1491, BT3238, and BACOVA_04502 are in the following referred to as OMV proteins. To examine whether these proteins are able to guide heterologous proteins into OMVs, each was genetically fused to the N-terminus of the luciferase reporter NanoLuc (NL, 19 kDa) with a C-terminal hexahistidine tag (His tag, H6) (SEQ ID NO: 49). As negative control, a His-tagged NL without the OMV protein was used. Previous studies in other bacteria have shown that fusion proteins with N-terminal OMV-proteins are exported more reliably than with C-terminal (83, 84).
The fusion proteins and the negative control were expressed in B. thetaiotaomicron, and the presence of NL in cell lysates as well as OMV-containing culture supernatants was determined. As depicted in
Bioinformatics-Based Approach to Identify Signal Peptides in B. thetaiotaomicron OMV Proteins.
Next, the OMV proteins were evolved towards efficient secretion with preferably small fusion tags. The relatively large proteins BT1491, BT3238, and BACOVA_04502 (256, 519, and 650 amino acids, respectively) elevate the risk of misfolding or steric hindrance of their fusion partners. This may impede their function or have a negative impact on secretion efficiency. Further, the intrinsic function of BT1491 is not elucidated and might have unexpected side effects or harm host cells in future applications.
It was reasoned that the N-terminal regions of the proteins are crucial for their secretion in OMVs, since certain N-terminal amino acid patterns, so called signal peptides, are essential and sufficient to direct proteins to the general secretory pathway, the periplasm, or membrane (142). To test this hypothesis, the amino acid sequences of the OMV proteins were analyzed for characteristic features and conserved sequence patterns using bioinformatics tools. Although no universal consensus sequence exists in signal peptides of Gram-negative bacteria, they typically display regions of distinct polarities or charges and a certain degree of conservation at the −3 and +1 positions relative to the cleavage site separating signal peptide and mature protein (143, 144). Further, secreted proteins cleaved by signal peptidase (SP)I behind an alanine are distinguished from lipoproteins cleaved by SPII before the cysteine that anchors the mature protein into the membrane by its attached lipid moiety.
Four bioinformatics software (LipoP 1.0 (144), SignalP 4.1 (145), Phobius (146), and SignalBLAST (147)) predicted N-terminal signal peptides in the three OMV proteins characteristic for proteins of the secretory pathway. Further, they were predicted to be lipoproteins with SPII specific cleavage sites (before a conserved cysteine) by the LipoP tool, which discriminates between signal peptides of secreted proteins and lipoproteins as well as N-terminal transmembrane helices in Gram-negative bacteria (144) (
To investigate whether lipoproteins are enriched in Bacteroides OMVs and whether a consensus sequence specific for OMV lipoproteins exists in their signal peptides, all B. thetaiotaomicron OMV proteins were compared with proteins exclusively detected in the OM in the study of Elhenawy et al. (74) using bioinformatics analysis. The protein BACOVA_04502 from B. ovatus that was used in this study was also included in the list of OMV proteins. Only 4 out of 61 Bacteroides OMV proteins had no predicted signal peptide by neither of the four tools. 31 OMV proteins were predicted lipoproteins by the LipoP software with cleavage sites occurring within the first 17-42 amino acids (aa) (
To compare amino acid sequences surrounding the SPII-cleavage site of OMV lipoproteins with those of other lipoproteins, sequence logos that indicate 4 aa before and 6 aa after the cleavage site were generated for both groups (
In conclusion, the sequence logos do not clearly point towards the existence of a conserved amino acids sequence adjacent to the cleavage site that targets particular lipoproteins to OMVs. However, a relative enrichment in aspartic acid was seen in the N-terminus of mature lipoproteins targeted to OMVs.
Truncated OMV Proteins Target NanoLuc to OMV-Containing Supernatants.
Based on the signal peptides predicted by bioinformatics tools, a short N-terminal sequence was experimentally tested to determine if it is sufficient for protein export into OMVs or if interactions mediated by the whole protein are required. To optimize the tags for small size and export efficiency, truncations of different length between 18 and 100 N-terminal amino acids were created (schemes in
For BT1491, this minimal length was 25 amino acids (BT1491Δ25); for BT3238 35 aa (BT3238Δ35) and for BACOVA_04052 28 aa (BACOVA_04052Δ28). BT1491Δ25, BT3238Δ35, and BACOVA_04502Δ28, hereinafter referred to as OMV tags, were chosen for the following experiments, as they feature a good balance between small size and export efficiency. Importantly, the tags showed higher export efficiencies than the FL proteins, with 5-, 2-, and 6-fold NL activity, respectively, suggesting that the rest of the protein does not mediate OMV translocation. Notably, for efficient export the tags required more amino acids than the predicted signal peptide only. Strikingly, the most efficient tags (BT1491Δ50, BT3238Δ35, BACOVA_04502Δ28) substantially reduced the intracellular luminescence showing that secretion via tags is an effective means of shedding proteins from the cell.
To conclude, heterologous proteins were enriched in OMV-containing supernatants after fusion to either of the three OMV proteins BT1491, BT3238, and BACOVA_04502, as well as optimized C-terminally truncated versions with increased export efficiency over the FL proteins. Further, a minimal sequence tag that is essential for secretion exists for each protein, indicating the existence of an N-terminal OMV signal peptide with little or no role for the rest of the protein.
Functional Heterologous Proteins Fused to OMV Tags Translocate to OMVs.
The previous experiments showed that the protein of interest was secreted in supernatants. However, the enrichment in culture supernatants does not distinguish between OMV- and soluble, non-OMV-mediated secretion. Hence, purified OMVs harvested from cell-free supernatants by high-speed centrifugation were analyzed next. The successful enrichment of OMVs in the pellet was confirmed by electron microscopy (
In conclusion, these results indicate that the NL reporter is secreted via OMVs by fusing it to B. thetaiotaomicron FL OMV proteins or optimized OMV tags without disturbing its luciferase functionality. These experiments show that the 25-35 N-terminal amino acids are sufficient for targeting to OMVs.
OMV Tags are Universal Across Bacteroides Spp.
OMV tag-NL fusions were expressed in B. fragilis and B. vulgatus to test whether the tags mediate export to OMVs in other Bacteroides species than B. thetaiotaomicron. While the B. thetaiotaomicron and B. fragilis strains are both purchased human isolates, B. vulgatus was isolated in-house from feces of Swiss Webstar mice. Purified OMVs of B. fragilis and B. vulgatus showed increased luminescence when expressing fusion proteins of OMV tags with NL compared to NL alone (
Since the three OMV tags are predicted to harbor signal peptides characteristic for lipoproteins, it is likely that they are anchored to the OMV membrane. The orientations of the OMV tag-NL fusion proteins in the membrane were investigated next. Although lipoproteins in Gram-negative bacteria have been predominantly found anchored in the inner leaflet of the outer membrane (150), a high percentage of Bacteroides lipoproteins have been found cell surface-exposed (151). BT3238 is a homolog of SusD, a starch-binding protein, and is presumably surface-exposed. Surface-exposed fusion proteins would facilitate interaction of protein therapeutics with cell surface receptors.
Proteinase K protection assays were performed on OMVs purified from B. thetaiotaomicron cultures expressing the three OMV tag-NL fusion proteins to examine whether they are accessible on the OMV surface or protected by the vesicle structure (
Consistent with the luminescence data, Western blots revealed lower protein amounts after incubation with proteinase K for 30 min compared to without treatment (
Fusion tags might be displayed at the cell surface and the OMV tags mediate anchorage in the outer leaflet of the OMV membrane. However, fusion proteins are substantially more stable when OMVs were intact. Up to 10% remained even after 24 hours PK digestion pointing towards a protective function of OMVs. Here, it is worth noting, that B. thetaiotaomicron produces a thick capsule that covers not only its cell envelope but also its OMVs (152) and might limit access of proteinase K to surface proteins leading to incomplete digestion. Alternatively, a proportion of OMV tags could have failed to display the fused NL on the surface and rather remain inside the lumen like shown for an E. coli fusion proteins in a previous study (84).
Having established an OMV-based secretion system for NL in the gut commensal B. thetaiotaomicron, it was utilized for the secretion of a therapeutic protein to treat intestinal disorders. Reduced levels of the immunoregulatory cytokine interleukin (IL)-10 have been linked to several intestinal diseases like inflammatory bowel diseases (IBD) (124, 125, 127, 153, 154). Therefore, the murine IL-10 (mIL-10) cytokine were genetically fused to the three OMV tags, and these fusion proteins were expressed in B. thetaiotaomicron and B. vulgatus. OMV tag-NL fusions and untagged mIL-10 which is not directed to OMVs were used as negative controls.
B. thetaiotaomicron cultures were grown for 6 h and B. vulgatus cultures for 20 h to obtain comparable OD600. Culture supernatants were concentrated using centrifugal filter tubes with either 10 or 100 kDa cut-off membranes, which retain large protein complexes and OMVs. Soluble dimeric IL-10 (37 kDa) is retained by the 10 kDa cut-off membrane, whereas it passes the 100 kDa membrane, and is therefore found in the >10 kDa but not in the >100 kDa fraction. Concentrates were analyzed for mIL-10 concentration [mIL-10] by sandwich enzyme-linked immunosorbent assay (ELISA) and normalized to total protein concentrations as determined by bradford assays (
To determine if the mIL-10 found in concentrated culture supernatants is associated with OMVs and is not a result of mIL-10 aggregates heavier than 100 kDa, OMVs were purified by high-speed centrifugation. [mIL-10] in purified OMVs were lower than in concentrated supernatants but still detectable (
Next, whether it will be conceivable to deploy Bacteroides spp. in vivo in future applications was assessed. Therefore, whether engineered Bacteroides species are able to stably colonize the intestine and whether they are tolerated by mice were investigated. As Bacteroides spp. are unable to colonize the gut of specific-pathogen-free (SPF) mice per se (28), colonization was promoted by antibiotic treatment for 7 days without sterilizing the intestine (155). Subsequently, mice were orally gavaged with B. thetaiotaomicron, B. fragilis, or B. vulgatus mouse isolate expressing the BT1491Δ25-NL fusion protein (Bt-BT1491Δ25-NL, Bf-BT1491Δ25-NL, Bvm-BT1491Δ25-NL). 4-7 days after bacterial gavage, experimental colitis was induced by administration of dextran sulfate sodium (DSS) in drinking water for 8 days. Simultaneously, IPTG was administered to induce the expression of the BT1491Δ25-NL fusion protein (
First, it was shown that mice were successfully colonized with Bacteroides spp. and that DSS and IPTG application did not interfere with colonization. Engineered Bacteroides were identified by plating mice feces on selective media (BHIS+Gm+Em) for Bacteroides and plasmid selection. B. vulgatus was still detected in feces 8 days after gavage, and B. thetaiotaomicron and B. fragilis 14 days after gavage (
To assess how well Bacteroides spp. are tolerated under inflammatory conditions in mice, the impact of Bacteroides colonization on intestinal inflammation in the DSS model of colitis was investigated. DSS-induced colitis is a widely used mouse model for experimental intestinal colitis which is thought to be induced by direct toxicity of DSS to colonic epithelial cells of the basal crypts (156, 157). Colitis severity was evaluated by the macroscopic feature of a reduced colon length and histological features of the colon such as inflammatory infiltrates, edema, and epithelial defects scored in a blinded fashion. Further, wellbeing of mice was monitored by recording the mouse weight over the experimental course.
Colitis was successfully induced, as non-colonized mice that where administered 3% DSS for 7 day in their drinking water revealed shortened colons (
Mice were stably colonized with Bacteroides spp. over the course of the experiment and expression of recombinant proteins was successfully induced. Bacteroides colonization was well tolerated and had no detrimental effect on intestinal inflammation in a DSS-induced colitis model.
The potential to deploy engineered microbes as intestinal drug delivery systems in human medicine brought forth an important emerging field of research. It enables a targeted, low-dose, and cost-effective treatment by using orally administered microbes that produce and deliver therapeutic agents directly to the site of action. The majority of work has been done in E. coli (131, 135, 136) and L. lactis (127, 132), which are not capable to continuously deliver drugs in high therapeutic doses due to their low abundance in the GI tract and their inability to colonize the gut, respectively. In this work, the basis for the development of a novel therapeutic protein delivery system to treat intestinal disorders was established. The aim was to enable the use of outer membrane vesicles (OMVs) produced by gut commensals of the Bacteroides species as delivery vehicles for therapeutic proteins to the intestine. Therefore, Bacteroides spp. that naturally reside in the intestinal mucus layer were engineered to generate OMVs that carry heterologous proteins which can be delivered to host cells in vivo.
Three OMV proteins—BT1491, BT3238, and BACOVA_04502—found in natural Bacteroides OMVs that translocated the reporter NanoLuc (NL) to OMVs when genetically fused to its N-terminus were identified. These proteins were further optimized by truncation to small fusion tags of 25-35 amino acids (OMV tags) while maintaining the ability to target NL for OMV-based secretion in B. thetaiotaomicron, B. fragilis, and a murine isolate of B. vulgatus. The three OMV tags were N-terminal regions of OMV lipoproteins comprising a predicted signal peptide and 4-16 additional amino acids. Our data suggest that the cargo is carried on the OMV surface via a lipid anchor in the outer leaflet of the OMV membrane. In a proteinase K assay, the speed of NL degradation was decreased by intact OMVs and 3-6% of the cargo remained 24 h after proteinase K addition. Additionally, translocation of the therapeutic protein IL-10 to OMVs in concentrations up to 0.3 ng OMV-associated IL-10 per mL culture supernatant in a proof-of-concept experiment were achieved. These results indicate that it is possible to target heterologous proteins to Bacteroides OMVs. For future in vivo applications, all three tested Bacteroides spp.—B. thetaiotaomicron, B. fragilis, and B. vulgatus mouse isolate—were identified as suitable chassis that stably colonized the gut and had no detrimental effects on mice in an experimental colitis model. This will be the basis for further in vivo work to examine therapeutic effects of OMV-associated proteins.
Gibson assembly master mix: 320 μL 5× Isothermal Master Mix (25% PEG-8000, 500 mM Tris-HCl pH 7.5, 50 mM MgCl2, 50 mM DTT, 5 mM NAD, 1 mM each of the four dNTPs), 0.64 μL 10 U/μL T5 exonuclease, 20 μL 2 U/μL Phusion High-fidelity DNA Polymerase, 0.16 μL 40 000 U/μL Taq DNA Ligase, 860 μL ddH2O
Supplemented brain-heart infusion (BHIS) medium: 37 g/L BHI, 5 g/L yeast extract, 10 mg/L hemin, 1 mg/L Vitamin K3, 0.5 g/L cysteine; diluted in ddH2O
Supplemented trypticase yeast extract glucose (TYG) medium: 10 g/L trypticase, 5 g/L yeast extract, 1 g/L Na2CO3, 10 mM glucose, 80 mM potassium phosphate buffer (pH 7.3), 20 mg/L MgSO4.7H2O, 400 mg/L NaHCO3, 80 mg/L NaCl, 0.0008% CaCl2, 4 μg/mL FeSO4.7H2O, 10 mg/L hemin, 1 mg/L Vitamin K3, 0.5 g/L cysteine; diluted in ddH2O
Bacteroides fragilis NCTC 9343,
Bacteroides thetaiotaomicron
Bacteroides vulgatus,
E. coli S17.1 λpir
Bacterial Growth Conditions.
Bacterial strains used in this study are listed above. E. coli S17.1 λpir was grown in LB medium or on LB agar supplemented with carbenicillin (100 μg/mL) for plasmid selection at 37° C. All Bacteroides strains were grown in BHIS or TYG media or plates in a Coy anaerobic chamber with 85% N2, 5% H2, and 10% CO2 at 37° C. Media and plates were pre-reduced overnight in an anaerobic atmosphere before culture inoculation. The antibiotics erythromycin (25 μg/mL) and gentamicin (200 μg/mL) were supplemented when necessary. Expression of fusion proteins was induced by addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) (100 μM).
Generation of Recombinant Bacteroides Strains.
pNBU1 constructs were generated in E. coli and then conjugated into Bacteroides strains. For IL-10-constructs, a codon-optimized sequence for murine IL-10 was purchased as gBlocks gene fragment from Integrated DNA Technologies and PCR-amplified. OMV-tag-NL/IL-10 fusion proteins were cloned into a pNBU1 backbone by Gibson assembly. The pNBU1 plasmid encodes the intN1 tyrosine integrase, which mediates sequence-specific recombination between the attN site of pNBU1 and one of two attBT sites of the tRNASer genes in the Bacteroides genome. Insertion inactivates the tRNASer gene. Simultaneous insertion into both tRNASer genes is unlikely due to the essentiality of tRNASer.
Polymerase Chain Reaction (PCR).
Primers were designed with the software Geneious® to have 22-38 base pairs annealing with the template with melting temperatures of 55-65° C., calculated with the OligoCalc melting temperature calculator (biotools.nubic.northwestern.edu/OligoCalc.html#helpthermo) (nearest neighbor). Homology regions for Gibson assembly were added to the 5′ end where necessary. 50 μL reactions with KAPA HiFi polymerase and 1 ng template DNA for 25 cycles were performed according to the instructions of the manufacturer. Extension durations were adjusted for each fragment to be 30 s/1 kbp. Primers and plasmids used in this study are listed in above.
DNA Gel Electrophoresis and Purification.
PCR products were analyzed by gel electrophoresis (135 V, 25 min) of 5 μl PCR product in DNA gel loading dye using 1% agarose gels stained with 1×SYBR safe DNA gel stain. DNA bands were visualized with the ChemiDoc imaging system. Remaining 45 μL PCR products were purified with the QIAquick PCR purification kit. DpnI digestion of purified PCR products was performed in CutSmart buffer for 1 h at 37° C. according to the manufacturer's instruction. DNA purification was repeated and DNA was eluted in 50 μL elution buffer.
Gibson Assembly.
Gibson assembly of DNA fragments with 30-60 base pair homology regions were done in 10 μL reactions containing 5 μL house-made Gibson assembly master mix and 5 μL mix of the DNA to be assembled, where DNA fragments should be in equimolar amounts (10-100 ng of each). Reaction mix was incubated for 1 h at 50° C.
Transformation.
Electrocompetent E. coli S17.1 λpir was transformed as follows. Bacteria were thawed on ice and 2 μL Gibson assembly mix with DNA was added. Cells were transferred into a pre-cooled electroporation cuvette (0.2 cm gap) and pulsed with 2500 V in an electroporator. 1 mL LB was immediately added and cells were recovered for 1 h at 37° C. while shaking. Cells were plated on pre-warmed LB agar plates containing carbenicillin (100 μg/mL) for plasmid selection and incubated overnight at 37° C.
Plasmid Purification and Sanger DNA Sequencing.
The presence of the correctly assembled DNA construct was verified by Sanger Sequencing. Therefore, DNA of a 3 mL overnight E. coli culture was isolated using the QIAprep spin miniprep kit according to the manufacturer's protocol. Plasmids were sequenced by QuintaraBio and sequences were analyzed with Geneious.
Conjugation of Constructs into Bacteroides Spp.
Constructs were conjugated from E. coli into B. thetaiotaomicron, B. fragilis, or B. vulgatus as follows. E. coli donor and Bacteroides recipient strains were grown overnight. 250 μL E. coli culture was washed once with PBS and cell pellet was resuspended in 1 mL Bacteroides culture (1:5 ratio). The mating mixture was pelleted, resuspended in 25 μL BHIS medium, and spotted onto pre-warmed BHIS agar plates. Plates were incubated upright aerobically overnight at 37° C. On day 2, cells were scraped off from the plate and fully resuspended in 1 mL BHIS medium. 250 μL suspension was plated on BHIS agar plates containing gentamicin (200 μg/mL) for Bacteroides selection and erythromycin (25 μg/mL) for plasmid selection (BHIS+Gm+Em). Plates were incubated anaerobically for 48 h at 37° C. On day 4, colonies were re-isolated on pre-reduced BHIS+Gm+Em plates. On day 5, colonies could be used for liquid overnight cultures in BHIS or TYG medium.
Preparation of Culture Supernatants, Concentrated Supernatants, and Outer Membrane Vesicle (OMVs).
Bacteroides cultures were grown overnight and subcultured 1:100 in prereduced BHIS or TYG media supplemented with IPTG (100 μM). After 6-20 h the optical densities at 600 nm (OD600) of 300 μl of cultures were measured and proceeded as follows.
Preparation of Culture Supernatants.
NL-producing Bacteroides cultures where centrifuged at 5,000×g for 5 min. Culture supernatants were saved and cell pellets were washed in 1 mL 1× phosphate-buffered saline (PBS) and resuspended in 1 mL PBS. Luciferase activity of culture supernatant and cells was measured as described in 6.2.5.
Preparation of Concentrated Supernatants.
mIL-10 producing Bacteroides cultures where centrifuged at 10,000×g for 15 min. Phenylmethyl-sulfonyl fluoride (PMSF) (1 mM) and proteinase inhibitor cocktail was added to culture supernatants according to the manufacturer's instructions. Supernatants were concentrated 20-25-fold using centrifugal filters with 10 kDa and 100 kDa cut-off membranes according to the manufacturer's instruction with 14,000×g for concentration spin 1,000×g for reverse spin. Concentrated supernatants were stored at −80° C. at least overnight and mIL-10 concentration were determined as described above.
OMV Purification.
50-100 mL of Bacteroides subcultures were grown for 6-20 h in BHIS or TYG medium supplemented with IPTG (100 μM) and centrifuged at 10,000×g for 15 min at 4° C. PMSF (1 mM) and proteinase inhibitor cocktail was added to culture supernatants according to the manufacturer's instructions. Supernatant was filtered through a 0.45 μm membrane and centrifuged at 70,000×g for 70 min at 4° C. The pellet was washed with PBS and the centrifugation step was repeated. The OMV-containing pellet was resuspended in 100 μL PBS and stored at 4° C. until usage if necessary. NL activity or mIL-10 concentrations of purified OMVs were determined as described in 6.2.5. and 6.2.7., respectively.
Purity of OMVs was assessed by transmission electron microscopy. Samples were absorbed onto carbon-coated copper grids, washed with ddH2O, and stained with 1% aqueous uranyl acetate. Samples were viewed on a JEOL 1200EX transmission electron microscope.
Bradford Protein Quantification Assay.
The quick start Bradford protein assay was used for colorimetric detection and quantification of protein concentrations in concentrated supernatants and OMV suspensions according to the microplate microassay protocol provided by the manufacturer. Briefly, samples were diluted in PBS according to expected protein concentration. Bovine serum albumin (BSA) was serially diluted in PBS to generate standards of known concentrations of 0-25 μg/mL. 150 μL standard or diluted sample were mixed with 150 μL of 1× dye reagent and incubated at room temperature for 5-10 min. Absorbance was measured at 595 nm on a microplate hybrid reader. Standards and samples were tested in triplicates. The concentration of each sample was determined based on the standard curve of known BSA concentrations.
Luciferase Assay.
Luciferase activities of cell suspensions, culture supernatants, or purified OMVs prepared from NL-producing Bacteroides cultures as described in 6.2.3. were determined by a Nano-Glo luciferase assay. NanoLuc luciferase present in samples catalyzes oxidation of the exogenously added substrate furimazine to generate a glow-type bioluminescence (λmax=460 nm) with a signal half-life of ca. 120 minutes. The working reagent was prepared by diluting luciferase assay substrate 1:50 in luciferase assay buffer. The reagent contains an integral lysis buffer that allows usage directly on cells expressing NanoLuc luciferase. 25 μL working reagent was mixed with 25 μL sample (cell suspensions, culture supernatants, purified OMVs, or feces homogenate) and luminescence was measured with an integration time of 1 second at a gain setting of 100 in a microplate hybrid reader. Luciferase activities were normalized to the OD600 of 300 μl of culture at the time of harvest.
Western Blot Analysis: Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE).
OMV suspensions were diluted in PBS to obtain same concentrations (as determined by Bradford assay) among all samples. 4 or 5 μg OMVs were mixed with 4×LDS sample buffer containing 10% β-mercaptoethanol, boiled for 10 minutes at 99° C., and loaded on 4-12% Bis-Tris NuPAGE gels. Proteins were separated by molecular weight at 200 V for 35 min in 1×MES SDS running buffer.
Western Blotting and Immunodetection.
Proteins were transferred onto nitrocellulose membranes by a dry electroblotting system using the iBlot™ gel transfer device and iBlot™ Gel transfer stacks with integrated nitrocellulose transfer membrane. Program P3 (20 V) was run for 6 or 7 minutes, depending on the protein size. Membranes were blocked with 1×PBS-T (0.05% Tween20 in PBS) containing 5% skim milk for at least 120 min at RT and gentle shaking. After washing with PBS-T, membranes were incubated with anti-6×His tag primary antibody (1:5,000) in PBS-T+5% milk overnight at 4° C. and gentle shaking. Membranes were then washed with PBS-T and incubated in horseradish peroxidase (HRP)-conjugated anti-mouse secondary antibody (1:5,000) in PBS-T+5% skim milk for 1 hour at RT. After washing with PBS-T, immunoreactivity was detected by the enhanced chemiluminescence (ECL) method using the Clarity™ Western ECL Substrate according the manufacturer's suggestions.
mIL-10 Quantification by Enzyme Linked Immunosorbent Assay (ELISA).
Mouse IL-10 of concentrated supernatants and OMVs was quantified by sandwich ELISA using the mouse IL-10 DuoSet ELISA kit according to the manufacturer's instructions. Recombinant mouse IL-10 standard was 2-fold serially diluted in reagent diluent (1% BSA in PBS) to generate eight standards of known concentrations of 0-2,000 pg/mL. Concentrated supernatants and OMVs were diluted 1:5 to 1:20 in reagent diluent according to expected mIL-10 concentration.
A 96-well microplate was coated with 100 μL of the α-mIL-10 capture antibody, diluted in PBS as instructed, per well and incubated overnight at room temperature. Wells were washed with 1× wash buffer (0.05% Tween20 in PBS) three times. After wash buffer was removed completely, wells were blocked by 300 μL reagent diluent for at least 60 min at room temperature and washing steps were repeated. 100 μL of sample or standards in reagent diluent were added to each well and incubated for 120 min at room temperature. After three washing steps, 100 μL of biotinylated α-mIL-10 detection antibody, diluted in reagent diluent as instructed, were added to each well and incubated for 120 min at room temperature. Microplate was washed three times, 100 μL of the working dilution of Streptavidin-HRP was added to each well, and incubated for 20 min at room temperature in the dark. After three washing steps, 100 μL of substrate solution (1:1 mixture of color reagent A (H2O2) and color reagent B (Tetramethylbenzidine)) per well were incubated for 20 minutes at room temperature in the dark. 50 μL of stop solution (2 N H2SO4) was added per well and gently mixed. Absorbance at 450 and 540 nm in each well was immediately measured using a microplate reader. Wavelength correction was done by subtracting the readings at 540 nm from the readings at 450 nm. All standards and samples were assayed in duplicates. mIL-10 concentrations in samples were determined based on the standard curve of known mIL-10 concentrations. mIL-10 concentrations were normalized to total protein concentrations as determined by Bradford assay.
Proteinase K Protection Assay.
To determine if OMV tag-NL fusion proteins are exposed on the OMV surface, accessibility to proteolytic activity was tested. Suspensions of 5 μg of OMVs in PBS were treated with 0.1 mg/mL proteinase K for various times between 30 min and 24 h at 37° C. in the presence or absence of 1% SDS (for western blot analysis) or 1% Triton-X100 (for luciferase assay). Following the incubation, all samples were placed on ice and proteolysis was stopped by addition of 1 mM phenylmethanesulfonyl fluoride (PMSF) when analyzed by western blot. The effects of proteinase K and detergents treatments on OMV-tag-NL loaded OMVs were determined by luciferase assay or western blot.
Animal study protocols were approved by the MIT Animal Care and Use Committee. 8-10 weeks-old male C57BL/6 mice were housed in non-sterile conditions with access to irradiated mouse chow and autoclaved water. 10 days prior to gavage of Bacteroides spp., mice were administered ciprofloxacin HCl (0.625 g/L) in sugar-sweetened drinking water and treated with metronidazole (1 mg/kg) by oral gavage every 24 h for 7 days. After antibiotic medication was stopped for 2 days, engineered Bacteroides spp. (5×108 CFU in 0.1 mL of 20% sucrose) were administered by oral gavage. Groups of mice (n=5) received either B. thetaiotaomicron, B. fragilis, B. vulgatus, or no bacteria. On day 4 or 7 after bacterial gavage acute colitis was induced by administration of 3% (w/v) dextran sulfate sodium (DSS, molecular weight 40-50 kDa) in drinking water for 8 days. Simultaneously, IPTG (25 mM) was administered in drinking water to induce the expression of the recombinant protein. Mice belonging to the same group were co-housed for the duration of the experiment. Body weight was recorded daily for the duration of the experiment. At the end of the study, mice were sacrificed and colonic length was measured, colon samples were taken for histological analysis, and fecal pellets were collected.
CFU and Luminescence in Feces.
Feces were weighted, 1 mL PBS was added, and were homogenized in a TissueLyser for 2 min at 25 Hz using sterile steel beads. Mixture was spun (30 s, 500×g) to pellet large debris. 1:10 serial dilutions were plated on selective BHIS+Gm+Em agar plates in six technical replicates. Plates were incubated anaerobically for 48 h at 37° C. and colony-forming units (CFU) were count. For measurement of luciferase activity, 25 μL of feces homogenate was mixed with 25 μL NanoLuc working solution and luminescence was read in a hybrid microplate reader as described above.
Histological Analysis.
Colon tissue was fixed in 10% (w/v) formalin, paraffin-embedded, sectioned, and stained with haematoxylin & eosin (H&E). Three sections (proximal, mid, and distal colon) per animal were microscopically scored on a scale of 0-4 with 0.5 increments for the following 7 criteria: inflammatory infiltrates, edema, epithelial defects, extent of epithelial defects/crypt atrophy, epithelial hyperplasia, dysplasia/neoplasia, and area of dysplasia/neoplasia. The total colitis score is the sum of the 7 sub-scores. Samples were read by a pathologist blinded to the identity of the samples.
The presence of N-terminal signal peptides and SPII cleavage sites were predicted with LipoP 1.0 (cbs.dtu.dk/services/LipoP/) (144), SignalP 4.1 (cbs.dtu.dk/services/SignalP/) (145), SignalBLAST (sigpep.services.came.sbg.ac.at/signalblast.html) (147), and Phobius (phobius.sbc.su.se/) (146) using default settings for Gram negative bacteria. Only LipoP discriminates between signal peptides of secreted proteins and lipoproteins as well as N-terminal transmembrane helices in Gram-negative bacteria (144). SignalBLAST predicts signal peptides based on sequence alignment techniques.
The UniProt database (uniprot.org/) (148) was used to search for ‘lipoprotein’ in Bacteroides thetaiotaomicron (strain ATCC 29148/DSM 2079/NCTC 10582/E50/VPI-5482). Of the 63 obtained results, only proteins that possessed lipoprotein signal peptides predicted by LipoP and were not detected in OMVs by Elhenawy et al. (74) were used for further analysis.
Protein sequence alignments were carried out using ClustalOmega (ebi.ac.uk/Tools/msa/clustalo/) or T-Coffee (ebi.ac.uk/Tools/msa/tcoffee/) software.
Sequence logos of 4 aa before and 6 aa after the SPII cleavage site as predicted by LipoP were generated using the WebLogo 3 software (weblogo.threeplusone.com/create.cgi) (178).
Data were analyzed using the GraphPad Prism 6.01 software and are presented as mean values±standard deviation (SD). Statistical significance was analyzed by two-tailed Student's t-tests with 95% confidence interval. The level of significance was set to p<0.05 for all experiments.
All definitions, as defined and used herein, should be understood to control over dictionary definitions, definitions in documents incorporated by reference, and/or ordinary meanings of the defined terms.
All references, patents and patent applications disclosed herein are incorporated by reference with respect to the subject matter for which each is cited, which in some cases may encompass the entirety of the document.
The indefinite articles “a” and “an,” as used herein in the specification and in the claims, unless clearly indicated to the contrary, should be understood to mean “at least one.”
It should also be understood that, unless clearly indicated to the contrary, in any methods claimed herein that include more than one step or act, the order of the steps or acts of the method is not necessarily limited to the order in which the steps or acts of the method are recited.
In the claims, as well as in the specification above, all transitional phrases such as “comprising,” “including,” “carrying,” “having,” “containing,” “involving,” “holding,” “composed of,” and the like are to be understood to be open-ended, i.e., to mean including but not limited to. Only the transitional phrases “consisting of” and “consisting essentially of” shall be closed or semi-closed transitional phrases, respectively, as set forth in the United States Patent Office Manual of Patent Examining Procedures, Section 2111.03.
This application claims the benefit under 35 U.S.C. §119(e) of U.S. provisional application No. 62/257,849, filed Nov. 20, 2015, and U.S. provisional application No. 62/413,398, filed Oct. 26, 2016, each of which is incorporated by reference herein in its entirety.
This invention was made with government support under Grant No. OD008435 awarded by National Institutes of Health. The government has certain rights in the invention.
| Number | Date | Country | |
|---|---|---|---|
| 62413398 | Oct 2016 | US | |
| 62257849 | Nov 2015 | US |
