Prophylactic immunization to Helicobacter pylori infection using spore vectored vaccines

Abstract Background Helicobacter pylori infection remains a major public health threat leading to gastrointestinal illness and increased risk of gastric cancer. Mostly affecting populations in developing countries no vaccines are yet available and the disease is controlled by antimicrobials which, in turn, are driving the emergence of AMR. Materials and Methods We have engineered spores of Bacillus subtilis to display putative H. pylori protective antigens, urease subunit A (UreA) and subunit B (UreB) on the spore surface. Following oral dosing of mice with these spores, we evaluated immunity and colonization in animals challenged with H. pylori. Results Oral immunization with spores expressing either UreA or UreB showed antigen‐specific mucosal responses (fecal sIgA) including seroconversion and hyperimmunity. Following challenge, colonization by H. pylori was significantly reduced by up to 1‐log. Conclusions This study demonstrates the utility of bacterial spores for mucosal vaccination to H. pylori infection. The heat stability and robustness of Bacillus spores coupled with their existing use as probiotics make them an attractive solution for either protection against H. pylori infection or potentially for therapy and control of active infection.

eradicate the underlying infection. Humans appear to acquire H. pylori in two ways, oral-oral being the most common route of transmission and vertical transmission from mother to child. 2 H. pylori is unusual in that it is able to colonize the stomach where this highly motile bacterium can penetrate mucus, cause inflammation, and degrade the stomach lining. Incidence rates appear higher in developing countries where there is poor sanitation, particularly Asia/SE Asia and in some countries such as Vietnam >70% of the population are carriers with up to 10%-25% exhibiting symptoms. [4][5][6] Infections are treated using a combination of antibiotics 2,7 but often multidrug regimens are required often in combination with proton pump inhibitors lasting for up to 14 days. This approach has often discouraged patient compliance. However, the prevalence of AMR (antimicrobial resistance; notably to clarithromycin and metronidazole 8 ) is so high that many infected patients are now considered as having fully resistant infections 9 and in some cases are unable to access antibiotic therapy. 10 Clarithromycin has been used around the world in standard triple therapy but clarithromycin-resistant H. pylori isolates have been rapidly increasing worldwide, for example, in China from ~15% in 2000 to ~53% in 2014. 11 In 2017, clarithromycinresistant H. pylori was included in the WHO's list of 12 antibioticresistant "priority pathogens" that pose the greatest threat to human health. 12 Importantly, antimicrobial therapy cannot protect against reinfection, and the rate of reinfection is as high as 15%-30% per year. 13 The greatest burden of H. pylori infection is in developing countries (notably China and SE Asia) where there is a clear link to an increased risk of gastric cancer. It has been suggested that a 10-year vaccination program might significantly reduce the impact of H. pylori infection both with regard to symptoms, gastric cancer, and the associated economic burden of disease management. 14 Conceptually, a vaccine would best be administered orally to enable the production of secretory IgA (sIgA) in the stomach mucosa preventing colonization. [15][16][17] However, other mucosal delivery routes (intranasal, rectal) have been successfully used. 15 and other delivery systems have been evaluated. [15][16][17] One of major problems with oral immunization is that resulting immunity is weak.
Accordingly, adjuvants such as cholera toxin (CT), the closely related heat-labile toxin (LT) of E. coli or the B subunit of CT (CTB) have been extensively evaluated. 18,19 Although there has been considerable effort in vaccine development, few human studies have demonstrated convincing levels of protective immunity. 16,17 The one promising exception being a recently described vaccine consisting of an orally-administered protein formulation comprised of UreB fused to LT. 20 Despite this there is a case for vaccination where even reduced efficacy might shorten existing treatment regimens and help protect against reinfection. 17 Here, we have evaluated bacterial spore vaccines using UreA and UreB as putative protective antigens. Using oral delivery spore vaccines induced antigen-specific mucosal IgA and in a mouse colonization model a 1-log reduction in stomach colonization was observed.

| General methods
Methods for B. subtilis including preparation of spores and extraction of spore coat proteins are described elsewhere. 21

| Construction of B. subtilis spores expressing urease antigens
A cloning method referred to as THY-X-CISE ®22 was used to intro-

| Detection of surface display of antigens using enzyme-linked immunosorbent assay
The whole-spore enzyme-linked immunosorbent assay (ELISA) protocol was followed as described elsewhere. 26 Briefly, spores were diluted to 2 × 10 8 spores/mL in PBS, and 50 μL of suspension was used to coat microplate wells (Greiner, high binding) overnight at 4°C. This was followed by blocking for 1 h at 37°C with PBS containing 0.05%  Cells were cultured in 400 mL LB medium at 37°C until an A 600 of approx. 0.8 was attained, and then, the temperature reduced to 16°C for 20 min before inducing protein expression by adding IPTG F I G U R E 1 Amino acid sequences of the fusion genes. Chimeric genes inserted at the thyA loci (proximal (thyA P ) and distal (thyB D ) segments are indicated) of B. subtilis are shown together with a schematic of the chromosomal region (hypo = hypothetical gene). Urease A (ureA) and urease B (ureB, C-terminal region) were fused in frame to the 3′-end of the spore coat protein gene. The MW of the chimeric proteins are shown.

| Determination of mucosal titers by indirect enzyme-linked immunosorbent assay
For analysis of immunological responses, fecal samples were col- and ODs read at 450 nm. Dilution curves were created for each sample and endpoint titers estimated as the maximum dilution that gave an absorbance reading above the average naive sample.

| Ethics approval
Murine studies were conducted with approval from Royal Holloway

| Statistical analysis
Statistical significance was assessed by the Mann-Whitney U-test or the Dunnett's test using Prism (GraphPad, Dotmatics).

| Display of urease antigens on the spore coat of B. subtilis
The complete urease A protein and the carboxy-terminus of urease B of H. pylori were expressed on the surface of B. subtilis spores by inframe fusion of the relevant ureA and ureB coding ORFs to the B. subtilis cotB gene (Figure 1; n.b., we were unable to express the complete UreB protein on the spore surface). Both UreA and UreB have been shown to confer protection when delivered orally. [28][29][30][31] Here, however, we used a truncated urease B that lacked the amino-terminal enzymatic domain. 23,24 The cotB gene has been used repeatedly for expression of heterologous antigens on B. subtilis spores and enables stable presentation of chimeric polypeptides. Our method for cloning used the THY-X-CISE® system that places the chimeric genes at the thyA (thymidylate synthetase A) gene of the prototrophic B. subtilis strain PY79. 22 This was followed by insertional disruption of the thyB locus resulting in strains (PK82 cotB-ureA and PK78 cotB-ureB CT ) that are unable to grow in the absence of thymine (or thymidine). Using a congenic strain (PK118) that carried insertional disruption of both thyA and thyB but carried no chimeric genes two immunological methods were used to verify surface display. First, western blotting using PAbs that recognize UreA and UreB and second, whole spore ELISA ( Figure 2).
Whole spore ELISA clearly demonstrated recognition of UreA and UreB on spores with some cross-reaction to PK118 spores ( Figure 2C,D). It should be noted that a tricistronic urease operon (ure-ABC) is present in most strains of B. subtilis with ureC corresponding to the enzymatic subunit that in H. pylori is named ureB. 32 B. subtilis UreA shares some homology with H. pylori UreA (~31%) and with UreC about 75% homology with H. pylori UreB. The operon is transcribed during ordinary vegetative cell growth but only at high levels during nitrogen-limited growth. 32,33 Using a standard agar-based biochemical method (Christensen's slant agar 34 ), we have confirmed that all three strains (PK118, PK78, and PK82) do not produce functional urease (not shown). Although we produced crops of spores, we cannot, however, rule out the possibility of the presence of low levels of B. subtilis-produced urease being present (possibly adsorbed to spores) and possibly accounting for this cross-reaction.
Blotting of size-fractionated spore coat extracts for PK82 (cotB-ureA) revealed three bands (~40, [diffuse], 64 and 70 kDa.) that were absent in PK118 spores (Figure 2A,B). One of these bands was in F I G U R E 2 Spore coat expression of urease proteins. B. subtilis vaccine strains carrying insertions at the thyA loci were examined by western blotting of SDS-PAGE (12% w/v) size-fractionated spore coat proteins extracted from preparations of pure spores (approx. 2 × 10 9 spores/extraction). Blots were probed with PAbs as shown (panels A, B). Surface expression was also determined by whole spore enzymelinked immunosorbent assay (ELISA) using microtiter plates coated with spores at 1 × 10 8 CFU/well (panels C, D). Panel A shows blots of coat proteins extracted from PK82 (cotB-ureA) and the isogenic parent strain PK118 (WT).  Figure 2A; the other bands most likely being multimeric or breakdown species). Western blotting of PK78 (cotB-ureB CT ) was less clean with cross-reacting bands in PK118 spores. However, one abundant band of the correct size for CotB-UreB (~61 kDa) was clearly present in PK78 and absent in PK118 spores ( Figure 2B).
Since the cross-reacting bands are associated with the spore coat the most likely explanation is that of cross-recognition with B. subtilis spore coat proteins. However, we were unable to identify any spore coat protein that showed significant levels of amino acid homology with either urease A or B.

| Immune responses in mice dosed with spore vaccines
Mice were dosed orally (i.g.) four times with spores of PK82 (CotB- Responses for both PK78 and PK82-dosed animals, at maximum, were significantly (p = 0.0001) greater than in mice dosed with PK118 spores or the naive group. Very low levels of UreA-specific sIgA were observed in PK118-dosed mice but these were not statistically significant.
Serum IgG responses measured at Day 61 also showed that both PK78 and PK82 were able to induce systemic immunity ( Figure 4).
Taken together, oral administration of spores expressing either UreA or UreB CT on the spore surface can elicit both mucosal and systemic responses. animals also showed a significant reduction (~72%, median values; p = 0.0001) in CFU but less so than PK78 dosed animals (89%, median values, p < 0.0001). Interestingly, animals dosed with "naked" spores (PK118), that is, spores displaying no H. pylori antigens, also showed a reduction (~40%, median, p < 0.01) in H. pylori CFU compared to naive animals. In conclusion, both spore vaccines expressing either UreA or UreB CT were able to confer protective immunity sufficient to reduce H. pylori colonization in mice.

| DISCUSS ION
Spores of B. subtilis have been used extensively as mucosal vaccine vectors where oral (intra-gastric or sublingual) or nasal administration efficiently induces mucosal immunity (typically sIgA) as well as a Th1 bias. [35][36][37] Antigens are displayed on the spore surface (diameter ~ 1 μm) typically fused (as chimeric fusions) to proteins associated with the outermost layers of the spore coat. Bacillus spores are dormant yet able to germinate and outgrow under favorable conditions. They are also particularly robust being able to survive exposure to extremes of heat, desiccation as well as noxious compounds including gastric fluids. 38 Remarkably, and as a rule, chimeric spore expression does not normally lead to significant degradation of the exposed heterologous protein. Spores are members of the aerobiome and also found in soil and vegetation. 39,40 As such animals and humans are exposed to a low level of Bacillus on a daily basis. 41,42 Considered together, the use of bacterial spores as oral vaccine vehicles is compelling.
For H. pylori vaccination, we evaluated two "classical" anti- noted that IgG is also present in mucosal samples 46 and we did not assess levels of this immunoglobulin. UreB has previously been used for H. pylori vaccination utilizing spores for oral delivery but this has incorporated the entire UreB polypeptide fused to the CotC spore coat anchor. 28 conferred a low level of protection (40% reduction in gastric CFU). These spores do not evoke antigen-specific sIgA so the most probable explanation is that of innate immunity. Bacillus spores have been well documented as being able to evoke innate immunity and for some pathogens such as influenza this can be protective. [48][49][50] This has included murine studies showing reduced colonization by Clostridium difficile following oral dosing with "naked" spores. 35 We suspect that repeat dosing with B. subtilis spores may trigger an innate immune response sufficient to exert some level of protection. A second point is that Zhou et al. 28 also evaluated a trimeric fusion protein comprising a CotC anchor fused to CTB (cholera toxin subunit B) and UreB. This vaccine provided the highest reduction in gastric CFU of ~90% and was thus similar to our data found here for CotB-UreB CT . 28  elements of the immune system may be required to achieve full sterilizing immunity. 15,52 It is well documented that cell-mediated immunity plays an important role in protection against H. pylori infection. 15,16 Gastric biopsy samples from infected patients display an increase in CD4+ T cells, 52 and a bias of Th1 cells has been considered necessary for protection. 53 In addition, UreB has been shown to induce Th17 cells that, in turn, are responsible for the production of the proinflammatory cytokines, IL-17, IL-17F, and IL-22. 54 Oral administration of Bacillus spores has been shown in mice to interact with components of the cellular immune system, notably toll-like receptors (TLRs), with in vivo induction of proinflammatory cytokines (TNFα and IL-6). 55 Potentially, these phenotypes may be linked with the abovementioned innate response but we suspect that the immunostimulatory properties of spores alone may also be contributing to the inhibition of H. pylori colonization. Lastly, as humans are exposed to low levels of Bacillus on a daily basis future development of the spore platform must consider and address the issue of tolerance and suppression of the immune response. 56

| CON CLUS ION
This work has shown the potential utility of spores for prophylactic vaccination to H. pylori infection. The use of a system that ensures containment of genetically modified probiotic spores is a further advantage primarily because the use of GMOs in humans remains contentious and biological containment using the approach reported here is assured. A therapeutic application of a H. pylori spore vaccine is also worthy of consideration and is under current investigation.
Finally, the spore platform enables other potential H. pylori antigens to be evaluated and potentially a multivalent vaccine to be formulated. Such an approach might further boost levels of protection.

AUTH O R CO NTR I B UTI O N S
PMK, TLPN, TKCN, GM, and DMDB conducted experimental studies. VDN, GC, HAH, and SMC designed studies. SMC wrote the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.