Skip to main content

Vaccination against Shigella and ETEC: novel antigens, novel approaches

Final Report Summary - STOPENTERICS (Vaccination against Shigella and ETEC: novel antigens, novel approaches)

Executive Summary:
Recent epidemiological evidences have shown that in low-income countries, behind rotavirus that is on the way of being controlled by large implementation of now available efficacious vaccines, a limited number of other pathogens account for the rest of the diarrheal disease burden in children below the age of 5 years. Among these pathogens, Shigella and enterotoxigenic Escherichia coli (ETEC) dominate. This “tandem” hence justifies renewed efforts to develop vaccines which, in association to rotavirus vaccination, would largely alleviate the disease burden of severe diarrhea that represents the second cause of child mortality, just behind acute infections of the lower respiratory tract, notwithstanding morbidity and the current trend to chronicity engaging a vicious circle of paediatric malnutrition.

Unfortunately, vaccine development in the field of Shigella and ETEC has not allowed so far the development of a convincing vaccine. Candidates were largely based on killed or live-attenuated strains administered orally, assuming that direct contact of the vaccine with the gut mucosal surface – the actual site of infection - would induce strong protective immunity. For many reasons, this has not been the case. At the same time, the development of polysaccharide conjugate vaccines administered parenterally provided the proof of concept of strong protective capacity against respiratory pathogens, including control of mucosal colonization.

STOPENTERICS was established as a way to validate this paradigm in the context of mucosal infection by Shigella and ETEC. STOPENTERICS was therefore established as a consortium of European, African and Asian groups with the aim of developing a novel paradigm encompassing:
-Introduction of the parenteral route of immunisation, at least for Shigella.
-Identification of Shigella and ETEC antigens, essentially proteins, preferably playing a role in pathogenesis that could help breaking the dogma of serotype-specificity of immune protection, i.e. polysaccharide side chains (O-Ag) of lipopolysaccharide/LPS in Shigella (five are prevalent: S.flexneri 1b, 2a, 3a, 6 and S.sonnei) and surface proteinaceous colonization factors in ETEC (numerous). In the latter, emphasis was put on ST toxin, a guanylate cyclase agonist that is now epidemiologically strongly associated with pathogenic ETEC isolates.

Advanced approaches were used to achieve this goal:
*Preparation/standardization of the production of vesicular outer membrane blebs (GMMA) and chemical synthesis of oligosaccharides, as motifs of the O-Ag, coupled, by conjugaison, to a protein carrier for Shigella vaccine development.
*Immunoproteomics and comparative genomics to identify relevant cross-protective proteins in Shigella and ETEC.
*Chemistry and molecular/structural biology to functionally attenuate and boost the immunogenicity of ETEC ST toxin.
*Development and/or adaptation of novel immunomonitoring assays, some of which (i.e. B-cell memory) were used in the course of the Shigella vaccine trial in Israel.

In summary, one can split the essential contributions of STOPENTERICS in four groups:

1-Development and validation in phase 1 clinical trials of two Shigella vaccine candidates:
*GMMA that bears a possible degree of cross-protection and offered promising results in the phase 1 study conducted in Paris.
*SF2aTT15 (synthetic carbohydrate-based conjugate) that showed outstanding immunogenicity in volunteers in a phase 1 study conducted in Israel.
Results of these two trials quickly became quite visible to the community and led the Bill and Melinda Gates Foundation (BMGF) to bring its support to further development of these candidates. In a sense, the essential goals of the FP7 call were fulfilled, since we could complete not only one, but two successful trials.

2-Development and implementation of chemical synthesis to the most relevant Shigella serotypes and consolidation of a pipeline of production, conjugation to carrier at GMP level.

3-Identification of a certain number of Shigella and ETEC surface proteins that could offer options for cross-protection, but clear difficulties to implement them in vaccine formulations at this stage.

4-Major progress made in combining rational attenuation and “antigenization” of ETEC ST toxin, with clear indications that the approach is feasible and may provide the expected breakthrough in ETEC vaccination in the future.

In brief, we consider that, at the cost of a necessary two-year extension, STOPENTERICS was a successful endeavour that fulfilled its major goals.
Project Context and Objectives:
STOPENTERICS aims are to develop novel vaccine candidates against Shigella and ETEC for children of the developing world. The consortium gathers a unique combination of laboratories, platforms, vaccinology centers from academia and industry in the North and in endemic areas. STOPENTERICS proposes novel solutions by imposing a paradigm switch (i) to break the dogma of serotype-specificity by seeking to induce cross-protective immunity. It exploits the potential of genomics, transcriptomics and proteomics to identify surface and virulence proteins conserved throughout Shigella or ETEC isolates. For ETEC, the development of a safe, immunogenic ST toxoid is also a major goal. For Shigella, the vaccine potential of bacterial outer membrane blebs (GMMA) is exploited; (ii) another approach is to improve the immunogenicity of Shigella glycoconjugates by using synthetic oligosaccharides mimicking the O-Ag moiety of the LPS for the key serotypes.

The ultimate aim is to optimize chances for best coverage by combining - possibly in a single formulation - cross-protective protein antigens and the most prevalent serotype-defining synthetic oligosaccharides. STOPENTERICS has developed two platforms, one dedicated to the identification and production of novel protein antigens, and one dedicated to the evaluation of the protective capacities of candidate antigens in animal models. However, the predictive value of these models shows limitations, due to the human-specificity of both pathogens, thus a strong incentive to also push the new vaccine candidates through clinical evaluation as soon as possible.

A key issue is also to define the optimal route of immunization. STOPENTERICS aims at parenterally-administered formulations. A significant effort is therefore invested in optimizing the delivery route and the associated adjuvant in order to obtain maximum titers of specific serum IgG combined with significant titers of mucosal sIgA. Particular attention will also be paid to the elicitation of strong T cell responses and B/T cell memory.

Project Results:
Work Package 1: “Clinical trials”
WP Leader: Armelle PHALIPON (IP) & Dani COHEN (TAU) [Beneficiary 1 – Institut Pasteur (IP)]

Activities of the WP1 included the successful completion of the manufacturing stages and "first in man" clinical and immunological evaluation of two novel subunit vaccines against shigellosis, namely the GMMA S. sonnei and SF2a-TT15 synthetic carbohydrate-based vaccines against S. flexneri serotype 2a (SF2a). Here is a brief report of the two phase 1 clinical trials.

GMMA: The GSK Vaccines Institute for Global Health (GVGH), former NGVH, developed a candidate Shigella sonnei vaccine (1790GAHB) using the Generalized Modules for Membrane Antigens (GMMA) approach. To evaluate the safety and immunogenicity of 1790GAHB, two parallel, phase 1, observer-blind, randomized, placebo-controlled, dose escalation studies were conducted in France ("study 1", funded by STOPENTERICS) and in the United Kingdom ("study 2") between February 2014 and April 2015 in 18-45years old volunteers (50 in study 1, 52 in study 2). Increasing doses of Alhydrogel adsorbed 1790AHB (respectively characterized by O Antigen (OAg) and protein concentrations), or placebo were given either by intramuscular route (0.059/1 0.29/5 1.5/25 2.9/50 5.9/100μg of OAg/μg of protein; study 1) or by intradermal (ID), intranasal (IN) or intramuscular (IM) route of immunization (0.0059/0.1 0.059/1 0.59/10μg ID, 0.29/5 1.2/20 4.8/80μg IN and 0.29/5μg IM, respectively; study 2). Vaccine induced immunogenicity was compared to anti-LPS antibody in a population naturally exposed to S. sonnei. 100% of vaccinees were responders after the first dose, with 4-fold or greater rise in titer. A parallel significant response was also found for IgA and IgG ASC and ALS to OAg, 7 days after the first injection. The vaccine induced an average of 15 fold increase in serum bactericidal antibody GMTs following immunization. All vaccinees had a significant increase in the percent of IgG memory B cells specific to LPS. Vaccines were well tolerated in both studies and no vaccine-related serious adverse events were reported. In study 1, doses ≥1.5/25μg elicited serum IgG median antibody greater than median level in convalescent subjects after the first dose. No vaccine group in study 2 achieved median antibody greater than the median convalescent antibody.
Conclusion: Intramuscularly administered Shigella sonnei GMMA vaccine is well tolerated, up to and including 5.9/100μg and induces antibody to the OAg of at least the same magnitude of those observed following natural exposure to the pathogen. Vaccine administered by ID or IN, although well tolerated, is poorly immunogenic at the doses delivered.
A first-in-human, single-blinded, observer-masked randomized, dose escalation (2 different doses of the sugar component), placebo-controlled study was conducted in healthy Israeli volunteers. Volunteers were selected from those who were pre-screened for SF2a LPS IgG serum antibody levels below 80 percentile and for HLA-B27 negativity. Sixty-four eligible subjects were assigned to one of two cohorts and randomized to receive the lower dose of 2 μg (with or without alum adjuvant) or matching placebo (cohort 1) and the higher dose of 10 μg (with or without alum adjuvant) or matching placebo (cohort 2). The study agents were administered by 3 intramuscular injections on days 0, 28 and 56. There were 9 follow-up visits for safety and immunogenicity with the last visit 3 month after dose 3. Immunogenicity was assessed using the parameters established and employed in the study of the immune response after natural Shigella infection.

SF2a-TT15 was shown to be safe and well-tolerated with no severe or serious adverse events reported. All possibly vaccine-related adverse events were classified mild except one classified as moderate (nausea).

At 2 μg, the first vaccine injection induced a significant rise in serum IgG anti-SF2a LPS GMTs of approximate 5 fold as compared with pre-vaccination or placebo recipients' levels. Alum enhanced the response after the second and third injection to 8.5 and 12 fold increases in GMT, respectively.
At 10 μg, the first injection of SF2a-TT15 elicited IgG anti-SF2a LPS rises in GMTs of 25 and 27 fold with or without alum, respectively, as compared with pre-vaccination levels, with no further increase after the second and third dose. All the vaccinees (100%) were responders after the first dose, with 4-fold or greater rise in titer. A parallel significant response was also found for IgA and IgG ASC and ALS to S. flexneri 2a LPS, 7 days after the first injection. The vaccine induced an average of 15 fold increase in serum bactericidal antibody GMTs following immunization (range of post-vaccination GMTs: 5918-14669) with a 4-fold or greater rise in serum bactericidal titers in 88% of the volunteers after the first injection. The SBA and avidity correlations with the levels serum IgG antibodies to S. flexneri 2a LPS indicate functional capabilities of these antibodies. Noteworthy, in the presence of alum, the anti-SF2a LPS IgG1 response is higher as compared to the non-adjuvanted vaccine. Urinary IgA levels were shown to be higher in the vaccines as compared to the placebo control, with no difference between the adjuvanted and non-adjuvanted vaccine. All vaccinees had a significant increase in the percent of IgG memory B cells specific to S. flexneri 2a LPS.
Measurement of T cell responses using multiparametric analysis of T-cells stimulated by the vaccine was performed at Institut Pasteur. Vaccine-stimulated T cell response, their effector functions and their polyfunctionality were analyzed at 3 time-points for each group. T-cell responses and polyfunctionality were measured in some volunteers, suggesting that the SF2a-TT15 vaccine primes for a specific T cell response. Measurement of specific T cell response on CD4+ and CD8+ T cell populations instead of global CD3+ T cells is underway. Vaccine-specific B-cell memory will also be analyzed in depth by multiparametric flow cytometry at IP on PBMC sample collections prepared by TAU.

Conclusions: The findings of the phase 1 study of the GMMA S. sonnei vaccine prototype and the interim data on the safety and immunogenicity of the SF2a-TT15 synthetic carbohydrate-based vaccine carried out in frame of STOPENTERICS, supported the inclusion of these vaccine candidates in the BMGF portfolio of accelerated Shigella vaccine development with corresponding prospective funding. The two monovalent prototype vaccines will be evaluated in CHIM studies for preliminary efficacy and tetravalent vaccines incorporating these SF2a and S. sonnei candidate vaccines are under development. Age-descending studies will be designed for further evaluation of both the monovalent and multivalent conjugate vaccines in low income countries. There will be also a long term immunological follow up, one and two years after the administration of the monovalent SF2a-TT15 aiming to examine the persistence of serum antibodies and B memory cell immunity. It is hoped that the immunogenicity induced by SF2a-TT15 in naïve adult volunteers will be maintained at high level when they will be given to very young children (less than 3 years old), thus improving the performance of the first generation of Shigella conjugates. The immunogenicity and further preliminary efficacy data generated by the phase 1, 2a and 2b studies of these vaccines will be bridged to those generated by phase 2 and 3 studies of the first generation of Shigella conjugate vaccines in children and young adults. It is hoped that all together these data will pave the way for pre-qualification by the WHO of at least one Shigella vaccine within the next 5-8 years.

Work Package 2: “Immunomonitoring and correlates of protection”
WP Leader: Dani COHEN (TAU) & A-M SVENNERHOLM (UGOT) [Beneficiary 9 – Tel Aviv University (TAU)]

Task 1: Developing innovative tools for systemic and mucosal immunomonitoring of Shigella parenteral candidate vaccines (Leader: D. Cohen, TAU)

Epidemiological Background:

We studied the epidemiological background of shigellosis in Israel using a sentinel laboratory-based surveillance of Shigella infections. We found and characterized cyclic biennal country-wide epidemics of S. sonnei shigellosis with the epicenter in Jewish orthodox crowded communities with households with a relative large number of children under 5 (1). In collaboration with Dr. Nick Thompson from the Sanger Institute and WP5 participants we used next generation sequencing of more than 300 S. sonnei isolates to understand the microevolution and transmission patterns within these domestic epidemic cycles and also internationally spread among similar communities (2,3, 4). Using archived isolates we also examined potential pressure of immunization with a potent Shigella conjugate vaccine on changes in the S. sonnei genome and particularly in the genes on the large invasiveness plasmid of 140Mdal encoding for S. sonnei LPS, the protective antigen (5).

We took a step further our previous observations on a distinctive force of infection of S. sonnei and S. flexneri and the assumption on potential differences in the infectious dose of the two species which may explain local and global trends in the epidemiology of shigellosis. To explore a potential link with acid-resistance traits we carried out a series of experiments and examined this parameter on representative strains of S. sonnei. S. flexneri 2a, S. flexneri 6 and Salmonella Newport (control). We found and documented for the first time significant inter-Shigella serogroup and serotype differences in the extent of acid-resistance which can explain the differences in the force of infection and recent trends of S. sonnei versus S. flexneri shigellosis worldwide. (6).

Study of the immune response to Shigella natural infection:

On top of the epidemiological and logistic infrastructure we established a network of hospitals and community clinics which made possible collection of clinical data and samples (stools, blood and urine) from children with culture-proven S. sonnei and S. flexneri shigellosis at different stages after infection and "healthy" controls. It has been shown that natural infection provides serotype specific protection against shigellosis at least for a short period of time and therefore we assumed that the study of the immune response induced by natural infection has to provide reference data for a similar or even stronger immune response of same parameters induced by a successful Shigella candidate vaccine.

We established a variety of quantitative and qualitative assays to examine a wide range of humoral and cellular parameters first following natural documented Shigella infection and later on, after immunization with the investigational Shigella flexneri 2a-TT15 conjugate vaccine (7).
Serum and peripheral blood mononuclear cells (PBMC) obtained from 147 subjects (children aged 0.5-17 years) were separated and processed. Eighty urine samples were also collected from cases of shigellosis and healthy controls.

Serum antibody levels to Shigella antigens (homologous and heterologous LPSs and ipa B and ipa D antigens) were measured by ELISA in the same samples. We re-documented the significant and specific response to S. sonnei and flexneri 2a LPSs and the crossreactivity of convalescent sera of patients with culture proven shigellosis when tested against ipaB and ipaD (strong response to ipaB and much weaker to ipaD). We reported in the past that urine antibodies against Shigella LPS in young adults are of mucosal origin (Cohen D et al. 1996). We measured urine IgA antibody levels to Shigella antigens in children using the same ELISA protocol. We found that 60% of children with S. sonnei shigellosis in early convalescence showed a significant response to the homologous LPS (levels above the GMT of controls plus 2 SD). We did not detect significant differences in the urine IgA antibody levels between acute (n=10) and late convalescence (n=35) stages, suggesting that the urine IgA anti-LPS response in children is of short duration. There was a good correlation between the urine IgA response and IgA B memory cell response (r=0.541 p<0.0001) and serum IgA response (r=0.621 p<0.008) to Shigella LPS (7).

We collaborated with Dr. Chris Gerke (WP1) on equivalenting serum IgG measurements according to protocols employed in Siena and in Tel Aviv. A reference titer of IgG anti-S. sonnei LPS was created to assess the immune response of the GMMA S. sonnei candidate vaccine ( 8).

We found a diverse susceptibility of S. sonnei phase 1 strains to serum killing with a clear and significant correlation with the level of S. sonnei LPS antibodies (r=0.495; p=0.014) in archived samples of young adults. Twenty eight percent and 46% of sera of children with S. sonnei shigellosis displayed bactericidal activity at early and late convalescence, respectively, showing also significant correlation with serum LPS antibodies. Interestingly, sera obtained from adult convalescent cases of S. sonnei shigellosis showed more frequently bactericidal anti-S. sonnei activity as compared with sera of children (75% vs 39,1%, p=0.02). Using the thiocyanate elution assay, we found that sera with bactericidal activity (n=17) had a significantly higher avidity index than sera with no bactericidal activity(n=9) against S. sonnei (Avidity index: 2.267 vs 1.76 p=0.048).

The ELISPOT method to measure B memory cell response (BM) as described by Crotty et al. for vaccinia (2004) was adapted at TAU to Shigella antigens and used for the first time to measure IgA and IgG BM response to Shigella sonnei and flexneri LPS and ipaB following natural Shigella infection. The IgA BM response to shigella homologous LPS was the highest at early convalescence, 18 days to 2.5 months after onset of disease. Nine of 12 subjects examined (75% with a GMT of 76 BM/106PBMC) had a significant IgA BM response which was still significantly persistent in 32% of children (GMT of 54 BM/106PBMC) at late convalescence, around 27 months after onset of disease. Only 4 (12%) of 33 controls (healthy children with no known previous exposure to Shigella) showed any B memory cell response to Shigella antigens. We found significant correlations between the IgA BM cell response and both IgA and IgG serum response to homologous LPS in children with S. sonnei and S. flexneri 2a shigellosis (n=84, Pearson corr.=0.723 P<0.01 for IgA and Pearson corr.=0.762 P< 0.01 for IgG). These data fit the epidemiological observations on the limited length of natural immunity conferred by natural Shigella infection.

Fresh PBMCs were stimulated in vitro with S. sonnei and S. flexneri 2a LPS, ipaB, ipaD and control antigens. The supernatants were semi-quantitatively screened for presence of 36 different cytokines by a dot-blot technology using the Proteome ProfilerTM- Cytokine ArrayR&D SYSTEMS® . Selected cytokines were further quantitatively analyzed using an image analysis software and by ELISA. In addition, the same PBMCs samples were also analyzed for gene expression of selected cytokines (IFN-γ, IL-6) after RNA extraction, reverse transcription and qPCR. Results have been compared and analyzed versus the time elapsed after the documented exposure.

Selected frozen supernatants from 8 subjects acutely and early infected with S. sonnei, 26 subjects late convalescent from S. sonnei infection and 11 healthy donors (HD) were sent to Institut Pasteur (Prof. Marie-Lise Goujeon) for parallel testing by the multianalyte profiling technology. These samples included supernatants from PBMC stimulated for 24h with various stimuli (LPS from S. sonnei , S. flexneri , IpaB, IpaD, or PHA). In HD, stimulation with LPS from S. sonnei or S. flexneri, IpaB , IpaD did not induce any cytokine/chemokine response, while the polyclonal PHA stimulation induced, as expected, a proinflammatory pattern. When specific response to LPS from S. sonnei were analyzed, in samples from infected sujects, an increase in the production of IL-2, IL-12, IL-17, IL-1ß, TNF-alpha and IL-6 was detected. This pattern is of particular interest as it corresponds to cytokines that have been shown to contribute to dendritic cell (DC) recruitment and activation in murine mucosal Shigella infection, and promoting Th17 and Th1 responses in the mucosa. These results identify for the first time correlates of mucosal immune response to natural Shigella infection in humans, and they confirm the triggering of mucosal human Th17 response that was previously detected at IP in volunteers immunized with live oral SD1 599 vacccine.

In collaboration with Dr. Lina Bernardini (WP6) we continued to measure pentraxin 3 (PTX3) levels in plasma of acute cases of shigellosis (n=55) and controls (n=77). PTX3 is synthesized in macrophages, DCs, PMNs, epithelial cells, fibroblasts etc. at the local site following exposure of the cells to LPS. PTX3 is a component of the humoral arm of innate immunity and acts similar to antibodies with the abilities of complement activation, opsonization, and regulation of inflammation. We found a significant association between presence and levels of PTX3 and early acute shigellosis and a significant trend of "dose-response" relationship with markers of severity of disease. No such associations were found with C reactive protein (short pentraxin) levels.

At NMIMR in Ghana, demographic, clinical data and samples were collected from 106 subjects with symptoms typical of dysentery of ages 0-5 (19%), 6-19 (16%) and 20-60 (65%). Paired samples were obtained from 45 of the cases of disease at the acute stage and 2-9 months later. In studies carried out in collaboration with the TAU group, of 80 stool specimens of patients with dysentery and tested by direct ipaH PCR, 32 (40%) were found positive for Shigella spp. or enteroinvasive E. coli. All positives were confirmed by sequencing of the amplicons. There was a significant serum IgA antibody response to S. sonnei, S flexneri 2a and S. flexneri 6 in 27%, 295 and 18%, respectively in the 45 paired sera tested. Forty two percent of the subjects had a signicant IgA response to at least one of the 3 LPSs. PCR detection rates of ipaH and serological data indicate that there is much exposure to Shigella and many of the cases of diarrhea/dysentery suffered most probably from shigellosis. Intensive training should be invested to make possible the isolation and characterization of shigellae from stool cultures.

SF2a-TT15 phase 1 clinical trial: investigational immune parameter measurements

Multiparametric analysis of T-cells stimulated by SF2a-TT15 was performed at Institut Pasteur combining antibodies specific for T-cell subsets, T-cell maturation markers and cytokines detected intracellularly, including IL-2, IFN-gammaγ and TNF-alpha. Vaccine-stimulated T cell response, their effector functions and their polyfunctionality were analyzed at 3 time-points irrespective of the groups to which volunteers belong. Strong T-cell responses and polyfunctionality were detected in some volunteers, increasing from baseline, suggesting that the SF2a-TT15 vaccine is priming T cell responses.
Further studies will be performed on additional humoral parameters, to identify the part of LPS recognized by T cells, the pattern of T-cell response in each group, and the correlates of immune response to SF2a-TT15 vaccine, by combining humoral and cellular responses. Vaccine-specific B-cell memory will be analyzed in depth by multiparametric flow cytometry at IP on PBMC sample collections prepared by TAU.

Towards the performance of the phase I clinical trial-related assays at TAU Research Lab (in frame of the STOPENTERICS WP1) a systematic expert training and guiding on establishment and implementation of a quality management system including quality assurance and quality control activities was carried out. The establishment of the quality management system was based on EMA reflection paper – "Reflection paper for laboratories that perform the analysis or evaluation of clinical trial samples" 2012

1. Cohen D, Bassal R, Goren S, Rouach T, Taran D, Schemberg B, Peled N, Keness Y, Ken-Dror S, Vasilev V, Nissan I, Agmon V, Shohat T. Recent trends in the epidemiology of shigellosis in Israel. Epidemiol Infect. 2014;142(12):2583-94.

2. Baker KS, Dallman TJ, Behar A, Weill FX, Gouali M, Sobel J, Fookes M, Valinsky L, Gal-Mor O, Connor TR, Nissan I, Bertrand S, Parkhill J, Jenkins C, Cohen D, Thomson NR. Travel- and Community-Based Transmission of Multidrug-Resistant Shigella sonnei Lineage among International Orthodox Jewish Communities. Emerg Infect Dis. 2016;22(9):1545-53.

3. Behar A, Baker KS, Bassal R, Valinsky L, Thomson NR, Cohen D. Microevolution and patterns of Shigella sonnei transmission within cyclic outbreaks of shigellosis in Israel. Israel Society of Microbiology Meeting. Vulcani Institute, Israel. 6 April 2017.

4. Cohen D, Behar A, Baker K, Korin H, Bassal R, Markovich M, Goren S, Thompson N. , Muhsen K Burden, risk factors and patterns of Shigella sonnei transmission in hyperendemic communities in a high-income country. Vaccines for Enteric Diseases Conference, Albofeira, Portugal, 9–11 October 2017.

5. Behar A, Fookes MC , Goren S, Thomson NR, Cohen D. Whole genome analysis to detect potential vaccine-induced changes on Shigella sonnei genome. Vaccine. 2015;33(26):2978-83.

6. Cohen D, Meron-Sudai S, Bialik A, Goren S, Hochberg A, Rubinstein U, Levy O, Bassal R, Ashkenazi S . Agent and host-related factors supporting the global emergence of S. sonnei shigellosis. Vaccines for Enteric Diseases Edinburgh, Scotland, 8–10 July 2015

7. Cohen D, Meron-Sudai S, Bialik A, Dorman A, Goren S, Hochberg A, Rubinstein U, Ashkenazi S. Characterization of the immune response to Shigella following natural infection towards identification of correlates of protection. Vaccines Against Shigella and ETEC (VASE) Conference. Washington DC, USA. 8-10 June 2016.

8. Launay O, Lewis DJM, Anemona A, Loulergue P, Leahy J, Sciré AS, Maugard A,
Marchetti E, Zancan S, Huo Z, Rondini S, Marhaba R, Finco O, Martin LB, Auerbach
J, Cohen D, Saul A, Gerke C, Podda A. Safety Profile and Immunologic Responses of a Novel Vaccine Against Shigella sonnei Administered Intramuscularly, Intradermally and Intranasally: Results From Two Parallel Randomized Phase 1 Clinical Studies in Healthy Adult Volunteers in Europe. EBioMedicine. 2017 Aug;22:164-172. doi: 10.1016/j.ebiom.2017.07.013. Epub 2017 Jul 15. PubMed PMID: 28735965.

9. Cohen D, Atsmon J, Artaud C, Meron-Sudai S, Gougeon M-L, Bialik A, Goren S, Asato V, Ariel-Cohen O, Reizis A, Dorman A, Volokhov I, Shaikevich D, Hoitink C.W.G Westdijk J, Ashkenazi S, Sansonetti P, Mulard LA, Phalipon A. A Phase I Dose Escalation Study to Assess the Safety and Immunogenicity of SF2a-TT15, a Synthetic Carbohydrate-Based Conjugate Vaccine against S. flexneri 2a in Healthy Adult Volunteers (Preliminary Results). Vaccines for Enteric Diseases Conference, Albofeira, Portugal, 9–11 October 2017.

Task 2: Developing innovative tools for mucosal and systemic immunomonitoring of ETEC oral candidate vaccines and infection; UGOT and icddrb (Leader: A-M Svennerholm, UGOT)

Considerable work towards the establishment of optimal methods for assessing in particular mucosally derived immune responses to oral vaccines, with a focus on ETEC vaccines, has been undertaken. For this purpose, samples collected from subjects participating in different Phase I trials of the oral inactivated multivalent ETEC vaccine (ETVAX) have been used. The vaccine, consisting of four inactivated recombinant E. coli strains overexpressing the major ETEC colonization factors (CFs) CFA/I, CS3, CS5 and CS6 mixed with an LT B-subunit related toxoid, LCTBA, has been shown to be broadly immunogenic when analyzed in different assays (1). Thus, ETVAX has, as tested by different assays modified in our laboratories, induced mucosal immune responses, i.e. IgA antibody secreting cell (ASC) responses as determined by the antibody in lymphocyte secretion (ALS) assay and fecal secretory IgA responses against all the major CFs and LTB in adults in Sweden (1).

A clinical Phase I/II trial of ETVAX in descending age groups (adults, toddlers, children 1.2 years and infants 6-12 months) in Bangladeshis is in progress. To identify optimal sampling time points of mucosa-derived samples in this trial, the kinetics of ALS responses after vaccination with a model vaccine (the oral cholera vaccine Dukoral) was assessed in Bangladeshi adults, toddlers and infants. Peak responses to cholera B subunit (antigen in the vaccine) were detected 5 days after the second immunization in all age groups (2), consistent with the kinetics of ALS responses observed in Swedish adults (1). The studies in Bangladesh also revealed that infants mounted significant fecal SIgA responses to CTB and that these responses were not masked by potential breast milk antibodies in feces (2). Recently, we have established a chemiluminescence assay using the Mesoscale discovery (MSD) platform, which has been shown to allow determination of immune responses in very small amounts of clinical specimens (eg samples collected from infants) for analyses of ETEC immune responses against vaccines and or/infection (Ahktar, Lundgren et al, to be published). The findings from the studies described above were used to establish the methods and sampling protocols for the ongoing Phase I/II trial in Bangladesh (Qadri et al).

Methods have also been developed allowing studies of functional mucosal IgA memory responses to ETEC vaccines. This has been shown by detection of significantly higher and earlier appearing intestine derived ALS responses to a single vaccine booster dose given to subjects who have been given ETEC vaccine 1-several years earlier compared to naïve individuals (3, 4). Studies have also been conducted in Bangladesh showing the appearance of antigen-specific memory B cells after ETEC infection (5)

To further evaluate the functional characteristics of antibody responses induced by ETVAX, methods have been established to evaluate if the vaccine can induce immunologically cross-reactive and high avidity antibody responses (6). Cross-reactive immune responses to CFs have been analyzed in mucosal (fecal and ALS) samples, and antibody avidity in serum and ALS samples, from both primary (two dose 2 weeks apart) and booster (after 1-2 years) vaccinated volunteers. These studies have shown that 60-90% of subjects who had previously responded to CFA/I in ALS or fecal specimens also developed cross-reactive antibodies to related CFs, i.e. CS1, CS14 and CS17, and that 70-90% of those responding to CS5 also responded to the closely related CS7. For subjects who had developed cross-reactive antibodies, the magnitudes of responses against vaccine CFs and related non-vaccine CFs were comparable. These results have indicate potential expanded ETEC strain coverage by ETVAX (6)

To enable measurement of antibody avidity, in particular for use on small volumes of mucosal samples, we have developed a simple method based on limiting antigen dilution. The results from this method correlated significantly with a more traditional chaotropic (KCN elution) ELISA avidity method, as well as with results from a functional toxin neutralization assay. Using limiting antigen dilution, we have shown that the avidity of serum and ALS antibodies to key vaccine antigens increased after a late booster dose compared to after a primary vaccination with ETVAX. (6)
To advance the knowledge on putative mechanisms supporting antibody production, and to evaluate potential new biomarkers of vaccine take and immunological memory, methods for analysis of different types of T helper cell responses to ETVAX have also been established. This was done by establishing a highly efficient protocol for freezing and thawing peripheral blood mononuclear cells (PBMCs) and following stimulation of the cryopreserved PBMCs with CFs or LTB determine the cytokine levels in culture supernatants by ELISA and the highly sensitive MSD approach. Using these methods, we have shown significantly higher production of the cytokines IFN-γ and IL-17A, but little IL-4, IL-5 or IL-13, indicating that ETVAX may induce both Th1 and Th17 type T cell responses (7).

A new protocol for flow cytometric detection of activated circulating T follicular helper (TFH) cells has also been established; such cells may provide essential support for B cell maturation and antibody production in germinal centers. Using this protocol, increased frequencies of CD4+ T cells with an activated TFH phenotype have been detected in blood after administration of ETVAX. A large proportion of the activated TFH cells were shown to express the gut homing marker β7 integrin, indicating that they were migrating to mucosal sites. Importantly, the magnitudes of the blood TFH responses to primary vaccination correlated significantly with the magnitudes of memory responses to ETVAX induced by booster vaccination, suggesting that TFH responses in peripheral blood may be used as marker for memory induction after mucosal vaccination (8).

Methods for assessing gut homing ASCs by magnetic bead selection of cells expressing gut-homing markers have also been established using samples from patients infected with ETEC. These studies have shown strongly elevated levels of β7 positive IgA producing cells in the circulation 7 days after onset of ETEC diarrhea and that 0.8-4.6 % of those cells were specific for the CF of the infecting strain. (9, 10),
Studies have also been undertaken to determine mucosal and systemic immune responses to ETEC natural infection including to evaluate ETEC specific memory B and memory T cell responses. Patients with ETEC diarrhoea have been enrolled and plasma and ALS specimens collected on days 2, 7 and 30 after onset of disease have been analyzed for immune responses to homologous CFs and LT; fecal extracts from the patients have been prepared and stored for further analyses (unpubl).
In conclusion the studies conducted in collaboration between UGOT and icddrb have resulted in establishment of a number of innovative tools for assessing B and T cell responses, including immunological memory, which can be used to determine immune responses against ETEC vaccines and ETEC infection in ongoing and also future studies. .

Collaboration and interactions among partners of WP2 and with other STOPENTERICS WPs
• Dr Svennerholm (UGOT) has extensive collaboration with Dr Qadri et al at icddr,b regarding collection of clinical specimens from ETEC patients, exchange of methods, ETEC epidemiological studies, clinical vaccine trials etc.
• Dr A Lundgren (UGOT) and Dr T Bhuyian (icddrb) have conducted several collaborative studies on T cell responses to oral enteric vaccines.
• One PhD student, M Akthar, works on collaborative projects at icddr,b and UGOT, as part of the STOPENTERICS project, under the supervision of Drs Qadri, Lundgren and Bhuiyan.
• ETEC strains and SOPs on immunoassays have been provided to other STOPENTERICS members.

1) Lundgren A, Bourgeois L, Carlin N, Clements J, Gustafsson B, Hartford M, Holmgren J, Petzold M, Walker R and Svennerholm AM: Safefety and immunogenicity of an improved oral inactivated multivalent enterotoxigenic Escherichia coli (ETEC) vaccine administered alone and together with dmLT adjuvant in a double-blind, randomized, placebo-controlled Phase I study. Vaccine 32(52):7077-7084, 2014.

2) Akhtar M, Qadri F, Bhuiyan TR, Akter S, Rafique TA, Khan A, et al. Kinetics of antibody-secreting cell and fecal IgA responses after oral cholera vaccination in different age groups in a cholera endemic country. Vaccine 2017 Jan 5;35(2):321-8

3) Susannah Leach, Anna Lundgren and Ann-Mari Svennerholm: Different kinetics of circulating antibody-secreting cell responses after primary and booster oral immunizations; a tool for assessing immunological memory.Vaccine 31: 3035-3038, 2013

4) Lundgren A, Jertborn M, Svennerholm AM. Induction of long term mucosal immunological memory in humans by an oral inactivated multivalent enterotoxigenic Escherichia coli vaccine. Vaccine 2016 Jun 8;34(27):3132-40. )

5) Alam MM, Aktar A, Afrin S, Rahman MA, Aktar S, Uddin T, Rahman MA, Al Mahbuba D, Chowdhury F, Khan AI, Bhuiyan TR, Begum YA, Ryan ET, Calderwood SB, Svennerholm AM and Qadri F: Antigen-specific memory B cell responses to enterotoxigenic Escherichia coli infection in Bangladeshi adults PLOS Neglected Trop Dis 8: e28

6) Leach S, Lundgren A, Carlin N, Lofstrand M, Svennerholm AM. Cross-reactivity and avidity of antibody responses induced in humans by the oral inactivated multivalent enterotoxigenic Escherichia coli (ETEC) vaccine ETVAX. Vaccine 2017 Jul 13;35(32):3966-73)

7) Lundgren A, Kaim J, Cardeno A, Leach S, Svennerholm A-M. Th1 and Th17 T cell responses induced by oral vaccination against enterotoxigenic Escherichia coli. Manuscript in preparation.

8) Cardeno A, Magnusson KM, Quiding-Jarbring M, Lundgren A. Activated T follicular helper-like cells are released into blood after oral vaccination and correlate with vaccine specific mucosal B-cell memory. Manuscript undergoing final review in Scientific Reports.

9) Bhuiyan TR, Hoq MR, Nishat NS, Al Mahbuba D, Rashu R, Islam K, Hossain L, Dey A, JB, Ryan ET, Calderwood SB, Svennerholm AM, Qadri F: Enumeration of Gut-Homing β7-Positive, Pathogen-Specific Antibody-Secreting Cells in Whole Blood from Enterotoxigenic Escherichia coli- and Vibrio cholerae-Infected Patients, Determined Using an Enzyme-Linked Immunosorbent Spot Assay Technique. Clin Vaccine Immunol. Oct 28;23(1):27-36, 2015

10) Bhuiyan TR, Hoq MR, Nishat NS, Al Mahbuba D, Rashu R, Islam K, et al. Assessing antigen specific HLA-DR+ antibody secreting cell (DR+ASC) responses in whole blood in enteric infections using an ELISPOT technique. Microbes Infect 2017 Nov 8.

Work Package 3: “Discovery and evaluation of novel Shigella cross protective protein antigens”
WP Leader: Christoph TANG [Beneficiary 17 – The University of Oxford (UOXF)]

- Task 1: In vivo transcriptomics (Leader: N. Thomson, WTSI)

In order to identify Shigella genes encoding possible novel protein candidates for cross-serotypic iprotection, we decided to proceed to transcriptomic analysis of bacteria exposed to in vivo conditions of growth in the course of an infection. Hence we undertook RNAseq analysis of Shigella flexneri and Shigella sonnei during growth in the mammalian intestine using the rabbit ligated ileal loop model. This model is the most validated non-primate model of intestinal shigellosis, and requires introduction exposure of the ileum of rabbits, and introduction of Shigella into the lumen. To differentiate mRNA of Shigella from closely related resident E. coli flora, Shigella was placed inside a dialysis bag, to allow bacteria access to nutrients/immune modulators in the lumen; wild-type Shigella was also introduced outside of the dialysis bags so they could interact with the intestinal lumen and elicit host responses which would change the local microenvironment. mRNA was collected from bacteria after 8 hrs in the gastrointestinal tract, as well as from bacteria grown in the laboratory for comparison.

The major challenges with this approach have been optimisation of the ileal loop model, obtaining sufficient RNA, and recovering high quality RNA from bacteria in the intestine. These issues were circumvented. We obtained high quality RNA from ileal loops that has been subjected to RNA sequencing. A key to success in these experiments is separating the cells and therefore RNA of Shigella from that of the host and the microbial flora. This would allow an accurate representation of the Shigella transcripts compared with resident commensal E. coli, and an excess of material from the host. This has, again, been achieved by introducing Shigella into ileal loops within dialysis tubing, to allow free passage of solutes and small molecules, without passage of cells.

Loops were infected with 109 Shigella and incubated in vivo for 8 hrs to allow for their adjustment to the environment in the GI tract. At this stage, bacteria were harvested, collected by centrifugation, resuspended in RNAlater, then frozen in liquid nitrogen. There was no significant degradation of the samples, and several conditions were compared. We have examined the behaviour for both S. flexneri and S. sonnei in the presence and absence of intestinal inflammation; in the presence of inflammation, invasive Shigella (108 CFU) was introduced into the GI lumen outside of the dialysis tube to provoke a host reaction at the mucosal surface. Results have shown that a relatively small subset of genes are subject to differential regulation during inflammation. Genes that are markedly enriched include those of currently unknown function, suggesting that their role in microbial physiology might be revealed by in vivo analyses. These experiments have been carried out in duplicate, on two separate occasions.

RNAseq was undertaken by Vertis, and the data analysed by Rockhopper. The results gave a list of over 300 genes which were differentially up- and down-regulated in vivo.

- Task 2: Analysis under in vivo simulated conditions (Leader: P. Sansonetti, IP)

As outlined in D3.1 we have obtained a list of surface exposed proteins encoded by genes which are upregulated in vivo and in conditions in the laboratory which simulate the in vivo situation (Task 1).

We have also compiled a list of genes which are differentially regulated in the laboratory in conditions which mimic in vivo growth i.e. under anaerobiasis, and genes which are regulated by Shigella FNR, the master regulator of anaerobic growth. The data published in BMC Genomics (1) acknowledges support from the Stopenterics consortium. The information was further integrated with proteomic results generated during the course of the project (Task 4, please see D3.5 International Journal of Medical Microbiology (2).

We undertook analysis of the genes identified in vitro and in vivo, and looked at conservation of potential antigens and their predicted surface location in a number of available Shigella genomes. We searched for genes present in Shigella but absent from E. coli, as we wanted an antigen to stimulate responses against pathogenic Shigella but not against members of the host microflora. We initially surveyed genomes from the strains selected for inclusion in the Stopenterics consortium (Task 3); the sequence of these genomes were published in Gut Pathogens (3) which acknowledges support from the Stopenterics consortium. However we have also access to extensive database of genome sequences through the collaborative networks established at the Sanger Centre. Analysis of genes in this extended set of strains demonstrated that there is considerable overlap in the accessory genomes of E. coli and Shigella.

Therefore there was a limited list of outer membrane proteins whose expression was upregulated in vivo and which retained some degree of specificity for Shigella. The identity of these proteins was passed onto the leader of WP6 for consideration for inclusion in the vaccine pipeline; some candidates were discarded because of patent issues.

- Task 3: Immunoproteomics (Leader: P. Sansonetti, IP)

Samples for proteomic analyses have been obtained from in vivo conditions. Furthermore antibodies are available from animals that have been immunised with outer membrane blebs (OMBs or GMMA; Task 4), and children with clinical shigellosis (WP 2, Professor Cohen). Proteomic samples were obtained using the same conditions as for RNA analysis to allow comparative analysis between the level of key proteins and gene expression.

The surface proteome of Shigella was derived from vesicles collected from bacteria during growth in the laboratory. Growth conditions were kept constant in a chemically defined medium until mid-log phase of growth. We utilised chemically defined media to avoid identification of peptides which are present in protease extracts in commercially available media. A deletion in tolR was introduced into strains from S. flexneri and S. sonnei. The inactivation resulted from deletion of the gene and replacement with a resistance marker on the chromosome. The tolR gene product is involved in the integrity of the outer membrane in Gram negative bacteria. Therefore tolR mutants have a propensity to shed outer membrane vesicles into the media during growth. This allows for an increased recovery and thence detection of membrane associated proteins from bacteria, separate from the cell pellet.

A total of over 2,300 proteins were identified by mass spectrometry from the total proteomes of S. flexneri and S. sonnei, with significant overlap between the species. The proteins were derived from corresponding peptides, and advantage was taken of whole genome sequences derived from the Stopenterics consortium, and the predicted proteins. We used standard algorithms to assign putative cellular locations to proteins.

Membrane associated proteins were significantly enriched in the proteomes of samples of outer membrane vesicles from both species, when compared with the proteomes of total cellular lysates from the same bacteria. This allowed us to generate a catalogue of the 40 most abundant surface associated proteins, and the presence of the corresponding genes was analysed in extended sequence database of E. coli and Shigella at the Sanger Centre.

The characterisation of the surface and total cellular proteomes from S. flexneri and S. sonnei were published in International Journal of Medical Microbiology (2), which cites support from the Stopenterics consortium.

Our results were integrated with other findings from WP3 to deliver a list of candidates for analysis.

- Task 4: Shigella OMB (Leader: C. Gerke, NVGH)

. Task 4.1: Identification of surface exposure and surface epitopes of antigens identified in Tasks 1 and 2.

As the surface digestion of GMMA proved unsuccessful in the first reporting period and was discontinued, a new approach to identify surface-exposed proteins was developed. Immunogenic surface proteins in GMMA were identified using a surface immunoprecipitation approach using live bacteria (see Task 4.2 below and Deliverable Report D3.3). A comparison with the initial results of the in vivo transcriptomics (carried out in Task 1) was performed and showed an overlap of the proteins identified by the two approaches.

Task 4.2: Characterization and optimization of Shigella OMB used as vaccine.

In the first reporting period we described that immunization with GMMA and modulated O-Antigen from either S. flexneri or S. sonnei (described in J Biol Chem (4) which cites Stopenterics support). protected mice against lethal challenge with S. flexneri (75%) but not against S. sonnei. The full proteome of S. sonnei and S. flexneri GMMA was determined to characterize the composition of GMMA that were protective against S. flexneri. S. sonnei were shown to be immunogenic, published in PLoS One (5) which cites Stopenterics support). Predominantly outer membrane proteins, outer membrane lipoproteins and periplasmic proteins were identified (see Deliverables Reports D3.3 and D3.4). To down-select candidates, experiments to determine the immune-proteome of GMMA were carried out using a surface immunoprecipitation approach. Live bacteria were incubated with sera raised against GMMA, washed, and subsequently lysed under conditions that maintain the antibody-target binding. The complexes were purified and analysed that mass spectrometry. Several outer membrane proteins were identified as immunogenic. Studies to investigate if the proteins elicit a protective response are planned.

In addition, in investigations to understand the different outcomes of the challenge studies with S. sonnei and S. flexneri after immunization with GMMA without O-Antigen, we found that S. sonnei expresses a capsule composed of the same polysaccharide as the outer antigen in the lipopolysaccharide. In contrast, S. flexneri does not possess a capsule. The capsule likely shields the targets for the immune response raised by GMMA proteins in S. sonnei and thus its identification will direct the further target selection. Understanding the process of immune evasion by the capsule was described in PLos Pathogens (6) and acknowledges the STOPENTERICS consortium.

In conclusion, this work package has produced a huge amount of unique transcriptomic and (immune)proteomic data concerning two major Shigella serotypes comparatively studied in different conditions. It, unfortunately did not provide the set of novel Shigella-specific proteins that may serve as cross-serotype protective antigens. The major reasons were the few additional proteins expressed in in vivo compared to in vitro conditions of growth. The even fewer number of proteins selected upon these criteria that were not also present on commensal E. coli strains. In addition, the few remaining candidates were either already identified and patented, or corresponded to transmembrane proteins that are notoriously extremely difficult to extract and purify, hence making them totally inappropriate for further development as candidates as part of a vaccine. These decisions were taken during our annual STOPERICS meetings, as well as following discussion with our partners from the vaccine industry.
This WP allowed to provide a good catalogue of protein immunogens that may be part of the immunogenicity and possible protective properties of the GMMA candidate vaccine.


1 – Vergara-Irigaray M, Fookes MC, Thomson NR, Tang CM
RNA-seq analysis of the influence of anaerobiosis and FNR on Shigella flexneri.
BMC Genomics. 2014 Jun 6;15:438. doi: 10.1186/1471-2164-15-438.

2 - Maggiore L, Yu L, Omasits U, Rossi O, Dougan G, Thomson NR, Saul A, Choudhary JS, Gerke C.
Quantitative proteomic analysis of Shigella flexneri and Shigella sonnei Generalized Modules for Membrane Antigens (GMMA) reveals highly pure preparations.
Int J Med Microbiol. 2016 Feb;306(2):99-108. doi: 10.1016/j.ijmm.2015.12.003. Epub 2015 Dec 23.

3 - Rossi O, Baker KS, Phalipon A, Weill FX, Citiulo F, Sansonetti P, Gerke C, Thomson NR.Draft genomes of Shigella strains used by the STOPENTERICS consortium. Gut Pathog. 2015 Jun 4;7:14. doi: 10.1186/s13099-015-0061-5. eCollection 2015

4 - Rossi O, Pesce I, Giannelli C, Aprea S, Caboni M, Citiulo F, Valentini S, Ferlenghi I, MacLennan CA, D'Oro U, Saul A, Gerke C. Modulation of endotoxicity of Shigella generalized modules for membrane antigens (GMMA) by genetic lipid A modifications: relative activation of TLR4 and TLR2 pathways in different mutants. J Biol Chem. 2014 Sep 5;289(36):24922-35. doi: 10.1074/jbc.M114.566570. Epub 2014 Jul 14.

5 - Gerke C, Colucci AM, Giannelli C, Sanzone S, Vitali CG, Sollai L, Rossi O, Martin LB, Auerbach J, Di Cioccio V, Saul A. Production of a Shigella sonnei Vaccine Based on Generalized Modules for Membrane Antigens (GMMA), 1790GAHB. PLoS One. 2015 Aug 6;10(8):e0134478. doi: 10.1371/journal.pone.0134478. eCollection 2015.

6 - Caboni M, Pédron T, Rossi O, Goulding D, Pickard D, Citiulo F, MacLennan CA, Dougan G, Thomson NR, Saul A, Sansonetti PJ, Gerke C. An O antigen capsule modulates bacterial pathogenesis in Shigella sonnei. PLoS Pathog. 2015 Mar 20;11(3):e1004749. doi: 10.1371/journal.ppat.1004749. eCollection 2015 Mar.

Work Package 4: “ETEC antigen discovery”
WP Leader: Halvor SOMMERFELT [Beneficiary 2 – Universitetet i Bergen (UiB)]

Infection with ETEC is one of the main causes of diarrheal diseases among young children in low- and middle-income countries, as well as among travellers to these countries. Human ETEC are defined as E. coli that are capable of producing one of both of the enterotoxins heat-stable toxin (ST) and heat-labile toxin (LT), and these enterotoxins induce diarrhea by binding to and activating the guanylate cyclase and the ganglioside GM1 receptors, respectively, found on the surface of human intestinal cells. There are two close to identical variants of ST, named human (STh) and porcine (STp), and the toxins and most of the other known ETEC virulence factors are encoded on large plasmids that often can be transferred to other E. coli. Several different ETEC families have emerged from the E. coli population through the initial transfer of the toxin genes, and close to all STh-producing ETEC, which are arguably the most pathogenic ETEC for young children, originate from 8 globally widespread and stable families.

To reduce the burden of childhood diarrhea, there is an ongoing effort to develop vaccines that may reduce the risk of ETEC infection and diarrhea. Since ETEC represents a diverse group of E. coli, the main challenge with the vaccine development is to identify suitable vaccine components that could offer a broad and specific protection. Since all human ETEC per definition produce one or both of ST and LT, part of the current strategy is to include detoxified versions of ST and LT in vaccines. While LT is large and immunogenic and has already been successfully targeted in vaccines offering some protection against diarrhea with LT-producing strains, ST by itself is too small to induce an immune response.

Part of the focus of Work Package 4 (WP4) has been to develop a strategy to present de-toxified ST molecules on the surface of larger, immunogenic molecules so that the immune response it triggers can neutralize the activity of native ST. For this work, we used a library of ST molecules where each of ST’s amino acids had been replaced with one of the 19 other amino acids most commonly used by E. coli. Given ST is 19 amino acids long, this gave a library of 361 single mutation ST molecules. After screening each mutant to identify those that no longer were toxic but still were recognizable by antibodies against native ST, we chemically coupled the most promising of these mutants to different larger carrier molecules and immunized mice. Thus, we identified a suitable carrier molecule and a coupling strategy that gave strong immune responses against native ST. While these results represent an important step towards developing effective ST-based vaccines, more work is needed to characterize the response, particularly in relation to degree of cross-reactivity with the endogenous guanylin and uroguanylin ligands.

As part of the effort to identify new ETEC antigen that could boost the effectiveness of new vaccine prototypes, WP4 has also focused on characterizing the different ETEC families that exists in the E. coli population, in particularly strains from the eight families that appear to be most often associated with childhood ETEC diarrhea. The goal has been, in part, to identify proteins that ETEC strains from these families produce and that could potentially contribute to inducing a broad and specific protective response if used in a vaccine.

To identify and characterize the proteins that these strains are capable of producing, we performed whole-genome sequencing of 25 ETEC strains from the eight above-mentioned families in order to obtain fragmented pictures of each strain’s genome. For one representative strain from each family, we also generated uninterrupted DNA sequences for each strain’s chromosome and for several of their plasmids. Having these uninterrupted sequences helped us to identify the location of genes (chromosome or plasmid), to identify stretches of DNA that differ between strains, and to identify groups of genes that may be functionally associated with each other.

In addition to the 25 ETEC strains, we also sequenced 19 ancestrally closely related E. coli strains. We used these in comparative genomics analyses to identify regions on the chromosomes or plasmids that differ between ETEC and other E. coli and which, therefore, could indicate an ETEC-specific trait that could be targeted by the vaccine development effort. By using a range of bioinformatic analyses, we identified 146 genes that we believe could be associated with ETEC colonization. To further characterize these proteins, their genes were cloned into an expression vector, and the expressed protein was subsequently purified.

In a parallel running project, human volunteers have been experimentally infected with the strain from which we had cloned the 146 genes, and we therefore expected that if any of these proteins were exposed to the immune system during the colonization process, we would find an increase in serum antibodies that target these proteins. We found elevated antibody responses against 16 of these proteins. More volunteers have now been experimentally infected with other strains we have characterized, and we are currently evaluating the degree of cross-reactivity in the immune response between proteins from different ETEC families are tested. The goal is to find proteins that could induce a strong and ETEC-specific response across several important ETEC families.

Work Package 5: “Antigen discovery: synthetic carbohydrates as immunogenic mimics of Shigella somatic antigen”
WP Leader: Laurence MULARD [Beneficiary – Institut Pasteur (IP)]

This WP aimed at accelerating the development of vaccine candidates against Shigella with focus on the polysaccharide O-side chains of the lipopolysaccharide (LPS), also called the somatic antigen (O-Ag). These antigenic structures of the bacterial surface are the major targets of immunological protection following natural infection, they are also eminently variable and on a general basis qualify as type-specific. Based on epidemiological data, this WP has considered the need for a vaccine providing broad coverage against S. flexneri and S. sonnei, the major circulating subgroups. Still, considering the number of involved subtypes and their substantial geographic and temporal variation in prevalence, a multivalent formula is required. Aware that complexity may turn out to be prohibitive for implementation of the target vaccine in most low- and middle-income countries, we reasoned that not all subtypes needed to be represented in the vaccine. As illustrated by available serotyping data, the chemistry of the S. flexneri O-Ags suggests that an O-Ag-based vaccine encompassing S. flexneri serotypes 1b, 2a, 3a and 6, would induce sera cross-reacting with most S. flexneri LPS, and therefore would provide coverage against not only the subtypes included in the vaccine but also other circulating subtypes. This assumption was later supported by epidemiological data in children under five, indicating that those four serotypes and cross-reactive ones represented 74% and 90%, respectively, of the total circulating S. flexneri strains. Therefore, investigation toward an O-Ag-based vaccine comprising a cocktail of four S. flexneri serotypes, as listed above, and S. sonnei was the selected strategy in WP5. Whilst the most prevalent subtypes are included in the vaccine, the expected coverage in the field is 83% of total Shigella.

Parenterally-administered detoxified LPS conjugates had a promising start, but met limitations in phase 3 trials conducted in infants and toddlers. Aiming at providing novel solutions by improving the immunogenicity of the selected O-Ags to fulfil requirements for vaccine efficiency in the first three years of life, WP5 was dedicated to the design of synthetic carbohydrate-based conjugate vaccine candidates, whereby synthetic oligosaccharides acting as antigenic, conformational, structural and functional mimics of the selected O-Ags are turned into potent immunogens following their site-selective conjugation onto an appropriate protein carrier. It is noteworthy that SF2a-TT15, which was designed to prevent S. flexneri 2a infection, is the first vaccine candidate to successfully illustrate this powerful approach up to a clinical trial (see WP1). As outlined below, achievements in WP5 further demonstrate the central contribution of glycochemistry and the potential of the implemented multidisciplinary rational strategy for the development of promising polysaccharide-based vaccines.

On the one hand, panels of protective IgG monoclonal antibodies required to identify antigenic mimics of the native Shigella O-Ags, were obtained against S. flexneri 1b, 3a and 6 by immunizing mice with whole bacteria and selecting based of binding to the homologous LPS.

On the other hand, large panels of well-defined oligosaccharides representing fragments of the O-Ags from S. sonnei and the three S. flexneri of interest were synthesized by use of finely tuned convergent routes, exploiting the potential of state-of-the art glycochemistry and beyond. Advantageously, the implemented strategy provided tools required to investigate in detail the contribution of the various O-Ag components to antigenicity and immunogenicity. Some 100 oligosaccharides were synthesized, purified and fully characterized. Key parameters taken into account included oligosaccharide chain length – from di- to 20mer and larger –, endchain residues since they are expected to be immunodominant when dealing with small and medium size haptens, stoichiometric and non-stoichiometric substitutions, which play a role in vaccine escape as well as in cross-reactivity among subtypes. As illustrated in the case of S. flexneri 3a, the most advanced serotype, a comprehensive study combining immunochemistry and biophysical analysis with external support from NMR and crystallography, led to identify synthetic oligosaccharides, fulfilling the required criteria in terms of antigenic, conformational and structural O-Ag mimicry.

Taking advantage of up-to-date sugar-protein carrier conjugation techniques, sets of glycoconjugates involving the most promising haptens, based on in vitro analysis, and tetanus toxoid as a carrier were engineered and evaluated in vivo. S. flexneri 3a vaccine candidates optimized in terms of hapten size and structure – in this case a 15mer or 20mer oligosaccharide –, hapten:tetanus toxoid loading, adjuvant and dose to be used to induce strong bactericidal antibody titers in mice were identified. Moreover, the proof of concept for a bivalent S. flexneri 2a/3a synthetic carbohydrate-based vaccine candidate was achieved in mice. Based on available pre-clinical data established in the frame of WP5 for a S. flexneri 2a/3a bivalent formulation and on the first disclosed data in human for SF2a-TT15 (see WP1), we propose the first S. flexneri bivalent glycoconjugate vaccine ready-to-go for clinical trial. Last but not least, achievements in the case of S. flexneri 1b, 6 and S. sonnei provide additional support, if needed, to the synthetic carbohydrate-based strategy to a Shigella vaccine explored in the frame of WP5. They pave the way to a broad coverage Shigella vaccine, fulfilling the increasing regulatory requirements for better-defined and safer molecules. The next step will consist in going beyond in vitro analysis to in vivo study so as to rapidly identify the most promising multivalent combination, and in fact only one, to push forward toward a phase I clinical trial.

Work Package 6: “Research and Development, from antigens to vaccine formulations and preclinical testing”
WP Leader: Maria Lina BERNARDINI [Beneficiary 6 – Universita degli studi di Roma La Sapienza (UNIROMA1)]

This WP is aimed at the preclinical studies of the vaccine candidates identified and developed in the consortium. This section includes both in vitro tests and in vivo studies to evaluate the immune response elicited by the vaccine candidates chosen by the network. In particular, in vitro tests were finalized to assess the protective efficacy of the antibodies raised against the relevant antigens in cell models mimicking the main virulence phenotypes of the pathogens. The in vivo studies were mainly focused on the development and the use of a novel animal model suitable to be exploited in the analysis of the protective efficacy of the relevant vaccine candidates.

Initially, a restricted number of Shigella proteins have been considered to be suitable for further studies as putative antigen candidates. This group included proteins secreted by the type 3 secretion system (T3SS) of Shigella, which are present in shigellae belonging to all serogroups and serotypes. However, for the majority of them the experimental procedures aimed at the protein purification failed due to the properties of the molecules: hydrophobicity or poor solubility. However, among them the IpaD antigen- secreted by the T3SS- seemed to be suitable for vaccine use and it was used in further analysis.

Shigella is a human pathogen and no animal is sensitive to the natural infection with it. To overcome this problem several animal models have been proposed and used, even if no animal model efficiently recapitulates all the features of shigellosis. Therefore, the choice of a suitable animal model has been a crucial decision for the network. Furthermore, we tried to choose an animal model adapted to assess the protective immunity elicited by the Shigella and the ETEC antigen candidates. The model of intra-rectal infection of guinea pig seemed the best choice to achieve both goals. Thus, major attention was paid to validate this model that was originally conceived only as a pathogenicity model to assess the virulence of Shigella. The limits of this animal model, which likely is the best non-primate model of infection to mimic some of the main features of shigellosis, reside in the poor characterization of the guinea pig immune responses and on the lack of knowledge about the guinea pig microbiota. In WP6 we set up a stringent protocol of Shigella infection via the intra-rectal route in these animals that would be reproducible for different species and serotypes of Shigella (S. flexneri 5 and 2 and S. sonnei). In a second step, this model was tested to assess the protective immune reaction upon immunization with the Shigella IpaD vaccine candidate (see later). With this aim, an appropriate immunization schedule and several macroscopic, microscopic and molecular parameters associated to the immune protection and suitable correlates of protection in sera were established.

Unfortunately, this model resulted unsuitable for the analysis of ETEC infection and it was abandoned for this pathogen.

In parallel we proceeded to in vitro studies by using polyclonal antibodies (rabbit immune serum) raised against Shigella IpaD. Several approaches were used. We assessed the antibody-mediated inhibition of epithelial cell invasion, of macrophage pyroptosis and of macrophage phagocytosis. S. sonnei and various serotypes of S. flexneri (5, 2a, 3, and 6) were tested in this assay. The anti-IpaD immune serum reduced bacterial cell invasion, and increased the rate of macrophage phagocytosis and pyroptosis of all strains to various extent, depending on the serotype.

This antigen was also tested for its capacity to protect the guinea pigs against S. flexneri 2a and S. sonnei challenges in the intra-rectal model of infection. The candidate was used in combination with an adjuvant –kindly provided by Sanofi Aventis- principally eliciting a Th1-oriented response. The results of these experiments showed that the antigen IpaD, only in combination with the adjuvant, protected animals from the disease provoked by pathogens S.flexneri 2a and S. sonnei. However, the crucial role of the adjuvant clearly emerged in the results, indicating that IpaD alone was a poor protective antigen. In conclusion, this model did not prove sensitive enough to validate a potential cross-serotype protective protein.
Similarly, when the synthetic OS-based conjugate vaccine, SF2a-TT15, was tested in this model, the balance between poor immunogenicity of the candidate in this model and the strong proinflammatory effect of the adjuvant made it clearly inadequate to predict protective immunity.

Work Package 7: “Training in vaccinology focusing on enteric vaccines”
WP Leader: Armelle PHALIPON [Beneficiary 1 – Institut Pasteur (IP)]

This WP was aimed at promoting training in vaccinology, focusing on some of the crucial steps in developing vaccines against enteric diseases. The proposed training has thus consisted in 3 different types of courses:
The 4wk-Vaccinology course organized at the Institut Pasteur (IP) , Paris, France, from 2011 and 2014, with a workshop activity on Shigella vaccine design on years 2013 and 2014;
the 2wk-course organized at Tel Aviv University (TAU), Tel Aviv, Israel, on the development and evaluation of enteric vaccines;
the 2 wk-course organized at ICDDR,b, Dhaka, Bangladesh on clinical trials and immunomonitoring of enteric vaccine candidates. Note that the workshop activity on Shigella vaccine design was also implemented during the Hong Kong Institut Pasteur Immunology course in 2014.

In summary, the learning objectives were:
IP Course: introduction to vaccinology, an integrative discipline, basic principles of preclinical and clinical trials, update of vaccine development for the major infectious diseases, and future challenges. A workshop dedicated to the design of a Shigella vaccine was implemented in 2013 and 2014 with experts belonging to the Stopenterics consortium (
TAU course: basics in vaccinology and detailed overview of the development of vaccines against enteric diseases, with success, drawbacks and challenges.
-ICDDR,b course: acquire knowledge on diarrheal diseases pathogenesis with emphasis on ETEC, cholera and shigellosis, implications of epidemiology and pathogenesis on the design of vaccines against the 3 corresponding pathogens.

Altogether, 165 trainees have benefited from these courses (100 for the IP course, 22 for the TAU course, 19 for the ICDDR,b course, and 24 for the IP Hong-Kong course), among which 22 were issued from laboratories belonging the Stopenterics consortium.

Besides this activity, a website dedicated to Stopenterics ( was created in order not only to promote interactions between the participants, but also to share expertise via a e-learning content and activities. The website was also key for the participants to the different courses for self-evaluation and preparing the Shigella vaccine design workshop activities before the courses started.
Further funding for updating this website would be key for the international enteric vaccine community with a particular focus on Shigella/ETEC vaccine development.

Potential Impact:
STOPENTERICS was framed around the concept of innovating in the field of Shigella and ETEC vaccine development, two enteric pathogens that, as etiological agents of severe diarrhea in children below 5 years in low income countries, weight together as much as rotavirus before implementation of rotavirus vaccination which has divided almost by half the incidence of diarrheal diseases in countries that have implemented this vaccination.

In other word, availability of a combined Shigella and ETEC vaccines, if implemented, would allow major further decrease of diarrheal diseases in the pediatric population, particularly in sub-saharan Africa, Indian subcontinent and south-east Asia. Based on current WHO figures, one could expect a further decrease of yearly death incidence of about 200 000, notwithstanding morbidity, particularly malnutrition (acute and chronic) associated with these infections.

STOPENTERICS has brought various degrees of proof of concept that this is achievable, particularly at this stage for Shigella prevention.
Prevention of infectious diseases is a major element warranting development of low income countries. We do hope STOPENTERICS will have contributed to this aim.

The main focus of Work Package 4 (WP4) has been on contributing to the development of new and effective vaccines against ETEC infection and diarrhea. The global ETEC vaccine development effort has several vaccines prototypes in the pipeline, but given that ETEC is comprised of a diverse group of E. coli, and that ETEC virulence factors tend to differ between different strains within and between families and between strains from different geographical locations, producing vaccines that can offer a broad and effective protection is challenging. The work done in WP4 will potentially impact the direction and speed of the ETEC vaccine development effort. This effort has traditionally been focused on inducing protection by immunizing with antigens that target the heat-labile toxin and the colonization factors, which usually are ETEC-specific plasmid-encoded fimbria that help anchor ETEC to the intestinal cell wall during colonization.

With parallel funding, we have managed to present a non-toxic version of the small heat-stable toxin (ST) which, after coupling to a carrier and administered as a conjugate to mice, induces a strong, neutralizing immune response against native ST. Most of the ETEC strains that contribute to the bulk of the ETEC disease burden in low and middle-income countries produce this toxin, and an effective vaccine-induced protection against ST-ETEC diarrhea is likely to reduce the risk of diarrhea and subsequent acute dehydration which often is life-threatening for young children. The work in the current project compared different carriers and concluded that a non-toxic derivative of ETEC’s heat-labile toxin subunit B, LT-B, was the most suitable carrier for the ST derivatives. Ultimately, we expect to produce a vaccine that can be used to protect young children living in low- and middle-income countries against ETEC diarrhea, in addition to provide an effective way to greatly reduce the risk of diarrheal diseases among travellers to these countries.

The work to characterize strains from the eight most epidemiologically important ETEC families contribute to the understanding of how ETEC functions and evolves, and it will help to identify potential new traits that can be targeted by vaccine development to develop broadly protective vaccines. At the same time, our assays used to screen for new potential new antigens should be of interest to ETEC researchers wishing to test promising vaccine antigens.

Although some of the methodology and results from this work will need design protection, the main dissemination activities for WP4 included presentations at public and scientific meetings, deposit of relevant data in publicly accessible databases and articles in peer-review scientific journals, some of which are already published.

New achievements in basic research generated through WP5 provide the innovation and fundamental knowledge to move the development of synthetic carbohydrate-based vaccines forward, in particular in the context of Shigella. Besides the identification of a promising bivalent vaccine candidate, major tools necessary for the design of the missing vaccine valences are now available. Remarkably, the results have encouraged additional dedicated financial implication from Institut Pasteur through its “Grand Programme Fédérateur” in Vaccinology (FlexBiVac project, PI: L. Mulard), and serve as a basis for ongoing discussion with the Bill and Melinda Gates Foundation to support further the development of a broad coverage synthetic carbohydrate-based Shigella vaccine (GlycoShig proposal, PI: L. Mulard).

WP5 has offered impressive training capacity for the young participating investigators by providing several fellowships as well as lab resources for postdoctoral fellows, PhD and master students. Data generated through WP5 are or will be made available to the scientific community as patents and by their publication in peer-reviewed scientific journals, conferences and seminars. They should also reach the general public through the Institut Pasteur and EU websites and possibly in the long run through the implementation of original Shigella diagnostic tools and/or vaccine.

Moreover, going beyond a Shigella vaccine, the results generated in WP5 combined to the demonstration that SF2a-TT15 was strongly immunogenic in naïve adults (see WP1), should contribute to novel breakthroughs in glycoscience and in vaccinology. Undoubtedly, they provide strong support to further promoting a carbohydrate-based medicinal chemistry approach for the development of powerful drug-like molecular vaccines. Owing to the relevance of carbohydrate antigens in antibacterial and more generally antimicrobial vaccination, the strategy illustrated in WP5 may find broad applications resulting in major societal benefit and therefore overall economic impact.

Through the different courses held both in high-income and low-income countries, 165 participants from more than 20 different nationalities have been trained in the field of vaccinology, including diarrheal diseases and vaccine development with a special focus on Shigella, ETEC and cholera. This training activity has therefore contributed to create an international community of young scientists who might become future actors in the field.

An excellence-oriented management system based on a two-level structure was put in place from the very beginning of the project in order to ensure the successful achievement of the STOPENTERICS ambitious objectives described in Annex I-Description of Work.

i) Decision-making/Strategic Management involved the Project/Scientific Coordinator, Prof.
Philippe Sansonetti (IP), and Steering Committee, composed of the Scientific Coordinator (Chair) and all Work Package Leaders, officially appointed at the STOPENTERICS Kick-Off Meeting held on 13th - 14th December 2010 in Abbaye de Royaumont, France.

ii) Day-to-Day operational management was led by the Project Coordinator, Prof. Philippe Sansonetti, with the support of Thierry Planchenault, Sophie Ablott, Thomas Wiest and Soizic Sergeant, STOPENTERICS Project Managers within the Grants Office of Institut Pasteur, and associated departments in the execution of all management-related, financial, legal and administrative tasks.

All project meetings, along with regular internal communication, were organised to discuss and ensure the scientific coherence and excellence of the project, and to follow its dissemination and communication activities:
The Kick-Off meeting was held on 13th – 14th December 2010 in Abbaye de Royaumont, France;
The 1st Annual meeting was held on 11th-14th December 2011, in Tel Aviv University, Israel;
The 2nd Annual meeting was organised on 11th-12th October 2012, Annecy, France;
The 3rd Annual meeting was held on 25th-26th September 2013, in Paris, France;
The 4th Annual meeting took place on 9th-10th October 2014, in Rome, Italy;
The 5th Annual / Final meeting took place on 28th-30th September 2015 in Oxford, United-Kingdom.

The Steering Committee met during these meetings to monitor the advancement of the project and prepare the next scientific and dissemination activities, including preparing post-grant activities in order to maintain and build on the project’s results. The Scientific Advisory Board also attended these meetings; their expertise and insight was fruitful, and their support for developing further funding strategies for the continuation of the STOPENTERICS project after its end was encouraging.

Aware of the EU Open Science policy, STOPENTERICS actively communicated on its activities to ensure the full exploitation of achieved results to the scientific community and the stakeholders, and foster a new generation of researchers in neglected infectious diseases. The consortium generated more than 40 high-ranking scientific publications in peer-reviewed journals, the majority of them having Open Access, and around 170 dissemination activities.
List of Websites: