Vaccines against gastroenteritis, current progress and challenges

ABSTRACT

Enteric viral and bacterial infections continue to be a leading cause of mortality and morbidity in young children in low-income and middle-income countries, the elderly, and immunocompromised individuals. Vaccines are considered an effective and practical preventive approach against the predominantly fecal-to-oral transmitted gastroenteritis particularly in the resource-limited countries or regions where implementation of sanitation systems and supply of safe drinking water are not quickly achievable. While vaccines are available for a few enteric pathogens including rotavirus and cholera, there are no vaccines licensed for many other enteric viral and bacterial pathogens. Challenges in enteric vaccine development include immunological heterogeneity among pathogen strains or isolates, a lack of animal challenge models to evaluate vaccine candidacy, undefined host immune correlates to protection, and a low protective efficacy among young children in endemic regions. In this article, we briefly updated the progress and challenges in vaccines and vaccine development for the leading enteric viral and bacterial pathogens including rotavirus, human calicivirus, Shigella, enterotoxigenic Escherichia coli (ETEC), cholera, nontyphoidal Salmonella, and Campylobacter, and introduced a novel epitope- and structure-based vaccinology platform known as MEFA (multiepitope fusion antigen) and the application of MEFA for developing broadly protective multivalent vaccines against heterogenous pathogens.

KEYWORDS:

• Enteric diseases
• vaccinology
• multiepitope fusion antigen (MEFA)
• enterotoxigenic Escherichia coli (ETEC)
• Vibrio cholerae
• Shigella
• nontyphoidal Salmonella
• Campylobacter
• rotavirus
• norovirus

Introduction

Enteric diseases remain a leading cause of mortality in young children in developing countries and morbidity in children worldwide.^1–5 Parasitic pathogens mainly Cryptosporidium spp., viral pathogens including rotavirus and norovirus, and bacterial pathogens Shigella, enterotoxigenic E. coli, Campylobacter, Vibrio cholerae, and Salmonella are the top causes of enteric infections.^2,6 These enteric pathogens are mostly fecal-to-oral transmitted. Infections are typically initiated with ingestion of contaminated food and water, leading to diseases including nausea, abdominal cramps, fever, and more commonly diarrhea and dehydration. Enteric infections are self-eliminated in healthy adults but without medical intervention can progress to life-threatening conditions and often death in the young children and immune compromised patients. Implementation of community-wide sanitation systems and supply of safe drinking water would effectively prevent enteric infections. However, for many resource-limited countries, water, sanitation, and hygiene (WASH) is not a goal quickly achievable. Handwashing with soap especially antibacterial soap reduces enteric infections in the short term, but it increases the risk of antimicrobial resistance built up in enteric pathogens. Vaccination, on the other hand, is considered currently more effective and more practical to prevent enteric diseases. Unfortunately, there are no vaccines licensed for many pathogens associated with enteric infections.

Challenges in developing effective vaccines against enteric infections include heterogeneity among strains (genotypes, serotypes, or pathotypes) of individual enteric pathogens, a lack of suitable animal models to assess vaccine efficacy in prior to human subject studies, a poor understanding at details of pathogen pathogenesis or disease mechanism (nontyphoidal Salmonella or Campylobacter), undefined host immune correlates to protection, short-lived protective immunity, absence of suitable or cost-effective cell culture systems to attenuate viral pathogens (human Calicivirus), and inability to achieve satisfied efficacy among young children in endemic regions or countries. Genetic and antigenic heterogeneity is perhaps the primary roadblock in developing broadly effective vaccines for many enteric pathogens. Additionally, multiple pathogens are often present at individual patients with enteric infections. This further complicates diagnostics and disease treatment and demands development of combination vaccines against two or more than two enteric pathogens for effective protection. Combination (combo) vaccines become attractive since the current WHO EPI (expanded program for immunization) is getting increasingly crowded with continuous introduction of new vaccines. In this review, we outlined the current status and challenges of vaccines or vaccine candidates for rotavirus, calicivirus, Shigella, enterotoxigenic E. coli (ETEC), cholera, nontyphoidal Salmonella, and Campylobacter, the top causes of enteric infections based on molecular diagnostics,^2,6 and described a new vaccine technology or vaccinology platform for the development of cross-protective multivalent vaccines.

I. Vaccines and vaccine development for viral enteric pathogens

Rotavirus (Rv)

A non-enveloped multi-layered virus is the top viral cause of enteric infections in children especially in countries or regions where rotavirus vaccines have not been introduced.^2,7 Nine or ten rotavirus species (A to I or J) have been identified. Four species (A, B, C, H), but predominantly species A (referred as rotavirus hereby), cause human infections or gastroenteritis. Rotavirus genome is composed of 11 double-stranded RNA (dsRNA) segments which encode 11 to 12 proteins, six viral structural proteins (VP1-VP4, VP6, VP7) and five or six non-structural proteins (NSP1-NSP5, or NSP6). Structural proteins VP7 glycoprotein (G antigen; G genotypes) and VP4 protease-sensitive protein (P antigen; P genotypes) located at the outmost layer of rotavirus virion are the host cell-attachment proteins and induce neutralizing antibodies. Six VP7 G genotypes (G1, G2, G3, G4, G9, G12) and three VP4 P genotypes (P4, P6, P8) are the most important associated with human disease, with six combinations or serotypes (G1P[8], G2P[4], G3P[8], G4P[8], G9P[8] and G12P[8]) responsible for over 90% of the severe rotavirus infections.^8–14 Rotavirus is transmitted primarily by ingestion of contaminated water and food, as well as physical contact of virus carriers including patients and contaminated surfaces, often leading to clinical outcomes ranging from mild watery diarrhea to acute gastroenteritis, dehydration, hypovolemic shock, and death. Oral rehydration (oral rehydration solutions) is the treatment commonly recommended before the introduction of rotavirus vaccines.¹⁵ Since the late 1990 s, several rotavirus vaccines have been introduced. Four vaccines have been prequalified or recommended by WHO for global licensing, and two are licensed by individual nations (Table 1).

Table 1. Licensed vaccines currently available for enteric pathogens.

	Vaccine (Manufacturer)	Type	Components	Availability	Indication
Viral disease
Rotavirus	Rotarix (GSK Biologicals)	Live oral	Attenuated human rotavirus strain (G1P[8])	Global	Infants >6 weeks age, two doses
Rotavirus	RotaTeq (Merck)	Live oral	5 human-bovine reassortant rotaviruses (G1, G2, G3 G4, P[8])	Global	Infants >6 weeks age, three doses
Rotavirus	Rotavac (Bharat Biotech)	Live oral	Attenuated human rotavirus strain (G9P[11])	India, Palestine	Infants >6 weeks age, three doses
Rotavirus	RotaSiil (Serum Institute of India)	Live oral	5 human-bovine (UK) reassortant rotaviruses (G1, G2, G3, G4, G9)	India	Infants >6 weeks age, three doses
Bacterial disease
Cholera	Dukoral (SBL)	Inactivated	Mix of inactivated V. cholerae O1 of Inaba classical, El Tor biotypes and Ogawa serotypes with rCTB subunit	Over 60 countries including Europe, Canada, Australia, and New Zealand, (not in USA)	Adults and children age ≥6 yrs: oral, two doses; Child 2–6 yrs: oral, three doses
Cholera	Shanchol (Shantha Biotechnics)	Inactivated	Mix of inactivated V. cholerae O1 Inaba E1 Tor strain Phil 6973, O1 Ogawa classical strain Cairo 50, O1 Inaba classical strain Cairo 48, O139 strain 4260B	Global Oral Cholera Vaccine stockpile	Adults and children age ≥1 yr: oral, two doses
Cholera	Euvichol (EuBiologics)	Inactivated	Mix of inactivated V. cholerae O1 Inaba Cairo 48 (H), O1 Inaba Phil 6973 El Tor, O1 Ogawa Cairo 50, and O139 4260B	Global Oral Cholera Vaccine stockpile	Adults and children >1 yr, 2 doses
Cholera	mORC-Vax (Vabiotech)	Inactivated	Mix of inactivated V. cholerae serogroups O1 classical and El Tor and O139	Vietnam	Adults and children age ≥1 yr: oral, two doses
Cholera	Vaxchora (PaxVax)	Live oral	V. cholerae strain CVD 103-HgR	USA	A single dose

Rotashield (human-rhesus RRV)

Rotashield is a quadrivalent vaccine product licensed in 1998. Rotashield consists of rhesus monkey backbone strain G3P[3] and three reassorted strains, with VP7 protein (reassorted) of G1, G2, and G4 to protect against VP7 serotypes G1 to G4. A three-dose regimen provided 70–90% efficacy against moderate-to-severe disease in infants.^9–11 However, this product was withdrawn from the market in 1999 due to a low risk of association with intestinal intussusception (IS).^16,17 Later studies showed that association of Rotashield vaccination with IS was age-related¹⁸ and that a regimen of two doses administered at the first and the second months yielded 64% protection without IS adverse effect.¹⁹

RotaTeq (RV5)

RotaTeq is a pentavalent bovine-human reassorted vaccine developed by Merck. RotaTeq has bovine rotavirus strain G6P[5] as the backbone and is composed of five reassorted strains (G1P[5], G2P[5], G3P[5], G4P[5] and G6P[8]). A regimen of three doses provided 74% efficacy against any rotavirus disease and 98% efficacy against severe disease in Europe and USA.^20–23 However, efficacy in developing countries ranged from only 51% to 64% for the first year and 20% to 46% in the second year.^24–26

Rotarix (RV1, RIX4414)

Rotarix is a monovalent vaccine developed by GSK. Rotarix carries tissue culture passage-attenuated human rotavirus strain G1P[8]²⁷ which represents the most common VP4 and VP7 antigens of human rotavirus. At a regimen of two doses, Rotarix yielded similar efficacy as RotaTeq, ranging from 38% to 97% against moderate-to-severe gastroenteritis (a score at ≥7 of Ruuska and Vesikari 20-point scoring system) in different countries or regions.^28–43

ROTAVAC

ROTAVAC is a monovalent vaccine developed by Bharat Biotech in India. It carries live naturally attenuated strain 116E (G9P[11]) isolated from children without symptoms. ROTAVAC is currently licensed in India and is shown 56% and 49% protection against hospitalization in the first and second year of life, and 35% efficacy against any rotavirus infection at a regimen of three doses.^44–48

ROTASIIL

Currently licensed in India, ROTASIIL is a bovine-human reassorted pentavalent vaccine (G1[P5], G2[P5], G3P[5], G4P[5], G9P[5]).^49–51 Developed by Serum Institute of India, ROTASIIL showed 33% efficacy against severe rotavirus gastroenteritis in the first year.⁵²

Additionally, two vaccine products are currently licensed by individual countries. LLR-85 (Lanzhou Lamb Rotavirus Vaccine) based on a lamb strain (G10P[12]) is licensed in China and showed 35–78% efficacy.^53–56 Rotavin-M1, an attenuated human strain (G1P[8]) is licensed in Vietnam.^57,58 Subunit rotavirus vaccines are also under development, including monovalent P2-VP8-P[8]⁵⁹ and trivalent VAC041 (P2-VP8-P[4]P[6]P[8]),^60,61 and efficacy will need to be evaluated in field trials.

Human calicivirus (HuCV)

Unlike rotavirus for which a few vaccines have been prequalified globally or regionally by WHO or individual nations, there is no vaccine licensed for the single-stranded RNA norovirus or other human calicivirus. Norovirus indeed exceeds rotavirus for global disease burden and becomes the most common viral cause of gastroenteritis in developed countries where rotavirus vaccines have been introduced.^6,62-65 Norovirus RNA genome consists of three open reading frames (ORFs). One ORF encodes a polyprotein cleaved into seven non-structural proteins (NP1 to NP7), whereas the other two ORFs encode minor capsid protein VP2 and major capsid protein VP1. VP1 is composed of a conservative shell domain (S) and a variable protruding domain (P). Seven norovirus genogroups and 40 genotypes have been identified, among them three genogroups (GI, GII, and GIV) and 29 genotypes are associated with human gastroenteritis.^66,67 However, genotype 4 in GII genogroup, GII.4, is by far the most prevalent and is responsible for a majority of acute gastroenteritis cases.⁶⁸ Therefore, GII.4 along with the initially identified Norwalk virus GI.1 are primarily targeted in norovirus vaccine development.

In contrast to rotavirus which exclusively infects children, norovirus causes gastroenteritis in people of all ages though mostly among young children and the elderly over 65 years. Like rotavirus, norovirus is mainly transmitted via ingestion of contaminated food and water, with outbreaks occur more commonly in close-contact settings including pre-school centers and schools, retirement care institutes, hospitals, as well as cruise ships. The typical clinical outcome is acute gastroenteritis but can be severe dehydration and death if infection progresses without intervention treatments.

There are several norovirus vaccine candidates under clinical or preclinical studies. Because norovirus is unculturable under current cell culture system, developing attenuated whole-cell vaccines becomes prohibitable. Consequently, calicivirus vaccine candidates under investigation are largely based on viral proteins. Norovirus major capsid protein VP1, when expressed in eukaryotic cells, forms virus-like particles (VLP) to exhibit antigenicity similarly to native viral particles. Additionally, the P domain of VP1 after being linked to a polypeptide and expressed in cell culture also aggregates into particles (P particle). Therefore, VP1 virus-like particles and P particles of global pandemic genotype GII.4 and the regional endemic GI genotypes are the primary antigens for norovirus vaccine development.

Bivalent GI.1/GII.4 VLP

GI.1/GII.4 VLP vaccine candidate under development by Takeda is a continuation of monovalent GI.1 VLP (Ligocyte) and is currently the one progressed furthest. GI.1 VLP was initially designed against Norwalk virus. When orally administered without adjuvant, GI.1 VLP showed 83% IgG seroconversion in volunteers.⁶⁹ Immunogenicity was enhanced when delivered intranasally with adjuvant monophosphoryl lipid A (MPL) and Al(OH)₃ at a two-dose regimen.^70,71 With GII.4 emerged as the most prevalent genotype, monovalent GI.1 VLP was improved to be a bivalent VLP candidate, by adding consensus GII.4 VLP and administered intramuscularly.⁷² Intramuscularly immunized adult volunteers developed GI.1- and GII.4-specific antibodies. Interestingly, GI.1/GII.4 VLP-induced antibodies showed an increasing neutralizing activity – antibody blocking histo-blood group antigens (HBGA).^73–75 When challenged with a GII.4 NoV strain, the vaccinated group showed reduction of disease severity but unfortunately without significant protection against overall gastroenteritis.⁷⁶ The lack of protection appeared to be caused mainly by a lower attack rate from the challenge strain (37.5% infection rate), presumably caused by preexisting anti-NoV antibodies in volunteer adults resulted from natural exposure to NoV infections. Efficacy of bivalent GI.1/GII.4 VLP against recurrent infection or infection from heterogenous genotypes needs to be further evaluated.

Other vaccine candidates including trivalent GI.3/GII.4/rotavirus rVP6,^77,78 P particle derived from a modified P domain of VP1,^79,80 and adenovirus- or alphavirus-vector expressing VP1 proteins^81,82 are also under development.

Challenges in developing norovirus vaccines are nicely summarized in two articles.^67,83 These challenges include virus heterogeneity and virus constant evolving, an absence of a conventional cell culture system, a lack of a small animal model, markers needed to separate host immune responses to infection versus vaccination, short-lived protective immunity, and the need to define immune correlates of protection to evaluate the efficacy of norovirus vaccine candidates. Currently, there are already three genogroups and 29 genotypes of noroviruses causing gastroenteritis. These viruses are continuously evolving through mutations as well as recombination from the circulating strains. That results in a constant emergence of new variants or strains⁸⁴ and potential escape of host immunity derived from vaccination or natural infection. On the other hand, the most vulnerable populations to norovirus gastroenteritis, infants, the elderly and immunocompromised patients have impaired immune function and often respond poorly to norovirus candidate vaccines, leading to a vaccine (otherwise a good candidate) with poor efficacy. The duration of host immunity against norovirus also needs to be better characterized. The short-lived protective immunity has largely discouraged efforts in developing norovirus vaccines since vaccines with short-term protection are only suit for travelers but not for the populations at the highest risk (young children, the elderly, and immunocompromised individuals) who need vaccination the most. Early studies suggested that host immunity last for only months after norovirus natural infection or administration,^85–88 though another study indicated that host immunity could last for at most several years.⁸⁹ Perhaps, the lack of an in vitro virus culture system and the absence of small animal models for immunogenicity and efficacy studies are key obstacles that prevent us from understanding pathogenesis of norovirus infection and making progress in vaccine development particularly of live attenuated vaccines. The lack of defined immune correlates to protection against norovirus infection is another key barrier for norovirus vaccine preclinical and clinical evaluation. Total serum antibodies specific to norovirus may not be necessarily correlated to protection against the disease. Only serum antibodies blocking the binding of virus-like particles to HBGA (histo-blood group antigen) were found to be associated with protection against norovirus.^{73–76,90,91} However, volunteers without or with low levels of HBGA due to an absence or a low expression of fucosyltransferase 2 gene (FUT2), despite they developed low serum antibodies after vaccination, were resistant to GI.1 or GII.4 norovirus infection,^91,92 misguiding or further complicating vaccine candidacy assessment.

Progress has been also made in vaccine development for the other enteric viruses including astroviruses (Astroviridae), adenoviruses (Adenoviridae), and sapoviruses (Caliciviridae). Other Astroviridae members such as VA-Like and MLB-like astroviruses, Picornaviridae (silivirus, cosavirus), and Parvoviridae families (bocaviruses, bufaviruses) are also isolated from patients (usually in infants and children) with gastroenteritis. Several subunit vaccines have been investigated for prevention against astrovirus infections. In particular, a trivalent subunit vaccine for hepatitis E virus, norovirus, and astrovirus was generated by fusing together the dimeric P domains of the three viruses to form a tetramer.⁹³ When intranasally administered to mice, this trivalent product induced significant neutralizing antibody responses to the P domains of all three viruses. Another subunit astrovirus vaccine candidate used the capsid protein (CP) of mink astrovirus elicited high levels of serum anti-CP antib odies and lymphoproliferation responses and also stimulated IFN-γ levels in mice.⁹⁴ Importantly, it was observed that virus shedding was suppressed and clinical signs including severe diarrhea were reduced in the litters born to the immunized mink mothers when challenged with a heterogeneous astrovirus strain.⁹⁴ Future human volunteer studies and clinical trials are needed to assess the efficacy of these vaccine candidates against viral gastroenteritis.

II. Vaccines and vaccine development for bacterial enteric pathogens

Shigella

Gram-negative non-motile, rod-shaped Shigella bacteria are the second and the first leading causes of diarrhea, respectively, in children aged 12–23 and 24–59 months in developing countries,² though a community-based study suggested the attribution of Shigella bacteria to children’s diarrhea may be less significant.⁶ Shigella is a genus of four species or groups (S. flexneri, S. sonnei, S. boydii, and S. dysenteriae) with more than 50 serotypes differentiated by lipopolysaccharide (LPS) O antigen specificity. After ingestion and passing through stomach and small intestines, Shigella bacteria invade host colon and rectum epithelial cells, multiply inside host cells, cause inflammation, induce cell death, spread to adjacent epithelial cells and destruct colon and rectum mucosal tissue,⁹⁵ resulting in shigellosis⁹⁶ or hemolytic uremic syndrome (HUS) with attribution from Shiga toxin (S. dysenteriae type 1).⁹⁷ Since only over ten (S. dysenteriae) to less than a couple of hundreds (S. sonnei & S. flexneri) of bacteria are needed to initiate shigellosis, person-to-person contact and even contact by flies can be important routes of transmission.^98–100

Identified virulence factors of Shigella include type III secretion system (T3SS) essential invasion plasmid antigens (Ipa A, B, C, D), intracellular spreading (Ics) proteins, LPS, and Shiga toxin. Ipa proteins kill macrophage and mediate bacterial adherence and invasion to colon and rectum epithelial cells, and Ics proteins (IcsA also called VirG, IcsB) assist Shigella bacteria for spreading to adjacent cells. Shigella LPS attributes to colon and rectum tissue damage and inflammatory responses, whereas Shiga toxin inactivates protein syntheses in eukaryotic cells. Clinical outcomes include fever, headache, vomiting, dehydration, and watery diarrhea but can progress to bloody and mucus-laden diarrhea, severe abdominal cramping, life-threaten dysentery, HUS, and death. Treatment of shigellosis is limited to supportive care to maintain hydration and to balance electrolyte levels. Antibiotic drugs are still prescribed to treat shigellosis for travelers and dysentery for children and immunocompromised individuals. However, antibiotics become less effective because Shigella spp. increasingly acquire antimicrobial resistance (AMR)^101–106or even worsen disease progression as some antibiotics promote Shiga toxin production.^107,108

There is no vaccine licensed for Shigella. The feasibility of vaccine protection against shigellosis, however, is suggested by field observation^109,110 and has been demonstrated in non-human primate¹¹¹ as well as controlled human infection studies,¹¹² evidenced by that pre-exposure to Shigella infection resulted in immunity and/or protection against subsequent infection or challenge. Moreover, serum antibodies to lipopolysaccharide (LPS) were revealed the correlates to protection.^113,114 Unfortunately, LPS O antigens are serotype specific. Heterogeneity of Shigella serotypes is indeed the main challenge in vaccine development. Fortunately, only couples of species and a handful of serotypes are highly prevalent in shigellosis. Of four Shigella spp., S. flexneri (mostly in low-income and lower middle-income countries; responsible for about two-thirds of all diarrheal episodes) and S. sonnei (middle-income and developed countries; responsible for about one-fourth of all episodes) attribute to 90% of shigellosis cases,¹¹⁵ with serotypes S. sonnei, S. flexneri 2a, 3a, and 6 for the vast majority of illnesses.^116,117 Current LPS O antigen-based Shigella vaccine candidates are primarily targeting on these serotypes. In contrast to the serotype-specific LPS O antigens, virulence factors Ipa proteins and VirG/IcsA are found conservative among Shigella species and serotypes.^116,118 Humans develop antibody responses specific to Ipa and VirG proteins after natural Shigella infection,^119–122 and antibodies derived from Ipa proteins were found to protect mice against pulmonary lethal challenge with different Shigella species or serotypes¹²³ or monkeys against severe diarrhea,¹²⁴ indicating Ipa and VirG alternative antigens for cross-protective Shigella vaccines.

Several Shigella vaccine candidates are under investigation. As briefly reviewed by Mani et al. recently,¹¹⁷ these include live attenuated or killed whole-cell candidates, glycoconjugate or bioconjugate candidates, as well as subunit vaccine candidates. While many are at the pre-clinical stage, a few candidates have proceeded to phase I, II, or even phase III studies.

WRSS (WRSS1, WRSs2, WRSs3)

WRSS1, a live attenuated S. sonnei strain (Mosley) with a part of virG gene deleted, was developed by Walter Reed Army Institute of Research (WRAIR). WRSS1 was demonstrated to protect against homologous challenge in guinea pig keratoconjunctivitis model.¹²⁵ WRSS1 was well tolerated and immunogenic in healthy adults from developed countries,^126,127 as well as adults and children in Bangladesh.¹²⁸ Unfortunately, WRSS1 at the regimen of a single dose of 1 × 10⁴ inocula showed no evident efficacy against homologous challenge in an endemic setting.¹²⁹ Subsequently, WRSs2 and WRSs3 were developed, by deleting enterotoxins (ShET1 and ShET2) or further reducing LPS endotoxicity of WSSR1.¹³⁰ WRSs2 and WRSs3 were well tolerated and induced immune responses in non-human primates¹³¹ and healthy adults,¹³² but they still showed reactogenicity though at a reduced rate.¹³² The efficacy of WRSs2 or WRSs3 against homologous (or heterologous) challenge is yet to be demonstrated.

CVD 1208S

CVD 1208S is derived from reconstruction of CVD 1208, a live attenuated S. flexneri 2a mutant with deletion of guaBA locus regulating synthesis of inosine 5ʹ-monophosphate dehydrogenase (encoded by guaB) and guanosine 5ʹ-monophosphate synthetase (encoded by guaA),¹³³ enterotoxin gene set (ShET1) and invasiveness plasmid gene sen (ShET2),¹³⁴ in animal-free medium.¹³³ CVD 1208 and 1208S were generally well tolerated in healthy adults, and induced robust anti-LPS IgA and IgG antibody-secreting cells (ASC) responses, moderate serum IgG or fecal IgA, but mild serum IgA responses. Currently, there is no published efficacy data available from either candidate strain.

Other live attenuated Shigella vaccine candidates currently under pre-clinical studies include typhoid vaccine Ty21a expressing Shigella LPS O-antigens (also ETEC adhesin and toxoid antigens)^135–137 and ShigETEC. ShigETEC applied a modified noninvasive Shigella strain Vadizen (Istrati-32) to express ETEC toxoid fusion antigen LTB-STa_N12S.¹³⁸

Inactivated Shigella whole-cell vaccine candidates

Inactivated whole-cell candidates were also investigated for protection against shigellosis. SsWC, formalin-inactivated S. sonnei, which showed to be immunogenic and protective against keratoconjunctivitis in guinea pigs when challenged with S. sonnei, was well tolerated in healthy adults.¹³⁹ Realized that SsWC is unlikely to provide cross-protection against other Shigella spp. or different serotypes, researchers added two more formalin-inactivated Shigella serotypes, S. flexneri 2a and 3a, to SsWC to produce a trivalent Shigella whole-cell vaccine candidate. This trivalent product induced anti-Ipa (B, C, D) and serotype-specific anti-LPS antibody responses and protected guinea pigs from three homologous serotypes in the keratoconjunctivitis model.¹⁴⁰ Additionally, two other inactivated whole-cell products were investigated. Inactivated ∆wzy Shigella flexneri 2a mutant which had O-antigen polysaccharide polymerase gene disrupted induced immunity cross-protective against serotypes 2a, 3a, and 6 in a mouse pulmonary model.¹⁴¹ Heat killed multi-serotype Shigella (HKMS), which combined six Shigella serotypes (S. dysenteriae type 1, S. flexneri 2a, 3a, 6, S. sonnei, S. boydii), protected guinea pigs in a recto-colitis model,¹⁴² neonatal mice passively,¹⁴³ and rabbits¹⁴⁴ against challenge of all six homologous strains.

Polysaccharide-based subunit vaccine candidates

Shigella LPS O-specific polysaccharide moiety (O-antigen) chemically or biologically conjugated to a carrier protein was attempted to improve O-antigen safety and immunogenicity in developing Shigella polysaccharide conjugate vaccines. Indeed, some conjugate vaccine candidates have been investigated more extensively or are advanced the most in clinical studies. S. sonnei-rEPA, LPS O-antigen specific to S. sonnei chemically conjugated to recombinant exoprotein A of Pseudomonas aeruginosa (rEPA), was found safe and immunogenic^145,146 and protective against S. sonnei shigellosis in healthy adults¹⁴⁷ as well as children above three years of age, but little protection was observed from the children aged less than three years.¹⁴⁸ Other carrier proteins including CRM9 of Corynebacterium diphtheriae and a succinylated nontoxic peptide of Clostridium difficile toxin A were examined for enhancement at Shigella O-antigen immunogenicity.^149,150

Different from chemical conjugation, bioconjugation uses E. coli N-linked protein glycosylation machinery to conjugate in vivo Shigella LPS O-antigens to a carrier protein.^151,152 To bioconjugate Shigella O-antigens to carrier protein exoprotein A of P. aeruginosa (EPA), a modified EPA gene (harbor N-glycosylation consensus sequences) and a Shigella O-antigen gene were expressed in E. coli. The O-antigen repeating units preassembled in cytoplasm flip to periplasm and polymerize into O-antigen chain. These O-antigen chains then link to EPA protein to form bioconjugates. Bioconjugates are finally released to outer membrane after cell lysis.¹⁵³ Bioconjugate vaccine candidates for S. dysenteriae (SD133) and S. flexneri 2a (Flexyn2a) were demonstrated safe and immunogenic in phase I trials.^154,155

Well-defined synthetic oligosaccharides were investigated as Shigella O-antigen surrogates to improve conjugate vaccine candidacy. It was reported that synthetic saccharides of S. dysenteriae induced anti-LPS antibodies after being conjugated to human serum albumin¹⁵⁶ and a synthetic pentadecasaccharide mimicking S. flexneri 2a O-antigen induced protective antibodies when was conjugated to tetanus toxoids.¹⁵⁷ However, the feasibility of using synthetic O-antigen surrogates in developing Shigella vaccines is yet to be fully established.¹⁵³ Additionally, S. sonnei 1790GAHB GMMA (generalized modules for membrane antigens) vaccine, blebs or native outer membrane vesicles shed by a genetically modified S. sonnei strain for a high level of production of LPS blebs,¹⁵⁸ was generally safe and induced robust antibody response to S. sonnei LPS in adults from an Africa endemic country.¹⁵⁹

Since Shigella O-antigen moieties of LPS are diverse due to variations of repeating units (two to six monosaccharides) thus are serotype specific, anti-LPS antibodies derived from an O-antigen of one Shigella serotype protect against homologous infections but cannot provide cross-protection against heterogeneous serotypes. In addition, conjugation of O-antigens with a carrier protein requires preparations and characterizations of the antigen and the carrier protein separately. Difficulties in conjugate purification and variations in bioconjugation process often lead to a lower conjugation efficiency, resulting in a lower production yield of bioconjugates. Moreover, the low efficiency of bioconjugation leaves a high proportion of carrier protein unglycosylated. This potentially mistakes host immune systems to respond to the carrier protein rather than the O-antigen, further hampering efforts in developing low-cost Shigella bioconjugate vaccines to be licensed in developing countries.

Protein-based subunit vaccine candidates

Protein-based and protein-LPS-based Shigella vaccine candidates under preclinical studies are mainly based on invasion plasmid antigens (Ipa proteins) or a complex of invasins composed of Ipa proteins and LPS. Like LPS, Ipa proteins (IpaB, C, D) induced protective antibodies against Shigella infections.^118,160,161 However, differed from the serotype-specific LPS, Ipa proteins are highly conserved among Shigella species and serotypes and even in other gram-negative bacteria,¹⁶² thus become leading antigens for the development of cross-protective Shigella vaccines. DBF, a genetic fusion of proteins IpaD and IpaB was demonstrated to induce antibodies not only broadly protective against lethal challenge of S. flexneri, S. sonnei, and S. dysenteriae in mice (pulmonary infection model)¹²³ but also against severe diarrhea in non-human primates (JD Clements, personal communication). Invaplex, a complex of invasins consisting IpaB, C, D proteins and LPS water-phase extracted from Shigella flexneri culture growth, was found to induce protective antibodies to the invasin complex and LPS in mice and guinea pigs.¹¹⁸ After initial examination of safety and immunogenicity, this protein-LPS-based subunit vaccine candidate was further modified to improve immunogenicity and more importantly efficacy.^161,163-165 However, the composition of the complex of invasins was never well defined and reproducibility remained as a challenge. A recent study revealed that Invaplex_AR carrying recombinant IpaB and IpaC proteins and LPS purified from S. flexneri 2a and S. sonnei had immunogenicity improved and protected mice better against body weight loss.¹⁶⁶ Data from future safety and immunogenicity studies from human subjects and eventual field studies will provide better assessment of Invaplex_AR or DBF subunit vaccine candidacy for Shigella.

Enterotoxigenic Escherichia coli (ETEC)

ETEC producing heat-stable toxin (STa) and/or heat-labile toxin (LT) is among the top five causes of children’s diarrhea from hospital-setting and community-setting surveillence.^2,6 ETEC is also the leading cause of diarrhea in international travelers including deployed civilian and military personnel (travelers’ diarrhea).^167–169 Bacterial adhesins (fimbrial and non-fimbrial) and enterotoxins are the two types of key virulence factors associated with ETEC infection.^170,171 ETEC bacteria produce over 25 adhesins to attach to different host receptors and two very distinctive enterotoxins (LT, STa) to disrupt homeostasis in host small intestinal epithelial cells to cause watery diarrhea. Ingested in contaminated food or water, ETEC bacteria pass host stomach, reach to small intestines, attach to host receptors, and colonize small intestines. Colonization brings ETEC in a proximity to effectively deliver enterotoxins into small intestinal epithelial cells. Enterotoxins elevate epithelial cells intracellular cyclic adenosine monophosphate (cAMP; by LT) and cyclic guanosine monophosphate (cGMP; by STa), leading to fluid hypersecretion and watery diarrhea or severe diarrhea like cholera. Without medical intervention, disease progresses to severe dehydration and causes acute death in young children.

There are no vaccines licensed for ETEC. Challenges in developing effective ETEC vaccines historically include immunological heterogeneity of ETEC pathotypes, identifying proper antigens to target enterotoxins especially the potent and poorly immunogenic STa toxin (until recently), and a lack of suitable animal models to assess vaccine efficacy.^171–175 An ideal ETEC vaccine would prevent all virulent ETEC pathotypes from colonizing host small intestines and neutralize enterotoxicity of both ETEC toxins. However, under current technologies, developing such a vaccine for ETEC seems unfeasible. A more practical strategy is to develop a vaccine cross-protective against the most prevalent and virulent ETEC pathotypes which cause moderate-to-severe infection.¹⁷³ ETEC pathotypes producing seven colonization factor antigen (CFA) or coli surface (CS) antigen adhesins [CFA/I, CFA/II (CS1-CS3), or CFA/IV (CS4-CS6)] and enterotoxin STa and/or LT are the most important pathotypes for children’s diarrhea and travelers’ diarrhea.^176–178 These pathotypes become the primary targets in current ETEC whole-cell (killed or live attenuated) and subunit vaccine development.

ETVAX

A killed whole-cell cocktail vaccine candidate is derived from rCTB-CF, an early killed whole-cell product.¹⁷⁹ Consisted of three ETEC strains expressing four adhesins (CFA/I, CS1, CS2, CS3) and an addition of recombinant B subunit of cholera toxin (rCT_B) initially, rCTB-CF was afterward modified to carry five killed strains expressing six adhesins (CFA/I, CS1-CS5) (plus recombinant rCT_B protein). Cholera toxin B subunit (CT_B) was the immunogen in cholera vaccine Dukoral which induced cross-protective immunity against LT-producing ETEC,¹⁷⁹ due to the homology between LT B subunit (LT_B) and CT_B (85% homology of amino acids).¹⁸⁰ Orally administered rCTB-CF was tolerated and induced antigen-specific antibody responses in healthy adults^179,181-183 and children aged above 18 months to two years in ETEC endemic countries.^184,185 When given to Egyptian children less than two years of age, rCTB-CF-induced responses in only 30% of the children.¹⁸⁴ Subsequent efficacy studies, however, showed that rCTB-CF provided little protection to Egyptian infants against ETEC diarrhea¹⁸⁶ or only protection against the severity of diarrhea but not the overall diarrhea rate in the US healthy adults traveling to Mexico and Guatemala.¹⁸⁷ To further improve anti-CF and anti-LT immunogenicity and to be suitable to vaccinate young children, ETVAX was developed as the next generation vaccine to replace rCTB-CF. ETVAX consists of four inactivated recombinant E. coli strains that hyper-express CFA/I, CS3, CS5, and CS6 adhesins and recombinant subunit protein LCTBA (CT_B subunit with seven amino acids replaced by the counterpart residues of LT_B subunit).^180,188 ETVAX was shown safe and broad immunogenic in healthy adults from developed countries and ETEC endemic countries.^189,190 Interestingly, ETVAX-induced antibodies were found to cross-react to CS1, CS14, CS17, and CS7 adhesins.¹⁹¹ A recently completed phase I study showed that ETVAX induced broad antigen-specific responses in Bangladesh children older than 12 months and moderate or mild responses in children less than 12 months of age.¹⁹² However, protective efficacy of ETVAX against ETEC diarrhea in the endemic setting for children especially those aged less than 12 months is yet to be evaluated. Because it does not induce protective immunity against the key STa toxin which plays a more important role in children’s diarrhea, ETVAX likely will encounter challenge in providing broad protection against ETEC diarrhea particularly diarrhea caused by STa+ ETEC strains.¹⁹³

ACE527

A live attenuated vaccine candidate consists of three live strains, ACAM 2025 (CFA/I, LT_B), ACAM 2022 (CS5/CS6, LT_B) and ACAM 2027 (CS1/CS2/CS3, LT_B).^194,195 When given two doses of about 10¹¹ CFUs (~3x10¹⁰ CFUs per strain), ACE527 induced vigorous immune responses but unfortunately also side effects, mainly vomiting in some adult subjects.¹⁹⁶ In a following double-blind placebo-controlled Phase IIb study, this live vaccine candidate (at a dose of 2 × 10¹¹ CFUs) reduced the severity of diarrhea and bacteria shedding in volunteers when challenged with ETEC strain H10407 producing STa and LT.¹⁹⁷ Because of the side effects and unsatisfied efficacy, ACE527 was withheld temporarily from further evaluation. A later study using a three-dose regimen at a lower dosage (10¹⁰ CFUs, ~3 × 10⁹ CFUs per strain) and the addition of adjuvant dmLT, however, showed 65.9% efficacy against severe diarrhea and 58.5% against diarrhea of any severity in the healthy adults challenged with H10407.¹⁹⁸ Interestingly, the same study found that adults vaccinated with ACE527 without adjuvant dmLT were not protected (23.1% efficacy against severe diarrhea). This brings up a question regarding the protective efficacy from the antibodies derived from adjuvant dmLT (which is also an antigen for an ETEC vaccine) against H10407 challenge and the definitive role dmLT played in this vaccine candidate (as an adjuvant, an antigen, or both). It seemed to us that dmLT played a more important role as an adjuvant for ACE527 because the additional anti-LT antibodies from dmLT adjuvant play little role against STa toxin from challenge strain H10407. Surely, ACE527 live attenuated ETEC vaccine candidate needs to be further evaluated for efficacy in adults and more importantly young children from ETEC endemic countries. However, ACE527, like the killed whole-cell candidate ETVAX, does not carry STa antigens to induce protective anti-STa immunity thus becomes limited in protecting against STa-producing ETEC infection.

ETEC subunit vaccine candidates

There are a few protein-based subunit vaccine candidates under development. CFA/I minor subunit tip adhesin CfaE protein was used for the development of a tip adhesin vaccine against ETEC producing CFA/I adhesin or adhesins antigenically homologous to CFA/I. Non-human primates intranasally immunized with CfaE tip adhesin protein developed antibodies that inhibit CFA/I agglutination to bovine erythrocytes¹⁹⁹ and further protect against ETEC H10407 challenge.²⁰⁰ Additionally, adults passively acquired hyperimmune bovine IgG raised against CfaE tip adhesin protein were significantly protected against H10407 challenge (63% efficacy).²⁰¹ These data suggested that tip adhesin CfaE-induced immunity protects the homologous challenge strain (H10407) and possibly ETEC strains expressing adhesins homologous to CFA/I, Class 5 ETEC fimbrial adhesins.²⁰² Unfortunately, CfaE tip adhesin vaccine is unlikely to protect against ETEC strains that express immunologically different adhesins.

MecVax, a multivalent ETEC vaccine candidate is composed of LT-STa toxoid fusion 3xSTa_N12S-mnLT_R192G/L211A protein and CFA/I/II/IV multiepitope fusion antigen (MEFA) protein.²⁰³ Toxoid fusion protein 3xSTa_N12S-mnLT_R192G/L211A, which carries three copies of STa toxoid STa_N12S (the 12^th residue asparagine was replaced with serine) and a monomeric peptide composed of one mutant LT A subunit and one LT B subunit,²⁰⁴ induced antibodies that neutralize STa and LT enterotoxicity^204-206 and protect against STa+ and LT+ ETEC diarrhea in a pig challenge model.^207,208 CFA/I/II/IV MEFA, an epitope- and structure-based recombinant protein carries neutralizing epitope of the seven most important ETEC adhesins (CFA/I, CS1-CS6), induced antibodies that inhibit adherence of ETEC expressing any of these seven adhesins.^209–211 Because they are thermostabile and importantly antigenic compatible, 3xSTa_N12S-mnLT_R192G/L211A and CFA/I/II/IV MEFA can be combined and co-administered,²¹² leading to the production of a multivalent ETEC subunit vaccine, MecVax. When parenterally (intramuscularly, intradermally, subcutaneously) administered, MecVax induced antibodies that neutralized STa and LT enterotoxicity and inhibited adherence of all seven adhesins (CFA/I, CS1-CS6). Most importantly, MecVax protected rabbits from ETEC colonization in small intestines and pigs from ETEC diarrhea.²⁰³ MecVax will be prepared under good manufacturing practice (GMP) production and evaluated in future efficacy studies to determine its ETEC vaccine candidacy.

Other ETEC subunit vaccine candidates currently under preclinical studies include newly identified conservative antigens (EtpA, EatA, EaeH, YghJ)^213,214 and STa toxoid conjugates.^215,216

Diarrheagenic E. coli other than ETEC, including enteropathogenic E. coli (EPEC) and enterohemorrhagic E. coli (EHEC), are also causes of enteric infections. EPEC strains are listed among the top ten causes of diarrhea in children under the age of two years,^2,6 but there are no vaccines developed for EPEC either. Ineffective diagnostics of EPEC disease outcomes and a lack of sufficient understanding at EPEC disease mechanisms and host immune responses are among the obstacles in EPEC vaccine development.²¹⁷ Unlike EPEC or ETEC, EHEC causes only sporadic outbreaks mainly due to consumption of undercooked beef products or raw vegetables. Though efforts are continuously made,²¹⁸ EHEC vaccine development for humans is discouraged because of a low incidence rate.

Cholera

An acute watery diarrhea caused by Vibrio cholerae remains a significant public health burden globally but more particularly in South Asia and sub-Sahara Africa.²¹⁹ Over 200 serogroups have been identified based on the variations of bacterial surface LPS. Historically, classical serogroup O1 caused pandemic cholera. Serogroup O139 emerged as a dominant cause of epidemic outbreaks in Asia in the early 1990s but became quiescent afterward, whereas O1 serogroup has reemerged as the primary agent of cholera. Serogroup O1 has two biotypes, classical and El Tor, and each biotype includes two serotypes (Inaba and Ogawa; rarely third serotype Hikojima). Transmitted through ingestion of contaminated water primarily and food as well, Vibrio cholera bacteria pass through stomach, colonize small intestine, deliver cholera toxin (CT; structurally and functionally homologous to heat-labile toxin of enterotoxigenic E. coli – ETEC) into intestinal epithelial cells to disrupt homeostasis in epithelial cells and to cause water hypersecretion and secretory diarrhea, often leading to dehydration and acute death if not treated.

Unlike ETEC or Shigella of which there are no vaccines licensed, cholera has three killed whole-cell oral vaccines (OCV) prequalified by WHO and licensed by many national regulatory authorities: Dukoral, Shanchol, and Euvichol (Table 1).

Dukoral

Also known as BS-WC, WC-BS, WC-rBS or CTB-WC, is a monovalent vaccine product initially prepared by Swedish National Bacteriological Laboratory (SBL). This whole-cell vaccine consists of 1 mg CT B subunit recombinant protein (rBS) and killed V. cholerae O1 (10¹¹ CFUs, four strains, two classical biotypes and two El Tor biotypes, 2.5 × 10¹⁰ CFUs of each biotype). Dukoral is orally administered in sodium bicarbonate buffer (against stomach acid degradation of CT_B). The initial field trial with a regimen of three doses at a six-week interval indicated that Dukoral provided 85% efficacy against cholera for six months in all children (2–5 years and above 5 years) and women above 15 years.²²⁰ During the first year following Dukoral vaccination, hospital visits for treatment of diarrhea were reduced by 26% and severe dehydrating and fatal diarrhea were decreased by 49%.²²¹ However, at 36 months following immunization, the efficacy of Dukoral decreased to 26% for children aged 2 to 5 years.²²²

Shanchol

From the field study of Dukoral, Clemens et al. observed that children aged 2–15 years and women above 15 years administered with the whole-cell vaccine without CT_B protein were also protected short term (six months) against cholera, though at a lower efficacy (58%).²²⁰ Without CT_B recombinant protein, bicarbonate buffer (which was used mainly to prevent CT_B protein from stomach acid degradation) was not needed for vaccine administration. This significantly eased the vaccine preparation and administration burden but also reduced vaccine manufacture complexity and cost. To produce an inexpensive oral whole-cell cholera vaccine against O1 and also O139 which emerged in the early 1990 s, VaBiotch (Vietnam) worked with International Vaccine Institute (IVI; South Korea) and developed a bivalent whole-cell cholera vaccine (ORC-Vax, mORCVAX), by adding the serogroup O139 component. ORC-Vax composed of four O1 serogroup strains (heat-killed classical Inaba and Ogawa, formalin-inactivated El Tor Inaba and classical Inaba) and one O139 serogroup strain, without CT_B recombinant protein, was included in EPI in Vietnam.²²³ Reformulated to remove toxin content,²²⁴ ORC-Vax was transferred to Shantha Biotechnics in India for good manufacture production and WHO pre-qualification,²²⁵ and was marketed as Shanchol afterward.²²⁶ Shanchol was demonstrated safe and immunogenic in adults and children aged over 12 months.²²⁶ At a regimen of two doses (14 days apart), Shanchol conferred 67%, 66%, and 65% cumulative protective efficacy, respectively, in two, three, and five years postvaccination among Kolkata residents aged one year and older.^227–229 The field study carried out in Dhaka, however, reported Shanchol of only 37% overall protective efficacy.²³⁰ When a single dose regimen was used, however, this vaccine failed in providing adequate protection to young Dhaka children (16% efficacy).²³¹

Euvichol (Euvichol-Plus)

To increase vaccine production for stockpile, IVI also transferred the killed oral cholera vaccine technology to EuBiologics in Korean for fermentation manufacture of a bivalent cholera vaccine based on the same formulation of Shanchol. The product was marketed as Euvichol. Euvichol was demonstrated to be well tolerated and immunogenic.^232,233 At a two-dose regimen, Euvichol induced vibriocidal responses at a level comparable to Shanchol.²³⁴ Modified by removing thimerosal preservative and packing in plastic tubes instead of glass vials, Euvichol becomes more efficient for vaccine stockpile, transportation, and administer. This modified Euvichol was subsequently labeled as Euvichol-Plus. With an easy delivery and at a lower cost, Euvichol-Plus gains attraction, though efficacy data from field trials are yet to be demonstrated. Euvichol-Plus is in the process to be transferred to Shantan Biotechnics and will be marketed as Vibochol.

Live attenuated cholera vaccines

A few live attenuated cholera strains were developed. CVD 103-HgR (Vaxchora), which was derived from an O1 classical strain with the deletion of cholera toxin A subunit (CT_A) and the inclusion of a mercury resistance gene, was shown initially not only tolerable and immunogenic in North American volunteers²³⁵ but also protective against cholera challenge.^236,237 However, CVD 103-HgR was found not protective against cholera in a subsequent large, randomized, double-blind, and placebo-controlled field study in Indonesia, possibly due to a low incidence of cholera in the country.²³⁸ Peru-15 (CholeraGarde) is another well-studied live attenuated oral cholera vaccine. Peru-15 is derived from O1 El Tor Inaba strain²³⁹ with a series of gene deletion.²⁴⁰ Peru-15 at a single dose (5x10⁸ – 1 × 10⁹ CFUs) was found to be well tolerated and immunogenic in adults, children, and infants from Bangladesh,^241–243 and showed an efficacy of over 90% against diarrhea from a challenge study in North American healthy adults.²⁴¹ However, Peru-15 needs to be studied in the cholera epidemic setting to assess better its candidacy against cholera.

Campylobacter

Campylobacter infections were reported about 14.3 cases per 100,000 population in the US annually.²⁴⁴ FoodNet confirmed that Campylobacter was the leading cause of foodborne illness in the US, accounting for the incidence rate of 19.5 per 100,000 population in 2018.²⁴⁵ Campylobacter infection rates vary, ranging from 5% to 18% in laboratory-based surveillance in developing countries.²⁴⁶ Major sources of Campylobacter infections are contaminated water and food including incorrectly prepared poultry meat and unpasteurized milk. Patients typically experience mild to severe diarrhea, bloody diarrhea, stomach pain, nausea, and/or vomiting lasting four to seven days. A dozen of Campylobacter species were isolated, but C. jejuni and C. coli are responsible for the most gastroenteritis cases.²⁴⁷ Infections are initiated with Campylobacter bacteria penetrating host gastrointestinal mucus and invading intestinal epithelial cells. That leads to epithelial tissue damages and diarrhea.

The identified virulence factors of C. jejuni include cell adhesion and invasion proteins, capsular polysaccharide (CPS), cytolethal distending toxin (CDT A, B, C), lipooligosaccharide (LOS), and flagella.²⁴⁸ Fibronectin-binding protein CadF,²⁴⁹ aspartate/glutamate-binding protein PEB1²⁵⁰ and surface-exposed lipoprotein JlpA²⁵¹ were also indicated to be involved in adhesion of Campylobacter bacteria to host cells. C. jejuni CPS, a unique antigenic determinant in Penner serotyping scheme for Campylobacter, evades host immune response thus is potentially a key vaccine component. CDT is a multi-subunit protein and induces DNA damage in host epithelial cells to cause apoptotic cell death.²⁵²

Preventing Campylobacter infections has been focused largely to decrease contamination at meat products during processing. Since poultry meat is the common source of infections, many preventive measures have been taken to inhibit Campylobacter colonization in the intestine tract of broiler chickens. Several vaccine candidates were tested in animal models, with a few advanced to phase 1 clinical trial.

Inactivated whole-cell vaccine candidates

Killed C. jejuni whole-cell vaccine (CWC) was developed and first evaluated in the early 1990s. CWC was prepared by formalin inactivation of C. jejuni strain 81–176 originally isolated from a child with acute diarrhea after consumption of contaminated raw milk.^253,254 A pre-clinical study showed that mice orally immunized with CWC adjuvanted with E. coli heat-labile enterotoxin (LT) (CWC-LT) developed high levels of intestinal sIgA and IgG responses to Campylobacter and LT.²⁵⁴ Moreover, inhibition of C. jejuni colonization was observed in the immunized mice when challenged with the homologous strain four weeks post CWC immunization.²⁵⁴ Rhesus monkeys immunized with CWC-LT developed systemic IgA and IgG to Campylobacter antigens. These promising findings led to safety and immunogenicity phase 1 clinical study.²⁵⁵ However, the followed controlled human infection model (CHIM) study suggested that this vaccine candidate was not protective against infection even from homologous strain C. jejuni 81–176.²⁵⁶

Protein-based subunit vaccine candidates

C. jejuni surface antigens were used for subunit vaccine development. The rFlaA-MBP, a flagellin-based vaccine candidate was prepared by fusing the conserved regions of C. coli VC167 FlaA flagellin with E. coli maltose-binding protein (MBP).²⁵⁷ Mice intranasally immunized with rFlaA-MBP developed antigen-specific serum IgG and intestinal secretory IgA antibodies. When challenged intranasally 26 days after immunization, mice were protected from Campylobacter colonization. However, a followed clinical trial failed to provide protection,^256,257 resulting in rFlaA-MBP abandoned from further development.

Protein-polysaccharide conjugate vaccine candidates

Protein-polysaccharide conjugate vaccines were prepared by conjugating Campylobacter polysaccharides to a protein carrier. CPS_81-176-CRM₁₉₇ conjugate vaccine candidate (CJCV1) had the capsule polysaccharide (CPS) of C. jejuni 81–176 (HS23/HS36 serotype complex) crosslinked to well-characterized diphtheria toxin mutant CRM₁₉₇.²⁵⁸ When administered subcutaneously in mice, CJCV1 (CPS_81-176-CRM₁₉₇) induced robust serum IgG response to CPS81-176. This vaccine candidate was found to significantly decrease illness in mice and to protect the immunized New World monkey (Aotus nancymaae) from diarrhea after intranasal or orogastric challenge with homologous strains C. jejuni.²⁵⁸ These results encouraged a phase 1 trial, in which 48 volunteers received two intramuscular vaccinations, 4 weeks apart, with or without adjuvant Alhydrogel.²⁵⁹ Unfortunately, results showed that CJCV1 did not elicit robust serum IgA and IgG against the CPS moiety,²⁵⁹ likely due to the lack of immunodominant nonstoichiometric O-methyl phosphoramidate (MeOPN) epitopes in CJCV1. Subsequently, new candidate CJCV2 was prepared from C. jejuni 81–176 strain by stably expressing phase-variable MeOPN transferases. CJCV2 is under manufacture for clinical evaluation.²⁵⁹

Nontyphoidal Salmonella

Nontyphoidal salmonellosis (NTS) refers to the disease caused by any serotypes of Salmonella except for typhoidal Salmonella such as S. Typhi and Paratyphi A, B, and C. Two main serovars associated with enteric infections are S. Typhimurium and Enteritidis, causing an estimated 153 million cases of gastroenteritis and an annual death rate of 57,000 deaths globally.²⁶⁰ In the United State, NTS is a leading cause of hospitalization and death due to foodborne illness, with 33% of the 1.2 million NTS cases are due to consumption of contaminated beef, pork, and poultry meat.²⁶¹ NTS infection results in self-limiting diarrhea in immunocompetent hosts; however, bacteremia occurs in approximately 5% of immunologically compromised individuals or patients with gastrointestinal illness.²⁶² Invasive NTS (iNTS) bacteremia often kills children who suffer from malnourishment or AIDS in the developing world, especially in sub-Saharan Africa, where multilocus sequence type ST313 has been dominant in association with bacteremia.^263,264 iNTS isolates within ST313 and S. Typhi might have similar virulence properties and process of adaption to the human host.²⁶⁵ iNTS isolates adhere to the mucosa, invade host small intestinal epithelial cells and then penetrate the mucosal barrier. Once iNTS bacteria are internalized through the M cells in the Peyer’s patches, they are translocated to the intestinal lymphoid follicles, spread to the liver, spleen, or bone-marrow, and disseminated throughout the body, leading to fatal illness by bacteremia.²⁶⁵

Whole-cell vaccine candidates

While licensed vaccines against typhoid Salmonella have been licensed and used globally, there are no vaccines available for NTS. Several vaccine candidates against S. Typhimurium and Enteritidis are under development, but only one NTS vaccine, a live attenuated WT05, has been completed for a Phase 1 study.²⁶⁶ Microscience Limited reported double mutant S. Typhimurium strain WT05 constructed by deleting aroC (encoding enzymes of the prechorismate biosynthetic pathway)²⁶⁷ and ssaV (encoding a protein of the SPI-2 that involves in invasion of host cells)²⁶⁶ genes were attenuated and immunogenic in humans. Adult volunteers receiving 10⁷ to 10⁹ CFU of WT05 showed no serious adverse effects and mounted S. Typhimurium LPS-specific antibody-secreting cell responses with variable serum IgG and IgA antibodies. However, prolonged stool shedding of WT05 was detected in healthy volunteers for up to 23 days post inoculation.²⁶⁶ Other live attenuated NTS vaccine strains include CVD 1921 and CVD 1923.²⁶⁸ CVD 1921 and CVD 1923 were derived from the invasive wild-type strains S. Typhimurium I77 by deleting genes guaBA (encoding guanine biosynthesis) and clpP (encoding a master protease regulator), or additionally fliD (encoding a capping protein). Oral LD₅₀ analyses confirmed that CVD 1921 (∆guaBA ∆clpP) or CVD 1923 (∆guaBA ∆clpP ∆fliD) were highly attenuated in mice; moreover, 80% to 86% of orally immunized mice were protected against lethal challenge from wild-type S. Typhimurium I77. Additionally, strain CVD 1941 with deletions of guaBA and clpP genes from wild-type S. Enteritidis R11 was also shown attenuated in mice and protected 76% of the immunized mice against S. Enteritidis infection.²⁶⁸

Protein-polysaccharide conjugate vaccine candidates

Glycoconjugate vaccines are also under development for NTS.²⁶⁹ O polysaccharides linked to core polysaccharides (COPS) of S. Enteritidis purified from the genetically attenuated strain CVD 1941 was conjugated to flagellin protein at various polysaccharide/flagellin ratios and examined for enhancement of polysaccharide immunogenicity.²⁶⁹ Mice intramuscularly immunized with COPS-flagellin conjugates developed high anti-LPS and anti-flagellin IgG, and the derived antibodies mediated opsonophagocytosis, an effective antibacterial mechanism against both complement-resistant and complement-sensitive NTS strains. Moreover, a followed challenge study showed mice immunized with COPS-flagellin conjugates, at different ratios of polysaccharide/flagellin, were protected against lethal challenge from wild-type S. Enteritidis R11 at the efficacy ranged from 83% to 100%.²⁶⁹

Protein-based subunit vaccine candidates

Genetic fusions of Salmonella pathogenicity island 1 (SPI-1) or 2 (SPI-2) tip and translocator proteins were applied to develop broadly protective subunit vaccines (S1 and S2) for NTS.²⁷⁰ S1 is the fusion of SipD and SipB, the tip and the translocator of the SPI-1 type three secretion system (T3SS), whereas S2 is the fusion of SseB and SseC, the tip and the translocator of the SPI-2 T3SS. When intramuscularly immunized and adjuvanted with Alhydrogel and monophosphoryl lipid A, S1 or S2 induced immunity to protect 60% of the immunized mice against S. enterica challenges. The immunized mice also had a reduction of cecal inflammation when challenged with homologous strain S. Typhimurium wild-type SL1344 or heterologous strain S. Enteritidis wild-type 125109. Both S1 and S2 candidates elicited high levels of IgG responses to SipB and SipD, or SseB and SseC, along with IgG-secreting cells in mouse spleens and bone marrow. However, no fecal IgA or antigen-specific IgA secreting cells in spleen or bone marrow were detected from the immunized mice.²⁷⁰

Vaccinology for multivalent vaccine development

Heterogeneity among viral genotypes or genogroups and bacterial pathotypes or serogroups is among the key challenges in developing vaccines against gastroenteritis. An effective vaccine needs to induce broad immunity cross-protective against heterogeneous viral or bacterial strains. To overcome this heterogeneity challenge, different vaccinology strategies have been attempted. This includes whole-cell cocktail vaccines which mix together several strains or isolates (inactivated or live attenuated) to expand vaccine coverage, vaccines which carry antigens conserved among different strains for generic immunity, and vaccines combining neutralizing epitopes or antigen domains representing various heterogeneous groups to induce an array of immunity for broad protection. Conventional whole-cell vaccine approach remains much practical for the pathogens of which disease mechanism is not fully characterized and virulence determinants are yet to be determined or safely included. With a serial of passages of pathogenic viral strains in vitro or in vivo, or systematic deletions of virulence genes, attenuate strains or mutant strains can be produced. Attenuated strains can also be obtained from animal originated strains (Jennerian vaccinology) or avirulent strains isolated from asymptomatic children or adults. These strains mimic the virulent strains in inducing host immune responses but do not cause infections, becoming whole-cell vaccine candidates. More specifically, systematically comparative examination of virulence significance among deletion mutants helps to identify pathogen virulence factors, leading to vaccines to target on specific virulence determinants. Whole-genome sequencing without prior knowledge of virulence determinants has reversed the conventional vaccine development process. Reverse vaccinology has a great potential to substantially accelerate vaccine development. By analyzing genome-wide sequence differences of virulent strains and avirulent strains, we can rapidly identify the putative virulence genes associated with pathogenicity and initiate vaccine development.^271,272 More recently, advance in structural biology and computational biology including protein modeling further allows us to in silico identify immuno-dominant B-cell and T-cell epitopes or peptide domains and to develop safe and precision vaccines (Figure 1).

Combining a few live attenuated or inactivated strains certainly broadens immunogenicity spectrum but can introduce unwanted effects. Excessive somatic antigens carried by a cocktail whole-cell vaccine product increase the risk of side effects and potentially deviate host immune responses away from targeting to specific virulence antigens. That leads to adverse effects and likely lower host immune responses, thus vaccine safety concern and unsatisfied protective efficacy. Additionally, a vaccine carrying multiple viral or bacterial strains introduces complexity and difficulties to formulation and manufacture, risking vaccine manufacturability and product affordability particularly to the most needed populations in the low-income countries. Using antigens conserved inclusively among virulent strains, particularly in the format of recombinant proteins, on the other hand, largely eliminates the negative impacts associated with excessive somatic antigens but also less likely alters host mucosal microbiome. However, identifying conservative antigens that induce protective immunity specific to enteric pathogens can be difficult potentially. Additionally, protein-based vaccines often need adjuvants (if the target recombinant proteins possess no adjuvanticity) and booster immunizations to improve induction of host immune responses, may also encounter challenges in inducing robust local mucosal immunity against enteric infections as well as concerns of vaccine manufacturability and affordability if multiple conservative antigens are needed.

An epitope-based and structure-based vaccinology named multiepitope-fusion-antigen (MEFA) is described recently for developing a broadly protective vaccine for ETEC.^209,273-277 Assisted with structural biology and computational biology, MEFA vaccinology uses a platform to combine structural vaccinology and epitope vaccinology and applies a backbone protein to present neutralizing epitopes of multiple heterogeneous strains to produce a single multivalent immunogen (Figure 2). Ideally, this backbone protein is a key virulence factor protein which is nontoxic and strongly immunogenic (better also exhibits adjuvanticity), possesses multiple well-separated epitopes and a stable secondary structure, and is easily expressed and purified. With substitutions of the backbone epitopes with foreign neutralizing epitopes, guided by protein modeling and molecular dynamic simulation, a MEFA protein mimics foreign epitope native antigenicity and becomes a multivalent antigen for a precision vaccine to provide cross-protection against heterogeneous pathotypes or pathogens. By applying this MEFA vaccinology platform, we have constructed multivalent proteins for broad protection against ETEC toxins,²⁰⁴ adhesins,^209,275 toxins and adhesins,^203,276,277 Shigella spp. and serotypes.²⁷⁸ Preclinical studies demonstrated ETEC adhesin MEFA CFA/I/II/IV and toxoid fusion 3xSTa_N12S-mnLT_R192G/L211A, when administered subcutaneously or intraperitoneally in mice and intramuscularly in pigs, induced broadly protective antibodies against adherence of the seven most important ETEC adhesins (CFA/I, CS1 to CS6)^209,275 or enterotoxicity of LT and STa.^204,207,208 Moreover, piglets born to the immunized sows acquired MEFA- and toxoid fusion-derived antibodies and were protected from diarrhea following STa+ or LT+ ETEC challenge.²⁰⁷ Additionally, a MEFA-based vaccine can further broaden its protective coverage by including another antigenically compatible MEFA immunogen to potentially protect against two or more than two groups of pathogens.

There are several advantages of the MEFA vaccinology platform in developing cross-protective vaccines. MEFA vaccinology is different from epitope vaccinology approach which stacks multiple epitopes into a linear peptide. Without assessing the secondary structure and native antigenicity of individual epitopes, epitope vaccines are often shown poor immunogenicity and a low efficacy. MEFAs mimic stable backbone secondary structure and epitope native antigenicity, thus induce strong immune responses to multiple virulence factors or strains represented by individual neutralizing epitopes, protecting against heterogeneous pathotypes. Like other protein-based vaccines, a MEFA-based vaccine consists of a purified protein produced by genetically engineered bacteria; therefore, the vaccine component is well defined. Since MEFA vaccine carries no somatic antigens and has no antibiotic-resistant genes involved, vaccines will not mask host immune responses or introduce the risk of antimicrobial resistance (AMR) to the environment. MEFA proteins can be produced in large quantities at a low cost compared to other strategies that require fermentation of multiple organisms to cover the different species and serotypes. Low-cost vaccines become essential to be licensed in low-income countries. Additionally, because it is protein-based, product purity, immunogenicity, thermostability, and safety characteristics can be easily documented and measured. Perhaps more important practically, MEFA vaccines can potentially be combined and co-administered with other vaccines to fit a single EPI (Expanded Program on Immunization), which becomes very much appreciated since the current EPI is getting overly crowded with the continuous introduction of new vaccines. Because MEFA vaccines would most likely be administered parenterally, they will thus be immunogenic even in infants in low-income to middle-income countries where oral vaccines are often less immunogenic and less protective.²⁷⁹ Furthermore, MEFA is a platform that can be adaptable for multivalent vaccine development for many pathogens including enteric Shigella, ETEC and other enteric pathogens, but potentially respiratory and other infectious pathogens as well.

Conclusion

Gastroenteritis remains a major threat to public health particularly in the low-income to middle-income countries. Parasitic, viral, and bacterial pathogens are the main causes of enteric infections. While WASH (clean drinking water, sanitation, and personal hygiene) may not be quickly implementable in resource-limited countries or regions, cost-effective vaccines and vaccination become a popular practice. However, though we have WHO prequalified vaccines for rotavirus, cholera, and Salmonella typhoid fever, we do not have vaccines licensed for parasites and many other enteric viral and bacterial pathogens including calicivirus, diarrheagenic E. coli, Shigella, Campylobacter, and no-typhoid Salmonella. Developing effective vaccines for enteric pathogens continues to be very challenging due to many reasons. A lack of understanding or characterization at immune correlates to protection from vaccines or vaccine candidates against enteric diseases hampers enteric vaccine development. While vaccine or vaccine-induced IgG antibody responses are more often measured, assessment of IgA particularly mucosal IgA responses should be emphasized. The greatest challenge in the current vaccine development is perhaps the heterogeneity among genotypes or serotypes of viral and bacterial pathogens. An effective vaccine would need to induce broad immunity to cross protect against heterogeneous genotypes or serotypes. Cocktail whole-cell vaccines which mix together a few heterogeneous genotypes or serotypes of an enteric viral or bacterial pathogen certainly broaden vaccine coverage. Whole-cell vaccines often encounter the challenge in inducing strong immunity in children aged less than two years to provide adequate protection to the most vulnerable group. LPS-based vaccines are typically serotype specific; thus, they often fall short in providing cross-protection against heterogeneous serotypes. While protein-based vaccines using conservative antigens for generic protection seem promising, antigens conserved inclusively in virulent strains or isolates may not be easily identified. The newly developed epitope- and structure-based MEFA (multiepitope fusion antigen) vaccinology platform may provide a means to develop broadly protective multivalent vaccines for heterogeneous pathogens or diseases.

Figure 1. A scheme to illustrate vaccinology evolution. From conventional vaccinology to reverse vaccinology and further to structural vaccinology, leading from whole-cell vaccines, to subunit vaccines and then structural epitope vaccines.

Figure 2. MEFA vaccinology applies a backbone immunogen to present neutralizing epitopes of virulence determinants from multiple heterogeneous strains, by substituting backbone surface exposed but less immunogenic backbone epitopes with foreign epitopes, and to mimic foreign epitope native antigenicity assisted with protein modeling and molecule dynamics simulation.

Disclosure of interest

The authors report no conflict of interest.

Acknowledgments

This manuscript is supported by NIH R01AI121067-01A1.

Supplementary

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was supported by the National Institutes of Health [R01AI121067-01A1].