Diarrhoeal Diseases (Updated February 2009)
Enterotoxigenic Escherichia coli (ETEC)
Introduction
Numerous types of diarrhoeagenic E coli strains have been identified worldwide, including enteropathogenic (EPEC), enterohaemorragic (EHEC), enteroinvasive (EIEC), enterotoxigenic (ETEC), Shiga toxin-secreting (STEC), diarrhoea-associated haemolytic (DHEC), entero-aggregative (EAAggEC), and cytolethal distending toxin-secreting (CDTEC) E coli strains. The prevalence of these strains and the burden of disease they cause are however unequal.
Some enterohaemorragic E coli strains have been lysogenized with phages encoding the toxin from Shigella dysenteriae [95] thus becoming Shiga toxin-secreting E coli (STEC). Such is the case of the strains belonging to serotype O157:H7, which cause a watery diarrhoea rapidly progressing to frank haemorragic colitis, itself leading in about 20% of paediatric patients to serious sequelae, including the haemolytic uremic syndrome (HUS) characterized by thrombocytopenia, hemolytic anemia and renal failure. The natural reservoir of the O157:H7 pathogen is cattle, which harbours the bacterium in its intestines: outbreaks in humans are linked to ingestion of meat or other foods contaminated by bovine faeces. Given the relatively low incidence of serious STEC infections and HUS worldwide, a vaccine against STEC is not thought of as a priority, but the development of Shiga toxoid as a vaccine against Shigella dysenteriae is receiving some attention (see Shigella vaccines below [46] . As most STEC infections are caused by serogroup O157, immunity to the O157 PS has been tested using a Pseudomonas aeruginosa exoprotein A-conjugated O157 PS vaccine, which was used in a Phase II clinical trial in children [96] , but protective efficacy remined doubtful.
In contrast with STEC, enterotoxigenic E coli (ETEC) strains remain a major cause of infantile diarrhoea in developing countries and of travellers’ diarrhoeas in visitors to these countries. The need to develop a vaccine against ETEC is therefore urgent.
Disease Burden
ETEC are an under-recognized but extremely important cause of diarrhoea in the developing world where there is non adequate clean water and poor sanitation [97] . They are the most commonly isolated bacterial enteropathogen in children below 5 years of age in developing countries, and account for several hundred million cases of diarrhoea and several ten of thousand deaths each year [98] . Disease caused by ETEC follows ingestion of contaminated food or water and is characterized by profuse watery diarrhoea lasting for several days that often leads to dehydration and malnutrition in young children. ETEC was thought to account for approximately 200 million diarrhoea episodes and 380 000 deaths annually [99] [100] . A more conservative estimate, about 170,000 deaths every year, was more recently suggested [101] . Repeated ETEC infections and persistent diarrhoeas in children in developing countries are not rare, as observed in infants living in the Nile delta area, who experienced between 4.6 and 8.8 diarrhoeal episodes per year, with ETEC accounting for 66% of these episodes [102] . The peak incidence of ETEC diarrhoea in developing country settings occurs in the first two years of life, with a declining incidence with age thereafter [103] . Surveillance of hospitalized cases of ETEC diarrhoea has shown that a large proportion of cases also occur in individuals over 10 years of age. ETEC are prevalent in surface water sources in developing countries such as Bengladesh, which explains the endemicity of the disease in these countries [104] . In young children, the tendency of ETEC to cause dehydrating diarrhoea is lower (approximately 5% of episodes) than that of rotavirus (approximately 36% of episodes). However, because the incidence of ETEC diarrhoea in children is considerably higher than that of rotavirus diarrhoea, the absolute number of dehydrating diarrhoea episodes due to ETEC is around 70% of that due to rotavirus.
ETEC also are the most common cause of travellers' diarrhoea that affects individuals from industrialized countries travelling to developing regions of the world [105] [106] . Each year, there are an estimated 10 million cases of LT-related ETEC travellers' diarrhoea worldwide [100]. Thus, up to 60% of US visitors to Mexico develop the disease, with travellers genetically predisposed to produce high levels of interleukin-10 (IL-10) being more likely to experience symptomatic ETEC travellers' diarrhoea [107].
Bacteriology
ETEC attach to specific receptors on the surface of enterocytes in the intestinal lumen by virtue of their hair-like fimbriae, which define strain-specific antigenicity. More than 20 types of fimbriae antigens, called E. coli surface antigens or colonization factor antigens (CFAs) have been described. Antibodies targeted to fimbriae are protective but show high serotype-specificity.
Once attached to the intestinal epithelium, ETEC elaborate both the heat-labile (LT) and the heat-stable (ST) toxins, which induce the watery diarrhoea. The heat-labile enterotoxin LT is an 86 kDa protein which, like the cholera toxin, with which it shares 82% amino acid homology, is made of five B subunits that bind to GM1 ganglioside receptors in the intestinal epithelium, and a single enzymatically active A subunit, the ADP-ribosylating activity of which leads to activation of cellular adenyl cyclase, efflux of Cl- ions and watery diarrhoea [108] . The heat-stable toxin (ST) is an 18-amino acid-long, highly folded peptide which also causes disruption of chloride channels in the cell and secretory diarrhoea. LT is expressed in about 66% of ETEC strains, either alone or in combination with ST, and thus is significantly responsible for the worldwide disease burden of ETEC [109] . Efforts are being made to identify major clonal groups among ETEC strains, largely based on O-antigen typing, CFA expression pattern and toxin profile [110].
Because of antigenic mimicry between CTB and LTB, short-term protection against ETEC disease has been documented in individuals immunized with CTB [73] [74] (see Cholera vaccines above).
Vaccine
Natural history studies of ETEC infections in children in developing countries suggest that these infections are immunizing, as reflected by declining rates of ETEC diarrhoea with age, lower ratios of symptomatic to asymptomatic ETEC infections with increasing age, and the protective relationships between initial ETEC infections and subsequent infections that have similar toxin and/or colonization factor phenotypes [111] [112] . These data suggest that immunization against ETEC early in life may be an effective preventive strategy. Pathogenesis of ETEC appears to be linked to CFAs and to the production of LT and/or ST. To provide broad-spectrum protection, an ETEC vaccine should, therefore, contain the most prevalent fimbrial antigens (CFA1 and CS1-CS6) and/or a LT toxoid [113] , although this view was recently challenged [114].
Killed oral vaccines
The oral killed WC/rBS cholera vaccine (DukoralTM) was found to prevent 23% of all diarrhoea episodes and 52% of episodes due to ETEC in Finnish tourists visiting Morocco. This protection was reported, however, not to last more than a few months [115] . In a retrospective study in Spain, vaccination with DukoralTM reduced by 43% the risk of travellers'diarrhoea [74] Vaccination against cholera and ETEC should be recommended to at-risk travellers [116].
The most successful ETEC vaccine approach so far, which was developed by investigators at the University of Göteborg (Sweden), is based on a killed, oral, whole-cell ETEC vaccine containing recombinant CTB together with five strains of formalin-killed ETEC cells that collectively express the colonization factors of greatest epidemiological importance in developing countries (CFA/I and CS1-CS6). Phase II studies of a 2-dose regimen of this vaccine have been conducted in Bangladesh, Egypt, Israel, Nicaragua, the USA and Europe and have found the vaccine to be safe and immunogenic, as manifested by induction of mucosal antibody responses to CTB and to the CFA components of the vaccine. A pilot efficacy trial of this vaccine in European tourists travelling to developing countries found the vaccine to confer about 80% protection against ST-ETEC diarrhoea (the only toxin phenotype detected in this study), although the small number of outcome events precluded statistically significant estimates of efficacy [117] . A 75% protection was reported against severe diarrhoea in US volunteers travelling to Mexico or Guatemala, although the vaccine did not reduce the overall rate of ETEC diarrhoeas in the vaccinees [118] . Trials of the oral killed vaccine efficacy are ongoing in travellers from the USA and Europe, as well as in Israeli military recruits, and Egyptian infants and young children, but vaccine efficacy in young children was found to be disappointingly low [114].
The oral killed vaccine approach is being pursued by several investigators [119] . E coli bacteria K12 over-expressing CFA/I have recently been engineered that could be useful as an oral killed CF-ETEC vaccine [120].
Live attenuated oral vaccines
Two live attenuated ETEC strains, PTL002 and PTL 003, which express the colonization factor CFA/II, were tested in a Phase I trial [121] . Based on its superior immunogenicity, PTL003 will be developed further as a component of a live, oral attenuated ETEC vaccine. Similarly, two nontoxinogenic ETEC strains that express CFA/I have been attenuated by mutagenesis of the aroC and ompR genes or the aroC, ompC, ompF and toxin genes, respectively. The latter strain, ACAM 2010, was found to be well tolerated and 73% immunogenic when fed to human volunteers. The strain will be developed as a live attenuated oral vaccine by Acambis, UK [122].
Another strategy, which is developed at the Center for Vaccine Development (CVD), University of Maryland (USA), is to use live attenuated Shigella vectors for expression of ETEC fimbrial and LT antigens. Such constructs might thereby protect against both Shigella and ETEC. Four lots of attenuated Shigella vaccine strain CVD 1204 (?guaBA) expressing ETEC fimbriae antigens CFA/1, CS2, CS3 and CS4, respectively, were found to be immunogenic in guinea pigs by the intranasal route. An additional strain was constructed that expressed a detoxified version of LT (LThK63). A mixed inoculum containing the five recombinant Shigella strains elicited immune responses to the five ETEC antigens plus the Shigella vector [123] . In a more recent approach, a combination of three ?guaBA attenuated vectors, S flexneri (CVD 1208), S sonnei (CVD1233) and S dysenteriae 1 (CVD 1252), was used instead of a single vector [124] , eliciting specific immune responses against each of the vectors as well as against each of the five ETEC antigens.
A similar approach is being followed by Microscience, UK, using their spi-VEC oral live attenuated typhoid vaccine as a vector for the delivery of ETEC antigens. The resulting oral vaccine, based on S typhi Ty2 derivative TSB7 harboring attenuation deletions in the ssaV and aroC genes and a chromosomally integrated copy of the E coli LT-B subunit, was shown to induce 67% and 97% immune responses to LT-B and S typhi lipopolysaccharide (LPS), respectively [125] . It might thus elicit protection against both ETEC diarrhoea and typhoid fever. Another live recombinant oral vaccine is developed that would cover traveller's diarrhoeal diseases due to Campylobacter, Shigella and ETEC.
The Walter Reed Army Institute of Research has similarly engineered an attenuated S flexneri 2a (SC608) vector derived from the well-characterized SC602 live attenuated vaccine strain which has undergone several clinical trials in human volunteers, to express the ETEC fimbriae subunit CfaB (CFA/I structural subunit) in combination with the LT-B subunit. Guinea pigs immunized by the intranasal route were subsequently protected from challenge with wild type S flexneri in a keratoconjunctivitis Sereny test [126] and serum antibodies from the vaccinated animals showed antitoxin (anti-ETEC) and agglutination (anti-S flexneri) activities.
Finally, the Peru15pCTB live attenuated oral cholera vaccine candidate developed by AVANT will address both V cholerae and ETEC diarrhoeas (see Cholera vaccines above)
Other ETEC vaccine approaches
A mixture of fimbrial antigen CS6 and LT was administered to human volunteers using a new delivery technology, the transcutaneous immunization patch [127] . An immune response to both antigens was elicited in about 50% of the volunteers. The presence of LT as an adjuvant was required for induction of responses to the CS6 antigen. Transcutaneous immunization with patches of 50 µg LT alone on days 0 and 21 was shown by IOMAI Corp, USA, to be safe and immunogenic. In a vaccination/challenge study, transcutaneous LT vaccination did not prevent but mitigated illness following high-dose oral challenge with a virulent LT+ ST+ ETEC strain [128] . In a Phase II randomized, double-blind, placebo-controlled field trial, healthy adults travelling to Mexico or Guatemala were 85% protected against severe diarrhoea and 75% protected against moderate-to-severe diarrhoea by vaccination with two LT patches given 2-3 weeks apart [129] . Meanwhile, disruption of the stratum corneum of the skin before vaccine patch application was found to result in significant increase of the immune neutralizing antitoxin response to LT: the process will likely be used systematically for transcutaneous immunization in the future [112].
Several other approaches are being pursued to develop specific ETEC vaccines including the use of purified colonization factors, of LT-only or LT-ST toxoids, or of edible transgenic plants that express the cholera toxin B subunit (CTB) or the LT B subunit [130].