Immune responses to cholera in children

Abstract

Cholera is a severe acute dehydrating diarrheal disease caused by Vibrio cholerae O1 or O139 infection, and is associated with significant mortality and morbidity globally. Although young children bear a high burden of the disease, currently available oral vaccines give a lower efficacy and shorter duration of protection in this group than in adults. According to the studies of natural infection, young children achieve comparable systemic anti-V. cholerae antigen-specific antibody, gut-homing antibody-secreting cell and memory B-cell responses as adults. Studies on innate and cell-mediated immune responses are lacking in children, and may offer important insights into differences in vaccine efficacy. The impact of host factors such as malnutrition, genetics and coinfection with other pathogens also remains to be fully defined.

Cholera is an acute dehydrating diarrheal disease caused by the Gram-negative bacterium Vibrio cholerae. Despite suboptimal surveillance and reporting, it is estimated that globally 2–3 million people have cholera each year, and more than 100,000 die from this infection annually [1,2]. Cholera is a disease of poverty, made worse in endemic urban areas by the overcrowding of informal housing settlements or slums [3]. The disease is endemic in over 50 countries; however, it also appears in epidemic form, both in endemic areas and in regions made susceptible by civil unrest and natural disasters, such as following the earthquake in Haiti in 2010 [4]. Although epidemic cholera is seen in all age groups in endemic areas, young children have a higher burden of disease [5–7].

The majority of cases of epidemic cholera are caused by infection with toxigenic strains of the O1 and O139 serogroups of V. cholerae. The O139 serogroup is differentiated from the O1 by the O antigen of the lipopolysaccharide (LPS). Although V. cholerae O139 has epidemic potential, for unknown reasons it has largely disappeared as a cause of cholera. The O1 serogroup is further divided into two biotypes: El Tor and classical. Although the classical biotype was the causative agent of earlier pandemics for which we have microbiologic data, the current seventh pandemic, which began in 1961, is now caused by the El Tor biotype.

In humans, ingestion of water or food contaminated with V. cholerae results in colonization of the small intestine, an attachment facilitated by the toxin coregulated pilus, a fimbrial protein involved in the formation of microcolonies [8]. Subsequent elaboration of cholera toxin (CT), and the action of the toxin on intestinal epithelial cells, causes the secretion of large amounts of chloride, sodium and water via activation of adenylate cyclase and increases in intracellular cyclic AMP [9]. Important factors affecting immune responses in V. cholerae infections are listed in BOX 1. In both adults and children, severity of infection can vary from asymptomatic or mild-to-moderate infection to death within hours of onset of massive diarrhea. Vomiting combined with purging of large volumes of stools resembling rice water can result in rapid dehydration and death due to hypovolemic shock [10]. In children, cholera can be complicated by severe hypoglycemia [11] and concomitant pneumonia [12]. The mainstay of treatment of patients with cholera is rapid fluid and electrolyte replacement, optimally in the form of oral rehydration solution containing salts, sugar and water, if the patient is able, including initiation of this at home upon onset of symptoms [10]. If available, the use of rice-based oral rehydration solution for the management of diarrhea owing to cholera having been shown to decrease volume of stools and is indicated for those 6 months and older [13]. Intravenous administration of isotonic fluids is often necessary for those who are severely dehydrated or unable to tolerate oral therapy. Antibiotic use is beneficial in severe cholera, shortening the duration of illness and thus lessening the amount of fluid replacement needed in both adults and children [14–16]. With prompt and appropriate treatment, the mortality rate of cholera is <1%, although higher mortality rates are common in complex emergencies and during the initial phase of an epidemic [17].

Box 1. Important microbial factors expressed by Vibrio cholerae.

• Lipopolysaccharide – T-cell-independent antigen, a serogrouping and serotyping determinant of Vibrio cholerae strains

• O-specific polysaccharide – T-cell-independent antigen, the determinant of serogroup and serotype

• Toxin coregulated pilus – T-cell-dependent group of antigens, involved in intestinal colonization

• Cholera toxin – T-cell-dependent antigen, causes intestinal secretion of electrolytes and water. A potent enterotoxin and immunoadjuvant. B subunit referred to as CTB or CtxB.

CTB: Cholera toxin B (CtxB).

In areas of the world endemic for cholera, although all age groups are susceptible, the largest burden of symptomatic cholera is often among children [5]. In a study of household contacts of patients with V. cholerae O1 infection in an urban area of Bangladesh, those who were ≤5 years of age had a significantly higher risk of acquiring infection than older family members in a 21-day observational period [18]. Despite the susceptibility and high burden of disease among young children in endemic areas, currently available cholera vaccines have shown lower protective efficacy and a shorter duration of protection in children under 5 years of age compared with older individuals [19]. Currently available cholera vaccines include two killed, whole-cell oral cholera vaccines (OCVs), both of which are WHO prequalified. One contains killed whole cells of a number of strains of classical and El Tor V. cholerae O1, supplemented with 1 mg/dose of recombinant CT B subunit (WC-rBS; Dukoral, Crucell, Sweden); the other is bivalent, containing killed strains of classical and El Tor V. cholerae O1 as well as O139, without CtxB supplementation (WC; Shanchol, Shantha Biotechnics, India). In children aged 2–5 years, the WC-rBS vaccine is administered as three doses orally at least 1 week apart. The WC-rBS vaccine is not recommended for children under 2 years of age. The WC vaccine is approved for children ≥1 year old, and given as two doses orally at least 14 days apart.

A recent review of studies of current OCVs and their predecessors demonstrated that vaccine efficacy in the first 2 years after vaccination was 66% in those >5 years old, but only 38% for children <5 years old [20]. A recently published report of the 3-year follow-up to a large field study of Shanchol in Kolkata, India, showed that while the overall efficacy was 66%, efficacy in those <5 years old was only 43%, with no significant protection in the third year of follow-up for this age group [21]. This is despite: the fact that after symptomatic infection, young children and older persons appear to achieve a similar reduction of at least 60–70% in disease lasting at least 3 years after an initial episode compared with controls [22]; that in volunteer studies, previous cholera is associated with 90–100% protection against subsequent challenge for the 3 years of the evaluation period [23]; that population-based field studies suggest that previous infection with classical cholera provides protection against subsequent disease lasting 6–10 years; and that previous infection with El Tor cholera provides protection against subsequent disease lasting 3–6 years [24]. It is important to note that protection against cholera following wild-type symptomatic disease appears to be similar among older individuals and young children. The reasons behind the differences between protection from natural infection and vaccination, and differences in vaccine efficacy between age groups, have yet to be determined, and such information could significantly contribute to an improved cholera vaccine or immunization strategy optimally effective in young children.

Immune responses to natural infection & vaccination in children

Innate immune responses

The innate immune response is involved in the initial defense against pathogens, including the triggering of an adaptive immune response. Young children, especially infants, have limited exposures to antigens and therefore limited ability to rely on rapidly deployed anamnestic immune responses; thus, compared to older children and adults, innate immune responses may be especially critical in young children for controlling infection during the initial exposure. Although V. cholerae is a noninvasive pathogen, and historically most investigations have focused on the adaptive immune responses, recent studies have shown that the innate immune response is upregulated in cholera [25–27].

Studies of patients with severe acute cholera have demonstrated that blood levels of mediators of the innate immune response, including leukotriene B4, lactoferrin, myeloperoxidase and nitric oxide, are elevated at the initial phase of infection in both children and adults compared with age-group-matched healthy controls [25,28]. Increases of such mediators in stool are also present.

Whole-genome microarray screening of duodenal biopsies from adults acutely infected with V. cholerae O1 shows that the majority of upregulated genes encode for proteins that are part of the innate response [27]. Histopathological studies of duodenal biopsies demonstrate that neutrophils infiltrate the mucosa, followed by an increase in degranulated mast cells and eosinophils during convalescence [26,29]. Other mediators of the innate response, including cytokines such as TNF-α and IL-1b, as well as bactericidal proteins, including lactoferrin, myeloperoxidase and defensins are also elevated [26,30]. Furthermore, recent studies suggest that the long palate, lung and nasal epithelium clone 1 protein expressed in Paneth cells also plays a role in modulating the innate response to V. cholerae LPS [31].

Studies of duodenal biopsies in children with cholera have not yet been performed. However, rectal biopsies from children with acute cholera show similar findings to those of adults, including an increase in neutrophils at onset of disease followed by an increase in mast cells at early convalescence [26]. These findings are associated with elevated expression of myeloperoxidase, lactoferrin and nitric oxide by immunohistochemistry, all of which persist up to 30 days longer than that seen in adults [25].

Many questions remain unanswered regarding the innate response in young children with V. cholerae infection. For instance, as innate immune responses, their signaling pathways and associated cytokine responses mature throughout early childhood [32,33], differences between young children and adults remain to be characterized in the context of V. cholerae infection and vaccination. Furthermore, the role of micronutrient supplementation on innate responses to V. cholerae, as has been demonstrated with zinc in enterotoxigenic Escherichia coli (ETEC) infection [34], needs to be evaluated.

Systemic antibody & memory B-cell responses

Evaluations of humoral immunity to V. cholerae infection have largely involved assessing systemic antibody responses, and studies in children with cholera and receiving cholera vaccines have also mostly focused on the vibriocidal antibody and antitoxin antibodies (although mucosal immune responses have been assessed in smaller cohorts, as described below).

Vibriocidal antibody responses

The serum vibriocidal antibody response is perhaps the most-studied immunologic marker of cholera infection. The antibodies are bactericidal in the presence of complement, and most of their activity is of the IgM isotype directed against LPS [35–37]. Titers increase with both symptomatic and asymptomatic infection [38], but levels wane quickly and fall to baseline levels within 9–12 months [39,40]. In endemic settings, the percentage of a population with detectable vibriocidal titers increases with increasing age, and 40–80% of individuals have detectable vibriocidal antibodies by 15 years of age [41,42]. Although higher titers have been associated with protection against V. cholerae O1 infection and disease [18,38,42,43], there is no threshold at which protection is complete [44].

A study of patients infected with V. cholerae O1 showed that children ≤12 years had a lower acute phase (day 2) vibriocidal antibody titer, but achieved a higher vibriocidal response to infection than those >12 years of age [45]. Similarly, a study of children hospitalized for V. cholerae O1 Ogawa infection showed that while young children (2–5 years of age) had significantly lower acute-phase titers than adults (18–60 years of age), young children had higher fold-increases in vibriocidal antibodies compared with adults, resulting in comparable peak titers during convalescence [46]. As such, it appears that young children are able to mount vibriodical responses following wild-type cholera infection that are comparable to those induced in older children and adults.

The few studies with age-group-specific comparisons of vibriocidal responses after oral vaccination have shown much heterogeneity in results depending on the type of vaccine and criteria of response. Interlaboratory variability also limits comparison across studies, although fold increase in response to infection or vaccination can be used as the important outcome. Overall, studies of immunogenicity as measured by fold change or responder frequency of vibriocidal titer show that young children have comparable immune responses to vaccination as older children and adults. A study of WC-BS, a predecessor of the current killed, whole-cell OCV WC-rBS, in Bangladesh, showed that in general, children aged 2–5 years had similar responder frequencies and fold changes as older children and adults [47,48], a response that was comparable regardless of whether or not a recombinant B subunit was included in the vaccine. Notably, young children, but not adults, had incremental increases in responder frequency with subsequent vaccine doses. These findings were also seen in a study of Shanchol vaccine, in which young children (both 2–5 years and 12–23 months of age) had similar responder rates as adults [49]. The live attenuated El Tor Inaba V. cholerae O1 OCV (Peru-15) also elicited comparable vibriocidal response rates in young children aged 2–5 years and infants aged 9–24 months as adults, although infants achieved the lowest rates [50].

Akin to what is seen following natural infection, studies of both live and killed formulations of cholera vaccines have demonstrated lower ‘seroconversion’ rates [51–53], and lower fold-increases [54], in children with higher baseline or acute-phase titers, largely reflecting a higher baseline/acute-phase level in endemic zones where exposure to V. cholerae is often repetitive.

Antigen-specific antibody responses

Immunoglobulin A

Secretory IgA (sIgA) is thought to be an important marker of mucosal humoral immunity. On the intestinal surface, sIgA is the predominant immunoglobulin, existing as a dimeric form produced in the mucosa that can neutralize intraluminal pathogens [55]. Systemically, IgA levels increase progressively in childhood, from 1% of adult levels in the newborn to 20% at 1 year, 50% at 5 years and 75% by 16 years of age [56]. In a study in Bangladesh among household contacts, including children, observed for 21 days after identification of the index case of cholera, levels of serum IgA specific to all three V. cholerae antigens examined were associated with protection against subsequent V. cholerae O1 infection during follow-up [18].

In children infected with V. cholerae O1 Ogawa, levels of serum IgA against LPS and CTB rise, with peak levels (at day 7), similar to that achieved in older children and adults [46]. By day 30 of convalescence, levels of IgA are still above baseline levels, although adults maintain LPS IgA levels that are significantly higher than those of younger children. In a study that included both O1 and O139 infections, patients >12 years had higher levels of homologous LPS IgA than those ≤12 years at both acute and convalescent phases of infection, whereas younger patients had higher CTB IgA responses, including those found in feces [45]. This may be due to younger children having more recent exposures to ETEC [57], a common childhood pathogen whose heat-labile enterotoxin is immunologically crossreactive with CT [58].

Most studies of serum IgA responses after vaccination have focused on the antitoxin response. Approximately 80% of children in Bangladesh aged 2–5 years achieved a more than or equal to twofold increase in CT IgA after two doses of a WC-rBS vaccine given 2 weeks apart [59], and a similar responder frequency rate was found in children <2 years of age given a WC-rBS vaccine in a subsequent study [60]. Antitoxin responses are not believed to confer protection against cholera [61], and indeed, the most recently licensed OCV, Shanchol, is not supplemented with CtxB.

Immunity against cholera is serogroup specific, as previous infection with V. cholerae O1 does not provide protection against O139 and vice versa [62]. Serogrouping reflects antigenic differences within the O-specific polysaccharide of LPS; unfortunately, few studies have examined IgA responses to LPS following vaccination (and none have yet evaluated O-specific polysaccharide responses), although the studies available suggest that young children may have lower anti-LPS responses than adults. In Bangladeshi children given Peru-15, a live attenuated Inaba vaccine, 54% of 2–5-year-old children achieved a more than or equal to twofold increase in homologous LPS IgA, while only 34% of those <2 years of age achieved such increases [50], both of which were lower than the responses seen in adults (88%) receiving the same vaccination in an earlier study [63]. Studies of the killed WC vaccine, also among Bangladeshi subjects, showed that while young children (2–5 years and 12–23 months) achieved comparable rates of response (defined as more than or equal to twofold increase in titer), the peak geometric mean titer decreased with age (LPS IgA against O1 Inaba, geometric mean titer of 171 in adults, 37 in those aged 2–5 years and 13 in those aged 12–23 months) [49]. LPS is a T-cell-independent antigen and despite the similar responder rates, the lower magnitude of the absolute antibody response may be a reflection of poorer humoral responses to T-cell-independent antigens in very young children [64].

Immunoglobulin G

Significant systemic IgG responses to V. cholerae antigens are detected following both natural infection and vaccination. In adults, elevations in serum IgG to CTB are detectable for at least 270 days following both natural infection [39] and two doses of WC-rBS vaccination [65]. However, studies of household contacts of cholera patients have not found levels of plasma antigen-specific IgG on exposure to be predictive of protection against subsequent cholera [39].

Children with V. cholerae infection also mount significant increases in plasma IgG against CTB and LPS by day 7 of infection [45,46]. Notably, baseline levels of plasma CTB IgG are significantly lower in adults than children, likely reflecting more recent exposure to ETEC in children. During V. cholerae O1 Ogawa infection, all age groups achieve similar magnitudes of CTB- and LPS-specific plasma IgG at convalescence [46], while a study including patients infected with V. cholerae O1 or O139 showed that adults mounted higher LPS IgG levels at convalescence, while children mounted higher CTB IgG levels [45]. In adults, systemic antibody responses in multiple subclasses of IgG have been demonstrated against both CT and LPS [66]. Such subclass studies have not been performed in children with cholera.

Antitoxin IgG responses have been demonstrated in children receiving OCV. In studies using a predecessor to the WC-rBS vaccine, WC-BS, which did not include recombinant CtxB but instead included CtxB isolated from cholera holotoxin (and therefore may have included residual amounts of the immunoadjuvant cholera holotoxin), children aged 2–5 years achieved a 2.6-fold rise in CT IgG compared with those receiving placebo, while those >15 years had a 4.7-fold rise [48]. A more recent study has shown that approximately 55% of Bangladeshi children aged 2–5 years receiving two doses of WC-rBS achieved a more than or equal to twofold increase in serum IgG antibody to CTB [59].

Immunoglobulin M

Antigen-specific IgM is the first antibody isotype to rise in the serum after exposure to antigen, and plays an important role in subsequent affinity maturation and isotype switching, giving way to other antibody isotypes such as high-affinity IgG [67]. IgM exists in pentameric form, giving it the ability to crosslink antigens and making it a strong activator of the complement system. On the mucosal surface, IgM is excreted by intestinal epithelia, contributing to the luminal defenses [68]. In newborns, IgM-producing plasma cells predominate in the mucosa, and secretory IgM is found in breast milk. Infants with selective IgM deficiency have an increased incidence of viral, Gram-negative bacterial, and polysaccharide-containing bacterial infections [69]. A study of African children showed that those with acute watery diarrhea had a tenfold greater intestinal IgM output in whole-gut lavage than controls [70].

In adults with V. cholerae O1 infection, IgM responses against LPS are elevated by day 7 after onset of illness, and remain persistently elevated above baseline for at least 30 days, the last period examined [71]. This finding is not unexpected, as the vibriocidal antibody is mostly composed of IgM directed against LPS [35]. Levels of IgM to CTB do not appear to change with cholera infection. Similarly, adults vaccinated with Peru-15 also produced low levels of serum IgM to CTB [63].

No studies are available characterizing IgM responses in children with cholera. Such investigations are needed given the importance of IgM antibody in humoral immunity against T-cell-independent antigens such as LPS [72], and its role in long-term immunity involving memory B cells (MBCs) [71].

MBCs in cholera

MBCs are found in the circulation after natural infection and vaccination and are thought to play a critical role in mediating long-term protective immunity by facilitating rapid anamnestic antibody responses upon re-exposure to antigen [73]. In adults hospitalized with natural infection, both IgA and IgG MBCs against V. cholerae antigens are detectable by day 30 after infection [39]. In fact, IgG MBC responses to T-cell-dependent antigens CTB and TcpA (a major pilus colonization factor of V. cholerae) persist for up to 1 year, longer than any other known marker of cholera immunity [39]. Despite the lack of increase in systemic IgM responses against CTB, IgM MBCs against LPS and CTB have also been described up to 30 days following acute infection with V. cholerae O1 [71].

Younger children with V. cholerae O1 Ogawa infection mount comparable CTB- and LPS-specific MBC responses by day 30 after infection as older children and adults [46], and there is a trend for younger children to mount higher levels of CTB IgG MBC responses than older children, likely reflecting more recent exposure to the crossreactive heat-labile toxin antigen of ETEC. Evaluation of the MBC responses in children with longer follow-up is needed.

In adults receiving WC-rBS vaccine, MBC responses to CTB and LPS are significantly shorter in duration and lower in magnitude than those in adults recovering from cholera [65]. These differences may partially account for the difference in duration of protection between vaccination and natural infection. The evaluation of MBC in children receiving cholera vaccination remains to be reported, and comparisons between vaccinated children and adults may uncover the role that MBCs play in determining the duration of vaccine protection. As of yet, there has also been no evaluation of memory responses targeting O-specific polysaccharide in children and adults, despite the fact that immunity to V. cholerae is serogroup specific.

Mucosal-specific adaptive responses

As V. cholerae is a noninvasive enteric pathogen, antigen-specific immune responses at the mucosal surface are believed to play a major role in protective immunity.

Secretory IgA

Secretory IgA is the predominant immunoglobulin at the mucosal surface, and may play a significant role in protection against cholera. In adults, antitoxin sIgA responses are detected in intestinal fluid, breast milk and saliva after both cholera infection and vaccination [74]. Anamnestic responses also likely contribute to protection, as evidenced by a rapid rise in intestinal lavage antitoxin IgA by day 3 in adults receiving a second course of immunization 15 months after primary vaccination [75], and by day 8 in previously infected adults given a repeated administration of V. cholerae O1 [23], both of which are the earliest times examined. However, such responses are unlikely to be the main mediators of immunity to cholera, as adults infected with V. cholerae O1 produce elevations in duodenal IgA against CTB and LPS up to day 30, but levels decrease to baseline by day 180, a shorter period of persistence than seen in circulating IgA [76].

Gut-homing antibody-secreting cells

Following antigen presentation in the gut mucosa, intestinal lymphocytes transiently migrate in the peripheral circulation before rehoming to the gut. These lymphocytes, termed antibody-secreting cells (ASCs), are detectable in blood even before antigen-specific sIgA is detectable, peaking at 7–10 days after mucosal challenge [77]. ASCs to LPS and CTB have been demonstrated in adults infected with both serotypes of V. cholerae [78]. ASCs are also detected in duodenal tissue after infection, and a recent study reported that LPS-specific IgA ASCs are significantly increased up to day 180 after infection [76], despite the absence of corresponding increases in mucosal LPS IgA antibody.

Young Bangladeshi children infected with V. cholerae O1 Ogawa mount ASC responses to CTB and LPS that peak by day 7 after onset of illness and are comparable to the magnitude seen in older children and adults [46], although there is a trend for older age groups to have higher LPS-specific IgA ASCs, likely the reflection of a poorer mucosal response to the T-independent antigen in young children. In vaccinated adults, ASCs to LPS have been reported to be present in the circulation 7 days after vaccination [63]; however, such responses have not been evaluated in vaccinated children.

Cell-mediated immunity

Although V. cholerae is a noninvasive mucosal pathogen, and pathogen-specific effector mucosal defense against V. cholerae is thought to be largely B-cell mediated, helper T cells likely play an important role in the development of B-cell immunity directed against protein antigens, as seen in the involvement of Th17 cells [79] and the chemokine receptor CXCR5 [80] in the generation of mucosal immune responses against CT. Adults recovering from cholera have increased frequencies of gut-homing, CD4⁺ expressing, effector and central memory T-cells that peak at day 7 [81–83]. In these patients, ex vivo stimulation of cells with a V. cholerae O1 Ogawa membrane protein preparation results in priming of both Th1 (IFN-γ) and Th17 (IL-17) cytokine responses, and an increase in the Th1 to Th2 CD4⁺ ratio [83]. By contrast, subjects given WC-rBS vaccine did not show these responses, although there was a trend toward an increase in IL-10 response. Furthermore, cytokines associated with a Th17 response are detectable in lamina propria samples during the acute stage of cholera using duodenal biopsy samples [83].

There are limited data on cellular immune responses in children with cholera. One study assessed aspects of cellular responses in infants aged 10–18 months given two doses of WC-rBS vaccine [84]. In these children, stimulation of post-vaccination lymphocytes ex vivo with a modified CTB resulted in increased CD4⁺ blast formation and IFN-γ production compared with responses elicited using prevaccination lymphocytes; however, in contrast with what occurs in adults, stimulation with a cholera membrane preparation did not produce any increases in children. Studies of T-cell responses in children with natural infection are needed.

Modifiers of immune responses in children

A number of host factors modify disease severity and immune responses following infection and vaccination, including genetic polymorphisms, nutrition, micronutrient status, blood group and coinfection [85].

Coinfection

In regions endemic for cholera, concomitant intestinal infection with parasites and bacteria is common, especially in children. Greater than 35% of children aged ≤10 years presenting with acute V. cholerae infection to a hospital in India had concomitant parasitic infection [49,86]. The impact of helminthic coinfection on blunting the mucosal immune response to cholera infection and vaccination has been demonstrated in both adults and children [87,88], and there is evidence that alterations in cell-mediated immunity are responsible for differences in mucosal responses. In patients presenting with severe cholera in Bangladesh, helminth coinfection was associated with decreased fecal and serum IgA responses to the T-dependent antigen CTB, but not to the T-independent antigen LPS [88]. Additionally, Ecuadorean children aged 13–17 years infected with Ascaris lumbricoides had a diminished Th1 cytokine response to vaccination with a live, attenuated OCV CVD 103-HgR compared with noninfected US controls, and this diminished response was partially reversed in a group treated with albendazole [89].

Coinfection with ETEC, an enteric bacterial pathogen commonly coendemic with V. cholerae, also alters the immune response to V. cholerae. In Bangladeshi adults and children hospitalized for diarrhea, those infected with both ETEC and V. cholerae (13% of patients) produced higher vibriocidal and higher antibody levels to CTB and LPS than those infected with V. cholerae alone [90].

Genetic factors

Recent studies have begun to address the role that genetics may play in the host immune response to cholera. It has long been known that patients with blood group O experience increased severity of cholera [91,92]. In a study of household contacts of index cholera patients in Bangladesh, familial segregation of susceptibility within households independent of blood group was observed, suggesting possible additional genetic contributions to cholera susceptibility and severity [18]. A family-based candidate gene association study in Dhaka, Bangladesh identified a variant in the promoter region of long palate, lung and nasal epithelium clone 1 to be associated with the disease [93]. Further exploration of genetic factors and their role in susceptibility to infection may uncover additional factors of importance affecting host immunity during cholera.

Malnutrition & micronutrient deficiency

Studies in animal models have demonstrated that protein deprivation results in impaired mucosal antitoxin responses [94] and severely diminished antibody responses to cholera vaccination [95]. In humans, malnutrition and intestinal infections are integrally related [96], and in Bangladesh, malnutrition is associated with up to half of deaths due to diarrhea in children under 5 years of age [97]. Low weight-for-age Z scores are associated with a 9.5 odds ratio for mortality from diarrhea [98], and in hospitalized patients with V. cholerae infection, individuals with protein–calorie malnutrition have increased stool losses and prolonged diarrhea [99].

Vitamin A, or retinol, and its metabolite, retinoic acid, are associated with host defense of infectious diseases, possibly through its effects on CD4⁺ T-cell function [100]. Vitamin A supplementation has been associated with reductions in mortality and morbidity, including incidence of diarrhea, when given to nonhospitalized young children in developing countries [101]. In patients hospitalized for cholera, retinol deficiency is more common in children <12 years than in older patients [45], and is associated with an increased risk of V. cholerae O1 infection and symptomatic disease [18]. Unfortunately, vitamin A supplementation of oral cholera vaccination in young children produced only small increases in vibriocidal antibody responses [102], and had no effect on antitoxin antibody responses [59,102].

Zinc deficiency in children is associated with an increased risk of diarrhea, and supplementation of children with zinc in developing countries is associated with reductions in both incidence and severity of diarrhea [103]. In children with V. cholerae infection, zinc supplementation decreases the duration of diarrhea and stool output [104], although zinc deficiency was not associated with increased susceptibility to infection in household contacts [18]. By contrast to vitamin A supplementation, zinc supplementation in children has a differential effect on the immunogenicity of OCV, in that it increases the vibriocidal antibody response [60,102] and IFN-γ production by CD4⁺ T cells [84], but the antitoxin responses were lowered [59]. In ETEC, a related acute toxigenic diarrheal infection, zinc supplementation resulted in an increase of complement C3 levels and in phagocytic functional activities [34]. The role that environmental or tropical enteropathy [105] may have on V. cholerae infection or vaccination and immune responses is unknown.

Expert commentary

Current correlates of protective immunity to V. cholerae infection, including the vibriocidal antibody and antigen-specific antibodies, are imperfect measures that fall to baseline within months of infection, while protection against recurrent, symptomatic cholera lasts for 3–7 years. MBC responses have recently been identified in cholera patients, and these responses are long lived. The lack of detailed immunologic studies of immune responses in children with cholera reflects in part difficulties performing clinical investigations in children, as well as difficulties in obtaining the required volumes of blood for immunologic studies, although these studies are needed given the inequity of vaccine efficacy between children and adults. The recent characterization of prominent innate mucosal immune responses in adults should be extended to children, despite the limitations of obtaining intestinal biopsies in this population. Given the advances in knowledge of cell-mediated immunity and the role of helper T cells in B-cell development, analysis of T-cell responses and their contribution to the induction of protective immunity following cholera are also needed. Analysis of polysaccharide responses are also warranted since immunity to V. cholerae is serogroup specific. Evaluating the effect of tropical enteropathy on both susceptibility and response to V. cholerae infection and vaccination is also required. As cholera continues to cause a large burden of mortality and morbidity in children globally, a better understanding of anti-V. cholerae immunity in young children will critically contribute to the development of an improved cholera vaccine and immunization strategy.

Five-year view

Recent advances in immunological methods have provided opportunities to look in detail at host responses to V. cholerae infection. In the next 5 years, we anticipate further advances in technology and methodology that will enable the evaluation of immune responses using much smaller volumes of biological specimens, thus further facilitating study of immune responses in infants and young children. Ongoing longitudinal studies of cholera patients and vaccinees will better inform our understanding of protective immunity following cholera. Progress on high-throughput techniques, including immunoproteomic and immunogenetic screening, will facilitate identification of novel antigens and factors contributing to host susceptibility. Currently available OCVs will set a benchmark for the development of future vaccines and immunization strategies, especially for children, who bear the largest burden of cholera globally.

Key issues.

• Cholera is a dehydrating diarrheal disease caused by Vibrio cholerae serogroups O1 and O139.

• In areas of the world endemic for cholera, children bear a large burden of infection.

• During cholera epidemics among immunologically naive populations, children and adults are equally affected by cholera.

• Current oral cholera vaccines have lower efficacy and shorter duration of protection in young children compared with adults.

• Innate immune responses occur during cholera in adults, although studies of innate immune responses to cholera have not been performed in children.

• Young children are able to mount comparable vibriocidal and toxin-specific antibody responses as adults to both infection and vaccination, but neither are sufficient predictors of protective immunity against cholera.

• Limited studies suggest that young children mount a lower magnitude of IgA responses to lipopolysaccharide than adults following vaccination.

• Anti-polysaccharide responses have not yet been evaluated during cholera, despite serogroup specificity of protection.

• Memory B cells mediate long-term protective immunity by facilitating anamnestic antibody responses, and such responses against V. cholerae antigens are comparable in young children and adults up to a month after infection.

• Current oral cholera vaccines have not been shown to induce memory responses comparable to those induced by infection.

• Investigations are lacking regarding long-term anti-V. cholerae memory responses in children following natural infection and vaccination.

• Th1 and Th17 responses may be involved in development of memory B-cell responses following cholera.

• Host factors such as helminth coinfection, blood group, enteropathy and micronutrient deficiencies affect immune responses in children with cholera.