OUP user menu

Shigella flexneri infection: pathogenesis and vaccine development

Amy V. Jennison, Naresh K. Verma
DOI: http://dx.doi.org/10.1016/j.femsre.2003.07.002 43-58 First published online: 1 February 2004


Shigella flexneri is a gram-negative bacterium which causes the most communicable of bacterial dysenteries, shigellosis. Shigellosis causes 1.1 million deaths and over 164 million cases each year, with the majority of cases occurring in the children of developing nations. The pathogenesis of S. flexneri is based on the bacteria's ability to invade and replicate within the colonic epithelium, which results in severe inflammation and epithelial destruction. The molecular mechanisms used by S. flexneri to cross the epithelial barrier, evade the host's immune response and enter epithelial cells have been studied extensively in both in vitro and in vivo models. Consequently, numerous virulence factors essential to bacterial invasion, intercellular spread and the induction of inflammation have been identified in S. flexneri. The inflammation produced by the host has been implicated in both the destruction of the colonic epithelium and in controlling and containing the Shigella infection. The host's humoral response to S. flexneri also appears to be important in protecting the host, whilst the role of the cellular immune response remains unclear. The host's immune response to shigellosis is serotype-specific and protective against reinfection by the same serotype, making vaccination a possibility. Since the 1940s vaccines for S. flexneri have been developed with little success, however, the growing understanding of S. flexneri's pathogenesis and the host's immune response is assisting in the generation of more refined vaccine strategies. Current research encompasses a variety of vaccine types, which despite disparity in their efficacy and safety in humans represent promising progress in S. flexneri vaccine development.

  • Shigella flexneri
  • Vaccine development
  • Pathogenesis

1 Introduction

Members of the genus Shigella are gram-negative facultative anaerobes that belong to the family, Enterobacteriaceae. They share common characteristics with members of the genus, Escherichia and the genetic relatedness clearly suggests that they are a subtype of E. coli [1, 2]. The genus is divided into four species, Shigella flexneri, Shigella boydii, Shigella sonnei and Shigella dysenteriae. These species are further divided into serotypes based on biochemical differences and variations in their O-antigen. Based on this classification scheme, Shigella flexneri is divided into 13 serotypes.

Shigella species invade the colonic and rectal epithelium of primates and humans, causing the acute mucosal inflammation characteristic of shigellosis. Infection is usually confined to the superficial layer of the colonic mucosa, where severe tissue damage leads to abscesses and ulceration. Destruction of the epithelial layer leads to the clinical symptoms of watery diarrhoea, severe abdominal pain and cramping, eventuating in the bloody mucoid stool characteristic of bacillary dysentery. In the absence of effective treatments, shigellosis patients may develop secondary complications such as septicaemia, pneumonia and haemolytic uremic syndrome [3].

Shigellosis occurs in an estimated 164.7 million people per year, of which 1.1 million cases result in death. 163.2 million annual cases occur in developing countries and 69% of all patients are children under the age of five [4].

S. flexneri is endemic in most developing countries and causes more mortality than any other Shigella species [5]. The predominant serotypes of S. flexneri in developing countries are serotypes 1b, 2a, 3a, 4a and 6, whilst in industrialised countries most isolates are 2a [4]. The high incidence of Shigella in developing countries is generally attributed to the lack of clean water, poor sanitation, malnutrition and cost of antibiotic treatment. Transmission is commonly via the faecal-oral route, which is augmented by poor hygiene and close personal contact.

Antibiotics can be used to treat shigellosis, reducing the period of bacterial excretion from the patient. However, S. flexneri is increasingly developing antibiotic resistance [6]. This escalation in resistance to the commonly used, cheaper antibiotics adds increased strain to the limited health services of developing countries. Consequently, the World Health Organisation has prioritised the development of a safe and effective vaccine against S. flexneri [4].

Numerous virulence genes have been identified in S. flexneri, with the majority of these genes being located on a 220 kb plasmid known as the virulence plasmid. At least three pathogenicity islands have also been located on the S. flexneri chromosome, encoding important virulence factors such as the lipopolysaccharide and genes for temperature-dependant regulation of the expression of virulence genes on the plasmid [710]. Two S. flexneri 2a genome sequences have recently been released, consisting of a chromosome of approximately 4 600 000 and 221 618 bp virulence plasmid [1, 2]. These sequences have confirmed that S. flexneri contains a number of bacteriophage-related genes. Some of the best-characterised S. flexneri phage genes are the serotype-conversion genes responsible for the serotype-specific modifications to the basic O-antigen structure [11].

The roles of many S. flexneri virulence genes have been studied in a variety of cell culture experiments such as invasion assays and plaque assays or through the use of in vivo animal models such as the Sereny test in the guinea pig, the mouse pulmonary model and rabbit ligated–intestinal–loop model [1215]. Continued research into S. flexneri virulence and pathogenesis will yield further understanding into the molecular basis of S. flexneri-mediated invasion and destruction of the intestinal mucosa, as well as the role of the host's subsequent innate, cellular and humoral immune responses. A comprehensive understanding into S. flexneri's disease-causing mechanisms will assist in vaccine development.

2 Pathogenesis of S. flexneri

S. flexneri is highly infectious, requiring as little as 100 cells to cause disease in adult volunteers [16]. This low infective dose is in part attributed to S. flexneri's ability to survive the low acidity of the host's stomach, via an up-regulation in acid resistance genes [17].

Once Shigella reach the colon, they begin to invade the mucosa, penetrating, replicating within and spreading between the mucosal epithelial cells. This behaviour and the subsequent inflammatory response of the host destroy the colonic epithelial layer generating the clinical symptoms of shigellosis (Fig. 1) [18].

Figure 1

Pathogenesis of S. flexneri. 1. Lumenal bacteria invade the colonic epithelial layer by three known mechanisms. S. flexneri can manipulate the tight junction proteins expressed by epithelial cells, allowing paracellular movement of bacteria into the sub-mucosa. 2. PMN cells recruited by IL-8 and IL-1β produced in response to S. flexneri invasion create gaps between epithelial cells, through which S. flexneri can transmigrate into the sub-mucosa. 3. Endocytic M cells transcytose bacteria, releasing them into an intraepithelial pocket filled with B and T lymphocytes and macrophages. 4. Macrophages phagocytose the bacteria. S. flexneri escapes the phagosome and induces the macrophage to undergo apoptosis. The apoptotic macrophage releases IL-1β. 5. Sub-mucosal S. flexneri contact the basolateral membrane of epithelial cells, activating secretion of proteins through their type-III secretion system. Proteins chaperoned in the cytosol of S. flexneri are secreted into the epithelial cell's cytoplasm through a pore formed by IpaB and IpaC. IpaC polymerises actin, IpgD dissociates the plasma membrane from the actin cytoskeleton, VirA destabilises microtubules and IpaA forms a complex with vinculin, depolymerising actin. This creates cell surface extensions which form around the bacterium, driving the epithelial cell to take up S. flexneri into a vacuole. 6. IpaB and IpaC lyse the vacuole, releasing S. flexneri into the epithelial cell's cytoplasm. The S. flexneri protein, IcsA is displayed on only one pole of the bacterium, creating a polymerised actin tail behind the bacterium. This propels S. flexneri through the cytoplasm until it contacts the plasma membrane, the force of the contact creates a protrusion into the neighbouring epithelial cell. Both membranes are lysed by IpaB and IpaC, releasing S. flexneri into the neighbouring epithelial cell. 7. Intracellular S. flexneri induces the epithelial cell to release IL-8. IL-8 and the IL-1β released from apopotic macrophages are chemotactic to PMN cells (represented by dotted arrows), attracting and inducing them to migrate through the epithelial layer to the lumen. This epithelial disruption amplifies S. flexneri invasion of the epithelial layer.

2.1 Crossing the colonic epithelial layer

Experiments using polarised cell lines have demonstrated that the majority of S. flexneri epithelial cell invasion occurs through the basolateral pole of colonic epithelial cells [19]. The epithelial layer acts as a barrier to pathogens found in the gut lumen. S. flexneri is able to penetrate the epithelial lining through the follicular associated epithelium (FAE). This is the epithelial layer found above the mucosa-associated lymph nodes, which contains highly endocytic M cells (Membranous epithelial cells).

M cells sample and transport lumenal antigen across the epithelial barrier, releasing it into an intraepithelial pocket formed by the basolateral membrane of the M cell. This pocket is filled with lymphocytes and macrophages waiting to take up any delivered lumenal antigen and initiate a mucosal immune response [20]. The transcytotic properties of M cells are exploited as a route for invasion of the impermeable epithelial lining by a number of pathogens, including S. flexneri (Fig. 1) [15].

Shigella appear to enter M cells by the same membrane “ruffling” seen in epithelial cell invasion [21]. Once internalised in an endocytic vacuole by the M cell, shigellae are moved rapidly through the cell and released into the intraepithelial pocket. Once the FAE is crossed, Shigella can access the basolateral membrane of the epithelial cells.

In the later stages of a Shigella infection, shigellae exploit the host's inflammatory response in order to amplify bacterial penetration of the colonic epithelium.

Macrophages infected by S. flexneri are induced to undergo apoptosis, releasing large amounts of IL-1, which is important in inducing inflammation and recruiting PMN cells to the site of infection. Additionally, the invasion of epithelial cells by Shigella activates the transcription and secretion of IL-8. IL-8 is chemotactic for PMN cells and plays a significant role in recruiting PMN cells to the infected subepithelial area, where they transmigrate through the epithelial lining to reach lumenal bacteria (Fig. 1) [22].

The influx of PMN cells across the epithelial layer in response to Shigella disrupts the integrity of the epithelium allowing lumenal bacteria to cross into the sub-mucosa in an M-cell independant mechanism [23]. This PMN recruitment has been demonstrated to be crucial for the generation of the inflammation and tissue destruction typical of shigellosis in a number of studies. Experiments in the rabbit ligated–intestinal–loop model of Shigella infection where either IL-1 or IL-8 is inhibited almost abolished inflammation, tissue destruction and notably decreased the amount of bacterial invasion [24, 25]. Additionally, the blocking of CD18, an adhesion molecule used by PMNs during migration, in the same animal model also diminished tissue damage and bacterial invasion [26].

Ironically, PMN-mediated interruption of the barrier function of the epithelial layer promotes the local spread of Shigella, whilst the same PMN cells appear to be responsible for restricting the infection to the submucosa and preventing systemic dissemination [25].

Recent research has revealed that the S. flexneri is capable of manipulating the tight-junction associated proteins of human intestinal epithelial cells, allowing bacterial paracellular movement through a model intestinal barrier. These results suggest that shigellae are also capable of penetrating the colonic epithelium via an M cell or PMN independant mechanism (Fig. 1) [27].

2.2 Macrophage apoptosis

Once released into the intraepithelial pocket of the M cell, bacteria are engulfed by resident macrophages, possibly through a bacterial driven macropinocytic event similar to Shigella entry of epithelial cells [28]. S. flexneri is able to evade the killing mechanisms of the macrophage by IpaB-mediated lysis of the phagocytic vacuole (Fig. 1). The membrane lysing properties of the virulence plasmid IpaB invasin allows the bacteria to gain free access to the cytoplasm [29]. Once in the macrophage cytosol, secreted IpaB binds and activates caspase-1, a member of the pro-apoptotic cysteine proteases [30]. Caspase-1 dependant apoptosis is not an immunologically silent cell death, as activated caspase-1 cleaves and activates the pro-inflammatory cytokines IL-1β and IL-18 [31]. Macrophage apoptosis occurs within four hours of in vivo Shigella infection, releasing bacteria into the sub-mucosa [32].

2.3 Adhesion to the basolateral membrane of colonic epithelial cells

It remains unclear exactly why S. flexneri preferentially invades epithelial cells thorough their basolateral membrane. The apical membrane of the colonic epithelial cells is covered with glycolipids which form a mucin layer. This layer may act as a physical barrier preventing S. flexneri access to the apical membrane, interfering with the type-III secretion systems delivery of the invasion plasmid antigens (Ipa), required for Shigella entry into epithelial cells [33].

Additionally the basolateral membrane of epithelial cells may display cellular components utilised by Shigella as cell adhesion receptors. The role of receptor-mediated epithelial cell adhesion in Shigella infection is incompletely understood. However, it is known that the Shigella Mxi/Spa secreton system requires contact with a host cell to trigger the secretion of the Ipa invasins [34]. A number of basolateral receptors capable of binding Shigella components have been identified. The Ipa proteins are capable of interaction with α5β1 integrin, a basolateral receptor which binds the extracellular matrix located beneath the epithelium [35]. IpaB is also able to bind the membrane receptor, CD44, which is the major cell surface receptor for hyaluronic acid and is found on the basolateral membrane of epithelial cells [36, 37]. Both α5β1 integrin and CD44 can act as cytoskeleton linkers, suggesting that upon the binding of Shigella they may contribute to the cytoskeletal reorganisation seen during epithelial invasion [38].

More recently, bacterial adherence to epithelial cells in a polarised model epithelium was shown to be dependant on the length and presence of the O-antigen [39]. These results suggest that the S. flexneri LPS may play a role in cell-bacteria interactions during epithelial cell invasion. A number of cell receptors capable of binding LPS have been characterised, including CD14 and the Toll family of receptors (TLRs), which are found on the basolateral membrane of epithelial cells [40, 41].

2.4 Uptake by the epithelial cell

S. flexneri invades epithelial cells through a macropinocytic process, where S. flexneri-induced rearrangements of the host cell cytoskeleton engulf the bacterium into a vacuole (Fig. 1).

The virulence plasmid of S. flexneri encodes two loci crucial to this invasive phenotype, the ipa locus and the mxi-spa locus. The ipa operon encodes the “invasion plasmid antigens”, IpaA, IpaB, IpaC and IpaD, which are the effectors of bacterial entry into the host cell. The mxi-spa operon encodes the components of a type-III secretion system, which is a flagella-like structure used to deliver proteins, such as the Ipa proteins, from the bacterial cytoplasm to the cytoplasmic membrane or even cytosol of the host cell [42]. The mxi-spa operon and IpaB, IpaC and IpaD are essential for in vitro epithelial cell invasion [43, 44].

The detailed mechanisms by which the Ipa proteins generate Shigella invasion are not completely understood. The Ipa proteins are synthesised and stored within the bacteria, where they are associated with chaperone proteins until secretion is activated by contact with a host cell [45, 46]. A complex formed by IpaB and IpaD may play a role in the regulation of this secretion [34]. Once secretion is activated by contact with an epithelial cell, the N-terminus of IpaC binds IpaB [47]. Both proteins are hydrophobic allowing this complex to insert into the membrane of the host cell to form a pore [48]. It is presumed that the other effector molecules, delivered by the type III secreton, are able to access the host cytoplasm through this pore.

The C-terminal domain of IpaC activates host cell Rho GTPases, triggering actin polymerisation and filopodial extensions in the vicinity of the bacteria [49]. IpaA is secreted into the cytosol of the host epithelial cell where it binds the cytoskeleton-associated protein vinculin. The IpaA–vinculin complex depolymerises actin filaments, organising an entry foci around the bacterium [50, 51]. IpgD is injected into the epithelial cell by the S. flexneri type III secretion system, where it acts as a phosphoinositide phosphatase, uncoupling the plasma membrane from the actin cytoskeleton, allowing membrane extensions to form [52].

VirA has recently been identified as an additional effector molecule of S. flexneri epithelial cell invasion. An interaction between VirA and tubulin within the host cytosol destabilises microtubules around the bacterial site of entry. It is proposed that this destabilisation could stimulate Rac1, a Rho family GTPase, creating lamellipodial extensions in the host cell [53].

The cytoskeletal rearrangements induced by the Shigella effector proteins results in the bacterium being internalised by the epithelial cell within a macropinocytic vacuole.

2.5 Replication within the epithelial cell and intracellular and intercellular spread

The macropinocytic vacuole containing the Shigella bacterium is rapidly lysed by the IpaB invasin, which acts as membranolytic toxin in the phagosome membrane, releasing Shigella into the host cell cytoplasm [29]. The lysis of the phagosome may also involve IpaC, which is able to disrupt phospholipid membranes upon insertion of its hydrophobic regions [54, 55].

S. flexneri can replicate inside the cytoplasm of epithelial cells in vitro with a doubling time of 40 minutes. Epithelial cells are observed undergoing necrotic-like death during shigellosis (Fig. 1) [56]. Although it was initially proposed that Shigella multiplication within the cytosol was the cause of epithelial cell lysis, it seems more likely that the cells are being destroyed by the host's inflammatory response [26]. In fact, Shigella would gain little advantage from killing the epithelial cell as whilst the bacteria are contained within the epithelial cell they are protected from immune cells and are in a favourable environment for replication [56].

mxiE, a gene located within the mxi/spa locus has recently been identified as a transcriptional regulator for a number of putative virulence factors required for virulence in the Sereny test. mxiE is only activated when the bacterium is within the epithelial cell cytosol suggesting that its role is to regulate virulence genes used in the post-invasion steps of infection [57].

Shigella is able to exploit the host cells actin assembly machinery to move through the host cell cytoplasm and into adjacent epithelial cells. This intra and intercellular spread is a crucial step in the virulence of Shigella and is driven by the outer membrane protein, IcsA (VirG) [5860]. IcsA is expressed in a unipolar fashion on the bacterial surface, with the greatest concentration localised to the old pole of the bacterium [61]. Newly synthesised IcsA appears to be directly targeted to the old pole by two internal regions, where it is autotransported to the outer membrane [62, 63]. The maintenance of IcsA's unipolar localisation is essential for intracellular movement and appears to be dependant on the structure of the LPS. Mutant S. flexneri strains missing, expressing partial O-antigen's or lacking a modal distribution of O-antigen chain length display non-polar surface localisation of IcsA and are unable to spread from cell to cell [6466]. It is possible that the LPS maintains IcsA polarity by forming interlocking microdomains with its O-antigen side chains on the surface of the bacterium, which would prevent IcsA from diffusing away from the old pole [67]. A recent study has revealed the need for S. flexneri to display short length O-antigen chains in order to prevent the blocking of IcsA's active sites by very long O-antigen chains. This finding suggests that S. flexneri has evolved to express two O-antigen chain lengths with each contributing to the virulence of the strain; short chains which allow IcsA to function and long chains which confer resistance to serum [68].

The Shigella protein DegP also appears to be required for efficient intracellular spread and polarised expression of IcsA. The exact role of DegP is unknown but it may be important in the delivery of IcsA to the bacterial surface [69].

IcsA at the bacterial pole interacts with the host protein neural Wiskott–Aldrich syndrome protein (N-WASP) and possibly with vinculin [7072]. IcsA specifically binds N-WASP and not other members of the WASP family, which stimulates actin-related protein (Arp) 2/3 complex-mediated actin polymerisation [73, 74]. This ligand specificity of IcsA–N-WASP may determine which host cells allow Shigella to use actin-based motility [74].

Actin polymerisation at the pole of the bacterium creates propulsive force, which drives the bacterium through the cytoplasm of the cell until it contacts the host cell membrane, forming a protrusion into the neighbouring epithelial cell [75]. The protrusion is actively endocytosed by the neighbouring cell in a myosin light chain kinase dependant mechanism, which also requires cadherin expression [76, 77]. The bacteria are then surrounded by two cellular membranes, which are lysed by secreted IpaB and IpaC [78]. Another protein, VacJ, has also been shown to be essential for freeing Shigella into the cytoplasm of the next cell [79]. Thus, Shigella is able to replicate and spread within the intestinal epithelial layer whilst avoiding exposure to the extracellular environment and its circulating immune cells.

3 The host's immune response to S. flexneri

3.1 Innate immunity

The severe inflammation generated by shigellosis can persist in the gut for over a month, with a general upregulation of a variety of cytokines (IL-1, TNF-α, IL-6, IFN-γ, TNF-β, IL-4, IL-10, TGF-β and IL-8) [80]. Although some of the clinical symptoms of shigellosis may actually be a direct consequence of the cytokines, they also assist in controlling and containing the infection.

Resident macrophages and infiltrating monocytes are unable to efficiently kill S. flexneri in their phagosomes and instead succumb to apoptosis [81, 82]. The IL-18 released by apoptotic macrophages can target NK cells and T lymphocytes, inducing production of IFN-γ[83]. IFN-γ deficient mice are five times more susceptible to a Shigella infection, as IFN-γ activates macrophages and fibroblast cells, which promote bacterial clearance and possibly inhibit bacterial replication within epithelial cells [84].

The most important consequence of the host's innate immune response appears to be the cytokine induced migration of PMN cells. The transcription factor NF-κ B is activated in Shigella-infected epithelial cells in an LPS-dependant mechanism, leading to the production and secretion of IL-8 by the infected cells [39]. IL-8 is a potent chemoattractant for PMN cells, as is the IL-1 released from apoptotic macrophages.

Shigella is unable to escape the phagocytic vacuole of PMN cells and are killed inside the phagosome [85]. Recent research has implicated the human neutrophil elastase (NE) as a key host defence protein of the neutrophil, capable of degrading Shigella virulence proteins within 10 min of Shigella infecting the neutrophil [86]. PMN cells ultimately play a crucial role in controlling the Shigella infection, confining extracellular bacteria to the mucosa, preventing deeper tissue invasion and systemic spread [25, 87].

Another host defence mechanism directed against Shigella has recently been discovered. The glycoprotein, lactoferrin, present in mucosal secretions, breast milk and phagocytic cells can impair the ability of S. flexneri to invade HeLa cells, exposing IpaB–IpaC complexes to protease degradation by disrupting the bacterial surface [88]. Additionally, a study in transgenic mice expressing a human intestinal defensin has demonstrated an important role for intestinally-secreted antibiotic peptides in controlling a Salmonella typhimurium enteric infection [89]. It is highly likely that intestinal defensins would display similar antibiotic properties against enteric S. flexneri.

3.2 Cellular immunity

Very little data is available on the host's cellular immune response to S. flexneri, especially in comparison to other intracellular bacteria. Studies have shown increased T cell activation in shigellosis patients and T cell clones have been isolated which proliferate in response to S. flexneri antigen [9093]. The cytokines induced by Shigella antigens in vaccine studies are suggestive of Th1 and Th2 lymphocyte responses [94, 95]. Additionally, the increased susceptibility of AIDS patients, deficient in CD4+ T cells, to shigellosis could suggest that cell-mediated immunity can play a protective role in shigellosis [96].

However, the contribution of T lymphocytes to the host's protective immunity to Shigella was studied in the mouse pulmonary model where mice deficient in T cells were vaccinated with attenuated S. flexneri. These mice were suitably protected from challenge with wild type bacteria despite their deficiency in T lymphocytes, suggesting that even if T cell responses develop to Shigella they are not essential for protection [97].

3.3 Humoral immunity

Information about the host's humoral response to Shigella infection has been collected from numerous serological studies of infected humans and experiments performed in animal models. The data suggests that the humoral immune response is a major component of protective immunity to shigellosis with both systemic and mucosal responses activated against the LPS and some virulence plasmid encoded proteins, including the Ipa proteins. The serotype-specific structure of the LPS is assumed to be the major target of the host's immune response as natural and experimental infections with Shigella confer serotype-specific immunity, where previous infection or vaccination provides little to no protection against heterologous serotypes [98100]. However, antibodies directed against epitopes shared between certain O-antigen structures do appear to show some cross-reactivity [101]. The protective significance of these cross-reactive antibodies is incompletely understood and is discussed in more detail in Section 4.6. However, the overall importance of an antibody response to Shigella infection has been confirmed in a study which showed that a reduced and delayed humoral immune response in comparison to adult patients is the likely cause of the increased susceptibility of children to shigellosis [102].

It appears that both the systemic and mucosal arms of the humoral response are activated as serum IgG, IgM and secretory IgA have all been implicated in the generation of serotype-specific immunity against S. flexneri.

Secretory IgA (sIgA) is made up of 2 IgA units and two polypeptides, the J chain and the secretory component (SC). sIgA transcytoses into the lumenal cavity of the intestine where the secretory component binds the mucosal coating of the epithelial cells, forming an antibody shield over the cells [103]. sIgA can also coat the outer membrane of lumenal bacteria, impeding invasion by preventing their attachment to the mucosal surfaces, mediate antibody-dependant cell-mediated cytotoxicity and interfere with bacterial utilisation of growth factors [104].

IgA, especially anti-LPS IgA have been detected in humans suffering natural shigellosis in a number of studies and is thought to play an important role in immunity to re-infection [105109]. Anti-LPS secretory IgA antibodies in the breast milk of mothers exposed to shigellosis appear to be responsible for the decreased severity of shigellosis in Shigella-infected infants [110]. Additionally, the implantation of a serotype-specific sIgA hybridoma on the back of mice protected them against intranasal challenge with a lethal dose of S. flexneri organisms [111]. This experiment suggests that a mucosal antibody directed against a single LPS epitope of Shigella could be sufficient for protective immunity against re-infection by the homologous serotype.

Despite shigellosis generally being a localised mucosal infection, serum antibodies IgG and IgM are detected in natural human infections directed against the LPS and virulence plasmid antigens [105107, 112, 113]. IgG and possibly IgM directed against the LPS have been shown to play a protective role in immunity to Shigella in mice studies. IgA deficient vaccinated mice are fully protected against pulmonary Shigella challenge, suggesting that IgG or IgM are able to provide immunity [114]. Immunised mice deficient in all T lymphocytes were protected from wild-type Shigella challenge by a predominantly anti-LPS IgM response [97]. It is still unclear what role serum antibodies directed against the LPS of S. flexneri are playing in the generation of serotype-specific immunity, although they may be stimulating complement killing of the bacteria or mediating antibody-dependant cellular cytotoxicity in the mucosal area [115117]. However, it must be stressed that the protective role of serum antibodies in controlling Shigella infection has been predominately characterised in mice and warrants further investigation as human vaccination data suggests that the parenteral stimulation of serum Ig does not correlate with protection [118121].

4 S. flexneri vaccine development

The cost of treating shigellosis with antibiotics, especially in the developing world, is unrealistic. The serotype-specific immunity generated by S. flexneri provides protection against reinfection by the homologous serotype, making vaccination a viable option for controlling shigellosis.

A suitable vaccine for shigellosis must fulfil certain requirements: the mucosal immune system must be activated and this immunity should be long-lasting, the vaccine must be cheap to manufacture, induce minimal side effects and be simple to administer, as children in developing countries will be the main recipients.

Since the 1940s a number of candidate vaccines for S. flexneri have been developed but as yet none have been successful enough for field release. Early attempts to develop S. flexneri vaccines consisted of inactivated bacteria delivered parenterally, which failed to induce a protective immune response, despite inducing a high titre of serum anti-LPS antibody [118121]. The lack of protection was most likely due to the failure of the parenteral vaccine in inducing a mucosal immune response.

Consequently, many recent vaccine strategies have concentrated on developing live vaccine strains which can be administered orally and will activate the effectors of mucosal immunity.

4.1 Subunit vaccines

Subunit Shigella vaccines may avoid the safety issues associated with live vaccines.

LPS can be complexed to proteosomes and delivered intranasally to humans. Clinical trials have revealed that a S. flexneri 2a LPS–proteosome vaccine is capable of generating a serotype-specific immune response in humans [122]. S. flexneri LPS has also been attached to proteins and delivered parenterally to volunteers as potential vaccines. These vaccines were safe in humans and induced strong serum antibody responses [123126].

Other subunit vaccines are yet to be evaluated in humans. Mice and guinea pigs were protected from S. flexneri challenge by mucosal immunisation with a purified complex of IpaB, IpaC, IpaD and LPS [127]. Ribosomal preparations from Shigella delivered parenterally can generate protective immunity in guinea pigs and monkeys [128]. The immune response is directed against O-antigen polysaccharides (l-hapten) purified with the ribosomal preparation. However, the O-antigen content in ribosomal preparations varies, making a consistent vaccine difficult to manufacture [129].

4.2 Killed oral vaccines

Early challenge experiments in monkeys revealed that orally administered acetone-killed and dried Shigella was unable to protect monkeys from infection [130]. More recently however, an oral heat-killed S. flexneri vaccine evaluated in a rabbit model was shown to be 100% protective [131]. Thus, further studies are required to determine the protective capabilities of killed oral vaccines for S. flexneri in humans.

4.3 Non-invasive live vaccines

Mutations in either the S. flexneri chromosome or the virulence plasmid have been used to generate non-invasive live vaccine strains. Most of these strains were safe in humans and were able to induce some degree of protective immunity in volunteers (Table 1). Probably the most successful of these vaccines is the invasion plasmid mutant, S. flexneri 2a Istrati T32 which is 100% safe in humans and provides up to an 85% protective efficacy. However, it must be administered in large (1 × 1011 CFU) multiple doses every six months which is expensive and difficult to implement in developing countries [132].

View this table:
Table 1

Live non-invasive oral S. flexneri vaccines which have been assessed in monkeys or humans

S. flexneri 2a 2457OSpontaneous avirulent mutant, virF inactivated by insertionReverts to virulence in humansMonkeys are protected. Caused dysentery in 34% of human volunteersReactogenic in humans[130, 146, 147]
S. flexneri streptomycin dependant strainsSpontaneous mutants incapable of growing in the absence of streptomycinReversion to streptomycin independence in volunteers. Diarrhoea and vomiting in 15–35% of volunteers (5 × 1010 CFU)Up to 90% protection in field trials with multiple doses. US trials found marginal protectionUnstable phenotype and inconsistent protective efficacy[98, 148, 149]
S. flexneri 2a Istrati T32Spontaneous deletion of ipaBCDA, invA and icsA (virG) from the virulence plasmidSafe in humans, mild adverse effects in very few volunteers at 2 × 1011 CFUAround 80% protection in humans when administered in 5 dosesProtection lasts 6 months. Reduces attack rate of heterologous Shigella serotypes[132, 150]

4.4 Invasive live vaccines

Invasive oral Shigella vaccine strain strategies are increasingly being explored as invasive strains deliver antigen to the mucosal immune system, provoking a strong immune response. As the genetic understanding of S. flexneri virulence has improved so have the strategies to construct safe invasive vaccines. Invasive vaccine strains are generally attenuated by mutations in either virulence genes necessary for pathogenesis after cell entry or in metabolic genes which prevent the bacteria from replicating and spreading in the host after invasion.

Mutations in either icsA and/or in a variety of metabolic genes have produced attenuated invasive vaccine strains which are safe and capable of up to 100% protection with multiple doses in monkeys (Table 2). A number of auxotrophic vaccine strains, some also carrying mutations in virulence genes, have been assessed for their safety and ability to induce a serotype-specific immune response in human volunteers in phase 1 clinical trials [94, 133135]. These strains vary in the levels of their attenuation and their immunogenicity (Table 2).

View this table:
Table 2

Live invasive oral S. flexneri vaccines which have been assessed in monkeys or humans

S. flexneri 5a SC5700Attenuated by insertions in icsA and iucMucoid diarrhoea in some monkeys with 5 × 1010 CFUAbout 70% protection in monkeys after 3 dosesSlightly reactogenic in monkeys[151]
S. flexneri 5a SC560Deletion in icsAMucoid diarrhoea and mild clinical symptoms in some monkeys with 5 × 1010 CFU100% protective in monkeys after 3 dosesSlightly reactogenic in monkeys[60]
S. flexneri 5a SC433Deletion in envZ and ompRMucoid diarrhoea and mild clinical symptoms in some monkeys with 5 × 1010 CFU100% protective in monkeys after 3 dosesSlightly reactogenic in monkeys[60]
S. flexneri 5a SC445Deletion in envZ, icsA and ompR100% safe in monkeys at 5 × 1010 CFUOne out of five monkeys challenged became sickLevel of attenuation may be too severe for consistent protection[60]
S. flexneri 2a vc77Auxotrophic for purine (Pur) and rifampicin resistantNo symptoms in adults or children up to 3 × 109 CFUDysentery in 2 out of 4 adults challengedPoor protective capabilities[152154]
S. flexneri Y TSF-21Thymine requiring (Thy) and temperature sensitive (Ts)No symptoms in monkeys at 1 × 1011 CFU100% protection in monkeys after 2 dosesTemperature sensitivity mutation has a high rate of reversion[155, 156]
S. flexneri Y SFL114Insertional inactivation of aroDNo symptoms in monkeys at 2–3 × 1010 CFU100% protection in monkeys after four dosesSome reversion of phenotype observed in laboratory conditions[157]
S. flexneri Y SFL124Deletion of aroDVery mild symptoms in some human volunteers at 2 × 109 CFU and children at up to 1 × 109 CFU100% protection in monkeys after three doses of 1 × 1011 CFUStrain is advantageous as it can be converted to new serotypes[135, 158160]
S. flexneri 2aSFL1070Deletion of aroDMild symptoms in volunteers at 1 × 107– 1 × 108 CFU. Increased clinical symptoms at 1 × 109 CFUAbout 85% protection in monkeys after 4 doses 1 × 1011 CFUFurther attenuation may be necessary[133, 161]
S. flexneri 2a CVD1203Deletions in icsA and aroANo symptoms in humans when administered in a single dose of 1 × 106 CFUChallenge only performed in guinea pigs, provided 83% protectionStrain may require further attenuation in order to reduce reactogenicity in humans[134, 162, 163]
S. flexneri 2aCVD1207Deletions in icsA, sen, set and guaABNo symptoms in humans when administered in a single dose of up to 1 × 108 CFUChallenge only performed in guinea pigs, provided 85% protectionImmunogenicity in humans may not be sufficient with just one dose[94, 144]
S. flexneri 2a SC602Deletions in icsA gene and the iuc locus.Mild diarrhoea and fever in humans at 1 × 104 CFU3 out of 7 volunteers challenged experienced mild diarrhoeaProvides protection or reduces symptoms to infection with S. flexneri 2a. Further attenuation may be required[136, 164]

The S. flexneri 2a vaccine strain, SC602 has proceeded to phase 2 clinical trials in humans. This strain carries deletions in icsA as well as the aerobactin iuc locus, which is involved in iron transport. SC602 is safe in humans at low doses (1 × 104 CFU) and capable of providing protection to immunised humans challenged with wild type 2a S. flexneri. However, the vaccine is only weakly attenuated causing symptoms such as diarrhoea and fever when administered in doses higher that 1 × 104 CFU [136]. Thus, despite promising results with S. flexneri invasive vaccine candidates, further work is required to achieve a balance between immunogenicity and safety in humans.

4.5 Hybrid vaccines

E. coli vaccine candidates have also been used to develop hybrid vaccines expressing Shigella antigens. Early attempts using Shigella–E. coli hybrid vaccines developed invasive vaccines which caused symptoms in human volunteers or which were not protective [137, 138]. Strains based on E. coli K12 carrying the group- and type-specific antigen of S. flexneri 2a and the virulence plasmid from S. flexneri 5 were unable to induce significant protection in immunised volunteers [139, 140].

Additionally, S. flexneri candidate vaccine strains are being engineered to express the O-antigen's of other Shigella species. The S. flexneri 2a vaccine strain T32 carrying a plasmid containing the gene cluster coding for S. sonnei O-antigen, was capable of providing 100% protection to mice against challenge with both virulent S. flexneri and S. sonnei [141]. The S. dysenteriae O-antigen biosynthesis genes were integrated into the SFL124 (serotype Y) vaccine strain, generating strains able to induce antibodies specific to both homologous and heterologous O-antigen structures in mice [142]. Similar approaches are also being used to generate vaccines protective against multiple S. flexneri serotypes and will be discussed below.

4.6 Multiple-serotype protection strategies

Because immunity to S. flexneri is serotype-specific, vaccination against one serotype will only provide protection to infection by the homologous serotype. The serotypes of S. flexneri differ in their distribution with up to four different serotypes prevalent in an endemic area. Thus, the ideal S. flexneri vaccine would provide protection to all prevalent serotypes of a particular geographical region.

All S. flexneri serotypes, with the exception of serotype 6, share a common O-antigen backbone. The addition of glucosyl and/or O-acetyl groups to the sugars of the backbone generates the type (I, II, IV, V and X) and group (3, 4, 6 and 7, 8) antigens that define the serotypes. Consequently, some serotypes share type and/or group antigens on their LPS [11]. Because the immune response is primarily directed against the LPS, some antibodies generated against the group or type antigen of one serotype should be cross-reactive to other serotypes. For example, antibody in human sera raised against S. flexneri 2a has been shown to cross-react with LPS from heterologous serotypes 1a, 2b, 5a and Y, which share type or group antigens with the serotype 2a O-antigen structure [101]. As cross-reactivity of the human sera to all of the different serotypes was not observed, it appears that the common group 1 antigen, which is shared by all S. flexneri serotypes [143] was not able to induce any sufficiently cross-reactive antibodies. This suggests that the group 1 antigen is poorly immunogenic and may not have a role in inducing protective antibodies against heterologous serotypes. It also remains unclear whether the cross-reactive antibodies directed against the other group and type antigens of the LPS, as mentioned above, can provide any protection against infection by heterologous serotypes. Therefore, further research is required to adequately establish the role of the cross-reactive O-antigen epitopes in human immunity against shigellosis.

However, animal studies have shown that mixing a number of S. flexneri vaccines of different serotypes into a vaccine cocktail can invoke an immune response with cross-reactive potential. A vaccine cocktail containing serotype 2a and 3a S. flexneri strains was administered to guinea pigs in the Sereny test, conferring significant protection to challenge by serotypes 1b, 2b, 5b and Y [144]. Thus, by combining a selection of S. flexneri serotypes into a vaccine cocktail, it may be possible to cross-protect against most S. flexneri serotypes.

Alternatively, single S. flexneri vaccine strains can be engineered to express the O-antigen of more than one serotype. Such strains should be capable of generating a protective immune response in the host directed against each of the serotype specific O-antigen structures. This lab has previously reported the serotype-conversion of the serotype Y S. flexneri candidate vaccine strain, SFL124 to serotype X by the insertion of the bacteriophage SfX serotype-conversion gene cluster [145]. This approach has been utilised to insert the serotype conversion gene cluster of bacteriophage SfV and the glucosyl transferase gene of bacteriophage SfII in tandem into the SFL124 chromosome. The resulting strain displayed the 3,4 group antigen and both the II and V type antigens as detected by monovalent antiserum and simultaneously induced a serotype-specific immune response to both serotypes 2a and 5a in the mouse pulmonary model (unpublished data, this lab). This strategy could be easily applied to any newly developed serotype Y vaccine candidates, ultimately generating a variety of S. flexneri polyvalent vaccine strains which could be combined into vaccine cocktails designed for specific geographical areas.

5 Conclusions

Through in vitro and in vivo studies we are beginning to develop a detailed picture of how S. flexneri invades the intestinal mucosa and causes disease. Further research into the host's immune response will ultimately reveal all inflammatory mediators involved in shigellosis and clarify whether cellular immunity plays an important role in the control of Shigella and in protection against reinfection. Virulence factors are crucial to S. flexneri for the development and maintenance of disease. Undoubtedly, many more remain to be identified and characterised. Discovery of new virulence factors and an understanding of gene regulation and conditional gene expression will be greatly assisted by the release of the S. flexneri 2a genome sequence. A number of the Shigella vaccine strategies mentioned above have shown promise in animal studies and initial human clinical trials, significantly advancing the status of Shigella vaccine development. However, to completely protect against natural Shigella infections, the development of vaccine cocktails and polyvalent vaccines must be addressed. Future vaccine research should encompass trials of mixed vaccines designed to confer protection against multiple serotypes.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
  51. [51].
  52. [52].
  53. [53].
  54. [54].
  55. [55].
  56. [56].
  57. [57].
  58. [58].
  59. [59].
  60. [60].
  61. [61].
  62. [62].
  63. [63].
  64. [64].
  65. [65].
  66. [66].
  67. [67].
  68. [68].
  69. [69].
  70. [70].
  71. [71].
  72. [72].
  73. [73].
  74. [74].
  75. [75].
  76. [76].
  77. [77].
  78. [78].
  79. [79].
  80. [80].
  81. [81].
  82. [82].
  83. [83].
  84. [84].
  85. [85].
  86. [86].
  87. [87].
  88. [88].
  89. [89].
  90. [90].
  91. [91].
  92. [92].
  93. [93].
  94. [94].
  95. [95].
  96. [96].
  97. [97].
  98. [98].
  99. [99].
  100. [100].
  101. [101].
  102. [102].
  103. [103].
  104. [104].
  105. [105].
  106. [106].
  107. [107].
  108. [108].
  109. [109].
  110. [110].
  111. [111].
  112. [112].
  113. [113].
  114. [114].
  115. [115].
  116. [116].
  117. [117].
  118. [118].
  119. [119].
  120. [120].
  121. [121].
  122. [122].
  123. [123].
  124. [124].
  125. [125].
  126. [126].
  127. [127].
  128. [128].
  129. [129].
  130. [130].
  131. [131].
  132. [132].
  133. [133].
  134. [134].
  135. [135].
  136. [136].
  137. [137].
  138. [138].
  139. [139].
  140. [140].
  141. [141].
  142. [142].
  143. [143].
  144. [144].
  145. [145].
  146. [146].
  147. [147].
  148. [148].
  149. [149].
  150. [150].
  151. [151].
  152. [152].
  153. [153].
  154. [154].
  155. [155].
  156. [156].
  157. [157].
  158. [158].
  159. [159].
  160. [160].
  161. [161].
  162. [162].
  163. [163].
  164. [164].
View Abstract