Colonization, localization, and inflammation: The roles of H. pylori chemotaxis in vivo

Abstract

Helicobacter pylori is a gram-negative bacterium that infects half of the world’s population, causing gastritis, peptic ulcers, and gastric cancer. To establish chronic stomach infection, H. pylori utilizes chemotaxis, driven by a conserved signal transduction system. Chemotaxis allows H. pylori to sense an array of environmental and bacterial signals within the stomach, guiding its motility towards its preferred niche within the gastric mucosa and glands. Fine-tuned localization, regulated by the chemotaxis system, enables robust colonization during the acute stage of infection. During chronic infection, chemotaxis helps maintain bacterial populations and modulates the host immune response. Given its importance in host colonization and disease, chemotaxis is an attractive target for future treatments against H. pylori infections.

Introduction

Helicobacter pylori is a gram-negative bacterium that has evolved a keen ability to chronically colonize the stomach, infecting roughly half of the world’s population [1]. Colonization can lead to the development of chronic gastritis, gastric and duodenal ulcers, and gastric cancers [2,3]. Colonization is also linked to potential health benefits, including protection against allergic asthma, and inflammatory bowel and esophageal diseases, presumably though modulating effector T-cell populations or gastric acid production [4–7]. The beneficial and pathogenic aspects of H. pylori require colonization, therefore an important goal is to understand the molecular basis of the bacterial factors that promote chronic infection.

H. pylori colonizes the primate stomach, a harsh environment. The stomach lumen ranges between pH 1–5 [8], conditions at which H. pylori is viable for only ~30min [9]. Additionally, stomach contents are cleared regularly, and the gastric mucosa undergoes constant turnover [10]. Accordingly, H. pylori must rapidly initiate colonization and localize where the environment is more hospitable: within 15μm from the gastric epithelial cells [11], and deep within gastric glands [12].

H. pylori colonization is promoted by chemotaxis, the focus of this review. Chemotaxis is a process that enables H. pylori to sense environmental signals and regulate its motility to move away from harmful conditions and towards favorable ones [13]. Chemotaxis promotes colonization and plays a role in modulating host immune responses [14–16]. In addition to chemotaxis, H. pylori utilizes a suite of colonization factors including urease, cell shape, adhesins, the Cag pathogenicity island type four secretion system, and the toxin VacA. Readers are referred to several excellent recent reviews on these topics [7,17–19].

The H. pylori signal transduction system transforms ligand presence into a swimming response

Chemotaxis signal transduction systems allow bacteria to direct their motility [13,20,21]. Such systems are widespread, present in ~50% of bacterial species, highlighting the strong fitness advantage they confer [20]. The chemotaxis system of H. pylori contains core chemotaxis proteins found in all chemotaxis systems, and auxiliary ones found in only some (Figure 1). The core chemotaxis proteins are the chemoreceptors (TlpA, TlpB, TlpC, and TlpD), the CheW coupling protein, the CheA kinase, and the CheY response regulator [13,22]. The H. pylori auxiliary chemotaxis proteins include three CheV-type coupling proteins (CheV1, CheV2, and CheV3), the CheZ phosphatase, and the unique chemotaxis protein ChePep, which localizes CheZ to the poles [12,23–28]. Environmental signals are sensed either directly or indirectly by chemoreceptors, and are relayed to the histidine kinase CheA via the CheW or CheV1 coupling proteins [13,27]. Currently, the role of CheV2 and CheV3 are unknown. Chemicals sensed as repellents activate CheA’s auto-phosphorylation, and this phosphoryl group is subsequently passed to CheY via histidine-to-aspartate phosphorelay [29]. Phosphorylated CheY interacts with the flagellar motor, causing it to rotate clockwise and the bacteria to reverse or change direction [30,31]. Alternatively, attractants squelch CheA’s auto-phosphorylation; non-phosphorylated CheY does not interact with the motor, and the bacteria swim straight, without direction changes. Mutants in any one of these proteins are non-chemotactic (Che⁻) to varying degrees and with different swimming behavior. Accordingly, cheW, cheA, cheY, cheV1, and cheV2 mutants are all Che⁻, displaying straight swimming phenotypes, likely because they do not produce phosphorylated CheY [24,25,32,33]. cheZ, chepep, cheV3 mutants are also Che⁻, but display hyper-reversal phenotypes, apparently because they produce high amounts of phosphorylated CheY [12,24–26]. Readers are referred to several excellent reviews of this system for more molecular details [18,34,35].

Figure 1.

H. pylori chemotaxis system. Chemotactic signals are sensed by the chemoreceptors TlpA, TlpB, TlpC, and TlpD. Signals are relayed through the coupling protein CheW (W) or the auxiliary CheV-type coupling protein CheV1 (V1) to the histidine kinase CheA (A). Repellents promote CheA auto-phosphorylation, while attractants squelch CheA auto-phosphorylation. Phosphorylated CheA passes its phosphoryl group to the CheY (Y) response regulator via histidine to aspartate phosphorelay. Phosphorylated CheY interacts with the flagellar motor and is dephosphorylated by CheZ phosphatase (Z), in complex with ChePep (Pep). CheV2 and CheV3 are not depicted as their role is this system is unknown.

H. pylori senses multiple host and bacterially-generated conditions as repellents and attractants

H. pylori senses specific chemicals and conditions via three transmembrane chemoreceptors with periplasmic sensing domains—TlpA, TlpB, TlpC— and one cytoplasmic receptor, TlpD [16,36–38]. The signals of these chemoreceptors and their distribution likely play a driving role in the localization of H. pylori in vivo. Thus, efforts have been made to determine H. pylori’s sensing profile.

H. pylori experiences multiple repellent conditions. Several of these are host generated, including acidic pH, reactive oxygen species (ROS), and bile [37,39,40]. The acidic stomach lumen is toxic to H. pylori [8,9]. Accordingly, acid is a potent chemorepellent, with TlpA, TlpB, and TlpD playing roles in sensing [37,41,42], and TlpC playing a role in modulating the acid response [43]. Currently, there is only a mechanistic proposal for how TlpB senses acid, via amino acids that are variably protonated [41]. Another signal, ROS [39], is produced by host epithelial and immune cells [44–46]. TlpD senses ROS via an unknown mechanism [39]. Bile acids are toxic to H. pylori and are sensed via unknown chemoreceptor(s) [40,47]. Bile is released from the gallbladder, into the duodenum [48]. The repellent and toxic properties of bile may explain why H. pylori does not colonize this region [40,47,49].

H. pylori is also repelled by the self-generated, quorum-sensing molecule autoinducer-2 (AI-2), and responds to its own electron transport chain (ETC) [38,50]. AI-2 is sensed by TlpB as a chemorepellent, in a manner dependent on the periplasmic proteins AibA and AibB [50,51]. A repellent response to AI-2 may promote dispersion of H. pylori in vivo. H. pylori also senses disruptions to the ETC via TlpD, but the physiological change required to induce this sensing is unknown [38,52,53]

In addition to chemorepellents, many of which are harmful to the bacterium, H. pylori also senses several beneficial chemoattractants. One of these chemoattractants, urea, is sensed by TlpB [40,54–56]. In non-infected individuals, urea is available at concentrations between 5–21mM within the stomach [57], and is hydrolyzed by H. pylori urease into ammonia and bicarbonate to buffer its local environment [7]. Arginine is sensed as a chemoattractant by TlpA [40,58,59], and is an essential amino acid for H. pylori [60,61]. In both urea and arginine chemotaxis, chemoattraction may help H. pylori find key nutrients, while also directing it toward the epithelial surface.

There are several additional chemotactic signals that have not been extensively studied. For example, H. pylori uses chemotaxis to migrate to the site of gastric damage [62]. However, the exact host-derived chemicals that direct this response are unknown. Another attractant that has not been mapped to a chemoreceptor is cholesterol [63].

Chemotaxis during in vivo colonization

Che⁻ H. pylori mutants have severe colonization defects (Figure 2), and appear to interact with host tissue differently based on their aberrant inflammatory responses (Figure 2). These aspects are discussed in the following sections.

Figure 2.

Chemotaxis promotes localization and modulates inflammation in vivo. Wild-type (WT) H. pylori robustly colonizes gastric glands early during infection, and can replicate and spread between glands. Conversely, non-chemotactic (Che⁻) H. pylori fail to colonize gastric glands to the same degree. Concurrently, WT H. pylori initiates a proinflammatory immune response, with a substantial T-helper 17 cell (Th17) response, while non-chemotactic H. pylori promotes a dampened immune response, skewed toward T-regulatory cells (Tregs).

Chemotaxis is required for wild-type colonization

Chemotaxis is required for wild-type (WT) level colonization of the stomach, based on studies using a variety of Che⁻ mutants, lacking CheA, CheY, CheW, or ChePep. In some cases, Che⁻ mutants did not colonize at all [16,32], and in others they colonized poorly [14,15,33,49,62,64]. Che⁻ mutants require 100-fold more bacteria to initiate colonization [33], thus some studies might have employed H. pylori doses that were below the infectious dose. Overall, these results show that chemotaxis is critical during the initial stages of infection, particularly with low infectious doses, such as could be experienced during human infection.

Chemotaxis during early infection

For roughly the first three months of an infection, chemotaxis is critical. During this period, Che⁻ mutants display significant colonization defects that are greater in one region, the antrum [33,49]. The stomach is extensively invaginated into over 25,000 glands, and Che⁻ mutants fail to robustly colonize these niches, especially within the antrum [12,64,65]. During the first month of a WT infection, the number of glands infected and the amount of H. pylori per gland increases. Che⁻ mutants, however, infect few glands initially and fail to promote this same population expansion between and within glands [64].

Chemotaxis during chronic infection

A hallmark of H. pylori colonization is its ability to achieve chronic infection. The role of chemotaxis during the chronic stage has been more difficult to evaluate. However, it seems that in the absence of challenge by WT, Che⁻ H. pylori are able to achieve WT levels of colonization as early as one month post infection, maintaining comparable levels up to six months post infection and presumably longer [14,33,49,64]. During the chronic stage of infection, Che⁻ H. pylori colonize gastric glands in the corpus and antrum, while WT populations also colonize the mucus layer of the corpus [64]. Additionally, Che⁻ mutants have been observed to be less closely associated with the gastric epithelium during this stage [14].

Competition

Che⁻ H. pylori have general colonization defects when they are the sole infecting strain, with even greater colonization defects observed during competition experiments. When mice are infected with equal doses of WT and Che⁻ H. pylori, the Che⁻ mutant is outcompeted by over 1000-fold [33,36]. If the infections are given sequentially, a secondary WT infection can displace a primary population of Che⁻ H. pylori, nearly displacing the entire gland population of Che⁻ bacteria. In contrast, a secondary WT infection is unable to displace a primary population of WT [33,64]. This finding suggests that chemotaxis is needed to maintain colonization during chronic colonization, especially when WT H. pylori are present.

Role of chemoreceptors for colonization

Chemoreceptors head the chemotaxis signal transduction system. Accordingly, several studies have examined the in vivo phenotypes of chemoreceptor mutants. Loss of individual chemoreceptors alters the sensing profile of H. pylori, but does not cause a complete loss of chemotactic ability. Specifically, bacteria would lose the ability to sense some compounds, perhaps biasing bacteria towards or away from signals sensed by the remaining chemoreceptors. Such a change could lead H. pylori towards more inhospitable environments.

tlpA and tlpC mutants colonize mice to WT levels during single strain infections, but are outcompeted by WT during competition infections [36]. This phenotype suggests that WT exacerbates the mutant’s defect in some way, possibly by competing for nutrients or causing a host response that the mutant cannot avoid. This idea would fit well with the function of TlpA for finding arginine [58,59].

tlpB mutants have either no or subtle defects that appear more pronounced later in infection [14,37,49,56]. Interestingly, tlpB mutants are not outcompeted by WT [14,16], suggesting that WT does not exacerbate the mutant’s defect. This outcome suggests TlpB’s signal, urea, is not a limiting substance for H. pylori, and fits well with its relatively high (mM) levels [57]. The late infection defects observed for tlpB mutants could be related to enhanced inflammation (see below) or to defects associated with AI-2 chemotaxis [14,50]

In single strain infections, tlpD mutants have substantial colonization defects early in infection, greater than Che⁻ H. pylori [49,52]. Additionally, tlpAD double mutants have an even greater colonization defect, and are defective in antral gland colonization [42]. Treatment of tlpAD-infected mice with omeprazole, which blocks acid production, rescues H. pylori levels to that seen in tlpD single strain infections [42,49]. This result suggest the inability to sense acid via TlpA and D reduces in vivo fitness, and indicates the remaining colonization defect in tlpD infections is potentially due to an inability to sense ROS or ETC conditions [38,39].

Role of auxiliary chemotaxis proteins in colonization

Loss of the auxiliary chemotaxis proteins creates H. pylori mutants that are either fully or partially Che⁻ [12,23,25,26]. Likewise, mutants lacking the CheVs or ChePep display colonization defects that range from severe, similar to fully Che⁻ mutants, to less severe [12,25,65]. Overall, data supports that auxiliary proteins can have as substantial an effect as core proteins.

Chemotaxis modulates host inflammation

Host inflammation is a major disease outcome of H. pylori colonization. Inflammation occurs upon recognition of microbial-associated molecular patterns (MAMPs) or damage associated molecular patterns (DAMPs) by local monocytes, macrophages, and epithelial cells. These cells release pro-inflammatory cytokines and chemokines, the latter of which recruits neutrophils and antigen presenting cells, including macrophages and dendritic cells, to the site of infection [66,67]. Dendritic cells release cytokines and, process and present H. pylori antigens, priming T-cell differentiation [7,66–71]. H. pylori colonization induces the differentiation of pro-inflammatory T-helper 1 (Th1) and T-helper 17 (Th17) cells and anti-inflammatory T-helper 2 (Th2) and T-regulatory cells (Tregs) [7,66,70,71]. The degree of host inflammation is modulated by the balance of effector T-cell populations. H. pylori induces inflammation via its intimate interaction with the gastric epithelium and the production of virulence factors such as the type 4 secretion system, CagA, NapA, and VacA [7,66,70]. The inflammatory response to H. pylori colonization, however, is also modulated by chemotaxis [14–16].

Chemotaxis was first shown to modulate inflammation using a tlpB mutant in a gerbil model of infection [16]. tlpB mutants colonized gerbils to WT levels at four weeks post infection, but induced only low gastric inflammation, as measured using histological enumeration of lymphocytes. While tlpB mutants induced less inflammation than WT, tlpB infected gerbils had high neutrophil recruitment to the gastric tissue, such as seen a few days after H. pylori infection. This phenotype suggested they were mimicking an early stage of infection [16].

Subsequent studies extended this work to the mouse model, examining the roles of all chemoreceptors [14]. Three months post-infection, Che⁻, tlpA, and tlpB mutants induced modestly less inflammation than WT. At later time points, six months post-infection, tlpA and tlpB mutants caused high inflammation, while Che⁻ mutants induced low inflammation. At both time points, all strains colonized to WT levels. Overall, these results made it clear that the inflammatory response was modulated by bacterial properties controlled by chemotaxis, and was affected independent of overall bacterial load [14].

To gain insight into why Che⁻ mutants induce less inflammation, Rolig et al. (2011) examined the specific immune cell populations recruited to the stomach. While Che⁻ mutants induce less histologically-evident inflammation, the total number of CD4+ T-cells recruited was equivalent between mice infected with WT or Che⁻ H. pylori two months post infection [15]. WT infections, however, were found to have higher amounts of pro-inflammatory Th17 cells, as assessed by expression of associated genes, and an overall elevated ratio of Th17/Treg cells compared with Che⁻ infections. A possible explanation for the ability of WT to trigger a Th17 response came from the observation that WT infections induced more gastric tissue apoptosis than did Che⁻ infections [15]. Because apoptosis is a host response that can lead to the differentiation of Th17 cells [72], it seemed plausible that the inability to trigger apoptosis partially explains why Che⁻ infections lack Th17 cells. It is still not fully understood why WT H. pylori induces more apoptosis and a more robust Th17 response, however, it is now known that Che⁻ mutants are mislocalized [12,14,64]. Mislocalization possibly induces modified interactions with distinct cell types compared to WT infections, modulating the resulting inflammatory response. The reason for the elevated inflammation of tlpA and tlpB mutants is still under study.

Conclusions

Chemotaxis is one of several colonization factors utilized by H. pylori to promote chronic infection of the stomach [7,12,14,16,25,32,33,49]. Work over the past 20 years strongly supports that chemotaxis helps H. pylori find nutrients such as urea and arginine, and avoid toxic substances such as acid and ROS. The list of critical compounds sensed by H. pylori will undoubtedly grow as more work is done to better understand the sensing profile of each chemoreceptor. Chemotaxis plays roles in localization and modulating the host immune response. The stomach landscape is comprised of multiple niches, and studies support that chemotaxis is particularly critical for colonization during early infection, especially in the antrum, and for spread into new glands [12,49,64]. During established infections, chemotaxis is less critical for colonization, perhaps because H. pylori has been able to stochastically access the glands and does not demand chemotaxis for growth in these pockets. Instead, in late-stage infections, the role of chemotaxis in inflammation control becomes apparent [14–16]. Understanding the mechanisms by which chemotaxis modulates host inflammation will facilitate our understanding of how bacteria control this important process, and could lead to future treatments to modify disease outcomes. Several pathogens are motile and chemotactic. H. pylori is leading our understanding of the roles for this important process in vivo, and it will be of great interest to assess how many of the same principles translate across bacterial systems and organs.

Highlights.

• The H. pylori chemotaxis system includes classical and auxiliary chemotaxis proteins

• H. pylori senses several host and bacterial ligands through four chemoreceptors

• Chemotaxis is critical for colonization during acute infection

• Chemotaxis modulates host inflammation during chronic infection