Airway immunometabolites fuel Pseudomonas aeruginosa infection

The human airway is replete with carbon sources for bacterial proliferation, essential microelements such as iron (Fe⁺⁺) and nutrients like amino acids and lipids [32]. Ubiquitous opportunists, such as P. aeruginosa commonly found in the environment, colonizing sinks and showers as well as streams, are often inadvertently inhaled. The vast majority of these organisms are rapidly cleared, but those that persist must forcibly adapt to the microenvironment provided by the airway.

Much of what is known about the composition of airway fluid has been obtained from studies of CF patients, focusing primarily upon the concentrations of Na⁺, Cl⁻, HCO3⁻ and water in the setting of infection [32‐34]. In studies using LPS as a surrogate for bacterial infection, macrophages very rapidly switch their resting metabolic activity from oxidative phosphorylation (OXPHOS) to glycolysis, with the resulting accumulation, oxidation and release of succinate and its byproduct ROS [35‐38]. Both metabolites activate the hypoxia induce factor 1α (HIF-1α) and IL-1β production initiating a proinflammatory response [35‐37] (Fig. 1a, b), which is associated with the recruitment and activation of phagocytes. Thus, IL-1β is an important proinflammatory cytokine expected to facilitate the eradication of P. aeruginosa. However, in several models of pneumonia, IL-1β, in fact, contributes to P. aeruginosa infection [39‐41]. Mice lacking a receptor for IL-1β (il1r^−/−) or those treated with its antagonist Anakinra actually clear P. aeruginosa significantly better than those with a robust IL-1β response. The inflamed CF airway with many activated macrophages and monocytes is enriched for succinate, in comparison to normal control broncheoalveolar lavage (BAL) fluid, and elevated amounts of succinate have also been observed in animal models mimicking the CF disease [30, 42, 43] (Fig. 1a, b). Thus, an important substrate for bacterial proliferation, succinate, is immediately available for bacterial consumption upon their entry into the airway, and is provided by activated myeloid cells.

Many different types of inhaled bacteria are likely to be entrapped by the dehydrated airway secretions that are prominent in CF; but in fact, only a few species cause persistent infection. These opportunists, most prominently P. aeruginosa must rapidly adapt to the available metabolites, including macrophage succinate. In contrast to many other bacteria that prefer glucose or amino acids, P. aeruginosa preferentially consumes succinate as directed by its crc locus, before utilizing other carbon sources [44‐48] (Fig. 2a). However, forced succinate consumption generates substantial oxidant stress for the organisms, which are already in an oxidant rich environment with activated phagocytes releasing ROS [14, 20, 24, 48]. As a response, P. aeruginosa variants are selected that utilize the anti-oxidant glyoxylate shunt (GS) to dissipate ROS and to generate biofilm, which itself has anti-oxidant properties [25, 26, 30, 49].

Laboratory strains of P. aeruginosa grown in high levels of succinate demonstrate metabolic changes that enable them to proliferate amidst high levels of oxidants [30] (Fig. 2b). These bacteria induce even more succinate in the airway as well as myeloid cell death, consistent with increased secretion of IL-1β and pyroptosis [30]. Succinate-exposed strains increased their glucose metabolism and utilization of threalose and acetate, which feed the GS, production of EPS and enable tolerance to oxidant stress. P. aeruginosa grown in high succinate were phenotypically different, with increased colony size, consistent with the abundant production of EPS. These strains caused significantly greater levels of infection in mouse models of pneumonia, decreased myeloid cell viability and promoted more IL-1β release. Importantly, the same constellation of metabolic and anti-oxidant changes was identified in a collection of clinical P. aeruginosa strains cultured from an adult with CF [30, 50, 51]; namely, altered carbon substrate utilization, increased use of the GS and expression of genes associated with EPS and biofilm formation. Both the clinical strains from CF and the P. aeruginosa grown in high succinate in vitro were more proficient in colonizing the airways of mice, causing persistent infection that lasted for days. Thus, the ability to metabolize succinate and modify their own metabolic activity in response to the ROS generated by this immunometabolite enables P. aeruginosa to proliferate and adapt to the pro-oxidant airway environment.

PTEN, by controlling the P3K/Akt/mammalian target of rapamycin (mTOR) pathway regulates cellular proliferation, glycolytic metabolism [52] and mitochondrial activity [53‐55]. PTEN participates in the regulation of mitochondrial respiration, especially controlling the function of cytochrome C[54] and the assimilation of the oxidative TCA cycle substrate isocitrate [30]. PTEN dysfunction induces increased expression of isocitrate dehydrogenase (IDH), which overstimulates mitochondrial complex I by providing with more nicotinamide adenine dinucleotide (NADH)[30]. Increased complex I activation induces more oxygen reduction and ROS, which, during LPS stimulation, is reinforced by electrons reversely transferred from complex II, succinate dehydrogenase (SDH)[37]. This excessive ROS production activates the anti-oxidant cell response, which, in part, is mediated by itaconate. This metabolite is the product of Irg1 (in humans Acod1), and its activity is linked to, for example, inhibition of SDH [56‐59]. Thus, PTEN deregulation induces oxidative stress in mitochondria and, as a compensatory mechanism, itaconate synthesis, which inhibits SDH and enable succinate accumulation and its release [30, 60]. For proper PTEN metabolic activity, this phosphatase associates with the C-terminal cytoplasmic tail of CFTR, which is known to increase its stability and activation by dephosphorylation [61]. In cells (and patients) with decreased membrane-associated CFTR, or in subjects harboring mutations in the CFTR C-terminal tail, there is also decreased PTEN numbers and increased succinate accumulation [42, 61] (Fig. 2). This deregulation is associated with more airway inflammation [61, 62]. P. aeruginosa infection of myeloid and epithelial cells induces both PTEN reduction and succinate release, and peripheral blood mononuclear cells (PBMCs) from individuals exhibiting CFTR-PTEN deficiency secrete much more succinate levels than controls after infection with these organisms [30, 61]. The airway of CFTR-PTEN deficient subjects showed tenfold succinate accumulation respect with healthy individuals, demonstrating that the metabolic activation status of epithelial and myeloid cells can impact the pro-oxidant composition of the airway metabolome. As the CFTR-PTEN association is not dependent upon the channel function of CFTR, therapeutic strategies that increase the membrane localization of CFTR should also increase PTEN and normalize succinate, inflammation and also reduce P. aeruginosa infection.

Itaconate, an electrophilic carboxylate, is also a prominent metabolite found in the infected airway [30, 59] (Figs. 1c, 3). As described above, itaconate is released by LPS activated macrophages in response to succinate oxidation by SDH, and functions to counter the proinflammatory response caused by ROS and IL-1β [63]. Itaconate and its derivatives 4-octyl itaconate and dimethyl-itaconate suppress inflammation by targeting Kelch-like ECH-associated protein 1 (KEAP1), which sequesters the anti-oxidant Nuclear Factor–Erythroid-2–Related Factor 2 (NRF2) transcription factor in the cytoplasm [64, 65]. Itaconate alkylates KEAP1, releasing NRF2 from proteosomal degradation and promoting its migration to the nucleus where it induces expression of multiple anti-oxidant genes such as HMOX1 (Heme-oxygenase 1), as well as suppress production of IFNβ [64]. Itaconate also suppresses secretion of inflammasome associated pro-inflammatory cytokines by dicarboxypropylation of pyrin domain-containing protein 3 (NLRP3), which reduces its interaction with NEK7 [66]. Reduced NLRP3-NEK7 complex formation abolishes the LPS-induced assembly of the inflammasome, preventing the production of, for example, IL-1β, and in PBMCs from subjects exhibiting the IL-1β based disease cryopyrin-associated periodic syndrome (CAPS) itaconate and 4-octyl itaconate reduce IL-1β production.

During Gram-negative infection of mucosal surfaces, itaconate participates in the metabolic control of inflammation by suppressing release of pro-inflammatory cytokines by interrupting the oxidation of succinate by SDH [63, 67]. Reduced transformation of succinate into fumarate diminishes the amounts of ROS required to inhibits prolyl hydroxylases (PHD), and impeding HIF1α migration to the nucleus to promote transcription of pro-glycolytic and pro-inflammatory genes [35]. Reduced SDH function due to itaconate accumulation safeguards the integrity of the mucosal tissue, as it induces less local damage and accelerates the process of tissue repair. Itaconate also interacts with macrophage aldolase, where it induces cysteine modifications that reduce its catalytic function and, eventually, inactivating glycolytic flux and the release of effector cytokines that might harm the lung [68]. Irg1 null mice are highly susceptible to the hyperinflammatory pathology associated with LPS [63]. In other pulmonary infections, the lack of itaconate enables massive recruitment of myeloid cells to the Mycobacterium tuberculosis infected lung, exacerbating inflammation, disrupting pulmonary function and reducing mouse survival [69]. Irg1 deficiency is associated with greater amounts of IL-1β accumulation in airway of mice infected with laboratory strains of P. aeruginosa [59], confirming the anti-inflammatory and protective role this metabolite has in the setting of pulmonary infection. The abundance of both anti—inflammatory itaconate and pro-inflammatory succinate in the airway suggests that successful pathogens adapt to both macrophage metabolites in order to persist.

In contrast to other organisms more susceptible to the electrophilic stress imposed by itaconate, such as Legionella pneumophila, S. aureus and Acinetobacter baumannii [70], P. aeruginosa has adapted to this metabolite, and are forced to degrade itaconate to survive. P. aeruginosa expresses 3 genes (ich, ict, icl) devoted to itaconate metabolism, enabling them to assimilate itaconate in the airway as carbon source [59, 70, 71] (Fig. 3a, b). This is a property shared by M. tuberculosis and the Aspergillus species [71, 72], suggesting an evolutive and conserved mechanism of adaption to host immunometabolites. The consumption of itaconate by the ict-ich-ccl locus generates acetyl-coA and pyruvate, which are feeders of the GS and TCA cycle, respectively and provide the components for biofilm production.

Clinical isolates of P. aeruginosa adapt to itaconate in vivo, with increased expression of the ict-ich-ccl locus, and induction of biofilm production. The clinical strains shift their preferred airway carbon source from succinate to itaconate. Itaconate stress causes increased bacterial outer membrane permeability, reducing the capacity of P. aeruginosa strains to transport LPS to their surface. These outer barrier function changes activate the algT-algD-MucA membrane stress repair system [73, 74], facilitating the replacement of LPS with more protective alginate polymers. Itaconate exposed laboratory strains of P. aeruginosa have diminished production of O-side chains of LPS [59], confirming that these organisms divert carbohydrates from endotoxin synthesis to the generation of EPS, in response to membrane stress. These same findings were apparent in clinical isolates of P. aeruginosa from CF patients, which had been chronically exposed to airway itaconate. These CF strains had developed mutations in the pathways involved in LPS surface display, specifically in the lptD locus [59]. LptD is the LPS transporter embedded in the outer membrane that flips and anchors the endotoxin towards the extracellular side [75, 76]. P. aeruginosa CF isolates from chronic infection fail to expose LPS nor do they induce macrophage release of succinate, as do environmental strains of P. aeruginosa. Of note, the EPS (alginate) produced by these isolates stimulates itaconate release from host immune cells [59]. The mechanisms by which P. aeruginosa EPS induces macrophage itaconate remains unclear. The changes in the display of EPS versus LPS on the surface of P. aeruginosa in chronic airway infection suggests that these organisms are forced to catabolize itaconate to prevent outer membrane disruption and biofilm clearance. The capacity of the ict-ich-ccl locus to provide P. aeruginosa with carbon structures such as acetyl-coA and pyruvate demonstrate its valuable contribution to the establishment of long-term lung colonization.

It is well appreciated that the induction of EPS moieties and biofilm production, especially the alginate overproducers characteristic of chronic infection in CF, correlate with the onset of intractable pulmonary infection [77]. Thus, the selection of these variants that induce itaconate production and then consume the metabolite, is of considerable clinical significance. The bacterial adaptive response, the production of EPS contributes to accumulation of phagocytes in the airways, but without resulting in bacterial clearance [78]. The prevailing dogma is that organisms enveloped in EPS, such as those with increased algD or pslA resist phagocytosis and are “less” immunogenic [24, 79, 80]. However, when itaconate-adapted clinical strains were examined for immunogenicity either in vivo or in vitro, they readily induced recruitment of immune cells into the airway [30, 59]. All of the strains stimulated variable amounts of IL-6 and TNFα, and recruited greater numbers of neutrophils and monocytes than control bacteria [30, 59]. The activation and release of substantial amounts of IL-1β was limited to the laboratory strain PAO1, perhaps due to its expression of LPS, flagella and toxins that directly activate the inflammasome and generate the release of succinate [39, 81] (Fig. 3c). The CF strains had lost expression of pathogen-associated molecular patterns (PAMPs) that are expected to induce IL-1β, specially LPS, flagella and the type 3 toxins [2, 16, 30, 59, 82], which was consistent with their failure to activate the release of succinate [35, 36] (Fig. 3d).

The immunogenicity of these P. aeruginosa CF isolates differed from that of the LPS-expressing laboratory strains, which have never been exposed to the itaconate-rich airway before [30, 59] (Fig. 3d). Although these CF strains failed to induce IL-1β, they stimulated myeloid cell reprograming generating itaconate. In sputum from CF patients infected with host-adapted P. aeruginosa, there is substantial accumulation of monocytes and macrophages that release itaconate. These findings indicate that the abundance of the immunometabolite itaconate has a major role in the selection of P. aeruginosa variants, promoting the display of EPS and not LPS on their surfaces and further contributing to the accumulation of itaconate and the establishment of intractable biofilms.