Abstract
Itaconate is a dicarboxylic acid that is able to inhibit the isocitrate lyase enzyme of the bacterial glyoxylate shunt. Activated macrophage have been shown to produce itaconate, suggesting that these immune cells may employ this metabolite as a weapon against invading bacteria. Here we demonstrate that itaconate can exhibit bactericidal effects under acidic conditions resembling the pH of a macrophage phagosome. In parallel, successful pathogens including Salmonella have acquired a genetic operon encoding itaconate degradation proteins, which are induced heavily in macrophage. We characterize the regulation of this operon by the neighbouring gene, itaR, in specific response to itaconate. Moreover, we develop an itaconate biosensor based on the operon promoter that can detect itaconate in a semi-quantitative manner and, when combined with the itaR gene, is sufficient for itaconate-regulated expression in E. coli. Using this biosensor with fluorescence microscopy, we observe bacteria responding to itaconate in the phagosomes of macrophage and provide additional evidence that interferon-γ stimulates macrophage itaconate synthesis and that mouse macrophage produce substantially more itaconate than human cells. In summary, we examine the role of itaconate as an antibacterial metabolite in mouse and human macrophage, characterize the regulation of Salmonella’s defense against it, and develop it as a convenient itaconate biosensor and inducible promoter system.
Importance In response to invading bacteria, immune cells can produce a molecule called itaconate, which can inhibit microbial metabolism. Here we show that itaconate can also directly kill Salmonella when combined with moderate acidity, further supporting itaconate’s role as an antibacterial weapon. We also discover how Salmonella recognizes itaconate and activates a defense to degrade it, and we harness this response to make a biosensor that detects the presence of itaconate. This biosensor is versatile, working in Salmonella or harmless E. coli, and can detect itaconate quantitatively in the environment and in immune cells. By understanding how immune cells kill bacteria and how the microbes defend themselves, we can better develop novel antibiotics to inhibit pathogens such as Salmonella.
Introduction
The mammalian immune system includes a multitude of weapons to defend against invading microbes and successful pathogens have evolved a plethora of mechanisms to evade, manipulate, or even benefit from these immune responses. One such pathogen, Salmonella enterica serovar Typhimurium (hereafter referred to as Salmonella), has acquired a number of Salmonella pathogenicity islands (SPI) that support its survival inside of a host organism. For instance, Salmonella employs SPI-1 to invade non-phagocytic cells, and SPI-2 allows the bacteria to survive intracellularly – including in macrophage – which is important for Salmonella virulence (1–4). These traits allow Salmonella to invade the gut epithelium and induce intestinal inflammation resulting in the characteristic gastroenteritis disease. Moreover, the induced inflammation is not merely a threat that Salmonella must survive, but it has adapted to thrive in the oxidative environment of the inflamed intestine and utilize inflammation-derived metabolites to outcompete resident microbiota (5–7).
Itaconate is a metabolite originally recognized in fungal species such as Aspergillus terreus and produced commercially for use in polymer industries (8, 9). A dicarboxylic acid comprised of succinate with a methylene group, itaconate is able to act as an inhibitor of the glyoxylate shunt enzyme AceA (isocitrate lyase) and can inhibit bacterial growth on carbon sources that require this pathway, such as acetate (10–12). Interestingly, it has been demonstrated that activated macrophage employ the IRG1 protein to produce itaconate, with higher concentrations being produced by mouse macrophage than human (12–14). Moreover, IRG1 closely associated with vesicles containing Legionella pneumophila and itaconate showed bactericidal activity against this pathogen in vitro (14). Itaconate was also found to inhibit Salmonella growth by reducing media pH and itaconate levels correlated with splenomegaly in Salmonella-infected mice (15). Cumulatively, these works suggest that itaconate acts as a weaponized metabolite that the immune system employs to inhibit or kill invading bacteria.
If itaconate is an antibacterial metabolite functioning for the immune system then it follows logically that successful pathogens must have methods to evade its effects. Indeed a genetic operon in Yersinia (ripABC for ‘required for intracellular proliferation’) encodes proteins catalyzing the degradation of itaconate into pyruvate and acetyl-CoA (16). This operon is not restricted to Yersinia and a variety of other pathogens encode homologs. These include the Salmonella genes STM3120-STM3118 that, with STM3117, comprise an operon in SPI-13 that we refer to as the ‘itaconate response operon’ (IRO). Interestingly, the IRO genes of Salmonella have been shown to be induced heavily in macrophage but not under any other condition tested, supporting a role in degrading macrophage-produced itaconate (17–19). High throughput screens have suggested that genes from this operon are important for Salmonella survival in mice (20–22). Furthermore, it has also been shown that SPI-13 is present in generalist but not human-restricted serovars of Salmonella, possibly due to reduced itaconate synthesis in human cells (23).
In this work we show that itaconate is bactericidal at low but not neutral pH and elucidate the regulation of the Salmonella IRO and its induction in mouse and human macrophage. We show that the promoter of the IRO (Pitac) is specifically induced by itaconate and that the LysR family transcriptional regulator encoded by the upstream gene, STM3121 (which we propose to name itaR for ‘itaconate regulator’), is both necessary and sufficient for this induction. Furthermore, using Pitac with a GFP reporter, we develop a semi-quantitative itaconate biosensor and employ it to show that the IRO is induced heavily in the J774 mouse macrophage cell line but requires interferon-γ (IFN-γ) stimulation to show a substantial response in the THP-1 human macrophage cell line.
Results
Itaconate is bactericidal at low pH
Previous works have demonstrated that itaconate can inhibit the function of the glyoxylate shunt enzyme, AceA, and act as a bacteriostatic agent when bacteria rely on carbon sources such as acetate that require this pathway (10–12). It has also been suggested that itaconate can inhibit bacteria by influencing media pH and at least one publication has demonstrated that it can have bactericidal activity (14, 15). To clarify this later phenotype we hypothesized that the dicarboxylic acid chemistry of itaconate would allow it to act in a bactericidal fashion at low pH by acting as a proton shuttle. In brief, the carboxyl groups of itaconate (pKa = 5.5 and 3.8) protonate and lose their charge at lower pH allowing them to traverse the bacterial membrane and release the protons in the more neutral pH of the cytoplasm, thereby exacerbating acid stress (Figure 1A).
To emulate the intracellular conditions that Salmonella may encounter in a Salmonella containing vacuole (SCV) of a macrophage we added itaconate to LPM media and then acidified to pH 4.4, 5.0, or 5.8 to cover a range from the most acidic to more regular estimates of SCV pH (24–26). Indeed we found that wild-type Salmonella showed a 1000-fold decrease in survival after 3h hours at pH 4.4 with itaconate (Figure 1B). This lethality was alleviated at higher pH and also occurred using a similar dicarboxylic acid, succinate. Importantly, the bactericidal effect was also dependent on the presence of itaconate (or succinate) and pH 4.4 LPM did not kill Salmonella in the absence of a dicarboxylic acid (Figure S1A). Interestingly, deletion of the entire IRO (STM3120-STM3117) or aceA had no effect, but deletion of the general stress response sigma factor, RpoS (σS), exacerbated Salmonella’s sensitivity at both pH 4.4 and 5.0 (Figures 1C and S1B). Cumulatively, these data demonstrate that itaconate or other dicarboxylic acids can act in a bactericidal fashion under acidic conditions by exacerbating acid stress.
The Salmonella IRO is induced specifically by itaconate in an itaR-dependent manner
The Salmonella pathogenicity island-13 includes genes encoding Ccl (STM3120), Ich (STM3119) and Ict (STM3118), whose homologs have been demonstrated to degrade itaconate into pyruvate and acetyl-CoA (16). This operon, which we refer to as the itaconate response operon (IRO), appears to also include the STM3117 gene (encoding a predicted glyoxalase-domain containing protein) and is adjacent to the STM3121 (itaR) gene on the reverse DNA strand (Figure 2). In light of its potential to degrade itaconate, we hypothesized that this metabolite may act as an inducer of IRO expression. To assess this, we constructed a plasmid-borne fusion of the operon’s promoter (Pitac) to superfolder GFP (sfGFP) as a reporter. Indeed, we found that the Pitac promoter was induced highly in the presence of itaconate and this response was entirely dependent on the presence of ItaR as neither a Salmonella itaR deletion mutant nor the same reporter plasmid in E. coli K12 (which does not encode itaR) showed induction (Figure 3A). In contrast, when the itaR gene was included on the reporter plasmid, itaconate-induced Pitac expression was restored in both the ΔitaR Salmonella strain and in E. coli. These data not only demonstrate that Pitac is induced in response to itaconate, but also that the neighbouring gene, itaR, is both necessary and sufficient for this induction.
To assess if the IRO promoter is induced specifically by itaconate, we examined induction of the pPitac reporter in media supplemented with a panel of similar metabolites. While mesaconate, citramalate and methylsuccinate (in order of induction strength) slightly induced expression, induction by itaconate was drastically more pronounced, suggesting that it is the principal inducer (Figure 3B). Notably, similar results were obtained in complex media (LB) and in MOPS minimal media with either glucose or glycerol as a carbon source, suggesting that the induction only requires itaconate and not additional factors in the media (Figure S2). Furthermore, induction by itaconate occurred in a dose dependent manner indicating that the reporter can be used to semi-quantitatively assess itaconate concentrations encountered by the bacteria (Figure S3).
The Salmonella IRO does not significantly contribute to survival in a mouse macrophage cell line
The inhibitory effect of itaconate on AceA, its bactericidal activity under acidic conditions, and the synthesis of itaconate in macrophage combine to support the concept that these immune cells may be employing itaconate as an antibacterial compound. As a successful pathogen, Salmonella has adapted to survive in activated macrophage and multiple previous works have examined how IRO genes may influence Salmonella survival and virulence (20–23). In our hands, we found no significant reduction in survival of ΔIRO or ΔitaR strains in the mouse J774 macrophage cell line (Figure S4). When the macrophage were pre-stimulated with IFN-γ, there was a slight reduction in survival relative to wild-type, but this was not significant when compared to an aceA mutant that showed no survival defect. In contrast, a phoP mutant control was readily inhibited by the macrophage even without IFN-γ.
Salmonella encounter itaconate in the phagosomes of mouse and IFN-γ-stimulated human macrophage
Multiple high throughput studies have demonstrated that the IRO genes are induced heavily in mouse macrophage (17–19). Additional studies have identified itaconate in both mouse and human macrophage but the mouse cells appear to produce significantly more of the metabolite (12, 13). Furthermore, it has been demonstrated that interferon-β (IFN-β) and IFN-γ can stimulate itaconate production in mouse macrophage (14, 27–29).
To examine itaconate levels encountered by intracellular bacteria inside macrophage, we employed our Pitac-sfGFP reporter plasmid as a biosensor. By including a constitutively expressed mCherry gene on the same plasmid, we could microscopically observe individual bacteria inside macrophage and obtain semi-quantitative data by generating a GFP/mCherry fluorescence ratio as an indicator of Pitac induction and accordingly itaconate levels. Using this system we observed strong induction of the Pitac promoter for wild-type Salmonella in J774 mouse macrophage and this signal was absent in the ΔitaR control (Figure 4). Furthermore, the response could also be observed in E. coli if ItaR was encoded on the same plasmid, demonstrating that bacteria that are poorly adapted to intracellular survival also encounter itaconate in macrophage.
In contrast to the mouse cell line, unstimulated human THP-1 macrophage showed negligible itaconate levels as very few of the bacterial reporters showed any green fluorescence above background levels (Figure 4). The bacteria did express the constitutive mCherry and could be observed in the macrophage, suggesting that the lack of green fluorescence was not due to decreased bacterial survival or protein expression. Intriguingly, the human cells did appear to synthesize itaconate when stimulated with human IFN-γ (M1 activation) leading to significant green fluorescence of the reporter bacteria in an ItaR-dependent fashion. In contrast, itaconate levels in THP-1 cells induced with IL-4 and IL-13 (M2 activation) resembled unstimulated cells (Figure S6).
Discussion
In this work we demonstrate that itaconate becomes bactericidal at acidic pH, suggesting an additional mechanism for itaconate to act as an antibacterial metabolite beyond inhibition of AceA. Thus, elevated itaconate levels in macrophage may act to inhibit bacterial metabolism while also exacerbating acid stress on microbes in the phagosome. Protonation of itaconate under acidic conditions also grant it increased access to the bacterial cytoplasm where de-protonation would trap the charged form close to its AceA target. This organic acid killing effect has been demonstrated previously, including in a recent work showing propionate inhibition of Salmonella in mice (30, 31). Of note, we also find that bacterial killing occurs with succinate, a metabolite similar to itaconate that similarly increases in concentration in activated macrophage (32). Our findings that these dicarboxylic acids can kill Salmonella at pH 4.4 but not higher may contribute to why Salmonella manipulates the SCV to maintain a pH closer to 5.0 in order to avoid this organic acid stress. Moreover, they imply that organic acids such as itaconate and succinate may contribute to the antibacterial activity of acidified phagosomes.
The antibacterial potential of itaconate, its synthesis in activated macrophage, and the localization of IRG1 to bacteria-containing vacuoles, support its potential role as a weapon against intracellular bacteria. Here we examined survival of Salmonella IRO or itaR deletion strains in the mouse macrophage J774 cell line but saw no significant decrease in survival, similar to a recent study examining SPI-13 in RAW264.7I macrophage (23). However, in that work, Espinoza et al. discovered that SPI-13 does play a role in Salmonella internalization into mouse – but not human – macrophage (23). Combined with previous works showing reduced survival of IRO mutants in mice, this operon may play a more significant survival role in the context of an animal infection (19–22).
Here we showed that Salmonella responds to itaconate in vitro and intracellularly by strongly inducing an operon encoding itaconate degradation proteins. This response is largely specific to itaconate and is entirely dependent on the neighbouring gene, STM3121, which we propose to rename itaR for ‘itaconate regulator’. The itaR gene product is predicated to be a LysR family transcriptional regulator, suggesting that itaconate induces IRO expression by interacting directly with the substrate binding domain of ItaR to activate it. Moreover, we find that ItaR is sufficient for itaconate induction of the Pitac promoter in E. coli, demonstrating its potential as a novel inducible expression system with over 50-fold higher transcription in the presence of the inexpensive and readily available inducer. A limitation of this expression system would be a requirement for growth on carbon sources independent of the glyoxylate shunt and also growth at neutral or alkaline pH, as we demonstrate that itaconate is bactericidal under acidic conditions. However, for many studies these conditions are met, adding Pitac to the repertoire of available inducible promoter systems.
Using our Pitac-sfGFP itaconate biosensor, we showed a pronounced response in unstimulated mouse macrophage whereas no induction was observed in the THP-1 human macrophage cell line without stimulation, suggesting that these cells are not producing itaconate to the same degree. Interferon has previously been demonstrated to stimulate itaconate production in mouse macrophage and we found that our biosensor was induced in human cells stimulated with IFN-γ (14, 27–29). Thus, while the human cell line was able to produce itaconate, it required auxiliary induction to do so and still produced less than the uninduced mouse macrophage. While it is possible that this reflects an artifact of the cell lines employed, it aligns well with previous works that quantified itaconate in both mouse and human cells (12, 13). Furthermore, Espinoza et al. recently determined that SPI-13 is abundant in generalist Salmonella serovars but not human-restricted ones (which instead encode SPI-8), suggesting the hypothesis that human-restricted serovars may not require an IRO because they do not encounter high levels of itaconate in humans (23).
While further work will have to be done to determine if human cells truly produce less itaconate than other species, itaconate could potentially open the door to novel antimicrobials. Indeed, a recent study demonstrated that small molecules can inhibit the activity of the IRO proteins and sensitize Salmonella to itaconate inhibition in minimal media (33). Such drugs could also sensitise other bacteria to itaconate including Yersinia, Pseudomonas and Mycobacteria species, which also encode an IRO (16, 33). Moreover, if human-restricted pathogens lack an IRO because human cells truly produce less itaconate, then they are potentially sensitive to it and itaconate itself could potentially be used as an antimicrobial against them. Our biosensor could prove invaluable in such studies for determining how much itaconate the bacteria are encountering, and the self-sufficiency of the biosensor allows it to be employed in a variety of species, providing added versatility.
In summary, here we present data that itaconate can act as a bactericidal metabolite at acidic but physiologically relevant pH. We identify the regulatory mechanism of an itaconate response operon in Salmonella and employ its promoter as a novel biosensor of relative itaconate concentrations in macrophage phagosomes. Finally, we provide further evidence that IFN-γ stimulates itaconate synthesis and moreover that human cells produce less of the metabolite than their mouse equivalents.
Materials and Methods
Bacterial strains and plasmids
All Salmonella strains used in experiments are derivatives of Salmonella enterica subsp. enterica serovar Typhimurium (S. Typhimurium) strain 14028s. As described previously, lambda red recombination and subsequent P22 phage transduction was used to generate all of the gene knockout mutants in this background (34–36). To allow for subsequent recombinations, the antibiotic resistance cassette was removed from the chromosome using the pCP20 plasmid encoding FLP recombinase (37). The heat-unstable pCP20 plasmid was eliminated by passaging overnight at 42°C and loss was confirmed by antibiotic treatment.
A reporter fusion of the IRO promoter to sfGFP (Pitac-sfGFP) was generated using Gibson cloning to insert the 333bp upstream of the STM3120 start codon (thereby including 25bp upstream of the predicted −35 box and the 5’ untranslated region) into the pXG10sf plasmid (replacing the existing promoter)(38–40). For the reporter construct including itaR, the same region was extended to 1570bp upstream of the STM3120 start codon to include the entire STM3121 ORF and a predicted transcriptional terminator following it. For fluorescence microscopy, constitutively expressed (PLtet0-1 promoter) mCherry was inserted into a transcriptionally independent region of the same plasmid. This variation of the plasmid was renamed ‘independent constitutive mCherry’ or pICM.
Metabolite induction of Pitac assay
Induction of the Salmonella Pitac promoter was assessed using a transcriptional fusion to sfGFP in either the pXG10sf or pICM plasmids. Data from the two plasmids were combined as the inducible region is identical and the plasmids only differ in the constitutively active mCherry expressed independently on pICM. Overnight LB cultures were used to inoculate (1/200 dilution) either LB or MOPS minimal media containing 0.2% of the indicated carbon source. Itaconate or other metabolites at neutral pH were supplemented to a concentration of 0.2%. Of note, for salts and hydrates the final 0.2% concentration reflects the percent of the carbon source itself; e.g. 0.2% succinate was made as 0.47% sodium succinate (dibasic) hexahydrate. Growth was conducted in a TECAN Infinite M200 plate reader at 37°C with shaking and OD600 and GFP fluorescence (475nm and 511nm excitation and emission wavelengths respectively) were read every 15 minutes. For clarity, bar graphs show fluorescence at 16h post inoculation. Chloramphenicol was included in all media at a concentration of 20μg/ml to maintain the plasmids.
Acidified media survival
LPM media was made as described previously (41, 42). Succinate or itaconate were added to 0.4% and the pH was then adjusted to 4.4, 5.0 or 5.8 as indicated. LB overnight cultures were resuspended in acidified media to an OD of 0.1 and incubated in a 37°C water bath. At time points, samples were taken, serial diluted and plated for colony forming units (CFU).
Intra-macrophage survival
The THP-1 human monocyte cell line and the J774 mouse macrophage cell line were maintained in RPMI Medium 1640 (with L-glutamine) supplemented with 10% FBS and 1% Glutamax, and grown at 37°C and 5% CO2. THP-1 cells were seeded in 96-well plates at 50,000 per well with 50nM PMA (phorbol 12-myristate 13-acetate) added to induce differentiation to adherent macrophage. After 48h, media was replaced with no-PMA growth media overnight with 100 U/ml human IFN-γ or IL-4 and IL-13 added if indicated. For J774 macrophage the cells were seeded in 96-well plates at 50,000 per well overnight with 100 U/ml mouse IFN-γ added if indicated. Salmonella in RPMI were added onto seeded cells at a multiplicity of infection (MOI) of approximately 20:1 and centrifuged for 10 minutes at 1000rpm to maximize cell contact. After centrifuging the samples were incubated at 37°C and 5% CO2 (time 0). After 30 minutes, cells were washed three times with PBS followed by fresh media containing 100 μg/ml gentamicin to kill extracellular Salmonella. At 2 hours the media was replaced with media containing gentamicin at 10 μg/ml. At timepoints, intracellular bacteria were recovered using PBS containing 1% Triton X-100 and vigorous pipetting. Samples were serially diluted and five 10μl spots were plated for CFU counting. Each sample included three separate wells as technical replicates (a total of 15 × 10μl spots counted per biological replicate).
Fluorescence microscopy
Fluorescence microscopy was conducted similarly to the macrophage survival assay with some exceptions: Cells were seeded in 24-well plates containing glass coverslips at 125,000 per well. Bacteria were infected at an MOI of approximately 100 to maximize the instances of macrophage containing bacteria. At timepoints the media was removed and cells were washed three times with PBS. They were then fixed for 10 minutes at room temperature in PBS + 4% paraformaldehyde (PFA). Following three more PBS washes the cells were permeabilized for 10 minutes in PBS + 0.2% Triton X-100 + 1% BSA. Coverslips were washed again, mounted on slides using 3μl mounting media containing DAPI and allowed to dry overnight in the dark. Slides where viewed using a Zeiss Observer.z1 microscope using a 100× oil immersion objective and the Zeiss Zen microscopy software. Images were taken with a Zeiss Axiocam 506 mono camera mounted on the microscope. For all samples a 2s exposure was used for mCherry and 1s exposure for sfGFP. ImageJ software was employed for quantification to calculate fluorescence intensities in the red and green channels relative to a neighbouring background region for each bacterium and a GFP/mCherry ratio was generated.
Acknowledgements
We would sincerely like to thank Dr. Scott Gray-Owen and members of his lab, in particular Dr. Ryan Gaudet, for their generous donation of technical expertise, macrophage cells lines, and use of their equipment. WWN was supported by an Operating Grant from the Canada Institutes for Health Research (MOP-86683) and a Natural Sciences and Engineering Research Council (NSERC) of Canada Grant (RGPIN 386286-10). SJH was supported by an NSERC Vanier Canada Graduate Scholarship.