The protozoan parasite Trypanosoma brucei is the causative agent of the neglected tropical disease human African trypanosomiasis, otherwise known as sleeping sickness. Trypanosomes have evolved many immune-evasion mechanisms to facilitate their own survival, as well as prolonging host survival to ensure completion of the parasitic life cycle. A key feature of the bloodstream form of T. brucei is the secretion of aromatic keto acids, which are metabolized from tryptophan. In this study, we describe an immunomodulatory role for one of these keto acids, indole-3-pyruvate (I3P). We demonstrate that I3P inhibits the production of PGs in activated macrophages. We also show that, despite the reduction in downstream PGs, I3P augments the expression of cyclooxygenase (COX2). This increase in COX2 expression is mediated in part via inhibition of PGs relieving a negative-feedback loop on COX2. Activation of the aryl hydrocarbon receptor also participates in this effect. However, the increase in COX2 expression is of little functionality, as we also provide evidence to suggest that I3P targets COX activity. This study therefore details an evasion strategy by which a trypanosome-secreted metabolite potently inhibits macrophage-derived PGs, which might promote host and trypanosome survival.
This article is featured in Top Reads, p.
Human African trypanosomiasis (HAT) is one of the world’s neglected diseases. Trypanosoma brucei, an extracellular parasitic protozoan of the genus Trypanosoma, is the causative agent. The disease is commonly known as sleeping sickness in humans and nagana in animals and can be fatal if left untreated. These diseases remain a major burden in sub-Saharan Africa for humans and livestock. Treatments remain limited, and drug resistance is an increasing feature (1).
A molecular aspect of the bloodstream stage of HAT is a perturbation of aromatic amino acid metabolism. Serum levels of tryptophan in infected animals are significantly decreased relative to healthy controls (2–5). This decrease is inversely linked to a simultaneous increase in aromatic keto acids, such as indole-3-pyruvate (I3P), phenylpyruvate, and hydroxyphenylpyruvate in the serum and urine (6–9). These keto acids are formed by the transamination of tryptophan by the parasitic enzyme aspartate aminotransferase (10, 11). The abnormal levels of aromatic keto acids in the blood and urine of infected animals, detectable by a pungent odor and red or brown color of the urine, have traditionally been used as a diagnostic tool to identify infected livestock (12).
During the course of infection, trypanosomes proliferate to extremely high numbers in the blood of infected individuals, at times reaching up to 0.2–1 × 109 cells/ml blood (13). The constant exposure of the parasite to the host immune system places the parasite under unique evolutionary pressure to both evade these unfavorable immune responses while also prolonging host survival to increase the likelihood of disease dissemination to other hosts via the tsetse fly. The most notable mechanism of immune evasion by trypanosomes is the rapid antigenic variation of the variable surface glycoproteins that entirely cover the surface of the trypanosome, which enables T. brucei to evade host Abs (14, 15). This interplay between host defense and parasite-evasion strategies gives rise to waves of parasitemia as the numbers of trypanosomes rise and fall cyclically, meaning that the concentrations of I3P will also fluctuate over the course of infection.
Macrophages are also known to be key effectors of host immune defense during infection. Recent reports have described how the keto acids that T. brucei secrete impair proinflammatory macrophage functions. I3P was found to decrease HIF-1α levels in LPS-stimulated macrophages, thereby impairing glycolysis and the proinflammatory cytokine IL-1β (9). Several T. brucei–derived keto acids, including I3P, were also shown to activate the NRF2 antioxidant pathway and reduce secretion of IL-6 from microglia (16). I3P has also been shown to activate the aryl hydrocarbon receptor (AhR) (17–19), which has multiple immunological roles.
Production of eicosanoids is another crucial arm of the innate immune response, and eicosanoids are divided into three classes: PGs, thromboxanes, and leukotrienes. These are lipid mediators, many of which have potent proinflammatory functions (20). PGs and thromboxanes are synthesized downstream of cyclooxygenase (COX) activity, the enzyme that nonsteroidal anti-inflammatory drugs, such as indomethacin and aspirin, target. There are two isoforms of COX. COX1 is ubiquitously and constitutively expressed, whereas COX2 is inducible by inflammatory stimuli, such as LPS. In particular, myeloid-derived PGE2 has been shown to be proinflammatory, promoting fever and even recently being implicated in aging-associated inflammation (21). PGE2 can bind to four different E prostanoid receptors (EP1–4) and through engagement of these receptors can activate other immune cells, including mast cells (22), Th1, and Th17 cells (23), thereby enabling PGE2 to both initiate and prolong inflammation. The significance of the proinflammatory impact of PGE2 and other PGs is highlighted by the widespread use of nonsteroidal anti-inflammatory drugs in the treatment of inflammatory diseases (24).
Alterations in PG secretion have been demonstrated during infection with T. brucei in mice. Macrophages isolated from infected mice during the first peak of parasitemia displayed enhanced secretion of PGE2, whereas macrophages taken at a later stage of infection had reduced capacity for both basal and LPS-induced PGE2 secretion (25). Inhibiting PG production could promote host survival by limiting the febrile response. In addition, PGs have been shown to be directly trypanocidal (26, 27). However, the mechanism of PG modulation by T. brucei has not been yet elucidated.
In this study, we describe a potential role for I3P during trypanosomiasis involving inhibition of PG production. Somewhat counterintuitively, I3P increases COX2 mRNA and protein, in a manner that is partially AhR-dependent and may also be the result of a feedback loop induced by low PG concentrations. However, this transcriptional boost of COX2 is of little functional consequence given that we also provide evidence that indicates that I3P blocks COX2 activity. We also show that this could have relevance for trypanosomiasis, as I3P also inhibits induction of PGs by T. brucei lysates, as well as inhibiting PG production in primary human macrophages. Our study therefore describes a potential immune-evasion mechanism by which a T. brucei–secreted metabolite suppresses host macrophage PG synthesis, possibly promoting survival of host and parasite.
Materials and Methods
LPS derived from Escherichia coli, serotype EH100 (Enzo Life Sciences), I3P, 3-methylcholanthrene, indomethacin, arachidonic acid (AA), AH6809, forskolin (Sigma-Aldrich), murine recombinant IFN-γ (ImmunoTools), and GW 627368X (Cayman Chemical) were used. Silencer Select control small interfering RNA (siRNA), Silencer Select AhR siRNA (assay ID s62162), and Lipofectamine RNAiMAX Transfection Reagent (Thermo Fisher Scientific) were also used. Abs used were anti-COX2 (Abcam) and anti–β-actin (Sigma-Aldrich), as well as anti-mouse IgG and anti-rabbit IgG secondary HRP-conjugated Abs (Jackson ImmunoResearch Laboratories). A PGE2 ELISA kit was also used (Enzo Life Sciences).
Mice and bone marrow–derived macrophage generation
Bone marrow–derived macrophages (BMDMs) were isolated from C57BL/6J mice (Harlan UK). Bones from AhR−/− mice and their matched wild-type mice were kindly provided by Prof. Brigitta Stockinger (Francis Crick Institute, London, U.K.). All animals were maintained under specific pathogen-free conditions in accordance with Irish and European Union regulations. All experiments were subject to prior ethical approval by Trinity College Dublin Animal Research Ethics Committee and the Health Products Regulatory Authority. Mice were euthanized in a carbon dioxide chamber, followed by cervical dislocation as confirmation of death. Bone marrow cells were flushed from the tibia, femur, and hip of the mice and differentiated in DMEM, which contained 1% penicillin/streptomycin, 10% FCS, and 20% L929 supernatant. After 6 d, the cells were counted and replated for experiments.
Primary human macrophage culture
Human PBMCs were isolated from buffy coats from healthy donors using Lymphoprep (Axis Shield). The blood was diluted 1:1 with PBS, and then 30 ml was layered on 20 ml Lymphoprep and spun for 20 min at 2000 rpm with no brake on. The middle layer of PBMCs was then transferred to a new tube and washed in PBS. To isolate the monocytes from these PBMCs, positive selection of CD14+ cells was carried out using anti-CD14–labeled magnetic beads (Miltenyi Biotec), according to the manufacturer’s instructions. The monocytes were then maintained in RPMI 1640 media, which contained 1% penicillin/streptomycin, 10% FCS, and 50 ng/ml M-CSF (ImmunoTools). After 6 d, the monocytes had differentiated into macrophages and were then used for experiments.
Trypanosome culture and lysis
Monomorphic MITat 1.1 bloodstream forms were cultured in HMI9 medium containing 10% FCS. Cells in the log phase of growth were harvested by centrifugation at 1500 × g for 5 min and lysed by a combination of osmotic shock and sonication. After centrifugation, the cells were resuspended at 1 × 109 cells/ml in PBS glucose buffer (3 mM NaH2PO4, 57 mM Na2HPO4, 44 mM NaCl, 5 mM KCl, and 1 mM MgCl2 [pH 8]) and subsequently diluted 1 in 10 with sterile water. The resulting suspension was subjected to sonication (30% power, 50% duration; Bandelin Sonopuls) for 1 min before being incubated at room temperature for 5 min. The lysate was centrifuged at 14,000 × g for 10 min to yield a soluble lysate fraction. This soluble lysate fraction was subsequently used to stimulate BMDMs.
siRNA transfection of BMDMs
Lipofectamine RNAiMAX Transfection Reagent (Thermo Fisher Scientific) was preincubated with the relevant siRNA and diluted in DMEM media containing no FCS or penicillin/streptomycin. This mixture was then used to treat BMDMs, yielding final concentrations of 5 μl/ml Lipofectamine RNAiMAX Transfection Reagent and 50 nM siRNA. The cells were then left for 8 h, before this media was replaced with DMEM containing 1% penicillin/streptomycin and 10% FCS. At 48 h posttransfection, the cells were used for experiments as required.
Cells were lysed in sample buffer (0.125 M Tris [pH 6.8], 10% [v/v] glycerol, and 0.02% SDS) and heated to 95°C for 5 min. The protein samples and Spectra BR protein ladder (Thermo Fisher Scientific) were then resolved on SDS-polyacrylamide gels, before being transferred to polyvinylidene fluoride membrane. Membranes were blocked for 1 h in 5% (w/v) dried milk in TBST before being incubated overnight at 4°C with the primary Ab. Following incubation for 1 h with the secondary Ab, the blots were developed using chemiluminescent substrate (Thermo Fisher Scientific).
Cells were lysed and RNA extracted using the PureLink RNA Minikit (Ambion). cDNA was then prepared using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems), according to the manufacturer’s instructions. Real-time quantitative PCR (qPCR) was then carried out on the resulting cDNA using a 7500 Fast Real-Time PCR System with PowerUp SYBR Green Master Mix (Applied Biosystems). The primer pair sequences used were as follows: Rps18, 5′-GGATGTGAAGGATGGGAAGT-3′ (forward) and 5′-CCCTCTATGGGCTCGAATTT-3′ (reverse); Ptgs2, 5′-CGGACTGGATTCTATGGTGAAA-3′ (forward) and 5′-CTTGAAGTGGGTCAGGATGTAG-3′ (reverse); Ptges, 5′-GGAAGAAGGCTTTTGCCAACC-3′ (forward) and 5′-CGAAGCCGAGGAAGAGGAAA-3′ (reverse); Pgis, 5′‐TCACCACGCACCCATGAG‐3′ (forward) and 5′‐TGGCGGAAGGTATGGAAAACATG-3′ (reverse); Tbxas1, 5′-CAGGTGTTGGTTGAGAACTT-3′ (forward) and 5′-ACTGAACCTGACCGTACATGTCACGTAAAAACAGAACG-3′ (reverse); and Ahr, 5′-GGCAGGATTTGCAAGAAGGAG-3′ (forward) and 5′-TGGATGACATCAGACTGCTGAA-3′ (reverse).
The cycling threshold method (2−ΔΔCt) was used to calculate relative quantification after normalization of each gene to murine Rps18. The calibrator used was the sample treated with neither LPS nor I3P. For knockdown experiments, the calibrator used was the sample treated with control siRNA but neither LPS nor I3P.
Cell supernatants were collected, and PG concentrations were measured using an ELISA kit for PGE2 (Enzo Life Sciences), according to the manufacturer’s instructions. Although this item is sold as a PGE2-specific ELISA kit, our data suggested that other COX-derived oxylipins were also detected with this kit. Therefore, we refer to measurements carried out using this ELISA kit throughout the article as quantifying PGs instead of PGE2.
To measure intracellular PG levels, cells were harvested in lysis buffer (1 mM EDTA in distilled water) and sonicated. Protein concentrations were normalized using the Pierce BCA Protein Assay kit (Thermo Fisher Scientific), as per the manufacturer’s instructions. PG concentrations were measured using the same ELISA kit as for supernatants.
Cell supernatants were collected and snap frozen in liquid nitrogen immediately. Samples were spiked with 2.1–2.9 ng PGE2-d4, PGD2-d4, 20-HETE-d6, 5-HETE-d8, 12-HETE-d8, 15-HETE-d8, 13-HODE-d4, and thromboxane B2 (TXB2)-d4 standards (Cayman Chemical) prior to extraction. Lipids were extracted by adding a 2.5-ml solvent mixture (1 M acetic acid/isopropanol/hexane; 2:20:30 [v/v/v]) to 1 ml supernatant in a glass extraction vial and vortexed for 30 s. A total of 2.5 ml hexane was added to samples, and after vortexing for 30 s, tubes were centrifuged (500 × g for 5 min at 4°C) to recover lipids in the upper hexane layer (aqueous phase), which was transferred to a clean tube. Aqueous samples were re-extracted as above by addition of 2.5 ml hexane, and upper layers were combined. Lipid extraction from the lower aqueous layer was then completed according to the Bligh and Dyer technique. Specifically, 3.75 ml of a 2:1 ratio of methanol/chloroform was added followed by vortexing for 30 s. Subsequent additions of 1.25 ml chloroform and 1.25 ml water were followed with a vortexing step for 30 s, and the lower layer was recovered following centrifugation as above and combined with the upper layers from the first stage of extraction. Solvent was dried under vacuum, and lipid extract was reconstituted in 100 μl HPLC-grade methanol. Lipids were separated by liquid chromatography (LC) using a gradient of 30–100% B over 20 min (A: water/Mob B 95:5 plus 0.1% acetic acid; B: acetonitrile/methanol 80:15 plus 0.1% acetic acid) on an Eclipse Plus C18 Column (Agilent Technologies) and analyzed on a Sciex QTRAP 6500 LC-MS/MS system. Source conditions were: temperature 475°C, ion spray voltage–4500, gas 1 60, gas 2 60, and curtain gas 35. Lipids were detected using multiple-reaction monitoring with the following precursor to product ion transitions: PGE2 and PGD2 [M-H]− 351.2/271.1, 15-deoxy-PGJ2 315.2/271.1, TXB2 369.2/169.1, 5-HETE 319.2/115.1, 8-HETE 319.2/155.101, 9-HETE 319.2/167.1, 11-HETE 319.2/167.102, 12-HETE 319.2/179.1, 15-HETE 319.2/219.1, 5-HEPE 317.2/115.1, 8-HEPE 317.2/155.1, 9-HEPE 317.2/167.1, 11-HEPE 317.2/167.101, 12-HEPE 317.2/179.1, 15-HEPE 317.2/219.1, 18-HEPE 317.2/259.1, 4-HDOHE 343.2/101.1, 7-HDOHE 343.2/141.1, 8-HDOHE 343.2/189.1, 10-HDOHE 343.2/133.101, 11-HDOHE 343.2/121.1, 13-HDOHE 343.2/193.1, 14-HDOHE 343.2/205.1, 16-HDOHE 343.2/233.101, 20-HDOHE 343.2/241.101, 9-HODE 295.2/171.1, 13-HODE 295.2/195.1, 5-HETrE 321.2/115.1, 15-HETrE 321.2/221.1, 9,10-DiHOME 313.2/201.1, 12,13-DiHOME 313.2/183.1, 8,9-DiHETrE 337.2/127.1, 11,12-DiHETrE 337.2/167.1, 14,15-DiHETrE 337.2/207.1, 14,15-DiHETE 335.201/207.1, and 17,18-DiHETE 335.2/247.1. Deuterated internal standards were monitored using precursor to product ion transitions of: TXB2-d4 [M-H]− 373.2/173.1, PGE2-d4 and PGD2-d4 355.2/275.1, 5-HETE-d8 327.2/116.1, 12-HETE-d8 327.2/184.1, 15-HETE-d8 327.2/226.1, 13-HODE-d4 299.2/198.1, and 20-HETE-d6 325.2/281.1. Chromatographic peaks were integrated using Multiquant 3.0.2 software (Sciex). The criteria for assigning a peak were signal/noise of at least 5:1 and with at least seven points across a peak. The ratio of analyte peak areas to internal standard was taken and lipids quantified using a standard curve made up and run at the same time as the samples.
COX activity assay
COX activity was measured using the COX Activity Assay Kit (Cayman Chemical) according to the manufacturer’s instructions. A total of 1 μg recombinant human COX2 (R&D Systems) was used per well, and for this experiment, DMSO was used as the solvent for I3P rather than cell culture media.
Statistical significance was determined by the one-way or two-way ANOVA methods as described in the figure legends. For the one-way ANOVA, data were analyzed with no matching or pairing. Gaussian distribution and equal SDs were assumed. Multiple comparisons were performed comparing the mean of each column to the mean of every other column. The Tukey test was used for correction for multiple comparisons. For the two-way ANOVA, multiple comparisons were performed comparing each cell mean with the other cell mean in that row. The Šidák test was used for correction for multiple comparisons. Data were expressed as mean ± SEM. Significance was defined as follows: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. GraphPad Prism v9 software was used for statistical analysis.
I3P inhibits LPS-induced PG and thromboxane secretion by macrophages
To investigate if I3P would modulate macrophage PG production, we treated murine BMDMs with I3P before stimulation with the TLR4 agonist LPS for 24 h. LPS, although derived from Gram-negative bacteria, is relevant in the study of T. brucei infection. It has been demonstrated that the concentration of LPS in the circulation is substantially elevated during trypanosomiasis (28–30), and it has been suggested that the presence of LPS contributes to disease pathology through enhancement of the host proinflammatory response (30). We first measured PGE2 using an ELISA and found a robust increase in response to LPS, which was potently inhibited by I3P pretreatment at concentrations as low as 62.5 µM (Fig. 1A). To confirm this result, a lipidomic screen of oxylipins was carried out using quantitative LC–tandem mass spectrometry (LC-MS/MS). In this study, LPS stimulation yielded a robust increase in PGE2 (Fig. 1B), although the levels were considerably lower for LC-MS/MS due to lower specificity of Ab-based methods for lipid quantitation. Using LC-MS/MS, we also found that LPS stimulated elevations in PGD2 (Fig. 1C), 15-deoxy-PGJ2 (Fig. 1D), and TXB2 (Fig. 1E). Generation of all of these was potently inhibited by I3P pretreatment, confirming and extending the ELISA result. Using LC-MS/MS, we also measured lipoxygenase-derived and cytochrome P450–derived oxylipins in our supernatants; however, due to some background levels of these in tissue culture serum, elevations in response to LPS were not clearly seen, and no consistent impact of I3P was observed (Supplemental Figs. 1–3). Thus, we concluded that I3P was blocking eicosanoids specifically at the level of COX activity, likely through preventing generation of PGs and TXB2. Because the ELISA is reporting ∼10-fold higher levels of PGE2 than LC-MS/MS, we considered it was reporting on COX-derived PGs more broadly. As confirmation, preincubation of BMDMs with the pan-COX inhibitor indomethacin completely blocked generation of lipids detected using this method (Fig. 1F).
I3P augments LPS-induced COX2 expression
We next investigated if the changes we had observed in PG production were due to changes in expression of any of the enzymes in the pathway. As COX2 is known to be potently upregulated by LPS stimulation (31) and is often referred to as the rate-limiting enzyme for this pathway, we first measured COX2 expression in BMDMs with I3P treatment. Although LPS increased expression of Ptgs2, the gene that encodes COX2, pretreatment with I3P further boosted Ptgs2 transcript (Fig. 2A). When protein levels of COX2 were examined by Western blot, a similar trend was observed with I3P further increasing LPS-induced COX2 expression at both the 6-h (lane 4 compared with lane 3) and 24-h time points (lane 6 compared with lane 5) (Fig. 2B, 2C). This I3P-induced increase in COX2 expression was seen at 1000 μM and 500 μM, but not at lower concentrations of I3P (Fig. 2D, 2E). The observation of this I3P-induced increase in COX2 expression was unexpected given that I3P had greatly reduced the secretion of PGs, which are dependent on COX2 expression in macrophages. We also measured the expression levels of a number of other enzymes involved in PG/thromboxane synthesis. We saw no changes with I3P treatment in the transcript levels of Ptges (Fig. 2F), which encodes PGE2 synthase, nor did we observe any changes in Pgis (Fig. 2G), which encodes PGI2 synthase. I3P caused a significant decrease in the gene encoding thromboxane A synthase, Tbxas1, at the 8-h time point (Fig. 2H). However, this modest reduction is unlikely to account for the potent decrease in TXB2 observed with I3P (Fig. 1E), nor could it account for the inhibition of PGs by I3P. We also tested whether I3P would affect intracellular PG levels in the same manner as secreted PGs. I3P reduced both intracellular and extracellular concentrations of PGs (Fig. 2I), implying that I3P is unlikely to affect PG transport. Thus, in terms of expression of enzymes in the PG synthesis pathway, I3P appears to significantly affect COX2 levels only, boosting its expression while inhibiting PG production.
I3P targets COX2 activity
Given the potent inhibitory effect that I3P had on LPS-induced PGs and thromboxanes, we wondered whether I3P may be a direct COX inhibitor. One of the best-characterized COX inhibitors, indomethacin, is also built on an indole-3-acetic acid framework and therefore shares structural similarities to I3P (Fig. 3A). As already shown, I3P was able to elicit a similar inhibitory effect compared with indomethacin, measuring PGs by ELISA (Fig. 1F). As had been previously reported in the literature (32, 33) and in a similar manner to I3P, indomethacin also boosted LPS-induced COX2 expression (Fig. 3B, 3C, lanes 5 and 6 compared with lane 4), albeit to a lesser degree than I3P (lanes 11 and 12 compared with lane 10). Due to these similarities to the archetypal COX inhibitor indomethacin, we hypothesized that I3P may be inhibiting COX2. First, we supplemented the media with AA, the substrate of COX2 (Fig. 3C), to elucidate if COX2 was the point of modulation. When exogenous AA was added, I3P could still potently inhibit LPS-induced PGs (Fig. 3E), which indirectly suggested that COX2 activity was being blocked by I3P, as the addition of AA was insufficient to overcome the effect. We next tested if I3P could directly inhibit COX2 activity, by adding varying concentrations of I3P to recombinant COX2 and assaying COX activity. We found that I3P potently inhibited COX activity with an IC50 of 68.97 μM (Fig. 3F), thus confirming that I3P functions as a COX2 inhibitor.
The increase in COX2 expression by I3P may depend on a PG feedback loop
We hypothesized that perhaps the I3P-mediated inhibition of COX2 activity and downstream reduction in PGs was part of a negative-feedback loop by which decreased levels of PGs would cause an upregulation of COX2 transcription. We found that the addition of exogenous PGE2 modestly attenuated the ability of I3P to boost LPS-induced COX2 expression (Fig. 4A, 4B). PGE2 can bind to four different EP receptors (EP1–4). The expression of EP2 has been particularly implicated in macrophage function (34, 35), and binding of PGE2 to the EP2 receptor or the EP4 receptor causes an increase in cAMP production (36–38). Treatment with the EP1/2 antagonist AH6809 augmented the increase in COX2 with I3P treatment (Fig. 4C, 4D, lanes 7 and 8 compared with lanes 3 and 4). Treatment with the EP4 antagonist GW 627368X also amplified the increase in COX2 with I3P treatment (Fig. 4E, 4F, lanes 7 and 8 compared with lanes 3 and 4). Furthermore, pretreatment with forskolin, which raises intracellular cAMP levels and thereby mimics engagement of the EP2 and EP4 receptors, gave rise to decreased COX2 expression in response to LPS and I3P (Fig. 4G, 4H, lanes 7 and 8 compared with lanes 3 and 4). These data suggest the possibility of a feedback loop by which low concentrations of PGs and lack of EP2 and EP4 engagement cause a transcriptional increase in COX2 expression.
The boost in COX2 expression by I3P is also partially AhR-dependent
As I3P has previously been shown to activate the AhR (17–19) and it is known that the COX2 promoter contains a corresponding xenobiotic-responsive element (39, 40), we reasoned that the AhR may also play a role in the I3P-induced increase in COX2 that we observed. Treatment with the AhR agonist 3-methylcholanthrene (3MC) increased LPS-induced COX2 expression in a similar manner to I3P (Fig. 5A, 5B, lane 4 compared with lane 6). However, 3MC had no effect on COX2 levels in AhR−/− BMDMs (Fig. 5A, 5B, lane 10 compared with lane 12). Similarly, the capacity of I3P to augment LPS-induced COX2 expression was partially impaired in the AhR−/− BMDMs, compared with the induction of COX2 observed in the wild-type BMDMs (Fig. 5C, 5D, lanes 11 and 12 compared with lanes 5 and 6). These data suggest that the increase in COX2 expression by I3P is partially dependent on activation of the AhR receptor. To test if AhR was also involved in PG inhibition by I3P, we knocked down AhR in BMDMs using siRNA, which successfully lowered transcript levels of Ahr (Fig. 5E). When AhR was silenced, I3P could no longer induce a boost in Ptgs2 transcript (Fig. 5F). Although loss of AhR yielded lower PG levels, I3P still significantly inhibited PG secretion (Fig. 5G), suggesting that the augmentation of COX2 expression by I3P is AhR-dependent but inhibition of PG production by I3P is not.
I3P inhibits trypanosome lysate-induced PGs and LPS-induced PGs in human macrophages
To demonstrate relevance during trypanosomiasis, we next investigated the effect of I3P on PGs induced by T. brucei lysate. It has previously been reported that for macrophages to respond to trypanosomes, they require stimulation with IFN-γ (9, 41), hence we cotreated the BMDMs with trypanosome lysate and IFN-γ, with or without I3P pretreatment. The cells that were pretreated with I3P prior to stimulation with IFN-γ and trypanosome lysate displayed enhanced levels of COX2 protein (Fig. 6A, 6B, lanes 8 and 9 compared with lanes 5 and 6) and secreted much less PGs (Fig. 6C). This mirrors the effects we have seen with I3P on LPS-induced COX2 and PGs. To investigate relevance of these findings for HAT, we tested if I3P had the same effects on human macrophages. The cells pretreated with I3P prior to stimulation with LPS displayed enhanced COX2 expression (Fig. 6D, 6E, lanes 4–6 compared with lane 3) and reduced PG production (Fig. 6F), thereby indicating that these effects may play a role in human infection. The secretion of I3P by T. brucei and the impact it has on PG/thromboxane production in macrophages through COX inhibition is depicted in (Fig. 6G.
Recently, the T. brucei–secreted metabolite I3P has emerged as an important immunomodulator, particularly regarding the regulation of host macrophage function (9, 16). In this study, we describe another mechanism by which I3P dampens the host immune response. We have provided evidence that I3P inhibits PG and thromboxane secretion from macrophages. We also have shown that I3P increases COX2 expression levels beyond those of LPS stimulation alone, contrary to the decrease in COX-dependent PGs. Our data indicate that the I3P-induced boost in COX2 may be mediated by two distinct mechanisms. We show that this COX2 boost is partially dependent on limiting a feedback loop involving endogenous PGE2 acting via EP2 and EP4 receptors. In addition, we provide evidence that the effect is, at least partially, dependent on AhR activation. However, it would seem that this increase in COX2 expression by I3P has little functional effect, as I3P dramatically reduces PGs and thromboxanes downstream of COX2, despite this transcriptional increase. In addition, the fact that higher concentrations of I3P are required to increase COX2 expression may call into question the physiological relevance of this observation during infection in vivo. In contrast, I3P inhibits PG production at concentrations as low as 62.5 μM and therefore is likely to yield similar effects in vivo. At these higher concentrations, the increase in COX2 expression might have other functional consequences because COX2 has been shown to interact with and inhibit p53, affecting cell survival (42). This requires further investigation in the context of trypanosomiasis. Despite this COX2 induction, we provide evidence showing that I3P inhibits COX2 activity and have demonstrated that this inhibition of PG production is also observed when macrophages are stimulated with trypanosome-derived lysates. Furthermore, we have shown that the effects of I3P on COX2 expression and PG production are also true for primary human macrophages.
Given its structural similarity to indomethacin, it is likely that I3P inhibits COX1 as well as COX2 but in macrophages, this is difficult to dissect. It has been reported that PGs in macrophages are more dependent on COX2 than COX1 (43), and indeed, our data show very little, if any, PG secretion without stimulation of the macrophages to induce COX2 upregulation. Therefore, we cannot conclude from our data if I3P inhibits COX2 activity only or both isoforms of the enzyme.
It was reported several decades ago that PG secretion was perturbed during trypanosomiasis. Macrophages taken from mice during an early stage of a T. brucei infection model displayed enhanced secretion of PGE2 compared with uninfected controls, whereas at later stages of infection, the isolated macrophages secreted reduced concentrations of PGE2, and their ability to respond to LPS was notably impaired (25). Our findings may shed light on the mechanism by which macrophage PGE2 secretion is hindered later in the course of disease. As the parasites proliferate to great numbers in the bloodstream (13), they will metabolize tryptophan to generate increasing concentrations of I3P (8), a metabolite that we have shown blocks COX2 activity and thereby lowers PGs. However, this proposed mechanism may in reality be complicated by the fact that trypanosomes have been reported to produce PGs of their own, which are known to interfere with host responses (27, 44).
Naturally, these data pose the question: why would T. brucei have evolved to inhibit host PGs? One possibility could be that it is beneficial to the parasite to inhibit the pyrogenic effects of PGE2, which are well characterized (45–47). This might aid host survival. Another potential reason could be to limit levels of PGD2. PGD2 and the PGD2-derived J series PGs have been shown to induce programmed cell death in trypanosomes in a manner dependent on reactive oxygen species (26, 27). Therefore, inhibition of PGs may be a method of self-preservation for T. brucei. However, as previously mentioned, the fact that trypanosomes make their own PGs (27, 44), which include PGD2, is intriguing. Perhaps it is advantageous for them to inhibit host-derived PGs so that they alone can control the levels of PGD2 and other PGs, potentially as a form of population control or quorum sensing.
Our results reveal a potential mechanism of immunomodulation by T. brucei, mediated by one of the trypanosome-secreted keto acids, I3P. These data provide insight into the interaction between T. brucei and the host immune system and could potentially inform new therapeutic strategies to limit trypanosomiasis via modulation of I3P.
We thank Professor Brigitta Stockinger (Francis Crick Institute, London, U.K.) for providing legs from AhR+/+ and AhR−/− mice.
This work was supported by the European Research Council (834370), Science Foundation Ireland (12/IA/1531), and the Wellcome Trust (205455). The work was also supported by a Sêr Cymru Project Sepsis grant funded by the Welsh Government/European Union European Regional Development Fund (to V.J.T. and V.B.O.). V.B.O. is a Royal Society Wolfson Research Merit Award Holder.
The online version of this article contains supplemental material.
Abbreviations used in this article
aryl hydrocarbon receptor
bone marrow–derived macrophage
human African trypanosomiasis
liquid chromatography–tandem mass spectrometry
small interfering RNA
The authors have no financial conflicts of interest.