Abstract
Influenza epidemics remain a leading cause of morbidity and mortality worldwide. In the current study, we investigated the impact of chronological ageing on tryptophan metabolism in response to influenza infection.
Examination of metabolites present in plasma collected from critically ill patients identified tryptophan metabolism as an important metabolic pathway utilised specifically in response to influenza. Using a murine model of influenza infection to further these findings illustrated that there was decreased production of kynurenine in aged lung in an indoleamine-pyrrole 2,3-dioxygenase-dependent manner that was associated with increased inflammatory and diminished regulatory responses. Specifically, within the first 7 days of influenza, there was a decrease in kynurenine pathway mediated metabolism of tryptophan, which resulted in a subsequent increase in ketone body catabolism in aged alveolar macrophages. Treatment of aged mice with mitoquinol, a mitochondrial targeted antioxidant, improved mitochondrial function and restored tryptophan metabolism.
Taken together, our data provide additional evidence as to why older persons are more susceptible to influenza and suggest a possible therapeutic to improve immunometabolic responses in this population.
Abstract
Alterations in mitochondrial gene expression in aged lung during influenza contribute to alterations in metabolic response pathways; specifically, decreased kynurenine pathway mediated tryptophan metabolism and increased ketone body catabolism https://bit.ly/3fb64lF
Introduction
Influenza epidemics still remain a leading cause of morbidity and mortality worldwide, with the highest incidence of hospitalisation and death occurring in persons aged >65 years [1]. The kynurenine pathway (KP) of tryptophan metabolism is a highly regulated pathway utilised by the immune system to promote immunosuppression in response to excessive inflammation [2–9]. Kynurenine catabolites, such as quinolinate, are essential for the production of nicotinamide and nicotinamide adenine dinucleotide (NAD+) as well as initiate involvement of type 2 T-helper cell mediated resolution [7]. In response to heightened levels of cellular stress or energy usage, KP-mediated oxidation of tryptophan can renew NAD+ levels [10]. Tryptophan metabolism by the KP is initiated by indoleamine-pyrrole 2,3-dioxygenase (IDO1). IDO1 expression is tightly regulated by the immune system; specifically, IDO1 is activated by pro-inflammatory cytokines and inhibited by regulatory, anti-inflammatory cytokines [2, 3, 11–15].
Despite the important role for KP-mediated tryptophan metabolism in the initiation and regulation of immune response to influenza, little is known regarding the impact of ageing and ageing-associated changes in mitochondrial stress on this metabolic pathway. In the current study, using plasma collected from critically ill patients, we show that tryptophan metabolism is an important metabolic pathway utilised specifically in response to influenza. Using a murine model of influenza infection to further these findings, our results illustrate decreased production of kynurenine in aged lung that was associated with increased inflammatory and diminished regulatory responses. Reduced IDO1 expression and activity in aged lung and alveolar macrophages corresponded with changes in metabolic gene regulation and the initiation of alternative tryptophan metabolism pathways. Our findings demonstrated that decreased IDO expression was associated with dysregulated mitochondrial gene expression and heightened generation of reactive oxygen species. Treatment of aged mice with mitoquinol, a mitochondrial targeted antioxidant, improved mitochondrial function and restored KP-mediated metabolism of tryptophan in alveolar macrophages and resulted in decreased inflammatory cytokine production and cellular recruitment to the aged lung.
Methods
Detailed methods are provided in the supplementary material.
Study approval and subjects
This study is an analysis of prospectively collected data from 39 subjects recruited from an intensive care unit cohort from New York Presbyterian Hospital/Weill Cornell Medical Center (institutional review board number 1405015116). Protocols for recruitment, data collection and sample processing have been described previously [16–20].
Mouse model of influenza
Animal experiments were performed in accordance with the animal guidelines of the institutional animal care and use committee at Weill Cornell Medicine (2016–0059). Young (2 months) and aged (18 months) male and female BALB/c mice were purchased from the National Institute on Aging rodent facility (Charles River Laboratories, Shrewsbury, MA, USA). Influenza viral stock (material 10100374, batch 4XP170531, 50% egg infective dose 1010.3 per mL was purchased from Charles River (Norwich, CT, USA). Mice were anaesthetised with isoflurane prior to intranasal instillation with 12.5 PFU influenza (50 μL volume in PBS). Mice received a 100 μL volume of PBS or 10–50 μg dose of mitoquinol (Cayman Chemical, Ann Arbor, MI, USA) intraperitoneally starting at day 3 post-influenza. Starting at day 0, mice received a daily i.p. injection containing PBS or BMS-986205 (200 μM) (Selleck Chemicals, Houston, TX, USA).
Statistics
Statistical analysis was performed using Prism (GraphPad, La Jolla, CA, USA). All samples were independent and contained the same sample size for analysis. p-values <0.05 were considered significant.
Results
Increased tryptophan metabolism in response to influenza infection
Using a principal component analysis to provide a high-level overview of the dataset illustrated limited clustering of metabolites in plasma collected from control, influenza A/B positive patients and patients with other viral infections (OVI) (figure 1a; patient demographics and PCR confirmed microbiology results table 1). Focusing on specific biochemical and pathway changes illustrated a distinct metabolic increase in several metabolic pathways, including tryptophan metabolism, that occurred in patients during the course of influenza infection (figure 1a and b). When compared to influenza A/B samples, there were significantly elevated levels of tryptophan and serotonin present in the plasma of patients with OVI (figure 1c and d). In contrast, there was a significant increase in kynurenine and kynurenate present in plasma from the influenza A/B groups when compared to control, with significantly higher levels of kynurenate in influenza A/B positive patients (figure 1e and f). To provide more insight, we evaluated kynurenine levels in plasma samples based on influenza severity (i.e. mild versus severe influenza). Severe influenza was defined as a positive influenza test result combined with acute respiratory distress syndrome (ARDS) requiring intubation. While patients with mild symptoms during influenza exhibited an age-associated trend in kynurenine expression, patients who exhibited severe influenza with ARDS had decreasing kynurenine and increasing tryptophan with increasing age (supplementary figure S1a). While these metabolites were also elevated in the OVI patients, the levels did not reach statistical significance, suggesting that influenza A/B infections cause the strongest inflammatory response. The further breakdown products quinolinate and xanthurenate were also examined, with significantly increased levels of quinolinate present in influenza A/B positive patients (figure 1g and h).
Tryptophan metabolism is altered in aged lung during influenza infection
In response to influenza, there was cellular infiltration in both young adult (aged 2 months) and aged adult (aged 18–20 months) murine lung, with marked levels of immune cells infiltrating into the aged lung (figure 2a). As the course of influenza infection progressed, there was increased inflammation present in the aged lung that corresponded with increased viral titres, morbidity and mortality (figure 2a and b, supplementary figure S1b and c). Examination of the total bronchoalveolar lavage (BAL) cell count illustrated significantly increased cellular infiltration in aged lung (figure 2c). Lung permeability was assessed by intranasal instillation of fluorescein isothiocyanate (FITC)-dextran. As shown in figure 2d, there was a significant increase in lung permeability, as illustrated by elevated FITC fluorescence in plasma, in aged lung at baseline as well as in response to influenza. Similarly, there was a significant increase in BAL protein that corresponded with elevated water accumulation in the aged lung during infection (figure 2e and f). Given the importance of NAD levels in mediating tissue injury, we examined changes in NAD+ expression in young and aged lung tissue. Our results illustrated an age-dependent decrease in NAD+ in lung tissue at baseline and during the course of infection (figure 2g). When compared to young tissue, there was increased expression of interleukin (IL)6 and IL1β by day 5, with significantly enhanced production by day 7 (figure 2h and i). Initial expression of resolution cytokines, such as IL10, was detectable in the BAL collected on day 7 post-infection, with a significantly higher level present in BAL isolated from young lung (figure 2j). We next examined whether heightened IL10 expression correlated with increased infiltration of CD4+CD25+ regulatory T-cells. When compared to young lung tissue, despite increased baseline CD4+CD25+ T-cell numbers in aged lung tissue, there were significantly decreased number of cells present in aged lung tissue in response to influenza (figure 2k).
We observed that metabolic profiles of young and aged lung samples cluster separately at baseline as well as during influenza (figure 3a). Focusing on the biochemical and pathway changes in the profiles illustrated an important role for kynurenine and tryptophan metabolism in modulating immune responses in the lung in response to influenza (figure 3b). Specifically, kynurenine was increasingly elevated over time in lung isolated from both young and aged adult mice, with significantly higher levels being detected in the young lung at day 7 post-infection (figure 3c). While levels of kynurenate were similar in young and aged lung, the kynurenine catabolite, quinolinate, was augmented in young lung at later time points of influenza infection (figure 3d and e). There was an increase in the presence of tryptophan in the aged lung, with significantly higher levels present at day 7 of influenza infection (figure 3f). In addition, increased expression of downstream tryptophan metabolites, indoleacetate and indolepropionate, were detected in aged lung at later time points of infection (figure 3g and h).
We next examined if changes in IDO1 expression and activity in aged lung might underlie changes in tryptophan metabolism in response to influenza. In response to influenza, there was significantly reduced IDO1-specific activity at days 5 and 7 post-infection in both lung and BAL collected from aged adult mice (figure 4a and b). Levels of kynurenine in lung and BAL illustrated a similar phenotype in aged mice, with significantly higher levels at baseline that decline during the course of influenza infection (figure 4c and d). A corresponding increase in tryptophan was also present in aged lung and BAL, with significantly increased expression at days 5 and 7 post-infection (figure 4e and f).
IDO1 inhibition in young macrophages and lung contribute to increased IL6 expression and inflammation during infection
To elucidate the importance of IDO1 for influenza-mediated production of inflammatory cytokines, such as IL6, we examined the impact of several IDO1 inhibitors on modulating IL6 production by young bone marrow derived macrophages (BMDM) in response to lipopolysaccharide (LPS) or polyinosinic-polycytidylic acid (poly I:C) stimulation. In response to stimulation with LPS (24 h) there was a significant increase in IL6 production by young BMDM treated with IDO1 inhibitors BMS-986205, 1-methyl-dl-tryptophan, NLG919 and IDO-1 inhibitor (supplementary figure S2a). In response to poly I:C stimulation (24 h), there was a significant increase in IL6 production by young BMDM treated with the IDO1 inhibitors BMS-986205, NLG919 and IDO-1 inhibitor (supplementary figure S2b). Based on these results, we chose to further examine the impact of IDO1 inhibition using the irreversible inhibitor BMS-986205. Treatment of young BMDM with BMS-986205 resulted in a significant decrease in IDO1 protein expression and activity, which corresponded with increased IL6 production (supplementary figure S2c–e). Daily in vivo treatment of young adult mice with BMS-986205 resulted in increased cellular infiltration and marked changes in the lung that corresponded with significant weight loss and increased viral titres (figure 5a–d). In response to BMS-986205, there was a significant increase in cells present in the BAL that corresponded with decreased IDO1 activity in lung and alveolar macrophage populations and significantly augmented levels of pro-inflammatory cytokines, such as IL6, and decreased production of pro-resolution cytokines, such as IL10 (figure 5e–i).
Age-associated mitochondrial dysfunction contributes to changes in IDO1 expression and KP-mediated metabolism of tryptophan
As shown in figure 6a, when compared to young lung, there was reduced Ido1 mRNA expression in aged lung at baseline and during the course of influenza infection. Interestingly, there was augmented expression of other metabolic pathways, such as betaine-homocysteine methyltransferase (Bhmt), which catalyses the conversion of homocysteine to methionine, and succinyl-CoA:3-ketoacid coenzyme A transferase 2A (Oxct2a), a key enzyme for ketone body catabolism, in aged lung by day 7 post-influenza (figure 6a, supplementary table S1). When compared to young lung, there was also diminished Ido1 mRNA expression in aged alveolar macrophages during the course of influenza infection (figure 6b, supplementary table S1). Decreased kynurenine 3-monooxygenase (Kmo) and nitric oxide synthase 2 (Nos2) mRNA expression, which was associated with augmented Oxct2a mRNA expression, was observed in aged alveolar macrophages by day 7 of influenza infection (figure 6b, supplementary table S1). As mitochondria play an intricate role in modulating metabolic processes, we next examined how different components of the mitochondria might respond to energy demands of the host during influenza infection. During influenza infection, there were distinct changes in gene regulation in both young and aged lung, with decreased gene expression in aged lung by day 7 of infection (figure 6c, supplementary table S2). Recent work has illustrated that kynurenine can bind to aryl hydrocarbon receptor (AhR) and result in heightened kynurenine activity [21]. Interestingly, in aged lung by day 7 of influenza, there was upregulated expression of AhR-interacting protein (Aip), a protein shown to stabilise and enhance AhR function, which corresponded with declining levels of kynurenine [22] (figure 6c, supplementary table S2).
Given these findings, we evaluated the impact of mitochondrial targeted antioxidants, mitoTEMPO-L, Trolox, and mitoquinol on IL6 production by LPS-stimulated BMDM. There was a significant reduction of IL6 at both 4 and 24 h in aged, mitoquinol-treated BMDM (supplementary figure S3a). Furthermore, in response to treatment of aged BMDM with mitoquinol, there was a significant increase in IDO1 expression, increased ratio of kynurenine to tryptophan post-stimulation with LPS and poly I:C, and a corresponding decrease in IL6 production (supplementary figure S3b–d). Based on these findings, we treated aged mice with daily injections of mitoquinol (50 μM·mouse·day−1) starting at a time point when clinical manifestations of influenza were detectable in aged mice (day 3 post-infection) (supplementary figure S4a). In response to mitoquinol, there was a marked decrease in weight loss and decreased viral titres (supplementary figure S4b, c). Furthermore, mitoquinol treatment dramatically impacted mitochondrial gene expression in aged lung on day 7 post-influenza (figure 6c, supplementary table S3). Specifically, mitoquinol treatment resulted in decreased expression of Aip (figure 6c, supplementary table S3). Superoxide formation in alveolar macrophages isolated from lung at select time points during influenza was also decreased in response to mitoquinol treatment (supplementary figure S4d). In response to daily treatment with mitoquinol there was a significant decrease in Bhmt and Oxct2a mRNA that was associated with increased IDO1 mRNA expression in aged lung by day 7 post-influenza (figure 6d, supplementary table S2). Examination of metabolic gene expression in aged alveolar macrophages at days 5 and 7 post-influenza infection, illustrated a decrease in Oxct2a mRNA and a corresponding increase in Ido1 and Kmo mRNA expression (figure 6e, supplementary table S2).
In response to different doses of mitoquinol, we detected increased IDO1 specific activity, which corresponded with significantly heightened kynurenine and reduced tryptophan production (figure 7a–c). Daily treatment with mitoquinol decreased cellular infiltration and inflammation in the aged lung (figure 7d, e). At day 7 post-influenza, there was a significant reduction of IL6 expression in BAL and lung homogenates collected from mitoquinol-treated mice (figure 7f). Furthermore, in response to daily treatment with mitoquinol, there was a dose-dependent increase in the number of CD4+CD25+ T-regulatory cells, which corresponded with increasing levels of IL10 production in aged lung by day 7 post-influenza (figure 7g, h).
Discussion
In the current study, we investigated the impact of chronological ageing on tryptophan metabolism in response to influenza infection. Examination of metabolites present in human plasma identified tryptophan metabolism as an important metabolic pathway utilised in response to influenza. Our results expand upon these findings and illustrate that decreased production of kynurenine in aged lung in response to influenza was associated with increased inflammatory and diminished regulatory responses. Reduced IDO1 activity in aged lung and alveolar macrophages corresponded with changes in metabolic gene regulation and the initiation of alternative tryptophan metabolism pathways. Our findings also demonstrated that decreased IDO1 expression was due to dysregulated mitochondrial gene expression and heightened generation of reactive oxygen species (ROS), as treatment with mitoquinol improved mitochondrial function and restored KP-mediated tryptophan metabolism. Taken together, our data provide additional evidence as to why older persons are more susceptible to influenza.
Using a murine model of infection, there was an increase in kynurenine in both young and aged lung starting at day 3 and continued to increase by day 5. These findings are in agreement with previous influenza studies investigating the metabolic responses in young adult mice [23]. However, by day 7, kynurenine levels were only heightened in young lung, with levels remaining unchanged in the aged lung. While kynurenine was produced by an inflammatory process, it has an anti-inflammatory function, often serving as a brake on the immune response. Downstream metabolites, such as quinolinate, play an important role in replenishing NAD+ levels to meet host energy demands in response to heightened cellular stress during influenza infection. Decreased KP-mediated tryptophan metabolism and production of quinolinate in aged lung corresponded with impaired NAD+ replenishment and an inability to meet host energy demands. As tryptophan metabolism is a negative feedback metabolism, an inefficacy of aged alveolar macrophages to utilise this metabolic pathway would result in decreased NAD+ replenishment and increased inflammation. Interestingly, kynurenine levels appeared to be higher in the older animals at baseline, but increased more dramatically in the young animals following infection. Previous work has illustrated that chronic low-grade inflammation can result in elevated levels of circulating kynurenine [24]. Therefore, it may be possible that the increased presence of inflammatory cytokines in aged hosts may have a significant impact on baseline kynurenine levels and resulting in changes in tryptophan metabolism in response to pathogenic stimuli.
During the course of influenza infection, there was a metabolic shift in aged lung and alveolar macrophages. When compared to young lung, there was an increase in Oxct2a, a key enzyme for ketone body catabolism, which corresponded with diminished mitochondrial gene expression. Our results illustrate that treatment with a mitochondrial targeted antioxidant improved mitochondrial gene regulation and diminished Oxct2a expression in aged lung and alveolar macrophages. Based on these findings, during the course of influenza, when there is decreased glycolytic ATP production, to meet host energy demands, there may be a change in metabolic gene expression, which may reflect increased ketone body catabolism. In the context of heightened cellular stress and energy deficiency that occurs in response to influenza, macrophage utilisation of ketone bodies may help to regulate tricarboxylic acid cycle flux, modulate pyruvate-derived gluconeogenesis and serve as an alternative source of ATP [25]. Metabolic reprogramming of macrophages in response to influenza highlights an age-associated impact on the phenotypic characteristics that may not only impact the antimicrobial properties, but also influence dysregulated tissue repair and remodelling.
Our current results illustrate that daily administration of mitoquinol to aged adult mice improves IDO1-mediated metabolism of tryptophan to kynurenine, thereby improving host innate immune responses and decreasing morbidity and mortality during the course of influenza. It is well appreciated that ROS signalling plays a key role in the initiation of innate immune responses during influenza; however, overly heightened ROS production can result in excessive immune responses and have a deleterious impact on host tissue systems. Designed as a lipophilic molecule bearing a cation moiety, mitoquinol can pass directly through the mitochondrial membrane to increase the mitochondrial antioxidant capacity and decrease mitochondrial oxidative damage [26]. Furthermore, mitoquinol can function as an antioxidant to prevent lipid peroxidation-induced apoptosis and protect mitochondria from oxidative damage [27]. Previously, mitoquinol has been shown to be efficacious in reducing mitochondrial oxidative damage in multiple diseases, such as sepsis, fatty liver disease and Alzheimer's disease [27, 28]. Future work will need to be performed to understand the exact mechanistic pathways that are altered in response to mitoquinol treatment and if therapeutic administration of mitoquinol in additional pulmonary viral infection models will also prove to be efficacious.
It is important to note that IDO1 expression by epithelial cells and fibroblasts can also influence regulatory responses at the site of inflammation [29, 30]. Recent work has illustrated that expression of IDO in lung parenchyma can inhibit acute lethal pulmonary inflammation [31]. Specifically, the production of kynurenine by lung epithelial cells as well as alveolar macrophages can suppress inflammatory activities within the lung [31]. In agreement with these findings, when young adult mice were treated with IDO1 inhibitor BMS-986205, there was a significant increase in inflammation and heightened lung injury during influenza infection. It is plausible that in the absence of IDO1 activity and diminished kynurenine production, there is an inability of epithelial and alveolar macrophages to inhibit inflammation, resulting in increased morbidity and mortality during influenza. Previous work has illustrated that IDO1-deficient and 1-methyltryptophan-treated mice are protected from morbidity during influenza A infection [4]. It is important to note that these studies examined the impact of IDO1 deficiency in young mice and it is possible that compensatory pathways may contribute to protection during influenza. In addition, our data in young adult mice using BMS-986205 illustrated an increase in morbidity during infection. We believe that this is due to the potent and selective IDO1 inhibitory properties of this compound when compared to 1-methyltryptophan, which has been shown to not induce effective in vivo IDO inhibition due to decreased potency and similar plasma concentrations of tryptophan post-treatment [32]. While our current work focused on alveolar macrophages, future studies will need to investigate the contribution of epithelial cell production of kynurenine during influenza and the impact of ageing on these responses.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary methods. ERJ-00443-2020.Supplement
Supplementary table S1 (related to Figure 6). Metabolic gene expression in young and aged murine lung and alveolar macrophages during influenza infection. ERJ-00443-2020.Table_S1
Supplementary table S2 (related to Figure 6). Mitochondrial gene expression in young and aged murine lung during influenza infection. ERJ-00443-2020.Table_S2
Supplementary figures. ERJ-00443-2020.Figures
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-00443-2020.Shareable
Footnotes
This article has supplementary material available from erj.ersjournals.com
Conflict of interest: S.J. Cho has nothing to disclose.
Conflict of interest: K.S. Hong has nothing to disclose.
Conflict of interest: E. Schenck has nothing to disclose.
Conflict of interest: S. Lee has nothing to disclose.
Conflict of interest: R. Harris has nothing to disclose.
Conflict of interest: J. Yang has nothing to disclose.
Conflict of interest: A.M.K. Choi has nothing to disclose.
Conflict of interest: H. Stout-Delgado has nothing to disclose.
Support statement: This work was supported by the National Heart, Lung, and Blood Institute (grant K08HL138285) and the National Institute on Aging (grants 5R01AG052530, 5R01AG056699). Funding information for this article has been deposited with the Crossref Funder Registry.
- Received February 26, 2020.
- Accepted November 8, 2020.
- Copyright ©ERS 2021. For reproduction rights and permissions contact permissions{at}ersnet.org