Decision letter: De novo synthesized polyunsaturated fatty acids operate as both host immunomodulators and nutrients for Mycobacterium tuberculosis

Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Successful control of Mycobacterium tuberculosis (Mtb) infection by macrophages relies on immunometabolic reprogramming, where the role of fatty acids (FAs) remains poorly understood. Recent studies unraveled Mtb's capacity to acquire saturated and monounsaturated FAs via the Mce1 importer. However, upon activation, macrophages produce polyunsaturated fatty acids (PUFAs), mammal-specific FAs mediating the generation of immunomodulatory eicosanoids. Here, we asked how Mtb modulates de novo synthesis of PUFAs in primary mouse macrophages and whether this benefits host or pathogen. Quantitative lipidomics revealed that Mtb infection selectively activates the biosynthesis of ω6 PUFAs upstream of the eicosanoid precursor arachidonic acid (AA) via transcriptional activation of Fads2. Inhibiting FADS2 in infected macrophages impaired their inflammatory and antimicrobial responses but had no effect on Mtb growth in host cells nor mice. Using a click-chemistry approach, we found that Mtb efficiently imports ω6 PUFAs via Mce1 in axenic culture, including AA. Further, Mtb preferentially internalized AA over all other FAs within infected macrophages by mechanisms partially depending on Mce1 and supporting intracellular persistence. Notably, IFNγ repressed de novo synthesis of AA by infected mouse macrophages and restricted AA import by intracellular Mtb. Together, these findings identify AA as a major FA substrate for intracellular Mtb, whose mobilization by innate immune responses is opportunistically hijacked by the pathogen and downregulated by IFNγ. Editor's evaluation In this study, the authors highlight a role for de novo biosynthesis of Poly-unsaturated Fatty Acids and the consequence effect of these metabolites on the production of arachidonic acid. The increased bio-availability of arachidonic acid seemingly promotes mycobacterial growth whilst inhibition of arachidonic acid formation, and its resultant downstream eicosanoid products, affect macrophage function but somewhat surprisingly do not affect growth of M. tuberculosis in macrophages or in mice. The uptake of the different classes of fatty acids in axenic culture as well as in macrophages is explored and the authors demonstrate that the Mce1 transporter is largely responsible for their uptake during in vitro growth but only plays a partial role in their uptake during growth of the pathogen in host cells. This work will be of interest to bacteriologists and those studying infectious diseases. https://doi.org/10.7554/eLife.71946.sa0 Decision letter eLife's review process Introduction Mycobacterium tuberculosis (Mtb), the causative agent of human tuberculosis (TB), caused 1.6 million deaths in 2017, and it is estimated that 23% of the world's population has a latent TB infection. This success is due to Mtb evolving sophisticated strategies to survive intracellularly in macrophages, its preferred habitat, for long periods of time (Bussi and Gutierrez, 2019). In particular, Mtb's capacity to import and metabolize host-derived lipids, including fatty acids (FAs) and cholesterol, contributes to long-term persistence in vivo (Nazarova et al., 2017; Nazarova et al., 2019; Pandey and Sassetti, 2008). At the macrophage level, Mtb infection triggers the formation of lipid droplets (LDs) whose FA content was proposed to serve as a nutrient source for intracellular Mtb (Daniel et al., 2011; Peyron et al., 2008; Singh et al., 2012). However, this view was challenged by a recent study showing that the IFNγ cytokine promotes LD accumulation by Mtb-infected macrophages while impairing the pathogen's capacity to acquire host-derived FAs (Knight et al., 2018). Whether Mtb infection and IFNγ signaling differentially impact on subcellular localization and dynamic redistribution of host FAs is largely unknown. Mtb was shown to import fluorescently labeled saturated and monounsaturated fatty acids (SFAs and MUFAs, respectively) via a dedicated protein machinery named Mce1, which is coordinated with Mce4-mediated import of cholesterol and plays an important role in Mtb lipid homeostasis (Laval et al., 2021; Lee et al., 2013; Nazarova et al., 2017; Nazarova et al., 2019; Wilburn et al., 2018). In addition to SFAs and MUFAs, mammalian cells produce the additional subset of polyunsaturated fatty acids (PUFAs), whose secondary metabolites shape macrophage effector functions (Dennis and Norris, 2015; Mayer-Barber and Sher, 2015). In particular, the catabolism of arachidonic acid (AA) by cyclooxygenases (COX) and lipoxygenases (LOX) yields prostaglandins and lipoxins/leukotrienes, products referred to as eicosanoids that are important signaling molecules modulating inflammation and apoptosis. Interestingly, Toll-like receptors (TLRs) induce signal-specific reprogramming of FA synthesis in macrophages, with differential impact on antibacterial immunity (Hsieh et al., 2020). How Mtb-driven stimulation of TLRs reprograms PUFA and eicosanoid biosynthesis by host macrophages, and whether this promotes anti-mycobacterial immune responses remained to be determined. Here, we combined quantitative lipidomics with genetic ablation and pharmacological inhibition approaches to assess the importance of the PUFA biosynthetic pathway in innate control of Mtb infection by macrophages. While PUFA biosynthesis contributed to the generation of inflammatory and antimicrobial responses in infected macrophages, its stimulatory effect on macrophage effector functions was not matched by an enhanced capacity to restrict intracellular Mtb infection. This led us to propose that newly generated PUFAs may serve as FA sources for intracellular Mtb. Our in vitro and cellular assays using traceable alkyne-FAs supported this hypothesis by showing that Mtb efficiently internalizes ω6 PUFAs in axenic cultures, and preferentially the eicosanoid precursor AA within macrophages. They also indicated that in cellulo, Mce1 partially contributes to Mtb's uptake of AA and supports intracellular persistence of the pathogen at late stages of infection. Together, our findings reveal the pro-inflammatory function of the PUFA biosynthetic pathway during Mtb infection and identify AA as a major FA source for intracellular Mtb. They support the view that Mtb draws on intracellular free AA generated by infection. Results Mtb infection stimulates the biosynthesis of SFAs, MUFAs, and upstream PUFAs by host macrophages Mtb's impact on host FA metabolism was investigated by infecting bone marrow-derived macrophages (BMDMs), extracting total and free cellular FAs, and quantifying each FA species by gas chromatography, with normalization to total DNA. Mtb triggered a significant increase in intracellular levels of free SFA palmitic acid (PA) and MUFAs (oleic acid [OA] and vaccenic acid [VA]) after 24 hr (Figure 1A and B). With regard to PUFAs, we detected an increased level of the ω6 precursor linoleic acid (LA) that was associated with elevated levels of its conversion product dihomo-gamma-linolenic acid (DGLA), but not the DGLA product AA (Figure 1C and D). On the ω3 PUFA side, α-linolenic acid (ALA) and eicosapentaenoic acid (EPA) were below detection limit, and the low levels of long-chain ω3 PUFAs docosapentaenoic acid (DPA) and docosahexaenoic acid (DHA) were not modulated by Mtb infection (Figure 1C and D). Infection-driven variations in free FA levels were associated with parallel trends in total FA levels (Figure 1—figure supplement 1). These changes were not specific of the Mtb pathogen as BMDM infection with the Mycobacterium bovis BCG vaccine induced comparable FA profiles (Figure 1B–D, Figure 1—figure supplement 1), and they were no longer observed 48 hr post infection. When we stimulated BMDMs with TLR2/4 agonists, levels of free and total FAs were similarly modulated (Figure 1B–D, Figure 1—figure supplement 1), suggesting that Mtb-driven changes in FA levels in host macrophages result from recognition of mycobacteria pattern by TLR2/4. Figure 1 with 1 supplement see all Download asset Open asset Mtb infection upregulates intracellular levels of free SFAs, MUFAs, and upstream PUFAs in host macrophages. (A) Schematics of SFA and MUFA biosynthetic pathways. PA, palmitic acid; SA, stearic acid; OA, oleic acid; VA, vaccenic acid. (B) Intracellular levels of free SFAs and MUFAs in BMDMs infected with M. bovis BCG (BCG) or M. tuberculosis H37Rv (Mtb) at the same multiplicity of infection (MOI) of 2:1, or treated with LPS or Pam3Csk4 (Pam3), or left untreated (Ctrl) for 24 hr. FA levels were normalized to total DNA content and are shown as fold change relative to Ctrl. (C) Schematics of PUFA biosynthetic pathways. LA, linoleic acid; DGLA, dihomo-γ-LA; AA, arachidonic acid; ALA, α-linolenic acid; DPA, docosapentaenoic acid; DHA, docosahexaenoic acid. (D) Intracellular levels of free PUFAs in BMDMs treated as in (B). All data are means ± SD (n = 3) and are representative of two independent experiments. *p<0.05, p<0.01, *p<0.001, unpaired Student's t-tests. De novo synthesis of SFAs is controlled by the FASN rate-limiting enzyme, and MUFAs are generated from SFAs by the SCD2 rate-limiting enzyme (Figure 1A; Guillou et al., 2010). In Mtb-infected BMDMs, increased levels of SFAs were associated with upregulation of Fasn transcript levels from 6 hr post infection (Figure 2A). Scd2 mRNA expression was upregulated at 24 hr post infection with Mtb (Figure 2A), and the OA:SA ratio reflecting the efficacy of SFA conversion into MUFAs was increased upon Mtb and BCG infection (Figure 2B). Conversion of ω6 and ω3 PUFA precursors into downstream PUFAs is jointly controlled by three enzymes: the FA desaturases (FADS)1 and FADS2, and the elongase ELOVL5 (Figure 1C). Fads1 and Elovl5 transcript levels were transiently repressed at 6hr post infection with Mtb, while on the opposite those of Fads2 were increased from 6hr until 24hr (Figure 2C). The inverse regulation of Fads1/Elovl5, compared to Fasn/Fads2, at 6hr post infection with Mtb was surprising since all genes are targets of the LXR and SREBP1 regulators of FA metabolism (Daemen et al., 2013; Joseph et al., 2002). It is interesting to note that the transient downregulation of Fads1 and Elovl5 gene expression correlated with a drop in LXR activity, reflected by decreased transcript levels of LXR target gene Abca1 after 6 hr of Mtb infection, despite significant induction of Nr1h3 gene expression. The sustained upregulation of Fasn and Fads2 gene expression was associated with transcriptional induction of Srebf1 and its target gene Dhcr24 (Figure 2D). Together, data in Figures 1 and 2 indicate that Mtb infection upregulates the biosynthesis of SFAs, MUFAs, and upstream PUFAs in host macrophages through activation of TLR2/4. They suggest that FA production results from activation of SREBP1, and that PUFA biosynthesis blockade downstream of FADS1 is due to a transient, post-transcriptional repression of LXR activity. Figure 2 Download asset Open asset Mtb infection and IFNγ signaling cooperate to stop host PUFA biosynthesis. (A) Relative mRNA expression of SFA/MUFA biosynthetic enzymes in BMDMs primed with IFNγ before infection with Mtb for the indicated times, as determined by qRT-PCR. (B) SCD activity in BMDMs, as estimated by the ratio of oleic acid (OA) to stearic acid (SA) levels, after 24 hr of infection with Mtb or BCG. (C–D) Relative mRNA expression of biosynthetic enzymes (C) or LXR/SREBP1 target genes (D) in BMDMs treated as in (A), as determined by qRT-PCR. All data are means ± SD (n = 3) and are representative of two independent experiments. *p<0.05, p<0.01, *p<0.001, ****p<0.0001, two-way ANOVA with Dunnett post-hoc multiple comparison tests (A, C, D) and unpaired Student's t-tests (B). IFNγ shuts down the biosynthesis of all FAs in Mtb-infected macrophages Macrophage ability to mount efficient anti-Mtb responses relies on their activation by the Th1 cell-derived cytokine IFNγ, recently shown to limit host FA intake by intracellular Mtb (Knight et al., 2018). We thus sought to determine if IFNγ modulates FA biosynthesis by infected macrophages. Stimulating BMDMs with IFNγ prior to infection prevented Mtb-induced expression of Fasn and Scd2 (Figure 2A and C), suggesting that IFNγ limits de novo synthesis of SFAs and MUFAs by infected macrophages. The effect of IFNγ on PUFA biosynthetic enzymes was more complex as the cytokine mitigated both the inhibitory effect of Mtb infection on Elovl5 and Fads1 gene expression after 6 hr, and its stimulatory effect on Fads2 gene expression (Figure 2C). IFNγ-induced decrease in Fasn and Fads2 expression correlated with reduced Abca1 and Dhcr24 transcript levels after 6 hr of infection (Figure 2D), suggesting that IFNγ prevents Mtb-induced stimulation of FA biosynthesis through downregulation of LXR and SREBP1 activity. To assess the effect of such transcriptional changes on PUFA biosynthesis, we quantified macrophage's ability to convert a deuterated derivative of the ω6 precursor LA into downstream products upon infection with Mtb, with or without IFNγ priming, between 6 and 24 hr post infection. All ω6 intermediates (i.e., 20:2-d11, DGLA-d11, and AA-d11) were quantified, allowing us to measure the activity of each enzyme of the PUFA biosynthetic pathway via product to substrate ratios (Figure 3A). The measured percentages of each ω6 PUFA, relative to total deuterated FA, are also shown in Figure 3—figure supplement 1A. FADS2 activity was not modulated by Mtb in the conditions of the experiment, suggesting that infection-induced upregulation of Fads2 expression (Figure 2C) takes more than 24 hr to translate into enhanced enzyme activity. In contrast, decreased Fads1 expression at 6 hr post infection (Figure 2C) correlated with a marked reduction of FADS1-mediated conversion of DGLA in AA (Figure 3A), irrespective of IFNγ stimulation. Likewise, IFNγ-driven repression of Fads2 gene expression in Mtb-infected BMDMs (Figure 2C) resulted in potent suppression of FADS2 activity (Figure 3A). Neither Mtb infection nor IFNγ stimulation affected the activity of ELOVL5. Therefore, the partial upregulation of the PUFA biosynthetic pathway that we observed in Mtb-infected macrophages was abrogated by cell exposure to IFNγ. In all, these results indicated that IFNγ shuts down the biosynthesis of all FAs in Mtb-infected macrophages. Figure 3 with 1 supplement see all Download asset Open asset FADS2 inhibition impairs the effector functions of macrophages during Mtb infection. (A) Activities of PUFA biosynthetic enzymes in resting or IFNγ-primed BMDMs, either left untreated (Ctrl), or infected with Mtb and treated with a FADS2 inhibitor (iFADS2) or vehicle control, as determined by a conversion assay from 6 to 24 hr post infection using the ω6 precursor LA-d11. Enzyme activities were estimated with the ratio of deuterated fatty acid product to substrate levels and are shown as fold change relative to Ctrl. nd, product not detected; n/a, not applicable (substrate and product not detected). (B) Schematics of biosynthetic pathways of arachidonic acid (AA)-derived eicosanoids. COX, cyclooxygenase; LOX, lipoxygenase; PG, prostaglandin; TX, thromboxane; HETE, hydroxyeicosatetraenoic acid; LX, lipoxin. (C, D) Secreted levels of COX- (C) and LOX-derived (D) metabolites of AA by BMDMs either uninfected (Ctrl) or infected with Mtb, and treated with iFADS2 or vehicle control for 24 or 48 hr. Data in (A), (C), and (D) are means ± SD (n = 3), *p<0.05, p<0.01, *p<0.001, ****p<0.0001, unpaired Student's t-tests (A) and two-way ANOVA with Dunnett post-hoc multiple comparison tests (C, D). (E, F) Heatmap of mRNA expression levels of inflammatory (E) and antimicrobial (F) genes determined by NanoString analysis of BMDMs treated as in (C) for 6 or 24 hr. Shown are genes that were significantly downregulated by iFADS2 treatment at 6 and/or 24 hr (fold change of at least 1.15 and FDR < 0.05, two-way ANOVA with Benjamini−Hochberg adjustment for multiple comparison). Source data are available in Figure 3—source data 1. Figure 3—source data 1 Complete list of normalized mRNA levels in BMDMs, either noninfected (NI Ctrl) or infected with Mtb, and treated with iFADS2 (Mtb iFADS2) or vehicle control (Mtb Veh), as determined by NanoString analysis. https://cdn.elifesciences.org/articles/71946/elife-71946-fig3-data1-v2.xlsx Download elife-71946-fig3-data1-v2.xlsx FADS2 inhibition impairs the effector functions of macrophages during Mtb infection Long-chain PUFAs can be mobilized by hydrolysis of phospholipids to fuel the production of lipid mediators of inflammation (Dennis and Norris, 2015). In particular, conversion of AA by the COX/LOX pathways generates eicosanoids (Figure 3B), modulating the ability of macrophages to control mycobacterial infection (Mayer-Barber and Sher, 2015). Although transcriptionally repressed, significant de novo synthesis of AA was maintained in Mtb-infected BMDMs (Figure 3—figure supplement 1A), suggesting a role in generation of eicosanoids. To test this, we used a selective inhibitor of FADS2 (SC-26196, hereafter named iFADS2) (Obukowicz et al., 1999). Exposing BMDMs to iFADS2 efficiently abrogated FADS2 activity in Mtb-infected, resting, and IFNγ-activated macrophages (Figure 3A), validating our experimental conditions. Consistent with previous studies (Chen et al., 2008; Knight et al., 2018), BMDMs infected with Mtb upregulated Ptgs2 expression (Figure 3—figure supplement 1B) and secreted higher amounts of AA metabolites deriving from both the COX (Figure 3C) and LOX pathways (Figure 3D) compared to noninfected controls. When BMDMs were infected with Mtb in the presence of iFADS2, the production of all COX/LOX-derived AA metabolites was significantly reduced (Figure 3C and D), suggesting that part of the infection-induced eicosanoids may originate from de novo synthesized PUFAs. Of note, upregulation of Ptgs2 expression and production of PGE2 were both potentiated by BMDM exposure to IFNγ prior to infection with Mtb, and the inhibitory effect of iFADS2 on Mtb-driven production of PGE2 production was maintained in IFNγ-activated macrophages (Figure 3—figure supplement 1B and C). PGE2 and LXA4 production by infected macrophages differentially influence the outcome of Mtb infection, promoting anti- or pro-mycobacterial responses via the induction of apoptotic or necrotic cell death, respectively (Chen et al., 2008; Divangahi et al., 2010; Mayer-Barber and Sher, 2015). Since iFADS2 treatment decreased Mtb-induced production of both PGE2 and LXA4 (Figure 3C and D), we tested how FADS2 inhibition affects the relative induction of apoptosis and necrosis in infected BMDMs (Figure 3—figure supplement 1D). Cell apoptosis and expression of Syt7, which is involved in lysosome-mediated membrane repair, were both downregulated by iFADS2 in Mtb-infected BMDMs, while necrosis levels remained unchanged (Figure 3—figure supplement 1D and E). We concluded that iFADS2-induced alterations of the PGE2/LXA4 balance results in a modest impairment of macrophage membrane repair and apoptosis during Mtb infection. To determine if FADS2 inhibition alters the innate immune functions of macrophages, we profiled the expression of a panel of genes involved in antimicrobial and inflammatory responses using a custom NanoString nCounter CodeSet (Supplementary file 1). BMDMs were infected with Mtb and treated or not with iFADS2, and gene expression was assessed at 6 and 24 hr post infection. We detected a decrease in Ptgs2 gene expression in iFADS2-treated BMDMs infected with Mtb, which may account for the observed decrease in PGE2 production (Figure 3C). Besides, FADS2 inhibition induced a significant downregulation of major inflammatory genes (Tnf, Il1b, Il6) (Figure 3E) and genes involved in antimicrobial responses of macrophages (Nos2, Nox1, Irg1, Sod2) (Figure 3F). Overall, our analyses of FADS2-inhibited macrophages indicated that the PUFA biosynthetic pathway promotes antimicrobial and inflammatory responses. Inhibiting FADS2 does not impact Mtb growth in vivo Our data in Figure 3 predicted that FADS2 inhibition should limit the macrophage capacity to restrict the intracellular growth of Mtb. To test this hypothesis, we infected resting or IFNγ-primed BMDMs with Mtb in the presence or absence of iFADS2 and monitored mycobacterial growth during 6 days by titrating colony-forming units (CFUs) in cell lysates. Pharmacological inhibition of FADS2 had no detectable effect on intracellular growth of Mtb (Figure 4A), neither in resting nor IFNγ-stimulated BMDMs. Similar results were obtained in BMDMs where Fads2 expression was silenced by siRNA-mediated knock-down (Figure 4—figure supplement 1A and B). To determine if a complete defect in FADS2 expression would impact Mtb intracellular we FADS2 in the human cell using the We three independent in 2 of FADS2 gene and FADS2 protein and a defect in PUFA conversion (Figure 4—figure supplement 1C and D). In with our data using iFADS2 and Mtb similarly in and (Figure Figure with 1 supplement see all Download asset Open asset Inhibiting FADS2 does not impact Mtb growth in macrophages nor mice. (A) Intracellular growth of Mtb resting or IFNγ-primed BMDMs treated with iFADS2 or with vehicle control, as determined by colony-forming at the indicated days post infection. Data are means ± SD (n = 3) and are representative of two independent experiments. *p<0.05, ***p<0.001, two-way ANOVA with post-hoc multiple comparison (B) Intracellular growth of Mtb or FADS2 as determined by at the indicated days post infection. (n = 3) in independent Data are representative of two independent experiments. (C, D) of Mtb in the (C) and (D) of treated with iFADS2 or with vehicle control during as determined by Data shown are means ± of two independent = = time (E) Relative mRNA expression of inflammatory and antimicrobial genes in the of treated as in (C), as determined by qRT-PCR. Data shown are means ± = or *p<0.05, in a two-way ANOVA with post-hoc multiple comparison Since lipid mediators and inflammatory genes modulated by iFADS2 are involved in immunity Mtb (Mayer-Barber and Sher, we investigated the effect of a inhibition of FADS2 in a mouse of infection with Mtb. treatment with iFADS2 not significantly Mtb growth in and in the conditions tested (Figure and D). However, iFADS2 treatment the transcriptional induction of inflammatory and antimicrobial genes in of Mtb-infected (Figure In with our data obtained in macrophages, these results indicated that FADS2 promotes the generation of anti-mycobacterial responses. PUFA biosynthesis in infected was not to Mtb Mtb efficiently imports ω6 PUFAs through the Mce1 transporter in axenic culture We that the anti-mycobacterial of PUFA biosynthesis on macrophage effector functions be by a pro-mycobacterial role of PUFAs as nutrient sources for Mtb. SFAs and MUFAs and shown to be and by Mtb in axenic and within macrophages (Nazarova et al., 2017; Nazarova et al., 2019). To determine if Mtb has the ability to import PUFAs, we used of FAs that can be detected by click-chemistry using et al., Figure Using this approach, we that PA, and to a are efficiently internalized by Mtb (Figure While ω3 PUFAs and uptake was we found that ω6 PUFAs and were by Mtb as efficiently as Figure with 1 supplement see all Download asset Open asset Mtb efficiently imports ω6 PUFAs through the Mce1 transporter in axenic (A) Schematics of the click-chemistry used to the uptake of SFAs, MUFAs and PUFAs by Mtb in axenic (B) of and uptake at a of by Mtb, as estimated by of the of Data shown are representative of three independent experiments. (C) Schematics of the of the and genes in Mtb (D) of by different Mtb for the expression of genes to relative to Mtb. Data are means ± SD (n = 3) and are representative of two independent **p<0.01, unpaired Student's t-tests. (E) of alkyne-FAs by Mtb and its relative to Data are means ± SD from three independent *p<0.05, p<0.01, *p<0.001, t-tests. (F) Relative uptake of and by Mtb in the presence of amounts of palmitic acid oleic acid linoleic acid or arachidonic acid Data shown are representative of at least two independent experiments. We assessed the role of Mce1 as of PUFAs in Mtb by in genes of by (Figure Figure supplement In with previous studies (Nazarova et al., 2017; Nazarova et al., or but not or resulted in in and import shown for in Figure Likewise, we found that Mtb's import of and was abrogated in the (Figure supplement 1B) and (Figure and that FA import was in the (Figure This Mtb's ability to import PUFAs via the Mce1 Since ω6 PUFAs and were efficiently by Mtb via Mce1, we tested if they with other FAs for and measured Mtb's uptake of alkyne-FAs as in the presence of amounts of of and was decreased by addition of their validating the conditions of the assay (Figure Interestingly, LA and AA with each other and with OA, but not with (Figure Together, data in Figure indicated that SFAs, MUFAs, and ω6 PUFAs are by Mtb via Mce1 in axenic culture, with They that ω6 PUFAs with OA, but not PA, for uptake by Mtb preferentially internalizes AA in the of macrophages We investigated if all FAs had comparable ability to to Mtb within infected macrophages. Here, BMDMs were infected with a of Mtb prior to a with of of PA, OA, LA, AA, or (Figure supplement Mtb was from infected and FAs by Mtb were by and quantified by (Figure supplement All alkyne-FAs be detected in intracellular Mtb after 24 hr, and at levels after hr (Figure from in vitro Mtb (Figure intracellular Mtb a marked for AA over all other FAs including (Figure This was by analysis of in Mtb-infected cells (Figure and of and in Mtb (Figure Figure 6 with 2 see all Download asset Open asset Mtb preferentially internalizes AA in the of macrophages. (A) uptake of alkyne-FAs by Mtb from BMDMs infected for 24 or hr, as measured by Data are means ± SD from two or three independent *p<0.05, t-tests. (B) of in Mtb-infected BMDMs at 24 hr post infection, as shown on representative = Mtb, = the relative of = = from to 1). = of the in the = 1 (C, D) of the in intracellular detected on (C) and in BMDMs either noninfected or Mtb-infected (D) using of BMDMs infected for 24 hr with a of Mtb. means ± for (C) and for not ****p<0.0001, unpaired Student's (C) and ANOVA with post-hoc multiple comparison tests (E) of alkyne-FAs by Mtb and its from BMDMs infected for 24 hr, relative to Mtb as by