To apply a holistic approach to assessing the role(s) of fatty acid metabolism on the intracellular growth of Mtb, we used a CRISPR genetic approach to knockout genes involved in lipid import (CD36, SLC27A1), lipid droplet formation (PLIN2), and fatty acid oxidation (CPT1A, CPT2) in murine primary macrophages (Figure 1A). Deletion of CD36 or CPT2 in mouse macrophages has been shown to impair intracellular growth of Mtb (Hawkes et al., 2010; Chandra et al., 2020). But the role of specialized long-chain fatty acid transporters (SLC27A1-6) on Mtb growth in macrophages is uncharacterized. SLC27A1 and SLC27A4 are the most abundant fatty acid transporter isoforms in macrophages (Nishiyama et al., 2018). PLIN2, or adipophilin, is known to be required for lipid droplet formation (Paul et al., 2008; Larigauderie et al., 2004). Five isoforms of mammalian perilipins (PLIN) are involved in lipid droplet biogenesis among which PLIN2 is the dominant isoform expressed in macrophages (Knight et al., 2018). However, macrophages derived from Plin2^-/- mice show no defects in the production of lipid droplets nor do they impair intracellular growth of Mtb (Knight et al., 2018). We targeted each of these genes with at least two sgRNAs in Hoxb8 Cas9⁺ conditionally immortalized myeloid progenitors (Kiritsy et al., 2021) to generate a panel of mutants that were deficient in the following candidates of interest; Slc27a1^-/-plin2^-/-, Cd36^-/-, Cpt1a^-/-, and Cpt2^-/-. Each individual sgRNA achieved >85% CRISPR-mediated deletion efficiency for all the five genes as analyzed by the Inference for CRISPR Edits (ICE) tool (Conant et al., 2022; Supplementary file 1). We verified the protein knockout phenotypes by flow cytometry and western blot analysis of differentiated macrophages derived from the CRISPR-deleted Hoxb8 myeloid precursors (Figure 1—figure supplement 1).

To confirm certain knockout phenotypes functionally, we checked lipid droplet biogenesis in Plin2^-/- macrophages in comparison to macrophages transduced with a non-targeting scramble sgRNA by confocal microscopy of BODIPY-stained cells. Cells were cultured for 24 hours in the presence of exogenous oleate to enhance the formation of lipid droplets (Listenberger and Brown, 2007). We observed a complete absence of lipid droplet formation in Plin2^-/- macrophages compared to controls (Figure 1—figure supplement 2A). This is contrary to previous observations in macrophages derived from PLIN2 knockout mice, which were reported to have no defect in lipid droplet formation (Knight et al., 2018). We also assessed the ability of Cpt2^-/- macrophages to oxidize fatty acids using the Agilent Seahorse XF Palmitate Oxidation Stress Test. Scrambled sgRNA and Cpt2^-/- macrophages were cultured in substrate-limiting conditions and supplied with either bovine serum albumin (BSA) or BSA-conjugated palmitate. As shown in Figure 1—figure supplement 2B, control macrophages were able to utilize and oxidize palmitate in substrate-limiting conditions indicated by a significant increase in oxygen consumption rates (OCRs) in contrast to cells supplied with BSA alone. Addition of the CPT1A inhibitor, etomoxir, inhibited the cell’s ability to use palmitate in these conditions (Figure 1—figure supplement 2B). CPT2 knockout in Cpt2^-/- macrophages impaired the cell’s ability to oxidize palmitate to a degree comparable to etomoxir treatment as evidenced by baseline OCRs compared to scrambled sgRNA control (Figure 1—figure supplement 2C).

We then assessed the different knockout mutant macrophages in their ability to support the intracellular growth of Mtb. We infected macrophages with Mtb Erdman at a multiplicity of infection (MOI) of 0.4 and assessed intracellular bacterial growth rates by counting colony-forming units (CFUs) 5 days post infection. All the five mutant macrophages significantly impaired Mtb growth rates compared to scrambled sgRNA as assessed by CFUs counts on day 5 (Figure 1B). Plin2^-/-, Slc27a1^-/-, and Cpt2^-/- macrophages displayed the strongest growth restriction phenotypes while Cd36^-/-macrophages had a moderate, but significant, impact on Mtb growth. In parallel, we quantified intracellular bacteria on day 0, 3 hours post infection (Figure 1C), to ascertain that subsequent differences on day 5 were not due to disparities in initial bacterial uptake. The moderate growth restriction phenotypes of Cd36^-/- macrophages were consistent with previous findings, which reported a similar impact on Mtb and M. marinum growth in macrophages derived from Cd36^-/- mice (Hawkes et al., 2010). Impaired growth of Mtb in Cpt1a^-/- and Cpt2^-/- macrophages is also consistent with previous reports that genetic and chemical inhibition of fatty acid oxidation is detrimental to the growth of Mtb within macrophages (Huang et al., 2018; Chandra et al., 2020).

Impairment of fatty acids metabolism by SLC27A1 knockout in macrophages rewires their substrate bias from fatty acids to glucose (Johnson et al., 2016). We reasoned that deletion of genes required for downstream processing of lipids (Figure 1A) could also reprogram macrophages and increase glycolysis, which could, in part, explain bacterial growth restriction. We analyzed the metabolic states of three knockout macrophages (Slc27a1^-/-, Plin2^-/-, and Cpt2^-/-) in uninfected and Mtb-infected conditions by monitoring OCRs and extracellular acidification rates (ECARs) using the Agilent Mito and Glucose Stress Test kits. All three mutant uninfected macrophages displayed reduced mitochondrial respiration as evidenced by lower basal and spare respiratory capacity (SRC) compared to scrambled sgRNA controls (Figure 2—figure supplement 1A and B). Mtb infection proportionally reduced basal and SRC rates across all the mutant macrophages and scrambled controls (Figure 2A and B) compared to uninfected macrophages, which is consistent with previous findings (Cumming et al., 2018). Plin2^-/- macrophages displayed the most marked reduction in mitochondrial activity in both uninfected and infected conditions, while Slc27a1^-/- macrophages were the least affected (Figure 2A, Figure 2—figure supplement 1A). As reported previously (Johnson et al., 2016), uninfected Slc27a1^-/- macrophages were more glycolytically active with higher basal and spare glycolytic capacity (SGC) compared to scrambled controls (Figure 2—figure supplement 1C and D). Uninfected Plin2^-/- and Cpt2^-/- macrophages were also more glycolytically active, but to a greater degree than Slc27a1^-/- (Figure 2—figure supplement 1C and D). Mtb infection increased the glycolytic capacity of all the three mutant macrophages (Figure 2C and D). Overall, Plin2^-/- macrophages exhibited the highest glycolytic capacity (Figure 2C, Figure 2—figure supplement 1D). Our data indicates that impairment of fatty acid metabolism at different steps significantly impacts mitochondrial respiration and reprograms cells toward glycolysis. Increased glycolytic flux in macrophages has been linked to the control of intracellular Mtb growth (Gleeson et al., 2016; Shi et al., 2015). Metabolic realignment as a consequence of interference with lipid homeostasis, which results in enhanced glycolysis may contribute to Mtb growth restriction in these mutant macrophages.

Mtb infection is known to induce increased glycolysis or the ‘Warburg effect’ in macrophages, mouse lungs, and human TB granulomas (Shi et al., 2015; Gleeson et al., 2016; Belton et al., 2016). Several studies have demonstrated that the Warburg effect is mediated by the master transcription factor hypoxia-inducible factor 1 (HIF1) (Courtnay et al., 2015). During Mtb infection, HIF1 is activated by the production of ROS, tricarboxylic cycle (TCA) intermediates and hypoxia in the cellular microenvironments as a consequence of altered metabolic activities and increased immune cell functions (Li et al., 2024; Shi et al., 2015; Gleeson et al., 2016; Belton et al., 2016). We assessed HIF1 stability in the three mutant macrophage lineages (Slc27a1^-/-, Plin2^-/-, and Cpt2^-/-) by monitoring total HIF1α protein levels by western blot, having confirmed that they were all more glycolytically active than the scrambled controls (Figure 2, Figure 2—figure supplement 1). Indeed, all the three mutant macrophages displayed significantly higher amounts of total HIF1α compared to scrambled controls after Mtb infection for 48 hours (Figure 3A, Figure 3—figure supplement 1A and B). Slc27a1^-/-, Plin2^-/- macrophages had increased levels of total HIF1α even in uninfected states (Figure 3A and B, Figure 3—figure supplement 1A and B). We also checked the phosphorylation status of the adenosine monophosphate kinase (AMPK), a master regulator of cell energy homeostasis (Garcia and Shaw, 2017), in the mutant macrophages since Seahorse flux analyses indicated that they had impaired mitochondrial activities (Figure 2, Figure 2—figure supplement 1). Western blot analysis revealed that in both Mtb-infected and uninfected conditions, impaired fatty acid metabolism in the mutant macrophages correlated with increased activation of AMPK as indicated by higher levels of phosphorylated AMPK compared to scramble (Figure 3C and D, Figure 3—figure supplement 1C and D). Interestingly, total AMPK was also increased, at least in Slc27a1^-/-and Plin2^-/- macrophages, in both uninfected and Mtb-infected conditions (Figure 3C and E, Figure 3—figure supplement 1C and D). These data point to a metabolic reprogramming of cells through activation of HIF1α and AMPK to promote glycolysis. In energetically stressed cellular environments, activated AMPK promotes catabolic processes such as autophagy to maintain nutrient supply and energy homeostasis (Garcia and Shaw, 2017). Autophagy is also an innate immune defense mechanism against intracellular Mtb in macrophages (Gutierrez et al., 2004). We examined the levels of autophagic flux in the mutant macrophages by monitoring LC3I to LC3II conversion by western blot and by qPCR analysis of selected autophagy genes (AMBRA1, ATG7, MAP1LC3B and ULK1). We observed an increase in autophagic flux by western blot analysis of LC3II/LC3I ratios in Mtb-infected Slc27a1^-/- and Cpt2^-/- macrophages (Figure 3—figure supplement 2A–C). Cpt2^-/- macrophages were more autophagic even in uninfected conditions (Figure 3—figure supplement 2A–C). Meanwhile, our qPCR analysis also revealed that the four autophagy genes were upregulated in both Mtb-infected and uninfected conditions in all the three mutants (Figure 3—figure supplement 2D and E). These data suggest that impaired fatty acid import and metabolism in macrophages could be restricting Mtb growth by promoting autophagy. These data agree with previous observations that inhibition of fatty acid oxidation enhances macrophage xenophagic activity, which leads to improved control of Mtb (Chandra et al., 2020).

The mycobacterial isocitrate lyase (icl1) acts as an isocitrate lyase in the glyoxylate shunt and as a methyl-isocitrate lyase in the methyl-citrate cycle (MCC) (Gould et al., 2006; McKinney et al., 2000). Mtb uses the MCC to convert propionyl CoA originating from the breakdown of cholesterol rings and β-oxidation of odd chain fatty acids into succinate and pyruvate, which are eventually assimilated into the tricarboxylic cycle (TCA) (Muñoz-Elías et al., 2006; Griffin et al., 2012). The buildup of propionyl CoA is toxic to Mtb and the bacteria relies on the MCC together with the incorporation of propionyl CoA to methyl-branched lipids in the cell wall as an internal detoxification system (Muñoz-Elías et al., 2006; Savvi et al., 2008). Mtb propionyl CoA toxicity is, in part, due to a cellular imbalance between propionyl CoA and acetyl CoA as an accumulation of the former or paucity of the latter results in the propionyl CoA-mediated inhibition of pyruvate dehydrogenase (Lee et al., 2013). Consequently, Mtb icl1-deficient mutants (Mtb Δicl1) are unable to grow in media supplemented with cholesterol or propionate, or intracellularly in macrophages (Lee et al., 2013). However, this growth inhibition could be rescued both in culture and in macrophages by exogenous supply of acetate or even chain fatty acids, which can be oxidized to acetyl-CoA (Lee et al., 2013). We took advantage of this metabolic knowledge to assess whether exogenous addition of the even chain fatty acid oleate can rescue the intracellular growth of Mtb Δicl1 mutants in our CRISPR knockout macrophages. Scrambled controls, Slc27a1^-/-, Plin2^-/-, and Cpt2^-/- macrophages in normal macrophage media or media supplemented with oleate were infected with an Mtb Δicl1 strain expressing mCherry at MOI 5. Bacterial growth measured by mCherry expression was recorded 5 days post infection. Consistent with previous observations (Lee et al., 2013), the Mtb Δicl1 mutant failed to replicate in both mutant and scramble macrophages that were grown in normal macrophage media as evidenced by baseline mCherry fluorescence (Figure 4A). Oleate supplementation successfully rescued the Mtb Δicl1 mutant in scrambled control macrophages. However, the growth restriction of the Mtb Δicl1 strain could not be alleviated by exogenous addition of oleate to the mutant macrophages (Slc27a1^-/-, Plin2^-/-, and Cpt2^-/-) (Figure 4A). These data suggest that impaired import (Slc27a1^-/-), sequestration (Plin2^-/-), or β-oxidation of fatty acids (Cpt2^-/-) blocks Mtb’s ability to access and use cellular lipids.

Oleate supplementation in macrophages induces the formation of lipid droplets, and we were able to confirm the inability to produce lipid droplets in Plin2^-/- macrophages using this approach (Figure 1—figure supplement 2A). As an indirect measure to track the fate of supplemented oleate in the mutant macrophages, we monitored lipid droplet biogenesis in Slc27a1^-/- and Cpt2^-/- macrophages to check if the inability to rescue the Mtb Δicl1 impaired growth phenotypes in these mutant macrophages could be possibly related to disruptions in lipid droplet formation. Confocal analysis of BODIPY-stained cells upon oleate supplementation revealed that Slc27a1^-/- macrophages also fail to generate lipid droplets (Figure 4B). In contrast, Cpt2^-/- macrophages produced more and larger lipid droplets in comparison to scrambled controls (Figure 4B). These data suggest that the inhibition of Mtb growth in these mutant macrophages is not merely through limitation of access to fatty acid nutrients.

We performed RNA sequencing of both host and bacteria in Mtb-infected mutant macrophages as a preliminary step in the identification of pathways restricting bacterial growth. We infected scrambled controls, Slc27a1^-/-, Plin2^-/-, and Cpt2^-/- macrophages with the Mtb smyc’::mCherry strain for 4 days and processed the samples for dual RNA sequencing as previously described (Simwela et al., 2024). Principal component analysis (PCA) of host transcriptomes revealed a clustering of all the three mutant macrophages away from scrambled controls (Figure 5A). Interestingly, there was a separation in transcriptional responses between the three mutant macrophages as Cpt2^-/- and Plin2^-/- macrophages clustered closer together and more distant from Slc27a1^-/- (Figure 5A). Overall, using an adjusted p-value<0.05 and absolute log₂ fold change >1.2, we identified 900 genes that were differentially expressed (DE) in Plin2^-/- macrophages (589 up, 311 down), 817 genes that were DE in Slc27a1^-/- macrophages (501 up, 315 down), and 189 genes that were DE in Cpt2^-/- (124 up, 65 down) (Supplementary file 2, Figure 5B). Consistent with the PCA, Venn diagram of the DE genes (Figure 5B) indicated divergent responses in the three mutant macrophage populations. We performed pathway enrichment analysis (Wu et al., 2021) of the DE genes to identify antimicrobial pathway candidates in the three mutant macrophages. We found that defects in fatty acid uptake in Slc27a1^-/--infected macrophages upregulated pro-inflammatory pathways involved in MAPK and ERK signaling and production of inflammatory cytokines (IFN‐γ, IL-6, IL-1α, β) (Figure 6A, Supplementary file 3). The pro-inflammatory signatures of the Slc27a1^-/- macrophages are consistent with previous observations that demonstrated that a deficiency in Slc27a1^-/- exacerbated macrophage activation in vitro and in vivo (Johnson et al., 2016). SLC27A1 is a solute carrier family member transporter and Mtb-infected Slc27a1^-/- macrophages showed reduced expression of other solute carrier transporters (SLC) such as SLC27A4, GLUT1 (SLC2A1), and eight SLC amino acid transporters (Figure 6—figure supplement 1A, B and 2A, Supplementary file 2). Interestingly, Mtb-infected Slc27a1^-/- macrophage transcriptomes exhibited upregulation of macrophage scavenger receptors (MSR1) and the ATP binding cassette transporter ABCC1 (Figure 6—figure supplement 1A), both of which can independently transport fatty acids into cells (Vogel et al., 2022; Raggers et al., 1999).

Meanwhile, Mtb infection of Plin2^-/- macrophages led to upregulation in pathways involved in ribosomal biology, MHC class 1 antigen presentation, canonical glycolysis, ATP metabolic processes, and type 1 interferon responses (Figure 6B, Supplementary file 3). In the downregulated Plin2^-/- DE gene set, enriched pathways included those involved in the production of pro-inflammatory cytokines; IL-6 and 8, IFN‐γ and IL-1 (Figure 6—figure supplement 2B, Supplementary file 3). Oxidative phosphorylation and processes involved in the respiratory chain electron transport were also significantly enriched in Plin2^-/- downregulated genes. This suggests that Mtb-infected Plin2^-/- macrophages increase glycolytic flux and decrease mitochondrial activities, which is consistent with our metabolic flux analysis data (Figure 2, Figure 2—figure supplement 1). Unlike Slc27a1^-/- macrophages, Plin2^-/- macrophages are, however, broadly anti-inflammatory as most pro-inflammatory genes were downregulated upon Mtb infection.

Further downstream in the lipid processing steps, inhibition of fatty acid oxidation in Cpt2^-/- macrophages upregulated pathways involved in MHC class 1 antigen presentation, response to IFN‐γ and IL-1 and T-cell-mediated immunity (Figure 6—figure supplement 3, Supplementary file 3). There was a limited overlap in enriched pathways in the upregulated genes between Mtb-infected Cpt2^-/- and Slc27a1^-/- macrophages such as those involved in the cellular responses to IL-1 and IFN‐γ. However, many pathways over-represented in Cpt2^-/- macrophages were common to Plin2^-/- macrophages (Figure 6B, Figure 6—figure supplement 3, Supplementary file 3). Similarly, both Mtb-infected Plin2^-/- and Cpt2^-/- macrophages were downregulated in the expression of genes involved in oxidative phosphorylation (Supplementary file 3). We confirmed expression levels of a selected gene by qPCR analysis of IL-1β and the type 1 interferon (IFN-β) response during the early time points of infection. Indeed, 4 hours post infection, IL-1β and IFN-β were both upregulated in Slc27a1^-/- macrophages compared to scrambled controls consistent with their pro-inflammatory phenotype (Figure 6—figure supplement 4A and B). On the contrary, Plin2^-/- macrophages downregulated IL-1β (Figure 6—figure supplement 4B). These data indicate that macrophages respond quite divergently to the deletion of the different steps in fatty acid uptake, which implies that the intracellular pressures to which Mtb is exposed may also differ.

We also analyzed transcriptomes from intracellular Mtb from scrambled controls, Slc27a1^-/-, Cpt2^-/-, and Plin2^-/- macrophages in parallel with host transcriptomes in Figure 5A. Using an adjusted p-value of <0.1 and an absolute log₂ fold change >1.4, 0 genes were DE in Slc27a1^-/- macrophages, 105 Mtb genes were DE in Plin2^-/- macrophages (69 up, 36 down), and 10 genes were DE in Cpt2^-/- macrophages (3 up, 7 down) (Supplementary file 4). Despite being restrictive to Mtb growth (Figure 1B) and appearing more pro-inflammatory (Figure 6A), Slc27a1^-/- macrophages did not elicit a detectable shift in the transcriptional response in Mtb compared to control host cells. We speculate that pro-inflammatory responses in Slc27a1^-/- macrophages could be enough to restrict the growth of bacteria, but the resulting compensatory responses as evidenced by the upregulation of macrophage scavenger receptors (Figure 6—figure supplement 1A) alleviate some of the stresses that a lack of fatty acid import could be duly exerting on the bacteria. Plin2^-/- macrophages appeared to elicit the strongest transcriptional response from Mtb, which is consistent with our CFU data (Figure 1B) as Plin2^-/- macrophages exhibited the strongest growth restriction. Among the DE genes in Mtb from Plin2^-/- macrophages (Supplementary file 4), a significant number of upregulated genes are involved in nutrient assimilation (Figure 7A). Mtb in Plin2^-/- macrophages upregulated CobU (Rv0254c), which is predicted to be involved in the bacteria’s cobalamin (vitamin B₁₂) biosynthesis. Vitamin B₁₂ is an important cofactor for the activity of Mtb genes required for cholesterol and fatty acid utilization (Campos-Pardos et al., 2024; Savvi et al., 2008). Genes involved in de novo long-chain fatty acid synthesis (AccE5, Rv281) (Bazet Lyonnet et al., 2014), cholesterol breakdown (HsaD, Rv3569c) (Lack et al., 2010), β-oxidation of fatty acids (EchA18, Rv3373; FadE22, Rv3061c) (Schnappinger et al., 2003), purine salvage (Apt, Rv2584c) (Warner et al., 2014), and tryptophan metabolism (Lott, 2020) were also upregulated in Mtb from Plin2^-/- macrophages (Figure 7A). This metabolic realignment response is seen most frequently under nutrient-limiting conditions (Huang et al., 2018; Pisu et al., 2020; Theriault et al., 2022). Mtb in Plin2^-/- macrophages also appears to experience a significant level of other cellular stresses as genes involved in DNA synthesis and repair, general response to oxidative stress and pH survival in the phagosome (DnaN, Rv0002; RecF, Rv0003; DinF, Rv2836c; Rv3242c, Rv1264) were upregulated (Figure 7B). Among the downregulated genes in Mtb in Plin2^-/- macrophages, FurA (Rv1909c), a KatG repressor was the most significant (Figure 7B, Supplementary file 4). FurA downregulation derepresses the catalase peroxidase, KatG, which promotes Mtb survival in oxidative stress conditions (Zahrt et al., 2001). These data suggest that Plin2^-/- macrophages could be, in part, restricting Mtb growth by increasing the production of ROS. The data also suggest that, contrary to a previous report (Knight et al., 2018), blocking lipid droplet formation in host macrophages does place increased nutritional and oxidative stress on intracellular Mtb.

Cpt2^-/- macrophages elicited a modest shift in transcriptional response from Mtb, and the majority of the DE genes (8 out of 10) were also DE in Plin2^-/- macrophages (Supplementary file 4). This is in agreement with the host transcription response as Cpt2^-/- and Plin2^-/- macrophages share similar candidate antibacterial responses (Figure 5A and B, Figure 6, Figure 6—figure supplement 3). In common with the Mtb transcriptome response in Plin2^-/- macrophages, upregulated genes in Mtb isolated from Cpt2^-/- macrophages included those involved in response to oxidative stress (DinF, Rv2836c) (Figure 7B, Supplementary file 4). To substantiate some of these pathways, we assessed the levels of total cellular ROS in Slc27a1^-/-, Plin2^-/-, and Cpt2^-/- macrophages in both Mtb-infected and uninfected conditions by staining the cells with the Invitrogen CellROX dye and confocal microscopy analysis of live stained cells. In both infected and uninfected conditions, all the three mutants displayed significantly elevated total cellular ROS compared to scramble controls (Figure 7—figure supplement 1A and B).

We next examined if pharmacological inhibitors would phenocopy the growth inhibition phenotypes we observed with specific gene deletions. Compounds that modulate lipid homeostasis are currently being investigated for HDT against TB, which is an area of considerable interest (Kim et al., 2020). Such compounds include metformin, a widely used antidiabetic drug that activates AMPK, inhibits fatty acid synthesis, and promotes β-oxidation of fatty acids (Fullerton et al., 2013; Singhal et al., 2014). Chemical inhibition of fatty acid β-oxidation is already known to promote macrophage control of Mtb (Chandra et al., 2020; Huang et al., 2018). We targeted macrophage lipid homeostasis with trimetazidine (TMZ), an inhibitor of β-oxidation of fatty acids, metformin, and an SLC27A1 inhibitor, FATP1 In (Matsufuji et al., 2013). We assessed the impact of these compounds on extracellular Mtb cultured in broth over 9 days in the presence of the inhibitors (DMSO, TMZ; 500 nM, metformin; 2 mM, FATP1 In; 10 μM, rifampicin; 0.5 μg/ml). None of the three lipid metabolism inhibitors had a measurable effect on Mtb growth in liquid culture media compared to DMSO controls (Figure 8A). Treatment with rifampicin completely blocked bacterial growth under the same conditions (Figure 8A). We next infected scrambled sgRNA control macrophages with Mtb at MOI 0.5. Inhibitors were added to infected macrophages 3 hours post infection, and bacterial CFUs were enumerated 4 days post treatment. Consistent with previous observations (Chandra et al., 2020; Singhal et al., 2014), TMZ and metformin significantly reduced bacterial loads in macrophages compared to DMSO controls (Figure 8B). Similarly, FATP1 In also impacted the intracellular growth of Mtb (Figure 8B). The results provide independent data that both genetic and chemical modulation of fatty acid metabolism at different steps in the process negatively impact intracellular growth of Mtb.

Generation of CRISPR mutant Hoxb8 macrophages was carried out as described previously (Simwela et al., 2024). Antibodies used for both western blot and flow cytometry were as follows: rat anti-mouse CD36:Alexa Fluor647 (Bio-Rad, 10 μl/million cells), rabbit anti-PLIN2 (Proteintech, 1:1000), rabbit anti-SLC27A1 (Affinity Biosciences, 1:1000), rabbit anti-CPT1A antibody (Proteintech, 1:1000), rabbit anti-CPT2 antibody (Proteintech, 1:1000), rabbit anti-HIF1α antibody (Proteintech, 1:1000), rabbit anti-AMPKα (1:1000, Cell Signalling Technology), rabbit anti-Phospho-AMPKα (1:500, Cell Signalling Technology), rabbit anti-LC3B (1:1000, Cell Signalling Technology), and rabbit anti-β-actin (1:1000, Cell Signalling Technology). For western blots, secondary antibodies used were anti-rabbit/mouse StarBright Blue 700 (1:2500, Bio-Rad). Blots were developed and imaged as described previously (Simwela et al., 2024).

Macrophages monolayers in Ibidi eight-well chambers were supplemented with exogenous 400 μM oleate for 24 hours to induce the formation of lipid droplets (Listenberger and Brown, 2007). Cells were then fixed in 4% paraformaldehyde and stained with BODIPY 493/503 (Invitrogen, 1 μg/ml) in 150 mM sodium chloride. Stained cells were mounted with media containing DAPI and imaged using a Leica SP5 confocal microscope.

A modified Seahorse mitochondrial stress test was used to measure macrophage’s ability to oxidize palmitate in substrate-limiting conditions. Two days before the assay, 1 × 10⁵ cells were plated in Seahorse cell culture mini plates. One day before the assay, macrophage media was replaced with the Seahorse substrate-limited growth media (DMEM without pyruvate supplemented with 0.5 mM glucose, 1 mM glutamine, 1% FBS, and 0.5 mM L-carnitine). On the day of the assay, substrate-limited media was replaced with assay media (DMEM without pyruvate supplemented with 2 mM glucose and 0.5 mM L-carnitine). In selected treatment conditions, cells were either supplied with BSA, BSA palmitate, or BSA palmitate plus etomoxir (4 μM). OCRs were measured using the Mito Stress Test assay conditions as described previously (Simwela et al., 2024).

The Mtb H37Rv Δicl1 mutant expressing mCherry (Lee et al., 2013) was used for the rescue experiments. The strain was maintained in 7H9 OADC broth as previously described (Simwela et al., 2024) in the presence of kanamycin (25 μg/ml) and hygromycin (50 μg/ml). 24 hours before infection, macrophages were cultured in normal macrophage media or media supplemented with 400 μM oleate. Cells were then infected with the Mtb Δicl1 mutant at MOI 5. The bacterial mCherry signal was measured on day 0 and day 5 post infection on an Envision plate reader (PerkinElmer). Oleate was maintained throughout the experiment in the rescue assay conditions.

Uninfected or Mtb-infected macrophages monolayers in Ibidi eight-well chambers were stained with the CellROX Deep Red dye (Invitrogen) as per the manufacturer’s staining protocol. Live stained cells were imaged on a Leica SP5 confocal microscope. Z-stacks were reconstructed in ImageJ from which mean fluorescence intensities (MFIs) for individual cells were obtained.