CRISPR Interference Reveals That All- Trans-Retinoic Acid Promotes Macrophage Control of Mycobacterium tuberculosis by Limiting Bacterial Access to Cholesterol and Propionyl Coenzyme A

Abstract

Macrophages are a protective replicative niche for Mycobacterium tuberculosis (Mtb) but can kill the infecting bacterium when appropriately activated. To identify mechanisms of clearance, we compared levels of bacterial restriction by human macrophages after treatment with 26 compounds, including some currently in clinical trials for tuberculosis. All-trans-retinoic acid (ATRA), an active metabolite of vitamin A, drove the greatest increase in Mtb control. Bacterial clearance was transcriptionally and functionally associated with changes in macrophage cholesterol trafficking and lipid metabolism. To determine how these macrophage changes affected bacterial control, we performed the first Mtb CRISPR interference screen in an infection model, identifying Mtb genes specifically required to survive in ATRA-activated macrophages. These data showed that ATRA treatment starves Mtb of cholesterol and the downstream metabolite propionyl coenzyme A (propionyl-CoA). Supplementation with sources of propionyl-CoA, including cholesterol, abrogated the restrictive effect of ATRA. This work demonstrates that targeting the coupled metabolism of Mtb and the macrophage improves control of infection and that it is possible to genetically map the mode of bacterial death using CRISPR interference. IMPORTANCE Tuberculosis, caused by the bacterium Mycobacterium tuberculosis, is a leading cause of death due to infectious disease. Improving the immune response to tuberculosis holds promise for fighting the disease but is limited by our lack of knowledge as to how the immune system kills M. tuberculosis. Our research identifies a potent way to make relevant immune cells more effective at fighting M. tuberculosis and then uses paired human and bacterial genomic methods to determine the mechanism of that improved bacterial clearance.

Keywords: CRISPR interference; Mycobacterium tuberculosis; cholesterol; macrophages; nutritional immunity; propionyl-CoA; retinoic acid.

FIG 1

All-trans-retinoic acid outperforms other activators in eliciting human macrophage control of M. tuberculosis infection. (a) Control by differentially matured MO-MCSF and MO-GMCSF (each point represents 1 donor) of autobioluminescent Mtb H37Rv at 5 days following infection and treatment with different activators, detailed in the compound key. (b) Growth in MO-GMCSF (3 donors) of autobioluminescent Mtb over time following treatment with different concentrations of ATRA or imatinib at 10 μM. (c) Comparison of Mtb loads in MO-GMCSF (3 donors) by CFU assay for ATRA and imatinib (10 μM) at 5 days. (d) Growth of autobioluminescent Mtb in MO-GMCSF (2 to 4 donors) at 5 days following treatment with ATRA or retinol (10 μM). All infections were performed at a multiplicity of infection of 2 bacteria per macrophage, with between 2 and 5 donors (a, c, d) or 3 donors (b) per condition. All summary data represent means ± standard deviations (SD). Any comparisons to the DMSO control not marked with statistical significance were not significant; statistics were performed using a 2-way analysis of variance (ANOVA) with Dunnett’s multiple-comparison test (a), an ordinary one-way ANOVA on area under the curve measurements with Šídák’s multiple-comparison test (b), a repeated-measures ANOVA with Dunnett’s multiple-comparison test (c), or an ordinary one-way ANOVA with Tukey’s multiple-comparison test (d). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FIG 2

Macrophage restriction of M. tuberculosis via all-trans-retinoic acid is associated with cholesterol limitation. (a) Comparison of Mtb loads in MO-GMCSF (3 donors) by CFU assay for ATRA and other pan-RAR agonists. (b to d) Volcano plots of differential gene expression by Mtb-infected MO-GMCSF (3 donors), comparing ATRA to a DMSO control (b) or to nonrestrictive pan-RAR agonists (c, d). (e) Clustered heatmap of gene expression relative to that in uninfected macrophages for all genes significantly differentially expressed during ATRA treatment relative to treatment with both nonrestrictive RAR agonists in infected MO-GMCSF. (f) Network plot of gene set clusters, the expression of which was significantly enriched during ATRA treatment relative to treatment with both of the nonrestrictive RAR agonists, in Mtb-infected MO-GMCSF. (g) Top 5 genes (largest absolute fold change) driving the enrichment of each gene set cluster shown in panel f and their differential expression values from the data displayed in panels c and d. (h) Changes in total cellular cholesterol 1 day after Mtb infection and treatment for MO-GMCSF (6 donors) treated with ATRA or nonrestrictive RAR agonists. All infections were performed at a multiplicity of infection of 2 bacteria per macrophage. All compounds were used at 10 μM in 0.2% DMSO. (a, h) Any comparisons to the DMSO control not marked with statistical significance were not significant in summary data representing means ± SD (a) and in lines representing the data median (h). Statistics were performed using a repeated-measures one-way ANOVA with Dunnett’s multiple-comparison test (a) or an ordinary one-way ANOVA with Dunnett’s multiple-comparison test (h). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FIG 3

Bacterial CRISPRi reveals increased reliance of M. tuberculosis on lipid import genes in macrophages treated with all-trans-retinoic acid. (a) Changes in gene-sgRNA set representation as measured by beta score (similar to log fold change), following CRISPRi library induction (+ATc) in axenic medium compared to the score for uninduced growth. Dark-blue points represent essential genes targeted with sets of strong (_s) or medium (_m) hypomorph sgRNAs (3 sgRNAs per point), light-blue points represent essential genes targeted with weak (_w) hypomorph sets of sgRNAs (3 sgRNAs per point), gray points represent nonessential genes, and black points represent individual negative (nontargeting) sgRNAs. (b) Schematic of CRISPRi experiments in macrophages and axenic broth culture. (c) Changes in gene-sgRNA set representation as measured by the beta score following CRISPRi library induction in axenic medium compared to the score for induced (+ATc) growth in primary mouse macrophages. Significance is measured by determining the Wald P value (equivalent to an adjusted P value). Point colors are as described for panel a; the green line is at the diagonal. (d, e) Volcano plots showing changes in gene-sgRNA set representation between 5 μM ATRA- and DMSO-treated primary macrophages from mouse (d) and human (e) (MO-GMCSF) sources. Red points represent gene-sgRNA sets that are significant at a Wald P value cutoff of <0.05. (f) MO-GMCSF control of autobioluminescent single-sgRNA clonal strains of Mtb at 5.5 days following infection and treatment with 5 μM ATRA, comparing strains with ATc-induced knockdown of the indicated Mtb genes to the same strains with no knockdown. Statistics were performed using a 2-way ANOVA with Šídák’s multiple-comparison test; color-coded adjusted P values are shown above each comparison (red indicates a P of <0.05, white indicates a P of 0.05, and blue indicates a P of >0.05 [continuous scale]). Infections were performed at a multiplicity of infection of 1 bacterium (a to e) or 2 bacteria (f) per macrophage.

FIG 4

Macrophage restriction of M. tuberculosis elicited by all-trans-retinoic acid can be relieved by cholesterol or odd-chain fatty acids. (a, b) Control of autobioluminescent Mtb treated with ATRA and various concentrations of water-soluble cholesterol, in MO-GMCSF at 5.5 days (a) or in BMDM at 9.5 days (b). (c to h) Comparison of Mtb loads by CFU assay at 5 days in MO-GMCSF (c to e) or BMDM (f to h). ATRA was added at 10 μM, and fatty acids were added complexed to essentially fatty acid-free bovine serum albumin (BSA) at a 3:1 ratio. All infections were performed at a multiplicity of infection of 2 bacteria per macrophage, with all samples in triplicate (3 donors for MO-GMCSF). DMSO was used at 0.1% in all comparisons. Statistics were performed using ordinary one-way ANOVAs with Šídák’s (a) or Dunnett’s (b) multiple-comparison test, ratio paired t tests (c to e), or unpaired t tests (f to h). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

FIG 5

ATRA-mediated restriction is dependent on limited propionyl-CoA and the methylcitrate cycle. (a, b) MO-GMCSF from 3 donors were infected with autobioluminescent Mtb at a multiplicity of infection of 2 bacteria per macrophage and treated with ATRA at 10 μM and/or propionate at 10 mM, with a baseline of 0.1% DMSO; luminescence was measured at 5.5 days. (c, d) BMDM (n = 3) were infected with the indicated Mtb strains (H37Rv background) at a multiplicity of infection of 2 bacteria per macrophage, with Mtb load enumerated by determining numbers of CFU at 5.5 days. Data are displayed as a ratio of the number of ATRA-treated CFU divided by the number of DMSO-treated CFU. Statistics were performed using a paired (a, b) or unpaired (c, d) t test. *, P < 0.05; ***, P < 0.001. WT, wild type.

FIG 6

M. tuberculosis increasingly requires anaplerotic assimilation of propionyl-CoA within macrophages treated with all-trans-retinoic acid. Model of cholesterol and fatty acids crossing the Mtb cell wall (purple) and progressing through propionyl-CoA metabolism, with conditional essentiality during ATRA treatment of macrophages for selected Mtb genes shown in heatmaps (data are from Table S3). Specific molecules are in blue boxes, and molecular classes are in gray boxes. Metabolic flux directionality was adapted from the published literature (53). Statistics were performed with MAGeCK MLE analysis for Wald P values.