Impaired fatty acid import or catabolism in macrophages restricts intracellular growth of Mycobacterium tuberculosis

Reviewer #1 (Public review):

Summary:

This study investigates the role of macrophage lipid metabolism in the intracellular growth of Mycobacterium tuberculosis. By using a CRISPR-Cas9 gene-editing approach, the authors knocked out key genes involved in fatty acid import, lipid droplet formation, and fatty acid oxidation in macrophages. Their results show that disrupting various stages of fatty acid metabolism significantly impairs the ability of Mtb to replicate inside macrophages. The mechanisms of growth restriction included increased glycolysis, oxidative stress, pro-inflammatory cytokine production, enhanced autophagy, and nutrient limitation. The study demonstrates that targeting fatty acid homeostasis at different stages of the lipid metabolic process could offer new strategies for host-directed therapies against tuberculosis.

The work is convincing and methodologically strong, combining genetic, metabolic, and transcriptomic analyses to provide deep insights into how host lipid metabolism affects bacterial survival.

Strengths:

The study uses a multifaceted approach, including CRISPR-Cas9 gene knockouts, metabolic assays, and dual RNA sequencing, to assess how various stages of macrophage lipid metabolism affect Mtb growth. The use of CRISPR-Cas9 to selectively knock out key genes involved in fatty acid metabolism enables precise investigation of how each step-lipid import, lipid droplet formation, and fatty acid oxidation affect Mtb survival. The study offers mechanistic insights into how different impairments in lipid metabolism lead to diverse antimicrobial responses, including glycolysis, oxidative stress, and autophagy. This deepens the understanding of macrophage function in immune defense.

The use of functional assays to validate findings (e.g., metabolic flux analyses, lipid droplet formation assays, and rescue experiments with fatty acid supplementation) strengthens the reliability and applicability of the results.

By highlighting potential targets for HDT that exploit macrophage lipid metabolism to restrict Mtb growth, the work has significant implications for developing new tuberculosis treatments.

Weaknesses:

The experiments were primarily conducted in vitro using CRISPR-modified macrophages. While these provide valuable insights, they may not fully replicate the complexity of the in vivo environment where multiple cell types and factors influence Mtb infection and immune responses.

Reviewer #2 (Public review):

Summary:

Host-derived lipids are an important factor during Mtb infection. In this study, using CRISPR knockouts of genes involved in fatty acid uptake and metabolism, the authors claim that a compromised uptake, storage, or metabolism of fatty acid restricts Mtb growth upon infection. Further, the authors claim that the mechanism involves increased glycolysis, autophagy, oxidative stress, pro-inflammatory cytokines, and nutrient limitation. The authors also claim that impaired lipid droplet formation restricts Mtb growth. However, promoting lipid droplet biogenesis does not reverse/promote Mtb growth.

Strengths:

The strength of the study is the use of clean HOXB8-derived primary mouse macrophage lines for generating CRISPR knockouts.

Weaknesses:

There are many weaknesses of this study, they are clubbed into four categories below

(1) Evidence and interpretations: The results shown in this study at several places do not support the interpretations made or are internally contradictory or inconsistent. There are several important observations, but none were taken forward for in-depth analysis. A
a) The phenotypes of PLIN2-/-, FATP1-/-, and CPT-/- are comparable in terms of bacterial growth restriction; however, their phenotype in terms of lipid body formation, IL1B expression, etc., are not consistent. These are interesting observations and suggest additional mechanisms specific to specific target genes; however, clubbing them all as altered fatty acid uptake or catabolism-dependent phenotypes takes away this important point. b) Finding the FATP1 transcript in the HOXB8-derived FATP1-/- CRISPR KO line is a bit confusing. There is less than a two-fold decrease in relative transcript abundance in the KO line compared to the WT line, leaving concerns regarding the robustness of other experiments as well using FATP1-/- cells.
c) No gene showing differential regulation in FATP-/- macrophages, which is very surprising.
d) ROS measurements should be done using flow cytometry and not by microscopy to nail the actual pattern.

(2) Experimental design: For a few assays, the experimental design is inappropriate
a) For autophagy flux assay, immunoblot of LC3II alone is not sufficient to make any interpretation regarding the state of autophagy. This assay must be done with BafA1 or CQ controls to assess the true state of autophagy.
b) Similarly, qPCR analyses of autophagy-related gene expression do not reflect anything on the state of autophagy flux.

(3) Using correlative observations as evidence:
a) Observations based on RNAseq analyses are presented as functional readouts, which is incorrect.
b) Claiming that the inability to generate lipid droplets in PLIN2-/- cells led to the upregulation of several pathways in the cells is purely correlative, and the causal relationship does not exist in the data presented.

(4) Novelty: A few main observations described in this study were previously reported. That includes Mtb growth restriction in PLIN2 and FATP1 deficient cells. Similarly, the impact of Metformin and TMZ on intracellular Mtb growth is well-reported. While that validates these observations in this study, it takes away any novelty from the study.

(5) Manuscript organisation: It will be very helpful to rearrange figures and supplementary figures.

Reviewer #3 (Public review):

Summary:

This study provides significant insights into how host metabolism, specifically lipids, influences the pathogenesis of Mycobacterium tuberculosis (Mtb). It builds on existing knowledge about Mtb's reliance on host lipids and emphasizes the potential of targeting fatty acid metabolism for therapeutic intervention.

Strengths:

To generate the data, the authors use CRISPR technology to precisely disrupt the genes involved in lipid import (CD36, FATP1), lipid droplet formation (PLIN2), and fatty acid oxidation (CPT1A, CPT2) in mouse primary macrophages. The Mtb Erdman strain is used to infect the macrophage mutants. The study, revealsspecific roles of different lipid-related genes. Importantly, results challenge previous assumptions about lipid droplet formation and show that macrophage responses to lipid metabolism impairments are complex and multifaceted. The experiments are well-controlled and the data is convincing.

Overall, this well-written paper makes a meaningful contribution to the field of tuberculosis research, particularly in the context of host-directed therapies (HDTs). It suggests that manipulating macrophage metabolism could be an effective strategy to limit Mtb growth.

Weaknesses:

None noted. The manuscript provides important new knowledge that will lead mpvel to host-directed therapies to control Mtb infections.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study investigates the role of macrophage lipid metabolism in the intracellular growth of Mycobacterium tuberculosis. By using a CRISPR-Cas9 gene-editing approach, the authors knocked out key genes involved in fatty acid import, lipid droplet formation, and fatty acid oxidation in macrophages. Their results show that disrupting various stages of fatty acid metabolism significantly impairs the ability of Mtb to replicate inside macrophages. The mechanisms of growth restriction included increased glycolysis, oxidative stress, pro-inflammatory cytokine production, enhanced autophagy, and nutrient limitation. The study demonstrates that targeting fatty acid homeostasis at different stages of the lipid metabolic process could offer new strategies for host-directed therapies against tuberculosis.

The work is convincing and methodologically strong, combining genetic, metabolic, and transcriptomic analyses to provide deep insights into how host lipid metabolism affects bacterial survival.

Strengths:

The study uses a multifaceted approach, including CRISPR-Cas9 gene knockouts, metabolic assays, and dual RNA sequencing, to assess how various stages of macrophage lipid metabolism affect Mtb growth. The use of CRISPR-Cas9 to selectively knock out key genes involved in fatty acid metabolism enables precise investigation of how each step-lipid import, lipid droplet formation, and fatty acid oxidation affect Mtb survival. The study offers mechanistic insights into how different impairments in lipid metabolism lead to diverse antimicrobial responses, including glycolysis, oxidative stress, and autophagy. This deepens the understanding of macrophage function in immune defense.

The use of functional assays to validate findings (e.g., metabolic flux analyses, lipid droplet formation assays, and rescue experiments with fatty acid supplementation) strengthens the reliability and applicability of the results.

By highlighting potential targets for HDT that exploit macrophage lipid metabolism to restrict Mtb growth, the work has significant implications for developing new tuberculosis treatments.

Weaknesses:

The experiments were primarily conducted in vitro using CRISPR-modified macrophages. While these provide valuable insights, they may not fully replicate the complexity of the in vivo environment where multiple cell types and factors influence Mtb infection and immune responses.

We thank the reviewer for pointing this out. We acknowledge that our in vitro system may indeed not fully replicate the complex in vivo environment in light of the heterogenous responses of macrophages to Mtb infection in whole animal models. We do believe, however, that the Hoxb8 in vitro model provides a powerful genetic tool to interrogate host-Mtb interactions using primary macrophages that represent the bone marrow-derived macrophage lineage. Reviewer #1 also made several helpful suggestions in their recommendations to authors relating to the reorganization of the data in our Figures in both the manuscript and the supplemental data. We will incorporate these suggestions into the revised version of the manuscript upon resubmission.

Reviewer #2 (Public review):

Summary:

Host-derived lipids are an important factor during Mtb infection. In this study, using CRISPR knockouts of genes involved in fatty acid uptake and metabolism, the authors claim that a compromised uptake, storage, or metabolism of fatty acid restricts Mtb growth upon infection. Further, the authors claim that the mechanism involves increased glycolysis, autophagy, oxidative stress, pro-inflammatory cytokines, and nutrient limitation. The authors also claim that impaired lipid droplet formation restricts Mtb growth. However, promoting lipid droplet biogenesis does not reverse/promote Mtb growth.

Strengths:

The strength of the study is the use of clean HOXB8-derived primary mouse macrophage lines for generating CRISPR knockouts.

Weaknesses:

There are many weaknesses of this study, they are clubbed into four categories below

(1) Evidence and interpretations: The results shown in this study at several places do not support the interpretations made or are internally contradictory or inconsistent. There are several important observations, but none were taken forward for in-depth analysis. A

a) The phenotypes of PLIN2-/-, FATP1-/-, and CPT-/- are comparable in terms of bacterial growth restriction; however, their phenotype in terms of lipid body formation, IL1B expression, etc., are not consistent. These are interesting observations and suggest additional mechanisms specific to specific target genes; however, clubbing them all as altered fatty acid uptake or catabolism-dependent phenotypes takes away this important point.

We thank the reviewer for highlighting this. Our main focus was on assessing the impact of manipulating lipid homeostasis in macrophages and the consequences this has on the intracellular growth of Mtb. It was never our intention to imply these mutants generated equivalent phenotypes, and we will modify the revised manuscript to reflect this point. We will stress that interfering with lipid processing at different stages in macrophages results in both shared and divergent anti-microbial conditions against Mtb.

b) Finding the FATP1 transcript in the HOXB8-derived FATP1-/- CRISPR KO line is a bit confusing. There is less than a two-fold decrease in relative transcript abundance in the KO line compared to the WT line, leaving concerns regarding the robustness of other experiments as well using FATP1^-/- cells.

CRISPR-Cas9 targeting of genes with single sgRNAs as is the case with our mutants generates insertions and deletions (INDELs) at the CRISPR cut site. These INDELs do not block mRNA transcription totally, and this is widely reported and accepted in the field. In these cases, RT-PCR or RNA-seq methods are not used to verify CRISPR knockouts as they are not sensitive enough to identify INDELs. We provide knockout efficiencies by ICE analysis in supplemental information file 1 for all the mutants used in the study. We also demonstrate protein depletion by western blot and flow cytometry for all the mutants (Figure 1 - figure supplement 1). Only mutants with greater than >90% protein depletion were used for subsequent characterization.

c) No gene showing differential regulation in FATP1^-/- macrophages, which is very surprising.

We assume the reviewer is referring to the Mtb transcriptome response in FATP1^-/- macrophages, which we agree was unexpected. However, we saw a significant compensatory response in the host cell (at transcriptional level) in FATP1-/- macrophages as evidenced by an upregulation of other fatty acid transporters (Figure 5 - figure supplement 1). We postulate that these compensatory responses could, in part, alleviate the stresses the bacteria experience within the cell, and these were discussed in the manuscript.

d) ROS measurements should be done using flow cytometry and not by microscopy to nail the actual pattern.

We thank the reviewer for the suggestion. However, confocal imaging is also widely used to measure ROS with similar quantitative power and individual cell resolution (PMID: 32636249, 35737799).

(2) Experimental design: For a few assays, the experimental design is inappropriate

a) For autophagy flux assay, immunoblot of LC3II alone is not sufficient to make any interpretation regarding the state of autophagy. This assay must be done with BafA1 or CQ controls to assess the true state of autophagy.

We would like to point out that monitoring LC3I to LC3II conversion by western blot, confocal imaging of LC3 puncta and qPCR analysis of autophagy related genes are all validated assays for monitoring autophagic flux in a wide variety of cells. We refer the reviewer to the latest extensive guidelines on the subject (PMID: 33634751). Furthermore, Bafilomycin A and chloroquine are not specific inhibitors of autophagy and therefore are of limited value as controls. BafA is an inhibitor of the proton-ATPase apparatus as well impacting autophagy through activity on the Ca-P60A/SERCA pathway. Chloroquine impacts vacuole acidification, autophagosome/lysosome fusion and slows phagosome maturation. So, while BafA and chloroquine will reduce autophagy their effects are pleotropic and their impact on Mtb is unknown.

b) Similarly, qPCR analyses of autophagy-related gene expression do not reflect anything on the state of autophagy flux.

See our response above.

(3) Using correlative observations as evidence:

a) Observations based on RNAseq analyses are presented as functional readouts, which is incorrect.

We are not entirely sure where we used our RNA-seq data sets as functional readouts. We used our transcriptome data to provide a preliminary identification of anti-microbial responses in the mutant macrophages infected with Mtb. Where applicable, we followed up and confirmed the more compelling RNA-seq data either by metabolic flux analyzes, qPCR, ROS measurements, and quantitative imaging.

b) Claiming that the inability to generate lipid droplets in PLIN2-/- cells led to the upregulation of several pathways in the cells is purely correlative, and the causal relationship does not exist in the data presented.

Again, it was not our intention to infer causality. Throughout the manuscript, we endeavor to present our data with a specific focus on describing the consequences of interfering with either fatty acid import, lipid droplet biogenesis and fatty acid oxidation on macrophage responses to Mtb. We will revisit the revised manuscript to remove any sections that imply causality.

(4) Novelty: A few main observations described in this study were previously reported. That includes Mtb growth restriction in PLIN2 and FATP1 deficient cells. Similarly, the impact of Metformin and TMZ on intracellular Mtb growth is well-reported. While that validates these observations in this study, it takes away any novelty from the study.

To the best of our knowledge, Mtb growth restrictions in PLIN2 and FATP1 deficient macrophages have not been reported elsewhere. To the contrary, PLIN2 knockout macrophages obtained from PLIN2 deficient mice have been reported to robustly support Mtb replication (PMID: 29370315), quite the opposite to our data. We extensively discuss these discrepancies in the manuscript. We also discuss and cite appropriate references where Mtb growth restriction for similar macrophage mutants have been reported (CD36^-/- and CPT2^-/-). Our aim was to carry out a systematic myeloid specific genetic interference of fatty acid import, storage and catabolism to assess the effect on Mtb growth at all stages of lipid handling instead of focusing on one target. In the chemical approach, we used TMZ and Metformin deliberately because they had already been reported as being active against intracellular Mtb and we wished to place our data in the context of existing literature. These studies were referenced extensively in the text.

(5) Manuscript organisation: It will be very helpful to rearrange figures and supplementary figures.

We will re-organize the figures in the manuscript revision as per the reviewer’s recommendation, and the recommendations of reviewer #1.

We will address the other concerns raised by reviewer #2 in the recommendations to authors during revision of the manuscript.

Reviewer #3 (Public review):

Summary:

This study provides significant insights into how host metabolism, specifically lipids, influences the pathogenesis of Mycobacterium tuberculosis (Mtb). It builds on existing knowledge about Mtb's reliance on host lipids and emphasizes the potential of targeting fatty acid metabolism for therapeutic intervention.

Strengths:

To generate the data, the authors use CRISPR technology to precisely disrupt the genes involved in lipid import (CD36, FATP1), lipid droplet formation (PLIN2), and fatty acid oxidation (CPT1A, CPT2) in mouse primary macrophages. The Mtb Erdman strain is used to infect the macrophage mutants. The study, reveals specific roles of different lipid-related genes. Importantly, results challenge previous assumptions about lipid droplet formation and show that macrophage responses to lipid metabolism impairments are complex and multifaceted. The experiments are well-controlled and the data is convincing.

Overall, this well-written paper makes a meaningful contribution to the field of tuberculosis research, particularly in the context of host-directed therapies (HDTs). It suggests that manipulating macrophage metabolism could be an effective strategy to limit Mtb growth.

Weaknesses:

None noted. The manuscript provides important new knowledge that will lead mpvel to host-directed therapies to control Mtb infections.