Noise in a Metabolic Pathway Leads to Persister Formation in Mycobacterium tuberculosis
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This manuscript is of interest to readers in the field of microbiology and antimicrobial resistance. It reports that genetically identical antibiotic-sensitive bacteria can escape killing upon exposure to an otherwise lethal concentration of antibiotics because they have a low ATP level. The authors further attempt to demonstrate that variation in the level of expression of some genes in energy-generating metabolic pathways allows for a wide range of ATP levels among the cells in a population, thus generating a subpopulation with low ATP levels that can survive antibiotic exposure.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
- Evaluated articles (eLife)
Abstract
M. tuberculosis infection requires the administration of multiple antibiotics for a prolonged period of time. Treatment difficulty is generally attributed to M. tuberculosis entrance into a nonreplicative, antibiotic-tolerant state. M. tuberculosis enters this nonreplicative state in response to immune stress.
Article activity feed
-
-
Evaluation Summary:
This manuscript is of interest to readers in the field of microbiology and antimicrobial resistance. It reports that genetically identical antibiotic-sensitive bacteria can escape killing upon exposure to an otherwise lethal concentration of antibiotics because they have a low ATP level. The authors further attempt to demonstrate that variation in the level of expression of some genes in energy-generating metabolic pathways allows for a wide range of ATP levels among the cells in a population, thus generating a subpopulation with low ATP levels that can survive antibiotic exposure.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
Quigley et al. provide new insights into persister cell formation in Mycobacterium tuberculosis, a notorious pathogen provoking infectious diseases worldwide. The study is most relevant as M. tuberculosis is particularly tolerant to antibiotic treatment because of a dormant state. The authors show, at the single-cell level, that low ATP Mt cells are killed more slowly by antibiotics compared to high ATP cells, indicating the critical importance of ATP levels in drug tolerance. Further, the authors claim that increased noise in ackA expression, involved in carbon metabolism, results in higher survival. However, this link is currently less clear and could be further strengthened to make solid proof of the underlying molecular mechanisms. To conclude, this manuscript is very interesting and may lead to improved …
Reviewer #1 (Public Review):
Quigley et al. provide new insights into persister cell formation in Mycobacterium tuberculosis, a notorious pathogen provoking infectious diseases worldwide. The study is most relevant as M. tuberculosis is particularly tolerant to antibiotic treatment because of a dormant state. The authors show, at the single-cell level, that low ATP Mt cells are killed more slowly by antibiotics compared to high ATP cells, indicating the critical importance of ATP levels in drug tolerance. Further, the authors claim that increased noise in ackA expression, involved in carbon metabolism, results in higher survival. However, this link is currently less clear and could be further strengthened to make solid proof of the underlying molecular mechanisms. To conclude, this manuscript is very interesting and may lead to improved understanding of persistence in relation to ATP levels and noise in an important pathogen.
-
Reviewer #2 (Public Review):
This manuscript by Quigley & Lewis investigates the impact of bacterial intracellular ATP concentration on the ability of Mycobacterium tuberculosis to survive-but not grow-upon exposure to antibiotics. This manuscript follows others (PMID 33872303, 34982597) from the same laboratory that convey similar findings but focuses on a different organism, Mycobacterium tuberculosis. First, the authors show that cells treated with bedaquiline (BDQ) (the concentration used is not indicated), a drug that leads to a decrease in the intracellular ATP pool, are better able to survive subsequent exposure to antibiotics of other classes. Although this is compelling, it should be noted that BDQ leads to other changes in treated cells, such as rapid collateral vulnerability of another metabolic enzyme, glutamine synthase …
Reviewer #2 (Public Review):
This manuscript by Quigley & Lewis investigates the impact of bacterial intracellular ATP concentration on the ability of Mycobacterium tuberculosis to survive-but not grow-upon exposure to antibiotics. This manuscript follows others (PMID 33872303, 34982597) from the same laboratory that convey similar findings but focuses on a different organism, Mycobacterium tuberculosis. First, the authors show that cells treated with bedaquiline (BDQ) (the concentration used is not indicated), a drug that leads to a decrease in the intracellular ATP pool, are better able to survive subsequent exposure to antibiotics of other classes. Although this is compelling, it should be noted that BDQ leads to other changes in treated cells, such as rapid collateral vulnerability of another metabolic enzyme, glutamine synthase (PMID 31501323); thus, the increase in the number of bacteria surviving antibiotic treatment following exposure to BDQ might not be solely due to BDQ's influence on intracellular ATP. Next, rather than measuring ATP of the bulk culture, the authors used a radiometric FRET-based ATP biosensor to measure ATP levels in single cells so that they could sort low-ATP from high-ATP cells and showed that the former survive exposure to a combination of rifampicin and streptomycin better than the latter; of note, the difference in survival between the two populations is much lower than the survival difference between cultures with and without BDQ pre-treatment, suggesting that additional effects of BDQ contribute to the generation of persisters. The authors further used an acetate-based minimal medium to identify enzymes contributing to energy production under the hypothesis that variations in levels of expression of these enzymes might allow isogenic bacterial cells to have heterogeneous ATP levels. The authors used this medium supplemented with different concentrations of acetate to show that growth on lower acetate concentrations led to a higher number of cells with low ATP levels, better survival upon exposure to a combination of rifampicin and streptomycin, and a wider distribution of ATP levels. This led the authors to hypothesize that a wider distribution of ATP levels, encompassing ATP levels low enough to allow for persister formation, stems from greater variation, or noise, in gene expression. To prove their hypothesis, the authors measured transcript levels of genes involved in the conversion of acetate into acetyl-CoA in single cells with either low or high ATP levels and showed that cells with low ATP levels have lower transcript levels than cells with high ATP levels and that ackA transcript levels vary among cells with high ATP levels. This is the only conclusion that can be extracted from this experiment; the statement in the discussion (lines 210-211) that says "This shows considerable noise in the expression of the acetate kinase AckA in low ATP M. tuberculosis cells when acetate is the sole carbon source (Figure 4A)." is erroneous. To prevent variability in AckA expression, the authors overexpressed AckA, which led to a decrease in the number of persisters upon exposure to a combination of streptomycin and rifampicin; the authors concluded from this experiment that noise in AckA expression was responsible for the generation of cells with low ATP levels that survive antibiotic treatment. As is, the data provided do not support these conclusions; in Figure 4B, the authors should show the distribution of ATP levels in single cells as they did in Figure 2 and verify that the proportion of cells with low ATP levels is decreased upon overexpression of AckA. AckA overexpression should be demonstrated by Western blot or qRTPCR data. In the same vein, the authors should quantify the extent to which the CV decreases when AckA is overexpressed as in Figure 3. Finally, showing that knocking down ackA leads to an increase in cells with low ATP levels and better survival upon to exposure to antibiotics would strengthen the authors' conclusions.
-
Reviewer #3 (Public Review):
The manuscript by Quigley et al. is a technically-ambitious and potentially important study linking noisy metabolic gene expression with ATP levels and persister formation in M. tuberculosis. This work addresses a critical question - the mechanism of persister formation in this important pathogen. Using a reporter for intracellular ATP and single cell expression analysis, the authors conclude that stochastic expression of the acetate kinase gene correlates with ATP levels and persister formation during growth on acetate. They then test the importance of this variation in expression level, showing that constitutive ackA transcription reduces the abundance of persister cells. Overall, this is a compelling hypothesis that is consistent with work in other bacterial systems, and could lay the foundation for new …
Reviewer #3 (Public Review):
The manuscript by Quigley et al. is a technically-ambitious and potentially important study linking noisy metabolic gene expression with ATP levels and persister formation in M. tuberculosis. This work addresses a critical question - the mechanism of persister formation in this important pathogen. Using a reporter for intracellular ATP and single cell expression analysis, the authors conclude that stochastic expression of the acetate kinase gene correlates with ATP levels and persister formation during growth on acetate. They then test the importance of this variation in expression level, showing that constitutive ackA transcription reduces the abundance of persister cells. Overall, this is a compelling hypothesis that is consistent with work in other bacterial systems, and could lay the foundation for new therapeutic strategies. Acknowledging the technical challenge of studying small cellular subpopulations and the effort that has already gone into this work, there are a number of issues that should be addressed:
Based on the effect of preexposure to BDQ (Figure 1A and B) the authors conclude that "Tolerance of these mechanistically unrelated antibiotics shows that a decrease in ATP causes multidrug tolerance." This conclusion is consistent with the data, but the authors did not show that other bacteriostatic antibiotics don't have the same effect.
The relationship between growth and persister formation in Figure 3 seems overinterpreted. Since the growth curve ends at the same time that the ATP measurements are made (1 week) it is difficult to know what growth phase these cultures are in. As a result, statements like the following are hard to support: "The level of ATP dropped in the order 20 - 10 - 5 mM acetate (Figure 3C, D), showing that growth is not affected by relatively small changes in ATP", and "Notably, apart from linking noise to persister formation, this experiment shows that growth rate per se does not determine tolerance." As the growth curves end at the time of analysis, how do the authors know if the low acetate cultures are entering stationary phase?
The correlation between ATP CV and persister fraction in figure 3i is very weak and driven by a single point that does not seem to be representative. I don't see a biologically-driven correlation here, and the statistics agree.
The populations sorted in figure 2A differ in a number of respects. They differ in ATP concentration dependent FRET signal, as intended, but they also differ by more than 10-fold in constitutive YFP fluorescence. This suggests a difference in plasmid copy number, and the inverse correlation between YFP and ATP signals suggests a dependency. One could hypothesize that the high ATP cells have low plasmid copy number because they are replicating more rapidly. This is mainly an observation that the authors may consider commenting on, and is relevant for the next point. The conclusion that low ATP cells are enriched for persisters is sound.
The single cell qPCR study is a great approach, but there are technical issues that limit interpretation. The mRNA levels were normalized to the origin of replication of the ATeam plasmid. Based on YFP fluorescence in Figure 2A, the copy number of this plasmid appears to vary between single cells and is correlated with ATP levels (see point 5). It seems quite plausible that low ATP cells have a low ackA/plasmid ratio because plasmid copy number is increased.
The effect of ackA expression on antibiotic tolerance (figure 4D) are quite compelling. However, it is not clear if this effect is related to ATP levels. The product of ackA (acetyl phosphate) could have unanticipated ATP-independent effects on the cell if the enzyme is overexpressed.
-
