Write a comment

Review of the preprint “Drug-induced differential culturability in diverse strains of Mycobacterium tuberculosis”

Authors: Valerie F. A. March, Nino Maghradze, Kakha Mchedlishvili, Teona Avaliani, Rusudan Aspindzelashvili, Zaza Avaliani, Maia Kipiani, Nestani Tukvadze, Levan Jugheli, Selim Bouaouina, Anna Doetsch, Galo A. Goig, Sebastien Gagneuxand Sonia Borrell

Posted on: bioRxiv

• URL: https://www.biorxiv.org/content/10.1101/2024.08.05.606579v1.full

• Supplementary material: https://www.biorxiv.org/content/10.1101/2024.08.05.606579v1.supplementary-material 

Brief summary of the study

This study investigates the phenomenon of differential culturability in Mycobacterium tuberculosis (Mtb) and shows that Mtb is more frequently recovered in liquid cultures than on solid media. These “differentially culturable tubercle bacteria” (DCTB) are caused by factors such as immune responses, hypoxia, starvation, low temperatures, and antibiotic treatment, the latter being the main focus of the study. The study shows with in vitro experiments by treating Mtb with the antibiotics isoniazid (INH) and rifampicin (RIF) that these drugs can cause DCTB.

The concept is relevant given that tuberculosis is a significant public health issue, and findings could contribute to more effective treatment and control strategies.

Major comments (validity or strength of the methodology, experiments and analyses, strength of the conclusions)

• Having controls without antibiotics for all experiments would be good for consistency 

• The reader would benefit from more explanations on the metabolic assay of choice

• Several crowd reviewers mentioned that the statistical analysis might need to be performed and explained more carefully. Some concerns were raised in certain images (see also below: miscellaneous comments)

• Some information about the patient cohort might be needed, e.g. underlying conditions such as presence of co-infections of chronic diseases

• “Time to positivity” needs to be explained early in the text

• The study seems to lack a strong conclusion, for example in the form of future experiments or in the form of possible recommendations to practitioners diagnosing tuberculosis by culturing methods. However, we also note that deductions made for application in clinical practice require careful interpretation

• The adopted method of using clinical samples and conducting in vitro experiments with different Mtb strains is appropriate. The conclusion also presented evidence to support that exposure to TB drugs affects the culturability of Mtb on solid medium, thereby influencing treatment monitoring and drug-pathogen interaction studies. 

• The use of more specific controls, especially in the in vitro experiments, may help isolate the effects of each drug more clearly.

Minor comments (clarifications to statements in the text, interpretation of the results, presentation of the data/figures)

• Details, background, and references related to the hypothesis might benefit from some elaboration

• It would be better to clarify if the study deals only with pulmonary TB

• The term MDR might benefit from a (operational) definition

• The authors might want to explain in the introduction the standard treatments for MDR TB vs non MDR TB

• The results were presented in a clear manner, although additional considerations regarding the specific clinical samples used could offer more explanation into the diversity of strains and the reproducibility of the study

• The authors need to provide clarifications on some statements, like how antibiotic-induced stress affects metabolic pathways in Mtb 

Comments on reporting (information on the statistical analyses or availability of data)

• Since two groups are being tested  with or without control - the authors would need to clarify this and in places where the two groups are being compared against control too, p value adjustment needs to be done

• The authors mention normality in the data, but it would be good if they could show it

• A more detailed description of the statistical methods used (e.g., specific tests for comparing means) would strengthen the reporting

Suggestions for future studies

• Future investigations could focus on the underlying molecular or genetic mechanisms that lead to the differential culturability of Mtb under the influence of TB drugs.

• Investigating the long-term effects of differential culturability on treatment outcomes in diverse populations, including those with co-infections (e.g., HIV), might also offer lasting solutions for TB management

• It could be interesting to perform transcriptomic and metabolic analysis of DCTB isolated from TB patients before and during treatment

• Studying a wider range of drug-resistant Mtb strains will provide more information on the phenomenon

• A mechanistic exploration and an in-depth discussion of the underlying molecular mechanisms

• Explore potential roles of specific genes or pathways involved in antibiotic stress response and persistence. 

• Alternative culture methods like the use of different culture conditions or media to assess culturability of Mtb under various stress conditions might be interesting

Miscellaneous comments

• We recommend that the authors say “ATP levels” instead of “ATP expression”

• Including more detailed statistical analysis in the figure captions would be beneficial

• In general, the description of the statistical analyses can be improved

• The figures can use a little more clarity with respect to the label/time normalization

• Fig 1(a) & 1(b) seemed to be switched. According to the legend, (b) is from n=35 drug-susceptible patients, and (a) from n=99 MDR patients. Shouldn't numbers be higher in (a) than in (b)?

• According to supplementary material, the range of drug concentration is 5X-160X MIC (and not 5X-180X MIC)

• Probably Figs. S1 & S2 need to be switched as S1 shows average values of S2?

• In Fig. 2, is the p value corrected for multiple testing? What test was used? From the figure 2B p=0.0076 is hard to believe. Was there a post-hoc test done?

• Why there is a “no drug” control in Fig. 2 and Fig. 4?

• The inclusion of additional control groups, such as untreated bacteria or bacteria treated with other antibiotics, could provide valuable comparative data

• In the “Metabolic viability assay” chapter, describing the characteristics of metabolic viability assay could be helpful for context. And: Why was RIF not used in the experiment?

• It is noted that the statement “all treatments led to some increase in the fold-change ATP expression compared to untreated controls” is a bit vague. Adding information such a number of data points, fold-increases, and statistical test could help the text and the readers!

• At one point, the author employ Kruskal Wallis test - there is a sufficient number of data points (>30) per group, so it's not clear why the authors choose a non-parametric test as opposed to a t-test

• How do you define metabolic dysregulation?

• In some (supplementary) figures, there seem to be only weak linear correlation

• Is fluorescence indicating bacterial load, or can it also indicate bacterial activity/viability?

• Antibiotic mechanisms of action can be elaborated briefly before the result and discussion. A more detailed explanation of the mechanisms of action of the selected antibiotics (INH and RIF) could strengthen the rationale for investigating their impact on culturability

• Shouldn’t “mid-log phase” be defined more precisely than “OD600 nm 0.4-0.8”?

• The results could be strengthened by providing more detailed statistical analysis, including effect sizes and confidence interval

• describe which test was used to test normality. If normal distribution was not seen, which non-parametric test was used and why?

Conflicts of interest of reviewers

• There are no conflicts of interest to declare.

Competing interests

The authors declare that they have no competing interests.