Cell cycle dependent variation in endocytosis drives phenotypic diversity in M. tuberculosis

Reviewer #1 (Public review):

Summary:

The manuscript- "Cell cycle-dependent variation in endocytosis drives phenotypic diversity in M. tuberculosis" by Subhash et al. demonstrates how host cell heterogeneity shapes intracellular pathogen phenotypes. The central and novel finding of this study (G2-phase cells have higher endocytic capacity and harbour more oxidised Mtb) highlights that a host cell cycle (interphase-driven) changes in endocytic capacity regulate bacterial redox states.

Strengths:

Overall, the study is well-executed and conceptually rich, establishing a causal link between host cell cycle progression, endocytic heterogeneity, and M. tuberculosis phenotypic diversity.

The combination of multiple modalities, including live-cell imaging, flow cytometry, scRNA-seq, and redox-sensitive bacterial reporters, supports these findings and substantially strengthens the biological relevance of the work.

The writing is generally clear, and the figures are well-organised.

This work will be of interest to readers across cell biology, microbiology, and infection biology

Weaknesses:

However, several central claims are only partially supported, the mechanistic depth is limited, and several experimental and analytical concerns need to be addressed.

Major Comments:

(1) The authors demonstrate a correlation between the G2 phase and elevated endocytic capacity. However, the mechanistic link (upstream molecular mechanism) between the cell cycle and endocytic upregulation remains largely unaddressed. The authors speculate that membrane biogenesis during volumetric expansion may drive increased endocytosis and note that lipid biosynthesis genes are upregulated in high-endocytic cells. It would substantially strengthen the paper to test this directly, by examining whether inhibition of lipid biosynthesis (e.g., with fatostatin or cerulenin) selectively reduces the G2-associated increase in endocytic capacity. Alternatively, cyclin-CDK axis perturbations (e.g., CDK1 inhibition with RO-3306 to specifically block G2/M entry) could be used to ask whether cells arrested in G2 maintain elevated endocytosis, helping distinguish cell-cycle-position-dependent from cell-cycle-progression-dependent effects.

(2) The current data show a clear association between high endocytic capacity and more oxidised Mtb, and the authors (consistent with their prior work) hint at lysosomal delivery as the likely mechanism. However, direct evidence for this in the current paper is limited. An experiment examining phagosomal pH or lysosomal fusion (e.g., using a pH-sensitive reporter or lysotracker) specifically in high- and low-endocytic-capacity cells after infection would help confirm this.

(3) Temporal resolution of Mtb redox dynamics. The plasticity experiment (Figure 6C-D) is elegant and shows that Mtb redox states revert as host cells divide and daughters enter G1. However, the experiment compares day 0 and day 3 post-sorting, which spans multiple cell divisions. While a finer time resolution (spanning 24h) would establish the causal relationship, the authors could discuss the possibility and consequences of multiple cell divisions between day 0 and day 3 used in the present study.

(4) Relevance of G2 percentages in differentiated macrophages. In Figure 7 and Supplementary Figure S7, only 4.4-5.7% of THP-1-derived macrophages and 5.7% of BMDMs are in G2. While the authors demonstrate statistically significant differences in Mtb redox states between G1 and G2 macrophages, the biological significance of such a small G2 fraction in a non-dividing population deserves discussion specifically with respect to: a) Are these cells re-entering the cycle? b) Is the G2 designation capturing a distinct functional state rather than active cycling? The authors should include additional markers (e.g., phospho-histone H3 for mitotic cells or BrdU incorporation to test for active S-phase) to characterise this population and clarify its identity and origin in differentiated macrophages, thereby meaningfully informing interpretation.

In conclusion, this is an important mechanism-driven study that highlights an important link in host-driven bacterial phenotypic heterogeneity. The experiments are thorough, the model is well-supported, and the study has implications for infection biology.

Reviewer #2 (Public review):

In this manuscript, the authors utilize a combination of techniques to show that macrophage endocytic capacity is partially dictated by cell-cycle stage, that Mycobacterium tuberculosis (Mtb) more readily infects macrophages that are in G2/M -phases, and that bacteria that are internalized by macrophages at different stages of the cell-cycle experience different levels of intracellular stress (as reported by the redox state of the bacteria). Furthermore, the authors provide evidence that terminally differentiated macrophages retain memory of the cell-cycle stage that they were in prior to differentiation, at least in the context of endocytic capacity.

This work provides evidence for the growing idea that fundamental heterogeneity in both host and bacterial organisms can alter the host-pathogen relationship in important ways. However, based on the current data, I am not convinced that the manuscript establishes endocytic capacity as the causal link between macrophage cell-cycle stage and bacterial state. The main issue is that fluorescence-based sorting for cell-cycle stage is likely to covary with cell size. Larger cells, including those later in the cell cycle, may be more likely to fall into the "high" fluorescence gate, while smaller cells may be enriched in the "low" population. Therefore, the observed phenotypes may still be cell-cycle-associated, but the causal determinant could be a correlated feature of cell-cycle progression rather than endocytic capacity itself. This is a significant caveat because nearly all the data, including the live-cell imaging following individual cells, rely on 'total' fluorescence, which will scale strongly with cell size.

If the authors' conclusion that endocytic capacity is cell-cycle regulated holds true after appropriate controls, this would significantly advance our understanding of the causal interplay between host cell-cycle state, endocytosis, and Mtb physiology. However, an alternative interpretation is that the observed differences in Mtb uptake and bacterial redox state are associated with cell-cycle stage but are not caused directly by differences in endocytic capacity. For example, they could instead reflect other cell-cycle-linked changes in macrophage physiology, such as cell size, intracellular volume, metabolic state, or some other mechanism important for Mtb pathogenesis. If the authors find that their data are best explained by cell-cycle stage independent of endocytic capacity, this would still represent an important advance. However, in that case, the manuscript should clearly distinguish the association with cell-cycle state from the downstream effector mechanisms, which would remain to be determined.

Strengths:

The authors utilize various macrophage models for their studies, which is important considering the variability in macrophage behavior, as well as the growing evidence that differences between mouse and human macrophages are relevant for Mtb infection.

Weaknesses:

The most important caveat is the covariance between fluorescence-based reporters and cell size. This concern applies to both the sorting experiments, which directly measure total fluorescence, and the time-lapse microscopy experiments, in which the authors show total fluorescence rather than mean, area-normalized fluorescence in Figure 3C. This could be explained by biomass accumulation alone, rather than by a specific cell-cycle-dependent increase in endocytic capacity. Without distinguishing total signal from concentration or activity per unit cell area/volume, it is difficult to conclude that endocytosis itself is regulated by cell-cycle stage rather than simply scaling with cell size.

Although the authors provide some evidence that the mean GFP intensity, which more closely reflects concentration, differs between the sorted populations in Figure 3B, they do not report statistics for this comparison. Moreover, this control is not carried through the rest of the manuscript, including in key experiments such as Figure 2B. As a result, it remains difficult to determine whether the observed differences between "high" and "low" populations reflect cell-cycle state specifically or instead reflect differences in total reporter fluorescence driven entirely by cell size.

The evidence for cell-cycle-dependent effects would be more convincing if the authors included additional controls. For example, they could:

(1) Plot both mean GFP intensity and total GFP intensity in Figure 3B, ideally alongside an unrelated fluorescent reporter that does not vary across the cell cycle. This would help distinguish changes in reporter concentration from changes driven by cell size or total fluorescence.

(2) Sort cells based on an unrelated fluorescent marker and test whether the same phenotypes - infectivity, dextran uptake, bacterial redox state, etc. - differ between high- and low-fluorescence populations. If these phenotypes are specific to the cell-cycle reporter and not observed with an unrelated marker, this would strengthen the conclusion that the effects are linked to cell-cycle state rather than to fluorescence intensity, cell size, or sorting artifacts.