1. CFP and YFP have pH-dependent brightness. Acetate production acidifies the cytoplasm. In some areas this could modestly impact the analysis (e.g. rates of switching in the mother machine experiment). However most analysis is unitless (distribution CV, fraction some number of SDs above the mean) so we don't suspect this will affect any of the paper's conclusions.
  2. Concentration of oxygen and other important components. We had some concern that there may be concentration gradients (especially in the mother machine) or changes over time in the long gel-pad experiments because of evaporation, oxygen depletion, etc). It would be helpful to have some back-of-the-envelope calculations in the supplemental material to address this (e.g. entering parameters for growth chamber volume, initial cell number, growth rate, and oxygen metabolism rate, how long would it take to significantly change the oxygen concentration in a sealed chamber?).
  3. Old-cell bias. The mother machine oversamples the sub-population of cells that would otherwise be rare in an exponentially growing culture: those with "old" cell poles. In the SucC supplemental movie it appears that the oldest cells might be more likely to switch on; the authors could test for any cell-pole-age-dependence in switching or consider some type of experiment using a "baby cell machine". However, this isn't necessary to justify the claims made from the corresponding experiments.
  4. We were not 100% convinced that sucC+ cells are an acetate-producing subpopulation. High pta expression in sucC+ cells is good evidence but not definitive. Acetate production lags sucC production in Fig 1C/D. Given that the long dwell times of the normal and sucC+ states at OD600 0.8, it should be possible to assay acetate production rates of sorted cells. This isn't necessary to prove the major claims of the paper, but phrases such as "These results suggested the possibility that acetate predominantly produced by sucC+ cells..." could be modified to address this uncertainty.

Technical notes

  1. Fig. 1B / S1: (1) It appears Fig. 1B shows data from OD600 0.8 (sucC) and OD600 2.0 (alsS), and that data was collected at other OD600 values for these genes and for those shown in Fig. S1. It would be great to have all of the data available for people interested in checking out how expression varies over time for all of these genes. (2) The choice of binning for several histograms in Fig. S1 seems too large. (3) A supplementary table listing listing distribution statistics for all promoters and ODs tested would be helpful.
  2. Some small things that would help: (1) scale bars and time codes to figures and movies, (2) quantify error in estimating CV, fraction of the population 2 s.d. above the mean,  (3) double check the figure references in the main text and correspondence between captions and figures, (4) having captions inline with figures in the preprint PDF would help, (5) it's sometimes unclear what range "±" indicates.

Acknowledgements

Thank you to attendees of the ITQB Preprint journal club whose comments contributed to this review, and to the authors of this paper for such a great example for the first meeting of our journal club!