Visualization and functional dissection of coaxial paired SpoIIIE channels across the sporulation septum

Reviewer #1:

I have only a few minor points. These points only need to be addressed by text additions and figure presentation changes, and there is no need of additional experiment or analysis.

1) It is now known that not 100% of the photoactivatable fluorescent proteins are fluorescent or photoactivatable, but some molecules remain dark. Please discuss how this will affect the quantification of monomer numbers in the SpoIIIE complex.

Even though our qPALM approach balances the over- and undercounting errors generated due to the blinking of PA-FPs (Lee et al., 2012), it is correct that less than 100% labeling and/or improperly folded PA-FPs can introduce errors in the counting. The label efficiency is much less of an issue in PALM because the labels are genetically encoded, and we used western blots to confirm that 100% of the SpoIIIE molecules were fused to PA-FPs, Dendra2 or tdEOS (Supplementary file 2). However, it is possible to imagine that some of the PA-FP domains may not have folded properly and so may remain dark. At this point in time we do not know any method to correct for this effect or to measure the fraction of unfolded or otherwise non-fluorescent proteins. Accordingly, we have now modified the text as follows in the Discussion section:

“… Although our qPALM approach balances the over- and undercounting errors, (Lee et al, 2012), the numbers of SpoIIIE molecules determined here may represent a lower bound to their true number if some of the photoactivatable fusion domains did not fold properly at the foci. Nevertheless, …”

2) The authors explain that P_spoIIQand P_spoIIDare only activated after polar septation and only in a compartment specific fashion. While this property of these promoters is widely known, a reference to this point when these promoters are mentioned might help the general reader.

We have introduced the references below to address this point:

a) For a description of spoIIQ expression:

Londoño-Vallejo J, Fréhel C, Stragier P. 1997. spoIIQ, a forespore-expressed gene required for engulfment in Bacillus subtilis. Molecular Microbiology 24:29-39. doi:10.1046/j.1365-2958.1997.3181680.x.

b) spoIID expression was first characterized in the following papers:

Rong S, Rosenkrantz MS, Sonenshein AL. 1986. Transcriptional control of the Bacillus subtilis spoIID gene. Journal of Bacteriology 165:771-9.

Clarke S, Lopez-Diaz I, Mandelstam J. 1986. Use of lacZ gene fusions to determine the dependence pattern of the sporulation gene spoIID in spo mutants of Bacillus subtilis. Journal of General Microbiology 132:2987-94. doi:10.1099/00221287-132-11-2987.

c) Microscopy images of sporangia expressing GFP from spoIIQ or spoIID promoters can be found in Supplemental figure 2 of the following paper:

Sharp MD, Pogliano K. 2002. Role of cell-specific SpoIIIE assembly in polarity of DNA transfer. Science 295:137–9. doi:10.1126/science.1066274.

3) The black bars in Figure 3B and Figure 6–figure supplement 1B showing the average are not very clearly visible.

Thank you for pointing this out. We have lightened the background color and changed the style of the black bars so that they are more visible.

Reviewer #2:

1) The cell-specific degradation technique is powerful. But the proof-of-concept data shows a few snap shots sparsely operated in time. A more dense time series would be more convincing.

We agree with the reviewer that a more continuous time series would be ideal to study degradation dynamics. We have attempted to address this point by performing time-lapse microscopy of individual sporangia, but our results have been compromised by phototoxicity (to which sporulating cells appear particularly sensitive) and GFP photobleaching. In the absence of solid timelapse data, we were left with the option of classifying sporangia according to the different engulfment stages that can be unambiguously distinguished using membrane dyes. We here used the five stages of engulfment that are used for this purpose (Sharp and Pogliano, 1999), although we did not include the two final stages, since at this point, degradation is complete. The addition of intermediate stages based simply on visual screening would be subjective. Nevertheless, we are continuing to develop the cell-specific degradation system, and a key priority is to improve the temporal resolution.

2) What should I make of the 19±11 and 15±8 molecules in the mother cell- and forespore-proximal complexes? Do the authors believe the true number is 12 and 12? What sets the limits of this counting technique? If the authors could lower the error it would be quite powerful.

Two main sources set the limit of the counting technique, the counting error (or uncertainty) and the inherent variation of the biological structure composition. Unfortunately, we do not have any direct method to decouple the different sources of the dispersion in the results. However, here we will try to do a back-of-the-envelope calculation according to our previous work based on the in vitro characterization of Dendra2 photophysics, where we demonstrated that the counting error depended on the number of molecules and the length of the experiment (Figure S8, Lee et al. 2012). We estimate that the molecular counting error contributes 3 and 2 monomers for populations peaked at ∼19 and ∼15, respectively (Figure 5D center and right), which is smaller than the standard deviation of the measured distribution (11 and 8). Then, the remaining standard deviation (after eliminating the contribution of the counting error), 10 and 7, should correspond to the inherent variation of the biological structure. This suggests that most of the variability in the number of monomers is due to biological variation in the stoichiometry of the SpoIIIE complex rather than counting errors, and that it would not be possible to narrow the variance by just lowering the counting uncertainty.

It is very likely that SpoIIIE molecules that are not integrated into a functional channel co-exist with the channels on both subcomplexes. Note that the effect of these extra “free” monomers would be to overestimate the number of monomers in the channel and would tend to compensate the underestimation arising from the improper folding of the PAFP’s described above (response to the Reviewer’s comment 1). Nonetheless, the contribution of these two sources of stochastic variation will necessarily add when estimating the variance of the distribution. Thus, we have more confidence in the first than in the second moment of the distribution.