Towards a Better Control of Engineered Circuit Transcription in Bacterial Genomes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Report on the manuscript: Towards a better control of engineered circuit expression in bacterial genomes, by Nazim Sarica, Laurent Janniere and Brian C. Jester

 

Overall, I find this work intriguing and valuable. However, the writing could benefit from additional information and clearer explanations to better support the ideas and conclusions presented. Below, I outline both general and specific comments that should be addressed.

General comments:

As a general comment, I find the use of 'gene expression' as a synonym for 'transcription' problematic, as these terms are not equivalent. Throughout the text, the authors refer to 'gene expression,' but their analysis is limited to transcription. This is particularly notable given that they use fluorescence reporters, which could easily allow for a more comprehensive analysis of gene expression using methods such as flow cytometry or plate readers. The use of these methods to directly monitor gene expression – that can be done even across different growth phases, eliminates the need for time-consuming quantifications, which could also introduce errors

In lines 502 to 504, the authors argue that 'by quantitating the protein end-product, results could be misleading due to other factors that influence final protein levels (post-transcriptional regulation, translational and post-translational regulation, and maturation of the fluorophore)'.  However, these factors would equally affect all constructs analyzed under the same conditions. Therefore, it remains unclear why fluorescence levels were never measured in this work.

I noticed a lack of important details regarding the parts and constructs used in this work. Specifically, the type of terminators employed is not mentioned, and although the length of the loop formed by the CI protein is not explicitly stated in Figure 3, it is mentioned in the conclusion section as 2.5 kb. I assume this refers to the construct containing the two reporter genes and the GRS sequence, but an earlier clarification of this point would greatly aid in understanding the experiments. Additionally, there is no discussion on whether such a relatively small loop holds biological significance.

 

I would also suggest including a supplementary file listing the 209 strains harboring the emGFP reporter at different chromosomal positions, as this would greatly enhance the clarity and reproducibility of the study.

 

The axis labels in all figures are illegible; please revise them for better readability.

 

 

Specific comments:

Lines 285-289

What were the reasons for selecting these specific carbon sources? I could not find evidence in the cited references (9, 11) regarding their influence on gene expression. Moreover, reference 42, cited below in lines 296-297, does not use these carbon sources. According to that study, genetic context affects reporter gene expression similarly across different growth conditions. They tested rich medium with casamino acids, glucose, succinate, and acetate—carbon sources that enter distinct metabolic pathways. The results presented here appear to contradict those reported in reference 42. I feel an explanation or discussion is missing on this point—how is it possible that such seemingly similar carbon sources have such a significant effect on transcription?

Line 301 and thereafter:

Are the positions rationally curated to be non-mutagenic related to those identified in the transposition experiment? Why are they located in the ori and ter regions? Given that these regions are expected to exhibit differential expression due to incomplete replication events, I find the selection of these sites for insertion problematic.

Line 304-309

It seems that variability among different insertions under the same growth conditions decreases when mutagenic effects are eliminated, particularly when glucose is used as the carbon source. Furthermore, a visual inspection suggests that variability is lower in the ter region compared to the ori region, possibly due to replication effects. A comprehensive statistical analysis of the data groups is needed to confirm whether these differences are statistically significant.

Line 313-315

Which macrodomains were explored in this study? Providing an explanation of this would help improve the clarity and understanding of the work.

Line 330: Considering that supercoiled….

Lines 345-352

This statement does not seem accurate. There is a clear difference between insertions in the ori region and those in the ter region. In the ori region, expression variability decreases significantly, while in the ter region, the decrease is less noticeable. I am not convinced by this joint conclusion without a more detailed analysis that separates insertions within both regions and includes statistical tests to determine whether their behaviors differ significantly.

Additionally, there appears to be an inconsistency in the ter region in Figure 2. The red points marked with a star (no-loop condition, glycerol) should correspond to the values of the same insertions in Figure 1, but they do not. For example, the insertion in region 2,180,000 has a value of 0.03 TPC in Figure 1, whereas in Figure 2, it is between 0.01 and 0.02 (is the star pointing the right position?).

 

Line 369

“e.g. without or with a DNA loop, respectively” shouldn't it be the other way around?

 

Line 375

 

Higher than…

 

 

 

Lines 376-380

Although I generally agree with this conclusion, there are discrepancies between these results and previous ones with the weak promoter regarding the increase or decrease in expression depending on loop formation. These discrepancies should be addressed and discussed.

Line 417

The sentence is unclear; I suggest reformulating it as follows: “Under the control of the p3 promoter and flanked by the CI binding sites”.

Line 434

This seems to be an overestimation of the effect; at least in the case of the tandem configuration, the difference is less than double. Do you consider these differences significant? Have you performed a statistical analysis to assess this?

Lines 436-437

Is the relatively low increase in Cherry transcription from the p10 promoter compared to the p1 promoter consistent with their expected strengths? In the graph, it seems that Cherry transcription increases similarly across all configurations. However, the Y-axis scale makes it difficult to discern the smaller bars. I suggest adjusting the Y-axis (dividing it into two parts) to better represent these smaller values. Nonetheless, it seems clear to me that this conclusion is not accurate, as the smallest increase occurs in the tandem configuration, likely only around two-fold. I believe an appropriate statistical analysis of the data presented in Figure 4 is essential for a more accurate evaluation and comparison of its significance.

Line 481

What was the configuration of the construct used in this experiment? Was it tandem? Please specify this for a clearer comparison with previous experiments.

Lines 493-495

Regarding the drastically different configurations of the nucleoid in the growth conditions used in this work please include the corresponding references, as I could not find this information in the cited literature.

Line 514-515

“It is possible that the Ori/Ter ratio is clause to one” If I understand correctly, you are suggesting that at an OD of 0.5, the Ori/Ter ratio is close to one. I find this unlikely, as cells at this stage are in the mid-exponential growth phase.

Line 519-529

Based on visual inspection, I do not observe this pattern. In the ter region, the differences range from a six-fold increase (seen in two of the 16 insertions) to five-, four-, and three-fold increases in three other points. In contrast, in the ori region, the highest increase (out of eight insertions) exceeds eight-fold, with additional increases of four-, three-, and two-fold observed. In any case, a statistical analysis is necessary to confirm whether these differences are significant.

 

Have you examined the corresponding insertions within the ter region to determine whether the surrounding operons are associated with highly expressed genes?

Line 533

As mentioned before, I observe that strain-to-strain variability is more significantly reduced in the ori region. I believe this should be highlighted and discussed further.

Line 559

For mCherry gene we found….

Line 557-560

Although it is expected that the PBS generated in the convergent orientation would prevent the expression of both genes, the divergent configuration seems to perform worse for GFP and Cherry, overall. Could you comment on this? Once again, a statistical analysis would be very helpful in this case.

Line 561

Strongly stimulated?

Line 573

To systematically

Line 605

DNA loops illustrates how cues?

Line 621-624

While I believe this study could contribute to the design of synthetic gene circuits that are better isolated from context sensitivity, I also feel that the conclusions are somewhat overstated and mention broader goals that may not be directly related to this specific contribution. Please rephrase the applicability of this contribution in a more realistic and less ambitious manner.

 

 

Author Response

Thank you for your comments. Please see the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript by Sarica et al. investigates the molecular mechanism of context sensitivity in E. coli and demonstrates that artificial DNA loops surrounding transcriptional units can function as insulating elements to mitigate position-specific effects on gene transcription. The study is well-written, and the results are presented in a logical and coherent manner.

 

 

Minor comments

 

There seems to be an issue with the font in the figures.

 

I could not find the exact locations of the synthetic DNA loops in the Ori and Ter regions. Are the loops inserted within the transcribed region or in intergenic regions? If located within a transcribed gene, the loops could potentially be affected by transcriptional interference from upstream or downstream promoters. While the authors include a transcriptional terminator next to the CI binding site within the loop, it may also be beneficial to include terminators outside the loop. This is particularly relevant, as DNA-bound lambda CI is sensitive to dislodgement by elongating RNA polymerase (Hao et al., NAR, 2016).

 

Thus, it may be worth exploring other DNA loops, such as those formed by the LacI repressor, which is more resistant to dislodgement by RNA polymerase (Hao et al., NAR, 2014). While this is likely beyond the scope of the current study, such experiments could provide further insights into the efficiency of DNA loops in mitigating context sensitivity.

Author Response

Thank you for your comments. Please find the attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I appreciate the effort put into the revised manuscript, including the clarifications in the text and the additions to the methodology and other sections, which have significantly improved the work. However, I still notice readability issues in some figures, which I assume will be addressed during the final editing process