Write a comment

Peer Review Report 

“Agrochemical-responsive gene expression control systems for modulating plant development and metabolism” 

Plant biologists and agricultural engineers need better tools to modulate gene expression amounts, locations, and timing. To modulate the timing of gene expression, researchers can use chemically-inducible systems. The inducers for these systems must be inexpensive, field-safe, and stable, and these systems should have the capacity to regulate both transgenes and native loci. Working to meet these criteria, the authors in the current study present two agrochemical-inducible systems. Both of these systems are based on fusions to the ecdysone receptor (EcR), which is induced by a widely used agrochemical called methoxyfenozide (MF). One of the systems fuses EcR to Cas9 for regulation of selected native loci, and the other system fuses EcR to Cre recombinase for regulation of desired transgenes. The authors performed initial validation and some optimization of these inducible systems in transient assays before characterizing the functionality of the systems in full transgenic plants. They were able to show moderate activation of fluorescent reporters after watering with MF, as well as modulation of metabolic flux and plant development.

The engineering tools and approaches presented in this study provide an important early demonstration of agrochemical-responsive perturbations to metabolic flux and plant development. It was exciting to see multi-component pathway engineering in plants, all under the control of the same field-safe inducer. Furthermore, it was interesting to see this paper’s demonstration of nuclear export sequences for reducing leak in plant inducible systems. While the transient infiltration data well supported the main claims at the beginning of the paper, some of the stable transgenics data would become more robust by screening more than 2 independent lines. Additional optimization work is required to increase the fold change of these systems and further lower leak. However, this is a promising initial demonstration of field-safe inducible tools for modulating metabolism and development in plants.

Major Points

1. In figure 1, panel b, the authors show a positive control with NLS-Cas9 (not inducible), but in panel e, the authors have no positive control. The authors could consider 1) a pre-flipped promoter with the Lox sites around the UBQ10 promoter in the forward direction, or 2) an NLS- Cre (not inducible) like they did for the Cas9-based system. This additional control would help add evidence to support the claims made later in the paper about binary responses of the Cre-based system (lines 330-335). If the positive control for the Cre system has a similar normalized reporter fluorescence to the EcR-Cre+MF, then it would suggest this system can more effectively ‘turn on’ or reach a maximum activation state, in a way that the Cas9 system cannot.

2. In the luminescence section, the authors found that 1 of 3 lines gave a 16.6% fold change with and without guides (figure 5, panel g) and that all 3 lines displayed significantly increased luminescence between uninduced and induced (figure S10A). The exact sample size isn’t reported, but it appears high. In order to support claims about these systems redirecting metabolic flux in a statistically significant and meaningful way, the authors could consider looking at effect size and measuring Cohen’s d. If they can show the effect size is large, then the main claims will be much stronger.

3. The authors state in the development section that “untreated DELLA activation lines were smaller than the untreated control plants” and that this is likely due to random T-DNA integration instead of leak because the “absolute levels of GAI and RGA were not significantly elevated in the untreated DELLA activation lines compared to the untreated no gRNA control lines (lines 677- 682). It would be very helpful if the authors could clarify whether the data in these 2 plots are from the same qPCR run and can be directly compared. Additional independent lines could also help strengthen this claim, as it is unlikely that every T-DNA integration location across 3-5+ independent lines would cause growth defects. Even small amounts of developmental regulators can vastly affect plant growth, so it is possible the system is leaky enough to decrease rosette size but not leaky enough to drastically alter detectable RNA levels.

Minor Points

1. The authors discuss how initial soil experiments in Arabidopsis were successful, but repeated assays using the same stable lines didn’t show the same results (lines 464-466). It is unclear if these subsequent experiments were performed with seeds from the same generation (T2, T3, etc), or if the initial experiments were done in T2, and the next experiments were done in T3. If the seeds were from the same generation, how much time passed between the initial and follow-up experiments? These details would help improve the clarity of the paper.

2. In figure 6, panel E, it appears the order of the leaves was flipped horizontally for the MF + treatment. In the MF - treatment and in supplementary figure S13 panel C, all the leaves go from youngest to oldest. Maintaining consistency of leaf age across figure panels make it easier for readers to understand.

3. For transient infiltrations in N. benthamiana, the authors watered half of the plants with MF one day after infiltrating and then quantified fluorescence from all the plants 3 days after infiltrating (lines 286-287). Clarification on why the authors decided to induce for 48 instead of 24 hours could be useful, especially if the authors tried 24 hours first or have any time-course data on induction. This could also increase the impact of the paper even more by helping the field better understand MF-based activation dynamics.

4. Throughout the paper, the Kernal density estimation (KDE) plots are used to visualize distributions and make claims about things like percent reduction in leak (figure 1 C and F, figure 2 B and D, figure 5 F, figure S10 B). All the plots have density on the y axis and normalized reporter fluorescence on the x axis, except for the plots in figure 1, which have flipped axes. Consistency throughout could improve clarity for the readers.

5. In panel D of figure 2, the unactivated reporter control is missing from the KDE plot. Including the control would help with transparency and clarity.

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.

I am one of 2 participants from the same research lab that reviewed this preprint as part of the Transparent and Accountable Peer Review training program. This report was written with the support of my research mentor.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.