Orthogonal regulation of phytochrome B abundance by stress-specific plastidial retrograde signaling metabolite

Plant survival necessitates constant monitoring of fluctuating light and balancing growth demands with adaptive responses, tasks mediated via interconnected sensing and signaling networks. Photoreceptor phytochrome B (phyB) and plastidial retrograde signaling metabolite methylerythritol cyclodiphosphate (MEcPP) are evolutionarily conserved sensing and signaling components eliciting responses through unknown connection(s). Here, via a suppressor screen, we identify two phyB mutant alleles that revert the dwarf and high salicylic acid phenotypes of the high MEcPP containing mutant ceh1. Biochemical analyses show high phyB protein levels in MEcPP-accumulating plants resulting from reduced expression of phyB antagonists and decreased auxin levels. We show that auxin treatment negatively regulates phyB abundance. Additional studies identify CAMTA3, a MEcPP-activated calcium-dependent transcriptional regulator, as critical for maintaining phyB abundance. These studies provide insights into biological organization fundamentals whereby a signal from a single plastidial metabolite is transduced into an ensemble of regulatory networks controlling the abundance of phyB, positioning plastids at the information apex directing adaptive responses.

MEcPP might be affecting phyB levels. Collectively, they propose a model in which the previously described alteration of auxin and CAMTA3 levels in ceh, and the newly found reduced PIF transcript levels, all regulate phyB levels via independent pathways. This is a timely and exciting finding that deepens our understanding of how stress-induced retrograde signaling impacts growth. However, in my opinion, the work would benefit from more direct evidence to support how phyB levels might be altered. Below are some comments and suggestions that might help improve it: -The role of CAMTA3 and auxin in the regulation of phyB abundance is not investigated. The possibility that CAMTA3 acts at the proteasome level requires further exploration. Do CAMTA3 and auxin act independently from each other? Are they affecting PIF abundance or function? Although addressing these questions in detail might be the subject of future work, adding some more mechanistic insight would support the model and significantly add to the manuscript. -The reduction in PIF transcript levels is quite small and might not be relevant at the protein level. I wonder if this really has an impact on phyB levels in ceh1.
-The effect of externally applied auxin suggests that altered levels of auxin in ceh1 impact phyB levels. Genetic evidence would be necessary to provide support for this finding.
-The transcriptomic analysis of camta3 suggests that there might be overlap with the phyB and PIF regulated gene network, and with hormone regulated pathways. It would be interesting to include comparisons with PIF and auxin regulated transcriptome that might suggest connections that are not explored now. Other points: -The reduction in SA levels in cehphyB is striking and would be interesting to discuss. -Fig 2a, 2b, 4a, 4e: The light conditions and sampling are not detailed. Are they all in LD? What time were samples collected? -Please describe "orthogonal regulation" in the context of this work.
-References need to be corrected. Citation of at least a few is incorrect and some relevant ones seem to be missing.
Reviewer #3 (Remarks to the Author): It is known, that chloroplast retrograde and light signaling use common components, and that the chloroplast impacts on light-induced seedling development. In this paper, Jiang et al. try to disentangle the interconnections of the chloroplast retrograde signaling metabolite MEcPP and photomorphogenesis. In the current study under review, a suppressor screen of ceh1 recovered two phyB mutant alleles which revert the ceh1 dwarf phenotype. Subsequent experiments show that the ceh1 mutant contains higher phyB protein levels compared to the WT. Thus it was hypothesized that the dwarf phenotype might be suppressed by lowering phyB protein levels in the ceh1 mutant. Indeed, the ceh1 dwarf phenotype is partially reverted, when the phyB levels are suppressed by (i) introducing oePIF4 or oePIF5 into ceh1, (ii) treating ceh1 with auxin, (iii) crossing ceh1 with the camta3 mutant. Moreover, very moderate phyB protein repression was achieved by (iv) treating ceh1 with an Ca2+ chelator or a Ca2+ channel blocker. More specific comments: Fig. 4: Overexpression of PIF4 and PIF5. Small differences of transcript accumulation measured in a real-time PCR experiment are not strongly supporting a downregulation. Moreover, the phyB-PIF interaction feedback-loop is leading to PIF PROTEIN degradation, PIF protein levels should be definitely measured here. Panel e. It is known that overexpression of PIF4 or PIF5 reduces phyB protein level. Thus emphasis should be on the extent of phyB protein level reduction in the ceh1 mutant. Numbers on phyB levels should be implemented in the figure. Fig. 5: Which auxin concentration was used? Fig. 6. RNA-Seq analysis. "Detailed analyses of the data determined that the expression of ~90% of robustly repressed genes in the ceh1 are CAMTA3-dependent, and are significantly overrepresented by genes in lightand auxin-response pathways (Fig. 6ab)." This is not a "detailed" analysis. The identified GOs are interesting. Which photosynthesis genes are down-regulated? Why are those GOs, which include responses to light representative for the DOWN-regulated genes since phyB protein level are up in the ceh1 mutant? How is the intersection with published phyB gain-of-function transcript data? The regulation of the RSRE cis element is CAMTA3-dependent and was identified in induced genes in the ceh1 mutant (Benn et al., PNAS 2016). Could you please comment also on the UP-regulated genes for the current RNA-Seq data? Scale for the GO analysis: Log10 p-value: are the values in minus? Otherwise the p-values would be very high. Fig. 7: "In addition, we have shown partial size recovery of ceh1 dwarf stature in the ceh1/camta3 double mutant lines." The authors cite their literature. However, it is sometimes not see whether results shown in this manuscript were already produced in former publications. Clearly, Fig. 7a was already produced in Benn et al. and it should be more obvious.
"The data illustrate compromised growth phenotypes in seedlings treated with these two agents most notably in P plants." This result is not meaningful for the establishment of a connection between phyB protein abundance and growth behaviour of ceh1. The ceh1 mutant does not look better under these treatments than under normal growth conditions. "The result collectively supports regulatory function of CAMTA3 in stabilizing phyB protein levels in a Ca2+-dependent manner." This is not very clearly demonstrated because the reduction of phyB in the calcium experiment is only by 20%. Overall: -The title should be re-written, because it is misleading -The language of the abstract and some other passages is of poor quality.
-Statistical significance testing should be indicated in the figures.
-One common theme of the ceh1/phyB, ceh1/oePIF and ceh1/camta3 mutants is the reduction of salicylic acid levels compared to ceh1. The performance of a ceh1 mutant in which the salicylicacid pathway was interrupted should be investigated. With the ceh1/eds16 mutant (Bjornson et al., Plant J 2017) the authors already have one suitable double mutant at hand.
-Further investigation of ceh1/phyB, ceh1/oe-PIF, ceh1/camta3 seedlings and the addition of Ca2+ to ceh1 is needed: a) The introduction section has one focus on SAR and the title is about photomorphogenesis. To underpin the statement "whereby a specific plastidial signal such as MEcPP is transduced into discrete biological responses that collectively regulate photomorphogenesis", it should be investigated at least whether in addition to the dwarf phenotype of the ceh1 mutant the hypocotyllength phenotype of ceh1 is reverted? How is flowering time altered?
b) In the model and the discussion section it is proposed that the "function of CAMTA3 is likely via impairment of proteasome-mediated degradation of the light-activated photoreceptor proteasome COP1-mediated proteolysis of phyB, achieved either through direct targeting of a component(s) of COP1 machinery and/or indirectly via targeting carbon status." This would be an important new key finding which should be definitely further investigated in this manuscript. In ceh1, phyB protein levels in the light stay as high as in darkness which is reminiscent of the cop1 mutant, indicating that proteasome-mediated degradation of phyB in ceh1 might indeed not be functional. The phyB Western blots of the ceh1 and ceh1/camta3 mutants and the auxin treated plants should also be shown in the presence of the proteasome inhibitor MG132.
The postulated phyB degradation by CAMTA3 should also be investigated by an in vitro assay similar to that shown in Figure 6, Jang et al., Plant Cell 2010 Minor comments: The Nam-Hai Chua group should be cited with their Jang et al, Plant Cell 2010, manuscript in which phyB polyubiquitination by COP1 in the nucleus is described.
Dear reviewers to address your collective concerns we have performed almost all of the additional experiments that are pertinent and within the scope of the current work. We truly believe that compliance with your collective inputs and suggestions has improved the overall quality of the manuscript and has brought to attention the central theme not previously discussed.
It should be noted that because of the substantial alterations, the changes are not highlighted in the text.
The specific response to each of your concerns are as follows:

Reviewer #1
1. The suppressor screen is not described anywhere in the manuscript. These details should be included in the methods section of the manuscript.
This information is now included in the method section. Fig. 7a, and similarly Fig. 1a and 4b need scale bars in order for claims about stunting reversal to be convincing.

A scale bar is provided in
Scale bar is now provided in all the plant images (Figs. 1 a-b, 4b, 5a, 6a, 7a, and c).
3. Instead of referring to "robustly" repressed or down-regulated genes, the actual fold-change in expression should be stated. This is relevant for page 8 of the main text and Fig. 6.
We have resolved this issue throughout the manuscript.
4. The text of the manuscript has to be proofread and edited for small grammatical errors that make it harder to read the manuscript. One example from page 5 is "These data collectively verify revertants as phyB mutant alleles and establish selected function of phyB in MEcPPmediated signal transduction pathways." What does "selected function of phyB" mean?
We have edited the entire manuscript and rewritten some sections including the one this reviewer is referring to. The text now reads: "The data collectively establishes phyB as a component of MEcPP signal transduction pathway involved in regulation of growth and SA level, but not in the induction of HPL expression.

The role of CAMTA3 and auxin in the regulation of phyB abundance is not investigated. The possibility that CAMTA3 acts at the proteasome level requires further exploration.
We have established the likely function of CAMTA3 in impairing the degradation machinery by examining phyB protein levels in mock and bortezomib treated seedlings (Figs. 8a-c & S5a-c).

Do CAMTA3 and auxin act independently from each other?
We had previously established that CAMTA3 does not alter auxin levels (Benn et al, PNAS 2016). Here we show the opposing functions of the two; CAMTA3 is stabilizing PhyB protein (Fig. 7a) whereas auxin reduces the levels (Figs. 5b and 6b).

Are they affecting PIF abundance or function?
We have not yet explored this possibility, but we will in future.

The reduction in PIF transcript levels is quite small and might not be relevant at the protein level. I wonder if this really has an impact on phyB levels in ceh1.
We agree with the reviewer's comment, however our genetic approach using PIF over-expressers has confirmed the role of PIFs in regulating phyB protein abundance in the ceh1 mutant background.

The effect of externally applied auxin suggests that altered levels of auxin in ceh1 impact phyB levels. Genetic evidence would be necessary to provide support for this finding.
To address this concern we have generated ceh1/tir1 double mutant line and grown them along with their respective parent backgrounds in the presence and absence of auxin. The results clearly show negative regulatory function of auxin in regulating phyB abundance (Fig. 6a-b) 6. The transcriptomic analysis of camta3 suggests that there might be overlap with the phyB and PIF regulated gene network, and with hormone regulated pathways. It would be interesting to include comparisons with PIF and auxin regulated transcriptome that might suggest connections that are not explored now.
We have performed bioinformatics analyses and presented the Vann diagrams along with the list of overlapping genes between ceh1, CAMTA3, PIF, IAA treated, and constitutive active allele of PhyB (YHB) (Figs. 3a-e and Table. S1-5).
Other points: -The reduction in SA levels in ceh/phyB is striking and would be interesting to discuss.
We have briefly discussed it in discussion section.
- Fig 2a, 2b, 4a, 4e: The light conditions and sampling are not detailed. Are they all in LD? What time were samples collected?
The pertinent information is now included.
MEcPP-mediates multifaceted and intersecting regulatory networks that regulate the abundance of the red light photoreceptor, phyB, via auxin signalling cascade and Ca 2+ -dependent CAMTA3-based transcriptional network.
-References need to be corrected. Citation of at least a few is incorrect and some relevant ones seem to be missing.
We apologies for this embarrassing oversight. We were not aware of the glitch in our Endnote system. The references are now corrected. Thanks for pointing out this important error.

Reviewer #3
1. Fig. 4 Our several attempts to examine PIF protein levels in ceh1 has proven unsuccessful. We agree with the reviewer's comments also shared by the reviewer #2. As such, we resorted to a genetic approach using ceh1/PIF overexpressers to overcome this shortcoming. We have included number on phyB immunoblots (Figs. 4e, 5b, 6b, 8a-b and S5a-b).

Fig. 5: Which auxin concentration was used?
We apologize for this omission, the concentrations are now included in the corresponding figure legends.
3. Fig. 6. RNA-Seq analysis. Detailed analyses of the data determined that the expression of ~90% of robustly repressed genes in the ceh1 are CAMTA3-dependent, and are significantly overrepresented by genes in light-and auxin-response pathways (Fig. 6ab). This is not a "detailed" analysis. The identified GOs are interesting. Which photosynthesis genes are down-regulated? Why are those GOs, which include responses to light representative for the DOWN-regulated genes since phyB protein level are up in the ceh1 mutant? How is the intersection with published phyB gain-of-function transcript data? The regulation of the RSRE cis element is CAMTA3-dependent and was identified in induced genes in the ceh1 mutant (Benn et al., PNAS 2016). Could you please comment also on the UPregulated genes for the current RNA-Seq data? Scale for the GO analysis: Log10 p-value: are the values in minus? Otherwise the p-values would be very high.
We have fully complied with the reviewer's suggestion, altered the text and included additional bioinformatics in supplemental data (Fig.s3a-e and Table S1-5). Fig. 7: "In addition, we have shown partial size recovery of ceh1 dwarf stature in the ceh1/camta3 double mutant lines." The authors cite their literature. However, it is sometimes not see whether results shown in this manuscript were already produced in former publications. Clearly, Fig. 7a was already produced in Benn et al. and it should be more obvious.

4.
The images shown here were not from the PNAS paper 2016, but the reconfirmation of the data by repeating the experiments. We have now made it fully clear in the text.

"
The data illustrate compromised growth phenotypes in seedlings treated with these two agents most notably in P plants." This result is not meaningful for the establishment of a connection between phyB protein abundance and growth behaviour of ceh1. The ceh1 mutant does not look better under these treatments than under normal growth conditions.
Fig 7c shows relative growth of P versus ceh1 plant in the presence and absence of LaCl3. In our view the image clearly shows that the agent negatively impacts the growth of P plants more notably than that of the ceh1. Hence, we believe that this image supports our claim.

"The result collectively supports regulatory function of CAMTA3 in stabilizing phyB protein levels in a Ca2+-dependent manner."
This is not very clearly demonstrated because the reduction of phyB in the calcium experiment is only by 20%.
We have repeated the experiment with a higher LaCl 3 concentration. The current result shows clear reduction in phyB protein levels (Fig. 7d). Fig. 8 We have complied with the reviewer's suggestion, see Fig. 9.

7.
Overall: -The title should be re-written, because it is misleading As suggested we have modified the title and replaced photomorphogenesis with phyB The language of the abstract and some other passages is of poor quality.
The abstract and many passages are rewritten.

-Statistical significance testing should be indicated in the figures.
We have complied, please see fig legends.
-One common theme of the ceh1/phyB, ceh1/oePIF and ceh1/camta3 mutants is the reduction of salicylic acid levels compared to ceh1. The performance of a ceh1 mutant in which the salicylicacid pathway was interrupted should be investigated. With the ceh1/eds16 mutant (Bjornson et al., Plant J 2017) the authors already have one suitable double mutant at hand.
This data was already included, please see Fig. S2a.
-Further investigation of ceh1/phyB, ceh1/oe-PIF, ceh1/camta3 seedlings and the addition of Ca2+ to ceh1 is needed: a) The introduction section has one focus on SAR and the title is about photomorphogenesis. To underpin the statement "whereby a specific plastidial signal such as MEcPP is transduced into discrete biological responses that collectively regulate photomorphogenesis", it should be investigated at least whether in addition to the dwarf phenotype of the ceh1 mutant the hypocotyl-length phenotype of ceh1 is reverted? How is flowering time altered?
We have now included images and measurements of the hypocotyl length (see Fig. 1b and S1a). Lastly, we disagree with the reviewer's comment concerning adding Ca 2+ to ceh1. We don't see any value in performing this experiment, nor the purpose and the conclusion one can draw from it.
The phyB Western blots of the ceh1 and ceh1/camta3 mutants and the auxin treated plants should also be shown in the presence of the proteasome inhibitor MG132.
We have complied and established the likely function of CAMTA3 in impairing the degradation machinery by examining phyB protein levels in mock and bortezomib treated seedlings (Figs. 8a-c & S5a-c).
The postulated phyB degradation by CAMTA3 should also be investigated by an in vitro assay similar to that shown in Figure 6, Jang et al., Plant Cell 2010. We disagree with this suggestion, since an in vitro assay is outside the current scope of this work and not necessary.

Minor comments:
The Nam-Hai Chua group should be cited with their Jang et al, Plant Cell 2010, manuscript in which phyB polyubiquitination by COP1 in the nucleus is described.
We agree and now we have included this important reference, sorry for the oversight.