Alternative transcription start site selection in Mr-OPY2 controls lifestyle transitions in the fungus Metarhizium robertsii

Metarhizium robertsii is a versatile fungus with saprophytic, plant symbiotic and insect pathogenic lifestyle options. Here we show that M. robertsii mediates the saprophyte-to-insect pathogen transition through modulation of the expression of a membrane protein, Mr-OPY2. Abundant Mr-OPY2 protein initiates appressorium formation, a prerequisite for infection, whereas reduced production of Mr-OPY2 elicits saprophytic growth and conidiation. The precise regulation of Mr-OPY2 protein production is achieved via alternative transcription start sites. During saprophytic growth, a single long transcript is produced with small upstream open reading frames in its 5′ untranslated region. Increased production of Mr-OPY2 protein on host cuticle is achieved by expression of a transcript variant lacking a small upstream open reading frame that would otherwise inhibit translation of Mr-OPY2. RNA-seq and qRT-PCR analyses show that Mr-OPY2 is a negative regulator of a transcription factor that we demonstrate is necessary for appressorial formation. These findings provide insights into the mechanisms regulating fungal lifestyle transitions.

Understanding fungal pathogens that have complex lifestyles in different hosts is an interesting and relevant question. Here the authors explore a signaling protein identified in an insect pathogenic fungus. They found two surprising things about the signaling protein that were not known: 1) its translation is regulated by an upstream element, which impacts its function and 2) the signaling protein interacts with a newly identified transcription factor to uniquely modulate its activity. There are many problems and errors with this paper: namely clarity in terms of helping the reader to understand the findings and put them into context. Below follow some suggestions: In the abstract 'follow on analysis' is not clear There are numerous typos, particularly in the figure legends. Is Opy2's RA domain that binds to Ste50 is conserved? Figure 2B is not that clear. Maybe it should be bigger? Hard to see the hyphal swelling. Can the CalW staining intensity be quantitated? Can the number of septum be counted?
What do the uninfected caterpillars look like?
The cross sections in panel 2C are not clear. Figure 2C is not mentioned in the text until much later in the paper. Figure 2E is not clear. I can see fluorescence in one of the merged images but not the unmerged image (for WT hydrophobic). Figure 3D, control for protein loading should be put next to the figure (or a subset of the bands from the gel). Same problem for 4A, Figure 4C and all other blots where this is done. Figure 4B, given that the secondary structures are not relevant, they should be moved to the supplement. Figure 4A, there is a background level by immunoblot in the ΔMr-OPY2 strain. Was this subtracted for the normalization of the other samples? Figure 4A; it looks like the lower band is of higher intensity than the upper band. Why would that be expected to happen?
Were the transcript levels checked for the data provided in Fig. 5A?
Cells lacking Opy2 are not virulent (Figure 1). Cells expressing more Opy2 are less virulent ( Figure  5). This is confusing and not clearly explained or discussed in this section by the authors. What are different assays used in Figs. 1 and 5?
The rationale for the experiments in several of the results sections is not well introduced. Why look at salt sensitivity? This needs to be introduced better.
In yeast, opy2 mutants are not sensitive to osmotic stress because of a redundant branch of the HOG pathway (SSK branch). Is there an SSK branch in this fungus?
A control is missing from the two hybrid: only containing one (bait) or the other (prey) plasmid to show that auto-activation is not occurring.
Yeast Opy2 does not directly interact with MAPK cascade components.
I'm not sure their model accurately captures the data, especially see Figure 7.
Reviewer #2 (Remarks to the Author): Alternative transcription start site selection of the Mr-OPY2 gene controls the transition from saprophyte to pathogen in Metarhizium robertsii The paper presents evidence that alternative transcription may play a role in egulation of the membrane protein-encoding gene Mr-OPY2. The Mr-OPY2 protein is produced abundantly and is required for appressorium formation in the insect pathogenic fungus Metarhizium, which is essential for insect infection. By contrast, low level expression of the gene is associated with saprophytic growth of the fungus. The authors suggest that during saprophytic growth the fungus produces a single long transcript that contains 5'UTR microORFs that suppress translation of the larger downstream major ORF. On insect cuticle, a shorter transcript leads to production of Mr-OPY2 protein. This appears to act as a negative regulator of the AFTF1 transcription factor involved in regulating appressorium development.
The paper is interesting and is a reasonably comprehensive study of the potential means by which the Mr-OPY2 is regulated. However, what is less clear to me is whether this gene is actually a pivotal regulator of insect infection, or whether this form of regulation is more widespread within a broader range of genes associated with insect infection. This leaves this reviewer wondering whether the key finding is the mode of regulation (which is actually quite well known in filamentous fungi) in the context of insect infection, or if it is the actual regulation of Mr-OPY2 that is the major funding. Mr-OPY2 serves a similar function in Metarhizium to the counterpart in yeast apart from its means of regulation and the putative interaction with AFTF1. Given that I find it hard to distinguish between the two possibilities, coupled with the fact that there are many technical issues with the paper, I am not really convinced that it warrants publication in Nature Communications.
Major concerns 1. Figure 1. It is not clear what the key (0.7) relates to precisely. The figure legends suggests this relates to the estimated number of amino acid substitutions, but how does this actually relate to the tree. Similarly, the figures in the Table within Figure 1B are not clear to me (statistics, but which?). 2. Figure 2B. This is too small to really show differences in appressorium morphology and they are not at all clear from the Figure. No scale bar. Figure 2C also needs a scale bar in both panels. The differences in mycelial colonization in the lower panels are not at all clear and no indication of replication or quantification of this difference is provided. 3. Figure 3. A No size marker or context at all. No indication of replication of experiment or quantification. Figure 3D No size marker or indication of replication. Figurev3E completely inadequate to show any differences. None of the differences referred to in the text are visible in this figure. No scale bar. Too small to be of any use, not consistent with text description. 4. Figure 4A. No size marker in lower panel. No context, no quantification, no indication of replication of the experiment. Figure 4C. No size marker. No loading control. No evidence of replication. 5. Figure 4D is mentioned in the text. Page 7, line 151 stating that Mr-OPY2 is 1.4 fold higher in hemolymph than in SDY. This panel is not present in the Figure. 6. Figure 5A. No loading control or size marker (not much use giving this in the supplement). No evidence of reproducibility. Figure 6B. No scale bar. Morphological differences stated in text not clear. Statistical analysis of conidiation differences not provided. 7. Figure 6A needs scale bars. Not really clear at all. The micrograph are no clear and too small. Figure 6C has no loading control or size markers. Figure 6D is not at all clear as a Y2H analysis. Where is dilution and high stringency versus medium/low stringency result. Inadequate control. 8. Figure 7. The differences in phosphorylation are key to the conclusions of the paper and yet this figure is really very poor. No indication of reproducibility is provided. No quantification of differences is provided apart from the 'ratios' which are not explained. This needs to be analyzed properly with replicates, given the critical nature of the data presented. 9. Supplemental Figures. I find the 'YES' 'NO' panels for each of the gels presented to be confusing/inadequate to illustrate the cloning strategies. No size markers on any of the gels shown also compounds the lack of clarity. No size markers provided on any of the protein gels.
For this reviewers, the lack of rigor in the experimental results presented are a real problem in interpreting the data. I cannot really decide whether the finding is novel and a 'new insight into fungal lifestyle transitions' given the poor quality of the data presented. I think they need to evidence the reproducibility of their findings and also provide evidence that Mr-OPY2 is the key gene in this context that they claim.

Point-to-point response to reviewer's comments
The reviewer's comments are in plain text and our responses are in bold.

Reviewer #1 (Remarks to the Author):
Understanding fungal pathogens that have complex lifestyles in different hosts is an interesting and relevant question. Here the authors explore a signaling protein identified in an insect pathogenic fungus. They found two surprising things about the signaling protein that were not known: 1) its translation is regulated by an upstream element, which impacts its function and 2) the signaling protein interacts with a newly identified transcription factor to uniquely modulate its activity. There are many problems and errors with this paper: namely clarity in terms of helping the reader to understand the findings and put them into context.

Response: To improve the clarity, we have made substantial changes to the text, figures and their legends.
Below follow some suggestions:

In the abstract 'follow on analysis' is not clear
Response: "follow-on analysis" mainly involved identification of the transcription factor AFTF1 and functional characterization of AFTF1 using gene disruption and qRT-PCR. We clarified this in the revised manuscript.
2. There are numerous typos, particularly in the figure legends.

Is
Opy2's RA domain that binds to Ste50 is conserved?

Response: Yes, we found CR-A and CR-D domains in Mr-OPY2, and this
information is provided in the revised manuscript.
6. Figure  The cross sections in panel 2C are not clear. Figure 2C is not mentioned in the text until much later in the paper. 8. Figure 2E is not clear. I can see fluorescence in one of the merged images but not the unmerged image (for WT hydrophobic). Figure 3E in the original version. We prepared a new figure (now figure 2) that is bigger than the original one.

Response: I assume the reviewer means
However, a point of this figure is that fluorescence will be weak when the WT is grown in 1/2SDY because the level of Mr-OPY2 protein is low as in 2E. In the nature of these experiments it is harder to see fluorescence in WT when it is grown in 1/2 SDY.
In the original picture (Fig.3E), the length of germlings grown in 1/2SDYwas shorter than the appressorium-forming germling. We redid the Histoimmunochemical staining experiments with hyphae grown for longer in 1/2SDY so that their length is similar to appressorium-forming germlings.
We found no difference in the fluorescent intensity of long hyphae (in the new images) and short hyphae (in the images in the first submission).
9. Figure 3D, control for protein loading should be put next to the figure (or a subset of the bands from the gel). Same problem for 4A, Figure 4C and all other blots where this is done.

Response: We moved all loading control gels (a subset of bands including the target protein) to be aligned in figs with their respective blots.
10. Figure 4B, given that the secondary structures are not relevant, they should be moved to the supplement.
Response: Agreed-moved to the supplement.
11. Figure 4A, there is a background level by immunoblot in the ΔMr-OPY2 strain.
Was this subtracted for the normalization of the other samples?
Response: No, the ΔMr-OPY2 strain was used as a negative control. WT was set to 1, and other samples except the ΔMr-OPY2 strain were relative to WT. This is mentioned in the revised legends.
12. Figure 4A; it looks like the lower band is of higher intensity than the upper band.
Why would that be expected to happen? This is now discussed in the discussion section.

Response
13. Were the transcript levels checked for the data provided in Fig. 5A?
Response: The transcript levels were checked with qRT-PCR and results are now provided in the revised manuscript.
14. Cells lacking Opy2 are not virulent (Figure 1). Cells expressing more Opy2 are less virulent ( Figure 5). This is confusing and not clearly explained or discussed in this section by the authors. What are different assays used in Figs. 1 and 5?
Response: Additional discussion has been added to clarify this point. which is why they are integrated in a single paper, and we see no reason to rank their importance. We think the reason for the reviewer's uncertainty about the novelty of this paper is that they judged our findings separately in different contexts, based on which the reviewer proposed their two "possibilities".
For the first "possibility", the reviewer thought that if the key finding is the regulation of translation by uORFs, then the paper is not novel because this regulatory mechanism is "quite well known in filamentous fungi". However, we do not say that the mode of regulation that applies to Mr-OPY2 is the major novel finding. How regulation of Mr-OPY2 effects Metarhizium's ability to infect insects is the thrust of this paper. 2. Figure 2B. This is too small to really show differences in appressorium morphology and they are not at all clear from the Figure. No scale bar. Figure 2C also needs a scale bar in both panels. The differences in mycelial colonization in the lower panels are not at all clear and no indication of replication or quantification of this difference is provided. 4. Figure 4A. No size marker in lower panel. No context, no quantification, no indication of replication of the experiment. Figure 4C. No size marker. No loading control. No evidence of replication.

Response
Response: The comments about figures have been addressed in our response above.
The loading control is moved from the supplement to the main figure.
5. Figure 4D is mentioned in the text. Page 7, line 151 stating that Mr-OPY2 is 1.4 fold higher in hemolymph than in SDY. This panel is not present in the Figure. Response: Apologies. We were referring to Figure 3D (now Figure 2D in the revised manuscript).
6. Figure 5A. No loading control or size marker (not much use giving this in the supplement). No evidence of reproducibility. Figure 6B. No scale bar.
Morphological differences stated in text not clear. Statistical analysis of conidiation differences not provided.

Response: The comments about figures have been addressed in our response
above.
Statistical analysis of conidiation differences is provided.
7. Figure 6A needs scale bars. Not really clear at all. The micrograph are no clear and too small. Figure 6C has no loading control or size markers. Figure 6D is not at all clear as a Y2H analysis. Where is dilution and high stringency versus medium/low stringency result. Inadequate control.

Response: The comments about figures have been addressed in our response
above.
The loading control is moved from the supplementary file to the main figure.
The yeast two hybrid assays have been repeated with more controls added. This was done under high stringency.
8. Figure 7. The differences in phosphorylation are key to the conclusions of the paper and yet this figure is really very poor. No indication of reproducibility is provided. No quantification of differences is provided apart from the 'ratios' which are not explained. This needs to be analyzed properly with replicates, given the critical nature of the data presented. For this reviewers, the lack of rigor in the experimental results presented are a real problem in interpreting the data. I cannot really decide whether the finding is novel and a 'new insight into fungal lifestyle transitions' given the poor quality of the data presented. I think they need to evidence the reproducibility of their findings and also provide evidence that Mr-OPY2 is the key gene in this context that they claim.

section.
Four lines of evidences confirm that Mr-OPY2 is a pivotal regulator of insect infection by M. robertsii. First, the deletion mutant of Mr-OPY2 was completely unable to infect insects. Secondly, we demonstrate that Mr-OPY2 protein levels are regulated by alternative transcription start site selection and manipulation of this process changes virulence. Thirdly, RNA-Seq analysis revealed that Mr-OPY2 regulates known virulence genes including cuticle degrading proteases. Finally, the transcription factor AFTF1 is regulated by Mr-OPY2, representing a new circuit regulating fungal infection.