Escribir un comentario

Manuscript Title: “Fungal volatiles drive lifestyle-dependent, systemic metabolic reprogramming in poplar”

Summary: In this paper, the authors provide novel evidence of systemic metabolic response of Populus x canescens to fungal volatile organic compounds. The primary research question was how plants respond to fungi of different ecological lifestyles—H. annosum (parasitic), L. bicolor (mutualistic), and P. placenta (saprotropic)—without physical contact, based only on their volatile organic compounds (VOCs). The experiment was continued for six weeks, during which volatilomic data was collected regularly and presented in a separate preprint. At the conclusion of this study, the authors collected leaves and roots for analysis. The authors sought to understand whether plants could sense and produce organized metabolic responses to fungi of different guilds. They position their study as a test of chemical communication display (CCD) in relation to plant-fungi interactions, a framework that claims that chemicals exuded into the environment can encode biologically useful information. This metabolomic analysis suggests that pathogen-associated fungal VOCs (H. annosum) may have suppression-dominant effects on poplar metabolism, while saprotropic fungi (P. placenta) promote lipid-centered metabolic adjustments, and mutualistic fungi trigger restrained metabolic response overall in compatibility-oriented programming over long time scales.

General Assessment: Overall, the study’s pot-in-pot system was methodologically clever and utilized rigorous analyses to investigate a novel question on systemic responses of plants to fungi of different guilds. The results are significant in demonstrating the potential role of fungal VOCs in triggering metabolic responses within plants prior to physical contact. Although prior studies have analyzed responses to single VOCs or at shorter time scales, this paper represents the first long-term study of VOCs from fungal guilds on poplar response. The methods were generally well-written and replicable. However, the study is limited by their single harvest endpoint. We also worry that the design may have permitted crosstalk of volatiles from the different fungal treatments as well as between plants because, as seen in figure 1, the top of the soil was open, and the plants were grown in close proximity to one another, which can potentially allow the escape of VOCs and could have a significant impact on poplar responses. The authors’ claims on lifestyle-specific signal response by plants in this experiment are not fully supported. Providing a single example species per plant is not sufficient evidence that these plant metabolic responses are due to fungal guilds. Although they address this as a limitation in their discussion, being more clear throughout the paper that the individual species are not fully generalizable to the guild would be useful. The concept of CCDs was also not clearly described in this paper, leaving the readers confused about how CCD relates to the metabolomics results shown.

Major concerns: The data from this experiment and Zhu et al. 2025, which measures and analyzes VOCs from this same experiment, are presented as separate papers. However, we believe the VOC component of this study is essential context and validation for the experimental setup. The relationship between the two papers is mentioned but ultimately glossed over. Instead, we believe the authors should integrate the two papers for a more cohesive and fully supported analysis. If this is not feasible, the authors could increase cross-discussion between the two papers to help support this preprint. In lines 75-84, which introduce the VOC preprint, this information should be emphasized and moved closer to the end of the introduction to increase clarity. Another option could be a combined figure that maps specific VOC chemical classes to the specific metabolic pathways that they are hypothesized to trigger, potentially expanding on the information given in table S1. By the same token, the ‘Term’ column in table S1 is not intuitive as it doesn’t give enough information to where the data is from, specifically the row labeled as ‘Treatment.’ While the Generalized Linear Models (GLMs) show that “SQT class” (sesquiterpenoids) predicts leaf responses, the table doesn’t specify which metabolic pathways those sesquiterpenoids are driving.

Second, a major revision of the experimental setup, or a tempering of the central claim, is needed. Most significantly, the use of a single fungal species per lifestyle category seems insufficient for claims of lifestyle-specific cue response in plants. At least three species per category, with strong taxonomic coverage, are necessary to support this claim. Fundamentally, it is impossible to distinguish if poplar was responding to a volatile bouquet of H. annosum or to a pathogenic lifestyle. Also, some aspects of the physical experiment setup are unclear. How was the flow of VOCs between pots or trays prevented, if at all? If this was not done, and treatments were kept in close proximity, which seems to be the case from Fig. 1A, then the authors cannot conclusively show that each plant was exclusively sensing VOCs from its associated fungal plate and not from aboveground VOCs.

Additionally, in the summary section, the key results list the discoveries as “pathogen VOCs produced a suppression-dominated systemic phenotype; saprotroph VOCs promoted lipid-centered metabolic activation; and mutualist VOCs elicited restrained, compatibility-consistent shifts with targeted pathway modulation,” but changing this to a less strongly worded claim would better match the data in this paper.

Finally, we are concerned that the study did not adequately test CCD. Given that only one observation was made after fungal VOCs had interacted with plants for weeks, we are unconvinced that the study provides strong evidence for CCD as contrasted with a more parsimonious view (e.g. that only some chemicals provide strong signals to the plant and that metabolic pathways are indifferent to ratios between them). An improvement here could be a Linear Discriminant Analysis of VOCs as predictors of metabolite response of the host, which would show how individual VOCs load metabolomic response.

Minor Concerns:

L47: In the introduction, the authors fail to explain the importance of VOCs in fungal signaling. There is rich literature available on this subject, but the authors only briefly explain them. It would significantly support the paper if the introduction provided more justification for the importance of VOCs in fungal signaling and plant responses to fungi. Specifically, details on the chemical makeup and distribution of VOCs are needed.

L51-52: “Reviews have highlighted their ecological roles as mediators of interspecific interactions” – This is vague, we think it would connect better to the introduction if the species involved in these interspecific interactions were mentioned.

L135: The methods state that trays were rotated every three days, but it is unclear in what manner they were rotated, or why. Otherwise, the controls for the setup seem strong.

L134, L148 – In the “pot-in-pot exposure system section,” n = 6 plants per treatment was mentioned, but in the “sample collection and metabolite extraction” section, the sample size for fungi treatments was n = 4 and for the control was n = 3. Please clarify why 2-3 replicates were excluded from the final LC-MS analysis.

L149: Should be “we checked weekly…” instead of “we weekly checked…"

L310, figure 3e: “...clusters, including a clear module of jointly up-regulated lipids.” – There are equally as many upregulated secondary metabolites as lipids in this cluster, and previously mentioned that the leaf metabolic networks were enriched in secondary metabolites and lipids. Please clarify if there is a reason secondary metabolites were not mentioned in the analysis.

L331: The text mentions ceramides and phosphatidic acids, but these aren’t listed on fig. 4. Further, how were these functional groups chosen to display?

L335-336: Would be clearer with listed examples of lipids that show P. placenta-specific accumulation.

Fi 1: Pot-in-pot is defined as PiP, but ‘PiP’ is only used once more, while ‘pot-in-pot’ appears 8 times.

Fig 1A: The photograph of the experimental setup doesn’t necessarily add anything to the paper that the diagram next to it already shows. This photo would be best placed in the paper’s supplement.

Fig 2A, Fig. 2D: The colors for the treatments should be in the same order on the figure and in the legend.

Fig 4A: Different color palettes should be used for functional groups and KEGG pathways. It’s currently a bit confusing to decipher as is.

Fig. 5: The graphic on the left is a little unclear as is. Right now, it seems to imply that plants are responding to a blend of volatiles from different fungi simultaneously. While this certainly happens in nature, it’s outside the scope of this paper to discuss this. Additionally, the mention of CCD could be clearer. We would suggest showing multiple plants each being exposed to different fungi. Then, you can visualize the different blends of compounds between treatments by showing different ratios of colored dots flowing to the plant, which would better emphasize the idea of broad chemical cues impacting plant responses.

Recommendation: Accept with major revisions

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.