Molecular dynamics simulations illuminate the role of sequence context in the ELF3-PrD-based temperature sensing mechanism in plants

Reviewer #1 (Public review):

Summary:

This manuscript explores the role of the Evening Complex (EC), specifically focusing on ELF3, a disordered protein component of the EC, and its temperature-dependent phase behavior. The study highlights the role of polyQ tracts in modulating temperature-sensitive condensate formation and provides a combination of computational approaches, including REST2 simulations and coarse-grained Martini simulations, to investigate how polyQ tract length and sequence context influence this behavior.

Strengths:

The study addresses a key question in plant biology - how temperature influences circadian clock-mediated growth regulation through protein phase behavior. The manuscript introduces the novel finding that polyQ tract length modulates the temperature-dependent formation of helices and condensates.

Weaknesses:

(1) Coarse-Grained Simulation Results Not Supported by Data:

The results presented in Figure 6A of the manuscript do not seem to show a clear trend in the number of clusters formed as a function of polyQ tract length. This is particularly evident in the comparison between 0Q and 7Q polyQ lengths, which display statistically similar values in terms of the number of clusters. The lack of distinction between these values raises questions about the sensitivity of the coarse-grained simulations to polyQ tract length, which the authors claim as a key modulator of condensate formation. This discrepancy weakens the argument that polyQ length directly impacts the clustering behavior in the simulations.

Suggested Analysis:

a) A more detailed statistical analysis should be performed to assess whether the observed differences between polyQ lengths are significant. This could involve hypothesis testing or the use of error bars in the graphs to better communicate the variability in the data.

b) Additionally, the authors should examine whether there are other features, such as cluster shape or internal structure, that might differentiate between different polyQ lengths, even if the total number of clusters is similar.

(2) Inconsistency in Cluster Size Across Temperatures (Figure 6B):

The results in Figure 6B show a striking difference in the size of the largest cluster between temperatures of 290K and 300K. This abrupt shift in behavior lacks a clear mechanistic explanation. Typically, phase transitions driven by temperature are more gradual, unless there is some underlying structural or chemical shift that the authors have not accounted for. Without a clear explanation, this sudden change in behavior reduces confidence in the simulation results.

Suggested Analysis:

a) The authors should explore possible explanations for the dramatic difference in cluster size between 290K and 300K. For example, they could investigate whether specific interactions (such as the breaking or formation of hydrogen bonds or hydrophobic contacts) might explain the behavior at higher temperatures.

b) It is important to check whether the coarse-grained simulation model has been adequately parameterized and scaled for accurate temperature dependence. Atomistic simulations of monomers and dimers with varying polyQ tract lengths could be used to fine-tune the coarse-grained model, ensuring it accurately reflects molecular behavior. The gross estimate of a 10% scaling factor might be insufficient and could lead to inaccurate representations of cluster formation.

(3) Scaling of Coarse-Grained Model with Atomistic Simulations:

As mentioned, the coarse-grained model used in the study may not have been properly scaled against atomistic data. A simple scaling factor of 10% may not be appropriate for accurately capturing the behavior of polyQ tracts across different lengths, especially considering their sensitivity to subtle changes in temperature. Without rigorous validation against atomistic simulations, the coarse-grained model's predictions could be skewed.

Suggested Analysis:

a) To address this, the authors should compare the coarse-grained model with atomistic simulations of monomeric and dimeric forms of ELF3 with different polyQ tract lengths. By comparing key structural parameters (e.g., radius of gyration, contact maps, and clustering propensity), the authors could adjust the coarse-grained model to more accurately reflect the atomistic behavior. The authors have wealth of atomistic simulation data that could afford such benchmarking and identification of scaling factor

b) Additionally, the authors should investigate whether the assumed scaling factor of 10% is appropriate for each polyQ length or whether it needs to be refined based on specific properties, such as the number of hydrophobic interactions or secondary structure stability.

(4) Lack of Analysis for Liquid-Like Behavior in Phase Separation:

The simulations presented in the manuscript do not analyze the liquid-like behavior of ELF3 condensates, which is a key characteristic of liquid-liquid phase separation (LLPS). In LLPS systems, condensates are often dynamic, with chains exchanging between clusters, indicating liquid-like rather than solid-like behavior. The authors fail to probe this crucial aspect, which is necessary to support the claim that ELF3 undergoes phase separation.

Suggested Analysis:

a) The authors should conduct additional analyses to probe the liquid-like nature of the clusters formed by ELF3. One approach would be to analyze the dynamics of chain exchange between clusters, measuring how frequently chains leave one cluster and join another over time. This analysis would reveal whether the condensates behave as liquid-like, dynamic structures or more static, solid-like aggregates.

b) Additionally, the temperature dependence of these exchange dynamics should be investigated. In true liquid-liquid phase separation, the rate of chain exchange is often sensitive to temperature. Observing how this rate changes between 290K and 300K, for instance, could help explain the abrupt shift in cluster size seen in Figure 6B.

c) The authors should also analyze whether the internal structures of the condensates are consistent with a liquid-like phase. For example, radial distribution functions and contact lifetimes could be calculated to reveal whether the clusters exhibit liquid-like organization.

(5) Lack of justification of polydispersity of polyQ:

The authors don't provide any rationale for choice of different copies of polyQ used in the manuscript for their chain-growth simulation studies. It will be more apt if it can be motivated via some precedent experimental observations.

(6) Lack of initiative to connect to Experiments:

While the computational models and simulations provide robust theoretical insights, the absence of direct experimental validation weakens the overall impact of the manuscript. For example, experimental data on how specific mutations in the polyQ tract influence ELF3 behavior in vivo would significantly bolster the authors' claims. The manuscript would benefit from either citing existing experimental studies that corroborate these findings or from suggesting future experimental directions.

Comments on revised version:

The authors have now adequately addressed to the key concerns of manuscript. The manuscript in the present form looks significantly improved.

Reviewer #2 (Public review):

Summary:

The authors investigate how ELF3, a disordered scaffolding protein in the plant circadian Evening Complex, responds to temperature by forming reversible nuclear condensates. They focus on the C-terminal prion-like domain and on a variable polyglutamine tract within it, asking how the tract length and surrounding sequence context tune temperature-responsive structural and condensation behavior. Using a tiered set of computational approaches, including sequence heuristics, hierarchical chain-growth ensembles, all-atom enhanced-sampling simulations, and coarse-grained condensate simulations of 100 monomers, they characterize wild-type, polyQ deletion, polyQ expansion, and an aromatic-disrupting F527A variant. In the revised manuscript, the central claim has been reframed so that polyQ length is now described as tuning condensate material properties rather than driving temperature-sensitive phase separation, with temperature-responsive condensation attributed primarily to a sticker-rich aromatic contact network.

Strengths:

The biological question is important and timely, and the multiscale computational strategy provides a fresh view of an intrinsically disordered protein and its variants. The all-atom enhanced sampling analyses identify a temperature-dependent long-range aromatic contact involving F527 and a methionine-tyrosine coordination motif, which are concrete and mechanistically interesting observations beyond what coarse-grained or sequence-only methods could provide. In response to the previous round of review the authors have added replicate averaged statistics with error bars on the new condensate analyses, introduced new dynamics observables including effective diffusivity, an anomalous diffusion exponent, the self van Hove function, shape anisotropy, per chain radius of gyration in the condensed phase, and a condensate lifetime, provided cluster size time series for transparency, justified the choice of polyQ tract lengths against published Arabidopsis polymorphisms, expanded the Methods with explicit formulas for the new analyses, and included a split half convergence check for the all atom ensembles. The reframing toward a sticker spacer interpretation is consistent with recent experimental work and represents a more cautious and defensible reading of the data.

Weaknesses:

Despite these substantive additions, several core concerns from the previous review remain only partially addressed, and, on close reading, the new supplementary analyses do not robustly support the reframed claim that polyQ length tunes condensate material properties. Error bars and replicate-averaged statistics were added to the new condensate panels, but the helical propensity and per-residue analyses throughout the rest of the manuscript still show only a single curve per temperature, so variability for these key observables remains unreported. Several of the newly added dynamics observables show that the variants are essentially indistinguishable within the reported uncertainty: the self van Hove distributions, the shape anisotropy distributions, and the per chain radius of gyration distributions in the condensed phase overlap almost entirely across variants, and the anomalous diffusion exponent has between replica spreads at low temperature that exceed the variant to variant differences, with variant orderings that change with temperature. The variant-dependent signal that does survive, namely a drop in condensate lifetime for the polyQ expansion and the aromatic mutant at the highest temperature studied, rests on a single temperature point, with replicate spreads spanning most of the metric's dynamic range.

The cluster size time series at higher temperatures shows the dominant cluster oscillating over a wide range across replicas, indicating intermittent dissolution and incomplete convergence in the very temperature regime where the variant-specific claims are made. The only convergence test provided is a split-half radius-of-gyration analysis for the all-atom ensembles, with no slab-geometry or coexistence-density check for the coarse-grained condensate simulations. The polyQ deletion variant forms dominant clusters comparable in size to wild type at low and intermediate temperatures, which on its own argues that variable polyQ presence is not a primary determinant of clustering and supports the earlier concern that the temperature sensitive behavior is dominated by generic chain length and aromatic sticker effects rather than polyQ specific sequence effects, a concern that the reframing softens but does not resolve. Statistical significance is not assessed anywhere, and with three replicas and largely overlapping error bars, claims of variant-specific differences would benefit from explicit statistical tests. Minor quality control issues are also visible in the supplementary material, including a mislabeling of the aromatic mutant in two analysis panels and an inconsistent trajectory length for one variant at one temperature.

Additional Context for Readers:

Readers should interpret the molecular mechanism proposed here with caution. The reframing from polyQ length driving temperature-sensitive phase separation to polyQ length tuning of condensate material properties is more scientifically measured and aligns with recent experimental work, but several of the supplementary observables introduced to support this revised claim indicate that the variants studied are statistically indistinguishable within the reported replicate uncertainty. The most robust observation in the revised work is that the prion-like domain undergoes a temperature-responsive break of an aromatic contact in all-atom simulations and that aromatic sticker contacts dominate inter-protein interactions in coarse-grained condensate simulations. The mechanistic role of the polyQ tract, beyond generic chain length and hydration effects, remains, as in the original submission, not clearly established by the simulations presented. Independent experimental validation of the proposed aromatic contact and of the predicted material-state differences between polyQ variants will be needed to establish the molecular mechanism, and improved condensate convergence tests, uniformly reported error bars across all simulation-derived figures, and explicit statistical tests of variant-versus-variant differences would substantially strengthen confidence in the conclusions.

Author response:

The following is the authors’ response to the original reviews.

eLife Assessment

In this potentially valuable computational study, the authors conducted atomistic and coarsegrained simulations to probe the temperature-dependent phase behaviors of ELF3, a disordered component of the evening complex in plant. The results aim to highlight the role of polyQ tracts in modulating the temperature sensitivity. The level of evidence is considered incomplete, due to the lack of systematic calibration of the coarse-grained model and limited statistical uncertainty analysis, especially considering the relatively subtle nature of the differences due to temperature change.

We agree that the subtle temperature dependence of ELF3-PrD condensation requires rigorous uncertainty reporting and careful interpretation of CG predictions. In the revised manuscript we therefore (i) report mean ± SEM across independent replicas for all CG observables and provide full time series in the Supplementary Information, and (ii) expand our CG analysis beyond cluster counting to include condensate stability (size), lifetime, internal mobility (D, α), dynamic heterogeneity (van Hove), and structural descriptors (anisotropy, singlechain compaction/density). These additions strengthen the robustness of the conclusions and even enable physical explanations of recent experimental measurements on ELF3-PrD condensates.

Public Reviews:

Reviewer #1 (Public review):

Summary:

This manuscript explores the role of the Evening Complex (EC), specifically focusing on ELF3, a disordered protein component of the EC, and its temperature-dependent phase behavior. The study highlights the role of polyQ tracts in modulating temperature-sensitive condensate formation and provides a combination of computational approaches, including REST2 simulations and coarse-grained Martini simulations, to investigate how polyQ tract length and sequence context influence this behavior.

Strengths:

The study addresses a key question in plant biology - how temperature influences circadian clock-mediated growth regulation through protein phase behavior. The manuscript introduces the novel finding that polyQ tract length modulates the temperature-dependent formation of helices and condensates.

Weaknesses:

(1) Coarse-Grained Simulation Results Not Supported by Data:

The results presented in Figure 6A of the manuscript do not seem to show a clear trend in the number of clusters formed as a function of polyQ tract length. This is particularly evident in the comparison between 0Q and 7Q polyQ lengths, which display statistically similar values in terms of the number of clusters. The lack of distinction between these values raises questions about the sensitivity of the coarse-grained simulations to polyQ tract length, which the authors claim as a key modulator of condensate formation. This discrepancy weakens the argument that polyQ length directly impacts the clustering behavior in the simulations.

Suggested Analysis:

A more detailed statistical analysis should be performed to assess whether the observed differences between polyQ lengths are significant. This could involve hypothesis testing or the use of error bars in the graphs to better communicate the variability in the data.

Additionally, the authors should examine whether there are other features, such as cluster shape or internal structure, that might differentiate between different polyQ lengths, even if the total number of clusters is similar.

We agree that the number of clusters in Fig. 6A does not show a strong or monotonic dependence on polyQ length (e.g., 0Q vs 7Q can overlap within uncertainty). The cluster number is highly sensitive to coarsening kinetics and rapidly approaches a late-time plateau, and therefore is not our primary discriminator of variant-dependent condensation behavior.

To address the reviewer’s request for statistical rigor and additional differentiating features, we have revised the analysis in two ways. First, we now report mean ± SEM across independent replicas for all key CG observables and provide full replicate time series in the Supplementary Information to make variability and convergence/coarsening explicit.

Second, we shift our main CG conclusions away from “cluster number” and toward more diagnostic observables of condensate robustness and material state, including: (i) stability via the late-time mean largest-cluster size, (ii) persistence/lifetime via the fraction of frames with largest cluster size greater than 50, (iii) internal dynamics via MSD-derived DDD and anomalous exponent ααα, (iv) dynamic heterogeneity via self van Hove distributions relative to a Gaussian reference, and (v) morphology/internal structure via κ² and Rg distributions.

Notably, the κ²/Rg distributions are broadly overlapping at 300 K, indicating that in our system variant differences are expressed more strongly in stability/persistence and internal dynamics (D/α/van Hove) than in a large shift in single-chain compaction at this temperature.

This revised framing also aligns our interpretation with the experimental picture put forward by Huntin et al -- polyQ length modestly affects onset-like behavior but more strongly tunes condensed-phase regimes and dynamics.

Relevant revisions have been made in the Results and the Discussion sections.

(2) Inconsistency in Cluster Size Across Temperatures (Figure 6B):

The results in Figure 6B show a striking difference in the size of the largest cluster between temperatures of 290K and 300K. This abrupt shift in behavior lacks a clear mechanistic explanation. Typically, phase transitions driven by temperature are more gradual, unless there is some underlying structural or chemical shift that the authors have not accounted for. Without a clear explanation, this sudden change in behavior reduces confidence in the simulation Results.

Suggested Analysis:

The authors should explore possible explanations for the dramatic difference in cluster size between 290K and 300K. For example, they could investigate whether specific interactions (such as the breaking or formation of hydrogen bonds or hydrophobic contacts) might explain the behavior at higher temperatures.

It is important to check whether the coarse-grained simulation model has been adequately parameterized and scaled for accurate temperature dependence. Atomistic simulations of monomers and dimers with varying polyQ tract lengths could be used to fine-tune the coarsegrained model, ensuring it accurately reflects molecular behavior. The gross estimate of a 10% scaling factor might be insufficient and could lead to inaccurate representations of cluster formation.

We agree that the apparently sharp change in largest-cluster size between 290 K and 300 K requires clearer interpretation. In the revised manuscript, we clarify that this behavior does not imply an abrupt thermodynamic phase transition; rather, in a finite (~100-chain) simulation box, the largest cluster size is sensitive to both (i) proximity to a coexistence boundary and (ii) coarsening kinetics. Consistent with this, all systems rapidly coarsen early and then approach a late-time plateau, so the dominant cluster size can change steeply when conditions shift the balance between one system-spanning droplet versus multiple long-lived subclusters.

To distinguish “true loss of condensation” from “differences in coarsening state,” we added replica-averaged stability and persistence metrics (mean ± SEM) and full time series. Importantly, the condensate lifetime (fraction of frames with largest aggregate-population > 50) is ~1 at both 290 K and 300 K, indicating that both temperatures correspond to a persistently condensed regime, not intermittent nucleation/dissolution. We therefore interpret the smaller dominant cluster at 290 K as reflecting slower coarsening / stronger kinetic arrest, where reduced chain mobility delays merger/annealing into a single large droplet within the simulated time window, leaving a larger satellite/dispersed population despite sustained condensation.

We further support this interpretation with mechanistic and dynamical analyses added in the revision. As temperature increases from 290 K to 300 K, we observe increased internal mobility (higher effective diffusivity, D) that would accelerate rearrangements and coalescence. In parallel, contact/desolvation analyses show progressive loss of protein-water contacts and gain of protein-protein contacts as clusters mature, and a residue-resolved comparison indicates net contact increases at 300 K relative to 290 K concentrated in aromatic-rich “sticker” regions, consistent with a strengthened intermolecular contact network that promotes more complete annealing at 300 K.

(We address the reviewer’s points regarding Martini temperature scaling/parameterization together with points (3)-(4) below.)

(3) Scaling of Coarse-Grained Model with Atomistic Simulations:

As mentioned, the coarse-grained model used in the study may not have been properly scaled against atomistic data. A simple scaling factor of 10% may not be appropriate for accurately capturing the behavior of polyQ tracts across different lengths, especially considering their sensitivity to subtle changes in temperature. Without rigorous validation against atomistic simulations, the coarse-grained model's predictions could be skewed.

(4) To address this, the authors should compare the coarse-grained model with atomistic simulations of monomeric and dimeric forms of ELF3 with different polyQ tract lengths. By comparing key structural parameters (e.g., radius of gyration, contact maps, and clustering propensity), the authors could adjust the coarse-grained model to more accurately reflect the atomistic behavior. The authors have wealth of atomistic simulation data that could afford such benchmarking and identification of scaling factor

Additionally, the authors should investigate whether the assumed scaling factor of 10% is appropriate for each polyQ length or whether it needs to be refined based on specific properties, such as the number of hydrophobic interactions or secondary structure stability.

We agree that temperature-dependent CG predictions must be interpreted carefully and that the interaction balance should be justified. In the revision, we therefore clarify both our calibration choice and the scope of inference.

We use Martini 3 with a single, literature-motivated adjustment: protein-water Lennard-Jones interactions are strengthened by 10 percent, following an established strategy shown to improve IDP/multidomain protein behavior in Martini 3. This scaling is applied uniformly to all residues, polyQ lengths, and temperatures to avoid introducing construct-specific parameters and to preserve a controlled comparison across variants.

We emphasize that our CG simulations are used in a comparative manner (how stability/dynamics/structure change with temperature and polyQ length under a fixed model), and we do not claim a quantitatively exact phase boundary or transition temperature for ELF3. In this spirit, and consistent with how Martini 3 has been used in prior work to probe thermally varying properties across temperature windows (while acknowledging documented limits to temperature transferability), we treat the temperature sweep as a comparative probe rather than an absolute calibration (https://doi.org/10.1063/5.0221199, 10.1021/acscentsci.5c00755, https://doi.org/10.1038/s41592-021-01098-3). Accordingly, we report replica uncertainty (mean ± SEM) for all CG observables and restrict conclusions to qualitative trends that are robust to replicate variability.

Finally, while we do not undertake a full ELF3-specific reparameterization, we include qualitative checks linking atomistic and CG behavior: the CG model reproduces the same qualitative features of single-chain reorganization inferred from atomistic simulations — notably the radiusof-gyration (Fig. S8) and the rearrangement/exposure of aromatic “sticker” regions that correlate with strengthened intermolecular contacts in the condensate. We emphasize that these comparisons are intended as qualitative sanity checks on trend direction, not as a quantitative validation or calibration of an absolute phase boundary.

(5) Lack of Analysis for Liquid-Like Behavior in Phase Separation:

The simulations presented in the manuscript do not analyze the liquid-like behavior of ELF3 condensates, which is a key characteristic of liquid-liquid phase separation (LLPS). In LLPS systems, condensates are often dynamic, with chains exchanging between clusters, indicating liquid-like rather than solid-like behavior. The authors fail to probe this crucial aspect, which is necessary to support the claim that ELF3 undergoes phase separation.

Suggested Analysis:

The authors should conduct additional analyses to probe the liquid-like nature of the clusters formed by ELF3. One approach would be to analyze the dynamics of chain exchange between clusters, measuring how frequently chains leave one cluster and join another over time. This analysis would reveal whether the condensates behave as liquid- like, dynamic structures or more static, solid-like aggregates.

Additionally, the temperature dependence of these exchange dynamics should be investigated. In true liquid-liquid phase separation, the rate of chain exchange is often sensitive to temperature. Observing how this rate changes between 290K and 300K, for instance, could help explain the abrupt shift in cluster size seen in Figure 6B.

The authors should also analyze whether the internal structures of the condensates are consistent with a liquid-like phase. For example, radial distribution functions and contact lifetimes could be calculated to reveal whether the clusters exhibit liquid-like organization.

We thank the reviewer for highlighting that liquid-like behavior is a key diagnostic for LLPS. We agree that our original manuscript did not explicitly quantify condensate material properties. In the revision, we therefore add several complementary analyses and figures that directly probe whether the condensed state in our simulations is liquid-like versus dynamically arrested, and how this depends on temperature and polyQ length.

(i) Condensate persistence vs temperature (stability and lifetime).

We now quantify two replica-averaged metrics with uncertainty (mean ± SEM): (a) stability, defined as the mean largest-cluster size over a late-time analysis window, and (b) lifetime, defined as the fraction of frames in which the dominant cluster exceeds a fixed size threshold. These results are shown in the new figures “Stability (Mean cluster size)” and “Lifetime (P [size > 50])”. In our system, both 290 K and 300 K correspond to a persistently condensed regime (lifetime ≈ 1 across variants), whereas at 340 K the lifetime drops substantially (≈0.3-0.5 depending on variant), indicating intermittent condensation / partial dissolution at high temperature. This directly demonstrates temperature-dependent persistence of the condensed state and clarifies that the key qualitative change at high temperature is loss of stability and intermittency, rather than a purely static cluster-size difference.

(ii) Internal mobility and viscoelasticity (D and α).

To probe liquid-like dynamics within the condensed state, we compute internal Mean squared displacement (MSD) and extract an effective internal diffusivity D(T) and anomalous exponent α(T) (new figures FIG X). In our system, D increases systematically with temperature for all variants, confirming that internal rearrangements accelerate at higher temperature. At the same time, α remains strongly subdiffusive (α ≈ 0.3-0.5), indicating constrained, non-Fickian motion rather than simple liquid diffusion. Importantly, we also observe variant-dependent mobility: around 300-320 K, 0Q exhibits markedly lower D than 19Q, consistent with stronger kinetic arrest in 0Q even when both variants are condensed. Together, these dynamics metrics show that our condensates are not ideal liquids, but instead occupy a viscoelastic / dynamically slowed regime with clear temperature dependence.

(ii) Dynamic heterogeneity (self van Hove).

We additionally compute the self van Hove displacement distributions (Fig. SX). In our system, the van Hove distributions deviate from a Gaussian reference matched to the MSD, with an excess of near-zero displacements relative to a simple Gaussian model. This non-Gaussian displacement statistics is consistent with heterogeneous/caging-like dynamics inside the condensed phase, further supporting a viscoelastic (gel-like) rather than purely liquid material state at the timescales accessible to simulation.

(iv) Internal structure and morphology (Rg and anisotropy).

Finally, we add structural descriptors as requested. The new Rg distribution and shape anisotropy (κ²) violin plots quantify single-chain compaction and heterogeneity in morphology within the condensed phase. In our system these structural distributions are broadly overlapping at 300 K, indicating that differences among variants are more strongly expressed in dynamics (D/α/van Hove) and stability/lifetime, rather than in a large change in single-chain compaction at this temperature. We report these results transparently and include them in the SI as additional mechanistic context.

We now explicitly frame our CG condensed phases as viscoelastic/dynamically slowed condensates rather than assuming fully liquid droplets. This interpretation is consistent with experimental observations on ELF3 PrLD that report very slow recovery/gel-like behavior under some conditions (i.e., condensates can age into low-mobility hydrogel states).

(6) Lack of justification of polydispersity of polyQ:

The authors don't provide any rationale for choice of different copies of polyQ used in the manuscript for their chain- growth simulation studies. It will be more apt if it can be motivated via some precedent experimental observations.

We agree and have clarified our rationale in the revised manuscript. ELF3’s polyQ tract is a naturally polymorphic short tandem repeat in Arabidopsis, reported to vary from roughly ~7 to ~29 glutamines in natural populations, and this variation has been linked to ELF3-dependent phenotypes and temperature-responsive growth (Undurraga et al.; Jung et al.). Importantly, recent ELF3 PrLD thermosensing/condensation experiments explicitly compare multiple polyQ lengths (including Q0, short/WT-like constructs such as Q7, and expanded tracts around ~Q20) and show that polyQ length tunes temperature-responsive phase behavior and condensate properties (Jung et al.; Hutin et al.).

Accordingly, for our chain-growth ensembles we chose a small, experimentally motivated set that brackets this range - 0Q (deletion), 7Q (WT-like short), and expanded lengths 13Q and 19Q (with 19Q closely matching the ~Q20 construct used experimentally), so that our simulations map onto established constructs and naturally occurring variation rather than arbitrary copy numbers.

The manuscript draft has been modified in the Results and Methods sections.

Jung J-H. et al. A prion-like domain in ELF3 functions as a thermosensor in Arabidopsis. Nature (2020).

Undurraga S. et al. Background-dependent effects of polyglutamine variation in the Arabidopsis thaliana gene ELF3. PNAS (2012), DOI: 10.1073/pnas.1211021109.

Hutin S. et al. Phase separation and molecular ordering of the prion-like domain of the Arabidopsis thermosensory protein EARLY FLOWERING 3. PNAS (2023).

(7) Lack of initiative to connect to Experiments:

While the computational models and simulations provide robust theoretical insights, the absence of direct experimental validation weakens the overall impact of the manuscript. For example, experimental data on how specific mutations in the polyQ tract influence ELF3 behavior in vivo would significantly bolster the authors' claims. The manuscript would benefit from either citing existing experimental studies that corroborate these findings or from suggesting future experimental directions.

We agree that our original submission did not make the experimental connections explicit enough, and we have strengthened this in the revision by (i) explicitly anchoring our results to published ELF3 thermosensing/condensation measurements and (ii) articulating concrete, experimentally testable mechanistic predictions that follow from the simulations.

(i) Explicit connection to published experimental benchmarks: We now cite and discuss key experimental studies that directly probe ELF3 temperature responsiveness and polyQ effects. Jung et al. demonstrated temperature-triggered ELF3 condensation/speckle formation in vivo and showed that polyQ length modulates thermoresponsive behavior. More recently, Hutin et al. compared ELF3 PrLD constructs spanning polyQ lengths (e.g., Q0, Q7, and ~Q20) and reported temperature-triggered phase separation, condition-dependent condensed-phase regimes (droplet-like versus more arrested/gel-/hydrogel-like), and reduced mobility/immobile fractions quantified by FRAP in some regimes. In the revised manuscript we explicitly map these observations onto our results: our coarse-grained simulations capture temperature-dependent condensation propensity, while our added condensate dynamics analyses (MSD-derived internal mobility DDD, anomalous exponent α\alphaα, and self van Hove displacement statistics) indicate dynamically slowed/heterogeneous condensates rather than assuming ideal liquid droplets—consistent with experimentally observed slow FRAP recovery and arrested behavior under some conditions.

(ii) Mechanistic Connections: While existing experiments establish that ELF3 condensation is temperature-triggered and tuned by polyQ length, they cannot directly resolve the molecular interaction changes that drive these macroscopic readouts. We therefore emphasize that our atomistic and coarse-grained analyses provide a mechanistic interpretation: temperature shifts reorganize and expose “sticker”-rich regions (notably aromatics), strengthening intermolecular contact networks that tune condensate stability and material properties. This framing aligns our conclusions with the experimental picture that polyQ length has modest effects on onset-like behavior but more strongly tunes condensed-phase robustness and dynamics (persistence, internal mobility, and arrest) across temperature

The modifications relevant to this are in the Discussion section.

Reviewer #2 (Public review):

Summary:

The authors aimed to explore how a key protein in the circadian clock of plants, ELF3, responds to temperature changes by forming molecular condensates. They focused on understanding the role of a specific region of the protein, a polyQ tract, in promoting temperature-sensitive structural changes and regulating the formation of condensates. Through a series of computational simulations, they sought to uncover the molecular basis for ELF3's temperature responsiveness and its broader implications for plant growth and adaptation to environmental conditions.

Strengths:

The study's strength lies in its focus on an important biological question: how plants sense and respond to temperature changes at the molecular level. The authors employed a variety of computational techniques, including coarse-grained simulations, to explore the role of specific molecular features in this process. These methods provide a multi-scale view of protein behavior and offer valuable insights into how molecular structures may influence biological function.

Weaknesses:

However, there are notable weaknesses in the evidence provided. While the authors present trends in molecular changes, such as shifts in helical propensity and the formation of condensates, these results seem subtle and are not strongly substantiated by statistical analysis. The lack of error bars in the figures makes it difficult to distinguish between meaningful signals and potential noise in the data. Furthermore, the temperature-sensitive behavior appears to be influenced more by chain length than by sequence-specific effects of the polyQ region, raising questions about whether the findings truly capture the molecular mechanisms responsible for temperature sensing. Additionally, some simulations, particularly those related to the formation of condensates, do not appear fully converged, which casts further doubt on the robustness of the results.

We appreciate the reviewer’s concerns regarding statistical support, sequence specificity, and convergence. In the revised manuscript we (i) report replicate-averaged means with uncertainty (mean ± SEM) for all key observables and add error bars/shaded bands to the relevant figures, (ii) provide the full time series plots in the Supplementary Information to make variability and equilibration transparent, and (iii) revise our interpretation to emphasize that polyQ length has only modest effects on onset-like metrics but more strongly tunes condensate stability and material state (lifetime, internal mobility (D), subdiffusion exponent (α), and non-Gaussian van Hove signatures). This revised framing is consistent with recent ELF3 PrLD experiments showing that polyQ variation can subtly affect onset while substantially modulating condensed-phase behavior and dynamics. Relevant changes to the main text have been made in the Results and Discussion section.

Additional Context for Readers:

Readers should interpret the results with caution, especially regarding the molecular mechanisms proposed for temperature sensing. While the study presents interesting trends, the evidence is not definitive, and the findings may be more reflective of general protein behavior (such as the effect of chain length on condensate formation) than specific sequence-driven responses to temperature. Further experimental studies and more converged simulations will be necessary to fully understand the role of ELF3 in temperature regulation.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

I already have listed my possible recommendations for authors for revising their manuscript in the review. By addressing these issues, the authors could significantly improve the robustness of their conclusions and provide stronger evidence for ELF3's role in temperature-responsive phase separation.