{"id":"66a4a1db-f9a8-4177-b6cf-22fd111b3f68","arxiv_id":"2412.14990","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Simulations show a Caprin1-ATP nanocondensate with an ATP core and protein shell, stabilized by a large negative zeta potential that drops as ATP levels rise and dissolve the droplet.","lead":"This study simulates how the protein Caprin1 and ATP form tiny liquid droplets with a solid ATP core wrapped in protein, and shows that a negative surface electric charge keeps the droplet stable. The findings connect surface electrostatics to droplet stability, which may help predict when and how small condensates form or dissolve in cells.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zeta potential estimate is geometrically inconsistent: the 30–40 Å slip-plane cut lies inside a 15–20 nm droplet rather than at its hydrodynamic shear surface, so the -110 mV stability claim is unsupported as stated.","rationale":"The central claim has two legs: (i) positive NS-ESP at low ATP matches NMR-PRE; (ii) a large negative zeta potential at moderate ATP explains stability. Leg (i) is supported after a disclosed empirical rescaling, but leg (ii) is the quantitative core of the paper's proposed mechanism and is not experimentally benchmarked. The weakest point is not the use of a coarse-grained force field per se—Martini has known limitations and the authors acknowledge them—but the geometric identification of the slip plane. A slip plane at 30-40 Å from the COM of a 15-20 nm droplet is ill-defined and likely inside the droplet. The paper itself invites experimental measurement of nanocondensate zeta potentials, which is appropriate, but the current estimate cannot carry the 'stable oil emulsion' conclusion without a physical derivation of the slip plane. The NS-ESP scaling (1.8) and ATP charge adjustment (-3e) are additional sources of uncertainty for the quantitative values, but they do not by themselves invalidate the sign change of the surface potential. I agree with the reader's conditional verdict; the requested clarification should focus on the slip-plane definition and the radial potential profile.","tokens_in":16558,"tokens_out":6699,"duration_ms":59183,"concrete_test":"Recompute the radial electrostatic potential φ(r) and cumulative charge Q(r) for the 10 mM condensate from the COM out to 120 Å, and define the slip plane physically, for example as the radius where water diffusivity or shear viscosity recovers its bulk value (or where the mobile-ion charge profile reaches its electrokinetic limit). Evaluate the zeta potential at that radius rather than at the fixed 30-40 Å cut. If the value differs from -110 mV by more than about 50 mV, or if φ(r) varies steeply across 30-40 Å, the reported zeta potential is an artifact of the chosen cut and the stability claim is unsupported; repeat for the 100 mM state.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central stability claim—that the Caprin1-ATP nanocondensate is stabilized by a large negative zeta potential, about -110 mV at 10 mM ATP and -25 mV at 100 mM—depends on identifying a slip plane and evaluating the electrostatic potential there (Fig. 5E). The authors place the 'slip-plane' grid points 30-40 Å from the condensate center of mass (Fig. 5D and Methods). However, the droplet is 15-20 nm in diameter, so its radius is 75-100 Å, and the authors' own Fig. 5C places the C-terminal protein chains between the Stern layer and the slip plane, i.e., the slip plane should lie near or outside the protein shell. A cut at 30-40 Å from the COM is therefore inside the immobilized-charge region unless the ATP core and protein shell are both within 40 Å, which is inconsistent with the stated droplet size and with the density profiles in Fig. 2D. The potential at 30-40 Å should be interpreted as an interior or Stern-layer potential, not a zeta potential. The physical slip plane is a hydrodynamic quantity (the surface where solvent tangential velocity vanishes), not an arbitrary radial distance; no such determination is provided. Moreover, the NMR-PRE data used for positive validation were measured on fused micron-scale condensates, not on the nanodroplet, so the negative NS-ESP and zeta values are simulation-only predictions. The reported 1.8 NS-ESP scaling and the -3e ATP charge adjustment calibrate the mixed-state comparison and do not validate the condensate electrostatics. Thus the quantitative basis for 'remarkable stability' is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents multiscale coarse-grained (Martini 3) molecular dynamics simulations, with back-mapping to all-atom resolution, of the C-terminal low-complexity domain of Caprin1 (103 residues) mixed with ATP at concentrations from 0 to 100 mM. The simulations show that at around 10 mM ATP the 10 Caprin1 chains spontaneously assemble into a 15-20 nm spherical condensate with an ATP-rich core (stabilized by pi-pi stacking and sodium counterions) and a protein shell enriched in the N-terminal arginine-rich region, while at 100 mM ATP the condensate dissolves into a mixed state. Using back-mapped structures, the authors compute per-residue near-surface electrostatic potentials (NS-ESP) and compare them with NMR-PRE data from Toyama et al. for the early mixed state. They then estimate a zeta potential for the condensate by evaluating the electrostatic potential at grid points 30-40 Å from the condensate's center of mass, obtaining about -110 mV at 10 mM ATP and about -25 mV at 100 mM ATP, which they interpret as evidence that the nanodroplet is electrostatically stabilized in a manner analogous to an oil-in-water emulsion.","tokens_in":16845,"tokens_out":4705,"duration_ms":39291,"significance":"If the central claims were fully supported, the paper would provide a valuable molecular picture of ATP-mediated condensation and reentrant dissolution of a disordered protein, with a concrete mechanism (an ATP-rich core, a protein shell, and a strong interfacial electrostatic potential) that could explain the stability of nanoscale condensates. The study is technically ambitious: it combines Martini 3 CG-MD, all-atom back-mapping, and NS-ESP calculations, and it includes a control simulation showing that ATP alone does not form oligomers at the relevant concentrations, which is a useful check on force-field artifacts. The qualitative findings on the core-shell organization and the role of N-terminal arginine-rich contacts are plausible and consistent with existing experimental data. However, the quantitative electrostatic claims are weakened by two post-hoc calibrations (ATP charge set to -3e and an NS-ESP scaling factor of 1.8 chosen to match the mixed-state experiments) and by an unvalidated and geometrically questionable definition of the zeta potential.","major_comments":[{"comment":"The reported 'excellent agreement' between the computed mixed-state NS-ESP and the NMR-PRE data is not an independent validation. The Methods state that the ATP charge was set to -3e and all computed NS-ESP values were scaled down by a factor of 1.8 'to ensure consistency with the experimental conditions reported by Toyama et al.' These are free parameters chosen so that the calculation matches the experimental magnitude. Consequently, the agreement in Fig. 4A is partly constructed by construction, and the paper's subsequent use of this agreement to lend credibility to the condensate predictions (which are not directly compared to experiment) is overstated. The authors should clearly label the procedure as a calibration rather than a prediction, and discuss how the conclusions would change if the scaling factor or charge state were varied within a reasonable range.","section":"§3, Fig. 4A and Methods (Electrostatic potential calculations)"},{"comment":"The zeta potential estimate is geometrically inconsistent with the stated condensate size. The droplet is 15-20 nm in diameter (radius ~75-100 Å), yet the slip plane is represented by grid points placed only 30-40 Å from the condensate's center of mass. This location lies inside the region occupied by the ATP core and the protein shell, as the density profiles in Fig. 2D indicate; it is not outside the immobilized charge layer. The physical slip plane is a hydrodynamic shear surface, and no calculation or simulation is provided to identify its location. As a result, the potential at 30-40 Å is an interior or Stern-layer potential, not a zeta potential, and the claimed -110 mV value does not support the conclusion that the nanodroplet behaves like a stable oil emulsion. To make this claim, the authors would need to determine the slip plane from the electrokinetic problem or from an explicit electrophoresis simulation, or they should reframe the result as an internal potential and remove the emulsion-stability interpretation.","section":"§3, Fig. 5D-E and Discussion"},{"comment":"The negative NS-ESP values for the condensate and the associated zeta potential are simulation-only predictions. The paper correctly notes that the NMR-PRE experiments for the condensed state were performed on fused micron-scale droplets that are not representative of the nanodroplet studied here. However, the abstract and discussion present the -110 mV zeta potential as if it were an experimentally validated stability mechanism. The text should explicitly state that, unlike the mixed state, the condensate electrostatics have not been benchmarked against any experimental measurement, and that the stability conclusion is a prediction of the model rather than an established finding.","section":"§3, Fig. 4B-C and Discussion"}],"minor_comments":[{"comment":"The phrase 'the the nanodroplet formation' contains a duplicated article; it should read 'the nanodroplet formation.'","section":"Abstract"},{"comment":"The word 'determinig' should be 'determining'.","section":"Introduction"},{"comment":"The text refers to 'Eq. (1)' but the equation is first defined only later in the Results section; please introduce and number the equation in the Methods, or use a consistent numbering scheme throughout.","section":"Methods (Electrostatic potential calculations)"},{"comment":"The cartoon reproduced from ref. [75] labels the Stern layer and slip plane, but the schematic is not drawn to the scale of the simulated condensate; the placement of the 30-40 Å grid points relative to this cartoon is unclear. A scale bar or a side-by-side density profile labeled with the same radial coordinates would help the reader see the claimed correspondence.","section":"Fig. 5C"},{"comment":"The CAPRIN1-CAPRIN1 RDFs and density profiles appear to be computed from a single 30 µs trajectory per concentration. Since condensation is a stochastic process, reporting results from multiple independent simulations (or at least stating that the findings are reproducible) would strengthen the statistical basis of the structural conclusions.","section":"Results (Condensation and dissolution)"},{"comment":"The phrase 'it seems to us that analogies to oil-in-water emulsions... is a fruitful line of investigation' is tentative, which is appropriate, but the later statement that the -110 mV value is 'commensurate with a stable oil emulsion' presents the analogy as quantitative. Please soften the language to reflect that this is a hypothesis that requires direct experimental measurement of nanodroplet zeta potentials.","section":"Discussion and Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a timely and important question, and the authors are clearly expert in the methods. The qualitative scenario (ATP-rich core, protein shell, electrostatic stabilization) is interesting and likely of interest to the physics.bio-ph readership. However, the two main quantitative pillars are in need of substantial revision: (i) the NS-ESP agreement is a calibration, not a validation, and the language should be adjusted accordingly; and (ii) the zeta potential estimate is based on an arbitrary and physically questionable slip-plane placement. I would encourage the editor to allow a major revision during which the authors either rework the zeta potential analysis using a proper electrokinetic definition or explicitly demote it to an interior Stern potential, and present the mixed-state agreement as an empirical calibration. With those changes, the paper could become a solid and valuable contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a serious simulation paper with a new structural model—ATP forms a stacked core capped by the Caprin1 N-terminus, giving a 15-20 nm core-shell nanocondensate—but the quantitative zeta-potential claim doesn't hold up as stated because the slip-plane cut sits inside the droplet, not at its surface.\n\nWhat's genuinely new: the core-shell architecture, with a solid-like ATP core stabilized by Na+ counterions and a Caprin1 protein shell, is a concrete structural hypothesis for how ATP promotes and dissolves Caprin1 condensates at the nanoscale. The observed sign flip of the NS-ESP across the phase-separation trajectory—positive in the mixed state, strongly negative in the condensate, weakly negative after reentry—is interesting and fits the qualitative electrostatic picture. The back-mapping to all-atom resolution and the residue-resolved NS-ESP calculation are well described, and the methods are detailed enough for expert re-implementation.\n\nThe soft spots. First, the 'excellent agreement' with the NMR-PRE data in the mixed state is partly constructed: the ATP charge is set to -3e and the computed NS-ESP values are scaled down by 1.8 to match the experimental buffer. That is a reasonable calibration, but it means the agreement does not independently validate the simulation. Second—and this is the load-bearing flaw—the zeta potential estimate for the condensate is based on grid points placed 30-40 Å from the center of mass. The droplet is 15-20 nm in diameter, so its radius is 75-100 Å. A cut at 30-40 Å lies inside the ATP core or at the inner edge of the protein shell, not at the hydrodynamic slip plane. The potential there is an interior or Stern-layer potential, not a zeta potential. Consequently, the -110 mV value and the 'stable oil emulsion' analogy are not supported. Third, each condition is a single 30 µs trajectory, with no error bars or replicates; the 5-10 mM phase boundary could be sensitive to finite-size effects. No code or data are shipped.\n\nNone of this destroys the qualitative story. The core-shell structure and the electrostatic sign flip are novel and worth pursuing. But the quantitative stability mechanism needs a proper calculation of the potential at the actual droplet interface, and ideally multiple independent simulations.\n\nWho should read it: anyone working on ATP as a hydrotrope or modulator of condensates, and simulators studying nanoscale phase separation. It deserves a serious referee—the structural model and the electrostatic analysis are important enough that the zeta-potential issue should be fixed rather than desk-rejected.\n\nRecommendation: send it to peer review, but flag the slip-plane placement and the calibration dependence. If those are addressed, the paper could be solid.","headline":"Core-shell Caprin1-ATP nanocondensate is a fresh structural idea, but the zeta-potential estimate rests on a slip-plane cut that sits inside the droplet, not at its hydrodynamic surface.","tokens_in":17433,"tokens_out":4721,"would_cite":false,"duration_ms":38839,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"ATP flips Caprin1's surface charge to build a stable nanocondensate: a solid ATP core, a negative zeta potential near -110 mV, and dissolution at higher ATP.","keywords":["Caprin1","ATP","liquid-liquid phase separation","nanocondensate","zeta potential","near-surface electrostatic potential","coarse-grained molecular dynamics","NS-ESP"],"falsifier":"Measure the electrophoretic mobility of 15-20 nm Caprin1-ATP droplets at 10 mM ATP and convert it to a zeta potential; a measured value far above -110 mV, or a positive value, would contradict the claimed electrostatic stabilization, as would an equivalent simulation in which a different slip-plane placement yields a near-zero potential.","tokens_in":1997,"feed_emoji":"⚡","tokens_out":2987,"duration_ms":86975,"temperature":0.7,"pith_summary":"Using coarse-grained simulations with all-atom back-mapping, the paper traces the ATP-driven phase behavior of the Caprin1 low-complexity domain through three states: a mixed state at low ATP, a 15-20 nm nanocondensate at about 10 mM ATP, and a reentrant mixed state at 100 mM ATP. It claims the condensate is an organized assembly rather than a uniformly mixed droplet: a solid-like core of ATP molecules stacked by pi-pi interactions and sodium counterions, coated by Caprin1 chains whose N-terminal arginine-rich region faces the ATP core. The central quantitative claim is electrostatic: the near-surface electrostatic potential (NS-ESP) of Caprin1 is positive (10-40 mV) in the initial mixed state, matching NMR paramagnetic relaxation enhancement experiments, but becomes highly negative at the condensate surface, corresponding to a Stern double layer and a zeta potential near -110 mV. This large negative zeta potential is offered as the explanation for the nanocondensate's stability against coalescence, by analogy to stable oil-in-water emulsions; raising ATP to 100 mM weakens the electrostatic anchoring and lowers the zeta potential to about -25 mV, coincident with dissolution. If correct, the picture makes ATP a dual agent in the same system, both scaffolding the droplet and, at higher concentration, dissolving it, and it gives a testable surface-electrostatics explanation for why nanometer-scale condensates resist fusion.","feed_headline":"Caprin1 droplets stay stable on a -110 millivolt shell","feed_subtitle":"ATP reverses the protein's surface charge to build an emulsion-like nanodroplet, then dissolves it at high concentration.","key_machinery":"The argument is carried by the near-surface electrostatic potential (NS-ESP), a per-residue quantity computed from back-mapped all-atom configurations by placing probe grid points around the backbone 1H nuclei and evaluating the electrostatic energy difference between positive and negative probes (Eq. (1) of the paper). To convert NS-ESP into a stability statement, the paper uses an oil-in-water emulsion analogy: the ATP core plus the adsorbed Caprin1 N-terminus form a Stern layer, the C-terminus contributes immobile charge inside the slip plane, and the zeta potential is estimated by placing ESP grid points 30-40 Å from the condensate's center of mass, outside the region of immobilized charge. That placement, together with the sign and magnitude of the ATP/Caprin1 charge layers, produces the quoted zeta potentials.","core_discovery":"The central discovery is that ATP does not merely trigger Caprin1 condensation; it organizes the condensate's charge. In the simulated nanocondensate, ATP molecules assemble into a dense inner core through pi-pi stacking of adenine groups, with sodium counterions bridging the triphosphate groups, and the positively charged N-terminal half of Caprin1 adsorbs onto this core while the C-terminus extends toward water. Back-mapping these coarse-grained structures to all-atom resolution, the paper computes per-residue near-surface electrostatic potentials. In the low-ATP mixed state these are positive and match the NMR-PRE measurements; in the condensate they become negative by hundreds of millivolts, so that the assembly resembles a charged double layer. Estimating the potential just outside the immobile charge region gives a zeta potential of about -110 mV at 10 mM ATP, a value the paper associates with emulsion stability, and about -25 mV after dissolution at 100 mM ATP.","pith_inferences":["Beyond the paper's stated conclusions, the oil-in-water analogy makes a direct electrophoretic prediction: the 15-20 nm condensate should migrate as a negatively charged colloid in an applied field, and measuring that mobility would test the -110 mV estimate without relying on the chosen slip-plane position.","The surface-dominated architecture implies that for nanoscale condensates, stability may be set by the interfacial charge layer rather than by bulk protein composition, which would connect these droplets to RNA- or polyphosphate-stabilized nanocondensates and to size-dependent differences in condensate behavior.","The paper itself notes that force-field limitations could inflate the magnitude of its zeta potential relative to micron-scale condensates; a systematic scan of slip-plane distance, or an explicit electrokinetic simulation, would show how sensitive the stability conclusion is to that modeling choice.","If the ATP-rich core is genuinely solid-like, the droplet should show slowed internal exchange of ATP and reduced fluidity on nanometer scales, a testable signature for fluorescence recovery or NMR experiments on small condensates."],"forward_implications":["At low ATP concentrations below about 5 mM, Caprin1 remains mixed and its surface NS-ESP stays positive (10-40 mV), consistent with the NMR-PRE experiments.","At moderate ATP concentrations around 10 mM, ATP self-assembles into a solid-like pi-stacked core with sodium counterions, and the resulting droplet is surface-dominated, with the N-terminal arginine-rich region of Caprin1 bound to the ATP core and the C-terminus facing water.","The condensate's charge organization is equivalent to a Stern double layer, giving an estimated zeta potential of about -110 mV, a value associated with stable emulsions and resistance to coalescence.","At 100 mM ATP, Caprin1-ATP interactions shift to weaker cation-pi and pi-pi contacts, the NS-ESP and zeta potential drop to about -25 mV, and the droplet dissolves back into a mixed state.","The same ATP molecule thus acts as both scaffold and hydrotrope: it stabilizes the nanodroplet at moderate concentrations and dissolves it at high concentrations."],"supporting_citations":[{"why":"Supplies the experimental NMR-PRE NS-ESP values that the paper's mixed-state and condensate calculations are matched against.","marker":"[34]"},{"why":"Provides the probe-grid protocol for computing per-residue near-surface electrostatic potentials that the paper adapts.","marker":"[33]"},{"why":"Establishes the relationship between condensate surface charge, zeta potential, and emulsion stability that the paper uses to interpret its zeta-potential estimates.","marker":"[36]"},{"why":"Provides the experimental and all-atom evidence that Mg-free ATP forms only small oligomers in solution, used to rule out an ATP-clustering artifact in the coarse-grained model.","marker":"[25]"},{"why":"Supplies the Martini 3 coarse-grained force field used for the 30 microsecond condensation simulations.","marker":"[51]"},{"why":"Supplies the Lennard-Jones scaling correction for disordered proteins that the simulations apply to improve Martini 3 accuracy.","marker":"[58]"},{"why":"Supplies the backward-mapping procedure that converts coarse-grained condensate configurations to all-atom structures for the electrostatic calculations.","marker":"[67]"},{"why":"Supplies experimental NMR evidence for the special role of the N-terminal arginine-rich region of Caprin1, which the contact analysis reproduces.","marker":"[74]"}],"fun_headline_variants":["ATP flips Caprin1's charge to form droplets","Charge reversal: ATP builds and dissolves Caprin1 droplets","From + to -: ATP's electrostatic switch controls Caprin1 droplets","Stable nanocondensates: Caprin1 relies on a -110 mV shell","ATP stacks into a charged core to stabilize Caprin1 droplets"],"cache_read_input_tokens":19456,"weakest_assumption_plain":"The stability claim rests on treating the nanocondensate like an oil droplet in water and on assuming that the relevant electric potential is measured just outside the layer of tightly bound charge, at 30-40 Å from the droplet's center; if that assumed measuring distance is wrong, or if the oil-droplet comparison fails at the nanometer scale, the -110 mV value and the stability conclusion do not follow.","fun_headline_variants_meta":{"raw":{"variants":["ATP flips Caprin1's charge to form droplets","Charge reversal: ATP builds and dissolves Caprin1 droplets","From + to -: ATP's electrostatic switch controls Caprin1 droplets","Stable nanocondensates: Caprin1 relies on a -110 mV shell","ATP stacks into a charged core to stabilize Caprin1 droplets"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000809,"raw_usage":{"total_tokens":3605,"prompt_tokens":1054,"completion_tokens":2551,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":670,"completion_tokens_details":{"reasoning_tokens":2460}},"tokens_in":670,"tokens_out":2551,"duration_ms":15650,"temperature":1.0,"reasoning_tokens":2460,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:44:17.623954+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the electrophoretic mobility of 15-20 nm Caprin1-ATP droplets at 10 mM ATP and convert it to a zeta potential; a measured value far above -110 mV, or a positive value, would contradict the claimed electrostatic stabilization, as would an equivalent simulation in which a different slip-plane placement yields a near-zero potential.","supporting_citations":[{"cited_title":"K.; Forman-Kay, J","cited_arxiv_id":null,"evidence_quote":"Supplies the experimental NMR-PRE NS-ESP values that the paper's mixed-state and condensate calculations are matched against."},{"cited_title":"C.; Pettitt, B","cited_arxiv_id":null,"evidence_quote":"Provides the probe-grid protocol for computing per-residue near-surface electrostatic potentials that the paper adapts."},{"cited_title":"J.; Krainer, G.; Espinosa, J","cited_arxiv_id":null,"evidence_quote":"Establishes the relationship between condensate surface charge, zeta potential, and emulsion stability that the paper uses to interpret its zeta-potential estimates."},{"cited_title":"Effects of Weak Nonspecific Interactions with ATP on Pro teins","cited_arxiv_id":null,"evidence_quote":"Provides the experimental and all-atom evidence that Mg-free ATP forms only small oligomers in solution, used to rule out an ATP-clustering artifact in the coarse-grained model."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Martini 3 coarse-grained force field used for the 30 microsecond condensation simulations."},{"cited_title":"E.; Pesce, F.; Roesgaard, M","cited_arxiv_id":null,"evidence_quote":"Supplies the Lennard-Jones scaling correction for disordered proteins that the simulations apply to improve Martini 3 accuracy."},{"cited_title":"A.; Pluhackova, K.; B¨ ockmann, R","cited_arxiv_id":null,"evidence_quote":"Supplies the backward-mapping procedure that converts coarse-grained condensate configurations to all-atom structures for the electrostatic calculations."},{"cited_title":"E.; Kim, T","cited_arxiv_id":null,"evidence_quote":"Supplies experimental NMR evidence for the special role of the N-terminal arginine-rich region of Caprin1, which the contact analysis reproduces."}],"review_version":1}