{"id":"5f25a122-581b-4218-9f19-dda557c84d0f","arxiv_id":"2411.16910","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Weak opposite membrane deformations repel by about 3 kBT, while strong opposite deformations attract by about 5.3 kBT.","lead":"Using a vesicle and optical tweezers, scientists measured how particles that bend a lipid membrane interact with a membrane tube bending it the opposite way. The interaction flips from repulsion to attraction as the particle's deformation gets stronger, giving numbers that theories can be tested against.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Partially wrapped particle 'attraction' may be direct NeutrAvidin–biotin adhesion to the tube, not a curvature-mediated interaction.","rationale":"I read the paper as making two distinct experimental claims: (i) a repulsive interaction between a fully wrapped particle and an outward tube, and (ii) an attractive well for a partially wrapped particle, with the sign switch attributed to deformation strength. The first claim is supported by a stable, fully wrapped particle and a control with non-deforming particles. The second claim is the load-bearing one for novelty. The partially wrapped particles are described as transient states caught in the wrapping transition, and—crucially—they are fully functionalized colloids, so the unwrapped part retains NeutrAvidin and can adhere directly to the biotinylated tube membrane. The paper's control (low NeutrAvidin density) is not an adequate control for this direct binding because it lacks the adhesive capacity. The reader's concern about the TPM equilibrium/constant-D assumption is real but secondary: it would affect the numerical values of both repulsion and attraction, whereas the adhesion confound would invalidate the attribution of the attraction altogether. The proposed soluble-biotin/passivated-Janus test discriminates between the two mechanisms and can be done with the existing setup. I therefore keep the reader's CONDITIONAL verdict unchanged: the paper should not be judged as establishing the sign-switch until this control is performed or the exposed surface is shown to be non-adhesive.","tokens_in":11899,"tokens_out":10690,"duration_ms":108978,"concrete_test":"Repeat the Sec. 3.2 measurement after passivating the exposed non-wrapped surface of the partially wrapped particle, e.g., by saturating exposed NeutrAvidin with soluble biotin (or using a Janus particle with a 50% adhesive patch and a passivated non-adhesive hemisphere). If the -5.3 kBT well disappears or is strongly reduced, the attraction is direct NeutrAvidin–biotin adhesion to the tube rather than a membrane-curvature-mediated interaction; if it persists unchanged, the curvature-mediated interpretation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central sign-switch result (Sec. 3.2) rests on the -5.3 kBT well between a partially wrapped particle and the pulled tube. The particle used there is the standard high-NeutrAvidin colloid that is only ~50% wrapped because the membrane lacks area; its non-wrapped hemisphere is therefore still covered with NeutrAvidin and can bind the biotinylated membrane of the tube. The control in Sec. 3.1 used particles with only 2.5% NeutrAvidin coverage, which cannot form such bonds; it therefore does not exclude a direct adhesive interaction between the exposed particle surface and the tube. The paper's conclusion that the interaction is 'solely due to the curvature' is not established for the partially wrapped case. This is more load-bearing than the constant-D/equilibrium assumption: even a perfectly equilibrated trajectory would show an attractive well if the particle adheres to the tube. The fully wrapped repulsion (no exposed NeutrAvidin) may be genuine, but the sign switch would then reflect adhesion, not deformation strength.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental study of membrane-mediated interactions between colloids and a pulled membrane tube in giant unilamellar vesicles. A membrane tube pulled outward creates a positive-curvature deformation, while particles wrapped on the inside create a negative-curvature deformation. The authors find that fully wrapped particles are repelled from the tube, with a repulsion of about 3 kBT at a distance of 0.5 µm and a power-law decay, whereas partially wrapped particles (about 50% coverage) are attracted to a potential well of -5.3 kBT located about 1 µm from the tube. They also report that Janus particles with a 67% adhesive patch attract each other with a maximum force of about 0.4 pN. The paper interprets these observations as evidence that the sign of curvature-mediated interactions depends on the strength of the deformation.","tokens_in":12100,"tokens_out":5220,"duration_ms":47837,"significance":"If the interpretation is correct, these experiments provide a quantitative demonstration that the sign of membrane-mediated interactions between oppositely curved deformations depends on deformation strength, a result that is relevant to cooperative protein assembly and microplastic interactions with cell membranes. The manuscript's strengths include the use of multiple independent techniques (transition-probability-based energy extraction and direct optical-tweezer force measurements), the inclusion of a non-deforming control, and the comparison with existing theoretical and simulation predictions. The central claim, however, rests on a control that may not exclude a direct adhesive interaction between the partially wrapped particle and the biotinylated tube, which makes the significance conditional on additional controls or analysis.","major_comments":[{"comment":"The attractive potential of -5.3 kBT for a partially wrapped particle may be caused by direct NeutrAvidin–biotin adhesion between the exposed surface of the particle and the biotinylated membrane of the tube, rather than by a curvature-mediated interaction. The partially wrapped particle is a standard, fully NeutrAvidin-functionalized colloid that is only about 50% covered by the membrane; its non-wrapped hemisphere is therefore still coated with NeutrAvidin and can bind biotin lipids on the tube. The control shown in Fig. 3b (green squares) is the same low-NeutrAvidin (2.5% surface coverage) particle used in Sec. 3.1, which cannot form such bonds. This control therefore does not rule out direct adhesion for the partially wrapped case. The paper's conclusion in Sec. 4 that \"the interaction was solely due to the curvature\" is not established for this system. Please provide a control using a partially wrapped particle whose non-wrapped hemisphere is passivated (e.g., with mPEG) or otherwise non-adhesive, or demonstrate from the trajectories that the particle does not contact the tube.","section":"Sec. 3.2 and Fig. 3b"},{"comment":"The transition-probability method used to extract interaction energies assumes a constant diffusion coefficient and equilibrium Boltzmann sampling. Near the tube, the membrane is strongly curved and the partially wrapped particle's local hydrodynamic coupling and available phase space may differ from the unperturbed membrane. If the diffusion coefficient decreases near the tube, the method would infer an apparent attractive well even in the absence of a real potential. The non-deforming control shows no measurable interaction, but that particle has a different geometry and may not experience the same position-dependent friction as a partially wrapped particle. Please test the constancy of the diffusion coefficient as a function of s from the trajectory data, or cross-validate the tube-particle potential with an independent force measurement of the type used for the Janus particles in Sec. 3.3.","section":"Sec. 3, energy extraction method"}],"minor_comments":[{"comment":"The height reconstruction assumes that the particle is confined to an ellipsoidal membrane, but near the tube the membrane shape is strongly perturbed; please quantify the uncertainty in the geodesic distance s that this approximation introduces for particles close to the tube.","section":"Sec. 2.7"},{"comment":"There are several typographical errors and incomplete references, for example \"25? –28\" in the introduction and \"? Upon\" in Sec. 3.2; these should be corrected.","section":"Sec. 1 and Sec. 3.2"},{"comment":"The sentence \"they do not interact significantly interact\" contains a duplicated verb and should be revised.","section":"Sec. 3.3"},{"comment":"The conclusion overstates the strength of the control experiment; please soften the claim that the interaction is \"solely due to the curvature\" in light of the control limitations discussed in the major comments.","section":"Sec. 4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses a timely and interesting question, and the experimental work appears careful. The main concern is that the attractive well for partially wrapped particles (the key evidence for the sign switch) may be attributable to direct NeutrAvidin–biotin adhesion rather than curvature-mediated interaction. This is a load-bearing issue that can be fixed with an additional control or by limiting the claim. The constant-diffusion assumption in the energy extraction is a second area that needs attention, but it is secondary to the adhesion concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives the first experimental numbers for interactions between opposite membrane deformations: fully wrapped particles are repelled from a pulled tube (about 3 kBT at 0.5 µm), partially wrapped particles show an attractive well (-5.3 kBT around 1 µm), and Janus particles attract with up to 0.4 pN. The repulsion result is the one I trust most. It is new, plausible, and backed by a non-deforming control that shows no interaction at low NeutrAvidin coverage. The authors also do a decent job engaging with prior simulations and explaining where their setup differs, including membrane tension and the shape of the deformation.\n\nThe real soft spot is the partially wrapped attraction. The particle exhibiting the -5.3 kBT well is a standard high-NeutrAvidin colloid that is only about 50% wrapped because the membrane lacks area. Its exposed hemisphere remains covered with NeutrAvidin, and the tube membrane is biotinylated. So there is a direct adhesive channel between the particle and the tube that has nothing to do with curvature. The control in Sec. 3.1 uses particles with only 2.5% NeutrAvidin coverage, which cannot form such bonds; it therefore does not exclude direct adhesion for the partially wrapped case. The paper's conclusion that the interaction is \"solely due to the curvature\" is not established for that system. That is more load-bearing than the constant-diffusion/Boltzmann extraction, which would bias the numbers but would not manufacture a well if the trajectory is sampling a bound state. The constant-D assumption near a strongly curved tube is a secondary worry, as is the ellipsoidal membrane assumption in the z-reconstruction.\n\nFor the fully wrapped repulsion, the arguments hold up well. The Janus-particle force measurement is also useful, though it addresses same-side deformations and is less central to the title claim. The figures would benefit from error bars, but that is a presentation issue, not the core problem.\n\nWho should read this: people working on membrane-mediated interactions, colloid-membrane systems, and curvature-based assembly. It deserves a serious referee. The repulsion measurement is novel and likely correct, and the flaw in the attraction claim is identifiable and fixable. I would send it to review and require an additional control: saturate exposed NeutrAvidin with free biotin, or use a Janus particle whose bare face is passivated, before trusting the sign switch.","headline":"First quantitative look at opposite-curvature membrane interactions, with a genuinely useful repulsion measurement for fully wrapped particles; the partially wrapped 'attraction' is compromised by a missing adhesion control.","tokens_in":12561,"tokens_out":2297,"would_cite":true,"duration_ms":23983,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Opposite membrane curvature repels when weak and attracts when strong.","keywords":["membrane-mediated interaction","curvature-mediated interaction","giant unilamellar vesicle","membrane tube","Janus particles","wrapping fraction","interaction potential","optical tweezers"],"falsifier":"Track a wrapped particle's short-time mean-squared displacement in narrow bins of distance from the tube: if the local diffusion coefficient varies beyond experimental error, the transition-probability energies (3 kBT and -5.3 kBT) are biased; equivalently, measure the force-distance curve directly with two optical traps and compare it with the negative gradient of the reported potential.","tokens_in":11712,"feed_emoji":"🧫","tokens_out":7914,"duration_ms":71125,"temperature":0.7,"pith_summary":"This paper asks whether two objects that deform a lipid membrane in opposite directions attract or repel each other through the membrane, and answers that the sign depends on how strongly they deform it. Using a giant unilamellar vesicle, the authors pair a particle wrapped inward by the membrane with an outward membrane tube pulled by optical tweezers. Fully wrapped particles, which imprint a weak deformation, are repelled from the tube with an energy of about 3 kBT at 0.5 µm. Particles wrapped over only about half their surface, which imprint a much stronger deformation, are instead attracted into a potential well 5.3 kBT deep centered about 1 µm from the tube. If the result holds, curvature-mediated interactions in cell membranes cannot be classified by curvature sign alone; the magnitude of the induced deformation matters.","feed_headline":"Opposite membrane curvature repels when weak, attracts when strong","feed_subtitle":"New GUV experiments show deformation strength, not just direction, decides membrane-mediated force sign.","key_machinery":"The load-bearing object is the geodesic distance s between the wrapped particle and the junction where the pulled tube meets the vesicle; the reported energies are functions of s. The experimental apparatus pairs controlled wrapping (NeutrAvidin–biotin adhesion density sets the contact area) with an optical trap that pulls a membrane tube outward, creating a well-defined positive-curvature deformation opposite the particle's inward deformation. Interaction energies are extracted from confocal trajectories with a transition-probability-matrix method that assumes a constant diffusion coefficient and equilibrium Boltzmann sampling, and the particle's out-of-plane position is reconstructed by assuming the vesicle is ellipsoidal. For the strongly attractive Janus pairs, where Boltzmann statistics cannot be sampled, the authors switch to two optical traps and measure the force directly.","core_discovery":"On the paper's own terms, the central discovery is a sign switch: the first experimental measurement of the membrane-mediated interaction between oppositely curved deformations shows repulsion for small deformations and attraction for large ones. A fully wrapped particle deforming a GUV membrane inward is repelled from an outward-deforming membrane tube, with a repulsion of about 3 kBT at a geodesic distance of 0.5 µm and a power-law decay with exponent -0.80 ± 0.08. A partially wrapped particle with about 53% coverage experiences an attractive potential well of -5.3 kBT centered roughly 1 µm from the tube, with a parabolic stiffness of 51.5 ± 0.5 fN/µm. Janus particles with 67% of their surface adhered attract each other with a maximum force of about 0.4 pN near contact, while particles with only 5% adhesion show no measurable interaction. Non-deforming adhered particles show no interaction with the tube, confirming that the forces are curvature-mediated rather than fluctuation-mediated.","pith_inferences":["One could map the repulsion-to-attraction switch by measuring energy profiles for a series of patch sizes between 5% and 67%; if the switch is monotonic in wrapping fraction, the threshold itself would be a measurable material property of the membrane-particle pair.","The attractive well at about 1 µm from the tube suggests a tube can act as a curvature trap that locally concentrates strongly deforming objects before they finish wrapping; this could be tested by seeding a partially wrapped particle near a tube and watching whether wrapping completes preferentially there.","In a multicomponent membrane, the sign switch implies that weakly and strongly curving proteins of the same curvature sign could self-segregate, since weak equal-sign deformations may repel while strong ones attract; the paper does not test this directly.","The assumption of constant diffusion coefficient near the tube is testable with local mean-squared-displacement measurements; if it fails, the absolute values of 3 kBT and -5.3 kBT would shift but the qualitative sign switch could still stand."],"forward_implications":["Opposite-curvature deformations are not universally repulsive; at fixed membrane tension, the sign of the membrane-mediated force is controlled by how strongly the object curves the membrane.","The measured energies (3 kBT repulsive, -5.3 kBT attractive) are of the same order as the equal-deformation attraction found earlier, so both channels can compete when proteins or colloids organize on a membrane.","Because non-deforming adhered particles show no interaction, the observed forces are attributable to membrane bending and tension rather than to thermal fluctuations or direct colloid-tube contact.","The direct optical-trap measurement gives a value of about 0.4 pN for the strongest same-side attraction in this system, a number that simulations and continuum theories of partially wrapped particles should reproduce.","The short-range power-law repulsion with exponent -0.80 provides a quantitative target for theories of two opposite deformations on a tense membrane."],"supporting_citations":[{"why":"Supplies the theoretical prediction that oppositely deforming objects attract at intermediate distances and repel at short range on a tense membrane, the framework this experiment tests.","marker":"[15]"},{"why":"Provides an analytical approach for the deformed membrane geometry that predicts attraction between objects on opposite sides of the membrane.","marker":"[18]"},{"why":"Gives a numerical calculation of the opposite-deformation potential well at finite tension, compared against the measured well depth and shape.","marker":"[24]"},{"why":"Reports coarse-grained simulations in which the interaction switches from repulsion to attraction as the deformation strength increases, matching the wrapped-versus-partially-wrapped sign switch observed here.","marker":"[26]"},{"why":"Supplies the GUV-plus-adhesive-colloid model system and the transition-probability method for extracting interaction energies, plus the -3.3 kBT equal-deformation baseline this work compares against.","marker":"[30]"},{"why":"Describes the vesicle electroformation and wrapping protocols used in sample preparation for these measurements.","marker":"[33]"},{"why":"Establishes the protocol for tuning NeutrAvidin surface density to control the wrapping fraction of the colloids.","marker":"[35]"},{"why":"Simulates membrane-deforming particles on an elongated vesicle and finds repulsion from positively curved regions, supporting the wrapped-particle-tube repulsion reported here.","marker":"[55]"},{"why":"Provides the two-optical-trap force measurement method used to quantify the 0.4 pN attraction between Janus particles.","marker":"[62]"}],"fun_headline_variants":["Opposite membrane curves: weak repels, strong attracts","Membrane deformation strength flips interaction sign","Weak opposing bends repel, strong ones attract in GUVs","Curvature magnitude decides membrane force direction","Repulsion to attraction with stronger opposing curvature"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reported energies are computed from particle trajectories assuming a constant diffusion coefficient and equilibrium Boltzmann sampling, even in the strongly curved region next to the tube where friction and available phase space may differ.","fun_headline_variants_meta":{"raw":{"variants":["Opposite membrane curves: weak repels, strong attracts","Membrane deformation strength flips interaction sign","Weak opposing bends repel, strong ones attract in GUVs","Curvature magnitude decides membrane force direction","Repulsion to attraction with stronger opposing curvature"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000163,"raw_usage":{"total_tokens":1264,"prompt_tokens":987,"completion_tokens":277,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":603,"completion_tokens_details":{"reasoning_tokens":204}},"tokens_in":603,"tokens_out":277,"duration_ms":3041,"temperature":1.0,"reasoning_tokens":204,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:44:41.099237+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Track a wrapped particle's short-time mean-squared displacement in narrow bins of distance from the tube: if the local diffusion coefficient varies beyond experimental error, the transition-probability energies (3 kBT and -5.3 kBT) are biased; equivalently, measure the force-distance curve directly with two optical traps and compare it with the negative gradient of the reported potential.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the theoretical prediction that oppositely deforming objects attract at intermediate distances and repel at short range on a tense membrane, the framework this experiment tests."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides an analytical approach for the deformed membrane geometry that predicts attraction between objects on opposite sides of the membrane."},{"cited_title":"Weitz and N","cited_arxiv_id":null,"evidence_quote":"Gives a numerical calculation of the opposite-deformation potential well at finite tension, compared against the measured well depth and shape."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports coarse-grained simulations in which the interaction switches from repulsion to attraction as the deformation strength increases, matching the wrapped-versus-partially-wrapped sign switch observed here."},{"cited_title":"van der Wel, A","cited_arxiv_id":null,"evidence_quote":"Supplies the GUV-plus-adhesive-colloid model system and the transition-probability method for extracting interaction energies, plus the -3.3 kBT equal-deformation baseline this work compares against."},{"cited_title":"Azadbakht, B","cited_arxiv_id":null,"evidence_quote":"Describes the vesicle electroformation and wrapping protocols used in sample preparation for these measurements."},{"cited_title":"van der Wel, N","cited_arxiv_id":null,"evidence_quote":"Establishes the protocol for tuning NeutrAvidin surface density to control the wrapping fraction of the colloids."},{"cited_title":"Vahid, A","cited_arxiv_id":null,"evidence_quote":"Simulates membrane-deforming particles on an elongated vesicle and finds repulsion from positively curved regions, supporting the wrapped-particle-tube repulsion reported here."},{"cited_title":"Azadbakht, T","cited_arxiv_id":null,"evidence_quote":"Provides the two-optical-trap force measurement method used to quantify the 0.4 pN attraction between Janus particles."}],"review_version":1}