{"id":"ae21970b-64d3-4250-8e24-9cde0fa19579","arxiv_id":"1908.00712","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For stiff membranes, increasing surface tension first slows, then speeds up, and then slows again the membrane pushed by a filament, because the bending energy cost of a protrusion peaks and then falls.","lead":"This paper studies a one-dimensional elastic membrane pushed by growing filaments and finds that raising surface tension can make membrane protrusions easier to form when the membrane is stiff. The result is a non-monotonic velocity-tension curve with a dip and a peak, explained by an interplay of surface and bending energy.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim rests on the linearized Helfrich Hamiltonian (Eq. 1); at low σ the protrusion reaches roughly 400–500 nm (Fig. 5c), and the neglected (∇h)⁴ terms are never quantified, so the nonmonotonic energy cost could be a linearization artifact.","rationale":"The paper's strongest claim is a direct simulation result: for moderate or large κ, V(σ) has a dip and a peak, and the average upward energy cost E_b^+(σ) has a peak. The direct simulation is the main evidence, and it is partially supported by L-scaling checks and by the d≠δ results in the Supplementary Material, so the absence of error bars is a reporting weakness rather than a decisive flaw. The load-bearing assumption is the linearized Helfrich Hamiltonian. The nonmonotonicity mechanism is explained through the discrete second and fourth derivatives h''_b and h''''_b in Eq. (2), and those derivatives are computed from the quadratic Hamiltonian. At σ=0.01 the height profile in Fig. 5(c) reaches hundreds of nanometers, and although the gradient may still be small if the horizontal lattice spacing is large, the paper never states the horizontal spacing quantitatively and never checks the size of the neglected quartic term. The approximate analytical shape in Eq. S-2 comes from the same linear equation, so it does not independently justify the linearization. The right test is to add the lowest-order nonlinear surface term and repeat the key measurements; this directly settles whether the peak in E_b^+ and the velocity dip-and-peak are robust properties of the Helfrich model or artifacts of the quadratic approximation. If the test passes, the conditional verdict can be upgraded; until then, conditional acceptance is appropriate.","tokens_in":13328,"tokens_out":20769,"duration_ms":220121,"concrete_test":"Run the same kinetic Monte Carlo model with the next-order correction to Eq. (1), replacing σΣ(h_i−h_{i−1})² by σΣ[(h_i−h_{i−1})² − (1/4)(h_i−h_{i−1})⁴], which is the low-gradient expansion of √(1+g²)−1, for κ=1.2 pN/nm, L=64, and σ∈[0.01,1] pN/nm. Also record the maximum |h_i−h_{i−1}| at σ=0.01. If the dip-and-peak in V and the peak in E_b^+ survive with shifts smaller than the σ-resolution, and the maximum gradient is below about 0.1, the linearization is adequate. If the peak disappears or moves substantially, the central claim is a quadratic-Hamiltonian artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—that E_b^+(σ) and V(σ) are non-monotonic because the membrane shape switches from a linear cusp to a rounded profile—is computed entirely from the quadratic Helfrich Hamiltonian in Eq. (1). The authors justify dropping nonlinear terms in Sec. II by asserting that height gradients are much less than unity, but they never report the actual maximum gradient or the magnitude of the next-order surface term, especially for σ=0.01 where Fig. 5(c) shows binding-site heights of roughly 400–500 nm over a few lattice sites. If |h_i−h_{i−1}| is O(0.1–1) on the lattice scale, the quadratic surface energy σΣ(h_i−h_{i−1})² overestimates the true Helfrich surface cost σΣ(√(1+(h_i−h_{i−1})²)−1), and the measured peak in E_b^+ and the resulting V dip-and-peak may be artifacts of this overestimate. The analytical shape calculation in Sec. III of the Supplementary Material uses the same linear operator (Eqs. S-1 and S-2), so it cannot validate the linearization. The reader's secondary point about missing error bars is less load-bearing than this, because the nonmonotonicity is large (about a factor of 3 in V) and smooth across many σ values; the linearization is a structural assumption underpinning every energy cost in Eqs. (2)–(5).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a one-dimensional lattice model of an elastic membrane pushed by polymerizing filaments, with the membrane height governed by a linearized Helfrich Hamiltonian that includes surface and bending terms. Using kinetic Monte Carlo simulations, the authors report that for fixed moderate or large bending rigidity κ, the average membrane velocity V as a function of surface tension σ is non-monotonic: V first decreases to a minimum, then increases to a maximum, and finally decays exponentially. They decompose the energy cost of a protrusion at the filament binding site into surface and bending contributions and show that the bending contribution, and hence the total cost, has a peak as a function of σ, so that over a range of σ increasing surface tension lowers the cost of protrusion. The mechanism is attributed to a qualitative change in the membrane shape around the binding site, from a sharply kinked linear profile at high σ to a rounded profile at low σ, which reduces the fourth derivative of the height. An approximate analytical calculation of the velocity from the energy costs is offered to explain the full V–σ curve.","tokens_in":13639,"tokens_out":12268,"duration_ms":111564,"significance":"If the reported effect is real, it is a counterintuitive and potentially important result for models of actin-driven membrane protrusion: surface tension, normally expected to suppress protrusions, can enhance them through its interplay with bending rigidity. The paper's strengths are its simple, clearly specified model; direct stochastic simulation of the full dynamics; checks of system-size dependence (Fig. S-5); and an explicit parameter set in physical units that makes the prediction testable. The central non-monotonicity is large (roughly a factor of 3 in V) and appears smoothly over many values of σ, which makes it unlikely to be a small-sample artifact. However, the physical relevance of the result depends on the validity of the linearized Hamiltonian in the low-σ regime, and the absence of statistical error bars weakens the quantitative support.","major_comments":[{"comment":"The linearization of the Helfrich Hamiltonian is asserted but never quantitatively validated. The non-monotonicity in E_b^+ and V occurs precisely at small σ where protrusion amplitudes are largest: Fig. 5(c) shows heights of roughly 400–500 nm, and the manuscript never reports the maximum value of |h_i − h_{i−1}| or the magnitude of the neglected quartic surface term relative to σ(h_i − h_{i−1})^2. Since Eqs. (2)–(5) and the Metropolis rates all derive from the quadratic Hamiltonian, if height gradients approach order unity on the lattice scale, the bending-energy peak and the velocity dip-and-peak could be artifacts of the linearization. The analytical check in Sec. III of the Supplementary Material (Eqs. S-1 and S-2) uses the same linearized operator and therefore cannot serve as independent validation. Please provide a quantitative gradient check over the full σ range, including σ = 0.01, and if necessary repeat the energy-cost calculation with the full nonlinear surface term.","section":"Sec. II, Eq. (1), Fig. 5(c)"},{"comment":"No error bars or number of independent runs are reported for any averaged quantity. The central claim is the non-monotonic dip-and-peak in V(σ) and the corresponding peak in E_b^+(σ); without statistical uncertainties the reader cannot judge whether the dip and peak are significant or whether the apparent decrease of E_b^+ after the peak is within noise. Please report standard errors over independent steady-state runs for the main curves (at least Figs. 3 and 4).","section":"Figs. 3–5 and supplementary figures"},{"comment":"The 'more quantitative explanation' replaces the fluctuating local energy costs E_b^± by their steady-state averages inside exponential rates. Because the exponential is nonlinear, this replacement is not generally a controlled approximation, and the manuscript presents no direct comparison between the predicted V(σ) obtained from Eqs. (3)–(5) and the simulated V(σ) in Fig. 3. Without such a comparison, Eqs. (3)–(5) constitute a consistency argument rather than a quantitative explanation. Please add the comparison or explicitly label the calculation as a qualitative consistency check.","section":"Sec. III.C, Eqs. (3)–(5)"}],"minor_comments":[{"comment":"In the large-σ limit E_b^+ ≈ E_b^-, so the downward thermal-fluctuation term does not vanish. Combining the terms in Eq. (3) gives a prefactor δ[(U0+2)p0−1] e^{−6βδ^2κ}, not δ(U0+1)p0 e^{−6βδ^2κ}. The exponential decay constant is unaffected, but the prefactor in Eq. (5) is incorrect as written.","section":"Eq. (5)"},{"comment":"The abstract and introduction state without qualification that protrusions are more easily formed as σ increases, but Sec. VII of the Supplementary Material shows that for d > δ the energy cost E_b^+(δ) is monotonic in σ (inset of Fig. S-6) and only the velocity retains a peak. Please qualify the claim to the d = δ case.","section":"Abstract and Sec. VII of Supplementary Material"},{"comment":"There are several typographical errors, including 'membran e' in the title and the line break in 'filem nt' in the abstract; these should be corrected.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The main risk to this paper is the linearization issue: the central claim is computed with a quadratic Hamiltonian, and the low-σ regime has the largest protrusions. If the authors can show numerically that max |∇h| stays small (or that the nonlinear term does not change the peak), the result is likely to be acceptable. The error-bar request is standard but less severe. The paper is within the journal's scope and the effect, if robust, is interesting."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version. This paper finds, in a 1D lattice model of a membrane pushed by growing filaments, that the average protrusion velocity as a function of surface tension is non-monotonic when the bending rigidity is moderate or large: velocity dips, then rises to a peak, then decays exponentially. The non-monotonicity is directly simulated, not derived from an uncontrolled approximation, and it is new compared with the authors' earlier work and the cited literature. The explanation, that the bending energy cost peaks because the membrane shape switches from a sharp cusp to a rounded profile, is clear and supported by measurements of the fourth derivative.\n\nWhat the paper does well: the model is simple and carefully described, the parameter values are physically motivated, and the authors test several auxiliary questions (system size, relative time scales, monomer-to-step size ratio). The analytic shape calculation in the supplement, despite using a simplified point-force version, captures the qualitative shape transition. The exponential tail at large sigma is checked numerically and matches the predicted decay constant.\n\nThe soft spots. The main one is the linearized Helfrich Hamiltonian. The authors justify neglecting nonlinear terms by asserting that height gradients are much less than unity, but the actual protrusion shapes at low sigma (Fig. 5c) reach roughly 400-500 nm in height over a few lattice sites, which on any physical mapping of the lattice spacing gives gradients of order 0.1-1, not clearly small. The analytic shape in the supplement uses the same linear operator, so it cannot validate the linearization. This matters because the peak in the bending energy—the mechanism behind the effect—could be suppressed or altered by the nonlinear bending energy at large slopes. The surface term, if anything, is overestimated by the quadratic approximation at large gradients, so the direction of the error is not obvious; the effect may survive in the nonlinear model, but the paper does not show it. Secondary issue: no error bars on the central simulation averages, though the dips and peaks are large and smooth, so this is a minor gap.\n\nThe citation pattern is fine. The paper properly builds on the authors' earlier work and other membrane-filament models.\n\nWho this is for: biophysicists working on membrane mechanics or actin-based protrusion. It is a reasonable candidate for peer review, not a desk reject. A serious referee should ask for a quantification of the maximum gradient in the simulations, and ideally a check of the nonlinear Hamiltonian, before the claim is taken as fully established. I'd accept it with major revisions, but I wouldn't block publication if the gradients turn out to be small on the relevant scale.","headline":"A directly simulated non-monotonic protrusion velocity vs surface tension, with a plausible shape-change mechanism, but the linearized Hamiltonian is used where gradients may be too large — needs a nonlinear check and error bars.","tokens_in":14147,"tokens_out":3818,"would_cite":true,"duration_ms":38383,"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":"For a fixed bending rigidity, a membrane pushed by growing filaments moves fastest at intermediate surface tension: the protrusion-energy cost itself is non-monotonic in tension.","keywords":["membrane protrusion formation","Helfrich Hamiltonian","surface tension","bending rigidity","actin polymerization","non-monotonic velocity","membrane shape transition","lattice Monte Carlo simulation"],"falsifier":"Rerun the lattice model with the full nonlinear Helfrich Hamiltonian at $\\kappa = 1.2$ pN/nm from $\\sigma = 0.01$ to $0.8$ pN/nm: if the bending-energy peak and the velocity rise vanish, the claimed effect is an artifact of linearization.","tokens_in":13122,"feed_emoji":"🔬","tokens_out":9124,"duration_ms":87173,"temperature":0.7,"pith_summary":"This paper claims that the usual intuition—surface tension always opposes membrane protrusion—fails when the membrane has a moderate or large bending rigidity. In a one-dimensional lattice model of a Helfrich membrane pushed by growing filaments, the membrane velocity as a function of $\\sigma$ is non-monotonic: it falls to a minimum, rises to a maximum, and then decays exponentially. The reason is that the bending-energy cost of creating a protrusion is itself non-monotonic: the surface part rises with tension, but the bending part peaks and then drops as the local membrane shape switches from a sharp linear kink to a rounded profile. If true, this provides a mechanical route by which stronger effective tension could make protrusion easier, with relevance for actin-driven cell motility and any elastic membrane deformed by a localized force.","feed_headline":"For stiff membranes, more tension can ease protrusion formation","feed_subtitle":"A model membrane pushed by filaments shows a velocity dip and peak as surface tension grows.","key_machinery":"The machinery is the discrete Helfrich Hamiltonian, $H = \\sigma \\sum_i (h_i-h_{i-1})^2 + \\kappa \\sum_i (h_{i-1}-2h_i+h_{i+1})^2$, where $h_i$ is a one-dimensional height field on a lattice and $\\sigma$, $\\kappa$ are proportional to surface tension and bending rigidity. From this Hamiltonian the paper derives the energy cost of raising or lowering the binding-site height by one polymer-monomer step $\\delta$, and decomposes it into a surface part $\\Sigma_b^\\pm$ controlled by the second difference $h''_b$ and a bending part $K_b^\\pm$ controlled by the discrete fourth derivative $h''''_b = h_{b-2} - 4h_{b-1} + 6h_b - 4h_{b+1} + h_{b+2}$. The fourth derivative is the load-bearing quantity: it rises as the protrusion kink sharpens and then falls when the membrane switches to a rounded profile, making $K_b^+$ peak with $\\sigma$. A linearized evolution equation with a point force, solved in the co-moving frame, gives an analytic height profile of the same shape-switching form used to support the mechanism.","core_discovery":"At a fixed bending rigidity $\\kappa$, the membrane velocity $V$ is not a monotonic function of surface tension $\\sigma$. For small $\\kappa$, $V$ decreases with $\\sigma$, as expected. For moderate or large $\\kappa$, $V$ first falls to a minimum, then rises to a maximum, and finally decays exponentially for large $\\sigma$. The paper locates the origin of this in the energy cost $E_b^+$ of raising the binding-site height by one monomer step: the surface-energy part $\\Sigma_b^+$ increases monotonically with $\\sigma$, but the bending-energy part $K_b^+$ has a peak. Because of this peak, there is a range of $\\sigma$ where the total protrusion-energy cost decreases as tension increases, making protrusion formation easier. The mechanism is a qualitative shape change near the binding site: at large tension the height profile is almost linear, at intermediate tension its slope steepens, and at low tension it switches to a rounded form with a smaller curvature contribution.","pith_inferences":["If the shape-switch mechanism is generic, the same non-monotonic energy cost should appear whenever a localized pusher deforms any interface governed by a quadratic surface-plus-bending energy; one test is to look for it in reconstituted lipid vesicles pushed by actin comet tails.","The two-dimensional version is left open by the authors; a concrete extension would be to simulate the 2D Helfrich membrane with pushing filaments near a flat barrier and measure whether the peak-tension position moves to higher or lower values when bending energy is larger.","The dependence of the peak position on the membrane/filament time-scale ratio suggests a dynamical criterion: tuning membrane fluidity or filament speed could move a cell in or out of the tension-helpful regime, which is a testable prediction even though the paper does not make it."],"forward_implications":["At moderate or large $\\kappa$, increasing $\\sigma$ from zero first slows the membrane, then accelerates it to a peak, then drives an exponential decay; the dip-and-peak is a prediction that can be looked for in force-velocity curves.","The peak tension $\\sigma^*$ shifts to smaller values as $\\kappa$ grows, and also as the membrane's relaxation time-scale $S/L$ grows, so stiffer or faster-fluctuating membranes show the rise at lower tension.","For small $\\kappa$ the effect disappears: the height profile stays linear at all tensions and $V$ falls monotonically with $\\sigma$.","If the filament monomer step $d$ is larger than the membrane fluctuation step $\\delta$, the peak in $V$ survives but the preceding minimum is absent, and the rise is driven by the tension dependence of downward membrane moves.","At large $\\sigma$, $V$ decays as $\\exp(-2\\beta\\delta^2\\sigma)$, consistent with a nearly flat membrane whose only remaining cost is surface energy."],"supporting_citations":[{"why":"The authors' earlier model of actin filaments growing against an elastic membrane; supplies the simulation framework and the low-bending-rigidity baseline where velocity decreases monotonically with tension.","marker":"[27]"},{"why":"Supplies the linearized height-field evolution equation and the point-force height profile used to explain the shape-switch analytically.","marker":"[43]"},{"why":"Provides the discretized Helfrich Hamiltonian form in Eq. (1) on which all energy-cost calculations are based.","marker":"[29–32]"},{"why":"Justifies dropping nonlinear terms in the Hamiltonian, i.e. the linearization that is the paper's weakest premise.","marker":"[24, 33, 34]"},{"why":"Experiments showing protrusion rate falls when membrane tension is increased; the expectation the paper's non-monotonic result is compared against.","marker":"[8, 9]"},{"why":"Shows a situation where tension enhances protrusion by directing actin polymerization; the antecedent phenomenon the proposed mechanism could explain.","marker":"[10]"},{"why":"Force-velocity measurement techniques the paper points to for experimental verification of the predicted dip-and-peak curve.","marker":"[44–46]"}],"fun_headline_variants":["Bending energy peak lets surface tension ease membrane protrusion","For stiff membranes, more tension can lower protrusion energy cost","Membrane velocity dip and peak arise from surface tension interplay","Surface tension's boost to protrusions hinges on bending energy","Interplay of surface and bending energy flips protrusion ease with tension"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The energy calculation is done with a linearized Helfrich Hamiltonian that assumes height gradients are small everywhere; at the lowest surface tensions the simulated protrusions are hundreds of nanometers high, so this assumption may not hold.","fun_headline_variants_meta":{"raw":{"variants":["Bending energy peak lets surface tension ease membrane protrusion","For stiff membranes, more tension can lower protrusion energy cost","Membrane velocity dip and peak arise from surface tension interplay","Surface tension's boost to protrusions hinges on bending energy","Interplay of surface and bending energy flips protrusion ease with tension"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000656,"raw_usage":{"total_tokens":2967,"prompt_tokens":870,"completion_tokens":2097,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":486,"completion_tokens_details":{"reasoning_tokens":2011}},"tokens_in":486,"tokens_out":2097,"duration_ms":15696,"temperature":1.0,"reasoning_tokens":2011,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:35:46.914011+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Rerun the lattice model with the full nonlinear Helfrich Hamiltonian at $\\kappa = 1.2$ pN/nm from $\\sigma = 0.01$ to $0.8$ pN/nm: if the bending-energy peak and the velocity rise vanish, the claimed effect is an artifact of linearization.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The authors' earlier model of actin filaments growing against an elastic membrane; supplies the simulation framework and the low-bending-rigidity baseline where velocity decreases monotonically with tension."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the linearized height-field evolution equation and the point-force height profile used to explain the shape-switch analytically."},{"cited_title":"Raucher and M","cited_arxiv_id":null,"evidence_quote":"Shows a situation where tension enhances protrusion by directing actin polymerization; the antecedent phenomenon the proposed mechanism could explain."}],"review_version":1}