{"id":"e8364327-9224-476c-80bc-0515f4b6f0c7","arxiv_id":"2505.18301","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Microtubule polymerization, not sliding, is the dominant driver of microtentacle formation, and curved, flexible, long microtentacles maximize contact with vessel walls.","lead":"Using FRAP experiments and simulations, the authors show that polymerization of microtubules, not motor-driven sliding, is the main force behind microtentacle formation in detached cells, and that this creates curved protrusions. Because longer, more flexible, more adhesive microtentacles make more wall contact, the result connects how microtentacles form to how circulating tumor cells might stick to vessel walls.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The curvature and kinesore-detachment mechanism depends on the unproven premise that McTN microtubules are anchored in the MTOC; the FRAP data alone do not establish this.","rationale":"The reader's weakest_assumption identifies exactly the premise I consider most load-bearing: MTs in McTNs remain anchored in the MTOC with plus ends toward the tip. The paper's own text in Section III.B admits that individual MTs could not be traced to the MTOC, and the evidence offered (EB3 anterograde bias, SEM continuity) is indirect. This matters not for the qualitative force-source claim, which is independently supported by the FRAP perturbations: paclitaxel slows recovery and kinesore reveals ROI1/ROI2 intensity losses. Instead, the anchorage premise is essential for the curvature mechanism proposed in Section III.F, where tip forces are said to bend McTNs by pushing against the MTOC, and for the interpretation that kinesore straightens McTNs by detaching MTs from the MTOC. If MTs were instead nucleated at the base or crosslinked along the membrane, the curved-McTN mechanism and the kinesore-detachment story would need substantive revision. The proposed test with minus-end markers would directly localize MT minus ends relative to the MTOC and settle whether the premise is correct. Because this is already reflected in the reader's conditional verdict and no part of the quantitative force-fitting evidence is invalidated by my objection, I recommend leaving the verdict unchanged.","tokens_in":21569,"tokens_out":9731,"duration_ms":94392,"concrete_test":"Perform super-resolution imaging (e.g., STED or SIM) of latrunculin-treated suspended RPE-1 cells co-stained for a microtubule minus-end marker such as CAMSAP2 and an MTOC marker such as gamma-tubulin, then quantify the fraction of McTN microtubule minus ends that colocalize with the MTOC. If a substantial fraction of minus ends are located along the McTN or at its base rather than at the MTOC, the anchorage premise is false; if they are predominantly at the MTOC, the premise is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III.B states that individual MTs could not be traced back to the MTOC because of the high MT density in suspended cells; the supporting evidence is an anterograde bias of EB3 comets and the absence of visible MT fragments in SEM images. Plus-end growth direction does not locate minus ends: MTs nucleated at the McTN base or along the bundle would also show anterograde EB3 movement, and continuity from the cell body is not continuity to the MTOC. This premise is load-bearing because Section III.F interprets paclitaxel-induced straightening as reduced tip force against the MTOC and kinesore-induced straightening as detachment from the MTOC; both interpretations collapse if MTs are not anchored. The novel kinesore-detachment claim rests on the same inference plus two indirect observations: bleached ROI displacement in about 20% of cells, and loss of a visible MTOC in adhered cells. If anchorage is not established, the central statement that polymerization of MTs anchored in the MTOC drives curved McTNs is overinterpreted, and the fitted ratio of polymerization to sliding force is not enough to rescue the curvature mechanism.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates the mechanism of microtentacle (McTN) formation in circulating tumor cells using a combination of fluorescence recovery after photobleaching (FRAP) experiments and simulations. The authors first show that latrunculin A-treated RPE-1 cells form MT-based protrusions that resemble McTNs and promote adhesion. They then use EB3 kymographs to show that MT plus-ends grow predominantly anterogradely toward the McTN tip, and FRAP experiments with paclitaxel and kinesore to argue that polymerization dominates over MT sliding. A two-state stochastic model of MT growth against a deformable membrane is used to reproduce the McTN length distribution and the FRAP recovery curves, yielding an asymptotic membrane force of about 28.7 pN and a sliding force of about 3.2 pN. The paper concludes that polymerization of MTs anchored at the MTOC drives the generation of curved McTNs, and that longer, more flexible McTNs enhance cell-wall contact. The authors propose that McTN length and curvature are functionally relevant parameters for CTC adhesion and metastasis.","tokens_in":21680,"tokens_out":8401,"duration_ms":64241,"significance":"If the central claims are correct, the paper identifies polymerization as the dominant force-generating mechanism for McTN formation and connects McTN morphology to adhesive function, which would be a meaningful advance for understanding CTC extravasation. Strengths include an independent perturbation design: paclitaxel slows FRAP recovery and kinesore produces clear ROI1/ROI2 intensity losses, providing qualitative support that does not rely on the simulation. The FRAP simulation is a genuine cross-check of the fitted model, and the adhesion simulations make falsifiable predictions about the role of McTN length and flexibility. The use of a non-cancerous cell line to argue that the mechanism is general is also a positive feature. However, the quantitative layer is fragile: the asymptotic force is fitted to the length distribution that it is later claimed to reproduce, and the sliding force is fitted to the same FRAP curves used for validation. The load-bearing assumption that McTNs remain anchored in the MTOC is supported only indirectly, and the curvature mechanism depends critically on this premise.","major_comments":[{"comment":"The asymptotic force F ≈ 28.7 pN is defined as the value that 'minimizes the error between simulated and experimental length distribution.' Consequently, the agreement between the simulated and experimental McTN length distributions shown in Fig. 5B is obtained by construction and does not independently validate the polymerization mechanism. The independent test is the FRAP recovery simulation; however, pure polymerization alone deviates by about 22%, and the improved 5% deviation is achieved by additionally fitting the sliding force to the same FRAP data. The text should explicitly distinguish fitted parameters from genuinely predicted quantities, and should present the 22% mismatch as a limitation of the pure-polymerization model rather than as confirmation.","section":"Section III.E, Eqs. (1)-(5), Suppl. Fig. S7B"},{"comment":"The claim that MTs in McTNs are anchored in the MTOC is load-bearing for the curvature mechanism, yet it rests on indirect evidence. The authors state in III.B that individual MTs could not be traced back to the MTOC due to high MT density; the supporting evidence is the anterograde bias of EB3 comets and the absence of visible MT fragments in SEM images. As noted, plus-end growth direction does not establish minus-end location: MTs nucleated near the McTN base or along the bundle would also produce anterograde EB3 movement, and SEM continuity from the cell body does not prove continuity to the MTOC. Section III.F uses this premise to interpret paclitaxel-induced straightening as reduced tip force against the MTOC and kinesore-induced straightening as detachment from the MTOC; both interpretations are invalid if MTs are not anchored. The authors should either provide direct evidence for minus-end anchoring (e.g., minus-end markers or photoactivation experiments) or substantially weaken the curvature and detachment claims.","section":"Section III.B and III.F"},{"comment":"The quantitative claim that 'the force exerted due to polymerization is ten times higher than the force generated by sliding' compares two fitted quantities: F = 28.7 pN is fitted to the length distribution, and the sliding force of 3.2 pN is fitted to the FRAP recovery curve. The reported ratio is therefore not a model prediction with quantified uncertainty; it is a statement about the best-fit values. The number of MTs per McTN, estimated as 10 from reference [56] and used to share the force in Section III.E, also affects the absolute values. Please provide a sensitivity analysis over these parameters and report confidence intervals for the fitted forces.","section":"Section III.E and Fig. 5"}],"minor_comments":[{"comment":"The assumption of negligible free-tubulin diffusion inside McTNs is stated without quantitative support: 'the highly confined geometry of McTNs leads to drastically reduced diffusion coefficients... and is therefore insufficient to affect the observed trends.' Please provide an estimate of the diffusion timescale for tubulin dimers in a ~1 µm diameter McTN or a reference supporting this claim.","section":"Section III.C"},{"comment":"There are numerous typographical and encoding issues, including 'imp ortant' in the title, ligature artifacts in 'fluorescence,' 'ˆA°C' for degrees Celsius, '1 1 µM' in the Supplementary Materials, and 'Comapring' in the p-value section. Please proofread carefully.","section":"Throughout"},{"comment":"The image-processing package is essential for McTN length and curvature quantification, but its details are deferred to a future publication. Please provide at least a brief validation of the length and curvature measurements (e.g., comparison to manual tracing or synthetic test images).","section":"Section II.E"},{"comment":"The description of the bleach region contains a redundant phrase: 'ROI* was manually located approximately at half the length of the McTN length.' Please clarify the selection criteria and the exact dimensions of the bleached region relative to the McTN diameter.","section":"Section II.G"},{"comment":"There is a suspicious duplication of p-values across different comparisons (e.g., 0.002117802 appears for both the 20-min adhesion time in Fig. 1G and the entering-vs-rescue comparison in Fig. 2B), suggesting a copy-paste error. Please verify all supplementary p-values.","section":"Supplementary Statistics"},{"comment":"The sliding velocity is set to equal the MT growth velocity 'as an approximation.' This choice directly affects the fitted sliding force and should be justified with a sensitivity analysis or a reference to measured sliding velocities.","section":"Section III.E, Suppl. Table S1"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a timely question, and the perturbation-based FRAP data are valuable and likely of interest to the biophysics and cell-biology readership. The main risks are the circularity of the length-distribution fit and the unproven MTOC-anchoring premise, both of which are load-bearing for the central claims. These are addressable through revised framing and additional experiments, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the qualitative claim—MT polymerization, not sliding, is the main driver of McTN formation—is well supported by the FRAP perturbation data. The quantitative force numbers are fit-derived and should not be quoted as measurements. The unproven MTOC anchorage is the main soft spot, but the paper flags it honestly.\n\nWhat's new: the FRAP design with ROI1/ROI*/ROI2 plus paclitaxel and kinesore is a clean way to separate polymerization from sliding. Control shows recovery in ROI*, no consistent loss in adjacent ROIs; paclitaxel slows recovery and produces a clear ROI2 loss; kinesore gives strong ROI1/ROI2 loss consistent with sliding. That is a solid independent perturbation logic. The EB3 kymographs are a reasonable supporting dataset. The model is standard two-state MT dynamics with a Brownian ratchet and force-dependent catastrophe, and the parameters are taken from their own kymographs. The simulations do reproduce the FRAP curves in the perturbed conditions with ~10% error, which gives some confidence the mechanism is captured.\n\nSoft spots: (1) The asymptotic membrane force F≈28.7 pN is fitted to the McTN length distribution it is then said to reproduce. The sliding force 3.2 pN is fitted to the FRAP recovery curves used for validation. So the tenfold ratio is an output of the fitting, not an independent measurement. The authors should provide uncertainties and ideally fit on one dataset, validate on another. (2) The curvature interpretation depends on MTs being anchored in the MTOC, which is stated as \"likely\" and supported by indirect evidence (EB3 anterograde bias, SEM continuity, and prior work). Plus-end growth direction alone doesn't establish minus-end location. This is a genuine limitation, but it is acknowledged; it should be pinned down experimentally before the curvature/adhesion model is taken as established. (3) The claim that free-tubulin diffusion is negligible rests on an assertion, not a measurement. It doesn't sink the qualitative argument, but it should be justified. (4) The vessel-wall framing uses RPE-1 cells on coated plastic; extrapolation to CTCs and endothelium is speculative.\n\nBottom line: this is a useful, honest paper. The central message—polymerization dominates over sliding in McTN formation—is probably right and is a meaningful advance over prior structural/adhesion work. The quantitative force ratio and the curvature mechanism need more support. I'd send it to peer review with the expectation of moderate revision, and I'd ask the authors to release their image-processing and simulation code. Who for: biophysicists and cell biologists studying microtubule-based protrusions, CTC adhesion, and force generation. Worth a serious referee.","headline":"FRAP perturbation data convincingly show polymerization dominates McTN formation, but the quantitative force ratio and curvature mechanism rest on assumptions that need pinning down.","tokens_in":22354,"tokens_out":4264,"would_cite":true,"duration_ms":31040,"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":"Microtubule polymerization, not microtubule sliding, is the force that builds the curved microtentacles circulating tumor cells use to attach to blood-vessel walls.","keywords":["microtentacles","circulating tumor cells","microtubule polymerization","microtubule sliding","FRAP","Brownian ratchet","cell adhesion","semi-flexible filament model"],"falsifier":"Using super-resolution or electron tomography to trace individual microtubules from a microtentacle tip back to the MTOC in latrunculin A-treated cells; if a substantial fraction of microtentacle microtubules are free fragments or have minus ends at the tip, the proposed tip-force-against-MTOC curvature mechanism and the kinesore-detachment interpretation both fail.","tokens_in":21247,"feed_emoji":"🧬","tokens_out":7997,"duration_ms":67186,"temperature":0.7,"pith_summary":"Circulating tumor cells can extend long, thin, microtubule-filled membrane protrusions called microtentacles, which help them grab blood-vessel walls during metastasis. This paper argues that the dominant force that builds these protrusions is microtubule polymerization, not the sliding of microtubules past one another, and that the reaction force of the deformed membrane on the growing tip, resisted by the microtubule-organizing center at the other end, gives microtentacles their characteristic curvature. The claim rests on FRAP experiments in non-cancerous RPE-1 cells whose actin cortex is weakened with latrunculin A, combined with simulations of two-state microtubule dynamics, force-dependent polymerization, and semi-flexible filaments. If correct, microtentacle length, flexibility, and curvature are not incidental shapes but functional adhesion parameters that determine how well a circulating tumor cell attaches to a vessel wall, which would give clinicians new readouts for metastatic potential and new targets for therapy.","feed_headline":"Polymerization builds the tentacles cancer cells cling with","feed_subtitle":"FRAP and simulation show growth forces run about 10-fold stronger than sliding; curved, flexible protrusions stick best to walls.","key_machinery":"The load-bearing machinery is a two-state model of microtubule dynamics with force-dependent polymerization and catastrophe rates, formulated through the Brownian ratchet relation and solved as master equations for the microtubule length distribution, coupled to a discrete semi-flexible filament model whose bending and extensional energies are minimized together with a Lennard-Jones adhesion energy when the microtentacle approaches a wall. The first model converts the measured polymerization parameters, growth and shrinkage velocities, catastrophe and rescue frequencies from EB3 kymographs, into a predicted protrusion-length distribution and FRAP recovery curves. The second converts length, bending rigidity, and adhesion strength into a wall-connection phase diagram. The bridge between them is the assumption that all microtubules in a microtentacle share the membrane force equally and remain anchored in the MTOC, so that tip forces translate into curvature rather than into sliding of the bundle.","core_discovery":"On the paper's own terms, the central discovery is that microtubule polymerization inside a microtubule-organizing-center-anchored bundle is the primary force generator for microtentacle formation: the membrane pushes back on the growing plus ends with a fitted asymptotic force of about 28.7 pN, while the best-fit contribution from forward microtubule sliding is only about 3.2 pN, roughly one tenth as large. Because the microtubules are anchored in the MTOC at the cell body, the reaction force of the deformed membrane on the tip bends the bundle and produces the curved microtentacles seen in cells. When a microtentacle approaches a wall, nonspecific adhesion competes with the filament's bending energy; the simulations map out a phase diagram in which long, flexible microtentacles with strong adhesion maximize the contact area, while short, stiff ones minimize it. This links the formation mechanism, polymerization-driven growth and curvature, to the function, adhesion to vessel walls, and identifies length and curvature as quantifiable determinants of circulating tumor cell attachment.","pith_inferences":["One testable extension: severing a microtentacle between tip and cell body with a laser should make the distal segment straighten in control cells but not in kinesore-treated cells, directly testing the anchoring assumption the authors could not resolve by imaging.","The model implies that modest changes in polymerization dynamics alone, without any adhesion change, should measurably shift the microtentacle curvature distribution; recording curvature histograms after low-dose drug treatments would probe this prediction.","If curvature is a functional adhesion parameter, then microtentacle length and curvature distributions in patient-derived circulating tumor cells might correlate with extravasation or clinical outcome; the paper does not test this, but its mechanism makes the correlation a reasonable hypothesis.","An implicit trade-off follows from the kinesore results: interventions that lengthen microtentacles by promoting sliding can also detach them from the MTOC and introduce kinks, so longer protrusions may be less adhesive; this suggests the optimal adhesive state is long, curved, and MTOC-anchored rather than simply maximally long."],"forward_implications":["Simulations reproduce the measured microtentacle length distribution and FRAP recovery only when polymerization is combined with weak forward sliding; the fitted membrane force is about 28.7 pN versus about 3.2 pN for sliding, a factor of roughly ten.","Suppressing polymerization with paclitaxel reduces FRAP recovery and yields straighter, slightly shorter microtentacles, while enhancing kinesin-1 sliding with kinesore produces longer but straighter microtentacles and causes bleached regions to move, indicating microtubules detach from the MTOC.","The wall-connection phase diagram shows that long, flexible microtentacles with strong nonspecific adhesion maximize the cell-wall contact area, while short, stiff microtentacles minimize it.","Kinks, modeled as localized reductions in bending rigidity, generally decrease wall connection, although the effect depends on kink position and relative stiffness."],"supporting_citations":[{"why":"Supplies the two-state growth/shrinkage model with catastrophe and rescue used for the master equations.","marker":"[42]"},{"why":"Supplies the Brownian ratchet relation that makes polymerization velocity decay with tip force.","marker":"[31]"},{"why":"Supplies the force dependence of catastrophe frequency used in eq. (2).","marker":"[50]"},{"why":"Shows how force is shared among microtubules in a bundle, the basis for dividing membrane force across roughly 10 microtubules.","marker":"[66]"},{"why":"Prior super-resolution work concluding that microtentacle microtubules connect to the MTOC and estimating about 10 microtubules per microtentacle.","marker":"[56]"},{"why":"Defines microtentacles and establishes their role in reattachment, the phenomenon this paper mechanistically explains.","marker":"[13]"},{"why":"Establishes kinesin-1-powered microtubule sliding, the activity targeted by paclitaxel and kinesore in the FRAP experiments.","marker":"[62]"},{"why":"Describes kinesore as a kinesin-1 activator, used to enhance sliding and detach microtubules from the MTOC.","marker":"[63]"},{"why":"Provides the FRAP intensity-recovery analysis protocol that the experiments follow.","marker":"[41]"},{"why":"Provides the Lennard-Jones potential used to model nonspecific microtentacle-wall adhesion in the phase diagram.","marker":"[70]"}],"fun_headline_variants":["Polymerization, not sliding, powers microtentacle growth","Growth forces, not sliding, build curved tentacles for cancer spread","Microtubule polymerization, not sliding, creates microtentacles for invasion","Polymerization-driven microtentacles help cancer cells stick to walls","Polymerization builds curved microtentacles that anchor circulating tumor cells"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the microtubules inside a microtentacle remain anchored in the microtubule-organizing center with their growing plus ends toward the tip; the authors could not trace individual microtubules to the MTOC because of high density, so if many microtubules are unanchored fragments, the curvature mechanism and the kinesore-detachment interpretation collapse.","fun_headline_variants_meta":{"raw":{"variants":["Polymerization, not sliding, powers microtentacle growth","Growth forces, not sliding, build curved tentacles for cancer spread","Microtubule polymerization, not sliding, creates microtentacles for invasion","Polymerization-driven microtentacles help cancer cells stick to walls","Polymerization builds curved microtentacles that anchor circulating tumor cells"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00221,"raw_usage":{"total_tokens":8579,"prompt_tokens":996,"completion_tokens":7583,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":612,"completion_tokens_details":{"reasoning_tokens":7488}},"tokens_in":612,"tokens_out":7583,"duration_ms":44996,"temperature":1.0,"reasoning_tokens":7488,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:33:39.262278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Using super-resolution or electron tomography to trace individual microtubules from a microtentacle tip back to the MTOC in latrunculin A-treated cells; if a substantial fraction of microtentacle microtubules are free fragments or have minus ends at the tip, the proposed tip-force-against-MTOC curvature mechanism and the kinesore-detachment interpretation both fail.","supporting_citations":[{"cited_title":"& Tabeling, P","cited_arxiv_id":null,"evidence_quote":"Supplies the two-state growth/shrinkage model with catastrophe and rescue used for the master equations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Brownian ratchet relation that makes polymerization velocity decay with tip force."},{"cited_title":"& Prost, J","cited_arxiv_id":null,"evidence_quote":"Supplies the force dependence of catastrophe frequency used in eq. (2)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows how force is shared among microtubules in a bundle, the basis for dividing membrane force across roughly 10 microtubules."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior super-resolution work concluding that microtentacle microtubules connect to the MTOC and estimating about 10 microtubules per microtentacle."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines microtentacles and establishes their role in reattachment, the phenomenon this paper mechanistically explains."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes kinesin-1-powered microtubule sliding, the activity targeted by paclitaxel and kinesore in the FRAP experiments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes kinesore as a kinesin-1 activator, used to enhance sliding and detach microtubules from the MTOC."},{"cited_title":"K., Marko, J","cited_arxiv_id":null,"evidence_quote":"Provides the FRAP intensity-recovery analysis protocol that the experiments follow."},{"cited_title":"R., Wysocki, A., Winkler, R., Gompper, G","cited_arxiv_id":null,"evidence_quote":"Provides the Lennard-Jones potential used to model nonspecific microtentacle-wall adhesion in the phase diagram."}],"review_version":1}