REVIEW 3 major objections 4 minor 19 references
Critical threshold for microtubule amplification through templated severing
T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read Microtubule copying flips on above a rescue threshold of ~0.36.
desk verdict Worth reading and citing: the bounded-growth threshold for templated severing is well supported by simulations and the two-crossovers theory lands for most parameters; the stress-test's offspring-count objection collapses on close reading. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The machine is a one-dimensional dynamic-instability model of a longitudinal microtubule in a periodic grid of stable transverse filaments, using the classical two-state growth-shrinkage kinetics with catastrophe and rescue rates. The load-bearing piece is the offspring-size criterion $M = [p_+ + (1-p_+) R^-_d(d)] M_+ > 1$, where $M_+$ is the mean number of severing descendants of a filament born in the growing state and $R^-_d(d)$ is the probability that a shrinking filament recovers the birth length $d$; from this condition the authors obtain an explicit formula for the critical $p_+^*$. To evaluate $M_+$ they develop a two-crossovers approximation in which only the first two transverse filaments matter, supplemented by a first-passage-time distribution for reaching a nearby crossover built from legal paths with at most one catastrophe-rescue cycle. The approximation's key output is the probability $p_{cr}(p_+)$ that a crossover at the first filament is erased by shrinkage after a severing at the second, which injects the $p_+$ dependence into the threshold.
What would settle it
In a bounded-growth reconstituted assay with transverse spacing fixed at 1.5 µm, tune $p_+$ (e.g., by varying CLASP concentration) and track individual severing lineages; the two-crossovers theory predicts extinction below $p_+^*=0.361$ and exponential amplification above it. Observing any lineage that survives repeated severing at $p_+=0.25$, or an extinction-to-amplification transition outside roughly 0.34–0.38, would falsify the predicted threshold.
Extended reading notes
Core claim
The central claim is that templated severing is self-amplifying in the unbounded-growth regime but requires a threshold probability of rescue-after-severing in the bounded-growth regime. Writing the mean number of severing descendants of a newborn filament as $M = [p_+ + (1-p_+) R^-_d(d)] M_+$, where $R^-_d(d)$ is the splitting probability that a shrinking filament recovers the birth length $d$ and $M_+$ is the offspring count for a filament born growing, the authors reduce the amplification condition to $M>1$ and solve it for a closed-form critical value $p_+^*$. The evaluation of $M_+$ relies on a two-crossovers approximation in which only the first two transverse filaments are retained and at most one catastrophe-rescue event is allowed before a crossover; this yields $p_+^*=0.361$ against a simulated $0.360$ for the reference parameters, with relative errors under 3% for most of the other tested parameter sets. In the unbounded-growth regime the average lifetime is infinite, so a filament can be severed without bound and descendants accumulate even when every lagging plus end shrinks, which removes the need for any threshold.
Load-bearing premise
The analytical threshold assumes the full grid of transverse microtubules can be replaced by just two filaments, and that at most one catastrophe-rescue event occurs before a crossover is reached; the paper's own comparisons show the resulting first-passage-time approximation fails for distant targets, so the closure is only locally valid.
Editorial extensions
If this is right
- In the bounded-growth regime, a newly severed filament and all its descendants go extinct with probability one when $p_+$ lies below the critical value, so the threshold is a true sharp phase boundary in the model.
- Above the threshold, the number of longitudinal microtubules grows exponentially; the growth rate and the probability that amplification succeeds both rise monotonically with $p_+$.
- In the unbounded-growth regime, amplification occurs even at $p_+=0$; rescue-after-severing only speeds up the process and lowers the extinction probability.
- Since tubulin depletion slows growth, a system that begins in the unbounded-growth regime can be pushed into the bounded-growth regime, where the threshold becomes decisive for whether reorientation completes.
Reading between the lines
- A testable corollary not drawn in the paper: the threshold should depend mainly on the probability that a severed lagging filament reaches the next transverse filament, so a single measurement of that crossing probability in vivo could predict whether reorientation will proceed.
- The one-catastrophe first-passage-time approximation is shown to fail for distant targets; this suggests the predicted threshold would drift if the transverse array is sparse (large spacing), a regime the paper does not simulate.
- The model freezes the transverse array as an inert template, but real transverse filaments are themselves dynamic and consume tubulin; extending the model to let the template erode during amplification would likely raise the effective threshold over time and could explain why reorientation slows before completion.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper investigates a stochastic one-dimensional model of cortical microtubule amplification via templated severing at crossovers with a stable transverse array. The model couples dynamic instability of longitudinal microtubules with a Gamma-distributed severing waiting time and a probability p+ of rescue-after-severing of the newly created lagging plus end. Stochastic simulations show that in the unbounded-growth regime amplification occurs even at p+ = 0, while in the bounded-growth regime amplification only occurs above a critical threshold p+_crit ≈ 0.36 for the reference parameters. The authors construct an approximate analytical theory based on truncating the number of crossovers considered (one- and two-crossovers approximations) and a first-passage-time distribution, and report that the two-crossovers theory predicts the simulated threshold with relative errors below 3% for most of the eleven parameter sets in Table II.
Significance. If valid, the result provides a quantitative, falsifiable prediction for when rescue-after-severing is required for light-induced microtubule array reorientation, connecting the CLASP-mediated rescue effect to a sharp threshold in the bounded-growth regime. The paper combines large-scale stochastic simulations (N = 10^5 trials) with an approximate analytical closure, and the central threshold phenomenon is robustly demonstrated in the simulations. The comparison across eleven parameter sets is a useful test of the analytical approximation. However, the analytical derivation of the critical point relies on a branching-process criterion whose formulation is not self-consistent, which undermines the theoretical explanation of the observed threshold even if the simulation result stands.
major comments (3)
- [Sec. III.B.2, Eq. (8)] The offspring count m_i is not a standard reproduction number. In Eq. (8), m_i = s_i + Σ_j c_ji, where c_ji counts the crossovers behind a severing event that may be resolved by severing on the lagging daughter microtubule. The text explicitly states that these lagging-daughter severings are treated as 'direct daughter microtubules of the mother.' In a branching process, the criticality condition is that the expected number of direct daughters per individual exceeds one; each severing event on a microtubule creates exactly one direct daughter (the lagging microtubule). Severing events on that lagging daughter are granddaughters and must be counted in the daughter's reproductive output, not the mother's. Adding them to m_i inflates M and makes the condition M > 1 (Eq. (5)) neither necessary nor sufficient for supercriticality. The two-crossovers correction in Eq. (11) subtracts only the fraction pcr of such crossovers resolved by shrinkage; it does not remove the fundamental misattribution. Therefore p+_crit,(2) is not derived from a valid reproduction number, and the agreement with simulated thresholds in Table II is not theoretically explained. The authors should either derive the threshold using the correct per-microtubule direct-daughter reproduction number or explicitly justify why their two-generation count acts as a valid criticality condition in this model.
- [Table II] The simulated critical values p+_crit are reported without error bars or confidence intervals. The relative errors of the two-crossovers theory range from 0% to 22.2% (row with v+=0.10, v-=0.250, rc=0.010, rr=0.020), with a second error of 15.7% (v+=0.10, v-=0.250, rc=0.015, rr=0.030). Without uncertainty estimates on the simulated thresholds, it is impossible to judge whether these deviations are statistically significant or whether the 'reasonable accuracy' claim holds uniformly across parameter sets. Please report standard errors computed from the simulation ensemble, or confidence intervals from a bootstrap analysis.
- [Sec. III.C.1, Eq. (14) and Fig. 4E-F] The FPTD approximation assumes at most one catastrophe-rescue event before reaching the target (Eq. (14)) and is shown to fail for distant targets (d = 30 and 60 μm in Fig. 4E,F). The authors argue that this failure is irrelevant because the arrival probability decays as exp(-d/l) for d ≫ l. However, this argument is not quantified for the parameter sets used in Table II, and the two-crossovers theory relies on FPTDs to both the first and second crossover. For parameter sets where the second crossover at 2d may be in a regime where the approximation degrades, the predicted p+_crit could inherit additional error beyond the branching-process issue. Please state the regime of validity of Eq. (14) for the parameters of Table II and discuss whether any of the reported predictions are affected by the FPTD failure.
minor comments (4)
- [Abstract / Discussion] The abstract claims the analytical theory predicts the critical threshold with 'reasonable accuracy,' but the relative errors in Table II include values of 15.7% and 22.2%. Reporting the range of errors in the abstract and discussion would give a more accurate picture.
- [Sec. III.B.2, Eq. (5)] The definition of M as 'the number of severing events that a newly-created microtubule undergoes' is ambiguous because the preceding text and Eq. (8) also count severing events on lagging daughters. A precise definition stating that each severing event creates exactly one lagging daughter, and clarifying what is included in M, would help the reader.
- [Eq. (14)] The Heaviside theta condition d(2/v+ + 1/v-) in Eq. (14) is not explained. A short derivation or reference would improve readability.
- [Appendix B] The correlation between s_i and ⟨1-δ_{c_i,0}⟩ is stated to be negligible without numerical support. Given that this factorization is used in Eq. (36), a brief test of its validity would strengthen the derivation.
Circularity Check
No significant circularity: the critical threshold emerges from the model equations and is checked against independent full-grid simulations.
full rationale
The paper's central prediction is the bounded-growth threshold p+_crit in p+. This threshold is not postulated or fitted: it is located in full-grid simulations (Fig. 3) and then independently estimated by solving the algebraic condition M(p+)=1 (Eqs. 6-11 and 24). The inputs M+_(1) and S entering Eq. (11) are measured from simulations of single microtubules, not from the amplification/extinction simulations that define p+_crit; they are not tuned to reproduce the threshold, and the p+-dependence enters through analytically derived pcr(p+) and crossover probabilities (Eqs. 22-23). The FPTD approximation (Eq. 14) is checked against independent strip simulations. Self-citations to Dogterom-Leibler [8], Mulder [13], and the authors' earlier work [5,6,9] supply standard model ingredients or prior experimental context; the threshold result does not reduce to any of these citations. The possible concern that m_i in Eq. (8) counts lagging-daughter severings as direct daughters is a modeling-validity issue about whether M is the true reproduction number, not a circular use of the target threshold, and therefore does not raise the circularity score.
Assumptions & free parameters
free parameters (3)
- p+ (probability of rescue-after-severing) =
tuned 0 to 1; critical ~0.36
- Gamma shape k =
7
- Gamma scale theta =
8.5 s
assumptions (8)
- domain assumption Dogterom-Leibler dynamic instability model with constant growth/shrinkage speeds and rates rc, rr.
- standard math Splitting probabilities and lifetime distributions from refs [13,14].
- domain assumption Transverse microtubules form a stable, inert grid with spacing d over the first 500s.
- ad hoc to paper Severing waiting times follow a Gamma distribution with k=7, theta=8.5s.
- ad hoc to paper At most one catastrophe-rescue event occurs before a crossover is reached.
- ad hoc to paper The full grid can be replaced by two transverse microtubules for the analytical theory.
- ad hoc to paper A microtubule cannot be severed twice at the same crossover.
- domain assumption Amplification occurs when the mean offspring number M exceeds 1.
Cite this review
Pith. "Pith review of Critical threshold for microtubule amplification through templated severing." pith.science (2026). https://pith.science/paper/33L54A3V
@misc{pith2026190811144,
author = {Pith},
title = {Pith review of: Critical threshold for microtubule amplification through templated severing},
year = {2026},
howpublished = {\url{https://pith.science/paper/33L54A3V}},
note = {Machine review of arXiv:1908.11144}
}
read the original abstract
The cortical microtubule array of dark-grown hypocotyl cells of plant seedlings undergoes a striking, and developmentally significant, reorientation upon exposure to light. This process is driven by the exponential amplification of a population of longitudinal microtubules, created by severing events localized at crossovers with the microtubules of the pre-existing transverse array. We present a dynamic one-dimensional model for microtubule amplification through this type of templated severing. We focus on the role of the probability of immediate rescue-after-severing of the newly-created lagging microtubule, observed to be a characteristic feature of the reorientation process. Employing stochastic simulations, we show that in the dynamic regime of unbounded microtubule growth, a finite value of this probability is not required for amplification to occur, but does strongly influence the degree of amplification, and hence the speed of the reorientation process. In contrast, in the regime of bounded microtubule growth, we show that amplification only occurs above a critical threshold. We construct an approximate analytical theory, based on a priori limiting the number of crossover events considered, which allows us to predict the observed critical value of the rescue-after-severing probability with reasonable accuracy.
Figures
Figures from the paper (3 more)
Reference graph
Works this paper leans on
-
[1]
Critical point in simulations By tuning the probability of rescue-after-severing p+ from 0 to 1, we observe two different behaviours, see Figure 3A: for lower values of p+ the average number of microtubules exponentially decays in time (extinction), whilst for higher values of p+ the number of microtubules exponentially increases (amplification). It follows...
-
[2]
If the size of such an offspring is, on average, greater than one, the output is amplification
Calculation of the critical point When microtubule is created through a severing event, either it shrinks to length zero and dies, or it is severed a sufficient number of times to create an offspring of new lagging microtubules. If the size of such an offspring is, on average, greater than one, the output is amplification. In other words, if M is the number of...
-
[3]
The first passage time distribution The creation of new crossovers for a microtubule undergoing dynamic instability is intimately linked to a FPTD problem for the same microtubule to reach a target. Here, we face this problem by making use of an approach where we consider all possible legal paths to reach the target, given the knowledge of the time needed ...
-
[4]
One-crossover theory Na¨ ıvely, one can think that once a crossover is created, the probabilityp(1) sev of resolving it with a severing event is given by the competition between two independent events: microtubule lifetime, expressed by the random variable T+ (x) with density function given by Eq. (29), and severing waiting time at the crossover, with ran...
-
[5]
Two-crossovers theory timer starts timer starts timer starts A B Figure 5. Schematic of the full one-crossover theory (A). The newly-created crossover can be resolved either by the shrinkage of the plus end below the crossover itself (lower blue square), or by the severing at crossover (upper blue square). Schematic of the full two-crossovers theory (B). ...
work page 2013
-
[6]
Bruce Alberts, Alexander Johnson, Julian Lewis, Martin Raff, Keith Roberts, and Peter Walter. Molecular Biology of the Cell. Garland Science, 2007
work page 2007
-
[7]
Dynamic instability of microtubule growth
Tim Mitchison and Marc Kirschner. Dynamic instability of microtubule growth. Nature, 312(5991):237–242, 11 1984
work page 1984
-
[8]
MAPs in plant cells: delineating microtubule growth dynamics and organization
John C Sedbrook. MAPs in plant cells: delineating microtubule growth dynamics and organization. Current Opinion in Plant Biology, 7(6):632–640, 12 2004
work page 2004
Show all 19 references
-
[9]
The cortical microtubule array: from dynamics to organization
Ram Dixit and Richard Cyr. The cortical microtubule array: from dynamics to organization. The Plant cell , 16(10):2546–52, 10 2004
2004
-
[10]
A mechanism for reorientation of cortical microtubule arrays driven by microtubule severing
Jelmer J Lindeboom, Masayoshi Nakamura, Anneke Hibbel, Kostya Shundyak, Ryan Gutierrez, Tijs Ketelaar, Anne Mie C Emons, Bela M Mulder, Viktor Kirik, and David W Ehrhardt. A mechanism for reorientation of cortical microtubule arrays driven by microtubule severing. Science (New...
2013
-
[11]
SPR2 protects minus ends to promote severing and reorientation of plant cortical microtubule arrays
Masayoshi Nakamura, Jelmer J Lindeboom, Marco Saltini, Bela M Mulder, and David W Ehrhardt. SPR2 protects minus ends to promote severing and reorientation of plant cortical microtubule arrays. The Journal of cell biology , 217(3):915–927, 3 2018
2018
-
[12]
Tran, R.A
P.T. Tran, R.A. Walker, and E.D. Salmon. A Metastable Intermediate State of Microtubule Dynamic Instability That Differs Significantly between Plus and Minus Ends. The Journal of Cell Biology , 138(1):105–117, 7 1997
1997
-
[13]
Physical aspects of the growth and regulation of microtubule structures
Marileen Dogterom and Stanislas Leibler. Physical aspects of the growth and regulation of microtubule structures. Physical Review Letters, 70(9):1347–1350, 3 1993
1993
-
[14]
CLASP stabilization of plus ends created by severing promotes microtubule creation and reorientation
Jelmer J Lindeboom, Masayoshi Nakamura, Marco Saltini, Anneke Hibbel, Ankit Walia, Tijs Ketelaar, Anne Mie C Emons, John C Sedbrook, Viktor Kirik, Bela M Mulder, and David W Ehrhardt. CLASP stabilization of plus ends created by severing promotes microtubule creation and reorie...
2018
-
[15]
Unnikrishna Pillai
Athanasios Papoulis and S. Unnikrishna Pillai. Probability, random variables, and stochastic processes . Tata McGraw-Hill, 2002
2002
-
[16]
Govindan and William B
Bindu S. Govindan and William B. Spillman. Steady states of a microtubule assembly in a confined geometry. Physical Review E, 70(3):032901, 9 2004
2004
-
[17]
search-and-capture\
Jan W. Vos, Marileen Dogterom, and Anne Mie C. Emons. Microtubules become more dynamic but not shorter during preprophase band formation: A possible \“search-and-capture\” mechanism for microtubule translocation. Cell Motility and the Cytoskeleton , 57(4):246–258, 4 2004
2004
-
[18]
Bela M. Mulder. Microtubules interacting with a boundary: Mean length and mean first-passage times. Physical Review E, 86(1):011902, 7 2012
2012
-
[19]
D. J. Bicout. Green’s functions and first passage time distributions for dynamic instability of microtubules. Physical Review E, 56(6):6656–6667, 12 1997
1997
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.