BGPT: Paper Review: VASP regulates the polar organization of adhesion-associated actin filaments

Explore by Goal

Quick Explanation Copied

Key finding

VASP loss shifts adhesion-associated actin toward less forward (barbed-end–toward-edge) polarity and disrupts adhesion dynamics, while zyxin loss primarily reduces FA-associated actin filament density; tropomyosin-decorated dorsal filaments remain spatially confined and largely independent of VASP/zyxin genotype in these experiments.

Long Explanation

Paper Review (BGPT): VASP regulates the polar organization of adhesion-associated actin filaments

DOI: 10.1101/2025.10.08.681099

Date of provided manuscript text: October 08, 2025 (preprint text in prompt)

Evidence backbone (what the paper actually measures)

Genetic perturbations: CRISPR/Cas9 KO of VASP, zyxin, or both in mouse embryonic fibroblasts (MEFs).
Dynamics readout: live-cell time-lapse of vinculin-Venus (FA turnover) using time-dependent pixel classification (persistent/new/lost).
Structural readout: cryo-ET of membrane-proximal FA regions with automated actin filament annotation (CrYOLO), 3D subtomogram reconstruction (RELION), and actin polarity determination via Actin Polarity Toolbox (APT).
Tropomyosin sub-phenotype: they reconstruct a tropomyosin-actin filament structure and quantify a subset (~6%) of actin filaments identified as tropomyosin-decorated; they then report dorsal localization metrics.

Sample sizes used (from provided text)

Cryo-ET tomograms total: 186
Cells per genotype: WT n=45; VASP−/− n=8; zyxin−/− n=29; DKO n=13.
Actin polarity segments analyzed: 661,369 (from 661,369 segments used for polarity determination).

Primary mechanistic claim tested

VASP is required for maintaining uniform forward actin polarity at FAs and impacts FA dynamics; zyxin (and DKO) further disrupts FA-associated actin filament density and polarity.

Tropomyosin decorates a dorsal filament pool whose localization (~110 nm above substrate) is largely unaffected by VASP/zyxin perturbation in absolute density.

Figure-style reconstruction: forward vs backward actin filament polarity at FAs

Note on units: the text provides absolute forward % and describes increases in backward filaments as “average enrichments of 10%, 20%, 26% over WT,” so the backward series is plotted in that reported form rather than converted to absolute backward %.

Adhesion-associated actin filament density (per µm³)

Tropomyosin-decorated actin: dorsal positioning and independence claim

This plot uses only the scalar values explicitly reported in the provided text: dorsal tropomyosin-actin is localized around ~110 nm above the substrate, and tropomyosin-decorated filaments constitute ~6% of analyzed actin filaments; the text states localization and (absolute) density are unaffected by VASP/zyxin perturbations.

Interpretation (with explicit uncertainty boundaries)

1) VASP affects FA dynamics and polarity more than it reduces FA-associated actin density

The paper reports that VASP loss slows vinculin turnover in adhesion sites (higher proportion/slowest decline of persistent pixels) while newly formed/lost rates are comparable to WT, and that VASP loss decreases the fraction of forward-oriented filaments at FAs (71% WT → 68% VASP−/−).

Crucially, the density of actin filaments at FAs is reported as ~4000 filaments/µm³ for WT and VASP−/−, suggesting that VASP’s polarity effects are not simply a consequence of globally fewer actin filaments at adhesions (in this experimental context).

2) Zyxin loss is associated with reduced FA-associated actin density and exacerbated polarity defects (especially in DKO)

The paper reports that zyxin loss does not significantly alter adhesion dynamics compared to WT in their vinculin turnover readout, yet zyxin KO (and DKO) reduce FA size/vinculin density and reduce dorsal/radial stress fiber densities.

Structurally, they report forward filament fractions decrease (71% WT → 65% zyxin−/−; 63% DKO) and backward filament enrichments rise (10/20/26% over WT for VASP−/−/zyxin−/−/DKO), alongside reduced actin density (WT/VASP−/− ~4000 vs zyxin−/− ~3000 and DKO ~2500 filaments/µm³).

3) Tropomyosin defines a dorsal actin pool that is spatially confined and largely independent of VASP/zyxin perturbations

The paper reports that most tropomyosin-decorated actin filaments localize to the dorsal aspect of the FA-associated actin bundle, ~20 nm above the center of actin volume, with predominant tropomyosin signal at ~110 nm above the substrate; the dorsal localization persists across KO lines, and absolute tropomyosin-decorated filament density is unchanged, while relative fraction increases in mutants due to decreased total actin density.

The mechanistic leap beyond these data is necessarily constrained: the study supports a spatial independence of this dorsal tropomyosin pool from VASP/zyxin, but “functional contractile integration” is a model interpretation rather than a direct measurement of force or tropomyosin dynamics in the provided text.

Skeptical critique (what could mislead, confound, or limit inference)

A) KO context and compensatory biology

The paper itself notes that differences in actin polarity between WT and VASP−/− are “likely due to compensatory activity from other Ena/VASP family members.” That is plausible, but it also means the observed VASP phenotype may be attenuated or partially buffered—limiting mechanistic specificity.

B) Static snapshots vs dynamics

The cryo-ET polarity/dorsal localization are structural snapshots, while the dynamics conclusions come from fluorescence turnover imaging. The combination is informative, but causality between “polarity shift” and “turnover change” can be underdetermined if multiple processes change concurrently (e.g., membrane-proximal actin assembly vs clutch engagement).

C) Sampling and selection: “membrane-proximal regions of each FA” and subset enrichment

Their cryo-ET is reported as membrane-proximal regions of each FA, and their tropomyosin-positive subset is ~6% of analyzed filaments. Both facts are fine experimentally, but they raise the possibility that region selection or detection sensitivity could bias which filaments are detectable/classifiable (especially for tropomyosin-decorated states).

D) Interpretation hierarchy: “independence” depends on what was measured

The claim that tropomyosin localization is “regulated independently” of VASP/zyxin is strongly supported in terms of the provided structural metrics (absolute density unchanged, dorsal localization persists, orientation distributions mirrored regardless of genotype). However, the study does not directly test whether VASP/zyxin alter tropomyosin biochemical state, isoform composition, or binding kinetics—so some molecular layers of the mechanism remain unresolvable from the reported readouts.

Reproducibility & data availability (from provided text)

They deposit EM structures of actin filaments and tropomyosin-actin filaments in EMDB (EMD-17742, EMD-17744) and tomograms for genotypes (EMD-17745, EMD-17746, EMD-17750, EMD-17752).
The Actin Polarity Toolbox code is available on GitHub at the provided URL.
Cryo-ET pipeline details (CrYOLO extraction, RELION refinement/classification, IMOD reconstruction, APT scoring thresholds like ccs(F) ≥ 0.6) are described, enabling technical replication of the analytic logic, though exact parameter defaults beyond the text can still matter.

One falsification-style check (what would most challenge the paper’s main claims)?

Rescue experiments: if reintroducing VASP or zyxin into the respective KO backgrounds does not restore the forward/backward polarity distribution and/or actin density patterns at FAs, then the genetic causality would be weakened. (This is an inference about what would be falsifying; the provided text primarily reports KO phenotypes rather than rescue.)
Tropomyosin invariance robustness: if more exhaustive sampling or alternative classification thresholds reveal that tropomyosin-positive filaments shift dorsal positioning or absolute density across genotypes, then the “independent” interpretation would be undermined.

Explore author-specific reviews on BGPT

This will let an AI agent iteratively analyze/triangulate the cryo-ET structural metrics and generate additional plots from the values explicitly provided in the manuscript text.

Feedback:

Updated: March 27, 2026