BGPT: Paper Review: Gut mucin fucosylation dictates the entry of botulinum toxin complexes

Fuel Your Discoveries

Quick Explanation Copied

Core finding

The paper argues that terminal α1,2-fucosylation of intestinal mucin (linked to FUT2 “secretor” status) changes the carbohydrate-binding spectrum of the botulinum toxin complex hemagglutinin (HA), thereby switching mucin penetration and intestinal entry routes (enterocytes vs microfold cells), which ultimately changes oral toxicity.

Evidence is built from mucin-binding/penetration assays, enzymatic removal of terminal glycans, FUT2-null mice, and structural/biochemical binding data.

Long Explanation

Paper Review (skeptical, evidence-based): Gut mucin fucosylation dictates the entry of botulinum toxin complexes

DOI: 10.1038/s41467-025-65384-w · Nature Communications (2025)

Key mechanistic claim (what the data are meant to show)

Hyper-oral-toxic L-PTC/B-Okra shows greater mucus penetration than non-hyper-oral-toxic L-PTC/A62A, and shifts intestinal entry toward enterocytes rather than microfold (M) cells.
Whether L-PTCs are trapped in mucin is primarily governed by HA (hemagglutinin) carbohydrate-binding activity.
Terminal α1,2-fucosylation (rather than GalNAcylation) is presented as a key glycan determinant that modulates HA binding/behavior; enzymatic removal increases HA/B-Okra binding to mucin.

Figure A — Reported competitive binding (IC50) for HA1/A62A vs HA1/B-Okra

Source values are directly taken from the paper’s reported IC50 comparisons for lactose vs α1,2-fucosyllactose.

Figure B — Reported KD change interpretation (multivalency/ligand specificity)

The paper reports a ~2.4-fold Kd increase for HA1/B-Okra when comparing lactose vs α1,2-fucosyllactose, while HA1/A62A shows similar affinities for lactose and α1,2-fucosyllactose.

Figure C — Mouse oral challenge design summary (group sizes)

The paper’s FUT2-null comparisons are: L-PTC/A62A (WT n=6, Fut2−/− n=4) and L-PTC/B-Okra (WT n=13, Fut2−/− n=9).

Figure D — Mechanistic model (as stated in the paper)

L-PTC/A62A: captured by α1,2-fucosylated mucin, resulting in mucus trapping and entry via M cells.
L-PTC/B-Okra: has a carbohydrate-binding spectrum enabling it to penetrate fucosylated mucus, leading to entry via enterocytes.
Host glycosylation tuning: changing terminal α1,2-fucose (e.g., through FUT2 status in mice; secretor status in humans; or enzymatic removal) changes toxin penetration probability and thus oral toxicity.

Evidence map (what each experimental block supports)

Claim tested	Experiment(s)	What it supports	Main uncertainty / skeptical note
Route difference (enterocytes vs M cells)	Ligated intestinal loops; immunofluorescence co-visualization with MUC2/GP2; binding/endocytosis readouts	Linking L-PTC/B-Okra to enterocyte uptake, and L-PTC/A62A to M-cell targeting	Imaging is descriptive; causality still depends on separating binding vs post-binding trafficking and whether imaging conditions match physiological ingestion dynamics.
Mucus barrier role	NAC mucus-depletion model; compare binding and oral toxicity shifts	NAC increases ability of L-PTC/A62A to bind villous epithelium and increases susceptibility, while NAC does not alter L-PTC/B-Okra toxicity	NAC can have multiple biological effects; paper argues via mucus liquefaction/inhibition of mucin synthesis, but off-target impacts on barrier integrity remain a possible confound.
HA determines mucin trapping/penetration	ELISA binding to mucin; chimeric L-PTCs swapping HA components	Oral toxicity depends on HA differences rather than BoNT/NTNHA when HA is swapped	Chimeras are powerful but may perturb complex assembly stability, multimerization, or effective concentrations in vivo even if “comparable BoNT activity” is reported.
α1,2-fucosylation is the key glycan determinant	Competition ELISAs with carbohydrates/lectins; glycan microarrays; AfcA and NAGA enzymatic removal; Fut2−/−	AfcA increases HA/B-Okra binding to mucin; Fut2−/− reduces susceptibility to oral L-PTC/B-Okra	Terminal glycan removal can alter global mucin properties (charge, hydration, steric accessibility). The paper isolates terminal α1,2-fucose vs GalNAc effects, but residue redistribution effects could still exist.
Structural mechanism	X-ray crystallography/ligand complexes; thermal shift assays	HA pocket architecture is consistent with differential acceptance/extension of α1,2-fucosylated ligands	Structures are static snapshots; multivalent mucin context is dynamic and may involve additional factors (mucus mesh properties, steric hindrance, co-factors).

Critical appraisal (what looks strong vs what remains uncertain)

Strengths

Multi-layered mechanistic triangulation: binding assays (mucin/PGM), penetration assays (Transwell/mucin-coated inserts), in vivo oral toxicity, glycan enzymatic editing (AfcA/NAGA), and structural ligand context are combined to argue causality, not just correlation.
HA-swap chimera logic is a relatively direct method to attribute oral toxicity differences to a specific component (HA) of the toxin complex.
Specific terminal glycan focus: the paper contrasts α1,2-fucose removal (AfcA) with α-GalNAc removal (NAGA) and finds differential effects, which reduces the plausibility that the phenotype is purely “generic mucin deglycosylation.”

Limitations / skeptical points

Human relevance is suggestive, not established. The paper discusses secretor status and FUT2 polymorphisms, but the central in vivo causality is demonstrated in mice. The paper notes epidemiological uncertainty.
Barriers are multi-factorial. NAC mucus depletion may alter more than fucose exposure, including physical mesh properties and barrier integrity. Even with a consistent pattern, the mechanistic attribution to fucosylation alone could be incomplete.
Static-binding assays vs dynamic mucin environments. ELISAs/microarrays measure binding to simplified glycans/mucins; mucin is a heterogeneous, densely packed, hydrated gel with dynamic turnover. Differences could arise from accessibility/sterics and not only residue recognition. The paper does address penetration in mucin-coated systems, which helps, but complete equivalence to in vivo mucus is still uncertain.
Sample sizes are modest for some in vivo comparisons. While several experiments are repeated, the FUT2 oral challenge group sizes are limited (e.g., n=4 for Fut2−/− vs n=6 for L-PTC/A62A). This affects precision for effect estimates, especially for non-significant or borderline differences (the paper reports at least one non-significant trend in humans).

What would most disprove or substantially revise the model?

Findings showing that terminal α1,2-fucosylation is not rate-limiting for mucus penetration in more physiologically relevant intestinal mucus (e.g., if mucin glycan editing changes fucose but toxins still penetrate similarly and enter via the same routes).
Evidence that HA binding specificity changes in vitro do not predict entry route or toxicity in vivo when toxin dosing, complex integrity, and mucus physical properties are controlled.

(These are model-critical falsification targets consistent with the paper’s causal chain.)

BGPT “go deeper” actions (bespoke next questions)

Author reviews (bespoke BGPT links)

Feedback:

Updated: April 14, 2026