# Integrative Mendelian randomization and experimental validation prioritize KLF4 in the gut microbiota–pyroptosis–barrier axis of ulcerative colitis

## TL;DR

The study identifies KLF4 as a key player in the gut microbiota-inflammation-barrier pathway in ulcerative colitis, suggesting it could be a new biomarker or treatment target.

## Contribution

The novel contribution is the integrative use of genetic and experimental approaches to prioritize KLF4 as a central mediator in ulcerative colitis.

## Key findings

- KLF4 is downregulated in UC and associated with pro-inflammatory signatures.
- KLF4 overexpression in a DSS model reduced disease activity and improved gut barrier integrity.
- Higher mucosal KLF4 levels predict response to anti-TNF-α therapy in UC patients.

## Abstract

Ulcerative colitis (UC) arises from complex crosstalk between gut microbiota, epithelial barrier integrity, and inflammatory cell death, yet causal mediators along this axis remain poorly defined. We aimed to delineate microbiota–pyroptosis–UC pathways and functionally validate key effectors, with a focus on KLF4.

A multistage framework integrating genome-wide association studies of gut microbiota (MiBioGen), plasma proteomics (deCODE), and UC (UK Biobank) was constructed to perform two-sample Mendelian randomization (MR). Pyroptosis-related proteins were screened for causal associations with UC, followed by MR of UC-associated microbial taxa on these proteins and two-step mediation analysis. KLF4 was further evaluated using bulk and single-cell transcriptomic datasets, including virtual knockout network perturbation. Its clinical relevance was tested in two cohorts of UC patients receiving anti-TNF-α therapy. Finally, the KLF4 function was validated in a dextran sulfate sodium (DSS)-induced colitis model with systemic AAV9-mediated KLF4 overexpression.

MR identified 35 pyroptosis-related plasma proteins and 23 microbial taxa with putative causal effects on UC. Mediation analysis highlighted MAPK11, PTEN, and KLF4 as dominant intermediates linking specific taxa to UC risk. KLF4 was consistently downregulated in UC, and low KLF4 expression was associated with enrichment of pro-inflammatory and immune-activation signatures. Higher mucosal KLF4 levels predicted response to anti-TNF-α therapy with moderate discriminatory performance. In DSS-induced colitis, KLF4 overexpression mitigated weight loss and disease activity, preserved colon length, improved histology, and reduced myeloperoxidase activity. KLF4 restored Claudin-1, Occludin, and Zo-1; suppressed Claudin-2; decreased intestinal permeability; limited gasdermin-D (GSDMD) cleavage; lowered IL-1β/IL-18 levels; and reshaped splenic leukocyte composition.

Our integrative genetic and experimental data position KLF4 as a central node in a gut microbiota–pyroptosis–barrier axis in UC, supporting KLF4 as a promising biomarker and therapeutic target for the precision management of UC.