Skin Organoids Link RUNX2 to Fibrotic Remodeling and Candidate Drug Response
A preprint combines human skin organoids, fibroblast perturbation and compound screening to investigate RUNX2-associated fibrosis. The candidate F0565-0303 reduced RUNX2 staining in organoids, while additional assays examined its effects in fibroblasts and mice.
Following Fibroblast States in Human Skin Tissue
A University of Alabama at Birmingham team used induced pluripotent stem cell-derived skin organoids to examine how fibroblasts change during fibrosis. The preprint connects tissue organization with gene-expression analysis and then uses those findings to investigate a candidate intervention. Its human model contains epidermal and dermal-like compartments, allowing stromal changes to be observed within a structured skin environment.
The organoids matured for 80 days before TGF-β stimulation. Untreated cultures were compared with different durations of stimulation at a shared endpoint. Collagen-related and α-smooth muscle actin signals increased during fibrotic remodeling, while single-cell RNA sequencing identified changing fibroblast populations. The authors compared these states with published scleroderma datasets to assess whether the experimental model captured disease-associated transcriptional features.
Testing RUNX2 beyond a Transcript Signature
RUNX2 expression increased in fibrotic organoids and in disease-associated fibroblast populations from the published human datasets. The team also examined immortalized and primary human skin fibroblasts after TGF-β stimulation. These two-dimensional experiments allowed the investigators to perturb RUNX2 directly and evaluate its relationship with profibrotic markers.
Stable RUNX2 knockdown reduced COL1A1 and α-smooth muscle actin expression in stimulated fibroblasts. Bulk RNA sequencing and CUT&RUN profiling then linked RUNX2 activity with programs involving extracellular matrix organization and cell–matrix adhesion. The binding analysis identified occupancy at fibrosis-associated loci, including RUNX1, CCN2 and LOXL2. Those experiments provide a molecular framework for the fibroblast-state observations in the organoids.
The model systems play different roles in that framework. Organoids establish the tissue context and spatial distribution of fibroblast states; fibroblast cultures support more direct genetic perturbation; genomic assays identify candidate regulatory relationships.
Prioritizing and Testing a Small Molecule
The team next screened approximately 113,000 commercially available compounds computationally, using models covering 38 selected targets and a separate fibrosis-associated phenotypic model. Thirteen candidates progressed to testing in stimulated primary human fibroblasts. F0565-0303 reduced RUNX2, COL1A1 and α-smooth muscle actin in this cellular comparison.
Subsequent experiments examined possible mechanisms. At 20 μM, the authors report an approximately 20% reduction in CK2α activity with F0565-0303 relative to the other two candidates, G756-1265 and Y509-4309. Computational structural analysis proposed interactions with CK2α and LOXL2. The drug-associated transcriptional response partly overlapped with RUNX2 knockdown. The binding predictions therefore remain proposed explanations alongside the measured cellular and enzyme results.
In the organoid experiment, F0565-0303 was introduced after fibrotic changes had been induced. RUNX2 immunofluorescence declined in two independent iPSC-derived organoid lines. This tissue experiment measured a specific molecular signal; the collagen and inflammatory responses assessed in primary fibroblasts, and the histological findings in the mouse experiments, are separate evidence supporting the wider investigation.
A Platform for Connecting Discovery with Tissue Responses
The work illustrates a progression from human disease-associated cell states to candidate prioritization and tissue-level testing. RUNX2 served both as a mechanistic lead and as a measurable response in the organoids. The results also show why a screening workflow benefits from retaining its intermediate assays: genetic perturbation, predicted binding, kinase activity and organoid staining each answer a different question about the candidate.
The authors identify missing immune and vascular components as limitations of the skin organoids. They also note that additional functional work is needed to define the full regulatory network and compound specificity. The available human organoid result supports suppression of RUNX2-associated signals under the tested conditions, with further therapeutic development remaining a separate task. This work was reported as an unreviewed preprint.