Here is a question that has nagged at those of us who live in the world of the threatened limb: when we see a circulating microRNA go up or down in a patient with a diabetic foot ulcer and peripheral artery disease (PAD), are we watching a driver of the disease, or just smoke from a fire lit somewhere else? Almost every study to date has been observational, which means it cannot tell the two apart. A new preprint takes a genuinely clever swing at that problem.
The authors pull together publicly available transcriptomic data (about 312 samples), run differential expression, and then do something most miRNA papers do not: they apply two-sample Mendelian randomization, using inherited genetic variants as instruments to ask whether a given miRNA is causally upstream of PAD rather than merely associated with it. Out of 63 dysregulated miRNAs, six survive as putative causal regulators.
Those six converge on a tidy little hub of usual suspects—VEGFA, AKT1, and HIF1A—anchoring the pathways we would expect to misbehave in a chronically ischemic, inflamed foot: PI3K-AKT signaling, the hypoxia response, and NF-kB-driven inflammation. miR-21-5p and miR-210-3p come out as risk-conferring; miR-126-3p reads as protective (its loss disinhibits the angiogenic machinery we badly want working in these patients).
Then comes the part that makes this fun to put in front of weird people with weird ideas: computational drug repurposing. Feed the hub-gene signature through drug-signature databases, dock the top candidates against the hub proteins in AutoDock Vina, and you get a short list of off-the-shelf agents—metformin, rapamycin, dexamethasone, atorvastatin, tofacitinib—with binding affinities reaching about minus 9.8 kcal/mol. Metformin shows the most consistent multi-target engagement across the network, and the authors make a reasonable argument that its vascular benefits in diabetes may run partly through miRNA-dependent mechanisms we have been under-crediting.
I like this paper for what it is: a coherent, testable scaffold that reframes circulating miRNAs as potentially actionable targets rather than passive biomarkers. That is a useful reframe. But let us keep our feet on the ground (so to speak). The genetic instruments come largely from European-ancestry cohorts, which is exactly the wrong place to be confident when the global DFU burden falls hardest on populations that are barely represented. The docking affinities are structural plausibility, not clinical endpoints. And there is no wet-lab validation here at all. This is a hypothesis-generating machine, and it should be read as one.
Still—as a way of deciding which old drug to drag into a diabetic-foot trial next, and why, this is a genuinely interesting potential advance. The honest next step is the unglamorous one: functional validation in primary DFU tissue, then an actual prospective study in a diverse cohort, ideally against endpoints we care about like wound healing and amputation-free survival. If even one of these repurposed candidates holds up, the computational detour will have paid for itself.
de Oliveira Andrade LJ, Matos de Oliveira GC, et al., “Circulating microRNAs as Putative Causal Regulators of Peripheral Arterial Disease in Diabetic Foot: An Integrative Multi-Omics and Drug Repurposing Study,” Research Square (preprint), 2026. DOI: 10.21203/rs.3.rs-9971778/v1
#DiabeticFoot #microRNA #miRNA #PAD #PeripheralArteryDisease #DrugRepurposing #Metformin #MendelianRandomization #Angiogenesis #LimbSalvage #MultiOmics
Thank you, Dr. Armstrong, for this material. Being a chemical guy. I enjoy the scientific debate and healthy wonder. Again, thank you for adding summer exploring project number 7 to my list of items to be researched and completed by September 4th 2026.