Fibrosis Signal BMP4 Drains Kidney Cells’ Mitochondrial Power in Diabetic Disease

Diabetic nephropathy, the kidney complication that quietly drives many patients toward dialysis, has long been framed as a story of scarring: excess sugar in the blood prods glomerular cells to pile up extracellular matrix until the delicate filtration units stiffen and fail. A new study from Tokushima University adds a metabolic twist to that narrative. Reporting in Frontiers in Clinical Diabetes and Healthcare, the researchers show that a well-known pro-fibrotic signal, bone morphogenetic protein 4, does not merely instruct mesangial cells to produce more matrix. It also strips those cells of a key mitochondrial quality-control protein called Parkin, leaving their power plants damaged, depleted, and underperforming.

Mesangial cells are the structural scaffolding of the glomerulus, the tuft of capillaries where blood is filtered. Keeping that scaffolding intact is energetically expensive, and the cells rely heavily on mitochondria to supply ATP. Earlier work from the same group had established that BMP4, a member of the TGF-beta signaling family, rises in diabetic kidneys and drives mesangial cells to overproduce type IV collagen and other matrix components through the ALK-Smad1 pathway. What remained unknown was whether BMP4 also interfered with the cells’ energy metabolism, and if so, by what mechanism.

To find out, the team turned to a classic animal model: mice given low-dose streptozotocin over five consecutive days to destroy their insulin-producing beta cells, producing sustained hyperglycemia. Six months after the first injection, mice with fasting glucose above 400 milligrams per deciliter were analyzed. Their kidneys showed the expected hallmarks of diabetic nephropathy, including elevated urinary albumin excretion and mesangial matrix expansion visible with silver staining. But electron microscopy also revealed something subtler: the mitochondria inside mesangial cells were swollen, and their internal cristae, the folded membranes where the electron transport chain operates, were disrupted.

The molecular readouts told a consistent story. Mitochondrial DNA copy number, a proxy for mitochondrial abundance, fell in the diabetic kidneys, as did the activity of cytochrome c oxidase, a key enzyme complex of the respiratory chain. Expression of Pink1 and Park2, the genes encoding the PINK1-Parkin mitophagy machinery that tags damaged mitochondria for destruction, was downregulated. So were Ppargc1a, the gene for the master mitochondrial biogenesis regulator PGC-1alpha, and its downstream targets Tfam, Cox4i1, and Atp5a1. Protein levels of Parkin and TOM20, a marker of the outer mitochondrial membrane, dropped in parallel.

In cultured mouse mesangial cells, the researchers could isolate BMP4’s contribution directly. Bathing the cells in recombinant BMP4 for 24 hours reproduced nearly the entire mitochondrial phenotype seen in the diabetic animals: mtDNA copy number fell, cytochrome c oxidase activity declined, oxygen consumption dropped, ATP production shrank, mitochondrial membrane potential collapsed, and cellular antioxidant capacity weakened. Parkin and TOM20 protein levels fell alongside the mRNA changes. The team also confirmed that high glucose alone, at 30 millimolar, impairs mesangial mitochondrial function, consistent with earlier reports in human mesangial cells, but BMP4 was tested as a specific upstream mediator whose levels rise in the diabetic kidney.

The causal chain became clearer when the researchers manipulated the pathway. Adding LDN-193189, a selective inhibitor of the BMP type I receptors ALK2 and ALK3, prevented BMP4 from suppressing Parkin and the mitochondrial genes, and preserved mtDNA copy number and membrane potential. It also blocked BMP4’s induction of the matrix genes Col4a1 and Col1a1, confirming that the inhibitor was effectively shutting down BMP signaling. Conversely, silencing Parkin with siRNA in otherwise healthy mesangial cells mimicked the BMP4 phenotype, reducing mtDNA copy number, respiratory chain activity, oxygen consumption, ATP levels, and membrane potential, and downregulating PGC-1alpha and its targets.

The most striking experiment involved Parkin overexpression. The team generated mesangial cell lines that constitutively produce Parkin from a transfected gene. In these cells, BMP4 could no longer push Parkin levels down, and the downstream consequences largely evaporated: PGC-1alpha and mitochondrial gene expression stayed up, mtDNA copy number was maintained, cytochrome c oxidase activity and ATP production were preserved. Because the overexpression was so strong, the authors note that quantitative comparisons of Parkin protein between overexpressing and control cells were not appropriate, and they caution that constitutive expression may alter normal regulatory modes, leaving open the possibility of post-transcriptional effects.

The findings position BMP4 as a dual-threat signal in diabetic kidney disease. On one hand, it promotes the structural pathology of mesangial expansion through Smad1-mediated matrix production. On the other, it appears to erode the metabolic foundation of the very cells doing that remodeling. The authors suggest that BMP4 may therefore be involved in both the structural and metabolic alterations of diabetic nephropathy, and that disrupted mitochondrial homeostasis accompanied by decreased Parkin expression could represent a candidate therapeutic pathway.

Therapeutically, the emphasis falls on restoring Parkin function rather than supplying the protein itself, since Parkin works as an intracellular E3 ubiquitin ligase. Prior studies have shown that the mitochondria-targeted antioxidant MitoQ can restore PINK1-Parkin expression through the Nrf2 pathway and ameliorate renal injury in diabetic kidney disease, while FoxO1 activation promotes PINK1-Parkin-mediated mitophagy and protects podocytes. Whether such interventions also influence upstream BMP4 expression remains to be determined.

The authors are careful about the limits of their short communication. The link between Parkin reduction and mitochondrial changes was assessed mainly at the level of gene expression and functional phenotypes, without directly measuring mitophagy flux or the intermediate molecular steps connecting Parkin to PGC-1alpha regulation. BMP signaling was probed only with a pharmacological inhibitor, so the specific receptor subtypes and downstream transcription factors responsible for Parkin suppression remain unidentified. The in vivo analysis examined whole kidney tissue rather than isolated mesangial cells, the temporal relationship between Parkin loss and mitochondrial dysfunction was not established, and no in vivo interventions were tested. Reactive oxygen species and inflammatory cytokines were not quantified, and a direct comparison between high-glucose and BMP4 stimulation was not performed. Still, the convergence of evidence from diabetic mouse kidneys, BMP4-treated cells, Parkin knockdown, and Parkin overexpression makes a compelling case that a fibrosis signal and a mitochondrial quality-control protein are entangled in the progression of diabetic kidney disease, and that untangling them may point toward new ways to protect the failing kidney.

Subject of Research: Role of BMP4-induced Parkin downregulation in mitochondrial dysfunction of mesangial cells in diabetic nephropathy

Article Title: Bone morphogenetic protein 4-associated mitochondrial dysfunction in diabetic nephropathy involves Parkin downregulation

Article References: Tamaki, M., Tominaga, T., Hasegawa, K., Kida, T., Miyakami, S., Shimizu, I., Yamaguchi, S., Minato, M., Nagai, K., & Wakino, S. (2026). Bone morphogenetic protein 4-associated mitochondrial dysfunction in diabetic nephropathy involves Parkin downregulation. Frontiers in Clinical Diabetes and Healthcare, 7, Article 1860424. https://doi.org/10.3389/fcdhc.2026.1860424

Image Credits: AI Generated

DOI: 10.3389/fcdhc.2026.1860424

Keywords: diabetic nephropathy, BMP4, Parkin, mitochondrial dysfunction, mesangial cells, mitophagy, PGC-1alpha, kidney disease, fibrosis, ATP production, ALK inhibitor, streptozotocin model