Glomerular nodular lesions, known as Kimmelstiel-Wilson nodules, are a pathological hallmark of progressive human diabetic nephropathy. We have induced severe diabetes in pigs carrying a dominant-negative mutant hepatocyte nuclear factor 1-alpha (HNF1α) P291fsinsC, a maturity-onset diabetes of the young type-3 (MODY3) gene in humans. In this model, glomerular pathology revealed that formation of diffuse glomerular nodules commenced as young as 1 month of age and increased in size and incidence until the age of 10 months, the end of the study period. Immunohistochemistry showed that the nodules consisted of various collagen types (I, III, IV, V and VI) with advanced glycation end-product (AGE) and Nε-carboxymethyl-lysine (CML) deposition, similar to those in human diabetic nodules, except for collagen type I. Transforming growth factor-beta (TGF-β) was also expressed exclusively in the nodules. The ultrastructure of the nodules comprised predominant interstitial-type collagen deposition arising from the mesangial matrices. Curiously, these nodules were found predominantly in the deep cortex. However, diabetic pigs failed to show any of the features characteristic of human diabetic nephropathy; e.g., proteinuria, glomerular basement membrane thickening, exudative lesions, mesangiolysis, tubular atrophy, interstitial fibrosis, and vascular hyalinosis. The pigs showed only Armanni-Ebstein lesions, a characteristic tubular manifestation in human diabetes. RT-PCR analysis showed that glomeruli in wild-type pigs did not express endogenous HNF1α and HNF1β, indicating that mutant HNF1α did not directly contribute to glomerular nodular formation in diabetic pigs. In conclusion, pigs harboring the dominant-negative mutant human MODY3 gene showed reproducible and distinct glomerular nodules, possibly due to AGE- and CML-based collagen accumulation. Although the pathology differed in several respects from that of human glomerular nodular lesions, the somewhat acute and constitutive formation of nodules in this mammalian model might provide information facilitating identification of the principal mechanism underlying diabetic nodular sclerosis.