Nonenzymatic glycation has been implicated in the pathogenesis of the dysregulated tissue remodeling that characterizes diabetic glomerulopathy, via the formation of advanced glycation end products (AGEs) and their binding to cell surface receptors. Several AGE-binding proteins have been identified so far, including p60, p90, and the adhesive and growth-regulating lectin galectin-3 (Gal-3), the components of the so-called AGE-receptor complex. This study aimed to evaluate the mesangial expression of the AGE-receptor complex and its modulation by the diabetic milieu, both in vivo, in non-diabetic versus streptozotocin-induced diabetic rats, and in vitro, in mesangial cells exposed to either normal glucose (NG) levels (5.5 mmol/l), as compared with high glucose (HG) levels (30 mmol/l) and iso-osmolar mannitol (M), or to native bovine serum albumin (BSA), as compared with glycated BSA with AGE formation (BSA-AGE) and glycated BSA in which AGE formation was prevented by aminoguanidine (BSA-AM). In vivo, Gal-3 protein and mRNA were not detectable in glomeruli from nondiabetic rats until 12 months after initiating the study. On the contrary, in diabetic rats, Gal-3 expression was observed at 2 months of disease duration, and it increased thereafter. Both p60 and p90 immunoreactivities were observed at the glomerular level with slightly increased expression of p90, but not p60, in diabetic versus nondiabetic animals. In vitro, Gal-3 was not detectable in mesangial cells cultured in NG (although it became evident after a certain number of passages in culture), whereas Gal-3 was detectable in cells grown on BSA. Prolonged exposure (2-4 weeks) of mesangial cells to HG but not to M, as well as growing cells on BSA-AGE and, to a lesser extent, BSA-AM, induced or significantly increased the expression of Gal-3, both protein (up to 2.65-fold) and mRNA (up to 3.10-fold) and its secretion in the medium (by approximately 50%). Both p60 and p90 were demonstrated in mesangial cells under NG conditions, and the expression of p90, but not p60, was upregulated by approximately 20% by HG or BSA-AGE. These results indicate that 1) under basal conditions, Gal-3, unlike p90 and p60, is not detectable in the mesangium but becomes expressed with aging and 2) the diabetic milieu induces or upregulates Gal-3 production, whereas it increases only slightly the expression of p90, but not p60. Gal-3 expression or overexpression may modulate the AGE-receptor-mediated events by modifying the function of the AGE-receptor complex. Additionally, it may exert direct effects on tissue remodeling by virtue of its adhesive and growth-regulating properties.