Vascular and parenchymal basement membranes (BMs) are thickened in diabetes, but alterations in individual BM components in diabetic eyes, especially in diabetic retinopathy (DR), are obscure. To identify abnormalities in the distribution of specific constituents, we analyzed cryostat sections of human eyes obtained at autopsy (seven normal, five diabetic without DR, and 13 diabetic with DR) by immunofluorescence with antibodies to 30 BM and extracellular matrix components. In non-DR eyes, no qualitative changes of ocular BM components were seen. In some DR corneas, epithelial BM was stained discontinuously for laminin-1, entactin/nidogen, and alpha3-alpha4 Type IV collagen, in contrast to non-DR corneas. Major BM alterations were found in DR retinas compared to normals and non-DR diabetics. The inner limiting membrane (retinal BM) of DR eyes had accumulations of fibronectin (including cellular) and Types I, III, IV (alpha1-alpha2), and V collagen. The BM zone of new retinal blood vessels in neovascularized areas accumulated tenascin and Type XII collagen, whereas normal, diabetic, and adjacent DR retinas showed only weak and irregular staining. In preretinal membranes, perlecan, bamacan, and Types VI, VIII, XII, and XIV collagen were newly identified. Diabetic BM thickening appears to involve qualitative alterations of specific BM markers at an advanced disease stage, with the appearance of DR.

SUMMARYCorneas of diabetic patients have abnormal healing and epithelial adhesion, which may be due to alterations of the corneal extracellular matrix (ECM) and basement membrane (BM). To identify such alterations, various ECM and BM components and integrin receptors were studied by immunofluorescence on sections of normal and diabetic human corneas. Age-matched corneas from 15 normal subjects, 10 diabetics without diabetic retinopathy (DR), and 12 diabetics with DR were used. In DR corneas, the composition of the central epithelial BM was markedly altered, compared to normal or non-DR diabetic corneas. In most cases the staining for entactin/nidogen and for chains of laminin-1 ( ␣1 ␤1 ␥1) and laminin-10 ( ␣5 ␤1 ␥1) was very weak, discontinuous, or absent over large areas. Other BM components displayed less frequent changes. The staining for ␣3 ␤1 (VLA-3) laminin binding integrin was also weak and discontinuous in DR corneal epithelium. Components of stromal ECM remained unchanged even in DR corneas. Therefore, distinct changes were identified in the composition of the epithelial BM in DR corneas. They may be due to increased degradation or decreased synthesis of BM components and related integrins. These alterations may directly contribute to the epithelial adhesion and wound healing abnormalities found in diabetic corneas. (J Histochem Cytochem 46:1033-1041, 1998)

These data suggest that TIMP-1 and TIMP-2 are important for scar formation and corneal remodeling, since they were found in increased amounts at radial keratotomy incision sites and keratoconus scars. The significance of the focal stromal defects in TIMP-3 staining, associated with absence of Bowman's layer on keratoconus corneas, needs to be elucidated. At the stages of disease examined in this study, gelatinase B may not play a significant role in these pathological processes, since it was not seen in any of the corneas examined.

The aim of this study was to evaluate whether recombinant human insulin-like growth factor I (rhIGF-I) could attenuate or prevent diaphragm (DIA) fiber atrophy with corticosteroid (CS) administration to emphysematous (EMP) hamsters. DIA muscle IGF-I responses to CS administration with and without exogenous rhIGF-I administration were evaluated. Three groups were studied: 1) EMP; 2) EMP + triamcinolone (T; 0.4 mg.kg-1.day-1 im); and 3) EMP + T + IGF-I (600 microg/day by constant infusion). After 4 wk, the DIA was analyzed histochemically and biochemically (IGF-I mRNA levels by RT-PCR and endogenous and exogenous IGF-I peptide levels immunochemically). Body weights of EMP-T progressively decreased, while those of EMP and EMP-T-IGF-I remained stable despite similarly reduced food intake in both T groups. DIA weight was reduced with T but preserved with rhIGF-I infusion. DIA fiber proportions were similar among the groups. The cross-sectional areas of types I, IIa, and IIx fibers were reduced (17 to 31%) with T administration but unchanged with rhIGF-I infusion. DIA IGF-I mRNA levels were similar across all groups. By contrast, the endogenous DIA IGF-I levels were reduced (41%) in the EMP-T-IGF-I animals. Total DIA IGF-I levels (endogenous + exogenous) were still significantly reduced. IGF-I immunoreactivity confirmed this reduction in all DIA fibers. We conclude that DIA fiber atrophy with T was completely prevented by exogenous rhIGF-I administration. This effect was likely mediated by the pharmacological influences of exogenously administered rhIGF-I. We speculate that this results from increased bioavailability of free IGF-I to react with muscle receptors. Reduced endogenous IGF-I levels in the DIA likely reflect a negative-feedback influence. These results may have important clinical implications for treatment options to offset the adverse effects of CS on the respiratory muscles in patients with chronic lung disorders.