Interplay between laminin-5 (Ln-5) and its integrin (Int) receptors alpha2beta1, alpha3beta1 and alpha6beta4 has been implicated in the progression and invasion of carcinomas. In this study we found abundant immunoreactivity for chains of Ln-5 (alpha3-beta3-gamma2) and Ln-10 (alpha5-beta1-gamma1), as well as for type VII collagen, in basement membranes (BM) of colorectal adenomas. In carcinomas of all differentiation grades, Lns were seen in tumor BMs, whereas type VII collagen was almost absent. Ln-5 appeared to accumulate along the invading edges of carcinomas, while Ln-10 was mostly absent. Immunoreactivity for Ln al chain, a component of Lns-1 and -3, was not seen in adenomas or carcinomas. Immunoreactivity for alpha2, alpha6, beta1 and beta4 Ints was found in all tumors and that for alpha3 Int in all adenomas and most of the carcinomas, often in colocalization with Ln-5. Immunoblotting of carcinoma tissues showed that the gamma2 chain of Ln-5 was present as typical Mr 105000 and 155000 isoforms. Immunoprecipitation experiments showed production of Ln-5 by cultured colon carcinoma cells. In quantitative cell adhesion experiments, function-blocking MAbs to alpha3 and beta1 Int subunits, but not those to Int alpha2 or alpha6 subunits, significantly inhibited the adhesion of cells to Ln-5. Our results suggest that BM composition in colorectal adenomas reflects the properties of surface epithelial BM of colorectal mucosa. In invading carcinomas, trimeric Ln-5, produced by carcinoma cells, is a major BM component and the cells use the alpha3beta1 Int complex for adhesion to Ln-5.

The production of nidogen by four renal cell carcinoma (RCC) and three pancreatic adenocarcinoma (PAc) cell lines has been studied in cell culture and in xenografted tumours in nude mice. In RCC cells, immunoreactivity for nidogen was seen only after exposure to monensin to induce cytoplasmic accumulation of secretory proteins. In PAc cells, immunoreaction was also detectable in control cells. Immunoblotting of control and monensin‐exposed cells and immunoprecipitation of culture media of radioactively labelled cells demonstrated the production of nidogen polypeptide of Mr ca. 150000 by six of the seven cell lines. Basement membranes (BMs) and stroma of the xenografted tumours derived from these six cell lines demonstrated immunoreactivity for both human and mouse nidogen, as revealed with species‐specific antibodies. The ability of the cells to produce nidogen in vitro and deposit in vivo was positively correlated with high histological grade of the xenografted tumours, although the small number of cell lines studied calls for further studies to confirm this. The distribution of nidogen in human RCC and PAc specimens was also studied by immunohistochemistry. There was strong immunoreactivity for nidogen in tumour stroma, BM of carcinoma cell nests, and endothelial basal lamina, but no conclusions could be drawn regarding histological grade and immunostaining patterns, because stromal production could not be ruled out. The results show that nidogen is produced by human carcinoma cells both in vitro and in vivo. Copyright © 1999 John Wiley & Sons, Ltd.

Retinoids and glucocorticoids are known to have a potential to modulate the expression of transforming growth factor-β (TGF-β). We investigated the effect of oral isotretinoin (13-cis-retinoic acid) on the expression of two distinct isoforms of TGF-β, TGF-β1 and TGF-β2, in suction blister fluid and serum in acne patients. We also investigated the effect of topical glucocorticoid (betamethasone-17-valerate) and age on suction blister fluid TGF-β1 in healthy volunteers. Six weeks of isotretinoin treatment caused a statistically significant 19% increase in suction blister fluid TGF-β1. The suction blister fluid TGF-β2 level remained below the sensitivity level of the immunoassay in many cases. Isotretinoin did not affect the serum TGF-β1 or TGF-β2 level. Betamethasone-17-valerate pretreatment for 3 days twice a day caused a statistically significant 17% decrease in suction blister fluid TGF-β1. The active form of TGF-β1 represented 5% of the total TGF-β1 in suction blister fluid. Our diffusion calculations suggest that all TGF-β1 and TGF-β2 detected in suction blister fluid have diffused from systemic circulation. The increase in suction blister fluid TGF-β1 after isotretinoin treatment seems to be of local origin, while the decrease in suction blister fluid TGF-β1 after glucocorticoid pretreatment seems to be due to glucocorticoid-induced vasoconstriction resulting in decreased diffusion of TGF-β1 from the circulation. Modulation of local interstitial fluid TGF-β1 concentration may be one mechanism by which isotretinoin and glucocorticoids mediate their effects in skin.