Experiments were carried out to evaluate the role of convection in the removal of large molecules from brain interstitial fluid. Radiolabeled test compounds were injected into the caudate nucleus of anesthetized rats through a guide cannula implanted 1 wk previously and the concentrations of isotope in brain and cerebrospinal fluid (CSF) determined at various times after injection. Control studies with 22Na indicate that the permeability of the blood-brain barrier is normal in tissue surrounding the intracerebral injection cannula. For 69,000 dalton serum albumin, 4,000 dalton polyethylene glycol, and 900 dalton polyethylene glycol, clearance from brain approximates a single exponential decay with half times of disappearance of 12.2, 12.6, and 14.4 h, respectively. Similarly in efflux rate, despite a fivefold difference in diffusion coefficient, is consistent with convective losses from brain, and the maximal rate of interstitial fluid removal estimated on the basis of these data is 0.11 microliter.g brain-1.min-1. Only 10-20% of total efflux is into bulk CSF withdrawn from the cisterna magna.

Fibronectin expression is of considerable importance in normal and fibrotic liver. Plasma fibronectin levels are correlated with good prognosis in liver failure, and cellular fibronectin plays a crucial role in fibrogenesis. In this study, we observed that the H4II rat hepatoma cell line does not express fibronectin. Furthermore, a recombinant vector (pFGH) containing the promoter elements of the fibronectin gene showed no promoter function when transfected into this cell line. However, pFGH was actively expressed in L-cells and rat skin fibroblasts, cell types that express large amounts of endogenous fibronectin. To study the mechanisms regulating fibronectin expression, we evaluated the transcriptional regulatory elements of the rat fibronectin gene by mutational analysis and DNA-protein binding studies. Deletional mutation analysis showed that the sequences between positions -164 and -90 are essential for promoter activity. This region contains the consensus binding sites for CCAAT and the cyclic AMP-responsive element. Gel retardation assays demonstrated that although the binding activity to the CCAAT site at -140 was essentially the same as that in extracts from L-cells, hepatoma cells and rat livers, substantially greater amounts and different patterns of binding to the adjacent cyclic AMP-responsive element were observed in the extracts from the expressing L-cells and rat livers compared with those in the nonexpressing hepatoma cell nuclear extracts. Furthermore, mutagenesis of the cyclic AMP-responsive element site dramatically reduced promoter activity in transient transfection assays. The cyclic AMP-responsive element at position -160 appears to play an important role in the constitutive expression of the rat fibronectin gene.

No abstract