With regard to tubulointerstitial fibrogenesis we are left with a variety informational gaps regarding nearly all aspects of this clinically important process. Table 2 summarizes a generalized version of fibrogenesis based primarily on investigations in other organs. Extrapolation of data obtained with other fibrogenic systems is useful, but only in so far as it motivates us to adapt and test many of the experimental principles within in the context of the kidney. This begins with a comprehensive examination of the in vivo state, the establishment of adequate animal models, and the dissections of the process in vitro. Key areas for the future are the characterization of the signals involved, the cellular responses to these signals, and the variations in interactions produced by differing inciting fibrogenic conditions.

The

One of several postulated roles for tissue transglutaminase (tTG) is the stabilization and assembly of extracellular matrix via peptide cross-linking. We previously determined that tTG activity increased in an animal model of hepatic fibrogenesis and in human liver disease. To further study the role of tTG in liver disease, we initiated investigations into the effect of a proinflammatory mediator, tumor necrosis factor (TNF)-α, on tTG activity in cultured liver cells. Treatment of human Hep G2 cells with 1 ng/ml TNF-α increased [14C]putrescine cross-linking to cellular proteins. An increase in tTG mRNA content was observed 1 h after addition of TNF-α, and levels of tTG mRNA remained elevated after 24 h. Hep G2 cells, transiently transfected with a luciferase reporter containing 1.67 kb of the human tTG promoter, showed an increase in reporter activity after addition of TNF-α. Gel shift experiments using nuclear extracts from TNF-α-treated cells and oligonucleotides containing the tTG nuclear factor (NF)-κB motif revealed increased binding, concordant with mRNA data. Transient transfections with a truncated reporter construct lacking the tTG NF-κB sequence showed an attenuated response to TNF-α treatment. Similar responses were seen in stably transfected HeLa cells. Primary hepatocytes isolated from a trangenic mouse line containing the mouse tTG promoter driving the β-galactosidase reporter, show similar time-dependent increases in promoter activity when treated with TNF-α. Furthermore, Hep G2 cells are incapable of upmodulating tTG promoter reporter activity in the presence of TNF-α when those cells overexpress a transdominant, negative mutant NF-κB subunit. Because TNF-α expression is upregulated in hepatic inflammation, the data suggest TNF-α-mediated increases in tTG expression may play an important role in the process of hepatic fibrogenesis.

The gene encoding the fast skeletal isoform of the chick troponin I (sTnI) protein has been sequenced and its organization into exons and introns established. The gene is approximately 4.5 kb in length and composed of 8 exons, the first of which contains solely 5' untranslated sequence. In addition to its major mRNA product, there is evidence that the sTnI gene encodes a second mRNA, present at low abundance levels in embryonic skeletal muscle. Sl nuclease protection and primer extension experiments indicate that the low abundance mRNA is initiated approximately 47 nucleotides upstream of the major transcriptional initiation site. Both mRNAs appear to encode identical sTnI polypeptides. A comparison of nucleotide sequence in the 5' flanking region of several muscle-specific genes, including the sTnI gene, reveals a heptanucleotide consensus sequence, 5'-CATTCCT-3', which is conserved in the 5' flanking regions of many vertebrate contractile protein genes.

Increased glomerular collagen IV mRNA in streptozotocin-diabetic rats and stimulation of matrix transcripts by high glucose levels in short-term mesangial cell culture provide evidence that stimulation of matrix synthesis is important in early diabetic glomerulopathy. To test whether transcriptional modulation of collagen IV genes is operative, we stably transfected a murine mesangial cell line with a "minigene" expressing luciferase driven by 5'-flanking and first-intron regions of the murine COL4A1 gene to assess the response to high glucose and the associated signaling pathway. Luciferase activity was stimulated in a dose- and time-dependent manner [near-maximal stimulation in 450 mg/dl glucose (G450) was more than twofold the level in 100 mg/dl (G100) at 48 h]; high concentrations of D-mannitol were without effect. Neither low (2 ng/ml) nor high doses (2 micrograms/ml) of insulin modified luciferase activity in either G100 or G450. We next studied whether activation of protein kinase C (PKC) mediates the effect of high glucose. Treatment with the active phorbol ester phorbol 12-myristate 13-acetate for 2-4 h or with a diacylglycerol analogue for 24 h significantly stimulated luciferase activity preferentially in G100; the PKC inhibitors staurosporine or calphostin C significantly reduced the activity preferentially in G450. Thus high glucose levels promote transcriptional activity of COL4A1 gene in this reporter mesangial cell line, perhaps through PKC activation.

No abstract

Stromelysin-1, matrix metalloproteinase-3 (MMP-3), is an important endopeptidase selectively expressed by somatic cells in organ tissues. The renal tubulointerstitium, for example, comprises tubular epithelium and interstitial fibroblasts forming the principal mass of the kidney. We observed that mRNA encoding stromelysin-1 is detectable in murine renal fibroblasts, but not in proximal tubular epithelium. Transcripts measured by RNase protection assay in renal fibroblasts increase following exposure to phorbol ester, and thereafter, activated stromelysin-1 protein can be detected in culture media by Western blotting. A 6.4 Kb genomic clone containing the putative stromelysin-1 promoter was isolated and a relevant 2.1 Kb PstI restriction fragment including 2.1 Kb of the immediate 5'-flanking region was sequenced on both strands. Two transcriptional start sites were identified by primer extension; the major start site corresponded to a previously established position in the rat promoter, and a second undescribed minor transcriptional start site was located 16 bp upstream of the primary site. A HiNF-A chromatin-activating element at -106 bp was found in the early promoter region of pR336 and an active AP-1 site at -72 bp with an Ets/PEA-3 motif at -203 bp was suggested by transient transfection of luciferase minigenes into renal fibroblasts responsive to phorbol ester. This Ets element was identical to a site in the early promoter of the fibroblast-specific gene FSP1. A baseline enhancement in activity of pR336 in fibroblasts was further observed with the addition of 5' flanking sequence out to -1980 bp. This additional region of flanking sequence contains two modular regions: one of multiple PEA-3 elements between -684 bp and -1955 bp and a second region between -1929 bp and -1980 bps containing a second AP-1 site at -1929 bp, a MBF-1/ MEP-1 metal binding site, and a PPAR peroxisome proliferator element at -1950 bp. Our findings implicate a gene structure with expected activity in a mesenchymal phenotype. The PKC-dependent regulation of the stromelysin-1 gene supports the notion that it may be modulated during inflammation or tissue remodeling.

We have been studying the expression of several homeobox genes in cultures of proximal tubular epithelium (MCT cells) harvested from adult mus musculus. Hox genes 2.1, 2.3, and 3.3, in particular, are all expressed at low levels in resting MCT cells. The expression of Hox 2.1 and 3.3 were not influenced by mitogenic (epidermal growth factor: EGF, and platelet-derived growth factor: PDGF) nor by hypertrophogenic cytokines (angiotensin II: Ang II) in serum-free media. Transcripts for Hox 2.3, however, were elevated in MCT cells by Ang II. EGF, and serum treatment, as early as 30 minutes after their addition, whereas no change, or slight reductions were observed with transforming growth factor beta (TGF beta), PDGF, and gamma-interferon (gamma IFN). Hox 2.3 was also super-induced by serum, in the presence of cycloheximide, in cells rested previously in serum-free media, suggesting that new protein synthesis was not required for expressive augmentation. The induction of Hox 2.3, moreover, was not specific for tubular epithelium, since the gene could be activated in tubulointerstitial fibroblasts after treatment with EGF. These experiments collectively represent a first report regarding the characterization of transcripts encoding homeoboxes in adult cells derived from renal tissue. The putative DNA-binding properties of homeobox proteins in general, the prompt and rapid induction of Hox 2.3 by morphogenic cytokines in tubulointerstitial cells, and the observed effect of cycloheximide on this gene, all indicate that Hox 2.3 might have a role in the general activation of mature somatic cells, as an immediate early event. probably in the capacity of a nuclear trans-acting factor.