Cartilage oligomeric matrix protein expression in systemic sclerosis reveals heterogeneity of dermal fibroblast responses to transforming growth factor beta.

OBJECTIVE: Cartilage oligomeric matrix protein (COMP) accumulates in systemic sclerosis (SSc) skin and is upregulated by transforming growth factor (TGF)beta. To further characterise the response to TGFbeta in SSc, we investigated TGFbeta1 and COMP expression and myofibroblast staining in SSc skin. METHODS: Skin biopsies from patients with diffuse cutaneous SSc (dSSc), limited cutaneous SSc (lSSc) and healthy controls were evaluated for COMP mRNA expression using real-time PCR. COMP, alpha-smooth muscle actin (SMA) and TGFbeta were assessed in skin sections and in cultured fibroblasts by immunohistochemistry. Clinical disease status was assessed by the modified Rodnan skin score (mRSS). RESULTS: Myofibroblasts expressing SMA and COMP were found coexpressed in many cells in dSSc dermis, but each also stained distinct cells in the dermis. Cultured SSc dermal fibroblasts also showed heterogeneity for COMP and SMA expression, with cells expressing SMA, COMP, both or neither. TGFbeta treatment increased COMP and SMA-expressing cells. COMP mRNA expression in lesional skin from patients with dSSc correlated with the mRSS and TGFbeta1 staining. CONCLUSION: These findings suggest that TGFbeta upregulation of COMP and/or SMA expression in subpopulations of fibroblasts contributes to different pathways of fibrosis and that multiple TGFbeta regulated genes may serve as biomarkers for the degree of SSc skin involvement.

Global expression analysis of the fibroblast transcriptional response to TGFbeta.

OBJECTIVES: Transforming Growth Factor-beta (TGFbeta) is the predominant cytokine in all forms of fibrotic reactions. As well as being secreted by immune modulators of fibrosis such as macrophages, it is involved in an autocrine feedback loop of fibroblast stimulation whose regulation is still poorly understood. We wished to gain some insight into the mechanisms of the fibroblast response to TGFbeta. METHODS: We undertook an exhaustive transcript profiling experiment using a widely validated restriction enzyme based method for identifying differentially expressed genes (GeneCalling). Transcriptional responses throughout a 24-hour time course were examined at multiple time points and classified. RESULTS: By 24 hours of TGF treatment over 1000 bands, representing a large number of transcripts, were down- or upregulated greater than 2-fold. All of the known genes responsive to TGFbeta, such as collagen and connective tissue growth factor, were upregulated. CONCLUSIONS: This encyclopedic method revealed many unknown transcriptional responses to TGFbeta including the upregulation of a variety of less expected cytoskeletal and matrix components, as well as interactions between the TGFbeta and tumor necrosis factor (TNF) pathways and alterations in cell death-related pathways. These may in part explain the idiosyncratic responses of mesenchymal cells to TGFbeta.

The hcKrox gene family regulates multiple extracellular matrix genes.

The transcription factor cKrox was originally identified as a protein that bound to a negative transcription regulatory element in the murine alpha1(I) collagen promoter. We recently reported the cloning and characterization of human cKrox (hcKrox). Overexpression of hcKrox in NIH3T3 fibroblasts efficiently repressed the promoters of the fibronectin and alpha1(I) collagen genes (70-90%) in transient transfection assays and suppressed the endogenous genes in hcKrox expressing permanent cell lines. We have now isolated genomic clones and cDNAs encoding two novel transcription factors related to hcKrox termed hcKrox-beta and hcKrox-gamma (the original clone is now referred to as hcKrox-alpha). Both contain three kruppel-like zinc-finger DNA binding motifs that are 71-78% identical to those of hcKrox-alpha. The NH(2)-terminus of all three proteins contains a POZ domain, a conserved 120 amino acid motif involved in transcriptional repression and protein dimerization. RT-PCR experiments demonstrate that all three hcKrox family members are expressed in foreskin and dermal fibroblasts. Transient transfection studies in NIH3T3 fibroblasts demonstrate that hcKrox-alpha -beta and -gamma, as well as the murine cKrox-beta homologue, LRF, suppress transcription driven by promoters for the alpha1(I) and alpha2(I) collagen, fibronectin and elastin genes. Electrophoretic mobility shift assays and coimmunoprecipitation studies suggest that homo- and heterodimerization occurs between cKrox family members. Dimer formation is influenced by amino acids in the NH(2)-terminal POZ domain and the Zn(+2)-finger region. Immunoprecipitation studies indicate that cKrox can form heterodimers in solution in the absence of DNA. Thus, a multi-gene family exists that can coordinately regulate several extracellular matrix genes and has the potential to form many heterodimeric transcription factors.

Regulation of matrix biosynthesis and degradation in systemic sclerosis.

The regulation of matrix biosynthesis in systemic sclerosis has been the focus of many studies, because excessive matrix synthesis causes pathologic changes, and because this would seem to be a good target for therapies aimed at ameliorating the disease. Possible targets for antifibrotic therapies include both matrix gene stimulatory and inhibitory pathways. Much recent progress has been made in understanding the mechanism of action of transforming growth factor-beta (TGF-beta), an important profibrotic cytokine with pleiotropic effects on fibroblasts. It appears that TGF-beta may use multiple signal transduction pathways in fibroblasts and it is possible that defects in any of these pathways may result in an abnormal response to TGF-beta, resulting in fibrosis. Studies on negative regulation of matrix gene expression have singled out the antifibrotic cytokines tumor necrosis factor-alpha and interferon-gamma. Finally, a new approach that compares mRNA expression in normal versus diseased fibroblasts has already lead to the discovery of genes that may play a role in the development of fibrosis. This represents an important advance because genes can be identified that have not previously been implicated in the control of matrix synthesis, and thus might not otherwise have been studied in this context.

FasL promoter activation by IL-2 through SP1 and NFAT but not Egr-2 and Egr-3.

Recently activated peripheral T cells treated with IL-2 for 4 days expressed Fas ligand (FasL)-mediated cytotoxicity. These IL-2-treated T cells had high nuclear expression of SP1 and NFAT, but lacked the Egr-2 and Egr-3 that could be induced by anti-CD3 stimulation and had been implicated in FasL gene activation. A minimal promoter region that responded to IL-2 was identified by transient transfection assays using deletion mutants. The data suggests that the GGGCGGAAA site present in the 5' end of the minimal FasL promoter is critical to IL-2-induced FasL gene activation. The GGGCGGAAA sequence contains an overlapping site used by two transcription factor families, one (GGGCGG) for the SP1 family and the other (GGAAA) for the NFAT family. FasL promoter activity was partially but statistically significantly reduced with constructs mutated at either site. More activity was lost with a construct mutated at both sites. In contrast, mutation at the Egr site had no effect on IL-2-induced FasL promoter activity. Our study identified a new FasL promoter site responding to IL-2-induced SP1 and NFAT factors. Furthermore, the nuclei of IL-2-treated cells express SP1 and NFAT, but not Egr-2 and Egr-3, for FasL gene activation.

Cloning and characterization of hcKrox, a transcriptional regulator of extracellular matrix gene expression.

cKrox is a novel zinc finger-containing transcription factor that binds to the alpha1(I) and alpha2(I) collagen gene promoters. The gene coding for cKrox is a new member of a family of early growth response genes, that play important roles in development. In the mouse, cKrox is expressed, beginning at 9.5 days of gestation and at 10.5 days in regions destined to become skin. In adult animals, expression is predominantly in skin, one of the two major organs where type I collagen is expressed. We have isolated cDNA clones for human cKrox. Theoretical translation of the nucleic acid sequence reveals 90% conservation of amino acids between the mouse and human proteins; however, the human gene product contains a 117-amino-acid N-terminal extension. The amino acid sequences of the zinc-finger DNA binding domains of mouse and human cKrox are identical. RT-PCR analysis of human fibroblasts indicates constitutive low-level expression of cKrox which can be transiently elevated by treatment with retinoic acid. Transient transfection assays indicate that hcKrox represses transcription of the alpha1(I) procollagen promoter, and electrophoretic mobility shift assays demonstrate that hcKrox binds to both the human and murine promoter DNA. Deletion derivatives of hcKrox demonstrate transcription-activating potential that is promoter-dependent. NIH3T3 cells permanently expressing hcKrox demonstrate a threefold and 10-fold decrease in alpha1(I) procollagen and fibronectin mRNA levels, respectively, compared to control cells. Consistent with this finding, a fibronectin promoter reporter construct is repressed more than 80% by hcKrox. These data suggest that hcKrox represses collagen transcription directly, and it may function as a repressor of fibronectin and possibly other matrix genes.

Functional characterization of the human biglycan 5'-flanking DNA and binding of the transcription factor c-Krox.

The transcriptional regulation of human biglycan expression under normal and pathological conditions was studied. The 5'-flanking regions of the human and mouse genes were isolated and analyzed; the two promoter regions share 81% identity. Both promoters are without a TATA and CAT box and contain multiple Sp1 sites. Human dermal fibroblasts were transiently transfected with progressive deletional human biglycan 5'-flanking DNA-CAT constructs, and a significant variation in activity among the individual constructs was found. A small deletion in several cases caused a more than 2-fold increase or decrease in promoter activity, thereby mapping the target sites for repressors or activators. Human biglycan expression is reduced in females with Ullrich-Turner syndrome (45,X) and increased in individuals with supernumerary sex chromosomes, and it has been speculated that biglycan plays a role in the short stature phenotype of Turner syndrome. Analysis of the transcriptional regulation of biglycan in individuals with sex chromosome anomalies showed that a -262 to -218 region of the biglycan promoter was differentially regulated. This region was extensively analyzed by DNAse footprinting and electrophoretic mobility shift assays, and a putative binding site for the transcription factor c-Krox was discovered. The binding of c-Krox to a site located at approximately -248 to -230 in the human biglycan promoter was confirmed by using extracts from COS cells expressing recombinant human c-Krox. The expression of c-Krox in bone was then examined by reverse-transcribed polymerase chain reaction and Northern blotting analysis; an approximately 3.4 kb transcript was detected in primary osteoblastic cells, in MG-63 cells, and in human bone marrow stromal cells. This is the first detection of c-Krox in bone cells, and it suggests that c-Krox, like another member of the Krox family, Krox-20, might play a regulatory role in bone.

Involvement of early growth response factor Egr-1 in apolipoprotein AI gene transcription.

Liver-specific expression of the apolipoprotein AI (apoAI) gene is mediated by transcription factors bound to three sites (A, B, and C) in the apoAI enhancer. Sites A and C bind various members of the nuclear receptor superfamily, including the orphan nuclear receptor apolipoprotein regulatory protein-1 (ARP-1); site B binds the liver-enriched factor hepatic nuclear factor-3. The immediate early growth response factor (Egr-1), which is transiently expressed in various pathophysiologic states of the liver, activates the apoAI enhancer and overcomes ARP-1-mediated repression of the enhancer in hepatoblastoma HepG2 cells. Deletion mapping analysis revealed two Egr-1 binding sites, E1 and E2, flanking site A. Erg-1 bound efficiently to both E1 and E2. Sp1 in HepG2 nuclear extracts bound to E2 but not E1. In HepG2 cells, E1 functioned as an Egr-1 response element, whereas E2 had high basal activity and was not further induced by Egr-1. Mutations that prevent Egr-1 binding to the apoAI enhancer abolished its responsiveness to Erg-1, while they had only minor effects on its constitutive activity. These mutations also diminished the ability of Egr-1 to overcome ARP-1-mediated repression. Elimination of transcription factor binding to sites A, B, or C reduced enhancer activity without affecting Egr-1-dependent activation. We argue that Egr-1 is recruited to the apoAI enhancer complex under unusual circumstances, such as those prevailing during liver regeneration, to maintain apoAI transcription levels by overriding prior transcriptional controls.

Repression by ARP-1 sensitizes apolipoprotein AI gene responsiveness to RXR alpha and retinoic acid.

The gene coding for apolipoprotein AI (apoAI), a lipid binding protein involved in the transport of cholesterol and other lipids in the plasma, is expressed in mammals predominantly in the liver and the intestine. Liver-specific expression is controlled by synergistic interactions between transcription factors bound to three separate sites, sites A (-214 to -192), B (-169 to -146), and C (-134 to -119), within a powerful liver-specific enhancer located between nucleotides -222 and -110 upstream of the apoAI gene transcription start site (+1). Previous studies in our laboratory have shown that ARP-1, a member of the nuclear receptor superfamily whose ligand is unknown (orphan receptor), binds to site A and represses transcription of the apoAI gene in liver cells. In a more recent series of experiments, we found that site A is a retinoic acid (RA) response element that responds preferentially to the recently identified RA-responsive receptor RXR alpha over the previously characterized RA receptors RAR alpha and RAR beta. In this study we investigated the combined effects of ARP-1 and RXR alpha on apoAI gene expression in liver cells. Transient transfection assays showed that site A is necessary and sufficient for RXR alpha-mediated transactivation of the apoAI gene basal promoter in human hepatoma HepG2 cells in the presence of RA and that this transactivation is abolished by increasing amounts of cotransfected ARP-1. Electrophoretic mobility shift assays and subsequent Scatchard analysis of the data revealed that ARP-1 and RXR alpha bind to site A with similar affinities. These assays also revealed that ARP-1 and RXR alpha bind to site A as heterodimers with an affinity approximately 10 times greater than that of either ARP-1 or RXR alpha alone. Further transfection assays in HepG2 cells, using as a reporter a construct containing the apoAI gene basal promoter and its upstream regulatory elements (including site A) in their natural context, revealed that RXR alpha has very little effect on the levels of expression regardless of the presence or absence of RA. However, while ARP-1 alone or ARP-1 and RXR alpha together dramatically repress expression in the absence of RA, the repression by ARP-1 and RXR alpha together, but not ARP-1 alone, is almost completely alleviated in the presence of RA. These results indicate that transcriptional repression by ARP-1 sensitizes apoAI gene responsiveness to RXR alpha and RA and suggest that the magnitude of this responsiveness is regulated by the intracellular ratio of ARP-1 to RXR alpha. These observations raise the possibility that transcriptional repression is a general mechanism for switching gene transcription between alternative transcription activation pathways.

A retinoic acid-responsive element in the apolipoprotein AI gene distinguishes between two different retinoic acid response pathways.

The gene coding for apolipoprotein AI, a plasma protein involved in the transport of cholesterol and other lipids in the plasma, is expressed predominantly in liver and intestine. Previous work in our laboratory has shown that hepatocyte-specific expression is determined by synergistic interactions between transcription factors bound to three separate sites, sites A (-214 to -192), B (-169 to -146), and C (-134 to -119), within a powerful liver-specific enhancer located in the region -222 to -110 nucleotides upstream of the apolipoprotein AI gene transcription start site (+1). In this study, it was found that site A is a highly selective retinoic acid-responsive element (RARE) that responds preferentially to the recently identified retinoic acid receptor RXR alpha over the previously characterized retinoic acid receptors RAR alpha and RAR beta. Control experiments indicated that a RARE in the regulatory region of the laminin B1 gene responds preferentially to RAR alpha and RAR beta over RXR alpha, while a previously described palindromic thyroid hormone-responsive element responds similarly to all three of these receptors. Gel retardation experiments showed that the activity of these RAREs is concordant with receptor binding. These results indicate that different RAREs may play a fundamental role in defining distinctive retinoic acid cellular response pathways and suggest that retinoic acid response pathways mediated by RXR alpha play an important role in cholesterol and retinoid transport and metabolism.

Synergistic interactions between transcription factors control expression of the apolipoprotein AI gene in liver cells.

The gene coding for apolipoprotein AI (apoAI), a plasma protein involved in the transport of cholesterol and other lipids in the plasma, is expressed predominantly in liver and intestine. Previous work in our laboratory has shown that different cis-acting elements in the 5'-flanking region of the human apoAI gene control its expression in human hepatoma (HepG2) and colon carcinoma (Caco-2) cells. Hepatocyte-specific expression is mediated by elements within the -256 to -41 DNA region relative to the apoAI gene transcription start site (+1). In this study it was found that the -222 to -110 apoAI gene region is necessary and sufficient for expression in HepG2 cells. It was also found that this DNA region functions as a powerful hepatocyte-specific transcriptional enhancer. Gel retardation and DNase I protection experiments showed that HepG2 cells contain proteins that bind to specific sites, sites A (-214 to -192), B (-169 to -146), and C (-134 to -119), within this enhancer. Site-directed mutagenesis that prevents binding of these proteins to individual or different combinations of these sites followed by functional analysis of these mutants in HepG2 cells revealed that protein binding to any one of these sites in the absence of binding to the others was not sufficient for expression. Binding to any two of these sites in any combination was sufficient for only low levels of expression. Binding to all three sites was essential for maximal expression. These results indicate that the transcriptional activity of the apoAI gene in liver cells is dependent on synergistic interactions between transcription factors bound to its enhancer.

Chromosomal organization of rRNA operons in Bacillus subtilis.

Integrative mapping with vectors containing ribosomal DNA sequences were used to complete the mapping of the 10 rRNA gene sets in the endospore forming bacterium Bacillus subtilis. Southern hybridizations allowed the assignment of nine operons to distinct BclI restriction fragments and their genetic locus identified by transductional crosses. Nine of the ten rRNA gene sets are located between 0 and 70 degrees on the genomic map. In the region surrounding cysA14, two sets of closely spaced tandem clusters are present. The first (rrnJ and rrnW) is located between purA16 and cysA14 closely linked to the latter; the second (rrnI, rrnH and rrnG) previously mapped within this area is located between attSPO2 and glpT6. The operons at or near the origin of replication (rrnO,rrnA and rrnJ,rrnW) represent "hot spots" of plasmid insertion.

Instability of rRNA operons in Bacillus subtilis.

Many laboratory strains of Bacillus subtilis contain 9 rather than 10 rRNA operons due to deletions occurring within the rrnJ-rrnW or rrnI-rrnH-rrnG gene cluster. These operons are members of two sets of closely spaced clusters located in the cysA-aroI region. Analysis of rescued DNA from integrants with insertions into rrnG and rrnH indicated that these tandemly arranged operons allowed frequent deletions of an rrn operon equivalent. These events may arise spontaneously by intrachromosomal recombination or by simultaneous double crossovers with a multimeric integrative plasmid.

Mapping of rRNA genes with integrable plasmids in Bacillus subtilis.

Integrable plasmids pGR102 and pWR103 containing ribosomal sequences from within the transcriptional units for 16S and 23S were used to transform Bacillus subtilis. To date, these plasmids integrated into 7 of 10 known rrn operons. Two such events occurred at unassigned operons, revealing a close linkage of the CAT gene of the plasmid to pha-1 situated between dal-1 and purB33 for rrnE and to thiA78 situated between glyB133 and re-12 for rrnD. All seven integration events that led to the loss of unique ribosomal BclI fragments can now be assigned to known rrn operons.