Smad4/DPC4-dependent Regulation of Biglycan Gene Expression by Transforming Growth Factor- ββββ in Pancreatic Tumor Cells

Overexpression of the small leucine-rich proteoglycan biglycan (BGN) in fibrosis and desmoplasia results from enhanced activity of transforming growth factor-beta (TGF-beta). In pancreatic adenocarcinoma, the tumor cells themselves may contribute to BGN synthesis in vivo, since 8 of 18 different pancreatic carcinoma cell lines constitutively expressed BGN mRNA, as shown by reverse transcription-PCR analysis. In PANC-1 cells, TGF-beta1 dramatically stimulated BGN mRNA accumulation through a BGN transcription-independent, cycloheximide-sensitive mechanism and strongly increased the synthesis and release of the proteoglycan form of BGN. The ability of TGF-beta1 to induce BGN mRNA was critically dependent on Smad signaling, since 1) the up-regulation of BGN mRNA was preceded by a marked increase in Smad2 phosphorylation in TGF-beta1-treated PANC-1 cells, 2) TGF-beta1 was unable to induce BGN mRNA in pancreatic carcinoma cell lines that carry homozygous deletions of the Smad4/DPC4 gene, 3) inhibition of the Smad pathway in PANC-1 cells by transfection with a dominant negative Smad4/DPC4 mutant significantly reduced TGF-beta1-induced BGN mRNA expression, 4) stable reintroduction of wild type Smad4/DPC4 into Smad4-null CFPAC-1 cells restored the TGF-beta1 effect, and 5) overexpression of Smad2 and Smad3 in PANC-1 cells augmented TGF-beta1 induction of BGN mRNA, whereas forced expression of Smad7, an inhibitory Smad, effectively blocked it. These results clearly show that a functional Smad pathway is crucial for TGF-beta regulation of BGN mRNA expression. Since BGN has been shown to inhibit growth of pancreatic cancer cells, the Smad4/DPC4 mediation of the TGF-beta effect may represent a novel tumor suppressor function for Smad4/DPC4: antiproliferation via expression of autoinhibitory BGN.


Introduction
The transforming growth factor-β (TGF-β) family comprises a large group of multifunctional cytokines with widespread distribution. They participate in a wide array of biological activities such as cell growth, differentiation, wound healing, apoptosis, and immunomodulation (1,2). TGF-β1, one of three mammalian TGF-β isoforms, (TGF-β1-3), is a potent inducer of extracellular matrix formation and has been implicated as the key mediator of fibrogenesis and desmoplasia in a variety of tissues (3). TGF-β1 promotes extracellular matrix accumulation primarily by inducing the synthesis of matrix proteins, such as collagens, fibronectin and proteoglycans.
Among the proteoglycans that are upregulated by TGF-β in vitro is biglycan (BGN), a prototype member of the small leucine-rich proteoglycan (SLRP) family (reviewed in Refs. [4][5][6]. BGN can be considered a marker gene for TGF-β activity which is reflected in vivo by the close spatial and temporal association of both proteins under physiological and various pathophysiological conditions (7)(8)(9)(10).
Due to its widespread, albeit cell type specific expression in the mammalian body, data on BGN function have been gathered from different organs and tissues and include regulation of matrix assembly, cellular adhesion (11), migration (12) and growth factor, e. g. TGF-β activity (13). The ability of BGN to bind TGF-β with high affinity (14) has been proposed to control the bioavailability of this growth factor.
Because of its pericellular localization BGN may function as a TGF-β binding protein that increases the probability of an interaction of TGF-β with its specific surface receptors. This scenario may have important implications for early progression of carcinomatous tumors, such as pancreatic carcinoma (10), since at this stage carcinoma cells are usually growth inhibited by TGF-β. Besides this indirect antiproliferative function, BGN may also directly inhibit growth of cancer cells in a TGF-β-independent manner, as Weber et al. (10) showed that exogenously administered recombinant BGN induced pancreatic cancer cells to arrest in the G1phase of the cell cycle.
The intracellular signaling mechanisms by which TGF-β controls the expression of BGN and other matrix-associated proteins remain poorly understood.
Signaling by TGF-β ligands requires two transmembrane serine-threonine kinase receptors (type II and type I). The ligand brings the two receptors together in a complex, and the constitutively active type II receptor kinase phosphorylates and activates the type I receptor kinase, which in turn activates downstream signaling pathways (15). Several signaling pathways are originating from the type I receptor, the most prominent one being the Smad pathway (16,17). This pathway is initiated when the type I receptor phosphorylates the receptor-regulated (R-)Smads, Smad2 or Smad3. Subsequently, the R-Smads heterodimerize with a co-Smad, Smad4. The Smad2/3-Smad4 heterodimer is then translocated to the nucleus where it binds directly or via other DNA binding proteins to the promoters of TGF-βresponsive genes to stimulate or repress their transcription (16,17). The antagonistic Smads, Smad6 and Smad7, on the other hand, inhibit the phosphorylation of R-Smads by the type I receptor and prevent the association of the R-Smads with Smad4/DPC4 (20)(21)(22).
Smad4 (also termed DPC4, for deleted in pancreatic carcinoma locus 4) has the characteristics of classical tumor suppressor gene, being mutated or deleted iñ 50% of pancreatic carcinomas and 15% of colorectal cancers (18,19). Since the discovery as a tumor suppressor much interest has focused on the role of Smad4/DPC4 as a mediator of TGF-β anti-proliferative signals which were originally thought to account for most, if not all, of its anti-oncogenic activity. However, several recent studies have confirmed that suppression of tumor formation and metastasis through this protein is more complex involving inhibition of tumor angiogenesis (23) and changes in the expression and activity of genes implicated in the control of cell adhesion and invasion (24). While all these effects may be directly or indirectly controlled by TGF-β , there are likely to be tumor suppressive activities that are unrelated to TGF-β signaling as inferred from the observation that inactivating DPC4 mutations occur together with mutations in the genes encoding TGF-β type II or type I receptor (25).
Our data indicate that TGF-β1 upregulation of BGN expression occurs through activation of Smad proteins and is critically dependent on a functional Smad4/DPC4. This is the first report demonstrating the involvement of Smad proteins in the TGF-β control of BGN and SLRP gene expression.

Experimental Procedures
Cell Lines and Cell Culture -The human pancreatic cancer cell lines PANC-1 and BxPC-3 and the osteoblastic osteosarcoma cell line MG-63 were purchased from the American Type Culture Collection (Rockville, MD). The pancreatic carcinoma cell lines CFPAC-1 and Hs766T were a kind gift of Dr. W. von Bernstorff (University of Kiel). The pancreatic carcinoma cell line COLO-357 and its supplier has been described previously (39). The COLO-357 cells used in this study harbor a wild type DPC4 gene (38) and are genetically distinct from COLO-357 cells obtained from another source, which have a homozygous deletion of the Smad4/DPC4 gene (40). All pancreatic cell lines were routinely maintained in RPMI 1640 supplemented with 10% and GAPDH mRNAs we carried out a competitive approach using an internal standard. The internal standard DNA for BGN of 386-bp was constructed by excision from the cDNA of an internal 114-bp BglII fragment, the standard DNA for PAI-1 was constructed by PCR using the PAI-1 antisense primer and a hybrid sense primer that contained nucleotides 357-378 followed by nucleotides 588-610 (5'- resulting in a 599-bp product upon amplification with PAI-1-sense and PAI-1antisense primers. Multiple reactions were run in parallel containing identical amounts of cDNA (corresponding to 100 or 200 ng of total RNA) but different concentrations of internal standard DNA. For this purpose, the standard DNA was serially diluted (0.9, 0.8, …, 0.09, 0.08, … and so forth). In order to keep reactions in the exponential phase, the number of cycles with an annealing temperature of 59°C was in the case of BGN adjusted to 16 cycles (PANC-1), 20 cycles (CFPAC-1), and 30 cycles (COLO-357) and for PAI-1 to 10 cycles, respectively. Following electrophoretic separation of PCR products on agarose gels and staining with ethidiumbromide photographs were taken and densitometrically scanned using the NIH Image software (version 1.62).
TGF-β induction of BGN mRNA was assessed from those reactions that showed an equimolar concentration of target and internal standard. The corresponding amount of target mRNA in these reactions was considered to be accurately determined when this ratio and the target to standard ratios of at least two neighboring reactions plotted against the corresponding standard dilutions on a semilogarithmic scale formed a linear relationship. To control for differences in cDNA synthesis the same cDNA was subjected to competitive PCR for GAPDH mRNA using primers GAPDH-sense: 5'-GGCGTCTTCACCACCATGGAG-3' and GAPDH-a n t i s e n s e :

Preparation of Proteinaceous Extracts, Partial Purification of Proteoglycans
and Chondroitinase ABC Digestion -Preparation of protein extracts was carried out as described previously (43). Briefly, pancreatic tumor cells were rinsed in PBS and   (46). From these data we conclude that BGN proteoglycan is strongly upregulated by TGF-β1 and properly secreted as a fully glycanated (and glycosylated) proteoglycan.

BGN mRNA is Upregulated by TGF-β1 in Cell Lines that Harbor a Wild Type
DPC4 Gene -Since results shown in Fig. 1Β indicated that BGN (core protein) synthesis in PANC-1 cells was strongly enhanced by TGF-β1, we tested by applying semiquantitative RT-PCR whether this was due to concomitant changes in BGN steady-state mRNA levels ( Fig. 2A). As expected, PANC-1 cells responded to TGF-

TGF-β Induction of BGN mRNA in PANC-1 Cells is Cycloheximide-sensitive and Does not Involve an Increase in BGN Promoter Activity -An earlier study has
shown that the TGF-β-induced rise in BGN mRNA was first detectable at 8h and peaked between 12 and 24 h after TGF-β addition to PANC-1 cells (10). To obtain some clues as to the underlying mechanism, we performed TGF-β stimulation experiments in the presence or absence of the protein synthesis inhibitor cycloheximide (CHX) (Fig. 3A). At both concentrations tested CHX effectively hence, did not express this C-terminal Smad4/DPC4 deletion mutant (see Fig. 5A).
Next, we evaluated the effect of the Smad4(1-514) mutant on endogenous BGN mRNA levels under the same conditions. In cells expressing Smad4(1-514) BGN mRNA induction was inhibited by 80% relative to vector controls (Fig. 5C).  Fig. 2) were retrovirally transduced with a full-length wild type Smad4/DPC4 cDNA. Successful restoration of Smad4 expression was verified by immunoblotting (Fig. 6A); the pool of transduced cells expressed nearly physiological level of this protein when compared to other cell lines with functional Smad4/DPC4, e. g. PANC-1 (Fig. 6A). Notably, when this pool of Smad4reconstituted CFPAC-1 cells was challenged with TGF-β1, BGN and PAI-1 mRNA levels markedly increased ( Fig. 6B and C). To test whether other known functions of Smad4/DPC4, e. g. mediation of the TGF-β antiproliferative effect, have been restored in these cells, we measured [ 3 H]-thymidine incorporation in TGF-β-treated CFPAC-1-DPC4 cells. As depicted in Fig. 6D, CFPAC-1-Smad4/DPC4 cells were not growth inhibited under conditions that efficiently arrested growth of PANC-1 cells.

Restoration of Smad4/DPC-4 Expression Renders
Together, these data clearly show that Smad4/DPC4 is involved in the induction of BGN and PAI-1 mRNA expression by TGF-β in pancreatic carcinoma cells.

The TGF−β effect on BGN mRNA Expression in PANC-1 Cells is Enhanced by Overexpression of Smad2 or Smad3 and Inhibited by Overexpression of Smad7 -
Since functional Smad4/DPC4 protein appeared to be necessary for TGF-β action on BGN mRNA, we reasoned that overexpression of R-Smads Smad2 and/or Smad3 should potentiate the TGF-β effect, whereas the inhibitory Smad Smad7 should interfere with it. The effects of Smad2, Smad3, and Smad7 were analyzed following their transient transfection into PANC-1 cells. Initially, functionality of the various Smad-encoding expression vectors was evaluated in a p6SBE reporter assay (Fig. 7A).
The p6SBE-lux plasmid contains 6 tandem repeats of the Smad-binding element (SBE) cloned in front of the luciferase reporter and has been used to specifically confer Smad4/DPC4-dependent transcriptional activation to a minimal promoter (40).
The TGF-β-induced reporter gene activity was strongly enhanced upon cotransfection with Smad2 or Smad3 but was repressed upon cotransfection with Smad7. The negative control plasmid p6MBE-lux containing 6 mutated SBEs was without any activity (data not shown). We then assessed the effect of these Smad proteins on BGN mRNA expression. Transient transfection of both Smad2 and Smad3 resulted in an 1.75-fold increase in TGF-β-induced BGN mRNA levels (Fig. 7B), whereas Smad1 under the same conditions had no effect (data not shown). In contrast, Smad7 potently inhibited the TGF-β effect on BGN mRNA (75% inhibition relative to vectortransfected controls, Fig. 7B). It should be mentioned that the transfection efficiency in these assays was ∼30% as determined by cotransfection with plasmid encoding We therefore consider it more likely that TGF-β1 exerts its effect in PANC-1 cells at a nuclear post-transcriptional level, e. g. pre-mRNA processing and/or nuclear export rather than on mature mRNA in the cytoplasm (41). Regarding the lack of transcriptional induction of the BGN gene Smad4/DPC4 may initially induce another gene(s) whose protein product(s) then translocates to the nucleus to mediate the BGN mRNA increase. This possibility is supported by our observation that TGF-β-induced accumulation of BGN mRNA in PANC-1 was effectively blocked by CHX. This results are in sharp contrast to equivalent data from another study where CHX had no effect (10). Further studies are required to elucidate the molecular mechanism of BGN mRNA accumulation and the immediate cellular targets of Smad4/DPC4.          A s P C -1 A 8 1 8 -4 B x P C -3 C a p a n -1 C a p a n -2 C F P A C -1 C o l o 3 5 7 H s 7 6 6 T H P A F M i a P a C a -2 M P a n c 9 6 P a C a -3 P a n c -1 P a n c -8 9 P a n c T u I P T 4 5 P 1 Q G P -1 T 3 M