Spelsberg, Thomas C.

We have previously demonstrated that β-catenin plays important roles in valve calcification with a specific osteogenic phenotype defined by increased bone mineral content and overall valve thickening. Recent studies indicate that KLF10 may be involved in mediating the Wnt signaling pathway in bone, which is known to play critical roles in osteoblast differentiation and mineralization. Therefore, we sought to test the role of KLF10 in mediating Wnt signaling, as well as differentiation and mineralization, in valve interstitial cells (VICs) isolated from porcine valves. Exposure of VICs to differentiation media led to increased expression of Runx2, Sox9 and osteocalcin. Differentiated cells also stained positive with Von Kossa while undifferentiated cells stained negative confirming the induction of an osteogenic phenotype. As expected, over-expression of both Lef1 and β-catenin led to activation of the top-flash reporter when transfected into VICs. Interestingly, over-expression of KLF10 also significantly up-regulated the top-flash reporter alone and further enhanced the activity of both Lef1 and β-catenin when co-transfected. We further confirmed the role of TIEG in an atherosclerotic mouse model ApoE/LRP5 double KO and found a two-fold increase in KLF10, Lrp6, and Runx2 expression in the cholesterol treated aortic valves as compared to controls. These data suggested that KLF10, Lef1 and β-catenin interact with each other to form a transcriptionally active protein complex leading to enhanced Wnt signaling in VICs. This possibility was further confirmed by the observation that KLF10 and β-catenin co-localize with one another in the nucleus of VICs following stimulation with LiCl and/or TGF-β. Taken together, these data implicate an important role for KLF10 in mediating Wnt signaling and Lef1 transcriptional activity in VICs, and implicate a potential role for canonical Wnt signaling in the observed osteogenic bone phenotype of cardiac aortic valves.

Background— Calcific aortic valve disease is the most common indication for surgical valve replacement in the United States. The cellular mechanisms of valve calcification are not well understood. We have previously shown that cellular proliferation and osteoblastogenesis are important in the development of valvular heart disease. Lrp5, a known low-density receptor-related protein, plays an essential role in cellular proliferation and osteoblastogenesis via the β-catenin signaling pathway. We hypothesize that Lrp5 also plays a role in aortic valve (AV) calcification in experimental hypercholesterolemia. Methods and Results— We examined the effects of cholesterol and atorvastatin in Watanabe rabbits (n=54). Group I (n=18) received a normal diet, group II (n=18) a 0.25% cholesterol diet, and group III (n=18) a 0.25% (w/w) cholesterol diet with atorvastatin for the development of calcification. The AVs were examined for cellular proliferation, Lrp5/β-catenin, and bone matrix markers. Bone formation was assessed by micro-computed tomography, calcein injection, and osteopontin expression. Low-density lipoprotein with and without atorvastatin was also tested in AV myofibroblasts for cellular proliferation and regulation of the Lrp5/β-catenin pathway. Our results demonstrate that the cholesterol diet induced complex bone formations in the calcified AVs with an increase in the Lrp5 receptors, osteopontin, and p42/44 expression. Atorvastatin reduced bone formation, cellular proliferation, and Lrp5/β-catenin protein levels in the AVs. In vitro analysis confirmed the Lrp5/β-catenin expression in myofibroblast cell proliferation. Conclusion— Hypercholesterolemic AV calcification is attenuated by atorvastatin and is mediated in part by the Lrp5/β-catenin pathway. This developmental pathway may be important in the signaling pathway of this disease.

Transforming growth factor (TGF)‐b signaling in naïve T cells induces expression of transcription factor Foxp3, a master regulator in the regulatory T cells (Tregs). However, the molecular mechanisms leading to Foxp3 induction remain unclear. Here we show that Itch−/− T cells are resistant to TGF‐b treatment with reduced Foxp3 expression. The E3 ubiquitin ligase Itch associates with and promotes ubiquitin conjugation to TGF‐b inducible early gene1 product (TIEG1). More importantly, Itch cooperates with TIEG1 to induce Foxp3 expression, which is reversed by TIEG1 deficiency. Functionally TGF‐b‐converted Tregs generated from TIEG1−/− mice are unable to inhibit naïve T cell proliferation in vitro and suppress airway inflammation in vivo. The results suggest an ubiquitin‐dependent non‐proteolytic pathway that is critical for Foxp3 expression and regulation of allergic responses.

Transcriptional control of Runx2 gene expression through two alternative promoters (P1 and P2) is critical for the execution of its function as an osteogenic cell fate determining factor. Using genomic DNaseI footprinting and in vivo DNA binding assays, we identified a helix‐loop‐helix (HLH) motif that overlaps with a Runx motif in the P1 promoter. The HLH mutation decreases P1 transcription 10‐fold in mesenchymal cells, while mutation of the Runx site enhances transcription 2 fold. Although the HLH element is not cell‐type specific, its function may depend on promoter context. Promoter deletion analysis shows that a GA‐rich region (−200 to −128) positively regulates Runx2 P1 transcription and synergizes with the HLH motif. Examination of the −200/−128 region reveals two tandemly repeated GA‐rich elements (GAY motifs) that each contain recognition motifs for purine‐rich DNA binding factors (i.e., SP1 and KLF10/TGFâ Inducible Factor [TIEG]). Forced expression of SP1 or TIEG modulates Runx2 P1 promoter activity through the GAY motifs. Taken together, our data suggest that the Runx2 gene transcription is regulated by synergistic interactions at multiple conserved composite promoter elements to control Runx2 expression and skeletal development.

Total cellular RNA was extracted from bone cells of three different femoral compartments of 2-mo-old rats. The intact femora were first incubated with collagenase to obtain periosteal cells. The bisected periosteum-free diaphyses and metaphyses were then incubated with collagenase to obtain enriched populations of endosteal and cancellous bone cells, respectively. The total cellular RNA from these three tissues was separated by size using agarose gel electrophoresis, transferred to nylon filters, hybridized to 32P-labeled cDNA probes for glyceraldehyde-3-phosphate dehydrogenase (GAP), pre-pro-alpha (I) type I collagen (collagen), osteocalcin (BGP), and alkaline phosphatase (AP), and the cDNA/mRNA hybrids were visualized by radioautography. Bone matrix deposition was measured in each tissue compartment by tetracycline-based dynamic bone histomorphometry. The bone formation and apposition rates were greatest in the periosteum and least in metaphysis. Mean mRNA levels for collagen and BGP were positively correlated with mean bone formation and mineral apposition rates. Interestingly, mean AP mRNA levels were not correlated with indexes of bone formation. These results demonstrate that the steady-state mRNA levels for bone matrix proteins in femora show pronounced site specificity and correlate with the rates of bone matrix deposition.

This report describes the relationship between bone formation and mRNA levels for selected bone proteins. Dynamic bone histomorphometry was used to measure bone formation in tibial periosteum of male rats from weanling (3 wk) to 52 wk old. Northern blot analysis of freshly isolated periosteal cells from the long bones was used to determine steady-state mRNA levels for the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAP), the bone matrix proteins osteocalcin (BGP), and prepro-alpha-2 (I) chain of type 1 precollagen (collagen), the osteoblast marker enzyme alkaline phosphatase (AP), and the osteoblast-derived signaling factor (growth factor) transforming growth factor-beta (TGF-beta). Radial growth at the tibial diaphysis achieved a maximum value in 8-wk-old rats and decreased progressively with age thereafter. This age-related decrease in the radial growth rate was initially due to reduced osteoblast activity; however, in older rats (greater than 17 wk old) reduced osteoblast number contributed to the decrease in bone formation. There was a strong correlation between the steady-state mRNA level for collagen and the periosteal bone formation rate. In contrast, the mRNA levels for the other bone proteins were more weakly correlated (TGF-beta and AP) or not correlated (BGP). These results suggest that the decreased bone matrix synthesis by periosteal cells in long bones of maturing rats is due to decreased expression of genes for bone matrix proteins.

Abstract The Transforming Growth Factor-β (TGFβ)/SMAD signaling pathway can function as either a tumor suppressor or metastasis promoter during tumor progression. In normal epithelial cells and early stages of epithelial tumorigenesis TGFβ functions as a tumor suppressor to decrease cell proliferation or induce apoptosis. However, during malignant progression tumor cells no longer respond to the anti-proliferative effects of TGFβ, but instead undergo an epithelial-to-mesenchymal transition (EMT) whereby cells acquire a migratory and invasive phenotype which promotes tumor metastasis. Resolution of the dichotomy in TGFβ function and a further understanding of its tumor suppressor and metastasis promoting functions may uncover new strategies for the treatment of epithelial cancers. In previous studies we demonstrated an important role of the TGFβ-Inducible Early Gene-1 (TIEG1)/Krüppel-like Factor-10 (KLF10) as a central regulator of TGFβ/SMAD signaling and the anti-proliferative functions of TGFβ. In this study we examined the role of KLF10 in controlling the TGFβ-induced EMT in human lung adenocarcinoma cells. We show that depletion of KLF10 results in a more pronounced induction of EMT in human A549 lung adenocarcinoma cells as assessed by quantitative real-time PCR, changes in cellular morphology and immunofluorescence microscopy, and transcriptome-wide (RNA-seq) analyses. Moreover, chromatin immunoprecipitation (ChIP) and chromatin immunoprecipitation-sequencing (ChIP-seq) analysis shows that KLF10 directly binds to GC-rich sequences in the promoter region of the EMT-promoting transcription factor SLUG/SNAI2 to repress its transcription. Consistent with these findings, an analysis of KLF10 and SNAI2 expression in lung cancer revealed that while KLF10 levels are decreased in lung cancer vs. normal samples, SNAI2 levels are increased. Additional ChIP studies showed that KLF10 recruits HDAC1 to the SNAI2 promoter and is required for the removal of activating histone acetylation marks. These findings reveal a previously unknown function of KLF10 in suppressing TGFβ-induced EMT and represent a significant advancement in the understanding the molecular dichotomy of TGFβ function during tumor progression. Citation Format: Vivek Kumar Mishra, Vijayalakshmi Kari, Malayannan Subramaniam, Simon J. Baumgart, Sankari Nagarajan, Florian Wegwitz, Thomas C. Spelsberg, John R. Hawse, Steven A. Johnsen. Krüppel-like Transcription Factor-10 (KLF10) suppresses the TGFβ-induced epithelial-to-mesenchymal transition. [abstract]. In: Proceedings of the AACR Special Conference on Chromatin and Epigenetics in Cancer; Sep 24-27, 2015; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2016;76(2 Suppl):Abstract nr A03.