Laboratory of Oral Connective Tissue Biology

答：我们假设参与控制PPi/Pi水平的因子，如ANK、NPP1、PHOSPHO1、TNAP，可能参与根的形成、胶凝发生。a.从Alpl KO小鼠获得的牙齿，除了牙骨质表型（无脱细胞牙骨质，PDL改变，牙齿脱落）外，表现出牙本质/牙釉质表型，模仿人类。通过TNAP酶替代，小鼠和人类（后者由Millan小组独立完成）修复牙釉质、牙本质和牙骨质缺陷。b.使用牙周缺损伤口模型，Ank KO小鼠的愈合速度更快。c.来自AlplxPhospho1的dKO小鼠的牙齿表现出比来自Alpl或Phospho1的KO的同类组织更深刻的牙本质表型，这表明了Phospho1在控制基质囊泡（MVs）中Pi水平的功能，以及TNAP和Phospho1在调节矿化方面的协同作用。最近的数据，与Millans小组合作，表明PHOSPHO1在牙骨质形成中起作用，这是一个意想不到的发现。因此，在体外，ANK和NPP1被证明是矿化驱动的基因/蛋白质，再加上体内的研究结果，证明了Pi/PPi调节对骨质形成的重要性，并为正在进行的研究提供了理论基础，这些研究使用体外方法来识别控制矿化的小分子，并在体内牙周伤口愈合模型中传递控制Pi/PPi的因素。有趣的是，通过检查Ank和Npp1 KO动物的牙周组织，发现了与基因表达相关的代偿机制，Ank KO组织中牙骨质中ENPP1蛋白水平升高，而Npp1 KO组织中Ank蛋白水平升高，但牙骨质表型并未改变。这很可能是由于TNAP在提供Pi的局部位点中发挥了重要作用。Npp1或Ank KO小鼠的骨水泥中OPN和DMP1的蛋白表达显著增加，进一步暴露于Pi的骨水泥母细胞在体外也显示出这两个基因的转录增加。这些数据为Pi/PPi比率如何调节骨质形成和特性提供了一个统一的假设，我们计划利用小鼠模型来刺激骨质再生。B.我们假设兄弟姐妹家族（如BSP、OPN和DMP1）及其相关整合素与磷酸盐代谢的关键调节因子之间存在关系，并且调节兄弟姐妹提供了再生骨质的治疗策略。a.骨唾液蛋白通过RGD整合素结合区影响细胞附着和信号传导，并通过成核羟基磷灰石晶体作为矿物沉淀的积极调节因子。为了验证我们的假设，我们分析了Bsp null（-/-）小鼠模型的牙齿发育。与Goldbergs小组合作的发育分析，通过组织学，组织化学和扫描电镜显示，Bsp -/-牙齿上的脱细胞牙骨质形成明显减少，牙骨质沉积低矿化。BSP的缺失导致PDL紊乱，上皮向下生长增加。Bsp -/-小鼠表现出广泛的根和牙槽骨吸收，这是由RANKL和破骨细胞增加介导的。与WT组织相比，KO小鼠组织中脱细胞形成区域的DMP1染色增加。结果表明，BSP在脱细胞牙骨质形成中起着重要而非多余的作用，可能参与了根表面矿化的启动。我们正在与NIDCR的Fisher一起探索整合素/BSP的关系。在与NIDCR的Young的一个相关项目中，我们假设蛋白聚糖的改变会损害牙周组织和关键兄弟姐妹的表达。两种slrp的dKO，纤维调节素和大聚糖，获得肌腱异位钙化和严重骨关节炎。偏振光研究显示dKO小鼠的牙槽骨存在异常的胶原原纤维和重构缺陷。IHC显示dKO小鼠PDL中slrp、asporin、lumican和decorin的染色增加，提示代偿机制。DMP1被认为是机械感觉骨细胞标志物，在dKO PDL中明显升高。RANKL表达的增加和TRAP阳性破骨细胞数量的增加进一步支持了dKO PDL稳态的破坏。为了阐明潜在的机制，我们用PCR阵列分析了dKO小鼠的PDL组织。结果显示，dKO组织中TGF β /BMP信号异常活跃，SMAD-4和-5基因表达显著增加，磷酸化SMAD5在dKO PDL中染色升高。因此，我们建议Fmod和Bgn阻止BMP与BMP受体的结合。从ECM中删除Fmod和Bgn的预测结果将是对TGF β /BMP的敏感性升高，从而影响BMP的信号转导，导致磷酸化SMAD5在PDL中的表达增加。C.我们假设PTH和1,25 d调节FGF23表达的能力涉及一个中间因子，特别是DMP1。1,25 d：使用小鼠成水泥细胞和骨细胞样细胞。在1,25D的存在下，Dmp1 mRNA水平下降了50%，而使用VDR激动剂和拮抗剂证实了维生素D受体途径的激活是必需的。进一步分析表明，组蛋白去乙酰化酶（HDAC）的募集是必要的。PTH：暴露于PTH（1-34）的成骨水泥细胞和骨水泥细胞表现出Dmp1 mRNA表达的剂量反应和时间依赖性。Western blot检测PTH治疗组DMP1下调30%。与对照组相比，PTH治疗使细胞内cAMP水平增加了10倍，而cAMP/PKA通路激动剂forskolin也降低了Dmp1 mRNA的表达，暗示cAMP/PKA通路。我们的研究结果表明，PTH可能通过几种途径调节FGF23，例如SOST和DMP-1，并且PTH和1,25 d介导的FGF23的作用可能具有一些共同的机制方面。未来使用RNAseq方法的研究将有助于确定参与Dmp1调控的特定DNA序列和转录因子。与DMP1启动子相关的研究一直与NIDCR的Collins和UW的Presland合作进行。

A. We hypothesized that factors involved in controlling PPi/Pi levels, e.g., ANK, NPP1, PHOSPHO1, TNAP, would be involved in root formation, cementogenesis. a. Teeth obtained from Alpl KO mice, beyond cementum phenotype (no acellular cementum, altered PDL, exfoliation of teeth) exhibited dentin/enamel phenotype, mimicking humans. Enamel, dentin and cementum defects were corrected by TNAP enzyme replacement, mice and humans (latter by Millan's group independently) b. Healing, using a periodontal defect wound model, occurred more rapidly in Ank KO mice, demonstrating significance of decreasing PPi levels at local sites toward promoting periodontal tissue repair c. Teeth from AlplxPhospho1 dKO mice exhibited a more profound dentin phenotype than comparable tissues from KO of Alpl or Phospho1 demonstrating the function of PHOSPHO1 in controlling Pi levels in matrix vesicles (MVs), and TNAP and PHOSPHO1 act synergistically in regulating mineralization. Recent data, in collaboration with Millans group, suggest that PHOSPHO1 plays a role in cementum formation, an unexpected finding. Thus, either PHOSPHO1 has a role in regulating Pi/PPi beyond MVs and/or cementum formation involves MVs d. In vitro, ANK and NPP1 were shown to be mineralized-driven genes/protein and coupled with in vivo findings, demonstrated the importance of Pi/PPi modulation for cementogenesis and provide the rationale for ongoing studies using in vitro methods to identify small molecules controlling mineralization and in vivo periodontal wound healing models to deliver factors controlling Pi/PPi, locally and systemically e. Intriguingly, examining periodontal tissues from Ank and Npp1 KO animals demonstrated a compensatory mechanism related to gene expression, ENPP1 protein levels increased in cementum in Ank KO tissues, while ANK protein levels increased in Npp1 KO tissues, yet the cementum phenotype was not altered, most likely due to the profound role of TNAP in providing Pi at local sites f. Cementum from of Npp1 or Ank KO mice exhibited a marked increase in protein expression for both OPN and DMP1, and further cementoblasts exposed to Pi in vitro also exhibited an increase in transcripts for these two genes. These data provide a unified hypothesis for how Pi/PPi ratio developmentally tunes cementum formation and properties, an insight that we plan to employ to stimulate cementum regeneration, using murine models. B. We hypothesized a relationship exists between the SIBLING family, e.g. BSP, OPN and DMP1, their associated integrins, and key regulators of phosphate metabolism and modulating SIBLINGS provides therapeutic strategy to regenerate cementum. a. Bone sialoprotein affects cell attachment and signaling through an RGD integrin-binding region, and acts as a positive regulator for mineral precipitation by nucleating hydroxyapatite crystals. To test our hypothesis we analyzed tooth development in a Bsp null (-/-) mouse model. Developmental analysis, in collaboration with Goldbergs group, by histology, histochemistry, and SEM revealed a marked reduction in acellular cementum formation on Bsp -/- teeth, and cementum deposited was hypomineralized.. Loss of BSP caused disorganized PDL and increased epithelial down-growth. Bsp -/- mice displayed extensive root and alveolar bone resorption, mediated by increased RANKL and osteoclasts. Increased staining for DMP1 in the region of acellular formation in tissues from KO mice vs. WT tissues was noted. Results suggest that BSP plays an essential and non-redundant role in acellular cementum formation, likely involved in initiating mineralization on the root surface. We are exploring the integrin/BSP relationship with Fisher at NIDCR b. In a related project with Young, NIDCR, we hypothesized that alterations in proteoglycans would compromise periodontal tissues and expression of key SIBLINGS. dKO for two SLRPs, fibromodulin and biglycan, acquire ectopic calcification of tendons and severe osteoarthritis. Polarized light studies demonstrated abnormal collagen fibrils and defective remodeling of alveolar bone of dKO mice. IHC revealed increased staining of SLRPs, asporin, lumican and decorin, in PDL of dKO mice, indicating compensatory mechanisms. DMP1, considered to be mechanosensory osteocyte marker, was increased markedly in dKO PDL. Disruption of homeostasis of dKO PDL was further supported by increased expression of RANKL and elevated numbers of TRAP positive osteoclasts. To elucidate underlying mechanisms, PDL tissues from dKO mice were analyzed by PCR array. Results indicated hyperactive TGF beta/BMP signaling in dKO tissues with significant increased expression of SMAD-4 and -5 genes, and elevated staining of phosphorylated SMAD5 in dKO PDL. Therefore we propose that Fmod and Bgn sequester BMP preventing binding to BMP receptors. The predicted outcome of deletion of Fmod and Bgn from ECM would be an elevated sensitivity to TGF beta/BMP which could affect BMPs signal transduction, causing an increased expression of phosphorylated SMAD5 in PDL. C. We hypothesized that the ability of PTH and 1,25D to modulate FGF23 expression involves an intermediate factor and specifically, DMP1. 1,25D: Murine cementoblasts and osteocyte-like cells were used. Dmp1 mRNA levels decreased by 50% in the presence of 1,25D, while use of a VDR agonist and antagonist confirmed that vitamin D receptor pathway activation was required. Further analysis showed that histone deacetylase (HDAC) recruitment was necessary. PTH: Cementoblasts and cementocytes exposed to PTH (1-34) exhibited a decrease in expression of Dmp1 mRNA in a dose-response and time-dependent manner. Western blot detected a 30% down-regulation of DMP1 with PTH treatment.Treatment with PTH increased intracellular cAMP levels by 10-fold compared to control, and forskolin, a cAMP/PKA pathway agonist, also decreased Dmp1 mRNA expression, implicating the cAMP/PKA pathway. Our results suggest that PTH may modulate FGF23 via several pathways e.g. SOST as well as DMP-1 and further that PTH and 1,25D-mediated effects on FGF23 may share some common mechanistic aspects. Future studies using RNAseq approaches will help determine the specific DNA sequences and transcription factor(s) involved in Dmp1 regulation. Research related to the DMP1 promoter has been in collaboration with Collins at NIDCR and Presland at UW.