Matrix Metalloproteinases: Remodeling of the Extracellular Matrix

我们先前已经证实，膜型金属蛋白酶(MT-MMPs)对小鼠骨骼发育是必不可少的，而小鼠的胶原酶活性关键依赖于MT1-MMPs。重要的是，与小鼠MT1-基质金属蛋白酶缺失相关的特征是人类消失性骨病温彻斯特综合征(OMIM#259600)的显著表型，该综合征现在被确认为MT1-基质金属蛋白酶基因座的纯合子突变。 由于MT1-MMPs缺陷的多效性，以及MT1-MMPs在骨、骨相关组织和非骨组织中的广泛表达模式，我们培育并利用了一种条件缺失突变小鼠品系。随后，我们在结缔组织中以逐渐的细胞成熟和发育阶段特异性的方式删除了MT1-MMP的活性，以分配MT1-MMP介导的细胞和组织周围蛋白分解的细胞和组织特异性功能。此外，我们还讨论了MT1-基质金属蛋白酶活性在骨髓来源的单核/巨噬细胞/破骨细胞间的作用，在那里MT1-基质金属蛋白酶表达的意义尚不清楚。 为了确定MT1-MMP活性在周细胞样细胞(其中一些已被证明是骨骼干细胞)中的作用，我们去除了SM22pha阳性细胞中的MT1-MMP活性。MT1-MMPs在这一亚群细胞中的丢失使人联想到普遍的MT1-MMPs消融，包括侏儒症、猖獗的骨吸收、骨形成减少、进行性消瘦、纤维化和早期死亡。这些观察表明，形成骨架的细胞是从血管周围细胞池中招募出来的，并利用MT1-MMP来发挥其功能。 通往成骨命运的一个连续步骤是转录因子SP7/Osterix(OSX)的表达。我们利用表达OSX-CRE的小鼠来去除早期定向成骨细胞中的MT1-MMPs。这会导致明显的骨骼畸形，其次是猖獗的骨吸收，这让人想起无条件的MT1-MMP缺乏症，但没有SM22pha特异性消融术观察到的那么严重。这一观察结果与MT1-MMP裂解RANKL形成生物活性的能力是一致的。MT1-基质金属蛋白酶的活性与RANKL诱饵-骨保护素(OPG)一起抑制破骨细胞的募集。这与MT1-MMP在表达1型胶原(Col1a1-2.3kb)的成熟成骨细胞中的功能形成了鲜明对比。与OSX特异性缺失不同，Col1介导的MT1-MMP消融导致小鼠大体正常，但成年后表现出骨量减少。因此，虽然祖细胞特异性缺失会导致畸形和吸收，但成熟的成骨细胞特异性消融主要影响骨的定位，而不是吸收。 在骨骼动态平衡和骨转换的背景下，破骨细胞是一种基本的细胞类型，它表达大量的MT1-MMPs。然而，MT1-MMPs在破骨细胞活性中的作用还不是很清楚。为此，我们利用LysM-CRE小鼠特异性地去除破骨细胞中的MT1-MMP，与其他细胞特异性缺失不同，这导致骨小梁含量增加。因此，破骨细胞特异性的MT1-基质金属蛋白酶缺乏不是观察到的MT1-基质金属蛋白酶缺陷小鼠骨质侵蚀的原因。MT1-基质金属蛋白酶缺陷型破骨细胞形成数量相等，表现出与野生型破骨细胞难以区分的形态，并保留了降解矿化基质的不妥协能力。然而，它们在降解骨衬骨膜的主要成分--纤维状胶原的能力上确实存在近乎完全的缺陷。这种结缔组织是破骨细胞和矿化骨之间的屏障，必须在矿物质溶解之前被降解。这些结果强调了破骨细胞中中性蛋白酶活性的作用，并解释了在这一细胞亚群中MT1-MMPs丢失后骨含量增加的原因。 与我们的合作者一起，我们一直在建立各种小鼠模型，用于研究骨纤维异常增殖症(FD)。FD是一种致残性骨骼疾病，其特征是用低矿化、无组织的骨和纤维化的骨髓取代正常的骨和骨髓，缺乏造血，导致受影响的骨骼畸形和骨折。这种疾病是由编码刺激性G蛋白的阿尔法亚单位Gs的基因的合子后突变(R201C，R201H)引起的。人类缺乏这种疾病的遗传被认为反映了生殖系传播的激活Gs-α突变的胚胎致死性，这种突变只能通过体细胞嵌合体存活下来。我们培育了多个原始性表达Gs&R201C的小鼠品系，并发现它们具有人类FD的遗传精确复制品。在所有组织和小鼠胚胎干细胞中强劲的转基因表达与骨骼组织的正常发育和骨骼细胞的分化有关。与人类疾病一样，小鼠的FD病变只在出生后出现。这些损伤被发现经历了三个不同的阶段：1)由过度的骨形成和正常的吸收定义的建模阶段；2)过度的重塑阶段；3)纤维发育不良阶段，在1岁的小鼠身上复制了人类骨骼病理的完全复制。因此，Gs-α突变足以引起FD，并与生殖系传播和小鼠正常胚胎发育相容。这些新的小鼠系构成了第一个FD模型。

We have previously established that membrane-type metalloproteinases (MT-MMPs) are essential for skeletal development in the mouse, where collagenolytic activity is critically dependent on MT1-MMP. Importantly, the traits associated with loss of MT1-MMP in the mouse are a remarkable phenocopy of the human vanishing bone disease, Winchester syndrome (OMIM # 259600), which now is identified as a homozygous mutation of the MT1-MMP locus. Due to the pleiotropic nature of MT1-MMP deficiency and the widespread expression pattern of MT1-MMP in bone, bone-associated tissues and non-bone tissues, we have generated and utilized a conditional deletion mutation mouse strain. We subsequently deleted MT1-MMP activity in a progressive cell-maturity and developmental stage-specific fashion in connective tissues to assign the cell- and tissue-specific functions of pericellular proteolysis mediated by MT1-MMP. Additionally we have addressed the role of MT1-MMP activity in the bone marrow derived monocyte/macrophage/osteoclast compartment where the significance of MT1-MMP expression is poorly understood. To establish the role of MT1-MMP activity in pericyte-like cells (some of which have been proven to be skeletal stem cells), we have ablated MT1-MMP activity in SM22alpha positive cells. Loss of MT1-MMP in this subset of cells is reminiscent of universal MT1-MMP ablation including dwarfism, rampant bone resorption diminished bone formation, progressive wasting, fibrosis and early demise. These observations demonstrate that cells forming the skeleton are recruited out of the perivascular cell pool and utilize MT1-MMP to exert their function. A successive step towards osteogenic fate is expression of the transcription factor, Sp7/Osterix (Osx). We utilized Osx-Cre expressing mice to ablate MT1-MMP in early committed osteoprogenitors. This leads to overt skeletal dysmorphism and secondarily, a rampant bone resorption reminiscent of unconditional MT1-MMP deficiency, yet less severe than that observed with SM22alpha-specific ablation. This observation is consistent with the ability of MT1-MMP to cleave and shed RANKL into a biologically inactive form. MT1-MMP activity in this capacity works to suppress osteoclast recruitment in conjunction the RANKL decoy, Osteoprotegerin (OPG). This was in stark contrast to the function of MT1-MMP in Type 1 Collagen (Col1a1-2.3kb)-expressing, mature osteoblasts. Unlike Osx-specific deletion, Col1-mediated ablation of MT1-MMP led to grossly normal mice, which however display diminished bone mass in adulthood. Thus, while progenitor-specific deletion results in dysmorphism and resorption, mature osteoblast-specific ablation mainly affects bone apposition, but not resorption. In the context of skeletal homeostasis and bone turnover, one of the essential cell types is an osteoclast, which expresses abundant levels of MT1-MMP. However, the role that MT1-MMP plays in osteoclastic activity is not well understood. For that reason, we utilized LysM-Cre mice to specifically ablate MT1-MMP in osteoclasts, and unlike other cell specific-deletions, this led to increased trabecular bone content. Osteoclast-specific MT1-MMP deficiency consequently is not the cause of the bone erosion observed in MT1-MMP deficient mice. MT1-MMP-deficient osteoclasts form in equivalent numbers, display a morphology indistinguishable from wild-type osteoclasts and retain an uncompromised ability to degrade mineralized matrix. They do, however, display a near complete defect in ability to degrade fibrillar collagen the major component of bone-lining periostea. This connective tissue is a barrier between the osteoclast and the mineralized bone that must be degraded prior to mineral dissolution. These results highlight the function of neutral proteinase activity in osteoclasts and explain the increase in bone content following loss of MT1-MMP in this cell subset. With our collaborators, we have been generating various mouse models for the study of fibrous dysplasia of bone (FD). FD is a crippling skeletal disease characterized by replacement of normal bone and marrow with hypomineralized, unorganized bone and a fibrotic marrow, devoid of hematopoiesis, leading to deformity and fracture of the affected bones. The disease is caused by post-zygotic mutations (R201C, R201H) of the gene encoding the alpha subunit of the stimulatory G protein, Gs. Lack of inheritance of the disease in humans is thought to reflect embryonic lethality of germline-transmitted activating Gs-alpha mutations, which would only survive through somatic mosaicism. Multiple lines of mice that express GsαR201C constitutively were generated and were found to develop an inherited, exact replica of human FD. Robust transgene expression in all tissues and murine embryonic stem cells was associated with normal development of skeletal tissues and differentiation of skeletal cells. As in the human diseases, FD lesions in mice developed only in the postnatal life. The lesions were found to develop via three distinct stages: 1) a modeling phase defined by excess bone formation and normal resorption; 2) an excessive remodeling phase; and 3) a fibrous dysplastic phase, which reproduced a complete replica of the human bone pathology in mice of age >1 year. Thus, Gs-alpha mutations are sufficient to cause FD, and are compatible with germline transmission and normal embryonic development in mice. These novel murine lines constitute the first model of FD.