Type I collagen is the most abundant protein in vertebrates. Its mutations typically result in Osteogenesis Imperfecta (OI), Ehlers-Danlos syndrome (EDS) or a combination of OI and EDS. Most OI mutations are substitutions of an obligatory glycine in the repeating Gly-X-Y triplets of the collagen triple helix. Disruption of the triple helix folding and structure by these mutations is clearly involved in the disease, but no relationship between different substitutions and OI severity has been found so far. We have established that the effect of Gly substitutions on the overall collagen stability depends on their location within different regions of the triple helix but not on the identity of the substituting residues. These regions appear to align with regions important for collagen folding, fibril assembly and ligand binding as well as some of the observed regional variations in OI phenotypes. In an ongoing study, we continue mapping of these regions and analysis of their association with OI phenotype variations. It has long been believed that bone pathology in OI results from abnormal collagen biosynthesis and function. However, recent discoveries by several research teams, including our group, are inconsistent with this idea. First, OI-like bone pathologies have also been found to be caused by deficiencies in other proteins, including: (a) Endoplasmic Reticulum (ER) chaperones involved in procollagen folding;(b) proteins important for maturation and function of osteoblasts but not directly involved in collagen biosynthesis (e.g., PERK, osteopotentia and osterix);and (c) proteins that affect osteoblast function from a distance, e.g., by altering serotonin synthesis in duodenum. Second, it has been demonstrated that normal bone homeostasis requires not only osteoblast synchronization with osteoclasts but also osteoblast coordination with other cells and organs. Based on our studies of OI mutations in collagen and other proteins, we argue that the primary cause of bone pathology in OI is osteoblast rather than collagen malfunction. Collagen mutations might be prevalent in OI simply because of their autosomal dominant inheritance and osteoblast malfunction associated with excessive ER stress response to procollagen misfolding. Collagen deficiency and/or malfunction is likely a modulating factor rather than the primary cause of the disease, potentially explaining why other connective tissues are usually less affected by collagen mutations than bones. Experimental testing of this hypothesis, which may open up new approaches to pharmacological OI treatment through ER stress targeting in osteoblasts, is currently under way. An interesting example of this paradigm shift is a newly discovered form of recessive OI associated with mutations resulting in pigment epithelium derived factor (PEDF) deficiency. PEDF is a collagen-binding protein and a close evolutionary relative of HSP47, which is a collagen-specific ER chaperone. Nevertheless, PEDF does not appear to be directly involved in type I collagen biosynthesis. During the last several months, we investigated collagen synthesis, folding and secretion by dermal fibroblasts from several OI patients with different PEDF deficiencies. Having found no abnormalities, we hypothesize that PEDF is involved in regulating osteoblast maturation and/or function, which is a subject of our ongoing study. To gain better understanding of osteoblast malfunction in skeletal disorders and develop novel approaches to treatment, in addition to cells and tissues from human patients, we utilize murine models. Our studies of mice with a Cys substitution for Gly-610 in the alpha-2 chain and mice with a Cys substitution for Gly-349 in the alpha-1 chain of type I collagen further support the idea of osteoblast malfunction as the primary cause of bone pathology in OI. We are now trying to translate this understanding into developing treatments that could reduce OI severity by normalizing osteoblast function. We are testing transplantation of bone marrow stromal cells, dietary changes as well as pharmacological approaches. Another important murine model of bone pathology associated with osteoblast malfunction is caudal vertebrae tumors in mice with deficiencies in different catalytic and regulatory subunits of protein kinase A, which is a crucial for cAMP signaling. In these tumors, we found accelerated bone matrix formation and deficient mineralization reminiscent of the McCune-Albright syndrome as well as very unusual collagen matrix organization and bone structures, which appear to be associated with improper maturation and/or function of osteoblasts. We hope that further studies of these animals will shed new light on the role of cAMP signaling in osteoblasts. One of our most significant advances in the past several years was the characterization of a collagenase-resistant isoform of type I collagen and its potential role in cancer, fibrosis, and other disorders. The normal isoform of type I collagen is a heterotrimer of two alpha-1 and one alpha-2 chains. However, homotrimers of three alpha-1 chains were found in some carcinomas, fibrotic tissues, and rare forms of OI and EDS associated with alpha-2 chain deficiency. Our studies of the homotrimeric collagen from an EDS patient revealed its resistance to cleavage by all major collagenolytic matrix metalloproteinases (MMPs), including MMP-1,2,8,13, and 14. A more detailed investigation showed this resistance to be related to an increased stability of the homotrimer triple helix at the primary MMP cleavage site, inhibiting unwinding of the helix at this site (necessary for the cleavage). MMP overexpression is a hallmark of invasive cancers;cleavage of stromal type I collagen fibers by cancer cells and recruited fibroblasts is an essential step in clearing an invasion path for cancer. Therefore, we hypothesized that synthesis of MMP-resistant collagen isoform may support cancer cell proliferation and tumor invasion;the MMP-resistant fibers laid down by these cells may serve as tracks, supporting outward cell migration and tumor growth. Our measurements confirmed the synthesis of a significant fraction of the homotrimers by cancer cells (20-40% in culture and even larger in vivo) but no homotrimer synthesis by normal mesenchymal cells or fibroblasts recruited into tumors. We found that more rigid homotrimeric type I collagen matrix supported faster proliferation and migration of cancer cells. Since the homotrimers do not appear to be produced in normal tissues, they may present a novel, appealing diagnostic and therapeutic target. Selective targeting of the homotrimers is a challenging project, since they do not contain unique peptide sequences. Yet, we hope to exploit differences in the stability of the homotrimeric and heterotrimeric triple helices that we discovered. In addition to carcinomas, type I homotrimers were reported in liver fibrosis and other fibrotic disorders. We established that their resistance to MMP is indeed involved in glomerular sclerosis in two different murine models. We found that it may also be a factor in the murine model of cAMP deficiency discussed above. However, homotrimeric type I collagen is not a general feature of fibrotic tissues. We did not find this collagen isoform in uterine fibroids or scleroderma skin lesions. We also did not find such molecules in amniotic membranes from normal or preterm deliveries, even though the homotrimer is believed to be a fetal isoform of type I collagen. Based on these and other studies, we hypothesize that the homotrimers are produced by undifferentiated, dedifferentiated, or transformed cells but not by normal or activated collagen-producing mesenchymal cells. We are now trying to understand the molecular mechanism regulating the homotrimer formation in different cells.
Showing the most recent 10 out of 28 publications
