Collagen-related diseases

Eunice Kennedy Shriver National Institute Of Child Health & Human Development

Investigators

Linked publications & trials

Abstract

Type I collagen (Col-I) is the most abundant human protein forming the structural scaffold of bone, skin and other tissues. Among pathologies associated with disruptions in Col-I biosynthesis are bone fragility in OI, CCS, and other skeletal dysplasias as well as skin and joint laxity in EDS. Over 80% of severe OI cases are caused by substitutions of glycine (Gly) required in every third position of collagen triple helix, the main functional unit of Col-I. By altering the helix folding, Gly substitutions cause malfunction of bone producing cells (osteoblasts) and alter formation and function of the collagen matrix in tissues. Our studies of mouse models and cells from OI patients revealed both effects to be pathogenic. We demonstrated that a key factor in bone failure is cell stress and resulting osteoblast malfunction due to Endoplasmic Reticulum (ER) disruption by collagen precursor (procollagen) misfolding. We have described noncanonical features of this stress, but its mechanism is still not fully understood. Recently, we identified ER-mitochondria contacts as the likely sensor of the ER disruption in an OI mouse model, suggesting novel stress mechanism and therapeutic targets. During the last year, our studies revealed similar ER disruption in fibroblasts responsible for synthesis and maintenance of lung central tendon, while fibroblasts of most other tissues appeared to be minimally affected. We have found that the resulting central tendon pathology is a crucial factor in perinatal lethality of severely affected animals. To understand and target these pathologies, we created and characterized several mouse models. In 2010, we reported a new G610C mouse model of OI, which mimics a Gly610 to Cys substitution in the alpha-2 chain of Col-I in a large group of OI patients. This model is now used for OI studies by many research groups in the US and around the world. It has been instrumental in our own studies demonstrating that Gly substitutions may trigger integrated stress response of the cell without activating canonical ER stress pathways. It revealed activation of a mitochondrial branch of integrated cell stress response (ISR) via disruption of ER-mitochondria contacts. This ISR branch is regulated by mitochondrial HSP70 and ATF5, which are paralogues of BIP (ER HSP70) and ATF4 regulators of the unfolded protein response branch of ISR. Our most recently developed mouse model, which we are still characterizing, is based on osteoblast-specific knockout of SEC24D. It mimics OI-like bone pathology in CCS type 2. Our study of this model appears to confirm activation of the mitochondrial ISR branch upon disruption of procollagen export from the ER through ER exit sites (ERESs). To further validate and better understand this novel type of cell stress response to ER disruption upon excessive secretory protein accumulation in the ER, we are developing additional mouse models by combining the G610C mutation, osteoblast specific SEC24D knockout, and global ATF5 knockout. Through understanding the role of ATF5 we hope to identify new therapeutic targets for alleviating ISR-related cellular malfunction. While OI is characterized by bone fragility and skeletal deformities, the main cause of perinatal mortality in OI is lung failure. Since we observed perinatal lethality of all homozygous and few heterozygous G610C animals due to lung failure, we initiated studies of molecular mechanisms underlying this pathology. We found lung pathology to be caused by multiple factors, including rib cage deformities, weakness of the lung parenchyma collagen matrix due to incorporation of mutant molecules, and disrupted collagen deposition and function in the central tendon of lungs. By combining structural characterization of the matrix with single-cell and spatially resolved RNA sequencing, we identified a previously unrecognized mechanism of lung pathology, which likely affects not only OI but many other skeletal dysplasia patients. Specifically, rib cage fractures and deformities combined with abnormal central tendon formation and function result in flattening and malfunction of the diaphragm, preventing proper lung inflation with amniotic fluid during fetal breathing. The deficient lung inflation disrupts development of saccular and alveolar structures as well as pneumocyte differentiation, resulting in lung hypoplasia of variable severity. In homozygous and most severely affected heterozygous G610C animals, the underdeveloped saccular structures and poor pneumocyte differentiation prevent breathing after birth and result in perinatal lethality. Less severely affected animals that compensate lung underdevelopment by hyperactive breathing, inhaling more air per minute than the weak collagen matrix of their lung parenchyma can support, which causes parenchymal injury, inflammation, and fibrosis. In surviving animals, the parenchymal injury accumulates during rapid animal growth. As the growth slows after puberty and stronger collagen matrix can be deposited, mice slowly repair the parenchymal injury and restore near normal lung function. Unlike mice, adult humans do not possess the ability to efficiently repair damaged lung parenchyma, likely causing continued accumulation of parenchymal damage and fibrosis, which may explain the progression of lung pathology in OI and other skeletal dysplasia patients. We therefore initiated collaboration with clinicians monitoring pregnancies, neonatal pathology, development, and subsequent progression of pathology in OI within OI lung consortium. This consortium includes scientists and clinicians from the New York Hospital for Special Surgery, National Jewish Health in Denver, UCLA, University of Arkansas, University of Maryland in Baltimore, and NIH. One of the key consortium goals is to develop approaches to early diagnostics and treatment of lung pathology in OI and other skeletal dysplasias. The specific focus of our research at NIH is in utero diagnostics of lung hypoplasia by ultrasound examination of lung inflation during fetal breathing, prevention and treatment of resulting neonatal pulmonary distress, and prevention and treatment of accumulating pulmonary damage and fibrosis caused by parenchymal collagen weakness. We hope that the recognition of likely lung hypoplasia at birth, its evaluation, and timely treatment may save and improve many lives. Our translational studies include collaborations with many basic and clinical researchers on mechanisms of collagen-related disorders. Over the years, we assisted Dr. Marini in discovering novel forms of OI and characterizing underlying pathology. Together with Dr. Byers, we investigated arginine substitutions in type I collagen, demonstrating features similar to Gly substitutions. We assisted Dr. Bonnemann in characterization of a complex connective tissue disorder involving pathology of multiple tissues due to deficient function of prolyl 4 hydroxylase 1, an enzyme hydroxylating proline in Col-I. In collaboration with Dr. Stratakis, we described abnormal maturation and function of osteoblasts caused by deficiencies in catalytic and regulatory subunits of protein kinase A that disrupt cAMP signaling, which was reminiscent of McCune-Albright syndrome (constitutively overactive cAMP signaling). In collaboration with Dr. Leppert, we described abnormal composition of collagen deposited in uterine fibromas, which could be involved in the dysregulation of uterine fibroblasts underlying this pathology. We are collaborating with Dr. Otsuru on studies of growth plate pathology and growth deficiency in G610C mice. The most recent example of these collaborations has been a study of osteochondrodysplasia associated with heterozygous variants in COL1A2 and TRPV4 published this year. We are also pursuing more fundamental cell biology of procollagen biosynthesis by osteoblasts and fibroblasts, the goal of which is understanding presently unknown molecular mechanisms for subsequent translation into clinical research. For instance, observations made by us and others suggest that disruptions in secretory trafficking and degradation of procollagen might be involved in many pathologies spanning the entire lifespan, from skeletal dysplasia in early development to osteoporosis in aging. Better understanding procollagen trafficking and degradation might therefore reveal new therapeutic targets and approaches. To study these mechanisms, we developed novel fluorescent constructs of procollagen for live cell imaging. Contrary to popular models, we found procollagen to be delivered to Golgi from ERESs by rapidly moving transport vesicles that have no COPII coat and are dependent on COPI coat formation. We also discovered that misfolded procollagen is recognized at ERES and rerouted from the secretory pathway to a novel autophagy (lysosomal degradation) pathway we termed ERES microautophagy, in which ERESs containing misfolded molecules are directly engulfed by lysosomes. We are investigating the mechanism of the lysosomal recruitment to ERES as a potential target for therapeutic applications. We are also investigating whether ERES microautophagy is a general quality control mechanism utilized by cells for many proteins and not just procollagen. To facilitate these studies, we have utilized CRISPR/CAS gene editing technology for creating several osteoblast cell lines, in which endogenous procollagen is fluorescently tagged and can be manipulated by Flp-recombinase to introduce mutations and change the fluorescent tags. We are using these cells to confirm and expand our knowledge of procollagen trafficking and autophagy. We are also providing them to researchers outside NIH as a useful resource for their studies of collagen-related pathologies. Together, our translational and fundamental studies suggest that autophagy of misfolded procollagen is a crucial adaptation mechanism in OI. To test this hypothesis, we created an OI mouse model for autophagy manipulation by altering Atg5 expression. Using this model, we discovered that osteoblasts recycle misfolded procollagen primarily by ERES micro-autophagy not just in cell culture but also in vivo. We also observed that 3-4-fold reduction in Atg5 increased perinatal lethality of heterozygous G610C mice due to lung failure from < 10% to 50%. We hypothesize that deficient overall animal adaptation to the reduced Atg5 and its effects on the fetal breathing movements cause the increased lethality. Our latest breakthrough has been the discovery that misfolded procollagen is recognized at ERES and directed to microautophagy by an ERES COPII coat protein SEC24D. Osteoblast specific knockout of this protein in mice that lack any mutations in procollagen or its chaperones prevents microautophagy of misfolded molecules forming under normal conditions. The resulting misfolded procollagen accumulation in osteoblast ER, severe ER disruption, cell stress and osteoblast malfunction are similar to the effects of OI mutations in procollagen. We have demonstrated that the bone pathology in Cole-Capenter syndrome 2, which is associated with autosomal recessive SEC24D mutations, is caused deficient misfolded procollagen microautophagy rather than deficient secretion of normally folded molecules. Understanding the adaptation of osteoblasts and other cells to procollagen misfolding and mechanisms of misfolded procollagen degradation may help finding new treatment targets for improving the survival of OI and CCS babies as well as minimizing long-term pathology in these and other bone disorders.