Mucin-type O-linked glycosylation is a widespread and evolutionarily conserved protein modification catalyzed by a family of enzymes (PGANTs in Drosophila or pGalNAcTs in mammals) that transfer the sugar N-acetylgalactosamine (GalNAc) to the hydroxyl group of serines and threonines in proteins that are destined to be membrane-bound or secreted. Defects in this type of glycosylation are responsible for the human diseases familial tumoral calcinosis and Tn syndrome. Additionally, changes in O-glycosylation have been associated with tumor progression and metastasis. More recently, genome-wide association studies have identified the genes encoding the enzymes that are responsible for initiating O-glycosylation among those associated with HDL-cholesterol levels, triglyceride levels, congenital heart defects, colon cancer and bone mineral density. From these studies, it is apparent that this conserved protein modification has a multitude of biological roles. The focus of our research group is to elucidate the mechanistic roles of O-glycans during development in order to understand how they contribute to disease susceptibility and progression. Previous work from our group demonstrated that O-linked glycosylation is essential for viability in Drosophila. Our studies have demonstrated roles for this protein modification in the secretion of extracellular matrix (ECM) proteins. Specifically, we found that loss of one PGANT family member alters secretion of an ECM protein, thereby influencing basement membrane composition and disrupting integrin-mediated cell adhesion during Drosophila wing development. Likewise, we demonstrated that O-glycosylation also modulates the composition of the ECM during mammalian organ development, influencing integrin and FGF signaling, thereby affecting cell proliferation and growth of the developing salivary glands. These results highlight a conserved role for O-glycosylation in secretion and in the establishment of cellular microenvironments. Recent studies by our group have focused on imaging regulated secretion in vivo and ex vivo to gain a mechanistic understanding. The factors that govern secretory vesicle formation and regulated, polarized secretion have been challenging to decipher, as many mammalian cells lose polarity and secretory capacity once excised and cultured ex vivo. Additionally, in many systems, the small size of vesicular structures and the lack of markers for cargo impede the ability to image actin dynamics and secretory vesicles with sufficient resolution. Furthermore, limited genetic tools in many systems have restricted the ability to interrogate the function of specific genes in these processes. To overcome these limitations, we developed a system for high-resolution, real-time imaging of discrete steps in secretion in a living, secreting organ. Using Drosophila salivary glands, which maintain their polarity and boast secretory granules 5-100 times larger than mammalian secretory granules, we are able to image the formation of secretory granules and recruitment of factors at single-granule resolution. Using fly lines expressing fluorescently-labeled molecules, we are able to directly image secretory vesicle cargo, apical membrane dynamics, and actin and myosin recruitment in real time to define the temporal sequence of events occurring during regulated exocytosis. By performing 4-dimensional (4D) imaging of actin dynamics during secretion, we show rearrangement of cortical actin at the point of vesicle fusion followed by coordinated recruitment of actin from the plasma membrane to the fused vesicle. In vivo knockdown of Arp2 or Arp3, both of which are involved in the synthesis of branched actin, resulted in similar phenotypes, where secretory vesicles fused with the plasma membrane but were unable to collapse and secrete their cargo. These vesicles grew very large and in some instances, separated from the plasma membrane and floated back into the cytoplasm. Additionally, we identified the Arp2/3 activator WASp, as an additional regulator of exocytosis using this system. Taken together, our results suggest that the vesicles secretory coat is formed in 2 distinct stages by 2 different actin structures. We propose that linear actin is initially recruited to vesicles followed by the subsequent recruitment of the Arp2/3 complex and the activator WASp to form the branched actin network necessary to generate compression forces driving vesicle collapse and membrane integration. Moreover, vesicle compression is essential for full release of large, highly glycosylated, mucin-like cargo in this system, as has been proposed for other large cargos in cell-based systems. These results have implications for the importance of branched actin formation in other systems such as the digestive tract, where secretion of large, highly glycosylated/cross-linked cargo is essential for organ function and protection of epithelial cell surfaces. In summary, our study highlights coordinated actin clearance and directional recruitment from the plasma membrane as well as essential roles for branched actin in regulated exocytosis in a living, secreting organ. The powerful combination of Drosophila genetics with in vivo and ex vivo imaging allows one to rapidly interrogate the role of factors in many aspects of secretion, including vesicle biogenesis, vesicle movement, fusion with the plasma membrane, release of granule cargo and expansion of cargo once in the lumen. Given that many genes and mechanisms involved in secretion are conserved across species, this system can provide detailed mechanistic information regarding the function of many components involved in mammalian secretion. Other collaborative studies (with the Tabak laboratory) have investigated the effects of loss of O-glycosylation on aspects of mammalian development. In this study, we found that Galnt1 was essential for normal cardiac function and valve development in mice. In summary, mice deficient for Galnt1 suffered from cardiac myopathy and valve thickening, along with aortic and pulmonary valve stenosis and regurgitation. Examination of developing tissues in these animals revealed enlarged valves in adults that stemmed from increased cell proliferation in the outflow tract cushion during embryogenesis (at E11.5). Developing valves in Galnt1-deficient animals had a dramatic reduction in O-glycans along with increased accumulation of extracellular matrix proteins. Additionally, developing valves had increased levels of intact versican and decreased levels of cleaved versican (versican normally undergoes cleavage during development). Interestingly, we found reduced levels of the proteases ADAMTS1 and ADAMTS5, which are responsible for versican cleavage. We also observed increases in BMP and MAPK signaling in the developing valves of Galnt1-deficient animals. Taken together, we propose a model in which the absence of Galnt1 alters the formation/processing/remodeling of the extracellular matrix, thereby altering conserved signaling pathways that regulate cell proliferation during valve development. Furthermore, this study provides the first evidence for the role of this protein modification in heart valve development and may represent a new model for idiopathic valve disease. In summary, we are using information gleaned from Drosophila to better focus on crucial aspects of development affected by O-glycosylation in more complex mammalian systems. We are also using real-time imaging within living organs to define the specific processes by which O-glycosylation influences secretion. Our hope is that the cumulative results of our research will elucidate the mechanisms by which this conserved protein modification operates in both normal development and disease susceptibility.
Showing the most recent 10 out of 25 publications
