Visualizing Protein Sorting in the Golgi Apparatus

In this work I investigated the mechanisms governing protein sorting in the Golgi. The cisternal maturation model of Golgi function postulates that anterograde movement of protein cargoes occurs passively during maturation while resident Golgi proteins are recycled via vesicular transport. Despite suggestive evidence, these predictions had not been tested. Cisternal maturation has been visualized in budding yeast; however, there has not been a tool to enable simultaneous visualization of Golgi cisternae and cargo proteins. Therefore, I first engineered a tool to visualize protein cargoes during Golgi maturation. The tool, Erv29- dependent Secretory Cargo (ESCargo), traps a fluorescent protein in aggregates within the ER. The aggregates are dissolved to release the fluorescent protein for synchronous transport through the secretory pathway. I demonstrated this tool behaves like a bona fide secretory cargo by visualizing its sequential localization with early and then late Golgi cisternae, tracking oligosaccharide modifications, and detecting cargo secreted into the medium. With this tool I verified a key prediction of the cisternal maturation model: secretory cargo remains within the lumen of maturing Golgi cisternae. Unexpectedly, I also discovered that secretory cargo can be recycled within the Golgi in a manner dependent on the AP-1 adaptor complex. Next I dissected how different classes of cargo proteins transit the Golgi. Most models postulate that all cargoes uniformly transit the Golgi to the latest stage of the trans-Golgi network (TGN) where they are sorted into vesicles. To test this hypothesis, I modified ESCargo for transport to the vacuole. I found that vacuolar cargoes exit the Golgi at the beginning of the late Golgi stage, significantly earlier than secretory cargoes. The departure was dependent on the Gga adaptors but not the AP-1 adaptor complex. Thus, the sorting of different cargoes is kinetically segregated during Golgi maturation. ESCargo utilizes evolutionarily conserved machinery and should function in other model organisms. I, along with several collaborators, tested the efficacy of ESCargo in Saccharomyces cerevisiae, mammalian tissue culture, Drosophila melanogaster, and the ciliate Tetrahymena thermophila. In all model organisms tested, ESCargo successfully generated a synchronized cohort of fluorescent secretory cargo - thus permitting comparative studies of secretory cargo trafficking between species. Finally I aimed to clarify the role of the AP-1 adaptor complex in intra-Golgi recycling. Using ESCargo, I found that secretory cargo recycling requires AP-1 and the mannosyltransferase Mnn1. Further, I demonstrated that inhibiting endocytosis in AP-1 mutants causes the mislocalization of resident Golgi enzymes and a severe defect in cisternal maturation. These results suggest that endocytosis acts as a backup recycling pathway for leaked late Golgi proteins in AP-1 mutants. Thus, AP-1 mediates an intra-Golgi recycling pathway that drives cisternal maturation. The majority of the data collected in this work is video microscopy. To appreciate the dynamic nature of protein sorting in the Golgi, it is essential to visualize these processes in real time. The movies associated with each chapter are included as supplementary files online. The first frame and a legend are located at the end of each chapter for reference