Abstract
Endosomal internalisation and subsequent lysosomal degradation of membrane proteins is important for regulation of multiple cellular processes, among these the termination of receptor signalling and degradation of misfolded membrane proteins. ESCRT (Endosomal sorting complex required for transport) proteins are vital for the sorting of ubiquitinated membrane proteins into multivesicular bodies for subsequent degradation in the lysosome. In this study we generated two stable cell lines expressing the EGFP tagged ESCRT proteins Hrs and hVps22. Our goal was to utilise these cell lines for investigations into ESCRT protein dynamics, the relative order of ESCRT protein recruitment to the endosomes, and the endosomal localisation of ESCRT proteins. However, though the EGFP-Hrs cell line seemed to express a functional Hrs protein, the EGFP-hVps22 protein was completely cytosolic and could not be visualised on endosomes. hVps4, and its mouse homologue Skd1 is an AAA-type ATPase shown to be necessary for release of ESCRT-III proteins from endosomal membranes. We examined whether transient transfection of a dominant negative Skd1 (Skd1 E235Q) could be used for hVps22 visualisation on endosomes. This led to the surprising discovery that EGFP-Hrs does not localise to Skd1 EQ positive compartments, indicating that this protein might be released from endosomes by an hVps4-independent mechanism. Further investigations of proteins from all ESCRT complexes, led us to discover that Hrs or ESCRT-0 might be special in its mechanism for endosomal release. Transient transfection with Skd1 E235Q led to visualisation of hVps22 on Skd1 EQ positive compartments, but the protein was immobilised and this approach cannot be used for studying protein dynamics. We went on to investigate whether lower expression levels of the dominant negative protein hVps4A E233Q could allow endosomal visualisation of the ESCRT protein Tsg101, while evading immobilisation on hVps4A positive compartments. We found that this approach could be used for visualising mCherry-Tsg101, but cannot be used for studying protein dynamics as it slows the on-off kinetics of endosomal recruitment and dissociation. It might be used for studying relative order of ESCRT protein dissociation from endosomes, but the method needs further testing and optimisation.