70 KD Heat Shock and their associated cofactors

Heart, Lung, And Blood Institute

Investigators

Linked publications & trials

Abstract

Our laboratory studies the mechanism of action of the 70-kDa class of heat shock proteins (Hsp70s), which have been termed molecular chaperones because they are involved in the folding and unfolding of proteins and in the formation and dissociation of protein complexes. In these studies we have concentrated on exploring the role of Hsp70 in clathrin-mediated endocytosis, in particular its ability to dissociate clathrin from clathrin-coated vesicles. In many of their activities the Hsp70s require cofactors known as J-domain proteins that induce protein substrates to bind to Hsp70, and we previously discovered that uncoating also requires a J-domain protein, the minor 100-kDa clathrin assembly protein (AP), auxilin. Auxilin is a nerve specific protein and we later discovered that the non-neuronal homolog of auxilin is the 150-kDa protein GAK that is very similar to auxilin but also contains an N-terminal kinase domain. We then showed that C. elegans has a single gene for auxilin and when auxilin expression is inhibited by RNA-mediated interference, there is a marked inhibition of clathrin-mediated endocytosis which in turn causes the worms to arrest during larval development. We also showed that yeast has a single gene for auxilin and that when this gene is deleted the resulting haploid yeast mutants showed an increase of clathrin associated with vesicles and a corresponding decrease in free clathrin in the cytosol. From these data, we concluded that Hsp70 and auxilin are required for a fundamental step in clathrin-mediated endocytosis. We then showed that in vivo this fundamental step not only involves dissociation of clathrin after clathrin-coated vesicles bud off but also clathrin exchange occurring during invagination of the clathrin-coated pit. Our results showed that clathrin and the adaptor protein, AP2, exchange with free clathrin and AP2 in the cytosol. Likewise we found that clathrin and AP1 on the trans-Golgi network exchanged with free clathrin and AP1 in the cytosol. From these data we concluded that clathrin-coated pits at both the plasma membrane and the trans-Golgi network are dynamic structures. We also concluded that clathrin exchange is required for the structural rearrangement of clathrin that occurs as clathrin-coated pits invaginate. We then used permeabilized cells to demonstrate that Hsc70 and auxilin are required for this clathrin exchange showing that they induce clathrin dissociation from clathrin-coated pits. However, clathrin did not rebind to the pits in the permeabilized cells unless cytosol was present. Likewise we found that cytosol was also required for both the dissociation and rebinding of AP2. We are currently investigating what factors in the cytosol are involved in the rebinding of clathrin and in the dissociation and rebinding of AP2. In particular, in regard to the latter question we are investigating the role of PIP2 and phosphorylation and dephosphorylation in the exchange of AP2. During the past year we have investigated two other aspects of Hsc70 and auxilin activity. First, we used optical reconstruction techniques to determine how Hsc70 dissociates individual clathrin triskelions from clathrin baskets. The clathrin baskets used in these structural studies were polymerized using C58J, a hybrid assembly protein that we designed in which the J-domain of auxilin is fused to a 58-kDa portion of the clathrin assembly protein, AP180. At pH 6 C58J-clathrin baskets bind three Hsc70s per clathrin triskelion but uncoating does not occur unless the pH is raised to 7. Our structural studies suggest that three Hsc70s bind to the vertex of each clathrin triskelion in these baskets and these Hsc70s disentangle the clathrin triskelion from the legs of adjacent clathrins. The Hsc70s then remain bound to the vertex of the clathrin triskelion after it dissociates. Second, we investigated the effect of GAK RNAi on clathrin-mediiated endocytosis in tissue culture cells. Interestingly, these studies suggested that, not only are Hsc70 and GAK required for uncoating of clathrin-coated vesicles, but they are also necessary to chaperone the dissociated clathrin so that it is able to return to the membrane and form new clathrin-coated pits. We found that, in the absence of GAK, the number of clathrin-coated pits markedly decreased and large clathrin aggregates appeared in the cytosol. There was no membrane associated with these aggregates suggesting that they were not simply clathrin-coated vesicles that had not been uncoated. Therefore these data suggest that the three Hsc70s bound to each dissociated clathrin triskelion may be required to chaperone the clathrin in the cytosol and may also be involved in clathrin rebinding to new clathrin-coated pits. During the past year our studies on clathrin associated with the TGN have also provided results that are relevant to a crucial question regarding the mechanism by which the trans-Golgi-network (TGN) partitions between the two daughter cells during mitosis. Our results suggest that clathrin and several TGN-associated membrane proteins do not simply fragment during mitosis but actually divide on the mitotic spindle during metaphase. Thus our results suggest that, in contrast to other portions of the Golgi, the TGN does not form de novo in daughter cells from elements dispersed throughout the cytosol during mitosis but rather is partly formed from vesicles that divide on the mitotic apparatus. Furthermore, our results suggest that, following mitosis, the TGN-derived vesicles coalesce into clathrin-coated TGN compartments at the spindle poles before the rest of the Golgi becomes functional. In addition to these studies on permeablized cells we have continued our studies on both the auxilin and GAK knock-out mice. Interestingly, the homozygous auxilin knock-out mice seem to be born with an eye defect suggesting that brain maturation is abnormal. They also seem to have a defect in coat color, a particularly interesting result since it is now known that in addition to being present in the nervous system auxilin is also found in melanosomes. In regard to the GAK knock-out mice, we have developed a fibroblast line from the flox-flox GAK mouse and when we infect these cells with adeno-cre virus we should be able to observe what happens to cells that are completely deprived of GAK. We will also be able to determine if GAK missing its N-terminal kinase domain but containing the part of the molecule homologous to auxilin can rescue these cells.