Topogenesis and intracellular sorting of membrane proteins
Proteins synthesized on cytosolic ribosomes must be sorted to the specific compartment(s) in which they perform their function. Most secretory and membrane proteins are first targeted to the endoplasmic reticulum (ER) and then distributed via the secretory pathway. Our research focuses on (1) how membrane proteins are inserted into the ER membrane and acquire a defined topology, and (2) how transport vesicles are formed at the trans-Golgi or endosomes, or in vitro from purified components and liposomes. We furthermore study the mechanism by which trafficking mutants of pro-vasopressin cause dominant diabetes insipidus.
Topogenesis of membrane proteins
The Sec61 translocon is a compact helix bundle that forms a pore for protein translocation with a lateral gate for the integration of transmembrane segments. Three distinct types of transmembrane domains can be distinguished with respect to the insertion mechanism: signal, stop-transfer, and reintegration sequences. We explore the requirements of these different types in terms of hydrophobicity, charged residues, and properties of flanking sequences by expressing model proteins in mammalian or yeast cells and anlyzing the resulting protein topologies. In yeast, we further analyze the translocon function by mutagenesis. In collaboration with Dominic Höpfner (Novartis), we characterize the action of novel translocation inhibitors to learn about the mechanism of action of the translocon.
Post-Golgi protein sorting
Little is known about how proteins exit the trans-Golgi. Sulfation is a trans-Golgi-specific modification useful to study post-Golgi traffic. To introduce sulfation sites, we have tagged proteins with short sequences for the attachment of (heavily sulfated) glycosaminoglycans (GAG). Interestingly, GAG attachment was found to affect protein traffic by inhibiting endocytosis and by accelerating trans-Golgi-to-cell surface transport both for secretory and membrane proteins. Since GAG chains are long, helical, semi-rigid polymers, we are now testing the hypothesis that molecular size affects transport mechanism.
Endosome identity, morphology, and transport are regulated by rab GTPases and their effectors. We are studying the role of rabaptin5, an effector of rab4 and rab5, and found it to interact with an initiator of autophagy. Our experiments suggest a role for rabaptin5 in endosomal quality control and autophagy.
Amyloid-like aggregation in protein sorting and disease
Autosomal dominant neurohypophyseal diabetes insipidus is a trafficking disease in which the hormone precursor provasopressin is mutated, retained in the ER, and causes progressive degeneration of vasopressinergic neurons. We have discovered that mutant provasopressin forms fibrillar aggregates in cell culture and in vitro. It has recently been proposed that prohormones form functional amyloids at the trans-Golgi to generate secretory granules. In this light, ER aggregation of mutant provasopressin may reflect a mislocalized physiological process. We identified two sequences independently responsible for fibrillar ER aggregation of provasopressin mutants: the N-terminal hormone sequence and the C-terminal glycopeptide. The same two sequences also contribute to sorting into granules, providing the first experimental in vivo evidence in support of granule formation by functional amyloids.
Post-Golgi protein sorting
Endosome identity, morphology, and transport are regulated by rab GTPases and their effectors. We are studying the role of rabaptin-5, an effector of rab4 and rab5, that associates with rabex5, the exchange factor of rab5. Based on mutational analysis, rabaptin-5 is found to control endosome morphology without affecting transferrin transport (determined by automated microscopy) in a manner that is incompatible with the prevailing model of rab5 feed-forward loop.
Little is known about how proteins exit the trans-Golgi. We use sulfation, a trans-Golgi-specific modification, to characterize the exit pathway and kinetics to the cell surface. If necessary, proteins of interest are tagged to introduce tyrosine-sulfation sites or short sequences for the attachment of (heavily sulfated) glycosaminoglycans (GAG). In this manner, we found GAG-attachment to accelerate exit kinetics and to change the exit pathway of model proteins. Similarly, the proteoglycan form of the amyloid precursor protein exits in a manner distinct from that of GAG-free splice variants.
Sulfation is a trans-Golgi-specific modification useful to study post-Golgi traffic. To introduce sulfation sites, we have tagged proteins with short sequences for the attachment of (heavily sulfated) glycosaminoglycans (GAG). Interestingly, GAG attachment was found to affect protein traffic by inhibiting endocytosis and by accelerating trans- Golgi-to-cell surface transport both for secretory and membrane proteins. We are analyzing the mechanistic and physiological implications for proteoglycan sorting. In endocrine cells, prohormones and granins are sorted at the trans-Golgi network into dense-core secretory granules by an entirely different mechanism. We found expression of granule cargo to be sufficient to generate granule-like structures in nonendocrine cells. Deletion analysis of chromogranin A showed that the same segments that are required for granule sorting in endocrine cells produce granule-like structures in fibroblasts. The results support the notion that self-aggregation is at the core of granule formation and sorting into the regulated pathway.
Diabetes insipidus: a degenerative trafficking disease
Autosomal dominant neurohypophyseal diabetes insipidus results from mutations in the precursor protein of the hormone vasopressin. Mutant precursors are retained in the ER of vasopressinergic neurons and cause cell degeneration. We discovered that pro-vasopressin mutants form disulfide-linked oligomers and develop large, fibrillar aggregations in fibroblast and neuronal cell lines (see Figure 2). Purified mutant pro-vasopressin spontaneously formed fibrils in vitro. Dominant diabetes insipidus thus belongs to the group of neurodegenerative diseases associated with fibrillar protein aggregates. We identified the vasopressin nonapeptide in the precursor sequence to be primarily responsible for aggregation in the ER, i.e. the same sequence that had been proposed to be responsible for amyloid aggregation into secretory granules at the trans-Golgi. The sequence physiologically important for cargo aggregation into the regulated secretory pathway thus is responsible for pathological aggregation of mutant precursors in the ER.