Membrane topology and function of Der3/Hrd1p as a ubiquitin-protein ligase (E3) involved in endoplasmic reticulum degradation.

The endoplasmic reticulum contains a protein quality control system that discovers malfolded or unassembled secretory proteins and subjects them to degradation in the cytosol. This requires retrograde transport of the respective proteins from the endoplasmic reticulum back to the cytosol via the Sec61 translocon. In addition, a fully competent ubiquitination machinery and the 26 S proteasome are necessary for retrotranslocation and degradation. Ubiquitination of mutated and malfolded proteins of the endoplasmic reticulum is dependent mainly on the ubiquitin-conjugating enzyme Ubc7p. In addition, several new membrane components of the endoplasmic reticulum are required for degradation. Here we present the topology of the previously discovered RING-H2 finger protein Der3/Hrd1p, one of the new components of the endoplasmic reticulum membrane. The protein spans the membrane six times. The amino terminus and the carboxyl terminus containing the RING finger domain face the cytoplasm. Altogether, RING finger-dependent ubiquitination of malfolded carboxypeptidase yscY in vivo, as well as of Der3/Hrd1p itself in vitro and RING finger-dependent binding of Ubc7p, uncovers Der3/Hrd1p as the ubiquitin-protein ligase (E3) of the endoplasmic reticulum-associated protein degradation process.

Genetic interactions of Hrd3p and Der3p/Hrd1p with Sec61p suggest a retro-translocation complex mediating protein transport for ER degradation.

The endoplasmic reticulum contains a quality control system that subjects misfolded or unassembled secretory proteins to rapid degradation via the cytosolic ubiquitin proteasome system. This requires retrograde protein transport from the endoplasmic reticulum back to the cytosol. The Sec61 pore, the central component of the protein import channel into the endoplasmic reticulum, was identified as the core subunit of the retro-translocon as well. As import of mutated proteins into the endoplasmic reticulum lumen is successfully terminated, a new targeting mechanism must exist that mediates re-entering of misfolded proteins into the Sec61 pore from the lumenal side de novo. The previously identified proteins Der3p/Hrd1p and, as we show here, Hrd3p of the yeast Saccharomyces cerevisiae, are localised in the endoplasmic reticulum membrane and are essential for the degradation of several substrates of the endoplasmic reticulum degradation machinery. Based on genetic studies we demonstrate that they functionally interact with each other and with Sec61p, probably establishing the central part of the retro-translocon. In the absence of Hrd3p, the otherwise stable protein Der3p/Hrd1p becomes rapidly degraded. This depends on a functional ubiquitin proteasome system and the presence of substrate molecules of the endoplasmic reticulum degradation system. When overexpressed, Der3p/Hrd1p accelerates CPY* degradation in Delta(hrd3) cells. Our data suggest a recycling process of Der3p/Hrd1p through Hrd3p. The retro-translocon seems to be build up at least by the Sec61 pore, Der3p/Hrd1p and Hrd3p and mediates both retrograde transport and ubiquitination of substrate molecules.

Re-entering the translocon from the lumenal side of the endoplasmic reticulum. Studies on mutated carboxypeptidase yscY species.

Misfolded or unassembled secretory proteins are retained in the endoplasmic reticulum (ER) and subsequently degraded by the cytosolic ubiquitin-proteasome system. This requires their retrograde transport from the ER lumen into the cytosol, which is mediated by the Sec61 translocon. It had remained a mystery whether ER-localised soluble proteins are at all capable of re-entering the Sec61 channel de novo or whether a permanent contact of the imported protein with the translocon is a prerequisite for retrograde transport. In this study we analysed two new variants of the mutated yeast carboxypeptidase yscY, CPY*: a carboxy-terminal fusion protein of CPY* and pig liver esterase and a CPY* species carrying an additional glycosylation site at its carboxy-terminus. With these constructs it can be demonstrated that the newly synthesised CPY* chain is not retained in the translocation channel but reaches its ER lumenal side completely. Our data indicate that the Sec61 channel provides the essential pore for protein transport through the ER membrane in either direction; persistent contact with the translocon after import seems not to be required for retrograde transport.