4SCIENTIFIC REVIEW Retrotranslocation: Endoplasmic Reticulum’s Junk Disposal Mechanism

The primary structure of polypeptides is converted to their final tertiary and quaternary structure by sequential maturation steps, and the endoplasmic reticulum (ER) provides the environment for the polypeptides to attain their proper 3-dimensional architecture. Proteins that misfold or fail to oligomerize with their partners (Chen et al., 1998; Wileman et al., 1990) are quickly degraded, as unfolded or unassembled proteins could interfere with normal cellular function. Retrotranslocation is the process by which terminally misfolded or unassembled ER proteins are translocated back into the cytosol for degradation mediated by the proteasomal machinery. Increasing amounts of evidence now support the fact that the same translocon pore that is involved in the translocation of polypeptides into the ER is also used for the retrotranslocation process. But questions, like how the misfolded proteins are recognized and targeted to the translocon pore, whether the process requires energy, and what pulls the polypeptides as they emerge out of the pore into the cytoplasm, remain to be elucidated. This review addresses our current knowledge about the retrotranslocation process. INTRODUCTION: HOW DOES THE ENDOPLASMIC RETICULUM DISPOSE OF MISFOLDED OR UNASSEMBLED PROTEINS? One hypothesis for the mechanism of protein disposal is selective degradation via the lysosomal pathway. However, a more plausible mechanism via a lysosome independent pathway has been recently verified by the degradation of many misfolded/unassembled proteins (Bonifacino et al., 1989) in the presence of lysosomal inhibitors. The other interesting aspect was that these misfolded proteins were endoglycosidase H sensitive and acquired neither sialic acid modification nor sensitivity to the endoglycosidase D (Klausner and Sitia, 1990), which pointed out that these proteins never left the endoplasmic reticulum (ER). Now, the question is how does the cell accomplish the pre–Golgi selective degradation of ER retained misfolded proteins or “ER associated degradation (ERAD)” (McCracken and Brodsky, 1996). The idea of degradation within the ER compartment is difficult to appreciate, as it would be detrimental for nascent polypeptides undergoing the process of folding. Initial reports about cytosol independent ER protein degradation by the “protease” ER-60 (Otsu et al., 1995) could not gain popularity, as later it was shown that ER-60 is in fact a molecular chaperone (Oliver et al., 1997). Also, the presence of a sub-domain in the ER specialized for degradation of misfolded or unassembled proteins could not be verified by electron microscopy (EM) (Klausner and Sitia, 1990). The other option for the ER to dispose the junk is to redirect the misfolded or unassembled proteins to the cytosol, where the ubiquitin-proteasomal pathway might degrade them . HOW DOES A PROTEIN THAT HAS ENTERED THE ENDOPLASMIC RETICULUM GET BACK TO THE CYTOPLASM? Identification of Sec61p associated with the retrotranslocation process came from the study of destruction of class I MHC heavy chain molecules in human cytomegalovirus (HMCV) infected cells (Wiertz et al., 1996a). The US2 gene product of HMCV pulls the class I MHC heavy chain out into the cytoplasm, where it is deglycosylated and destroyed by the proteasomal machinery. The deglycosylated breakdown product was shown to be associated with Sec61 complex. Retrotranslocation was shown to be non-specific for virus, when non-viral cells, which do not express viral gene product US2, showed a large proportion of class I MHC heavy chain associated with the Sec61 complex only when misfolding was induced by treatment with dithiothreitol (DTT) (Wiertz et al., 1996b). Also, in yeast, a misfolded protein was shown to be associated with the Sec61p by using an ER to cytosol export defective temperature sensitive Sec61p strain (Pilon et al., 1997). However, it should be noted that ATPase enzymes like FtsH in E. Coli (Kihara et al., 1999) and AAA proteases Yta10/Yta12 and Yme1 in mitochondrial inner membrane (Langer, 2000; Leonhard et al., 2000) have been identified that can degrade transmembrane proteins by extracting them from the lipid bilayer without the involvement of a translocon pore. Thus, a translocon independent pathway for retrotranslocation cannot be ruled out. HOW TO CONTROL THE TWO-WAY TRAFFIC? If Sec61p, a translocon channel, is involved in both the translocation and retrotranslocation process, then how does the channel regulate this two-way traffic? One possibility might be that there are at least two different classes of specialized translocon channels, where one