Protein-folding homeostasis in the endoplasmic reticulum and nutritional regulation

Abstract

The flux of newly synthesized proteins entering the endoplasmic reticulum (ER) is under negative regulation by the ER-localized PKR-like ER kinase (PERK). PERK is activated by unfolded protein stress in the ER lumen and inhibits new protein synthesis by the phosphorylation of translation initiation factor eIF2α. This homeostatic mechanism, shared by all animal cells, has proven to be especially important to the well-being of professional secretory cells, notably the endocrine pancreas. PERK, its downstream effectors, and the allied branches of the unfolded protein response intersect broadly with signaling pathways that regulate nutrient assimilation, and ER stress and the response to it have been implicated in the development of the metabolic syndrome accompanying obesity in mammals. Here we review our current understanding of the cell biology underlying these relationships.

Figure 1.

A schematic overview of the eIF2α-phosphorylation-dependent gene-expression program in the context of the UPR. In addition to attenuating global protein synthesis, eIF2(αP) also up-regulates the translation of rare mRNA, exemplified by those encoding the transcription factors ATF4 and ATF5. This couples increased eIF2 phosphorylation to a gene expression program whose targets include the transcription factor CHOP, which activates a regulatory subunit of an eIF2(αP)-directed phosphatase, PPP1R15A/GADD34, and to enhanced expression of genes involved in amino acid import and assimilation and other more conventional target genes of the UPR that are coregulated by the two other branches of the UPR (IRE1, XBP1, and ATF6). The net effect of this program is to dephosphorylate eIF2α (a task that is aided by the constitutively expressed PPP1R15B/CReP) and thereby reverse translational repression and to enhance the capacity of the cell to synthesize and secrete proteins by its effects on amino acid metabolism and the ER. The near-immediate repression of protein synthesis upon PERK activation defends the cell against ER stress; however, the gene expression program has conflicting effects on levels of ER stress and appears to have evolved primarily to restore synthesis and secretion of proteins.

Figure 2.

Translational control in β cells. (A) Schema of the interplay between glycemic excursions and translational control in β cells. High serum glucose (i) stimulates pro-insulin translation in the β cell, increasing ER client protein load and promoting a physiological ER stress. The latter is modulated by PERK-mediated eIF2α phosphorylation (ii), which protects the β cell from ER stress while capping insulin synthesis (iii). Insulin resistance, which impairs glycemic control (iv), burdens the β cell by increased signaling in i. (B) A comparison of the effects of glucose excursions on pro-insulin translation in islets of Langerhans isolated from wild-type and PERK knockout mice. The arrows point to the ∼30% difference (Δ) in pro-insulin translation affected by PERK signaling (point ii in A). (C) Photomicrographs of pancreas from wild-type and PERK knockout mice of the indicated age stained with antiserum to insulin and glucagon. Note the profound loss of insulin-positive β cells in the PERK mutant. (Panel from Harding et al. 2001a; reprinted, with permission, from the authors.)

Figure 3.

Failure of homeostasis in the UPR. Faced with conventional levels of unfolded protein load (A), homeostasis in the ER lumen is maintained by a balance between factors that favor chaperone reserve by restraining protein synthesis, such as IRE1-mediated RIDD and PERK-mediated eIF2α phosphorylation and factors that favor protein synthesis [such as amino transporters, tRNA synthetases, or eIF2(αP)-phosphatases, PPP1R15A]. This balance was honed by years of natural selection. However, under conditions of usually heavy unfolded protein load, for example, caused by a mutation in an abundant secreted protein such as the Ins2^AKITA, homeostasis is favored by a different balance, by one with less PPP1R15A and more eIF2(αP) (B). This propensity for failure of homeostasis at the UPR’s normal set points explains the beneficial effects of PPP1R15A inactivation in certain unusual stressful situations.