Hepatic cytochromes P450: structural degrons and barcodes, posttranslational modifications and cellular adapters in the ERAD-endgame

Abstract

The endoplasmic reticulum (ER)-anchored hepatic cytochromes P450 (P450s) are enzymes that metabolize endo- and xenobiotics i.e. drugs, carcinogens, toxins, natural and chemical products. These agents modulate liver P450 content through increased synthesis or reduction via inactivation and/or proteolytic degradation, resulting in clinically significant drug-drug interactions. P450 proteolytic degradation occurs via ER-associated degradation (ERAD) involving either of two distinct routes: Ubiquitin (Ub)-dependent 26S proteasomal degradation (ERAD/UPD) or autophagic lysosomal degradation (ERAD/ALD). CYP3A4, the major human liver/intestinal P450, and the fast-turnover CYP2E1 species are degraded via ERAD/UPD entailing multisite protein phosphorylation and subsequent ubiquitination by gp78 and CHIP E3 Ub-ligases. We are gaining insight into the nature of the structural determinants involved in CYP3A4 and CYP2E1 molecular recognition in ERAD/UPD [i.e. K48-linked polyUb chains and linear and/or "conformational" phosphodegrons consisting either of consecutive sequences on surface loops and/or disordered regions, or structurally-assembled surface clusters of negatively charged acidic (Asp/Glu) and phosphorylated (Ser/Thr) residues, within or vicinal to which, Lys-residues are targeted for ubiquitination]. Structural inspection of select human liver P450s reveals that such linear or conformational phosphodegrons may indeed be a common P450-ERAD/UPD feature. By contrast, although many P450s such as the slow-turnover CYP2E1 species and rat liver CYP2B1 and CYP2C11 are degraded via ERAD/ALD, little is known about the mechanism of their ALD-targeting. On the basis of our current knowledge of ALD-substrate targeting, we propose a tripartite conjunction of K63-linked Ub-chains, P450 structural "LIR" motifs and selective cellular "cargo receptors" as plausible P450-ALD determinants.

Keywords: 26S proteasome; CHIP; Cytochromes P450; E2/E3 ubiquitin ligases; ERAD; K48 & K63 ubiquitination; autophagic lysosomal degradation; gp78 E3; p62; p97; phosphodegrons.

Fig. 1. CYP3A4 ERAD/UPD: Known cellular participants

See discussion for details.

Fig. 2. gp78-mediated ubiquitination of CYP2E1: Involvement of the same gp78 basic patches involved in its intermolecular electronic interactions with CYP3A4

This was documented in in vitro functionally reconstituted E1/E2/E3 systems in the presence or absence of CYP3A4 or ATP, using either recombinant gp78 wild type (WT) or its patch mutant (PT) wherein its positively charged residues R307-R308-R310-H312-K313/Q584-R585-K586 (known to interact with certain CYP3A4 DEST clusters) were mutated to Ala via site-directed mutagenesis (Wang et al., 2015). Color code wheel: Red>orange>yellow>green>blue> indigo>violet.

Fig. 3. Autophagy in mammalian cells

The major steps in ALD of cellular proteins and organelles are illustrated. ALD-inhibitors that block the various steps and serve as diagnostic probes are also shown. See discussion for details.

Fig. 4. ³⁵S-pulse-chase analyses of CYP3A and CYP2B time-course in cultured rat hepatocytes

Cells were pulsed upon pretreatment for 4 days with either phenobarbital, 1 mM (A) or dexamethasone, 10 μM (B), following which they were pulse-chased with ³⁵S-methionine/cysteine for 1 h in a methionine-cysteine-free WEM culture medium as described previously (Kim et al., 2010). The culture medium was then exchanged with a medium containing cold methionine/cysteine, and harvested at 0, 1, 3 and 6 h thereafter. In parallel, similarly ³⁵S-pulse-chased cultures were treated at time 0 h with either MG132 (10 μM) or 3-MA (5 mM)/NH₄Cl (50 mM). Cell lysates were immunopreciptated with either anti-CYP2B1 (A) or anti-CYP3A23 (B) antibodies. Aliquots of immunoprecipitates were subjected to SDS-PAGE and fluorography with Typhoon imaging (left panels). Smaller aliquots (10 μL) were subjected to liquid scintillation counting (right panels). CT, control. Arrows depict the parent 50 kDa ³⁵S-labeled P450 species.

Fig. 5. CYP2E1 and CYP3A4 structures with linear and “conformational” phosphodegrons

Previously identified phosphorylated S/T-residues (green) and ubiquitinated K-residues are shown. The K-residues ubiquitinated by gp78-complex are colored red, those ubiquitinated by CHIP-complex in blue, and those ubiquitinated by both E3-complexes in magenta. Note that these residues lie within or in the vicinity of linear loops or clusters of pS/pT (green), D and E residues colored cyan and orange, respectively.

Fig. 6. Plausible linear and “conformational” phosphodegrons of other major human liver P450s

Surface DEST clusters are shown with vicinal K-residues shown in yellow, S/T phosphorylation sites in green, D and E residues in cyan and orange, respectively. Note that although neither the ubiquitination nor the phosphorylation sites have yet to be conclusively identified, their situation within or vicinal to DEST clusters is highly intriguing.

Fig. 7. Plausible mechanisms of P450 autophagic targeting

Roles of hepatic p97/Npl4/Ufd1 complexes in ER-extraction, p62 and NBR1 heteromers in autophagic targeting of ubiquitinated P450s via their LIR- and UBA-domains to the phagophore, and of p97/p47 complexes in subsequent autophagosome biogenesis through fusion of the phagophore ends.

Fig. 8. CYP2B4 versus CYP2B6 surface LIR-motifs

Two different orientations of CYP2B4, and CYP2B6 are shown based on their individual structures (See discussion). The critical W₀ aromatic residue of the essential LIR W/F/YXXL/I/V core motif is depicted as follows: Trp (blue), Phe (green), and Tyr (magenta), and the other LIR residues in yellow. The extended LIR motif (D/E/X-D/E/X-D/E/X-W/F/Y-D/E/X-D/E/X-L/I/V) is also depicted in yellow, when present in the P450 primary sequence. Note that these critical W₀-residues are largely buried in CYP2B6, but not CYP2B4.

Fig. 9. Surface LIR-motifs of other human P450s

Two different orientations of CYP2C9 and CYP2D6 (A) and CYP2E1 and CYP3A4 (B) are shown based on their individual structures (See discussion). The critical W₀ aromatic residue and the other LIR residues of the essential LIR W/F/YXXL/I/V core motif are depicted as follows: Trp (blue), Phe (green), and Tyr (magenta), and the other LIR residues in yellow. The extended LIR motif (D/E/X-D/E/X-D/E/X-W/F/Y-D/E/X-D/E/X-L/I/V) is also depicted in yellow, when present in the P450 primary sequence. Note that these critical W₀-residues are largely buried in CYP3A4.

Fig. 9. Surface LIR-motifs of other human P450s

Two different orientations of CYP2C9 and CYP2D6 (A) and CYP2E1 and CYP3A4 (B) are shown based on their individual structures (See discussion). The critical W₀ aromatic residue and the other LIR residues of the essential LIR W/F/YXXL/I/V core motif are depicted as follows: Trp (blue), Phe (green), and Tyr (magenta), and the other LIR residues in yellow. The extended LIR motif (D/E/X-D/E/X-D/E/X-W/F/Y-D/E/X-D/E/X-L/I/V) is also depicted in yellow, when present in the P450 primary sequence. Note that these critical W₀-residues are largely buried in CYP3A4.

Fig. 10. K₄₈- versus K₆₃-linked P450 ubiquitination in HepG2 cells

Cells were transfected for 48 h with a C-terminally tagged Myc/His₆ P450 plasmid. The vector-containing medium was removed, cells washed, and then fresh medium added with MG-132, or 3MA/NH₄Cl or no inhibitor (Control, CT). Six h later, cells were harvested and lysates immunoprecipitated with antibodies to anti-Myc antibodies, and immunoblotted with specific anti-K₄₈-Ub or anti-K₆₃-Ub IgGs. Note that all these P450s undergo both types of ubiquitination. However, K₄₈-linked Ub-chains accumulate upon inhibition with the proteasomal inhibitor MG-132, whereas K₆₃-linked Ub-chains accumulate upon 3MA/NH₄Cl-mediated ALD inhibition. This may indicate that the type of ubiquitination per se is not a determinant of P450 UPD or ALD. Additional factors (structural degrons, barcodes or adapters) must be involved.

Fig. 11. Cycloheximide-chase analyses of P450 turnover with UPD- or ALD-diagnostic probes

HepG2 cells were cultured and transfected with each P450 plasmid vector for 48 h as in Fig. 10. After washing and medium exchange, cells were treated with cycloheximide (100 μM) with or without MG132 or 3MA/NH₄Cl as in Fig. 10, followed by harvesting of cells at various times, and immunoblotting of lysates with an anti-Myc antibody. A. Temporal plots of parent 50 kDa P450 species, following densitometric quantification. B. Overexposure of the same immunoblots to reveal HMM ubiquitinated Myc-tagged-species. Note that all these P450s are stabilized by the UPD-inhibitor, MG132 to some extent. However, only CYP2D6 and CYP2E1 parent and HMM-species are significantly stabilized by the ALD probe 3MA/NH₄Cl. Even CYP2C9 ( 50 kDa species) shows some stabilization upon 3MA/NH₄Cl-treatment at 24 h. No stabilization of CYP2B6 is appreciably detected.

Fig. 12. Relative role of CHIP- or gp78-E3 ligases in human P450 degradation

Myc/His6-tagged P450 plasmids were cotransfected with either HA-tagged CHIP or gp78 E3 ligases or their structural deletion constructs in HepG2 cells. As previously, CYP2E1 and CYP3A4 degradation is known to require both E3 ligase complexes for maximal degradation, and served as controls. Degradation of CYP2C9 and CYP2D6 similarly required both functionally relevant CHIP and gp78 E3-ligases, as the deletion of the CHIP catalytic U-box or Hsp70-interacting domain or of the gp78-E2-interacting or Cue-domains aborted their degradation. CYP2B6 is ubiquitinated (Fig. 10), but its degradation appears to be independent of either E3 ligase.