Spatial regulation of UBXD8 and p97/VCP controls ATGL-mediated lipid droplet turnover

Abstract

UBXD8 is a membrane-embedded recruitment factor for the p97/VCP segregase that has been previously linked to endoplasmic reticulum (ER)-associated degradation and to the control of triacylglycerol synthesis in the ER. UBXD8 also has been identified as a component of cytoplasmic lipid droplets (LDs), but neither the mechanisms that control its trafficking between the ER and LDs nor its functions in the latter organelle have been investigated previously. Here we report that association of UBXD8 with the ER-resident rhomboid pseudoprotease UBAC2 specifically restricts trafficking of UBXD8 to LDs, and that the steady-state partitioning of UBXD8 between the ER and LDs can be experimentally manipulated by controlling the relative expression of these two proteins. We exploit this interaction to show that UBXD8-mediated recruitment of p97/VCP to LDs increases LD size by inhibiting the activity of adipose triglyceride lipase (ATGL), the rate-limiting enzyme in triacylglycerol hydrolysis. Our findings show that UBXD8 binds directly to ATGL and promotes dissociation of its endogenous coactivator, CGI-58. These data indicate that UBXD8 and p97/VCP play central integrative roles in cellular energy homeostasis.

Fig. 1.

UBAC2 regulates the partitioning of UBXD8 between the ER and LD. (A and B) Immunolocalization of endogenous UBXD8 (A; red) or exogenously expressed UBXD8-S (B; red) and LDs (BODIPY 493/503; green) in untreated or oleate-treated HeLa cells. (C) UBXD8-S (green) immunolocalization is restricted to the ER in HeLa cells coexpressing UBAC2-myc (red). (D) UBXD8-S partitioning into LD fractions is suppressed by coexpression of UBAC2-myc. Immunoblot analysis of oleate-treated HEK293 cells expressing the indicated constructs fractionated into ER-enriched (P) and LD-enriched (LD) fractions. Equivalent percentages of each fraction were analyzed. (E) Validation of shRNA depletion of endogenous UBAC2. (F and G) Depletion of endogenous UBAC2 promotes trafficking of endogenous UBXD8 to LDs, as evaluated by immunoblot analysis of equivalent percentages of P and LD fractions (F) and by immunofluorescence microscopy (G). In the micrographs, white boxes indicate the magnified regions. (Scale bars: 10 μm.)

Fig. 2.

UBXD8-mediated recruitment of p97/VCP to LDs inhibits LD turnover. (A and B) HeLa cells transfected with the indicated constructs were analyzed by immunofluorescence microscopy and BODIPY 493/503 staining (A), and the size and number of LDs were quantified (B). Cells are outlined in black (untransfected) or red (transfected). (C) UBAC2 depletion increases total LD area under basal conditions. LD content in HeLa cells expressing control or UBAC2 shRNA was analyzed by BODIPY 493/503 staining, and the area per cell was quantified. (D) UBXD8 recruits p97/VCP to LDs via its UBX domain. The distribution of endogenous p97/VCP (red) was analyzed in the presence of the indicated UBXD8-S constructs (green) in HeLa cells. Nuclei were stained with DAPI (blue). (E) p97/VCP ATPase function is required for LD homeostasis. U2OS cells expressing inducible p97/VCP(WT) or p97/VCP(EQ) were incubated in the presence or absence of 200 µM oleate and doxycycline, BODIPY 493/503-stained LDs were analyzed by immunofluorescence microscopy, and the area per cell was quantified. Representative images are shown in Fig. S4B. All graphical data are quantified as mean ± SEM. An asterisk indicates a significant difference (P < 0.05, t test) from the control based on n = 500–600 droplets from each of three independent biological replicates. In the micrographs, white boxes indicate the magnified regions. (Scale bars: 10 μm.)

Fig. 3.

LD-localized UBXD8 inhibits lipolytic degradation of LDs. (A) Timeline for oleate (200 µM) and triacsin C treatments (10 µM) for the experiments shown in B–E. (B–D) LD area per cell was quantified at time 0 to measure LD synthesis (B) or at the indicated time points to measure LD turnover (C and D) in transfected HeLa cells stained with BODIPY 493/503. (E) HeLa cells expressing the indicated constructs were treated according to the timeline shown in A, except that ¹⁴C-oleate (0.3 µCi) was included during the oleate pulse. Cellular lipids from the 0 and 3 h time points were extracted and separated by TLC, and the amount of ¹⁴C-oleate–labeled TAG was determined by scintillation counting. All graphical data are quantified as mean ± SEM. An asterisk indicates a significant difference (P < 0.05, t test) from the control based on three independent biological replicates. (Scale bars: 10 μm.)

Fig. 4.

UBXD8 binds ATGL and impairs ATGL-dependent LD turnover. (A) UBXD8 suppressed LD degradation induced by ATGL overexpression. Total LD area per cell was quantified in HeLa cells expressing the indicated constructs. (B) ATGL is required for LD up-regulation by UBXD8. LD area per cell was quantified in WT or ATGL^−/− MEFs expressing a control plasmid or UBXD8-S. (C) UBXD8 interacts with ATGL. Immunoblot analysis of UBXD8 complexes immunopurified from detergent-solubilized LD-enriched fractions isolated from oleate-treated HEK293 cells. (D) Assessment of the ATGL–UBXD8 interaction by bimolecular fluorescence complementation. HEK293 cells transiently transfected with the indicated C-terminal (Yc) or N-terminal (Yn) “split YFP” constructs were treated with 200 µM oleate for 24 h and analyzed by flow cytometry.

Fig. 5.

UBXD8 levels do not impact the proteasomal degradation of ATGL. (A) Oleate suppresses ATGL proteasomal degradation. Steady-state levels of endogenous ATGL in HEK293 cells incubated in the presence or absence of 200 μM oleate for 24 h were evaluated by immunoblot analysis. Where indicated, 10 µM MG132 was included for the final 12 h. (B) ATGL is a constitutive proteasome substrate. Cycloheximide (CHX) shutoff analysis of endogenous ATGL turnover in HEK293 cells in the presence or absence of 10 µM MG132. (C) UBXD8 does not influence ATGL levels. Steady-state levels of endogenous ATGL in HEK293 cells or HEK293 cells stably expressing UBXD8-S or UBAC2-S incubated in the presence or absence of 200 μM oleate for 24 h were evaluated by immunoblot analysis. (D) UBXD8 does not influence ATGL turnover. CHX shutoff analysis of endogenous ATGL turnover in control HEK293 cells or HEK293 cells stably expressing UBXD8-S or UBAC2-S.

Fig. 6.

UBXD8 regulates the association of ATGL with its activator CGI-58. (A) Endogenous ATGL is preferentially associated with UBXD8 over CGI-58. Immunoblot analysis of endogenous ATGL, UBXD8, or CGI-58 in LD-enriched fractions from oleate-treated HEK293 cells immunodepleted with the indicated antibodies. (B) Depletion of endogenous UBXD8 from LDs promotes the interaction of CGI-58 and ATGL. Immunoblot analysis of endogenous CGI-58 complexes from LD-enriched fractions isolated from oleate-treated control or UBAC2-S–expressing HEK293 cells. (C) Assessment of the ATGL–CGI-58 interaction by bimolecular fluorescence complementation. HEK293 cells transiently transfected with the indicated C-terminal (Yc) or N-terminal (Yn) “split YFP” constructs were treated with 200 µM oleate for 24 h and analyzed by flow cytometry. (D) LD-associated UBXD8 modulates the interaction between ATGL and CGI-58. Quantification of flow cytometry data of ATGL(S47A)-Yc and Yn-CGI-58 bimolecular fluorescence complementation in oleate-treated HEK293 cells expressing the indicated constructs. Normalized YFP fluorescence levels (from four independent experiments) are presented as mean ± SEM. An asterisk indicates a significant difference (P < 0.05, t test) from the control cells. (E) Model illustrating the role of UBXD8 and p97/VCP as regulators of cellular fat storage. The UBXD8–UBAC2 interaction establishes a dominant tethering mechanism that restricts a pool of UBXD8 from trafficking to LDs. In the ER, UBXD8 mediates the oleate-regulated inhibition of lipogenic proteins. Oleate treatment induces UBXD8 trafficking to LDs and stabilizes ATGL by inhibiting its constitutive proteasomal degradation. On LDs, UBXD8 binds ATGL and, through its recruitment of p97/VCP, functions as a noncompetitive inhibitor that mediates dissociation of the activated ATGL–CGI-58 complex.