Fig 2: Identification and validation of HREs in the ADRP gene promoter. (A) Sketch map of the ADRP promoter region. The putative HREs are indicated by arrows. The nucleotide sequences are numbered in relation to the transcription initiation site, which is designated '+1'. P1, P2, P3 and P4 indicate the primers used for polmterase chain reaction amplification of the immunoprecipitated chromatin fragments in Fig. 2E. (B) Comparison of the conserved HRE and flanking nucleotides identified in the human, rat and mouse ADRP gene promoter regions. (C and D) Luciferase reporter assays were performed in MCF7 cells transfected with the constructs containing the indicated sequences from the human ADRP gene promoter region. Each transfection experiment was performed in triplicate. The relative mean luciferase activity in the cells under hypoxia is shown as the fold over the mean activity in the cells under normoxia. Error bars represent the mean ± standard deviation. pGL3-pro and EPO represent negative and positive control, respectively. (E) Mutant sequence of the putative HREs in the construct of Fig. 2D are indicated. (F) ChIP-PCR assays were performed using the indicated primers and antibodies, to demonstrate binding of HIF1α to the -33 HRE of ADRP in the MCF7 cells under normoxic and hypoxic conditions. PPP1R3C was used as a positive control. ADRP, adipose differentiation-related protein; HIF1, hypoxia-inducible factor-1; N, normoxiaa; H, hypoxia; HRE, hypoxia response element; EPO, erythropoietin; WT, wild-type; Mut, mutant; PPP1R3C, protein phosphatase 1, regulatory subunit 3C; Rel Luc Act, relative luciferase activity. *P<0.05 and #P>0.05.

Fig 3: ADRP is induced by hypoxia in an HIF-dependent manner. (A) mRNA levels of ADRP were analyzed using qPCR in MCF7 cells cultured in normoxia or hypoxia, or in medium containing 100 µM DFO for 24 h. The mRNA levels of ADRP were normalized to those of β-actin. The relative mRNA expression of ADRP is presented as the values in hypoxia, relative to normoxia. Data are expressed as the mean ± standard deviation.(B) Protein levels of HIF1α, ADRP and β-actin in the MCF7 cells were determined using western blotting under the different treatment conditions. Anti-β-actin antibody was used as a control for equal protein loading. (C) mRNA levels of ADRP and HIF1α were determined using qPCR in MCF7 cells transfected with siHIF1A and siCON. The relative mRNA expression levels are presented as the fold values of mRNA levels in cells transfected with siHIF1A/cells transfected with siCON. Data are expressed as the mean ± standard deviation (D) Protein levels of HIF1α, ADRP and β-actin in the transfected MCF7 cells were determined using western blot assays, as above. *P<0.05. ADRP, Adipose differentiation-related protein; HIF1, hypoxia-inducible factor-1; N, normoxia; H, hypoxia; DFO, deferoxamine mesylate salt; si, small interfering RNA; CON, control.

Fig 4: Detection and validation of ADRP knockdown. (A) mRNA levels of ADRP and HIF1A were determined using qPCR. The mRNA levels of ADRP and HIF1α were normalized to those of β-actin. The relative expression level of each mRNA is presented as the fold values in cells transfected with siADRP/siCON. Data are presented as the mean ± standard deviation of three independent experiments. (B) Protein levels of ADRP, HIF1A and β-actin were determined using western blot assays. β-actin was used as a loading control. *P<0.05 and #P>0.05. ADRP, adipose differentiation-related protein; HIF1, hypoxia-inducible factor-1; si, small interfering RNA; CON, control.

Fig 5: LD membrane proteins co-translationally inserted into the ER reach LDs. (A) HSD17B11 with an N-terminal OPG2 tag and a C-terminal FLAG epitope (O-HSD17B11-F) was expressed in HepG2 cells, which were subsequently loaded with oleic acid in the presence of 50 µM zVAD-fmk to inhibit cytosolic N-glycanase (Misaghi et al., 2004). Cells were homogenised, LDs isolated from the post-nuclear supernatant by flotation through a sucrose gradient, and fractions analysed for the presence of selected intracellular compartments by western blotting with the following organelle-specific markers: BAP31 (ER), ADRP (LDs) and tubulin (cytosol). Migration of ectopically expressed OPG2-HSD17B11-FLAG was established by western blotting with the anti-FLAG antibody. ‘1g’ indicates singly N-glycosylated OPG2-HSD17B11-FLAG, and ‘*’ denotes an EndoH-resistant species most likely crossreacting with the anti-FLAG antibody. Inset shows products recovered in fraction 1 with and without EndoH treatment prior to analysis as described. (B) The proportion of N-glycosylated OPG2-HSD17B11-FLAG in the LD fraction (see panel A) that corresponds to contaminating ER membranes and to authentic LD-associated species was calculated based on the relative levels of an ER marker, BAP31, in each fraction (see Materials and Methods). Data are mean±s.e.m. ****P<0.0001 (two-tailed unpaired t-test). n=4 biological replicates. (C) A schematic representation of the ectopically expressed LD membrane proteins and experimental setup used to address their accessibility to proteinase K (PK). Indicated LD membrane proteins and a control (fully membrane-spanning protein Sec61β) tagged with the FLAG epitope (‘F’) and a variant of the OPG2 tag with mutated N-glycan consensus sites (‘O*’) were transiently expressed in HeLa cells. The plasma membrane was then selectively permeabilised and protease accessibility tested either in the absence or presence of 1% (v/v) Triton X-100 (TX). Samples were resolved by SDS-PAGE and results were visualised by western blotting using anti-rhodopsin (OP) and anti-FLAG antibodies. Western blotting for an endogenous ER membrane protein, calnexin (CNX), was used to compare ER membrane integrity between samples and provide an internal control. Filled red square and black circles indicate candidate membrane-protected protease-resistant fragments.