MEM-alpha cell culture medium, OptiMEM reduced-serum medium, fetal bovine serum (FBS), penicillin and streptomycin were obtained from Gibco. Rosiglitazone, T0070907, MG132, bafilomycin A1, chloroquine diphosphate, pepstatin A and E64d were from Tocris. Anti-LRP1 (ab92544); anti-LDLR (ab52818) anti-β-actin (ab49900); secondary antibodies: HRP-conjugated goat anti-rabbit IgG (ab97051), HRP-conjugated goat anti-mouse IgG (ab97023), Alexa Fluor 488 conjugated goat anti-rabbit IgG (ab150077) and Alexa Fluor 594 conjugated goat anti-mouse IgG (ab150116), and concanamycin A were from Abcam. The anti-LC3α/β antibody (SC-398822) was purchased from Santa Cruz Biotechnologies. The anti-LAMP1 (D4O1S) and anti-caveolin-1(D46G3) were purchased from Cell Signaling. 6-Carboxyfluorescein-labeled (FAM) Aβ_1-42 was obtained from Anaspec. TRIzol reagent and the transcriptor first strand cDNA synthesis kit were from Thermo Fisher and Roche, respectively. DMSO and other reagents were from Sigma–Aldrich.

Hepatocellular carcinoma HepG2 cells were obtained from the American Type Culture Collection (HB-8065, ATCC). Cells were seeded in MEM-alpha medium, supplemented with 10% FBS and 1% penicillin/streptomycin. Cells were then grown in a humidified environment with 5% CO₂ until they were confluent. For experiments, HepG2 cells were seeded into 12-well (2.5 × 10⁵ cells/well) or 6-well (5 × 10⁵ cells/well) plates and were cultured for 48 h, for uptake and protein analysis experiments, respectively. Cells were then exposed to different concentrations of RGZ (3 nM–30 μM) in reduced-serum medium (OptiMEM). The final concentration of DMSO in each experiment was less than 0.5%. Other experiments were carried out in MEM containing 10% FBS, as explained in the Supplementary Information Section.

Following treatments, RNA was extracted from cells with TRIzol reagent. Isolated RNA was quantified utilizing a NanoDrop 2000 spectrophotometer (Thermo Fisher). cDNA was then prepared with the transcriptor first strand cDNA synthesis kit according to the manufacturer’s protocol. The final cDNA solution was stored at -20°C prior to analysis.

cDNA from samples was analyzed and quantified by RT-qPCR using TaqMan assay probes for LRP1 (Hs00233856_m1), PPARG (Hs00234592_m1) and ACTB (Hs01060665_g1) in a StepOne Plus station (Applied Biosystems). This reaction was carried out in triplicate in microamp optical 96-well plates in a total volume of 10 μL/well. The relative LRP1 expression levels were calculated as the fold-change determined by the ΔΔCt method with ACTB representing a housekeeping gene.

HepG2 cells were seeded into 96-well plates and were exposed to different concentrations of RGZ. Treatments were carried out with the control group consisting of DMSO-treated cells. Cell viability was determined by the colorimetric method utilizing 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS, CellTiter 96^®AQueous One Solution, Promega). Cell viability was determined at 490 nm and was calculated as percent of control.

HepG2 cells were rinsed twice with ice-cold PBS and proteins were extracted with M-PER and MEM-PER, for whole cell lysis and membrane isolation, respectively (both from Thermo Fisher). These lysis buffers contained Halt protease, phosphatase inhibitors and EDTA (Thermo Fisher). The protein concentration was determined by the colorimetric bicinchoninic acid assay (BCA assay, Thermo Fisher). Equal amounts of total protein from cell lysates were separated by SDS-PAGE (20 and 40 μg for LRP1 and LC3, respectively). Proteins from the gel were then electro-transferred onto 0.45 μm nitrocellulose and 0.2 μm PVDF membranes for LRP1 and LC3, respectively.

The membranes were then blocked for 1 h at room temperature (rt), with either 5% non-fat powdered milk dissolved in TBS-T or 5% bovine serum albumin in TBS-T, for the nitrocellulose and PVDF membranes, respectively. Following blocking, membranes were incubated overnight at 4°C with the primary antibodies anti-LRP1 (1:10,000), anti-LDLR (1:2,000), anti-caveolin1 (1:1,000) and anti-LC3 (1:500). The membranes were subsequently washed three times with TBS-T and were then exposed to HRP-conjugated goat anti-rabbit IgG (1:2,000) or HRP-conjugated goat anti-mouse IgG (1:2,000) for 1 h at rt. β-Actin was used as the loading control; the membranes were incubated with HRP-conjugated mouse polyclonal anti-β-actin (1:10,000) for 1 h at rt. SuperSignal West Pico chemiluminescent substrate (Thermo Fisher) was added to the membranes and they were incubated for 5 min. Bands were visualized with the C-DiGit blot scanner (Licor Technologies).

HepG2 cells were grown in MEM alpha medium. The LRP1 construct (Mini LRP-IV-EGFP) was kindly provided by Dr. Michel Khrestchatisky and Dr. Marion David (Université d’Aix-Marseille and Vect-Horus, Marseille, France, respectively; Perrot et al., 2012). Cells were transiently transfected with 1.5 μg of pDNA using Lipofectamine 2000 (Thermo Fisher), following manufacturer’s protocol.

For immunocytochemistry, HepG2 cells were cultured into 8-well Lab-Tek^TM II Chamber Slides (Nunc^TM) and were then treated with either RGZ or inhibitors. Cells were rinsed twice with ice-cold PBS, fixed with 4% paraformaldehyde in PBS (PFA, Boston Bioproducts) for 15 min at rt, and were permeabilized with 0.1% Triton-X (Sigma–Aldrich) for 10 min. The slides were blocked with 10% goat-serum, and 0.3 M glycine in PBS-Tween 20 (0.1%) for 1 h at rt. Subsequently, the antibodies anti-LRP1 (1:100), anti-LAMP1 (1:50) and anti-LC3 (1:50) were added and slides were incubated overnight at 4°C. The slides were then washed 3 times for 5 min each with PBS-T and were incubated with Alexa Fluor 488 conjugated goat anti-rabbit IgG and Alexa Fluor 594 conjugated goat anti-mouse IgG (each at 1:400 dilution) for 1 h at rt. Following immunostaining, slides were mounted with diamond mounting medium containing DAPI (Thermo Fisher). Slides were then visualized with the Leica TCS SP8 confocal microscopy station and the pictures were digitized with the Leica Application Suite X software. Recordings of the 3D-reconstruction can be found in the Supplementary Information Section. For time-lapse imaging microscopy, HepG2 cells were cultured in 35 mm μ-Dishes (Ibidi) overnight. Following transfection with the mini-LRP-IV-EGFP plasmid, cells were treated with either vehicle or RGZ (30 μM) for 22 h. Deep red lysotracker (Thermo Fisher) was added to visualize the lysosomal compartment. Single z-plane images were recorded every 3 min for the next 2 h from both control and RGZ-treated cells. Cells were visualized with the Leica TCS SP8 confocal microscopy station and pictures were digitized with the Leica Application Suite X software. Recordings of time-lapse imaging can be found in the Supplementary Information Section.

Synthetic FAM-Aβ_1-42 peptide was resuspended as previously described (Fuentealba et al., 2010; Kanekiyo et al., 2013). Briefly, FAM-Aβ_1-42 was resuspended in 1,1,1,3,3,3-hexafluoroisopropanol and aliquoted into amber tubes. The solvent was evaporated with a stream of nitrogen gas and tubes containing the dried film were stored at -20°C until they were utilized for uptake studies. For experiments, the films were resuspended in DMSO to obtain a 100 μM stock solution and were then diluted to 500 nM FAM-Aβ_1-42 in OptiMEM. HepG2 cells were treated with RGZ (30 μM) for 24 or 48 h. Following this treatment, HepG2 cells were incubated with fresh FAM-Aβ_1-42 (500 nM) for 30 min, 1 and 2 h. Cells were then lysed with M-PER, as above, and fluorescence intensity was monitored with the Synergy HT plate reader (Biotek), with excitation at 485/20 nm and emission at 528/20 nm. FAM-Aβ_1-42 uptake is expressed as relative fluorescence units per mg protein (RFU/mg).

All statistical analyses were conducted using GraphPad Prism 6. Data are reported as mean ± SEM for at least three independent experiments. Comparisons between two or multiple groups were performed by a two-tailed Student’s t-test and one-way analysis of variance (ANOVA), respectively. The ANOVA was followed by a post hoc Dunnet’s test, in which experimental groups were compared to controls (cells treated with vehicle). A value of p < 0.05 was considered to demonstrate a significant difference between treatments.

Previous studies demonstrated the modulation of LRP1 by the transcription factor PPARγ in cells treated with RGZ (Gauthier et al., 2003; Moon et al., 2012a). In order to confirm the RGZ effect on LRP1 expression, we replicated the experiments previously reported, which utilized HepG2 cells (Moon et al., 2012a). In agreement with the literature, we observed the transcriptional activation of LRP1 mRNA in cells treated with RGZ at 3 and 10 μM for 24 h (Figure 1A). However, we did not observe increases in LRP1 protein levels following either 24 or 48 h exposures to these concentrations of RGZ (Figures 1C,D). Other variables were also tested, such as incubation time, RGZ concentration and presence of serum in the treatments, but we did not observe the LRP1 protein upregulation previously described (Moon et al., 2012a, Figure 1 and Supplementary Figure 1). In contrast, higher concentrations than 10 μM RGZ decreased both mRNA and protein LRP1 levels significantly after 24 and 48 h treatments. In addition, the effect of RGZ seemed to be specific to reduce LRP1, since LDLR (another member of the LDL receptor family that undergoes clathrin-dependent endocytosis) and caveolin-1 (a marker for caveolae-dependent endocytosis) were not altered in the presence of RGZ (Figures 1C,D). We established that these effects were not due to cytotoxicity, as cell viability was maintained at these RGZ concentrations (Figure 1B and Supplementary Figure 1).

The LRP1 gene contains the PPRE that is recognized by activated PPARγ to begin the transcription of LRP1 (Gauthier et al., 2003; Moon et al., 2012a,b; Kim et al., 2014). However, we observed that higher concentrations than 10 μM RGZ reduced LRP1 mRNA levels in a time- and concentration-dependent manner (Figure 1A). Previous studies demonstrated that high concentrations of PPARγ agonists, including RGZ and pioglitazone, reduce PPARγ levels (Hauser et al., 2000; Kilroy et al., 2009). We observed that LRP1 mRNA reduction is a consequence of the PPARγ downregulation resulting from exposures of cells to RGZ at concentrations higher than 10 μM in a time- and concentration-dependent manner, as shown in Figure 2A. In addition, we tested whether PPARγ antagonism by T0070907 (T007) affects LRP1 protein levels. Our observation was that PPARγ antagonism reduced LRP1 protein levels in a concentration-dependent manner (Figure 2B).

While reduced LRP1 mRNA levels can be explained by PPARγ downregulation, we observed that high concentrations of RGZ reduced LRP1 protein levels to a greater extent than LRP1 transcription (Figure 1). This suggests that in addition to the transcriptional effect on LRP1 levels, a possible proteolytic mechanism might be affecting LRP1 protein levels in RGZ-treated HepG2 cells. We then investigated two proteolytic degradation mechanisms as possible explanations for these observations: the ubiquitin proteasome system (UPS) and lysosomal degradation.

The ubiquitin proteasome system (UPS) is the primary mechanism for protein degradation in eukaryotic cells (Vilchez et al., 2014). Previous studies have demonstrated that the pharmacological activation of PPARγ promotes its own proteolytic degradation via UPS (Hauser et al., 2000; Kilroy et al., 2009). In addition, it has been proposed that UPS is the key regulator for LRP1 degradation in a number of cell lines, including HepG2 (Melman et al., 2002; Cal et al., 2013). The use of MG132, a potent UPS inhibitor, in HepG2 cells treated with RGZ (30 μM) for 24 h did not prevent LRP1 protein degradation, as shown in Figure 2C. We also observed that MG132 reduced LRP1 protein levels in a concentration-dependent manner (Figure 2C and Supplementary Figure 2). These results excluded the possibility that LRP1 degradation occurred by the UPS pathway in RGZ-treated HepG2 cells. This suggested that RGZ reduced the amount of LRP1 present by a mechanism other than UPS.

Since LRP1 has been demonstrated to be involved in the transport of its ligands to the lysosomal compartment for degradation (Laatsch et al., 2012; Gan et al., 2014; Zhao et al., 2015), we decided to study the status of lysosome proteolytic degradation in RGZ-treated HepG2 cells. Cells were incubated with bafilomycin A1 (BAF), a vacuolar H⁺-ATPase inhibitor which blocks lysosomal activity, in the presence or absence of RGZ (30 μM) for 24 h. Western blot analysis demonstrated increased LRP1 protein levels, up to a 4-fold change, in cells co-treated with BAF (10 nM) and RGZ (30 μM) compared to the control, as depicted in Figure 2D. These results suggest that during RGZ treatment, LRP1 undergoes proteolytic degradation through the lysosomal system.

A long exposure of RGZ-treated HepG2 cells to BAF might affect other non-specific cellular events, such as UPS activation and protein aggregation, as previously reported (Korolchuk et al., 2009; Rubinsztein et al., 2009). We reduced BAF incubation time to the last 4 h of the experiment (total time of 24 h). In addition, cells were exposed to a mixture of the cathepsin inhibitors pepstatin A (PepA) and E64d, each at 30 μM. The purpose of the PepA/E64d mixture was to reduce lysosomal activity. As expected, RGZ reduced LRP1 levels, determined both by immunocytochemistry and Western blot analysis. On one hand, BAF treatment significantly increased LRP1 signal both in the presence or absence of RGZ, but PepA/E64d did not prevent LRP1 degradation, probably because PepA and E64d act downstream from BAF (Figures 3A, 4A). This suggests that BAF is capable to prevent the lysosomal degradation of LRP1 caused by RGZ.

BAF inhibits the last stage of autophagy, consisting of lysosomal degradation (Klionsky et al., 2016); and RGZ has been associated with autophagy in cancer cell lines (Zhou et al., 2009; Vara et al., 2013). To assess whether autophagy induced by RGZ is responsible for LRP1 degradation, the autophagosome marker LC3 was analyzed by immunocytochemistry and Western blot. Our observations demonstrated that RGZ (30 μM) induced autophagy by the presence of LC3 puncta, monitored by immunocytochemistry and LC3-II generation by Western blot. Moreover, the presence of BAF induced LC3 accumulation (Figures 3A, 4A). In addition, reduced levels of LRP1 were also observed in EBSS-starved HepG2 cells, but BAF prevented LRP1 degradation (Supplementary Figure 3). LRP1 did not co-localize with LC3, suggesting that LRP1 degradation induced by RGZ is not an autophagy-dependent process, but RGZ might promote LRP1 lysosomal degradation (Figure 4A).

We found that BAF prevented LRP1 proteolytic degradation induced by RGZ (Figures 2D, 3A). We then tested whether other lysosomal activity inhibitors would prevent LRP1 degradation in 30 μM RGZ-treated cells. Our observations were that 10 nM concanamycin A (CMA), another vacuolar H⁺-ATPase inhibitor, had an effect similar to that of BAF in RGZ-treated cells. On the other hand, the lysosomotropic agent chloroquine (CQ, known to increase the lysosomal pH and thus reduce activity of some lysosome proteolytic enzymes) at 30 μM did not prevent RGZ-induced LRP1 degradation (Figure 3B). These observations suggest that lysosomal enzymes which are affected by CQ and PepA/E64d, are not responsible for LRP1 degradation. Taken together these results suggest that BAF and CMA, which act more upstream than CQ and PepA/E64d, prevented LRP1 lysosomal degradation caused by 30 μM RGZ.

Since previous observations suggested that 30 μM RGZ might promote LRP1 lysosomal protein degradation (Figures 2D, 3B), we immunostained LRP1 and the lysosomal marker LAMP1 to determine LRP1 lysosomal localization. Our findings confirmed that LRP1 co-localizes with LAMP1 in the 3D-reconstruction we obtained using confocal microscopy (Figure 4B and Supplementary Figure 4). In addition, HepG2 cells were transfected with a functional LRP1 construct (Mini LRP-IV-EGFP; Perrot et al., 2012). Transfected HepG2 cells were then treated with 30 μM of RGZ for 22 h and we performed time-lapse imaging to monitor this fluorescent construct in the presence of lysotracker over a 1 h period. Lysotracker is a fluorescent dye used to label and to monitor acidic cellular compartments, including lysosomes. Our findings demonstrated that the mini-LRP-IV-EGFP construct co-localized with lysotracker to a greater extent in RGZ-treated cells than in the control. Moreover, the fluorescence intensity was constant in the control, while RGZ-treated cells displayed a reduced fluorescence intensity (Figure 4C). This suggested that RGZ induces lysosomal degradation of both endogenous LRP1 protein and ectopically overexpressed fluorescent Mini-LRP-IV-EGFP.

RGZ (30 μM) was demonstrated to reduce both LRP1 mRNA and protein levels, by the mechanisms of downregulation of PPARγ and protein lysosomal degradation, respectively (Figures 1–3). Given that the primary role of LRP1 is to endocytose its ligands at the cell surface (Perrot et al., 2012; Kanekiyo et al., 2013; Ramanathan et al., 2015) we evaluated LRP1 plasma membrane (PM) levels in cells treated with different concentrations of RGZ. Our results demonstrate that 30 μM RGZ reduced LRP1 PM levels, whereas 3 and 10 μM did not affect LRP1 PM levels. In addition, LDLR PM levels were not affected by RGZ, suggesting the specificity of RGZ to regulate LRP1 levels (Figure 5A).

The fluorescent peptide Aβ_1-42 (FAM-Aβ_1-42) is one of the hallmarks of Alzheimer’s disease patients and this peptide was previously demonstrated to have a high affinity for LRP1 (Deane et al., 2004; Kanekiyo et al., 2013) Moreover, hepatic cells have an important role in the clearance and uptake of systemic Aβ via LRP1 endocytosis (Tamaki et al., 2006, 2007; Sagare et al., 2011). Thus, we tested whether the LRP1 reduction caused by 30 μM RGZ affected FAM-Aβ_1-42 uptake in HepG2 cells. Our findings demonstrated that following 24 and 48 h incubations with RGZ, only the 2 h exposure of FAM-Aβ_1-42, following the 48 h incubation, resulted in a significant difference in FAM-Aβ_1-42 uptake compared to control (p < 0.001; Figure 5B). This impaired uptake correlated with lower LRP1 levels in the whole cell lysates and PM, presented in Figures 1, 2, and 5A.