LXR activation restores cholesterol homeostasis-related gene expression and induces cholesterol efflux from BPDCN. (A) BPDCN cell lines (CAL-1, GEN2.2) and blood samples from 4 patients diagnosed with BPDCN (LPDC #4, #7, #8, #10) were treated with 1 µM LXR agonists, T0901317 (T09) or GW3965 (GW), for 24 hours. ABCA1 and ABCG1 mRNA levels were determined by qRT-PCR in CAL-1 (n = 8), GEN2.2 (n = 5), and in blood samples containing >75% of leukemic PDCs (n = 4). Levels of mRNA were normalized to those of GAPDH for each sample and then expressed as fold change relative to the average value for vehicle-treated cells (*P < .05, **P < .01, ****P < .0001, Mann-Whitney). (B) CAL-1 cells were treated with 1 µM T0901317 (T09) or GW3965 (GW), or vehicle control alone, for 24 hours. LXRα isoform and ABCA1 protein expression were evaluated by western blot. The human hepatocellular carcinoma cell line HepG2 was used as control and reference for LXRα and ABCA1 expression, whereas untreated CAL-1 lysates were used as reference for actin expression. Expression of LXR protein was compared with actin expression with the vehicle condition being considered arbitrary as 1. Results of 1 experiment of 3 are shown. (C) Primary BPDCN cells from 1 patient (LPDC #4) were treated with 1 µM LXR agonists T0901317 (T09) or GW3965 (GW) for 24 hours. Expression of CD123 (as a BPDCN-specific marker) and ABCA1 was assessed by confocal microscopy. Nuclei were stained with DAPI. (D) This cartoon, adapted from Oram and Vaughan,⁸  represents 1 of the current accepted models of cholesterol efflux, illustrating cholesterol efflux experiments performed thereafter. Cholesterol is excluded from cells through ABCA1 and/or ABCG1 transporters. Lipid-poor APOA1 accepts cholesterol (yellow symbols) from cells through ABCA1-mediated cholesterol efflux. This induces nascent HDL formation which then accepts supplemental cholesterol loading via ABCG1-mediated efflux. Addition of HDL2 implicates only ABCG1-mediated cholesterol efflux. (E) Cholesterol efflux was assessed using [³H]-cholesterol-acetylated LDL-loaded CAL-1 cells treated with 1 µM T0901317 (T09) or GW3965 (GW) for 24 hours. Cholesterol efflux was triggered by the addition of 20 µg/mL HDL2 (right panel, n = 4) or 10 µg/mL APOA1 (left panel, n = 3). Data were expressed as percentage of cholesterol efflux (mean ± SEM of n experiments), as described in supplemental Methods (*P < .05, Mann-Whitney). (F) Cholesterol content of BPDCN cells was assessed after treatment with T0901317 (T09) or GW3965 (1 µM) for 24 hours followed by a 4-hour incubation with APOA1 cholesterol acceptor (10 µg/mL). Cellular cholesterol content was determined using filipin staining analyzed by confocal microscopy. One representative experiment of 3 for CAL-1 cells, 1 of 2 for GEN2.2 cells, is shown. LPDC represents data of a blood sample from 1 BPDCN patient tested of 4 (LPDC #5, #6, #8, #9). The PDC marker CD123 allows the identification of leukemic PDCs in blood samples. Cumulative filipin fluorescence intensity from the 4 different BPDCN samples was expressed as mean ± SEM (bottom right panel, *P < .05, Mann-Whitney). Fluorescence intensity of the vehicle condition is considered arbitrary as 1.