The HDL particle composition determines its antitumor activity in pancreatic cancer

Figures

• 

  Figure 1.. Pancreatic-cancer cells are sensitive toward cholesterol depletion and rHDL particles. (A) Panc02 cells were grown in RPMI medium containing either 2% FCS, 2% LPDS, or 2% LPDS with the addition of 5 μM lovastatin and 100 μM mevalonate. Viability of cells was determined every 24 h using a WST1 viability assay (n = 4/4/4; *P < 0.05; one-way ANOVA). (B) Viability was determined of Panc02, 6066, and BxPC3 pancreatic-cancer cells after culture of cells for 72 h in either RPMI medium with 2% LPDS or cholesterol depleted medium (n = 4/4; *P < 0.05; unpaired t test). (C) Panc02 cells cultured in RPMI medium with increasing concentrations of FCS were treated for 48 h with PBS, high-density lipoprotein (HDL) isolated from human plasma of a healthy donor (75 μg/ml) or reconstituted HDL (75 μg/ml, molar ratio of PC:C:CE:APOA1 of 100:12.5:0:1, ZLB Behring), and viability was determined using a WST1 assay (n = 4/4/4; *P < 0.05; one-way ANOVA). (D) HDL particles were reconstituted using the indicated molar ratios of phosphatidylcholine (PC), cholesterol (C), cholesterol ester (CE), and apolipoprotein A-I (APOA1). (E) Electrophoretic mobility of HDL particles was tested on a 1% agarose gel. (F)…

• 

  Figure 2.. The high-density lipoprotein (HDL) particle composition differentially affects pancreatic-cancer cell growth characteristics. (A) After a 24-h starvation period, Panc02 and 6066 cells were cultured in RPMI medium with 2% FCS in the presence of PBS, native human HDL (nHDL), or the indicated rHDL particles (75 μg/ml) for 48 h. Afterward, viability was determined using a WST1 assay (n = 4 biological replicates per group; *P < 0.05; unpaired t test). (B) Representative flow cytometry blots of Panc02 cells treated with different HDL particles (75 μg/ml) stained for the detection of apoptosis using 7-AAD and Annexin V. (B, C, D, E) Quantification of live, early apoptotic, and late apoptotic cells from flow cytometry experiments shown in (B) (n = 3 biological replicates per group, *P < 0.05; one-way ANOVA). (F, G, H) Panc02 cells were starved for 8 h and afterward treated with indicated concentrations of different HDL particles for 16 h. qPCR experiments show relative mRNA levels of Hmgcr, Hmgcs, and the Ldlr. Data are expressed as fold change over PBS-treated control cells (n = 3 replicates per group; *P < 0.05; one-way ANOVA).

• 

  Figure 3.. SR-B1 influences cholesterol efflux and antitumor properties of rHDL particles. (A) 3H cholesterol-loaded Panc02 and 6066 cells were subjected to cholesterol efflux assays with indicated high-density lipoprotein (HDL) particles (10 μg/ml) for 8 h. Cholesterol efflux is shown as % of transferred tracer from cells to HDL particles compared with control conditions (n = 4; *P < 0.05; one-way ANOVA). (B) relative, particle-specific cholesterol efflux capacity was analyzed in the absence or the presence of BLT1 (n = 4; *P < 0.05; one-way ANOVA). (C) Schematic representation of tracer-labeled rHDL particles. (D, E) 14C cholesterol and 3H cholesteryl ester influx, respectively, from rHDL to Panc02 cells in the absence or presence of BLT1(n = 4; *P < 0.05; one-way ANOVA). (F) Cellular cholesterol content was measured by GC in Panc02 cells treated for 48 h with different HDL species (75 μg/ml) in the presence of 2% FCS and 5 μM lovastatin (n = 2, *P < 0.05; one-way ANOVA). (G, H) quantification of early and (H) late apoptotic cells upon treatment of Panc02 cells with indicated HDL particles in the absence or presence of BLT1 (n = 3; *P < 0.05; unpaired t test).

• 

  Figure S1.. mRNA expression analysis of important lipid/cholesterol transport proteins in Panc02 cells. Quantitative PCR analysis was performed to compare relative expression levels of Scarb1 (Sr-b1), Abca1, Abcg1, and Abcg4. Hprt was used as a housekeeping gene, and expression levels of target genes are compared with Scarb1 (n = 3).

• 

  Figure 4.. The LXR agonist TO901317 increases cholesterol efflux and apoptosis-inducing properties of small, lipid-poor rHDL. (A) TO901317 induces ABCA1 mRNA and protein levels in Panc02 cells (n = 3; *P < 0.05; one-way ANOVA). (B) Relative, particle-specific cholesterol efflux from Panc02 cells in the absence or presence of TO901317 (n = 4; *P < 0.05; unpaired t test). (C) 3H cholesterol accumulation in Panc02 cells in the presence of TO901317 after efflux to indicated high-density lipoprotein particles (n = 4; *P < 0.05; one-way ANOVA). (D) Relative apoptosis rates of TO901317-treated Panc02 cells induced by either rHDL1 or rHDL2 in the absence or presence of BLT1 (n = 3; P < 0.05; unpaired t test).

• 

  Figure S2.. Apoa1 KO mice show similar tumor growth rates compared with their wildtype littermates. WT or Apoa1 KO C57Bl6/J mice were inoculated with 0.5 × 106 syngeneic tumor cells either subcutaneously in the right flank (Panc02, B16F10, lung carcinoma cell) or orthotopically in the second mammary fat pad (E0771). (A) Tumor growth kinetics and tumor weight at experimental end stage of the Panc02 pancreatic-cancer model (n = 7/7). (B) Tumor growth kinetics and tumor weight at experimental end stage of the B16F10 melanoma model (n = 13/10 WT and Apoa1 KO mice, respectively). (C) Tumor growth kinetics and tumor weight at experimental end stage of the lung carcinoma cell lung cancer model (n = 9/8). (D) Tumor growth kinetics and tumor weight at experimental end stage of the E0771 breast cancer model (n = 10/13; *P < 0.05; unpaired t test for A, B, C, D).

• 

  Figure 5.. AAV-mediated APOA1 reconstitution/rHDL injection reduces Panc02 tumor growth in Apoa1 KO mice. (A) mRNA expression levels of APOA1 in liver tissue of Panc02-bearing WT, Apoa1 KO, and AAV-APOA1–reconstituted Apoa1 KO mice at experimental end stage (n = 8/6/7). (B) Western blot analysis of APOA1 protein levels in plasma of Apoa1 KO mice prior and post AAV-APOA1 injection. (C) High-density lipoprotein (HDL)-C levels in Panc02 tumor-bearing mice at the experimental end stage (n = 8/5/7; *P < 0.05; one-way ANOVA). (D, E) Panc02 tumor growth kinetics and (E) tumor weight at experimental end stage (n = 8/5/7; *P < 0.05; unpaired t test). (F) Panc02 tumor weight in WT mice either receiving AAV-Ctrl or AAV-APOA1 particles (n = 8/4, unpaired t test). (G) Panc02 tumor weight of WT, Apoa1 KO, and rHDL (0.2 mg per injection; PC:C:CE:APOA1 = 100:12.5:0:1, ZLB Behring)-injected Apoa1 KO mice at an experimental end stage (n = 7/7/7; *P < 0.05; unpaired t test). (H) Western blot analysis of Panc02 cell lysates upon treatment of cells with indicated HDL particles (75 μg/ml) for 18 h. (I, J) Quantification of band intensities of pAKT (Thr308) and pAKT (S473), respectively, normalized to β-actin.…

• 

  Figure S3.. AAV-mediated reconstitution of APOA1 does not change intratumoral hypoxia or immune cell infiltration. (A) Representative images showing hypoxic, pimonidazole+ areas (brown) of WT, Apoa1 KO, and AAV-APOA1–reconstituted Apoa1 KO mice. Images were acquired using a Zeiss Axio Scope using a 10x objective. Scale bar, 200 μm. (B) Quantification of hypoxic areas as a percentage of the total tumor area (n = 8/8/11; *P < 0.05; one-way ANOVA). (C, D, E) Representative FACS blots and quantification of frequencies of tumor-resident MRC1+ macrophages, granulocytic MDSC, CD8+ cytotoxic T cells, and CD4+ T-helper cells were determined using flow cytometry (n = 8/8/11; *P < 0.05; one-way ANOVA).

• 

  Figure 6.. Pancreatic-cancer patients show decreased plasma high-density lipoprotein-C and cholesterol efflux capacities compared with healthy donors and increased survival probability with lower tumor-associated low-density lipoprotein receptor and HMGCS expression levels. (A, B, C) Plasma total triglyceride levels, (B) total cholesterol levels, and (C) high-density lipoprotein-C were determined in plasma samples from a cohort of healthy volunteers and pancreatic-cancer patients in late stages of their disease (n = 19/19; *P < 0.05; unpaired t test). (D, E) Nonnormalized plasma efflux capacity and (E) plasma efflux capacity normalized to APOA1 levels (n = 19/19; *P < 0.05; unpaired t test). (F) APOA1-to-SAA1 plasma protein ratio after ELISA measurement of protein concentrations. (G) Cholesterol efflux in plasma samples of patients with low (<13 μg/ml) or high (>13 μg/ml) SAA1 plasma concentrations. (H, I) Kaplan–Meier overall survival plots for low-density lipoprotein receptor and HMGCS expression, respectively, in the GDA TCGA pancreatic-cancer cohort (log-rank test, *P < 0.05, UCSC Xena).

• 

  Figure S4.. Candidate genes that regulate intracellular cholesterol accumulation correlate with overall survival of cancer patients. (A, B) SCARB1 and (B) ABCA1 expression do not correlate with patient overall survival in the GCD TCGA pancreatic-cancer cohort. (C, D) Low-density lipoprotein receptor and (D) HMGCS expression inversely correlate with patient overall survival in the TCGA Pan-cancer cohort. (E) MYLIP gene expression positively correlates with overall survival in the TCGA Pan-cancer cohort (A, B, C, D, E, log-rank test, *P < 0.05, UCSC Xena). (F) Heat map analysis demonstrates an inverse correlation of low-density lipoprotein receptor and MYLIP expression patterns in the TCGA Pan-cancer cohort.