Hypercholesterolemia and reduced HDL-C promote hematopoietic stem cell proliferation and monocytosis: studies in mice and FH children

Abstract

Previous studies have shown that mice with defects in cellular cholesterol efflux show hematopoietic stem cell (HSPC) and myeloid proliferation, contributing to atherogenesis. We hypothesized that the combination of hypercholesterolemia and defective cholesterol efflux would promote HSPC expansion and leukocytosis more prominently than either alone. We crossed Ldlr(-/-) with Apoa1(-/-) mice and found that compared to Ldlr(-/-) mice, Ldlr(-/-)/Apoa1(+/-) mice, with similar LDL-cholesterol levels but reduced HDL-cholesterol (HDL-C) levels, had expansion of HSPCs, monocytosis and neutrophilia. Ex vivo studies showed that HSPCs expressed high levels of Ldlr, Scarb1 (Srb1), and Lrp1 and were able to take up both native and oxidized LDL. Native LDL directly stimulated HSPC proliferation, while co-incubation with reconstituted HDL attenuated this effect. We also assessed the impact of HDL-C levels on monocytes in children with familial hypercholesterolemia (FH) (n = 49) and found that subjects with the lowest level of HDL-C, had increased monocyte counts compared to the mid and higher HDL-C levels. Overall, HDL-C was inversely correlated with the monocyte count. These data suggest that in mice, a balance of cholesterol uptake and efflux mechanisms may be one factor in driving HSPC proliferation and monocytosis. Higher monocyte counts in children with FH and low HDL-cholesterol suggest a similar pattern in humans.

Trial registration: ClinicalTrials.gov NCT01078675.

Copyright © 2013 Elsevier Ireland Ltd. All rights reserved.

Figures

Figure 1

LdLr−/−/ Apoa1+/− mice on WTD develop monocytosis, with increases in total, Ly-6 Chi and Ly-6 Clo monocytes, and neutrophilia in the blood. 8-week-old mice were placed on a WTD for 6 weeks and blood was then collected and stained for indicated cell types. (A) Total monocytes, monocyte subsets and neutrophils were analyzed by flow cytometry as seen by representative dot plots with group percent means and (B) expressed as absolute cell counts calculated from percentages applied to total leukocyte count determined by automated cell count. *P<0.05 vs. all other genotypes and ^P<0.05 vs. Apoa1−/−, Ldlr+/−/ Apoa1−/−. Data are mean with SEM, n=10–13. (C) HDL to total cholesterol ratio (ug/mL) is plotted with associated monocyte percentage for all mice in the study.

Figure 2

Ldlr−/−/ Apoa1+/− mice on a WTD have expansion of HSPCs in the bone marrow and increased proliferation in vivo, as well as a trend towards increased atherosclerosis. (A&B) BM cells were isolated and analyzed by flow cytometry to identify percent HSPCs. (C) Prior to sacrifice, mice were injected with EdU. In vivo proliferation was determined by EdU incorporation via flow cytometry. *P<0.05 vs. all other genotypes. (D) Hearts were placed in paraffin, stained with H&E and plaque area was determined by averaging plaque at aortic root over serial sections for each animal. P=0.1. Data are mean with SEM, n=10–13.

Figure 3

WT and LdLr−/− BM HSPCs take up nLDL and oxLDL. (A&B) HSPCs were isolated from BM by FACS, cDNA was prepared and mRNA expression was quantified by real-time PCR. Data is mean with SEM, n=3–5. (C&D) Cultured HSPCs were loaded with increasing concentrations of BODIPY-nLDL alone and in presence of excess unlabeled nLDL (200μg/mL) in WT and LdLr−/− BM cells. Uptake was quantified by flow cytometry and data points are averages of mean fluorescence intensity (MFI) of BODIPY in HSPCs. (E) HSPCs were cultured in media with LPDS serum, IL-3 and GM-CSF in the presence of either BODIPY labeled nLDL or DiI labeled oxLDL. Uptake was determined by flow cytometry at 6 and 12 hours.

Figure 4

LDL directly stimulates HSPC proliferation, which can be reversed by reconstituted HDL. (A) BM from Ldlr−/− mice was cultured and loaded overnight with increasing nLDL concentration (μg/mL). In vitro cell cycling in HSPCs was determined by FACS using DAPI and is expressed by the percent of HSPCs in SG2M phase. Ldlr−/− BM cells co-incubated with nLDL (50μg/mL) and increasing concentrations of rHDL. HSPCs were then sorted by FACS and the average total number of cells per well are shown (B). rHDL concentrations are reflective of protein. ^P<0.05 vs LDL alone.

Figure 5

Children with familial hypercholesterolemia and low HDL have increased circulating blood monocytes. (A) Children with homozygous FH were grouped based on plasma HDL concentration. Low HDL was defined as ≤ 40mg/dL, mid HDL as 40–50mg/dL and high HDL as ≤ 50mg/dL. Percent monocyte was measured by automated blood count. Data is mean with SEM, n=49. *P <0.01 vs low HDL group. (B) HDL plotted with associated monocyte count for all individuals in cohort.