Very-low and low-density lipoproteins induce neutral lipid accumulation and impair migration in monocyte subsets

Abstract

Blood monocytes are heterogeneous effector cells of the innate immune system. In circulation these cells are constantly in contact with lipid-rich lipoproteins, yet this interaction is poorly characterised. Our aim was to examine the functional effect of hyperlipidaemia on blood monocytes. In the Ldlr(-/-) mouse monocytes rapidly accumulate cytoplasmic neutral lipid vesicles during hyperlipidaemia. Functional analysis in vivo revealed impaired monocyte chemotaxis towards peritonitis following high fat diet due to retention of monocytes in the greater omentum. In vitro assays using human monocytes confirmed neutral lipid vesicle accumulation after exposure to LDL or VLDL. Neutral lipid accumulation did not inhibit phagocytosis, endothelial adhesion, intravascular crawling and transmigration. However, lipid loading led to a migratory defect towards C5a and disruption of cytoskeletal rearrangement, including an inhibition of RHOA signaling. These data demonstrate distinct effects of hyperlipidaemia on the chemotaxis and cytoskeletal regulation of monocyte subpopulations. These data emphasise the functional consequences of blood monocyte lipid accumulation and reveal important implications for treating inflammation, infection and atherosclerosis in the context of dyslipidaemia.

Figure 1. Blood monocytes from Ldlr^−/− mice are lipid loaded.

Blood monocytes from Ldlr^−/− mice fed chow or high fat diet (HFD) for 8 weeks were analysed by flow cytometry. (A) A representative FACS dot plot example of SSC^high/SSC^low monocytes on chow and after HFD, gated from CD115^pos cells. (B) Quantification of GR1^high and GR1^low monocyte SSC with (grey bars) or without HFD (black bars). (C) Percentage of GR1^high/GR1^low monocytes that are SSC^high on chow/HFD. (D) Percentage of monocytes that are CD11c^pos on chow or HFD, gated from CD115^pos CD11b^pos cells. (E) Representative histogram of GR1^high or GR1^low monocyte CD11c expression, gated from CD115^pos CD11b^pos cells. (F) Percentage of SSC^high/SSC^low monocytes that are CD11c^pos on chow or HFD. (G) Neutral lipid staining of Ly6c^hi and Ly6c^low blood monocytes after 16 weeks HFD. Scale bar represents 10 μm. Staining is quantified as (H) vesicles per cell (20–25 cells per condition) and (I) LipidTOX vesicle median fluorescence intensity (MFI) (20–25 cells per subset) in HFD only. Error bars show the mean±SEM. *, ** and *** represents P < 0.05, P < 0.01 and P < 0.001 respectively analysed by Mann–Whitney U test. n = 3–4 mice per group. See also Sup. Fig. 1.

Figure 2. Sterile peritonitis model in dyslipidemic Ldlr^−/− mice.

Thioglycollate peritonitis (72 hours post-thioglycollate injection; Thio) or controls was induced in Ldlr^−/− mice (n = 4 per group) with or without 16 week high fat diet (HFD) and peritoneal lavage collected. (A) Monocytes and macrophages/ml gated as CD115+ CD11b+ (B) Inflammatory gene expression in peritoneal cells from Thio peritonitis (fold change in HFD versus chow). (C) Intravenous injection of 1 μm latex beads was used to track monocyte migration out of the blood during peritonitis. CD115⁺CD11b⁺Bead⁺ cells/ml in peritoneal lavage from Ldlr^−/− mice with or without HFD (n = 4 per group). (D) Cytoskeletal gene expression in peritoneal cells from Thio peritonitis (fold change in HFD versus chow). (E) Representative flow cytometry plots of CD115 and F4/80 staining in the omentum of Ldlr^−/− 72 hours post Thio, gated from all CD45+ cells. (n = 4–5 mice per group). (F-G) Leukocyte populations/gram of tissue from (E): (F) CD45⁺ leukocytes, (G) CD115+ F4/80^low monocytes. * and ** represents P < 0.05 and P < 0.01 respectively analysed by Mann–Whitney U test. See also Sup. Fig. 2.

Figure 3. Neutral lipid accumulation in LDL and VLDL treated monocytes.

Human CD16^pos and CD16^neg monocytes were purified and treated with LDL or VLDL (100 μg/ml) and intracellular neutral lipid was quantified using LipidTOX by confocal microscopy. (A) Percentage cells neutral lipid positive after 150 mins VLDL treatment (n = 10 fields). (B) Number of neutral lipid vesicles per cell over time after VDL treatment (n = 25–50 cells). (C,D) Same as (A,B) respectively with LDL treatment. (E) Representative examples of neutral lipid positive monocytes from LDL and VLDL treatments. Scale bar represents 20 μm. (F) Membrane cholesterol content in monocytes after 120 mins LDL or VLDL treatment, quantified by filipin staining intensity. Error bars show the mean ± SEM. *, ** and *** represents P < 0.05, P < 0.01 and P < 0.001 respectively analysed by Mann–Whitney U test. See also Sup. Fig. 3.

Figure 4. Effects of VLDL treatment on monocyte migration.

(A) Transwell migration of CD16^pos and CD16^neg monocytes to C5a (250 ng/ml) with or without VLDL pre-treatment (2 hrs; 100 μg/ml). Data represents 12 fields of view from 3 independent experiments. (B) Representative cells/field migrated shown in (A) (scale bar = 100 μm). (C) Representative track projections from CD16^pos and CD16^neg monocytes with or without VLDL pre-treatment (2 hrs; 100 ug/ml) in a 2D real-time chemotaxis assay towards a C5a gradient (n = 3 independent experiments). Tracks were analysed for (D) confinement ratio (E) displacement (μm) and (F) speed (μm/sec) (n = 120–800 cells) (G) CD16^pos and CD16^neg cell surface expression of CD88 after 2 hours LDL or VLDL treatment, assessed by flow cytometry (n = 2 donors in duplicate). (H) Adhesion of CD16^pos and CD16^neg monocytes to HUVECS from only monocytes pre-treated with VLDL (2 hrs; 100 μg/ml), normalised to CD16^neg PBS treated (n = 4 donors), (I) Same as (H) with only HUVECS treated with VLDL. (J) Transmigration of CD16^pos and CD16^neg monocytes pre-treated with (2hrs; 100 μg/ml) or without VLDL through TNF activated HUVEC: cells per field, normalised to CD16^neg PBS treated. (n = 4 donors) Error bars show the mean ± SEM. *** represents P < 0.001 analysed by Mann–Whitney U test. See also Sup. Fig. 4.

Figure 5. Intravital imaging of monocytes in chow or high fat diet (HFD) fed Cx3cr1^gfp/+ mice.

Cx3cr1^gfp/+ mice were fed chow or HFD for 6 weeks (A) Blood total and LDL cholesterol levels of chow vs HFD mice. (n = 4 per group) (B) Representative neutral lipid staining of Gr1^hi or Gr1^low monocytes from Cx3cr1^gfp/+ mice on HFD. Scale bar represents 10 μm. GFP^highGr1^low monocyte vascular crawling was then assessed by intravital microscopy. (n = 4 per group; n = 20–25 cells per condition) (C) Track straightness (D) Track length (μm) (E) Track duration (seconds) (F) Track speed (μm/sec) and (G) Track displacement (μm) length of intravascular monocytes during approximately 45 minutes of imaging. (H) Track projections of patrolling monocytes from Cx3cr1^gfp/+ mice fed HFD or chow during ~45 minutes intravital imaging. Error bars show the mean ± SEM. * and ** represents P < 0.05 and P < 0.01 respectively analysed by Mann–Whitney U test.

Figure 6. Cytoskeletal dynamics in VLDL treated human monocytes.

Human CD16^pos and CD16^neg monocytes were treated with or without VLDL (100 μg/ml; 2 hrs) and morphology analysed using phalloidin staining. (n = 20 cells per condition). (A) Cell area (um²) (B) circularity and (C) representative images of CD16^pos and CD16^neg monocytes pre-treated with PBS or VLDL. Scale bar = 10 μm. (D) Representative spatial localization of lipid-droplets in VLDL treated monocyte, assessed by brightfield microscopy. Red arrows indicate lipid vesicle. Scale bar = 10 μm (E) RHOA activation in CD16^pos and CD16^neg monocytes after treatment with PBS, VLDL or VLDL with or without CN03 (2 μg/ml). (n = 3 donors). Error bars show the mean ± SEM. * and ** represents P < 0.05 and P < 0.01 respectively analysed by Mann–Whitney U test. See also Sup. Fig 4.