Persistent Effect of Advanced Glycated Albumin Driving Inflammation and Disturbances in Cholesterol Efflux in Macrophages

Abstract

Advanced glycated albumin (AGE-albumin) impairs cholesterol efflux and contributes to inflammation in macrophages. The current study evaluated: (1) the persistence of the deleterious effect of AGE-albumin in cholesterol efflux and in inflammation, and (2) how metabolic control in diabetes mellitus (DM) contributes to attenuate the deleterious role of AGE-albumin in macrophage cholesterol homeostasis. Methods: AGE-albumin was produced in vitro or isolated from uncontrolled DM subjects' serum before (bGC) and after improved glycemic control (aGC). Albumin samples were incubated with bone marrow-derived macrophages and ¹⁴C-cholesterol efflux or LPS- induced cytokine secretion were determined immediately, or after cell resting in culture media alone. The ABCA-1 degradation rate was determined after cell incubation with cycloheximide, and ABCA1 protein level by immunoblot. Oil Red O staining was used to assess intracellular lipid accumulation. Results: A persistent effect of AGE-albumin was observed in macrophages in terms of the secretion of inflammatory cytokines and reduced cholesterol efflux. HDL-mediated ¹⁴C-cholesterol efflux was at least two times higher in macrophages treated with aCG-albumin as compared to bGC-albumin, and intracellular lipid content was significantly reduced in aGC-albumin-treated cells. As compared to bGC-albumin, the ABCA-1 protein content in whole cell bulk was 94% higher in aCG-albumin. A 20% increased ABCA-1 decay rate was observed in macrophages treated with albumin from poorly controlled DM. AGE-albumin has a persistent deleterious effect on macrophage lipid homeostasis and inflammation. The reduction of AGEs in albumin ameliorates cholesterol efflux.

Keywords: ABCA-1; advanced glycated albumin; atherosclerosis; cholesterol efflux; diabetes mellitus; inflammation.

Figure 1

Persistence of the effect of GAD-albumin on ¹⁴C-cholesterol efflux impairment in macrophages. ¹⁴C-cholesterol and acetylated LDL overloaded BMDMs were incubated for 48 h with low-glucose DMEM containing 1 mg/mL medium of control (C-albumin) or glycolaldehyde-treated albumin (GAD-albumin). After this period, cells were washed with PBS/FAFA and immediately tested for cholesterol efflux mediated by HDL₂ (panels A and B) or HDL₃ (panels C and D), or after resting in DMEM/FAFA for different periods of time (0 to 12 h). Efflux experiments were performed for 6 h. In panels B and D, the efflux rate is expressed as ratio of GAD-albumin/C-albumin, considering three independent experiments. Comparisons were done by Student’s t test * p < 0.05 as compared to C-albumin (n = 3); # p < 0.05 as compared to other time periods.

Figure 2

Persistence of the effect of glycolaldehyde (GAD)-modified albumin on the secretion of TNF and IL-6 by LPS-challenged macrophages. Cholesterol-overloaded BMDMs were incubated with low-glucose DMEM containing C or GAD-albumin (1 mg/mL), for 48 h. After this period, cells were washed with PBS/FAFA and incubated with DMEM/FAFA alone for different periods of time. After treatment with lipopolysaccharide (LPS, 1 μg/mL), for 24 h, TNF and IL-6 concentration were determined in the medium by ELISA. (Panel A): TNF secretion after 2 h, 4 h, 6 h, and 8 h of cell resting in DMEM/FAFA prior to LPS challenge (absolute values); (panel B): area under the curve (AUC) of TNF secretion; (panel C): TNF secretion (absolute values) after 2 h, 12 h, and 24 h of cell resting in DMEM/FAFA prior to LPS challenge; (panel D): independent experiments (from panels A and C) were considered together and expressed as the Alb-GAD/Alb-C ratio for TNF secretion; (Panel E): IL-6 secretion (absolute values); (panel F): area under the curve (AUC) of the IL-6 secretion along time. Comparisons were done by Student’s t test. * p < 0.05 in comparison to C-albumin (n = 3); # p < 0.05 in comparison to other time periods.

Figure 2

Persistence of the effect of glycolaldehyde (GAD)-modified albumin on the secretion of TNF and IL-6 by LPS-challenged macrophages. Cholesterol-overloaded BMDMs were incubated with low-glucose DMEM containing C or GAD-albumin (1 mg/mL), for 48 h. After this period, cells were washed with PBS/FAFA and incubated with DMEM/FAFA alone for different periods of time. After treatment with lipopolysaccharide (LPS, 1 μg/mL), for 24 h, TNF and IL-6 concentration were determined in the medium by ELISA. (Panel A): TNF secretion after 2 h, 4 h, 6 h, and 8 h of cell resting in DMEM/FAFA prior to LPS challenge (absolute values); (panel B): area under the curve (AUC) of TNF secretion; (panel C): TNF secretion (absolute values) after 2 h, 12 h, and 24 h of cell resting in DMEM/FAFA prior to LPS challenge; (panel D): independent experiments (from panels A and C) were considered together and expressed as the Alb-GAD/Alb-C ratio for TNF secretion; (Panel E): IL-6 secretion (absolute values); (panel F): area under the curve (AUC) of the IL-6 secretion along time. Comparisons were done by Student’s t test. * p < 0.05 in comparison to C-albumin (n = 3); # p < 0.05 in comparison to other time periods.

Figure 3

¹⁴C-cholesterol efflux from macrophages treated with serum albumin isolated from subjects with DM before (bCG) and after improvement of glycemic control (aGC). BMDMs were overloaded with acetylated LDL and ¹⁴C-cholesterol. After washing and equilibration in DMEM/FAFA, BMDMs were treated with 1 mg/mL of serum albumin isolated from subjects with DM bCG and aGC, for 48 h. After rinsing, cells were incubated with HDL₂, for 6 h (panel A: subjects 1 to 5) or apoA-I for 8 h (panel C: subjects 1, 3 and 4). The % of cholesterol efflux was calculated as: ¹⁴C-cholesterol in medium/¹⁴C-cholesterol in medium + ¹⁴C-cholesterol in cell × 100. The specific efflux mediated by HDL or apoA-I was calculated by subtracting values of total efflux (incubations with DMEM/FAFA plus HDL₂ or apoA-I) from those obtained in incubations with DMEM/FAFA alone (basal efflux). In panels B and D, data are presented with all subjects together, respectively, for efflux mediated by HDL₂ and apoA-I. Comparisons were done by Student’s t test (mean ± SE; n = 4); * p < 0.05.

Figure 4

Intracellular lipid content in macrophages treated with serum albumin isolated from DM subjects before (bCG) and after improvement of glycemic control (aGC). BMDMs were overloaded with acetylated LDL and after washing treated with 1 mg/mL of serum albumin isolated from DM subjects bCG and aGC, for 18 h. Cells were rinsed and then incubated with apoA-I or HDL₂ (6 h) as cholesterol acceptors. Intracellular lipid content was determined after staining with Oil Red O (representative photographs for each experimental condition). (Panel A): subject 1; (Panel B): subject 3. Comparisons were done by Student’s t test (mean ± SE; n = 4); * p < 0.05.

Figure 5

Persistence of the effect of albumin isolated before (bGC) and after (aGC) improvement of glycemic in subjects with DM in the inflammatory response elicited by LPS in macrophages. Cholesterol-overloaded BMDMs were incubated with low-glucose DMEM containing albumin isolated from subjects with DM (subjects 2 and 6) bGC and aGC (1 mg/mL), for 48 h. After this period, cells were washed with PBS/FAFA and incubated with DMEM/FAFA alone for different periods of time. After treatment with lipopolysaccharide (LPS 1 μg/mL), for 24 h, TNF concentration was determined in the medium by ELISA. Data are expressed as percentage in relation to incubations with control albumin. (Panels A and C): TNF secretion (absolute values) in incubations with albumins (bCG and aGC) from subjects 2 and 6, respectively. (Panels B and D): area under the curve (AUC) for TNF secretion in incubations with albumins (bGC and aGC) from subjects 2 and 6, respectively. Comparisons were done by Student’s t test * p < 0.05 in comparison to bCG (n = 3).

Figure 6

ABCA-1 protein level in macrophages treated with albumin isolated before (bGC) and after (aGC) improvement of glycemic control. ABCA1 protein levels and representative immunoblot. After 18 h-incubation with 2 mg/mL albumin isolated from subjects 1, 3, and 6 bGC and aGC, J774 macrophages were scrapped and equal amounts of cell protein loaded into a 6% polyacrylamide gel. Immunoblot was performed using anti-ABCA-1 antibody. Data are presented as arbitrary units (AU) corrected by beta-actin. Samples derived from the same experiment were processed in same gel/blot. Comparisons were done by Student’s t test (n = 3, mean ± SE). * p < 0.05.

Figure 7

ABCA-1 decay rate in J774 macrophages treated with albumin isolated from poorly controlled T1DM and control subjects. J774 macrophages were incubated with T0901317 overnight and then exposed to a pool of serum albumin (2 mg/mL) isolated from non-DM control subjects (C; n = 5) or poorly controlled T1DM individuals (n = 4) in the presence of cycloheximide over time. Immunoblot was performed using anti ABCA-1 antibody and bands were corrected by beta-actin. Three independent experiments were performed, and a representative blot is shown in panel A. The ABCA-1 decay rate was calculated by the area under the curve (AUC) of ABCA-1 level (panel B). Comparisons were done by Student’s t test (n = 3, mean ± SE). * p < 0.05.