Featured Article: Depletion of HDL3 high density lipoprotein and altered functionality of HDL2 in blood from sickle cell patients

Abstract

In sickle cell disease (SCD), alterations of cholesterol metabolism is in part related to abnormal levels and activity of plasma proteins such as lecithin cholesterol acyltransferase (LCAT), and apolipoprotein A-I (ApoA-I). In addition, the size distribution of ApoA-I high density lipoproteins (HDL) differs from normal blood. The ratio of the amount of HDL₂ particle relative to the smaller higher density pre-β HDL (HDL₃) particle was shifted toward HDL₂. This lipoprotein imbalance is exacerbated during acute vaso-occlusive episodes (VOE) as the relative levels of HDL₃ decrease. HDL₃ deficiency in SCD plasma was found to relate to a slower ApoA-I exchange rate, which suggests an impaired ABCA1-mediated cholesterol efflux in SCD. HDL₂ isolated from SCD plasma displayed an antioxidant capacity normally associated with HDL₃, providing evidence for a change in function of HDL₂ in SCD as compared to HDL₂ in normal plasma. Although SCD plasma is depleted in HDL₃, this altered capacity of HDL₂ could account for the lack of difference in pro-inflammatory HDL levels in SCD as compared to normal. Exposure of human umbilical vein endothelial cells to HDL₂ isolated from SCD plasma resulted in higher mRNA levels of the acute phase protein long pentraxin 3 (PTX3) as compared to incubation with HDL₂ from control plasma. Addition of the heme-scavenger hemopexin protein prevented increased expression of PTX3 in sickle HDL₂-treated cells. These findings suggest that ApoA-I lipoprotein composition and functions are altered in SCD plasma, and that whole blood transfusion may be considered as a blood replacement therapy in SCD. Impact statement Our study adds to the growing evidence that the dysfunctional red blood cell (RBC) in sickle cell disease (SCD) affects the plasma environment, which contributes significantly in the vasculopathy that defines the disease. Remodeling of anti-inflammatory high density lipoprotein (HDL) to pro-inflammatory entities can occur during the acute phase response. SCD plasma is depleted of the pre-β particle (HDL₃), which is essential for stimulation of reverse cholesterol from macrophages, and the function of the larger HDL₂ particle is altered. These dysfunctions are exacerbated during vaso-occlusive episodes. Interaction of lipoproteins with endothelium increases formation of inflammatory mediators, a process counteracted by the heme-scavenger hemopexin. This links hemolysis to lipoprotein-mediated inflammation in SCD, and hemopexin treatment could be considered. The use of RBC concentrates in transfusion therapy of SCD patients underestimates the importance of the dysfunctional plasma compartment, and transfusion of whole blood or plasma may be warranted.

Keywords: Inflammation; acute phase proteins; hemopexin; lipoproteins; sickle disease.

Figure 1

Non-denaturing gel electrophoretic lipoproteins separation. Twenty microliters of ApoB-depleted plasma mixed with 20 µL of saccharose loading buffer containing 5 mg/mL of the lipid stain Sudan Black were loaded in each lane of a polyacrylamide 2.5–18% gradient gel. Samples were run in a Tris–glycine–saccharose system (see the Materials and methods section) for 18 h at 4℃ at 60 V. (a) A representative image of a gel with two normal plasma samples, one SCD plasma sample of a patient at baseline, and two samples of patients at day 1 of a VOE onset. The position of the High Molecular Weight Calibration kit for electrophoresis (GE Healthcare) bands is shown on the right. (b) Quantification of the amount of HDL₂ relative to HDL₃, in plasma samples of eight normal plasmas, 15 SCD plasmas at baseline and 9 SCD plasmas at day 1 of a VOE are shown. (c) The HDL₂/HDL₃ ratio was analyzed for five of the sickle patients at baseline (open circles) and the matching sample during VOE (filled circles). Two-tailed P values were calculated with GraphPadPrism software and are indicated on the panels

Figure 2

HDL distribution and HDL-ApoA-I exchange of SCD plasma. Five hundred microliters of fresh washed RBC were mixed with 900 µL of normal or SCD plasma and incubated at 37℃. At the indicated time, 420 µL of the mixture was transferred to a new tube and centrifuged at low speed to remove the RBC. The plasma (300 µL) was collected and stored at −80℃ until analysis. Quantification of the relative amount of HDL₂/HDL₃ was performed as described in Figure 1 and is shown on the left axis. The EPR nitroxide signals, relative to an instrument reference peak, obtained by addition of lipid-free ApoA-I-spin labeled probe mixed with an increasing volume of ApoB-depleted plasma (0.6 µL to 14.4 µL) were performed in triplicate as previously described. The plasma unquenching capacity (PUC) of the samples was calculated with GraphPadPrism software and is shown on the right axis. The results are presented as relative to the value obtained with the plasma sample mixed with RBC (or saline) without incubation (t = 0 min)

Figure 3

Dichlorofluorescein assay. Results obtained with the di-acetate esterified reduced form of DCFH (DCFHDA) and of the di-acetate esterified oxidized form of DCF (DCFDA) are shown on panels (a) and (b), respectively. A stock solution of DCFHDA was prepared in methanol as described in the Materials and methods section and was either diluted in water (DCFHDA/MeOH; black symbols) or treated with NaOH (0.01 N) for 30 min (DCFHDA/NaOH; green symbols). The chemical was then added at a final concentration of 1 µM in 100 µL Tris-HCl 40 mM pH 7.4. Real-time fluorescence measurements were performed in 96-well plates at 37℃. Assays in panel (a) were performed in buffer alone (asterisk), buffer with 20 µM CuCl₂ (triangle) or with 50 µg/mL POVPC (circle). Assays in panel (b) were performed with plasma (0.4 µg cHDL) (open circle), plasma with 5 µM CuCl₂ (filled circle) or buffer with 5 µM CuCl₂ (asterisk). (A color version of this figure is available in the online journal.)

Figure 4

Reducing potential of plasma. Measurements were performed with DCFH in 100 µL Tris-HCl 40 mM pH 7.4 at 37℃. (a) Reactions were performed with three plasma samples (A, B, C) and with plasma sample (A) in presence of 5 µM CuCl₂ or 100 µM H₂O₂. One assay was performed with twice the amount of sample (indicated by 2×). (b) Reactions were performed with SCD (n = 31) and normal (n = 12) plasma samples in the presence of 5 µM CuCl₂. The slopes of the linear fit of DCF fluorescence values as a function of time were calculated with GraphPadPrism software. Data are presented relative to the values obtained in buffer in the absence of plasma in panel A and relative to the values obtained with 5 µM CuCl₂ in the absence of plasma in panel B. Error bars represent the standard deviations of three measurements. The median of the values for normal (0.1295) and SCD (0.1218) is indicated by a red line. (A color version of this figure is available in the online journal.)

Figure 5

Anti-oxidant potential of HDL₂ and HDL₃. HDL₂ and HDL₃ were isolated from SCD and control plasma and assayed for their ability to prevent (<1) or enhance (>1) oxidation of DCFH. Measurements were performed in triplicate as described in Figure 4 in the presence of CuCl₂ with HDL particles at a final concentration of 0.05–0.1–0.2 µg/mL (panels a, b and c). In panel (d), the mean (±SD) of the values obtained in each of the four groups (sickle HDL₂, sickle HDL₃, normal HDL₂, normal HDL₃) were calculated and plotted as a function of the HDL concentration. Note that at the highest concentration tested of 0.2 µg/mL, HDL2 isolated from control plasma was neither anti- or pro-oxidant (ratio of 1) but that sickle HDL2 was as effective as the control HDL3 particles in preventing DCFH oxidation (<1)

Figure 6

PTX3 mRNA quantification in HDL-treated HUVEC. Human umbilical vein endothelial cells were grown in complete medium to 70% confluence, and then cultured for 3 h in serum-poor medium prior co-culture with HDL diluted in serum-poor medium at the indicated concentration (see the Materials and methods section). HDL particles were isolated from SCD and normal plasma as shown in Figure 5. Total RNAs of treated cells were isolated and expression of PTX3 mRNA was quantified. Fold change of expression is presented as the ratio of the level of values obtained in the presence of HDL relative to those obtained in their absence. The bar graph in panel A shows the mean of the fold change values obtained with the samples in the normal (n = 5) or sickle (n = 8) cohort at different HDL₂ concentration (30–50–80–200 µg/mL). Error bars represent the standard deviation of the mean. Panels (b) and (c) show the fold change of expression of each sample obtained after treatment with HDL₃ and HDL₂ particles isolated from normal and SCD plasma at a concentration of 30 µg/mL and 200 µg/mL, respectively. The means and error bars, representing the min-max in each treatment group, are also shown. Two-tailed P values were calculated with GraphPadPrism software

Figure 7

Hemopexin interference with HDL function. (a) PTX3 mRNA expression was quantified as described in the legend of Figure 6 in HUVEC treated with HDL₂ particles isolated from normal or SCD plasma at a final concentration of 30 µg/mL in the absence or presence of 12.5 µM Hx. The effect of Hx on the reducing potential of plasma (panel B) and of HDL₂ and HDL₃ (panel C) was performed by monitoring the oxidation of DCFH in the presence of 20 µM CuCl₂ as described in the legend of Figure 4. Measurements were triplicated and performed with ApoB-depleted plasma (0.4 µg cHDL) (panel b) or with 0.2 µg/mL of HDL (panel c)

Figure 8

Working model. The high level of intravascular hemolysis in SCD leads to increased cell-free hemoglobin (CfHb) and consequently an increase of heme released from hemoglobin. Both haptoglobin (Hp) and hemopexin (Hx) are taxed to remove these redox active agents. The steady-state concentration of these compounds is found to be low in sickle cell plasma. As a result, heme bound to HDL is not efficiently Hx bound and initiates inflammatory reactions, including the formation of PTX3