Circulating Lipoprotein Concentrations Correlate with Total but Not Free Heme in Different Sickle Cell Disease Genotypes

In chronic hemolytic states, hemoglobin is released in the bloodstream and free heme is formed upon hemoglobin oxidation. Free heme has been shown to trigger vaso-occlusion in animal models of sickle cell disease (SCD) by activating endothelial cells through toll-like receptor 4 (TLR4). Serum levels of glycoproteins specialized in clearing hemoglobin and heme, such as haptoglobin and hemopexin, are typically depleted in SCD plasma. High-density (HDL) and low-density lipoproteins (LDL) can also bind heme and work as the first line of defense against heme toxicity. Heme can react with LDL and generate oxidized LDL (ox-LDL), another TLR4 ligand.We aimed to investigate the importance of lipoproteins as heme scavengers by comparing SCD patients with homozygous sickle cell anemia (SS) patients and double heterozygous hemoglobin SC disease (SC) patients, who present with higher and lower hemolytic rates, respectively.

Eighty-three patients (aged 18-68 years old, 44 female) participated in this study: 43 SS (12 receiving hydroxyurea - HU) and 40 SC. Exclusion criteria were pregnancy, blood transfusion, pain crisis, or use of lipid lowering drugs in the past 3 months. Peripheral venous blood samples were collected in tubes with EDTA for complete blood counts and without antiocoagulant for serum separation after centrifugation at 500g for 20 min. Serum lactate dehydrogenase, total cholesterol and HDL were measured by conventional enzymatic-colorimetric methods on a Roche Hitachi Modular automation system (Roche, Germany). LDL was calculated by Friedewald equation. Serum oxidized LDL (ox-LDL), total and free heme were determined with commercially available ELISA assays (Mercodia Oxidized LDL ELISA, Heme QuantiChrom Assay, Bioassay Systems, USA). Statistical analysis was performed with SPSS v.22.0.

LDH and reticulocyte count were significantly higher in SS patients than SC, and hemoglobin was significantly lower (P=0.001). Total heme was significantly higher in SS patients taking HU (SSHU) (90 ± 59 mM) compared with SC patients (66 ± 18 mM), P=0.01. Total cholesterol was higher in SC than in SS patients, 141 ± 28 vs. 107 ± 20 mg/dL, respectively, P =0.01). We found no significant differences between groups regarding LDL (P=0.08), HDL (P=0.10), ox-LDL (P=0.52). Total heme correlated negatively with total cholesterol (r=-0.33, P=0.037) and HDL (r=-0.49, P=0.001) in the whole population, regardless of genotype or treatment with HU. Surprisingly, free heme did not differ between SS, SSHU, or SC groups of patients (P=0.19), and did not correlate with cholesterol levels.

The negative correlation between total heme and total cholesterol supports that lipoproteins contribute to defend the body against heme toxicity. Despite HDL, LDL, and ox-LDL not being significantly different among the groups, the stronger negative correlation between total heme and HDL suggests that this cholesterol fraction may be more important in heme scavenging. We hypothesize that more intense hemolysis may lead to the production of heme capable of oxidizing lipoproteins, which can then be rapidly cleared by macrophages, explaining hypocholesterolemia in SS patients. Similar concentrations of free heme and ox-LDL in all groups suggests that, in steady state, their concentrations are kept relatively constant regardless of hemolytic rate. Since circulating concentrations of these TLR4 ligands did not differ between SC and SS patients in our study, this might indicate that the contribution of hemolysis to endothelial dysfunction and to differences in severity and in the variety of complications between these genotypes may be better explained by TLR4 activation by bound heme rather than by activation by free heme or ox-LDL.

Financial support: FAPESP/CAPES/CNPq