Anti-Inflammatory Effects of Hydroxyurea in a Murine Model of Chronic Intravascular Hemolysis

Intravascular hemolysis is associated with a number of disorders, including sickle cell anemia (SCA). In addition to diminishing red blood cell (RBC) counts, clinical manifestations associated with intravascular hemolysis include leg ulcers, pulmonary hypertension and priapism. Recent studies indicate that intravascular hemolysis propagates the inflammatory processes associated with many hemolytic diseases. This study aimed to establish a murine model of chronic hemolysis (CH) that elevates circulating cell-free heme to concentrations similar to those seen in SCA and observe the effects of CH on inflammatory parameters. We also evaluated the effects of concomitant administration of hydroxyurea (HU), a drug commonly used as therapy for SCA, on hemolytic inflammation. For induction of CH, C57BL/6 mice received twice weekly low concentration doses (10 mg/Kg/day) of phenylhydrazine (PHZ, i.p.) for four weeks (wks). Control mice received saline (CON) and some mice also received HU (50 mg/Kg/day, 5 times/wk for 4 wks, i.p.). Plasma cell-free hemoglobin (Hb) and heme were elevated to 0.4±0.1 g/L and 51.4±8.1 µM, respectively (N=4, P<0.05 compared to CON), after four wks of PHZ; concentrations similar to those observed in humans and mice with SCD (data not shown). Four wks of CH induced significant leukocytosis and reduced RBC counts (WBC; 6.1±0.8, 20.2±3.2 x10³/µL: RBC; 6.6±0.4, 4.8±0.3 x10⁶/µL for CON and CH, respect., P<0.01, N=4-5), and also significantly increased spleen weight in mice (0.33±0.01, 1.00±0.09 % of body weight for CON and CH, respect., P<0.0001, N=4-5). CH mice displayed substantial depletion of the Hb- and heme-binding proteins, haptoglobin (HP) and hemopexin (HPX) (HP; 129.9±5.2, 1.01±0.2 µg/ml: HPX; 1.98±0.22, 0.44±0.13 mg/ml for CON and CH, respect., P<0.0001, N=5), and increased plasma hemoxygenase-1 (HO-1) (17.4±1.7, 37.8±3.3 ng/ml, P<0.0001, N=5). With regard to inflammatory markers, four wks of CH significantly increased circulating levels of cytokines IL-1β and IL-10 (IL-1β; 3.77±0.53, 6.06±1.14 pg/mL: IL-10, 6.75±0.28, 8.46±0.87 pg/mL for CON and CH, respect., P<0. 05, N=4-5), but not TNF-α (0.70±0.17, 0.76±0.24 pg/mL for CON and CH, respect., N=4-5). Western blotting demonstrated that CH elevated hepatic protein expressions of CD163 (Hb-HP complex receptor), HO-1 and LRP1 (CD91, Heme-HPX complex membrane receptor), while depleting hepatic HPX (CD163; 0.3±0.02, 1.8±0.4 AU: LRP1; 0.4±.0.1, 1.2±0.4 AU: HO-1; 0.93±0.03, 2.66±1.01 AU: HPX; 0.7±0.2, 0.14±0.02 AU relative to GAPDH, for CON and CH, respect., N=3-8, P<0.05 for CD163, HO-1, HPX). Concomitant treatment of mice with 50 mg/Kg/day HU throughout hemolysis induction (CHHU) did not significantly alter hematological data, spleen weight or plasma markers of hemolysis, compared to CH (WBC, 20.6±1.91 x10³/µL: RBC, 4.6±0.1 x10⁶/µL: spleen weight; 1.01±0.06 % body weight: HP; 0.71±0.04 pg/mL: HPX; 0.34±0.64 mg/mL, respect., P>0.05 compared to CH, N=5), although a slight decrease in plasma HO-1 was observed (30.7 ± 2.2 ng/ml for CHHU, P>0.05). Notably, HU therapy significantly decreased plasma levels of cytokines IL-1β and IL-10, in mice undergoing CH (IL-1β; 2.22±0.24 pg/mL: IL-10; 5.81±0.29 pg/mL, N=5, P<0.01 compared to CH) and partially abolished the effects of CH on hepatic protein levels of HO-1 (1.39±0.23 AU, N=3) and HPX (0.4±0.08 AU, N= 7, P<0.05 compared to CH). In summary, using repeated low dose PHZ administrations, we successfully induced moderate and continuous hemolysis in mice, similar to that observed in SCA. This CH was associated with the depletion of hemoglobin and heme scavenger proteins, and with the production of pro- and anti-inflammatory cytokines, in addition to higher expression of scavenger receptors in the livers of these mice. Concomitant therapy of mice with HU, frequently used to induce fetal hemoglobin (HbF) in SCA patients, did not alter hematological data or markers of hemolysis, indicating that this drug does not inhibit hemolysis in our CH model. However, HU had significant anti-inflammatory effects, reducing markers of inflammation and elevating hepatic HPX protein expression. As such, HU has anti-inflammatory effects during long term use that are independent of its HbF-inducing properties, providing further evidence of the mechanism of action HU in SCA and further supporting its use in other hemolytic disorders.