Synthesis and pharmacological evaluation of pomalidomide derivatives useful for sickle cell disease treatment

Abstract

Fetal hemoglobin (HbF) induction constitutes a valuable and validated approach to treat the symptoms of sickle cell disease (SCD). Here, we synthesized pomalidomide-nitric oxide (NO) donor derivatives (3a-f) and evaluated their suitability as novel HbF inducers. All compounds demonstrated different capacities of releasing NO, ranging 0.3-30.3%. Compound 3d was the most effective HbF inducer for CD34⁺ cells, exhibiting an effect similar to that of hydroxyurea. We investigated the mode of action of compound 3d for HbF induction by studying the in vitro alterations in the levels of transcription factors (BCL11A, IKAROS, and LRF), inhibition of histone deacetylase enzymes (HDAC-1 and HDAC-2), and measurement of cGMP levels. Additionally, compound 3d exhibited a potent anti-inflammatory effect similar to that of pomalidomide by reducing the TNF-α levels in human mononuclear cells treated with lipopolysaccharides up to 58.6%. Chemical hydrolysis studies revealed that compound 3d was stable at pH 7.4 up to 24 h. These results suggest that compound 3d is a novel HbF inducer prototype with the potential to treat SCD symptoms.

Keywords: Epigenetics; Fetal hemoglobin inducers; NO-donors; Nitric oxide; Pomalidomide; Sickle Cell Disease.

Figure 1.

Drug design of pomalidomide–NO donor compounds (3a–f).

Figure 2.

Expression of γ-globin in CD34⁺ cells. (A) CD34⁺ cells were treated for 14 days with different concentrations of pomalidomide and compounds 3b, 3d, and 3f. (B) CD34⁺ cells were treated for 14 days with 1μM pomalidomide (POM), 10 μM hydroxyurea (HU), and 2.5 μM compound 3d. DMSO (0.1%) was used as a negative control. Cell lysates were analyzed by western blot using γ-, β-, and α-globin antibodies. The mean relative γ-globin and β-globin expression levels were normalized to that of GAPDH by analyzing the band density using ImageJ software. The error bars represent the SD from three individual experiments. *p < 0.05 compared with DMSO (ANOVA followed by Dunnet test).

Figure 3.

Quantification of hemoglobin in CD34⁺ cells. Hemoglobin production was assessed in supernatants from H₂O-lysed cells, using high-performance liquid chromatography (HPLC). Blotplots from DMSO, pomalidomide, HU, and 3d-treated cultures. The area under the curve (AUC) was calculated for each sample, and relative hemoglobin composition was expressed as the mean percent (A) HbF/(HbF + HbA) ±SEM (N = 3), *p < 0.05 and (B) the mean percent HbA/(HbA + HbF) ±SEM, *p < 0.05 compared with DMSO (ANOVA followed by Dunnet test).

Figure 4.

Growth curves (number of cells) for DMSO-, HU-, pomalidomide (POM)-, and 3d-treated cultures at days 4, 7, 11, and 14. The data are shown as mean cells/mL ±SEM (N = 3). *p < 0.05 compared with DMSO (ANOVA followed by Tukey test).

Figure 5.

Flow cytometric characterization of erythroid precursors and terminal erythroblast differentiation. Terminal differentiation was monitored via α4-integrin and band 3 levels of GPA^pos cells on indicated days. Cells with α4-integrin^hi/band 3 ^lo represent less mature erythroblasts, whereas those with α4-integrin^lo/band3 ^hi are further differentiated (N = 3).

Figure 6.

(A) Western blot for BCL11A and IKAROS in CD34⁺ cells treated with 0.1% DMSO (negative control), 1 μM POM, 10 μM HU, and 2.5 μM compound 3d. Cell lysates were analyzed on day 4 of the treatment. The mean relative BCL11A and IKAROS expression levels were normalized to that of GAPDH by analyzing the band density using the ImageJ software. The error bars represent the SD from three individual experiments, *p < 0.05 compared with DMSO (ANOVA followed by Dunnet test). (B) Western blot for LRF in CD34⁺ cells treated with 0.1% DMSO, 1 μM POM, 10 μM HU, and 2.51 μM 3d at day 4 of culture. The mean relative LRF expression level was normalized to that of GAPDH by analyzing the band density using the ImageJ software. The error bars represent the SD from three individual experiments, *p < 0.05 compared with DMSO (ANOVA followed by Dunnet test).

Figure 7.

Effects 0.1% DMSO, 1 μM POM, 10 μM HU, and 2.5 μM compound 3d on intracellular cGMP levels in HUVEC cells. HUVEC cells were used at the confluence, and the results were measured after 48 h of treatment. Values represent means ±SEM (n = 3). *p < 0.05

Figure 8.

Level of the proinflammatory cytokine, TNF-α, determined using ELISA in the supernatant of human mononuclear culture treated with LPS and co-incubated with test drugs. Pomalidomide (POM) was used as positive anti-inflammatory controls. LPS at 50 mg/mL was used as an inflammatory stimulus. Data are represented as the mean ± SEM; p < 0.05 compared with lipopolysaccharide (LPS) (ANOVA followed by Tukey’s test); N = 3 experiments; C(−) : DMSO.

Figure 9.

In vitro determination of chemical stability for compound 3d in buffers at pH 1.0 and 7.4. Data are represented as means ±SEMs and expressed as %.

Scheme 1.

Synthesis of pomalidomide derivatives with NO-donor properties (3a–f). Reagents and conditions: a) acetone, Pd/C, H₂; b) aminobenzohydrazides, ethanol, H⁺, room temperature, 4–24 h; c) 3-aminophthalic anhydride (5), EDC, DMAP, DMF anhydrous, N₂ atmosphere, room temperature, 24–30 h.