Mitapivat reprograms the RBC metabolome and improves anemia in a mouse model of hereditary spherocytosis

Abstract

Hereditary spherocytosis (HS) is the most common, nonimmune, hereditary, chronic hemolytic anemia after hemoglobinopathies. The genetic defects in membrane function causing HS lead to perturbation of the RBC metabolome, with altered glycolysis. In mice genetically lacking protein 4.2 (4.2-/-; Epb42), a murine model of HS, we showed increased expression of pyruvate kinase (PK) isoforms in whole and fractioned RBCs in conjunction with abnormalities in the glycolytic pathway and in the glutathione (GSH) system. Mitapivat, a PK activator, metabolically reprogrammed 4.2-/- mouse RBCs with amelioration of glycolysis and the GSH cycle. This resulted in improved osmotic fragility, reduced phosphatidylserine positivity, amelioration of RBC cation content, reduction of Na/K/Cl cotransport and Na/H-exchange overactivation, and decrease in erythroid vesicles release in vitro. Mitapivat treatment significantly decreased erythrophagocytosis and beneficially affected iron homeostasis. In mild-to-moderate HS, the beneficial effect of splenectomy is still controversial. Here, we showed that splenectomy improves anemia in 4.2-/- mice and that mitapivat is noninferior to splenectomy. An additional benefit of mitapivat treatment was lower expression of markers of inflammatory vasculopathy in 4.2-/- mice with or without splenectomy, indicating a multisystemic action of mitapivat. These findings support the notion that mitapivat treatment should be considered for symptomatic HS.

Keywords: Genetic diseases; Hematology; Mouse models; Therapeutics.

Figure 1. RBCs from 4.2^–/– mice are characterized by an abnormal metabolomic profile.

(A) IB analysis using specific Abs against Pklr and Pkm2 in unfractionated RBCs (whole RBCs) and fractionated RBCs, according to density in F1(density N1.074, corresponding to a young and reticulocyte-enriched fraction) and F2 (density N1.092, corresponding to older RBCs) from WT and 4.2^–/– mice. Protein (75 μg) was loaded on an 8% T, 2.5%C polyacrylamide gel; catalase was the protein loading control. One representative gel from 3 with similar results is shown. Densitometric analysis of IBs is shown in the bar graphs. Data are reported as mean ± SEM (n = 4). *P < 0.05 compared with WT animals by 1-way ANOVA. (B) 2D PCA scores plot demonstrating statistical clustering of WT and 4.2^–/– RBC metabolomic profiles (n = 4–6). The 15 metabolites contributing most to the separation of groups are reflected by high variable importance in projection (VIP) scores (bottom graph). These metabolites include intermediates of glycolysis, TCA, and GSH pathways. (C) Heatmap of the 30 most significant different features identified by t test (P < 0.005; n = 4–6). The heatmap scale ranges from –2 to 2 (Kyoto Encyclopedia of Genes and Genomes pathway metabolites) was expressed on a log2 scale. Figures were created using MetaboAnalyst 5.0. Wb: Western-blot; DU: density unit; VIP: variable importance in projection; S-ribosyl-L- ho, S-ribosyl-L- homocysteine; Geranyl pyrophosphate a, Geranyl-pyrophosphate 2-isopropylmatic acid.

Figure 2. Mitapivat improves anemia by metabolic reprogramming of 4.2^–/– mouse RBCs.

(A) Hct, Hb, and Hb to RDW ratio, as marker of spherocytosis and reticulocyte count in WT and 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) up to 6 months of age. The red arrow indicates the starting point for mitapivat administration. Data are reported as mean ± SEM (WT, n = 5; 4.2^–/– vehicle, n = 9; 4.2^–/– mitapivat, n = 12). *P < 0.05 compared with WT; °P < 0.05 compared with vehicle-treated 4.2^–/– mice by 1 way ANOVA with Dunnett’s longitudinal comparison. (B) 2D PCA scores obtained from the analysis of untargeted metabolites of RBCs from WT and 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) for 6 months. The value of each biological replicate was normalized. Data on WT animals are in red, on 4.2^–/– mice are in green, and on mitapivat-treated 4.2^–/– mice are in blue (n = 4–6). The variable importance in projection (VIP) score plot for the top 15 most important metabolite features identified by partial least squares discriminant analysis (PLS-DA). The box indicates the relative concentration from vehicle and mitapivat groups (right panel). Figures were created using MetaboAnalyst 5.0. (C and D) Plasma EPO, total bilirubin, and LDH, as markers of hemolysis, in WT and 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) for 6 months (n = 5). *P < 0.05 compared with WT; °P < 0.05 compared with vehicle-treated 4.2^–/– mice by 1-way ANOVA. dCMP, deoxycytidine monophosphate; dTMP, deoxytimidine monophosphate; AMP, adenosine monophosphate; UMP, uridine 5-monophosphate.

Figure 3. Mitapivat improves RBC osmotic fragility, reduces the amount of circulating annexin-V⁺ erythrocytes, and decreases the release of erythroid vesicles in 4.2^–/– mice.

(A) Representative scatter plots (left panel) of the osmotic fragility test determined by flow cytometry and percentage of RBC lysis (right panel) at 192 mOsm of RBCs from WT and 4.2^–/– mice treated with either vehicle or mitapivat (100 mg/kg/d) for 6 months. Results are reported as mean ± SEM from 3–4 mice/group. *P < 0.05 compared with WT; °P < 0.05 compared with vehicle-treated 4.2^–/– mice by 1-way ANOVA. Fsc, forward scatter; Ssc, side scatter. (B) Annexin-V⁺ RBCs from WT and 4.2^–/– mice treated with either vehicle or mitapivat (100 mg/kg/d) for 6 months. Results are mean ± SEM from 11–19 mice/group. *P < 0.05 compared with WT; °P < 0.05 compared with vehicle-treated 4.2^–/– mice by 1-way ANOVA. (C) Plasma erythroid microvesicles determined by flow cytometry from WT and 4.2^–/– mice treated with either vehicle or mitapivat (100 mg/kg/d) for 6 months. Results are reported as mean ± SEM from 5–6 mice/group. *P < 0.05 compared with WT by t test. (D) The bar chart shows flow cytometric analysis results of erythroid vesicles in vitro released under shared stress conditions from 4.2^–/– RBCs incubated for 50 minutes in the presence of vehicle or mitapivat (2 μM). Results are reported as mean ± SEM; n = 6. °P < 0.05 compared with vehicle-treated mice, determined by t test. Representative Coomassie-stained gel and IB analysis using specific Abs against band 3 and peroxiredoxin-2 (Prx-2) of erythroid vesicles in vitro released under shared stress conditions from WT and 4.2^–/– erythrocytes incubated for 50 minutes in the presence of either vehicle or mitapivat (2 μM). Results are reported mean ± SEM; n = 3–5. °P < 0.05 compared with vehicle-treated RBCs, determined by t test. Wb, Western blot; DU: density unit.

Figure 4. Mitapivat significantly reduces erythrophagocytosis, promotes a proresolving profile of splenic macrophages, and protects against hemolysis-induced inflammatory vasculopathy in 4.2^–/– mice.

(A) H&E staining (left column) and iron staining (Perls’ Prussian blue; right column) in spleens from WT and 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) for 6 months. Original magnification, ×200. Scale bars: 100 μM. One representative image from 4 with similar results is shown. (B) Percentage of Ter-119/F4/80 double-positive splenic macrophages isolated from 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) for 6 months, determined by flow cytometry. In parallel, surface expression of the M1-like marker CD80 was determined (see gating strategies for intracellular and surface staining analysis in left-side plots). Results are reported as mean ± SD from 10 mice/group. °P < 0.05 compared with vehicle-treated 4.2^–/– mice (unpaired t test). (C) IB analysis using specific Abs against Vcam1, Icam1, and Tbx in isolated aortas from WT and 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) for 6 months. Protein (50 μg) loaded on an 8% T, 2.5%C polyacrylamide gel. Actin was the protein loading control. One representative gel from 4–5 with similar results is shown. Densitometric analysis of IBs is shown in Supplemental Figure 6B. FSC, forward scatter; SSC, side scatter; Wb, Western blot.

Figure 5. Mitapivat-treated 4.2^–/– mice had reduced liver iron accumulation and protection against liver oxidation.

(A) H&E staining (left column) and iron staining (Perls’ Prussian blue; right column) in liver tissue from WT and 4.2^–/– mice treated with either vehicle or mitapivat (100 mg/kg/d) for 6 months. Original magnification, ×200. Scale bars: 100 μM. One representative image from 4 with similar results is shown. (B) Quantification of Perls’ iron staining of liver tissue (left bar graph) and the non-heme liver iron content determined using the bathophenanthroline staining method (right bar chart) in WT and 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) for 6 months. Data are reported as mean ± SEM (n = 3–7). *P < 0.05 compared with WT mice and °P < 0.05 compared with vehicle-treated mice by 1-way ANOVA. (C) OxyBlot analysis of the soluble fractions of liver from WT and 4.2^–/– mice treated as in A. The carbonylated proteins (1 mg) were detected by treating with 2,4-dinitrophenylhydrazine and blotted with anti–DNP Ab. Gapdh was the protein loading control. Quantification of band area is shown in Supplemental Figure 7B. (D) IB analysis using specific Abs against phosphorylated (p-)NF-κB p65, NF-κB p65, (p-)Nrf2, and Nrf2 in liver tissue from WT and 4.2^–/– mice treated as in A. Protein (75 μg) was loaded on an 8% T, 2.5%C polyacrylamide gel. Gapdh was the protein loading control. One representative gel from 4 with similar results is shown. Densitometric analysis of IBs is shown in Supplemental Figure 7C. (E) IB analysis using specific Abs against HO-1 and Gpx1 in liver from WT and 4.2^–/– mice treated as in A. Protein (50 μg) loaded on an 11% T, 2.5%C polyacrylamide gel. Gapdh was the protein loading control. One representative gel from 4 with similar results is shown. Densitometric analysis of IBs is shown in Supplemental Figure 7D. Wb, Western blot.

Figure 6. Noninferiority of mitapivat versus splenectomy in 4.2–/– mice is associated with protection against hemolysis-induced inflammatory vasculopathy.

(A) Experimental design to study the effect of splenectomy in combination with mitapivat (100 mg/kg/d) in 4.2^–/– mice. (B) Hb levels and reticulocyte count in 4.2^–/– mice treated with vehicle or mitapivat (100 mg/kg/d) or splenectomized (spleenx) monitored up to 7 months of age. Data are reported as mean ± SEM; 4.2^–/– vehicle, n = 5–7; 4.2^–/– mitapivat, n = 5-7; 4.2^–/– spleenx, n = 6). °P < 0.05 compared with vehicle-treated 4.2^–/– mice; §P < 0.05 compared with mitapivat-treated 4.2^–/– mice (1-way ANOVA with Dunnett’s longitudinal comparison). (C) Plasma LDH and total bilirubin (Tbil) levels in 4.2^–/– mice treated as described in A. Data are reported as mean ± SEM (n = 4); °P < 0.05 compared with vehicle-treated 4.2^–/– mice (1-way ANOVA). (D and E) Hb to RDW ratio and annexin-V⁺ RBCs from WT and 4.2^–/– mice with or without spleen and treated with either vehicle or mitapivat (100 mg/kg/d). Data are reported as mean ± SEM (n = 4–15). *P < 0.05 compared with WT mice, °P < 0.05 compared with vehicle-treated 4.2^–/– mice; §P < 0.05 compared with mitapivat-treated 4.2^–/– mice; ^P < 0.05 compared with spleenx 4.2^–/– mice treated with vehicle (2-way ANOVA). (F) IB analysis using specific Abs against Vcam1, Icam1, and Tbx in isolated aortas from 4.2^–/– mice with or without spleen and treated with either vehicle or mitapivat. Protein (50 μg/μL) loaded on an 8% T, 2.5%C polyacrylamide gel. Actin was the protein loading control. One representative gel from 3–5 with similar results is shown. Densitometric analysis of IBs is shown in Supplemental Figure 9B. Wb, Western blot.

Figure 7. A proposed mode of action of mitapivat in murine HS.

In 4.2^–/– mice, a model of HS, mitapivat improves anemia by metabolic reprograming of HS erythrocytes. This results in amelioration of HS RBC features with reduction of erythrophagocytosis and modulation of splenic macrophages toward a proresolving pattern with decreased splenic conditioning and reduced erythroid vesicle release. In 4.2^–/– mice, mitapivat induced improvement of chronic hemolysis reduces liver iron overload, modulates c-duodenum iron absorption via Dmt1 downregulation, and protects against hemolysis-induced inflammatory vasculopathy.