Red Blood Cell Pyruvate Kinase Properties in Sickle Cell Disease - of Mice and Men

Background: Townes and Berkeley mouse models of sickle cell disease (SCD) are commonly used to study SCD pathophysiology and to develop novel therapies. Although the pathophysiology in these models is similar to the human disease, there are distinct differences. For example, Townes SS mice have higher ATP and lower 2,3-diphosphyglycerate (2,3-DPG) levels compared to Townes AA mice. This contrasts with SCD patients, who show lower ATP and higher 2,3-DPG levels, suggesting potential differences between human and murine red blood cell (RBC) metabolism. Pyruvate kinase (PK), a key ATP-generating enzyme in glycolysis, is a target for novel SCD therapies. While PK activators have been shown to increase hemoglobin in clinical trials, such treatments do not consistently lead to the same hematological change in mouse models, further indicating that PK properties may differ between SCD mouse models and human patients. A detailed comparative characterization of PK and glycolytic metabolites in murine SCD models and SCD patients is therefore important.

Aim: To study PK properties in murine and human SCD RBCs.

Methods: Wild-type (C57BL/6J), Townes (AA and SS) and Berkeley mice (non-sickling and sickling) were used. Human blood was obtained from healthy controls (HbAA) and untreated SCD patients (HbSS). Hematological parameters were measured using the Sysmex hematology analyzer. PK (Vmax, 5mM substrate) and hexokinase (HK) activity, and PK thermostability (53°C) were measured in purified RBCs. Levels of ATP, 2,3-DPG, as well as reduced glutathione (GSH) were measured by targeted LC-MS/MS and spectrophotometry, respectively. P50 (oxygen tension at which 50% of hemoglobin is saturated with oxygen) was measured with the Hemox analyzer (TCS). Statistical analysis was performed in Graphpad Prism 9.4.1. by Welch's t-test.

Results: Samples from 17 wild-type mice, 21 Townes mice (10 AA, 11 SS), 15 Berkeley mice (7 non-sickling, 8 sickling) and 12 human subjects (6 HbAA, 6 HbSS) were included. Both Townes and Berkeley SCD mice showed increased PK activity (Table 1; p<0.0001). This may be attributed to their high reticulocyte counts (Townes SS vs. AA: 43% vs. 4%; Berkeley sickling vs. non-sickling: 30% vs. 6%), because when normalizing PK activity to that of HK, another RBC age-dependent enzyme, no genotype differences were found in these mouse models. This lack of PK/HK difference was in sharp contrast with our finding in human controls vs. SCD patients (Table 1). Notably, both Berkeley sickling and non-sickling mice had lower PK/HK ratio than their Townes counterparts (p<0.0001). Mouse PK thermostability was markedly reduced compared to that of human, and PK from the two sickling mice was less stable than their non-sickling controls (Fig. 1). Moreover, both sickling mice showed higher ATP/2,3-DPG ratio (Table 1; p<0.01). This was again in direct contrast with our observation using human blood, where SCD patients had lower ATP/2,3-DPG ratio than healthy controls (p<0.05). Even though there was no difference in ATP/2,3-DPG ratio between the two SCD models, Townes SS mice had significantly lower 2,3-DPG and higher ATP levels than Townes AA mice, whereas Berkeley sickling mice only had significantly higher ATP level compared to their controls. Furthermore, sickling mice had lower P50 value compared to non-sickling mice, reflecting a higher oxygen affinity. This also contrasted with human SCD patients, who had higher P50 value than healthy controls (Table 1). GSH levels were increased in both SCD patients and sickling mice, although the elevation was more pronounced in sickling mice (Table 1).

Conclusions: This study reveals important distinctions between mouse models and humans: no differences in PK/HK ratio between sickling and non-sickling mice, and an overall lower PK thermostability in mice. Furthermore, this study shows significant changes in 2,3-DPG content (important for oxygen affinity), energy level (ATP), oxygen affinity (P50) and antioxidant level (GSH) when comparing sickling to non-sickling mice. Sickling mice also have higher ATP/2,3-DPG ratio, which contrasts with SCD patients. These differences should be considered when studying PK activators in SCD mouse models and extrapolating results from SCD mouse models to humans.