Secondary Deficiency of Pyruvate Kinase and Altered Red Cell Metabolism in Hereditary Xerocytosis Caused By PIEZO1 and KCNN4 Defects

INTRODUCTION

Hereditary xerocytosis (HX) is a rare red blood cell (RBC) disorder associated with hemolysis and iron overload. HX is caused by mutations in PIEZO1, the gene coding for the mechanosensitive cation channel PIEZO1, or KCNN4, the gene encoding for the Gardos channel, a calcium-dependent potassium channel. In both cases, inappropriate Gardos channel activation leads to loss of potassium and water, thereby causing dehydration of the red cell. The mechanism by which HX RBCs are cleared from the circulation is unknown. In addition, to date no therapy is available for HX. Previously we reported that decreased activity and thermostability of pyruvate kinase (PK), a key regulatory enzyme of red cell glycolysis, is a feature of various forms of hereditary hemolytic anemia.¹ Here we report on the study on PK properties and associated metabolic alterations in HX patients with PIEZO1 or KCNN4 defects.

METHODS

Enzymatic activities of RBC PK and hexokinase (HK) were measured together with PK-thermostability in order to assess relative PK activity and enzyme stability. RBC PK protein levels were measured by Mesoscale assay. 2,3-DPG and ATP levels were measured on whole blood by LC-MS/MS. Metabolomics was performed by untargeted Direct-infusion High Resolution MS on dried blood spots.

RESULTS

Ten PIEZO1-HX patients, 3 KCNN4-HX patients and 8 healthy controls (HC) were studied. Hereditary deficiency of PK was excluded by DNA sequence analysis of PKLR. PK activity was significantly decreased compared to HK in PIEZO-1 HX (median ratio 4.6) and KCNN4-HX patients (median ratio 5.8) compared to HC (median ratio 9.4) (Fig 1A). This relative deficiency of PK was accompanied by a significant substantial decrease in PK thermostability in HX RBCs (median residual activity 39% and 61% for PIEZO1-HX and KCNN4-HX, respectively) compared to HC (median residual activity 78%, Fig 1B). In both cases, the decrease was most pronounced in PIEZO1-HX RBCs. PIEZO1-HX RBCs also had lower PK protein levels (129 μg/gHb, Fig 1C) compared to HCs (174 μg/gHb), whereas levels resembled HCs in KCNN4-HX RBCs (165 μg/gHb). Subsequent analysis of red cell metabolites showed statistically decreased levels of 2,3-DPG in RBCs from in PIEZO1 HX patients, whereas they were normal in KCNN4-HX RBCs (Figure 1D). This was confirmed by untargeted metabolomics on dried blood spots (data not shown). In both patient groups RBC ATP levels were not different from HCs (data not shown).

CONCLUSION

PK enzyme activity and stability are compromised in HX, in particular in PIEZO1-HX patient RBCs. In PIEZO1-HX, but not KCNN4-HX RBCs, this was associated with decreased levels of PK protein, indicating different mechanisms causing secondary partial PK deficiency in the two forms of HX. Metabolic analysis of PIEZO1-HX RBCs showed normal ATP and decreased 2,3-DPG levels. This contrasts with the decrease in ATP and increase in 2,3-DPG as the key metabolic hallmarks of hereditary PK deficiency, suggesting complex metabolic (dys)regulation in RBCs with defective PIEZO1 function. Since decreased levels of 2,3-DPG increases oxygen affinity of hemoglobin, this may explain the mild erythrocytosis seen in some PIEZO1-HX patients. Interestingly, in KCNN4-HX RBCs the loss of PK activity and thermostability was not associated with altered levels of 2,3-DPG and ATP, highlighting an yet unknown difference between the two forms of HX. Current in depth studies are in progress to map the metabolic changes in PIEZO1- and KCNN4-HX in more detail.

Our results indicate that compromised PK function, possibly resulting from altered PIEZO1 function, could contribute to the pathophysiology of PIEZO1-HX by affecting RBC glycolysis and probably also other metabolic pathways. Our findings may enhance our understanding of HX pathophysiology, and offer novel opportunities for therapeutic strategies in HX.