Regulation of Mature Erythrocyte Proteasomal Machinery Under Hypoxia

Proteasome machinery is a conserved cellular component to maintain normal protein homeostasis. Hypoxia is well known to control hypoxia inducible factor levels by proteasomal machinery in nucleated cells. However, the specific targets and regulation of proteasomal machinery in non-nucleated mature erythrocytes under hypoxia remain poorly understood. To determine if hypoxia regulates erythrocyte proteasomal machinery, we conducted Western blot to detect total ubiquitinated and K48 specific ubiquitinated proteins on the erythrocyte membrane in both human and mice with or without sickle cell disease (SCD), a hemolytic genetic disease with a high mortality, morbidity and frequently facing hypoxia. We found that ubiquitinated, especially K48 specific ubiquitinated proteins were significantly accumulated in both SCD Berkeley mice and humans compared to WT mice and normal controls, indicating that proteasomal machinery is impaired in SCD. Next, to determine specific ubiquitinated proteins accumulated on the membrane of human sickle erythrocytes (sRBC), we conducted immunoprecipitation of sRBC membrane proteins with total ubiquitin antibody followed by an robust and nonbiased proteomic profiling. We found significant accumulation of several categories of ubiquitinated proteins on human mature sRBC membrane, including cytoskeleton proteins (Spectrin, Actin, Ankryin), glycolytic enzymes (GAPDH, 2,3-BPG mutase, Pyruvate Kinase, G6PD), transporters (Band3, large neutral AA transporter, calcium transporter, ENT1), reactive oxygen species (ROS) related enzyme (catalase), components of proteasome machinery [E2, E3 ligases, and valosin-containing protein (p97)]. Mechanistically, we revealed that the impaired proteasomal machinery found in mature sRBC was due to the blockage of trafficking of p97 bound ubiquitinated-proteins from membrane to cytosolic proteasome. As such, inhibition of p97 by CB-5083 or proteasome by MG132 led to further induction of hypoxia-induced ubiquitinated membrane proteins and sickling in cultured human sRBC. Given the fact that sphingosine-1-phosphate (S1P) contributes to sickling by binding with deoxygenated sickle Hb (deoxy-HbS), triggering deoxy-HbS anchoring membrane and releasing glycolytic enzymes, we immediately hypothesized that S1P may be involved in proteasomal machinery by regulating trafficking of p97-bound ubiquitinated proteins from membrane to cytosol in sRBC. To test this intriguing possibility, we generated SCD/Sphk1^-/- mouse. Intriguingly, we found that the genetic deletion of SphK1 attenuated impaired proteasomal machinery in sRBC with less accumulation of p97 and ubiquitinated proteins on sRBC membrane, indicating that elevated S1P is detrimental in sRBC by inducing accumulation of p97 and ubiquitinated proteins on the membrane. Moreover, to determine if S1P-regulated p97 trafficking from membrane to the cytosol is unique to sRBC, we exposed wild type and SphK1^-/- mice to 8% hypoxia up to 72 hours. In contrast to sRBC, we found that genetic deletion of SphK1 abolished p97 trafficking from membrane to cytosol in normal erythrocytes under hypoxia. Finally, we conducted in vitro proof-of-principle genetic studies to determine if S1P directly involves in translocation of membrane anchored p97 to cytosol using inverted ghost membrane (IGM). We demonstrated that S1P treatment induces deoxy-HbA translocation from cytosol to membrane and in turn restoring p97 release from membrane to the cytosol in IGM isolated from SphK1^-/- mice only under hypoxia but not normoxia. Thus, we have provided both human and mouse genetic evidence supporting a working model: in normal individuals under hypoxia, S1P is a key factor regulating the efficient proteasomal machinery by binding deoxy-HbA, promoting deoxy-HbA anchoring membrane and in turn triggering release of p97-ubiqutinated proteins to cytosol for its degradation. With mutation in HbS, S1P promotes deoxy-HbS anchoring the membrane and forms polymers, thus blocks membrane bound p97 release and impairs proteasomal machinery in sRBC. Overall, our findings identified that S1P is a missing key component of proteasome machinery by its differential mechanism regulating p97 trafficking from membrane to cytosol in normal erythrocyte physiology under hypoxia and the pathophysiology of SCD that open up new and differential therapies for the SCD and normal individuals facing hypoxia.