Proteomic Analysis of RBC Ghosts from the Beta-Adducin and Protein 4.2 Knock-Out Mouse Models of Inherited Hemolytic Anemia.

Abstract

Inherited hemolytic anemia (spherocytosis or elliptocytosis) is one of the most common inherited diseases with an incidence of 1:2500 to 1:5000 in populations of Northern European descent. While it is known that mild to severe inherited hemolytic anemias can arise from defects in the red blood cell (RBC) membrane skeleton, fundamental questions remain unanswered surrounding the clinical variability and non-erythroid effects of known RBC membrane skeleton mutations. To identify proteins that may be involved in disease severity and secondary effects, we used shotgun proteomics to globally profile proteins in RBC ghosts (i.e., RBC membrane skeleton and associated proteins) from well-defined mouse models of inherited hemolytic anemia. A peptide level ‘bottom-up' analysis was performed on RBCs from normal mice, beta-adducin knock-out mice (Add2-KO, compensated anemia), and protein 4.2 knock-out mice (4.2-KO, mild anemia). For each genotype, whole blood was taken from independent biological replicates and RBCs were purified using cellulose acetate chromatography. The isolated RBCs were lysed to generate RBC ghosts whose protein complements were digested with trypsin. For each biological replicate, five replicate runs utilizing 0.1 ug digested protein were performed via microcapillary liquid chromatography coupled with tandem mass spectrometry. Normal versus diseased comparisons were made using a protein profile found consistently across all independent samples for each genotype. In total, 435 unique proteins were identified for the normal mouse RBC ghost. In contrast, 731 and 848 unique proteins were identified for the Add2-KO and 4.2-KO mice RBC ghosts respectively. Previously identified membrane skeleton proteins were found for all three genotypes with the predicted absence of the knock-out proteins. In addition to well-known membrane proteins, a surprising number of proteins were found involved in processes such as protein repair, protein degradation, Ras oncogene biology, and glycolysis. For both knock-out mice, a large number of proteins involved in translation were identified most likely reflecting their elevated reticulocytosis status. Comparison of the normal and Add2-KO RBC profiles revealed 5 proteins present only in normal the RBC while 53 proteins were present only in the diseased RBC. Likewise, normal vs 4.2 KO comparison revealed 6 proteins present only in the normal RBC while 111 were only in the diseased RBC. Comparison between the two KO mice revealed 34 proteins present only in the Add2-KO and 88 proteins present only in the 4.2 KO. Some of the identified differences are proteins with unknown functions (example, SH3-binding domain glutamic acid-rich protein like). Other differences involve proteins associated with diverse processes such as protein folding (Bcl2-associated athanogene 2), protein modification (magnesium-dependent phosphatase-1), protein transport (RAB35), metabolism (N-acetylneuraminic acid phosphatase), signal transduction (Prohibitin 2), and apoptosis (Rho GTPase activating protein 1). We report that tandem mass spec analysis of disease model RBC ghosts have demonstrated differences in their proteomes and that these identified differences potentially represent candidate proteins involved in disease severity and secondary effects.