Structural and Functional Interactions of KCl Cotransport Proteins KCC1 and KCC3 in Sickle and Normal Erythrocyte Membranes

Abstract

The KCl cotransporter (KCC) mediates volume reduction in normal reticulocytes and exaggerated KCC activity in sickle red blood cells (SS RBC) (Joiner et al, Blood 109:1728, 2007) contributes to pathological dehydration that potentiates sickling. Three separate genes (KCC1, KCC3, KCC4) are expressed in RBC (Crable et al, Exp. Hem. 33:624, 2005). KCC1 and KCC3 proteins have been shown to interact in ex vivo expression systems (Simard et al, JBC 282(25):18083, 2007), and co-expression of an N-terminal truncation of KCC1 reduces KCC activity mediated by full-length KCC1 or KCC3 (Casula et al. JBC 276:41870, 2001), suggesting functional interaction. We show here via western blot analysis that SS RBC membranes contain more KCC1 protein (relative to KCC3) than AA RBC, independent of the reticulocytosis of sickle blood. Immunoprecipitation of solubilized SS RBC membranes with KCC3-specific antibody yielded a band at 125 kD on SDS PAGE which contained KCC1, as identified by western blotting with KCC1-specific antibody and by TOF mass spectroscopy. The effect of co-expression of KCC1 and KCC3 on KCC activity was assessed by measuring NEM-stimulated, Cl-dependent, (ouabain + bumetanide)-insensitive Rb uptake in HEK 293 cells. The Flip-In T-rex HEK 293 cell line (Invitrogen) containing a tetracycline-response promoter was transfected with a pcDNA5a plasmid containing KCC3a cDNA. Recombination of the plasmid with the integrated tet-promoter construct inserts the KCC3a gene under control of a tetracycline-responsive promoter. These cells were subsequently transduced with a retroviral vector (SF-91. Hildinger et at, Gene Ther . 5:1575, 1998) containing KCC1 cDNA linked to a GFP cassette. Control cells contained SF-91 vector lacking KCC1. Cells were selected for GFP expression and grown in the absence (un-induced, no KCC3a expression) or presence of tetracycline (induced, KCC3a expression). From this binary matrix, four types of cells were obtained: Cells with no additional KCC expression, representing endogenous KCC activity; cells with only KCC1 or KCC3a expression; cells with both KCC1 and KCC3a expression. Western blots indicated similar KCC1 expression in cells with KCC1 only and [KCC1 + KCC3] and similar KCC3 expression in cells with KCC3 only and [KCC1 + KCC3]. Thus, the expression of neither isoform was affected by the presence of the other. KCC activity in cells overexpressing KCC1 only was similar to endogenous activity in HEK 293 cells; i.e., transport activity of KCC1 alone was minimal. Cells overexpressing KCC3 only had a 5-fold increase in KCC activity over endogenous levels. When KCC1 was co-expressed with KCC3 in [KCC1 + KCC3] cells, an additional 50% increase in KCC activity was observed (p < 0.05 by paired t-test, N=4), despite similar levels of KCC3 expression by western blot analysis. This synergistic effect was dependent on the cytoplasmic N-terminus of KCC1, as it was not seen when the first 39 amino acids of KCC1 were removed. Interestingly, removal of the entire cytoplasmic N-terminal domain (117 aa) produced an inhibitory effect when co-expressed with KCC3a in HEK cells, as previously reported in Xenopus oocytes (Casula et al.). These data indicate that KCC1 and KCC3 interact structurally and functionally in RBC membranes, and provide another potential mechanism for regulation of KCC activity via multimeric associations between KCC isoforms. Thus, KCC activity could be modulated not only by transcriptional mechanisms and post-translational modification (phosphorylation), but also by altering the ratios of KCC isoforms or the kinetics of their association. We speculate that higher levels of KCC1 protein relative to KCC3 in SS RBC membranes could account for higher KCC activity in these cells relative to AA RBC.