(A) Phylogenetic tree of 147 unique members of the CaCA superfamily is shown. Five major branches in the tree are labelled, and representative members are shown (GenBank® accession numbers are given in parentheses): EcoYRBG, the YRBG protein from Escherichia coli (NP_417663); CerVCX, the vacuolar Ca2+/H+ exchanger from Saccharomyces cerevisiae, (NP_010155); AthCAX1 and AthCAX2, Ca2+/H+ exchangers from Arabidopsis thaliana (NP_181352 and NP_566452); NCX1, NCX2 and NCX3, Na+/Ca2+ exchangers from Homo sapiens (NP_066920, NP_055878 and NP_150287); NCKX1, NCKX2, NCKX3, NCKX4 and NCKX5, Na+/Ca2+–K+ exchangers from Homo sapiens (NP_004718, NP_065077, NP_065740, NP_705932 and NP_995322); NCLX/NCKX6 from Homo sapiens (NP_079235). (B) Putative general topology model for proteins of the CaCA superfamily. Hydrophobic presumed transmembrane helical segments are illustrated as cylinders and are linked to show the topological connectivity. The conserved α-repeat regions are shown boxed and oppositely oriented with respect to the membrane. The N-terminal transmembrane segment (light grey) is absent from some, and is thought to be removed by a co- or post-synthetic proteolytic cleavage event in other, members. The C-terminal two-transmembrane spans (light grey) are variably present in different family members. Additionally, and not shown, the CCX branch is predicted to have an additional hairpin pair of transmembrane helices between the central cytosolic loop and the α-2 motif. Modified from  with permission from Oxford University Press.
The experimentally supported model for NCX1 has nine proposed transmembrane-spanning segments (TMSs), a cleaved ‘signal peptide’ (TMS0, with the putative signal peptide cleavage site shown as SPase?) and re-entrant loops, one of which may be partially helical, as part of the conserved α-repeat motifs (dark blue, labelled α1 and α2). Critical amino acids in the α-repeat regions are highlighted in red (Glu113, Asn143, Asp814 and Asn842). The single N-linked glycosylation site is shown in green (N-CHO). Highlighted regions of the cytosolic loop include the XIP, the two CBDs (CBD1 and CBD2) and the region subject to alternative splicing. Proposed sites involved in regulatory Ca2+ binding are also shown.
Cartoon representations of the two CBD structures [59,60] (CBD1 is green, and CBD2 is magenta, with the region subject to alternative splicing, Thr570–Ala609, in yellow) are shown juxtaposed as they may pack in the native protein. The membrane plane lies at the top, orthogonal to the orientation of the Figure, connected to the CBDs via their labelled N- and C-terminal ends. The orientation is similar to that represented in Figure 2. Red spheres indicate the location of bound Ca2+ ions, and the two key amino acids that define the difference in properties between the mutually exclusive alternatively spliced exons A and B are shown as stick representations in blue (Asp578 and Lys585). Figure prepared with the PDB files 2FWS, 2FWU and 2DPK using PyMOL software (http://www.pymol.org). The 2FWS and 2FWU structures, as well as the location of the four Ca2+ ions in CBD1 from 2DPK, were used to prepare a model illustrating a possible packing arrangement, according to . Note that, although based on structural data, this Figure represents a possible composite model and not an actual structure. An interactive three-dimensional version of this Figure is available at http://www.BiochemJ.org/bj/406/0365/bj4060365add1.htm.
Figure 4Proposed transport cycle for Na+/Ca2+ exchangers
(A) The NCX1 cycle. The empty exchanger is shown to bind either one Ca2+ ion or three Na+ ions, in a mutually exclusive manner. Either binding event allows a conformational change that reorients the accessibility of the ion-binding site across the membrane barrier. Forward-mode Ca2+ extrusion in exchange for Na+ entry occurs by sequential steps in a clockwise direction. Conversely, reverse-mode Ca2+ influx in exchange for Na+ exit proceeds around the cycle in an anti-clockwise direction. (B) The NCKX cycle. As in (A), except the empty transporter binds either one Ca2+ ion plus one K+ ion or four Na+ ions. The binding of Ca2+ and K+ is shown ordered based on unpublished work (F. Visser and J. Lytton) and because kinetic considerations under normal physiological ion concentrations suggest that the cycle is likely to be limited by Ca2+ binding to an exchanger with a cytoplasmically oriented site primed with bound K+. An animated version of this Figure is available at http://www.BiochemJ.org/bj/406/0365/bj4060365add2.htm.
The experimentally supported topology model for NCKX2 has ten transmembrane-spanning segments (TMSs), a cleaved ‘signal peptide’ (TMS0, with the putative signal peptide cleavage site shown as SPase?), but no re-entrant loops. The conserved α-repeat motifs are shown in dark blue (labelled α1 and α2), and the single N-linked glycosylation site is shown in green (N-CHO). Critical amino acids in the α-repeat regions are highlighted in red (Glu188, Asn215, Asp548, Thr551 and Asp575). In the cytosolic loop, the region subject to alternative splicing is indicated, as well as three consensus PKC phosphorylation sites thought to be involved in regulation of activity (Thr166, Thr476 and Ser504) and a cysteine residue (Cys395) thought to be important for redox regulation of NCKX2 activity and higher-order oligomer formation.