Deshpande CN et al. (2018), Calcium is an essential cofactor for metal effl...

XB-ART-55599

Nat Commun 2018 Aug 06;91:3075. doi: 10.1038/s41467-018-05446-4.

Show Gene links Show Anatomy links

Calcium is an essential cofactor for metal efflux by the ferroportin transporter family.

Deshpande CN , Ruwe TA , Shawki A , Xin V , Vieth KR , Valore EV , Qiao B , Ganz T , Nemeth E , Mackenzie B , Jormakka M .

???displayArticle.abstract???
Ferroportin (Fpn)-the only known cellular iron exporter-transports dietary and recycled iron into the blood plasma, and transfers iron across the placenta. Despite its central role in iron metabolism, our molecular understanding of Fpn-mediated iron efflux remains incomplete. Here, we report that Ca2+ is required for human Fpn transport activity. Whereas iron efflux is stimulated by extracellular Ca2+ in the physiological range, Ca2+ is not transported. We determine the crystal structure of a Ca2+-bound BbFpn, a prokaryotic orthologue, and find that Ca2+ is a cofactor that facilitates a conformational change critical to the transport cycle. We also identify a substrate pocket accommodating a divalent transition metal complexed with a chelator. These findings support a model of iron export by Fpn and suggest a link between plasma calcium and iron homeostasis.

???displayArticle.pubmedLink??? 30082682
???displayArticle.pmcLink??? PMC6079014
???displayArticle.link??? Nat Commun
???displayArticle.grants??? [+]

???attribute.lit??? ???displayArticles.show???

	Fig. 1. Calcium activates Fpn-mediated iron efflux in Xenopus oocytes. a First-order rate constants (k) for 55Fe efflux from control oocytes and oocytes expressing human Fpn, in the presence (+Na+o) or absence (−Na+o) of 100 mM extracellular Na+; n = 10, 14, 16, 17 oocytes per group (left to right). Two-way ANOVA revealed no interaction (P = 0.86); within Fpn, +Na+o did not differ from −Na+o (P = 0.73). b 55Fe efflux from control oocytes and oocytes expressing human Fpn in the presence of 2 mM extracellular Ca2+ or in its absence (i.e. 0 mM Ca2+ plus 1 mM EGTA) (n = 16, 16, 15, 16). Two-way ANOVA: interaction (P < 0.001). c 55Fe efflux from control oocytes (red) and oocytes expressing Fpn (black) as a function of [Ca2+]o (control, n = 30; Fpn, n = 84; i.e. 6–9 oocytes per group at each [Ca2+]o). Fpn data were fit by Eq. (2): \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{0.5}^{{\mathrm{Ca}}}$$\end{document}K0.5Ca = 0.8 ± (SEM) 0.2 mM, Vmax occurred at k = (16 ± 2) × 10−3 min−1 (r2 = 0.88, P < 0.001). d 57Co efflux from control oocytes and oocytes expressing Fpn in the presence of 2 mM extracellular Ca2+ or in its absence (i.e. 0 mM Ca2+ plus 1 mM EGTA) (n = 7, 7, 6, 6). Two-way ANOVA: interaction (P < 0.001). e 65Zn efflux from control oocytes and oocytes expressing Fpn injected with 50 nl of 50 µM 65Zn in the presence of 2 mM extracellular Ca2+ or in its absence (i.e. 0 mM Ca2+ plus 1 mM EGTA) (n = 19, 21, 20, 19). Two-way ANOVA: interaction (P < 0.001). f 63Ni efflux from control oocytes and oocytes expressing Fpn in the presence of 2 mM extracellular Ca2+ or in its absence (i.e. 0 mM Ca2+ plus 1 mM EGTA) (n = 10, 9, 12, 12). Two-way ANOVA: interaction (P < 0.001). All bar graphs indicate mean, SD, and individual scores; data in (c) are mean, SD
	Fig. 2. Alkaline-earth metal selectivity. a 55Fe efflux from control oocytes and oocytes expressing Fpn, in the presence of extracellular calcium (2 mM) and magnesium (1 mM), the absence of extracellular calcium (–Ca2+oi.e. 0 mM Ca2+ plus 1 mM EGTA), or the absence of extracellular magnesium (–Mg2+o)i.e. 0 mM Mg2+ plus 1 mM EDTA) (n = 12 in each group). Two-way ANOVA: interaction (P < 0.001); within Fpn, all groups differ from one another (P ≤ 0.004). b 55Fe efflux from control oocytes and oocytes expressing Fpn in the presence of 3 mM extracellular magnesium (+Mg2+), or 1 mM Mg2+ plus 2 mM calcium (+Ca2+), strontium (+Sr2+), or barium (+Ba2+) (n = 30, 39, 28, 32, 27, 36, 25, 32). Two-way ANOVA: interaction (P < 0.001). In multiple pairwise comparisons vs. control: Fpn differed from control in the case of Ca2+ and Sr2+ (P < 0.001), but not Mg2+ (P = 0.14) or Ba2+ (P = 0.082); within Fpn, all conditions differed from ‘+Ca2+’ (P < 0.001). Bar graphs indicate mean, SD, and individual scores
	Fig. 3. No evidence to support that calcium is a transported substrate of Fpn. a Uptake of 100 μM 45Ca2+ in control oocytes and oocytes expressing Fpn that were injected with either vehicle alone (−Fei) or 50 nl of 50 µM Fe (+Fei) (n = 12, 12, 13, 14, left to right). Two-way ANOVA: no interaction (P = 0.70). b Uptake of 100 μM 45Ca2+ in control oocytes and oocytes expressing Fpn in the absence of extracellular Na+ (Na+ was replaced by choline) (n = 12, 8, 11, 12). Oocytes were injected with either vehicle alone (−Fei) or 50 nl of 50 µM Fe (+Fei). Two-way ANOVA: interaction (P = 0.008); within Fpn, +Fei did not differ from −Fei (P = 0.069). c Effect of 1 µM ionomycin (IMN) or 2 µM thapsigargin (THG) on 55Fe efflux (n = 15, 14, 14, 16, 15, 14). All media contained 0.2% DMSO. Two-way ANOVA: no interaction (P = 0.13). d Efflux of 55Fe or 45Ca efflux from control oocytes and oocytes expressing Fpn (n = 11, 12, 9, 10). Two-way ANOVA: interaction (P < 0.001); within 45Ca, Fpn did not differ from control (P = 0.98). All bar graphs indicate mean, SD, and individual scores
	Fig. 4. Calcium activates Fpn-mediated iron efflux in a HEK cell expression system. HEK293T cells expressing Fpn were loaded with 55Fe, and 55Fe efflux measured in media containing 1 mM Ca2+ (black) or 0 Ca2+ +1 mM EGTA (green). Data, normalized by 55Fe content at time = 0, are displayed as induced minus uninduced in the same preparation. Each symbol-and-line plot represents one of four independent preparations. Two-way RM ANOVA: interaction of [Ca2+]o and time (P < 0.001); Ca2+ significantly activated 55Fe efflux over 6 h (P = 0.002) and 24 h (P < 0.001), but not at earlier time points (P ≥ 0.24)
	Fig. 5. Crystal structure of a Ca2+-bound BbFpn reveals a conserved binding site. a Overall structure of the Ca2+ -bound inward facing conformation of BbFpn and close up view of the Ca2+ site (right). Approximate location of the membrane bilayer is illustrated with red (outside) and blue (inside) dots, generated using the PPM server41. TM7 is highlighted in orange. The Ca2+ ion is colored in green, and coordinating residues are shown in ball-and-stick. Some of the distances are longer than the reported ideal Ca2+–O distance (~2.5 Å)42, although this is not unprecedented considering the resolution of the structure17. b Intracellular view (left) and Fo–Fc omit map of the Ca site (right; contoured at 5.5σ). c, d Binding of the Ca2+ ion leads to a kink forming in TM6, likely promoted by the Ca2+ coordinating N196 and E203. e Representative ITC experiment between wild type BbFpn/Ca2+. Stoichiometry (N) and dissociation constant (Kd) of each interaction are indicated. The variations in free energy (ΔG), enthalpy (ΔH), and entropy (−TΔS) are respectively, in blue, green, and red in the histogram
	Fig. 6. Mutagenesis of putative Ca2+-coordinating residues in human Fpn. a 55Fe efflux from control oocytes and oocytes expressing wildtype or mutant Fpn in media containing 2 or 20 mM Ca2+ (n = 77, 54, 96, 69, 37, 35, 37, 35, 56, 39, 34, 29, left to right). Two-way ANOVA: interaction, P < 0.001; Q99A and N212Q differed from control (P < 0.001) but D39E and E219D did not differ from control (P ≥ 0.78). Raising [Ca2+]o from 2 to 20 mM increased 55Fe efflux for Q99A and N212Q (P ≤ 0.001) but not for any other mutant (P ≥ 0.61). Relative expression levels were estimated from GFP fluorescence (Supplementary Fig. 7A). b 55Fe efflux from control oocytes (red, n = 22) and oocytes expressing wildtype Fpn (black, n = 62) or Q99A (blue, n = 76) as a function of [Ca2+]o (i.e. 6–9 oocytes at each concentration). Data were fit by Eq. (2): wildtype Fpn, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{0.5}^{{\mathrm{Ca}}}$$\end{document}K0.5Ca = 0.9 ± (SEM) 0.1 mM, Vmax occurred at k = (24 ± 1) × 10−3 min−1 (r2 = 0.97, P < 0.001); and Q99A, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{0.5}^{{\mathrm{Ca}}}$$\end{document}K0.5Ca = 2.4 ± 0.4 mM, Vmax occurred at k = (20 ± 1) × 10−3 min−1 (r2 = 0.83, P < 0.001). c 55Fe efflux from control oocytes and oocytes expressing wildtype or mutant Fpn (n = 12, 11, 12, 12, 11, 12, 12, 10). Two-way ANOVA: interaction, P = 0.009; neither mutant differed from control at either 2 or 20 mM Ca2+ (P ≥ 0.76). Raising [Ca2+]o from 2 to 20 mM increased 55Fe efflux for wildtype Fpn (P < 0.001) but not for either mutant (P ≥ 0.73). All bar graphs indicate mean, SD, and individual scores; data in b are mean, SD
	Fig. 7. Putative substrate-binding site located in the C-terminal domain. a Surface representation of the C-terminal domain of BbFpn, viewed from the central cavity. The pocket formed by the non-continuous TM7 is in the center of the domain, with the Ni-EDTA molecule shown in ball-and-stick. Insets show Fo–Fc omit (top) and 2Fo–Fc map (bottom) of the Ni-EDTA site. b Close-up view of the Ni-EDTA-binding site and coordination. Main hydrogen bonds are indicated by dotted lines. R348 forms a salt-bridge to one of the carboxylate groups of the EDTA moiety, shown in yellow, whereas H261 is a direct ligand to the metal (green sphere). c Superposition of the Ni-EDTA bound structure and the previously determined open inward conformation (gray; PDB entry 5AYO) illustrating conformational change of H261. d Effects of mutating residues within the putative metal-binding site of human Fpn. 55Fe efflux (black) and 63Ni efflux (green) from control oocytes and oocytes expressing wtFpn or one of several amino-acid substitutions at residue D325 (n = 24, 29, 32, 26, 30, 22, 28, 35, 33, 35, left to right). The bar graph indicates mean, SD, and individual scores. Two-way ANOVA revealed an interaction (P < 0.001). Within 55Fe efflux, each mutant differed from wtFpn (P < 0.001); and D325N (P < 0.001) but neither D325A (P = 0.78) nor D325H (P = 0.025) differed from control. Within 63Ni, all groups differed from control (P < 0.001); D325A and D325H differed from wtFpn (P < 0.001) but D325N did not differ from wtFpn (P = 0.14). e–g Working model to illustrate Ca2+-activated Fpn-mediated iron efflux

References [+] :

Abramson, Structure and mechanism of the lactose permease of Escherichia coli. 2003, Pubmed