Ca2+/H+ exchange by acidic organelles regulates cell migration in vivo

Abstract

Increasing evidence implicates Ca(2+) in the control of cell migration. However, the underlying mechanisms are incompletely understood. Acidic Ca(2+) stores are fast emerging as signaling centers. But how Ca(2+) is taken up by these organelles in metazoans and the physiological relevance for migration is unclear. Here, we identify a vertebrate Ca(2+)/H(+)exchanger (CAX) as part of a widespread family of homologues in animals. CAX is expressed in neural crest cells and required for their migration in vivo. It localizes to acidic organelles, tempers evoked Ca(2+) signals, and regulates cell-matrix adhesion during migration. Our data provide new molecular insight into how Ca(2+) is handled by acidic organelles and link this to migration, thereby underscoring the role of noncanonical Ca(2+) stores in the control of Ca(2+)-dependent function.

Figure 1.

CAXs are widespread in the animal kingdom. (a) Proposed consensus topology of animal CAXs depicting a cytosolic N terminus, 13 transmembrane regions (cylinders), and a luminal C terminus. An N-terminal domain of unknown function (DUF 307) and a Ca²⁺/H⁺ antiporter (ChaA) domain comprising 11 transmembrane regions numbered 0–10 are highlighted. (b) CAX phylogeny in major metazoan phyla. Organisms depicted from left to right are Aplysia californica, Capitella teleta, Strongylocentrotus purpuratus, Branchiostoma floridae, Takifugu rubripes, Latimeria chalumna, Xenopus tropicalis, Anolis carolinensis, Gallus gallus, Ornithorhynchus anatinus, and Sarcophilus harrisii. See Table S2 for accession numbers. (c) Expression of animal CAXs. Agarose gel analysis of PCR reactions using primers specific for S. purpuratus (left) or Xenopus (right) CAX and templates prepared from RNA isolated from eggs and stage 16–18 embryos, respectively. Reactions were performed either with (+) or without (−) reverse transcription (RT). (d) Sequence comparison of animal, plant, and yeast CAXs. ClustalW2 multiple sequence alignment of TM2 and TM7 of the α1 and α2 repeats of CAX, respectively, from the sea urchin (SpuCAX), frog (XlaCAX), yeast (SceVCX1), and plant (AthCAX). *, glutamate residues conserved in both repeats. (e) Structural model of animal CAX. Magnified view of TM2 and 7 of Xenopus CAX highlighting E420 and E648, likely important for ion coordination. (f and g) Effect of CAX on Ca²⁺ homeostasis in yeast. (f) Inductively coupled plasma emission spectroscopy of total Ca²⁺ content. Data (n = 9) are from wild-type (WT) or Δvcx1Δpmc1 yeast grown in 10 mM Ca²⁺ and transformed with empty vector (Vec.) or vectors expressing CAX, CAXΔN, and CAX^E420A. (g) Ca²⁺/H⁺ exchange measurements using endomembrane-enriched membrane vesicles and acridine orange fluorescence recovery in response to 50 µM Ca²⁺. Data (n = 4 preparations) are from wild-type or Δvcx1Δvnx1 yeast transformed with the indicated vector. Error bars represent SEM. **, P < 0.01.

Figure 2.

CAX is required for cell migration in vivo. (a) Tissue distribution of frog CAX. In situ hybridization of CAX and the neural crest marker Twist in stage 26/27 Xenopus embryos. Arrowheads mark the neural crest streams. (b–d) Effect of CAX knockdown on neural crest migration in vivo. (b) In situ hybridization analyses of Twist and the additional neural crest marker Snail2 in exemplar stage 23/24 embryos injected with two antisense morpholinos (10 ng AMO1 or 20 ng AMO2). Dorsal views (left) and views along the anterior–posterior axis (right) for both the uninjected (control) and antisense morpholino–injected (*) sides of embryos (delineated by the dashed lines) are shown. Phenotypes were considered mild or severe. (c) Summary data quantifying the proportion of embryos displaying the indicated phenotype in response to increasing amounts of antisense morpholino. Results are from a single experiment where all six experimental manipulations were performed in parallel on the same batch of embryos (∼1,000 injections). (d) Summary data from seven knockdowns (using 10 ng AMO1 and Twist) quantifying the length of all three neural crest streams, normalized to that in the uninjected side of the same embryo. Results are from CAX-depleted morphant embryos coinjected with increasing amounts of mRNA (0.6 or 1.2 ng; color coded) for GFP-CAX (left) or from control embryos injected with mRNA alone (right). Bars, 500 µm. Error bars represent SEM. **, P < 0.01; ***, P < 0.001.

Figure 3.

CAX localizes to acidic organelles. (a and b) Subcellular distribution of CAX in neural crest cells. (a) Confocal (top) and transmitted light (bottom) micrographs of exemplar neural crest cells expressing GFP-CAX. Inset shows a magnified image highlighting large (open triangles) and small (closed triangles) GFP-positive structures. (b) Images of GFP-CAX–expressing cells labeled with Lysotracker red (Lyso.). (Left) Images of large vesicles from cells loaded with 30 nM Lysotracker. (Right) Magnified images of small vesicles from independent cells loaded with 300 nM Lysotracker. Intensity plots of red and green fluorescence between the white lines in the overlaid image are shown below the images. (c–f) Subcellular distribution of CAX in human cells. (c) Confocal micrographs of the indicated cell type transiently transfected with GFP-CAX and fixed before imaging. (d) Distribution of GFP-CAX in live HeLa cells labeled with 100 nM Lysotracker red. (e) Distribution of GFP-CAX in fixed HeLa cells coexpressing LAMP1-mRFP. (f) Immunocytochemistry analysis using an antibody raised to LAMP1 in SHSY5Y cells or fibroblasts expressing GFP-CAX. Bars: (a and b) 5 µm; (c–f) 10 µm.

Figure 4.

CAX regulates cytosolic Ca²⁺ signaling. (a–f) Effect of CAX overexpression on agonist-evoked Ca²⁺ signals. (a) Cytosolic Ca²⁺ levels of SHSY5Y cells loaded with the Ca²⁺ indicator Fura-2 and stimulated with 100 µM carbachol. Cells were from cultures transiently transfected with mRFP-CAX or LAMP1-mRFP, and data were segregated according to whether they were mRFP positive (transfected) or not (untransfected). (b) Pooled data (from seven transfections) quantifying the area under the curve (integral) and the peak response. (c) Effect of mRFP-CAX on carbachol-evoked Ca²⁺ signals in the presence of GPN. Cultures were treated with DMSO or 200 µM GPN for 30 min before stimulation. (d) Pooled data from four transfections. (e) Cytosolic Ca²⁺ responses of cells stimulated with 10 µM carbachol from cultures transiently transfected with LAMP1-mRFP, CAX-mRFP, CAXΔN-mRFP, or mRFP-CAX^E420A. Insets are expanded views of the initial responses presented as line plots without error bars (traces from cells expressing LAMP1 and the CAX mutants overlap). (f) Pooled data from six transfections. (g and h) Effect of CAX knockdown on agonist-evoked Ca²⁺ signals. (g) Cytosolic Ca²⁺ levels of neural crest cells expressing the Ca²⁺ indicator R-GECO1 and stimulated with 1 mM carbachol. Explants were from control embryos and embryos injected with 10 ng AMO1. (h) Pooled data from five knockdowns quantifying the proportion of cells that responded to agonist. Error bars represent SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 5.

CAX regulates cell substrate attachment. (a and b) Effect of CAX knockdown on chemotaxis. (a) Tracks of individual explants migrating toward beads soaked in Sdf-1 over the entire period of recording (left) or the initial 100 min (right). (b) Summary data (from three knockdowns) quantifying forward motion index, persistence, and velocity of migration. (c–g) Effect of CAX knockdown on cell dispersion. (c) Summary data from three knockdowns quantifying the size of control and morphant explants 30 min after plating. (d) Representative color-coded triangulation diagrams at time 0 and 6 h from control or morphant explants. (e) Time courses (from three knockdowns) quantifying triangle area. Data are normalized to control explants at time 0. (f) Transmitted light micrographs showing spreading of control or morphant explants over a 20-min period. Panels on the right show an overlay of the area occupied by the explants where red represents the extent of spreading. Bar, 500 µm. (g) Summary data from three knockdowns quantifying spread area. (h and i) Effect of Ca²⁺ buffering on cell dispersion. Triangulation (h) and spreading (i) analysis of explants treated with cell-permeable Ca²⁺ chelators (50 µM BAPTA/EGTA-AM) or DMSO 30 min before imaging. (j and k) Live cell imaging of CAX. Stills from time-lapse confocal imaging of explants coexpressing mRFP-CAX and either membrane GFP (j; bar, 10 µm) or focal adhesion kinase–GFP (k). Arrowheads mark small CAX-positive vesicles. (k) Bars: (top) 4 µm; (bottom) 1 µm. Heat map summarizes the mean projection for CAX-positive vesicles relative to the centroid of focal adhesions (intersection of white lines) within 5 × 5–µm regions of interest. Data are from 28 focal adhesions from two expression experiments. (l–n) Effect of CAX knockdown on focal adhesions. (l) Expression of focal adhesion kinase–GFP (left) and immunocytochemistry analysis using an antibody to phosphopaxillin (middle) in control and morphant explants. Overlays (right) show costaining of phalloidin. Bar, 20 µm. (m) Summary data (94–422 cells from two to three knockdowns) quantifying the number and size of labeled structures. (n) Lifetime analysis of focal adhesion kinase–GFP in control and morphant explants (30–58 focal adhesions from two knockdowns). 10 ng AMO1 was used for all knockdowns. Error bars represent SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001.