ATP4a Is Required for Wnt-Dependent Foxj1 Expression ... - Cell Press

Cell Reports

Article ATP4a Is Required for Wnt-Dependent Foxj1 Expression and Leftward Flow in Xenopus Left-Right Development Peter Walentek,1 Tina Beyer,1 Thomas Thumberger,1,2 Axel Schweickert,1 and Martin Blum1,* 1University

of Hohenheim, Institute of Zoology, Garbenstrasse 30, 70593 Stuttgart, Germany address: Centre for Organismal Studies, Im Neuenheimer Feld 230, 69120 Heidelberg, Germany *Correspondence: [email protected] DOI 10.1016/j.celrep.2012.03.005 2Current

SUMMARY

Most vertebrate embryos break symmetry by a ciliadriven leftward flow during neurulation. In the frog Xenopus asymmetric expression of the ion pump ATP4a was reported at the 4-cell stage. The ‘‘ionflux’’ model postulates that symmetry is broken flow-independently through an ATP4-generated asymmetric voltage gradient that drives serotonin through gap junctions to one side of the embryo. Here, we show that ATP4a is symmetrically expressed. Gene knockdown or pharmacological inhibition compromised organ situs, asymmetric marker gene expression, and leftward flow. The gastrocoel roof plate (GRP), where flow in frog occurs, revealed fewer, shortened, and misaligned cilia. Foxj1, a master control gene of motile cilia, was downregulated in the superficial mesoderm, from which the GRP develops. Specifically, ATP4 was required for Wnt/b-catenin-regulated Foxj1 induction and Wnt/ PCP-dependent cilia polarization. Our work argues for evolutionary conservation of symmetry breakage in the vertebrates.

INTRODUCTION Lateral body plan asymmetry is found in animal phyla of both protostomes and deuterostomes (Levin and Palmer, 2007; Basu and Brueckner, 2008). In all cases, inner organs adopt an asymmetric position with respect to the dorsal midline (organ situs). Snails display shell chirality and left- or right-handed torsion of the visceral organs (Grande and Patel, 2009). In sea urchins, the adult develops from one side of the paired coelomic sacs of the bilateral symmetrical larva (Hibino et al., 2006). In vertebrates, many organs such as the heart, lung, and stomach are asymmetrically localized in the chest and abdomen (Hamada et al., 2002). Throughout the animal kingdom, an asymmetric gene cascade—consisting of the TGFß-type growth factor nodal, its feedback inhibitor lefty, and the homeobox transcription factor Pitx2—governs the correct establishment of organ 516 Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors

situs, demonstrating evolutionary conservation of the central pathway (Schweickert et al., 2012). The molecular mechanisms upstream of the nodal cascade, however, seem to vary between and within phyla (Tabin, 2006). The two models of symmetry breakage discussed at present approach the underlying problem, how in a bilateral symmetrical embryo gene induction occurs only on the left and not on the right side, in a completely different way (Levin and Palmer, 2007; Blum et al., 2009). In the cilia-based model of symmetry breakage, anterior-posterior cues are used during early neurulation to polarize cilia to the posterior pole of cells (Hirokawa et al., 2006). Clockwise rotation and the convex curvature of cell surfaces therefore result in leftward flow of extracellular fluids, which is what has been described in fish Kupffer’s vesicle (KV; Essner et al., 2005), amphibian gastrocoel roof plate (GRP; Schweickert et al., 2007), and mammalian node/posterior notochord (PNC; Nonaka et al., 1998). Flow induces the nodal cascade by downregulation of the nodal inhibitor Coco on the left side of the GRP (Schweickert et al., 2010). In this model, left-right (LR) axis specification is a consequence of the previously established anterior-posterior and dorsal-ventral polarities in an otherwise perfectly bilaterally symmetrical embryo (Schweickert et al., 2012). The second model assumes asymmetries upstream of the leftspecific induction of nodal and, in principle, requires a maternally derived LR-specific cue present in the zygote. Central to this hypothesis, the so-called ‘‘ion-flux’’ model (Levin and Palmer, 2007), are three components. (1) An asymmetric activity of the gastric proton-potassium ion pump ATP4 (also known as gastric H+/K+-ATPase) at the four-cell stage was proposed to set up an asymmetric voltage gradient (Levin et al., 2002). This gradient would cause the passage of (2) serotonin (Fukumoto et al., 2005) through (3) gap junctions (GJs; Esser et al., 2006) to become asymmetrically enriched in one or few blastomeres at the 32–64 cell stage (i.e., much earlier than flow). ATP4a and serotonin have both been reported to be asymmetrically localized in frog embryos (Levin et al., 2002; Fukumoto et al., 2005), and inhibition of GJ communication affects laterality in vertebrates (Levin and Mercola, 1998; 1999). In addition, ATP4 has been involved in the induction of the asymmetric nodal cascade in sea urchins and in chick as well (Levin et al., 2002; Hibino et al., 2006), where leftward flow and cilia involvement have so far not been reported, suggesting that ATP4 represents the central

component of an evolutionary conserved mechanism of symmetry breakage upstream of asymmetric nodal induction. We have recently reported that, contrary to a previous report, maternal serotonin in Xenopus embryos is symmetrically distributed with respect to the LR axis throughout early development. Specifically, serotonin signaling acts as a competence factor for Wnt in the specification of the superficial mesoderm (SM; Beyer et al., 2012a), from which the ciliated GRP derives (Shook et al., 2004). In addition we and others recently showed that GJ communication is required for the transfer of asymmetric cue(s) from the midline to the lateral plate mesoderm (LPM) in frog and mouse (Beyer et al., 2012b; Viotti et al., 2012). This, however, still leaves the formal possibility that other small and charged molecules could become asymmetrically distributed due to the action of ATP4. Here we show that symmetrically distributed ATP4 is required for leftward flow during neurulation, which renders the ‘‘ion-flux’’ model implausible and reinforces flow as the conserved mode of symmetry breakage in Xenopus. RESULTS Symmetric ATP4 Expression Regulates Laterality on the Dorsal Side of the Frog Embryo ATP4 is a member of the P-type superfamily of membranespanning ion pumps (Axelsen and Palmgren, 1998). In an electroneutral manner, it transports protons across membranes in exchange for potassium ions. ATP4 functions as a heterotetramer consisting of two catalytic alpha subunits (ATP4a) and two accessory beta subunits (ATP4b; Geering, 2001; Shin et al., 2009). As a first step to analyzing a potential ATP4 function for leftward flow, we reinvestigated the mRNA expression pattern during early development. Levin et al. (2002) reported consistent asymmetric expression of ATP4a in the ventral-right blastomere at the four-cell stage. We failed to detect any LR asymmetries in mRNA localization of ATP4a during cleavage stages (Figures 1A–1C). Maternally deposited mRNAs were detected throughout the cytoplasm of the animal hemisphere (Figures 1A–1C; data not shown). Zygotic transcripts became detectable from stage 43 onward in prospective gastric tissue (Figure 1D). A sense probe was negative (Figure 1E; data not shown). As Aw et al. (2008, p. 354) observed ‘‘significant variability of in situ signal in embryos from different females,’’ we analyzed a total of 320 embryos from the 2–32 cell stage. Embryos were derived from eight females from our own animal facility as well as two females from a colony in Heidelberg, Germany. In situ signals in vegetal hemispheres or LR asymmetries of any kind were not detected. Semiquantitative RT-PCR revealed that ATP4a mRNA levels decreased from cleavage through gastrula/neurula stages but were still detectable at stage 17 (Figure S1A available online). A parallel decrease of protein levels was not observed. Immunohistochemistry demonstrated ubiquitous ATP4a expression in all germ layers at gastrula and neurula stages. Signals were found at the plasma membrane and in vesicle-like structures in the cytoplasm (Figures S1B–S1D). Next we analyzed LR development in embryos in which ATP4 function was impaired. An antisense morpholino oligonucleotide (MO) was designed to interfere with the translational start site of

ATP4a mRNA. ATP4aMO or control MO (CoMO) was injected into the prospective marginal zone of ventral or dorsal blastomeres (VMZ and DMZ, respectively) at the four-cell stage (Figure S1E), and embryos were cultured to stage 32 or 45 for the assessment of Pitx2c expression or organ situs, respectively (Figures S1F and S1G). While VMZ or CoMO injections did not affect laterality, Pitx2c expression and organ situs were compromised in morphants injected into the DMZ region, i.e., the area from which the GRP is derived (Figures 1F and 1G; Figure S1E). More lateral injections into the C2 and C3 lineages reduced the effects on Pitx2c expression significantly (Figure 1F; Figure S1E), strongly suggesting that ATP4a was required in the dorsalmedial cell lineage. Assessment of organ placement at stage 45 was complicated by high mortality (data not shown) and by about one third of embryos developing fluid-filled cysts in the head and trunk, which prevented determination of situs (Figure S1G). Cysts indicated a malfunction of the pronephros (Wessely and Tran, 2011); indeed, mRNA expression of ATP1b1 revealed disturbed or absent pronephric tubules in morphants (Figure S3D). Pronephros malfunction likely reduced the viability of larvae at stages used for the determination of organ situs, offering an explanation why situs defects were encountered less frequently than alterations of Pitx2c expression patterns (Figures 1F and 1G). MO specificity was demonstrated by rescue experiments. Coinjection of both mRNA or DNA constructs together with ATP4aMO rescued Pitx2c expression (Figure 1F), further demonstrating that ATP4 was required for LR development following the onset of zygotic transcription at stages 8-9 (midblastula transition; MBT). The original study on ATP4 and LR by Levin et al. (2002) used a number of pharmacological inhibitors. To confirm our MO results, we tested the most efficient drug in the previous study, the imidazopyridine compound SCH28080 (Shin and Sachs, 2006). SCH28080 efficiently altered both Pitx2c mRNA expression patterns and organ situs (Figures 1F and 1G). Treated specimens frequently developed cysts (data not shown). In addition to the previous report, we show that SCH28080 disrupted LR development when specimens were treated starting from gastrula stages onward, i.e., SCH28080 interfered with laterality specification well beyond cleavage stages (Figure 1G). In summary, these results demonstrated that ATP4a was expressed in a LR symmetrical manner and required for laterality determination post-MBT and only on the dorsal side of the embryo. ATP4 Is Required for Leftward Flow In order to investigate whether ATP4 was required for cilia-driven leftward flow across the GRP, we prepared dorsal explants from wild-type and manipulated embryos, fluorescent beads were added, and flow was analyzed as described elsewhere (Schweickert et al., 2007; Vick et al., 2009). Robust leftward flow was observed in wild-type, CoMO-injected, and DMSOtreated specimens (Figures 2A and 2B; Movie S1). Directionality of flow was described using the dimensionless number rho (r). Rho was calculated from time-lapse movies and represents the mean resultant directionality of particle trails (Rayleigh’s test of uniformity). Rho values can range from 1, when all trajectories point in the same direction, to 0, when particles move randomly. Control flow reached r values of 0.76 ± 0.08, Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors 517

Figure 1. ATP4a Is Symmetrically Expressed and Required Post-MBT for LR Development (A–E) Whole-mount in situ hybridization of staged embryos at the (A) two-cell, (B) four-cell, and (C) 32-cell stages. Note that ATP4a mRNA was enriched in the animal cytoplasm, as shown in the bisected four-cell embryo in (B), viewed from the inner surface in (B0 ) and (B’’), but symmetrically expressed with respect to the LR axis. (D) Zygotic ATP4a expression in the stomach of a stage (st.) 45 tadpole; see plane of section in (D0 ) indicated by dashed line in (D). Arrowhead indicates gastric epithelium in blow-up. (E) Four-cell embryo stained with sense control RNA.

518 Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors

0.7 ± 0.16, and 0.84 ± 0.12 in untreated, CoMO-injected, and DMSO-treated specimens, respectively (Figures 2A, 2C, 2E, and S2). In contrast, flow was disturbed in ATP4a morphants and SCH28080-incubated specimens with r values of 0.45 ± 0.22 and 0.6 ± 0.21, respectively (Figures 2A, 2D, 2F, and S2; Movie S1). Aberrant flow in morphants was confirmed by analyzing flow velocity: the mean velocity of moving particles was reduced to 1.16 ± 0.93 mm/s in ATP4a morphants and 1.78 ± 0.69 mm/s in SCH28080-treated specimens, as compared to 2.5 ± 1.09 mm/s, 2.43 ± 0.95 mm/s, and 2.98 ± 0.78 mm/s in wild-type, CoMO-injected, and DMSO-treated explants, respectively (Figures 2B–2F). Flow thus was attenuated and nondirectional following interference with ATP4 function. Alterations of flow can result from absence of cilia, reduced cilia length, lack of posterior polarization, or lack of cilia motility (Nonaka et al., 1998; Maisonneuve et al., 2009; Vick et al., 2009; Beyer et al., 2012a). Time-lapse movies indicated that cilia in morphants and SCH28080-treated embryos were motile (Movie S1). Ciliation of GRP cells was investigated by scanning electron microscopy (SEM). The overall morphology of the GRP appeared normal in morphant specimens (Figures 2G and 2H; data not shown). The GRP was triangular in shape and consisted of small cells, as compared to the flanking endodermal cells (Blum et al., 2007; Figures 2G and 2H). Morphants, however, revealed fewer and shorter cilia, which were less frequently polarized to the posterior pole of cells (Figures 2G–2K), in good agreement with the observed flow distortions. Turbulent Flow in ATP4a Morphants Results in Bilateral Induction of the Nodal Cascade Impaired flow in ATP4a morphants and SCH28080-treated embryos resulted in predominantly bilateral expression of Pitx2c (Figure 1F; Levin et al., 2002). Bilateral induction of asymmetric genes in the LPM is thought to be generally accompanied by midline defects, i.e., disturbed expression of lefty/antivin in the floor plate, notochord, and dorsal endoderm (Lenhart et al., 2011). It has been proposed that lefty/antivin provides a barrier to prevent LPM nodal to cross from the left to the right side (Cheng et al., 2000). It is surprising that, in the case of ATP4a morphants, the embryonic midline was unaffected, as judged by lefty/antivin and Xbra mRNA expression (data not shown). We have previously shown that when flow was ablated in Xenopus embryos through inhibition of ciliary motility or by application of viscous media, nodal cascade induction was blocked, resulting in predominantly absent marker gene expression. Absent flow in these cases was characterized by less than 25 directed particles in time-lapse movies (Schweickert et al., 2007; Vick et al., 2009; Beyer et al., 2012a). Careful examination of flow in ATP4a morphants and upon SCH28080 treatment revealed that considerably more than 25 particles were directed beyond Brownian movement, though generally less than in control samples (Figure S2). Significantly, particles moved toward the left as well as to the right side of the GRP (Figures 3A–3C; Movie

S1). These data suggest that attenuated and nondirectional fluid flow was sufficient to induce the nodal cascade on either side of the GRP, independent of midline barrier defects. To further elucidate whether such a direct and midlineindependent mechanism of right-sided nodal cascade induction was active in ATP4-manipulated embryos, we analyzed molecular targets downstream of flow and upstream of LPM Pitx2c. As a direct response to flow, the nodal inhibitor Coco becomes downregulated on the left side of the GRP, resulting in left-sided derepression of Xnr1 at postflow stages (Vonica and Brivanlou, 2007; Schweickert et al., 2010). Preflow, Coco is coexpressed with Xnr1 in the lateralmost cells of the GRP, which sense flow (Figure 3D). In ATP4a morphants, left-sided downregulation of Coco was not observed, i.e., equal staining intensities of Coco were encountered on both sides of the GRP (Figures 3D and 3E; Schweickert et al., 2007). In addition GRP-Xnr1 expression was unaffected in morphants (Figure 3D). Loss of asymmetric Coco expression was previously reported as a consequence of flow ablation, resulting in a lack of left-sided Xnr1 derepression (Schweickert et al., 2010). Bilateral expression of Pitx2c, however, implies that Xnr1 derepression occurred bilaterally, i.e., that equal Coco signals represent equal downregulation and bilateral Xnr1 derepression on either side of the GRP. In order to test this hypothesis, epistatic loss-of-function experiments were performed. When Xnr1 was knocked down in the GRP by MO injection, nodal cascade induction in the LPM was inhibited, as previously reported (Schweickert et al., 2010; Figure 3F1). Parallel knockdown of ATP4a led to the same result (Figure 3F2), demonstrating that ATP4 acted upstream of GRP-Xnr1. Xnr1 knockdown exclusively on the left side of the GRP prevented Pitx2c expression in the left LPM (Figure 3F3). Additional bilateral knockdown of ATP4a, however, resulted in >50% of embryos displaying right-asymmetric expression of Pitx2c (Figure 3F4). Thus, attenuation of ATP4— even in the absence of normal left-sided activation of the nodal cascade—resulted in right-only induction of Pitx2c expression (Figure 3F4). This was true for the left side as well, as Xnr1 knockdown on the right side of the GRP did not interfere with wild-type Pitx2c expression (Figure 3F5) and rescued expression patterns in ATP4a morphants (Figure 3F6). The observed reduction in cilia length and bead velocity made us wonder whether cilia motility was required for the right- and left-sided induction of Pitx2c in ATP4a morphants. To address this question, ciliary motility was inhibited by MO-mediated knockdown of dynein motor protein dnah9 (dnah9-SB-MO; Vick et al., 2009). As shown in Figure 3F7, loss of ciliary motility in parallel to ATP4a knockdown indeed switched Pitx2c expression patterns from bilateral to absent. We previously showed that flow was only required on the left side of the GRP, i.e., right-sided knockdown of dnah9 resulted in wild-type Pitx2c expression (Vick et al., 2009). Parallel knockdown of ATP4a throughout the GRP and of ciliary motility exclusively on the right or left side prevented the induction of Pitx2c on the very side where ciliary motility was ablated (Figures

(F and G) (F) Pitx2c expression and (G) organ situs in wild-type embryos at tadpole stages (26–32) and in specimens injected or treated as specified. The image in (D0 ) was reconstructed from several images taken from the same histological section. **Highly significant. ***Very highly significant. ns, not significant; n, number of embryos analyzed; wt, wild-type. See also Figure S1.

Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors 519

Figure 2. Impaired Ciliation and Leftward Flow upon ATP4 Loss of Function (A–F) Flow analysis. (A) Directionality and (B) velocity of fluorescent beads added to GRP explants at stage 17 were drastically reduced in ATP4a morphants or SCH28080-treated specimens, as compared to wild-type, CoMO-injected, or DMSO-treated embryos. n represents number of explants analyzed. (C–F) Frequency distribution of trajectory angles in representative explants injected with (C) CoMO, (D) ATP4aMO, (E) DMSO, and (F) SCH28080. Dashed circles


3F8 and 3F9). When a triple knockdown was performed such that ATP4 was absent from the entire GRP (bilateral Pitx2c when applied alone), Xnr1 was unilaterally deleted on the left side (absence of left-sided Pitx2c) and ciliary motility by dnah9MO was exclusively on the right side, Pitx2c was no longer induced on either side (Figure 3F10). These results strongly suggest that residual ciliary length and motility in ATP4a morphants were sufficient to induce the LPM nodal cascade on either side of the GRP. This conclusion gained further support from analyzing Coco expression in flow and ATP4 compromised specimens: Left-sided ablation of ciliary motility in ATP4a morphants led to an expansion of the left-sided Coco domain, demonstrating that the residual ciliary motility in morphants was sufficient to result in a repression of Coco expression on the right side of the GRP (Figures 3D and 3E). In summary, these experiments show that bilateral induction of the nodal cascade can occur in the absence of midline defects through an overall attenuated and turbulent flow. ATP4 Is Required for Both Canonical and Noncanonical Wnt Signaling Next, we wondered how ATP4 might impact on ciliation and cilia polarization in the GRP. Examination of ATP4a morphant tadpoles at stage 45 revealed phenotypes that were reminiscent of well-known Wnt pathway deficiencies (Figure 4A): Morphants displayed shortened AP axes (Figure S3B); small eyes and reduced pigmentation; and anterior head defects and malfunction of the embryonic kidney, the pronephros (Figure S3D), resulting in cyst formation (Tao et al., 2005; Te´telin and Jones, 2010; Figure S1G). Polarization defects of GRP cilia (cf. Figures 2H and 2K) likely represent Wnt defects as well, as it has recently been shown that noncanonical Wnt signaling controls posterior polarization of LR cilia in mouse and frog (Maisonneuve et al., 2009; Antic et al., 2010; Song et al., 2010). A link between Wnt signaling and another proton pump, the vacuolar H+-ATPase (ATP6), has been recently established, in which acidification of Wnt signalosomes was shown to be crucial for both canonical and noncanonical Wnt signaling (Cruciat et al., 2010; Hermle et al., 2010). Canonical and noncanonical Wnt target genes and processes were therefore analyzed to investigate a more general role of ATP4 in Wnt pathway activation. To assess canonical Wnt signaling, we investigated expression of the target gene En2 (McGrew et al., 1999) and double axis induction (Sokol et al., 1991). En2 was reduced at the mid-/hindbrain boundary of ATP4a morphants (Figure 4B). Expression was rescued to wild-type levels upon coinjection of a DNA expression construct encoding b-catenin (b-cat; data

not shown). Secondary axis formation was induced in 80%– 90% of specimens injected into ventral blastomeres at the four-cell stage with mRNAs encoding Xwnt8a, Dvl2, or b-cat (Figure 4C). Twinning was significantly reduced when ATP4aMO was coinjected with either Xwnt8a or Dvl2; however, b-cat-induced secondary axis formation was unaffected (Figure 4C), indicating an interaction of ATP4 with Wnt signaling at the level of the plasma membrane, as described for ATP6 (Cruciat et al., 2010). A function of ATP4 in noncanonical Wnt signaling was analyzed in animal cap explant cultures, a widely used assay for planar cell polarity (PCP)-mediated convergent extension (CE) in the frog Xenopus (Wang et al., 2008; Buechling et al., 2010). Elongation of activin-treated animal cap explants (Green and Smith, 1990) was greatly reduced when ATP4aMO was injected into the animal region of blastomeres at the four-cell stage and rescued upon coinjection of constitutively active RhoA mRNA (Paterson et al., 1990), a downstream target of noncanonical Wnt signaling (Song et al., 2010; Figure 4D). A dominant-negative RhoA construct did not rescue CE (data not shown). Four additional lines of evidence supported the notion that ATP4 acted on noncanonical Wnt signaling as well (Figure S3). Neural tube closure was affected in morphants as was elongation of the AP axis (Figures S3A and S3B). Both phenotypes were partially rescued upon coinjection of ATP4a mRNA (Figures S3A and S3B). In addition, Twist1-positive neural crest cells failed to migrate in morphants (Figure S3C) and fewer premigratory cells were specified, as judged by Twist1 signal intensity (Alfandari et al., 2010; Ossipova and Sokol, 2011; Figure S3C). Finally, formation of pronephric tubules was compromised in morphants, a defect ascribed to noncanonical Wnt signaling as well (Zhang et al., 2011). Taken together, these experiments suggest that ATP4 acts in both canonical and noncanonical Wnt pathway activation upstream of the intracellular effectors b-cat and RhoA. ATP4 Is Required Post-MBT for Canonical WntMediated Induction of Foxj1 and Wnt/PCP-Dependent Cilia Polarization Reduction of cilia length and number of ciliated GRP cells (Figures 2I and 2J) indicated that the activity of Foxj1, the master control gene of motile cilia (Tamakoshi et al., 2006; Stubbs et al., 2008), was affected in ATP4a morphants. Foxj1 is first transcribed in the SM of gastrula stage embryos, which is situated on top of the dorsal lip and will form the GRP following involution (Pohl and Kno¨chel, 2004; Stubbs et al., 2008; Beyer et al., 2012a; Figure 5A). Foxj1 expression in ATP4a morphants or SCH28080treated specimens was markedly reduced (Figures 5A and 5D),

indicate maximum frequency in histogram specified in percent. a, anterior; l, left; n, number of particles above threshold; p, posterior; r, right; v, average velocity of particles; r, quality of flow. (G–K) SEM analysis of GRP ciliation and morphology. (G and H) Representative dorsal explants reveal shorter cilia, fewer ciliated cells, and polarization defects in (H) ATP4a morphant (H) as compared to (G) CoMO-injected specimen. Cell boundaries are indicated by dashed orange lines in higher magnification of SEM pictures in (G0 ) and (H0 ). Blowups in (G0 ) and (H0 ) illustrate a long and posteriorly polarized cilium in (G0 ), indicated by a green arrowhead, and two short cilia in (H0 ), of which one emerges in a central position, indicated by a yellow arrowhead. (G’’ and H’’) Evaluation of cilia polarization. Green, posterior; yellow, other; red, no cilium. (I–K) Quantification of (I) ciliation rate and (J) cilia length of GRP cilia in defined areas, as indicated by white squares in (G) and (H). (K) Cilia polarization was assessed in areas of defined size, as indicated by dashed boxes in (G) and (H). **Highly significant. ***Very highly significant. ns, not significant. Numbers indicate number of dorsal explants (in parentheses) or cilia (in brackets) analyzed. See also Figure S2 and Movie S1.


Figure 3. Turbulent Flow in ATP4a Morphants Causes Bilateral Nodal Cascade Induction (A–C) Analysis of bead trajectories in time-lapse movies of dorsal explants from representative (A) CoMO-injected or (B) ATP4aMO-injected embryos and (C) specimen treated with SCH28080. Flow is displayed as GTTs of 25 s length; cf. color bar in (A). Note that trajectories in (B) and (C) project to the left side (indicated with blue arrows) and right side (indicated with pink arrows) of the GRP, whereas GTTs in (A) point uniformly to the left. White arrows represent trajectories running anteriorly or posteriorly. (D) Whole-mount in situ hybridization of stage 20 dorsal explants with probes specific for (top) Coco and (bottom) Xnr1. Left-sided Coco repression was lost in (middle) ATP4a morphants and inverted upon (right) parallel left-sided knockdown of flow. Xnr1 expression was unaffected. (E) Quantification of Coco expression patterns. (F) Quantification of Pitx2c expression patterns in stage 26–32 tadpoles following MO injections into the C1 lineage (dorsal midline-GRP) of four-cell embryos as indicated. Note that the bilateral induction (cf. Figure 1F) in ATP4a morphants was dependent on both the presence of GRP-Xnr1 and ciliary motility. See also Movie S1.

without affecting organizer function or dorsal mesoderm induction, as judged by mRNA expression of Gsc, Xnr3, Xbra, and Not (data not shown). The ATP4-Wnt link established above made us wonder whether Foxj1 was under Wnt control as well. 522 Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors

Indeed, coinjection of a b-cat DNA expression construct rescued Foxj1 expression in the SM of morphants (Figures 5B and 5D). Significantly, ectopic expression of DNA constructs encoding b-cat or a constitutively active form of the Wnt co-receptor

LRP6 (LRP6DE1-4; data not shown) into the ventral marginal zone of four-cell embryos resulted in an expansion of the Foxj1 expression domain to reach all around the blastopore (Figure 5C). These data clearly showed that Foxj1 transcription in the SM was under the control of canonical Wnt signaling post-MBT, which in turn relied on ATP4. To confirm the functional relevance of ATP4/ Wnt-dependent Foxj1 induction for LR axis formation, a Foxj1 DNA expression construct (Stubbs et al., 2008) was coinjected with ATP4aMO into the prospective GRP lineage at the fourcell stage, and tadpoles were analyzed for Pitx2c expression. In these experiments, asymmetric Pitx2c transcription was partially rescued (Figure 5E), demonstrating that ATP4/Wnt-dependent Foxj1 expression in the SM was required for specification of the LR axis. The involvement of Foxj1 in ATP4-mediated LR axis specification afforded the opportunity of testing whether the ATP4 effects on canonical and noncanonical Wnt signaling could be experimentally separated. To that end, Foxj1 was assayed for its potential to rescue ATP4a-induced ciliation phenotypes. As depicted in Figure S4, Foxj1 rescued the ciliation rate in morphant GRPs, while cilia polarization defects did not improve. This experiment demonstrates that ATP4a was required for both canonical and noncanonical Wnt signaling upstream of ciliation and cilia polarization. In summary, our analysis of ATP4a in LR axis formation in Xenopus revealed a symmetric function in Wnt/b-cat-regulated Foxj1 induction and Wnt/PCPdependent cilia polarization, i.e., in the setup of leftward flow during neurulation. DISCUSSION Pharmacological inhibitors have implicated ATP4 in LR development throughout the chordates (Levin et al., 2002; Hibino et al., 2006; Shimeld and Levin, 2006; Gros et al., 2009). Our present work in the frog Xenopus demonstrates that ATP4 acts on leftward flow during early neurula stages. ATP4 is involved in two distinct steps, both of which occur post-MBT: (1) canonical Wnt signaling-dependent induction of Foxj1 transcription in the SM of the gastrula embryo, which forms the GRP following involution over the dorsal lip of the blastopore (canonical Wntdependent Foxj1 expression in the zebrafish KV has very recently been reported as well [Caron et al., 2012], suggesting evolutionary conservation of mechanisms); and (2) posterior polarization of cilia in the GRP proper, which was recently shown to depend on noncanonical Wnt/PCP. Our detection of polarization defects probably owes to the only partial knockdown of Foxj1 in ATP4a morphants, as previous work has shown that GRP cilia were more severely affected upon Foxj1 knockdown (Stubbs et al., 2008). Flow in ATP4a morphants was much reduced in strength and velocity, and particles moved in all directions. Unlike in other cases when flow was absent or severely hampered and the nodal cascade was not induced at all (Schweickert et al., 2007; Vick et al., 2009; Beyer et al., 2012a; also in Foxj1 morphants, data not shown), this residual and aberrant flow unexpectedly resulted in predominantly bilateral expression of Pitx2c in the left and right LPM. In accordance with this pattern, we found particles moving toward the left and right margins of the GRP upon

Figure 4. ATP4 Is Required for Canonical and Noncanonical Wnt Signaling (A) Phenotypes of ATP4a morphant tadpoles, which were less pigmented, displayed shortened AP axes, small heads, and reduced eyes. (B and C) Canonical Wnt signaling. (B) Reduced En2 expression at the midhindbrain boundary of (bottom) morphant stage 26 tadpole. (C) Xwnt8- and Dvl2-mediated, but not b-cat-induced, twinning requires ATP4a. (D) Noncanonical Wnt signaling. Convergent-extension movements of activininduced animal cap explants: Reduced elongation in ATP4aMO-injected explants was rescued by coinjection of constitutively active (CA) RhoA. ***Very highly significant. ns, not significant. See also Figure S3.

ATP4 loss of function. It is interesting that the embryonic midline, floor plate, and notochord were present and that lefty/antivin was expressed normally in ATP4a morphants. This is remarkable and, to our knowledge, the first case in which bilateral expression of left-asymmetric genes was found in the absence of midline barrier (i.e., lefty/antivin) defects. It might thus be worth re-evaluating mouse and fish mutants with bilateral marker gene expression for aberrant flow. Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors 523

Figure 5. ATP4 Is Required for Wnt/b-cat-Mediated Induction of Foxj1 During Gastrulation (A–C). Reduced mRNA expression of (A) Foxj1 in the SM of ATP4a-morphant or SCH28080-treated embryos was rescued upon coinjection of (B) a b-cat DNA expression construct. (C) Ectopic expression of Foxj1 on the ventral lip following injection of a b-cat DNA expression construct into ventral blastomeres at the four-cell stage (VMZ lineage). (D) Quantification of results. (E) Partial rescue of Pitx2c expression in ATP4a morphants upon coinjection of a Foxj1 DNA expression construct. See also Figure S4.

In any case, ATP4 impacts on Wnt signaling upstream of the branching into the canonical and noncanonical pathway. It is tempting to speculate that ATP4 is required for acidification of Wnt signalosomes and thus functions in much the same way as shown for ATP6 (Cruciat et al., 2010; Hermle et al., 2010). The localization of ATP4a in vesicle-like structures supports this notion. It is interesting that, when ATP6 was pharmacologically inhibited in Xenopus, chick, and zebrafish, LR defects ensued as well, including ciliation defects in the zebrafish KV (Adams et al., 2006). LR defects were also reported when the potassium channels Kir and KCNQ, which supply the counter ions for ATP4-dependent acidification across membranes, were targeted (Adams et al., 2006; Aw et al., 2008; Morokuma et al., 2008). All four channels, ATP4, ATP6, Kir, and KNCQ were implemented in the ‘‘ion-flux’’ model of symmetry breakage, which was proposed to act during very early cleavage stages. In this scheme, symmetry is broken through asymmetric passage of serotonin through GJs along a voltage gradient established by asymmetric activity of the ion pump ATP4 (Levin et al., 2002; Levin and Palmer, 2007). Our present work unequivocally demonstrates that ATP4 action on LR development happens at a later stage, post-MBT: (1) ATP4a morphants were rescued using DNA expression constructs, which only become transcribed after onset of zygotic transcription; (2) LR defects occurred following inhibition of ATP4 by SCH28080 524 Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors

treatment starting at gastrulation; (3) we were unable to reproduce the reported asymmetry of maternal ATP4a mRNA transcripts in the ventral right blastomere of the four-cell embryo (Levin et al., 2002), which constitutes the initial observation and the base of the ‘‘ion-flux’’ model. The validity of this model in the frog Xenopus has been questioned by our recent work on serotonin (Beyer et al., 2012a), which showed that serotonin was symmetrically distributed throughout cleavage and that serotonin signaling acted as a competence factor for Wnt signaling in the specification of the SM. Thus, both serotonin and ATP4 impact on Wnt signaling, Foxj1 expression in the SM, and flow at the GRP. Despite these common features, a number of distinct characteristics can be ascribed to these two LR factors as well: (1) While ATP4 and serotonin acted on both canonical and noncanonical Wnt signaling, early targets of maternal Wnt signaling (e.g., Xnr3) were only affected upon loss of serotonin signaling (Beyer et al., 2012a; data not shown); (2) flow was absent upon interference with serotonin signaling but randomized in ATP4a morphants, resulting in absent and bilateral LPM marker gene expression, respectively. Residual flow in ATP4a morphants likely is due to the only partial repression of Foxj1 in the SM, which is more pronounced upon serotonin loss of function; (3) the gross morphology of GRP cells was altered upon loss of serotonin signaling (altered cell size in addition to ciliation defects), while the GRP appeared normal in

shape and cellular morphology in ATP4a morphants. These characteristics suggest that serotonin signaling and ATP4 have common and distinct functions in Wnt signaling and specification of the SM/GRP. The third crucial component of the ‘‘ion-flux’’ model, gap junctional communication (Levin and Mercola, 1998, 1999), has very recently been shown to be involved in the transfer of the asymmetric cue(s) from the GRP to the LPM both in frog and mouse (Beyer et al., 2012b; Viotti et al., 2012). Together with our work on serotonin and ATP4, these data render the ‘‘ion-flux’’ scheme of symmetry breakage highly unlikely in the frog Xenopus. There are, however, a number of vertebrate and nonvertebrate chordates with a clearly conserved nodal cascade that lack leftward flow (Tabin, 2006; Gros et al., 2009). This is particularly true for the chick, in which molecular asymmetry was first described in 1995 (Levin et al., 1995) and which has been studied extensively in many laboratories, but holds for pig embryos as well, where no cilia have been found in the node/PNC (Gros et al., 2009), i.e., the equivalent of the GRP in the frog or the KV in fish. In chick and pig, the node/organizer displays morphological asymmetries, and it was recently reported that this asymmetry is generated by asymmetric cell movements during gastrulation in chick (Gros et al., 2009). Remarkably, this asymmetric cell migration is dependent on ATP4, as SCH28080 prevented migration (Gros et al., 2009). In the tunicate Ciona and in sea urchins, SCH28080 treatments of embryos have shown that molecular asymmetries are under control of ATP4 as well, strongly suggesting that ATP4 function is as conserved as the nodal cascade in deuterostome LR axis specification (Hibino et al., 2006; Shimeld and Levin, 2006). It should be noted that studies in the chick have implicated an ATP4 function on symmetry breakage during gastrulation (Levin et al., 2002), which is in perfect agreement with our study in Xenopus. How then could one reconcile a conserved ATP4 function during gastrulation and conserved nodal cascade induction in the left LPM with absence of leftward flow in birds and some (at least one) mammals? Evolutionary considerations led us to propose that symmetry breakage in the chordates should be conserved throughout (Blum et al., 2009). The common denominator should be Wnt signaling upstream of the nodal cascade. It is important to note that a serotonin function has been documented in chick LR development as well (Fukumoto et al., 2005), integrating this factor into the conserved LR tool box. The same holds true for ATP6 (Adams et al., 2006), and expression patterns of Wnt ligands fit to this proposal as well (e.g., Hardy et al., 2008). One possible scenario therefore could be that, in species that lack flow, asymmetric cell migration at the node/organizer are induced under the control of Wnt/PCP (Zhang and Levin, 2009). This scenario would, however, not solve the question of biased symmetry breakage. We like to suggest a different and simpler solution, namely that motile and sensory cilia do play a role even in species in which flow was not detected so far. For quite some time, cilia and flow went undetected in the frog Xenopus, mostly because of difficulties in detecting the ciliated tissue. This is a general problem in the vertebrates, as these tissues (GRP, KV, PNC/node) differ considerably in shape and dimension (Blum et al., 2007). They share as a common characteristic their only very transient existence, which complicates

matters further. Our present work strongly indicates that there is no need for a vigorous and strong cilia-driven flow in order to induce the nodal cascade. The residual flow in ATP4a morphants was sufficient to trigger the cascade. A recent paper by Hamada and colleagues has reported the same basic finding in mutant mouse embryos, in which as few as two cilia were sufficient for nodal cascade induction (Shinohara et al., 2012). In an effort to study the conserved LR tool box in the chick, we have recently detected Foxj1 expression prior to asymmetric cell migration (P.W., Simone Geyer, and M.B., unpublished data). As Foxj1 is linked to motile cilia in the chick as well (Cruz et al., 2010), it may be conceivable that an ATP4/Foxj1-regulated ciliary function controls symmetry breakage upstream of asymmetric cell migration. Further experiments will show to what extent variations of the common theme do exist in vertebrates, chordates, and beyond. EXPERIMENTAL PROCEDURES Statistical Evaluation of Results Statistical evaluation of experiments represented by bar graphs were performed using chi-square tests (http://www.physics.csbsju.edu/stats/ contingency.html). Statistics of experiments represented by box plots were calculated by Wilcoxon sum of ranks (Mann-Whitney) tests (http://www.fon. hum.uva.nl/Service/Statistics/Wilcoxon_Test.html). Manipulation of Embryos Embryos were injected at the two- to eight-cell stage using a Harvard Apparatus setup in 13 modified Barth’s solution (MBSH) with 4% Ficoll (BioChemica) and transferred to 0.13 MBSH 15 min after injection. Drop size was calibrated to about 7–8 nl per injection. Rhodamine-B or Cascade Blue dextran (0.5–1.0 mg/ml; Molecular Probes) was coinjected and used as lineage tracer. Morpholino concentrations of ATP4aMO (50 -GTCATATTGTTCCTTTTTCCC CATC-30 ) or CoMO (random control oligo; Gene Tools) used in cases not specified in the figure legends were as follows: 23 0.5 pmol (Figure 1G; Figures S1F and S1G; Figures 2, 3A, 3B, 3D, and 3E; Figure S4; Figures 5A and 5E); 13 1 pmol (Figures S3C and S3D); 23 1 pmol (Figures 4A and 4B; Figures S3A and S3B); 43 1 pmol (Figure 4D). mRNAs were prepared using the Ambion message machine kit and diluted to the following concentrations: 30 ng/ml (RhoA-CA; RhoA V14 in (Paterson et al., 1990), 50 or 80 ng/ml (ATP4a), 80 ng/ml (Dvl2; Sokol, 1996), 100 ng/ml (Xwnt8a [Sokol et al., 1991] and b-catGFP [Miller and Moon, 1997]). DNAs were purified using the PureYield Plasmid Midiprep kit (Promega) and diluted to a concentration of 0.5 ng/ml (Foxj1CS2+), 1 ng/ml (ATP4a-CS2+MT and b-cat-GFP-CS2+). SCH28080 (SigmaAldrich) was dissolved in DMSO (AppliChem) and used at concentrations of 100–200 mM as indicated. DMSO without dissolved SCH28080 served as control. Flow and GRP Analysis Embryos were coinjected with lineage tracer in MO experiments to control for correct targeting of the GRP. Data processing was as described elsewhere (Schweickert et al., 2007; Vick et al., 2009; Beyer et al., 2012a). The whiskers of the box plots extend to maximal 1.53 IQR, and outliers are displayed as circles. To determine cilia parameters of GRP cells, a comparable area at the center of the GRP was selected in SEM pictures (magnification, 500fold) for manual analysis of cilia number, length (Figures 2G and 2H, gray box), and polarization (Figures 2G and 2H, dotted black box), which were analyzed in ImageJ (Beyer et al., 2012a). The whiskers of the box plots extend to maximal 1.53 IQR, outliers are displayed as circles. Animal Cap Explant Culture Animal cap assays were performed according to Green and Smith (1990). All cells of the four-cell embryo were injected into the animal pole, and animal caps were cut at stage 9. Recombinant human Activin A (R&D Systems) was


added immediately after dissection, and embryos were cultured until control specimens reached stage 22–30.

Buechling, T., Bartscherer, K., Ohkawara, B., Chaudhary, V., Spirohn, K., Niehrs, C., and Boutros, M. (2010). Wnt/Frizzled signaling requires dPRR, the Drosophila homolog of the prorenin receptor. Curr. Biol. 20, 1263–1268.

SUPPLEMENTAL INFORMATION

Caron, A., Xu, X., and Lin, X. (2012). Wnt/b-catenin signaling directly regulates Foxj1 expression and ciliogenesis in zebrafish Kupffer’s vesicle. Development 139, 514–524.

Supplemental Information includes Extended Experimental Procedures four figures, and one movie and can be found with this article online at doi:10.1016/j.celrep.2012.03.005. LICENSING INFORMATION This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 Unported License (CC-BY-NC-ND; http://creativecommons.org/licenses/by-nc-nd/3. 0/legalcode). ACKNOWLEDGMENTS Valuable discussions with colleagues J. Christian, A. Fainsod, C. Kintner, C. Niehrs, and O. Wessely throughout this project are gratefully acknowledged. K. Geehring, R. Harland, C. Kintner, M. Levin, R. Moon, C. Niehrs, H. Steinbeisser, and O. Wessely provided reagents. I. Schneider, B. Ulmer, and S. Bogusch helped with some of the experiments. Work in the Blum lab was supported by Deutsche Forschungsgemeinschaft (DFG) Grant BL285/ 9-1. P.W., T.B., and T.T. were recipients of PhD fellowships from the Landesgraduiertenfo¨rderung Baden-Wu¨rttemberg. Received: December 21, 2011 Revised: March 6, 2012 Accepted: March 21, 2012 Published online: April 19, 2012

Chen, P.X., Mathews, P.M., Good, P.J., Rossier, B.C., and Geering, K. (1998). Unusual degradation of alpha-beta complexes in Xenopus oocytes by betasubunits of Xenopus gastric H-K-ATPase. Am. J. Physiol. 275, C139–C145. Cheng, A.M., Thisse, B., Thisse, C., and Wright, C.V. (2000). The lefty-related factor Xatv acts as a feedback inhibitor of nodal signaling in mesoderm induction and L-R axis development in xenopus. Development 127, 1049–1061. Cruciat, C.-M., Ohkawara, B., Acebron, S.P., Karaulanov, E., Reinhard, C., Ingelfinger, D., Boutros, M., and Niehrs, C. (2010). Requirement of prorenin receptor and vacuolar H+-ATPase-mediated acidification for Wnt signaling. Science 327, 459–463. Cruz, C., Ribes, V., Kutejova, E., Cayuso, J., Lawson, V., Norris, D., Stevens, J., Davey, M., Blight, K., Bangs, F., et al. (2010). Foxj1 regulates floor plate cilia architecture and modifies the response of cells to sonic hedgehog signalling. Development 137, 4271–4282. Esser, A.T., Smith, K.C., Weaver, J.C., and Levin, M. (2006). Mathematical model of morphogen electrophoresis through gap junctions. Dev. Dyn. 235, 2144–2159. Essner, J.J., Amack, J.D., Nyholm, M.K., Harris, E.B., and Yost, H.J. (2005). Kupffer’s vesicle is a ciliated organ of asymmetry in the zebrafish embryo that initiates left-right development of the brain, heart and gut. Development 132, 1247–1260. Fukumoto, T., Kema, I.P., and Levin, M. (2005). Serotonin signaling is a very early step in patterning of the left-right axis in chick and frog embryos. Curr. Biol. 15, 794–803.

REFERENCES

Geering, K. (2001). The functional role of beta subunits in oligomeric P-type ATPases. J. Bioenerg. Biomembr. 33, 425–438.

Adams, D.S., Robinson, K.R., Fukumoto, T., Yuan, S., Albertson, R.C., Yelick, P., Kuo, L., McSweeney, M., and Levin, M. (2006). Early, H+-V-ATPasedependent proton flux is necessary for consistent left-right patterning of non-mammalian vertebrates. Development 133, 1657–1671.

Grande, C., and Patel, N.H. (2009). Nodal signalling is involved in left-right asymmetry in snails. Nature 457, 1007–1011.

Alfandari, D., Cousin, H., and Marsden, M. (2010). Mechanism of Xenopus cranial neural crest cell migration. Cell Adhes. Migr. 4, 553–560. Antic, D., Stubbs, J.L., Suyama, K., Kintner, C., Scott, M.P., and Axelrod, J.D. (2010). Planar cell polarity enables posterior localization of nodal cilia and leftright axis determination during mouse and Xenopus embryogenesis. PLoS ONE 5, e8999. Aw, S., Adams, D.S., Qiu, D., and Levin, M. (2008). H,K-ATPase protein localization and Kir4.1 function reveal concordance of three axes during early determination of left-right asymmetry. Mech. Dev. 125, 353–372. Axelsen, K.B., and Palmgren, M.G. (1998). Evolution of substrate specificities in the P-type ATPase superfamily. J. Mol. Evol. 46, 84–101. Basu, B., and Brueckner, M. (2008). Cilia multifunctional organelles at the center of vertebrate left-right asymmetry. Curr. Top. Dev. Biol. 85, 151–174. Beyer, T., Danilchik, M., Thumberger, T., Vick, P., Tisler, M., Schneider, I., Bogusch, S., Andre, P., Ulmer, B., Walentek, P., et al. (2012a). Serotonin signaling is required for Wnt-dependent GRP specification and leftward flow in Xenopus. Curr. Biol. 22, 33–39. Beyer, T., Thumberger, T., Schweickert, A., and Blum, M. (2012b). Connexin26-mediated transfer of laterality cues in Xenopus. Biology Open, in press. Published online April 3, 2012. 10.1242/bio.2012760.

Green, J.B., and Smith, J.C. (1990). Graded changes in dose of a Xenopus activin A homologue elicit stepwise transitions in embryonic cell fate. Nature 347, 391–394. Gros, J., Feistel, K., Viebahn, C., Blum, M., and Tabin, C.J. (2009). Cell movements at Hensen’s node establish left/right asymmetric gene expression in the chick. Science 324, 941–944. Hamada, H., Meno, C., Watanabe, D., and Saijoh, Y. (2002). Establishment of vertebrate left-right asymmetry. Nat. Rev. Genet. 3, 103–113. Hardy, K.M., Garriock, R.J., Yatskievych, T.A., D’Agostino, S.L., Antin, P.B., and Krieg, P.A. (2008). Non-canonical Wnt signaling through Wnt5a/b and a novel Wnt11 gene, Wnt11b, regulates cell migration during avian gastrulation. Dev. Biol. 320, 391–401. Harland, R.M. (1991). In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol. 36, 685–695. Hermle, T., Saltukoglu, D., Gru¨newald, J., Walz, G., and Simons, M. (2010). Regulation of Frizzled-dependent planar polarity signaling by a V-ATPase subunit. Curr. Biol. 20, 1269–1276. Hibino, T., Ishii, Y., Levin, M., and Nishino, A. (2006). Ion flow regulates leftright asymmetry in sea urchin development. Dev. Genes Evol. 216, 265–276. Hirokawa, N., Tanaka, Y., Okada, Y., and Takeda, S. (2006). Nodal flow and the generation of left-right asymmetry. Cell 125, 33–45.

Blum, M., Andre, P., Muders, K., Schweickert, A., Fischer, A., Bitzer, E., Bogusch, S., Beyer, T., van Straaten, H.W.M., and Viebahn, C. (2007). Ciliation and gene expression distinguish between node and posterior notochord in the mammalian embryo. Differentiation 75, 133–146.

Lenhart, K.F., Lin, S.-Y., Titus, T.A., Postlethwait, J.H., and Burdine, R.D. (2011). Two additional midline barriers function with midline lefty1 expression to maintain asymmetric Nodal signaling during left-right axis specification in zebrafish. Development 138, 4405–4410.

Blum, M., Weber, T., Beyer, T., and Vick, P. (2009). Evolution of leftward flow. Semin. Cell Dev. Biol. 20, 464–471.

Levin, M., and Mercola, M. (1998). Gap junctions are involved in the early generation of left-right asymmetry. Dev. Biol. 203, 90–105.


Levin, M., and Mercola, M. (1999). Gap junction-mediated transfer of left-right patterning signals in the early chick blastoderm is upstream of Shh asymmetry in the node. Development 126, 4703–4714. Levin, M., and Palmer, A.R. (2007). Left-right patterning from the inside out: widespread evidence for intracellular control. Bioessays 29, 271–287.

Shin, J.M., and Sachs, G. (2006). Gastric H,K-ATPase as a drug target. Dig. Dis. Sci. 51, 823–833. Shin, J.M., Munson, K., Vagin, O., and Sachs, G. (2009). The gastric HKATPase: structure, function, and inhibition. Pflugers Arch. 457, 609–622.

Levin, M., Johnson, R.L., Stern, C.D., Kuehn, M., and Tabin, C. (1995). A molecular pathway determining left-right asymmetry in chick embryogenesis. Cell 82, 803–814.

Shinohara, K., Kawasumi, A., Takamatsu, A., Yoshiba, S., Botilde, Y., Motoyama, N., Reith, W., Durand, B., Shiratori, H., and Hamada, H. (2012). Two rotating cilia in the node cavity are sufficient to break left-right symmetry in the mouse embryo. Nature Communications 3, 622.

Levin, M., Thorlin, T., Robinson, K.R., Nogi, T., and Mercola, M. (2002). Asymmetries in H+/K+-ATPase and cell membrane potentials comprise a very early step in left-right patterning. Cell 111, 77–89.

Shook, D.R., Majer, C., and Keller, R. (2004). Pattern and morphogenesis of presumptive superficial mesoderm in two closely related species, Xenopus laevis and Xenopus tropicalis. Dev. Biol. 270, 163–185.

Maisonneuve, C., Guilleret, I., Vick, P., Weber, T., Andre, P., Beyer, T., Blum, M., and Constam, D.B. (2009). Bicaudal C, a novel regulator of Dvl signaling abutting RNA-processing bodies, controls cilia orientation and leftward flow. Development 136, 3019–3030.

Sokol, S., Christian, J.L., Moon, R.T., and Melton, D.A. (1991). Injected Wnt RNA induces a complete body axis in Xenopus embryos. Cell 67, 741–752.

McGrew, L.L., Takemaru, K., Bates, R., and Moon, R.T. (1999). Direct regulation of the Xenopus engrailed-2 promoter by the Wnt signaling pathway, and a molecular screen for Wnt-responsive genes, confirm a role for Wnt signaling during neural patterning in Xenopus. Mech. Dev. 87, 21–32. Meijering, E., Jacob, M., Sarria, J.-C.F., Steiner, P., Hirling, H., and Unser, M. (2004). Design and validation of a tool for neurite tracing and analysis in fluorescence microscopy images. Cytometry A 58, 167–176. Miller, J.R., and Moon, R.T. (1997). Analysis of the signaling activities of localization mutants of beta-catenin during axis specification in Xenopus. J. Cell Biol. 139, 229–243. Morokuma, J., Blackiston, D., and Levin, M. (2008). KCNQ1 and KCNE1 K+ channel components are involved in early left-right patterning in Xenopus laevis embryos. Cell. Physiol. Biochem. 21, 357–372. Nonaka, S., Tanaka, Y., Okada, Y., Takeda, S., Harada, A., Kanai, Y., Kido, M., and Hirokawa, N. (1998). Randomization of left-right asymmetry due to loss of nodal cilia generating leftward flow of extraembryonic fluid in mice lacking KIF3B motor protein. Cell 95, 829–837. Ossipova, O., and Sokol, S.Y. (2011). Neural crest specification by noncanonical Wnt signaling and PAR-1. Development 138, 5441–5450. Paterson, H.F., Self, A.J., Garrett, M.D., Just, I., Aktories, K., and Hall, A. (1990). Microinjection of recombinant p21rho induces rapid changes in cell morphology. J. Cell Biol. 111, 1001–1007. Pohl, B.S., and Kno¨chel, W. (2004). Isolation and developmental expression of Xenopus FoxJ1 and FoxK1. Dev. Genes Evol. 214, 200–205. Schweickert, A., Campione, M., Steinbeisser, H., and Blum, M. (2000). Pitx2 isoforms: involvement of Pitx2c but not Pitx2a or Pitx2b in vertebrate left-right asymmetry. Mech. Dev. 90, 41–51.

Sokol, S.Y. (1996). Analysis of Dishevelled signalling pathways during Xenopus development. Curr. Biol. 6, 1456–1467. Song, H., Hu, J., Chen, W., Elliott, G., Andre, P., Gao, B., and Yang, Y. (2010). Planar cell polarity breaks bilateral symmetry by controlling ciliary positioning. Nature 466, 378–382. Stubbs, J.L., Oishi, I., Izpisu´a Belmonte, J.C., and Kintner, C. (2008). The forkhead protein Foxj1 specifies node-like cilia in Xenopus and zebrafish embryos. Nat. Genet. 40, 1454–1460. Tabin, C.J. (2006). The key to left-right asymmetry. Cell 127, 27–32. Tamakoshi, T., Itakura, T., Chandra, A., Uezato, T., Yang, Z., Xue, X.-D., Wang, B., Hackett, B.P., Yokoyama, T., and Miura, N. (2006). Roles of the Foxj1 and Inv genes in the left-right determination of internal organs in mice. Biochem. Biophys. Res. Commun. 339, 932–938. Tao, Q., Yokota, C., Puck, H., Kofron, M., Birsoy, B., Yan, D., Asashima, M., Wylie, C.C., Lin, X., and Heasman, J. (2005). Maternal wnt11 activates the canonical wnt signaling pathway required for axis formation in Xenopus embryos. Cell 120, 857–871. Te´telin, S., and Jones, E.A. (2010). Xenopus Wnt11b is identified as a potential pronephric inducer. Dev. Dyn. 239, 148–159. Tran, U., Pickney, L.M., Ozpolat, B.D., and Wessely, O. (2007). Xenopus Bicaudal-C is required for the differentiation of the amphibian pronephros. Dev. Biol. 307, 152–164. Vick, P., Schweickert, A., Weber, T., Eberhardt, M., Mencl, S., Shcherbakov, D., Beyer, T., and Blum, M. (2009). Flow on the right side of the gastrocoel roof plate is dispensable for symmetry breakage in the frog Xenopus laevis. Dev. Biol. 331, 281–291. Viotti, M., Niu, L., Shi, S.-H., and Hadjantonakis, A.-K. (2012). Role of the gut endoderm in relaying left-right patterning in mice. PLoS Biol. 10, e1001276.

Schweickert, A., Weber, T., Beyer, T., Vick, P., Bogusch, S., Feistel, K., and Blum, M. (2007). Cilia-driven leftward flow determines laterality in Xenopus. Curr. Biol. 17, 60–66.

Vonica, A., and Brivanlou, A.H. (2007). The left-right axis is regulated by the interplay of Coco, Xnr1 and derrie`re in Xenopus embryos. Dev. Biol. 303, 281–294.

Schweickert, A., Vick, P., Getwan, M., Weber, T., Schneider, I., Eberhardt, M., Beyer, T., Pachur, A., and Blum, M. (2010). The nodal inhibitor Coco is a critical target of leftward flow in Xenopus. Curr. Biol. 20, 738–743.

Wang, Y., Janicki, P., Ko¨ster, I., Berger, C.D., Wenzl, C., Grosshans, J., and Steinbeisser, H. (2008). Xenopus Paraxial Protocadherin regulates morphogenesis by antagonizing Sprouty. Genes Dev. 22, 878–883.

Schweickert, A., Walentek, P., Thumberger, T., and Danilchik, M. (2012). Linking early determinants and cilia-driven leftward flow in left-right axis specification of Xenopus laevis: A theoretical approach. Differentiation 83, S67– S77. Published online December 1, 2011. PMID: 22136958.

Wessely, O., and Tran, U. (2011). Xenopus pronephros development—past, present, and future. Pediatr. Nephrol. 26, 1545–1551.

Shimeld, S.M., and Levin, M. (2006). Evidence for the regulation of left-right asymmetry in Ciona intestinalis by ion flux. Dev. Dyn. 235, 1543–1553.

Zhang, Y., and Levin, M. (2009). Left-right asymmetry in the chick embryo requires core planar cell polarity protein Vangl2. Genesis 47, 719–728.

Zhang, B., Tran, U., and Wessely, O. (2011). Expression of Wnt signaling components during Xenopus pronephros development. PLoS ONE 6, e26533.


Supplemental Information EXTENDED EXPERIMENTAL PROCEDURES Cloning of Constructs A full-length ATP4a cDNA was cloned following PCR from adult gastric cDNA using primers matching NM_001090874: forward 50 ATGGGGAAAAAGGAACAATATG-30 , reverse 50 -TTAATAATACATCTCCTTGTCGAAC-30 . The full-length cDNA was subcloned into expression vector CS2+MT. Whole-Mount In Situ Hybridization Embryos were fixed in MEMFA for 1–2 hr and processed following standard protocols. Digoxigenin-labeled (Roche) RNA probes were prepared from linearized plasmids using SP6, T3, or T7 RNA polymerase (Promega). In situ hybridization was conducted according to Harland (1991). Probes used were as follows: ATP4a (this work), Pitx2c (Schweickert et al., 2000), ATP1b1 (Tran et al., 2007), Xnr1 and Coco (Schweickert et al., 2010), En2 (cloned from NM_001101743.1), Xtwist1 (cloned from NM_001085883), Foxj1 (Stubbs et al., 2008), and Sox3 (cloned from BC072222). RT-PCR RT-PCR was performed as described by Beyer et al. (2012a) using ATP4a-specific primers: forward 50 -TTATGTCCCT GAGGGTCTGC-30 , reverse 50 -AGTCGTTGACACCATCACCA-30 . Cytoskeletal actin type-8 served as loading control using primers matching M24770.1: forward 50 -AGGGTGTAATGGTTGGTATGG-30 , reverse 50 -ACCTTCTACAATGAACTTCGTG-30 . Immunohistochemistry Immunohistochemistry was performed as described by Vick et al. (2009) using the following antibodies: Anti-ATP4a (rabbit, 1:500) (Chen et al., 1998); Anti-Tubulin Acetylated (mouse, 1:700; Sigma); anti-rabbit Alexa 555 (goat, 1:250; Invitrogen); anti-mouse DyLight 488 (rabbit, 1:250; Jackson Immuno Research); and anti-mouse Cy3 (sheep, 1:250; Sigma). Cell boundaries were visualized by Alexa 488-conjugated phalloidin (Invitrogen), which stained the actin cytoskeleton. For histological analysis, embryos were embedded in gelatin-albumin and sectioned on a vibratome (30 mm). SEM analysis was performed as described by Schweickert et al. (2007). Convergent Extension Assays To analyze neural tube closure, we measured the width of the anterior neural plate at its widest point (as indicated by dashed lines in Figure S3A) using the ImageJ plugin NeuronJ (Meijering et al., 2004). AP length of tadpoles was analyzed accordingly (cf. Figure S3B). SUPPLEMENTAL REFERENCES Aw, S., Adams, D.S., Qiu, D., and Levin, M. (2008). H,K-ATPase protein localization and Kir4.1 function reveal concordance of three axes during early determination of left-right asymmetry. Mech. Dev. 125, 353–372. Beyer, T., Danilchik, M., Thumberger, T., Vick, P., Tisler, M., Schneider, I., Bogusch, S., Andre, P., Ulmer, B., Walentek, P., et al. (2012). Serotonin signaling is required for Wnt-dependent GRP specification and leftward flow in Xenopus. Curr. Biol. 22, 33–39. Chen, P.X., Mathews, P.M., Good, P.J., Rossier, B.C., and Geering, K. (1998). Unusual degradation of alpha-beta complexes in Xenopus oocytes by betasubunits of Xenopus gastric H-K-ATPase. Am. J. Physiol. 275, C139–C145. Harland, R.M. (1991). In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol. 36, 685–695. Meijering, E., Jacob, M., Sarria, J.-C.F., Steiner, P., Hirling, H., and Unser, M. (2004). Design and validation of a tool for neurite tracing and analysis in fluorescence microscopy images. Cytometry A 58, 167–176. Schweickert, A., Campione1, M., Steinbeisser, H., and Blum, M. (2000). Pitx2 isoforms: involvement of Pitx2c but not Pitx2a or Pitx2b in vertebrate left-right asymmetry. Mech. Dev. 90, 41–51. Schweickert, A., Vick, P., Getwan, M., Weber, T., Schneider, I., Eberhardt, M., Beyer, T., Pachur, A., and Blum, M. (2010). The nodal inhibitor Coco is a critical target of leftward flow in Xenopus. Curr. Biol. 20, 738–743. Schweickert, A., Weber, T., Beyer, T., Vick, P., Bogusch, S., Feistel, K., and Blum, M. (2007). Cilia-driven leftward flow determines laterality in Xenopus. Curr. Biol. 17, 60–66. Stubbs, J.L., Oishi, I., Izpisu´a Belmonte, J.C., and Kintner, C. (2008). The forkhead protein Foxj1 specifies node-like cilia in Xenopus and zebrafish embryos. Nat. Genet. 40, 1454–1460. Tran, U., Pickney, L.M., Ozpolat, B.D., and Wessely, O. (2007). Xenopus Bicaudal-C is required for the differentiation of the amphibian pronephros. Dev. Biol. 307, 152–164. Vick, P., Schweickert, A., Weber, T., Eberhardt, M., Mencl, S., Shcherbakov, D., Beyer, T., and Blum, M. (2009). Flow on the right side of the gastrocoel roof plate is dispensable for symmetry breakage in the frog Xenopus laevis. Dev. Biol. 331, 281–291.

Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors S1

Figure S1. ATP4a Expression, MO Targeting and Morphant Phenotypes, Related to Figure 1 (A–D) Expression of ATP4a mRNA (A) and protein (B-D). (A) RT-PCR analysis. Actin served as loading control. Note that mRNA levels decreased but expression persisted through stage 17. (B-D) Immunohistochemistry with an antibody previously used for ATP4a localization in Xenopus (Aw et al., 2008). (B, C) Z-stack projections; (D) individual frames. (B) Gastrula embryo at stage 10.5. ATP4a (B0 ) was ubiquitously expressed, including the superficial mesoderm on top of the dorsal lip of the blastopore (dotted line). (C) Dorsal explant at stage 17. Ventral view of the GRP at the boundary between the GRP and lateral endodermal crest (LEC) cells (dotted line). Double staining with an antibody against acetylated tubulin (green) in (C0 ) revealed cilia at the GRP and midbodies in the area of the LECs (shown at higher magnification in insets). (D) Optical sections of region outlined in (C) at levels indicated in schematic drawing demonstrated localized staining at membranes as well as in vesicle-like cytoplasmic structures. (E) Injection of lineage tracer rhodamine-B dextran into the marginal region (prospective C-tier of 32-cell embryo) of 4-cell embryos revealed specific targeting of GRP tissue only when dorsal blastomeres were injected close to the dorsal pole (dorsal marginal zone, DMZ, top). More lateral injections of dorsal blastomeres (C2 lineage) or injection of ventral blastomeres (C3 and C4) targeted the intermediate mesoderm (C2/C3), lateral plate mesoderm (C2/C3), and ventral mesoderm (VMZ), respectively. (F and G) Pitx2c expression patterns (F) and organ situs (G) encountered in ATP4a morphants. The outflow tract of the heart and the position of the gall bladder are indicated by green and red arrowheads, respectively, and the direction of gut looping is marked by yellow arrows. Note that morphants occasionally (and dosedependently) developed cysts, and therefore organ situs could not be determined.

S2 Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors

Figure S2. Frequency Distribution of Trajectory Angles in Dorsal Explants, Related to Figure 2 (A) Uninjected controls. (B) CoMo-injected controls. (C) ATP4aMO-injected specimens. (D) DMSO- and (E) SCH28080-treated embryos. Dashed circles mark maximum frequency in histogram specified in percent. a, anterior; l, left; n, number of particles above threshold; p, posterior; r, right; v, average velocity of particles; rho, quality of flow.


Figure S3. ATP4a Acts on Noncanonical Wnt Signaling, Related to Figure 4 (A and B) Convergent extension. (A) Neural tube closure: widening of the neural tube in ATP4a morphants. Embryos were injected unilaterally into the animal right blastomeres at the 4 cell stage, fixed and processed for Sox3 expression to visualize the neural plate by WMISH at stage 18. Staging was according to the progress of neural tube closure on the uninjected (left) side. Note that the widening of the neural plate was partially rescued upon ATP4a co-injection. (B) Shortening of the anterior-posterior (AP) axis. Embryos were bilaterally injected into the DMZ at the 4 cell stage and the AP extension was determined at stage 32. Note that MO-induced shortening was partially rescued by ATP4a co-injection. The width of the neural plate (A) and the length of embryos (B) were measured (as indicated by dashed lines). Results are depicted as box plots. (C) Neural crest cell specification and migration: right-sided ATP4aMO-injection resulted in reduced and altered Twist1 mRNA expression at stage 26 (n1-n3/4; neural crest migratory streams). (D) Right-sided pronephric tubule (pnt) defects upon unilaterally right-sided ATP4aMO-injection, as demonstrated by WMISH using the pronephros marker gene ATP1b1. d, dorsal; l, left; pnd, pronephric duct; pnt, pronephric tubule; r, right; v, ventral. ***, very highly significant. Numbers in brackets represent number of analyzed specimens.

S4 Cell Reports 1, 516–527, May 31, 2012 ª2012 The Authors

Figure S4. Rescue of Ciliation Rate but Not Cilia Polarization by Foxj1 in ATP4a Morphants, Related to Figure 5 Embryos were injected at the 4-cell stage into the DMZ and dorsal explants were prepared at stage 17. Specimens were processed for IHC to assess cilia polarization (A-D), or for SEM analysis to determine the GRP ciliation rate (E). (A–C) IHC using antibodies against acetylated tubulin to visualize cilia (red) and actin (green) to outline cell boundaries. (A) Control uninjected specimen. (B) ATP4a morphant. (C) Coinjection of ATP4aMo and Foxj1 mRNA. (A0 -C0 ) Evaluation of results. (D) Quantification of cilia polarization. Note that co-injection of Foxj1 aggravated polarization defects. (E) Ciliation rate. Note that ciliation rate was partially rescued by Foxj1 co-injection.
