Fig. 1. Inversin is required for pronephric tubule development in X. laevis. (A and B) Bilateral injection of Invs-Mo at the four-cell stage resulted in severe edema. (C) The pronephros-specific antibodies 3G8 and 4A6 stain the entire pronephric tubule on the uninjected side. (D) On the Invs-Mo–injected side of the same embryo, staining is absent in the distal and intermediate tubule segment but is maintained in the proximal segment. (E and F) Enlarged view of the pronephros in C and D, respectively. (G) Percentages of embryos with unobstructed tubular excretion on both sides. (H) Fluorescent dextran (70 kDa) excretion shows tubule patency in Invs-Mo–injected embryos. (I–L) Enlarged views of the pronephros at the phase of maximal dextran excretion. Note the reduction in tubular coiling on the Invs-Mo–injected side (J) compared with the uninjected side (I). (K and L) Injection of control-Mo (CTL-Mo) did not interfere with tubule patency or length. (M) Tubule-area size was measured as depicted. d.v., dorso-ventral; a.p., anterior-posterior. (N) Ratios of injected vs. uninjected tubule areas of the same embryos.
Fig. 3. Impaired pronephros morphogenesis caused by Inversin depletion is largely independent of proliferation or apoptosis. (A) Embryos were in situ hybridized against Na-K-ATPase (inverted differential interference contrast, green) and immunostained with an anti–phospho-Histone H3 (p-Histone H3) antibody (red) and DAPI (blue). The area of the proximal pronephros was magnified (Right). (B) The percentage of mitotic (p-Histone H3-positive) cells in the pronephros was determined in five sections of 10 embryos; no significant difference was detected (P = 0.36, Student t test). (C) Section of a camptothecin-treated embryo; camptothecin almost completely abrogates mitosis. The area of the proximal pronephros was magnified (Right). (D) Positive TUNEL staining is indicated by a black arrow. The area of the proximal pronephros was magnified (Right). (E) Immunostaining with anti–Caspase-3 (green), acetylated tubulin (red), and DAPI (blue). Note that both methods did not detect increased apoptosis on the Invs-Mo–injected side. (F and G) Mitotic inhibition does not prevent ventral extension defects of pronephric loops after Invs-Mo injection. (F) Embryos were treated at stage 31 with DMSO or camptothecin and processed for in situ hybridization against Na-K-ATPase at stage 36. Brackets mark the boundaries of ventral extension of the pronephric loop. (G) Quantification of ventral extension in millimeters. Error bars represent SD (*P < 0.001). The red arrow points to a reduction of the ventral pronephros extension in Invs-Mo–injected embryos that significantly exceeds the reduction caused by camptothecin treatment alone.
Fig. 4. Inversin acts downstream of Frizzled-8 in convergent extension movements and pronephros development. (A–D) Embryos were unilaterally injected with Fzd-8–Mo. In situ hybridization against Na-K-ATPase shows reduced ventral extension of the intermediate tubule. (C and D) Enlarged view of the pronephros region in A and B. (E) Dorsal injection of dominant-negative Xenopus Frizzled-8 encompassing the extracellular domain of Frizzled-8 (ECD8). Axis extension defects indicative of impaired convergent extension (ICE) were scored as indicated on a scale from 1 to 3. Coexpression of Inversin mRNA partially rescued the convergent extension defects. (F–N) Staining of the pronephros with the tubule-specific antibodies 3G8 and 4A6. (H–M) Unilateral injection of Fzd-8–Mo resulted in a reduction of 3G8 and 4A6 staining ranging from mild (I) to severe (K). (L–N) Coinjection of Inversin mRNA (1 ng) rescued the defective 3G8/4A6 immunoreactivity in Fzd-8–Mo-injected embryos (*P < 0.05).
Fig. 2. Inversin affects morphogenetic movements during ventral extension
of the proximal pronephros. (A–E) Abnormal ventral extension of the
Inversin-depleted pronephros. (A–D) Time-lapse confocal microscopy of
pronephros morphogenesis was performed in parallel between CTL-Mo–
and Invs-Mo–injected Xenopus embryos to monitor cell movements within
the developing proximal pronephros. Cells were labeled by membrane-GFP
and nuclear-RFP; only cells clearly identifiable as part of the pronephros
were tracked (white dots in A and C). Clustering of cells with similar
movements (green, red, yellow, and blue balls in B and D) was performed
automatically; their trajectories over the indicated time are depicted in B
and D. Cells that left the focal plane are depicted as crosses. (Scale bar: 100
μm.) (E) The distance between green (corresponding to the proximal segment) and yellow (intermediate segment) cell clusters in micrometers over
time (hours). In contrast to CTL-Mo–injected embryos (blue line), the distance
between these two cell clusters failed to increase in Invs-Mo–injected embryos during the first 6 h (green line), causing the abnormal ventral extension of the Inversin-depleted pronephros. (F–I) Distal to proximal cell
migration is unaffected by Inversin depletion. (F) The photo-convertible
fluorophore Kaede was broadly expressed in Xenopus embryos injected with
either CTL- or Invs-Mo. Photoconversion from green to red was restricted to
a stripe at stage 34. (G) At stage 39 (18 h after photoconversion), red fluorescent cells extended from the red stripe in a proximal direction. (I) Enlargement of the boxed area in G. The dashed line marks the left-side
boundary between the photo-converted area and surrounding tissue. The
white arrowhead indicates red fluorescent distal tubule cells that migrated
in a proximal direction. The arrow points to the pronephric tubule convolute. (H) Measurements of anterior migration in micrometers in CTL- and
Invs-Mo–injected Xenopus embryos. Error bars represent SEM.
Fig. 5. Inversin is required for membrane localization of Dishevelled. (A)
Mosaic cell clusters in animal caps were generated by sequential injection at
the two- and four-cell stage. All cells received a first injection at the two-cell
stage that included Dishevelled-GFP. A subset of these cells received a second
injection at the four-cell stage; only cells modified by the second injection are
marked by red nuclei expressing histone-2B–red fluorescent protein (H2BRFP). Thus, neighboring mosaic cell clusters with different injections can be
compared based on the presence or absence of red nuclei. (B) Dishevelled-GFP,
localized in a typically punctuate pattern (first injection, all cells), is recruited
to the plasma membrane in Frizzled-8 (Fzd-8)–injected cells (second injection,
cells with red nuclei). (C) Fzd-8–induced membrane localization of Dishevelled
(first injection, all cells) is abolished in the subset of cells that received the InvsMo in a second injection (second injection, cells with red nuclei). (D) Defective
membrane recruitment of Dishevelled after injection of Dishevelled-GFP, Fzd8, and Invs-Mo (first injection, all cells) can be rescued by mouse Inversin RNA
injection (second injection, cells with red nuclei). DAPI staining (blue) visualizes
all cell nuclei. White arrows point to membrane-localized Dishevelled. (E and
F) In immunostaining of stage-36 embryos, Dishevelled-2 (green) localizes to
the membrane of tubular epithelial cells indicated by the presence of cilia
(acetylated tubulin in red, white arrowheads). A dashed line indicates the
apical surface of the pronephric tubule; nuclei are stained with DAPI (blue).
The Invs-Mo–injected side shows a strong reduction of Dishevelled-2 staining
at the apical membrane of the pronephric tubules. The Dishevelled channel is
depicted alone in E′ and F′.
Fig. S1. Histological sections of Invs-Mo–injected embryos. (A) Schematic of targeted injection site into four to eight cell-stage Xenopus embryos. In unilateral
injections, membrane GFP (mem GFP) was included as a lineage tracer to confirm correct targeting into the pronephros region. (B) The uninjected side shows
background fluorescence. (C) GFP is detectable along the pronephros (red arrow) on the injected side. (D–I) H&E staining and histological sections of embryos
at stage 44. nt, neural tube; so, somites; pn, pronephros; en, endoderm. (E and H) Enlarged view of the pronephros of uninjected (E) and Invs-Mo–injected (H)
embryos. The pronephros seems smaller in Invs-Mo–injected embryos. Double arrow, proximal tubule; single arrows, intermediate tubules. (F and I) Distal
tubules in a separate section (arrowheads) of uninjected (F) and Invs-Mo–injected (I) embryos. (Scale bar: 200 μm.)
Fig. S2. Inversin depletion reduces ventral extension of the pronephric tubule. (A–J) In situ hybridization against Na-K-ATPase was performed at different
stages. The right column shows the Invs-Mo–injected side; the left column shows the noninjected side of the same embryo. (C and D) Enlarged view of the
boxed areas in A and B, respectively. (C, E, G, and I) The primary loop extends ventrally on the uninjected side. (D, F, H, and J) In Invs-Mo–injected sides,
extension of the ventral pronephros is reduced, leading to a limited elongation of the tubular convolute. (K–L) In situ hybridization against chloride channel
ClC-Ka was performed on continuous serial sections of E17.5 mouse kidneys (8 μm) to reconstruct ClC-Ka–positive nephron segments of wild-type and Invs (−/−)
kidneys. Black bars indicate 200 μm. (M) Quantification of the average volume per tubule in millimeters cubed (P = 0.057; n = 4 kidneys, t test).
Fig. S3. Transcription factors characterizing the pronephros anlagen are unaffected by depletion of Inversin. (A–L) Unilateral injection of Invs-Mo at the fourcell stage was followed by whole-mount in situ hybridization against transcription factors that characterize the pronephric anlagen, including lim-1 (A and B),
pax-8 (C and D), hepatocyte nuclear factor 1 homeobox B (vHNF1) (E–H), and the tubule-specific transcription factor GATA binding protein 3 (GATA-3) (I–L)
at the indicated stages. No substantial differences were detected compared with the noninjected side. (M) In situ hybridization against Wnt4 visualized in
a 200-μm vibratome section of an embryo. (N) The white arrow points to the in situ signal localized in the somatic layer (so) of the intermediate mesoderm. No
expression was detected in the splanchnic layer (sp) or the endoderm (ed). This indicated correct polarization of the pronephric anlagen.
Fig. S4. Normal segmentation of the Inversin-deficient pronephros. (A) Schematic diagram of segment-specific expression of in situ markers. PT, proximal
tubule; IT, intermediate tubule; DT, distal tubule. (B–K) After unilateral injection of Invs-Mo, whole-mount in situ hybridization was performed at stage 38. The
sodium/glucose cotransporter (SGLT-1K) is expressed in the proximal tubule and nephrostomes (B and C). The sodium bicarbonate cotransporter NBC1 is
expressed in the proximal tubule and early distal tubule (DT1) but not the intermediate tubule (D and E). Note that the gap between proximal and distal
expression was clearly visible in Invs-Mo–injected embryos. The sodium/potassium/chloride transporter NKCC2 is expressed in the intermediate segment (F and
G), the thiazide-sensitive NaCl cotransporter NCC is expressed in the distal tubule (H and I), and ClC-K is expressed in the intermediate segment (J and K).
Intermediate tubules appear smaller and less extended on the injected side (arrowheads in G and K). Brackets in J and K mark the boundaries of ventral
extension of the intermediate tubule. (L and M) The glomerular marker Nephrin is equally expressed on both sides. (N) Dorso-ventral extension of the pronephric tubule measured in embryos stained for ClC-K in millimeters is shown. Coinjection of Inversin mRNA rescued the defects of Invs-Mo (*P ≤ 0.001, rank
sum test; the error bars represent SEM)
Fig. S5. Wnt target genes are not altered after knockdown of Inversin. Invs-Mo–injected embryos were fixed at indicated stages; in situ hybridization was
performed against the canonical Wnt target genes axin2 (A–F) and fibronectin (G–L)
Fig. S6. Dishevelled knockdown mimics the Inversin phenotype. (A and B) Overexpression of mInvs did not affect pronephros formation. (C and D) Combined
knockdown of Dishevelled-1 and -3 or expression of the dominant-negative Dishevelled mutant Xdd1 (E and F) resulted in a shortened tubule with impaired
ventral extension. (G) β-Catenin was detectable at the plasma membrane but not within the nucleus. (H–K) 3D tilted projections of the confocal stacks of G.
White arrows point to cilia stained for acetylated tubulin.
Fig. S7. Inversin knockdown does not affect JNK signaling. (A) Overexpression of Fzd-8 does not rescue the pronephros defects of Invs-Mo injection. (B)
Activated phospho-JNK 1 and 2 were detected by immunostaining. Nuclear localization and intensity were not altered in Invs-Mo–injected embryos. (C)
Overexpression of the Jun N-terminal/mixed lineage kinase xMLK did not rescue the defects in tubule extension.
Fig. S8. Inversin is required but not sufficient for membrane localization of Dishevelled. (A, D, G, and J) Subcellular localization of Dishevelled-GFP (0.2 ng) in
animal cap cells. (B, E, H, and K) DIC images were obtained to visualize cell borders. (C, F, I, and L) Overlay. (A–C) Dishevelled-GFP alone. (D–F) Dishevelled-GFP
localizes to the plasma membrane on coexpression of Fzd-8. (G–I) Dishevelled-GFP accumulates in the cytoplasm if Invs-Mo and Fzd-8 mRNA are coinjected. (J–
L) Dishevelled-GFP accumulates in cytoplasmic vesicles in animal cap cells that were only coinjected with Inversin-RNA. (M) Lateral view of a z stack of mosaic
animal cap cells. All cells were injected with Dvl-GFP and Fzd-8; cells with red nuclei also received Invs-Mo and reveal reduced membrane-associated Dvl-GFP.
(N) Optical section of the apical surface. (O) xInvs transcripts were detectable in animal cap cells by RT-PCR.