An inwardly rectifying K+ channel is required for patterning

Abstract

Mutations that disrupt function of the human inwardly rectifying potassium channel KIR2.1 are associated with the craniofacial and digital defects of Andersen-Tawil Syndrome, but the contribution of Kir channels to development is undefined. Deletion of mouse Kir2.1 also causes cleft palate and digital defects. These defects are strikingly similar to phenotypes that result from disrupted TGFβ/BMP signaling. We use Drosophila melanogaster to show that a Kir2.1 homolog, Irk2, affects development by disrupting BMP signaling. Phenotypes of irk2 deficient lines, a mutant irk2 allele, irk2 siRNA and expression of a dominant-negative Irk2 subunit (Irk2DN) all demonstrate that Irk2 function is necessary for development of the adult wing. Compromised Irk2 function causes wing-patterning defects similar to those found when signaling through a Drosophila BMP homolog, Decapentaplegic (Dpp), is disrupted. To determine whether Irk2 plays a role in the Dpp pathway, we generated flies in which both Irk2 and Dpp functions are reduced. Irk2DN phenotypes are enhanced by decreased Dpp signaling. In wild-type flies, Dpp signaling can be detected in stripes along the anterior/posterior boundary of the larval imaginal wing disc. Reducing function of Irk2 with siRNA, an irk2 deletion, or expression of Irk2DN reduces the Dpp signal in the wing disc. As Irk channels contribute to Dpp signaling in flies, a similar role for Kir2.1 in BMP signaling may explain the morphological defects of Andersen-Tawil Syndrome and the Kir2.1 knockout mouse.

Fig. 1.

Knockout of mouse Kir2.1 causes cleft of the secondary palate and patterning defects of the skeletal digits. (A-D) Ventral view of Alizarin Red (bone) staining of palates of newborn wild-type (A), Kir2.1 knockout (B) and Kir2.1 heterozygous knockout (C,D) pups. (E,F) Ventral view of wild-type palate (E) and Kir2.1 knockout cleft of secondary palate (F). (G,H) Forelimb of wild-type (G) and Kir2.1 knockout (H). (I,J) Ventral view of a wild-type (I) and Kir2.1 knockout (J) forepaw. (K,L) Whole-mount forelimb skeletons from newborn animals stained with Alcian Blue and Alizarin Red: wild-type (K); preaxial digit duplication of the forelimb is shown in the Kir2.1 knockout (L). The dorsal view of the right limb is shown. Anterior is upwards.

Fig. 2.

Reduced Irk2 function wing phenotypes. (A) Quantification of wing defects when irk2 function is decreased (n≥100 flies). Error bars indicate s.e.m. (B) Bifurcation of L2 and L3 wing veins from Irk2DfA/Irk2DfB female. (C) Reduced wing size, angled hinge and bifurcation from L4 of Irk2DfA/Irk2DfB female. (D) Bristle transformations and incomplete L5 from male irk2^G8696. (E) Wild-type male wing L2-5 longitudinal veins; anterior (a) and posterior (p) crossveins are labeled. (F) irk2DfA/irk2DfB male fly wing. (G) Wing expansion defects from daughterless-GAL4 Irk2 siRNA female. (H) Ubiquitous expression of irk2 siRNA causes patches of wing tissue necrosis. (I) Small wing and hinge defects from MS1096-GAL4 Irk2 siRNA female. (J) Small wing, hinge and venation defects from MS1096-GAL4 Irk2 siRNA male. (K) Rescued male irk2DfA/irk2DfB; engrailed-GAL4 UAS-irk2WT fly wing. Arrows indicate wing hinge defects. Scale bars: 100 μm.

Fig. 3.

Phenotypes of flies expressing Irk2DN. (A,B) Graphs of adult (A) and larval (B) survival of actin-GAL4 UAS-Irk2WT, actin-GAL4 UAS-Irk2DN and siblings without transgene. Error bars indicate s.e.m. (C) L2-3 fusion in wing from MS1096-GAL4 Irk2DN5.1 female at 25°C. (D) Bifurcated wing from MS1096-GAL4 Irk2DN5.1 at 25°C. (E) engrailed-GAL4 UAS-irk2WT. (F) Wing of MS1096-GAL4 Irk2DN5.1 male at 25°C. (G) MS1096-GAL4 UAS Irk2DN5.1 at 29°C. (H) engrailed-GAL4 UAS Irk2DN5.1 at 25°C.

Fig. 4.

Irk1 and Irk3 siRNA phenotypes. (A) Ubiquitous expression of Irk1 and Irk3 siRNA causes lethality. (B) Wing-directed Irk1 and Irk3 siRNA causes wing venation defects and a reduction in wing size. (C) Irk1-AAA or Irk3-AAA ubiquitous expression does not decrease survival. (D) Quantification of Irk1 mRNA in wild-type, actin-GAL4-UAS Irk1WT and actin-GAL4-UAS Irk1AAA. (E) Quantification of Irk2 mRNA in wild-type, actin-GAL4-UAS Irk2WT and actin-GAL4-UAS Irk2AAA. (F) Quantification of Irk3 mRNA in wild-type, actin-GAL4-UAS Irk3WT and actin-GAL4-UAS Irk3AAA. Data are mean±s.e.m.

Fig. 5.

Reduced Dpp or Thickveins function enhances Irk2DN phenotypes. (A,B) Irk2DN5.2 female incomplete L5. (C) Thickened, bifurcated veins of dpp^hr92/+ female wing. (D) Irk2DN5.2/+; dpp^hr92/+ female wing is small and missing veins. (E) Thickened, bifurcated L2, L3 and L4 veins, and incomplete anterior crossvein of tkv⁷/+ female wing. (F) Irk2DN5.2/+; tkv⁷/+ female wing is small and missing veins. (G) Quantification of wing defects of MS1096-GAL4 UAS-irk2WT, UAS-Irk2DN5.1, UAS-Irk2DN 5.2 and with dpp^hr92/+ or tkv⁷/+. n>100 for all genotypes. Data are mean±s.e.m.

Fig. 6.

Reduced Mad phosphorylation in Irk2DfA/Irk2DfB and Irk2 siRNA wing discs. (A-D) TUNEL-stained wing discs. Anterior is rightwards. (E-H) Anti-p-Mad stained wing discs. (I-L) Relative fluorescence intensity across a posterior to anterior cross-section of the anti-p-Mad-stained wing disc shown in E-H. (M-O) Graphs of average peak intensity of control and irk2DfA/irk2DfB (M), control and MS1096-GAL4 irk2 siRNA (N), and control and daughterless-GAL4 Irk2 siRNA (O). Control and experimental discs were stained and imaged in parallel. Graphs represent average peak intensities for n>7 anti-p-Mad-stained discs. Peak intensity is determined by subtracting minimum from maximum fluorescence intensity in a posterior to anterior cross-section of the anti-p-Mad stained wing disc. Data are mean±s.e.m. Scale bars: 50 μm.

Fig. 7.

Immunohistochemistry demonstrates that Dpp signaling is reduced in Irk2DN and Irk2DfA/Irk2DfB. (A-I) Wild-type wing discs (A-C), irk2DfA/irk2DfB (D,E) and Irk2DN5.1 (G-I) stained with anti-Spalt (A,D,G), anti-Wingless (B,E,H) and anti-Achaete (C,F,I). Scale bars: 50 μm.

Fig. 8.

Irk2DN expression causes apoptosis and eliminates p-Mad staining in the third instar larval imaginal wing disc. (A-C) TUNEL stained wing disc. (D-F) Anti-p-Mad stained wing disc. (A,D) MS1096-GAL4-Irk2WT. (B,E) MS1096-GAL4-Irk2DN. (C,F) MS1096-GAL4-UAS-Irk2DN; P35. n>10 discs. (G) MS1096-GAL4-UAS-Irk2DN; P35 male wing. (H) Control MS1096-GAL4-UAS-Irk2DN male wing. Scale bars: 50 μm.

Fig. 9.

Model for Irk channels in Dpp signaling. (A) With functional Irk channels, Dpp binds type 1 (Thickveins) and type 2 (Punt) kinase receptors that are stabilized by proteoglycans (Dally, not shown). Upon Dpp binding, activated type 1 receptors phosphorylate Mad. P-Mad binds Medea and enters the nucleus to affect transcription. (B) Blocking Irk channels hinders Mad phosphorylation. Irk channels could be necessary for Dpp production/distribution, receptor complex stabilization or Tkv/Mad phosphorylation.