Planar polarity specification through asymmetric subcellular localization of Fat and Dachsous

Abstract

Two pathways regulate planar polarity: the core proteins and the Fat-Dachsous-Four-jointed (Ft-Ds-Fj) system. Morphogens specify complementary expression patterns of Ds and Fj that potentially act as polarizing cues. It has been suggested that Ft-Ds-Fj-mediated cues are weak and that the core proteins amplify them. Another view is that the two pathways act independently to generate and propagate polarity: if correct, this raises the question of how gradients of Ft and Ds expression or activity might be interpreted to provide strong cellular polarizing cues and how such cues are propagated from cell to cell. Here, we demonstrate that the complementary expression of Ds and Fj results in biased Ft and Ds protein distribution across cells, with Ft and Ds accumulating on opposite edges. Furthermore, boundaries of Ft and Ds expression result in subcellular asymmetries in protein distribution that are transmitted to neighboring cells, and asymmetric Ds localization results in a corresponding asymmetric distribution of the myosin Dachs. We show that the generation of subcellular asymmetries of Ft and Ds and the core proteins is largely independent in the wing disc and additionally that ommatidial polarity in the eye can be determined without input from the Ft-Ds-Fj system, consistent with the two pathways acting in parallel.

Figure 1

Dachs Asymmetry and Its Regulation by Fj and Ds (A–E) Confocal images of WT third-instar wing discs showing Dachs distribution revealed by immunofluorescence using an antibody raised against Dachs. (B–E) Magnifications of specific regions of the disc as marked in (A). To quantify the degree of Dachs asymmetry, we calculated the ratio of intensity of immunofluorescence on PD/AP cell junctions for the different regions, showing more on PD boundaries in regions (B), (C) and (D) (PD/AP ratios of 1.3, 1.7, and 1.24 respectively) but a negligible difference in region (E) (PD/AP ratio of 1.05). Dorsal is top and anterior left for all wing discs. (F) Wing disc containing dachs¹ clones near the dorsal hinge (marked by lack of β-Gal, green) labeled for Dachs (magenta). Arrowheads point to Dachs at distal junctions. Arrow points to proximal cell junctions where Dachs is reduced. (G) Wing disc containing ft^G-rv clone near the hinge (marked by lack of β-Gal, green) labeled for Dachs (magenta). (H) Wing disc containing clone overexpressing Ds (Act>stop>GAL4, UAS-lacZ/UAS-ds, marked with β-Gal, green) labeled for Dachs (magenta). Dachs levels are increased on the boundary of the clone and 1–2 cells away (arrowheads). (I–L) EGFP-Dachs expressed in patches in the wing disc near to the dorsal hinge (I and J) and ventral eye disc (K and L). Arrowheads point to distal and arrows point to proximal cell junctions in (I) and (J). Arrowheads point to equatorial and arrows point to polar cell junctions in (K) and (L). Equator is top and anterior left. (I and K) WT (w hsFlp; Act>stop>EGFP-dachs). (J and L) No Ds and Fj gradients (w hsFlp; ds^UA071 fj^d1 Act>stop>EGFP-dachs/ ds^38K fj^P1; tub-GAL4/UAS-ds). (M) Mean fluorescence intensity of EGFP-Dachs staining at cell junctions on the edges of clones in genotypes in (I)–(L). WT (blue bars) and no Ds and Fj gradients (red bars). Values normalized to WT levels on distal or equatorial cell junctions. Error bars show SEM. One-way analysis of variance (ANOVA) tests were applied (^∗∗∗p < 0.001, ^∗p < 0.05, comparisons between columns linked by bars). EGFP-Dachs is significantly higher on distal cell junctions in the wing and significantly higher on equatorial and posterior cell junctions in the eye. No statistical difference was found between levels on cell junctions in the absence of Ds and Fj gradients (red bars). (N) Ds (magenta) and Fj (yellow) expression patterns in the wing and eye disc and corresponding direction of Dachs (green) asymmetry. In the wing disc, Ds levels are high in the hinge with lower levels in the wing pouch resulting in a strong boundary of Ds expression. Fj is expressed in the center of the wing pouch with a gradient toward the edge of the pouch. Dachs asymmetry forms on the PD axis (arrows) in alignment with these Ds and Fj boundaries or gradients with high levels of Dachs on distal cell junctions. In the eye, Ds is expressed in a gradient that is high at the poles but also slightly higher anteriorly. Fj is expressed in an opposing gradient with high expression at the equator. Dachs is polarized in alignment with these gradients with high Dachs on the equatorial and posterior side of the cells. Labels refer to dorsal (D), ventral (V), anterior (A), and posterior (P). See also Figure S1.

Figure 2

Asymmetry of Ft and Ds Distribution (A and B) Confocal images of apicolateral junctions in third-instar wing discs near the dorsal hinge. Labeled for (A) Dachs (magenta) and Ds (green) and (B) Dachs (magenta) and Ft (green). Arrowheads point to Dachs and Ds or Ft colocalization on PD cell boundaries. (C) Images of patches of Ds-EGFP expression in the wing and dorsal eye disc showing Ds-EGFP (green or white) and Arm (magenta). By using a fly strain in which ds is tagged at its endogenous locus with EGFP, we generated patches of Ds-EGFP expression abutting cells expressing untagged Ds but with no change in gene dosage (ds-EGFP FRT40/ ds⁺ FRT40). Ds-EGFP is enriched at distal cell edges in the wing disc (distal is bottom) and equatorial and posterior cell edges in the eye disc (equator is top and posterior right). Arrowheads point to Ds enrichment and arrows point to reduced Ds. (D) Twenty-eight hr pupal wing containing a clone overexpressing Ds (Act>stop>GAL4/UAS-ds) labeled for Ds (green) and Dachs (magenta). Arrowheads point to cell boundaries outside the overexpression clone where Dachs and Ds colocalize. Note that as Dachs and Ds asymmetry is lost by this stage of development ([2]; data not shown), the Dachs and Ds asymmetry generated around the clone is due to Ds overexpression. Additionally asymmetric localization is more easily observed in the pupal wing, because cells are larger and more uniform than in wing and eye discs and cell division is no longer occurring. (E) Ratio of mean fluorescence intensity of Actin, Dachs, Ft, and Ds labeling on PD compared to AP cell boundaries in wing discs close to the dorsal hinge. Error bars show SEM between wing discs (n = 10). A one-way ANOVA test was applied (significance levels [^∗∗∗p < 0.001, ^∗∗p < 0.01, ^∗p < 0.05] between columns linked by bars). Dachs, Ds, and Ft show enrichment on PD boundaries compared to cortical actin. Enrichment of Ft (^∗p > 0.05) and Ds (^∗∗p < 0.01) is not as strong as Dachs. (F and G) Mean fluorescence intensity levels of Ds-EGFP staining at cell junctions on the edges of clones in wing (F) and eye discs (G) normalized to WT levels on distal or equatorial cell junctions. Error bars show SEM. One-way ANOVA tests were applied (significant differences linked by bars). In wing discs, Ds-EGFP is significantly higher on distal cell junctions than posterior, anterior, or proximal junctions. In the eye, Ds-EGFP is signficantly higher on equatorial and posterior cell junctions than polar or anterior boundaries. (H) Mean fluorescence intensity of Dachs (blue), Ds (red), and Arm (yellow) on cell boundaries parallel to UAS-ds overexpression clones in the pupal wing. Intensity was normalized to mean levels on boundaries away from the clone. The cell boundary at the clone edge is labeled 0 (zero). Error bars show SEM between clones (n = 10). A one-way ANOVA test was applied comparing each column with values on cell boundaries away from the clone (column 6). When averaged over ten clones, significant increases in Dachs and Ds were only found around 1–2 cells away from the clone boundary, although in some clones, differences are visible by eye 3–4 cells away; for example, see the clone in (D). See also Figure S2.

Figure 3

Relationship of Ft-Ds-Dachs and Core Protein Polarization (A–E) Images of EGFP-Dachs expressed in patches under the actin promoter in the wing disc near to the dorsal hinge in WT (F) and fz^P21 mutant background (G) labeled for GFP (green) and Arm (magenta). (B–E) Shown are magnifications of specific regions of the disc as marked in (A). Arrows show the direction of Fz-EYFP asymmetry pointing distally (toward the center of the wing pouch). Numbers refer to strength of Fz-EYFP asymmetry in individual clones as a ratio of fluorescence levels on cell boundaries on distal edges of clones versus proximal edges. There is a high degree of variation in the strength of asymmetry between clones, with an average asymmetry of 6.0 ± 0.96 (SEM). (F and G) Images of EGFP-Dachs expressed in patches under the actin promoter in the wing disc near to the dorsal hinge in WT (F) and fz^P21 mutant background (G) labeled for GFP (green) and Arm (magenta). Arrowheads point to distal cell junctions where EGFP-Dachs is enriched and arrows point to proximal junctions with less Dachs. Asymmetry of EGFP-Dachs was measured and no significant difference detected between WT and fz mutant (ratio of distal to proximal EGFP-Dachs WT = 2.9 ± 0.35 [SEM] and fz^P21 = 2.5 ± 0.21 [SEM], >40 clones). (H–L) Images of Fz-EYFP expressed in patches under the actin promoter in wing discs in a dachs^GC13/dachs¹ mutant background labeled for GFP (green) and Arm (magenta). (I–L) Shown are magnifications of specific regions of the disc as marked in (H). In dachs^GC13/dachs¹ discs, Fz-EYFP predominately points distally except in dorsal clones close to the hinge and distant from the AP boundary (L) where the direction of Fz-EYFP asymmetry follows the pouch-hinge boundary. This was not the case in WT (E). In (I)–(K), numbers refer to strength of Fz-EYFP asymmetry in individual clones as ratio of distal/proximal fluorescence levels. In (L), the ratio of Fz-EYFP fluorescence relative to the axis of the pouch-hinge boundary is indicated: PD asymmetry is negligible in these clones (PD ratios are 1.2 and 0.7). (M) Mapped orientation of Fz-EYFP asymmetry in the dorsal half of wing discs in WT, dachs^GC13/dachs¹ and ft^G-rv dachs^GC13/ ft⁸ dachs¹. Red arrows indicate Fz-EYFP clones in which asymmetry was misoriented (pointing away from the AP boundary). See also Figure S3.

Figure 4

Regulation of Ommatidial Polarity by the Ft-Ds Pathway and Dachs (A–F) Diagrams representing the chirality and orientation of ommatidia in the adult eye. Dorsal chiral form of ommatida (red), ventral chiral form (green), achiral ommatidia (blue), and disrupted ommatidia (^∗) marked. Eyes wholly mutant for specific genotypes were generated using the eyeless-GAL4 UAS-FLP (“EGUF”) system except in (E). (A) FRT40. (B) ft^G-rv FRT40. (C) dachs^GC13 FRT40. (D) ft^G-rv dachs^GC13FRT40. (E) ft^G-rv dachs^GC13 UAS-wts / ft⁸ dachs¹ tub-GAL4. (F) ft^G-rv dachs^GC13 FRT40 UAS-yki. (G) Quantification of polarity defects. Columns 1–6 are the same genotypes as in (A)–(F). Additional genotypes are ds^UA071 /ds^38k (column 7) and ds^UA071 d^GC13 / ds^38k d¹ (column 8). Error bars show SEM between eyes (n > 6). A one-way ANOVA test was applied (^∗∗∗p < 0.001, ^∗∗p < 0.01, ^∗p < 0.05 comparisons between columns linked by bars). ft^G-rv/ ft⁸ eyes could not be examined for comparison with genotype in (E) because the combination is lethal. However, ft⁸ EGUF eyes have a level of inversions (data not shown) similar to ft^G-rv eyes, suggesting that the polarity phenotype is suppressed in (E). (H–J) Diagrams representing orientation of ommatidia around overexpression clones (yellow). (H) Act≫GAL4/UAS-ft. (I) Act≫GAL4/UAS-ds. (J) dachs^GC13/dachs¹ Act≫GAL4 UAS-ds. Equator is toward the bottom and pole toward the top of the images. Nonautonomous inversions in polarity occur on the equatorial side of UAS-ft clones (H) and on the polar side of UAS-ds clones (I). UAS-ds clones in a dachs mutant (J) produce polar inversions at the same level with an average of number of inversions per clone in WT of 2.8 (n = 4) and 2.7 in dachs (n = 7). See also Figure S4.