Frizzled2 receives WntA signaling during butterfly wing pattern formation

ABSTRACT Butterfly color patterns provide visible and biodiverse phenotypic readouts of the patterning processes. Although the secreted ligand WntA has been shown to instruct the color pattern formation in butterflies, its mode of reception remains elusive. Butterfly genomes encode four homologs of the Frizzled-family of Wnt receptors. Here, we show that CRISPR mosaic knockouts of frizzled2 (fz2) phenocopy the color pattern effects of WntA loss of function in multiple nymphalids. Whereas WntA mosaic clones result in intermediate patterns of reduced size, fz2 clones are cell-autonomous, consistent with a morphogen function. Shifts in expression of WntA and fz2 in WntA crispant pupae show that they are under positive and negative feedback, respectively. Fz1 is required for Wnt-independent planar cell polarity in the wing epithelium. Fz3 and Fz4 show phenotypes consistent with Wnt competitive-antagonist functions in vein formation (Fz3 and Fz4), wing margin specification (Fz3), and color patterning in the Discalis and Marginal Band Systems (Fz4). Overall, these data show that the WntA/Frizzled2 morphogen-receptor pair forms a signaling axis that instructs butterfly color patterning and shed light on the functional diversity of insect Frizzled receptors.

, situated between the R 5 -M 3 veins, in late fifth instar wing disks.This fBOc expression persists in pupal stages (Fig. 2), and has not been described in nymphalids other than V.

Fig. S9 .
Fig. S9.Size regulation of Discalis elements may involve Wg/Fz2 signaling in ventral V. cardui and A. incarnata forewings.(A) The D 1 and anterior D 2 (aD 2 ) elements are unaffected by WntA mKOs.(B) D 1 and aD 2 show an expansion of black in fz2 KOs.(C) R/patternize heatmap comparing WntA vs. fz2 mKO forewings.Black scale expansions were consistently observed in D 1 and D 2 across all the fz2 mKO samples (arrows).The upper row features black pixel distributions across WT (N = 20 images), WntA mKO (N= 20), and fz2 mKO (N= 40).(D) Pupal forewing expression of wg marks presumptive aD 2 and D 1 patterns, suggesting exaggerated fz2 phenotypes may be due to a role of Wg/ Fz2 signaling in inhibiting their size.(E) Pupal hindwing expression of wg marks the Basalis (B), D 2 , and D 1 elements.These three elements remain after WntA and fz2 loss-of-function.(F) Pupal forewing of A. incarnata is marked by wg in the presumptive D 2 and D 1 patterns similar to in V. cardui.(G) Pupal hindwing expression of wg is only in the presumptive B element.(H-K) In A. incarnata 15-17% pupal wings, wg is most prominent in the forewing D 2 (split across the anterior and posterior compartments, here marked with the M 2 vein as dotted line), weakly expressed in the forewing D 1 , and absent from hindwings where D elements are missing.Scale bars = 1 mm

Fig. S10 .
Fig. S10.Conserved functions of WntA and fz2 in a third nymphalid butterfly.(A) J. coenia wild type with annotated NGP elements and fz2 mKO mutant.fz2 crispants show a complete loss of the Basalis (B) and CSS elements, as well as a distalization and reshaping of the dPf elements, most visible in the forewings (box).(B) ISH of WntA, fz2 and wg in the proximal regions of the J. coenia WT forewing at 17% pupal development (n = 5 replicates).WntA induces the B and CSS elements, with visible fz2 depletion at this stage, consistently with a negative feedback of WntA/Fz2 signaling on fz2; wg is strongly expressed in D 1 and D 2 .(C) Expression of WntA and depletion of fz2 in the hindwing CSS (cyan) and MBS region (green) is consistent with a role of the WntA/Fz2 signaling on inducing the CSS and positioning peripheral patterns.

Fig. S11 .
Fig. S11.Mosaic fz1 crispants show wings with disorganized scale arrays, short forewings, and no pattern defects in J. coenia.PCP phenotypes are visible in these whole-specimen views as zones of apparent wing wear.Left: ventral side ; right : dorsal side.Black arrowheads : extensive PCP phenotypes ; gray arrowheads : wing surface with small PCP clones ; no arrowhead : surface with WT phenotype.Asterisks denote forewings that are WT on ventral and dorsal sides, and contralateral to a PCP-mutant forewing.Wing length (magenta) and width (green) between vein landmarks were measured in these PCP-asymmetric forewing pairs (N=7).Mutant wings are significantly shorter than WT in length (Wilcoxon signed rank test, p = 0.039) but not in width (p = 1).

Fig. S12 .
Fig. S12.Pupal wing defects with incomplete wing growth and cuticularization in fz1 crispants of V. cardui.(A) WT pupa in lateral (top) and ventral (bottom) views, with insets showing the distal portion of the pupal forewing (fw) and hindwing (hw) suturing with thoracic and abdominal cuticle.(B) Two examples of pupal phenotypes observed in V. cardui fz1 crispants (N= 13), with incomplete wing growth and distal sealing.Antennae, proboscis and legs, which bundle around the ventral midline, appeared unaffected.Scale bars = 1 mm.

Fig. S13 .
Fig. S13.Pupal RNAi electroporation knockdowns of fz1 in J. coenia.Ventral views of single specimens electroporated with fz1 DsiRNA on the ventral right forewing, shown here next to their contralateral wild-type control (WT-CL).PCP-like phenotypes are visible as fields of straightened scales (lacking normal curvature) and lacking regular organization under high magnification in all five replicates (as shown in E', bottom panel) and were not observed in controls.PCP-like phenotypes were pronounced in A. incarnata knockdown experiments (Fig. 4G-J).Scale bars: E' = 200 µm.

Fig. S14 .
Fig. S14.Pupal RNAi electroporation sham-negative controls and description of stress artefacts in J. coenia.Examples of artefacts observed in sham electroporation procedures with GFP DsiRNA (A-B, ventral forewings ; C, ventral hindwing), and following electroporation with the electrode directly in contact with the wing tissue (D-F).See Methods section for details.

Fig. S15 .
Fig. S15.Pupal RNAi electroporation knockdowns of positive controls fz2 and Ubx in J. coenia.Representative examples of knockdown phenotypes obtained in forewings with fz2 DsiRNA (A-C) and in hindwings with Ubx DsiRNA (D-G).Effects on CSS reduction (fz2) and partial homeoses of hindwings into forewings (Ubx) are consistent with expectations from CRISPR-induced KO phenotypes and validate our implementation of the RNAi electroporation method.

Fig. S21 .
Fig. S21.Genotyping of G 0 crispants.(A) Detection of a single frameshift fz2 mutation clone in a V. cardui crispant by direct Sanger sequencing.No wildtype sequence was recovered in this individual, implying the existence of a second mutant allele that was not amplified by PCR.DNA was directly amplified with the Phire Tissue Direct PCR Master Mix.(B) Detection of multiple frameshift mutations in two D. plexippus fz2 crispants following colony PCR.(C-D) Detection ofindels at the predicted sgRNA cutting sites in V. cardui fz1 and fz3 crispants using sequence trace decomposition with the TIDE tool (Brinkman et al., 2014).Aberrant sequencing signals after the predicted cut site are due to the presence of indel alleles, resulting in mixed chromatograms.(E) Detection of a single frameshift fz4 mutation clone in a V. cardui crispant by direct Sanger sequencing.