Julia Ainsworth

The evolution of complex brains required the incorporation of newly evolved neurons into existing circuits, yet the genetic and developmental mechanisms enabling their integration remain poorly understood. Butterflies expanded their color vision by adding a second R7 photoreceptor per ommatidium, a rare departure from the conserved insect eye ground plan. Because each R7 makes an independent stochastic fate choice, this duplication increased the number of ommatidial types from two to three. We identified the genetic basis of this expansion and demonstrated how the brain accommodates the new input. R7 duplication was achieved by converting R3/4 cells to an R1/6 fate, triggering the R7 recruitment program on both sides of the ommatidium. By manipulating transcription factor expression in Drosophila , we engineered a “butterfly-fly” that recapitulates this ancient R7 duplication and three-type stochastic mosaic. In the fly brain, two R7 subtypes connect to specific types of Dm8 neurons, which are born in excess and undergo apoptosis if they fail to find synaptic partners. In our butterfly-fly model, these surplus Dm8s are immediately rescued and form subtype-appropriate connections. These findings suggest that population-level variation maintains a reservoir of potentially interacting neurons, providing a developmental substrate that allows the brain to immediately accommodate newly evolved neurons.

The evolution of complex brains required existing neurons and neural circuits to accommodate new inputs. The genetic and developmental mechanisms that enable such integration are largely unknown. Butterflies evolved more complex retinal mosaics through the addition of a second R7 color photoreceptor per ommatidium (unit eye). In Drosophila , the unique R7 makes a stochastic choice to express one of two opsin genes. In butterflies, the two R7s make independent stochastic cell fate choices in each ommatidium, producing three ommatidial types instead of two. Here, we investigate the developmental basis of this change and how the butterfly brain accommodates expanded sensory receptor input. We first identified the changes in gene expression that cause a second R7 cell to be specified. We then modified Drosophila retinas to have butterfly-like transcription factor expression, causing recruitment of an additional R7. The two R7s make independent stochastic choices, like butterflies, leading to three stochastically distributed ommatidial types. In Drosophila , the two R7 subtypes connect to their target neurons, either yDm8 or pDm8. Dm8 neurons of both types are born in excess and Dm8s that do not find connections with their cognate y or pR7s undergo apoptosis. In the presence of extra R7s in butterfly-like fly retinas, additional Dm8s are retained, leading to two Dm8s per medulla column that make appropriate connections with the matching R7 subtypes, facilitating the expansion of color vision. We propose that the presence of cells that would otherwise die provide developmental flexibility that can allow brains to accommodate newly evolved inputs.

A spectacular diversity of forms and features allow species to thrive in different environments, yet some structures remain relatively unchanged. Insect compound eyes are easily recognizable despite dramatic differences in visual abilities across species. It is unknown whether distant insect species use similar or different mechanisms to pattern their eyes or what types of genetic changes produce diversity of form and function. We find that flies, mosquitos, butterflies, moths, beetles, wasps, honeybees, and crickets use homologous developmental programs to pattern their retinas. Transcription factor expression can be used to establish homology of different photoreceptor (PR) types across the insects: Prospero (Pros) for R7, Spalt (Sal) for R7+R8, and Defective proventriculus (Dve) for R1-6. Using gene knockout (CRISPR/Cas9) in houseflies, butterflies, and crickets and gene knockdown (RNAi) in beetles, we found that like Drosophila , EGFR and Sevenless (Sev) signaling pathways are required to recruit motion and color vision PRs, though Drosophila have a decreased reliance on Sev signaling relative to other insects. Despite morphological and physiological variation across species, retina development passes through a highly conserved phylotypic stage when the unit eyes (ommatidia) are first patterned. This patterning process likely represents an “insect eye ground plan” that is established by an ancient developmental program. We identify three types of developmental patterning modifications (ground plan modification, nonstochastic patterns, and specialized regions) that allow for the diversification of insect eyes. We suggest that developmental divergence after the ground plan is established is responsible for the exceptional diversity observed across insect visual systems.

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.

Butterfly color patterns provide visible and biodiverse phenotypic readouts of the patterning processes that occur in a developing epithelium. While the secreted ligand WntA was shown to instruct the color pattern formation in butterflies, its modes of reception and signal transduction remain elusive. Butterfly genomes encode four homologues of the Frizzled-family of Wnt receptors. Here we show that CRISPR mosaic knock-outs of frizzled2 ( fz2 ) phenocopy the color pattern effects of WntA loss-of-function in multiple nymphalids. While WntA mosaic clones result in intermediate patterns of reduced size, consistently with a morphogen function, fz2 clones are cell-autonomous. Shifts in pupal expression in WntA crispants show that WntA and fz2 are under positive and negative feedback, respectively. Fz1 is required for Wnt-independent planar cell polarity (PCP) 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.