Ultrabithorax modifies a regulatory network of genes essential for butterfly eyespot development in a wing sector-specific manner

Abstract

SummaryNymphalid butterfly species often have a different number of eyespots in forewings and hindwings but how the hindwing identity gene, Ultrabithorax (Ubx), drives this asymmetry, is not fully understood. It is also unclear why eyespot serial homologs originated initially only in hindwings. To address these questions, we examined a three-gene regulatory network (GRN) for eyespot development in the hindwings of Bicyclus anynana butterflies, and compared it to the same network previously described for forewings. We also examined how Ubx interacted with each of these three eyespot-essential genes. We found similar genetic interactions between the three genes in fore and hindwings, but we discovered three regulatory differences: Antennapedia (Antp) merely enhances spalt (sal) expression in the eyespot foci in hindwings but is not essential for sal activation, as in forewings; Ubx up- regulates Antp in all hindwing eyespot foci but represses Antp outside these wing regions and; Ubx is regulating sal expression in a wing-sector specific manner, i.e., it does not affect sal expression in the wing sectors that contain eyespots in both fore and hindwings, but it positively regulates sal expression in the sectors that have hindwing-specific eyespots. These results suggest that Ubx, or its downstream targets, might have paid a role in the origin of eyespots, restricted initially to hindwings, via the positive regulation of sal, an essential gene for eyespot development. We propose that Antp got co-opted into the eyespot GRN at a later stage by coming under Ubx regulation. This Hox gene redundancy, together with a novel positive feedback loop between sal and Antp, might have allowed Antp to functionally replace Ubx in forewings and lead to the origin of forewing eyespots. Outside the eyespot focal regions, we discovered that Ubx is up-regulated by Distal-less along the wing margin. We propose a model for how the regulatory connections between these genes might have evolved to produce wing- and sector-specific variation in eyespot number.