Developmental Evolution: Downsizing Wings in the Flightless Emu

Abstract

The vestigial wings of emus are a striking illustration of morphological evolution. A new study points to reduced activity of an essential signaling pathway as a factor in the evolution of the emu’s stunted wings.

The reduction or loss of body structures is a repeated theme in evolution. Examples, such as armor loss in sticklebacks and eye reduction in subterranean cavefish, have become better understood through modern genetic and genomic approaches [1,2]. In birds, powered flight was a key innovation that was subsequently lost in multiple avian lineages [3–5]. Among the avian groups that evolved flight loss are ratites, an assemblage of flightless birds that includes emus, cassowaries, kiwis, rheas, ostriches, and the extinct moa. Phylogenetic studies indicate that tinamous, a group of flight-capable birds, are nested within the ratite lineage and are most closely related to the moa [6]. For this reason, it is thought that the last common ancestor of ratites and tinamous was most likely capable of flight, with separate ratite lineages then independently losing powered flight. The evolution of flight loss in these birds is associated with a reduction in wing size [7]. This reduction is most extreme in the wingless moa, least extreme in the ostrich, and intermediate in the emu, which has small, vestigial wings. A new study by Young and colleagues [8], published in this issue of Current Biology, takes a comprehensive look at emu forelimb development and gene expression to home in on the mechanisms that cause the emu’s wings to develop differently from those of flighted birds.

The chicken, a flight-capable bird, has long been a central model organism for limb research and serves as an excellent reference for investigations of limb reduction in emus and other flightless avians. Comparative embryology shows that alterations in the development of emu wing buds are apparent early, with a clear delay in wing bud outgrowth relative to the chicken and other flighted birds [9]. While simple external views of developing emu embryos demonstrate visible delays in the formation of the wing bud, the first cellular hallmarks of limb development begin even earlier. For the first time, Young et al. have examined these early cellular events in the emu and, in doing so, have discovered something unexpected – these events proceed in the emu with the same developmental timing that is observed in chickens. More specifically, the generation and recruitment of the progenitor cells that will form the wing skeleton and musculature are initiated normally. Instead, it is the subsequent cellular proliferation of progenitors that is delayed and accounts for the small size of the emu wing. But what developmental mechanisms are responsible for this stalled cellular growth?

Prior work suggested that delayed expression of the Tbx5 gene, which encodes a transcription factor essential for initiating forelimb development, might contribute to the stunted development of emu wings [10]. Further work by Farlie and coworkers also implicated Nkx2.5, a gene that isn’t normally active during early limb development but that displays novel expression in the early wing bud of emu embryos [11]. Young et al. systematically reinvestigated gene expression patterns in emu embryos to find out why cell growth is delayed during emu wing development. Since the legs of emus develop without experiencing a delay in cellular proliferation, the authors compared the transcriptomes of emu forelimb and hindlimb progenitor cells to find genes that are expressed differently between these cell populations. Some genes have conserved forelimb- or hindlimb-biased expression and are expected to be differentially expressed between these cell populations in both emu and chick. For example, Tbx4 and Pitx1 are well-known hindlimb-specific factors that show a hindlimb-biased expression in both species. Of greatest interest, however, are those genes that show altered expression only in emu forelimb cells, and transcriptome analyses revealed several genes in the fibroblast growth factor (Fgf) family that fall into this category. Since Fgf signaling plays a pivotal role in driving proliferation of the cells that contribute to the forelimbs and hindlimbs, these are intriguing limb reduction candidate genes.

Among the Fgf candidate genes identified, Fgf10 and Fgf8 were of particular interest. Each is expressed in a separate field of cells with Fgf8 expressed in the cells of the limb ectoderm and Fgf10 expressed in cells of the lateral plate mesoderm (LPM), which lie beneath the ectoderm [12]. Production of these Fgf signaling factors creates a positive feedback loop between these cell populations that maintains expression of the Fgf genes in their respective expression domains and promotes cellular proliferation. While these genes are expressed at similar levels in the chick forelimb and hindlimb, the emu shows diminished forelimb expression of both genes. Directly visualizing the expression of these genes during early emu wing development demonstrated that Fgf8 is initially not expressed in the emu forelimb, and the domain of Fgf10 expression is more restricted in emu than in chick. These deficiencies are transient, as both genes show a chicken-like expression pattern at later stages of development. When the authors transplanted Fgf10-expressing LPM cells from the chick into the emu, they found that this was sufficient to stimulate the overlying emu forelimb ectoderm to express Fgf8. Furthermore, experimentally forcing high levels of Fgf10 in the emu cells was also sufficient to induce early Fgf8 expression and the precocious development of wing buds. This suggests that reduced Fgf10 expression in emu forelimb progenitors causes delayed Fgf8 expression, resulting in lower cell proliferation and a reduced wing bud. However, this raises the question of what causes the reduced Fgf10 gene expression in the emu.

Enhancers are regulatory DNA sequences that modulate the expression of genes within specific cell types. In order to become active and allow transcription factors to bind, enhancers must adopt an accessible or open chromatin state. These regions of open chromatin can be identified via ATAC-seq [13], and the authors used this method to find species-specific chromatin differences between the forelimbs and hindlimbs of emus and chicks. Although many thousands of open chromatin regions were found in each tissue, they narrowed down their list to a small set of regions that are open in the chick forelimb, chick hindlimb, and emu hindlimb, but are in a closed chromatin state in the emu forelimb. Ultimately the authors zeroed in on a dozen regions of interest, representing potential enhancers that might explain differences in emu forelimb gene expression. One of these potential enhancers is located near Fgf10, suggesting that altered activity of this element might be the cause of the emu’s diminished Fgf10 expression.

ATAC-seq data can also identify ‘footprints’ of transcription factor binding present in regions of open chromatin. Using this footprinting method, Young et al. discovered that Ets family transcription factors were among the few motifs consistently identified in the emu hindlimb, chick hindlimb, and chick forelimb but not in the emu forelimb. This apparent absence of Ets binding in emu forelimb progenitors is notable because Fgf signaling activates Ets transcription factors. To investigate the importance of Ets factors in driving early limb gene expression, the authors focused on a limb enhancer for the Sall1 gene that is in an open state in chick forelimb and in a closed state in emu forelimb. Mutation of an Ets site within the Sall1 enhancer ablated its activity during early chicken forelimb development, though activity was recovered later in development. In addition, when placed in the emu wing bud, the wild-type chick enhancer failed to drive limb expression. These results support a model where both Fgf signaling and intact Ets binding sites are required for the activation of key enhancers during early limb development.

While previous studies have investigated the basis of wing size reduction in emus, Young and coworkers provide multiple converging lines of evidence that tell a remarkably consistent story – a reduction in Fgf signaling in early forelimb progenitor cells leads to a delay in cellular proliferation and wing bud outgrowth. As the authors point out, this does not necessarily mean that changes in Fgf signaling caused the initial evolution of flight loss. It is likely, however, that changes in Fgf signaling are one of the primary developmental mechanisms responsible for the subsequent evolution of diminutive wings in the emu. A recent investigation of wing size reduction in the flightless Galapagos cormorant implicated a preponderance of coding mutations in cilia-related genes as contributing factors in the small winged phenotype of this species [5]. Thus, different wing-reduced birds may have convergently evolved undersized forelimbs through different genetic mechanisms. Though it remains to be discovered what genetic mutations are triggering shifts in emu gene expression and enhancer activity, this study significantly expands our understanding of what sets emu wings apart from the wings of flighted birds.