In this post I will discuss our recent paper entitled “Leg length and bristle density, both necessary for water surface locomotion, are genetically correlated in water striders” . Probably all of you have ever encountered one of those elegant semi-aquatic bugs that, as their vernacular names reflect, glide, measure, scoot, skate, skim, skip, stride, tread over the surface of all water surfaces on the planet (ocean included).
Semi-aquatic bugs (or Gerromorpha) likely derived from a terrestrial ancestor that evolved the ability to stand and move on the water-air interface about 200 million years ago. Access to previously unexploited ecological opportunities is associated with phenotypic evolution and often results in significant lineage diversification . Our favorite bugs are no exception to the rule. Early-diverging lineages of semi-aquatic bugs occupy transitional zones and walk both on land and water, whereas derived lineages evolved rowing as a novel mode of locomotion on open-water surface [3,4]. Water surface invasion is commonly viewed as a stepwise process that involved both the diversification of leg morphologies and the evolution of densely arranged water-repellent ‘hairs’ that allow the insects to exploit surface tension. However, the mechanisms by which species develop traits adapted to the new ecological niche are not well understood.
Previous studies having predominantly focused on leg morphologies, we decided to explore the other side of the coin — the leg ‘hairs’. We first shown that the leg ‘hairs’ are nothing else than bristles . Then the central question remains still on the table: how do semi-aquatic bugs evolve such a high density of bristles on their legs? Here started our investigation to find out what makes a water strider so hairy. After several years of leg-focused research, the Khila lab was quite excited and well determined to enter the bristle world.
Candidate gene approach
We first compiled a list of more than 120 genes known to play a role in bristle development in Drosophila, and searched for gene duplication events in the Gerromorpha. We found that the gene Beadex has two copies in the Gerromorpha. Whereas the copie BxA plays a role in bristle development in Gerris buenoi (= species with high bristle density), BxA is not significantly expressed during embryogenesis in Mesovelia mulsanti (= species with low bristle density). This result suggests that differences in leg bristle density might be partly attributable to differential expression level of BxA. We also identified two copies of the gene taxi in the semi-aquatic bugs that result from a duplication in the lineage leading to the Gerromorpha. We found that the gene taxiB evolves faster than taxiA and we detected positive selection along the taxiB lineage, suggesting functional divergence of the two copies. We depleted taxiA or taxiB transcripts using RNAi, and we observed defects in the development of leg bristles. However, we found that only ds-taxiB knockdown individuals exhibited shorter legs. Again, we were here to focus on the bristles, so we did not linger on the leg phenotype.
Genes involved leg bristle development in Gerromorpha…
By combining comparative transcriptomics and RNA interference, we identified six genes whose role in bristle development had never been documented in model organisms. We unraveled the role of the GPN-loop GTPase 2, the c-Myc binding protein MYCBP, and a protein-glutamate O-methyltransferase in bristle patterning; the role of the bHLH transcription factor Net and the MAP kinase signal responder protein Dodo in bristle elongation; and the role of the actin binding protein Simiate in bristle orientation and size.
… played also a role in leg growth!
Looking for bristle genes, finding bristle genes! It sounded like a pretty straightforward research project. Well, the truth turned out to be much more complicated and exciting. We found that suppressing the expression of bristle-related genes resulted in a leg shortening. This time, the leg phenotype was so predominant that we could not overlook it. We spent a substantial number of hours with our pictures of legs and our ruler. The quantification was unequivocal: most of the legs in the knockdown individuals were shorter. In brief, shorter and barer, or barer and shorter!
Cell division as a shared molecular mechanism
This hypothesis remained to be verified. To this end, we stained all nuclei with DAPI, and M-phase nuclei with anti-PH3 antibodies in the developing embryonic legs. We then compared the cell division rate between control and ds-taxiB knockdown individuals. We found that the legs T2, which are shorter in ds-taxiB individuals, showed a reduced mitotic activity. By doing so, we noticed that the orientation of cell division might also be affected. Thus, we measured the angle of division relative to the proximo-distal (PD) axis of the leg. We found that epithelial cells bias their orientation perpendicular to the PD axis in knockdown individuals, whereas these cells predominantly orient their divisions parallel to the PD axis in the control embryos.
Pleiotropy as a facilitator of diversification
The key message of the project began to take shape: leg length and bristle density, both necessary for water surface locomotion, are genetically correlated in the semi-aquatic bugs. We had the feeling we could go even a bit further. We plotted the bristle density against the leg length for controls and knockdown individuals of G. buenoi: linear correlation, checked! Because bristle density and leg length vary a lot across the semi-aquatic bugs, we extended our correlation analysis to all the species we had in the laboratory. The correlation between our two favorite traits still stands at the infraorder level.
The classical view of the origin of the semi-aquatic bugs implies the stepwise evolution of longer legs and denser leg bristles. Our findings suggest that the invasion of water surface might have been more ‘straightforward’ because of genetic pleiotropy. However, even if leg length and bristle density are both necessary for water surface locomotion, it does not necessarily mean these two traits have been selected for. Indeed, we cannot exclude that only one trait was selected whereas the other trait was simply a spandrel or by-product . Our study represents the first step towards the deciphering of the molecular mechanism of high bristle density in the Gerromorpha, and further investigation will be needed to build a better picture of this ecologically relevant trait. Nevertheless, one thing is certain: in Gerromorpha, it is all about leg length!
 Finet C, Decaras E, Rutkowska M, Roux P, Collaudin S, Joncour P, Viala S, Khila A. (2022). Leg length and bristle density, both necessary for water surface locomotion, are genetically correlated in water striders. PNAS 119: e2119210119.
 Schluter D. (2000). The ecology of adaptive radiation. Oxford University Press, Oxford.
 Andersen NM. (1976). A Comparative Study of Locomotion on the Water Surface in Semiaquatic Bugs (Insecta, Hemiptera, Gerromorpha). Vidensk. Meddel. Dansk Naturhist. Foren. Kjobenhavn 139: 337-396.
 Crumière AJJ, Santos ME, Sémon M, Armisen D, Moreira FFF, and Khila A. (2016). Diversity in Morphology and Locomotory Behavior Is Associated with Niche Expansion in the Semi-aquatic Bugs. Curr. Biol. 26: 3336-3342.
 Finet C, Decaras, Armisén D, Khila A. (2018). The achaete–scute complex contains a single gene that controls bristle development in the semi-aquatic bugs. Proc. R. Soc. B 285: 20182387.
 Gould SJ, Lewontin RC. (1979). The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proc. R. Soc. B 205: 581-598.