View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Dispatch R125

the locust and modelled on a robot insect pitches, rolls or rotates, 6. Ijspeert, A.J., Crespi, A., and Cabelguen, J.M. (2005). Simulation and [11]. The same principle has been ventral optic flow will be distorted. robotics studies of salamander used to enable a small free model Can the animal measure and locomotion. Applying neurobiological aircraft to avoid collisions with discount these movements, or are principles to the control of locomotion in robots. Neuroinformatics 3, walls [12]. other sensorimotor loops, such as 171–196. Exploiting some aspect of optic the optomotor reflex, deployed 7. Collins, S., Ruina, A., Tedrake, R., and Wisse, M. (2005). Efficient bipedal robots flow directly to control locomotion simultaneously to minimise them based on passive dynamic walkers. is often discussed, following [14]? Is there any evidence of Science 307, 1082–1085. Gibson [13], as an example of an systematic difference, for example 8. Webb, B. (2000). What does robotics offer animal behaviour? Anim. Behav. 60, ‘affordance’. The exact meaning of in the sensitivity range, of 545–558. this term is subject to debate, but elementary motion detectors 9. Baird, E., Srinivasan, M., Zhang, S., and Cowling, A. (2005). Visual control of flight essentially it is the idea that what pointing at different parts of the speed in honeybees. J. Exp. Biol. 208, animals are designed to perceive visual field [15] that would fit with 3895–3905. are opportunities for action, rather the proposed difference in control 10. Srinivasan, M.V., Chahl, J.S., Weber, K., Nagle, M.G., and Zhang, S.W. (1999). than action-neutral properties of function? What exactly are the Robot navigation inspired by principles of the environment around them. wing movements that need to be insect systems. Robotics Auton. Sys. 26, 203–216. Instead of seeing shape, size and controlled [16] and might these 11. Blanchard, M., Rind, F.C., and distance of an object, for example, also be ‘matched’ to specific Verschure, P.F.M.J. (2000). Collision we observe its graspability. This control problems? Will avoidance using a model of the locust LGMD neuron. Robotics Auton. Sys. 30, influential, and sometimes understanding the basic control 17–38. controversial, view of perception is rules help us to trace out the neural 12. Zufferey, J.C., and Floreano, D. (2006). Fly-inspired visual steering of an ultralight particularly relevant to robotics, pathways that support this indoor aircraft. IEEE Trans. Robotics 22, where specialised sensory behaviour? The combination of 137–146. systems for cues such as optic behavioural experiments and 13. Gibson, J.J. (1966). The senses considered as perceptual systems flow have often proved more useful robot models is likely to be an (Boston: Houghton Mifflin). than conventional computer important tool in future 14. Neumann, T.R. and Bu¨ lthoff, H.H., (2001). Insect inspired visual control of vision. In one sense these may be discoveries. translatory flight. Proceedings of thought of as tricks or short-cuts Advances in Artificial Life: 6th European that enable the animal or robot to References Conference, ECAL 2001, Prague, Czech 1. Wehner, R. (1987). ‘Matched filters’ — Republic, September 10–14. avoid difficult measurements neural models of the external world. 15. Egelhaaf, M., Kern, R., and Krapp, H.G. and calculations. But the concept J. Comp. Physiol. A. 161, 511–531. (2002). Neural encoding of 2. Franceschini, N., Ruffier, F., and Serres, J. behaviourally relevant visual-motion of action-oriented perception (2007). A bio-inspired flying robot sheds information in the fly. Trends Neurosci. may also be important in light on insect piloting abilities. Curr. Biol. 25, 96–102. understanding higher level 17, 329–335. 16. Dickinson, M.H. (2006). Insect flight. Curr. 3. Heran, P., and Lindauer, M. (1963). Biol. 16, 309–314. cognitive skills, as it strongly Windkompensation und determines how we structure the Seitenwindkorrektur der Bienen beim Flug Uber} Wasser. Z. vergl. Physiol. 47, Institute of Perception, Action and world around us. 39–55. Behaviour, University of Edinburgh, For insect flight control, many 4. Srinivasan, M.V., Zhang, S., Chahl, J.S., JCMB, Kings Buildings, Mayfield Road, issues remain to be resolved. Barth, E., and Venkatesh, S. (2000). Edinburgh, Eh9 3JZ, UK. How honeybees make grazing landings Ventral optic flow can be easily on flat surfaces. Biol. Cyb. 83, 171–183. E-mail: [email protected] detected if the animal is flying 5. Hu, D.L., Chan, B., and Bush, J.W.M. (2003). The hydrodynamics of water straight ahead and the sensor is strider locomotion. Nature 424, pointing straight down. But if the 663–666. DOI: 10.1016/j.cub.2006.12.008

Speciation Genetics: Epistasis, hybrid sterility? As with many problems, Darwin struggled with Conflict and the Origin of Species the genetic details but, in the end, got the basics right: hybrid sterility ‘‘is not a specially endowed Evolutionary biologists have long recognized that the sterility and quality, but is incidental on other inviability of species hybrids must involve incompatible epistatic acquired differences,’’ (p. 245) interactions between two (or more) genes. The first pair of such and is caused by a hybrid’s hybrid incompatibility genes has now been identified. ‘‘organization having been disturbed by two organizations Daven C. Presgraves Species [1] to hybrid sterility having been compounded into because, being scrupulous, he one’’ (p. 266). Species are often reproductively wished to confront the possible Fifty years would pass before isolated from one another by the shortcomings of his theory head Bateson [2] and later Dobzhansky intrinsic sterility or inviability of on: why would natural selection, [3] and Muller [4,5] devised their hybrids. Darwin devoted an which acts to increase individual a genetic model for the evolution entire chapter of his Origin of fitness, cause the evolution of of such hybrid fitness problems Current Biology Vol 17 No 4 R126

Alfred Sturtevant discovered in evolving protein with putative aabb 1919 [8], is that D. melanogaster DNA-binding properties. Ancestor can be crossed to its sister species Introducing the wild-type D. simulans. The bad news is that, D. melanogaster allele, Hmrmel, to Sturtevant’s great kills hybrids [13–15]. disappointment [9], all hybrids With Hmr identified, Brideau between the two species are sterile et al. [7] turned to Lhr, Hmr’s or dead [10]. Crosses between putative partner gene in this hybrid D. melanogaster females and incompatibility. After crudely AAbb X aaBB Species 1 Species 2 D. simulans males, for example, mapping Lhr to a region spanning produce only sterile hybrid approximately 37 genes, the daughters, as hybrid sons die authors narrowed the candidates AaBb during the transition from larvae according to two criteria: elevated Hybrid: to pupae (Figure 2). molecular divergence and sterile / inviable? An important break came in predicted DNA-binding Current Biology 1979, when Takao Watanabe [11] properties. The gene CG18468 discovered a strain of D. simulans fit the bill. Figure 1. The Dobzhansky–Muller model that, when crossed to Tellingly, the two rescue strains for loss of fitness in hybrids between D. melanogaster females, of Lhr have conspicuous lesions recently diverged species. produces viable hybrid sons. in CG18468. First, Lhr1 has a 4 A common ancestor species with the two-locus genotype, aabb, splits into Watanabe mapped this Lethal kilobase retrotransposon insertion 0 two independently evolving lineages. In hybrid rescue (Lhr) gene to the in the 5 untranslated region that one, mutation A arises and spreads to second chromosome by simply cripples CG18468 expression. fixation; in the other, mutation B arises crossing rescue (Lhr) and Second, in addition to 20% amino and spreads to fixation. The A and B mu- non-rescue (Lhr+) D. simulans acid divergence, the protein tations have never coexisted in the same strains and scoring recombinants product of wild-type CG18468sim genome and thus may not be function- ally compatible when combined in spe- for their ability to sire hybrid sons. also has a 16 amino acid insertion mel cies hybrids. These results opened new that is absent from CG18468 ; possibilities for the genetic Lhr2 precisely lacks this insertion, (Figure 1). Today we would say that analysis of hybrid inviability suggesting that these 16 amino hybrid sterility and inviability are between D. melanogaster and D. acids may be responsible for caused by incompatible epistatic simulans in two ways. For one, hybrid inviability. In the key interactions between genes that because F1 hybrid inviability can be experiment, expressing wild-type have functionally diverged rescued by a single mutation, its CG18468sim suppressed Lhr1 between species. The so-called genetic basis might be simple. For rescue of hybrid males: that is, Dobzhansky–Muller model shows another, rescue mutations could be CG18468sim complements Lhr1, how such hybrid incompatibilities genetically mapped and identified proving that CG18468 sim is evolve as incidental byproducts within species rather than in Lhrsim. of adaptive or neutral genetic hybrids. Similar searches in Other individual incompatibility divergence (Figure 1). This simple D. melanogaster turned up the genes have been identified [16], model now forms the basis of X-linked rescue mutation, Hybrid but this is the first time that two virtually all work on hybrid sterility male rescue [12], which (as its epistatically interacting partners and inviability [6]. name implies) also rescues hybrid of a hybrid incompatibility have Another fifty years would pass sons. been identified: Hmrmel is before speciation geneticists were From the beginning [10] incompatible with Lhrsim, causing able to replace the As and Bs of (Figure 2), it has been clear that F1 hybrid male inviability Figure 1 with actual genes. In hybrid inviability involves (Figure 2). a recent paper, Brideau et al. [7] a recessive X-linked allele from But speciation genetics is about report the identification of the D. melanogaster that is more than just filling in the As and second of a pair of interacting incompatible with a dominant the Bs of the Dobzhansky–Muller hybrid incompatibility genes. To autosomal allele from D. simulans. model. Identifying hybrid appreciate this achievement, It was therefore promising that the incompatibility genes answers we have to understand why, two rescue mutations mapped to questions that were previously until now, the genetics of the D. melanogaster X(Hmr) and inaccessible. First, we can say speciation has lagged so far a D. simulans autosome (Lhr): something about the evolutionary behind the genetics of everything could these rescue mutations be histories of the genes. The else. compatible alleles at the otherwise sequences of Hmr and Lhr, like First, it is difficult to do genetics incompatible loci? A series of those of other incompatibility in species hybrids — hybrids that studies by Barbash and genes [17–19], show signatures of are often sterile or dead. Second, colleagues, culminating in the adaptive evolution. This is one of this problem is particularly acute in new Brideau et al. [7] study, shows the most important findings of the the workhorse of evolutionary that they are. In previous work, last decade of speciation research: genetics, the fruitfly Drosophila Barbash et al. [13] showed that hybrid fitness problems are melanogaster. The good news, as wild-type Hmr encodes a rapidly incidental byproducts of adaptive Dispatch R127

3. Dobzhansky, T. (1937). Genetics and the Origin of Species (New York: Columbia D. melanogaster D. simulans University Press). 4. Muller, H.J. (1940). Bearing of the Drosophila work on systematics. In The New Systematics, J.S. Huxley, ed. x (Oxford: Clarendon Press), pp. 185–268. 5. Muller, H.J. (1942). Isolating mechanisms, evolution, and temperature. Biol. Symp. 6, 71–125. 6. Coyne, J.A., and Orr, H.A. (2004). Speciation (Sunderland, Massachusetts: Hmr mel Lhr mel Hmr mel Lhr mel Sinauer). 7. Brideau, N.J., Flores, H.A., Wang, J., Maheshwari, S., Wang, X., and Barbash, D.A. (2006). Two Dobzhansky- Muller genes interact to cause hybrid Hmr sim Lhr sim Lhr sim lethality in Drosophila. Science 314, 1292–1295. 8. Sturtevant, A.H. (1919). A new species closely resembling Drosophila melanogaster. Psyche. 26, 153–155. Lethal incompatibility 9. Provine, W.B. (1991). Alfred Henry Current Biology Sturtevant and crosses between Drosophila melanogaster and Drosophila Figure 2. Inviability of hybrids between D. melanogaster and D. simulans. simulans. Genetics 129, 1–5. 10. Sturtevant, A.H. (1920). Genetic studies When D. melanogaster females (red) are crossed to D. simulans males (blue), only ster- on Drosophila simulans. I. Introduction. ile hybrid daughters are produced because hybrid sons die. Left bars, sex chromo- Hybrids with Drosophila melanogaster. somes; right bars, second chromosome; small hooked bar, Y chromosome. Hmrmel Genetics 5, 488–500. is incompatible with Lhrsim, causing lethality of hybrid males. Hybrid daughters are 11. Watanabe, T.K. (1979). A gene that mel sim rescues the lethal hybrids between viable because they are heterozygous Hmr /Hmr . Drosophila melanogaster and D. simulans. Japn. J. Genet. 54, 325–331. divergence, just as Darwin be required for Hmrmel–Lhrsim 12. Hutter, P., and Ashburner, M. (1987). Genetic rescue of inviable hybrids imagined. hybrid inviability. But if this were between Drosophila melanogaster and Second, besides telling us about true, it seems likely that a third its sibling species. Nature 327, the genetics of speciation, the locus would have turned up in 331–333. 13. Barbash, D.A., Siino, D.F., Tarone, A.M., identities of hybrid incompatibility previous screens for hybrid rescue and Roote, J. (2003). A rapidly evolving genes also say something about mutations. Myb-related protein causes species isolation in Drosophila. Proc. Natl. Acad. the biological basis of functional An alternative possibility, Sci. USA 100, 5302–5307. divergence. Hmr and Lhr encode favored by Brideau et al. [7], is that 14. Barbash, D.A., Roote, J., and putative DNA-binding proteins. Hmrmel–Lhrsim lethality depends, Ashburner, M. (2000). The Drosophila melanogaster Hybrid male rescue gene Brideau et al. [7] show that the LHR not on a third major locus, but on causes inviability in male and female protein interacts and co-localizes the cumulative effects of many species hybrids. Genetics 154, 1747–1771. with Heterochromatin Protein 1 differences in heterochromatic 15. Orr, H.A., and Irving, S. (2000). Genetic (HP1) in heterochromatic regions repeats and transposons. This is analysis of the Hybrid male rescue of the genome, especially at an interesting hypothesis but one locus of Drosophila. Genetics 155, 225–231. centromeres, telomeres and the that may prove difficult to test. 16. Orr, H.A., Masly, J.P., and dot-like fourth chromosome. Nevertheless, these new findings, Presgraves, D.C. (2004). Speciation genes. Curr. Opin. Genet. Dev. 14, Notably, these regions are along with those from other hybrid 675–679. characterized by a high rate of incompatibility genes [20], suggest 17. Barbash, D.A., Awadalla, P., and evolutionary turnover of something that Darwin could Tarone, A.M. (2004). Functional divergence caused by ancient positive repetitive sequences and never have imagined. In particular, selection of a Drosophila hybrid transposable elements. The genetic conflicts — involving incompatibility locus. PLoS Biol. 2, 839–848. authors speculate that Lhr’s selfish transposons, repetitive 18. Presgraves, D.C., Balagopalan, L., divergence may result from its DNAs, drive elements and Abmayr, S.M., and Orr, H.A. (2003). coevolution with these so forth— rather than good old- Adaptive evolution drives divergence of a hybrid inviability gene between two heterochromatic sequences. fashioned ecological adaptation species of Drosophila. Nature 423, This leads us to one of the could drive the divergence that 715–719. 19. Ting, C.-T., Tsaur, S.-C., Wu, M.-L., and surprises of the study. Based on in turn gives rise to hybrid Wu, C.-I. (1998). A rapidly evolving a simple two-locus model, Hmrmel incompatibilities and, homeobox at the site of a hybrid sterility should be incompatible with Lhrsim, consequently, to species gene. Science 282, 1501–1504. 20. Presgraves, D.C., and Stephan, W. (2007). causing inviability; however, differentiation. Pervasive adaptive evolution among driving Lhrsim expression in interactors of the Drosophila hybrid a pure D. melanogaster genetic References inviability gene, Nup96. Mol. Biol. Evol. 24, 1. Darwin, C. (1859). On the Origin 306–314. background does not cause of Species by Means of Natural sim inviability — Lhr only kills in Selection or the Preservation of Favored Department of Biology, University of a hybrid genetic background. This Races in the Struggle for Life (London: Rochester, Rochester, New York 14627, Murray). implies that a simple two-locus 2. Bateson, W. (1909). Heredity and USA. model is not sufficient for this variation in modern lights. In Darwin E-mail: [email protected] and Modern Science, A.C. Seward, ed. incompatibility. The appropriate (Cambridge: Cambridge University genotype at some third locus may Press), pp. 85–101. DOI: 10.1016/j.cub.2006.12.030