Consequences of Hoxb1 Duplication in Teleost Fish
Total Page：16
File Type：pdf, Size：1020Kb
EVOLUTION & DEVELOPMENT 9:6, 540–554 (2007) Consequences of Hoxb1 duplication in teleost fish Imogen A. Hurley,a Jean-Luc Scemama,b and Victoria E. Princea,c,Ã aDepartment of Organismal Biology and Anatomy, The University of Chicago, 1027 East 57th Street, IL 60637, USA bDepartment of Biology, Howell Science Complex, East Carolina University, Greenville, NC 27858, USA cCommittees on Developmental Biology, Neurobiology and Evolutionary Biology, The University of Chicago, 1027 East 57th Street, Chicago, IL 60637, USA ÃAuthor for correspondence (email: [email protected]) SUMMARY Vertebrate evolution is characterized by gene bass. Consistent with this theory, we found that the ancestral and genome duplication events. There is strong evidence that Hoxb1 expression pattern is subdivided between duplicate a whole-genome duplication occurred in the lineage leading to genes in a largely similar fashion in zebrafish, medaka, and the teleost fishes. We have focused on the teleost hoxb1 striped bass. Further, our analysis of hoxb1 genes reveals that duplicate genes as a paradigm to investigate the conse- sequence changes in cis-regulatory regions may underlie sub- quences of gene duplication. Previous analysis of the functionalization in all teleosts, although the specific changes duplicated zebrafish hoxb1 genes suggested they have vary between species. It was previously shown that zebrafish subfunctionalized. The combined expression pattern of the hoxb1 duplicates have also evolved different functional two zebrafish hoxb1 genes recapitulates the expression capacities. We used misexpression to compare the functions pattern of the single Hoxb1 gene of tetrapods, possibly due of hoxb1 duplicates from zebrafish, medaka and striped bass. to degenerative changes in complementary cis-regulatory Unexpectedly, we found that some biochemical properties, elements of the duplicates. Here we have tested the which were paralog specific in zebrafish, are conserved in both hypothesis that all teleost duplicates had a similar fate post duplicates of other species. This work suggests that the fate of duplication, by examining hoxb1 genes in medaka and striped duplicate genes varies across the teleost group. INTRODUCTION remodeling post duplication is the antifreeze proteins in Ant- arctic fish (Chen et al. 1997; Cheng and Chen 1999). Another Duplication events can occur at the level of individual genes, example is the duplication of the RNAse1 gene in a leaf chromosomal segments, and even entire genomes. Recently, eating monkey (Zhang et al. 2002). Gene or genome duplica- the availability of genome-wide sequence data has led to the tion would therefore provide the raw genetic material for identification of whole-genome duplication events during the the evolution of novelty. However, the likelihood of ne- evolution of organisms as diverse as plants (e.g., rice, Yu et al. ofunctionalization is predicted to be insufficient to account for 2005), fungi (e.g., yeast, Kellis et al. 2004), and animals (e.g., the scale of duplicate retention identified in recent analyses fish, Jaillon et al. 2004). We now have evidence that wide- (Lynch and Conery 2000; Lynch and Force 2000). spread duplication events are an integral part of genome evo- Neofunctionalization events are thought to be exceptional lution. What is less clear is the fate of preserved duplicates cases rather than the norm because beneficial mutations are following the duplication event. much less likely than degenerative mutations. The most likely fate for a duplicate gene is nonfunction- Alternative hypotheses have been proposed to explain how alization, whereby one of a pair of duplicates acquires de- duplicate genes might be preserved. Following a duplication generate mutations leading to its transformation into a event, both duplicates could acquire complementary degen- pseudogene, or its eventual complete disappearance from erative mutations so that both genes are required to fulfill all the genome. Classically, if both duplicates were retained, it the functions of the single ancestral gene (reviewed by Prince was assumed that one gene had acquired beneficial mutations and Pickett 2002). Degenerative mutations could be fixed in a which led to a new, positively selected function, in a process reciprocal and neutral manner in the cis-regulatory or coding known as neofunctionalization (Ohno 1970). The other du- regions of genes, leading to the loss of ancestral gene sub- plicate gene, its paralog, would also be retained because it functions. This process is known as subfunctionalization and continued to perform the function of the single ancestral gene. the preservation of duplicates in this way has been formalized An example of neofunctionalization through extensive protein in the Duplication, Degeneration and Complementation 540 & 2007 The Author(s) Journal compilation & 2007 Blackwell Publishing Ltd. Hurley et al. hoxb1 duplication in teleost ¢sh 541 (DDC) model (Force et al. 1999). Following preservation by (Tetraodon nigroviridis and Takifugu rubripes,orderTetra- DDC processes, the duplicate genes would no longer be under dontiformes). Unfortunately, restricted access to pufferfish the same selection pressures as one another and would there- embryos is a limitation of these species as an experimental fore be free to evolve along their own novel trajectories. system. We have chosen to perform a detailed study of dupli- In this way, subfunctionalization could facilitate neo- cate gene evolution in three other teleost species that are more functionalization (Mazet and Shimeld 2002; Prince and Pick- amendable to experimentation (Fig. 1A): the ostariophysan ett 2002) or the fine tuning of the duplicate members of a pair zebrafish (Danio rerio, order Cypriniformes) and the to their specific subfunctions (Roth et al. 2007). Equally, du- acanthopterygian fishes, medaka (Oryzias latipes,orderBel- plicates, which were originally preserved by neofunctionalizat- oniformes), and striped bass (Morone saxatilis, order Per- ion, may retain redundant subfunctions and therefore have ciformes). The divergence events which split the Ostariophysi the potential to undergo subfunctionalization after their initial and Acanthopterygii occurred at least 150 Ma according to preservation. Examples of duplicate genes believed to be sub- fossil evidence (reviewed by Benton and Donoghue 2007) functionalized are engrailed 1 (Force et al. 1999), Mitf (Lister whereas molecular estimates suggest that this divergence oc- et al. 2001; Altschmied et al. 2002), sox9 (Cresko et al. 2003), curred at least 100 Ma earlier (e.g., Peng et al. 2006). and POMC (de Souza et al. 2005). We have focused on the Hox genes to investigate the con- Individual examples of subfunctionalization continue to be sequences of gene duplication. Hox genes encode transcription described, but a comparative analysis is necessary to assess factors that are responsible for providing positional identity whether duplicates are subfunctionalized in the same manner along the anteroposterior axis of the developing embryo in different lineages following a duplication event. This is an (McGinnis and Krumlauf 1992; de Rosa et al. 1999). These important question to address because research typically fo- genes are arranged in clusters and the original evidence for the cuses on a few model organisms, which we assume are rep- presence of a fish-specific whole-genome duplication came resentative of a wider range of species. Also, it is clear that from the observation that zebrafish has almost twice the num- duplication events play an extensive role in genome evolution ber of Hox clusters as mammals (7 vs. 4; Amores et al. 1998; and it has been hypothesized that these events could facilitate Prince et al. 1998a). Hox cluster duplication has been followed the evolution of phenotypic novelty. If we are to understand by gene loss in all teleosts and different species have retained this process, we must first understand how genes evolve post different complements of Hox genes (reviewed by Prohaska duplication in different species. and Stadler 2004). However, many duplicate Hox genes have Fishes are an excellent model in which to explore this phe- been retained, making these genes ideal systems to study the nomenon (reviewed by Hurley et al. 2007). It is now widely fate of duplicate genes following the duplication event. accepted that a whole-genome duplication event occurred in Subfunctionalization of duplicate teleost Hox genes has the ray-finned fish lineage following its split from the lobe- previously been identified by comparison of the duplicates finned fishes. Supporting evidence comes from whole-genome with the single ancestral gene, which must be inferred by sequencing projects in multiple fish species, which have iden- comparison with an outgroup species. This principle is exem- tified blocks of anciently duplicated genes (pufferfish, Jaillon et plified by a study of hoxb5 duplicates in zebrafish (Bruce et al. al. 2004; medaka, Naruse et al. 2004; zebrafish, Woods et al. 2001), where the combined expression of hoxb5a and hoxb5b 2005). We also now have an improved knowledge of the timing resembles the expression of the single Hoxb5 gene in tetra- and phylogenetic position of this event. It had been estimated pods, suggesting that both duplicates were necessary to reca- that the whole-genome duplication event occurred in excess of pitulate the ancestral expression pattern (Bruce et al. 2001). 270 million years (Ma) ago (Taylor et al. 2001; Christoffels et The biochemical functions of the duplicates