oskar is the only gene in the animal kingdom necessary and sufficient for specifying functional germ cells. However, oskar has only been identified in holometabolous ("higher") insects that specify their germline using specialized cytoplasm called germ plasm. Here we show that oskar evolved before the divergence of higher insects and provide evidence that its germline role is a recent evolutionary innovation. We identify an oskar ortholog in a basally branching insect, the cricket Gryllus bimaculatus. In contrast to Drosophila oskar, Gb-oskar is not required for germ cell formation or axial patterning. Instead, Gb-oskar is expressed in neuroblasts of the brain and CNS and is required for neural development. Taken together with reports of a neural role for Drosophila oskar, our data demonstrate that oskar arose nearly 50 million years earlier in insect evolution than previously thought, where it may have played an ancestral neural role, and was co-opted to its well-known essential germline role in holometabolous insects.

The Statistical Inference of Function Through Evolutionary Relationships (SIFTER) framework uses a statistical graphical model that applies phylogenetic principles to automate precise protein function prediction. Here we present a revised approach (SIFTER version 2.0) that enables annotations on a genomic scale. SIFTER 2.0 produces equivalently precise predictions compared to the earlier version on a carefully studied family and on a collection of 100 protein families. We have added an approximation method to SIFTER 2.0 and show a 500-fold improvement in speed with minimal impact on prediction results in the functionally diverse sulfotransferase protein family. On the Nudix protein family, previously inaccessible to the SIFTER framework because of the 66 possible molecular functions, SIFTER achieved 47.4% accuracy on experimental data (where BLAST achieved 34.0%). Finally, we used SIFTER to annotate all of the Schizosaccharomyces pombe proteins with experimental functional characterizations, based on annotations from proteins in 46 fungal genomes. SIFTER precisely predicted molecular function for 45.5% of the characterized proteins in this genome, as compared with four current function prediction methods that precisely predicted function for 62.6%, 30.6%, 6.0%, and 5.7% of these proteins. We use both precision-recall curves and ROC analyses to compare these genome-scale predictions across the different methods and to assess performance on different types of applications. SIFTER 2.0 is capable of predicting protein molecular function for large and functionally diverse protein families using an approximate statistical model, enabling phylogenetics-based protein function prediction for genome-wide analyses. The code for SIFTER and protein family data are available at http://sifter.berkeley.edu. [Supplemental material is available for this article.]Automated protein function prediction is an important challenge for computational biology because protein function is difficult to describe and represent, protein databases are littered with annotation errors, and our understanding of how molecular functions arise and mutate over evolutionary time is far from complete. Because biologists depend on protein function annotations for insight and analysis, automated methods have been used extensively to compensate for the relative dearth of experimental characterizations. Although there are 10 7 protein sequences in the comprehensive UniProt database (The UniProt Consortium 2010), <5% have annotations from the Gene Ontology Annotation (GOA) database (Barrell et al. 2009). Far fewer (0.2%) have been manually annotated, and only 0.25% of those manual annotations are from the molecular function ontology in Gene Ontology (GO) (The Gene Ontology Consortium 2010) and are based on experimental evidence. Because of the need for so many annotations, function prediction methods are often assessed based on annotation quantity rather than quality, increasing the number of false positive function annotations and polluting databases (Galper...

The Nudix homology clan encompasses over 80,000 protein domains from all three domains of life, defined by homology to each other. Proteins with a domain from this clan fall into four general functional classes: pyrophosphohydrolases, isopentenyl diphosphate isomerases (IDIs), adenine/guanine mismatch-specific adenine glycosylases (A/G-specific adenine glycosylases), and non-enzymatic activities such as protein/protein interaction and transcriptional regulation. The largest group, pyrophosphohydrolases, encompasses more than 100 distinct hydrolase specificities. To understand the evolution of this vast number of activities, we assembled and analyzed experimental and structural data for 205 Nudix proteins collected from the literature. We corrected erroneous functions or provided more appropriate descriptions for 53 annotations described in the Gene Ontology Annotation database in this family, and propose 275 new experimentally-based annotations. We manually constructed a structure-guided sequence alignment of 78 Nudix proteins. Using the structural alignment as a seed, we then made an alignment of 347 “select” Nudix homology domains, curated from structurally determined, functionally characterized, or phylogenetically important Nudix domains. Based on our review of Nudix pyrophosphohydrolase structures and specificities, we further analyzed a loop region downstream of the Nudix hydrolase motif previously shown to contact the substrate molecule and possess known functional motifs. This loop region provides a potential structural basis for the functional radiation and evolution of substrate specificity within the hydrolase family. Finally, phylogenetic analyses of the 347 select protein domains and of the complete Nudix homology clan revealed general monophyly with regard to function and a few instances of probable homoplasy.

How do proteins evolve novel functions? To address this question, we are studying the evolution of a mammalian toxin, the serine protease BLTX [1], from the salivary glands of the North American shrew Blarina brevicauda. Here, we examine the molecular changes responsible for promoting BLTX toxicity. First, we show that regulatory loops surrounding the BLTX active site have evolved adaptively via acquisition of small insertions and subsequent accelerated sequence evolution. Second, these mutations introduce a novel chemical environment into the catalytic cleft of BLTX. Third, molecular-dynamic simulations show that the observed changes create a novel chemical and physical topology consistent with increased enzyme catalysis. Finally, we show that a toxic serine protease from the Mexican beaded lizard (GTX) [2] has evolved convergently through almost identical functional changes. Together, these results suggest that the evolution of toxicity might be predictable-arising via adaptive structural modification of analogous labile regulatory loops of an ancestral serine protease-and thus might aid in the identification of other toxic proteins.

No abstract