Janelia Farm Research Campus
Learn about the new HHMI research campus located in Virginia. More

Summary: Paul Sternberg is interested in how genes and genomes specify development and control stereotyped behavior.

To understand how a genome specifies the properties of an organism, we focus on the nematode Caenorhabditis elegans, which by virtue of its small cell number and its stereotyped anatomy, development, and behavior is amenable to intense genetic analysis. Since we know its complete genome sequence, this worm also serves as a model for using genomic information to glean biological insight. We seek to understand how signals between cells are integrated to coordinate organ formation and how genes and neural circuits control the ability to execute stereotyped behavior in response to environmental and nematode-produced signals. Our strategies include identification of genes through genetic and molecular screens, detailed observation of cell and organism behavior, and cycles of computational and experimental analyses. We also use comparative analysis to take advantage of conservation to define key elements of the genome, of regulatory circuits, and of divergence to understand unique features of a species. Many of the genes we identified are the nematode counterparts of human genes, and our experience is that many of our findings apply to human genes as well.

We are interested in genetic regulatory networks. We continue to construct doubly mutant strains of C. elegans to infer relationships among genes. For example, quantitative measurement of the phenotypes of double mutants allowed us to define three WNT signaling pathways acting during organogensis, in particular on the orientation of vulval 2° lineage precursor cells. Two pathways work in parallel, and the third acts antagonistically to the other two. One of these parallel signaling pathways involves LIN-18, ortholog of human RYK, a novel WNT receptor whose signal transduction pathway is unknown. The other parallel pathway involves a classical WNT receptor, LIN-17. The third pathway involves CAM-1, ortholog of the ROR transmembrane tyrosine kinase; VANG-1, ortholog of the membrane protein Van Gogh/Strabismus; and LRP-2, homolog of Arrow/LRP. These receptors respond to distinct WNT ligands. The WNT EGL-20 is expressed in the tail and orients the 2° lineages to the posterior. The other two WNTs are expressed in the anchor cell and provide a localized signal to orient the polarity of the posterior 2° lineage toward the center, apparently overriding the posterior WNT signal. These three WNT pathways converge on asymmetric localization of β-catenin, and we are trying to elucidate how CAM-1 signals and how it acts antagonistically to the other two WNT pathways.

Our past work used integration of data from multiple types of experiments and multiple organisms to predict genetic interactions. Although this approach was strikingly successful, one limitation was the lack of sufficient data. We thus are developing rapid, quantitative assays for phenotypes to compare the effects of gene mutation or RNA interference (RNAi) knockdown of gene activity. We then test selected combinations of perturbations to test potential interactions. We are using behavioral assays—worm locomotion, feeding behavior, and male body posture in response to pharmacological agents—to obtain profiles of genes relevant to nervous system function and dysfunction, particularly addiction.

A major part of the information in a genome is cis-regulatory sequences, and we (along with many others) want to understand how to interpret regulatory sequence because many of our developmental biology projects stalled out at the level of cis-regulation. Our approach is threefold. We are using comparative genomics to identify potential regulatory regions. We are testing relatively short sequences in vivo in enhancer assays and using the information to tune how we predict regulatory elements. We are also profiling the transcriptomes of particular cell types to obtain large sets of genes expressed in those cells.

We started a new project on cell migration to understand both normal organogenesis and potential migratory programs that might be accessed by metastatic tumor cells. The C. elegans male linker cell (LC) undergoes a complex migration with changes in direction, speed, and morphology. An initial functional screen for genes involved in LC migration identified the Tlx ortholog nhr-67 as being necessary for the middle parts of the migratory program, such as negative regulation of the netrin receptor unc-5 to allow a ventral turn. We have profiled the transcriptome of individual LCs by microdissection, amplification, and cDNA deep sequencing. This study identified about 800 LC-enriched genes whose functions we are now analyzing, including a number of conserved proteins of unknown function that we predict will have roles in migration in human cells.

We discovered that an epidermal growth factor (EGF) receptor–phospholipase C signaling pathway regulates C. elegans sleep, defined as behavioral quiescence and increased latency to arousal (they take longer to respond to aversive stimuli). We are screening by RNAi for genes that are necessary for induction of sleep by LIN-3 (EGF). One such gene is the UNC-103 ERG potassium channel that René García (HHMI, Texas A&M University) had previously implicated in regulation of C. elegans male spicule protraction. This gene is the ortholog of a gene involved in human cardiac arrhythmia. We are now using calcium imaging to identify the neurons that are affected in this sleep-like state.

Nematodes are major parasites and agricultural pests, and we seek to leverage our understanding of C. elegans biology to nematodes of broader interest. To accomplish this we are studying nematode-specific processes in C. elegans, and parasitic-relevant behaviors in other nematodes. Host finding is a general feature of nematode parasites, and we have found that the infective juveniles of two types of insect-killing nematodes, Heterorhabditis bacteriophora and Steinernema carpocapsae, are attracted to carbon dioxide. We found that C. elegans adults rapidly avoid CO2 above ambient levels and that this aversive response requires a pair of sensory neurons, the BAG neurons. In collaboration with Bob Horvitz’s laboratory (HHMI, Massachusetts Institute of Technology), we have found by calcium imaging that the BAG neurons are activated directly by CO2 and that this activation requires a guanylyl cyclase, GCY-9, that we propose to be the receptor for CO2. By contrast with C. elegans adults, C. elegans dauer larvae, analogous to infective juveniles of parasitic species, are attracted to CO2.

We identified and killed the BAG neurons in H. bacteriophora and S. carpocapsae and find that the neural circuitry, at least at the level of the sensory neurons, is conserved. H. bacteriophora and S. carpocapsae are attracted to a broad range of insect hosts. Using gas chromatography to separate volatile compounds and mass spectrometry to identify them by their mass, we have identified some of the odors released by a set of insect hosts. Some of these odors, as well as a number of others, are attractive to either or both of these nematodes. We thus believe that the broad host range of these nematodes relies in part on a common signal, CO2, and in part on their ability to respond to a broad range of odors.

We would like to use molecular genetics to study aspects of these parasites. To support this effort, we are analyzing their genomes. We have established RNAi for H. bacteriophora. We are collaborating with several other groups to analyze the H. bacteriophora genome sequenced by Washington University Genome Sequencing Center, and we sequenced in-house a polymorphic strain to obtain single-nucleotide polymorphisms (SNPs), which we will use to generate a genetic/physical map.

To demonstrate the feasibility of de novo assemblies of nematode genomes from short-read sequences in-house, we assembled the genome of Caenorhabditis angaria from 75 nucleotide paired-end reads. This particular genome was fragmented, and we developed a way to use deep cDNA sequencing to scaffold the genomic DNA into larger than gene-size pieces. We used the C. angaria genome to identify highly conserved sequences that are candidate regulatory elements. We have assembled and annotated five Steinernema genomes, including the S. carpocapsae genome, and Panagrellus redivivus as an outgroup. These genomes are being used for molecular genetic analysis and to identify Steinernema-enriched gene families. Also, in collaboration with Robin Gasser’s group (University of Melbourne), we are working on the larger genome of the sheep parasite Haemonchus contortus and the pig parasite Ascaris suum.

We have previously studied particular aspects of the sensory response of the male nematode to contact with mating partners, and we have also developed an assay for hermaphrodite (or female, depending upon the species) attraction of males. In collaboration with Arthur Edison (University of Florida) and Frank Schroeder (Cornell University), we used these assays to purify several chemicals that constitute the C. elegans hermaphrodite-mating cue. These chemicals, called ascarosides, are related members of a family of small molecules that are derivatives of the dideoxy sugar ascarylose, first identified from the nematode Ascaris. The ascarosides have lipid chains of different lengths. The potential diversity of ascarosides leads us to hypothesize that ascarosides are a general family of nematode social-signaling molecules that are analogous to bacterial quorum-sensing signals. We are testing this hypothesis by purifying mating pheromone from other nematodes.

We examined several molecular aspects of nematode life-cycle decisions. We used L1 larval arrest to study nutritional control of these decisions, and went on to use microarrays and NanoString technology to examine transcriptional changes. We were early adopters of chromatin immunoprecipitation analyzed by deep sequencing (ChIP-seq) and discovered that RNA polymerase accumulates at the 5' end of transcriptional units during L1 arrest. We then examined the genomic organization related to arrested states and the transition back to growth. We used this analysis to develop a model for the selective advantage of operons in metazoans, namely that operons decrease the need for transcriptional resources in the initial stages of transition to growth, either release from L1 arrest or recovery from dauer larvae. We are now examining how the entry into dauer is controlled by dauer pheromones (mixture of ascarosides) and steroid hormones (dafachronic acid). We collaborated with Adam Anteb (Max-Planck-Institute for Biology of Ageing) to analyze the role of dafachronic acid in pheromone response, in particular how worms respond to a shift form high to low pheromones when larvae are deciding to undergo reproductive or dauer development.

We continue to organize, store, and display information about C. elegans and to extend these efforts to other nematodes. With our international team of collaborators, we present this information in an Internet-accessible database, WormBase (www.wormbase.org). Our major contribution is to extract information from the literature, focusing on gene, protein, and cell function; gene expression; gene-gene interactions; and functional genomics data. Annotation of gene function includes use of the Gene Ontology (GO; www.geneontology.org, and we are developing these ontologies as part of the GO Consortium. To facilitate these processes, we have developed Textpresso (www.textpresso.org), a search engine for biological literature. In collaboration with other model organism databases, we have applied Textpresso to the literature of C. elegans, Drosophila, Arabidopsis, mouse, and several human diseases. We use this system to automate some steps in the extraction of information from full-text papers. Extension of Textpresso to neuroscience is part of the Neuroscience Information Framework.

Grants from the National Human Genome Research Institute, the National Institute of Drug Abuse, and the National Institute of General Medical Sciences provided partial support for these projects.

As of May 30, 2012