Mihail Sarov et al. (2007) International Worm Meeting "Recombineering based high throughput protein tagging in C. elegans."

Mihail Sarov, Francis Stewart, Anthony Hyman

[International Worm Meeting, 2007]

The methods for protein tagging in C. elegans so far have been inefficient and largely dependent on artificial cDNA based constructs, which can lack important regulatory elements. The recent generation of a genomic fosmid library[1] opened the possibility for direct modification of any gene of interest in its native genomic environment by recombineering in E. coli. We recently published a method for precise in vivo engineering of large genomic clones into GFP tagged transgenes for ballistic transformation[2]. Using such transgenes, we reproduce known and document new expression patterns. We further demonstrate that the tagged protein can complement the function of its endogenous counterpart. Applying the latest developments of the recombineering field and introducing several innovations of our own we have now streamlined the protocol to allow rapid 96 well format liquid culture processing in a continuous recombineering pipeline that takes just four days. This unprecedented throughput and the achieved economy of scale allow us to rapidly produce thousands of tagged transgenes. Our method has become the basis for a large-scale project aimed at comprehensive description of the expression pattern (through fluorescent reporters) and the DNA binding sites (through affinity tag based chromatin immunoprecipitation) of more than 400 C. elegans transcription factor genes (as part of the NIH funded ModENCODE program). At the meeting we will discus the advantages of a distributed, community based collaboration model that will bring this technology to any worm lab and will make the genome wide protein tagging a feasible goal. 1. Perkins, J., K. Wong, R. Warren, J.E. Schein, J. Stott, R. Holt, S. Jones, M.A. Marra, and D. Moerman. 2005. A Caenorhabditis elegans fosmid library. In 15th International C. elegans Conference, University of California, Los Angeles. 2. Sarov, M., S. Schneider, A. Pozniakovski, A. Roguev, S. Ernst, Y. Zhang, A.A. Hyman, and A.F. Stewart. 2006. A recombineering pipeline for functional genomics applied to Caenorhabditis elegans. Nature Methods 3: 839-844.

Mihail Sarov et al. (2006) European Worm Meeting "Generic Method for C. elegans Gene Tagging by in vivo Recombineering in E. coli"

Francis Stewart, Mihail Sarov, Anthony Hyman

[European Worm Meeting, 2006]

Mihail Sarov1, Anthony Hyman2, Francis Stewart1. Here we present a fast and generic protocol for production of tagged transgenes for bombardment-based transformation in C. elegans by engineering of genomic BAC clones. The tag is inserted with minimal disturbance to the endogenous genomic context. The gene is then subcloned into an unc-119 based shuttle vector. All steps are done by in vivo recombineering in E. coli. Large DNA sequences can be modified in this way, and the method is applicable for virtually any gene of interest. All regulatory regions from 5 and 3 UTRs, introns etc. are present in the transgenic construct, and are likely to result in a correct expression pattern.. As a proof of principle we used the RPCI-94 Caenorhabditis briggsae BAC library. We GFP tagged and subcloned a relatively large (16 kbp) gene. Transformation efficiency by bombardment was similar to that of the commonly used pAZ132 vector. GFP expression patterns were reproducible and similar to that of a published promoter:GFP reporter construct. To show that the C. briggsae transgene is functional in C. elegans we knocked down the endogenous gene by RNAi. This makes the transgene the only expressed copy of the gene, essentially simulating a knock-in situation. The RNAi induced phenotype was rescued by the transgene in most of the animals.. While the ability to knock down the endogenous gene is a big advantage of the cross-species transgenes, C. elegans clones would still be preferable for most studies. Recently a genomic fosmid library for C. elegans was announced. Our protocol can be directly applied to these clones. The method is fast - tagging and subcloning can be accomplished within a week. The high efficiency of recombination in E. coli makes it possible do all steps in liquid culture, only checking the final construct for correct recombination. We are currently setting up the conditions for 96 well format tagging that would allow us to automatically process a large number of genes in parallel. We aim to establish a library of tagged transgenic constructs covering the whole genome as a community resource.

Mihail Sarov et al. (2005) International Worm Meeting "Generic method for protein tagging in Caenorhabditis elegans"

Francis Stewart, Mihail Sarov, Assen Roguev, Anthony Hyman

[International Worm Meeting, 2005]

One of the biggest challenges of the post-genomic era is to describe the gene expression patterns and protein-protein interactions on genome wide scale. Protein tagging has been used extensively for localization and complex purification in yeast, where homologous gene targeting is straightforward. To extend this research to Caenorhabditis elegans we have developed a method for fast and efficient gene tagging using homologous recombination in bacteria.The region of interest is subcloned in a shuttle vector from genomic BAC clone and the tag is inserted precisely where desired. Transgenic worm lines are then generated by ballistic transformation. Large DNA sequences can be modified in this way and the method is applicable even for very long genes. Preserving the endogenous genomic context ensures exact reconstitution of the time, space and level of expression of the wild type gene. We designed and tested series of subcloning vectors containing the unc-119 gene, a low-copy replication origin and a selectable marker for bacteria. The tag is added as a cassette with selection marker, which is flanked by frt sites and can be removed upon expression of Flp. We have generated targeting cassettes for several GFP variants and double affinity tags.Caenorhabditis briggsae genomic BAC clones are available from the BACPAC Resource Center, Oakland, California. To map the clones we wrote a program that BLASTs the BAC end sequences to the genome, filters the highest scoring matches and exports them in a format that can be visualized with the genome browser of Wormbase. Use of these BACs is not limited to C. briggsae as most proteins are highly conserved between the different Caenorhabditis species. On DNA level however, the genes are divergent enough to allow specific knockdown by RNAi of the endogenous, untagged allele. This opens an interesting opportunity for comparative cross-species studies.To confirm the validity of our approach we have tagged several proteins with known expression patterns and protein-protein interactions. Data will be presented on the GFP localization and protein complex purification of the Trithorax group protein lin-59.

Mihail Sarov et al. (2008) European Worm Meeting "A first version of the C. elegans Transgeneome: genome scale set of tagged fosmid transgenes"

Vlarie Reinke, Stewart Kim, Anthony Hyman, Francis Stewart, John Murray, Robert Waterston, Michael Snyder, Mihail Sarov, Mei Zhong

[European Worm Meeting, 2008]

Protein tagging with fluorescent and/or affinity epitopes provides a. generic way to address gene function. Recently we have developed a 96 well. format recombineering pipeline that allows us to rapidly convert large. genomic DNA clones such as BACs or fosmids into tagged transgenes [1]. Due. to their large size, these transgenes usually contain most cis regulatory. elements and correctly recapitulate the endogenous patterns of gene. expression. Ballistic transformation of fosmid transgenes allows us to. routinely generate low copy integrated worm lines, with stable and. reproducible expression of the tagged protein in variety of tissues and. developmental stages including germline and early embryos.. The high efficiency and fidelity of this approach allows us to engineer. genome scale sets of fosmid transgenes with unprecedented throughput.. Generation of a comprehensive "Transgeneome" resource with a tagged. transgene for most C. elegans genes is now within reach. As a first step. towards this goal we started with several targeted projects that aim to. determine the physical interactions and the localization of more than 1000. proteins involved in transcription regulation, chromatin structure and. function and cell division. As a part of the ModENCODE consortium we have. already generated hundreds of transgenic constructs and tens of integrated. transgenic worm lines. That allowed us to test more than 20 different tag. combinations and we now have validated tags that perform very well for. protein localization, purification and chromatin immunoprecipitation. At. the meeting we will present results illustrating the power of the approach. as well as some of the lessons we learned from the large-scale application. of the method. We will discus how the community can access these resources. and we hope to initiate collaborations that will translate this research to. other functional gene sets.. 1. Sarov, M., S. Schneider, A. Pozniakovski, A. Roguev, S. Ernst, Y. Zhang,. A.A Hyman, and A.F Stewart. 2006. A recombineering pipeline for. functional genomics applied to Caenorhabditis elegans. Nature Methods 3:. 839-844.

Lee M-H et al. (1997) International C. elegans Meeting "TOWARD THE IDENTIFICATION OF IN VIVO RNA TARGETS OF GLD-1."

Lee M-H, Schedl TB

[International C. elegans Meeting, 1997]

gld-1 is a female germ cell specific tumor suppressor gene that is essential for normal oocyte differentiation and meiotic prophase progression (Francis et al., 1995a,b). GLD-1 is a member of a small family of proteins, including the mouse/human Sam68 and Quaking proteins, that are highly related over an ~ 200 amino acid region which contains a 70 to 100 amino acid RNA binding domain called the KH motif (Jones and Schedl, 1995). Extensive mutational analysis of gld-1 has demonstrated the importance of conserved sequences for its in vivo function. GLD-1 is a cytoplasmic protein and therefore is likely to control mRNA translation or RNA stability in the cytoplasm (Jones et al., 1996). Currently, no obvious RNA targets have been identified from genetic analysis. We have employed a biochemical approach to identify in vivo RNA targets of GLD-1. The strategy is based on the ability of anti GLD-1 antibodies to immunoprecipitate (IP) GLD-1 from a worm lysate and the strong likelihood that GLD-1 present in the lysate is functional and bound to RNAs. GLD-1 was IPed with affinity purified polyclonal antibodies from a cytosol extract of adult hermaphrodites or with rabbit IgG as a control. RNAs co-IP with GLD-1, as well as with rabbit IgG, were converted into cDNAs. Non-specifically trapped RNAs from the GLD-1 IP were eliminated by subtracting cDNAs from the GLD-1 IP with cDNAs from the control IP. The difference product after 4 rounds of subtraction (DP4) was used to screen a Lambda ZAP II cDNA library constructed with the cDNAs from the GLD-1 IP. Duplicate filters were also screened with a probe made from control IP cDNA. Clones that are positive only with the DP4 probe are currently being characterized. Francis et al., 1995a Genetics 139, 579-606; Francis et al., 1995b Genetics 139, 607-630; Jones and Schedl, 1995 Genes Dev. 9, 1491-1504; Jones et al., 1996 Dev. Biol. 180, 165-183. This work was supported by NIH Grant HD25614.

J Rogers et al. (1999) International C. elegans Meeting "Feeding For Phenotypes"

J Hubbard, J Rogers

[International C. elegans Meeting, 1999]

We were intrigued by the recent results of Timmons and Fire (1998) suggesting that RNAi-induced phenotypes could be observed by simply feeding worms with bacteria that are producing the dsRNA of interest. Given the potency of RNAi in the germ line, we wondered if this technique could be used to produce germline-defective phenotypes of interest to our lab. In addition, we sought to apply the technique to identify genes involved in germline function that may otherwise cause a loss-of-function embryonic lethal phenotype. To test this possibility, we collected several cDNAs from genes that produce loss-of-function germline-defective phenotypes (some of which also cause embryonic or larval lethatity) and introduced them into the "double T7" vector described by Timmons and Fire. We are testing these constructs by feeding synchronous populations of L1 larvae with bacteria that express dsRNA and scoring them in the adult for germline defects. Preliminary data suggests that germline phenotypes can be obtained in this manner, but that they may be restricted to events that occur relatively late in germline development. For example, gld-1 dsRNA fed to L1 larvae produced an abnormal oocyte phenotype characteristic of partial loss-of-function alleles of gld-1 (class E and F; Francis et al., 1995) at 60% penetrance. Progeny of unaffected adults were then grown on RNAi-producing bacteria. In these animals we observed the more severe tumorous phenotype in which progression through meiotic prophase is altered and oogenesis does not occur (Francis et al., 1995).

Frederic Horndli et al. (2008) C.elegans Neuronal Development Meeting "Identification of ACR-16 Specific Trafficking and Targeting Components"

Andres Maricq, Frederic Horndli, Michael Jensen

[C.elegans Neuronal Development Meeting, 2008]

Three types of receptors are present at the neuromuscular junction in C. elegans: the levamisole acetylcholine receptor (AChR) complex LevR (UNC-29, UNC-38, UNC-63, LEV-1 and LEV-8), the nicotinic AChR ACR-16 and the GABA receptor UNC-49. Both ACR-16 and the LevR complex respond to cholinergic inputs, with ACR-16 contributing to 70% of the EPSP (Francis M.M.., et al., 2005). Although a mechanism for stabilization and clustering of ACR-16 has been described (Francis MM., et al., 2005), very little is known, about how AChR receptors are transported and inserted at the NMJ. Using RNAi targeting of known transport motors and cargo adaptor proteins, in combination with either unc-29 (x29) or acr-16 (ok789) deletion backgrounds we have successfully identified the kinesin-1 transport complex (composed of unc-116, klc-2, unc-16 and unc-14 (Byrd DT., et aI., 2001, Sakamoto R., et al., 2004)) as an ACR-16 specific transport complex. Indeed, RNAi and deletion mutants for all elements of this complex in combination with unc-29 (x29), exhibit almost complete paralysis as well as ACR-16 GFP mislocalization. In addition, the same RNAi and deletion mutants for the elements of the kinesin-1 complex do not affect acr-16(ok789) movement nor UNC-29 GFP and UNC-49 GFP localization. Encouraged by these results we have started a whole genome RNAi screening approach using the unc-29(x29) and acr-16(ok789) background to discover more elements related to ACR-16 and UNC-29 trafficking, targeting, insertion and stabilization.

Sri Koduri et al. (2006) Neuronal Development, Synaptic Function, and Behavior Meeting "Role of Stargazin in Regulating Glutamate Gated Currents"

Sri Koduri, Andres Maricq, David Madsen, Craig Walker

[Neuronal Development, Synaptic Function, and Behavior Meeting, 2006]

The majority of excitatory neurotransmitter function in vertebrates is mediated by ionotropic glutamate receptors (iGluRs). Reconstitution of iGluRs from C. elegans in Xenopus oocytes requires the auxiliary subunits SOL-1 and STG-1 (see abstracts by Francis et al., and Walker et al.). STG-1 shares weak sequence identity with vertebrate stargazin - a member of the family of glutamate receptor regulatory proteins known as TARPs (Transmembrane AMPA Receptor Regulatory Proteins). We have cloned stargazin homologues from Apis mellifera and Drosophila and co-expressed them with ionotropic glutamate receptor subunits in Xenopus oocytes. We found that the kinetics of the glutamate gated current were dependent upon the combination of the glutamate receptor subunit and the stargazin homologue. Apis STG, when co expressed with vertebrate GluR1, caused the channels to become non-desensitizing. We are currently analyzing chimeras made from Apis STG and vertebrate stargazin to determine which domain of the protein modifies the desensitization kinetics.

Nicholas Duclos et al. (2006) Neuronal Development, Synaptic Function, and Behavior Meeting "Identifying Genes Required for Signaling via the CAM-1 Receptor Tyrosine Kinase"

Nicholas Duclos, Andres Maricq, Michael Francis, Penelope Brockie

[Neuronal Development, Synaptic Function, and Behavior Meeting, 2006]

Signaling at synapses requires that cells within the nervous system adopt the correct cell identity. Thus, both pre- and post-synaptic cells must express a specific compliment of molecules and establish the appropriate synaptic connections. We would like to understand how the C. elegans Ror receptor tyrosine kinase (RTK), CAM-1, contributes to these processes. Ror RTKs are highly expressed in the vertebrate nervous system as well as other tissues, yet the mechanism of signaling by these RTKs remains unclear. cam-1 encodes the only Ror RTK in C. elegans and is normally expressed in both the nervous system and in body wall muscles. In neurons, it has been associated with EGL-20, a Wnt signaling molecule, in regulating nervous system development (Forrester et al.). At the neuromuscular junction, it has been implicated in the stabilization and localization of the acetylcholine receptor subunit ACR-16 (Francis et al.). Overexpression of CAM-1 in either muscles or neurons produces dramatic overgrowth of cellular processes. In order to identify gene products that regulate Ror RTKs, we have initiated a forward screen for modifiers of CAM-1 function that takes advantage of this overexpression phenotype. Overexpression of CAM-1 in the command interneurons produces a dramatic change in cell morphology that is easily identifiable by the presence of numerous highly branched ectopic neuronal outgrowths. To identify genes that function in the same pathway as cam-1, we are screening for suppressors of this phenotype. We expect to find genes that encode upstream regulators of CAM-1 signaling, such as potential ligands, as well as downstream components of the CAM-1 signaling pathway. The identification of these genes will yield valuable insights in to the mechanisms of Ror RTK signaling. Francis MM, Evans SP, Jensen M, Madsen DM, Mancuso J, Norman KR, Maricq AV. The Ror Receptor Tyrosine Kinase CAM-1 Is Required for ACR-16-Mediated Synaptic Transmission at the C. elegans Neuromuscular Junction. Neuron 46: 581-594Forrester WC, Kim C, Garriga G. The C. elegans Ror RTK CAM-1 collaborates with EGL-20/Wnt to regulate cell migration. 168: 1951-1962

Clifford R et al. (1996) Midwest Worm Meeting "Searching for Proteins that Associate with GLD-1"

Clifford R, Giorgini F, Schedl TB

[Midwest Worm Meeting, 1996]

The gld-1 gene is involved in multiple aspects of C. elegans germline development. The primary, essential function of gld-1 is to direct oogenesis. In the absence of gld-1 activity, germ cells that would otherwise develop into oocytes enter meiosis normally but exit early pachytene to re-enter a mitotic cell cycle and form a tumor. The gene also plays nonessential roles in inhibiting premeiotic germ cell proliferation and promoting germ cells to adopt the male sexual fate [1,2]. gld-1 encodes a 463 amino acid protein that contains an evolutionarily conserved domain of ~200 amino acids that is also found in the mammalian Quaking and Sam68 proteins. A KH RNA binding motif lies within this domain [3]. Since GLD-1 appears to be localized exclusively to the cytoplasm, the protein is likely to regulate mRNA translation or stability [4]. To better understand how gld-1 acts in the control of germline development, we are using the yeast two-hybrid system [5] to isolate proteins that physically associate with the gene product. Through this approach we hope to identify regulators and cofactors of GLD-1. By screening an oligo(dT)-primed cDNA library (gift of R. Barstead) with a fragment of GLD-1 that includes the amino terminal half of the conserved domain, we identified the F28E10.1 gene product as a protein that specifically binds GLD-1. The F28E10.1 locus, which was identified by the C. elegans genome sequencing consortium, is predicted to encode a novel, hydrophilic protein of 118 kD. This gene maps to the region of chromosome IV that is deleted by mDf4 and may correspond to one of the lethal or sterile loci uncovered by the deficiency [6,7]. We are currently testing the biological relevance of the interaction between GLD-1 and F28E10.1 through in vivo assays. Progress will be reported. [1] Francis et al. 1995a. Genetics 139: 579-606; [2] Francis et al. 1995b. Genetics 139: 607-630; [3] Jones and Schedl. 1995. Genes Dev 9: 1491-1504; [4] Jones et al., in preparation; [5] Chien et al. 1991. PNAS 88: 9578-9582; [6] Rogalski and Riddle. 1988. Genetics 118: 61-74; [7] Cassada et al. 1981. Dev Biol 84: 193-205.