Sivaramakrishnan, Priya et al. (2021) International Worm Meeting "Transcription rates in the early embryo"

Sivaramakrishnan, Priya, MURRAY, JOHN ISAAC, Watkins, Cameron

[International Worm Meeting, 2021]

Rapid changes in gene expression are necessary for cells to respond to environmental, immunological or developmental signals. As a critical first step, high transcription synthesis rates are required during these responses. In the rapidly dividing C. elegans embryo, where cell cycles are as short as 15 minutes, temporal constraints on transcription are particularly severe. Early cell fate specification genes have to be transcribed at very high rates to accumulate to threshold levels within short developmental time windows. Studies in other labs have shown that reducing the normally high expression levels of some endodermal regulators results in aberrant intestinal fate specification. We hypothesize that high rates of transcription could be required for the fate decisions in several additional early cells. We estimated transcription rates in the early embryo by analyzing single cell RNA sequencing data and found that maximal rates vary across cell types, lineages, and embryonic stages. We have validated some of these by single molecule RNA fluorescence in situ hybridization (smFISH) and identified features of rapidly transcribed genes such as short gene length, shorter and fewer introns, and lack of trans-splicing. We found several motifs enriched in the core promoters of high-rate genes including the initiator element (inr), sites for the SP1 transcription factor, and binding sites for lineage-specific regulators. Using the high-rate mesoderm specification gene ceh-51 as a test case, we measured the effect of these promoter elements on transcription rates using CRISPR-mediated promoter mutagenesis and smFISH. Our analysis shows significant redundancies built into the ceh-51 core promoter with modest contributions to rate from each motif, suggesting that control of transcription rates is surprisingly robust. We are currently designing experiments to test for a direct impact of transcription rate on cell fate. Ultimately, we hope to understand how transcription parameters such as bursting are influenced by rate control and how this contributes to gene expression precision underlying the invariant C. elegans embryonic lineage.

Brendan Galvin et al. (2001) International C. elegans Meeting "A ced-4 SUPPRESSION SCREEN TO IDENTIFY NEW GENES INVOLVED IN PROGRAMMED CELL DEATH"

Brendan Galvin, Bob Horvitz, Scott Cameron

[International C. elegans Meeting, 2001]

Many genes involved in programmed cell death in C. elegans have been identified by screening for mutations that allow survival of cells that are normally destined to die. The opposite approach, screening for mutations that result in an increase in the number of cell deaths, has been comparatively unexplored, in part because of the lack of efficient methods to identify such mutations. A lin-11::gfp reporter (generated by Scott Cameron) expresses in the Pn.aap cells of the ventral cord. In wild-type animals, the six Pn.aap cells of the P3-P8 lineages survive, differentiate into VC neurons, and express lin-11::gfp . In strong cell-death defective mutants, all 12 Pn.aap cells survive and express GFP. In the lin-11::gfp reporter strain the amount of cell death can be quantified by scoring the number of fluorescent nuclei in the ventral cord using a fluorescence-equipped dissecting microscope. Using the lin-11::gfp reporter, I have performed a screen for suppressors of a partial loss-of-function ced-4 mutant by looking for mutants with a reduction in the number of GFP-positive Pn.aap cells. To date approximately 40,500 mutagenized genomes have been screened; 5,000 were screened clonally and 35,500 non-clonally. Two strong suppressors were obtained that reduce the number of GFP-positive Pn.aap cells. The first suppressor, identified in the clonal screen, is recessive and is recessively sterile. The second suppressor, identified in the non-clonal screen, is dominant and is recessively sterile. Further characterization of these mutants will be described.

S Cameron et al. (1999) International C. elegans Meeting "A screen for mutations that affect programmed cell death in the ventral cord"

S Cameron, N Tsung, B Horvitz

[International C. elegans Meeting, 1999]

We have developed an efficient screen for mutations that affect the survival or death of specific ventral cord cells during normal development. A lin-11::gfp reporter construct is expressed in the six VC motor neurons that arise from the P3-P8 lineages as Pn.aap cells. The lineally equivalent Pn.aap cells from the W, P1, P2 and P9-12 lineages die by programmed cell death. In ced-3 animals, in which cell death is prevented, all the "undead" Pn.aap nuclei express GFP and fluoresce brightly. We are using this reporter to screen for mutations that result in the survival of Pn.aap cells in lineages in which they normally die, or the death of Pn.aap cells in lineages in which they normally survive. Hermaphrodites are mutagenized with EMS, and the F2 progeny are screened using a stereomicroscope for animals with an abnormal number or pattern of fluorescing nuclei. To date, 20,000 haploid genomes have been screened, and 70 mutants isolated. Mutants with abnormal survival of Pn.aap cells include new loss-of-function mutations in ced-3 , ced-4 , and egl-1 . New pag-3 alleles have been recovered (see abstract by Cameron et al. ). Mutants with abnormal death of Pn.aap cells include alleles of at least one gene not previously identified as affecting Pn.aap survival. As the Pn.aap cells are sexually dimorphic, weak, self-fertile Tra mutations are also recovered. Mutants isolated in this screen will be characterized with the goal of identifying the factors that determine whether individual Pn.aap cells survive or die.

EJ Aamodt et al. (1999) International C. elegans Meeting "Conservation of sequence and intron/exon structure between the homologous pag-3 genes from C. elegans and C. briggsae"

L Shen, EJ Aamodt, J McDermott, B Rose

[International C. elegans Meeting, 1999]

The pag-3 gene encodes a zinc finger protein involved in controlling neuron specific gene expression. pag-3 mutants were first isolated based on misexpression of touch receptor genes in the BDU interneurons, the lineal sisters to the ALM touch neurons. pag-3 mutants are lethargic, kink when trying to move backward (reverse kinker Unc) and have lineage defects in the P-neuroblast lineages (see abstract by Cameron et al., this meeting). We identified a fosmid containing the C. briggsae homologue of the C. elegans pag-3 gene and the Washington University Genome Sequencing Group sequenced it. When transformed into a C. elegans pag-3 mutant, the C. briggsae pag-3 gene rescued the pag-3 reverse kinker and lethargic phenotypes. The C. elegans pag-3 gene fused to lacZ was expressed in the same pattern in C. elegans and C. briggsae . Unlike many gene homologues compared between C. elegans and C. briggsae , extensive sequence conservation was found in the noncoding regions upstream of the pag-3 exons, in several of the introns and in the downstream noncoding region. Furthermore, the splice acceptor and splice donor sites were conserved and the size of the introns and exons was surprisingly similar. Based on the genomic sequence, the sequence of the C. briggsae PAG-3 protein was predicted. The predicted protein sequence was 85% identical to the sequence of C. elegans PAG-3. These results showed the evolutionary conservation of the pag-3 gene sequence, its expression and function. Because so much of the non-coding regions of pag-3 are conserved, the control of pag-3 may be quite complex, involving the binding of many transacting factors.

PW Reddien et al. (1999) International C. elegans Meeting "The engulfment of dying cells contributes to the killing process of programmed cell death"

HR Horvitz, JS Cameron, PW Reddien

[International C. elegans Meeting, 1999]

The engulfment of a cell corpse by a neighboring cell typically occurs rapidly after the generation of a cell programmed to die. This engulfment can initiate before the completion of cell division, as shown by electron microscopy (1). Because cell corpses are generated in mutants defective in engulfment, the proposed function of engulfment has long been solely the removal of unwanted apoptotic cell bodies. We have discovered, however, that in addition to functioning in cell-corpse removal, engulfment assists in the killing of dying cells. Animals with strong loss-of-function mutations in the killer gene ced-3 lack most if not all programmed cell deaths. However, weaker ced-3 mutants exist that lack only a small percentage of programmed cell deaths. We found that such a partial block in the execution of programmed cell death is enhanced by mutations in genes that are involved in the engulfment process ( ced-1 , ced-2 , ced-5 , ced-6 , ced-7 , and ced-10 ). Furthermore, mutations in these engulfment genes alone are sufficient to result in a low-penetrance survival of some cells that normally die in the ventral cord. Lineage analysis shows that cells that fail to die initially show some morphological characteristics of programmed cell deaths but ultimately appear morphologically indistinguishable, using Nomarski optics, from living cells. Surviving Pn.aap cells in these ventral cord lineages are capable of expressing the cell-type specific reporter lin-11::gfp (see Cameron, Tsung, and Horvitz, this meeting), suggesting that they are capable of differentiating. We are analyzing the relative contributions of the different engulfment genes to cell killing, how the engulfment role in killing interacts genetically with ced-8 and ced-9 , and the contribution of the killing role of engulfment to the death of different cell types. 1. Sulston, J.E., Albertson, D.G., and Thomson, J.N. (1980). Dev. Biol. 78 , 542-576.

Erich M Schwarz et al. (2003) International Worm Meeting "Comparative analysis of cis-regulatory sequences using four Caenorhabditis species."

Nora Mullaney, Erich M Schwarz, Tristan De Buysscher, Paul W Sternberg, Barbara J Wold, John A DeModena, Eunpyo Moon, Hiroaki Shizuya

[International Worm Meeting, 2003]

Comparing homologous cis-regulatory DNA sequences from three or more genomes has advantages over pairwise comparison of only two. Cis-regulatory sequences are short (6-20 bp) and tolerate substantial variation. Purely random pairing of unrelated 100-bp DNA segments is expected to yield two perfect 6 bp matches. Alignment of a third or fourth sequence should greatly lower the frequency of false positive regions, allowing small but real cis-regulatory sequences to be efficiently detected. This increased resolution should also allow direct comparison between phylogenetically conserved sequences and statistically overrepresented sequences, which may yield complementary views of regulatory elements. In the Caenorhabditis genus, C. remanei appears to be most closely related to C. briggsae; two other species, CB5161 and PS1010, comprise the two closest known and culturable Caenorhabditis species outside the elegans-briggsae group (Fitch, 2000). CB5161 is closest to C. elegans, and PS1010 the next most divergent; these two species thus provide an evolutionarily graded series. We have constructed fosmid libraries from CB5161 and PS1010, and begun sequencing individual fosmids for comparative analysis of genes involved in vulval or sensory neuron development. At the same time, we have devised the Mussa software package to adapt the algorithms of Davidson and coworkers (Brown et al., 2002) to multiple sequence analysis. At this writing, we have sequence data from the egl-30, lin-11, and mab-5 loci of both CB5161 and PS1010. Initial results of sequencing and comparative sequence analysis will be presented. References: Brown, C.T., Rust, A.G., Clarke, P.J., Pan, Z., Schilstra, M.J., De Buysscher, T., Griffin, G., Wold, B.J., Cameron, R.A., Davidson, E.H., and Bolouri, H. (2002). New computational approaches for analysis of cis-regulatory networks. Dev. Biol. 246, 86-102; Fitch, D.H.A. (2000). Evolution of Rhabditidae and the male tail. J. Nematol. 32, 235-244.

Galvin BD et al. (2000) East Coast Worm Meeting "Two screens for new genes involved in programmed cell death"

Galvin BD, Bob Horvitz, Cameron S

[East Coast Worm Meeting, 2000]

Many genes involved in programmed cell death in C. elegans have been identified by screening for mutations that allow survival of cells that are normally destined to die. The opposite approach, screening for mutations that result in the deaths of cells, has been comparatively unexplored, in part because of a lack of efficient methods to identify such mutations. A lin-11::gfp reporter (see abstract by Cameron and Horvitz) expresses in Pn.aap cells. In wild-type animals, only the six Pn.aap cells of the P3-P8 lineages survive and express GFP, while in mutants defective for cell killing all 12 Pn.aap cells survive and express GFP. A strain carrying this reporter allows one to monitor the survival of cells using a dissecting microscope rather than Nomarski optics, greatly facilitating screening. In this way, I am performing a ced-4 suppression screen by looking for a reduction in the number GFP-positive cells and thus an increase in cell death in the ventral cords of partial loss-of-function ced-4 mutants. To identify substrates of the caspase CED-3, which is required for programmed cell death, I am using a modified two-hybrid screen. An active caspase is composed of two subunits proteolytically derived from an inactive procaspase. The active site of the caspase contains a cysteine, which when mutated prevents the cleavage activity of the enzyme. In this screen the two active subunits of the CED-3 caspase are expressed separately, with the active-site cysteine changed to a serine. This amino acid substitution should permit an interaction with substrates but prevent their cleavage, allowing for the identification of interacting proteins. Other groups have identified substrates of mammalian caspases using this approach (1, 2). Candidates from the screen will be analyzed by testing in vitro cleavage by CED-3 and assayed by RNAi to examine their roles in programmed cell death. 1. Van Criekinge, W., van Gurp, M., Decoster, E., Schotte, P., Van de Craen, M., Fiers, W., Vandenabeele, P., and Beyaert, R. Analytical Biochemistry. 263, 62-66 (1998). 2. Kamada, S, Kusano, H., Fujita, H., Ohtsu, M., Koya, R. C., Kuzumaki, N., and Tsujimoto, Y. Proc. Natl. Acad. Sci. USA. 95, 8532-8537 (1998).

S Cameron et al. (1999) International C. elegans Meeting "pag-3 may specify both neuroblast cell fate and terminal fates during development of the ventral cord"

B Horvitz, JB McDermott, S Cameron, E Aamodt

[International C. elegans Meeting, 1999]

Pathways controlling the programmed deaths of specific cells in C. elegans are likely to be evolutionarily conserved, and human homologues of genes in these pathways are candidate disease genes. We recovered two alleles of the gene pag-3 as mutations that result in increased numbers of cell corpses in the ventral cord (see abstract by Cameron, Tsung, and Horvitz). pag-3 is a 336 amino acid Zn-finger protein with extensive identity to the mammalian proto-oncogene Gfi-1 . We have shown that in pag-3 animals the Pn.aaa neuroblast reiterates the fate of its mother, Pn.aa, generating supernumerary Pn.aap cells. These cells can undergo programmed cell death, resulting in an increased number of cell corpses in the ventral cord. These findings establish a role for pag-3 in determining neuronal cell fate by regulating cell lineage. pag-3 may also function to regulate terminal aspects of neuronal differentiation. To begin to test this idea we used an affinity-purified antiserum to define the expression pattern of PAG-3 and, in particular, to determine whether it is present during terminal differentiation. We found that PAG-3 protein is present in the touch neurons, the BDU neurons, and many neurons in the head and tail of embryos and adults. In the ventral cord, protein is first detected in the Pn.aa neuroblast, consistent with the lineage findings that descendants of this cell are abnormal. PAG-3 protein remains present through the birth of the mature VA, VB, and VC motor neurons derived from the Pn.aa neuroblast, then rapidly disappears from all but six cells in the mature ventral cord. Expression persists in adults in VA11 and VA12 as well as in four cells of the retrovesicular ganglion. PAG-3 expression during terminal differentiation of ventral cord neurons and its persistence through adulthood in specific subsets of ventral cord neurons suggest that pag-3 may regulate both neuroblast lineage and specific aspects of neuronal differentiation.

JB McDermott et al. (1999) International C. elegans Meeting "Searching for Targets of pag-3"

JB McDermott, EJ Aamodt

[International C. elegans Meeting, 1999]

Mutations in pag-3 cause pleiotropic defects including a lethargic phenotype, a reverse kinker uncoordinated phenotype, lineage defects (see abstract by Cameron et al.), and misexpression of touch neuron-specific genes in the BDU interneurons. pag-3 encodes a zinc finger protein related to the vertebrate proto-oncoprotein Gfi-1. Gfi-1 functions as a transcriptional repressor in transfected cells, and is expressed in lymphoid tissues. pag-3 is expressed in several neural cell types, including the touch cells, the BDU cells, and cells of the VA, VB and VC motor lineages. We are taking three strategies to identify targets of PAG-3. First, we are trying to identify authentic PAG-3 binding sites in touch neuron genes such as mec-7 . Second, we are using DNA chip technology to identify genes whose expression is altered in pag-3 mutants. Third, we are screening for suppressers of the pag-3 uncoordinated phenotype. To identify PAG-3 binding sites we have focused on the mec-7 gene. mec-7 is one of the touch neuron genes misexpressed in the BDU neurons of pag-3 mutants. PAG-3 expressed in bacteria binds well to the consensus Gfi-1 binding site and to mec-7 sites containing the core (AAT/GC) of the Gfi-1 consensus site. To correlate the binding of PAG-3 to changes in target gene expression, we transformed worms with mec-7lacZ mutated at the strongest PAG-3 binding sites. Since PAG-3 may act directly or indirectly on mec-7 , we have transformed worms with deletions of mec-7lacZ to identify the element responsible for misexpression in the BDU neurons. For the DNA microarray screen, we have prepared mRNA from pag-3(ls64) and N2 L1 larvae. The ls64 allele contains a nonsense mutation just upstream of the zinc fingers. Northern analysis will be used to confirm that the abundance of specific mRNAs is altered in pag-3 mutants. Finally, we are looking for suppressers of the uncoordinated phenotype by screening for improved motility in mutagenized pag-3(ls20) worms.

Cameron S et al. (1998) East Coast Worm Meeting "pag-3 FUNCTIONS IN THE VENTRAL CORD LINEAGES TO SPECIFY NEUROBLAST CELL FATE"

Horvitz HR, Cameron S

[East Coast Worm Meeting, 1998]

How specific cells are programmed to die during development is poorly understood. Vertebrate homologues of genes involved in controlling this process are candidate disease genes. We therefore sought to identify genes that regulate cell death in specific cells during C. elegans development. We conducted a screen using Nomarski microscopy of mutagenized ced-1 animals to search for mutations that affect the 10 cell deaths that normally occur in the ventral cord during the late L1 and early L2 stages (1). Among the mutations we recovered was one, n3098, that results in increased numbers of cell corpses in those P-cell lineages in which cell deaths normally occur. Phenotypes associated with n3098 include a reverse kinker-Unc movement defect and occasional misexpression of the neurotransmitter GABA in specific head and tail neurons. n3098 maps on the right arm of LGX. We examined the P cell lineages in n3098 animals and discovered that the Pn.aaa neuroblast reiterates the fate of its mother, Pn.aa. The posterior daughter cell of these reiterated divisions can undergo programmed cell death, as can the analogous cells, P(1-2).aap and P(9-12).aap, in wild-type lineages, and these deaths in reiterated divisions account entirely for the increased number of cell corpses observed in the ventral cord of n3098 animals. Since the mutant phenotype is recessive, it is likely that the wild-type gene product identified by the n3098 mutation is required for the determination of the Pn.aaa neuroblast fate. Our model predicts that in n3098 animals the reiterated P(3-8).aa-cell lineages should result in supernumerary VC-like motor neurons (P(3-8).aap) and should contain a single cell with the characteristics of Pn.aaa at the expense of the normal VA and VB motor neurons produced by this neuroblast. We confirmed the presence of supernumerary VC-like neurons in n3098 animals using anti-FMRF-amide antisera (kindly provided by Chris Li). An absence of normal VA and VB motor neurons in n3098 animals could account for their Unc phenotype. The map location, Unc phenotype, and GABA-misexpression phenotype of n3098 animals suggested that this mutation might be an allele of pag-3, a gene isolated in the Aamodt laboratory in screens for mutations that result in misexpression of a mec-7::lacZ reporter (2). pag-3 encodes a 336 amino acid Zn-finger protein with extensive similarity to the vertebrate protein Gfi-1 (3). n3098 fails to complement mutations in pag-3 and contains a stop codon just N-terminal to the Zn-finger domains of PAG-3. In transgenic mice Gfi-1 can promote oncogenesis when overexpressed, possibly by inhibiting the transcription of proapoptotic genes such as Bax (4). Since other genes in C. elegans, such as unc-86 and lin-29, function not only to determine cell fate through regulating cell lineage but also to directly control specific aspects of terminal differentiation, we are testing whether pag-3 may function similarly, i.e., both in the determination of Pn.aaa neuroblast identity and in the regulation of terminal cell fates in the Pn.aa lineages. As programmed cell death is one terminal cell fate expressed in these lineages, we are specifically examining whether pag-3 may directly regulate transcription of the ced (cell death) genes. Demonstrating such a role for pag-3 would suggest that programmed cell death in the ventral cord may be controlled using mechanisms shared with other programs of normal neuronal development. 1. Cameron, S. and Horvitz, H. R., 1997 International C. elegans Meeting Abstract p. 628. 2. Jia, Y. et al., Genetics 142: 141-147, 1996. 3. Jia, Y. et al., Development 124: 2063-2073, 1997. 4. Gilks, C. B. et al., Molec. and Cell. Biology 13: 1759-1768, 1993.