J Liu et al. (1999) International C. elegans Meeting "CDK-Activating Kinase (CAK): A genomic search"

ET Kipreos, J Liu

[International C. elegans Meeting, 1999]

The major cell cycle regulator, cyclin-dependent kinases (CDKS) require phosphorylation by a CDK-Activating Kinase (CAK) to be active. Cell cycle regulatory proteins are highly conserved in eukaryotes, therefore, one might expect that the central regulator CAK would also be conserved. In human, xenopus and clams, the protein CDK7 was proposed to be a CAK by biochemical experiments. However, those experiments didn't show that CDK7 functions as a CAK in vivo. Recently, a Drosophila cdk7 temperature sensitive mutant was obtained and was shown to be defective in CAK activity in vivo, but only for Cdc2/CyclinA and Cdc2/CyclinB complexes and not for the Cdk2/CyclinE complex (Larochelle et al., 1998, Genes & Dev. 12: 370). This result raises two questions. First, what is the CAK for the Cdk2/cyclinE complex? Second, since CDK7 is also the RNA polymerase II CTD kinase, was the cdk7 phenotype merely due to impaired transcription of a real CAK? In S. cerevisiae, in contrast, another protein, CAK1/CIV1, has been shown convincingly by genetic and biochemical experiments to be the essential CAK. CAK1 is a diverged CDK family member. At present no CAK1 homolog has been found. The sequenced C elegans genome enabled us to search for a CAK1 ortholog. We identified 12 C. elegans CDKs by using BLAST and PROFILE searches of the C. elegans genomic database with the six budding yeast CDKs and 16 human CDKs. We then used maximum likelihood, parsimony, and neighbor-joining methods to study the phylogeny of these 34 CDKs. All thre approaches produced similar trees. We determined that C. elegans has a CDK7 ortholog, but no obvious CAK1 ortholog; other major categories of human CDKs have their counterparts in C. elegans. One C. elegans CDK, H01G02.2, could not be definitively grouped with other CDKs. Although H01G02.2 is more similar to other yeast CDKs than to CAK1, it is theoretically possible that CAK1 and H01G02.2 are orthologs that have diverged by unequal rates of evolution. We are currently employing an RNAi approach to determine the loss-of-function phenotypes of H01G02.2 and CDK7. If our phylogeny results are confirmed, it would indicate that one of the most central aspects of cell cycle regulation, the CDK-Activating Kinase, has not been conserved from yeast to higher eukaryotes.

Ji Liu et al. (2001) International C. elegans Meeting "The regulation of cell cycle exit by cul-1 & lin-23 genes"

Edward Kipreos, Ji Liu

[International C. elegans Meeting, 2001]

Hyperplasia, the excess proliferation of cells in a tissue, often precedes neoplasia in cancer development. In C. elegans, inactivation of either the cul-1 or lin-23 gene results in hyperplasia in diverse tissues, with mutant cells unable to exit the cell cycle in response to developmental cues. The exact function of cul-1 and lin-23 in the regulation of cell cycle exit is not understood. Orthologs of CUL-1 and LIN-23 in other organisms function together as components of SCF ubiquitin-ligase complexes that target proteins for degradation. We hypothesize two mechanisms by which CUL-1 and LIN-23 promote cell cycle exit: A. CUL-1 and LIN-23 could directly target a positive cell cycle regulator for degradation, e.g., the vertebrate CUL1 has been shown to degrade cyclin E. B. CUL-1 and LIN-23 could act in the Wnt/wingless pathway to degrade the signaling molecule b-catenin, which functions as an oncogene in vertebrates. lin-23 orthologs slmb (Drosophila) and b-TrCP (vertebrate) have been shown to function in this pathway. To study CUL-1 and LIN-23 function, two approaches are being used: candidate substrates testing, and extragenic suppressor screens. Genetic interactions between cul-1, lin-23, cycE, and b-catenin mutants are being studied. There are three b-catenins in C. elegans: bar-1, wrm-1, and hmp-2. We have constructed double mutants of lin-23 with bar-1 and hmp-2. If lin-23 mutants' hyperplasia phenotype was caused by elevated levels of BAR-1 or HMP-2 proteins, then inactivating them in the double mutants should rescue hyperplasia. However, we failed to observe such a rescue, indicating that BAR-1 and HMP-2 are probably not the critical substrates for LIN-23 in terms of cell cycle exit control. We are currently measuring the protein levels of cyclin E and b-catenin in cul-1 and lin-23 mutants by immunofluorescence and Western blot. Extragenic suppressor screens are on-going to isolate genes that genetically interact with lin-23. Our strategy is to mutagenize the strain lin-23(rh293) dpy-2(e8) exc-7(rh252)/mnC1 and look for Dpy Exc non-Lin animals among the F2 progeny. Such animals would indicate a lin-23 exc-7 dpy-2 homozygote with a suppression of the Lin phenotype. Prospective suppressors include LIN-23 substrates, other components of the SCF complex, and other regulators of cell cycle exit. I will report my results of both approaches at the meeting.

K Liu et al. (1999) International C. elegans Meeting "Molecular, functional and behavioral characterization of unc-11 mutant alleles."

EB Landies, A Holgado, A Kurani, A Alfonso, K Liu, R Golan

[International C. elegans Meeting, 1999]

unc-11 encodes the nematode homolog of the mammalian neuronal specific clathrin assembly protein AP180 (Nonet et al. , 1999). Five alternatively spliced cDNAs and three protein isoforms can be detected in the wild type strain N2. Like its mammalian counterpart, the UNC-11 protein(s) is(are) enriched in the nervous system. However, they are also expressed in the intestine, the coelomocytes and the gonad. unc-11 has at least two functions in the nervous system: 1) to sort/recycle the synaptic vesicle protein synaptobrevin and 2) to regulate clathrin cage sizes (Nonet et al. , 1999). Therefore, UNC-11 functions affect both exocytosis and endocytosis of synaptic membrane components and consequently mutants are defective in synaptic transmission (Nonet et al. , 1999, Holgado and Alfonso, 1999 C. elegans meeting). To address the biological relevance of the individual cDNAs and determine whether the proteins they encode have different functions and subcellular localizations in vivo , we have created transgenic lines expressing individual isoforms. We present our findings on the functional, biochemical and subcellular characterization of three different isoforms. To determine structure-function correlations, the available 15 mutant alleles were sequenced. The molecular null alleles are viable, supporting the conclusion that unc-11 function is not essential for synaptic transmission (Nonet et al. , 1999). Most of the mutations result in frameshifts and these early truncations affect all putative isoforms derived from the gene. Since UNC-11 is expressed in the intestine, gonad and the coelomocytes, we are interested in determining possible functions of UNC-11 in tissues other than the nervous system. We present data on several mutant alleles defective in neuronal function, in which mutant proteins are detectable and their immunolocalization suggests interesting novel functions. Supported by NS32449 to A.A. Nonet, Holgado, Brewer, Serpe, Norbeck, Holleran, Wei, Hartweig, Jorgensen and Alfonso, 1999 (submitted). UNC-11, a C. elegans AP180 homolog, regulates the size and protein composition of synaptic vesicles.

Jun Liu et al. (2001) International C. elegans Meeting "Mesodermal cell fate specification in the postembryonic M lineage"

Jun Liu, Andrew Fire

[International C. elegans Meeting, 2001]

We are interested in understanding how mesodermal cell fates are specified. The postembryonic mesodermal lineage, the M lineage, provides a valuable system for these studies. During postembryonic development, the single blast cell M divides characteristically and reproducibly to generate 14 striated bodywall muscles, 16 non-striated sex muscles (vulval and uterine muscles), and two non-muscle cells (coelomocytes). Previous work from our lab and others has shown that correct patterning of the M lineage involves both lineage specific intrinsic factors and cell-cell signaling. We are interested in identifying and characterizing the functions of these factors in M lineage fate specification. At the start of the M lineage, two Hox cluster genes lin-39 and mab-5 , as well as their cofactor, a homeodomain protein CEH-20 specify cellular identity and myogenic potential of M. We have found an overlapping function for lin-39 and mab-5 in diversification of different cell fates within the lineage, and have shown that the C. elegans twist ortholog hlh-8 is a direct target of these factors in the M lineage. Using gfp-based mutagenesis screens, we have isolated 26 additional mutants that have defects in M lineage patterning (12 of them came from a screen carried out by Judy Yanowitz in the lab). Three of these mutants turned out to be allelic to sma-9 , and showed intriguing genetic interactions with lin-12(0) . We are in the process of characterizing these and other mutants isolated from the screens.

Liu J et al. (1993) International C. elegans Meeting "Isolation and analysis of lin-3 suppressors."

Sternberg PW, Liu J

[International C. elegans Meeting, 1993]

Liu DWC et al. (1991) International C. elegans Meeting "Mutations Affecting Periodicity of Defecation."

Liu DWC, Thomas JH

[International C. elegans Meeting, 1991]

Defecation is a periodic behavior in C. elegans, occurring about every 45 seconds in well-fed, adult wild-type worms. The cycle periodicity (CP) may be under the control of a neuronal pattern generator. Extensive screens for constipated (Con) mutants yielded only one mutation in one gene (dec-1) which caused a lengthening of the defecation CP (Thomas, Genetics 124:855-872, 1990). We reasoned that Con screens might fail to reveal many mutants with long cycles and should miss mutants with shorter than normal CP. To isolate short and long-cycle mutants we undertook a direct screen for animals with altered CP. We screened for mutant candidates in the second generation after EMS treatment. CP was timed in individual animals, and a mutant candidate was picked: 1 ) if it repeated the behavior twice or more in one minute, 2) if it repeated the behavior less often than once per minute. We anticipated finding several classes of mutants; 'sick' long- period, pumping defective long-period, 'healthy' long-period, and short-period. Defecation is regulated by food intake. Wildtype worms will stop defecating off of food. CP is correlated with the rate at which the pharynx pumps in food; pumping defective worms defecate less often. The screen has led to the identification of four mutations which cause CP lengthening. 1 ) One mutation (sa86) causes rapid shallow pumping; less effective pumping and lower food intake might explain the long CP. 2) & 3) One mutation has lowered brood size (sa88), and another (sa91) has smaller broods and retarded development. We are concerned that long CP could result from non-specific affects on health, rather than mutations specifically affecting the CP generator. However, many Unc and Con mutants have normal CP. 4) Finally, animals carrying the sa73 mutation have an average period of 100 seconds, and appear healthy otherwise. Furthermore we have isolated one mutation (sa92) which leads to fast cycling. Animals carrying sa92 have a CP of about 30 seconds. Epistasis analysis among the defecation cycle (Dec) genes, as well with other defecation (Con) genes will be presented.

Liu J et al. (1995) International C. elegans Meeting "GENETIC AND MOLECULAR ANALYSIS OF LIN-3"

Liu J, Sternberg PW

[International C. elegans Meeting, 1995]

lin-3 is the vulval inducing signal made by the anchor cell during C. elegans vulval development. It encodes a transmembrane protein with one extracellular EGF repeat. Two different lin-3 cDNAs exist by alternative splicing, they may encode one membrane-bound form and one secreted form of the protein. The secreted form is sufficient to induce the vulva. Reduction of lin-3 activity causes a Vulvaless (Vul) phenotype, overexpression of lin-3 in transgenic animals results in a Multivulva (Muv) phenotype. lin-3 expression in the somatic gonad is restricted to the anchor cell, beginning at the stage when vulva induction initiates. lin-3 also affects viability, fertility, and male spicule development. To isolate new genes encoding regulators required for lin-3 transcription, protein modification and activity, we screened for suppressors of 1) syIs1, which is a chromosomal integrant of a lin-3 transgene and has a dominant and fully penetrant Muv phenotype, and 2) lin-3 (n378/n1059), a highly Vul strain. Four new loci are defined by the syIs1 suppressors. At least two act upstream the lin-3 receptor in the vulva induction pathway. We are using several lin-3 trangenic constructs to determine whether these genes act at lin-3 transcription or protein level. At least one new locus was found in the suppressors of lin-3 (n378/n1059). It may be a negative regulator of lin-3 expression and function. We sequenced the lin-3 mutant alleles, which provide information about the functions of different domains of the protein. We are also making lin-3 constructs in which only one of the two alternative splicing is used. These constructs will be used to study the different functions of the two forms of protein.

Liu DWC et al. (1993) International C. elegans Meeting "Behavioral and genetic analysis of defecation periodicity."

Liu DWC, Thomas JH

[International C. elegans Meeting, 1993]

Liu KK et al. (1993) International C. elegans Meeting "Neuronal control of male mating behavior."

Liu KK, Sternberg PW

[International C. elegans Meeting, 1993]

Christine Liu et al. (2001) International C. elegans Meeting "Defining the structural requirements for unc-64(md130)'s general anesthetic resistance"

C Michael Crowder, Christine Liu, Stephen J Hunt

[International C. elegans Meeting, 2001]

unc-64 encodes a homolog of vertebrate syntaxin 1A, which is expressed ubiquitously in the C. elegans nervous system. Syntaxin is a t-SNARE composed of 5 functional domains (Ha, Hb, Hc, H3, and TM) and is involved in membrane fusion of synaptic vesicles. Its H3 helical domain is known to interact with other SNARE proteins during neurotransmission. We have shown previously that mutations in unc-64 profoundly alter the volatile anesthetic (VA) sensitivity of C. elegans . Animals with the unc-64( md130) mutation are semidominantly resistant to isoflurane and halothane while worms carrying other hypomorphic unc-64 mutations are 30 times more sensitive to these anesthetics. The md130 lesion is a G to A mutation at the splice donor site of intron 6 of unc-64 . By RT-PCR, md130 produces a small amount of wild type-mRNA along with truncated forms lacking half of the H3 domain and the entire TM domain, rendering them unable to properly interact with other SNARE proteins. We want to define the structural requirements for md130 's resistance to VAs. First, we transformed N2 worms with a plasmid containing genomic unc-64 carrying the md130 lesion and found that these animals exhibit the same behavioral and VA phenotypes as unc-64(md130) homozygous animals. unc-64(null/+) animals transformed with the same plasmid are also VA resistant, and transformed unc-64(null) homozygotes are viable. Another plasmid with an added stop codon that produces only the truncated md130 product and no wild type product conferred VA resistance similar to unc-64(md130) and predictably failed to rescue the lethality of unc-64(null) . N2 transformed with syntaxin containing only the H3 and TM domains (delta-Habc H3 TM) are very Unc (too Unc to test their anesthetic sensitivity), slow growing and have small brood sizes. N2 animals transformed with unc-64 lacking only the H3 domain (Habc delta-H3 TM) are uncoordinated and slower growing compared to N2 and are not VA resistant. This plasmid also does not rescue the lethality of unc-64 (null) . Other unc-64 deletion constructs and amino acid substitution mutations are currently being made.