SEL-5, A Serine/Threonine Kinase That Facilitates lin-12 Activity in Caenorhabditis elegans

Corresponding author: Iva Greenwald, 701 West 168th St., HHSC Rm. 720, Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032., greenwald{at}cuccfa.ccc.columbia.edu (E-mail)

Communicating editor: R. K. HERMAN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Ligands present on neighboring cells activate receptors of the LIN-12/Notch family by inducing a proteolytic cleavage event that releases the intracellular domain. Mutations that appear to eliminate sel-5 activity are able to suppress constitutive activity of lin-12(d) mutations that are point mutations in the extracellular domain of LIN-12, but cannot suppress lin-12(intra), the untethered intracellular domain. These results suggest that sel-5 acts prior to or during ligand-dependent release of the intracellular domain. In addition, sel-5 suppression of lin-12(d) mutations is tissue specific: loss of sel-5 activity can suppress defects in the anchor cell/ventral uterine precursor cell fate decision and a sex myoblast/coelomocyte decision, but cannot suppress defects in two different ventral hypodermal cell fate decisions in hermaphrodites and males. sel-5 encodes at least two proteins, from alternatively spliced mRNAs, that share an amino-terminal region and differ in the carboxy-terminal region. The amino-terminal region contains the hallmarks of a serine/threonine kinase domain, which is most similar to mammalian GAK1 and yeast Pak1p.

DURING development, cells with equivalent potential adopt different fates as a consequence of cell-cell interactions. Many such interactions are mediated by receptors of the LIN-12/Notch family (reviewed in GREENWALD 1998 ) binding to transmembrane protein ligands of the Delta/Serrate/LAG-2 (DSL) family (GREENWALD 1998 ). Ligand binding appears to induce cleavage in or near the transmembrane domain to release the intracellular domain (LIEBER et al. 1993 ; STRUHL et al. 1993 ; KOPAN et al. 1994 ; SCHROETER et al. 1998 ; STRUHL and ADACHI 1998 ). The intracellular domain of LIN-12/Notch proteins translocates to the nucleus, where it complexes with transcription factors of the CBF1/Suppressor of Hairless/LAG-1 family and participates in transcriptional regulation of target genes (JARRIAULT et al. 1995 ; KOPAN et al. 1996 ; CHEN et al. 1997 ; EASTMAN et al. 1997 ; STRUHL and ADACHI 1998 ).

Other members of the LIN-12/Notch pathway, and factors that influence the activity of the LIN-12/Notch pathway, have been conserved evolutionarily. Some of these components have been identified in genetic screens based on suppression or enhancement of lin-12 mutations (sel genes). Screens that rely on suppressing missense mutations that cause constitutive LIN-12 activity have yielded at least seven genes. Three of these genes have been characterized molecularly and have been found to be conserved components that are important for LIN-12/Notch activity. lag-2, a ligand gene, was identified in such screens by antimorphic alleles (TAX et al. 1994 , TAX et al. 1997 ). sel-12, a presenilin gene, was defined by loss-of-function mutations (LEVITAN and GREENWALD 1995 ). Presenilin, which was identified independently by mutations that cause familial early-onset Alzheimer's disease by altering the processing of ß-amyloid precursor protein (SELKOE 1998 ), appears to be critical for the transmembrane proteolytic processing event that constitutes LIN-12/Notch signal transduction (DESTROOPER et al. 1999 ; STRUHL and GREENWALD 1999 ). Finally, sup-17, which encodes a transmembrane disintegrin/metalloprotease of the ADAM family, was defined by loss-of-function mutations (TAX et al. 1997 ; WEN et al. 1997 ); the Drosophila homolog of this gene, kuzbanian, was identified independently in genetic screens on the basis of its Notch^- phenotype (ROOKE et al. 1996 ).

We report the characterization of another gene, sel-5, which was identified in a screen for suppressors of missense mutations that activate LIN-12 (TAX et al. 1997 ). TAX et al. 1997 identified two alleles of sel-5 in their suppressor screen and performed genetic mosaic analysis, which suggested that sel-5 suppression is cell autonomous. Here, we have performed additional genetic analysis and have shown that sel-5 encodes two proteins containing a common serine/threonine kinase domain. We discuss possible ways sel-5 activity may influence lin-12 activity.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

General Caenorhabditis elegans methods and strains:
General methods for the handling and maintenance of C. elegans are as described previously (BRENNER 1974 ). The wild-type parent for all strains was C. elegans var. Bristol N2 (BRENNER 1974 ). The alleles used in this work are:

• Linkage group (LG) I: smg-1(r861) (HODGKIN et al. 1989 ), unc-54(r293) (PULAK and ANDERSON 1993 ).

• LG III: unc-79(e1068) (SEDENSKY and MENEELY 1987 ), mab-21(bx53) (BAIRD et al. 1991 ), dpy-17(e164) (BRENNER 1974 ), ncl-1(e1865) (HEDGECOCK and HERMAN 1995 ), unc-36(e251) (BRENNER 1974 ), unc-32(e189) (BRENNER 1974 ), lin-12(n302) (GREENWALD et al. 1983 ), lin-12(n676) (GREENWALD et al. 1983 ), lin-12(n950) (GREENWALD et al. 1983 ), lin-12(n137) (GREENWALD et al. 1983 ), lin-12(ar170) (HUBBARD et al. 1996 ), glp-1(e2141) (PRIESS et al. 1987 ), glp-1(e2142) (PRIESS et al. 1987 ), glp-1(ar202) (J. HUBBARD and I. GREENWALD, unpublished observations). sDp3 is a free duplication of part of this chromosome (ROSENBLUTH et al. 1985 ). sDf121(s2098) is a homozygous lethal deletion, which is rescued by sDp3, of part of this chromosome (STEWART et al. 1998 ).

• LG IV: dpy-20(e1282) (HOSONO et al. 1982 ).

• LG V: him-5(e1467) and him-5(e1490) (HODGKIN et al. 1979 ).

• LG X: sel-12(ar131) (LEVITAN and GREENWALD 1995 ).

• Transgenes: arIs12 [lin-12(intra)] (STRUHL et al. 1993 ) expresses the intracellular domain of LIN-12 under the control of lin-12 regulatory sequences and is marked with the dominant marker rol-6(su1006). arIs13 [lag-2::lacZ] (WILKINSON et al. 1994 ) carries the reporter lag-2::lacZ and is marked with rol-6(su1006). arIs41 (LEVITAN and GREENWALD 1998 ) expresses a functional LIN-12::GFP fusion protein under the control of lin-12 regulatory sequences and is marked with rol-6(su1006). arEx29 (K. FITZGERALD and I. GREENWALD, unpublished results) is an extrachromosomal array carrying multiple copies of the lin-12(+) genomic region and is marked with rol-6(su1006).

Genetic mapping of the sel-5 locus:
To map sel-5 relative to mab-21, we examined recombinants segregating from unc-79(e1068) mab-21(bx53) dpy-17(e164)/sel-5(n1254) lin-12(n302); him-5(e1490) hermaphrodite parents. Of the 34 Unc non-Dpy hermaphrodites picked, 16 had the recombinant chromosome unc-79(e1068) sel-5(n1254) lin-12(n302), 16 had the chromosome unc-79(e1068) mab-21(bx53) lin-12(n302), and 2 had the chromosome unc-79(e1068) lin-12(n302). This last class of recombinants places sel-5 to the left of the cloned gene mab-21 (CHOW et al. 1995 ; Figure 1).

View larger version (6K): In this window In a new window Download PPT slide   Figure 1. The sel-5 region of LG III. Only the relevant genes, cosmids, and rearrangements mentioned in the text are shown. Both sel-5 and mab-21 are contained on cosmid F35G12 (30, 709 kb). unc-79 is ~0.5 cM to the left of sel-5, while dpy-17 is ~1.5 cM to the right of sel-5. The left endpoint of sDf121 was found to lie between sequences in F35G12 and F10G11 (see text).

To map sel-5 relative to sDf121, we mated sel-5(n1254) lin-12(n302); him-5(e1490) males to dpy-17(e164) sDf121(s2098) unc-32(e189)/dpy-17(e164) ncl-1(e1865) unc-36(e251) hermaphrodites. Non-Dpy F₁ hermaphrodite progeny were picked to individual plates and scored for egg laying, and their genotype was deduced from markers present in their progeny (F₂).

Determining the left endpoint of sDf121:
The left endpoint of sDf121 was determined using the polymerase chain reaction (PCR) with test primers in that region on unhatched eggs laid by the strain dpy-17(e164) sDf121(s2098) unc-32(e189); sDp3, essentially as was previously done for the right endpoint of this deletion (STEWART et al. 1998 ). The main difference is that in every reaction, we included a positive control primer pair (known to be outside the region of the deletion; GCGATTGGGCGAACTGGTAACCACAG and CCGTCGAGCCAGCCAAGCGACAACCATCGC) and a negative control primer pair (from a region known to be deleted in sDf121 and is covered by sDp3, GGTGGTATTATTGTATCCATAAACGCand AGTATTGACACCCAAAGAATATAAC), in addition to the test primer pair (TGATTACTGTAAGTTGCTACAAGATA and AATGTCTTCAGTATGTAGTTTGTTAC for F35G12; CTAACATCATTCTATTGAGCTGCTTG and TTGGACAATGTGCCGAAAGTTCAGAC for F56F3). More than 10 eggs were scored for the presence of the test primer pair expected band (~400 bp) on an agarose gel when the expected band (~200 bp) was detected for the positive control primer pair (signifying the presence of DNA in the reaction) and no band of the expected size (~600 bp) was detected for the negative control primer pair (signifying that the unhatched egg did not have sDp3).

Analysis of mutant phenotypes:
Two independent isolates of each genotype were tested; in every case, they gave similar results, and the data are pooled in the tables. The presence of sel-5 mutations was confirmed by sequencing and/or complementation analysis (see Table 1) in strains in which these alleles did not have a visible effect. For all assays, Egl⁺ hermaphrodites were allowed to lay eggs for a timed interval (usually overnight) and all progeny produced during that interval were analyzed; for egg-laying defective (Egl) hermaphrodites, all progeny were analyzed. Similar results were seen with both methods. To score egg laying, L4 hermaphrodites were placed on a single seeded plate and checked for 2 (20°) or 4 (15°) more days. To check the anchor cell, the presence of a vulva, or the existence of dorsal coelomocytes, worms at the appropriate stage were checked at 630x magnification with a Zeiss Axiophot compound microscope. The number of pseudovulvae in hermaphrodites and males and the presence of ectopic hooks in male tails were all scored with a dissection microscope.

Molecular analysis:
Standard methods were used for the manipulation of recombinant DNA (SAMBROOK et al. 1989 ), unless otherwise indicated. All enzymes were from New England Biolabs (Beverly, MA), unless otherwise indicated. PCR was done using the Expand long template PCR system (Boehringer Mannheim, Mannheim, Germany) according to the manufacturer's instructions.

sel-5 cDNA and mutant alleles sequence analysis:
To determine the 5' end of sel-5, we performed PCR on a C. elegans cDNA library (BARSTEAD and WATERSTON 1989 ) using the primers PstSL1 (ACACTGCAGGGTTTAATTACCCAAGTTTGAG), containing a PstI site (underlined) followed by the SL1 trans-spliced leader; (KRAUSE and HIRSH 1987 ) and Hind-F35N (ACAAAGCTTCTACAGTTCCAGAACCGGATGGCATTG), containing a HindIII site (underlined) followed by sel-5 sequence from an exon in the cDNA clone yk7a7. The 1.1-kb PCR fragment was digested with PstI and HindIII and subcloned into the same sites in the pBluescript SK⁺ vector (Stratagene, La Jolla, CA) to give plasmid pJF93.

To determine the complete sel-5A cDNA sequence, we linked the overlapping sequences of the yk7a7 (generously provided by Y. Kohara) and pJF93 inserts. To determine the complete sel-5B cDNA sequence, we linked the overlapping sequences of the yk13e1 (generously provided by Y. Kohara) and pJF93 inserts; we also sequenced the insert in plasmid pJF99 (see below).

The lesions associated with sel-5(n1250) and sel-5(n1254) were determined by sequencing several PCR products from single-stranded templates (ALLARD et al. 1991 ; KALTENBOECK et al. 1992 ), by use of internal primers to cover all exons and exon/intron boundaries. To determine the extent of the deletion in sel-5(ok149), we did single worm PCR (WILLIAMS et al. 1992 ) on these worms using the primers internal to sel-5 (except for the added restriction sites) PstIR1 (ACACTGCAGTTTCTTCCAGGTGGATTTGC; the added PstI site is underlined) and BamIL1 (ACAGGATTCCAAACACATCATCCACCACC; the added BamHI site is underlined). The 1.3-kb PCR fragment was digested with PstI and BamHI, subcloned into the same sites of pBluescript SK⁺, and sequenced. sel-5(ok149) has 1.977 kb deleted and the sequence AAATTCAAATGACGAGAT inserted in its place.

Sequence comparisons and alignments were obtained using the Blast program (ALTSCHUL et al. 1990 ) through the NCBI Web site and GCG version 8 programs (DEVEREUX et al. 1984 ).

Plasmid constructions:
The vector litmus38D1 was made by digesting litmus38 (New England Biolabs) with MfeI and EcoRI and religating the compatible ends, thus removing all restriction sites between them. The vector PIN2 drives inserted sequences under the control of sel-12 regulatory sequences; it contains unique BamHI and NotI sites inserted at the second amino acid of a sel-12 rescuing genomic fragment containing 2.8 kb of 5' flanking region (D. LEVITAN and I. GREENWALD, unpublished observations).

Plasmid pJF98 is the 13.5-kb NheI fragment from cosmid F35G12 (generously provided by A. Coulson) subcloned into the same site in litmus38D1. This plasmid contains a genomic fragment of sel-5 with 6.237 kb of 5' flanking sequence and 0.589 kb of 3' flanking sequence (downstream of the sel-5A polyadenylation site).

pJF99 is the 2.08-kb PCR fragment amplified from a C. elegans cDNA library (BARSTEAD and WATERSTON 1989 ) using primers F35A1 (ACACGGCCGATGCCTCTAGGGCTTTTCAGCTCTGGAAAAG; the added EagI site is underlined and is followed by the initiation codon of sel-5 shown in boldface) and F35B1(ACACGGCCGCTAAACTTGAAAACCACGAGAAGTGGTTC; the added EagI site is underlined and is followed by the stop codon of sel-5B shown in boldface type), restriction digested with EagI, and inserted into the same site in litmus38D1. The pJF99 insert was completely sequenced to confirm it contains the sel-5B cDNA. pJF99 was digested with SnaBI and ApaI (sites present only in the vector), blunt ended with T4 DNA polymerase, and religated, thus removing all the restriction sites in between and creating plasmid pJF101. The 2.08-kb EagI fragment from pJF99 was subcloned into the NotI site of PIN2 yielding plasmid pJF120 (sel-5B under the control of sel-12). PCR was done on plasmid KSGFPS65T (LEVITAN and GREENWALD 1998 ) using primers SpeGFP1 (ACAACTAGTCCCATGAGTAAAGGAGAAGAACTTTTCACTGG; the added SpeI site is underlined) and SpeGFP2 (ACAACTAGTTTTGTATAGTTCATCCATGCCATGTC; the added SpeI site is underlined) yielding a 0.7-kb fragment. This fragment was digested with SpeI and inserted into the same site of pJF101, in frame, to get a translational fusion, thus yielding plasmid pJF104. The 2.7-kb EagI fragment from pJF104 was subcloned into the NotI site of PIN2 yielding plasmid pJF110, which therefore has SEL-5B::GFP(S65T) under the control of sel-12 sequences.

pJF103 is the 1.64-kb EagI-SpeI fragment from pJF101 ligated with the 5.4-kb EagI-SpeI fragment from pyk7a7 (circularized plasmid from yk7a7), thus reconstituting the sel-5A cDNA (ATG to stop codon with 3' untranslated sequences). PCR was done using pJF103 as template and the primers F35A1 and F35A2 (ACACGGCCGTTACAAGTCGGTTGGATCATCATGATCTTCC; the added EagI site is underlined and is followed by the stop codon of sel-5A shown in boldface type) and yielded a 3.2-kb fragment. This fragment was digested with EagI and inserted into the same site in litmus38D1 yielding plasmid pJF105, which contains the sel-5A cDNA (ATG to stop codon). pJF106 is pJF105 digested with SnaBI and ApaI (sites in vector), blunt ended with T4 DNA polymerase, and recircularized, thus removing the restriction sites between these two enzymes.

The 3.2-kb EagI insert from pJF106 was inserted into the NotI site of PIN2, thus yielding plasmid pJF113, which has sel-5A cDNA under the control of sel-12 regulatory sequences. pJF109 is the 0.7-kb PCR fragment (template KSGFPS65T, primers SpeGFP1 and SpeGFP2, see above) carrying GFP-(S65T), digested with SpeI, and inserted into the same site in pJF106, in frame. The 3.9-kb EagI fragment from pJF109 was subcloned into the NotI site of PIN2 to make plasmid pJF111, which therefore has SEL-5A::GFP(S65T) under the control of sel-12.

To place GFP at the C terminus of SEL-5A, PCR was done on KSGFPS65T using primers BamGFP1 (ACAGGATCCCATGAGTAAAGGAGAAGAACTTTTCACTGG; the added BamHI site is underlined) and BamGFP2 (ACAGGATCCTTTGTATAGTTCATCCATGCCATGTG; the added BamHI site is underlined) yielding a 0.7-kb fragment. This fragment was digested with BamHI and used to replace the 0.178-kb BamHI fragment in pJF105 to give plasmid pJF107. The 3.7-kb EagI insert from pJF107 was subcloned into the NotI site of PIN2, thus making plasmid pJF108, which has SEL-5A::GFP(S65T) (GFP at the C terminus; see RESULTS) under the control of sel-12.

Worm transformation:
Microinjection of DNA into the germ line of C. elegans hermaphrodites was done essentially as previously described (FIRE 1986 ; MELLO et al. 1991 ). Plasmid DNA was injected at 1–50 µg/ml. For most experiments, pRF4 (MELLO et al. 1991 ), carrying rol-6(su1006), was injected at 150 µg/ml as a cotransformation marker. For the SEL-5::GFP localization experiments, pMH86 (HAN and STERNBERG 1991 ), carrying wild-type dpy-20, was injected at 20 µg/ml into a dpy-20(e1282) background as a cotransformation marker.

Double-stranded RNA synthesis and microinjection:
Double-stranded RNA was synthesized in vitro essentially as described previously (FIRE et al. 1998 ). Briefly, RNA was synthesized from phagemid clones using an RNA transcription kit (Stratagene) in which both the T3 and the T7 polymerase were added to the same reaction. The RNA was purified on RNeasy columns (Qiagen, Valencia, CA), eluted in water, diluted in injection buffer (FIRE et al. 1998 ), and allowed to anneal at 37° for 10–30 min.

The RNA was microinjected into the pseudocoelomic space of young adult hermaphrodites. Injected worms were placed on individual seeded plates and the phenotype(s) of their F₁ progeny was checked.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The AC/VU decision and lin-12 genetics (background):
Two gonadal cells, named Z1.ppp and Z4.aaa, are initially equivalent in their developmental potential in that each has an equal chance of becoming the anchor cell (AC), a terminally differentiated cell type, or a ventral uterine precursor cell (VU), which contributes descendants to the ventral uterus (KIMBLE and HIRSH 1979 ). However, in any given hermaphrodite, only one of these cells will become the AC, while the other becomes a VU, depending on lin-12-mediated interactions between them (KIMBLE 1981 ; GREENWALD et al. 1983 ; SEYDOUX and GREENWALD 1989 ). This process is termed the "AC/VU decision." Studies of genetic mosaics and reporter genes have suggested that a stochastic variation in lag-2 (ligand) and lin-12 (receptor) activity between Z1.ppp and Z4.aaa is amplified by a feedback mechanism in both cells, so that lag-2 expression becomes restricted to the presumptive AC and lin-12 expression becomes restricted to the presumptive VU (SEYDOUX and GREENWALD 1989 ; WILKINSON et al. 1994 ).

Mutations that eliminate lin-12 activity result in two ACs (GREENWALD et al. 1983 ). Mutations that constitutively activate lin-12 cause the absence of an AC and hence such mutants lack a vulva (GREENWALD et al. 1983 ); this phenotype is referred to below as "0 AC-Egl." There are two different types of constitutively activated lin-12 alleles: lin-12(d) mutations, a set of missense mutations in the extracellular domain (GREENWALD and SEYDOUX 1990 ), and lin-12(intra), a transgene that encodes just the intracellular domain (STRUHL et al. 1993 ). The key difference between these activated forms is that LIN-12(d) proteins are transmembrane proteins that presumably must undergo proteolytic processing to release the intracellular domain for signal transduction, whereas LIN-12(intra) does not.

Genetic analysis of sel-5 in the AC/VU decision:
Two alleles of sel-5, sel-5(n1250) and sel-5(n1254), were identified in screens for suppressors of the 0 AC-Egl defect caused by the lin-12(d) allele lin-12(n950) (TAX et al. 1997 ). A new allele, sel-5(ok149), was generated by the C. elegans gene knockout consortium when the coding region was identified (as described below). The sel-5 mutations are essentially recessive to sel-5(+), but sel-5/sDf121 and heteroallelic combinations display suppressor activity (Table 1), suggesting that sel-5 mutations are loss-of-function. This inference has been confirmed by sequence analysis of sel-5 mutations, as described below.

sel-5 mutations appear to reduce lin-12 activity, as all three sel-5 alleles suppress the 0 AC-Egl phenotype of lin-12(d) alleles by restoring one AC (Table 1). The proportion of hermaphrodites with one AC depends on two factors:

1. Temperature: All three sel-5 alleles also appear to be cold sensitive, i.e., suppression of lin-12(d) is greater at low temperature (15°). As sel-5(n1254) and sel-5(ok149) are likely to be molecular null alleles, this observation suggests that the process in which sel-5 functions is cold sensitive.

2. The starting level of lin-12 activity: sel-5 mutations are more efficient suppressors when the level of constitutive lin-12 activity is lower. For example, sel-5 is an efficient suppressor of the 0 AC-Egl defect of lin-12(n302), a "weaker" activated allele, than lin-12(n137), a "stronger" activated allele. In addition, sel-5 homozygotes, and heteroallelic combinations, are more efficient suppressors of lin-12(n302)/+ than lin-12(n302).

To gain insight into the interactions between sel-5 and lin-12 in the AC/VU decision, we examined the interaction between sel-5 and lin-12(intra). lin-12(intra), like lin-12(d) alleles, results in constitutive lin-12 activity (STRUHL et al. 1993 ). However, sel-5 is unable to suppress the 0 AC defect associated with arIs12 [lin-12(intra)] (Table 2). We think that the failure of sel-5 to suppress lin-12(intra) is not likely due simply to differences in the degree of constitutive activity (point 2 above), as lin-12(intra) has a lower penetrance of the 0 AC defect and hence appears to have lower constitutive activity than lin-12(n302) and lin-12(n676). As described above, LIN-12(intra) is a cytosolic protein that is not associated with the plasma membrane, whereas LIN-12(d) mutations are transmembrane proteins. The interactions between sel-5 and different activated forms therefore suggest that sel-5 does not act in signal transduction by activated LIN-12 and instead suggest a role for SEL-5 prior to or during ligand-dependent release of the intracellular domain.

None of the sel-5 mutations cause defects in the number of anchor cells (Table 3). Mutations that reduce lin-12 activity might enhance the 2 AC defect of lin-12 hypomorphic alleles, or might cause a synthetic 2 AC defect when combined with mutations in genes encoding other factors that facilitate lin-12 activity. However, we have not seen any evidence for such interactions for lin-12(ar170), a lin-12 hypomorphic mutation (HUBBARD et al. 1996 ) (Table 3). We do not know if these negative results mean that full-length LIN-12 must be constitutively active for sel-5 to influence its activity, as in lin-12(d) mutants or when LIN-12(+) is overexpressed, or whether there is a difference in the degree to which lin-12 activity must be lowered to see an effect [i.e., lin-12(d) suppression may require less of a reduction in lin-12 activity than causing a 2 AC phenotype under these conditions].

We also investigated the potential involvement of sel-5 in transcriptional control of lag-2 and lin-12. In wild-type hermaphrodites, lag-2::lacZ and lin-12::lacZ transcriptional reporter genes are initially expressed in both Z1.ppp and Z4.aaa, and a stochastic fluctuation is amplified by a feedback mechanism so that only the presumptive anchor cell expresses LAG-2, while the presumptive VU expresses only LIN-12 (WILKINSON et al. 1994 ). A similar result is seen with a lin-12::gfp translational reporter (LEVITAN and GREENWALD 1998 ). We examined the expression of the lag-2::lacZ and lin-12::gfp in sel-5(n1254) hermaphrodites and saw no change in transcription patterns as compared to sel-5(+) (data not shown). These observations suggest that sel-5 is not involved in the transcriptional feedback mechanism that operates during the AC/VU decision.

Molecular cloning of sel-5:
sel-5 had been mapped previously to chromosome III between ced-4 and dpy-17 (TAX et al. 1997 ). We obtained additional map data placing sel-5 to the left of mab-21 and within the deficiency sDf121 (see Table 1 and MATERIALS AND METHODS). To correlate the genetic and physical maps in this area, we determined that the left endpoint of sDf121 deleted sequences within cosmid F35G12 (see MATERIALS AND METHODS) but did not delete sequences in cosmid F56F3 (see MATERIALS AND METHODS) or F10G11 (STEWART et al. 1998 ), thus placing sel-5 within a few open reading frames (ORFs) to the left of mab-21 (Figure 1).

We were not successful in our attempts to identify sel-5 sequences by an antisuppression assay: transgenic lines carrying cosmids and yeast artificial chromosomes (YACs) were generated in a sel-5(n1254) lin-12(n302) background, and the presence of egg-laying-defective hermaphrodites would have been an indication that the transgenes carried sel-5(+). Since sel-5 mutations appear to be loss-of-function mutations, we tried an alternative approach, RNA-mediated interference (FIRE et al. 1998 ), to deplete the activity of candidate genes. We injected double-stranded RNA (dsRNA) corresponding to various ORFs in the region into lin-12(n302) adult hermaphrodites and looked for F₁ progeny that were able to lay eggs. Injection of dsRNA corresponding to two independent cDNA clones, yk13e1 and yk7a7, yielded a high percentage of such egg-laying F₁ progeny (Table 4).

yk13e1 and yk7a7 both correspond to the same ORF, F35G12.3, of which there are two splice variants sharing a common N terminus (see below). To confirm that F35G12.3 is sel-5, we sequenced the predicted exons and exon/intron boundaries of F35G12.3 in the two original sel-5 alleles, sel-5(n1250) and sel-5(n1254), and found a single base change in each allele (see below). In addition, a deletion within F35G12.3 was generated by the C. elegans gene knockout consortium and behaves in complementation tests as a sel-5 allele [designated sel-5(ok149)] (MATERIALS AND METHODS). These results indicate that F35G12.3 corresponds to sel-5.

Molecular analysis of the sel-5 coding region and sel-5 mutations:
We sequenced yk7a7, yk13e1, and PCR products (MATERIALS AND METHODS) and found that all the exons and introns were as predicted by AceDB (EECKMAN and DURBIN 1995 ) and reported in sequence database (GenEmbl no. Z46242). The cDNA analysis indicated that there are two alternatively spliced variants of sel-5; yk7a7 corresponds to the longer cDNA (referred to as sel-5A) and yk13e1 corresponds to the shorter cDNA (referred to as sel-5B). The presence of these two splice variants was also confirmed by the appearance of two equally abundant bands of the appropriate sizes (3.7 and 2.4 kb) on a Northern blot of total N2 (wild-type) RNA probed with the complete sel-5 gene (data not shown). In the mRNA, SL1 is transpliced 10 bp upstream of the initiating AUG; a polyadenosine tail is added 352 bp after the stop codon in sel-5(A) and 298 bp after the stop codon of sel-5B.

sel-5A and sel-5B encode predicted products of 1077 and 690 amino acids, respectively. The first 653 amino acids of these products are identical and at the amino termini include a region of 325 amino acids that is homologous to the catalytic site of serine/threonine kinases (Figure 2; HANKS and HUNTER 1995 ). The putative kinase domain of SEL-5 is most homologous to the GAK1 kinase in rats and humans and their homologues in budding yeast (Figure 3). No other significant similarity to any known protein was detected outside of the kinase domain.

View larger version (195K): In this window In a new window Download PPT slide   Figure 2. sel-5 predicted proteins and mutant alterations. (A) Schematic representation of the two SEL-5 proteins. Similar shading indicates regions of identical amino acid sequence. The positions of the various mutations for the different alleles are indicated (see text for details). (B) Relevant genomic sequence of sel-5 and predicted protein sequences. Coding nucleotide (nt) sequence is shown in uppercase letters; noncoding sequence is shown in lowercase letters. Numbering of the nucleotide sequence starts at the A (+1) of the initiation codon. The positions of the SL1 trans-spliced leader sequence and the poly-adenosine tails for the alternatively spliced mRNAs are indicated by arrowheads above the nucleotide sequence. The nucleotide changes in sel-5(n1250) and sel-5(n1254) are also shown above the nucleotide sequence. The nucleotides underlined are deleted in sel-5(ok149) and replaced by the sequence AAATTCAAATGACGAGAT. The predicted SEL-5A protein sequence is shown below the nucleotide sequence. SEL-5B has the same amino terminus; the divergent SEL-5B predicted sequence is shown above the nucleotide sequence. The predicted end of the putative kinase domain is indicated. The double underlined amino acids (929–987) were replaced by GFP in the SEL-5::GFP construct (see text).

View larger version (71K): In this window In a new window Download PPT slide   Figure 3. SEL-5 kinase domain similarity. (A) Sequence alignment of the SEL-5 kinase domain to that from Saccharomyces cerevisiae Pak1p (GenEmbl no. U24167), rat rGAK1 (GenEmbl no. D38560), and human hGAK1 (GenEmbl no. D88435). The 12 conserved subdomains of the kinase catalytic site are numbered above the sequences. Identical amino acids are reverse contrasted. (B) The percentage relatedness of the different proteins to SEL-5 was calculated by dividing the number of identical (or identical plus similar) amino acids by the longer of the two sequences. The following amino acids were considered similar: I/L/V, S/T, R/K, N/Q, and D/E. (Top) The proteins most similar to SEL-5. These include S. cerevisiae Pak1p, S. cerevisiae YNL020C (GenEmbl no. Z71296) and YBR059C (GenEmbl no. Z35928), Schizosaccharomyces pombe SPBC6B1.02 (GenEmbl no. AL021838), Drosophila heteroneura partial sequence AF052296 (GenEmbl no. AF052296), rat GAK1 and human GAK1. (Bottom) A comparison of SEL-5 to representative members from other families of ser/thr kinases: S. cerevisiae Cdc28p (GenEmbl no. X00257), rat PKCa (GenEmbl no. X07286), and human c-src (GenEmbl no. 59932).

The sel-5(n1254) G-to-A transition destroys the acceptor splice site at the end of intron 3, and the sel-5(n1250) G-to-A transition destroys the donor splice site at the beginning of intron 8 (Figure 2). Both mutations are predicted to result in premature termination of both SEL-5A and SEL-5B. In the absence of cryptic alternative splicing, sel-5(n1254), which behaves by genetic criteria as a null allele (Table 1), is predicted to terminate prior to the kinase domain, while sel-5(n1250) is predicted to terminate after the kinase domain. sel-5(ok149) contains a deletion of sequences from within exon 5 to sequences within exon 10 that would remove amino acids 153–582 of both proteins. Translation of the predicted mRNA results in a premature stop codon due to a shift in the ORF in the kinase domain (the alternative amino acids QIQMTRFDQSERWTGECIYDG are predicted before the stop codon).

Both SEL-5A and SEL-5B can complement sel-5(n1254):
sel-5(n1254) lin-12(n302) hermaphrodites carrying transgenes corresponding to the sel-5 gene (complete ORF with 6.277-kb upstream sequences and 0.589-kb downstream sequences) remained egg-laying proficient; we observed what appears to be transient rescue in some lines in early generations that was lost in later ones (data not shown). Cosmid F35G12, which contains sel-5(+), also shows this behavior (data not shown), so we believe the lack of antisuppression reflects low expression of sel-5, perhaps because of some property of extrachromosomal arrays carrying sel-5 sequences.

To assess the function of SEL-5A and SEL-5B in the AC/VU decision, we therefore used heterologous regulatory sequences, from the sel-12 gene, to drive their expression (see MATERIALS AND METHODS). Hermaphrodites of genotype sel-5(n1254) lin-12(n302) carrying extrachromosomal arrays expressing either SEL-5A or SEL-5B under the control of sel-12 regulatory sequences were egg-laying defective (Table 5). This reversal of suppression by both products indicates that both SEL-5A and SEL-5B can function in the AC/VU decision and that the sequences unique to SEL-5A are not necessary for this function of sel-5.

SEL-5A subcellular localization:
To investigate the subcellular localization of SEL-5A, we replaced 59 amino acids at its carboxyl-terminus with GFP (CHALFIE et al. 1994 ) and expressed the resulting SEL-5A::GFP hybrid protein under the control of sel-12 regulatory sequences. The SEL-5A::GFP protein is functional, as sel-5(n1254) lin-12(n302) hermaphrodites carrying transgenes expressing SEL-5A::GFP are egg-laying defective (Table 5). We examined the subcellular localization of the hybrid protein in sel-5(+) lin-12(+) hermaphrodites. The hybrid protein seems to localize predominantly to the cytoplasm of various cells (including Z1.ppp and Z4.aaa; data not shown) and is excluded from the nucleus (Figure 4); several independent transgenic lines showed a similar pattern of subcellular localization.

View larger version (32K): In this window In a new window Download PPT slide   Figure 4. Intracellular distribution of SEL-5::GFP protein. Shown are epifluorescent micrographs of L4 (A) or L3 (B) stage dpy-20(e1282) hermaphrodites with an extrachromosomal array carrying the wild-type dpy-20 gene and the PIN-2::SEL-5A plasmid (sel-5A::gfp under the control of sel-12 sequences). (A) Cytoplasmic staining (nuclear excluded) in head neurons. (B) Cytoplasmic staining (nuclear excluded) in the VPC (arrow) and neuronal processes in the ventral midcord (arrowheads).

sel-5 displays tissue-specific interactions with lin-12:
We assessed the ability of sel-5 mutations to suppress phenotypes other than the 0 AC-Egl phenotype associated with lin-12(d) mutations (Table 6). In hermaphrodites, strong lin-12(d) mutations cause a highly penetrant Multivulva phenotype, because the cells P3.p-P8.p, also called the vulval precursor cells (VPC), adopt a particular vulval fate, termed "2^o", and generate pseudovulvae. In addition, hermaphrodites carrying strong lin-12(d) mutations are missing dorsal coelomocytes, because the cells M.dlpa and M.drpa instead become sex myoblasts. In males, strong lin-12(d) mutations cause ectopic hooks (sensory structures), because the cells P9.p-P11.p adopt the male equivalent of the "2^o" fate, and cause P3.p-P6.p to generate pseudovulvae inappropriately. We saw no effect on the Multivulva phenotype of lin-12(d) hermaphrodites or the generation of ectopic hooks or pseudovulvae in males. However, we did see suppression of the transformation in fate of dorsal coelomocytes to sex myoblasts.

We also explored potential interactions between sel-5 and glp-1, another C. elegans lin-12/Notch gene (YOCHEM and GREENWALD 1989 ). lin-12 and glp-1 are functionally redundant for certain cell fate decisions (LAMBIE and KIMBLE 1991 ), and GLP-1 can efficiently substitute for LIN-12 when expressed under the control of lin-12 regulatory sequences (FITZGERALD et al. 1993 ). We combined sel-5(n1254) with glp-1(e2141) and glp-1(e2142), which are partial loss-of-function alleles at 15° (AUSTIN and KIMBLE 1987 ; PRIESS et al. 1987 ), and saw no effect on total brood size, maternal effect lethality, or fertility (data not shown). We also injected glp-1(ar202) hermaphrodites with sel-5 yk7a7 dsRNA and did not see any suppression of the gain-of-function mitotic proliferation defect of this allele (data not shown).

These observations suggest that sel-5 influences lin-12 activity in a tissue-specific manner. We do not know whether sel-5 is expressed in a tissue-specific manner or whether sel-5 activity influences lin-12 activity only under certain conditions.

RNA-mediated interference analysis of selected genes pertaining to GAK1 or GAK1 interacters:
The GAK1 family of proteins has been shown to interact either functionally or physically with numerous other genes and proteins, most notably, cyclin G and p53 (THIAGALINGAM et al. 1995 ; KANAOKA et al. 1997 ). The similarity of the SEL-5 kinase domain to the GAK1 family suggested that other GAK1-interacting proteins might play a role in the AC/VU decision. To test this hypothesis, we used RNA-mediated interference (RNAi) to reduce endogenous gene activity of C. elegans relatives of HMG1, p53-binding protein, Rb, c-abl, Ref-1, cyclin G, Cdk5, and PP2A. We injected dsRNA for these genes into lin-12(n302) adult hermaphrodites and looked for suppression of the egg-laying defect in the F₁ progeny. We saw no evidence for suppression (Table 7), in contrast to the efficient suppression observed when sel-5 dsRNA is injected (Table 4).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Previous work suggested that sel-5 facilitates lin-12 activity cell autonomously (TAX et al. 1997 ). In this study, we have performed additional genetic analysis to explore the effect of sel-5 on lin-12 activity, and below, we speculate about the possible function of sel-5 in lin-12-mediated cell fate decisions.

We have also cloned the sel-5 gene and showed that it encodes two alternatively spliced products, which share a serine/threonine kinase domain. The kinase domain of SEL-5 is most similar to that of mammalian GAK1 and yeast Pak1p. GAK1 was isolated as a cyclin G-interacting protein and was also shown to co-immunoprecipitate with CDK5 (KANAOKA et al. 1997 ). Pak1p was isolated in yeast as a multicopy suppressor of a yeast mutation defective in p53-mediated transcriptional activation (THIAGALINGAM et al. 1995 ). p53 is known to transcriptionally regulate the expression and/or activity of several genes and to interact with some others (reviewed by KO and PRIVES 1996 ; LEVINE 1997 ), including cyclin G, CDK5, phosphatase 2A ß' subunit, and HMG1. p53 is also thought to exert part of its effect on the cell cycle through its regulation of the phosphorylation of Rb (reviewed by KO and PRIVES 1996 ; LEVINE 1997 ). However, we have not been able to detect any obvious links between sel-5 or lin-12 and the p53 or Rb pathway in the AC/VU decision.

We have shown that loss of sel-5 activity can suppress the AC fate transformation associated with constitutive activity of lin-12(d) mutations, which are point mutations in the extracellular domain of LIN-12, but not of lin-12(intra), the untethered intracellular domain. This result suggests that sel-5 acts prior to or during ligand-dependent release of the intracellular domain.

The genetic interactions between sel-5 and constitutively active lin-12 alleles are in many ways reminiscent of the interactions between lin-12 and sel-12, a presenilin (LEVITAN and GREENWALD 1995 , LEVITAN and GREENWALD 1998 ), and between lin-12 and sup-17, a metalloprotease homolog of ADAM10/Kuzbanian (WEN et al. 1997 ). SEL-12/presenilin has been implicated in the ligand-dependent release of the intracellular domain (DESTROOPER et al. 1999 ; STRUHL and GREENWALD 1999 ), and SUP-17/ADAM10 has been proposed to be involved in a cleavage event that occurs upon ligand binding (LOGEAT et al. 1998 ). The similar genetic interactions raise the possibility that sel-5 is involved in modulating one of these proteolytic processing events, perhaps by activating some component by phosphorylation. The genetic interactions between sel-5 and the two classes of constitutively active lin-12 alleles would be consistent also with a role for SEL-5 in the trafficking of LIN-12 (or processing factors) to the cell surface. However, in this context, we note that we see no evidence for a change in the subcellular distribution or accumulation of a LIN-12::GFP hybrid protein in a sel-5 mutant background (H. FARES, unpublished observations).

In contrast to sup-17 and sel-12, the ability of sel-5 to suppress lin-12(d) mutations is tissue specific: loss of sel-5 activity can suppress defects in the AC/VU decision of the somatic gonad and the sex myoblast/coelomocyte decision, but cannot suppress defects in ventral hypodermal cell fate decisions. It is possible that sel-5 is not expressed in ventral hypodermal cells or that its function in ventral hypodermal cells is masked by a redundant activity or process in these cells. Alternatively, it is possible that a more interesting biological difference underlies the tissue specificity. For example, one difference between the lin-12-mediated cell fate decisions that are not affected by sel-5 activity (in the ventral hypodermis) vs. the cell fate decisions that are affected by sel-5 activity (the AC/VU decision and sex myoblast/coelomocyte decision) is that the unaffected decisions involve ectodermal derivatives, whereas the affected decisions involve mesodermal derivatives. Another obvious difference is that the unaffected cells are epithelial cells with a well-defined apical/basolateral axis of polarity. Where the epithelial cell polarity machinery is not active, perhaps contact between LIN-12 and its ligand(s) results in cell polarization or specific membrane microdomains; if so, then sel-5 might be involved in defining the axis of polarity or might influence the transport to or modification of a component at the region of cell contact.

Loss of sel-5 activity does not cause any cell fate transformations associated with loss of lin-12 (or glp-1) activity. The absence of a visible phenotype has been a characteristic of many genes recovered in suppressor/enhancer screens for genes that influence lin-12 activity in C. elegans, and in principle might reflect functional redundancy due to related genes or of different mechanisms that influence receptor activity. In the case of sel-12, null mutants lack the hallmark lin-12 cell fate transformations affecting the AC and VPC (LEVITAN and GREENWALD 1995 ) because a strict requirement for presenilin activity is masked by the presence of a second, functionally redundant, presenilin gene, hop-1 (LI and GREENWALD 1997 ). At this time, there is no obvious candidate for a gene that may be affording functional redundancy for sel-5, as there is no gene with high similarity in the C. elegans genome sequence database. We therefore favor the view that the lack of a sel-5 null mutant phenotype reflects a relatively small effect on lin-12 activity, possibly because other mechanisms afford some redundant influence.

	ACKNOWLEDGMENTS

We gratefully acknowledge the generosity of Jim Thomas and David Baillie for unpublished material and information. The sel-5 deletion allele ok149 was generated and generously provided by the C. elegans Gene Knockout Consortium (Oklahoma Medical Research Foundation). Much credit is due to Bob Barstead, Alan Coulson, Yuji Kohara, Gary Moulder, John Sulston, Bob Waterston, and their colleagues for their vital work on the C. elegans genome and cDNA sequence projects. We also thank Richard Ruiz and Ilya Temkin for their excellent technical assistance, Barth Grant and Xiajun Li for comments on the manuscript, and past and present members of the laboratory for very helpful discussions. H.F. is a Postdoctoral Associate and I.G. is an Investigator of the Howard Hughes Medical Institute.

Manuscript received July 1, 1999; Accepted for publication August 9, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALLARD, M. W., D. L. ELLSWORTH, and R. L. HONEYCUTT, 1991  The production of single-stranded DNA suitable for sequencing using the polymerase chain reaction. BioTechniques 10:23-26.

ALTSCHUL, S. F., W. GISH, W. MILLER, E. W. MEYERS, and D. J. LIPMAN, 1990  Basic local alignment search tools. J. Mol. Biol. 215:403-410[Medline].

AUSTIN, J. and J. KIMBLE, 1987  glp-1 is required in the germ line for the regulation of the decision between mitosis and meiosis in C. elegans.. Cell 51:589-599[Medline].

BAIRD, S. E., D. H. FITCH, I. A. KASSEM, and S. W. EMMONS, 1991  Pattern formation in the nematode epidermis: determination of the arrangement of peripheral sense organs in the C. elegans male tail. Development 113:515-526[Abstract].

BARSTEAD, R. J. and R. H. WATERSTON, 1989  The basal component of the nematode dense-body is vinculin. J. Biol. Chem. 264:10177-10185[Abstract/Free Full Text].

BRENNER, S., 1974  The genetics of Caenorhabditis elegans.. Genetics 77:71-94[Abstract/Free Full Text].

CHALFIE, M., Y. TU, G. EUSKIRCHEN, W. W. WARD, and D. C. PRASHER, 1994  Green fluorescent protein as a marker for gene expression. Science 263:802-804[Abstract/Free Full Text].

CHEN, Y., W. H. FISCHER, and G. N. GILL, 1997  Regulation of the ERBB-2 promoter by RBPJk and NOTCH. J. Biol. Chem. 272:14110-14114[Abstract/Free Full Text].

CHOW, K. L., D. HALL, and S. W. EMMONS, 1995  The mab-21 gene of C. elegans encodes a novel protein required for choice of alternate cell fates. Development 121:3615-3626[Abstract].

DESTROOPER, B., W. ANNAERT, P. CUPERS, P. SAFTIG, and K. CRAESSAERTS et al., 1999  A presenilin-1-dependent gamma-secretase-like protease mediates release of Notch intracellular domain. Nature 398:518-522[Medline].

DEVEREUX, J., P. HAEBERLI, and O. SMITHIES, 1984  A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387-395.

EASTMAN, D. S., R. SLEE, E. SKOUFOS, L. BANGALORE, and S. BRAY et al., 1997  Synergy between Suppressor of Hairless and Notch in regulation of Enhancer of split m gamma and m delta expression. Mol. Cell. Biol. 17:5620-5628[Abstract].

EECKMAN, F. H. and R. DURBIN, 1995  ACeDB and Macace. Methods Cell Biol. 48:583-605[Medline].

FIRE, A., 1986  Integrative transformation of Caenorhabditis elegans.. EMBO J. 5:2673-2680[Medline].

FIRE, A., S. XU, M. K. MONTGOMERY, S. A. KOSTAS, and S. E. DRIVER et al., 1998  Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans.. Nature 391:806-811[Medline].

FITZGERALD, K., H. A. WILKINSON, and I. GREENWALD, 1993  glp-1 can substitute for lin-12 in specifying cell fate decisions in Caenorhabditis elegans.. Development 119:1019-1027[Abstract].

GREENWALD, I., 1998  LIN-12/Notch signaling: lessons from worms and flies. Genes Dev. 12:1751-1762[Free Full Text].

GREENWALD, I. and G. SEYDOUX, 1990  Analysis of gain-of-function mutations of the lin-12 gene of Caenorhabditis elegans.. Nature 346:197-199[Medline].

GREENWALD, I., P. STERNBERG, and H. R. HORVITZ, 1983  The lin-12 locus specifies cell fates in Caenorhabditis elegans.. Cell 34:435-444[Medline].

HAN, M. and P. W. STERNBERG, 1991  Analysis of dominant-negative mutations of the Caenorhabditis elegans let-60 ras gene. Genes Dev. 5:2188-2198[Abstract/Free Full Text].

HANKS, S. K. and T. HUNTER, 1995  The eukaryotic protein kinase superfamily: kinase (catalytic) domain structure and classification. FASEB J. 9:576-596[Abstract].

HEDGECOCK, E. M. and R. K. HERMAN, 1995  The ncl-1 gene and genetic mosaics of Caenorhabditis elegans.. Genetics 141:989-1006[Abstract].

HODGKIN, J., A. PAPP, R. PULAK, V. AMBROS, and P. ANDERSON, 1989  A new kind of informational suppression in the nematode C. elegans.. Genetics 123:301-313[Abstract/Free Full Text].

HODGKIN, J. A., H. R. HORVITZ, and S. BRENNER, 1979  Nondisjunction mutants of the nematode Caenorhabditis elegans.. Genetics 91:67-94[Abstract/Free Full Text].

HOSONO, R., K. HIRAHARA, S. KUNO, and T. KURIHARA, 1982  Mutants of Caenorhabditis elegans with dumpy and rounded head phenotype. J. Exp. Zool. 235:409-421.

HUBBARD, E. J. A., Q. DONG, and I. GREENWALD, 1996  Evidence for physical and function association between EMB-5 and LIN-12 in Caenorhabditis elegans.. Science 273:112-115[Abstract].

JARRIAULT, S., S. BROU, F. LOGEAT, E. H. SCHROETER, and R. KOPAN et al., 1995  Signalling downstream of activated mammalian Notch. Nature 377:355-358[Medline].

KALTENBOECK, B., J. W. SPATAFORA, X. ZHANG, K. G. KOUSOULAS, and M. BLACKWELL et al., 1992  Efficient production of single-stranded DNA as long as 2 kb for sequencing of PCR-amplified DNA. BioTechniques 12:164-171[Medline].

KANAOKA, Y., S. H. KIMURA, I. OKAZAKI, M. IKEDA, and H. NOJIMA, 1997  GAK: a cyclin G associated kinase contains a tensin/auxilin-like domain. FEBS Lett. 402:73-80[Medline].

KIMBLE, J., 1981  Alterations in cell lineage following laser ablation of cells in the somatic gonad of Caenorhabditis elegans.. Dev. Biol. 87:286-300[Medline].

KIMBLE, J. and D. HIRSH, 1979  The post-embryonic cell lineages of the hermaphrodites and male gonads in Caenorhabditis elegans.. Dev. Biol. 87:396-417.

KO, L. J. and C. PRIVES, 1996  p53: puzzle and paradigm. Genes Dev. 10:1054-1072[Free Full Text].

KOPAN, R., J. S. NYE, and H. WEINTRAUB, 1994  The intracellular domain of mouse Notch: a constitutively activated repressor of myogenesis at the basic helix-loop-helix region of MyoD. Development 120:2385-2396[Abstract/Free Full Text].

KOPAN, R., E. H. SCHROETER, H. WEINTRAUB, and J. S. NYE, 1996  Signal transduction by activated mNotch: importance of proteolytic processing and its regulation by the extracellular domain. Proc. Natl. Acad. Sci. USA 93:1683-1688[Abstract/Free Full Text].

KRAUSE, M. and D. HIRSH, 1987  A trans-spliced leader sequence on actin mRNA in C. elegans.. Cell 49:753-761[Medline].

LAMBIE, E. J. and J. KIMBLE, 1991  Two homologous regulatory genes, lin-12 and glp-1, have overlapping functions. Development 112:231-239[Abstract].

LEVINE, A., 1997  p53, the cellular gatekeeper for growth and division. Cell 88:323-331[Medline].

LEVITAN, D. and I. GREENWALD, 1995  Facilitation of lin-12-mediated signalling by sel-12, a Caenorhabditis elegans S182 Alzheimer's disease gene. Nature 377:351-354[Medline].

LEVITAN, D. and I. GREENWALD, 1998  LIN-12 protein expression and localization during vulval development in C. elegans.. Development 125:3101-3109[Abstract].

LI, X. and I. GREENWALD, 1997  HOP-1, a Caenorhabditis elegans presenilin, appears to be functionally redundant with SEL-12 presenilin and to facilitate LIN-12 and GLP-1 signaling. Proc. Natl. Acad. Sci. USA 94:12204-12209[Abstract/Free Full Text].

LIEBER, T., S. KIDD, E. ALCAMO, V. CORBIN, and M. W. YOUNG, 1993  Antineurogenic phenotypes induced by truncated Notch proteins indicate a role in signal transduction and may point to a novel function for Notch in nuclei. Genes Dev. 7:1949-1965[Abstract/Free Full Text].

LOGEAT, F., C. BESSIA, C. BROU, O. LEBAIL, and S. JARRIAULT et al., 1998  The Notch1 receptor is cleaved constitutively by a furin-like convertase. Proc. Natl. Acad. Sci. USA 95:8108-8112[Abstract/Free Full Text].

MELLO, C. C., J. M. KRAMER, D. T. STINCHCOMB, and V. A. AMBROS, 1991  Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 10:3959-3970[Medline].

PRIESS, J. R., H. SCHNABEL, and R. SCHNABEL, 1987  The glp-1 locus and cellular interactions in the early C. elegans embryos. Cell 51:601-611[Medline].

PULAK, R. and P. ANDERSON, 1993  mRNA surveillance by the Caenorhabditis elegans smg genes. Genes Dev. 7:1885-1897[Abstract/Free Full Text].

ROOKE, J., D. PAN, T. XU, and G. M. RUBIN, 1996  KUZ, a conserved metalloprotease-disintegrin protein with two roles in Drosophila neurogenesis. Science 273:1227-1231[Abstract].

ROSENBLUTH, R. E., C. CUDDEFORD, and D. L. BAILLIE, 1985  Mutagenesis in Caenorhabditis elegans. II. A spectrum of mutational events induced with 1500 R of gamma-radiation. Genetics 109:493-511[Abstract/Free Full Text].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Ed. 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SCHROETER, E. H., J. A. KISSLINGER, and R. KOPAN, 1998  Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature 393:381-386.

SEDENSKY, M. M. and P. M. MENEELY, 1987  Genetic analysis of halothane sensitivity in Caenorhabditis elegans.. Science 236:952-954[Abstract/Free Full Text].

SELKOE, D. J., 1998  The cell biology of beta-amyloid precursor protein and presenilin in Alzheimer's disease. Trends Cell Biol. 8:447-453[Medline].

SEYDOUX, G. and I. GREENWALD, 1989  Cell autonomy of lin-12 function in a cell fate decision in C. elegans.. Cell 57:1237-1245[Medline].

STEWART, H. I., N. J. O'NEIL, D. L. JANKE, N. W. FRANZ, and H. M. CHAMBERLAIN et al., 1998  Lethal mutations defining 112 complementation groups on a 4.5 Mb sequenced region of Caenorhabditis elegans chromosome III. Mol. Gen. Genet. 260:280-288[Medline].

STRUHL, G. and A. ADACHI, 1998  Nuclear access and action of Notch in vivo. Cell 93:649-660[Medline].

STRUHL, G. and I. GREENWALD, 1999  Presenilin is required for activity and nuclear access of Notch in Drosophila. Nature 398:522-525[Medline].

STRUHL, G., K. FITZGERALD, and I. GREENWALD, 1993  Intrinsic activity of the Lin-12 and Notch intracellular domains in vivo. Cell 74:331-345[Medline].

TAX, F. E., J. J. YEARGERS, and J. H. THOMAS, 1994  Sequence of C. elegans lag-2 reveals a cell-s