sel-7, a Positive Regulator of lin-12 Activity, Encodes a Novel Nuclear Protein in Caenorhabditis elegans

Corresponding author: Iva Greenwald, 701 W. 168th St., Room 720, College of Physicians and Surgeons, Columbia University, New York, NY 10032., greenwald{at}cancercenter.columbia.edu (E-mail)

Communicating editor: B. J. MEYER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Suppressor genetics in C. elegans has identified key components of the LIN-12/Notch signaling pathway. Here, we describe a genetic and molecular characterization of the suppressor gene sel-7. We show that reducing or eliminating sel-7 activity suppresses the effects of constitutive lin-12 activity, enhances the effects of partially reduced lin-12 activity, and causes a synthetic Lin-12(0) phenotype when combined with a null mutation in the sel-12 presenilin gene. These observations suggest that sel-7 is a positive regulator of lin-12 activity. We also show that SEL-7 encodes a novel nuclear protein. Through yeast two-hybrid screening, we identified an apparent interaction partner, K08E3.8, that also interacts with SEL-8, a known component of the nuclear complex that forms upon LIN-12 activation. Our data suggest potential roles for SEL-7 in the assembly or function of this nuclear complex.

CELLS with equivalent potentials are specified via cell-cell interactions to adopt distinct fates during animal development. Many interactions are mediated by the receptor LIN-12/Notch (GREENWALD 1998 ). LIN-12/Notch is activated by binding of a ligand of the DSL (Delta/Serrate/LAG-2) family. Ligand binding induces at least two proteolytic cleavages, one in the extracellular domain of the receptor and the other in the transmembrane domain. The latter cleavage releases the intracellular domain of the receptor, which translocates into the nucleus and forms a transcriptional activation complex with LAG-1 (known as Suppressor of Hairless in Drosophila and CBF1 or RBP-J in mammals) and other cofactors (WEINMASTER 2000 ).

The anchor cell (AC)/ventral uterine precourser cell (VU) decision is a well-characterized cell fate decision that is mediated by LIN-12 in Caenorhabditis elegans (GREENWALD 1998 ). In wild-type hermaphrodites, two gonadal cells interact with each other so that one becomes the AC and the other becomes the VU. Missense mutations or truncations that cause constitutive lin-12 activity cause both cells to become VUs, so that an AC is missing; the absence of an AC results in an egg-laying defective (Egl) phenotype that is easily visualized in the dissecting microscope. Extragenic suppressors defined by reversion of the 0 AC-Egl defect have identified a number of sel (suppressor/enhancer of lin-12) genes (TAX et al. 1997 ; L. VALLIER, I. KATIC, J. CHEN and I. GREENWALD, unpublished data). Characterized sel genes have been found to encode components important for LIN-12/Notch activity. Among these are two membrane-associated proteases that mediate proteolytic cleavage of LIN-12/Notch after ligand binding: SEL-12, an ortholog of human presenilin, which mediates transmembrane cleavage of LIN-12/Notch (LEVITAN and GREENWALD 1995 ; DE STROOPER et al. 1999 ; STRUHL and GREENWALD 1999 ), and SUP-17, an ortholog of Drosophila Kuzbanian and a mammalian ADAM protein, which is likely to mediate cleavage of the extracellular domain (WEN et al. 1997 ). Another sel gene, sel-5, encodes a cytoplasmic serine/threonine kinase that may influence LIN-12 trafficking (FARES and GREENWALD 1999 ).

In contrast to the SEL proteins that appear to have enzymatic activities, SEL-8 is a novel, glutamine-rich nuclear protein (DOYLE et al. 2000 ; PETCHERSKI and KIMBLE 2000A ). SEL-8 is required for LIN-12/Notch activity in C. elegans and appears to form a bridge between LAG-1 and the intracellular domain of LIN-12 (DOYLE et al. 2000 ; PETCHERSKI and KIMBLE 2000A ). There is no primary sequence homology with non-nematode proteins; however, SEL-8 has been proposed to serve as the counterpart to the glutamine-rich protein Mastermind of Drosophila and mammals (PETCHERSKI and KIMBLE 2000B ). Mastermind has been shown to be a component of the transcriptional activation complex that contains Suppressor of Hairless and the Notch intracellular domain (WU et al. 2000 ; KITAGAWA et al. 2001 ).

Here we report the characterization of another sel gene, sel-7 (first identified in TAX et al. 1997 ), which, like sel-8, also encodes a novel nuclear protein. Our analysis is consistent with a role for SEL-7 in the formation or function of the nuclear complex that activates expression of LIN-12/Notch target genes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

General C. elegans methods and strains:
Standard methods as described in BRENNER 1974 were used for handling, maintenance, and genetic analysis of C. elegans. Experiments were performed at 20° unless otherwise indicated. The wild-type parent for all strains was C. elegans var. Bristol N2 (BRENNER 1974 ), except for mapping experiments done with the Hawaiian strain CB4856. The main alleles used in this work are LG I: smg-1(r861), unc-54(r293) (HODGKIN et al. 1989 ); LG III: lin-12(n302, n137, n950) (GREENWALD et al. 1983 ), lin-12(ar170) (HUBBARD et al. 1996 ), glp-1(ar202) (PEPPER et al. 2003 ), glp-1 (e2141, e2142) (PRIESS et al. 1987 ); and LG X: sel-7 (n1253, ar516, ar523, ar586) (TAX et al. 1997 and this study), sel-12 (ar171) (LEVITAN and GREENWALD 1995 ).

Transgenes:
arIs12[lin-12(intra)] 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) (STRUHL et al. 1993 ). arIs37 expresses a secreted form of green fluorescent protein (GFP) under the control of a muscle-specific promoter, so that GFP is secreted into the pseudocoelom; uptake of GFP from the pseudocoelom can be used to help visualize coelomocytes (FARES and GREENWALD 2001 ). arIs51 [cdh-3::gfp] (KARP and GREENWALD 2003 ) expresses GFP under the control of cdh-3 regulatory sequences and can be used to help visualize anchor cells (PETTITT et al. 1996 ).

Additional information about the alleles listed above and about unc-1(e538), unc-3(e151), unc-84(e1410), and other markers used for mapping and for facilitating genetic analyses in this work can be found via Wormbase at http://www.wormbase.org.

Mutant analysis and scoring:
Strains were grown and scored at 20° unless otherwise specified. Phenotypes were assessed as follows. For egg-laying defect, L4 larvae were picked onto individual plates and scored 2 or 3 days later. An animal was scored as Egl⁺ if it laid eggs and as Egl if it did not lay eggs or formed a bag of larvae during this period of time. The egg-laying ability of sterile individuals could not be assessed. For ACs, strains generally contained arIs51, which expresses GFP in the AC. L3 larvae were scored using Nomarski optics, and ACs were identified by both morphology and GFP expression. For multivulva, L4 larvae were picked onto individual plates and checked for the number of pseudovulvae the next day. Egg-laying-competent hermaphrodites were excluded from the final tallies, since the presence of an AC causes a normal vulva to form. For coelomocytes, the number of coelomocytes was determined in L4 larvae or young adults both by their characteristic morphology under Nomarski optics and by accumulation of GFP. For germline proliferation, individual glp-1 (ar202) L4 larvae grown at 20° were transferred to the nonpermissive temperature, 25°, and scored for progeny production. For embryonic lethality, individual glp-1(e2142) L4 larvae were transferred to fresh plates every day for 3 consecutive days, and the eggs laid on the plates were scored for hatching. This assay was conducted at 15°.

In addition to canonical lin-12 phenotypes, we also observed that the double mutant sel-7(n1253) unc-3(e151) formed dauers at 80% penetrance (25/30) at 25°, whereas neither single mutant did (data not shown). Loss of lin-12 activity caused by alleles such as lin-12(oz48) has been observed to enhance the frequency of dauer formation in dauer-constitutive mutant backgrounds (LEVITAN and GREENWALD 1998 ). To assess whether the synthetic dauer-constitutive phenotype resulted from reduced lin-12 activity in a sel-7 mutant, we examined lin-12(oz48); unc-3(e151) and found that it does not show a daf-c phenotype (0/51). In addition, it has been observed that unc-3 becomes dauer constitutive at 27° (AILION and THOMAS 2000 ). We therefore tested whether sel-7 forms dauers at 27°, but found that it does not. We did not attempt to determine the basis for the synergy between sel-7 and unc-3.

Genetic mapping of sel-7 locus and sequencing:
sel-7 has been previously mapped between unc-84 and unc-3 on LG X (TAX et al. 1997 ). We used single-nucleotide polymorphisms (SNPs) on LG X to refine the map position of sel-7(n1253). We identified the following SNPs: xl1 on cosmid F16B12; xl2F1R1 on K04C1; xl3DoaF4R7 on C11H1; xl4F7R7A and xl5F7R7B on T14C1; and xl6F2R2 on M163; xl7F2R2, jlF3R3, and jcF4R4 on K04G11. Males of the Hawaiian strain CB4856 were crossed to unc-84 sel-7(n1253) unc-3 hermaphrodites, and F₁ heterozygotes segregated Unc-84 non-Unc-3 F₂ recombinants. The presence of sel-7(n1253) in the recombinants was assessed later by the ability to suppress lin-12(n302). A total of 101 recombinants were obtained and analyses of the SNPs in these recombinants enabled us to map sel-7 between jlF3R3 and jcF4R4 on K04G11. We then sequenced the predicted genes on this cosmid. PCR product using one pair of primers, K04G11.2F1: 5'-TCTCTCGGATGGTACATATCGG-3' and K04G11.2R1: 5'-TACTTCAAACGTAAGGGGTG-3' contains a C-to-T mutation in the predicted gene K04G11.2.

cDNA cloning and correction of the GENEFINDER prediction for K04G11.2:
To determine the gene structure of sel-7 in the absence of any available expressed sequence tags, we screened three cDNA libraries and subsequently used the ProQuest cDNA library (GIBCO BRL, Gaithersburg, MD, catalog no. 10835) to amplify pieces of sel-7 cDNA. Two overlapping fragments were amplified by using primer sets (1) 5pPC86 (5'-TATAACGCGTTTGGAATCACT-3', recommended by the manufacturer) and K2C2 (5'-CTCGGTTTCCTTCGGCGATTG-3') and (2) K2C.33 (5'-CGAGGATTCTGATGGCGACTTAG-3') and 3pPC86 (5'-GTAAATTTCTGGCAAGGTAGAC-3', recommended by the manufacturer). The two pieces were then pasted together by PCR. Sequencing of the cDNA thus obtained revealed that the predicted gene K04G11.2 does not contain the entire coding region of sel-7. The first four exons of sel-7 correspond to K04G11.2 (rapid amplification of cDNA ends experiments showed that sel-7 was transpliced to a nonclassic spliced leader, 5'-GGTTTAATTACCCAAGTTTGAG-3') and then a 5-kb intergenic region was spliced out between the fourth and the fifth exon. The last three exons (fifth, sixth, and seventh) encompass what had been predicted as K04G11.6. This gene structure was further confirmed by RNA-mediated interference (RNAi) and the identification of a mutation in the fifth exon in another sel-7 allele, ar516.

RNAi :
For dsR NA m icro -inj ecti on: Doub le-s tranded RNA (dsRNA) was synthesized in vitro using the cloned PCR fragment amplified from N2 genomic DNA as template. To confirm the cloning and the gene structure of sel-7, the following pairs of primers were used to amplify different regions of sel-7 genomic sequences: (1) K04G11.2F1 (5'-TCTCTCGGATGGTACATATCGG-3') and K2R3 (5'-CGTGTTGTCTGCTTTCAATCG-3'), (2) 905R1(5'-GGTTTAAGAATTCCCGCACCG-3') and VKF0 (5'-TGAAGAGATAATGCTAGGCGAG-3'), (3) 905R1 and VKF1 (5'-CTAGATCAACGTTCACTGGTGG-3'), and (4) K2C.36 (5'-AGCCAGCTGAGTAAAGC-3') and VKF0. Both T7 and T3 polymerases were added to the same in vitro transcription reaction, or each strand of RNA was synthesized with T7 or T3 polymerase alone. The RNA was purified on RNeasy columns (QIAGEN, Valencia, CA) and then allowed to anneal in 0.5x injection buffer (MELLO and FIRE 1995 ). dsRNA was micro-injected into the pseudocoelomic space of young adult hermaphrodites. Injected worms were individually placed and their F₁ progeny were scored. For RNA feeding, full cDNA of sel-7 (see above) and the cDNA clones of F13B10.1B and K08E3.8 pulled out from the yeast two-hybrid screens (see below) were cloned into vector pPD129.36 (TIMMONS et al. 2001 ) and then transformed into bacterial strain HT115 (TIMMONS and FIRE 1998 ; TIMMONS et al. 2001 ). Eggs or arrested L1 larvae were placed onto a lawn of such bacteria and relevant phenotypes were later scored.

Plasmid construction:
sel-7 cDNA clone: sel-7 cDNA was obtained as described above. A SalI site was generated before the stop codon by PCR with a pair of complementary primers (5'-GACCCAAGTCTATTCAAGGTTGTCGACTAAAATCCTCGCCTAGC-3'; the underlined sequence indicates the SalI site). GFP was excised from a KSGFPS65T vector (pPD114.38; A. FIRE, personal communication) with SalI and inserted in frame into the cDNA construct. The 5' untranslated region of sel-12 was excised with SacI and NotI from the vector superPin2 (a modified version of Pin2; see FARES and GREENWALD 1999 , which has additional cloning sites), which contains the whole genomic region of sel-12, and then inserted before the sel-7 cDNA::gfp construct. For yeast two-hybrid plasmids, vectors were provided by ProQuest. pPC86 contains the GAL4 activation domain, and pDBLeu contains the GAL4 DNA-binding domain. Full-length cDNAs of lag-1, lin-12(intra), sel-7, sel-8, and skip (ZHOU et al. 2000 ) were cloned either between SalI and NotI sites or between SalI and SpeI sites on both plasmids except for the following constructs. Full-length cDNA of lag-1 was used for pPC86, but a truncated form that runs from Met630 of LAG-1 until the end of the protein was used for insertion into pDBLeu. Two forms of LIN-12(intra) were used; both start at Pro930, but end differently. For pPC86, LIN-12(intra) ends at the end of the protein. For pDBLeu, the insertion ends at Val1320 and a stop codon was inserted after the valine. For yeast three-hybrid plasmids, the third, non-GAL4-domain-containing yeast expression vector pRST was generated on the basis of pRS416. The promoter P_ADH and the terminator T_ADH were excised from pPC86 with KpnI, HindIII, and SmaI, BamHI, respectively, and inserted into pRS416 between the corresponding sites. Full-length cDNAs of lag-1, lin-12(intra) (Pro930 till TAA), and K08E3.8 were then cloned into the pRST vector.

Yeast two-hybrid and three-hybrid assays:
Plasmids, yeast cells, and protocols for the yeast two-hybrid and three-hybrid assays were provided by the ProQuest two-hybrid system (GIBCO BRL, catalog no. 10835). The yeast strain used for all the experiments in this study was MaV203, and all yeast cells were cultured at 30°. For yeast transformation, cells were streaked out from frozen stock and four single colonies were picked and suspended in YPAD medium and inoculated at 30° for overnight. On the next morning, the cells were diluted to OD₆₀₀ = 0.1 and cultured until OD₆₀₀ reached 0.4. Cells were collected by centrifuge and transformed with desired DNA according to the protocol provided by the manufacturer. Cells were then spread on proper nutrient-deficient plates (SC-Leu-Trp for two-hybrid assays and SC-Leu-Trp-Ura for three-hybrid assays) for plasmid selection and incubated at 30° for 72 hr. For interaction detection, four colonies from each plate were patched onto a new plate, allowed to grow for 18 hr, and then replica plated onto an 82-mm nitrocellulose membrane. Cells were left on the membrane on top of YPAD medium for another 24 hr at 30°, and the interaction was then tested by ß-galactosidase assay described as follows. The membrane was lifted off the plate, frozen at -80° for 20–30 min, transferred onto two layers of filter paper preabsorbed in the reaction buffer (5 ml Z-buffer, 30 µl ß-mercaptoethanol, 300 µl 3% X-Gal for each membrane), and then incubated at 37° for the reaction to take place. The appearance of blue colony color was assessed every hour for 1–4 hr and then again after overnight incubation.

Yeast two-hybrid screen:
The ProQuest two-hybrid C. elegans cDNA library was amplified according to the protocol provided by the manufacturer. pDBLeu-sel-7cDNA (see the plasmid construction above) was first transformed into yeast strain MaV203, and a single colony was picked and cultured in 2 ml SC-Leu medium at 30° overnight. On the next day, the cells were diluted into 50 ml SC-Leu medium and cultured again at 30° overnight. On the third morning, cells were diluted into 300 ml SC-Leu medium so that the OD₆₀₀ was 0.1 and then cultured at 30° until OD₆₀₀ reached 0.5. Cells were collected and transformed with 25 µl cDNA library DNA (750 µg/ml) and spread on 15-cm SC-Leu-Trp plates. Plates were incubated at 30° for 72 hr, and colonies were replica plated onto 13.52-cm nitrocellulose membrane filters. Positive colonies were selected by ß-galactosidase assay as follows: Membranes were frozen at -80° for 30 min, transferred onto two layers of filter paper preabsorbed in the reaction buffer (15 ml Z-buffer, 39 µl ß-mercaptoethanol, 300 µl 3% X-Gal for each membrane) and then incubated at 30° overnight. Blue colonies were then picked and streaked onto SC-Leu-Trp plates. Single colonies were picked and DNA was extracted with DNA-Pure plasmid mini-prep kit (CPG) and sequenced for its identity. We did two independent screens, and 15 blue colonies were obtained. Of these 15, 12 regrew on SC-Leu-Trp plates; these 12 colonies represented three different genes. We then tested their self-activation by transforming them with an empty pDBLeu plasmid using the methods described above. One of the three genes turned out to be self-activating and was removed from the positive list. Therefore, we obtained two different sel-7 interacting genes, F13B10.1B and K08E3.8, described further in the text.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

sel-7 is a positive regulator of lin-12 activity in the AC/VU decision:
In wild-type hermaphrodites, two initially equivalent cells, Z1.ppp and Z4.aaa, interact with each other so that one becomes the AC and the other becomes the VU; each cell has an equal chance to become the AC. In lin-12(d) mutants such as lin-12(n302), LIN-12 is constitutively active (GREENWALD and SEYDOUX 1990 ), both Z1.ppp and Z4.aaa become VUs, and the worm has no AC and hence is vulvaless and egg-laying defective (0 AC-Egl). sel-7(n1253) was originally identified by TAX et al. 1997 as a suppressor of the 0 AC-Egl defect of lin-12(n302), and we identified additional alleles of sel-7 in a similar screen (L. VALLIER, I. KATIC, J. CHEN and I. GREENWALD, unpublished observations). We have used sel-7(n1253) for most genetic experiments; sequence analysis described below suggests that it is likely to be a strong loss-of-function or null allele.

We did not observe any overt phenotype of sel-7 in an otherwise wild-type background, which is consistent with the observation by TAX et al. 1997 . However, genetic interactions with different classes of lin-12 alleles suggest that sel-7 is a positive regulator of lin-12 activity (Table 1 and Table 2). First, sel-7(n1253) suppresses the 0 AC-Egl phenotype caused by constitutively active LIN-12. For lin-12(d) missense mutations, which can be arranged into an allelic series (GREENWALD et al. 1983 ), the degree of suppression roughly correlates with the strength of the lin-12(d) allele: sel-7(n1253) is a more efficient suppressor of lin-12(n302), which has a relatively lower level of constitutive lin-12 activity, than of lin-12(n137), which is relatively "stronger." Second, sel-7 enhances the 2AC defect of lin-12(ar170), a partial loss-of-function allele.

In addition, loss of sel-7 activity displays synergy with loss of sel-12 activity (Table 2). sel-12 is another positive regulator of the LIN-12/Notch signal transduction pathway. sel-12 encodes a homolog of human presenilin and mediates the transmembrane cleavage that releases the LIN-12 intracellular domain from its membrane tether. Neither sel-12 nor sel-7 alone displays any defect in the AC/VU decision, yet the sel-12 sel-7 double mutant displays an incompletely penetrant 2AC phenotype, reminiscent of lin-12 partial loss-of-function alleles (Table 2). This result suggests that lin-12 activity is lower in a sel-7 mutant background, although not low enough to cause defects in the AC/VU decision.

Loss of sel-7 activity also suppresses the 0 AC phenotype caused by lin-12(intra), which, like lin-12(d) mutations, causes constitutive LIN-12 activity. LIN-12(intra) resembles the cleavage product produced after LIN-12 activation, so the ability of sel-7 to suppress lin-12(intra) is consistent with a role for sel-7 downstream of the activated receptor (Table 1).

sel-7 activity in other lin-12- or glp-1-mediated cell fate decisions:
We examined the ability of sel-7 to suppress lin-12(d) mutations for defects in the two cell fate decisions that have served as the basis for other genetic studies of lin-12, vulval precursor cell (VPC) specification and the decision made by certain mesodermal cells between the coelomocyte (cc) and sex myoblast fate. We also examined the ability of sel-7 to suppress or enhance glp-1 alleles for defects in germline proliferation and embryonic development.

VPC specification: There are six VPCs, consecutively numbered P3.p–P8.p. LIN-12/Notch signaling, along with other signaling events, specifies their fates (reviewed in GREENWALD 1998 ). In a wild-type hermaphrodite, only P5.p and P7.p adopt a fate called "2°." However, in a strong lin-12(d) background, such as lin-12(n137) and lin-12(n950), all six VPCs adopt the 2° fate, resulting in a distinctive Multivulva (Muv) phenotype. sel-7 does not suppress the Muv phenotype caused by lin-12(d) mutations, suggesting that sel-7 does not function in the VPCs or that loss of sel-7 activity does not reduce lin-12 activity sufficiently to suppress the Muv phenotype; alternatively, there may be functional redundancy between SEL-7 and another, unknown, protein (Table 3).

Sex mesoblast/coelomocyte specification: Two cells, M.drpa and M.dlpa, give rise postembryonically to coelomocytes that are located dorsally and generally posterior to the gonad. In lin-12(d) mutants, M.drpa and M.dlpa are transformed into their ventral equivalents and instead become sex mesoblasts. sel-7 partially suppresses this transformation (Table 3).

glp-1-mediated decisions: TAX et al. 1997 reported that sel-7(n1253) enhanced the effects of the glp-1(e2141) loss-of-function allele for embryonic lethality and brood size, suggesting that sel-7 plays a role in cell fate decisions mediated by glp-1. However, we did not see significant interaction of sel-7(n1253) with other glp-1 alleles. sel-7 did not enhance the embryonic lethality of glp-1(e2142) (PRIESS et al. 1987 ): 30% (38/125) of glp-1(e2142) embryos arrested at 15°, similar to 21% (25/120) of glp-1(e2142); sel-7(n1253) and 32% (55/172) of glp-1(e2142); sel-7(ar516) embryos. sel-7 did not suppress the sterility caused by the constitutively active allele glp-1(ar202) (PEPPER et al. 2003 ): 96% (54/56) of glp-1(ar202) hermaphrodites were sterile at 25°, as were 91% (52/57) of glp-1(ar202); sel-7(n1253) hermaphrodites.

Molecular identification of sel-7:
sel-7 had been previously mapped between unc-84 and unc-3 on LG X (TAX et al. 1997 ). We identified new SNPs (see MATERIALS AND METHODS) and mapped sel-7 onto one cosmid, K04G11 (Fig 1A). We began sequencing DNA prepared from sel-7(n1253) worms for predicted genes on this cosmid and found a point mutation in a predicted gene called K04G11.2, suggesting that that this gene corresponded to sel-7.

View larger version (118K): In this window In a new window Download PPT slide   Figure 1. Molecular cloning and DNA sequence analysis of sel-7. (A) Mapping and cloning of sel-7. Only the relevant genes and cosmids are shown. (B) Schematic of the SEL-7 protein. The SEL-7 cDNA sequence is in GenBank (accession no. AY398504). (C) Protein alignment of C. elegans SEL-7 and its apparent orthologs in C. briggsae (CBP15881), in A. ceylanicum (CB175283), in O. volvulus (AW208316), and in B. malayi (AA071502). Each of the SEL-7 homologs was aligned with SEL-7 sequence using program ClustalW, and the alignments are superimposed here.

Our analysis of cDNAs (see MATERIALS AND METHODS), however, revealed that K04G11.2 was actually joined to another predicted gene that lay in the same orientation, then called "K04G11.6." Within the 5-kb interval separating K04G11.2 and K04G11.6 was another predicted gene in the opposite orientation. The composite gene, now listed as K04G11.2 in Wormbase, contains seven exons; the first four were originally from K04G11.2. Corroborating evidence for this structure is described in the next section.

Corroborating evidence that K04G11.2 is sel-7 and the sel-7 null phenotype:
We have verified that the composite K04G11.2/K04G11.6 (now listed as K04G11.2) corresponds to sel-7 by three criteria. The first two of these criteria also indicate that the null phenotype of sel-7 is Sel; i.e., K04G11.2 acts as a suppressor/enhancer of lin-12 and does not cause a phenotype that is readily detected in an otherwise wild-type background.

First, we found sequence alterations associated with other sel-7 alleles throughout the composite gene (Fig 1B). The sequence analysis also suggests that most or all of the sel-7 alleles that we obtained are strong loss-of-function or null mutations. sel-7(n1253), the reference allele used for most of our genetic analysis, is predicted to truncate the SEL-7 protein about halfway through, at amino acid 171; although it seems likely, we cannot conclude from the sequence information that sel-7(n1253) is a molecular null. sel-7(ar523) appears likely to be a molecular null allele. sel-7(ar523) is predicted to truncate SEL-7 at amino acid 55 and behaves like sel-7(n1253) in terms of suppression of lin-12(n302) (Table 1). sel-7(n1253) does not have any visible phenotype or any effect on the number of ACs in an otherwise wild-type background.

Second, we injected four different dsRNAs, corresponding to different regions of the composite gene, into lin-12(n302) hermaphrodites and assayed the progeny of injected hermaphrodites for egg-laying ability (see Fig 1A). All four dsRNAs suppressed the 0 AC-Egl phenotype of lin-12(n302): For a dsRNA corresponding to exons 1–4, 22/29 injected hermaphrodites gave egg-laying progeny; for exons 5 and 6, 2/2; for exons 6 and 7, 5/5; and for exons 5–7, 8/9. The fact that depletion of sel-7 activity by RNAi suppresses lin-12(n302) corroborates the inference from sequence analysis that sel-7 mutations reduce or eliminate sel-7 activity.

Third, we obtained evidence that a cDNA corresponding to the corrected K04G11.2 sequence can complement sel-7(n1253). We were unable to achieve good expression or rescue with a genomic fragment corresponding to the sel-7 region in simple arrays (data not shown), a problem also encountered in the analysis of sel-5 (FARES and GREENWALD 1999 ) and other genes (J. HSIEH, S. KOSTAS and A. FIRE, personal communication). We therefore expressed a composite K04G11.2::gfp cDNA under the control of sel-12 regulatory sequences in a complex array. As shown in Table 1, lin-12(ar170); sel-7(n1253) hermaphrodites display a completely penetrant 2AC phenotype. We crossed arEx434, an extrachromosomal array that gives detectable expression of GFP-tagged SEL-7, into the lin-12(ar170); sel-7(n1253) background and observed that 50/96 hermaphrodites had 1AC. This observation supports the identification of K04G11.2 as sel-7 and suggests furthermore that the SEL-7::GFP protein is functional.

SEL-7 is a nuclear protein:
We analyzed the predicted SEL-7 protein using two programs, SMART (http://smart.embl-heidelberg.de/) and PESTFIND (http://bioweb.pasteur.fr/seqanal/interfaces/pestfind-simple.html) and found that SEL-7 contains two predicted PEST sequences, but no other distinctive amino acid sequence motifs (Fig 1B). BLAST searches identified highly similar sequences in Caenorhabditis briggsae and in three parasitic nematode species, Brugia malayi, Oncocerca volvulus, and Ancylostoma ceylanicum (Fig 1C), but did not identify similar sequences in other organisms.

The carboxy-terminal-tagged SEL-7::GFP fusion protein expressed from arEx434, under the control of sel-12 regulatory sequences, can be visualized in living worms. It is nuclear in the developing gonad (Fig 2) and elsewhere (data not shown). The sequence analysis and subcellular localization of SEL-7 taken together indicate that SEL-7 is a novel nuclear protein.

View larger version (162K): In this window In a new window Download PPT slide   Figure 2. SEL-7 is restricted to the nuclei. Shown here is an L3 larva with arrows pointing to the GFP expression in the gonad. The worm is shown under (A) GFP filter and (B) Nomarski optics.

SEL-7::GFP fusion protein, tagged at the amino-terminus, expressed from sel-7 genomic sequences, is also nuclear (data not shown). Unfortunately, transgenes containing these constructs are poorly expressed and become silenced, so we were not able to assess the function of these proteins or determine a meaningful sel-7 expression profile using them. However, we note that these transgenes did display nuclear GFP in the VPCs; perhaps there is a function of SEL-7 in the VPCs that is not detectable as suppression of the Multivulva phenotype of lin-12(n950) by loss of sel-7 activity.

Analysis of SEL-7 protein-protein interactions with components of the LIN-12(intra) nuclear complex:
The apparent nuclear localization of SEL-7 and its function as a positive regulator of lin-12 activity suggested the possibility that SEL-7 is a component of the nuclear complex that transduces LIN-12/Notch activity. We therefore used the yeast two-hybrid system to explore whether SEL-7 displays physical associations with the following known components of the nuclear complex: LIN-12(intra), LAG-1, SEL-8, and SKIP. Constructs that encode SEL-7 fused to the GAL4 DNA-binding domain or to the GAL4 activation domain were cotransformed into yeast with constructs expressing hybrid proteins containing one of the LIN-12 pathway nuclear components named above or SEL-7 itself. We found that SEL-7 appears to self-associate, but does not display any interaction with other components of the LIN-12 nuclear complex (Table 4).

As SEL-8 can be detected in a complex with LAG-1 and LIN-12(intra) only when both are present (PETCHERSKI and KIMBLE 2000A ), using a three-hybrid assay we assessed whether this was also the case for SEL-7 (MATERIALS AND METHODS). SEL-7 did not show any interaction even when both LAG-1 and LIN-12(intra) were provided (Table 5). We also tested interaction of SEL-7 and SKIP (ZHOU et al. 2000 ), providing either LAG-1 or LIN-12(intra) in the hybrid vector. Again, we did not observe any interaction. Thus, if SEL-7 is a component of the nuclear complex, it may not interact directly with any of these known components.

Identification of potential SEL-7-binding proteins using yeast two-hybrid screening:
We performed a yeast two-hybrid screen for proteins that interact with SEL-7, hoping to find an interaction partner that would illuminate SEL-7 function. After screening 7.5 x 10⁵ transformants, we identified seven positive clones that satisfied our screening criteria (see MATERIALS AND METHODS). One clone corresponded to the predicted gene F13B10.1b, which encodes a protein with two sterile -motif domains; the other six clones corresponded to the predicted gene K08E3.8, which encodes a protein with a prion (asparagine- or glutamine-rich) domain.

We evaluated the function of these two putative interacting proteins using RNAi. First, we fed wild-type hermaphrodites with bacteria expressing dsRNA for each of the two genes, but observed no obvious phenotypes (data not shown). Then, we performed RNAi assays in sensitized backgrounds and looked for effects on the AC/VU decision comparable to those caused by loss of sel-7 activity (Table 6). Neither F13B10.1B nor K08E3.8 suppresses the 0 AC-Egl phenotype of lin-12(n302) or displays a synergistic effect with sel-4. (sel-4, an as-yet-uncloned sel gene, displays a synthetic 2AC phenotype in combination with sel-7; J. CHEN, I. KATIC and I. GREENWALD, unpublished observations). However, K08E3.8 weakly enhanced the 2AC phenotype of lin-12(ar170) to a degree comparable to sel-7 RNAi, suggesting that K08E3.8 may function along with SEL-7 (Table 7).

We also used the yeast two-hybrid system to test whether F13B10.1B and K08E3.8 might interact with LIN-12(intra), LAG-1, SEL-8, and SKIP, the known nuclear complex components. Both F13B10.1B and K08E3.8, like SEL-7, self-associate. The only other interaction observed was between K08E3.8 and SEL-8. Like SEL-8, K08E3.8 is also glutamine rich, so we wondered if, like SEL-8, K08E3.8 would interact with LAG-1 or LIN-12(intra). However, K08E3.8 did not show any interaction when both LAG-1 and LIN-12(intra) were present in a three-hybrid assay.

Given that K08E3.8 interacts with both SEL-7 and SEL-8, but SEL-7 and SEL-8 do not interact, we wondered whether K08E3.8 can serve as a bridge between the two proteins. We therefore performed a three-hybrid assay using these three proteins with SEL-7 fused to the GAL4-DB domain, SEL-8 to GAL4-AD domain, and K08E3.8 in pRST, but did not observe any interaction (data not shown).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have shown that alleles that reduce or eliminate sel-7 activity lead to a reduction in the level of lin-12 activity. These observations suggest that sel-7 is a positive regulator of lin-12 activity. Loss of sel-7 activity can suppress the effects of two different constitutively activated forms of LIN-12, caused by either missense mutations affecting the extracellular domain or expression of the LIN-12 intracellular domain, which mimics the proteolytic cleavage product. This observation is consistent with, but does not prove, sel-7 action downstream of activated LIN-12.

sel-7 encodes a novel nuclear protein. By sequence analysis, there appear to be orthologs in other nematodes, but not in more distant phyla. This observation is reminiscent of SEL-8, which also encodes a novel nuclear protein (DOYLE et al. 2000 ; PETCHERSKI and KIMBLE 2000A ). However, on the basis of the presence of SEL-8 in a ternary complex with LAG-1 and the intracellular domain of GLP-1, a plausible case can be made that SEL-8 performs a function similar to that of Mastermind in Drosophila and mammals (PETCHERSKI and KIMBLE 2000A ; WU et al. 2000 ; KITAGAWA et al. 2001 ). If there is a counterpart to SEL-7 in other organisms, it must be highly diverged in primary sequence.

In the case of sel-8, an integral role in LIN-12/Notch signaling was evident from phenotypic analysis (DOYLE et al. 2000 ). In the case of sel-7, the case is less clear. The effect of sel-7 on lin-12 activity is seen only for a subset of lin-12- or glp-1-mediated cell fate decisions in sensitized genetic backgrounds; a sel-7 null allele does not display any overt phenotypic defects (TAX et al. 1997 and this study). We do not know whether the restricted effect of sel-7 in sensitized backgrounds and the lack of a visible phenotype in the absence of sel-7 activity reflect functional redundancy with another component or complex or a more limited role for sel-7 as an effector of lin-12 activity. Although there is no predicted protein in the C. elegans genome that seems highly similar to sel-7, it may be that functional redundancy is afforded by a protein that is not related in primary sequence.

We speculate that SEL-7 might have a role in the assembly or function of the nuclear complex that includes LAG-1, LIN-12(intra), and SEL-8 (Fig 3). Although we have not detected a direct physical interaction with known components of the nuclear complex, we did identify one protein, K08E3.8, which can interact physically with SEL-7 and SEL-8, and which genetically appears able to influence lin-12 activity in a sensitized background. Although it would be premature to infer a role for SEL-7 or K08E3.8 on this basis, we note that, although our own analyses did not identify clear relatives of K08E3.8 (data not shown), Wormbase Release WS109 (http://www.wormbase.org) lists related human and yeast proteins that have functions in transcriptional regulation. At least five additional sel genes are defined by conventional genetic screens (TAX et al. 1997 ; L. VALLIER, I. KATIC, J. CHEN and I. GREENWALD, unpublished observations) and many more identified by RNAi screens (C. BAIS and I. GREENWALD, unpublished observations). Among these as-yet-uncharacterized sel genes there may be components that link SEL-7 to the nuclear complex or to other aspects of lin-12/Notch signaling.

View larger version (31K): In this window In a new window Download PPT slide   Figure 3. Models for SEL-7 function. (K, K08E3.8). (A) Component of nuclear complex. (B) Facilitated assembly of nuclear complex.

	FOOTNOTES

¹ Present address: Department of Genetics, Harvard Medical School, Boston, MA 02115.

	ACKNOWLEDGMENTS

We thank Iskra Katic and Natalie De Souza for helpful comments on the manuscript. This work was supported by National Institutes of Health grant CA-095389 (to I.G.). I.G. is an Investigator of the Howard Hughes Medical Institute.

Manuscript received September 2, 2003; Accepted for publication October 1, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AILION, M. and J. H. THOMAS, 2000  Dauer formation induced by high temperatures in Caenorhabditis elegans.. Genetics 156:1047-1067.[Abstract/Free Full Text]

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

DE STROOPER, 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.[CrossRef][Medline]

DOYLE, T. G., C. WEN, and I. GREENWALD, 2000  SEL-8, a nuclear protein required for LIN-12 and GLP-1 signaling in Caenorhabditis elegans.. Proc. Natl. Acad. Sci. USA 97:7877-7881.[Abstract/Free Full Text]

FARES, H. and I. GREENWALD, 1999  SEL-5, a serine/threonine kinase that facilitates lin-12 activity in Caenorhabditis elegans. Genetics 153:1641-1654.[Abstract/Free Full Text]

FARES, H. and I. GREENWALD, 2001  Genetic analysis of endocytosis in Caenorhabditis elegans: coelomocyte uptake defective mutants. Genetics 159:133-145.[Abstract/Free Full Text]

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.[CrossRef][Medline]

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

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

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

KARP, X. and I. GREENWALD, 2003  Post-transcriptional regulation of the E/Daughterless ortholog HLH-2, negative feedback, and birth order bias during the AC/VU decision in C. elegans. Genes Dev. 17:3100-3111.[Abstract/Free Full Text]

KITAGAWA, M., T. OYAMA, T. KAWASHIMA, B. YEDVOBNICK, and A. KUMAR et al., 2001  A human protein with sequence similarity to Drosophila Mastermind coordinates the nuclear form of Notch and a CSL protein to build a transcriptional activator complex on target promoters. Mol. Cell. Biol. 21:4337-4346.[Abstract/Free Full Text]

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.[CrossRef][Medline]

LEVITAN, D. and I. GREENWALD, 1998  Effects of SEL-12 presenilin on LIN-12 localization and function in Caenorhabditis elegans.. Development 125:3599-3606.[Abstract]

MELLO, C. and A. FIRE, 1995  DNA transformation. Methods Cell Biol. 48:451-482.[Medline]

PEPPER, A. S., D. J. KILLIAN, and E. J. HUBBARD, 2003  Genetic analysis of Caenorhabditis elegans glp-1 mutants suggests receptor interaction or competition. Genetics 163:115-132.[Abstract/Free Full Text]

PETCHERSKI, A. G. and J. KIMBLE, 2000a  LAG-3 is a putative transcriptional activator in the C. elegans Notch pathway. Nature 405:364-368.[CrossRef][Medline]

PETCHERSKI, A. G. and J. KIMBLE, 2000b  Mastermind is a putative activator for Notch. Curr. Biol. 10:R471-R473.[CrossRef][Medline]

PETTITT, J., W. B. WOOD, and R. H. PLASTERK, 1996  cdh-3, a gene encoding a member of the cadherin superfamily, functions in epithelial cell morphogenesis in Caenorhabditis elegans.. Development 122:4149-4157.[Abstract]

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

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

TAX, F. E., J. H. THOMAS, E. L. FERGUSON, and H. R. HORVITZ, 1997  Identification and characterization of genes that interact with lin-12 in Caenorhabditis elegans.. Genetics 147:1675-1695.[Abstract]

TIMMONS, L. and A. FIRE, 1998  Specific interference by ingested dsRNA. Nature 395:854.[CrossRef][Medline]

TIMMONS, L., D. L. COURT, and A. FIRE, 2001  Ingestion of bacterially expressed dsRNAs can produce specific and potent genetic interference in Caenorhabditis elegans.. Gene 263:103-112.[CrossRef][Medline]

WEINMASTER, G., 2000  Notch signal transduction: a real rip and more. Curr. Opin. Genet. Dev. 10:363-369.[CrossRef][Medline]

WEN, C., M. M. METZSTEIN, and I. GREENWALD, 1997  SUP-17, a Caenorhabditis elegans ADAM protein related to Drosophila KUZBANIAN, and its role in LIN-12/NOTCH signalling. Development 124:4759-4767.[Abstract]

WU, L., J. C. ASTER, S. C. BLACKLOW, R. LAKE, and S. ARTAVANIS-TSAKONAS et al., 2000  MAML1, a human homologue of Drosophila Mastermind, is a transcriptional co-activator for NOTCH receptors. Nat. Genet. 26:484-489.[CrossRef][Medline]

ZHOU, S., M. FUJIMURO, J. J. HSIEH, L. CHEN, and S. D. HAYWARD, 2000  A role for SKIP in EBNA2 activation of CBF1-repressed promoters. J. Virol. 74:1939-1947.[Abstract/Free Full Text]

This article has been cited by other articles:

				M. W. Kankel, G. D. Hurlbut, G. Upadhyay, V. Yajnik, B. Yedvobnick, and S. Artavanis-Tsakonas Investigating the Genetic Circuitry of Mastermind in Drosophila, a Notch Signal Effector Genetics, December 1, 2007; 177(4): 2493 - 2505. [Abstract] [Full Text] [PDF]

				Y. Liu and E. M. Maine The Bro1-Domain Protein, EGO-2, Promotes Notch Signaling in Caenorhabditis elegans Genetics, August 1, 2007; 176(4): 2265 - 2277. [Abstract] [Full Text] [PDF]

				I. Katic, L. G. Vallier, and I. Greenwald New Positive Regulators of lin-12 Activity in Caenorhabditis elegans Include the BRE-5/Brainiac Glycosphingolipid Biosynthesis Enzyme Genetics, December 1, 2005; 171(4): 1605 - 1615. [Abstract] [Full Text] [PDF]