Caenorhabditis elegans (C. elegans) is an ideal model organism to study cell cycle regulation during the development of multicellular organisms, owing to its invariant cell lineage and well defined cell anatomy (Deppe et al., 1978; Sulston and Horvitz, 1977; Sulston et al., 1983). The intestine is an essential organ for multiple biological processes including the distribution of nutrition, immunity, stress response, and aging (An and Blackwell, 2003; Libina et al., 2003; Mallo et al., 2002; Sharrock, 1983). In C. elegans, the intestinal cells originate from a single embryonic blastomere, E, undergoing four rounds of nearly symmetrical cell divisions during the embryogenesis to produce 16 intestinal cells, termed the 16E cells (Deppe et al., 1978; Leung et al., 1999; Sulston et al., 1983). Subsequently, another round of cell divisions occurs in the two anterior and the two posterior intestinal cells. As a result, the intestine consists of 20 cells at the point of hatching (Leung et al., 1999; Sulston et al., 1983). After the hatching, intestinal cells adopt specialized division cycles. At the first (L1) larval stage, some, but not all, intestinal nuclei undergo nuclear divisions without cytokinesis, resulting in binucleated cells (Sulston and Horvitz, 1977), called as binucleation. As a result, worms in the late L1 stage possess 30 to 34 intestinal nuclei in the 20 cells. Thereafter, all the intestinal nuclei undergo endoreduplication before each larval molting to double their DNA contents without nuclear division or cytokinesis (Hedgecock and White, 1985). Because intestinal nuclei endoreduplicate four times up to the four larval stages, they finally contain 32n DNA contents per nucleus at the adult stage. In this way, the intestinal divisions in C. elegans change their cell cycle modes from cell divisions, binucleations to endoreduplications. However, the timely regulation of the transitions of these division modes still remains elusive.

Cdc25 phosphatase promotes cell cycle through dephosphorylation of cyclin-dependent kinases (Cdks) (Fantes, 1979). There are four cdc25 homologs in C. elegans; cdc-25.1, cdc-25.2, cdc-25.3, and cdc-25.4 (Ashcroft et al., 1998). Prolonged CDC-25.1 activity in the cdc-25.1 gain-of-function (gf) mutants is reported and is usually required up to the 16 E cell stage in wild type, caused extra intestinal cell divisions during the embryogenesis (Clucas et al., 2002; Kostić and Roy, 2002). The sequence analysis revealed that these cdc-25.1(gf) mutants contained mutations in the conserved DSG(X)_2+nS motif (Clucas et al., 2002; Kostić and Roy, 2002), which is recognized by a beta-transducin repeats-containing protein (β-TrCP), the substrate recognition subunit of the SCF^β-TrCP E3 ubiquitin ligase, when both of the two serine (S) residues in the motif are phosphorylated (Fuchs et al., 2004). The SCF^β-TrCP E3 ubiquitin ligases play pivotal roles in cell-cycle regulation by targeting critical cellcycle regulators (Fuchs et al., 2004). Genetic studies suggest that the gf mutations in the motif prevented the recognition of mutated CDC-25.1 by LIN-23, a C. elegans β-TrCP ortholog, which permitted the prolonged presence of the CDC-25.1 protein, resulting in intestinal hyperplasia in the cdc-25.1(gf) mutants (Hebeisen and Roy, 2008). In addition, CDC-25.2 is reported to be required for both the intestinal cell divisions after the 16E cell stage during the embryogenesis and binucleations during the L1 larval stage (Lee et al., 2016). We further revealed that the intestinal hyperplasia in lin-23(e1883) mutants was effectively suppressed by RNAi depletion of cdc-25.2 (Lee et al., 2016). This result suggests that LIN-23 targets CDC-25.2 for ubiquitination, and CDC-25.2 activity repressed by LIN-23 is required to change the intestinal cell-cycle mode from binucleation to endoreduplication. These findings indicate that timely repression of CDC-25.1 and CDC-25.2 activities by LIN-23 is essential for proper transitions of division modes during C. elegans intestinal development. Nevertheless, LIN-23, the C. elegans β-TrCP ortholog, indeed recognizes CDC-25.1 and CDC-25.2 for ubiquitination has not yet been investigated. In this study, LIN-23 is demonstrated to exert most significant contribution to the transition from binucleation to endoreduplication among the C. elegans E3 ligase components. Moreover, we show that LIN-23 physically interacts with CDC-25.1 and CDC-25.2, and facilitates their ubiquitination.

Caenorhabditis elegans strains

Nematodes were cultured and handled at 20°C using standard methods (Brenner, 1974). C. elegans strains used in this study are as follows: N2 Bristol and MR156: rrIs01[Pelt-2::GFP; unc-119(+)] X, which were used as wild type. CB3514: lin-23(e1883)/dpy-10(e128) II, YHS47: lin-23(e1883)/dpy-10(e128) II; rrIs01[Pelt-2::GFP + unc-119(+)] X, DH2: zyg-11(b2) II. DS97: mat-2(ax76) II, EU593: mel-26(or184) I, EU640: cul-2(or209) III, UP63: mat-3(ku233) III, NJ582: cul-1(e1756)/unc-69(e587) III, VC141: zif-1(gk117) III, VC370: rfp-1(ok572)/eT1 III; +/eT1 V, VC562: rbx-1(ok782) V/nT1[qls51] (IV;V), VC684: rbpl-1(ok907) I/hT2[bli-4(e937)let-?(q782)gls48] (I;III), VC954: rnf-113(ok1401) III/hT2[bli-4(e937) let-?(q782)qls48] (I;III), VC1241: skr-1(ok1696) I, VC1439: skr-2(ok1938) I/hT2[bli-4(e937) let-?(q782)qls48] (I;III), YHS171: rfp-1(ok572) III/hT2[bli-4(e937)let-?(q782)gls48] (I;III); rrIs01[Pelt-2::GFP + unc-119(+)] X, YHS161: rbx-1(ok782) V/nT1[qls51] (IV;V); rrIs01[Pelt-2::GFP + unc-119(+)] X, and YHS163: skr-1(ok1696) I; rrIs01[Pelt-2::GFP + unc-119(+)] X.

Quantification of intestinal nuclei and microscopy

To examine the number of intestinal nuclei marked by GFP using an elt-2::GFP transgene, worms were transferred to 0.2 mM tetramisole in M9 buffer on a poly-L-lysine-coated slide glass, covered with a coverslip and observed using a fluorescence microscope (Zeiss Axioskop 2, Carl Zeiss, Germany). Otherwise, DNA of some strains, whose intestine is not marked with intestinal GFP, was stained with Hoechst 33342 solution (40 mM NaCl, 10 mM Tris-HCl (pH7.5), 1 mM EDTA, 20% Glycerol, 20 μg/ml Hoechst 33342). The worms were placed on a water drop on a poly-L-lysine-coated slide glass and fixed by quickly dehydrating on an alcohol lamp. The fixed worms were covered with a coverslip containing Hoechst 33342 solution and gelutol. The samples prepared on a slide glass were observed using a fluorescence microscope (Zeiss Axioskop 2, Carl Zeiss, Germany). Images were taken using an Orca-ERG digital camera (Hamamatsu, Japan) and NIS-elements software (Nikon, Japan).

RNA interference

RNAi depletion of cdc-25.2 in mutants of E3 ubiquitin-ligase complex genes was performed by the soaking RNAi method as previously described (Lee et al., 2016). The cdc-25.2 dsRNA was prepared by in vitro transcription using a cdc-25.2 cDNA clone, yk472b2, as the DNA template. Worms synchronized at the first larval (L1) stage were soaked into the dsRNA solution and incubated for 48 h at 20°C. Then the soaked animals were transferred to OP50-seeded NGM plates and allowed to grow until the young adult stage. The efficiency of RNAi treatment was confirmed by the expression germ-line proliferation defective (Glp) phenotype, which is one of the characteristic cdc-25.2 RNAi phenotypes, and when confirmed, the number of intestinal nuclei was counted to determine the possible suppression of the mutant intestinal phenotype by cdc-25.2 RNAi.

Cell culture

HeLa cells were maintained in Dulbecco’s Modified Eagle’s Medium (Welgene, Korea), supplemented with 10% fetal bovine serum (Welgene), 100 U/ml penicillin (Welgene), and 100 μg/mL streptomycin (Welgene) in a humidified atmosphere with 5% CO₂ at 37°C.

Plasmid construction and transfection

To generate Myc-tagged CDC-25.1 and CDC-25.2 fusion proteins at their N-terminal, cDNA sequences of cdc-25.1 and cdc-25.2 were cloned into the pCMV-Myc vector (Clontech, USA). The cDNAs of cdc-25.1 and cdc-25.2 were PCR-amplified from a C. elegans mixed-stage cDNA library by using the following primer pairs: 5′-ggtacc ATG GCT ACC ACC GGG GAA AAA G-3′ and 5′-gcggccgc TTA TTC GGC GTC GTC AGA AAT C-3′ for cdc-25.1 cDNA; 5′-gg ggtacc ATG AAT AGG CCT TCA CAA ATA TC-3′ and 5′-attt gcggccgc TTA TGA AAA ACG AGG ACA TGT C-3′ for cdc-25.2 cDNA. KpnI and NotI cleavage sites underlined in the primer sequences were used for cloning. To express GFP-tagged LIN-23 in mammalian cells, lin-23 cDNA sequence was PCR-amplified from a C. elegans mixed-stage cDNA library using the following primer pair: 5′-gg ggtacc ATG TCT TCA CCG CAC CG-3′ and 5′-tcc cccggg TTA TGG GCC ACC ATC TG-3′. Using the restriction sites of KpnI and SmaI underlined in the sequences, cDNA of lin-23 was ligated into the pEGFP-C1 plasmid DNA (Clontech). The DNA sequence of the cloned plasmid construct was confirmed by sequencing (Macrogen, Korea). To construct FLAG-tagged LIN-23 plasmid, cDNA sequence of lin-23 was cloned into the pCMV-Myc vector (Clontech). Using Myc-tagged LIN-23 plasmid as a template, inverse PCR was performed with primers: 5′-ATG GAC TAC AAG GAC GAC GAT GAC AAG GGT ACC ATG TCT TCA CCG CAC-3′ and 5′-GT CCT TGT AGT CCA TGG TGG GCC CGC GGG TAC-3′. The Myc sequence in the Myc-tagged LIN-23 plasmid was replaced with FLAG sequence using In-Fusion HD cloning kit (Clontech) to generate FLAG-tagged LIN-23 plasmid. The HA-tagged ubiquitin under the control of the CMV promoter was generated as previously described (Yoo et al., 2014). HeLa cells were transfected with 1 μg of each plasmid DNA using jetPRIME reagent (polyplus, France), following the manufacturer protocol. Same amount of empty vector was transfected to negative control samples as a loading control of the transfection. At 48 h after the transfection, the cells were harvested for further analyses.

Co-immunoprecipitation and Western blot analysis

Cell pellets were resuspended in NP-40 lysis buffer (150 mM NaCl, 1.0% NP-40, 50 mM Tris-Cl, pH 8.0) containing protease inhibitor mixture (Roche Applied Bioscience, Germany) and incubated for 30 min at 4°C, and sonicated (CosmoBio, Japan). Then, 650 μg of cell lysates were incubated with protein G-sepharose (Sigma, USA) at 4°C overnight, and 2 μg of monoclonal anti-c-Myc antibody (9E10, Santa Cruz, USA) was added to the samples. The samples were incubated for 4 h at 4°C with gentle rotation. After washing with NP-40 lysis buffer 4 times, samples were boiled in sodium dodecyl sulfate (SDS) loading buffer for 10 min. Boiled cell lysate proteins were separated by 7–10% SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Millipore, USA) using a semidry transfer apparatus (Hoefer, USA). The membrane was blocked with 5% non-fat dried milk/2% BSA in TBS-T (150 mM NaCl, 50 mM Tris-Cl, 0.1% Tween 20) for 30 min at RT. The blots on the membrane were first incubated with the following primary antibodies at 4°C overnight: anti-HA antibody (1:1,000, Y-11, Santa Cruz); anti-Myc antibody (1:1,000, 9E10, Santa Cruz); anti-GFP antibody (1:1,000, A11122, Invitrogen, USA); monoclonal anti-FLAG M2 antibody (1:500, F1804, Sigma); monoclonal anti-α-Tubulin antibody (1:5,000, T9026, Sigma). The blots on the membrane were then incubated with the secondary antibody, horseradish peroxidase-conjugated anti-mouse IgG (1:5,000, Santa Cruz) or horseradish peroxidase-conjugated anti-rabbit IgG (1:5,000, Santa Cruz), at RT for 2 h. Finally, immunoreactive blots were detected as chemiluminescence signals by using ECL reagent (Amersham® Biotechnology, UK) and a Bio-imaging Analyzer LAS-4000 (Fuji-film, Japan).

In vivo ubiquitination assay

HeLa cells were co-transfected with plasmid DNAs expressing HA-tagged ubiquitin, Myc-tagged CDC-25.1 or CDC-25.2, and GFP-tagged LIN-23 for 48 h and then treated with 25 μM of MG132 (Sigma) for 6 h to inhibit the activity of proteasome and then harvested. The harvested cell pellets were then lysed in the NP-40 lysis buffer with protease inhibitors as previously described. Myc-tagged CDC-25.1 or Myc-tagged CDC-25.2 was immunoprecipitated with anti-Myc antibody (9E10, Santa Cruz) and protein G sepharose beads (Sigma). The immunoprecipitates were washed four times with NP-40 lysis buffer and boiled in SDS loading buffer for 10 min. The immunoprecipitated and boiled samples were then analyzed by Western blot analysis.

Statistical analysis

The bar graphs are shown as the average ± S.D. Box-and-whisker plots are displayed using Tukey whiskers, extending up to data points 1.5 interquartile ranges away from the first and the third quartile. P value was calculated by Student’s t-test to identify statistically significant difference. A P-value of < 0.05 was considered significant. Most data were obtained from more than three independent experiments with multiple sample numbers (n), indicated in the figure legends.

Extra intestinal nuclear divisions were observed in several mutants of E3 ubiquitin-ligase component genes, but the highest number of intestinal nuclei was observed in lin-23 mutants

In a previous genome-wide RNAi screen of ubiquitination-related genes, RNAi depletion of 17 genes induced excess intestinal cell divisions during the embryogenesis (Du et al., 2015). Among the 17 genes, lin-23 was included. We previously reported that intestinal hyperplasia in lin-23(e1883) mutants was effectively suppressed by the RNAi depletion of cdc-25.2 (Lee et al., 2016). We hypothesized that if the activity of CDC- 25.2 is negatively regulated not only by LIN-23 but also by other E3 ubiquitin ligase components during the postembryonic intestinal development, loss-of-function (lf) mutants of these genes would display excess intestinal nuclear division phenotype during the larval development, and further this excess intestinal nuclear division phenotype would be suppressed by RNAi depletion of cdc-25.2 as seen in lin-23(e1883) mutants. Therefore, first we examined the number of intestinal nuclei in the lf mutants of the ubiquitination-related genes at their adult stage (Fig. 1). Among the 17 genes, RNAi depletion previously reported to show excess intestinal cells (Du et al., 2015), 14 genes were tested, because their lf mutant strains were available from the Caenorhabditis Genetic Center (Table. 1 and Fig. 1A). Among them, for cul-1(e1756), skr-2(ok1938), rbx-1(ok782), rnf-113(ok1401), and rbpl-1(ok907), their heterozygous mutant animals were examined instead of homozygous mutant ones (Fig. 1A), because their homozygous animals could not develop to the adult stage. They showed a normal number of intestinal nuclei compared to the wild type control, suggesting that these genes are haplosufficient in the intestinal development (Fig. 1A). Among the 14 ubiquitination-related genes, lin-23(e1883), cul-2(or209), mat-2(ax76), mat-3(ku233), and rfp-1(ok572) homozygous lf mutant animals showed statistically significant increase in the number of intestinal nuclei compared to the wild type (Fig. 1A). For example, immunofluorescent microscopic images of GFP-marked intestinal nuclei in wild-type, lin-23(e1883), skr-1(ok1696), and rfp-1(ok572) mutant adult animals are shown in Fig. 1B. Notably, although statistically significant, the increase in the number of intestinal nuclei in the cul-2, mat-2, mat-3, and rfp-1 lf mutants was much less than those in lin-23(e1883) mutants. The four mutants contained only 1 to 5 more intestinal nuclei than the wild type, whereas lin-23(e1883) mutants contained around 17 more intestinal nuclei than the wild type (Fig. 1A). Therefore, LIN-23 was considered as the major E3 ubiquitin ligase component that represses the CDC-25.2 activity during the postembryonic intestinal development.

RNAi depletion of cdc-25.2 suppressed extra intestinal nuclear divisions in mutants of E3 ubiquitin-ligase component genes

In lin-23(e1883) mutants, additional intestinal nuclear divisions occur in the place of endoreduplication at the L1 to L2 molting (Kipreos et al., 2000). Furthermore, this intestinal hyperplasia in the lin-23 mutants was almost completely suppressed by the RNAi depletion of cdc-25.2 (Lee et al., 2016). Because intestinal hyperplasia was also observed in lf mutants of several E3 ubiquitin-ligase component genes besides the lin-23 mutants (Fig. 1), we hypothesized that these E3 ubiquitin-ligase components may also repress the CDC-25.2 activity to shift the intestinal division mode from nuclear division to endoreduplication. In such case, intestinal hyperplasia phenotype in these lf mutants would be suppressed by the RNAi depletion of cdc-25.2, as observed for the lin-23 mutants. To test this possibility, we treated RNAi depletion of cdc-25.2 to the lf mutants of cul-2, mat-2, mat-3, and rfp-1 to investigate whether their intestinal hyperplasia phenotype was suppressed (Fig. 2). cdc-25.2 RNAi indeed found to suppress the intestinal hyperplasia phenotype of the mutants of cul-2, mat-2, mat-3, and rfp-1 at statistically significant levels (Figs. 2A and 2B). However, their levels of suppression (p < 0.05) were statistically less significant than the level of suppression observed in lin-23(e1883) mutants (p < 0.00005). These results support our view that LIN-23 is the major E3 ubiquitin ligase component that represses the CDC-25.2 activity during the postembryonic intestinal development, although other E3 ubiquitin ligase components may also participate in this repression.

CDC-25.2 and CDC-25.1 interact with LIN-23

To examine whether LIN-23, a C. elegans β-TrCP ortholog, physically interacts with CDC-25.2, an in vivo mammalian cell culture system was used. HeLa cells were co-transfected with Myc-tagged CDC-25.2 and GFP-tagged LIN-23 (Fig. 3A), or with Myc-tagged CDC-25.2 and FLAG-tagged LIN-23 (Fig. 3B). Then, immunoprecipitation was performed using anti-Myc antibody. The immunoprecipitates were analyzed by western immunoblotting with either anti-GFP antibody (Fig. 3A) or anti-FLAG antibody (Fig. 3B) to investigate the co-immunoprecipitation of GFP or FLAG-tagged LIN-23 with Myc-CDC-25.2 proteins. Previous genetic studies in cdc-25.1(gf) mutants suggested that a mutation of β-TrCP motif causes intestinal hyperplasia, and the possibility that LIN-23 targets the CDC-25.1 by recognizing the β-TrCP concensus sequence was proposed (Hebeisen and Roy, 2008). However, there is no molecular evidence for the interaction between CDC-25.1 and LIN-23. Therefore, in this study, the interaction between CDC25.1 and LIN-23 was also analyzed. GFP-tagged LIN-23 was also co-immunoprecipitated with Myc-CDC-25.1 (Fig. 3C). The immunoprecipitates were also analyzed with anti-Myc antibody to check the efficiency of immunoprecipitation. In both the CDC-25.1 and CDC-25.2 immunoprecipitates, GFP-LIN-23 protein band of ~100 kDa in size (Figs. 3A and 3C), or FLAG-LIN-23 protein band of ~80 kDa in size (Fig. 3B) was successfully detected (Figs. 3A–3C; the rightmost lanes). Input cell lysates were also examined, confirming the expression of respective recombinant proteins. These results indicate that, although CDC-25.2 does not contain a β-TrCP recognition consensus motif (DSG(X)_2+nS; see Discussion), both CDC-25.1 and CDC-25.2 physically interacted with LIN-23 in the mammalian cell culture system.

LIN-23 facilitated ubiquitination of CDC-25.1 and CDC-25.2

To examine whether the observed physical interaction between LIN-23 and CDC-25.1 and that between LIN-23 and CDC-25.2 actually increased the ubiquitination of CDC-25.1 and CDC-25.2, respectively, an in vivo ubiquitination assay in HeLa cells was performed (Fig. 4). HeLa cells were transfected with several plasmid DNAs to express HA-tagged ubiquitin (Ub), Myc-tagged CDC-25.1 or Myc-tagged CDC-25.2, and GFP-tagged LIN-23. Then, the transfected HeLa cells were treated with, a proteasome inhibitor, MG132, to avoid the degradation of polyubiquitinated proteins (see “Materials and Methods” for details). Further, cell lysates prepared from the transfected cells were immunoprecipitated with anti-Myc antibody. Finally, the anti-Myc immunoprecipitates were western immunoblotted with anti-HA antibody to evaluate the levels of ubiquitination of respective immunoprecipitates. HA-positive ladder-like bands of various sizes were detected, which are indicative of polyubiquitination of proteins, in all the anti-Myc immunoprecipitates (Figs. 4A and 4C). The intensity of HA-positive ubiquitination signals was significantly stronger when GFP-LIN-23 was co-expressed in the cells compared to the cells, in which GFP-LIN-23 was not co-expressed (Figs. 4A and 4C). In the presence of GFP-LIN-23, the relative intensities of HA-positive ubiquitination signals, which were normalized with those of Myc-positive signals, were about three times stronger in the Myc-CDC-25.1 immunoprecipitates, and about twice stronger in the Myc-CDC-25.2 immunoprecipitates, than in the absence of GFP-LIN-23, in which GFP was co-expressed as a mock (Figs. 4B and 4D). These results indicate that LIN-23 facilitated ubiquitination of CDC-25.1 and CDC-25.2 in the mammalian cell culture system.

C. elegans utilizes three different cell-cycle modes for its intestinal development. First, normal cell divisions occur during the embryogenesis to generate 20 intestinal cells (Deppe et al., 1978; Leung et al., 1999; Sulston et al., 1983). Second, binucleations, the nuclear divisions without cytokinesis, occur during the L1 larval stage to produce 30–34 intestinal nuclei in the 20 cells (Sulston and Horvitz, 1977). Finally, endoreduplications, the DNA replications without nuclear divisions, occur in each intestinal nucleus at respective four larval moltings to increase their DNA contents. Previous genetic studies indicated that CDC-25.1 is specifically required for embryonic intestinal cell divisions up to the 16E cell stage, and then CDC-25.2 takes over the role of CDC-25.1 up to the stage of binucleations at the L1 larval stage, resulting in 30–34 nuclei in 20 intestinal cells (Lee et al., 2016). Furthermore, LIN-23, the C. elegans β-TrCP ortholog, was genetically identified as a repressor of CDC-25.1 and CDC-25.2 to prevent excess embryonic intestinal cell divisions and excess larval intestinal nuclear divisions, respectively (Hebeisen and Roy, 2008; Lee et al., 2016). Nevertheless, whether LIN-23 actually targets CDC-25.1 and CDC-25.2 for ubiquitination as a component of E3 ubiquitin ligase has not been investigated previously. In this study, by using an in vivo mammalian cell culture co-transfection system, for the first time, we demonstrated that LIN-23 physically interacts with CDC-25.1 and CDC-25.2, and facilitates their ubiquitination.

LIN-23 is a C. elegans ortholog of β-TrCP, a class of F-box protein, functioning as a substrate-recognizing subunit in the SCF (Skp1-Cullin-F-box protein) E3 ubiquitin ligase complex (Kipreos et al., 2000). In the complex, Cullin offers the scaffold to the complex and Skp1 serves as an adaptor protein that links Cullin and an F-box protein. β-TrCP interacts with a substrate that is phosphorylated within the DSG(X)_2+nS destruction motif (Maniatis, 1999). The SCF^β-TrCP E3 ubiquitin ligase mediates the ubiquitination and proteasomal degradation of the phosphorylated substrates that play a key role in the signal transduction or cell-cycle regulation (Fuchs et al., 2004). Previous studies revealed that cdc-25.1(gf) mutants contained mutations in the conserved DSG(X)_2+nS destruction motif (Clucas et al., 2002; Kostić and Roy, 2002). These findings strongly support the idea that, in wild type, CDC-25.1 protein is indeed targeted by LIN- 23-containing E3 ubiquitin ligase for ubiquitination and proteasomal degradation. In contrast, CDC-25.2 protein does not contain an amino-acid sequence that is completely matching to the DSG(X)_2+nS destruction motif sequence. Nevertheless, in this study, we found that GFP-LIN-23 or FLAG-LIN-23 was specifically co-immunoprecipitated with Myc-CDC-25.2 as with Myc-CDC-25.1, and that the ubiquitination of Myc-CDC-25.2 significantly increased in the presence of GFP-LIN-23 as that of Myc-CDC-25.1 in an in vivo mammalian cell culture system. Therefore, CDC-25.2 was assumed to be recognized by LIN-23 either using a sequence that does not completely match to the destruction motif or by an unidentified mechanism.

A previous bioinformatics study revealed 866 potential ubiquitination-related genes in C. elegans genome, including 836 E3 ubiquitin ligase component genes (Du et al., 2015). Through a systematic RNAi screening covering 80% of these potential ubiquitination-related genes, the above study identified 17 genes, RNAi of which induced excess intestinal cell divisions during the embryogenesis (Du et al., 2015). This finding suggests that these 17 gene products potentially function as repressors of embryonic intestinal cell divisions by targeting a critical cell-cycle regulator such as CDC-25.1 for ubiquitination. Indeed, lin-23 was included among the 17 genes (Du et al., 2015). In this study, to identify possible counterpart components that function together with LIN-23 in the same E3 ubiquitin ligase complex, or to find additional substrate-recognizing subunits that target CDC-25.2, 14 genes out of above 17 genes were tested to see whether viable lf mutants of them also show excess intestinal nuclear divisions during the larval development. Among the Cullin and Skp1 ortholog genes, cul-2 homozygous mutants showed increased numbers of intestinal nuclei. Nevertheless, the level of increase was not so significant compared to that in lin-23 homozygous mutants. In contrast, cul-1 and skr-2 homozygous mutants could not be examined, because they could not develop to the adult stage. Furthermore, their heterozygous mutants did not show any significant difference in the number of intestinal nuclei from wild-type control. Therefore, counterpart components that likely function together with LIN-23 in the same E3 ubiquitin ligase complex could not be identified. Among the ortholog genes that encode components of different types of E3 ubiquitin ligase, mat-2, mat-3, and rfp-1 mutants showed increased numbers of intestinal nuclei. mat-2 and mat-3 encode the components of Anaphase-Promoting Complex (APC/C), the E3 ubiquitin ligase that has a critical role in cell-cycle regulation (Golden et al., 2000). rfp-1 encodes a component of RING-type E3 ligase (Crowe and Candido, 2004). Nevertheless, in these three mutants, the levels of increase in the number of intestinal nuclei and the levels of suppression of the number of intestinal nuclei by cdc-25.2 RNAi were also not as significant as those in lin-23 mutants. Therefore, we concluded that LIN-23 is the major E3 ubiquitin ligase component that recognizes both CDC-25.1 and CDC-25.2 to repress their activities for proper transition of cell-cycle modes during the C. elegans intestinal development, although other E3 ubiquitin ligase components may also participate in this regulation.