Development 127, 5071-5082 (2000) 5071

Printed in Great Britain © The Company of Biologists Limited 2000

DEV5405

The C. elegans F-box/WD-repeat protein LIN-23 functions to limit cell division

during development

Edward T. Kipreos1,*, Sonya P. Gohel2 and Edward M. Hedgecock2

1Department of Cellular Biology, The University of Georgia, Athens, Georgia 30602, USA
2Department of Biology, The Johns Hopkins University, Baltimore, Maryland 21218, USA
*Author for correspondence (e-mail: [email protected])

Accepted 11 September; published on WWW 2 November 2000

SUMMARY

In multicellular eukaryotes, a complex program of that is orthologous to the Saccharomyces cerevisiae gene
developmental signals regulates cell growth and division by MET30, the Drosophila melanogaster gene slmb and the
controlling the synthesis, activation and degradation of G1 human gene βTRCP, all of which function as components
cell cycle regulators. Here we describe the lin-23 gene of of SCF ubiquitin-ligase complexes. Loss of function of
Caenorhabditis elegans, which is required to restrain cell the Drosophila slmb gene causes the growth of ectopic
proliferation in response to developmental cues. In lin-23 appendages in a non-cell autonomous manner. In contrast,
null mutants, all postembryonic blast cells undergo lin-23 functions cell autonomously to negatively regulate cell
extra divisions, creating supernumerary cells that can cycle progression, thereby allowing cell cycle exit in
differentiate and function normally. In contrast to the response to developmental signals.
inability to regulate the extent of blast cell division in lin-23
mutants, the timing of initial cell cycle entry of blast cells is Key words: Ubiquitin proteolytic pathway, SCF, Hyperplasia, G1 to
not affected. lin-23 encodes an F-box/WD-repeat protein G0 transition, cul-1, Caenorhabditis elegans

INTRODUCTION methods are contributing to our understanding of the role of

cell cycle regulators in higher eukaryotes, most notably
Development of multicellular animals requires the careful in Drosophila melanogaster (Edgar and Lehner, 1996).
interplay of cell division, differentiation, and morphogenesis. In Caenorhabditis elegans, with the advantage of invariant, visible
all animals, cell division is regulated to a large extent by the cell lineages that allow cell division to be followed at single cell
action of cyclin-dependent kinases (CDKs) and their associated resolution, has only recently begun to be widely used as a model
cyclin subunits, which are required for CDK activation (Pines, in the study of cell cycle regulation. In most cases, the cell cycle
1995; Edgar and Lehner, 1996). CDK/cyclin complexes are regulators identified in Drosophila and C. elegans have had
regulated by three general mechanisms (Pines, 1995). First, yeast counterparts that also function in cell cycle regulation,
transcriptional regulation of cyclins and CDK/cyclin inhibitory although the cell cycle function of the multicellular eukaryotic
molecules (cyclin-dependent kinase inhibitors or CKIs). homolog has generally evolved to fit a developmental context.
Second, ubiquitin-mediated degradation of cell cycle regulators, Recessive mutations that affect the proliferation of multiple
in particular cyclins and CKIs. And third, activation and tissues often identify genes that regulate the cell cycle machinery
inactivation of the CDK subunit by phosphorylation. directly, while tissue-specific proliferation phenotypes are often
The basic framework of cell cycle regulation, as well as the indicative of genes regulating tissue-specific signal transduction
functions of many cell cycle regulators, has been conserved in pathways. We have isolated a gene, lin-23, that is required for
all eukaryotes. Nevertheless, there are significant differences limiting the cell proliferation of all tissues during development
between cell cycle regulation in animals and yeast. In particular, of C. elegans. lin-23 shares homology with the MET30 gene of
for a given cell cycle regulator in yeast there is often an budding yeast, which has been shown to function in the
increased number of homologs in animals, presumably to allow ubiquitin-mediated degradation of the CDK-inactivating kinase
greater flexibility in regulating cell division with developmental Swe1 and to indirectly regulate G1 cyclin mRNA stability
or tissue-specific events (Pines, 1995; Liu and Kipreos, 2000). (Kaiser et al., 1998; Patton et al., 2000). Met30p functions as
Genetic methods have been instrumental in the identification an F-box component of an SCF ubiquitin-ligase (E3) complex
of genes involved in cell cycle regulation. The majority of cell (Patton et al., 1998; Kaiser et al., 1998). SCF complexes contain
cycle regulators have been identified through genetic screens Skp1, a Cullin, Rbx1, and an F-box containing protein that
in budding and fission yeast with metazoan homologs directly binds the substrate (see Deshaies, 1999).
subsequently studied. To an increasing extent, forward genetic In higher eukaryotes, lin-23 orthologs have been identified in
5072 E. T. Kipreos, S. P. Gohel and E. M. Hedgecock

humans, Xenopus laevis and Drosophila (Margottin et al., 1998; likelihood and parsimony methods were used to create phylogenies.
Spevak et al., 1993; Jiang and Struhl, 1998; Theodosiou et al., For neighbor-joining analysis, a distance matrix, created with the
1998). However, none of these genes has yet been shown to ProtML program of MOLPHY version 2.3 using the JTT model
function in cell cycle regulation. The lin-23 mutant phenotype (Adachi and Hasegawa, 1996), was used to create a tree with the
therefore reveals a new function for this class of metazoan F- NEIGHBOR program of PHYLIP version 3.573c (Felsenstein, 1993).
Maximum likelihood analysis was performed with the ProtML
box genes, allowing cell cycle exit in response to developmental program of MOLPHY version 2.3 (Adachi and Hasegawa, 1996),
cues. using the exhaustive search method with the JTT model. Parsimony
analysis was performed with PAUP* version 4.b2 (Swofford, 1999)
using the exhaustive search method.
MATERIALS AND METHODS Bootstrap support values derive from: parsimony bootstrap analysis
with 1000 replicates using default Branch and Bound settings;
Genetic analysis maximum likelihood RELL bootstrap analysis with 10,000 replicates;
Alleles e1883 and e1925 were isolated by clonally screening EMS- and neighbor-joining bootstrap analysis with 1000 replicates.
treated hermaphrodites for self-progeny with extra postembryonic
cells. Alleles rh293 and rh294 were isolated in a clonal Antisense RNA analysis
non-complementation screen of ethylmethanesulfonate (EMS)- Antisense lin-23 RNA was synthesized from the 1.1 kb 3′ cDNA using
mutagenized dpy-10(e128)/let-22(mn22). 3200 F1 animals were the Ambion MEGAScript kit. RNA was diluted to 0.75 mg/ml with
screened. F1 clones that failed to segregate Dpy progeny were tested diethylpyrocarbonate-treated water, and injected into both ovaries of
for non-complementation to lin-23(e1883). Alleles e1521, oz107, wild-type adult hermaphrodites (Mello and Fire, 1995). Embryo cell
m731, and rh194 were gifts from J. Hodgkin, T. Schedl, M. L. Edgley numbers were obtained by counting nuclei of squashed embryos
and J. D. Plenefisch, respectively. lin-23 allele e1883 was mapped by stained with DAPI, as described by Kipreos et al. (1996).
3-factor cross between lin-4(e912) and dpy-10(e128). A total of
17 recombinants from lin-23(e1883)/lin-4(e912) dpy-10(e128) Molecular analysis
heterozygotes gave the relative distances: lin-4 (10/17) lin-23 (7/17) Genomic DNA containing lin-23 was identified by complementation
dpy-10. The deficiency mnDf30 failed to complement lin-23; and lin- of the mutant phenotype. To assay for lin-23(+) activity, 50 µg/ml of
23(e1883)/mnDf30 heterozygotes had a phenotype indistinguishable cosmids or cosmid subclones along with 150 µg/ml plasmid PRF4,
from lin-23(e1883) homozygotes. Analysis of cell numbers in embryos carrying the dominant marker rol-6 (su1006) allele (Mello and Fire,
and larvae was as described by Kipreos et al. (1996). 1995), were microinjected into the distal gonad arms of lin-
Double homozygous mutants of lin-23 and sel-10 were obtained from 23(e1883) dpy-10(e128)/lin-4(e912) using published methods (Mello
the heterozygous lin-23 and homozygous sel-10 strain lin-23(e1883)/+; and Fire, 1995). Transformants carrying the injected DNA as stable
sel-10(ar41) lon-3(e2175) and had both Lin and Lon phenotypes. A extrachromosomal arrays were cloned and their progeny were scored
clonal analysis of progeny from lin-23(e1883)/dpy-10(e128); cul- for Dpy non-Lin-23 individuals. Rare recombinants with this same
1(e1756)/unc-69(e587) double heterozygotes was performed with 327 phenotype were excluded by testing for Dpy Lin-4 self progeny.
embryos. The progeny were wild type (77 observed/82 expected); Unc Cosmids T25H2, F58F12, and a 10.7 kb subclone of T25H2,
(43/41); Dpy (45/41); Lin – both cul-1 and lin-23 homozygotes (85/82); T∆ScaI, that extends from an internal ScaI site to the end of the
Dpy Unc (19/20); Dpy Lin (19/20); Unc Lin (20/20); and the remainder genomic insert, were each found to rescue lin-23(e1883) (Fig. 3).
(19/20), inferred to be Lin Cul, were 5 arrested embryos (pre-comma Removing a 1.5 kb ScaI-StuI fragment from the end of T∆ScaI, to
stage) and 14 L1 and L2 arrested animals with a range of cell division produce T∆StuI, severely reduced lin-23(+) activity; while removal of
from no postembryonic divisions to hyperplasia. a further 1.5 kb StuI-BamHI fragment abolished lin-23(+) activity (Fig.
Mosaic analysis was performed on lin-23(oz107), ncl-1(e1865) 3). The 9.2 kb T∆StuI subclone was used as a hybridization probe to
homozygotes containing the array ekEx11. The strain was created from isolate an incomplete, 1.1 kb lin-23 cDNA clone from a phage library
lin-23(oz107)/mnC1, ncl-1(e1865) by injection of a mixture of 20
ng/µl cosmid T25H2 (containing lin-23(+)), 20 ng/µl cosmid C33C3
(containing ncl-1(+); Miller and Shakes, 1995), 40 ng/µl plasmid Fig. 1. Postembryonic cell lineages of lin-23(e1883) hermaphrodites.
pBK48 (containing GFP driven by the let-858 promoter; Kelly et al., Larval stages are indicated on the left. Wild-type lineages, shown for
1997) and 120 ng/µl of N2 wild-type genomic DNA cut with SmaI. A comparison, are from Sulston and Horvitz (1977), and Kimble and
homozygous lin-23(oz107), ncl-1(e1865) line, ET84, was obtained. Hirsh (1979). (A) Lineage of the lateral H and V hypodermoblasts in
L4-stage ET84 animals were examined by epifluorescence microscopy, a single lin-23 hermaphrodite. Syncytial hypodermis (H), seam (S),
with GFP epifluorescence indicating the presence of the lin-23(+) neuron/neuroglia (N), and apoptosis (X). (B) Lineage of the
transgenic extrachromosomal array. The point in the lineage of array neuroblast QL in three lin-23 hermaphrodites. Arrows indicate
loss was deduced from the pattern of fluorescing vulval cells, which anteriorly (leftward) or posteriorly (rightward) directed cell
maintained their normal positions in the vulva. Confirmatory movements. (C) Lineage of the ventral hypodermoblasts P9 and P10
information was obtained by scoring for the Ncl (large nucleoli) in three lin-23 hermaphrodites (a single individual lin-23
phenotype of the neuronal P5.a, P6.a and P7.a descendants for those hermaphrodite was followed for the P9 lineage and leftmost P10
cases in which the array was lost prior to the creation of the vulval lineage). Y-shaped lineage branches indicate a failed cytokinesis
precursor cells (VPCs) (in these neuronal cells the let-858 promoter followed by nuclear refusion. (D) Lineage of vulval hypodermoblasts
was not active). in a single lin-23 hermaphrodite. The position of the uterine anchor
Quantitation of intestine cell ploidy was performed by serial cell is shown for reference. Origins of the vulval precursors were not
confocal sectioning of propidium iodide-stained adults and L2 stage determined in this individual. (E) Lineage of the mesoblasts Z1 and
larvae using hypodermal and neuronal cells as internal 4C and 2C Z4 in two lin-23 hermaphrodites. Distal-tip cell (DTC). (F) Lineage
standards, respectively, as described by Feng et al. (1999). of intestinal cells in a single lin-23 hermaphrodite. Endoreplications
(S) and nuclear divisions are shown for wild type (Hedgecock and
Phylogenetic analysis White, 1985) and lin-23(e1883). Intestinal nuclei in the int2R and
Initial protein sequence alignments were made with the CLUSTAL X int9L cells were observed to fuse together in late mitosis in this
program (Thompson et al., 1994). The alignment was then optimized lineage; this fusion is also observed occasionally in wild-type
by hand to minimize insertions/deletions. Neighbor-joining, maximum intestinal cells (data not shown).
 lin-23 negatively regulates cell division 5073

A wild type
hrs H1 H2 V1 V2 V3 V4 V5 V6
0

H H H H H H
H
10 NH

L1
S H
H H H H H H H H H
20 H H
X
L2
H H H H H H H H H H H H H
30

L3
HS HS HS HS HS HS HS HS HS HS HS HS HS

lin-23
hrs H1 H2 V1 V2 V3 V4 V5 V6
0

H H H H H H
H H
10 HH

L1

20 H
S H N
H H H H HH H H
HH H H NN NN
H H H H H
H H
L2 HH HH
NNX
30 H HH H
H H H H H H
H H H H H
H H H
H H H H H
H

40
L3
H H
H H H H H H
H H H H H H
S S S S S S S S HS S H SS S S S S S S S S

wild type lin-23 lin-23 lin-23 C wild type lin-23 lin-23 lin-23
QL QL QL QL P9 P10 P9 P10 P10 P10

H H
L1
XN X N H H H
L1 X NN X NN H
NN X N NN NN X NN X NN NN
X NN
X N X N NN NN NN
N
N XN N NN
X NN N
X X X X X
NN
X
NN L2 NN N X
X NNNN
N
NN

D wild type E wild type lin-23 lin-23

AC Z1 Z4 Z1 Z4 Z1 Z4
P3p P4p P5p P6p P7p P8p

HH L1
DTC

DTC

L3 HH HH
DTC

DTC

Vulva
L2
DTC

DTC
DTC

lin-23
AC

H
H
HH
HH F wild type lin-23
L3 int1 int2 int3-9 int1 int2 int3-9
HH H
L1 S S S
HH S S
Vulva
5074 E. T. Kipreos, S. P. Gohel and E. M. Hedgecock

(Barstead and Waterston, 1989). The missing 5′ region was obtained motile, if sometimes misoriented (Fig. 1B). Interestingly, the
by PCR amplification from the library using primers to unique lin-23 programmed asymmetry observed in certain neuroblast
sequence (GTCGAA-TCGTGTTATCACT) and vector sequences. divisions, e.g., Q.a and Q.p, is reduced or lost (data not shown).
PCR products were cloned and two independent isolates were Ventral hypodermoblasts generate several motorneurons plus
sequenced. The first seven bases of the lin-23 cDNA match the SL1 a hypodermal cell during the L1 stage (Sulston and Horvitz,
trans-spliced leader (Krause and Hirsh, 1987), indicating that the
cDNA contains the entire 5′ end. Missing 3′ untranslated lin-23 cDNA
1977). The central six hypodermal cells, P3.p-P8.p, form the
sequence was obtained by sequencing the EST clone yk22d11 (a kind vulval primordium, while others fuse with the syncytial
gift fromY. Kohara). The lin-23 cDNA sequence has been deposited in hypodermis (Fig. 1D). The committed vulval precursors divide
GenBank (accession number AF275253). during the L3 stage, responding in part to an inductive signal
Standard molecular biology methods were used throughout from the uterine anchor cell. In lin-23, ventral hypodermoblasts
(Sambrook et al., 1989). RNA samples of the various developmental divide excessively during the L1 stage, producing extra
stages were isolated and processed for northern analysis as described motorneurons, but usually a single hypodermal cell (Fig. 1C).
by Kipreos et al. (1996). Quantitation of northern signals was During the L3 stage, these cells divide excessively, generating
performed on a 300A Molecular Dynamics Densitometer. In situ 42±6 nuclei per vulva (n=21) on average, compared to 22 in
hybridization with lin-23 sense and antisense RNA was performed as wild type. The precursors nearest the uterine anchor cell divide
described by Feng et al. (1999). Immunofluorescence with monoclonal
antibody MH27 (Francis and Waterston, 1991) was performed
most extensively, suggesting that they remain responsive to its
according to the Freeze-Crack protocol (Miller and Shakes, 1995). inductive signal (Fig. 1D). Subsequent invagination and
differentiation of the vulva appear nearly normal (Fig. 2A,B).
Hypodermoblasts G1, G2, W and K also divide excessively (data
not shown).
RESULTS During the L1 stage, mesoblasts Z1 and Z4 each generate 5
presumptive gonadoblasts plus a motile distal-tip cell that
The lin-23 null phenotype determines the shape of the ovary and regulates germline
Four lin-23 alleles were isolated in screens for mutants with proliferation by maintaining the distal germ cells in a mitotic state
excess postembryonic cell divisions, and four additional alleles (Kimble and Hirsh, 1979; Kimble and White, 1981). During the
were kind gifts from T. Schedl, J. Hodgkin, M. L. Edgley and L3 stage, the committed gonadoblasts divide to produce the ovary,
J. D. Plenefisch. These eight alleles, which are all recessive, spermatheca and uterus. In lin-23, Z1 and Z4 often undergo one
appear phenotypically indistinguishable. Embryogenesis is or more extra divisions during the L1 stage, generating extra
normal in homozygous lin-23 progeny of heterozygous gonadoblasts and distal-tip cells (Fig. 1E). In particular, about 38%
hermaphrodites. After hatching, all 53 somatic blast cells of these gonads (94/250) have one extra distal-tip cell, arising from
remain quiescent until stimulated by normal developmental either Z1 or Z4, while another 2% (6/250) have two extra cells,
signals. However, upon entering the mitotic cycle, lin-23 blast one from each mesoblast. These extra distal-tip cells direct the
cells divide excessively (detailed below), generating formation of extra gonadal arms and are competent to keep distal
differentiated cells of appropriate tissue types. None of the 503 germ cells mitotic (Fig. 2C). During the L3 stage, the committed
postmitotic cells created during embryogenesis was observed to gonadoblasts divide excessively, producing an epithelium with
divide in these mutants. normal regional differentiation, including spermathecae with the
Lateral hypodermoblasts, or seam cells, are hypodermal stem characteristic high concentration of tight junctions (Fig. 2D,E).
cells that divide nearly synchronously at each larval stage The non-gonadal mesoblast M also divides excessively, producing
(Sulston and Horvitz, 1977); after each division, non-seam extra coelomocytes, body wall muscles and sex muscles (data not
daughters endoreplicate and then fuse with the syncytial shown).
hypodermis (Hedgecock and White, 1985) (Fig. 1A). In lin-23, Intestinal cells int3-int9 generally undergo nuclear division,
the seam cells divide excessively while retaining their stem cell i.e., mitosis without cytokinesis, at L1 lethargus, followed
character and coordination with the larval molt cycle (Fig. 1A). immediately by endoreplication, while the anterior intestinal
Several other developmental constraints on the seam cell cycle cells int1 and int2 do not enter mitosis and instead undergo a
appear unaffected, e.g., the H0 cell remains quiescent and all single endoreplication at the L1 lethargus (Sulston and
cell divisions cease at the adult stage (data not shown). There Horvitz, 1977; Hedgecock and White, 1985). Additional
is considerable variation in the precise number and timing of endoreplications occur for all intestine cells at each subsequent
the extra cell divisions among hypodermoblasts, even within the molt, so that the adult intestine has binucleate cells with a 32C
same individual (Fig. 1A, data not shown). This variability is DNA content in each nucleus. In lin-23 mutants, int3-int9 cells
also observed for other lineages (Fig. 1B,C,E, data not shown). undergo two successive nuclear divisions at the L1 lethargus,
Neuroblasts Q, V5.pa and T.p, arising from the lateral becoming tetranucleate, while int2 cells often undergo a single
hypodermoblasts, generate a small number of peripheral nuclear division (Fig. 1F). Typically in lin-23 adults, int1 cells
neurons and neuroglia (Sulston and Horvitz, 1977). At are mononucleate with 32C DNA content (32.1±5.1C, n=10),
characteristic positions within each neuroblast lineage, the while binucleate int2 and tetranucleate int3-9 cells have 16C
mitotic spindle must shift asymmetrically during anaphase to DNA content per nucleus (16.9±2.9C, n=10), indicating that the
create daughter cells of unequal sizes. In lin-23, the cell fate absence of lin-23 does not cause excess endoreplicative cycles
decision to generate neuroblasts from hypodermal precursors is and that the extra divisions in lin-23 at the L2 lethargus are
maintained. The committed neuroblasts generated, however, conversions of endoreplicative cycles to mitotic cycles.
produce almost double their normal number of descendants In contrast to somatic blast cell divisions, germ cell number
(Fig. 1A,B, data not shown). These cells generally resemble is not affected in lin-23 mutants. A count of germ cell number
their normal counterparts, e.g., many of the Q descendants are in newly hatched young adults showed no differences between
 lin-23 negatively regulates cell division 5075
lin-23 and wild type, with 665±126 and 684±133 germ cells mutations were mapped by chemical cleavage of mismatched
(n=10), respectively. bases (Cotton et al., 1988) and sequenced. Each allele has a
single GC to AT transition creating a missense codon or
lin-23 maternal product nonsense codon (Fig. 5). Allele rh294, which terminates at
lin-23 hermaphrodites are sterile, producing only a few, inviable codon 12, is evidently a molecular null. All of the alleles
embryos. These embryos appear normal as early blastula, but appear to be functional nulls, as each allele is phenotypically
eventually arrest, usually with excess cells and no overt indistinguishable from rh294.
morphogenesis (Fig. 2F,G). Using DNA
squashes, we have counted over 850 cells
in lin-23 embryos compared to the 558
cells present in wild-type embryos at
hatching (Sulston et al., 1983). To
determine whether lin-23(+) is required
within the germline for normal
embryonic development, we injected lin-
23 antisense RNA (derived from lin-23
cDNA, see below) into the germ cell
syncytia of wild-type hermaphrodites.
For many embryonically acting genes in
C. elegans, antisense and double-strand
RNAs have been shown to eliminate both
maternal and zygotic products (Fire et
al., 1998). Eggs fertilized in the first few
hours after injection of lin-23 antisense
RNA completed embryogenesis, but
exhibited postembryonic hyperplasia
characteristic of lin-23 larvae. Eggs
fertilized later invariably arrested
during embryogenesis, phenocopying
embryos from homozygous lin-23
hermaphrodites (data not shown). We
conclude that lin-23(+) is required for
development beyond blastula, but that
maternal products suffice through the
end of embryogenesis.

Molecular analysis of lin-23

lin-23 was mapped between lin-4 and
dpy-10 on chromosome II. Genomic
clones from this region were assayed as
transgenes for their ability to complement
lin-23(e1883). lin-23(+) activity was
localized to a 9.2 kb fragment of cosmid
T25H2 and this fragment was used to
isolate cDNA clones (Fig. 3). The lin-23
cDNA hybridizes to a single transcript of
approx. 2.8 kb with the highest level of
expression in embryos and gravid adults
(Fig. 4A). In situ hybridization revealed
high levels of lin-23 mRNA in the germ
line of adults and early embryos (Fig.
4B,C). Zygotes have high levels of lin-23
mRNA indicating the presence of
maternally provided mRNA (Fig. 4C).
The level of lin-23 mRNA decreases Fig. 2. Hyperplasia in lin-23 mutants. (A-C,F,G) Differential interference contrast and (D,E)
immunofluorescence micrographs. Developing vulval, v, and uterine, u, epithelia in (A) wild-
during embryogenesis and was type and (B) lin-23(e1883) L4 hermaphrodites. (C) Gonad with three distal arms (arrowheads)
undetectable above background levels in in lin-23(e1883) L3 hermaphrodite. Spermathecae (arrowheads) in (D) wild-type and (E) lin-
larvae (Fig. 4C, data not shown). 23(e1883) adult hermaphrodites stained with monoclonal antibody MH27 to zonula adherens
The full-length lin-23 mRNA is (Francis and Waterston, 1991). Extra spermathecae in lin-23 mutants are associated with the
predicted to encode a cytosolic protein extra gonadal arms. (F) Wild-type embryo near hatching (approx. 700 minutes) and (G) arrested
of 665 amino acids. Six lin-23 lin-23 embryo (>700 minutes). Scale bars are 10 µm.
5076 E. T. Kipreos, S. P. Gohel and E. M. Hedgecock
lin-4 lin-23 dpy-10

0.1 map unit

lin-23(+) # rescued
Activity lines/total
Cosmids
T22G1 - 0/4
C29F5 - 0/2
T25H2 + 17/18
F58F12 + 1/1

Clones
T25H2 + 17/18
T ScaL - 0/1
T KpnI + 0/2
T KpnI-ScaI
10.7 kb
- 0/6
T ScaI + 7/7
9.2 kb
T StuI +/ - 3/25
7.7 kb
T BamHI - 0/22

ScaI StuI BamHI

lin-23 mRNA:
1kb

Fig. 3. Cloning the lin-23 gene by complementation. A genetic map Fig. 4. lin-23 developmental mRNA expression. (A) Northern blot of
of the lin-4 to dpy-10 interval of chromosome II is shown at the top. total RNA probed initially for lin-23 mRNA (upper) and then 18S
The approximate extent of cosmids used in complementation rRNA (lower). Relative intensities of the lin-23 signals, using rRNA
experiments is denoted by dashed lines. The first column on the right to correct for loading differences, were 96% (E, mixed embryos),
indicates whether a clone was found active (plus) or inactive (minus) 34% (L1 larvae), 33% (L2), 27% (L3), 33% (L4), 44% (yA, young
in rescuing lin-23 homozygotes. The second column gives the number adults), and 100% (grA, gravid adults). (B) In situ hybridization of
of stable transformed lines that rescued lin-23 relative to the total adult wild-type hermaphrodites with control lin-23 sense RNA (top)
number of stable lines for a given clone. Cosmid T25H2 and its or antisense RNA (bottom). Note that eggs inside hermaphrodites are
subclones are shown in expanded scale. A schematic of the 10.7 kb generally not permeable to RNA probes. (C) Wild-type zygote
T∆ScaI fragment with the position of lin-23 exons is shown at the hybridized with lin-23 sense RNA (top) and wild-type zygote and 60-
bottom. The position of the StuI and BamHI sites in T∆ScaI cell stage embryo hybridized with lin-23 antisense RNA (bottom).
correspond to the 5′ ends of the T∆StuI and T∆BamHI subclones.
Subclone T∆StuI (lacking the first exon) gave a limited rescue of lin-
23 (3 partially rescuing lines/25 stable lines), presumably because 505) (Fig. 6A). In contrast to the low level of homology among
transcription initiation occurred from cryptic promoters/enhancers or WD-repeat proteins generally (Neer et al., 1994), this high level
extragenic promoters/enhancers present in the extrachromosomal of homology suggests that slmb and the two highly similar and
array (Mello and Fire, 1995). apparently recently duplicated genes βTRCP and βTRCP2 are
the Drosophila and human orthologs of lin-23.
LIN-23 contains an F-box motif (Bai et al., 1996; Kumar and In budding yeast, two genes encode proteins with both an F-
Paietta, 1995) followed by seven tandem WD repeats (Neer et box and WD repeats, MET30 and CDC4. Both Met30p and
al., 1994) (Figs 5, 6A). The F-box motif is found in a number Cdc4p function as F-box components of SCF ubiquitin-ligase
of cell cycle regulators and has been shown capable of complexes (see Deshaies, 1999). LIN-23 is more similar to
mediating protein-protein interaction (Bai et al., 1996). WD Met30p than to Cdc4p, with 27% identity to Met30p and 23%
repeats are found in proteins with diverse functions (Neer et al., identity to Cdc4p in the F-box and WD-repeat regions (Fig.
1994). Seven WD repeats form a β propeller structure, which 6A). Another C. elegans gene, sel-10 (Hubbard et al., 1997),
is thought to function as a scaffold for protein-protein has a higher degree of homology to CDC4 than to MET30, with
interactions (Sondek et al., 1996; Garcia-Higuera et al., 1996). 31% identity to Cdc4p and 23% identity to Met30p in the F-
box and WD-repeat regions (Fig. 6A). Phylogenetic analysis
lin-23 homologs suggests that sel-10 is more evolutionarily related to CDC4,
Comparison of LIN-23 to genomic databases reveals a high while lin-23 is more related to MET30 (Fig. 6B). As both the
degree of homology to the F-box/WD-repeat proteins S. cerevisiae and C. elegans genomes have been sequenced, and
βTRCP/BTRC and βTRCP2/FBXW1B/HOS in humans, and there are no other C. elegans genes that are more closely related
Slmb in Drosophila (Margottin et al., 1998; Cenciarelli et al., to MET30 and CDC4; it appears that lin-23 and MET30 are
1999; Winston et al., 1999a; Koike et al., 2000; Jiang and Struhl, orthologs, and that sel-10 and CDC4 are orthologs. Both
1998; Theodosiou et al., 1998). LIN-23 is 36-40% identical to Met30p and Cdc4p form SCF complexes containing the cullin
the three proteins in the F-box region (LIN-23 residues 56-124); Cdc53p (see Deshaies, 1999). The C. elegans homolog of
and 81-83% identical in the WD-repeat region (residues 172- CDC53, cul-1, has a hyperplasia mutant phenotype similar to
 lin-23 negatively regulates cell division 5077

Fig. 5. lin-23 mRNA and protein sequence. The first seven nucleotides of the lin-23 cDNA derive from a trans-spliced SL1 leader (overlined) and
indicate that the cDNA is full-length (Krause and Hirsh, 1987). lin-23 exon junctions are marked by triangles. The predicted LIN-23 protein
sequence is shown below the nucleotide sequence. The F-box sequence at residues 87 to 127 is boxed. The seven WD-repeat sequences are
denoted by brackets below the protein sequence. A potential bipartite nuclear localization sequence (Dingwall and Laskey, 1991) at residues 160
to 174 is underlined. Alleles in which GC to AT transitions create terminator codons are circled: rh294 (Q12 to TAG), e1925 (W327 to TGA),
e1883 (W450 to TGA), oz107 (W450 to TAG]) and rh194 (W499 to TAG). Allele rh293, in which a GC to AT transition created a missense
mutation (G411 to R), is boxed.

lin-23 (Kipreos et al., 1996); and the cul-1, lin-23 double not arise in slmb(−) cells, but occurs in surrounding tissues due
homozygote arrests development earlier than either single to the ectopic expression of the morphogens Decapentaplegic
homozygote. (Dpp) and Wingless (Wg; Jiang and Struhl, 1998; Theodosiou
To test whether lin-23 functions redundantly with the et al., 1998). To test whether lin-23 functions in a cell
CDC4 ortholog sel-10, we constructed a heterozygous lin-23, autonomous or non-autonomous manner, we carried out mosaic
homozygous sel-10 strain. sel-10 encodes a suppressor of lin- analysis. Our analysis focused on the vulva, as the most is
12 loss-of-function alleles (Sundaram and Greenwald, 1993). known about the regulation of cell proliferation in the vulva
sel-10 homozygous null animals are essentially wild type with relative to other C. elegans tissues. During the L3 stage, the
only mild impenetrant phenotypes that are attributable to mild vulval precursor cells (VPCs) are induced to proliferate in 1°,
overactivity of the lin-12 receptor, e.g., 1% of animals lack an 2°, or 3° cell lineage patterns based on an initial inductive signal
anchor cell and 4% of animals have gonad migration defects secreted by the anchor cell of the uterus (Greenwald, 1997).
(Hubbard et al., 1997). The double lin-23, sel-10 homozygote Mosaic analysis was performed on a lin-23, ncl-1 homozygous
progeny of this strain appeared to have a phenotype identical to line that contained an extrachromosomal array containing the
lin-23 homozygotes, with the minor exception that the number lin-23(+) gene, the ncl-1(+) gene, and a plasmid in which the
of animals with extra distal tip cells decreased relative to lin-23 let-858 promoter drove the expression of green fluorescent
homozygotes from 40±7% (n=250) to 6±2% (n=102). This protein (GFP). Cells that had the array were lin-23(+), ncl-1(+),
experiment suggests that there is no significant functional and had green epifluorescence. Cells that lost the array were lin-
redundancy between lin-23 and sel-10. 23(−), ncl-1(−) with large nucleoli, and had no epifluorescence.
The uterine precursor cells arise from the somatic gonadal
lin-23 functions cell autonomously precursors Z1 and Z4, which both arise from the embryonic MS
Loss-of-function mutations of the Drosophila ortholog of lin- blast cell. The VPCs P5.p, P6.p, and P7.p arise from the
23, slmb, also result in excessive proliferation (Jiang and Struhl, embryonic blast cell ABp. Loss of the array in MS produces
1998; Theodosiou et al., 1998). However, the proliferation does uterine precursors that lack lin-23(+) and that exhibit uterine
5078 E. T. Kipreos, S. P. Gohel and E. M. Hedgecock

A
**************************************************************************
LIN-23: MD K I V H R L S H Y Q L G K V D N F I R P M L Q R D F I S N L P A - - - - H L V E L I L F N V N S D S L K S C E E V S T S W R C A L A R - G Q H W K K L I
bTrCP: VE H L I S Q M C H Y Q H G H I N S Y L K P M L Q R D F I T A L P A R G L D H I A E N I L S Y L D A K S L C A A E L V C K E W Y R V T S D - G M L W K K L I
Slimb: VE H L L S R M C H Y Q H G Q I N A Y L K P M L Q R D F I T L L P I K G L D H I G E N I L S Y L D A E S L K S S E L V C K E W L R V I S E - G M L W K K L I
Met30p: LD G I L S T S C F P Q L S Y I S S L V T H M I K I D F I S I L P Q - - - - E L S L K I L S Y L D C Q S L C N A T R V C R K W Q K L A D D - D R V W Y H M C
Cdc4p: LF R L V A N M D R S E L S D L G T L I K D N L K R D L I T S L P F - - - - E I S L K I F N Y L Q F E D I I N S L G V S Q N W N K I I R K S T S L W K K L L
SEL-10: FT R L L Q E S N M T N I R Q L R A I I E P H F Q R D F L S C L P V - - - - E L G M K I L H N L T G Y D L L K V A Q V S K N W K L I S E I - D K I W K S L G

LIN-23: EKNVRSDSLWWGLSEKRQWDKFLNISRDMSVRRICEKFNYDVNIKRDKLDQLILMHVFYSKLYPKIIRDIHNIDNNWK
bTrCP: ERMVRTDSLWRGLAERRGWGQYLF--------------------KNKPPDGNAPPNSFYRALYPKIIQDIETIESNWR
Slimb: ERKVRTDSLWRGLAERRNWMQYLF---------------------KPRPGQTQRPHSFHRELFPKIMNDIDSIENNWR
Met30p: EQHIDRKCPNCGWGLPLLHMKRARIQQ-----------------NSTGSSSNADIQTQTTRPWKVIYRERFKVESNWR
Cdc4p: IS E N F V S P K G F N S - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - L N L K L S Q K Y P K L S Q Q D R L - R L S F L E N I F I L K N W Y
SEL-10: VE E F K H H P D P T D R V T G A W Q G T A I A A G V T I P D H I Q P C D L N - - - - - V H R F L K L Q K F G D I F E R A A D K S R Y L R A D K I E K N W N

LIN-23 (2 0 7 ) : RG N Y K M T R I N C Q S E N S K G V Y - C L Q Y D D D K I V S G L R D N T I K I W D
bTrCP (2 5 2 ) : CG R H S L Q R I H C R S E T S K G V Y - C L Q Y D D Q K I V S G L R D N T I K I W D
Slimb (1 9 5 ) : TG R H M L R R I N C R S E N S K G V Y - C L Q Y D D G K I V S G L R D N T I K I W D
Met30p (2 9 0 ) : -- - K G H C R I Q E F K G H M D G V L T - L Q F N Y R L L F T G S Y D S T I G I W D
Cdc4p (3 6 7 ) : -N P K F V P Q R T T L R G H M T S V I T C L Q F E D N Y V I T G A D D K M I R V Y D
SEL-10 (2 4 2 ) : -- A N P I M G S A V L R G H E D H V I T C M Q I H D D V L V T G S D D N T L K V W C

LIN-23 (2 4 9 ) : RK D Y S C S R I L S G H T G S V L - - C L Q Y D N R V I I S G S S D A T V R V W D
bTrCP (2 9 4 ) : KN T L E C K R I L T G H T G S V L - - C L Q Y D E R V I I T G S S D S T V R V W D
Slimb (2 3 7 ) : RT D L Q C V K T L M G H T G S V L - - C L Q Y D D K V I I S G S S D S T V R V W D
Met30p (3 2 9 ) : LF T G K L I R R L S G H S D G V K - - T L Y F D D R K L I T G S L D K T I R V W N
Cdc4p (4 0 9 ) : SI N K K F L L Q L S G H D G G V W - A L K Y A H G G I L V S G S T D R T V R V W D
SEL-10 (2 8 3 ) : ID K G E V M Y T L V G H T G G V W T S Q I S Q C G R Y I V S G S T D R T V K V W S

LIN-23 (2 8 9 ) : VETGECIKTLIHHCEAVLHLRFANG----IMVTCSKDRSIAVWD
bTrCP (3 3 4 ) : VNTGEMLNTLIHHCEAVLHLRFNNG----MMVTCSKDRSIAVWD
Slimb (2 7 7 ) : VNTGEMVNTLIHHCEAVLHLRFNNG----MMVTCSKDRSIAVWD
Met30p (3 6 9 ) : YI T G E C I S T Y R G H S D S V L S V D S Y Q K - - - - V I V S G S A D K T V K V W H
Cdc4p (4 5 0 ) : IK K G C C T H V F E G H N S T V R C L D I V E Y K N I K Y I V T G S R D N T L H V W K
SEL-10 (3 2 5 ) : TV D G S L L H T L Q G H T S T V R C M A M A G S - - - - I L V T G S R D T T L R V W D

LIN-23 (3 2 9 ) : --------------------MVSPRDITIRRVLVGHRAAVNVVDFDD--RYIVSASGDRTIKVWS
bTrCP (3 7 4 ) : --------------------MASPTDITLRRVLVGHRAAVNVVDFDD--KYIVSASGDRTIKVWN
Slimb (3 1 7 ) : --------------------MTSPSEITLRRVLVGHRAAVNVVDFDE--KYIVSASGDRTIKVWS
Met30p (4 0 9 ) : ------------------------VESRTCYTLRGHTEWVNCVKLHPKSFSCFSCSDDTTIRMWD
Cdc4p (4 9 4 ) : LPKESSVPDHGEEHDYPLVFHTPEENPYFVGVLRGHMASVRTVSGHG--NIVVSGSYDNTLIVWD
SEL-10 (3 6 5 ) : -----------------------VESGRHLATLHGHHAAVRCVQFDG--TTVVSGGYDFTVKIWN

LIN-23 (3 7 2 ) : MDTLEFVRTLAGHRRGIACLQY----------RGRLVVSGSSDNTIRLWD
bTrCP (4 1 2 ) : TSTCEFVRTLNGHKRGIACLQY----------RDRLVVSGSSDNTIRLWD
Slimb (3 6 0 ) : TSTCEFVRTLNGHKRGIACLQY----------RDRLVVSGSSDNSIRLWD
Met30p (4 5 0 ) : IRTNSCLKVFRGHVGQVQKIIPLTIKDVENLATDNTSDGSSPQDDPTMTD
Cdc4p (5 5 7 ) : VAQMKCLYILSGHTDRIYSTIY--------DHERKRCISASMDTTIRIWD
SEL-10 (4 0 5 ) : AHTGRCIRTLTGHNNRVYSLLF--------ESERSIVCSGSLDTSIRVWD

LIN-23 (4 1 2 ) : --------------------IHSGVCLRVLEGHEELVRCIRFDEKRIVSGAYDGKIKVWD
bTrCP (4 5 7 ) : --------------------IECGACLRVLEGHEELVRCIRFDNKRIVSGAYDGKIKVWD
Slimb (4 0 0 ) : --------------------IECGACLRVLEGHEELVRCIRFDTKRIVSGAYDGKIKVWD
Met30p (5 0 0 ) : --------------------GADESDTPSNEQETVLDENIPYPT-HLLSCGLDNTIKLWD
Cdc4p (5 9 9 ) : LENIWNNGECSYATNSASPCAKILGAMYTLQGHTALVGLLRLSDKFLVSAAADGSIRGWD
SEL-10 (4 4 7 ) : -----------------FTRPEGQECVALLQGHTSLTSGMQLRGNILVSCNADSHVRVWD

LIN-23 (4 5 2 ) : LQAALDPRALSSEICLCSLVQHTGRVFRLQ---FDDFQIVSSSHDDTILIWD
bTrCP (4 9 7 ) : LVAALDPRAPAGTLCLRTLVEHSGRVFRLQ---FDEFQIVSSSHDDTILIWD
Slimb (4 4 0 ) : LVAALDPRAASNTLCLNTLVEHTGRVFRLQ---FDEFQIVSSSHDDTILIWD
Met30p (5 3 9 ) : -- - - - - - - - V K T G K C I R T Q F G H V E G V W D I A - - - A D N F R I I S G S H D G S I K V W D
Cdc4p (6 5 9 ) : -- - - - - - - - - A N D Y S R K F S Y H H T N L S A I T T F - - Y V S D N I L V S G S E N Q F N I Y N
SEL-10 (4 9 0 ) : -- - - - - - - - I H E G T C V H M L S G H - - R S A I T S L Q W F G R N M V A T S S D D G T V K L W D

B βTRCP2
100/82/99
79/66/83 βTRCP
100/52/100 Slmb
100/100/100 LIN-23
Met30p
99/100/88
SEL-10
Cdc4p
 lin-23 negatively regulates cell division 5079

Fig. 6. (A) Sequence comparison of C. elegans LIN-23 and SEL-10, particular, in 27 of 27 cases in which the vulva was lin-23(+)
H. sapiens βTRCP, D. melanogaster slmb, and S. cerevisiae Met30p and the uterus was lin-23(−), the vulva had normal cell
and Cdc4p (Hubbard et al., 1997; Spevak et al., 1993; Jiang and proliferation and morphogenesis (Fig. 7A,B). In two cases with
Struhl, 1998; Theodosiou et al., 1998; Thomas et al., 1995; Yochem a lin-23(+) vulva and lin-23(−) uterus, we observed a multivulva
and Byers, 1987). The alignment begins with the extended F-box of phenotype in which an extra vulva was created distant from the
Kumar and Paietta (1995) (stars above sequence), and extends normal site, apparently the result of an extra anchor cell (AC)
through the seven WD-repeats. The standard F-box motif of Bai et al.
(1996), is boxed and the seven WD repeats are offset. Gaps are shown
being created in the uterus (data not shown). Finally, we
as dashes. Sequences were aligned using the Clustal W program observed 12 animals showing mosaicism among the VPCs or
(Thompson et al., 1994) and then adjusted manually. Residues their descendants. In all cases, descendants lacking lin-23
identical in three or more proteins are shown in bold. exhibited hyperplasia, while those with lin-23(+) expression
(B) Phylogenetic analysis using the sequences above plus the H. had normal cell divisions (Fig. 7C-E). These results indicate
sapiens βTRCP2 protein, which is 92% identical to βTRCP in the that lin-23 functions cell autonomously to regulate cell
region aligned above. Regions of the alignment that were used in the proliferation.
phylogeny are underlined in the SEL-10 sequence. Rooted at its
midpoint, this cladogram depicts the tree obtained by Parsimony,
Maximum Likelihood, and Neighbor-joining methods. Exhaustive DISCUSSION
search methods were used for both Parsimony and Maximum
Likelihood, indicating that this is the best tree for both methods.
Bootstrap support values are given in the nodes for Parsimony, In this paper we present evidence that the lin-23 gene functions
Neighbor-joining, and Maximum Likelihood methods, respectively as a negative cell cycle regulator. From heterozygous parents,
(separated by slash marks). lin-23 maternal product, which is provided at least in part as
mRNA, is sufficient for normal embryogenesis. In the absence
hyperplasia, while loss of the array in ABp produces VPCs that of maternal product, embryos arrest development with excess
lack lin-23(+) and that exhibit hyperplasia of the vulva. Loss of cell numbers, suggesting that lin-23 is required to restrain cell
the array in the uterus did not promote vulval hyperplasia, in proliferation during embryonic development. lin-23 maternal

Fig. 7. Analysis of lin-23 vulval mosaics. Mosaic analysis was performed on lin-23(oz107); ncl-1(e1865) homozygous animals carrying an
extrachromosomal array, ekEx11, that contains the lin-23(+) gene, the ncl-1(+)gene, and a plasmid with GFP expressed by the let-858 promoter. Cells
containing the extrachromosomal array were scored for the presence of epifluorescence and absence of Ncl phenotype (large nucleoli). (A,B) Loss of
the array in the uterus does not affect vulva development. (A) DIC image of the vulva and uterus in a late-L4 stage lin-23(oz107); ncl-1(e1865)
hermaphrodite mosaic for array ekEx11. The number of nuclei in each half of the vulva are given; note that wild type have 11 nuclei in each half.
(B) Epifluorescent image of the same animal, with the number of cells fluorescing in each half of the vulva; not all cells are visible in this focal plane.
The uterus, which lacks the array, exhibits hyperplasia. (C-E) Cells that lose the array within the vulva exhibit hyperplasia. The lineages of the VPCs
P5.p, P6.p, and P7.p (Sulston and Horvitz, 1977) are shown (C, top). To demonstrate the location of the VPC descendants in the L4 vulva, lines
connect the terminal vulval cells of the lineage with the location of their nuclei in the L4 vulva (Sulston and Horvitz, 1977) (C, bottom). The position
of the filled circles next to the lineage indicates the VPC or VPC descendant in which the ekEx11 extrachromosomal array was lost; in one case the
array was apparently lost in a precursor to both P6.p and P7.p. In every case, cells that lost the array produced extra cells (6.0±4.2 extra cells per loss of
array, n=12). In all cases, neighboring vulva cells that did not lose the array had normal numbers of cell divisions. (D,E) An example of loss of the
ekEx11 array in P5.p. (D) DIC image of the uterus and vulva of a mid-L4 stage lin-23(oz107); ncl-1(e1865) hermaphrodite that is mosaic for array
ekEx11. The number of cells for each half of the vulva are given: 17 in the anterior and 11 in the posterior. (E) Epifluorescent image of the same
animal. The number of fluorescent cells for each half of the vulva are given; the 4 fluorescent cells in the anterior were found in the location of wild-
type P6.pa descendants. The 13 cells without the array were in the region of the vulva normally populated by P5.p descendants. Scale bar is 10 µm.
5080 E. T. Kipreos, S. P. Gohel and E. M. Hedgecock

product appears to be exhausted by the L1 larval stage. In the L1 conjugating enzyme Cdc34p (Deshaies, 1999). Both Cdc4p and
and subsequent larval stages, blast cells divide supernumerary Met30p are involved in the degradation of cell cycle regulators.
times to produce excess cells. All tissue types are represented Cdc4p is required for the degradation of G1 cyclins (Cln1p and
among the blast cells that divide excessively in lin-23 mutants, Cln2p), cyclin-dependent kinase inhibitors (Sic1p and Far1p),
indicating that lin-23 is required to restrain mitotic cell division and the DNA replication protein Cdc6p (see Deshaies, 1999).
irrespective of tissue type. Met30p is required for the degradation of the CDK inhibitory
In lin-23 mutants, cells appear to require normal signals to kinase Swe1p, and for the repression of the sulfur network
start cell division. Only blast cells and their descendants divide genes by inactivating the transcription factor Met4p (Kaiser et
in lin-23 larval stages, while postmitotic non-blast cells do not al., 1998; Thomas et al., 1995). Increased activity of Met4p in
divide. Blast cells do not divide precociously, but rather wait met30 mutants produces a G1 arrest with increased turnover of
until the appropriate developmental stage to divide. Further, the mRNA for the G1 cyclins CLN1, CLN2 and PCL2 (Patton
among the VPCs, those closest to the AC, which secretes an et al., 2000).
inductive/proliferative signal (see Greenwald, 1997), divide Parsimony analysis suggests that LIN-23 is more closely
more than those further from the AC, as is true for wild type. related to Met30p, while another C. elegans protein SEL-10 is
Final differentiation states of lin-23 cells appear normal. more closely related to Cdc4p (Fig. 6). SEL-10 functions to
Distinct neuronal, hypodermal, somatic gonad, muscle, and negatively regulate LIN-12 activity and has been found to
intestine phenotypes are readily observed by DIC microscopy. physically interact with the intracellular domain of LIN-12,
The spermathecae in lin-23 mutants expresses adherens suggesting that SEL-10 functions in the turnover of LIN-12
junctions, as revealed by staining with the MH27 antibody, (Hubbard et al., 1997). Surprisingly, a null allele of sel-10 has
similar to differentiated wild-type spermathecae. Finally, the only minor, impenetrant phenotypic consequences in a wild-
extra distal tip cells produced in lin-23 mutants are both type lin-12 genetic background (Hubbard et al., 1997). sel-10
competent to lead out the developing gonadal arm and to keep is the apparent ortholog of the yeast cell cycle regulator CDC4.
distal germ cells in a mitotic state. While the sel-10 mutant phenotype does not indicate a role in
Morphogenesis is disorganized in several larval tissues in lin- cell cycle regulation, it is possible that another gene functions
23 mutants relative to wild type, e.g., spermathecae and uterus, redundantly with sel-10 to effect cell cycle regulation. However,
and in embryos lacking lin-23 maternal product. lin-23 may a double mutant of lin-23 and sel-10 failed to uncover synthetic
be required for morphogenic processes, alternatively, the cell cycle phenotypes, suggesting that lin-23 does not share
disorganization may be the byproduct of excess cell critical functions with sel-10. Finally, the dissimilar mutant
proliferation. In both cki-1 and cul-1 loss-of-function animals, phenotypes of lin-23 (defective cell cycle exit) and met30 (cell
which also exhibit hyperplasia, there is a similar disorganization cycle arrest) further suggests that the cellular functions of SCF
of larval tissues and lack of clear morphogenesis in embryos complexes have not been conserved between yeast and metazoa.
(Kipreos et al., 1996; Feng et al., 1999; Hong et al., 1998; data lin-23 has a mutant hyperplasia phenotype similar to that of
not shown). the cul-1 gene, which also encodes an SCF component (Kipreos
In contrast to mitotic cell cycles, lin-23 is not required for et al., 1996). Orthologs of lin-23 and cul-1 function together in
restraining purely endoreplicative cycles. The ploidy of SCF complexes in both yeast and humans, making it likely that
intestine cells, which endoreplicate in wild type to achieve 32n, LIN-23 and CUL-1 also form an SCF complex in C. elegans.
does not increase beyond 32n in lin-23 mutants. The first The phenotypes of cul-1 and lin-23, while similar, do show
endoreplication for most wild-type intestine cells is preceded differences. Whereas lin-23 maternal products perdure only
by a nuclear division (segregation of chromosomes but no through embryogenesis, cul-1 maternal products can suffice
cytokinesis). Interestingly, the first endoreplication after the through the L1 stage. Despite the later onset of the cul-1 mutant
nuclear division is converted to a second nuclear division in lin- phenotype, cul-1 larvae arrest development earlier with a more
23 mutants, suggesting that lin-23 is required to allow these severe hyperplasia, e.g., cul-1 mutants exhibit on average twice
cells to bypass mitosis. Later endoreplications are not converted as much vulval hyperplasia as lin-23 mutants (Kipreos et al.,
to nuclear divisions in lin-23 mutants, suggesting that mitotic 1996; this study).
bypass is maintained by a LIN-23-independent mechanism, Cullins have been found to interact with multiple F-box
potentially via the downregulation of mitotic cyclin, as is proteins to target different substrates. The phenotype of
observed in Drosophila endoreplicative cycles (Edgar and inactivation of the cullin component should be equivalent to the
Lehner, 1996). sum of the inactivation of the separate F-box proteins with
We cloned lin-23 and determined that it encodes a putative which it functions (assuming that the F-box proteins do not have
protein with an F-box motif and seven WD repeats. lin-23 is independent functions). Our data therefore predicts that an SCF
expressed at its highest levels in embryos and the adult complex containing CUL-1 and LIN-23 contributes part of the
germline, but has some expression in all developmental stages. negative cell cycle regulatory capability of CUL-1, but that
The complete S. cerevisiae genome has just two F-box/WD- CUL-1 (with a more severe hyperplasia phenotype) will also
repeat proteins, Cdc4p and Met30p (Bai et al., 1996). Both function with other F-box protein(s) to negatively regulate the
Cdc4p and Met30p function as components of an SCF E3 cell cycle. Higher eukaryotic orthologs of lin-23 have been
complex to facilitate the recognition of substrates by a identified in Xenopus (βTRCP), humans (βTRCP and βTRCP2),
ubiquitin-conjugating enzyme (E2) for ubiquitin-mediated and Drosophila (slmb) (Spevak et al., 1993; Margottin et al.,
proteolysis. Met30p and Cdc4p function as the substrate 1998; Suzuki et al., 1999; Jiang and Struhl, 1998). The human
recognition subunits of different SCF complexes, SCFMet30 and ortholog, h-βTRCP, was found to be co-opted by the Human
SCFCdc4. Each complex also includes Skp1p, the cullin Cdc53p, Immunodeficiency Virus Vpu protein to target the CD4 receptor
and Rbx1p/Roc1p/Hrt1p, and interacts with the ubiquitin- for degradation (Margottin et al., 1998); and it also functions in
 lin-23 negatively regulates cell division 5081
the degradation of IκBα and β-catenin (Yaron et al., 1998; and osmium tetroxide and its application to the study of mutations. Proc.
Shirane et al., 1999; Winston et al., 1999b; Hart et al., 1999; Natl. Acad. Sci. USA 85, 4397-4401.
Deshaies, R. J. (1999). SCF and Cullin/Ring H2-based ubiquitin ligases. Annu.
Latres et al., 1999; Fuchs et al., 1999; Kroll et al., 1999;
Rev. Cell. Dev. Biol. 15, 435-467.
Hatakeyama et al., 1999; Spencer et al., 1999; Suzuki et al., Dingwall, C. and Laskey, R. A. (1991). Nuclear targeting sequences – a
1999; Tan et al., 1999; Wu and Ghosh, 1999; Vuillard et al., consensus? Trends Biochem. Sci. 16, 478-481.
1999). Edgar, B. A. and Lehner, C. F. (1996). Developmental control of cell cycle
Cells lacking the Drosophila ortholog of lin-23, slmb, regulators: a fly’s perspective. Science 274, 1646-1652.
Felsenstein, J. (1993). PHYLIP (Phylogeny Inference Package) version 3.5c.
ectopically express the morphogens Wg and Dpp (Jiang and Distributed by the author. Department of Genetics, University of Washington,
Struhl, 1998; Theodosiou et al., 1998). Ectopic intersection of Seattle.
Wg- and Dpp-expressing cells in the same compartment of Feng, H., Zhong, W., Punkosdy, G., Gu, S., Zhou, L., Seabolt, E. K. and
imaginal discs induces formation of ectopic appendages (see Kipreos, E. T. (1999). CUL-2 is required for the G1-to-S phase transition
and mitotic chromosome condensation. Nat. Cell Biol. 1, 486-492.
Serrano and O’Farrell, 1997). Animals with mosaic patches of Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., Driver, S. E. and Mello,
slmb(−) cells have extra appendages that are attributable to the C. C. (1998). Potent and specific genetic interference by double-stranded
ectopic expression of Wg and Dpp (Jiang and Struhl, 1998; RNA in Caenorhabditis elegans. Nature 391, 806-811.
Theodosiou et al., 1998). The excess cell proliferation that is Francis, R. and Waterston, R. H. (1991). Muscle cell attachment in
observed in slmb mosaic animals derives from the non-cell Caenorhabditis elegans. J. Cell Biol. 114, 465-479.
Fuchs, S. Y., Chen, A., Xiong, Y., Pan, Z. Q. and Ronai, Z. (1999). HOS, a
autonomous growth of wild-type imaginal disc cells that are human homolog of Slimb, forms an SCF complex with Skp1 and Cullin1
induced to create ectopic appendages by the secretion of and targets the phosphorylation-dependent degradation of IkappaB and beta-
morphogens by the slmb mutant cells. Interestingly, clones of catenin. Oncogene 18, 2039-2046.
cells with null slmb alleles, although initially present in the Garcia-Higuera, I., Fenoglio, J., Li, Y., Lewis, C., Panchenko, M. P., Reiner,
O., Smith, T. F. and Neer, E. J. (1996). Folding of proteins with WD-
imaginal disc, are not observed in the ectopic appendage that repeats: comparison of six members of the WD-repeat superfamily to the G
they later induce, suggesting either decreased cell proliferation protein β subunit. Biochem. J. 35, 13985-13994.
or cell lethality (Theodosiou et al., 1998). Greenwald, I. (1997). Development of the vulva. In C. elegans II (ed. D. L.
Hyperplasia in lin-23 mutants does not appear to be due to Riddle), pp. 519-542. Plainview, NY: Cold Spring Harbor Laboratory Press.
morphogen secretion, as the hyperplasia occurs throughout Hart, M., Concordet, J. P., Lassot, I., Albert, I., del los Santos, R., Durand,
H., Perret, C., Rubinfeld, B., Margottin, F., Benarous, R. et al. (1999).
development in all tissue types, and mosaic analysis indicates The F-box protein beta-TrCP associates with phosphorylated beta-catenin
that lin-23 mutant hyperplasia occurs through a cell and regulates its activity in the cell. Curr. Biol. 9, 207-210.
autonomous mechanism. It is currently unclear why the two Hatakeyama, S., Kitagawa, M., Nakayama, K., Shirane, M., Matsumoto,
orthologs have such different modes of action in Drosophila and M., Hattori, K., Higashi, H., Nakano, H., Okumura, K., Onoe, K. et
al. (1999). Ubiquitin-dependent degradation of IΚBα is mediated by a
C. elegans. It is tempting to speculate that both genes evolved ubiquitin ligase Skp1/Cul1/F-box protein FWD1. Proc. Natl. Acad. Sci.
to negatively regulate cell proliferation, however, differences in USA 96, 3859-3863.
the way cell proliferation is regulated in Drosophila, via the Hedgecock, E. M. and White, J. G. (1985). Polyploid tissues in the nematode
action of morphogens to pattern surrounding cells, compared to Caenorhabditis elegans. Dev. Biol. 107, 128-133.
C. elegans, where cell intrinsic decisions are much more Hong, G., Roy, R. and Ambros, V. (1998). Developmental regulation of a
cyclin-dependent kinase inhibitor controls postembryonic cell cycle
important, have shaped the mechanism of action of the two gene progression in C. elegans. Development 125, 3585-3597.
products. Hubbard, E. J., Wu, G., Kitajewski, J. and Greenwald, I. (1997). sel-10, a
negative regulator of lin-12 activity in Caenorhabditis elegans, encodes a
We thank Tim Schedl, Jonathon Hodgkin, Mark Edgley, and John member of the CDC4 family of proteins. Genes Dev. 11, 3182-3193.
Plenefisch for kind gifts of additional lin-23 alleles; Robert Barstead Jiang, J. and Struhl, G. (1998). Regulation of the Hedgehog and Wingless
for providing a cDNA library; Alan Coulson for providing cosmids; signalling pathways by the F-box/WD40-repeat protein Slimb. Nature 391,
Yuji Kohara for providing cDNAs; and William Kelly for providing 493-496.
Kaiser, P., Sia, R. A. L., Bardes, E. G. S., Lew, D. J. and Reed, S. I. (1998).
plasmid pBK48. Some C. elegans strains were provided by the Cdc34 and the F-box protein Met30 are required for degradation of the cdk-
Caenorhabditis Genetics Center (funded by the NIH National Center inhibitory kinase Swe1. Genes Dev. 12, 2587-2597.
for Research Resources). E. T. K. is supported by grants from the NIH Kelly, W. G., Xu, S., Montgomery, M. K. and Fire, A. (1997). Distinct
(GM55297) and the Human Frontiers Science Program Organization requirements for somatic and germline expression of a generally expressed
(RG-229/98), and E. M. H. is supported by a grant from the NIH Caenorhabditis elegans gene. Genetics 146, 227-238.
(NS26295). Kimble, J. and Hirsh, D. (1979). The postembryonic cell lineages of the
hermaphrodite and male gonads in Caenorhabditis elegans. Dev. Biol. 70,
396-417.
Kimble, J. E. and White, J. G. (1981). On the control of germ cell
REFERENCES development in Caenorhabditis elegans. Dev. Biol. 81, 208-219.
Kipreos, E. T., Lander, L. E., Wing, J. P., He, W. W. and Hedgecock, E. M.
Adachi, J. and Hasegawa, M. (1996). MOLPHY version 2.3: Programs for (1996). cul-1 is required for cell cycle exit in C. elegans and identifies a novel
molecular phylogenetics based on maximum likelihood. Computer Science gene family. Cell 85, 829-839.
Monographs, No. 28. Institute of Statistical Mathematics, Tokyo. Koike, J., Sagara, N., Kirikoshi, H., Takagi, A., Miwa, T., Hirai, M. and
Bai, C., Sen, P., Hofmann, K., Ma, L., Goebl, M., Harper, J. W. and Elledge, Katoh, M. (2000). Molecular cloning and genomic structure of the
S. J. (1996). SKP1 connects cell cycle regulators to the ubiquitin proteolysis betaTRCP2 gene on chromosome 5q35.1. Biochem. Biophys. Res. Commun.
machinery through a novel motif, the F-box. Cell 86, 263-274. 269, 103-109.
Barstead, R. J. and Waterston, R. H. (1989). The basal component of the Krause, M. and Hirsh, D. (1987). A trans-spliced leader on actin mRNA in
nematode dense-body is vinculin. J. Biol. Chem. 264, 10177-10185. C. elegans. Cell 49, 753-761.
Cenciarelli, C., Chiaur, D. S., Guardavaccaro, D., Parks, W., Vidal, M. and Kroll, M., Margottin, F., Kohl, A., Renard, P., Durand, H., Concordet, J.
Pagano, M. (1999). Identification of a family of human F-box proteins. Curr. P., Bachelerie, F., Arenzana-Seisdedos, F. and Benarous, R. (1999).
Biol. 9, 1177-1179. Inducible degradation of IκBα by the proteasome requires interaction with
Cotton, R. G. H., Rodrigues, N. R. and Campbell, R. D. (1988). Reactivity the F-box protein h-betaTrCP. J. Biol. Chem. 274, 7941-7945.
of cytosine and thymine in single-base-pair mismatches with hydroxylamine Kumar, A. and Paietta, J. V. (1995). The sulfur controller-2 negative
5082 E. T. Kipreos, S. P. Gohel and E. M. Hedgecock

regulatory gene of Neurospora crassa encodes a protein with beta-transducin Sulston, J. E. and Horvitz, H. R. (1977). Post-embryonic cell lineages of the
repeats. Proc. Natl. Acad. Sci. USA 92, 3343-3347. nematode, Caenorhabditis elegans. Dev. Biol. 56, 110-156.
Latres, E., Chiaur, D. S. and Pagano, M. (1999). The human F box protein Sulston, J. E., Schierenberg, E., White, J. G. and Thomson, J. N. (1983).
β-Trcp associates with the Cul1/Skp1 complex and regulates the stability of The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev.
β-catenin. Oncogene 18, 849-854. Biol. 100, 64-119.
Liu, J. and Kipreos, E. T. (2000). Evolution of cyclin-dependent kinases Sundaram, M. and Greenwald, I. (1993). Suppressors of a lin-12 hypomorph
(CDKs) and CDK-activating kinases (CAKs): differential conservation of define genes that interact with both lin-12 and glp-1 in Caenorhabditis
CAKs in yeast and metazoa. Mol. Biol. Evol. 17, 1061-1074. elegans. Genetics 135, 765-783.
Margottin, F., Bour, S. P., Durand, H., Selig, L., Benichou, S., Richard, V., Suzuki, H., Chiba, T., Kobayashi, M., Takeuchi, M., Suzuki, T., Ichiyama,
Thomas, D., Strebel, K. and Benarous, R. (1998). A novel human WD A., Ikenoue, T., Omata, M., Furuichi, K. and Tanaka, K. (1999). IκBα
protein, h-beta TrCp, that interacts with HIV-1 Vpu connects CD4 to the ER ubiquitination is catalyzed by an SCF-like complex containing Skp1, cullin-
degradation pathway through an F-box motif. Mol. Cell 1, 565-574. 1, and two F-box/WD40-repeat proteins, betaTrCP1 and betaTrCP2.
Mello, C. and Fire, A. (1995). DNA transformation. In Caenorhabditis elegans: Biochem. Biophys. Res. Commun. 256, 127-132.
Modern Biological Analysis of an Organism. Vol. 48 (ed. H. F. Epstein and Swofford, D. L. (1999). PAUP*. Phylogenetic Analysis Using Parsimony (*and
D. C. Shakes), pp. 451-482. San Diego: Academic Press. Other Methods). Version 4, Sunderland, MA: Sinauer Associates.
Miller, D. M. and Shakes, D. C. (1995). Immunofluorescence microscopy. In Tan, P., Fuchs, S. Y., Chen, A., Wu, K., Gomez, C., Ronai, Z. and Pan, Z.-
Caenorhabditis elegans: Modern Biological Analysis of an Organism (ed. H. Q. (1999). Recruitment of a ROC1-CUL1 ubiquitin ligase by Skp1 and HOS
F. Epstein and D. C. Shakes), pp. 365-394. San Diego: Academic Press. to catalyze the ubiquitination of IΚBα. Mol. Cell 3, 527-533.
Neer, E. J., Schmidt, C. J., Nambudripad, R. and Smith, T. (1994). The ancient Theodosiou, N. A., Zhang, S., Wang, W. Y. and Xu, T. (1998). slimb
regulatory-protein family of WD-repeat proteins. Nature 371, 297-300. coordinates wg and dpp expression in the dorsal-ventral and anterior-
Patton, E. E., Peyraud, C., Rouillon, A., Surdin-Kerjan, Y., Tyers, M. posterior axes during limb development. Development 125, 3411-3416.
and Thomas, D. (2000). SCF(Met30)-mediated control of the Thomas, D., Kuras, L., Barbey, R., Cherest, H., Blaiseau, P.-L. and Surdin-
transcriptional activator Met4 is required for the G(1)-S transition. EMBO Kerjan, Y. (1995). Met30p, a yeast transcriptional inhibitor that responds to
J. 19, 1613-1624. S-adenosylmethionine, is an essential protein with WD40 repeats. Mol. Cell.
Patton, E. E., Willems, A. R., Sa, D., Kuras, L., Thomas, D., Craig, K. L. Biol. 15, 6526-6534.
and Tyers, M. (1998). Cdc53 is a scaffold protein for multiple Thompson, J. D., Higgins, D. G. and Gibson, T. J. (1994). CLUSTAL W:
Cdc34/Skp1/F-box protein complexes that regulate cell division and improving the sensitivity of progressive multiple sequence alignment through
methionine biosynthesis in yeast. Genes Dev. 12, 692-705. sequence weighting, positions-specific gap penalties and weight matrix
Pines, J. (1995). Cyclins and cyclin-dependent kinases: a biochemical view. choice. Nucleic Acids Res. 22, 4673-4680.
Biochem. J. 308, 697-711. Vuillard, L., Nicholson, J. and Hay, R. T. (1999). A complex containing
Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning. betaTrCP recruits Cdc34 to catalyse ubiquitination of IκBα. FEBS Lett. 455,
Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press. 311-314.
Serrano, N. and O’Farrell, P. H. (1997). Limb morphogenesis: connections Winston, J. T., Koepp, D. M., Zhu, C., Elledge, S. J. and Harper, J. W.
between patterning and growth. Curr. Biol. 7, R186-R195. (1999a). A family of mammalian F-box proteins. Curr. Biol. 9, 1180-1182.
Shirane, M., Hatakeyama, S., Hattori, K. and Nakayama, K. (1999). Winston, J. T., Strack, P., Beer-Romero, P., Chu, C. Y., Elledge, S. J. and
Common pathway for the ubiquitination of IκBα, IκBβ, and IκBε mediated Harper, J. W. (1999b). The SCFβ-TRCP-ubiquitin ligase complex associates
by the F-box protein FWD1. J. Biol. Chem. 274, 28169-28174. specifically with phosphorylated destruction box motifs in IΚBα and β-
Sondek, J., Bohm, A., Lambright, D. G., Hamm, H. E. and Sigler, P. B. catenin and stimulates IΚBα ubiquitination in vitro. Genes Dev. 13, 270-283.
(1996). Crystal structure of a GA protein βγ dimer at 2.1 A resolution. Nature Wu, C. and Ghosh, S. (1999). beta-TrCP mediates the signal-induced
379, 369-374. ubiquitination of IκBβ. J. Biol. Chem. 274, 29591-29594.
Spencer, E., Jiang, J. and Chen, Z. J. (1999). Signal-induced ubiquitination Yaron, A., Hatzubai, A., Davis, M., Lavon, I., Amit, S., Manning, A. M.,
of IkappaBalpha by the F-box protein Slimb/beta-TrCP. Genes Dev. 13, 284- Andersen, J. S., Mann, M., Mercurio, F. and Ben-Neriah, Y. (1998).
294. Identification of the receptor component of the IκBα-ubiquitin ligase. Nature
Spevak, W., Keiper, B. D., Stratowa, C. and Castañón, M. J. (1993). 396, 590-594.
Saccharomyces cerevisiae cdc15 mutants arrested at a late stage in anaphase Yochem, J. and Byers, B. (1987). Structural comparison of the yeast cell
are rescued by Xenopus cDNAs encoding N-ras or a protein with β- division cycle gene CDC4 and a related pseudogene. J. Mol. Biol. 195, 233-
transducin repeats. Mol. Cell. Biol. 13, 4953-4966. 245.