genes required for glp-1 asymmetry in the early ......genes. of the blastomeres,the embryo was...

DEVELOPMENTAL BIOLOGY 181, 36–46 (1997)ARTICLE NO. DB968413

Genes Required for GLP-1 Asymmetry in the EarlyCaenorhabditis elegans Embryo

Sarah L. Crittenden,* David Rudel,† Joe Binder,* Thomas C. Evans,‡,1

and Judith Kimble*,†,‡,§,2

*Howard Hughes Medical Institute, Departments of †Biochemistry and §Medical Genetics,and ‡Laboratory of Molecular Biology, University of Wisconsin–Madison,Madison, Wisconsin 53706

The translation of maternal glp-1 mRNA is regulated both temporally and spatially in the early Caenorhabditis elegansembryo (T. C. Evans, S. L. Crittenden, V. Kodoyianni, and J. Kimble, Cell 77, 183–194, 1994). To investigate the controlof embryonic glp-1 expression, we have examined the distribution of GLP-1 protein in selected maternal effect mutantsthat affect pattern or fate in the early embryo. We find that mutants that disrupt anterior–posterior asymmetry in theearly embryo (par-1–par-6, emb-8, Par(q537)) disrupt the spatial but not temporal control of GLP-1 expression: GLP-1 isobserved at the normal stage of embryogenesis in par-like mutants; however, it is uniformly distributed. In contrast,mutants that alter blastomere identity (skn-1, pie-1, mex-1, apx-1) do not affect the normal GLP-1 pattern. We concludethat genes controlling the asymmetry of cellular components, including P granules, also control GLP-1 asymmetry in theearly embryo. The finding that mutants that disrupt anterior–posterior asymmetry translate GLP-1 in all blastomeressuggests that loss of embryonic asymmetry causes translational activation of GLP-1 in the posterior. q 1997 Academic Press

INTRODUCTION rior P1 blastomere; GLP-1 continues to be expressed in ABdescendants through the 28-cell stage.

Both temporal and spatial aspects of the maternal GLP-1The specification of cell fates during metazoan develop- pattern rely on translational control (Evans et al., 1994).

ment requires the precise temporal and spatial control of Reporter mRNAs carrying the glp-1 3* untranslated regionregulatory proteins. GLP-1, a transmembrane receptor re- (UTR) are expressed in a pattern similar to that of GLP-1.lated to Drosophila Notch, mediates several inductive in- Furthermore, a small deletion of only 61 nt from the middleteractions in the early Caenorhabditis elegans embryo of the glp-1 3*UTR eliminated spatial control but left tem-(Priess et al., 1987; Austin and Kimble, 1987; Yochem and poral regulation intact: the reporter was expressed at theGreenwald, 1989; Austin and Kimble, 1989; Roehl and Kim- correct stage but not in the right place (Evans et al., 1994).ble, 1993; Hutter and Schnabel, 1994, 1995; Mango et al., Because reporter RNA regulated by this mutant 3*UTR was1994a; Mello et al., 1994; Moskowitz et al., 1994; Shelton expressed ectopically in posterior P1 descendants, we pre-and Bowerman, 1996; see Fig. 1). The expression of GLP- dicted the existence of a trans-acting repressor localized in1 protein from maternal glp-1 mRNA is controlled both the P1 blastomere (Evans et al., 1994).temporally and spatially (Evans et al., 1994; see Fig. 1). We have begun to investigate the genetic control of GLP-Whereas glp-1 mRNA is uniformly distributed in both oo- 1 asymmetry in the early C. elegans embryo. Such genescytes and early embryos (1 to 4 cells), GLP-1 protein is first might regulate glp-1 mRNA specifically or they might beseen at the 2-cell stage, in the anterior AB but not the poste- required more generally for establishing embryonic polarity.

In either case, genes regulating the early embryonic patternof GLP-1 are predicted to act maternally because GLP-1asymmetry is visible early in embryogenesis, at the 2-cell1 Present address: University of Colorado Health Sciences Center,stage. Furthermore, mutations in such genes are predictedDenver, CO 80262.to cause lethality because release of glp-1 mRNA from con-2 To whom reprint requests should be addressed. Fax: (608) 265-

5820. E-mail: [email protected]. trol by its 3*UTR results in lethality (T.E., unpublished).

36

0012-1606/97 $25.00Copyright q 1997 by Academic Press

All rights of reproduction in any form reserved.

AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas

37Control of GLP-1 Asymmetry

(AB) and a smaller posterior daughter (P1) (Fig. 1). AB andP1 differ in their molecular components as well as in theirpatterns of division and the types of tissues they generate(Laufer et al., 1980; Sulston et al., 1983; Cowan and McIn-tosh, 1985; Priess and Thomson, 1987; see Guo andKemphues, 1996, for review).

The maternal effect par genes (for partitioning defective)play a central role in controlling the earliest embryonicasymmetries (Kemphues et al., 1988; Kirby et al., 1990;Morton et al., 1992; Cheng et al., 1995; Guo and Kemphues,1996). The phenotypes of par-1–4 include disruption of theasymmetry of the actin cytoskeleton, failure to localize Pgranules to the posterior, and loss of the normal asymmetriccleavage pattern. par-1 also disrupts the localization of aposterior determinant, SKN-1 (Bowerman et al., 1993).Three par genes have been characterized at the molecularlevel. PAR-1 is a serine/threonine kinase (Guo andKemphues, 1995), PAR-2 protein contains a putative ATP-binding site and a zinc binding domain of the ‘‘RING finger’’class (Levitan et al., 1994), and PAR-3 is a novel cytoplasmicprotein (Etemad-Moghadam et al., 1995). PAR-1, PAR-2,and PAR-3 are localized in the 1-cell embryo soon afterfertilization: both PAR-1 and PAR-2 localize to the posteriorcortex (Guo and Kemphues, 1995; Boyd et al., 1996), whilePAR-3 is found in the anterior cortex. PAR-3 is restricted

FIG. 1. GLP-1 translation is regulated both temporally and spa- to the anterior by PAR-2 (Cheng et al., 1995; Etemad-tially in the early C. elegans embryo. Thick lines represent GLP- Moghadam et al., 1995), and PAR-2 and PAR-3, in turn, are1 in the membrane, shading represents GLP-1 in the cytoplasm.

required for localization of PAR-1 to the posterior (Etemad-GLP-1 translation is repressed in oocytes and 1-cell embryos (tem-Moghadam et al., 1995). The asymmetric distribution ofporal repression). GLP-1 translation is repressed in P1 descendantsPAR proteins in the cell cortex of the zygote is thought tofrom the 2-cell stage (spatial repression). At the 4-cell stage, GLP-play an important role in polarizing the embryonic cytoskel-1 in ABp is thought to interact with APX-1 in P2 to promote ABp

fate (Mango et al., 1994a; Mello et al., 1994; Hutter and Schnabel, eton, which establishes multiple cellular and molecular1994; Moskowitz et al., 1994; Mickey et al., 1996; Shelton and asymmetries (Guo and Kemphues, 1996).Bowerman, 1996). At the 12-cell stage, GLP-1 in ABal descendants While the par genes play a broad role in specifying ante-and ABar descendants is thought to interact with an unknown rior–posterior asymmetry, a separate set of maternal genesligand in MS to promote development of the anterior pharynx (Gen- specify the identities of individual blastomeres. SKN-1, adreau et al., 1994; Hutter and Schnabel, 1994; Mango et al., 1994b; transcription factor found predominantly in the P1 blasto-Moskowitz et al., 1994). Arrows indicate inductive interactions.

mere at the 2-cell stage, is required for proper specificationof one of the daughters of P1, called EMS (Bowerman et al.,1992, 1993; Blackwell et al., 1994). Localization of SKN-1to P1 depends on both par-1 and mex-1 activity: in bothIn addition to GLP-1 asymmetry, several other asymmet-

ries are established soon after fertilization in the early C. mutants, SKN-1 is found ectopically in AB descendants(Bowerman et al., 1993; Mello et al., 1992). Activity of theelegans embryo. The point of sperm entry determines the

future posterior of the embryo (Goldstein and Hird, 1996). pie-1 gene is required for specification of the P2 blastomere(Mello et al., 1992, 1996). In addition, GLP-1 signaling isAfter fertilization, cytoplasmic rearrangements occur (Hird

and White, 1993; Hird, 1996) and P granules, putative germ- aberrant in both skn-1 and pie-1 mutants (Bowerman et al.,1992; Mango et al., 1994a; Shelton and Bowerman, 1996).line determinants, are localized to the posterior end of the

zygote (Strome and Wood, 1982, 1983). Disruption of the Finally, APX-1, a transmembrane protein and potential li-gand for GLP-1, is required for signaling from P2 to ABpactin cytoskeleton at this point has profound effects on the

distribution of P granules as well as on the ability of the (Mello et al., 1994; Mango et al., 1994a; Mickey et al., 1996;Shelton and Bowerman, 1996).treated embryo to cleave properly and to produce specific

tissue types (Hill and Strome, 1988, 1990). This period of To identify genes involved in regulating GLP-1 asymme-try, we analyzed GLP-1 in a broad spectrum of maternalcytoplasmic rearrangement preceding first cleavage is likely

to be crucial for the asymmetric distribution of many regu- effect lethal mutant embryos. GLP-1 is mislocalized in parmutants, but not in mutants with altered blastomere iden-latory proteins, including GLP-1. The first cleavage division

splits the fertilized zygote into a larger anterior daughter tity or defective GLP-1 signaling. In addition, we identified

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas

38 Crittenden et al.

was scored as mislocalized. If the level of GLP-1 was lower in twonew mutants with phenotypes similar to those of the parof the blastomeres, the embryo was scored as having a mislocalizedgenes.but still asymmetric distribution of GLP-1. For wild-type 8- to 12-cell embryos we detect levels of GLP-1 in E and MS that are dis-tinctly lower than in AB descendants. If an 8- to 15-cell mutantMATERIALS AND METHODS embryo had GLP-1 in all blastomeres at levels higher than thosein wild-type embryos, then it was scored as mislocalized. Embryos

C. elegans Strains and Maintenance were only included if the distribution of GLP-1 could unambigu-ously be scored as either mislocalized or wild-type.

C. elegans strains were maintained as described by Brenner (1974).Mutants and strains used in this study were as follows: LGI, emb-6(hc65), emb-12(g5), emb-20(g27), zyg-2(b10), par-6(zu170)unc-13(e450)/hT2; LGII, emb-21(g31), emb-23(g39), zyg-9(b244), emb- RESULTS27(g48), zyg-1(b1), rol-1(e91)mex-1(zu121)/mnC1, rol-1(e91)mex-1(zu120)/mnC1; LGIII, emb-5(hc61), emb-8(hc69), emb-13(h6), emb- To identify genes required for proper GLP-1 expression16(g19), emb-25(g45), emb-30(g53), emb-33(g60), emb-1(hc62), emb-

in the early embryo, we examined a variety of maternal7(hc66), par-2(lw32)unc-45(e286)/sC1, par-2(it5ts), lon-1(e185) par-effect lethal (Mel) mutant embryos. In the wild-type adult3(it71)/qC1, pie-1(zu154)unc-25(e156)/qC1; LGIV, emb-3(hc59), emb-germ line, GLP-1 is not detected in meiotic cells or matur-11(g1), emb-26(g47), emb-31(g55), par-5(it121) unc-22(e66)/DnT1,ing oocytes (Crittenden et al., 1994); in wild-type embryos,skn-1(zu67)/DnT1; LGV, emb-4(hc60), emb-18(g21), rol-4(sc8)par-

1(b274)/DnT1, dpy-21(e428)par-4(it33)/ozDf2, par-4(it57ts), apx-1(or- GLP-1 is not observed at the 1-cell stage and is found in AB3)dpy-11(e224)/DnT1. See text for references. Temperature sensitive descendants by the 4-cell stage (Evans et al., 1994). Somemutants were grown at 157C to maintain stocks; other strains were 2-cell embryos possess weak GLP-1 staining that also isgrown at 207C unless noted otherwise. restricted to AB, suggesting that translation of maternal glp-

1 mRNA first occurs at the 2-cell stage in AB blastomeres.After the 4-cell stage, strong GLP-1 staining persists in ABScreen for New Strict Maternal Effect Lethaldescendants until the 28-cell stage (Evans et al., 1994),Mutantswhile weak GLP-1 is detected in membranes between cer-

N2 hermaphrodites were mutagenized with ethylmethanesulfo- tain P1 descendants beginning at the 8-cell stage (betweennate as described (Brenner, 1974). From mutagenized L4 hermaphro- E and MS and between E and P3, data not shown; R. Feich-dites, F1 self-progeny were plated individually, and to identify ma- tinger and R. Schnabel, personal communication).ternal effect lethal mutants, F2 were then picked to individual To examine GLP-1 expression in Mel mutants, we stainedplates and grown at 257C. F2 that laid all or mostly dead eggs mutant embryos derived from homozygous mutant her-were crossed with wild-type males to identify strict maternal effect

maphrodites with anti-GLP-1 and anti-P granule antibodies.lethal (Mel) mutants (those not rescued by zygotic expression ofWe refer to such embryos as ‘‘x embryos’’ where x representsthe sperm’s genome). From 4734 haploid genomes screened, wethe gene. For example, a par-1 embryo refers to a homozy-obtained 182 independently isolated strict Mel mutants. Of these,gous par-1 mutant embryo derived from a homozygous mu-three mislocalized GLP-1 (see Results). Two of these, q514 andtant par-1 hermaphrodite.q516, mapped to linkage group V and failed to complement par-

4(it33). The third, q537, mapped to linkage group II and comple- To score spatial regulation of GLP-1, we examined GLP-ments mex-1(zu121). 1 in 4- to 15-cell embryos. Mutant embryos were scored as

having a wild-type GLP-1 pattern if the levels of GLP-1 inAB and P1 descendants were similar to those in wild-type.

Immunofluorescence Embryos were scored as having mislocalized GLP-1 if GLP-1 was present in P1 descendants at a level higher than thatTo detect GLP-1, we used a mixture of anti-EGFL, anti-LNG,in wild-type embryos (see Materials and Methods for moreand anti-ANK polyclonal antibodies (Crittenden et al., 1994); to

detect P granules, we used either OICID4 or K76 monoclonal anti- details). To score temporal regulation of GLP-1, we lookedbodies (Strome and Wood, 1983). Embryos and germ lines were for the presence or absence of GLP-1 in oocytes and 1-cellstained as described (Evans et al., 1994). Secondary antibodies were embryos, when it is not normally detected.labeled with FITC, LRSC, Cy3, or Cy5. Images were collected onBio-Rad MRC1000 and 1024 laser scanning confocal microscopes.To stain temperature-sensitive mutants, L4 hermaphrodites were Mutations in par-1 and par-4 Always Disruptgrown at 257C for 16–24 hr after which embryos were processed GLP-1 Asymmetryfor staining.

In par-1 embryos, the first cleavage division is symmetri-cal and P granules are not localized to P1 but instead are

Scoring Embryos found in both AB and P1 (Kemphues et al., 1988; Guo andKemphues, 1995). In par-4 embryos, the first cleavage oftenIn wild-type 4-cell embryos we do not detect GLP-1 in P1 descen-generates large AB and small P1 blastomeres as in wild-type,dants. If all blastomeres of a mutant 4-cell embryo had levels of

GLP-1 comparable to those found in wild-type AB blastomeres, it but P granules are distributed to both AB and P1 (Kemphues


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas


et al., 1988; Morton et al., 1992). We stained embryos homo- see also Kemphues et al., 1988), while 25% had P granulesonly in P1 descendants. We also noticed two par-3 embryoszygous for a strong par-1 allele, par-1(b274) (Kemphues et

al., 1988) or for a strong par-4 allele, par-4(it33). All par-1 with a very low level of GLP-1 in P1 descendants, but withP granules distributed equally in all four blastomeres (Fig.embryos (n Å 30) and par-4 embryos (n Å 30) mislocalized

GLP-1. 2G). Adult germ lines (n Å 14), oocytes (n Å 13), and 1-cellembryos (n Å 2) had normal distributions of GLP-1.In 100% of both par-1 and par-4 mutant embryos we de-

tected GLP-1 in all blastomeres (Fig. 2I). For par-1, all em- Thus, in par-2 and par-3 mutant embryos, GLP-1 is mislo-calized in less than 100% of the embryos, indicating thatbryos had approximately equal levels of GLP-1 expression

in all blastomeres (Fig. 2B). In embryos where P granules the activities of PAR-2 and PAR-3 are not essential for GLP-1 asymmetry.could be scored, they were distributed equally among the

blastomeres (n Å 4; Fig. 2B; see also Kemphues et al., 1988; We also stained weak alleles of two par genes that are lesswell characterized. In par-5(it121) mutant embryos, whichGuo and Kemphues, 1995). Adult germ lines (nÅ 9), oocytes

(n Å 9), and 1-cell embryos (n Å 8) had a normal GLP-1 have a phenotype similar to that of par-2 (K. Kemphues,personal communication), 14% (1/7) had mislocalized GLP-pattern. For par-4, 90% (27/30) had similar levels of GLP-1

in all blastomeres, while three 4-cell embryos had notice- 1. In par-6(zu170) mutant embryos, which have a phenotypesimilar to that of par-3 (Watts et al., 1996), 5% (1/20) hadably lower levels of GLP-1 in two blastomeres. P granules

stained weakly and were present in all blastomeres (n Å 4; mislocalized GLP-1.Fig. 2C; see also Kemphues et al., 1988). Adult germ lines(n Å 20), oocytes (n Å 12), and 1-cell embryos (n Å 5) had emb-8 Is Required for both GLP-1 and P Granulea normal GLP-1 pattern.

AsymmetryThus, both par-1 and par-4 are required for spatial butnot temporal regulation of GLP-1 asymmetry. The time of In an attempt to identify other Mel mutants affecting

GLP-1 asymmetry, we screened a collection of 24 previouslyonset of GLP-1 translation appears to be normal in thesemutant embryos, and the presence of GLP-1 in P1 descen- isolated temperature-sensitive strict Mel mutants (Wood et

al., 1980; Miwa et al., 1980; Cassada et al., 1981) with anti-dants is likely to reflect ectopic translation there.GLP-1 and anti-P granule antibodies. Of these only one,emb-8(hc69) mislocalized GLP-1 (see below). Four other

Mutations in par-2 and par-3 Often Disrupt GLP-1 mutants arrested at the 1-cell stage with no detectable GLP-Asymmetry 1 (emb-1(hc62), n Å 3; emb-7(hc66), n Å 3; emb-27(g48), n

Å 75; zyg-1(b1), n Å 49), indicating that temporal controlMany par-2 and par-3 embryos segregate P granules prop-erly to P1 at the first division; however, all of them exhibit of GLP-1 expression is normal. However, since embryos

arrested at the 1-cell stage, the asymmetric distribution ofsynchronous and symmetric first cleavage divisions. Thesecond cleavage planes differ between the two mutants GLP-1 could not be scored. Finally, 19 localized GLP-1 prop-

erly (emb-3(hc59), n Å 3; emb-4(hc60), n Å 1; emb-5(hc61),(Kemphues et al., 1988; Cheng et al., 1995).We stained embryos homozygous for the nonsense muta- n Å 16; emb-6(hc65), n Å 25; emb-11(g1), n Å 26; emb-

12(g5), nÅ 4; emb-13(g6), nÅ 6; emb-16(g19), nÅ 42; emb-tion par-2(lw32) (Levitan et al., 1994). Thirty-seven percent(12/32) mislocalized GLP-1 (Fig. 2F). Among embryos mislo- 18(g21), n Å 4; emb-20(g27), n Å 9; emb-21(g31), n Å 9;

emb-23(g39), n Å 8; emb-25(g45), n Å 22; emb-26(g47), ncalizing GLP-1, three 4-cell embryos had lower levels ofGLP-1 in P1 descendants. In embryos where P granules could Å 3; emb-30(g53), n Å 2; emb-31(g55), n Å 2; emb-33(g60),

n Å 10; zyg-2(b10), n Å 16; zyg-9(b244), n Å 13).be scored, 86% (12/14) had P granules localized to P1 descen-dants (Fig. 2E), while 14% (2/14) had P granules in AB as For emb-8(hc69) embryos, 68% (13/19) had a normal dis-

tribution of GLP-1 (Figs. 3A, 3B, and 3E). Thirty-two percentwell as P1 descendants (Fig. 2D; see also Kemphues et al.,1988). One embryo that had GLP-1 only in AB descendants (6/19) of the embryos stained for GLP-1 in both AB and P1

descendants at the 4- to 12-cell stages (Figs. 3C and 3E). Inalso had a large number of P granules in both AB and P1

descendants (Fig. 2D), raising the possibility that anterior all six of the 4-cell embryos that mislocalized GLP-1, thelevel of GLP-1 was higher in two of the four blastomeres,localization of GLP-1 may not depend on proper posterior

localization of P granules. Adult germ lines (nÅ 7), oocytes indicating that some asymmetry still exists (Fig. 3C). Inthose emb-8(hc69) embryos scored for both GLP-1 and P(nÅ 7), and 1-cell embryos (n Å 3) had normal distributions

of GLP-1. granules, 70% (7/10) had P granules only in P1 descendants(Fig. 3A) and 30% (3/10) had P granules in both AB and P1We stained embryos homozygous for the null mutant par-

3(it71) (Etemad-Moghadam et al., 1995). Seventy-seven per- descendants (Figs. 3B and 3E). Interestingly, in several 4-cell embryos, P granules were found either mostly in EMScent (23/30) mislocalized GLP-1 (Figs. 2G, 2H, and 2I).

Among embryos mislocalizing GLP-1, five 4-cell embryos (Fig. 3A) or in all blastomeres except EMS (Fig. 3B; alsoobserved by L. A. Khan and S. Siddiqui, personal communi-had lower levels of GLP-1 in P1 descendants (Figs. 2G and

2H). In embryos where P granules could be scored, 75% had cation). In two 4-cell embryos, GLP-1 was localized properlyto AB descendants even though P granules were present inP granules in both AB and P1 descendants (Figs. 2G and 2H;


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas


2


AID DB 8413 / 6x16h$8413 12-16-96 21:34:07 dbas


3

4


AID DB 8413 / 6x16h$8413 12-16-96 21:34:07 dbas


both AB and P1 descendants (Fig. 3B). All blastomeres of par-4 since they map to the approximate position of par-4on chromosome V and fail to complement par-4.many 4-cell emb-8 embryos had similar sizes and nuclei

appeared to be at the same stage of the cell cycle, suggesting The other mutant, q537, appears to be an allele of a newpar gene. Par(q537) maps to chromosome II; none of the 6that emb-8 embryos may have synchronous and symmetri-

cal early divisions like par mutants. known par genes is found on chromosome II. For Par(q537),39% (23/38; Fig. 3E) had GLP-1 in both AB and P1 descen-dants. Of these, 10 had less GLP-1 in P1 descendants than

Screen for New Genes That Control GLP-1 in AB descendants (Fig. 3D). In those embryos in which PAsymmetry granule staining was scored, 37% (10/27) had P granules in

both AB and P1 descendants. P granules were often clusteredTo identify new genes that control GLP-1 asymmetry, wein the middle of the embryo (Fig. 3D).screened for strict Mel mutants and stained them to exam-

The lack of mutants that specifically disrupt GLP-1 asym-ine GLP-1 expression. To determine whether mutants af-metry can be interpreted in several ways. One possibilityfected other embryonic asymmetries, we also stained withis that only one or a few genes can mutate to this specificanti-P granule antibodies and noted relative size and orien-phenotype and that our screen was not large enough to re-tation of blastomeres in the stained embryos.cover these mutants. Alternatively, such a regulator mayWe screened 4734 haploid genomes and isolated 182 stricthave a zygotic phenotype if it is also required for the propermaternal effect lethal mutants. GLP-1 was detectable in allregulation of GLP-1 (or other regulatory proteins) at other182 strict Mels. Therefore, we did not identify any candidatetimes or in different tissues.activators of GLP-1 translation. However, 3/182 mutants

expressed GLP-1 in both AB and P1 descendants at the 4-Genes That Affect Blastomere Identity and/or GLP-cell stage. All three new mutants also mislocalized P gran-1-Mediated Inductions Do Not Affect GLP-1 Levelules and appeared to have altered early cleavage patternsor Distribution(Fig. 3D and data not shown). GLP-1 was not detected in

oocytes or 1-cell embryos of any of these mutants. Two of A number of maternal effect genes affect the identity ofindividual blastomeres at the 4-cell stage. In skn-1 and pie-the three new mutants, q514 and q516 are new alleles of

FIG. 2. par gene activity is required for establishing GLP-1 asymmetry. Wild-type and par embryos were double stained with anti-GLP-1 and anti-P granule antibodies. GLP-1 is green; P granules are red. Embryos are approximately 40 mm long. (A) Wild-type 4-cell embryo.GLP-1 is detected in membranes and cytoplasms of ABa and ABp, but not in EMS or P2. Levels of GLP-1 staining are variable in bothwild-type and par mutants. (B) par-1(b274) 4-cell embryo. GLP-1 is present in membranes and cytoplasms of all four cells at levelscomparable to that of wild-type AB blastomeres. There does appear to be slightly less GLP-1 in the membrane between EMS and P2. Pgranules are present in the cytoplasm of all four blastomeres. (C) par-4(it33) 4-cell embryo. The pattern of GLP-1 is similar to that of par-1. The membrane between EMS and P2 is stained; however, it is not completely visible in this particular embryo. P granules are foundin all blastomeres. (D–F) Three different par-2(lw32) 4-cell embryos showing different phenotypes. The second cleavage plane in both ABand P1 in par-2 embryos is transverse, thus their blastomere arrangement is somewhat different from that of wild-type. (D) GLP-1 islocalized properly, but P granules occur in all blastomeres. (E) Both GLP-1 and P granules are asymmetric: P granules are present in bothpresumed P1 descendants. Therefore, P granules were segregated correctly at the 2-cell stage, but not thereafter. (F) GLP-1 is mislocalized.P granules were not stained in this embryo. (G and H) Two different par-3(it71) embryos. Both embryos have GLP-1 in posterior blastomeres,but the levels are not as high as those of par-1 or par-4. P granules are found in all blastomeres in both embryos. While P granules appearto be symmetrically distributed, GLP-1 is still somewhat asymmetric. (I) Summary of staining results with par embryos. Percentages of4- to 15-cell embryos mislocalizing GLP-1 and/or P granules are shown.FIG. 3. emb-8 and Par(q537) activities are required for both GLP-1 and P granule asymmetry. emb-8(hc69) and Par(q537) embryos weredouble stained with anti-GLP-1 and anti-P granule antibodies. GLP-1 is green; P granules are red. emb-8 embryos are fragile making itdifficult to preserve morphology; membranes are often less distinct than wild-type. Embryos are approximately 40 mm long. (A–C) Threedifferent emb-8(hc69) embryos are shown, each representing a distinct phenotype. (A) Both GLP-1 and P granules are asymmetricallydistributed, though P granules are detected only in EMS. (B) GLP-1 is asymmetric, but P granules are found in both AB and P1 descendants.In contrast to A, P granules are found in all blastomeres except EMS. (C) GLP-1 occurs in all membranes; P granules were not stained.(D) A Par(q537) embryo. GLP-1 is present in all membranes; P granules are present in three of four cells and are clustered together in oneregion. (E) Summary of results with emb-8 and Par(q537) embryos. Percentages of 4- to 12-cell embryos mislocalizing GLP-1 and/or Pgranules are shown.FIG. 4. GLP-1 asymmetry is independent of signaling and posterior blastomere identity. apx-1 (A), pie-1 (B), skn-1 (C), and mex-1 (D)mutant 4-cell embryos were double stained with anti-GLP-1 and anti-P granule antibodies. GLP-1 is green; P granules are red. In all cases,GLP-1 is similar to wild-type; it is detected in the membranes and cytoplasm of AB descendants and is not detected in either the membranesof cytoplasms of P1 descendants. Variability in brightness of the cytoplasm occurs in wild-type embryos and does not seem to be a featureof any particular mutant. P granules are distributed normally in all mutants except mex-1, where they are found in both P1 descendants.Embryos are approximately 40 mm long.


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas


1 mutants, fates of the two P1 descendants, EMS and P2, asymmetries. par-1 and par-4 which strongly affect GLP-1and P granule asymmetry only weakly affect cytoplasmicare altered; in mex-1 mutants, AB fates are altered at least

in part due to mislocalization of SKN-1. In addition, P gran- streaming, the position of pronuclear fusion, and the asym-metric distribution of actin in the 1-cell embryo (Kemphuesules are not segregated normally in mex-1 embryos (Mello

et al., 1992; Schnabel et al., 1996). We stained 4- to 12-cell et al., 1988; Kirby et al., 1990). In addition, the first divisionin par-4 mutants is asymmetric (Kemphues et al., 1988;skn-1(zu67) embryos (nÅ 18; Fig. 4A), pie-1(zu154) embryos

(n Å 27; Fig. 4B), and embryos from two alleles of mex-1, Morton et al., 1992), indicating that proper placement ofthe first cleavage plane is not sufficient for proper GLP-1mex-1(zu121) (n Å 17; Fig. 4C), and mex-1(zu120) (n Å 18;

data not shown) and found that GLP-1 staining was similar asymmetry. In contrast, par-2 and par-3, which have aweaker effect on GLP-1 and P granule asymmetry, haveto wild-type. Thus, the correct fates of AB and P1 descen-

dants are not essential for asymmetric GLP-1 expression. strong effects on the other early embryonic asymmetriesthat par-1 and par-4 only affect weakly (Kemphues et al.,The apx-1 gene encodes a signaling ligand for GLP-1 that

is used in P2 to induce ABp at the 4-cell stage (Mello et al., 1988; Kirby et al., 1990). Thus, par-1 and par-4 play a centralrole in controlling the asymmetric distribution of proteins,1994; Mango et al., 1994a; Mickey et al., 1996; Shelton and

Bowerman, 1996). To ask whether the level of GLP-1 in AB including GLP-1, within the early embryo. In contrast, par-2 and par-3 appear to play a central role in cytoskeletaldescendants is affected by the absence of its signaling li-

gand, we scored apx-1(or3) (n Å 17; Fig. 4D) 4- to 12-cell asymmetry.A pathway for par gene activity has been proposed inembryos. GLP-1 staining was similar to wild-type at all

stages. which PAR-2 and PAR-3 are required for the proper localiza-tion of PAR-1 (Etemad-Moghadam et al., 1995). However,in contrast to PAR-1, PAR-2 and PAR-3 are not absolutelyrequired for GLP-1 asymmetry. This dependence on PAR-1DISCUSSIONsuggests that sufficient PAR-1 activity remains in manypar-2 and par-3 mutant embryos to allow proper regulationEvans et al. (1994) showed that maternal glp-1 mRNA is

translationally regulated by elements in its 3*UTR to of GLP-1 asymmetry. Consistent with this, it has been pro-posed that PAR-1 localization may not be absolutely re-achieve an asymmetric expression of GLP-1 protein in the

early embryo. In this paper, we have begun to identify genes quired for its activity (Boyd et al., 1996).required for GLP-1 asymmetry. Specifically, we have identi-fied eight genes, par-1–par-6, emb-8, and Par(q537), that

Temporal Regulation Can Be Separated frominfluence GLP-1 asymmetry in the early embryo and haveSpatial Regulation in the Early Embryoshown that many others do not affect GLP-1 expression.

Our results have a number of implications for the control of The glp-1 3*UTR mediates both temporal and spatial reg-ulation of GLP-1 translation in the early embryo (Evans etGLP-1 asymmetry in the early embryo, which are discussed

below. al., 1994). Deletion analysis of the 3*UTR raised the possi-bility that temporal and spatial regulation were mediatedby different regions of the 3*UTR, possibly with separate

The Initial Establishment of Polarity Is Linked regulators binding each element. Consistent with this idea,with the Establishment of GLP-1 Asymmetry GLP-1 appears to be translated at the normal time in par

mutants; however, its spatial asymmetry is lost. Alterna-We have found that the activity of genes regulating manyaspects of anterior–posterior asymmetry in the early em- tively, a single element in the 3*UTR may direct both tem-

poral and spatial regulation, but temporal and spatial regula-bryo, par-1–6, emb-8, and Par(q537), are required for theproper spatial but not temporal regulation of maternal GLP- tion may be controlled by distinct factors.1 expression. In par mutant embryos, we observe extra GLP-1 expression in posterior blastomeres, indicating that GLP-

Translational Regulation of glp-1 in the Early1 can be translated in all blastomeres in the absence of parEmbryo: Repression or Activation?activity. In contrast, we have found that GLP-1 distribution

in the early embryo is independent of both GLP-1-mediated Two simple ways to create GLP-1 asymmetry are (1) pres-ence of a translational activator in the anterior of the em-signaling and the proper specification of anterior and poste-

rior blastomere identity. bryo, or (2) presence of a translational repressor in the poste-rior of the embryo (see Fig. 5). In addition, temporal controlof GLP-1 translation is necessary to ensure that GLP-1 pro-

Several Individual Aspects of Embryonic Polarity tein is not synthesized in a 1-cell embryo and subsequentlyCan Be Uncoupled from the Establishment inherited by all embryonic blastomeres. The role of the parof GLP-1 Asymmetry genes in GLP-1 translational control is not understood;

however, because par gene activity is required for a widePhenotypic differences among the par mutants indicatethat GLP-1 asymmetry is not coupled to all early embryonic range of embryonic asymmetries, it is likely that the main


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas


ules. We have also observed several embryos in which GLP-1 is properly localized, but P granules are found in both ABand P1 descendants. These results raise the possibility thatproper P granule localization may not be absolutely requiredfor proper regulation of GLP-1 asymmetry.

Signaling Is Not Required for GLP-1 Expression inthe Early Embryo

Positive feedback does not appear to play a role in main-taining GLP-1 expression levels in the early embryo: neitheractive ligand nor active GLP-1 itself (S.C., unpublished) isrequired for proper expression. In contrast to the embryo,glp-1 activity is required for continued GLP-1 expression inthe germ line (S.C., unpublished), and positive autoregula-tion has been suggested (Kodoyianni et al., 1992; Chris-tensen et al., 1996; Berry and Schedl, in press). Positive

FIG. 5. Two models for how par genes could affect GLP-1 asym- feedback has also been proposed to maintain expression ofmetry. Shading indicates presence of GLP-1; / indicates presence a C. elegans GLP-1 homolog, LIN-12, in the larval somaticof an activator, 0 indicates presence of repressor. 0 with an x gonad (Seydoux and Greenwald, 1989; Wilkinson et al.,through it indicates an inactive repressor. See discussion for more 1994). While positive feedback does not appear to controldetails. (A) In wild-type embryos, an activator of GLP-1 translation levels of GLP-1 in 2- to 28-cell embryos, activity of maternalmay be restricted to the anterior blastomere, AB. In par mutants

GLP-1 appears to activate LIN-12 expression in AB descen-this activator would be found in both AB and P1. (B) In wild-typedants later in embryonic development (Moskowitz andembryos a repressor of GLP-1 translation may be restricted to theRothman, 1996).posterior blastomere, P1. In par mutants this repressor would be

found in both AB and P1; however, it must not be active since GLP-1 is present in all blastomeres.

Blastomere Identity Is Established after GLP-1Asymmetry

GLP-1 asymmetry is established early and requires therequirement for par activity is to properly localize a regula- activity of the par genes. However, genes that establish blas-tor of GLP-1 translation (Fig. 5). They might localize a trans- tomere identity do not affect GLP-1 asymmetry. For exam-lational activator (or its RNA) to the anterior (Fig. 5A) or a ple, the presence of GLP-1 in mex-1(zu120) AB blastomerestranslational repressor (or its RNA) to the posterior (Fig. indicates that even though these blastomeres contain SKN-5B). In the latter case, the repressor would have to be inac- 1 (Bowerman et al., 1993) and produce tissue types typicaltive when not localized properly since GLP-1 is present in of MS (Mello et al., 1992), they still regulate GLP-1 in anall blastomeres in par mutant embryos. This situation oc- AB-specific manner. In addition, SKN-1 is neither necessarycurs in Drosophila where Nanos, a translational repressor nor sufficient to repress GLP-1 expression. Therefore, SKN-of hunchback mRNA, is not translated unless its mRNA 1 specifies EMS identity after GLP-1 asymmetry is estab-is localized properly (Gavis and Lehmann, 1994). lished and establishment of blastomere identity is not

closely linked to regulation of GLP-1 asymmetry in theearly embryo.Is a Repressor of glp-1 Translation Associated with

P Granules?Is There a Separate Pathway for Regulation ofIn Drosophila, the proper formation and posterior local-GLP-1 Asymmetry?ization of polar granules are required for the posterior re-

pression of hunchback translation and the determination The mechanisms for generating the asymmetric localiza-tion of certain regulators in the early embryo can be sepa-of abdominal and germ cell fates (see Lehmann, 1995, for

review). By analogy, it is possible that factors required for rated genetically. Thus, mex-1(zu120) affects SKN-1 but notGLP-1 asymmetry (Bowerman et al., 1993; this paper), andthe asymmetric expression of embryonic proteins are asso-

ciated with P granules in C. elegans. We have found that mex-3 affects PAL-1 asymmetry but not GLP-1, SKN-1, orAPX-1 asymmetries (B. Draper and J. Priess, 1996; C. Hunterin many par mutant embryos there is a correlation between

P granule mislocalization and GLP-1 mislocalization. The and C. Kenyon, 1996). Furthermore, par-2–6 may not affectGLP-1 and SKN-1 asymmetry to the same extent (B. Bow-presence of P granules in the same blastomeres as GLP-1

suggests that, at least in par mutant embryos, an active erman, personal communication; this paper). Thus, thereare both global regulators of embryonic asymmetry, suchrepressor of GLP-1 translation is not associated with P gran-


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas


ing in Caenorhabditis elegans, is homologous to human CBF1as par-1, and regulators affecting a more restricted group ofand Drosophila Su(H). Development 122, 1373–1383.asymmetries, such as mex-3. Specific regulators of GLP-1

Cowan, A. E., and McIntosh, J. R. (1985). Mapping the distributionasymmetry have not yet been identified, and their existenceof differentiation potential for intestine, muscle, and hypodermisremains an open question for future investigation.during early development in Caenorhabditis elegans. Cell 41,923–932.

Crittenden, S. L., Troemel, E. R., Evans, T. C., and Kimble, J. (1994).ACKNOWLEDGMENTS GLP-1 is localized to the mitotic region of the C. elegans germ

line. Development 120, 2901–2911.We thank Deena Schuster, Phil Balandyk, Kim Oas, and Eileen Draper, B. W., Mello, C. M., Bowerman, B., Hardin, J., and Priess,

Durkin for technical assistance, Ken Kemphues, Bruce Bowerman, J. R. (1996). MEX-3 is a KH domain protein that regulates blasto-Susana Guedes, Craig Mello, and past and present members of the mere identity in early C. elegans embryos. Cell, in press.Kimble lab for helpful discussions, Lisa Kadyk, Maria Gallegos, Etemad-Moghadam, B., Guo, S., and Kemphues, K. J. (1995). Asym-and the reviewers for comments on the manuscript, and Bruce metrically distributed PAR-3 protein contributes to cell polarityBowerman, Laura Berry, Tim Schedl, Richard Feichtinger, Ralf and spindle alignment in early C. elegans embryos. Cell 83, 743–Schnabel, L. A. Khan, and S. Siddiqui for sharing unpublished data. 752.We also thank Ken Kemphues, Lynn Boyd, Bruce Bowerman, and Evans, T. C., Crittenden, S. L., Kodoyianni, V., and Kimble, J.especially the Caenorhabditis Genetics Center for sending strains. (1994). Translational control of maternal glp-1 mRNA establishesAnti-P granule monoclonal antibodies K76 and OICID4 were pro- an asymmetry in the C. elegans embryo. Cell 77, 183–194.vided by the Developmental Studies Hybridoma Bank maintained Gavis, E. R., and Lehmann, R. (1994). Localization of nanos RNAby a contract from NICHD (N01-HD-2-3144). This work was sup- controls embryonic polarity. Nature 369, 315–318.ported by grants from NIH and NSF to J.K. and Wisconsin Alumni Gendreau, S. B., Moskowitz, I. P. G., Terns, R. M., and Rothman,Research Foundation and NIH Molecular Biosciences Training J. H. (1994). The potential to differentiate epidermis is unequallygrant fellowships for D.R. J.K. is an investigator with the Howard distributed in the AB lineage during early embryonic develop-Hughes Medical Institute. ment in C. elegans. Dev. Biol. 166, 770–781.

Goldstein, B., and Hird, S. N. (1996). Specification of the antero-posterior axis in Caenorhabditis elegans. Development 122,1467–1474.REFERENCES

Guo, S., and Kemphues, K. J. (1995). par-1, a gene required for estab-lishing polarity in C. elegans embryos, encodes a putative Ser/Austin, J., and Kimble, J. (1987). glp-1 is required in the germ lineThr kinase that is asymmetrically distributed. Cell 81, 611–620.for regulation of the decision between mitosis and meiosis in C.

Guo, S., and Kemphues, K. J. (1996). Molecular genetics of asym-elegans. Cell 51, 589–599.metric cleavage in the early Caenorhabditis elegans embryo.Austin, J., and Kimble, J. (1989). Transcript analysis of glp-1 andCurr. Opin. Genet. Dev. 6, 408–415.lin-12 homologous genes required for cell interactions during

Hill, D. P., and Strome, S. (1988). An analysis of the role of micro-development of C. elegans. Cell 58, 565–571.filaments in the establishment and maintenance of asymmetryBerry, L. W., Westlund, B., and Schedl, T. (1996). Germline tumorin Caenorhabditis elegans zygotes. Dev. Biol. 125, 75 –84.formation caused by activation of glp-1, a member of the Notch

Hill, D. P., and Strome, S. (1990). Brief cytochalasin-induced disrup-family of receptors. Development, in press.tion of microfilaments during a critical interval in 1-cell C. eleg-Blackwell, T. K., Bowerman, B., Priess, J. R., and Weintraub, H.ans embryos alters the partitioning of developmental instruc-(1994). Formation of a monomeric DNA binding domain by Skn-tions to the 2-cell embryo. Development 108, 159–172.1 bZIP and homeodomain elements. Science 266, 621–628.

Hird, S. (1996). Cortical actin movements during the first cell cycleBowerman, B., Draper, B. W., Mello, C. C., and Priess, J. F. (1993).of the Caenorhabditis elegans embryo. J. Cell Sci. 109, 525–533.The maternal gene skn-1 encodes a protein that is distributed

Hird, S. N., and White, J. G. (1993). Cortical and cytoplasmic flowunequally in early C. elegans embryos. Cell 74, 443–452.polarity in early embryonic cells of Caenorhabditis elegans. J.Bowerman, B., Eaton, B. A., and Priess, J. R. (1992). skn-1, a mater-Cell Biol. 121, 1343–1355.nally expressed gene required to specify the fate of ventral blasto-

Hunter, C. P., and Kenyon, C. (1996). Spatial and temporal controlsmeres in the early C. elegans embryo. Cell 68, 1061–1075.target pal-I blastomere specification activity to a single blasto-Boyd, L., Guo, S., Levitan, D., Stinchcomb, D. T., and Kemphues,mere lineage in C. elegans embryos. Cell 87, 217–226.K. J. (1996). PAR-2 is asymmetrically distributed and promotes

Hutter, H., and Schnabel, R. (1994). glp-1 and inductions establish-association of P granules and PAR-1 with the cortex in C. elegansing embryonic axes in C. elegans. Development 120, 2051–2064.embryos. Development 122, 3075–3084.

Hutter, H., and Schnabel, R. (1995). Establishment of left-rightBrenner, S. (1974). The genetics of Caenorhabditis elegans. Genet-asymmetry in the Caenorhabditis elegans embryo: A multistepics 77, 71–94.process involving a series of inductive events. Development 121,Cassada, R., Isnenghi, E., Culotti, M., and von Ehrenstein, G. (1981).3417–3424.Genetic analysis of temperature-sensitive embryogenesis mu-

Kemphues, K. J., Priess, J. R., Morton, D. G., and Cheng, N. (1988).tants in Caenorhabditis elegans. Dev. Biol. 84, 193 –205.Identification of genes required for cytoplasmic localization ofCheng, N. N., Kirby, C. M., and Kemphues, K. J. (1995). Control ofearly C. elegans embryos. Cell 52, 311–320.cleavage spindle orientation in Caenorhabditis elegans: The role

Kirby, C., Kusch, M., and Kemphues, K. (1990). Mutations in theof the genes par-2 and par-3. Genetics 139, 549–559.par genes of Caenorhabditis elegans affect cytoplasmic reorgani-Christensen, S., Kodoyianni, V., Bosenberg, M., Friedman, L., and

Kimble, J. (1996). lag-1, a gene required for lin-12 and glp-1 signal- zation during the first cell cycle. Dev. Biol. 142, 203–215.


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas


Kodoyianni, V., Maine, E. M., and Kimble, J. (1992). Molecular basis required zygotically for early embryonic cellular interactions andare regulated by maternal GLP-1 signalling in Caenorhabditisof loss-of-function mutations in the glp-1 gene of Caenorhabditis

elegans. Mol. Biol. Cell 3, 1199–1213. elegans. Development 122, 4105–4117.Priess, J. R., Schnabel, H., and Schnabel, R. (1987). The glp-1 locusLaufer, J. S., Bazzicalupo, P., and Wood, W. B. (1980). Segregation

and cellular interactions in early C. elegans embryos. Cell 51,of developmental potential in early embryos of Caenorhabditis601–611.elegans. Cell 19, 569–577.

Priess, J. R., and Thomson, J. N. (1987). Cellular interactions inLehmann, R. (1995). Establishment of embryonic polarity duringearly C. elegans embryos. Cell 48, 241–250.Drosophila oogenesis. Semin. Dev. Biol. 6, 25–38.

Roehl, H., and Kimble, J. (1993). Control of cell fate in C. elegansLevitan, D. J., Boyd, L., Mello, C. C., Kemphues, K. J., and Stinch-by a GLP-1 peptide consisting primarily of ankyrin repeats. Na-comb, D. T. (1994). par-2, a gene required for blastomere asym-ture 364, 632–635.metry in Caenorhabditis elegans, encodes zinc-finger and ATP-

Schnabel, R., Weigner, C., Hutter, H., Feichtinger, R., and Schnabel,binding motifs. Proc. Natl. Acad. Sci. USA 91, 6108–6112.H. (1996). mex-1 and the general partitioning of cell fate in theMango, S. E., Thorpe, C. J., Martin, P. R., Chamberlain, S. H., andearly C. elegans embryo. Mech. Dev. 54, 133–147.Bowerman, B. (1994a). Two maternal genes, apx-1 and pie-1, are

Seydoux, G., and Greenwald, I. (1989). Cell autonomy of lin-12required to distinguish the fates of equivalent blastomeres in thefunction in a cell fate decision in C. elegans. Cell 57, 1237–1245.early Caenorhabditis elegans embryo. Development 120, 2305–

Shelton, C. A., and Bowerman, B. (1996). Time-dependent re-2315.sponses to glp-1 mediated inductions in early C. elegans embryos.Mango, S. E., Lambie, E. J., and Kimble, J. (1994b). The pha-4 geneDevelopment 122, 2043–2050.is required to generate the pharyngeal primordium of Caenorhab-

Strome, S., and Wood, W. B. (1982). Immunofluorescence visualiza-ditis elegans. Development 120, 3019–3031.tion of germ-line-specific cytoplasmic granules in embryos, lar-Mello, C. C., Draper, B. W., Krause, M., Weintraub, H., and Preiss,vae, and adults of Caenorhabditis elegans. Proc. Natl. Acad. Sci.J. R. (1992). The pie-1 and mex-1 genes and maternal control ofUSA 79, 1558–1562.blastomere identity in early C.elegans embryos. Cell 70, 163–

Strome, S., and Wood, W. B. (1983). Generation of asymmetry and176.segregation of germ-line granules in early C. elegans embryos.Mello, C. C., Draper, B. W., and Priess, J. R. (1994). The maternalCell 35, 15–25.genes apx-1 and glp-1 and establishment of dorsal–ventral polar-

Sulston, J. E., Schierenberg, E., White, J. G., and Thomson, J. N.ity in the early C. elegans embryo. Cell 77, 95 –106.(1983). The embryonic lineage of the nematode CaenorhabditisMello, C. C., Schubert, C., Draper, B., Zhang, W., Lobel, R., andelegans. Dev. Biol. 100, 64 –119.Priess, J. R. (1996). The PIE-1 protein and germline specification

Watts, J. L., Etemad-Moghadam, B., Guo, S., Boyd, L., Draper, B. W.,in C. elegans embryos. Nature 382, 710–712.Mello, C. C., Priess, J. R., and Kemphues, K. K. (1996). par-6, aMickey, K. M., Mello, C. C., Montgomery, M. K., Fire, A., andgene involved in the establishment of asymmetry in early C.Priess, J. R. (1996). An inductive interaction in 4-cell stage C.elegans embryos, mediates the asymmetric localization of PAR-elegans embryos involves APX-1 expression in the signaling cell.3. Development 122, 3133–3140.Development 122, 1791–1798.

Wilkinson, H. A., Fitzgerald, K., and Greenwald, I. (1994). Recipro-Miwa, J., Schierenberg, E., Miwa, S., and von Ehrenstein, G. (1980).cal changes in expression of the receptor LIN-12 and its ligandGenetics and mode of expression of temperature-sensitive muta-LAG-2 prior to commitment in a C. elegans cell fate decision.tions arresting embryonic development in Caenorhabditis eleg-Cell 79, 1187–1198.ans. Dev. Biol. 76, 160–174.

Wood, W. B., Hecht, R., Carr, S., Vanderslice, R., Wolf, N., andMorton, D. G., Roos, J. M., and Kemphues, K. J. (1992). par-4, aHirsh, D. (1980). Parental effects and phenotypic characterizationgene required for cytoplasmic localization and determination ofof mutations that affect early development in Caenorhabditisspecific cell types in Caenorhabditis elegans embryogenesis. Ge-elegans. Dev. Biol. 74, 446–469.netics 130, 771 –790.

Yochem, J., and Greenwald, I. (1989). glp-1 and lin-12, genes impli-Moskowitz, I. P. G., Gendreau, S. B., and Rothman, J. H. (1994).cated in distinct cell–cell interactions in C. elegans, encode simi-Combinatorial specification of blastomere identity by glp-1 de-lar transmembrane proteins. Cell 58, 553–563.pendent cellular interactions in the nematode Caenorhabditis

elegans. Development 120, 3325–3338. Received for publication August 23, 1996Accepted September 23, 1996Moskowitz, I. P. G., and Rothman, J. (1996). LIN-12 and GLP-1 are


AID DB 8413 / 6x16h$$541 12-16-96 21:34:07 dbas

genes required for glp-1 asymmetry in the early ......genes. of the blastomeres,the embryo was...

Documents