Early embryogenesis in Caenorhabditis elegans is characterized by a series of unequal cleavages that mark the stepwise separation of somatic and germ lineages. We have developed an in situ hybridization protocol to examine the localization of specific maternal and embryonically transcribed messenger RNAs during these early cleavages. We detected three classes of maternal RNAs: RNAs that are maintained in all cells, RNAs that are maintained in germline cells but are lost from somatic cells, and a population of RNAs that are associated with the germlinespecific P granules. We observed embryonically transcribed RNAs in somatic cells as early as the 4-cell stage. These transcripts were not detected in germline cells. These observations suggest that mechanisms which distinguish between soma and germline cause asymmetries in mRNA stability and transcription within the first few cleavages of C. elegans embryogenesis.

The distinction between soma and germline is essential to the development of multicellular organisms (e.g. Denis and Lacroix, 1993). In the nematode Caenorhabditis elegans, the separation of somatic and germ lineages occurs during the first four cleavages (Fig. 1; Sulston et al., 1983). The zygote, P0, can be considered the first germline cell (e.g. Schierenberg and Strome, 1992). The first cleavage divides P0 into a larger somatic blastomere AB and a smaller germline cell P1. The AB blastomere divides equally to generate somatic descendants that divide in synchrony (AB lineage). In contrast, the P1 germline cell divides unequally to give rise to the somatic blastomere EMS and the germline cell P2. Unequal germline cell divisions continue until the five somatic founder lineages (AB, MS, E, C and D) and the primordial germ cell P4 are formed (24-cell stage, Fig. 1).

Fig. 1.

Early cell lineage. P0 is the zygote and can be thought of as the first germ cell. Unequal cleavages of the germline cells (P0-P3) lead to five somatic founder cells (AB, MS, E, C, D) and the germline progenitor cell P4. Thickened lines indicate the segregation of P granules.

Fig. 1.

Early cell lineage. P0 is the zygote and can be thought of as the first germ cell. Unequal cleavages of the germline cells (P0-P3) lead to five somatic founder cells (AB, MS, E, C, D) and the germline progenitor cell P4. Thickened lines indicate the segregation of P granules.

The fates of somatic blastomeres are established by mechanisms involving both cell-cell interactions and the asymmetric activity of maternally encoded components (reviewed in Wood and Edgar, 1994). The mechanisms that distinguish somatic cells from germline cells are less understood. Cytoplasm centrifugation and extrusion experiments have suggested that a posteriorly located ‘germline factor’ is sufficient for certain aspects of germline identity in nematodes (Boveri, 1910; Schierenberg, 1988). Direct evidence for asymmetrically localized factors first came from the analysis of the germline-specific P granules. P granules are cytoplasmic particles of unknown composition which can be visualized by immunofluorescence (Strome and Wood, 1982) or by electron microscopy (Wolf et al., 1983). P granules are uniformly distributed at the time of fertilization, but become localized to the posterior region of the 1-cell embryo prior to the first cleavage. During this and subsequent divisions, the P granules remain associated with germline cells (Strome and Wood, 1982; Fig. 1). Although this behavior suggests that P granules may function in some aspect of germline development, their function is not yet known. Nevertheless, the existence of such asymmetrically localized particles suggests that the distinction between soma and germline may depend on asymmetric partitioning of germline and/or soma-specifying components.

In this study, we have investigated the consequences of early asymmetry on the distribution of maternal and embryonically transcribed mRNAs. In particular, we have asked two questions. (1) Do specific maternal mRNAs exhibit asymmetric distributions in the early embryo? (2) When do asymmetries in embryonic transcription first appear? To address these questions, we have developed a high resolution in situ hybridization protocol to visualize mRNAs in whole-mount embryos. We have used this protocol to determine the expression patterns of 21 genes and to examine the general distribution of poly(A)+ RNAs during embryogenesis. We find that several maternal and embryonically transcribed mRNAs exhibit asymmetric distributions during early embryonic cleavages.

Strains

We used wild-type C. elegans (N2) and the following mutant strains. SP 756: mnDf90 / mnC1 [dpy-10(e128) unc-52(e444)] II. mnDf90 deletes the gene cey-1 (Sigurdson et al, 1984; Jantsch-Plunger, 1993); mnC1 [dpy-10(e128) unc-52(e444)] is a balancer chromosome for LGII (Herman, 1978).

PD 9593: mDf4 / DnT1 [unc-?(n754) let-?] (IV; V). mDf4 deletes ama-1 (ama-1 encodes the large subunit of RNA polymerase II; Bird and Riddle, 1989). DnT1 [unc-?(n754) let-?] is a reciprocal translocation (LG IV-V) which carries a dominant Unc mutation and a recessive lethal mutation (Ferguson and Horvitz, 1985).

BW 163: ctDf1 / DnT1 [unc-?(n754) let-?] (IV; V). ctDf1 deletes dpy-30 (Hsu and Meyer, 1994; Manser and Wood, 1990).

MT 5734: nDf41 / DnT1[unc-?(n754) let-?] (IV; V): nDf41 deletes skn-1 (Bowerman et al., 1992, 1993).

KK 300: par-4 (it57ts). par-4 (it57ts) is a temperature-sensitive maternal effect mutation which causes embryos to divide abnormally at 25°C (Morton et al, 1992).

PD 182: glp-1(e2141ts); ccIs182. ccIs182 is a chromosomally integrated transgenic array containing the glp-1 gene (Yochem and Greenwald, 1989), and a glp-1::lacZ translational fusion (pPD26.99, A. F. unpublished results).

PD 9556: ccIs9556. ccIs9556 is a chromosomally integrated transgenic array containing rol-6(su1006dom) (Mello et al., 1991) and a pes-10::lacZ translational fusion (pGS15.24; G. S. unpublished results).

Clones

The following clones were used to generate gene-specific probes. cm12b2 (act-1), cm04g2 (ncc-1; cdc2 related), cm04h4 (tubulin α-related), cm01d4 (tubulin β-related), cm12d4 (EF1α-related), cm01h11 (HSP70A-related), cm04g7 (HSP90-related), cm14g5 (histone H1-related) (Waterston et al., 1992);

pBSZ (lacZ; Nipam Patel, personal communication); pPCA (crf-2; Ann Sluder, personal communication); pUZ1200 (hlh-1; Krause et al., 1990);

SK-/b7 (lin-19; Edward Kipreos and Edward Hedgecock, personal communication);

JP600 and JP601 (skn-1; Bowerman et al., 1992);

MC334 (cey-1) and MC422 (cey-2) (cey: C. elegans Y box homolog; Jantsch-Plunger, 1993);

pDH38 (dpy-30; Dave Hsu and Barbara Meyer, personal communication; Hsu, 1992);

pDB33 (ama-1; Bird and Riddle, 1989); pCeIF-1800 (CeIF, eIF4-A-related; Roussell and Bennett, 1992); pPD19.60 (unc-54), pPD53.02 (glp-1), pPD33.24 (RP21C-related) and pGS17.09 (pes-10; Seydoux and Fire, unpublished data; pes: pattern expression site).

Probe synthesis

cDNA-derived probes

Digoxigenin (DIG)-labeled single-stranded DNA probes were synthesized by multiple cycles of primer extension in the presence of DIG-dUTP, using genomic or cDNA clones as templates (Patel and Goodman, 1992). Probes were used at a concentration of 0.5-2.5 μg/ml in hybridization buffer. Sense orientation control probes were tested only for the following genes: pes-10, cey-1, cey-2, skn-1 and ama-1. No signal was detected with any of these control probes (data not shown).

Oligonucleotide probes

We used the following single stranded oligonucleotide probes: oligo(dT) (30mer), oligo(dA) (30 mer), anti-SL1 (CTCAAACTTGGGTAAT-TAAACC), anti-SL1 precursor (GCTAACGCCAAATTTCTTTGGG).

These probes were end-labeled with terminal transferase and digoxigenin-ddUTP (Boehringer Mannheim) and used at a concentration of 0.5 μg/ml in hybridization buffer.

Permeabilization and fixation

Gravid hermaphrodites and their laid eggs were harvested in water, digested twice for three minutes in hypochlorite solution (1 N NaOH, 10% commercial bleach), and washed in PBS (140 mM NaCl, 3 mM KCl, 4 mM Na2HPO4, 1.5 mM KH2PO4). Embryos are protected by their egg shells from hypochlorite digestion, while larvae and adults are dissolved by this treatment. The earliest embryos isolated by this method were in the 1-cell stage at the time of pronuclear meeting, approximately 30 minutes after fertilization.

Embryos (approx. 1000) in PBS (10–20 μl) were pipetted onto a 14×14 mm polylysine-coated well (Cel-Line Associates Inc., Cat# 10-2066), overlayed with a glass coverslip and frozen on an aluminum block on dry ice. After cracking off coverslips, the slides were immersed in 100% methanol at −20°C for 5 minutes, placed in 100% methanol at room temperature for 5 minutes and rehydrated at room temperature using the following methanol series: 90% MeOH in H2O, 70% MeOH in PBS; 50% MeOH in PBS (1 minute washes), followed by two 5 minute washes in PTw (1× PBS, 0.1% Tween 20).

At this point, in some experiments, a Proteinase K digestion step was included: embryos were incubated for 15 minutes in 1 μg/ml or 10 μg/ml of proteinase K in PTw at room temperature. Proteinase K digestion was stopped by incubating for 2 minutes in 2 mg/ml glycine in PTw.

After two 5 minute washes in PTw, the embryos were fixed for 20 minutes in fixative (1× PBS, 0.08 M Hepes (pH 6.9), 1.6 mM MgSO4, 0.8 mM EGTA, 3.7% formaldehyde), and washed twice for 5 minutes in PTw, once for 5 minutes in 2 mg/ml glycine in PTw, and three times for 5 minutes in PTw.

Hybridization using cDNA-derived probes

Hybridization buffer consisted of 50% formamide, 5× SSC, 100 μg/ml autoclaved salmon sperm DNA, 50 μg/ml heparin, 0.1% Tween 20. Fixed embryos on slides were prehybridized in three steps: (1) 10 minutes at room temperature in a 1:1 mix of PTw and hybridization buffer; (2) 10 minutes at room temperature in undiluted hybridization buffer; (3) 1–4 hours at 48°C in previously boiled hybridization buffer.

30 μl of probe was applied and embryos overlayed with a parafilm coverslip, sealed with rubber cement and incubated overnight at 48°C. After hybridization, embryos were washed at 48°C twice with hybridization buffer, twice with 3:2 [hybridization buffer:PTw], twice with 1:4 [hybridization buffer:PTw], and twice in PTw (15 minute washes). Embryos were then washed twice (20 minutes) in PBT (1× PBS, 0.1% BSA, 0.1% Triton X-100) at room temperature.

Hybridization using oligonucleotide probes

Prehybridization and hybridization were performed at 37°C as above with adjusted concentrations of SSC and formamide as follows: oligo(dT) and oligo(dA) probes: 3× SSC; 3% formamide; anti-SL1: 2× SSC, 7.7% formamide; anti-SL1 precursor: 2× SSC, 14.3% formamide. These conditions are predicted to give optimal hybridization for each oligonucleotide (Davis et al., 1986). After hybridization, embryos were washed for 10 minutes in hybridization solution at 37°C, and twice for 10 minutes in TTBS (150 mM NaCl, 50 mM Tris-HCl pH 7.8, 0.1% BSA, 0.1% Tween-20) at room temperature.

Detection

Alkaline phosphatase (AP)-mediated detection was carried out using an AP-anti-DIG antibody conjugate (Boehringer Mannheim). The AP color reaction was performed in staining solution (100 mM NaCl, 5 mM MgCl2, 100 mM Tris, pH 9.5; 0.1% Tween 20, 1 mM Levamisole) containing 4.5 μl NBT/ml, 3.5 μl X-phosphate/ml and 1 μg DAPI/ml, and stopped by washing twice in PBS. The embryos were mounted in 5 –10 μl of a 70% glycerol solution under a glass coverslip.

Fluorescent detection was carried out using rhodamine-labeled or unlabeled anti-DIG antibodies (mouse monoclonal or sheep poly-clonal, Boehringer Mannheim) and appropriate fluorescent secondary antibodies, as indicated. Embryos were mounted in 5 –10 μl of mounting media (70% glycerol, 1 mg/ml p-phenylenediamine pH 9, 0.02% NaN3) under a glass coverslip.

Double-labeling using anti-P granule antibody

After the detection step, embryos were incubated for 2 hours at room temperature in a 1:400 dilution of mouse monoclonal OIC1D4 anti-P granule antibody (Strome, 1986) conjugated to 5(and 6)-carboxyfluorescein, succinimidyl ester (FSE; the conjugated antibody was a gift of J. Paulsen and S. Strome). Embryos were washed twice for 5 minutes with TTBS before mounting.

Staging of embryos and blastomere identification

The precise developmental stage of individual embryos was determined by counting DAPI-stained nuclei. Blastomeres ABa, ABp, EMS, C, P1, P2 and P3 were identified by their characteristic positions. P4, Z2 and Z3 were identified using anti-P granule antibodies. In 26-cell embryos, Ca and Cp were identified based on their position relative to P4 and on their characteristic mitotic nuclei (they are the only dividing cells at that stage). D, Da and Dp were identified based on position relative to P4.

Analysis of the embryonic expression pattern of 21 genes

We have developed an in situ hybridization protocol for the detection of RNA in whole-mount C. elegans embryos (for a detailed manual see Seydoux and Fire, 1994). This protocol is based on C. elegans protocols for the permeabilization and fixation of populations of whole-mount embryos (Albertson, 1984; Krause et al., 1990; Ahnn and Fire, 1994) and on Drosophila protocols for in situ hybridization (Tautz and Pfeifle, 1989; Patel and Goodman, 1992 and personal communication). Briefly, mixed-stage embryos are permeabilized by freeze-cracking on slides, fixed with methanol and formaldehyde, and hybridized with a single-stranded DNA probe labeled with digoxigenin-dUTP. Hybridized probe is then visualized using an anti-digoxigenin antibody coupled to either an enzymatic or a fluorescent marker.

We have used this protocol to analyze the embryonic expression of 21 genes. Some of these genes encode genetically characterized regulators of embryonic development (e.g. glp-1 and skn-1; Priess et al., 1987; Bowerman et al., 1992), while others encode ubiquitous proteins (e.g. ama-1, encoding the large subunit of RNA polymerase II; Bird and Riddle, 1989). The observed distributions of transcripts derived from these genes are presented in Table 1. We found that the in situ hybridization protocol allows visualization of RNA from the 1-cell stage to the end of embryogenesis (pretzel stage; 554 cells). The specificity of the hybridization was tested by comparing staining obtained from sense and antisense probes derived from the same gene, and from antisense probes derived from different genes: no staining was detected with sense probes (data not shown); while antisense probes derived from different genes gave distinct patterns (Table 1).

Table 1.

Embryonic expression pattern of 21 genes

Embryonic expression pattern of 21 genes
Embryonic expression pattern of 21 genes

Distinguishing between maternal and embryonically transcribed RNAs

In our survey of RNA distribution patterns, we found some RNAs that were present from the 1-cell stage, and others that appeared de novo at later stages of development (Table 1). The earliest RNAs are likely to be maternal (transcribed by the maternal genome and incorporated in the egg during oogenesis), while RNAs appearing de novo are likely to be embryonically transcribed. To distinguish between the contributions of maternal and embryonic transcription for specific genes, we made use of strains heterozygous for chromosomal deletions that remove relevant loci (Table 2). The cey-1 gene provides a particularly good example of both maternal and embryonically transcribed RNAs. Deficiency mnDf90 has been shown to remove the cey-1 gene and flanking markers (Sigurdson et al., 1984; Jantsch-Plunger, 1993). Deficiency heterozygotes (mnDf90/+) were allowed to self-fertilize to generate a population consisting of 75% of embryos that have one or two copies of the cey-1 gene (mnDf90/+ or +/+), and 25% of embryos that have no copy of cey-1 (mDf90/mDf90). Examining this population for the presence of the cey-1 RNA, we found that 99% of 1-to 4-cell-stage embryos were positive for cey-1 RNA, indicating that this early RNA is independent of embryonic transcription (Table 2). In contrast, only 75% of 200-to 550-cell-stage embryos derived from mnDf90/+ hermaphrodites were positive for cey-1 RNA, indicating that cey-1 RNA present at this stage depends on embryonic transcription of the cey-1 gene. (Wild-type populations of embryos are 100% positive for cey-1 staining, Table 2). These data are consistent with the hypothesis that RNAs present at the 1-cell stage are maternal, while RNAs arising later in development are embryonically transcribed.

Table 2.

Distinguishing maternal from embryonically transcribed RNAs

Distinguishing maternal from embryonically transcribed RNAs
Distinguishing maternal from embryonically transcribed RNAs

Maternal RNAs

All of the maternal RNAs detected in our survey (14/14) were evenly distributed in the 1-cell embryo. After the onset of cell division, however, we observed two distinct distribution patterns. A subset of maternal RNAs (8/14) continued to show a uniform distribution in all cells. We refer to these RNAs as class I maternal RNAs.

A second class of maternal RNAs in our survey (6/14 ; designated class II maternal RNAs) showed distinctive asymmetry between somatic and germline blastomeres following the onset of cell division: at each P cell division, these RNAs were segregated to both daughters, but appeared to be subsequently lost from the somatic daughter and/or its progeny, while being maintained in the P daughter. The most rapid rate of loss was observed for the cey-2 RNA, which by the late 2-cell stage was detected at a lower level in the somatic AB blastomere compared to the germline P1 blastomere (Fig. 2). Slower rates of somatic loss were observed for the other RNAs (Fig. 2). In all cases, these maternal RNAs appeared absent after the 24-cell stage from all somatic lineages, but could still be detected in P4 (Fig. 2).

Fig. 2.

Localization of maternal RNAs in early embryos. Embryos were hybridized to dpy-30, cey-2, skn-1 or glp-1 digoxigenin-labeled antisense probes. Probes were visualized using an antidigoxigenin antibody coupled to alkaline phosphatase, which was reacted with chromogenic substrates to give a dark color. These maternal RNAs appear predominantly cytoplasmic, although we have not excluded the possibility that they are also present in nuclei. In this and all other figures, anterior is to the left and dorsal is up. Embryos measure approximately 50 μm along their anteriorposterior axis.

Fig. 2.

Localization of maternal RNAs in early embryos. Embryos were hybridized to dpy-30, cey-2, skn-1 or glp-1 digoxigenin-labeled antisense probes. Probes were visualized using an antidigoxigenin antibody coupled to alkaline phosphatase, which was reacted with chromogenic substrates to give a dark color. These maternal RNAs appear predominantly cytoplasmic, although we have not excluded the possibility that they are also present in nuclei. In this and all other figures, anterior is to the left and dorsal is up. Embryos measure approximately 50 μm along their anteriorposterior axis.

We used appropriate deficiency chromosomes to confirm that the observed distributions of class I and class II RNAs reflected the distributions of maternal transcripts. This analysis was carried out for two class I genes (dpy-30 and ama-1) and one class II gene (skn-1). In each case, we found RNA distributions in early (1-to 28-cell) stages to be identical between wild-type embryos and deficiency homozygoytes lacking both genomic copies of the probed gene (Table 2).

Because our method of detection relies on the ability of an exogenous probe to hybridize to RNA in situ, it was important to verify that the apparent disappearance of class II RNAs from somatic cells was due to actual loss of RNA by degradation, rather than to the formation of double-stranded RNA or ribonucleoprotein complexes that were inaccessible to the probe. The quantitative masking of maternal mRNA by antisense RNA seems unlikely, given that we detected no antisense transcripts (nuclear or cytoplasmic) using appropriate sense probes for cey-1, cey-2 and skn-1.

To address the possibility of masking by ribonucleoprotein complexes, we performed in situ hybridization experiments to cey-2 and CeIF RNAs following digestion of embryos with proteinase K. Proteinase K treatment at a low concentration (1 μg/ml) did not change the hybridization patterns of the cey-2 or CeIF probes at any stage (data not shown). Proteinase K treatment at a higher concentration (10 μg/ml) was sufficient to cause significant damage to the morphology of embryos, but still did not affect the distribution of cey-2 and CeIF RNAs in 12-cell and older embryos (cell integrity in earlier embryos was too damaged for cell boundary identification; data not shown). Although we cannot rule out the possibility that some ribonucleoprotein complexes are still present under these conditions, these results are consistent with loss of RNA by degradation. This interpretation is further supported by experiments reported by Ahringer et al. (1992). Their northern analyses of RNA derived from populations of staged embryos indicate that the level of glp-1 RNA, another class II maternal RNA, decreases dramatically during the first two hours of embryogenesis.

Embryonically transcribed RNAs

For 12 of the 21 genes in our survey, embryonically transcribed RNAs were easily recognized as new signals arising after a period when no RNA was detected. The timing of the onset of embryonic transcription varied greatly for different genes, from before gastrulation (which begins at the 26-cell stage) to after the onset of morphogenesis (550-cell stage), with no apparent preference for any particular developmental stage (Table 1). For a subset of genes in our survey (8/21), the onset of embryonic transcription could not be determined because newly transcribed RNAs could not be distinguished from perduring maternal RNAs.

The earliest embryonically transcribed RNA detected in our survey was derived from pes-10. pes-10 is a gene of unknown function that we identified in a promoter-trap screen (carried out as in Hope, 1991) designed to identify transcripts expressed in early embryos (details of the screen will be published elsewhere). We found that pes-10 RNA is present transiently in each somatic lineage starting in the 4-cell stage (Fig. 3). Because no pes-10 RNA was detected at the 1-cell stage, and because the earliest pes-10 RNA was detected in nuclei rather than cytoplasm, we believe this RNA to be dependent on embryonic transcription. To analyze the onset of pes-10 transcription with a greater degree of sensitivity, we have taken advantage of the amplified signal derived from a chromosomally integrated, multicopy pes10::lacZ transgene. The pes-10::lacZ RNA is detected sequentially in each somatic lineage shortly after each lineage diverges from the germ lineage (Fig. 3). In each somatic blastomere, the pes-10:: lacZ RNA initially accumulates in two nuclear foci, which may correspond to the site of transcription on the two homologous chromosomes that carry the multi-copy transgene. These ‘double dots’ are absent from cells undergoing mitosis, suggesting that transcription may not be occurring during that time. In each somatic lineage, pes-10 and pes-10::lacZ RNAs were detected for one to four cell cycles, before being rapidly degraded.

Fig. 3.

Localization of pes-10 and pes-10::lacZ embryonically transcribed RNAs in early embryos. Wild-type embryos (first column) or embryos transformed with a pes-10::lacZ transgene (second column) were hybridized, respectively, to pes-10 or lacZ anti-sense probes. Probes were visualized as described in the legend of Fig. 2. Arrows point to nuclei of recently arisen somatic founder blastomeres where the pes-10::lacZ message has just started to accumulate in two foci. In contrast to the endogenous pes-10 message, which quickly becomes cytoplasmic (see 7/8-cellstage), the pes-10::lacZ RNAs remains predominantly nuclear until the 26-cell stage. This apparent block in nuclear export (or cytoplasmic instability) has also been observed for other lacZ fusion RNAs (data not shown), and could be due to the long intronless coding region of lacZ. Cs refer to the 4 C-derived blastomeres: Caa, Cap, Cpa, Cpp.

Fig. 3.

Localization of pes-10 and pes-10::lacZ embryonically transcribed RNAs in early embryos. Wild-type embryos (first column) or embryos transformed with a pes-10::lacZ transgene (second column) were hybridized, respectively, to pes-10 or lacZ anti-sense probes. Probes were visualized as described in the legend of Fig. 2. Arrows point to nuclei of recently arisen somatic founder blastomeres where the pes-10::lacZ message has just started to accumulate in two foci. In contrast to the endogenous pes-10 message, which quickly becomes cytoplasmic (see 7/8-cellstage), the pes-10::lacZ RNAs remains predominantly nuclear until the 26-cell stage. This apparent block in nuclear export (or cytoplasmic instability) has also been observed for other lacZ fusion RNAs (data not shown), and could be due to the long intronless coding region of lacZ. Cs refer to the 4 C-derived blastomeres: Caa, Cap, Cpa, Cpp.

Transcription of pes-10 and of the pes-10::lacZ transgene was never detected in the germ lineage (P0-P4) (Figs 3 and 4). Interestingly, embryonically transcribed RNAs from the other genes in our survey also appeared to be absent from the germ lineage in the early embryo. A particularly striking example comes from the lin-19 RNA which is expressed in all somatic blastomeres from the 15-cell stage to the 60-cell stage, but was never detected in the germ lineage (Fig. 4).

Fig. 4.

pes-10 and lin-19 embryonically transcribed RNAs are absent from the germ lineage. Wild-type embryos were hybridized to pes-10 P3 or lin-19 antisense probes and stained with anti-P granule antibody, which identifies blastomeres from the germ lineage. (A) 24-cellaembryo hybridized to pes-10 antisense probe. (C) 28-cell embryo hybridized to lin-19 antisense probe. (B,D) Same embryos as in A or C, respectively, showing anti-P granule staining in P4.

Fig. 4.

pes-10 and lin-19 embryonically transcribed RNAs are absent from the germ lineage. Wild-type embryos were hybridized to pes-10 P3 or lin-19 antisense probes and stained with anti-P granule antibody, which identifies blastomeres from the germ lineage. (A) 24-cellaembryo hybridized to pes-10 antisense probe. (C) 28-cell embryo hybridized to lin-19 antisense probe. (B,D) Same embryos as in A or C, respectively, showing anti-P granule staining in P4.

When do embryonically transcribed RNAs first appear in the germ lineage? As discussed above, the presence of perduring maternal RNAs, particularly in the germ lineage, makes it difficult to determine whether embryonically transcribed RNAs are also present. For example, the act-1 RNA is present in the germ lineage continuously from the 1-cell stage to the end of embryogenesis (Table 1). Because there are no available deficiencies that remove the act-1 gene, we were not able to determine when this RNA becomes dependent on embryonic transcription. In one case, however, we detected a newly arisen, and therefore presumably embryonically transcribed, RNA in the germ lineage: this transcript was derived from an array containing multiple copies of the glp-1 gene and of a glp-1::lacZ fusion, and was first detected in Z2 and Z3 in the bean stage, after the onset of morphogenesis (Table 1). This result suggests that, at least by that stage, the germ cells Z2 and Z3 have become competent for transcription.

The par-4 gene is required for the asymmetric distributions of maternal and embryonically transcribed RNAs

The generation of morphological asymmetry in early embryos has been shown to involve the products of the par-1, par-2, par-3 and par-4 genes (Kemphues et al., of poly(A)+ RNAs using a digoxigenin-labeled oligo(dT) probe. We found that this probe hybridized uniformly to the cytoplasm of all somatic cells. In the P blastomeres, we detected additional clusters of strong staining above a uniform background (Fig. 6). In P0, P1, and P2, these clusters were preferentially associated with the posterior half of the cell cortex prior to each division. In P3 and P4, the clusters coalesced and 1988). These genes were identified by maternal-effect mutations which lead to defects in early cleavage patterns and mislocalization of the germline-specific P granules. We used a mutation in the par-4 gene to address the possibility that par-dependent mechanisms also function in the generation of asymmetry in maternal and embryonically transcribed RNA distributions. In par-4 embryos, after a normal asymmetric first cleavage, both the AB and P1 blastomeres divide symmetrically and synchronously (Kemphues et al., 1988; Morton et al., 1992). This and other characteristics led to the proposal that the P (germ) lineage is transformed to an AB-like lineage in par-4 embryos (Morton et al., 1992). We found that the distributions of both the cey-2 and pes-10 RNAs in par-4 embryos reflect this proposed P1 to AB transformation. The cey-2 RNA, a maternal transcript which in wildtype embryos is lost from the AB lineage but is maintained in the P lineage, was found to be lost rapidly from all blastomeres in par-4 embryos (Fig. 5). Conversely, the pes-10 RNA, a embryonically transcribed RNA which in wild-type is detected in the AB lineage but not in the P lineage, was detected in all blastomeres in par-4 embryos (Fig. 5). These results indicate that the mechanism disrupted in par-4 embryos is required for the asymmetric distributions of both maternal and embryonically transcribed RNAs.

Fig. 5.

Distribution of cey-2 and pes-10 RNAs in wild-type and par-4 embryos. Early embryonic lineages are depicted for both wild-type and par-4 embryos. Thick black lines indicate cells where cey-2 (maternal) or pes-10 (embryonically transcribed) cytoplasmic RNAs were detected. Thick gray lines represent cells where nuclear pes-10 RNA was detected.

Fig. 5.

Distribution of cey-2 and pes-10 RNAs in wild-type and par-4 embryos. Early embryonic lineages are depicted for both wild-type and par-4 embryos. Thick black lines indicate cells where cey-2 (maternal) or pes-10 (embryonically transcribed) cytoplasmic RNAs were detected. Thick gray lines represent cells where nuclear pes-10 RNA was detected.

Fig. 6.

Distribution of poly(A)+ RNAs. Arrows point to representative P granule-like staining. (A-C,E,F) Wild-type embryos hybridized to an oligo(dT) probe visualized using an antidigoxigenin antibody coupled to rhodamine. (D) Same embryo as in C labeled with a FSE-conjugated anti-P granule antibody. Embryos stages are as follows: (A) 1-cell, (B) 2-cell, (C,D) 4-cell, (E) 8-cell, (F)100-cell. (G,H) Wild-type 2- and 6-cell embryos hybridized to an antisense cey-2 probe visualized using an unlabeled anti-digoxigenin antibody and a 2° antibody coupled to fluorescein.

Fig. 6.

Distribution of poly(A)+ RNAs. Arrows point to representative P granule-like staining. (A-C,E,F) Wild-type embryos hybridized to an oligo(dT) probe visualized using an antidigoxigenin antibody coupled to rhodamine. (D) Same embryo as in C labeled with a FSE-conjugated anti-P granule antibody. Embryos stages are as follows: (A) 1-cell, (B) 2-cell, (C,D) 4-cell, (E) 8-cell, (F)100-cell. (G,H) Wild-type 2- and 6-cell embryos hybridized to an antisense cey-2 probe visualized using an unlabeled anti-digoxigenin antibody and a 2° antibody coupled to fluorescein.

Clusters of poly(A)+ RNAs co-localize with P granules

So far, our analysis of maternal and embryonically transcribed RNAs has been limited to the 21 genes included in our survey. To complement this analysis, we have examined the overall distribution became peri-nuclear. This dynamic pattern of localization is identical to the distribution of P granules (Strome and Wood, 1982). To determine whether the oligo(dT)-hybridizing clusters were coincident with P granules, we performed double label in situ experiments with P granules labeled with an FSE-conjugated anti-P granule antibody (OIC1D4; Strome, 1986). In these experiments, we found that the majority of the oligo(dT)-hybridizing clusters co-localized with P granules (compare Figs 6C and D). These results suggest that poly(A)+ RNA is present in P granules.

As a control for the specificity of the oligo(dT) staining pattern, we have used an oligo-dA probe, which gave no staining (Table 3). In addition, pre-treatment of embryos with RNAse A or RNAse I eliminated staining with oligo(dT) (Table 3).

Table 3.

Specificity of oligo(dT) hybridization

Specificity of oligo(dT) hybridization
Specificity of oligo(dT) hybridization

To characterize further the P granule-associated RNAs, we used a probe complementary to the 22 nucleotides of SL1, a short sequence which is trans-spliced upstream of many RNAs (Krause and Hirsh, 1987; Bektesh et al., 1988). Like the oligo(dT) probe, this anti-SL1 oligonucleotide probe hybridized to the cytoplasm of all cells and to cytoplasmic granules in the P cells (presumably P granules; data not shown). In contrast, an oligonucleotide complementary to the 3′ region of the unspliced SL1 precursor RNA did not show any staining (data not shown). We have also used a monoclonal antibody (K121; Krainer, 1988) that recognizes the trimethylguanosine cap found at the 5′ end of many small RNAs, including SL1 (Liou and Blumenthal, 1990; Van Doren and Hirsh, 1990). Like the oligo(dT) and anti-SL1 probes, this antibody recognizes putative P granules (data not shown). Together, these results suggest that SL1-containing RNAs are present in P granules.

What is the identity of the RNAs associated with P granules? Because these RNAs are detected as early as the 1-cell stage, we presume that they are maternal RNAs. Although all the specific maternal RNAs included in our survey were present in the P lineage, none appeared to be preferentially associated with P granules in the 1-to 4-cell stage (eg. compare Fig. 6B,G). Later, when the P granules start to coalesce around the nucleus in P3 and P4, class II maternal RNAs were often found to also adopt this perinuclear localization (eg. Figs 2, 6H), but the significance of this late association remains unclear. These results suggest that RNAs associated with P granules in early embryos belong to a class of maternal RNAs not represented in the 21 genes chosen for our survey. We refer to these RNAs as class III maternal RNAs.

We have developed an in situ hybridization protocol suitable for the detection of mRNA in whole-mount C. elegans embryos. We have used this protocol to analyze the expression pattern of 21 genes (summarized in Fig. 7) and the overall distribution of poly(A)+ RNAs. We detect asymmetry in the distribution of maternal and embryonically transcribed RNAs as early as the 2- and 4-cell stages.

Fig. 7.

Embryonic expression of 21 genes as determined by in situ hybridization. Horizontal lines delineate the period during which a specific RNA was detected as listed in Table 1. Flags indicate changes in tissue specificity. Tissue specificity is indicated above each line (only major tissues are listed, UC refers to expression restricted to unidentified cells); otherwise expression is ubiquitous. Genes are grouped according to the type of maternal transcripts detected. (A) Genes producing class I maternal RNAs: maternal transcripts that appear to be maintained equally in all cells at least up to the 28-cell stage. The onset of embryonic transcription (if any) could not be determined for these genes because of the perdurance of maternal transcripts. (B) Genes producing class II maternal RNAs: maternal transcripts that appear to be lost from somatic cells but are maintained in germline cells. These maternal RNAs are represented with jagged lines to distinguish them from the later arising embryonically transcribed RNAs (straight lines). (C) Genes producing no maternal RNAs. Only transcripts arising after the 1-cell stage (presumably embryonically transcribed) were detected for these genes.

Fig. 7.

Embryonic expression of 21 genes as determined by in situ hybridization. Horizontal lines delineate the period during which a specific RNA was detected as listed in Table 1. Flags indicate changes in tissue specificity. Tissue specificity is indicated above each line (only major tissues are listed, UC refers to expression restricted to unidentified cells); otherwise expression is ubiquitous. Genes are grouped according to the type of maternal transcripts detected. (A) Genes producing class I maternal RNAs: maternal transcripts that appear to be maintained equally in all cells at least up to the 28-cell stage. The onset of embryonic transcription (if any) could not be determined for these genes because of the perdurance of maternal transcripts. (B) Genes producing class II maternal RNAs: maternal transcripts that appear to be lost from somatic cells but are maintained in germline cells. These maternal RNAs are represented with jagged lines to distinguish them from the later arising embryonically transcribed RNAs (straight lines). (C) Genes producing no maternal RNAs. Only transcripts arising after the 1-cell stage (presumably embryonically transcribed) were detected for these genes.

The fate of maternal RNAs in early embryos

Among the 21 genes included in our survey, we observed two distinct classes of maternal RNAs: maternal RNAs that are maintained equally in all cells during early cleavages (class I RNAs), and maternal RNAs that are maintained preferentially in germ cells and are rapidly lost in somatic cells (class II RNAs).

Why are some maternal RNAs rapidly degraded in somatic cells while others are not? Among RNAs that are degraded, we find genes that function in processes ranging from mRNA translation (CeIF, Roussell and Bennett, 1992) to blastomere fate determination (glp-1, Priess et al., 1987; skn-1, Bowerman et al., 1992). Proteins encoded by this class of RNAs include some that are present in all cells in the early embryo (CEY-1, Jantsch-Plunger, 1993), and others that are localized to specific blastomeres (SKN-1, Bowerman, 1993; GLP-1, Evans et al., 1994). From these observations, it is difficult to find a common functional property that distin-guishes these class II maternal RNAs from class I maternal RNAs. One possibility is that maternal RNAs that are required only in early development (e.g. skn-1, glp-1;Priess et al., 1987; Bowerman et al., 1992,) or for which embryonically transcribed RNAs arise early (cey-1, CeIF) are degraded rapidly, while maternal RNAs whose products are required throughout development (e.g. ama-1; Bird and Riddle, 1989) are maintained in all cells to ensure continuous expression.

Why are maternal RNAs protected from rapid degradation in the germline? This phenomenon has also been observed in Drosophila for the maternal transcript of the Hsp83 gene. This transcript is initially uniformly distributed throughout the embryo. One hour after fertilization, it is degraded rapidly everywhere except in the polar plasm at the posterior pole of the syncytial embryo (Ding et al, 1993). Components of the pole plasm, in particular the RNA-containing polar granules, function in the determination of pole cells, the Drosophila primordial germ cells (Illmensee and Mahowald, 1974). One possibility is that protection of RNA from degradation is an essential property of the polar plasm to maintain RNAs that are essential for the development of germ cells. A similar phenomenon may be occurring in C. elegans. As discussed below,

the germline cells of C. elegans may also contain specialized RNAs. Mechanisms that evolved to maintain these RNAs may protect all RNAs present in germline cells from rapid degradation.

Clusters of poly(A)+ RNAs are associated with P granules

All class I and class II RNAs detected in our survey using gene-specific probes were uniformly distributed in the 1-cell stage embryo. Using an oligo(dT) probe, however, we found evidence for a third class of maternal RNAs. We observed clusters of poly(A)+ RNAs that co-localized with the germline-specific P granules. Because these RNAs are present at the 1-cell stage, they are likely to be maternal in origin. In addition, because none of the class I and class II maternal RNAs detected in our survey appeared to be associated with P granules, the poly(A)+ P granule-associated RNAs likely belong to a third class of maternal RNAs not included in our survey. Whether these class III RNAs are exclusively associated with P granules or are also present in the cytoplasm of somatic cells remains to be determined.

What is the function of P granules and their associated class III maternal RNAs? The C. elegans P granules are similar to the polar granules, dense bodies or nuage that have been observed in the germline of many organisms (Eddy, 1975). Much like the C. elegans P granules, the Drosophila polar granules contain RNA and are localized to the posterior pole of the early embryo prior to their segregation in germ cells (Mahowald, 1962, 1971). As mentioned above, the Drosophila polar granules and their associated polar plasm and RNAs are known to be required for the development of posterior structures including germ cells. No such function, however, has yet been demonstrated for the P granules of C. elegans. Nevertheless, it remains a reasonable hypothesis that P granules and class III RNAs function in some aspect of germ line and/or posterior development. Asymmetrically localized cytoplasmic factors have been postulated to function in the determination of the anterior-posterior axis in the 1-cell embryo (Schieren-berg and Strome, 1992). Isolation of class III RNAs and subsequent genetic analysis will help determine what role if any these RNAs have in these processes.

Embryonic transcription can be detected in somatic cells as early as the 4-cell stage

Using a probe specific for the pes-10 RNA, we detected embryonically transcribed RNAs in the nuclei of somatic cells as early as the 4-cell stage. Similarly, Edgar et al. (1994) were able to detect [3H]UTP incorporation in permeabilized embryos as early as the 8-cell stage (background levels precluded detection in earlier stages). These observations indicate that, in C. elegans, embryonic transcription is not delayed by the rapid first mitotic cycles (also see Edgar et al., 1994). In this regard, C. elegans differs from other organisms with rapid initial mitoses such as Drosophila and Xenopus, where the bulk of embryonic transcription starts only after cell cycles slow down (Newport and Kirschner, 1982; Edgar and Schubiger, 1986). One possibility is that the relatively small size of C. elegans genes permits them to be transcribed during rapid mitotic cycles.

Asymmetry in the distribution of newly transcribed RNAs was apparent from the onset of embryonic transcription. Newly transcribed RNAs were detected in all somatic blastomeres from the 4-cell stage onward, but were not detected in the germ lineage during early cleavages. A lack of embryonic transcription in the germ lineage of the early embryo has been observed in Drosophila, where germ cell nuclei failed to incorporate [3H]UTP (Zalokar, 1976) and to hybridize with [3H]poly(U) (Lamb and Laird, 1976) in the blastoderm stage. The mechanisms that inhibit transcription in early germ cells are not known. Although we cannot rule out the production of specific transcripts in the C. elegans embryonic germ lineage, our observations support a model where factor(s) required for the repression (or activation) of transcription are asymmetrically segregated at each unequal germline cell division. An intriguing possibility is that such factors might also function to maintain (or degrade) maternal RNAs.

Mechanisms required for unequal divisions in the early embryo affect the distributions of maternal and embryonically transcribed RNAs

As described in the Introduction, previous studies have shown that somatic and germline cells differ in size, cleavage pattern, and P granule content. This study now shows that somatic and germline cells also differ in three aspects of RNA metabolism. First, germline cells protect maternal RNAs from the rapid degradation observed in somatic cells. Second, germline cells contain clusters of poly(A)+ RNAs not seen in somatic cells. Third, as early as the 4-cell stage, somatic cells transcribe certain RNAs, which are off in germ cells. Using a mutation in the par-4 gene, we have shown that a failure to maintain unequal cell division in the early embryo affects at least two of these processes (maternal RNA degradation and embryonic transcription). [par-4 embryos also missegregate P granules (Morton et al., 1992) and would therefore also be expected to mislocalize poly(A)+ RNA clusters, although this was not tested.] These results indicate that mechanisms required for the establishment of asymmetry in the early embryo have immediate effects on the distribution of RNAs. Whether these effects in turn result in the determination or execution of distinct germline and somatic fates will be the subject of further studies.

We thank Nipam Patel for help and encouragment in developing in situ hybridization for C. elegans; Donna Albertson, Tom Evans, David Greenstein, Vincent Guacci, Mike Krause, Shohei Mitani, Pete Okkema, Neela Patel and Jorge Mancillas for additional protocols and advice; Susan Strome, Janet Paulsen and Joe Gall for antibodies; David Bird, Edward Kipreos, Verena Jantsch-Plunger, Jim Priess, Ann Sluder, Dave Hsu, Barbara Meyer, Deborah Roussell, Karen Bennett, Patty Wohldmann and Bob Waterston for DNA clones; and Donald Brown, Susan Dymecki, Joe Gall, Tim Schedl, Bill Kelly, Lihsia Chen, Mary Montgomery and Joo-Hong Ahnn for comments on the manuscript. G. S. is a Helen Hay Whitney Fellow. A. F. is a Rita Allen Foundation Scholar. This work was supported by NIH (Grant R01-GM37706) and the Carnegie Institution of Washington. Some strains used in this work were provided by the Caenorhabditis Genetics Center (funded by NIH Center for Research Resources).

Ahnn
,
J.
and
Fire
,
A.
(
1994
).
A screen for genetic loci required for body-wall muscle development during embryogenesis in Caenorhabditis elegans
.
Genetics
137
,
483
498
.
Albertson
,
D. G.
(
1984
).
Formation of the first cleavage spindle in nematode embryos
.
Dev. Biol
.
101
,
61
72
.
Ahringer
,
J.
,
Rosenquist
,
T. A.
,
Lawson
,
D. N.
and
Kimble
,
J.
(
1992
).
The Caenorhabditis elegans sex determining gene fem-3 is regulated post-transcriptionally
.
EMBO J
.
11
,
2303
2310
.
Bektesh
S.
,
Van Doren
K.
and
Hirsh
D.
(
1988
).
Presence of the Caenorhabditis elegans spliced leader on different mRNAs and in different genera of nematodes
.
Genes Dev
.
2
,
1277
1283
.
Bird
,
D. M.
and
Riddle
,
D. L.
(
1989
).
Molecular cloning and sequencing of ama-1, the gene encoding the largest subunit of Caenorhabditis elegans RNA polymerase II
.
Molec. Cell Biol
.
9-10
,
4119
4130
.
Boveri
,
T.
(
1910
).
Ueber die Teilung centrifugierte Eier von Ascaris megalocephala
.
Wilhelm Roux Arch. EntwMech. Org
.
31
,
101
125
.
Bowerman
,
B.
,
Eaton
,
B. A.
and
Priess
,
J. R.
(
1992
).
skn-1, a maternally expressed gene required to specify the fate of ventral blastomeres in the early C. elegans embryo
.
Cell
68
,
1061
1075
Bowerman
,
B.
,
Draper
,
B. W.
,
Mello
,
C. C.
and
Priess
,
J. R.
(
1993
)
The maternal gene skn-1 encodes a protein that is distributed unequally in early C. elegans embryos
.
Cell
74
,
443
452
Davis
,
L. G.
,
Dibner
,
M. D.
and
Battey
,
J. F.
(
1986
).
In Basic Methods in Molecular Biology
.
Elsevier
:
New York
.
Denis
,
H.
and
Lacroix
,
J.
(
1993
).
The dichotomy between germ line and somatic line, and the origin of cell mortality
.
Trends Genetics
,
9-1
,
7
11
.
Ding
,
D.
,
Parkhurst
,
S. M.
,
Halsell
,
S. R.
and
Lipshitz
,
H. D.
(
1993
).
Dynamic Hsp83 RNA localization during Drosophila oogenesis and embryogenesis
.
Molec. Cell. Biol
.
13-6
,
3773
3781
.
Eddy
,
E. M.
(
1975
).
Germ plasm and the differentiation of the germ cell line
.
Intern. Rev. Cytol
.
43
,
229
280
.
Edgar
,
B. A.
and
Schubiger
,
G.
(
1986
).
Parameters controlling transcriptional activation during early Drosophila development
.
Cell
44
,
871
877
.
Edgar
,
L. G.
,
Wolf
,
N.
and
Wood
,
W. B.
(
1994
).
Early transcription in Caenorhabditis elegans embryos
.
Development
120-2
,
443
451
.
Evans
,
T. C.
,
Crittenden
,
S. L.
,
Kodoyianni
,
L.
and
Kimble.
J
. (
1994
).
Translational control of maternal glp-1 mRNA establishes an asymmetry in the C. elegans embryo
.
Cell
77
,
183
194
.
Ferguson
,
E. L.
and
Horvitz
,
H. R.
(
1985
).
Identification and characterization of 22 genes that affect the vulval cell lineages of the nematode C. elegans
.
Genetics
110
,
17
72
.
Herman
,
R. K.
(
1978
).
Crossover suppressors and balanced recessive lethals in C. elegans
.
Genetics
88
,
49
65
.
Hope
,
I. A.
(
1991
).
Promoter trapping in Caenorhabditis elegans
.
Development
113
,
399
408
.
Hsu
,
D. R.
(
1992
).
Genetic and molecular analysis of dosage compensation
in Caenorhabditis elegans
.
Ph. D. Thesis, MIT
.
Hsu
,
D. R.
and
Meyer
,
B. J.
(
1994
).
The dpy-30 gene encodes an essential component of the C. elegans dosage compensation machinery
.
Genetics
, in press.
Illmensee
,
K.
and
Mahowald
,
A. P.
(
1974
).
Transplantation of posterior plasm in Drosophila. Induction of germ cells at the anterior pole of the egg
.
Proc. Natl. Acad. Sci USA
71
,
1016
1020
.
Jantsch-Plunger
,
V.
(
1993
).
Gene expression and regulation of the unc-54 myosin heavy chain gene in Caenorhabditis elegans: characterization of an enhancer element and enhancer binding factors
.
Ph. D. Thesis, University of Vienna
.
Kemphues
,
K. J.
,
Priess
,
J. R.
,
Morton
,
D. G.
and
Cheng
,
N.
(
1988
).
Identification of genes required for cytoplasmic localization in early C. elegans embryos
.
Cell
52
,
311
320
.
Krainer
,
A.
(
1988
).
Pre-mRNA splicing by complementation with purified human U1, U2, U4/U6 and U5 snRNPs
.
Nucleic Acids Research
16
,
9415
9429
.
Krause
,
M.
and
Hirsh
,
D.
(
1987
).
A trans-spliced leader sequence on actin mRNA in C. elegans
.
Cell
49
,
753
761
.
Krause
,
M.
,
Fire
,
A.
,
Harrison
,
S.W.
,
Priess
,
J.
and
Weintraub
,
H.
(
1990
).
CeMyoD accumulation defines the body wall muscle cell fate during C. elegans embryogenesis
.
Cell
63
,
907
919
.
Lamb
,
M. M.
and
Laird
,
C. D.
(
1976
).
Increase in nuclear poly(A)-containing RNA at syncytial blastoderm in Drosophila melanogaster embryos
.
Dev. Biol
.
54
,
31
42
.
Liou
,
R. F.
and
Blumenthal
,
T.
(
1990
).
Trans-spliced Caenorhabditis elegans messenger RNAs retain trimethylguanosine caps
.
Molec. Cell. Biol
.
10
,
1764
1768
.
Mahowald
,
A. P.
(
1962
)
Fine structure of pole cells and polar granules in Drosophila melanogaster
.
J. Exp. Morph
.
151
,
201
215
.
Mahowald
,
A. P.
(
1971
)
Polar granules of Drosophila. IV. Cytochemical studies showing loss of RNA from polar granules during early stages of embryogenesis
.
J. Exp. Morph
.
176
,
345
352
.
Manser
,
J.
and
Wood
,
W. B.
(
1990
).
Mutations affecting embryonic cell migrations in Caenorhabditis elegans
.
Dev. Genetics
11
,
49
64
.
Mello
,
C. C.
,
Kramer
,
J. M.
,
Stinchcomb
,
D.
and
Ambros
,
V.
(
1991
).
Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences
.
EMBO J
.
10-12
,
3959
3970
.
Morton
,
D. G.
,
Roos
,
M.
and
Kemphues
,
K. J.
(
1992
).
par-4, a gene required for cytoplasmic localization and determination of specific cell types in Caenorhabditis elegans embryogenesis
.
Genetics
130
,
771
790
.
Newport
,
J.
and
Kirschner
,
M.
(
1982
).
A major developmental transition in early Xenopus embryos: II. Control of the onset of transcription
.
Cell
30
,
687
696
.
Patel
,
N. H.
and
Goodman
,
C. S.
(
1992
).
Preparation of digoxigenin-labeled single-stranded DNA probes
.
In Non-radioactive Labeling and Detection of Biomolecules
. (ed.
C.
Kessler
).
Springer-Verlag
,
Berlin
.
Priess
,
J. R.
,
Schnabel
,
H.
and
Schnabel
,
R.
(
1987
).
The glp-1 locus and cellular interactions in early C. elegans embryos
.
Cell
51
,
601
611
.
Roussell
,
D. L.
and
Bennett
,
K. L.
(
1992
).
Caenorhabditis cDNA encodes an eIF-4A-like protein
.
Nucleic Acids Res
.
20
,
3783
.
Schierenberg
,
E.
(
1988
).
Localization and segregation of lineage-specific cleavage potential in embryos of Caenorhabditis elegans
.
Roux’s Arch. Dev. Biol
.
197
,
282
293
.
Schierenberg
,
E.
and
Strome
,
S.
(
1992
).
The establishment of embryonic axes and determination of cell fates in embryos of the nematode Caenorhabditis elegans
.
Seminars in Developmental Biology
3
,
25
33
.
Seydoux
,
G.
and
Fire
,
A.
(
1994
).
Whole-mount in situ hybridization for the detection of RNA in C. elegans embryos
.
In C. elegans: Modern Biological Analysis of an Organism. Methods in Cell Biology
(ed.
H.
Epstein
and
D.
Shakes
).
Academic Press
,
San Diego
(in press).
Sigurdson
,
D. C.
,
Spanier
,
G. J.
and
Herman
,
R. K.
(
1984
).
C. elegans deficiency mapping
.
Genetics
108
,
331
345
.
Strome
,
S.
(
1986
).
Asymmetric movments of cytolasmic components in Caenorhabditis elegans zygotes
.
J. Embryol. Exp. Morph
.
97
,
15
29
.
Strome
,
S.
and
Wood
,
W. B.
(
1982
).
Immunofluorescence visualization of germ-line-specific cytoplasmic granules in embryos, larvae, and adults of C. elegans
.
Proc. Natl Acad. Sci. USA
79
,
1558
1562
.
Sulston
,
J.E.
,
Schierenberg
,
E.
,
White
,
J. G.
and
Thomson
,
J. N.
(
1983
).
The embryonic cell lineage of the nematode C. elegans
.
Dev. Biol
.
100
,
64
119
.
Tautz
,
D.
and
Pfeifle
,
C.
(
1989
).
A nonradioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback
.
Chromosoma
98
,
81
85
.
Van Doren
,
K.
and
Hirsh
,
D.
(
1990
).
mRNAs that mature through trans-splicing in Caenorhabditis elegans have a trimethylguanosine cap at their 5′ termini
.
Molec. Cell. Biol
.
10
,
1769
1772
.
Waterston
,
R.
,
Martin
,
C.
,
Craxton
,
M.
,
Huynh
,
C.
,
Coulson
,
A.
,
Hillier
,
L.
,
Durbin
,
R.
,
Green
,
P.
,
Shownkeen
,
R.
,
Halooran
,
N.
,
Metzstein
,
M.
,
Hawkins
,
T.
,
Wilson
,
R.
,
Berks
,
M.
,
Du
,
Z.
,
Thomas
,
J.
,
Thierry-Mieg
,
J.
and
Sulston
,
J.
(
1992
).
A survey of expressed genes in Caenorhabditis elegans
.
Nature Genetics
1
,
114
123
.
Wolf
N.
,
Priess
J.
and
Hirsh
D.
(
1983
).
Segregation of germline granules in early embryos of C. elegans: An electron microscopic analysis
.
J. Embryol. Exp. Morph
.
73
,
297
306
.
Wood
,
W. B.
and
Edgar
,
L. G.
(
1994
).
Patterning in the C. elegans embryo
.
Trends Genetics
10-2
,
49
53
.
Yochem
,
J.
and
Greenwald
,
I.
(
1989
).
glp-1 and lin-12, genes implicated in distinct cell-cell interactions in C. elegans, encode similar transmembrane proteins
.
Cell
58
,
553
563
.
Zalokar
,
M.
(
1976
).
Autoradiographic study of protein and RNA formation during early development of Drosophila eggs
.
Dev. Biol
.
49
,
425
437
.