SL1 Trans-Soliced Leader RNA Performs an Essintial Embrvonic Function in Caenorhabditis Elegans That Can Also Be Supplied by SL2 RNA

Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

The SL1 trans-soliced leader RNA performs an essintial embrvonic function In Caenorhabditis elegans that can also be supplied by SL2 RNA

Kimberly C. Ferguson, Paul 1. Heid, and Joel H. ~othman' Department of Biochemistry, University of Wisconsin, Madison, Wisconsin 53706 USA

Covalent joining of leader RNA exons to pre-mRNAs by trans-splicing has been observed in protists and invertebrates, and can occur in cultured mammalian cells. In the nematode Caenorhabditis elegans, -60% of mRNA species are trans-spliced to the 22-nucleotide SLI leader, and another -10% of mRNAs receive the 22-nucleotide SL2 leader. We have isolated deletions that remove the rrs-1 cluster, a gene complex that contains -110 tandem copies of a repeat encoding both SL1 RNA and 5s rRNA. An SL1-encoding gene alone rescues the embryonic lethality caused by these deletions. Mutations within the Sm-binding site of SL1 RNA, which is required for trans-splicing, eliminate rescue, suggesting that the ability of the SL1 leader to be trans-spliced is required for its essential activity. We observe pleiotropic defects in embryos lacking SL1 RNA, suggesting that multiple mRNAs may be affected by the absence of an SL1 leader. We found, however, that SL1-receiving messages are expressed without an SL1 leader. Surprisingly, when overexpressed, SL2 RNA, which performs a distinct function from that of SL1 RNA in wild-type animals, can rescue the lethality of embryos lacking SL1 RNA. Moreover, in these mutant embryos, we detect SL2 instead of SL1 leaders on normally SL1-trans-spliced messages; this result suggests that the mechanism that discriminates between SL1 and SL2-trans-splicing may involve competition between SL1 and SL2-specific trans-splicing. Our findings demonstrate that SL1 RNA is essential for embryogenesis in C. elegans and that SL2 RNA can substitute for SL1 RNA in vivo. [Key Words: trans-splicing; C. elegans; RNA processing; embryogenesis; ribosomal RNA] Received March 28, 1996; revised version accepted April 24, 1996.

Trans-splicing of a small 22- to 39-nucleotide RNA Although the trans-splicing process and certain struc- leader sequence onto the 5' end of mRNAs is a mecha- tural features of the SL RNAs are conserved in organisms nism of mRNA maturation that occurs in lower eukary- that perform trans-splicing, the functions of trans-splic- otes, including trypanosomes, nematodes, trematodes, ing are not fully understood. In some cases, however, and Euglena (for review, see Agabian 1990; Nilsen 1993). trans-splicing is known to result in the production of The spliced leader RNA (SL RNA), the precursor RNA functional mRNAs. For example, trans-splicing in trypa- containing the trans-spliced leader sequence, appears to nosomes apparently serves to process polycistronic participate in trans-splicing as an SL ribonucleoprotein mRNAs into monocistronic units; as a result, all (RNP)(Bruzik et al. 1988; Thomas et al. 1988; Van Doren mRNAs receive a spliced leader (Walder et al. 1986; and Hirsh 1988; Maroney et al. 1990; Michaeli et al. Agabian 1990).Similarly, in the nematode Caenorha bdi- 1990; Palfi et al. 1991). The nematode SL RNP protein tis elegans, processing of polycistronic messages derived components include the Sm proteins, which also associ- from the fraction of transcription units that are orga- ate with most U snRNAs involved in cis-splicing (Bruzik nized into operons appears to occur by trans-splicing of et al. 1988; Mattaj 1988; Thomas et al. 1988; Van Doren the downstream messages to the minor spliced leader and Hirsh 1988; Maroney et al. 1990). Other RNP com- SL2 (for review, see Blumenthal 1995). ponents known to participate in cis-splicing, such as the Although it is clear that trans-splicing acts to process U2 and U6 snRNPs, are also required for trans-splicing polycistronic messages, it is also likely to perform other in both trypanosomes and nematodes (Tschudi and Ullu functions. Only -10% of all mRNAs in C. elegans re- 1990; Hannon et al. 1991; Watkins and Agabian 199 1). ceive SL2 and only -25% of all C. elegans genes appear to be organized into operons (Zorio et al. 1994; Ross et al. 1995).Nematode genera other than Caenorhabchtis have 'Corresponding author. not been found to contain operons or an SL2-like leader

GENES & DEVELOPMENT 10:1543-1556 0 1996 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/96 $5.00 1543 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Eerguson et al.

RNA; mRNAs from most of these species are trans- mally SL1 -trans-spliced mRNAs produce functional spliced to a leader that is identical to the major trans- products even in the absence of zygotic SL1. spliced leader, SL1, in C. elegans (Krauseand Hirsh 1987; Nilsen 1993). Although the function of SL1-trans-splicing is not known, -60% of C. elegans messages are Results trans-spliced to this leader (Zorio et al. 1994))implying Identification of mutations that delete the rrs-1 gene that SL1 performs other functions beyond the processing cluster and isolation of rescuing clones of operons. In addition to SL1 and SL2, other spliced leaders have been characterized recently in C. elegans In an effort to identify zygotically transcribed genes es- (Ross et al. 1995). These novel spliced leaders are most sential for early stages of embryonic development in C. similar to SL2, but their functions are not generally un- elegans, we isolated two allelic loss-of-function muta- derstood (Ross et al. 1995).Given the existence of these tions (e2482 and wl ) that lead to early defects in embryo- other SL RNAs, it is conceivable that there may be ad- genesis and late embryonic arrest. To investigate the ditional roles for trans-splicing in C. elegans. function of the corresponding genomic region, we iden- Although the trans-splicing specificity of some of the tified molecular clones that rescued these mutations. SL RNAs has not been well characterized, mechanisms e2482 and wl were localized to an -500-kb interval by clearly exist to discriminate between SL1 and SL2-trans- mapping them relative to restriction site polymorphisms splicing. SL2 is not found on the most upstream mRNAs (Fig. lA,B). Pools of cosmid clones corresponding to this derived from polycistronic messages, and SL1 is not gen- interval were transformed into e2482/+ animals and erally found on downstream mRNAs of such operons tested for their ability to rescue embryonic lethality. (Spieth et al. 1993; Zorio et al. 1994).In addition, unlike From the one rescuing pool (pool 2, Fig. 1B; Table l), a SL2, SL1 is found on many monocistronic messages. The single cosmid (K07H12)was identified that is sufficient only requirement for trans-splicing of SL1 onto a pre- for rescue (Table 1). Rescue was indicated by a decrease mRNA appears to be the presence of an outron, an AU- in the fraction of arrested embryos produced by trans- rich sequence followed by a splice acceptor site but con- formed heterozygotes and the presence of viable e2482 taining no splice donor site (Conrad et al. 1991, 1993, homozygous larvae (Table 1). As expected, rescue was 1995).In contrast, splicing of SL2 onto downstream mes- never complete, as the transformed DNA assembles into sages in operons may be coupled to polyadenylation of an extrachromosomal array that is inefficiently trans- the upstream mRNA in the polycistronic message mitted during cell division (Mello et al. 199 1).Some res- (Spieth et al. 1993). cued animals survived to adulthood, however, and it was A number of observations suggest that the SL1 leader possible to propagate homozygous mutants for many might facilitate the expression of trans-spliced mRNAs. generations that were transgenic for the rescuing se- For example, in many cases, SL1 is spliced close to the quences. initiation codon of the protein coded by a trans-spliced Fragments derived from K07H12 were further tested mRNA (Bektesh et al. 1988), and it has therefore been for rescue, and an - 1-kb fragment (1-kblK07H12; Table proposed that SL1 may serve to promote efficient trans- 1)was found to rescue both e2482 and wl. The sequence lation of these messages (Bektesh et al. 1988; Spieth et al. of a 985-bp subclone of this fragment (not shown) was 1993). Consistent with this notion, recent studies have found to be nearly identical to the published sequence of shown that the SL1 leader can enhance translation in a 980-bp repeat from the 110-kb rrs-1 cluster (Nelson and vitro (Maroney et al. 1995).Additional studies have be- Honda 1985). Previous analysis of repeats from the rrs-1 gun to identify the RNA sequences and protein compo- cluster showed that each contains one gene encoding 5s nents required for trans-splicing in vitro and in vivo (for rRNA and one encoding SL1 RNA, a 105-nucleotide review, see Agabian 1990; Nilsen 1993; Blumenthal RNA that includes the 22-nucleotide SL1 leader (Nelson 1995); however, the requirement for neither trans-splic- and Honda 1985; Krause and Hirsh 1987)(Fig. 1C).There ing nor a trans-spliced leader have been demonstrated were 21 bp that differed between the 985-bp sequence of directly. our rescuing subclone and the published sequence of the We describe the characterization of mutations that 980-bp repeat. Significantly, the 5s rRNA and SL1 RNA eliminate the tandemly duplicated copies of the SL1 coding regions were identical between the two se- RNA gene and demonstrate that SL1 RNA is essential quences. These observations suggest that the 985-bp sub- for normal embryonic development and viability. Fur- clone that rescues e2482 and wl does not encode an ac- thermore, we show that the essential function of SL1 tivity that is unique from other rrs-1 repeats, consistent RNA can be provided by SL2 RNA. Elimination of zy- with our finding that other repeats from the cluster were gotic SL1 RNA results in the inappropriate trans-splicing also able to rescue these mutations (not shown). These of the SL2 leader onto normally SL1-trans-spliced results also suggested that wl and e2482 might be dele- mRNAs, suggesting that the mechanism that discrimi- tions that remove some or all of the rrs-1 repeats. nates between SL1 and SL2-trans-splicing may involve To examine whether wl and e2482 are deletions of the competition between the two RNAs for trans-splicing rrs-1 cluster, genomic DNA prepared from homozygous machinery. This observation may explain our finding mutant embryos was analyzed. Southern blot analysis that whereas the lack of zygotic SL1 RNA affects many showed that few or no rrs-1 repeats were present in ei- aspects of embryonic development, a number of nor- ther mutant (Fig. 2A-C). Moreover, analysis of genomic

1544 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

C. elegans SL1 RNA is essential for embryogenesis

VR unit Figure 1. Genetic and physical maps of him-5 unc-76 rrs-1 dpy-21 par4 - - - - the region to which e2482 and wl map. (A) 0. Genetic map of a portion of the right arm 0 . .. 0 . .. of linkage group V. e2482 and wl map ap- 0 ...... proximately halfway between the genes unc-76 and dpy-21. The location of the R12B9 rrs-1 cluster was determined previously by -W04GlO rrs-1 cluster -F16GS mapping of the position of a polymorphism -50kb that is immediately adjacent to the cluster --K07H12 I I (Nelsonand Honda 1985).(B) Physical map u - WP~ ,I'\ stplos WP~ of a segment of the region shown in A. wP1, stP108, wP3, and wP2 are polymorphism~used to map e2482 and wl. The / \ \ cosmid clones contained in the rescuing ,/ SL1 5s rRNA \ RESCUE \ pool 2 (K07H12, F16G5, W04G10, and 4-- + R12B9) and the position of the 110-kbrrs-1 cluster are shown above the line. (C)Map 1 SI;I site 985 of the 985-bp rescuing subclone derived from cosmid K07H12. The fragment is ar- bitrarily numbered from nucleotide 1, the first nucleotide of the left BamliI site, to nucleotide 985, the last nucleotide of the A SmA3 AATGG - Sm Site AATTTTGG right BamHI site. The location (boxes)and b~rn-~luAA GACGCGTC TGG - direction of transcription (arrows)of the SL1 (nucleotides 127-231) and 5s rRNA (nucleotides408-526) transcription units, the position of the 435-bp SLl+, 5s-rescuing fragment, and the position of the 740-bp SL1 -, 5s + fragment are shown. (B)BamHI; (Bm)BsmI; (Br)BseRI; (T)ThaI. The table (right)indicates whether the clones shown in C and D are able to rescue the embryonic lethality of e2482. (D)Sequences of the wild-type Sm-binding site in SL1 and the mutant Sm sites in the constructs SmA3 and Sm-Mlu. Shadowing indicates extra nucleotides derived from ligation of the MluI linker.

DNA digested with a restriction enzyme that releases sufficient quantities of maternally supplied 5s rRNA to the rrs-1 cluster as high molecular weight fragments de- complete embryogenesis and hatch. In contrast, a 740-bp tected no remnants of the rrs-1 cluster in either wl or fragment containing the 5s rRNA gene alone (SL1- / e2482 embryos (Fig. 2D).Thus, both mutations appear to 5s +, Fig. 1C)did not rescue embryonic lethality (Table delete the entire rrs-1 cluster. 1).In addition, the development of 12 homozygous mutant embryos from these transgenic lines was followed and none of the mutant phenotypes appeared to be de- The SL1 RNA-encoding gene is necessary to rescue tectably altered by the presence of the 5s rDNA (Fig. the embryonic lethality of deletions of the rrs-1 cluster 3BID).Thus, zygotic 5S rRNA is neither necessary nor To test whether the embryonic lethality caused by these sufficient to rescue these mutants. rrs-1 deletions results from the lack of SL1 RNA, 5s rRNA, or both, constructs encoding each of these RNAs The Sm binding site in SLI RNA is essential were tested individually. A construct containing a 435- bp sequence from the subcloned rrs-1 repeat of K07H12, for embryonic viability including the entire SLl gene and only 25 nucleotides of SL1 RNA contains a highly conserved Sm binding site the 5s rRNA coding region (SL1+ 15s - , Fig. 1C), was that binds nematode and mammalian Sm proteins found to be sufficient to rescue the embryonic lethality (Bruzik et al. 1988; Thomas et al. 1988; Van Doren and of both alleles (Table 1).Embryonic lethality was rescued Hirsh 1988). The Sm binding site is present in many with approximately the same efficiency as with the en- trans-spliced leader RNAs and in most snRNAs that di- tire 1-kb rrs-1 repeat (Table 1). Unlike the animals res- rect cis-splicing (Bruzik et al. 1988; Mattaj 1988; Nilsen cued with the 1-kb repeat, however, most of the defi- et al. 1989; Ross et al. 1995). The Sm site, and the pro- ciency homozygotes rescued with the SL1 gene alone teins that bind to this site, appear to be required for var- died early in larval development with no other obvious ious aspects of small nuclear RNP (snRNP) assembly, phenotypes, and none survived to adulthood. Of 10 res- localization, and stability (Mattaj 1988; Jones and Guth- cued embryos followed throughout embryogenesis, all rie 1990; Rymond 1993; Roy et al. 1995).Mutation of the embryos looked similar to wild-type embryos and under- Sm-binding site in the SL RNA of the nematode Ascaris went normal differentiation and morphogenesis, demon- lumbricoides abolishes trans-splicing in vitro (Maroney strating that the SL1 gene rescues all conspicuous em- et al. 1990). Additional experiments demonstrated that bryonic defects (Fig. 3A,C). This observation also dem- two elements in the Ascaris SL RNA are sufficient for onstrates that zygotic 5s rRNA is not required until trans-splicing activity in vitro; one of these elements postembryonic stages, implying that embryos contain includes the Sm-binding site (Hannon et al. 1992). We

GENES & DEVELOPMENT 1545 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Ferguson et al.

Table 1. Transformation rescue experiments

------. - - - Percent rescue of embryonic lethality No. of rescued lines1 Clone tested total no. of lines e2482 (no.)" wl (no.)" Embryonic rescueb

None N.A. 0 (1688) N.A. Pool 2 212 N.D. + K07H12 212 N.D. + 1 kbIK07H12 10113 36 (2287) + SL1+ 15s - 415 35 (1476) + SLl - /5S + 017 N.D. - SmA3 0120 N.D. Sm-M lu 018 N.D. Heterozygous e2482 or wl animals were used for transformation experiments (see Materials and methods). The average percentages of arrested embryos for the parental lines were 25.3% for e2482 and 25.9% for wl,as expected of heterozygous animals carrying a recessive lethal mutation. Percent rescue was calculated using the formula lOO[l - (average percent arrested embryos observed1 average percent arrested embryos from the parental strain)].(N.A.) Not applicable; (N.D.)not determined. "Total number of progeny counted, including arrested embryos and larvae. b~mbryonicrescue was scored as + if the average percentage of arrested embryos was below 20%, and arrested elongated embryos andlor extra Unc-76 larvae (significantly greater than the -2% Unc-76 recombinants usually observed in the parental strain) were observed (see Materials and methods). "Lines that showed no apparent rescue with the 1-kblK07H12 and SL1+ 5s fragments were not used in the calculation of percent rescue. d~lthoughthe 740-bp fragment used in this transformation experiment does not by itself rescue the embryonic lethality of e2482, it appears to provide sufficient levels of 5s rRNA to rescue the absence of 5s rRNA in the deletion mutants. This was demonstrated by transforming the 740-bp fragment together with the 435-bp SL1 fragment. In all seven lines obtained from transformation of the two fragments, transformants gave rise to extra fertile Unc-76 adults. These animals proved to be rescued homomzygous mutants, as they produced a higher percentage (>40%)of arrested e2482 embryos, than do the Unc-76 recombinants (-25% arrested embryos).

created two mutations in the Sm binding site of the C. genes were selected for these studies. The messages en- elegans SL1 RNA (Fig. ID) to test whether this con- coded by each of these genes are normally trans-spliced served sequence is required for rescue of the rrs-1 dele- to SL1 (Krause et al. 1990; Spieth et al. 1991; Kennedy et tion mutants. One mutation is a small deletion that re- al. 1993; Okkema et al. 1993),and each is embryonically moves 3 of the 8 bases comprising this site (SmA3); the transcribed (Waterston 1989; Krause et al. 1990; Spieth other is a substitution of these 3 bases with an 8-base et al. 1991; Kennedy et al. 1993). Moreover, there is no sequence (Sm-Mlu) (Fig. ID). We found that both SmA3 evidence of a maternal requirement for any of these gene and Sm-Mlu abolished rescue in many lines tested (Ta- products. Thus, analysis of these messages should reveal ble 1).These results imply that SL1 RNA is the relevant whether zygotic messages are SL1-trans-spliced. The fate rescuing activity within the 435-bp fragment and dem- of each message was followed in mutant embryos by use onstrate that the Sm site is required for its function, of a PCR-based assay that detects SL1-trans-spliced mes- consistent with the notion that trans-splicing of the SL1 sages. We found that although SL1 was always detected leader to mRNAs is an essential process. Thus, SL1 RNA on each of these messages in wild-type RNA samples, in appears to be both necessary and sufficient for rescue of most samples of mutant embryo RNA tested, SL1 was the embryonic lethality of rrs-1 deletions. not detected on any of these messages (Fig. 4; Table 2). Thus, the absence of zygotic SL1 RNA correlates with the absence of an SLl leader sequence from at least four The SLI leader is absent from normally SL1 -trans- normally SL 1-trans-spliced mRNAs. spliced messages in mutants lacking zygotic SLI RNA The two rrs-1 deletions appear to remove all SLl-encod- Absence of zygotic SLI RNA leads to pleiotropic ing genes and presumably eliminate all zygotically sup- defects during embryonic development plied SL1 RNA. There appears to be a substantial pool of The finding that SL1-trans-splicing does not appear to maternal SL1 RNA in early embryos (G. Seydoux, pers. occur in rrs-1 deletion mutants suggested that the func- comm.), however, and it is conceivable that this pool tion of zygotic messages requiring SL1-trans-splicing for might be sufficient to sustain some SL1-trans-splicing of their proper expression may be reflected in the pheno- zygotically transcribed messages. To determine whether type of embryos carrying these deletions. Consistent SL1-trans-splicing occurs in embryos lacking zygotic with the prevalence of messages that are trans-spliced to SL1 RNA, we investigated whether the SL1 leader was the SL1 leader in C. elegans [-60% of all mature present on messages that are normally SL1-trans-spliced. mRNAs (Zorio et al. 1994)],we observe widespread de- The products of the ges-1, elt-1, hlh-1, and myo-3 fects in mutant embryos lacking the SL1-encoding genes

1546 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

C elegms SL1 RNA is essential for embryogenesis

Figure 3. Phenotypes of wild-type embryos, arrested mutant embryos, and embryos rescued with an SL1 RNA-encoding construct. (A) Wild-type (N2]pretzel stage embryo. This embryo has elongated into a worm that will hatch shortly after this stage. (B) Homozygous e2482 terminally arrested embryo. A number of differentiated cell types can be seen. (Arrows) Epi- Figure 2. Analysis of the rrs-I cluster in wl, e2482, and wild- dermal nuclei; (arrowhead]programmed cell death corpse. The type animals. (A]Southern blot of genomic DNA probed with phenotypes of wl embryos are not distinguishable from those of the 985-bp rescuing repeat sequence. Genomic DNA from ani- e2482 embryos (not shown]. (C)e2482 embryo rescued with the mals of the indicated genotype [wl, e2482, N2 (wild-type)]was 435 bp SL1+ 15s - fragment. The embryo shown is at the pret- digested with BamHI, which fragments the rrs-1 cluster into zel stage, as in B. This embryo hatched and showed an Unc-76 -1-kb repeats (arrow).Quantitation of the amount of 1-kb re- phenotype, implying that it carried e2482, which is marked peat from each mutant indicated that the wl DNA preparation with the closely linked unc-76(e911)mutation. After hatching, contained -2% of the wild-type level of 1-kb repeat, and the this animal died early in larval development, as expected of an e2482 preparation contained -8% of the wild-type level (not e2482 homozygote rescued with the SL1 gene, but without the shown). The presence of rrs-l sequences in the mutant DNA is 5s rRNA gene. (Dl A representative e2482 embryo produced by attributable to contaminating wild-type DNA in the mutant an e2482 heterozygote transgenic for the 740-bp SL1- 15S+ preparations. (B]A different region of the blot shown in A ex- construct. It was not practical to determine whether this par- posed for a longer period to allow detection of another (low-copy ticular embryo carries the transformed DNA, as the lethal phe- number) genomic sequence that hybridizes to the repeat se- notype precludes scoring of the transgenic marker (rol-6)gene. quence. This low-copy sequence was used as a control to deter- However, like the untransformed parent strain, the transformed mine whether any or all of the repeats detected in A derive from heterozygotes produced -25% arrested embryos (Table 1), and contaminating wild-type DNA in the mutant embryo DNA all of the arrested embryos showed terminal phenotypes that preparations. Shown is a representative control fragment (2.8 were indistinguishable from the phenotypes of nontransformed kb, arrow) that flanks the cluster, as determined by Southern e2482 embryos. Bar, - 10 pm. analysis of flanking cosmids (not shown). The amount of this fragment in the mutant DNA preparations corresponded to much less than one copy per diploid genome (not shown). This indicates that this sequence is deleted in both mutants and can (Figs. 3B, 5, and 6). The earliest phenotypes that we dis- therefore be used as a control for the amount of contaminating cern in both e2482 and wl mutants include defects in wild-type DNA in the mutant DNA preparations. Quantita- cell migrations associated with gastrulation (Fig. SA-F] tions showed that the ratio of the amount of this fragment in the mutant DNA to that in wild-type was comparable to the and premature cell division in the early endoderm lin- ratio seen for the 1-kb rrs-l repeat. This finding indicates that eage. Both phenotypes appear reproducibly at the onset both mutants contain very few, or no, rrs-1 repeats. Positions of of gastrulation (-28-cell stage), indicating an early re- molecular weight markers are shown at the left in kilobases. (C] quirement for zygotic SL1 RNA. Interestingly, it is at The blot shown in A and B was stripped and reprobed with approximately this stage in embryogenesis that the ma- cosmid ZC130, which contains sequences from chromosome I. ternal pool of SL1 RNA begins to decrease significantly This provided a control that was used to normalize for the total (G. Seydoux, pers. comm.). We also observe a defect in amount of DNA loaded in each lane. (DlAnalysis of wl,e2482, compaction during mid-embryogenesis: In mutant em- and wild-type (N2)DNA preparations digested with PstI, which bryos, blastomeres fail to adhere properly at the stage in does not cut within a typical rrs-1 repeat. Positions of molecular which they normally become tightly packed in wild-type weight markers are shown at the left in kilobases. Small amounts of wild-type high molecular weight fragments were embryos (Fig. SG,H).In addition, defects in specification detected after longer exposures (not shown); this is attributable of certain cell types are observed: Some cells (e.g., the to contaminating wild-type DNA. Because neither allele con- locomotory muscles of the body wall] are made at re- tained any smaller rrs-1-hybridizing fragments detectable even duced numbers, and others (e.g., muscles of the feeding after long exposures, the rrs-1 cluster appears to be deleted com- organ, or pharynx] are often absent entirely (Fig. 6A-Dl. pletely in both mutants. Finally, the mutant embryos uniformly fail to undergo

GENES & DEVELOPMENT 1547 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Ferguson et al.

controls wl embryos [e.g., Fig. 3BJ, even though ced-3 mRNA is wt mut normally SL1-trans-spliced (Yuan et al. 1993).These results indicated that the SL1 leader may not be essential RT I+ +'++Ir= I+ +I for the expression of some messages that are normally SL 1-trans-spliced.

Normally SLI -trans-spliced messages are expressed in the absence of the SL1 leader As reported above, at least some zygotic messages that are trans-spliced to SL1 in wild-type embryos do not re- hlh- 1 ceive SL1 in mutants lacking the SLl-encoding genes. To I test whether these messages are translated in the absence of the SLl leader, we analyzed for the presence of the hlh-I, myo-3, and ges-1 gene products in mutant embryos. High levels of immunoreactive protein encoded by the hlh-1 and myo-3 genes were seen in the appropriate cells [Fig. 7A-D). Moreover, the gut esterase encoded by the ges-1 gene (Edgar and McGhee 1986) was observed at high levels in the intestine (Fig. 7E,F). These results indicate that the lack of zygotic SL1 RNA does not dra- matically attenuate the expression of these mRNAs.

The selectivity of SL2-trans-splicingis altered in mutants lacking zygotic SLl RNA The ability of the ges-1, myo-3, and hlh-1 messages to be Figure 4. RT-PCR analysis of SLl -trans-splicing of embryonic messages in wild-type (N2)and e2482 mutant embryos. For each message, two representative samples are shown for each geno- Table 2. RT-PCR analysis of SLI -trans-splicing in wild-type type. A + or - sign above the blot indicates whether a reverse and mutant embryos transcription was performed. - RT indicates RNA was used di- No. of samples positive for rectly in a PCR reaction. The same RNA samples were used for SL1-trans-splicingltotalno. samplesa the + RT and - RT reactions. Controls in which only E. coli was used to make the extracts are indicated as - emb. RT-PCR mRNA tested N2 (wild type) e248zb was then performed on these mock extracts (see Materials and methods). For elt-1 and myo-3, arrows point to the positions of ges-l 515 014 the expected products; the origin of the other product in each hlh-1 4/4 0/9 case is not known. [Lanes 1, 2)Wild-type SL1; (lanes 34, e2482 myo-3 12/12 116' SLl; (lanes 5,6) wild-type SL1, - RT controlsi (lanes 7,8) e2482 elt-ld 8/8 2/1OC SL1, - RT controls, (lanes 9,10) no embryonic RNA added, + "For each wild-type and e2482 sample listed, a portion of the RT-PCR controls. Positions of molecular weight markers of the same RNA sample was used in a control PCR reaction in which indicated size in base pairs are shown at the right. the reverse transcriptase (RT)step was omitted ( - RT lanes, Fig. 4; see Materials and methods). In addition, two control reactions were performed for each message, in which a portion of a normal morphogenesis later in embryogenesis and arrest mock extract containing no embryonic RNA was added to the RT-PCR reaction ( - emb lanes, Fig. 4; see Materials and meth- as unelongated masses of differentiated cells that are not ods). A product was not detected in any of these controls. fully enclosed by epidermal tissue (Fig.3BJ. This range of b~oreach e2482 sample in which an SL1 product was not de- phenotypes is consistent with an essential role of zygotic tected, a product was found to be amplified either with primers SL1 in the expression or metabolism of multiple embry- specific for an internal portion of the messages (not shown), or onic mRNAs. with SL2 (Fig. 8; Table 31, confirming that these messages were Although embryos lacking the SL1-encoding gene present in the samples. cluster show severe developmental defects, some of the 'The SL1 products detected in the e2482 samples for elt-1 and phenotypes of e2482 and wl embryos suggest that ex- myo-3 messages may reflect infrequent SL1-trans-splicing of pression of a subset of mRNAs normally receiving SL1 these messages or might instead be the result of contaminating may not be profoundly affected in these mutants. Some wild-type embryos in the selected population (seeMaterials and cell types, such as the germ-line precursors and the in- methods). *1n one set of experiments for elt-1 (218 wild-type reactions and testine, appear to be properly specified in e2482 and wl 2/10 e2482 reactions), DNase was not added. In subsequent ex- embryos (e.g., Fig. 6E-HI. In addition, programmed cell periments, products were detected in the -RT control reac- deaths, which require the zygotically expressed ced-3 tions for some other messages, so the DNase step was added (see gene product (Ellis and Horvitz 1986),occur in e2482 and Materials and methods].

1548 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

C. elegans SL1 RNA is essential for embryogenesis

Alternatively, another trans-spliced leader, such as SL2, might substitute for SL1 in its absence. Although SLl and SL2 are spliced to the same acceptor site consensus sequence, SL2 is thought to be specifically involved in processing downstream cistrons from polycistronic messages; it is not found on the 5' end of the first mRNA derived from a polycistronic message (Huang and Hirsh

Figure 5. Migration and adhesion defects associated with the lack of zygotic SL1 RNA. (A-D] Migration of the endoderm progenitor cells. The daughters of the endoderm progenitor cell (Ea and Ep] and the germ-line precursor (P,] have begun migrat- ing into the interior of a wild-type embryo (A)but remain on the surface of an e2482 embryo (B], -100 min postfertllization. (C] A wild-type embryo -120 min postfertilization after Ea and Ep have divided; all four granddaughters of E have moved into the interior of the embryo. (D)An e2482 embryo at the same stage as shown in C; all E granddaughters remain on the surface of the embryo (Eaa and Epa are out of the plane of focus]. (E,F]Tracings of positions of the muscle cell progenitor D and its descendants (Da, Daa, Daaa, and Daaap], during development from 100-225 Figure 6. (A-H) Analysis of cell-type specific markers. min after first cleavage (wild-type]and 10W00 min after first [(A,C,E,G] wild-type embryo; (B,D,F,H) terminally arrested cleavage (~2482).The largest circle represents the largest cell c2482 embryo]. JA,B]Embryos stained with antibody NE814C6, (Dl; the smallest circle represents the smallest cell (Daaap).Ar- which recognizes all body-wall muscle cells. Muscles are orga- rows indicate the general direction of cell movements. (El Mi- nized into rows at this stage in wild-type embryos. Reduced gration of D cells in a wild-type embryo; as shown, the cells numbers of body-wall muscle cells are observed in mutant em- move anteriorly and also into the interior of the embryo. (F]D bryos, and cells are located primarily in the posterior region of cell migration in an e2482 embryo; the movement of the cells the embryo. (C,D] Embryos stained with antibody 3NB12, appears largely random, and cells remained on the surface of the which recognizes a subset of pharyngeal muscle cells. Few or no embryo throughout most of the period followed. This migration 3NB12-positive cells are observed in mutant embryos; two are defect was observed in all eight e2482 embryos examined. (G,H) seen in this embryo (arrows; Dl. (E,F] Embryos stained with Cell adhesion defects. (G)Cells on the surface of wild-type em- antibody 1CB4, which recognizes all 20 intestinal cells, intes- bryos at -180 min postfertilization appear to undergo a com- tinal-rectal valve cells (arrow), pharyngeal gland cells (not paction process. (H)Cells on the surface of e2482 embryos at shown), and a few neurons (not shown). Approximately the about the same time as in G appear uncompacted and rounded proper number of intestinal cells are observed in mutant em- up. Bar, -10 pm. bryos, however, morphogenesis of the intestine is aberrant. In- testinal-rectal valve cells are present. (G,H) Embryos stained with antibody OIClD4, which stains the two germ-line precursors (arrows).Although these precursors are present in mutant expressed in the absence of an SL1 leader may indicate embryos they do not migrate to their proper location. Bar, that trans-splicing is not required for their expression. -10 km.

GENES &. DEVELOPMENT 1549 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Eerguson et al.

bryo, muscle cells are clustered in the posterior, presumably because of cell migration defects {cf. Fig. 5F). The mutant embryos express fewer than the wild-type number of hlh-I-positive muscle cells. On the basis of similar results with other markers (e.g., Fig. 6B), this defect does not appear to be specific for the hlh-I product. (C,D]Expression of the myo-3 product in a wild-type comma stage embryo and a terminally arrested e2482 embryo, respec- tively. The muscles are clustered at the posterior of the mutant embryo, and fewer than normal numbers are seen, as described above. (E,F)Embryos stained for the gut esterase activity of the ges-l product. Expression is seen in a wild-type comma stage embryo and a terminally arrested e2482 embryo. Bar, - 10 pm.

1989; Spieth et al. 1993; Zorio et al. 1994). In contrast, are instead provided by SL2 in its absence, thereby al- SL1 is spliced onto a larger subset of messages but is lowing the expression of the aberrantly SL2-trans-spliced infrequently spliced onto downstream messages in oper- messages. The embryonic lethality associated with dele- ons (Spieth et al. 1993; Zorio et al. 1994). Thus, a selection of the SL1-encoding genes, however, demonstrates tion mechanism can apparently discriminate between that although SL2 is trans-spliced onto messages that SL1 and SL2-trans-splicing, depending on the location of the splice site in the pre-mRNA sequence (Spieth et al. 1993).The mechanism of this discrimination process is unknown. To determine whether the discrimination of SL2- trans-splicing is altered in cells lacking SL1, we examined messages that lack SL1 for the presence of the SL2 leader, using the PCR-based assay described above. Pre- vious studies have found that although SL1 is detectable on ges-1, hlh-1, elt-I, and myo-3 messages in wild-type (N2) embryos, SL2 is not detected on these messages (Krause et al. 1990; Spieth et al. 1991; Kennedy et al. 1993; Okkema et al. 1993) (Fig. 8; Table 3; data not shown). Surprisingly, however, we found that SL2 is fre- quently detectable on the elt-1 and myo-3 messages (Fig. 8; Table 3) in embryos lacking zygotic SL1 RNA. We also detected SL2 on the hlh-I and ges-1 messages, albeit less consistently (data not shown]. The finding that the SL2 leader is trans-spliced inappropriately in rrs-1 deletion mutants indicates that the selection mechanism that Figure 8. RT-PCR analysis of SL2-tmns-splicing in wild-type prevents trans-splicing of SL2 onto the 5' end of unproc- (N2)and e2482 mutant embryos. For each message, two repre- essed transcripts can be overridden in the absence of SL1 sentative samples are shown for each genotype. The arrows RNA. point to the correct size product; the origin of the other products in each case is not known. The reverse transcription reactions SL2 can substitute for SL1 when expressed from a were the same as were used for the SL1 analysis (Fig. 4). (See multicopy extrachromosomal array Table 3 for an explanation of the controls.)(Lanes 1,2)wild-type SL2; (lanes3,4) e2482 SL2. Positions of molecular weight mark- It is conceivable that functions normally supplied by SL1 ers of the indicated size in base pairs are shown at the left.

1550 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

C. eleguns SL1 RNA is essential for embryogenesis

Table 3: RT-PCR analysis of SL2-trans-splicingin wild-type thality of e2482 embryos with approximately the same and mutant embryos efficiency as the SLI-encoding fragment (Fig. 9; Table 4). No. of samples positive for As with the SL1 gene alone, most rescued animals ar- SL2-trans-splicingltotalno. samplesa rested as early larvae. Primer extension analysis of RNA prepared from one rescuing transgenic line confirmed mRNA tested N2 (wild type) e2482 that SL2 is overexpressed from the extrachromosomal array (data not shown). In addition, when transgenic lines containing the SL2-encoding construct, as well as the 740-bp 5s rRNA fragment, were established, a frac- "The same RT reactions used to generate the SL1 data in Table tion of the rescued animals reached adulthood; these 2 and Fig. 4 were used in these analyses. The - RT controls for adults were invariably sterile. These findings imply that these experiments were performed using the SL1 primer and appropriate downstream primer as described in Table 2. Two SL2 RNA can substitute functionally for SL1 RNA dur- controls were performed for each message in which a portion of ing embryonic and most of postembryonic development a mock extract containing no embryonic RNA was added to the when expressed at higher than normal levels. The ability RT-PCR reaction using the SL2 primer and the appropriate of SL2 to substitute for SL1, however, appears to be in- downstream primer; a product was not detected in any of these complete; unlike an SL1-encoding transgene, the SL2 controls (see Materials and methods]. gene apparently cannot provide the function required for b~hedetection of SL2 in two of the samples might be attribut- fertility. able to rare trans-splicing of SL2 onto myo-3. This has not been reported previously (Okkema et al. 1993); however, such a rare trans-splicing event might be detectable in our experiments be- Discussion cause of the large number of samples analyzed. In two additional wild-type samples, a product was detected; however, the We report five major findings: (1)Zygotic SLl is essential signal was not reproducible in multiple PCR attempts. There- for a number of events in embryogenesis and for embry- fore, these samples were excluded from the above data. onic viability in C. elegans; (2)5s rRNA is not required 'For each e2482 sample in which a product was not detected, zygotically in the embryo but is required for postembry- control amplifications were performed using primers specific onic development; (3)some messages that are normally for an internal portion of the messages to confirm the presence SL1-trans-spliced are expressed even without an SL1 of the elt-1 messages in the samples (not shown). leader; (4)in the absence of SL1 RNA, SL2 is inappropriately trans-spliced onto messages that are normally trans-spliced only to SLl; and (5) SL2 can substitute functionally for SL1. normally receive only SL1, SL2 cannot substitute fully for SL1 when expressed at normal levels. Because SL2 appears to be trans-spliced less efficiently onto some Trans-splicing of SLI appears to be essential messages such as hlh-1 andges-1, it is possible that some for embryogenesis messages may not receive enough SL2 to be properly We have demonstrated that zygotic SL1 RNA performs expressed in rrs-1 deletion mutants. Unlike the SL1-en- an essential function beginning early in C. elegans em- coding genes, which are present in >lo0 tandemly re- bryogenesis. An SL1 RNA-encoding clone is necessary peated copies, there are few copies of the SL2 gene in the and sufficient to rescue the embryonic lethality of dele- C. elegans genome (Huang and Hirsh 1989; Ross et al. tions that encompass the entire 110-kb rrs-1 cluster. Our 1995);the failure of SL2 RNA to substitute for SL1 RNA findings also show that the maternal contribution of 5s might simply reflect lower levels of the former. Alterna- rRNA to the oocyte appears to be sufficient to sustain tively, since SL2 is only -45% identical to SL1 (Huang embryonic development to completion. and Hirsh 1989), the differences between the sequences of these two leaders might account for the inability of SL2 to substitute for SL1 in rrs-1 deletion mutants. To investigate further whether SL1 RNA supplies a function that cannot be provided by SL2 RNA, we asked whether an SL2-encoding gene could rescue deletions of the rrs-1 cluster when expressed at elevated levels from an extrachromosomal array. When exogenous DNA is introduced into worms by transformation, it is generally in- corporated into an extrachromosomal element carrying multiple copies of the DNA; transgenes are presumably overexpressed from such extrachromosomal arrays. Figure 9. Rescue of e2482 mutant embryos with an SL2-encoding construct. (A) An e2482 embryo rescued with the SL2 A construct encoding the 110-nucleotide SL2 RNA RNA-encoding construct. The embryo shown is at the pretzel was introduced into e2482/+ animals and stable trans- stage as in Fig. 3, A and C. This embryo hatched, exhibited an genic lines established. Transgenic lines were tested for Unc-76 phenotype, and died early in larval development (c.f. Fig. rescue of embryonic lethality. Unexpectedly, we found 3C).(B] Terminally arrested e2482 embryo, shown for compar- that the SL2-encoding gene rescued the embryonic le- ison. Bar, -10 pm.

GENES & DEVELOPMENT 1551 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Ferguson et al.

Table 4: SL2 transformation rescue experiments trans-splicing onto all normally SL1-trans-spliced messages, or SL2 may not recognize SL1 splice acceptor sites Percent rescue of Clone No. of rescued lines/ e2482 embryonic Embryonic efficiently, as suggested by our observations that SL2 is tested total no. of lines lethality (no.] rescue found only sporadically on some messages in rrs-1 deletion mutants (not shown). None N.A. 0 (2039) N.A. SL2 15/22 42 (4769) + -- Zygotic SLI RNA is necessary for multiple events Scoring of embryonic rescue was performed and percent rescue in em bryogenesis was calculated as described in Table 1. (N.A.)Not applicable. We found that multiple processes during embryogenesis are affected in mutants lacking zygotic SL1 RNA. For example, some cell types, such as epidermis (not shown) It is possible that unspliced SL1 RNA may perform an and body-wall muscle (Fig. 61, are made at fewer than essential function distinct from its role in SL1-trans- normal numbers as determined by analysis of cell fate- splicing per se, as has been shown to be the case for the specific markers. In the case of body-wall muscle, at U1 snRNP, which participates in cis-splicing (Wassar- least three muscle-specific markers, including the prod- man and Steitz 1993). Our results and previous data, ucts of the SL1-trans-spliced hlh-I and myo-3 mRNAs, however, are consistent with the view that trans-splic- are expressed; however, there are uniformly too few ing of the SL1 leader onto embryonic mRNAs is neces- muscle cells in mutant embryos. Experiments in which sary for its essential function. We have shown that res- individual muscle cell precursors were laser ablated in- cue of embryonic lethality depends on a functional Sm- dicated that this failure to make muscles is preferen- binding site. This site is well-conserved in all nematode tially restricted to particular muscle cell lineages [J.H. SL RNAs and in U snRNAs from fungi to vertebrates Rothman, unpubl.). Approximately normal numbers of (Bruzik et al. 1988; Guthrie and Patterson 1988; Mattaj muscle cells are made from the C and D muscle precur- 1988; Huang and Hirsh 1989; Nilsen et al. 1989) and is sors in these mutants; however, the MS muscle progen- required for trans-splicing in vitro (Maroney et al. 1990; itor, which normally gives rise to 28 muscle cells, pro- Hannon et al. 1992). Previous studies have determined duces very few muscles in mutant embryos. Schnabel that -70% of mRNAs receive a spliced leader (Zorio et (1995)reported that cellular interactions are required to al. 1994);we have shown that at least some messages fail specify muscles arising from MS. Thus, there may be a to be trans-spliced to SL1 in rrs-1 deletion embryos. SL 1-trans-spliced message required in this signaling pro- Therefore, our findings demonstrate that trans-splicing cess that is affected by the absence of SL1 in rrs-1 dele- does not occur properly in rrs-l deletion embryos, likely tion mutants. Other phenotypes that we have character- accounting for the phenotypes that we observe. ized, such as cell migration and cell adhesion defects, suggest that cell adhesion molecules may not be properly expressed in these mutants. Although SL1 has not been shown directly to regulate a cell-type-specific event, it is SL1 vs. SL2-trans-splicing noteworthy that another trans-spliced leader in C. ele- In rrs-1 deletion embryos, SL2 is spliced onto messages gans, SL4, is expressed specifically in the epidermis that normally receive only SL1 (Fig. 8). The selection (Ross et al. 1995), perhaps indicative of a cell-type-spe- process that governs SL1 versus SL2-trans-splicing in cific regulatory process. wild-type animals may involve competition between Our findings that SL1 is required for many aspects of SL1 RNPs and SL2 RNPs, the former having a greater C. elegans embryogenesis are consistent with the wide- affinity for the 5'-most splice acceptor site of a primary spread occurrence and apparent conservation of the transcript. Our results indicate that when SL1 is absent, trans-splicing process. RNA trans-splicing occurs natu- the SL2 RNP is able to recognize splice acceptor sites in rally in trypanosomes, Euglena, nematodes, and trema- the context of sequences to which the SL1 RNP is nor- todes (for review, see Agabian 1990; Nilsen 1993), and mally recruited; this observation is consistent with such mammalian cells are also competent to perform trans- a competition model. splicing when provided with the appropriate substrates The finding that SL2 can be spliced onto messages that (Bruzik and Maniatis 1992).Certain secondary structural are normally SL1-trans-spliced may account in part for elements are common to many known SL RNAs (Bruzik its ability to substitute functionally for SL1 in the em- et al. 1988; Nilsen et al. 1989; Nilsen 1993), and all bryo when overexpressed, despite the divergence in the known trans-spliced leaders in nematodes are identical sequence of the two spliced leaders. We found, however, to the C. elegans SL1 leader (Nilsen 1993), with the ex- that the SL2-rescued mutants grow more slowly than ception of the leader from the nematode Meloidogyne those rescued with SL1 (K. Ferguson et al., unpubl.) and incognita (which differs from SL1 at one position), C. that adults from the rescued lines are always sterile. elegans SL2, and a family of SL2-related spliced leaders Therefore, SL2 may not be capable of performing all of in C. elegans (Huang and Hirsh 1989; Ray et al. 1994; the functions normally provided by SL1. Alternatively, Ross et al. 1995). The observation that multiple pro- overexpression of SL2 from an extrachromosomal array cesses occurring during embryogenesis are affected in may not provide sufficient levels of SL2 RNA to allow mutants lacking zygotic SL1 RNA is consistent with a

1552 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

C. elegans SL1 RNA is essential for embryogenesis requirement for SL1 in the metabolism of multiple em- leader sequence may improve translatability of a mes- bryonic messages; the conservation of trans-splicing sug- sage (Maroney et al. 1995). Alternatively, trans-splicing gests that this requirement is likely to be true for other per se might function to remove inhibitory sequences in species. the 5'-untranslated region of primary transcripts, thereby allowing them to function. The availability of mutants lacking zygotic SL1 will make it possible to SL1 is not required for expression of some SL1- examine the sequences required for the proper function trans-spliced messages of the leader and to determine what role it might have in Given that -60% of messages receive SL1, it is not sur- metabolism or translation of a message in vivo. prising that the lack of SL1 RNA causes severe pleiotropic defects in the C. elegans embryo. Not all aspects of embryogenesis, however, are affected. Although mutant Materials and methods embryos develop more slowly than wild-type embryos, they arrest after the normal number of cell divisions Worm culture, strains, and anti bodies have occurred and appear healthy many hours after their The mutations described here were isolated from C. elegans var. arrest (not shown).In addition, many cell types appear to Bristol strain N2 (Brenner 1974). Culturing, mutagenesis, and be properly specified in these mutants (Fig. 6).Thus, dif- genetic manipulation of C, elegans were as described by Brenner ferentiation and other cellular functions, such as cell di- (1974). Nematodes were grown at 20°C in all experiments. vision and programmed cell death, are not generally Methods for mounting and viewing C, elegans embryos by Nomarski microscopy have been described previously (Sulston blocked. In addition, we have shown that products of et al. 1983). Embryos were fixed and stained for immunofluo- some normally SL1-trans-spliced messages are ex- rescence as described (Sulston and Hodgkin 1988).Mutant em- pressed. bryos were collected for Nomarski microscopy and immunoflu- There are several possible explanations for our find- orescence -14-18 hr after being laid, unless otherwise indi- ings that normally SL1 -trans-spliced messages appear to cated. Wild-type siblings have hatched by this time. Antibodies be expressed in rrs-1 deletion mutants. First, it is con- NE814C6 (Goh and Bogaert 1991 ), 3NB12 (Priess and Thomson ceivable that some normally trans-spliced messages, like 1987),and 1CB4 (Okamoto and Thomson 1985),were obtained the 30% of mRNAs that are not trans-spliced, may be from the MRC-Cambridge collection. OIClD4 (Strome and expressed efficiently without a trans-spliced leader, as Wood 1982) was a gift from Susan Strome. Antibody 5-6.1.1.1, has been observed for a modified version of the normally which recognizes the myo-3 gene product (Miller et al. 1983), was provided by David Miller 111, and the hlh-1 antibody (Krause SL1-trans-spliced rol-6 mRNA (Conrad et al. 1993)) or et al. 1990) was a gift from Michael Krause. Anti-rabbit and the absence of the SL1 leader may cause subtle defects in anti-mouse fluorescein-conjugated antibodies were obtained expression of these messages that cannot be detected by from Sigma. Embryos were stained for gut esterase activity, in- our methods. Although zygotic SL1 is essential, there is dicating the presence of the ges-1 gene product, as described as yet no direct evidence that any SL1-trans-spliced mes- previously (Edgar and McGhee 1986). The following mutations sage has an absolute requirement for a trans-spliced on linkage group V were used for mapping and balancing the leader to be expressed in vivo; analysis of additional gene rrs-1 deletion mutations: sma-1 (e30);rol-4(sc8); unc-61 (e228); products in the rrs-1 deletion mutants may identify mes- unc-76(e911);dpy-21 (e428). sages whose expression requires an SL1 leader. Second, some messages may be properly expressed in these mutants by receiving SL1 from the maternal pool of SL1 Isolation and mapping of the mutations RNA; however, this pool is not detectable in somatic e2482 was isolated in a general, genome-wide screen for ethyl- nuclei past the 50-cell stage (G. Seydoux, pers. comm.). methanesulfonate (EMS)-inducedzygotic embryonic lethal mu- Finally, the inappropriate trans-splicing of SL2 onto nor- tations that affect early events in embryogenesis. e2482 was mally SL1-trans-spliced messages, such as myo-3 and mapped between unc-76 and dpy-21 by standard three-factor elt-1, in rrs-1 deletion mutants may allow them to be recombination experiments (Sulston and Hodgkin 1988). Re- combinants between unc-76 and dpy-21 placed e2482 14/30 of the expressed. distance from unc-76 to dpy-21. wl was isolated in a targeted embryonic lethal screen designed to identify EMS-induced le- What is the essential role of SL1! thal mutations in the vicinity of the e2482 locus. In this latter screen, L4-stage rol-4(sc8) unc-61(e228)/sma-I (e30) unc-76(e911) Although no clear role for SL1 has been demonstrated, hermaphrodites were mutagenized with 50 mM EMS, and F, our results show that it is essential for viability. What progeny were scored for the absence of either the RolUnc or function does the SL1 leader contribute to trans-spliced SmaUnc phenotypes. Absence of either marker indicated the mRNAs? Trans-splicing of SL1 onto messages may pro- presence of a zygotic lethal mutation. A total of -1 1,000 hap- vide cis-acting sequences that actively control the sta- loid genomes was screened. This screen identified 209 zygotic bility, localization, or translation of messages. Our find- lethal mutations on the right arm of chromosome V. Comple- mentation tests against e2482 were performed, and one of the ings that SL2, which is quite divergent from the other 209 mutations (wl)was found to be allelic to e2482. nematode trans-spliced leaders, can substitute for SL1 e2482 and wl were mapped relative to the Tcl polymorphism during embryogenesis, suggests that there may not be a stP108 (Williams et al. 1992),which is specific for the Bergerac rigid structural requirement for the primary sequence of (Bo) strain. Heterozygotes containing a Bristol (N2) chromo- the leader per se. Recent evidence suggests that the SL1 some in trans to a Bergerac (Bo)chromosome were constructed,

GENES & DEVELOPMENT 1553 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Ferguson et al. and recombinants between unc-76 and the lethal mutation and Microin jection techniques dpy-21 and the lethal mutation were mapped by PCR analysis as Microinjection of DNAs was performed as described elsewhere described elsewhere (Williams et al. 1992).All 269 such recom- (Mello et al. 1991). Constructs used in the transformation ex- binants carried the Bo stPIO8 pattern, indicating tight linkage of periments are described in Figure 1, with the exception of the the lethal mutations to that polymorphism. Additional restric- SL2 construct, which is a subclone of a 3.8-kb BamHI fragment tion site polymorphisms were identified by Southern blot anal- containing one copy of the gene encoding SL2a RNA. One to 10 ysis of N2 and Bo genomic DNA digested with various restric- pglml of the indicated construct was coinjected with 100 pg/ tion enzymes. Several cosmids in the region were used as ml of rol-6(su1006) DNA (a dominant marker used to select probes, and three cosmids were found to detect three Hind111 transformants) into unc-76(e911) e2482/unc-61 (e228) adult her- polymorphisms (wP1, wP2, and wP3). A portion of the recom- maphrodites. F, progeny that showed a Rol phenotype were binants described above was analyzed for each polymorphism. cloned onto individual plates. F, broods that contained Rol an- Of 38 unc-76 recombinants, 37 carried the Bo wP1 pattern; of 19 imals, indicating stable transmission of the injected DNA, were dpy-21 recombinants, 17 carried the Bo wP3 pattern; and of 21 used to establish transgenic lines. To test each of the indicated dpy-21 recombinants, 17 carried the Bo wP2 pattern. These re- constructs for rescue of the wl allele, extrachromosomal arrays sults placed the e2482 lesion between wP1 and wP3. wl mapped from two of the rescued e2482 lines were crossed into wl het- to an interval consistent with this location (wl recombinants erozygotes. For each line, the number of arrested embryos and were not mapped with respect to wP1). the total number of progeny were scored from at least 10 adult animals in the F, or later generations. For each experiment in Plasmid constructions which rescue of embryonic lethality was observed, animals ex- hibiting an Unc-76 phenotype were also observed [as is expected The 985-bp BamHI fragment from K07H12 was subcloned into of rescued animals because the unc-76(e911) mutation is the Stratagene Bluescript I1 SK vector. One subclone was chosen closely linked to e2482 in the strains assayed for rescue]. This for sequence analysis and transformation experiments. Of the provided additional evidence that homozygous mutant embryos 21 bp that differ between this subclone and the published repeat were rescued. For each construct tested, the average percentage sequence, 20 of these differences are clustered in the spacer of arrested embryos was calculated by determination of the per- region between the 3' ends of the 5s rRNA and SL1 RNA coding centage of arrested embryos produced by each line, and averag- regions. The other difference is located upstream of the 5s ing these percentages. For all cases in which rescue was ob- rRNA and SLl RNA coding regions. The 985-bp subclone was served, these percentages were significantly different from that digested with BsmI and XmnI (a site within the vector sequence) observed for the parental strain, and in most cases, the percent- to yield a fragment containing 435 bp of the 985-bp insert (as ages were highly statistically significant (X0.002). shown) plus an additional 1945 bp from the vector, which was gel purified. The 740-bp fragment was purified from a ThaI di- gest of the 985-bp subclone insert. The Sm mutations were con- Isolation of embryonic RNA and RT-PCR analysis structed by digestion of the 985-bp subclone with BseRI, which Total RNA isolation and RT-PCR was performed essentially as cuts in the middle of the Sm site. The 3' protruding ends were described elsewhere (Conrad et al. 1991; Spieth et al. 1993),with converted to blunt ends by digestion with T4 polymerase and the following modifications or additions: For preparation of em- ligated to create SmA3. Alternatively, the ends were ligated to bryonic RNA, -10-30 elongating wild-type embryos or -10- an 8-bp MluI linker, and the ligation mixture was digested with 30 homozygous mutant embryos -8-10 hr postfertilization MluI and religated to form the Sm-Mlu mutant. The sequence were hand-picked into 50 p1 of sterile water. Mutant embryos alterations of the mutations were confirmed by sequence anal- were identified under the dissecting microscope as those em- ysis. bryos that had not elongated 8-10 hr postfertilization; this method has potential limitations, as it is difficult to confirm that all embryos picked are homozygous rrs-1 deletion mutants. Isolation of genomic DNAs and Southern analysis Embryonic extracts and total RNA were prepared, and RNA was Genomic DNA was prepared from a mixed population of wild- resuspended in 1 x DNase buffer. One unit of RQ1 DNase type animals as described previously (Pilgrim 1993). Arrested (Promega) was added and samples incubated at 37°C for -10 homozygous wl and e2482 embryos were isolated based on pre- min. After phenol-chloroform extraction, RNA was precipi- viously described procedures (Shamu 1989). Briefly, animals tated with two volumes of ethanol. The DNase step and second heterozygous for each mutation were grown in liquid cultures. phenol-chloroform extraction were found to be important in Worms were allowed to settle, and the pellet was resuspended some cases to prevent spurious signals that were detected in the in M9 buffer. The worms were treated with 1.6 N NaOH, 6% - RT controls (see below) when these steps were omitted. One- NaOC1, to kill the adults but not the embryos. The embryos eighth of the RNA was used directly in a control PCR reaction were pelleted by centrifugation and washed three times in M9 ( - RT), performed as described below, and one-half of the sam- buffer. After 24 hr, the embryos were pelleted by centrifugation, ple used in an annealing reaction, followed by reverse transcrip- resuspended in M9 buffer, and treated with 0.25 M KOH, 1.2% tion ( + RT). Additional controls ( - emb) were performed by NaOC1, to dissolve hatched larval worms and adult carcasses, preparation of extracts as above containing Escherichia coli leaving arrested embryos. Embryos were pelleted by centrifuga- alone (which is normally transferred along with the embryos), tion and washed in M9 buffer. Genomic DNA was prepared, and addition of one-half of the sample to an annealing reaction, and -3 pg of each DNA was digested and loaded onto a 1% agarose performance of RT-PCR as described below. Annealings were gel (for analysis of the rrs-1 repeat) or a 0.6% agarose gel (for performed at 50°C for 75 min with 200 ng of the appropriate analysis of the high molecular weight fragments). Gels were downstream primers (described below). One-half of each anneal- blotted and filters probed with the rrs-1-rescuing subclone that ing reaction was added to a reverse transcription reaction mix was labeled with [CY-~~P]~ATPby random priming with the consisting of 33 mM Tris-HC1, at pH 8.3, 40 mM NaC1, 6.7 mM Prime-a-Gene kit from Promega. Quantitation was performed DTT, 20 mM magnesium acetate, 6.7 mM of each dNTP, 20 on a Molecular Dynamics PhosphorImager model 425E. units of RNAsin ribonuclease inhibitor (Promega),and 8 units

1554 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

C. elegans SL1 RNA is essential for embryogenesis of AMV reverse transcriptase (Promega).Reactions were incu- References bated at 42°C for 30 min. Reactions were diluted to 50 p1 with water, and 5 p1 was used in a PCR reaction. PCR was performed Agabian, N. 1990. Trans-splicing of nuclear pre-mRNAs. Cell with 50 ng of the appropriate upstream and downstream primers 61: 1157-1160. (described below) for 35 cycles under the following conditions: Bektesh, S., K. Van Doren, and D. Hirsh. 1988. Presence of the 94°C for 30 sec, 50°C for SL1 reactions and 45°C for SL2 reac- Caenorhabditis elegans spliced leader on different mRNAs tions for 1 min, and 72°C for 1 min. Products were separated on and in different genera of nematodes. Genes & Dev. 2: 1277- 1.3% agarose gels, gels blotted, and filters probed with 1283. [a-32P]dATP-labeledprobes corresponding to an internal portion Blumenthal, T. 1995. Trans-splicing and polycistronic tran- of the messages tested, which were generated by PCR. To gen- scription in Caenorha bditis elegans. Trends Genet. 11: 132- erate the probes, %o of an initial PCR reaction, performed ac- 136. cording to the conditions described below and either -2 pg of Brenner, S. 1974. The genetics of Caenorhabditis elegans. Ge- mixed stage genomic DNA or 5 ng of cloned hlh-1 cDNA (a gift netics 77: 71-94. from Michael Krause) as the original template, was used in a Bruzik, J.P. and T. Maniatis. 1992. Spliced leader RNAs from subsequent PCR reaction in a standard buffer provided by the lower eukaryotes are trans-spliced in mammalian cells. Na- manufacturer (Perkin Elmer) with 50 ng of each primer (de- ture 360: 692-695. scribed below), 25 pCi of [a-32P]dATP,and 2.5 units of Taq Bruzik, J.P., K. Van Doren, D. Hirsh, and J.A. Steitz. 1988. polymerase (Perkin Elmer). PCR was performed for 20 cycles Trans-splicing involves a novel form of small nuclear ribo- under the following conditions: 94°C for 30 sec, 48°C for 1 min, nucleoprotein particles. Nature 335: 559-562. and 72°C for 1 min. After hybridization, blots were washed Conrad, R., J. Thomas, J. Spieth, and T. Blumenthal. 1991. In- twice with 2 x SSPE, 0.1 % SDS, at room temperature, and twice sertion of part of an intron into the 5' untranslated region of with 1x SSPE, 0.1 % SDS, at 65°C. The following oligonucle- a Caenorhabditis elegans gene converts it into a trans- otides were used as downstream primers in the reverse tran- spliced gene. Mol. Cell Biol. 11: 1921-1926. scription and PCR (numbers correspond to complementary po- Conrad, R., R.F. Liou, and T. Blumenthal. 1993. Conversion of sitions of the GenBank sequences): ges-l 4900-4876 (exon 3), a trans-spliced C. elegans gene into a conventional gene by (Kennedy et al. 1993); hlh-1 53 13-5292 (exon 3), (Krause et al. introduction of a splice donor site. EMBO J. 12: 1249-1255. 1990);myo-3 4230-421 1 (exon 2), (Dibb et al. 1989);elt-1 1052- Conrad, R., K. Lea, and T. Blumenthal. 1995. SL1 trans-splicing 1034 (exon 2) (Spieth et al. 1991).Oligonucleotides correspond- specified by AU-rich synthetic RNA inserted at the 5' end of ing to the first 20 nucleotides of either the SL1 or SL2 leader Caenorhabditis elegans pre-mRNA. RNA 1: 164-1 70. were used as upstream primers (Spieth et al. 1993).The follow- Dibb, N.J., I.N. Maruyama, M. Krause, and J. Karn. 1989. Se- ing oligonucleotides were used to generate the probes for South- quence analysis of the complete Caenorhabditis elegans ern analysis of the RT-PCR products [numbers indicate posi- myosin heavy chain gene family. J. Mol. Biol. 205: 603-613. tions in the GenBank sequence (upstream primers) or comple- Edgar, L.G. and J.D. McGhee. 1986. Embryonic expression of a mentary positions (downstream primers)]: ges-1 upstream gut-specific esterase in Caenorhabditis elegans. Dev. Biol. 3314-3329, downstream 3366-335 1 (exon 1); hlh-1 upstream 114: 109-118. 503 1-5056, downstream 5 180-5 164 (exon 2); myo-3 upstream Ellis, H.M. and H.R. Horvitz. 1986. Genetic control of pro- 2971-2987, downstream 3036-30 16 (exon 1); elt-1 upstream grammed cell death in the nematode Caenorhabditis ele- 740-757, downstream 857-841 (exon 1). gans. Cell 44: 8 17-829. Guthrie, C. and B. Patterson. 1988. Spliceosomal snRNAs. Annu. Rev. Genet. 22: 387-419. Goh, P.-Y. and T. Bogaert. 1991. Positioning and maintenance of Acknowledgments embryonic body wall muscle attachments in C. elegans requires the mup-1 gene. Development 11 1: 667-68 1. J.H.R. is grateful to Dr. John White for providing space in his Hannon, G.J., P.A. Maroney, and T.W. Nilsen. 1991. U small laboratory at the Medical Research Council (MRC),Cambridge, nuclear ribonucleoprotein requirements for nematode cis- UK, during initial phases of this work, for encouragement, and and trans-splicing in vitro. J. Biol. Chem. 266: 22792-22795. for stimulating discussions. We thank Phil Anderson, Dave Hannon, G.J., P.A. Maroney, Y.-T. Yu, G.E. Hannon, and T.W. Brow, Judith Kimble, Michael Terns, and Marv Wickens, as well Nilsen. 1992. Interaction of U6 snRNA with a sequence re- as members of the Rothman laboratory, for comments on a quired for function of the nematode SL RNA in trans-splic- previous version of the manuscript. We thank Tom Blumenthal ing. Science 258: 1775-1 780. and Kevin Van Doren for the SL2 construct, the Blumenthal Huang, X.Y. and D. Hirsh. 1989. A second trans-spliced RNA laboratory for primers, protocols, and helpful discussions, and leader sequence in the nematode Caenorhabditis elegans. Jim McGhee and Michael Krause for protocols and primers. Proc. Natl. Acad. Sci. 86: 8640-8644. Some of the strains used in this work were provided by the Jones, M.H. and C. Guthrie. 1990. Unexpected flexibility in an Caenorhabditis Genetics Center, which is funded by the Na- evolutionarily conserved protein-RNA interaction: Genetic tional Center for Research Resources of the National Institutes analysis of the Sm binding site. EMBO J. 9: 2555-2561. of Health. K.C.F. and P.J.H. were supported by a Cell and Mo- Kennedy, B.P., E.J. Aamodt, F.L. Allen, M.A. Chung, M.F.P. He- lecular Biology Predoctoral Training Grant from the National schl, and J.D. McGhee. 1993. The gut esterase (ges-1)from Institutes of Health (NIH).This work was supported by a grant the nematodes Caenorhabditis elegans and Caenorhabdtis from the NIH (GM 48137), a Searle Scholars Award from the briggsae. J. Mol. Biol. 229: 890-908. Chicago Community Trust, and a Shaw Scientists Award from Krause, M. and D. Hirsh. 1987. A trans-spliced leader sequence the Milwaukee Foundation to J.H.R. on actin mRNA in C. elegans. Cell 49: 753-761. The publication costs of this article were defrayed in part by Krause, M., A. Fire, S. Harrison-White, J. Priess, and H. Wein- payment of page charges. This article must therefore be hereby traub. 1990. CeMyoD accumulation defines the body wall marked "advertisement" in accordance with 18 USC section muscle cell fate during C. elegans embryogenesis. Cell 1734 solely to indicate this fact. 63: 907-919.

GENES & DEVELOPMENT 1555 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Ferguson et al.

Maroney, P.A., G.J. Hannon, J.A. Denker, and T.W. Nilsen. Shamu, C.E. 1989. "Molecular characterization of the mab-3 1990. The nematode spliced leader RNA participates in region of the C. elegans genome." M. Phil. thesis. Trinity trans-splicing as an Sm snRNP. EMBO 1. 9: 3667-3673. College, Cambridge University, UK. Maroney, P.A., J.A. Denker, E. Darzynkiewicz, R. Laneve, and Spieth, J., Y.H. Shim, K. Lea, R. Conrad, and T. Blumenthal. T.W. Nilsen. 1995. Most mRNAs in the nematode Ascaris 1991. elt-1, an embryonically expressed Caenorhabditis el- lumbricoides are trans-spliced: A role for spliced leader ad- egans gene homologous to the GATA transcription factor dition in translational efficiency. RNA 1: 714-723. family. Mol. Cell Biol. 11: 4651-4659. Mattaj, I.W. 1988. U snRNP assembly and transport. In Struc- Spieth, J., G. Brooke, S. Kuersten, K. Lea, and T. Blumenthal. ture and function of major and minor small nuclear ribo- 1993. Operons in C. elegans: Polycistronic mRNA precur- nucleoprotein particles (ed. M.L. Birnstiel), pp. 100-1 14. sors are processed by trans-splicing of SL2 to downstream Springer-Verlag, New York, NY. coding regions. Cell 73: 521-532. Mello, C.C., J.M. Kramer, D. Stinchcomb, and V. Ambros. 1991. Strome, S. and W.B. Wood. 1982. Immunofluorescence visual- Efficient gene transfer in C. elegans: Extrachromosomal ization of germ-line specific cytoplasmic granules in em- maintenance and integration of transforming sequences. bryos, larvae, and adults of Caenorhabditis elegans. Proc. EMBO J. 10: 3959-3970. Natl. Acad. Sci. 79: 1558-1562. Michaeli, S., T.G. Roberts, K.P. Watkins, and N. Agabian. 1990. Sulston, J. and J. Hodgkin. 1988. Methods. In The nematode Isolation of distinct small ribonucleoprotein particles con- Caenorhabditis elegans (ed. W.B. Wood), pp. 587-606. Cold taining the spliced leader and U2 RNAs of Trypanosoma Spring Harbor Laboratory, Cold Spring Harbor, NY. brucei. 1. Biol. Chem. 265: 10582-10588. Sulston, j., E. Schierenberg, J. White, and J. Thomson. 1983. The Miller, D.M., I. Ortiz, G. Berliner, and H.F. Epstein. 1983. Dif- embryonic cell lineage of the nematode C. elegans. Dev. ferential location of two myosins within nematode thick Biol. 100: 64-1 19. filaments. Cell 34: 477-490. Thomas, J.D., R.C. Conrad, and T. Blumenthal. 1988. The C. Nelson, D.W. and B.M. Honda. 1985. Genes coding for 5S ribo- elegans trans-spliced leader RNA is bound to Sm and has a somal RNA of the nematode Caenorhabditis elegans. Gene trimethylguanosine cap. Cell 54: 533-539. 38: 245-25 1. Tschudi, C. and E. Ullu. 1990. Destruction of U2, U4, or U6 Nilsen, T.W. 1993. Trans-splicing of nematode premessenger small nuclear RNAs blocks trans-splicing in trypanosome RNA. Annu. Rev. Microbiol. 47: 413-440. cells. Cell 61: 459-466. Nilsen, T.W., J. Shambaugh, J.A. Denker, G. Chubb, C. Faser, L. Van Doren, K. and D. Hirsh. 1988. Trans-spliced leader RNA Putnam, and K. Bennett. 1989. Characterization and expres- exists as small nuclear ribonucleoprotein particles in Cae- sion of a spliced leader RNA in the parasitic nematode As- norhabditis elegans. Nature 335: 556-559. caris lumbricoides var. suum. Mol. Cell Biol. 9: 3543-3547. Walder, J.A., P.S. Eder, D.M. Engman, S.T. Brentano, R.Y. Okamoto, H. and J.N. Thomson. 1985. Monoclonal antibodies Walder, D.S. Knutzon, D.M. Dorfman, and J.E. Donelson. which distinguish certain classes of neuronal and supporting 1986. The 35-nucleotide spliced leader sequence is common cells in the nervous tissue of the nematode Caenorhabditis to all trypanosome messenger RNAs. Science 233: 569-571. elegans. 1. Neurosci. 5: 643-653. Wassarman, K.M. and J.A. Steitz. 1993. Association with termi- Okkema, P.G., S. Harrison-White, V. Plunger, A. Aryana, and A. nal exons in pre-mRNAs: A new role for the U1 snRNP? Fire. 1993. Sequence requirements for myosin gene expres- Genes & Dev. 7: 647-659. sion and regulation in Caenorha bditis elegans. Genetics Waterston, R.W. 1989. The minor myosin heavy chain, mhcA, 135: 385-404. of Caenorhabditis elegans is necessary for the initiation of Palfi, Z., A. Giinzl, M. Cross, and A. Bindereif. 1991. Affinity thick filament assembly. EMBO 1. 8: 3429-3436. purification of Trypanosoma brucei snRNPs reveals com- Watkins, K.P. and N. Agabian. 1991. In vivo UV cross-linking of mon and specific protein components. Proc. Natl. Acad. Sci. U snRNAs that participate in trypanosome trans-splicing. 88: 9097-9101. Genes & Dev. 5: 1859-1869. Pilgrim, D. 1993. The genetic and RFLP characterization of the Williams, B.D., B. Schrank, C. Huynh, R. Shownkeen, and R.H. left end of linkage group I11 in Caenorhabditis elegans. Ge- Waterston. 1992. A genetic mapping system in Caenorhab- nome 36: 712-724. ditis elegans based on polymorphic sequence-tagged sites. Priess, J.R. and J.N. Thomson. 1987. Cellular interactions in Genetics 131: 609-624. early C. elegans embryos. Cell 48: 241-250. Yuan, J., S. Shaham, S. Ledoux, H.M. Ellis, and H.R. Horvitz. Ray, C., A.G. Abbott, and R.S. Hussey. 1994. Trans-splicing of a 1993. The C. elegans cell death gene ced-3 encodes a protein Meloidogyne incognita mRNA encoding a putative esopha- similar to mammalian interleukin- 1@-converting enzyme. geal gland protein. Mol. Biochem. Parasitol. 68: 93-101. Cell 75: 641-652. Ross, L.H., J.H. Freedman, and C.S. Rubin. 1995. Structure and Zorio, D.A.R., N.N. Cheng, T. Blumenthal, and J. Spieth. 1994. expression of novel spliced leader RNA genes in Caenorhab- Operons as a common form of chromosomal organization in ditis elegans. 1. Biol. Chem. 270: 22066-22075. C. elegans. Nature 372: 270-272. Roy, J., B. Zheng, B.C. Rymond, and J.L. Woolford. 1995. Struc- turally related but functionally distinct yeast Sm D core small nuclear ribonucleoprotein particle proteins. Mol. Cell Biol. 15: 445-455. Rymond, B.C. 1993. Convergent transcripts of the yeast PRP38- SMDl locus encode two essential splicing factors, including the D 1 core polypeptide of small ribonucleoprotein particles. Proc. Natl. Acad. Sci. 90: 848-852. Schnabel, R. 1995. Duels without obvious sense: Counteracting inductions involved in body wall muscle development in the Caenorhabditis elegans embryo. Development 121: 2219- 2232.

1556 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

The SL1 trans-spliced leader RNA performs an essential embryonic function in Caenorhabditis elegans that can also be supplied by SL2 RNA.

K C Ferguson, P J Heid and J H Rothman

Genes Dev. 1996, 10: Access the most recent version at doi:10.1101/gad.10.12.1543

References This article cites 52 articles, 23 of which can be accessed free at: http://genesdev.cshlp.org/content/10/12/1543.full.html#ref-list-1

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here.