HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 279, Issue 44, 45573-45585, October 29, 2004

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 279/44/45573    most recent M407475200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lall, S. Articles by Davis, R. E. Search for Related Content PubMed PubMed Citation Articles by Lall, S. Articles by Davis, R. E. Related Collections Papers Of The Week Social Bookmarking                      What's this?

Contribution of Trans-splicing, 5' -Leader Length, Cap-Poly(A) Synergism, and Initiation Factors to Nematode Translation in an Ascaris suum Embryo Cell-free System^*

Sabbi Lall, Cassandra C. Friedman, Marzena Jankowska-Anyszka¶, Janusz Stepinski||, Edward Darzynkiewicz||, and Richard E. Davis**

From the Department of Biology, City University of New York Graduate Center, College of Staten Island, Staten Island, New York 10314 and the ^¶Department of Chemistry and ^||Department of Biophysics, University of Warsaw, 02-093 Warsaw, Poland

Received for publication, July 6, 2004 , and in revised form, August 18, 2004.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Trans-splicing introduces a common 5' 22-nucleotide sequence with an N-2,2,7-trimethylguanosine cap (m ^2,2,7₃GpppG or TMG-cap) to more than 70% of transcripts in the nematodes Caenorhabditis elegans and Ascaris suum. Using an Ascaris embryo cell-free translation system, we found that the TMG-cap and spliced leader sequence synergistically collaborate to promote efficient translation, whereas addition of either a TMG-cap or spliced leader sequence alone decreased reporter activity. We cloned an A. suum embryo eIF4E homolog and demonstrate that this recombinant protein can bind m⁷G- and TMG-capped mRNAs in cross-linking assays and that binding is enhanced by eIF4G. Both the cap structure and the spliced leader (SL) sequence affect levels of A. suum eIF4E cross-linking to mRNA. Furthermore, the differential binding of eIF4E to a TMG-cap and to trans-spliced and non-trans-spliced RNAs is commensurate with the translational activity of reporter RNAs observed in the cell-free extract. Together, these binding data and translation assays with competitor cap analogs suggest that A. suum eIF4E-3 activity may be sufficient to mediate translation of both trans-spliced and non-trans-spliced mRNAs. Bioinformatic analyses demonstrate the SL sequence tends to trans-splice close to the start codon in a diversity of nematodes. This evolutionary conservation is functionally reflected in the optimal SL to AUG distance for reporter mRNA translation in the cell-free system. Therefore, trans-splicing of the SL1 leader sequence may serve at least two functions in nematodes, generation of an optimal 5'-untranslated region length and a specific sequence context (SL1) for optimal translation of trimethylguanosine capped transcripts.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Spliced leader (SL)¹ trans-splicing is present in diverse metazoan phyla including the chordates, cnidarians, platyhelminths, and nematodes (reviewed in Ref. 1). Spliced leader addition forms the mature 5' end of the mRNA providing an unusual cap structure and common sequence to the mRNA. In the case of Caenorhabditis elegans, 70% of transcripts have one of two 22-base sequences, spliced leader 1 or spliced leader 2 (Ref. 2 and reviewed in Ref. 3). Because the spliced leader donor RNA exon has a 2,2,7-trimethylguanosine cap, rather than the usual monomethylguanosine cap, the product of the trans-splicing reaction is a TMG-capped mRNA (4–7). Transsplicing may therefore affect translation initiation.

The major function of SL2 trans-splicing in C. elegans is to contribute to the resolution of polycistronic transcripts to generate mature mRNAs (8–10). Trans-splicing also plays this role in some other organisms, including the kinetoplastid protozoans, and perhaps platyhelminths (11, 12). However, it is not clear that resolution of polycistronic pre-mRNAs is the sole role for spliced leader trans-splicing, particularly in platyhelminths (see 13).² The predominant form of trans-splicing in C. elegans and other nematodes is SL1 trans-splicing (57% of C. elegans and at least 70% of Ascaris suum transcripts) (2, 14). However, the major role(s) of SL1 trans-splicing has yet to be defined in nematodes, and the role of trans-splicing in other organisms such as Ciona intestinalis and Hydra vulgaris also remains unknown (15, 16).

Since trans-splicing has a profound effect on the 5'-UTR of mRNAs, it has long been postulated that SL1 and TMG-cap addition might influence translational efficiency of mRNAs (see Ref. 17). Indeed a survey of SL-containing transcripts showed that trans-splicing tends to occur close to the start codon of mRNAs, providing circumstantial evidence for a relationship between trans-splicing and translation (3). While the nematodes are a major focus of both genomic and developmental biology research, in vitro translation studies have not been conducted in C. elegans because of the lack of a competent cell-free system. Embryos of the parasitic nematode Ascaris are highly suitable for preparation of several types of biochemically competent extracts that have been used to examine multiple molecular processes in nematodes (18–22). Ascaris translation extracts were used to show that 80–90% of transcripts are trans-spliced (14). This same study found that the TMG-cap and SL1 sequence collaborate to increase translational efficiency in a micrococcal nuclease treated Ascaris extract.

In recent years, development and characterization of in vitro eukaryotic translation systems have used cap dependence and synergy between the mRNA cap and poly(A)-tail as an important characteristic for validation of cell-free systems (Ref. 23 and reviewed in Ref. 24). Cap dependence and cap/poly(A) synergy are in vivo properties of translation, and some in vitro systems do not recapitulate these characteristics. Indeed only a few systems exhibiting cap/poly(A)-tail synergy exist, none of them allowing translation studies in a nematode cell-free system (25–29). The effect of trans-splicing on protein synthesis in nematodes in a system known to reflect in vivo conditions has not been carried out. In fact, little has been done to functionally characterize many attributes of nematode translation, including 5'-UTR length and start codon choice, factors known to be important in mammalian and yeast translation systems (30–33).

Translation initiation in yeast and mammals involves the binding of a multiprotein complex, eIF4F to the 5' end of the mRNA (reviewed in Refs. 34 and 35). eIF4F consists of eIF4E, eIF4G, and eIF4A. Whereas eIF4E recognizes the 5' mRNA cap structure, eIF4G acts as a coordinating scaffold, connecting eIF4E, eIF4A, and the poly(A)-binding protein (36–38). The interaction of eIF4G with eIF4E has been shown to enhance binding of the latter to the mRNA cap (39, 40). eIF4G is involved in recruiting the S 40 preinitiation complex to the mRNA, and the interaction of eIF4G with the poly(A)-binding protein may be the basis of the synergistic effect of a cap and poly(A)-tail upon translational efficiency (reviewed in Ref. 34). Of these events, only the binding of eIF4E proteins to methylated nucleoside triphosphates (as a proxy for the mRNA cap structure) has been examined in nematodes. Whereas many organisms have one to two homologs of eIF4E, C. elegans has five (named ife-1 through 5, Ref. 41) that differ in their binding specificity for mono- and trimethyl-cap analogs. Of these proteins, C. elegans IFE-3 and IFE-4 most closely resemble the two human eIF4E proteins in primary structure, and both bind to an m⁷GTP-Sepharose column. IFE-1, -2, and -5 are capable of binding both m⁷GTP- and m ^2,2,7₃GTP-Sepharose matrices (41). Such binding experiments, along with structural modeling, were used to predict and define residues that might enable IFE-1, -2, and -5 binding to m ^2,2,7₃GTP-Sepharose (42). These binding studies and in vivo functional experiments (43) were used to support the hypothesis that C. elegans IFE-3 and -4 might be responsible for m⁷GpppG-based translation (with IFE-4 playing a non-essential role). IFE-1, -2, and -5 were predicted to play a role in m ^2,2,7₃GpppG-cap based translation. However, conservation of these TMG-binding proteins across nematodes, differential binding of IFEs to a capped RNA (i.e. the in vivo substrate) rather than a cap analog, and their contribution to the translation of trans-spliced transcripts has not been examined.

In the current study, we describe and use an Ascaris cell-free system to examine the effect of trans-splicing on translation. We show that the Ascaris in vitro system is mRNA cap-dependent and retains translational synergy between the cap structure (both m⁷G and TMG) and the poly(A)-tail, an effect that we have observed in vivo in separate studies.³ We find that in our Ascaris extract, non-trans-spliced transcripts are most efficiently translated. Whereas either the TMG-cap or SL sequence alone have a detrimental effect on protein synthesis, together they promote translation of a reporter mRNA. We then show that Ascaris eIF4E-3 is capable of binding both m⁷G- and TMG-capped RNA. The cross-linking efficiency of Ascaris eIF4E-3 to mRNA may explain the translational activity of trans-spliced versus non-trans-spliced transcripts in the in vitro system. We have also used a bioinformatic analysis to show that the spliced leader (SL1) tends to splice close to the translation start codon across nematode species. This correlates in part with the most efficient SL to start codon distance for translation in our cell-free system and may therefore reflect evolutionary selection for optimal expression. The Ascaris system should be valuable for further characterizing and understanding nematode translation and the role of trans-splicing in post-transcriptional gene expression. The parasite cell-free extract may also prove useful for drug testing and identification of targets for rational drug design.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
A. suum Extract Preparation—Embryos were collected from the proximal third (4 cm) of A. suum uteri (Carolina Biological Supply). After rinsing in cold 1x PBS (140 mM NaCl, 2.7 mM KCl, 8 mM Na₂HPO₄, 2 mM KHPO₄ (pH 7.5)), the uterine tissue was dissolved in 0.5 N NaOH. Embryos were washed extensively in cold H₂O, twice in cold 1x PBS (pH 2.0), and stored in 5 volumes of PBS (pH 2.0) at 4 °C until required. Development was initiated by agitating embryos at 110 rpm, 30 °C in 1x PBS (pH 2.0). At day 5 (32–64 cell stage, determined by 4',6-diamidino-2-phenylindole staining), embryos were spun for 5 min at 750 x g, and the eggshells were removed by a 90-min 0.4 N KOH, 1.4% sodium hypochlorite treatment at 30 °C. After four washes in 10–20 volumes of cold 1x PBS (pH 7.4), the embryo pellet was resuspended in 0.4 volume of buffer A (10 mM KCl, 1.5 mM MgCl₂, 20 mM Tris (pH 7.9), 1 mM DTT, 1x Roche Applied Science EDTA-free complete protease inhibitors, 0.1 mM PMSF, 2.5 µg/ml leupeptin, and 2 µg/ml pepstatin). The embryo suspension was homogenized within 5 min with 8–10 passes in a 15 ml Wheaton Dura-Grind® stainless steel dounce (Millville, NJ) and the homogenate centrifuged at 27,000 x g, 4 °C, 20 min. The intermediate layer (soluble non-lipid layer) was dialyzed twice against 1000–2000 volumes of ice-cold buffer (50 mM KCl, 20 mM Tris-HCl (pH 7.9), 0.2 mM EDTA, 1 mM DTT, 0.5 mM PMSF, 20% glycerol). The extract was then centrifuged (21,000 x g, 4 °C, 10 min), frozen in liquid nitrogen, and stored at–80 °C until use.

In Vitro Transcription—PCR-generated templates for in vitro transcription were synthesized using the primers in supplemental Table I (which introduce a T7 promoter), from either pRL-null for Renilla luciferase or pGL3-Basic for firefly luciferase (Promega). The in vitro transcription reactions (T7 Megascript, Ambion) typically contained 1 µg of template, 10 mM ATP, 10 mM CTP, 10 mM UTP, 12 mM cap analogue (ApppG, m⁷GpppG, or m ^2,2,7₃GpppG, Ref. 44), 1.5 mM GTP, 1x Reaction buffer, 1x T7 polymerase mix and was carried out at 37 °C for 4 h. Uncapped transcription reactions lacked cap analog. After DNase I treatment, the mRNA transcript was purified, qualified by formaldehyde denaturing gel electrophoresis, and quantified using UV spectroscopy and the Ribogreen reagent (Molecular Probes). Analysis of cap orientation on transcripts was carried out as described (45) and indicated that 90% of the m⁷GpppG and 80% of the m ^2,2,7₃GpppG caps were added in the correct orientation. These analyses demonstrated that differences in the translation efficiencies observed were not a function of differences in cap orientation on the transcripts.

Uniformly labeled transcripts for UV cross-linking were synthesized with the Promega T7 Riboprobe kit according to the manufacturer's instructions using 0.5 µg of template in a 20-µl reaction containing 50 µCi of [-³²P]GTP (3000 Ci/mmol, PerkinElmer Life Sciences), a 0.5 mM concentration each of CTP, UTP, and ATP, 0.2 mM GTP, and 1.6 mM cap analog. Uncapped transcripts were similarly generated without cap analog. The transcripts were gel-purified from 5% denaturing PAGE gels. Cap-labeled RNAs were prepared from uncapped RNA substrates using [-³²P]GTP (PerkinElmer Life Sciences) and recombinant vaccinia RNA guanylyltransferase and (guanine-N⁷)-methyltransferase (generously provided by Stewart Shuman) or human guanylyltransferase (generously provided by Aaron Shatkin) (46). RNA sequences used for translation and cross-linking are provided in supplemental Table II.

In Vitro Translation and Reporter Quantification—Ten-microliter translation reactions contained 5–12.5 µg/ml reporter RNA, 50% Ascaris embryo extract (125–150 µg of extract protein), 10 mM Tris-HCl (pH 7.9), 125 mM KOAc, 25 mM KCl, 2.6 mM Mg(OAc)₂, 0.5 mM GTP, 0.5 mM ATP, 10 mM creatine phosphate, 50 µg/ml creatine kinase, 2.5 mM DTT, 0.1 mM EDTA, 50 µM Promega amino acid mix, 0.25 mM PMSF, and 10% glycerol. After 30–60 min at 30 °C, the translation reaction was diluted in the appropriate lysis buffer (Promega) and frozen on dry ice. Typically, 0.1% of the reaction was read in a Sirius luminometer v.2.2 (Berthold Detection Systems). Dual assay experiments were performed by mixing 62.5 ng of test mRNA (Renilla luciferase) with 62.5 ng m⁷G-firefly luciferase. After 1 h, translation was assayed using the Promega dual luciferase assay kit. Statistical analysis was carried out using Prism 4.0.

Cloning and Sequence Analysis of eIF4E Genes—A. suum eIF4E-3 (GenBank^TM Accession number AY599493 [GenBank] ) was cloned from 32–64 cell cDNA (synthesized as described below) (12, 47–50) by 5' RACE using a degenerate primer (supplemental Table I) based on conserved residues identified in nematode eIF4E sequences. RACE products were directionally cloned into the pAMP vector (Invitrogen), as described by the manufacturer. 3' RACE was then used to obtain the entire open reading frame of A. suum eIF4E-3 (primers in supplemental Table I). C. elegans ife-1 and ife-3 coding regions were obtained from mixed stage oligo(dT) or random primed C. elegans cDNA using the primers in supplemental Table I. Phylogenetic analysis was performed using neighbor-joining and Bayesian methods (see overviews in Refs. 51–53).

Recombinant Protein Production and Nucleotide Binding Studies— The eIF4E open reading frames were cloned into the NdeI/BamHI sites of pET16b and transformed into Rosetta(DE3) bacteria (Novagen). Protein production was induced overnight (25 °C) using 0.4 mM isopropyl -D-thiogalactopyranoside. The bacterial pellet was frozen at–80 °C, then sonicated in lysis buffer (20 mM HEPES-KOH (pH 7.5), 100 mM KCl, 2 mM EDTA, 1 mM DTT, 10% glycerol, 0.4x Roche Applied Science EDTA-free protease inhibitors), and the fusion protein was purified on aNi²⁺-nitrilotriacetic acid-agarose column (Qiagen). The bound protein was eluted using an imidizole gradient (10–500 mM), dialyzed against buffer B (50 mM KCl, 20 mM Tris (pH 7.5), 0.2 mM EDTA, 1 mM DTT, 0.5 mM PMSF, 20% glycerol) and stored at –80 °C.

For cap analog binding studies, recombinant protein was added to m⁷GTP-Sepharose or m ^2,2,7₃GTP-Sepharose 4B beads (41) and incubated for 45 min at 4 °C on a rotating platform. The beads were collected, the supernatant removed (unbound fraction), and the beads washed three times in buffer B (above). Protein bound to the beads and in the unbound fraction were analyzed by 12.5% SDS-PAGE to determine protein binding efficiency to the matrices. Proteins were detected by Coomassie Blue staining or SYPRO Ruby labeling (Molecular Probes).

UV cross-linking was performed by incubating 500 ng of recombinant protein with 250,000 dpm of uniformly labeled RNA (or 100,000 dpm of cap-labeled RNA) for 10 min at 30 °C in 20 µl of buffer (100 mM KOAc, 2mM Mg(OAc)₂, 20 mM HEPES-KOH (pH 7.5), 2 mM DTT, 3% glycerol). Cross-linking was then performed in a Hoefer UV Cross-linker, at a distance of 4 cm (120,000 µJ/cm²) for 30 min on ice. The cross-linked mRNA was then treated with 20 µg of RNase A (United States Biochemical, 37 °C for 30 min) and analyzed by SDS-PAGE on a 4–15% Tris-HCl gel (Bio-Rad), and the gel was exposed to Kodak BioMax film overnight at –80 °C. Cross-linked proteins were also imaged and analyzed using the STORM 860 and ImageQuant programs (Amersham Biosciences).

Search for Spliced Leaders in Ascaris Embryos—To assess the presence of SL sequences on Ascaris embryo mRNAs, an oligonucleotidecapped cDNA library was constructed from 32–64 cell stage A. suum embryo RNA (12, 47–50). In brief, total RNA was dephosphorylated with alkaline phosphatase (Roche Applied Science), decapped using tobacco acid pyrophosphatase (Epicenter Technologies), and an RNA oligonucleotide (see supplemental Table I) ligated to the 5' end of RNAs using T4 RNA ligase (New England Biolabs) under the conditions described by the manufacturers. First strand cDNA was synthesized using an adaptor oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen) at 42 °C for 30 min, followed by addition of more Superscript enzyme and a further incubation at 50 °C for 30 min. Reverse transcription-PCR products were generated using primers complementary to the RNA oligonucleotide and the 3' adaptor sequence. PCR products 500–1000 and 1000–3000 bp were gel-purified and directionally cloned using ligation-independent cloning with the pAMP system (Invitrogen). The 5' ends (only 30 nucleotides) of 1500 randomly selected clones were directly sequenced from bacterial lysates using ³²P-end-labeled primers and cycle sequencing.

Bioinformatic Analysis of SL to AUG Distance—The position of the spliced leader sequence relative to the start codon in trans-spliced mRNAs was ascertained using cDNA and EST sequences from available developmental stages in current databases. We took the following steps to ensure that only cDNA sequences that truly contain a spliced leader were assessed. Nematode cDNA sequences that contain an entire spliced leader sequence at their 5' end were identified using a BLASTN search of the NCBI nr data base. Additional sequences were found using BLASTN searches of EST data bases at NCBI and the NEM-BLAST server (nema.cap.ed.ac.uk/ncbi_blast.html). Libraries based on PCR amplification with an SL containing primer were not used (to avoid sequences where an SL had been fortuitously introduced by PCR primers). The predicted start codon was either that specified in the data base entry or the most 5' AUG of the longest open reading frame in the sequence (in the case of ESTs).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The A. suum in Vitro Translation Extract Is Cap-dependent and Shows Synergy between the Cap and Poly(A)-tail—A. suum embryonic extracts have previously been used to study translation in vitro (14). However, these experiments were carried out following micrococcal nuclease treatment of endogenous mRNAs and several characteristics of the cell-free system, including potential cap dependence and cap/poly(A)-tail synergism, were not described. To further develop and characterize a nematode in vitro system, we modified the extract preparation (see "Experimental Procedures"), the micrococcal nuclease treatment was excluded and demonstrated that the extracts were cap-dependent and exhibited cap/poly(A)-tail synergism (see below). Our preparation yielded extracts with increased protein concentration and allowed evaluation of reporter translation under conditions where exogenously added transcripts compete with endogenous mRNAs for translation complexes.

There was a linear relationship between reporter transcript concentration up to at least 400 ng (m⁷GpppG-Ren-A₈₅) and translation as measured by Renilla activity (Fig. 1A). This linear message dependence extended up to 1 µg of RNA for the firefly luciferase reporter (data not shown). We assayed 50–125 ng of transcript in subsequent experiments to evaluate translation at the lower end of the linear range of the dose-response relationship. Time courses of reporter translation showed there is a 5-min lag followed by Renilla luciferase accumulation with time, leveling off at around 30 min (Fig. 1B). The leveling off was observed with different transcripts and seems to be due to diminishing extract activity rather than mRNA instability insofar as 70% of uniformly labeled full-length mRNAs persisted until at least 60 min (data not shown).

View larger version (36K): [in this window] [in a new window]   FIG. 1. Characterization of the A. suum in vitro translation system. A, Renilla luciferase mRNA had a linear dose-response relationship for the given concentration range when incubated in translation extract at 30 °C for 1 h. RLU/sec, relative light units/second of 0.1% of the translation reaction. B, a time course of reporter accumulation from 125 ng of m7GpppG-Ren-A85 transcript indicated that reporter accumulation occured until at least 30 min after mRNA addition. C, ApppG-capped transcript translational efficiency relative to an uncapped mRNA (ppp-RNA). D, translational efficiency in the A. suum extract is methylation dependent, as shown by reporter activity of m7G- or TMG-capped transcripts relative to an ApppG-capped Renilla luciferase mRNA. A GpppG-capped transcript is translated in the extract, except in the presence of 0.5 mM SAH (a methylation inhibitor) (see "Results"). SAH itself has no significant effect on the translation of m7GpppG capped RNA (data not shown). E, effect of transcript cap and poly(A)-tract on translation in Ascaris extracts. Translational efficiency in the A. suum extract was methylated cap-dependent, as shown by reporter activity of m7GpppG- or m 2,2,73GpppG-capped transcripts relative to an ApppG-capped Renilla luciferase mRNA. Translational efficiency was also poly(A)-dependent, as shown by increased Renilla luciferase production when an A85 tail was added. The cap and poly(A)-tail had a synergistic effect on translation in A. suum extract, shown by the greater than additive effect on translation of endowing a transcript with an m7GpppG- or TMG-cap as well as a tail. Numbers in E are from a representative experiment with two replicates that was repeated on at least five occasions (combined data shown in C and D). All translation experiments illustrated were carried out in duplicate at least twice using independently prepared extracts and RNA transcripts. Footnotes in E: a, activity of given reporter/activity of ApppG-capped mRNA with same tail; b, activity of given reporter/activity of capped mRNA without tail; c, activity of given reporter/activity of ApppG-capped and untailed mRNA; d, synergistic effect of cap and tail (c/(a+b)) for given transcript.

We then tested the effect of a hypermethylated (TMG, m ^2,2,7₃GpppG) cap on translation. TMG-caps dramatically increase the efficiency of reporter translation, up to 195-fold relative to ApppG-capped mRNA (in the presence of SL1 and an A₈₅ tail, Fig. 1E). Together this shows that the cell-free system is extremely methylated cap-dependent and that both the m⁷G- and TMG-cap promote reporter production.

Poly(A)-tail dependence was assessed by comparing the translational efficiency of m⁷G- and TMG-capped Renilla luciferase transcripts with and without a poly(A)-tract (A₃₀ and A₈₅). The effect of the poly(A)-tail on translation was found to be similar for both m⁷G- and TMG-capped transcripts, with an increase in reporter activity upon addition of an A₈₅ tract (Fig. 1E). Addition of an A₈₅ tail to the transcript led to a 65- and 157-fold increase in reporter production from m⁷G- and TMG-capped transcripts, respectively. Transcripts with a poly(A)-tract of 85 nucleotides demonstrated a higher translation efficiency in the extract (2.5-fold) than transcripts with a shorter 30 nucleotide poly(A)-tract.

We then tested how translation is affected by addition of both a cap and poly(A)-tail. The presence of the m⁷GpppG-cap with an A₈₅ tail increased Renilla luciferase reporter translation by up to 17,362-fold, compared with an ApppG-capped transcript without a tail (ApppG-Ren, Fig. 1E). This is much greater than the predicted additive effect based on reporter transcripts carrying an m⁷GpppG-cap or A₈₅ tail alone. We therefore conclude that this is a synergistic effect (Fig. 1E) based on the equation: effect on translation of both a cap and tail/(effect of cap alone + effect of tail alone). By this equation, the m⁷G-cap and poly(A)-tail have a synergistic effect of 48. The TMG-cap and poly(A)-tail also drastically promoted translation when compared with ApppG-SLRen (17,979-fold increase). The TMG-cap and poly(A)-tail had a synergistic effect of 87 on reporter production. The same effects were seen with multiple preparations of mRNA as well as in independently prepared extracts. Therefore both the m⁷G- and TMG-cap synergistically enhance translation when combined with a poly(A)-tail in the A. suum embryonic extract.

The Effect of Cap Structure and the Spliced Leader Sequence on Translation in the A. suum Cell-free System—Since the discovery of trans-splicing, it has been speculated that spliced leader addition might have a profound effect on mRNA translation. Therefore, we set out to test the hypothesis that transsplicing affects translational efficiency in A. suum. To test the translational efficiency of mRNAs with different 5'-leaders in our system, we synthesized the transcripts represented in Fig. 2A. In vitro translation of TMG-Ren-A₈₅ transcripts was considerably less efficient than the equivalent m⁷G-capped transcript (Fig. 2, B and C). Addition of the SL1 sequence increased translational efficiency of TMG-capped transcripts almost 7-fold to 64% of m⁷G-capped mRNA. Interestingly, the SL1 sequence had the opposite effect on the mean efficiency of m⁷G-based translation, reducing reporter protein accumulation to 73%. Therefore, the SL1 sequence specifically enhances expression from TMG-capped transcripts but leads to a decrease in translation of m⁷G-capped mRNAs. By following decay of uniformly radiolabeled reporters, we observed that the effects of the m ^2,2,7₃GpppG-cap and SL1 sequence on reporter production were not a function of differential transcript stability (data not shown).

View larger version (42K): [in this window] [in a new window]   FIG. 2. The SL1 sequence synergistically enhances translation of TMG-capped transcripts but negatively affects translation of m7G-capped transcripts. A, transcripts used to study the effect of cap/SL1 combinations in A. suum translation extracts. The left column indicates names and abbreviated names (in parentheses) used to refer to transcripts. B, while a TMG cap alone reduced translation of the reporter mRNA (relative to m7G-Ren-A85), the SL1 sequence significantly enhanced the translation levels of TMG-capped RNAs. However, the SL1 sequence diminished relative translational efficiency of an m7G-capped reporter transcript. Error bars reflect variation in five experiments using independently prepared extracts and RNA transcripts. C, interaction of the TMG-cap with the spliced leader sequence was synergistic, as shown in a representative experiment involving two sets of replicates. Footnotes in C: a, activity of given reporter/Activity of ApppG-capped reporter with same 5'-UTR; b, activity of given reporter/activity of methylated, capped reporter without SL1; c, activity of given reporter/Activity of uncapped reporter without SL1; d, synergistic effect of cap and SL sequence (c/(a+b)) for given transcript.

A. suum eIF4E-3 Binds Both Monomethylated and Trimethylated Capped RNA and Cap Analogs—To examine the basis of cap recognition on trans-spliced versus non-trans-spliced transcripts in nematode embryos, we cloned an A. suum eIF4E cDNA using a combination of degenerate PCR and RACE from 32–64 cell stage oligonucleotide-capped RNA. All of the independent clones characterized corresponded to a single eIF4E sequence. Amino acid alignments and phylogenetic analysis revealed that this A. suum eIF4E is most closely related to C. elegans IFE-3 (Fig. 3, supplemental Table III), and we have named it eIF4E-3. Since A. suum eIF4E-3 appears homologous to the eIF4E with a well defined role in translation in most eukaryotes and is the only ife isoform proven to be present in 32–64 cell embryos (also see below), we focused our attention on its functional characterization.

View larger version (141K): [in this window] [in a new window]   FIG. 3. Multiple sequence alignment and phylogenetic analyses of the nematode eIF4E proteins reveal that A. suum eIF4E-3 is most closely related to C. elegans IFE-3. A, alignment of the A. suum eIF4E-3 with other nematode homologs. Only the conserved region of eIF4E (Ascaris amino acid 50–230) are illustrated in the alignment. Circled letters indicate residues previously implicated in specific functions, G = eIF4G interacting, C = cap binding, and p = phosphorylation site as described (34, 40). Boxed residues at 60–61 are thought to be involved in TMG-cap binding specificity as described for C. elegans eIF4E-5 (42). Arrows indicate the region used in phylogenetic analyses. B, phylogenetic analysis using the Neighbor-joining method (and strongly supported by Bayesian analysis of the corresponding nucleotide sequence, data not shown) supported the hypothesis that at the sequence level A. suum eIF4E is most closely related to C. elegans IFE-3. Starred nodes are supported with a posterior probability of >0.83 in Bayesian likelihood analysis, indicating strong support at these positions using multiple phylogenetic analysis tools. Species are Brugia malayi, Hemonchus contortus, C. elegans, and A. suum.

View larger version (27K): [in this window] [in a new window]   FIG. 4. A. suum eIF4E binds both m7G- and TMG-capped mRNA in a manner that is enhanced by eIF4G and influenced by the presence of SL1. A, cap-analog binding and competition experiments demonstrate A. suum eIF4E binds to m7GTP-Sepharose 4B and that binding is reduced in the presence of either 200 µM m7GTP or m 2,2,73GTP. The binding and competition data for Ascaris eIF4E to m7GTP-Sepharose 4B resemble those previously described for C. elegans eIFE-1 but not eIFE-3 (43). B, binding of recombinant eIF4E proteins to m7GTP-Sepharose and m 2,2,73GTP-Sepharose 4B revealed that A. suum eIF4E-3 binds both cap analogs and thus has cap analog binding properties resembling C. elegans IFE-1, not IFE-3. C, either m7GTP or m 2,2,73GTP can act as a competitor for Ascaris eIF4E-3 binding to m32,2,7GTP-Sepharose 4B. Overall, the data in A–C demonstrate that Ascaris eIF4E-3 can bind both m7GTP and m32,2,7GTP cap analogs, and this binding can be competed against with either m7GTP or m32,2,7GTP. D, A. suum eIF4E cross-linking to mRNA is affected by cap identity and 5'-UTR sequence. A. suum eIF4E cross-linked more strongly to uniformly labeled m7G- than TMG-capped Renilla luciferase RNA. Addition of the SL1 sequence to a TMG-capped transcript increased A. suum eIF4E cross-linking, while the presence of SL1 decreased the interaction with an m7G-capped mRNA. The bottom panel demonstrates cross-linking to cap-labeled RNA substrates where the 32P is located in the first phosphate of the triphosphate linkage (m7G32pppG-RNA). eIF4E-3 cross-linking to uniformly or cap-labeled m7G-capped transcripts decreases upon addition of SL1 sequence. E, UV cross-linking of A. suum eIF4E to TMG- or m7G-capped RNA is increased in the presence of eIF4G. In addition to CBP80, bovine serum albumin and His-tag-purified -galactosidase did not increase cross-linking. Sepharose binding and cross-linking experiments were carried out multiple times with independently prepared transcripts and protein preparations. F, increasing levels of m7GTP and m 2,2,73GTP competitors similarly inhibited translation of both m7GpppG-Renilla luciferase A85 and TMG-SL firefly luciferase A85 reporter transcripts in the A. suum cell-free system. Equal amounts of the two reporters were added to the translation system and translation of both reporters simultaneously carried out in the presence of the indicated inhibitor. Following 30 min of translation, the translated proteins were assayed using the dual luciferase assay. Data illustrated represent 0.1% of reporter activity assayed.

Cap binding of A. suum eIF4E-3 was further examined using cross-linking to determine whether recombinant C. elegans eIF4G (GenBank^TM NM_063328 [GenBank] ) influences the nature and affinity of cap binding. eIF4G enhanced UV cross-linking (3–4-fold) of A. suum eIF4E-3 to substrates with either a TMG- or an m⁷GpppG-cap structure (Fig. 4E).

Ascaris eIF4E-3 May Translate Both m⁷G- and TMG-capped mRNAs—The ability of A. suum eIF4E-3 to bind both m⁷G- and TMG-capped substrates led us to hypothesize that Ascaris embryo translation of trans-spliced and non-trans-spliced mRNAs might be carried out by A. suum eIF4E-3 and thus involve fewer eIF4E homologs than predicted for C. elegans. Preliminary translation experiments directed toward immunodepletion or m⁷GTP-Sepharose depletion of eIF4E from the extracts followed by reconstitution with recombinant A. suum eIF4E-3 were not successful. As an alternative approach, we therefore examined the effect of m⁷GTP and m ^2,2,7₃GTP inhibition on translation of m⁷GpppG-firefly luciferase and TMG-Renilla luciferase mRNAs simultaneously introduced into the A. suum translation extract. We predicted that if A. suum had eIF4E homologs with different cap specificities (as in C. elegans), translation of the m⁷G- and TMG-capped substrates would be differentially affected by the competitor nucleotides. As shown in Fig. 4F, m⁷GTP inhibited the translation of both m⁷G- and TMG-capped mRNAs in the extract with a similar inhibition curve. This is to be expected from the previous analyses of the C. elegans IFE proteins, which all bound m⁷GTP-Sepharose. However, m ^2,2,7₃GTP also inhibited translation of both m⁷G- and TMG-capped transcripts. Previous m ^2,2,7₃GTP-Sepharose binding studies using C. elegans IFE-3 and -4 suggest these proteins would not be inhibited by m ^2,2,7₃GTP (43). Together with the binding and cross-linking studies, our evidence therefore indicates that a single protein, A. suum eIF4E-3, might be fully capable of translating both trans-spliced and non-transspliced transcripts.

A. suum and Brugia malayi May Have Fewer eIF4E Isoforms than C. elegans—As C. elegans has five ife isoforms, we searched the Ascaris EST data base (40,000 ESTs from a variety of different stages and tissues) for other A. suum eIF4E cDNAs. Only one other eIF4E was identified and is present in a gut-derived cDNA library (e.g. GenBank^TM BQ380672 [GenBank] ). Both multiple sequence alignment and phylogenetic analyses show that A. suum eIF4E-4 has closest identity to C. elegans IFE-4 at the amino acid level (Fig. 3, supplemental Table III). The whole genome sequence (5-fold coverage) of the filarial nematode (www.tigr.org/tdb/e2k1/bma1) contains three eIF4E corresponding by phylogenetic analysis to a C. elegans ife-3, ife-4, and one form of the ife-1, -2, and -5 group (Fig. 3). A similar set of ife genes is found in another nematode, H. contortus. Thus, we have identified eIF4E genes resembling IFE-3 and -4 in all of the nematodes examined but have been unable to identify multiple copies of genes related to ife-1, -2, and -5 in nematodes outside of the Caenorhabditis group. This suggests that fewer eIF4E isoforms might be present in nematodes other than C. elegans.

The SL Sequence Trans-splices Close to the Start Codon of Transcripts—It has previously been observed that the SL1 sequence tends to splice close to the start codon (3), a phenomenon that will affect the sequence and length of the 5'-UTR of trans-spliced transcripts. As previous analysis had not examined SL1 and SL2 sequences separately, we also addressed whether both SL1 and SL2 would be found close to the start codon. We identified a set of C. elegans cDNAs in the data bases that contain an SL1 or SL2 sequence and noted the SL to AUG distance (as described under "Experimental Procedures"). We found that both SL1 and SL2 do indeed tend to splice close to the start codon, with 49% of transcripts containing a spliced leader sequence within 10 nucleotides of the initiator AUG (Fig. 5A). Even within the set of cDNAs where the SL1 to AUG distance is less than 10 nucleotides, there is a tendency for the spliced leader to lie close to the start codon (Fig. 5B). When we extended this analysis to Ascaris and other nematode species, we found that the SL1 sequence trans-splices close to the AUG in nematodes other than C. elegans. The sample size is somewhat small as many nematode ESTs do not contain the entire spliced leader sequence. However, in species that phylogenetically group with A. suum (Clade III nematodes), at least 45% of cDNAs contain an SL1 sequence within 10 nucleotides of the start codon (Fig. 5C and see supplemental Fig. 1).

View larger version (18K): [in this window] [in a new window]   FIG. 5. SL sequences tend to trans-splice close to the start codon in nematode transcripts. A, a survey of SL1 and SL2 transspliced C. elegans transcripts revealed that trans-splicing tends to occur close to the start AUG. B, examination of 59 C. elegans cDNAs with less than 10 nucleotides separating the SL1 sequence and start codon reveals that greater than 70% of the transcripts have the SL1 sequence within four nucleotides of the predicted start AUG. C, the SL1 sequence also tends to trans-splice close to the start codon in 182 mRNAs from nematodes including Ascaris and five species that are phylogenetically related to A. suum (so-called clade III nematodes including Dirofilaria immitis, Onchocerca volvulus, Litomosoides sigmodontis, B. malayi, and Toxocara canis). Data for each nematode included in the compilation are illustrated in supplementary Fig. 1.

The SL1 Sequence Has an Optimal Distance for Translation from the Start Codon—To test the hypothesis that the position of SL1 trans-splicing affects translation, we synthesized the transcripts illustrated in Fig. 6A containing an increasing number of random bases in the 5'-UTR (Ns, where N = A, U, G, or C; derived from degenerate oligonucleotide primers between the spliced leader and the start AUG) (see supplemental Table II). It has previously been shown that the three nucleotides upstream of the start codon, particularly the –3 position, affect mammalian translation (31). To eliminate the effect of randomizing the three nucleotides at the –3 to –1 positions, we initially fixed these as ACC (nucleotides that are found at these positions when nematode transcripts are surveyed, as well as the Kozak consensus for these positions). To facilitate normalization of our translation assays, an m⁷GpppG-capped firefly luciferase mRNA was included as an internal control with the Renilla luciferase test transcript. The translation of both mRNAs was measured using the dual luciferase assay system and luminescence from the latter reporter normalized relative to the former. When the distance between the SL1 sequence and the start codon was increased by three nucleotides (sequence ACC), translational efficiency increased (up to 7.5-fold, Fig. 6B, panel 1, first two bars). When seven additional nucleotides were added to the ACC (increasing SL to AUG distance to 10 bases), there was a drop in translational efficiency (Fig. 6B). As SL to AUG distance was further increased, reporter activity remained relatively low. This may explain the tendency of SL1 to splice close to the start codon described above (Fig. 5, A–C).

View larger version (23K): [in this window] [in a new window]   FIG. 6. Analysis of SL1 sequence to start codon distance indicates there is an optimal SL1 to start codon distance for translation of reporter transcripts. A, transcripts used to analyze the effect of SL1 to AUG distance on in vitro translation in A. suum embryonic extracts. In a series of test transcripts, SL1 sequence was separated from the AUG by an increasing number of random bases derived from degenerate oligonucleotides synthesized with equal proportions of each base at a position (see "Results" and supplemental Table II). The effect of 5'-UTR length and cap identity was also tested, using a series of constructs with an increasing number of random bases between the cap and the start codon (see supplemental Fig. 2). (B) The placement of "ACC" between the SL1 sequence and the start AUG enhanced translation (two left bars, panel 1), an effect that diminished in transcripts as the SL1 to AUG distance was increased (left group of bars). Renilla luciferase activity was normalized to an internal control (m7GpppG-capped firefly luciferase reporter). The drop in translational efficiency with distance was not simply due to increasing 5'-UTR length or the TMG cap to AUG distance, as translation was enhanced with 5'-UTR length (up to a certain distance) in the absence of SL1 sequence (panels 2 and 3). Although the relative differences between reporter activity vary from experiment to experiment, the illustrated trend was consistently seen. C, the effect of the –3 residue on translation efficiency. Constructs used to test the effect of –3 residue identity relative to the start codon are illustrated and the translation efficiencies of the three constructs shown. The presence of anAatthe –3 position enhanced reporter production over a G or U in the nematode in vitro translation extract. All data shown (n = 2) are representative of results from experiments repeated at least six times with independently synthesized mRNAs and prepared translation extracts.

To test whether this effect was due to the SL1 sequence itself or the increasing length of the 5'-UTR in these experiments, we replaced TMG-SL1 with m⁷G-N₂₁. The optimal 5'-UTR length for SL1 containing mRNAs is distinct from that observed for m⁷G-ACC-Ren, where the optimal total length is 31 nucleotides (Fig. 6B, panel 2; see also supplemental Fig. 2). Thus the drop in reporter activity with increasing SL1 to AUG distance was not due to increasing 5'-UTR length (or the presence of randomized nucleotide sequence) but was related to either the SL1 sequence or TMG-cap. Note also that the optimal TMG-cap to start codon distance is greater when SL1 is replaced with random nucleotides (Fig. 6B, panel 3). Consequently, changes in translation efficiency were specifically due to varying distance between the SL sequence and the start codon and were not simply a function of 5'-UTR length or cap identity. The tendency of SL to trans-splice close to the start codon may therefore reflect evolutionary selection for optimal expression (see "Discussion").

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Insights into Nematode Translation and the Role of the Spliced Leader Sequence—We have described a nematode in vitro translation system that displays attributes of in vivo translation. Translation in the A. suum cell-free extract is cap-dependent, increases with poly(A)-tail length, and shows a synergistic interaction between the cap and poly(A)-tail. These and other features of the cell-free system reflect our in vivo translation data obtained following biolistic introduction of the same reporter transcripts in Ascaris embryos.³ Start codon context and 5'-UTR length were shown to contribute to translation efficiency. In particular, an adenine at the –3 position in the transcripts analyzed leads to higher translation than either G or U at this position. Aspects of translation defined in other metazoa are conserved in nematodes including the contribution of start codon context and 5'-UTR length and cap and poly(A)-tail synergism on translation efficiency. The synergy between the TMG-cap and poly(A)-tail indicates that the same interactions occur in trans-spliced transcripts. Notably, the synergism between the TMG-SL and poly(A)-tail was greater than that observed for the m⁷G-cap and poly(A)-tail. Cross-linking of eIF4E to m⁷G- and TMG-capped RNAs was also enhanced by the presence of eIF4G, further demonstrating that interactions between translation initiation factors are conserved in nematodes.

While the presence of either the spliced leader or a trimethylguanosine cap alone to a test transcript leads to a significant drop in translation efficiency, together these features synergistically enhance translation. Previous studies also observed a collaborative effect between the TMG-cap and SL sequence and a detrimental effect of either the TMG cap or SL sequence alone on an endogenous trans-spliced mRNA (14). Our experiments have extended these studies using an optimal 5'-UTR length and by examining multiple reporters to eliminate sequence specific effects on translation by the body or 3'-UTR of the transcript. In contrast to Maroney et al. (14), we find that non-trans-spliced transcripts are more efficiently translated than trans-spliced mRNAs, an effect consistent with our in vivo observations.³ We attribute the differences observed to 1) extract preparation, 2) potential contributions of the endogenous Ascaris mRNA sequences to translation efficiency in the earlier study, and 3) differences in the 5'-UTR sequences. In addition, our translation assays measured luciferase activity derived through addition of an exogenous mRNA that must compete for translation with all endogenous mRNAs in the extract preparation. In contrast, previous studies removed endogenous RNAs by micrococcal nuclease treatment and measured [³⁵S]methionine incorporation into protein.

We have further found that the optimal 5'-UTR length for trans-spliced test transcripts is 25 nucleotides (including the SL) and 31 nucleotides for non-trans-spliced test transcripts. Our analyses (see Fig. 5 and supplemental Fig. 1) indicate that transcripts with a non-optimal SL to AUG distance are definitely present in vivo. This is not unexpected as factors affecting translation of each mRNA are complex. Indeed, mRNAs are unlikely to be translated at similar levels but their expression finely tuned to produce the required levels of protein by a variety of intrinsic and extrinsic factors.

The SL1 sequence is remarkably conserved across nematode species. This sequence conservation may reflect several roles for the SL1 sequence. For example, the SL sequence has been shown to be a promoter element for SL RNA transcription (20). Although lethality caused by a deletion of the SL1 donor genes can be rescued by introduction of SL sequences that are substantially changed or even extended in length (57, 58), more directed and subtle assays show that mutations in the SL1 sequence can significantly alter the translatability of a transspliced mRNA (14). These SL1 mutations presumably affect the ability of the spliced leader to synergistically collaborate with the TMG-cap to promote translation. Overall, these and our data suggest that the SL1 sequence itself influences translation.

Bioinformatics data indicate that the SL sequence tends to trans-splice close to the start codon in a diversity of nematode species and that this evolutionary conservation is functionally reflected in general with the optimal SL to AUG distance for reporter mRNA translation in the cell-free system. Our data also show a correlation between translational efficiency of TMG-SL1 containing mRNAs and cross-linking levels to A. suum eIF4E-3. This suggests that the TMG-cap and SL1 together might influence interactions between the transcript and the translation initiation complex, perhaps reflecting specific requirements for translation initiation of trans-spliced transcripts. This model is consistent with recent evidence that the spliced leader, as well as the so-called cap 4 methylation status, can affect association of transcripts with polysomes in the kinetoplastid Leishmania tarentolae (59).

The SL1 sequence has a specific effect on optimal 5'-UTR length when compared with random sequence. Thus, addition of the spliced leader sequence may also provide a relatively short, uniform 5'-UTR that is free of small upstream open reading frames and optimal for ribosomal scanning. This model might predict that trans-spliced nematode primary transcripts may not be optimal for translation. We have attempted to examine this question by identifying the site of transcription initiation for several trans-spliced A. suum genes to define and examine the region upstream of the AUG in primary transcripts. This has proven difficult as multiple techniques fail to consistently identify a discrete transcription start site for two genes (v-ATPase and the ribosomal gene L23a, data not shown), although these assays can define a transcription start site for a non-trans-spliced gene (glutathione S-transferase). Notably, transcription initiation sites for trans-spliced C. elegans genes also have not been definitively characterized. Our inability to discretely define transcription initiation sites could be consistent with the hypothesis that transcription initiation may be heterogeneous for such genes (17). Trans-splicing might then function to trim these heterogeneous primary transcripts, generating uniform 5'-UTRs that are more optimal for translation.

Overall our data suggest that the translation machinery in nematodes may have evolved to translate trans-spliced mRNAs with 5'-UTRs of a particular length containing both the spliced leader and its TMG-cap. Therefore, trans-splicing of the SL1 sequence may serve at least two functions in nematodes, generation of an optimal 5'-UTR length as well as the sequence context for translation of trimethylguanosine capped transcripts.

Translation Initiation and the Evolution of the eIF4E Genes in Trans-splicing Organisms—We have cloned eIF4E-3 from A. suum and our cross-linking studies indicate that its cap-binding characteristics may explain differential activity of TMG-versus m⁷GpppG-capped reporters in the extract, as well as the effect of the SL1 sequence on translation. A. suum eIF4E-3 can bind both immobilized m⁷GTP- and m ^2,2,7₃GTP-Sepharose, as well as m⁷G and TMG-capped mRNAs. Our results further indicate that while A. suum eIF4E-3 is most closely related to C. elegans IFE-3 in sequence, its binding properties resemble those of C. elegans IFE-1, -2, or -5. Interestingly, A. suum eIF4E-3 binds TMG-cap even though residues previously postulated as being critical to C. elegans IFE-3 m⁷GTP-selective cap binding are conserved in the two proteins. In particular, mutational analysis of C. elegans IFE-5 (⁶⁴NV⁶⁵ YL; see equivalent residues 60–61 boxed in Fig. 4) increased IFE-5 m⁷GTP binding, indicating that these residues play a role in cap-binding selectivity (42). C. elegans eIF4E isoforms with YL residues at this position were reported to be less efficient at binding m ^2,2,7₃GTP. C. elegans IFE-3, which only binds m⁷GTP, has the YL residues. However, A. suum eIF4E-3 has residues YL at the equivalent position but binds both m⁷GTP- and m ^2,2,7₃GTP-Sepharose and cross-links to the corresponding RNAs. This indicates the YL residues may not be as important for selective binding to m⁷GTP in the Ascaris eIF4E-3, and additional residues that have yet to be defined may contribute to cap-binding specificity.

Regions of H. sapiens eIF4E are relatively unstructured until nucleotide binding (60, 61). In addition, both eIF4E and eIF4G undergo conformational changes on interaction with each other and on nucleotide binding (40). Steric flexibility and conformational changes might mean that although A. suum eIF4E-3 has a similar sequence to C. elegans IFE-3, it can still accommodate a trimethylated cap nucleotide. The structure of the A. suum eIF4E-3 binding pocket containing a nucleotide is therefore predicted to differ in its dimensions from that defined for the human protein and predicted for C. elegans IFE-3 (42). Indeed preliminary structural homology modeling indicates that the dimensions of the A. suum eIF4E-3 might be larger than those of its closest sequence homolog IFE-3, a prediction that could explain why the A. suum protein can accommodate the TMG-cap while C. elegans IFE-3 cannot.²

Our characterization of A. suum IFE-3 suggests that additional eIF4E homologs such as those observed in C. elegans (e.g. ife-1, -2, and -5) might not be necessary for TMG-cap binding of transcripts in Ascaris embryos. Analysis of the sequenced genomes of two other trans-splicing metazoa, C. intestinalis and Schistosoma mansoni, suggests that they do not have a distinct ife-1, -2, and -5 class homolog based on molecular phylogenetic analyses. Therefore, current data indicate there might not be an ancestral eIF4E homolog specifically shared by trans-splicing organisms, as proposed previously (16). In addition, it has become clear that many non-trans-splicing organisms, including H. sapiens, have multiple eIF4E isoforms. For example, spliced leader trans-splicing has not been detected in Drosophila melanogaster, yet this organism has 6 eIF4E-like genes (62).

Trans-splicing and Polycistronic Transcripts—While operons have not yet been shown to exist in A. suum, the apparent absence of an SL2 sequence suggests that if polycistronic transcripts do arise they may be resolved by SL1 trans-splicing. Indeed SL1 has been shown to resolve polycistronic transcripts in C. elegans (63). The genome sequence of B. malayi is currently available (www.tigr.org/tdb/e2k1/bma1/) and will provide information on whether additional spliced leaders are present in this nematode and whether gene organization suggests that polycistronic transcripts are likely to be present. However, the origins of SL trans-splicing and its ancestral role (in resolving polycistronic transcripts versus promoting translation) within the nematodes will likely only be revealed if basal species (those potentially closer to the common ancestor of nematodes) are examined. In addition, functional analysis of trans-splicing in flatworms (S. mansoni), H. vulgaris, and C. intestinalis will shed light on whether trans-splicing arose in a common ancestor or has independently evolved in multiple lineages. Such issues will eventually be resolved by identifying the full phylogenetic extent of trans-splicing, coupled with functional analysis in multiple species.

	FOOTNOTES

 
^* This work was supported by National Institutes of Health Grants AI49558, AI043300, and AI039714 and City University of New York-College of Staten Island startup funds (to R. E. D.) and by Grants PBZ-KBN-059/T09/10 and KBN 3 P04A 021 25 from the Polish Committee for Scientific Research to (E. D.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. 1 and 2 and Tables 1–3.

This article was selected as a Paper of the Week.

Current address: Dept. of Biology, New York University, New York, NY 10012.

^** To whom correspondence should be addressed. Current address: UCHSC at Fitzsimons, Depts. of Biochemistry and Molecular Genetics and Pediatrics, Mail Stop 8101, P. O. Box 6511, 12801 East 17th Ave., Aurora, CO 80045. Tel.: 303-724-3226; Fax: 303-724-3215; E-mail: richard.davis{at}uchsc.edu.

¹ The abbreviations used are: SL, spliced leader; m₃^2,2,7GpppG (or TMG), 2,2,7-trimethylguanosine; m⁷GpppG (or m⁷G), monomethylguanosine; UTR, untranslated region; PBS, phosphate-buffered saline; DTT, dithiothreitol; PMSF, phenylmethylsulfonyl fluoride; RACE, rapid amplification of cDNA ends; EST, expressed sequence tag; SAH, S-adenosylhomocysteine.

² R. E. Davis, unpublished data.

³ L. S. Cohen, C. Mihkli, M. Jankowska-Anyszka, J. Stepinski, E. Darzynkiewicz, and R. E. Davis, manuscript in preparation.

	ACKNOWLEDGMENTS

 
We acknowledge Leah Cohen, Claudette Mikhli, David Ndegwa, Yubin Zhang, Scott Hodgson, Eugenia Ribeiro, and Angelica Parra for c