Unique tRNA Introns of an Enslaved Algal Cell
http://www.100md.com
分子生物学进展 2005年第8期
Cell Biology, Philipps-University Marburg, Karl-von-Frisch Sta?e 8, 35032 Marburg, Germany
E-mail: maier@staff.uni-marburg.de.
Abstract
Nucleomorphs are remnant nuclei of eukaryotic, secondary endosymbionts exclusively found in cryptophytes and chlorarachniophytes. The nucleomorph of the cryptophyte Guillardia theta codes for 36 transfer RNA (tRNA) genes, 15 of them predicted to contain introns and 1 pseudo-tRNA. Some of the predicted intervening sequences are manifested at positions not known in Eukarya, even tRNAs with more than one intron were suggested. By isolating reverse transcriptase–polymerase chain reaction products of the spliced tRNAs we verify the processing of all predicted intron-harboring tRNAs and demonstrate the splicing of the smallest introns (3 nt) investigated so far. However, the spliced intervening sequences are in some cases shifted in respect to the predicted ones. Moreover, we show that introns, if inserted into the B-box of tRNA genes in the nucleomorph of cryptophytes, mimic promoter regions and do not abolish transcription by RNA polymerase III. Consequently, internal nucleomorph-encoded tRNA promoter regions are in some cases dissected from the sequence of the mature tRNAs. By reanalyzing tRNA introns of a recently sequenced red algae we furthermore show that splicing of introns at unusual positions may be introduced in cryptophytes by its secondary endosymbiont. However, in contrast to the rest of the symbiont genome, introns are not minimized in quantity but are instead scattered along the tRNA genes.
Key Words: Guillardia theta ? Cyanidioschyzon merolae ? tRNA ? nucleomorph ? secondary endosymbiosis
Introduction
It is generally accepted that the genetic system of eukaryotes is related to that of Archaea and seems to be different from the transcription-translation machinery of Bacteria (Martin and Müller 1998; Martin et al. 2001). Such a difference can be observed in the splicing mechanisms of transfer RNA (tRNA) introns as well. These types of introns are common in all three domains, Bacteria, Archaea, and Eukarya (Marck and Grosjean 2002), but Bacteria evolved self-splicing tRNA introns, whereas tRNA introns from Archaea and Eukarya were processed in an enzymatic catalyzed reaction (Abelson, Trotta, and Li 1998). tRNA introns, if present, could be located at different positions in Archaea (Belfort and Weiner 1997) and are restricted to a conserved position between the first and second base 3' of the anticodon in Eukarya (Marck and Grosjean 2002) (fig. 1). Moreover, as recently reported, red algae may encode tRNAs with introns, which are predicted at positions outside the conserved position near the anticodon (Matsuzaki et al. 2004) (fig. 1). In Bacteria self-splicing group I introns are described, which are located 3' of the anticodon or within the anticodon (Kuhsel, Strickland, and Palmer 1990; Xu et al. 1990; Reinhold-Hurek and Shub 1992; Biniszkiewicz, Cesnaviciene, and Shub 1994) (fig. 1).
FIG. 1.— Comparison of intron positions in the nucleomorph, other eukaryotes, Bacteria, and Archaea. Arrows show intron positions. The thick arrow in the scheme of nucleomorph tRNA indicates the region from which a "variable loop" is removed by splicing. Data for the intron positions of Archaea-specific introns are from Marck and Grosjean (2002); for eubacterial group I introns see text.
Several algae evolved by the enslavement of a phototrophic eukaryotic cell by another, except heterotrophic eukaryote (secondary endosymbiosis) (Delwiche and Palmer 1998; McFadden 1999; Maier, Douglas, and Cavalier-Smith 2000). During the evolution of a "cell within a cell," the symbiont organelles were reduced or eliminated. As in some secondary evolved algae the reduction led to a plastid surrounded by additional membranes, and an intermediate stage in the reduction of the symbiont organelles is found in cryptophytes and chlorarachniophytes (Cavalier-Smith 2002; Stoebe and Maier 2002). Here a remnant cytoplasm of the eukaryotic symbiont including its pigmy nucleus, the nucleomorph, is maintained. The nucleomorph genome of one cryptomonad, Guillardia theta, was sequenced, thereby lightning up a reduced eukaryotic genome organized in three small chromosomes (Zauner et al. 2000; Douglas et al. 2001). This miniaturized genome is maintained to encode genes whose products have important functions in the plastid. To synthesize these functions, a transcription and translation apparatus has to be expressed in the secondary symbiont cytoplasm.
By scanning the nucleomorph genomic sequence for tRNA-encoding genes, we showed that 35 transfer DNAs (tDNAs) should be encoded by the nucleomorph (Douglas et al. 2001). However, the nucleomorph should not be autonomous in respect to tRNAs due to a missing tRNA for glutamate (Douglas et al. 2001). In addition, we described one tRNA intron, located at a position otherwise not known from Eukarya (Zauner et al. 2000).
In the paper at hand, we show that splicing of tRNA introns, which are unique for eukaryotic organisms in respect to its localization, lead to bona fide tRNAs. In this respect, splicing of tRNA introns, located 3' or 5' of the anticodon, in the variable loop, TC-loop, or D-loop, is demonstrated. Moreover, we demonstrate splicing of the smallest eukaryotic introns (3 nt) and show that insertion in the internal promoter regions of tRNA genes do not abolish the transcription and splicing of tRNAs. Our data indicate that the ability to splice tRNA introns located at unique positions within the nucleomorph genome was most likely the legacy of the red algae progenitor. However, in the nucleomorph genome tRNA introns were shortened, but in contrast to other noncoding chromosomal elements, spread over most possible tRNA gene localizations, and contradict therefore the general tendency of streamlining the nucleomorph genome.
Materials and Methods
Reverse Transcriptase–Polymerase Chain Reaction and Sequencing
RNA from G. theta was isolated and reverse transcriptase–polymerase chain (RT-PCR) reactions were done according to manufacturer's instructions (MBI Fermentas, St. Leon-Rot, Germany). Primer sequences are given in supplementary table 1 (Supplementary Material online). RT-PCR products were isolated from 17% denaturing polyacrylamid gels and cloned into pGEM-T? (Promega, Mannheim, Germany). Individual clones were sequenced on a LI-COR 4200 sequencer using labeled M13-20 standard primers. In some cases, we detected additional splicing variants harboring insertions-deletions as well as exchanges of bases ("editing"). In the case in which the variants were isolated only one time or show no secondary structure specific for tRNAs, they were not included in our analyses. Instead, they were considered as PCR artifacts or as improper transcripts, which were detected in nucleomorph-specific messenger RNAs (mRNAs) in another study as well (Fraunholz, Moerschel, and Maier 1998).
tDNA Detection and Intron Prediction
tDNA detection and intron prediction were done with tRNAscan-SE provided at http://www.genetics.wustl.edu/eddy/tRNAscan-SE/ using relaxed parameters (cove only). tDNA promotor prediction was performed with POLIIISCAN provided at http://wwwmgs.bionet.nsc.ru/mgs/programs/poliiiscan/.
Characterization of Cotranscription
Cotranscription of tRNA genes with upstream-located genes was identified in the case of hlip-tRNA LeuUAG and uce-E2-tRNA PheGAA in expressed sequence tag clones. Two further cotranscribed tRNAs were isolated by RT-PCR experiments. For these, specific primers for the upstream-located gene (snrpD and tcpZ1) and for the tRNA genes (tRNA ValUAC and tRNA IleUAU) were used for RT-PCR experiments, respectively.
To isolate complementary DNAs (cDNAs) covering hlip (high light-induced protein), our cDNA library was screened with a gene-specific probe. Positive clones were sequenced as described above.
Results
The Nucleomorph Is Expected to Encode tRNAs with Unique Introns
As predicted by tRNAscan-SE (Lowe and Eddy 1997), 36 tRNAs and 1 related pseudogene should be encoded in the nucleomorph genome of G. theta. This lead to a set of tRNAs offering all amino acids with the exception of a tRNA glutamate (Glu, E), which seems to be missing in the nucleomorph genome of G. theta (Douglas et al. 2001). The predicted tRNAs show proper anticodons flanked 5' by a T and 3' by a purine, as it is common in eukaryotes except for initiator tDNA iMetCAU in higher eukaryotes (Marck and Grosjean 2002).
As in other eukaryotes, the tRNA genes (tDNAs) are dispersed in the nucleomorph genome and not clustered as it is found predominately in bacteria (Marck and Grosjean 2002). Furthermore, 3'-CAA is not part of the tDNA (Tomita and Weiner 2001), and A- and B-boxes, representing the internal promoter, can be identified in the predicted tDNAs (Spargue 1995) (fig. 2).
FIG. 2.— Sequence of tRNA genes and processed tRNAs. The sequence of tRNA genes (tDNAs) and the processed tRNAs are shown. The first line in the columns indicates the tDNA. Green positions show the B-box, blue the A-box, and red the anticodon. The predicted introns are shown in bold letters. Value of the A- and B-box refers to the F value predicted by POLIIISCAN. (i) = intermediate form; (i2) = second intermediate form. tRNAs are indicted in the three-letter code. The red colored A refers to the appearance of a tRNA GluUUC transcript with an additional A at position 53. (See also supplementary table 2, Supplementary Material online.)
Further on tRNAscan-SE (Lowe and Eddy 1997) predict introns at the highly conserved position between the first two bases 3' of the anticodon in eukaryotes. In addition the prediction program indicates introns at unusual positions (fig. 2). Out of the 36 nucleomorph-encoded tRNAs, 15 were predicted to encode at least one intervening sequence. These potential nonsplicesomal introns vary in length from only 3 up to 24 nt (fig. 2). Some of them would be unique for the domain Eukarya in respect to its localization and should emphasize the unique nature of some nucleomorph-encoded tRNA genes: seven of the tDNA introns are predicted to be located in the conserved position between the first two bases 3' of the anticodon, seven in the D-loop, and eight in the TC-loop (fig. 1). Interestingly, six tDNAs should encode pre-tRNAs with more than one intron. Five of them are predicted to harbor one intron in the D- and another one in TC-loop, which would represent a unique arrangement of intron insertions into a eukaryotic tDNA (fig. 1). Surprisingly, some of the introns are inserted into the B-box of tRNA genes, which would interrupt the promotor structure and therefore suppress transcription of these genes by RNA polymerase III (fig. 2).
Furthermore, nucleomorph chromosomes are built up of terminal repeats bordering a single-copy region (Zauner et al. 2000). Housekeeping genes located within the single-copy regions show an average A/T content of 77%. The A/T content of the tDNAs is remarkably lower (55%), reflecting the structural constrains of the two- and three-dimensional structures of the tRNAs. On the other hand A/T content is increasing in the predicted tRNA introns significantly (81% A/T), which shows the preference of A/T in noncoding regions.
tRNAs of the Symbiont of Cryptophytes are Encoded in the Nucleomorph
As the extremely condensed nucleomorph genome of the cryptophyte G. theta has only a limited coding capacity (Douglas et al. 2001), it is possible that tRNAs are imported from the host compartments to supplement missing functions as it is known from other organisms (Tan et al. 2002; Esseiva et al. 2004). In order to prove if tRNAs of the secondary endosymbiont originate exclusively from nucleomorph genes or from a second gene located in the cell nucleus, we amplified exemplarily four tRNA genes (tDNAs for AlaUGC, CysACA, LeuUAG, and PheGAA) from genomic DNA of G. theta. As the tDNAs for AlaUGC are predicted to encode one intron and the genes for tRNA CysACA and PheGAA even two (see later), primers were designed such that neither a tDNA with nor a possible nuclear-encoded tDNA without the predicted intron(s) is favored in PCR reactions. These experiments generated solely nucleomorph-specific amplificates thereby demonstrating that only nucleomorph-specific tDNA serves as matrix for spliced tRNAs. As, therefore, no indications for a nucleus-encoded symbiont-specific tRNA exist, our experiments exclude an import of nucleus-encoded tRNAs into the symbiont in the analyzed cases.
Introns of the Anticodon Loop are Spliced by the Secondary Endosymbiont
In order to prove whether the eukaryotic symbiont of cryptophytes is able to excise the predicted introns or the intron-harboring tDNAs are pseudogenes, we performed RT-PCRs. As control, tRNA ArgUCG and tRNA GlnCUG, whose tDNAs show no evidence for introns, were amplified from G. theta total RNA. As expected, both tRNAs are collinear with the genomic sequence (fig. 2). Next, we amplified the nucleomorph tRNAs ArgCCU, ArgUCU, TrpCCA, and TyrGUA (fig. 2). The tDNAs possess a predicted intron at the conserved position between the first two bases 3' of the anticodon. The RT-products confirmed that introns are encoded in the tDNAs and that they are removed in the mature tRNAs. The predicted localization of the introns concurs with the obtained RT-products in the tRNAs ArgUCU and TrpCCA. Surprisingly, the introns of ArgCCU and TyrGUA are shifted downstream in respect to its predictions. In the case of ArgCCU the intron is also extended for four bases (fig. 2). Therefore, introns in the tDNAs for ArgCCU and TyrGUA are inserted at positions otherwise not known from Eukarya. However, the mature tRNAs show bona fide anticodons as well as the conserved sides for the three-dimensional structure (fig. 3). Interestingly, the anticodon stem in tRNAs ArgCCU and TyrGUA shows one mismatch. At the moment, we cannot decide whether this secondary structure leads to a nucleomorph-specific functional tRNA or is a characteristic of nonfunctional tRNA.
FIG. 3.— Scheme of conserved tRNA bases found in the processed nucleomorph-specific tRNAs. Clamps indicate the predicted internal base pairing in the tertiary structure. Lines 1–3: conventionally used numbering system for tRNA positions. Line 4: secondary structure of the tRNA cloverleaf model. Solid boxes indicate the anticodon (AC). V: variable loop positions 44–48 (X: UGAGGUAAUCA; X1: GGUUACAACC; X2: GGGUACUACCU). The putative intron in MetCAU not found to be spliced is indicated as A*(UUGUACGCG).
Pseudogenes, the Missing tRNA for Glutamate and Introns 5' of the Anticodon
In nucleomorph chromosome III one stretch of bases is predicted to encode a pseudogene with some similarities to a potential tDNA (fig. 2). No anticodon was predicted precisely in this genomic region, but we noticed that conserved bases necessary for three-dimensional folding as well as possible A- and B-boxes for transcription initiation are present. Characterization of the maturated transcript of this region showed a bona fide tRNA LeuUAA (fig. 2), which is created by removing a 10-nt-long intron 5' of the anticodon and a 3-nt one in the TC-loop. The latter is one of the smallest introns identified so far and, interestingly, integrated within the B-box (see later).
As mentioned in Douglas et al. (2001), we noticed that the nucleomorph of G. theta should encode at least one tRNA for the whole set of amino acids with the exception of glutamate (Glu, E), for which no nucleomorph-located tDNA could be identified. On the other hand, we detected two tDNAs encoding a tRNA for cysteine. One of them, CysACA, is anticipated to harbor an 18-nt insertion at the conserved intron position 3' of the anticodon (fig. 2). However, the RT-PCR product of the predicted tRNA CysACA differs remarkably in respect to the intron positioning and anticodon prediction. As indicated in figure 2, splicing of a 10-nt-long intron containing the predicted anticodon occurred. This creates a potential tRNA with a "new" anticodon prediction for GluUUC harboring an additional intron located in the conserved position 3' of the UUC anticodon. This second intron is spliced as well (fig. 2). Thus, the missing tRNA providing glutamate is hidden in a tDNA predicted to encode tRNA CysACA.
The tDNA GluUUC is not only exceptional by harboring two introns. By studying the genomic version as well as the different intermediate forms, we detected that after splicing the 5' intron three different forms exist. In one, the intron 3' of the anticodon is spliced, in the second an additional a G to A transversion was detected, and in the third form an A is inserted (supplementary table 2, Supplementary Material online). However, in the tRNA from which both introns are spliced, only the additional A is present (fig. 1). As the different forms were isolated in independent experiments several times, a complicated maturation of this tDNA GluUUC may be indicated by the different intermediates.
Single Introns Located Within the TC- or the D-Loop
Previously, we have shown that the symbiont-specific tRNA SerAGA, whose gene is encoded in the nucleomorph of G. theta, contains two introns (Zauner et al. 2000). One is located 3' of the anticodon and the other one in the D-loop; the latter one was the first description of an intron in this loop in a eukaryotic encoded tDNA. An extended in silico search for further unusual tRNA introns suggested five nucleomorph-specific intron positions otherwise not known from eukaryotes: three in the TC-loop (tDNA AlaUGC, IleUAU, and LeuCAA) and two in the D-loop (tDNA AsnGUU and ValAAC) (fig. 2). We amplified the spliced molecules of these tDNAs. The tRNAs AlaUGC, IleUAU, and LeuCAA were processed as predicted (fig. 2). Interestingly, as in tRNA LeuUAA, a 3-nt intron is correctly spliced from the primary transcript of tRNA AlaUGC. In all three genes, although the predicted introns are located within the B-box, these tDNAs are transcribed. In the case of the D-loop–located introns of tRNAs AsnGUU and ValAAC the intron prediction differs from in vivo situation. Whereas in the tRNA AsnGUU the intron is shifted for one base, the intron of tRNA ValAAC is shifted and extended by 1 nt. Moreover, we detected in the tDNA ValAAC an additional intron, which is again 3-nt-long and inserted into the B-box region (fig. 2).
Splicing of tRNAs with Multiple Predicted Introns
As shown above, the nucleomorph harbors tDNAs including introns that are unique for the domain Eukarya in respect to its localization. tRNAScan-SE predicts further nucleomorph-specific tDNAs, which may harbor more than one intron. This is true for two tDNAs (PheGAA and CysGCA) which are predicted to encode two introns, one in the D- and one in the TC-loop, as well as for elongator tDNA eMetCAU with a prediction of three introns, one in the conserved position 3' of the anticodon, one in the D-, and one in the TC-loop.
In the case of tDNA PheGAA our experiments show that both introns are processed. Interestingly, both introns were discovered to be shifted two bases to the 5'; the TC-loop intron is furthermore extended for one base. Again, the TC-loop intron is located within the B-box (fig. 2). An important consequence of the splicing reaction in the D-loop is shown in figures 2 and 3. The conserved GG at positions 18 and 19 in respect to the mature tRNA, necessary for correct three-dimensional interaction of the tRNA (Marck and Grosjean 2002), is missing in the gene structure as well as in the potential tRNA after the removal of the predicted intron. However, splicing the authentic intron creates this conserved twin G, thereby indicating that the tDNA PheGAA encodes a tRNA with a correct three-dimensional structure.
tRNAScan-SE detects two tDNAs specific for cysteine in the nucleomorph of G. theta. As mentioned above, one of these tDNAs is transcribed and spliced to a tRNA with specificity for glutamate. Therefore, the other tDNA, predicted as specific for cysteine (tRNA CysGCA) should provide this amino acid. The in silico data predict that the gene for this tRNA contains two introns, one in the TC- and the other in the D-loop (fig. 2). In our experiments, three introns were found to be spliced: one is located within the D-loop, one was predicted as the variable loop, and the third is again a 3-nt intron located within the B-box. Thus, our findings indicate a tDNA with three introns at positions otherwise not known from eukaryotes.
Whereas the initiatior tRNA iMetCAU is encoded by the nucleomorph without any in silico-detectable intron, elongator tRNA eMetCAU is predicted to harbor three introns. RT-PCR analyses confirmed that the introns located in the D- (shifted and extended for one base) and TC-loop, which is located within the B-box, are removed from the pre-tRNA (fig. 2). Moreover, we detected intermediates, from which only the D-loop intron is spliced, whereas the intron inserted in the TC-loop is still present. To our surprise no products were detected from which the third intron is removed. As this intron seems to be the most common one in this tDNA, we repeated the experiment with different primers. By analyzing more than 100 different transcripts none was detected in which the intron at the conserved position is removed. Therefore, no mature tRNA eMetCAU was verified. This result could reflect either very small amounts of the mature tRNA or technical problems, which did not occur in our analyses of all other nucleomorph-specific tDNA introns. Another possibility is that the host provide the elongator tRNA, thereby getting the control of the translation in the symbiont compartment.
The Enigmatic Introns in the B-box
Our analyses of the mature nucleomorph-specific tRNAs and their encoding genes clearly demonstrate that in seven cases introns are inserted within the B-box. It is known that cotranscription of genes is common for nucleomorphs (Gilson and McFadden 1996; Fraunholz, Moerschel, and Maier 1998). Having this in mind, a hypothesis to explain the presence of introns in the B-box is cotranscription of tRNA genes with upstream-located protein-encoding ones. This would change a RNA polymerase III transcript into one which is expressed by the RNA polymerase II. In order to prove this, we determined exemplarily the cotranscription of four tRNAs with upstream sequences. These were the genes for hlip (high light-inducible polypeptide) and the downstream-located, intron-less tRNA LeuUAG, the gene for the small nucleolar ribonucleoprotein D and tRNA ValUAC (one intron in the D-loop), T-complex protein 1, zeta subunit, and tRNA IleUAU (one intron in the B-box), and finally the gene for a ubiquitin-conjugating enzyme E2 and the downstream-located tRNA PheGAA (two introns). Cotranscription was detected in all examples. Nevertheless, this could speak either for the correctness of the hypothesis or for an extended transcription at the 3' end of protein-encoding mRNAs. As in some cases where the tRNAs overlap with their upstream-located protein-encoding gene (fig. 4), we analyzed the maturation of the primary transcripts. In the case of the cotranscript of hlip with tRNA LeuUAG, the upstream gene overlaps by two bases with tRNA LeuUAG. One would expect that the mature transcript of hlip and a complete tRNA could be generated by an endonucleolytic step and subsequent polyadenylation (fig. 4). An endonucleolytic cut 5' of the GA from the stop codon TGA of hlip would generate the correct 5' end of the tRNA, and subsequent polyadenylation of the 3' end of the hlip-transcript would then generate a TAA stop codon (fig. 4). Therefore, a hlip-transcript with the secondarily created stop codon (TAA) should be detectable. However, by analyzing 12 cDNAs specific for the hlip-sequence (from which 10 represent independent transcripts as demonstrated by different 3' nontranslated regions) we exclusively detected the unmodified cotranscriptional form. Thus, the "cotranscription hypothesis" is attractive, but most likely wrong. Hereupon we analyzed the intron sequences again and identified that the introns inserted into the B-box mimic the promoter region with high predicted values (fig. 2). Therefore, the encoded sequence of promoter regions is not part of the mature tRNAs, which is unique for eukaryotes.
FIG. 4.— Alternative transcription model of tRNAs in the nucleomorph of Guillardia theta. The hlip gene overlaps with tDNA for LeuUAG at the 3' end by two bases. After endonucleolytic processing the transcript, the disrupted stop codon of the hlip-transcript could be restored by polyadenylation. The two arrows labeled with Met indicate putative start codons of orf101 downstream of the tDNA for LeuUAG.
The Red Alga Dowry
The genetic system of eukaryotes is more related to archaea than to that of bacteria (Martin and Müller 1998; Martin et al. 2001). This is also seen in the presence of tDNA introns in archaea and eukaryotes, which are rare and of another type in bacteria. Due to projects on the nucleomorph of G. theta and the genome of the red alga Cyanidioschyzon merolae we noticed that contrary to earlier views, introns in tDNA can be similarly distributed in eukaryotes than in archaea. As red algae and the symbiont of the cryptophytes have a common ancestor (Van de Peer et al. 1996), similarities in respect to intron structure and localization can be expected. Therefore, we have reanalyzed the red algal data and confirm that some of the indicated C. merolae tDNA introns, which are generally longer than the nucleomorph G. theta ones, should be positioned at unique sites within tRNA genes, too (table 1). The predicted tDNA introns from C. merolae are inserted into the variable loop and the D- and TC-stems. Otherwise, no introns were predicted in the anticodon loop except the conserved intron localized 3' of the anticodon. There is also no intron prediction concerning the B-box region. A comparative analysis showed that none of the tDNAs with a unique intron localization encoded by the nucleomorph genome of G. theta has a counterpart in C. merolae (fig. 1 and table 1). Taken together, the nucleomorph-specific tDNA introns are shorter than the red algal ones, but occupied additional localizations within the tRNA genes.
Table 1 Comparison of Guillardia theta and Cyanidioschyzon merolae
Similar to the nucleomorphs of cryptophytes, red algae genomes are poor in splicesomal introns and in comparison to other eukaryotes enriched in tRNA introns. Therefore, the presence and unique distribution of nucleomorph-specific tRNA introns indicate a red alga dowry. Nevertheless, it was unexpected that the nucleomorph genome on the one hand streamlines and minimizes its genome, but on the other hand tolerates introns within nearly all positions of its tRNA genes. This may be explained either by the selfish nature of these introns or by structural-functional reasons. Structural-functional reasons of tRNA introns are discussed for Archaea. Here intervening sequences could play an essential role in tRNA maturation, especially in the modification of pre-tRNAs by intron-dependent tRNA modification enzymes (Motorin and Grosjean 1999). Moreover, introns may also be essential in the stepwise folding of the tRNAs by avoiding false pairing in tRNA maturation (Dennis, Omer, and Lowe 2001).
Conclusions
The complete nucleomorph sequence of the cryptomonad G. theta exhibited a streamlined eukaryotic genome with a gene density similar to bacterial genomes (Douglas et al. 2001). Such a reduction of noncoding sequences is manifested e.g., in splicesomal introns as shown by the detection of only 17 splicesomal introns in approximately 450 nucleomorph-specific genes or open reading frames (Douglas et al. 2001). One exception of genome streamlining is shown in this paper. As mentioned, 15 of 36 tDNAs plus one region which is annotated as a pseudogene encode tRNAs with one or more introns. Moreover, most of these introns are unique for Eukarya in respect to its localization and size (fig. 2). Therefore, the cryptophytic symbiont had successfully evolved additional mechanisms to detect and process these unusual tRNA introns. This capacity may be bequeathed from the red alga progenitor of the secondary symbiont of cryptophytes, as seen for tRNA introns detected in a recently sequenced red algae. Both the existence of introns and the evolution of uncommon targeting mechanisms for a splicing apparatus seem to be contradictory for a streamlined genome. Thus, the existence of introns in nucleomorph-coded tRNAs in the conserved position 3' of the anticodon as well as those at unique positions should have structural and/or functional reasons or may indicate selfish DNA. Anyway, biochemical and structural analyses on the unusual introns of the secondary endosymbiont will be of major importance and will show new mechanisms of RNA metabolism. Moreover, new genome projects on other protists will highlight if unusual intron insertions are common to other organisms. Such a situation could imply that earliest eukaryotes, like the red algae, once harbored introns outside the 3' conserved region of the anticodon, which were differentially lost in the course of evolution in most eukaryotes studied so far.
Supplementary Material
Supplementary tables 1 and 2 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
Acknowledgements
Our research is supported by the Deutsche Forschungsgemeinschaft (SFB-TR1).
References
Abelson, J., C. R. Trotta, and H. Li. 1998. tRNA splicing. J. Biol. Chem. 273:12685–12688.
Belfort, M., and A. Weiner. 1997. Another bridge between kingdoms: tRNA splicing in archaea and eukaryotes. Cell 89:1003–1006.
Biniszkiewicz, D., E. Cesnaviciene, and D. A. Shub. 1994. Self-splicing group I intron in cyanobacterial initiator methionine tRNA: evidence for lateral transfer of introns in bacteria. EMBO J. 13:4629–4635.
Cavalier-Smith, T. 2002. Nucleomorphs: enslaved algal nuclei. Curr. Opin. Microbiol. 5:612–619.
Delwiche, C. F., and J. D. Palmer. 1998. The origin of plastids and their spread via secondary symbiosis. Plant Sys. Evol. Suppl. 11:51–86.
Dennis, P. P., A. Omer, and T. Lowe. 2001. A guided tour: small RNA function in Archaea. Mol. Microbiol. 40:509–519.
Douglas, S., S. Zauner, M. Fraunholz, M. Beaton, S. Penny, L. T. Deng, X. Wu, M. Reith, T. Cavalier-Smith, and U. G. Maier. 2001. The highly reduced genome of an enslaved algal nucleus. Nature 410:1091–1096.
Esseiva, A. C., A. Naguleswaran, A. Hemphill, and A. Schneider. 2004. Mitochondrial tRNA import in Toxoplasma gondii. J. Biol. Chem. 279:42363–42368.
Fraunholz, M. J., E. Moerschel, and U. G. Maier. 1998. The chloroplast division protein FtsZ is encoded by a nucleomorph gene in cryptomonads. Mol. Gen. Genet. 260:207–211.
Gilson, P. R., and G. I. McFadden. 1996. The miniaturized nuclear genome of eukaryotic endosymbiont contains genes that overlap, genes that are cotranscribed, and the smallest known spliceosomal introns. Proc. Natl. Acad. Sci. USA 93:7737–7742.
Kuhsel, M. G., R. Strickland, and J. D. Palmer. 1990. An ancient group I intron shared by eubacteria and chloroplasts. Science 250:1570–1573.
Lowe, T. M., and S. R. Eddy. 1997. tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Res. 25:955–964.
Maier, U. G., S. E. Douglas, and T. Cavalier-Smith. 2000. The nucleomorph genomes of cryptophytes and chlorarachniophytes. Protist 151:103–109.
Marck, C., and H. Grosjean. 2002. tRNomics: analysis of tRNA genes from 50 genomes of Eukarya, Archaea, and Bacteria reveals anticodon-sparing strategies and domain-specific features. RNA 8:1189–1232.
Martin, W., M. Hoffmeister, C. Rotte, and K. Henze. 2001. An overview of endosymbiotic models for the origins of eukaryotes, their ATP-producing organelles (mitochondria and hydrogenosomes), and their heterotrophic lifestyle. Biol. Chem. 382:1521–1539.
Martin, W., and M. Müller. 1998. The hydrogen hypothesis for the first eukaryote. Nature 392:37–41.
Matsuzaki, M., O. Misumi, I. T. Shin et al. (42 co-authors). 2004. Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428:653–657.
McFadden, G. I. 1999. Plastids and protein targeting. J. Eukaryot. Microbiol. 46:339–346.
Motorin, Y., and H. Grosjean. 1999. Multisite-specific tRNA: m5C-methyltransferase (Trm4) in yeast Saccharomyces cerevisiae: identification of the gene and substrate specificity of the enzyme. RNA 5:1105–1118.
Reinhold-Hurek, B., and D. A. Shub. 1992. Self-splicing introns in tRNA genes of widely divergent bacteria. Nature 357:173–176.
Sprague, K. U. 1995. Transcription of eukaryotic tRNA genes. Pp. 31–50 in D. S?ll and U. RajBhandary, eds. Transfer RNA. ASM Press, Washington, D.C.
Stoebe, B., and U. G. Maier. 2002. One, two, three: nature's tool box for building plastids. Protoplasma 219:123–130.
Tan, T. H., R. Pach, A. Crausaz, A. Ivens, and A. Schneider. 2002. tRNAs in Trypanosoma brucei: genomic organization, expression, and mitochondrial import. Mol. Cell. Biol. 22:3707–3717.
Tomita, K., and A. M. Weiner. 2001. Collaboration between CC- and A-adding enzymes to build and repair the 3'-terminal CCA of tRNA in Aquifex aeolicus. Science 294:1334–1336.
Van de Peer, Y., S. A. Rensing, U. G. Maier, and R. De Wachter. 1996. Substitution rate calibration of small subunit ribosomal RNA identifies chlorarachniophyte endosymbionts as remnants of green algae. Proc. Natl. Acad. Sci. USA 93:7732–7736.
Xu, M.-Q., S. D. Kathe, H. Goodrich-Blair, S. A. Nierzwicki-Bauer, and D. A. Shub. 1990. Bacterial origin of a chloroplast intron: conserved self-splicing group I introns in cyanobacteria. Science 250:1566–1570.
Zauner, S., M. Fraunholz, J. Wastl, S. Penny, M. Beaton, T. Cavalier-Smith, U. G. Maier, and S. Douglas. 2000. Chloroplast protein and centrosomal genes, a tRNA intron, and odd telomeres in an unusually compact eukaryotic genome, the cryptomonad nucleomorph. Proc. Natl. Acad. Sci. USA 97:200–205.(Oliver Kawach, Christine )
E-mail: maier@staff.uni-marburg.de.
Abstract
Nucleomorphs are remnant nuclei of eukaryotic, secondary endosymbionts exclusively found in cryptophytes and chlorarachniophytes. The nucleomorph of the cryptophyte Guillardia theta codes for 36 transfer RNA (tRNA) genes, 15 of them predicted to contain introns and 1 pseudo-tRNA. Some of the predicted intervening sequences are manifested at positions not known in Eukarya, even tRNAs with more than one intron were suggested. By isolating reverse transcriptase–polymerase chain reaction products of the spliced tRNAs we verify the processing of all predicted intron-harboring tRNAs and demonstrate the splicing of the smallest introns (3 nt) investigated so far. However, the spliced intervening sequences are in some cases shifted in respect to the predicted ones. Moreover, we show that introns, if inserted into the B-box of tRNA genes in the nucleomorph of cryptophytes, mimic promoter regions and do not abolish transcription by RNA polymerase III. Consequently, internal nucleomorph-encoded tRNA promoter regions are in some cases dissected from the sequence of the mature tRNAs. By reanalyzing tRNA introns of a recently sequenced red algae we furthermore show that splicing of introns at unusual positions may be introduced in cryptophytes by its secondary endosymbiont. However, in contrast to the rest of the symbiont genome, introns are not minimized in quantity but are instead scattered along the tRNA genes.
Key Words: Guillardia theta ? Cyanidioschyzon merolae ? tRNA ? nucleomorph ? secondary endosymbiosis
Introduction
It is generally accepted that the genetic system of eukaryotes is related to that of Archaea and seems to be different from the transcription-translation machinery of Bacteria (Martin and Müller 1998; Martin et al. 2001). Such a difference can be observed in the splicing mechanisms of transfer RNA (tRNA) introns as well. These types of introns are common in all three domains, Bacteria, Archaea, and Eukarya (Marck and Grosjean 2002), but Bacteria evolved self-splicing tRNA introns, whereas tRNA introns from Archaea and Eukarya were processed in an enzymatic catalyzed reaction (Abelson, Trotta, and Li 1998). tRNA introns, if present, could be located at different positions in Archaea (Belfort and Weiner 1997) and are restricted to a conserved position between the first and second base 3' of the anticodon in Eukarya (Marck and Grosjean 2002) (fig. 1). Moreover, as recently reported, red algae may encode tRNAs with introns, which are predicted at positions outside the conserved position near the anticodon (Matsuzaki et al. 2004) (fig. 1). In Bacteria self-splicing group I introns are described, which are located 3' of the anticodon or within the anticodon (Kuhsel, Strickland, and Palmer 1990; Xu et al. 1990; Reinhold-Hurek and Shub 1992; Biniszkiewicz, Cesnaviciene, and Shub 1994) (fig. 1).
FIG. 1.— Comparison of intron positions in the nucleomorph, other eukaryotes, Bacteria, and Archaea. Arrows show intron positions. The thick arrow in the scheme of nucleomorph tRNA indicates the region from which a "variable loop" is removed by splicing. Data for the intron positions of Archaea-specific introns are from Marck and Grosjean (2002); for eubacterial group I introns see text.
Several algae evolved by the enslavement of a phototrophic eukaryotic cell by another, except heterotrophic eukaryote (secondary endosymbiosis) (Delwiche and Palmer 1998; McFadden 1999; Maier, Douglas, and Cavalier-Smith 2000). During the evolution of a "cell within a cell," the symbiont organelles were reduced or eliminated. As in some secondary evolved algae the reduction led to a plastid surrounded by additional membranes, and an intermediate stage in the reduction of the symbiont organelles is found in cryptophytes and chlorarachniophytes (Cavalier-Smith 2002; Stoebe and Maier 2002). Here a remnant cytoplasm of the eukaryotic symbiont including its pigmy nucleus, the nucleomorph, is maintained. The nucleomorph genome of one cryptomonad, Guillardia theta, was sequenced, thereby lightning up a reduced eukaryotic genome organized in three small chromosomes (Zauner et al. 2000; Douglas et al. 2001). This miniaturized genome is maintained to encode genes whose products have important functions in the plastid. To synthesize these functions, a transcription and translation apparatus has to be expressed in the secondary symbiont cytoplasm.
By scanning the nucleomorph genomic sequence for tRNA-encoding genes, we showed that 35 transfer DNAs (tDNAs) should be encoded by the nucleomorph (Douglas et al. 2001). However, the nucleomorph should not be autonomous in respect to tRNAs due to a missing tRNA for glutamate (Douglas et al. 2001). In addition, we described one tRNA intron, located at a position otherwise not known from Eukarya (Zauner et al. 2000).
In the paper at hand, we show that splicing of tRNA introns, which are unique for eukaryotic organisms in respect to its localization, lead to bona fide tRNAs. In this respect, splicing of tRNA introns, located 3' or 5' of the anticodon, in the variable loop, TC-loop, or D-loop, is demonstrated. Moreover, we demonstrate splicing of the smallest eukaryotic introns (3 nt) and show that insertion in the internal promoter regions of tRNA genes do not abolish the transcription and splicing of tRNAs. Our data indicate that the ability to splice tRNA introns located at unique positions within the nucleomorph genome was most likely the legacy of the red algae progenitor. However, in the nucleomorph genome tRNA introns were shortened, but in contrast to other noncoding chromosomal elements, spread over most possible tRNA gene localizations, and contradict therefore the general tendency of streamlining the nucleomorph genome.
Materials and Methods
Reverse Transcriptase–Polymerase Chain Reaction and Sequencing
RNA from G. theta was isolated and reverse transcriptase–polymerase chain (RT-PCR) reactions were done according to manufacturer's instructions (MBI Fermentas, St. Leon-Rot, Germany). Primer sequences are given in supplementary table 1 (Supplementary Material online). RT-PCR products were isolated from 17% denaturing polyacrylamid gels and cloned into pGEM-T? (Promega, Mannheim, Germany). Individual clones were sequenced on a LI-COR 4200 sequencer using labeled M13-20 standard primers. In some cases, we detected additional splicing variants harboring insertions-deletions as well as exchanges of bases ("editing"). In the case in which the variants were isolated only one time or show no secondary structure specific for tRNAs, they were not included in our analyses. Instead, they were considered as PCR artifacts or as improper transcripts, which were detected in nucleomorph-specific messenger RNAs (mRNAs) in another study as well (Fraunholz, Moerschel, and Maier 1998).
tDNA Detection and Intron Prediction
tDNA detection and intron prediction were done with tRNAscan-SE provided at http://www.genetics.wustl.edu/eddy/tRNAscan-SE/ using relaxed parameters (cove only). tDNA promotor prediction was performed with POLIIISCAN provided at http://wwwmgs.bionet.nsc.ru/mgs/programs/poliiiscan/.
Characterization of Cotranscription
Cotranscription of tRNA genes with upstream-located genes was identified in the case of hlip-tRNA LeuUAG and uce-E2-tRNA PheGAA in expressed sequence tag clones. Two further cotranscribed tRNAs were isolated by RT-PCR experiments. For these, specific primers for the upstream-located gene (snrpD and tcpZ1) and for the tRNA genes (tRNA ValUAC and tRNA IleUAU) were used for RT-PCR experiments, respectively.
To isolate complementary DNAs (cDNAs) covering hlip (high light-induced protein), our cDNA library was screened with a gene-specific probe. Positive clones were sequenced as described above.
Results
The Nucleomorph Is Expected to Encode tRNAs with Unique Introns
As predicted by tRNAscan-SE (Lowe and Eddy 1997), 36 tRNAs and 1 related pseudogene should be encoded in the nucleomorph genome of G. theta. This lead to a set of tRNAs offering all amino acids with the exception of a tRNA glutamate (Glu, E), which seems to be missing in the nucleomorph genome of G. theta (Douglas et al. 2001). The predicted tRNAs show proper anticodons flanked 5' by a T and 3' by a purine, as it is common in eukaryotes except for initiator tDNA iMetCAU in higher eukaryotes (Marck and Grosjean 2002).
As in other eukaryotes, the tRNA genes (tDNAs) are dispersed in the nucleomorph genome and not clustered as it is found predominately in bacteria (Marck and Grosjean 2002). Furthermore, 3'-CAA is not part of the tDNA (Tomita and Weiner 2001), and A- and B-boxes, representing the internal promoter, can be identified in the predicted tDNAs (Spargue 1995) (fig. 2).
FIG. 2.— Sequence of tRNA genes and processed tRNAs. The sequence of tRNA genes (tDNAs) and the processed tRNAs are shown. The first line in the columns indicates the tDNA. Green positions show the B-box, blue the A-box, and red the anticodon. The predicted introns are shown in bold letters. Value of the A- and B-box refers to the F value predicted by POLIIISCAN. (i) = intermediate form; (i2) = second intermediate form. tRNAs are indicted in the three-letter code. The red colored A refers to the appearance of a tRNA GluUUC transcript with an additional A at position 53. (See also supplementary table 2, Supplementary Material online.)
Further on tRNAscan-SE (Lowe and Eddy 1997) predict introns at the highly conserved position between the first two bases 3' of the anticodon in eukaryotes. In addition the prediction program indicates introns at unusual positions (fig. 2). Out of the 36 nucleomorph-encoded tRNAs, 15 were predicted to encode at least one intervening sequence. These potential nonsplicesomal introns vary in length from only 3 up to 24 nt (fig. 2). Some of them would be unique for the domain Eukarya in respect to its localization and should emphasize the unique nature of some nucleomorph-encoded tRNA genes: seven of the tDNA introns are predicted to be located in the conserved position between the first two bases 3' of the anticodon, seven in the D-loop, and eight in the TC-loop (fig. 1). Interestingly, six tDNAs should encode pre-tRNAs with more than one intron. Five of them are predicted to harbor one intron in the D- and another one in TC-loop, which would represent a unique arrangement of intron insertions into a eukaryotic tDNA (fig. 1). Surprisingly, some of the introns are inserted into the B-box of tRNA genes, which would interrupt the promotor structure and therefore suppress transcription of these genes by RNA polymerase III (fig. 2).
Furthermore, nucleomorph chromosomes are built up of terminal repeats bordering a single-copy region (Zauner et al. 2000). Housekeeping genes located within the single-copy regions show an average A/T content of 77%. The A/T content of the tDNAs is remarkably lower (55%), reflecting the structural constrains of the two- and three-dimensional structures of the tRNAs. On the other hand A/T content is increasing in the predicted tRNA introns significantly (81% A/T), which shows the preference of A/T in noncoding regions.
tRNAs of the Symbiont of Cryptophytes are Encoded in the Nucleomorph
As the extremely condensed nucleomorph genome of the cryptophyte G. theta has only a limited coding capacity (Douglas et al. 2001), it is possible that tRNAs are imported from the host compartments to supplement missing functions as it is known from other organisms (Tan et al. 2002; Esseiva et al. 2004). In order to prove if tRNAs of the secondary endosymbiont originate exclusively from nucleomorph genes or from a second gene located in the cell nucleus, we amplified exemplarily four tRNA genes (tDNAs for AlaUGC, CysACA, LeuUAG, and PheGAA) from genomic DNA of G. theta. As the tDNAs for AlaUGC are predicted to encode one intron and the genes for tRNA CysACA and PheGAA even two (see later), primers were designed such that neither a tDNA with nor a possible nuclear-encoded tDNA without the predicted intron(s) is favored in PCR reactions. These experiments generated solely nucleomorph-specific amplificates thereby demonstrating that only nucleomorph-specific tDNA serves as matrix for spliced tRNAs. As, therefore, no indications for a nucleus-encoded symbiont-specific tRNA exist, our experiments exclude an import of nucleus-encoded tRNAs into the symbiont in the analyzed cases.
Introns of the Anticodon Loop are Spliced by the Secondary Endosymbiont
In order to prove whether the eukaryotic symbiont of cryptophytes is able to excise the predicted introns or the intron-harboring tDNAs are pseudogenes, we performed RT-PCRs. As control, tRNA ArgUCG and tRNA GlnCUG, whose tDNAs show no evidence for introns, were amplified from G. theta total RNA. As expected, both tRNAs are collinear with the genomic sequence (fig. 2). Next, we amplified the nucleomorph tRNAs ArgCCU, ArgUCU, TrpCCA, and TyrGUA (fig. 2). The tDNAs possess a predicted intron at the conserved position between the first two bases 3' of the anticodon. The RT-products confirmed that introns are encoded in the tDNAs and that they are removed in the mature tRNAs. The predicted localization of the introns concurs with the obtained RT-products in the tRNAs ArgUCU and TrpCCA. Surprisingly, the introns of ArgCCU and TyrGUA are shifted downstream in respect to its predictions. In the case of ArgCCU the intron is also extended for four bases (fig. 2). Therefore, introns in the tDNAs for ArgCCU and TyrGUA are inserted at positions otherwise not known from Eukarya. However, the mature tRNAs show bona fide anticodons as well as the conserved sides for the three-dimensional structure (fig. 3). Interestingly, the anticodon stem in tRNAs ArgCCU and TyrGUA shows one mismatch. At the moment, we cannot decide whether this secondary structure leads to a nucleomorph-specific functional tRNA or is a characteristic of nonfunctional tRNA.
FIG. 3.— Scheme of conserved tRNA bases found in the processed nucleomorph-specific tRNAs. Clamps indicate the predicted internal base pairing in the tertiary structure. Lines 1–3: conventionally used numbering system for tRNA positions. Line 4: secondary structure of the tRNA cloverleaf model. Solid boxes indicate the anticodon (AC). V: variable loop positions 44–48 (X: UGAGGUAAUCA; X1: GGUUACAACC; X2: GGGUACUACCU). The putative intron in MetCAU not found to be spliced is indicated as A*(UUGUACGCG).
Pseudogenes, the Missing tRNA for Glutamate and Introns 5' of the Anticodon
In nucleomorph chromosome III one stretch of bases is predicted to encode a pseudogene with some similarities to a potential tDNA (fig. 2). No anticodon was predicted precisely in this genomic region, but we noticed that conserved bases necessary for three-dimensional folding as well as possible A- and B-boxes for transcription initiation are present. Characterization of the maturated transcript of this region showed a bona fide tRNA LeuUAA (fig. 2), which is created by removing a 10-nt-long intron 5' of the anticodon and a 3-nt one in the TC-loop. The latter is one of the smallest introns identified so far and, interestingly, integrated within the B-box (see later).
As mentioned in Douglas et al. (2001), we noticed that the nucleomorph of G. theta should encode at least one tRNA for the whole set of amino acids with the exception of glutamate (Glu, E), for which no nucleomorph-located tDNA could be identified. On the other hand, we detected two tDNAs encoding a tRNA for cysteine. One of them, CysACA, is anticipated to harbor an 18-nt insertion at the conserved intron position 3' of the anticodon (fig. 2). However, the RT-PCR product of the predicted tRNA CysACA differs remarkably in respect to the intron positioning and anticodon prediction. As indicated in figure 2, splicing of a 10-nt-long intron containing the predicted anticodon occurred. This creates a potential tRNA with a "new" anticodon prediction for GluUUC harboring an additional intron located in the conserved position 3' of the UUC anticodon. This second intron is spliced as well (fig. 2). Thus, the missing tRNA providing glutamate is hidden in a tDNA predicted to encode tRNA CysACA.
The tDNA GluUUC is not only exceptional by harboring two introns. By studying the genomic version as well as the different intermediate forms, we detected that after splicing the 5' intron three different forms exist. In one, the intron 3' of the anticodon is spliced, in the second an additional a G to A transversion was detected, and in the third form an A is inserted (supplementary table 2, Supplementary Material online). However, in the tRNA from which both introns are spliced, only the additional A is present (fig. 1). As the different forms were isolated in independent experiments several times, a complicated maturation of this tDNA GluUUC may be indicated by the different intermediates.
Single Introns Located Within the TC- or the D-Loop
Previously, we have shown that the symbiont-specific tRNA SerAGA, whose gene is encoded in the nucleomorph of G. theta, contains two introns (Zauner et al. 2000). One is located 3' of the anticodon and the other one in the D-loop; the latter one was the first description of an intron in this loop in a eukaryotic encoded tDNA. An extended in silico search for further unusual tRNA introns suggested five nucleomorph-specific intron positions otherwise not known from eukaryotes: three in the TC-loop (tDNA AlaUGC, IleUAU, and LeuCAA) and two in the D-loop (tDNA AsnGUU and ValAAC) (fig. 2). We amplified the spliced molecules of these tDNAs. The tRNAs AlaUGC, IleUAU, and LeuCAA were processed as predicted (fig. 2). Interestingly, as in tRNA LeuUAA, a 3-nt intron is correctly spliced from the primary transcript of tRNA AlaUGC. In all three genes, although the predicted introns are located within the B-box, these tDNAs are transcribed. In the case of the D-loop–located introns of tRNAs AsnGUU and ValAAC the intron prediction differs from in vivo situation. Whereas in the tRNA AsnGUU the intron is shifted for one base, the intron of tRNA ValAAC is shifted and extended by 1 nt. Moreover, we detected in the tDNA ValAAC an additional intron, which is again 3-nt-long and inserted into the B-box region (fig. 2).
Splicing of tRNAs with Multiple Predicted Introns
As shown above, the nucleomorph harbors tDNAs including introns that are unique for the domain Eukarya in respect to its localization. tRNAScan-SE predicts further nucleomorph-specific tDNAs, which may harbor more than one intron. This is true for two tDNAs (PheGAA and CysGCA) which are predicted to encode two introns, one in the D- and one in the TC-loop, as well as for elongator tDNA eMetCAU with a prediction of three introns, one in the conserved position 3' of the anticodon, one in the D-, and one in the TC-loop.
In the case of tDNA PheGAA our experiments show that both introns are processed. Interestingly, both introns were discovered to be shifted two bases to the 5'; the TC-loop intron is furthermore extended for one base. Again, the TC-loop intron is located within the B-box (fig. 2). An important consequence of the splicing reaction in the D-loop is shown in figures 2 and 3. The conserved GG at positions 18 and 19 in respect to the mature tRNA, necessary for correct three-dimensional interaction of the tRNA (Marck and Grosjean 2002), is missing in the gene structure as well as in the potential tRNA after the removal of the predicted intron. However, splicing the authentic intron creates this conserved twin G, thereby indicating that the tDNA PheGAA encodes a tRNA with a correct three-dimensional structure.
tRNAScan-SE detects two tDNAs specific for cysteine in the nucleomorph of G. theta. As mentioned above, one of these tDNAs is transcribed and spliced to a tRNA with specificity for glutamate. Therefore, the other tDNA, predicted as specific for cysteine (tRNA CysGCA) should provide this amino acid. The in silico data predict that the gene for this tRNA contains two introns, one in the TC- and the other in the D-loop (fig. 2). In our experiments, three introns were found to be spliced: one is located within the D-loop, one was predicted as the variable loop, and the third is again a 3-nt intron located within the B-box. Thus, our findings indicate a tDNA with three introns at positions otherwise not known from eukaryotes.
Whereas the initiatior tRNA iMetCAU is encoded by the nucleomorph without any in silico-detectable intron, elongator tRNA eMetCAU is predicted to harbor three introns. RT-PCR analyses confirmed that the introns located in the D- (shifted and extended for one base) and TC-loop, which is located within the B-box, are removed from the pre-tRNA (fig. 2). Moreover, we detected intermediates, from which only the D-loop intron is spliced, whereas the intron inserted in the TC-loop is still present. To our surprise no products were detected from which the third intron is removed. As this intron seems to be the most common one in this tDNA, we repeated the experiment with different primers. By analyzing more than 100 different transcripts none was detected in which the intron at the conserved position is removed. Therefore, no mature tRNA eMetCAU was verified. This result could reflect either very small amounts of the mature tRNA or technical problems, which did not occur in our analyses of all other nucleomorph-specific tDNA introns. Another possibility is that the host provide the elongator tRNA, thereby getting the control of the translation in the symbiont compartment.
The Enigmatic Introns in the B-box
Our analyses of the mature nucleomorph-specific tRNAs and their encoding genes clearly demonstrate that in seven cases introns are inserted within the B-box. It is known that cotranscription of genes is common for nucleomorphs (Gilson and McFadden 1996; Fraunholz, Moerschel, and Maier 1998). Having this in mind, a hypothesis to explain the presence of introns in the B-box is cotranscription of tRNA genes with upstream-located protein-encoding ones. This would change a RNA polymerase III transcript into one which is expressed by the RNA polymerase II. In order to prove this, we determined exemplarily the cotranscription of four tRNAs with upstream sequences. These were the genes for hlip (high light-inducible polypeptide) and the downstream-located, intron-less tRNA LeuUAG, the gene for the small nucleolar ribonucleoprotein D and tRNA ValUAC (one intron in the D-loop), T-complex protein 1, zeta subunit, and tRNA IleUAU (one intron in the B-box), and finally the gene for a ubiquitin-conjugating enzyme E2 and the downstream-located tRNA PheGAA (two introns). Cotranscription was detected in all examples. Nevertheless, this could speak either for the correctness of the hypothesis or for an extended transcription at the 3' end of protein-encoding mRNAs. As in some cases where the tRNAs overlap with their upstream-located protein-encoding gene (fig. 4), we analyzed the maturation of the primary transcripts. In the case of the cotranscript of hlip with tRNA LeuUAG, the upstream gene overlaps by two bases with tRNA LeuUAG. One would expect that the mature transcript of hlip and a complete tRNA could be generated by an endonucleolytic step and subsequent polyadenylation (fig. 4). An endonucleolytic cut 5' of the GA from the stop codon TGA of hlip would generate the correct 5' end of the tRNA, and subsequent polyadenylation of the 3' end of the hlip-transcript would then generate a TAA stop codon (fig. 4). Therefore, a hlip-transcript with the secondarily created stop codon (TAA) should be detectable. However, by analyzing 12 cDNAs specific for the hlip-sequence (from which 10 represent independent transcripts as demonstrated by different 3' nontranslated regions) we exclusively detected the unmodified cotranscriptional form. Thus, the "cotranscription hypothesis" is attractive, but most likely wrong. Hereupon we analyzed the intron sequences again and identified that the introns inserted into the B-box mimic the promoter region with high predicted values (fig. 2). Therefore, the encoded sequence of promoter regions is not part of the mature tRNAs, which is unique for eukaryotes.
FIG. 4.— Alternative transcription model of tRNAs in the nucleomorph of Guillardia theta. The hlip gene overlaps with tDNA for LeuUAG at the 3' end by two bases. After endonucleolytic processing the transcript, the disrupted stop codon of the hlip-transcript could be restored by polyadenylation. The two arrows labeled with Met indicate putative start codons of orf101 downstream of the tDNA for LeuUAG.
The Red Alga Dowry
The genetic system of eukaryotes is more related to archaea than to that of bacteria (Martin and Müller 1998; Martin et al. 2001). This is also seen in the presence of tDNA introns in archaea and eukaryotes, which are rare and of another type in bacteria. Due to projects on the nucleomorph of G. theta and the genome of the red alga Cyanidioschyzon merolae we noticed that contrary to earlier views, introns in tDNA can be similarly distributed in eukaryotes than in archaea. As red algae and the symbiont of the cryptophytes have a common ancestor (Van de Peer et al. 1996), similarities in respect to intron structure and localization can be expected. Therefore, we have reanalyzed the red algal data and confirm that some of the indicated C. merolae tDNA introns, which are generally longer than the nucleomorph G. theta ones, should be positioned at unique sites within tRNA genes, too (table 1). The predicted tDNA introns from C. merolae are inserted into the variable loop and the D- and TC-stems. Otherwise, no introns were predicted in the anticodon loop except the conserved intron localized 3' of the anticodon. There is also no intron prediction concerning the B-box region. A comparative analysis showed that none of the tDNAs with a unique intron localization encoded by the nucleomorph genome of G. theta has a counterpart in C. merolae (fig. 1 and table 1). Taken together, the nucleomorph-specific tDNA introns are shorter than the red algal ones, but occupied additional localizations within the tRNA genes.
Table 1 Comparison of Guillardia theta and Cyanidioschyzon merolae
Similar to the nucleomorphs of cryptophytes, red algae genomes are poor in splicesomal introns and in comparison to other eukaryotes enriched in tRNA introns. Therefore, the presence and unique distribution of nucleomorph-specific tRNA introns indicate a red alga dowry. Nevertheless, it was unexpected that the nucleomorph genome on the one hand streamlines and minimizes its genome, but on the other hand tolerates introns within nearly all positions of its tRNA genes. This may be explained either by the selfish nature of these introns or by structural-functional reasons. Structural-functional reasons of tRNA introns are discussed for Archaea. Here intervening sequences could play an essential role in tRNA maturation, especially in the modification of pre-tRNAs by intron-dependent tRNA modification enzymes (Motorin and Grosjean 1999). Moreover, introns may also be essential in the stepwise folding of the tRNAs by avoiding false pairing in tRNA maturation (Dennis, Omer, and Lowe 2001).
Conclusions
The complete nucleomorph sequence of the cryptomonad G. theta exhibited a streamlined eukaryotic genome with a gene density similar to bacterial genomes (Douglas et al. 2001). Such a reduction of noncoding sequences is manifested e.g., in splicesomal introns as shown by the detection of only 17 splicesomal introns in approximately 450 nucleomorph-specific genes or open reading frames (Douglas et al. 2001). One exception of genome streamlining is shown in this paper. As mentioned, 15 of 36 tDNAs plus one region which is annotated as a pseudogene encode tRNAs with one or more introns. Moreover, most of these introns are unique for Eukarya in respect to its localization and size (fig. 2). Therefore, the cryptophytic symbiont had successfully evolved additional mechanisms to detect and process these unusual tRNA introns. This capacity may be bequeathed from the red alga progenitor of the secondary symbiont of cryptophytes, as seen for tRNA introns detected in a recently sequenced red algae. Both the existence of introns and the evolution of uncommon targeting mechanisms for a splicing apparatus seem to be contradictory for a streamlined genome. Thus, the existence of introns in nucleomorph-coded tRNAs in the conserved position 3' of the anticodon as well as those at unique positions should have structural and/or functional reasons or may indicate selfish DNA. Anyway, biochemical and structural analyses on the unusual introns of the secondary endosymbiont will be of major importance and will show new mechanisms of RNA metabolism. Moreover, new genome projects on other protists will highlight if unusual intron insertions are common to other organisms. Such a situation could imply that earliest eukaryotes, like the red algae, once harbored introns outside the 3' conserved region of the anticodon, which were differentially lost in the course of evolution in most eukaryotes studied so far.
Supplementary Material
Supplementary tables 1 and 2 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
Acknowledgements
Our research is supported by the Deutsche Forschungsgemeinschaft (SFB-TR1).
References
Abelson, J., C. R. Trotta, and H. Li. 1998. tRNA splicing. J. Biol. Chem. 273:12685–12688.
Belfort, M., and A. Weiner. 1997. Another bridge between kingdoms: tRNA splicing in archaea and eukaryotes. Cell 89:1003–1006.
Biniszkiewicz, D., E. Cesnaviciene, and D. A. Shub. 1994. Self-splicing group I intron in cyanobacterial initiator methionine tRNA: evidence for lateral transfer of introns in bacteria. EMBO J. 13:4629–4635.
Cavalier-Smith, T. 2002. Nucleomorphs: enslaved algal nuclei. Curr. Opin. Microbiol. 5:612–619.
Delwiche, C. F., and J. D. Palmer. 1998. The origin of plastids and their spread via secondary symbiosis. Plant Sys. Evol. Suppl. 11:51–86.
Dennis, P. P., A. Omer, and T. Lowe. 2001. A guided tour: small RNA function in Archaea. Mol. Microbiol. 40:509–519.
Douglas, S., S. Zauner, M. Fraunholz, M. Beaton, S. Penny, L. T. Deng, X. Wu, M. Reith, T. Cavalier-Smith, and U. G. Maier. 2001. The highly reduced genome of an enslaved algal nucleus. Nature 410:1091–1096.
Esseiva, A. C., A. Naguleswaran, A. Hemphill, and A. Schneider. 2004. Mitochondrial tRNA import in Toxoplasma gondii. J. Biol. Chem. 279:42363–42368.
Fraunholz, M. J., E. Moerschel, and U. G. Maier. 1998. The chloroplast division protein FtsZ is encoded by a nucleomorph gene in cryptomonads. Mol. Gen. Genet. 260:207–211.
Gilson, P. R., and G. I. McFadden. 1996. The miniaturized nuclear genome of eukaryotic endosymbiont contains genes that overlap, genes that are cotranscribed, and the smallest known spliceosomal introns. Proc. Natl. Acad. Sci. USA 93:7737–7742.
Kuhsel, M. G., R. Strickland, and J. D. Palmer. 1990. An ancient group I intron shared by eubacteria and chloroplasts. Science 250:1570–1573.
Lowe, T. M., and S. R. Eddy. 1997. tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Res. 25:955–964.
Maier, U. G., S. E. Douglas, and T. Cavalier-Smith. 2000. The nucleomorph genomes of cryptophytes and chlorarachniophytes. Protist 151:103–109.
Marck, C., and H. Grosjean. 2002. tRNomics: analysis of tRNA genes from 50 genomes of Eukarya, Archaea, and Bacteria reveals anticodon-sparing strategies and domain-specific features. RNA 8:1189–1232.
Martin, W., M. Hoffmeister, C. Rotte, and K. Henze. 2001. An overview of endosymbiotic models for the origins of eukaryotes, their ATP-producing organelles (mitochondria and hydrogenosomes), and their heterotrophic lifestyle. Biol. Chem. 382:1521–1539.
Martin, W., and M. Müller. 1998. The hydrogen hypothesis for the first eukaryote. Nature 392:37–41.
Matsuzaki, M., O. Misumi, I. T. Shin et al. (42 co-authors). 2004. Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428:653–657.
McFadden, G. I. 1999. Plastids and protein targeting. J. Eukaryot. Microbiol. 46:339–346.
Motorin, Y., and H. Grosjean. 1999. Multisite-specific tRNA: m5C-methyltransferase (Trm4) in yeast Saccharomyces cerevisiae: identification of the gene and substrate specificity of the enzyme. RNA 5:1105–1118.
Reinhold-Hurek, B., and D. A. Shub. 1992. Self-splicing introns in tRNA genes of widely divergent bacteria. Nature 357:173–176.
Sprague, K. U. 1995. Transcription of eukaryotic tRNA genes. Pp. 31–50 in D. S?ll and U. RajBhandary, eds. Transfer RNA. ASM Press, Washington, D.C.
Stoebe, B., and U. G. Maier. 2002. One, two, three: nature's tool box for building plastids. Protoplasma 219:123–130.
Tan, T. H., R. Pach, A. Crausaz, A. Ivens, and A. Schneider. 2002. tRNAs in Trypanosoma brucei: genomic organization, expression, and mitochondrial import. Mol. Cell. Biol. 22:3707–3717.
Tomita, K., and A. M. Weiner. 2001. Collaboration between CC- and A-adding enzymes to build and repair the 3'-terminal CCA of tRNA in Aquifex aeolicus. Science 294:1334–1336.
Van de Peer, Y., S. A. Rensing, U. G. Maier, and R. De Wachter. 1996. Substitution rate calibration of small subunit ribosomal RNA identifies chlorarachniophyte endosymbionts as remnants of green algae. Proc. Natl. Acad. Sci. USA 93:7732–7736.
Xu, M.-Q., S. D. Kathe, H. Goodrich-Blair, S. A. Nierzwicki-Bauer, and D. A. Shub. 1990. Bacterial origin of a chloroplast intron: conserved self-splicing group I introns in cyanobacteria. Science 250:1566–1570.
Zauner, S., M. Fraunholz, J. Wastl, S. Penny, M. Beaton, T. Cavalier-Smith, U. G. Maier, and S. Douglas. 2000. Chloroplast protein and centrosomal genes, a tRNA intron, and odd telomeres in an unusually compact eukaryotic genome, the cryptomonad nucleomorph. Proc. Natl. Acad. Sci. USA 97:200–205.(Oliver Kawach, Christine )