INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Group I introns are a class of RNA elements that share a secondary structure which allows the intron to undergo self-splicing from the primary transcript (5). While most group I introns are located in the genes of mitochondria and chloroplasts of lower eucaryotes, some are found in nuclear genes. Interestingly, nuclear group I introns reside only in rDNA, the gene encoding rRNA, and when present they occupy all of the ca. 200 rDNA copies typical of eucaryotic organisms. Some group I introns are mobile genetic elements. They encode a site-specific endonuclease that recognizes and cleaves a DNA sequence at or near the intron insertion site of the intron-lacking allele. The double-strand break is then repaired by replication of the intron into the intron-lacking allele, thus converting all intron-lacking alleles into intron-containing alleles. This process is termed intron homing due to its high specificity (3).

Among the ca. 150 nuclear group I introns reported so far, only three have been shown or have been inferred to be mobile: DiSSU1 from the slime mold Didymium iridis (6, 21, 22), NaSSU1 from the protist Naegleria andersoni and other Naegleria species (8), and PpLSU3 from the slime mold Physarum polycephalum. Originally found in the large-subunit rDNA gene of the Carolina strain (29), PpLSU3 contains the open reading frame (ORF) for the homing endonuclease I-PpoI (for nomenclature of intron-encoded endonucleases, see reference 7) in its 5' half and the ribozyme part in its 3' half. The sequence of the ribozyme part of PpLSU3 is 70% identical to the Tetrahymena thermophila intron TtLSU1, which is inserted at the same location as PpLSU3, suggesting a common evolutionary origin. Previous work has shown that PpLSU3 RNA not only undergoes self-splicing but also cleaves itself at an internal processing site (IPS), thus separating the I-PpoI ORF and the ribozyme (38). I-PpoI recognizes a 13- to 15-bp DNA sequence in a portion of the large-subunit rRNA gene that is 100% identical in all eucaryotes (9, 52). When a plasmid-borne PpLSU3 is transformed into Saccharomyces cerevisiae, most cells die upon the induction of I-PpoI expression, because I-PpoI makes double-strand breaks in the ca. 120 rDNA repeats on chromosome XII. Of the cells that survive, most do so because of disruption of the cleavage site in all rDNA copies, due either to acquisition of point mutations at the cleavage site or to homing of PpLSU3 into this site (30). Homing of PpLSU3 into yeast rDNA provides a unique experimental system with which to study PpLSU3 and I-PpoI expression in vivo.

In all known eucaryotes, protein-encoding genes are known to be transcribed only by RNA polymerase II (pol II). The repeated genes for three of the four rRNAs are transcribed by pol I, giving rise to a pre-rRNA that is further processed to yield mature 28S, 5.8S, and 18S rRNAs (reviewed in reference 33). A few attempts have been made to express protein under the pol I promoter control on a plasmid. In some cases, a cryptic pol II promoter nearby on the plasmid was utilized to make the mRNA, thus complicating the interpretation of the data (25). In cases where protein synthesis was attributed to the pol I-made transcript, the expression level was much lower than that from a pol II transcript (11, 15, 44). Low expression is probably due to the fact that the pol I transcript lacks a 5' cap and a 3' poly(A) tail, which are important for export, stabilization, and efficient translation of mRNAs (reviewed in reference 18). Recently, Lo et al. (24) reported that the HIS4 RNA synthesized from the pol I promoter is uncapped but does have a poly(A) tail at its 3' end. The pol I-made HIS4 RNA is both unstable and translated inefficiently, giving rise to 3% as much HIS4 protein as the HIS4 mRNA transcribed by pol II from the wild-type gene (24). Pol I is not absolutely required for the expression of rRNA. Nogi et al. have demonstrated that transcription of the 35S rDNA from the GAL7 promoter on a plasmid can provide the sole source of rRNA in a pol I temperature-sensitive (ts) yeast strain (31, 32). Furthermore, functional rRNA can be synthesized from separate 18S and 5.8S/28S transcriptional units driven by the GAL7 promoter (23).

The endonucleases encoded by the three mobile nuclear group I introns constitute a unique example of naturally occurring protein expression from the rDNA gene. Interestingly, these endonucleases (I-PpoI, I-DirI, and I-NanI) also comprise a distinct family based on a common sequence motif of 30 amino acids containing histidine and cysteine residues (20). I-PpoI is the only member of this group for which a genetic system is readily available to study how the endonuclease is expressed from the rDNA gene. In this study, we sought to answer two questions about the expression of I-PpoI. First, is pol I actually responsible for the synthesis of the mRNA for I-PpoI, as might be surmised from the location of the intron in rDNA? A ts pol I strain carrying PpLSU3 in chromosomal rDNA showed no I-PpoI activity at the nonpermissive temperature, directly implicating this RNA polymerase in endonuclease expression. Second, is the mRNA for I-PpoI the full-length excised intron RNA or the processed 5' half intron RNA? A PpLSU3 mutant with greatly reduced levels of the processed 5' half intron RNA showed increased levels of I-PpoI, strongly suggesting that the full-length intron is, or gives rise to, the mRNA for the endonuclease.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Protein extraction and endonuclease activity assay. Yeast cells were grown to an optical density at 600 nm (OD₆₀₀) of 1.0 to 1.5. Cells from a 5-ml culture were spun down and resuspended in 100 µl of breakage buffer (10% glycerol, 200 mM Tris-HCl [pH 8.0], 10 mM MgCl₂, 1 mM phenylmethylsulfonyl fluoride [PMSF]). One hundred microliters of glass beads (425 to 600 µm; Sigma) was added, and the cells were vortexed for 5 min at 4°C. After centrifugation at 4°C for 15 min, the supernatant was used as crude protein extract. Physarum microplasmodia were grown in shaking flasks at 26°C in SDM medium (1). For Physarum protein extracts, each gram (wet weight) of Physarum was homogenized in 2 ml of ice-cold buffer F (50 mM Tris-HCl [pH 8.0], 10 mM EDTA, 1 mM dithiothreitol [DTT], 0.5 mM PMSF, 10 µM leupeptin, 5% glycerol). The homogenate was centrifuged at 80,000 × g for 90 min. The supernatant was dialyzed at 4°C against buffer G (50 mM Tris-HCl [pH 7.5], 1 mM DTT, 0.5 mM PMSF, 5 µM leupeptin, 10% glycerol) for 24 h. Total protein was quantified with the Bio-Rad protein assay dye reagent kit. The I-PpoI endonuclease activity assay was performed as described previously (9). Plasmid p42 was linearized with AvaII and used as the substrate. Yeast crude protein extract was incubated with 300 ng of linearized p42 in I-PpoI buffer (50 mM Tris-HCl [pH 8.0], 10 mM MgCl₂, 50 mM NaCl) at 37°C for 1 h. The reaction was stopped by adding EDTA to a final concentration of 20 mM. Yeast tRNA (25 µg), 1/10 volume of 3 M sodium acetate (NaOAc), and 2 volumes of ethanol were then added to precipitate the DNA. The DNA was resuspended in gel loading buffer (0.05% bromophenol, 0.05% xylene cyanol FF, 6% glycerol) and run on 1.0% agarose gels. For I-PpoI endonuclease activity assays with radiolabeled substrate, primers JL83 (5'TCACCCCGGAATTGGTTTATCC3') and JL84 (5'CGAATGGGACCTTGAATGC3') were used to amplify a 950-bp rDNA fragment in the presence of 20 µCi of [-³²P]dATP in a 100-µl PCR. Twenty nanograms of the PCR product was used in a 50-µl I-PpoI endonuclease assay reaction. The reaction also included 4 µg of poly(dI-dC). The DNA was run on a 1.5% agarose gel, and the gel was dried and exposed to film.

Primer extension. Twenty picomoles of primer JL68 (5'TCTCGCAACATGCACGATGC3') was end labeled with 20 µCi of [-³²P]ATP, using T4 polynucleotide kinase (Boehringer Mannheim Corporation). The labeled primer was purified by passing it through a Sephadex G-25 column. The purified primer was precipitated with 20 µg of total RNA from yeast strain INVSc2/I3 and resuspended in reverse transcriptase buffer (Ambion Inc.). The reaction mixture was incubated at 90°C for 5 min and cooled to room temperature to allow annealing. One microliter of 10 mM deoxynucleoside triphosphates, 15 U of Moloney murine leukemia virus reverse transcriptase (Ambion), and 20 U of RNase inhibitor (Boehringer Mannheim) were added. The reaction mixture (20 µl) was incubated at 37°C for 1 h and ethanol precipitated. The product was fractionated on a 8% polyacrylamide-urea gel. Dideoxy sequencing reactions using the same primer and performed with the dsDNA Cycle Sequencing system (GIBCO BRL Life Technologies, Inc.) were run in parallel. The gel was dried and exposed to film.

RNA preparation and Northern blot analysis. Yeast cells were grown to an OD₆₀₀ of 1.0 to 1.5. Cells from a 10-ml culture were resuspended in 400 µl of AE buffer (50 mM NaOAc [pH 5.3], 10 mM EDTA [pH 8.0]) and 40 µl of 10% sodium dodecyl sulfate (SDS). An equal volume of phenol (450 µl, equilibrated with AE buffer) was added, and the cells were incubated at 65°C for 4 min. The cells were then snap-frozen in an ethanol-dry ice bath and spun for 6 min. The aqueous phase was extracted with phenol-chloroform (equilibrated with AE buffer) twice. RNA was precipitated by adding 1/10 volume of 3 M NaOAc (pH 5.3) and 2.5 volumes of ethanol to the aqueous phase and resuspended in H₂O previously treated with diethylpyrocarbonate.

Plasmid construction. Plasmids were constructed by standard cloning methods (40). Mutations were introduced by PCR in two steps with the mutations included in the primers (47). Plasmid pCPIPpo was constructed by cloning the EcoRI-SphI fragment from pGal IPpo (30) into the EcoRI and SphI sites of the URA3-based plasmid yCP50 (37). Plasmids pJLI3, pJLI3/IPS1, and pJLI3DIPpo contain PpLSU3 with 170 bp of 5' yeast rDNA exon sequence and 130 bp of 3' yeast rDNA exon sequence cloned into the EcoRI and SalI sites of the HIS3-based plasmid pRS423 (43). pJLDIPpo contains two point mutations, at positions 57 and 58 of PpLSU3, to introduce two stop codons. pJLI3IPS1 has mutations at IPS1 changing G/U (the slash indicates the cleavage site) to AA. pJLTtLSU1 was constructed by cloning the EcoRI-HindIII fragment of pSW012 (53) into the EcoRI and SalI sites of pRS423. pJLI3ORF was constructed by cloning the EcoRI-SalI fragment of pI3ORFTZ (35) into the EcoRI and SalI sites of pRS423. It has the ribozyme part of PpLSU3 with 378 bp of Physarum upstream exon and 24 bp of downstream exon. Plasmid pRSIPpo was constructed by cloning the EcoRI-SphI fragment of pGal IPpo into the LEU2-based plasmid pRS415 (43).

Yeast strains and media. Yeast strains (Table 1) were grown according to standard procedures (42) in YEPD medium at 30°C. When selection of plasmids was required, cells were grown in synthetic minimal medium supplemented with amino acids. For detection of I-PpoI activity in strain NOY401/I3, cells were transformed with pNOY103R by the lithium acetate method (2). Transformants were streaked on SD or SGal (42) plates and incubated at either 23 or 37°C, and cells were subsequently inoculated into liquid SD or SGal medium to make the protein extracts.

Transintegration. Plasmid pCPIPpo and one of the pJL series plasmids were cotransformed into yeast strain INVSc2 (Invitrogen Corporation) by the lithium acetate method (2), and the cells were plated on SD-Ura-His plates. Transformants were streaked on SGal-Ura-His plates to induce the expression of I-PpoI. Single colonies on SGal-Ura-His plates were grown in SGal-Ura-His liquid medium. DNA was extracted from the above culture, and PCR was performed to screen for intron-integrated colonies. Primers JL9 (5'CGTGAATTCAACTTAGAACTGGTACG3') and JL8 (5'TATATCGATTCTGCCAAGCCCGT3'), which span the PpLSU3 insertion site, were used for this PCR analysis. Cells were cured of plasmids by being grown in liquid YEPD for approximately 10 generations and then plated on YEPD plates. The loss of the plasmids was checked by replicating colonies on YEPD plates onto SD-Ura and SD-His plates.

Subcellular fractionation of yeast. Subcellular fractionation was performed according to reference 17. Yeast cells were grown to an OD₆₀₀ of 1.0 to 1.5 in YEPD. Cells from 200 ml of culture were harvested, washed with H₂O, and incubated in 10 ml of 100 mM Tris-HCl (pH 9.4)-10 mM DTT at room temperature for 10 min. The cells were spun down and washed in 10 ml of spheroplast buffer (1.2 M sorbitol, 20 mM KPO₄ [pH 7.4]). The cells were incubated in 5 ml of spheroplast buffer with 4 mg of Lyticase (Sigma) at 25°C for 20 min. The spheroplasts were centrifuged and resuspended in 5 ml of homogenization buffer (18% Ficoll DL400 [Sigma], 0.5 mM MgCl₂, 20 mM KPO₄ [pH 6.45]) and homogenized by 10 strokes with a Dounce homogenizer. Five milliliters of sorbitol buffer (2.4 M sorbitol, 0.5 mM MgCl₂, 20 mM KPO₄ [pH 6.45]) was added to the homogenate. The homogenate was centrifuged at 3,000 rpm for 10 min at 4°C. The supernatant was centrifuged again at 12,000 rpm at 4°C for 25 min. The supernatant from the high-speed centrifugation was frozen on dry ice immediately, and RNA was extracted from it later. The pellet from the 12,000-rpm spin was resuspended in 16.6% Ficoll-0.3 M sucrose-0.5 mM MgCl₂-20 mM KPO₄ (pH 6.45) and homogenized again by three strokes. The homogenate was spun at 3,000 rpm for 5 min, and the supernatant was loaded on a sucrose step gradient (2.0, 1.8, 1.5, 1.3, and 1.2 M sucrose in 0.5 mM MgCl₂-20 mM KPO₄ [pH 6.45]). The gradient was spun in an SW60 rotor at 25,000 rpm at 4°C for 1 h, and the nuclear fraction, which is at the interphase of the 1.5 and 1.8 M sucrose steps, was used for RNA extraction.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Accumulation of PpLSU3 RNA and I-PpoI protein in yeast cells carrying the intron in rDNA. To begin to understand how I-PpoI is expressed in yeast, we transformed plasmid pGALI3, which has PpLSU3 flanked by 376 bp of Physarum upstream rDNA sequence and 27 bp of downstream sequence cloned under the GAL1,10 promoter (30), into yeast strain INVSc2 (Table 1). This transformation yielded numerous colonies with PpLSU3 integrated into every rDNA copy, as evidenced by PCR analysis. The new yeast strain derived after curing of the plasmid, called INVSc2/I3, was mildly compromised in growth, with a doubling time of approximately 3 h in YEPD medium, compared with 1.5 h for the wild-type parent strain (data not shown). To identify PpLSU3 RNA species in yeast and to determine their steady-state levels, total RNA was analyzed by Northern blotting with a probe that covers both the I-PpoI ORF and the ribozyme part of PpLSU3 (Fig. 1A). The blot showed expected bands corresponding to the full-length intron RNA, the 5' half intron RNA containing the I-PpoI ORF, and the 3' half intron RNA containing the ribozyme (Fig. 1B), as well as a band smaller than the 3' half intron RNA. Probing the blot separately with the 5' or the 3' half of the intron showed that the extra band was derived from the ribozyme part of the intron (data not shown). For ease of discussion, we call these two processed PpLSU3 RNA species 3' half (L) and 3' half (S), respectively. By comparison with known amounts of PpLSU3 RNA transcribed in vitro from plasmid pI3TZ (35), the amount of PpLSU RNA in yeast was estimated to be 2% of total RNA (data not shown). Consistent with this estimate, the 5' half RNA, 3' half (L), and 3' half (S) could be seen readily on an ethidium bromide-stained polyacrylamide-8 M urea gel (Fig. 5B). These results imply that PpLSU3 RNA is very stable in yeast. Since total mRNA in a cell comprises only a few percent of total RNA (49), PpLSU3 RNA thus is present at a level much higher than that of any actively expressed mRNA species.

View larger version (33K): [in this window] [in a new window]   FIG. 1.   Detection of PpLSU3 RNA and I-PpoI endonuclease activity in the PpLSU3 integrated yeast strain INVSc2/I3. (A) Schematic drawing of the P. polycephalum rRNA gene with PpLSU3 inserted in position 1925 (Escherichia coli reference sequence) of the large subunit. Exon sequences are indicated by filled boxes, and intron sequences are indicated by open boxes. IPS1, internal processing site observed in vitro; IPS2, internal processing site observed in yeast only. The probe used for Northern blot analysis (10) is indicated as a bar below. (B) Northern blot analysis of PpLSU3 RNA in yeast. Lanes: Y, 6 µg of total RNA from INVSc2/I3 (Table 1) run on a 5% polyacrylamide-8 M urea gel followed by Northern blot analysis using a ribonucleotide probe transcribed from plasmid pd55SX; T, in vitro-transcribed PpLSU3 from plasmid pI3TZ submitted to self-splicing conditions; M, Ambion Century RNA marker. (C) I-PpoI endonuclease activity assay. Lanes: M, 1-kb DNA ladder (New England Biolabs); P42, substrate plasmid p42 linearized with AvaII (the band above the major band is incompletely digested product); , linearized p42 incubated with protein extract from intronless yeast strain INVSc2; +, linearized p42 incubated with 100 pg of purified I-PpoI; I3, linearized p42 incubated with 20 or 50 ng of total protein from INVSc2/I3.

Cleavage of intron RNA at two internal processing sites. PpLSU3 processing differs in vitro and in yeast, in that a second 3' fragment of the intron RNA is generated in vivo (Fig. 1B). To further identify this RNA species, we carried out primer extension experiments to map the 5' end of this new RNA species. An additional primer extension product corresponding to the 5' end of the yeast-specific 3' intron RNA was detected 15 bp downstream of the original internal processing site (Fig. 2A) which had been mapped in vitro (38). We call the original site IPS1 and the yeast-specific site IPS2. Previous results had shown that cleavage at IPS1 is mediated by the ribozyme (38). As in the first step of intron splicing, an exogenous G is added to the 5' end of the 3' RNA fragment when cleavage at IPS1 takes place with naked RNA in vitro. The primer extension product therefore extends one nucleotide beyond the actual cleavage site. It is likely that cleavage at IPS2 is also mediated by the ribozyme. This inference is based on the finding that when mutations abolishing ribozyme function were introduced into a plasmid-borne PpLSU3, no cleavage at IPS1 or at IPS2 was observed in yeast (data not shown). However, we cannot rigorously exclude that the yeast-specific cleavage is mediated by a cellular nuclease, nor do we know whether IPS2 cleavage is a G-addition reaction. If it is, IPS2 would map to the sequence GAGAG/AAAA (the slash indicates the cleavage site). In Fig. 2B, the positions of IPS1 and IPS2 are shown in the schematic drawing of part of the secondary structure of PpLSU3. IPS2 is at bottom of the P1 stem, which is formed as in all group I introns by base pairing between the 5' exon and the intron internal guide sequence.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   PpLSU3 RNA is cleaved at two IPSs in yeast. (A) The 5' ends of the 3' half PpLSU3 RNA species in yeast were mapped by primer extension. Note that the sequence from the primer extension reads the complementary strand to the RNA. (B) Schematic drawing of part of the PpLSU3 secondary structure showing the two IPSs as mapped in panel A. Exon sequences are in lowercase, and intron sequences are in uppercase. The 5' splice site is indicated (5'SS).

I-PpoI protein is synthesized from a pol I transcript. At least two hypotheses could explain the low level of I-PpoI protein compared with the high level of PpLSU3 RNA. In the first, the intron RNA is synthesized as a precursor along with the other rRNAs, i.e., by pol I. Intron RNA is then spliced from the precursor and processed further to yield the full-length, 5' half, and 3' half RNA species detected by Northern blotting. I-PpoI protein is translated from one of these pol I-derived RNAs. Lacking a 5' cap and a 3' poly(A) tail, the pol I transcript is translated inefficiently, but because of the abundance of intron RNAs, I-PpoI enzymatic activity nevertheless is easily detected. In the second hypothesis, a cryptic pol II promoter in the 5' rDNA exon directs the synthesis of a low level of I-PpoI mRNA. Such a low level might be undetectable or indistinguishable because of the vast excess of pol I-derived RNA species.

View larger version (39K): [in this window] [in a new window]   FIG. 3.   I-PpoI mRNA is synthesized by pol I. (A) Experimental strategy used to determine which RNA polymerase is responsible for I-PpoI expression. Exon sequences are indicated by filled boxes; intron sequences are indicated by open boxes. Mutations at the I-PpoI target site on plasmid pNOY103R are indicated by an X in the 25S rDNA gene. (B) I-PpoI activity assay showing that I-PpoI protein is not expressed in the pol I ts strain at the restrictive temperature. Lanes: 1, p42 linearized with AvaII; 2 and 3, linearized p42 incubated with protein extract from wild-type (WT) strain INVSc2/I3/pNOY103R grown at 23°C; 4 and 5, linearized p42 incubated with protein extract from NOY401/I3/pNOY103R grown at 23°C; 6 and 7, linearized p42 incubated with protein extract from INVSc2/I3/pNOY103R grown at 37°C; 8 and 9, linearized p42 incubated with protein extract from NOY401/I3/pNOY103R grown at 37°C. For NOY401/I3/pNOY103R grown at 37°C, 1 and 2.5 µg of total protein were used. For the other protein extracts, 200 and 500 ng of total protein were used.

Integration of mutant PpLSU3 into chromosomal rDNA repeats by expression of I-PpoI in trans. To further study how I-PpoI is expressed from the rRNA gene, we wanted to move mutant forms of PpLSU3, including mutants that might not express I-PpoI and thus would themselves be unable to home, into chromosomal rDNA repeats. To overcome this problem, we developed a method called transintegration (Fig. 4A), in which homing of the mutant intron is driven by expression of I-PpoI from a separate plasmid, pCPIPpo. In this plasmid the I-PpoI ORF, by itself without associated intron sequences, is under GAL1,10 promoter control. Most colonies that grow on galactose and hence are resistant to I-PpoI either have acquired point mutations at the I-PpoI target sites in rDNA on chromosome XII or have acquired the intron, thus disrupting that site. To identify survivors that have acquired the intron, PCR with two flanking exon primers can be performed. After the identification of colonies that have acquired the intron, both the donor plasmid carrying the mutant PpLSU3 and the I-PpoI expression plasmid pCPIPpo are cured (see Materials and Methods). I-PpoI endonuclease activity assays then are performed to assess the effect of mutation.

View larger version (35K): [in this window] [in a new window]   FIG. 4.   Integration of mutant forms of PpLSU3 as well as TtLSU1 into chromosomal rDNA repeats. (A) Diagram of the transintegration method. Exon sequences are indicated by filled boxes; intron sequences are indicated by open boxes. Mutations at I-PpoI target sites are indicated by an X on rDNA repeats. (B) Polyacrylamide-8 M urea gel run with total RNA from yeast strains integrated with the constructs described in Materials and Methods. The gel was stained with 0.5 µg of ethidium bromide per ml. INVSc2 is the parental intronless strain. For each strain, 2 and 6 µg of total RNA were loaded.

The IPS1 mutant yeast strain has greatly decreased 5' half PpLSU3 RNA but increased I-PpoI expression. Two major forms of I-PpoI ORF-containing PpLSU3 RNA exist in the cells: the full-length intron RNA and the 5' half RNA. In principle, either or both might be the mRNA for the endonuclease. To determine which RNA species is translated or gives rise to an mRNA, we sought to prevent cleavage at IPS1 by mutating this site from G/U to AA (the slash indicates the actual cleavage site) (Fig. 2B and 4B). This mutant form of PpLSU3 was successfully integrated into rDNA repeats by means of the transintegration method, and then I-PpoI activity was assayed in crude extracts and RNA species were analyzed by Northern blotting. The experiment was based on the prediction that if the 5' half RNA is the mRNA, mutation that abolishes cleavage at the IPS should decrease I-PpoI expression, whereas if the full-length PpLSU3 RNA is the messenger, the IPS1 mutant strain should have the same or increased I-PpoI activity. Northern blot analysis showed that cleavage at IPS1 was abolished, as indicated by the disappearance of the 3' half RNA (L) band (Fig. 5A). However, the 3' half RNA (S) band was still present in the IPS1 mutant, indicating that cleavage at IPS2 is independent of IPS1. Surprisingly, no 5' half PpLSU3 RNA was detected in the IPS1 mutant. Since cleavage at IPS2 did take place, as evidenced by the presence of the complementary 3' half RNA, we conclude that the 5' half PpLSU3 RNA generated by cleavage at IPS2 [5' half (L)] must be unstable. In fact, a low amount of this species was visible on Northern blots when total RNA was analyzed on formaldehyde-agarose gels, which in our hands detected lower levels of RNA than polyacrylamide gels (data not shown).

View larger version (55K): [in this window] [in a new window]   FIG. 5.   Cleavage at IPS1 is abolished in the IPS1 mutant yeast strain, but I-PpoI expression is increased. (A) Northern blot analysis of the IPS1 mutant yeast strain. Lanes: 1 and 2, total RNA from wild-type (WT) strain INVSc2/I3; 3 and 4, total RNA from INVSc2/IPS1. For each strain, 2 and 6 µg of total RNA were loaded. Plasmid pd55SX was used to make the riboprobe. The I-PpoI ORF is indicated by a filled box; the rest of the intron sequence is indicated by black bars. The dotted box represents the 5' half (L) RNA, which is not detectable on this gel. (B) I-PpoI activity assay for the IPS1 mutant yeast strain. Lanes: M, 1-kb DNA ladders (New England Biolabs); P42, p42 linearized with AvaII; , linearized p42 incubated with protein extract from the intronless strain INVSc2; WT, linearized p42 incubated with protein extract from INVSc2/I3. IPS, linearized p42 incubated with protein extract from INVSc2/IPS1. For each strain, 50, 200, and 500 ng of total protein were used.

Subcellular distribution of PpLSU3 RNA species in yeast. To gain further support for the conclusion that the full-length PpLSU3 RNA is or gives rise to the mRNA, we performed subcellular fractionation experiments with the yeast strain INVSc2/I3. Broken yeast cells were separated into nuclear and cytoplasmic fractions as described in Materials and Methods. RNA was prepared from both fractions and subjected to Northern blot analysis along with total RNA. As marker for the integrity of nuclei, radioactive RNA complementary to the small nucleolar RNA U3 was used to probe the same fractions (Fig. 6). The majority of U3 was found in the nuclei as expected, but a small portion (approximately 8%) was detected in the cytoplasm. To facilitate the detection of small amounts of RNA, Northern blot analysis was performed with formaldehyde-agarose gels, on which the 3' half (L) and 3' half (S) RNA species are not resolved from one another. We found that all PpLSU3 RNA species were present in both nuclear and cytoplasmic fractions (Fig. 6). About 38, 65, and 70% (averages from two independent experiments) of the full-length, 5' half, and 3' half PpLSU3 RNAs, respectively, were present in the cytoplasmic fraction. These data by themselves do not help clarify the nature of the mRNA. However, subcellular fractionation with the IPS1 mutant strain INVSc2/IPS1 showed a similar distribution of intron RNA species, except that the 5' half (L) RNA, which we showed previously was unstable, was about 2% of the wild-type level in both the nuclear and cytoplasmic fractions (data not shown). Given the increased levels of endonuclease in this mutant, the greatly reduced levels of cytoplasmic 5' half RNA suggest strongly that this species cannot serve as the mRNA. Taken together, we interpret these data to support the model that the intact excised intron RNA is, or gives rise to, the mRNA. Our data do not address the possibility that a minor RNA species derived from the full-length RNA, but not the full-length RNA itself, is the mRNA. This possibility is difficult to test since such a minor species might be undetectable.

View larger version (64K): [in this window] [in a new window]   FIG. 6.   Subcellular distribution of PpLSU3 RNA species in yeast. Yeast cells were fractionated as described in Materials and Methods. RNA was prepared from the nuclear and cytoplasmic fractions and subjected to Northern blot analysis on a 1.5% formaldehyde-agarose gel, using a riboprobe synthesized from pd55SX. Lanes: T, total RNA from INVSc2/I3; N, nuclear fraction; C, cytoplasmic fraction. U3, the same fractions probed with a riboprobe complementary to U3.

Measurement of PpLSU3 RNA and I-PpoI expression in Physarum. To investigate if the results of these studies with yeast also apply to I-PpoI expression from the intron in its natural host, the acellular slime mold P. polycephalum, we performed Northern blot analysis with total RNA from Physarum microplasmodia. PpLSU3 RNA was found to accumulate to a much lower level in Physarum than in yeast (Fig. 7A). Among the three PpLSU3 RNA species, the 5' half RNA was at a much higher level than the full-length RNA and the 3' half RNA. The lower level of the 3' half intron RNA indicates that it is less stable than the 5' half intron RNA, in contrast to the relative stabilities of these species in yeast. The apparent low level of the full-length RNA in Physarum may be due to the more efficient cleavage at the IPS in the natural host for this intron. Quantitation of the Northern blot data showed that the amounts of the full-length, 5' half, and 3' half PpLSU3 RNAs were 250-, 30-, and 200-fold, respectively, lower than those in yeast. Therefore, it appears that all PpLSU3-derived RNA species are less stable in Physarum than in yeast.

View larger version (41K): [in this window] [in a new window]   FIG. 7.   Both PpLSU3 RNA and I-PpoI protein accumulate at a low level in Physarum. (A) RNA samples were run on a 1.5% formaldehyde-agarose gel, and Northern blot analysis was carried out as described in the legend to Fig. 6. Lanes: Y, 100 ng of total RNA from yeast strain INVSc2/I3; P, 8 µg of total RNA from Physarum. (B) Detection of I-PpoI activity in Physarum protein extract. A PCR product containing the I-PpoI target site amplified in the presence of [-32P]dATP was used as the substrate to be incubated with total protein extract from Physarum and yeast.

4

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Expression of I-PpoI from a pol I transcript. We have provided direct evidence that the I-PpoI protein is translated from the pol I transcript from the chromosomal rDNA locus of rDNA of yeast. The rule that protein-encoding genes are transcribed by eucaryotic pol II is commonly accepted. A functional consequence of pol II transcription is that (with few exceptions) mRNAs contain a 5' cap and a 3' poly(A). These modifications serve to enhance translation efficiency in several ways, such as promoting export of mRNA to the cytoplasm, stabilizing mRNA, and stimulating translation initiation. Besides the three examples of mobile nuclear group I introns, a possible exception to this rule is provided by the variable surface glycoprotein gene of Trypanosoma brucei, which has been suggested to be transcribed by pol I (55). But in this case, a small RNA synthesized by pol II is ligated to the 5' end of the pol I transcript by trans splicing, so that the mature mRNA still contains a 5' cap (28, 46, 55). Furthermore, the mRNA is polyadenylated at the 3' end. A second possible exception is offered by retrotransposons in the rDNA gene of insects, exemplified by the R2 element in Bombyx mori, which are inserted into multiple copies of rDNA repeats in many insects. But unlike the mobile group I introns in rDNA genes, where every rDNA copy has the intron, the R2 element never occupies more than about 30% of all rDNA repeats (19). Moreover, the rDNA copies carrying R2 elements are transcriptionally inactive. However, since mobility of the R2 element is observed occasionally, transcription of the retrotransposon is inferred to occur at a low level. The nature of the RNA polymerase that transcribes the R2 element has not been investigated in detail. I-PpoI, therefore, is the first well-documented case of protein expression by a pol I transcript from the chromosomal rDNA locus.

Cleavage at the IPS as a mechanism to regulate I-PpoI expression. The IPS1 mutant has increased I-PpoI activity, suggesting that the full-length PpLSU3 RNA is the real messenger. Perhaps cleavage of the full-length RNA at the IPS has evolved as a mechanism to destroy the mRNA, in order to downregulate expression of I-PpoI. Downregulation of homing endonuclease expression is a common feature of mobile group I introns. Group I intron-encoded endonucleases recognize long DNA sequences of 12 to 40 bp, which ensures that the endonuclease has only one target site in the genome. However, at least under experimental conditions in vitro, these endonucleases can tolerate base changes in their recognition sites (52). Therefore, overexpression of the endonuclease potentially can be deleterious to the host cell. This is especially critical for nuclear endonucleases, since nuclear genomes are vastly larger than organellar genomes. In the case of the mitochondrially encoded yeast nucleases I-SceII, I-SceIII, and I-SceIV, endonuclease activity comes from a fusion protein that is in frame with the upstream exon (16, 27, 41, 50). Expression is therefore downregulated by splicing of the intron from pre-mRNA. I-SceI, the endonuclease encoded by the  intron, is expressed from a minor species derived from cleavage at a certain site (54). Endonucleases of T-even phage group I introns are expressed only at later stages of phage infection (14).

A translation enhancer element in the full-length intron RNA? The results presented in this paper strongly suggest that the full-length RNA is or gives rise to the real mRNA. How does the cellular translation machinery determine to translate the full-length RNA but not the 5' half RNA? Perhaps the ribozyme part of PpLSU3 RNA, as a highly structured RNA, acts as a translation enhancer element to bind to protein factors that augment translation initiation. There are numerous examples of translation control elements in the 3' untranslated region (UTR). Typical mRNAs have a 3' poly(A) tail that enhances translation by stabilizing mRNA and by increasing translation initiation, acting synergistically with 5' cap (12). The 3' UTR of histone mRNAs enhances translation by serving as an exporting signal and by stabilizing the RNA (26, 45). The 3' UTR region of PAV barley yellow dwarf virus acts as a translational enhancer that mimics a 5' cap in facilitating translation of uncapped mRNA (48). Recently, the  intron RNA of yeast has been shown to be in ribonucleoprotein complexes of 50S (34). Our preliminary data show that the majority of PpLSU3 RNA also is not free, indicating the presence of bound proteins. Identifications of these proteins may help elucidate the role of the 3' sequence of PpLSU3 RNA in translation.