The group I-like ribozyme DiGIR1 mediates alternative processing of pre-rRNA transcripts in Didymium iridis Anna Vader 1,2 , Steinar Johansen 2 and Henrik Nielsen 1 1 Department of Medical Biochemistry and Genetics, The Panum Institute, Copenhagen, Denmark; 2 Department of Molecular Biotechnology, Institute of Medical Biology, University of Tromsø, Norway During starvation induced encystment, cells of the myxo- mycete Didymium iridis accumulate a 7.5-kb RNA that is the result of alternative processing of pre-rRNA. The 5¢ end corresponds to an internal processing site cleaved by the group I-like ribozyme DiGIR1, located within the twin- ribozyme intron Dir.S956-1. The RNA retains the majority of Dir.S956-1 including the homing endonuclease gene and a small spliceosomal intron, the internal transcribed spacers ITS1 and ITS2, and the large subunit rRNA lacking its two group I introns. The formation of this RNA implies clea- vage by DiGIR1 in a new RNA context, and presents a new example of the cost to the host of intron load. This is because the formation of the 7.5-kb RNA is incompatible with the formation of functional ribosomal RNA from the same transcript. In the formation of the 7.5-kb RNA, DiGIR1 catalysed cleavage takes place without prior splicing per- formed by DiGIR2. This contrasts with the processing order leading to mature rRNA and I-DirI mRNA in growing cells, suggesting an interplay between the two ribozymes of a twin- ribozyme intron. Keywords: Didymium iridis; group I intron; ribozyme; pre- rRNA processing. Group I introns contain a conserved set of sequences and structural elements that are involved in the removal of the intron by splicing. They constitute one class out of fewer than 10 classes of naturally occurring ribozymes [1]. Group I introns vary considerably in complexity. Most introns contain only the sequence information required for splicing, whereas others contain large extensions of the peripheral domains. Some of the larger group I introns contain an open reading frame, usually represented by a homing endonuclease gene (HEG). HEGs are found in different configurations, e.g. fused in frame with the upstream exon or as an independent expression unit [2]. The most complex group I introns are the twin-ribozyme introns that in addition contain a group I-like cleavage ribozyme (GIR1) involved in the expression of the intron HEG [3]. The complex structure of the twin-ribozyme introns suggests a complex biology. This has been demonstrated in the case of the Dir.S956-1 (former DiSSU1; the recently introduced nomenclature for group I introns [4] is used throughout this paper) intron found in the small subunit ribosomal RNA (SSU rRNA) gene in the myxomycete Didymium iridis (Fig. 1). One of the ribozymes (DiGIR2) catalyses intron excision and exon ligation (Fig. 1, left panel). In addition, this ribozyme displays a pronounced 3¢ splice site hydrolysis activity, which induces the formation of full-length intron RNA circles using a processing pathway that is distinctly different from splicing ([5]; unpublished data]. The other ribozyme (DiGIR1), which along with the I-DirI HEG is inserted in DiGIR2, carries out hydrolysis at two internal processing sites (IPS1 and IPS2) located at its 3¢ end [5,6]. In vivo, this cleavage results in the formation of the 5¢ end of the I-DirImRNAandis followed by cleavage at an in vivo specific internal processing site (IPS3) downstream of the HEG and by polyadenylation (summarized in Fig. 1, left panel). Finally, a 51-nucleotide spliceosomal intron (I51) within the HEG RNA is removed before the resulting I-DirI mRNA is transported to the cytoplasm where it associates with the polysomes [7]. Homing activity of the I-DirI protein has been demonstra- ted by Dir.S956-1 intron mobility studies involving genetic crosses between intron-containing and intron-lacking Didymium isolates [8]. During our work on the in vivo expression of Dir.S956-1, we noted the presence of an I-DirI HEG-containing RNA species that migrated similarly to the 7.46-kb ladder band on a denaturing agarose gel, but did not hybridize to an SSU probe. This observation, as well as reverse transcription/ PCR analyses which showed that Dir.S956-1 produces full- length intron circles in vitro [9] and in vivo [10], led us to believe that this unknown RNA represented a circular species that was retarded in the gel during electrophoresis. We have subsequently observed that the signal intensity of the 7.5-kb band varies greatly according to the state of the D. iridis culture when the RNA was isolated. To address the question of its formation and possible function in cellular I-DirI expression, we here investigate its identity and distribution. We demonstrate that the RNA is a 7.5-kb Correspondence to H. Nielsen, Department of Biochemistry and Genetics, Laboratory B, The Panum Institute, Blegdamsvej 3, DK-2200 N, Denmark. Fax: +45 35 32 77 32, Tel.: +45 35 32 77 63, E-mail: hamra@imbg.ku.dk Abbreviations:DiGIR,Didymium group I ribozyme; SSU, small subunit; LSU, large subunit; ETS, external transcribed spacer; ITS, internal transcribed spacer; HEG, homing endonuclease gene; IPS, internal processing site. (Received 24 July 2002, revised 24 September 2002, accepted 30 September 2002) Eur. J. Biochem. 269, 5804–5812 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03283.x linear species generated by an unusual pre-rRNA processing event mediated by DiGIR1, and that it accumulates during starvation-induced encystment in D. iridis. EXPERIMENTAL PROCEDURES Cell cultivation, RNA isolation and Northern blotting analysis The intron-containing D. iridis strain Lat3-5, derived from the Pan2-44 isolate, has been described previously [8]. The cells were cultured at 26 °C in liquid media (DS/2; 1mgÆmL )1 D -glucose, 0.5 mgÆmL )1 yeast extract, 0.1 mgÆmL )1 MgSO 4 ,1mgÆmL )1 KH 2 PO 4 ,1.5mgÆmL )1 K 2 HPO 4 ) containing Escherichia coli cells. Cells and cysts were counted in a Tu ¨ rk chamber or electronically in a Coulter Multisizer (Coulter Electronics Ltd). Cysts were scored by their ability to resist lysis in 0.5% Nonidet P-40 [11]. Cysts were stained by the addition of 1 vol. 0.25% Trypan Blue in standard NaCl/P i . For RNA extraction, a total of 10 7 Didymium cells were harvested by centrifugation at 400 g for 5 min The pellet was dissolved in 1 mL Trizol Reagent (Gibco-BRL) and RNA extracted according to the manufacturer’s instructions. Aliqouts of RNA were denatured for 15 min at 65 °C in loading buffer (1 · Mops, 17.8% formalde- hyde, 50% formamide, 12 ngÆlL )1 ethidium bromide) and fractionated on a 5.3% formaldehyde/1% agarose gel in 5.3% formaldehyde/1 · Mops (40 m M Mops, 10 m M NaAc, 2 m M EDTA, 0.04% HAc). The RNA was then transferred to a nylon membrane by Northern blotting using capillary action. Hybridization was carried out either in Rapid Hyb solution (AP Biotech) or in Ultrahyb solution (Ambion) according to the manufacturer’s instructions. The external transcribed spacer (ETS), GIR1, HEG, GIR2, internal transcribed spacer (ITS)1, ITS2 and large subunit (LSU)1 probes were amplified from Lat3-5 genomic DNA [8] by PCR using the oligo pairs OP448/OP449, OP20/C78, OP1/OP2, OP11/OP180, SSU6/SSU7, C231/ C232 and OP65/OP169, respectively. The sequences of the oligonucleotides are: OP1, 5¢-CACTTCTAGAACCA TGGTGAAAGGAACG-3¢;OP2,5¢-TGTCTGGATCCT CATCTG-3¢;OP4,5¢-TGTTGAAGTGCACAGATT-3¢; OP11, 5¢-GACTAGTTGACTTCTCACAGA-3¢; OP20, 5¢-TTGAACACTTAATTGGGT-3¢;OP65,5¢-GGAG Fig. 1. Homing endonuclease gene expression and life cycle of D. iridis. Proposed processing pathways in the formation of RNA species encoded by the Dir.S956-1 homing endonuclease gene (HEG). In vegetatively growing D. iridis, the Dir.S956-1 intron is spliced out from pre-rRNA and further processed into an I-DirI endonuclease mRNA (left panel; see [7] for details). Starvation/encystment results in an alternative processing pathway of the intron (right panel), induced by DiGIR1 ribozyme cleavage at an internal intron processing site. Subsequently, a 7.5-kb linear RNA is formed after the excision of the LSU rRNA introns Dir.L1949 and Dir.L2449. The accumulated 7.5-kb RNA contains all of the Dir.S956-1 sequences except those encoding the cleavage ribozyme DiGIR1. A possible functional role of the 7.5-kb RNA is as an alternative precursor for the endonuclease mRNA during excystment. Here, the HEG RNA might be separated from the remaining 7.5-kb RNA sequences by cleavage at the host induced IPS3 or the ribozyme induced 3¢ splice site. (A) 1.46 kb RNA (the full length intron after splicing). (B) 1.23 kb RNA (resulting from GIR1 cleavage). (C) 0.90 kb RNA (resulting from cleavage at IPS3). (D, E) Nuclear and cytoplasmic form of the 0.85 kb RNA also referred to as the I-DirI mRNA. (Inset) Life cycle of the myxomycete D. iridis. Haploid amoebae or swarm cells can transform into dormant cysts under unfavourable environmental conditions. This process is reversible. Alternatively, two compatible amoebae or swarm cells can act as gametes and fuse to produce a diploid zygote. Growth of the zygote is accompanied by a series of nuclear divisions, leading to the formation of a multinucleated plasmodium. Eventually, the plasmodium transforms into fruiting bodies, which release haploid spores. Germination of the spores completes the life cycle, in that vegetative amoebae or swarm cells are formed again. Ó FEBS 2002 Ribozyme mediated processing of pre-rRNA (Eur. J. Biochem. 269) 5805 GTTCAGAGACTATA-3¢;OP169,5¢-ACCTAAGGC GGACGTTACTG-3¢;OP180,5¢-GCCTCCCTTGGGA TAT-3¢; OP448, 5¢-AACCGAACAATGAGACTGAA-3¢; OP449, 5¢-CTCGTATTCGAAGGCATGCA-3¢;C78, 5¢-TGCTTCCTTTCGGAACGA-3¢; C231, 5¢-ATTCCGA TATCGTGCTCTA-3¢; C232, 5¢-AAGAGGTTGGCCAA GGAA-3¢; SSU6, 5¢-CGAATTCAGGGGCAACATCGG TTC-3¢;SSU7,5¢-CGAATTCACCGAGGTTACAAG GCA. The ETS, GIR1, HEG, GIR2 and LSU1 PCR products were purified on S-300 spin-columns (Pharmacia) prior to labelling by random priming using the Mega Prime kit (Amersham) and [a- 32 P]dCTP (3000 CiÆmmol )1 ; Amersham). The ITS1 and ITS2 PCR products were cloned using the Topo TA cloning kit (version J, Invitrogen) according to the manufacturer’s instruction. Plasmids harbouring the ITS1 insert in the correct orientation were linearized by HindIII digestion, while the ITS2 insert was further subcloned into the XbaI/HindIII site of the pBlue- script+ vector (Stratagene) to obtain the correct orienta- tion. The resulting pBluescript plasmid was linearized with XbaI. Riboprobes were transcribed from 500 ng linearized template DNA using 500 l M each of rATP, rCTP and rGTP, 25 l M rUTP, 0.5 l M [a- 32 P]UTP (3000 CiÆmmol )1 ; Amersham), 10 m M dithiothreitol and 50 U T7 RNA polymerase (Stratagene) in 20 lLof1· the supplied buffer at 37 °Cfor1h. RNaseH analysis A mix consisting of 6 lg RNA and 50 pmol oligo was heated in 1 · RNaseH buffer (GibcoBRL) at 80 °Cfor 1min.At45°C, 20 U RNasin (Pharmacia) was added, and the sample incubated for 10 min. After transfer to ice, 0.5 U RNaseH (GibcoBRL) was added to produce a total volume of 10 lL. The sample was then incubated at 30 °Cfor 5 min, prior to analysis by Northern blotting (see above). Primer extension For primer extension, gel-purified OP4 was labelled with [a- 32 P]ATP (3000 CiÆmmol )1 , Amersham) using T4 poly- nucleotide kinase (Gibco-BRL). RNA was added to 2 pmol labelled oligo in 1 · RT buffer (50 m M Tris/HCl at pH 8, 60 m M KCl, 10 m M MgCl 2 ,1m M dithiothreitol) in a total volume of 5 lL, denatured at 80 °Cfor2minand incubated at 45 °C for 10 min. Subsequently 4 lLRNA/ oligo mixture was added to a tube containing 1 U AMV reverse transcriptase (RT; Pharmacia), 1 U RNasin (Promega), 0.2 m M dATP, dCTP and dTTP and 0.4 m M dGTP (Pharmacia) in 1 · RT buffer. The reaction was incubated1 hat40°C before being stopped by the addition of 5 lL formamide loading buffer. The primer extension product was denatured by heating at 100 °Cfor1min before separation on an 8 M urea/8% polyacrylamide gel. Cell fractionation and sucrose gradients DS/2 (see above) was added to 2 · 10 7 Lat3-5 cells to a total volume of 2 · 14 mL and centrifuged in two tubes at 300 g for 5 min The pellet from one tube was dissolved in 1 mL Trizol (see above) for extraction of total RNA. The cells in the other tube were resuspended in 250 lL ice-cold lysis buffer (20 m M Tris/HCl pH 8.0, 1.5 m M MgCl 2 ,140m M KCl, 1.5 m M dithiothreitol, 1 m M CaCl 2 ,0.1m M EDTA, 0.16 m M cycloheximide, 0.5% Nonidet P-40, 500 UÆmL )1 RNasin), incubated for 5 min in ice/water to allow lysis of the cells and centrifuged at 10 000 g,4°C for 10 min. The pelleted nuclei were dissolved in Trizol (nuclear RNA), and the supernatant was extracted with phenol/chloroform and precipitated by EtOH (cytosolic RNA). For sucrose gradients, 250 lg whole cell RNA was heated to 70 °C for 5 min, cooled on ice and centrifuged at 13 000 g,4°C for 5 min The supernatant was loaded onto a linear 15–40% sucrose gradient in 10 m M Tris/HCl at pH 7.5, 100 m M LiCl, 10 m M EDTA and 0.2% SDS and centrifuged for 20 h at 4 °C and 25 000 r.p.m. in a Beckman SW27.1 rotor. Fractions of approximately 1 mL were collected and RNA was isolated by phenol/chloroform extraction. RESULTS A 7.5-kb I-DirI HEG RNA signal is enriched upon starvation of Didymium cells The life-cycle of a typical myxomycete can be roughly divided into a diploid macroscopic stage consisting of a plasmodium and the fruiting bodies that develop from it, and a haploid microscopic stage (see Fig. 1B). Haploid myxomycete cells are uninucleate and exist in two intercon- vertible active states; nonpolarized amoebae and polarized flagellated swarm cells. The particular form in which a given cell exists depends largely upon the amount of water in the environment, with swarm cells tending to dominate under aqueous conditions. In nature, myxamoebae or swarm cells feed by phagocytosis of bacteria. Under conditions unfa- vourable for continued growth, such as starvation, the vegetative cells will develop into dormant cysts. Cysts can remain viable for long periods of time, and have been suggested to be very important for the survival of myxomycetes in some habitats. To examine whether the cellular amount of the 7.5-kb I-DirI HEG RNA correlates with food availability, intron- harbouring Lat3-5 amoebae were grown in monoxenic culture using E. coli as a food source (Fig. 2A). As the cells grow and multiply, food is depleted (time points 1–5) and the myxamoebae gradually transform into very active swarm cells (points 5–6). Eventually, activity ceases and the starving cells develop into dormant cysts (point 7). As encystment is defined by the formation of a cell wall, we have chosen to score cysts by their ability to resist lysis in 0.5% Nonidet P40 [11]. However, it is important to keep in mind that cyst formation is most likely committed biochemically long before this time. Examination of whole cell RNA from a time course of a Didymium culture shows that the 7.5-kb RNA is hardly detectable at the first time points when food is plentiful, but becomes abundant when the cells are starved and the culture reaches the stationary phase (Fig. 2B). At the last time points the 7.5-kb RNA is the predominant HEG RNA in the cells. While the amounts of some of the other HEG RNA species also vary, none exhibits the same pattern. It is interesting to note that another prominent signal corresponding to a 3.9- kb RNA, which comigrates with the LSU rRNA, decrea- ses as the 7.5 kb signal increases. The 3.9 kb RNA appears 5806 A. Vader et al. (Eur. J. Biochem. 269) Ó FEBS 2002 to be a nuclear species [7], and a similar RNA has been observed when whole cell RNA from the Didymium CR8 isolate was probed with the Dir.S956-2 group I intron [10]. The fact that the Dir.S956-1 and Dir.S956-2 group I introns are unrelated [10], suggests that the formation of the 3.9 kb RNA is independent of the intron, and results from a more general alternative pathway of Didymium pre- rRNA processing. The 7.5-kb signal is a linear RNA made by alternative processing of the pre-rRNA To confirm that the 7.5-kb signal indeed represented a circular form of the Dir.S956-1 intron RNA, the following experiments were carried out. First, RNA from the time course experiment shown in Fig. 2, was analysed on a denaturing 4% polyacrylamide gel in diluted electro- phoresis buffer (0.4 · TBE). Under these conditions we know ) from repeated experiments using in vitro tran- scribed and processed RNA ) that circles are retarded and thus efficiently separated from the corresponding linear form of the intron RNA. Northern blotting analysis showed that an RNA with the same migration as a Dir.S956-1 circle is indeed present in Didymium cells, but that this RNA is enriched at the start of the time course rather than during starvation (data not shown). Second, the circular and linear forms of Dir.S956-1 RNA from an in vitro splicing reaction were separated on a denaturing polyacrylamide gel, cut out and recovered after elution. The RNA species were analysed on a denaturing agarose gel. The results showed that the Dir.S956-1 circle is only slightly retarded on an agarose gel (data not shown). Thus, contrary to our previous Fig. 2. Analysis of RNA from D. iridis Lat3-5 cells harvested from a time course growth experiemt. (A) Time course of culture growth, showing starvationandsubsequentencystmentofvegetativeD. iridis Lat 3-5 cells. The time points when total number of Didymium cells (d), number of encysted cells (s)oramountofE. coli food (j) was measured are indicated. The numbered arrows denote the time points when RNA samples were obtained. (B) Northern blot of Lat3-5 whole cell RNA. Didymium cells (5 · 10 6 ) were harvested at the time points indicated in (A). After extraction, the RNA was separated on a 1% denaturing agarose gel and analysed by Northern blotting analysis using the HEG probe described in Fig. 3A. An overview of the observed intron RNA species is shown to the right. Exon, open reading frame and intron sequences are indicated in black, grey and white, respectively. The position of the 5¢ and 3¢ splice sites (SS) as well as internal processing sites (IPS) are indicated. The 1.46-kb RNA is the full- length excised Dir.S956-1 intron, while 1.23-kb RNA and 0.85-kb RNA represent processed forms of the intron RNA [7]. The 7.5-kb signal under investigation is denoted**, while the identity of the signal marked * is discussed in the text. The size indications are derived from the 0.24–9.5 kb ladder (GibcoBRL) visualized by ethidium bromide staining. Ó FEBS 2002 Ribozyme mediated processing of pre-rRNA (Eur. J. Biochem. 269) 5807 suggestion, the 7.5-kb signal does not correspond to the full length circular intron RNA. Next, we hypothesized that the 7.5-kb RNA is formed by an alternative processing of the pre-rRNA in which the SSU rRNA sequence upstream of Dir.S956-1 is removed. This hypothesis would similarly be consistent with the previously published observation of lack of hybridization of an upstream SSU probe to the 7.5-kb RNA [7]. Considering the low abundance of this RNA compared with ribosomal RNAspecies,andthefactthatitcoexistswithRNAs containing the same structural elements, we decided to deduce its structure by analysis of preparations of whole cell RNA rather than to isolate it. Whole cell RNA was isolated from Didymium cellsharvestedearlyandlateinatime course (corresponding to time points 1 and 6 in Fig. 2). These RNAs were analysed by Northern blotting and RNaseH cleavage. In the Northern blotting analysis, parallel filters were hybridized with a panel of probes complementary to different parts of the Didymium pre- rRNA including ETS, HEG, ITS1 and ITS2 (Fig. 3A). A signal of 9.5 kb was detected only by the non-Dir.S956-1 probes (i.e. ETS, ITS1 and ITS2; Fig. 3B), suggesting that it represents the pre-rRNA subsequent to Dir.S956-1 excision. The observation of the 9.5 kb RNA is in agreement with splicing being one of the earliest events in pre-rRNA processing, as previously shown for the Tth.L1925 intron in Tetrahymena thermophila [12]. Although information on the precise location of the 5¢ and 3¢ ends of Didymium pre- rRNA is not available, the size of 9.5 kb for this RNA is in reasonable agreement with the expected size based on reported ribosomal DNA sequences from different Didymium isolates. Fig. 3. Characterization of the 7.5-kb RNA signal. (A) Schematic presentation of the D. iridis (Lat3-5 strain) rDNA. The upper panel shows the SSU and LSU rRNA genes, as well as the position of the Dir.S956-1, Dir.L1949 and Dir.L2449 group I introns [9,18]. Exon and intron sequences are denoted in black and white, respectively. The localization of the probes applied for Northern blotting analyses is indicated with thick black lines. The lower panel is an enlarged segment of the upper panel, and shows the positions of the oligonucleotides used in RNaseH analysis. (B) Northern blotting analysis of whole cell RNA. Parallel filters containing RNA from Didymium cells harvested at an early (E) and late (L) time point (corresponding to positions 1 and 6 in Fig. 2A) were hybridized with the probes indicated. The size indications are derived from the 0.24- to 9.5-kb ladder (GibcoBRL) visualized by ethidium bromide staining. (C) RNaseH analysis of the 7.5-kb RNA signal. Whole cell RNA from a late time point (corresponding to point 6 in Fig. 2A) was fractionated on a 15–30% sucrose gradient (see Fig. 5). Fractions enriched in the 7.5-kb RNA, as determined by Northern blotting analysis, were pooled and the extracted RNA subjected to RNaseH analysis with the oligonucleotides indicated. The resulting RNA was analysed by Northern blotting using the HEG probe shown in (A). The size indications are derived from the High Range RNA ladder (Fermentas) visualized by ethidium bromide staining. 5808 A. Vader et al. (Eur. J. Biochem. 269) Ó FEBS 2002 The RNA species that are involved in expression of the HEG, and particularly those that are involved in the formation of the 7.5-kb RNA, are considerably less abundant than most of the ribosomal RNA precursors. For this reason, detection of HEG RNAs with non-HEG probes was technically difficult. Nevertheless, a 7.5-kb signal specific for late RNA was identified by the ITS1 and ITS2 probes, but was absent when a 5¢ ETS probe was used (Fig. 3B). The 7.5-kb RNA was furthermore detected by exonic LSU probes but was not detected by probes targeted towards the LSU introns (data not shown). The absense of the LSU introns in the 7.5-kb RNA implies a different splicing order during the formation of this RNA compared with normal processing of the pre-rRNA in which the Dir.S956-1 sequences are removed prior to the LSU introns (see above). The expected size of an RNA composed of the elements suggested by the Northern blotting experiment is  7.6 kb which is in good agreement with the observed size of 7.5 kb. In the RNaseH analysis, the parallel RNA samples were hybridized to oligonucleotides complementary to different parts of the Didymium pre-rRNA (see Fig. 3A). RNaseH, which will degrade the RNA strand of the resulting RNA : DNA heteroduplexes, was then added. The resulting RNA fragments were visualized by Northern blotting analysis using a HEG probe. The experiment showed that oligonucleotides complementary to ITS-1, ITS-2, LSU as well as the SSU-sequence downstream of Dir.S956-1 will induce RNaseH-catalysed cleavage of the 7.5-kb RNA and result in cleavage products of the expected size (Fig. 3C). Oligonucleotides complementary to SSU sequences upstream of Dir.S956-1, on the other hand, had no effect (data not shown), showing that these sequences are not part of the 7.5-kb RNA. Taken together, the data from Northern blotting and RNaseH analyses are consistent with the hypothesis that the 7.5-kb RNA signal is a linear species produced by alternative processing of the pre-rRNA. Only parts of the Dir.S956-1 intron are included in the 7.5-kb RNA To determine if all of the Dir.S956-1 intron is included in the 7.5-kb RNA, filters containing whole cell RNA from early and late points in a time course (see above) were hybridized with probes complementary to GIR1, HEG and GIR2 (Fig. 4A). Surprisingly, the Northern blotting results dem- onstrated that GIR1 is absent from the 7.5-kb RNA (Fig. 4B). RNaseH analyses substantiated this conclusion in that oligonucleotides complementary to GIR1 did not cleave the 7.5-kb RNA, whereas oligonucleotides that hybridize to different parts of the HEG or GIR2 led to cleavage (data not shown). The sizes of the upstream cleavage products, detected by their hybridization to a HEG probe in Northern blotting analysis, suggested that the 5¢ end of the 7.5-kb RNA was located at or near the internal processing sites IPS1/2. RNaseH analysis was also used to examine whether the 51-nucleotide spliceosomal intron within the HEG is present in the 7.5-kb RNA. This intron is removed during maturation of the I-DirImRNAfromtheexcised Dir.S956-1 RNA [7]. Cleavage with an oligonucleotide complementary to the spliceosomal intron sequences showed that these are retained in the 7.5-kb RNA (data not shown). The DiGIR1 ribozyme catalyses 5¢ end formation of the linear 7.5-kb RNA To map precisely the 5¢ end of the 7.5-kb RNA, it was necessary to separate it from other Dir.S956-1-containing RNAs in the cell. Whole cell RNA from a late time point was fractionated by ultracentrifugation through a 15–40% sucrose gradient (Fig. 5A). Screening of the collected fractions by Northern blotting analysis, showed that the HEG-containing RNAs could successfully be separated according to size (Fig. 5B). Primer extension with an oligo complementary to the 5¢ end of the HEG produced only one signal which corresponded to IPS2 (Fig. 5C). The intensity of the signal through the fractions suggested that two populations of HEG RNAs with this end exist in the Fig. 4. Characterization of the intronic part of the linear 7.5-kb RNA. (A) Schematic presentation of the Dir.S956-1 intron, showing the position of the 5¢ and 3¢ splice sites (SS) as well as the internal pro- cessing sites (IPS). The exon, open reading frame and intron sequences are shown in black, grey and white, respectively. The localization of the probes used for Northern blotting analysis are indicated with thick black lines. (B) Northern blotting analysis of whole cell RNA. Parallel filters containing RNA from an early (E) and a late (L) time point (corresponding to positions 1 and 6 in Fig. 2A) were hybridized with the indicated GIR1, HEG and GIR2 probes. The G319 RNA marker (Promega) was used as a size marker. Ó FEBS 2002 Ribozyme mediated processing of pre-rRNA (Eur. J. Biochem. 269) 5809 cells. As expected, IPS2-terminated RNAs were found in the low molecular mass fractions where the processed forms of the excised Dir.S956-1 (1.23-kb RNA and 0.85- kb RNA; see Fig. 2) are located. Another IPS2-termin- ated HEG RNA exists in the high molecular fractions where the 7.5-kb RNA is the predominant HEG RNA. This implies that the 5¢ end of the 7.5-kb RNA corresponds to IPS2. While DiGIR1 cleaves at two processing sites (IPS1 and IPS2) in an obligate sequential order in vitro [6], only IPS2-cleaved RNA has been detected in vivo [7]. We infer that 5¢ end formation of the 7.5-kb RNA is a result of DiGIR1 catalysis. The critical involvement of DiGIR1 in the formation of the 5¢ end of the RNA could be tested by the introduction of mutations in the catalytic site of the ribozyme [3]. Unfortunately, a transformation protocol for Didymium is currently not available. The 7.5-kb RNA is located in the nucleus Previous studies have shown that the different I-DirIHEG RNAs differ in their intracellular distribution. While the fully processed 0.85-kb RNA is located almost exclusively in the cytosol, all other examined HEG RNAs (3.9 kb, 1.46 kb, and 1.23 kb RNA, respectively; see Fig. 2) are nuclear [7]. In order to examine the subcellular location of the 7.5-kb RNA, Didymium Lat3-5 cells from a late time point (corresponding to 6 in Fig. 2A) were lysed and fractionated by centrifugation and RNA isolated from the nuclear and cytosolic fractions. Although we had some difficulties recovering large RNAs after the fractionation procedure, probably due to high nuclease activity in the encysting cells, the resulting Northern blot indicated a nuclear localization for the 7.5-kb RNA (Fig. 6). As expected, the 0.85-kb RNA was the only HEG RNA species to be found in the cytosol. As such, it provided an internal control for the success of the fractionation procedure. DISCUSSION We have previously noted the presence of an I-DirIHEG RNA signal corresponding to 7.5 kb when D. iridis Lat3-5 RNA was analysed by Northern blotting [7]. We show here that this RNA is a linear species, produced by alternative processing of the pre-rRNA. The 5¢ end of the RNA corresponds to an internal processing site in Dir.S956-1 (IPS2), implying that it is formed through cleavage by the DiGIR1 ribozyme. The 3¢ end has not been mapped but is assumed to correspond approximately to the 3¢ end of the pre-rRNA. Apart from two group I introns in the LSU rRNA (Dir.L1949 and Dir.L2449), the 7.5-kb RNA Fig. 6. Intracellular localization of the 7.5-kb RNA. Total(T),nuclear (N) and cytosolic (C) RNA from 5 · 10 5 Lat3-5 cells was run on a 1% denaturing agarose gel and analysed by hybridization using the HEG probe described in Fig. 3A. The identity of the observed signals is indicated on the right. The size indications are derived from the High Range RNA ladder (Fermentas). Fig. 5. Mapping of the 5¢ end of the linear 7.5-kb RNA. Whole cell RNA (250 lg) from a late time point (corresponding to position 6 in Fig. 2A) was fractionated on a 15–40% sucrose gradient. In addition to the pelleted material that had run through the gradient (P), 22 fractions were collected (1–22). (A) Denaturing agarose gel of fract- ionated RNA. RNA was recovered from the collected fractions and analysed on a 1% agarose gel stained with ethidium bromide. The positions of the SSU and LSU rRNAs are indicated. The 0.24- to 9.5- kb ladder (GibcoBRL) was used as a size marker (L). (B) Northern blotting analysis of fractionated RNA. The gel shown in (A) was analysed by hybridization using the HEG probe indicated in Fig. 4A. The positions of the 7.5-kb RNA and the processed forms of the intron (1.46-kb, 1.23-kb, and 0.85-kb RNA, respectively) are shown. The size indications are derived from the ladder shown in (A). (C) Primer extension analysis of fractionated RNA. RNA recovered from the collected fractions was analysed using OP4 (an 18-mer complementary to a sequence 35–52 nucleotides downstream of IPS2) as a primer. OP4 wasalsousedtomakeaDNAsequencingladderwhichwasusedto determine the exact position of the primer extension stop at IPS2 as indicated. 5810 A. Vader et al. (Eur. J. Biochem. 269) Ó FEBS 2002 contains all sequence elements downstream of IPS2 (see Fig. 1, right panel). These include the LSU rRNA exons, the internal transcribed spacers (ITS1 and ITS2) and the part of the SSU rRNA that is located downstream of Dir.S956-1. All intron sequences downstream of IPS2, including the spliceosomal intron within the I-DirIprotein coding region, are also present. The RNA is localized in the nucleus, and accumulates when Didymium cells are starved. In cells about to form dormant cysts, the 7.5-kb RNA is the predominant I-DirIHEGRNA.Whenviewedasa nonribosomal RNA, it is relatively abundant and remark- ably stable considering the turn-over of other RNAs, including ribosomal RNAs, that take place during encystment. Theexistenceofthe7.5-kbI-DirIHEGRNAis surprising, as the molecule contains several intrinsic activities and signals that would be expected to result in its disappearance. Firstly, even though the cleavage by DiGIR1 at IPS2 removes the 5¢ splice site site of Dir.S956-1, the catalytic core of the DiGIR2 splicing ribozyme remains intact. DiGIR2 RNAs with even larger truncations in the 5¢ endhavepreviouslybeenshownto perform efficient hydrolysis at the 3¢ splice site of Dir.S956-1 [5]. Secondly, an internal processing site (IPS3) is located downstream of the HEG region. The site is cleaved during maturation of the mRNA from the excised intron [7]. Thirdly, the 7.5-kb RNA contains a polyadenylation signal, which in mRNA formation indu- ces cleavage and polyadenylation of the I-DirImRNA. Finally, a spliceosomal intron (I51) harboured by the I-DirI HEG is removed in the mature I-DirI mRNA [7]. All of the activities mentioned above would be expected to act against the preservation of the 7.5-kb RNA. The composition of the 7.5-kb RNA suggests how it is formed and no new activities need to be postulated to account for its structure. Instead, its accumulation can be explained by an alteration of the relative rates of known processing activities. All of the experiments carried out in the present study aim at analysing the steady-state level of the RNAs involved. It is frequently observed that the processing of pre-rRNA and pre-mRNA slows down when cells are starved. In the ciliate Tetrahymena pyrifor- mis, the pre-rRNA processing rate has been reported to be decreased 36-fold during starvation [13] and a similar 12-fold decrease has been observed in T. thermophila [14]. As a result, precursors and processing intermediates tend to show increased steady-state levels under such conditions. In the present case, this could explain the inclusion of ITS1 and ITS2 in the 7.5-kb RNA, as well as the presence of the spliceosomal intron and the failure to use the polyadenylation signals. The observed nuclear localization of the 7.5-kb RNA is expected as the RNA retains several putative nuclear retention signals, e.g. a spliceosomal intron. The accumulation of the 7.5-kb RNA during starva- tion does not necessarily imply that this RNA is specific for starved cells. It is possible that the 7.5-kb RNA is a default intermediate in the formation of the I-DirI mRNA that is rapidly turned over during normal exponential growth and thus not detected by the methods applied in the present study. This would leave the excised intron as a dead end rather than as a precursor in the formation of I-DirI mRNA. We are currently unable to exclude this possibility but consider it unlikely in view of the high and comparable levels of the excised intron and precursors for the I-DirI mRNA found in exponentially growing cells [7]. In any case, the accumulation of the 7.5-kb RNA implies an alternative processing pathway for pre-rRNA. Interestingly, the relative activities of the splicing and cleavage ribozymes seem to be altered when the formation of the 7.5-kb RNA is compared to the formation of other HEG RNAs. In the formation of mature rRNA from pre- rRNA, excision of the optional Dir.S956-1 intron precedes that of the two conserved LSU introns. Furthermore, DiGIR1 seems to be active only after splicing by DiGIR2 has taken place [5,7,9]. In the 7.5-kb RNA, on the other hand, Dir.L1949 and Dir.L2449 have been removed, while the catalytic part of DiGIR2 is still present and DiGIR1 is active without prior splicing. In addition, we have recently discovered that full-length circles are formed in a reaction pathway catalysed by DiGIR2 and that DiGIR1 is com- pletely inactive in this pathway (unpublished data). These observations suggest that some mechanism is in operation to modulate the activities of DiGIR1 and DiGIR2 with respect to each other leading to different processing products. What can be the function of the 7.5-kb RNA? During cyst formation, cells actively degrade intracellular sub- stances. In a study of the myxomycete Physarum flavico- mum it was found that the content of protein, neutral hexose and RNA decreased by 40, 41 and 21%, respect- ively [15]. At the same time, excystment has been reported to occur even in the presence of 300 lgÆmL )1 actinomy- cin D, suggesting that new RNA synthesis is not required [16]. Rather, stable messengers are stored by the cysts in preparation for rapid emergence. Protein synthesis seems to be required for excystment [11,16], although these findings have been disputed [17]. Although it is quite possible that the 7.5-kb RNA is a dead-end product, it remains a possibility that it functions as a precursor for the I-DirI mRNA during excystment. Preliminary experi- ments aimed at following the 7.5-kb RNA through germination of cysts have failed due to lack of synchrony in cultures of germinating cysts. In conclusion, we have described a new RNA species that accumulates by alternative processing of pre-rRNA during starvation-induced encystment in Didymium.Themost interesting aspect of this RNA is perhaps the implications of its formation. It provides a new example of the cost of intron load on the host cell because the formation of this RNA is incompatible with the formation of ribosomal RNA from the same transcript. It shows DiGIR1 activity in a different RNA context than previously demonstrated and shows that the activities of the splicing and cleavage ribozymes of a twin ribozyme intron can be modulated with respect to each other resulting in different processing products. ACKNOWLEDGEMENTS We thank F. Frenzel for technical assistance and J. Christiansen for helpful suggestions to the experiments. This work was supported by grants from the Norwegian Research Council (AV), the Danish Research Council for Natural Sciences (HN), and The NOVO Foundation (HN). Ó FEBS 2002 Ribozyme mediated processing of pre-rRNA (Eur. J. Biochem. 269) 5811 REFERENCES 1. Cech, T.R. & Golden, B.L. (1999) In The RNA World. (Gesteland, R.F., Cech, T.R. & Atkins, J.F., eds), pp. 321–349. Cold Spring Harbor Press, New York. 2. Lambowitz, A.M. & Belfort, M. (1993) Introns as mobile genetic elements. Ann. Rev. Biochem. 62, 587–622. 3. Einvik, C., Elde, M. & Johansen, S. (1998) Group I twintrons: genetic elements in myxomycetes and schizopyrenid amoebo- flagellate ribosomal DNAs. J. Biotechnol. 64, 63–74. 4. Johansen, S. & Haugen, P. (2001) A new nomenclature of group I introns in ribosomal DNA. RNA 7, 935–936. 5. Decatur, W.A., Einvik, C., Johansen, S. & Vogt, V.M. (1995) Two group I ribozymes with different functions in a nuclear rDNA intron. EMBO J. 14, 4558–4568. 6. Einvik, C., Nielsen, H., Westhof, E., Michel, F. & Johansen, S. (1998) Group I-like ribozymes with a novel core organization perform obligate sequential hydrolytic cleavages at two processing sites. RNA 4, 530–541. 7. Vader, A., Nielsen, H. & Johansen, S. (1999) In vivo expression of the nucleolar group I intron-encoded I-DirI homing endonuclease involves the removal of a spliceosomal intron. EMBO J. 18, 1003– 1013. 8. Johansen, S., Elde, M., Vader, A., Haugen, P., Haugli, K. & Haugli, F. (1997) In vivo mobility of a group I twintron in nuclear ribosomal DNA of the myxomycete Didymium iridis. Mol. Microbiol. 24, 737–745. 9. Johansen, S. & Vogt, V.M. (1994) An intron in the nuclear ribo- somal DNA of Didymium iridis codes for a group I ribozyme and a novel ribozyme that cooperate in self-splicing. Cell 76, 725–734. 10. Vader, A. (1998) Nuclear group I introns of the myxomycetes; or- ganization, expression and evolution.PhDThesis.Universityof Tromsø, Norway. 11. Gorman, J.A., Dove, W.F. & Shaibe, E. (1977) Anisomycin sen- sitive mutants of Physarum polycephalum isolated by cyst selec- tion. Mol. Gen. Genet. 151, 253–259. 12. Din, N., Engberg, J., Kaffenberger, W. & Eckert, W.A. (1979) The intervening sequence in the 26S rRNA coding region in T. ther- mophila is transcribed within the largest stable precursor for rRNA. Cell 18, 525–532. 13. Eckert, W.A. & Kaffenberger, W. (1980) Regulation of rRNA metabolism in Tetrahymena pyriformis. I. Nutritional shift-down. Eur. J. Cell Biol. 21, 53–62. 14. Sutton, C.A., Sylvan, P. & Hallberg, R.L. (1979) Ribosome bio- synthesis in Tetrahymena thermophila. IV. Regulation of riboso- mal RNA synthesis in growing and growth arrested cells. J. Cell Physiol. 101, 503–513. 15. Henney, H.R. Jr & Chu, P. (1977) Differentiation of Physarum flavicomum: metabolic patterns and the role of amino acids in the control of encystment. Exp. Mycol. 1, 41–51. 16. Raub, T.J. & Aldrich, H.C. (1981) Ultrastructural events and kinetics of microcyst germination in the myxomycete Didymium iridis. Arch. Microbiol. 128, 384–389. 17. Haars, A., Zaar, K. & Huttermann, A. (1979) Induction of syn- chronous differentiation (encystment) in Physarum polycephalum myxamoeba. Arch. Microbiol. 120, 105–112. 18. Johansen, S., Johansen, T. & Haugli, F. (1992) Structure and evolution of myxomycete nuclear group I introns: a model for horizontal transfer by intron homing. Curr. Genet. 22, 297–304. 5812 A. Vader et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . alternative processing of pre-rRNA. The 5¢ end corresponds to an internal processing site cleaved by the group I-like ribozyme DiGIR1, located within the twin- ribozyme. of DiGIR1 catalysis. The critical involvement of DiGIR1 in the formation of the 5¢ end of the RNA could be tested by the introduction of mutations in the