Abstract

The second intron in the mitochondrial cytb gene of Saccharomyces capensis, belonging to group I, encodes a 280 amino acid protein containing two LAGLIDADG motifs. Genetic and molecular studies have previously shown that this protein has a dual function in the wild-type strain. It acts as a specific homing endonuclease I-ScaI promoting intron mobility and as a maturase promoting intron splicing. Here we describe the synthesis of a universal code equivalent to the mitochondrial sequence coding for this protein and the in vitro characterization of I-ScaI endonuclease activity, using a truncated mutant form of the protein p28bi2 produced in Escherichia coli. We have also determined the cleavage pattern as well as the recognition site of p28bi2. It was found that p28bi2 generates a double-strand cleavage downstream from the intron insertion site with 4 nt long 3-overhangs. Mutational analysis of the DNA target site shows that p28bi2 recognizes a 16–19 bp sequence from positions –11 to +8 with respect to the intron insertion site.

Received September 29, 1999; Revised and Accepted December 23, 1999.

INTRODUCTION

Homingendonucleases encoded by introns and inteins are enzymes able to promote the mobility of the DNA sequences which encode them, resulting in the transfer of these sequences to recipient allele genes lacking these genetic elements (14). This unidirectional transfer is initiated by an intron encoded protein which has a specific double-strand DNA endonuclease activity that cleaves an intronless recipient genome at (or in the vicinity of) the intron insertion site.

Homing endonucleases form four families according to specific motifs present in their sequence: LAGLIDADG, GIY-YIG, H-N-H and the His-cys box (5,6). Recently the crystallographic structures of four of these endonucleases have been determined providing the structural basis for understanding the functions of these molecules: PI-SceI (7), I-CreI (8), I-PpoI (9) and I-DmoI (10). I-PpoI belongs to the His-cys box group while the others are members of the LAGLIDADG family. The secondary structure of the catalytic domain of I-TevI, a member of the GIY-YIG family, has also recently been determined using NMR spectro­scopy (11).

Several proteins of the yeast mitochondrial genome, which are encoded by group I introns and which contain in their sequence two LAGLIDADG motifs, function in general either as maturases or as endonucleases (2). At present, two cases have been reported showing that the same molecule displays both enzymatic activities. This was demonstrated by genetic analysis for the protein encoded in the Saccharomyces capensis bi2 intron (12) and by in vitro studies for the protein encoded by the unique intron present in the cytb gene of Aspergillus nidulans (13). It has also been shown that the second function, either maturase or endonuclease, can be activated by mutation. It was demonstrated that the endo­nuclease I-SceII encoded by the ai4 intron of the Saccharomyces cerevisiaecoxI gene can gain a latent maturase activity after a single amino acid substitution in its sequence (14) and two amino acid replacements in the S.cerevisiae bi2 maturase lead to the acquisition of a homing endonuclease function (12).

In the present work we have studied the endonuclease activity of the bifunctional protein encoded by the S.capensis bi2 intron. The universal code equivalent of the mitochondrial gene coding the I-ScaI endonuclease was synthesized de novo and expressed in Escherichia coli. The I-ScaI activity was characterized in vitro using a recombinant truncated form of the protein. The specific DNA interactions and endonucleolytic properties of this bifunctional molecule are compared to those observed for the homing endonucleases previously studied.

MATERIALS AND METHODS

Synthesis of a universal code equivalent to the bi2 intronic ORF

Construction of the synthetic gene encoding the I-ScaI endo­nuclease was performed using a recursive PCR method described by Dillon and Rosen (15). A set of seven oligonucleo­tides, 76–85 nt long with an overlap of 15–23 nt was used for each strand synthesis. The position of the oligonucleotides used for each strand are given in Figure 1A and their sequences can be seen in Figure 1B. This sequence was designed using, where possible, the most commonly represented codons in E.coli (16); some minor codons were introduced in order to create restriction sites. The synthesis consisted of two steps. For the first step we used 0.5–1.5 µl from a mixture of the 14 oligonucleotides at a concentration 2 × 10–5 M each in a 100 µl reaction which contained 10 mM Tris–HCl pH 8.3, 1.5 mM MgCl2, 50 mM KCl, 0.2 mM dNTPs, 0.005% Tween-20, 0.005% NP-40, 0.01% BSA with 2.5 U Taq polymerase (Bioprobe Systems). The PCR program consisted of five cycles of 95°C for 40 s, 42°C for 60 s, 65°C for 5 s and 74°C for 120 s, followed by 20 cycles of 95°C for 30 s, 50°C for 60 s, 65°C for 5 s and 74°C for 150 s. In the second step we used 0.5–1 µl from the previous step as template in a 50 µl reaction buffered as above with two flanking primers (pa1 and pb1) at concentrations of 1 µM each (Fig. 1A). The PCR program consisted of three cycles of 95°C for 40 s, 42°C for 60 s and 74°C for 150 s, followed by 30 cycles of 95°C for 30 s, 50°C for 60 s, 65°C for 5 s and 74°C for 150 s. The PCR reaction products were analyzed by 1% agarose gel electrophoresis. The 870 bp synthesized DNA was recovered from a 0.4% low melting point agarose gel electrophoresis (FMC BioProducts) following the manufacturer’s protocol. DNA was digested at the flanking restriction sites, KpnI and BamHI, ligated into the same sites of vector pBluescript II sk+ (Stratagene) and the recombinant plasmid was then used to transform DHα E.coli cells (Bethesda Research Laboratories). Among the clones obtained, 11 were sequenced on both strands using the Sequenase v.2 kit (US Biochemicals). The gene was reassembled from two clones and the final assembly, BS4552, was checked again by sequencing.

Construction of the I-ScaI truncated form p28bi2

The p28bi2 coding sequence was generated by removing the sequence at the 5′-terminus of the intronic ORF up to the StyI restriction site (Fig. 1A). The shortened sequence was obtained by the PCR method using BS4552 as template with two primers. The upstream primer had the sequence (from positions 122 to 139, Fig. 1B) corresponding to the beginning of the truncated variant with KpnI and NdeI restriction sites added at the 5′-end. The downstream primer was pb1, which includes the HindIII and BamHI cloning sites (Fig. 1A). The PCR-amplified DNA was digested with KpnI and BamHI and ligated into a pBluescript II sk+ vector open at the same sites. After transformation and cloning, the plasmid sequence between StyI and HindIII was exchanged with the equivalent part taken from BS4552 containing the exact sequence. Afterwards the plasmid was digested with NdeI and BamHI and the insert was recloned into the pET11 vector (Novagen) for expression in E.coli BL21/DE3.

Expression and purification of I-ScaI p28bi2

To produce the protein, 1 liter of LB medium with ampicillin (100 µg/ml) was inoculated with 1 ml of an overnight culture of pET11-p28bi2-transformed E.coli BL21/DE3. Cells were grown at 37°C until reaching an A600 value between 0.5 and 0.6. Cultures were transferred to 30°C and expression induced by addition of IPTG to 1 mM. Cells were grown for an additional 3 h, harvested by centrifugation and frozen at –75°C. When needed the culture pellet was thawed at 0°C and redispersed in 30 ml of 50 mM Tris–HCl pH 8, 1 mM EDTA, 25% (w/v) sucrose. Inclusion bodies were then extracted following the procedure described by Nagai and Thøgersen (17). The pellet obtained after this treatment was washed twice by redispersion and centrifugation in 50 mM Tris–HCl, pH 8, 8 M urea. The final pellet was then incubated in 6 ml of 6 M guanidine–HCl (GdmCl), 5 mM dl-dithiothreitol (DDT) overnight at 8°C. The extract was centrifuged at 20 000 g and the supernatant used in the following procedure. A sample of 1 ml of the 6 M GdmCl extract which had an A280 value in the range 20–30, depending on the preparation, was diluted 20-fold in buffer A (20 mM Na phosphate, pH 7, 8 M urea), the solution was centrifuged and the supernatant was mixed with 5 ml of S-Sepharose FF (Pharmacia Biotech) equilibrated in buffer A at 8°C. Urea was deionized by treatment with resin AG 501-X8 (Bio-Rad). The suspension was swirled for 15 min, poured into a column and the resin left to settle. The solution was allowed to flow through the resin which was then washed successively with 5 gel bed volumes of each of the following solutions: buffer B (20 mM Na phosphate, pH 7), buffer B + 1 M NaCl, buffer B + 2 M NaCl and buffer B. p28bi2 interacted very strongly with the resin and was still bound to it after the above treatment. The protein was released by washing the resin with 15 ml of 6 M GdmCl, which was reconcentrated to 1 ml using a centrifugal filter device (Ultrafree Biomax; Millipore). This protein solution, which had an A280 value between 7 and 10, was then chromatographed on a Superdex 200 (Pharmacia Biotech) column (1.6 × 60 cm) equilibrated in buffer C (50 mM Tris–HCl pH 7.5, 6 M GdmCl, 3 mM 2-mercaptoethanol) at a flow rate of 1 ml/min at 5°C. A major peak of absorbance at 280 nm was detected. The pool of fractions collected under the central region of this peak and which contained from 4.5 to 6 A280 units was used in the following renaturation procedure conducted at 4°C. The protein solution was adjusted to an A280 value of 0.1 in 6 M GdmCl, then rapidly diluted to 4 M GdmCl using water containing glycerol at 20% (v/v) and then dialyzed three times against 100 vol of 10 mM Na phosphate pH 7, 3 mM 2-mercapto­ethanol, 10% (v/v) glycerol, for 20 h each. Approximately 55% of the dialyzed material was recovered in the soluble fraction after centrifugation of the renatured protein solution at 35 000 g for 1 h. DDT was then added to a concentration of 2 mM and the preparation was stored at 4°C. The homogeneity of the protein preparation was evaluated by SDS–12% PAGE followed by staining with Coomassie brillant blue R250. The molar extinction coefficient of p28bi2 in 6 M GdmCl at 280 nm (5.18 × 104 M–1 cm–1) was established on the basis of the protein concentration determined using a tryptophan fluorescence titration method described by Pajot (18) and taking four tryptophan residues present in one molecule of the Mr 28 336 protein. The protein concentrations in 6 M GdmCl solution were evaluated by absorbance measurements at 280 nm. For diluted protein solutions in non-denaturing buffer, samples were first concentrated 10 times, before determining the concentration using the Bio-Rad (Bradford) protein assay procedure.

Determination of the cleavage site and mutagenese analysis

The cleavage site of p28bi2 was determined using a modified primer extension method (19). Plasmid pYGT28 (6 µg) containing the endonuclease target sequence (12) served as a double-strand template using the forward and reverse primers 1212 and 1233 (New England Biolabs) with Sequenase (Amersham). The reaction was either dideoxyterminated to generate a control ladder or extended with 80 µM dNTPs to produce [α-35S]-labeled substrate for endonuclease digestion. The latter reaction was stopped by adding 95% formamide, then phenolyzed and ethanol precipitated. The DNA was redissolved in water and used for the cleavage reaction with p28bi2. After phenol extraction and ethanol precipitation the digestion products were analyzed on a 6% polyacrylamide–urea sequencing gel alongside the sequence ladder.

For random mutagenese analysis, two oligonucleotides were synthesized containing degenerate endonuclease target site sequences using 90% of the bases present in the wild-type sequence with 3.3% for each of the other three bases. Their 5′-flanking sequences were designed to create 5′-overhangs for the BamHI and HindIII restriction sites. The mutagenized target sequences were inserted into pUC13 previously opened with BamHI and HindIII and the resulting recombinant plasmid used to transform DHα cells. Plasmids purified by CsCl gradient were sequenced using the 1212 and 1233 primers with the Sequenase v.2 kit, then linearized with SspI and used for the cleavage reaction with p28bi2.

I-ScaI p28bi2 endonuclease activity assays

The plasmid used for endonuclease reactions, pYGT28, is a pUC13 vector containing a 28 bp I-ScaI target site inserted between the BamHI and HindIII sites (12). The plasmid was first linearized by digestion with SspI to be used as substrate. For this plasmid 100 ng correspond to 0.057 pmol of the recognition sequence. The assays under the optimal conditions were performed in 60 µl of 50 mM Tris–HCl pH 8.7, 8 mM MgCl2, 45 mM NaCl, 4% (v/v) glycerol, 100 µg/ml BSA with 200–400 ng of DNA substrate at 30°C for 30 min. The reaction was stopped by addition of EDTA to 25 mM, then treated with proteinase K at 150 µg/ml for 30 min at room temperature. The reaction products were analyzed by 1% agarose gel electrophoresis, stained with BET after migration and the cleavage efficiencies determined as described by Monteilhet (20). The optimal conditions for DNA cleavage were determined by assays where 200 ng of DNA substrate were digested with 2 U p28bi2 for 30 min in 60 µl of the standard solution at various temperatures or at 30°C in the standard solution containing increasing amounts of NaCl (or KCl) or MgCl2 (or MnCl2) to analyze the effects of different concentrations of monovalent and divalent cations on the reaction. The following buffer systems were used at a concentration of 50 mM to analyze the effect of pH on the reaction, imidazol–HCl between pH 6.2 and 8, Tris–HCl between pH 7.2 and 9 and diethanolamine–HCl between pH 8 and 10. One unit of activity was taken as the amount of enzyme preparation required to cleave 200 ng of substrate to 50% completion under the conditions of the standard assay.

RESULTS

Synthesis of the universal code equivalent to the bi2 intronic ORF and construction of truncated forms of the synthetic gene

The genetic code in S.cerevisiae mitochondria differs from the universal code by the use of the stop codon TGA to specify tryptophan (21), the use of the isoleucine codon ATA to specify methionine (22) and the use of the leucine codon family CTN to specify threonine (23). The bi2 intron ORF contains 26 non-universal codons (20 ATA, three TGA and three CTN) as well as a number of codons (~40% of all codons) that are very rarely used by the E.coli translation machinery. In order to produce the protein in E.coli, we first prepared a universal code equivalent to the mitochondrial ORF employing the codons used most frequently in E.coli. Since a total of 204 changes distributed over the entire sequence of the bi2 ORF would have to be introduced (Fig. 1B), we instead synthesized the entire gene by a PCR method as described in Materials and Methods. Among the clones obtained from the synthesized DNA, 11 were sequenced on both strands. The observed error rate was 1.1%. The gene was reassembled by ligating the KpnI–MluI fragment taken from clone BS45, which contains two mutations downstream from the MluI restriction site (Fig. 1B), and the MluI–BamHI fragment from clone BS52, containing a two base deletion upstream from MluI. The final assembly, BS4552, was checked again and found to be error free.

Because of the low level of protein expression obtained with the complete ORF (see below), a truncated form of the protein, p28bi2, was constructed, following the strategy described in Materials and Methods. p28bi2, which produced protein more efficiently in E.coli, yields a protein of 28 kDa corresponding to 243 amino acids of the C-terminal region of the bi2 intron encoded protein. Another truncated form of the protein, p25bi2, was also constructed (Fig. 1B) to determine whether the shorter sequence also maintained the same specific activity. p25bi2, which gives a protein of 25 kDa corresponding to 213 amino acids of the C-terminal region of the bi2 intron encoded protein, was also expressed in E.coli and purified (Fig. 2).

As both p28bi2 and p25bi2 were found to be active as specific endonucleases, we decided to use only the longest form, p28bi2, for the first characterization of I-ScaI endo­nuclease activity by in vitro studies.

Expression and purification of I-ScaI p28bi2

The T7 expression system with the pET11 plasmid were used to produce the recombinant proteins in E.coli BL21/DE3. A low level of production was observed when the entire synthetic gene was used. This can be explained by the fact that the N-terminal region of the encoded protein contains a long stretch of hydrophobic amino acids (Ile13–Met30). This feature has been implicated in toxic effects towards the host and results in low yields of recombinant protein. To check the effect of removing this part of the sequence, a construction in which the first 36 codons at the 5′-terminus of the ORF were missing was used to produce a shortened form of the protein (p28bi2). When used in the same expression system, a significant improvment in the production of protein, ~5- to 10-fold, was observed.

Purification of I-ScaI was performed according to the protocol described in Materials and Methods using E.coli BL21/DE3 transformed with pET11-p28bi2. The homogeneity of the protein preparations is shown in Figure 2. The recombinant protein was produced as insoluble aggregates which could be resolubilized in GdmCl. No specific endonuclease activity was detected in the soluble fraction following cell extraction and no improvment in the solubility of p28bi2 was observed by fusing the protein to thioredoxin in vector pET32. As p28bi2 was found to interact very strongly with the chromatographic resins under non-denaturing conditions, all the chromatographic procedures were conducted in the presence of denaturing agents. The preparations of renatured p28bi2 were all free of non-specific nucleolytic activities and, when stored at 5°C, the specific endonuclease activity was found to be stable for several weeks. Approximately 40 000 U of specific activity could be obtained from a 1 liter culture.

Characterization of the I-ScaI p28bi2 endonuclase activity

Plasmid pYGT28, containing the I-ScaI recognition site, was linearized with SspI and used as substrate so that specific cleavage would generate two bands at 2094 and 613 bp (Fig. 3). The effect of temperature and pH and the nature of divalent and monovalent cations were analyzed in order to determine the conditions for optimal cleavage.

Magnesium (or manganese) was required for cleavage to occur (Fig. 3). The optimal concentration of the divalent cation was 8 mM and the use of higher concentrations of Mg2+ (or Mn2+) was progressively detrimental to the cleavage reaction, with no cleavage occuring at or above 30 mM (Fig. 4A). Several other divalent cations were tested as potential cofactors (Ni2+, Co2+, Cu2+, Zn2+ and Ca2+), but none was found to be active. The optimal activity was observed at temperatures between 28 and 40°C (Fig. 4B) and no cleavage was observed below 10°C on the time scale of the standard assays. Nevertheless, we observed that the cleavage reaction was able to proceed at 5°C, albeit slowly, with optimal cleavage obtained after 25–30 h incubation (data not shown). The presence of NaCl was found to stimulate the reaction with an optimal effect at 45 mM. Identical results were observed when using KCl instead of NaCl (Fig. 4C). The effect of pH was found to depend upon the nature of the divalent cation. With Mg2+, the activity was stimulated at alkaline pH with an optimum between pH 8.5 and 9, followed by a rapid decrease above pH 9 (Fig. 4D). The level of activity was relatively low (15%) at neutral pH. When Mn2+ was used as cofactor, activity was observed throughout a broad range of pH, between 6.2 and 8.5, with an optimum at pH 7.3 (data not shown). The main effect observed with Mn2+ was a reduced specificity of p28bi2. Several mutants of the recognition site (see Fig. 8), T(–6)A, C(+4)G, T(+5)G, A(+6)T and A(+6)G, which were not efficiently cleavable with Mg2+, could be cleaved nearly as well as the wild-type sequence in the presence of Mn2+.

Although the linearized plasmid, pYGT28, was routinely used as substrate, the supercoiled form of the plasmid was also efficiently cleaved. The results of the comparative assays presented in Figure 5 show that supercoiled pYGT28 was cleaved 2–2.5 times faster than the linearized plasmid.

Complete cleavage of the substrate by p28bi2 was not observed in any of the assays. The optimal cleavage efficiencies of the linearized plasmid, depending upon the enzyme preparation, were between 75 and 90%. p28bi2 proved to be very unstable when exposed to temperatures >15°C. When incubated alone at 30°C in the reaction buffer and in the absence of magnesium, 85% of the original cleavage capacity of the enzyme was lost after 10 min (Fig. 6). Similar protection of the protein against inactivation was observed when p28bi2 was incubated with identical concentrations of DNA substrate (pYGT28) or with non-specific DNA (plasmid pET22 or oligonucleotide duplex) or with a non-specific RNA stem–loop (Fig. 6), indicating that the stabilization was mainly the result of non-specific interactions.

Determination of the cleavage site and recognition sequence of I-ScaI p28bi2

The location of cleavage effected by p28bi2 on each strand of the recognition site was determined by the method of primer extension (19). The products of the primer extension were digested with p28bi2 and analyzed as described in Materials and Methods. As shown in Figure 7, a single major band is observed for each strand and no band can be seen at the same position for the non-digested sample. Cleavage occurs at 5 and 1 bp downstream from the intron insertion site for the coding and non-coding strand, respectively. The cleavage generates 3′-overhangs of 4 nt. When supercoiled pYGT28 was digested with p28bi2 the two cohesive ends created by the specific cleavage could be religated using T4 DNA ligase (data not shown).

It was shown in a previous study that a 28 bp sequence extending from position –14 to +14, in relation to the intron insertion site, could be cleaved specifically using mitochondrial extracts from the S.capensis wild-typestrain YB4234 (12). Mutations were generated in this sequence by random mutagenesis and also by introducing mutations at specific positions. A total of 112 clones obtained from the random mutagenesis have been sequenced. Among them, 11 were found to have the wild-type sequence, 20 had a single mutation, 42 were double mutants and the others contained three or more substitutions or had short deletions. To establish the cleavage site of the I-ScaI endonuclease, 16 single, 26 double and one deletion mutant obtained from random mutagenesis were used. Furthermore, this analysis was completed by the construction of directed mutations, among which five were localized at the 3′-end and 12 others at the 5′-end of the target site. All the mutagenized sequences were used as substrates in standard assays with 2 U of purified p28bi2. The results are summarized in Figure 8 and show that mutations which diminish or block cleavage by p28bi2 are concentrated in a 16 bp region between positions –6 and +10 with respect to the intron insertion site, whereas mutations localized outside the –8 to +11 region have no effect. For each position in the 16 bp region, there is at least one mutant which affects the cleavage reaction and for eight of these positions (–5, –4, –3, –2, +1, +2, +7 and +8) cleavage can be completely blocked. Positions near the intron insertion site appear to be critical, as a majority of the mutations blocking cleavage are found between positions –3 and +2.

DISCUSSION

We have demonstrated in this paper that E.coli recombinant protein encoded by the second intron of the mitochondrial cytb gene of S.capensis can recognize invitro a sequence encompassing the cytb gene exon B2–exon B3 junction and catalyze a double-strand break of the DNA close to the intron insertion site. We have constructed a truncated form of I-ScaI endonuclease, p28bi2, which is shortened by 36 residues at the N-terminal part of I-ScaI encoded by intron bi2 and have shown its enzymatic properties. Thus, it appears that active molecules endowed with specific endonucleolytic activity can be formed in the absence of the N-terminal region of the intron encoded sequence. Also, it would suggest that in vivo active I-ScaI molecules might be a result of proteolytic processing from a precursor, since the mitochondrially translated protein is likely to be a hybrid upstream exons–intron product. This type of processing was shown to occur in the case of the S.cerevisiae bi4 intron encoded maturase in which the coding sequence is also fused in-frame with the upstream exons of the cytb gene (2427).

The characteristics of the cleavage reaction catalyzed by p28bi2 are reminiscent of those already reported for other homing endonucleases belonging to the LAGLIDADG family. The similarities are a requirement for Mg2+ (or Mn2+) for cleavage to take place, optimal activity at alkaline pH and stimulation of the reaction by moderate concentrations of the monovalent cation NaCl (or KCl). Manganese, when used as a cofactor, was found to relax the specificity of p28bi2, as observed in the case of the intein encoded PI-SceI (28) and the HO endonuclease (29). The properties associated with manganese, and initially reported for a type II restriction enzyme (30), may be related to the results of numerous studies showing that manganese and other metals of the transition series interact more strongly than the alkaline earth metals with nucleotides, inducing destabilizing effects on the DNA structure (31). The structure of the DNA embedding the p28bi2 target site also appears to be a factor which influences the cleavage reaction. The enzyme was found to work faster if the recognition site was contained in a supercoiled plasmid as compared to the linearized form, a characteristic previously reported for PI-SceI (32).

p28bi2 catalyzes a double-strand cleavage in the vicinity of the intron insertion site, creating 3′-overhangs of 4 nt, a feature common to all the LAGLIDADG homing endonucleases studied so far. The cleavage occurs 5 and 1 nt dowstream from the intron insertion site for the coding and non-coding strands, respectively, a pattern also observed for I-CeuI (33), I-CpaII (34), I-CreI (35), I-PorI (36) and I-SceII (37). p28bi2 recognizes an asymmetrical 16–19 bp sequence spanning the intron insertion site. The mutational analysis of the I-ScaI target site presented in this work shows that all mutants impair, to various degrees, the cleavage reaction. A high proportion of the 15 mutations which completely block cleavage are concentrated near the intron insertion site, suggesting that this region is critical for specific recognition and/or cleavage. These results also indicate that the sequence requirement of p28bi2 is rather stringent, as is observed for I-SceI (38) and PI-SceI (39).

p28bi2 was found to give incomplete cleavage of its DNA substrate. This could be due to the instability of p28bi2 when not bound to its substrate or it could reflect slow release of the DNA cleavage products, as was demonstrated in the case of I-SceI (40). We also observed that p28bi2 was equally well protected against inactivation by DNA containing the recognition site or non-specific DNA or RNA, which indicates that p28bi2 also binds non-specifically to nucleic acids, as was observed in the case of I-CreI (41).

Although the properties associated with the activity of I-ScaI p28bi2, described in this work, are very similar to those described for other homing endonucleases of the LAGLIDADG family, it has been shown previously that in living cells the S.capensis bi2 intron encoded protein not only ensures intron propagation but also functions as an RNA maturase promoting its own intron splicing (12). It remains to be seen whether the purified form of p28bi2 produced in E.coli has kept or lost maturase activity. Further study of the S.capensis bi2 intron encoded protein, which displays both enzymatic activities in the same molecule, should contribute to progress in understanding the link between endonuclease and maturase functions.

ACKNOWLEDGEMENTS

We wish to thankL. Aggerbeck for helping with the chromato­grahic purifications, for critical reading of the manuscript and for correcting the English. We also thank B. Sargueil for the generous gift of RNA samples and D. Menay for the synthesis of oligonucleotides. D.D. was the recipient of a short post-doctoral fellowship from the CNRS within the framework of the Jumelage Franco-Polonais.

*

To whom correspondence should be addressed. Tel: +33 1 69 82 31 87; Fax: +33 1 69 07 55 39; Email: monteilhet@cgm.cnrs-gif.fr

Figure 1. Strategy for synthesis and sequence of a universal code equivalent for the ORF encoded by the second intron of the cytb gene of S.capensis. (A) (Top) Schematic of the orientation and overlapping of the oligonucleotides used for ORF synthesis; the 5′- and 3′-ends of the oligonucleotides are numbered according to their positions in the synthesized sequence presented in (B). (Bottom) Positions of the unique restriction sites present in the synthesized ORF. (B) Sequence of the coding strand with the sequence of the translated protein in the one letter code (amino acids in bold were specified by the non-universal codons in the mitochondrial sequence); restriction sites in the flanking regions used for cloning are in italic; changes introduced in the codons relative to the mitochondrial sequence are shown in lower case; brackets indicate the 5′-end of the sequences used to construct the p28bi2 and p25bi2 truncated forms of the endonuclease; the two conserved dodecapeptide (or LAGLIDADG) motifs are underlined.

Figure 1. Strategy for synthesis and sequence of a universal code equivalent for the ORF encoded by the second intron of the cytb gene of S.capensis. (A) (Top) Schematic of the orientation and overlapping of the oligonucleotides used for ORF synthesis; the 5′- and 3′-ends of the oligonucleotides are numbered according to their positions in the synthesized sequence presented in (B). (Bottom) Positions of the unique restriction sites present in the synthesized ORF. (B) Sequence of the coding strand with the sequence of the translated protein in the one letter code (amino acids in bold were specified by the non-universal codons in the mitochondrial sequence); restriction sites in the flanking regions used for cloning are in italic; changes introduced in the codons relative to the mitochondrial sequence are shown in lower case; brackets indicate the 5′-end of the sequences used to construct the p28bi2 and p25bi2 truncated forms of the endonuclease; the two conserved dodecapeptide (or LAGLIDADG) motifs are underlined.

Figure 2. SDS–PAGE of purified p28bi2 and p25bi2 truncated forms of the I-ScaI endonuclease.

Figure 2. SDS–PAGE of purified p28bi2 and p25bi2 truncated forms of the I-ScaI endonuclease.

Figure 3. p28bi2 endonuclease activity assays. (A) Structure of the plasmid pYGT28 used as substrate. pYGT28 was linearized with SspI before cleavage by p28bi2, which generates two DNA fragments of 2094 and 613 bp, respectively. (B) Lanes 1–4, 200 ng of substrate were incubated in the standard solution at 30°C for 30 min with 0.5, 1.0, 1.5 and 2 U p28bi2, respectively; lane 0, conditions as in lane 4 but Mg2+ was omitted. Optimal cleavage efficiency 90% (lane 4).

Figure 3. p28bi2 endonuclease activity assays. (A) Structure of the plasmid pYGT28 used as substrate. pYGT28 was linearized with SspI before cleavage by p28bi2, which generates two DNA fragments of 2094 and 613 bp, respectively. (B) Lanes 1–4, 200 ng of substrate were incubated in the standard solution at 30°C for 30 min with 0.5, 1.0, 1.5 and 2 U p28bi2, respectively; lane 0, conditions as in lane 4 but Mg2+ was omitted. Optimal cleavage efficiency 90% (lane 4).

Figure 4. Determination of the optimal conditions for substrate cleavage by p28bi2. The graph shows the effects of (A) temperature, (B) pH, (C) MgCl2 concentration and (D) NaCl concentration on the cleavage reaction. The conditions of the assays are described in Materials and Methods.

Figure 4. Determination of the optimal conditions for substrate cleavage by p28bi2. The graph shows the effects of (A) temperature, (B) pH, (C) MgCl2 concentration and (D) NaCl concentration on the cleavage reaction. The conditions of the assays are described in Materials and Methods.

Figure 5. Velocity of cleavage reactions using the supercoiled or the linearized substrate. Supercoiled (open circles) or linearized (closed circles) pYGT28 (2.5 µg) was incubated with 10 U p28bi2 in 400 µl of standard solution at 30°C. Aliquots (40 µl) were taken at selected time points, the reactions were stopped and the products analyzed as described in Materials and Methods.

Figure 5. Velocity of cleavage reactions using the supercoiled or the linearized substrate. Supercoiled (open circles) or linearized (closed circles) pYGT28 (2.5 µg) was incubated with 10 U p28bi2 in 400 µl of standard solution at 30°C. Aliquots (40 µl) were taken at selected time points, the reactions were stopped and the products analyzed as described in Materials and Methods.

Figure 6. Stabilization of p28bi2 by interactions with DNA or RNA. p28bi2 (2 U) was incubated at 30°C for 10 min in the standard solution minus Mg2+ either without DNA or with 300 ng of the specific or non-specific substrate. After this preincubation, 8 mM Mg2+ and 300 ng of substrate pGYT28 were added to all the assays except the reaction shown in lane 3, where only the addition of Mg2+ was necessary, then assays were incubated for a further 30 min at 30°C, stopped and analyzed. Lane 1, standard reaction shown as reference (300 ng of pYGT28 substrate and 2 U enzyme); lane 2, preincubation without DNA; lane 3, with pYGT28; lane 4, with linearized plasmid pET22 (5.4 kb); lane 5, with the 30 nt long oligoduplex; lane 6, with the 26 nt RNA stem–loop.

Figure 6. Stabilization of p28bi2 by interactions with DNA or RNA. p28bi2 (2 U) was incubated at 30°C for 10 min in the standard solution minus Mg2+ either without DNA or with 300 ng of the specific or non-specific substrate. After this preincubation, 8 mM Mg2+ and 300 ng of substrate pGYT28 were added to all the assays except the reaction shown in lane 3, where only the addition of Mg2+ was necessary, then assays were incubated for a further 30 min at 30°C, stopped and analyzed. Lane 1, standard reaction shown as reference (300 ng of pYGT28 substrate and 2 U enzyme); lane 2, preincubation without DNA; lane 3, with pYGT28; lane 4, with linearized plasmid pET22 (5.4 kb); lane 5, with the 30 nt long oligoduplex; lane 6, with the 26 nt RNA stem–loop.

Figure 7. Determination of the cleavage site. Autoradiogram showing the extension polymerization products digested or not with p28bi2 and the corresponding sequence ladder. The cleavage sites are indicated by arrows.

Figure 7. Determination of the cleavage site. Autoradiogram showing the extension polymerization products digested or not with p28bi2 and the corresponding sequence ladder. The cleavage sites are indicated by arrows.

Figure 8. Mutational analysis of the p28bi2 recognition sequence. The wild-type sequence of the target site is shown on top with the cleavage site indicated by a staggered line and the intron insertion site by an arrow. The three possible base substitutions for each position are indicated on the left side of the grid. closed circles, the wild-type sequence; +, mutant cleaved as well as the wild-type; ε, reduced cleavage; –, no cleavage.

Figure 8. Mutational analysis of the p28bi2 recognition sequence. The wild-type sequence of the target site is shown on top with the cleavage site indicated by a staggered line and the intron insertion site by an arrow. The three possible base substitutions for each position are indicated on the left side of the grid. closed circles, the wild-type sequence; +, mutant cleaved as well as the wild-type; ε, reduced cleavage; –, no cleavage.

References

1 Dujon,B. (
1989
)
Gene
 ,
82
,
91
–114.
2 Lambowitz,A.M. and Belfort,M. (
1993
)
Annu. Rev. Biochem.
 ,
62
,
587
–622.
3 Belfort,M. and Perlman,P.S. (
1995
)
J. Biol. Chem.
 ,
270
,
30237
–30240.
4 Cooper,A.A. and Stevens,T.H. (
1995
)
Trends Biochem. Sci.
 ,
20
,
351
–356.
5 Belfort,M. and Robert,R.J. (
1997
)
Nucleic Acids Res.
 ,
25
,
3379
–3388.
6 Jurica,M.S. and Stoddard,B.L. (
1999
)
Cell. Mol. Life Sci.
 ,
55
,
1304
–1326.
7 Duan,X., Gimble,F.S. and Quiocho,F.A. (
1997
)
Cell
 ,
89
,
555
–564.
8 Health,P.J., Stephens,K.M., Monnat,R.J. and Stoddard,B.L. (
1997
)
Nature Struct. Biol.
 ,
4
,
468
–476.
9 Flick,K.E., Jurica,M.S., Monnat,R.J. and Stoddard,B.L. (
1998
)
Nature
 ,
394
,
96
–101.
10 Silva,G.H., Dalgaard,J.Z., Belfort,M. and Van Roey,P.V. (
1999
)
J. Mol. Biol.
 ,
286
,
1123
–1136.
11 Kovalski,J.C., Belfort,M., Stapleton,M.A., Holpert,M., Dansereau,J.T., Pietrokovski,S., Baxter,S.M. and Derbyshire,V. (
1999
)
Nucleic Acids Res.
 ,
27
,
2115
–2125.
12 Szczepanek,T. and Lazowska,J. (
1996
)
EMBO J.
 ,
15
,
3758
–3767.
13 Ho,Y., Kim,S.J. and Waring,R.B. (
1997
)
Proc. Natl Acad. Sci. USA
 ,
94
,
8994
–8999.
14 Dujardin,G., Jacq,C. and Slonimski,P.P. (
1982
)
Nature
 ,
298
,
628
–632.
15 Dillon,P.J. and Rosen,C.A. (
1993
) In White,B.A. (ed.), PCR Protocols: Current Methods and Applications. Humana Press, Totowa, NJ, Vol. 15, pp.
263
–267.
16 Wada,K., Wada,Y., Doi,H., Ishibashi,F., Gojobori,T. and Ikemura,T. (
1991
)
Nucleic Acids Res.
 ,
19
(suppl.),
1981
–1986.
17 Nagai,K. and Thøgersen,H.C. (
1987
)
Methods Enzymol.
 ,
153
,
461
–481.
18 Pajot,P. (
1976
)
Eur. J. Biochem.
 ,
63
,
263
–269.
19 Wenzlau,J.M., Saldanha,R.J., Butow,R.A. and Perlman,P.S. (
1989
)
Cell
 ,
56
,
421
–430.
20 Monteilhet,C., Perrin,A., Thierry,A., Colleaux,L. and Dujon,B. (
1990
)
Nucleic Acids Res.
 ,
18
,
1407
–1412.
21 Fox,T.D. (
1979
)
Proc. Natl Acad. Sci. USA
 ,
76
,
6534
–6538.
22 Hudspeth,M.E.S., Ainley,W.M., Schumard,D.S., Butow,R.A. and Grossman,L.I. (
1982
)
Cell
 ,
30
,
617
–626.
23 Li,M. and Tzagoloff,A. (
1979
)
Cell
 ,
18
,
47
–53.
24 De la Salle,H., Jacq,C. and Slonimski,P.P (
1982
)
Cell
 ,
28
,
721
–732.
25 Weiss-Brummer,B., Rödel,G., Schweyen,R.J. and Kandewitz,F. (
1982
)
Cell
 ,
29
,
527
–536.
26 Anziano,P.Q., Manson,D.K., Mahler,H.R. and Perlman,P.S. (
1982
)
Cell
 ,
30
,
925
–932.
27 Jacq,C., Banroques,J., Becam,A.M., Slonimski,P.P., Guiso,N. and Danchin,A. (
1984
)
EMBO J.
 ,
3
,
1567
–1572.
28 Gimble,F.S. and Thorner,J. (
1992
)
Nature
 ,
357
,
301
–306.
29 Kostriken,R. and Heffron,F. (
1984
)
Cold Spring Harbor Symp. Quant. Biol.
 ,
49
,
89
–96.
30 Hsu,M. and Berg,P. (
1978
)
Biochemistry
 ,
17
,
131
–138.
31 Duguid,J., Bloomfield,V.A., Benevides,J. and Thomas,G.J.,Jr (
1993
)
Biophys. J.
 ,
65
,
1916
–1928.
32 Wende,W., Grindl,W., Christ,F., Pingoud,A. and Pingoud,V. (
1996
)
Nucleic Acids Res.
 ,
24
,
4123
–4132.
33 Marshall,P. and Lemieux,C. (
1991
)
Gene
 ,
104
,
241
–245.
34 Turmel,M., Mercier,J.P., Côté,V., Otis,C. and Lemieux,C. (
1995
)
Nucleic Acids Res.
 ,
23
,
2519
–2525.
35 Dürrenberger,F. and Rochaix,J.D. (
1993
)
Mol. Gen. Genet.
 ,
236
,
409
–414.
36 Lykke-Andersen,J., Thi-Ngoc,H.P. and Garrett,R.A. (
1994
)
Nucleic Acids Res.
 ,
22
,
4583
–4590.
37 Delahodde,A., Goguel,V., Becam,A.M., Creusot,F., Perea,J., Banroques,J. and Jacq,C. (
1989
)
Cell
 ,
56
,
431
–441.
38 Colleaux,L., D’Auriol,L., Galibert,F. and Dujon,B. (
1988
)
Proc. Natl Acad. Sci. USA
 ,
85
,
6022
–6026.
39 Gimble,F.S. and Wang,J. (
1996
)
J. Mol. Biol.
 ,
263
,
163
–180.
40 Perrin,A., Buckle,M. and Dujon,B. (
1993
)
EMBO J.
 ,
12
,
2939
–2947.
41 Wang,J., Kim,H.H., Yuan,X. and Herrin,D.L. (
1997
)
Nucleic Acids Res.
 ,
25
,
3767
–3776.

Comments

0 Comments