A Y639F/H784A T7 RNA polymerase double mutant displays superior properties for synthesizing RNAs with non‐canonical NTPsPadilla, Robert;Sousa, Rui
doi: 10.1093/nar/gnf138pmid: 12490729
Abstract A T7 RNA polymerase in which Tyr639 is mutated to Phe readily utilizes 2′‐deoxy, 2′‐NH2 and 2′‐F NTPs as substrates and has been widely used to synthesize modified RNAs for a variety of applications. This mutant does not readily utilize NTPs with bulkier 2′‐substituents, nor does it facilitate incorporation of NTPs with modifications at other positions. Introduction of a second mutation (H784A) into the Y639F background markedly enhances utilization of NTPs with bulky 2′‐substituents (2′‐OMe and 2′‐N3), and may also enhance use of NTPs with modifications at other than the 2′‐position. The Y639F/H784A double mutant may therefore be exceptionally useful for incorporation of a variety of non‐canonical NMPs into RNA. Received August 14, 2002; Revised October 4, 2002; Accepted October 20, 2002 INTRODUCTION The robust activity and strict promoter specificity of the single‐subunit phage RNA polymerases (RNAPs) makes them ideal for synthesizing specific RNAs in vitro (1). Mutation of phage T7 RNAP Tyr639 to Phe eliminates discrimination of the hydrogen‐bonding potential of the 2′‐substituent of the substrate NTP (2). This mutant enzyme can be used to incorporate 2′‐dNMPs, 2′‐F‐NMPs and 2′‐amino‐NMPs into transcripts (3) to make them RNase resistant (4), for studies of RNA structure–function relationships (5) or to expand the chemical repertoire of ribozymes. However, this mutant is not as useful for incorporating NMPs with bulkier substituents (2′‐OMe groups) into transcripts (6). The barrier to utilizing such substrates may not be in incorporating them into the RNA, but in subsequently extending the transcripts, i.e. many non‐canonical NMPs may act as effective chain terminators. The Y639F mutation does not overcome this barrier because the tyrosine is positioned so as to discriminate the structure of the substrate NTP (Fig. 1) (7), but is not involved in sensing the structure of the 3′‐NMP of the RNA. The crystal structure of a T7 RNAP transcription complex (7) reveals that the side chain of His784 occupies the minor groove side of the 3′‐rNMP:template base pair (Fig. 1). Though this histidine is well conserved in the phage RNAPs, an H784A mutant has near wild‐type activity but exhibits enhanced extension of RNAs with mispaired 3′‐termini (8). These observations suggest that the H784A mutation might relax the barrier to extension of transcripts containing non‐canonical NMPs at the 3′‐end of the RNA. In particular, replacement of the bulky histidine side chain with alanine might make room for extending RNAs with 3′‐NMPs carrying bulky minor groove substituents. MATERIALS AND METHODS RNAPs were prepared as described (8). Transcription reactions were carried out for 30 min at room temperature in 40 mM Tris–HCl pH 8.0, 6 mM MgCl2, 10 mM NaCl and 2 mM spermidine, with templates at 10–7 M and RNAPs at 2 × 10–7 M. All NTPs were used at 0.5 mM, and transcripts were labeled by inclusion of 1% (v/v) 800 Ci/mM [α‐32P]GTP in the reaction. Modified NTPs were from Trilink Biotechnologies. Reactions were stopped by addition of an equal volume of 90% formamide, 50 mM EDTA and 0.01% xylene cyanol, resolved by electrophoresis in 20% acrylamide, 1% bisacrylamide, 1× TBE gels, and visualized on a Molecular Dynamics phosphorimager. RESULTS We tested the ability of the wild‐type and the Y639F and Y639F/H784A mutants to utilize 2′‐OMe‐UTP, 2′‐OMe‐CTP, 2′‐azidoUTP and 2′‐azidoCTP in RNA synthesis. On a template [BglII cut pPK5 (9)] in which CTP and UTP are first incorporated at +14 and +15, respectively, the wild‐type enzyme can synthesize 34 nt run‐off transcripts in reactions in which UTP is replaced with 2′‐OMeUTP (Fig. 2A, lane 2) or in reactions in which CTP is replaced with 2′‐OMeCTP (Fig. 2A, lane 3). Relative to the reaction with four NTPs (Fig. 2A, lane 1), yields of run‐off transcripts are reduced 6‐ and 21‐fold in the 2′‐OMeUTP and 2′‐OMeCTP reactions, respectively. When both UTP and CTP are replaced with the corresponding 2′‐OMeNTPs, run‐off transcription with the wild‐type enzyme is undetectable (Fig. 2A, lane 4). Terminated transcripts are seen in these reactions at positions where 2′‐OMeUMP or 2′‐OMeCMP are incorporated (Fig. 2A, lanes 2–4). The Y639F mutation enhances utilization of the 2′‐OMeNTPs: relative to reactions with four NTPs (Fig. 2A, lane 5), run‐off transcription with Y639F is reduced, respectively, 2‐, 8‐ and 55‐fold in reactions with 2′‐OMeUTP or 2′‐OMeCTP or both (Fig. 2A, lanes 6–8). With the Y639F/H784A double mutant the reduction in run‐off transcript synthesis is <15% in reactions with 2′‐OMeUTP (Fig. 2A, lane 10) and only 3‐ and 6‐fold in reactions with, respectively, 2′‐OMeCTP or both 2′‐OMeUTP and 2′‐OMeCTP (Fig. 2A, lanes 11 and 12; data on the relative activity of the mutant enzymes with non‐canonical NTPs is summarized in Table 1). Identical experiments were carried out with 2′‐azidoUTP or 2′‐azidoCTP replacing UTP or CTP, respectively (Fig. 2B). With the wild‐type enzyme, run‐off transcription was reduced 15‐ and 50‐fold in reactions with 2′‐azidoUTP or 2′‐azidoCTP, respectively (Fig. 2B, lanes 2 and 3), and was undetectable in reactions with both 2′‐azidoUTP and 2′‐azidoCTP (Fig. 2B, lane 4). With Y639F, use of 2′‐azidoUTP or 2′‐azidoCTP reduced run‐off transcription by 70% (Fig. 2B, lanes 6 and 7) compared to reactions with four NTPs (Fig. 2B, lane 5), while use of both 2′‐azidoUTP and 2′‐azidoCTP reduced run‐off transcription by 11‐fold (Fig. 2B, lane 8). With Y639F/H784A use of a single azido‐modified NTP reduced run‐off transcription by only 40% (Fig. 2B, lanes 10 and 11), while use of two 2′‐azidoNTPs reduced it by 3‐fold (Fig. 2B, lane 12). With the pPK5 template UMP and CMP are not incorporated into the first 9 nt of RNA. Incorporation of non‐canonical NMPs during synthesis of the first 9 nt of RNA is expected to cause greater reductions in run‐off transcript synthesis because the RNA is readily released from the transcription complex during this initial phase of transcription. Therefore, if incorporation of a non‐canonical NMP is slow, the RNA will usually dissociate before the NMP is incorporated. Once the RNA is >9 nt in length the transcription complex becomes much more stable, and is unlikely to release the RNA during the time required for incorporation of a non‐canonical NMP. To see whether NMPs with bulky 2′‐substituents could be incorporated during the initial phase of transcription we used HindIII linearized pT75 (10). On this template, UMP is first incorporated at +14 while CMP is incorporated at +6 and +7. In reactions with 2′‐OMeUTP run‐off transcript synthesis is reduced 11‐, 5‐ and <2‐fold with the wild‐type and Y639F and Y639F/H784A mutants, respectively (Fig. 2C, lanes 2, 6 and 10). However, in reactions with 2′‐OMeCTP run‐off transcription is undetectable with the wild‐type or Y639F enzymes (Fig. 2C, lanes 3, 4, 7 and 8), and is reduced ∼700‐fold with Y639F/H784A (Fig. 2C, lanes 11 and 12). Run‐off transcript yields in reactions with HindIII cut pT75 and 2′‐azidoUTP are reduced 18‐, 8.5‐ and <2‐fold with the wild‐type and Y639F and Y639F/H784 mutants, respectively (Fig. 2D, lanes 2, 6 and 10), relative to reactions with four NTPs (Fig. 2D, lanes 1, 5 and 9). In reactions with 2′‐azidoCTP or both 2′‐azidoCTP and 2′‐azidoUTP, run‐off transcription is undetectable with the wild‐type enzyme (Fig. 2D, lanes 3 and 4), is reduced 65‐fold with Y639F (Fig. 2D, lanes 7 and 8), but is reduced only 7‐fold with Y639F/H784A (Fig. 2D, lanes 11 and 12). DISCUSSION By introducing an H784A substitution into a Y639F background, we obtain a polymerase with an enhanced ability to incorporate NMPs with bulky 2′‐substituents into RNA. In reactions with 2′‐OMe‐ or 2′‐azido‐modified NTPs yields of run‐off transcripts, relative to reactions with the four canonical NTPs, are markedly increased with the double mutant and premature termination products are greatly reduced or eliminated. Incorporation of the modified NMP is more efficient if it occurs during elongation, rather than during the poorly processive initiation phase of transcription when the RNA is <9 nt in length. However, the double mutant does incorporate azido‐NMPs efficiently, even during initial transcription (Fig. 2D, lanes 11 and 12). Given the position of the His784 side chain in the crystal structure of an initial transcript complex (7), it is likely that the effect of the H784A mutation is due primarily to enhanced extension of RNAs containing a modified NMP at the 3′‐terminus. This would be consistent with the observation that the H784A mutation enhances extension of RNAs with mispaired 3′‐termini, but does not increase incorporation of mispaired NMPs (8). This also suggests that the H784A mutant, either alone or in a Y639F background, might be useful in synthesizing RNAs with NTPs containing modifications at other than the 2′‐position. In this study we tested 2′‐OMe‐NTPs and 2′‐azidoNTPs because these are commercially available and are not readily used by either the wild‐type or Y639F polymerase. Evaluation of the utility of the H784A and Y639F/H784A mutants in synthesis with differently modified NTPs is probably best done on a case‐by‐case basis. ACKNOWLEDGEMENTS This work was supported by NIH grant GM52522 (to R. S.) and funds from the Welch Foundation. View largeDownload slide Figure 1. The templating base (cyan), 3′rNMP (red):template (blue) base pair and Tyr639 (yellow) and His784 (green) side chains as seen in the structure of a T7 RNAP initial transcription complex (PDB 1QLN) (7). View largeDownload slide Figure 1. The templating base (cyan), 3′rNMP (red):template (blue) base pair and Tyr639 (yellow) and His784 (green) side chains as seen in the structure of a T7 RNAP initial transcription complex (PDB 1QLN) (7). View largeDownload slide Figure 2. Incorporation of non‐canonical NMPs with the wild‐type, Y639F and Y639F/H784A enzymes. The NTPs (at 0.5 mM) present in each reaction are specified over each gel lane. (A) BglII cut pPK5 (9) as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. (B) BglII cut pPK5 as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. (C) HindIII cut pT75 (10) as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. (D) HindIII cut pT75 as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. The sequences of the transcripts obtained from the pPK5 and pT75 templates are presented to the left of the gels. View largeDownload slide Figure 2. Incorporation of non‐canonical NMPs with the wild‐type, Y639F and Y639F/H784A enzymes. The NTPs (at 0.5 mM) present in each reaction are specified over each gel lane. (A) BglII cut pPK5 (9) as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. (B) BglII cut pPK5 as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. (C) HindIII cut pT75 (10) as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. (D) HindIII cut pT75 as template. Lanes 1–4, wild‐type; lanes 5–8, Y639F; lanes 9–12, Y639F/H784A. The sequences of the transcripts obtained from the pPK5 and pT75 templates are presented to the left of the gels. Table 1. Activity of mutant RNAPs in reactions with 2′‐OMe‐ or 2′‐azidoNTPs Wild‐type Y639F Y639F/H784A Template: BglII cut pPK5 4 rNTPs 100 100 100 2′‐OmeU 17 50 90 2′‐OmeC 5 13 35 2′OMeC+2′OmeU n.d.a 2 17 2′‐AzU 6 30 60 2′‐AzC 2 30 60 2′AzU+2′AzC n.d. 9 35 Template: HindII‐cut pT75 4 rNTPs 100 100 100 2′‐OmeU 9 20 60 2′‐OmeC n.d. n.d. 0.14 2′‐AzU 6 12 60 2′‐AzC n.d. 1.5 14 2′AzU+2′AzC n.d. 1.5 14 Wild‐type Y639F Y639F/H784A Template: BglII cut pPK5 4 rNTPs 100 100 100 2′‐OmeU 17 50 90 2′‐OmeC 5 13 35 2′OMeC+2′OmeU n.d.a 2 17 2′‐AzU 6 30 60 2′‐AzC 2 30 60 2′AzU+2′AzC n.d. 9 35 Template: HindII‐cut pT75 4 rNTPs 100 100 100 2′‐OmeU 9 20 60 2′‐OmeC n.d. n.d. 0.14 2′‐AzU 6 12 60 2′‐AzC n.d. 1.5 14 2′AzU+2′AzC n.d. 1.5 14 ‘Activity’ reflects synthesis of full‐length run‐off transcripts and is normalized to that seen in reactions with four rNTPs which is assigned a value of 100 in arbitrary units. an.d., not detectable. View Large References 1. Milligan,J.F., Groebe,D.R., Witherell,G.W. and Uhlenbeck,O.C. ( 1987) Oligoribonucleotide synthesis using T7 RNA polymerase and synthetic DNA templates. Nucleic Acids Res. , 15, 8783–8798. Google Scholar 2. Sousa,R. and Padilla,R. ( 1995) A mutant T7 RNA polymerase as a DNA polymerase. EMBO J. , 14, 4609–4621. Google Scholar 3. Huang,Y., Eckstein,F., Padilla,R. and Sousa,R. ( 1997) Mechanism of ribose 2′‐group discrimination by an RNA polymerase. Biochemistry , 36, 8231–8242. Google Scholar 4. Padilla,R. and Sousa,R. ( 1999) Efficient synthesis of nucleic acids heavily modified with non‐canonical ribose 2′‐groups using a mutant T7 RNA polymerase. Nucleic Acids Res. , 27, 1561–1563. Google Scholar 5. Ryder,S.P., Ortoleva‐Donnelly,L., Kosek,A.B. and Strobel,S.A. ( 2000) Chemical probing of RNA by nucleotide analog interference mapping. Methods Enzymol. , 317, 92–109. Google Scholar 6. Sioud,M. and Sorensen,D.R. ( 1998) A nuclease‐resistant protein kinase C alpha ribozyme blocks glioma cell growth. Nat. Biotechnol. , 16, 556–561. Google Scholar 7. Cheetham,G.M.T. and Steitz,T.A. ( 1999) Structure of a transcribing T7 RNA polymerase initiation complex. Science , 286, 2305–2308. Google Scholar 8. Brieba,L.G. and Sousa,R. ( 2000) Roles of Y639 and H784 in ribose discrimination by T7 RNAP. Biochemistry , 39, 919–923. Google Scholar 9. Mentesanas,P.E., Chin‐Bow,S.T., Sousa,R. and McAllister,W.T. ( 2000) Characterization of bacteriophage T7 RNA polymerase elongation complexes. J. Mol. Biol. , 302, 1049–1062. Google Scholar 10. Tabor,S. and Richardson,C.C. ( 1985) A bacteriophage T7 RNA polymerase/promoter system for controlled exclusive expression of specific genes. Proc. Natl Acad. Sci. USA , 82, 1074–1078. Google Scholar
T7 RNA polymerase as a self‐replicating label for antigen quantificationTannous, Bakhos A.;Laios, Eleftheria;Christopoulos, Theodore K.
doi: 10.1093/nar/gnf140pmid: 12490731
Abstract Enzymes are used widely as labels in binding assays for protein analytes, because they provide signal amplification. Efforts at improving the assay sensitivity have been focused mainly on the synthesis of novel substrates, e.g. fluorogenic and chemiluminogenic ones. We report the investigation of T7 RNA polymerase (T7RP) as a label with unique characteristics for antigen quantification. In an in vitro, coupled (one‐step) transcription/translation reaction, T7RP catalyzes the expression of an enzyme‐coding DNA template to produce free enzyme (luciferase) in solution. We demonstrate that the generated luciferase is linearly related to the input T7RP in a range covering over four orders of magnitude. It is also shown that T7RP exhibits a significant level of self‐replication (100‐fold) in vitro by acting on a DNA template comprising the T7RP cDNA downstream of a T7 promoter. By combining the self‐replication reaction with the expression of luciferase DNA, as low as 1400 T7RP molecules are detectable. Furthermore, the T7RP is biotinylated, complexed with streptavidin and used for antigen quantification in a microtiter well‐based assay with high sensitivity and reproducibility. Received August 21, 2002; Revised and Accepted October 30, 2002 INTRODUCTION Whole‐genome sequencing projects have led to the identification of thousands of new genes. The challenge ahead is to unravel gene function and regulation on a genome‐wide scale. Most studies of gene function are based on the comparison of expression profiles between control and perturbed states, which allows for the identification of genes whose expression is induced or suppressed. DNA microarrays provide valuable information on gene expression at the mRNA level (1,2). Gene function, however, is manifested through the activity of the encoded protein. mRNA abundances do not always correlate with protein concentrations due to significant post‐translational regulation (3). Consequently, the direct quantitative analysis of proteins provides more accurate information about biological systems. Moreover, the comparison of protein expression profiles in patients and normal samples (differential profiling) reveals potential biomarkers for diagnosis, prognosis and monitoring of disease progression, as well as new therapeutic targets. The challenge, however, lies in the fact that proteins present at low concentrations are usually the ones that mediate the cellular response to various stimuli and are involved in the early stages of pathological processes. A recent study has shown that half of the yeast proteome was undetectable using two‐dimensional electrophoresis followed by mass spectrometry (4). Thus, high sensitivity, along with specificity, are essential requirements for any new technique in the field of proteomics, because they permit quantification of minute amounts of antigen and/or the use of smaller numbers of cells. Furthermore, these qualities must be combined with the ability for automation and high‐throughput protein analysis, in order to exploit the information provided by large‐scale sequencing projects. Target amplification techniques analogous to PCR that offer exquisite sensitivity to nucleic acid analysis are not available for protein analytes. The most sensitive protein assays are based on the interaction of the analyte with a specific binder (antibody, receptor or peptide) that is linked to a signal‐generating molecule (label). The assay sensitivity is determined, mainly, by the detectability of the label and the affinity of the binder. DNA fragments have been used as labels that provide signal amplification through replication [PCR (5) or rolling circle DNA replication (6)] or expression (7). However, the most widely used labels are enzymes (alkaline phosphatase, horseradish peroxidase, etc.) because they provide signal amplification through the turnover of many substrate molecules to detectable product. For almost 30 years, research efforts have been focused on the synthesis of novel substrates to allow more sensitive detection of enzyme labels. Thus, chromogenic substrates were gradually replaced by fluorogenic (8) and, more recently, chemiluminogenic ones (9). In contrast, this work introduces an enzyme label, T7 RNA polymerase (T7RP), which (i) has the unique ability to self‐replicate in vitro and (ii) catalyzes the in vitro synthesis of a second enzyme (firefly luciferase). The resulting signal amplification is due to the generation of many enzyme molecules in solution. The assay allows for antigen quantification with high sensitivity, wide dynamic range and very good reproducibility. Because it is performed in microtiter wells, it is amenable to automation and high‐throughput analysis. MATERIALS AND METHODS In vitro coupled transcription/translation The reaction mixture contained rabbit reticulocyte lysate (TNT) from Promega Corp., Madison, WI, supplemented with amino acids, but lacked T7RP. The appropriate DNA templates were added to the mixture. Determination of firefly luciferase A 2 µl aliquot of the transcription/translation reaction mixture was added to 50 µl of luciferase substrate buffer (20 mmol/l Tricine, pH 7.8, 1.1 mmol/l magnesium carbonate pentahydrate, 2.7 mmol/l MgSO4, 0.1 mmol/l EDTA, 33 mmol/l dithiothreitol, 270 µmol/l coenzyme A, 530 µmol/l ATP and 470 µmol/l luciferin) (10). The luminescence was monitored for 1 min using a liquid scintillation counter (model LS‐6500; Beckman Instruments, Fullerton, CA) in the single‐photon monitoring mode. Biotinylation of T7 RNA polymerase An aliquot of 1 mg (1.8 µmol) of sulfo‐N‐hydroxysuccinimide ester of biotin (NHS‐LC‐biotin; Pierce, Rockford, IL) was dissolved in 3 ml of dimethyl sulfoxide and then diluted to 15 µmol/l in T7RP buffer (20 mmol/l magnesium phosphate, pH 7.7, 0.1 mol/l NaCl, 1 mmol/l dithiothreitol and 1 mmol/l EDTA). A sample of 2 µl of the NHS‐LC‐biotin solution was mixed with 1 µl (6 pmol) of T7RP (Stratagene, La Jolla, CA) and incubated for 1 h at 4°C. The volume was increased to 50 µl with T7RP buffer containing 0.2 g/l bovine serum albumin (BSA). The biotinylated T7RP (BT7RP) was purified from free biotin by size exclusion chromatography using NAP columns (Amersham Pharmacia Biotech, Piscataway, NJ). The enzyme was eluted with 1 ml of sodium phosphate buffer pH 6.8. An aliquot of 100 µl of 10× concentrated T7RP buffer containing 1.4 g/l BSA was added to the purified BT7RP solution and the mixture was concentrated by ultrafiltration using microcon‐30 filters (mol. wt cut‐off = 30 000; Amicon, Beverly, MA) Preparation of streptavidin‐biotinylated T7 RNA polymerase complex (SA–BT7RP) Purified BT7RP (3 pmol) was mixed with 4.8 pmol streptavidin (Sigma, St Louis, MO), diluted in T7RP buffer (final volume 150 µl). The complexation reaction was allowed to proceed for 10 min at room temperature and the SA–BT7RP complex was used without purification. Biotinylation of monoclonal anti‐prostate‐specific antigen (PSA) antibody The monoclonal anti‐PSA antibody solution (catalog no. 8311; Diagnostic Systems Laboratories, Webster, TX) was dialyzed overnight against 3.5 l of 0.1 mol/l sodium bicarbonate at 4°C. A sample of 0.2 mg of the antibody was diluted with 0.5 mol/l carbonate buffer, pH 9.1, to a final concentration of 0.5 g/l. For biotinylation, 1 mg NHS‐LC‐biotin was dissolved in 50 µl of dimethyl sulfoxide and a 12.5 µl (0.25 mg) aliquot was added to the antibody solution. The mixture was incubated for 2 h at room temperature. The biotinylated antibody was stored at 4°C and used without purification. T7 RNA polymerase as a label for antigen quantification U‐bottom, polystyrene microtiter wells (Nunc Maxisorp; Life Technologies, Burlington, Ontario, Canada) were coated overnight at room temperature with 25 µl of 5 mg/l capture anti‐PSA antibody (catalog no. 8301; Diagnostic Systems Laboratories) diluted in 50 mmol/l Tris, pH 7.8, and 0.5 g/l NaN3. Before use, the wells were washed six times with wash solution (50 mmol/l Tris, pH 7.4, 150 mmol/l NaCl and 1 ml/l Tween‐20). A 10 µl aliquot of PSA standard (Scripps Laboratories, CA) diluted in 50 mmol/l Tris, pH 7.8, and 60 g/l BSA, along with 15 µl of 0.5 mg/l biotinylated anti‐PSA antibody, diluted in assay buffer (50 mmol/l Tris, pH 7.8, 60 g/l BSA, 0.5 mmol/l KCl, 0.5 g/l NaN3 and 0.5 g/l Triton X‐100), were added to each well. The immunoreaction was allowed to proceed for 1 h with continuous shaking. At the end of the incubation, any unbound biotinylated anti‐PSA antibody was removed by washing the wells six times as above. Afterwards, 25 µl of 2.4 nmol/l SA–BT7RP complex (diluted in T7RP buffer containing 1% fat‐free dry skim milk) was added to each well and incubated for 10 min. The wells were then washed six times followed by twice with 50 mmol/l potassium acetate. Subsequently, 25 µl of transcription/translation mixture containing 52.5 fmol luciferase cDNA (Luc‐DNA) (4.3 kb plasmid containing the T7RP promoter upstream from the firefly luciferase gene) was added to each well. The coupled in vitro transcription/translation reaction was allowed to proceed for 90 min at 30°C and the activity of synthesized firefly luciferase was measured as described above. Antigen quantification using a self‐replicating T7 RNA polymerase label The formation of the immunocomplex on microtiter wells and the binding of the SA–T7RP complex were carried out as described above. Subsequently, 23.5 µl of transcription/translation mixture containing 37.5 fmol T7RP‐DNA [plasmid pT7G1 (11), a kind gift from J. A. Wolff, Departments of Pediatrics and Medical Genetics, Waisman Center, Madison, WI] and 150 ng of salmon testes DNA (Sigma) was added to each well and incubated for 60 min (self‐replication phase). Afterwards, 1.5 µl of Luc‐DNA (26 fmol) was added to the wells and incubated for another 60 min (detection phase). The activity of synthesized firefly luciferase was measured as above. RESULTS AND DISCUSSION T7RP was chosen for this study because it is one of the simplest DNA‐dependent enzymes, capable of transcribing a complete gene without the need of additional proteins. Moreover, it is a single polypeptide chain (mol. wt 98 000), specific for its promoter. T7RP has been cloned and overexpressed in Escherichia coli (12). The principle of antigen quantification using T7RP as a label is illustrated in Figure 1. Two approaches for measuring T7RP, based on in vitro transcription/translation (with and without self‐replication), are also shown diagrammatically in Figure 1. Quantification of T7 RNA polymerase by coupled in vitro transcription and translation The goal of these experiments was to establish a relationship between the input T7RP and the synthesized protein in an in vitro transcription/translation system. The expression of firefly luciferase was chosen because this enzyme can be detected with high sensitivity by using its characteristic bioluminogenic reaction (13). Various amounts of T7RP were added to a coupled (one‐step) transcription/translation reaction (rabbit reticulocyte mixture, final volume 12.5 µl) that contained the firefly Luc‐DNA under the control of the T7 promoter. The reaction was allowed to proceed for 90 min at 30°C and then the activity of synthesized luciferase was measured by adding 2 µl of the expression mixture to 50 µl of luciferin substrate solution. It was observed (Fig. 2) that the luminescence was linearly related to the number of T7RP molecules in a range extending over four orders of magnitude (5.2 × 104–8 × 108 molecules of T7RP). The signal‐to‐background (S/B) ratio for the 5.2 × 104 molecules was 2.8. The coupled transcription/translation process consists of a series of complex reactions that require the concerted action of numerous factors, such as RNA polymerase, ribosomal subunits, translation initiation, elongation and termination factors, aminoacyl‐tRNA synthetases, etc. Nevertheless, our data demonstrate that the final outcome is a simple linear relationship between input T7RP and the in vitro synthesized protein over a wide range of T7RP concentrations. This forms the basis for the development of a T7RP‐based signal amplification system exploiting T7RP as a label. Enhancing the detectability of T7 RNA polymerase through self‐replication A self‐replication system for T7RP was designed as follows. Two consecutive 90 min in vitro expression reactions were carried out (12.5 µl each). In the first reaction the T7RP catalyzed the transcription of its cognate gene positioned downstream of the T7 promoter (T7RP‐DNA) (11,14) and the generated RNA was translated simultaneously into active T7RP molecules. The newly synthesized T7RP also acted on the T7RP‐DNA template to produce more of the enzyme (self‐replication). The T7RP was then measured by transferring 2 µl into another expression reaction, containing 35 fmol Luc‐DNA, and monitoring the synthesized luciferase. The extent of self‐replication is a function of the T7RP‐DNA level, as indicated by the increase in the luminescence as the T7RP‐DNA concentration increases (Fig. 3A). At the optimum level of T7RP‐DNA, the self‐replication process caused a 110‐fold increase in the signal compared to a reaction that contained no T7RP‐DNA. Similar experiments with decreasing amounts of T7RP (aimed at estimating the detectability of the polymerase) revealed a low level of ‘illegitimate’ transcription of T7RP‐DNA in the absence of T7RP. This was attributed to a eukaryotic RNA polymerase activity that is present in the rabbit reticulocyte extract and initiates a low level of transcription of T7RP‐DNA, generating a few T7RP molecules which, in turn, are amplified by entering the self‐replication cycle. This activity was not detectable in the absence of T7RP‐DNA. Because illegitimate transcription compromises the detectability of T7RP, we carried out experiments to minimize it by adding various amounts of salmon DNA to the rabbit reticulocyte extract. Addition of 100 ng salmon DNA suppressed illegitimate transcription by 98%, whereas it caused only a 15% decrease in the T7RP‐catalyzed transcription of T7RP‐DNA (Fig. 3B). In order to estimate the detectability of T7RP in a self‐replication system, various amounts of the enzyme were added to the first expression reaction containing 25 fmol T7RP‐DNA followed by a separate expression reaction containing 35 fmol Luc‐DNA. The linearity extends from 1.4 × 103–107 molecules (Fig. 2). The S/B ratio at 1400 T7RP molecules was 2.5. We investigated the possibility of combining self‐replication with the detection of T7RP in a single reaction mixture containing both T7RP‐DNA and Luc‐DNA templates. However, it was observed that self‐replication was suppressed dramatically due to competition between the two templates for binding to a limited number of T7RP molecules. Therefore, a delayed addition protocol was designed in which the T7RP was first allowed to act on its cognate gene (self‐replication) followed by the addition of Luc‐DNA, in the same reaction mixture. The T7RP‐DNA/Luc‐DNA ratio, as well as the incubation times required before and after the addition of Luc‐DNA, were optimized to ensure efficient self‐replication and detection with the delayed addition protocol. The T7RP‐DNA/Luc‐DNA molar ratio was studied in the range 0.05–3 at three levels of Luc‐DNA (Fig. 3C). The luminescence increases as the ratio becomes greater, due to increased self‐replication, and reaches a plateau when the molar ratio of the two templates becomes 1–1.5. For the same molar ratio, the signal increases by increasing the concentration of Luc‐DNA. Various combinations of reaction times before and after Luc‐DNA addition were studied (Fig. 3D). The maximum signal was achieved with 60 min–60 min, respectively, giving a 50‐fold enhancement over the assay that contains no T7RP‐DNA. Other combinations, such as 30–60, 30–90 and 60–30, compromised either the yield of the self‐replication reaction or the detection reaction, thus giving a lower signal. In particular, the 0–90 combination (both templates added simultaneously at the beginning of expression) gave the lowest yield of self‐replication (Fig. 3D). The detectability of T7RP using the optimized single expression reaction protocol (delayed addition protocol) was 8.5 × 103 molecules, with a S/B ratio of 4.3. The luminescence was a linear function of the amount of T7RP, up to 107 molecules (Fig. 2). Biotinylation of T7 RNA polymerase, complexation with streptavidin and application to antigen quantification The effect of biotinylation on the activity of T7RP was studied by reacting with increasing concentrations of the sulfo‐N‐hydroxysuccinimide ester of biotin (NHS‐LC‐biotin) at pH 7.7 and 9.0. All reactions were incubated for 1 h at 4°C and the activity of T7RP was measured by in vitro coupled transcription/translation. Inactivation of T7RP becomes significant at biotin:T7RP molar ratios greater than 5 (Fig. 4A), due to modification of free amino groups that are necessary for full activity. The inactivation was more extensive at pH 9.0 than pH 7.7 because the biotinylation reaction is more efficient when the NH2 groups are deprotonated. A microtiter well‐based ‘two‐site’ immunoassay was developed for PSA, as a model (Fig. 1). The antigen was bound both by an immobilized capture antibody and a biotinylated detection antibody. The two anti‐PSA antibodies are directed to different epitopes. BT7RP was complexed to streptavidin and added to the immunocomplex. The solid phase‐bound T7RP was measured by in vitro expression. The complexation of streptavidin with BT7RP was studied at SA:BT7RP molar ratios ranging from 0.5 to 12 (Fig. 4B) and the complexes were applied directly to the assay of 20 fmol antigen. The luminescence reached a maximum at a SA:BT7RP molar ratio of 1.5. The signal dropped sharply at lower or higher ratios. When BT7RP is in excess, all four biotin‐binding sites of streptavidin are occupied and the complexes cannot bind to the biotinylated antibody on the solid phase. On the other hand, when streptavidin is in excess, free streptavidin competes with SA–BT7RP complex for binding to the well (15). The time‐course of the in vitro transcription/translation process with immobilized T7RP was studied up to 180 min (Fig. 4C). The luminescence increased with time but the S/B ratio reached a plateau at 120 min. The background was defined as the luminescence obtained when no antigen was present in the well. The concentration of SA–BT7RP added to the well also affected both the signal and the background of the assay and therefore its detectability (Fig. 4D). The signal increased with the SA–BT7RP concentration and a plateau was reached at 7 nmol/l. The S/B ratio, however, was highest at 3 nmol/l and then dropped because of the increasing non‐specific binding of the complex to the solid phase. The sensitivity and dynamic range of the optimized immunoassay were assessed by analyzing serial dilutions of the antigen, in the appropriate buffer. The S/B ratio was plotted as a function of the amount of PSA in the assay mixture (Fig. 5). In the absence of T7RP‐DNA (no self‐replication), 195 amol antigen were detected with a S/B ratio of 2.3. When the self‐replication system was used (single expression reaction with a delayed addition of Luc‐DNA), as low as 12 amol PSA could be detected with a S/B ratio of 1.9. The dynamic range of the assay extended up to 50 000 amol (Fig. 5). For comparison of detectabilities, we also performed a classical ELISA assay in which the immunocomplex was detected by using a streptavidin–alkaline phosphatase conjugate (New England BioLabs) and the enzymic activity was determined by adding p‐nitrophenylphosphate as a substrate (Sigma) and measuring the absorbance at 405 nm on a microplate photometer (model EL‐307C; BioTek Instruments, Winooski, VT). As shown in Figure 5, a S/B ratio of 2.1 was obtained for 7300 amol antigen. Consequently, the proposed assay offers ∼600‐fold higher detectability and a much wider analytical range. To assess the reproducibility of the proposed immunoassay (including all the steps, i.e. coating of the wells, immunocomplex formation, binding of SA–BT7RP, in vitro coupled transcription/translation and luciferase measurement), we analyzed samples containing 0.1, 1 and 10 fmol PSA. The percent coefficients of variation were 7.8, 7.5 and 8.1, respectively (n = 7). It should be noted that the assay is carried out in three steps, namely immunocomplex formation, reaction with the SA–T7RP complex and finally self‐replication of T7RP and luciferase synthesis. The wells are washed at the end of each step to remove all unbound components of the samples as well as the excess of reagents. Consequently, the transcription/translation reaction mixture does not come into contact with the sample components (removed during first washing). It has been shown previously that T7RP transcripts that are neither capped nor polyadenylated can be efficiently translated in eukaryotic systems (16,17). Optimizing the structure of T7RP‐DNA and Luc‐DNA templates may further enhance the sensitivity of the system. For instance, suitable enhancer and transcription termination sequences may be incorporated to increase the yield of both self‐replication of T7RP and luciferase synthesis. Insertion of both T7RP‐DNA and Luc‐DNA templates, under the control of the T7 promoter, into a single vector may also be investigated for higher yields. Besides firefly luciferase DNA, cDNAs for other highly detectable proteins may be employed, e.g. green fluorescent protein (18), alkaline phosphatase, aequorin (19), etc. A eukaryotic (rabbit reticulocyte) coupled transcription/translation system was used throughout this work. However, prokaryotic (E.coli) systems may also be tested by using DNA templates containing a T7 promoter and a Shine–Dalgarno sequence for ribosome binding. We have used streptavidin as a linker between the biotinylated detection antibody and BT7RP. An alternative means of biotinylation would be to express a recombinant T7RP/strep‐tag (II) fusion protein. Strep‐tag (II) is an eight amino acid polypeptide, generated by combinatorial engineering, which exhibits binding properties for streptavidin or for a mutant form of streptavidin called strep‐tactin (20). The assay is amenable to automation because it is performed in microtiter wells. Although we used the 96‐well format, the technology is transferable to microtiter plates with larger numbers of wells as well as to microfabricated wells for high throughput parallel analysis of multiple proteins. While this manuscript was in progress, another technique for protein analysis was reported (21) that exploited the T7RP reaction for signal amplification (IDAT, immunodetection amplified by T7RP). However, the two concepts are fundamentally different. In IDAT, a double‐stranded DNA fragment (the substrate for T7RP) was used as the antibody label. Following immunocomplex formation, the label was transcribed for 4 h by excess T7RP to generate multiple RNA copies. Since RNA by itself is not a suitable reporter molecule, a 32P‐labeled ribonucleotide was incorporated during transcription in order to confer high detectability. The labeled transcripts were then separated by electrophoresis on a denaturing (urea) polyacrylamide gel and measured by autoradiography. Conversely, the present work uses the enzyme T7RP as a label that by acting on both T7 cDNA and Luc cDNA templates produces the corresponding mRNAs that upon translation generate T7RP (self‐replication of T7RP) and firefly luciferase (indicator enzyme). Although at present the proposed assay does not achieve the reported detectability of IDAT, it offers a number of significant advantages. (i) The assay does not use radioactive isotopes. During the last two decades there have been intense research efforts in the fields of immunoassays and DNA/RNA hybridization assays towards replacing radioactive labels with non‐radioactive alternatives, in order to avoid the hazards associated with the use and disposal of radioactivity. As a consequence, most clinical laboratories now use exclusively non‐radioactive assays and the use of non‐radioactive detection systems in research laboratories is expanding rapidly. (ii) In contrast to IDAT, which requires tedious denaturing gel electrophoresis and autoradiography, the present assay is performed entirely in microtiter wells, thereby allowing for automation and high‐throughput analysis. (iii) A quantitative relationship is established between the luminescence signal and the amount of antigen with a dynamic range covering almost four orders of magnitude. (iv) Compared to IDAT, the proposed assay is much shorter. Indeed, after immunocomplex formation, IDAT requires a 4 h transcription step followed by time‐consuming electrophoresis and autoradiography. In contrast, the present technique requires ∼2 h for quantification of the immunocomplexes. (v) Because of the self‐replication reaction, amplification in the proposed system is exponential, whereas IDAT involves linear amplification of the label. In recent years, research efforts have focused increasingly on the identification of binders, with the requisite affinity and specificity, for a large number of protein analytes. Various approaches include recombinant antibodies selected by phage (22) or ribosome display (23), RNA or DNA aptamers (24) and small organic compounds selected through combinatorial library methods (25,26). The extent and/or the position of biotinylation of proteins, nucleic acids and small molecules can be controlled to achieve minimal interference with the interaction between the binder and the protein analyte. Consequently, SA–BT7RP is a universal detection reagent that can bind with high affinity (Kd = 10–14 M) to any biotinylated binder–analyte complex. ACKNOWLEDGEMENTS We would like to thank J. A. Wollf for providing the plasmid pT7G1. This work was supported by grants from the National Science and Engineering Research Council of Canada (NSERC). View largeDownload slide Figure 1. Assay configuration for quantification of antigens using T7RP as a label. The antigen (Ag) is bound simultaneously to an immobilized capture antibody and a biotinylated detection antibody. BT7RP complexed with streptavidin (SA) is then added to the immunocomplex. The bound T7RP is determined by in vitro coupled transcription/translation. Two approaches were explored. (a) T7RP acts on firefly Luc‐DNA, located downstream of the T7 promoter, to produce several molecules of active luciferase which is measured by its characteristic bioluminogenic reaction. (b) T7RP acts on T7RP cDNA (T7RP‐DNA), positioned downstream of the T7 promoter, to generate several T7RP molecules (self‐replication phase) which, in turn, act on Luc‐DNA to produce luciferase (detection phase). B, biotin. The T7 promoter is represented by a hatched square. View largeDownload slide Figure 1. Assay configuration for quantification of antigens using T7RP as a label. The antigen (Ag) is bound simultaneously to an immobilized capture antibody and a biotinylated detection antibody. BT7RP complexed with streptavidin (SA) is then added to the immunocomplex. The bound T7RP is determined by in vitro coupled transcription/translation. Two approaches were explored. (a) T7RP acts on firefly Luc‐DNA, located downstream of the T7 promoter, to produce several molecules of active luciferase which is measured by its characteristic bioluminogenic reaction. (b) T7RP acts on T7RP cDNA (T7RP‐DNA), positioned downstream of the T7 promoter, to generate several T7RP molecules (self‐replication phase) which, in turn, act on Luc‐DNA to produce luciferase (detection phase). B, biotin. The T7 promoter is represented by a hatched square. View largeDownload slide Figure 2. Establishing a quantitative relationship between the input T7RP and the synthesized firefly luciferase in a coupled in vitro transcription/translation system with and without self‐replication of T7RP. Squares, the transcription/translation mixture contained only Luc‐DNA as template (no self‐replication). Triangles, two expression reactions (12.5 µl each) were carried out. T7RP‐DNA, placed downstream of the T7 promoter, served as the template for the first reaction (self‐replication). Then, 2 µl were transferred to the second expression reaction in which the Luc‐DNA served as template (detection). Circles, a single expression reaction was carried out with delayed addition of Luc‐DNA. Transcription/translation was allowed to proceed for 60 min with T7RP‐DNA as template (self‐replication) and then Luc‐DNA was added and the reaction proceeded for another 60 min prior to luciferase measurement. cpm, counts per min. View largeDownload slide Figure 2. Establishing a quantitative relationship between the input T7RP and the synthesized firefly luciferase in a coupled in vitro transcription/translation system with and without self‐replication of T7RP. Squares, the transcription/translation mixture contained only Luc‐DNA as template (no self‐replication). Triangles, two expression reactions (12.5 µl each) were carried out. T7RP‐DNA, placed downstream of the T7 promoter, served as the template for the first reaction (self‐replication). Then, 2 µl were transferred to the second expression reaction in which the Luc‐DNA served as template (detection). Circles, a single expression reaction was carried out with delayed addition of Luc‐DNA. Transcription/translation was allowed to proceed for 60 min with T7RP‐DNA as template (self‐replication) and then Luc‐DNA was added and the reaction proceeded for another 60 min prior to luciferase measurement. cpm, counts per min. View largeDownload slide Figure 3. (A) Effect of the amount of T7RP‐DNA on the extent of self‐ replication of T7RP. Two consecutive transcription/translation reactions (12.5 µl each) were performed. In the first reaction T7RP acted on various concentrations of T7RP‐DNA (self‐replication). Then 2 µl of the reaction mixture was transferred into a second expression reaction containing a constant amount of Luc‐DNA (35 fmol). The synthesized firefly luciferase was measured as described in Materials and Methods. The luminescence is plotted against the amount of T7RP‐DNA in the transcription/translation reaction. The arrow indicates the signal obtained without self‐replication (absence of T7RP‐DNA). (B) Effect of the concentration of salmon DNA on the extent of ‘illegitimate’ expression of T7RP‐DNA in the absence of T7RP (solid line) and on the extent of T7RP‐catalyzed expression (dashed line). Experiments were performed as above using two consecutive expression reactions, the first reaction containing 25 fmol T7RP‐DNA and various amounts of salmon testes DNA and the second reaction containing 35 fmol Luc‐DNA. The percent luminescence is plotted as a function of the amount of salmon DNA in the transcription/translation reaction mixture. The value of 100% is defined as the luminescence obtained with no salmon DNA present. (C) Study of the effect of the T7RP‐DNA:Luc‐DNA molar ratio on the yield of an expression reaction that combines self‐replication of T7RP and luciferase synthesis. A delayed addition protocol was performed (60 min–60 min). Increasing amounts of T7RP‐DNA were used with 8.75 (squares), 17.5 (circles) and 35 fmol (triangles) Luc‐DNA. (D) Effect of the reaction times before and after the addition of Luc‐DNA (delayed addition protocol) on the yield of an expression reaction mixture that contains T7RP and T7RP‐DNA. The reaction starts with the addition of 25 fmol T7RP‐DNA. The first and second numbers of each pair on the x‐axis correspond to the incubation time before and after the addition of 35 fmol Luc‐DNA, respectively. The first column (90 min) represents the signal obtained in the absence of self‐replication (no T7RP‐DNA). View largeDownload slide Figure 3. (A) Effect of the amount of T7RP‐DNA on the extent of self‐ replication of T7RP. Two consecutive transcription/translation reactions (12.5 µl each) were performed. In the first reaction T7RP acted on various concentrations of T7RP‐DNA (self‐replication). Then 2 µl of the reaction mixture was transferred into a second expression reaction containing a constant amount of Luc‐DNA (35 fmol). The synthesized firefly luciferase was measured as described in Materials and Methods. The luminescence is plotted against the amount of T7RP‐DNA in the transcription/translation reaction. The arrow indicates the signal obtained without self‐replication (absence of T7RP‐DNA). (B) Effect of the concentration of salmon DNA on the extent of ‘illegitimate’ expression of T7RP‐DNA in the absence of T7RP (solid line) and on the extent of T7RP‐catalyzed expression (dashed line). Experiments were performed as above using two consecutive expression reactions, the first reaction containing 25 fmol T7RP‐DNA and various amounts of salmon testes DNA and the second reaction containing 35 fmol Luc‐DNA. The percent luminescence is plotted as a function of the amount of salmon DNA in the transcription/translation reaction mixture. The value of 100% is defined as the luminescence obtained with no salmon DNA present. (C) Study of the effect of the T7RP‐DNA:Luc‐DNA molar ratio on the yield of an expression reaction that combines self‐replication of T7RP and luciferase synthesis. A delayed addition protocol was performed (60 min–60 min). Increasing amounts of T7RP‐DNA were used with 8.75 (squares), 17.5 (circles) and 35 fmol (triangles) Luc‐DNA. (D) Effect of the reaction times before and after the addition of Luc‐DNA (delayed addition protocol) on the yield of an expression reaction mixture that contains T7RP and T7RP‐DNA. The reaction starts with the addition of 25 fmol T7RP‐DNA. The first and second numbers of each pair on the x‐axis correspond to the incubation time before and after the addition of 35 fmol Luc‐DNA, respectively. The first column (90 min) represents the signal obtained in the absence of self‐replication (no T7RP‐DNA). View largeDownload slide Figure 4. (A) Study of the effect of biotinylation on the activity of T7RP. Biotinylation was performed at various NHS‐LC‐biotin:T7RP molar ratios at pH 7.7 and 9.0, as described in Materials and Methods. T7RP was then determined by in vitro expression using 35 fmol Luc‐DNA as template. The percent luminescence is plotted versus the molar ratio NHS‐LC‐biotin:T7RP at pH 7.7 (solid line) and pH 9.0 (dashed line). The value 100% is defined as the signal obtained from non‐biotinylated T7RP. (B) Optimization of the streptavidin SA:BT7RP molar ratio for the preparation of the SA–BT7RP complex. The complex was prepared as described in Materials and Methods and 60 fmol were used (without prior purification) for antigen quantification (20 fmol PSA). The solid and dashed lines correspond to the signal and S/B ratio, respectively. The background is defined as the luminescence obtained in the absence of antigen. (C) Time dependence of the transcription/translation reaction with T7RP immobilized on the solid phase. The immunoassay was performed as described in Materials and Methods. Following reaction with the SA–BT7RP complex, a 25 µl transcription/translation mixture was added containing 52.5 fmol Luc‐DNA as template. Expression was allowed to proceed for various time intervals (up to 180 min). The solid and dashed lines correspond to the signal and S/B ratio, respectively. The background is defined as the luminescence obtained in the absence of antigen. (D) Effect of the concentration of SA–BT7RP complex on the luminescence (solid line) and the S/B ratio (dashed line) obtained from the assay of 10 fmol antigen. Luc‐DNA was used as template for T7RP. The immunoassay was performed as described in Materials and Methods. View largeDownload slide Figure 4. (A) Study of the effect of biotinylation on the activity of T7RP. Biotinylation was performed at various NHS‐LC‐biotin:T7RP molar ratios at pH 7.7 and 9.0, as described in Materials and Methods. T7RP was then determined by in vitro expression using 35 fmol Luc‐DNA as template. The percent luminescence is plotted versus the molar ratio NHS‐LC‐biotin:T7RP at pH 7.7 (solid line) and pH 9.0 (dashed line). The value 100% is defined as the signal obtained from non‐biotinylated T7RP. (B) Optimization of the streptavidin SA:BT7RP molar ratio for the preparation of the SA–BT7RP complex. The complex was prepared as described in Materials and Methods and 60 fmol were used (without prior purification) for antigen quantification (20 fmol PSA). The solid and dashed lines correspond to the signal and S/B ratio, respectively. The background is defined as the luminescence obtained in the absence of antigen. (C) Time dependence of the transcription/translation reaction with T7RP immobilized on the solid phase. The immunoassay was performed as described in Materials and Methods. Following reaction with the SA–BT7RP complex, a 25 µl transcription/translation mixture was added containing 52.5 fmol Luc‐DNA as template. Expression was allowed to proceed for various time intervals (up to 180 min). The solid and dashed lines correspond to the signal and S/B ratio, respectively. The background is defined as the luminescence obtained in the absence of antigen. (D) Effect of the concentration of SA–BT7RP complex on the luminescence (solid line) and the S/B ratio (dashed line) obtained from the assay of 10 fmol antigen. Luc‐DNA was used as template for T7RP. The immunoassay was performed as described in Materials and Methods. View largeDownload slide Figure 5. Assessing the sensitivity and analytical range for antigen quantification using T7RP as a label, (squares) without self‐replication (absence of T7RP‐DNA) and (circles) with self‐replication of T7RP. For self‐replication, T7RP‐DNA was included in the expression reaction mixture and the delayed addition protocol was employed. The immunoassays were carried out as described in Materials and Methods. The luminescence (corrected for the background) is plotted against the concentration of PSA present in the well. The background is defined as the signal obtained in the absence of antigen. Also, shown (triangles) are data for a classical ELISA that uses a streptavidin–alkaline phosphatase conjugate (SA–AP) for detection and p‐nitrophenylphosphate as a chromogenic substrate. Following immunocomplex formation (as described in Materials and Methods), SA–AP was added (1000 U), instead of SA–BT7RP, and incubated for 15 min. After washing out the excess of conjugate, the substrate was added for 30 min in the dark, followed by absorbance measurement at 405 nm. View largeDownload slide Figure 5. Assessing the sensitivity and analytical range for antigen quantification using T7RP as a label, (squares) without self‐replication (absence of T7RP‐DNA) and (circles) with self‐replication of T7RP. For self‐replication, T7RP‐DNA was included in the expression reaction mixture and the delayed addition protocol was employed. The immunoassays were carried out as described in Materials and Methods. The luminescence (corrected for the background) is plotted against the concentration of PSA present in the well. The background is defined as the signal obtained in the absence of antigen. Also, shown (triangles) are data for a classical ELISA that uses a streptavidin–alkaline phosphatase conjugate (SA–AP) for detection and p‐nitrophenylphosphate as a chromogenic substrate. Following immunocomplex formation (as described in Materials and Methods), SA–AP was added (1000 U), instead of SA–BT7RP, and incubated for 15 min. After washing out the excess of conjugate, the substrate was added for 30 min in the dark, followed by absorbance measurement at 405 nm. References 1. Wodicka,L., Dong,H., Mittmann,M., Ho,M.H. and Lockhart,D.J. ( 1997) Genome‐wide expression monitoring in Saccharomyces cerevisiae. Nat. Biotechnol. , 15, 1359–11367. Google Scholar 2. Schena,M., Shalon,D., Heller,R., Chai,A., Brown,P.O. and Davis,R.W. ( 1996) Parallel human genome analysis: microarray‐based expression monitoring of 1000 genes. Proc. Natl Acad. Sci. USA , 93, 10614–10619. Google Scholar 3. Gygi,S.P., Rochon,Y., Franza,B.R. and Aebersold,R. ( 1999) Correlation between protein and mRNA abundance in yeast. Mol. Cell. Biol. , 19, 1720–1730. Google Scholar 4. Gygi,S.P., Corthals,G.L., Zhang,Y., Rochon,Y. and Aebersold,R. ( 2000) Evaluation of two‐dimensional gel‐electrophoresis‐based proteome analysis technology. Proc. Natl Acad. Sci. USA , 97, 9390–9395. Google Scholar 5. Sano,T., Smith,C.L. and Cantor,C.R. ( 1992) Immuno‐PCR: very sensitive antigen detection by means of specific antibody‐DNA conjugates. Science , 258, 120–122. Google Scholar 6. Schweitzer,B., Wiltshire,S., Lambert,J., O’Malley,S., Kukanskis,K., Zhu,Z., Kingsmore,S.F., Lizardi,P.M. and Ward,D.C. ( 2000) Immunoassays with rolling circle DNA amplification: a versatile platform for ultrasensitive antigen detection. Proc. Natl Acad. Sci. USA , 97, 10113–10119. Google Scholar 7. Chiu,N.H.L. and Christopoulos,T.K. ( 1999) Two‐site expression immunoassay using a firefly luciferase‐coding DNA label. Clin. Chem. , 45, 1954–1959. Google Scholar 8. Christopoulos,T.K. and Diamandis,E.P. ( 1996) Fluorescence immunoassays. In Diamandis,E.P. and Christopoulos,T.K. (eds), Immunoassay. Academic Press, San Diego, CA, pp. 309–335. Google Scholar 9. Kricka,L.J. ( 1996) Chemiluminescence immunoassay. In Diamandis,E.P. and Christopoulos,T.K. (eds), Immunoassay. Academic Press, San Diego, CA, pp. 337–353. Google Scholar 10. Titus,D.E. ( 1991) Promega Protocols and Applications Guide, 2nd Edn. Promega Corp., Madison, WI. Google Scholar 11. Deng,H. and Wolff,J.A. ( 1994) Self‐amplifying expression from the T7 promoter in 3T3 mouse fibroblasts. Gene , 143, 245–249. Google Scholar 12. Davanloo,P., Rosenberg,A.H., Dunn,J.J. and Studier,F.W. ( 1984) Cloning and expression of the gene for bacteriophage T7 RNA polymerase. Proc. Natl Acad. Sci. USA , 81, 2035–2039. Google Scholar 13. Wilson,T. and Hastings,J.W. ( 1998) Bioluminescence. Annu. Rev. Cell. Dev. Biol. , 14, 197–230. Google Scholar 14. Dubendorff,J.W. and Studier,F.W. ( 1991) Creation of a T7 autogene. Cloning and expression of the gene for bacteriophage T7 RNA polymerase under control of its cognate promoter. J. Mol. Biol. , 219, 61–68. Google Scholar 15. Diamandis,E.P. and Christopoulos,T.K. ( 1991) The biotin‐(strept)avidin system: principles and applications in biotechnology. Clin. Chem. , 37, 625–636. Google Scholar 16. Gurevich,V.V., Pokrovskaya,I.D., Obukhova,T.A. and Zozulya,S.A. ( 1991) Preparative in vitro mRNA synthesis using SP6 and T7 RNA polymerases. Anal. Biochem. , 195, 207–213. Google Scholar 17. Craig,D., Howell,M.T., Gibbs,C.L., Hunt,T. and Jackson,R.J. ( 1992) Plasmid cDNA‐directed protein synthesis in a coupled eukaryotic in vitro transcription‐translation system. Nucleic Acids Res. , 20, 4987–4995. Google Scholar 18. Tsien,Y. ( 1998) The green fluorescent protein. Annu. Rev. Biochem. , 67, 509–544. Google Scholar 19. White,S.R. and Christopoulos,T.K. ( 1999) Signal amplification system for DNA hybridization assays based on in vitro expression of a DNA label encoding apoaequorin. Nucleic Acids Res. , 27, e25. Google Scholar 20. Skerra,A. and Schmidt,T.G.M. ( 1999) Applications of a peptide ligand for streptavidin: the strep‐tag. Biomol. Eng. , 16, 79–86. Google Scholar 21. Zhang,H.T., Kacharmina,J.E., Miyashiro,K., Greene,M.I. and Eberwine,J. ( 2001) Protein quantification from complex protein mixtures using a proteomics methodology with single‐cell resolution. Proc. Natl Acad. Sci. USA , 98, 5497–5502. Google Scholar 22. de Wildt,R.M.T., Mundy,C.R., Gorick,B.D. and Tomlinson,I.M. ( 2000) Antibody arrays for high‐throughput screening of antibody‐antigen interactions. Nat. Biotechnol. , 18, 989–994. Google Scholar 23. Hanes,J., Schaffitzel,C., Knappik,A. and Pluckthun,A. ( 2000) Picomolar affinity antibodies from a fully synthetic naive library selected and evolved by ribosome display. Nat. Biotechnol. , 18, 1287–1292. Google Scholar 24. Hesselberth,J., Robertson,M.P., Jhaveri,S. and Ellington,A.D. ( 2000) In vitro selection of nucleic acids for diagnostic applications. Rev. Mol. Biotechnol. , 74, 15–25. Google Scholar 25. Han,Y. and Kodadek,T. ( 2000) Peptides selected to bind the Ga180 repressor are potent transcriptional activation domains in yeast. J. Biol. Chem. , 275, 14979–14984. Google Scholar 26. Shuker,S.B., Hajduk,P.J., Meadows,R.P. and Fesik,S.W. ( 1996) Discovering high‐affinity ligands for proteins: SAR by NMR. Science , 274, 1531–1534. Google Scholar
A Y639F/H784A T7 RNA polymerase double mutant displays superior properties for synthesizing RNAs with non‐canonical NTPsPadilla, Robert; Sousa, Rui
doi: N/Apmid: N/A
A T7 RNA polymerase in which Tyr639 is mutated to Phe readily utilizes 2′‐deoxy, 2′‐NH2 and 2′‐F NTPs as substrates and has been widely used to synthesize modified RNAs for a variety of applications. This mutant does not readily utilize NTPs with bulkier 2′‐substituents, nor does it facilitate incorporation of NTPs with modifications at other positions. Introduction of a second mutation (H784A) into the Y639F background markedly enhances utilization of NTPs with bulky 2′‐substituents (2′‐OMe and 2′‐N3), and may also enhance use of NTPs with modifications at other than the 2′‐position. The Y639F/H784A double mutant may therefore be exceptionally useful for incorporation of a variety of non‐canonical NMPs into RNA.
Random DNA fragmentation with endonuclease V: application to DNA shufflingMiyazaki, Kentaro
doi: 10.1093/nar/gnf139pmid: 12490730
Abstract The enzyme endonuclease V nicks uracil‐containing DNA at the second or third phosphodiester bond 3′ to uracil sites. I applied the enzyme to random fragmentation of DNA to revise the complex DNA shuffling protocol. The merit of using endonuclease V is that cleavage occurs at random sites and the length of the fragments can easily be adjusted by varying the concentration of dUTP in the polymerase chain reaction. Unlike the conventional method using DNase I, no partial digestion or gel separation of fragments is required. Therefore, labor is dramatically reduced and reproducibility ensured. I applied this method to recombine two truncated green fluorescent protein (GFP) genes and demonstrated successful DNA shuffling by the appearance of the fluorescent full‐length GFP genes. Received August 9, 2002; Revised September 25, 2002; Accepted October 28, 2002 INTRODUCTION DNA recombination in vitro as an evolution strategy was first reported by Stemmer in 1994 (1). Stemmer’s method, which he called DNA shuffling, uses enzymatic digestion (most commonly with DNase I) of parent genes to generate a pool of random DNA fragments. These fragments can be assembled by iterative cycles of denaturation, annealing and extension with thermostable DNA polymerase—known as the polymerase chain reaction (PCR). This reaction generates a mixture of products in length and combination, from which full‐length genes are amplified by PCR with flanking primers. Although the method has already proven to be extremely useful in directed evolution (1,2), the experimental procedure requires careful monitoring of each reaction step and, hence, is extremely labor intensive and time consuming. In particular, DNA fragmentation is most problematic since the reaction has to be carefully controlled in order to obtain fragments of appropriate length. In addition, the length of the fragments obtained by DNase I digestion varies greatly with minor changes in conditions, including the amount of nuclease, the source (supplier) or lot of nuclease, the reaction temperature and the purity of DNA substrates. This makes the digestion experiment non‐reproducible. Furthermore, it is known that DNase I digests double‐stranded DNA preferentially at sites adjacent to pyrimidine nucleotides (3). Therefore, using fragments generated by DNase I digestion may induce a sequence bias into the recombination (4). To overcome these drawbacks, I attempted to develop an alternative random digestion method whereby complex libraries can be obtained through a simple and reproducible approach. In this study, I used endonuclease V, which is known to nick uracil‐containing DNA (and other damaged DNA having inosine, abasic sites, and so on) at the second or third phosphodiester bond 3′ to uracil sites (5,6). Although the cleavage sites are always two or three bases downstream of a thymidine (substituted by uracil) site, this method is expected to produce much fewer hot and cold spots that decrease random representation of fragments which occur with DNase I cleavage (3,4). Using the DNA fragments generated by endonuclease V digestion, successful DNA shuffling was achieved with shuffling efficiency equivalent to DNase I. MATERIALS AND METHODS Reagents Endonuclease V was purchased from Trevigen (Gaithersburg, MD), Taq2000 DNA polymerase was from Stratagene (La Jolla, CA), deoxynucleotides were from Amersham (Piscataway, NJ). Plasmid pGFPuv was obtained from Clontech (Palo Alto, CA). Competent Escherichia coli JM109 cells were purchased from Toyobo (Osaka, Japan). A plasmid, pTGuv1, contained the green fluorescent protein (GFP) gene (2) under control of a tac promoter. Preparation of uracil‐containing recombination templates Uracil‐containing recombination templates were prepared by PCR in the presence of dUTP (7). The PCR mixture contained 10 mM Tris–HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each of dATP, dGTP and dCTP, and 0.2 mM of a dTTP/dUTP mixture, 25 pmol each of primers, 50 ng of each template plasmid, and 1.25 U of Taq2000 DNA polymerase in a total volume of 50 µl. The concentration of dTTP/dUTP was varied (0.2 mM/0 mM, 0.15 mM/0.05 mM, 0.1 mM/0.1 mM, 0.05 mM/0.15 mM or 0 mM/0.2 mM) to find a condition which gives fragments of appropriate length. The PCR mixture was heated at 95°C for 1 min, followed by 25 cycles of incubation at 94°C for 30 s, 55°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 5 min. Amplified DNA (∼1.5 µg) was separated by agarose gel electrophoresis, purified in a QIAquick spin column (Qiagen, Tokyo, Japan), and dissolved in 10 mM N‐2‐hydroxyethylpiperazine‐N′‐2‐ethanesulfonate (HEPES)–KOH pH 7.4, 50 mM NaCl, 0.5 mM MnCl2. Endonuclease V digestion of the recombination templates To the solution, 1 U of endonuclease V was added and incubated at 37°C for 12 h, followed by heating at 95°C for 10 min. A portion of the products (3 µl) was used for agarose gel (1.6%) electrophoresis to check the fragment lengths. The rest of the sample was purified in a QIAquick spin column, eluted in 30 µl of water, and used for the subsequent assembly reaction. Assembly reaction The assembly reaction of DNA fragments was carried out in 10 mM Tris–HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each of dNTP, 5 µl each of DNA fragments, and 1 U of Taq2000 DNA polymerase in a total volume of 20 µl. Reaction mixtures were heated at 95°C for 1 min, followed by 30 cycles of incubation at 94°C for 30 s, 55°C for 30 s, 72°C for 30 s, and a final incubation at 72°C for 5 min. Amplification of full‐length genes Full‐length genes were amplified using a set of flanking primers. The PCR mixture contained 10 mM Tris–HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each of dNTP, 25 pmol each of flanking primers, 0.5 µl of assembly product, and 1.25 U of Taq2000 DNA polymerase in a total volume of 50 µl. Reaction mixtures were heated at 95°C for 1 min, followed by 25 cycles of incubation at 94°C for 30 s, 55°C for 30 s, 72°C for 30 s, and a final incubation at 72°C for 5 min. Library production and screening Amplified fragments were purified and cloned back into the expression plasmid by ligation. Competent E.coli JM109 cells were transformed with the ligation mixture and grown overnight at 37°C on LB agar plates containing 100 µg/ml ampicillin and 0.1 mM isopropyl‐β‐d‐thiogalactopyranoside. The total number of colonies was counted under white light and the number of fluorescent colonies was counted under UV light (∼365 nm). RESULTS AND DISCUSSION In order to overcome the technical problems of DNase I digestion and to achieve high shuffling efficiency, Kikuchi et al. (8) used frequent‐cutter restriction enzymes. Their method is effective at shuffling and is simple and reproducible. However, it requires DNA sequence information a priori in order to select appropriate restriction enzymes and hence it is not versatile. In addition, the cleavage sites and the length of the obtained fragments are not random. Therefore, the resultant library will severely lack complexity. To overcome these drawbacks, I attempted to develop an alternative random digestion method whereby complex libraries can be obtained through a simple and reproducible approach. In this study, I used endonuclease V, which is known to nick uracil‐containing DNA (and other damaged DNA having inosine, abasic sites and so on) at the second or third phosphodiester bond 3′ to uracil sites (5,6). Although the endonuclease V specifically recognizes uracil, the PCR technique allows uracil to distribute at random over the sequence. In addition, the cleavage site is not adjacent to the uracil site but at the second or third 3′ bond to the uracil site. For these reasons, digestion sites are considered to be random, as with DNase I. Another determinant for efficient shuffling is the length of fragments. With endonuclease V, the length can be controlled by adjusting the number of uracil residues in the DNA simply by varying the concentration of dUTP in the PCR. All these features unique to endonuclease V are suitable for random DNA fragmentation, compared with the conventional method using DNase I. Based on this idea, I tested the validity of endonuclease V‐mediated fragmentation in DNA shuffling. For recombination templates, two truncated GFP genes were prepared in which the full‐length GFP sequence (∼750 bp) was truncated at amino acids Leu60 and Val120, respectively, by replacing the amino acid codons with a stop codon (TAA). Recovery of fluorescence requires recombination between two variants (GFPuv60* and GFPuv120*, respectively) to restore a full‐length GFP gene, and this system allows a quick assay for DNA shuffling (Fig. 1) (9). In the experiment, the ratio of dTTP to dUTP was first varied in the PCR preparation of full‐length recombination templates. After treatment with endonuclease V, a portion of the products (1/10) was separated by agarose gel electrophoresis (Fig. 2). As predicted, no digestion occurred for the sample prepared in the absence of dUTP (Fig. 2, lane 1), and the average length of fragments decreased as the ratio of dTTP to dUTP decreased. The full‐length gene was visible for the samples prepared in 0.2 mM dTTP/0 mM dUTP, 0.15 mM dTTP/0.05 mM dUTP and 0.1 mM dTTP/0.1 mM dUTP (Fig. 2, lanes 1–3), suggesting that these conditions were not suitable for DNA shuffling. DNA fragments prepared in 0.05 mM dTTP/0.15 mM dUTP gave 100–300‐bp fragments (Fig. 2, lane 4) and the condition was considered to be appropriate for DNA shuffling. If dTTP was completely substituted by dUTP, the resultant fragments were hardly visible, probably because the fragments became too short (Fig. 2, lane 5). This condition was also considered to be unsuitable for DNA shuffling. In order to survey the appropriate preparation of the fragments for DNA shuffling, the fragments were purified and subjected to subsequent assembly and amplification reactions. At this stage, no products were produced from the fragments prepared in the absence of dTTP (complete substitution to dUTP), indicating the unsuitability for DNA shuffling. The rest of the preparations gave a full‐size gene. The gene was then cloned and mutant libraries were created. As for the result of screening for fluorescence, libraries created by using the fragments prepared in dTTP/dUTP of 0.2 mM/0 mM, 0.15 mM/0.05 mM and 0.1 mM/0.1 mM gave virtually no fluorescent colonies (<0.1%), indicating the contamination of the full‐length gene was problematic. In contrast, DNA fragments prepared in 0.05 mM dTTP/0.15 mM dUTP gave fluorescent colonies at a high frequency; 10 out of 100 transformants were fluorescent. This fraction of fluorescent colonies (10%) was smaller than the theoretical prediction of 25% (10), but was close to the value obtained by the DNase I method, which was done in nearly the same assay system (10). Five fluorescent colonies were picked and their DNA sequences checked. In all the colonies, the sequence of the GFP gene was identical to that of the parent GFP gene. In my case, fragments of appropriate length were obtained in the 0.05 mM dTTP/0.15 mM dUTP condition. However, this does not mean that the condition is immediately applicable to any other genes. This is because base composition varies from sequence to sequence. In our case, the G+C content of the GFP gene was 41%, slightly AT‐rich. If the G+C content is high, as is often seen in some bacteria such as Streptomyces and Thermus, it will be necessary to increase the ratio of dUTP to dTTP accordingly. Besides the base composition, the incorporation of uracil also depends upon the PCR conditions (e.g. DNA polymerase, number of cycles, amount of template) (7). For these reasons, it is always necessary to optimize the PCR conditions to get fragments of appropriate length. In conclusion, random DNA fragmentation by endonuclease V is a handy and reproducible method and can be used instead of fragmentation by DNase I, which is technically problematic, without sacrificing shuffling efficiency. View largeDownload slide Figure 1. Recombination assay system. Two GFP genes having stop codons (GFPuv60* and GFPuv120*) were used for recombination templates. If recombination occurs between the two stop codons and the resultant gene does not contain stop codons, the gene recovers the fluorescence. The fluorescent gene is shown as white bars and non‐fluorescent variants having stop codon(s) are shown as shaded bars. View largeDownload slide Figure 1. Recombination assay system. Two GFP genes having stop codons (GFPuv60* and GFPuv120*) were used for recombination templates. If recombination occurs between the two stop codons and the resultant gene does not contain stop codons, the gene recovers the fluorescence. The fluorescent gene is shown as white bars and non‐fluorescent variants having stop codon(s) are shown as shaded bars. View largeDownload slide Figure 2. Monitoring the endonuclease V digestion by agarose gel (1.6%) electrophoresis. M, DNA size standard (1 kb DNA ladder; Takara, Tokyo, Japan); lane 1, 0.2 mM dTTP/0 mM dUTP; lane 2, 0.15 mM dTTP/ 0.05 mM dUTP; lane 3, 0.1 mM dTTP/0.1 mM dUTP; lane 4, 0.05 mM dTTP/0.15 mM dUTP; lane 5, 0 mM dTTP/0.2 mM dUTP. See text for the other components of the reaction mixture. View largeDownload slide Figure 2. Monitoring the endonuclease V digestion by agarose gel (1.6%) electrophoresis. M, DNA size standard (1 kb DNA ladder; Takara, Tokyo, Japan); lane 1, 0.2 mM dTTP/0 mM dUTP; lane 2, 0.15 mM dTTP/ 0.05 mM dUTP; lane 3, 0.1 mM dTTP/0.1 mM dUTP; lane 4, 0.05 mM dTTP/0.15 mM dUTP; lane 5, 0 mM dTTP/0.2 mM dUTP. See text for the other components of the reaction mixture. References 1. Stemmer,W.P. ( 1994) Rapid evolution of a protein in vitro by DNA shuffling. Nature , 370, 389–391. Google Scholar 2. Crameri,A., Whitehorn,E.A., Tate,E. and Stemmer,W.P. ( 1996) Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotechnol. , 14, 315–319. Google Scholar 3. Sambrook,J. and Russell,D.W. ( 2001) Molecular Cloning: A Laboratory Manual, 3rd Edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Google Scholar 4. Joern,J.M., Meinhold,P. and Arnold,F.H. ( 2002) Analysis of shuffled gene libraries. J. Mol. Biol. , 316, 643–656. Google Scholar 5. Huang,J., Lu,J., Barany,F. and Cao,W. ( 2001) Multiple cleavage activities of endonuclease V from Thermotoga maritima: recognition and strand nicking mechanism. Biochemistry , 40, 8738–8748. Google Scholar 6. Yao,M. and Kow,Y.W. ( 1997) Further characterization of Escherichia coli endonuclease V. Mechanism of recognition for deoxyinosine, deoxyuridine and base mismatches in DNA. J. Biol. Chem. , 272, 30774–30779. Google Scholar 7. Slupphaug,G., Alseth,I., Eftedal,I., Volden,G. and Krokan,H.E. ( 1993) Low incorporation of dUMP by some thermostable DNA polymerases may limit their use in PCR amplifications. Anal. Biochem. , 211, 164–169. Google Scholar 8. Kikuchi,M., Ohnishi,K. and Harayama,S. ( 1999) Novel family shuffling methods for the in vitro evolution of enzymes. Gene , 236, 159–167. Google Scholar 9. Volkov,A.A., Shao,Z. and Arnold,F.H. ( 2000) Recombination and chimeragenesis by in vitro heteroduplex formation and in vivo repair. Nucleic Acids Res. , 27, e18. Google Scholar 10. Volkov,A.A. and Arnold,F.H. ( 2000) Methods for in vitro DNA recombination and random chimeragenesis. Methods Enzymol. , 328, 447–456. Google Scholar
Optimization of trans‐splicing ribozyme efficiency and specificity by in vivo genetic selectionAyre, Brian G.; Köhler, Uwe; Turgeon, Robert; Haseloff, Jim
doi: N/Apmid: N/A
Trans‐splicing ribozymes are RNA‐based catalysts capable of splicing RNA sequences from one transcript specifically into a separate target transcript. In doing so, a chimeric mRNA can be produced, and new gene activities triggered in living cells dependent on the presence of the target mRNA. Based on this ability of trans‐splicing ribozymes to deliver new gene activities, a simple and versatile plating assay was developed in Saccharomyces cerevisiae for assessing and optimizing constructs in vivo. Trans‐splicing ribozymes were used to splice sequences encoding a GAL4‐derived transcription activator into a target transcript from a prevalent viral pathogen. The transcription activator translated from this new mRNA in turn triggered the expression of genes under the regulatory control of GAL4 upstream‐activating sequences. Two of the activated genes complemented metabolic deficiencies in the host strain, and allowed growth on selective media. A simple genetic assay based on phenotypic conversion from auxotrophy to prototrophy was established to select efficient and specific trans‐splicing ribozymes from a ribozyme library. This simple assay may prove valuable for selecting optimal target sites for therapeutic agents such as ribozymes, antisense RNA and antisense oligodeoxyribonucleotides, and for optimizing the design of the therapeutic agents themselves, in higher eukaryotes.
Optimization of trans‐splicing ribozyme efficiency and specificity by in vivo genetic selectionAyre, Brian G.;Köhler, Uwe;Turgeon, Robert;Haseloff, Jim
doi: 10.1093/nar/gnf141pmid: 12490732
Abstract Trans‐splicing ribozymes are RNA‐based catalysts capable of splicing RNA sequences from one transcript specifically into a separate target transcript. In doing so, a chimeric mRNA can be produced, and new gene activities triggered in living cells dependent on the presence of the target mRNA. Based on this ability of trans‐splicing ribozymes to deliver new gene activities, a simple and versatile plating assay was developed in Saccharomyces cerevisiae for assessing and optimizing constructs in vivo. Trans‐splicing ribozymes were used to splice sequences encoding a GAL4‐derived transcription activator into a target transcript from a prevalent viral pathogen. The transcription activator translated from this new mRNA in turn triggered the expression of genes under the regulatory control of GAL4 upstream‐activating sequences. Two of the activated genes complemented metabolic deficiencies in the host strain, and allowed growth on selective media. A simple genetic assay based on phenotypic conversion from auxotrophy to prototrophy was established to select efficient and specific trans‐splicing ribozymes from a ribozyme library. This simple assay may prove valuable for selecting optimal target sites for therapeutic agents such as ribozymes, antisense RNA and antisense oligodeoxyribonucleotides, and for optimizing the design of the therapeutic agents themselves, in higher eukaryotes. Received August 16, 2002; Revised October 9, 2002; Accepted October 20, 2002 INTRODUCTION RNA molecules with catalytic activities, commonly referred to as ribozymes, are potential tools for genetic manipulation. Various naturally occurring ribozymes have been dissected and used as endoribonucleases to cleave specific target transcripts (1–3). In general, however, these ribozymes demonstrate poor efficiency in vivo, and optimization has been hampered by the need to detect small reductions in target‐gene activity (4,5). An alternative approach to ribozyme technologies might exploit the coupled cleavage and ligation reactions of naturally occurring RNA catalysts involved in splicing. In splicing reactions, intervening regions that disrupt the coding sequence of a gene are removed, and result in gene activation. Gene activation presents tremendous potential for the development of RNA‐based technologies since detection of a de novo gene activity above negligible background is intrinsically easier to study than reductions in the levels of an existing gene, and the new activity can have profound effects on the phenotype of the host organism. Group I introns are natural splicing catalysts that autocatalyze the removal of intron sequences and ligation of the exon sequences via two sequential transesterification reactions (6). Group I introns have been bisected, and co‐expression of the individual components in vivo resulted in trans‐splicing of exon sequences from one molecule into sequences of another, to create a unique chimeric transcript (7–9). Translation of the fusion RNA resulted in de novo gene activation. As one example, wild type globin sequences were trans‐spliced into sickle globin mRNA to create ‘repaired’ transcripts in vivo (10). In a second example, an endotoxin was activated by trans‐splicing endotoxin sequences into the coat protein mRNA of a prevalent plant virus (11). The translated product of this fusion mRNA resulted in inhibition of cellular proliferation, and was limited to cells expressing the viral target sequences. The latter example incorporated design parameters shown to improve the specificity of the ribozyme for its target, and the efficiency of trans‐splicing, relative to other designs (7,11,12). The inclusion of an ‘extended antisense sequence’ (EAS) at the 5′ end of the ribozyme improved both the efficiency and specificity for the intended target (11). Trans‐splicing efficiency was further improved by retaining a potential P10 duplex at the 3′ end of the catalytic core (7). Notwithstanding these improved design parameters, it was recognized that trans‐splicing efficiency and specificity could be further improved. In this paper, an in vivo, genetic approach for selecting optimized ribozymes from a library of constructs is described. Selection is by a simple Saccaromyces cerevisiae plating assay, and is based on ribozyme‐mediated conversion from auxotrophy to prototrophy. Start‐codon‐deficient mRNA encoding a GAL4‐derived transcription activator was trans‐spliced in frame with the desired target transcript, which, for the experiments described, was derived from the coat protein of cucumber mosaic virus (CMV) (13). The translated product of this chimeric transcript activated downstream genes that were under the regulatory control of GAL4 upstream activating sequences (UASs): the HIS3 gene to confer histidine prototrophy, the ADE2 gene to confer adenine prototrophy and the lacZ gene to confer β‐galactosidase activity. Constructs demonstrating efficient splicing in vivo were selected by their ability to complement the metabolic deficiencies in the host strain, and allow growth on selective media. These constructs were then subjected to a second plating assay that assured that the observed prototrophy was the result of trans‐splicing with the desired target transcript. A more rigorous test for specificity, in which the GAL4‐derived sequences are replaced with sequences encoding a potent cytotoxin, is also described. This in vivo genetic selection scheme, unlike existing computer‐based and in vitro screening approaches (14), enables rapid identification of ribozymes optimized for a particular target from large libraries in the complex cellular environment. MATERIALS AND METHODS Plasmid construction Plasmids were constructed by standard procedures (15,16). A CMV coat protein (CP)‐green fluorescent protein (GFP) target gene and a mutant derivative were isolated as BamHI–SacI fragments from plasmids pCMV‐GFP and pMUT‐GFP, respectively (7,11). Each was sub‐cloned in to the same sites of pVT103‐U (17) between the ADH1 promoter and terminator sequences to create pCMV‐GFP(URA3) and pMUT‐GFP(URA3). For expression of the ribozymes, an SphI fragment containing the ADH1 promoter and terminator sequences from pVT103‐U was inserted into the PvuII sites of pRS424 (18) by blunt‐end ligation. In the resulting plasmid, pADH424, the ADH1 promoter directs transcription from the f1‐ strand (i.e. transcription is from the same strand as the TRP1 allele). The previously described ribozymes genes (11) with 3′ exon sequences derived from diphtheria toxin A (DTA) chain were inserted into pADH424 as BamHI–SacI fragments to create p2ATcisRzDTA, p2ATΔP5RzDTA, p2AT9RzDTA, p2AT54RzDTA and p2AT302RzDTA. A new 3′ exon encoding amino acids 2–147 of the yeast GAL4 protein fused to amino acids 412–490 of the herpes simplex virus VP16 protein (19) was PCR amplified from plasmid pBin35S‐mGAL4/VP16+UASmGFP5er (J. Haseloff, unpublished) using a 5′ primer with sequence ACTAGGTACCCAACAATAAAGCTCCTGTCCTCC and a 3′ primer with sequence ACTAACTAGTGGATCCTACCCACCGTACTCG. The PCR product was digested with KpnI and SpeI (italicized), sub‐cloned into the same sites of pET17B (Novagen), and subsequently digested with KpnI and PstI. This gene fragment then replaced the DTA 3′ exons in the above plasmids as KpnI–PstI fragments to create pcisRzGVP, pΔP5RzGVP, p9RzGVP, p54RzGVP and p302RzGVP. A XhoI site internal to the GAL4‐VP16 3′ exon (GVP) was removed from p54RzGVP by site‐directed mutagenesis (20) using an oligonucleotide with sequence CCTCCTGATCTTCCCTaGAGAGGACCTCGACATGATCC (mutated nucleotide is in lower case). In doing so, codon usage was improved for yeast, without altering the amino acid sequence. The RzGVP fusion gene was then PCR amplified with a 5′ primer corresponding to a sequence in the ADH1 promoter, GCACAATATTTCAAGCTAT, and a 3′ primer with sequence CCAGCTGCTGCAGACCCACCGTACTCGTCAATTCCAAG, removing the BamHI site from the 3′ end of the gene fragment. This PCR product was digested with XhoI (downstream of the 5′ primer binding site in the template) and PstI (italicized), and sub‐cloned into the same sites of pcisRzDTA (11) to create pcisRzGVP(ΔXΔB). This cis‐splicing construct was introduced into yeast, and the GVP 3′ exon was functionally tested (not shown). The functional exon was then sub‐cloned into p54RzGVP as a KpnI–PstI fragment to create p54RzGVP(ΔXΔB). Ten additional ribozyme constructs, pARzGVP, pBARzGVP, pCBARzGVP, pDCBARzGVP, pBRzGVP, pCBRzGVP, pDCBRzGVP, pCRzGVP, pDCRzGVP and pDRzGVP, were created by introducing new EAS regions into p54RzGVP(ΔXΔB). The new sequences were created by PCR using pMUT‐GFP as the template, and appropriate combinations of the following oligonucleotides: A, GACTTAACTCGAGACCAGTGCTGGTCGTAACCG; B, GACTTAACTCGAGAAAACTACCTGTTCCATGGC; C, GAC TTAACTCGAGCCCTCGTCAACAGGATCGAGC; D, GAC TTAACTCGAGAATACTCCAATTGGCGATG; A′, CTTAAGGATCCTCCAGTAGTGCAAATAAATTTAAGG; B′, CTTAAGGATCCTGTCTCCCTCAAACTTGAC; C′, CTTAAGGATCCTTTTGTTGATAATGATCAGCG; and D′, CT TAAGGATCCCTTATTTGTATAGTTCATCC (i.e. A and A′ created the ‘A’ EAS, A and B′ created the ‘BA’ EAS, A and C′ created the ‘CBA’ EAS, etc.). The resulting PCR products were digested with BamHI and XhoI (italicized), and sub‐cloned into the same site of p54RzGVP(ΔXΔB) (see Fig. 1 for a graphic representation of target and ribozyme genes used in this study). Test for prototrophy Yeast manipulations were by standard protocols (21,22). Yeast strain PJ69‐4a (MATatrp1‐901 leu2‐3 112 ura3‐52 his3‐200 gal4D gal80D LYS2::GAL1‐HIS3 GAL2‐ADE2 met2::GAL7‐lacZ) (23) was cultured on supplemented minimal medium (SMM; synthetic dextrose minimal medium supplemented with adenine sulfate, uracil, l‐tryptophan, l‐histidine and l‐methionine, each at 20 mg/l and l‐leucine at 100 mg/l) (21). 5‐Fluoro‐orotic acid (5‐FOA) (Sigma) was used at 1 g/l, and X‐gal indicator plates, and β‐galactosidase assays, were as described previously (15). PJ69‐4a was initially transformed with pCMV‐GFP(URA3) and pMUT‐GFP(URA3), and selected on –URA medium. The RzGVP constructs were then individually introduced, and selected on –URA –TRP medium. Single colonies were inoculated into –URA –TRP broth, and grown overnight at 30°C. The cultures were then diluted, and 1 × 104 cells of each strain were spotted on the selective media indicated in Figure 3. RNA was isolated (15) from cells grown overnight in –HIS medium, and reverse transcribed with Superscript RT II (Life Technologies) using the manufacturer’s instructions. Trans‐spliced cDNA was PCR amplified using a 5′ primer corresponding to a sequence in the CMV target gene, TCCAAGGAGATATATAACAATGC, and a 3′ primer corresponding to a sequence in GVP region of the ribozymes, ACGTCCTCGCCGTCTAAGTGGAG. Amplified cDNA was sequenced at the Cornell BioResource Center with the CMV target primer. Genetic selection Equal aliquots of the 13 trans‐splicing RzGVP constructs (Fig. 2B) were mixed to create a RzGVP library and transformed en masse into PJ69‐4a harboring either pCMV‐GFP(URA3) or pMUT‐GFP(URA3). Five percent of the transformation mixture (∼1 × 106 cells) was plated to –URA –TRP to determine transformation efficiency, and the remainders were plated to –URA –TRP –HIS medium to select for effective splicing events. After 5 days of growth on –URA –TRP –HIS, 24 colonies were randomly selected, patched to –HIS medium, and grown for 2 days. Each was then patched to –HIS +5‐FOA and assayed for target‐independent growth after 5 days. 5‐FOA is a toxic analog of a compound in the uracil biosynthetic pathway, and only cells that have lost the URA3‐encoding target plasmid are able to grow in its presence. To characterize the RzGVP genes that conferred prototrophy for histidine, yeast strains from –HIS plates were inoculated into –TRP broth and grown overnight. Total DNA was isolated, and transformed into Escherichia coli strain KC8 (pyrF, leuB600, trpC, hisB463) (15). KC8 cells transformed with RzGVP constructs were selected for tryptophan prototrophy on M9 minimal medium supplemented with uracil, l‐histidine and l‐leucine (40 mg/l each), and ampicillin (100 mg/l). Plasmid was isolated and analyzed by restriction digest from four KC8 colonies obtained from each yeast strain prototrophic for histidine. Stringent test for specificity The GVP 3′ exons in constructs pARzGVP, pBARzGVP, pCBARzGVP and pDCBARzGVP were replaced with the DTA 3′ exon from pcisRzDTA (11) as KpnI–PstI fragments to create constructs pARzDTA, pBARzDTA, pCBARzDTA and pDCBARzDTA. Growth rates of PJ69‐4a cultures harboring these constructs, as well as constructs pADH424, p2ATcis RzDTA, p2ATΔP5RzDTA, p2AT9RzDTA, p2AT54RzDTA and p2AT302RzDTA (see above), were measured by taking OD600 readings of –TRP broth cultures maintained in log phase for at least six generations. RESULTS Ribozyme design The trans‐splicing ribozymes used are based on the group I intron of the large ribosomal RNA subunit of Tetrahymenathermophila. This intron autocatalyzes its own excision and ligation of the exon sequences via two sequential transesterification reactions (6). The reaction mechanism of trans‐splicing is assumed to be the same as for cis‐splicing, with the exception that the 5′ exon is on a separate molecule from the intron and 3′ exon, necessitating an intermolecular association via base pairing (compare Fig. 1A and B). General principles for the design and construction of trans‐splicing ribozymes, and details of the reaction mechanism, were described previously (7,11). Ribozymes designed to splice GAL4‐derived transcription activator sequences into the CP mRNA of CMV were created in both cis‐ and trans‐splicing configurations (Fig. 1A and B). Ribozymes targeted to CMV sequences are desirable as CMV is a prevalent agricultural pathogen with a host range exceeding 800 species, and is recognized as the causal agent of numerous disease epidemics worldwide (13). To facilitate monitoring the target in the yeast assay, the CP sequences were fused in frame to sequences encoding the GFP (CMV‐GFP; Fig. 1B). The sole sequence requirement in the target mRNA is a uridine that forms a wobble base pair with a guanosine in the internal guide sequence (IGS) of the ribozyme, thereby defining the 5′ splice site (24,25). A uridine 10 nt downstream of the CP AUG initiation codon was chosen as the 5′ splice site. Selecting a uridine close to the initiation codon minimizes the N‐terminal fusion produced after splicing to the 3′ exon. Furthermore, selecting a splice site close to the 5′ terminus of the target transcript may facilitate helix formation between the target and the ribozyme (4). An inactive control target was also created to gauge ribozyme specificity by mutating this uridine to a guanosine, thereby preventing recognition of the 5′ splice site (MUT‐GFP). Sequences in the intron IGS were altered from wild type Tetrahymena sequences to allow formation of an artificial helix P1 with the CMV target while maintaining the essential U:G wobble base pair at the 5′ splice site (Fig. 1A and B). A splicing‐deficient control ribozyme was created to monitor splicing‐independent expression of the 3′ exon. This construct, pΔP5RzGVP, is identical to pcisRzGVP (Fig. 1A), except that the P5abc region of the intron core, which is essential for splicing avtivity (26), is deleted. The 3′ exon encoding a GAL4‐derived transcription activator was fused downstream of the ribozyme so that splicing would result in an in frame fusion to the target coding sequence (Fig. 2C). The GAL4‐derived 3′ exon does not contain an AUG start codon which otherwise may contribute to spurious expression. Sequences at the 5′ end of the 3′ exon and in the IGS were adjusted to allow formation of a P10 duplex (Fig. 1A and B), which is hypothesized to facilitate alignment of the exons (27), and is demonstrated to improve trans‐splicing efficiency (7). Previous studies indicated that increasing the length of the EAS improves both trans‐splicing efficiency and specificity (7,11). To test the universality of this finding, the EAS regions in this work ranged from 0 nt (p9RzGVP) to 764 nt (pDCBARzGVP) (Fig. 1B). Furthermore, previous trans‐splicing ribozyme designs incorporated an unpaired region between helix P1 and the EAS/target duplex to potentially facilitate formation of helix P10 after the first transesterification reaction. To gauge the tolerated gap between helix P1 and the EAS/target duplex, the EASs of the ribozymes were designed to leave target sequences ranging from 5 to 605 nt unpaired, while the unpaired ribozymes sequences were maintained at 10 nt in all constructs containing an EAS (Fig. 1B). Complementation of metabolic deficiencies Yeast was selected as a eukaryotic model organism for developing and assaying the trans‐splicing ribozymes since yeast is easy to manipulate, grows rapidly and well developed genetic tools are available. In addition, yeast shares many aspects of gene expression and cellular organization with higher eukaryotes that may affect ribozyme efficacy, but cannot be addressed in prokaryotic systems. The yeast strain used, PJ69‐4a (23), has selectable marker genes downstream of GAL4‐recognized UASs that are expressed only in the presence of a GAL4‐derived transcription activator: the GAL1‐HIS3 gene to confer histidine prototrophy, the GAL2‐ADE2 gene to confer adenine prototrophy and the GAL7‐lacZ gene to enable β‐galactosidase assays (Fig. 1D). PJ69‐4a cells harboring either the target construct, pCMV‐GFP(URA3), or mutant target construct, pMUT‐GFP(URA3), were individually transformed with each of the 15 RzGVP plasmids, and selected on –URA –TRP. Individual colonies were inoculated into –URA –TRP broth, grown overnight and 1 × 104 cells of each were spotted onto selective media (Fig. 2). Growth of all strains was rapid on –URA –TRP medium, as expected since selection is for the presence of target and ribozyme plasmids only (Fig. 2B). The growth rate of strains harboring pcisRzGVP was reduced relative to the others (not shown), suggesting expression of the GAL4‐VP16 transcription activator from a cis‐splicing construct is strong enough to be mildly toxic. A reduction in growth rate resulting from high expression levels of GAL4‐VP16 is documented (28). On media indicative of GAL4 activity, cells harboring pcisRzGVP were prototrophic for histidine and adenine, and β‐galactosidase activity was observed (Fig. 2C, D and E, respectively), demonstrating efficient activation of the GAL1‐HIS3, GAL2‐ADE2 and GAL7‐lacZ indicator genes. As expected for this cis‐splicing construct, GAL4 activity was evident in the presence (Fig. 2C–E) and absence of target construct, pCMV‐GFP(URA3), as evident by growth on indicator medium containing 5‐FOA (Fig. 2F). Importantly, growth on media indicative of GAL4 activity was not observed with cells expressing the inactive, internally deleted ribozyme, ΔP5RzGVP, indicating that complementation of the metabolic deficiencies was strictly dependent on delivery of GAL4 activity via splicing. These results demonstrate that (i) the modifications made to sequences flanking the intron catalytic core do not prevent splicing activity, (ii) splicing yields a translatable chimeric mRNA encoding a GAL4‐derived transcription activator and (iii) the transcription activator with a short CMV CP‐derived peptide fusion is active in binding DNA at GAL4‐recognized UASs, and is able to promote transcription of the downstream genes. For cells harboring trans‐splicing ribozymes, growth on media indicative of GAL4 activity was observed in conjunction with the proper target (Fig. 2C and D), but not with the mutant target [pMUT‐GFP(URA3); not shown], nor in the absence of target (Fig. 2F). These results demonstrate a strict reliance on the target mRNA for expression of the GAL4 sequences fused to the trans‐splicing ribozymes. On –URA –TRP –HIS medium (Fig. 2C), the relative growth rates of strains exhibiting growth were p302RzGVP > pARzGVP > pBARzGVP > p54RzGVP. The EAS lengths in these constructs are 293, 200, 401 and 45 nt, respectively. No growth was evident in the other strains. On –URA –TRP –ADE medium (Fig. 2D), growth was only observed with the trans‐splicing constructs p302RzGVP, and pARzGVP (p302RzGVP > pARzGVP). The GAL2‐ADE2 gene of PJ69‐4a is a more stringent marker than the GAL1‐HIS3 gene (23), and it is thus not surprising that strains with limited growth on –URA –TRP –HIS medium were unable to propagate on –URA –TRP –ADE medium. Figure 2E demonstrates that delivery of the GAL4 activity via trans‐splicing also activates the lacZ gene from the GAL7 promoter. Values obtained from quantitative β‐galactosidase assays supported the findings obtained by observing growth rates on –URA –TRP –ADE medium: p302RzGVP confers greater GAL4 activity in the presence of the target than the other trans‐splicing constructs. To establish the accuracy of trans‐splicing, total RNA was isolated from cultures harboring the target construct pCMV‐GFP(URA3), and either p302RzGVP or pARzGVP. RT–PCR was performed with a 5′ primer corresponding to a sequence in the 5′ exon of the target, and a 3′ primer corresponding to a sequence in the GVP 3′ exon of the ribozyme. The amplified cDNA products were sequenced, and found to contain the expected sequence at the splice junction (AUG GAC AAA U/UU AGG UAC C; Fig. 1C). Genetic selection assay Based on the finding that the ribozymes conferred target‐dependent growth with different efficiencies, a genetic screen for selecting optimized trans‐splicing constructs from a ribozyme library was established. It was predicted that colonies harboring the most efficient ribozyme constructs would predominate on selective media, relative to less efficient constructs. The 13 trans‐splicing ribozyme constructs were mixed in equal proportions to create a ribozyme library. This library was then introduced to PJ69‐4a expressing either the proper target or the mutant target, and the number of colonies growing on selective media was scored (Table 1). Colony development on –URA –TRP indicated that 1.3 × 105 total transformants were obtained for strains harboring either the proper or mutant target. For the pMUT‐GFP(URA3) mutant target strain, no colonies developed on –URA –TRP –HIS medium, as expected from the established specificity of the trans‐splicing reaction. For the pCMV‐GFP(URA3) target strain, 1140 colonies became established after 5 days of growth. Twenty‐four of these were selected at random and patched to –HIS medium, where growth was rapid (Fig. 3A). Each was then transferred to –HIS +5‐FOA medium to assay for the maintenance of histidine prototrophy after loss of the target plasmid (Fig. 3B). Twenty‐three strains did not show growth after 5 days incubation, indicating a reliance on the URA3 encoding target plasmid for complementation of the histidine deficiency. One strain became established on –HIS +5‐FOA medium, indicating target‐independent histidine prototrophy. Since each of the plasmids in the ribozyme library was individually demonstrated to confer histidine protrophy in a target‐dependent fashion, prototrophy in this strain probably resulted from a rare DNA rearrangement. Ribozyme‐encoding plasmids were rescued from each of the 24 histidine prototrophs, and identified by restriction digest analysis. The 23 strains dependent on the target for histidine prototrophy contained the most efficient construct: p302RzGVP. This result demonstrates the effectiveness of phenotypic conversion to prototrophy for selecting the most efficient ribozymes from a library of constructs. The splice junction was characterized in two of the cultures that demonstrated target‐dependent growth on –HIS medium by RT–PCR and sequencing. The amplified cDNA had the expected sequence (AUG GAC AAA U/UU AGG UAC C; Fig. 1C). Counter selection against indiscriminate splicing The 13 trans‐splicing ribozymes analyzed in this study did not complement the metabolic deficiencies of the yeast host strain in the absence of the intended target (Fig. 2F). Therefore, based on this criterion of phenotypic conversion to prototrophy, each demonstrated high specificity. However, in a library of greater complexity, target‐independent expression of the 3′ exon might occur by either non‐specific cis‐ or trans‐splicing. For non‐specific cis‐splicing, at least three criteria must be met: (i) a region upstream of the ribozyme IGS (i.e. the EAS) must be able to fold back and base pair with the IGS to form a P1 helix; (ii) this unintended helix must contain a U wobble base paired with a G in the IGS at position 4, 5 or 6 of the P1 stem to define the 5′‐splice site (25,29); and (iii) for translation of the spliced RNA, sequences upstream of the spurious 5′ splice site must contain a potential AUG start codon positioned such that it is in frame with the 3′ exon following splicing. Implicit in the third criterion is that there are no stop codons between the AUG and the 3′ exon‐encoded open reading frame of the splice product. For non‐specific trans‐splicing to occur, an unintended helix P1 must be able to form between the ribozyme IGS and a separate RNA molecule, followed by criteria (ii) and (iii) discussed for cis‐splicing. It is worth noting that available evidence suggests that the presence of an EAS minimizes indiscriminate interactions between the ribozyme IGS and other transcripts (11). Nevertheless, since expression of the 3′ exon‐encoded gene could theoretically occur via indiscriminate splicing, a practical genetic selection scheme for optimized trans‐splicing constructs must be able to differentiate between target‐specific and non‐specific events. Therefore, the target gene was maintained on a plasmid encoding the URA3 gene, and could be eliminated by plating to medium containing 5‐FOA. Strains emerging from a ribozyme library on –HIS medium, which are also able to grow on –HIS +5‐FOA medium, would harbor indiscriminate ribozymes, whereas those unable to grow on –HIS +5‐FOA medium would contain ribozymes exhibiting high target specificity and efficiency. To test whether counter selection on 5‐FOA medium can effectively differentiate between desirable target‐dependent ribozymes, and undesirable indiscriminate ones, the trans‐splicing ribozyme library was ‘spiked’ with the self‐splicing construct, pcisRzGVP, and the genetic selection assay was repeated (Table 1). Six colonies growing on –URA –TRP –HIS medium were streaked to –HIS media, and subsequently to –HIS +5‐FOA medium (Fig. 3C and D). Of these six strains prototrophic for histidine, two did not grow on –HIS +5‐FOA medium, indicating a dependence on the target plasmid for prototrophy. The ribozyme plasmid identified in these two was the most efficient trans‐splicing construct, p302RzGVP. Therefore, target‐dependent and ‐independent expression of the GAL4 encoding 3′ exon is readily differentiated by counter selection for the target plasmid. Negative selection Although the trans‐splicing ribozymes described are specific enough to prevent target‐independent growth on selective media (Fig. 2F), a more stringent assay for scrutinizing spurious expression of the 3′ exon may be desirable in certain applications. Previous work established that low levels of 3′ exon expression could be monitored by using a 3′ exon encoding the DTA chain polypeptide, and measuring growth rates in liquid media relative to appropriate controls (11). In this study, we used this growth‐rate assay to establish the effect of the EAS on minimizing 3′ exon expression, since it is hypothesized that increasing EAS length improves specificity by ‘shielding’ the active site from non‐specific transcripts (11). The GVP 3′ exons were therefore replaced with DTA sequences, and the growth rates of cultures harboring p2ATΔP5DTA, p2AT9RzDTA, p2AT54RzDTA, pARzDTA, p2AT302RzDTA, pBARzDTA, pCBARzDTA and pDCBARz DTA were established in the absence of target relative to a strain harboring the pADH424 parental plasmid (Fig. 4). The doubling time of cultures harboring p2ATΔP5DTA was 5% greater than those harboring pADH424. Ribozymes derived from the T.thermophila rRNA intron that lack the P5abc region are inactive (11,26). The increase in doubling time is therefore independent of splicing. Cultures expressing ribozymes fused to GFP sequences do not demonstrate growth reductions relative to controls, indicating that the ribozyme itself is not toxic (B. G. Ayre, unpublished). The growth rate reduction in cultures harboring p2ATΔP5DTA is thus attributed to spurious expression of the DTA sequences. Translation initiation at a ribozyme internal AUG sequence, in conjunction with frame shifting events, is the simplest explanation for ‘leaky’ expression of the 3′ exon. For the trans‐splicing ribozymes, an inverse trend was observed between EAS length and growth rate: shorter EAS sequences had longer doubling times (Fig. 4). Cultures expressing ribozymes with EAS lengths of 0 nt (p2ATRzDTA), 45 nt (p2AT54RzDTA) and 200 nt (pARzDTA) grew slowly relative to strains expressing the inactive ribozyme (p2ATΔP5DTA). These reductions in growth rates are thus ribozyme dependent, but target independent, and may result from indiscriminate splicing. Cultures expressing a ribozyme with 293 nt of EAS (p2AT302RzDTA) grew at a rate equivalent to p2ATΔP5DTA cultures, suggesting that indiscriminate splicing is minimized. Significantly, strains expressing ribozymes with 401, 601 and 764 nt of EAS (pBARzDTA, pCDARzDTA and pDCBARzDTA, respectively) grew faster than the inactive ribozyme control, suggesting that the EAS region not only limits non‐specific splicing but may also buffer potential toxicity by minimizing read‐through effects. These results demonstrate the utility of this growth‐rate assay for monitoring subtle differences in 3′ exon expression among different constructs with high sensitivity. DISCUSSION This work arose from the question of whether or not trans‐splicing ribozymes could be used as a general mechanism for delivering new gene activities dependent on the presence of a particular target transcript. We previously demonstrated that similar ribozymes could deliver sufficient cytotoxic gene activity to prevent colony formation in yeast, and could therefore potentially be used to specifically ablate virus‐infected or malignant cells (11). For the work presented here, it was rationalized that transcription factors, like cytotoxins, are sensitive biological indicators, with a small number of molecules being sufficient to activate high levels of downstream gene expression. If trans‐splicing ribozymes could deliver sufficient quantities of a transcription factor, the transcription factor could in turn drive the expression of virtually any gene of interest at biologically relevant levels. To explore this possibility, ribozymes targeted against CMV CP sequences were created with 3′ exon sequences encoding a GAL4‐derived transcription factor. Splicing was expected to result in a chimeric mRNA consisting of the transcription factor sequences fused in frame with the AUG start codon of the CMV CP target. The translation product of this mRNA was in turn expected to bind at GAL4‐recognized UASs, and induce transcription of the downstream genes. The downstream genes of interest were HIS3 to confer histidine prototrophy, ADE2 to confer prototrophy for adenine and lacZ to confer β‐galactosidase activity (Figs 1 and 2). To establish the assay, and demonstrate its utility for rapidly assessing ribozyme function in eukaryotic cells, a series of ribozymes were constructed with EASs ranging from 0 nt (p9RzGVP) to 764 nt (pDCBARzGVP) (Fig. 1B). Previous work indicated that increasing the length of EAS improved both trans‐splicing efficiency (7) and specificity (11). Con sistent with previous findings, increasing the length of the EAS to 293 nt resulted in increased levels of gene delivery. However, beyond this length, efficiency dropped. Trans‐splicing activity was observed with an EAS of 401 nt, but not with 601 nt or 764 nt (Fig. 2C). From this result, we conclude that the length of antisense sequence does not contribute directly to ribozyme efficiency. More likely, the EAS provides one or more short stretches of solute‐exposed nucleotides which are able to base pair with solute‐exposed nucleotides in the target transcript. This initial pairing would anchor the ribozyme to the target, and duplex formation between the two molecules would then align the 5′ splice site of the target with the ribozyme active site to enable the splicing reaction. Similar mechanisms of helix formation across important regulatory sequences from an initial site of nucleation are well characterized in natural antisense interactions (30). Based on this theory, we propose that the ribozyme with 293 nt of extended antisense sequence is the most efficient because it serendipitously contains nucleotides which are best able to interact with the target. Longer antisense sequences would contain these sequences, as might shorter sequences, however the nucleotides may not be available for interaction with the target due to unique secondary and tertiary structures within the transcripts. If this model is correct, more efficient ribozymes with shorter antisense sequences should be possible, providing that the antisense sequences are solute‐exposed and complementary to solute‐exposed target sequences. However, empirical design and in vivo testing of ribozymes is laborious, and although several computer‐based and in vitro technologies have been applied to optimizing ribozymes (14,31), they cannot predict the behavior of a transcript in the complex cellular environment. A more efficient approach would be to create a population of constructs and use in vivo genetic selection to identify individuals with desirable characteristics from the pool. Phenotypic conversion from histidine auxotrophy to prototrophy in yeast was therefore used to select ribozymes with high affinity for the CMV CP‐derived target from a library of constructs. Of the strains emerging from the genetic selection assay, 24 were analyzed in detail, and 23 harbored p302RzGVP. This result correlates well with those obtained when the constructs were tested individually: yeast strains harboring the target plasmid and p302RzGVP became established on selective media more rapidly than those harboring the other trans‐splicing constructs (Fig. 2C). Selection for adenine prototrophs could similarly be used (Fig. 2D), and assaying β‐galactosidase activity (Fig. 2E) provides a further measure of relative ribozyme activity. The trans‐splicing ribozyme library was created by directional cloning of PCR‐amplified target gene segments, such that each ribozyme contained target sequences in the antisense orientation. This approach allowed precise analysis of defined constructs, but limited library complexity. Libraries of greater complexity may contain ribozymes with improved target affinity since each unique transcript is expected to have different solute‐exposed regions due to variations in secondary and tertiary structures, or protein interactions. To create libraries of greater complexity, target sequences could be sheared, or digested with Dnase I, and cloned upstream of the ribozyme catalytic core. Alternatively, a random sequence library could be constructed from synthetic oligonucleotides. The same library could then be used to select efficient interactions with any target transcript of interest, and the interactions would not necessarily rely on strict Watson–Crick base pairings. Non‐canonical interactions, such as tetra‐loop interactions with the minor groove of double‐stranded RNA (32,33), can have greater stability than standard Watson–Crick interactions. Furthermore, most naturally occurring antisense interactions ‘display’ antisense sequences in a loop structure terminating a duplexed stem (30,34). Displaying random sequences at the end of a stem–loop structure incorporated into the EAS region may similarly improve interaction between the ribozyme and the target. Random oligonucleotide libraries for selecting optimal target sites in vitro are described with 5 × 109 individual ribozyme genes (14). Libraries for in vivo screening could in principle be scaled to any size, but in practice would be subject to the same constraints that apply to yeast ‘two‐hybrid’ protein interaction screens (23). Initial experiments with random sequence libraries are encouraging (B. G. Ayre, unpublished). Although a number of strategies could be used to create a ribozyme library, some individuals within the library may exhibit target‐independent cis‐ or trans‐splicing, and result in false positives. A practical genetic selection scheme for identifying efficient trans‐splicing constructs from a large pool must therefore readily differentiate target‐dependent trans‐splicing from non‐specific events. For this reason, the CMV‐GFP target gene was cloned into a yeast plasmid encoding the URA3 gene, and could be selected against by plating cells on media containing 5‐FOA. In an experiment to select efficient trans‐splicing ribozymes from a library containing cis‐splicing constructs, it was found that transferring colonies from –HIS medium to –HIS +5‐FOA medium was effective in differentiating target‐independent events from legitimate trans‐splicing (Fig. 3). Yeast strains that were not able to grow on –HIS +5‐FOA medium harbored the most efficient trans‐splicing construct, p302RzGVP (Table 1). Trans‐splicing ribozymes incorporating GAL4‐derived sequences as the 3′ exon did not confer growth to yeast on selective media in the absence of the specific target (Fig. 2F), and are therefore specific as defined by the criterion of phenotypic conversion to prototrophy. However, to assess non‐specific expression of the 3′ exon with greater stringency, a very sensitive and straightforward assay was employed by replacing GAL4 3′ exon sequences with those encoding DTA chain, and measuring growth rates in liquid culture. In this assay, spurious expression of the 3′ exon was evident as increases in culture doubling time (Fig. 4). Some of this expression was independent of ribozyme activity, as indicated by the increased doubling time of strains containing the splicing‐deficient construct, p2ATΔP5DTA. Ribozyme‐ independent expression may result from non‐canonical initiation, stop codon read‐through, or frame‐shift translation of the 3′ exon (there is no in frame AUG). With active trans‐splicing RzDTA constructs, a general inverse trend was observed between the length of the EAS and spurious 3′ exon expression. Yeast harboring constructs with relatively short EAS regions (i.e. p2AT9RzDTA, p2AT54RzDTA and pARzDTA) demonstrated 3′ exon expression exceeding that of the splicing‐deficient control, suggesting indiscriminate splicing between these ribozymes and non‐target transcripts. Ribozymes with longer EAS regions demonstrated 3′ exon expression equivalent to, or less than, the splicing‐deficient control, supporting the premise that increasing the length of the EAS minimizes ribozyme interaction with non‐targeted sequences. It was previously argued that longer EAS regions may form fortuitous secondary and tertiary interactions that exclude non‐specific mRNAs from entering the active site. In contrast, interaction between the intended target and the EAS would initiate formation of an extended duplex and resolve these structures. The 5′ splice site of the intended target would then enter the ribozyme active site, and accurate splicing would proceed (11). It is noteworthy that ribozymes with the longest EAS regions, pBARzDTA, pCBARzDTA and pDCBARzDTA, demonstrated 3′ exon expression levels less than that of the splicing‐deficient control, indicating the EAS also minimizes splicing‐independent effects. To summarize, we demonstrate that trans‐splicing ribozymes can be used to deliver new transcription factors to living cells dependent on the presence of an intended mRNA target. The transcription factor can in turn drive the expression of a chosen gene, or genes, at biologically relevant levels, and have profound effects on the host organism. The array of genetic tools available and ease of manipulation make yeast an attractive model for developing RNA‐based therapeutics, yet yeast has proved recalcitrant to trans‐acting ribozyme technologies (35). Our results demonstrate functionality in yeast, and the specific delivery of growth‐promoting gene activities presents a practical approach to applying in vivo genetic selection to ribozyme development in a eukaryotic host. Optimized ribozyme designs could then be readily introduced to higher eukaryotes for toxin‐mediated ablation, or induction of gene activity by a trans‐activator, dependent on the presence of a specific target mRNA. Work is currently underway to test the trans‐splicing ribozyme in higher plants. We anticipate that this technology will have broad applications in diverse fields such as biotechnology, therapeutics and genomic studies. Furthermore, the genetic screen, as presented, identifies efficient and specific associations between the target and the antisense sequences of the ribozyme. Interaction with an intended target is the first and probably rate‐limiting step in trans‐splicing and trans‐cleaving ribozyme technologies, as well as in RNA‐ and DNA‐based antisense technologies. Trans‐splicing ribozymes for genetic selection of optimal target interactions could thus be adapted to these other applications. For example, a trans‐splicing ribozyme could be converted to a cleaving ribozyme by incorporating a hammerhead domain into the extended antisense region, or the group I intron could simply be deleted to create a provocative antisense design. This technology may therefore provide an in vivo alternative to computer prediction and in vitro selection of ‘optimized’ target sites (14), which do not take into account the inherent structure of RNA in the cellular environment, nor the extent to which proteins prevent RNA–RNA interactions. Apart from optimizing target–ribozyme interactions, the screen could also be used as an alternative to biochemical optimization of the ribozyme itself (36). ACKNOWLEDGEMENTS We thank Andrew Newman and Robert Arkowitz for helpful discussions and materials, and Roisin McGarry for comments on the manuscript. This work was supported by grants from AgrEvo (J.H.) and the USDA (9801617; R.T.). View largeDownload slide Figure 1. Ribozyme design and gene activation. (A) Schematic diagram of the cis‐splicing control ribozyme in pcisRzGVP. 5′ exon sequences are represented in blue, 3′ exon sequences are indicated in green and intron sequences are black, except for the IGS which is indicated in red. The modified P1 helix is shaded, and the modified P10 helix is boxed. Watson–Crick base pairs are indicated by vertical dashes, and G:U wobble base pairs are indicated by dots. Arrows indicate the 5′ and 3′ splice sites. Restriction endonuclease sites: B, BamHI; X, XhoI; K, KpnI; P, PstI. (B) Schematic diagram of the trans‐splicing ribozymes. The horizontal blue line represents the target transcript. Sequences surrounding the 5′ splice site are indicated, with the 5′ exon in uppercase. Regions of the target complementary to the EAS regions of the ribozymes are indicated by A, B, C and D, with the limits of complementarity indicated by nucleotide number (relative to the AUG start codon). The trans‐splicing ribozymes are represented below the target. The name of each construct is on the left, and the corresponding EAS is represented on the right as horizontal bars. The numbers flanking each EAS region correspond to the complementary sequences in the target. Other symbols are as described for (A). (C) Nucleotide and amino acid sequence at the anticipated splice junction (indicated by arrows). 5′ exon sequences are in blue and 3′ exon sequences are in green. (D) Activation of downstream genes after translation of the spliced mRNA. The GAL4‐VP16‐derived proteins bind to GAL4‐recognized UASs (open arrows) and promote transcription of HIS3 to confer histidine prototrophy, ADE2 to confer adenine prototrophy and lacZ for detectable β‐galactosidase activity (solid bars). The figures are not drawn to scale. View largeDownload slide Figure 1. Ribozyme design and gene activation. (A) Schematic diagram of the cis‐splicing control ribozyme in pcisRzGVP. 5′ exon sequences are represented in blue, 3′ exon sequences are indicated in green and intron sequences are black, except for the IGS which is indicated in red. The modified P1 helix is shaded, and the modified P10 helix is boxed. Watson–Crick base pairs are indicated by vertical dashes, and G:U wobble base pairs are indicated by dots. Arrows indicate the 5′ and 3′ splice sites. Restriction endonuclease sites: B, BamHI; X, XhoI; K, KpnI; P, PstI. (B) Schematic diagram of the trans‐splicing ribozymes. The horizontal blue line represents the target transcript. Sequences surrounding the 5′ splice site are indicated, with the 5′ exon in uppercase. Regions of the target complementary to the EAS regions of the ribozymes are indicated by A, B, C and D, with the limits of complementarity indicated by nucleotide number (relative to the AUG start codon). The trans‐splicing ribozymes are represented below the target. The name of each construct is on the left, and the corresponding EAS is represented on the right as horizontal bars. The numbers flanking each EAS region correspond to the complementary sequences in the target. Other symbols are as described for (A). (C) Nucleotide and amino acid sequence at the anticipated splice junction (indicated by arrows). 5′ exon sequences are in blue and 3′ exon sequences are in green. (D) Activation of downstream genes after translation of the spliced mRNA. The GAL4‐VP16‐derived proteins bind to GAL4‐recognized UASs (open arrows) and promote transcription of HIS3 to confer histidine prototrophy, ADE2 to confer adenine prototrophy and lacZ for detectable β‐galactosidase activity (solid bars). The figures are not drawn to scale. View largeDownload slide Figure 2. Ribozyme‐mediated phenotype alteration on indicator media. (A) Legend for the position of each yeast strain harboring pCMV‐GFP(URA3), and the indicated ribozyme construct. (B‐F) Growth of yeast strains on the indicated plates. (E) Quantitative β‐galactosidase values were obtained from four experiments, and variation is expressed as standard deviation. Plates were photographed after 5 days incubation at 30°C, except (E), which was photographed after an additional 2 weeks incubation at 4°C. View largeDownload slide Figure 2. Ribozyme‐mediated phenotype alteration on indicator media. (A) Legend for the position of each yeast strain harboring pCMV‐GFP(URA3), and the indicated ribozyme construct. (B‐F) Growth of yeast strains on the indicated plates. (E) Quantitative β‐galactosidase values were obtained from four experiments, and variation is expressed as standard deviation. Plates were photographed after 5 days incubation at 30°C, except (E), which was photographed after an additional 2 weeks incubation at 4°C. View largeDownload slide Figure 3. Test for target‐independent growth on selective media. (A) Twenty‐four strains transformed with pCMV‐GFP(URA3) and the trans‐splicing ribozyme library patched to –HIS medium. Strains were chosen randomly from the colonies that became established after introducing the ribozyme library to pCMV‐GFP(URA3) cultures and selecting for histidine prototrophy on –URA –TRP –HIS medium. (B) Yeast stains from (A) patched to –HIS +5‐FOA medium. One strain that exhibited target‐independent growth (i.e. growth on medium containing 5‐FOA) is indicated by the arrow. (C) Six strains transformed with pCMV‐GFP(URA3) and the cis/trans‐splicing ribozyme library streaked to –HIS medium. Strains were chosen randomly from those that became established after introducing the ribozyme library to pCMV‐GFP(URA3) cultures and selecting for histidine prototrophy on –URA –TRP –HIS medium. (D) Yeast stains from (C) streaked to –HIS +5‐FOA medium. The two strains that required the target plasmid for histidine prototrophy (i.e. failure to grow in the presense of 5‐FOA) are indicated by asterisks. View largeDownload slide Figure 3. Test for target‐independent growth on selective media. (A) Twenty‐four strains transformed with pCMV‐GFP(URA3) and the trans‐splicing ribozyme library patched to –HIS medium. Strains were chosen randomly from the colonies that became established after introducing the ribozyme library to pCMV‐GFP(URA3) cultures and selecting for histidine prototrophy on –URA –TRP –HIS medium. (B) Yeast stains from (A) patched to –HIS +5‐FOA medium. One strain that exhibited target‐independent growth (i.e. growth on medium containing 5‐FOA) is indicated by the arrow. (C) Six strains transformed with pCMV‐GFP(URA3) and the cis/trans‐splicing ribozyme library streaked to –HIS medium. Strains were chosen randomly from those that became established after introducing the ribozyme library to pCMV‐GFP(URA3) cultures and selecting for histidine prototrophy on –URA –TRP –HIS medium. (D) Yeast stains from (C) streaked to –HIS +5‐FOA medium. The two strains that required the target plasmid for histidine prototrophy (i.e. failure to grow in the presense of 5‐FOA) are indicated by asterisks. View largeDownload slide Figure 4. Relative growth rate of strains expressing ribozymes with DTA sequences as the 3′ exon. Relative doubling times of yeast strains harboring the indicated RzDTA constructs normalized to the parent construct, pADH424 (100%), in the absence of the intended target. The trans‐splicing constructs are arranged by increasing length of EAS. Results are from three independent trials and variation is expressed as standard error of the mean. View largeDownload slide Figure 4. Relative growth rate of strains expressing ribozymes with DTA sequences as the 3′ exon. Relative doubling times of yeast strains harboring the indicated RzDTA constructs normalized to the parent construct, pADH424 (100%), in the absence of the intended target. The trans‐splicing constructs are arranged by increasing length of EAS. Results are from three independent trials and variation is expressed as standard error of the mean. Table 1. Colony formation of strains harboring either pMUT‐GFP(URA3) or pCMV‐GFP(URA3) target plasmids on selective media after introduction of the indicated ribozyme library Plasmids introduced Colony development on SMM media Plasmids recovereda –URA –TRP –URA –TRP –HIS –HIS +5‐FOAb pCMV‐MUT(URA3) + trans Rz library 1.3 × 105 0 n.a. n.a. pCMV‐GFP(URA3) + trans Rz library 1.3 × 105 1140 1/24 23/23 p302Rz‐GVP pCMV‐GFP(URA3) + cis/trans Rz library 1.3 × 105 1200 4/6 2/2 p302Rz‐GVP Plasmids introduced Colony development on SMM media Plasmids recovereda –URA –TRP –URA –TRP –HIS –HIS +5‐FOAb pCMV‐MUT(URA3) + trans Rz library 1.3 × 105 0 n.a. n.a. pCMV‐GFP(URA3) + trans Rz library 1.3 × 105 1140 1/24 23/23 p302Rz‐GVP pCMV‐GFP(URA3) + cis/trans Rz library 1.3 × 105 1200 4/6 2/2 p302Rz‐GVP n.a., not applicable. aProportion of Rz plasmids identified from strains demonstrating target‐dependent growth on –HIS medium. bProportion of analyzed strains prototrophic for histidine independent of the URA‐encoding target plasmid. View Large References 1. Branch,A.D. and Robertson,H.D. ( 1991) Efficient trans‐cleavage and a common structural motif for the ribozymes of the human hepatitis‐delta agent. Proc. Natl Acad. Sci. USA , 88, 10163–10167. Google Scholar 2. Haseloff,J. and Gerlach,W.L. ( 1988) Simple RNA enzymes with new and highly specific endoribonuclease activities. Nature , 334, 585–591. Google Scholar 3. Ojwang,J.O., Hampel,A., Looney,D.J., Wongstaal,F. and Rappaport,J. ( 1992) Inhibition of human‐immunodeficiency‐virus type‐1 expression by a hairpin ribozyme. Proc. Natl Acad. Sci. USA , 89, 10802–10806. Google Scholar 4. Bertrand,E., Pictet,R. and Grange,T. ( 1994) Can hammerhead ribozymes be efficient tools to inactivate gene‐function? Nucleic Acids Res. , 22, 293–300. Google Scholar 5. Branch,A.D. ( 1998) A good antisense molecule is hard to find. Trends Biochem. Sci. , 23, 45–50. Google Scholar 6. Cech,T.R. ( 1990) Self‐splicing of group I introns. Annu. Rev. Biochem. , 59, 543–568. Google Scholar 7. Köhler,U., Ayre,B.G., Goodman,H.M. and Haseloff,J. ( 1999) Trans‐splicing ribozymes for targeted gene delivery. J. Mol. Biol. , 285, 1935–1950. Google Scholar 8. Galloway‐Salvo,J.L., Coetzee,T. and Belfort,M. ( 1990) Deletion tolerance and trans‐splicing of the bacteriophage T4 td intron. Analysis of the P6‐L6a region. J. Mol. Biol. , 211, 537–549. Google Scholar 9. Sullenger,B.A. and Cech,T.R. ( 1994) Ribozyme‐mediated repair of defective mRNA by targeted trans‐splicing. Nature , 371, 619–622. Google Scholar 10. Lan,N., Howrey,R.P., Lee,S.W., Smith,C.A. and Sullenger,B.A. ( 1998) Ribozyme‐mediated repair of sickle β‐globin mRNAs in erythrocyte precursors. Science , 280, 1593–1596. Google Scholar 11. Ayre,B.G., Kohler,U., Goodman,H.M. and Haseloff,J. ( 1999) Design of highly specific cytotoxins by using trans‐splicing ribozymes. Proc. Natl Acad. Sci. USA , 96, 3507–3512. Google Scholar 12. Jones,J.T., Lee,S.W. and Sullenger,B.A. ( 1996) Tagging ribozyme reaction sites to follow trans‐splicing in mammalian cells. Nature Med. , 2, 643–648. Google Scholar 13. Palukaitis,P., Roossinck,M.J., Dietzgen,R.G. and Francki,R.I.B. ( 1992) Cucumber mosaic virus. Adv. Virus Res. , 41, 281–348. Google Scholar 14. Turner,P.C. (ed.) ( 1997) Ribozyme protocols. Methods in Molecular Biology, Vol. 74. Humana Press Inc, Totowa, NJ. Google Scholar 15. Ausubel,F., Brent,R., Kingston,R.E., Moore,D.D., Seidmam,J.G. and Smith,J.A. (eds) ( 1995) Short Protocols in Molecular Biology. 3rd Edn. John Wiley and Sons, Inc., NY. Google Scholar 16. Sambrook,J., Fritsch,E.F. and Maniatis,T. ( 1989) Molecular Cloning. Cold Spring Harbor Laboratory Press, NY. Google Scholar 17. Vernet,T., Dignard,D. and Thomas,D.Y. ( 1987) A family of yeast expression vectors containing the phage f1 intergenic region. Gene , 52, 225–233. Google Scholar 18. Christianson,T.W., Sikorski,R.S., Dante,M., Shero,J.H. and Hieter,P. ( 1992) Multifunctional yeast high copy number shuttle vectors. Gene , 110, 119–122. Google Scholar 19. Sadowski,I., Ma,J., Triezenberg,S. and Ptashne,M. ( 1988) GAL4‐VP16 is an unusually potent transcriptional activator. Nature , 335, 563–564. Google Scholar 20. Kunkel,T.A., Roberts,J.D. and Zakour,R.A. ( 1987) Rapid and efficient site‐specific mutagenesis without phenotypic selection. Methods Enzymol. , 154, 367–382. Google Scholar 21. Kaiser,C., Michaelis,S. and Mitchell,A. ( 1994) Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Google Scholar 22. Guthrie,C. and Fink,G.J. (eds) ( 1991) Guide to Yeast Genetics and Molecular Biology. Methods in Enzymology, Vol. 194. Academic Press, Inc., San Diego. Google Scholar 23. James,P., Halladay,J. and Craig,E.A. ( 1996) Genomic libraries and a host strain designed for highly efficient two‐hybrid selection in yeast. Genetics , 144, 1425–1436. Google Scholar 24. Barfod,E.T. and Cech,T.R. ( 1989) The conserved U‐G pair in the 5′ splice‐site duplex of a group I intron is required in the 1st but not the 2nd step of self‐splicing. Mol. Cell. Biol. , 9, 3657–3666. Google Scholar 25. Doudna,J.A., Cormack,B.P. and Szostak,J.W. ( 1989) RNA structure, not sequence, determines the 5′ splice‐site specificity of a group I intron. Proc. Natl Acad. Sci. USA , 86, 7402–7406. Google Scholar 26. Joyce,G.F., Vanderhorst,G. and Inoue,T. ( 1989) Catalytic activity is retained in the Tetrahymena group I intron despite removal of the large extension of element P5. Nucleic Acids Res. , 17, 7879–7889. Google Scholar 27. Waring,R.B., Towner,P., Minter,S.J. and Davies,R.W. ( 1986) Splice‐site selection by a self‐splicing RNA of Tetrahymena. Nature , 321, 133–139. Google Scholar 28. Berger,S.L., Pina,B., Silverman,N., Marcus,G.A., Agapite,J., Regier,J.L., Triezenberg,S.J. and Guarente,L. ( 1992) Genetic isolation of ADA2—a potential transcriptional adapter required for function of certain acidic activation domains. Cell , 70, 251–265. Google Scholar 29. Murphy,F.L. and Cech,T.R. ( 1989) Alteration of substrate specificity for the endoribonucleolytic cleavage of RNA by the Tetrahymena ribozyme. Proc. Natl Acad. Sci. USA , 86, 9218–9222. Google Scholar 30. Zeiler,B.N. and Simons,R.W. ( 1998) Antisense RNA Structure and Function. In Simons,R.W. and Grunberg‐Manago,M. (eds), RNA Structure and Function. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 437–464. Google Scholar 31. Lan,N., Rooney,B.L., Lee,S.W., Howrey,R.P., Smith,C.A. and Sullenger,B.A. ( 2000) Enhancing RNA repair efficiency by combining trans‐splicing ribozymes that recognize different accessible sites on a target RNA. Mol. Ther. , 2, 245–255. Google Scholar 32. Cate,J., Gooding,A., Podell,E., Zhou,K., Golden,B., Kundrot,C., Cech,T. and Doudna,J. ( 1996) Crystal‐structure of a group‐I ribozyme domain—principles of RNA packing. Science , 273, 1678–1685. Google Scholar 33. Pley,H.W., Flaherty,K.M. and McKay,D.B. ( 1994) Model for an RNA tertiary interaction from the structure of an intermolecular complex between a GAAA tetraloop and an RNA helix. Nature , 372, 111–113. Google Scholar 34. Franch,T., Petersen,M., Wagner,E.G.H., Jacobsen,J.P. and Gerdes,K. ( 1999) Antisense RNA regulation in prokaryotes: Rapid RNA/RNA interaction facilitated by a general U‐turn loop structure. J. Mol. Biol. , 294, 1115–1125. Google Scholar 35. Arndt,G.M. and Atkins,D. ( 1996) RNA catalysis in fungi. In Eckstein,F. and Lilley,D.M.J. (eds), Nucleic Acids and Molecular Biology. Springer‐Verlag, Berlin, Vol. 10, pp. 343–366. Google Scholar 36. Zarrinkar,P.P. and Sullenger,B.A. ( 1999) Optimizing the substrate specificity of a group I intron ribozyme. Biochemistry , 38, 3426–3432. Google Scholar
Promoter‐trapping in Saccharomyces cerevisiae by radiation‐assisted fragment insertionKiechle, Markus;Manivasakam, Palaniyandi;Eckardt‐Schupp, Friederike;Schiestl, Robert H.;Friedl, Anna A.
doi: 10.1093/nar/gnf136pmid: 12490727
Abstract Non‐homologous insertion (NHI) of DNA fragments into genomic DNA is a method widely used in insertional mutagenesis screens. In the yeast Saccharomyces cerevisiae, the efficiency of NHI is very low. Here we report that its efficiency can be increased by γ‐irradiation of recipient cells at the time of transformation. Radiation‐assisted NHI depends on YKU70, but its efficiency is not improved by inactivation of RAD5 or RAD52. In a pilot study, we generated 102 transformant clones expressing a lacZ reporter gene under standard conditions (30°C, rich medium). The site of insertion was determined in a subset of eight clones in which lacZ expression was altered by UV‐irradiation. A comparison with published data revealed that three of the eight genes identified in our screen have not been targeted by large‐scale transposon‐based insertion screens. This suggests that radiation‐assisted NHI offers a more homogeneous coverage of the genome than methods relying on transposons or retroviral elements. Received October 9, 2002; Revised and Accepted October 19, 2002 INTRODUCTION Random insertional mutagenesis is a powerful tool in the analysis of gene function that is amenable to genome‐wide analyses and that has been applied in a variety of organisms (1–4). In addition to studying the phenotype associated with disruption of the target sequence, by insertion of cassettes containing appropriate elements it is also possible to conduct gene expression analyses and to insert tag sequences that can be used for protein isolation or localisation studies. Insertion cassettes are often derived from mobile DNA elements, such as retroviruses or transposons, and efficient insertion into genomic DNA is then mediated by the respective mobilising apparatus (e.g. integrases or transposases). As most mobile elements exhibit a certain degree of target site specificity, uniform coverage of all genes in the genome may be difficult to achieve with this approach. For example, by insertional mutagenesis using Tn3‐derived mini‐transposons a large collection of Saccharomyces cerevisiae mutant clones has been established (5,6). Although, until the end of the year 2001, the site of insertion has been characterised in more than 22 000 insertion clones, less than two‐thirds of the about 6200 yeast genes are represented in this collection (7). In addition to gene‐size dependent biases in targeting efficiency, non‐random insertion of Tn3‐derived transposons (8) and unequal representation of genes in the yeast library mutagenised may account for this effect. To achieve complete coverage of the genome, the introduction of complementary approaches is necessary. One such approach is non‐homologous insertion (NHI; 9) of DNA cassettes not derived from mobile elements, a process which appears to exploit the cellular machinery that normally repairs DNA double‐strand breaks (DSB) by non‐homologous end joining (NHEJ; 10). Successful application of NHI strategies in large‐scale insertional mutagenesis projects has, for example, been described in mammalian cells, plants and Schizosaccharomyces pombe (11–13). Similar approaches in S.cerevisiae have been precluded by the low efficiency of NHI in this organism (14). We sought to establish an efficient NHI‐based method for random insertion of a promoter‐trap cassette containing a lacZ reporter gene in S.cerevisiae. Previous work demonstrated that NHI efficiency increases upon co‐transformation of restriction endonucleases (14–16). Restriction‐enzyme‐mediated integration (REMI) was subsequently used for insertion mutagenesis in a variety of organisms (17). To avoid a sequence bias towards genomic recognition sites of the co‐transformed endonuclease, we here investigate whether γ‐irradiation is suitable for facilitating NHI. Within the yeast genome, DSB induced by sparsely ionising radiation appear to be distributed randomly (18), although small‐scale variations due to radical shielding by DNA‐associated proteins cannot be excluded. Irradiation should therefore result in a largely random distribution of insertion sites. Here, we report that γ‐irradiation shortly before or after the heat‐shock step of transformation enhances the rate of NHI sufficiently to allow its application in a promoter‐trap scheme. To demonstrate proof‐of‐principle, we analysed 2000 NHI transformants for lacZ expression. Among the 102 lacZ expressing clones eight transformants were isolated that exhibited altered lacZ expression after UV‐irradiation. Interestingly, three of the eight insertion sites were located in genes that so far have not been found hit in the more than 22 000 clones obtained in the large‐scale transposon‐based mutagenesis screen (5–7). Thus, radiation‐assisted NHI of linearised plasmids may be suitable as a complementary approach in order to allow full coverage of the S.cerevisiae genome. MATERIALS AND METHODS For optimising radiation‐assisted NHI, plasmids pM150 and pM151 (16) were used. These pUC18 derivatives contain the URA3 gene and differ in the restriction sites present in the MCS. For constructing plasmid pFA6‐kanMX6‐lacZ, a 3.1 kb BamHI/SalI fragment containing the lacZ gene was isolated from plasmid pCM159 (19) and inserted in BamHI/SalI‐digested plasmid pFA6‐kanMX6 (20). Plasmid MKM20 was generated by PCR amplification of the lacZ gene, using pFA6‐kanMX6‐lacZ as a template, with primers lacZ‐C‐BamHI (5′‐CGA ATT CGG ATC CAG CTG AAG CTT CGT ACG‐3′) and lacZ‐N‐BglII (5′‐CGA ATT CAG ATC TAC TGG CCG TCG TTT TAC‐3′), and insertion of the BamHI/BglII‐digested product into the BglII site of pM151. The lacZ allele thus obtained lacks the first 20 nucleotides, including the start codon, and lacZ expression is obtained only upon in‐frame insertion of the BglII‐linearised plasmid MKM20 into expressed genes. NHI experiments were performed in haploid strain RSY12 (MATa leu2‐3,112 his3‐11,15 ura3Δ::HIS3), in which the entire URA3 open reading frame was replaced by the HIS3 gene (14). Strain RSY12 rad5 was generated by transformation of a rad5Δ::kanMX‐lacZ deletion cassette generated by PCR, using primers rad5‐kanMX‐C (5′‐CTA TTC AAA CAG CAT CTG GAT TTC TTC AAT TCT CCT TTT TCT GGG TCA CCC GGC CAG CG‐3′) and rad5‐lacZ‐N (5′‐ATG AGT CAT ATT GAA GAA AGG AAG TTT TTT AAC GAT CCC GTC GTT TTA CAA CG‐3′) on template pAF6‐kanMX6‐lacZ. Generation of strains RSY12 yku70 and RSY12 rad52 has been described earlier (15,21). Yeast cells were transformed using a high efficiency method (22). For transformation, plasmids pM151 or MKM20 were digested to completion with BglII, extracted with phenol, precipitated and dissolved in sterile water. Control transformations, using the same batch of cells, were performed with circular plasmid YEplac195 (23). For radiation‐assisted NHI, transformation reaction mixes were γ‐irradiated at room temperature, using a 60Co‐γ‐source (Atomic Energy of Canada, Ltd) at a dose rate of 10 Gy min–1. If not stated otherwise, irradiation took place immediately after the heat‐shock step of the transformation procedure. Relative transformation frequencies were determined by normalising the rate of transformation of the linearised plasmid (per microgram of DNA) relative to the rate of transformation with the control plasmid. Integrated plasmids and adjacent genomic sequences were recovered by plasmid rescue as described (21), except that several restriction enzymes that do not cut within the plasmid sequence (BamHI, BspEI, NheI and XhoI) were used in parallel digestions of each DNA sample, in order to enhance the efficiency of plasmid rescue. Plasmids with flanking chromosomal DNA, as identified by migration behaviour in agarose gels, were sequenced by Medigenomix GmbH (Martinsried, Germany) using sequencing primers P1 (5′‐AGC GGA TAA CAA TTT CAC ACA GGA‐3′) and P2 (5′‐ACT CCA GCC AGC TTT CC‐3′). Sequencing data were analysed by a BLASTN search on the Saccharomyces Genome Database (SGD; http://genome‐www.stanford.edu/Saccharomyces/). To identify lacZ‐expressing transformants, patches of cells on selection plates were covered with ∼10 ml chloroform and incubated for 5 min to permeabilize cells. After removing the chloroform, plates were air‐dried for 5 min, before they were covered with X‐Gal suspension (1 mg/ml X‐Gal in 1% LMP‐agarose solution, dissolved at 60°C). After solidification of the agarose, plates were incubated at 30°C up to 24 h to allow for colour development. To quantify UV‐induced lacZ expression, cells were grown to logarithmic phase (5 × 106 cells/ml), harvested by centrifugation and re‐suspended at a titre of 1 × 108 cells/ml in liquid holding recovery (LHR) buffer (67 mM K2HPO4/KH2PO4, 100 mM glucose, pH 5). Aliquots of 2 ml were UV‐irradiated (254 nm) in the dark in plastic Petri dishes under constant stirring. Post‐irradiation incubation of irradiated suspensions was performed in the dark at 30°C under constant shaking. At given time points, 1 ml aliquots were removed, and after centrifugation cells were re‐suspended in 0.5 ml Z‐buffer (75 mM Na2HPO4, 50 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4) and immediately frozen at –80°C, where they were kept until detection. For detection, all samples of an experimental series were thawed simultaneously and each 50 µl of the suspensions were used to determine the OD600. The remaining cells were incubated for 1 h at 37°C with 100 µl zymolyase solution (zymolyase T100, Seikagaku Corp., Japan; 0.5 mg in 1 ml Z‐buffer). After completion of lysis, 100 µl of CPRG solution (4 mg/ml Chloro‐Phenol‐Red‐Galactopyranosid in Z‐buffer) were added and incubated at room temperature. Incubation times (tinc) varied between 1 and 180 min, according to the different levels of lacZ expression. Reactions were stopped by adding 200 µl 1 M Na2CO3. After 15 min centrifugation at 15 000 r.p.m., the optical density of the supernatant was measured at 574 nm and 634 nm. The β‐galactosidase activity in Miller units was calculated as: [MU] = [(OD574 – OD634) × 1000] / (OD600 × tinc). Semi‐quantitative screening of lacZ expression was performed similarly, except that cell titres at the time of irradiation were not determined. RESULTS The frequency of NHI events increases upon irradiation We investigated the influence of irradiation on the frequency of NHI of BglII‐linearised plasmid pM151 (16), which contains the yeast URA3 gene, into the genome of strain RSY12, in which the URA3 locus had been deleted (14). The frequency of transformation with plasmid pM151 was normalised with respect to parallel control transformations with a circular episomal plasmid. γ‐Irradiation of cells immediately after the heat‐shock step of the transformation procedure caused an increase in NHI frequency of up to ∼15‐fold (Fig. 1A). Thus, about 150 transformants can be obtained in one transformation experiment, using 7 µg of linearised plasmid. We refer to this method as radiation‐assisted NHI. Comparable effects were seen when cells were irradiated immediately before the heat shock, while the effect declined with increasing duration of the period between irradiation and transformation, until, with periods exceeding 4 h, no positive effect of irradiation could be seen anymore (data not shown). Additional experiments using plasmids linearised with other enzymes showed that radiation‐assisted NHI is more pronounced with fragments terminating in single‐stranded overhangs than with plasmids terminating in blunt ends (data not shown). An increase of NHI frequency with irradiation is not seen in Ku70‐deficient mutants [Fig. 1A; (21)], demonstrating that the insertion of transformed fragments into broken chromosomal DNA involves NHEJ. In general, NHEJ contributes little to the repair of radiation‐induced chromosomal breaks in yeast (10). We hypothesised that the efficiency of radiation‐induced NHI may be increased in mutant strains that exhibit more potent NHEJ. For example, the Rad5 protein, inactivation of which confers moderate radiosensitivity, has been invoked in the regulation of DSB repair pathways in S.cerevisiae. While wild‐type cells repair plasmids gapped within a sequence that is also present on a chromosome almost exclusively by homologous recombination with the chromosomal donor sequence, rad5 mutants rely predominantly on NHEJ‐type repair of these plasmids (24). In contrast to its influence in plasmid repair, however, the rad5 mutation does not appear to have a role in NHI, as, both with and without irradiation, the relative transformation efficiency was not significantly altered in rad5 mutants as compared with the wild‐type strain (Fig.1A). It has been proposed that NHEJ processes and homologous recombination competes for repair of DNA DSB and that the decision of which pathway is used is, at least in part, determined by the proteins initially binding to the DNA ends (25). Thus, by inactivation of Rad52, a DNA end‐binding protein essential for homologous recombination, more ends may become accessible to Ku proteins and the relative frequency of NHEJ events may increase. Indeed, as compared with wild‐type cells, we observe a strong increase in the relative frequency of NHI events in rad52 mutants both with and without irradiation (Fig. 1B). However, since the control transformation with circular plasmids is about 50 times less efficient in the rad52 mutant than in wild‐type [(4.2 + –3.6) × 102 versus (1.9 + –0.5) × 104 transformants/µg plasmid DNA in rad52 mutants and wild‐type, respectively], the absolute number of transformants is reduced in the mutant. Reduced transformation with circular plasmids in rad52 mutant strains has been observed previously (15). In addition, the mutant is much more sensitive to the cell killing effects of radiation (survival of 14 versus 45% at 100 Gy), further reducing the number of transformants obtainable. Application of radiation‐assisted NHI in a promoter‐trapping scheme The bacterial β‐galactosidase gene lacZ is a widely used reporter gene, the expression of which can easily be monitored qualitatively and quantitatively (26). Insertion of a BglII/BamHI‐digested PCR fragment encompassing the lacZ gene into the BglII site of pM151 created plasmid MKM20. The lacZ allele used in our studies lacks the first 25 nucleotides including the start codon. By targeted in‐frame insertion into the chromosomal RAD54 gene, a gene well known for its UV inducibility (27), we verified the suitability of this allele to serve as a reporter gene in analysis of UV‐induced gene induction (data not shown). To use for promoter trapping, plasmid MKM20 was digested with BglII and transformed into strain RSY12 using a high efficiency transformation protocol. Immediately after the heat‐shock step, the cells were irradiated with 50 Gy. A dose of 50 Gy was chosen to achieve both sufficient cell survival (∼60%) and high efficiency of NHI. At this dose, on average 2 DSB/cell are induced (18). Transformation efficiency with plasmid MKM20 was comparable with that of pM151. A total of about 2000 Ura+ transformants were generated and tested by X‐Gal overlay for in‐frame insertion of MKM20 into yeast genes expressed under standard growth conditions (YPD, 30°C). Detectable expression of the inserted lacZ gene was found in 102 transformants (5%). To identify those transformants in which the reporter gene expression was affected by UV‐irradiation, a semi‐quantitative screen was conducted. Cells grown to logarithmic growth phase were re‐suspended in non‐growth buffer, half of the sample was UV‐irradiated (60 J/m2) and the other half served as mock‐treated control. The non‐growth condition was chosen so that changes in gene expression are not due to any secondary effect of radiation‐induced cell cycle arrest. Cells were lysed 2 h after treatment and β‐galactosidase activity was determined using a colourimetric assay. Out of 102 lacZ‐expressing clones, 12 candidates differing at least 2‐fold in expression in treated versus untreated cells were identified and chosen for further analysis. In eight of these, the site of insertion of the promoter‐trapping construct was successfully determined by plasmid rescue (see Table 1). In the remaining clones, plasmid rescue and/or sequencing were not successful (three cases) or the construct was found to be fused to a fragment from the 2 µm plasmid which obviously enabled autonomous replication (one case). In six of the eight events indicated in Table 1, insertion of the reporter construct occurred in‐frame within a coding region. In clone 42, insertion occurred in‐frame within the intron of the UBC13 gene, 46 bp upstream of the branch‐point (28). In clone 66, insertion occurred 22 bp upstream of the BUD4 initiator codon. The next in‐frame ATG codon is located at position –165 (as counted from the BUD4 initiator codon); since translational frameshift events occur readily in yeast (6), alternatively an out‐of‐frame ATG at position –36 may have been used to allow expression of the reporter gene. To investigate whether insertion events differ from events observed spontaneously, we sequenced both insertion junctions (Fig. 2). Similar to events seen in spontaneous NHI (9), in the majority (6/8) of the insertion events at least one junction apparently was facilitated by microhomology between plasmid end and genomic sequence. In five of the eight integrations, no or only one nucleotide of the genomic sequence was lost. Only one event (clone 60) was associated with a large genomic rearrangement (inversion or duplication). An analysis of the kinetics of expression of the lacZ‐fusion proteins in response to UV irradiation with 60 J/m2 verified strong induction (>2‐fold) for fusions involving MRP8, SDS24, NVJ1 and YBR052c, and 1.5‐fold induction for APG1 (Fig. 3). These genes have been found by microarray analyses to be generally stress‐inducible (29–31). For the fusion involving UBC13, β‐galactosidase activity in UV‐irradiated and mock‐treated samples differed by >1.5‐fold, but instead of showing a clear induction, the treated samples rather exhibited a slower decline of β‐galactosidase activity during the post‐treatment incubation than the mock‐treated samples (Fig. 3). In microarray and northern analyses, however, increased amount of UBC13 transcripts after DNA damage have been described (29–33). We assume that the different conditions during post‐treatment incubation (non‐growth versus growth conditions) are responsible for the discrepancies. A slightly accelerated decline of β‐galactosidase activity in treated versus untreated sample was observed during post‐treatment incubation in a clone involving a fusion with BUD4, a gene which by others was described as repressed in response to DNA damage due to cell cycle arrest (30). In contrast to the results of the semi‐quantitative analysis, no significant difference between treated and untreated samples was observed concerning the β‐galactosidase activity of FIR1‐lacZ fusions when assaying the kinetics of expression. Of the genes affected by insertion, only UBC13 is known to be involved in response to DNA damage. To test whether the other genes have an as yet uncharacterised function in damage resistance, UV and γ‐ray sensitivity was tested for all insertion clones. None of the clones (except for the ubc13 mutant) exhibited increased sensitivity (data not shown), confirming a recent report that only a very small proportion of the genes whose expression is altered after DNA damaging treatments actually is involved in conferring resistance to these treatments (34). DISCUSSION The method used to insert a mutagenising fragment into the genome may be a critical factor for successful applica tion of insertional mutagenesis screens. Use of mobile element‐derived mechanisms is highly efficient, but often accompanied by a target bias that precludes uniform coverage of all genomic loci. The alternative method, random insertion of fragments without using mobilising machinery, is still poorly characterised in terms of mechanistic details. Recent evidence supports the notion that NHI is mediated by components of the NHEJ pathway of DSB repair (15,21,35; this work). Here we demonstrate that radiation‐assisted NHI, similar to REMI (9), results in an increased yield of transformants. Optimal enhancement of NHI was seen when irradiation took place around the time of the heat‐shock step in transformation, i.e. the time when the transformed DNA enters the cell. Similar to spontaneous and restriction enzyme‐mediated NHI, radiation‐assisted NHI works better on fragments terminating in single‐stranded overhangs than in blunt‐ended fragments. Also, the junction sites observed in radiation‐assisted NHI were very similar to those observed after spontaneous or enzyme‐mediated NHI in their dependence on microhomologies and generation of small deletions or insertions. Whether large genomic alterations (as seen in clone 60) occur more often in radiation‐assisted NHI than in spontaneous NHI (13) remains to be tested. Our attempts to manipulate the regulatory system of DSB repair by inactivating RAD5 or RAD52 did not improve radiation‐assisted NHI. In line with our results, it has recently been found that impairment of homologous recombination does not enhance NHEJ efficiency in yeast (36). Whether, as suggested by the data of these authors, the frequency of radiation‐assisted NHI is increased in G1/G0 cell populations as compared with logarithmically growing populations, remains to be tested. Expression of the lacZ reporter gene under standard growth conditions (30°C, YPD) was observed in 5% of our transformants. This is about the frequency expected, considering the proportion of coding DNA in S.cerevisiae [about 0.7; (37)], the probability of in‐frame insertion in the correct orientation (0.33 × 0.5), the frequency of essential genes [0.19; (38)] and the fact that expression of some genes under standard conditions may be too low to allow detection. Among the clones expressing lacZ, ∼8% were found to increase expression upon UV‐irradiation. Considering the experimental differences, this value compares well with results obtained in expression studies using microarrays (29–31,34). Our investigation differs from these studies in that we performed post‐irradiation incubation of treated samples and mock‐treated controls under non‐growth conditions. In addition, depending on the site of insertion of the reporter construct, the amount of lacZ‐fusion proteins may not only reflect promoter activity, but, at least in part, also post‐transcriptional and post‐translational regulation of expression. It is known that in many cases protein levels do not correlate with mRNA abundance (39). The most interesting aspect of this pilot study is that three of the eight trapped genes identified (FIR1, BUD4, MRP8) have not yet been targeted in a large‐scale transposon‐based insertional mutagenesis screen in spite of the impressive number of more than 22 000 insertion events analysed so far (5–7). We infer this from the fact that these genes are not listed in the TRIPLES database [http://ygac.med.yale.edu/triples/triples.htm; (7)], which describes all insertional mutagenesis events generated by the transposon‐based approach. Another two of the genes trapped in our study (NVJ1, UBC13) have been targeted only once by transposon‐based trapping. In contrast, many other genes, including APG1, SDS24 and YBR052, have by now been hit multiple times. We conclude that by radiation‐assisted NHI a more complete coverage of the genome can be obtained and that this method is suitable to complement any existing large‐scale mutagenesis screen. A major concern may, however, arise from using a clastogenic agent that has a potential of introducing additional genetic and genomic alterations. At present, we cannot estimate how often the analysis of insertion‐associated phenotypes will suffer from artefacts caused by radiation‐induced additional alterations. It should be noted, however, that good experimental practice requires that in any insertional mutagenesis screen, including transposon‐based screens, phenotypes be carefully checked for segregation with given insertion alleles. We propose that radiation‐assisted NHI is a suitable means to complement existing large‐scale studies in functional genomics in S.cerevisiae and any other organisms that exhibits a low‐efficiency NHEJ system. It would, of course, be desirable to use multi‐purpose fragments in transformation that allow, after insertion, to simultaneously investigate a variety of endpoints. ACKNOWLEDGEMENTS We gratefully acknowledge expert technical assistance by K. Winkler, financial support by Deutsche Forschungsge meinschaft (KI 796/1) and European Commission (FIS5‐1999‐00066), and travel grants by the Neuherberger Forschungsförderung. View large Download slide View large Download slide Figure 1. Frequency of transformation with linearised plasmid pM151 when cells were treated with γ‐irradiation immediately after the heat‐shock step of the transformation procedure. The number of transformants per microgram plasmid DNA was normalised with respect to parallel control transformations with circular plasmid YEplac195. (A) Comparison of wild‐type strain RSY12 (white) and its rad5 (light grey) and yku70 (dark grey) mutant derivatives. Indicated are means and standard deviations from three independent experiments (except RSY12 rad5, 200 Gy: one experiment). (B) Comparison of RSY12 [white; same data as in (A)] and its rad52 (grey) mutant derivative. Note that the scale of the y‐axis has changed. Data for RSY12 rad52 are from two experiments (0 and 50 Gy) or one experiment (100 and 200 Gy). View large Download slide View large Download slide Figure 1. Frequency of transformation with linearised plasmid pM151 when cells were treated with γ‐irradiation immediately after the heat‐shock step of the transformation procedure. The number of transformants per microgram plasmid DNA was normalised with respect to parallel control transformations with circular plasmid YEplac195. (A) Comparison of wild‐type strain RSY12 (white) and its rad5 (light grey) and yku70 (dark grey) mutant derivatives. Indicated are means and standard deviations from three independent experiments (except RSY12 rad5, 200 Gy: one experiment). (B) Comparison of RSY12 [white; same data as in (A)] and its rad52 (grey) mutant derivative. Note that the scale of the y‐axis has changed. Data for RSY12 rad52 are from two experiments (0 and 50 Gy) or one experiment (100 and 200 Gy). View largeDownload slide Figure 2. Junctions between plasmid MKM20 and genomic DNA, as determined by plasmid rescue and sequencing with primers P1 and P2. Remnants of the single‐stranded overhangs present in the transformed plasmid after restriction digest are indicated in bold. For clarification, the first nucleotides of genomic sequence at the junction sites are labelled by shading. Regions of microhomology at the junctions are underlined; nucleotides that are inserted after plasmid integration are framed. Numbers in the chromosomal sequences refer to chromosomal positions as indicated in the SGD database. Note that in clone MK66, 16 nt of plasmid sequence were absent at the P1 junction. View largeDownload slide Figure 2. Junctions between plasmid MKM20 and genomic DNA, as determined by plasmid rescue and sequencing with primers P1 and P2. Remnants of the single‐stranded overhangs present in the transformed plasmid after restriction digest are indicated in bold. For clarification, the first nucleotides of genomic sequence at the junction sites are labelled by shading. Regions of microhomology at the junctions are underlined; nucleotides that are inserted after plasmid integration are framed. Numbers in the chromosomal sequences refer to chromosomal positions as indicated in the SGD database. Note that in clone MK66, 16 nt of plasmid sequence were absent at the P1 junction. View largeDownload slide Figure 3.LacZ expression in samples treated with 60 J/m2 UV radiation (filled squares) and in mock‐treated samples (open squares) as a function of post‐irradiation incubation under non‐growth conditions. LacZ activity is indicated in Miller units. View largeDownload slide Figure 3.LacZ expression in samples treated with 60 J/m2 UV radiation (filled squares) and in mock‐treated samples (open squares) as a function of post‐irradiation incubation under non‐growth conditions. LacZ activity is indicated in Miller units. Table 1. Insertion of the lacZ reporter relative to the trapped gene in clones displaying differential expression in a semi‐quantitative screen after UV‐irradiation Clone ID Gene Insertion of lacZ N‐terminus MK12 APG1 In‐frame within ORF MK33 FIR1 In‐frame within ORF MK35 MRP8 In‐frame within ORF MK38 SDS24 In‐frame within ORF MK42 UBC13 In‐frame in intron MK60 NVJ1 In‐frame within ORF MK66 BUD4 22 bp upstream MK82 YBR052c In‐frame within ORF Clone ID Gene Insertion of lacZ N‐terminus MK12 APG1 In‐frame within ORF MK33 FIR1 In‐frame within ORF MK35 MRP8 In‐frame within ORF MK38 SDS24 In‐frame within ORF MK42 UBC13 In‐frame in intron MK60 NVJ1 In‐frame within ORF MK66 BUD4 22 bp upstream MK82 YBR052c In‐frame within ORF View Large References 1. Parinov,S. and Sundaresan,V. ( 2000) Functional genomics in Arabidospis: large‐scale insertional mutagenesis complements the genome sequencing project. Curr. Opin. Biotech. , 11, 157–161. Google Scholar 2. Stanford,W.L., Cohn,J.B. and Cordes,S.P. ( 2001) Gene‐trap mutagenesis: past, present and beyond. Nature Rev. Genet. , 2, 756–768. Google Scholar 3. Vidan,S. and Snyder,M. ( 2001) Large‐scale mutagenesis: yeast genetics in the genome era. Curr. Opin. Biotech. , 12, 28–34. Google Scholar 4. Golling,G., Amsterdam,A., Sun,Z., Antonelli,M., Maldonado,E., Chen,W., Burgess,S., Haldi,M., Artzt,K., Farrington,S. et al. ( 2002) Insertional mutagenesis in zebrafish rapidly identifies genes essential for early vertebrate development. Nature Genet. , 31, 135–140. Google Scholar 5. Ross‐Macdonald,P., Sheehan,A., Roeder,G.S. and Snyder,M. ( 1997) A multipurpose transposon system for analyzing protein production, localization, and function in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA , 94, 190–195. Google Scholar 6. Ross‐Macdonald,P., Coelho,P.S.R., Roemer,T., Agarwal,S., Kumar,A., Jansen,R., Cheung,K.‐H., Sheehan,A., Symoniatis,D., Umansky,L. et al. ( 1999) Large‐scale analysis of the yeast genome by transposon tagging and gene disruption. Nature , 402, 413–418. Google Scholar 7. Kumar,A., Cheung,K.‐H., Tosches,N., Masiar,P., Liu,Y., Miller,P. and Snyder,M. ( 2002) The TRIPLES database: a community resource for yeast molecular biology. Nucleic Acids Res. , 30, 73–75. Google Scholar 8. Davies,C.J. and Hutchinson,C.A. ( 1995) Insertion site specificity of the transposon Tn3. Nucleic Acids Res. , 23, 507–514. Google Scholar 9. Schiestl,R.H., Dominska,M. and Petes,T.D. ( 1993) Transformation of Saccharomyces cerevisiae with nonhomologous DNA: illegitimate integration of transforming DNA into yeast chromosomes and in vivo ligation of transforming DNA to mitochondrial DNA sequences. Mol. Cell. Biol. , 13, 2697–2705. Google Scholar 10. Lewis,L.K. and Resnick,M.A. ( 2000) Tying up loose ends: non‐homologous end‐joining in Saccharomyces cerevisiae. Mutat. Res. , 451, 71–89. Google Scholar 11. Stanford,W.L., Caruana,G., Vallis,K.A., Inamdar,M., Hidaka,M., Bautch,V.L. and Bernstein,A. ( 1998) Expression trapping: identification of novel genes expressed in hematopoetic and endothelial lineages by gene trapping in ES cells. Blood , 92, 4622–4631. Google Scholar 12. Budziszewski,G.J., Lewis,S.P., Glover,L.W., Reineke,J., Jones,G., Ziemnik,L.S., Lonowski,J., Nyfeler,B., Aux,G., Zhou,Q. et al. ( 2001) Arabidopsis genes essential for seedling viability: isolation of insertional mutants and molecular cloning. Genetics , 159, 1765–1778. Google Scholar 13. Chua,G., Taricani,L., Stangle,W. and Young,P.G. ( 2000) Insertional mutagenesis based on illegitimate recombination in Schizosaccharomyces pombe. Nucleic Acids Res. , 28, e53. Google Scholar 14. Schiestl,R.H. and Petes,T.D. ( 1991) Integration of DNA fragments by illegitimate recombination in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA , 88, 7585–7589. Google Scholar 15. Schiestl,R.H., Zhu,J. and Petes,T.D. ( 1994) Effect of mutations in genes affecting homologous recombination on restriction enzyme‐mediated and illegitimate recombination in Saccharomyces cerevisiae. Mol. Cell. Biol. , 14, 4493–4500. Google Scholar 16. Manivasakam,P. and Schiestl,R.H. ( 1998) Nonhomologous end joining during restriction enzyme‐mediated DNA integration in Saccharomyces cerevisiae. Mol. Cell. Biol. , 18, 1736–1745. Google Scholar 17. Riggle,P.J. and Kumamoto,C.A. ( 1998) Genetic analysis in fungi using restriction‐enzyme‐mediated integration. Curr. Opin. Microbiol. , 1, 395–399. Google Scholar 18. Kraxenberger,A., Friedl,A.A. and Kellerer,A.M. ( 1994) Computer simulation of pulsed field gel runs allows the quantitation of radiation‐induced double‐strand breaks in yeast. Electrophoresis , 15, 128–136. Google Scholar 19. Gari,E., Piedrafita,L., Aldea,M. and Herrero,E. ( 1997) A set of vectors with a tetracycline‐regulated promoter system to modulate gene expression in Saccharomyces cerevisiae. Yeast , 13, 837–848. Google Scholar 20. Wach,A., Brachat,A., Alberti‐Segui,C., Rebischung,C. and Philippsen,P. ( 1997) Heterologous HIS3 marker and GFP reporter modules for PCR‐targeting in Saccharomyces cerevisiae. Yeast , 13, 1065–1075. Google Scholar 21. Kiechle,M., Friedl,A.A., Manivasakam,P., Eckardt‐Schupp,F. and Schiestl,R.H. ( 2000) DNA integration by Ty integrase in yku70 mutant Saccharomyces cerevisiae cells. Mol. Cell. Biol. , 20, 8836–8844. Google Scholar 22. Gietz,D., St. Jean,A., Woods,R.A. and Schiestl,R.H. ( 1992) Improved method for high efficiency transformation of intact yeast cells. Nucleic Acids Res. , 20, 1425. Google Scholar 23. Gietz,R.D. and Sugino,A. ( 1988) New yeast‐Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six‐base pair restriction sites. Gene , 74, 527–534. Google Scholar 24. Ahne,F., Jha,B. and Eckardt‐Schupp,F. ( 1997) The RAD5 gene product is involved in the avoidance of non‐homologous end‐joining of DNA double strand breaks in the yeast Saccharomyces cerevisiae. Nucleic Acids Res. , 25, 743–499. Google Scholar 25. van Dyck,E., Stasiak,A.Z., Stasiak,A. and West,S.C. ( 1999) Binding of double‐strand breaks in DNA by human Rad52 protein. Nature , 398, 728–731. Google Scholar 26. Mount,R.C., Jordan,B.E. and Hadfield,C. ( 1996) Reporter gene systems for assaying gene expression in yeast. Methods Mol. Biol. , 53, 239–248. Google Scholar 27. Averbeck,D. and Averbeck,S. ( 1994) Induction of the genes RAD54 and RNR2 by various DNA damaging agents in Saccharomyces cerevisiae. Mutat. Res. , 315, 123–138. Google Scholar 28. Davis,C.A., Grate,L., Spingola,M. and Ares,M. ( 2000) Test of intron predictions reveals novel splice sites, alternatively spliced mRNAs and new introns in meiotically regulated genes of yeast. Nucleic Acids Res. , 28, 1700–1706. Google Scholar 29. Gasch,A.P., Spellman,P.T., Kao,C.M., Carmel‐Harel,O., Eisen,M.B., Storz,G., Botstein,D. and Brown,P.O. ( 2000) Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell , 11, 4241–4257. Google Scholar 30. Gasch,A.P., Huang,M., Metzner,S., Botstein,D., Elledge,S.J. and Brown,P.O. ( 2001) Genomic expression response to DNA‐damaging agents and the regulatory role of the yeast ATR homolog Mec1p. Mol. Biol. Cell , 12, 2987–3003. Google Scholar 31. Jelinsky,S.A., Estep,P., Church,G.M. and Samson,L.D. ( 2000) Regulatory network revealed by transcriptional profiling of damaged Saccharomyces cerevisiae cells: Rpn4 links base excision repair with proteasomes. Mol. Cell. Biol. , 20, 8157–8167. Google Scholar 32. Brusky,J., Zhu,Y. and Xiao,W. ( 2000) UBC13, a DNA‐damage‐inducible gene, is a member of the error‐free postreplication repair pathway in Saccharomyces cerevisiae. Curr. Genet. , 37, 168–174. Google Scholar 33. Ulrich,H.D. ( 2001) The srs2 suppressor of UV sensitivity acts specifically on the RAD5‐ and MMS2‐dependent branch of the RAD6 pathway. Nucleic Acids Res. , 29, 3487–3494. Google Scholar 34. Birrell,G.W., Brown,J.A., Wu,H.I., Giaever,G., Chu,A.M., Davis,R.W. and Brown,J.M. ( 2002) Transcriptional response of Saccharomyces cerevisiae to DNA‐damaging agents does not identify the genes that protect against these agents. Proc. Natl Acad. Sci. USA , 99, 8778–8783. Google Scholar 35. Manivasakam,P., Aubrecht,J., Sidhom,S. and Schiestl,R.H. ( 2001) Restriction enzymes increase efficiencies of illegitimate DNA integration but decrease homologous integration in mammalian cells. Nucleic Acids Res. , 29, 4826–4833. Google Scholar 36. Karathanasis,E. and Wilson,T.E. ( 2002) Enhancement of Saccharomyces cerevisiae end‐joining efficiency by cell growth stage but not by impairment of recombination. Genetics , 161, 1015–1027. Google Scholar 37. Harrison,P.M., Kumar,A., Lang,N., Snyder,M. and Gerstein,M. ( 2002) A question of size: the eukaryotic proteome and the problems in defining it. Nucleic Acids Res. , 30, 1083–1090. Google Scholar 38. Giaever,G., Chu,A.M., Ni,L., Connelly,C., Riles,L., Véronneau,S., Dow,S., Lucau‐Danila,A., Anderson,A., André,B. et al. ( 2002) Functional profiling of the Saccharomyces cerevisiae genome. Nature , 418, 387–391. Google Scholar 39. Griffin,T.J., Gygi,S.P., Ideker,T., Rist.,B., Eng,J., Hood,L., Aebersold,R. ( 2002) Complementary profiling of gene expression at the transcriptome and proteome levels in Saccharomyces cerevisiae. Mol. Cell . Proteomics , 1, 323–333. Google Scholar
T7 RNA polymerase as a self‐replicating label for antigen quantificationTannous, Bakhos A.; Laios, Eleftheria; Christopoulos, Theodore K.
doi: N/Apmid: N/A
Enzymes are used widely as labels in binding assays for protein analytes, because they provide signal amplification. Efforts at improving the assay sensitivity have been focused mainly on the synthesis of novel substrates, e.g. fluorogenic and chemiluminogenic ones. We report the investigation of T7 RNA polymerase (T7RP) as a label with unique characteristics for antigen quantification. In an in vitro, coupled (one‐step) transcription/translation reaction, T7RP catalyzes the expression of an enzyme‐coding DNA template to produce free enzyme (luciferase) in solution. We demonstrate that the generated luciferase is linearly related to the input T7RP in a range covering over four orders of magnitude. It is also shown that T7RP exhibits a significant level of self‐replication (100‐fold) in vitro by acting on a DNA template comprising the T7RP cDNA downstream of a T7 promoter. By combining the self‐replication reaction with the expression of luciferase DNA, as low as 1400 T7RP molecules are detectable. Furthermore, the T7RP is biotinylated, complexed with streptavidin and used for antigen quantification in a microtiter well‐based assay with high sensitivity and reproducibility.