- Email: [email protected]
Random mutagenesis in a plant viral genome using a DNA repair-deficient mutator Escherichia coli strain
Journal of Virological Methods 94 (2001) 37 – 43 www.elsevier.com/locate/jviromet
Random mutagenesis in a plant viral genome using a DNA repair-deficient mutator Escherichia coli strain Xiaoyun Lu, H. Hirata, Y. Yamaji, M. Ugaki, S. Namba * Laboratory of Bioresource Technology, Graduate School of Frontier Sciences, The Uni6ersity of Tokyo, 1 -1 -1 Yayoi, Bunkyo-ku, Tokyo 113 -8657, Japan Received 29 June 2000; received in revised form 17 January 2001; accepted 18 January 2001
Abstract Random mutagenesis in a plant viral genome is valuable for generating attenuated strains or for analyzing viral gene function at the molecular level. A DNA repair-deficient mutator Escherichia coli strain was used for random mutagenesis of a plant viral genome. A full-length infectious cDNA clone of Citrus tatter leaf 6irus (genus Capillo6irus) L strain (CTLV-L) genomic RNA under the T7 promoter sequence (pITCL) was introduced into the mutator E. coli strain XL1-Red and mutagenized overnight. To fix mutations, the mixture of plasmid DNA isolated from colonies of the mutator bacteria was introduced into another E. coli strain, JM109, which has normal DNA repair function. Infectious viral genomic RNA was transcribed in vitro from each mutagenized pITCL clone and inoculated on host plants. Phenotypic mutants were selected for altered pathogenicity in the inoculated plants. Nucleotide sequence analysis of each mutant revealed that mutations were introduced randomly into the CTLV-L genome regardless of the function of the viral gene. The nucleotide substitutions were biased towards single point mutations, which consisted of more transitions than transversions or single-base frameshifts. These mutations were preserved stably in plants subject to sequential mechanical inoculation. The strategy presented below is a simple and very efficient way to generate virus mutants for analyzing the functions of viral genes. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Citrus tatter leaf 6irus; Capillo6irus; Random mutagenesis; Full-length infectious cDNA clone; A mutator E. coil strain XL1-Red; Pathogenicity; Transition; Transversion
1. Introduction Attenuated plant virus strains have attracted much attention since McKinney (1929) first described cross-protection, a phenomenon in which * Corresponding author. Tel.: + 81-424-693125; fax: +81424-698786. E-mail address: [email protected] (S. Namba).
infection with an attenuated strain suppresses the disease symptoms caused by subsequent infection by a virulent strain. Attenuated strains can result from natural mutations in the field or from those induced by temperature, chemical, or radiation treatment in the laboratory. Attenuated viruses can also be obtained by mutagenesis of cloned viral genomes. Mutagenesis of a viral genome is
0166-0934/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0166-0934(01)00270-1
38
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
valuable, not only for producing attenuated viruses, but also for analyzing the function of viral genes and the proteins that they encode. There are two common strategies for mutagenesis of cloned viral genomes: site-directed and random mutagenesis. Site-directed mutagenesis is a powerful method, in which specific nucleotides of viral genes are altered so that the nature and function of the nucleotides and the genes can be assesed critically. To exploit this technique, however, detailed structural information is required about the targeted gene or other related genes. Unfortunately, sufficient information is not always available. In addition, if we rely solely on the existing information about the target gene in order to determine which nucleotides to mutagenize, then we will fall to detect other important nucleotides in the gene. On the other hand, random mutagenesis is effective for identifying the location and boundaries of sequences with a particular function in genes within a cloned DNA fragment, and thus is valuable for analyzing viral gene function at the molecular level. Successful random mutagenesis is exemplified by the in vitro directed evolution of proteins/enzymes. In this process, a gene coding for the protein is randomly mutagenized, expressed in vitro or in vivo, and desired characteristics of the protein, such as its activity, specificity or stability, are selected for (Kuchner et al., 1997; Harayama 1998; Shibata et al., 1998; Jaeger and Reetz, 2000; Reetz and Jaeger, 2000). The random mutagenesis method can even be used when no nucleotide sequence information is available, and is readily used when simple genetic screening is available, because mutations can be introduced anywhere in cloned DNA. Random mutagenesis can be accomplished by several different methods, such as altering the sequences within restriction endonuclease sites, inserting an oligonucleotide linker randomly into a plasmid, damaging plasmid DNA in vitro with chemicals, or incorporating incorrect nucleotides during in vitro DNA synthesis. Various methods have been used to generate random mutagenesis libraries, including chemical mutagenesis (Stassen et al., 1992), PCR-driven mutagenesis (Shibata et al., 1998), and DNA shuffling (Harayama 1998).
However, these methods are reportedly time-consuming, laborious, and expensive, and they generate only a limited number of random mutations at a time. In this study, random mutations were incorporated into the cloned genome of Citrus tatter leaf (genus Capillo6irus) L strain (CTLV-L) using the E. coli mutator strain XL1-Red (Greener et al., 1996, 1997), which lacks the primary DNA repair pathways. This is the first time that an E. coli mutator strain has been used for mutagenesis of a plant virus genome. The efficiency of this method and the nature of the mutations induced are described and its usefulness in the mutagenesis of viral genome is discussed.
2. Materials and methods
2.1. The 6irus strain and the infectious full-length cDNA clone of its genomic RNA CTLV-L, which was used in this experiment, is a filamentous virus with a 6.5 kb long linear positive-sense ssRNA genome. A full-length infectious clone of CTLV-L, pITCL, was constructed by inserting the full-length cDNA of the genome of CTLV-L with a T7 promoter between the SalI and SmaI sites of pUC18 as described previously (Ohira et al., 1995).
2.2. Mutagenesis of the CTLV-L genome Random mutations were generated in pITCL as illustrated in Fig. 1. Approximately 0.3 mg of pITCL plasmid DNA was transformed into 100 ml of competent cells of E. Coli mutator strain XL1-Red (endA1, gyrA96, thi-1, hsdR17, supE44, relA1, lac, mutD5, mutS, mutT, Tn10 (Tetr)) in the presence of 1.7 ml of mercaptoethanol following the manufacturer’s protocol (Stratagene, La Jolla, CA). After the mixture was incubated on ice for 30 min, it was heat shocked at 42°C for 45 s. SOC medium (0.9 ml) was added to the mixture, which was then incubated at 37°C for 1 h with vigorous shaking. Two hundred microliters of culture were plated onto an LB/ampicillin agar plate, which was incubated at 37°C overnight.
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
About 50 colonies which appeared on the plate were mixed and a mixture of mutagenized plasmids
39
was directly isolated from the mixed colonies by the alkaline lysis method (Sambrook et al., 1989).
Fig. 1. Diagram indicating the steps performed for mutagenesis using the E. coli XL1-Red mutator strain and selecting mutants with altered pathogenicity on plants.
40
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
To fix mutations, the mixture of plasmids was introduced into competent cells of E. coli JM109 with a normal DNA repair function (e14 (McrA− ), recA1, endA1, gyrA96, thi-1, hsdR17 (rk− mk+ ), supE44, relAl, delta (lac− proAB), F%[traD36, proAB, lacIq, lacZdeltaM15 ]) (Wako, Japan) as described above. Selected colonies were cultured on a mid-scale (10 ml) and plasmids were isolated using a QIAGEN Piasmid Kit (QIAGEN, Germany) following the manufacturer’s protocol.
2.3. In 6itro synthesis of infectious transcripts Approximately 10 mg of each plasmid DNA was linearized by complete digestion with Not I, and then the reaction mixture was incubated for 30 min at 37°C with proteinase K to eliminate RNase activity. After extracted with phenol– chloroform, the plasmid DNA was ethanol-precipitated and resuspended in 10 ml RNase-free sterile distilled water. In vitro transcripts were synthesized from 1 mg of this DNA template with T7 RNA polymerase using an mCAP RNA Capping Kit (Stratagene, USA) in a 25 ml reaction according to the manufacturer’s instructions.
2.4. Inoculation of plants A host plant, Chenopodium quinoa, was grown in a greenhouse at 20– 25°C, and used for inoculation experiments when it reached the stage of 5–10 leaves. Two leaves per plant were inoculated. Approximately 3 ml of transcripts were rubbed onto each leaf, which had been sprinkled with carborundum before inoculation. The plants were observed for signs of disease, starting 4 days post inoculation.
2.6. RT-PCR and direct sequencing To analyze the viral RNA in infected plants, the total RNA was extracted from inoculated and upper leaves as described by Chomczynski and Sacchi (1987) with slight modification, and 1–2 kb overlapping regions covering the whole CTLV genomic RNA were reverse-transcribed and PCRamplified by using the synthesized primers and the TaKaRa RNA PCR Kit (Takara, Japan) following the manufacturer’s protocol. The PCR products were directly sequenced by the synthesized primers and the kit described above.
3. Results
3.1. Selection of plasmid mutants by their pathogenicity in inoculated plants Infectious RNA transcribed from the original pITCL induced typical symptoms of CTLV-L consisting of necrotic local lesions on inoculated leaves ca. 4 days post inoculation and systemic chlorotic mottling and upper leaf malformation on the next day. However, transcripts synthesized from the 50 plasmid clones produced four different types of symptoms. Six transcripts (RM4, RM12, RM17, RM22, RM24, and RM46) did not produce any signals of systemic viral infection, whereas RM11 and RM19 caused symptoms that were milder than those of pITCL. RM21 was almost symptomless and very mild and delayed symptoms were observed later. The other 41 transcripts produced the same symptoms as the original pITCL.
3.2. Complete genomic sequence of each mutant 2.5. Nucleotide sequence analysis The nucleotide sequence of each plasmid mutant was determined using either universal primers ( − 21 M13 forward and M13 reverse primers) or 26 synthesized internal primers designed from the complete sequence of pITCL (Ohira et al., 1995) using BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, USA) following the manufacturer’s protocol.
Of the 50 inoculated clones, above-mentioned nine clones produced altered symptoms on C. quinoa compared to that of the original pITCL. The complete genomic sequence of these nine clones as well as four clones showing the same symptoms as pITCL were determined. Nine clones with altered symptoms contained one to three nucleotide mutations in the genome (Fig. 2), while no mutations were found in four clones
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
41
Fig. 2. Schematic representation of each mutant. Boxes represent ORFs, lines represent untranslated regions (UTR). CP, capsid protein; MP, movement protein; a, methytransferase domain; b, papain-like proteinase domain; c, NTP-binding helicase domain; d, RNA polymerase domain. The position of each mutation is depicted as vertical or oblique lines.
showing the original symptoms. The number, type, and the nature of the mutations are given in Table 1. Clones RM17, RM19, RM21, and RM24 each contained a single point mutation. Clones RM19 and RM21 possessed silent mutations, while different point mutations that induced amino acid changes were found in clones RM11 and RM22. Clones RM4 and RM46 had one nucleotide deletion, while clone RM12 had one nucleotide insertion. No clones had more than two consecutive insertions or deletions. Among the 14 point mutations, transitions (93%) predominated over transversions (7%). The frequency of individual transition was 29 for T– C, 18 for C– T, 24 for G–A, and 5% for A– G. There was one T – G transversions and no G– T, C – G, or G – C transversions. The frequencies of insertions and deletions were low.
3.3. Stability of each mutation in infected plants RT-PCR and sequence analyses were used to examine the stability of each mutation in the viral genome in infected plants. Sequence analysis of three mutant clones (RM11, RM19 and RM21) and the original clone (pITCL) revealed that the whole genomic RNA sequence of each clone was completely identical to the original sequence, including the corresponding mutated nucleotides, of each infectious cDNA clone used. In addition, the symptoms induced by these three mutant clones and the original clone did not change after at least three passages using inocula from successively infected plants.
4. Discussion The E. coli mutator strain XL1-Red was engineered to be deficient in three independent DNA repair pathways, mutS, mutD, and mutT (Greener et al., 1996, 1997). It has been applied for random mutagenesis of cloned genes encoding 6%-N aminoglycoside acetyltransferase from a multiresistance transposon (Panaite and Tolmasky, 1998), an esterase from Pseudomonas fluorescens (PFE) (Bornscheuer et al., 1998; Henke and Bornscheuer, 1999), and a hepatitis B-specific single chain Fv (scFv) (Coia et al., 1997). In this study, this system was used to mutagenize randomly a plant viral genome. The plasmid pITCL, which contains an infectious full-length cDNA of a capillovirus CTLV-L genomic RNA, was mutagenized in the mutator strain and subsequently the mutations were fixed in a normal E. coli strain JM109. The infectious viral RNA transcribed in vitro from the cDNA clones were inoculated onto host plants. It was that several mutated clones had altered pathogenicity on the host plant by observing the symptoms of infection and by sequencing the plasmid DNA. The results indicate that E. coli mutator strain XL1-Red can be used to generate useful, new genetic mutants of plant viruses that have varying pathogenicity in plants. The mutation rate averaged 1.77 mutations per 6496 bp-long CTLV-L genome when the bacteria was cultured overnight. In this system, it is reported that the longer the culture period is, the more mutations are introduced (Greener et al., 1996, 1997). Thus it is
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
42
expected that we could control the approximate number of mutations to be introduced by simply changing the length of the culture period. Almost every possible type of mutation was observed (Table 1), including rare transversions and single-base frameshifts. Single nucleotide mutations were more common than multiple nucleotide mutations, and substitutions dominated deletions or insertions. Transitions occurred more frequently than transversions. Most of the transitions involved pyrimidines (thymine to cytosine and vice versa), although a few occurred between purines (adenine to guanine and vice versa). Notably, the rate of transversions using XL1-Red was low, and only one T– G transversion was observed. The mutations were introduced evenly throughout the viral genome as shown in Fig. 2. The pathogenicity of some mutants was different from that of pITCL. Interestingly, both RM19 and RM21 had one silent mutation caused by a single nucleotide substitution, yet RM19 produced symptoms similar to those of pITCL infection, while RM21 induced mild symptoms later. Both RM17 and RM24 had single amino acid changes resulting from single nucleotide sub-
stitutions in the capsid protein region of the genome of CTLV-L. No symptoms were induced, and the virus could not be detected in the host by RT-PCR. Both RM11 and RM22 had two changed amino acids, yet the former produced symptoms similar to those of pITCL infection, while the latter did not produce any symptoms and no virus was detected. As expected, RM4, RM12, and RM46 did not produce any symptoms because they contained frameshift mutations in their genes and could not produce functional viral proteins. It is expected that further study of functions of viral genes can be possible by using numerous mutant viruses readily produced by this method. The results clearly demonstrate that in vivo mutagenesis in the E. coli mutator strain is an efficient alternative to the methods used for mutagenesis of plant virus genes. This method is fast, easy to use, requires no special equipment, and the steps required to obtain new viral mutants are not hampered by the problems inherent in other mutagenesis methods. The assay employed has several advantages: it does not require other special substrates, easy visual identification is possi-
Table 1 Characterization of mutants Mutant
Symptoma
No. of mutation
Mutated basesb
Type of mutation Point mutation Transition
Transversion
Insertion
Deletion
RM4
ND
3
2
0
0
1
RM11 RM12
M ND
2 3
2 2
0 0
0 1
0 0
RM17 RM19 RM21 RM22 RM24 RM46
ND M + ND ND ND
1 1 1 2 1 3
1 1 1 2 1 1
0 0 0 0 0 1
0 0 0 0 0 0
0 0 0 0 0 1
a b
AAT6 AAC6 GA6 C GG6 C GGG6 GG6 GT6 C GC6 C GG6 C GA6 C TGG6 TGA6 G6 AA A6 AA T6 CA TTCA CC6 C CT6 C GGC6 GGT6 GGT6 GGC6 GT6 T GC6 T AG6 T AA6 T CC6 A CT6 A T6 TC C6 TC ACC6 AC6 TT6 G TG6 G
M, Milder symptoms than the original clone; +, Symptomless infection; ND, No symptom and no virus detected. Changed bases are underlined.
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
ble by investigating the pathogenicity of each mutant on plants, and a large number of mutants can be investigated within a short time. In practice, it takes about 3 weeks to generate new virus mutants with altered properties: 1 week for mutagenesis of the virus genome using XL1-Red strain and retransformation into JM109 strain, 1 week for selecting mutants by plant inoculation, and 1 week for determining the mutations in each variant based on sequencing. In addition, large mutant libraries can be created, enabling the detailed investigation of functional domains. One possible disadvantage of the method is that an infectious full-length cDNA clone of the viral genomic RNA must be available beforehand. However, the availability of many virus sequences and the long PCR technique make it easier to obtain infectious cDNA clones. This study used the T7 promoter to transcribe viral cDNA in vitro. However, the use of functional plant promoters that transcribe cDNA in vivo should make the method more convenient by eliminating the in vitro transcription step.
Acknowledgements We are grateful to Shigern Hatano for his excellent technical assistance.
References Bornscheuer, U.T., Altenbuchner, J., Meyer, H., 1998. Directed evolution of an esterase for the stereoselective resolution of a key intermediate in the synthesis of epothilones. Biotechnol. Bioeng. 58, 554 –559. Chomczynski, P., Sacchi, N., 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate – phenol– chloroform extraction. Anal. Biochem. 162, 156 – 159.
.
43
Coia, G., Ayres, A., Lilley, G.G., Hudson, P.J., Irving, R.A., 1997. Use of mutator cells as a means for increasing production levels of a recombinant antibody directed against hepatitis B. Gene 201, 203 – 209. Greener, A., Callahan, M., Jerpseth, B., 1997. An efficient random mutagenesis technique using an E. coli mutator strain. Mol. Biotechnol. 7, 189 – 195. Greener, A., Callahan, M., Jerpseth, B., 1996. An efficient random mutagenesis technique using an E. coil mutator strain. Methods Mol. Biol. 57, 375 – 385. Harayama, S., 1998. Artificial evolution by DNA shuffling. Trends Biotechnol. 16, 76 – 82. Henke, E., Bornscheuer, U.T., 1999. Directed evolution of an esterase from pseudomonas fluorescens. Random mutagenesis by error-prone PCR or mutator strain and identification of mutants showing enhanced enantioselectivity by a resorufin-based fluorescence assay. Biol. Chem. 380, 1029 – 1033. Jaeger, K.E., Reetz, M.T., 2000. Directed evolution of enantioselective enzymes for organic chemistry. Curr. Opin. Chem. Biol. 4, 68 – 73. Kuchner, O., Arnold, F.H., 1997. Directed evolution of enzyme catalysts. Trends Biotechnol. 15, 523 – 530. McKinney, H.H., 1929. Mosaic diseases in the Canary Islands, West Africa and Gibraltar. J. Agric. Res. 39, 557 – 578. Ohira, K., Namba, S., Rozanov, M., Kusumi, T., Tsuchizaki, T., 1995. Complete sequence of an infectious full-length cDNA clone of citrus tatter leaf capillovirus: comparative sequence analysis of capillovirus genomes. J. Gen. Virol. 76, 2305 – 2309. Panaite, D.M., Tolmasky, M.E., 1998. Characterization of mutants of the 6%-N-acetyltransferase encoded by the multiresistance transposon Tn1331: effect of phen171-to-Leu171 and Tyr80-to-Cys80 substitutions. Plasmid 39, 123 – 133. Reetz, M.T., Jaeger, K.E., 2000. Enantioselective enzymes for organic synthesis created by directed evolution. Chemistry 6, 407 – 412. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning. Cold Spring Harbor Laboratory Press, pp. 1.25 – 1.27. Shibata, H., Kato, H., Oda, J., 1998. Random mutagenesis on the pseudomonas lipase activator protein, LipB: exploring amino acid residues required for its function. Protein Eng. 11, 467 – 472. Stassen, A.P., Zaman, G.J., van Deursen, J.M., Schoenmakers, J.G., Konings, R.N., 1992. Selection and characterization of randomly produced mutants of gene V protein of bacteriophage M13. Eur. J. Biochem. 204, 1003 – 1004.
Random mutagenesis in a plant viral genome using a DNA repair-deficient mutator Escherichia coli strain Xiaoyun Lu, H. Hirata, Y. Yamaji, M. Ugaki, S. Namba * Laboratory of Bioresource Technology, Graduate School of Frontier Sciences, The Uni6ersity of Tokyo, 1 -1 -1 Yayoi, Bunkyo-ku, Tokyo 113 -8657, Japan Received 29 June 2000; received in revised form 17 January 2001; accepted 18 January 2001
Abstract Random mutagenesis in a plant viral genome is valuable for generating attenuated strains or for analyzing viral gene function at the molecular level. A DNA repair-deficient mutator Escherichia coli strain was used for random mutagenesis of a plant viral genome. A full-length infectious cDNA clone of Citrus tatter leaf 6irus (genus Capillo6irus) L strain (CTLV-L) genomic RNA under the T7 promoter sequence (pITCL) was introduced into the mutator E. coli strain XL1-Red and mutagenized overnight. To fix mutations, the mixture of plasmid DNA isolated from colonies of the mutator bacteria was introduced into another E. coli strain, JM109, which has normal DNA repair function. Infectious viral genomic RNA was transcribed in vitro from each mutagenized pITCL clone and inoculated on host plants. Phenotypic mutants were selected for altered pathogenicity in the inoculated plants. Nucleotide sequence analysis of each mutant revealed that mutations were introduced randomly into the CTLV-L genome regardless of the function of the viral gene. The nucleotide substitutions were biased towards single point mutations, which consisted of more transitions than transversions or single-base frameshifts. These mutations were preserved stably in plants subject to sequential mechanical inoculation. The strategy presented below is a simple and very efficient way to generate virus mutants for analyzing the functions of viral genes. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Citrus tatter leaf 6irus; Capillo6irus; Random mutagenesis; Full-length infectious cDNA clone; A mutator E. coil strain XL1-Red; Pathogenicity; Transition; Transversion
1. Introduction Attenuated plant virus strains have attracted much attention since McKinney (1929) first described cross-protection, a phenomenon in which * Corresponding author. Tel.: + 81-424-693125; fax: +81424-698786. E-mail address: [email protected] (S. Namba).
infection with an attenuated strain suppresses the disease symptoms caused by subsequent infection by a virulent strain. Attenuated strains can result from natural mutations in the field or from those induced by temperature, chemical, or radiation treatment in the laboratory. Attenuated viruses can also be obtained by mutagenesis of cloned viral genomes. Mutagenesis of a viral genome is
0166-0934/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0166-0934(01)00270-1
38
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
valuable, not only for producing attenuated viruses, but also for analyzing the function of viral genes and the proteins that they encode. There are two common strategies for mutagenesis of cloned viral genomes: site-directed and random mutagenesis. Site-directed mutagenesis is a powerful method, in which specific nucleotides of viral genes are altered so that the nature and function of the nucleotides and the genes can be assesed critically. To exploit this technique, however, detailed structural information is required about the targeted gene or other related genes. Unfortunately, sufficient information is not always available. In addition, if we rely solely on the existing information about the target gene in order to determine which nucleotides to mutagenize, then we will fall to detect other important nucleotides in the gene. On the other hand, random mutagenesis is effective for identifying the location and boundaries of sequences with a particular function in genes within a cloned DNA fragment, and thus is valuable for analyzing viral gene function at the molecular level. Successful random mutagenesis is exemplified by the in vitro directed evolution of proteins/enzymes. In this process, a gene coding for the protein is randomly mutagenized, expressed in vitro or in vivo, and desired characteristics of the protein, such as its activity, specificity or stability, are selected for (Kuchner et al., 1997; Harayama 1998; Shibata et al., 1998; Jaeger and Reetz, 2000; Reetz and Jaeger, 2000). The random mutagenesis method can even be used when no nucleotide sequence information is available, and is readily used when simple genetic screening is available, because mutations can be introduced anywhere in cloned DNA. Random mutagenesis can be accomplished by several different methods, such as altering the sequences within restriction endonuclease sites, inserting an oligonucleotide linker randomly into a plasmid, damaging plasmid DNA in vitro with chemicals, or incorporating incorrect nucleotides during in vitro DNA synthesis. Various methods have been used to generate random mutagenesis libraries, including chemical mutagenesis (Stassen et al., 1992), PCR-driven mutagenesis (Shibata et al., 1998), and DNA shuffling (Harayama 1998).
However, these methods are reportedly time-consuming, laborious, and expensive, and they generate only a limited number of random mutations at a time. In this study, random mutations were incorporated into the cloned genome of Citrus tatter leaf (genus Capillo6irus) L strain (CTLV-L) using the E. coli mutator strain XL1-Red (Greener et al., 1996, 1997), which lacks the primary DNA repair pathways. This is the first time that an E. coli mutator strain has been used for mutagenesis of a plant virus genome. The efficiency of this method and the nature of the mutations induced are described and its usefulness in the mutagenesis of viral genome is discussed.
2. Materials and methods
2.1. The 6irus strain and the infectious full-length cDNA clone of its genomic RNA CTLV-L, which was used in this experiment, is a filamentous virus with a 6.5 kb long linear positive-sense ssRNA genome. A full-length infectious clone of CTLV-L, pITCL, was constructed by inserting the full-length cDNA of the genome of CTLV-L with a T7 promoter between the SalI and SmaI sites of pUC18 as described previously (Ohira et al., 1995).
2.2. Mutagenesis of the CTLV-L genome Random mutations were generated in pITCL as illustrated in Fig. 1. Approximately 0.3 mg of pITCL plasmid DNA was transformed into 100 ml of competent cells of E. Coli mutator strain XL1-Red (endA1, gyrA96, thi-1, hsdR17, supE44, relA1, lac, mutD5, mutS, mutT, Tn10 (Tetr)) in the presence of 1.7 ml of mercaptoethanol following the manufacturer’s protocol (Stratagene, La Jolla, CA). After the mixture was incubated on ice for 30 min, it was heat shocked at 42°C for 45 s. SOC medium (0.9 ml) was added to the mixture, which was then incubated at 37°C for 1 h with vigorous shaking. Two hundred microliters of culture were plated onto an LB/ampicillin agar plate, which was incubated at 37°C overnight.
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
About 50 colonies which appeared on the plate were mixed and a mixture of mutagenized plasmids
39
was directly isolated from the mixed colonies by the alkaline lysis method (Sambrook et al., 1989).
Fig. 1. Diagram indicating the steps performed for mutagenesis using the E. coli XL1-Red mutator strain and selecting mutants with altered pathogenicity on plants.
40
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
To fix mutations, the mixture of plasmids was introduced into competent cells of E. coli JM109 with a normal DNA repair function (e14 (McrA− ), recA1, endA1, gyrA96, thi-1, hsdR17 (rk− mk+ ), supE44, relAl, delta (lac− proAB), F%[traD36, proAB, lacIq, lacZdeltaM15 ]) (Wako, Japan) as described above. Selected colonies were cultured on a mid-scale (10 ml) and plasmids were isolated using a QIAGEN Piasmid Kit (QIAGEN, Germany) following the manufacturer’s protocol.
2.3. In 6itro synthesis of infectious transcripts Approximately 10 mg of each plasmid DNA was linearized by complete digestion with Not I, and then the reaction mixture was incubated for 30 min at 37°C with proteinase K to eliminate RNase activity. After extracted with phenol– chloroform, the plasmid DNA was ethanol-precipitated and resuspended in 10 ml RNase-free sterile distilled water. In vitro transcripts were synthesized from 1 mg of this DNA template with T7 RNA polymerase using an mCAP RNA Capping Kit (Stratagene, USA) in a 25 ml reaction according to the manufacturer’s instructions.
2.4. Inoculation of plants A host plant, Chenopodium quinoa, was grown in a greenhouse at 20– 25°C, and used for inoculation experiments when it reached the stage of 5–10 leaves. Two leaves per plant were inoculated. Approximately 3 ml of transcripts were rubbed onto each leaf, which had been sprinkled with carborundum before inoculation. The plants were observed for signs of disease, starting 4 days post inoculation.
2.6. RT-PCR and direct sequencing To analyze the viral RNA in infected plants, the total RNA was extracted from inoculated and upper leaves as described by Chomczynski and Sacchi (1987) with slight modification, and 1–2 kb overlapping regions covering the whole CTLV genomic RNA were reverse-transcribed and PCRamplified by using the synthesized primers and the TaKaRa RNA PCR Kit (Takara, Japan) following the manufacturer’s protocol. The PCR products were directly sequenced by the synthesized primers and the kit described above.
3. Results
3.1. Selection of plasmid mutants by their pathogenicity in inoculated plants Infectious RNA transcribed from the original pITCL induced typical symptoms of CTLV-L consisting of necrotic local lesions on inoculated leaves ca. 4 days post inoculation and systemic chlorotic mottling and upper leaf malformation on the next day. However, transcripts synthesized from the 50 plasmid clones produced four different types of symptoms. Six transcripts (RM4, RM12, RM17, RM22, RM24, and RM46) did not produce any signals of systemic viral infection, whereas RM11 and RM19 caused symptoms that were milder than those of pITCL. RM21 was almost symptomless and very mild and delayed symptoms were observed later. The other 41 transcripts produced the same symptoms as the original pITCL.
3.2. Complete genomic sequence of each mutant 2.5. Nucleotide sequence analysis The nucleotide sequence of each plasmid mutant was determined using either universal primers ( − 21 M13 forward and M13 reverse primers) or 26 synthesized internal primers designed from the complete sequence of pITCL (Ohira et al., 1995) using BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, USA) following the manufacturer’s protocol.
Of the 50 inoculated clones, above-mentioned nine clones produced altered symptoms on C. quinoa compared to that of the original pITCL. The complete genomic sequence of these nine clones as well as four clones showing the same symptoms as pITCL were determined. Nine clones with altered symptoms contained one to three nucleotide mutations in the genome (Fig. 2), while no mutations were found in four clones
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
41
Fig. 2. Schematic representation of each mutant. Boxes represent ORFs, lines represent untranslated regions (UTR). CP, capsid protein; MP, movement protein; a, methytransferase domain; b, papain-like proteinase domain; c, NTP-binding helicase domain; d, RNA polymerase domain. The position of each mutation is depicted as vertical or oblique lines.
showing the original symptoms. The number, type, and the nature of the mutations are given in Table 1. Clones RM17, RM19, RM21, and RM24 each contained a single point mutation. Clones RM19 and RM21 possessed silent mutations, while different point mutations that induced amino acid changes were found in clones RM11 and RM22. Clones RM4 and RM46 had one nucleotide deletion, while clone RM12 had one nucleotide insertion. No clones had more than two consecutive insertions or deletions. Among the 14 point mutations, transitions (93%) predominated over transversions (7%). The frequency of individual transition was 29 for T– C, 18 for C– T, 24 for G–A, and 5% for A– G. There was one T – G transversions and no G– T, C – G, or G – C transversions. The frequencies of insertions and deletions were low.
3.3. Stability of each mutation in infected plants RT-PCR and sequence analyses were used to examine the stability of each mutation in the viral genome in infected plants. Sequence analysis of three mutant clones (RM11, RM19 and RM21) and the original clone (pITCL) revealed that the whole genomic RNA sequence of each clone was completely identical to the original sequence, including the corresponding mutated nucleotides, of each infectious cDNA clone used. In addition, the symptoms induced by these three mutant clones and the original clone did not change after at least three passages using inocula from successively infected plants.
4. Discussion The E. coli mutator strain XL1-Red was engineered to be deficient in three independent DNA repair pathways, mutS, mutD, and mutT (Greener et al., 1996, 1997). It has been applied for random mutagenesis of cloned genes encoding 6%-N aminoglycoside acetyltransferase from a multiresistance transposon (Panaite and Tolmasky, 1998), an esterase from Pseudomonas fluorescens (PFE) (Bornscheuer et al., 1998; Henke and Bornscheuer, 1999), and a hepatitis B-specific single chain Fv (scFv) (Coia et al., 1997). In this study, this system was used to mutagenize randomly a plant viral genome. The plasmid pITCL, which contains an infectious full-length cDNA of a capillovirus CTLV-L genomic RNA, was mutagenized in the mutator strain and subsequently the mutations were fixed in a normal E. coli strain JM109. The infectious viral RNA transcribed in vitro from the cDNA clones were inoculated onto host plants. It was that several mutated clones had altered pathogenicity on the host plant by observing the symptoms of infection and by sequencing the plasmid DNA. The results indicate that E. coli mutator strain XL1-Red can be used to generate useful, new genetic mutants of plant viruses that have varying pathogenicity in plants. The mutation rate averaged 1.77 mutations per 6496 bp-long CTLV-L genome when the bacteria was cultured overnight. In this system, it is reported that the longer the culture period is, the more mutations are introduced (Greener et al., 1996, 1997). Thus it is
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
42
expected that we could control the approximate number of mutations to be introduced by simply changing the length of the culture period. Almost every possible type of mutation was observed (Table 1), including rare transversions and single-base frameshifts. Single nucleotide mutations were more common than multiple nucleotide mutations, and substitutions dominated deletions or insertions. Transitions occurred more frequently than transversions. Most of the transitions involved pyrimidines (thymine to cytosine and vice versa), although a few occurred between purines (adenine to guanine and vice versa). Notably, the rate of transversions using XL1-Red was low, and only one T– G transversion was observed. The mutations were introduced evenly throughout the viral genome as shown in Fig. 2. The pathogenicity of some mutants was different from that of pITCL. Interestingly, both RM19 and RM21 had one silent mutation caused by a single nucleotide substitution, yet RM19 produced symptoms similar to those of pITCL infection, while RM21 induced mild symptoms later. Both RM17 and RM24 had single amino acid changes resulting from single nucleotide sub-
stitutions in the capsid protein region of the genome of CTLV-L. No symptoms were induced, and the virus could not be detected in the host by RT-PCR. Both RM11 and RM22 had two changed amino acids, yet the former produced symptoms similar to those of pITCL infection, while the latter did not produce any symptoms and no virus was detected. As expected, RM4, RM12, and RM46 did not produce any symptoms because they contained frameshift mutations in their genes and could not produce functional viral proteins. It is expected that further study of functions of viral genes can be possible by using numerous mutant viruses readily produced by this method. The results clearly demonstrate that in vivo mutagenesis in the E. coli mutator strain is an efficient alternative to the methods used for mutagenesis of plant virus genes. This method is fast, easy to use, requires no special equipment, and the steps required to obtain new viral mutants are not hampered by the problems inherent in other mutagenesis methods. The assay employed has several advantages: it does not require other special substrates, easy visual identification is possi-
Table 1 Characterization of mutants Mutant
Symptoma
No. of mutation
Mutated basesb
Type of mutation Point mutation Transition
Transversion
Insertion
Deletion
RM4
ND
3
2
0
0
1
RM11 RM12
M ND
2 3
2 2
0 0
0 1
0 0
RM17 RM19 RM21 RM22 RM24 RM46
ND M + ND ND ND
1 1 1 2 1 3
1 1 1 2 1 1
0 0 0 0 0 1
0 0 0 0 0 0
0 0 0 0 0 1
a b
AAT6 AAC6 GA6 C GG6 C GGG6 GG6 GT6 C GC6 C GG6 C GA6 C TGG6 TGA6 G6 AA A6 AA T6 CA TTCA CC6 C CT6 C GGC6 GGT6 GGT6 GGC6 GT6 T GC6 T AG6 T AA6 T CC6 A CT6 A T6 TC C6 TC ACC6 AC6 TT6 G TG6 G
M, Milder symptoms than the original clone; +, Symptomless infection; ND, No symptom and no virus detected. Changed bases are underlined.
X. Lu et al. / Journal of Virological Methods 94 (2001) 37–43
ble by investigating the pathogenicity of each mutant on plants, and a large number of mutants can be investigated within a short time. In practice, it takes about 3 weeks to generate new virus mutants with altered properties: 1 week for mutagenesis of the virus genome using XL1-Red strain and retransformation into JM109 strain, 1 week for selecting mutants by plant inoculation, and 1 week for determining the mutations in each variant based on sequencing. In addition, large mutant libraries can be created, enabling the detailed investigation of functional domains. One possible disadvantage of the method is that an infectious full-length cDNA clone of the viral genomic RNA must be available beforehand. However, the availability of many virus sequences and the long PCR technique make it easier to obtain infectious cDNA clones. This study used the T7 promoter to transcribe viral cDNA in vitro. However, the use of functional plant promoters that transcribe cDNA in vivo should make the method more convenient by eliminating the in vitro transcription step.
Acknowledgements We are grateful to Shigern Hatano for his excellent technical assistance.
References Bornscheuer, U.T., Altenbuchner, J., Meyer, H., 1998. Directed evolution of an esterase for the stereoselective resolution of a key intermediate in the synthesis of epothilones. Biotechnol. Bioeng. 58, 554 –559. Chomczynski, P., Sacchi, N., 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate – phenol– chloroform extraction. Anal. Biochem. 162, 156 – 159.
.
43
Coia, G., Ayres, A., Lilley, G.G., Hudson, P.J., Irving, R.A., 1997. Use of mutator cells as a means for increasing production levels of a recombinant antibody directed against hepatitis B. Gene 201, 203 – 209. Greener, A., Callahan, M., Jerpseth, B., 1997. An efficient random mutagenesis technique using an E. coli mutator strain. Mol. Biotechnol. 7, 189 – 195. Greener, A., Callahan, M., Jerpseth, B., 1996. An efficient random mutagenesis technique using an E. coil mutator strain. Methods Mol. Biol. 57, 375 – 385. Harayama, S., 1998. Artificial evolution by DNA shuffling. Trends Biotechnol. 16, 76 – 82. Henke, E., Bornscheuer, U.T., 1999. Directed evolution of an esterase from pseudomonas fluorescens. Random mutagenesis by error-prone PCR or mutator strain and identification of mutants showing enhanced enantioselectivity by a resorufin-based fluorescence assay. Biol. Chem. 380, 1029 – 1033. Jaeger, K.E., Reetz, M.T., 2000. Directed evolution of enantioselective enzymes for organic chemistry. Curr. Opin. Chem. Biol. 4, 68 – 73. Kuchner, O., Arnold, F.H., 1997. Directed evolution of enzyme catalysts. Trends Biotechnol. 15, 523 – 530. McKinney, H.H., 1929. Mosaic diseases in the Canary Islands, West Africa and Gibraltar. J. Agric. Res. 39, 557 – 578. Ohira, K., Namba, S., Rozanov, M., Kusumi, T., Tsuchizaki, T., 1995. Complete sequence of an infectious full-length cDNA clone of citrus tatter leaf capillovirus: comparative sequence analysis of capillovirus genomes. J. Gen. Virol. 76, 2305 – 2309. Panaite, D.M., Tolmasky, M.E., 1998. Characterization of mutants of the 6%-N-acetyltransferase encoded by the multiresistance transposon Tn1331: effect of phen171-to-Leu171 and Tyr80-to-Cys80 substitutions. Plasmid 39, 123 – 133. Reetz, M.T., Jaeger, K.E., 2000. Enantioselective enzymes for organic synthesis created by directed evolution. Chemistry 6, 407 – 412. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning. Cold Spring Harbor Laboratory Press, pp. 1.25 – 1.27. Shibata, H., Kato, H., Oda, J., 1998. Random mutagenesis on the pseudomonas lipase activator protein, LipB: exploring amino acid residues required for its function. Protein Eng. 11, 467 – 472. Stassen, A.P., Zaman, G.J., van Deursen, J.M., Schoenmakers, J.G., Konings, R.N., 1992. Selection and characterization of randomly produced mutants of gene V protein of bacteriophage M13. Eur. J. Biochem. 204, 1003 – 1004.