日 植 病 報 64: 244-248 (1998)

Ann. Phytopathol. Soc. Jpn. 64: 244-248 (1998)

An Efficient Cloning Strategy for Viral , Double-stranded RNAs with Unknown Sequences*

Masamichi ISOGAI**,Ichiro UYEDA**and Tatsuji HATAYA**

Abstract

A strategy was designed to efficiently clone double-stranded RNAs (dsRNAs) of unknown sequences

and very low availability. From rice plants infected with rice black streaked dwarf fijivirus (RBSDV), ten

dsRNA genomic segments were extracted directly. The 3•Œ ends of the plus and minus strands of the

dsRNAs were polyadenylated and then used as templates for an initial reverse transcription using an

oligo-dT-containing adapter primer (AP). The first-strand cDNAs of both polarities were annealed, filled

in and amplified by the polymerase chain reaction using one primer containing an adapter region

sequence identical to that in the AP. The amplified cDNA products corresponded in size to the full

lengths of RBSDV S5, S6, S7, S8, S9 and S10; full length cDNA clones to RBSDV S8, S9 and S10

containing both terminal nucleotide sequences were obtained. Moreover, the nucleotide sequences of six

full length cDNA clones were the same as those previously reported. These data indicate that this method

may be applicable to full length cDNA cloning of dsRNAs.

(Received January 6, 1998; Accepted March 2, 1998)

Key words: cDNA cloning, rice black streaked dwarf virus, double-stranded RNA.

In this paper, rice black streaked dwarf fijivirus

INTRODUCTION (RBSDV), a ten-segmented dsRNA virus, was used to develope an efficient method for obtaining full length Cashdollar et al.1) first reported a method for cloning cDNAs to dsRNAs with unknown sequences by combin- double-stranded RNA (dsRNA) genomes of reoviruses. ing PCR with the methods of Cashdollar et al.1). The strategy made full use of the double-stranded nature of the genome. The full length ds cDNA mole- MATERIALS AND METHODS cules were synthesized by the addition of oligo (C) to the RNA template, reverse-transcribing them with oligo Source of virus and direct extraction of genomic

(dG) primer, hybridizing the cognate plus and minus dsRNAs from infected rice plants The isolate of cDNA strands, and filling the ends with DNA polymer- RBSDV used in this study had been maintained in rice ase I. This method was later modified by first plants for more than twenty years3). polyadenylating the 3•Œ ends of the RNA instead of The dsRNA genomes of RBSDV were extracted adding oligo (C) and is widely used for cloning viral directly by the method in Murao et al.7). In brief, total genomes in the Reoviridae. Because the method requires nucleic acids were extracted from about 0.5g of white 10-50ƒÊg of template RNAs for the first-strand cDNA tumorous stalks of infected rice plants with phenol- synthesis, it is not applicable when only a small amount chloroform and then ethanol precipitated. The ssRNA of the genomic RNAs is available. After commercial was precipitated with 2M LiCl and bound with CC41 kits for gene cloning of single-stranded RNA (ssRNA) cellulose powder (Whatman, UK) to further purify the became available, they have been widely used to synthe- dsRNA. About 2ƒÊg of RBSDV dsRNA genome was size first-strand cDNAs to dsRNAs after denaturation. obtained. In addition, due to the accumulation of nucleotide Preparation of cDNA The 3•Œ termini of both sequence data, polymerase chain reaction (PCR) tech- strands of the dsRNA were first polyadenylated. To 2ƒÊg niques can now be used to make ds cDNAs from a very of total RBSDV genome in 30ƒÊl of H2O, 20ƒÊl of dimeth- small quantity of the template. However, when the yl sulf oxide was added, and the mixture was incubated nucleotide sequence is unknown and only a very small at 95•Ž for 5min to melt the dsRNAs. The denatured quantity of the dsRNA genome is available, full length dsRNA was precipitated with 3vol of ethanol and 1/10 cDNAs are still difficult to clone. vol of 3M sodium acetate, pH 5.2. The precipitate was

* This work was supported in part by Research Fellowships of the Japan Society for the Promotion of Science for Young

Scientists. ** Faculty of Agriculture , Hokkaido University, Kita 9, Nishi 9, Kita-ku, SapPoro 060-8589, Japan 北 海 道 大 学 農 学 部 Ann. Phytopathol. Soc. Jpn. 64 (4). August, 1998 245

suspended in 51ƒÊl of H2O; and 40ƒÊl of polyadenylation Sequencing The cloned cDNA was sequenced by reaction mixture (125mM Tris-HCI, pH 7.9, 500mM the dideoxynucleotide chain terminator method9) using a NaCl, 25mM MgCl2, 5mM MnCl2, 2.5mM dithio- Thermo Sequenase fluorescent labeled primer cycle threitol), 2ƒÊl of 100mM ATP and 2ƒÊl of poly (A) sequencing kit with 7-deaza-dGTP for Li-Cor sequencer polymerase (1.2U/ƒÊl) were added. After incubation at (Amersham). Deduced nucleotide sequences were assem- 37•Ž for 30min, polyadenylated RNA was extracted bled and analyzed by the computer program DNASIS with phenol/chloroform and precipitated with 2.5vol of (Hitachi Software Engineering Co., Ltd.). ethanol and 1/2vol of 7.5M ammonium acetate. It was then suspended in 12ƒÊl of H2O; and 1ƒÊl of 10ƒÊM RESULTS oligo-dT-containing adapter primer (AP), 5•ŒCGATGGT- ACCTGCAGGCGCGCC(T)17 3•Œ, was added. After incu- PCR amplification bation at 95•Ž for 2min, the mixture was quickly chilled The strategy used to clone RBSDV genes was on ice. To this mixture was added 2ƒÊl of 10•~reverse designed to synthesize full length ds cDNA copies of transcriptase buffer (200mM Tris-HCl, pH 8.4, 500mM unknown dsRNA genes. A flow diagram of the proce- KCl), 2ƒÊl of 25mM MgCl2, 1ƒÊl of 10mM dNTP mixture dure is shown in Fig. 1. The 3•Œtermini of both strands of and 2ƒÊl of 0.1M dithiothreitol. After preincubation at the RBSDV dsRNA were polyadenylated using poly (A)

42•Ž for 2min, 1ƒÊl of SuperScriptTM reverse tran- polymerase after denaturation in dimethyl sulf oxide. scriptase (200U/ƒÊl; GIBCO BRL) was added, and the The polyadenylated products were used as a template mixture was then incubated for 50min at 42•Ž. The for an initial reverse transcription using AP, the oligo- reaction was terminated by heating at 70•Ž for 15min, dT-containing adapter primer. The AP initiated cDNA and the mixture was placed on ice for 3min. One ƒÊl of synthesis at the polyadenylated regions. Products were RNase H (2U/ƒÊl) was then added, and the mixture was amplified with PCR using AUAP containing an adapter incubated for 20min at 37•Ž. The cDNA was passed region sequence identical to that in the AP. through a Sephadex G-50 column to remove unincorpo- In the early phase of PCR, the first-strand cDNAs rated nucleotides and short cDNAs. Eluted cDNAs were from both strands (+ and -) were annealed to each precipitated by ethanol and suspended in 100ƒÊl of H2O. other and extended to the 3•Œ ends. As a result, the A 20-ƒÊl portion of the suspension was used for sequence of the adapter region at the 5•Œ end and the amplification with 2ƒÊl of 10ƒÊM primer containing complementary sequences of the adapter region at the 3 adapter region sequences identical to that in the AP end were added to both strands of the first-strand cDNA.

(AUAP), 5•ŒCGATGGTACCTGCAGGCGCGCC 3•Œ, for 30 In the late phase of PCR, ds cDNA to full length dsRNA cycles at 95•Ž (1min), 60•Ž (2min), and 72•Ž (3min) in 50 was amplified only by AUAP. The products of the PCR l Takara EX Taq polymerase buffer with 5U of were analyzed by electrophoresis in a Tris-borate- Takara EX Taq (Takara). buffered agarose gel (Fig. 2). Five size classes of cDNA

Fig. 1. Strategy for full length cDNA cloning of the dsRNA. AP and AUAP are the oligo-dT-containing adapter primer and the primer containing adapter region sequences, respectively. 246 日本植物病理学会報 第64巻 第4号 平成10年8月

1 2 A

Fig. 2. PCR products analyzed on a 1% agarose gel stained with ethidium bromide. Lane 1, RBSDV genome. Lane 2, amplified cDNAs of RBSDV genome.

molecules were clearly detected. The sizes of these cDNAs were the same as those of RBSDV S5-S10. According to Kawano et al.4), the molecular weight of RBSDV S5 was estimated to be about 2.0•~106 daltons

(about 3kbp), showing that dsRNAs of 3kbp can be converted to full length cDNAs. Full length cloning The PCR products were electrophoresed on 5% poly- acrylamide gel in Tris-borate buffer, and then two bands, one corresponding to RBSDV S8 and the other corresponding to a mixture of S9 and S10, were excised Fig. 3. Typical sequence analysis of the terminal and eluted from the gel. Eluted DNAs with both termini domains of a full length cDNA clone (RBH14) containing an adapter region with Pst I were inserted corresponding to RBSDV S10. (A) Analysis of into the Pst I site of pBluescript II SK- (Stratagene, the 5•Œ terminus of the S10 coding strand. (B) USA), and then Escherichia coli strain MV1184 was Analysis of the 5•Œterminus of S10 noncoding transformed. The inserts were confirmed initially by strand. The homopolymer tails added during sequencing across the plasmid/insert junction. Inserts of cDNA synthesis can be observed below the full length cDNA to RBSDV S8-S10 were followed by indicated sequences. sequences of the adapter region and added poly (A) (Fig. 3). Comparison of nucleotide sequences of six cDNA T at the 11th position from the 5•Œend, which was exactly clones to RBSDV S10 the same position as those of RBH13 and RBH16 (Fig.

Six clones of full length cDNAs to RBSDV S10 4A). Therefore, the insertion of T at the 11th position

(RBH11, RBH13, RBH14, RBH16, RBH17 and RBH19) from the 5•Œ end was not a characteristic of MRDV S10; were sequenced. These clones contained the complete some molecules with T insertion at the 11th position terminal regions of RBSDV S10 reported by Uyeda et from the 5•Œend should exist in the population of RBSDV al.8). Clones RBH11, RBH14, RBH17 and RBH19 had the S10. same 1801 nucleotides (nt) as RBSDV S10 reported by Heterogeneity in RBSDV S10 was also found; more Uyeda et al.8). However, RBH13 and RBH16 were 1802 than two of the six clones had the same substitution at nt long. Considering error rate per nucleotide polymer- two positions (Fig. 4B). A nucleotide substitution is also ized by a reverse transcriptase and a Taq DNA unlikely to have occurred at the same position in two of polymerase2,10), an insertion is unlikely to have occurred the six clones. One was at the 549th position from the 5 at the same position in two of the six clones. Besides, end and the other was at the 557th position. At position this change was in the 5•Œuntranslated region and did not 549 in the clones of RBH14 and RBH17, aspartic acid cause a frameshift of the open reading frame (Fig. 4A). was changed glycine. And at position 557 in the clones of Furthermore, maize rough dwarf fijivirus (MRDV) S10 RBH14, RBH17 and RBH19, proline was changed serine. has 88% nucleotide sequence homology to RBSDV S10 Furthermore, substitutions at 11 positions had substitu- and is 1802nt long as analyzed by Marzachi et al.5). tions in only one of the 6 clones. However, only three of When sequences of MRDV S10 were compared with the 11 positions resulted in amino acid changes (data not those of the cDNA clones, MRDV S10 had an additional shown). Ann. Phytopathol. Soc. Jpn. 64 (4). August, 1998 247

A clones to full length RBSDV 510 which were cloned by the above method. There were insertions and substitu- tions at the same position in more than two of them. Based on the error rate per nucleotide polymerized by a reverse transcriptase and a Taq DNA polymerase2,10), they should actually exist in the S10 population from only one RBSDV isolate. Furthermore, there were 11 positions where only one clone had a different nucleotide from the other 5 clones. Although nucleotide changes at these positions could possibly result from errors in B cloning, only three of them resulted in amino acid changes. Therefore, most of them are considered to be nucleotide sequence variations within the population of one RBSDV isolate, and errors during cloning were low. Replication of ssRNA genome viruses except for retroviruses proceeds from a reaction of RNA- dependent RNA polymerase encoded by a virus genome. In the process of synthesizing a complementary strand, dsRNAs of a virus replicative form are produced6). For Fig. 4. Comparison of the nucleotide sequences of full this reason, most ssRNA genome viruses have dsRNAs length cDNA to S10 of RBH11, RBH13, RBH14, in infected cells. The cloning method in this paper might RBH16, RBH17 and RBH19. Only nucleotides also be applicable for cloning other RNA viruses. different from those of RBH11 are shown. (A)

Comparison of the 5•Œ terminal sequences of the

coding strand. Start colons of the open reading Literature cited

frame are underlined. (B) The same substitution 1. Cashdollar, L.W., Esparza, J., Hudson, G.R., Chmelo, R., at two positions, 548nt and 556nt from the 5•Œ Lee, P.W.K. and Joklik, W.K. (1982). Cloning of the terminal of the plus strand, in two of six cDNA double-stranded RNA genes of reovirus: Sequence of clones. (C) Position of nucleotide change in only the cloned S2 gene. Proc. Natl. Acad. Sci. USA 79: 7644- one clone. This change caused an amino acid 7648. change of a remarkable disposition. The 2. Domingo, E. and Holland, J. (1988). High error rates, predicted amino acid sequences are shown population equilibrium and evolution of RNA replica- under the nucleotide sequence, and amino acid tion system. In RNA Genetics (Holl, J.E., Domingo, E. changes caused by the nucleotide change are and Ahiquist, P. eds.), CRC Press, Boca Raton, FL., Vol. indicated by open letters (B and C). 3, pp. 3-35. 3. Isogai, M., Azuhata, F., Uyeda, I., Shikata, E. and Kimura, I. (1995). Genomic relationships between rice DISCUSSION black streaked dwarf and maize rough dwarf fijiviruses detected by nucleic acid hybridization. Ann. Phyto- In this work, total RBSDV genome was amplified by pathol. Soc. Jpn. 61: 513-518. a new cloning method, and amplified cDNA bands with 4. Kawano, S., Uyeda, I. and Shikata, E. (1984). Particle the same migration pattern as RBSDV S5-S10 were structure and double-stranded RNA of rice ragged stunt recognized by agarose gel electrophoresis. Respective virus. J. Fac. Agric. Hokkaido Univ. 61: 408-418. amplified bands corresponding to RBSDV S8, S9 and S10 5. Marzachf, C., Boccardo, G., Milne, R., Isogai, M. and were excised and eluted from the gel, and eluted bands Uyeda, I. (1995). Genome structure and variability of with both termini containing an adapter region with Fijiviruses. Semin. Virol. 6: 103-108. restriction sites were inserted into the plasmid vectors. 6. Matthews, R.E.F. (1992). Fundamentals of Plant Vi- However, cDNA bands corresponding to RBSDV S1-S4 rology, Academic Press, California, pp. 91-182. were not detected. Low-molecular-weight genome seg- 7. Murao, K., Suda, N., Uyeda, I., Isogai, M., Suga, H., Yamada, N., Kimura, I. and Shikata, E. (1994). ments may have been preferentially amplified when Genomic heterogeneity of rice dwarf Phytoreovirus field total viral genome segments were used as templates. isolates and nucleotide sequences of variants of genome Another possibility is that the PCR conditions of the segment 12. J. Gen. Virol. 75: 1843-1848. cloning were not optimized for a long and accurate PCR. 8. Uyeda, I., Azuhata, F. and Shikata, E. (1990). Nucleo- For cloning high-molecular-weight genome segments tide sequence of rice black-streaked dwarf virus genome such as S1-S4, only the desired single genome segment segment 10. Proc. Jpn. Acad. 66: 37-40. should be extracted, and then the cloning procedure 9. Sanger, F., Nicklen, S. and Coulson, A.R. (1977). DNA should be started. Furthermore, optimizing conditions sequencing with chain-terminating inhibitors. Proc. for a long and accurate PCR might be necessary. Natl. Acad. Sci. USA 74: 5463-5467. Sequence data were obtained from the six cDNA 10. Tindall, K.R. and Kunkel, T.A. (1988). Fidelity of DNA 248 日本植物病理学会報 第64巻 第4号 平成10年8月

synthesis of the thermus aquaticus DNA polymerase. ゴdT配 列 を持 つ ア ダ プ タ ー プ ラ イ マ ー(AP)を 用 い た 逆 転 写 Biochemistry 27: 6008-6013. 反 応 の 鋳 型 と し た 。 続 い て,作 製 し た フ ァ ー ス トス ト ラ ン ド cDNAを 鋳 型 と し て,AP中 の ア ダ プ タ ー 領 域 と同 じ 配 列 を 持 和 文 摘 要 っ プ ラ イ マ ー を用 い てPCRを 行 っ た 。その 結 果, RBSDV S5か らS10の 全 長 に対 応 す る増 幅 産 物 が ア ガ ロ ー ス ゲ ル 電 気 泳 動 で 磯 貝 雅 道 ・上田 一 郎 ・畑谷 達 児:未 知 の ウ イ ル ス ニ 本 鎖RNA 確 認 さ れ た 。プ ラ ス ミ ドベ ク タ ー に ク ロー ニ ン グ し た 結 果,S8, ゲ ノ ム の効 率 的 な ク ロ ー ニ ン グ法 S9, S10の 全 長cDNAク ロ ー ン を作 製 す る こ とが で き た 。 さ ら に,S10の 全 長cDNAを 含 む6ク ロ ー ン の 塩 基 配 列 解 析 か ら, 未 知 の塩 基 配 列 を持 つ 二 本 鎖RNAを 効 率 的 に ク ロ ー ニ ン グ 本 ク ロ ー ニ ン グ方 法 は 変 異 を起 こす 可 能 性 が 少 な い こ と が 明 ら す る 方 法 を 開 発 し た 。 材 料 と し て,イ ネ 黒 条 萎 縮 ウ イ ル ス か と な っ た 。 こ れ ら の 結 果 か ら,こ の方 法 は未 知 の 二 本 鎖RNA (RBSDV)に 感 染 し た イ ネ よ り直 接 抽 出 し た10分 節 二 本 鎖 の効 率 的 な全 長 ク ロ ー ニ ン グ に 応 用 で き る と考 え ら れ た 。 RNAゲ ノ ム を 用 い た 。 二 本 鎖RNAゲ ノ ム の+鎖 お よ び-鎖 の3′ 末 端 に ポ リAポ リ メ ラ ー ゼ で ポ リA配 列 を 付 加 し,オ リ