Quick viewing(Text Mode)

Identification and Characterization of Two Tandem Repeat Sequences (Trsb and Trsc) and a Retrotransposon (RIRE1) As Genome-General Sequences in Rice

Identification and Characterization of Two Tandem Repeat Sequences (Trsb and Trsc) and a Retrotransposon (RIRE1) As Genome-General Sequences in Rice

Genes Genet. Syst. (1996) 71, p. 373–382 Identification and characterization of two tandem repeat sequences (TrsB and TrsC) and a (RIRE1) as -general sequences in rice

Reiko Nakajima, Kenichi Noma, Hisako Ohtsubo and Eiichi Ohtsubo* Institute of Molecular and Cellular Biosciences, The University of Tokyo, Yayoi 1-1-1, Bunkyo-ku, Tokyo 113, Japan

(Received 17 August 1996, accepted 27 December 1996)

Three kinds of DNA sequences (here called TrsB, TrsC and RIRE1) have been previously reported to be those repeated in tandem specifically in the wild rice species with FF, CC or EE genome, respectively. To characterize these genome type-specific sequences, we carried out PCR using a pair of primers, which hybridize to a restricted region in the repeating unit sequence and prime DNA synthesis in both directions. Gel electrophoresis and DNA sequencing revealed that PCR using primers for TrsB (or TrsC) amplified the fragments with an integral series of a unit length not only from total DNA of the rice strain with FF (or CC) genome, but also from those of the rice strains with non-FF (or non-CC) genome. TrsB or TrsC was, however, found to be repeated in an extraordinary number of copies in the species with FF or CC genome, respectively, in which the TrsB (or TrsC) sequence has been originally identified. PCR using primers for RIRE1 produced various sizes of fragments from total DNA of the rice strains with EE genome. The fragments, however, showed no progression at interval of the unit length characteristic for tandem repeats. Nucleotide sequencing of the amplified fragments revealed that they were not the sequences repeated in tandem, but were those interspersed as an element having partial homology with the LTR sequences of , Wis- 2-1A in wheat and BARE-1 in barley. RIRE1 was present in the rice species with any types of , but in the species with EE genome in an extraordinary number of copies.

In rice species belonging to the genus Oryza, which have INTRODUCTION been classified into six diploid genome types (AA, BB, CC, Genomes of higher contain a large fraction of DD, EE and FF) and two tetraploid genome types (BBCC DNA sequences repeated in tandem or interspersed. and CCDD) (Chang, 1984), there exist genome-specific Besides the ribosomal RNA and DNA, tandem repeats as well as genome-general tandem repeats various kinds of tandem repeat sequences have been found (Zhao et al., 1989; Wu and Wu, 1992). These tandem re- in Secale cereale (Bedbrook et al., 1980; Appels et al., peat sequences have been identified by partially digesting 1986), Arabidopsis thaliana (Martinez-Zapater et al., total genomic DNA with a restriction enzyme and carrying 1986), Avena sativa (Fabijanski et al., 1990), Oryza sativa out Southern hybridization using each sequence as a (Wu and Wu, 1987; Ohtsubo et al., 1991; De Kochko et al., probe, which demonstrates a unique ladder band pattern 1991), and Actinidia deliciosa (Crowhurst and Gardner, characteristic for the tandem repeat sequences. In this 1991). Interspersed sequences identified are various study, we employed a method based on polymerase chain kinds of elements, such as transposable DNA elements reaction (PCR) used to identify and characterize a tandem (Nevers et al., 1986; Peterson, 1987; Coen et al., 1989; repeat sequence in rice, named TrsA, which is repeated Fedoroff, 1989; Gierl and Saedler, 1992), retrotransposons even in a few copies (Ohtsubo and Ohtsubo, 1994). Using (Flavell et al., 1992; Moore et al., 1991; Voytas et al., 1992; this method, we demonstrated that the three kinds of ge- Hirochika and Hirochika, 1993) and nome-specific tandem repeat sequences previously found (Shepherd at al., 1984; Mochizuki et al., 1992; Yoshioka in rice, pOb1, pOo2 and pOa4 (Zhao et al., 1989) (here et al., 1993; Deragon et al., 1994). called TrsB, TrsC and RIRE1, respectively), were actually genome-general sequences, and that one of them is not a * Corresponding author. tandem repeat sequence, but a retrotransposon inter- 374 R. NAKAJIMA et al.

spersed in rice genomes. [0.1% Ficoll (Pharmacia), 0.1% polyvinylpyrrolidon (Nakarai), 0.1% (w/v) bovine serum albumin (Seikagaku Kogyo)], 0.5% (w/v) sodium dodecyl sulfate (SDS), and 100 MATERIALS AND METHODS µg of sonicated salmon sperm DNA (Sigma)/ml. Then the Rice strains. Rice strains used are listed in Table 1. solution was exchanged for the hybridization solution Total genomic DNA of most of wild rice strains, which which contained the 32P-labeled oligonucleotide or cloned were isolated as described previously (Lichtenstein and DNA at 105 cpm/ml, and hybridization was carried out for Draper, 1985), was obtained from Y. Fukuta (Hokuriku 12 to 15 h at 65°C. The filter was washed sequentially in National Agricultural Experiment Station). Total 2 × SSC, 1 × SSC and 0.1 × SSC, each containing 0.1% SDS, genomic DNA of several wild rice strains was obtained at the temperature calculated from each of the probe se- from H. Hirochika (National Institute of Agrobiological quences in Table 2 as described by Meikoth and Wahl Resources), and Y. Sano (National Institute of ). (1984).

Cloning and DNA sequencing of the PCR-amplified Slot-blot hybridization. The copy number of a se- fragments. PCR was carried out by the standard quence was estimated by slot-blot hybridization described method (Saiki et al., 1988) using 2.5 units of AmpliTaq previously (Tenzen et al., 1994) using the DNA of DNA polymerase (Perkin Elmer), 0.5 µg total rice DNA, pRN43 carrying the TrsB sequence, pRN59 carrying the and 1 µM of each pair of primers (Table 2). Thirty cycles TrsC sequence, or pRN28 carrying a portion of the RIRE1 of amplification were carried out under the following con- sequence. For the copy number calculation, the sizes of dition; denaturation for 1 min at 93°C, annealing for 1 min the haploid genome of O. sativa and O. brachyantha were at 55°C, and DNA synthesis for 2 min at 72°C. The PCR- taken as 430 Mb, and the sizes of the haploid genome of O. amplified fragments were ligated to a linear pCRII vector officinalis and O. australiensis were taken as 550 and 960 in a TA cloning kit (Invitrogen) as recommended by the Mb, respectively (Martinez et al., 1994). supplier. DNA sequencing was carried out by the dideoxynucleotide chain termination method (Messing, Accession numbers. The nucleotide sequence data re- 1983; Sanger et al., 1977) using Sequenase Ver. 2.0 ported appear in the DDBJ, EMBL and GenBank nucle- (United States Biochemicals). The frequency of muta- otide sequence databases under the accession numbers, tions induced by PCR under the condition described above D85598~85604. is 1 per 642 bp (Tenzen et al., 1994).

RESULTS Southern hybridization. The PCR-amplified frag- ments were electrophoresed in a 1.8% agarose gel and Amplification of the repeated sequences in rice by stained with ethidium bromide. DNA was transferred to PCR. Three kinds of DNA sequences, pOb1 (159 bp) in a nylon membrane (Hybond-N+; Amersham) by alkaline O. brachyantha with FF genome, pOo2 (366 bp) in O. blotting under the condition recommended by the supplier. officinalis with CC genome and pOa4 (511 bp) in O. The filter was preincubated at 65°C for 1 h in 5 ml/100 cm2 australiensis with EE genome, have been reported as ge- of the hybridization solution containing 6 × SSC (0.9 M nome-specific tandem repeats (Zhao et al., 1989). To NaCl, 0.09 M trisodium citrate), 5 × Denhardt’s solution identify and characterize these sequences, we carried out

Table 1. Rice strains used Strain Species Genome type Source or reference

Nipponbare O. sativa AA Mochizuki et al. (1993) T65 O. sativa AA Mochizuki et al. (1993) W1581 O. barthii AA Y. Fukuta W1192 O. glumaepatula AA Y. Fukuta W1625 O. meridionalis AA Y. Fukuta W1577 O. punctata BB Y. Sano W1582 O. punctata BB Y. Fukuta W0002 O. officinalis CC Y. Fukuta W1521 O. eichingeri CC Y. Fukuta W1538 O. australiensis EE Y. Fukuta W0008 O. australiensis EE Y. Sano W1401 O. brachyantha FF Y. Fukuta WM5 O. alta CCDD H. Hirochika W0046 O. latifolia CCDD H. Hirochika Tandem repeat sequences and retrotransposon in rice 375

Table 2. Oligonucleotides used

Primer Sequenceb Origin Positionc or probea Ob1f 5'-gagctcTGACCTTGGATTCATTCCAT-3' pOb1 (=TrsB) 1- 20 Ob1b 5'-gagctcCCCTTCCAGCCCGAGAGTTT-3' pOb1 (=TrsB) 158-149 Oo2f 5'-ggatccGGATATGCGATGCGTTTTAG-3' pOo2 (=TrsC) 1- 20 Oo2b 5'-ggatccAAAGGAATGTGCCAACCATG-3' pOo2 (=TrsC) 366-357 Oa4f 5'-ggatccATCCAACAATGTGTTCTCTA-3' pOa4 (=RIRE1) 316-297 Oa4b 5'-ggatccAGACCCACCATGAGACACTA-3' pOa4 (=RIRE1) 317-337 F1 5'-TCCGTCATCAACTTCACC-3' TrsB 16- 33 F1' 5'-GAGCACGAAGTGTAAACC-3' TrsB 119-102 F2 5'-AGATTCGACTATTAACTTC-3' TrsB 53- 71 C1 5'-GGGCAATGTCCTTAAAGT-3' TrsC 20- 37 C1' 5'-AACAAGGTGAGGAATGTG-3' TrsC 353-336 C2 5'-GAACGACATTGGACGGGCTA-3' TrsC 143-162 E1 5'-ATCCAACAATGTGTTCTCT-3' RIRE1 316-298 E5' 5'-TGTGCCATTAGTTGGTCTCA-3' RIRE1 347-366 E6 5'-ATCCCTTTGCCTCACGAT-3' RIRE1 725-709 E6' 5'-TGTTAGTGATATGCCCT-3' RIRE1 1- 16 E4 5'-TCTCCATGTTGCCTCTAGGG-3' RIRE1 33- 14 E4' 5'-CGAGTGAATGACTAAGCT-3' RIRE1 650-667 a A pair of primers Ob1f/Ob1b were used to amplify the fragments containing the pOb1 (= TrsB) sequence. Primers Oo2f/Oo2b and Oa4f/Oa4b were used to amplify the pOo2 (= TrsC) and pOa4 (= RIRE1) sequences, respectively. Primers F1/F1' and C1/C1' were de- signed on the consensus sequences of TrsB and TrsC and used to amplify each Trs sequence from total DNA of the various rice strains. Primers E1/E6' and E5'/E6 were used to amplify the DNA fragments of the two regions, 1-316 and 349-727, within LTR of RIRE1 (see Figs. 4 and 5A), respectively. F2 and C2 were used as probes for TrsB and TrsC, respectively. E4 and E4' were used as probes for RIRE1. b Lowercase letters indicate linker sequences. c Numbers are coordinates given to the consensus sequences shown in Figs. 2 and 4.

PCR using a pair of primers, which hybridize to a re- neous, and did not show any unique ladder band pattern stricted region within each repeated DNA sequence and characteristic for the tandem repeat sequence (Fig. 1C). prime DNA synthesis in both directions from the region. This suggests that the pOa4 sequence is not repeated in If the repeating unit sequences are present in tandem in a tandem in O. australiensis. We will demonstrate later genome, PCR amplifies the fragments with the integral that it is a portion of the LTR sequence of a retro- series of a unit length, depending upon the repeating units transposon, here called RIRE1. annealed by the primers. Such fragments form a unique ladder of bands, representing monomers, dimers, trimers, Nucleotide sequence analysis of TrsB and TrsC. etc., in an agarose gel upon electrophoresis, as demon- We cloned three fragments amplified by PCR using prim- strated previously for a 360-bp tandem repeat sequence, ers for TrsB and sequenced them (see pRN43, 44 and 52 named TrsA (Ohtsubo and Ohtsubo, 1994). When total shown in Fig. 2A). A consensus sequence of 158 bp could DNA prepared from O. brachyantha or O. officinalis was be derived from nine monomer units identified (Fig. 2A). used as a template, the PCR fragments, which were ampli- Each unit sequence showed 88.5 - 98.8% similarity to the fied using a pair of primers for pOb1 or pOo2, formed consensus sequence. There seems a tendency of base bands showing a unique ladder pattern, respectively (Fig. substitutions occurring at the same position in every three 1A and B). This is consistent with the previous result unit, suggesting the presence of trimer units in a small or that pOb1 and pOo2 are present in tandem in each rice large region of repetition in O. brachyantha with FF species (Zhao et al., 1989). We call pOb1 and pOo2 se- genome. Similarly, cloning and sequencing of the three quences here as TrsB and TrsC (for Tandem Repeat Se- fragments amplified by PCR using primers for TrsC could quence B and C), respectively, since they are shown later derive a consensus sequence of 366 bp from the sequences to be not the genome-specific tandem repeats, but the ge- of the six units in O. officinalis with CC genome (Fig. 2B). nome-general tandem repeats. Each unit sequence showed 90.0 - 96.7% similarity to the When total DNA prepared from an O. australiensis consensus sequence. Both TrsB and TrsC sequences strain was used as a template, the fragments amplified by were highly homologous to those of pOb1 (90.0 - 92.0%) PCR using primers for pOa4 were found to be heteroge- and pOo2 (96.0%), respectively (Fig. 2). 376 R. NAKAJIMA et al.

Fig. 1. Agarose gels (1.8 %) showing fragments amplified by PCR. A Fragments amplified by PCR using a pair of primers for the pOb1 (= TrsB) sequence (Ob1f and Ob1b; Table 2) from total DNA of O. brachyantha W1401. Note that the fragments form a unique ladder of bands as indicated by bars. The length of the smallest fragment corresponding to a monomer unit of TrsB, 158 bp (see Fig. 2), is only shown. B Fragments amplified using a pair of primers for the pOo2 sequence (= TrsC) (Oo2f and Oo2b; Table 2) from total DNA of O. officinalis W0002. Note that the fragments form a unique ladder of bands as indicated by bars. The length of only the smallest frag- ment corresponding to a monomer unit of TrsC, 366 bp (see Fig. 2), is shown. C Fragments amplified using a pair of primers for the pOa4 (= RIRE1) sequence (Oa4f and Oa4b; Table 2) from total DNA of O. australiensis W1538. Note that the fragments are heteroge- neous in size and do not form any unique ladder of bands indicative of tandem repeats. Numbers indicate lengths of major fragments in bp.

TrsB and TrsC as genome type-general sequences. We then determined copy numbers of TrsB and TrsC in To examine whether TrsB or TrsC is present in the species various species by slot-blot hybridization. The copy num- with non-FF or non-CC genome, respectively, we carried ber of TrsB in O. brachyantha with FF genome was calcu- out PCR using total DNA prepared from various wild rice lated to be 9.2 × 104 per haploid genome, which is almost strains as templates. In this experiment, we used a pair consistent with the number reported by Zhao et al. (1989), of primers, which were derived from the consensus se- which was re-calculated to be 6.6 × 104 by taking 430 Mb quence for TrsB or TrsC (see Table 2), since homologous for the size of the haploid genome of O. brachyantha.On sequences to TrsB or TrsC, if any in the various rice the other hand, the copy number of TrsB in the species strains, may have diverged, but not much from the consen- with non-FF genome was less than 50 per haploid genome. sus sequence. The fragments amplified by PCR using The copy number of TrsC in O. officinalis with CC genome such primers for TrsB (or TrsC) were found to form the was 4.1 × 104 per haploid genome, which is almost consis- same ladder of bands as those from the rice strain with FF tent with the number reported by Zhao et al. (1989), which (or CC) genome (data not shown). We then carried out was re-calculated to be 7.8 × 104 by taking 550 Mb for the Southern blot analysis, and found that the probes for TrsB size of the haploid genome of O. officinalis. On the other (or TrsC) hybridized with the bands of the fragments ob- hand, the copy number of TrsC in the species with non-CC served above (Fig. 3A). These results demonstrate that genome was less than 100 per haploid genome. These re- TrsB (or TrsC) sequences is present not only in the species sults indicate that TrsB or TrsC is extraordinarily ampli- with FF (or CC) genome but also in the species with non- fied in the species with FF or CC genome, respectively. FF (or non-CC) genome. We cloned and sequenced the fragments amplified for pOa4, a portion of the LTR sequences of a retro- TrsB from the species with AA, BB, CC or EE genome. transposon, RIRE1. As described earlier, PCR using a The sequences determined were homologous to one an- particular pair of primers for the EE genome-specific se- other with sequence variation, 2 - 14 %, when they were quence, pOa4, produced the fragments which did not show compared with the consensus sequence derived from those the integral series of a 511-bp unit length (see Fig. 1C). with FF genome (Fig. 2A). This further confirms the re- Cloning and sequencing of the amplified fragments re- sult obtained above that TrsB is present not only in the vealed that each of the six clones analyzed contained a species with FF genome but also in the species with non- non-pOa4 sequence in addition to a partial or entire pOa4 FF genome. sequence (Figs. 4 and 5A). pRN28 contained the largest Tandem repeat sequences and retrotransposon in rice 377 378 R. NAKAJIMA et al.

Fig. 3. Presence of TrsB, TrsC, and RIRE1 in various rice strains. A Autoradiograms of the fragments of TrsB (a) and TrsC (b), showing a unique ladder band pattern (bars) (see Fig. 1A and B). The fragments were amplified by PCR using primers for TrsB (F1 and F1'; Table 2) or TrsC (C1 and C1'; Table 2) from total DNA of various rice strains and electrophoresed in 1.8% agarose gels. They were transferred to a nylon membrane and hybridized with the 32P-labeled probe (F2 or C2; Table 2). Lanes 1-9 show the DNA samples from total DNA of the rice strains, O. sativa cv. Nipponbare (AA), O. barthii W1581 (AA), O. glumaepatula W1192 (AA), O. meridionalis W1625 (AA), O. punctata W1582 (BB), O. officinalis W0002 (CC), O. eichingeri W1521 (CC), O. australiensis W1538 (EE), and O. brachyantha W1401 (FF) (see Table 1). B (a) An ethidium bromide-stained agarose gel (1.8%), showing the fragments of the 1-316 region of the RIRE1 element amplified by PCR using primers E1 and E6' (Table 2). (b) An autoradiogram showing the result of Southern hybridiza- tion of the sample shown in a with a probe for RIRE1 (E4; Table 2). Lanes 1-14 show the DNA samples from total DNA of the rice strains, O. sativa cv. Nipponbare (AA), O. sativa cv. T65 (AA), O. barthii W1581 (AA), O. glumaepatula W1192 (AA), O. meridionalis W1625 (AA), O. punctata W1582 (BB), O. punctata W1577 (BB), O. officinalis W0002 (CC), O. eichingeri W1521 (CC), O. australiensis W1538 (EE), O. australiensis W0008 (EE), and O. brachyantha W1401 (FF), O. alta WM5 (CCDD), and O. latifolia W0046 (CCDD) (see Table 1).

fragment whose sequence is homologous to that of the Database search revealed that the sequence in pRN28 fragment in any one of the other clones (Figs. 4 and 5A). was about 58 % homologous to both retrotransposons Wis- Interestingly, the right end regions of a 316 bp in length 2-1A in wheat (Harberd et al., 1987) and BARE-1 in barley are in common and consisted of a non-pOa4 sequence, 1 - (Manninen and Schulman, 1993). In the homologous se- 38 bp (by the coordinates given to pRN28 in Figs. 4 and 5A) quences, the 1 - 316 region in clone pRN28, which con- and the pOa4 sequence, 39 - 316 bp. The left end regions tained the common non-pOa4 sequence of 38 bp and the consisted of only a portion of the pOa4 sequence, 317 - 514 pOa4 sequence of 278 bp (Figs. 4 and 5), corresponded to bp, or the pOa4 sequence plus a non-pOa4 sequence, 515 - the region at the 5' end of the LTR sequences of Wis-2-1A 997 bp (see Figs. 4 and 5A). A 511-bp region in the con- and of BARE-1. The remaining 317 - 997 region in sensus sequence showed 94% similarity to the pOa4 se- pRN28 (see Figs. 4 and 5A) corresponded to the inside re- quence (Fig. 4). gion of the LTR sequences. We thus named the sequence,

Fig. 2. Nucleotide sequences of TrsB and TrsC. A TrsB. A consensus sequence (CON-FF) is shown by boldface letters. This se- quence was derived from the repeating unit sequences in three clones (pRN43, 44 and 52) from O. brachyantha W1401 with FF genome. Another consensus sequence (CON-ABCE) is also shown. This sequence was derived from the repeating unit sequences in four clones (AA1, BB1, CC1 and EE1) from O. barthii W1581 with AA genome, O. punctata W1577 with BB genome, O. officinalis W0002 with CC genome and O. australiensis W1538 with EE genome, respectively. The repeating unit sequences of pOb1 (Zhao et al. 1989) are shown for comparison. B TrsC. A consensus sequence (CON-CC) is shown by boldface letters. This sequence was derived from the repeat- ing unit sequences in three clones (pRN15, 20 and 29) from O. officinalis W0002 with CC genome. The repeating unit sequence of pOo2 (Zhao et al. 1989) is shown for comparison. Nucleotides identical to those in the consensus sequence are represented by short bars, and deleted nucleotides by slashes. Asterisks indicate positions of primers which are shown in parentheses (see Table 2). Numbers are coordinates given to each consensus sequence. Tandem repeat sequences and retrotransposon in rice 379

Fig. 4. Nucleotide sequences of the PCR-amplified fragments using primers for pOa4 (= RIRE1). A consensus sequence (CON), which is shown by boldface letters, was derived from the sequences in five clones named pRN. The pOa4 sequence (Zhao et al. 1989) is shown for comparison. Nucleotides identical to those in the consensus sequence are represented by short bars, and deleted nucleotides by slashes. Asterisks indicate positions of primers (Oa4f and Oa4; Table 2) used to amplify the fragments from total DNA of O. australiensis W1538. Numbers are coordinates given to the consensus sequence. See Fig. 5A for schematic representations of the sequences in the pRN clones. E5', E6' E1 and E6 indicate positions of primers used to amplify two fragments of RIRE1 (see the text). from which pRN28 (or pOa4) was derived, as RIRE1 (for using a pair of primers (E1 and E6'; Table 2), which am- RIce RetroElement 1). Fig. 6 shows nucleotide sequence plify a region inside of RIRE1, 1 - 316 bp (see Fig. 4). All homology in the 5' regions of LTR in RIRE1, BARE-1 and of the fragments amplified from total DNA of the rice Wis-2-1A. strains with non-EE genome were the same in size as To see whether the RIRE1 sequence is present in the those from total DNA of an O. australiensis strain with EE species with non-EE genome or not, we carried out PCR genome (Fig. 3Ba). A probe for RIRE1 hybridized to 380 R. NAKAJIMA et al.

Fig. 5. Structures of the fragments amplified by PCR using primers for pOa4 (= RIRE1), and a model for amplification of the fragments by PCR. A Schematic representations of structures of the PCR-amplified fragments from total DNA of O. australiensis W1538. Solid boxes indicate the regions corre- sponding to the pOa4 sequence; open boxes indicate the non-pOa4 sequences. Note that a sequence at position 249 in clone pRN35 is duplicated in tandem (see Fig. 4). B The possible structure of RIRE1 and a model for amplification of the RIRE1 fragments by PCR. RIRE1 may have the structure of retrotransposon with two LTRs (open boxes) which flank an internal region (thick line). There may exist an RIRE1 element with an insertion of another RIRE1 element at a site within an LTR, as shown. The resulting RIRE1 has a complex form of LTR with an inner LTR region a which is connected with an LTR end region b. PCR using a pair of primers, which hybridize to the pOa4 sequence (shadowed boxes) in the LTR sequence and prime DNA synthesis as shown, amplifies the fragment shown at the bottom. Small arrows indi- cate primers (Oa4f and Or4b; Table 2) used to amplify the frag- ments from total DNA of the O. australiensis strain used. Numbers are coordinates given to the consensus sequence of RIRE1 (see Fig. 4).

Fig. 6. Nucleotide sequence homology among BARE-1, Wis-2-1A and RIRE1. The sequences in the regions at the 5' ends are compared. Numbers at top indicate coordinates to RIRE1 (see Figs. 4 and 5A). Asterisks indicate homologous bases.

these fragments (Fig. 3Bb). This shows that a portion of a wide range of sizes which appeared as a smear (data not RIRE1 is present in the species with any genome types. shown), indicating that RIRE1 is present in an extraordi- We also carried out PCR using another pair of primers (E5' nary number of copies as an interspersed element in the O. and E6; Table 2), which amplify a different region inside of australiensis genome. Digestion with EcoRV generated RIRE1, 349 - 727 bp (see Fig. 4). All the fragments am- some fragments forming distinct bands on a smear. plified from total DNA of the rice strains with non-EE ge- There existed a band of the EcoRV fragment of about 0.5 nome were the same in size as those from total DNA of the kb in length (data not shown), which may correspond to O. australiensis strain with EE genome, and a probe for the pOa4 fragment (511 bp in length) cloned previously by RIRE1 hybridized to these fragments (data not shown). Zhao et al. (1989) from the O. australiensis DNA digested This shows that RIRE1 is present in the species with any with EcoRV. genome types. We then determined the number of copies of RIRE1 in We carried out Southern hybridization using total DNA O. sativa L. cv. Nipponbare and O. australiensis by slot- which were digested with the restriction enzyme, EcoRV, blot hybridization (see Materials and Methods) to be about BamHI, HindIII or EcoRI. The 32P-labeled RIRE1 probe 1.8 × 102 and 7.5 × 103 per haploid genome, respectively. hybridized to many O. australiensis DNA fragments with The latter number was significantly lower than that re- Tandem repeat sequences and retrotransposon in rice 381 ported by Zhao et al. (1989), which was re-calculated to be RIRE1 element at a different site inside of them during 6.4 × 104 by taking 960 Mb for the size of the haploid ge- retroposition of RIRE1 in the species with EE genome. nome of O. australiensis. This difference may be due to This could generate complex forms of LTRs with another the strain used in this study, which is different from that LTR at one or another site inside of the LTR sequence in used by Zhao et al. (1989). Note here, however, that both the first RIRE1 elements (Fig. 5B). Some complex forms O. australiensis strains did contain an extraordinary num- might have an LTR region with the 5'-end segment of LTR, ber of copies of RIRE1, particularly when we compared which was short enough for PCR amplification by the with the copy number of RIRE1, 1.8 × 102 per haploid ge- primers, which hybridized to both the 5'-end segment and nome of O. sativa L. cv. Nipponbare. a region inside of LTR, as shown in Fig. 5B. We have also shown in this paper that the RIRE1 ele- ments are present in all the rice species with various ge- DISCUSSION nome types. We carried out Southern hybridization us- We have shown in this paper that PCR using a pair of ing total DNA from wheat and barley, and found that primers, which hybridize to a restricted region and prime these plants also have a significant number of RIRE1 cop- DNA synthesis in both directions from the region, is useful ies in their (our unpublished results). to identify tandem repeat sequences, TrsB and TrsC, and Considering the results that Wis-2-1A retroelements are to characterize them as a genome-general sequence. distributed in Gramineae, such as barley, rye and oat, by a Each of these sequences was present in an extraordinary mechanism of horizontal spread (Moore et al., 1991), number of copies in the species with a particular genome RIRE1 seems to be also distributed in Gramineae by the in which each sequence has been found. This leads us to same mechanism. assume that both TrsB and TrsC are the sequences present as tandem repeats in a few copies in an ancestral We thank Y. Fukuta, H. Hirochika and Y. Sano for providing us rice strain, but each has been unusually amplified in the total genomic DNA from several rice strains. This work was sup- ported by a Grant-in Aid for Scientific Research from Ministry of species with a particular genome during divergence of the Education, Science, and Culture of Japan, and by a grant Pioneer- rice species belonging to the genus Oryza. ing Project in Biotechnology from the Ministry of Agriculture, We have also shown here that pOa4, which is thought to Forestry, and Fishery of Japan. be an EE genome-specific tandem repeat with a unit length of 5 ~ 16 kb (Zhao et al., 1989), is actually an inter- spersed element, named RIRE1, having partial homology REFERENCES with the LTR sequences in retrotransposons, Wis-2-1A in Appels, R., Dennis, E., Smythe, D. and Peacock, W. (1981) Two wheat and BARE-1 in barley. We have recently deter- repeated DNA sequences from heterochromatic regions of rye mined the entire sequence of RIRE1 which contains the (Secale cereale) chromosomes. Chromosoma 84, 265–277. Bedbrook, J. R., Jones, J., O’Dell, M., Thompson, R. D. and LTR sequences of 1614 bp and the internal region of 5277 Flavell, R. B. (1980) A molecular description of telomeric het- bp flanked by two LTRs (Noma et al., personal communi- erochromatin in Secale species. 19, 545–560. cation). The internal region shows high homology with Chang, T. (1984) Conservation of rice genetic resources: Luxury or those in BARE-1 and Wis-2-1A (Noma et al., personal com- necessity? Science 224, 251–256. munication). Coen, E., Robbins, T. P., Almeida, J., Hudson, A. and Carpenter, R. (1989) Consequences and Mechanisms of Transposition in In the present study, we identified the RIRE1 sequences Antirrhinum majus. In: Mobile DNA (eds.: D. E. Berg and M. in the PCR-amplified fragments forming a heterogeneous M. Howe) pp 413–436. American Society for Microbiology, bands instead of a unique ladder bands characteristic for Washington, DC. the tandem repeat sequences. It is an interesting ques- Crowhurst, R. and Gardner, R. (1991) A genome-specific repeat tion why such fragments with a portion of LTR could be sequence from Kiwifruit (Actinidia deliciosa var. deliciosa). Theor. Appl. Genet. 81, 71–78. obtained by the PCR method initially designed for the De Kochko, A., Kiefer, M. C., Cordesse, F., Reddy, A. S. and identification and characterization of tandem repeat Delseny, M. (1991) Distribution and organization of a sequences. Moore et al. (1991) have described that there tandemly repeated 352-bp sequence in the Oryzae family. exist many copies of Wis-2-1A and even solo LTRs, which Theor. Appl. Genet. 82, 57–64. are adjacent one another in the wheat genome. Even if Deragon, J.-M., Landry, B. S., Pelissier, T., Tutois, S., Tourmente, S. and Picard, G. (1994) An analysis of retroposition in plants RIRE1 and solo LTRs of RIRE1 are present in the rice based on a family of SINEs from Brassica napus. J. Mol. strain with EE genome, it does not explain the reason why Evol. 39, 378–386. the fragments with a portion of LTR of RIRE1 could be Fabijanski, S., Fedak, G., Armstrong, K. and Altossar, I. (1990). A amplified by PCR. Considering the structures of the probe for the C genome in Avena (oat). PCR-amplified fragments with a common 5'-end sequence Theor. Appl. Genet. 79, 1–7. Fedoroff, N. V. (1989). Maize Transposable Elements. In: Mobile of LTR and variable lengths of DNA segments inside of DNA (ed.: D. E. Berg and M. M. Howe), pp 375–411. Ameri- LTR (see Figs. 4 and 5A), we propose that there exist some can Society for Microbiology, Washington, DC. RIRE1 elements which accepted an insertion of another Flavell, A. J., Dunbar, E., Anderson, R., Pearce, S., Hartley, R. 382 R. NAKAJIMA et al.

and Kumar, A. (1992) Ty1-copia group retrotransposons are able elements. Adv. Bot. Res. 12, 103–203. ubiquitous and heterogeneous in higher plants. Nucleic Ac- Ohtsubo, H. and Ohtsubo, E. (1994) Involvement of transposition ids Res. 20, 3639–3644. in dispersion of tandem repeat sequences (TrsA) in rice Gierl, A. and Saedler, H. (1992). -transposable elements and genomes. Mol. Gen. Genet. 245, 449–455. tagging. Plant Mol. Biol. 19, 39–49. Ohtsubo, H., Umeda, M. and Ohtsubo, E. (1991) Organization of Harberd, N. P., Flavell, R. B. and Thompson, R. D. (1987) Identifi- DNA sequences highly repeated in tandem in rice genomes. cation of a transposon-like insertion in a Glu-1 allele of Jpn. J. Genet. 66, 241–254. wheat. Mol. Gen. Genet. 209, 326–332. Peterson, P. (1987). Mobile elements in plants. CRC Crit. Rev. Hirochika, H. and Hirochika, R. (1993) Ty1-copia group retro- Plant Sci. 6, 104–208. transposons as ubiquitous components of plant genomes. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R., Jpn. J. Genet. 68, 35–46. Horn, G. T., Mullis, K. B. and Erlich, H. A. (1988) Primer- Lichtenstein, C. P. and Draper, J. (1985) Genetic engineering of directed enzymatic amplification of DNA with a thermostable plants. In: DNA cloning, Vol. II (ed.: D. M. Glover), pp 67– DNA polymerase. Science 239, 487–491. 119. IRL Press, Oxford, Washington, DC. Sanger, F., Nicklen, S. and Coulson, A. R. (1977) DNA sequencing Manninen, I. and Schulman, A. H. (1993) BARE-1, a copia-like with chain-terminating inhibitors. Proc. Natl. Acad. Sci. retroelement in barley (Hordeum vulgare L.). Plant Mol. USA 74, 5463–5467. Biol. 22, 829–846. Shepherd, N. S., Schwartz-Sommer, Z., Vel Spalve, J. B., Gupta, Martinez, C. P., Arumuganathan, K., Kikuchi, H. and Earle, E. D. M., Wiehand, U. and Saedler, H. (1984) Similarity of the Cin1 (1994) Nuclear DNA content of ten rice species as determined repetitive family of Zea mays to eukaryotic transposable by flow cytometry. Jpn. J. Genet. 69, 513–523. elements. Nature 307, 185–187. Martinez-Zapater, J., Estelle, M. and Somerville, C. (1986) A Tenzen, T., Matsuda, Y., Ohtsubo, H. and Ohtsubo, E. (1994) highly repeated DNA sequence in Arabidopsis thaliana. Transposition of Tnr1 in rice genomes to 5'-PuTAPy-3', dupli- Mol. Gen. Genet. 204, 417–423. cating the TA sequence. Mol. Gen. Genet. 245, 441–448. Meikoth, J. and Wahl, G. (1984) Hybridization of nucleic acids Voytas, D. F., Cummings, M. P., Konieczny, A., Ausubel, F. M. immobilized on solid supports. Anal. Biochem. 138, 267– and Rodermel, S. R. (1992) copia-like retrotransposons are 284. ubiquitous among plants. Proc. Natl. Acad. Sci. USA 89, Messing, J. (1983). New M13 vectors in cloning. Methods 7124–7128. Enzymol. 101, 20–78. Wu, T. and Wu, R. (1987) A new rice repetitive DNA shows se- Mochizuki, K., Ohtsubo, H., Hirano, Y., Sano, Y. and Ohtsubo, E. quence homology to both 5s RNA and tRNA. Nucleic Acids (1993) Classification and relationships of rice strains with AA Res. 15, 5913–5923. genome by identification of transposable elements at nine Wu, T. and Wu, R. (1992) A novel repetitive DNA sequence in the loci. Jpn. J. Genet. 68, 205–217. genus Oryza. Theor. Appl. Genet. 84, 136–144. Mochizuki, K., Umeda, M., Ohtsubo, H. and Ohtsubo, E. (1992) Yoshioka, Y., Matsumoto, S., Kojima, S., Ohshima, K., Okada, N. Characterization of a plant SINE, p-SINE1, in rice genomes. and Machida, Y. (1993) Molecular characterization of a short Jpn. J. Genet. 67, 155–166. interspersed repetitive element from tobacco that exhibits Moore, G., Lucas, H., Batty, N. and Flavell, R. B. (1991) A family sequence homology to specific tRNAs. Proc. Natl. Acad. Sci. of retrotransposon and associated genomic variation in USA 90, 6562–6566. wheat. Genomics 10, 461–468. Zhao, X., Wu, T., Xie, Y. and Wu, R. (1989) Genome-specific re- Nevers, P., Shepherd, N. S. and Saedler, H. (1986) Plant transpos- petitive sequences in the genus Oryza. Theor. Appl. Genet. 78, 201–209.

Top View