JOBNAME: horts 43#3 2008 PAGE: 1 OUTPUT: April 23 08:32:38 2008 tsp/horts/163067/02585 HORTSCIENCE 43(3):962–964. 2008. oligo probes or DNA probes complementary to PG. DNA oligo probes were ordered from IDT; a ‘‘+’’ symbol preceding represents an The Flavr Savr Tomato, an Early LNA-modified nucleotide. Sequences used Example of RNAi Technology were as follows: Elysia K. Krieger1, Edwards Allen, Larry A. Gilbertson, A1: AAAC+ATA+TGA+TAATATT+GCA+ TTT+GAGCAAG+CAT+GGA+ATGAA and James K. Roberts A2: ATTA+TAA+TGG+AGAATAT+AAA+ Monsanto Company, Chesterfield Campus, 700 Chesterfield Parkway, TTA+GTAGGGG+AAA+GTG+GAAAA Chesterfield, MO 63017 B1: ATGC+AAT+ATT+ATCATAT+GTT+ TTT+CCATCAC+CCT+TAG+CTCCA William Hiatt and Rick A. Sanders B2: ATAG+TCT+ATA+ATTATGG+GAT+ Monsanto Company, Calgene Campus, 1920 5th Street, Davis, CA 95616 ACT+TAACGTC+TTG+CAT+TTCCA Additional index words. Flavr Savr tomato, polygalacturonase, RNAi, T-DNA linkage Oligos were isotopically labeled with Abstract. The Flavr Savr tomato was introduced as the first genetically engineered whole [g-32P]ATP as described in Allen et al. food in 1994. The commercial event, resulting from transformation with an antisense (2005). DNA probes: Probe C—First 483 expression cassette of the endogenous polygalacturonase gene, was sequenced and found bp of the NptII sequence. Probe D—Last to contain two contiguous, linked, transfer DNA insertions. We found polygalacturonase 420 bp of the PG sequence. DNA probes suppression correlates with accumulation of ’21-nt small interfering RNAs, the were labeled with dCTP using the RadPrime hallmark of an RNA interference-mediated suppression mechanism. DNA-labeling system (catalog no. 18428- 011, Invitrogen) according to the manufac- turer’s directions. The Flavr Savr tomato has increased Southern blots and inverse PCRs. Geno- Results storage life through suppression of the to- mic DNA was isolated from tomato leaves, mato polygalacturonase (PG) gene, resulting and EcoRV Southern blots were done accord- Commercial Flavr Savr locus contains from transformation of an antisense expres- ing to Redenbaugh et al. (1992). To make two linked T-DNAs. Plants from commer- sion cassette of the PG cDNA (pCGN1436) inverse PCR libraries, 10 mg of genomic cial hybrid seed (CR3-1436-613:@.12/ (Sheehy et al., 1988). Efficacious but non- DNA was digested separately with XhoI, CF5013), not included in the previous anal- commercial pCGN1436 transformation NcoI, EcoRV, HindIII, and PstI and ligated ysis (Sanders and Hiatt, 2005), were grown events contain multiple T-DNA insertions to itself. Primers complementary to regions and selfed to obtain positive (4523, 3951, linked by right borders (Redenbaugh et al., of the T-DNA were used to walk out across 3949) and null (4193, 4131, 3975) sibling 1992; Sanders and Hiatt, 2005), leading to the the T-DNA inserts as well as into genomic plants for molecular analysis. Genomic DNA hypothesis that readthrough transcription of flanking regions. PCRs were done with Extaq from these plants was digested with EcoRV, the PG-linked antisense (PGAS) cassettes re- (Takara Bio, Shiga, Japan), and products and Southern blot hybridization was per- sulted in RNA interference (RNAi)-mediated were cloned and sequenced using pCR2.1- formed to determine T-DNA copy number PG suppression, rather than antisense- TOPO (Invitrogen Corp., Carlsbad, CA). and linkage (Fig. 1A). Bands were observed mediated sequestration, of the endogenous PCR primers. in the transgenic plants of 3698- and 5162-bp PG transcript (Watson et al., 2005). To test 1: TATACCCGCAGTCCGCTCACCCTACC corresponding to the PG region of the locus, this hypothesis, we determined if RNAi sup- 2: ACACGAAACTCATGAGGGAGGAGATG in addition to a 2-kb band of the native pression is responsible for the Flavr Savr phe- 3: GAGACCGATGTTCGTTCCGGAACCTT PG gene present in all lanes. This is consis- notype by characterizing the T-DNA locus of 4: AAGGTTCCGGAACGAACATCGGTCTC tent with the presence of two copies of the the Flavr Savr tomato event and analyzing for 5: ATTCAAAAGTCGTTAATGGCTGCGG PG region of the T-DNA. The transgene the presence of 21-nt PG small interfering ATCAAG locus and flanking genomic sequence were RNAs (siRNAs) diagnostic of RNAi. 6: CAATTGTAAATGGCTTCATGTCCGG obtained by inverse PCR amplification, clon- GAAATC ing, and sequence analysis of resulting frag- Materials and Methods ments. The locus sequence confirmed the PG enzyme activity assays. PG enzyme T-DNA linkage at the left border (LB) of Tomato plants. CF5013 is the tomato activity assays, measured as an A540 absor- one of the two T-DNAs; the second T-DNA is cultivar that resulted from a nonconventional bance, were done in replicates of six accord- truncated, removing all of the LB sequences cross between the homozygous PGAS inbred ing to Redenbaugh et al. (1992). The average and part of the MAS promoter (Fig. 1B). line CR3-1436-613:@.12 with the tomato enzyme activity of the three transgenic toma- These results were further confirmed by hybrid Mountain Spring. CF5013 is hetero- toes (4523, 3951, 3949) was 0.058, 0.073, amplifying the T-DNAs with flanking geno- zygous for the PGAS gene which is dominant and 0.075, respectively, compared with the mic and unique junction region primers and was sold commercially as the Flavr Savr nulls (4193, 4131, 3975), which were 1.032, (Fig. 1C) in transgenic (4523) and null tomato. These plants were selfed, and result- 1.11, and 0.98. (4193) siblings. The 6974-bp band (lane 1) ing sibling plants were screened by taqman RNA purification and Northern blots. contains the truncated (left) T-DNA. The PCR for NptII. Of the 64 plants screened, the Total RNA was purified from ripe tomatoes 7725-bp band (lane 3) contains the full-length ratio was 43 to 21 NptII positive to negative using Trizol (Invitrogen) and further cleaned (right) T-DNA. We were unable to amplify plants, which is consistent with a 3:1 ratio of using the RNA/DNA Midi Kit (Qiagen, across the insert (14,674-bp, lane 5), likely transgenic to non-transgenic tomato plants Valencia, CA), according to procedures pro- due to the inverted repeat sequence. In the (c2 P = 0.15). vided by the manufacturer. Low molecular null sibling, these primers amplify a 1058-bp weight Northern blots were done with 17% band (lane 6) verifying insert location. The Received for publication 23 Dec. 2007. Accepted polyacrylamide gels (Llave et al., 2002a), for publication 6 Jan. 2008. 764-bp band (lane 7) is the product across We thank Daniel Free and Christina Kavanaugh and high molecular weight Northern blots the unique T-DNA junction sequence. Taken for greenhouse support. were done with 1.5% agarose gels (Llave together, these results show the commercial 1To whom reprint requests should be addressed; et al., 2002b). Blots were hybridized with Flavr Savr tomato contained two T-DNAs e-mail [email protected] radioactive 42-nt isotopically labeled LNA linked in an inverted orientation. 962 HORTSCIENCE VOL. 43(3) JUNE 2008 JOBNAME: horts 43#3 2008 PAGE: 2 OUTPUT: April 23 08:32:39 2008 tsp/horts/163067/02585 an abundance of 21-nt siRNAs associated with sequence-specific mRNA degradation and translational repression (Brodersen and Voinnet, 2006; Fagard and Vaucheret, 2000; Hamilton et al., 2002; Watson et al., 2005). The 24-nt class, associated with transcrip- tional repression (Brodersen and Voinnet, 2006; Fagard and Vaucheret, 2000; Hamilton et al., 2002; Watson et al., 2005), was less abundant and some PG mRNA was still detectable in Flavr Savr tomatoes. If suppres- sion of PG activity was due to sequestration from translation by binding to antisense RNA (Metzlaff et al., 1997), the accumulation of sense PG transcript should be unaffected and antisense transcripts would be detected. In contrast, if suppression were due to trans- criptional or post-transcriptional gene silenc- ing, accumulation of the message would be reduced, suggesting transgene-associated suppression was not due to repression of endogenous PG mRNA transcription but to RNAi (Brodersen and Voinnet, 2006). It is interesting that siRNAs were detected for PGAS and NptII, but not for any other regions of the T-DNA. This suggests that Fig. 1. Characterization of siblings of the Flavr Savr tomato. (A) Southern blot confirmation of T-DNA suppression requires transcription from the number using probe D of the PG gene. Transgenic (4523, 3951, and 3949) and non-transgenic (4193, associated PG and NptII promoters. Using 4131, and 3975) siblings were analyzed. The T-DNA insertion corresponds to the 3.6- and 5.1-kb flanking sequence as a query in searches with hybridizing bands. (B) Diagram of the sequenced 13,621-bp T-DNA insertion and flanking genomic BLAST, we identified a region 1197-bp regions. DNA probes (bars C and D) and primers (arrows 1–6) used for confirmation PCRs are downstream of the T-DNA insert containing indicated. A putative CYP82 family cytochrome P450 mono-oxygenase gene in the adjacent genomic sequence is indicated (gray box). (C) Confirmation of linked T-DNA orientation by PCR with genomic a putative cytochrome P450 gene (Fig. 1B). flank and unique junction sequence primers. PCRs are shown in pairs; odd-numbered lanes use Potential promoter or enhancer elements as- transgenic sibling 4523 genomic DNA, even-numbered lanes use null sibling 4193 genomic DNA. sociated with this gene may be associated Sizes of the marker bands (lane L) are indicated at right. Left T-DNA, lanes 1 and 2 (primers 1 and 4); with transcription across the T-DNA locus right T-DNA, lanes 3 and 4 (primers 2 and 3); full insert, lanes 5 and 6 (primers 1 and 2); unique and could be responsible for contributing junction region, lanes 7 and 8 (primers 5 and 6). to the initial production of a dsRNA for the coding regions, thus resulting in the siRNAs from transgenes accumulate in that the level of NptII mRNA was low, con- production of the siRNAs.