Translation start sequences affect the efficiency of silencing of Agrobacterium tumefaciens T-DNA oncogenes

Abstract

Agrobacterium tumefaciens oncogenes cause transformed plant cells to overproduce auxin and cytokinin. Two oncogenes encode enzymes that convert tryptophan to indole-3-acetic acid (auxin): iaaM (tryptophan mono-oxygenase) and iaaH (indole-3-acetamide hydrolase). A third oncogene (ipt) encodes AMP isopentenyl transferase, which produces cytokinin (isopentenyl-AMP). Inactivation of ipt and iaaM (or iaaH) abolishes tumorigenesis. Because adequate means do not exist to control crown gall, we created resistant plants by introducing transgenes designed to elicit posttranscriptional gene silencing (PTGS) of iaaM and ipt. Transgenes that elicit silencing trigger sequence-specific destruction of the inducing RNA and messenger RNAs with related sequences. Although PTGS has proven effective against a variety of target genes, we found that a much higher percentage of transgenic lines silenced iaaM than ipt, suggesting that transgene sequences influenced the effectiveness of PTGS. Sequences required for oncogene silencing included a translation start site. A transgene encoding a translatable sense-strand RNA from the 5' end of iaaM silenced the iaaM oncogene, but deletion of the translation start site abolished the ability of the transgene to silence iaaM. Silencing A. tumefaciens T-DNA oncogenes is a new and effective method to produce plants resistant to crown gall disease.

Figure 1.

Transformation vector and transgene constructs. A, The T-DNA portion of the plant transformation vector pPEV6. Boxes labeled LB and RB indicate left- and right-hand border sequences; the right border includes an overdrive sequence. The box labeled CaMV represents the CaMV 35S promoter, and the lines labeled 5′-UTR and 3′-UTR indicate the 5′- and 3′-UTRs from CaMV. Boxes labeled NOS and Term denote the nopaline synthase promoter and 3′-UTR/polyadenylation signal, which facilitate expression of the neomycin phosphotransferase gene, labeled nptII. Arrowheads show the BamHI restriction site used to insert transgenes downstream of the CaMV 35S promoter and the EcoRI site used to estimate T-DNA copy number. B, The iaaM-stop (solid line) and ipt-stop (shaded line) transgenes. Sequences indicate start codons (ATG) in bold italic letters and inframe or native stop codons in bold letters.

Figure 2.

Silencing of the iaaM and ipt oncogenes in transgenic tobacco. A, The responses of iaaM-silencing tobacco line TDP1-B7 to infection with Ti-plasmidless (A136), wild-type (A348), and ipt-mutant (338) A. tumefaciens. B, The responses of non-silencing line TDP1-B31 to infection with the same strains. C, The responses of ipt-silencing tobacco line CW1-K27 to infection with Ti-plasmidless, wild-type, and iaaM-mutant (328) A. tumefaciens. D, The responses of ipt::iaaM-silencing line CW4-B30 to infection with these A. tumefaciens strains.

Figure 3.

Southern-blot analysis of transgene structure and copy number in transgenic tobacco. A, Transgene structure iaaM-stop (TDP1) and ipt::iaaM-stop (CW4) lines. Genomic DNA was digested with BamHI (lanes 2–6 and 10 [5 μg lane^–1]; lanes 1, 7, and 8 [10 μg lane^–1]; lane 9 [4 μg]); blots were probed with labeled iaaM-stop DNA. Lanes 1 through 4 contained DNA isolated from CW4 lines B1 (lane 1), B11 (lane 2), B22 (lane 3), and B30 (lane 4). Lanes 5 through 8 contained DNA from TDP1 lines B7 (lane 5), B17 (lane 6), B27 (lane 7), and B31 (lane 8). Lanes 9 and 10 contained DNA from PEV6 (vector-only) lines B2 (lane 9) and B14 (lane 10). B, Transgene copy number in TDP1 and CW4 tobacco lines. Genomic DNA was digested with EcoRI (5 μg lane^–1); blots were probed with labeled iaaM-stop DNA. Lanes 1, 2, 7, and 8 contained DNA from TDP1 (iaaM-stop) lines B7, B27, B17, and B31, respectively. Lanes 3 through 6 contained DNA from CW4 (ipt::iaaM-stop) lines B1, B11, B22, and B30, respectively. Lanes 9 and 10 contained DNA from PEV6 (vector-only) lines B2 and B14. Lane M contained phage λ-DNA digested with HindIII; numbers beside this lane indicated fragment size in base pairs. C, Transgene copy number and structure in ipt-stop (CW1) tobacco lines. Genomic DNA (10 μg lane^–1) from lines CW1-K27 (lanes 1 and 4) and CW1-K52 (lanes 2 and 5) was digested with either Hinf1 (lanes 1 and 2) or EcoRI (lanes 4 and 5). Lane 3 contained 100 ng of pUC119-ipt plasmid DNA (equivalent to 5 copies genome^–1) digested with Hinf1. Numbers on the left side of the gel indicate the size (in base pairs) and location of λ-HindIII restriction fragments.

Figure 4.

Northern-blot analysis of sense-strand RNA accumulation in iaaM-stop (TDP1) and ipt::iaaM-stop (CW4) tobacco lines. Each lane contains 10 μg of total RNA isolated from lines TDP1-B7, -B17, -B27, and -B31 in lanes 1–4, and lines CW4-B1, -B11, -B22, and -B30 in lanes 5–8. Lanes 9 and 10 contain RNA from the vector-only lines PEV6-B2 and B14. A, Northern blot probed with radiolabeled iaaM antisense RNA. B, The same filter stained with methylene blue to reveal ribosomal RNAs. C, The same northern blot probed with radiolabeled ubiquitin DNA.

Figure 5.

Northern-blot analysis of antisense RNA accumulation in iaaM-stop (TDP1) tobacco lines. Each lane contains 10 μg of total RNA isolated from iaaM-stop lines TDP1-B7, -B17, -B27, and -B31 in lanes 1 through 4, respectively. Lanes 5 and 6 contain total RNA from vector-only lines PEV6-B2 and B14. A, A northern blot probed with radiolabeled iaaM-stop DNA (both strands). B, Ribosomal RNA on the same filter stained with methylene blue.

Figure 6.

Northern-blot and real-time RT-PCR analyses of sense-strand RNA accumulation in ipt-stop (CW1) tobacco lines. A, Northern blot probed with radiolabeled ipt DNA; B, total RNA stained with methylene blue. Each lane contained 10 μg of total RNA isolated from ipt-stop lines CW1-K27 (lane 1) and CW1-K52 (lane 2). C, The results of real-time RT-PCR performed on RNA isolated from nuclei of an ipt-silencing line (CW1-K27), a non-silencing line (CW1-K52), and a vector-only control (PEV6-B14). RNA levels were normalized to the amount of 5.8S rRNA detected in each sample; pg RNA indicates picograms of ipt RNA per picogram of 5.8S rRNA. No ipt RNA (<0.07 fg) was detected in nuclei from ipt-silencing line CW1-K27 or vector-only line PEV6-B14, which lacked ipt sequences.

Figure 7.

Transgenes encoding sense and antisense RNA. RB and LB indicate right and left T-DNA borders, respectively. Arrows labeled peCaMV and peFMV indicate the location and orientation of enhanced CaMV 35S and FMV promoters. The arrowhead labeled HSP70L represents the 5′-untranslated leader sequence from the petunia (Petunia hybrida) heat shock protein 70 gene. Arrows labeled iaaM-stop and ipt-stop show the location, orientation, and length of oncogene-silencing transgenes containing iaaM and ipt sequences. Numbers in parentheses list the portions of the iaaM-stop transgene contained in pJP17 and pJC14. The arrow labeled NptII shows the position and orientation of the neomycin phosphotransferase gene, which is flanked by the nopaline synthase promoter and 3′-UTR and polyadenylation signal.

Figure 8.

Gene silencing during mixed infections. K. daigremontiana stems (A and B) and potato discs (C and D) co-inoculated with wild-type A. tumefaciens (A348) and disarmed strain EHA101 containing oncogene-silencing plasmid pJP17 (A and C) or an empty vector pJP21 (B and D).