436 Brief Communication
Transcriptional and posttranscriptional gene silencing are mechanistically related Titia Sijen*, Irma Vijn†, Alexandra Rebocho, Rik van Blokland‡, Dick Roelofs§, Joseph N.M. Mol and Jan M. Kooter Two distinct gene-silencing phenomena are observed in plants: transcriptional gene silencing (TGS), which involves decreased RNA synthesis because of promoter methylation, and posttranscriptional gene silencing (PTGS), which involves sequence-specific RNA degradation. PTGS is induced by deliberate [1–4] or fortuitous production (R.v.B., unpublished data) of double-stranded RNA (dsRNA). TGS could be the result of DNA pairing [5], but could also be the result of dsRNA, as was shown by the dsRNAinduced inactivation of a transgenic promoter [6]. Here, we show that when targeting flower pigmentation genes in Petunia, transgenes expressing dsRNA can induce PTGS when coding sequences are used and TGS when promoter sequences are taken. For both types of silencing, small RNA species are found, which are thought to be dsRNA decay products [7] and determine the sequence specificity of the silencing process [8, 9]. Furthermore, silencing is accompanied by the methylation of DNA sequences that are homologous to dsRNA. DNA methylation is assumed to be essential for regulating TGS and important for reinforcing PTGS [10]. Therefore, we conclude that TGS and PTGS are mechanistically related. In addition, we show that dsRNA-induced TGS provides an efficient tool to generate gene knockouts, because not only does the TGS of a PTGS-inducing transgene fully revert the PTGS phenotype, but also an endogenous gene can be transcriptionally silenced by dsRNA corresponding to its promoter. Address: Department of Developmental Genetics, Institute for Molecular Biological Sciences, BioCentrum Amsterdam, Vrije Universiteit, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands. Present addresses: * Hubrecht Laboratory, Netherlands Institute for Developmental Biology, Uppsalalaan 8, 3584 CT Utrecht, † Department of Phytopathology, Wageningen University, Binnenhaven 9, 6709 PD Wageningen, ‡ ChromaGenics B.V., Plantage Muidergracht 12, 1018 TV Amsterdam, and § Promega Benelux, Kenauweg 34, 2331 BB Leiden, The Netherlands. Correspondence: Jan M. Kooter E-mail:
[email protected] Received: 22 December 2000 Revised: 9 February 2001 Accepted: 9 February 2001 Published: 20 March 2001 Current Biology 2001, 11:436–440 0960-9822/01/$ – see front matter 2001 Elsevier Science Ltd. All rights reserved.
Results and discussion Coding sequence dsRNAs induce PTGS with high efficiency
The introduction of transgenes harboring coding sequences of the flower pigmentation gene chalcone synthaseA (chsA) into Petunia induces PTGS as indicated by white flower tissue [11–13]. To determine whether dsRNA triggers chsA PTGS, the constructs 35S-IR.chs and IR.chs were transformed into Petunia. They carry two chsA cDNAs that are arranged as inverted repeats (IR). In 35S-IR.chs, they are controlled by an enhanced 35S promoter (Figure 1a), whereas in IR.chs they are promoterless. DNA blot analysis identified transformants carrying at least one intact T-DNA (data not shown). For IR.chs, three out of ten transformants produced flowers that were completely or partially white (Table 1). For 35S-IR.chs, all 11 transformants showed chsA silencing (Table 1; Figure 1b), although the extent of silencing varied from fully white flowers (Figure 1b, plants #5, #7, and #10) to flowers with only some white spots (Figure 1b, plant #4). To confirm that the lack of pigmentation was because of the PTGS of chsA, quantitative RT-PCR analysis was performed. PTGS is known to target only mature mRNAs and not unspliced nuclear precursor mRNAs. In accordance with PTGS, the amount of unspliced chsA mRNA in 35S-IR.chs plants was comparable to that in wild-type plants, while mature chsA mRNA levels were reduced (data not shown). RNaseA/ T1 protection assays on RNA that was pretreated with RNaseA under high salt conditions were used to examine silenced IR.chs and 35S-IR.chs plants for the production of dsRNA, which would be produced unintentionally from the IR.chs construct and deliberately from the 35S-IR.chs construct. All silenced plants that were tested contained dsRNA, while none of the nonsilenced plants did (Table 1, IR.chs plants; Figure 1b). Thus, chsA PTGS is associated with the presence of chsA dsRNA. A hallmark of PTGS is the presence of small RNAs of both sense and antisense polarity [14]. RNA blot analysis of the 35SIR.chs plants showed the presence of sense and antisense small RNAs in both flowers (Figure 1c) and leaves (data not shown).
TGS and PTGS are mechanistically related
To show that dsRNA-containing promoter sequences can trigger TGS in Petunia, we constructed transgenes in which the mannopine synthase promoter (Pmas) is driving IRs of different parts of the 35S promoter, namely the full promoter (Pmas-IR.35Sfull), the enhancer sequences (Pmas-IR.35Senh), and the minimal promoter (Pmas-
Brief Communication 437
Figure 1
Coding sequence dsRNA–induced PTGS of chs expression. (a) Map of the constructs IR.chs and 35S-IR.chs. The arrows inside the boxes labeled “chsA” indicate the orientation of the cloned cDNAs; the hooked arrows mark transcription start sites. Probe I was used for the RNaseA/T1 protection assays shown in Figure 1b, and gives rise to a protected fragment of 318 nt; probe II was used for the protection assays shown in Figure 2c, and gives rise to a fragment of 398 nt. (b) Flower phenotypes of eight transgenic plants harboring the construct 35S-IR.chs, and the RNaseA/T1 protection assay showing the presence of chsA dsRNA. The transgene copy number for each transformant is indicated at the top whereby ␦ refers to a T-DNA locus containing the nptII gene but lacking an intact chsA-IR. As controls, a nonsilenced wild-type plant and the posttranscriptionally silenced transformant PSE19-1 [23] were used. (c) RNA blot analysis on 15% polyacrylamide gels to show the presence of chsAspecific small RNAs of sense and antisense polarity. The size marker is a DNA oligonucleotide of 23 nt.
IR.35Smini; Figure 2a). These constructs were introduced into transformant 35S-IR.chs#3 (Figure 2b, panel I). Double transformants containing both the 35S-IR.chs#3 locus and an intact Pmas-IR.35Sfull (re.full), Pmas-IR.35Senh (re.enh), or Pmas-IR.35Smini transgene (re.mini) were identified by DNA blot analysis (data not shown). All four re.full plants showed partial or almost full reversion to
the wild-type pigmentation as indicated by the purple flowers occasionally containing a white spot (Figure 2b, panel II). The three re.enh plants were fully wild type (Figure 2b, panel III). Two of the five re.mini plants remained silenced, while the other three showed some reversion (Figure 2b, panel IV). Nuclear RNA of the plants was analyzed for the presence of 35S dsRNA and chsA dsRNA. While no 35S dsRNA was detected in the re.mini plants (Figure 2c, lane 5), it was clearly present in the reverted re.full and re.enh plants (Figure 2c, lanes 3 and 4). In the re.enh plants, a protected fragment of about the expected length (261 nt) was found (Figure 2c, lane 4). In the re.full plants, the protected fragment was much shorter (about 80 nt) than expected (359 nt; Figure 2c, lane 3). It is unclear how the dsRNA that gives rise to this short protected fragment is produced, but it is conceivable that events at the IR-35S promoters may interfere with transcription from the Pmas promoter. The presence of 35S dsRNA in the reverted re.full and re.enh plants correlates with the absence of chsA dsRNA (Figure 2c, lanes 9 and 10). Reversion of the chs-silenced phenotype by 35S dsRNA–induced repression of chsA dsRNA production indicates that chs silencing is mediated by chsA dsRNA. Consistent with the absence of chsA dsRNA, no chsA-specific small RNAs were detected in the flowers of the re.enh and re.full plants (data not shown). Given that TGS is known to be due to the methylation of promoter sequences, we examined the methylation status of the 35S promoters in the transformant 35SIR.chs#3 and in the double-transformed revertant plants by digesting DNA with the m5C-methylation-sensitive restriction enzyme, AluI. In 35S-IR.chs#3, which contains two copies of the construct, the AluI sites in the 35S promoters were partially methylated, as three 35S-hybridizing fragments were found (Figure 2d, lane 4). The largest corresponds to a fully methylated promoter, whereas the smaller ones correspond to partially methylated promoters. Apparently, this partial methylation does not prevent the 35S promoter from being active, because the plant produces chsA dsRNA (Figure 1b, plant #3). In the re.full and re.enh plants, the methylation of the 35S promoter of the 35S-IR.chs transgenes was enhanced, as only the fragment corresponding to the complete promoter was detected (Figure 2d, lanes 5 and 6, blue “m”). Also, the 35S sequences of the Pmas-IR.35Sfull and PmasIR.35Senh transgenes, in re.full and re.enh plants, respec-
Table 1 PTGS and TGS correlate with the presence of dsRNA. Transgene construct
Plants with intact transgene
Silenced plants
dsRNA/silenced plants tested
dsRNA/nonsilenced plants tested
IR.chs 35S-IR.chs 35S-IR.Pdfr
10 11 6
3 11 6
2/2 7/7 6/6
0/3 n.a. n.a.
n.a.: not applicable
438 Current Biology Vol 11 No 6
Figure 2
Promoter dsRNA–induced TGS of a PTGS-inducing transgene. (a) Maps of three different Pmas-IR.35S constructs. The arrows inside the boxes indicate the orientation of the 35S promoter sequence; hooked arrows mark transcription start sites. The indicated probe was used in RNaseA/T1 protection assays, and is expected to result in a protected fragment of 359 nt with the construct Pmas-IR.35Sfull, a fragment of 261 nt with the construct Pmas-IR.35Senh, and a fragment of 123 nt with the construct Pmas-IR.35Smini. (b) Typical phenotypes of double transformants: 35S-IR.chs#3 (I), re.full (II), re.enh (III), and re.mini (IV). (c) Detection of dsRNA by RNaseA/T1 protection assays in RNaseA-pretreated nuclear RNA of a wild-type plant and the plants described in (b). To detect 35S dsRNA, the probe depicted in (a) was used, while for the detection of chsA dsRNA, probe II indicated in Figure 1a was used. (d) DNA blot analysis showing increased methylation of AluI sites in the 35S promoter in re.full and re.enh plants. The DNA of transformants or of a mixture of wild-type and plasmid was digested with AluI and hybridized with
a 35S promoter. For each construct, the expected fragments are indicated below the DNA blot, and the size (in bp) is indicated above each fragment. Fragments for the 35S-IR.chs construct are presented in blue, those for the Pmas-IR.35Sfull construct in green, and those for Pmas-IR.35Senh in red. “u” indicates the fragments expected from an unmethylated 35S promoter, “p” indicates fragments from a partially methylated promoter, “m” indicates fragments from a fully methylated promoter, and “e” indicates the fragments obtained when sequences outside the 35S promoter are methylated. “A” indicates an AluI site. In the DNA blot next to the lanes are the fragments marked whereby the colors refer to the constructs shown below the blot, whereas the letters refer to the methylation status. (e) RNaseA/ T1 protection assay on cytoplasmic RNA of wild-type, 35S-IR.chs#3, re.enh, and re.full plants to detect 35S-specific small RNAs. The probe used is indicated in (a). Size markers were two RNA oligonucleotides of 23 and 27 nt.
tively were methylated. The fragments detected were longer (Figure 2d, lanes 5 and 6, marked in green or red) than those from the unmethylated plasmid DNAs (Figure 2d, lanes 2 and 3, green and red “u”). Also, AluI sites bordering the 35S promoters were partially methylated, given that also fragments (Figure 2d, lanes 5 and 6, green or red “e”) that were longer than those corresponding to fully methylated 35S promoter fragments (Figure 2d, lanes 5 and 6, green or red “m”) were observed. Beside DNA methylation, dsRNA-induced silencing is accompanied by the production of small RNAs (Figure 1d) [6, 8, 9, 15]. By RNaseA/T1 protection assays, we analyzed cytoplasmic fractions of the revertant plants for the presence of 35S small RNAs. Figure 2e (lanes 3 and 4) shows
that re.enh and re.full plants indeed contain 35S small RNAs. The level in the re.full plants was lower than in the re.enh plants, which is probably because of the much shorter dsRNA derived from the Pmas-IR.35Sfull transgene (80 nt) than that from Pmas-IR.35Senh (270 nt; Figure 2c, lanes 3 and 4). The observations that both TGS and PTGS can be induced by dsRNA, that small RNAs are produced, and that cognate DNAs become methylated suggest that TGS and PTGS are mechanistically linked. The cellular machinery does not seem to discriminate between dsRNAs consisting of coding and promoter sequences, and both are cleaved into small RNAs [7]. When homologous
Brief Communication 439
Figure 3
Models for PTGS and TGS with a central role for dsRNA. (a) PTGS is induced when the dsRNA consists of coding sequences. dsRNAs are cleaved into guide RNAs of 23 to 21 nt in length. When complementary mRNAs are present, these guide RNAs may pair with the mRNA and mediate their degradation, as indicated by the red cross. This decay may occur after the elongation of the guide RNAs by an RdRP (indicated in green). In addition, the dsRNAs or the small RNAs direct the methylation of homologous DNA sequences as indicated by black ovals. (b) TGS is induced when the dsRNA consists of promoter sequences. dsRNAs are cleaved into small RNAs of 23 to 21 nt, and either these small RNAs or the dsRNA itself direct the methylation of corresponding DNA sequences (black ovals). This promoter methylation inhibits transcription, as indicated by the red cross.
mRNAs are present, the small antisense RNAs may anneal and guide the degradation of these mRNAs (Figure 3a). Prior to degradation, an RNA-dependent RNA polymerase (RdRP) [16–18] may elongate some of the guide RNAs on the mRNA, after which the dsRNA part is degraded. In addition, either guide RNA or dsRNA upon pairing or looping-in [19] may trigger the methylation of homologous DNA. The methylation of coding sequences does not seem to affect transcription, while the methylation of promoter sequences usually results in promoter inactivation (Figure 3b), probably by histone deacetylation and chromatin condensation. dsRNA-induced TGS of an endogenous gene
To determine whether dsRNA could be used to transcriptionally downregulate endogenous genes, we raised transgenic plants that were designed to express dsRNA for the promoter of the flower pigmentation gene dihydroflavonol 4-reductaseA (dfrA). These plants carried a 35S-IR.Pdfr
Figure 4
Promoter dsRNA–induced TGS of the endogenous dfrA gene. (a) Schematic representation of T-DNA construct 35S-IR.Pdfr. The arrows inside the boxes labeled “Pdfr” indicate the orientation of the dfrA promoter; the hooked arrows mark transcription start sites. The indicated probe was used in RNaseA/T1 protection assays and is expected to result in a protected fragment of 304 nt with the construct 35S-IR.Pdfr. (b) Flower phenotypes of two dsPdfr plants, I and II. (c) Detection of dsRNA by an RNaseA/T1 protection assay on RNaseA-pretreated RNA of wild-type, dsPdfr.I, and dsPdfr.II plants. The probe used is indicated in (a). (d) Semiquantitative RT-PCR analysis on wild-type, dsPdfr.I, and dsPdfr.II plants and the posttranscriptionally silenced transformant PSEdfr-chs.1c. Forward primers that are specific for gapdh, the second chsA exon, the chsA intron, the fifth dfrA exon, and the fourth dfrA intron were used. The number of amplification cycles is indicated.
transgene composed of two inversely oriented dfrA promoters (without a transcription start site and TATA box, denoted Pdfr) and controlled by an enhanced 35S promoter (Figure 4a). Transformants carrying at least one intact transgene were identified by DNA blot analysis (data not shown) and named dsPdfr. The overall pigmentation of the flowers of dsPdfr plants was severely reduced relative to wild-type flowers (Figure 4b), suggesting that Pdfr dsRNA had silenced the dfrA gene. The flowers contained sectors of white and light purple cells (Figure 4b), which appeared to be of clonal origin. This pattern is very different from the nonclonal patterns that were observed with posttranscriptionally silenced pigmentation
440 Current Biology Vol 11 No 6
genes (for example, Figure 1b) [20], and resembles the pattern observed in flowers, in which a pigmentationinducing transgene was transcriptionally silenced by position effects [21]. To confirm that the dfrA gene was transcriptionally silenced in the dsPdfr plants, quantitative RT-PCR analysis was performed to measure the accumulation of mature and precursor dfrA mRNAs. In the case of TGS, it was expected that the mature as well as the precursor mRNAs would be absent or reduced. This was indeed found for the dsPdfr plants (Figure 4c). In transformant PSEdfr-chs.1c, in which the chsA and dfrA genes are posttranscriptionally silenced [22], only the levels of the mature dfrA and chsA mRNAs were reduced (Figure 4c). Thus, the reduced pigmentation in the dsPdfr plants is because of transcriptional silencing of the dfrA gene. RNaseA/T1 protection assays showed that these transcriptionally silenced dsPdfr plants contain Pdfr dsRNA (for example, Figure 4d, lanes 2 and 3). RNA blot analysis on RNaseA-pretreated RNA revealed that the 35S-IR.Pdfr transgene is transcribed into full-length Pdfr dsRNA of approximately 1800 bp (data not shown). Also in dsPdfr plants, dsRNA-induced silencing is accompanied by the production of small RNAs as assessed by RNaseA/T1 protection assays of untreated cytoplasmic RNA (data not shown). DNA blot analysis using methylation-sensitive enzymes showed that the endogenous dfrA promoter in the silenced dsPdfr plants was more highly methylated than in wildtype plants (see Figure S1 in Supplementary material published with this article on the Internet). Also, the Pdfr transgene sequences were severely methylated. It is likely that the hypermethylation of the endogenous dfrA promoter is responsible for the TGS of the dfrA gene. These results indicate that dsRNA-induced TGS of endogenous genes offers an additional way to study gene function, in particular for the analysis of gene family members whose promoters may be more divergent than their coding regions.
4.
5. 6.
7. 8. 9. 10.
11.
12.
13.
14. 15. 16.
17.
18. 19.
Supplementary material Supplementary material including methodological details and a figure showing the increased methylation of the endogenous dfr promoter is available at http://current-biology.com/supmat/supmatin.htm.
20. 21.
Acknowledgements This work was funded by the European Union Biotechnology Program Framework IV, projects BIO4-CT97-2300 (T.S.) and BIO4-CT96-0253 (I.V. and D.R.), by The Netherlands Organization for the Advancement of Research (NWO-STW; R.v.B), and by an Erasmus scholarship to A.R.
References 1. Waterhouse PM, Graham HW, Wang MB: Virus resistance and gene silencing in plants can be induced by simultaneous expression of sense and antisense RNA. Proc Natl Acad Sci USA 1998, 95:13959-13964. 2. Chuang CF, Meyerowitz EM: Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. Proc Natl Acad Sci USA 2000, 97:4985-4990. 3. Smith NA, Singh SP, Wang MB, Stoutjesdijk PA, Graan AG,
22. 23.
Waterhouse PM: Total silencing by intron-spliced hairpin RNAs. Nature 2000, 407:319-320. Schweizer P, Pokorny J, Schulze-Lefert P, Dudler R: Technical advance: double-stranded RNA interferes with gene function at the single-cell level in cereals. Plant J 2000, 24:895903. Baulcombe DC, English JJ: Ectopic pairing of homologous DNA and post-transcriptional silencing in transgenic plants. Curr Opin Biotechnol 1996, 7:173-180. Mette MF, Aufsatz W, van Der Winden J, Matzke MA, Matzke AJ: Transcriptional silencing and promoter methylation triggered by double-stranded RNA. EMBO J 2000, 19:51945201. Bernstein E, Caudy AA, Hammond SM, Hannon GJ: Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 2001, 409:363-366. Zamore PD, Tuschl T, Sharp PA, Bartel DP: RNAi: double-stranded RNA directs the ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell 2000, 101:25-33. Elbashir SM, Lendeckel W, Tuschl T: RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev 2001, 15:188-200. Morel J, Mourrain P, Beclin C, Vaucheret H: DNA methylation and chromatin structure affect transcriptional and posttranscriptional transgene silencing in Arabidopsis. Curr Biol 2000, 10:1591-1594. van der Krol AR, Mur LA, Beld M, Mol JNM, Stuitje AR: Flavonoid genes in petunia: addition of a limited number of gene copies may lead to a suppression of gene expression. Plant Cell 1990, 2:291-299. Napoli C, Lemieux C, Jorgensen R: Introduction of a chimeric chalcone synthase gene into petunia results in reversible co-suppression of homologous genes in trans. Plant Cell 1990, 2:279-289. van Blokland R, van der Geest N, Mol JNM, Kooter JM: Transgenemediated suppression of chalcone synthase expression in Petunia hybrida results from an increase in RNA turnover. Plant J 1994, 6:861-877. Hamilton AJ, Baulcombe DC: A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 1999, 286:950-952. Hammond SM, Bernstein E, Beach D, Hannon GJ: An RNA-directed nuclease mediates post-transcriptional gene silencing in Drosophila cells. Nature 2000, 404:293-296. Dalmay T, Hamilton A, Rudd S, Angell S, Baulcombe DC: An RNAdependent RNA polymerase gene in Arabidopsis is required for posttranscriptional gene silencing mediated by a transgene but not by a virus. Cell 2000, 101:543-553. Mourrain P, Beclin C, Elmayan T, Feuerbach F, Godon C, Morel JB, et al.: Arabidopsis SGS2 and SGS3 genes are required for posttranscriptional gene silencing and natural virus resistance. Cell 2000, 101:533-542. Voinnet O, Lederer C, Baulcombe DC: A viral movement protein prevents spread of the gene silencing signal in Nicotiana benthamiana. Cell 2000, 103:157-167. Wassenegger M, Heimes S, Riedel L, Sanger HL: RNA-directed de novo methylation of genomic sequences in plants. Cell 1994, 76:567-576. Jorgensen RA: Cosuppression, flower color patterns, and metastable gene expression states. Science 1995, 268:686691. Meyer P, Heidmann I, Niedenhof I: Differences in DNAmethylation are associated with a paramutation phenomenon in transgenic petunia. Plant J 1993, 4:89-100. Stam M: Post-transcriptional silencing of flower pigmentation genes in Petunia hybrida by (trans)gene repeats. PhD thesis. Vrije Universiteit, Amsterdam, The Netherlands: 1997. Stam M, de Bruin R, Kenter S, van der Hoorn RAL, van Blokland R, Mol JNM, et al.: Post-transcriptional silencing of chalcone synthase in Petunia by inverted transgene repeats. Plant J 1997, 12:63-82.
