Paramutation of tobacco transgenes by small RNA-mediated transcriptional gene silencing.

It has been well established that trans-acting small RNAs guide promoter methylation leading to its inactivation and gene silencing at the transcriptional level (TGS). Here we addressed the question of the influence of the locus structure and epigenetic modifications of the target locus on its susceptibility for being paramutated by trans-acting small RNA molecules. Silencing was induced by crossing a 35S promoter silencer locus 271 with two different 35S-driven transgene loci, locus 2 containing a highly expressed single copy gene and locus 1 containing an inverted posttranscriptionally silenced (PTGS) repeat of this gene. Three generations of exposure to RNA signals from the 271 locus were required to complete silencing and methylation of the 35S promoter within locus 2. Segregating methylated locus 2 epialleles were obtained only from the third generation of hybrids, and this methylation was not correlated with silencing. Strikingly, only one generation was required for the PTGS locus 1 to acquire complete TGS and 35S promoter methylation. In this case, paramutated locus 1 epialleles bearing methylated and inactive 35S promoters segregated already from the first generation of hybrids. The results support the hypothesis that PTGS loci containing a palindrome structure and methylation in the coding region are more sensitive to paramutation by small RNAs and exhibit a strong tendency to formation of meiotically transmissible TGS epialleles. These features contrast with a non-methylated single copy transgenic locus that required several generations of contact with RNA silencing molecules to become imprinted in a stable epiallele.

Gene silencing induced by hairpin or inverted repeated sense transgenes varies among promoters and cell types.

*In transgenic calli and different tissues of Arabidopsis thaliana plants, the in trans silencing capacity of a 35S-beta-glucuronidase (GUS) hairpin RNA construct was investigated on a target GUS gene, under the control of the 35S, a WRKY or several cell cycle-specific promoters. *GUS histochemical staining patterns were analyzed in all tissues of the parental lines and supertransformants harboring the hairpin construct. Quantitative GUS activity measurements determined GUS suppression by a 35S-GUS hairpin or inverted repeated GUS transgenes in leaves and calli. *In some supertransformants, GUS-based staining disappeared in all tissues, including calli. In most supertransformants, however, a significant reduction was found in mature roots and leaves, but residual GUS activity was observed in the root tips, young leaves and calli. In leaves of most hairpin RNA supertransformants, the GUS activity was reduced by c. 1000-fold or more, but, in derived calli, generally by less than 200-fold. The silencing efficiency of inverted repeated sense transgenes was similar to that of a hairpin RNA construct in leaves, but weaker in calli. *These results imply that the tissue type, nature of the silencing inducer locus and the differential expression of the targeted gene codetermine the silencing efficiency.

High frequency of single-copy T-DNA transformants produced by floral dip in CRE-expressing Arabidopsis plants.

For genetic transformation of plants, floral dip with Agrobacterium often results in integration of multiple T-DNA copies at a single locus and frequently in low and unstable transgene expression. To obtain efficient single-copy T-DNA transformants, two CRE/loxP recombinase-based simplifying strategies for complex T-DNA loci were compared. A T-DNA vector with oppositely oriented loxP sites was transformed into CRE-expressing and wild-type control Arabidopsis thaliana plants. Of the primary CRE-expressing transformants, 55% harboured a single copy of the introduced T-DNA, but only 15% in the wild-type plants. However, 73% of the single-copy transformants in the CRE background showed continuous somatic inversion of the DNA segment between the two loxP sites. To avoid inversion of the loxP-flanked T-DNA segment, two T-DNA vectors harbouring only one loxP site were investigated for their suitability for CRE/loxP recombinase-mediated resolution upon floral-dip transformation into CRE-expressing plants. On average, 70% of the transformants in the CRE background were single-copy transformants, whereas the single-copy T-DNA frequency was only 11% for both vectors in the wild-type background. Both resolution strategies yielded mostly Cre transformants in which the 35S-driven transgene expression was stable and uniform in the progeny and remarkably, also in Cre transformants with multiple T-DNA copies. Therefore, a role is proposed for the CRE recombinase in preventing inverted T-DNA repeat formation or modifying the locus chromatin structure, resulting in a reduced sensitivity for silencing.

Trans-generation inheritance of methylation patterns in a tobacco transgene following a post-transcriptional silencing event.

We have studied the inheritance of the epigenetic state of tobacco transgenes whose expression was post-transcriptionally silenced by an invertedly repeated silencer locus. We show that, in hybrids, the coding region of the target neomycin phosphotransferase (nptII) gene was almost exclusively methylated at CG configurations, and dense non-CG methylation occurred in the 3' untranslated region. Homologous sequences in the silencer locus were heavily methylated at both CG and non-CG motifs. After segregation of the silencer locus, the CG methylation but not the non-CG methylation of the target genes was transmitted to the progeny. In the segregants, we observed an overall increase of CG methylation in the target genes, associated with a re-distribution from the 3' end of the coding region towards the middle. This pattern was inherited with some fluctuation for at least two additional generations in the absence of a detectable T-DNA-derived small RNA fraction. Thus CG methylation is not cleared during meiosis and may be inherited over generations without RNA signals being present. These epi-allelic variants re-expressed the reporter gene immediately after segregation of the trigger, showing that relatively dense CG methylation (approximately 60-80%) imprinted on most of the coding region (>500 bp) did not reduce expression compared with the parental non-methylated locus. We propose that the genic CG methylation seen in euchromatic regions of the genome may originate from ancient post-transcriptional gene silencing events as a result of adventitiously produced methylation-directing RNA molecules.

Development and application of novel constructs to score C:G-to-T:A transitions and homologous recombination in Arabidopsis.

We report on the development of five missense mutants and one recombination substrate of the beta-glucuronidase (GUS)-encoding gene of Escherichia coli and their use for detecting mutation and recombination events in transgenic Arabidopsis (Arabidopsis thaliana) plants by reactivation of GUS activity in clonal sectors. The missense mutants were designed to find C:G-to-T:A transitions in a symmetrical sequence context and are in that respect complementary to previously published GUS point mutants. Small peptide tags (hemagglutinin tag and Strep tag II) and green fluorescent protein were translationally fused to GUS, which offers possibilities to check for mutant GUS production levels. We show that spontaneous mutation and recombination events took place. Mutagenic treatment of the plants with ethyl methanesulfonate and ultraviolet-C increased the number of mutations, validating the use of these constructs to measure mutation and recombination frequencies in plants exposed to biotic or abiotic stress conditions, or in response to different genetic backgrounds. Plants were also subjected to heavy metals, methyl jasmonate, salicylic acid, and heat stress, for which no effect could be seen. Together with an ethyl methanesulfonate mutation induction level much higher than previously described, the need is illustrated for many available scoring systems in parallel. Because all GUS missense mutants were cloned in a bacterial expression vector, they can also be used to score mutation events in E. coli.

Evaluation of CRE-mediated excision approaches in Arabidopsis thaliana.

The ability of the CRE recombinase to catalyze excision of a DNA fragment flanked by directly repeated lox sites has been exploited to modify gene expression and proved to function well in particular case studies. However, very often variability in CRE expression and differences in efficiency of CRE-mediated recombination are observed. Here, various approaches were investigated to reproducibly obtain optimal CRE activity. CRE recombination was analyzed either by transforming the CRE T-DNA into plants containing a lox-flanked fragment or by transforming a T-DNA harboring a lox-flanked fragment into plants producing the CRE recombinase. Although somatic CRE-mediated excision of a lox-flanked fragment was obtained in all transformants, a variable amount of germline-transmitted deletions was found among different independent transformants, irrespective of the orientation of transformation. Also, the efficiency of CRE-mediated excision correlated well with the CRE mRNA level. In addition, CRE-mediated fragment excision was compared after floral dip and after root tissue transformation when transforming in a CRE-expressing background. Importantly, less CRE activity was needed to excise the lox-flanked fragment from the transferred T-DNA after root tissue transformation than after floral dip transformation. We hypothesize that this is correlated with the lower T-DNA copy number inserted during root transformation as compared to floral dip transformation.

Down-regulation of endogenes mediated by a transitive silencing signal.

Some RNA silencing systems in plants, nematodes, and fungi show spreading of silencing along target sequences, termed transitive silencing. Here, we address the question of whether endogenous targets can be silenced by a transitive silencing signal in plants. In transgenic Arabidopsis thaliana plants that harbored a silencing-inducing locus and a transgenic chimeric primary target, silencing of a secondary transgenic target occurred and the expression of the endogenous catalase genes was down-regulated, coinciding with a knock-down phenotype. Strikingly, the efficiency of the catalase silencing appeared to be correlated with the zygosity of the primary target locus and, to a lesser extent, with that of the silencing-inducing locus. These data suggest that silencing of an endogene induced by transgenic secondary small interfering RNAs (siRNAs) might depend on the amount of primary target transcripts that can act as template for the production of an efficient transitive silencing signal.

The frequency and efficiency of endogene suppression by transitive silencing signals is influenced by the length of sequence homology.

Transitivity, the spread of RNA silencing along primary target sequences, leads to the degradation of secondary targets that have no sequence homology to the initial silencing trigger. We demonstrate that increasing the distance between direct and adjacent target sequences in a transgenic primary target delays the onset of silencing of a secondary target gene. Silencing can spread in a 3' to 5' direction over a distance of at least 500 nucleotides (nt), but this requires consistently more time compared to a distance of 98 nt or 250 nt. The efficiency and frequency of transitive silencing of an endogene depends on the length of its sequence homology with the primary target. With a length of 500 nt, efficient silencing can eventually be established in all plants, whereas lengths of 250 nt and 98 nt homology result in less efficient and less frequent suppression. These results suggest that amplification of secondary small interfering RNAs (siRNAs) is a time-requiring process that gradually expands the population of siRNAs until a steady-state level is reached. Moreover, the length of the sequence homology in the primary target providing secondary siRNAs determines whether this steady-state level readily exceeds the threshold necessary for efficient silencing.

Insights into recognition of the T-DNA border repeats as termination sites for T-strand synthesis by Agrobacterium tumefaciens.

The recognition of the T-DNA left border (LB) repeat is affected by its surrounding sequences. Here, the LB regions were further characterized by molecular analysis of transgenic plants, obtained after Agrobacterium tumefaciens-mediated transformation with T-DNA vectors that had been modified in this LB region. At least the 24-bp LB repeat by itself was insufficient to terminate the T-strand synthesis. Addition of the natural inner and/or outer border regions to at least the LB repeat, even when present at a distance, enhanced the correct recognition of the LB repeat, reducing the number of plants containing vector backbone sequences. In tandem occurrence of both the octopine and nopaline LB regions with their repeats terminated the T-strand synthesis most efficiently at the LB, yielding a reproducibly high number of plants containing only the T-DNA. Furthermore, T-strand synthesis did not terminate efficiently at the right border (RB) repeat, which might indicate that signals in the outer RB region inhibit the termination of T-strand synthesis at the RB repeat.

The trans-silencing capacity of invertedly repeated transgenes depends on their epigenetic state in tobacco.

We studied the in trans-silencing capacities of a transgene locus that carried the neomycin phosphotransferase II reporter gene linked to the 35S promoter in an inverted repeat (IR). This transgene locus was originally posttranscriptionally silenced but switched to a transcriptionally silenced epiallele after in vitro tissue culture. Here, we show that both epialleles were strongly methylated in the coding region and IR center. However, by genomic sequencing, we found that the 1.0 kb region around the transcription start site was heavily methylated in symmetrical and non-symmetrical contexts in transcriptionally but not in posttranscriptionally silenced epilallele. Also, the posttranscriptionally silenced epiallele could trans-silence and trans-methylate homologous transgene loci irrespective of their genomic organization. We demonstrate that this in trans-silencing was accompanied by the production of small RNA molecules. On the other hand, the transcriptionally silenced variant could neither trans-silence nor trans-methylate homologous sequences, even after being in the same genetic background for generations and meiotic cycles. Interestingly, 5-aza-2-deoxy-cytidine-induced hypomethylation could partially restore signaling from the transcriptionally silenced epiallele. These results are consistent with the hypothesis that non-transcribed highly methylated IRs are poor silencers of homologous loci at non-allelic positions even across two generations and that transcription of the inverted sequences is essential for their trans-silencing potential.

The preferred route for the degradation of silencing target RNAs in transgenic plants depends on pre-established silencing conditions.

RNA silencing can be initiated upon dsRNA accumulation and results in homology-dependent degradation of target RNAs mediated by 21-23 nt small interfering RNAs (siRNAs). These small regulatory RNAs can direct RNA degradation via different routes such as the RdRP/Dicer- and the RNA-induced silencing complex (RISC)-catalysed pathways. The relative contribution of both pathways to degradation of target RNAs is not understood. To gain further insight in the process of target selection and degradation, we analysed production of siRNAs characteristic for Dicer-mediated RNA degradation during silencing of mRNAs and chimeric viral RNAs in protoplasts from plants of a transgenic tobacco silencing model line. We show that small RNA accumulation is limited to silencing target regions during steady-state mRNA silencing. For chimeric viral RNAs, siRNA production appears dependent on pre-established cellular silencing conditions. The observed siRNA accumulation profiles imply that silencing of viral target RNAs in pre-silenced protoplasts occurs mainly via a RISC-mediated pathway, guided by (pre-existing) siRNAs derived from cellular mRNAs. In cells that are not silenced at the time of infection, viral RNA degradation seems to involve Dicer action directly on the viral RNAs. This suggests that the silencing mechanism flexibly deploys different components of the RNA degradation machinery in function of the prevailing silencing status.

Spreading of post-transcriptional gene silencing along the target gene promotes systemic silencing.

Transitive silencing and grafting-induced gene silencing phenomena were combined to investigate whether a primary target beta-glucuronidase (gus) gene could promote the generation of systemic transitive silencing signals. Tobacco plants with hemizygous or homozygous silencer locus and in trans silenced primary target were used as a source of post-transcriptionally silenced rootstocks and tobacco plants with or without a secondary target locus as scion source. The silencer locus harbored two identical neomycin phosphotransferase II (nptII)-containing T-DNAs, integrated as an inverted repeat. The primary target locus carried a gus gene with homology to the transcribed region of the nptII gene only in the 3' untranslated region, whereas the secondary target locus had two or more copies of a gus gene without homology to transcribed sequences of the silencer locus. The upstream region of the initially targeted sequences of the in trans silenced gus gene could induce the production of a systemic signal. This signal was capable of triggering post-transcriptional gene silencing (PTGS) of the secondary target gus genes in the scion. In addition, the induction of systemic silencing was strikingly dosage dependent for the silencer as well as the primary target loci in the rootstock. Moreover, in the scions, the secondary target gus genes had to be present to generate detectable amounts of short interfering RNAs.

Epigenetic switch from posttranscriptional to transcriptional silencing is correlated with promoter hypermethylation.

Changes in the distribution of methylcytosine residues along a transgene locus of tobacco (Nicotiana tabacum) in relation to the type of gene silencing were studied in parental plant leaves, calli, and regenerated plants derived thereof. Parental-silenced HeLo1 (hemizygous for locus 1) plants show posttranscriptional silencing of the residing nptII (neomycin phosphotransferase II) transgene and cytosine methylation restricted to the 3' end and center part of the transcribed region. Here, we report that with an increasing number of cell cycles, DNA methylation changes gradually, and methylation is introduced into the promoter during cell culture and more slowly in vegetatively propagated plants. After 24 months of callus in vitro cultivation, an epigenetic variant, designated locus 1E, was obtained in which cytosine methylation of symmetrical (CG and CNG) sites was almost complete within the 5' end of the nptII-transcribed region and the 35S promoter. Further, methylation of nonsymmetrical sites appeared de novo in the promoter, whereas this type of methylation was significantly reduced in the 3' end of the transcribed region when compared with locus 1. The newly established epigenetic patterns were stably transmitted from calli into regenerated plants and their progeny. The protein and steady-state RNA levels remained low in locus 1E, whereas with nuclear run-on assays, no detectable amounts of primary transcripts were found along the nptII gene, indicating that the methylated promoter became inactivated. The results suggest that a switch between posttranscriptional and transcriptional gene silencing could be a mechanism leading to irrevocable shut down of gene expression within a finite number of generations.

RNA target sequences promote spreading of RNA silencing.

It is generally recognized that a silencing-inducing locus can efficiently reduce the expression of genes that give rise to transcripts partially homologous to those produced by the silencing-inducing locus (primary targets). Interestingly, the expression of genes that produce transcripts without homology to the silencing-inducing locus (secondary targets) can also be decreased dramatically via transitive RNA silencing. This phenomenon requires primary target RNAs that contain sequences homologous to secondary target RNAs. Sequences upstream from the region homologous to the silencing inducer in the primary target transcripts give rise to approximately 22-nucleotide small RNAs, coinciding with the region homologous to the secondary target. The presence of these small RNAs corresponds with reduced expression of the secondary target whose transcripts are not homologous to the silencing inducer. The data suggest that in transgenic plants, targets of RNA silencing are involved in the expansion of the pool of functional small interfering RNAs. Furthermore, methylation of target genes in sequences without homology to the initial silencing inducer indicates not only that RNA silencing can expand across target RNAs but also that methylation can spread along target genes.

An active role for endogenous beta-1,3-glucanase genes in transgene-mediated co-suppression in tobacco.

Post-transcriptional gene silencing (PTGS) is characterized by the accumulation of short interfering RNAs that are proposed to mediate sequence-specific degradation of cognate and secondary target mRNAs. In plants, it is unclear to what extent endogenous genes contribute to this process. Here, we address the role of the endogenous target genes in transgene-mediated PTGS of beta-1,3-glucanases in tobacco. We found that mRNA sequences of the endogenous glucanase glb gene with varying degrees of homology to the Nicotiana plumbaginifolia gn1 transgene are targeted by the silencing machinery, although less efficiently than corresponding transgene regions. Importantly, we show that endogene-specific nucleotides in the glb sequence provide specificity to the silencing process. Consistent with this finding, small sense and antisense 21- to 23-nucleotide RNAs homologous to the endogenous glb gene were detected. Combined, these data demonstrate that a co-suppressed endogenous glucan ase gene is involved in signal amplification and selection of homologous targets, and show that endogenous genes can actively participate in PTGS in plants. The findings are introduced as a further sophistication of the post-transciptional silencing model.