Contrasting epigenetic control of transgenes and endogenous genes promotes post-transcriptional transgene silencing in Arabidopsis

Butel, Nicolas; Yu, Agnès; Le Masson, Ivan; Borges, Filipe; Elmayan, Taline; Taochy, Christelle; Gursanscky, Nial R.; Cao, Jiangling; Bi, Shengnan; Sawyer, Anne; Carroll, Bernard J.; Vaucheret, Hervé

doi:10.1038/s41467-021-22995-3

Download PDF

Article
Open access
Published: 13 May 2021

Contrasting epigenetic control of transgenes and endogenous genes promotes post-transcriptional transgene silencing in Arabidopsis

Nature Communications volume 12, Article number: 2787 (2021) Cite this article

6790 Accesses
5 Citations
27 Altmetric
Metrics details

Subjects

Abstract

Transgenes that are stably expressed in plant genomes over many generations could be assumed to behave epigenetically the same as endogenous genes. Here, we report that whereas the histone H3K9me2 demethylase IBM1, but not the histone H3K4me3 demethylase JMJ14, counteracts DNA methylation of Arabidopsis endogenous genes, JMJ14, but not IBM1, counteracts DNA methylation of expressed transgenes. Additionally, JMJ14-mediated specific attenuation of transgene DNA methylation enhances the production of aberrant RNAs that readily induce systemic post-transcriptional transgene silencing (PTGS). Thus, the JMJ14 chromatin modifying complex maintains expressed transgenes in a probationary state of susceptibility to PTGS, suggesting that the host plant genome does not immediately accept expressed transgenes as being epigenetically the same as endogenous genes.

Ectopic targeting of CG DNA methylation in Arabidopsis with the bacterial SssI methyltransferase

Article Open access 25 May 2021

Wanlu Liu, Javier Gallego-Bartolomé, … Steven E. Jacobsen

DNA methylation-free Arabidopsis reveals crucial roles of DNA methylation in regulating gene expression and development

Article Open access 14 March 2022

Li He, Huan Huang, … Jian-Kang Zhu

RNA interference-independent reprogramming of DNA methylation in Arabidopsis

Article 30 November 2020

Taiko Kim To, Yuichiro Nishizawa, … Tetsuji Kakutani

Introduction

RNA-mediated gene silencing is a highly conserved eukaryotic mechanism that regulates endogenous gene expression and acts as a defense mechanism against viruses and transposons^1,2,3. It occurs when double-stranded RNA (dsRNA) is formed by transcription of an inverted repeat, the copying of single-stranded RNA (ssRNA) by viral and host RNA-dependent RNA polymerases (RDRs), or the simultaneous transcription of both the positive and negative strands of DNA. DICER-like (DCL) proteins convert the dsRNA into microRNAs (miRNAs) or small-interfering RNAs (siRNAs), which then guide ARGONAUTE-like (AGO) proteins to mediate sequence-specific gene silencing.

In plants, 24 nucleotide (nt) siRNAs guide transcriptional gene silencing (TGS) and RNA-directed DNA methylation (RdDM)^4,5,6. The process of de novo RdDM also requires the cytosine methyltransferase DRM2⁷. Following the establishment of DNA methylation and through subsequent rounds of DNA replication, cytosine methylation at CG and CHG sites can be maintained in the absence of siRNA by the cytosine methyltransferases MET1, and CMT2 and CMT3, respectively, whereas maintenance of CHH methylation requires the continual presence of 24 nt siRNAs and DRM2^7,8.

Abundant siRNAs are also derived from viral RNA and RNA Polymerase II (Pol II) transcripts that have been converted into dsRNA by viral or host RDRs. In plants, RDR6 plays the major role in producing these dsRNAs, which are processed by DCL4 and DCL2 into 21 nt and 22 nt siRNAs, respectively. These siRNAs form a complex with AGO1 and guide silencing of complementary RNA viruses and post-transcriptional gene silencing (PTGS) of complementary mRNAs⁹. Once PTGS is triggered, the biogenesis of 21 and 22 nt siRNAs usually extends along the entire length of the mRNA in a process known as transitivity^10,11,12. Furthermore, once induced, PTGS moves systemically throughout the plant^2,13,14.

It is well established that over-expression of aberrant mRNAs lacking a 5′ cap^15,16 or poly(A) tail¹⁷ induces RDR6-dependent PTGS in plants. The transcription of aberrant RNAs is a by-product of RNA Pol II transcription in all eukaryotes, but normally, aberrant RNAs are intercepted and removed by the highly conserved RNA Quality Control (RQC) pathway. In addition to removing aberrant RNAs, the RQC pathway modulates gene expression via mRNA turnover. The mechanism of mRNA turnover has been studied in much more detail than the fate of aberrantly transcribed RNA, and the first step in mRNA decay is the removal of the polyA tail, followed by the removal of the 5′ cap by the decapping complex, which is composed of DCP1, DCP2 and VARICOSE (VCS) in plants¹⁸. Once the stabilizing 5′ and 3′ modifications have been removed from the mRNA, the exposed RNA is degraded 5′-3′ by XRN exoribonucleases, and 3′-5′ by the exosome in all eukaryotes¹⁹. In Arabidopsis, dpc1, dcp2, vcs, xrn, and exosome mutations that compromise RQC allow aberrant RNA to accumulate and result in spontaneous RDR6-dependent PTGS of both transgenes^{15,16,20,21,22,23,24} and endogenous genes^21,25.

Mutants impaired in the histone H3K4me3 demethylase JMJ14 have been identified in two independent forward genetic screens for PTGS-deficient mutants^26,27. Molecular analysis of jmj14 plants compared to wild-type (WT) plants revealed a decrease in transgene siRNA accumulation, a decrease in transgene H3K4m3 and an increase in CHG methylation at the transgene promoter^26,28. Thusfar, DNA methylation has only been mechanistically implicated in TGS of promoters in most eukaryotic lineages⁷, inhibition of transcriptional elongation in fungi²⁹, and most recently, inhibition of aberrant transcription from the gene body, i.e., the intragenic regions, of plants^30,31 and mammalian embryonic stem cells³². Therefore, JMJ14 could link DNA methylation to aberrant RNAs and PTGS. However, the increase of transgene DNA methylation in jmj14 mutants contrasted the decreased methylation of CHH and CHG sites observed in two endogenous transposons in jmj14 mutants²⁷, and resembled more the effect of ibm1 mutations on endogenous sequences. IBM1 encodes a histone H3K9me2 demethylase, and ibm1 mutants have been reported to cause increased CHG and CHH methylation in the gene body of endogenous genes^33,34,35. Therefore, the effect of jmj14 and ibm1 mutations on the epigenetic features of transgenic versus endogenous sequences required further investigation.

In this work, we perform genome-wide DNA methylation analysis of a jmj14 mutant and show that very few endogenous loci exhibit a change in DNA methylation in the mutant compared to wild type, and the DNA methylation changes in the jmj14 mutant are restricted to genomic sequences that normally exhibit highly variable DNA methylation levels^36,37. In contrast, we detect increased methylation at all sequence contexts (CG, CHG, CHH) in transgenic sequences of the jmj14 mutant, but not the ibm1 mutant. We also show that JMJ14 promotes and IBM1 prevents transgene PTGS, respectively, and that the susceptibility of transgenes to PTGS correlates with low levels of DNA methylation in the transgene and high levels of transgene aberrant RNA, indicating that JMJ14 in combination with other chromatin modifying proteins plays a crucial role in establishing and maintaining stably expressed transgenes in an epigenetic state that is distinct from endogenous genes.

Results

A genetic screen for impaired systemic PTGS identifies additional jmj14 alleles

Using the transgenic line 10027-3³⁸, a genetic screen recovered three jmj14 mutants, #38, #90 and #148, that showed defects in root-to-shoot PTGS transmission (Fig. 1A). The 10027-3 GFP reporter system involves a constitutive p35S:GFP transgene linked to another transgene that drives root tip-specific expression of a GF hairpin RNA homologous to GFP³⁸. In 10027-3 WT plants, PTGS of GFP is initiated in the root apex during embryogenesis, and then as the seedlings germinate, silencing spreads into the shoot apex such that all true leaves that form show complete PTGS of GFP³⁸ (Fig. 1A). The systemic spreading of PTGS in this GFP reporter system requires RDR6-dependent amplification of dsRNA, using the constitutive GFP mRNA target as a template³⁸. The three 10027-3 jmj14 mutants were identified by candidate gene sequencing on a collection of EMS-induced 10027-3 mutants showing defects in systemic PTGS³⁹. Mutants #38 and #90 carry a nonsense and splice site mutation in JMJ14, respectively, and both mutants showed a lack of GFP silencing in the shoot apex but as the seedlings develop, PTGS of GFP spreads into the petiole and mid-vein of mature leaves (Fig. 1A). In contrast, mutant #148 contains a missense mutation in JMJ14 and showed a complete loss of GFP silencing (Fig. 1A).

**Fig. 1: Three new alleles of *jmj14* show defects in systemic PTGS of *GFP*.**

F1 plants from crosses between mutants #38 and #90 (Fig. 1B), and between mutants #148 and #38 (Fig. 1C) displayed a defect in systemic PTGS, thereby confirming that the jmj14 mutations were the causative genetic defect in these mutants. These three mutants further substantiated the role of JMJ14 in RDR6-dependent PTGS²⁶, and suggest the possible involvement of JMJ14 in systemic spreading of PTGS. Hereafter, the #38, #90 and #148 mutants will be referred to as jmj14-5, jmj14-6 and jmj14-7, respectively (Fig. 1D).

JMJ14 promotes aberrant RNA-based RDR6-dependent PTGS

The results presented above suggest that JMJ14 plays a role in the systemic spreading of PTGS. However, they do not exclude that JMJ14 also plays a role in the execution of PTGS. To resolve these two possibilities, the two-component GUS-silencing system 6b4-306 was used. This transgenic line consists of the p35S:GUS transgene locus called 6b4, which never triggers PTGS alone in WT plants, plus an unlinked p35S:hpGU hairpin transgene locus called 306, which produces dsRNA and siRNAs homologous to the first half of the GUS coding sequence⁴⁰. In contrast to the 10027-3 GFP reporter system, both the 6b4 GUS transgene and 306 the hairpin transgene are expressed constitutively, resulting in the execution of PTGS in every cell, and therefore does not require RDR6-dependent spreading of PTGS to silence the GUS transgene throughout the plant⁴⁰. Crossing the 6b4-306 double transgenic line to the jmj14-4 mutant (P789S; Fig. 1D) produced triple homozygous 6b4-306 jmj14 plants in which PTGS occurred as efficiently as in 6b4-306 WT controls (Fig. 2A). Thus, JMJ14 is specifically required for RDR6-dependent PTGS, and not for the execution of PTGS when siRNAs are derived from a constitutively expressed hairpin transgene.

**Fig. 2: JMJ14 promotes aberrant RNA-induced PTGS, but is not required for PTGS induced by hairpin dsRNA.**

We next investigated the possibility that JMJ14 plays a role in the initiation of RDR6-dependent PTGS by enhancing the production of aberrant RNA from transgene loci. To address this question, we used the 6b4 transgenic line that does not carry any additional T-DNA loci. As mentioned above, the 6b4 locus carries a p35S:GUS transgene that never shows spontaneous PTGS in WT plants, but the introduction of 6b4 in various RQC-deficient mutant backgrounds results in the spontaneous triggering of PTGS^16,21,22,24. This result is explained by the 6b4 GUS transgene producing low amounts of aberrant RNAs that are efficiently degraded by the cellular RQC pathways in WT plants. However, when RQC is impaired, aberrant RNAs derived from the 6b4 GUS transgene are converted into dsRNA by RDR6 and processed into 21- and 22 nt siRNAs by DCL4 and DCL2, respectively, thus activating PTGS. To determine if JMJ14 was contributing to the production of aberrant RNA from the 6b4 GUS transgene, we utilized the mRNA decapping-defective vcs-9 mutant to generate a 6b4 vcs jmj14 double mutant, which was compared to 6b4 WT, 6b4 jmj14 and 6b4 vcs plants for the level of PTGS. The hypomorphic vcs-9 allele was used because vcs-9 mutants are viable and fertile, and because spontaneous PTGS occurs with 100% efficiency in 6b4 vcs-9 plants²¹. We observed an increase in GUS activity and a reduction in GUS siRNA accumulation in 6b4 vcs jmj14 compared to 6b4 vcs plants (Fig. 2B). Given that JMJ14 is not involved in the production of siRNA derived from constitutively expressed hairpin RNA, these results strongly suggest that JMJ14 promotes PTGS by enhancing the production of aberrant RNA from the 6b4 GUS transgene.

JMJ14 impairment decreases the production of GUS aberrant RNAs

We previously identified an uncapped RNA, antisense to the GUS mRNA, hereafter referred to as aberrant SUG (abSUG) RNA¹¹. Although it is not certain that this abSUG RNA is the aberrant RNA that triggers PTGS, its abundance correlates perfectly with the efficiency of PTGS in the two p35S:GUS reference lines 6b4 and L1. Indeed, it is detected at very low levels in 6b4 plants, which do not trigger PTGS spontaneously. It is also detected at only low levels in 6b4 xrn3 xrn4 and L1 plants, which trigger PTGS spontaneously with 100% efficiency, most likely because the abSUG RNA is degraded by PTGS, similar to GUS mRNA. However, abSUG RNA accumulates to higher levels in 6b4 xrn3 xrn4 rdr6 and L1 rdr6 plants, but remains low in 6b4 rdr6 plants. Given that the p35S:GUS transgene at the L1 locus is transcribed at a higher level than its identical transgene counterpart at the 6b4 locus, these results suggest that the production of abSUG at a high level in L1 plants or the absence of its degradation in 6b4 xrn3 xrn4 plants could explain the capacity of these plants to trigger PTGS¹¹.

To further test the hypothesis that jmj14 mutations limits the production of aberrant RNAs, the abundance of abSUG RNA was measured in L1, L1 rdr6 and L1 jmj14 rdr6 plants. To reliably assess the effect of JMJ14 on the levels of the abSUG RNA, the rdr6 mutant background was used to avoid RDR6-dependent PTGS of the abSUG RNA. The amount of abSUG RNA was very low in L1 WT plants compared to L1 rdr6 plants, thereby confirming that abSUG RNA is degraded by PTGS in L1 WT plants (Fig. 3). However, the amount of abSUG RNA in the L1 rdr6 mutant was at least double the amount in the L1 jmj14 rdr6 double mutant (Fig. 3), which further supports the hypothesis that JMJ14 promotes transgene-derived aberrant RNA production.

**Fig. 3: Impairing JMJ14 decreases the steady state level of aberrant *abSUG* RNA.**

A synergistic relationship between systemic PTGS signals and aberrant RNA in the induction of RDR6-dependent PTGS

The results described above strongly suggest that JMJ14 promotes the production of aberrant RNA from transgene loci, and that JMJ14 also plays a role in systemic PTGS. To investigate the possibility of a collaborative relationship between aberrant RNA produced from transgene loci and systemic PTGS signals in the induction of RDR6-dependent PTGS, we conducted reciprocal grafting experiments using WT and jmj14 mutants as rootstocks and scions in our GFP and GUS reporter systems for graft-transmissible PTGS^13,38.

When GFP-expressing 10027-3 jmj14-7 scions were grafted onto GFP-silenced 10027-3 WT rootstocks, graft-transmissible PTGS failed to be initiated (Fig. 4A), suggesting that the jmj14-7 mutation completely abolished the capacity of the scion to respond to the systemic PTGS signal transmitted from the 10027-3 WT roots. A feature of PTGS in plants is the predominance of 21 nt siRNAs produced by DCL4, along with a much lower abundance of 22 nt siRNA produced by DCL2. Nevertheless, DCL2 and its 22 nt siRNA play a more important role in systemic PTGS than DCL4^38,41. Therefore, we also grafted 10027-3 jmj14-7 scions onto 10027-3 dcl4-5 rootstocks that produce predominantly DCL2-dependent 22 nt siRNAs³⁸, and again, graft-transmissible PTGS failed to be initiated in the jmj14-7 scions. These results suggest that jmj14 scions are incapable of responding to either DCL4-dependent 21 nt or DCL2-dependent 22 nt siRNAs as mobile silencing signals.

**Fig. 4: JMJ14 is required for reception of graft-transmissible PTGS in scions and efficient transmission of PTGS from rootstocks.**

We also grafted GFP-expressing scions of transgenic line 214 (Supplementary Fig. 1) onto 10027-3 or 10027-3 jmj14 mutant lines as rootstocks¹³. Whereas GFP PTGS is efficiently transmitted from 10027-3 rootstocks to 214 scions, we observed a decrease in systemic transmission of PTGS when 214 scions were grafted onto 10027-3 jmj14 rootstocks (Fig. 4A).

These findings were confirmed using the jmj14-4 mutant allele in the 6b4 GUS reporter background. Whereas grafting 6b4 WT scions grafted onto L1 WT rootstocks triggered efficient GUS PTGS in 6b4 scions, 6b4 jmj14-4 scions grafted onto L1 WT rootstocks did not trigger PTGS (Fig. 4B, C). In addition, 6b4 jmj14-4 scions also failed to initiate PTGS when grafted onto 6b4-306 rootstocks (Fig. 4C). Furthermore, similar to what was observed for jmj14-5 and jmj14-6 rootstocks using GFP as the reporter for PTGS (Fig. 4A), when GUS-expressing 6b4 WT scions were grafted onto L1 jmj14-4 rootstocks, the onset of graft-transmissible PTGS in the scions was delayed (Fig. 4B). Altogether, these results indicate that JMJ14 plays an essential role in the perception of the PTGS systemic signal.

Transgene DNA methylation counteracts aberrant RNA-induced PTGS

Southern blot analysis using a methylation-sensitive enzyme previously revealed that L1 jmj14-3, L1 jmj14-4, 6b4 jmj14-3, and 6b4 jmj14-4 plants exhibit increased DNA methylation at CHG sites in the promoter of the p35S:GUS transgene²⁶. To determine the methylation profile of the entire transgene locus in a jmj14 mutant background, whole-genome bisulfite sequencing (WGBS) analysis was performed on plants carrying the 6b4 locus. WGBS revealed that the transgenic line 6b4 carried a single T-DNA insertion. However, the left border did not delineate the 6b4 T-DNA insertion and the entire binary vector was also found to have inserted adjacent to the left border (Supplementary Fig. 2). This phenomenon of co-transfer of non-T-DNA vector sequences occurs frequently with T-DNA integration events into plant chromosomes⁴². WGBS analysis revealed a high level of CG methylation in the GUS coding sequence of 6b4 plants (Fig. 5), which most likely reflects the highly transcribed status of the p35S:GUS transgene in this transgenic line. Of note, almost no CG methylation was found in the adjacent NPTII transgene (Supplementary Fig. 3), consistent with a low level of NPTII expression and the 6b4 line being barely resistant to kanamycin¹¹. However, a surprisingly high level of CG methylation was observed in the non-T-DNA vector sequences inserted along with the T-DNA (Supplementary Fig. 3), suggesting that these sequences had been transcribed by a plant RNA polymerase prior to, and/or after integration into the chromosome. In contrast to CG methylation, the level of CHH and CHG methylation was very low in both T-DNA and vector sequences of the 6b4 locus in the WT genetic background (Fig. 5 and Supplementary Fig. 3). The situation was very different in 6b4 jmj14-4 plants where the GUS coding sequence exhibited CHG hypermethylation and the 35S promoter showed increased CHG and CHH methylation (Fig. 5). Of note, CHH and CHG hypermethylation in 6b4 jmj14 was not observed in the adjacent NPTII transgene but was observed in the vector sequences flanking the left border (Supplementary Fig. 3A), suggesting that CHH and CHG hypermethylation in jmj14 occurs mostly at sequences that show high levels of CG methylation in the WT genetic background.

**Fig. 5: JMJ14 and NAC50/NAC52 prevent CHG and CHH methylation of the *p35S:GUS* transgene.**

These results prompted us to hypothesize that CHH and CHG hypermethylation of the p35S:GUS transgene in 6b4 jmj14 plants could prevent, or at least reduce, the production of aberrant RNAs (Fig. 3), thus explaining why jmj14 scions cannot trigger PTGS upon grafting onto silenced rootstocks (Fig. 4). To test this hypothesis, 6b4 jmj14 drm2 cmt3 plants were generated, and we confirmed the absence of CHH and CHG methylation in the p35S:GUS transgene and the entire 6b4 locus in this line (Fig. 5 and Supplementary Fig. 3). When grafted onto GUS-silenced rootstocks, 6b4 jmj14 drm2 cmt3 plants triggered PTGS as efficiently as grafted 6b4 WT controls (Table 1), supporting the hypothesis that CHH and CHG hypermethylation is instrumental in preventing systemic PTGS in 6b4 jmj14 plants.

Table 1 Frequency of spontaneous and graft-induced PTGS.

Full size table

There is a low but detectable level of CHG methylation (but not CHH methylation) in the GUS coding sequence of 6b4 WT plants (Fig. 5), and we asked if this residual level of CHG methylation prevents spontaneous induction of PTGS in these plants. To test this hypothesis, 6b4 drm2 cmt3 plants were generated. Our analysis revealed that CHG methylation is reduced in the GUS coding sequence of 6b4 drm2 cmt3 plants compared to 6b4 WT plants (Fig. 5A), and indeed, ~10% of these plants trigger spontaneous PTGS (Table 1). These results strongly support the hypothesis that transgene CHG methylation strongly limits the production of aberrant RNAs that would otherwise induce spontaneous PTGS of the homologous GUS coding sequence. In 6b4 drm2 cmt3 plants, we propose that the lower level of CHG methylation of the GUS coding sequence increases the transcription of aberrant RNAs, thereby allowing spontaneous PTGS to occur. While 6b4 WT plants produce insufficient aberrant RNAs to trigger spontaneous PTGS of the GUS coding sequence, it is sufficient for graft-induced PTGS to occur due to a synergistic relationship between aberrant RNA expressed from the GUS transgene of 6b4 WT scions and systemic PTGS signals derived from GUS-silenced rootstocks. In contrast, in 6b4 jmj14 scions, increased CHG methylation further limits the production of aberrant RNA from the 6b4 GUS transgene, rendering it non-responsive to graft-transmissible PTGS signals from silenced GUS rootstocks.

Remarkably, WGBS of 6b4 WT and 6b4 jmj14-4 plants revealed that loss of JMJ14 had a limited effect on the extent of DNA methylation at endogenous loci (Fig. 6). A summary of all endogenous DMRs identified between 6b4 WT and 6b4 jmj14-4 is presented in Supplementary Data 1. Only one endogenous hyper CHG DMR was identified in jmj14, and this endogenous region did not show hyper CG and hyper CHH methylation, whereas the transgenic sequences (both the T-DNA and the co-integrated backbone sequences of the binary vector) exhibited hyper CG, hyper CHG and hyper CHH methylation (Fig. 5 and Supplementary Fig. 3).

**Fig. 6: JMJ14 and NAC50/NAC52 do not affect CHG and CHH methylation of endogenous genes and TEs.**

Both JMJ14 and NAC52 interact with the p35S:GUS transgene and are required for systemic PTGS

Mutations in JMJ14 were previously shown to adversely affect Pol II occupancy and H3K4me3 levels in the promoter of the p35S:GUS transgene in transgenic line 6b4²⁶. However, it was not determined if this effect was due to a direct interaction between JMJ14 and chromatin at the 35S promoter. To address this question, chromatin immunoprecipitation (ChIP) of JMJ14 was performed on the 6b4 jmj14-4 line complemented with a pJMJ14:3xFlag-JMJ14 transgene⁴³. ChIP was conducted using Flag antibodies and followed by quantitative PCR (qPCR) using primer pairs located in the 35S promoter, and in the 5′ and 3′ regions within the GUS transgene body. Quantitative ChIP-PCR revealed a significant enrichment of JMJ14 primarily at the 35S promoter (Fig. 7A). These results indicate that JMJ14 is a component of chromatin at the promoter of the p35S:GUS transgene, and are consistent with Pol II occupancy, high levels of H3K4me3²⁶ and lack of DNA methylation (Fig. 5) within the 35S promoter of 6b4 WT plants.

**Fig. 7: Hierarchical binding of JMJ14 and NAC52 to the promoter and coding sequence of the *p35S:GUS* transgene, respectively.**

JMJ14 has also been shown to interact with the NAC-domain transcription factor NAC52 and its close relative NAC50^44,45. NAC50 and NAC52 most likely play redundant roles, but NAC52 appears more important than NAC50 in Arabidopsis due to its higher level of expression. Indeed, a genetic screen for PTGS impaired mutants identified a mutation, originally named sgs1, which turned out to be a mutant allele of NAC52, whereas reverse genetics showed that a nac50 knockout mutation had no effect on PTGS²⁸. To determine whether NAC50 or NAC52 participates in graft-induced PTGS of the 6b4 GUS transgene, a 6b4 nac50 nac52 double mutant was generated and grafted onto GUS-silenced rootstocks. Similar to 6b4 jmj14, the 6b4 nac50 nac52 scions were incapable of triggering PTGS of GUS when grafted onto GUS-silenced rootstocks (Table 1), indicating that both JMJ14 and NAC50/NAC52 are required for reception of graft-induced PTGS of the 6b4 GUS transgene. In addition, WGBS analysis of non-grafted 6b4 nac50 nac52 plants revealed transgene CHH and CHG hypermethylation similar to non-grafted 6b4 jmj14 plants (Fig. 5B and Supplementary Fig. 3B), confirming the correlation between CHH and CHG hypermethylation and decreased induction of PTGS.

Importantly, WGBS of 6b4 WT and 6b4 nac50 nac52 plants also confirmed that like the loss of JMJ14, loss of both NAC50 and NAC52 had a limited effect on the extent of DNA methylation at endogenous loci. Indeed, only one endogenous hyper CHG DMR was identified in jmj14 and only one endogenous hyper CHG DMR was identified in nac50 nac52 (Supplementary Data 1). These two DMRs were located on chromosome 5 and chromosome 3, respectively, indicating that jmj14- and nac50 nac52-dependent hyper CHG methylation is only found in the transgenic sequences (both the T-DNA and the co-integrated backbone sequences of the binary vector). Moreover, transgenic sequences not only exhibited hyper CHG methylation but also hyper CG and hyper CHH methylation (Fig. 5 and Supplementary Fig. 3). The only endogenous hyper CHG methylation found in jmj14 did not show hyper CG and hyper CHH methylation, while the only endogenous hyper CHG methylation found in nac50 nac52 showed hyper CHH methylation but not hyper CG methylation (Supplementary Data 1). Therefore, it is most likely that the endogenous DMRs observed between wild type and either jmj14 or nac50 nac52 correspond to endogenous genomic sequences that exhibit highly variable methylation levels^36,37. Together, WGBS analysis of WT, jmj14 and nac50 nac52 plants indicates that JMJ14 and NAC50/NAC52 specifically limit the level of DNA methylation of transgenes and have essentially no effect on the DNA methylation status of endogenous genes.

To further address the role of NAC50/NAC52 in PTGS, a pUBQ10:NAC52-GFP transgene was generated and introduced into the 6b4 WT line. The pUBQ10 promoter drives constitutive expression of transgenes in Arabidopsis. At first, we tested whether the pUBQ10:NAC52-GFP constructs expressed a functional protein that actually interacts with JMJ14. To do so, co-immunoprecipitation (co-IP) experiments were performed on plants resulting from a cross between a pJMJ14:JMJ14-Myc line⁴⁶ and the 6b4 + pUBQ10:NAC52-GFP line. As expected, JMJ14-Myc and NAC52-GFP proteins co-precipitated in pull-down assays using a GFP antibody (Supplementary Fig. 4). Furthermore, ChIP was conducted on the 6b4 + pUBQ10:NAC52-GFP line using GFP antibodies, followed by qPCR using primer pairs located in the 35S promoter, and in the 5′ and 3′ regions within the GUS transgene body. The ChIP-qPCR analysis revealed a significant enrichment of NAC52 in the 35S promoter and the 5′ region of the GUS coding sequence (Fig. 7B).

Finally, to address the hierarchical action of JMJ14 and NAC50/NAC52 in PTGS, we generated 6b4 jmj14 nac52 + pJMJ14:3xFlag-JMJ14⁴³ and 6b4 jmj14 + pUBQ10:NAC52-GFP lines. ChIP-qPCR revealed that the association of JMJ14 with chromatin at the 35S:GUS transgene was not affected by the nac52(sgs1) mutation (Fig. 7A), whereas the association of NAC52 was impaired by the jmj14 mutation (Fig. 7B). Altogether, these results strongly suggest that JMJ14 binds to chromatin at the promoter of the p35S:GUS transgene, which promotes the binding of NAC50/NAC52 to the 35S promoter and the downstream GUS coding sequence. Our results are consistent with JMJ14 and NAC50/NAC52 acting in a chromatin complex to prevent CHG methylation throughout the entire transgene, thus allowing aberrant RNAs to be produced, which in turn induce PTGS.

IBM1 impairment slightly decreases CHH and CHG methylation in the transgene body and promotes PTGS

To test further the hypothesis that transgene methylation influences PTGS capacity, the 6b4 locus was introduced into the ibm1 mutant background. IBM1 encodes an H3K9me2 demethylase that suppresses CHH and CHG methylation in the body of about half of the endogenous genes of Arabidopsis^33,34,35,47. We therefore expected 6b4 ibm1 plants to exhibit increased CHH and CHG methylation in the GUS coding sequence and suppress induction of PTGS when grafted onto silenced GUS rootstocks. However, while WGBS analysis confirmed increased CHH and CHG methylation in the body of 9000 endogenous genes (Fig. 6 and Supplementary Data 1), CHH and CHG methylation remained the same or slightly decreased in transgenic sequences in 6b4 ibm1 plants (Fig. 5 and Supplementary Fig. 3B). The decrease in CHG methylation in transgenic sequences of 6b4 ibm1 plants was most obvious in the GUS coding sequence (Fig. 5B). These results confirmed that actively expressed transgenes and endogenous genes exhibit major epigenetic differences. Moreover, spontaneous triggering of GUS PTGS occured in ~20% of 6b4 ibm1 plants (Table 1 and Supplementary Fig. 5), indicating that JMJ14 and IBM1 have opposite effects on transgene PTGS.

Discussion

Endogenous protein-coding genes produce aberrant RNAs that are efficiently degraded by RNA quality control (RQC) pathways. In plants, this is essential to prevent the conversion of aberrant RNAs into dsRNA by RDRs, which would result in the production of siRNAs that could destroy homologous and essential mRNAs. Indeed, RQC impairment causes lethality in Arabidopsis, which can be rescued by suppressing RDR6 activity^21,25. Similarly, transgenes produce aberrant RNAs in proportion to the level of transcript, and provided the aberrant RNAs do not exceed an abundance threshold within the cell, they are also eliminated by RQC. However, if RQC is impaired or saturated by high levels of transcription, both transgenes and endogenous genes can be subjected to RDR6-dependent PTGS^15,16. Occasionally, endogenous loci in plants can be subjected to PTGS despite a functioning RQC pathway being present. However, these loci generally exhibit genomic rearrangements involving gene duplication events that allow the production of dsRNA and induction of PTGS^{48,49,50,51,52,53,54}. It is assumed that such rearrangements involving endogenous genes are tolerated because they only adversely affect dispensable genes. Obviously, transgenes are also dispensable because they correspond to sequences that do not exist naturally in plants, and the transgene loci that induce PTGS the most efficiently involve strong promoters and/or tandem insertion of multiple copies of the transgene, often in inverted orientations⁵⁵. High rates of transcription from multiple transgenes increases the chances of aberrant RNA by-products forming, and only a fraction of this aberrant RNA needs to escape degradation by RQC to be recruited by RDR6, and converted into dsRNA to induce PTGS.

Despite these similarities, actively expressed transgenes are generally more susceptible to PTGS than endogenous genes. These differences between actively expressed transgenes and endogenous genes have long remained a mystery. Our finding that transgenes and endogenous genes exhibit different epigenetic characteristics is particularly interesting because transgenes are transferred into the host genome as naked DNA or naked DNA associated with Agrobacterium proteins, and once integrated into the chromosome, must associate with histones to form chromatin. At present, the type of eukaryotic histones that associate with naked or newly integrated foreign DNA is not known. One could imagine that the insertion of extrachromosomal DNA into the chromosome attracts particular histone marks to label this DNA as ‘new or foreign’. It seems likely that what happens for transgenes also occurs when a new copy of a retrotransposon integrates into the genome. Supporting this hypothesis, it has been shown that activation of retrotransposons triggers a PTGS response, just like transgenes do^56,57,58. Although this hypothesis is attractive, our findings suggest that there are striking epigenetic differences between transgenes and endogenous loci, including genes and stabilized transposons. Here, we showed that transgenes and endogenous loci (genes and stabilized transposons) behave differently when the histone H3K4me3 demethylase JMJ14, and two cooperating transcription factors that interact with JMJ14, NAC50, and NAC52, are impaired. Indeed, in jmj14 or nac50 nac52 mutants, an increase in CHG methylation, and CHH methylation to a lesser extent, is observed at the 6b4 transgene locus (Fig. 5 and Supplementary Fig. 3). An increase in CHG methylation was particularly evident in the promoter and gene body of the p35S:GUS transgene, and also parts of the integrated Agrobacterium vector sequences (Fig. 5 and Supplementary Fig. 3), whereas no change in DNA methylation profiles was observed at any endogenous genomic loci in these mutants (Fig. 6 and Supplementary Data 1). Remarkably, the jmj14 and nac50nac52 mutations also suppress the triggering of PTGS of the p35S:GUS and p35S:GFP when these transgenic mutant lines were used as scions grafted onto rootstocks showing PTGS (Fig. 4 and Table 1). Moreover, CHH and CHG hypermethylation, and not simply the absence of JMJ14 or NAC50/NAC52, appears causal in preventing PTGS because 6b4 jmj14 drm2 cmt3 scions trigger PTGS when grafted onto silenced rootstocks (Table 1), indicating that the impairment of DRM2-CMT3-dependent CHH and CHG methylation in a jmj14 background restores the susceptibility of transgenes to graft-transmissible PTGS. As JMJ14 and NAC52 bind to the p35S:GUS transgene within the 6b4 locus (Fig. 7), we propose that the interacting complex limits CHG and CHH methylation of the GUS coding sequence, allowing the production of aberrant RNAs that are required to trigger and maintain PTGS. Supporting this hypothesis, the 6b4 drm2 cmt3 transgenic line lacks CHH and CHG methylation of the GUS coding sequence and triggers PTGS spontaneously, whereas the 6b4 WT line exhibits a low CHH and CHG methylation level of the GUS coding sequence that is nevertheless sufficient to block the spontaneous triggering of PTGS.

The 6b4 T-DNA insertion locus contains not only an intact T-DNA but also the entire Agrobacterium binary vector sequence alongside the left border of the T-DNA (Supplementary Fig. 2). Remarkably, the CHG and CHH methylation profiles of the vector sequences were similar to the p35S:GUS transgene in WT plants and increased in the jmj14 mutant in a DRM2-CMT3-dependent manner (Supplementary Fig. 3). These results suggest that not just promoter-driven transgenes, but also promoterless foreign DNA sequences could be subject to JMJ14-dependent suppression of CHG and CHH methylation, resulting in the production of aberrant RNA and potential PTGS of homologous mRNA. It is interesting to note that bombardment of transgenic tobacco expressing a p35S:GFP transgene with promoterless GFP DNA sequences can induce PTGS⁵⁹. Adding to this earlier report, our data are consistent with promoter-independent aberrant RNA transcripts being produced from foreign DNA sequences prior to or immediately after integration into the plant chromosome. In contrast to the p35S:GUS transgene and Agrobacterium vector sequences, CHG and CHH methylation of the pNOS:NPTII:tNOS transgene within the 6b4 locus remained low in the jmj14 and nac50 nac52 mutant backgrounds (Supplementary Fig. 3). This could be explained by JMJ14 not being able to bind to the Agrobacterium NOS promoter, which may have been selected for during the evolution of Agrobacterium as a pathogen to ensure that the transferred oncogenes on the T-DNA were not epigenetically distinguishable from endogenous plant genes.

H3K4 trimethylation and DNA methylation are considered to have antagonistic influences activating and repressing expression of endogenous genes, respectively⁴⁶. Therefore, it was expected that for endogenous loci in jmj14 mutants, where H3K4me3 levels increase, DNA methylation would decrease. However, our WGBS analysis revealed that there were minimal changes in DNA methylation levels within endogenous genes of jmj14 (Fig. 6), further emphasizing the difference between transgenes and endogenous sequences. Nevertheless, we asked whether the binding of JMJ14 could promote the production of aberrant RNAs from endogenous genes and their subsequent transformation into siRNAs, like it does at the p35S:GUS transgene of the 6b4 locus. If this was the case, one would expect that the fraction of endogenous genes that produce siRNAs when RQC is impaired^21,25,60 would be enriched for genes that bind JMJ14⁴⁵. However, the fraction of siRNA-producing endogenous genes that bind JMJ14 is exactly that expected by chance (Supplementary Fig. 6), whether considering the siRNA-producing endogenous genes identified in the decapping mutants dcp2 and vcs²¹, the siRNA-producing endogenous genes identified in the exonuclease double mutant xrn4 ski2²⁵, or the siRNA-producing endogenous genes identified when plants are infected by viruses that inhibit RQC⁶⁰. Therefore, unlike the p35S:GUS transgene of the 6b4 locus, which binds JMJ14, endogenous genes that bind JMJ14 do not appear particularly prone to producing endogenous siRNAs when RQC is compromised. Altogether with our results, these results reinforce the idea that transgene DNA hypermethylation caused by the absence of JMJ14 is specific to transgenes and not endogenous genes.

We not only found that transgenes and endogenous genes exhibit different epigenetic behavior with regard to the H3K4me3 demethylase JMJ14, but also to the histone H3K9me2 demethylase IBM1^33,34,35,47. Indeed, in jmj14 mutant plants, an increase in DNA methylation is observed in the newly inserted sequences of the 6b4 locus but not in endogenous genes, whereas in ibm1 mutant plants, an increase in DNA methylation is observed in endogenous genes but not in the newly inserted sequences of the 6b4 locus (Figs. 5 and 6 and Supplementary Fig. 3). Moreover, whereas 6b4 plants do not trigger PTGS spontaneously but undergo systemic PTGS when grafted on silenced rootstocks, 6b4 jmj14 plants do not trigger PTGS spontaneously and are incapable of undergoing systemic PTGS, while 6b4 ibm1 plants trigger PTGS spontaneously, similar to 6b4 drm2 cmt3 plants (Table 1 and Supplementary Fig. 5), indicating that JMJ14 and IBM1 have opposite effects on transgene PTGS. It may be assumed that actively expressed transgenes behave epigenetically the same as endogenous genes. However, our data suggest that despite being stably expressed in a WT background, transgenes exhibit epigenetic features that strongly differs from that of endogenous genes with regard to JMJ14 and IBM1 regulation.

To conclude, our research shows that JMJ14 maintains expressed transgenes in an epigenetic state that allows a residual level of aberrant RNA to be produced from the transgene body, thereby rendering the transgene particularly susceptible to PTGS. This epigenetic phenomenon could therefore represent an evolutionary probation period for the expressed transgene, until it is either completely silenced, or alternatively, epigenetically accepted as an endogenous locus in the new host genome.

Methods

Plant material

The transgenic reporter lines 10027-3, 214, L1, 6b4 and 6b4-306, and the mutants jmj14-4, nac52^sgs1, nac50 nac52, rdr6^sgs2-1, cmt3-7, drm2-3, ibm1-1, vcs-9 all are in the Arabidopsis accession Columbia^{13,21,26,38,40,44,47,61,62}. The transgenic lines carrying the tagged constructs pJMJ14:JMJ14-Myc, pJMJ14:Flag-JMJ14 and pUBQ10:NAC52-GFP also are in the Arabidopsis accession Columbia^28,43,46. Primers used for genotyping are listed in Supplementary Table 1.

Growth conditions and grafting techniques

Arabidopsis seeds were surface-sterilized, sown on a nutritive medium (1.3% S-medium Duchefa, 1% Phytoblend agar), vernalized at 4 °C for 2 days, and then placed in a culture chamber at 23 °C, 70% humidity, 120 µE m⁻² light with a 16 h light/8 h dark (long-days) or 8 h light/16 h dark (short-days) photoperiod. Seedlings grown under long-day conditions were transferred to soil after 2 weeks. Seedlings grown under short-day conditions were used for grafting expertiments as described in ref. ⁶³. Briefly, 6 days after gemination, seedlings were cut transversely across the hypocotyl using a razor blade (90° butt graft). Then, scions and rootstocks were placed on a nitrocellulose filter (Hybond). Hypocotyls of scions and rootstock were introduced into a silicon microtube (2 mm long) to connect them to each other, and incubated under short-day conditions (8 h light, 16 h dark) for 7 to 14 days. Grafted seedlings that did not show adventitious roots were transferred to soil and grown under a short-days photoperiod.

GUS activity and GUS RNA analysis

GUS protein was extracted from plant leaves, and GUS activity was quantified by monitoring the quantity of 4-methylumbelliferone produced by cleavage of the substrate 4-methylumbelliferyl-b-D-glucuronide (Duchefa) on a fluorometer (Thermo Scientific fluoroskan ascent)⁶⁴.

RNA extraction and HMW or LMW RNA gel blot analyses were performed using 5–10 µg of total RNA and GUS, U6 and 25S probes¹⁶. For the reverse transcription, RNA was treated with DNaseI (Invitrogen) and 1 μg of DNA‐free RNA was reverse transcribed with the primer called RT_ASGUS_Linker using the RevertAid H Minus Reverse Transcriptase (ThermoFisher, http://www.thermofisher.com/) Amplification was performed by using the LK and RbcS1Rev primers¹¹, and qPCR results were normalized with EiF1a¹¹.

Whole-genome bisulfite sequencing and DNA methylation analysis

Plants were grown under short-day conditions for five weeks (experiment 1: 6b4-1, 6b4 jmj14, 6b4 jmj14 cmt3 drm2, 6b4 cmt3 drm2) or eight weeks (experiment 2: 6b4-2, 6b4 nac50 nac52, 6b4 ibm1). Genomic DNA was isolated from leaf tissue from a mix of eight plants per genotype using the Nucleospin Plant II Maxi kit (Macharey Nagel), and library preparation and sequencing was performed by BGI Genomics (Hong Kong). Briefly, genomic DNA was fragmented to 100–300 bp by sonication, end-repaired, and ligated to methylated adaptors. Bisulfite treatment was then performed using the EZ DNA Methylation-Gold kit (Zymo), and the bisulfite-treated fragments were PCR amplified and sequenced as paired-end 100 bp reads (PE100) with DNBSEQ technology. Pre-processed high-quality reads were mapped to the TAIR10 genome using bismark with default settings for paired-end libraries⁶⁵, and all downstream analysis were performed using custom R scripts. All figures presenting bisulfite data were generated using R version 4.0.2. A summary of all bisulfite sequencing data generated in this study is presented in Supplementary Table 2, and is accessible through NCBI’s Gene Expression Omnibus (GSE152584).

ChIP-qPCR analysis

ChiP was performed on chromatin using 2 g of crosslinked in vitro shoots from 15-day-old seedlings⁶⁶. Chromatin was sonicated (30 s ‘’ON”/30 s ‘’OFF”) with a Bioruptor UCD200 (Diagenode). The chromatin solution was diluted 10 fold with ChiP dilution buffer. Fifty microliters of Dynabeads Protein G (invitrogen) and 25 µL of GFP-Trap Dynabeads (chromotek) was washed twice with ChiP dilution buffer. Nineteen micrograms of Flag (Sigma F3165) antibodies were added to G Protein and incubated at least 2 h at 4 °C with gentle rotation. Beads were washed three times with ChiP dilution buffer. Then, 1 mL of the diluted chromatin was added to the beads/antibodies and incubated overnight at 4 °C with gentle agitation. Beads were washed as decribed⁶⁶. After the last TE wash, reverse crosslinking (at least 4 h at 65 °C) and elution were performed using an IPure kit (Diagenode (AL-100-0100)). The final elution was performed in 60 µL and the chromatin was stored at −20 °C until analysis.

Using the Biorad-CFX-Maestro Software, the ChIP was analyzed by qPCR using 2 µL of chromatin in triplicate. Primers are listed in Supplementary Table 1. The mean of three qPCR results (with SD < 0.4 cycle threshold) was used for each point. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)) was used as an internal ref. ⁶⁷. Results are represented as fold change: normalized expression (ΔΔCq) given by the ratio of Relative Quantity of the sample (2^{(Cq 6b4-Cq sample)} for each identical oligo with 100% efficiency) divided by the Relative Quantity of GAPDH. At least two biological replicates were analyzed each time. Results show the mean and SD of the independent biological replicates.

Co-immunoprecipitation experiment

For each bulk of plants 15 days after germination, 2 g of fresh tissue was ground in CoIP Buffer (50 mM Tris pH 7.5, 100 mM EDTA, 15% Glycerol, 1% NP40, 1% Triton X100, 1x cOmplete^TM Protease Inhibitor Cocktail (Merck). After 30 min on ice, samples were centrifugated at 7650 × g at 4 °C and the supernatant quantified after miracloth filtration. Five micrograms of this crude extract was submitted to preclear on 50 μL of magnetic beads coated with protein A for 2 h at 4 °C. After magnetic decantation, 50 μL of GFP-Trap_M (gtm-20, Chromotek) was rinsed (2x PBS0.1% and 1x CoIP buffer) and added to the supernatant. This was incubated overnight at 4 °C with gentle agitation. The GFP beads were then rinsed 6x in CoIP buffer, and finally adjusted to 40 μL of CoIP buffer, 5 μL of 6x Laemmli buffer and incubated at 95 °C for 5 min. After vortexing, 40 μL of this final supernatant was loaded onto a 10% polyacrylamide gel. Following sodium dodecyl sulphate–polyacrylamide gel electrophoresis, western blotting was performed using anti myc (mouse 9E10, sigma-M4439), anti GFP (Rabbit Abcam-Ab290) or anti tubulin (mouse B-5-1-2, sigma-T5168) antibodies. An ImageQuant-LAS4000 (GE Healthcare) was used for detection.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

Data supporting the findings of this work are available within the paper and its Supplementary Information files. A reporting summary for this Article is available as a Supplementary Information file. The datasets and plant materials generated and analyzed during the current study are available from the corresponding author upon request. Bisulfite sequencing data are accessible through NCBI’s Gene Expression Omnibus GSE152584. The source data underlying Figs. 1A–C, 2B, and 4C, as well as Supplementary Figs. 4 and 5 are provided as a Source Data file. Source data are provided with this paper.

References

Bartel, D. P. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116, 281–297 (2004).
Article CAS PubMed Google Scholar
Baulcombe, D. RNA silencing in plants. Nature 431, 356–363 (2004).
Article ADS CAS PubMed Google Scholar
Lippman, Z. & Martienssen, R. The role of RNA interference in heterochromatic silencing. Nature 431, 364–370 (2004).
Article ADS CAS PubMed Google Scholar
Havecker, E. R. et al. The Arabidopsis RNA-directed DNA methylation argonautes functionally diverge based on their expression and interaction with target loci. Plant Cell 22, 321–334 (2010).
Article CAS PubMed PubMed Central Google Scholar
Xie, Z. et al. Genetic and functional diversification of small RNA pathways in plants. PLoS Biol. 2, 642–652 (2004).
Article CAS Google Scholar
Zhang, Z., Liu, X., Guo, X., Wang, X. J. & Zhang, X. Arabidopsis AGO3 predominantly recruits 24-nt small RNAs to regulate epigenetic silencing. Nat. Plants 2, 16049 (2016).
Article CAS PubMed Google Scholar
Law, J. A. & Jacobsen, S. E. Establishing, maintaining and modifying DNA methylation patterns in plants and animals. Nat. Rev. Genet. 11, 204–220 (2010).
Article CAS PubMed PubMed Central Google Scholar
Zemach, A. et al. The Arabidopsis nucleosome remodeler DDM1 allows DNA methyltransferases to access H1-containing heterochromatin. Cell 153, 193–205 (2013).
Article CAS PubMed PubMed Central Google Scholar
Martinez de Alba, A. E., Elvira-Matelot, E. & Vaucheret, H. Gene silencing in plants: a diversity of pathways. Biochim. Biophys. Acta 1829, 1300–1308 (2013).
Article CAS PubMed Google Scholar
Mlotshwa, S. et al. DICER-LIKE2 plays a primary role in transitive silencing of transgenes in Arabidopsis. PLoS ONE 3, e1755 (2008).
Article ADS PubMed PubMed Central CAS Google Scholar
Parent, J. S. et al. Post-transcriptional gene silencing triggered by sense transgenes involves uncapped antisense RNA and differs from silencing intentionally triggered by antisense transgenes. Nucleic Acids Res. 43, 8464–8475 (2015).
Article CAS PubMed PubMed Central Google Scholar
Vaistij, F. E., Jones, L. & Baulcombe, D. C. Spreading of RNA targeting and DNA methylation in RNA silencing requires transcription of the target gene and a putative RNA-dependent RNA polymerase. Plant Cell 14, 857–867 (2002).
Article CAS PubMed PubMed Central Google Scholar
Brosnan, C. A. et al. Nuclear gene silencing directs reception of long-distance mRNA silencing in Arabidopsis. Proc. Natl Acad. Sci. USA 104, 14741–14746 (2007).
Article ADS CAS PubMed PubMed Central Google Scholar
Palauqui, J. C., Elmayan, T., Pollien, J. M. & Vaucheret, H. Systemic acquired silencing: transgene-specific post-transcriptional silencing is transmitted by grafting from silenced stocks to non-silenced scions. EMBO J. 16, 4738–4745 (1997).
Article CAS PubMed PubMed Central Google Scholar
Gazzani, S., Lawrenson, T., Woodward, C., Headon, D. & Sablowski, R. A link between mRNA turnover and RNA interference in Arabidopsis. Science 306, 1046–1048 (2004).
Article ADS CAS PubMed Google Scholar
Gy, I. et al. Arabidopsis FIERY1, XRN2, and XRN3 are endogenous RNA silencing suppressors. Plant Cell 19, 3451–3461 (2007).
Article CAS PubMed PubMed Central Google Scholar
Luo, Z. & Chen, Z. Improperly terminated, unpolyadenylated mRNA of sense transgenes is targeted by RDR6-mediated RNA silencing in Arabidopsis. Plant Cell 19, 943–958 (2007).
Article CAS PubMed PubMed Central Google Scholar
Goeres, D. C. et al. Components of the Arabidopsis mRNA decapping complex are required for early seedling development. Plant Cell 19, 1549–1564 (2007).
Article CAS PubMed PubMed Central Google Scholar
Chen, C. Y. & Shyu, A. B. Mechanisms of deadenylation-dependent decay. Wiley Interdiscip. Rev. RNA 2, 167–183 (2011).
Article CAS PubMed Google Scholar
Gregory, B. D. et al. A link between RNA metabolism and silencing affecting Arabidopsis development. Dev. Cell 14, 854–866 (2008).
Article CAS PubMed Google Scholar
Martinez de Alba, A. E. et al. In plants, decapping prevents RDR6-dependent production of small interfering RNAs from endogenous mRNAs. Nucleic Acids Res. 43, 2902–2913 (2015).
Article CAS PubMed PubMed Central Google Scholar
Moreno, A. B. et al. Cytoplasmic and nuclear quality control and turnover of single-stranded RNA modulate post-transcriptional gene silencing in plants. Nucleic Acids Res. 41, 4699–4708 (2013).
Article CAS PubMed PubMed Central Google Scholar
Thran, M., Link, K. & Sonnewald, U. The Arabidopsis DCP2 gene is required for proper mRNA turnover and prevents transgene silencing in Arabidopsis. Plant J. 72, 368–377 (2012).
Article CAS PubMed Google Scholar
Yu, A. et al. Second-site mutagenesis of a hypomorphic argonaute1 allele identifies SUPERKILLER3 as an endogenous suppressor of transgene posttranscriptional gene silencing. Plant Physiol. 169, 1266–1274 (2015).
Article PubMed PubMed Central CAS Google Scholar
Zhang, X. et al. Plant biology. Suppression of endogenous gene silencing by bidirectional cytoplasmic RNA decay in Arabidopsis. Science 348, 120–123 (2015).
Article ADS CAS PubMed Google Scholar
Le Masson, I. et al. Mutations in the Arabidopsis H3K4me2/3 demethylase JMJ14 suppress posttranscriptional gene silencing by decreasing transgene transcription. Plant Cell 24, 3603–3612 (2012).
Article PubMed PubMed Central CAS Google Scholar
Searle, I. R., Pontes, O., Melnyk, C. W., Smith, L. M. & Baulcombe, D. C. JMJ14, a JmjC domain protein, is required for RNA silencing and cell-to-cell movement of an RNA silencing signal in Arabidopsis. Genes Dev. 24, 986–991 (2010).
Article CAS PubMed PubMed Central Google Scholar
Butel, N., Le Masson, I., Bouteiller, N., Vaucheret, H. & Elmayan, T. sgs1: a neomorphic nac52 allele impairing post-transcriptional gene silencing through SGS3 downregulation. Plant J. 90, 505–519 (2017).
Article CAS PubMed Google Scholar
Rountree, M. R. & Selker, E. U. DNA methylation inhibits elongation but not initiation of transcription in Neurospora crassa. Genes Dev. 11, 2383–2395 (1997).
Article CAS PubMed PubMed Central Google Scholar
Bewick, A. J. & Schmitz, R. J. Gene body DNA methylation in plants. Curr. Opin. Plant Biol. 36, 103–110 (2017).
Article CAS PubMed PubMed Central Google Scholar
Choi, J., Lyons, D. B., Kim, M. Y., Moore, J. D. & Zilberman, D. DNA methylation and histone H1 jointly repress transposable elements and aberrant intragenic transcripts. Mol. Cell 77, 310–323 (2020). e7.
Article CAS PubMed Google Scholar
Neri, F. et al. Intragenic DNA methylation prevents spurious transcription initiation. Nature 543, 72–77 (2017).
Article ADS CAS PubMed Google Scholar
Inagaki, S. et al. Autocatalytic differentiation of epigenetic modifications within the Arabidopsis genome. EMBO J. 29, 3496–3506 (2010).
Article CAS PubMed PubMed Central Google Scholar
Miura, A. et al. An Arabidopsis jmjC domain protein protects transcribed genes from DNA methylation at CHG sites. EMBO J. 28, 1078–1086 (2009).
Article CAS PubMed PubMed Central Google Scholar
Wang, X. et al. RNA-binding protein regulates plant DNA methylation by controlling mRNA processing at the intronic heterochromatin-containing gene IBM1. Proc. Natl Acad. Sci. USA 110, 15467–15472 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Becker, C. et al. Spontaneous epigenetic variation in the Arabidopsis thaliana methylome. Nature 480, 245–249 (2011).
Article ADS CAS PubMed Google Scholar
Schmitz, R. J. et al. Transgenerational epigenetic instability is a source of novel methylation variants. Science 334, 369–373 (2011).
Article ADS CAS PubMed PubMed Central Google Scholar
Taochy, C. et al. A genetic screen for impaired systemic RNAi highlights the crucial role of DICER-LIKE 2. Plant Physiol. 175, 1424–1437 (2017).
Article CAS PubMed PubMed Central Google Scholar
Cao, J. et al. Can-Seq: a PCR and DNA sequencing strategy for identifying new alleles of known and candidate genes. Plant Methods 16, 16 (2020).
Article CAS PubMed PubMed Central Google Scholar
Beclin, C., Boutet, S., Waterhouse, P. & Vaucheret, H. A branched pathway for transgene-induced RNA silencing in plants. Curr. Biol. 12, 684–688 (2002).
Article CAS PubMed Google Scholar
Parent, J. S., Bouteiller, N., Elmayan, T. & Vaucheret, H. Respective contributions of Arabidopsis DCL2 and DCL4 to RNA silencing. Plant J. 81, 223–232 (2015).
Article CAS PubMed Google Scholar
Gelvin, S. B. Integration of Agrobacterium T-DNA into the Plant Genome. Annu. Rev. Genet. 51, 195–217 (2017).
Article CAS PubMed Google Scholar
Jeong, J. H. et al. Repression of FLOWERING LOCUS T chromatin by functionally redundant histone H3 lysine 4 demethylases in Arabidopsis. PLoS ONE 4, e8033 (2009).
Article ADS PubMed PubMed Central CAS Google Scholar
Ning, Y. Q. et al. Two novel NAC transcription factors regulate gene expression and flowering time by associating with the histone demethylase JMJ14. Nucleic Acids Res. 43, 1469–1484 (2015).
Article CAS PubMed PubMed Central Google Scholar
Zhang, S. et al. C-terminal domains of a histone demethylase interact with a pair of transcription factors and mediate specific chromatin association. Cell Discov. 1, 15003 (2015).
Deleris, A. et al. Involvement of a Jumonji-C domain-containing histone demethylase in DRM2-mediated maintenance of DNA methylation. EMBO Rep. 11, 950–955 (2010).
Article CAS PubMed PubMed Central Google Scholar
Saze, H., Shiraishi, A., Miura, A. & Kakutani, T. Control of genic DNA methylation by a jmjC domain-containing protein in Arabidopsis thaliana. Science 319, 462–465 (2008).
Article ADS CAS PubMed Google Scholar
Clough, S. J. et al. Features of a 103-kb gene-rich region in soybean include an inverted perfect repeat cluster of CHS genes comprising the I locus. Genome 47, 819–831 (2004).
Article CAS PubMed Google Scholar
Coen, E. S. & Carpenter, R. A semi-dominant allele, niv-525, acts in trans to inhibit expression of its wild-type homologue in Antirrhinum majus. EMBO J. 7, 877–883 (1988).
Article CAS PubMed PubMed Central Google Scholar
Della Vedova, C. B. et al. The dominant inhibitory chalcone synthase allele C2-Idf (inhibitor diffuse) from Zea mays (L.) acts via an endogenous RNA silencing mechanism. Genetics 170, 1989–2002 (2005).
Article PubMed PubMed Central CAS Google Scholar
Durand, S., Bouche, N., Perez Strand, E., Loudet, O. & Camilleri, C. Rapid establishment of genetic incompatibility through natural epigenetic variation. Curr. Biol. 22, 326–331 (2012).
Article CAS PubMed Google Scholar
Melquist, S. & Bender, J. Transcription from an upstream promoter controls methylation signaling from an inverted repeat of endogenous genes in Arabidopsis. Genes Dev. 17, 2036–2047 (2003).
Article CAS PubMed PubMed Central Google Scholar
Tuteja, J. H., Clough, S. J., Chan, W. C. & Vodkin, L. O. Tissue-specific gene silencing mediated by a naturally occurring chalcone synthase gene cluster in Glycine max. Plant Cell 16, 819–835 (2004).
Article CAS PubMed PubMed Central Google Scholar
Tuteja, J. H., Zabala, G., Varala, K., Hudson, M. & Vodkin, L. O. Endogenous, tissue-specific short interfering RNAs silence the chalcone synthase gene family in glycine max seed coats. Plant Cell 21, 3063–3077 (2009).
Article CAS PubMed PubMed Central Google Scholar
Sidorenko, L. V. et al. GC-rich coding sequences reduce transposon-like, small RNA-mediated transgene silencing. Nat. Plants 3, 875–884 (2017).
Article CAS PubMed Google Scholar
Creasey, K. M. et al. miRNAs trigger widespread epigenetically activated siRNAs from transposons in Arabidopsis. Nature 508, 411–415 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Mari-Ordonez, A. et al. Reconstructing de novo silencing of an active plant retrotransposon. Nat. Genet. 45, 1029–1039 (2013).
Article CAS PubMed Google Scholar
Nuthikattu, S. et al. The initiation of epigenetic silencing of active transposable elements is triggered by RDR6 and 21-22 nucleotide small interfering RNAs. Plant Physiol. 162, 116–131 (2013).
Article CAS PubMed PubMed Central Google Scholar
Voinnet, O., Vain, P., Angell, S. & Baulcombe, D. C. Systemic spread of sequence-specific transgene RNA degradation in plants is initiated by localized introduction of ectopic promoterless DNA. Cell 95, 177–187 (1998).
Article CAS PubMed Google Scholar
Cao, M. et al. Virus infection triggers widespread silencing of host genes by a distinct class of endogenous siRNAs in Arabidopsis. Proc. Natl Acad. Sci. USA 111, 14613–14618 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Elmayan, T. et al. Arabidopsis mutants impaired in cosuppression. Plant Cell 10, 1747–1758 (1998).
Article CAS PubMed PubMed Central Google Scholar
Lindroth, A. M. et al. Requirement of CHROMOMETHYLASE3 for maintenance of CpXpG methylation. Science 292, 2077–2080 (2001).
Article CAS PubMed Google Scholar
Turnbull, C. G., Booker, J. P. & Leyser, H. M. Micrografting techniques for testing long-distance signalling in Arabidopsis. Plant J. 32, 255–262 (2002).
Article CAS PubMed Google Scholar
Elmayan, T. & Vaucheret, H. Expression of single copies of a strongly expressed 35S transgene can be silenced post-transcriptionally. Plant J. 9, 787–797 (1996).
Article CAS Google Scholar
Krueger, F. & Andrews, S. R. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics 27, 1571–1572 (2011).
Article CAS PubMed PubMed Central Google Scholar
Gendrel, A. V., Lippman, Z., Martienssen, R. & Colot, V. Profiling histone modification patterns in plants using genomic tiling microarrays. Nat. Methods 2, 213–218 (2005).
Article CAS PubMed Google Scholar
Czechowski, T., Stitt, M., Altmann, T., Udvardi, M. K. & Scheible, W. R. Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol. 139, 5–17 (2005).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank Dr. Angélique Déléris for providing the pJMJ14:JMJ14-Myc line, Dr. Yoo-Sun Noh for providing the pJMJ14:Flag-JMJ14 line, and Dr. Xin-Jian He for providing the nac50 nac52 mutant.

Author information

These authors contributed equally: Nicolas Butel, Agnès Yu.
These authors jointly supervised this work: Bernard J. Carroll, Hervé Vaucheret.

Authors and Affiliations

Institut Jean-Pierre Bourgin, UMR 1318, INRAE, AgroParisTech, Université Paris-Saclay, Versailles, France
Nicolas Butel, Agnès Yu, Ivan Le Masson, Filipe Borges, Taline Elmayan, Christelle Taochy & Hervé Vaucheret
Université Paris-Sud, Université Paris-Saclay, Orsay, France
Nicolas Butel
School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, QLD, Australia
Christelle Taochy, Nial R. Gursanscky, Jiangling Cao, Shengnan Bi, Anne Sawyer & Bernard J. Carroll
Queensland Alliance for Agriculture and Food Innovation, The University of Queensland, Brisbane, QLD, Australia
Anne Sawyer

Authors

Nicolas Butel
View author publications
You can also search for this author in PubMed Google Scholar
Agnès Yu
View author publications
You can also search for this author in PubMed Google Scholar
Ivan Le Masson
View author publications
You can also search for this author in PubMed Google Scholar
Filipe Borges
View author publications
You can also search for this author in PubMed Google Scholar
Taline Elmayan
View author publications
You can also search for this author in PubMed Google Scholar
Christelle Taochy
View author publications
You can also search for this author in PubMed Google Scholar
Nial R. Gursanscky
View author publications
You can also search for this author in PubMed Google Scholar
Jiangling Cao
View author publications
You can also search for this author in PubMed Google Scholar
Shengnan Bi
View author publications
You can also search for this author in PubMed Google Scholar
Anne Sawyer
View author publications
You can also search for this author in PubMed Google Scholar
Bernard J. Carroll
View author publications
You can also search for this author in PubMed Google Scholar
Hervé Vaucheret
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

N.B., A.Y., I.L.M., C.T., N.R.G., J.C., and S.B. conducted the experiments; F.B., T.E., A.S., B.J.C., and H.V. analyzed the data; B.J.C. and H.V. designed the experiments and wrote the paper.

Corresponding authors

Correspondence to Bernard J. Carroll or Hervé Vaucheret.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks Brian Gregory and other, anonymous, reviewers for their contributions to the peer review of this work. Peer review reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Supplementary Data 1

Description of Additional Supplementary Files

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Butel, N., Yu, A., Le Masson, I. et al. Contrasting epigenetic control of transgenes and endogenous genes promotes post-transcriptional transgene silencing in Arabidopsis. Nat Commun 12, 2787 (2021). https://doi.org/10.1038/s41467-021-22995-3

Download citation

Received: 21 September 2020
Accepted: 06 April 2021
Published: 13 May 2021
DOI: https://doi.org/10.1038/s41467-021-22995-3

This article is cited by

Cold acclimation diversity in Arabidopsis thaliana: CRISPR/Cas9 as a tool to fine analysis of Tandem Gene Arrays, application to CBF genes
- Carine Géry
- Evelyne Téoulé
Development Genes and Evolution (2022)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.