Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence a distinct phase of Polycomb silencing to destabilize long-term epigenetic memory at Arabidopsis FLC Julia I. Qüesta,1,2 Rea L. Antoniou-Kourounioti,2 Stefanie Rosa, Peijin Li, Susan Duncan, Charles Whittaker, Martin Howard, and Caroline Dean John Innes Centre, Norwich Research Park, Norwich NR4 7UH, United Kingdom In Arabidopsis thaliana, the cold-induced epigenetic regulation of FLOWERING LOCUS C (FLC) involves distinct phases of Polycomb repressive complex 2 (PRC2) silencing. During cold, a PHD–PRC2 complex metastably and digitally nucleates H3K27me3 within FLC. On return to warm, PHD–PRC2 spreads across the locus delivering H3K27me3 to maintain long-term silencing. Here, we studied natural variation in this process in Arabidopsis ac- cessions, exploring Lov-1, which shows FLC reactivation on return to warm, a feature characteristic of FLC in pe- rennial Brassicaceae. This analysis identifies an additional phase in this Polycomb silencing mechanism downstream from H3K27me3 spreading. In this long-term silencing (perpetuated) phase, the PHD proteins are lost from the nucleation region and silencing is likely maintained by the read-write feedbacks associated with H3K27me3. A combination of noncoding SNPs in the nucleation region mediates instability in this long-term si- lencing phase with the result that Lov-1 FLC frequently digitally reactivates in individual cells, with a probability that diminishes with increasing cold duration. We propose that this decrease in reactivation probability is due to reduced DNA replication after flowering. Overall, this work defines an additional phase in the Polycomb mechanism instrumental in natural variation of silencing, and provides avenues to dissect broader evolutionary changes at FLC. [Keywords: FLOWERING LOCUS C; Polycomb; epigenetics; mathematical modeling; natural variation; vernalization] Supplemental material is available for this article. Received September 27, 2019; revised version accepted January 6, 2020. In annual Arabidopsis thaliana, timing of the transition Vernalization relies on the activity of a Polycomb re- to flowering is controlled by a range of environmental pressive complex 2 (PHD–PRC2), containing the core and endogenous signals. The prolonged cold aligns flower- components VERNALIZATION2 [VRN2/Su(z)12], MSI1, ing with spring in a process known as vernalization. An FIE/EED, and the E(z) homologs SWINGER (SWN) and important factor in vernalization is the MADS-box tran- CURLY LEAF (CLF) (Gendall et al. 2001; Wood et al. scription factor gene FLOWERING LOCUS C (FLC). 2006; Finnegan and Dennis 2007; De Lucia et al. 2008; An- FLC encodes a floral repressor that directly represses gel et al. 2011), and the PHD proteins VERNALIZATION5 many genes including FLOWERING LOCUS T (FT) (VRN5), VERNALIZATION5/VIN3-LIKE 1 (VEL1), and (Searle et al. 2006). Winter cold silences FLC expression, the cold-inducible VERNALIZATION INSENSITIVE3 and this repression is epigenetically maintained providing (VIN3) (Mylne et al. 2004; Sung and Amasino 2004; Greb memory of having been exposed to low temperatures and et al. 2007). The silencing is through a cis-based chroma- allowing the plant to respond to the long days of spring tin switching mechanism at FLC (Berry et al. 2015), through activation of FT, which moves via the phloem which occurs in distinct phases: (1) specific nucleation to the shoot apical meristem to activate the floral transi- of H3K27me3 close to the FLC transcription start site dur- tion (Corbesier et al. 2007; Jaeger and Wigge 2007). ing cold exposure, and (2) subsequent spreading of H3K27me3 to cover the entire FLC locus when plants are returned to the warm (Finnegan and Dennis 2007; 1 Present address: Centre for Research in Agricultural Genomics (CRAG), De Lucia et al. 2008). Nucleation confers reversible meta- CSIC-IRTA-UAB-UB, Bellaterra, Barcelona 08193, Spain 2 These authors contributed equally to this work. stable epigenetic memory, while the H3K27me3 spread Corresponding authors: caroline.dean@jic.ac.uk, martin.howard@jic.ac. uk Article published online ahead of print. Article and publication date are © 2020 Qüesta et al. This article, published in Genes & Development, is online at http://www.genesdev.org/cgi/doi/10.1101/gad.333245.119. Free- available under a Creative Commons License (Attribution-NonCommer- ly available online through the Genes & Development Open Access cial 4.0 International), as described at http://creativecommons.org/licens- option. es/by-nc/4.0/. 446 GENES & DEVELOPMENT 34:446–461 Published by Cold Spring Harbor Laboratory Press; ISSN 0890-9369/20; www.genesdev.org Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory state is widely associated with long-term epigenetic si- completely stably repressed (Fig. 1A). Cold exposure lencing of FLC (Yang et al. 2017). Increasing cold exposure (5°C) caused down-regulation of FLC mRNA that was leads to a higher proportion of FLC loci switched to the maintained at low levels for 10 d after transfer to warm Polycomb silenced state, reflected in increasing average temperatures (22°C) (Fig. 1A), but not after 30 d. Even after H3K27me3 levels in the plant (Angel et al. 2011). a 12-wk cold treatment, Lov-1 FLC expression reactivated Dynamics of FLC silencing vary in Arabidopsis acces- to some extent (Fig. 1A), unlike the rapid vernalizing geno- sions collected from different locations (Shindo et al. type Col FRI. Reduced levels of FLC allow activation of FT 2006; Coustham et al. 2012; Li et al. 2014; Duncan et al. mRNA, therefore, to examine the consequences of FLC 2015). In a Northern Swedish accession, Lov-1, FLC silenc- reactivation we measured the levels of FT after vernaliza- ing requires many more weeks of cold exposure and this tion. Some degree of FT activation was observed in Col was found to be the result of cis polymorphisms (SNPs) FRI plants after any cold treatment (Fig. 1B). However, in that quantitatively modulate the silencing (Coustham the conditions tested, FT activation was not detected after et al. 2012). Similarly, cis variation between annual and pe- 4 wk of cold in Lov-1 (Fig. 1B). Longer cold exposure was re- rennial Arabis alpina FLC homologs has been found to quired in Lov-1 to activate FT, which then correlated with modulate FLC silencing (Kiefer et al. 2017) and this influ- flowering time. In Col FRI, all tested cold exposures effec- ences stability of H3K27me3 levels over FLC in both the tively accelerated flowering upon return to warm temper- A. alpina and Arabidopsis halleri (Wang et al. 2009; Nishio atures (Fig. 1C). As found previously (Shindo et al. 2006), et al. 2016). How and at what stage these noncoding SNPs Lov-1 did not flower after 4-wk vernalization, but required affect Polycomb silencing of FLC is unknown. longer cold treatments to accelerate flowering as demon- Here, we provide an in-depth experimental and theoret- strated by the 8- and 12-wk cold treatments (Fig. 1C). ical analysis of the influence of the Lov-1 noncoding SNPs on FLC epigenetic silencing. Our analysis reveals a further distinct phase in the sequence of events in Polycomb si- H3K27me3 can be lost from Lov-1 FLC even after fully lencing, with the SNPs causing instability in epigenetic spreading across the locus memory even after full spreading of H3K27me3 across H3K27me3 accumulation at FLC differs significantly be- the locus. We analyzed the SNPs and show how their ep- tween rapid and slower vernalizing A. thaliana accessions istatic combination mediates this instability. Our work (Shindo et al. 2006; Coustham et al. 2012; Duncan et al. defines an additional mechanistic phase in Polycomb- 2015). To provide a more detailed view of H3K27me3 dy- based epigenetic memory and provides findings likely to namics over the different phases of vernalization, we car- have implications for flowering time across plants with ried out ChIP analysis, comparing Col FRI and Lov-1 FLC, a broad range of reproductive strategies. including a higher density of primers and more time points during and after cold treatment (Fig. 2). At the cell population level, H3K27me3 gradually accumulated Results during the cold at the nucleation region of FLC in Col FRI (Fig. 2), consistent with the PHD–PRC2 complex ac- Lov-1 FLC is never completely stably repressed cumulating at the same region during cold exposure (De by vernalization Lucia et al. 2008; Yang et al. 2017). In Lov-1, as in previous A detailed analysis of FLC silencing after different lengths observations, the population level H3K27me3 at FLC is of cold treatment revealed that Lov-1 FLC is never significantly lower in NV conditions (P-value < 0.0001), A Figure 1. FLC in Lov-1 reactivates to differ- ent levels after different periods of cold. (A) Expression of FLC mRNA before the cold and after 4, 8, and 12 wk of vernalization. (NV) Nonvernalized. (B) Expression of FT mRNA in the same samples. (A,B) Values are means ± SD of three biological replicates. B (C ) Flowering time was recorded as days to flower after vernalization after the indicated period of cold. Triangles indicate individual measurements (n = 12–48). (DNF) Did not flower. C GENES & DEVELOPMENT 447 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. A cold. In contrast, the H3K27me3 that had accumulated in the cold and spread across the gene body was lost from the Lov-1 FLC chromatin after 30 d in the warm (Fig. 2B, right column). Such a phenomenon has previous- ly been reported for FLC and its orthologs in perennial spe- cies (Wang et al. 2009; Nishio et al. 2016). Reduction of H3K27me3 and therefore lack of maintenance of stable B FLC repression had also been reported in mature leaves of Col FRI, when 10-wk-old vegetative plants were vernal- ized for 4 wk (Finnegan and Dennis 2007). However, in the latter case it is not clear whether H3K27me3 had fully spread along the locus before FLC reactivation. It is worth noting that in young seedlings and under the conditions tested in the current work, we never detected postcold FLC reactivation in Col FRI plants (Figs. 1A, 3B–D, 5A, C, 6B; Supplemental Figs. S3, S9A). In parallel, we tested the behavior of H3K36me3 (Supplemental Fig. S2), a chromatin modification that antagonizes H3K27me3 at the FLC nucleation region dur- ing vernalization (Yang et al. 2014). A high peak of H3K36me3 was found at the FLC transcriptional start site in NV conditions both in Col FRI and Lov-1 (Supple- mental Fig. S2A). Cell population H3K36me3 levels de- C creased during cold in both genotypes (Supplemental Fig. S2B) and the reduced levels were maintained 10 d after cold. Together with Lov-1 FLC transcript reactivation (Fig. 1A), H3K36me3 levels rose at 30 d after 4 and 8 wk of cold in Lov-1 FLC, whereas H3K36me3 levels remained low in the stably silenced Col FRI FLC (Supplemental Fig. S2B). Only after 12 wk of cold were no differences ob- served in H3K36me3 levels between the two genotypes. We next considered what might underlie these differ- ences in Lov-1 FLC regulation. Lov-1 FLC effectively nu- cleates H3K27me3 in the cold, thereby differing from the behavior of FLC in the PHD–PRC2 mutants vin3, Figure 2. Cell population level H3K27me3 initially accumu- vrn5 (Supplemental Fig. S3), and vrn2 (Chandler et al. lates at Lov-1 FLC, but can subsequently be lost. (A) Comparison of Col FRI and Lov-1 H3K27me3 levels along the FLC locus using 1996; Gendall et al. 2001; Sung and Amasino 2004; Greb ChIP qPCR with primers in the locations indicated after 10 d of et al. 2007), all of which block nucleation of H3K27me3. growth at 22°C. (B) As for A, showing samples following the indi- Although similar in terms of FLC mRNA reactivation, cated vernalization and postcold treatment. (C ) The data of B H3K27me3 initially spreads across the gene body in Lov- (with additional time points 20 d warm after 4 and 12 wk of 1, unlike in lhp1 and clf mutants, which nucleate but can- cold) shown as the average value of all primers in the nucleation not spread H3K27me3 (Fig. 2; Supplemental Fig. S3; region (fifth through eighth primers of Supplemental Table S1) Mylne et al. 2006; Sung et al. 2006; De Lucia et al. 2008; and body region (denoted as “spreading”) (12th–13th primers of Yang et al. 2017). These data suggest that the components Supplemental Table S1). (A,B) Values are means ± SEM of two required for transcriptional down-regulation and Poly- to four biological replicates. (C) Error bars indicate SEM. comb silencing during the cold are functional and not af- fected by the cis-polymorphisms present in the Lov-1 and after 2 wk (P-value < 0.0001) and 4 wk (P-value = FLC allele. It is only later, in the long-term maintenance 0.0018) of cold (Fig. 2; Supplemental Fig. S1). Lov-1 re- phase, that these elements mediate the reactivation phe- quired at least 8 wk of cold for H3K27me3 to attain similar notype of the Lov-1 accession. Previous models have pro- levels as Col FRI (Fig. 2B,C; Supplemental Fig. S1). posed that spreading of H3K27me3 along the FLC locus is In Col FRI, spreading of H3K27me3 across the whole sufficient to maintain the FLC repressed state through FLC locus occurred after the plants were moved to subsequent cell divisions (Angel et al. 2011). For Lov-1 warm temperatures, though some spreading occurred in FLC, this is true for only a limited number of days follow- the cold with longer treatments (Fig. 2B,C). Similarly, ing transfer to warm temperatures. H3K27me3 spread across the FLC locus in Lov-1 (Fig. 2B,C), after 10 d of growth in warm temperatures regard- Reactivation is the loss of silencing at individual FLC loci less of the duration of the cold treatment (Fig. 2B, middle column). In all cases, after 10-d warm the gene body Vernalization is a cell-autonomous process with each H3K27me3 levels were proportional to the duration of copy of FLC in individual cells switching from an 448 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory A B C E F D Figure 3. Stochastic and cell autonomous reactivation of Lov-1 FLC. (A) Schematic depicting the experimental design. Col FRI and Lov-1 plants were pregrown for 7 d (NV) and then vernalized for 4 wk. Cells with active (on) or inactive (off) FLC expression are represented with yellow and red circles, respectively. After vernalization, plants were moved to warm conditions. At 7 d after cold, vernalized plants were divided into two groups: (1) Leaf samples were collected (4WT7) for RNA extraction, and (2) all leaves were marked with a pen and the plants were allowed to grow for two additional weeks. At the end of this period (4WT21), leaves that were present at 4WT7 (old leaves, dark green) and leaves that arose after 4WT7 (new leaves, pale green) were sampled separately. xWTy refers to y days of warm following x weeks of cold treatment. (B) Expression of FLC mRNA in Col FRI and Lov-1 plants at 7 d (4WT7) and 21 d (4WT21) after cold. White bars represent 4WT7 data. Dark-green and pale-green bars represent FLC levels in old and new leaves at 4WT21 (dashed line), respectively. Numbers on X-axes (1–4) depict the four individual plants tested at 4WT21. One-tailed, paired Student’s t-test results: P-value = 0.9578 testing whether new leaves have higher FLC than old leaves at 4WT21 in Lov-1; P-value = 0.4939 for the same in Col FRI. One-tailed, two-sample t-test results: P-value =3.572 × 10−6 testing whether pooled 4WT21 leaves have higher FLC than 4WT7 in Lov-1; P-value = 0.9916 for the same in Col FRI. (C ) smRNA FISH showing FLC mRNA in isolated root meristematic cells from Lov-1 and Col FRI plants, either nonvernalized (NV) or treated with 4 (4WT0) and 8 (8WT0) wk of cold, and then 30 d (8WT30) in postcold warm. Cells showing stochastic reactivation (arrowheads). Scale bars, 10 μm. (D) Quantification of the number of cells with (+) and without (−) FLC mRNA signal at the time points depicted in C. N = number of cells analyzed. (E) Schematic structure of chromosome 5 of Col FRI (gray), Lov-1 (black) and F1 plants generated from reciprocal crosses between Col FRI and Lov-1. Dashed lines depict the FLC region. (F) ChIP-qPCR profiles of H3K27me3 at 30 d following a 4-wk vernalization treatment (4WT30) in Col FRI, Lov-1, and crosses. Lov-1 and Col FRI data are means ± SEM of two biological replicates. Data for one biological replicate is shown for each type of F1. Leaves of 24 individual plants were sampled for each biological replicate. epigenetically on to off state during cold exposure (Angel lenced FLC to off during the cold could then revert to et al. 2011; Berry et al. 2015). In the analysis described the on state, a more similar scenario to the meristem re- above, FLC expression measurements and ChIP experi- version observed in some perennial species (Aikawa ments were undertaken on whole plant samples. Thus, et al. 2010). the Lov-1 FLC reactivation and H3K27me3 profiles reflect We designed an experiment to distinguish between cell population averages and do not directly reveal FLC re- these two hypotheses (Fig. 3A). We vernalized Col FRI activation dynamics at individual loci. We reasoned that, and Lov-1 plants for 4 wk (Fig. 3A). Seven days after the following vernalization, the higher postcold FLC tran- cold treatment (4WT7), we collected all rosette leaves script levels detected in Lov-1 could be achieved in two from a pool of four Col FRI and Lov-1 plants to determine different ways. One possibility was that the increased lev- FLC expression levels (Fig. 3A,B). In parallel, we marked els of FLC are the result of a new population of cells that the leaves of four individual Col FRI and Lov-1 plants to never experienced cold temperatures, but instead arise identify the ones present already at that time and allowed from the division of “FLC-active” cells after cold. A sec- them to grow for an additional 14 d in the warm. At ond option would be that cells that had epigenetically si- 4WT21, we sampled separately those leaves that were GENES & DEVELOPMENT 449 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. present at 4WT7 (the marked leaves) (dark-green leaves in copies of the FLC gene. To this end, we performed recipro- Fig. 3A) and the newly formed leaves (the unmarked cal crosses between Lov-1 and Col FRI (Fig. 3E), and ana- leaves) (pale-green leaves in Fig. 3A) for RNA extraction. lyzed the H3K27me3 ChIP profiles of F1 plants after If hypothesis 1 is correct, only the postcold developed vernalization (4WT30). The intermediate accumulation leaves would produce FLC mRNA. Thus, we should ob- of H3K27me3 in F1 plants (Fig. 3F) suggests that while serve a significant difference in FLC expression levels be- the Col FRI allele is covered by H3K27me3, the Lov-1 al- tween the old and the newly formed leaves. In contrast, a lele has lost this histone modification. Tracking two FLC more even FLC expression in both leaf types would favor copies with different fluorescent tags previously showed hypothesis 2. Although we cannot exclude the possibility that one copy of FLC may be switched off, while the other that the pool of “new leaves” could also contain some remains switched on, when located in different chromo- leaves that arose from leaf primordia that developed dur- somes within the same cell (Berry et al. 2015). Here, our ing the cold, the “old leaves” pool certainly will not con- data support the hypothesis that memory of FLC activity tain leaves formed after the cold, and so under hypothesis after vernalization can be stored locally in the chromatin 1 they will not reactivate. Therefore, this experimental environment of the FLC gene itself, even when the two design should still allow us to distinguish between the copies are located in allelic positions. two hypotheses, depending on whether we observed dif- ferences in FLC expression levels between the two types Modeling FLC reactivation of pool (“old leaves” present during cold vs. “new leaves” mixture of cold and postcold leaves). As expected, we ob- The Lov-1 FLC allele remains silenced for ∼10 d after cold served that postcold stable repression of FLC occurs in after 4 wk of vernalization (Fig. 1A). After that time, cell- Col FRI (Fig. 3B). In Lov-1, although FLC levels increase autonomous stochastic reactivation occurs (Fig. 3). This from 4WT7 to 4WT21, there is no evidence that the delay in reactivation is not straightforward to explain, FLC levels at 4WT21 differ significantly between the since the temperature is constant for the whole postcold two different types of leaves tested (Fig. 3B, P-value = period. Furthermore, the reactivation rate after this delay 0.9578 comparing old and new leaves), suggesting that is relatively high, as FLC levels rise back to NV levels after the reactivation observed is more likely the result of the 30-d warm following 4-wk cold (Fig. 1A). postcold loss of chromatin silencing at individual FLC Our previous work had revealed that local nucleation of loci (hypothesis 2). PHD–PRC2 (without subsequent spreading) gave a meta- We were able to confirm this result by performing sin- stable memory state, which transitions to a long-term epi- gle-molecule RNA FISH (smRNA FISH) in root meriste- genetically silenced state associated with H3K27me3 matic cells (Duncan et al. 2016). We designed probes to spread across the whole locus (Yang et al. 2017). To ex- specifically visualize single FLC mRNA molecules in indi- plain the Lov-1 results, we propose that there is an addi- vidual cells of Arabidopsis root meristematic tissue dur- tional, later phase in the silencing mechanism, which ing and after cold treatment. Lov-1 and Col FRI exhibit we term the perpetuated state (Fig. 4). In this state, the high levels of FLC mRNA expression before the cold PHD proteins are lost from the nucleation region, with si- (NV in Fig. 3C), and thus all cells analyzed showed FLC lencing maintained potentially by the positive read–write mRNA signal. As expected, the cold-induced FLC repres- feedbacks associated with H3K27me3 (Margueron et al. sion detected in whole plants (Fig. 1A) is also observed at 2009). The SNPs in Lov-1 FLC would then lead to instabil- the level of individual cells (4WT0 and 8WT0, Fig. 3C), ity of the perpetuated state, explaining the observed delay with Lov-1 and Col FRI cells losing FLC mRNA signal dur- in reactivation, as multiple transitions back to the high ing cold. Thirty days after the cold (8WT30), Col FRI FLC expression H state must occur before expression is reacti- is stably silenced with no cells exhibiting FLC mRNA ex- vated. The spread and perpetuated states are also observ- pression (Fig. 3C). In contrast, we observed individual Lov- able in Col FRI FLC, but the changes in H3K27me3 are 1 root meristematic cells showing FLC reactivation (ar- more subtle in that case. In our detailed ChIP profile of rowheads, Fig. 3C). Lov-1 FLC mRNA reactivation may Figure 2, we observed a reduction of H3K27me3 in the nu- also be observed in a single cell out of many cells in a tissue cleation region after 30 d after cold in Col FRI, with the section (Supplemental Fig. S4A–D), again favoring hypoth- levels coming down to those of the body region. However, esis 2 stated above. In addition, we also detected sectors of the perpetuated state in Col FRI remains stably silenced FLC mRNA reactivation within other tissue samples (Sup- because the Lov-1 SNPs are absent. This result suggests plemental Fig. S4E). We reasoned that these sectors might that the same processes are occurring in both Lov-1 and be the result of clonal division of cells that had reactivated Col FRI. Consistently, the levels of the metastable mem- FLC expression. To show that the behavior described ory protein VRN5 are also decreased at the FLC locus after above is specific to FLC, we also tested expression of cold in Col FRI (Yang et al. 2017). PP2A mRNA, which was evenly expressed at 9WT30 We developed a mathematical model incorporating the both in Col FRI and Lov-1 cells (Supplemental Fig. S4F). epigenetic phases shown in Figure 4 (Supplemental Fig. Thus, our results strongly suggest that Lov-1 FLC reactiva- S5), in an effort to reproduce the reactivation of Lov-1 tion is an on/off (digital), cell-autonomous process. FLC after 4 wk of vernalization. This model is an exten- So far, we have provided evidence that Lov-1 FLC reac- sion of the vernalization model presented in Antoniou- tivation occurs stochastically in individual cells. We next Kourounioti et al. (2018), adapted to Lov-1 and fitted to asked whether the loss of H3K27me3 occurs at individual H3K27me3 and H3K36me3 ChIP data, as well as FLC 450 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory treatments. This observation suggested that a potential factor involved in FLC postcold stable repression also changes its expression after longer vernalization treat- ments. An obvious candidate factor is FT, a direct target of FLC, given its postcold expression profile. FT mRNA expression was not activated in Lov-1 after a 4-wk vernal- ization treatment (Fig. 1B); instead, longer cold exposures of 8 and 12 wk were required for FT activation. In addition, it had previously been reported that FT feeds back to re- press FLC at both vegetative and reproductive stages (Chen et al. 2014; Chen and Penfield 2018; Luo et al. 2019). Therefore, we wondered whether the lack of FT ac- tivation in Lov-1 after 4 wk of cold could explain FLC re- activation in that case and, more generally, the differences in the reactivation probability (Fig. 1A,B). We tested this hypothesis by transferring Col FRI and Lov-1 plants to the warm in either long-day (LD) or short-day (SD) photoperiods, conditions that either pro- mote or prevent FT activation, respectively (Fig. 5). After Figure 4. Model of FLC reactivation in Lov-1. The proposed 5 wk of vernalization, further FLC reduction was observed phases of FLC silencing and reactivation. Schematic of the FLC in Col FRI at 10 and 20 d after cold, independent of the gene, showing the histone modifications in different epigenetic photoperiod (Fig. 5A,C). As expected, FT activation only states. (Red flags) H3K27me3; (green flags) H3K36me3. VIN3 is occurred in LD (Fig. 5D) but not in SD photoperiods (Fig. necessary for the nucleation of PHD–PRC2, which is associated 5B). A similar result was found for FT in Lov-1, although with H3K27me3 in the nucleation region of FLC. Spreading of these plants required a longer treatment of 10 wk of cold the PRC2 results in H3K27me3 modifications along the whole to induce FT (Fig. 5D). Nevertheless, FLC reactivation gene. The green arrow at the right shows the Lov-1-specific reac- was detected after 10 wk in Lov-1 at 20 d after cold under tivation step. Transitions between these states are simulated in both SD and LD conditions, and therefore regardless of the the mathematical model. The blue background indicates the cold-dependent transitions. absence or presence of FT transcripts, respectively (Fig. 5A,C). Indeed, in LD, where there was more FT, the reac- tivation was higher, unlike what we would expect if FT expression data. This model broadly reproduced all these was stabilizing FLC repression. Thus, our data indicate data, except for the continued reduction of the FLC levels that FT levels do not influence FLC epigenetic stability after the cold in Col FRI (Supplemental Fig. S6). in Lov-1. Another possibility is that the Lov-1 polymor- We also explored why the extent of Lov-1 FLC reactiva- phisms could disrupt either or both of the previously tion is different depending on the duration of cold. We found that the model introduced above actually gives a higher rate of reactivation after longer cold, contrary to the data. To illustrate this effect, we constructed a simpler A B model, ignoring for simplicity the delay in reactivation. This model describes a single transition from off to on in the warm, and can be written using a single ordinary differential equation (see the Materials and Methods for details). In such a simple model with a constant probabil- ity for any off cell to reactivate, the reactivation rate is proportional to the fraction of off cells at the end of the C D cold. Therefore, the reactivation rate will always be higher for the 8- and 12-wk cold-treated compared with the 4-wk cold-treated because more copies are available to reacti- vate at those times. This is the opposite of what we observed. We conclude that the reactivation probability for off cells must change with time to explain this discrep- ancy. Below we will discuss reasons for this effect and in- corporate it into the model. Figure 5. FT is not required for FLC stable repression after ver- nalization. (A–D) Expression of FLC mRNA (A,C ) and FT FT is not responsible for the change in the reactivation mRNA (B,D) in Col FRI and Lov-1 at 0 d (black bars), 10 d (gray probability bars), and 20 d (white bars) after 5 and 10 wk (only for Lov-1) of vernalization treatment. After vernalization, plants were moved We sought to determine why the Lov-1 FLC reactivation to 22°C at either short day (SD; A,B) or long day (LD; C,D) photo- probability diminishes with increasing duration of cold periods. Values are means ± SEM of three biological replicates. GENES & DEVELOPMENT 451 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. reported FT-binding sites at FLC (Chen and Penfield 2018; occurs in both Lov-1 and Col FRI. However, after many Luo et al. 2019). However, if FT was necessary for stable rounds of replication, only Lov-1 FLC reactivates, while silencing of FLC, we would expect reactivation also in Col FRI FLC remains stably silent, suggesting that FLC si- Col FRI in SD conditions where there is no FT. The lack lencing in Col FRI is intrinsically more stable and there- of such FLC postcold reactivation (Fig. 5A) argues against fore it can withstand the continued challenge of any role of FT in stabilizing FLC after vernalization. Con- replication in high growth conditions. We next addressed sistently, it has been observed that the FT feedback repres- the differences between Lov-1 and Col FRI FLC that could sion of FLC can be overcome by vernalization (Chen and confer this difference. Penfield 2018). Lov-1 FLC polymorphisms influence stability of the Earlier flowering after longer cold is correlated with less perpetuated silenced state reactivation Analysis of chimeric FLC fusions had previously shown Given the lack of a role for FT, we continued to explore that cis, noncoding polymorphisms quantitatively modu- what might influence the FLC reactivation rate, focusing late chromatin silencing of Lov-1 FLC (Coustham et al. on a role for flowering itself. To further probe this aspect, 2012). To test whether these polymorphisms are responsi- Lov-1 plants were grown for longer after cold (beyond 30 d) ble for the instability of the perpetuated state in Lov-1 in SD conditions, where flowering is suppressed. Our re- FLC, we performed ChIP experiments using near isogenic sults show that, in the absence of floral induction, Lov-1 lines (NIL1) (Fig. 6A; Coustham et al. 2012). Similar to plants continued to reactivate their FLC even after 12 Lov-1, FLC expression is reactivated in NIL1 plants after wk of vernalization (Supplemental Fig. S7A). This is con- 4 wk of cold, although this reactivation is not as strong sistent with results from perennials that showed more re- as that observed in Lov-1 (Fig. 6B). Nevertheless, NIL1 activation in meristems that continued vegetative growth FLC reactivation occurred even after 8 wk of cold and compared with flowering meristems (Lazaro et al. 2018). postcold FT activation was delayed in both Lov-1 and We then considered that the transition to flowering sup- NIL1 compared with Col FRI (Supplemental Fig. S9). Nu- pressed postcold production of leaves and that this was the cleation and spreading of H3K27me3 occurred in all the factor that influenced the level of FLC reactivation. Con- plants tested (Fig. 6C), but Lov-1 and NIL1 showed a sig- tinued vegetative growth after cold involves continued nificant loss of H3K27me3 at 30 d after cold and accumu- cell division/DNA replication. DNA replication repre- lation of H3K36me3 (Fig. 6C), with the result for sents the major challenge to maintenance of chromatin H3K27me3 confirmed in a second NIL line (NIL2) (Sup- states, by disrupting histone modifications (Alabert and plemental Fig. S10; Duncan et al. 2015). The intermediate Groth 2012). Fewer replication events therefore mean a re- phenotype of the NILs between Col FRI and Lov-1 may re- duced need for the Polycomb apparatus that rebuilds the flect factors other than the cis-localized SNPs contribut- silenced state, especially after DNA replication. Indeed, ing to instability in FLC silencing. we see a reduction in the levels of the core PRC2 methyl- Among all the noncoding polymorphisms between the transferase components SWN and CLF at the FLC locus Col-0 and Lov-1 FLC alleles, four single-nucleotide chang- when the plants had flowered and the rosette had stopped es (Lov-1 key SNPs) at the 5′ end of the gene are linked growing (Supplemental Fig. S7B,C), as assayed by ChIP ex- to the differential FLC silencing of Lov-1 and Col FRI periments in a Col FRI background at T30. (Coustham et al. 2012). These Lov-1 SNPs are located at Plants that had been vernalized for longer, flowered positions −121, −56, +326, and +598 bp from the transcrip- sooner, thus adding fewer new rosette leaves (Supplemen- tional start site (TSS) of the gene (Fig. 7A). We sought to tal Fig. S7D,E). In contrast, after short vernalization treat- determine the contribution of these four SNPs to the dis- ments, more leaves are produced before flowering, so ruption of the long-term epigenetic memory observed in there is more cell division/DNA replication. In the case Lov-1 FLC after the cold. To address this question, we of Lov-1, this could be the cause of reactivation of FLC started by comparing the vernalization response of Lov-1 and therefore delay flowering even further, such that with another accession, Edi-0. The reason for choosing most Lov-1 plants never flower after a short 4-wk vernal- Edi-0 was that it shares three of the key SNPs with Lov- ization (Fig. 1C). The continued leaf production and 1 (−121, −56, and +326) (Fig. 7A; Li et al. 2014), even the many rounds of cell division/DNA replication eventu- though it is stably silenced after relatively short cold peri- ally cause FLC reactivation to almost NV starting levels ods, with a rapid vernalization (RV) response type equiva- (Fig. 1A). lent to Col FRI (Shindo et al. 2006; Li et al. 2014). We Incorporating this hypothesis into our full model, we compared the accumulation of H3K27me3 in Lov-1 and were able to reproduce a reduction in the reactivation Edi-0 before (Fig. 7B) and after 4 wk of cold treatment rate of Lov-1 FLC dependent on the duration of cold expo- (Fig. 7C). Edi-0 FLC maintains high levels of H3K27me3 sure (Supplemental Fig. S8). The model captured all the after 4 wk of cold exposure (Fig. 7C), suggesting that the qualitative features of the FLC mRNA, H3K27me3 and +598 SNP is the causative SNP for the Lov-1 phenotype. H3K36me3 levels in the cold and subsequent warm for To further explore whether +598 prevents stable FLC si- various cold durations, except for the continued reduction lencing, we generated the following transgenic lines in a of the FLC levels after the cold in Col FRI, as previously. common Col FRI background (Fig. 7D): (1) Lov FLC lines The reduced growth of the rosette in response to flowering that carry a Lov-1 FLC transgene, and (2) Col FLC 452 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory A B Figure 6. Lov-1 cis, noncoding polymorphisms lead to instability of perpetuated silenced state at Lov-1 FLC. (A) Schematic structure of chromo- some 5 of Col FRI (gray), Lov-1 (black) and the nearly isogenic line NIL1 (Coustham et al. 2012). Black boxes represent the introgressed Lov-1 genomic segments in NIL1. (B) Expression of FLC mRNA in Col FRI, Lov-1 and NIL1 at C 0, 10, 20, and 30 d of growth in the warm after a 4-wk vernalization treatment. Values are means ± SEM of three to six biological replicates. (C ) Accumulation of H3K27me3 and H3K36me3 in the same samples and time points as described in B. Values are means ± SEM of two to four bio- logical replicates. A598T lines that carry a Col-0 type FLC construct in To further test the effect of the +598 SNP in FLC silenc- which we introduced the A-to-T change at the position ing, we generated two additional transgenic constructs in a +598 (A598T). To account for the variation that may arise Col FRI background (Fig. 7D): (1) LEL FLC A598T that due to variable transgene copy number and insertion sites carries a LEL FLC chimera in which we substituted the in the new transgenic lines generated, the experiments A to T at position +598, and (2) ELE FLC T598A that carries were carried out using pools of lines. One rosette leaf of an ELE FLC with T mutated to A at position +598. Interest- each independent transgenic line (36 Lov FLC and 24 ingly, introducing the +598 SNP into the LEL chimera re- Col FLC A598T) was collected and pooled for RNA extrac- sults in FLC reactivation after 4 wk of cold similar to tion to test FLC expression in plants exposed to 4 wk of ELE, Lov FLC, and the Lov-1 accession (Fig. 7E). In con- cold (Fig. 7E). FLC mRNA levels of Lov FLC and Col trast, removing the Lov-1 +598 T of the ELE FLC results FLC A598T were compared with those of a previously in stable silencing similar to LEL, Col FLC, Col FRI, and characterized transgenic line that carries the Col-0 FLC Edi-0 (Fig. 7E). In these cases, 60 and 74 independent trans- transgene (Col FLC) (Csorba et al. 2014). As expected, genic lines were pooled for LEL FLC A598T and ELE FLC FLC was reactivated in the Lov FLC lines (Fig. 7E), rein- T598A, respectively. These results show that SNP +598 forcing that Lov-1 cis polymorphisms are needed for has a strong effect on Lov-1 FLC reactivation, but only FLC unstable silencing. However, FLC was stably re- when combined with the other three Lov-1 SNPs. Thus, pressed in both Col FLC and Col FLC A598T lines (Fig. the noncoding SNPs interact epistatically to influence 7E) indicating that introducing the +598 SNP alone in the stability of the perpetuated silenced state at FLC. the Col-0 FLC background is not enough to disrupt FLC si- The transgenic analysis presented in this study was per- lencing after cold. formed in the common genetic background Col FRI to Previous work from our laboratory had demonstrated specifically test how the Lov-1 noncoding SNPs interfere that swapping the FLC region that contains the four with the FLC perpetuated silencing following vernaliza- Lov-1 SNPs reciprocally between Lov-1 and Edi-0 FLC al- tion. However, we cannot exclude that other trans-factors leles can revert the vernalization response of these two in addition to the noncoding SNPs contribute to Lov-1 FLC Arabidopsis accessions (Li et al. 2014). Transgenic lines instability (Fig. 6B,C; Supplemental Fig. S10). carrying the chimeric Lov-1/Edi-0/Lov-1 (LEL) FLC allele The alignment of the FLC region covering the four Lov- (LEL FLC, a pool of 77 independent transgenic lines) in a 1 SNPs (Supplemental Fig. S11) shows that SNPs −121, Col FRI background (Fig. 7D) showed stable FLC silencing −56, and +598 are variable among all the taxa evaluated after 4 wk of cold (Fig. 7E; Li et al. 2014). In contrast, lines (A. thaliana, A. halleri, and A. alpina), with no clear evo- carrying the Edi-0/Lov-1/Edi-0 (ELE) FLC construct (ELE lutionary conservation between long-term stability (Col-0 FLC, a pool of 61 independent transgenic lines) in a Col and Edi-0 FLCs) or instability (Lov-1 FLC, A. halleri FLC, FRI background (Fig. 7D) showed FLC reactivation (Fig. and A. alpina PEP1) of epigenetic silencing of the FLC 7E; Li et al. 2014), which results in a slow vernalization re- orthologs. As suggested previously by Kiefer et al. sponse similar to the Lov FLC line. Thus, the ELE FLC (2017), the Lov-1 SNP +326 seems to be ancestral, showing vernalization response matches that of the Lov-1 acces- conservation among Lov-1 and Edi-0 accessions of A. sion rather than the Edi-0 accession, while the LEL FLC thaliana, A. halleri, and A. alpina, only being different chimera shows a vernalization response similar to Edi-0 in the FLC allele of A. thaliana Col-0. However, as it is and Col FRI. conserved both in Edi-0 and Lov-1, it does not correlate GENES & DEVELOPMENT 453 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. A B Analysis of H3K27me3 dynamics using Lov-1 NILs al- lowed us to demonstrate that A. thaliana long-term epige- netic memory is coordinated in cis by noncoding SNPs along the FLC locus. Furthermore, using a transgenic ap- proach, we showed that the +598 Lov-1 SNP contributes to postcold FLC instability, but only when in combination with the other three Lov-1 SNPs. These SNPs are within C the nucleation site, yet specifically affect the long-term memory state. The nucleation region could therefore be viewed as a Polycomb Response Element (PRE) by analogy with flies, where a PRE is required to effectively propagate H3K27me3 once silencing has become established (Cole- man and Struhl 2017; Laprell et al. 2017). At FLC, the nu- cleation region (PRE) is sufficient to nucleate PHD–PRC2 D E and confer metastable silencing, and our work here has shown that this process is not affected by the Lov-1 FLC SNPs. The Lov-1 SNPs do, however, affect long-term epi- genetic silencing of the perpetuated state, in a process that is potentially mediated by a histone read–write mecha- nism. In contrast, polymorphisms at the FLC locus in Col FRI might confer greater stability to this perpetuated state. The instability of the perpetuated FLC silencing state caused by Lov-1 SNPs is likely equivalent to the Figure 7. Lov-1 key SNPs prevent long-term epigenetic memory PRC2-maintained state in Drosophila that cannot hold at FLC. (A) Schematic representation of FLC genomic structures memory through DNA replication after excision of the (not to scale) of the A. thaliana accessions Col-0 (gray), Edi-0 PRE (Coleman and Struhl 2017; Laprell et al. 2017). In (green), and Lov-1 (red). The region of the Lov-1 key SNPs (Cous- mammalian cells, the JARID2/MTF2–PRC2 nucleated tham et al. 2012) is depicted with a dotted black line. (Black let- state may be comparable with the PHD–PRC 2 metastable ters) Lov-1 SNPs; (white letters) Col-0 SNPs; (i) indel; (RV) rapid state (Oksuz et al. 2018; Perino et al. 2018). vernalization response; (SV) slow vernalization response. (B,C) The absence of PRC trans-factors can result in loss of Accumulation of H3K27me3 in Lov-1 (red) and Edi-0 (green) be- epigenetic silencing through cell division (Gaydos et al. fore (NV; B) and at 0, 10, and 30 d in the warm after a 4-wk vernal- ization treatment (C). Values are means ± SEM of two (Lov-1) and 2014; Audergon et al. 2015; Ragunathan et al. 2015; Cole- three (Edi-0) biological replicates. (D) Schematic of different man and Struhl 2017; Laprell et al. 2017). However, it is transgenic lines carrying wild type and mutated Col-0 FLC trans- unlikely that the Lov-1 SNPs differentially affect binding genes (gray), and Lov-1 (red) and chimeric Lov-1/Edi-0 (red/green) of the PHD proteins VIN3 and VRN5, as they disassociate FLC transgenes. (Black letters) Lov-1 SNPs; (white letters) Col-0 from the FLC locus within 10 d following transfer from SNPs. (E) Expression of FLC mRNA at 10 (black bars) or 30 (white cold (Yang et al. 2017), and both the nucleation and bars) days in the warm after a 4-wk vernalization treatment in the spreading processes are unaffected by the SNPs. Potential different transgenic lines from D. Data are shown normalized to candidates to influence maintenance of the perpetuated T10. Values are means ± SEM of four biological replicates. state could be the H3K27me3 methyltransferases SWN and CLF, as we demonstrated that they transiently bind with unstable FLC silencing after vernalization. There- to FLC after cold (Supplemental Fig. S7B,C). Similarly, it fore, our data suggest that different specific combinations remains to be investigated whether the SNPs modulate of noncoding SNPs at FLC have evolved to influence the LHP1 activity at FLC. Another possibility is that stability of the long-term PRC2 silenced state. H3K27me3 is actively removed from the FLC locus and a candidate for such a role would be the H3K27 demethy- Discussion lase EARLY FLOWERING 6 (ELF6) (Crevillén et al. 2014). This could potentially also explain the difference in start- Our work on the mechanistic dissection of natural varia- ing FLC mRNA levels between Lov-1 and Col FRI (Fig. tion in the vernalization response of A. thaliana acces- 1A). However, the ELF6 binding profile is not limited to sions has enabled us to elaborate the phases of PRC2 the area covered by the Lov-1 SNPs (Yang et al. 2014). silencing, an understanding that may be widely applica- Stronger association of a trans-factor that promotes tran- ble. Previous models have proposed a separate nucleation scription could also explain the observed phenotype as state that confers metastable memory, followed by spread- transcription opposes Polycomb silencing (Berry et al. ing of PRC2 and H3K27me3 across the locus for long-term 2017). In all these cases, the perpetuated silent state would silencing (Yang et al. 2017). We now add a further phase become more unstable in Lov-1 compared with Col FRI, where the metastable memory element is lost and silenc- and therefore more susceptible to switching due to the ing may be maintained predominantly by the histone challenge of replication. modification feedbacks, in a configuration we term the Our work has not only helped to dissect the fundamen- perpetuated state. tal Polycomb silencing mechanism, but has also enabled a 454 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory broader dissection of the evolutionary changes at FLC. In- ferred to as RSR and SRS, respectively, in Li et al. 2014) were stability in silencing of an FLC homolog was first observed mutagenized using overlap extension PCR cloning to obtain the in the perennial species A. alpina (Wang et al. 2009). Based ELE FLC T598A and LEL FLC A598T lines, using the primers de- on the parallels with Lov-1 FLC behavior, it is possible tailed in Supplemental Table S2. The single Col FLC transgenic line used in the current work had been selected previously that instability of a perpetuated state may be the basis of (Csorba et al. 2014) for its similarity with Col FRI regarding this FLC reactivation in perennial plants. The molecular FLC mRNA expression and flowering time. All the other con- analysis of FLC behavior in perennial plants shows structs (Lov FLC, Col FLC A598T, ELE FLC T598A, and LEL many phases in parallel to A. thaliana FLC. For example, FLC A598T) were transformed into flc-2 FRI plants (Michaels analysis of A. halleri AhgFLC expression showed three and Amasino 1999). Multiple independent transgenic lines were distinct phases in a natural environment: (1) quantitative used for RNA expression analysis (36 for Lov FLC, 24 for Col down-regulation during winter, (2) maintenance at low FLC A598T, 74 for ELE FLC T598A, 60 for LEL FLC A598T, 61 levels for ∼8 wk, and (3) reactivation to NV levels to reset for ELE FLC, and 77 for LEL FLC) (Li et al. 2014). All of them com- AhgFLC for the following flowering season (Aikawa et al. plemented the early flowering phenotype of flc-2 FRI. SWN-YFP 2010). The latter profile suggests that nucleation and (Wang et al. 2006) and 35S::GFP-CLF/clf-28 (Schubert et al. 2006) were both described previously. spreading of H3K27me3 at AhgFLC chromatin transiently maintain low levels of the floral repressor following win- ter, generating a time window for flowering. The subse- Growth conditions quent loss of the repressive modification would trigger Seeds were surface sterilized and sown on MS medium plates AhgFLC reactivation, which coincides with meristem re- without glucose and stratified for 48 h at 4°C. For nonvernalized version. It remains to be investigated whether noncoding (NV) conditions, plants were grown for 10 d in long-day condi- SNPs at AhgFLC are responsible for its seasonal expres- tions (LD; 16 h light at 22°C, 8 h darkness at 18°C). For vernaliza- sion variation, but for A. alpina, replacing the endogenous tion experiments, plants were pregrown for 7 d, and then PEP1 with the FLC from its annual relative was sufficient transferred to 5°C (unless indicated otherwise) under short-day to give stable silencing and an annual flowering response conditions (SD; 8 h light, 16 h darkness) for different durations (Hyun et al. 2019). Furthermore, increasing evidence sup- of cold before being returned to either LD (16 h light at 22°C, 8 ports the hypothesis that cis-acting sequence variation in- h darkness at 18°C) or SD (8 h light, 16 h darkness, at 20°C). For fluences postcold FLC stability in the Brassicaceae (Irwin flowering time measurements as well as for ChIP and RNA ex- pression experiments at 20 and 30 d after cold, plants were trans- et al. 2016; Kiefer et al. 2017). We envisage that the molec- ferred from plates to soil 10 d after vernalization and grown in ular understanding described here, together with the controlled environment chambers in long-day conditions (16 h mathematical model, will help unpick the relative contri- light at 22°C, 8 h darkness at 20°C). butions of SNPs, a perpetuated silenced state, and DNA Flowering was counted as days to flower after the cold treat- replication-dependent reactivation, on the behavior of ment, from moving plants to warm until bolting, but did not in- FLC orthologs in plants with a range of reproductive clude pregrowth or time spent in the cold. Bolting was scored strategies. when stems reached 2 cm in height. For single molecule RNA FISH (smRNA FISH) imaging, plants were grown and vernalized in vertically oriented Petri dishes con- Materials and methods taining MS medium. Nonvernalized plants were grown for 7 d at 22°C with a 16-h photoperiod and vernalized plants were grown Plant material under the conditions mentioned above for different lengths of time. Following vernalization, plants were transferred to larger All mutants and transgenic lines were in Col FRI sf2 background plates (one plant per plate) containing fresh MS medium and (referred to as Col FRI throughout this work) (Lee et al. 1994). Mu- grown horizontally for an additional 30 d at warm conditions tant alleles of vin3 (vin3-4) (Bond et al. 2009), vrn5 (vrn5-8) (Greb (LD, 16 h light at 22°C). et al. 2007), clf (clf-81) (Kim et al. 1998), lhp1 (lhp1-3) (Larsson et al. 1998) in Col FRI background were generated previously (Yang et al. 2017). Arabidopsis thaliana natural accessions Lov- RNA preparation and qPCR 1 and Edi-0 (Shindo et al. 2006) as well as the Lov-1 introgressed lines NIL1 (Coustham et al. 2012) and NIL2 (Duncan et al. RNA was extracted as described (Box et al. 2011), using phenol 2015) were described elsewhere. equilibrated to pH8, followed by lithium chloride precipitation. Col FLC (Csorba et al. 2014) and Lov FLC (Coustham et al. RNA was DNase-treated with Turbo DNA-Free DNase (Life 2012) transgenic lines were described previously. Briefly, a SacI Technologies). cDNA was synthesized with SuperScript III re- fragment (∼12 kb) that comprises ∼6 kb of the genomic sequence verse transcriptase (Life Technologies) using either gene-specific of the FLC (Col or Lov-1) gene with its flanking natural 5′ (∼3.5 kb) primers or oligo dT (12–18). qPCR was performed using SYBR and 3′ (∼2.6 kb) sequences were cloned into pBLUESCRIPT Green master mix II on a LightCycler 480 II (both Roche). Primer (pBluescript-FLCprom::FLC), subcloned into a pENTRY by sequences are listed in Supplemental Table S2. UBC was used as SacI/XhoI digestion (pENTRY-FLCprom::FLC), and subsequently the normalization gene control. subcloned into a pSLJ-DEST vector (a modified version of the pSLJ series) (Jones et al. 1992) using Gateway Cloning technology Chromatin immunoprecipitation (Thermo Fisher Scientific). The pENTRY-FLCprom::FLC was mutagenized to generate the Col FLC A598T construct, following Nuclei were extracted using Honda buffer as described previously the Megaprimer method of site-directed mutagenesis (Sarkar and (Sun et al. 2013), using 3 g of crosslinked plant material for Sommer 1990), using the primers detailed in Supplemental Table histone ChIP and at least 5 g for SWN-YFP/CLF-GFP ChIP. In S2. The previously reported ELE FLC and LEL FLC constructs (re- all histone ChIP reactions, sonication, immunoprecipitation, GENES & DEVELOPMENT 455 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. DNA recovery, and purification were performed as previously de- is not greater than in old leaves (Fig. 3). Additionally, the 4WT21 scribed (Angel et al. 2011). The antibodies used were anti-H3 samples were pooled and compared against the 4WT7 results for (Abcam, ab1791), anti-H3K27me3 (Millipore 07-449), and anti- Col FRI and Lov-1, using a one-tailed, two-sample t-test (function H3K36me3 (Abcam ab9050). All ChIP experiments were quanti- ttest2 from Matlab R2018b). fied by quantitative PCR (qPCR) with appropriate primers (Supplemental Table S1). For H3K27me3 analysis, SHOOT MER- Alignment of FLC orthologs ISTEMLESS (STM) was used as the internal control and data are represented as the ratio of (H3K27me3 FLC/H3 FLC) to Genomic sequences of FLC orthologs were aligned using (H3K27me3 STM/H3 STM). In the case of H3K36me3, ACTIN ClustalW. was used as the internal control and the data are represented as the following: ratio of (H3K36me3 at FLC/H3 FLC) to (H3K36me3 ACT/H3 ACT). Modeling For SWN-YFP/CLF-GFP ChIP, purified nuclei were resuspend- Simple FLC expression model showing that the reactivation probability ed in RIPA buffer (1X PBS, 1% NP-40, 1% sodium deoxycholate, decreases with duration of time in the cold We introduced a simple 0.1% SDS, Roche Complete tablets, 1mM PMSF). Two rounds of model for FLC reactivation. In the warm, we had the transition sonication were performed in a Bioruptor (Diagenode) at medium M  A, where M represents the fraction of FLC off gene copies setting. Sonication I: four times during 5 min (30 sec on/30 sec and A represents the fraction of FLC on gene copies. We assumed off); sonication II: twice for 5 min (30 sec on/30 sec off). After that this transition happens with a constant probability a. each round of sonication, samples were centrifuged at 12,000 rpm for 10 min at 4°C, and the supernatant was collected in a dA dM fresh tube for the subsequent step. Protein A agarose/salmon = aM = − . dt dt sperm DNA (Millipore) and anti-GFP (Abcam, ab290) antibodies Therefore, the reactivation rate is proportional to the fraction of were used for immunoprecipitation. The data are represented as off copies M at all times. Because more FLC copies are silenced IP to input. with longer cold durations, this rate will become greater with more vernalization. This is the opposite of what was observed in our experimental results (Fig. 1A), leading us to reject our as- Single-molecule RNA FISH (smRNA FISH) sumption of a constant probability of reactivation. smRNA FISH was carried out as described by Duncan et al. (2017). Briefly, roots were fixed for 30 min in 4% paraformalde- Full FLC model Our model of FLC reactivation (Fig. 4; Supple- hyde. Samples were prepared on slides by squashing root tissue mental Fig. S5) built on previous work (Antoniou-Kourounioti under coverslips, ethanol permeabilized, washed, and incubated et al. 2018), which modeled a sequence of cell-autonomous overnight at 37°C with RNA FISH probes. smRNA FISH probes switches between digital states. Here, we kept the first two states (Supplemental Tables S3–S5) were used at a final concentration the same as before—H (high transcription) and I (inactive)—but of 25 nM. A Zeiss Elyra PS1 inverted microscope was used for im- we separated the epigenetic state E (epigenetically silenced) into aging. A x100 oil-immersion objective (1.46 NA) and cooled EM- N (nucleated), S (spread), and the newly proposed state P (perpet- CCD Andor iXon 897 camera (512512 QE > 90%) were used to ob- uated). These new states allowed us to compare the model output tain all images. Quasar 570 and 670 probes were excited by 561- against the ChIP data, as well as with the FLC expression data. and 642-nm laser lines, with signals detected between 570 and Switches between these states included the VIN3-independent 640 nm and between 655 and 710 nm, respectively. DAPI signal pathway switch (H ↔ I) as in the previous work, as well as the was acquired with a 405-nm laser excitation and signal detection VIN3-dependent nucleation switch (now the I  N and H  N between 420 and 480 nm. Maximum projections and analysis of switches). The newly added switches were spreading (N  S), three-dimensional pictures were performed using Fiji (an imple- loss of nucleated PHD–PRC2 (S  P), and reactivation (P  I), mentation of ImageJ, a public domain program by W. Rasband the latter being specific to Lov-1. available from http://rsb.info.nih.gov/ij). Z-stacks of Supplemen- A further complication that must be considered in this model is tal Figure S4 were deconvolved using AutoQuant X2 (Media that an active cell cycle is required for spreading (see below and Cybernetics). Yang et al. 2017) and presumably also for reactivation. Because only a restricted number of cells are progressing through the cell cycle (and so are dividing) in plants, in the model we must Statistical analysis consider two populations of FLC gene copies, those in dividing cells and those in nondividing cells. The normalized number of Comparing H3K27me3 levels in Col FRI and Lov-1 A two-tailed gene copies in each of the FLC states in dividing cells was repre- paired Student’s t-test was applied (paired t-test function in sented by the variables Hs, Is, Ns, Ss, and Ps, and for nondividing Graphpad) to compare whether the levels of H3K27me3 were sig- cells, the variables were H, I, N, S, and P, respectively. In both cas- nificantly different between Lov-1 and Col FRI under each cold es, these numbers were normalized to the initial number of total treatment (see Supplemental Fig. S1). The primer pairs are gene copies in dividing cells, so that the dividing copy variables “paired” according to position between the genotypes to remove could be defined as fractions (as the total number of these cells the effect of the primer position on the H3K27me3 levels and in- does not change). This normalization was for simplicity and did stead only look at the effect of the genotype. Five comparisons not affect our results. were performed, one for each time point in Supplemental Figure A simplified model for replication was implemented: The divi- S1, and so we performed the Bonferroni correction to adjust the sion of a dividing cell was assumed to lead to a new dividing cell significance level to α = 0.01. that replaces the old, and a constant number (dn) (defined in Sup- plemental Table S6) of nondividing cells that all appear instanta- Comparing reactivation of Lov-1 FLC in old and new leaves A one- neously. This structure was based on the fact that nonstem cells tailed paired t-test (function ttest from Matlab R2018b) was only divide a fixed number of times (approximately five). Our es- used to compare the reactivation in old and new leaves in each timate ignored endoreduplicating cells for simplicity, but includ- plant, with the null hypothesis that in new leaves the reactivation ing this population would give a limited number of further 456 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory events, and we found that even a large change in the dn parameter the dividing cells are capable of spreading, the apparent rate of did not affect our results. In dividing cells, the gene copies under- spreading is diluted by the nondividing cell population whose go switching between FLC states without their numbers being number is dependent on the duration of pregrowth. changed after a cycle of replication and division, and where we as- sumed that the state does not affect the likelihood of division. Early transcriptional shutdown and VIN3-dependent silencingThe Therefore, the number of total gene copies in dividing cells was “VIN3-independent pathway” (Hepworth et al. 2018) and the constant. Gene copies in nondividing cells were generated by “VIN3-dependent” nucleation switch (including the dynamics the division of dividing cells and they can also undergo transi- of VIN3) were modeled as described in Antoniou-Kourounioti tions according to Supplemental Figure S5. et al. (2018). The same temperature sensitivity was also used, The spreading (N  S) switch is known to depend on an though in this study only 5°C (cold) and 22°C (warm) were active cell cycle (Yang et al. 2017). The loss of nucleated PHD– relevant. PRC2 (S  P) and reactivation (P  I) switches are also assumed Therefore, in the model here, the VIN3-independent transition to depend on an active cell cycle, presumably replication, based rates for H ↔ I were on evidence showing that, in the absence of spreading, the meta- stable memory is lost only in cycling cells (Yang et al. 2017). This s1 = ps1 is because replication is the major challenge to the stability of si- lencing, presumably due to loss of histone modifications and oth- and er factors defining the silenced state. To incorporate these aspects ⎧ ⎪ ⎪ 0, Tn ≤ Tr1 in our model, the N  S, S  P, and P  I transitions could only ⎨ Tn − Tr1 occur in dividing cells. Furthermore, at every division of a nucle- r1 (Tn ) = pr1 , Tr1 , Tn , Tr2 , ⎪ ⎪ Tr2 − Tr1 ated gene copy (Ns), all offspring were assumed to have spread (Ss ⎩ pr1 , Tn ≥ Tr2 and S) and so the spreading rate was the same as the division rate. For the other transitions (S  P and P  I), only a proportion of where ps1, Tr1, Tr2, and pr1 are parameters defined in Supplemen- the dividing cells switches during division (Supplemental Fig. tal Table S6, and Tn is the mean nighttime temperature, where S5). Ordinary differential equations were used to describe the dy- night was defined previously as the time from 6 h before midnight namics of the system: to 6 h after midnight (Antoniou-Kourounioti et al. 2018). In the present work, the temperature was constant throughout the dHs day–night cycle, and therefore Tn was simply the temperature = −s1 Hs + r1 Is − ps s2 Hs , dt of the treatment. dIs = s1 Hs − r1 Is − s2 Is + r2 Ps , The VIN3-dependent transition rate (for I  N) is dt  dNs 0, T ≤ T1 or T ≥ T2 = s2 (Is + ps Hs ) − gNs , s2 (V, T) = , dt ps2 V (T − T1 )(T2 − T), T1 , T , T2 dSs = gNs − s3 Ss , where T is temperature, V is the VIN3 concentration, calculated dt according to the LSCD model of Antoniou-Kourounioti et al. dPs = s3 Ss − r2 Ps , (2018), and T1, T2, and ps2 are parameters defined in Supplemental dt Table S6. dH = dn gHs − s1 H + r1 I − ps s2 H, The newly defined transitions N  S, S  P, and P  I only oc- dt cur in dividing cells and therefore have rates that depend on the dI = dn r2 Ps + dn gIs + s1 H − r1 I − s2 I, temperature-dependent growth rate dt  dN pg1 , warm (22◦ C) = s2 (I + ps H), g(T) = , dt pg2 , cold (5◦ C) dS s3 = ps3 g, = dn gNs + dn (g − s3 )Ss , dt and and r2 = pr2 g, dP = dn s3 Ss + dn (g − r2 )Ps , where pg1, pg2, ps3, and pr2 are parameters defined in Supplemen- dt tal Table S6. where s1, r1, s2, r2, s3, and g are functions that determine the rates of the transitions and are explained in detail in the following sec- tions. The rate g is the division rate as well as the spreading rate. Initial conditions The initial conditions of our system were cho- Rates ps and dn are constants given in Supplemental Table S6. sen to match the observed initial conditions of the ChIP data, with the additional constraint that, in Col FRI, the relative num- Temperature input The temperature conditions [T(t)] used in the bers of gene copies in each of the FLC states, as well as the nor- model simulation matched the experimental conditions with malized FLC mRNA concentration ([FLC], defined in the next the exception of the pregrowth duration (parameter tg in Supple- section), are at steady state in the warm, with the latter equal mental Table S6). The function T(t) takes the form to 1. Therefore, at t = 0, ⎧ ◦ ⎨ 22 C, −tg ≤ t ≤ 0 pr1 T(t) = 5◦ C, 0 , t ≤ vernalization duration . Hs = H0 = (1 − P0 ) , ⎩ ◦ s1 + pr1 22 C, vernalization duration ,t ≤ 120 s1 Is = I0 = (1 − P0 ) , s1 + pr1 tg was fitted to the data since the growth rate in the first few days is not consistent with the later growth rate. This parameter, Ns = 0, Ss = 0, Ps = P0, and [FLC] = 1, where P0 describes the relative together with the growth rate and temperature dependence of number of gene copies that are off before vernalization and it is growth, also affected the rate of spreading after cold. Since only defined in Supplemental Table S6. This parameter value was GENES & DEVELOPMENT 457 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. fitted to the levels of H3K27me3 in the body region at the NV 2018), but where the H in that work is replaced by (Hs + H)/n to time point. Also important is that we defined the starting levels account for the different populations of cells. of the gene copies in nondividing cells to be equal to 0 (H = I = N = S = P = 0), so that their relative amounts before the cold match Comparison of numerical model output with data For comparison of the relative amounts of the corresponding dividing cell variables the model output with the ChIP data, we chose parameters pA, pn, that produce them. and pN (defined in Supplemental Table S6) and compared each of the histone modification measurements with the corresponding Model of FLC concentration The FLC concentration is defined as value, as indicated. the amount of FLC divided by the volume, which is approximated H3K36me3 in nucleation region: here as being proportional to the total number of gene copies (n), where the constant of proportionality is absorbed into the oth- Hs + H er parameters. n is a monotonically increasing variable calculated pA . n by n = Hs + Is + Ns + Ss + Ps + H + I + N + S + P = 1 + H + I + N + S + P. H3K27me3 in nucleation region: We modeled the FLC amount according to previous work (Antoniou-Kourounioti et al. 2018), but adjusting for the separate Ns + N Ss + S Ps + P dividing and nondividing cells as described here. FLC is tran- pn + pA + pN . n n n scribed only in the H state of the FLC gene, so the FLC mRNA amount (NFLC) obeys H3K27me3 in body region:   dNFLC Ss + S Ps + P = p f1 (Hs + H) − pf NFLC , pN + . dt n n where pf1 and pf are the transcription and degradation rates Matlab version R2018b was used to solve the models numerical- of NFLC, respectively. The FLC mRNA concentration (F ) is ly, with solver ode15s (Shampine and Reichelt 1997). NFLC F= , n Reactivation probability decreasing through stalled growth and flower- ing Flowering stalls growth in the rosette and therefore also and therefore stalls the cell cycle. As an active cell cycle is necessary for spread- dF (dNFLC /dt)n − NFLC (dn/dt) p f1 (Hs + H) NFLC dn/dt NFLC ing and reactivation in the model, we might expect flowering to = = − pf − dt n2  n n n n have an effect on the reactivation rate. Since the FLC concentra- . Hs + H dn/dt NFLC Hs + H dn/dt tion affects flowering time, FLC can effectively feedback on itself = p f1 − pf + = p f1 − pf + F n n n n n (Supplemental Fig. S8E). In Supplemental Figure S8, we propose a decreasing probability For simplicity, we ignored the effect of dilution on FLC mRNA for reactivation related to flowering time after the cold. We imple- (dn/dt)/n, since the dynamics of the mRNA due to its relatively mented this decreasing reactivation probability by changing the fast degradation (half-life of ∼6 h) (Ietswaart et al. 2017) will be growth rate function so that it depends on the history of the much faster than dilution due to growth of the plant [pf ≫ (dn/ FLC concentration. Flowering time is also controlled by the cur- dt)/n]. In fact, while the plants are very small (and so n is small), rent temperature the plant is experiencing, so this was also in- the two rates are comparable. However, the dilution rate decreas- cluded in the changing growth rate function. Specifically, the es very quickly with size so that by the end of the pregrowth pe- growth rate becomes riod (t = 0) we can approximate  rpg1 , warm (22◦ C) g= , dn/dt rpg2 , cold (5◦ C) pf + ≈ pf , n where r is the ratio of the current growth rate to the initial growth and the equation for dF/dt becomes rate and is given by dF Hs + H dr = p f1 − pf F. = −k([FLC], T)r. dt n dt We chose the initial value of F to be at steady state: At t = 0, r0 = 1 and r decays exponentially in permissive condi- tions determined by the function p f1 H0 p f1 H0 = pf F0 ⇒ F0 = .  pf pk , T . T2 and [FLC] , pFLC k= , 0, otherwise, The FLC concentration normalized to this initial level ([FLC] = F/F0) therefore obeys where pk−1 defines the timescale over which the growth rate de- cays. pFLC is the FLC threshold below which the flowering transi- p f1 (Hs + H) tion may begin (defined in Supplemental Table S6). Therefore, d[FLC] n F = − pf ⇒ while the temperature conditions are inductive (warm: T > T2) dt p f1 H0 F0 and the FLC concentration is permissive of flowering, the plant  pf  flowering is accelerated and therefore on average the growth d[FLC] Hs + H = pf − [FLC] . rate is decreasing exponentially in time. dt nH0 During the postcold warm, and while the FLC is below the With our choice of initial conditions and with the dynamical threshold, r will be decreasing while simultaneously FLC will transitions permitted in the warm in Col FRI, the FLC concentra- be reactivating. For short cold treatments, FLC will be above tion then remains at steady state in the initial warm period. the threshold so r will not decrease and reactivation will not The previous equation reduces to the equation for FLC mRNA slow down (Supplemental Fig. S8F). Conversely, for longer cold concentration previously used (Antoniou-Kourounioti et al. treatments, where the FLC levels will be lower at the start of 458 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory the warm, r will be decreasing while FLC remains under the Box MS, Coustham V, Dean C, Mylne JS. 2011. Protocol: a simple threshold, leading to less reactivation (Supplemental Fig. S8G). phenol-based method for 96-well extraction of high quality A more quantitative comparison of the slowing growth rate and RNA from Arabidopsis. Plant Methods 7: 7. doi:10.1186/ the change in reactivation will require more accurate measure- 1746-4811-7-7 ments of growth and division/endoreduplication at a cellular lev- Chandler J, Wilson A, Dean C. 1996. Arabidopsis mutants show- el. Indeed, assuming that it reflects replication, the difference ing an altered response to vernalization. Plant J 10: 637–644. between leaf addition rates observed in Supplemental Figure doi:10.1046/j.1365-313X.1996.10040637.x S7D,E can only explain the difference in reactivation rate at a Chen M, Penfield S. 2018. Feedback regulation of COOLAIR ex- qualitative not quantitative level. pression controls seed dormancy and flowering time. Science 360: 1014–1017. doi:10.1126/science.aar7361 Chen M, MacGregor DR, Dave A, Florance H, Moore K, Paszkie- Acknowledgments wicz K, Smirnoff N, Graham IA, Penfield S. 2014. Maternal temperature history activates Flowering Locus T in fruits to We thank all members of the Dean and Howard groups for discus- control progeny dormancy according to time of year. Proc sions. We also thank Scott Berry for comments and discussions at Natl Acad Sci 111: 18787–18792. doi:10.1073/pnas the early phases of this project, Giuseppe Facchetti for discus- .1412274111 sions on the modeling work, and Cecilia Lövkvist for critically Coleman RT, Struhl G. 2017. Causal role for inheritance of reading the manuscript. The project was supported by European H3K27me3 in maintaining the off state of a Drosophila Research Council Advanced Investigator grant MEXTIM and HOX gene. Science 356: eaai8236. doi:10.1126/science UK Biotechnology and Biological Sciences Research Council In- .aai8236 stitute Strategic Programme grant GEN (BB/P013511/1). Corbesier L, Vincent C, Jang S, Fornara F, Fan Q, Searle I, Gia- Author contributions: J.I.Q. and R.L.A.-K. performed all exper- kountis A, Farrona S, Gissot L, Turnbull C, et al. 2007. FT pro- iments and analyzed the data. S.R. and S.D. performed smRNA tein movement contributes to long-distance signaling in floral FISH experiments. J.I.Q. and P.L. generated Arabidopsis transgen- induction of Arabidopsis. Science 316: 1030–1033. doi:10 ic lines, and C.W. contributed to the characterization of their phe- .1126/science.1141752 notype. R.L.A.-K. and M.H. performed the modeling. J.I.Q., Coustham V, Li P, Strange A, Lister C, Song J, Dean C. 2012. R.L.A.-K, M.H., and C.D. designed the experiments, interpreted Quantitative modulation of Polycomb silencing underlies the data, and wrote the manuscript with input from all the other natural variation in vernalization. Science 337: 584–587. authors. doi:10.1126/science.1221881 Crevillén P, Yang H, Cui X, Greeff C, Trick M, Qiu Q, Cao X, Dean C. 2014. Epigenetic reprogramming that prevents trans- References generational inheritance of the vernalized state. Nature 515: 587–590. doi:10.1038/nature13722 Aikawa S, Kobayashi MJ, Satake A, Shimizu KK, Kudoh H. 2010. Csorba T, Questa JI, Sun Q, Dean C. 2014. Antisense COOLAIR Robust control of the seasonal expression of the Arabidopsis mediates the coordinated switching of chromatin states at FLC gene in a fluctuating environment. Proc Natl Acad Sci FLC during vernalization. Proc Natl Acad Sci 111: 16160– 107: 11632–11637. doi:10.1073/pnas.0914293107 16165. doi:10.1073/pnas.1419030111 Alabert C, Groth A. 2012. Chromatin replication and epigenome De Lucia F, Crevillen P, Jones AM, Greb T, Dean C. 2008. A PHD– maintenance. Nat Rev Mol Cell Biol 13: 153–167. doi:10 Polycomb repressive complex 2 triggers the epigenetic silenc- .1038/nrm3288 ing of FLC during vernalization. Proc Natl Acad Sci 105: Angel A, Song J, Dean C, Howard M. 2011. A Polycomb-based 16831–16836. doi:10.1073/pnas.0808687105 switch underlying quantitative epigenetic memory. Nature Duncan S, Holm S, Questa J, Irwin J, Grant A, Dean C. 2015. Sea- 476: 105–108. doi:10.1038/nature10241 sonal shift in timing of vernalization as an adaptation to ex- Antoniou-Kourounioti RL, Hepworth J, Heckmann A, Duncan S, treme winter. Elife 4: e06620. doi:10.7554/eLife.06620 Qüesta J, Rosa S, Säll T, Holm S, Dean C, Howard M. 2018. Duncan S, Olsson TSG, Hartley M, Dean C, Rosa S. 2016. A Temperature sensing is distributed throughout the regulatory method for detecting single mRNA molecules in Arabidopsis network that controls FLC epigenetic silencing in vernaliza- thaliana. Plant Methods 12: 13. doi:10.1186/s13007-016- tion. Cell Syst 7: 643–655.e9. doi:10.1016/j.cels.2018.10.011 0114-x Audergon PN, Catania S, Kagansky A, Tong P, Shukla M, Pidoux Duncan S, Olsson TSG, Hartley M, Dean C, Rosa S. 2017. Single AL, Allshire RC. 2015. Epigenetics. Restricted epigenetic in- molecule RNA FISH in Arabidopsis root cells. Bio Protoc 7: heritance of H3K9 methylation. Science 348: 132–135. e2240. doi:10.21769/BioProtoc.2240 doi:10.1126/science.1260638 Finnegan EJ, Dennis ES. 2007. Vernalization-induced trimethyla- Berry S, Hartley M, Olsson TSG, Dean C, Howard M. 2015. Local tion of histone H3 lysine 27 at FLC is not maintained in mitot- chromatin environment of a Polycomb target gene instructs ically quiescent cells. Curr Biol 17: 1978–1983. doi:10.1016/j its own epigenetic inheritance. Elife 4: e07205. doi:10.7554/ .cub.2007.10.026 eLife.07205 Gaydos LJ, Wang W, Strome S. 2014. Gene repression. H3K27me Berry S, Dean C, Howard M. 2017. Slow chromatin dynamics al- and PRC2 transmit a memory of repression across generations low Polycomb target genes to filter fluctuations in transcrip- and during development. Science 345: 1515–1518. doi:10 tion factor activity. Cell Syst 4: 445–457.e8. doi:10.1016/j .1126/science.1255023 .cels.2017.02.013 Gendall AR, Levy YY, Wilson A, Dean C. 2001. The VERNALI- Bond DM, Wilson IW, Dennis ES, Pogson BJ, Jean Finnegan E. ZATION2 gene mediates the epigenetic regulation of vernal- 2009. VERNALIZATION INSENSITIVE 3 (VIN3) is required ization in Arabidopsis. Cell 107: 525–535. doi:10.1016/S0092- for the response of Arabidopsis thaliana seedlings exposed 8674(01)00573-6 to low oxygen conditions. Plant J 59: 576–587. doi:10.1111/j Greb T, Mylne JS, Crevillen P, Geraldo N, An H, Gendall AR, .1365-313X.2009.03891.x Dean C. 2007. The PHD finger protein VRN5 functions in GENES & DEVELOPMENT 459 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Qüesta et al. the epigenetic silencing of Arabidopsis FLC. Curr Biol 17: 73– Margueron R, Justin N, Ohno K, Sharpe ML, Son J, Drury WJ III, 78. doi:10.1016/j.cub.2006.11.052 Voigt P, Martin SR, Taylor WR, De Marco V, et al. 2009. Role Hepworth J, Antoniou-Kourounioti RL, Bloomer RH, Selga C, of the polycomb protein EED in the propagation of repressive Berggren K, Cox D, Collier Harris BR, Irwin JA, Holm S, Säll histone marks. Nature 461: 762–767. doi:10.1038/ T, et al. 2018. Absence of warmth permits epigenetic memory nature08398 of winter in Arabidopsis. Nat Commun 9: 639. doi:10.1038/ Michaels SD, Amasino RM. 1999. FLOWERING LOCUS C en- s41467-018-03065-7 codes a novel MADS domain protein that acts as a repressor Hyun Y, Vincent C, Tilmes V, Bergonzi S, Kiefer C, Richter R, of flowering. Plant Cell 11: 949–956. doi:10.1105/tpc.11.5.949 Martinez-Gallegos R, Severing E, Coupland G. 2019. A regula- Mylne J, Greb T, Lister C, Dean C. 2004. Epigenetic regulation in tory circuit conferring varied flowering response to cold in an- the control of flowering. Cold Spring Harb Symp Quant Biol nual and perennial plants. Science 363: 409–412. doi:10.1126/ 69: 457–464. doi:10.1101/sqb.2004.69.457 science.aau8197 Mylne JS, Barrett L, Tessadori F, Mesnage S, Johnson L, Bernata- Ietswaart R, Rosa S, Wu Z, Dean C, Howard M. 2017. Cell-size-de- vichute YV, Jacobsen SE, Fransz P, Dean C. 2006. LHP1, the pendent transcription of FLC and its antisense long non-cod- Arabidopsis homologue of HETEROCHROMATIN PRO- ing RNA COOLAIR explain cell-to-cell expression TEIN1, is required for epigenetic silencing of FLC. Proc Natl variation. Cell Syst 4: 622–635.e9. doi:10.1016/j.cels.2017.05 Acad Sci 103: 5012–5017. doi:10.1073/pnas.0507427103 .010 Nishio H, Buzas DM, Nagano AJ, Suzuki Y, Sugano S, Ito M, Mor- Irwin JA, Soumpourou E, Lister C, Ligthart J-D, Kennedy S, Dean inaga S, Kudoh H. 2016. From the laboratory to the field: as- C. 2016. Nucleotide polymorphism affecting FLC expression saying histone methylation at FLOWERING LOCUS C in underpins heading date variation in horticultural brassicas. naturally growing Arabidopsis halleri. Genes Genet Syst 91: Plant J 87: 597–605. doi:10.1111/tpj.13221 15–26. doi:10.1266/ggs.15-00071 Jaeger KE, Wigge PA. 2007. FT protein acts as a long-range signal Oksuz O, Narendra V, Lee CH, Descostes N, LeRoy G, Raviram in Arabidopsis. Curr Biol 17: 1050–1054. doi:10.1016/j.cub R, Blumenberg L, Karch K, Rocha PP, Garcia BA, et al. 2018. .2007.05.008 Capturing the onset of PRC2-mediated repressive domain for- Jones JD, Shlumukov L, Carland F, English J, Scofield SR, Bishop mation. Mol Cell 70: 1149–1162.e5. doi:10.1016/j.molcel GJ, Harrison K. 1992. Effective vectors for transformation, ex- .2018.05.023 pression of heterologous genes, and assaying transposon exci- Perino M, van Mierlo G, Karemaker ID, van Genesen S, Vermeu- sion in transgenic plants. Transgenic Res 1: 285–297. doi:10 len M, Marks H, van Heeringen SJ, Veenstra GJC. 2018. MTF2 .1007/BF02525170 recruits Polycomb repressive complex 2 by helical-shape- Kiefer C, Severing E, Karl R, Bergonzi S, Koch M, Tresch A, Coup- selective DNA binding. Nature Genet 50: 1002–1010. doi:10 land G. 2017. Divergence of annual and perennial species in .1038/s41588-018-0134-8 the Brassicaceae and the contribution of cis-acting variation Ragunathan K, Jih G, Moazed D. 2015. Epigenetics. Epigenetic in- at FLC orthologues. Mol Ecol 26: 3437–3457. doi:10.1111/ heritance uncoupled from sequence-specific recruitment. Sci- mec.14084 ence 348: 1258699. doi:10.1126/science.1258699 Kim GT, Tsukaya H, Uchimiya H. 1998. The CURLY LEAF gene Sarkar G, Sommer SS. 1990. The ‘megaprimer’ method of site-di- controls both division and elongation of cells during the ex- rected mutagenesis. BioTechniques 8: 404–407. pansion of the leaf blade in Arabidopsis thaliana. Planta Schubert D, Primavesi L, Bishopp A, Roberts G, Doonan J, Jenu- 206: 175–183. doi:10.1007/s004250050389 wein T, Goodrich J. 2006. Silencing by plant Polycomb-group Laprell F, Finkl K, Müller J. 2017. Propagation of Polycomb-re- genes requires dispersed trimethylation of histone H3 at ly- pressed chromatin requires sequence-specific recruitment to sine 27. EMBO J 25: 4638–4649. doi:10.1038/sj.emboj DNA. Science 356: 85–88. doi:10.1126/science.aai8266 .7601311 Larsson AS, Landberg K, Meeks-Wagner DR. 1998. The TERMI- Searle I, He Y, Turck F, Vincent C, Fornara F, Kröber S, Amasino NAL FLOWER2 (TFL2) gene controls the reproductive transi- RA, Coupland G. 2006. The transcription factor FLC confers a tion and meristem identity in Arabidopsis thaliana. Genetics flowering response to vernalization by repressing meristem 149: 597–605. competence and systemic signaling in Arabidopsis. Genes Lazaro A, Obeng-Hinneh E, Albani MC. 2018. Extended vernali- Dev 20: 898–912. doi:10.1101/gad.373506 zation regulates inflorescence fate in Arabis alpina by stably Shampine LF, Reichelt MW. 1997. The Matlab ODE suite. SIAM J silencing PERPETUAL FLOWERING1. Plant Physiol 176: Sci Comput 18: 1–22. doi:10.1137/S1064827594276424 2819–2833. doi:10.1104/pp.17.01754 Shindo C, Lister C, Crevillen P, Nordborg M, Dean C. 2006. Var- Lee I, Michaels SD, Masshardt AS, Amasino RM. 1994. The late- iation in the epigenetic silencing of FLC contributes to natural flowering phenotype of FRIGIDA and mutations in LUMINI- variation in Arabidopsis vernalization response. Genes Dev DEPENDENS is suppressed in the Landsberg erecta strain of 20: 3079–3083. doi:10.1101/gad.405306 Arabidopsis. Plant J 6: 903–909. doi:10.1046/j.1365-313X Sun Q, Csorba T, Skourti-Stathaki K, Proudfoot NJ, Dean C. 2013. .1994.6060903.x R-loop stabilization represses antisense transcription at the Li P, Filiault D, Box MS, Kerdaffrec E, van Oosterhout C, Wilczek Arabidopsis FLC locus. Science 340: 619–621. doi:10.1126/sci AM, Schmitt J, McMullan M, Bergelson J, Nordborg M, et al. ence.1234848 2014. Multiple FLC haplotypes defined by independent cis- Sung S, Amasino RM. 2004. Vernalization in Arabidopsis thali- regulatory variation underpin life history diversity in Arabi- ana is mediated by the PHD finger protein VIN3. Nature dopsis thaliana. Genes Dev 28: 1635–1640. doi:10.1101/gad 427: 159–164. doi:10.1038/nature02195 .245993.114 Sung S, He Y, Eshoo TW, Tamada Y, Johnson L, Nakahigashi K, Luo X, Chen T, Zeng X, He D, He Y. 2019. Feedback regulation of Goto K, Jacobsen SE, Amasino RM. 2006. Epigenetic mainte- FLC by FLOWERING LOCUS T (FT) and FD through a 5′ FLC nance of the vernalized state in Arabidopsis thaliana requires promoter region in Arabidopsis. Mol Plant 12: 285–288. doi:10 LIKE HETEROCHROMATIN PROTEIN 1. Nat Genet 38: .1016/j.molp.2019.01.013 706–710. doi:10.1038/ng1795 460 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence long-term memory Wang D, Tyson MD, Jackson SS, Yadegari R. 2006. Partially re- sponse requires a polycomb-like protein complex that also dundant functions of two SET-domain Polycomb-group pro- includes VERNALIZATION INSENSITIVE 3. Proc Natl teins in controlling initiation of seed development in Acad Sci 103: 14631–14636. doi:10.1073/pnas.0606385103 Arabidopsis. Proc Natl Acad Sci 103: 13244–13249. doi:10 Yang H, Howard M, Dean C. 2014. Antagonistic roles for .1073/pnas.0605551103 H3K36me3 and H3K27me3 in the cold-induced epigenetic Wang R, Farrona S, Vincent C, Joecker A, Schoof H, Turck F, switch at Arabidopsis FLC. Curr Biol 24: 1793–1797. doi:10 Alonso-Blanco C, Coupland G, Albani MC. 2009. PEP1 regu- .1016/j.cub.2014.06.047 lates perennial flowering in Arabis alpina. Nature 459: 423– Yang H, Berry S, Olsson TSG, Hartley M, Howard M, Dean C. 427. doi:10.1038/nature07988 2017. Distinct phases of Polycomb silencing to hold epigenet- Wood CC, Robertson M, Tanner G, Peacock WJ, Dennis ES, Hel- ic memory of cold in Arabidopsis. Science 357: 1142–1145. liwell CA. 2006. The Arabidopsis thaliana vernalization re- doi:10.1126/science.aan1121 GENES & DEVELOPMENT 461 Downloaded from genesdev.cshlp.org on March 22, 2021 - Published by Cold Spring Harbor Laboratory Press Noncoding SNPs influence a distinct phase of Polycomb silencing to destabilize long-term epigenetic memory at Arabidopsis FLC Julia I. Qüesta, Rea L. Antoniou-Kourounioti, Stefanie Rosa, et al. Genes Dev. 2020, 34: originally published online January 30, 2020 Access the most recent version at doi:10.1101/gad.333245.119 Supplemental http://genesdev.cshlp.org/content/suppl/2020/01/28/gad.333245.119.DC1 Material References This article cites 60 articles, 25 of which can be accessed free at: http://genesdev.cshlp.org/content/34/5-6/446.full.html#ref-list-1 Creative This article, published in Genes & Development, is available under a Creative Commons Commons License (Attribution-NonCommercial 4.0 International), as described at License http://creativecommons.org/licenses/by-nc/4.0/. Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here. © 2020 Qüesta et al.; Published by Cold Spring Harbor Laboratory Press