SEARCH

SEARCH BY CITATION

Keywords:

  • seed germination;
  • abscisic acid (ABA);
  • ABI5;
  • chromatin remodelling;
  • PKL;
  • embryogenesis

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Embryogenesis in Arabidopsis results in mature, osmotolerant embryos within dry seeds. Late-embryogenesis programs involve the transcription factors ABI3 and ABI5, which are necessary for osmotolerance. ABI3 and ABI5 are degraded in seeds initiating germination, abolishing their protected state. However, during an early stage of germination, strong osmotic stresses, or ABA exposure maintain ABI3 and ABI5 expression, leading to growth arrest, and osmotolerance. Mild stress stimuli delay ABI3 and ABI5 disappearance, retarding germination but not preventing eventual closure of embryogenesis programs. PICKLE (PKL), a putative chromatin modifier, is necessary to repress ABI3 and ABI5 expression during germination in response to ABA. We show that pkl mutants display persistent high expression of ABI3 and ABI5 upon ABA stimulation. In turn, maintenance of ABI5 expression leads to hypersensitive germination responses to ABA in pkl seeds. We provide evidence that ABI3 and ABI5 are less associated with repressed chromatin in pkl mutants. Our results provide evidence that PKL-dependent repression of embryonic gene expression extends to late-embryogenesis genes and is associated with changes in chromatin. We suggest that effective closure of embryogenesis omostolerance programs during germination prevents excessive plant reactions to stress.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

In Arabidopsis thaliana, embryos in mature seeds are in a dormant (germination-incompetent) and metabolically near-silent state. They store food reserves and resist desiccation (Kroj et al., 2003; Vicente-Carbajosa and Carbonero, 2005). Maturation probably evolved as an adaptation to the terrestrial environment, protecting the plant until favorable conditions allow seed germination and a transition toward the autotrophic phase. Transcription factors (TFs) such as LEC1, FUS3, and ABI3 are essential for orchestrating maturation (To et al., 2006), and abscisic acid (ABA) signaling is important for regulating their expression and activity (Finkelstein et al., 2002).

LEC1 and FUS3 expression ceases prior to final embryo maturation (Nambara et al., 2000; Parcy et al., 1997; Raz et al., 2001), consistent with an early function in embryogenesis (illustrated in Figure 1). ABI3 expression, however, persists until final maturation stages, and positively regulates late embryogenesis abundant (LEA) genes (e.g. AtEm1 and AtEm6), whose products have been associated with desiccation tolerance (Parcy et al., 1994). ABI3 operates with other TFs, such as the basic leucine-zipper TF ABI5, which binds to ABA-responsive elements (ABRE) present in various LEA promoter genes (Bensmihen et al., 2002; Carles et al., 2002; Lopez-Molina et al., 2002).

image

Figure 1.  Summary of embryonic expression during embryogenesis and germination. Induction of ABI3 and ABI5 expression upon seed imbibition is possible if stress is applied during the germination time window but not thereafter.

Download figure to PowerPoint

For after-ripened (i.e. non-dormant) seeds, imbibition of water is sufficient to trigger germination (Finch-Savage and Leubner-Metzger, 2006); this requires gibberellin (GA) synthesis (Debeaujon and Koornneef, 2000; Peng and Harberd, 2002). Germination as such is a developmental transition towards autotrophy and is subject to a variety of environmental influences (e.g. light quality, osmolarity).

Under normal conditions, germination is a period during which ABI3 and ABI5 mRNA and protein rapidly decay from their high levels in dry seeds. However, germination can be arrested or delayed by osmotic stress or exogenous ABA during a time window of about 48 h after seed imbibition (Figure 1), i.e. prior to cotyledon greening (Lopez-Molina et al., 2001, 2002). Such stress signals stimulate and maintain ABI3 and ABI5 expression, leading to sustained germination arrest (i.e. growth arrest) and increased osmotolerance (Figure 1). We have suggested that the time window permitting ABA-induced growth arrest represents a last protective embryonic checkpoint prior to irreversible commitment to autotrophy (Lopez-Molina et al., 2001). In the case of mild stress, low ABA levels or delayed signals, ABI3 and ABI5 expression is extended but is short-lived (Lopez-Molina et al., 2001, 2002). Thus, a germination delay occurs but no growth arrest. This phenomenon marks the end of ABA-dependent ABI3 and ABI5 expression and of embryonic gene expression programs. Terminating ABI5 expression is essential to ensure plant growth: transgenic plants constitutively expressing ABI5 are hypersensitive to stress and cannot grow normally under low stress conditions (Lopez-Molina et al., 2001). During germination, the balance of the antagonistic effects of ABA and GA signaling influences the final outcome of growth arrest, delay or unabated growth (Finch-Savage and Leubner-Metzger, 2006).

Ogas et al. (1999) reported that pickle (pkl) mutants abnormally express LEC1 and FUS3 upon seed imbibition, which suggests that PKL normally acts to keep these genes repressed at that time. In addition, pkl mutants display embryonic characteristics in adult plants, i.e. foci of neutral lipid accumulation in primary root tips, called ‘pickle’ roots (Ogas et al., 1997). PKL encodes a putative SWI/SNF class chromatin-remodeling factor of the CHD3 type (harboring a chromodomain, a SNF2-related helicase/ATPase domain, a DNA binding domain and a PHD zinc finger; Ogas et al., 1999). In animals, factors encoded by PKL-related genes are found in Mi-2/NuRD complexes (myositis-2/nucleosome remodeling and histone deacetylase) that have been shown to support remodeling events rendering chromatin transcriptionally non-permissive (Bowen et al., 2004; Nilasena et al., 1995; Tyler and Kadonaga, 1999). These complexes repress embryonic programs during developmental transitions triggered by hormonal cues. For example, in Caenorhabditis elegans, they maintain somatic differentiation by repressing germline-specific genes (Unhavaithaya et al., 2002). In Arabidopsis, PKL expression is absent in dry seeds but is rapidly initiated upon seed imbibition (Henderson et al., 2004; Li et al., 2005).

Thus, PKL-mediated chromatin-remodeling events probably take place during germination and repress early-embryonic gene expression programs. However, the role of PKL in repressing late-embryogenesis programs involving ABI3 and ABI5 has not been investigated. PKL-dependent changes in chromatin have not been reported either. Here, we show that pkl mutants over express ABI3 and ABI5 in response to ABA. ABI5 over expression in turn leads to hypersensitive germination responses. Moreover, we found changes in histone markers associated with ABI3 and ABI5 promoter sequences, which are consistent with their chromatin being less silent in pkl seeds. Taken together, our work suggests that PKL-dependent repression of embryogenesis gene expression extends to late-maturation gene expression programs.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

pkl seeds display hypersensitive responses to exogenous ABA

Seed germination responses to ABA were evaluated by measuring embryo radical tip emergence and chlorophyll accumulation. Figure 2(a) shows ABA dose–response curves for radical tip emergence frequency in wild-type (WT) and pkl seeds. In the absence of ABA, tip emergence frequencies were similar in WT and pkl plants, whereas exacerbated responses to ABA could be observed in pkl mutants seeds, particularly at low ABA concentrations (0.25–1 μm ABA) to which WT seeds are only moderately sensitive (Figure 2a). Similarly, chlorophyll accumulation was markedly lower in pkl plants treated with ABA than in WT (Figure 2b).

image

Figure 2. pkl mutants display hypersensitive responses to ABA. (a) Effect of exogenous ABA on seed germination. WT (Col and Ws ecotypes), pkl, pkl/abi5-4 and abi5-4 seeds were grown with or without ABA. Germination was scored 48 h after stratification in two independent seed batches. (b) Chlorophyll levels in seedlings 4 days after stratification in the presence or absence of ABA. Chlorophyll levels for each genotype were measured relative to those found in the absence of ABA (normalized to 100).

Download figure to PowerPoint

Thus, ABA elicits stronger inhibition responses in germinating pkl seeds, implying that PKL is a negative regulator of ABA signaling during seed germination. As ABI3 and ABI5 expression levels determine the strength of the growth arrest, we examined the expression of these genes in response to ABA in pkl mutant seeds.

ABI3 and ABI5 expression are increased and sustained in pkl mutants in response to ABA

Figure 3(b,c) shows time courses of ABI3 and ABI5 expression in WT and pkl mutant seeds upon stratification. In the absence of ABA, downregulation of mRNA and protein from high levels in dry seeds normally occurred in both genotypes, although it was slightly but reproducibly slower in pkl mutants, notably for ABI3. Such differences were not observed by Rider et al. (2004), but this could be due to the later time point that they used and different culture conditions. In contrast, exposure to ABA (0.25 μm) led to a more marked and sustained expression (mRNA and protein) of both genes in pkl seeds compared with WT (Figure 3b,c). It should be stressed that PKL mRNA expression, which is normally initiated upon seed imbibition (Henderson et al., 2004; Li et al., 2005), was not affected by ABA (Figure 4).

image

Figure 3.  Sustained ABI3 and ABI5 expression in pkl mutants treated with low concentrations of ABA. (a) Seeds were stratified in the absence or presence of ABA as indicated. DS, dry seeds. (b, c) ABI3 and ABI5 mRNA and protein expression was recorded for WT and pkl material isolated at the indicated times. Hybridizations were performed using the same blots. Loading controls for the Western blots are shown in Figure S1.

Download figure to PowerPoint

image

Figure 4. PKL mRNA expression is not affected by ABA. PKL mRNA expression was similar in arrested, ABA-treated WT seeds and non-treated germinating seedlings (MS). Northern blots were performed as described in Figure 3 using a PKL ORF probe isolated from a plasmid kindly provided by J. Ogas, Purdue University. DS, dry seeds.

Download figure to PowerPoint

Thus, PKL prevents inappropriate maintenance of ABI3 and ABI5 expression under mild stress signals. As ABI5 operates downstream of ABI3 to arrest growth (Lopez-Molina et al., 2002), we performed further experiments to determine whether persistently elevated ABI5 accumulation in pkl mutants is related to their exacerbated germination responses to ABA.

pkl hypersensitivity to ABA requires ABI5

pkl/abi5 double mutants were generated to assess the role of ABI5 in responses of pkl mutants to ABA. The data in Figure 2 show that pkl/abi5 double mutants lost hypersensitive responses to ABA. Together with the observation that WT transgenic lines over expressing ABI5 also display hypersensitive germination responses to ABA (see Figure S7; Lopez-Molina et al., 2001), our data suggest that ABI5 over-accumulation is necessary for exacerbated pkl germination responses. In turn, high ABI5 expression in pkl may result from de-repressed chromatin associated with ABI5, or from higher ABI3 accumulation, which positively regulates ABI5 expression (Lopez-Molina et al., 2002), or a combination of both. Changes in chromatin may occur if PKL, in the context of its general influence on keeping silent the chromatin associated with embryogenesis genes, promotes ABI3 and ABI5 transcriptional repression. We examined this possibility using chromatin immunoprecipitation (ChIP) experiments.

PKL promotes ABI3 and ABI5 transcriptional repression

ChIP experiments (Bowler et al., 2004; Solomon et al., 1988) were used to compare dimethylation levels of Lys9 or Lys27 at histone H3 (H3-K9, H3-K27) between WT and pkl seeds treated or not with a low concentration of ABA (0.25 μm ABA, see Experimental procedures). High methylation levels at such histone markers are usually associated with silent, transcription-non-permissive chromatin (Fuchs et al., 2006). We included the dimethylated H3-K27 marker because it is often associated with polycomb-group (PcG) protein-mediated chromatin repression (Fischer et al., 2006; Tariq and Paszkowski, 2004), and because of available evidence that it might participate in chromatin-remodeling changes influenced by PKL (Li et al., 2005; Makarevich et al., 2006; see Discussion).

We measured DNA amounts present in ChIP precipitates using cycle threshold (Ct) values from real-time PCR amplification curves (see Experimental procedures). Ta2, a peri-centromeric retrotransposon, was used as an internal reference gene for normalization of DNA amounts, as it is constitutively repressed in all cells (thus the amount of ABI3 or ABI5 DNA in a given ChIP precipitate is divided by the amount of Ta2 DNA in the same ChIP precipitate). A series of independent controls indicated that PKL does not influence the state of the chromatin associated with Ta2 (detailed in Experimental procedures and Figure S5). We similarly used UBIQUITIN (UBQ) DNA so as to normalize DNA amounts relative to a gene constitutively expressed in all cells (see Figure S2).

ChIP experiments were performed during germination as well as after its completion, i.e in 7-day-old seedlings, a time at which WT and pkl plants express significantly lower levels of ABI3 and ABI5 (data not shown).

ABA-treated pkl mutant seeds had 2- to 2.5-fold lower H3-K9 and H3-K27 methylation levels at ABI3 and ABI5 promoter DNA sequences relative to WT plants (Figure 5). In the absence of ABA, pkl mutant seeds had lower H3-K9 and H3-K27 methylation levels relative to WT plants, but less markedly (Figure 5; about 1.2- to 1.8-fold less in the pkl mutant, with no significant difference for the amount of H3-K27 associated with ABI3). In contrast, 7-day-old WT and pkl seedlings displayed mostly no difference in H3-K9 and H3-K27 association with ABI DNA (Figure 5). Similar results were obtained using UBIQUITIN (UBQ) as a internal reference to normalize ABI DNA amounts in ChIP precipitates, although the differences between ABA-treated WT and pkl seeds for the H3-K9 marker were less pronounced (Figure S5) and variability between seed batches was more apparent. The higher variability may be due to lower UBQ association with silent chromatin (as UBQ is expressed in all cells), which would yield less robust and reproducible signals in PCR amplifications.

image

Figure 5.  Sustained ABI3 and ABI5 expression in pkl-1 mutants under low ABA conditions correlates with lower methylation levels in histone markers for repressed chromatin. Fold increases in WT histone H3 methylation levels relative to pkl were obtained by dividing the normalized DNA amounts (using Ta2 as a reference gene) recovered in ChIP experiments with WT plant material by those recovered in ChIP experiments with pkl plant material. ChIP experiments were performed on plant material treated with 0.25 μm ABA harvested after seed stratification, as indicated (see Experimental procedures). DS, dry seeds.

Download figure to PowerPoint

These results indicate that PKL exerts an influence over the chromatin associated with ABI genes by reducing its repressive markers, particularly in ABA-treated seeds. In turn, the lower ABI5 association with repressed chromatin found in ABA-treated pkl seeds is consistent with the exacerbated ABI expression in response to ABA (Figure 3). These data are consistent with the previously proposed notion that PKL plays a specific role in repressing embryonic gene expression during germination. For example, Rider et al. (2004) and Henderson et al. (2004) have shown that LEC1 expression in pkl mutants mainly occurs during germination and is not elevated in leaves despite PKL expression in this tissue (see also Figure S3).

We also explored PKL-dependent DNA methylation events as additional markers for repressed chromatin. No DNA methylation events could be identified in ABI3 or ABI5 sequences using methylation-sensitive restriction enzymes or sodium bisulfite sequencing (data not shown, see Experimental procedures). On the other hand, we found PKL-dependent DNA methylation events in LEC1 promoter sequences (see Figure S4 and Discussion), consistent with the PKL-dependent repression of LEC1 during germination (Ogas et al., 1999).

Taken together, our ChIP results suggest that PKL is necessary to maintain ABI3 and ABI5 chromatin in a repressed state on ABA during germination.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

A mutant lacking expression of the putative chromatin-remodeling factor-encoding gene PKL has been shown to abnormally express early-embryonic genes during seed germination (e.g. LEC1, FUS3) as well as in the adult plant (e.g. PHERES1; Li et al., 2005; Ogas et al., 1999; Rider et al., 2004).

Here we extended the study of PKL-dependent repression of embryogenesis gene expression by examining late-embryogenesis genes whose expression is ABA-dependent during germination. We focused on ABI3 and ABI5 due to their role in establishing protective growth arrest (Lopez-Molina et al., 2001, 2002).

During seed germination of pkl mutants in the absence of ABA, the rapid decreases in ABI3 and ABI5 expression seen in WT seeds were delayed by about 1 day (Figure 3), suggesting that PKL influences repression of their expression at that time. This delay was not associated with slower germination, presumably because ABI3 and ABI5 growth inhibitory activity requires ABA (Lopez-Molina et al., 2001, 2002). De-repressed expression of ABI3 and ABI5 was best revealed by exposing pkl mutant seeds to low ABA concentrations, as high concentrations saturate ABI3 and ABI5 expression levels and totally prevent germination. Low ABA concentrations prolonged the expression of ABI3 and ABI5 by several days in pkl mutant seeds. Prolonged ABI5 expression led to exarcerbated germination delays (Figure 2). Thus, the influence of PKL on repression of embryogenesis gene expression also extends to ABI3 and ABI5 expression during the time interval when they are responsive to ABA (i.e. during germination). Indeed, their expression is no longer induced by ABA in pkl mutant green tissues, as in WT (data not shown). This suggests that the process preventing ABI3 and ABI5 induction by ABA in green tissues does not involve PKL, which is consistent with specific de-repression of LEC1 and FUS3 during germination in pkl plants (Figure S3).

The molecular basis of PKL function is suggested by its sequence, which reveals a chromodomain, a SNF2-related helicase/ATPase domain, a DNA binding domain and a PHD zinc finger (Ogas et al., 1999), domains that are typically found in chromatin-remodeling factors. Consistent with this hypothesis, persistence of ABI3 and ABI5 expression in the pkl mutant correlated with lower methylation levels in histone markers for silent chromatin (Figure 5). This could suggest that PKL acts directly on ABI3 and ABI5 to promote formation of repressed chromatin. How PKL exerts its influence on chromatin remains an open issue, and the observed changes in histone and DNA methylation may result from general changes in chromatin caused by the pkl mutation. Our ChIP studies are restricted to small DNA elements upstream of the transcription start site of ABI genes, which raises the question of the specificity of the reported differences. Recent genome-wide studies have shown that H3-K27 and H3-K9 methylation is found in chromatin of silent genes in regions extending −10 kbp and +10 kbp from the transcription start site (Barski et al., 2007). This indicates that the influence of PKL on the state of chromatin associated with ABI genes may extend to promoter, genic and 3′ genomic regions as well. Furthermore, the absence of notable differences between WT and pkl chromatin in 7-day-old seedlings, a time when PKL does not play an essential role in ABI expression as ABA-responsiveness has disappeared, supports the notion that the DNA elements used in our ChIP studies provide information on chromatin events associated with ABI genes, which are under the influence of PKL.

Previous studies suggested that various chromatin-remodeling complexes can also influence the expression of genes de-repressed in pkl mutants. For example, MEDEA and CURLY LEAF/SWINGER are embryonic PcG proteins inhibiting FUS3 expression. MEDEA is also necessary to inhibit PHERES1 expression during embryogenesis, while PKL inhibits its expression in adult plants (Kohler et al., 2005; Li et al., 2005; Makarevich et al., 2006). We therefore speculate that PKL operates with distinct chromatin-remodeling complexes during germination. In the case of ABI3 and ABI5, lower H3-K27 methylation levels in pkl mutants indicate that PKL exerts its repressive activities with unidentified PcG proteins. In addition, lower H3-K9 methylation in pkl mutants may indicate that PKL recruits an H3-K9 methyltransferase such as SUVH4/KYP or its homologues (Ebbs and Bender, 2006).

Thus, our work indicates that ABA (i.e. stress) in a germinating seed increases ABI3 and ABI5 gene expression, while the general influence of PKL is to limit the transcription potential of embryogenesis genes, probably by promoting silencing of their chromatin. These apparently contradictory influences on ABI gene expression probably reflect the developmental transition taking place during germination. ABI gene expression can still be stimulated upon stress prior to autotrophy, while general embryogenesis gene expression becomes permanently silenced.

However, the ABI3 and ABI5 mRNA and protein over-accumulation observed in pkl plants may not necessarily originate from the same cells expressing ABI3 and ABI5 in WT germinating seeds. Indeed, the germinating embryo consists of multiple types of cells in various stages of differentiation, some of which may not significantly express ABI3 and ABI5. Thus, ABI3 and ABI5 de-repressed expression in pkl seeds may result from certain cells in which vegetative differentiation proceeds normally while embryogenesis gene expression is effectively repressed by PKL. Evidence that ABI5 is not expressed in all the cells of a germinating embryo comes from studies with reporter transgenes bearing ABI5 promoter sequences fused to GUS (Lopez-Molina et al., 2002). Whereas blue staining is observed in all cells of dry seed embryos (Lopez-Molina et al., 2002), only the radicle tip is stained in germinating seeds, irrespective of the presence or absence of ABA (see Figure S5A). Similar studies apparently contradict our findings. Indeed, Penfield et al, 2006. used promoter ABI5 sequences fused to GUS and revealed blue staining in all cells even in germinating embryos (Brocard et al., 2002; Penfield et al., 2006). However, their findings coincide with ours in that GUS activity is strongly diminished upon imbibition (Brocard et al., 2002). Thus, in an embryo where GUS expression, driven by ABI5 promoter, diminishes rapidly, the timing for blue staining and duration of the GUS enzymatic assay may explain the discrepancies in blue staining throughout the embryo.

Consistent with this hypothesis, a few observations indicate that PKL does indeed not exert repressive activities in all embryonic cells. Firstly, our DNA methylation studies show that only a subset of LEC1 promoter DNA, and therefore of cells, is methylated (see Figure S4). In these cells, the role of PKL in maintaining DNA methylation is essential, given that no DNA methylation is observed in pkl mutants. This could indicate that LEC1 repression through chromatin in a repressed state takes place in a subset of cells. Secondly, and consistently with the latter hypothesis, pkl plants transformed with LEC1 promoter DNA sequences fused to the GUS reporter gene show staining predominantly in the radicle 24 h after imbibition (see Figure S5B). WT plants transformed with the same construct did not display significant expression (Figure S5B). Taken together, these data indicate that de-repression of embryonic gene expression in pkl mutant plants takes place in a subset of embryonic cells.

Finally, the ABA-dependent growth arrest mediated by late-embryonic genes during germination occurs while the embryo transits towards its autotrophic green phase and must be understood in this larger context (Lopez-Molina et al., 2002). When fully implemented, the arrest program must completely counteract GA signaling, which promotes differentiation towards vegetative cells while repressing their embryonic character. When endogenous GA synthesis is diminished during imbibition, using specific inhibitors, germination is delayed. This delay is sharply enhanced in pkl mutant seeds, suggesting that PKL activity normally contributes in some way to promoting GA signaling (Henderson et al., 2004; L.L.-M., unpublished results). Stimulating the GA pathway could be achieved by preventing its repression, possibly involving RGL2, a GA-response negative regulator of seed germination (Lee et al., 2002). In a complex situation of hormonal competition, the activity of PKL in attenuating ABA responses and potentiating GA responses may reflect its integrative role: maintaining acceleration and retardation of growth in balance by modulating the association of key genes with repressed chromatin.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant material

Seeds and tissues from Arabidopsis thaliana ecotype Columbia (Col-0), the pkl-1 mutant (in the Col-0 background; kindly provided by J. Ogas) and the abi5-4 mutant (Wassilewskija background, Lopez-Molina and Chua, 2000) were used in all experiments. Seeds were surface-sterilized and sown on agar plates containing Murashige and Skoog (MS) salt solution supplemented or not with ABA as described previously (Lopez-Molina and Chua, 2000). Seeds were stratified by 3 days incubation at 4°C in the dark, and then transferred to 22°C with a 16-h photoperiod.

ABA responses

Endosperm rupture (defined as ‘germination’) was scored 72 h after seed stratification. Germination events are expressed as a percentage of the total number of seeds. The results are the means of two independent experiments using two different seed batches of WT and pkl plants grown in parallel. Chlorophyll measurements were performed on 50 4-day-old seedlings ground in 50 μl of 50 mm Tris, pH 8, with 1% SDS at room temperature. Aliquots (7 μl) of the resulting extract were mixed with 993 μl of 80% acetone (4°C) and centrifuged for 5 min at 18 000 g. The resulting supernatant was used for total chlorophyll quantification by measuring optical density at 652 nm. Total chlorophyll content (μg μl−1) was calculated as OD652 nm × 4.25. The results presented are the mean of two independent experiments.

Transgenic lines

The reporter gene β-glucuronidase (GUS) was fused in-frame to the second LEC1 ATG as shown in Figure S4 by amplifying a 2.85-kbp fragment of LEC1 upstream sequences using the primers 5′-CGGTCGACAATTAGAAATGGTTGCTCAGAAATCAG-3′ and 5′-CGGGATCCTGGGGTTGGGTTGTTGCTGTGCTGG-3′, and cloning it into the BamHI and SalI sites of the GUS-containing binary vector PBI101 (Clontech, http://www.clontech.com/). The plasmid was introduced into Agrobacterium tumefaciens strain GV3101, and 3-week-old Col-0 and pkl-1 plants were transformed using the floral dipping method (Clough and Bent, 1998).

β-glucuronidase assay

Embryos excised from dry seeds as well as 2- and 5-day-old seedlings were incubated for 3 h at 37°C in the following buffer: 10 mm Na2EDTA, 0.1 mm KFe(CN)2, 50 mm phosphate buffer, pH 7, and 2 mm X-Gluc. The material was fixed in 5% formaldehyde, 5% acetic acid, 45% ethanol for 1 h, and washed several times in 70% ethanol. Tissues were examined using a Zeiss STEMI 2000-C stereomicroscope (http://www.zeiss.com/).

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) experiments were performed during germination at the earliest time possible, which coincided with full testa rupture (thus avoiding formaldehyde cross-linking difficulties). Testa rupture in WT and pkl mutant seeds is complete 24 and 32 h, respectively, after stratification in the absence and presence of ABA (0.25 μm). Chromatin samples were immunoprecipitated [using dimethylated H3-K9 (catalog no. 07-441) and H3-K27 (catalog no. 07-452), Upstate Biotechnologies; http://www.upstate.com) and ChIP experiments were performed as described previously (Bowler et al., 2004). The immunoprecipitated DNA was purified using a silica-gel membrane (Qiagen, http://www.qiagen.com/) and analyzed by quantitative real-time PCR. Aliquots (5 μl) of purified DNA were used in 20 μl real-time PCR reactions using a Bio-Rad ICycler IQ (http://www.bio-rad.com/) and Absolute QPCR SYBR Green Mix (ABGene; http://www.abgene.com). PCR primers were designed to amplify 100 bp fragments. The primers pairs used were as follows: 5′-TTGGGTCATCATTATATCCGTCT-3′ and 5′-GGTCTTCGTTTTCTGGTCCTT-3′ (ABI5 promoter), 5′-GTTTAAGAACCACCGCTTGG-3′ and 5′-CTCCTCGTGCCGCTAGTATC-3′ (ABI3 promoter), 5′-CTAAACGCATCGATTTCATCATG-3′ and 5′-AGGTCCCATCATCGGGCT-3′ (Ta2 retrotransposon). The amounts of DNA were derived from cycle threshold (Ct) values using the standard curve method [(DNA) = 10 × (C0Ct)/a, where C0 and a are determined experimentally]. The amounts of DNA obtained in WT and pkl using the ABI3 and ABI5 primers were normalized to that for Ta2, a putative transposable element at peri-centromeric sequences associated with silent chromatin. DNA methylation patterns in centromeric DNA repeats were unaffected in pkl mutant plants, further indicating that PKL does not exert a significant influence on heterochromatin (Figure S3; Mathieu et al., 2005). Ta2 was highly associated with dimethylated H3-K9 relative to dimethylated H3-K27 (Mathieu et al., 2005) and its chromatin is unaffected in pkl plants. Further evidence that Ta2 chromatin is unaffected in pkl came from observations in ChIP experiments starting with an equal amount of WT and pkl material. For example, ChIP on material obtained from 2 g of seeds isolated 24 h after seed stratification yielded the following Ct values for Ta2: H3-K9, Ct (WT) = 21.85 ([DNA] = 1.99), Ct (pkl) = 21.8 ([DNA] = 2.06); H3-K27, Ct (WT) = 27.8 ([DNA] = 0.031), H3-K27, Ct (pkl) = 27.8 ([DNA] = 0.031). A ChIP experiment that measured Ta2 DNA amounts associated with H3-K9 in WT (Zue ecotype) and ddm1 mutants (devoid of heterochromatin; Mathieu et al., 2005) was also performed. WT material yielded DNA amounts two orders of magnitude higher than ddm1 mutants; for example, Ct (WT) = 24.7 ([DNA] = 0.26) and Ct (ddm1) = 32.3 ([DNA] = 0.0011). ChIP experiments lacking antibodies failed to amplify any sequence significantly. All ChIP results were calculated from the mean of three independent experiments, with PCR reactions performed at least twice. The significance of the results was evaluated using the t-test (< 0.05).

Southern blot analysis

Southern blots of genomic DNA from Col-0 and pkl-1 germinating seeds were performed according to standard procedures. Aliquots (5 μg) of genomic DNA were digested by enzymes that are not affected by DNA methylation (HindIII), are impaired by CpG methylation (HpyCH4IV, HpaII, AciI, BstBI), or are impaired by CpNpG methylation (MspI, Fnu4HI, Hpy188III, Sau96I; New England Biolabs; http://www.neb.com). Blots were hybridized with various 32P-labeled probes encompassing ABI3, ABI5, LEC1, and FUS3 genes prepared from amplicons of Col-0 genomic DNA. For the LEC1 promoter probe, the following primers were used: 5′-CATGCACGGGTATATTCAATGAACAC-3′ and 5′-AAACGTGATAACTCACTTTCGTAACG-3′.

Sodium bisulfite sequencing

Aliquots (2 μg) of Col-0 and pkl-1 genomic DNA were digested with SpeI and NheI, which cut outside of the region of interest. Sodium bisulfite treatments were performed as described previously (Jacobsen et al., 2000). Aliquots (5 μl) of the bisulfite-treated DNA were used for PCR reactions with the following primers: 5′-TAATATTTAATTGTAATATGTATTATT-3′ and 5′-AATAAAAAAAAAATTTTTATACATATAAA-3′. PCR products were purified using a silica-gel membrane (Qiagen) and sequenced directly using the primers used for PCR amplification.

Northern and Western blot analysis

Total RNA from Arabidopsis was isolated as described previously (Vicient and Delseny, 1999). Blots were hybridized with ABI3 and ABI5 full-length ORF probes according to standard procedures. rRNA was stained with methylene blue. Western blot analysis was performed as previously described (Lopez-Molina et al., 2001, 2002).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We are especially grateful to Pierre Vassalli for his help and support at all the stages of the work presented here. We thank Jerzy Paszkowski and Ueli Schibler for advice and criticism. We thank Joe Ogas (Université de Lausanne, Lausanne, Switzerland) for providing pkl-1 seeds, and Hans Reinke for advice on real-time quantitative PCR. Caroline Gutjahr, Urszula Piskurewicz, Christophe Belin and Jean-David Rochaix provided critical comments on the manuscript and numerous suggestions. We thank Anne Vassalli (Purdue University, Indiana, USA) for valued improvement to the manuscript. This work was supported by grants from the Swiss National Science Foundation and the US Department of Agriculture to L.L.-M. and by the State of Geneva.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Figure 1 Red Ponceau staining and coomassie blue staning of the protein samples used for western blots shown in Figure 3

Figure 2 A: PKL-dependent DNA methylation in the LEC1 promoter region. LEC1 gene structure (NM_102046) and methylation sites are indicated. The location of the probe used for Southern blot analysis is indicated as well as the corresponding hybridization bands (“a”, “b”, “c” and “d”).

B: Sodium bisulfite sequencing of the LEC1 promoter region. DNA was isolated from seeds 24 h after stratification. The sequenced region is indicated in A and the sequencing results shown in B. PKL-dependent methylated cytosines are indicated in red. The HpyCh4IV (ACGT) restriction sites are underlined.

Figure 3 Centromeric DNA methylation is not affected in pkl mutants. WT and pkl genomic DNA was digested with MspI (CCGG) and AciI (CCGC), enabling detection of CpNpG and CpG methylation, respectively. Southern blot analysis was performed using a 180-bp centromeric repeat probe

Figure 4 A: Promoter-ABI5::GUS blue staining 24 hours after seed imbibition ((Lopez-Molina et al. 2002).

B: Promoter-LEC1::GUS blue staining 24 hours after seed imbibition (WT and pkl mutant background)

Figure 5 Same experiment as in Figure 5 except that Ubiquitin is used for normalization.

Figure 6LEC1 and FUS3 transient mRNA expression during germination in stratified and non-stratified seeds. Northern blots were performed as described in Figs 3.

Figure 735S::HA-ABI5 plants display hypersensitive germination responses to ABA. Effect of exogenous ABA on seed germination. WT (Ler) and 35S::HA-ABI5 (Ler ecotype, Lopez-Molina et al. 2001), seeds were grown with or without ABA. Germination was scored 48 h after stratification with two independent seed batches.

Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
TPJ_3288_sm_figs.pdf5215KSupporting info item
TPJ_3288_sm_legends.doc28KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.