In Posidonia oceanica cadmium induces changes in DNA methylation and chromatin patterning.

In mammals, cadmium is widely considered as a non-genotoxic carcinogen acting through a methylation-dependent epigenetic mechanism. Here, the effects of Cd treatment on the DNA methylation patten are examined together with its effect on chromatin reconfiguration in Posidonia oceanica. DNA methylation level and pattern were analysed in actively growing organs, under short- (6 h) and long- (2 d or 4 d) term and low (10 μM) and high (50 μM) doses of Cd, through a Methylation-Sensitive Amplification Polymorphism technique and an immunocytological approach, respectively. The expression of one member of the CHROMOMETHYLASE (CMT) family, a DNA methyltransferase, was also assessed by qRT-PCR. Nuclear chromatin ultrastructure was investigated by transmission electron microscopy. Cd treatment induced a DNA hypermethylation, as well as an up-regulation of CMT, indicating that de novo methylation did indeed occur. Moreover, a high dose of Cd led to a progressive heterochromatinization of interphase nuclei and apoptotic figures were also observed after long-term treatment. The data demonstrate that Cd perturbs the DNA methylation status through the involvement of a specific methyltransferase. Such changes are linked to nuclear chromatin reconfiguration likely to establish a new balance of expressed/repressed chromatin. Overall, the data show an epigenetic basis to the mechanism underlying Cd toxicity in plants.

Introduction

In the Mediterranean coastal ecosystem, the endemic seagrass Posidonia oceanica (L.) Delile plays a relevant role by ensuring primary production, water oxygenation and provides niches for some animals, besides counteracting coastal erosion through its widespread meadows (Ott, 1980; Piazzi ; Alcoverro ). There is also considerable evidence that P. oceanica plants are able to absorb and accumulate metals from sediments (Sanchiz ; Pergent-Martini, 1998; Maserti ) thus influencing metal bioavailability in the marine ecosystem. For this reason, this seagrass is widely considered to be a metal bioindicator species (Maserti ; Pergent ; Lafabrie ). Cd is one of most widespread heavy metals in both terrestrial and marine environments.

Although not essential for plant growth, in terrestrial plants, Cd is readily absorbed by roots and translocated into aerial organs while, in acquatic plants, it is directly taken up by leaves. In plants, Cd absorption induces complex changes at the genetic, biochemical and physiological levels which ultimately account for its toxicity (Valle and Ulmer, 1972; Sanit‡ di Toppi and Gabrielli, 1999; Benavides ; Weber ; Liu ). The most obvious symptom of Cd toxicity is a reduction in plant growth due to an inhibition of photosynthesis, respiration, and nitrogen metabolism, as well as a reduction in water and mineral uptake (Ouzonidou ; Perfus-Barbeoch ; Shukla ; Sobkowiak and Deckert, 2003).

At the genetic level, in both animals and plants, Cd can induce chromosomal aberrations, abnormalities in chromatin structure, and nuclear apoptotic changes (Zhang and Xiao, 1998; Foitova and Kovarik, 2000; Pulido and Parrish, 2003; Banfalvi ; Deckert, 2005). Moreover, in mammals, Cd is widely recognized as a carcinogen (International Agency for Research on Cancer, 1993; National Toxicology Program, 2000). Although the mechanisms underlying such effects need to be further investigated, the mechanism of Cd may well have a non-genotoxic (Goering ; Waalkes and Misra, 1996; Waisberg ) or epigenetic basis involving DNA methylation (Benbrahim-Tallaa ; Jiang ). In mammals, cytosine DNA methylation occurs almost exclusively in the symmetric CG context (Ehrlich, 1982) while, in plants, it occurs in all sequence contexts: symmetric CG and CHG (in which H=A, T or C) and asymmetric CHH (Henderson and Jacobsen, 2007). In plants, three different families of DNA methyltransferases with distinct substrate specificities and different modes of action account for DNA methylation. Firstly, DNA METHYLTRANSFERASE 1 (MET1, also known as DMT1), whose members act as maintenance methyltransferases, introduces methyl groups specifically into CG sequences (Finnegan and Kovac, 2000). Secondly, DOMAINS REARRANGED METHYLTRANSFERASES (DRMs), a homologue of the mammalian DNA methyltransferase 3 (DNMT3) family, maintains asymmetric CHH methylation through persistent de novo methylation (Cao and Jacobsen, 2002a, b; Wada ).

Thirdly, CHROMOMETHYLASES (CMTs), a plant-specific DNA methyltransferase family, are involved primarily in the maintenance of symmetrical CHG methylation (Lindroth ; Papa ). Furthermore, CMTs also play a role in de novo methylation (Cao ; Chan ). CMT is characterized by a chromatin-associated domain (chromodomain) embedded within the catalytic motifs I and IV of the protein (Henikoff and Comai, 1998). The presence of this plant-specific methyltransferase provides an explanation for the high levels of CHG methylation in plant genomes relative to animals. In Arabidopsis, the targeting of AtCMT3 methylation is accomplished by short interfering RNA (siRNA) pathways (Zilberman ; Chan , 2006) and histone methylation H3K9, H3K27 (Jackson ; Lindroth, 2004). Interestingly, in Arabidopsis, the establishment of heterochromatic nuclear domains, that represent densely methylated areas, are marked by the occurrence of both of these histone modifications (Fischer ).

DNA methylation, together with other epigenetic marks, regulates gene transcriptional activity in response to both endogenous factors and external stimuli, therefore playing a significant role in plant development (Finnegan ; Dennis ; Bitonti ; Fojtova ). Accordingly, in several species, stressful conditions, which are known to affect plant growth and development strongly, are associated with variations in genome methylation. In particular, altered DNA methylation (i.e. hypermethylation and hypomethylation), can occur in relation to low temperature, osmotic stress and drought but also by exposure to heavy metals (Finnegan ; Labra ; Aina ; Dyachenko ; Hashida ; Boyko ; Choi and Sano, 2007). Regarding Cd, while, in mammals, Cd-induced carcinogenesis was associated with a global DNA hypermethylation and over-expression of maintainance/de novo DNA methyltransferases (Takiguchi ; Benbrahim-Tallaa ; Jiang ), in plants, the extent of DNA methylation in Cd toxicity is limited to one species, Raphanus sativus L. (Yang ).

The aim of the work reported here was to determine the influence of short- and long-term and low- and high-dose cadmium treatments (CdCl2 at 10 μM or 50 μM for 6 h, 2 d or 4 d) on the DNA methylation status of P. oceanica plants. Firstly, the DNA methylation level and the nuclear distribution of DNA methylated sites were estimated in actively growing organs (i.e. apical tips and young leaves) through Methylation-Sensitive Amplification Polymorphisms (MSAPs) (Xiong ; Xu ) and immunocytolabelling, respectively. Secondly, the transcript level of one member of the CMT family was analysed due to its involvement in both maintenance and de novo methylation activity. In addition, regarding the effects of Cd-heavy metal on chromatin reconfiguration (Banfalvi ; Ma ), nuclear chromatin ultrastructure was also analysed in the shoot apical meristem of Cd-treated plants. A Cd-induced DNA hypermethylation and an up-regulation of CMT expression associated with a specific nuclear pattern of DNA methylation was observed as well as progressive changes in chromatin structure.

Materials and methods

Plant material

Posidonia oceanica (L.) Delile (2n=2x=20) has a 2C-value of 6.25 pg DNA (den Hartog ; Kuo ; Dolenc Koce et al., 2003). Individual shoots of P. oceanica were randomly sampled by SCUBA diving from preserved meadows (Acunto ) of Calabria (San Nicola Arcella, 39°48′;45.71'' N, 15°47′53.74'' E), Southern Italy. Shoots were collected about every 10 m along linear transects to obviate sampling within the same clonal patch, then transferred to the laboratory and acclimated for 1 week in three different aquaria, each containing 100 l of seawater, pH 7.8, 16.5±0.5°C, a 16/8 h light/dark cycle, and ensuring water flow-through and fluxing oxygen throughout. Thereafter, one aquarium was maintained in the same conditions as the control, while, in the second and third aquaria, CdCl2 was added to reach final concentrations of 10 μM or 50 μM of Cd, respectively. From each aquarium, about 20–25 individual shoots were collected at 6 h, 2 d or 4 d. Analyses were performed on apical tips (which includes the shoot apical meristem, leaf primordium, and young leaflets) and leaf (which includes short leaves (<5 cm) and the basal region of intermediate leaves (5 cm

Markers of the Cd treatment: chlorophyll content

Chlorophyll a (Chl a) and chlorophyll b (Chl b) content was evaluated in Cd-treated (Cd-tP at 10 μM or 50 μM for 6 h, 2 d or 4 d) and control plants (CP for 6 h, 2 d or 4 d). Excised leaves (0.5 g fresh weight) were immediately ground in a mortar with liquid nitrogen and immersed in 80% acetone 1:5 (w/v) for 3 h in darkness, followed by centrifugation at 3000 rpm for 15 min at 4 °C and the supernatant was recovered. Chlorophyll amounts were determined using a spectrophotometer (model Cary 50Bio, Varian, Turin, Italy). A646.8 and A663.2 were determined and used for calculation of the total contents of Chl a and b according to Lichtenthaler (1987). Three replicates were performed and, for each replicate, six measurement were carried out on each sample. Statistical analysis was performed using a one-way ANOVA followed by the Neuman–Keul’s post-hoc test.

Quantitative real-time PCR (qRT-PCR) of an important metal tolerance gene: Posidonia oceanica Metallothionein 2k (PoMT2k)

Total RNA isolation and reverse transcription:

Total RNA was extracted from apical tips and leaves of Cd-tP (10 μM or 50 μM for 6 h, 2 d or 4 d) and CP (6 h, 2 d or 4 d), according to Bruno . RNA was resuspended in RNase-free water and treated with DNase I (Roche Diagnostic Mannheim, Germany) for 15 min at 37 °C. Quality and quantity of RNA was verified using a NanoDrop® spectrophotometer ND-1000, the integrity was checked by agarose 0.8% gel electrophoresis. About 2–3 μg of total RNA were retro-transcribed using a First Strand cDNA Synthesis Kit (Fermentas, Milan, Italy) according to the manufacturer’s instructions.

Primer design

For PoMT2k (AJ628145) mRNA level analysis, specific oligonucleotide primers were designed, using PRIMER3 software (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi, accessed 11 January 2006) according to the strategies set up by Yokoyama and Nishitani (2001). PoMT2k primer sequences are: PoMT-FW (5′-ATGGAGAAGAGCACCACCAC-3′) and PoMT-BW (5′-TAGCAGTTGCAGGGATCACA-3′). After checking independent trials of housekeeping genes, the Po5.8S rRNA gene (AJ225091) produced the most reproducible results across various cDNAs, and was used as a normalization control. The primer sequences are Po5.8S-FW (5′-TCCGAACACTTCCGGTACAT-3′) and Po5.8SBW (5′-CAATCAACGCAATTCACCAC-3′). Each primer pair used was designed to obtain a final PCR product of about 150 bp length, and was tested according to different parameters: (i) robustness, successful amplification over a range of annealing temperatures; (ii) specificity, the generation of a single significant peak in the melting curve; and (iii) the consistency of highly reproducible CT values within the reactions of a triplicate. The average efficiency of the primer pairs used ranged between 0.95 and 1.0.

Amplification conditions

qRT-PCR amplification conditions were performed on a Bio-Rad Miniopticon (Bio-Rad, Milan, Italy) Single colour thermocycler with Bio-Rad SYBR Green Supermix (Cat. No.170–8884), according to Bruno .

Data analysis

The results of qRT-PCR were analysed using an Opticon Monitor: quantification real-time PCR Detection System (Bio-Rad), a program that facilitates the analysis of the kinetics of each performed reaction. Cycle threshold (CT) values were obtained with the Genex software (Bio-Rad) and data were analysed with the 2−ΔΔCTmethod (Livak and Schmittgen, 2001). The mean of PoMT2k expression levels was calculated from three biological repeats, obtained from three independent experiments.

Methylation-Sensitive Amplification Polymorphism (MSAP) analysis

MSAP analysis was performed according to Xu . Total genomic DNA was extracted according to Serra from apical tips and leaves of Cd-tP (10 μM or 50 μM for 6 h, 2 d or 4 d) and CP (6 h, 2 d or 4 d). Digestion of RNA was performed by incubating the DNA with 0.1 mg ml−1 RNase A (Roche Diagnostic Mannheim, Germany) for 60 min at 37 °C. The quality and quantity of DNA was verified using a NanoDrop® spectrophotometer ND-1000, the integrity was checked by agarose 0.8% gel electrophoresis.

MSAP analysis is a modified Amplified Fragment Length Polymorphism (AFLP) technique (Vos, 1995) comprising three major parts.

(i) Digestion and ligation reactions:

DNA samples were digested with MspI and HpaII (Promega, Milan, Italy), a methylation-sensitive restriction enzyme used instead of MseI as a ‘frequent cutter’ enzyme. In particular, about 1 μg of genomic DNA was double digested O.N. at 37 °C with EcoRI–MspI or EcoRI–HpaII (7 U each) in a final volume of 45 μl containing Multi-core buffer (Promega). The digested DNA fragments, from the two reactions, were ligated separately with an equal volume of the adapter/ligation solution at 16 °C overnight, containing either the MspI–HpaII adapter (50 pmol) or the EcoRI adapter (5 pmol) and 1.5 units of T4 DNA ligase (Promega) in a final volume of 50 μl.

The adapters (EcoRI and HpaII–MspI) used in the MSAP analysis are shown in Supplementary Table S1 at JXB online and were prepared according to Xu by mixing equimolar amounts of the two DNA strands, keeping the mixture at 65 °C for 10 min, and then cooling it down to room temperature.

(ii) Pre-amplification and selective amplification reaction:

After ligation, the reaction mixture was diluted 5-fold to 250 μl with 10 mM TRIS-HCl, 0.1 mM EDTA pH 8.0 (TE) and the primers therein used as templates for the pre-selective amplification. PCR reactions (Primus 96 Plus Thermal Cyclers MWG AG Biotech) were performed for 20 cycles with the following cycle profile: a 45 s denaturation step at 94 °C, a 1 min annealing step at 54 °C, and a 1 min extension step at 72 °C. After a final elongation step (10 min at 72 °C) the pre-amplificated product was diluted 15-fold with TE and further used as template for the selective amplification, according to classical AFLP cycling parameters (Vos, 1995). Primers used in the pre-selective and selective amplification have core sequences complementary to the EcoRI and HpaII/MspI adapters, with selective 3' nucleotide (see Supplementary Table S1 at JXB online). In particular, four and 24 different primer combinations were used for pre-selective and selective amplification, respectively, which are sufficient to give an optimum number of scoreable polymorphic fragments per P. oceanica genome size (see Supplementary Table S2 at JXB online).

(iii) Detection reactions:

PCR products were separated electrophoretically as follows: 5 μl of formamide loading buffer (90% formamide, 10 mM EDTA pH 8.0, and bromophenol blue and xylene cyanol as tracking dyes) added to 5 μl of the MSAP reaction mixture, were denatured for 2 min at 95 °C, and then cooled rapidly on ice. Samples were loaded onto a 5% denaturing polyacrylamide Criterion Precast Gel (Bio-Rad, Milan, Italy) and electrophoresed in TBE electrophoresis buffer for 3 h at 200 V. Gels were pre-run at 150 V for 1 h in a Bio-Rad Vertical mini electrophoresis apparatus (Criterium cell, Bio-Rad). The products were silver-stained as described in Qu , and fingerprint patterns were visualized using a phosphoimage analysis system, GS-700 Imaging Densitometer (Bio-Rad).

Each amplified band represents a recognition site, cleaved by one or both of the methylation-sensitive restriction endonucleases. The methylation levels were deduced by counting the reproducible polymorphic bands produced within each DNA by the two different enzymes, and were calculated from three biological repeats, obtained from three independent experiments.

Characterization of polymorphic bands

Polymorphic amplified fragments were excised from the gel, eluted with QUIAEX II Gel extraction kit (Qiagen, Hilden, Germany) and then amplified with the same primers under the conditions used for selective amplification. Sequence information was obtained by cloning a cluster of polymorphic fragments in the pGEM-T easy vector system according to the manufacturer’s instructions (Promega, Milan, Italy) and sequencing individual clones (Genelab ENEA, Rome, Italy). The sequences obtained were compared with nucleotide sequences in the publicly available databases using BLAST (http://www.arabidopsis.org/Blast/index.jsp; http://blast.ncbi.nlm.nih.gov/Blast.cgi).

5-methylcytidine-immunocytolabelling

For 5-methylcytidine immunolabelling, apical tips (n=10 for each sample) of Cd-tP (10 μM or 50 μM for 6 h, 2 d or 4 d) and CP (6 h, 2 d or 4 d) were excised and fixed in 0.5% (v/v) gluteraldehyde and 3% (w/v) paraformaldehyde mixture in PBS (135 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, 8 mM K2HPO4, pH 7.2). Samples were dehydrated and embedded in Technovit 8100 resin. Sections were cut at 3 μm with a tungsten knife using a Leica 2155 microtome and mounted on slides. Immunolabelling was applied as described in Chiappetta . A labelling index (L.I.) of DNA methylation was expressed as number of silver grains/nuclear area unit and statistical differences were evaluated by ANOVA followed by the Neuman–Keuls post hoc test. The number and the pattern of distribution of silver grains per nucleus were estimated by scoring all the serial sections using a Leica DMRB microscope and a Leica Q500/W image analyser equipped with a CCD camera.

Posidonia oceanica CHROMOMETHYLASE (PoCMT) features

Isolation and sequence analysis of putative PoCMT:

A PCR-based strategy using degenerate primers was used to amplify cDNA obtained by the reverse transcription of total RNA from apical tips through a First Strand cDNA Synthesis Kit (Fermentas, Milan, Italy). The design of primers (see Supplementary Table S3 at JXB online) anchored to conserved structural motifs of the CMT family were based on Arabidopsis and other species for which sequences were available. The 5' and 3' ends of the transcripts were recovered through the 5'- and 3'- RACE System (Invitrogen, Milan, Italy) using the kit anchor primers provided by the manufacturer and the specific primers, FW3'-RACE and BW5'-RACE (see Supplementary Table S3 at JXB online). All PCR fragments were cloned into pGEM-T easy vector system (Promega, Milan, Italy), sequenced (Genelab ENEA, Rome, Italy), and confirmed to share high identity with higher plant CMTs. The complete coding region of the P. oceanica CMT gene was assembled and then verified by overlapping PCR with specific primers (see Supplementary Table S3 at JXB online) and sequenced in both directions.

PCR components were: cDNA (20 ng), 1.5 μM each primers, 0.5 mM dNTs, Taq DNA polymerase (Go Taq, Promega) 2.5 U, 1/10 of Taq buffer 10×, 2.5 mM MgCl2, in a final volume of 50 μl.

PCR conditions were: starting cycle at 95 °C for 3 min; 35 cycles at 95 °C for 60 s, 54 °C for 50 s and 72 °C for 60 s, with a final extension at 72 °C for 10 min.

Alignments and phylogenetic analysis:

The PoCMT amino acid sequence was aligned with other CMT proteins using ClustalW (http://www.ebi.ac.uk/clustalw) and optimized by visual inspection (PILEUP program). Identification of the characteristic structural motifs within the protein sequences was conducted using Protein Blast (http://blast.ncbi.nlm.nih.gov/), InterProScan (http://www.ebi.ac.uk/Tools/InterProScan/), and ExPASy-PROSITE (http://expasy.org/prosite/) (Falquet ).

Phylogenetic trees were constructed in MegaBlast4 (based on the minimum evolutionary criterion), using bootstrap values performed on 1000 replicates and the 50% value was accepted as indicative of a well-supported branch.

Southern blot analysis:

The probe used to determine the copy number of PoCMT was amplified from genomic P. oceanica DNA. For the CMT family, the primers targeted highly conserved motifs to maximize the likelihood of detecting other members of the gene family. The identity of the probe fragment was verified by cloning and sequencing. Ten μg of genomic DNA extracted from leaves was restricted O.N. at 37 °C with 60 U of either EcoRI, EcoRV and HindIII (Fermentas, Milan, Italy) which do not cut in the probe, in a 200 ml final volume. The digested DNA was precipitated at –20 °C O.N., resuspended in 50 ml volume and a 5 ml aliquot was rapidly checked by electrophoresis. DNA fragments were separated O.N. on an 0.8% agarose gel and transferred to a Hybond-N+ membrane (Amersham Pharmacia Biotech, Milan, Italy) by a Vacuum Blotter (Bio-Rad) according to the manufacturer’s specifications. Southern blots were performed with a digoxigenin-labelled probe obtained after amplification of gDNA with the PCR DIG Probe Synthesis Kit (Roche Diagnostic Mannheim, Germany). The digoxygenin-labelled probe spanning about 1.2 kb of DNA was obtained using the primer CMT.S-FW 5′-GCAGAGCACTTTCTGGCTCT-3′ and CMT.S-BW 5′-AAACGGTCTTAAAGACTTCCC-3'. Membrane prehybridization and hybridization were performed according to Bruno .

qRT-PCR of the PoCMT gene:

The expression level of PoCMT was tested in apical tips and leaves of Cd-tP (10 μM or 50 μM for 6 h, 2 d or 4 d) and CP (6 h, 2 d or 4 d). Specific oligonucleotide primers for PoCMT were designed using PRIMER3 software. Primer sequences are CMT.REAL-FW (5′-TGTAATCAATAGCAAAGTCTTCATGTC-3′) and CMT.REAL-BW (5′-AATGACCGATCAATCTTTTGC-3′). 5.8S rRNA (AJ225091) was chosen as a reference. Each primer pair used for qRT-PCR was designed to obtain a final PCR product of about 150 bp length and tested according to different parameters as described above. The average efficiency of all the primer pairs used ranged between 0.95 and 1.0. qRT-PCR amplification conditions were performed as mentioned before for PoMT2k expression analysis. Data analysis was performed according to Bruno .

Results

Markers of cadmium stress

As potential markers of Cd phytotoxicity, leaf chlorophyll (chl a/b) content, which is known to be negatively affected by Cd treatment (Stobart ; Shukla ; Ekmekci ) was estimated. PoMT2k expression levels were also measured (AJ628145) (Cobbett and Goldsbrough, 2002; Dal Corso Verbruggen ).

Leaf chlorophyll content is affected by Cd treatment

In leaves of control plants (CP), the value of Chl a/b did not change over time and hence the values at 6 h, 2 d and 4 d were pooled. However, a significant decrease in total Chl amount was detected in the leaves of Cd-tP compared with CP regardless of Cd concentration (10 μM or 50 μM). However, the effect of Cd on Chl a and Chl b was quite different. In particular, the reduction was higher for Chl a (65%) than Chl b (45%). In addition, for Chl a the strongest reduction (about 55% at 10 μM or 50 μM) was observed after short-term (6 h) Cd treatments, while Chl b progressively decreased from 6 h to 4 d of Cd exposure (see Supplementary Fig. S1 at JXB online). Therefore, the Chl a:b ratio ranged from 1.1 (6 h) to 1.4 (4 d) whatever treatment was applied.

Cd-treatment induces transcription of PoMT2k

PoMT2k expression level was analysed in both apical tips and leaves of P. oceanica. For CP, the expression level of PoMT2k was similar regardless of sampling time and hence the mean values for 6 h, 2 and 4 d were pooled. At 6 h, 2 d or 4 d Cd-10 μM treatment PoMT2k expression in the apical tips was not significantly different from the CP. However, Cd-50 μM treatment induced a significant increase in PoMT2k expression at 6 h and 2 d but it did not persist in the 4 d treatment (see Supplementary Fig. S2 at JXB online).

In leaves, PoMT2k expression at Cd-10 μM was not significantly different from CP at 6 h but significantly increased at 2 d and 4 d. However, once again 50 μM-Cd induced a significant increase in PoMT2k expression at 6 h and then a highly significant increase at both 2 d and 4 d (37-fold and 40- fold higher than in CP, respectively) (see Supplementary Fig. S2 at JXB online). Taken together, these results are consistent with a Cd-induced response in treated plants.

MSAP technique is a useful tool to assess DNA methylation status in P. oceanica

MSAP analysis allowed us to verify if changes in DNA methylation occurred in P. oceanica plants exposed to Cd stress and if these putative changes could target specific classes of genes. Based on the use of the restriction endonucleases HpaII–MspI that recognize the same tetranucleotide sequence (5′-CCGG-3′) but show differential sensitivity to cytosine methylation, specific signal-band profiles, which reflected the status and level of methylation of this specific sequence, were generated (see Supplementary Fig. S3 at JXB online).

Firstly, the efficiency of each primer combination in generating polymorphisms was evaluated, and specific 3′ extensions were the most efficient (see Supplementary Table S2 at JXB online).

According to Lu , four band-profile classes were defined, and were related to methylation status and enzyme sensitivity to site-specific methylation (Table 1). Namely, HpaII is inactive if both cytosines are fully methylated (two strand methylation), while it cleaves the hemi-methylated sequences (one strand methylation) at a lower rate than the unmethylated ones; MspI cuts in the case of inner cytosine methylation (5′- C 5mCGG-3′), but not in the case of outer cytosine methylation (5′-5mCCGG-3′).

Table 1.

Band-profiles classes defined on the basis of methylation sensitivity and restriction pattern of isoschizomers HpaII and MspI. HpaII and MspI recognize the same tetranucleotide sequence (5'-CCGG-3'), but display differential sensitivity to DNA methylation. Underlined cytosine is methylated.

Methylation status	MSAP band-profile methylation		Sensitivities to site-specific
	HpaII	MspI	HpaII	MspI
Class I	CCGG CCGG	Present	Present	Active
	GGCC GGCC
Class II	CCGG	Present	Absent	Inactive
	GGCC
Class III	CCGG	Absent	Present	Active
	GGCC
Class IV	CCGG	Absent	Absent	Inactive
	GGCC

In Cd-tP plants MSAP signal-band profiles changed in a time and dose- dependent manner

To quantify the differences in signal band profiles between CP and Cd-tP plants, the ratio between total and methylated bands (M/T (%)=methylation ratio %) and the ratio between total and fully methylated bands (FM/T (%)=fully methylation ratio %) was estimated (Table 2).

Table 2.

Summary of DNA methylation levels detected through MSAP analysis in Posidonia oceanica genomic DNA isolated from apical tips and leaves of control plants (CP) and Cd-tP (10 μM or 50 μM CdCl2) after 6 h, 2 d, and 4 d from the beginning of treatment. DNA methylation content was separately evaluated on apical tips and leaves and later on pooled to obtain the mean DNA methylation levels between the two tissues.

Polymorphic bands were scored as type I, II, III, and IV (as described in Table 1).

	Control plants				Cadmium 10 μM				Cadmium 50 μM
	6 h	2 d	4 d	Total	6 h	2 d	4 d	Total	6 h	2 d	4 d	Total
I	274	267	192	733	137	142	145	424	131	137	139	407
II	27	29	31	87	38	39	33	109	35	41	28	104
III	24	25	24	73	22	25	27	74	23	28	26	77
IV	3	4	7	14	2	5	2	9	2	2	3	7
T (I+II+III)	325	321	247	893	197	206	205	608	189	206	193	588
M (II+III+IV)	54	58	62	174	62	69	62	193	60	71	57	188
M/T (%)	16.6	18	25	19.8a	31.5	33.5	30.2	31.7a	31.7	34.5	29.5	32a
FM (III+IV)	27	29	31	87	24	30	29	83	25	30	29	84
FM/T (%)	8.3	9.0	12.5	9.9a	12.2	14.5	14.1	13.6a	13.2	14.5	15.0	14.3a

T, total amplified bands; M, total methylated bands; M/T (%), methylation ratio (%); FM, Fully methylated bands;

FM/T (%), fully methylation ratio (%).

Highlighted in grey are the highest value during the time of M/T (%) and FM/T (%) in CP and Cd-tP.

Representing a mean level of M/T (%) and FM/T (%), during the time, respectively.

Note that, although MSAP analysis was performed separately on apical tips and leaves, the results were pooled since both the methylation ratio % (M/T%) and the fully methylation ratio % (FM/T%) was similar between the two organs analysed and there was no significant differences in the level of 5mC between data sets.

Six hours from the beginning of treatment in aquarium conditions, the M/T was 16.6% in CP, while in Cd-tP exposed for the corresponding times at 10 μM-Cd and 50 μM-Cd the methylated ratio was 31.5% and 31.7%, respectively (Table 2). After the 2nd day of aquarium conditions, a slight increase of these ratios occurred in both CP (18%) and Cd-tP plants (33.5% at 10 μM and 34.5% at 50 μM). On the 4th day, these ratios decreased in Cd-tP (30.2% and 29.5% for 10 μM and 50 μM) and increased in CP plants (25%), while maintaining higher values in the former compared with the latter (Table 2).

A different trend was observed for fully methylated ratios, (FM/T%) which constantly increased during the analysed period in both Cd-tP, especially at 50 μM, and in CP plants. However, the values were significantly higher in Cd-tP than in CP plants, whatever treatment duration was considered (Table 2). Hence, a rapid increase in methylation level occurred in plants exposed to both low and high doses even after short-term Cd treatments.

Concerning the frequency of different profile classes (I–III) in Table 1, CP class I was strongly represented compared with the other two classes, ranging from 84% at 6 h to 78% at 4 d. In Cd-tP plants (10 μM and 50 μM), the frequency of class I dropped abruptly to about 69% soon after treatment (6 h) and the frequency of classes II (18%) and III (12%) increased. Further variations were not detected after longer-lasting (2 d or 4 d) Cd exposure. Moreover, hemi-methylation of the outer cytosine (class II) occurred more often than full methylation of internal cytosine (class III) not only in CP but also in Cd-tP plants (see Supplementary Fig. S4 at JXB online).

Methylation-related polymorphisms involved specific genes

To obtain more information on the polymorphisms detected using MSAP, a cluster of fragments that were differentially amplified in Cd-tP vs CP plants were randomly selected, cloned, and sequenced. Specifically, 99 out of 521 polymorphic fragments (19%) were analysed for Cd-tP versus CP plants.

By detecting, through BLASTX sequence homology with the GenBank database, more than 70% of the analysed fragments were annotated as hypothetical or unknown genes probably corresponding to intergenic DNA sequences, introns, promoters, and other junk-DNA (Fig. 1). However, and very interestingly, some genes of known function were also identified accounting for the remaining fraction of analysed polymorphic fragments (Fig. 1; Table 3).

Fig. 1.

Categories of differentially methylated genes in Posidonia oceanica identified by comparing plants either grown in an aquarium in the presence of sea water (CP for 6 h, 2 d or 4 d) or in two different aquaria with CdCl2 (10 μM or 50 μM for 6 h, 2 d or 4 d). (This figure is available in colour at JXB online.)

Table 3.

Predicted function of methylation related polimorphyc fragments detected between Posidonia oceanica plants grown in three different aquaria in the presence of sea water (CP) or CdCl2 at different concentration (10 μM or 50 μM) for different times (6 h, 2 or 4 d) established through BLASTX allignment (http://www.arabidopsis.org/Blast/index.jsp; http://blast.ncbi.nlm.nih.gov/Blast.cgi)

Fragment	Length (bp)	Accession no	Predicted function
Cd1	811	JF787622	YCF2; Hypothetical chloroplast RF2
Cd2	260	JF787625	Nucleic-acid-binding; zinc ion binding
Cd3	358	JF811732	Nucleic-acid-binding; zinc ion binding
Cd4	507	Jf811741	NUDX26; Nudix hydrolase homologue 26
Cd5	542	JF811728	nudx15; Nudix hydrolase homologue 15
Cd6	603	JF811727	HSP70; Heat shock protein 70 family protein
Cd7	290	JF811748	SAUL1; Senescence-associated e3 ubiquitin ligase 1
Cd8	895	JF787626	ADH1; Alcohol dehydrogenase 1
Cd9	449	JF811736	MAPKKK10; Mitogen-activated protein kinase kinase kinase
Cd10	350	JF811739	Leucine-rich-receptor-like-protein-kinase-family
Cd11	531	JF787627	Leucine-rich-repeat-protein-kinase-family
Cd12	406	JF811730	Ccysteine-rich-rlk (receptor-like protein kinase) 8
Cd13	496	JF811734	HMA5; Heavy metal atpase 5
Cd14	423	JF811740	CAX2; Cation exchanger 2
Cd15	346	JF811743	ALMT9; Aluminium-activated malate transporter 9
Cd16	275	JF811744	T14P4.9; Nucleoside transporter family protein
Cd17	354	JF811737	ACBP4; Acyl-CoA binding protein 4
Cd18	387	JF811735	SS4; Starch synthase 4
Cd19	435	JF787623	Retropepsins; pepsin-like aspartate proteases
Cd20	395	JF811731	Acyl-CoA N-acyltransferases (nat) superfamily protein
Cd21	379	JF811747	MWD22.21; Had superfamily subfamily IIIB acid phosphatase
Cd22	419	JF811733	HDA1; Histone deacetylase 1
Cd23	335	JF811738	FLC; flowering locus C
Cd24	208	JF811742	HAC5; Histone acetyltransferase of the CBP family 5
Cd25	479	JF811746	DNA/RNA-polymerases
Cd26	779	JF811745	Reverse-transcriptase (RNA-dependent DNA polymerase)
Cd27	626	JF811729	Reverse-transcriptase (RNA-dependent DNA polymerase)
Cd28	206	JF787624	Transposable-element-gene

Sequences (n=28) belonging to functional genes included: stress responsive (e.g. P. oceanica HSC70, P. oceanica ADH1); transporter proteins (e.g. P. oceanica CAX2, P. oceanica HMA5), epigenetic regulatory proteins (e.g. P. oceanica HDA1); zinc finger, signal transduction proteins, mitogen-activated protein kinase, and others (Table 3).

A specific pattern of 5-methylcytidine-immunolabelling marks apical tips of Cd-tP

The DNA methylation pattern was investigated through an immunocytological approach on interphase nuclei of apical tips. A clear increase of labeling index (number of silver grains/nuclear area unit) was observed in the Cd treatment compared with CP. In particular, significant DNA hypermethylation was observed after 6 h of 50 μM-Cd exposure. DNA methylation reached its highest level after 2 d in both low- and high-dose Cd (10 μM or 50 μM) treatments, but decreased significantly in the longer term treatment of 4 d, and was not significantly different from that in CP at this time (Fig. 2).

Fig. 2.

Labelled Index % (L.I. %) expressed as number of silver grain/nuclear area (μm2) evaluated in apical tips of Posidonia oceanica collected from a preserved meadow and grown in an aquarium in the presence of sea water (CP for 6 h, 2 d or 4d) or treated with CdCl2 (Cd-tP) at different concentrations (10 μM or 50 μM) for different times (6 h, 2 d or 4 d). Statistical analysis: ANOVA followed by the Neuman–Keul's post-hoc test. **P <0.01; ***P <0.001.

Moreover in the Cd treatment, the increase in DNA methylation was associated with a particular distribution of methylation sites (Fig. 3). Notably, there was an increased localization of silver grains along the nuclear membrane to form a dense methylated area. This was particularly evident after 2 d of Cd treatment (10 μM or 50 μM), when the methylation level reached the maximum value (Fig. 3).

Fig. 3.

Nuclear methylation pattern by 5-methylcytidine-immunocytolabelling in apical tips of Posidonia oceanica collected from preserved meadow and grown in an aquarium in the presence of sea-water (CP for 6 h, 2 d or 4 d) or treated with CdCl2 (Cd-tP) at different concentrations (10 μM or 50 μM) for different times (6 h, 2 d or 4 d). Changes in nuclear methylation pattern were detected in relation to both Cd concentration and treatment duration. Bars (6 h, 2 d and 4 d, aquarium untreated sample; 6 h CdCl2 10 μM, 4 d CdCl2 50 μM)=6.5 μm; (4 d CdCl2 10 μM, 6 h CdCl2 50 μM)=6.8 μm; (2 d CdCl2 10 μM and 50 μM)=7.4 μm. (This figure is available in colour at JXB online.)

The expression level of Posidonia oceanica CHROMOMETHYLASE is modulated under Cd treatment

In order to investigate whether DNA hypermethylation was related to variations in the expression level of the CMT gene which encodes plant DNA methyltranferase, a full-length cDNA encoding a P. oceanica CMT homologue was isolated preliminarly.

PoCMT genomic features and structural analysis of the deduced protein product

A complete cDNA of PoCMT consisting of 2685 bp with an ORF of 2409 bp was obtained, which encoded a predicted protein of 802 amino acid residues with a calculated mass of 91.06 kDa. Comparison of the deduced amino acid sequence with other higher plant CMTs revealed the highest identity with Eleais guineensis CMT1 (60%), followed by Arabidopsis thaliana CMT2 (53%) and Oryza sativa MET2c (52%) (Fig. 4C). All the characteristic motifs I, IV, VI, VIII, IX, and X (motifs defined by Posfai ; Kumar ) of the CMT family, including the signature chromodomain between motifs I and IV, were identified (see Supplementary Fig. S5 at JXB online) and so the identified gene was designated as Posidonia oceanica CMT1 (PoCMT1) and submitted to EMBL/Genbank (accession number JF787621).

Fig. 4.

Posidonia oceanica PoCMT1 genomic features and its protein genealogy. (A) Scheme of the genomic organization of PoCMT1. Start and stop codons and polyadenylation signals are typed. Degenerate and gene specific forward (FW) and backward (BW) directed primers used in PoCMT1 characterization are shown by arrows. The RACE System using the 5′- and 3′-anchor primers provided by the manufacturer and the specific PoCMT1 primers, FW3′-RACE and BW5'-RACE was used to recovered 5′ and 3′ ends of the transcripts. The Southern probe fragment is represented by the black bar. UTR, untranslated regions. (B) Southern analysis: genomic DNA was digested with EcoRI (EI), EcoRV (EV), and HindIII (HIII) endonucleases, electrophoresed on a 0.8% agarose gel, blotted, and hybridized with a digoxygenin-labelled probe, which targeted highly conserved CMT motifs. The molecular weights of a co-migrating DNA marker are in kilo base pairs (kb). (C) Phylogenetic tree of selected plant CMTs. Deduced protein sequence was aligned by the multiple alignment program CLUSTALW. The MEGA4 program was used for the construction of phylogenetic trees. CMT genes from P. oceanica are boxed. Bootstraps values (at branch points) are given for major nodes and are percentages of 1000 replicates. The maintenance DNA methyltransferase (A. thaliana MET1 and Homo sapiens DNMT1a), de novo DNA methyltransferase (Homo sapiens DNMT3a/b), and a chromatin remodelling factor (Homo sapiens CHD1) were used to create an outgroup (accession numbers are given in parentheses): Arabidopsis thaliana CMT1 (NP_565245.1), CMT2 (NP_193637.2), CMT3 (NP_177135.1), MET1 (NP_199727.1); Brassica rapa CMT1 (BAF34637.1); Elaeis guineensis CMT1 (ABW96889.1); Hordeum vulgare CMT1 (CAJ01708.1); Nicotiana sylvestris CMT3 (CAQ18903.1); Oryza sativa OsMET2a (BAH37019.1), OsMET2b (BAH37020.1), OsMET2c (BAH37021.1); Posidonia oceanica CMT1 (JF787621); Zea mays CMT1 (Q9AXT8), Zea mays CMT2 (Q9ARI6); Homo sapiens DNMT1a (NP_001124295.1), DNMT3a (NP_072046.2), DNMT3b (NP_787046.1), CHD1 (AAB87381.1).

Genomic organization

Southern blot was used to estimate membership of the CMT gene family in P. oceanica. The probe fragment for PoCMT1 covers the distal end of the chromodomain and conserved motifs IV–VIII and was therefore expected to identify all members of this gene family (Fig. 4A). Figure 4B shows that a genomic DNA probe hybridized two fragments, independent of the restriction enzyme used, indicating the presence in the P. oceanica genome of a small multigene CMT family with at least two members.

The expression level of PoCMT1 increases in Cd-treated plants

The expression level of the PoCMT1 gene was investigated in both apical tips and leaves (Fig. 5). In neither organ did the low Cd concentration (10 μM) treatment induce significant increases in the level of PoCMT1 transcripts; indeed the 4 d 10 μM-Cd treatment induced a small but significant decrease in expression level only in the leaves (Fig. 5).

Fig. 5.

Mean ±SE relative PoCMT1 (Posidonia oceanica CHROMOMETHYLASE1) mRNA level estimated through qRT-PCR in (A) apical tips and (B) leaves of Posidonia oceanica collected from a preserved meadow and grown in an aquarium in the presence of sea water (CP) or treated with CdCl2 (Cd-tP) at different concentrations (10 μM or 50 μM) for different times (6 h, 2 d or 4 d). Expression was determined through triplicate measurements of three independent replicates.

Conversely, the higher Cd concentration (50 μM) treatment resulted in significant upregulation of PoCMT1 in apical tips after 6 h and even more so (10-fold compared with CP) after 2 d of treatment (Fig. 5A). However, transcript levels fell abruptly at 4 d to be only 2-fold more than CP at this time (Fig. 5A). In leaves, 6 h of the 50 μM treatment induced the highest level of PoCMT1 expression but thereafter there was a progressive decrease, although at both 2 d and 4 d, its expression continued to be significantly higher than CP (Fig. 5B). In general, Cd-induced PoCMT1 would be consistent with a clear mechanism of de novo DNA methylation in leaves and apical tips which were particularly sensitive at 6 h and 2 d of 50 μM-Cd, respectively.

Chromatin reconfiguration and apoptotic features under Cd treatment

Chromatin ultrastructure was investigated by analysing interphase nuclei of shoot apical meristems in both CP and Cd-tP plants. In CP, nuclei showed a uniform distribution of euchromatic matrix, with dense chromatin areas called chromocentres, spread across the entire nucleus (Fig. 6). This pattern of chromatin configuration was typical of CP cells regardless of sampling time. The 10 μM-Cd treatment resulted in a similar chromatin structural profile to that observed in CP irrespective of treatment period (from 6 h to 4 d; Fig. 6), and the same pattern was evident after 6 h of 50 μM-Cd (Fig. 6). However, the 2 d 50μM-Cd treatment induced a progressive condensation of chromatin quite unlike that observed in CP (Fig. 6). Indeed, a predominance of strong heterochromatic regions with a consequent reduction in euchromatic matrix was typically observed, together with noticeable heterochromatic deposits on the nuclear envelope. In addition, the 4 d 50 μM-Cd treatment induced a strong heterocromatic protrusion (Fig. 6) together with evidence of nuclear envelope disintegration, perhaps as a prelude to, or coincident with, nuclear apoptotic disorganization (Fig. 6).

Fig. 6.

Chromatin ultrastructure organization by transmission electron microscope of 10 or 50μM CdCl2 treated plants of Posidonia oceanica compared with the control plants (CP 2 d). Black and white arrows indicate examples of chromocenters and nucleoli, respectively. Bar (2 d, aquarium untreated sample, 6 h CdCl2 10 μM, 4 d CdCl2 50 μM)=3.2 μm; (4 d CdCl2 10 μM, 6 h CdCl2 50 μM)=3.4 μm; (2 d CdCl2 50 μM)=3.7 μm.

Discussion

The primary role of DNA methylation in the control of chromatin structure and gene expression has been well documented (Finnegan ; Matzke ). Moreover, abnormal DNA methylation occurs both in animals and plants following different stresses, highlighting the importance of this epigenetic regulatory mechanism for the establishment of a new order of chromatin ‘status’ not only in relation to developmental events but also in relation to stress responses (Arnholdt-Schmitt, 2004; Henderson and Jacobsen, 2007; Boyko and Kovalchuk, 2008; Chinnusamy and Zhu, 2009; Zhang ). In the present work, a tight interface between DNA methylation pattern and Cd action in Posidonia oceanica is demonstrated.

Cd toxicity has been extensively studied in both mammals and plants but its mechanism of action is not unequivocal. So far, in mammals, changes in the level of global and gene-specific DNA methylation following Cd exposure have been reported in four independent studies (Takiguchi ; Benbrahim-Tallaa ; Huang ; Jiang ), featuring DNA methylation as an important mechanism in cadmium-induced cancer progression (Calvisi ; Choi ; Shimizu ). In some species, Cd toxicity can be partially related to the establishment of oxidative stress conditions (Balestrasse , 2006; Benavides ; Gratao ; Garnier ) despite Cd being a non-redox metal. However, at this time, many conflicting data exist and very few data are available in plants on the relationship between Cd toxicity and an epigenetic mechanism such as DNA methylation.

It is demonstrated here that, in P. oceanica, Cd exposure causes genomic DNA hypermethylation. In plant genomes, the overall 5mC content depends on genome size; methylation levels are different in various species and even within different tissues of the same species (Teyssier ). For example, the percentage of the genome that is 5mC-rich is 6% in Arabidopsis thaliana (0.33 pg 2C−1) but 25% in Zea mays (5.57 pg 2C−1) (Ergle and Katterman, 1961). For P. oceanica (6.25 pg 2C−1), a 5mC mean content of approximately 19.8% is reported in actively growing organs, as evaluated through MSAP analysis. Moreover, about a 2-fold increase in DNA methylation level was induced by low-(10 μM) and high-dose (50 μM) Cd treatments. Up-regulation of PoCMT1, a putative member of the plant-specific CMT family, is also shown in apical tips and leaves exposed to Cd. In A. thaliana, CMT genes are generally expressed at moderate levels in all organs of the plant (McCallum ). In Solanum lycopersicum, a progressive decrease in the expression level of putative CMT genes occurs during plant and fruit development (Teyssier ). In our work, and in line with Arabidopsis data, PoCMT1 was moderately expressed in all organs (data not shown) but was strongly up-regulated by a high-dose Cd treatment even in a short-term treatment of 6 h, while such an up-regulation was not observed in a low-dose Cd treatment. On this basis, a preferential initial involvement of DOMAINS REARRANGED METHYLTRANSFERASES (DRMs) de novo methyltransferases is probably in the increase of the 5mC level observed in P. oceanica plants soon after the 10 μM Cd treatment, followed by CMT activities. This is in agreement with published data showing that the initial methylation imprint on DNA in response to stress can be created by DRMs at asymmetric sites and then perpetuated at symmetric CG and CHG by MET1 and CMT3, respectively (Cao and Jacobsen, 2002, ; Chan ; Matzke ; Naumann ). Notably, in mammals, Cd-induced carcinogenesis is associated with an over-expression of both de novo and maintenance DNA METHYLTRANSFERASES (DNMTs) resulting in a relative increase in DNA methylation at the global level (Benbrahim-Tallaa ; Jiang ).

In line with the role of DNA methylation in modulating the chromatin structure, DNA hypermethylation and the up-regulation of PoCMT1 methyltransferase in P. oceanica plants in the 50 μM-Cd treatment, were associated with pronounced chromatin condensation. This reconfiguration of chromatin structure is entirely consistent with the role of CMT3, which is preferentially to methylate transposon-related and repeated sequences and thus increase the level of heterochromatinization (Bartee ; Kato ; Tran, 2005; Chan ). Notably, immunocytological analysis also allowed us to verify that de novo methylation strongly involved heterochromatic areas tightly related to the nuclear envelope. Therefore, it may be safely assumed that the Cd-induced up-regulation of PoCMT1 results in the increase of the DNA methylation ratio, together with an increased heterochromatic nuclear fraction.

Further investigation is now necessary to define why and how the expression levels of PoCMT1 are altered by Cd treatment. In mammals, the transcription of DNMTs is stimulated by the transcription factors, Sp1 and Sp3 (Kishikawa ; Jinawath ), whose expression was induced by abiotic stress (Sato ). The extent to which a similar mechanism exists in plants is currently unknown. Moreover non-CpG methylation is unique to plants, so it may represent an additional degree of complexity of the epigenetic regulatory pathways used by plants. In this context, clear structural abnormalities, such as disorganization of the nuclear membrane and the appearance of apoptotic bodies, all related to chromatin condensation and chromatin damage, were also visible in a small fraction of P. oceanica nuclei. It is well known that the appearance of condensed chromatin as a result of heterochromatin aggregation is one of the common feagures of early apotosis (Yamada ; Doonan and Cotter, 2008; Banfalvi, 2009). Therefore, chromatin structural changes visible in P. oceanica nuclei under a high dose of Cd could be the early steps in the apoptotic degradation of the chromatin compound rather than epigenetic regulation.

These abnormalities are consistent with the phytotoxic effects of Cd at the cellular level. Indeed, exceeding the tolerable level of Cd concentration interferes with DNA repair and synthesis activities and, as a consequence, cells undergo programmed cell death (Banfalvi, 2000; Fojtova and Kovarik, 2000; Banfalvi ; Ma ). Moreover, alterations of DNA methylation induced by Cd exposure may cause aberrant expression of some genes important in apoptotic mechanisms (Deckert, 2005; Bertin and Averbeck, 2006).

The sequencing of random (19 %) MSAP-polymorphic fragments verified that, in P. oceanica plants exposed to Cd, both non-coding sequences and specific genes were targeted by DNA methylation changes. This fraction of differentially methylated genes might include those with a role in cell proliferation, in cell signalling, in proteasome-mediated actions, and in the stress response and, therefore, with a direct effect on cell physiology, thus supporting a complex network of changes underlying Cd toxicity.

In conclusion, CHROMOMETHYLASE, a gene involved in both maintenance and de novo DNA methylation has been characterized in P. oceanica, whose genome information is still very limited. It has been demonstrated that exposure to Cd can perturb P. oceanica at the level of DNA methylation possibly through the involvement of this specific encoded DNA methyltransferase. Moreover, clear evidence is provided that methylation-related chromatin reconfiguration is a very strong feature of the Cd-induced stress event probably accounting for the establishment of a new balance of expressed/repressed chromatin. Further investigations are now required to define the involvement of other MET gene families and, above all, to define whether these changes in DNA methylation pattern could be part of the mechanism established by the plant to overcome Cd stress. However, considering that, at a high dose, Cd exposure induced nuclear structural disorder tightly linked to chromatin remodelling, an epigenetic basis of the Cd toxicity response is highlighted for the first time in plants.