Hypomethylation of cytosine residues in cold-sensitive regions of Cestrum strigilatum (Solanaceae)

Heterochromatin comprises a fraction of the genome usually with highly repeated DNA sequences and lacks of functional genes. This region can be revealed by using Giemsa C-banding, fluorochrome staining and cytomolecular tools. Some plant species are of particular interest through having a special type of heterochromatin denominated the cold-sensitive region (CSR). Independent of other chromosomal regions, when biological materials are subjected to low temperatures (about 0 °C), CSRs appear slightly stained and decondensed. In this study, we used Cestrum strigilatum (Solanaceae) to understand some aspects of CSR condensation associated with cytosine methylation levels, and to compare the behavior of different heterochromatin types of this species, when subjected to low temperatures.


Introduction
The major portion of plant genomes is composed of repetitive DNAs, which are organized in families, according to base-pair length, mechanisms of accumulation/reduction, dispersion and their role in genome behavior. The most important site of satDNA occurrence is heterochromatin, and the genomes of some plant species exhibit large portions of heterochromatin, in addition to other repetitive families, for example, Scilla (Deumling and Greilhuber, 1992) and Citrus (Silva et al., 2010). Furthermore, heterochromatin itself can also be associated with various repetitive DNA families (Grewal and Jia, 2007;Gaeta et al., 2010), which can be revealed after Giemsa C-banding, differentiated as AT-rich (DAPI + ) or GC-rich (CMA + ) by fluorescence staining (Guerra, 2000), and recognized by molecular techniques (Chang et al., 2008). A good example of this is in a study by Urdampilleta et al. (2009), who isolated and located a satDNA (AT-rich) with a length of 725 bp, in the chromosomes of Urvillea chacoensis (Sapindaceae).
The genomes of species of Cestrum (Solanaceae) also possess numerous and diverse repetitive DNA families, many of which associated with heterochromatin. These, en-countered as dots or large blocks, can be classified into i) C-Giemsa + bands, ii) CMA + bands associated or not with NORs, (iii) DAPI + bands associated or not with CSRs (cold-sensitive regions), and (iv) CMA + /DAPI + bands Greilhuber, 1992, 1993a,b;Fregonezi et al., 2006;Fernandes et al., 2009). Cold-sensitive regions are revealed when meristematic tissues are exposed to low temperatures (about 0°C). They appear as decondensed chromatin, weakly stained after exposure to a nonspecific dye, such as Giemsa or orcein. CSRs have been described in approximately 50 plant-species belonging to different families (see Punina et al., 2001). Eight species of Cestrum revealed CSRs. Cestrum parqui, C. strigilatum, C. fasciculatum, C. elegans and C. aurantiacum were studied by Greilhuber (1992, 1993a,b). Fregonezi et al. (2006), when studying C. amictum, C. intermedium, C. sendtnerianum and C. strigilatum, showed that, in these species, the principal CSRs are DAPI + . Even so, in the case of C. strigilatum, not all DAPI + heterochromatin corresponds to CSRs. Although, according to Punina et al. (2001), most CSRs are composed of DAPI + heterochromatin, there are species of the genus Sambucus and Viburnum with CMA + /CSRs (Benko-Iseppon and Morawetz, 1993).
To better understand the decondensed behavior of DAPI + /CSRs in Cestrum strigilatum, we examined root tips treated at low temperatures (= cold treatment) using se-quential fluorescence banding (CMA/DAPI) and immunolabeling with an anti-5-methylcytosine antibody.

Material and Methods
Seedlings of Cestrum strigilatum (10) were collected in five different localities at Londrina, Paraná, Brazil, for subsequent cultivation in tubes in the Laboratório de Biodiversidade e Restauração de Ecossistemas (LABRE), Centro de Ciências Biológicas, Universidade Estadual de Londrina, Londrina, Paraná, Brazil. Vouchers were deposited at the FUEL herbarium. Samples were pre-treated in two ways: i) with 0.05% colchicine at room temperature for 4 h to obtain conventional chromosomes and ii) with distilled water at 0°C for 25 h to obtain CSRs. In both cases, samples were fixed in a solution of ethanol/acetic acid (3:1, v:v), for up to 24 h and stored at -20°C. For conventional staining, samples were softened in 4% cellulase plus 40% pectinase (w: v) at 37°C for 1 h, hydrolyzed in 1 N HCl for 10 min at 60°C, dissected in a drop of 60% acetic acid, and then squashed. After removal of the coverslips by freezing in liquid nitrogen, samples were stained in 2% Giemsa and mounted with Entellan (Merck).
Chromosome banding was according to Schweizer (1976). For both pre-treatments, root tips were first softened in an enzyme solution, dissected and squashed, as previously described. After removal of the coverslips by freezing in liquid nitrogen, samples were aged for 3 days, stained with 0.5 mg/mL of chromomycin A 3 (CMA) for 90 min and 2 mg/mL of 4-6-diamidino-2-phenylindole (DAPI) for 30 min, and mounted in 1:1 glycerol/McIlvaine buffer (pH 7.0) plus 2.5 mM MgCl 2 .
Immunodetection was performed in root tips previously treated with colchicine and with low temperatures. Samples were softened in 4% cellulase plus 40% pectinase (w: v) in 1x PBS buffer at 37°C for 2 h, dissected and squashed, as previously described. After removal of the coverslips by freezing in liquid nitrogen, slides were kept in 1x PBS for 5 min. The material was blocked in a solution of 3% BSA (Inlab) in 1x PBS plus 0.2% Tween 20 at room temperature for 10 min. Afterwards, samples were incubated in a moist chamber at 4°C for 12 h with 25 mL of a solution (1:100, v:v) of mouse anti-5-methylcytosine (Eurogentec) and 3% BSA (Inlab) in 1x PBS plus 0.2% Tween 20. After three washes in 1x PBS for 5 min each, samples were incubated with the FITC-conjugated secondary goat anti-mouse IgG, diluted 1:100 (v:v) in 3% BSA in 1x PBS plus 0.2% Tween 20, at 37°C for 1 h. Washes were carried out in 1x PBS at room temperature. Slides were mounted with 25 mL of a solution composed of glycerol (90%), DABCO (2.3%), 20 mM Tris-HCl, pH 8.0 (2%), 2.5 mM MgCl 2 (4%), and distilled water (1.7%), plus 1 mL of 2 mg/mL DAPI.
Images were acquired separately in grayscale mode using a Leica DM 4500 B microscope equipped with a DFC 300FX camera, and overlapped using a Leica IM50 4.0 software, with the colors blue for DAPI, yellow for CMA 3 , and greenish-yellow for FITC.

Results and Discussion
Conventional Giemsa staining and CMA/DAPI banding showed that the samples of Cestrum strigilatum examined here had a chromosome number of 2n = 16, with five metacentric and three submetacentric pairs ( Figure 1A), as well as large AT-rich blocks in pairs 2, 3, 4, 6, 7 and 8 (Figure 1C) and fewer GC-rich blocks in pairs 7 and 8 (Figure 1E), as previously reported (Berg and Greilhuber, 1993a;Fregonezi et al., 2006). Cold treatment showed that the DAPI + /CSRs of pairs 2, 4, 6, 7 and 8 were negatively stained with CMA ( Figures 1B, D, F and 2), as reported by Berg and Greilhuber (1993a) and Fregonezi et al. (2006). Notwithstanding, some fine AT-rich blocks that appeared as CMA -( Figure 1D and F, indicated by arrowheads), and also in pair 4 (Figure 2), were not cold-sensitive. The decondensed appearance of DAPI + /CSRs was more evident when compared to adjacent GC-rich blocks (Figures 2 and  3, pair 8) and non-adjacent blocks (Figures 2 -pair 7 and 3pair 6). This difference in chromatin condensation certainly depends on epigenetic control, which basically consists of the modification of histones and/or DNA methylation (Fuks, 2005).
In general, when compared to other chromosome regions, heterochromatic regions have high levels of cytosine methylation, as also occurs in non-cold-treated chromosomes of C. strigilatum ( Figure 4A-F). The use of the anti-5-methylcytosine antibody in plants has been effective in differentiating highly methylated cytosine residues in chromosomes (Cremonini et al., 2003). Differences in methylation levels between euchromatic and heterochromatic regions have been reported by immunolabeling with anti-5-methylcytosine antibody in several plant groups, such as Haplopappus gracilis (Castiglione et al., 2008), and Costus spiralis and Eleutherini bulbosa (Feitosa and Guerra, 2011).
Immunodetection using anti-5-methylcytosine antibodies in cold-treated samples revealed hypomethylation in the DAPI + /CSR regions ( Figure 5). The differences in methylation levels can be better visualized, on comparing both DAPI + regions of non-cold-treated samples (Figure 4) and the terminal CMA + block on the short arm of pair 7. Notwithstanding also showing hypermethylation ( Figure  5G-H, head and arrowhead), thereby indicating that ATrich sequences of C. strigilatum are more inclined to hypomethylation than the GC-rich. It is assumed here that this relationship is due to DAPI preferentially binding to ATrich and CMA to GC-rich regions (see Guerra, 2000). Interestingly, Cremonini et al. (2003) obtained CSRs in those chromosomes of Zingeria biebersteiniana that were not fully compatible with the heterochromatic regions revealed by chromosome banding. Our results with C. strigilatum seem to be more consistent with the hypothesis proposed by Punina et al. (2001) that cold-sensitive regions are AT-rich heterochromatic regions that decondense in response to a strong cold-shock. It is proposed that cold treatment could interfere in the stability of DAPI + /CSRs in C. strigilatum, thereby provoking modifications in normal heterochromatin condensation. According to Avramova (2002), DNA methylation is important in stabilizing heterochromatin structures.
Species of Cestrum show karyotypical differences in the number of sites and amount of CSRs. In spite of this discrepancy, it is important to note that in our results, as well as for DAPI + /CSRs bands in species of Cestrum (Berg and Greilhuber, 1993a;Fregonezi et al., 2006), and CMA + /CSRs bands in species of Sambucus and Viburnum (Benko-Iseppon and Morawetz, 1993), only larger heterochromatic blocks are not only more easily visualized, but Hypomethylation in CSRs of Cestrum strigilatum 457   appear evidently decondensed, when samples are treated at low temperatures. Thus, there is possibly a minimum size for heterochromatin blocks capable of detecting changes in methylation levels, when samples are exposed to cold-treatment. This, together with the examination of small and hypomethylated heterochromatic regions being difficult through technical limitations, could possibly be alternative explanations for the occurrence of cold-sensitive regions containing both DAPI + and CMA + . However, it is important to mention that anti-5-methylcytosine detection has not been undertaken in Sambucus and Viburnum species.
To better understand the behavior of CSRs, as well as the evolutionary significance, epigenetic role and molecular drive of these DAPI + /CSRs in Cestrum, it will be necessary, not only to determine the sequences of these satDNA regions, but also test other antibodies against proteins capable of controlling heterochromatin dynamics, such as post-translational modifications associated with histone phosphorylation, acetylation, ubiquitylation and/or methylation. Figure 4 -Immunolabeling with an anti-5-methylcytosine antibody in the prometaphasic chromosomes of C. strigilatum obtained from control experiments (non-cold-induced). A, C and E correspond to pairs 2, 4 and 8, respectively, after DAPI staining. B, D and F correspond to pairs 2, 4 and 8, respectively, after immunolabeling and detection with an FITC-conjugated secondary antibody. Arrows indicate hypermethylated DAPI + bands. For comparison, observe the decondensed and hypomethylated regions in Figure 5.
Figure 5 -Immunolabeling with an anti-5-methylcytosine antibody in the prometaphase and metaphasic chromosomes of C. strigilatum obtained from cold-induction experiments. A) DAPI stained partial prometaphase. Arrows indicate CSRs. This image was acquired in grayscale format, whereupon the DAPI bands were converted to blue. B) The same prometaphase after immunolabeling and detection with an FITC-conjugated secondary antibody. Arrows indicate CSRs. C-D, E-F and G-H show pairs 4, 6 and 7, respectively. Note the hypomethylation status of the DAPI + /CSR (arrows). For comparison, observe the terminal region of the short arm in pair 7, where there is a hypermethylated NOR-associated CMA + band, which shows negative DAPI staining.