*2.4. Chromatin*/*DNA Accessibility Assay*

For the Linola and genetically modified GM-CHS plants (W92.40), the chromatin/DNA accessibility assay was conducted. The chromatin isolated from in vitro cultured plants was digested with particular restriction enzymes: AatII and PvuI (particular sites for digestion were localized in the CHS gene, around the +1273 CCGG motif). Subsequently, DNA derived from the isolated chromatin was also incubated with restriction enzymes. The nucleic acid was purified and the effect of DNA digestion was assessed by real-time PCR at the methylation CCGG site (+1273). Figure 5 presents the relative quantity of the product obtained for the motif +1273 for undigested control, chromatin and DNA digested with restriction enzymes for Linola and GM-CHS plants. The results of this study have shown that between Linola and genetically modified plants, significant differences in the quantity of the product were obtained. For the GM-CHS flax with the stable methylation in the +1273 CCGG motif an increased level of the product was noted; thus, for transgenic plants, less accessibility of chromatin for restriction enzymes in comparison to the non-transgenic Linola was assumed. No relevant changes were observed in the +1273 motif for the DNA accessibility between the Linola and GM-CHS plants.

the Linola and GM-CHS plants.

ODN, except ODN6 with 6 plants). Despite the percentage of maintaining induced changes being

**Table 3.** Maintenance of gene expression modulation of CHS gene in F1 and F2 generations of plants

**Generation Type of Plant Percentage of Modulation Maintenance (%)** 

ODN1 met 80 ODN6 met 50 ODN11 met 80 ODN11 pto 100

ODN6 met 50 ODN11 met 38 ODN11 pto 47

For the Linola and genetically modified GM-CHS plants (W92.40), the chromatin/DNA accessibility assay was conducted. The chromatin isolated from in vitro cultured plants was digested with particular restriction enzymes: AatII and PvuI (particular sites for digestion were localized in the CHS gene, around the +1273 CCGG motif). Subsequently, DNA derived from the isolated chromatin was also incubated with restriction enzymes. The nucleic acid was purified and the effect of DNA digestion was assessed by real-time PCR at the methylation CCGG site (+1273). Figure 5 presents the relative quantity of the product obtained for the motif +1273 for undigested control, chromatin and DNA digested with restriction enzymes for Linola and GM-CHS plants. The results of this study have shown that between Linola and genetically modified plants, significant differences in the quantity of the product were obtained. For the GM-CHS flax with the stable methylation in the +1273 CCGG motif an increased level of the product was noted; thus, for transgenic plants, less accessibility of chromatin for restriction enzymes in comparison to the non-transgenic Linola was

lower, the results of F2 analysis were more significant.

treated with modified ODNs (met and pto).

F1

F2

2.4. *Chromatin/DNA Accessibility Assay*

**Figure 5.** Chromatin/DNA accessibility assay. Chromatin from non-transgenic (Linola) and GM-CHS flax (W92.40) was isolated using a dedicated technical kit. The chromatin and chromatin-derived DNA (primarily treated with proteinase) were incubated with restriction enzymes AatII and PvuI (particular sites for digestion were localized in the CHS gene). After being cut, the DNA was purified and the effect of DNA digestion was assessed by real-time PCR at the methylation CCGG site (+1273). The values are referred to the reference gene expression actin. The relative quantity (RQ) presents the transcript level in comparison to the samples not incubated with the restriction enzymes (undigested, **Figure 5.** Chromatin/DNA accessibility assay. Chromatin from non-transgenic (Linola) and GM-CHS flax (W92.40) was isolated using a dedicated technical kit. The chromatin and chromatin-derived DNA (primarily treated with proteinase) were incubated with restriction enzymes AatII and PvuI (particular sites for digestion were localized in the CHS gene). After being cut, the DNA was purified and the effect of DNA digestion was assessed by real-time PCR at the methylation CCGG site (+1273). The values are referred to the reference gene expression actin. The relative quantity (RQ) presents the transcript level in comparison to the samples not incubated with the restriction enzymes (undigested, set as 1 for each line). Data represent the mean value ± SD from at least three independent experiments. The significance of the differences between the samples was determined by Student's *t*-test: a—the reference of the methylated motif to unmethylated, b—in reference to the undigested sample for unmethylated motif, c—in reference to the undigested sample for methylated motif, d—the reference of the DNA to chromatin, for the appropriate restriction enzyme. Asterisk indicates \* *p* < 0.05.

#### **3. Discussion**

Genetic engineering methods allow precise modification of the phenotype by introducing changes into the sequence of the plant genome. A well-known and commonly used method of plant genetic modification is *Agrobacterium*-mediated plant transformation. An important aspect of this method is the possibility to acquire stable transformants. Unfortunately, the practical application of transgenic plants for industrial purposes is limited in most European countries, due to restrictive legal regulations. The main reason is the lack of trust of potential GM plant consumers in relation to the harmless impact on people and animals. Another issue would be the risk of the disturbance to the natural environment by transgenic crops. The lack of public acceptance of GM crops encourages scientists to search for new methods that enable induction of organism variability without the need to change the genome [22].

Recently, a lot of attention has been devoted to methods based on the use of site-specific nucleases (SSNs), including TALENs, ZFNs and CRISPR/Cas9 [23]. Although the principle of SSN technology is based only on the editing of endogenous genes, the legal regulations regarding this method are not entirely resolved. Some genome editions can be considered as mutations and thus might be controlled by legal restrictions that apply to GMO [24]. Methods that will allow the use of the natural cell genetic repertoire are eagerly in demand. Epigenetic changes are modifications that take place in nature, mainly during the plant adjustment to constantly changing environmental conditions. In addition, the epigenetic DNA pattern can be maintained during cell division and be inherited [7,25]. Therefore, epigenetic modifications have a large potential for application in the phenotype variability of plants since they are naturally occurring changes.

Plants are considered as the best model for epigenetic regulation research. All the epigenetic mechanisms described so far occur in plants. In the plant genome, methylation in the 5th position of the cytosine in a DNA chain is a significant epigenetic modification. In plants, 5-methylcytosine can apply to three nucleotide contexts: CpG, CpHpG and CpHpH. The equilibrium of the methylation pattern is obtained through the interaction of de novo methylation processes, maintenance of methylation and demethylation. Recently, the methylation of adenine in the N6 position (N6-mA) has also been demonstrated in the genome of *Arabidopsis thaliana*. It has been shown that this modification is essential for plant development and occurs in actively expressed genes [26]. DNA methylation plays an important role in many processes in plants, such as vegetative development, reproduction, fertilization or gametogenesis. Additionally, DNA methylation specifically interacts with other epigenetic mechanisms: posttranslational histone modifications and processes related to non-coding RNA molecules. Due to the constant need to adapt to changing environmental conditions, the "plasticity" of the plant epigenome enables phenotypic traits to be adjusted by specifically regulating gene expression [6].

The use of short oligodeoxynucleotides (ODNs) proved to be a precise method of inducing changes at the epigenomic level. The ODN technology is based on the introduction of short oligodeoxynucleotides of size 12 to 25 nucleotides into cells, which are complementary to the corresponding regions of the target gene. The technology has proven to be an effective method of studying gene function and transformation in animal cells. The method has gained popularity due to the fact that it may impact the modulation of the gene expression of a particular gene and does not lead to modification of the genomic sequence. Currently, ODN technology is optimized to be used in plant research. The favorable property of this method for plant research is the possibility of direct introduction of ODNs to plant cells and limitation of the pleiotropic effects. Although the mechanism of ODN action has not been fully explained yet, it was suggested that the following processes are involved: RNA interference, RNA activation and RNA-dependent DNA methylation [8,9,27]. The majority of reports indicate that the silencing of gene expression is probably the result of the degradation of RNA:DNA duplexes by RNase H [28]. The formation of a triple complex consisting of a DNA double helix and ODN sequence is also not excluded; hence the ability to form a transcript may be limited [29].

The main purpose of our research was to induce significant changes in the transcript level of the CHS gene, which would be maintained after cell divisions. Previously [9], we indicated that the ODN technology is effective in modulation of CHS gene expression and DNA methylation. The changes triggered by the new technology were similar to those observed initially in genetically modified flax (GM-CHS) [9].

Modified ODNs proved to be effective in prolonging the time of maintaining induced changes in methylation of cytosines in the flax genome. Among the two used modifications, methylation of cytosines present in the ODN sequence proved to be more effective than the unmodified version. Analysis of subsequent generations of plants treated with ODNs confirmed the possibility of maintaining induced epigenetic changes during cell division [5]. Due to the declining number of F2 generation plants that maintained the changes induced by unmodified ODNs, F1 generation plants treated with modified oligonucleotides were obtained. The changes induced through ODN6 were maintained in the F1 and F2 generations with the lowest percentage of maintenance. In plants growing in vitro after treatment with ODN6, the tendency of primarily induced changes remains despite subcultures (Figure S1), which confirms the transmission of epigenetic changes during vegetative reproduction [30]. However, at the same time, we cannot exclude that the elevated level of genomic methylation, and hence the reduced expression of the CHS gene, might be triggered by repeated in vitro sub-culturing of plants [31]. More importantly, the modification of ODN 6 through cytosines' methylation led to improvement of the perpetuation of changes in subsequent generations.

Analysis of methylation of CCGG motifs in the F1 generation of plants treated with ODNs showed the maintenance of the methylation pattern in two of the three analyzed motifs. Instability of methylation status observed in motif +1219/+775 concerned not only plants treated with ODNs, but also control plants. Thus, we presume that particular CCGG motifs may be more susceptible to methylation/demethylation driven by environmental factors, just as the cytosines localized in the transcription factor binding region can be more susceptible to methylation [32]. Despite the fact that in

the F1 plants treated with ODN6, the repression of CHS gene was not observed, the methylation pattern of these plants remained after cell divisions. The possible reason for not keeping the CHS expression decreased after the action of ODN6 might be the compensation for effects driven by environmental conditions, and hence increased expression of the CHS gene was observed.

In order to confirm the specific nature of the ODN action, we also determined whether modulation of CHS gene expression displays any changes to the transcript level of the other genes encoding enzymes related to the phenylpropanoid pathway (Figure S2). In plants treated with ODN1, there were no significant changes in the expression of analyzed genes other than *CHS* involved in the synthesis of phenylpropanoids. Only *PAL* and *HCT* showed a minor reduction in gene expression. In the *Arabidopsis thaliana* model plant, HCT gene repression led to a decrease in lignin levels, which is associated with increased CHS gene activity and redirection of metabolic changes to the flavonoid compound synthesis pathway [33]. However, in the flax with CHS gene repression, decreased levels of lignin and HCT gene repression were observed, which did not affect the flavonoid content [19]. Regarding the flax treated with oligodeoxynucleotides modified by methylation, a positive relation was also observed between expression of genes CHS and HCT, where an elevated transcript level of CHS gene was accompanied by increased HCT gene expression. Despite the prolonged maintenance of changes in DNA methylation induced by ODN with thiophosphate bonding, pleiotropic effects concerning the genes related to the synthesis of flavonoids and lignin were observed. Analyzed genes displayed elevated transcript levels. According to a research by [19], enzymes involved in the biosynthesis of flavonoid compounds and lignin occur in the form of multienzyme complexes in which the main regulating component is possibly CHS [19]. Therefore, modulation of CHS gene expression may lead to control of the expression of other genes encoding the components of these complexes. Moreover, in potato transgenic plants overexpressing the CHS gene, effects on other metabolic pathways were also observed, which may be caused by competition for substrates or the effect of the synthesized flavonoids on plant metabolism [34]. However, in plants treated with ODN1 met, no changes in the expression of other genes encoding enzymes involved in the synthesis of phenylpropanoids (Supplemental Figure S2) were observed, which may suggest that this modification not only stabilizes the modulation of target gene expression, but also increases the specificity of ODN. The group of Khorova have proven that methylation of oligonucleotides increases nucleic acid resistance to nuclease degradation and increases binding strength when interacting with the complementary sequence [35].

DNA methylation is known to cooperate with the positioning of nucleosomes to regulate the structure of chromatin, which leads to an appropriate modulation of gene expression [36]. We have already presented the analysis of the nucleosome energy diagram, which showed that methylation of variable motifs in the CpG island region results in the formation of less stable nucleosomes with higher energy and their shift towards the 30 end of the DNA [9]. In the vicinity of one of the methylated motifs, sites recognized by AatII and PvuI endonucleases were identified; hence experimental verification of the is silico analysis was performed. In the chromatin structure, DNA regions bound to the histone proteins are less available for endonucleases [37]. Thus, DNA present in nucleosomes is protected from degradation by restriction enzymes. Analysis of chromatin digestion and DNA showed that there is probably a change in the positioning of nucleosomes as a result of the increase in the level of methylation in CCGG +1273 sites.

The present results confirmed that ODN technology induces changes in the plant genome that can be inherited. These data contribute to the statement that the use of natural genetic repertoire through epigenetic variability can be a valuable alternative to improve crop quality.

#### **4. Materials and Methods**

## *4.1. Designing Short Oligodeoxynucleotide Sequences (ODN) Homologous to CHS Gene Regions*

On the basis of our previous work [9], for this study, three ODNs homologues to the particular CHS gene regions were selected: ODN1 (50UTR), ODN6 (intron) and ODN11 (30UTR). The ODN sequences were designed using Mfold software (version 3.2, Genetics Computer Group, Madison, WI, USA) in the antisense orientation and correspond simultaneously to two CHS gene isoforms: LuCHS6 (Phytozome data base: Lus10033717) and LuCHS7 (Phytozome database: Lus10031622). Detailed information about used ODN sequences were reported by [9] (Table 2, Supplementary Table S1). In order to enhance the resistance of the sequences to the degradation by nucleases, the sequences of the three mentioned oligonucleotides were modified by methylation (all cytosines occurred in the sequences) and thiophosphate modification (at each phosphodiester bond). The oligodeoxynucleotides were synthesized by Genomed S.A. (Warszawa, Poland).

#### *4.2. Plant Material*

Flax seeds *(Linum usitatissimum* L., cv. Linola) were obtained from the Flax and Hemp Collection of the Institute of Natural Fibers (Pozna ´n, Poland).

#### Flax Growing Conditions for ODN Technology

To investigate whether modified ODNs are more effective in the stability of the induced epigenetic modulation, flax plants were cultured in in vitro conditions in the phytotron at 16 h light (22 ◦C), 8 h darkness (16 ◦C). The seeds were first sterilized for 15 min with 50% PPM—a broad-spectrum biocide/fungicide for plant tissue culture (Plant Preservative Mixture; Plant Cell Technology, Washington, DC, USA) and then germinated on Murashige and Skoog (MS) basal medium (Sigma-Aldrich, St. Louis, MO, USA), supplemented with 2% sucrose, pH = 5.8, solidified with 0.8% agar on Petri dishes. Mature flax plants were cultured on the MS basal medium, supplemented with 2% sucrose, pH = 5.8, solidified with 0.8% agar in sterile glass jars. In order to minimize plant infection by pathogens, the medium was complemented with 0.1% PPM™ (Plant Preservative Mixture) Plant Cell Technology (Washington, USA).

In order to obtain seeds for raising the next generation of plants in the field, *Linola* flax seeds were germinated in the soil in the phytotron conditions in 16 h light (22 ◦C), 8 h darkness (16 ◦C), before they were treated with ODNs.

#### *4.3. Identification of cDNA Sequences*

The isolated mRNA from Linola flax was submitted to sequencing. Necessary sample preparations, sequencing and data processing were performed by Genomed S.A. (Warszawa, Poland). Data from three biological replicates were analyzed.

#### *4.4. Flax Treatment with ODNs by Infiltration*

Four-week-old Linola flax cultured in in vitro conditions were used for the treatment with modified ODNs. The plants were cut above their roots and submerged in water solution of the particular ODN in the 10 µM concentration. The incubation was carried out in the vacuum chamber for 20 min. After the treatment, the plants were put into an MS medium. The material for the analysis was harvested 10 days after infiltration. The harvesting time was indicated by previous experiments [8] and enabled determination of the ODNs' activity in comparison to the unmodified sequences.

For the field cultivation, whole two-week-old seedlings germinated in the soil were treated with the ODNs by the forced osmosis method. It was described in [38] that the sugar starvation method of introducing ODNs is effective in the induction of changes in DNA methylation. First, the plants were transferred from the soil into the water and kept in the darkness in the phytotron chamber in order to deplete endogenous sucrose. Then seedlings were put in the water solution of 100 mM sucrose in the presence of 10 µM antisense ODNs. After 24 h of incubation in the darkness with particular oligodeoxynucleotides, plants were put into the soil and cultivated in the experimentation field. The matured seed capsules were harvested approximately four months later. Collected seeds were stored in dry, cool conditions. In order to determine the stability of induced modifications in the subsequent generations, some of the seeds were sterilized and cultured in in vitro conditions.
