Elevated levels of cadmium (Cd) were found in the soil in several regions of the world (Khan et al., 2011; Wang et al., 2015; Dumitrel et al., 2017). Contamination of the environment with this metal possesses serious threat to plants growth as Cd is highly mobile and toxic. Importantly, it can be absorbed by crop plants and in this mode enter human organisms leading to serious disorders. Cd is considered as class I carcinogen. The main targets of its toxicity are kidneys, lungs, and skeleton system (Kah et al., 2012).

In recent years, significant effort has been made to elucidate Cd impact on plants. The research has been focused on the toxicity mechanisms, Cd sensing, and the development of metal tolerance. The most universal responses to this toxic element, not only limited to plants but also found in bacteria and animals, are over-production of ROS (reviewed in Chmielowska-BąK and Deckert, 2012). ROS are “double-faced” molecules, which on one hand can lead to significant damage of cellular compounds but on the other hand are crucial components of signaling network indispensable for the activation of the defense (Cuypers et al., 2016; Sewelam et al., 2016). Recently it has been proposed that ROS signal is transmitted by the products of the oxidation of biological molecules including lipids, proteins, and nucleic acids (reviewed in Chmielowska-BąK et al., 2015).

Indeed, various studies carried out on human, animal, and plant system demonstrated correlation between oxidation of certain species of mRNA and decrease in the level of encoded proteins, indicating that the process constitutes a post-transcriptional gene regulatory mechanism (Shan et al., 2003, 2007; Tanaka et al., 2007; Chang et al., 2008; Bazin et al., 2011; Gao et al., 2013). The decrease in the amount of proteins encoded by oxidized transcripts is most probably dependent on recently described ribosome stalling. In vitro studies carried out on reconstituted bacterial system demonstrated that occurrence of 8-OHG, the most common oxidative modification of RNA, causes slowing down of translation process by 2–4 magnitudes. The effect has been observed regardless of the position of oxidized bases in the codon, even in the wobble position. In eukaryotic extracts, translation was nearly completely inhibited by the presence of 8-OHG. It was suggested that occurrence of 8-OHG in transcripts leads to alterations in RNA–RNA interactions and prevents adaptation of active conformation in the decoding center. At the same time, it has been shown that oxidized transcripts are subjected to ribosome-based quality control and are predestined for degradation through No-Go decay pathway (NGD) (Simms et al., 2014). In concordance, yeast mutants with alerted decapping system leading to less efficient mRNA degradation showed elevated level of 8-OHG in the transcripts accompanied by premature death of the cells. The yeast mutants were also characterized by higher frequency of reversion from Trp^- (tryptophan minus) phenotype (Stirpe et al., 2016).

Another major discovery is the fact that mRNA oxidation is a selective process. Study carried out on postmortem isolated brain tissues of patients suffering from Alzheimer’s disease demonstrated that oxidation did not occur in the most abundant transcripts such as β-actin. On the other hand, the most prominent oxidation was always noted in specific mRNAs encoding proteins engaged in signal transduction, cellular transport, gene expression regulation, and response to alerted ROS metabolism (Shan et al., 2003). These findings have been confirmed by other reports carried out on human, animal, and plant models (Shan et al., 2007; Tanaka et al., 2007; Chang et al., 2008; Bazin et al., 2011; Gao et al., 2013). Despite the importance of transcript oxidation in genes regulation and cell functioning so far, only two studies were dedicated to elucidation of this phenomenon in plants. It was demonstrated on sunflower seeds that alleviation of seed dormancy during dry after-ripening was associated with increase in the 8-OHG in mRNA. The observed oxidation was limited to 24 definite transcripts encoding proteins associated with metabolism, response to stress factors, and transport (Bazin et al., 2011). Similarly, studies on wheat showed selective oxidation of certain transcripts associated with changes in protein levels and release from seed dormancy (Gao et al., 2013).

Beside the 8-OHG, oxidation might lead to the formation of numerous other modified bases in nucleic acids (Barciszewski et al., 1999). Studies using in vitro translation system showed that the most common oxidative modifications of RNA, namely 8-OHG, 5-hydroxyuridine (5-OHU), 5-hydroxycytidine (5-OHC), 8-oxo-7,8-dihydroadenosine (8-OHA), 1,N6-ethenoadenosine (ε-A), 3,N4-ethenocytidine (ε-C), and abasic sites (AP), result in slowing down or complete inhibition of the translation process (Calabretta et al., 2015). In turn, oxidation of DNA is associated with decrease in its stability and enhanced mutation rate. High level of oxidatively modified bases has been noted in various types of cancer (Sedelnikova et al., 2010). The repair of DNA lesions induced by oxidation is carried out by the Base Excision Repair (BER) pathway. The initial step of BER is recognition and excision of the modified bases by DNA glycosylases leading to the formation of AP-sites (abasic sites), which are considered markers of oxidative stress (Čolak, 2008; Antoniali et al., 2017). Recently, a method of abasic sites detection has been successfully applied in the evaluation of RNA oxidation (Tanaka et al., 2011). However, so far the changes in the level AP-sites in transcripts of plants exposed to stress factors have not been described.

The aim of present study was examination of the influence of Cd in two concentrations (10 and 25 mgl^-1) on the intensity of oxidative stress and frequency of RNA oxidation-dependent modifications – 8-OHG and AP-sites. Our previous research showed that in the earliest period of Cd stress (3 h), ROS are engaged in the regulation of gene expression, while strong accumulation of O₂^- and H₂O₂ has been marked after 24 h (Chmielowska-BąK et al., 2017). Therefore, these two time points, 3 and 24 h, were applied in the present study.

Exposure of seedlings to Cd for 3 h had no effect on their morphology, growth, or the amount of dead cells (Figures 1A,C,D). In turn, 24 h of Cd stress resulted in roots browning and inhibition of their growth (Figures 1B,C). Exposure for 24 h to higher Cd concentration led to significant increase in the amount of dead cells (Figure 1D).

The level of 8-OHG (oxidatively modified base) was approximately 5 times higher in mRNA in relation to the total RNA (Figures 2A,B, 3A,B). In response to Cd, significant increase in the level of 8-OHG has been noted in the total RNA and mRNA after 3 h of treatment with lower concentration (10 mgl^-1) (Figures 2A,B). This effect was not observed after 24 h or in response to higher Cd concentration (25 mgl^-1) (Figures 3A,B).

The opposite tendency has been noted in the case of ROS accumulation. General ROS were detected with specific dye, CM-H₂DCFD, which in response to oxidation emits green fluorescent signal. The fluorescence signal was generally lower after 3 than after 24 h. No differences in fluorescence intensity have been observed between the roots of control and Cd-stressed seedlings after 3 h (Figure 2C). However, a visible increase in the fluorescence signal has been noted in roots of the seedling treated with Cd for 24 h in relation to the control (Figure 3C).

The ROS over-production was correlated in time with increased lipid peroxidation (Figure 3D) and AP-sites frequency in the mRNA (Figure 3E and Supplementary Figure S1). After 3 h, no significant changes in TBARS or AP-sites level were observed between Cd-stressed and control seedlings (Figures 2D,E). In turn, after 24 h, the level of TBARS and AP-sites was significantly higher in the roots of seedlings treated with Cd at higher concentration (25 mgl^-1) (Figures 3D,E and Supplementary Figure S1).

Exposure of plants to Cd leads to the development of various symptoms of toxicity (reviewed in Gallego et al., 2012). In the present study, 3 h long treatment with this metal did not affect the morphology and growth of soybean seedlings roots. At this time point, no statistically significant differences in cells viability were noted (Figures 1A,C,D). However, already after 24 h, browning and shortening of the roots in response to Cd treatment has been observed (Figures 1A,C). Additionally, 24 h long exposure to higher Cd concentration led to significant increase in the amount of dead cells (Figure 1D).

Cadmium (Cd) toxicity might be, at least partially, mediated by ROS. These molecules, which include hydrogen peroxide (H₂O₂), hydroxyl radical (HO), superoxide anion (O₂^-), and singlet oxygen (¹O₂), are highly reactive and mediate oxidation of various cellular compounds. Numerous reports showed that exposure to stresses leads to increase in the level of ROS accompanied by oxidation of proteins and membrane lipids (Gill and Tuteja, 2010; Cuypers et al., 2016; Sewelam et al., 2016). In fact, accumulation of the product of lipid peroxidation, malondialdehyde (MDA) and TBARS, is considered a typical symptom of oxidative stress. However, so far little attention has been given to the oxidation of nucleic acids in plant exposed to unfavorable condition. Moreover, the studies were limited only to changes in the level of DNA oxidation markers (Macovei et al., 2010; Yin et al., 2010), while there is no information concerning ROS impact on RNA. It is worth highlighting that studies on bacteria showed higher susceptibility of RNA to oxidation when compared to DNA. Under the same conditions, the level of oxidatively modified RNA exceeded level of oxidized DNA (Liu et al., 2012). This might be explained by the fact that DNA molecules are more protected due to localization in the nucleus, higher level of packing, and association with numerous proteins. Among the main RNA types, mRNA seems to be the most vulnerable to the oxidation (Bazin et al., 2011).

Indeed in the present study, the frequency of the most common oxidative modification, 8-OHG, was approximately 5 times higher in mRNA than in the total RNA (Figures 2, 3). Another modification associated with oxidation processes, abasic sites (AP-sites), were undetectable in the total RNA using same or even higher concentrations of sample as in the case of mRNA (data not shown). The results indicate that also this modification occurs more frequently in the transcripts than other RNA types. Interestingly, Cd-dependent induction of 8-OHG and AP-sites were separated in time. Significantly, higher levels of 8-OHG in response to Cd were noted only after 3 h of treatment with lower metal concentration (Figures 2A,B). In turn, increase in the level of AP-sites was observed after 24 h of exposure to the higher concentration (Figure 3E) and was accompanied by increase in the level of other markers of oxidative stress – strong ROS over-production (Figure 3C) and enhanced lipid peroxidation (Figure 3D).

The observed variable in time effect of Cd-dependent ROS signal is in concordance with other studies. For example, several reports indicate that Cd stress leads to the generation of differing in the time of occurrence ROS waves (Garnier et al., 2006; Peréz-Chaca et al., 2014; Lv et al., 2017). In the case of the study on tobacco suspension cells, it has been shown that exposure to this metal leads to generation of three ROS waves, whereas the earliest one occurred already within the first hour of metal treatment and was dependent on the activity of membrane bond enzyme, NADPH oxidase. Same research reported that Cd cytotoxicity was associated with the second ROS wave resulting from disturbances in mitochondria functioning (Garnier et al., 2006). Also our earlier research showed that ROS signal in soybean seedlings exposed to Cd differs in time. The activity of NADPH oxidase modulated expression of signaling associated genes after 3 h of metal treatment, while significant ROS accumulation and lipid peroxidation were noted only after 24 h (Chmielowska-BąK et al., 2017). Apparently, the role of ROS in plant cells is temporal, species, and spatial specific (Miller et al., 2008; Cuypers et al., 2016). For example, it has been evidenced in Arabidopsis thaliana that ROS signal originating from peroxisomes and chloroplasts has distinct effect on the transcriptome (Sewelam et al., 2014). Another study using agents, which induce distinct ROS types, showed that some transcripts are modulated specifically by O₂^-, H₂O₂, or ¹O₂ (Gadjev et al., 2006).

Reactive oxygen species might play various roles in plants exposed to stress conditions. On one hand, these molecules are responsible for oxidative damage of membrane lipids, proteins, and nucleic acids (Das and Roychoudhury, 2014; Rio, 2015). On the other hand, ROS are engaged in signaling network and defense mechanism (Cuypers et al., 2016). ROS-dependent signaling includes direct sensing, for example, through transcription factors or serine/threonine protein kinase oxidative signal-inducible 1 (OXI1) and indirect modulation of signal through changes in cellular redox status and/or interaction with other signaling molecules such as nitric oxide, calcium ions, mitogen-activated protein kinases, or plant hormones (Neill et al., 2002; Chi et al., 2013; Kopczewski and Kuźniak, 2013; Wrzaczek et al., 2013; Waszczak et al., 2015; Cuypers et al., 2016). Recently, it has been proposed that ROS signal might be also transmitted by products of oxidation such as oxylipins, peptides derived from protein oxidation, and oxidatively modifies nucleic acids (reviewed in Chmielowska-BąK et al., 2015). One of the important future challenges in research concerning the role of ROS in plants response to stresses is elucidation of the exact role of specific ROS signals.

In the case of present research, the observed Cd-dependent induction of AP sites is most probably a symptom of oxidative stress (Figure 3E and Supplementary Figure S1). The assumption is based on the fact that AP sites induction was correlated in time with significant increase in ROS level and lipid peroxdiation (Figures 3C,D). However, at the present stage of research, it is difficult to explain the exact role of rapid induction of 8-OHG formation in RNA noted already after 3 h (Figures 2A,B), when the symptoms of oxidative stress were still not detectable (Figures 2C,D). Studies carried out on animal, human, and plant models showed that 8-OHG formation is not a random but highly selective process limited to defined transcripts, although the mechanism of its selectivity has not been yet discovered. High rate of 8-OHG in transcripts leads to ribosome stalling and in consequence to the decrease in the amount of encoded proteins (Shan et al., 2007; Tanaka et al., 2007; Chang et al., 2008; Bazin et al., 2011; Gao et al., 2013; Simms et al., 2014). The selective nature of 8-OHG formation and its impact on protein biosynthesis indicate that this process constitutes a newly discovered mechanism of post-transcriptional gene regulation. Interestingly, it seems that the gene regulatory function of transcripts abundant in 8-OHG plays distinct role in animals and plants. In the case of animal and human models, the high 8-OHG level in mRNA has been shown to be associated with the development of neurodegenerative disorders (Shan et al., 2003; Chang et al., 2008; Kong and Lin, 2010). In turn in plants, this oxidative modification of transcripts is essential for regulation of the level of certain proteins and alleviation of seed dormancy – a natural process in plants’ life cycle (Bazin et al., 2011; Gao et al., 2013).

In summary, this is the first report showing increased oxidation of total RNA and mRNA in plants exposed to stress. The observed increase in the level of AP-site was correlated in time with strong ROS accumulation and lipid peroxidation indicating that this mRNA modification constitutes a marker of oxidative challenge. However, the rapid induction of 8-OHG is puzzling and its exact role in plants response to unfavorable conditions needs further elucidation.

JC-B and JD designed the research. JC-B and AE-G carried out the cultivation of material, collection of samples, and RNA and mRNA isolation. JC-B and KI conducted the evaluation of ROS and AP-sites level. JC-B, MB, and AE-G performed the measurements of 8-OHG level including isolation of mRNA. JC-B carried out the examination of lipid peroxidation. All authors analyzed the obtained results. JC-B wrote the manuscript. All authors critically read, corrected, and approved the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.