Phytohormones such as, abscisic acid (ABA), jasmonic acid (JA), and salicylic acid (SA) play vital roles in stress responses (Reje et al., 2014). For example, ABA is active when plants are exposed to heavy metals, drought, extremes in temperature, high salinity, or radiation. It also functions in various developmental processes, including seed germination and dormancy, and the closure of stomata (Vishwakarma et al., 2017). Both JA and SA are involved in responses to biotic stresses. The JA-signaling pathway is critical to the development of resistance to necrotrophic pathogens that derive nutrients from dead or dying cells, while the SA-signaling pathway is mainly activated against biotrophic pathogens that obtain nutrients from living host tissues (Glazebrook, 2005). In plants resistant to pathogen infections, the JA- and SA-defense pathways generally interact antagonistically (Verhage et al., 2010).

When plants are stressed, their hormone signaling is altered, which then changes the expression of responsive genes. In the gene regulation network, RNA binding proteins (RBPs) participate at the transcriptional and post-transcriptional levels, which is a powerful strategy of gene expression regulation. The most common post-transcriptional modifications are alternative splicing and polyadenylation, which change RNA sequences, protein functioning, and the availability of RNA-binding sites (Filichkin et al., 2010; Di Giammartino et al., 2011). The RBPs bind with target RNAs via their RNA-binding domains. To date, approximately 40 types of RNA-binding domains have been identified from various RBPs (Kishore et al., 2010). The novel YTH (for YT521-B homology) domain and YTH domain-containing RNA binding proteins (YTPs) have been described in yeast, animals (including humans; Homo sapiens), and plants (Stoilov et al., 2002).

The first identified YTP was YT521, found in the astrocytes of rat (Rattus norvegicus). The mRNA level of YT521 can be depressed under oxygen-deficient conditions but recovered after reoxygenation. Proteins that interact with YT521 are also RNA-binding proteins, e.g., RA301, SC35, and SF2, which indicates that YT521 is involved in post-transcriptional regulation (Imai et al., 1998). Another YTP gene, YT521-B, cloned from the brain library of rat, can alter the selection of alternative splicing sites in a concentration-dependent manner (Hartmann et al., 1999). Mmi1, a YTP in the model system Schizosaccharomyces pombe (fission yeast), is important in the molecular regulation of meiosis and controls sexual differentiation in fission yeast (Harigaya et al., 2006; Chen et al., 2011; Hiriart et al., 2012). However, meiosis-specific mRNAs are eliminated by binding to Mmi1 during the vegetative growth phase (Harigaya et al., 2006). Four YTPs occur in humans, and various YTHDF2 loci are associated with human longevity (Cardelli et al., 2006). Recently, researchers are focusing on functions of YTHDF2 and YTHDF3 in binding with the methylated N6 position of selected internal adenines in mRNAs and non-coding RNAs, which affect the translation status and lifetime of mRNA (Li F. et al., 2014; Zhu et al., 2014; Shi et al., 2017). These studies with yeast and animal YTPs have primarily focused on interactions between YTPs and target RNAs, and have not examined their roles in stress responses.

However, the information about plant YTPs are limited in contrast with researches on animals and yeast. The first plant YTP—cleavage and polyadenylation specificity factor 30 (CPSF30)—was discovered in Arabidopsis, where it participates in defense and oxidative-stress responses, as well as cellular signaling (Addepalli and Hunt, 2007; Chakrabarti and Hunt, 2015). Li D. et al. (2014) have identified YTP family members in Arabidopsis and analyzed their functions in abiotic-stress responses. Fifteen members of the Malus YTP family have been reported and their expression has been investigated via quantitative real-time PCR (qRT-PCR) experiments during leaf senescence and under various abiotic-stress conditions (Wang et al., 2014). Although Malus hupehensis (Pamp.) Rehd. YTP1 and YTP2 (MhYTP1 and MhYTP2) have functions in leaf senescence and fruit ripening (Wang et al., 2017), their roles in stressed plants have not been confirmed.

Here, we investigated the responses of MhYTP1 and MhYTP2 to various environmental stimuli in Arabidopsis thaliana, based on their promoter cis-elements. After fusion with the β-glucuronidase (GUS) gene, their promoters were transferred into Arabidopsis plants. We also treated plants that over-expressed MhYTP1 or MhYTP2 and tested the responses of those transgenics to SA, methyl jasmonate (MeJA), ABA, methyl viologen (MV), exposure to 4 or 40°C, salt (NaCl), or mannitol.

There are limited researches on plant YTPs. Although, MhYTP1 and MhYTP2 could promote leaf senescence and fruit ripening (Wang et al., 2017), their roles in stressed plants have not been confirmed. Each of their promoter regions contains several stress-related cis-elements indicating their roles in stress responses. MhYTP1 and MhYTP2 are homologous and share more than 90% similarity (Wang et al., 2017), but they differ in their responses to various environmental stimuli.

The TCA-Element occurs in the MhYTP1 promoter region but not in the MhYTP2 promoter. However, the latter contains a Box-W1 element (fungal elicitor responsive element). When we compared the responses to exogenous applied SA, expression was reduced when GUS was fused to either the MhYTP1 or MhYTP2 promoter, and all OE lines proved resistant to treatment with 10 μM SA. This indicated that the two genes are active in SA signaling. Another important participant in the network of plant pathogen defenses is JA. Here, GUS expression was induced by treatment with MeJA, especially in plants transformed with the MhYTP2 promoter. Moreover, gene overexpression plants, particularly the MhYTP2-OE lines, are more resistant to MeJA. All of these results led us to conclude that MhYTP1 plays a more important role in SA signaling and while MhYTP2 has a greater function in JA signaling.

Three phytohormones—SA, JA, and ethylene (ET)—are crucial to the regulation of signal transduction pathways. However, such control is not achieved through the isolated activation of only a single hormonal pathway but by a complex regulatory network that connects the different pathways. This enables each hormone to either assist or antagonize the others as required to fine-tune the defense response to an individual pathogen. Whereas, the SA-signaling pathway is mainly activated against biotrophic pathogens, the JA-signaling pathway is associated with the development of resistance to necrotrophic pathogens (Glazebrook, 2005; Adie et al., 2007; Verhage et al., 2010). Thus, our findings suggest that MhYTP1 has a leading role in defenses against biotrophic organisms while MhYTP2 functions in resistance to necrotrophic pathogens. The presence of an ET-responsive element (ERE) in the MhYTP2 promoter provides further evidence of a relationship between that gene and hormone signaling, again implying cooperation between JA- and ET-signaling in protecting plants against pathogen infections.

Under various biotic and abiotic stresses, e.g., drought, cold, salinity, and pathogen attacks, ABA signaling regulates the genes responsible for seed germination and stomatal movement (Adie et al., 2007; Beck et al., 2007; Hirayama and Shinozaki, 2007). The promoter regions of MhYTP1 or MhYTP2 do not contain any ABA- or drought-related cis-elements. Nevertheless, GUS expression levels of ABA treated plants transferred with MhYTP1 promoter were induced. However, the GUS expression levels of ABA treated plants transferred with MhYTP2 promoter were not significantly different from that of untreated plants. Overexpression plants are more resistant to ABA treatment. Thus, both of the two genes participated in ABA signaling and MhYTP1 may play a leading role. Then MhYTP1 and MhYTP2 regulated ABA signaling related stress responses.

Almost all promoter regions for Malus domestica YTP genes contain at least one anaerobic stress-related cis-element (Wang et al., 2014). We found that the promoter region of MhYTP1 has an ARE while that region of MhYTP2 have two ARE and a GC-motif. In examining how their expression is regulated when plants are under hypoxia stress due to elevated ROS production, we found that expression of GUS fused with either of those promoters increased after MV treatment. This was particularly true for the MhYTP2 promoter. Likewise, expression of the rat homolog YT521-B is inhibited under oxygen-deficient conditions but is recovered after re-oxygenation (Imai et al., 1998). The first YTP identified in a plant, AtCPSF30, also participates in the response to oxidative stress (Chakrabarti and Hunt, 2015). All of these reports suggest that YTPs are functionally homologous in plants and animals. After growing on normal MS or MS media supplemented with 1 μM MV for 7 days, MhYTP1-OE and, especially, MhYTP2-OE plants performed better that the WT. This difference might have been a result of the MhYTP2 promoter region carrying more anaerobic stress-related cis-elements. Therefore, we can conclude that both genes function in the oxidative stress response, with MhYTP2 having the dominant role.

The presence of HSEs in the MhYTP2 promoter region implies that expression of that gene is induced by changes, especially an increase, in temperature. Consistently, the GUS gene fused with the MhYTP2 promoter was induced by exposure to heat, i.e., 40°C, but not when fused with the MhYTP1 promoter. In contrast, expression of GUS, when fused with the promotor of either gene, was induced by chilling at 4°C. After 4 days of treatment at either temperature, overexpression of these YTP genes conferred resistant to both chilling and heat. This demonstrated that the function of a gene is determined not only by its promoter but also by its structure and that of its protein. As RBPs, the functions of MhYTP1 and MhYTP2 are also determined by their target RNAs. In addition, our transgenic plants also showed greater resistance to high salinity and mannitol (simulated drought) when compared with similarly treated WT plants.

From these results stated above, we can see that MhYTP1 and MhYTP2 participate in multiple stress responses. While these sources of stress can be present simultaneously, so also do MhYTP1 and MhYTP2 interact with each other, making it difficult to identify individual stress responses. One possible outcome of being exposed to multiple stresses is that plants are better able to develop resistance to more than one type of stress (Bowler and Fluhr, 2000). This phenomenon is called cross-tolerance, which is achieved through a powerful regulatory system that allows plants to adapt quickly to a changing environment (Bowler and Fluhr, 2000; Capiati et al., 2006; Fujita et al., 2006; Suzuki et al., 2012). JA, SA, and ET signaling interact with each other and all play roles in responses to both biotic and abiotic stresses. Miura and Tada (2014) have described the importance of SA in the drought response. This hormone also improves the resistance of Hordeum vulgare to water deficits (Bandurska and Stroiński, 2005). The pathogen related (PR) proteins are crucial for resistance against pathogens, and their expression is strongly up-regulated when plants are attacked (Seo et al., 2010). Moreover, many PR genes are induced when plants are exposed to abiotic stress (Seo et al., 2010).

We propose that MhYTP1 and MhYTP2 participate in such a variety of stress responses because of their YTH domains. This type of domain can bind with a short, degenerated, single-stranded RNA sequence motif that can be described only by a weight matrix (Zhang et al., 2010). However, an accurate motif sequence has not yet been confirmed. Thus, target RNAs of MhYTP1 and MhYTP2 are numerous, which may be the reason of their wide-spectrum of cellular functions. The way in which RBPs function in stress responses is determined by their promoter cis-elements, gene structures, protein structures, and target RNAs. Here, we analyzed the promoter regions of MhYTP1 and MhYTP2 and focused only on how each is induced by stresses. More research involving gene and protein structures, as well as target RNAs, is needed if we aim to improve our understanding of the exact roles of these proteins in stress responses.

In summary, we investigated the responses of MhYTP1 and MhYTP2 to various environmental stimuli by studying their promoter cis-elements. Although, MhYTP1 and MhYTP2 are homologous genes in apple, they still differ in their responses to various treatments. Both genes are active in SA, JA, and ABA signaling and play roles in plant responses to oxidative stress, chilling, heat, salinity, and mannitol-induced physiological drought stress. While MhYTP1 has the more dominant role in SA and ABA signaling, MhYTP2 is the main gene involved in JA signaling and oxidative stress responses.

NW: Experiment design and implementation, manuscript composition; TG, PW, XS, YS, XJ, and BL: Experiment implementation and assistance; XG: Purchase and management of reagents, maintenance of laboratory apparatus; FM: Experiment design and implementation, composition and review of manuscript, financial support for experiments and laboratory apparatus.