*3.6. Heterologous Expression of SlBAG9 Conferred Sensitivity to Drought, Salt, and ABA*

Our previous studies showed that *SlBAG9* with a higher expression level at the transcriptional and protein levels in response to high temperature stress was involved in the negative regulation of thermotolerance [22,23]. The expression analysis of *SlBAG9* showed that it might be involved in the response to drought, salt, and ABA (Figure 5). To further investigate the biological functions of *SlBAG9*, heterologous overexpression lines of slbag9 in Arabidopsis (2–12, 4–9) were used here, which have been shown to be two homozygous lines [22]. We evaluated the seed germination and growth on 1/2 MS medium containing 200 and 300 mM mannitol, 175 and 200 mM NaCl, and 1.0 and 1.5 μM ABA treatments. No significant difference between the overexpression lines and WT was observed on 1/2 MS medium without treatment. However, the seed germination rate and seedling growth of *SlBAG9*-overexpressing lines were inhibited by mannitol, salt, and ABA compared to that of WT (Figures 6 and S2). For germination, after five days on MS medium containing 300 mM mannitol, about 85% of wild-type seeds, but less than 60% and 30% of the transgenic seeds (2–12, 4–9, respectively) germinated (Figure 6A). After ten days, more than 90% wild-type cotyledons, but less than 15% transgenic cotyledons turned green (Figure 6B), and the post-germination growth of transgenic materials was much worse than wild-type ones (Figure 6C). Two different treatments (200 and 300 mM mannitol) showed a certain concentration effect (Figure 6). For seedling growth, seven-day-old seedlings of wild-type, 2–12, and 4–9 were transferred to 1/2 MS containing 300 mM mannitol for

14 days and the growth pattern was observed. Under normal conditions, no significant difference was found in the aspects of seedlings growth of all the seedlings. Drought inhibited the growth of all seedlings, but the inhibition degree of 2–12, and 4–9 lines was greater than that of wild-type seedlings (Figure S2). A similar phenotype was also observed in NaCl treatment (Figure 6, Figure S2). These results demonstrated that *SlBAG9* conferred plant more sensitive to drought and salt stress, negatively regulating osmotic stress.

**Figure 6.** The increased sensitivity of *SlBAG9*-overexpressing Arabidopsis to mannitol, NaCl, and ABA stress. (**A**) The germination rate of *SlBAG9*-overexpressing lines (2–12 and 4–9) and wild type (WT) on 1/2 MS medium containing 200 and 300 mM mannitol (M1, M2), 175 and 200 mM NaCl (N1, N2), and 1.0 and 1.5 μM ABA (A1, A2) treatments for 5 days. (**B**) Cotyledon greening rate of *SlBAG9* overexpressing lines and WT on 1/2 MS medium containing 200 and 300 mM mannitol (M1, M2), 175 and 200 mM NaCl (N1, N2), and 1.0 and 1.5 μM ABA (A1, A2) treatment for 10 days. Error bars indicate the SD of three replicated experiments. (**C**) The seedling growth of *SlBAG9*-overexpressing lines and WT Arabidopsis on 1/2 MS plates containing 300 mM mannitol (M2), 175 mM NaCl (N1), and 1.5 μM ABA (A2) treatments for 10 days.

ABA is an important mediator of drought and salt and senescence [26]. To further study whether the ABA was involved in *SlBAG9*-mediated response, we observed the phenotype of *SlBAG9*-overexpressing lines grown under ABA treatment. In total, 100% of wild-type seeds, but about 40% transgenic seeds germinated after five days on 1/2 MS medium supplemented with 1.5 μM ABA (Figure 6A). More than 80% cotyledons of wild-type cotyledons, but less than 50% of transgenic lines turned green as a result of ABA treatment for seven days (Figure 6B,C). For seedling growth, ABA treatment decreased leaf chlorophyll content, while the leaves of the WT exhibited an obvious stay-green phenotype (Figure S2).

#### *3.7. SlBAG9 Downregulated Stress/ABA-Responsive Genes*

To better understand the mechanism of *SlBAG9*-mediated osmotic stress, mRNA levels of two ABA-responsive genes *ABI3*, and *RD29A* and four stress-responsive genes *DREB2A*, *P5CS1*, *FSD1*, and *CAT1* were determined by qRT-PCR. The RNA transcript levels of the six genes were not significantly altered under normal growth conditions (Figure 7). However, the expression levels of these genes were upregulated in both WT and transgenic plants under osmotic stress or ABA treatment, but the increase was not as pronounced in transgenic plants as in WT plants (Figure 7). These results suggested that these genes in the *SlBAG9*-overexpressing plants are more sensitive to osmotic stress than WT plants, and that *SlBAG9* may be involved in the regulation of stress-regulated gene expression. The expression levels of two key ROS scavenging enzyme gene *FSD1*, and *CAT1* were also regulated by *SlBAG9*, indicating that *SlBAG9* modulates the ROS scavenging system. In view of this, it was speculated that the ROS scavenger-related enzyme activity was regulated by *SlBAG9*. Therefore, in the next section we further analyze the possible effect of *SlBAG9* on the oxidative damage and the activities of SOD and CAT.

**Figure 7.** Relative expression of stress/ABA-responsive genes in *SlBAG9*-overexpressing Arabidopsis exposed to mannitol, NaCl, and ABA stress. The seven-day-old Arabidopsis plantlets of wild-type, 2–12, and 4–9 were transferred to 1/2 MS containing 0, 300 mM mannitol, 175 mM NaCl, and 1.5 μM ABA for three days. Total RNAs were extracted, and qRT-PCR analyses were performed. Error bars indicate the SD of three replicated experiments.

#### *3.8. SlBAG9 Overexpression Aggravated Oxidative Damage under Drought, Salt, and ABA*

Stress conditions trigger the accumulation of ROS molecules, resulting in oxidative damage to plant organelles [27]. To uncover whether *SlBAG9* was involved in drought, salt, and ABA-induced oxidative stress, the H2O2 level was estimated. Excessive accumulation of H2O2 was observed in the *SlBAG9*-overexpressing lines compared with the wild-type plants under drought, salt, and ABA conditions (Figure 8A). MDA is considered to be an effective marker of membrane damage caused by oxidative stress [28]. In accordance with the stressed phenotype, *SlBAG9*-overexpressing line 2–12 and 4–9 accumulated more MDA compared with the wild type under drought, salt, and ABA treatment (Figure 8B). These results indicated that *SlBAG9* can sensitively respond to drought, salt, and ABA to induce H2O2 generation and cause membrane lipid peroxidation, resulting in hypersensitivity responses. Since the above data suggested that *SlBAG9* mediates oxidative damage, and downregulate ROS scavenger-related gene *FSD1* and *CAT1*, the activities of SOD and POD were determined. There was little difference in SOD and CAT activities in WT, 2–12, and 4–9 genotypes under normal conditions. However, the SOD and CAT activities were lower

in the 2–12 and 4–9 lines compared with the WT (Figure 8C,D). This result indicated that *SlBAG9* overexpression is implicated in ROS clearance and reduces tolerance to drought, salt, and ABA-induced oxidative stress by regulating antioxidant enzyme activity, which could support different change patterns of H2O2 and MDA (Figure 8).

**Figure 8.** *SlBAG9* overexpression aggravated oxidative damage in Arabidopsis under mannitol, NaCl, and ABA treatment. The seven-day-old Arabidopsis plantlets of wild-type, 2–12, and 4–9 were transferred to 1/2 MS containing 0, 300 mM mannitol, 175 mM NaCl, and 1.5 μM ABA for three days for physiological change evaluation. The H2O2 content (**A**), MDA content (**B**), SOD activity (**C**), and CAT activity (**D**) were analyzed quantitatively in WT and SlBAG9-overexpressing lines 2–12 and 4–9. Error bars indicate the SD of three independent experiments.

#### **4. Discussion**

Cultivated tomato is easily affected by various environmental factors. Researchers have been studying potential resistance genes in plants for a long time. With the available tomato genome, the study of gene function is becoming more and more important [8]. Our previous research made us very interested in the tomato *SlBAG* gene family [22,29]. Many studies have shown that *BAG* gene plays an important role in plant growth, development, and stress response (reviewed by Thanthrige et al. [7]). Therefore, it is necessary to identify *SlBAG* genes and their biological function in tomato. Lately, Irfan et al. [24] identified 11 BAG genes using the tomato database (ITAG2.4). However, the database now includes three versions (2.4, 3.0, 4.0), in which different BAG gene information is presented. The present study aimed to a comprehensive genome-wide functional characterization of *SlBAG* genes and proteins in tomato. We obtained 10 BAG genes through sequence alignment and multi database comparison, cloned all ten CDS sequences, and finally determined the information of these genes (Table 1 and Table S2). Compared with the newly published genetic information obtained by Irfan et al. [24], *SlBAG6*, *SlBAG8*, *SlBAG10*, *SlBAG4*, and *SlBAG1* have new CDS sequences and eight gene structural sequences were new in addition to *SlBAG7*, and *SlBAG9*.

For the ten BAG proteins in this study, their protein structures and evolutionary relationships with BAG proteins in tomato, Arabidopsis and rice were analyzed. The proteins were divided into two groups, and the difference is mainly at their N-terminal. In addition to BAG domain, the first group has a UBL domain. A UBL domain can interact with 26S proteasome and is an indispensable part of BAG1 in stress response [30]. UBL domain in the group I suggests that they may also participate in the degradation of some proteins as molecular bridges. The second group has a specific CaM-binding domain IQ motif near the BD domain in plants, indicating that it may be involved in unique biological processes [31]. Irfan et al. [24] found that SlBAG7 and SlBAG9 had IQ domain. In the present study, three SlBAGs (SlBAG4, SlBAG7, SlBAG9) in the group II have IQ domain, indicating that their biological functions may be related to the Ca2+ signal. IQ motif can bind CaM and affect the formation of complex between CaM and targeted protein [32]. In vitro studies have shown that Ca2+ can affect the binding affinity of AtBAG6 and CaM and regulate the process of cell death mediated by AtBAG6. The CaM-binding motif and the BD are required for AtBAG6-mediated cell death [32]. As a signaling hub, AtBAG5 connects the Ca2+ signaling network with the Hsc70 chaperone system to regulate plant senescence. The IQ motif mutant retains the association between AtBAG5 with Hsc70 while disrupting the association of AtBAG5 with apo-CaM [11]. It is possible that the increase in Ca2+ in the mitochondrial matrix may protect mitochondria from senescence through a combination of Ca2+ and apo-CAM, so as to promote the release of Hsc70 from CaM/AtBAG5/Hsc70 signal complex and to inhibit ROS production [11,12].

The BAG family is widely distributed in the plant kingdom (Figure 1). The BAG family also exists in various plant tissues and organs. Results from many experiments have shown that the BAG gene family plays an important role in plant growth and development [3]. *AtBAG4* and *AtBAG6* expressions were detected in the root, stem, leaf, and flower of Arabidopsis. In the whole development process, *AtBAG2* and *AtBAG6* genes are expressed in various tissues in overlapping or specific expression pattern [33]. Arabidopsis knockout mutant *atbag4* or *atbag6* has the phenotype of early flowers and multi branching inflorescences, with a shortened life cycle and early aging. It was found that the rosette diameter of the 4-week-old *atbag2* mutant was larger than that of the wild type [34]. Arabidopsis plants overexpressing *AtBAG6* are shorter than wild-type plants [3]. The tissue-specific expression experiment of rice showed that *OsBAG1*, *OsBAG3*, and *OsBAG4* had the highest expression in roots, stems, and internodes, indicating that they may be involved in cell elongation and expansion. Rice OsBAG4 and EBR1 form a protein complex, which makes EBR1 control its protein stability level through ubiquitination of OsBAG4, inhibit growth and development, resulting in plant dwarfism [17]. Irfan et al. [24] showed that several *SlBAG* genes such as *SlBAG1*, *SlBAG3*, *SlBAG6*, and *SlBAG9* had differential expression during fruit development, which suggested that they might have a role in fruit development as well. In this study, many *SlBAGs* in tomato showed specific expression patterns in organs, indicating their important roles (Figure 4). Recently, He et al. [8] showed that *BAG2* and *BAG5b* were highly expressed in tomato stem and flower. Here, the corresponding name was *SlBAG7* and *SlBAG9*, respectively, which was not only highly expressed in flowers, but also in fruits (Figure 4).

The regulatory elements in plant promoters play an important regulatory role at the transcriptional level [35]. Here, it was found that in all *SlBAG* promoters there are a series of elements related to abiotic stress and hormones, such as MYC, MBS, DRE, HSE, W-box, and ABA and SA- responsive cis-elements (Figure S1). These results were consistent with the previous study that most of the same cis elements were found in 2000 bp upstream regions of *SlBAG* genes [24] and ~1000 bp upstream regions of all Arabidopsis BAG family genes, suggesting that they play a role in coping with different environmental stresses such as cold, drought and high salinity [4]. Take W-box as an example, AtBAG7 translocates from the ER to the nucleus, where it interacts with the transcription factor WRKY29, which then binds to the W-box in the promoter of *AtB*AG7 to initiate the transcription of *AtBAG7* and other chaperones to promote stress tolerance [36]. The accumulated evidence shows that BAG expression can be regulated by various abiotic stresses [3,7,24,36]. Accordingly, here we monitored the transcriptional response of tomato *SlBAG* genes with the exposure to various stress conditions including drought, salt, HT, cold, as well as ABA and H2O2 signals (Figure 5). Plant *BAG* gene expression is related to its function to some extent. Cold upregulated the expression of *AtBAG4* and *AtBAG4* overexpression increased tobacco plants tolerance to cold, salt, UV, and oxidative stress [3]. Heat stress significantly upregulated *AtBAG6* gene and protein level and the *atbag6* mutant is sensitive heat stress [4,33]. Transgenic rice plants overexpressing salt-induced *OsBAG4* showed tolerance to NaCl stress [25]. *AtBAG6* transcript levels were significantly upregulated by H2O2 [3]. The expression of several *SlBAG* genes was also induced by ACC, the precursor of ripening hormone ethylene and ABA, suggesting that *SlBAG* genes are potentially involved in the fruit ripening regulation and stress response [24]. In this study, some *SlBAG* genes also showed similar expression patterns. However, the involvement of H2O2 in the BAG-mediated biological function is unclear. The above results suggested that *SlBAG* family is involved in the response of tomato plants to abiotic stresses such as salt, drought, cold and HT, and ABA and H2O2 signals may be involved in these pathways.

*SlBAG9* was noteworthy because in our previous studies it showed higher gene expression level and protein abundance under high temperature stress [22]. Overexpression of *SlBAG9* decreased the tolerance to HT [22,29]. Cis-elements and expression analysis indicated that *SlBAG9* may be involved in drought, salt, and ABA stress (Figures 5 and S1). This study further showed that *SlBAG9*-overexpressing Arabidopsis exhibited increased sensitivity to mannitol, salt, and ABA treatment (Figures 8 and S1). In terms of salt stress, many studies have shown that BAG can positively regulate plant salt tolerance. Arabidopsis with overexpression of *TaBAG* and *TaBAG2* showed significant enhancement of salt tolerance [6]. The *atbag4* mutant was more sensitive to salt stress [3]. Transgenic rice plants heterologously expressing *AtBAG4* showed higher salt tolerance than WT [37]. Recently, Wang et al. [25] reported that OsBAG4 functioned as a bridge between OsMYB106 and OsSUVH7 under salt stress to regulate OsHKT1;5 expression, so as to improve salt tolerance. Pan et al. [38] showed that salt suppressed *BAG6* and *BAG7* expression but addition of ACC (1-aminocyclo-propane-1-carboxylic acid) in the salt treatment could re-activate *BAG6* and *BAG7* expression, indicating that BAG genes are involved in the process of plant cell death induced by salt stress. The negative regulation mechanism of *SlBAG9* on salt tolerance remains to be studied. As far as drought and ABA are concerned, there are few studies on BAG gene function so far. Arabidopsis leaves with low-level *AtBAG4* overexpression appeared to be drought tolerant [3]. AtBAG4 interacted with potassium influx channel protein KAT1 in guard cells to regulate stomatal movement [39]. However, our observation was consistent with the latest data from Arabidopsis *AtBAG2* and *AtBAG6* [33]. Germination of *atbag2*, *atbag6*, and *atbag2atbag6* seeds was less sensitive to ABA compared to WT, whereas *AtBAG2* and *AtBAG6* overexpression lines showed the opposite results for ABA. The survival rate of *atbag2*, *atbag6*, and *atbag2atbag6* plants was higher than that of the WT under drought stress. In addition, these mutants showed differential expression of several stress and ABA-related genes and low ROS levels after drought and ABA treatment [33]. In this study, *ABI3*, *RD29A*, *DREB2A*, and *P5CS1* expression of transgenic plants was lower than that of WT plants under osmotic stress. In *Arabidopsis thaliana,* these genes had all been shown to be inducible by drought, salinity, or ABA. The higher expression of these stress genes was related to plant tolerance [40]. All these data indicated that the decreased transcription levels of these genes in *SlBAG9*-overexpressing lines lead to enhanced sensitivity to osmotic stress, which might be mediated by ABA. There are three points of view that support our hypothesis. First, the *SlBAG9* expression was induced by exogenous ABA (Figure 6). Second, *SlBAG9* overexpression in Arabidopsis resulted in hypersensitivity to ABA (Figures 7 and S2). Third, under ABA and osmotic stress, the *SlBAG9* overexpression significantly downregulated the expression levels of ABA signaling pathway genes such as *ABI3* and *RD29A* (Figure 8), which were reported to play positive regulators in ABA-associated abiotic stress [41]. The decrease in *ABI3* and *RD29A* expression of *SlBAG9* overexpression lines might be one of the reasons for the increased sensitivity to ABA and osmotic stress. The specific signaling pathway of ABA mediated by *SlBAG9* remains to be unveiled.

It is known that a variety of environmental stresses can promote ROS production and the senescence process will also increase the accumulation of ROS. Therefore, the regulation of ROS production plays a key role in senescence and stress response [42]. It has been reported that *atbag2*, *atbag6*, and *atbag2atbag6* seedlings lines accumulated lower H2O2 when treated with drought and ABA [33]. The transgenic plants overexpressing *AtBAG5* showed greater H2O2 accumulation than WT, indicating that *AtBAG5* is involved in leaf senescence by regulating the production of ROS. In the present study, overexpression of *SlBAG9* in Arabidopsis could induce the excess production of H2O2 in response to drought, salt, and ABA (Figure 8), suggesting that *AtBAG5* may be involved in regulating these stress responses though the proliferation of ROS. Excessive production of ROS will lead to oxidative stress and cell death in growing plants [43] H2O2 and MDA content is recorded with the extent of oxidative stress [28]. Consistent with the results of H2O2, *SlBAG9* overexpressing seedlings exhibited significantly induced H2O2 and MDA accumulation compared to the WT seedlings after stress treatment. (Figure 8). These results corresponded well with the phenotype of these *SlBAG9* overexpression plants in response to drought, salt, and ABA, which suggested that *SlBAG9* accelerate H2O2 excess production, leading to a deeper degree of oxidative damage. Correspondingly, gene expression analyses showed *SlBAG9* overexpression downregulated the expression of key ROS scavenger-related genes *FSD1* and *CAT1* (Figure 7). *FSD1* encodes a chloroplast/nuclei/cytosol localized SOD that utilizes Fe as the cofactor (FeSOD) and FSD1 presents osmoprotection in Arabidopsis [44]. *CAT1* encodes a peroxisomal catalase (CAT1), which is implicated in the drought and salt stress responses [45]. SOD and CAT scavenge ROS by converting superoxide to H2O2 and H2O2 to oxygen and water, sequentially. There is a closely correlation between the expression of both two genes and enzyme activity. Li et al. [46] found that the expression of *CAT1* and *FSD1* were upregulated in *PpDHN*-overexpressing plants under drought and salt stress and elevated levels of SOD and CAT enzyme activities were detected accompanying the trends in expression of these genes. In the transgenic overexpressing *IpASR*, *CAT1* and *FSD1* expression and SOD and CAT enzyme activities showed the same trend under salt and drought stress [47]. In the current study, it was clearly shown that the expression of *FSD1* and *CAT1* (Figure 7) and the activities of SOD and CAT (Figure 8C,D) were not altered among wild-type and over-expressing *SIBAG9* lines; however, a difference was only detected under stress conditions or in response to exogenous ABA supply. Taken together all the above indicate that *SIBAG9* is involved in regulation of stress responses, which needs to be further explored. ABA signaling pathway is an important way for plants to deal with abiotic stress and ABA also regulates senescence [26]. Whether ABA was involved in SlBAG9-mediated antioxidant protection pathway still needed further study. Taken together, it was suggested that the increased sensitivity of *SlBAG9* overexpression to drought, salt, and ABA might be related to the oxidative damage regulated by antioxidant enzyme system such as SOD and CAT. This complex mechanism needs to be deeply investigated in the future.
