*2.8. Expression Analysis of MePOD Genes in Response to Various Abiotic and Biotic Stresses and Related Signals*

To test the transcription of *MePOD* genes upon exposure to methyl jasmonate (MeJA), salicylic acid (SA), abscisic acid (ABA), H2O2, salt, osmotic stress (by mannitol treatment), cold stress, and *Xanthomonas axonopodis* pv *manihotis* (*Xam*), nine genes (*MePOD-13, -16, -17, -19, -23, -68, -74, -85, -86*) that were induced by drought stress in at least two tissues in all three varieties were selected for quantitative real-time PCR (qRT-PCR) analysis (Figure 7; Table S6). We define upregulation as: log2-based fold change > 1. With MeJA treatment, the expression of eight genes (*MePOD-13, -17, -19, -23, -68, -74, -85,* and *-86)* was induced, with particularly high levels expressed 24 h after treatment. ABA treatment resulted in increased transcript levels of *MePOD-17* and *-85*. With SA treatment, the expression levels of *MePOD-13, -16, -19, -23, -68, -85,* and *-86* were amplified. H2O<sup>2</sup> treatment led to the induction of *MePOD-13, -17, -23,* and *-85*. Salt treatment induced the expression of *MePOD17* after two and six hours, and three days of treatment, but the gene was repressed after 14 days of treatment. Under osmotic stress, *MePOD-13, -17, -19, -23, -85,* and *-86* were upregulated, among which *MePOD-85* was induced throughout the entire treatment time. In response to cold treatment, *MePOD17* was upregulated at 2 and 15 h. Exposure to the pathogen *Xam* led to the upregulated expression of six *POD* genes (*MePOD-13, -16, -17, -74, -85,* and *-86*) at a minimum of two time points. Together, these results demonstrate that the *POD* genes of cassava respond to multiple stresses and related signals (Figure 7).

*Int. J. Mol. Sci.* **2019**, *20*, x FOR PEER REVIEW 10 of 18

224 **Figure 7.** Expression profiles of cassava *POD* genes in the leaves of Arg7 after exposure to MeJA, SA, 225 ABA, H2O2, salt, osmotic stress (mannitol treatment), cold stress, and *Xam*. Log2-based qRT-PCR fold 226 changes were used to build the heat map with Mev4.9.0 software. The changes in color represent the 227 relative gene expression level. **Figure 7.** Expression profiles of cassava *POD* genes in the leaves of Arg7 after exposure to MeJA, SA, ABA, H2O<sup>2</sup> , salt, osmotic stress (mannitol treatment), cold stress, and *Xam*. Log2-based qRT-PCR fold changes were used to build the heat map with Mev4.9.0 software. The changes in color represent the relative gene expression level.

#### 228 **3. Discussion 3. Discussion**

223

229 Given the significant role of PODs in various physiological processes, including responses to 230 biotic and abiotic stresses, it was necessary to scientifically investigate the potential functions of *POD* 231 genes in cassava, which is an important crop. In this study, we identified 91 PODs in the cassava 232 genome (Figure 1); thus, cassava has more POD members than Arabidopsis but fewer than rice, 233 *Populus trichocarpa, Medicago sativa*, maize, and *Pyrus bretschneideri* [9,24–30]. We found that 92% 234 (84/91) of the MePODs have a molecular mass in the range of 30 to 45 kDa, which is in accordance 235 with previous studies [2,7]. Most of the *POD* genes (89/91) in cassava harbor more than one exon 236 (Figure 3), which is similar to the proportion of single-exon *POD* genes in *Pyrus bretschneideri* and *Zea*  237 *mays* (*PbPRX* (90/94) and *ZmPRX* (89/107), respectively) [29,30]. The similarities in gene structure and 238 motif composition among the members in each MePOD subgroup support the phylogenetic 239 classification presented here. Given the significant role of PODs in various physiological processes, including responses to biotic and abiotic stresses, it was necessary to scientifically investigate the potential functions of *POD* genes in cassava, which is an important crop. In this study, we identified 91 PODs in the cassava genome (Figure 1); thus, cassava has more POD members than Arabidopsis but fewer than rice, *Populus trichocarpa, Medicago sativa*, maize, and *Pyrus bretschneideri* [9,24–30]. We found that 92% (84/91) of the MePODs have a molecular mass in the range of 30 to 45 kDa, which is in accordance with previous studies [2,7]. Most of the *POD* genes (89/91) in cassava harbor more than one exon (Figure 3), which is similar to the proportion of single-exon *POD* genes in *Pyrus bretschneideri* and *Zea mays* (*PbPRX* (90/94) and *ZmPRX* (89/107), respectively) [29,30]. The similarities in gene structure and motif composition among the members in each MePOD subgroup support the phylogenetic classification presented here.

240 The expansion of a gene family primarily occurs via three kinds of modes: segmental duplication 241 of multiple genes, tandem duplication of individual genes, and whole-genome duplication [36,37]. 242 To analyze the duplication modes of the *POD* genes in cassava, we first identified the chromosomal The expansion of a gene family primarily occurs via three kinds of modes: segmental duplication of multiple genes, tandem duplication of individual genes, and whole-genome duplication [36,37]. To analyze the duplication modes of the *POD* genes in cassava, we first identified the chromosomal

243 locations of the *MePOD* genes. Chromosomal mapping revealed that these genes are widely 244 distributed among 17 chromosomes and one scaffold (cassava has 18 chromosomes in total) (Figure

locations of the *MePOD* genes. Chromosomal mapping revealed that these genes are widely distributed among 17 chromosomes and one scaffold (cassava has 18 chromosomes in total) (Figure 4), which is in accordance with the wide chromosomal distribution of PODs in Arabidopsis, rice, *Populus trichocarpa*, maize, and *Pyrus bretschneideri* [9,24,27,29,30]. Secondly, 16 paralogous *POD* genes were characterized in the cassava genome, indicating that tandem duplication contributed to MePOD expansion. Accumulated evidence has demonstrated that duplication events have been important for gene expansion in the POD family. A total of 37 PRX genes in *Populus* and 24 *POD* genes in maize were identified as tandem duplications, further supporting that tandem duplication has been a significant means of *POD* gene expansion [27,29]. Almost all these paralogous MePODs had low or no expression after drought treatment, but 63% (10 out of 16) from the postharvest transcriptome were expressed (Table S6), of which *MePOD-2, -30, -32, -33, -39,* and *-44* were significantly upregulated and *MePOD-34, -56,* and *-62* were repressed at some time point during the PPD process (Figure 6C). These results indicate that most of the *MePOD* genes resulting from tandem duplication-driven expansion are involved in the PPD process of cassava storage roots. Paralogous PRX genes were also found to be involved in other biological processes. In maize, paralogous genes *ZmPRX-26, -42,* and *-75* were induced after NaCl, PEG, SA, or H2O<sup>2</sup> treatment [29]. In Chinese pear, the expression of paralogous genes *PbPRX-42* and *-64* increased during fruit development [30].

The POD family is positively related to the reduced production of hydrogen peroxide and the decreased formation of reactive oxygen species, and the suppression of these species increases plant resistance to stresses [4,5,11,12,19,20]. In this study, the total number of *POD* genes responding to drought (log2-based fold change > 1) was greater in both the roots and leaves of W14 than that in Arg7 and SC124, suggesting the comprehensive activation of PODs in response to drought in W14 (Figure 6B). The wild ancestor W14 has been previously confirmed to be more resistant to drought than the two cultivars SC124 and Arg7 [38,39]. Accumulated evidence suggests that the overexpression of *POD* genes results in increased plant tolerance to drought and osmotic stresses [19,20,40,41]. The activity of POD enzyme was significantly enhanced under drought stress [42]. Consequently, we conclude that the high ratio of MePODs induced by drought in W14 might contribute to its strong drought tolerance.

Previous studies have suggested that ROS production results in the deterioration process in cassava during the postharvest period, and a reduction in ROS accumulation could delay the PPD process [35,43]. The POD family mainly participates in the peroxidative cycle and hydroxylic cycle, resulting in the reduced production of H2O<sup>2</sup> and the decreased formation of ROS [4,5,11,12]. Some PRXs have been shown to change in expression during the fruit storage process [44,45]. The activity of POD enzyme significantly increased during cassava PPD process, suggesting their possible role during the postharvest period of cassava [35,44]. In this study, we found that 78% (71 out of 91) of PODs (log2-based fold change > 1) were upregulated in the storage roots of SC124 (Figure 6C). Interestingly, 13% (12 out of 91) of PODs (log2-based fold change > 1) were induced at all points. Collectively, these results indicate that *MePOD* genes are involved in the PPD process in cassava storage roots.

Previous research has indicated that PODs can extensively participate in plants' responsse to biotic and abiotic stresses [18–23]. Here, we selected nine genes (*MePOD-13*, *-16*, *-17*, *-19*, *-23*, *-68*, *-74*, *-85*, and *-86*) to further examine their expression levels after various treatments (Figure 7; Table S6). These genes are located on different regions of chr7, 13, 3, 17, 8, 15, 10, 9, and 18, respectively (Figure 4). Phylogenetic analysis indicates that *MePOD-16*, *-68*, *-74*, and *-85* belong to subgroup A; *MePOD-17* belongs to subgroup D*;* and *MePOD-13*, *-19*, *-23*, and *-86* belong to subgroup E (Figure 1). The results show that all nine of the analyzed MePODs were upregulated in response to at least two types of treatments. *MePOD17* and *MePOD85* (log2-based fold change > 1) were induced by six treatments (MeJA, salt, cold stress, osmotic stress, ABA, and *Xam* and MeJA, osmotic stress, SA, ABA, H2O2, and *Xam*, respectively); *MePOD13* was upregulated by five treatments (MeJA, osmotic stress, ABA, SA, H2O2, and *Xam*); and *MePOD23* and *MePOD86* were upregulated by four treatments (MeJA, osmotic stress, SA, and H2O<sup>2</sup> and MeJA, SA, H2O2, and *Xam*, respectively). Of these, *MePOD13* and

*MePOD23* were induced after H2O<sup>2</sup> treatment in cassava leaves (Figure 7) but exhibited the opposite trend of expression during the PPD process in storage roots (Figure 6C), suggesting their differential roles in diverse tissues. *MePOD-13, -19, -23, -68,* and *-86* (belonging to subgroup E, except for *MePOD68*) were upregulated by MeJA and SA treatments but downregulated by ABA treatment (Figures 1 and 7). The expression of some PODs has been induced by MeJA and SA treatments in other plant species [2,46]. The opposite direction of expression of these *POD* genes between MeJA and SA treatments and ABA treatment may be due to the antagonism between MeJA/SA and ABA [47,48]. Whereas ABA plays a prominent role in plants' tolerance to drought stress [38], MeJA- and SA-mediated signaling pathways are also activated under drought stress [49,50]. The induction of these genes by MeJA, SA, and drought suggests their possible involvement in MeJA- and SA-mediated drought responses in cassava. The responses of *POD* genes to multiple treatments have been observed in other plants. In Arabidopsis, *AtPrx33* and *AtPrx34* were upregulated after H2O<sup>2</sup> and flg22 treatments [51]. In maize, *ZmPRX-26*, *-42*, and *-71* were induced by H2O2, salt, and PEG treatments [29]. Phylogenetic analysis of MePODs with AtPrx-33 and -34 and ZmPRX-26, -42, and -71 found that MePOD86 shares a close phylogenetic relationship with ZmPRX71 (Figure S1), suggesting their functional conservation in multiple treatments. Multiple stresses, such as cold, salt, or PEG, induced the activity of POD enzyme, demonstrating the response of *POD* genes to environmental stress [52–54]. These results suggest that MePODs participate in the response to multiple stresses or related signals and are candidate targets for the genetic improvement of cassava.
