Transcriptome-Wide Identification and Characterization of the JAZ Gene Family in Mentha canadensis L.

Jasmonate ZIM-domain (JAZ) proteins are the crucial transcriptional repressors in the jasmonic acid (JA) signaling process, and they play pervasive roles in plant development, defense, and plant specialized metabolism. Although numerous JAZ gene families have been discovered across several plants, our knowledge about the JAZ gene family remains limited in the economically and medicinally important Chinese herb Mentha canadensis L. Here, seven non-redundant JAZ genes named McJAZ1–McJAZ7 were identified from our reported M. canadensis transcriptome data. Structural, amino acid composition, and phylogenetic analysis showed that seven McJAZ proteins contained the typical zinc-finger inflorescence meristem (ZIM) domain and JA-associated (Jas) domain as conserved as those in other plants, and they were clustered into four groups (A-D) and distributed into five subgroups (A1, A2, B1, B2, and D). Quantitative real-time PCR (qRT-PCR) analysis showed that seven McJAZ genes displayed differential expression patterns in M. canadensis tissues, and preferentially expressed in flowers. Furthermore, the McJAZ genes expression was differentially induced after Methyl jasmonate (MeJA) treatment, and their transcripts were variable and up- or down-regulated under abscisic acid (ABA), drought, and salt treatments. Subcellular localization analysis revealed that McJAZ proteins are localized in the nucleus or cytoplasm. Yeast two-hybrid (Y2H) assays demonstrated that McJAZ1-5 interacted with McCOI1a, a homolog of Arabidopsis JA receptor AtCOI1, in a coronatine-dependent manner, and most of McJAZ proteins could also form homo- or heterodimers. This present study provides valuable basis for functional analysis and exploitation of the potential candidate McJAZ genes for developing efficient strategies for genetic improvement of M. canadensis.

JA, ABA, and abiotic stress treatments. In addition, our results showed that McJAZ proteins were localized to the nucleus or cytoplasm, and most of them interacted with McCOI1a in a coronatine-dependent manner, and also formed homo-or heterodimers in M. canadensis. Thus, the current results enrich our knowledge of JAZ family genes and provide a valuable foundation for functional analysis and exploitation of the candidate McJAZ genes for genetic improvement of M. canadensis.

Identification and Molecular Cloning of McJAZ Genes from M. canadensis
To identify the putative McJAZ family genes in M. canadensis, in total, twelve wellknown JAZ protein sequences from A. thaliana [19] and nine reported JAZ protein sequences from A. annua [24] were collected and used as queries to perform a BLASTP search based on the previous transcriptome data (SRP132644) at the National Center for Biotechnology Information (NCBI) [67]. Finally, we identified and annotated eight non-redundant fulllength McJAZ genes in the M. canadensis transcriptome database. Next, we obtained seven amplicons corresponding to seven distinct full-length McJAZ genes through PCR amplification with gene specific primers. Subsequently, the purified amplicons were ligated with pClone007 Blunt Simple vector, and the resulting products were transformed into Escherichia coli DH5α and sequenced, respectively. After this, we successfully cloned seven full-length McJAZ genes, and denoted them as McJAZ1-McJAZ7 ( Figure 1 and Table 1). As expected, all of these McJAZ proteins contained a N-terminus conserved TIFY domain and a C-terminus Jas domain (also named CCT_2 motif). These findings were further corroborated by using the online SMART tools (http://smart.embl-heidelberg.de/, accessed on 25 April 2021) with the default parameters ( Figure 1 and Table 1). Next, sequence analysis showed that the length of the McJAZ genes coding sequences (CDs) ranged from 372 (McJAZ7) to 1041 (McJAZ3) base pairs (bp), which was similar in length to the known AtJAZ CDs ranging from 396 (AtJAZ8) to 1059 (AtJAZ3) bp in A. thaliana [19]. Furthermore, these candidate McJAZ genes encoded predicted products which varied in length from 123 (McJAZ7) to 346 (McJAZ3) amino acids (aa) residues, with molecular weights (M W ) ranging from 13.87 kilodalton (kDa) (McJAZ7) to 36.89 kDa (McJAZ3), and the isoelectric point (pI) ranged from 6.12 (McJAZ2) to 9.96 (McJAZ6) ( Table 1). Notably, the pI features of most of McJAZ proteins except for McJAZ2 were greater than 7, indicating that most of the McJAZ proteins were basic proteins. Taken together, all of these results provide an earlier basis for further characterization of JAZ family proteins or related networks in M. canadensis.   1 Genbank Accession number; 2 the length of coding sequences; 3 the length of amino acid sequences; 4 the molecular weight of protein (kDa); 5 the isoelectric point (pI); 6 the core sequence feature of TIFY motif; 7 the position of Jas domain within the amino acid sequences; aa: amino acid.

Sequence Alignment and Phylogenetic Analysis of McJAZ Genes
To analyze the similarity of the homologous sequences among the candidate McJAZ proteins, multiple sequence alignment analysis of the seven McJAZ proteins or the pairwise McJAZ proteins was carried out. The results showed that except for McJAZ6 and McJAZ7, which shared 80.80% identity at the amino acid sequence level (Supplemental Table S1), the rest of pairwise McJAZ proteins contained differential amino acid identity

Sequence Alignment and Phylogenetic Analysis of McJAZ Genes
To analyze the similarity of the homologous sequences among the candidate McJAZ proteins, multiple sequence alignment analysis of the seven McJAZ proteins or the pairwise McJAZ proteins was carried out. The results showed that except for McJAZ6 and McJAZ7, which shared 80.80% identity at the amino acid sequence level (Supplemental Table S1), the rest of pairwise McJAZ proteins contained differential amino acid identity ranging from 6.94% to 27.80% shared identity, analyzed by local DNAMAN 6.0 software (Figure 2A and Supplemental Table S1). In contrast to the sequences overall, the typical TIFY and Jas domains were conserved in all McJAZ proteins in M. canadensis (Figure 2A). Similarly, the yield sequence logos drawn using the MEME tool also showed highly conserved amino acid sequence within the TIFY and Jas domains ( Figure 2B). In addition, we also found that McJAZ6 and McJAZ7 possessed the typical EAR (LELRL) motif in the N-terminal region that is widely present in JAZ proteins across different plants ( Figure 2A). Interestingly, evidence suggests that the TIFY or Jas motif is responsible for mediating protein-protein interactions for itself or other partners [19,34]. Overall, these results showed the possibility that the function of the seven McJAZ proteins may be conserved within M. canadensis. served amino acid sequence within the TIFY and Jas domains ( Figure 2B). In addition, we also found that McJAZ6 and McJAZ7 possessed the typical EAR (LELRL) motif in the Nterminal region that is widely present in JAZ proteins across different plants ( Figure 2A). Interestingly, evidence suggests that the TIFY or Jas motif is responsible for mediating protein-protein interactions for itself or other partners [19,34]. Overall, these results showed the possibility that the function of the seven McJAZ proteins may be conserved within M. canadensis. To understand the evolutionary and phylogenetic relationships of the JAZ family genes between M. canadensis and other known species including A. thaliana, O. sativa, and A. annua, a phylogenetic tree was constructed based on an unrooted neighbor-joining (NJ) method using twelve AtJAZ proteins from A. thaliana, fifteen OsJAZ proteins from O. sativa, nine AaJAZ proteins from A. annua, and seven McJAZ proteins from M. canadensis. The results showed that all the JAZ proteins were clustered into four groups (Groups A-D) and further classified into A1, A2, B1-B3, C, and D subgroups according to the reported phylogenetic analysis of AtJAZ proteins in Arabidopsis [34] Figure 3). To understand the evolutionary and phylogenetic relationships of the JAZ family genes between M. canadensis and other known species including A. thaliana, O. sativa, and A. annua, a phylogenetic tree was constructed based on an unrooted neighbor-joining (NJ) method using twelve AtJAZ proteins from A. thaliana, fifteen OsJAZ proteins from O. sativa, nine AaJAZ proteins from A. annua, and seven McJAZ proteins from M. canadensis. The results showed that all the JAZ proteins were clustered into four groups (Groups A-D) and further classified into A1, A2, B1-B3, C, and D subgroups according to the reported phylogenetic analysis of AtJAZ proteins in Arabidopsis [34]

Tissue Expression Patterns of McJAZ Genes
To further investigate expression patterns of McJAZ genes in M. canadensis, we analyzed the expression levels of each McJAZ gene across different tissues ( Figure 4A) including roots, stems, young leaves, and flowers using quantitative real-time PCR (qRT-PCR) assays. The results showed that all McJAZ genes were broadly expressed in all evaluated tissues and displayed the highest expression in flowers ( Figure 4). More specifically, expression patterns fluctuated across tissues and were significantly different for each McJAZ gene and could be classified into three types. As shown in Figure 4B

Tissue Expression Patterns of McJAZ Genes
To further investigate expression patterns of McJAZ genes in M. canadensis, we analyzed the expression levels of each McJAZ gene across different tissues ( Figure 4A) including roots, stems, young leaves, and flowers using quantitative real-time PCR (qRT-PCR) assays. The results showed that all McJAZ genes were broadly expressed in all evaluated tissues and displayed the highest expression in flowers ( Figure 4). More specifically, expression patterns fluctuated across tissues and were significantly different for each McJAZ gene and could be classified into three types. As shown in Figure 4B with the highest levels in flowers, followed by young leaves, stems, and relatively low expression in roots compared with other tissues. In addition, McJAZ3 was found to be most highly expressed in flowers, and its expression was successively reduced in young leaves, roots, and stems. We next found that McJAZ7 showed highest expression in flowers, moderately expression in stems and young leaves, and low expression in roots ( Figure 4B). These results suggest that the seven McJAZ genes studied here were constitutively expressed in all four M. canadensis tissues, and most of them were highly abundant in the flower and young leaf tissues. low expression in roots compared with other tissues. In addition, McJAZ3 was found to be most highly expressed in flowers, and its expression was successively reduced in young leaves, roots, and stems. We next found that McJAZ7 showed highest expression in flowers, moderately expression in stems and young leaves, and low expression in roots ( Figure 4B). These results suggest that the seven McJAZ genes studied here were constitutively expressed in all four M. canadensis tissues, and most of them were highly abundant in the flower and young leaf tissues.

Expression Profiles of McJAZ Genes in Response to JA and ABA Treatments
Numerous reports have shown that MeJA and ABA play a vital role in regulating plant development, stress response, and various physiological processes through controlling the expression patterns of downstream genes [8,18,68,69]. Hence, to next investigate the potential roles of McJAZ genes in response to MeJA and ABA, qRT-PCR was performed to analyze the relative expression profiles of these McJAZ genes in M. canadensis that had been treated with MeJA or ABA. Based on qRT-PCR results, we found that the expression pattern of each McJAZ gene was variable under MeJA or ABA treatment (

Expression Profiles of McJAZ Genes in Response to JA and ABA Treatments
Numerous reports have shown that MeJA and ABA play a vital role in regulating plant development, stress response, and various physiological processes through controlling the expression patterns of downstream genes [8,18,68,69]. Hence, to next investigate the potential roles of McJAZ genes in response to MeJA and ABA, qRT-PCR was performed to analyze the relative expression profiles of these McJAZ genes in M. canadensis that had been treated with MeJA or ABA. Based on qRT-PCR results, we found that the expres-sion pattern of each McJAZ gene was variable under MeJA or ABA treatment ( Figure 5). Under MeJA treatment, the transcript levels of seven McJAZ genes were distinctly upregulated ( Figure 5A), similar to previously reported transcriptome data (Supplemental Figure S1) [67]. In general, McJAZ1, McJAZ2, McJAZ6, and McJAZ7 transcript levels displayed a rapid increase, peaking at 1 h after MeJA exposure, and then gradually decreased at 3, 6, and 12 h, and returned to approximately control levels (CK 0 h) after prolonged MeJA exposure (24 h). In contrast, McJAZ3, McJAZ4, and McJAZ5 expression increased at 1 h after MeJA exposure, and significantly increased until a peak after 3 h, then gradually decreased after prolonged MeJA exposure (6, 12, and 24 h) ( Figure 5A).
In addition, we found that except for McJAZ4, the transcripts of McJAZ genes were also dynamically affected by exogenous ABA treatment. Among them, McJAZ2, McJAZ5, and McJAZ6 transcript levels were significantly up-regulated at 1, 3, and 24 h for McJAZ2, at 1, 12, and 24 h for McJAZ5, and at 1 and 24 h for McJAZ6 after ABA exposure ( Figure  5B). In particular, the expression levels of McJAZ2, McJAZ5, and McJAZ6 reached their highest levels at 14, 21, and 7 times higher than the control (CK 0 h) at 1 h respectively. Additionally, the expression of McJAZ1 was slightly up-regulated at 1 h, but down-regulated subsequently (3, 6, 12, and 24 h), and reached its lowest level at 6 h after ABA exposure. The transcript level of McJAZ3 was not obvious different at 1, 3, and 24 h, but was down-regulated at 6 and 12 h, compared with control (CK 0 h). McJAZ7 expression was slightly decreased after 1, 3, 6, and 12 h of ABA exposure, and then returned to a level comparable to that seen at the control time point (CK 0 h). These results revealed that McJAZ genes expression responds to JA and ABA, and may be involved in JA or ABA signaling pathway in M. canadensis.   Figure 5B). In particular, the expression levels of McJAZ2, McJAZ5, and McJAZ6 reached their highest levels at 14, 21, and 7 times higher than the control (CK 0 h) at 1 h respectively. Additionally, the expression of McJAZ1 was slightly up-regulated at 1 h, but down-regulated subsequently (3, 6, 12, and 24 h), and reached its lowest level at 6 h after ABA exposure. The transcript level of McJAZ3 was not obvious different at 1, 3, and 24 h, but was downregulated at 6 and 12 h, compared with control (CK 0 h). McJAZ7 expression was slightly decreased after 1, 3, 6, and 12 h of ABA exposure, and then returned to a level comparable to that seen at the control time point (CK 0 h). These results revealed that McJAZ genes expression responds to JA and ABA, and may be involved in JA or ABA signaling pathway in M. canadensis.

McJAZ Genes Are Involved in Response to Drought and Salt Stress
Abiotic stress is known to severely limit plant growth and development, and affect multiple physiological processes, and the species' geographical distribution. To further examine the potential functions of McJAZ genes in response to abiotic stresses, qRT-PCR assays were employed to investigate the expression patterns of Abiotic stress is known to severely limit plant growth and development, and affect multiple physiological processes, and the species' geographical distribution. To further examine the potential functions of McJAZ genes in response to abiotic stresses, qRT-PCR assays were employed to investigate the expression patterns of McJAZ genes in M. canadensis treated with drought and salt stress. The results showed that, under drought stress treatment, the expression levels of the seven McJAZ genes were significantly up-regulated at the individual time point, and expression levels varied by gene ( Figure 6A) Figure 6A).
In addition, seven McJAZ genes had a fluctuating and significantly different transcriptional response to salt treatment ( Figure 6B   were up-regulated, and peaked at more than 7-fold or 9-fold compared with the control at 3 h, gradually decreased after prolonged salt treatment (6 and 12 h), then returned to levels similar to control (CK 0 h) at 24 h after salt treatment. These results implied that most of the McJAZ genes were sensitive to drought and salt stress treatments, and are likely involved in drought and salt stress related signaling networks in M. canadensis.

Subcellular Localization of McJAZ Proteins
To further determine subcellular localization of McJAZ proteins, subcellular localization was first predicted using an online tool (http://cello.life.nctu.edu.tw/, accessed on 30 April 2021). Preliminary results showed that seven McJAZ proteins were predicted to be targeted to nuclei (Table 1). Accordingly, we next constructed an McJAZs-GFP (Green Fluorescence Protein) fusion expression vector under control of the CaMV 35S promoter, and subsequently these fusion proteins or the GFP control were transiently expressed in Nicotiana benthamiana leaf cells. The subcellular localization results showed that, in contrast to GFP, which was localized uniformly throughout the cell, most of the McJAZ proteins were exclusively located in the nucleus marked with the red arrow in the leaves of N. benthamiana (Figure 7), which is consistent with the previously predicted subcellular localization (Table 1). Meanwhile, McJAZ2 and McJAZ3 were also observed with cytoplasmic localization apart from nuclear ( Figure 7). This phenomenon is consistent with recent reports that some JAZ proteins of Lycoris aurea including LaJAZ1 show both nuclear and cytosolic localization [70].

Subcellular Localization of McJAZ Proteins
To further determine subcellular localization of McJAZ proteins, subcellular localization was first predicted using an online tool (http://cello.life.nctu.edu.tw/, 30 April 2021). Preliminary results showed that seven McJAZ proteins were predicted to be targeted to nuclei (Table 1). Accordingly, we next constructed an McJAZs-GFP (Green Fluorescence Protein) fusion expression vector under control of the CaMV 35S promoter, and subsequently these fusion proteins or the GFP control were transiently expressed in Nicotiana benthamiana leaf cells. The subcellular localization results showed that, in contrast to GFP, which was localized uniformly throughout the cell, most of the McJAZ proteins were exclusively located in the nucleus marked with the red arrow in the leaves of N. benthamiana (Figure 7), which is consistent with the previously predicted subcellular localization (Table 1). Meanwhile, McJAZ2 and McJAZ3 were also observed with cytoplasmic localization apart from nuclear ( Figure 7). This phenomenon is consistent with recent reports that some JAZ proteins of Lycoris aurea including LaJAZ1 show both nuclear and cytosolic localization [70].

McJAZ Proteins Interact with McCOI1a
Previous studies report that JAZ proteins play a vital role in JA response through interacting with the JA receptor, CORONATINE INSENSITIVE 1 (COI1) in a JA-Ile-or coronatine (COR, a functional mimic of bioactive JA-Ile)-dependent manner rather than a JA-or MeJA-dependent manner [19,23,71,72]. Some JAZ proteins in non-model plants may also interact with AtCOI1 [71]. To determine the role of McJAZ proteins in the JA response in M. canadensis, we first examined the interaction between AtCOI1 and McJAZ

McJAZ Proteins Interact with McCOI1a
Previous studies report that JAZ proteins play a vital role in JA response through interacting with the JA receptor, CORONATINE INSENSITIVE 1 (COI1) in a JA-Ile-or coronatine (COR, a functional mimic of bioactive JA-Ile)-dependent manner rather than a JA-or MeJA-dependent manner [19,23,71,72]. Some JAZ proteins in non-model plants may also interact with AtCOI1 [71]. To determine the role of McJAZ proteins in the JA response in M. canadensis, we first examined the interaction between AtCOI1 and McJAZ proteins by yeast two-hybrid (Y2H) assays. As shown in Figure 8, all of the yeast transformants grew normally on the control medium (DDO, Figure 8A), whereas none of the McJAZ proteins interacted with AtCOI1 regardless of whether the selection medium (QDO) was supplemented with the JA-Ile mimic, COR ( Figure 8B,C). Subsequently, given that proteinprotein interaction might occur in a context-dependent manner in different plant species, we next cloned the homologous protein of AtCOI1, designated this as proteins by yeast two-hybrid (Y2H) assays. As shown in Figure 8, all of the yeast transformants grew normally on the control medium (DDO, Figure 8A), whereas none of the McJAZ proteins interacted with AtCOI1 regardless of whether the selection medium (QDO) was supplemented with the JA-Ile mimic, COR ( Figure 8B,C). Subsequently, given that protein-protein interaction might occur in a context-dependent manner in different plant species, we next cloned the homologous protein of AtCOI1, designated this as

Homo-and Heterodimeric Interaction of McJAZ proteins
Our previous results showed that all of the McJAZ proteins have a conserved TIFY domain (Figures 1 and 2), and previous evidence has also demonstrated that the TIFY domain is involved in regulating homo-and heterodimeric interactions among JAZ proteins [34]. Thus, we further examined the possibility that the seven McJAZ proteins could form homo-or heterodimers by testing 49 combinations among the seven McJAZ proteins in the Y2H assays ( Figure 9A

Homo-and Heterodimeric Interaction of McJAZ Proteins
Our previous results showed that all of the McJAZ proteins have a conserved TIFY domain (Figures 1 and 2), and previous evidence has also demonstrated that the TIFY domain is involved in regulating homo-and heterodimeric interactions among JAZ proteins [34]. Thus, we further examined the possibility that the seven McJAZ proteins could form homo-or heterodimers by testing 49 combinations among the seven McJAZ proteins in the Y2H assays ( Figure 9A

Discussion
JAs play diverse functions across the JA signaling pathway in plants [8]. Notably, JAZ proteins, core negative regulators of JA signaling, are widely reported to play a vital role in JA-mediated biological processes including plant growth, defense against pathogen infection, insect attack, abiotic stress, and plant specialized metabolism [20,41,73]. To date, JAZ family proteins have been systematically identified in various plants including A. thaliana [33], T. aestivum [37], H. brasiliensis [39], and L. aurea [70]. However, research regarding the JAZ gene family has not yet been reported in the economically and medicinally important plant species M. canadensis. Based on the available transcriptome data [67], we performed a comprehensive identification of JAZ proteins in M. canadensis and further investigated their sequence characteristics and protein structure, evolutionary relationships, expression profiles, subcellular locations, and protein-protein interactions. This work provided a basis for subsequent functional analysis of JAZ genes to deepen our understanding of JA signaling in the Lamiaceae family, especially in M. canadensis.
We isolated seven McJAZ genes from M. canadensis, and found that the length of CDs in the McJAZ gene family ranged from 372 to 1041 bp ( Figure 1 and Table 1), which was in line with the reported CDs length in the AtJAZ gene family [19]. Moreover, most McJAZ proteins were basic proteins with pI values greater than 7, except for McJAZ2 with a pI value of 6.12, which was attributed to a higher proportion of acidic amino acids in McJAZ2 such as 21 aspartic acids and 12 glutamic acids than those in the rest of McJAZ proteins, and this phenomenon was ubiquitous in JAZ proteins of other plants like Camellia sinensis [74]. In general, JAZ protein at least contains one conserved TIFY and a typical Jas domain at its Nand C-terminal, and some JAZ proteins including Arabidopsis AtJAZ7 and AtJAZ8 have an LxLxL type of EAR motif at the N-terminus [33]. Here, although the amino acid similarity among McJAZ proteins is low (Figure 2 and Supplemental Table S1), each identified McJAZ protein contained one conserved N-terminus TIFY and a C-terminus Jas domain, which further indicated evolutionary conservation of these structures. Furthermore, McJAZ6 and McJAZ7 have an additional EAR motif (LxLxL) in the N-terminus (Figure 2). TIFY has dual functions in regulating homo-and heterodimeric interactions among different JAZ proteins, and recruiting TPL/TPRs by interacting with NINJA protein [22]. The Jas domain is responsible for interacting with COI1 and various TFs including the core regulator MYC2 to complete the JA response [19], and is involved in subcellular localizations of JAZ proteins [75]. The EAR motif in some JAZ proteins is used to directly recruit the TPL-coupled inhibitory complex to attenuate JA response in the absence of NINJA [36]. Given that protein function is tightly associated with conserved domain structure, all of these results suggest that these seven McJAZ proteins may share some common strategy to regulate the perception of JA and JA response in M. canadensis, much like JAZ proteins in different plant species. Notably, the Jas domain of the JAZ protein usually contained a conserved sequence (SLX 2 FX 2 KRX 2 RX 5 PY), but our results showed that McJAZ2 has a truncated Jas motif (DLPIARKNSLARFLEKRRDR) without the X 5 PY sequence (Figure 2). Previous studies found that some deviation existed in the Jas domain owing to alternative splicing (AS). For example, AtJAZ10.3, a splice variant of AtJAZ10 protein in Arabidopsis, has a truncated Jas motif (DLPIARRKSLQRFLEKRKER) lacking LVSTSPY (X 5 PY) sequence in the whole Jas domain of AtJAZ10.1 [76]. Thus, whether McJAZ2 is a splice variant like AtJAZ10.3 merits further investigation.
The tissue expression patterns of target genes are closely tied to gene function. Specifically, JAZ family genes were constitutively and differentially expressed in different tissues of various plants such as wheat [37], rubber tree [39], and L. aurea [70]. Furthermore, the widespread expression of JAZ genes in different plant tissues is involved in regulating plant growth and development [1,41]. Here, we found that seven McJAZ genes were constitutively expressed in the tested tissues, and showed a variable expression profile (Figure 4). These results suggest that different McJAZ genes may play an important role in regulating the growth and development of different tissues in M. canadensis. It is intriguing that seven McJAZ genes showed the highest expression in flowers over all other tissues studied ( Figure 4). Among these, the transcripts of McJAZ5/6/7 reached 33-fold, 180-fold and 70-fold within flowers relative to root tissue. Moreover, earlier reports showed that JAZ genes played roles in flower and stamen development in Arabidopsis and in rice [1,19]. Thus, McJAZ genes may share similar functions with the above-mentioned JAZ genes in flower development. In addition, we found that McJAZ1/2/4/5/6 showed similar expression patterns ( Figure 4B), which, taken with the conservation of TIFY and Jas domains among them, suggests some functional redundancy in the McJAZ1/2/4/5/6 in M. canadensis.
The defense-related hormones JA and ABA induce fluctuation in transcript abundance of JAZ genes in various plants, and this critical transcriptional regulation of JAZ genes is connected with the potential roles of JAZ including JA-and ABA-mediated plant development, defense, and specialized metabolism [19,29,37,38,45,47]. For instance, JA treatment and other environmental elicitors rapidly induce JAZ gene expression, which in turn attenuates response to JA [20]. In addition, OsJAZ1 was induced by ABA and involved in negative regulation of drought tolerance in rice [47]. Here, the seven studied  Figure 5B). Notably, cross-talk between JA and ABA has been reported in plants [29]. Our study shows that McJAZ2, McJAZ5, and McJAZ6 are also co-induced by MeJA and ABA, indicating that these McJAZ genes may regulate cross-talk between these two phytohormones. In addition, studies show that the JAZ family genes exhibited varied expression in response to one or more abiotic stressors in different plants. JAZ family genes are differentially induced by drought, salt, and low temperature in wheat, rice, Camellia sinensis, and Brassica rapa, and upregulation of these genes is involved in salt and drought tolerance [37,38,74,77]. Here, we found that the identified seven McJAZ genes showed differential upregulation to cope with drought stress, and In plants, JAZ proteins participate in JA signaling pathways by interacting with JA receptor COI1 or themselves [19,34]. In our study, some of the McJAZ proteins were found to interact with McCOI1a in a COR-dependent manner (Figure 8). Furthermore, McJAZ proteins showed variable capacities to form homo-or heterodimers via interacting with each other in a COR-independent manner (Figure 9). These results suggest that the functions of McJAZ gene in M. canadensis might be similar to those in Arabidopsis. Previous studies have shown that AtJAZ1 can interact with AtCOI1 or SlCOI1 in the presence of COR [32]. However, the identified McJAZ proteins only interacted with McCOI1a rather than AtCOI1 in the presence of COR, suggesting that regulation of the interactions of JAZ-COI1 may differ across species. In addition, the phylogenetic relationship among the JAZ genes reveals the putative function of them as homologous genes often share similar biological roles. In this study, we found that McJAZ1 and McJAZ2 clustered with AtJAZ1/2, or AtJAZ5/6 into the subgroup B2 or B1. McJAZ3, McJAZ4, and McJAZ5 clustered with AtJAZ3/4 or AtJAZ10 into the subgroup A2 or A1. McJAZ6/7 clustered with AtJAZ7/8 into the group D ( Figure 3). It was clear that AtJAZ proteins typically interacted with AtMYC2 to regulate diverse physiological functions including plant metabolism and defense [78,79]. Addition-ally, AtJAZ3 interacted with AtSPL9 to regulate plant resistance to insect herbivores [80]. AtJAZ1/4/9 interacted with AtICE1/2 to regulate freezing tolerance in Arabidopsis [25]. AtJAZ1/8/11 interacted with and attenuated the transcriptional functions of MYB21 and MYB24 to regulate stamen development [81]. AtJAZ1/2/5/6/8/9/10/11 also regulated JA-mediated anthocyanin biosynthesis and trichome development by interacting with the WD-Repeat/bHLH/MYB complexes in Arabidopsis [14]. Given the functional specificity, diversity, and redundancy of AtJAZ genes in Arabidopsis [82], McJAZ proteins may regulate plant development, defense, and specialized metabolism including EO biosynthesis by interacting with the as-yet unidentified McMYC2 or additional TFs in M. canadensis ( Figure 10). Further investigation needs to be conducted to determine the complete regulatory network of McJAZ proteins.
genes reveals the putative function of them as homologous genes often share similar biological roles. In this study, we found that McJAZ1 and McJAZ2 clustered with AtJAZ1/2, or AtJAZ5/6 into the subgroup B2 or B1. McJAZ3, McJAZ4, and McJAZ5 clustered with AtJAZ3/4 or AtJAZ10 into the subgroup A2 or A1. McJAZ6/7 clustered with AtJAZ7/8 into the group D ( Figure 3). It was clear that AtJAZ proteins typically interacted with AtMYC2 to regulate diverse physiological functions including plant metabolism and defense [78,79]. Additionally, AtJAZ3 interacted with AtSPL9 to regulate plant resistance to insect herbivores [80]. AtJAZ1/4/9 interacted with AtICE1/2 to regulate freezing tolerance in Arabidopsis [25]. AtJAZ1/8/11 interacted with and attenuated the transcriptional functions of MYB21 and MYB24 to regulate stamen development [81]. AtJAZ1/2/5/6/8/9/10/11 also regulated JA-mediated anthocyanin biosynthesis and trichome development by interacting with the WD-Repeat/bHLH/MYB complexes in Arabidopsis [14]. Given the functional specificity, diversity, and redundancy of AtJAZ genes in Arabidopsis [82], McJAZ proteins may regulate plant development, defense, and specialized metabolism including EO biosynthesis by interacting with the as-yet unidentified McMYC2 or additional TFs in M. canadensis ( Figure 10). Further investigation needs to be conducted to determine the complete regulatory network of McJAZ proteins.  Notably, JA treatment promoted biosynthesis of EO in two Mentha plants [58,67], and EO from Mentha plants could be used to control plant bacterial disease caused by A. tumefaciens in tomato [83], enhance neighboring plants' resistance to insect herbivores [56], and act as a natural insecticide for control of insects including Tribolium castaneum, Lasioderma serricorne, and Liposcelis bostrychophila adults [84]. This suggests that manipulation of JA signaling in Mentha plants might have dual roles in regulating EO biosynthesis and affecting the anti-insect or anti-pathogen ability of neighboring plants. JAZ proteins, the key components of the JA signaling pathway, acted as important regulators involved in mediating biosynthesis of specialized metabolism such as anthocyanin [14,85], tanshinone [49], and artemisinin [24] through employing the different regulatory mechanisms. However, the detailed molecular mechanisms of JA-mediated EO biosynthesis by McJAZ genes or a McJAZ-related network remain uninvestigated within the mint family. Thus, subsequent research into functional analysis of McJAZ genes or McJAZ-associated regulatory networks in Mentha plants will deepen our understanding of the regulatory mechanisms of JA-mediated EO biosynthesis, and help us to develop efficient strategies for genetic improvement of EO content, and even enhance plant defense against pathogen infection, herbivorous insect attack, and abiotic stress in M. canadensis.

Plant Materials and Treatments
M. canadensis was conserved and planted at the Germplasm Nursery in the Institute of Botany, Jiangsu Province, and Chinese Academy of Sciences, Nanjing, Jiangsu Province [67], for use in each assay described here. Cuttings of M. canadensis were propagated by inserting into water until roots emerged, and then planted into plastic pots containing a mixture of nutrient soil and vermiculite (2:1, v/v). For in vitro culture of N. benthamiana, seeds were sown on soil and 10-day-old seedlings were transplanted into plastic pots containing a mixture of nutrient soil and vermiculite (2:1, v/v). Next, M. canadensis and N. benthamiana plants were grown in a light chamber at 23 ± 2 • C with the desired light regime, and a 16-h light/8-h dark photoperiod.
For hormone and salt stress treatments, two-week-old M. canadensis seedlings were sprayed with 100 µM MeJA (Sigma-Aldrich (Shanghai) Trading Co. Ltd., Shanghai, China), 20 µM ABA (Sigma-Aldrich (Shanghai) Trading Co. Ltd., Shanghai, China), and 150 mM NaCl for 1, 3, 6, 12, and 24 hours (h) according to methods adapted from previous studies [12,86]. For drought treatment, two-week-old hydroponically cultured M. canadensis seedlings were exposed to dry filter paper for 1, 3, 6, 12, and 24 h at room temperature according to methods adapted from previous studies [87]. Untreated plants were used as the control (labeled as CK 0 h). Then, leaves of the control and treated seedlings, of matching positions, were collected at the indicated time point and frozen in liquid nitrogen and stored at −80 • C for further analysis.

Isolation and Cloning of M. canadensis McJAZ Family Genes
The isolation of M. canadensis McJAZ genes were performed as described previously [88]. In short, the transcriptome data (SRP132644) of M. canadensis was downloaded from the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih. gov/, accessed on 20 March 2020). The well-known 12 A. thaliana [34], and 9 A. annua JAZ protein sequences were downloaded from TAIR (https://www.arabidopsis.org/, accessed on 20 January 2021) or the annotated A. annua genome [89], respectively. Then, each of them was employed as queries to search for JAZ proteins in M. canadensis protein databases by the local BLASTP method with an e-value of 1 × 10 −5 . The screened sequences were submitted to ORF Finder (https://www.ncbi.nlm.nih.gov/orffinder, accessed on 15 March 2021) to gain the complete coding sequences, and the candidate sequences were subjected to SMART (http://smart.embl-heidelberg.de/, accessed on 25 April 2021) tools with default parameters to substantiate the existence of the conserved TIFY and Jas domains [33], and determine their precise positions. The candidate McJAZ gene sequences with both TIFY and Jas domains were selected for further analysis.
To clone the McJAZ genes, the young leaves of two-week-old M. canadensis were collected for RNA extraction using the reagent Trizol (Cat # 9108, Takara Biomedical Technology (Beijing) Co., Ltd., Beijing, China), and cDNA was synthesized using PrimeScript RT reagent Kit with gDNA Eraser (RR047A, Takara Biomedical Technology (Beijing) Co., Ltd., Beijing, China). The candidate McJAZ genes were amplified using PCR with defined primers (Supplemental Table S2) in a 25 µL volume including 1 µL 5x cDNA template, 0.5 µL of each forward and reverse primer, 5 µL 5x PrimeSTAR DNA Polymerase buffer, 2 µL PrimeSTAR ® DNA Polymerase, and 16 µL sterile ddH 2 O. The amplification procedure was as follows: 95 • C for 10 min followed by 35 cycles of denaturation for 15 s at 95 • C, annealing at designated temperature for 20 s, and extension at 72 • C for 1 min per kilobase (kb). Subsequently, the full-length coding sequences of McJAZ genes were gel-purified using Gel DNA Mini Purification Kit (TSINGKE Biotechnology, Beijing, China), and the purified amplicons were ligated with pClone007 Blunt Simple vector (TSV-007B, TSINGKE Biotechnology, Beijing, China), and then the resulting products were transformed into Escherichia coli DH5α and sequenced in TSINGKE Biotechnology Co., Ltd (TSINGKE Biotechnology, Chengdu, China).

Sequence Analysis and Phylogenetic Tree Analysis
The sequence and physio-chemical properties of candidate McJAZ proteins were predicted using the online ExPASy server (http://us.expasy.org/tools, accessed on 3 May 2021) [90], including the length of amino acid (aa) sequences, molecular weight (M W ), and theoretical protein isoelectric points (pIs). The CELLO v.2.5 tool (http://cello.life.nctu.edu. tw/, accessed on 30 April 2021) was employed to predict the subcellular localization of candidate McJAZ proteins.
Multiple sequence alignment analysis of the seven McJAZ proteins was performed to analyze the characteristics of the conserved sequences using Clustal X2 software with default settings [91]. Meanwhile, multiple sequence alignment of the pairwise McJAZ proteins to identify the amino acid sequence similarity was done using DNAMAN 6.0 (Lynnon Biosoft, Quebec City, QC, Canada) software with default parameters. The frequency of the conserved amino acids at each site within the TIFY and Jas domain was analyzed using online MEME tools (http://meme-suite.org/tools/meme, accessed on 8 May 2021) using the default settings. The composition and position of the conserved motifs, and the sequence logos of the TIFY and Jas domains were created by online MEME tools. Next, to check out the evolutionary relationships of McJAZ family proteins, the JAZ protein sequences including 12 A. thaliana [20], 15 O. sativa [38] and 9 A. annua [24] homologous proteins were selected and acquired from the databases Phytozome v12.1 (https://phytozome.jgi.doe.gov/, accessed on 20 January 2021) and NCBI, respectively. The MEGA 5.0 program was adopted to construct the phylogenetic tree using a neighborjoining (NJ) approach with the following parameters: Pairwise deletion and 2000 bootstrap replicates [92].

Subcellular Localization of McJAZ Proteins
The full-length coding sequences of each McJAZ gene without its stop codon was PCR-amplified using gene-specific primers (Supplemental Table S2), and the amplicons were subcloned into the N-terminal side of the pCAMBIA1300 vector containing the green fluorescent protein (GFP) reporter gene to produce the recombinant fusion construct pCAMBIA1300-McJAZs-GFP under the control of the CaMV 35S promoter. Then the fusion constructs were transformed into Agrobacterium strains GV3101 through the conventional freezing-thawing method, and the GV3101 strains harboring each pCAMBIA1300-McJAZs-GFP or empty pCAMBIA1300-GFP were transiently infiltrated into 5-week-old N. benthamiana leaves according to previous report [24]. GFP signals were detected 2 to 3 days after infiltration using a confocal laser-scanning microscope (Leica TCS SP5-II, Wetzlar, Germany).

RNA Extraction and Quantitative Real-Time PCR (qRT-PCR) Analysis
To study tissue expression patterns, different tissues of 3-month-old M. canadensis, including roots, stems, young leaves, and flowers were harvested and stored at −80 • C. Total RNA of the different tissues (labeled in Figure 4A) and the aforementioned samples was isolated using the total RNAprep pure Extraction Kit (Tiangen Biotech, Beijing, China) according to the manufacturer's recommendations. cDNA was synthesized from 1.0 µg total RNA using the PrimeScript 1st Strand cDNA Synthesis Kit (Takara Biomedical Technology (Beijing) Co., Ltd., Beijing, China) according to the manufacturer's instructions. The qRT-PCR was performed using the SYBR Premix Ex Taq II Kit (Tiangen Biotech, Beijing, China) according to the manufacturer's instructions and methods previously described [93]. All PCR reactions were performed on at least three replicates under identical conditions. Expression level of each gene was normalized to the expression of β-actin gene (KM044035) used as internal control in the mint family in the previous study [61], and the relative mRNA expression for each gene was calculated using the 2 −∆∆Ct method [94]. All primers are listed in Supplemental Table S2.

Yeast Two-Hybrid (Y2H) Experiments
Yeast two-hybrid assays were used to test the paired physical interactions among McJAZ proteins, or the interactions between McJAZ proteins and AtCOI1, or its homolog McCOI1a. The full-length coding sequences of McJAZ genes were cloned into the prey vector pGADT7 (Activation domain, AD) or bait vector pGBKT7 (Binding domain, BD) (Takara, Japan), respectively, and the full-length coding sequences of McCOI1a and AtCOI1 were amplified using gene-specific primers (Supplemental Table S2), and inserted into the bait vector pGBKT7 (BD). The indicated combinations were co-transformed into the yeast strain AH109 using the lithium acetate method according to the manufacturer's instructions (Cat # 630440, Takara Biomedical Technology (Beijing) Co., Ltd., Beijing, China), and bait-only or prey-only was tested with empty AD or BD as negative controls.
To investigate protein-protein interactions, the transformed AH109 yeast cells were first cultivated in liquid control medium SD/-Leu/-Trp (DDO) over 24 h, and then 5 µL of each yeast suspension was taken to inoculate the solid control medium SD/-Leu/-Trp (DDO) and selection medium SD/-Leu/-Trp/-Ade/-His (QDO) with or without 35 µM coronatine (COR), a structural and functional mimic of JA-Ile obtained from several strains of Pseudomonas syringae according to previous studies [19,72,95]. The yeast cells were grown at 30 • C for 4 days before imaging.

Statistical Analysis
All experiments were conducted with at least three biological replicates. Significant differences between the control versus each treatment were analyzed by Student's t-test using GraphPad Prism 5 software. The values are mean ± standard error (SE) for replicates in each group. * p values ≤ 0.05 or ** p values ≤ 0.01 were considered as significant or extremely significant.

Accession Numbers
The gene sequences data mentioned in this article can be found in the NCBI database under the accession numbers: