Genome-Wide Identification and Expression Analysis of the Dof Transcription Factor in Annual Alfalfa Medicago polymorpha

The Dof transcription factor is a plant-specific transcription gene family that plays various biological functions in plant development and stress response. However, no relevant research has been conducted on Medicago polymorpha. Here, 36 MpDof genes were identified in the M. polymorpha genome and further divided into 10 groups based on the comparative phylogenetic analysis. The essential information of MpDof genes, such as chromosomal localization, gene structure, conserved motifs, and selective pressures were systematically analyzed. All 36 MpDof genes were predicted to contain more cis-acting elements related to hormone response. MpDof24 and MpDof25 were predicted to interact with MpDof11 and MpDof26 to involve in the photoperiod blooms process. The MpDof genes showed a diverse expression pattern in different tissues. Notably, MpDof29 and MpDof31 were specifically expressed in the large pod and root, respectively, suggesting their crucial role in the pod and root development. qRT-PCR analysis indicated that the expression levels of MpDof10, MpDof25, MpDof26, and MpDof29 were obviously up-regulated under drought, salt, and cold stress. Collectively, genome-wide identification, evolutionary, and expression analysis of the Dof transcription gene family in M. polymorpha will provide new information to further understand and utilize the function of these Dof genes in Medicago plants.


Introduction
Transcription factors (TFs), also known as trans-acting factors, interact with cis-acting elements in the promoter region to regulate gene expression [1]. DNA-binding One Zinc Finger (DOF) is a plant-specific TF and plays a crucial role in the process of plant growth and development. The N-terminal of the Dof proteins has a highly conserved single zinc finger DNA-binding domain consisting of 50-52 amino acid residues, which can specifically bind to 5 -(T/A) AAAG-3 sequences in the target gene promoter [2,3]. The C-terminal domain of Dof proteins can interact with various proteins and participate in the activation of gene expression, which makes Dof protein show a variety of functions, including tissue development, seed germination, light regulation, and response to biotic and abiotic stresses [4,5].
The first Dof gene was cloned from Zea mays [6]. To date, numerous Dof TFs have been predicted or cloned from higher plants, algae, and bryophytes. The number of Dof TFs varied among different species. The highest number of 117 Dof genes were identified in Brassica napus [7], while only one Dof gene was identified in the Chlamydomonas reinhardt [8]. It thus can be seen that the Dof gene family has expanded dramatically during the evolution from lower plants to higher plants, which may be related to the adaptation to complex environmental conditions. For example, Arabidopsis Dof5.6 has been shown to be involved in vascular tissue development and interfascicular cambium formation [9]. pression patterns. Our results provide a valuable basis for further understanding the gene expansion and evolution model of the plant Dof gene family and analyzing the potential function of the Dof gene in plant abiotic stress regulation.

Identification and Characteristics of the MpDof Gene Family
A total of 36 Dof genes of M. polymorpha were identified by using 108 Dof protein sequences of O. sativa, Arabidopsis, and M. truncatula as a query to perform Blastp, and MpDof01 was renamed to MpDof36, as shown in Table 1. In addition, the physicochemical properties, including the chromosome location, protein length, protein hydrophilic, isoelectric point, protein molecular weight, DNA molecular weight, and orthologous genes of 36 MpDof proteins, were detected. The amino acid sequence lengths of these MpDofs ranged from 157 to 495 amino acids (Table 1). Correspondingly, the protein molecular weights centered in the range of 17. 66-54.15 kDa. The GRAVY values of −0.987 to −0.447 represented that 36 MpDof proteins were hydrophilic. Isoelectric point analysis showed that the 7 MpDof proteins were acidic (pI < 7), and the remaining 25 MpDof proteins were basic proteins (pI > 7).

Comparative Phylogenetic Analysis of Dof Genes in M. polymorpha
To comprehensively understand the number difference of Dof genes among four species, phylogenetic relationship analysis and orthologous group classification were performed. The predicted Dof proteins were assigned into ten groups named A, B1, B2, C1, C2.1, C2.2, C3, D1, D2, and E as in the Arabidopsis classification ( Figure 1 and Table S1). The results also indicated that the number of groups varies greatly among the four tested species. Group B1, B2, C1, C2.1, C2.2, and D1 were found in all four species (Figure 2). Groups A, C3, D2, and E were lineage-specific, while Group E only existed in O. sativa. Group B1 and D1 contained more genes than other groups, with 21 and 35 Dof genes, respectively. The Dof gene numbers in these groups were distinct among different species, for example, two, five, and nine Dof genes of group A in Arabidopsis, M. polymorpha, and M. truncatula, respectively. In addition, the phylogenetic analysis based on the genomes of four tested species supported the closest genetic relationship between M. polymorpha and M. truncatula. The phylogeny also showed that the divergence time between M. polymorpha and M. truncatula had been estimated to be~15.3 MYA. The above results suggested that the Dof gene family expansions and homologous Dof genes gain or lose in M. polymorpha, O. sativa, Arabidopsis, and M. truncatula during evolution. Furthermore, the analysis of dicotyledonous model plant Arabidopsis, monocotyledonous model plant rice, alfalfa model plant M. truncatula, and M. polymorpha can provide a basis for studying the evolution of the Dof gene in dicotyledonous plants.

Gene Structure and Conserved Motifs Analysis of the MpDofs
To further explore the structural discrepancy among MpDofs, exon-intron structures and conserved motifs were analyzed. The phylogenetic relationships of 36 MpDofs are shown in Figure 3A, and exon-intron structures are shown in Figure 3B. As the results show, the vast majority of genes in a given subgroup revealed a similar exon-intron structure. Thirty-six MpDof genes contained zero or one intron. In subgroup I, nine MpDof genes had one intron, and two MpDof genes (MpDof05 and MpDof31) had none. Seven MpDof genes included in subgroup II had no introns and MpDof07 had one intron. Among the 17 subgroups III and VI MpDof genes, 4 (MpDof13, MpDof28, MpDof19, and MpDof21) contained none, and the remaining 13 contained 1.

Gene Structure and Conserved Motifs Analysis of the MpDofs
To further explore the structural discrepancy among MpDofs, exon-intron structures and conserved motifs were analyzed. The phylogenetic relationships of 36 MpDofs are shown in Figure 3A, and exon-intron structures are shown in Figure 3B. As the results show, the vast majority of genes in a given subgroup revealed a similar exon-intron structure. Thirty-six MpDof genes contained zero or one intron. In subgroup I, nine MpDof genes had one intron, and two MpDof genes (MpDof05 and MpDof31) had none. Seven MpDof genes included in subgroup II had no introns and MpDof07 had one intron. Among the 17 subgroups III and VI MpDof genes, 4 (MpDof13, MpDof28, MpDof19, and MpDof21) contained none, and the remaining 13 contained 1.

Gene Structure and Conserved Motifs Analysis of the MpDofs
To further explore the structural discrepancy among MpDofs, exon-intron structures and conserved motifs were analyzed. The phylogenetic relationships of 36 MpDofs are shown in Figure 3A, and exon-intron structures are shown in Figure 3B. As the results show, the vast majority of genes in a given subgroup revealed a similar exon-intron structure. Thirty-six MpDof genes contained zero or one intron. In subgroup I, nine MpDof genes had one intron, and two MpDof genes (MpDof05 and MpDof31) had none. Seven MpDof genes included in subgroup II had no introns and MpDof07 had one intron. Among the 17 subgroups III and VI MpDof genes, 4 (MpDof13, MpDof28, MpDof19, and MpDof21) contained none, and the remaining 13 contained 1.  As shown in Figure 3C, the number of motifs located on each MpDof gene varied from 1 to 11. All 36 MpDof proteins shared a common motif 1, which corresponds to the Dofconserved domain (Figures 4 and S1). Multiple sequence analysis indicated that 36 MpDof proteins contained a highly conserved CX2CX21CX2C single zinc-finger structure, which is significant for the function of MpDof genes. A total of 27 of 54 amino acid residues were highly conserved (100% identical in all 36 MpDof proteins) amino acids CPRC-S-TKFCY-NNY-QPR-FC-C-R-WT-GG-R-G-R. Moreover, some of the MpDof proteins had extra motifs, for example, only MpDof24 and MpDof25 contained motif 7 and motif 12, which may be related to specific functions. The motif components of MpDofs from subgroup I were the most complex: motif 3, motif 12, motif 19, motif 5, motif 18, motif 14, motif 6, motif 4, motif 17, motif 20, motif 10, motif 7, motif 2, and motif 9 were specific for them. Subgroup III Dof proteins possessed relatively simple motif components compared with subgroup I; motif 11 and motif 8 were specific to it. Members belonging to subgroups II and IV contained only motif 1, except for MpDof02 and MpDof20, which both contained motif1 and motif 16. Thus, the results suggested that differences in exon-intron distribution and motif components between subgroups may lead to functional diversity. group I were the most complex: motif 3, motif 12, motif 19, motif 5, motif 18, motif 14, motif 6, motif 4, motif 17, motif 20, motif 10, motif 7, motif 2, and motif 9 were specific for them. Subgroup III Dof proteins possessed relatively simple motif components compared with subgroup I; motif 11 and motif 8 were specific to it. Members belonging to subgroups II and IV contained only motif 1, except for MpDof02 and MpDof20, which both contained motif1 and motif 16. Thus, the results suggested that differences in exon-intron distribution and motif components between subgroups may lead to functional diversity.

The Location on Chromosome, Gene Duplication, and Syntenic Analysis
The location of MpDof genes on chromosomes and synteny in the M. polymorpha genome are shown in Figure 5 and Table S2. The distribution of MpDof family members on 7 chromosomes was uneven and each chromosome possessed MpDof genes ranging from 2 to 7 ( Figure 5). Chromosomes 4 and 7 had the fewest number of MpDof genes with two, while chromosomes 1, 3, and 6 had the largest number of Dof genes with 7. We found a total of 15 pairs of MpDof gene duplications which included 13 segmental and 2 tandem duplications ( Figure 5). We can see that some MpDof genes, such as MpDof11, have been duplicated three times. Tandem duplications led to MpDof gene clusters or hot regions, such as the MpDof gene clusters located on chromosomes 5 and 7, whereas segmental duplications caused homologous genes between chromosomes that may expand the number of MpDof genes groups. Moreover, the synteny relationship between M. polymorpha and M. truncatula was stronger ( Figure 6), while the synteny relationship with rice was the simplest. This can provide information on the origin and evolution of species.  Table 2). In addition, we can also observe that the Ka and Ks values of tandem-duplicated gene pairs (MpDof24/MpDof25; MpDof35/MpDof36) were the lowest, and the divergent time of them was later than 13 segmental-duplicated gene pairs.    The non-synonymous substitution rate (Ka), synonymous substitution rate (Ks), and Ka/Ks value of 15 duplicated gene pairs were calculated to report the natural selection pressure acting on Dof gene pairs and to estimate the date of MpDof gene duplication. We found that the Ka/Ks ratios of all 15 MpDof duplicated gene pairs were less than one ( Table 2). This suggested that these duplicated gene pairs were under negative selection pressure. The Ka values were used to calculate the date on which 15 MpDof gene pairs duplicated. The results showed that the duplicated events of 15 MpDof gene pairs spanned approximately from 1.79 to 93.36 MYA (millions of years ago) ( Table 2). In addition, we can also observe that the Ka and Ks values of tandem-duplicated gene pairs (MpDof24/MpDof25; MpDof35/MpDof36) were the lowest, and the divergent time of them was later than 13 segmental-duplicated gene pairs.

Cis-Acting Elements Analysis
We mainly analyzed light-responsive, hormone-responsive, abiotic-stress-responsive, and development-related elements (Table S3), which are the key regulatory factors involved in plant growth and development or other important physiological processes. The cisacting elements were distributed among these predicted MpDof genes. Analysis of the cis-acting elements can be conducted to estimate the putative functions of MpDof genes. Compared with cis-acting elements related to abiotic stress and development, there were more hormone-responsive elements in the promoter regions of 36 MpDof genes, including those related to auxin, gibberellin, salicylic acid, abscisic acid, and methyl jasmonic acid, while abscisic acid-and methyl jasmonic acid-responsive elements were the most abundant ( Figure 7 and Table S3). It is suggested that the Dof genes may regulate plant growth and development or other physiological processes mainly by responding to plant hormone signals, especially abscisic acid and methyl jasmonic signals. We predicted that cis-acting elements related to abiotic stress, including DRE, LTR, WUN-motif, GC-motif, MBSI, TCrich repeats, and MBS, and elements involved in defense and stress-responsiveness and drought inducibility (TC-rich repeats and MBS) were relatively abundant. Among the development-related elements identified, the number of elements involved in endosperm expression, meristem expression, and meristem expression were dominant. In addition, it was also found that there are few light-responsive elements in the Dof gene family of M. polymorpha. motif, MBSI, TC-rich repeats, and MBS, and elements involved in defense and stress-responsiveness and drought inducibility (TC-rich repeats and MBS) were relatively abundant. Among the development-related elements identified, the number of elements involved in endosperm expression, meristem expression, and meristem expression were dominant. In addition, it was also found that there are few light-responsive elements in the Dof gene family of M. polymorpha.

MpDofs Protein-Protein Interaction Network Analysis
The STRING online tool was employed to predict the functional relationships of MpDof proteins by analyzing protein-protein interactions among the Arabidopsis homologs of MpDofs. The amino acid sequence of DOF5.5 was highly similar to those of MpDof24 and MpDof25; the sequence of CDF2 was highly similar to those of MpDof05,

MpDofs Protein-Protein Interaction Network Analysis
The STRING online tool was employed to predict the functional relationships of MpDof proteins by analyzing protein-protein interactions among the Arabidopsis homologs of MpDofs. The amino acid sequence of DOF5.5 was highly similar to those of MpDof24 and MpDof25; the sequence of CDF2 was highly similar to those of MpDof05, MpDof14, MpDof17, MpDof27, MpDof35; the sequence of CDF3 was highly similar to those of MpDof11, MpDof26, MpDof36; and the sequence of AT5G02460 was highly similar to those of MpDof10 and MpDof16 (Table S4). The DOF5.5, CDF2, CDF3, and AT5G02460 all had complex interactions with FKF1 ( Figure 8). In addition, there were interactions between MpDof24 and MpDof25 and MpDof11, MpDof26 and MpDof36. Similarly, OBP1 and AT2G34140, which were highly similar to MpDof03 and MpDof31, respectively, were predicted to exist in interactions with TGA4. Moreover, OBP3 (MpDof08, MpDof13, MpDof15, MpDof28, MpDof34) interacted with TGA4 and FKF1.

Expression Patterns of MpDof Genes at Different Development Stages, Tissues, and Abiotic Stresses in Medicago Plants
The MpDof gene expression data for different developmental stages (seeding stage; early flowering stage; late flowering stage) were visualized and analyzed. As shown in Figure 9A and Table S5, there were divergences among 36 MpDof genes for expression levels in different developmental stages. MpDof26 showed the maximum expression level at the early flowering stage, while the minimum expression level was shown at the seeding stage. MpDof01 had a comparatively high expression level in all three growth stages. Some MpDof genes, such as MpDof08, MpDof17, and MpDof25, were not expressed in the three growth stages. We predicted the expression pattern of MpDof genes in different developmental tissues (leaf bud, large pod, leaf, medium pod, flower, petiole, root, small pod, stem). The expression patterns suggested that MpDof genes have distinct transcript levels in different tissues. The 36 MpDof genes were classified into 4 subgroups based on their expression patterns. As shown in Figure 9B and Table S6, subgroup A contained seven MpDof genes that were expressed in any tissue. In subgroup B, MpDof29 was specifically expressed in large pods and MpDof31 showed specific expression in the leaf, while the remaining five MpDof genes were not expressed in all nine tissues. Subgroup C contained 11 MpDof genes, which were mainly expressed in the flower, root, and stem, but the expression level was relatively low. Subgroup D showed that MpDof27 and MpDof28 were expressed in the leaf and root, and that MpDof32 was highly expressed in the flower and large pod.
We assessed the expression patterns of the M. sativa homologous Dof genes in leaves and roots under salt and drought stress. Under drought stress ( Figure 9C and Table S7),

Expression Patterns of MpDof Genes at Different Development Stages, Tissues, and Abiotic Stresses in Medicago Plants
The MpDof gene expression data for different developmental stages (seeding stage; early flowering stage; late flowering stage) were visualized and analyzed. As shown in Figure 9A and Table S5, there were divergences among 36 MpDof genes for expression levels in different developmental stages. MpDof26 showed the maximum expression level at the early flowering stage, while the minimum expression level was shown at the seeding stage. MpDof01 had a comparatively high expression level in all three growth stages. Some MpDof genes, such as MpDof08, MpDof17, and MpDof25, were not expressed in the three growth stages. We predicted the expression pattern of MpDof genes in different developmental tissues (leaf bud, large pod, leaf, medium pod, flower, petiole, root, small pod, stem). The expression patterns suggested that MpDof genes have distinct transcript levels in different tissues. The 36 MpDof genes were classified into 4 subgroups based on their expression patterns. As shown in Figure 9B and Table S6, subgroup A contained seven MpDof genes that were expressed in any tissue. In subgroup B, MpDof29 was specifically expressed in large pods and MpDof31 showed specific expression in the leaf, while the remaining five MpDof genes were not expressed in all nine tissues. Subgroup C contained 11 MpDof genes, which were mainly expressed in the flower, root, and stem, but the expression level was relatively low. Subgroup D showed that MpDof27 and MpDof28 were expressed in the leaf and root, and that MpDof32 was highly expressed in the flower and large pod.
We assessed the expression patterns of the M. sativa homologous Dof genes in leaves and roots under salt and drought stress. Under drought stress ( Figure 9C and Table S7), MpDof01 was highly expressed in both leaf and root tissues. It is worth noting that most MpDof genes had significantly low expression levels in leaves after drought treatment. However, some MpDofs were significantly up-regulated after drought stress in roots, such as MpDof02, MpDof03, MpDof15, MpDof20, and MpDof27. MpDof11, MpDof26, and MpDof27 were obviously down-regulated in both tissues under salt stress ( Figure 9D and Table S7). In addition, the expression of MpDof02, MpDof03, MpDof09, MpDof10, and MpDof34 in the root was clearly higher than in the leaf. However, the peak expression of these genes occurred at different time points after salt treatment: the expression peaks of MpDof02 and MpDof09 appeared at 27 h, while the expression peaks of MpDof03, MpDof10, and MpDof34 appeared at 3 h.
MpDof01 was highly expressed in both leaf and root tissues. It is worth noting that most MpDof genes had significantly low expression levels in leaves after drought treatment. However, some MpDofs were significantly up-regulated after drought stress in roots, such as MpDof02, MpDof03, MpDof15, MpDof20, and MpDof27. MpDof11, MpDof26, and MpDof27 were obviously down-regulated in both tissues under salt stress ( Figure 9D and Table S7). In addition, the expression of MpDof02, MpDof03, MpDof09, MpDof10, and MpDof34 in the root was clearly higher than in the leaf. However, the peak expression of these genes occurred at different time points after salt treatment: the expression peaks of MpDof02 and MpDof09 appeared at 27 h, while the expression peaks of MpDof03, MpDof10, and MpDof34 appeared at 3 h.

Validation of MpDof Genes Expression Patterns under Abiotic Stress
All eight MpDof genes had different responses to drought, salt, and low-temperature stress. The changes in five genes (MpDof10, MpDof22, MpDof23, MpDof26, and MpDof29) in drought response were essentially consistent, they were obviously up-regulated at 12 h, but their expression levels were not significantly different from those of the control (0 h) at 24 h. In addition, the expression of MpDof03 gradually increased under drought conditions  Figure 10). Compared with the control, the expression level of MpDof03 was up-regulated by 3 times at 24 h. There was no significant difference in the expression of MpDof11 at other drought treatment time points compared to the control. MpDof25 was highly induced at 3 h, and the relative expression level was about 140 times that of the control. During salt treatment, the expression levels of eight MpDof genes at 12 h and 24 h were significantly higher than those at 0 h and 3 h. MpDof11 was significantly down-regulated after the low-temperature treatment. MpDof11 was down-regulated about 5 times compared with the control after 24 h of low-temperature treatment. MpDof03, MpDof22, and MpDof23 were not induced by cold stress. The expression peaks of MpDof10 and MpDof25 both appeared at 3 h, which were about 2.5 times and 4 times that of the control, respectively. Furthermore, the peaks of MpDof26 and MpDof29 occurred at 12 h, and then gradually decreased, which were about 36 times and 18 times that of the control, respectively. in drought response were essentially consistent, they were obviously up-regulated at h, but their expression levels were not significantly different from those of the control h) at 24 h. In addition, the expression of MpDof03 gradually increased under drought co ditions ( Figure 10). Compared with the control, the expression level of MpDof03 was u regulated by 3 times at 24 h. There was no significant difference in the expression MpDof11 at other drought treatment time points compared to the control. MpDof25 w highly induced at 3 h, and the relative expression level was about 140 times that of t control. During salt treatment, the expression levels of eight MpDof genes at 12 h and 24 were significantly higher than those at 0 h and 3 h. MpDof11 was significantly down-re ulated after the low-temperature treatment. MpDof11 was down-regulated about 5 tim compared with the control after 24 h of low-temperature treatment. MpDof03, MpDof2 and MpDof23 were not induced by cold stress. The expression peaks of MpDof10 an MpDof25 both appeared at 3 h, which were about 2.5 times and 4 times that of the contr respectively. Furthermore, the peaks of MpDof26 and MpDof29 occurred at 12 h, and th gradually decreased, which were about 36 times and 18 times that of the control, respe tively. Figure 10. The relative expression levels of MpDof genes in leaves under drought, salt, and lowtemperature stress. Relative expression data were calculated using Mt-ubiquitin as the reference. Different lowercase letters represent significant differences between different treatment time points (p < 0.05).

Discussion
The Dof TFs have been shown to play transcriptional regulatory roles in numerous important physiological and biochemical processes in plants, such as nutrient transport, carbon and nitrogen metabolism, and plant hormones [36][37][38]. As a multifunctional vegetable, M. polymorpha has an important application value [29]. However, due to its limited distribution, some adverse environmental conditions in many areas are not conducive to its normal growth and development. To provide new gene resources for the cultivation of new resistant varieties of M. polymorpha, we need to conduct a comprehensive investigation of the Dof gene family. In this study, we discussed the possible evolutionary patterns of the Dof family of M. polymorpha and determined their potential functions under abiotic stress. Thirty-six Dof genes were identified in M. polymorpha. Forty-two, thirty-six, and thirty Dof genes were identified in M. truncatula [39], Arabidopsis [40], and O. sativa [33], respectively. However, the genome sizes of M. polymorpha, M. truncatula, Arabidopsis, and rice are about 457 Mb, 500 Mb, 125 Mb, and 466 Mb, respectively, which means that the number of Dof TF family members does not increase with the size of the genome. The above results also suggested that Dof genes have shown greater diversity in long-term evolution.
The number of Dof genes of M. polymorpha and many angiosperms such as M. truncatula, M. sativa, Arabidopsis, and rice are close to each other [39,41,42]. However, previous studies found that a single Dof gene was found in Chlamydomonas Rheinis, 19 Dof genes were found in moss Physcomitrella patens, and Dof genes were not identified in the red algae Cyanidioschyzon merolae and the diatom Thalassiosira pseudonana [43], suggesting that Dof genes may originate from the earlier ancestors before the differentiation between Chlamydomonas and terrestrial plants. In addition, 12 and 10 Dof genes were identified in fern Selaginella moellendorffiii and gymnosperms Pinus taeda, respectively [43]. Therefore, we speculate that ferns and gymnosperms have Dof gene loss events after the differentiation of moss lineage and vascular plants, while large-scale gene duplication events of Dof genes occur during the evolution of angiosperms, which may be closely related to the development of complex regulatory mechanism networks of angiosperms.
The phylogenetic relationship analysis is conducive to our preliminary understanding of the evolution of gene families and gene functions [44]. Similar to previous studies [40,41,45], the 36 Dof genes in M. polymorpha were mainly divided into 10 groups. MpDof gene numbers of the orthologous groups are different among four species, indicating that the Dof gene family may experience various evolutionary pathways in distinct species. Group D1 contained the largest number of Dof genes, with 35, while groups C2.2 and D2 contained only 8 and 4 Dof genes, respectively, suggesting that these orthologous groups had different degrees of gene amplification and loss. Phylogenetic analysis showed that MpDof genes had the closest genetic relationship with MtDof genes, but were further related to Arabidopsis and O. sativa. These MpDof genes may be produced by further genomic expansion after the divergence of different plant species, thus legume plants are more closely related. Group E only included 12 OsDof genes, indicating that group E evolved separately during the differentiation of monocotyledons and dicotyledons. The above results also support the theory that genes continue to experience random loss or obtain events in the process of evolution [31,46]. The Dof members in the same group may have similar functionality [4]. Arabidopsis Dof members (AtDof9, AtDof15, AtDof30, and AtDof32) included in group C1 have been shown to be involved in plant vascular tissue and seed development and are highly expressed in root tissues [9,47,48]. It was also found that MpDof09, MpDof18, and MpDof33 included in group C1 were involved in extensive tissue expression. We speculate that the function of MpDofs is similar to that of Arabidopsis CDF protein included in D1 group, which is a key factor in regulating photoperiod flowering in Arabidopsis and plays an important role in different abiotic stress responses [49][50][51][52]. Furthermore, protein network interaction predicted that the MpDof protein in group D1 may play a role in plant photoperiod flowering and abiotic stress response through interaction.
Further duplication analysis showed that segmental duplication played a major role in the expansion of Dof gene in M. polymorpha. This result is similar to the Dof family reported in M. truncatula and Tartary buckwheat [39,53]. However, some gene families (CAB family) had high-level tandem duplications, which functionally differ from these (MYB TF) with high segmental duplication levels [54]. In addition, divergence times of all the segmental duplication gene pairs ranged from 21.45 to 93.36 MYA. It can be concluded that all the segmental-duplicated gene pairs may occur after the differentiation of Arabidopsis and legumes and before the differentiation of M. polymorpha and M. truncatula. However, tandem-duplicated gene pairs occur after the formation of M. polymorpha and may play a special role in M. polymorpha. Furthermore, the MpDof genes have undergone negative selection during the evolution of M. polymorpha. The diversity of the Ka/Ks value indicates that the evolution rate of duplicated genes is different when facing the pressure of purification selection, which may lead to the functional differentiation of the Dof gene [55]. There are great differences in tissue expression between MpDof14 and MpDof25, suggesting that some duplicated genes had functional differentiation. The expression patterns of MpDof11, MpDof14, and MpDof27 in different developmental stages of M. polymorpha are similar, indicating that some genes may have functional redundancy after duplication.
The genes containing the same motif are most likely caused by gene duplication events [5]. Furthermore, the result that the Dof gene of the same subgroup has a similar exon-intron structure can further support the close evolutionary relationship among the Dof members in the evolutionary tree. The Dof gene of M. polymorpha contains very few introns (zero to one intron), which is similar to Arabidopsis, tea, and cassava Dof genes [40,45,56]. However, Chlamydomonas, Physcomitrella patens, Selaginella moellendorffiii, and Pinus taeda contain 4, 6, 5, and 4 introns, respectively [8]. In Medicago plants, the number of introns in M. truncatula and M. sativa Dofs varies from zero to four, and zero to seven, respectively [39,41]. The results revealed that intron loss occurred during the evolution of M. polymorpha. Studies have shown that the evolution of poor introns and intron-free genes may be related to vascular system and seed development, which promotes the complexity of structure and regulatory pathways during the evolution of higher plants [57]. In addition, previous studies of gene families have found that genes with no intron and intron deficiency (three or less) are more likely to play a role in abiotic stress responses such as drought and salt than intron-rich genes [58]. However, further experiments are needed to analyze the specific functions of poor intron family genes in plant growth and development and resistance to abiotic stress. The MpDof gene family can be used as a resource for a poor-intron gene family, which provides important value for exploring the origin, evolution, and function of plants.
In recent decades, numerous studies have reported that the plant Dof TFs are involved in the plant's response to various abiotic stresses. The results revealed that the overexpression of SlCDF1 or SlCDF3 enhanced Arabidopsis plant resistance to drought and salt, and stress-responsive genes such as COR15, RD29A, and RD10 were transcriptionally activated, suggesting that SlCDF may be a significant upstream regulator of the plant's response to drought and salt stress [59]. The OsDof1 overexpression line ensured the seed-setting rate under low-temperature stress, demonstrating that it participated in the biological process of enhancing cold tolerance in rice [60]. MpDof01 was expressed by drought in leaves and roots, indicating that MpDof01 may help to improve drought tolerance in plants. Furthermore, some MpDof genes may have diverse regulatory mechanisms in different tissues and abiotic stress responses. For example, the expression of MpDof03 is up-regulated in roots and down-regulated in leaves under salt stress, while MpDof03 is down-regulated in roots after salt stress. qRT-PCR results further showed that the expression levels of most MpDofs changed significantly under drought, salt, and cold stress, indicating that these genes play a key role in improving abiotic stress resistance. The validated eight MpDofs were differentially expressed under three types of abiotic stress. For example, MpDof11 had no significant change under drought treatment, was significantly up-regulated under salt stress, and was obviously down-regulated under low-temperature conditions, suggesting that the regulation mechanism of MpDof genes were different under different stresses. In Brassica napus, the expression level of BnCDF1 was not affected by drought or low-temperature stress, while the BnCDF1 transgenic Arabidopsis showed higher freezing resistance [23]. Salt treatment significantly increased the expression of all eight MpDof genes, showing that MpDofs may be important regulatory factors for M. polymorpha in response to salt stress. Arabidopsis that overexpressing MtDof32, which was divided into group D1, had stronger salt tolerance than the wild-type plants [61]. In addition, soybean GmDof41 positively regulated salt tolerance by binding with the promoter of GmDREB2A [62]. It is worth noting that MpDof25 was induced by drought, salt, and low temperature, but its expression increased by about 140 times in 3 h of drought treatment, which suggested that MpDof25 may be a drought-sensitive gene. Similarly, MpDof26 was considerably up-regulated under three stress treatments, indicating that it may be a crucial TF for plants to respond to diverse abiotic stresses and can be used as a potential target for plant stress resistance breeding. Overexpression of AT3G47500.1 (AtCDF3), which is highly homologous to MpDof26, improves the tolerance of Arabidopsis to drought, salt, and cold stress [51]. These results show that MpDof TFs are induced by multiple abiotic stresses; thus, we speculate that MpDofs may have complex regulatory functions in abiotic stresses.
To sum up, this study contributes by further improving the evolutionary path of the Dof gene family in dicotyledons and provides a basis for the study of the structural and functional diversity of Dof genes. In addition, our results found that MpDof10, MpDof25, MpDof26, and MpDof29 may be good candidates for genetic engineering breeding with regard to resisting various abiotic stresses.

Phylogenetic Analysis of Dof Family Members and Orthologous Groups Identification
To comprehensively acquaint the evolutionary relationships of the MpDof gene family, a multiple-sequence alignment of 141 Dof protein sequences from Arabidopsis, O. sativa, M. truncatula, and the putative MpDof protein sequences was conducted using ClustalW with the default parameters. Subsequently, the phylogenetic tree was constructed by the neighbor joining method with 1000 bootstrap replicates in MEGA7 software (with a Poisson model and Pairwise deletion).
In this study, diamond software was used to perform an all-versus-all BlastP search with parameters: E value 1 × 10 −3 as the input file for the OrthoFinder software [64], and orthologous groups were identified by previously published methods [65]. Then, Tajima's D values of all orthologous groups were calculated by DnaSP 5.0 to evaluate the selective forces among orthologous groups [66].

Detection of Gene Structures and Conserved Motifs of MpDofs
Both CDS sequences and genomic sequences of MpDof were used to predict the MpDof gene structure by utilizing the Gene Structure Display Server (http://gsds.gao-lab.org/, accessed on 14 November 2021). The intron and exon distribution were exhibited in the gene structure. The highly conserved Dof domain was identified based on multiple sequence alignment using DNAMAN version 7 software. The MpDof protein sequences were identified using the MEME online program v5.4.1 (https://meme-suite.org/meme/, accessed on 14 November 2021), and the maximum number of motifs was 20. Ultimately, the gene structure and motifs were visualized using TBtools software [67].

Chromosomal Localization, Gene Duplication Events, and Synteny Analysis of the MpDof Genes
The chromosomal position of the MpDof gene was obtained from the gff3 file in the Genome Warehouse in National Genomics Data Center (https://ngdc.cncb.ac.cn/search/ ?dbId=gwh&q=Medicago_polymorpha, accessed on 2 September 2021). The "MCScanX" function was used to analyze gene duplication events. Then, the chromosome location of 36 Dof genes and synteny relationships in the M. polymorpha genome were visualized with "Advanced Circos" tools. Syntenic analysis images of M. polymorpha between Arabidopsis, M. truncatula, and O. sativa were conducted by "Dual Systeny Plot" view software in TBtools.
The nonsynonymous (Ka), synonymous (Ks), and Ka/Ks ratios for 15 duplicated MpDof gene pairs were calculated by performing the Simple Ka/Ks Calculator function of TBtools software. It is generally assumed that Ka/Ks > 1, Ka/Ks = 1, and Ka/Ks < 1 indicate positive selection pressure, natural selection pressure, and negative selection pressure, respectively [68]. Then, the time of duplication (T) was estimated using the formula T = Ks/2λ, where λ = 1.5 × 10 −8 represents the rate of replacement of each locus per dicotyledon year [69].

Identification of Promoter Cis-Acting Elements
The promoter of 2000 bp upstream sequences from the initiation codon of the putative MpDof genes was extracted. Then, the cis-acting elements in these sequences were predicted in the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/ plantcare/html/, accessed on 14 October 2021) and visualized utilizing TBtools software [70].

Functional Protein Interaction Network Prediction of MpDofs
In the study, Arabidopsis was used as the association background to analyze the initial function of MpDofs. The putative MpDof protein sequences were located in the STRING server (https://cn.string-db.org/ accessed on 12 November 2021) to predict the functional interaction networks and functional annotations. The minimum required interaction score was 0.70, and the first and second shells were set to no more than 10 interactors. The network edges were set as line colors to represent the protein-protein interaction types.

Expression Analysis of Dof Genes among Different Tissues at Different Development Stages
The expression pattern of each MpDof gene in the seedling stage, early flowering stage, and late flowering stage were obtained from the papers published by our research team [31].
As a model plant in legumes, M. truncatula has a high sequence similarity with M. polymorpha. The gene expression data of M. truncatula in different organs including leaf bud, large pod, leaf, medium pod, flower, petiole, root, small pod, and stem were downloaded from LegumeIP v3 Browser (https://www.zhaolab.org/LegumeIP/gdp/13/gene/profile/ 5?sessionid=Plant_organs, accessed on 5 December 2021) to predict the organ-specific expression patterns of MpDof genes at diverse development stages [71]. The expression profile was shown in the heatmap function in TBtools.

Expression Pattern Analysis of MpDof Genes under Abiotic Stress
Because Medicago sativa is a perennial Medicago plant, Alfalfa Gene Editing Database (http://alfalfagedb.liu-lab.com/heatmap/heatmap/, accessed on 12 February 2023) was used to retrieve M. sativa Dof genes homologous to MpDof genes and to evaluate the expression profiles of these MsDof genes under abiotic stress. The heatmap drawing tool is the same as above.
To further explore the expression level of MpDof genes under different abiotic stress treatments, eight MpDof genes were randomly selected according to the orthologous genes responding to abiotic stress for qPCR verification. M. polymorpha seedlings were collected after drought (PEG 6000, 20%), salt (220 mM), and low-temperature (4 • C) treatments for 0, 3, 12, and 24 h, respectively, and three replicates of each sample were applied to perform qRT-PCR. The total RNA was extracted using a Vazyme Total RNA Isolation Kit (Vazyme, Nanjing, China). The total RNA was synthesized into the first strand of cDNA using the cDNA synthesis kit (Vazyme, Nanjing, China). Eight pairs of gene-specific primers for qRT-PCR analysis were designed by PerlPrimer v1.1.21 software and displayed in Table  S8. A 10 µL reaction volume for each sample contained 5 µL of 2 × AceQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China), 0.2 µL of each primer, 1 µL of diluted cDNA product, and 3.6 µL of ddH 2 O. The qRT-PCR reaction procedures recommended in the qPCR kit, which were carried out on the platform supported by the QuantStudio 3 system (Thermo Fisher Scientific, Waltham, MA USA), were as follows: 5 min at 95 • C for DNA polymerase activation, denaturation, and anneal/extension at 95 • C for 10 s and 60 • C for 30 s, respectively, for a total of 40 cycles. Mt-ubiquitin was selected as the internal control to calculate the relative expression data according to the 2 −∆∆CT method.

Conclusions
In this study, we identified 36 Dof genes in M. polymorpha, and the results showed that these Dof genes all have highly conserved single zinc finger domains. At the same time, phylogenetic relationships, conserved motifs, the exon-intron structure, synteny, gene duplication events, cis-acting elements, protein interaction, and expression profiles were comprehensively analyzed. The protein-protein interaction network analysis indicated that multiple MpDof proteins may participate in regulating photoperiod flowering. The expression profiling analysis showed that MpDof genes possessed different expression patterns in distinct tissues and developmental stages. Combined with cis-acting element analysis, the possible regulatory expressions of MpDof genes were further detected. In addition, the expression of most MpDofs presents different dynamic reaction processes under drought, salt, and low-temperature conditions. The selected eight MpDof genes were significantly up-regulated under salt treatment, which proved that the MpDof genes played an important role for M. polymorpha in responding to salt stress. Our results would provide references for the further study of various functions and regulatory networks of Dof genes in M. polymorpha.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/plants12091831/s1, Figure S1: Twenty motif sequence logos of MpDof proteins; Table S1: The Dofs from four species; Table S2: The position of genes in chromosomes; Table S3: Functions of predicted cis-acting elements; Table S4: Annotation information of proteins in Arabidopsis thaliana highly matches the query sequence;