Genome-wide identification and comprehensive analysis of the NAC transcription factor family in Sesamum indicum

The NAM, ATAF1/2, and CUC2 (NAC) family constitutes a large family of plant-specific transcription factors, involved in many aspects of physiological processes and a variety of abiotic stresses. There is little information concerning the NAC family in Sesamum indicum. In this study, 87 sesame NAC genes were identified and phylogenetically clustered into 12 groups with Arabidopsis NAC genes. A total of 83 SiNAC genes were distributed non-randomly on the 16 linkage groups in sesame. Four and 49 SiNACs were found to be tandemly and segmentally duplicated, respectively. Expression profiles of SiNAC genes in different tissues (root, stem, leaf, flower, seed, and capsule) and in response to drought and waterlogging stresses by using RNA-seq data demonstrated that 23 genes were highly expressed in all tissues, 18 and 31 SiNACs respond strongly to drought and waterlogging stresses, respectively. In addition, the expression of 30 SiNAC genes distributed in different subgroups was analyzed with quantitative real-time RT-PCR under cold, osmotic, and salt stresses, revealed that their expression patterns vary in response to abiotic stresses. SiNAC genes displayed diverse expression patterns among the different tissues and stress treatments, suggested that their contribution to plant growth and development in sesame and multiple stress resistance in sesame. In this study, NAC transcription factors were analyzed in sesame and some specific candidate SiNAC genes in response to abiotic stress for functional study were identified. This study provides valuable information to deepen our understanding of the abiotic stress responses by NAC transcription factors in sesame.


Introduction
Sesame (Sesamum indicum), one of the most important oilseed crops worldwide, has long been considered as a representative health food in China, Korea, Japan, and other East Asian countries [1]. The yield and quality production of sesame is seriously compromised by a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 multiple detrimental environmental factors, such as waterlogging, drought, salinity, pests, and diseases. Transcription factors (TFs) are types of regulatory proteins that control plant growth and development, as well as their responses to environmental changes, by regulating the expression of downstream genes [2][3][4]. For example, it has been demonstrated that some MYB TFs play significant roles in different biological processes, such as hormone signal transduction, organ development, metabolism, and responses to biotic and abiotic stresses [5][6][7]. Expression profiles analysis in different species have shown that many WRKY genes are involved in responses to various abiotic stresses, like drought, waterlogging, wounding, and salt stress [8][9][10][11]. The identification and functional characterization of TFs is a major area of transcription machinery research. The release of a high-quality genome sequence of sesame offers an excellent research opportunity to genome-wide analysis of sesame gene family [12].
The NAC (NAM, ATAF, and CUC) domain protein family is one of the largest plant-specific TF families. Its name was originally derived from the three proteins that contain a similar DNA-binding domain: no apical meristem (NAM), ATAF1-2, and cup-shaped cotyledon (CUC) [4,13]. Typically, NAC proteins possess a conserved NAM domain at the N-terminus and a divergent transcription regulatory region at the C-terminus [14]. The N-terminal NAC domain usually comprises nearly 160 amino acid (aa) residues that are divided into five subdomains: A-E [14]. Subdomains A, C, and D are commonly highly conserved, whereas subdomains B and E are highly divergent and might confer functional diversity to NAC TFs [3,14]. Subdomain A may play an important role in the formation of functional dimers with other NAC domain proteins, and subdomains C and D bind to DNA. In contrast, the C-terminal transcription regulatory regions are highly divergent and operate as functional domains by conferring regulation diversity of transcriptional activation activity [3,14]. Moreover, some NAC TFs contain transmembrane motifs at the C-terminal end, which function in plasma membrane or endoplasmic anchoring [15].
NAC was firstly identified in Arabidopsis thaliana as a key gene for pattern formation in embryo and flowers in Petunia [13] and organ separation [4]. Later, it was found that NAC proteins regulated a variety of plant developmental processes, such as the formation of adventitious shoots [16], development of the shoot apical meristem [4,13] [19,21,23], and grain nutrient remobilization [24]. Although the exact regulatory mechanism of NAC genes is not much known, their activity can be regulated through different processes: binding of specific TFs to NAC regulatory region in the promoter at transcriptional level; miRNA164-mediated cleavage of NAC genes at post-transcriptional level; NAC protein degradation mediated by ubiquitins, dimerization and interaction with other non-NAC proteins at post-translational level. In recent years, increasing evidences have shown that a large number of NAC genes play important roles in the regulation of plant tolerance to abiotic stresses [3]. In Arabidopsis, AtNAC2, AtNAC3, ANAC019, and RD26 regulate the expression of stress-responsive genes involved in multi-abiotic stress tolerance [25][26][27][28]. Transgenic Arabidopsis overexpressing ATAF1, MlNAC9, or CarNAC4 and transgenic rice overexpressing SNAC1 or OsNAC2 exhibits enhanced resistance to salt, drought and other stresses, respectively [29][30][31][32][33]. It was reported that some NAC membrane-bound TFs (NTLs) are mostly regulated in protecting plants against abiotic stresses [18,21,34]. For example, Arabidopsis NTL4 is a membrane-bound NAC transcription factor which promotes reactive oxygen species production to during drought stress [11]. In Soybean, the heteroexpression of active membrane-bound GmNTL1/GmNLT11 proteins was found to cause improved tolerance to abiotic stresses [35]. All these studies indicate that specific NAC TFs are very important in plants' physiological signaling and multi-abiotic stress tolerance as well as plant development.
Recently, 117 NAC genes in Arabidopsis [36], 74 in grape (Vitis vinifera) [37], 163 in Populus trichocarpa [38], 152 in Nicotiana tabacum [39], 151 in O. sativa [36], 101 in Brachypodium distachyon [40], and 152 in Zea mays [41]were identified by genome-wide analyses, respectively. Sesame is one of the most important oil crop plants and affected seriously by multiple biotic and abiotic stresses [42]. Although a lot of NAC TFs have been functionally described in Arabidopsis, O. sativa, and other plants, only one sesame NAC gene (SiNST1/SiNAC58, SIN_1005755) was found significantly associated with content of lignin and seed coat thickness [43]. The functions of most of NAC members in sesame were unknown. In this study, we performed a comprehensive investigation of the NAC gene family in sesame through their identification, molecular characterization, phylogenetic analysis, and expression profiling in different tissues (root, stem, leaf, flower, seed, and capsule) and various abiotic stresses (drought, waterlogging, cold, osmotic, and salt). The results may lay the foundation for future functional characterizations of NAC TFs in sesame.

Identification of NAC members in sesame
To identify the NAC family members in sesame, both Hidden Markov Model (HMM) and BLASTP searches were performed in the sesame genome with Arabidopsis and O. sativa NAC sequences as queries (S1 Table). A total of 87 putative NAC TFs with a conserved NAM domain were found in the complete sesame genome (S2 Table). Owing to the lack of a designated standard annotation for the 87 NAC genes in sesame, we named them SiNAC01-SiNAC87 based on their position from top to bottom in the sesame linkage groups (LGs). The NAC genes identified in sesame encoded proteins ranging from 130 (SiNAC56) to 631 (SiNAC85) aa residues in length, with an average of 345 aa (S2 Table). Detailed information of sesame NAC genes, including accession numbers, precise positions, and similarities to their Arabidopsis orthologs are listed in S2 Table. Unequal distribution and gene duplication of SiNAC genes in the sesame genome A total of 83 members of the SiNAC gene family were distributed non-randomly on the 16 LGs in sesame, and four SiNAC genes (SiNAC84-SiNAC87) were mapped to unanchored scaffolds (Fig 1, S2 Table). LG 3 contained the largest number (10,~11.5%) of sesame NAC genes, followed by LG 4, LG 8, and LG 12, each with eight members (~9.2%). In contrast, LG 7 and LG 16 contained only one gene each.
Four and 49 SiNACs were found to be tandemly and segmentally duplicated, respectively which might have contributed significantly to the expansion of the SiNAC family (S2 Table). The 49 SiNACs were located on duplicated segments on LGs 1-15 as illustrated in S1 Fig. The duplicated segments on LG 8 contains the highest number of SiNACs (8), followed by six genes on the duplicated segments located on LG 12. In contrast, only one SiNAC each was found on the duplicated segments located on LGs 1, 7, and 13 (S1 Fig). The syntenic analysis of sesame, Arabidopsis and V. vinifera subgenomes showed 13 sesame NAC orthologous gene pairs between sesame and Arabidopsis, and 34 pairs between sesame and V. vinifera (S2 Fig, S2 and S3 Tables).

Conserved motifs and gene structure of SiNACs
Multiple sequence alignment of sesame NAC proteins revealed that all SiNAC proteins included a highly conserved N-terminal NAC region (S3 Fig). Most of the SiNACs had a complete NAC domain consisting of five subdomains: A, B, C, D and E. However, SiNAC32, SiNAC47, and SiNAC56 lacked A and B subdomains, and SiNAC49 lacked D and E subdomains. To study the evolutionary relationship of these SiNACs, a neighbor-joining (NJ) tree was created from alignments of 87 NAC protein full-length sequences. The SiNAC TFs were divided into 10 groups, designated as subgroups a-j (Fig 2A). The subgroup h was the largest one, containing 17 members, whereas the subgroup d had only three members. In addition, most of the duplicated NAC members were clustered in the same subgroup. For instance, two duplicated NAC genes (SiNAC08 and SiNAC36) were assigned to the subgroup c, and the other three (SiNAC30, SiNAC54, and SiNAC60) were classified into the subgroup g with high bootstrap values. The MEME program was used to predict putative motifs to get a better understanding of the diversity of NAC TFs from sesame. In total, 20 distinct motifs were identified ( Fig 2B and S4 Fig). It is obvious that most of the closely related members in the phylogenetic tree exhibited common motif compositions, suggesting that the NAC members clustered in the same subgroup may have similar biological functions. The motif distribution analyses of the NAC proteins revealed that the N-termini of most SiNAC TFs contained five highly conserved subdomains (A-E), which conferred DNA-binding activity [14], except for most members of subfamily j which had no subdomains A and B (Fig 2B). In some specific subgroups, some conserved motifs were identified in C-terminal regions, such as motifs 10 and 18 in subgroup h and motif 15 in subgroup e (Fig 2B), suggesting that the specific functions of different subgroups could be due to specific motifs.
The exon/intron organization in the SiNACs coding sequences were compared using online Gene Structure Display Server 2.0 (http://gsds.cbi.pku.edu.cn/). The result revealed that sesame NAC genes contain 0 to 6 introns, and the majority of SiNAC genes contained three exons. The detailed gene structure is presented in Fig 2C and S4 Table. Among these SiNACs,  those in most subfamilies harbored 2-5 introns, with the exception of SiNAC16 and SiNAC35 which had no intron. Overall, SiNACs in the same phylogenetic group shared highly similar exon/intron structure including intron number, intron phase, and exon length.

Phylogenetic analysis of SiNACs
To explore the evolutionary relationships of NAC proteins from sesame and Arabidopsis, a NJ phylogenetic tree was created using full protein sequences of 203 NACs from sesame and Arabidopsis. The 87 SiNAC proteins formed 12 clades together with NACs from Arabidopsis, and were designated as subgroups I-XII (Fig 3). The largest clade was subgroup VII, which contained 16 SiNAC members; whereas the smallest was subgroup XII, with only one member. Subgroup XI contained only members from Arabidopsis, which implies that the homologs of these genes may have been lost in sesame following the divergence during the evolutionary process. Remarkably, the Arabidopsis NAC genes with the same functions exhibited a tendency to cluster into the same subgroup. For example, the VNDs (VND1-VND7) and the well-characterized NAC TFs involved in shoot organ boundary delimitation (CUC1, CUC2, and CUC3) were mainly located in subgroup I and II, respectively. Subgroup VIII encompassed many well-known stress-responsive Arabidopsis NAC genes, including ANAC019, AtNAC2, AtNAC3, ATAF1, ATAF2, and RD26. Therefore, these three subgroups were named according to their orthologous groups in Arabidopsis-VND (secondary wall synthesis NAC, subgroup I), CUC (development-related NAC, subgroup II), and SNAC (stress-related NAC, subgroup VIII) (Fig 3).

Membrane-bound SiNAC subfamily
Recently, NAC membrane-bound TFs (MTFs) have been identified in many plants and implicated in response to abiotic stress [18,21,34]. In the present study, six SiNAC proteins containing α-helical transmembrane motifs at C-terminal ends were identified using the SMART web server, and named SiNTL1-SiNTL6 (Table 1). A NJ phylogenetic tree of NAC MTFs from sesame, Arabidopsis, and O. sativa was performed (Fig 4). The phylogenetic tree revealed that sesame MTFs had a close relationship with Arabidopsis MTFs suggesting similar functions of these genes in both species (Fig 4).

Expression profiles of SiNAC genes in response to drought
To determine the expression patterns of SiNAC genes in response to drought, transcriptome data of drought-tolerant (DT) cv. ZZM0635 and drought-sensitive (DS) cv. ZZM4782 under drought stress were analyzed. When the soil water content at 35, 15, 9, and 6%-d0 (control), d1, d2, and d3 treatments, respectively-the root tissues of DT and DS plants were harvested for RNA-seq analysis [45]. A total of 79 SiNACs common to DT and DS plants were expressed differentially after drought treatment (Fig 6). Of these, nine (SiNAC07, SiNAC10, SiNAC39, SiNAC41, SiNAC56, SiNAC61, SiNAC62, SiNAC75, and SiNAC80) were significantly up-regulated [log 2 fold-change (FC) > 1] and nine (SiNAC03, SiNAC05, SiNAC09, SiNAC15, SiNAC40, SiNAC49, SiNAC51, SiNAC54, and SiNAC55) were down-regulated (log 2 FC < -1) in roots of both DT and DS plants at all the three time points during drought treatment (Fig 6). Many SiNACs (especially some duplicated SiNAC genes) from the same phylogenetic subgroup showed similar expression profiles under drought stress between the DT and DS cultivars, suggesting that those SiNACs may have similar functions in the process of sesame response to drought (Fig 6). In particular, 8 SiNACs (SiNAC07, SiNAC10, SiNAC24, SiNAC56, SiNAC61, SiNAC62, SiNAC75, and SiNAC80) from the subgroup SNAC were identified to be significantly up-regulated in the roots of both DT and DS plants at least one time point during drought treatment. However, some genes showed different expression profiles between DS and DT cultivars. Expressions of SiNAC21, SiNAC23, SiNAC32, SiNAC38, SiNAC46, and SiNAC76 were down-regulated in roots of DS plants, but increased or were unchanged in DT plants after drought treatment. These NAC genes responding to drought treatment may play important roles in regulation of sesame drought tolerance.

Expression profiles of SiNAC genes under waterlogging stress
To determine the expression profiles of SiNACs under waterlogging stress, the 87 NAC genes were analyzed using transcriptome data obtained from root samples of waterlogging-tolerant (WT) cv. Zhongzhi No. 13 and waterlogging-sensitive (WS) cv. ZZM0563 when waterlogged for 0, 3, 9, and 15 h [42]. Similar to the expression profiles in response to drought stress, most of the SiNACs from the same phylogenetic subgroup also showed similar expression profiles in response to waterlogging stress between WT and WS genotypes (Fig 7). Among them, 11 SiNACs were up-regulated (log 2 FC > 1) and 20 SiNACs were down-regulated (log 2 FC < -1) in roots of both DT and DS plants for at least two time points during waterlogging treatment. Obviously, 3 SiNAC genes (SiNAC12, SiNAC41, and SiNAC77) of the subgroup II and 6 SiNACs (SiNAC07, SiNAC10, SiNAC24, SiNAC61, SiNAC62, and SiNAC75) of the subgroup SNAC were identified to be significantly up-regulated in the roots under drought and waterlogging treatments (Fig 7). However, the expression patterns of some genes between WS and WT plants differed. Notably, SiNAC04, SiNAC06, SiNAC35, SiNAC53, and SiNAC83 were upregulated in WT roots, but did not significantly change in WS plants after waterlogging treatment (Fig 7). In addition, SiNAC10, SiNAC16, SiNAC17, SiNAC44, SiNAC50, SiNAC66, SiNAC72, SiNAC75, and SiNAC81 were up-regulated in WS roots, but detected down-regulated or unaffected in WT roots after waterlogging treatment (Fig 7). The functions of those waterlogging-responsive SiNACs may enhance resistance of sesame in response to waterlogging stress.

Expression profiles of SiNAC genes upon exposure to various abiotic stresses
To extend our understanding of SiNAC genes in response to various environmental stresses, 30 SiNACs (including all SiNTL genes, all members of subgroup SNAC, and one or two members of other subgroups) distributed in different subgroups were selected for further study of their responses to cold, osmotic, and salinity treatments at transcriptional level using quantitative real-time PCR (qRT-PCR). Heatmaps representation of expression FC in responses to the three stresses are shown in Fig 8 (original data were shown in S5 Fig). Under cold treatment, the majority of detected SiNACs were down-regulated at two or three time points. Some SiNACs were also induced by low temperature. Among them, SiNAC13, SiNAC47, SiNAC66, and SiNAC79 were highly up-regulated with low-temperature condition (Fig 8). Most of the detected SiNACs were up-regulated at all the three time points after osmotic treatment (Fig 8).
Among them, SiNAC75, SiNAC10, SiNAC62, SiNAC80, and SiNAC63 were highly up-regulated at 12 h after osmotic treatment (Fig 8). Notably, almost all of subgroup SNAC members were greatly sensitive to osmotic treatment, indicated that their functions in enhancing tolerance to abiotic stresses. After salinity treatment, the expression of most detected SiNACs was up-regulated at two or three time points (Fig 8). Among them, expression of 15 genes was significantly up-regulated from 6 to 12 h suggesting that they may participate in sesame salt-stress tolerance. In contrast, SiNAC06, SiNAC13, SiNAC47, and SiNAC55, were down-regulated at least at two time points after salinity treatment (Fig 8).
The heat maps showed that all analyzed SiNACs displayed variations of expression levels in responses to one or more stresses at the different time points after treatments. In addition, the majority of the detected SiNACs were up-regulated under osmotic and salt conditions, but down-regulated under cold treatment (Fig 8). Notably, SiNAC07, SiNAC21, and SiNAC79 were  up-regulated after cold, osmotic, and salt treatments (Fig 8). Overall, those NAC genes responding to one or more treatments may play a significant role in the regulation of sesame tolerance to different abiotic stresses.

Discussion
Sesame is an important oilseed crop and its growth and productivity are severely affected by abiotic stresses, particularly drought and waterlogging. The NAC gene family is one of the largest families of TFs and plays important roles in plant growth, development, and response to abiotic and biotic stresses. NAC TF genes have been identified and characterized in many species [14, 36,37], but there is little information about this family in sesame. In this study, 87 NAC genes were identified in the sesame genome. The sesame NAC gene family (350 Mb, 87 NAC genes) is small compared with those of Arabidopsis (125 Mb, 117 NAC genes) and O. sativa (466 Mb, 151 NAC genes) [12, 36,46,47]. However, it is large compared with its closest related woody perennial V. vinifera (490 Mb, 74 NAC genes) [37,48]. We speculate that more NAC genes were identified in the sesame genome may reveal that more need for these genes of sesame in transcriptional regulations. Gene duplication is one of the main driving forces in the genetic systems and evolution of genomic [49], and previous studies revealed that segmental duplication was the main method of NAC family expansion in several species [38]. In this study, 56.3% (49/87) of SiNACs were found to be segmentally duplicated (S2 Table and  Molecular characterizations of SiNAC TFs revealed that motif compositions and gene structures differed among the subgroups; however, the NAC proteins that clustered in the same subgroup showed a similar motif composition and gene structure, which was in agreement with reports in Arabidopsis, O. sativa, and V. vinifera [14,36,37] (Fig 2). The diversity of conserved motifs and gene structures may also indicate that SiNAC TFs have diverse biological functions. Structurally similar NAC proteins within species were found to be functionally orthologous [36,50]. When compared with the well-studied species Arabidopsis, an evolutionary relationship analysis of the NAC subfamily in sesame was performed. In this study, some SiNAC TFs which clustered in the same subgroup, especially shared similar gene expression profiles in different tissues of sesame or in response to various abiotic stresses, are likely to possess similar biological functions. For example, the Arabidopsis NAC proteins (NSTs and VNDs) of subgroup VND are involved in secondary wall synthesis [51]. Four SiNAC genes (SiNAC08, SiNAC36, SiNAC58, and SiNAC82) from the same subgroup VND were found highly expressed in the capsules, indicating that these genes may regulate secondary wall synthesis in sesame capsule. Furthermore, SiNST1/SiNAC58 was found significantly associated with lignin content and seed coat thickness in sesame [43], which is consistent with the function of its Arabidopsis homologous gene NST1 [50]. Subgroup CUC encompassed the NAC proteins involved in shoot organ boundary delimitation [16, 52] and subgroup SNAC includes NAC genes such as AtNAC2, AtNAC3, ATAF1, RD26, and ANAC019, which have been shown to be induced by abiotic stresses including drought and salinity [23,[25][26][27][28]. Consequently, phylogeny-based functional prediction helpful for SiNACs' functional characterization and should be a focus in research on functions of sesame NAC genes in the future.
In this study, tissue specific and stress responsive SiNACs were identified with available transcriptome sequencing data and qRT-PCR. The different expression patterns of SiNACs provide vital information for determining the functions of SiNAC TFs. The transcriptome sequencing data from different tissues of cv. Zhongzhi No. 13 showed that 23 SiNACs were expressed in all organs, and several were expressed specifically in different tissues, suggesting that these sesame NAC genes may have specific functional roles in sesame organ growth and development. Notably, SiNAC58 was highly expressed in seeds and capsules, consistent with its function of increasing content of lignin and seed coat thickness [43]. Evidence suggests that NAC genes play significant roles in the regulation of responses to various abiotic stresses in plants [23,[25][26][27]32]. In this study, we identified many sesame NAC genes induced by one or more abiotic stresses. Almost all members of the subgroup SNAC were responsive to various abiotic stresses. Especially, three SiNACs (SiNAC07, SiNAC21, and SiNAC79) were up-regulated under drought, waterlogging, cold, osmotic, and salt treatments, implying their roles in multiple abiotic stresses resistance. SiNAC61 and SiNAC62 were highly expressed after drought and salt treatments and may function to improve the endurance of sesame to drought and salt stresses. In addition, we found some members of other subgroups to be highly active in abiotic stress response. For example, expressions of SiNAC04 (subgroup VII), SiNAC06 (subgroup I), SiNAC53 (subgroup VII), and SiNAC83 (subgroup IX) increased in WT plants but did not obviously change in WS plants at 9 h after waterlogging, indicating that these four SiNAC genes might play significant roles in sesame waterlogging tolerance.
Previous studies show that the dormant form of NTLs were activated by proteolytic cleavage once stimulated, and the activated TFs regulate the target genes expression after the degradation of cytoplasmic anchors [15, 34,53]. It is known that some of Arabidopsis NTLs, like NTM2, NTL4, NTL6, NTL8 and NTL9, involved in response to abiotic stresses [18, [54][55][56][57]. Phylogenetic analysis of sesame NTLs with Arabidopsis and rice NTLs (Fig 4) suggesting similar functions of these genes and provide a reference for investigation of putative functions of sesame NTLs in the future. In this study, the six detected SiNTL genes were induced rapidly by at least one type of stress treatment-waterlogging, drought, cold, osmotic, and salt stressesconsistent with previous reports [3,18,34]. Notably, SiNTL2/SiNAC18, SiNTL3/SiNAC19, and SiNTL4/SiNAC28 (homology of Arabidopsis AtNTL4) were significantly up-regulated after salt-stress treatment, but down-regulated after cold treatment, implying that these SiNTLs are functionally associated with the salt and cold stress responses. Overall, the expression profiles of SiNACs under various stresses suggest that different SiNACs may be involved in various signaling pathways and stress responses, and that one SiNAC gene also involved in multiple abiotic stress responses and stress resistance regulation in sesame.

Conclusions
In this study, 87 sesame NAC genes were identified and the distribution, basic classification, gene structure, and evolutionary characteristics of them were investigated. The expression profiles of SiNACs in different tissues can assist in understanding the molecular basis sesame development and growth. Furthermore, analyses of the expression profiles of SiNAC following various abiotic stress treatments indicated that they are highly active in responses to drought, waterlogging, cold, osmotic, and salinity treatments, reveal that sesame NACs may represent convergence points of different signaling pathways. This work may provide a strong basis for future functional recearch of NAC TFs in responses to abiotic stresses in sesame.

Physical positions and gene duplication analyses of NAC genes
The physical positions of SiNACs were mapped to the 16 sesame LGs using a GFF file downloaded from the Sinbase database [12]. An in-house Perl script was then used to draw graphic NAC genes into the LGs of the sesame genome [61]. For nomenclature, the prefix 'Si' for S. indicum was added followed by NAC and numbered refer to its physical position from top to bottom on sesame LGs 1-16.
Genome and NAC sequences of Arabidopsis and V. vinifera were extracted from the TIGR database (http://www.tigr.org/tdb/e2k1/ath1/) and Genoscope (http://www.genoscope.cns.fr). MCScanX was employed to analyze segmental duplication in the sesame genome and collinear analysis among the sesame, Arabidopsis, and V. vinifera genomes with default parameters [62]. Members of the NAC gene family with segmental duplication and collinear analysis were retrieved from the above data sets for further analysis. Circos was used to construct the diagram [63].

Phylogenetics analysis and molecular characterization
Clustal X 2.1 and MEGA 5.0 were used to construct NJ phylogenetic trees based on aa sequence of NAC members from sesame and Arabidopsis with 1000 bootstrap replicates [64,65]. A MEME v4.10.1 utility (Multiple Em for Motif Elicitation, http://meme-suite.org/index. html) was used to display the motifs of NAC proteins from sesame [66]. The exon/intron organizations of sesame NAC genes were performed from Gene Structure Display Server (http:// gsds.cbi.pku.edu.cn/) [67]. The membrane-bound SiNAC protein predictions were determined by SMART web server (http://smart.embl-heidelberg.de/) [59].

Plant growth, stress treatment, and RNA extraction
The sesame seeds were selected from sesame germplasm provided by the Oil Crops Research Institute, Chinese Academy of Agricultural Sciences, Wuhan, China. Uniform seeds of sesame cv. Zhongzhi No. 13 were sterilized with 3% sodium hypochlorite and washed four times using sterile water. The seeds were germinated on tow filter papers with sterile water in an illuminated incubator using a 16/8 h light/dark cycle at 28˚C. Three days later, seedlings were placed in half-strength Hoagland solution. For cold stress, 14-day-old seedlings were transferred to a illumination incubator at 4˚C. For osmotic-and salt-stress treatments, 14-day-old seedlings were treated with 15% PEG 6000 and 150 mM NaCl, respectively. The leaves of treated seedlings were harvested at 0 (control or CK), 2, 6, and 12 h for assays. All samples were immediately placed in liquid nitrogen and stored at -80˚C until use. Total RNA of seedlings was extracted using an EASYspin Plus kit (Aidlab, Beijing, China). The RNA of samples was reverse transcribed using a HiScript II 1st Strand cDNA Synthesis kit (Vazyme Biotech, Nanjing, China) with oligo (dT23) primer.

Expression analyses of SiNAC genes
The expression pattern of SiNAC genes was analyzed using three groups of transcriptome sequencing data obtained earlier by our group. For the expression analysis of SiNAC genes in different tissues, the transcriptome data were obtained from root, stem, leaf, flower, seed, and capsule of cv. Zhongzhi No. 13 under normal growth conditions. For the transcriptome data of drought stress treatment, samples of the roots of DT cv. ZZM0635 and DS cv. ZZM4782 were harvested when the soil water content was at 35% (d0 controls), 15% (d1 treatment), 9% (d2 treatment), and 6% (d3 treatment) at the early anthesis stage [45]. To determine the expression of sesame NAC genes in response to waterlogging, transcriptome data were obtained from roots of WT cv. Zhongzhi No. 13 and WS cv. ZZM0563 at 0, 3, 9, and 15 h after applying waterlogging [42]. Transcript abundance was calculated by RPKM. The hierarchical cluster analyses and heat maps were generated by MultiExperiment Viewer using log-transformed RPKM values [68].
To study the expression profiles of sesame NAC genes in response to cold (4˚C), osmotic (15% PEG 6000), and salinity (150 mM NaCl) stresses, qRT-PCR programs were performed on a LightCycler480 Real-Time PCR System using ChamQ™ SYBR 1 qPCR Master Mix (Vazyme Biotech, Nanjing, China) according to the manufacturer's instructions. Expression of 30 sesame NAC genes distributed in different subgroups were detected by qRT-PCR with three biological replications, and the sesame Histone H3.3 gene (SIN_1004293) was used as an internal control [69]. The gene-specific primers are listed in S5 Table. Relative gene expression data were analyzed with the 2 -ΔΔCT method [69].