The Responses of the Lipoxygenase Gene Family to Salt and Drought Stress in Foxtail Millet (Setaria italica)

Plant lipoxygenases (LOXs), a kind of non-heme iron-containing dioxygenases, participate plant physiological activities (especially in response to biotic and abiotic stresses) through oxidizing various lipids. However, there was few investigations on LOXs in foxtail millet (Setaria italica). In this study, we identified the LOX gene family in foxtail millet, and divided the total 12 members into three sub-families on the basis of their phylogenetic relationships. Under salt and drought stress, LOX genes showed different expression patterns. Among them, only SiLOX7 showed up-regulated expression in Yugu1 (YG1) and Qinhuang2 (QH2), two stress-tolerant varieties, indicating that SiLOX7 may play an important role in responses to abiotic stress. Our research provides a basis for further investigation of the role of LOX genes in the adaptation to abiotic stresses and other possible biological functions in foxtail millet.


Introduction
Drought and salt are both the main abiotic factors limiting the growth and development of plants, and thereby crop yield [1]. In many parts of the globe, the land has been suffering from salinization and increased drought condition, which will aggravate the negative impact on crops and endanger food production [2]. Drought and salt stress may cause membrane disorder, protein denaturation or inactivation by imposing osmotic stress on plants, and produce excessive reactive oxygen species leading to membrane oxidative damage, thereby causing plant growth disorders, photosynthesis inhibition and premature aging [3].
Lipoxygenases (oxygen oxidoreductase, EC 1.13.11.12; LOX) belong to non-heme ironcontaining fatty dioxygenases which catalyze the degradation of polyunsaturated fatty acids [4,5]. This kind of iron-containing dioxygenase was widely present in fungi, animals, plants, algae, protozoa and bacteria cells [6,7]. The major substrates of LOX enzymes are linoleic acid and linolenic acid in plants [8]. Lipoxygenases catalyze the production of oxylipins, which play a key role during the whole growth period of plants, such as seed germination and plant growth, especially against biotic and abiotic stresses [9][10][11][12]. The down-regulation of lipoxygenase activity in rice seeds under storage conditions reduced the deterioration of seed quality. Additionally, rice seeds with reduced LOX enzyme activity showed higher germination efficiency after storage [13]. The metabolites of LOX pathway are the fatty acid oxidation products mediated by LOX, known as oxylipins, such as jasmonates (JA), green leaf volatiles (GLVs) and death acid (10-OPDA, 10-oxo-11phytodienoic acid, and 10-OPEA, 10-oxo-11-phytoenoic acid, as well as a series of related portal.html; accessed on 1 October 2020) [39]. LOX sequences of Arabidopsis thaliana were accessed from TAIR website (https://Arabidopsis.org/; accessed on 1 October 2020) [40]. The LOX sequences of rice were downloaded from CHINA RICE DATA CENTER website (http://www.ricedata.cn/; accessed on 1 October 2020). Two main methods were used for the identification of the foxtail millet LOX gene family. First, with the BLAST program in TBtools [41], the Arabidopsis LOX protein sequences were used as queries to identify foxtail millet LOX genes. The superfluous sequences with the same chromosome locus were removed from the BLAST result. Second, we scanned of the proteome for the presence of "lipoxygenase" and "PLAT/LH2" domains. Then we excluded sequences that only had "lipoxygenase" or "PLAT/LH2". All predicted protein sequences were identified.
The TBtools software was used to visualize the location of the acquired genes on the chromosomes [41]. The ExPASy bioinformatics website (https://www.expasy.org/; accessed on 1 October 2020) was used to predict the molecular weight and isoelectric point of the foxtail millet LOX proteins.

Multiple Sequence Alignment and Phylogenetic Tree Construction
The MUSCLE program [42], with default settings, was used to align the LOX protein sequences. Editing and visualization of alignment was on the Jalview [43].Then, six Arabidopsis thaliana LOX, twelve rice LOX and thirteen foxtail millet LOX proteins were used for building a phylogenetic tree. The phylogenetic tree was constructed with the Molecular Evolutionary Genetics Analysis (MEGA 7) software and the Neighbor-Joining method using 1000 bootstrap values [44]. Using the same methods generated another phylogenetic tree, only based on twelve foxtail millet LOX proteins.

Conserved Motifs, Protein Secondary Structure, and Subcellular Localizations Analyses
The MEME Suite (http://meme-suite.org/index.html; accessed on 1 October 2020) [45] was applied to search for motifs, and with default settings it was expected the maximum number to be found was set to 20. The TBtools [41] software was used to analyze the phylogenetic tree and motifs of foxtail millet LOX. In addition, sequence logos for the conserved LOX domain of Foxtail millet LOX proteins were constructed by WebLogo [46].

Plant Material and Stress Treatment
In this study, we used three varieties of foxtail millet with different sensitivities to salt and drought stress, 'Qinhuang2' (QH2, stress-tolerant), 'Yugu1' (YG1, stress-tolerant) and 'AN04' (stress-sensitive). Mature non-dormant seeds were sown into pots with nutrient soil and vermiculite (4:1, v/v). Plants were cultivated manual climatic box at Shanxi Agricultural University, and grown under controlled environmental setting of 28°C day/22°C night, 16 h light/8 h dark photoperiod, the light intensity was 14,000 lx. Seeds were sowed three rows per rack, one variety (20 plants) per row, four racks per treatment to ensure sufficient material. Stress treatment were performed when seedlings had six true leaves. For the drought stress treatments, when the water content of soil was less than 15%, start harvest samples. For the salt stress treatments, the seedlings were watered by 500 mmol/L NaCl solution. This certainly did not change the environment setting of manual climatic box. The above-ground part of a seedling was collected for physiological parameters determination and qRT-PCR. The materials at each time point were harvested three seedlings as biological replicates. All the samples were promptly frozen by liquid nitrogen after being harvested, and stored at −80°C ultra-low temperature freezer to prevent RNA degradation.

Total RNA Extraction, Inverse Transcription, qRT-PCR Analysis
To extract total RNA, the frozen foxtail millet seedling samples were pestled with liquid nitrogen to powder. Total RNA from all simples were extracted with RNaiso Plus reagent (Takara Biotechnology, Beijing, China). To check the quality of the RNA samples, the ultra-low volume spectrometer (BioDrop, Cambridge, UK) was used to measure the concentration of RNA samples and A260/A280. Select RNA samples with A260/280 ratio between 1.8-2.1 for subsequent RNA reverse transcribed into cDNA. Used PrimeScript™ RT reagent Kit with gDNA Eraser (Takara Biotechnology, Beijing, China) and follow the manufacturer's instructions to make RNA reverse transcribed into cDNA, and cleaned the gDNA. We diluted cDNA five times for qRT-PCR.
qRT-PCR was performed on a Bio-Rad CFX96 Real Time PCR machine, using TB Green ® Premix Ex Taq™ II (Takara Biotechnology, Dalian, China). qRT-PCR conditions: 95°C for 30 s (initial denaturation), 95°C for 5 s (denaturation),60°C for 30 s (annealing and extending), step denaturation to annealing and extending for 40 cycles, 60°C to 95°C and increment 0.5°C for 5 s (melt curve). Relative gene expression levels were calculated using the ∆∆Cq method [47] by Bio-Rad Manager 3.1 software, and ACTIN was used as an internal control. Gene-specific primers were designed using Primer Premier 5.0 software (Supplementary Table S1).

Determination of Physiological Parameters Related to Salt and Drought Stress
The activities of superoxide dismutase (SOD), peroxidase (POD) and concentration of malondialdehyde (MDA) were measured by spectrophotometer. Fresh seedling of 0.1 g were homogenized in 1.5 mL 0.1% trichloroacetic acid (TCA), and centrifuged at 12,000 rpm for 15 min at 4 • C; the resulting supernatant was used to measure MDA by 2-thiobarbituric acid (TBA) reaction [48]. We used 0.1 g of seedling samples homogenized in 1 mL pH 7.8, 50 mM phosphate buffer. Afterwards, the homogenate was centrifuged at 12,000 rpm for 15 min at 4 • C, and we used the supernatant to determine enzyme activities. SOD activity were measured using the nitrotetrazolium blue chloride Illumination method [49], and the activity of POD was measured in the light of the guaiacol method [50].

Identification and Characterization of LOX Genes in Foxtail millet
To identify the LOX gene family in foxtail millet (SiLOXs), we used protein sequence homology searching and scanning of the proteome for the presence of "lipoxygenase" and "PLAT/LH2" domains. Firstly, we compared the protein sequences of Arabidopsis LOXs with the proteome of foxtail millet. As a result, 13 SiLOXs were identified and designated as SiLOX1-13 (Supplemental Table S2). Through protein structure analysis, SiLOX12 was excluded, as it lacked the PLAT/LH2 domain. All the other 12 members were identified as SiLOX genes, with the standard encoding sequences of LOX genes [51].
Foxtail millet LOXs were located on 7 chromosomes, except Chromosomes 2 and 8, with the most LOXs located on Chromosome 9. Among them, chromosome 9 had 5 LOX genes, chromosome 7 had 2 LOX genes, chromosome 1,3,4,6 only had one LOX gene ( Figure 1). The CDS length of SiLOX genes varied between 1829-2970 bp, the encoded SiLOX proteins had 599 to 973 amino acids and the predicted molecular weight ranged from 69,190.23 Da to 106,479.19 Da. The isoelectric point (PI) ranged from 5.55 to 8.59 (Supplementary Table S2). Using the GO entries in the Phytozome database, we searched different classifi of GO function annotations of LOX gene in foxtail millet (Table 1). Among all S genes, SiLOX1, SiLOX2, SiLOX3, SiLOX5, SiLOX7, SiLOX8, SiLOX9, SiLOX10 an LOX13 all contained GO:0005515, GO:0016702, GO:0046872, GO:0055114. But SiLOX LOX6 and SiLOX11 excluded GO:0005515. The GO items enriched in the SiLOX f were mainly concentrated in the molecular functions of oxidoreductase activity, met binding, and the biological process of oxidation-reduction process. In addition, som LOX genes were annotated in molecular function of protein binding.

Phylogenetic Analysis of the SiLOX Gene Family
To show the evolutionary relationships of the LOX family members in differen cies, we constructed a Neighbor-Joining phylogenetic tree with 1000 bootstrap of 31 LOX amino acid sequences including 12 SiLOXs, 12 OsLOXs (from rice) and 6 At (from Arabidopsis). Phylogenetic tree categorizes the genes into 3 kinds of LOXs LOX,13-LOX Type Ⅰ and 13-LOX Type Ⅱ ( Figure 2). Among 12 SiLOXs, 8 SiLOXs (SiL SiLOX3, SiLOX4, SiLOX5, SiLOX7, SiLOX9 and SiLOX13) were characterized into 9 enzymes group with 2 AtLOXs (AtLOX1 and AtLOX5) and 6 OsLOXs (L-2, r9-L Using the GO entries in the Phytozome database, we searched different classification of GO function annotations of LOX gene in foxtail millet (Table 1). Among all SiLOX genes, SiLOX1, SiLOX2, SiLOX3, SiLOX5, SiLOX7, SiLOX8, SiLOX9, SiLOX10 and SiLOX13 all contained GO:0005515, GO:0016702, GO:0046872, GO:0055114. But SiLOX4, SiLOX6 and SiLOX11 excluded GO:0005515. The GO items enriched in the SiLOX family were mainly concentrated in the molecular functions of oxidoreductase activity, metal ion binding, and the biological process of oxidation-reduction process. In addition, some SiLOX genes were annotated in molecular function of protein binding. Table 1. Statistics list of GO classification of LOX gene in foxtail millet.

Conserved Protein Domain and Motifs of SiLOXs
PLAT/LH2(IPR001024) and lipoxygenase (IPR013819) domain were identified in almost all 12 SiLOX proteins ( Table 2). To further understand the relationship among SiLOXs, we used 12 SiLOXs protein sequences for multiple alignment, and constructed a separate phylogenetic tree based on these multiple alignments. In addition, we used MEME Suite (Ver.5.3.3) to predict motifs in SiLOXs. SiLOXs were divided into 3 sub-families, as described above (9-LOX, 13-LOX I and 13-LOX II). We also found 20 motifs in SiLOXs, and nineteen of 20 motifs were shared among all SiLOXs protein (Figure 3a). Almost all 9-LOX members had all 20 motifs except SiLOX4 (which lacked 2 motifs). protein secondary structures were mainly composed of random coil and α-helix, with random coils in higher proportion. In addition, these secondary structures also contain extended strand and β-turn (Table 3).  SiLOXs. The first H of motif1 in SiLOX4 and SiLOX11 was replaced by N and Q ( Figure  3b,c), All 12 SiLOX protein secondary structures were mainly composed of random coil and α-helix, with random coils in higher proportion. In addition, these secondary structures also contain extended strand and β-turn (Table 3). For the prediction of the subcellular localization of the SiLOX proteins, SOPMA program was used. The results showed that 7 SiLOX proteins were located at cytoplasm (Si-LOX1, SiLOX2, SiLOX3, SiLOX4, SiLOX7, SiLOX8 and SiLOX9) and 6 SiLOX proteins in chloroplast. As previously described, all 13-LOX type Ⅱ contained chloroplast transit peptide, whereas 13-LOX type Ⅰ might be localized in cytoplasm (Supplementary Table S2).  For the prediction of the subcellular localization of the SiLOX proteins, SOPMA program was used. The results showed that 7 SiLOX proteins were located at cytoplasm (SiLOX1, SiLOX2, SiLOX3, SiLOX4, SiLOX7, SiLOX8 and SiLOX9) and 6 SiLOX proteins in chloroplast. As previously described, all 13-LOX type II contained chloroplast transit peptide, whereas 13-LOX type I might be localized in cytoplasm (Supplementary Table S2).

Physiological Responses of Different Varieties of Foxtail millet to Abiotic Stresses
We measured the physiological parameters related to stress in three varieties foxtail millet under two abiotic stresses, salinity and drought conditions. The results revealed that the trends of physiological indexes of the three varieties had an obvious relationship with their respective tolerance. Under drought stress, the MDA content of QH2 and YG1 was lower than that of AN04, while the MDA content of YG1 increased sharply from 4 to 6 days, then decreased to lower than that of AN04 from 6 to 8 days. Similarly, under salt stress, MDA content showed a similar variation trend, but the MDA content of QH2 increases from 0-2 d and is higher than that of AN04, and then began to decrease after 2 d and gradually falls below that of AN04. The content of MDA was a reflection of the peroxidation degree of the plant cell membrane, and was considered as one of the indicators of plant exposure to abiotic organisms. Compared with the tolerant varieties YG1 and QH2, the sensitive variety AN04 had a significantly higher MDA content and suffered more. Abiotic stress increases the content of ROS (mainly O 2− and H 2 O 2 ) in plants, which can damage plant cells. Peroxidase (POD) and superoxide dismutase (SOD) were antioxidant enzymes in plants, which eliminate peroxidase damage by scavenging ROS such as H 2 O 2 in plants. The results showed that the POD activity firstly increased and then decreased under salt stress and drought stress. As expected, POD activity of tolerant varieties YG1 and QH2 was higher than that of sensitive variety AN04. The POD activity of YG1 was significantly higher than that of QH2 at 4 d, 6 d (drought stress) and 4 h (salt stress). The activities of SOD and POD showed a similar trend under the two stress conditions, and the SOD activity in AN04 was lower than that in YG1 and QH2. Under drought stress, SOD activity of QH2 increased faster than that in YG1, and then decreased. In addition, SOD activity of QH2 was almost always lower than that in YG1 under salt stress (Figure 4).

Expression Analysis of SiLOX Genes under Salt and Drought Stress
To explore the potential roles of LOX genes in foxtail millet against abiotic stresses, we investigated the expressions of SiLOXs under salt and drought treatments. The results of qRT-PCR showed that the expressions of all detected SiLOX genes were different under each stress condition ( Figure 5). To eliminate the influence of plant circadian rhythm and biological clock on gene expression levels during drought stress, leaves were sampled at 10 a.m. for every sampling points. For salt stress treatment during 12 h on the same day, we detected the gene expression level of untreated foxtail millet, and the expression pattern of SiLOXs were displayed by relative expression level which was a compared percentage of expression level of treated plants and control.

Expression Analysis of SiLOX Genes under Salt and Drought Stress
To explore the potential roles of LOX genes in foxtail millet against abiotic we investigated the expressions of SiLOXs under salt and drought treatments. Th of qRT-PCR showed that the expressions of all detected SiLOX genes were differen each stress condition ( Figure 5). To eliminate the influence of plant circadian rhy biological clock on gene expression levels during drought stress, leaves were sam 10 a.m. for every sampling points. For salt stress treatment during 12 h on the sa we detected the gene expression level of untreated foxtail millet, and the express tern of SiLOXs were displayed by relative expression level which was a compa centage of expression level of treated plants and control.
Under both salt and drought stress, these genes showed changed expressio SiLOX5 showed the tendency of significant increasing expression after salt and stress. SiLOX1 in the three varieties showed high expression in all periods of drou salt stress. SiLOX2 showed a tendency to significantly reduced gene expressi drought stress, but increased under salt treatment. However, the expression of was significantly reduced under drought stress, with opposite pattern under sa Under drought stress, the expression of SiLOX8 and SiLOX10 showed a downwar Under salt stress, the expression of SiLOX8 increased first and then decreased, the expression of SiLOX10 decreased first and then increased.
In drought-tolerant varieties YG1 and QH2, the expression levels of SiLO significantly upregulated from by salinity stress, and this increasing trend start after treatment. The expression level of SiLOX7 was increased sharply at 4d and r maximum at 6d in QH2 under drought treatment. However, in YG1, the express kept a high level after 2d, which was similar to the 9-LOX gene subfamily in cotto Under both salt and drought stress, these genes showed changed expression level. SiLOX5 showed the tendency of significant increasing expression after salt and drought stress. SiLOX1 in the three varieties showed high expression in all periods of drought and salt stress. SiLOX2 showed a tendency to significantly reduced gene expression after drought stress, but increased under salt treatment. However, the expression of SiLOX9 was significantly reduced under drought stress, with opposite pattern under salt stress. Under drought stress, the expression of SiLOX8 and SiLOX10 showed a downward trend. Under salt stress, the expression of SiLOX8 increased first and then decreased, whereas the expression of SiLOX10 decreased first and then increased.
In drought-tolerant varieties YG1 and QH2, the expression levels of SiLOX7 were significantly upregulated from by salinity stress, and this increasing trend started at 2 h after treatment. The expression level of SiLOX7 was increased sharply at 4 d and reach the maximum at 6 d in QH2 under drought treatment. However, in YG1, the expression level kept a high level after 2 d, which was similar to the 9-LOX gene subfamily in cotton under drought and salt stress [33]. Interestingly, SiLOX11 expression was considerably upregulated in both salt and drought stress only in QH2. In addition, SiLOX6 was upregulated in drought-sensitive variety AN04. We analyzed the correlation between the expression of some LOX genes (SiLOX6, Si-LOX7 and SiLOX11) and physiological parameters under salt stress. The expression levels of these genes were significantly positively correlated with the content of POD, and SiLOX7 was the most significant. (Supplemantary Figure S1). These results indicated that SiLOX7 and SiLOX11 may regulate the responses to abiotic stress in foxtail millet.
with the content of POD, and SiLOX7 was the most significant. (Supplemantary S1). These results indicated that SiLOX7 and SiLOX11 may regulate the responses otic stress in foxtail millet.

Discussion
With the development of genome sequencing technology, more and more LO have been identified in different species. Moreover, research on the growth and d ment of the LOX gene family and its response to biological and abiotic stresses ha ually been extensive. However, relevant research in the past involved nothing ab foxtail millet LOX gene family. Almost all the plant LOX gene families that hav identified are based on the 6 members of the model plant Arabidopsis LOX gene [54], such as 64 putative LOXs in 4 species of cotton [33], 14 deduced LOXs in toma 9 LOXs in sorghum [4], 13 LOXs in maize [29], 14 LOXs in rice [5] and more. In search, we identified the LOX gene family in foxtail millet, which included 12 m more than twice as many members of the Arabidopsis LOX gene family, but one le the members of the LOX gene family in rice. Interestingly, the LOX genes in mi maize were also about twice as those in Arabidopsis. The differences in the num members and their distribution on the chromosomes might indicate that the LO

Discussion
With the development of genome sequencing technology, more and more LOX genes have been identified in different species. Moreover, research on the growth and development of the LOX gene family and its response to biological and abiotic stresses has gradually been extensive. However, relevant research in the past involved nothing about the foxtail millet LOX gene family. Almost all the plant LOX gene families that have been identified are based on the 6 members of the model plant Arabidopsis LOX gene family [54], such as 64 putative LOXs in 4 species of cotton [33], 14 deduced LOXs in tomato [11], 9 LOXs in sorghum [4], 13 LOXs in maize [29], 14 LOXs in rice [5] and more. In this research, we identified the LOX gene family in foxtail millet, which included 12 members more than twice as many members of the Arabidopsis LOX gene family, but one less than the members of the LOX gene family in rice. Interestingly, the LOX genes in millet and maize were also about twice as those in Arabidopsis. The differences in the number of members and their distribution on the chromosomes might indicate that the LOX gene had not been conserved in the evolutionary process.
Through phylogenetic analysis, the foxtail millet LOX genes were classified into three subfamilies based on protein structure and sequence similarity. Consistent with the description of Feussner and Wasternack [51], the chloroplast transit peptide was present at the N-terminus of the Type-II LOX protein sequences and excluded from the Type-I LOX proteins. In addition, Type-II LOX only contain 13-LOX Type II, but 9-LOX and 13-LOX Type-I together constitute the Type-I LOX. Through the multiple sequence alignment of LOX protein sequences in foxtail millet, we observed some conserved motifs. Like tomato and cotton LOX genes, there was a highly conserved histidine-rich motif His-(X)4-His-(X)4-His-(X)17-His-(X)8-His in these conserved motifs, which contains 38 amino acids [11,33]. Interestingly, we found that SiLOX4 and SiLOX10 lacked the first His in this motif among 12 predicted SiLOXs. Therefore, we speculate that SiLOX4 and SiLOX10 may have different enzymatic activities from other members. This variation may be due to the diversity of LOX in the evolution process. Previous studies had shown that replacing the C-terminal isoleucine with valine can significantly increase the activity of lipoxygenase, but lipoxygenase may be inactivated due to the substitution of other amino acids. Among all the predicted amino acid sequences of SiLOX, 12 members have isoleucine at the C-terminus, while SiLOX4 had an asparagine substitution. Based on the above findings, SiLOX4 was more likely to have different enzyme activities from the other SiLOXs. According to the identification rules by Feussner and Wasternack, SiLOX12 was excluded from the LOX gene family, as it contained only the lipoxygenase domain [51].
LOX enzymes play key roles in almost all stages of plant life. In addition to participating in plant germination, growth, development and other biological activities, it also plays an important role in the response of plants to resist adversity [9,10,25]. The predicted LOX genes in foxtail millet are all annotated to the JA biosynthetic pathway. JA improves the tolerance of plants to salt and drought stress [20,21], suggesting that the members of the LOX gene family in foxtail millet might be related to improve the stress tolerance.
After analysis of the expression patterns of SiLOX genes under two abiotic stresses by qRT-PCR, we found that some of the genes show an up-regulation trend under drought or salt stress, which indicates that there may be similar tolerance mechanisms in foxtail millet to respond to drought and salt stress. In cotton, most GhLOX genes are related to heat and salt stress [33]. The similar expression pattern of LOX genes under different stresses may reflect the cross adaptation of plants. Some SiLOX genes showed different expression patterns under drought and salt stress. SiLOX2, SiLOX6, SiLOX8 and SiLOX9 were up-regulated under salt stress and down-regulated under drought stress. In QH2, SiLOX10 and SiLOX11 are down-regulated under drought stress and up-regulated under salt stress. It was noteworthy that SiLOX7 in stress-tolerant varieties QH2 and YG1 showed a tendency of up-regulation under both stresses, especially in QH2. However, no significant changes were observed in sensitive variety AN04. The homologous genes of SiLOX7 in Arabidopsis were AtLOX1 and AtLOX5. Both the Arabidopsis mutants lox1 and lox5 have reduced stress tolerance with the increased MDA content [55]. Therefore, we speculated that SiLOX7 plays an important role in improving the stress tolerance of foxtail millet.
By integrative analysis of the expression pattern of SiLOX genes, MDA content, the activities of SOD and POD both under drought and salt stress the possible mechanism of the influence of the SiLOX genes on the tolerance was explored in foxtail millet. In the three varieties of foxtail millet, the activities of SOD and POD all showed a trend of increasing at first then decreasing, whereas MDA generally rose slowly. It was observed from the results that the physiological parameters of the three varieties under stresses were correlated with the expression of SiLOX7 to some extent. The activities of POD and SOD in YG1 and QH2 with high SiLOX7 expression was significantly higher than those of AN04 with low SiLOX7 expression, especially the activities of POD, which were more obvious ( Figure S1). The MDA contents in AN04 were generally higher than those in QH2 and YG1. Plants produced ROS when suffered by abiotic stresses, which will damage the structure of plants in different aspects. Enzymes such as POD, SOD and some non-enzymatic antioxidants improved the stress tolerance of plants by scavenged ROS and free radicals produced by plants [56]. Therefore, the higher enzyme activities of POD and SOD in YG1 and QH2 than in AN04 may be related to their tolerance to stress. In addition, the content and increase scale of MDA in these two tolerant varieties were generally less than those in AN04, indicating that the antioxidant capacity of YG1 and QH2 was higher than that in AN04. We speculated that the trend of MDA in tolerant varieties was related to the increases in the activities of POD and SOD and the up-regulation of LOX genes such as SiLOX7, thereby reducing the degree of oxidation.

Conclusions
This research is the first genome-wide identification of the LOX gene family in foxtail millet. A total of 12 SiLOX genes were identified in foxtail millet, and they were highly conserved. The similarity and specificity of SiLOX genes in different varieties under drought and salt stress were found. Comprehensive analysis of gene expression patterns and physiological parameters related to stress suggested that SiLOX7 could be involved in the stress tolerance in foxtail millet, such as affecting the activities of some antioxidant enzymes to resist the damage of peroxides to plants. The above research may support further analysis the function of foxtail millet LOX genes.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/life11111169/s1, Table S1: Primers information for qRT-PCR; Table S2: Information of the SiLOX genes. Including gene names and gene ID, chromosome location, gene length, CDS length, protein length, physical and chemical properties and prediction of subcellular location; Figure S1.