Genetic Transformation of LoHDZ2 and Analysis of its Function to Enhance Stress Resistance in Larch

To study the function of LoHDZ2 in larch, we rst constructed a p1302-LoHDZ2::GUS overexpression vector. Through Agrobacterium-mediated infection, the expression vector was transferred into a larch embryogenic cell line. A stable resistant cell line was subsequently screened, and mature embryos were induced to grow until they developed into seedlings. Antagonistic cell lines were identied at both the DNA and RNA levels. The transgenic cell lines were then subjected to GUS staining, and transgenic cell lines were ultimately identied and obtained. These transgenic cell lines were sequenced to identify differentially expressed genes, and a cluster analysis was performed. The resistant cell lines were cultured under stress conditions involving 20% PEG 6000 and 200 mM NaCl proliferation media (1/10-BM). After the stress treatment, the contents of peroxidase (POD), malondialdehyde (MDA) and superoxide dismutase (SOD) in both wild-type and transgenic cell lines were measured. The results are summarized below. the transgenic cell line. The transgenic cell lines showed a trend of "rst decreasing, then increasing and then decreasing again", and the change trend of the three transgenic cell lines was very consistent. When the calli were treated for 96 h, the values of OE2 and OE4 were the same as those of the wild-type calli, but the value of OE1 was higher than that of the wild-type and the other two transgenic cell lines. These results show that under drought stress, the POD enzyme activity of the transgenic cell line is stronger than that of the wild


Introduction
Larix olgensis is an important coniferous tree species in China. Its trunk is straight, and its wood is hard, of good quality and durable [1] . In this study, through the exploration of the internal molecular mechanism of Larix olgensis, growth stress resistance-related genes of Larix olgensis were identi ed, and fastgrowing, stress-resistant Larix olgensis was developed by genetic improvement.
Homeodomain leucine zippers (HD-Zips) are kinds of plant-speci c transcription factors belonging to the homeobox protein family [2] . Based on characteristics such as sequence conservation, gene structure and physiological function, HD-Zip transcription factors can be divided into four subclasses: HD-ZipI, HD-ZipII, HD-ZipIII and HD-ZipIV subclasses [3] . Due to the differences in genetic sequence and protein structure of members of the different subfamilies, HD-Zip transcription factors participate in different plant development processes and regulate different metabolic processes.
Based on previous laboratory-based research results of HD-ZipII subfamily genes, to further study the function of these genes, we cloned LoHDZ2 and overexpressed it in larch calli by the Agrobacteriummediated method to obtain transgenic larch. The changes in physiological and biochemical indices of transgenic larch under drought and salt stress were then evaluated via transcriptome sequencing, and the function of LoHDZ2 was preliminarily studied.

Plant materials
Embryogenic cell lines of Larix olgensis preserved in our laboratory were used as research objects, and fresh embryogenic calli cultured for approximately 14 days were used as plant materials. The wild-type larix olgensis cell lines used in the experiment were induced in the early stage of the laboratory and stored in the State Key Laboratory of Forest Tree Genetics and Breeding(Northeast Forestry University).The basic medium used was 1/10-BM proliferation medium, and dark culture was carried out at room temperature at 25 ℃.
According to the LoHDZ2 sequence of Larix olgensis (the gene number in the NCBI database is MW206675), ampli cation fragments for speci c primers were designed, and a p1302-LoHDZ2::GUS plant overexpression vector was constructed. The construct was successfully transformed into competent Agrobacterium tumefaciens cells for plant genetic transformation [4] .

Genetic transformation of larch
Fresh embryogenic callus lines were selected from 1/10-BM callus proliferation media for genetic introduction of p1302-LoHDZ2::GUS. The genetic transformation of larch was carried out by Agrobacterium-mediated infection. The calli of larch were infected with infection solution with an OD600 of 0.5 for 20 minutes and cocultured on media for 2 days. After coculture, the calli were cleaned with sterilized water three times and then transferred to culture media consisting of 500 mg/L Cef for strict sterilization two times (each time for 5 min). After cleaning, the calli were placed on sterile lter paper to absorb excess water. Then, they were transferred to screening media consisting of 4 mg/L Hyg to screen for resistant larch calli. After three resistance screenings, transgenic larch calli were obtained [5,6] . Then, the transgenic embryogenic cell line was induced to develop into somatic embryos via somatic induction media, after which the somatic embryos were induced to develop into transgenic larch seedlings by 1/2strength MS rooting media [7] .

Acquisition and detection of resistant calli
The DNA of wild-type and transgenic embryogenic calli was extracted, and the following full-length primers were used: LoHDZ2-F (ATGGAAGAGATGAAGAACAAGCA) and LoHDZ2-R (TTAGCAAGCTGCAGACTGTTGG). PCR detection showed that the ampli ed fragment was approximately 1000 bp. The PCR reaction mixture comprised 2 µl of template DNA, 1 µl of upstream and downstream primers, 1 µl of EasyTaq ® DNA polymerase, and 20 µl of P-mother liquor (P-mother liquor was prepared by mixing the following reagents: 10× EasyTaq ® buffer:2.5 mM dNTPs:H 2 O=5:3:32). The reaction procedure was as follows: 94 °C for 3 min, 35 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 1 min, 72 °C for 7 min, and then a pause at 16 °C.
RNA from transgenic larch calli was extracted and reverse transcribed into cDNA for qRT-PCR. Primer 5 software was used to design gene quantitative primers (LoHDZ2-RT-F, CTTGGCGTTGGTGTGTCTATG, LoHDZ2-RT-R, TGGGCATGAACCAAAGAAAC). An ABI 7500 uorescence qRT-PCR instrument was used, and a dissolution curve was generated according to the standard procedure of the ABI 7500 instrument. The reaction conditions were 94 ℃ for 30 s, 94 ℃ for 5 s, 60 ℃ for 15 s and 72 ℃ for 10 s. Forty cycles were required from the second step to the fth step, and the difference among the three CT values was less than 1. Microsoft Excel 2016 was used for data analysis, with the -ΔΔCT formula used for calculations. Figures were constructed with GraphPad Prism 5 software. The internal reference gene was a Larix olgensis gene (the gene number on the NCBI database is MF278617.1), and the wild-type was used as a control. The expression of LoHDZ2 in the transgenic calli was analyzed on the basis of the qRT-PCR results.
In terms of growing conditions for GUS-stained plants, the transgenic resistant embryogenic cell line was cultured in 1/10-BM media for 10 days. With respect to GUS staining, tissue from resistant plants was removed and immersed in an appropriate amount of X-gluc (Sigma) staining solution, which was then subjected to a vacuum until there were no bubbles in the staining solution, afterward, the tissue was incubated at 37 ℃ for 12-16 h [8] . GUS staining was observed under an Olympus microscope BX51 (Japan). Four transgenic lines were selected, and 0.6 g of each transgenic line was used for GUS staining, which was repeated 3 times. Wild-type tissues were used as negative controls to observe the color development of the calli.

RNA-seq analysis
Transcriptome sequencing and bioinformatics analysis The total RNA of three transgenic cell lines (OE1, OE2 and OE4) and the wild-type callus sequencing samples was extracted and sent to Lianchuan Biotechnology Company for total RNA detection and HiSeq sequencing. The method was the same as that in Jianzhong Hu's study on the Arabidopsis transcriptome [9] .
Clean reads were obtained from the data after ltering. The original sequencing data clean reads were assembled de novo into a unigene sequence set by using Trinity assembly software. Mapping data were obtained by comparing the unigenes with the clean reads for library quality evaluation [10] . Differential gene expression analysis, functional annotation and enrichment analysis of the differentially expressed genes were carried out according to the gene expression differences between different sample groups [11] .

qRT-PCR-based veri cation
Taking the cDNA of the transcriptome sample returned by the company as the template, 10 genes (5 upregulated genes and 5 downregulated genes) were selected from the differentially expressed genes revealed by the transcriptome sequencing for data veri cation. qRT-PCR primers were designed by Primer 5.0. The internal reference gene used was LoB80280 (the gene number in the NCBI database is MF278617.1). The quantitative primers of the 10 differentially expressed genes are shown in Table 1. The method used was the same as that above.

Growth of and gene expression in transgenic larch calli under different treatments
Fresh wild-type and transgenic embryogenic callus lines growing for 10 days were selected and cultured on 20% PEG 6000 and on proliferation media supplemented with 200 mM NaCl (1/10-BM) (prepared according to the ratio of 1:20 g/ml). The materials were removed at 0 h, 24 h, 48 h, 72 h and 96 h after stress treatment, frozen in liquid nitrogen, and stored at -80 °C for subsequent tests. Each treatment included three replications, with three callus lines placed in each plate, and the fresh weight of each callus was 0.6 g (three replications per plate). Then, the RNA of the treated transgenic calli was extracted and reverse transcribed into cDNA for real-time qRT-PCR. The speci c method used was the same as that above.
The samples were stored at -80 °C, and the contents of peroxidase (POD), malondialdehyde (MDA) and superoxide dismutase (SOD) were determined. A Suzhou Grace Biotechnology Co., Ltd., test kit (spectrophotometer method) was used according to the manufacturer's instructions.

Genetic transformation of larch
Embryogenic calli on 1/10-BM proliferation media were selected for genetic transformation of larch. The genetically transformed calli were cocultured for approximately 2 days and then transferred to culture media consisting of 500 mg/L Cef for strict sterilization. After three separate cultures on the screening media, ve resistant cell lines were obtained. The resistant cell lines were then cultured in proliferation media for 10 days, after which they then in 1/4-BM transition media for 10 days. Afterward, the calli were placed in somatic embryo induction media for somatic embryo induction to obtain larch transgenic somatic embryos (Fig. 1). Then, the somatic embryos were placed in 1/2-strength MS rooting media to obtain transgenic larch seedlings, as shown in Fig. 2.

Molecular detection of transgenic LoHDZ2 calli
Five transgenic callus cell lines growing on 1/10-BM were randomly selected to extract plant genomic DNA. The DNA was subsequently used as template for PCR detection. p1302-LoHDZ2::GUS plasmids were used as positive controls, and the wild-type was used as a negative control. The PCR detection results of p1302-LoHDZ2::GUS-transformed embryogenic calli are shown in Fig. 3. The results showed that four transgenic calli produced bands at corresponding positions, which preliminarily showed that p1302-LoHDZ2::GUS had been integrated successfully into the larch genome.
The ve abovementioned resistant callus cell lines growing on 1/10-BM media and one wild-type callus cell line were collected, and their RNA was extracted and then reverse transcribed into cDNA. The cDNA was used as a template for qRT-PCR-based detection, and wild-type cDNA was used as a control. The results showed that the expression levels of transgenic lines OE1, OE2 and OE4 were higher, approximately 1.9, 2.9 and 3.6 times higher, respectively, than that of the wild-type cell lines (Fig.  4).
To further con rm that p1302-LoHDZ2::GUS was integrated into the resistant calli, GUS staining solution was applied to wild-type larch calli and resistant calli. The results showed that four of the transgenic calli turned blue, while the wild-type callus in the control group exhibited no color change, indicating that the recombinant p1302-LoHDZ2::GUS plasmid had been successfully integrated into the larch calli (Fig. 5,   Fig. 6).
According to the detection of the DNA and RNA in the transgenic cell lines and the analysis of the GUS staining results, the OE1, OE2 and OE4 transgenic cell lines were ultimately selected as follow-up research materials.

RNA-seq analysis
Unigene functional annotation The assembled unigenes do not have corresponding functional annotations, so we added corresponding functional annotations using Diamond, a new comparison software similar to BLASTX. Because the sequences of similar functional genes (nucleic acid sequences or protein sequences) are highly conserved among different species, we selected six authoritative databases, namely, the NCBI NR, GO, KEGG, Pfam, SwissProt and eggNOG databases [12] (Tab.2).

qRT-PCR-based veri cation
To verify the accuracy and reliability of the transcriptome data, 10 differentially expressed genes (5 upregulated and 5 downregulated ones) were selected for qRT-PCR-based veri cation. The results show that these genes have different expression fold changes, as detected by RNA-seq and qRT-PCR (Fig.  7). This may be because the sensitivity of the two detection methods is different. Nonetheless, the expression trends are consistent, indicating that the transcriptome data are reliable and can be used for further gene functional analysis.

Screening of differentially expressed genes
In this study, DESeq2 was used to screen differentially expressed genes. The differentially expressed genes among different varieties were screened on the basis of their expression fold change (|log2(fold change)|>1) and signi cance level (pvalue < 0.05). The results showed that there were 304 differentially expressed genes in transgenic larch cell lines compared with wild-type cell lines: 167 upregulated differentially expressed genes and 140 downregulated differentially expressed genes (Fig. 8).

Cluster analysis of differentially expressed genes
To compare the clustering patterns of the differentially expressed gene expression pro les of the transgenic and wild-type calli, a clustering heatmap of differentially expressed genes was constructed for the genes with a large variance in expression in different samples (Fig. 9). The results showed that compared with that in the wild-type resistant calli, the expression of many genes in the transgenic resistant calli changed, the number of upregulated genes especially changed, which was approximately 176. There were 140 downregulated genes, indicating that the overexpression of LoHDZ2 has a positive regulatory effect on downstream genes. The heatmap shows that the transcriptome sequencing of the three biological replicates of each sample have good consistency, indicating that the sequencing data are relatively reliable.
To further explore the function of LoHDZ2, the differentially expressed genes with the largest variance in the transgenic and wild-type samples were used to construct a cluster heatmap of differentially expressed genes, and 46 upregulated genes and 30 downregulated genes with large expression fold changes were selected for NCBI BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi). When the results were compared with the information in the various aforementioned databases, relevant annotation and species information was obtained .
Among the 46 differentially expressed genes, including 22.0 kDa class IV, heat-shock protein-like 22.0 kDa IV, heat stress transcription factor A-1-like, heat-shock proteins, and heat-shock cognate 70 kDa protein, most have been annotated in broad-leaved tree species [13] . The annotated LRR receptors, such as the serine/threonine protein kinase FLS2 LRR receptor and serine/threonine protein kinase FLS2, may inhibit cell proliferation and stimulate extracellular matrix synthesis depending on the cell type. In addition, the xyloglucan galactosyltransferase GT11, which affects the internal mechanisms of plant growth and development, was also annotated [14] .
Among the 30 downregulated differentially expressed genes, poly[ADP ribose] polymerase 3 isoform X2, a late-embryogenesis abundant protein, was identi ed [15] . Moreover, this protein plays an important role in plant growth and development, in disease resistance, and in response to hormones and stress [16] and is involved with E3 ubiquitin-protein ligase SINAT2-like protein.
GO functional annotation and enrichment analysis GO functional annotation was carried out for the differentially expressed genes in the different samples. The differentially expressed genes were annotated to the three different GO classi cation categories: cellular components, biological processes and molecular functions. The number of differentially expressed genes between the different samples and the wild-type is annotated to the three branches. After GO functional classi cation annotation, 769 genes were annotated. Among them, 243 gene pathways were enriched in cellular components, 194 molecular function pathways, and 334 genes were annotated to biological process pathways. The enriched differentially expressed genes and their information are shown in Table 3.
In the cellular component pathway, the most abundant genes were associated with terms such as nucleus (39), cytoplasm (31), chloroplast (17), cytosol (15), plasma membrane and integral components of membrane (16), mitochondria (8), and cell wall and plasmodesmata (6). In the molecular function pathway, most (43) genes were related to binding proteins, followed by catalytic activity and molecular function, so binding and catalytic activity are the main molecular pathway processes. Among the pathways involved in biological processes, the largest number of enriched genes was associated with the response to various self-processes (40), followed by transcriptional regulatory processes (23) and then metabolic processes (12). These differentially expressed genes screened by GO enrichment analysis are involved in biological processes such as larch metabolism and response to stimuli. Moreover, these genes may be closely involved in regulating the response of larch to external stimuli and larch growth processes.

KEGG functional annotation and enrichment analysis
The annotation results of the differentially expressed genes were classi ed according to type of KEGG pathway. The results showed that the differentially expressed genes were enriched in 20 metabolic pathways (Fig. 10). The metabolic pathways were divided into ve pathway types, namely, organic systems, environmental information processing, metabolism and genetic information processing and cellular processes. Among them, the pathways enriched in metabolism were the most enriched, including 10 metabolic types, accounting for 50% of the whole enrichment pathway, these pathways mainly included those associated with carbohydrate synthesis, nucleotides, amino acid metabolism, lipid metabolism and energy metabolism. This shows that, on the basis of the premise of normal biological growth, different genes may participate in adaptations associated with plant growth and development via various metabolic pathways.
The genetic information processing pathway had the largest number of genes, including genes related to genetic information processes involving DNA transcription, protein translation, protein folding, classi cation and degradation. There are two types of environmental information processing pathways: signal transduction and membrane transport. The other two pathways have only one type, namely, environmental adaptation, transportation and catabolism. A variety of metabolic pathways involved in the growth, development and stress resistance of larch, and these metabolic pathways are coordinated by a variety of transcription factor-encoding genes. However, the speci c cooperation needs to be further studied and veri ed.

Determination of physiological and biochemical indices under different treatments
Effects of different stress treatments on the POD activity of transgenic calli POD activity re ects the ability of plants to scavenge H 2 O 2 and other reactive oxygen species. POD is an enzyme closely related to energy and respiratory metabolism, and POD activity has effects of antioxidation and delaying aging [17] . The higher the activity of POD is, the stronger the physiological metabolism and antioxidant capacity of plant tissue, which can accelerate the removal of active oxygen such as H 2 O 2 and lead to better adaptability to adverse conditions [18] .
It can be seen in Figure 11, Under PEG 6000 simulated drought stress, the POD activity of wildtype calli was very weak with increased stress duration and was always lower than that of the transgenic cell line. The transgenic cell lines showed a trend of " rst decreasing, then increasing and then decreasing again", and the change trend of the three transgenic cell lines was very consistent. When the calli were treated for 96 h, the values of OE2 and OE4 were the same as those of the wild-type calli, but the value of OE1 was higher than that of the wild-type and the other two transgenic cell lines. These results show that under drought stress, the POD enzyme activity of the transgenic cell line is stronger than that of the wild type, which means that POD can effectively remove the peroxide produced by stress, increase drought resistance.
Under NaCl stress, the POD activity of the wild-type calli was stable with increasing stress duration, but it was lower than that of transgenic cell line. The transgenic cell lines generally showed the trend of "increasing rst and then decreasing", and the value was the highest at 48 hours of stress treatment. These results indicated that 48 hours of treatment is a node, and the expression trend of the three transgenic cell lines was very consistent, indicating that the transgenic cell lines have a certain degree of salt tolerance.

Effects of different stress treatments on the MDA content of transgenic calli
It can be seen in Figure 12, MDA is the main product of membrane lipid peroxidation in plants under stress [19], and its amount can indirectly re ect the degree of oxidative damage to plants caused by stress. The higher the MDA content is, the greater the degree of oxidative damage [20] . At the same time, the protective enzyme system composed of SOD (superoxide dismutase) and POD can not only remove excess reactive oxygen species over time but also remove excessive amounts of MDA, which can reduce membrane lipid peroxidation and protect membrane structure [20] .
Under PEG 6000 simulated drought stress, the MDA content of the wild-type calli increased rst and then decreased with the prolonging of stress and reached the maximum at 48 h. The transgenic cell lines also showed a similar trend, but the MDA content reached the highest after 48 hours of stress. Moreover, the MDA content of the transgenic cell lines was lower than that of the wild-type cell lines after 24 hours and 48 hours of stress, and the accumulation degree of harmful substances was low, indicating that the degree of oxidative damage of the transgenic cell lines was less than that of the wild-type cell lines.
Under NaCl stress, the MDA content of wild-type and transgenic cell lines decreased with the extension of stress time. In general, the MDA content of the transgenic cell lines was lower than that of the wild-type cell lines, indicating that the degree of oxidative damage of the transgenic cell lines was lower than that of the wild-type cell lines.

Effects of different stress treatments on the content of superoxide dismutase (SOD) in calli
It can be seen in Figure 13, As an important antioxidant enzyme, SOD catalyzes the disproportionation of superoxide anion radicals to produce hydrogen peroxide and oxygen to remove reactive oxygen species and ultimately improve the ability of plants to resist drought stress [5] .
Under PEG 6000 simulated drought stress, the SOD content of the wild-type calli showed a trend of "increasing rst and then decreasing" with the extension of stress duration, peaking at 24 hours. The transgenic cell lines also expressed a similar trend, in which the SOD content of OE1 and OE2 reached the maximum after 48 hours of treatment, while that of OE4 reached the maximum after 24 hours of treatment. Nonetheless, both of these maximum contents were higher than the those of the wild-type cell lines, indicating that the drought resistance of the transgenic cell lines was enhanced.
Under NaCl stress, the SOD content of wild-type cell lines decreased linearly after 24 hours of stress treatment and decreased continuously until 96 hours of treatment. The SOD content of the three transgenic cell lines was higher than that of the wild-type, and the SOD content of the three transgenic cell lines was relatively high, indicating that the transgenic cell lines had greater salt stress resistance than the wild-type cell lines did.

Discussion
At present, there are few reports on the successful genetic transformation of conifers such as larch, which is mainly due to the di culty of foreign gene transformation and integration, imperfect regeneration systems and so on [21] . In this study, the embryogenic cell line of Larix olgensis was stably transformed by Agrobacterium-mediated genetic transformation, and a cell line overexpressing the LoHDZ2 gene of the HD-ZipII subfamily was obtained. Furthermore, the transgenic cells were successfully induced to develop into somatic embryos and then into seedlings.
LoHDZ2 is a gene of HD-ZipII subfamily. HD-ZipII subfamily genes are mainly involved in the light avoidance response and respond to changes in light quality [22] . Expression analyses have con rmed that some of these HD-ZipII genes are indeed regulated by auxin in wheat. Together, our results suggest that HD-ZipII subfamily transcription factors regulate plant development, possibly through the auxin pathway in plants [23]. In early research, the LoHDZ2 gene was successfully transformed into tobacco, and transgenic tobacco was obtained. After determining the phenotypic characteristics and physiological and biochemical indices of the transgenic tobacco, researchers found that transgenic tobacco exhibited plant dwar ng, leaf enlargement and early owering phenomena. It was thus preliminarily speculated that this gene may be related to plant growth and development. In the present study, transgenic and wild-type tobacco were stressed with PEG 6000 and NaCl, and their physiological and biochemical indices were measured. According to the physiological and biochemical indices, the transgenic cell lines had stronger resistance to external stress than the wild-type cell lines did. Many studies have shown that there is a close relationship between plant growth and resistance. For example, Wu Yuepeng and others used transgenic technology to import stress resistance-related genes to obtain transgenic plants; these plants exhibited improved plant stress resistance to a certain extent, but there were also problems such as stunted growth and development and dwar ng [24] .
Li Jianbo et al. studied the growth and development of poplar and Arabidopsis leaves, stems and adventitious roots and plant stress resistance by transforming poplar with Ptowusa, Ptowox4a, Ptowox5a and Ptrhsp17.8. They found that the growth of plants was affected, which were dwarf-like but exhibited increased plant resistance [25] . Perennial ryegrass with DREB1A gene (DREB) and perennial ryegrass with BADH-CMO gene (BC) showed different degrees of dwar ng. It was found that the leaf cells of transgenic plants and nontransformed plants were signi cantly shorter than those of the nontransformed plants, and the perennial ryegrass with the DREB1A gene was resistant to high temperature; thus, the DREB1A gene and the BADH-CMO gene improved the drought resistance of perennial ryegrass [26] . In response to biological stress, a gene silencing pathway, hormone-mediated signal transduction pathway and metabolic regulatory pathway usually function in plants. These response mechanisms are coordinated by hormone signals and other small molecular signals [22] .
HD-Zip genes can improve the activity of antioxidant enzymes and the accumulation of some soluble organic substances [27,28] . For example, transgenic pepper, cotton and poplar plants expressing HDG11 have high levels of proline, soluble sugars, and antioxidant enzymes (SOD and CAT) under high concentrations of NaCl, which reduces the oxidative damage of plants and is conducive to tolerance to salt stress by regulating osmotic homeostasis. Moreover, HDG11 transgenic plants show a low content of MDA. As an end-product of lipid peroxidation, the lower the MDA content is, the more helpful it is to maintain the homeostasis of membranes and proteins in vivo [29,30] . Plant growth regulatory networks are very complex and have mutual in uences. Therefore, it is preliminarily speculated that LoHDZ2 of the HD-ZipII subfamily may be involved in plant growth, metabolism and stress resistance.
Temperature stress includes high-temperature, low-temperature and severe variable-temperature stress [31] .
HD-Zips are mostly studied under high-temperature and low-temperature stress [32] . Under these two stresses, most enzyme activities are decreased, resulting in abnormal biochemical reactions and possibly cell death [33] . The expression of some HD-Zip transcription factors can be induced under both high-an low-temperature stresses; for example, some genes of cucumber and potato are upregulated under these conditions [34] . In warm (20-30 ℃), dry environments, HaHB4 induced the expression of redox-related and heat-shock protein-coding genes in transgenic soybean, indicating the potential function of HaHB4 in the heat resistance mechanism [35] . Studies have shown that heavy metals are inducers of lipid peroxidation. When plants are polluted with heavy metals, especially toxic heavy metals, these metals can disrupt the structure and function of membranes and affect plant metabolism. The content of some heavy metal elements such as iron (Fe) needs to be maintained at steady state for plants to develop normally. Athb1 negatively regulates the expression of cafer1 in response to iron by binding to MYB transcription factors, inhibiting the overexpression of iron-related genes and participating in the regulation of iron homeostasis [36,37] . The HD-Zip subfamily member HAT1 inhibits resistance against yellow mosaic virus (CMV) by inhibiting the expression of defense-related genes in Arabidopsis, such as PR1, PR2, and the ROS-related drugresistant protein glutathione transferase (GST) [3] . In an experiment involving HAT1, HAT2 and HAT3 lossof-function mutants inoculated with a virus, HAT1, HAT2 and HAT3 were found to be negative regulators of virus resistance [38] . According to various reports, in Arabidopsis, compared with other proteins, HD-Zip and proteins are more sensitive to changes in light quality. When phytochrome senses the light signal reaching the canopy, it mainly participates in the light avoidance reaction by inducing stem growth [39] . In Arabidopsis, shade avoidance is regulated by positive (PIF) and negative (HERL/SISCI) regulators of gene expression to ensure the rapid remodeling of plants to achieve the most suitable architecture for Page 13/23 growth in a particular environment [40] . According to the transcriptome sequencing analysis, it was found that the expression of the differentially expressed genes corresponding to heat stress in the transgenic cell line was upregulated, and the genes expressed in response to cadmium ion and other metal ions were also upregulated. At the same time, this line also responded to injury, bacteria, and viruses, indicating that the transgenic cell line has a certain ability to resist external pressure and stresses.         Supplementary Files