Over-expression of transcription factor ARK1 gene leads to down-regulation of lignin synthesis related genes in hybrid poplar ‘717’

Improving wood growth rate and wood quality are worthy goals in forest genetics and breeding research. The ARK1 gene is one member of the ARBORKNOX family in all plants, which play an essential role in the process of plant growth and development, but the mechanism associated with its gene network regulation is poorly investigated. In order to generate over-expression transgenic hybrid poplar, the agrobacterium-mediated transformation was used to obtain transgenic hybrid poplar ‘717’ plants to provide insight into the function of the ARK1 gene in poplar. Moreover, the morphology of transgenic plants was observed, and transcriptome analysis was performed to explore the ARK1 gene function. The results showed that there were significant differences in pitch, stem diameter, petiole length, leaf width, leaf length and seedling height between ARK1 transgenic seedlings and non-transgenic seedlings. The transgenic seedlings usually had multiple branches and slender leaves, with some leaves not being fully developed. The results of transcriptome analysis showed that the differentially expressed genes were involved in the growth of poplars, including proteins, transcription factors and protein kinases. Genes related to the positive regulation in plant hormone signal transduction pathways were up-regulated, and the genes related to lignin synthesis were down-regulated. The RT-qPCR analysis confirmed the expression levels of the genes involved in the plant hormone signal transduction pathways and phenylpropanoid pathway. In conclusion, the ARK1 gene had a positive regulatory effect on plant growth, and the gene’s coding enzymes related to lignin synthesis were down-regulated.

factors regulate plant growth and development in a variety of ways, interact with hormone pathways mediated by auxin, gibberellin (GA) and mitogen (CK) to activate signal pathways in plants, and have been shown to regulate the genes encoding GA biosynthesis directly 11,12 . GA has been reported to affect the lignification of stem cell walls 8 .
Advances in molecular biology provide new methods to study secondary growth and cambium function in forest trees. There are many economically and ecologically important species in the genus Populus, with some developed as popular models for molecular biology in angiosperm trees. The Populus genome has been sequenced and would be exceedingly useful in the study of secondary growth and cambium function. The creation of gene over-expression stable lines is widely used in protein engineering, drug discovery, gene functional analysis, and other basic researches. RNA-Seq, a deep-sequencing technology, is a useful method for transcriptome profiling. Studies using this approach have already expanded our view on the complexity of poplar transcriptomes. Yao et al. (2018) used RNA-Seq to screen differentially expressed genes (DEGs) and detect the NAC family in poplar leaves 15 . In an analysis of the overexpression of AtGolS3 and CsRFS in poplar, La Mantia et al. (2018) found that transcriptome analysis and qRT-PCR validation revealed the genetic network of the defence response to poplar leaf rust disease 16 . After transcriptome analysis of MYB165-and MYB194-overexpressing poplars, Ma et al. (2018) found that MYB165 and MYB194 were negatively related to many phenylpropanoid enzyme genes and shikimate pathway enzyme genes 17 .
This paper uses transgenic technology and high flux sequencing techniques to sequence and analyse the transcriptome of transgenic and non-transgenic hybrid poplar '717' . This technology is adopted to identify critical metabolic pathways and genes involved in poplar secondary growth, explore the effect of transcription factor ARK1 on secondary growth of woody plants and study the functions of secondary growth-related genes in poplar. Moreover, the functional annotation, functional classification and metabolic pathway enrichment of the differentially expressed genes (DEGs) were studied.

Materials and methods
Plant and bacterial materials. The explants used in the experiment were taken from the hybrid poplar '717' (INRA 717-1B4, a female P.tremula × P.alba) grown in Southwest Forestry University, Kunming, China. The young leaves and stem segments of hydroponics and root sprouts were selected as explants. Shanxi Bored Biotechnology Co. Ltd. (Shanxi, China) synthesised the ARK1 gene, which was then constructed into a binary vector, pCAMBIA 1300. Agrobacterium tumefaciens strain LBA4404 was preserved in our laboratory.
Transformation. The leaves of hybrid poplar '717' infected by A. tumefaciens were inoculated on a callus induction medium (MS + 1.0 mg/L NAA + 1.0 mg/L 6-BA) and co-cultured at 28 °C for 2-4 days. The co-cultured calli were washed with aseptic water three times, dried with aseptic paper, and then transferred to an aseptic differentiation medium (MS + 1.0 mg/L 6-BA + 0.4 mg/L ZT) containing carbenicillin and kanamycin. The selective culture was carried out under 8 to 16-hour photoperiod conditions at 28 °C. Approximately 28 days later, the medium was changed to induce new calli to form and sprout. When the adventitious buds grew to 2 to 3 cm, they were moved to a rooting medium (1/2 MS + 0.02 mg/L NAA + 0.6 mg/L IBA) containing carbenicillin and kanamycin for root culture. When the adventitious roots grew to 2 to 3 cm, the plants were moved to a greenhouse.
Identification of transformants. PCR analysis. The DNA was extracted with an HiPure SF plant DNA mini kit (Magen Company, New York, USA) from transgenic and non-transgenic seedlings. The sequences of the primers were 5′-AAGATCCAGCCCTTGACCAA-3′ and 5′-CATTGCCATCACCACAACCA-3′. Then the PCR reaction was carried out in a GeneAmp RCR System 9600 (Perkin Elmer, Foster City, CA, USA) under the PCR conditions of 94 °C for 3 min; 94 °C for 30 sec, 55 °C for 30 sec, 72 °C for 5 min, 35 cycles; 72 °C for 10 min.
Measurement of morphological changes in transgenic plants. The morphological differences of tissue-cultured seedlings between three different transgenic lines with a specific PCR band and three non-transgenic plants that underwent the same conditions were compared. The diameter of stem segments, the number of internodes and the length of internodes at the same growth stage were measured. Transcriptome analysis. The experimental materials were taken from the stem segments under the fifth leaf to the sixth leaf of three transgenic lines and three non-transgenic seedlings. The sampling time was at 11: 00 am.
The purified samples were sequenced using the HiSeq high-throughput sequencing platform by Shanxi Bored Biotechnology Co., Ltd. (Shanxi Province, China). The genome of hybrid poplar '717' was used as the reference genome, and the download address was http://aspendb.uga.edu/index.php/databases/spta-717-genome.
DESeq was used to analyse the differential expression among different groups 18 . The DEGs between the two biological conditions were obtained, and the DEGs were classified. Then the phyper function of R software was used for enrichment analysis. Those DEGs with fold change ≥2 and FDR < 0.01 were regarded as significant enrichment.
Additionally, a DEG pathway annotation analysis was used to analyse the functions of the genes further. Those DEGs with fold change ≥2 and FDR < 0.01 were regarded as significant.
RT-qPCR analysis. This study selected the transcriptomic expression levels of ten genes including the ARK1 gene and nine genes involved in the plant hormone signal transduction pathways and the phenylpropanoid pathway for validation by RT-qPCR analysis in three transgenic lines and three non-transgenic plants used in the transcriptome analysis. The elongation factor gene EF1β was used as an internal control 19 . The total RNA was extracted using the Qiagen RNeasy Mini Kit (Qiagen Inc., Valencia, CA), and then reversely transcribed into cDNA by random primers. The RT-qPCR analysis was conducted according to a previous report 20 . Gene-specific Comparison of the growth between ARK1 transgenic and non-transgenic seedlings. The transgenic and non-transgenic seedlings were weighted at the same growth stage (45 days). Compared to non-transgenic plants, transgenic plants had slender stem segments (usually fasciculated and multi-branched), slender leaves and undeveloped leaves (Fig. 1). The appearance of ARK1 transgenic seedlings was consistent with the ones obtained by Groover et al. 21 . There were significant differences between transgenic and non-transgenic seedlings in node space, stem segment diameter and leaf width. The internode distances, stem segment diameters, and width and length of leaves were measured in the fifth and sixth leaves. There were significant differences in node spacing, stem diameter, petiole length, leaf width, leaf length and seedling height between transgenic and non-transgenic seedlings ( Table 2).
Analysis of transcriptome data. After the original data were filtered by quality control, the redundant sequences and low-quality reads were removed, and a total of 45.8 GB clean reads were obtained. The percentage of Q30 bases was higher than 91.85%, and the average GC content of the six samples was 47.69%, indicating that the sequence quality was good and met the requirements of database construction. Sequences were aligned between the clean reads and the reference genome of P. tomentosa, and alignment efficiency varied from 55.61% to 60.61%. Reference genomes could annotate approximately 57.53% of the sequences. The clean reads alignment rate of the reference sequence was 56.61%.
The total mapping ratio of the two groups compared with the previous reference genome was 58.81%. The lowest was 55.61%, and the highest was 60.61%. The average clean reads ratio of the two groups at a specific position of the reference genome was 74.11%, and the unique alignment between the groups was uniform, with the lowest at 73.46% and the highest at 74.86%. The comparison results showed that the comparison efficiency between the reads of each group and the reference genome ranged from 55.61% to 60.61%, and the selected reference genome met the requirement for analysis. A total of 641 DEGs were identified, of which 389 were up-regulated, and 252 were down-regulated. www.nature.com/scientificreports www.nature.com/scientificreports/ Through Gene Ontology (GO) enrichment analysis of 641 common differentially expressed genes in transgenic and non-transgenic plantlets (Fig. 2), a total of 496 DEGs were obtained in the enrichment entry of GO, and 428 DEGs belonged to the biological process, of which 237 were up-regulated, and 191 were down-regulated. Three hundred and eight DEGs belonged to the cell composition, of which 201 were up-regulated, and 107 were down-regulated. Moreover, 426 DEGs belonged to molecular function, of which 236 were up-regulated, and 190 were down-regulated. In the biological process category, there were 31 DEGs involved in protein phosphorylation, followed by 25 DEGs in the cellular metabolic process. Nineteen DEGs belonged to cell proliferation. Regarding cell locations, most DEGs were in the nucleus with 58 DEGs, followed by an integral component of the membrane, and the whole component of the membrane. There were 49 DEGs in the plasma membrane. In the molecular function category, most DEGs were ATP binding, followed by DNA binding and transcription factor activity. Furthermore, 24 DEGs belonged to sequence-specific deoxyribonucleic-acid-binding, including transcription factor activity, sequence-specific DNA binding and microtubule-binding.

Functional analysis of differentially expressed genes. Next, pathway enrichment analysis of 641
DEGs was carried out using the KEGG database. Metabolic pathways annotated were the organic system, environmental information processing, cell processing, metabolic and genetic information processing (Fig. 3). In this enrichment process, the metabolic pathway was the most significantly enriched pathway with 112 DEGs, followed by genetic information processing with 21 DEGs.
Results of annotation comparison showed that 160 significantly different genes in the hybrid poplar '717' were annotated and enriched into 57 metabolic pathways. Thirteen differentially expressed genes were expressed in the metabolic pathway (KO: ko01100) pathway, and 11 DEGs were enriched in the plant hormone signal transduction (KO: ko04075) pathway. There were ten DEGs in the DNA replication (KO: ko03030) pathway and nine DEGs in the starch and sucrose metabolism (KO: ko00500) pathway. There were six DEGs in the amino sugar and nucleotides glucose metabolism (KO: ko00520) pathway and six DEGs in the purine metabolism (KO: ko00230) pathway.

Screening and analysis of growth-related DeGs. GO enrichment analysis and pathway functional
annotation was applied to screen the expression and regulation of DEGs related to the secondary growth. The cell tip growth pathway (GO:0009932), meristem growth pathway (GO:0010075), plant hormone signal transduction (Ko04075) and phenylpropanoid biosynthesis (Ko00940) were selected for analysis. There were six DEGs related   Table 2. Growth data related to transgenic and non-transgenic seedlings. **Indicated that the difference was extremely significant, and * indicated that the difference was significant (*P-value < 0.05, **P-value < 0.01).
The genes involved in the plant hormone signal transduction pathways, including XM_024591467.1 (EIN4-like protein), XM_024610415.1 (ethylene receptor), XM_024607193.1 (ethylene response transcription factor 1B), XM_002311147.3 (phosphatase 2 C) and XM_024581841.1 (phosphate protease 2 C 51), were down-regulated. EIN4-like protein and ethylene receptors play a role in two-component system-based signal transductions in plant growth and development, acting as negative regulators of ethylene signal transduction in plant embryos, etiolated seedlings, leaves, roots, inflorescences and stamens. EIN4-like protein was expressed in pollen and tapetum and moderately expressed in carpels 23 . Ethylene response transcription factor 1B was reported to play an important part in plant growth and development and organ formation 24 . Phosphatase 2C 37 and phosphate protease 2C 51 both catalyse the dephosphorylation of phosphate serine and threonine phosphate residues of specific protein substrates, which can regulate the reversible phosphorylation of proteins in a variety of signal transduction pathways. Phosphatase plays a major role in plant growth and development and is mainly involved in the development of plant ears 25,26 . PYR1 abscisic acid receptor proteins are receptors involved in signal transduction. They bind to abscisic acid (ABA) and mediate its signal transduction. After binding to ABA, these proteins interact with 2C protein phosphatase and inhibit its activity 27 .
There were both positive and negative regulatory genes among these growth-related genes screened. Up-regulated DEGs were mainly concerned with positive regulation, while the down-regulated DEGs were mainly concerned with negative regulation in the transgenic seedlings. These results indicated that the ARK1 gene had a positive regulatory effect on plant growth. www.nature.com/scientificreports www.nature.com/scientificreports/ The genes with increased expression in transgenic plants in the phenylpropanoid biosynthesis pathway (Fig. 4b) included XM_002304914.3 (β-glucosidase, 5.05050992570199) and XM_024598534.1 (β-glucosidase, 3.0702239239688). β-glucosidase, a glycosyl hydrolase, participates in carbohydrate transport and metabolism, including plant morphogenesis and energy metabolism, and plays an essential role in plant development 28,29 . The genes with decreased expression included XM_024590232.1 (trans-cinnamic acid-hydroxylase, -2.08868270819994), XM_002310515.3 (peroxidase 17, -1.3784859676881) and XM_024605163.1 (peroxidase 73, -2.0529527985568). Trans-cinnamic acid-hydroxylase was found to catalyse the formation of P-coumaric www.nature.com/scientificreports www.nature.com/scientificreports/ acid and p-coumaroyl-CoA (Fig. 5). Moreover, the phenylpropanoid pathway was reported to provide a variety of secondary metabolites in plants, participate in plant tissue differentiation and protect plant tissue from environmental stress 30,31 . Peroxidase belongs to class III of the heme-dependent peroxidase superfamily in plants. All members of the superfamily shared heme-repair groups and catalysed multistep oxidation involving hydrogen peroxide as electron receptors 32 . Peroxidases catalyse the removal of H 2 O 2 and are involved in the oxidation of toxic reductants, lignin biosynthesis and degradation of thrombus, catabolism of auxin and responses to environmental stresses, such as injury, pathogen attack and oxidative stress 33 . The down-regulated genes -peroxidase 17 and peroxidase 73 -were found to catalyse the formation of p-hydroxyphenyl lignin, guaiacyl lignin, 5-hydroxyguaiacyl lignin and lilac lignin (Fig. 5).
RT-qPCR validation. The RT-qPCR analysis results showed that the expression levels of all the selected genes were consistent with the transcriptomic analysis results (Fig. 5). The findings confirmed the increased expression levels of the positive regulatory genes in the plant hormone signal transduction pathways. Additionally, results showed that the genes coding enzymes related to lignin synthesis in the phenylpropanoid pathway were down-regulated in transgenic plantlets compared to non-transgenic plantlets.

Discussion
Foreign gene transformation mediated by A. tumefaciens is the result of the interaction between bacteria and plant cells, which can usually affect the infection ability of A. tumefaciens. All the factors that can affect the infection ability of A. tumefaciens, the ability of plant cell transformation response and the ability of transformant regeneration will affect the transformation effect. Confalonieri et al. (2003) found that the transformation rate of poplar × P. tomentosa backcross hybrid was 1.22% and 2.59% respectively, which was easier to transform than that of the P. tomentosa male plant -the transformation rate of 1,319 male plants was only 0.34% 34 . This experiment obtained 35 candidates of transformed seedlings of hybrid poplar '717' . Overall, a total of 11 seedlings were positive in PCR detection, and the transformation rate was 31.43%.
In this study, secondary growth-related gene ARK1 was transformed into hybrid poplar '717' mediated by A. tumefaciens. There were significant differences in node spacing, stem diameter, petiole length, leaf width, leaf length and seedling height between ARK1 transgenic and non-transgenic seedlings. Similar to this study, it was reported that the hybrid poplar with ARK1 over-expression grew vigorously, and the branching ability was very strong, which was usually characterised by multiple branches in a single node 12 . Chuck et al. found that the genes in the KNOX gene family in Arabidopsis thaliana were expressed in stem apical meristem rather than in mature organs 35 . Over-expression of KNOTTED1 (KN1) gene in transformed seedlings could change normal leaves into lobed leaves, jag from the base of the leaves, and display leaves that do not fully develop or expand or are without slender petioles. In this study, leaves that were not fully developed or expanded also appeared in the ARK1 transgenic poplar lines, and the specific function of secondary growth-related genes could be changed through the abnormal expression of transcription factors (ARK1). ARK1 (ARBORKNOX1) and ARK2 (ARBORKNOX2) genes are poplar homologous genes of Arabidopsis STM and BP (BREVIPEDICELLUS), which play an essential role in regulating cambium cell differentiation 11,21 . ARK1 was widely expressed in the apical meristem (SAM) and vascular cambium region, and down-regulated in the terminal differentiation cells of the leaves and secondary vascular tissues of the apical meristem. Groover et al. (2006) were the first to clone the homologous gene ARK1 of Arabidopsis STM in Populus tomentosa 21 and found that ARK1 was mainly expressed in the cambium. Over-expression of ARK1 or STM was reported to inhibit the differentiation of xylem and phloem fibres, inhibit leaf development and shorten the length of internodes 21 . Combined with the expression analysis of ARK1 in the process of adventitious bud and adventitious root formation, it was found that ARK1 was mainly involved in primordium formation and further differentiation of www.nature.com/scientificreports www.nature.com/scientificreports/ meristem cells in the late primordium 36 . ARK1 was also found to play a role in the differentiation of different meristem in the stem tip, root tip and formation layer 36 . The result of secondary growth is the division of coordination cells in the meristem region of xylem and the differentiation of progeny cells in endodermis and wood tissue 37 . The results of microarray analysis showed that there was a good correlation between the transcriptional level of genes and the function in cell division and differentiation at specific stages of wood development 12,38 . It indicated that ARK1 was vital to regulating poplar growth. The current study selected 27 DEGs involved in poplar growth by analysing the DEGs in the pathways of plant hormone signal transduction (Ko04075), phenylpropanoid biosynthesis (Ko00940), cell tip growth (GO:0009932) and regulation of meristem growth (GO:0010075). These genes included coding proteins, transcription factors and protein kinases, which were related to plant growth and development and lignin regulation. It illustrated that ARK1 was also crucial for regulating poplar growth.
This study found that the enzymes related to lignin synthesis in the phenylpropanoid pathway were down-regulated. Lignin is one of the three main chemical components of wood (lignin, cellulose and hemicellulose) and has essential biological functions in plants 39,40 . Lignin limits the development of the paper industry due to environmental pollution and the need for a large amount of energy for wood production in the process of paper-making. The reduction of lignin content of trees can not only improve the economic and environmental benefits of the pulp and paper industry but also promote the decomposition of lignocellulose and improve the conversion efficiency of sugar 41 . The changes of lignin content and composition had no adverse effect on the growth of transgenic plants but increased the biomass of transgenic plants, such as stem diameter, plant height and internode length. Zhou et al. (2018) found that when the lignin content of C3H and HCT transgenic hybrid poplars (P. alba × P. glandulosa '84 K') decreased, the phenotype of plants showed abnormal growth in height, diameter, and so on 42 . Su et al. (2019) found that when the content of lignin reduced, the content of lignin deposited in the cell wall decreased, which easily led to the abnormal phenotype of tissue culture seedling 43 . This study proposes that the abnormal phenotype of ARK1 transgenic poplar in node spacing, stem diameter, petiole length, leaf width, leaf length and seedling height is due to the down-regulation of the enzymes related to lignin synthesis.
Chromatin immunoprecipitation sequencing (ChIP-seq) technology was employed to identify ARK1 binding loci. Findings showed that ARK1 is a vital transcription factor of the vascular cambium and cell differentiation regulation in Populus 12 . This study also found that ARK1 is a key regulator of cell differentiation in Populus. However, Liu et al. (2015) did not report a relationship between the expression of ARK1 and the expression of enzymes related to lignin synthesis 12 . Our results also showed that when the expression level of genes related to the lignin content of plants reduced, the plants would grow abnormally in terms of height and diameter. Groover et al. (2006) reported a similar phenotype of ARK1 transgenic poplar after analysing transcriptome data (using microarray). However, more cell-wall associated GO terms were found in Groover and colleagues' study. The current study found lignin biosynthesis genes to be mostly down-regulated in ARK1-overexpressing lines, whereas Groover et al. (2006) found 35S::ARK1 trees to have increased lignin, which is paradoxical. Of course, the reduction in lignin gene expression was only based on two genes; thus, this finding requires further support via future studies.

conclusions
The ARK1 was transformed into hybrid poplar '717' . PCR detection showed that the positive rate was 31.43%. There were significant differences in node spacing, stem diameter, petiole length, leaf width, leaf length and seedling height between transgenic and non-transgenic seedlings. The stem segments of transgenic '717' hybrid poplar seedlings were slender, fasciculated and multi-branched. The leaves were slender, and some leaves were not fully developed.
Twenty-seven DEGs involved in poplar growth and development were screened out, including proteins, transcription factors and protein kinases. The up-regulated DEGs were mainly positive regulatory genes, while the down-regulated DEGs were mainly negative regulatory genes. The ARK1 gene had a positive regulatory effect on plant growth, and the gene's coding enzymes related to lignin synthesis were down-regulated.

Data availability
RNA-seq data were presented at the Genome Sequence Archive of National Genomics Data Center, Beijing Institute of Genomics (accession number CRA002209).