Introduction

Gibberellins (GAs) are plant hormones that regulate the growth and development of plants, including seed germination, stem elongation, leaf expansion, flower induction and development, and fruit set. GA is a large group of tetracyclic diterpenoid carboxylic acid synthesized from 2-oxoglutarate-dependent dioxygenases. The GA20-oxidases (GA20ox) and GA3-oxidases (GA3ox) mediate the last stage of GA biosynthesis by precursor GA12, which controls the levels of bioactive GAs (GA1, GA3, GA4, etc.). Bioactive GAs levels can be deactivated by GA2-oxidases (GA2ox) (Hedden and Thomas 2012; Salazar-Cerezo et al. 2018). JcGA2ox transgenic plants led to lower GAs levels and displayed a typical dwarf phenotype in Arabidopsis(Hu et al. 2017). In pea, the PsGA3ox1 overexpression plants had higher GAs concentrations, while altering GA biosynthesis and catabolism gene expression and plant phenotype(Reinecke et al. 2013). The GA deficient ga1-3 mutant and AtGA2ox1 overexpression plants had a high expression level of DELLA protein(Middleton et al. 2012). The expression of GA20ox, GA3ox and GA2ox are maintained by a feedback regulation to respond to levels of bioactive GAs, which through development and environmental stimuli(Fukazawa et al. 2017).

The DELLA protein subfamily belongs to a large GRAS (named after GIBBERELLIN INSENSITIVE [GAI], REPRESSOR OF gal-3 [RGA], and SCARECROW [SCR]) transcription factor family with repressor activity(Daviere and Achard 2013; Hauvermale et al. 2012). Arabidopsis has five DELLA proteins: GAI (GAIN SENSITIVE), RGA (REPRESSOR-OF-GA), RGL1 (RGA-LIKE1), RGL2 (RGA-LIKE2), and RGL3 (RGA-LIKE3). Previous studies demonstrated that GAI, RGL1 and RGA play an essential role in regulating hypocotyl growth, stem and root elongation(Zheng et al. 2018; Dill and Sun 2001). PmRGL2 in poplar was a negative regulator of GA responses that regulated bud dormancy and plant height(Lv et al. 2018). In oilseed rape, DELLA protein also negatively regulated plant height through GA(Zhao et al. 2017; Liu et al. 2010). In tomato, SlGAI could mediate the control of leaf form(Jasinski et al. 2008), SlDELLA operated as a growth repressor during fruit development by removing the response to GA and repressing cell expansion(Marti et al. 2007). Therefore, DELLA protein has the potential to actively coordinate plant growth and development.

DELLA protein is a key transcription factor in the GA signaling pathway (Daviere et al. 2008), which acts as a repressor of plant growth(Dill et al. 2001; Silverstone et al. 1997). The GA-GID1 (Gibberellin Insensitive Dwarf)-DELLA complexes are polyubiquitinated and degraded by 26 S proteasome(Dill et al. 2004; Ueguchi-Tanaka et al. 2007b; Murase et al. 2008; Ariizumi et al. 2011; Achard and Genschik 2009). It has been shown that DELLAs can regulate the expression of GA biosynthesis and catabolism genes. AtGA20ox2 and AtGA3ox1 were identified as targets of DELLA, which activate the expression of these genes(Zentella,2007). SlDELLA activated the expression of GA biosynthesis genes GA20ox1 and GA3ox1 in tomato(Hu et al. 2018). The DELLA-GAF1 (GAI-ASSOCIATED FACTOR1) complex was the main component in GA feedback regulation of GA20ox2 and GA3ox1(Wen et al. 2021).

Up to date, it is little known about the DELLA gene involving tuber morphogenesis in tuber crops. In potato, GA modulated host defense in response to phytoplasma infection via DELLA and SA signaling pathways(Ding et al. 2013). In cassava, MeDELLA silenced plants decreased the resistance to cassava bacterial blight, and MeDELLA and MeGAI were involved in GA signal resistance to drought stress (Li et al. 2018b). The involvement of GA3 in yam tuber enlargement and yield was described(Kim et al. 2003; Yoshida et al. 2008). In our previous reports, high-level endogenous GA3 had been identified in the yam early tuber expansion stage, and the application of GA3 promoted tuber expansion(Gong et al. 2015, 2016). This implies that GA3 affects the development of yam tuber, and other components of the GA signaling cascade may regulate tuber development.

In our previous study, exogenous GA3 increased tuber weight, causing a significant in Guihuai 16 and Guihuai 5 tuber yield(Gong et al. 2015), and we isolated the DELLA protein gene for yam tuber and functionally characterized it in phylogenetic analysis(Zhou et al. 2021). Expression patterns were shown the transcripts of DoDELLA1 were in response to GA3 treatment in leaf, stem and tuber. In this study, a transgenic method was used to characterize DELLA protein function by overexpressing DoDELLA1. The possible mechanism underlying the regulation of DoDELLA1 in plant growth and development was studied.

Materials and methods

Field experiment and sampling

The yam cultivars Guihuai 16 (GH16, with long and small tuber) and Guihuai 5 (GH5, with large tuber) were planted at the farm of Guangxi University (Fig.S1). Its healthy tubers’ germination and planting patterns were as previously described(Gong et al. 2015). Young leaves and stems were defined as those borne on the first, second, third and fourth nodes (from the shoot tip), while mature leaves and stems were sampled from the 15th node after field planting for approximately 40d. Tuber samples were collected at tuber initiation, expansion and mature stage after field planting for approximately 40d, 90d, and 150d, respectively. Exogenous GA3 and GA biosynthesis inhibitor paclobutrazol (PP333) at 200 mg/L treatments were done as previously described(Gong et al. 2016), and summarized as below. The treatments were sprayed on the leaves of GH16 on June 20, and GH5 on July 20. Water-treated served as the control. All treatments were replicated three times, using 60 plants as a plot and arranged in a randomized complete block design. Mature leaves, mature stems and tubers were sampled 30d after treatment. Twelve plants were selected randomly from every repetition each time. All samples were immediately frozen in liquid nitrogen, and stored at -80℃. All samples were collected before 09:00 to minimize circadian effects on gene expression during 2019.

Yeast two-hybrid assay

The yeast transactivation assay was performed using Yeast and Transformants System(Hernandez-Garcia et al. 2019). The open reading frame and different fragments of DoDELLA1 sequences were amplified and fused into the NdeI-BamHI sites of the pGBKT7 bait vector (GAL4 binding-domain, DBD). Primer pairs used were listed in Table S1. pGBKT7-53 and pGADT7-T vector acted as a positive control, pGBKT7-Lam and pGADT7-T vector acted as a negative control. Vectors (100ng) were introduced into Y2H gold yeast strains and cultured on SD medium without Leu containing AbA and X-α-gal at 30°C for 3d.

Plant materials, plasmid constructions, plant transformation and growth conditions

Tobacco plants (Nicotiana tabacum) were grown in tissue culture under a day photoperiod (14 h light and 8 h darkness) on half-strength Murashige and Skoog’s (MS) medium containing 3%(w/v) sucrose. pBI121-DoDELLA1, which is a fragment cloned version of the yam DELLA1 gene, together with a cauliflower mosaic virus(CaMV) 35 S promoter and a nopaline synthase gene (NOS) terminator. Primer pairs were listed in Table S1. Tobacco leaf discs were infected with Agrobacterium tumefaciens strain AH105 harboring pBI121-DoDELLA1. The transformed plants were selected on MS medium containing 300 mg L− 1 kanamycin and 500 mg L− 1 claforan, and transferred to fresh medium monthly. Transgenic plants were transferred to pots filled containing potting soil with nutrients and moved to an environmental chamber (24℃C, 16 h light/8 h dark). Independent transgenic seeds were germinated on MS medium containing 100 mg L− 1 kanamycin and were cultured at the chamber(24℃C, 16 h light/8 h dark). After a two-week selection, kanamycin-resistant plantlets were transferred to a fresh medium, followed by one month of culture under the same conditions. The plants were transferred to the soil under the same conditions for the isolation of homozygous line. T1 plants were self-pollinated to obtain the T2 generation. T2 seedlings were tested by kanamycin segregation, and a 3:1 segregation ratio was obtained for T3 generation. Homozygous lines were detected by kanamycin segregation and were determined by qRT-PCR (primer listed in Table S2). A total of six independent transgenic T3 lines were obtained (Fig.S2).

Quantitative real-time PCR analyses

Total RNA was extracted using the MiniBEST reagent (TaKaRa, Dalian, China), and reverse transcribed into cDNA using the PrimeScript RT reagent Kit (TaKaRa, Dalian, China). The expression levels of DoDELLA1, NtGA20ox1-1, NtGA20ox1-3, NtGA3ox4-1, NtGA2ox1-3, NtGA2ox1-8 and NtGA2ox2-3 were measured by quantitative real-time PCR using SYBR Green Supermix on the CFX96 Real-Time PCR Detection System (BIO-RAD), according to the manufacturer’s protocol (primer listed in Table S2). All reactions were performed in triplicate. The 2 −ΔΔCt method was used to estimate relative expression level, and DoActin was used as internal controls for the yam tuber sample, NtActin was used as internal controls for different tobacco plants. Young leaves and stems were defined as those borne on the first, second, third and fourth nodes (from the shoot tip), while mature leaves and stems were sampled from the 15th node after field planting for approximately 20d as a control sample, respectively. The tuber control sample was collected after field planting for approximately 30d. The tobacco control sample was collected after germination.

Transgenic tobacco phenotype analyses

The wild-type and the third generations of transgenic tobacco plants were grown on agar medium, transplanted to the soil after germination. Samples were collected at 30d, 60d, and 120d after transplanting to the soil. Exogenous GA3 and PP333 at 200 mg/L were sprayed at 30d, respectively. Each tobacco plant was used to measure plant height, internode length, leaf length and width, flower and fruit. Phenotypic measurements of tobacco plants were undertaken using three independent lines (three plants from each line). All phenotype analyses were collected from 30 plants, using Student’s t-test. Mature, fully expanded functional leaves of wild-type and transgenic tobacco were collected during the growth period for the measurement of hormone levels. All samples were collected before 09:00 to minimize circadian effects on gene expression and hormone contents. The extraction method and quantitation of endogenous hormone were carried out as previously described(Gong et al. 2016). The extracted and purified samples were subjected to high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) analysis.

Results

Expression patterns of DoDELLA1 in yam

In a previous study, GH5 and GH16 are known for reproducible individual differences in the response of stem and tuber to GA3 (Fig.S1), and we cloned the full-length cDNA sequence of DoDELLA1 (MV377793). DoDELLA1 was redundantly expressed in all tissues of GH5 and GH16. DoDELLA1 had a higher expression level in GH5 than in GH16 except in mature stems and initiation stage tuber without significant difference. DoDELLA1 had a higher expression level in mature leaves, young stems and expansion stage tuber, but had lower expression in mature stems and mature stage tuber in GH5. In GH16, the DoDELLA1 expression level had the highest in young stems compared to other tissues, and the lowest expression in mature stage tuber. Interestingly, expression levels of DoDELLA1 increased in the expansion stage and gradually declined with tuber development in GH5, but decreased gradually in GH16 with tuber development (Fig. 1 A).

Fig. 1
figure 1

Expression levels and transcriptional activation analysis of DoDELLA1 in two yam cultivars. (A) Expression levels of DoDELLA1. (B) Effects of GA3 and PP333 treatments on expression level of DoDELLA1. Transcripts accumulation was determined using qRT-PCR on three biological and three technical replicates. Letters indicate significantly different (Student’s t-test; P < / > 0.05). Le, leaf; St, stem; Tu, tuber; Y, young; M, mature; IS, initiation stage; ES, expansion stage; MS, mature stage. (C) Schematic representation of the constructs used for transcriptional activation assay in yeast. (D) Transcriptional activation assay in yeast. PC: positive control, NC: negative control

Full size image

GA3 treatment down regulated DoDELLA1 expression levels in leaf, stem and tuber of two cultivars, while PP333 treatment enhanced DoDELLA1 expression levels except for stem. However, the degree of change in response to GA3 and PP333 of leaf and tuber was higher than that of stem in two cultivars (Fig. 1B).

Transcriptional activation of DoDELLA1 N-terminal domain

We examined the transcriptional activation of DoDELLA1 using yeast expression system. Yeast strain Y2H Gold was transformed with a plasmid carrying the DNA-binding domain of the yeast GAL4 transcription factor (DBDGAL4) fused with different fragments of DoDELLA1 protein (Fig. 1 C). Yeast cells carrying full-length DBDGAL4-DoDELLA1 (plasmid D) grew blue in the SDO/X/A medium. To further analyze which conserved domain of DoDELLA1 is responsible for the transcriptional activation, we segmented the DoDELLA1 further into 231-634aa (plasmid d1), 285-634aa (plasmid d2), 114-634aa (plasmid d3), 35-284aa (plasmid d4) and 35-113aa (plasmid d5). Yeast cells carrying plasmid d3, plasmid d4 and plasmid d5 grew blue in SDO/X/A medium (Fig. 1D), suggesting that DELLA and VHYNP domain are involved in transcriptional activity, LH, VHIID and SAW domain are not involved in transcriptional activity.

The changes of characteristics of phenotype, physiology and gene expression in DoDELLA1 transgenic tobacco line

DoDELLA1 transgenic tobacco lines were used to evaluate the role of DoDELLA1. The morphological characteristics of independent transgenic lines were obtained and DoDELLA1 expression levels were detected in each transgenic tobacco line (Fig. 2 and Fig.S3). Compared to wild-type tobacco plants, over-expression DoDELLA1 transgenic tobacco plants showed obvious dwarfing and late flowering. DoDELLA1-2, DoDELLA1-3 and DoDELLA1-4 lines showed more serious dwarf and no flowering. The expression levels of DoDELLA1 in DoDELLA1-2, DoDELLA1-3 and DoDELLA1-4 lines were higher than other lines, suggesting that the dwarfing and no flowering phenotype of transgenic plants was related to the expression level of DoDELLA1. Since DoDELLA1-2, DoDELLA1-3 and DoDELLA1-4 did not blossom and could not harvest seeds, we selected the DoDELLA1-1 line in T3 plants for further research in this experiment.

Fig. 2
figure 2

Plant height (A) and DoDELLA1 expression (B) in transgenic tobacco lines and wild-type (WT) tobacco. Plant height measurements are the means (± SE) of 30 seedlings. Transcripts accumulation was determined using qRT-PCR on three biological and three technical replicates. Letters indicate significantly different (Student’s t-test; P < / > 0.05)

Full size image

The DoDELLA1-1 transgenic plants exhibited significant dwarfism, compact leaf type, and late flowering after 120d of growth (Fig. 3 A). The height of DoDELLA1-1 transgenic plants were 44.22% shorter than that in wild-type plants (Table 1). Besides, the number of internode and leaf in DoDELLA1-1 transgenic plants was 30.08% and 43.66% lower than wild-type plants, respectively. The weight of fruit was 33.33% lighter than that of wild-type plants. It shows that over-expression of DoDELLA1-1 in tobacco inhibits plant growth.

Fig. 3
figure 3

Phenotypes and hormone contents in wild-type (WT) and DoDELLA1-1 tobacco line. A. Phenotypes of two tobacco lines at 120-day-old, bar = 10 cm. B-F. Hormone contents (GA1, GA3, ZT, IAA, ABA) of two tobacco lines in different developmental stages. Hormone accumulation was determined using HPLC-MS/MS on three biological and three technical replicates. Letters indicate significantly different (Student’s t-test; P < / > 0.05)

Full size image
Table 1 Vegetative and reproductive growth characteristics of wild-type (WT) and DoDELLA1-1 tobacco lines under GA3 and PP333 treatment
Full size table

In the wild-type tobacco plants, the content of endogenous GA1 and GA3 showed two obvious peaks at 30d and 90d, respectively. GA1 and GA3 contents were the highest at 30d, which were 30.18 ng/g and 153.79 ng/g, respectively (Fig. 3B, C). ZT (Zeatin) content was the highest at 30d, which was 0.75ng/g, and gradually decreased with growth (Fig. 3D). IAA (Auxin) content gradually increased with growth, reaching the highest level of 1.34 ng/g at 120d (Fig. 3E). ABA (Abscisic acid) content increased first and then decreased with the growth, reaching the highest at 60d (Fig. 3 F).

In the DoDELLA1-1 transgenic tobacco plants, GA1, GA3 and ZT contents were the highest at 30d, which were 4.56 ng/g, 35.27 ng/g and 1.76 ng/g, respectively, and gradually decreased with growth (Fig. 3B, C). IAA content increased first and then decreased with the growth, reaching the highest at 90d, which was 0.95 ng/g (Fig. 3E). ABA content changed little at 30d, 60d and 90d, but decreased at 120d (Fig. 3 F). Compared to wild-type tobacco plants, GA1 and GA3 contents were significantly lower in the DoDELLA1-1 transgenic tobacco plants over 30-120d (Fig. 3B, C), IAA contents were significantly lower at 120d (Fig. 3E), ABA contents were significantly lower at 60d and 90d, and ZT contents were significantly higher at 30d (Fig. 3 F).

The expression of some tobacco genes involved in GA metabolism were assessed (Fig. 4). The expressions of NtGA20ox1-1, NtGA20ox1-3 and NtGA3ox4-1 were exhibited similar expression patterns in wild-type and DoDELLA1-1 transgenic tobacco plants, it decreased first and then increased along with the growth. NtGA2ox1-3 expression increased in both tobacco lines. NtGA2ox1-8 and NtGA2ox2-3 expression tended to be stable in wild-type lines. NtGA2ox1-8 expression decreased first and then increased in DoDELLA1-1 transgenic lines, while NtGA2ox2-3 expression followed a down-up-down pattern. The expressions of NtGA20ox1-1, NtGA20ox1-3 and NtGA3ox4-1 were lower in DoDELLA1-1 transgenic lines than in wild-type lines. While NtGA2ox1-3, NtGA2ox1-8 and NtGA2ox2-3 enhanced the expression levels in DoDELLA1-1 transgenic lines compared to wild-type lines. These results suggest that the over-expression of DoDELLA1-1 decreased the levels of GA1 and GA3 in transgenic tobacco plants by down regulating the expressions of NtGA20ox1-1, NtGA20ox1-3, NtGA3ox4-1, and up regulating the expressions of NtGA2ox1-3, NtGA2ox1-8 and NtGA2ox2-3. It shows that DoDELLA1-1 can regulate the expressions of GA synthesis-related genes.

Fig. 4
figure 4

The expression levels of tobacco GA-metabolism and signaling-related genes in wild-type (WT) and DoDELLA1-1 tobacco lines in different developmental stages. Transcripts accumulation was determined using qRT-PCR on three biological and three technical replicates. Letters indicate significantly different (Student’s t-test; P < / > 0.05)

Full size image

Effect of exogenous GA 3 and PP 333 on the characteristics of phenotype, physiology and gene expression in DoDELLA1 -1 transgenic tobacco lines

The DoDELLA1-1 transgenic and wild-type tobacco plant were sprayed 200 mg/L exogenous GA3 and PP333 at 30d after transplanting to the greenhouse (Table 1; Fig. 5 A). Compared to wild-type plants without treatment, wild-type plants exhibited slender growth associated with slender internode, stem and leaf, while plant height was enhanced by 17.16%, and delayed flowering by GA3 treatment. Wild-type plants exhibited compact growth concomitant with a severe decrease in the height, stem diameter, internode length and number, leaf length, width and number, fruit weight and number, flower number by PP333 treatment. GA3 treatment rescued to a certain extent DoDELLA1-1 transgenic plant height mainly due to extending internode length with no considerable changes in internode number. Moreover, GA3 application restored leaf type, flower number, flowering time, fruit number and weight. Compared to the untreated DoDELLA1-1 transgenic lines, GA3 application caused a significant increase in internodes length, leaf length and width, flower number, fruit weight and number, while plant height was enhanced by 44.48%, and precocious flowering in DoDELLA1-1 transgenic lines. PP333 treatment showed the strongest compact growth phenotype in DoDELLA1-1 transgenic lines.

Fig. 5
figure 5

Phenotypes and hormone contents of wild-type (WT) and DoDELLA1-1tobacco lines under GA3 and PP333 treatments. A-B. Phenotypes of two tobacco lines at 120-day-old, bar = 10 cm. C-G. Hormone contents (GA1, GA3, ZT, IAA, ABA) of two tobacco lines at 120-day-old. Hormone accumulation was determined using HPLC-MS/MS on three biological and three technical replicates. Letters indicate significantly different (Student’s t-test; P < / > 0.05)

Full size image

Compared to untreated wild-type, GA3 treatment significantly increased GA1, GA3, ZT, IAA and ABA contents by 8.85 fold, 6.46 fold, 3.71 fold, 0.76 fold and 2.78 fold in the wild-type plants, respectively (Fig. 5). PP333 treatment significantly decreased GA1, GA3 and IAA contents by 1.02 fold, 1.26 fold and 1.05 fold in the wild-type plants, but increased ABA content by 0.41 fold, respectively. Compared to untreated DoDELLA1-1 transgenic lines, GA3 treatment significantly increased GA1 and GA3 contents by 1.76 fold and 2.63 fold in DoDELLA1-1 transgenic lines respectively, and increased ZT, IAA and ABA contents, but had no significant difference. PP333 treatment significantly decreased GA3 content by 0.56 fold in DoDELLA1-1 transgenic lines, but increased GA1, ZT and ABA contents by 0.51 fold, 0.92 fold and 0.90 fold, respectively. GA1, ZT, IAA and ABA contents of DoDELLA1-1 transgenic lines after GA3 treatment were restored to the wild-type plants level, and GA3 content was significantly higher than the wild-type plants.

GA3 treatment decreased DoDELLA1 expression level, while PP333 treatment enhanced their expression level in DoDELLA1-1 transgenic plants (Fig. 6 A). Furthermore, GA3 and PP333 treatment decreased the expression levels of NtGA20ox1-1, NtGA20ox1-3, NtGA2ox1-3 and NtGA2ox1-8 in wild-type plants respectively, and GA3 was more effective than PP333. GA3 treatment decreased significantly the expression levels of NtGA20ox1-1, NtGA20ox1-3, NtGA2ox1-3 and NtGA2ox1-8 in DoDELLA1-1 transgenic plants, increased significantly the expression levels of NtGA3ox4-1 and NtGA2ox2-3, respectively. PP333 treatment increased significantly the expression levels of NtGA20ox1-1, NtGA20ox1-3 and NtGA2ox2-3 in DoDELLA1-1 transgenic plants, decreased significant NtGA2ox1-3 and NtGA2ox1-8 expression, respectively (Fig. 6B).

Fig. 6
figure 6

The expression levels of genes in two tobacco lines under GA3 and PP333 treatments. (A) The expression levels of DoDELLA1 in DoDELLA1-1 tobacco lines. (B) The expression levels of tobacco GA metabolism-related genes in wild-type (WT) and DoDELLA1-1 tobacco lines. Transcripts accumulation was determined using qRT-PCR on three biological and three technical replicates. Letters indicate significantly different (Student’s t-test; P < / > 0.05)

Full size image

Discussion

DELLA proteins, which belong to the transcription factor in GA-dependent growth processes, have been reported in plant species, and designated as SLR1 in rice (Oryza sativa), DWARF8 (D8) and DWARF9 (D9) in maize (Zea mays), MhGAI in tea (Malus hupehensis), Rht in wheat (Triticum aestivum), and FvRGA1 in strawberry (Fragaria vesca)(Li et al. 2018a; Hirano et al. 2012; Jusovic et al. 2018; Wang et al. 2012; Lawit et al. 2010). Previous studies have indicated that DELLA proteins play an important role in plant growth, such as plant height, stem elongation, seed germination, floral and root development(Vera-Sirera et al. 2016). In this study, the differences in expression levels of DoDELLA1 in two yam cultivars may be the main factor regulating the varietal differences. Although DoDELLA1 studies have been conducted functional in GA signaling by DoGID1s-DoDELLA1 in yam(Zhou et al. 2021), the regulatory role in yam is poorly understood. Therefore, the functions of the DoDELLA1 protein were investigated in tobacco development.

A transcriptional activation function of the N-terminal domains of DoDELLA1

DELLA proteins are a subfamily of the plant-specific GRAS family of transcriptional regulators that mediates gibberellin signaling. DELLA proteins have a unique structure divided into two major domains. The C-terminal GRAS functional domain is characterized by several motifs: two leucine heptad repeats (LH), VHIID and SAW, the N-terminal GA perception domain has two specific motifs: the DELLA and TVHYNP domains(Daviere and Achard 2013; Hauvermale et al. 2012). In previous studies, DoDELLA1 has two highly conserved C-terminal and N-terminal domains, and GA3-DoGID1s-DoDELLA1 interaction was established(Zhou et al. 2021), which suggests the DoDELLA1 may indeed establish functional in GA signaling. Interestingly, although DoDELLA1 was indeed establishing functional interactions with DoGID1s, the N-terminal domains of DoDELLA1 had transcriptional activation (Fig. 1D). Previously, it has been suggested that DELLA and TVHYNP domains that were essential for this activity and GA response(Daviere and Achard 2016; Sun et al. 2010; Hirano et al. 2012). Additionally, DELLA proteins have been reported to act as a transcriptional coactivator in certain development contexts. AtGAI binding AtGAF co-transcriptionally activated GA synthesis pathway AtGA20ox, AtGA3ox and AtGID1b expression to regulate GA homeostasis(Fukazawa et al. 2014, 2017).

DoDELLA1 plays an important role in the growth regulation of tobacco.

The DELLA proteins were considered as master negative regulators of GA signaling, due to the dwarfism phenotypes in the overexpression of function, and tall or slender phenotypes characterized the loss of function(Daviere and Achard 2016; Hauvermale et al. 2012). Different plant species have different numbers of DELLA proteins and regulate different plant phenotypes. AtRGA and AtGAI perturbed overall plants growth behavior, including rooting capacity, plants architecture, and general vegetative and reproductive growth (Dill and Sun 2001; Feng et al. 2008; de Lucas et al. 2008), AtRGL2 was the major inhibitor of seed germination(Lee et al. 2002), AtRGA, AtRGL1 and AtRGL2 together exhibited a significant reduction in silique length, seed number and disorder flower structure(Cheng et al. 2004; Tyler et al. 2004). The DWARF8 (D8) and DWARF9 (D9) in maize showed the same severe phenotype(Lawit et al. 2010). Overexpressing PslDELLA in Arabidopsis had strong dwarf plant height, short internodes, smaller flower size and seeds(El-Sharkawy et al. 2017).

In this study, we determined 6 transgenic lines by qRT-PCR and the expression level of DoDELLA1 was higher, while the plant height was a severe dwarf (Fig. 2). Wild-type plants displayed the typical phenotype, whereas DoDELLA1 transgenic plants were shorter in stature and leaf, late-flowering or no flowering, suggesting that DoDELLA1 protein function is consistent with other plant DELLA protein, and perturb tobacco plants growth behavior. SIDELLA inhibited tomato fruit size, but exogenous GA3 rescued its phenotype(Marti et al. 2007). GA treatment rescued to a certain extent PslGAI and PslRGL plants height mainly due to extending internode length(El-Sharkawy et al. 2017). In this study, GA3 treatment was rescued to DoDELLA1 transgenic plant height and leaf size (Table 1; Fig. 5). It suggests that DoDELLA1 is phenotypically characterized for GA-dependent traits and in response to GA3 treatment.

The GA3 levels in DoDELLA1-1 transgenic lines are significantly lower than in wild-type lines (Fig. 5). The significant different accumulation of GA3 levels can be due to higher transcription DoDELLA1 still impact GA metabolism to decrease GA3 levels in DoDELLA1-1 transgenic lines (Fig. 6). PP333 treatment significantly increased GA1, ZT and ABA contents (Fig. 5), but exhibited the strong compact growth phenotype in DoDELLA1-1 transgenic lines (Table 1; Fig. 5), which had opposite effects with GA3 treatment. There are maybe two reasons. First, ABA content increased by PP333 treatment can inhibit the growth of DoDELLA1-1 transgenic lines. Second, DELLA proteins not only are degraded with the classical proteolytic-dependent pathway in GA signaling, but also other alternative pathways. Partly mutant alleles of sly could not germinate, as they impaired the degradation and accumulated high levels of DELLA, and overexpression of GID1 could rescue germination without DELLA degradation, in which the formation of the GA-GID1-DELLA complex could inactivate the DELLA protein without degrading it (Ariizumi et al. 2008). SPINDLY is an O-linked-N-acetylglucosamine transferase that regulated the activity of the DELLA protein through glycosylation in Arabidopsis, rice and barley(Filardo et al. 2009; Shimada et al. 2006; Swain et al. 2001).

Mechanism of DoDELLA1 regulates tobacco growth.

Bioactive GA plays a crucial role in coordinating different plant growth, and DELLA proteins may regulate GA biosynthesis modulating the expression of corresponding genes(Zentella et al. 2007). In this study, it was found that GA1 and GA3 content decreased in over-expression DoDELLA1-1 transgenic tobacco over 30-120d, and the expressions of NtGA20ox1-1, NtGA20ox1-3 and NtGA3ox4-1 were lower, while those of NtGA2ox1-3, NtGA2ox1-8 and NtGA2ox2-3 were higher during the same time (Fig. 4). These findings are consistent with overexpression of PmRGL2 in poplars was reduced PmGA20ox2, PmGA3ox1 expression which reduced the biosynthesis of GA4 to release bud dormancy(Lv et al. 2018). AtGA20ox2 and AtGA3ox1 were identified as targets of DELLA, which activate the expression of these genes(Zentella,2007). SlDELLA might activate the expression of GA biosynthesis genes GA20ox1 and GA3ox1 in tomato fruit initiation(Hu et al. 2018). The DELLA-GAF1 complex was the main component in GA feedback regulation of GA20ox2 and GA3ox1 to elaborate leaf margin formation in Medicago truncatula (Wen et al. 2021). It suggests that DoDELLA1 is conserved and involved in the regulation of GA metabolism genes to impact GAs levels, which perturbs tobacco plants’ growth.

The application of GA3 increased endogenous GA3 levels, and PP333 decreased endogenous GA3 levels in wild-type and DoDELLA1-1 transgenic plants. In the meanwhile, GA3 treatment decreased the expression levels of NtGA20ox1-1, NtGA20ox1-3, and PP333 treatment increased the expression levels of NtGA20ox1-1, NtGA20ox1-3 in wild-type and DoDELLA1 transgenic plants. It means that DoDELLA1 transgenic plants have feedback and feed-forward mechanism, whereby bioactive GAs reduces GA synthesis and speeds up GA deactivation. AtGAI binding AtGAF co-transcriptionally activated GA synthesis pathway AtGA20ox, AtGA3ox and AtGID1b expression to regulate GA homeostasis through feedback(Fukazawa et al. 2014, 2017). The expression of NtGA20ox1-1, NtGA20ox1-3 are maintained by a feedback regulation to respond to levels of bioactive GA3 in DoDELLA1 transgenic plants. However, NtGA3ox4-1, NtGA2ox1-3, NtGA2ox1-8 and NtGA2ox2-3 expression levels were not consistent with feedback and feed-forward mechanism in wild-type and DoDELLA1-1 transgenic lines under GA3 and PP333 treatment (Fig. 6). Auxin and brassinosteroid regulated GA3ox and GA2ox expression mediating GA content(Hedden and Thomas 2012). It is speculated that other hormones are involved in the regulation of GA metabolism process. The nature of the differential expression between wild-type and DoDELLA1-1 transgenic lines is yet unclear and will require further analyses.

DELLA protein alleles Vvgai1 inhibited plant height, leaf and flower development, a DELLA-centered feedback mechanism that maintains GA homeostasis in Vvgai1 transgenic soybean plants was found, but it was believed that the intricate interactions of DELLAs with numerous transcription factors control plant development and growth, which involved the growth of stem, leaf and flower, such as VvGRF (GROETH REGULATING FACTOR), VvPIF1 (PHYTOCHROME INTERACTING FACTOR 1), VvEXP1 (EXPANSIN), VvCO (CONSTANS) and VvTFL1 (CENTRORADIALIS)(Arro et al. 2019). In this study, DoDELLA1 proteins reduced internode length and height in tobacco, and impacted GA metabolism genes to regulate GA3 level. However, it could not be an identity that DoDELLA1 protein is involved in GA homeostasis regulation directly. Since previous studies indicate that DELLAs repress plant growth by binding to a wide variety of proteins or transcriptional regulators by GA signaling, while involved in other plant hormones. DELLA proteins interacted with ARF6 (AUXIN RESPONSE FACTOR), BZR1 (BRASSINAZOLE-RESISTANT1) and PIF4 (PHYTOCHROME INTERACTING FACTOR 4), which involved promoting hypocotyl cell elongation in Arabidopsis thaliana seedling growth by integrating GA, brassinosteroid, auxin, and light/temperature environmental stimuli(de Lucas et al. 2008; Feng et al. 2008; Gallego-Bartolome et al. 2012; Oh et al. 2014). AtDELLA, AtABI3 (ABA INSENSITIVE 3), and AtABI5 (ABA INSENSITIVE 5) formed a protein complex that bound the promoter and activated the transcription of target genes, such as SOMNUS (SOM), a C3H-type zinc finger that negatively regulated seed germination(Lim et al. 2013). In this study, GA1, GA3 and ABA contents were significantly reduced in the DoDELLA1-1transgenic lines, while GA1, ZT, IAA and ABA contents were restored to the wild-type plants level after GA3 treatment (Fig. 5). The mechanism of DoDELLA1 regulating GA metabolism and signaling pathways in an integrated manner with other transcription factors by ZT, IAA and ABA, is still unknown, and needs further research.

Conclusions

In conclusion, our results suggest that DoDELLA1, a considered major negative regulator of GA signaling gene from yam tuber, inhibited the growth and development of transgenic tobacco plants. The observed changes in the dwarf phenotype of the transgenic tobacco plants may result from GA content changes via the regulation of the expression of GA-related genes by DoDELLA1.