SiFBA5, a cold-responsive factor from Saussurea involucrata promotes cold resilience and biomass increase in transgenic tomato plants under cold stress

Background Saussurea involucrata survives in extreme arctic conditions and is very cold-resistant. This species grows in rocky, mountainous areas with elevations of 2400–4100 m, which are snow-covered year-round and are subject to freezing temperatures. S. involucrata’s ability to survive in an extreme low-temperature environment suggests that it has particularly high photosynthetic efficiency, providing a magnificent model, and rich gene pool, for the analysis of plant cold stress response. Fructose-1, 6-bisphosphate aldolase (FBA) is a key enzyme in the photosynthesis process and also mediates the conversion of fructose 1, 6-bisphosphate (FBP) into dihydroxyacetone phosphate (DHAP) and glycerol triphosphate (GAP) during glycolysis and gluconeogenesis. The molecular mechanisms underlying S. involucrata’s cold tolerance are still unclear; therefore, our work aims to investigate the role of FBA in plant cold-stress response. Results In this study, we identified a cold-responsive gene, SiFBA5, based on a preliminary low-temperature, genome-wide transcriptional profiling of S. involucrata. Expression analysis indicated that cold temperatures rapidly induced transcriptional expression of SiFBA5, suggesting that SiFBA5 participates in the initial stress response. Subcellular localization analysis revealed that SiFBA5 is localized to the chloroplast. Transgenic tomato plants that overexpressed SiFBA5 were generated using a CaMV 35S promoter. Phenotypic observation suggested that the transgenic plants displayed increased cold tolerance and photosynthetic efficiency in comparison with wild-type plants. Conclusion Cold stress has a detrimental impact on crop yield. Our results demonstrated that SiFBA5 positively regulates plant response to cold stress, which is of great significance for increasing crop yield under cold stress conditions. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-021-02851-8.


Background
The Calvin cycle is a critical part of photosynthetic carbon fixation, providing essential compounds for plant survival [1,2]. This process is associated with a variety of catalytic reactions and divided into three distinct phases: carbon fixation, reduction, and regeneration of the starting molecule [3]. It has been reported that plants grown in moderate light and temperature conditions can withstand more than a 50% reduction in ribulose-1, 5-biphosphate carboxylase-oxygenase (Rubisco) activity with little effect on photosynthesis. In contrast, under high light conditions, decreased Rubisco activity was related to the reduction of photosynthesis [4]. Rubisco is not always the most active enzyme during the Calvin cycle, other enzymes probably exert greater control over carbon flow during photosynthesis, specifically, sedoheptulose-1, 7-bisphosphatase (SBPase), transketolase (TK), and fructose-1,6-bisphosphate aldolase (FBA). Therefore, these enzymes are candidates for study in efforts to increase photosynthetic carbon fixation [3,5].
FBA is a key metabolic enzyme that mediates the reversible conversion of fructose-1,6-bisphosphate (FBP) into dihydroxyacetone phosphate (DHAP) and glyceraldehyde-3-phosphate (GAP) during glycolysis and gluconeogenesis [6,7]. In the Calvin cycle, FBA can also catalyze the condensation of sedoheptulose-1,7-bisphosphate from erythrose-4-phosphate and GAP [8,9]. Antisense inhibition and photosynthesis modeling studies have shown that FBA may play a major role in controlling carbon metabolism rates and ribulose-1,5biphosphate (RuBP) regeneration in the Calvin cycle [10][11][12]. Experiments in tobacco (Nicotiana tabacum) indicated that overexpression of FBA can promote RuBP regeneration while slightly increasing growth and biomass productivity [5]. Additionally, a study on tomato (Solanum lycopersicum) showed that SlFBA7 overexpression increased the activity of several major enzymes in the Calvin cycle, as well as the net photosynthetic rate (Pn), seed size, and stem diameter [13].
FBA is also involved in plant cold-stress response. For instance, cold stress upregulates FBA family member ClAldC expression in Codonopsis lanceolata [14]. Interestingly, the FBA family responds differently to cold treatment in Arabidopsis and rice (Oryza sativa) [13,15]. Despite extensive research on several model plants and commercial crops, few studies have explored the role of FBA in more extreme habitats.
S. involucrata is a perennial alpine plant resistant to extreme cold stress [16]. Despite having a short life cycle, S. involucrata can accumulate substantial biomass and has evolved numerous genes specifically associated with cold tolerance [17]. Recently, transcriptome sequencing and extensive bioinformatics analysis were completed for S. involucrata, providing us with an opportunity to determine the specific FBA gene associated with S. involucrata cold stress tolerance [18]. In the present study, we used Agrobacterium-mediated transformation to construct transgenic lines of tomato plants to study the putative function of SiFBA5 genes [19,20].

Isolation of siFBA5 from S. involucrata
SiFBA5 cDNA was identified from transcriptome sequencing of S. involucrata. Full-length cDNA was obtained using gene-specific primers [17]. Sequencing analysis revealed that the open reading frame (ORF) of SiFBA5 cDNA (1179 bp) encodes a 392-amino-acid protein, with a molecular mass of 42.57 kDa and isoelectric point (pI) of 8.84 ( Figure S1-S2). Multiple sequence alignment revealed that SiFBA5 shared 75.2%, 74.9%, and 70.2% similarity to Compositae homologs CcFBA3, LsFBA3, and HaFBA3, respectively; thus, the gene was highly conserved (Fig. 1a). The gene also exhibited the canonical lysine residue of Class I aldolases involved in aldol condensation. Phylogenetic analysis indicated that SiFBA5 is closely related to the homolog AtFBA3, a chloroplastic FBA (Fig. 1b). This outcome suggests that SiFBA5 localizes to the chloroplast.

Expression patterns of SiFBA5
To investigate how SiFBA5 responds to cold stress, S. involucrata rosettes at the five-leaf stage were exposed to 4°C for 1, 3, 6, 12, and 24 h. After 1 h, the relative expression of SiFBA5 increased 10-fold, before gradually declining (Fig. 2a). This rapid and strong induction of expression suggested that SiFBA5 participates in the initial response to cold stress.

Subcellular localization of SiFBA5
WoLF PSORT (https://wolfpsort.hgc.jp/), YLoc (www. multiloc.org/YLoc), and MultiLoc2 (http://www-bs. informatik.uni-tuebingen.de/Services/MultiLoc2) predicted that SiFBA5 subcellular localization was in the chloroplast. To test the robustness of this prediction, we constructed a SiFBA5-GFP fusion protein and used confocal microscopy to assay its presence in Nicotiana benthamiana leaf cells. We confirmed that SiFBA5-GFP was localized to the chloroplast (Fig. 2b). Therefore, SiFBA5 likely participates in the Calvin cycle and may be related to downstream gene regulation during photosynthetic carbon flux.

Cold stress responses of transgenic SiFBA5 overexpression lines
To reveal the functional roles of SiFBA5, we constructed tomato overexpression lines under the control of the 35S promoter. Analysis with PCR verified transgene integration. Under normal temperature conditions (25°C), growth did not differ between overexpression lines and wild type, suggesting that excess SiFBA5 does not cause an aberrant phenotype (Fig. 3). However, during cold tolerance assays (4°C), wild-type plants exhibited thinning and drooping leaves, along with bent side-branches, whereas overexpressing SiFBA5 plants did not. Furthermore, at 0°C, wild-type leaves curled and deepened in color, while transgenic leaves were no longer viable (Fig. 3).

Physiological changes in transgenic plants under cold stress
Cold stress causes the overproduction of reactive oxygen species (ROS) and oxidative damage in plant cells. Under  From left to right: channels of GFP fluorescence, bright field, chloroplast autofluorescence and an overlay of all three. Scale bar = 20 μm. One-way ANOVA was used to compare the statistical difference between measurements (P < 0.05) cold stress conditions (4°C), the overexpression lines exhibited significantly less malondialdehyde (MDA) accumulation than wild type (P < 0.05) (Fig. 4a). Since MDA is the product of ROS-induced lipid peroxidation, this outcome indicated that the plants were subject to less oxidative stress when SiFBA5 was overexpressed. We then assayed the activities of three major antioxidant enzymes (superoxide dismutase, SOD; catalase, CAT; and peroxidase, POD) to investigate the cause of this change. There was no significant difference in SOD activity between transgenic and wild-type plants under normal conditions, but under cold stress, SOD activity was significantly enhanced (P < 0.05) (Fig. 4b). Additionally, we found that POD activity increased significantly in the transgenic lines compared with the wild type under cold stress (P < 0.05) (Fig. 4c). As temperatures decreased, the corresponding decrease in CAT activity was more pronounced than the decreases in SOD and POD activity. Transgenic and wild type plants again differed significantly, with less CAT activity in the former than in the latter (P < 0.05) (Fig. 4d). Our results suggest that SiFBA5 overexpression inhibits ROS damage under cold stress by both decreasing MDA accumulation and enhancing antioxidant enzyme activity.

Effects of SiFBA5 on photosynthetic and chlorophyll fluorescence factors
We determined changes in net photosynthetic rate (Pn) to delve deeper into how cold stress alters photosynthetic efficiency. Our findings showed that transgenic plants and wild-type plants did not differ significantly in Pn under normal conditions. However, at 10°C and 4°C, overexpression lines displayed significantly higher Pn than wild type (P < 0.01) (Fig. 4g). In line with this observation of elevated photosynthesis under SiFBA5 overexpression, shoot and root dry biomass were respectively 84.34 and 40.14% greater in transgenic plants than in wild type (P < 0.01) (Fig. 4h). We also observed significantly higher total chlorophyll content in overexpression lines than in wild type at 16°C and 10°C (P < 0.05) (Fig. 4e). Finally, we measured the maximum efficiency of PSII photochemistry (Fv/Fm), which reflects PSII photoinhibition. In the SiFBA5 overexpression lines, Fv/Fm was greater than in wild-type plants (P < 0.05) (Fig. 4f), indicating that cold stress did not damage the PSII complex in transgenic plants.

Discussion
Plants are sessile and cannot avoid damage in their environment by moving, so they have evolved a series of genes to respond to stress. A previous study found that the use of Agrobacterium-mediated transformation of Persian walnut (Juglans regia) somatic embryos can enhance the resistance response of the plants and contribute to better appearance of the kernels for consumers [21][22][23]. Freezing injury is a major cause of crop loss and a limiting factor. There are some studies regarding cold hardiness mechanisms and methods of cold adaptation in walnut [24][25][26].
Accordingly, this study is designed to exploit the FBA gene from S. involucrata and transfer it into tomato plants. Fructose-1,6-bisphosphate aldolase is critical to carbon metabolism in plants, playing a role in processes such as glycolysis, gluconeogenesis, pentose phosphate pathway, and the Calvin cycle [7]. Studies in multiple plant species have amply demonstrated the critical role of FBA in regulating plant stress-tolerance, growth, and development [5,14,27,28]. To the best of our knowledge, the present study is one of the first studies to examine FBA function in a plant species endemic to an extreme habitat, specifically S. involucrata, which can survive even at temperatures below − 21°C [29]. Transcriptome analysis of S. involucrata identified 15 FBA gene transcripts, three of which were upregulated in response to cold stress (data not shown). We selected SiFBA5 here for further analysis. A bioinformatics analysis revealed that SiFBA5 proteins are highly similar to FBA homologs in other species. Consistent with the low-temperature transcriptome profile, cold stress , and e total chlorophyll content. f Maximal efficiency of PSII photochemistry (Fv/Fm), g net photosynthetic rate (Pn), h biomass. Mean ± SD was obtained in one experiment with three biological replicates. Asterisks indicate significant differences between WT and transgenic plants (*P < 0.05; **P < 0.01). One-way ANOVA was used to compare the statistical difference between measurements (P < 0.05) significantly increased SiFBA5 transcript abundance. Phylogenetic analysis of SiFBA5 revealed highly homologous relationships with chloroplast FBAs in plants such as Arabidopsis, rice, tobacco, and tomato [5,13,15]. In higher plants, FBAs have two isozymes: cytosolic and chloroplast [15], the latter is involved in the Calvin cycle. In potatoes, suppression of chloroplast FBA inhibits photosynthesis [30]. Additionally, the SiFBA5-GFP fusion protein was localized to the chloroplast (Fig. 2b). Our SiFBA5-overexpression transgenic lines revealed improved photosynthetic efficiency and cold tolerance compared with wild-type plants under low-temperature exposure.
We also successfully identified some of the physiological processes involved in the cold stress response, such as increased MDA production. Therefore, we tested the MDA content in wild type and SiFBA5overexpression transgenic lines under cold stress, which revealed a decrease in the MDA content of SiFBA5-overexpression transgenic lines relative to wild type (P < 0.05). These results indicated that SiFBA5-overexpressing plants had lower oxidative stress (Fig. 4a) [31]. Furthermore, under cold stress, transgenic plants had significantly higher SOD, POD, and CAT activity than in the wild type. These three enzymes are part of the antioxidant defense system that evolved in plants to counteract ROS-induced oxidative damage (e.g., lipid peroxidation) [32]. Given that abiotic stress increases the activity of these enzymes; we speculated that overexpression of SiFBA5 Values are mean ± SD (n = 3). One-way ANOVA was used to compare the statistical difference between measurements (P < 0.05) can eliminate excess ROS from cells, offering protection against oxidative damage.
Low temperatures decrease plant capacity to use light energy and increase photoinhibition in cold-sensitive plants [33]. Some data suggest that photosynthesis is inhibited through reduced aldolase expression [10]. Another study found that AtFBA overexpression specifically promoted RuBP regeneration, CO 2 fixation efficiency, and photosynthetic rate in transgenic tobacco [5]. Our results demonstrated that SiFBA5 overexpression lines under cold stress constitutively elevated FBA expression in chloroplasts and promoted Rubisco expression, which enhanced RuBP regeneration, leading to accelerated RuBP metabolism. These changes then promote the entire photosynthetic pathway. Variation in the individual FBA of transgenic plants showed that the regenerative capacity of the C3 cycle places limitations on photosynthetic rate [3].
Cold stress decreases Fv/Fm, making the latter an important variable in screening for cold tolerance [34]. Our study confirmed that cold stress reduces Fv/Fm, resulting in plants that cannot efficiently use excitation energy and thus exhibit reduced PSII photochemical efficiency. However, transgenic plants had higher Fv/Fm than wild type under cold stress. The ability to sustain proportionately higher Fv/Fm could enhance photosynthesis after a cold-stress-induced slowdown. We also found that overexpression of chloroplast FBA slightly increased Pn under cold stress, but not under normal temperatures. Thus, prolonged exposure to cold stress appeared to reduce SiFBA5 transcription, but the overexpression of SiFBA5 partially compensated for FBA loss, which is in line with data from potatoes [30] and tomatoes [35].

Conclusion
We successfully cloned SiFBA5 from S. involucrata and identified the gene product as an aldolase. Overexpressing SiFBA5 in tomato alleviated cold-stress-induced cellular damage, improved cold tolerance, and improved photosynthetic capacity under cold stress conditions. Our data provide a theoretical foundation and experimental basis for using SiFBA5 in the development of S. involucrata germplasm resources. Moreover, the gene is a strong candidate for genetic engineering to increase crop production and stress resistance in cold climates.

Plant materials and treatments
S. involucrate seeds were a gift from Zhaosu County Forestry Bureau, Yili Prefecture, Xinjiang. Seedlings of S. involucrata were cultured using plant tissue culture technology and placed in a glasshouse at 19°C with a 16 h light/8 h dark photoperiod.
Nicotiana benthamiana seeds were grown in pots with 3:1 peat moss and vermiculite soil (v/v) under a photoperiod of 12 h light at 26°C/12 h dark at 22°C, with5 0% relative humidity. Tomato (Yaxin 87-5) seeds were surface-sterilized in 75% ethanol for 1 min, 2% v/v sodium hypochlorite for 15 min, and rinsed six times with sterile distilled water. Seeds were sown on 1/2 Murashige and Skoog (MS) medium. For a week, seeds were incubated at 25°C under a 16 h light/8 h dark cycle with fluorescent lights. Cotyledon explants of 7-d-old seedlings were used for Agrobacterium transformation. For cold treatment, rosettes at the five-leaf stage were exposed to 4°C for 1, 3, 6, 12, and 24 h, respectively. All tissue samples were immediately frozen in liquid nitrogen and stored at − 80°C.

RNA isolation and qRT-PCR
Total RNA was isolated with TRIzol from S. involucrate and tomato, following manufacturer protocol. PrimeScriptTM RT reagent Kit (TaKaRa, China) was used for first-strand cDNA synthesis. Quantitative real-time PCR (qRT-PCR) was performed in the Light Cycler 480 system (Roche Diagnostics, Germany) using LightCycler® 480 SYBR Green I master mix (Roche Diagnostics). The thermocycling protocol was 40 cycles of 95°C for 10 s, 60°C for 20 s, and 72°C for 30 s. Relative expression was calculated with the 2 -ΔΔCt method. Each expression profile was independently verified in triplicate. qRT-PCR primers are shown in Supplementary Table S1. We used S. involucrat cDNA to amplify the FBA gene and cDNA from tomato to detect the expression of related genes. The experiment is represented as means ± SE of three replications.

Isolation of SiFBA5 gene from S. involucrata
The SiFBA5 gene sequence was obtained from Saussure a K B a s e ( h t t p : / / w w w . s h e n g t i n g b i o l o g y . c o m / SaussureaKBase/index.jsp.). Putative full-length SiFBA5 cDNA and genomic DNA were amplified using genespecific primers SiFBA5-F and -R (Supplementary Table  S1). SiFBA5 cDNA was ligated into the pMD®19-T simple vector (TaKaRa, China) and E. coli DH5α competent cells (TransGen, China); successful cloning was verified via Sanger sequencing (BGI, Beijing, China). All subsequent constructs were made using this clone as a template. SiFBA5 gene homologs were identified using BlastP in NCBI (https://blast.ncbi.nlm.nih.gov/Blast. cgi). Multiple sequence alignments were performed in ClustalX2 with default parameters and DNAMAN. Alignments were then adjusted for constructing the phylogenetic tree in MEGA 5.0.

DNA constructs
The coding region of SiFBA5 was amplified using primers G-SiFBA5-F and -R, with appropriate restriction sequences (Table S1). PCR products were digested with BamH I and Sal I, then ligated with the multiple cloning site (MCS) of pCAMBIA2300 binary vector, containing GFP as the reporter under the control of cauliflower mosaic virus 35S (CaMV 35S) promoter [36]. The construct for SiFBA5 subcellular location was generated through the in-frame fusion of complete SiFBA5 CDS (stop codon TAG deleted) with GFP. To generate the overexpression construct, SiFBA5 ORF was cloned into the pCAMBIA2300 vector under the control of the CaMV 35S promoter. The insert was released from pMD19-T-SiFBA5 through Sma I and Sal I digestion, and then ligated into pCAMBIA2300 MCS. Constructs were introduced into Agrobacterium strain GV3101 through electroporation using the GENE PULSER II system (Bio-Rad, Hercules, CA, USA).

Plant transformation
The 35S::SiFBA5 vectors were introduced into a tomato for overexpression lines, respectively. Select sterile, robust tomato seedlings with fully expanded cotyledons, cut off the two ends of the cotyledons, and then divide them into two from the middle of the leaves. Use the leaf discs as genetically transformed explants, and spread them on MS + 2 mg/L 6-BA+ 0.15 mg/L IAA medium with the backside down, dark culture for 2 days. Transformation into the tomato, with cotyledons as explants, followed published protocols [37]. Callus induction and shoot regeneration were induced. Next, inoculated explants were screened on MS medium supplemented with different concentrations of plant growth regulators and antibiotics. Shoots were regenerated on a selective medium containing kanamycin (50 mg L − 1 ). Rooted kanamycin-resistant primary transformants (5-10 cm high) were transferred to potting soil; SiFBA5 presence was confirmed via PCR. The 35S::SiFBA5-GFP vectors were introduced into tobacco for subcellular location experiments, the third to fifth youngest leaves of the three-to four-week-old Nicotiana benthamiana (approximately 8-10 leaf stage) was injected with Agrobacterium-expressing GFP-fusion protein using a 1 mL syringe. Following 2-3 d incubation in growth chambers, a laser scanning confocal microscope (Zeiss, Germany) was used to image GFP fluorescence in tobacco leaves.

Stress treatment
Stress treatments were initiated after plants were grown for 9 weeks in the soil at 25°C under a 16 h/8 h (light/ dark) photoperiod. To investigate the role of SiFBA5 in cold tolerance, wild-type and transgenic tomato plants were incubated (Percival Scientific, USA) for a week under a 12/12 h light (25°C)/dark (22°C) cycle with moderate light intensity and 50-60% relative humidity. Then, wild-type and transgenic tomato plants were subjected to cold stress at 16, 10, and 4°C, each for 3 d. Subsequently, plants were placed overnight in a chamber set to 0°C.

Physiological measurements
Physiological traits were measured at different stages after cold stress treatments. Analysis of stress-induced oxidative damage used an MDA assay as previously described [38]. The nitro blue tetrazolium (NBT) photoreduction method was used to estimate SOD activity [39]. Next, POD activity was determined through monitoring increases in absorbance at 470 nm during guaiacol oxidation [40]. Finally, CAT activity was assayed through monitoring decreases in H 2 O 2 absorbance (240 nm) within 1 min [41]. Total chlorophyll contents were measured following a previous method with some modifications [33]. We weighed out 0.1 g of leaf tissue into pieces, total chlorophyll was extracted using ethanol: acetone: H 2 O (4.5:4.5:1) and analyzed by UV spectrophotometry.

Gas-exchange and chlorophyll fluorescence
Net photosynthesis and stomatal conductance were detected using the portable photosynthesis system LI-6400 (LI-COR, USA). Chlorophyll fluorescence was measured with a Mini PAM chlorophyll fluorometer (Waltz, Germany) [42].

Statistical analysis
All data are represented as means ± SE of three replications. SPSS 13.0 software (SPSS, Chicago, USA) was used to compare MDA content, SOD activity, POD activity, CAT activity, or relative gene expression in the control and transgenic lines. One-way analysis of variance followed by Tukey's honestly significant difference (P < 0.05 and P < 0.01) multiple comparison tests were used to determine the significant differences. Error bars in all figures represent standard deviations from the mean. SigmaPlot 12.0 was used to generate figures.