Background

Light plays a pivotal role in plant growth and development. It provides solar energy for photosynthesis, and doubles as an environmental signal, orchestrating morphological and physiological trade-offs throughout the plants’ cycle [1]. However, shade conditions pose a significant barrier in agricultural production [2] and urban greening [3]. Lower vegetation in the intercrop** or vertical planting system particularly receives limited light [4, 5], especially ground covers and lawns [6,7,8].

Filtered by the upper plants, the light intensity at the bottom is reduced. Red (R, λ = 600–700 nm) and blue (B, λ = 400 ~ 500 nm) light are primarily captured by chloroplasts, while far red light (FR, λ = 700 ~ 800 nm) is partially preserved due to reflection of surroundings [8, 9]. This disproportionate reduction results in a depressed R: FR ratio and low B, which photoreceptors perceive as shade signals, leading to two contending strategies: shade avoiding or shade tolerant [10]. The shade-avoidance syndrome (SAS), characterized by rapid elongation of stems and petioles and accelerated flowering, has been extensively studied [11,12,13,14]. However, this strategy is limited to plants of similar stature situated in open habitats. For understory plants, the height disparity renders elongation ineffective in esca** shade [15]. Instead, these plants have evolved shade-tolerant strategies, suppressing SAS and reallocating resources towards optimizing photosynthesis and bolstering physical defenses [16,17,18]. Despite this, our understanding of shade tolerance remains fragmented [15, 19], particularly concerning understory herbs [20, 21], a topic which has so far received limited attention.

Given the partial overlapped of shading perception and response in both shade and non-shade plants [22, 23], it was considered that molecular regulatory components shared between shade tolerance and shade avoidance. The divergence in strategy, however, arises from different signal transduction pathways. Within molecular cascades of shade response, Phytochrome-Interacting Factors (PIF4, PIF5, PIF7) act as core signaling hubs coordinating the bulk of the downstream events [2, 13]. At low R: FR ratio, PIF degradation slows on account of impaired phosphorylation via the inactivation of phyB [24]. Resultantly, an accumulation of PIFs, particularly PIF7 [25, 26], triggers the auxin network, thereby prompting elongation [27]. Furthermore, cryptochromes (cry1 and cry2) perceive reduced blue light, and cry inactivation enhances PIF abundance (PIF4, PIF5) [28, 29]. Notably, PIF activity is moderated by inhibitors in a feedback-regulated method, including HFR1, PAR1, PAR2 [30], and DELLA protein [31]. Antagonistic factors that obstruct the regulatory pathway of the Shade Avoidance Syndrome (SAS) may hint at shade tolerance mechanisms [9].

In response to low light, acclimations also consist of modifications in leaf anatomical structure and chloroplast ultrastructure at cellular and organismal levels. Shade leaves display a higher specific leaf area [32], a larger proportion of spongy tissue [33], and a greater level of grana thylakoid stacking [34] when compared to sun leaves. Additionally, photosynthesis is notably modified under low light [35, 36], with enzyme concentrations in the Calvin-Benson cycle varying with light intensity [37]. The PSII/PSI ratio and LHCII heighten under low light exposure [38]. To acclimate to such low-irradiance environment, post-translational modifications, which mainly occur in the activity of metabolic enzymes, are typically swift and immediate. Examples of these modifications include protein phosphorylation (e.g., Nitrate reductase, EC:1.7.1.1) [39] and sulfhydryl reduction (e.g., Rubisco EC:4.1.1.39, fructose-1,6- bisphosphatase EC:3.1.3.11, sedoheptulose-1,7-bisphosphatase EC:3.1.3.37) [40]. Following environmental perturbation, new equilibrium is established between the photosynthetic electron transport [41], its correlated Calvin-Benson cycle [42], nitrogen assimilation [43], and reactive oxygen species metabolism [44, 45]. This primarily results from the fact that ATP and reducing power needed for these intertwined biochemical processes is drawn from absorbed light energy. A decrease in light brings about more intense competition among these processes.

Bermudagrass (Cynodon dactylon (L.) Pers.) is a perennial warm-season (C4) grass of the NAD-ME biochemical subtype [46, 47]. Its extensive usage as turfgrass or forage stem from exceptional resistance to abiotic stresses, yet it proves sensitive to shade [48]. C4 plants, characterized by the CO2-concentrating mechanism, demonstrate ecological dominance in warm, high-light environments [49]. However, limited survival of C4 plants in shaded conditions can be attributed to their relatively restrained plasticity and high energy consumption [50, 51]. NAD-ME type, in particular, is the most susceptible to low light among C4 biochemical isoforms [51]. Shading acclimations in C3 photosynthesis have been a focus of extensive studies over the past several decades, whereas comparable research concerning C4 is nearly nonexistent.

In order to investigate long-term photosynthetic acclimation to shade for C4 grass, bermudagrass was subjected to shade via a shading net for a duration of one week. By placing the photosynthetic electron transport at the center of our focus, we linked ROS metabolism with C/N assimilation. This method enabled the tracking of the energy absorption, transport, and utilization process on the thylakoid membrane. Such approach allowed for a comprehensive analysis of the shade-tolerant adaptability of bermudagrass from a holistic perspective.

Results

Alterations of photosynthetic pigments under shading conditions

Gradual shading conditions (Group B to E) led to linear changes in the photosynthetic pigment content within Bermudagrass. Lower light intensity corresponded with decreased total pigment, predominantly due to the significant reduction of chlorophyll b (Fig. 1A). The density of chlorophyll b in group E was almost half that of group B. Notably, chlorophyll b showed an increase in Group B relative to the control (Group A), thereby reducing the ratio of chlorophyll a to chlorophyll b. However, subsequent shade conditions induced an increase in carotenoid content (from 1.63% in group B to 12.27% in group E) and a significant surge in chlorophyll a level. Examination of pigment biosynthesis gene expression levels in selected groups (A, B, and E) provided insights into the observed pigment alterations. Lower light levels induced a down-regulation in the expression of HEMA, a gene coding for the enzyme catalyzing the initial step in chlorophyll synthesis (Fig. 1B). This may had led to a decrease in the total amount of chlorophyll. Additionally, it was observed that shading increased the transcript levels of light-harvesting complexes (LHCB2, LHCA2) and enzyme PDS, a crucial player in carotenoid synthesis, under light shade while a decrease was observed under heavy shade (Fig. 1C, D). An interesting observation was a sharp up-regulation of PSY under shading (Fig. 1D). Similarly, PORA’s transcription level saw a 3.9-fold increase in Group E relative to the control (Fig. 1B). All these changes suggest that the alteration in photosynthetic pigment synthesis and pigment-binding protein levels play a significant role in modulating the composition of photosynthetic pigments under different shading conditions.

Fig. 1
figure 1

The adjustment of light harvesting capacity of Bermudagrass under shade. (A) Photosynthetic pigment content (per gram of fresh weight). The numbers and letters enclosed in boxes refer to the mean values and statistical differences, respectively. The capital letters along the X axis categorize the increasing severity of the shade, ranging from Group A to Group E, indicating the control (100%, 500 µmol m− 2 s− 1) and subsequent shading treatments of 50% (250 µmol m− 2 s− 1), 25% (125 µmol m− 2 s− 1), 12.5% (62.5 µmol m− 2 s− 1), and 6.25% (31.25 µmol m− 2 s− 1), respectively. This study examined the transcript levels of key enzymes responsible for chlorophyll (B) and carotenoid synthesis (D), along with the light-harvesting complex subunits (C). The genes encoding these proteins are HEMA (glutamyl-tRNA reductase), PORA (Protochlorophyllide reductase A), PSY (Phytoene synthase), PDS (15-cis-phytoene desaturase), LHCB2 (Photosystem II light harvesting complex gene 2), and LHCA2 (Photosystem I chlorophyll a/b-binding protein 2). The expression levels of these genes were normalized, with the control (Group A) defined as the baseline. The statistical differences were identified using one-way ANOVA and Student-Newman-Keuls multiple range tests, columns marking the letters were used to signify Mean ± SD and statistically significant variation (P < 0.05). The study used five biological replicates to ensure accuracy

Full size image

Enhanced electron transfer beyond QA- was observed under shaded conditions

Fig. 2
figure 2

Illustration of shading effects on fast chlorophyll fluorescence transients and cyclic electron transport in Bermudagrass. (A) Polyphasic rise of chlorophyll fluorescence under varying shading conditions; (B, C) The apex variable fluorescence and initial fluorescence of PSII; (D) Transcription levels of PGR5L1A, PGR5L1B, and PGR5, which encode PGR5-like protein 1 A, PGR5-like protein 1B, and Protein PROTON GRADIENT REGULATION 5 respectively. Here, A group is defined as 1. We utilized an ANOVA with Student-Newman-Keuls (P < 0.05) in our study. Columns marking the letters were used to signify Mean ± SD and statistically significant variation (P < 0.05). Five biological replicates were considered

Full size image
Fig. 3
figure 3

Radar plots of parameters derived from the JIP-test. Information in detail was shown in Table 1

Full size image
Table 1 The variation of fluorescence transient parameters of Bermudagrass under shade
Full size table

Increased shading appeared to remodel chlorophyll a fluorescence (O-J-I-P) transient curve (Fig. 2A). The variable fluorescence (FV) elevated from group A to D and plummeted in group E (Fig. 2B; Table 1). We separated the curve into two segments, O-J and J-P phase, based on the count of QA− reductions. The O-J phase witnessed a noticeable increase in the initial fluorescence (F0) from group A to C, seemingly indicating an enhancement of light-harvesting capacity, given the parallel increase in LHC transcriptional levels and photosynthetic pigments as shown in Fig. 1A. As J-P phase variation surpassed that of the O-J phase (Fig. 2A), we focused our analysis on electron transfer parameters beyond QA−. Electron transport flux further than QA− per RC (ET0/RC) visibly surged in group D and E treatment compared to the control.

φE0 (efficiency/probability that an electron moves further than QA−), Ψ0 (quantum yield of electron transport), Sm and N (the pool size of the electron acceptor beyond QA−), and PIABS (Performance index for energy conservation from exciton to the reduction of intersystem electron acceptors) demonstrated a comparable trend (Fig. 3A, B; Table 1). These findings suggest an enhancement in electron transport under intense shade stress. Furthermore, the transcription levels of PGR5/PGR5L, a cyclic electron transport key component, were triggered by shade signal in groups A to C (Fig. 2D). Concurrently, parameters φP0,

ABS/RC and TR0/RC remained unaffected by shade, implying that the absorption and trap** of light energy per RC remains stable despite shading.

The accumulation and transport of carbohydrates were impeded under shade

Our analysis showed that the content of soluble sugar in leaves displayed a U-shaped pattern, reaching its lowest point in the C group as the shading increased, as shown in Fig. 4A. Conversely, root sugar content linearly diminished with each subsequent group (Fig. 4A). Following this, we investigated the enzyme activity and transcription levels of key components entailed in the synthesis, transport, and decomposition of carbohydrates like sucrose and starch to gain insight into the reasoning behind this. The in vitro enzyme activity tests showed that the activity of PEPC (Phosphoenolpyruvate Carboxylase) was decreased by shade (Fig. 4B). However, it recovered with increasing degree of shading (Fig. 4B). Moreover, an observed down-regulation was witnessed with the transcription level of TPT, SPS, SUT1, SS with increasing shade levels (Fig. 4C). This obstruction in leaf sucrose metabolism could very likely have prompted the observed decrease in soluble root sugar, given that the transport of carbohydrates in plants primarily occurs as sucrose. Additionally, a unique up-regulation was discovered in Beta amylose (BAM) transcription levels (Fig. 4D). This may well have been a corollary of the escalating soluble sugar content in groups D and E. However, the degradation of starch did not counterbalance the continuous decline in the soluble sugar content in the roots that came with increasing shade, as observed in Fig. 4D.

Fig. 4
figure 4

The carbon metabolism of Bermudagrass under shade. (A) Soluble sugar content in leaves (L) and roots (R); (B) Key enzymes (PEPC, Phosphoenolpyruvate carboxylase) activities associated with carbon assimilation in leaves; Transcript levels of genes connected with sucrose metabolism (C) and starch decomposition (D) in leaves. Protein names and functions of the corresponding genes: TPT (Triose phosphate/phosphate translocator, providing sucrose synthesis precursors), SPS (Sucrose-phosphate synthase, key rate-limiting enzyme in sucrose biosynthesis), SUT1 (Sucrose transport protein, transporting sucrose to the sieve element-companion cell for phloem loading), SS (Sucrose synthase, participating in the breakdown of sucrose), GWD (Alpha-glucan water dikinase, required for starch degradation), PWD (Phosphoglucan-water dikinase, required for starch degradation), BAM (Beta-amylase, required for starch degradation). Group A was considered as 1. A one-way ANOVA with the Student-Newman-Keuls test (P < 0.05) was carried out between various points on the same line or between different bars in the same set of histograms. Columns or lines marking the letters were ascribed to signify Mean ± SD and significant discrepancies. Five biological replicates were utilized

Full size image

The nitrogen assimilation adjustments of plant in shade surrounding

Fig. 5
figure 5

The nitrogen metabolisms in Bermudagrass under shade. (A) Soluble protein and crude protein content in leaf; (B, C, D) Transcriptional levels of genes related to nitrate reduction and ammonia assimilation. In the legend, “A” and “R-A” respectively denote leaves and roots of group A, and a similar nomenclature applies to groups C and E. Protein names of the corresponding genes are: FNRL2 (Ferredoxin-NADP reductase, leaf isozyme 2), NR (nitrate reductase), FD-NIR (Ferredoxin-nitrite reductase), FD (Ferredoxin), GS1 (Glutamine synthetase, cytosolic isozyme), GS2 (Glutamine synthetase, leaf isozyme, chloroplastic), Fd-GOGAT (glutamine oxoglutarate aminotransferase). The statistical methods and difference labeling mirror those of Fig. 4, with group A defined as 1

Full size image

Carbon and nitrogen metabolism in plants was inherently coupled and competitive, which form the material basis of crop yield and quality respectively. Observations uncovered increased crude protein content (an essential measure of forage quality) per leaf dry weight within mild shade (Fig. 5A). However, soluble protein content appeared to decline in low shade and then subsequently rebound as shading intensified (Fig. 5A). This trend indicates a shift in the allocation of nitrogen within Bermudagrass throughout different shading conditions. Organic nitrogen stems from the transformation of inorganic nitrogen (NO3−, NH4+) through nitrogen assimilation. Our current study examined the transcriptional level of enzymes involved in this process. As the shading deepened, the transcript levels of FNRL2, NR, and NIR exhibited a gradual decrease (Fig. 5B). This suggests that shading might inhibit the conversion of nitrate to ammonium in leaves. The expression level of FD increased in light shade and decreased in heavy shade. Interestingly, GS2 and Fd-GOGAT transcripts mirrored similar patterns as the level of FD (Fig. 5B, C). This could be because ferredoxin delivers the reducing capacity for the latter two. Simultaneously, shading resulted in an elevation of NR and GS1 levels within the root (Fig. 5D). Taken together, our findings propose that both the diminished reduction force and deficiency in carbon assimilation caused by shedding light intensity might contribute to the alteration of nitrogen assimilation in shaded conditions. Accordingly, roots shoulder more nitrogen assimilation responsibilities than chloroplasts under shade.

The contrasting antioxidant enzyme activities of leaves and roots in response to shading

Fig. 6
figure 6

Antioxidant enzyme activities (A, B) in roots and leaves under shading. In the legend, “SOD” and “R-SOD” refer to Superoxide dismutase activities in leaves and roots, respectively. It applies equally to POD (Peroxisome) and CAT (Catalase). A group was defined as 1. One-way ANOVA with Student-Newman-Keuls (P < 0.05) was performed between different points on the same line. Lines marking the letters indicated Mean ± SD and statistically significant differences. Five biological replicates were utilized for the experiments

Full size image

It was found that enzymatic activities changed in direct opposition under different light intensities in leaves compared to roots (Fig. 6A, B). Specifically, the SOD, POD, and CAT enzyme activities in leaves presented a U-shaped curve with light intensity reductions and estimated to the lowest in group C (Fig. 6A). Even though enzyme activities appeared to recover in groups D and E, they remained below their initial levels. By contrast, SOD, POD, and CAT enzyme activities in roots followed a parabolic trajectory with increased shading degree, peaking in group C (Fig. 6B).

Discussion

It is widely accepted that shade-tolerant plants display a lower chlorophyll a/b ratio but a higher PSII/PSI ratio than plants in sunny environments [15]. Teramoto et al. [52] and Wu et al. [53] indicated that the transcription levels of genes encoding LHCII proteins (including LHCB2), in Chlamydomonas reinhardtii or Camellia oleifera, increased in low-light environments in low light (50 µmol m− 2 s− 1) compared to medium light (200 µmol m− 2 s− 1). Notably, Chlorophyll b is instrumental in controlling the antenna size of the photosynthetic apparatus and ensuring LHCII stability [54,55,56]. Our study demonstrated that the chlorophyll b/aratio and the level of LHC transcripts initially rise and subsequently decline with diminishing light intensity (Fig. 1A, C). We hypothesize that the increasing LHC may correspond with the decrease in chla/chlb observed from group E to B. Furthermore, we observed a steady augmentation in carotenoid content as the shading grew deeper (Fig. 1D). Previous research suggested that PIFs suppress HEMA expression [57] and PSY [58]. However, PAR1 prevents PIF1 from inhibiting PSY1 expression in shaded conditions [59]. This alteration in PSY transcription levels may explain the increased carotenoid content recorded. Shading creates a low-light environment that substantially changes the content and composition of photosynthetic pigments, thereby profoundly affecting the input of light energy in photosynthesis. Based on these observations, we tentatively suggest that bermudagrass resistance to progressively deeper shade can transition from an active adaptation to passive tolerance.

State transitions have been suggested as a way to maximize light harvesting efficiency at low light intensities [60, 61]. A high proportion of far-red light under shade conditions dephosphorylates LHCII and shifts the photosynthetic apparatus to state 1 [62], where which LHCII detaches from PSI and rebinds to PSII. This process triggers an increase in overall chlorophyll fluorescence (ChlF) yield in “state 1” and a decrease in “state 2”, as the ChlF yield of PSI was much lower than that of PSII at room temperature [63]. In our study, we noted that F0, Fm and FV increase with deeper shading from group A to D (Figs. 2A, B and C and 3A; Table 1), which aligned with the previous reports.

Under low light conditions, plants aim to maximize light harvesting which consequently makes them vulnerable to slight surges in light intensity. Non-photochemical quenching (NPQ) serves as a protective buffer during the activation of photosynthesis in low light until CO2 assimilation is adequately stimulated [64]. The induction of NPQ is partially due to the swift initiation of the cyclic electron flow (CEF), a process largely governed by proton gradient regulation 5 (PGR5) and the PGR5-like photosynthetic phenotype 1 (PGRL1) [44]. The intermolecular disulfide bonds of PGRL1, formed in the dark, are reduced with the commencement of low light, accompanied by a brief surge in NPQ [65]. The transcription levels of PGR5 and PGR5LA, PGR5LB, as depicted in Fig. 2D, exhibit an upregulation under shade in Group C. On the other hand, linear electron transfer generates a static ATP to NADPH ratio of approximately 1.3–1.5, which is inadequate to meet the energy demands of C4 photosynthesis. This ATP demand rises proportionally with the quantity of CO2 that escapes from the bundle sheath back to the mesophyll cells in shaded conditions. To counterbalance this additional ATP demand in PEP regeneration, C4 plants initiate cyclic electron flow (CET) around photosystem I (PSI) [66]. Following this reasoning, bermudagrass could potentially respond timely to the excessive energy induced by light fluctuations and uncoupling of carbon assimilation in shaded conditions by fortifying the PGR5-facilitated NPQ pathway. Simultaneously, this could account for the diversion of photosynthetic electron transfer over QA− under shade, leading to a significant increment in ET0/RC, φE0, and Ψ0 in groups B to E, relative to the control, as observed in this study (Fig. 3A, B; Table 1).

The enzyme NADPH: protochlorophyllide (Pchlide) oxidoreductase (POR) is known to catalyze the transformation of Pchlide into chlorophyllide under illumination and ultimately into chlorophyll. In our research, we reported that the transcript levels of PORA showed an increase corresponding with progressive shading (Fig. 1B). This observation is consistent with prior studies, wherein PORA was found to accumulate at protein levels during skotomorphogenesis [67]. We speculated that this phenomenon might be associated with the ability of the functional POR: Pchlide complex to mitigate 1O2 production risk during greening, thereby preventing photobleaching post-illumination [68]. Functionally, this seems reminiscent of the defense mechanism provided by PGR5 against sudden photo-oxidative damage.

C4 plant species exhibit distinct morpho-anatomical and biochemical variances in carbon fixation mechanisms compared to their C3 counterparts [69]. The C4 photosynthesis involves a process of carbon fixation that utilizes a CO2-concentrating mechanism (CCM), operational in the mesophyll cells (MS) and the bundle sheath cells (BSC). Initially, the enzyme phosphoenolpyruvate carboxylase (PEPC) catalyses the formation of oxaloacetate (OAA) from CO2 and PEP, which is subsequently reduced to malic acid. This malic acid permeates into the BSC, leading to the decarboxylation of CO2, which then enters the Calvin cycle. Consequently, CCM elevates the concentration of CO2 around Rubisco, substantially decreasing Rubisco’s oxygenase activity. Under conventional conditions, the efficiency of carbon assimilation in C4 plants typically surpasses that of C3 plants. This is attributed to the higher energy demands of photorespiration in C3 plants when compared to the operation of CCM in C4 plants [70].

Contrarily, under shade or fluctuating light conditions, C4 species exhibit reduced photosynthetic capacity and phenotypic plasticity in comparison to C3 species. This can mainly be attributed to the requirement of C4 species for more enzymatic steps than C3 species to be activated by light, which is important for promoting the metabolite gradient between MC and BSC [71]. Consequently, this leads to lesser CO2 assimilation (or CO2 leakage) [72], and an inhibited ability to use inconsistent light effectively [50]. Depending on the primary decarboxylase in the BSC, C4 plants are classified into three biochemical isoforms, which are NAD-ME, NADP-ME, and PCK. Among them, the NAD-ME plants exhibit the least adaptation to low light [51]. Past research has demonstrated that shadow diminishes PEPC activity and initial Rubisco activity in all C4 grass subtypes. However, the PEPC inactivation in NAD-ME experiences minimum reduction due to shading [51, 73, 74]. The PEPC enzyme undergoes activation via phosphorylation. This process is facilitated by calcium-independent serine/threonine protein kinase (PEPC-PK), whose transcription and protein synthesis are reliant upon light [75]. This phenomenon serves as a pivotal explanation for the decline in PEPC activity under shaded conditions (Fig. 4B). In our investigation, we additionally observed a resurgence in PEPC activity with the deepening of shade (Fig. 4B), and we propose several potential rationales for this observation. Phosphorylated PEPC exhibits greater resistance to proteolysis in comparison to its dephosphorylated counterpart. Furthermore, phosphorylation engenders specific docking sites for protein-protein interactions, and the involvement of 14-3-3 protein may induce conformational alterations or influence the interactions of target molecules upon binding to distinct phosphorylation sites on various target proteins [75]. We posit that 14-3-3 protein might modulate protein stability through its engagement in protein interactions in shade. Moreover, previous research has demonstrated that both 14-3-3 protein and PEPC serve as binding proteins for phosphatidic acid (PA) [76], an inhibitor of PEPC activity. Thus, 14-3-3 protein could potentially restore PEPC activity in shaded environments by competitively binding to PA.

Sugar serves not only as an energy source and vital component of structural material in plants but also as a signal that regulates the expression of genes and enzymatic activities [77]. The Sucrose Non-Fermentation-Associated Kinase (SnRK1) is part of a protein kinase family pivotal to energy and metabolic homeostasis [78]. Past research indicated that SnRK1 phosphorylates and inactivates Sucrose Phosphate Synthase (SPS) and Nitrate Reductase (NR) [79], and is essential for the transcription of genes such as Sucrose Synthase (SS) and α-Amylase (α-AMY) [80, The assay of soluble sugar and protein

A 0.20 g portion of either freshly chopped leaves or roots was boiled in distilled water for 30 min. Following filtration, the volume of the filtrate was fixed, totaling 25 ml. 0.5 ml filtrate was used to detect soluble sugar content according to phenol method with sucrose as the standard [98].

In regards to the crude protein content, 0.20 g of dried leaves were digested with 10 ml H2SO4 via a graphite digester (SH220N, Hannon, China). The previous solution was then examined with an entirely automated Kjeldahl nitrogen determination instrument (K9860, Hannon, China), as the given instructions. Crude protein content (%) = nitrogen content × 6.25 × 100%. Soluble protein was determined by extracting the supernatant from 0.2 g of fresh leaves ground with liquid nitrogen using a pre-chilled phosphate buffer (pH = 7.8). The conditions for centrifugation were 12,000 × g at 4 °C for 20 min. 20 µl of the supernatant was mixed with coomassie brilliant blue (G-250) dye solution for 2 min. The light absorption at 595 nm was applied to estimate the soluble protein content [95], using bovine serum albumin as the protein standard.

Enzyme activity analysis

The enzymes were extracted from 0.2 g of fresh leaves or roots through liquid nitrogen grinding and centrifugation. The activities of SOD (EC:1.15.1.1) and POD (EC:1.11.1.7) enzymes were analyzed as reported earlier with slight modifications. The 3 ml reaction mixture for SOD consisted of 50 mM phosphate buffer (pH 7.8), 195 mM Methionine, 20 mM riboflavin, 100 µM EDTA-Na2, 750 µM NBT, and 0.2 mL of the crude enzyme. The reaction system was illuminated for 30 min at 72 µmol m− 2 s− 1, and then, the absorbance at 560 nm was recorded. For POD, the reaction mixture included 20 mM guaiacol, 100 mM PBS (pH 6.0), 40 mM H2O2, and 0.2 mL of enzyme extract; its mean change in absorbance per minute was measured at 470 nm. The CAT (EC:1.11.1.6) reaction process involved 0.1 ml of the crude enzyme solution mixed with 1.9 ml of the 50 mmol/L phosphate buffer (pH 7.4) and 1 ml of 45 mmol/L H2O2. After sufficient mixing, the absorbance at 240 nm was measured using a spectrophotometer (TU-1901, Persee general instrument Co. Bei**g.) The reaction system using distilled water instead of the crude enzyme served as a control. Measurements were taken at 1-minute intervals for 3 min, and a decrease of OD240 by 0.01 in 1 min was defined as one unit (U) of enzyme activity.

The enzymatic activities of PEPCase (Phosphoenolpyruvate carboxylase, EC:4.1.1.31) was determined as follows. The crude enzyme was derived from 0.2 g of fresh leaves using an extraction buffer that comprised 0.1 M Tricine-HCl (pH 8.4), 10 mM MgCl2, 1 mM EDTA, 7 mM β-mercaptoethanol, 5% glycerol (v/v), and 1% PVP, and then centrifuged at 15,000 × g, at 4 °C, for 10 min. A 2.5 ml reaction system was developed for gauging PEPCase activity. This involved 100 mmol Tris-HCl (pH 9.2), 10 mmol MgCl2, 10 mmol NaHCO3, 0.16 mmol NADH, 0.5 mmol PEP, malate dehydrogenase (15 U), and 0.5 ml of the crude enzyme. After maintaining a constant water bath at 28 °C for 10 min, the reaction was activated by PEP. Eventually, the activity of PEPCase was assessed via the decreasing rate of NADH at 340 nm.

RNA isolation and quantitative real-time PCR

The extraction of mRNA was undertaken by applying the Plant Total RNA Purification Kit’s instructions (Gmbiolab. Co., Ltd, Taiwan) to 0.1 g of fresh leaf, which was later reverse transcribed to cDNA utilizing the Hifair™ II 1st Strand cDNA Synthesis Kit (YEASEN, Shanghai, China). The gene sequences utilized in this study were obtained from transcriptome sequencing data, and the primer sequences, designed by the Oligo7 software, are provided in Additional File 1.

Each reaction system, consisting of 2 µl cDNA template, 10 µl of SYBR Green master mix with low Rox (Yeasen, China), 0.5 µl of forward primers, 0.5 µl of reverse primers, and 7 µl of nuclease-free water, totalled 20 µl. The ABI Quantstudio 6 Flex real-time PCR system (Applied Biosystems, Foster City, CA) was employed to operate the qRT-PCR protocol, which included conducting a melting curves inspection at the end of each reaction. The ACTIN gene served as the reference, and five replications were made for each reaction.

Statistical analysis

Each assay was carried out at least five times independently. ANOVA with Student-Newman-Keuls was used to determine the significance of the differences, with a P-value of less than 0.05 considered statistically significant. The dots or bars in the graph represent the mean plus or minus the standard deviation. The levels of significance were represented by a series of letters (a, b, c, d).