Acclimatization of photosynthetic apparatus and antioxidant metabolism to excess soil cadmium in Buddleja spp.

Heavy metal (HM) pollutants can cause serious phytotoxicity or oxidative stress in plants. Buddleja L., commonly known as “butterfly bushes”, are frequently found growing on HM-contaminated land. However, to date, few studies have focused on the physiological and biochemical responses of Buddleja species to HM stress. In this study, potted seedlings of B. asiatica Lour. and B. macrostachya Wall. ex Benth. were subjected to various cadmium (Cd) concentrations (0, 25, 50, 100, and 200 mg kg−1) for 90 days. Both studied Buddleja species showed restricted Cd translocation capacity. Exposure to Cd, non-significant differences (p > 0.05) were observed, including quantum yield of photosystem II (PSII), effective quantum yield of PSII, photochemical quenching and non-photochemical quenching in both species between all studied Cd concentrations. Moreover, levels of cellular reactive oxygen species (ROS) significantly declined (p < 0.05) with low malondialdehyde concentrations. In B. asiatica, high superoxide dismutase and significantly enhanced (p < 0.05) peroxidase (POD) activity contributed greatly to the detoxification of excess ROS, while markedly enhanced POD activity was observed in B. macrostachya. Additionally, B. macrostachya showed higher membership function values than did B. asiatica. These results suggested that both Buddleja species exhibited high Cd resistance and acclimatization.


Results
Cd uptake, transfer, and bioaccumulation. The Cd concentrations, the calculated translocation and bioconcentration factors in B. asiatica and B. macrostachya are listed in Table 1. After being cultivated for 90 days, the amount of Cd that had accumulated in both Buddleja species was slightly higher when the plants had been grown on increased soil Cd concentrations. In this study, large amounts of Cd were taken up by the roots, 18.17 ± 0.89-75.94 ± 9.71 and 8.54 ± 0.03-184.31 ± 0.27 mg kg −1 DW in B. asiatica and B. macrostachya, respectively (Table 1). In B. asiatica, the roots accumulated 75.94 ± 9.71 mg kg −1 DW following the 200 mg kg −1 Cd treatments, but only 26.15 ± 1.52 and 28.23 ± 1.18 mg kg −1 DW Cd had accumulated in the stems and leaves, respectively (Table 1). Following the 200 mg kg −1 Cd treatment, B. macrostachya had accumulated 184.31 ± 0.27, 32.11 ± 1.89, and 5.43 ± 0.03 mg Cd kg −1 DW in the roots, stems, and leaves, respectively ( Table 1). The amounts of accumulated Cd in plant tissues (including root, stem and leaf) markedly increased with increasing soil Cd concentrations (Table 1). In both study species, the maximum concentrations of Cd in leaves were found following the 50 mg kg −1 Cd treatment, with 29.25 ± 2.34 and 6.54 ± 0.13 mg Cd kg −1 DW in B. asiatica and B. macrostachya, respectively (Table 1).
To evaluate the Cd accumulation capacity of the two Buddleja species, the bioconcentration factor (BCF) and translocation factor (TF) were calculated (Table 1). At soil treatments of 25-200 mg Cd kg −1 , both B. asiatica and B. macrostachya showed low BCFs, ranging from 0.12 ± 0.00-5.86 ± 0.01 (Table 1). Moreover, the roots of B. macrostachya had greater BCFs than those of B. asiatica, while, B. asiatica shoots presented greater BCFs than those of B. macrostachya at the same Cd concentrations (Table 1). In B. asiatica, the shoots had relatively greater BCFs than their roots; in contrast, B. macrostachya roots had higher BCFs than did the shoots (Table 1).  (Table 1). At soil Cd concentrations of 50 mg Cd kg −1 and higher, neither of our study plant species were able to uptake, accumulate, or translocate Cd to the same extent as hyperaccumulators with low TFs (< 1.0). However, at soil Cd concentrations of 50 mg kg −1 and less, the TFs values in B. asiatica plants were greater than one, ranging from 1.91 ± 025-1.20 ± 0.18 (Table 1). These, however, dramatically declined (p > 0.05) with increasing soil Cd concentrations.

Photosynthetic activity.
To assess the effect of soil Cd on Buddleja photosynthesis, leaf chlorophyll a fluorescence was determined in our two Buddleja study species, B. asiatica and B. macrostachya (Fig. 1). Both species showed similar response patterns to different Cd concentrations within a species, while there were divergent patterns between species (Fig. 1). Both B. asiatica and B. macrostachya showed great quantum yield of PSII (Fv/Fm), 0.86 ± 0.02-0.87 ± 0.02 and 0.77 ± 0.01-0.78 ± 0.01, respectively (Fig. 1a). Following exposure to soil Cd, though B. asiatica demonstrated a larger effective quantum yield of PSII (ΦPSII) and greater photochemical quenching (Qp) than did B. macrostachya with significant differences (p < 0.05), the two species showed similar response patterns to increasing Cd concentrations (Fig. 1b,c). Furthermore, there were no significant differences (p > 0.05) between treatments with different soil Cd concentrations at a species level.
Following exposure to Cd, B. asiatica plants showed lower non-photochemical quenching (NPQ), quantum of regulated energy dissipation [Y(NPQ)], and quantum yield of non-regulated energy dissipation [Y(NO)] than that in B. macrostachya (Fig. 1d-f). Furthermore, there were significant differences (p < 0.05) in NPQ, Y(NPQ) and Y(NO) between the two studied Buddleja species at all investigated soil Cd concentrations. However, Cd exposure did not increase their NPQ, Y(NPQ) and Y(NO) levels, with only nonsignificant differences (p > 0.05).

Membrane lipid peroxidation.
After exposure to Cd for 90 days, seedlings of both Buddleja species survived and appeared to be growing normally. To evaluate the membrane lipid peroxidation, we measured the malondialdehyde (MDA) content in the leaves (Fig. 2a). There were significant differences (p < 0.05) in the leaf MDA concentrations between B. asiatica and B. macrostachya at all concentrations of Cd studied (Fig. 2a). In our study, B. macrostachya leaves showed consistently lower MDA concentrations (always lower than 0.002 μmol g −1 ) than that did B. asiatica leaves, which had MDA concentrations ranging from 0.36 ± 0.02 to 0.49 ± 0.01 μmol g −1 (Fig. 2a).
The two Buddleja species showed different patterns of MDA response to increasing soil Cd concentrations ( Fig. 2a). At Cd concentrations up to 100 mg kg −1 , the concentration of MDA in B. asiatica leaves significantly decreased (p < 0.05) with increasing soil Cd concentrations (Fig. 2a). However, leaf MDA concentrations in these plants markedly increased up to a relatively high levels (200 mg Cd kg −1 ) of Cd (Fig. 2a). However, the leaf MDA concentrations in B. macrostachya leaves remained at a low level with no significant differences following increasing soil Cd concentrations (p > 0.05; Fig. 2a).

Antioxidant enzyme activities.
To evaluate the antioxidant enzyme defense system in our two Buddleja study species in response to different soil Cd concentrations, we measured the activities of several enzymatic antioxidants in the leaves (Fig. 2b-d). Following exposure to Cd, B. asiatica and B. macrostachya showed different response patterns in SOD and POD activities (Fig. 2b,c). Compared with B. macrostachya (62.39 ± 5.05 to 96.17 ± 4.71 U g −1 ), B. asiatica presented high SOD activity (307.65 ± 3.94-391.69 ± 17.92 U g −1 ) at all investigated soil Cd concentrations (Fig. 2b). Moreover, SOD activity significantly declined (p < 0.05) with increasing soil Cd concentrations in B. asiatica species, besides the markedly increase at 100 mg Cd kg −1 (Fig. 2b). The highest SOD activity occurred following the 100 mg Cd kg −1 and control treatments, 391.69 ± 17.92 and 388.83 ± 33.78 U g −1 , respectively (Fig. 2b). In contrast, B. macrostachya showed much lower SOD activity (p < 0.05) compared to B. asiatica at all Cd concentrations. Furthermore, there were no significant differences (p > 0.05) between any of the investigated Cd treatments with increasing soil Cd concentrations (Fig. 2b). SOD activity in B. macrostachya peaked following the 25 mg Cd kg −1 treatment, with a value of 96.17 ± 4.71 U g −1 (Fig. 2b).
In contrast, POD activity responded very differently to increasing Cd concentrations than did SOD activity in both B. asiatica and B. macrostachya (Fig. 2b,c). After exposure to soil Cd, B. macrostachya showed greater POD activity than B. asiatica at all Cd concentrations, with values ranging from 943.75 ± 168.14-2,002.92 ± 253.57 U g −1 min −1 and 57.87 ± 1.63-137.24 ± 1.37 U g −1 min −1 , respectively (Fig. 2c). Up to 200 mg Cd kg −1 , POD activity in B. macrostachya showed a gradual increase with increasing soil Cd concentrations, but the differences between all Cd treatments were nonsignificant (p > 0.05). In B. asiatica, the POD activity increased at low Cd concentrations (25 mg Cd kg −1 ) with significantly (p < 0.05) differences compared to the control (Fig. 2c). However, POD activity declined significantly (p < 0.05) following exposure to Cd concentrations above 25 mg kg −1 (Fig. 2c).
After exposure to low concentrations of Cd, both B. asiatica and B. macrostachya presented low catalase (CAT) activities with a similar response pattern (Fig. 2d), and at Cd concentrations up to 50 mg kg −1 , CAT activity was similar in both between and within species with nonsignificant differences (p < 0.05; Fig. 2d). However, following exposure to high soil Cd concentrations (100 and 200 mg Cd kg −1 ), B. asiatica plants showed high CAT activities with nonsignificant differences (p < 0.05), while those in B. macrostachya markedly decreased (Fig. 2d).

Membership function.
To evaluate the Cd resistance and accumulation characteristics in B. asiatica and B. macrostachya, a membership function based on all the investigated physiological and biochemical parameters was calculated (

Discussion
Phytoremediation is considered to be an excellent choice for environmental HM management. Plant species have evolved different mechanisms to cope with HM exposure, either acting as contaminant accumulators or excluders 27 . The HM accumulation capacity of hyperaccumulator and hypertolerant plant species has been widely discussed 5,11,17,23,28 . In our present study, both Buddleja species extracted and accumulated large amounts of Cd in the roots, but not in the stems or leaves (Table 1). Moreover, the Cd concentrations in all studied plant tissues markedly increased with increasing soil Cd concentrations. According to the standard of Sun et al. 17 , the Cd concentrations in neither roots nor shoots in either Buddleja species reached the criterion to qualify as  www.nature.com/scientificreports/ a Cd-hyperaccumulator. Thus, our results suggest that neither of our studied Buddleja species would qualify as Cd hyperaccumulators.
To further evaluate the potential for Cd phytoextraction and translocation, translocation factor (TF) values, which can estimate the efficiency of translocation of HMs from plant roots to shoots have often been used 17,28,29 . In our study, B. asiatica presented greater TF values than B. macrostachya at all studied Cd concentrations (Table 1). Furthermore, at low soil Cd concentrations (e.g., 25 and 50 mg kg −1 ), B. asiatica showed a high TF that was larger than one, but this was not the case in B. macrostachya (Table 1). This is consistent with Waranusantigul et al. 26 , in which B. asiatica showed high Cd translocation potential with large TF values. In addition, BCF plays a crucial role in evaluating the potential for phytoextraction of a given species 17,23,29 . BCF is a measure of the capacity of plants to extract HMs from the soil and transfer them to plant tissues. Although all our data imply that neither of our Buddleja study species meets the criteria to qualify as Cd-hyperaccumulators, they still have great potential in HM-contaminated land as typical woody pioneer plants with great soil Cd hypertolerance.
Cd-induced phytotoxicity or redox reactions often causes the damage to the photosynthetic apparatus [6][7][8]23 . In angiosperms, the damage to the photosynthetic apparatus causes an increase in the Fo value and a decline in the Fv/Fm ratio 30 . The Fv/Fm ratio, which usually ranges between 0.75 and 0.85, has been used as an efficiency indicator to evaluate abiotic stress on PSII [30][31][32] . In the present study, although the Fv/Fm ratio declined slightly with increasing soil Cd concentrations, both B. asiatica (0.86 ± 0.02 to 0.87 ± 0.02) and B. macrostachya (0.77 ± 0.01 to 0.78 ± 0.01) continued to show high Fv/Fm ratios (Fig. 1a). Moreover, no significant differences (p > 0.05) were found between different Cd treatments within each species. This suggested that both B. asiatica and B. macrostachya performed well when grown in artificially Cd polluted soil. Furthermore, this implies that both studied Buddleja species showed good Cd resistance and tolerance.
Chlorophyll fluorescence competes for excitation energy with photochemical processes, and it can efficiently reflect the leaf photochemical efficiency and growth adaptability of plant species to environmental stresses 22,24 . ΦPSII and Qp are often used to account for practical photosynthetic performance and photosynthetic activity, respectively 22,32 . Compared to B. macrostachya, B. asiatica plants exhibited higher ΦPSII and Qp with significant differences (p < 0.01) at all Cd treatments (Fig. 1b,c). Moreover, there was a positive correlation between ΦPSII and Qp in both Buddleja species (Fig. 1b,c). However, no significant differences (p > 0.05) were found between the different Cd treatments within each species. Wu et al. 23 found that Cd-hypertolerant plants or hyperaccumulators may either show an increase in plant net biomass, or exhibit an improved photosynthetic capacity after exposure to Cd. Thus, all our data suggest that Cd exposure did not reduce either plant photosynthetic activity or performance in either Buddleja species. Furthermore, both B. asiatica and B. macrostachya plants presented great photochemical light energy utilization.
Plant species not only can utilize most of the energy in light for photosynthesis, but also can efficiently dissipate the excessive excitation energy into harmless heat 33,34 . The extent of heat dissipation is closely related to the NPQ capacity 35 . The efficiency of the dissipation of excessive excitation energy into harmless heat is also reflected by Y(NPQ) 33,34 . A high Y(NPQ) indicates that the absorbed light energy is excessive and that the excessive excitation energy is efficiently dissipated into harmless heat 33 . In our study, B. macrostachya presented greater NPQ and Y(NPQ) capacities compared to B. asiatica plants, with significant differences (p < 0.01) between the two species (Fig. 1d,e). Nonsignificant differences (p > 0.05) were found between different Cd treatments within each species. This suggested that Cd exposure did not reduce the capacity to dissipate excessive excitation energy in either Buddleja species. Moreover, B. macrostachya presented a greater photoprotection capacity than did B. asiatica species.
However, not all PSII energy dissipation can be regulated through NPQ 33,34 . For instance, the non-regulated Y(NO) is made up of nonphotochemical quenching due to photoinactivation and constitutive thermal dissipation 35 . Generally, the value of Y(NO) has been found to be positively associated with the extent of photodamage 33,34 In this study, Cd exposure did not significantly (p > 0.05) increase the values of Y(NO) in both studied Buddleja species under any of the Cd treatments. Compared with B. asiatica, B. macrostachya plants exhibited high Y(NO) values (as NPQ and Y(NPQ)) (Fig. 1f). This implied that exposure to Cd did not increase the extent of photodamage in either of the Buddleja species, and that their protective regulatory mechanisms were efficient. Furthermore, our results suggested that both B. asiatica and B. macrostachya showed high tolerance and acclimatization to artificial soil Cd stress.
Environmental stress often causes oxidative damage and can lead to the excessive production of ROS harming the cell membrane 13,36 . Both stressed and unstressed plant cells might produce ROS, but HM-tolerant plants generally have a well-developed antioxidant system for the removal of ROS 11,23,36 . Malondialdehyde, a product of lipid peroxidation has been used as a key indicator for the determination of oxidative damage in plants 8,11,23,36 . In this study, leaves from both Buddleja study species produced low amounts of MDA following treatment with elevated Cd concentrations (Fig. 2a). This implies that both studied Buddleja species can efficiently activate their antioxidant defense systems and quench free radicals. Furthermore, these efficient ROS scavenging mechanisms protect Buddleja species from destructive reactions under soil Cd stress.
Plants have evolved diverse mechanisms and various protective antioxidant defense systems to eradicate oxidative stress 16,18,20,21 . Antioxidants' response against Cd toxicity varies among different plant species and experimental conditions 17,19,23,36,37 . Enzymatic antioxidants are predominantly produced in the sensitive foliage and protect plants from damage by quenching free radicals. Compared with Cd-sensitive plants, Cd-tolerant plants response positively in antioxidant enzyme activities to Cd-induced stress 38 16,17 . In this study, both two studied Buddleja species can efficiently regulate the steady-state level of cellular ROS through different antioxidant mechanisms. After exposure to soil Cd stress, B. asiatica maintained high SOD levels at all soil Cd concentrations, while high POD levels were maintained in B. macrostachya (Fig. 2b,c). Buddleja asiatica Scientific Reports | (2020) 10:21439 | https://doi.org/10.1038/s41598-020-78593-8 www.nature.com/scientificreports/ markedly reduced the activities of SOD and POD following exposure to Cd, however, POD activity significantly improved (p < 0.01) following the 25 and 50 mg Cd kg −1 treatments (Fig. 2b,c). In contrast, POD activity in B. macrostachya gradually increased following exposure to soil Cd, and a stable low level of SOD activity was maintained throughout (Fig. 2b,c). All our results suggest that both our Buddleja study species were equipped with efficient mechanism to manage cellular redox homeostasis at its optimum after exposure to soil Cd stress. These enzymatic antioxidants improved plant tolerance to soil Cd stress, and preserved normal plant growth and metabolism in these species following increasing soil Cd concentration.
In addition, to evaluate overall plant resistance and acclimatization to environmental stresses, the comprehensive membership function value (D) was also calculated 39,40 . The D value is used to identify Cd-resistant plants, where a large D value is positively correlated with high resistance and acclimatization 39,40 . After exposure to soil Cd stress, both B. asiatica and B. macrostachya exhibited not only different response mechanisms and growth performance, but also excellent acclimatization with high D values (Table 2). Combined with their capacities to accumulate Cd in their tissues, our results suggest that B. asiatica and B. macrostachya show species-specific responses in their resistance and acclimatization to Cd biotoxicity. Moreover, they have efficient ROS detoxification mechanisms in response to soil Cd stress, and are potentially Cd-hypertolerant plants.

Conclusion
Buddleja asiatica and B. macrostachya plants growing in Cd-contaminated soil presented high tolerance to Cd stress with limited Cd accumulation capacity. In neither Buddleja species did Cd exposure affect their photosynthetic activity or photoprotection capacity. On the contrary, levels of ROS significantly declined, the plants preserving cellular redox homeostasis through different enzymatic antioxidant mechanisms. In B. asiatica, SOD and POD activities contributed greatly to the detoxification mechanism in response to oxidative stress, while B. macrostachya relied only on POD activity. Moreover, both species showed great Cd tolerance and resistance, and both are therefore potential Cd-hypertolerant plants for use in phytoremediation.

Cultivation of plants and soil treatments.
This study was carried out between April 2017 and February 2019 in the greenhouses of Honghe University, Yunnan, China. Cuttings were taken from a mature B. asiatica specimen and biennial seedlings of B. macrostachya were used. Before the experiments, all individuals (about 40) were collected and cultivated in a greenhouse at Honghe University for 1 week. Twenty-five individuals of each species, having similar heights and growth vigor were then selected for use in the experiments.
Soil was collected from the base of the mature B. asiatica specimen and was used in each of the experiments. The organic carbon content of the soil sample was 32.85 ± 8.18 mg g −1 , the total nitrogen was 0.34 ± 0.10 mg g −1 , and the total phosphorous was 0.20 ± 0.02 mg g −1 . The soil sample was slightly Cd-contaminated, and Cd content was ~ 4.60 mg kg −1 . Soil samples were air-dried and sieved through a 2 mm mesh, then subpackaged into separate bags with 10 kg in each bag. The levels of Cd in the soil were then adjusted to 0, 25, 50, 100, and 200 mg kg −1 DW using Cd supplied as Cd(NO 3 ) 2 ·4H 2 O. All soil samples were then watered with distilled water and incubated in bags for 20 days with ~ 60% humidity 41 . Plants were then transplanted into the soil and were cultivated under normal greenhouse conditions for 90 days.
Determination of Cd concentration. Following exposure to Cd for 90 days, the Cd concentrations in different plant tissues were investigated. Plant samples were washed clean of soil and were rinsed with deionized water. Subsequently, plant roots, stems, and leaves were separately oven-dried at 70 °C for a week. The dried tissues were weighed, ground, and sieved through a 2 mm stainless steel mesh.
Plant tissues (roots, stems, and leaves) were separately microwave digested using the Speedwave ENTRY system (Berghof, Germany) with a mixed solution of HNO 3 + HClO 4 (v/v) = 3:1. The Cd content was determined using Flame Atomic Absorbance Spectrometry (FAAS; TAS-990, Beijing Purkinje General Instrument Co., Ltd, China) and a standard Cd solution (GSB04-1721-2004) provided by the National Center for Reference Materials.

Evaluation of Cd translocation potential. To evaluate the potential of the test plants to bioaccumulate
Cd, the translocation factor (TF) and bioconcentration factor (BCF) indexes were further calculated following Liu et al. 29 : Evaluation of photosynthetic activity. Chlorophyll a fluorescence traits were recorded using a Li-6400XT with a fluorescent leaf chamber (LiCOR Inc., USA). Ten mature leaves from each treatment were selected for the measurement of chlorophyll a fluorescence and were marked. After being dark-adapted for 30 min, Fo and Fm were recorded. The ratio of variable to maximal fluorescence [Fv/Fm; Fv/Fm = (Fm − Fo)/ Fm], which characterizes the quantum yield of photosystem II, was determined following Figueroa et al. 31 .
The following day (9:00-11:00), chlorophyll a fluorescence parameters were measured using the same leaf. Before the test, all marked leaves were light-adapted with natural light for half an hour. A saturating light pulse (1200 µmol quanta m 2 s −1 , 1 s) was applied for closing all reaction centers. Then Fo′, Fm′, and Fs were recorded 5  Determination of lipid peroxidation. To evaluate the extent of lipid peroxidation of plants, the concentrations of malondialdehyde (MDA) in the leaves were measured following Sun et al. 17 . The seventh or eighth healthy mature leaves from the apex of the branches were used for analysis. About 0.5 g of fresh leaf sample was ground and dissolved in 10 ml cold trichloroacetic acid (TCA; 10%). The homogenate was then centrifuged for 20 min at 4000 rpm and 4 °C for 10 min. The liquid supernatant was taken and used for analysis. Three replicates for each Cd treatment were performed.
Assay of antioxidant enzymes. To evaluate the antioxidant enzyme defense system in the study plants, the antioxidant metabolism was investigated. The seventh or eighth healthy mature leaves from the apex of the branches were used for analysis. About 0.2 g of fresh leaf tissues was ground and dissolved in 10.0 ml of cold phosphatic buffer solution (PBS) at PH 7.0. The homogenate was then centrifuged at 4000 rpm for 15 min. The supernatant liquid was taken and used for analysis. Superoxide dismutase (SOD), peroxidase (POD), and catalase (CAT) activities were subsequently determined following Sun et al. 17 and Sidhu et al. 11 . Three replicates for each Cd treatment were performed. The SOD activity was determined by measuring the inhibition in the photoreduction of nitroblue tetrazolium (NBT). The POD activity was evaluated using the guaiacol oxidation assay. The CAT activity was determined by monitoring the disappearance of H 2 O 2 .

Statistical analyses.
To evaluate the performance and acclimatization of plants under different Cd stresses, the membership function value of Cd resistance in each Buddleja species was also calculated using the modified equations following Liu et al. 39 as follows: where u(X j ) is the membership function value for adaptability of the trait (j); X j , X min , and X max are the means, minimum and maximum values of the trait (j), respectively; P j is the contribution rate of the trait (j) based on principal component analysis; W j is rate of the contribution of the trait (j) in all studied traits, and D is the comprehensive membership function value for the adaptability of the species. One-way analysis of variance (ANOVA) was performed using PAST version 2.0 to reveal any differences between species or different Cd treatments. Fisher's least significant difference (LSD) tests were performed using PAST version 2.0 44 to analyze statistically significant differences among different samples at the level of 0.05 and 0.01. A significant difference was considered at two different levels (p < 0.05 or < 0.01). Unless indicated, data presented represent the mean value plus or minus the standard error (± SE). To reduce the heterogeneity of variances, the data were log 10 -transformed, if necessary. All bar charts were created by using ORIGIN (Version 2019b, OriginLab Corporation, USA) and then adjusted and assembled using Adobe Illustrator CS4 software (Adobe Systems, San Jose, CA) as shown in Figs. 1 and 2. (1) u X j = X j − X min (X max − X min ) , j = 1, 2, . . . , n (2) W j = P j / n j=1 P j , j = 1, 2, . . . , n (3) D = n j=1 u(X j ) × W j , j = 1, 2, . . . , n