Uptake, accumulation and metabolic response of ferricyanide in weeping willows

Abstract

The remediation potential and metabolic responses of plants to ferricyanide were investigated using pre-rooted weeping willows (Salix babylonica L.) grown hydroponically in growth chambers and treated with potassium ferricyanide. Positive responses were observed for the plants exposed to ≤ 274.13 mg CN L⁻¹ as ferricyanide, exhibiting higher chlorophylls and soluble proteins compared with the controls. Visible toxic symptoms were only noted for the treatment exposed to 506.67 mg CN L⁻¹ after 120 h of incubation. Activity of superoxide dismutases (SOD) in leaves showed a slight change to ferricyanide exposure in most treatments. Catalase (CAT) and peroxidase (POD) activities were negatively correlated to the concentrations of ferricyanide. Of all the selected parameters measured, soluble proteins of plants were the most sensitive to ferricyanide, showing a significant linear correlation (R² = 0.952). Between 6.90 and 12.66% of the applied ferricyanide were removed by plants from the hydroponic solution at different treatments over the 192 h of exposure. Small amounts of the applied chemical taken up from the hydroponic solutions were detected in all parts of plant materials: the highest concentration was associated with roots in all treatments, followed by stems; the lowest was observed in leaves. The mass balance analysis showed that the total cyanide recovered in plant biomass was constant in all treatments, indicating that transport is a major limiting step for the uptake of ferricyanide by plants. The majority of the ferricyanide taken up from the growth media was possibly assimilated during transport through plants. The velocity of the removal processes can be described by Michaelis-Menten kinetics, and the half-saturation constant (K_M) and the maximum removal capacity (v_max) were estimated to be 228.1 mg CN L⁻¹ and 36.43 mg CN kg⁻¹d⁻¹, respectively, using non-linear regression methods. These results suggest that weeping willows can take up, transport and assimilate ferricyanide; and phytoremediation is an option for cleaning up the environmental sites contaminated with cyanide complexes.

---

1 Introduction

Cyanide (CN⁻) is a common chemical used in extraction of ores, metallurgy and the photographic industry¹ and frequently found in industrial wastes. More than 100 000 tons of cyanide out of the annual 1.4 million tons produced enter the environment² through industrial wastewater and related activities on an annual basis. The lethal dose of sodium cyanide to human (Homo sapiens) is estimated to be 3.0 mg NaCN kg⁻¹ body weight.³ Indeed, complete destruction of cyanide released into the environment is necessary to protect the ecological health of the environment. Conventional methods available for elimination of CN⁻ include physical adsorption, biological accumulation, and chemical oxidation involving chlorine, H₂O₂, ozone, ferrate(IV), TiO₂ or SO₂ in the presence of an appropriate catalyst.^4–6 These techniques are often being limited by the required stringent conditions and also by the effectiveness of the treatment processes for application of in situ contamination treatment in complex matrices, including groundwater and sediment.

Cyanide concentration is generally very low in the environment because free CN⁻ reacts readily with metal ions available to form stable complexes, e.g., ferrocyanide Fe^II(CN)₆⁴⁻ and ferricyanide Fe^III(CN)₆³⁻.^7,8 These dominant chemical forms of cyanide complexes can be > 97% of the total cyanide in soil and groundwater⁹ and contaminated sites may contain >1% complexed cyanides on a dry weight basis.¹⁰ Photo-decomposition of cyanide complexes to free cyanide can be achieved¹¹ and phytoremediation of free cyanide by a number of plants from three continents and climate zones^12–18 has shown the great potential of in situ bioremediation of sediment and groundwater contaminated with cyanide and/or cyanide complexes by plant metabolism. Phytotoxicity of cyanide was also determined,¹⁵ in which KCN dissociates almost completely in water and was toxic to willows at CN ion concentrations above 20 mg L⁻¹.

More than 1000 plant species are cyanogenic and possess detoxifying enzyme systems for cyanide metabolism^14,16 and the enzyme β-cyanoalanine synthase (CAS)^19,20 catalyzes the biochemical reaction to produce asparagine as the only metabolic product when ¹⁴C⁻labeled ¹⁴CN⁻ was used in a study including both cyanogenic and non-cyanogenic plants.²¹ Beavis and Vercesi²² described a mitochondrial anion channel capable of transporting ferrocyanide and the main physiological role of CAS, which was exclusively located in mitochondria, was the detoxification of toxic cyanide.²³ It has been found that iron and cyanide were taken up together as a chemical complex by willows exposed to ferrocyanide and ferricyanide, and cyanide was liberated from the complex in plant materials and was metabolized.²⁴ However, little information is available about phytotoxicity of ferricyanide to terrestrial plants. Therefore, the objectives of this study were to assess the assimilation of ferricyanide by weeping willow and to determine the associated oxidative stress responses by enzymes of plant leaves under different concentrations of ferricyanide treatments.

Materials and methods

Trees specimens and exposure regimes

Weeping willows (Salix babylonica L.) grown on the campus of Hunan Agricultural University, P.R. China, were used in this investigation. Tree cuttings of 40 cm in length were removed from a matured tree and all cuttings used in this study were obtained from a single tree. They were then placed in buckets of tap water kept at room temperature of 15 to 18 °C under natural sunlight until new roots and leaves appeared. After a two-month period of growth, each young rooted cutting was transferred to a 250 mL Erlenmeyer flask filled with approximately 200 mL modified ISO 8692 nutrient solution (Table 1), which was prepared from reagent grade chemicals. The flasks were all sealed with cork stoppers and silicon sealant (Dow Chemical Co, Midland, Michigan) between the tree stem and the flask to prevent escape of water or chemicals, and wrapped with aluminium foil completely to prevent the potential growth of algae. For each treatment concentration, nine replicates were prepared. All flasks were housed in a climate control chamber kept at a constant temperature of 24.0 ± 1 °C under natural sunlight (light: dark cycle 14:10 h). The plants were conditioned for 48 h first to allow adaptation to the new environmental conditions. Then, the weight of each plant-flask system was measured and recorded. The flasks including the tree cuttings were weighed again after 24 h. By doing this way, the transpiration rate was calculated for each flask. Trees with a similar transpiration rate were selected and grouped in a treatment of the tests. The nutrient solution in each flask was replaced by a spiked solution, except for the controls. Ferricyanide used was in the form of potassium ferricyanide [K₃Fe(CN)₆] of analytical grade with ≥ 95% purity. It should be noted that 1 mg K₃Fe(CN)₆ equals to 0.474 mg CN. Several concentrations (70.10, 137.46, 274.13, 410.26 and 506.66 mg CN L⁻¹) were prepared. Two sets of controls were made: One control was with ferricyanide, but no plant cuttings to quantify the effects of loss during handling, volatilization, hydrolysis and/or degradation by microorganism; the other was with trees in the nutrient solution without addition of ferricyanide to quantify the transpiration rate of the non-exposed control trees. It should be noted that all concentrations of ferricyanide complexes and of total cyanide in this study are referred to as mg CN L⁻¹.

Table 1 Composition of the modified ISO 8692 nutrient solution used in this study

Macronutrients/µmol L^−1		Micronutrients/nmol L^−1
NaNO3	2823.9	H3BO3	2992.1
MgCl2·6H2O	59.0	MnCl2·4H2O	2097.0
CaCl2·2H2O	122.4	ZnCl2	22.0
MgSO4·7H2O	60.9	CoCl2·6H2O	6.3
KH2PO4	246.0	CuCl2·2H2O	0.1
NaHCO3	1785.5	NaMoO4·2H2O	28.9

---

The toxic effect was quantified by measuring the transpiration rate of the young rooted trees in flasks. The transpiration rate of plants is coupled to the photosynthesis, and an inhibition of transpiration is a reliable and quick measurement of toxic responses by plants.²⁵ The weight loss of the plant-flask system was expressed as the transpiration rate.

Chlorophyll measurement

The chlorophyll content in leaves was determined spectrophotometrically at the end of the experiments (192 h). Plant leaves were cut into small pieces, precisely weighed (0.5 g fresh weight) and placed in 25 mL flasks. Then, 80% acetone was used to fill to the 25 mL volume. Three separate flasks were conducted for each treatment group. All flasks were placed in the dark for 24 h. During this period, flasks were mixed by shaking twice. The absorption of light at 645 and 663 nm was measured in a cell with an optical path of 10 mm against 80% acetone as a blank. The amount of chlorophyll a and chlorophyllb in plant leaves was calculated by the following formulae according to Maclachalam and Zalik:²⁶Where C_a is the concentration of chlorophyll a (mg g⁻¹FW), C_b is the concentration of chlorophyllb (mg g⁻¹FW), D is the optical density (OD) at the specific wave length indicated, V is the final volume (mL), W is the fresh weight of leaf materials (g), and d is the length of the light path in cm.

Enzyme activity measurement

The activities of three antioxidant enzymes SOD, CAT and POD were measured in fresh leaves at the end of the experiment. Fresh leaves were taken from the shoot. Leaves of 0.3 g (fresh weight) was precisely weighed and placed in a tissue grinder. Phosphate buffer solution of 1.4 mL (pH 7.8, containing NaH₂PO₄, Na₂HPO₄, PVPP, EDTA and mercapto-ethanol) was added before grinding. The process was performed in an ice-bath and then centrifuged at 8 000 rpm for 15 min, the supernatant was collected and stored at 4 °C and employed in the enzyme assays. Each enzyme was measured independently. SOD, POD and CAT activities in the leaf cells of plants were determined spectrophotometrically as described previously by Jin and Ding.²⁷

Assay of SOD activity. The reaction mixture (3 mL) was composed of 13 mM methionine, 0.075 mM NBT, 0.1 mM EDTA, 0.002 mM riboflavin, and 0.1 mL of enzyme extract in 50 mM phosphate buffer (pH 7.8). The mixture in the test tube was placed on a rotating tube holder at 25 °C for 10 min. The absorbance was measured spectrophotometrically at 550 nm in a cell with an optical path length of 10 mm against the reaction mixture without enzyme extract. The unit of SOD activity (U g⁻¹FW) was defined as the amount of enzyme, which caused 50% inhibition of the initial rate of reaction in the absence of the enzyme.

Assay of CAT activity. The enzyme extract (0.1 mL) was added to a 2 mL assay mixture (50 mM Tris-HCl buffer pH 6.8, containing 5 mM H₂O₂). The reaction was terminated by adding 0.1 mL of 20% titanic tetrachloride after incubation for 1 min at 25 °C. The absorbance of the reaction solutions was measured at 405 nm spectrophotometrically against the reaction mixture without enzyme extract. One unit of CAT activity (U g⁻¹FW) was defined as the amount of CAT, which decomposed 1 µmol hydrogen peroxide in one minute at 25 °C.

Assay of POD activity. The reaction mixture (3 mL) was composed of 100 mM potassium phosphate buffer (pH 7.0), 20 mM guaiacol, 65 mM H₂O₂ and 0.1 mL enzyme extract. Changes in absorbance were recorded spectrophotometrically at 470 nm against the reaction mixture without enzyme extract for 3 min at 25 °C. One activity unit of POD (U g⁻¹FW) was defined as the amount of enzyme that caused an increase of 0.001 absorbance unit per minute.

Soluble protein measurement

The soluble protein content was determined spectrophotometrically in fresh leaves from the top shoot as described by Jin and Ding.²⁷ Three separate measurements were conducted for each treatment. At the end of the experiments (192 h), 0.5 g of tissue materials (fresh weight) was precisely weighed and placed in a tissue grinder. Phosphate buffer solution of 2.5 mL (65 mM, pH 7.8) containing 0.4% mercapto-ethanol (v/v) was added before grinding. The process was performed in an ice-bath and then centrifuged at 12 000 × g for 15 min. The supernatant was stored at 4 °C before analyzing the soluble protein in leaves. 0.1 mL aliquot of the samples was pippetted into a vessel and 5.0 mL Coomassie Brilliant Blue G-250 solution (Sigma-Aldrich Inc., St. Louis, Missouri) were added. After mixing, the vessel was left standing for 2 min. The absorption of light at 595 nm was measured spectrophotometrically against de-ionized water as a reference. Albumin Bovine V solution from bovine serum (Sigma-Aldrich Inc., St. Louis, Missouri) was used as a standard. The Coomassie Brilliant Blue G-250 solution was prepared as follow: 100 mg Coomassie Brilliant Blue G-250 were dissolved in 50 mL 90% ethanol, followed by 100 mL of 85% phosphoric acid (v/v), and then diluted to 1000 mL with demineralized water as described by Cheung and Gu.²⁸ This solution was finally filtered through 0.45µm membrane filter (Gelman Science, Ann Arbor, Michigan) before use and stored at room temperature with a maximum holding time of one month.

Chemical analysis

Total cyanide in solution. Total cyanide is the sum of free cyanide (HCN and CN⁻) and complexed cyanide. The total cyanide in solution was analyzed with a standard method (State Environmental Protection Administration of P.R. China). Ten millitres of 1% NaOH was added into the absorption vessel of the distillation unit. Five mL of the spiked solution were placed in a 500 mL round bottom flask, and then 200 mL of distilled water was added. Then 10 mL of sodium ethylenediamine tetraacetate (EDTA) with a concentration of 10% (m/v) and 10 mL of phosphoric acid (≥85% purity) were added before heating and mixing. Approximately 100 mL distilled solution containing cyanide from plant materials were collected, quantitatively transferred to a 100 mL volumetric flask and made up to the volume with de-ionized water. The solution was stored at below 6 °C until the concentration of cyanide was determined. The samples were all analyzed within a maximum hold time of 4 h.

One to 5 mL of solution samples were pipetted into a 25 mL colorimetric cylinder (depending on the concentrations of cyanide in solution), and 0.1% NaOH was added to the 10 mL volume. Then 5.0 mL of a buffer solution with potassium dihydrogen phosphate and sodium phosphate were added. Quickly 0.2 mL of 1% (w/v) chloramine-T solution were introduced. The vessel was sealed with a stopper and left standing for 3 to 5 minutes. Five mL of the colour reagent consisting of isonicotinic acid and 3-methyl-1-phenyl-5-pyrazolone were then added. The content was diluted with de-ionized water to the mark (25 mL) and mixed thoroughly. Finally, the colorimetric cylinders were all kept in a water bath at a temperature of 32 °C for 40 min. The absorption of light at 638 nm was measured in a cell of optical path length of 10 mm against a water as reference. All chemicals used were > 99.5% purity, except potassium cyanide and nicotinic acid, which were of technical grade (92–95% purity); but the stock solution and the standard solution of KCN used in this test were calibrated by a standard solution of AgNO₃, which was also calibrated using a standard solution of NaCl (standard method from SEPA, P. R. China). The detection limit of this method was determined to be between 0.004 and 0.25 mg CN L⁻¹, depending on the volume of the sample used. The sample preparation methods used were also checked against spiking samples with calibrated solution standards and the mean recovery was 98.46%.

Total cyanide in plant materials. The total cyanide in plant materials was analyzed according to the method by Yu et al.²⁹ Plant materials from the treated and the non-treated plants were harvested after 192 h of experiments. Fresh plant biomass (2.5 to 15 g FW, depending on the harvested weight of plant materials) cut into small pieces was used instead of 5 mL of the spiked sample. The remaining procedure was identical to those described earlier.

It should be noted that the concentrations of cyanide in the text and figures refer to the total cyanide concentrations, not free cyanide.

Statistical methods

The students t-test (two-tailed) and the Pearson's product-moment correlation and regression were conducted. The significance of the correlations was judged using tabled values for the critical r (degree of freedom n-2, significance level was 0.01 or 0.05) from Sachs.³⁰

Results

Effects of ferricyanide on the transpiration rate of weeping willows

Transpiration rate of weeping willows grown in hydroponic solution spiked with various concentrations of ferricyanide was measured (Table 2). Compared with the non-treated control plants, negligible reduction in transpiration rate was detected for willows exposed to ferrocyanide ≤ 274.13 mg CN L⁻¹. When exposed to higher concentrations of ferricyanide ≥ 410.26 mg CN L⁻¹, the treated willows showed a marked decrease in transpiration rate (> 13%), but the difference was not significant (p > 0.05). Visible toxic symptoms, e.g., chlorosis of leaves, were observed for the treatment group exposed to 506.66 mg CN L⁻¹ after 120 h of exposure.

Table 2 Effects of various ferricyanide treatments on transpiration rate, soluble protein, chlorophyll contents, and activities of SOD, CAT and POD. The exposure period was 192 h. Values are the mean of three replicates for both the treated and non-treated control plants, except the transpiration rate (nine replicates for the treated plants and six for the non-treated plants), numeric values in brackets represent standard deviation, FW = fresh weight

Characteristic	Cyanide concentrations (mg CN L^−1)
	Control	70.10	137.46	274.13	410.26	506.66
a Significantly different to the controls on 95% significance level (two-tailed t-test).
Transpiration rate(g d^−1)	3.28(1.248)	3.20(1.050)	3.22(0.961)	3.21(1.004)	2.85(0.800)	2.72(0.771)
Chlorophyll a(mg g^−1FW)	0.28(0.022)	0.38 (0.051)	0.34(0.082)	0.32(0.024)	0.32(0.079)	0.24(0.034)
Chlorophyll b(mg g^−1FW)	0.28(0.019)	0.36(0.058)	0.36(0.087)	0.34(0.027)	0.33(0.078)	0.24(0.032)
Soluble protein(mg g ^−1FW)	4.82(1.715)	7.31^a(0.689)	6.55^a(0.676)	6.07^a(0.425)	5.69^a(0.253)	5.14(0.976)
Superoxide dismutases(U g^−1FW)	250.22(13.079)	250.61(19.582)	236.49(21.791)	243.92(16.497)	244.14(23.243)	214.92(35.757)
Peroxidase(U g^−1FW)	39.04(5.261)	49.46(3.959)	46.52(7.394)	40.92(8.231)	33.60(2.839)	19.60^a(1.633)
Catalase(U g^−1FW)	195.91(47.869)	197.96(40.763)	175.01(22.468)	177.07(23.707)	121.73(24.529)	127.92(33.371)

---

Effects of ferricyanide on the content of chlorophylls in plant leaves

Chlorophyll contents in the leaves of the plants were affected by ferricyanide application (Table 2). The change of both chlorophyllsa and b due to ferricyanide application followed a similar declining trend. Increases of 35.7 and 28.6% were obtained for chlorophyllsa and b, respectively, at the treatment exposed to 70.10 mg CN L⁻¹ in comparison with the controls, but the difference was not statistically significant (p > 0.05). When increased up to 410.26 mg CN L⁻¹, ferricyanide also produced positive effects on both chlorophylls. At 506.66 mg CN L⁻¹, a reduction of 14.3% was observed for both chlorophylls, and, chlorosis of leaves was also found in the same treatment group after 120 h of exposure.

Effects of ferricyanide on the content of soluble proteins in plant leaves

Table 2 shows the measured soluble protein contents of plants under different ferricyanide treatments after 192 h of exposure. Soluble protein contents in plant leaves were significantly affected by the presence of ferricyanide (p < 0.05) except for the plants exposed to 506.66 mg CN L⁻¹, in which an increase of 6.64% was detected (p > 0.05). Although positive effects of ferricyanide application were observed in all treated plants, a declining trend was apparent over the increase of ferricyanide concentration.

Effects of ferricyanide on the activities of antioxidant enzymes in plant leaves

Activities of SOD, CAT and POD in leaves of weeping willows were also determined at the end of ferricyanide exposure. For non-treated control plants, activity of SOD in leaves was 250.22 U g ⁻¹FW, whereas the activities of SOD for the treatments exposed to ferricyanide ranging from 70.10 to 410.26 mg CN L⁻¹ remained at the same level with an average of 243.79 U g ⁻¹FW (SD 13.844, n = 4) (Table 2). Such a slight difference in the SOD activity between the treated and the non-treated plants was probably due to the differences in individual tree specimens used and fewer free superoxide radicals formed in plant cells. When exposed to 506.66 mg CN L⁻¹, activity of SOD in leaves was severely inhibited with a reduction of 14.22% compared to the controls, but the difference between the treatment and the controls was statistically insignificant (p > 0.05).

Activity of POD in leaves varied with different treatments of ferricyanide to the plants (Table 2). POD activity was inversely proportional to the concentration of ferricyanide up to 274.13 mg CN L⁻¹. The values of POD activities in those treatments were still higher than that of the non-treated control plants (39.04 U g⁻¹FW). At 70.10 mg CN L⁻¹, activity of POD in leaves was 26.69% higher than the controls. With ferricyanide concentrations increased from 410.26 to 506.66 mg CN L⁻¹, activities of POD were strongly inhibited showing a reduction of 13.93 and 49.80% in comparison with the controls, respectively. The difference was significant statistically (p < 0.05).

A declining trend for the activities of CAT in leaves was observed for plants exposed to different treatment concentrations of ferricyanide (Table 2). At 70.10 mg CN L⁻¹, a negligible increase in CAT activity of plants was detected compared with the controls. With a further increase of ferricyanide concentrations, a strong inhibition of CAT activities was observed, particularly at the treatments exposed to 410.26 and 506.66 mg CN L⁻¹ in which approximately 35% reduction of CAT activities in leaves was detected for both treatments compared with the controls. The difference was insignificant (p > 0.05).

Removal of ferricyanide from hydroponic solution by weeping willows

Fig. 1 shows the measured total ferricyanide concentrations in the hydroponic solutions at different treatments after 192 h of exposure in this study. In controls with the absence of plants, no change of ferricyanide concentrations was observed over the entire period of incubation (data not shown), implying the dissociation of ferricyanide did not occur within the designed system in the absence of light. The amounts of applied ferricyanide removed from the hydroponic solution by the presence of weeping willows in all treatments were quantified. Between 6.90 and 12.66% of the applied ferricyanide were removed by plants, but majority of the chemical remained in the hydroponic solutions. A higher removal rate of ferricyanide by basket willows was reported,²⁴ most likely due to the difference in the duration of exposure periods.

Fig. 1 Measured total cyanide concentrations (mg CN l⁻¹) in aqueous solution at different treatments. The exposure period was 192 h. The values are the mean of three replicates. Vertical lines represent standard deviations (I, initial concentration; F, final concentration).

Mass balance of total cyanide

Ferricyanide contents in the various plant components are presented in Fig. 2. Cyanide was detected in plant materials of all treatment groups, confirming uptake and transport of ferricyanide into plants. The background of total cyanide in non-exposed control trees was 0.023 mg CN kg⁻¹ for roots, 0.038 mg CN kg⁻¹ for leaves and 0.015 mg CN kg⁻¹ for stems (n = 2 for all controls), and the concentrations in the respective solutions were below the detection limit of 0.004 mg CN L⁻¹. Total cyanide levels in all treated plant materials at the end of ferricyanide exposure were significantly higher than the controls. Concentrations of total cyanide in different plant materials were strongly correlated to the applied ferricyanide concentrations in the solution, judged by the critical r for a given n (significant at α = 0.05)³⁰ (Fig. 3). However, substantial differences existed in the distribution of cyanide in plant materials after exposure to various concentrations of ferricyanide, indicating transport and assimilation of ferricyanide within plant materials. The highest concentration of total cyanide was observed in roots, followed by stems, and the lowest in leaves of all treatments.

Fig. 2 Measured total cyanide concentrations (mg CN kg⁻¹FW) in root, stems and leaves of weeping willows (Salix babylonica L.) at different treatment concentrations. The exposure period was 192 h. The values are the mean of three replicates. Vertical lines represent standard deviations; FW = fresh weight.

Fig. 3 Correlations between the applied ferricyanide concentrations (mg CN L⁻¹) and the measured total cyanide concentrations (mg Cr kg⁻¹FW) in roots, stems and leaves of weeping willows (Salix babylonica L.); linear regression with R² values added.

A mass balance analysis of total cyanide was made, based on the measurements of the total cyanide in tissues and the solution data (Table 3). Between 6.90–12.66% of the applied ferricyanide was removed from the hydroponic solution by plants after 192 h of exposure, while approximately 0.36% of the total ferricyanide applied (SD 0.073, n = 5) was recovered in plant materials. Most likely, 95.21% of the ferrocyanide (SD 1.985, n = 5) taken up from the hydroponic solution was assimilated during transport within the willows.

Table 3 Mass balance for ferricyanide in the hydroponic systems containing weeping willows (Salix babylonica L.) The exposure period was 192 h. Values are mean of three replicates, in brackets: standard deviation

Treatment(mg CN L^−1)	Mass in solution (mg CN)		Mass in tissues^a (µg CN)			Estimated metabolic rate (mg CN kg^−1d^−1)
	Initial	Final	Root	Stem	Leaf
a The measured concentrations of total cyanide in plant materials were significantly higher than that of the background cyanide in non-exposed control trees. The background cyanide was not subtracted from concentrations in root, stems and leaves for the mass balance. b Significantly different to the controls on 95% significance level (two-tailed t-test).
70.10	17.53	15.48(0.210)	38.09^b(8.823)	23.06^b(9.174)	3.30^b(1.505)	8.17(2.596)
137.46	34.37	30.01(0.642)	45.81^b(12.860)	106.59^b(23.128)	5.49^b(1.444)	12.16(1.624)
274.13	68.53	63.23(0.099)	77.27^b(34.987)	99.61^b(38.986)	4.40^b(0.615)	22.56(1.676)
410.26	102.57	95.99(3.872)	113.701^b(7.683)	227.52^b(67.252)	10.42^b(4.321)	23.84(4.925)
506.66	126.67	118.95(0.914)	142.68^b(52.982)	332.03^b(69.115)	18.23^b(4.733)	23.57(3.201)

---

Determination of enzyme kinetics (K_M and v_max)

The relationship between the assimilation rates of ferricyanide by plants and the applied ferricyanide concentrations was plotted (Fig. 4), showing Michaelis-Menten kinetics, which can be described by the following equation³¹where v (mg kg⁻¹) is the removal rate of the substrate concentration C (mg L⁻¹), v_max (mg kg⁻¹d⁻¹) is the maximal removal velocity, K_M (mg L⁻¹) is the half-saturation constant (the substrate concentration where the disappearance velocity is half the maximum) and M (kg) is the mass of the plant compartment. K_M and v_max for an enzyme reaction can be obtained by the analysis of initial velocities v at a series of substrate concentrations. Then, K_M and v_max can be fitted to the measured v. The most commonly used transformation is the double-reciprocal plot, which was formulated by Hans Lineweaver and Dean Burk.³²

Fig. 4 Estimated ferricyanide metabolic rate (mg CN kg⁻¹FW d⁻¹) of weeping willows (Salix babylonica L.) for different treatments. The exposure period was 192 h. The values are the mean of three replicates. Vertical lines represent standard deviations.

In most cases, the disadvantage of the Lineweaver-Burk plot is that experimental measurements of C are not distributed homogeneously on the graph. Additionally, small errors in v lead to large errors in 1/v and hence to large errors in K_M and v_max.³¹ Realistic values of the half-saturation constant, K_M, were 265.9 mg CN L⁻¹; the maximum metabolic capacity, v_max, was 38.61 mg CN kg⁻¹d⁻¹ using Lineweaver-Burk plots. Kinetic data were also analyzed with a computer program using non-linear regression treatments, provided by GraphPad Software Inc. Slightly low values for K_M and v_max were obtained to be 228.1 mg CN L⁻¹ and 36.43 mg CN kg⁻¹d⁻¹, respectively.

Discussion

Visible toxic symptoms, e.g., chlorosis of leaves, were only observed for the plants exposed to the highest concentration of ferricyanide in this study. The results of all measured toxic effects over the concentrations tested were plotted and analyzed (data not shown). All linear trends were significant, except for the activity of SOD, judged by the critical r for given n (α = 0.05).³⁰ Of the selected parameters, the best correlation was obtained for the soluble protein contents in leaves (R² = 0.969, significant at α = 0.05), indicating that the soluble protein contents were the most sensitive indicator to ferricyanide exposure. The susceptibility of these parameters to the change of ferricyanide concentrations was in the order: soluble proteins > POD > CAT > transpiration rate > chlorophyll a > chlorophyllb > SOD.

Positive effects were observed in this study when plants were exposed to ferricyanide ≤ 274.13 mg CN L⁻¹, exhibiting higher measured values of chlorophyll contents and soluble proteins in plants than those of the non-treated control plants. This is largely due to the well-established assimilation systems in weeping willows as well as the enhanced antioxidation responses by biochemical processes.

Ferricyanide is purportedly membrane impermeable,¹³ therefore in vivo dissociation or biodegradation of ferricyanide to ferrocyanide or free cyanide would be a prerequisite before entering to the plants. In a reported work by Larsen and Trapp,²⁴ferricyanide was taken up by basket willows and the amounts of dissolved cyanide were assimilated and metabolized inside the plants. Larsen and Trapp²⁴ also found that ferricyanide can be extracellular reduced by the ferrireductase to ferrocyanide in the plant growth media containing nutrient compounds, whereas ferricyanide remained the dominating species in the distilled water largely due to the lack of electron acceptor.

Willows assimilate iron through solubilization of Fe^(III) by extracellular acidification, first reduction of Fe^(III) to Fe^(II) by a plasmamembrane redox system named ferrireductase, then uptake of Fe²⁺ can be achieved by a specific transporter located in membrane.^33–35 This mechanism requires the obligatory reduction of extracellular Fe^(III)-chelate complexes before the splitting of the complex and then uptake of the released Fe^(II).

Free cyanide (HCN or CN⁻) is known to be converted to the metabolite asparagine in the presence of the enzymes β-cyanoalanine synthase and β-cyanoalanine hydrolase.^20,36 Other earlier findings also suggested that an assimilation pathway for ferrocyanide largely existed in plants.^10,13,29 It was shown that free cyanide did not accumulate in healthy trees.¹⁵ In this study, amounts of total cyanide detected in plant materials were significantly higher than the background level, implying that ferricyanide was probably still in the same original complex form in the plant materials. The same observation was also made by Larsen and Trapp.²⁴ Additionally, the ¹⁵N from the labeled ferrocyanide was not detected in willow tissues as cyanide using a stable nitrogen isotope.¹³ Therefore, there are two other possibilities for the assimilation of ferricyanide by plants: plants either assimilate ferricyanide directly as a substrate or use the intermediate ferrocyanide as a substrate without any additional conversion. However, further in-depth studies are needed to elucidate the detailed degradation biochemical pathway of ferricyanide in plants and the specific chemical form of ions transported into the plant initially.

The variations in ferricyanide application strongly affected both the kinetics of uptake and metabolism and the transpiration rate. In this study, the correlation between the metabolic rates and the initial ferricyanide concentrations followed Michaelis-Menten kinetics as shown in Fig. 4. Indeed, this coincided with the change of plant transpiration rates. When exposed to higher concentrations of ferricyanide ≥ 410.26 CN L⁻¹, a decrease in transpiration rate was evident. As a result, the loss of ferricyanide from hydroponic solution into plants was lowered, consequently metabolic rates of ferricyanide by plants were unable to increase further. Larsen and Trapp²⁴ also found that ferricyanide uptake was strongly connected to the transpiration. Additionally, the transpiration stream concentration factor (TSCF) for ferricyanide was estimated to be 0.42, which indicated a close connection between the uptake of water and of iron cyanide, but not a linear relationship.²⁴

The conversion of ferricyanide to ferrocyanide is most likely to take place prior to the uptake by plant roots. However, Larsen and Trapp²⁴ observed that there was little support from the experimental results that ferricyanide was taken up as Fe^(II) ion after reduction in solution. This transformation pathway may operate in plant materials during the transport of ferricyanide. Here, it is interesting to note that the removal rates of ferricyanide from the hydroponic solution by plants between the five treatments were quite different in this study, but the recovery of total cyanide in plant materials was very similar for all treatment groups. This indicated that the availability of the enzyme capable of assimilating ferricyanide within plant materials is a major limiting factor for the conversion of ferricyanide by plants.

Less than 13% of the applied ferricyanide was removed from the plant growth media by plants and the kinetics in this study showed signs of enzyme-limitation and followed Michaelis-Menten kinetics. Ferricyanide has long been considered membrane impermeable,¹³ which implied that a specific carrier or transporter with low affinity for ferricyanide may exist in root membrane. Previously, the Michaelis-Menten kinetics of cyanide removal by 12 plant species out of 9 families were determined¹⁸ and the values of v_max and K_M were between 6.68 and 21.91 mg CN kg⁻¹ h⁻¹ and 0.90 to 3.15 mg CN L⁻¹, respectively. In this study, realistic values of K_M and v_max were estimated to be 228.1 mg CN L⁻¹ and 36.43 mg CN kg⁻¹d⁻¹ using non-linear regression treatments, which was significantly different to the data obtained with free cyanide. This could be a reliable explanation for the higher uptake and metabolic rate for free cyanide than ferricyanide in willows.

Conclusions

The added ferricyanide did not induce deleterious lesions on the plant physiological functions over 192 h of exposure. Positive effects were observed at low concentrations of ferricyanide. Visible toxic symptoms were observed for the plants exposed to 506.66 mg CN L⁻¹. Some parameters for toxicity determination showed a dose-dependent manner. Soluble proteins in leaves were the most sensitive indictor for the plants exposed to ferricyanide application. Although a fraction of the initially applied ferricyanide was removed from the hydroponic solution in the presence of plants, the majority remained in the solution. Small amounts of the applied ferricyanide were recovered in plant materials and the distribution in biomass varied significantly in different treatment groups. The estimated metabolic rates at different ferricyanide treatment concentrations can be described by Michaelis-Menten kinetics. The information suggests that the assimilation of ferricyanide in plants is possible and phytoremediation of ferricyanide is an option for treating environmental sites contaminated with cyanide complexes.

Acknowledgements

This work was supported by a Ph.D. studentship from The University of Hong Kong. Thanks to Sheng-Zhuo Huang, Luan Li and Shuo Liu for their technical assistance.

References

1. M. Barclay, A. Hart, C. J. Knowles, J. C. L. Meeussen and V. A. Tetdl, Enzyme Microb Technol., 1998, 22, 223–231.
2. T. Mudder and M. Botz, Mining Environ Manag., 2001, 9, 8–12.
3. G. C. Miller and C. A. Pritsos, in Cyanide: Social, Industrial and Economic Aspects 2003, ed. C. A. Yong, L. G. Tidwell and C. G. Anderson,Warrendale: TMS (The Minerals, Metals & Materials Society), pp. 73–81.
4. M. M. Figueira, V. S. T. Ciminelli, M. C. Andrade and V. R. Linardi, Can J Microbiol., 1996, 42, 519–523.
5. V. Augugliaro, V. Loddo, G. Marci, L. Palmisano and M. J. L. Munoz, J Catal., 1997, 166, 272–283.
6. V. K. Sharma, W. Rivera, J. O. Smith and B. O'Brien, Environ Sci Technol., 1998, 32, 2608–2613.
7. T. L. Theis and M. J. West, Environ Technol Lett., 1986, 7, 309–316.
8. J. L. Meeussen, M. G. Keizer and W. H. Van-Riemsdijk, Environ Sci Technol., 1992, 26, 1832–38.
9. T. L. Theis, T. C. Young, M. Huang and K. C. Knutsen, Environ. Sci Technol., 1994, 28, 99–106.
10. M. Samiotakis and S. D. Ebbs, Environ Pollut., 2004, 127, 169–173.
11. P. Kjeldsen, Wat Air Soil Pollut., 1999, 115, 279–307.
12. S. Trapp and H. Christiansen, in Phytoremediation: Transformation and control of contaminants ed. S. C. McCutcheon and J. L. Schnoor, Hoboken: John Wiley & Sons, 2003, pp. 829–862.
13. S. Ebbs, J. Bushey, S. Poston, D. Kosma, M. Samiotakis and D. Dzombak, Plant Cell Environ., 2003, 26, 1467–1478.
14. M. Larsen, S. Trapp and A. Pirandello, Chemosphere, 2004, 54, 325–333.
15. M. Larsen, A. Ucisik and S. Trapp, Environ Sci Technol., 2005, 39, 2135–2142.
16. X. Z. Yu, S. Trapp, P. H. Zhou, C. Wang and X. S. Zhou, Chemosphere, 2004, 56, 121–126.
17. X. Z. Yu, S. Trapp, P. H. Zhou and H. Hu, Chemosphere, 2005, 59, 1099–1104.
18. X. Z. Yu, P. H. Zhou, X. S. Zhou and Y. D. Liu, Environ Sci Pollut Res., 2005, 12, 227–232.
19. J. M. Miller and E. E. Conn, Plant Physiol., 1980, 65, 1199–1202.
20. A. Maruyama, K. Saito and K. Ishizawam, Plant Mol Biol., 2001, 46, 749–760.
21. K. Manning, in Cyanide compounds in biology, ed. Ciba Foundation, Chichester: John Wiley & Sons, 1988, pp. 92–110.
22. A. D. Beavis and A. E. Vercesi, J Biol Chem., 1992, 267, 3079–3087.
23. H. R. Hendrickson and E. E. Conn, J Biol Chem., 1969, 244, 2632–2640.
24. M. Larsen and S. Trapp, Environ Sci Technol., 2006, 40, 1956–1961.
25. S. Trapp, K. C. Zambrano, K. O. Kusk and U. Karlson, Arch Environ Contam Toxicol., 2000, 39, 154–160.
26. S. Maclachalam and S. Zalik, Can J Bot., 1963, 41, 1053–1062.
27. J. H. Jin and Z. R. Ding, Methods of Plants Biochemistry Analysis. Beijing: Chinese Science Press, 1981.
28. K. H. Cheung and J. D. Gu, World J Micro Biotechnol., 2005, 21, 213–219.
29. X. Z. Yu, P. H. Zhou and Y. M. Yang, Ecotoxicol., 2006, 15, 461–467.
30. L. Sachs, Angewandte Statistik. Berlin: Springer, 1992.
31. A. Cornish-Bowden, Fundamentals of Enzyme Kinetics. London: Portland Press, 1995.
32. P. J. F. Henderson, in Enzyme Assays, ed. R. Eisenthal and M. J. Danson, Oxford: IRL Press, 1992.
33. M. L. Guerinot and Y. Ying, Plant Physiol., 1994, 104, 815–820.
34. H. Marschner and V. Römheld, Plant Soil., 1994, 165, 261–274.
35. D. Eide, Curr Opin Cell Biol., 1997, 9, 573–577.
36. P. A. Castric, K. J. F. Farnden and E. E. Conn, Arch Biochem Biophys., 1972, 152, 62–69.

---