Environ. Sci. Technol. 2007, 41, 4271-4276
Formation and Degradation Kinetics of the Biofumigant Benzyl Isothiocyanate in Soil A N N E L O U I S E G I M S I N G , * ,† J E S L E I S G A A R D P O U L S E N , †,‡ HENRIK LAURBERG PEDERSEN,† AND HANS CHRISTIAN BRUUN HANSEN† Department of Natural Sciences, Faculty of Life Sciences, University of Copenhagen, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark, and Plant Biochemistry Laboratory, Faculty of Life Sciences, University of Copenhagen, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark
Glucosinolates (GSLs) are produced by plants of the Capparales order. Upon enzymatic hydrolysis the GSLs can be transformed to the toxic isothiocyanates (ITCs), which can be used as biofumigants for the control of soilborne pests. The rates of ITC formation and degradation are critical to both biofumigation and the toxicity and leaching of GSLs and ITCs in soil. Degradation kinetics of benzyl GSL and benzyl ITC in a sandy and clayey surface and subsoil at 8-9 °C at natural moisture contents were investigated, as was the rate of formation of ITC from the GSL. Degradation of GSL followed logistic kinetics with t1/2 ) 0.79.1 days. Degradation was faster in clayey soil compared to sandy soil, and faster in surface soil compared to subsoil. In surface soils, up to 25% of added GSL was detected as ITC, while only 1-6% were detected in the subsoils. ITC degradation followed first-order kinetics with t1/2 ) 0.3-1.7 days, with faster degradation in subsoils than in surface soils. Based on the data for GSL hydrolysis and ITC degradation, the concentration of ITC following GSL application was successfully modeled assuming complete conversion of glucosinolate to isothiocyanate and firstorder degradation of isothiocyanate.
Introduction Substitution of naturally produced bioactive compounds for synthetic pesticides is gaining increasing interest as a novel form of pest control in agriculture (1-5). Also in soil fumigation there is an increasing interest in substituting naturally produced isothiocyanates (ITCs) for the synthetic fumigants methyl bromide and methyl isothiocyanate ITC (6-8). The natural ITCs are hydrolysis products of GSLs (Figure 1), which are compounds produced by plants of the Capparales order including Brassica plants like rape and indian mustard (9, 10). GSLs are sulfur-containing glucosides and, besides from the ITCs, other hydrolysis products may form depending on the specific GSL and environmental factors like pH (10). The enzyme catalyzing the hydrolysis is * Corresponding author phone: +45 3528 2413; fax: +45 3528 2398; e-mail:
[email protected]. † Department of Natural Sciences, Faculty of Life Sciences, University of Copenhagen. ‡ Plant Biochemistry Laboratory, Faculty of Life Sciences, University of Copenhagen. 10.1021/es061987t CCC: $37.00 Published on Web 05/19/2007
2007 American Chemical Society
a β-thio glucosidase called myrosinase, which is present in all plants producing GSLs but it is also present in soil (1012). Upon disruption of plant tissue, the GSLs come in contact with myrosinase and ITCs start to form (10). Utilization of these natural fumigants can be achieved by a technique called “biofumigation” where Brassica plants are mulched and incorporated into the surface soil where the ITCs are produced, and subsequently, depress soil-borne pests (7, 8). Optimization of biofumigation by addition of GSLs requires proper description and control of the rates of GSL hydrolysis and ITC dissipation to obtain the appropriate doses and to quantify the risks associated with the utilization. Naturally produced compounds for pest control are usually conceived as more environmentally safe than man made pesticides simply because they are natural. However, natural compounds can be just as toxic as synthetic compounds, and some of the most toxic compounds known are of natural origin. An example is the lethal toxin ricin produced by castor bean (Ricinus communis); after (13), other examples comprise cytisin in laburnum, the carcinogenic glucoside ptaquiloside present in bracken, cyanogenic glucosides in different crop plants and trees, and different mycotoxins of which some, like the afflatoxins, are highly toxic and some are carcinogenic. In the case of ITC biofumigants, it has been documented that these compounds are toxic to a wide range of organisms including nematodes, insects, bacteria, and fungi (7, 14-20), and since the toxicity is not specific (16), it is likely that nontarget as well as target organisms will be affected. Naturally produced pesticides may spread in the environment like conventional pesticides due to leaching from soils thereby contaminating drinking water reservoirs and surface waters. Thus, if natural compounds are going to be used as pesticides, it is critical to investigate the environmental fate and effects of these compounds as is done for conventional pesticides (5). Currently, very little is known about the environmental chemistry of natural compounds in soil. However, it appears that glucosidic natural compounds hardly sorb in soil as seen for GSLs, cyanogenic glucosides, and the sesquiterpenoid ptaquiloside (11, 21, 22). The aim of this study has been to quantify the formation kinetics of benzyl ITC in soil by hydrolysis of benzyl GSL as well as to quantify the degradation kinetics of both benzyl GSL and benzyl ITC. The experiments were done at natural soil moisture contents and at 8-9 °C with A and B horizons of a sandy and a clayey soil. The kinetics of all reactions were mathematically described to gain additional insight into the mechanisms and to obtain tools for future optimization of biofumigation and risk assessment.
Experimental Section Soils. A sandy and a clayey Danish soil were used in the experiments. Both soils are agricultural soils, and no GSL containing crops have been grown on the soils for at least the last 5 years. The soils were chosen because of their different soil characteristics and because they represent two of the most common soil types found in Denmark and Northern Europe. The sandy soil is from Jyndevad in the southwestern part of Denmark, and it has been developed on coarse sandy glacifluvial material and it is classified as a humic psammentic dystrudept (23). The clayey soil is from Sjællands Odde (Sj. Odde) in the northeastern part of Denmark and has been developed on calcareous clayey lodgement and melt-out till from the Weichselian glaciation, and it is classified as a typic agriudoll (23). Selected properties of the soils are given in VOL. 41, NO. 12, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4271
FIGURE 1. Hydrolysis of benzyl glucosinolate and the formation of benzyl isothiocyanate. The reaction is catalyzed by the enzyme myrosinase.
TABLE 1. Selected Properties for the Two Soils Used soil
depth (cm)
pHCaCl2
Nb (%)
Cb (%)
claya (%)
silta (%)
sand (%)
Sj. Odde A Sj. Odde B Jyndevad A Jyndevad B
0-30 30-65 0-30 30-65
7.2 7.6 6.9 6.1
0.14 0.05 0.12 0.04
1.25 0.30 2.43 1.00
19 31 5 5
18 26 3 1
63 43 92 94
a
Clay: < 2 µm, silt 2-20 µm.
b
9
respiration (µg C‚h-1g-1)
91.2 116.0 81.8 49.8
0.458 0.171 0.405 0.073
N Is Total Nitrogen and C Is Total Organic Carbon. The Water Content Was Measured by Drying at 110 °C
Table 1. The soil respiration rate was measured as described by Gimsing et al. (24). After the soils were sampled, they were stored at -18 °C at the moisture contents at time of sampling. Before use, the soils were thawed in a refrigerator (5 °C) and sieved through a 2 mm sieve. All experiments were done with the original water content (Table 1). Degradation Kinetics of GSL and ITC and Formation of ITC from GSL. The degradation kinetics of benzyl GSL was investigated by weighing 2 g of soil into a 50 mL centrifuge tube with screw cap (TPP) and adding 100 µL benzyl GSL (2 mg mL-1) giving 100 µg g-1 soil or 214.8 nmol g-1 soil. The benzyl GSL was purchased from www.glucosinolates.com (Faculty of Life Scences, University of Copenhagen, Denmark) and used as received. The samples were incubated in an 8-9 °C water bath and remaining GSL extracted after 0.25, 0.50, 1, 2, 4, 6, 8, 10, 12 h and 1, 2, 3, 4, 6, 8, 10, 12, 14, 21, and 28 days. The samples were extracted by adding 5 mL cold MilliQ-water (Millipore) vortexed for a few seconds and centrifuging for 10 min at 8000g in a precooled centrifuge. After centrifugation, the supernatant was decanted into a 10 mL syringe and passed through a 0.45 µm regenerated cellulose filter (Mikrolab Aarhus A/S) to a 1.5 mL Eppendorf tube. The samples were stored at -18 °C until measurement with capillary electrophoresis (CE) as described by Gimsing et al. (11). Formation of benzyl ITC from benzyl GSL in the soils was followed by weighing 2 g of soil into a 100 mL glass flask and adding 100 µL of the standard solution of benzyl GSL (2 mg mL-1) giving 100 µg g-1 soil or 214.8 nmol g-1 soil of benzyl GSL. The flasks were sealed with airtight rubber lids and incubated in a water bath (8-9 °C) and sampled after 0.50, 1, 2, 4, 6, 8, 10, 12 h and 1, 2, 3, 4, 6, 8, 10, 12, 14, and 21 days. The extraction of ITC present in both the aqueous phase, gas phase and reversibly sorbed to the solid phase of the soil was performed by adding 5 mL of ethyl acetate through the rubber lid with a syringe. The next hour the samples were shaken a couple of times, and then the liquid phase was subsequently decanted into at glass beaker. Addition of 5 mL ethyl acetate to the soil samples was repeated two more times. After combining the ethyl acetate extracts, 72 µL of phenylethyl ITC (559 µmol L-1) was added as internal standard. The samples were evaporated to approximately 5 mL in a fume hood and passed through a Pasteur pipet packed with anhydrous magnesium sulfate to remove water. Samples were 4272
water cont. mg g-1 soil
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 12, 2007
now evaporated to 1-2 mL in a fume hood and transferred to vials and quantified with GC-MS (PolarisGCQ, ThermoFinnigan, Austin). The GC column was 30 m × 0.25 mm i.d., coated with 0.25 µm fused silica Rtx-5MS (Restek, Bellefonte). Helium was used as a carrier gas at a constant flow of 1.2 mL min-1. The temperature program was 40 °C (2 min) to 285 °C at 15 °C min-1. The transfer line was 275 °C, the MS ion source 200 °C and the scan range was from 50 to 200 m/z. Quantification was made by selective ion monitoring of the 91 m/z fragment for benzyl ITC and 163 m/z fragment for phenylethyl ITC. The injection temperature was 200 °C in splitless mode, and the injection volume was 2 µL sample and 1 µL air. LOD was determined to be 2.3 µM for benzyl ITC and 0.28 µM for phenylethyl ITC. The degradation kinetics of benzyl ITC was determined by weighing 2 g of soil into 250 mL glass flasks with an airtight rubber lid. The benzyl ITC solution used for the experiment was made by adjusting the pH of 2 mL of the benzyl GSL solution (2 mg mL-1) to 8 with a few µL of 0.1 M KOH and then adding a small amount of myrosinase (EC 3.2.3.1 purified from Sinapis alba, Sigma). The mixture was incubated at room temperature for 1.5 h. Benzyl ITC production resulted in a grayish color of the solution. Measurement of the benzyl ITC concentration in this solution showed that all the GSL had been converted to ITC. 100 µL of this benzyl ITC solution was added to the soil giving an initial concentration of 32.1 µg g-1 soil or 214.8 nmol g-1 soil of the ITC. Samples were taken after 1 and 3 h and 1.5, 5, 7, 9, and 14 days. The samples were extracted and measured as described above for the experiments where ITC formation was studied. The effect of ITC on soil respiration was investigated by repeating the ITC degradation experiment, but instead of extracting the samples, the CO2 production was measured on a GC regularly for 14 days. Samples without ITC were included as control samples. All experiments were done in triplicate. Modeling. The aim of the modeling is to describe the time-dependent variation in the concentration of benzyl ITC in the soil when benzyl GSL is added to the soil (Figure 2) based on the hydrolysis kinetics for GSL and the dissipation kinetics of ITC determined in independent experiments. The change in the concentration of benzyl ITC versus time, [ITC]′(t), depends on ITC produced by the hydrolysis of benzyl
FIGURE 2. Formation of benzyl isothiocyanate following the application of benzyl glucosinolate to soil. The bars show the standard deviation.
FIGURE 3. Degradation of benzyl glucosinolate in the A and B horizons of the two soils. The lines are the fit of the data to a logistic degradation curve. The bars show the standard deviation.
GSL and ITC removed by degradation and irreversible sorption. This leads to the following differential equation:
[ITC]′(t) ) r1[GSL](t) - r2[ITC](t)
(1)
where t is time in days, [ITC](t) is the ITC concentration at a given time, [GSL](t) is the GSL concentration at a given time, r1 is the rate at which GSL is hydrolyzed to ITC and r2 is the rate of ITC dissipation. To model the formation of ITC according to eq 1, the function [GSL](t) should be known. Here, a logistic model has been used as it is able to fit GSL hydrolysis data from both A and B horizons. The general form of the logistic model is
[GSL](t) )
K 1 + cert
(2)
where [GSL](t) is the GSL concentration at time t and K, c, and r are positive parameters found by fitting using the data for the degradation of GSL in each of the horizons and types of soil. The rate coefficient r2 in eq 1 is estimated from the experiments where ITC dissipation was followed. The dissipation kinetics of ITC can be successfully fitted by a firstorder model:
[ITC](t) ) ae-r2t
(3)
where the parameters a and r2 are found by fitting in a separate experiment. Based on these considerations, the time-dependent concentration of ITC in the experiments where GSL has been added can be calculated by solving the differential eq 1, where [GSL](t) is of the form (2) and where r2 is the rate from (3). The solution [ITC](t), for which [ITC](0) ) 0 is given by the following formula:
[ITC](t) ) r1
∫e t
0
K dse-r2t 1 + cers
r2s
(4)
The parameter r1 is estimated by fitting. This is done for each horizon and each type of soil so that there are already etimates for r2, K, r, and c. More on this in the Supporting Information.
Results and Discussion Degradation Kinetics of GSL and ITC and Formation of ITC from GSL. From a biofumigation point of view it is desirable that all GSLs in plant material incorporated into the soil are hydrolyzed to the biologically active ITCs to obtain the maximum effect against pathogens (25). GSLs may also
FIGURE 4. Degradation of benzyl isothiocyanate in the A and B horizons of the two soils. The lines are the fit of the data to a first-order degradation curve. The bars show the standard deviation.
TABLE 2. The Two Rate Constants r1 and r2 from the Differential Equation (eq 1) Describing the Formation and Degradation of ITC from GSLa soil
r1
r2
r
Sj. Odde A Sj. Odde B Jyndevad A Jyndevad B
0.427 (0.023) 0.018 (0.001) 0.204 (0.008) 0.022 (0.002)
-0.412 (0.077) -2.270 (0.374) -0.482 (0.075) -0.684 (0.109)
1.47 0.24 0.65 0.50
t1/2 t1/2 GSL d ITC d 0.73 8.13 1.86 9.04
1.68 0.31 1.44 1.01
a The constant r is the rate constant from the logistic model describing the degradation of GSL, eq 2. the half-lives for ITC and GSL are also given. Number in brackets is the standard error.
enter the soil from the roots of growing plants, or in the future, pure GSLs may be used as biofumigant precursors and in this situation the rate of hydrolysis will depend on the myrosinase activity in the treated soil (11). In this study, the rate of formation of ITC differed widely between soils (Figure 2). In the A-horizons a maximum of 22-25% of the added benzyl GSL could be found as benzyl ITC, whereas in the B-horizons only 0.9-5.6% could be detected. The detected maximum amount corresponds to 50 nmol g-1 soil of benzyl ITC. These amounts are lower than what has been found in studies where GSLs have been added as part of plant material or rapeseed meal, because in these cases, the myrosinase activities are higher resulting in considerably higher rates of ITC formation. In a study by Elberson et al. (26) with rapeseed meal-amended soil, a maximum of between 36 and 67% of the GSLs were recovered as ITCs. In another study, 2-propenyl GSL and myrosinase was added together to soils and around 80% was recovered VOL. 41, NO. 12, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4273
FIGURE 5. Results of the modeling (solid lines) of the formation of isothiocyanate and the degradation of glucosinolate shown together with data. as 2-propenyl ITC (27). A biofumigation field study achieved an ITC release efficiency of 56% 30 min after application of Indian mustard (Brassica juncea) to the soil (28). Another reason for the low recovery of ITC seen in our study, especially in the B-horizons, could be due to formation of other GSL hydrolysis products, such as nitriles and thiocyanates. In fact, in the subsoils, benzyl nitrile and other unknown products were detected. However, they were only present in minor amounts (
