In recent years, emerging pollutants, including pharmaceuticals, personal care products, and endocrine disrupting substances have been widely detected in surface waters. The potential environmental and public health impact of these chemicals is of scientific and regulatory concern 1–6. Major sources for environmental exposure to these chemicals can include the effluent of sewage treatment plants, land application of municipal biosolids (sewage sludge), municipal landfill leachate, and irrigation with reclaimed wastewater 7.
Liquid and dewatered municipal biosolids have been widely used in agriculture as a source of crop nutrients and organic matter for soil improvement 8. In simulated precipitation experiments, we have been characterizing the movement of micropollutants from land fertilized with biosolids in surface runoff water 2, 4. In more recent experiments, we have detected the tricyclic antidepressant amitriptyline (Fig. 1) and the transformation product 10-hydroxyamitriptyline in surface runoff from land fertilized with biosolids 9. Biosolids used in these experiments contained 448 ng amitriptyline g dry weight−1, and 23 ng 10-hydroxyamitriptyline g dry weight−19. Amitriptyline is among the most widely used tricyclic antidepressants, with approximately six tonnes purchased annually in Canada 10. Since its introduction in the 1960s, the drug has been used widely in treating depression, to control anxiety, and to treat a variety of chronic pain syndromes 11, 12. Amitriptyline has been detected at µg L−1 concentrations in surface water 13 and in wastewater effluent 5.
Amitriptyline acts mainly by inhibiting the neurotransmitters norepinephrine and serotonin reuptake in the central nervous system. The metabolism of amitriptyline has been thoroughly investigated in mammals (including rats, mice, and humans) and fungi 14–21. In humans, the main pathway for amitriptyline metabolism is demethylation in liver microsomes to the pharmacologically active metabolite nortriptyline 11, 20. Nortriptyline is also an antidepressant available through prescription; approximately 500 kg are sold annually in Canada 10. Both amitriptyline and nortriptyline undergo benzylic hydroxylation, mainly at the 10-position 11. Both parent and transformation products are excreted intact 13. The hydroxylated metabolites of amitriptyline and nortriptyline are excreted in the urine mainly as glucuronide conjugates. The demethylation of amitriptyline and nortriptyline is catalyzed mainly by CYP2C19 at low amitriptyline concentrations, whereas CYP3A4 may be more important at higher amitriptyline concentrations 11, 18, 20. The hydroxylation of amitriptyline and nortriptyline into their active metabolite, 10-hydroxyamitriptyline and 10-hydroxynortriptyline is catalyzed mainly by CYP2D6 11, 20.
Given the very large volume use of amitriptyline and nortriptyline, their widespread detection in surface water, their detection in biosolids, and the widespread use of biosolids in agriculture, the present study aimed to elucidate the persistence characteristics and dissipation pathways of amitriptyline and nortriptyline in agricultural soils. Furthermore, given that micropollutants will reach agricultural land through the application of biosolids, the effect of this material on soil dissipation was determined. To our knowledge, this is the first study to examine the fate of these drugs in soil.
Materials and chemicals
Amitriptyline-[N-methyl-3H] hydrochloride (radioactive purity >99%, specific activity, 60 mCi mmol−1) was purchased from American Radiolabeled Chemicals. Amitriptyline hydrochloride (purity >98%), nortriptyline hydrochloride (purity >98%), and (± )-E-10-hydroxylated amitriptyline were purchased from Sigma-Aldrich, Canada (Fig. 1). High performance liquid chromatography (HPLC) grade acetonitrile, ethanol, methanol were purchased from Caledon. All other chemicals were obtained from Sigma-Aldrich, Canada. Stock solutions of 3H-labeled (final radioactive concentration of 1,500 k dpm 100 µl−1) and unlabeled (1 mg ml−1) chemicals were prepared in ethanol and stored at −20°C until later use.
Three agriculture soils (sampling depth, 0–20 cm) that varied in texture and chemical properties were used in the present study: a loam obtained from the Agriculture and Agri-Food Canada research farm at London, Ontario (42°59′N, 81°15′W); a sandy loam soil obtained from the Agriculture and Agri-Food Canada research farm at Delhi, Ontario (42°51′N, 80°29′W); and a clay loam soil obtained from the Essex Region Conservation Authority research farm in Holiday Beach, Ontario (42°2′N, 83°3′W). Key properties of these soils are described in Al-Rajab et al. 22. All soils were obtained from areas that were under sod and had never received biosolids. Soils were sieved to a maximum particle size of 2 mm and were stored at −7°C prior to experimentation.
The liquid municipal biosolids (LMB) used in the present study were obtained from the Adelaide pollution control plant in London, Ontario. The biosolids were returned activated sludge from secondary treatment and had been aerobically digested. The LMB used had the following key characteristics: pH 7, dry matter content = 0.5%, organic matter content = 0.4%, and carbon-to-nitrogen ratio = 6:1.
Fifty-gram portions of the soils were incubated in laboratory microcosms as described in Al-Rajab et al. 22 and Sabourin et al. 23. Briefly, microcosms consisted of 150 ml baby-food jars incubated in sealable glass 1-L mason jars. A scintillation vial containing 10 ml of water was placed in each jar to maintain soil moisture and prevent desiccation of the soil. Soils were supplemented with 3H-labeled and unlabeled amitriptyline by adding stock solutions in ethanol to 1-g portions of pulverized air-dried soil, allowing the solvent to evaporate, and then thoroughly mixing this into 49 g (moist wet) of soil to give a total of 50 g. Soils received 1 µg amitriptyline g−1 and 30 k dpm g−1 moist soil, and soil moisture were normalized to 15% in all treatments. Triplicate microcosms were prepared for each treatment, and these microcosms were incubated in darkness at 30°C, unless otherwise indicated. The effects of LMB (10% v/v) on drug dissipation were evaluated only in the loam soil. In experiments evaluating nortriptyline, dissipation soils received 10 µg nortriptyline g−1 moist loam soil; otherwise, the incubation conditions and extraction procedures were the same as described for amitriptyline.
In preliminary experiments, we determined the efficiency of various solvent mixtures to extract radioactivity from soils supplemented with 3H-amitriptyline. Because amitriptyline and nortriptyline are basic, NH4OH was added to the organic solvent to improve the extraction efficiency. Acetonitrile with 2% NH4OH was considered satisfactory to extract amitriptyline from the three soils. The recovery of amitriptyline from the three soils ranged from 69.3 to 77.7% (Table 1).
Table 1. The extraction recoveries (%) from loam soil, clay soil, and sandy soil using various solventsa
Triplicates for each data point.
na = data not available.
Acetonitrile + 2% NH4OH
77.2 ± 5.1
69.3 ± 5.2
77.7 ± 3.4
Hexane/ethyl acetate/NH4OH (90:8:2)
54.8 ± 1.3
62.5 ± 1.2
Four-gram portions of soil were removed periodically from each microcosm with a spatula and stored at −20°C until extraction. Soils were extracted three times with 15 ml of organic solvent each time. For each extraction, samples were vigorously shaken automatically for 20 min on a wrist-action shaker. The samples were then centrifuged for 15 min at 1,100 g in an 2150 swinging bucket rotor in a Hereaus-Christ Labofuge 3000 centrifuge (Fisher Scientific). The supernatants were transferred to a clean glass vial, reduced to dryness under nitrogen in a 35°C water bath, taken up in 200 µl of acetonitrile-water mixture (1:1, v/v) and stored at −20°C until analyzed.
Radioactivity in soil extracts was determined by liquid scintillation counting (LSC) using a Beckman Coulter Model LS 6500 instrument. Each sample was added to 10 ml UniverSol scintillation cocktail (ICN) in plastic scintillation vials. Data were corrected automatically for quenching.
The 3H-amitriptyline and potential radioactive transformation products in 50 µl portions of soil extracts were analyzed by HPLC, consisting of an Agilent 1260 Infinity HPLC instrument, autosampler and fraction collector (Agilent), with ultraviolet (UV) detection and radioactivity detection (EG&G Berthold LB509 Radioflow Detector, Berthold GMBH & Co. KG). An Agilent Zorbax Eclipse extended XDB C-18 column, (4.6 × 250 mm, 5 µm pore size), coupled with a guard column with the same packing material (2.1 × 12.5 mm, 5 µm), was used for the separation. The mobile phase consisted of acetonitrile and aqueous ammonium acetate (20 mM) containing 0.1% formic acid (33:67, v/v) delivered at a flow rate of 0.8 ml/min. The UV detector wavelength was set at 210 nm. Under these conditions, the retention times of nortriptyline, amitriptyline, and amitriptyline-N-oxide were 11.2 min, 12.6 min, and 16.1 min, respectively. Fractions from the solvent outflow of the HPLC column were collected at 0.5 min intervals into glass autosampler vials. The radioactivity of each fraction was determined by LSC.
To identify potential amitriptyline and nortriptyline transformation products, selected soil extracts were analyzed by HPLC-mass spectrometry (HPLC-MS). The HPLC-MS system consisted of an Alliance 2690 HPLC with an autosampler coupled to a time-of-flight mass spectrometer with LCT orthogonal acceleration (Waters/Micromass). Samples were separated with a Synergy Hydro-RP column (150 × 2 mm, 4 µm; Phenomenex). The mobile phase consisted of A (water containing 10% acetonitrile and 0.1% formic acid) and B (acetonitrile containing 10% water and 0.1% formic acid). The B content changed linearly from 20 to 100% over 15 min. The flow rate was 0.2 ml/min, with an injection volume of 20 µl. The total effluent from the column was introduced into the source of the mass spectrometer. The mass spectrometer was operated using electrospray ionization in positive ion mode. The MS conditions were as follows: capillary, 3000 V; cone, 20 V; desolvation temperature, 250°C; source temperature, 80°C; and desolvation gas flow, 450 L/h. Mass spectra were acquired over the range 100 to 400 m/z. Instrument control, data acquisition, and processing were accomplished using MassLynx 4.0 (Waters/Micromass). Extracts for HPLC-MS analysis were obtained from the soil incubations with unlabeled amitriptyline or nortriptyline.
Calculations and statistics
Extractable radioactivity is presented as a percentage of the initial radioactivity added. Dissipation rates for parent compounds were estimated on the basis of removal of total radioactivity from the extractable phase. The half-life of the parent compounds (DT50) in loam soil was expressed as the days to dissipate 50% of the initially added radioactivity from the parent compounds or the initially added unlabeled parent compound, that is, nortriptyline. Total extractable radioactivity was used to estimate DT50 in clay and sandy soil, where the metabolites of amitriptyline were less than 3% of the supplemented parent compound. In experiments with nortriptyline-supplemented soil, extractable parent compound and potential transformation products were evaluated by LC-TOF-MS/UV. Raw data analysis was conducted using Microsoft Excel 2007 (Microsoft). Dissipation curves were plotted using Microsoft Excel 2007. Data in the figures represent the mean and standard deviation of triplicate samples. Statistically significant differences between treatments were established by Student's t tests at 95% confidence. Curves were fitted to a first-order equation (exponential decay).
RESULTS AND DISCUSSION
Kinetics of amitriptyline dissipation in three agriculture soils
Total extractable radioactive residues declined only slowly in the three agriculture soils (Fig. 2). During 77 d of incubation at 30°C, the total extractable radioactivity decreased from 60.3 to 26.6% in the loam; from 60 to 30.9% in the sandy loam; and from 81.2 to 47.2% in the clay loam (Fig. 2). Soil extracts were fractionated by HPLC, and the distribution of radioactivity in parent and transformation products was quantified by LSC. During 77 d of incubation, recovered radioactivity co-migrated with standards of amitriptyline and nortriptyline and a compound which, on the basis of LC-TOF-MS, was determined to be amitriptyline-N-oxide. The portion of radioactivity that co-migrated with nortriptyline and amitriptyline-N-oxide increased slowly and linearly, and each never exceeded 10% of the initial radioactivity applied (Fig. 3). The portion of radioactivity co-migrated with nortriptyline and amitriptyline-N-oxide in clay soil was very low (<1%), which was close to the detection limit (the purity of 3H-amitriptyline is ∼99%). It must be noted that the loss of one methyl group from amitriptyline-[N-methyl-3H] yields a nortriptyline-[N-methyl-3H] product with half the specific activity of the parent. Thus, on a molar basis, the nortriptyline yields are double those represented by the accumulated radioactivity, namely 11.2, 1.6, and 5.2% in the loam, clay loam, and sandy loam, respectively. In parallel experiments, the dissipation of unlabeled amitriptyline and accumulation of nortriptyline and nortriptyline-N-oxide in soil supplemented with 1 µg g−1 moist soil of amitriptyline was determined by HPLC-UV. Results were entirely consistent with those observed with 3H-amitriptyline (data not shown).
From Figures 2 and 3, we see significant differences exist in the dissipation kinetics of amitriptyline in the three different agriculture soils. The dissipation of amitriptyline in loam soil is much faster than that in clay and sandy soil. The reason may be due to the higher microorganism activities in loam soil, which has higher organic content.
The fit to a first-order dissipation kinetics model was good, with r2 values ranging from 0.92 to 0.99 (Table 2). Using this model, DT50s of amitriptyline ranged from 34.1 ± 3.2 d for the loam soil to 85.3 ± 3.2 d for the sandy soil. As mentioned in the Experiment section, the radioactivity carried in amitriptyline was used to calculate the DT50 in loam soil. The total extractable radioactivity was used to estimate the DT50 in clay and sandy soil (Supplemental Data, Fig. S1), because the total amount of metabolites was less than 3% of added parent compound in clay and sandy soil (Fig. 3).
Table 2. Kinetic parameters for amitriptyline dissipation in three agriculture soils with varying texture using a first-order modela
The removal of extractable radioactivity was used to estimate the time (in days) to dissipate 50% of the added parent compound (DT50).
k = rate constant; r2 = linear correlation coefficient
34.1 ± 3.2
0.0172 ± 0.0005
0.9781 ± 0.0025
56.1 ± 2.5
0.0124 ± 0.0006
0.9713 ± 0.0281
85.3 ± 3.2
0.0081 ± 0.0003
0.9203 ± 0.0267
Three distinct peaks were detected in extracts of the loam soil by HPLC-TOF-MS (Fig. 4). Under the analytical conditions employed, the retention times of authentic amitriptyline and nortriptyline standards were 11.35 min and 11.10 min, respectively (Fig. 4). The protonated molecular ion ([MH]+) of authentic amitriptyline and nortriptyline was at m/z of 278 and 264, respectively. Comparing the retention times and mass spectra of the soil extracts with those of authentic standards, peaks 1 and 2 were identified as nortriptyline and amitriptyline, respectively.
Peak 3, eluted after amitriptyline (Fig. 4), was identified as amitriptyline-N-oxide. Peak 3 has m/z of 294, which is 16 amu heavier than that of amitriptyline. It is suspected to be hydroxyamitriptyline or amitriptyline N-oxide. Hydroxyamitriptyline is more polar than amitriptyline, thus, it eluted earlier than amitriptyline and nortriptyline (retention time = 5 min). By increasing the cone voltage of TOF-MS, the ion fragmentation increased. The main fragments of protonated hydroxyamitriptyline (m/z 294) are ions of m/z 276 (loss of water, MH+-H2O) and 231 (loss of water and neutral N group, MH+-H2O-NH(CH3)2). By comparing retention times and mass spectra of standard hydroxyamitriptyline, peak 3 was unambiguously determined not to be hydroxyamitriptyline. The main fragment ion of peak 3 (m/z 294) is m/z of 233. The relative intensity of ion m/z 233 increased when the cone voltage increased (Supplemental Data, Fig. S2). The same fragment ion m/z 233 was observed for amitriptyline and notriptyline when the cone voltage increased, which confirms that peak 3 and amitriptyline and nortriptyline have the same dissociation pathways, namely, loss of the nitrogen containing group as a neutral molecule from the protonated compounds. Holman et al. 24 reported the diagnostic losses of N, N-dimethylamine, and N, N-dimethylhydroxylamine from protonated dialkyl tertiary amine-containing pharmaceuticals and dialkyl tertiary amine-N-oxides, respectively, by electrospray ionization-low energy collision-induced dissociation tandem mass spectrometry. They concluded that oxidation of the dialkyl tertiary amine group did not change the dissociation behavior of the compound. The main fragment ions of protonated amitriptyline and amitriptyline-N-oxide are m/z of 233, which is formed through the loss of the nitrogen-containing group 24. From the data above, peak 3 was identified as amitriptyline-N-oxide. The mass fragmentation pathways (with electrospray ionization mass spectrometry) of amitriptyline and its metabolites are shown in Figure 5.
There were no target analytes (amitriptyline, nortriptyline, and amitriptyline-N-oxide) detected in soil that was not supplemented with amitriptyline (Supplemental Data, Fig. S3). This confirms that nortriptyline and amitriptyline-N-oxide were the transformation products of amitriptyline in soils and are not contaminants in the soil matrix.
Adding the liquid municipal biosolids to the soil had no effect on the dissipation of amitriptyline in the loam soil (Fig. 6). Assuming first-order kinetics, the total extractable radioactivity dissipated with TD50s of 46.2 ± 2.7 d in the presence and 49.5 ± 3.5 d in the absence of biosolids. The transformation products nortriptyline and amitriptyline-N-oxide were also identified in LMB-supplemented soils, and the LC-TOF-MS chromatograms were similar to that of soil extracts without the addition of LMB (Supplemental Data, Fig. S3).
Kinetics and pathways of nortriptyline dissipation in a loam soil
Given the prominence of nortriptyline as a key transformation product and its neurological bioactivity, its persistence was explored further in the loam soil (Fig. 7). Following 50 d of incubation at 30°C, LC-TOF-MS/UV analysis of soil extracts indicated that about 43% of the initially applied nortriptyline was recovered. The DT50 for nortriptyline was determined to be 40.5 ± 3.2 d. Soil extracts obtained on day 50 contained two hydroxylated metabolites, each representing less than 10% of the nortriptyline added.
Extracts prepared at the end of an incubation of the loam soil with nortriptyline contained three main peaks resolvable by LC-TOF-MS with retention times of 11.02, 11.94, and 12.53 min, respectively (Fig. 8). The peak with the retention time of 12.53 min was identified as nortriptyline by comparing the retention time and mass spectra with that of a standard. Peak 1 and 2 have the same ions of m/z 280, 16 amu higher than protonated parent nortriptyline ion; thus, peaks 1 and 2 are oxidative biotransformation products (hydroxylation–oxidation of aliphatic or aromatic carbon) or N-oxidation (N-oxides) of nortriptyline. In the mass spectra of peak 1 (Fig. 8), there are two major ions, m/z 280 [MH+] and m/z 262 (loss of water, MH+-H2O). When increasing the cone voltage, the intensity of m/z 262 increased, the ion of m/z 280 decreased, and a new peak at m/z of 219 (MH+-H2O-C2H5N) appeared. In the spectra of peak 2, the ion with m/z 280 was the dominate ion, accompanying with a small ion of m/z 237. When increasing the cone voltage, the intensity of m/z 237 increased and a new peak at m/z 219 appeared. Based on the mass spectra data and literature data, peaks 1 and 2 were identified as 10-hydroxynortriptyline (10-OH-nortriptyline) and 2-OH-nortriptyline, respectively.
It has been reported that E-10-OH-nortriptyline is the major metabolite of nortriptyline in animals and humans 11, 17, 25, 26. The E-2-OH-nortriptyline follows E-10-OH-nortriptyline in abundance as a nortriptyline metabolite in rat bile and also occurs in rabbit urine 17. There is a predominant loss of water from metabolites with aliphatic hydroxylation, while the loss of water was not favored when hydroxylation was phenolic 27. Aliphatic hydroxylation could be distinguished readily from aromatic hydroxylation based on the extent of water loss 27. N-oxides produced distinct [M + H - O]+ ions 28–30. These [M + H - O]+ ions were not produced in the mass spectra of hydroxylated metabolites. From the extensive water loss of peak 1, it is assigned as 10-OH-nortriptyline. Peak 2 is assigned as 2-OH-nortriptyline. Overall, the rationale for concluding that peak 2 was not nortriptyline-N-oxide are as follows. First, peak 2 does not have the distinct [M + H - O]+ ion. Second, peak 2 also does not have the distinct ion at m/z of 233 (loss of N-containing neutral group) as amitriptyline-N-oxide 24. Third, the fragment ion of m/z 237 was from loss of -C2H5N from protonated nortriptyline [MH-C2H5N]+, which cannot be formed from nortriptyline-N-oxide. Fourth, ion m/z 236 ([M-C2H5N]+]) has been reported as a main fragment of amitriptyline-N-oxide by electron impact ionization 17. The mass fragmentation scheme of hydroxylated nortriptyline with electrospray ionization mass spectrometry is shown in Figure 9.
Amitriptyline dissipated in three soils with DT50s ranging from 34.1 to 85.3 d at 30°C. Two biologically active transformation products, nortriptyline and amitriptyline-N-oxide, were detected with the former reaching a maximum concentration of 11% of the parent. Nortriptyline was dissipated in the loam soil with a DT50 of 40.5 ± 3.2 d. Less than 10% of nortriptyline was metabolized to hydroxylated compounds following a 50-d incubation. In the present study, N-demethylation and N-oxidation of amitriptyline were the main metabolic pathways of amitriptyline, and C-hydroxylation of nortriptyline was the main metabolic pathway of nortriptyline. Adding municipal biosolids to the soil had no effects on the amitriptyline dissipation rates.
Amitriptyline and nortriptyline are basic tertiary amines with pKas of 9.4 and 10.0, respectively 31. Using current methods, environmental risk assessments for medicines that are ionized at environmental pH values are problematic in that they assume soil sorption characteristics (i.e., KOC, KOW) that are not realistic and that will lead to predicting environmental fate incorrectly. A case study with the serotonin reuptake inhibitor fluoxetine concluded that new methods to predict the fate of ionisable compounds in soils are needed, as well as further research concerning factors determining environmental exposure to these compounds 32. The present study establishes under controlled laboratory conditions kinetics and pathways of dissipation of these psychiatric drugs and that the products of soil amitriptyline transformation are bioactive.
This research was funded by Agriculture and Agri-Food Canada and by Health Canada (New Substances Assessment and Control Bureau). H. Li was funded through the Natural Sciences and Engineering Research Council of Canada Visiting Fellowship in Government Laboratories program. We thank L. Sabourin for obtaining the biosolids samples and A. Scott for assisting with the experiments.
Fig. S1. Plot of ln(Ct/C0) for experiments to determine the dissipation of amitriptyline in sandy and clay soil. Soils were supplemented with 1 µg g−1 and 200 k dpm g−1 moist soil of 3H-amitriptyline. Soil moisture content was adjusted to 15% before incubation. C0, Ct = total extractable radioactivity in the soil at time 0 and t, respectively.
Fig. S2. Cone voltage effects on the fragments patterns of amitriptyline-N-oxide, amitriptyline and nortriptyline.
Fig. S3. Chromatograms of soil extracts from soils incubated for 0 and 35 d, blank soil (no spiking of amitriptyline), and soils incubated for 35 d in the presence of biosolids. (A) Blank soil; (B) soil at the beginning of incubation (day 0); (C) soil incubated for 35 d; (D) soil incubated for 35 d in the presence of biosolids. (82 KB DOC).