Effects of erythromycin, tetracycline and ibuprofen on the growth of Synechocystis sp. and Lemna minor

Aquatic Toxicology 67 (2004) 387–396

Effects of erythromycin, tetracycline and ibuprofen on the growth of Synechocystis sp. and Lemna minor Francesco Pomati a,b , Andrew G. Netting a , Davide Calamari b , Brett A. Neilan a,∗ a

School of Biotechnology and Biomolecular Sciences, The University of New South Wales, Sydney 2052, NSW, Australia b Environmental Research Group, DBSF, University of Insubria, Via J.H. Dunant 3, 21100 Varese, Italy Received 26 October 2003; received in revised form 3 February 2004; accepted 4 February 2004

Abstract Pharmaceutically active substances have recently been recognised as an emerging environmental problem. Human and veterinarian therapeutic agents can contaminate aquatic ecosystems via sewage discharges (human and animal excretion), improper disposal or industrial waste. Very little is known on the effects of pharmaceutical pollutants on aquatic photosynthetic organisms. In this study the effects of erythromycin, tetracycline and ibuprofen on the growth of the cyanobacterium Synechocystis sp. PCC6803 and the duckweed Lemna minor FBR006 were studied at concentrations of 1–1000 ␮g l−1 . At dosage of 1 mg l−1 , erythromycin affected the growth of both Synechocystis and Lemna with a maximum inhibition of 70 and 20%, respectively. Tetracycline had inhibitory effects (20–22% reduction in growth) on Synechocystis at intermediate dosages. The same aminoglycoside antibiotic promoted growth in Lemna by 26% at 10 ␮g l−1 , while frond development was reduced at 1 mg l−1 (tetracycline). The anti-inflammatory ibuprofen strongly stimulated the growth of Synechocystis at all concentrations tested (72% increase at 10 ␮g l−1 ) although inhibited Lemna in a linear dose-dependent manner with a 25% reduction over control levels at a dosage of 1 mg l−1 . The 7 days effective concentration (EC50 ) calculated for Lemna were 5.6, 1 and 4 g l−1 , respectively, for erythromycin, tetracycline and ibuprofen. Moreover, exposure to the three pharmaceuticals resulted in the production of the stress hormone, abscisic acid (ABA), in Lemna. Erythromycin and tetracycline were more effective in promoting ABA synthesis compared to ibuprofen. The effects shown by the three therapeutic drugs on Synechocystis and Lemna growth may have potential implications in the assessments of residual environmental risks associated with the presence of pharmaceuticals in freshwater ecosystems. Promotion of ABA synthesis in Lemna by the two antibiotics and by copper suggests that the plant hormone could be a suitable (additional) indicator for future evaluation of phytotoxicity that results in plant senescence. © 2004 Elsevier B.V. All rights reserved. Keywords: ABA; Cyanobacteria; Duckweed; Erythromycin; Ibuprofen; Tetracycline

1. Introduction

∗

Corresponding author. Tel.: +61-2-9385-3235; fax: +61-2-9385-1591. E-mail address: [email protected] (B.A. Neilan).

Some hormonally active chemicals, capable of affecting reproduction and/or inducing carcinogenesis in wildlife as well as humans, have been found to be globally ubiquitous contaminants in the last decade

0166-445X/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2004.02.001

388

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

(Fenner-Crisp, 1997). Furthermore, pharmaceutically active substances now appear to be an emerging problem. This class of pollutants includes human and veterinarian therapeutic agents and active ingredients in cosmetic products (for review, see Daughton and Ternes, 1999). Such compounds are disposed or discharged into waters primarily via domestic/industrial sewage and terrestrial runoff (Halling-Sorensen et al., 1998). Erythromycin, tetracycline and ibuprofen are pharmaceutical agents used worldwide in both human medicine and in veterinary practice. The annual environmental load in Italy for 1997 has been estimated in the range of tens of tons for antibiotics and in the range of hundreds of kilograms for anti-inflammatory drugs (Zuccato et al., 2001). These compounds can be excreted with urine or faeces, partially metabolised or unmodified. Detectable quantities of erythromycin, tetracycline and ibuprofen have been found in several typologies of European surface waters and sediments, with concentrations ranging from ng l−1 to ␮g l−1 (Richardson and Bowron, 1985; Stumpf et al., 1996; Stan and Heberer, 1997; Hirsh et al., 1999; Zuccato et al., 2000). Here, we report the effect of these three pharmaceuticals on a planktonic prokaryote and an aquatic plant. Pharmaceuticals are specifically designed to penetrate biological membranes and reach universal molecular systems (enzymes, receptors, etc.), thereby increasing the probability of unexpected consequences across a number of species. The potential effects of these agents on non-target organisms are mostly unknown. Recently it has been suggested that there is a need to develop ecotoxicity screening procedures that take into consideration the modes of action of therapeutic drugs on non-target species (Daughton and Ternes, 1999), such as algae and aquatic plants (Wang, 1991; Cleuvers and Ratte, 2002; Frankart et al., 2002). The objective of this study was to determine the effects of erythromycin, tetracycline, and ibuprofen on the growth of the cyanobacterium Synechocystis sp. PCC6803 and the duckweed Lemna minor FBR006. We also investigated the effects of the three pharmaceuticals on the production of the plant stress-hormone abscisic acid (ABA). The strategic aim was to determine whether ABA can be used as an indicator of phytotoxicity in duckweed over a 90 min exposure test. ABA is a phytohormone that has a

regulatory role in many plant physiological processes (for review, see Fedoroff, 2002). ABA mediates stress tolerance responses, being a key signal compound controlling stomatal aperture and is implicated in mediating seed dormancy, growth regulation, leaf senescence and abscission. ABA is a sesquiterpenoid (15-carbon, Fig. 4) which is partially synthesised via the mevalonic pathway in chloroplasts and other plastids primarily of leaves. Biosynthesis occurs indirectly through the production of the carotenoid violaxanthin by isomerisation and division followed by an oxidation reaction. The one molecule of xanthonin produced in this way is unstable and spontaneously reconfigures to the ABA aldehyde, which is then further oxidised to free ABA. ABA is produced by a variety of organisms and microorganisms and it is considered a universal Ca++ agonist across taxonomic kingdoms (Minorsky, 2002). Current evidence, however, doesn’t indicate a defined physiological role for this compound in cyanobacteria.

2. Materials and methods 2.1. Chemicals Erythromycin, tetracycline and ibuprofen were obtained from ICN (ICN Biomedicals Australasia Pty Ltd., NSW, Australia). Solutions were prepared on day prior to use in each experiment, and then diluted in the culture media to reach the final tested concentrations. CuSO4 (Fluka, Buchs, Switzerland) was also used as positive control in duckweed toxicity tests at a concentration of 0.3 mg l−1 . Analytical chemicals were from Fluka (Buchs, Switzerland). Solvents for use in the PrepStation and with the gas chromatograph/mass spectrometer (GC/MS) were of ‘Omnisolv’ grade from EM Science (Gibbstown, NJ). 2.2. Tested organisms The cyanobacterium Synechocystis sp. PCC6803 was obtained from the Pasteur Culture Collection, Paris, France. The strain FBR006 of L. minor was isolated from wild samples of duckweed collected in Centennial Park, Sydney, Australia, during December 2001. Plants were disinfected by immersing the fronds in 70% ethanol and sterile MQ-water as previ-

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

ously reported (Jenner and Jassen-Mommen, 1989). The samples were disinfected and inoculated in fresh medium every second day, and the isolate FBR006 was ultimately chosen for the toxicity tests. 2.3. Experimental conditions Both Synechocystis sp. PCC6803 and L. minor FBR006 were grown in BG-11 freshwater medium (Rippka et al., 1979) adjusted to pH 7.0. Cultures of the two organisms were maintained in 50 and 100 ml volumes in sterilised flasks and beakers, respectively, under the standard conditions of 26 ◦ C, continuous white light (daylight neon lamp 60 W) at an intensity of 25 ␮mol photon m−2 s−1 (QSL-2100 Scalar PAR Irradiance Sensor, Biospherical Instruments Inc., San Diego, CA).

389

manner. Three replicates were prepared by transferring the plants into sterilised beakers with 100 ml BG-11 medium dosed to the final concentration with each pharmaceutical. Plants of similar size were selected and a total of 12 fronds per test chamber were prepared. Lemna was exposed to the test chemicals for a period of 7 days with the growth medium and test solution replaced 5 days after the onset of the experiment to prevent nutrient limitation or depletion of the tested pharmaceuticals. Test chambers were inspected for changes in frond number at the beginning of the exposure period (day 0) and on days 3, 5 and 7. On day 7, the total number of fronds present in the untreated controls was used as a reference to estimate the effect of the test agents on duckweed. 2.5. Statistical analyses

2.4. Toxicity tests The incubation time chosen to investigate the effect of the three pharmaceuticals on the growth of Synechocystis was 5 days, which is comparable to previous studies on the effect of other chemical agents on cyanobacterial growth (Suzuki et al., 2000). Growth of Synechocystis cultures was monitored in 1 cm disposable cuvettes by recording the optical density at 750 nm (OD750 ) with an Ultrospec II UV/VS spectrometer (LKB Biochrom Ltd, Cambridge, UK). Culture in late-logarithmic growth phase was used as inocula for new batch cultures (50 ml) and exposed to erythromycin, tetracycline and ibuprofen at 0 (control), 1, 10, 100 and 1000 ␮g l−1 . Culture densities were monitored by means of OD750 measured at day 0, 2 and 5. The experiments were performed in triplicate and the initial cell density of the replicates was set to approximately 0.1 OD750 . Toxicity tests on L. minor were essentially performed as suggested in the guidelines proposed by the United States Environmental Protection Agency (1996). Briefly, Lemna was dosed with erythromycin, tetracycline and ibuprofen in concentration series of 0 (control), 1, 10, 100 and 1000 ␮g l−1 . CuSO4 at 0.3 mg l−1 , which has been previously shown to represent the EC50 in the duckweed 7 day toxicity test (unpublished results), was used as the positive control in all experiments to determine if the Lemna fronds were responding to a known chemical in the expected

The dose response plots and effective concentration calculations were performed by means of probit analysis (Finney, 1971, 1978) using the software PriProbit ver. 1.62 (1996–2000, Masayuki Sakuma, Japan) under all or nothing complementary LOG–LOG model parameters. The fiducial limits of potency were elaborated utilising the SAS-equivalent method and the LOG-likelihood function maximized by the Variable Metric method (Davidon–Fletcher–Powell algorithm). The number of responders needed for the elaboration of dose-response curves and toxicological parameters was calculated as the average number of control fronds minus the average number of fronds in the sample. All other statistical analyses and plots were performed using the software for PC Origin 5.0 (Microcal Software, Northampton, MA). 2.6. Short term toxicity test, ABA extraction, sample preparation, derivatisation, and gas chromatography/mass spectrometry analysis Experiments designed to evaluate the influence of pharmaceuticals on ABA production were performed essentially as the standard Lemna test described above. Plants of similar size were selected and a total of 12 fronds per test chamber were prepared. Lemna was transferred in duplicates to sterilised beakers with 100 ml BG-11 medium and let to adapt to the test chamber for 2 h. CuSO4 , erythromycin, tetracy-

390

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

cline, and ibuprofen were then diluted into the growth medium to reach the final concentrations of 0.3, 5.6, 1 and 100 mg l−1 , respectively. These concentrations represented the corresponding EC50 as estimated below, while 100 mg l−1 of ibuprofen was chosen being the maximum solubility of this agent in water at 25 ◦ C. CuSO4 has been previously shown to promote ABA synthesis in Lemna leaves (unpublished results) and therefore utilised in this study as positive control. Duplicates of untreated samples were used as negative controls. In this experiment, Lemna was exposed to test concentrations for a period of 90 min and then the fronds were collected and extracted for ABA analysis as previously reported (Duffield and Netting, 2001). Briefly, plants were homogenised for 90 s in 2 ml 0.1 M tetraethylammoniumformate (TEAF) buffer (pH 6.0): acetone, 1:9. Duckweed extracts were then filtered (Millipore AP, Millipore Australia Pty. Ltd., North Ryde, NSW, Australia) and evaporated to dryness under a N2 stream. The pellets were resuspended in 1 ml of citrate buffer (0.2 M, pH 6.3) followed by the addition at 1 ml of 0.02 M cetyltrimethylammonium bromide (CTMA Br). Samples were then extracted three times with dichloromethane (DCM):iso-propanol (IPA), 4:1, with the organic acids, including ABA, extracted into the organic phase as ion pairs. The three organic phases were combined and evaporated to dryness. The excess CTMA Br was removed after adding 1 ml of 0.01 M H2 SO4 and extracting three times with diethyl ether. Evaporation of the ether rendered the free ABA fraction. For sample preparation, the following procedure was completed on a PrepStation (SPE Module: Hewlett Packard, Wilmington, Delaware, USA). The dry ABA fractions from the duckweed extracts were dissolved in 10 ␮l dimethyl acetamide (DMA):tetramethylammonium hydroxide (TMAOH), 5:1. This was followed by the addition of 10 ␮l DMA:pentafluorobenzyl (PFB) bromide, 10:3. The total volume was mixed for 1 min at room temperature. Water:butanol:hexane, 2:1:10 (260 ␮l) was added and mixed for 12 s. The hexane layer was extracted and evaporated to give PFB–ABA. A silica column (0.1 g) for each sample was washed with 200 ␮l hexane and the PFB–ABA, in 200 ␮l hexane, was transferred onto it. The column was then washed with 700 ␮l dichloromethane and the PFB–ABA

eluted with 150 ␮l t-butylmethyl ether, which was then evaporated. This method gave a yield of 80% based on the recovery of [G-3H] ABA in PFB–ABA as previously published (Duffield and Netting, 2001). A ThermoQuest (Austin, Texas, USA) GCQ gas chromatograph/mass spectrometer, operating in the negative chemical ionisation mode with methane as the reagent gas, was used to detect free ABA as PFB–ABA in the duckweed extracts. For use with the standard cool-on-column (COC) injector, 20 ␮l of toluene was added to the sample and 1 ␮l injected. The injector was held at 60 ◦ C for 0.1 min and then programmed to increase by 100 ◦ C min−1 to a final temperature of 290 ◦ C and then held for 1.6 min. The column oven was also initially held at 60 ◦ C for 0.1 min and then programmed to rise by 40 ◦ C min−1 to 250 ◦ C, hold for 2 min and then increase by 10 ◦ C min−1 to 300 ◦ C and held for 3.15 min. The column used was a Restek Rtx-5MS, 220 ␮m × 25 m (Chromalytic Technology, Boronia, VIC, Australia). Electronic pressure control was used so that a constant linear velocity of helium of 400 mm s−1 was obtained while mass spectra were collected from 5 to 15 min. An AS 2000 autosampler (ThermoQuest CE Instruments, Rodano, Italy) was used for injecting samples with a 530 ␮m diameter retention gap from 2 to 5 m length. In the relevant cases PFB–ABA was quantified by measuring the peak areas for m/z = 263, over the analytical detection limit of 0.63 ng, and comparing them to the ABA standard curve previously determined (Duffield and Netting, 2001).

3. Results 3.1. Effect of erythromycin on growth Growth of Synechocystis sp. PCC6803 was reduced at 1, 100 and 1000 ␮g l−1 erythromycin over a 5 days period of exposure (Fig. 1A and B). One micron per liter of the antibiotic resulted in 15% lower density compared to the control cultures, while dosages of 100 and 1000 ␮g l−1 showed a decrease of 66 and 70%, respectively (Fig. 1B). On the other hand, exposure of the cyanobacterial cultures to erythromycin at 10 ␮g l−1 promoted growth by an average of 20% compared to the controls after the fifth day of incubation.

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396 40

(A)

391

(B)

20 0

% variation / control

-20 -40 -60 -80

15 10 5

0

1

3

2

4

5

(C)

1 (D)

10

1

10

100

1000

0 -5 -10 -15 -20 -25 -30

0

1

2

3

4

5

6

7

days

8

100

µg/L

1000

Fig. 1. Effects of erythromycin at 1 (䊐), 10 (䊊), 100 () and 1000 () ␮g l−1 on the growth of Synechocystis sp. (A and B) and L. minor (C and D). CuSO4 at 0.3 mg l−1 (×) was used as a positive control in the duckweed test. (A and C): Time courses of erythromycin exposure; (B and D): dose-response plots for data at day 5 and 7, respectively.

The aquatic plant L. minor responded to erythromycin with a linear and dose-dependent decrease in growth (Fig. 1C and D) over the 7 day period of the test. While exposure of fronds to 1 ␮g l−1 of the antibiotic resulted in no detectable negative effects on fronds compared to the controls, 100 and 1000 ␮g l−1 doses inhibited Lemna growth by 15 and 20%, respectively. Exposure of duckweed to erythromycin at 10 ␮g l−1 promoted the growth of one of the three experimental replicates in comparison with the untreated plants, while the other two were negatively affected. CuSO4 , employed at 0.3 mg l−1 as positive control in the phytotoxicity test, inhibited Lemna growth by 22% after 7 days of exposure (Fig. 1C). 3.2. Effect of tetracycline on growth Tetracycline inhibited the growth of Synechocystis in comparison with unexposed cultures for doses of

10 and 100 ␮g l−1 (Fig. 2A and B), giving an average percentage reduction in cell density of 20 and 22% for the two concentrations, respectively. Exposure of cyanobacterial cells for 5 days to 1 ␮g l−1 of the antibiotic showed no evident effect in comparison with the controls, while the highest dose tested (1000 ␮g l−1 ) resulted in an average Synechocystis culture density 9% higher than control levels. The effect of tetracycline on the growth of Lemna was markedly stimulatory for exposure to levels of 1 and 10 ␮g l−1 , resulting in 18 and 26% increase compared to untreated plants, respectively (Fig. 2C and D). Dosing of duckweed with 100 ␮g l−1 of the antibiotic slightly promoted frond growth (2.0%) after 7 days, while tetracycline at 1000 ␮g l−1 , on the other hand, strongly inhibited Lemna cultures development (−43%) (Fig. 2D). The positive control, CuSO4 at 0.3 mg l−1 , reduced Lemna growth by 26% in comparison to the untreated cultures.

392

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396 20

(A)

(B)

10 0

% variation / control

-10 -20 -30 -40 0

50 40 30 20 10 0 -10 -20 -30 -40 -50

1

2

3

4

5

1

0

10

100

1000

(D)

(C)

1

2

3

4

5

6

7

days

8

1

10

µg/L

100

1000

Fig. 2. Effects of tetracycline at 1 (䊐), 10 (䊊), 100 () and 1000 () ␮g l−1 on the growth of Synechocystis sp. (A and B) and L. minor (C and D). CuSO4 at 0.3 mg l−1 (×) was used as a positive control in the duckweed test. (A and C): Time courses of tetracycline exposure; (B and D): dose-response plots for data at day 5 and 7, respectively.

3.3. Effect of ibuprofen on growth The non-steroidal anti-inflammatory drug ibuprofen, at concentrations of 1–1000 ␮g l−1 , stimulated the growth of Synechocystis sp. PCC6803 over the 5 days of exposure (Fig. 3A and B). The highest increase in average culture density, expressed as percentage over control levels, was detected at 10 ␮g l−1 ibuprofen (72%), followed by 100 ␮g l−1 (42%), 1000 ␮g l−1 (31%) and 1 ␮g l−1 (12%) (Fig. 3B). With L. minor, ibuprofen inhibited the growth after 7 days of exposure (Fig. 3C and D) at all concentrations tested, with the strongest effect observed at 1000 ␮g l−1 which is equal to 25% reduction over the control levels. Dosing of duckweed with 1 ␮g l−1 of the anti-inflammatory agent resulted in no negative effect on growth compared to the other treatments over the first 5 days of the test (4%) (Fig. 3C). After the replacement of test solutions on day 5, however, the average percentage of growth dropped to −14%. Growth of duckweed samples exposed to CuSO4 at

0.3 mg l−1 were inhibited by 30% after 7 days of treatment. 3.4. Dose-response analysis of Lemna The growth inhibition data, collected for erythromycin, tetracycline, and ibuprofen at day 7 of the Lemna experiments, were employed to perform standard toxicity analyses. The values utilised for probit and logistic analysis were computed by means of all or nothing complementary LOG–LOG model (SASequivalent), corrected for the natural response rate, which was estimated to be zero for erythromycin and ibuprofen and below 1.65 ␮g l−1 for tetracycline (at 99% limits). By means of the above models the effective doses were estimated, together with the slopes representing the angular coefficients of the logistic plots for the three pharmaceuticals. The 7 days EC50 (99% confidence limits) values for erythromycin, tetracycline and ibuprofen were 5.62 (slope 1.19), 1.06 (slope 3.47), and 4.010 (slope 0.19) mg l−1 , respectively.

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

80

(A)

393

(B)

60

% variation / control

40 20 0 -20

0

5

1

3

2

4

1

5

10

100

1000

100

1000

(D)

(C)

0 -5

-10 -15 -20 -25 -30 -35

0

1

2

3

4

5

6

7

8

1

days

10

µg/L

Fig. 3. Effects of ibuprofen at 1 (䊐), 10 (䊊), 100 () and 1000 () ␮g l−1 on the growth of Synechocystis sp. (A and B) and L. minor (C and D). CuSO4 at 0.3 mg l−1 (×) was used as a positive control in the duckweed test. (A and C): Time courses of ibuprofen exposure; (B and D): dose-response plots for data at day 5 and 7, respectively.

3.5. Effect of erythromycin, tetracycline and ibuprofen on ABA production in Lemna Fig. 4 shows the mass GC/MS chromatogram at m/z = 263 for PFB–ABA corresponding to an extract

of Lemna exposed to CuSO4 . ABA copper sulphate was quantified by GC/MS at 5.78 ± 0.86 pg on column over the analytical detection limit. In unexposed duckweed fronds free ABA could not be detected. Ibuprofen had little effect on ABA production, which could only be detected in traces in the samples extracted after exposure to the anti-inflammatory drug. ABA was quantified in Lemna exposed to erythromycin and tetracycline as 4.42 (±0.54) and 1.85 (±0.21) pg of total ABA on column, respectively.

4. Discussion

Fig. 4. Representative GC/MS chromatogram of L. minor exposed for 90 min to CuSO4 at 0.3 mg l−1 . Abscisic acid (ABA) was traced at m/z = 263 in the negative chemical ionisation mode, using methane as the reagent gas.

Erythromycin, a macrolide antibiotic that blocks bacterial protein synthesis, had negative effects on the growth of both Synechocystis and Lemna at all concentrations tested, other than 10 ␮g l−1 . At that dose, the antibiotic seemed to stimulate the growth of the cyanobacterial cultures. Erythromycin has been shown to have a similar effect on some other filamentous

394

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

cyanobacteria, in which growth was stimulated only for certain intermediate concentrations (unpublished results). The expected inhibitory effect of the antibiotic at 10 ␮g l−1 was observed both in Synechocystis and Lemna at day 2, although turned into a stimulating effect after 5 days. In Lemna, the medium was renewed at day 5, and the inhibitory effect due to the addition of freshly prepared erythromycin is evident at day 7 (Fig. 1). Since erythromycin is considered to be stable in water solution for long periods (Richardson and Bowron, 1985), these results suggest that the antibiotic can be metabolised by the cyanobacteria and the duckweed. Erythromycin might be degraded into metabolites that are no longer toxic for these organisms, or even growth stimulating. Further investigation is needed to support the possibility of positive effects of breakdown products of these antibiotic on aquatic phototrophs. The inhibitory effect on Lemna exerted by erythromycin was indeed novel. To date, no reports are available in the literature on the effect of this antibiotic on aquatic plants. Tetracycline is an aminoglycoside that is active against a broad spectrum of prokaryotes via its inhibition of protein synthesis by preventing the association of aminoacyl-tRNAs with the bacterial ribosomes. The antibiotic inhibited the growth of Synechocystis at 10 and 100 ␮g l−1 while no negative effect on growth was seen at 1000 ␮g l−1 (Fig. 2B). Such an high tetracycline concentration may have induced the cyanobacterial production of exoenzymes able to convert the antibiotic into breakdown molecules that had a subtle stimulating influence on Synechocystis growth. Although this could be the result of natural variability in tetracycline resistance, the antibiotic seemed slightly stimulatory to L. minor growth for doses between 1 and 100 ␮g l−1 . On the contrary, a tetracycline concentration of 1 mg l−1 was markedly inhibitory for Lemna growth. The sensitivity and type of response to tetracycline are known to be dependent upon the transport of tetracycline molecules across the cytoplasmic membrane (Prescott et al., 1996). The opposing responses found for the two organisms tested may, therefore, suggest differences in this uptake mechanism. Ibuprofen is a non-steroidal anti-inflammatory drug that has been shown to significantly affect the growth of several bacterial and fungal species (Sanyal et al.,

1993; Chowdhury et al., 1996). Also in this study, the growth of both Synechocystis and Lemna was inhibited by freshly added ibuprofen (Fig. 3A and C). The effect, however, turned into a growth stimulation after the second day. Medium replacement at day 5 renewed the inhibitory effect of ibuprofen in Lemna. Ibuprofen is an instable chemical, degraded in aquatic environments with a t50 < 1 day (Richardson and Bowron, 1985). Our evidence therefore suggests that ibuprofen metabolites are non toxic for aquatic organisms tested, and they may also have growth stimulating properties. The optimum growth promoting concentration, however, was found to be 10 ␮g l−1 of the anti-inflammatory, with lower growth stimulation detected for higher doses. The growth stimulating effect of ibuprofen in Synechocystis was similar, though stronger, to those previously reported on cyanobacteria for other synthetic pharmaceuticals (Suzuki et al., 2000). On the other hand, compared to the ibuprofen non-observed effective concentrations (NOEC) of tens and hundreds of mg l−1 published for algae, Daphnia, mysid and fish (Knoll BASF, 1995), duckweed was demonstrated to be the most sensitive eukaryotic organism tested against this agent. Only a slight decrease of plant growth along with increasing ibuprofen concentrations was also observed, suggesting a very sensitive response of Lemna to the anti-inflammatory at low concentrations. During the course of the present study we also investigated the effect of erythromycin, tetracycline and ibuprofen on ABA production by duckweed. The quantity of ABA in a plant extract is thought to correlate with the degree of general metabolic stress suffered by the organism, resulting in reduced growth rates (Netting, 2000). However, the tested pharmaceuticals elicited different responses on ABA production by L. minor. Both the antibiotics promoted release of the stress hormone as seen in comparison with the effect of CuSO4 , with erythromycin being the most effective in stimulating ABA biosynthesis. Copper sulphate has been previously found to induce the production of ABA in Lemna (unpublished results). Copper is known to activate cellular responses involved in oxidative stress (Mira et al., 2002), which triggers the release of free ABA (Fedoroff, 2002). Although the modes of action of erythromycin and tetracycline on bacteria are well known, no consis-

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

tent information is available on their effect on aquatic plants. It is possible that these two antibiotics have adverse effects on Lemna metabolism, leading to intracellular oxidative imbalance and ABA release in a similar way to what is seen after CuSO4 exposure. Ibuprofen, however, had little effect on ABA synthesis. This anti-inflammatory agent has low solubility in water compared to the other two compounds and was used in the duckweed test at a concentration of 100 mg l−1 , much lower than the EC50 (4.01 g l−1 ). The low stimulation of ABA production due to ibuprofen, as seen by its elevated EC50 value and its low solubility in water suggest that not all pharmaceutical agents could be screened for toxicity using a single test based on the detection and quantification of ABA. Due to the low concentrations of ABA analysed in this study, the results presented should be considered only semi-quantitative and indicative of the activation of ABA synthesis. However, the sensitivity of the method proposed can be widely improved by increasing the initial number of test fronds. Future sensitivity improvement and testing of more contaminating compounds may validate the use, for some classes of pollutants, of ABA as an indicator of toxicity in duckweed over a 90 min exposure assay. The use of an ELISA kit for analysis and quantification of the hormone would lead to a quick and sensitive ABA-toxicity test that could be performed directly in the field. In conclusion, during the course of this study we observed varying and sometimes opposite effects of erythromycin, tetracycline and ibuprofen on the growth of Synechocystis sp. PCC6803 and L. minor FBR006. Divergent responses of higher plants and algae have already been reported for certain environmental contaminants (Wang, 1991). Both organisms, however, responded to the presence of pharmaceuticals in water and were found to be among the most sensitive species so far tested against therapeutic drugs. Eukaryotic algae have been found to be affected by antibiotics in the range of 10–100 mg l−1 with higher organisms being even less sensitive (Kumpel et al., 2001). For ibuprofen, both species used in this study were found to be at least an order of magnitude more sensitive than previously studied organisms, including bacteria and vertebrates (Sanyal et al., 1993; Knoll BASF, 1995). The results presented in this study, however, suggest that the effects of the three pharmaceuticals

395

investigated cannot be easily interpreted without taking into account their persistence in the environment and the toxicity of their possible metabolites. Unfortunately, the literature regarding breakdown products of therapeutic agents and their biological effects lacks of information. Many questions cannot be answered at this stage. Further investigation is needed to elucidate the effect of degradation products of pharmaceuticals in the environment, as well the effect of drug mixtures on aquatic life. One possible future strategy could be the analysis of gene expression in model aquatic organisms exposed to commercial drugs and their metabolites. This molecular biology approach could be the most sensitive in detecting the effects of such chemicals at dosages (ng l−1 ) comparable to those found in aquatic environments (Zuccato et al., 2000). In terms of the ecological impact of therapeutic drugs on aquatic environments, the results presented here may have potential implications for the future assessment of risk associated with freshwater ecosystems and consumption of water. In some cases, concentrations of 1–6 ␮g l−1 have been detected for antibiotics and anti-inflammatory agents in surface waters (Richardson and Bowron, 1985; Hirsh et al., 1999). In this study, the EC50 values found for Lemna with erythromycin, tetracycline and ibuprofen were an order of magnitude higher than the environmental levels detected in previous studies (Zuccato et al., 2000). Our results, however, suggest that even ␮g l−1 concentrations of therapeutic drugs can affect the growth of aquatic phototrophs. Moreover, the antibiotics tested were found to promote ABA synthesis and, although the concentrations screened were higher than those detected in surface waters, the plant hormone was shown to be a suitable indicator of stress in duckweed for the future assessment of pharmaceutical phytotoxicity and water quality.

Acknowledgements The authors are grateful to Rikke Stausbøll and Betina Klint Neilsen for the useful experimental assistance. Scholarships from the School of Biotechnology and Biomolecular Sciences and The University of New South Wales to F.P. are also acknowledged.

396

F. Pomati et al. / Aquatic Toxicology 67 (2004) 387–396

References Chowdhury, B., Roy, D., Chavan, U., Mukhopadhyay, S., 1996. The anti-inflammatory, antipyretic, analgesic compound ibuprofen also has antibacterial activity against Gram-positive bacteria. Med. Sci. Res. 24, 801–802. Cleuvers, M., Ratte, H.T., 2002. Phytotoxicity of coloured substances: is Lemna duckweed an alternative to the algal growth inhibition test? Chemosphere 49, 9–15. Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ. Health Perspect. 107, 907–942. Duffield, P.H., Netting, A.G., 2001. Methods for quantitation of abscisic acid and its precursors from plant tissues. Anal. Biochem. 289, 251–259. Fedoroff, N.V., 2002. Cross-talk in abscisic acid signaling. Sci. STKE 9, RE10. Fenner-Crisp, P.A., 1997. Endocrine disruptor risk characterization: an EPA perspective. Regul. Toxicol. Pharmacol. 26, 70–73. Finney, D.J., 1971. Probit Analysis, third ed. Cambridge University Press, Cambridge, UK. Finney, D.J., 1978. Statistical Method in Biological Assay, third ed. Charles Griffin and Company, London, UK. Frankart, C., Eullaffroy, P., Vernet, G., 2002. Photosynthetic responses of Lemna minor exposed to xenobiotics, copper, and their combination. Ecotoxicol. Environ. Safe. 53, 439–445. Halling-Sorensen, B., Nielsen, S.N., Lanzky, P.F., Ingerslev, F., Lutzhoft, H.C., Jorgensen, S.E., 1998. Occurrence, fate and effects of pharmaceutical substances in the environment. Chemosphere 36, 357–393. Hirsh, R., Ternes, T., Haberer, K., Kratz, K.L., 1999. Occurrence of antibiotics in the aquatic environment. Sci. Total Environ. 225, 109–118. Jenner, H.A., Jassen-Mommen, J.P.M., 1989. Phytomonitoring of pulverized fuel ash leachates by the duckweed Lemna minor. Hydrobiologia 188–189, 361–366. Knoll BASF, 1995. Pharma date sheet ibuprofen. Knoll Pharmaceuticals, Notthingham, NG2 3AA, UK. Issue/revision date 6 April 1995. Kumpel, T., Alexy, R., Kummerer, K., 2001. What do we know about antibiotics in the environment? In: Kummerer, K. (Ed.), Pharmaceuticals in the Environment. Springer-Verlag, Berlin, Heidelberg, Germany, p. 80.

Minorsky, P.V., 2002. Abscisic acid: a universal signalling factor? Plant Physiol. 128, 788–789. Mira, H., Martinez, N., Penarrubia, L., 2002. Expression of a vegetative-storage-protein gene from Arabidopsis is regulated by copper, senescence and ozone. Planta 214, 939–946. Netting, A.G., 2000. pH, abscisic acid and the integration of metabolism in plants under stressed and non-stressed conditions: cellular responses to stress and their implication for plant water relations. J. Exp. Botany 343, 147–158. Prescott, L.M., Harley, J.P., Klein, D.A., (Eds.), 1996. Microbiology, third ed., Times Mirror Higher Education Group, Inc., IA, USA. Richardson, M.L., Bowron, J.M., 1985. The fate of pharmaceuticals in the aquatic environment. J. Pharm. Pharmacol. 37, 1–12. Rippka, R., Deruelles, J., Waterbury, J., Herdman, M., Stanier, R., 1979. Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J. Gen. Microbiol. 111, 1–61. Sanyal, A.K., Roy, D., Chowdhury, B., Banerjee, A.B., 1993. Ibuprofen, a unique anti-inflammatory compound with antifungal activity against dermotophytes. Lett. Appl. Microbiol. 17, 109–111. Stan, H.J., Heberer, T., 1997. Pharmaceuticals in the aquatic environment. Anal. Mag. 25, 20–23. Stumpf, M., Ternes, T.A., Haberer, K., Seel, P., Bauman, W., 1996. Determination of pharmaceutics in sewage plants and river water. Vom Wasser 86, 291–303. Suzuki, T., Ezure, T., Yamaguchi, T., Domen, H., Ishida, M., Schmidt, W., 2000. Stimulatory effect of procaine on the growth of several microalgae and cyanobacteria. J. Pharm. Pharmacol. 52, 243–251. United States Environmental Protection Agency, 1996. Prevention, Pesticides and Toxic Substances (7101), EPA712-C-96-156, Ecological Effects Test Guidelines OPPTS 850.4400, Aquatic Plant Toxicity Test Using Lemna spp., Tiers I and II. Wang, W., 1991. Literature review on higher plants for toxicity testing. Water, Air, Soil Pollut. 59, 381–400. Zuccato, E., Calamari, D., Natangelo, M., Fanelli, R., 2000. Presence of therapeutic drugs in the environment. Lancet 355, 1789–1790. Zuccato, E., Bagnati, R., Fioretti, F., Natangelo, M., Calamari, D., Fanelli, R., 2001. Environmental loads and detection of pharmaceuticals in Italy. In: Kummerer, K. (Ed.), Pharmaceuticals in the environment. Springer-Verlag, Berlin, Heidelberg, Germany, p. 65.