Environmental Pollution 202 (2015) 177e186
Contents lists available at ScienceDirect
Environmental Pollution journal homepage: www.elsevier.com/locate/envpol
Genotoxic and immunotoxic potential effects of selected psychotropic drugs and antibiotics on blue mussel (Mytilus edulis) hemocytes delucq a, b, Marle ne Fortier a, Pauline Brousseau a, Emilie Lacaze a, *, Julie Pe le ne Budzinski b, Michel Fournier a Michel Auffret c, He a b c
INRS, Institut Armand-Frappier, 531 des Prairies Blvd., Laval, H7V 1B7 QC, Canada EPOC-LPTC, UMR 5805, Universit e Bordeaux 1, 351 Cours de la Lib eration, 33405 Talence, France ^le Brest-Iroise, 29 280 Plouzane, France LEMAR UMR CNRS 6539, Institut Universitaire Europ een de la Mer, Technopo
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 January 2015 Received in revised form 11 March 2015 Accepted 13 March 2015 Available online
The potential toxicity of pharmaceuticals towards aquatic invertebrates is still poorly understood and sometimes controversial. This study aims to document the in vitro genotoxicity and immunotoxicity of psychotropic drugs and antibiotics on Mytilus edulis. Mussel hemocytes were exposed to fluoxetine, paroxetine, venlafaxine, carbamazepine, sulfamethoxazole, trimethoprim and erythromycin, at concentrations ranging from mg/L to mg/L. Paroxetine at 1.5 mg/L led to DNA damage while the same concentration of venlafaxine caused immunomodulation. Fluoxetine exposure resulted in genotoxicity, immunotoxicity and cytotoxicity. In the case of antibiotics, trimethoprim was genotoxic at 200 mg/L and immunotoxic at 20 mg/L whereas erythromycin elicited same detrimental effects at higher concentrations. DNA metabolism seems to be a highly sensitive target for psychotropic drugs and antibiotics. Furthermore, these compounds affect the immune system of bivalves, with varying intensity. This attests the relevance of these endpoints to assess the toxic mode of action of pharmaceuticals in the aquatic environment. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Pharmaceuticals Invertebrate Hemolymph Comet assay Phagocytosis Intracellular ROS
1. Introduction The presence of pharmaceuticals in wastewater effluents and in surface water has been documented for more than thirty years ago (Hignite and Azarnoff, 1977). The improvement of analytical techniques in the mid 90s helped the detection of pharmaceuticals, globally raising awareness in the scientific community (FattaKassinos et al., 2011). Nowadays, the worldwide presence of pharmaceuticals including antibiotics, hormonal and anti-hormonal drugs, anti-inflammatory agents, anticonvulsants, b-blockers and antidepressant in the aquatic environment has been attested (Kümmerer, 2009). Due to persistence of pharmaceuticals after wastewater treatments, these substances are continuously introduced in surface waters (Fent et al., 2006). In this respect, pharmaceutical substances have been found in effluents from wastewater treatment plants at low mg/L concentrations, in surface
* Corresponding author. E-mail addresses: [email protected] (E. Lacaze), julie.pedelucq@iaf. delucq), [email protected] (H. Budzinski), michel. inrs.ca (J. Pe [email protected] (M. Fournier). http://dx.doi.org/10.1016/j.envpol.2015.03.025 0269-7491/© 2015 Elsevier Ltd. All rights reserved.
freshwater, groundwater and marine environment in the high ng/L vier et al., 2013; Heberer, 2002; and even in drinking water (De Santos et al., 2010; Togola and Budzinski, 2008). Furthermore, several drugs have already been dosed in wild organisms including native freshwater mussels and caged marine mussels (Bringolf et al., 2010; McEneff et al., 2014). Increasing concern towards reduced effect of human activities in aquatic ecosystems requests advanced knowledge on possible adverse effects of these compounds, including invertebrates. Pharmaceuticals are biologically active substances exerting a specific mode of action in humans and animals. The development of drugs is based on extensive toxicological studies in order to design substances interacting with specific pathways in target organisms with a limited toxicity. However, it has been demonstrated that pharmaceutical compounds can interact with counterparts targets that could be present in invertebrates, and potentially affect comparable pathways (Christen et al., 2010). These highly active molecules can also be potentially toxic for invertebrates due to differences in physiology and pharmacodynamics (Besse and Garric, 2008). Mammals receptors may be indeed conserved in invertebrates but not always their specific function, leading to
178
E. Lacaze et al. / Environmental Pollution 202 (2015) 177e186
unexpected chronic effects in these organisms (Crane et al., 2006). The effects of pharmaceuticals exposure on freshwater and marine bivalves ranged from the molecular level, such as disturbed enzyme functions, DNA damage, and modulated gene expression (Contardo-Jara et al., 2011; Franzellitti et al., 2013; Gonzalez-Rey and Bebianno, 2013; Parolini et al., 2013; Quinn et al., 2011), to cellular level with increased oxidative stress, cytotoxicity and immunotoxicity (Aguirre-Martínez et al., 2013; Binelli et al., 2009; et al., 2006a; Gust et al., 2012; MartinCanesi et al., 2007; Gagne Diaz et al., 2009; Matozzo et al., 2012; Munari et al., 2014; Parolini et al., 2011; Tsiaka et al., 2013) to the fitness of organisms with embryotoxicity (Fabbri et al., 2014; Hazelton et al., 2013; Di Poi et al., 2014), reprotoxicity (Bringolf et al., 2010; Cortez et al., 2012; Fong and Ford, 2014; Hazelton et al., 2014; Lazzara et al., et al., 2010). While the 2012) and decrease of feeding rate (Sole effect characterization has already been addressed as one of the big question of pharmaceuticals in the environment (Boxall et al., 2012), their effects on invertebrates remained sometimes controversial. It thus appears justified to identify additional effects and modes of action that cannot be predicted from the approach of mammal extrapolation for the main therapeutic pharmaceutical classes. Short-term in vitro experiments are recommended as the first step to understand the mechanisms involved in molecular and cellular responses towards contaminants such as pharmaceuticals that exert plural and complex modes of action. In this context, we attempt to identify the genotoxic and immunotoxic effects of several pharmaceuticals from various therapeutic classes (three antidepressants: paroxetine, fluoxetine and venlafaxine and one anticonvulsant: carbamazepine, and three antibiotics: erythromycin, trimethoprim and sulfamethoxazole) on Mytilus edulis hemocytes in vitro. Fluoxetine and Venlafaxine are the most widely used antidepressants worldwide (Metcalfe et al., 2010). Paroxetine is a highly studied molecule, while it is probably less persistent than fluoxetine and venlafaxine in the aquatic environment due to a significant biodegradation of the drug entering a wastewater treatment plant (Cunningham et al., 2004). Carbamazepine has also been selected since it has been proposed as chemical tracers for urban pollution (Clara et al., 2005; Ternes et al., 2001). Based on previous prioritization lists, the antibiotics sulfamethoxazole, erythromycin and trimethoprim were chosen due to their high toxicity to microorganisms, and because they are among the most frequently detected group of potentially toxic pharmaceuticals (Besse and Garric, 2008). The blue mussel M. edulis represents a good model organism to address effects of pharmaceuticals. As an important seafood product for human consumption, its physiological system and hormone-regulation are relatively well understood. As filter feeders it has been proved, in laboratory and in field-deployed organisms, that mussels can bioaccumulate pharmaceuticals (Benotti and Brownawell, 2007; McEneff et al., 2014; Thomas and Hilton, 2004). Hemocytes are in charge of the transport of nutrients, shell repair, host defence and immunity (Cheng, 1981). Immunity in bivalves is a nonspecific inflammatory response mediated by humoral components circulating in the hemolymph and hemocytes that are capable of encapsulation, phagocytosis, intracellular production of reactive oxygen metabolites and releasing degradative enzymes (Pipe and Coles, 1995; Pipe et al., 1997). Therefore, the endpoints chosen in this study were hemocytes viability, phagocytosis efficiency, ROS production (produced after phagocytosis of foreign particles or micro-organisms) and DNA damage. Immunotoxic and genotoxic endpoints have been chosen to bring knowledge on the mode of action of pharmaceuticals, and because immunotoxic and genotoxic effects can be translated into traditional ecologically relevant endpoints, as they can lead to implications at higher biological levels in invertebrates (Galloway and
Depledge, 2001; Lacaze et al., 2011; Lewis and Galloway, 2009). 2. Materials and methods 2.1. Chemicals and drug solutions Psychotropic drugs paroxetine (CAS 61869-08-7), fluoxetine (CAS 54910-89-3), venlafaxine (CAS 93413-69-5) and carbamazepine (CAS 298-46-4) and three antibiotics trimethoprim (CAS 73870-5), erythromycin (CAS 114-07-8), sulfamethoxazole (CAS 72346-6) were purchased from SigmaeAldrich Chemical Co., ON, Canada, as well as all others chemicals, unless otherwise stated. All drugs were dissolved in dimethyl sulfoxide DMSO. Several in vitro assays were performed on hemolymph with an increasing dose of DMSO ranging from 0.001% to 5% to evaluate the eventual impact of this vehicle on hemocytes, as recommended by Hutchinson et al., 2006. The concentration of 0.5% did not cause any effects on cells and was finally chosen (data not shown). Solvent concentration never exceeds 0.5% in the final culture medium and solvent controls at 0.5% DMSO were used in all exposures. Concentrations of pharmaceuticals ranged from mg/L to mg/L. The very highest concentration tested potentially resulted in cytotoxicity, according to solubility limits. The Table 1 summarized concentrations of each drug measured in municipal influents, effluents and surface water worldwide. For each drug, a stock solution was prepared in DMSO and kept at 20 C. 2.2. Test organism Mussels were collected from Pleasance Bay, on a high seas farm, in the Madeleine Islands, St Lawrence estuary (47 290 N, 61870 W). These islands are expected to be exempt of anthropogenic contamination. Mussels were shipped to the laboratory and immediately transferred to 20 L tanks filled with artificial marine water (Instant Ocean®, Reef Crystal, Cincinnati, OH, USA, salinity 31-1 psu). Organisms were acclimated 2 weeks before experiments in oxygenated water at 15 C under a 12/12 h photoperiod, fed three times a week with commercial phytoplankton (Phytoplex®, Kent Marine, Franklin, WI, USA). Water was changed three times a week. Before the collection of hemolymph, mussels were weighed and measured to determine the condition index (CI) expressed as weight (g)/shell length (mm). Only organisms with same CI (0.3 ± 0.04, shell length ¼ 62 ± 5 mm) were selected. 2.3. Hemolymph collection, viability and exposure Hemolymph was collected individually from the posterior adductor muscle using a sterile 3 mL syringe with a 23G needle. Hemocyte counts and mortality rate were determined by flow cytometry (PCA Guava Cytometer, Guava Technologies®, Hayward, CA, USA) using the ViaCount® solution kit. This assay is based on differential permeability of two DNA-binding dyes. Briefly, an aliquot of 50 mL of hemolymph was mixed with 200 mL of Viacount and at least 5000 events were recorded. Viability rate was calculated as follow: viability ¼ 100% mortality rate. Six mussels were selected per drug, whose cellularity of hemolymph was sufficient to test the entire concentration range of the drug as well as a negative control and a carrier control. Hemolymph was diluted to 200 000 hemocytes per mL with sterile artificial marine water. For the exposure, stock solutions of drugs were diluted in artificial marine water at 10, 50, 5000 and 50 000 times for having the 5$X, X, X$102 and X$104 concentrations, respectively (Table 2). 25 mL of each diluted pharmaceutical solution was added to the 475 mL of cells to obtain the final concentration. Cells were exposed for 21 h at 15 C in the dark under gentle agitation. For phagocytosis, after
E. Lacaze et al. / Environmental Pollution 202 (2015) 177e186
179
Table 1 Review of the concentrations of the investigated psychotropic drugs and antibiotics measured in the municipal WWTP influent, effluent, and in the aquatic environment. Minimal concentrationemaximal concentration in ng/L (bold: maximal concentration reported in literature). Concentration reported (ng/L)
Erythromycin
Trimethoprim
Sulfamethoxazole
Fluoxetine
Paroxetine
Carbamazepine
Venlafaxine
WWTP influent
WWTP effluent
Aquatic environment
226e1537 (Lin and Tsai, 2009) 1800 (Verlicchi et al., 2012)
50 (Zuccato et al., 2005) 30e2000 (Verlicchi et al., 2012) 20e420 (Terzi c et al., 2008) 8.9e294 (Kim et al., 2007) 203e415 (Lin and Tsai, 2009) 20e8000 (Verlicchi et al., 2012)
Contents lists available at ScienceDirect
Environmental Pollution journal homepage: www.elsevier.com/locate/envpol
Genotoxic and immunotoxic potential effects of selected psychotropic drugs and antibiotics on blue mussel (Mytilus edulis) hemocytes delucq a, b, Marle ne Fortier a, Pauline Brousseau a, Emilie Lacaze a, *, Julie Pe le ne Budzinski b, Michel Fournier a Michel Auffret c, He a b c
INRS, Institut Armand-Frappier, 531 des Prairies Blvd., Laval, H7V 1B7 QC, Canada EPOC-LPTC, UMR 5805, Universit e Bordeaux 1, 351 Cours de la Lib eration, 33405 Talence, France ^le Brest-Iroise, 29 280 Plouzane, France LEMAR UMR CNRS 6539, Institut Universitaire Europ een de la Mer, Technopo
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 January 2015 Received in revised form 11 March 2015 Accepted 13 March 2015 Available online
The potential toxicity of pharmaceuticals towards aquatic invertebrates is still poorly understood and sometimes controversial. This study aims to document the in vitro genotoxicity and immunotoxicity of psychotropic drugs and antibiotics on Mytilus edulis. Mussel hemocytes were exposed to fluoxetine, paroxetine, venlafaxine, carbamazepine, sulfamethoxazole, trimethoprim and erythromycin, at concentrations ranging from mg/L to mg/L. Paroxetine at 1.5 mg/L led to DNA damage while the same concentration of venlafaxine caused immunomodulation. Fluoxetine exposure resulted in genotoxicity, immunotoxicity and cytotoxicity. In the case of antibiotics, trimethoprim was genotoxic at 200 mg/L and immunotoxic at 20 mg/L whereas erythromycin elicited same detrimental effects at higher concentrations. DNA metabolism seems to be a highly sensitive target for psychotropic drugs and antibiotics. Furthermore, these compounds affect the immune system of bivalves, with varying intensity. This attests the relevance of these endpoints to assess the toxic mode of action of pharmaceuticals in the aquatic environment. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Pharmaceuticals Invertebrate Hemolymph Comet assay Phagocytosis Intracellular ROS
1. Introduction The presence of pharmaceuticals in wastewater effluents and in surface water has been documented for more than thirty years ago (Hignite and Azarnoff, 1977). The improvement of analytical techniques in the mid 90s helped the detection of pharmaceuticals, globally raising awareness in the scientific community (FattaKassinos et al., 2011). Nowadays, the worldwide presence of pharmaceuticals including antibiotics, hormonal and anti-hormonal drugs, anti-inflammatory agents, anticonvulsants, b-blockers and antidepressant in the aquatic environment has been attested (Kümmerer, 2009). Due to persistence of pharmaceuticals after wastewater treatments, these substances are continuously introduced in surface waters (Fent et al., 2006). In this respect, pharmaceutical substances have been found in effluents from wastewater treatment plants at low mg/L concentrations, in surface
* Corresponding author. E-mail addresses: [email protected] (E. Lacaze), julie.pedelucq@iaf. delucq), [email protected] (H. Budzinski), michel. inrs.ca (J. Pe [email protected] (M. Fournier). http://dx.doi.org/10.1016/j.envpol.2015.03.025 0269-7491/© 2015 Elsevier Ltd. All rights reserved.
freshwater, groundwater and marine environment in the high ng/L vier et al., 2013; Heberer, 2002; and even in drinking water (De Santos et al., 2010; Togola and Budzinski, 2008). Furthermore, several drugs have already been dosed in wild organisms including native freshwater mussels and caged marine mussels (Bringolf et al., 2010; McEneff et al., 2014). Increasing concern towards reduced effect of human activities in aquatic ecosystems requests advanced knowledge on possible adverse effects of these compounds, including invertebrates. Pharmaceuticals are biologically active substances exerting a specific mode of action in humans and animals. The development of drugs is based on extensive toxicological studies in order to design substances interacting with specific pathways in target organisms with a limited toxicity. However, it has been demonstrated that pharmaceutical compounds can interact with counterparts targets that could be present in invertebrates, and potentially affect comparable pathways (Christen et al., 2010). These highly active molecules can also be potentially toxic for invertebrates due to differences in physiology and pharmacodynamics (Besse and Garric, 2008). Mammals receptors may be indeed conserved in invertebrates but not always their specific function, leading to
178
E. Lacaze et al. / Environmental Pollution 202 (2015) 177e186
unexpected chronic effects in these organisms (Crane et al., 2006). The effects of pharmaceuticals exposure on freshwater and marine bivalves ranged from the molecular level, such as disturbed enzyme functions, DNA damage, and modulated gene expression (Contardo-Jara et al., 2011; Franzellitti et al., 2013; Gonzalez-Rey and Bebianno, 2013; Parolini et al., 2013; Quinn et al., 2011), to cellular level with increased oxidative stress, cytotoxicity and immunotoxicity (Aguirre-Martínez et al., 2013; Binelli et al., 2009; et al., 2006a; Gust et al., 2012; MartinCanesi et al., 2007; Gagne Diaz et al., 2009; Matozzo et al., 2012; Munari et al., 2014; Parolini et al., 2011; Tsiaka et al., 2013) to the fitness of organisms with embryotoxicity (Fabbri et al., 2014; Hazelton et al., 2013; Di Poi et al., 2014), reprotoxicity (Bringolf et al., 2010; Cortez et al., 2012; Fong and Ford, 2014; Hazelton et al., 2014; Lazzara et al., et al., 2010). While the 2012) and decrease of feeding rate (Sole effect characterization has already been addressed as one of the big question of pharmaceuticals in the environment (Boxall et al., 2012), their effects on invertebrates remained sometimes controversial. It thus appears justified to identify additional effects and modes of action that cannot be predicted from the approach of mammal extrapolation for the main therapeutic pharmaceutical classes. Short-term in vitro experiments are recommended as the first step to understand the mechanisms involved in molecular and cellular responses towards contaminants such as pharmaceuticals that exert plural and complex modes of action. In this context, we attempt to identify the genotoxic and immunotoxic effects of several pharmaceuticals from various therapeutic classes (three antidepressants: paroxetine, fluoxetine and venlafaxine and one anticonvulsant: carbamazepine, and three antibiotics: erythromycin, trimethoprim and sulfamethoxazole) on Mytilus edulis hemocytes in vitro. Fluoxetine and Venlafaxine are the most widely used antidepressants worldwide (Metcalfe et al., 2010). Paroxetine is a highly studied molecule, while it is probably less persistent than fluoxetine and venlafaxine in the aquatic environment due to a significant biodegradation of the drug entering a wastewater treatment plant (Cunningham et al., 2004). Carbamazepine has also been selected since it has been proposed as chemical tracers for urban pollution (Clara et al., 2005; Ternes et al., 2001). Based on previous prioritization lists, the antibiotics sulfamethoxazole, erythromycin and trimethoprim were chosen due to their high toxicity to microorganisms, and because they are among the most frequently detected group of potentially toxic pharmaceuticals (Besse and Garric, 2008). The blue mussel M. edulis represents a good model organism to address effects of pharmaceuticals. As an important seafood product for human consumption, its physiological system and hormone-regulation are relatively well understood. As filter feeders it has been proved, in laboratory and in field-deployed organisms, that mussels can bioaccumulate pharmaceuticals (Benotti and Brownawell, 2007; McEneff et al., 2014; Thomas and Hilton, 2004). Hemocytes are in charge of the transport of nutrients, shell repair, host defence and immunity (Cheng, 1981). Immunity in bivalves is a nonspecific inflammatory response mediated by humoral components circulating in the hemolymph and hemocytes that are capable of encapsulation, phagocytosis, intracellular production of reactive oxygen metabolites and releasing degradative enzymes (Pipe and Coles, 1995; Pipe et al., 1997). Therefore, the endpoints chosen in this study were hemocytes viability, phagocytosis efficiency, ROS production (produced after phagocytosis of foreign particles or micro-organisms) and DNA damage. Immunotoxic and genotoxic endpoints have been chosen to bring knowledge on the mode of action of pharmaceuticals, and because immunotoxic and genotoxic effects can be translated into traditional ecologically relevant endpoints, as they can lead to implications at higher biological levels in invertebrates (Galloway and
Depledge, 2001; Lacaze et al., 2011; Lewis and Galloway, 2009). 2. Materials and methods 2.1. Chemicals and drug solutions Psychotropic drugs paroxetine (CAS 61869-08-7), fluoxetine (CAS 54910-89-3), venlafaxine (CAS 93413-69-5) and carbamazepine (CAS 298-46-4) and three antibiotics trimethoprim (CAS 73870-5), erythromycin (CAS 114-07-8), sulfamethoxazole (CAS 72346-6) were purchased from SigmaeAldrich Chemical Co., ON, Canada, as well as all others chemicals, unless otherwise stated. All drugs were dissolved in dimethyl sulfoxide DMSO. Several in vitro assays were performed on hemolymph with an increasing dose of DMSO ranging from 0.001% to 5% to evaluate the eventual impact of this vehicle on hemocytes, as recommended by Hutchinson et al., 2006. The concentration of 0.5% did not cause any effects on cells and was finally chosen (data not shown). Solvent concentration never exceeds 0.5% in the final culture medium and solvent controls at 0.5% DMSO were used in all exposures. Concentrations of pharmaceuticals ranged from mg/L to mg/L. The very highest concentration tested potentially resulted in cytotoxicity, according to solubility limits. The Table 1 summarized concentrations of each drug measured in municipal influents, effluents and surface water worldwide. For each drug, a stock solution was prepared in DMSO and kept at 20 C. 2.2. Test organism Mussels were collected from Pleasance Bay, on a high seas farm, in the Madeleine Islands, St Lawrence estuary (47 290 N, 61870 W). These islands are expected to be exempt of anthropogenic contamination. Mussels were shipped to the laboratory and immediately transferred to 20 L tanks filled with artificial marine water (Instant Ocean®, Reef Crystal, Cincinnati, OH, USA, salinity 31-1 psu). Organisms were acclimated 2 weeks before experiments in oxygenated water at 15 C under a 12/12 h photoperiod, fed three times a week with commercial phytoplankton (Phytoplex®, Kent Marine, Franklin, WI, USA). Water was changed three times a week. Before the collection of hemolymph, mussels were weighed and measured to determine the condition index (CI) expressed as weight (g)/shell length (mm). Only organisms with same CI (0.3 ± 0.04, shell length ¼ 62 ± 5 mm) were selected. 2.3. Hemolymph collection, viability and exposure Hemolymph was collected individually from the posterior adductor muscle using a sterile 3 mL syringe with a 23G needle. Hemocyte counts and mortality rate were determined by flow cytometry (PCA Guava Cytometer, Guava Technologies®, Hayward, CA, USA) using the ViaCount® solution kit. This assay is based on differential permeability of two DNA-binding dyes. Briefly, an aliquot of 50 mL of hemolymph was mixed with 200 mL of Viacount and at least 5000 events were recorded. Viability rate was calculated as follow: viability ¼ 100% mortality rate. Six mussels were selected per drug, whose cellularity of hemolymph was sufficient to test the entire concentration range of the drug as well as a negative control and a carrier control. Hemolymph was diluted to 200 000 hemocytes per mL with sterile artificial marine water. For the exposure, stock solutions of drugs were diluted in artificial marine water at 10, 50, 5000 and 50 000 times for having the 5$X, X, X$102 and X$104 concentrations, respectively (Table 2). 25 mL of each diluted pharmaceutical solution was added to the 475 mL of cells to obtain the final concentration. Cells were exposed for 21 h at 15 C in the dark under gentle agitation. For phagocytosis, after
E. Lacaze et al. / Environmental Pollution 202 (2015) 177e186
179
Table 1 Review of the concentrations of the investigated psychotropic drugs and antibiotics measured in the municipal WWTP influent, effluent, and in the aquatic environment. Minimal concentrationemaximal concentration in ng/L (bold: maximal concentration reported in literature). Concentration reported (ng/L)
Erythromycin
Trimethoprim
Sulfamethoxazole
Fluoxetine
Paroxetine
Carbamazepine
Venlafaxine
WWTP influent
WWTP effluent
Aquatic environment
226e1537 (Lin and Tsai, 2009) 1800 (Verlicchi et al., 2012)
50 (Zuccato et al., 2005) 30e2000 (Verlicchi et al., 2012) 20e420 (Terzi c et al., 2008) 8.9e294 (Kim et al., 2007) 203e415 (Lin and Tsai, 2009) 20e8000 (Verlicchi et al., 2012)
