Effects of Dietary Cyanobacteria of Two Different Sources on Growth and Recovery of Hybrid Tilapia (Oreochromis Niloticus � O

Toxicon 54 (2009) 208–216

Contents lists available at ScienceDirect

Toxicon

journal homepage: www.elsevier.com/locate/toxicon

Effects of dietary cyanobacteria of two different sources on growth and recovery of hybrid tilapia (Oreochromis niloticus O. aureus)

Guifang Dong a,b, Xiaoming Zhu a,*, Dong Han a, Yunxia Yang a, Lirong Song a, Shouqi Xie a,c a State Key Laboratory of Freshwater Ecology and Biotechnology, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan, Hubei 430072, China b Graduate University of Chinese Academy of Sciences, Beijing 100039, China c Aquaculture Divisions, E-Institute of Shanghai Universities, Shanghai, China article info abstract

Article history: A 115 days feeding trial was conducted to evaluate the effect of dietary cyanobacteria on Received 22 December 2008 growth, microcystins (MCs) accumulation in hybrid tilapia (Oreochromis niloticus O. Received in revised form 28 March 2009 aureus) and the recovery when the fish were free of cyanobacteria. Three experimental Accepted 30 March 2009 diets were formulated: the control (cyanobacteria free diet); one test diet with cyano- Available online 15 April 2009 bacteria from Lake Taihu (AMt, 80.0 mg MCs g1 diet) and one with cyanobacteria from Lake Dianchi (AMd, 410.0 mg MCs g1 diet). Each diet was fed to fish for 60 days and then Keywords: all fish were free of cyanobacteria for another 55 days. Cyanobacteria Growth A significant increase in feeding rate (FR) was observed in fish fed AMd diet after a first st Recovery 30-day exposure (1 EP), and in fish fed both AMt diet and AMd diet after a second 30-day nd Hybrid tilapia exposure (2 EP). Specific growth rates (SGR) of fish fed AMt diet and AMd diet were both obviously affected after the first 30-day exposure, but SGR was only significantly affected in fish fed AMt diet after the second 30-day exposure. After a 55-day recovery, there were no significant differences among diets in the indices mentioned above. Much higher concentrations of MCs were accumulated in tissues of all fish exposed to cyanobacteria. After the 55-day recovery, MC concentrations in fish tissues were significantly lower than those on day 60. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction serious cyanobacterial blooms have occurred frequently in many freshwater lakes (e.g. Lake Dianchi, Lake Taihu, Lake Eutrophication in freshwater has induced the occurrence Chaohu, etc.) and most blooms produced MCs at high of dense cyanobacterial blooms (Carmichael, 1994). Some concentrations (Li et al., 2001; Song et al., 1998). Previous cyanobacteria including Anabaena, Aphanocapsa, Hapalosi- studies indicate that MCs accumulated in tissues of aquatic phon, Nostoc, Pseudanabaena, Planktothrix and Microcystis animals such as snails (Bellamya aeruginosa), shrimps could produce cyclic peptide toxins – microcystins (MCs) (de (Palaemon modestus and Macrobrachium nipponensis), and Figueiredo et al., 2004; Izaguirre et al., 2007) – which can fishes in those freshwater lakes (Chen and Xie, 2005; Chen cause liver failure in animals, livestock and aquatic life et al., 2005; Xie et al., 2005; Chen et al., 2007). (Carmichael,1994; Sivonen and Jones,1999), and even cause The acute effects of cyanobacteria or MCs on fish species human diseases or death (Azevedo et al., 2002). have been well investigated (Bury et al., 1995; Li et al., Fish deaths have been reported to be related to severe 2005). Generally, aquatic animals might be exposed to MCs cyanobacterial blooms (Rodger et al., 1994). In China, via the consumption of toxic cyanobacteria (Soares et al., 2004) or through gills (Zimba et al., 2001). Cyanobacteria could also be an important dietary component for many * Corresponding author. Tel.: þ86 27 68780060; fax: þ86 27 68780667. fish species including tilapia (Zurawell et al., 2005). E-mail address: [email protected] (X. Zhu). However, there is still a lack of studies on the chronic toxic

0041-0101/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.toxicon.2009.03.031 G. Dong et al. / Toxicon 54 (2009) 208–216 209 effects of dietary cyanobacteria on fishes (Zhao et al., 2005, Table 1 1 2006a). It is also unclear if or how much the fish could Formulation and chemical composition of experimental diets (g 100 g in dry matter). recover when they are free of cyanobacteria. MCs not only affect the growth of fish, but also accu- Control AMt AMd mulate in their tissues (Magalha˜es et al., 2003; Mohamed Ingredients et al., 2003; Soares et al., 2004; Deblois et al., 2008). MCs Algae meal 0.0 43.6 50.8 contaminated fish may be a risk to human health from the White fishmeal (USA) 20.0 20.0 20.0 Soybean meal (oil-extracted) 49.9 0.0 0.0 food chain. To reduce potential risks, a tolerable daily Corn starch 10.2 10.2 10.2 1 intake (TDI) of 0.04 mgkg total MCs per kilogram body Fish oil 5.3 5.8 6.0 weight per day as fish meat was used as a provisional a-Starch 6.0 6.0 6.0 a guideline (Chorus and Bartram, 1999). Mineral premix 0.4 0.4 0.4 Vitamin premixb 5.0 5.0 5.0 Previous studies in our laboratory demonstrated that Choline chloride 0.1 0.1 0.1 chronic exposure to lower levels of dietary cyanobacteria Cr2O3 0.5 0.5 0.5 increased the growth rate of Nile tilapia without impact on Cellulose 2.7 8.4 1.0 food conversion efficiency but high toxins were found to Chemical composition (g 100 g1 in dry matter) accumulate in fish muscle and liver (Zhao et al., 2006). The Crude protein (%) 41.4 43.1 44.0 toxin level reported by Zhao et al. (2006) was even higher Crude lipid (%) 7.8 7.9 7.9 than the tolerable daily intake. If these contaminated fish Crude ash (%) 10.7 10.2 11.0 Gross energy (kJ g1) 19.7 20.8 20.6 could recover when they were free of toxins is very Microcystins (mgg1) 0.0 80.0 410.0 important for food safety. a 1 $ The purpose of the present study was to investigate the Mineral premix (mg kg diet, H440): NaCl, 500; MgSO4 7H2O, 7500; NaH PO $2H O, 12,500; KH PO , 16,000; Ca(H PO ) $2H O,10,000; FeSO , chronic effects of dietary cyanobacteria on hybrid tilapia 2 4 2 2 4 2 4 2 4 1250; C6H10CaO6$5H2O, 1750; ZnSO4$7H2O,176.5; MnSO4.$4H2O, 81; and if or how much the fish could recover when they were CuSO4.$5H2O, 15.5; CoSO4$6H2O, 0.5; KI, 1.5; starch, 225. free of cyanobacteria in their diet. b Vitamin premix (mg kg1 diet, NRC, 1993): Thiamin, 20; riboflavin, 20; pyridoxine, 20; cyanocobalamine, 2; folic acid, 5; calcium patotheniate, 50; inositol, 100; niacin, 100; biotin, 5; starch, 3226; vitamin A (ROVIMIX 2. Materials and methods A-1000), 110; vitamin D3, 20; vitamin E, 100; vitamin K3, 10.

2.1. Fish, cyanobacteria and experimental diets transferred from the rearing tank to the flow-through Hybrid tilapia (Oreochromis niloticus O. aureus)were system for acclimation. At the beginning of the trial, fish obtained from the tilapia hatchery farm in Puqi, Hubei, were fasted for 1 day to empty the gut. Fifty fish of similar China and acclimated in two cylindrical fiberglass tanks size (initial body weight 2.2 g) were randomly selected, (diameter 150 cm height 120 cm, water volume 1500 L) weighed and stocked in each tank. Six tanks were randomly for 20 days prior to the experiment. During the acclimation assigned to each diet. During the experiment, aeration was period, fish were fed to satiation twice daily (09:00 and provided to each tank to maintain dissolved oxygen level 15:00) with the control diet. above 7 mg L1; the photoperiod was 12 h light:12 h dark Two batches of fresh cyanobacteria were collected from with the light period from 08:00 to 20:00. Light intensity at Lake Taihu in Jiangsu and Lake Dianchi in Yunnan, China. the water surface was around 200 lx. Water temperature The cyanobacteria from Lake Taihu were composed of 50% was recorded daily and was maintained at 26 2 C, pH Microcystis aeruginosa and 50% Microcystis wesenbegii was about 7.0. Ammonia-N was monitored once a week and whereas the cyanobacteria from Lake Dianchi mainly con- was less than 0.5 mg L1 and residual chloride was less sisted of Microcystis aeruginosa (95%). The fresh cyanobac- than 0.01 mg L1. Fish were hand-fed to apparent satiation teria were air-dried before use. twice daily (09:00 and 15:00). The daily food supplied was Three experimental diets were formulated to be recorded and uneaten diets were siphoned 1 h after approximately isonitrogenous (crude protein: 42%) and feeding, dried to constant weight at 70 C and reweighed. isocaloric (gross energy: 20 kJ g1)(Table 1). In the control Leaching rates (potential loss of uneaten diet) of uneaten diet, 20% fishmeal and 49.9% soybean meal were used as the diets were estimated by placing weighed food in tanks protein source while 43.6% of cyanobacteria from Lake without fish for 1 h and then collecting, drying and Taihu were used to replace soybean meal to formulate AMt reweighing. Leaching rate was used to calibrate the and 50.8% of cyanobacteria from Lake Dianchi were used to uneaten diets. The accurate total food intake was calculated replace soybean meal to formulate AMd. The MCs concen- from the difference between food intake and the calibrated trations in the AMt and AMd diets were 80.0 mg MCs g1 diet uneaten diet. Fresh and intact faeces were siphoned 1 h and 410.0 mg MCs g1 diet, respectively. All diets were made after the collection of uneaten diets from the 11th day of into pellets (1 mm in diameter) with a laboratory presser, the experiment throughout the experimental period. The oven-dried at 60 C and stored at 4 C before use. faeces samples were divided into two samples and used for digestibility or MCs determination. The faecal samples used 2.2. Experimental procedure for digestibility were dried at 60 C and kept at 4 C until chemical analysis. The faecal samples used for MCs analysis The trial was conducted in a flow-through system con- were stored at 20 C until freeze-dried for MCs determi- taining 18 conical fiberglass tanks (diameter: 80 cm, water nation. The trial was composed of four periods (first volume: 300 L). Two weeks before the trial, fish were exposure period: 1st EP, first 30 days; second exposure 210 G. Dong et al. / Toxicon 54 (2009) 208–216 period, 2nd EP, second 30 days; first recovery period: 1st RP, All samples were extracted first with 5% acetic acid third 30 days; second recovery period: 2nd RP, last 25 days). (10 mL) for 20 min, then with 80% methanol aqueous Fish were fed with three experimental diets during the solution (20 mL) for 1 h with sufficient mixing by exposure period and then all were fed with the control diet a magnetic stirrer, and finally centrifugation (15 min, during the recovery period. All fish in each tank were 8000 r min1) with residue re-extracted for 45 min. The batch-weighed, and sampled after 1 day of food depriva- three extracts were combined, diluted, and passed through tion at the end of each period. Three fish from each tank a pre-conditioned Sep-Pak C18 cartridge (Waters, Milford, were randomly sampled at each sampling. Sampled fish MA). The cartridge was first rinsed with 100% methanol were dissected immediately and liver, kidney, whole solution and then ultrapure water. MCs were eluted with intestine and dorsal white muscle (from posterior edge of 10 mL of 100% methanol. After evaporation at 35 Cto operculum to end of dorsal-fin base above the lateral line) dryness in a rotary evaporator, the residue containing MCs were dissected and stored at 20 C until freeze-dried for was dissolved in 1 mL of ultrapure water. The samples MCs determination. At the end of exposure period (EP) and obtained were stored at 20 C for MCs analysis. recovery period (RP), three fish from each tank were The ELISA kit and technology was kindly provided by randomly sampled for analysis of whole body chemical Prof. Song Lirong (Institute of Hydrobiology, Chinese composition. Those sampled fish in each tank were auto- Academy of Sciences, Wuhan, China). The ELISA method claved at 120 C, dried to constant weight at 70 C and was used to determine MCs according to the previous homogenized before whole body composition analysis method by Hu et al. (2008). In brief, a microtiter plate (including dry matter, protein, lipid, ash and energy). (Nunc, Roskilde, Denmark) was coated with 100 mL of (MC- LR)-BSA (bovine serum albumin) at 4 C overnight, the 2.3. Chemical analysis unbound antigen was washed away with 0.01 M PBST (phosphate buffered saline containing 0.05% Tween-20, pH Dry matter contents , crude protein, crude lipid, crude 7.4) with a model 1575 immunowash apparatus (BioRad, ash and gross energy were determined for diets and whole USA), and the well was washed again with 0.01 M PBST fish bodies. Crude protein and gross energy were deter- followed by blocking with 0.01 M PBS (pH 7.4) containing mined for faecal samples. All these parameters were 0.5% gelatin at 4 C overnight. Fifty milliliters of standard or determined according to the conventional method (AOAC, sample were pipetted into the coated wells, and then 50 mL 1984): dry matter was determined by drying to constant of monoclonal antibody (mAb) was added into each well. weight in an oven at 105 1 C. Dry matter content was After incubation at 37 C in the model 237 microplate calculated according to the equation: dry matter incubator (BioRad, USA) for 90 min followed by washing (%) ¼ 100% 100 [initial sample weight (g) final three times with PBST, 100 mL of goat anti-mouse IgG sample weight after drying to constant weight (g)]/initial (1:3000 dilution with PBS containing 0.5% gelatin) was sample weight (g). Nitrogen was determined by the Kjel- added into each well and incubated at 37 C for 30 min dahl method, and protein content was calculated from the followed by washing five times with PBST. Then, 100 mLof nitrogen content multiplied by 6.25. Crude lipid was 3,30,5,50-tetramethylbenzidine (TMB) substrate solution determined by ether extraction using a Soxtec system was added, the plate was incubated at 37 C for 10 min and (Soxtec System HT6, Tecator, Hoganas, Sweden). Crude lipid the reaction was stopped by addition of 50 mLof1MH2SO4. content was calculated according to the equation: crude The optical density (450 nm) of the ELISA assay was lipid (%) ¼ 100 extracted oil weight in initial sample (g)/ measured using a microplate reader (model 550 microplate initial sample weight (g). Crude ash content was deter- reader, BioRad Laboratories, Hercules, CA). mined by incineration at 550 C in a muffle furnace. Crude ash content was calculated according to the equation: crude ash (%) ¼ 100 [initial sample weight (g) final 2.5. Calculations and statistical analysis sample weight after incinerated (g)]/initial sample weight (g). Gross energy content was measured by combustion in Feeding rate (FR, % body weight/day) ¼ 100 total food a microbomb calorimeter (Phillipson microbomb calorim- intake (dry matter, g)/days/[(initial body weight (wet eter, Gentry Instruments Inc., Aiken, SC, USA). A compar- weight, g) þ final body weight (wet weight, g))/2]. ison of combustion values between sample and standard Specific growth rate (SGR, %/day) ¼ 100 [ln final body sample (benzoic acid) was used to calculate the gross weight (wet weight, g) ln initial body weight (wet energy. Contents of Cr2O3 in diets and faeces were deter- weight, g)]/days. mined as described by Bolin et al. (1952). At least two Food conversion efficiency (FCE, %) ¼ 100 [final body measurements were made for each sample. weight (wet weight, g) initial body weight (wet weight, g)]/food intake (dry matter, g). 2.4. Microcystins analysis in fish tissue Protein retention efficiency (PRE, %) ¼ 100 protein gain in fish/protein intake (dry matter, g). Fish liver, kidney, whole intestine and dorsal white Energy retention efficiency (ERE, %) ¼ 100 energy muscle samples were freeze-dried and all samples were gain in fish/energy intake (dry matter, kJ g1). ground into powder (diameter: 250 mm). The recovery of all Hepatosomatic index (HSI, %) ¼ 100 hepatopancreas samples ranged from 89 to 95%, coefficient of variation (CV) weight (g)/somatic weight (g). ranged from 2.33 to 11.8%. The MCs extraction procedure ADC of dry matter (ADCd,%)¼ (1 %Cr2O3 in diet/% for tissue samples was as follows: Cr2O3 in faeces) 100. G. Dong et al. / Toxicon 54 (2009) 208–216 211

ADC of protein (ADCp,%)¼ [1 % protein in faeces % Cr2O3 in diets/(% protein in diets %Cr2O3 in faeces)] 100. ADC of energy (ADCe,%)¼ [1 % energy content in faeces %Cr2O3 in diets/(% energy content in diets % Cr2O3 in faeces)] 100. The initial body weight at the beginning of each period was used as the covariance. The normality and homoge- neity of variances among data were tested and then analysis of covariance (ANCOVA) was used to adjust the fish initial body weight. All data were subjected to one-way ANOVA. If significances (P < 0.05) were identified, Dun- can’s multiple range tests were used to test differences between diets. Statistica 6.0 for windows was used for statistical analysis. Fig. 1. Effect of dietary cyanobacteria on feeding rate (FR) of hybrid tilapia. Mean values during the experimental period with different superscript are 3. Results significantly different (P < 0.05).

No mortality of fish was observed during the experiment. of the control during the exposure period (P < 0.05) while no significant difference was observed between groups (P > 0.05) (Fig. 3). 3.1. Growth and feed utilization During the exposure period, protein retention efficiency (PRE) (Fig. 4), energy retention efficiency (ERE) (Fig. 5) and Table 2 showed that there was no significant difference hepatosomatic index (HSI) (Fig. 6) were significantly lower between body weights at the start of the experiment in the fish fed the diets with cyanobacteria (P < 0.05) while (P > 0.05) and the body weight of the fish fed the diets with there was no significant difference between groups during cyanobacteria (AMt or AMd) was significantly lower than the recovery period (P > 0.05). that of the control during the whole experiment (P < 0.05). Effects of dietary cyanobacteria on growth and feed utilization are shown in Figs. 1–6. During the first exposure 3.2. Apparent digestibility coefficient period (1st EP), feeding rate (FR) of the fish fed with AMd was significantly higher (P < 0.05) (Fig. 1). During the During the exposure period, apparent digestibility of nd second exposure period (2 EP), FRs of the fish fed with the dry matter (ADCd)(Fig. 7), protein (ADCp)(Fig. 8) and diets with cyanobacteria (AMd and AMt) were significantly energy (ADCe)(Fig. 9) were significantly lower in the fish higher than that of the control and this continued until the fed the diets with cyanobacteria (P < 0.05) while there was end of the experiment including the two recovery periods no significant difference between groups with or without (P < 0.05). cyanobacteria (P > 0.05). During the recovery period, ADCp Fig. 2 shows that the specific growth rate (SGR) of the fish of the AMd was significantly higher than that of AMt fed with the diets with cyanobacteria (AMd and AMt) was (P < 0.05) (Fig. 8). significantly lower than the control during the first exposure period (P < 0.05). During the second exposure period, SGR of the fish fed with AMt was significantly lower than that of the 3.3. Microcystins accumulation and depuration in fish tissue control (P < 0.05) while SGR of the fish fed with AMd was similar to the control (P > 0.05). During the first recovery Duringtheexposureperiods,microcystins(MCs)accu- period, the fish previously exposed to cyanobacteria showed mulated in fish liver, kidney and intestines increased with significantly higher SGR than that of the control (P < 0.05) dietary MCs and were also time-dependent (P < 0.05) while there was no significant difference between groups (Fig. 10A–C). The clearance of MCs in fish liver, kidney and during the second recovery period (P > 0.05). intestines was also time-dependent while the concentrations Food conversion efficiencies (FCE) of the fish fed the of the fish previously exposed to higher MCs were still higher diets with cyanobacteria were significantly lower than that than those exposed to lower MCs (P < 0.05) (Fig. 10A–C).

Table 2 Body weight of hybrid tilapia at different sampling time points (mean SE).

Diets Body weight (g)

Initial Day 30 Day 60 Day 90 Day 115 Control 2.21 0.01 8.44 0.07a 23.76 0.41a 52.85 1.00a 82.67 2.35a AMt 2.23 0.03 4.94 0.15b 11.18 0.83b 35.31 1.81b 63.95 3.23b AMd 2.23 0.01 6.36 0.26c 16.85 0.67c 43.34 1.57c 72.34 1.36c

Mean values with different superscript letters in the same column are signiﬁcantly different (P < 0.05). 212 G. Dong et al. / Toxicon 54 (2009) 208–216

Fig. 2. Effect of dietary cyanobacteria on specific growth rate (SGR) of hybrid Fig. 4. Effect of dietary cyanobacteria on protein retention efficiency (PRE) of tilapia. Mean values during the experimental period with different super- hybrid tilapia. Mean values during the experimental period with different script are significantly different (P < 0.05). superscript are significantly different (P < 0.05).

During the exposure period, MCs accumulation in fish Growth performance in terms of weight gain (WG) or muscle was both dose- and time-dependent (P < 0.05) specific growth rate (SGR) is an important parameter for (Fig. 10D). During the first recovery period, the muscle MCs fish. Fish exposed to or feeding on toxic cyanobacterial cells in the fish previously exposed to lower MCs continued to have been reported to show reduced growth (Bury et al., increase and were even higher than in the fish previously 1995; Kamjunke et al., 2002a,b). Decreased growth rate exposed to higher MCs on day 90 (P < 0.05). The MCs in the was also observed in common carp exposed to Microcystis muscle of the fish previously exposed to higher MCs by feeding with bloom scum (Li and Chung, 2004) and in showed time-dependence (P < 0.05). planktivorous fish by naturally ingested Microcystis aeru- Faecal MCs concentration of the fish fed with higher ginosa (Rai, 2000). Also, a dose-dependent decrease in the MCs (AMd) was higher than that of those fed lower MCs growth rate of medaka fish embryos was observed at (AMt) (P < 0.05) (Fig. 11). higher concentrations of MCs (Jacquet et al., 2004). The present study found that feeding fish with cyano- 4. Discussion bacteria could result in lower body weight and growth. In contrast, our previous study showed that chronic 4.1. Effect of dietary cyanobacteria on fish growth consumption of dietary cyanobacteria containing MCs increased the growth of Nile tilapia and had no impact on Previous studies demonstrated that cyanobacterial food conversion efficiency (Zhao et al., 2006). The difference exposure through the gastrointestinal system could cause between these two studies might be due to: (1) adminis- toxic effects on many fish species such as common carp tration of different MCs doses. The dietary MCs in Zhao et al. (Fischer and Dietrich, 2000), silver carp (Xie et al., 2007), (2006) were 1.253–5.460 mgg1, which was much lower rainbow trout (Tencalla and Dietrich, 1996), tilapia than the present study; (2) the difference between species (Mohamed et al., 2003; Mohamed and Hussein, 2006). (Zhao, 2006). In the present study, hybrid tilapia exposed to cyanobacteria showed decreased HSI. It is possible that the

Fig. 3. Effect of dietary cyanobacteria on food conversion efficiency (FCE) of Fig. 5. Effect of dietary cyanobacteria on energy retention efficiency (ERE) of hybrid tilapia. Mean values during the experimental period with different hybrid tilapia. Mean values during the experimental period with different superscript are significantly different (P < 0.05). superscript are significantly different (P < 0.05). G. Dong et al. / Toxicon 54 (2009) 208–216 213

Fig. 6. Effect of dietary cyanobacteria on hepatosomatic index (HSI) of hybrid tilapia. Mean values at different sampling time points with different superscript are signiﬁcantly different (P < 0.05).

Fig. 8. Effect of dietary cyanobacteria on ADC of protein (ADCp) of hybrid fish exposed to higher levels of dietary MCs could not repair tilapia. Mean values during the experimental period with different super- the damage as fast it occurred, resulting in a decrease in script are significantly different (P < 0.05). liver weight and HSI (Casarett et al., 1996). Calabrese. (1999) found that lower doses of toxins may present study and there is still a need for extensive research induce a range of stimulatory responses and the magnitude to identify how long the fish require to recover so as to of stimulatory responses is typically 40–60% greater than catch up with normal fish. the control. Therefore, the maximum magnitude of stimulatory response in SGR of Nile tilapia in the study by Zhao et al. (2006) was 15.8% greater than the control and the 4.3. Microcystins accumulation and depuration in fish tissue magnitude was lower than in general results. Accumulation and depuration of MCs in fish and other aquatic organisms have been investigated in numerous 4.2. Compensatory growth during the recovery period studies (Xie et al., 2004; Pires et al., 2004; Pereira et al., 2004). In the present study, MCs accumulated in liver, In the present study, hybrid tilapia exposed to higher kidney, whole intestine and dorsal white muscle of hybrid MCs obtained higher body weight than those exposed to tilapia. This was consistent with previous results in many lower MCs. This could be due to the hormesis (Calabrese, fish species (Tencalla et al., 1994; Soares et al., 2004; Zhao 1999) and resulted in higher feeding rate and higher et al., 2006). Williams et al. (1997) and Smith and Haney. protein retention. Partial compensatory growth (Ali et al., (2006) demonstrated that MCs concentrations depurated 2003) was observed during the first recovery period in the in several organs (liver and muscle) in a time-dependent manner. However, in our study, MCs contents showed

Fig. 7. Effect of dietary cyanobacteria on ADC of dry matter (ADCd) of hybrid Fig. 9. Effect of dietary cyanobacteria on ADC of energy (ADCe) of hybrid tilapia. Mean values during the experimental period with different super- tilapia. Mean values during the experimental period with different superscript are signiﬁcantly different (P < 0.05). script are signiﬁcantly different (P < 0.05). 214 G. Dong et al. / Toxicon 54 (2009) 208–216

Fig. 10. Microcystin accumulation and depuration in tissues of hybrid tilapia. (A) Microcystin contents in liver at each sampling; (B) microcystin contents in kidney at each sampling; (C) microcystin contents in whole intestine at each sampling; (D) microcystin contents in muscle at each sampling. Mean values with different lowercase letters represent the significance of AMt diet at each sampling (P < 0.05); mean values with different capital letters represent the significance of AMd diet at each sampling (P < 0.05); asterisks (*) represent significance between AMt and AMd diets at each sampling (P < 0.05). a clear decrease only in liver, kidney and total intestine phenomenon was observed by Soares et al. (2004) in Tilapia during the recovery period, but not in muscle. Interestingly, rendalli. It could be a consequence of the metabolization of the highest MCs concentration in muscle was observed protein phosphatases, a turnover that would lead to the after the first 30 days of recovery in fish previously fed AMt release of these toxins, allowing their detection by the diet. A similar result was also found in yellow catfish in our ELISA method. laboratory (Dong et al., unpublished). This could be due to Mohamed and Hussein (2006) reported that tilapia can differential transfer from liver to muscle. Honkanen et al. depurate and excrete MCs into the bile and surrounding (1990) suggested that there was a strong interaction water as a way to avoid toxicity from these hepatotoxins. In between MC-LR and the catalytic subunit of PP2A. The our study, hybrid tilapia excreted high amounts of MCs ELISA method used in our study made it possible to detect through faeces during the accumulation period. When MCs bound to PP1 and PP2A. Therefore, muscle MCs compared with previous studies in common carp and gibel included free MCs plus bound MCs during the depuration carp, this suggests that hybrid tilapia are more resistant to period and the total MCs in those tissues were to some MCs than common carp and gibel carp due to the asso- extent overestimated. Vasconcelos (1995) and Amorim and ciated way of avoiding toxicity from MCs. Vasconcelos (1999) also observed higher MCs concentra- Cyanobacterial blooms have been occurring frequently tions in mussels in depuration periods. A similar in the aquatic systems of aquaculture and fisheries. Therefore, commercial fish from those environments could pose a potential threat to public health. WHO has proposed a tolerable daily intake (TDI) of 0.04 mgkg1 body weight per day for MC-LR (Chorus and Bartram, 1999). The present study showed that muscle MCs concentrations in fish fed both AMt diet and AMd diet were more than 600 ng g1 dry weight before and after the entire recovery period. There- fore, fish muscle could still not be safe for human consumption after a period of depuration. Although dietary MCs concentrations in the present study were higher than those in freshwater, detailed study should be conducted to estimate how long is required for clearance of the toxin, depending on exposure time and dose, so as to guarantee the safety of food for human consumption. Fig. 11. Faecal microcystin concentration in hybrid tilapia during the exposure period. Mean values with different superscript are significantly In conclusion, dietary cyanobacteria from Lake Taihu different (P < 0.05). and Lake Dianchi showed negative effects on growth, feed G. Dong et al. / Toxicon 54 (2009) 208–216 215 utilization and nutrient retention of hybrid tilapia during Honkanen, R.E., Zwiller, J., Moore, R.E., Daily, S.L., Khatra, B.S., Dukelow, M., the exposure period. Fish showed recovery in growth when Boynton, A.L., 1990. Characterization of microcystin-LR, a potential inhibitor of type 1 and type 2A protein phosphatases. J. Biol. Chem. they were free of dietary cyanobacteria, but the clearance of 265, 19401–19404. microcystins in fish muscles was still slow. Hu, C.L., Gan, N.Q., He, Z.K., Song, L.R., 2008. A novel chemiluminescent immunoassay for microcystin (MC) detection based on gold nano- particles label and its application to MC analysis in aquatic environ- Acknowledgements mental samples. Int. J. Environ. Anal. Chem. 88, 267–277. Izaguirre, G., Jungblut, A.D., Neilan, B.A., 2007. Benthic cyanobacteria (Oscillatoriaceae) that produce microcystin-LR, isolated from four The research was supported by the National Natural reservoirs in southern California. Water Res. 41, 492–498. Science Foundation of China (project no. 30750110022) and Jacquet, C., Thermes, V., de Luze, A., Puiseux-Dao, S., Bernard, C., Joly, J.S., the National Basic Research Program (973 no. Bourrat, F., Edery, M., 2004. Effects of MC-LR on development of medaka fish embryos (Oryzias latipes). Toxicon 43, 141–147. 2008CB418006), and partly by the National Key Technology Kamjunke, N., Schmidt, K., Pflugmacher, S., Mehner, T., 2002a. R&D Program (no. 2007BAD37B02-0404) and the innova- Consumption of cyanobacteria by roach (Rutilus rutilus): useful or tion project of the Chinese Academy of Sciences (KSCX2- harmful to the fish? Freshw. Biol. 47, 243–250. Kamjunke, N., Mendonca, R., Hardewig, I., Mehner, T., 2002b. Assimilation SW-129). The authors thank Mr Chenlin Hu and Mr of different cyanobacteria as food and the consequences for internal Guanghan Nie for their technical help. energy stores of juvenile roach. J. Fish Biol. 60, 731–738. Li, X.Y., Chung, I.K., 2004. Subchronic oral toxicity of microcystin in common carp (Cyprinus carpio L.) exposed to Microcystis under Conflict of interest laboratory conditions. Toxicon 44, 821–827. Li, X.Y., Liu, Y.D., Song, L.R., 2001. Cytological alterations in isolated The authors declare that there are no conflicts of hepatocytes from common carp (Cyprinus carpio L.) exposed to microcystin-LR. Environ. Toxicol. 16, 517–522. interest. Li, L., Xie, P., Chen, J., 2005. In vivo studies on toxin accumulation in liver and ultrastructural changes of hepatocytes of the phytoplanktivorous bighead carp i.p.injected with extracted microcystins. Toxicon 46, References 533–545. Magalha˜es, V.F., Marinho, M.M., Domingos, P., Oliveira, A.C., Costa, S.M., Ali, M., Niceieza, A., Wootton, R.J., 2003. Compensatory growth in fishes: Azevedo, L.O., Azevedo, S.M.F.O., 2003. Microcystins (cyanobacteria a response to growth depression. Fish Fish. 4, 147–190. hepatotoxins) bioaccumulation in fish and crustaceans from Sepetiba Amorim, A´ ., Vasconcelos, V., 1999. Dynamics of microcystins in the Bay (Brasil, RJ). Toxicon 42, 289–295. mussel Mytilus galloprovincialis. Toxicon 37, 1041–1052. Mohamed, Z.A., Hussein, A.A., 2006. Depuration of microcystins in tilapia AOAC (Association of Official Analytical Chemists), 1984. In: Williams, S. fish exposed to natural populations of toxic cyanobacteria: a labora- (Ed.), Official Methods of Analysis, fourteenth ed. AOAC, Inc., Arling- tory study. Ecotoxicol. Environ. Saf. 63, 424–429. ton, VA, Washington, DC, p. 1018. Mohamed, Z.A., Carmichael, W.W., Hussein, A.A., 2003. Estimation of Azevedo, S.M.F.O., Carmichael, W.W., Jochimsen, E.M., Rinehart, K.L., microcystins in the freshwater fish Oreochromis niluticus in an Egyptian Lau, S., Shaw, G.R., Eaglesham, G.K., 2002. Human intoxication by fish farm containing a Microcystis bloom. Environ. Toxicol. 18, 137–141. microcystins during renal dialysis treatment in Caruaru-Brazil. Toxi- Pereira, P., Dias, E., Franca, S., Pereira, E., Carolino, M., Vasconcelos, V., cology 181-182, 441–446. 2004. Accumulation and depuration of cyanobacterial paralytic Bolin, D.W., King, R.P., Klosterman, E.W., 1952. A simplified method for shellfish toxins by the freshwater mussel Anodonta cygnea. Aquat. determination of chromic oxide (Cr2O3) when used as an inert Toxicol. 68, 339–350. substance. Science 116, 634–635. Pires, L.M.D., Karlsson, K.M., Meriluoto, J.A.O., Kardinaal, E., Visser, P.M., Bury, N.R., Eddy, F.B., Codd, G.A., 1995. The effects of the cyanobacterium Siewertsen, K., Donk, E.V., Ibelings, B.W., 2004. Assimilation and Microcystis aeruginosa, the cyanobacterial hepatotoxin in MC-LR, and depuration of microcystin-LR by the zebra mussel, Dreissena poly- ammonia on growth rate and ionic regulation of brown trout. J. Fish morpha. Aquat. Toxicol. 69, 385–396. Biol. 46, 1042–1054. Rai, A.K., 2000. Evaluation of natural food for planktivorous fish in Calabrese, E.J., 1999. Evidence that hormesis represents an ‘‘Over- lakes Phewa, Begnas and Rupa in Pokhara Valley, Nepal. Limnology compensation’’ response to a disruption in homeostasis. Ecotoxicol. 1, 81–89. Environ. Saf. 42, 135–137. Rodger, H.D., Turnbull, T., Edwards, C., Codd, G.A., 1994. Cyanobacterial Carmichael, W.W., 1994. The toxins of cyanobacteria. Sci. Am. 270, 78–86. (blue–green algal) bloom associated pathology in brown trout (Salmo Casarett, L.J., Amdur, M.O., Klaassen, C.D., Doull, J., 1996. Casarett and trutta L.) in Loch Leven. Scotland. J. Fish Dis. 17, 177–181. Doull’s Toxicology: The Basic Science of Poisons, fifth ed. McGraw- Sivonen, K., Jones, G., 1999. Chapter 3. Cyanobacterial toxins, in: Chorus, I., Hill, New York. p.1111. Bartram, J., (Eds.), Toxic Cyanobacteria in Water: A Guide to Their Chen, J., Xie, P., 2005. Tissue distributions and seasonal dynamics of the Public Health Consequences, Monitoring and Management. E & FN hepatotoxic microcystins-LR and -RR in two freshwater shrimps, Spon on behalf of WHO, London, pp. 55–124. Palaemon modestus and Macrobrachium nipponensis, from a large Smith, J.L., Haney, J.F., 2006. Foodweb transfer, accumulation, and depu- shallow, eutrophic lake of the subtropical China. Toxicon 45, 615–625. ration of microcystins, a cyanobacterial toxin, in pumpkinseed sunfish Chen, J., Xie, P., Guo, L.G., 2005. Tissue distributions and seasonal (Lepomis gibbosus. Toxicon 48, 580–589. dynamics of the hepatotoxic microcystins-LR and -RR in a freshwater Soares, R.M., Magalha˜es, V.F., Azevedo, S.M.F.O., 2004. Accumulation and snail (Bellamya aeruginosa) from a large shallow, eutrophic lake of the depuration of microcystins (cyanobacteria hepatotoxins) in Tilapia subtropical China. Environ. Pollut. 134, 423–430. rendalli (Cichlidae) under laboratory conditions. Aquat. Toxicol. 70,1–10. Chen, J., Xie, P., Zhan, D.W., Lei, H.H., 2007. In situ studies on the distri- Song, L.R., Sano, T., Li, R.H., Watanabe, M.M., Liu, Y.D., Kaya, K., 1998. bution patterns and dynamics of microcystins in a biomanipulation Microcystin production of Microcystis viridis (cyanobacteria) under fish – bighead carp (Aristichthys nobilis). Environ. Pollut. 147, 150–157. different culture conditions. Phycol. Res. 42 (Suppl), 19–23. Chorus, I., Bartram, J., 1999. Toxic Cyanobacteria in Water. A Guide to Their Tencalla, F., Dietrich, D., 1996. Biochemical characterization of micro- Public Health Consequences, Monitoring and Management. Published cystin toxicity in rainbow trout (Oncorhynchus mykiss). Toxicon 35, on behalf of WHO by F & FN Spon, London, p. 416. 583–595. Deblois, C.P., Rodriguez, R.A., Giani, A., Bird, D.F., 2008. Microcystin Tencalla, F., Dietrich, D., Schlatter, C., 1994. Toxicity of Microcystis aeru- accumulation in liver and muscle of tilapia in two large Brazilian ginosa peptide toxin to yearling rainbow trout (Oncorhynchus mykiss). hydroelectric reservoirs. Toxicon 51, 435–448. Aquat. Toxicol. 30, 215–224. de Figueiredo, D.R., Azeiteiro, U.M., Esteves, S.M., Goncalves, F.J.M., Vasconcelos, V., 1995. Uptake and depuration of the heptapeptide toxin Pereira, M.J., 2004. Microcystin-producing blooms – a serious global microcystin-LR in Mytilus galloprovincialis. Aquat. Toxicol. 32, 227–237. public health issue. Ecotoxicol. Environ. Saf. 59, 151–163. Williams, D.E., Daw, S.C., Kent, M.L., Anderson, R.J., Craig, M., Holmes, C.F.B., Fischer, W.J., Dietrich, D.R., 2000. Pathological and biochemical charac- 1997. Bioaccumulation and clearance of microcystins from salt water terization of microcystin-induced hepatopancreas and kidney damage mussels, Mytilus edulis, and in vivo evidence for covalently bound in carp (Cyprinus carpio). Toxicol. Appl. Pharmacol. 164, 73–81. microcystins in mussel tissues. Toxicon 35, 1617–1625. 216 G. Dong et al. / Toxicon 54 (2009) 208–216

Xie, L.Q., Xie, P., Ozawa, K., Honma, T., Yokoyama, A., Park, H.D., 2004. Zhao, M., Xie, S.Q., Zhu, X.M., Yang, Y.X., Gan, L.Q., Song, L.R., 2005. Effect Dynamics of microcystins-LR and -RR in the phytoplanktivorous of inclusion of blue–green algae meal on growth and accumulation of sliver carp in a sub-chronic toxicity experiment. Environ. Pollut. 127, microcystins in gibel carp (Carassius auratus gibelio). J. Appl. Ichthyol. 431–439. 22, 72–78. Xie, L.Q., Xie, P., Guo, L., Li, L., Miyabara, Y., Park, H.D., 2005. Organ Zhao, M., Xie, S.Q., Zhu, X.M., Yang, Y.X., Gan, L.Q., Song, L.R., 2006. distribution and bioaccumulation of microcystins in freshwater fishes Effect of dietary cyanobacteria on growth and accumulation of with different trophic levels from the eutrophic Lake Chaohu, China. microcystins in Nile tilapia (Oreochromis niloticus). Aquaculture 261, Environ. Toxicol. 20, 293–300. 960–966. Xie, L., Yokoyama, A., Nakamura, K., Park, H., 2007. Accumulation of Zimba, P.V., Khoo, L., Gaunt, P.S., Brittan, S., Carmichael, W.W., 2001. microcystins in various organs of the freshwater snail Sinotaia histrica Confirmation of catfish, Ictalurus punctatus (Rafinesque), mortality and three fishes in a temperate lake, the eutrophic Lake Suwa, Japan. from Microcystis toxins. J. Fish Dis. 24, 41–47. Toxicon 49, 646–652. Zurawell, R.W., Chen, H., Burke, J.M., Prepas, E.E., 2005. Hepatotoxic Zhao, M., 2006. The comparative study of effects of blue–green algae meal cyanobacteria: a review of the biological importance of micro- inclusion on fishes of different food habit. Doctoral thesis, Chinese cystins in freshwater environments. J. Toxicol. Environ. Health B 8, Academy of Sciences, China, 43–109. 1–37.