UDC: [504.064:577.25](564.141) MOLECULAR RESPONSES OF THE BIVALVE MOLLUSKS FROM THE COOLING POND AS A MODEL FOR PREDICTION OF CONTEMPORARY ENVIRONMENTAL CHALLENGES

To reveal a potential combined effect of pollution and continuos warming towards freshwater organisms, the molecular responses of stress and toxicity were compared du­ ring two years in the bivalve mollusks Anodonta anatina from the cooling pond of Khmel­ nytskyi Nuclear Power Plant (N­group) and from the ponds in the forestry (F­group) and agricultural (A-group) areas. Using Classification and Regression Tree (CART) software on the basis of a set of studied markers, characteristics for the distinguishing of N­group were selected. Firstly, it was the constantly elevated level of the vitellogenin­like protein in gonads and hemolymph of male specimens that witness the effect of endocrine dis­ ruption. The mussels of N­group demonstrated on oxidative injury with highest values of the Index of inter­seasonal variability for oxidized glutathione and metallothionein­related thiols among three groups, and constantly high level of apoptotic activity in the digestive gland. Remarkably higher levels of copper and cadmium within the tissues were detected in the N-group reflecting the industrial pollution. These metals were preferentially bound to metal­buffering protein metallothionein. A combination of these characteristics demon­ strates that stable elevated temperature jointly with industrial pollution caused a partial apoptotic elimination of cells with the nuclear lesions and provoked the exceeding of resistance of stress­related responses and reproduction­related functions in the mussels.


INTRODUCTION
Traditionally, at the assessment of environmental toxicity main attention is paid to individual hazardous substances that derived from pointed sources and caused big accidental spills [8]. However, novel environmental challenges biorisks are mainly connected to the complexity of harmful effects [6,24,35,38,45]. It is inconstant in time, composition, and its effect is difficult to be predicted. In these circumstances, typical responses of habitants to adverse effects can be distorted. Moreover, the resilience of portion of the Nichlava River near the boroughs of the city of Borshchiv (48°48′ N, 26°00′ E) that receives effluents from the region with intense agricultural activity and also from the city, in which a creamery is the most prominent source of pollution and the waste water plants are absent. The site with stable higher temperature of water (N) is located on the southern bank of the cooling pond of Khmelnytskyi Nuclear Power Plant (NPP) in Netishyn (in forestry area on the tributary of river Horyn, 50°21′ N, 26°38′ E). The connection by water between these three sites is absent, distance is about 100-200 km, so we may confirm that mollusks from three sites represent three different populations. The samplings were carried out simultaneously in all sites. Individuals were trans ported to the laboratory in cages with native water and treated within a day after the sampling procedure. Water samples were also collected and transferred to the labora tory in iced packs for the determination of chemical parameters. For each biochemical parameters eight digestive glands, gills and hemolymph samples were prepared indivi dually. Each procedure of tissue analysis was carried out at a temperature around 4 °C. Micronuclei were determined in the hemocytes, vitellogeninlike protein -in the gonads and hemolymph, all other biomarkers -in the digestive gland and/or gills.
Tissues were homogenized (1/10 w/v) in 0.1 M pH 7.4 phosphate buffer containing 100 mM KCl and 1 mM EDTA as well 0.1 mM PMSF for the inhibition of proteolysis. Homogenates were centrifuged at 6 000×g for 10 min and the resulting supernatant was used immediately for measurement.
For the measuring of enzyme activities, tissue was homogenized (1/10 w/v) in 0.1 M pH 7.4 phosphate buffer containing 100 mM KCl and 1 mM EDTA as well 0.1 mM PMSF for the inhibition of proteolysis. Homogenates were centrifuged at 6 000×g for 10 min and the resulting supernatant was used immediately for measurement. The protein concentration in the supernatant was determined by the method of Lowry et al. (1951) using bovine serum albumin as the standard [30]. Chemicals. Products: 5,5′dithiobis(2nitrobenzoic acid) (DTNB), thiobarbituric acid (TBA), reduced glutathione (GSH), glutathione reductase, 2vinylpyridine, serum albumin, phenylmethylsulfonyl fluoride (PMSF), βmercaptoethanol, NADH, NADPH, Sephadex G50, acetylAspGluValAsp pnitroanilide, tbutyl methyl ether were pur chased from Sigma. All other chemicals were of analytical grade.
Measurement of Body Indices. In each group individual's length, total, whole soft body, shell and separate tissues weight were recorded. The condition index (CI) of the tissues was calculated as the ratio: (drained mass of organ/soft tissues weight) × 100, the condition factor (CF) as the ratio: (total weight of mollusks/(shell length) 3 ) × 100.
Total glutathione concentration was quantified by the glutathione reductase recy cling assay [2]. To estimate the oxidized glutathione (GSSG) level, the protein free sam ple was treated with 2-vinylpyridine prior to assay run (60 min) at 2 % final concentration [23]. The rate of 5thionitrobenzoic acid formation was monitored spectrophotometri cally at 412 nm. The redox-index of glutathione (RI GSH) as the ratio of concentrations ([Total glutathione]-[GSSG])/[Total glutathione] was also calculated. Standards were prepared from reduced glutathione, and concentrations were expressed as µmol per g wet weight.
The activity of lactate dehydrogenase (EC 1.1.1.27) was determined in digestive gland using the UV assay with pyruvate by determining of the amount of NADH oxida tion at 340 nm [7]. The extinction was read after every minute interval for a period of 4 min. A molar extinction coefficient of 6.22·10 6 M 1 ·cm 1 was used.
Markers of Exposure. Vitellogeninlike proteins (VtgLP) were evaluated in go nads and hemolymph as alkalilabile phosphate level according to Blaise et al. (1999) [8]. Briefly, 200 ml of gonad homogenates or 1.5 ml hemolymph was mixed with t-butyl methyl ether for 30 min at room temperature. The ether phase was thoroughly mixed with 1 M NaOH for 60 min at 37 °C. The levels of free phosphates were determined by the phosphomolybdenum assay.
Metallothioneins (MTs) were determined from thiols measure with DTNB accord ing to the method of Viarengo et al. (1997) after the ethanol/chloroform extraction. The levels of MTs were calculated assuming the relationship: 1 mol MTs = 20 mol GSH and expressed as mg of MTs per gram of FW tissues [45].
Caspase3 colorimetric assay was based on the hydrolysis of peptide acetylAsp GluValAsp pnitroanilide (AcDEVDpNA) by caspase3, resulting in a release of the pnitroaniline (pNA) moiety. pNitroaniline was detected at 405 nm (ε mM =10.5). The con centration of the pNA released from the substrate was calculated from the absorbance values at 405 nm [9].
To assay the genotoxicity of water contaminants, the micronucleated hemocytes in hemolymph (MN) were calculated by the presence of miniature nuclei in the cytoplasm of postmitotic cells [4]. The frequencies of MN were expressed per 1,000 cells stud ied. Also nuclear lesions were scored into other categories: lobed nuclei (L), dumbbell shaped or segmented nuclei (S), and kidneyshaped nuclei (K). Cells with other than MN abnormalities were counted and the result was expressed as a mean value of the sums (L+S+K) for all individual lesions observed per 1,000 cells. Not less than 2,000 cells were calculated in each sample.
Metal Determination. Zn, Cu and Cd concentration was measured in samples of weighed tissue of digestive gland (250 mg) of each specimen, and pooled eluate of low weight thermostable proteins after chromatography (10 mL) in each group of mussels (in triplicate). Chromatography of thermostable extract was accomplished on Sephadex G50, as described by   [15]. The samples were dried for 24 h at 105 °C, and then digested with 5 ml HNO 3 for 3 h at 105 °C under pressure, using an acid-cleaned Teflon bomb. The Cu and Zn content was analyzed by atomic absorption spectrophotometry against certified standards on spectrometer C-115, ("Lomo", Russia) and Cd, on graphite furnace atomic absorption spectrometer S-600 ("Selmi", Ukraine). The reliability of the measurements towards several of the selected elements was as sessed by analyzing ERM-CE 278 certified reference material (mussel tissue); recover ies of metals were between 90 % and 110 %.
Metal concentration in tissues and MTs forms was expressed as μg·g 1 fresh weight (FW), in MTs also -as nmol·g 1 FW.
Statistical Analysis. All measurements were expressed as means ± standard deviation (SD) of 8 individuals for each biochemical parameters and of 14 individu als for each morphological index. To assess which biochemical indicator was the most affected by local conditions the index of inter-seasonal variability (VI) defined as the ratio between the maximal and minimal mean responses (IP = Mmax / Mmin) and was calculated for each biochemical parameters as average values during three seasons by Vidal M.L. et al. (2002) [46].
Since data were not normally distributed (Lilliefors' test), nonparametric tests (Kruskall-Wallis ANOVA and Mann-Whitney U-test) were performed (significant at p < 0.05). Data were subjected to principal component analysis (PCA) to evaluate the biomarkers relation both in different sites and different seasons. Classification tree was built using the classification and regression tree (CART) software on the basis of all determined biological characteristics jointly in three seasons. CART builds the tree by recursively splitting the variable space based on the impurity of the variables to deter mine the split till the termination condition is met. Set of biological parameters were considered as predictor (independent) variables and site was considered as predic tive (dependent) variable. The partitioning intends to increase the homogeneity of the two resulting subsets or nodes, based on the response variable. The partitioning stops when no splitting rule can improve the homogeneity of the nodes significantly.
All statistical calculations were performed with Statistica v 8.0 and Excel for Win dows2000.

RESULTS
Measurement of the morphologic indices of the mollusks showed a decrease in weightrelated indexes during the summer at sites F, and especially N, when compared to the values in spring, with a further recovery or elevation in sites A and N, but it re mained low during autumn at site F (Table 1). In the site N, most indices in the same seasons (periods III and IV) were not differed significantly. Comments: a -statistically significant differences between sites; b -statistically significant differences between periods of sampling compared to I values. Always P < 0.05. Примітки: a -відмінність між сайтами статистично вірогідна, b -відмінність між періодами відбору статистично вірогідна порівняно зі значенням періоду I, Р <0,05.
Study revealed seasonal variability of the indices of oxidative stress and Lactate dehydrogenase activity in all studied groups. The catalase activity was higher in the A group, particularly in autumn. In the Ngroup from the cooling pond, its value was lesser among the groups of comparison in most cases (Fig. 2) On the other hand, the level of TBARS in the N-group belonged to highest levels detected in this study, particularly in the summer.
The activity of LDH in the digestive gland demonstrated the intersite variability in summer and autumn. In autumn it reached up to twotimes higher level in the mussels from cooling pond. The level of GSH remained rather high in the Ngroup, whilst in the group from pol luted agricultural area it was low in the most groups of comparison (Fig. 3). However, the levels of GSSG and Redox Index of GSH were in some cases extremely different in these mussels from other groups with higher value of GSSG and low value of RI GSH. The calculation of the Index of Variability of the characteristics of stress response demonstrated that the variability of GSSG concentration was most prominent in the N group and belonged to main characteristics of variability for this site among all studied indices ( Table 2).  Table 2. The assessment of the markers of exposure showed that the level of MT was high est in the Ngroup in the two seasons (Fig. 4). Besides that, MT level belongs to most variable indices for this group of mussels ( Table 2). One of most remarkable molecular characteristics of the Ngroup was stable elevated activity of caspase3 (Fig. 4). It was accompanied by elevated level of the hemocytes with micronuclei and nuclear abnor malities, as compare to the reference Fgroup (Fig. 5). The level of caspase3 and genotoxicity was particularly high in the second year of study.

Parameters
The total and MTrelated concentrations of Cu, Zn and Cd in the digestive gland are represented in Fig. 6. Prominent seasonal differences were shown for metal concentra tion in the tissue and in MTs. In summer, in groups F and A, highest levels of metals in MTs were indicated, and in the group N, lowest their values were correspondent to this season. When three groups were compared, the concentration of Cu in MTs was always highest in group N. In spring and autumn periods, MTs in this group contained also highest concentrations of Zn and Cd. Calculation of metal concentrations in MTs forms showed that the rate of Zn in MTs was up to tenfold higher than Cu, and only in group N in summer the rate of Cu in MTs was 1.5 fold higher than of Zn. The concentra tion of MTs elevated prominently from spring to summer-autumn, especially in group N. In summer and autumn, intersite differences of MT concentration were detected with highest value for group N and lower for group A.
According to PCA (Fig. 7, A), 60.65 % of data belonged to Factors 1 and 2. The specimens from site N were located jointly, whilst the groups from other sites were distributed in seasondependent regularity between different locations. So, despite the variability of GSSG level, the similarity of the responses of MT and caspase3 deter mined the constant general characteristic of the specimens from this site. To elucidate main partitioning markers, we used the CART algorithm (Fig. 7, B). When all of the biological parameters at the three groups were compared, VtgLP level in the gonads was represented at the node as partitioning criterion for the groups F and N. F and A groups were separated by catalase activity in the digestive gland. The resulting confu sion matrix showed an overall classification accuracy of 90 %. Any of the terminal nodes did not contained misclassified mussels. The best classification was predicted for group N, followed by groups F and A.

DISCUSSION
Despite an unique experience of the effect of stable elevated temperature in the cooling ponds of NPP, the data concerning the molecular responses of aquatic animals in such sites are scant. The mussels from typical polluted and reference field sites were studied previously, and the particularities of their responses were stressed [15]. For the newly selected group of mussels, derived from the Nsite, the industrial pollution, effect of low doses of radiation and also constantly elevated ambient temperature were sus pected to be the reasons for the peculiarities of responses.
Elevated temperature provoked the activation of stressrelated systems, as it is evident from our data and other results on mollusks [17,42] and fish [31]. Two main characteristics of oxidative stress were distinguished in the mussels from site N. Firstly, a high lipid peroxidation level was found in summer period. Obviously, this peculiarity was caused by particularly high temperatures. It is known that high temperatures and temperaturerelated hypoxia in the aquatic animals affect the antioxidant defense sys tem [42]. For example, similarly to the response of oxidative stress markers at site N in summer, longterm temperature stress has been associated with oxidative damage, decreased values of GSH and RI GSH and also increased activities of antioxidants in the gills of Perna viridis [43]. High rates of injury to mollusks at site N was confirmed by the fact that decreased weight and CI of gills in summer were only found here. We can speculate that the limited ability to withstand high temperatures is because of high levels of anaerobic activity (high LDH activity). The metabolic depression and a shift from aerobic to anaerobic metabolism is a typical response to unfavorable conditions in mussels [3,42,49]. Pollution by toxic metals at this site may contribute to adverse environmental effect. The effect of increased temperature and exposure to Cd appeared to override a capability for aerobic energy production in a model marine bivalve, the eastern oyster Crassostrea virginica, resulting in impaired stress tolerance causing low condition index and elevated LPO [28]. Low weight intracellular thiols, GSH and MT, were selected by highest Index of Variability in the present study. It can be explained by as the most targeted their relation crucial biological functions in general stress response, particularly scavenging of reac tive oxygen species, storage and transport of metal ions [10,44]. Upregulation of these stressrelated thiols is considered as one of the mechanisms involved in the adaptive response to low dose radiation exposure in aquatic animals [10,33]. On the other hand, the decreased level of GSH under the influence of low-level radiation (below 20 cSv), accompanied by increase of the level of lipid peroxidation products in human plasma, was revealed long after the accident (4-7 years) in the liquidators and children of ex posed mothers from Chornobyl atomic power station area [26]. Incidentally, an opposite effect but also with high magnitude compare to the reference values, the elimination of MT but elevated GSH level was detected in the snails from the pond in the vicinity of Chornobyl area [21].
As reported Verlecar et al. (2007), increased activities of superoxide dismutase and other antioxidant enzymes but decreased values of GSH were indicated in the digestive glands of Perna viridis under longterm (14 days) exposure to elevated temperature [43]. Contradictory results in the present study could be explained by different regimes of heating. An increase of anaerobic glycolysis is well approved sign of toxic effect in mollusks [3]. This effect was shown in our study for the mussels from Nsite.
In mollusks from site N, the function of MTs was activated in autumn due to their defense against toxic metals. Group N was also distinguished by Cuenriched MTs. It was correspondent to highly elevated MT concentration in most cases. This group was also characterized by high levels of caspase3 (constantly). The peculiarities of the environment in the sites related to the cooling ponds of NPP are studied scant. High level of Cu in the environment could be one of their attributes [16]. To date, twoyear study in the vicinity of a NPP in northeastern France shown high levels of Cu in the area low sensitivity of indigenous Dreissena polymorpha to the spatial effects [24]. Elevated ambient temperature in the N-site can provoke specific effects such as high MT and caspase3 levels. These results are well coordinated with the information concerning the expression of heat shock proteins in goldfish inhabiting Gaobeidian Lake in Beijing, China. Water of this lake was reported to be moderately polluted and have an elevated temperature. The up-regulation of the heat shock proteins suggested that fish under these specific environmental conditions were experiencing a complex stress process [48]. Piano et al. (2004) reported that the levels of MT were significantly increased in the tissues of Ostrea edulis individuals exposed to thermal stress (35 °C, 1 hour) and allowed to recover at 18 °C for 24 hours [34].
Separate studies have been undertaken in Ukraine in the vicinity of NPPs, and absence of the effects of heated discharge waters of NPP on the zoobenthos was de mon strated [41]. However these studies were mostly conducted on the composition of populations and histochemical criterions. Elucidation of reasons for the biochemical abnormalities detected in our study at site N, needs a more prolonged and comprehen sive study.
Dreissena polymorpha populations in water basins near the ChNPS also showed no significant effect of radiations [18]. Besides, we explored a more comprehensive bat tery of biomarkers that showed an early warning of the effects of a nearby NPP whereas histological and morphological effects were less evident.
Despite information concerning the activity of caspases in bivalve mollusks being scant, caspase3 activity seems to represent a valuable marker for the determination of toxic effects. Romero et al. (2011) reported that caspase genes in Mytilus galloprovincialis have extremely high expression levels within the gland and gills, where the apoptotic process is highly active due to the clearance of damaged cells, and the role of apoptosis in the defense against pathogens, contributes to the reasoning for including this assay in the set of biomarkers for mollusks [27,29,37]. In general, the activation of MTs and apoptotic functions in this group of mussels witness about successful response to heat stress in the mussels.
A high seasonal dependence for the biochemical markers of pollution is well known in marine mollusks, especially for oxidative stress related indices and MTs [21,39,47]. This can be explained by better conditions for growth and a faster metabolism due to optimal surrounding conditions (food availability, temperature). Studies of seasonal ef fects in freshwater mollusks are fragmental [5,46]. Nevertheless, seasonal regularity in the mussels from cooling pond was distorted as compare to the other sites, both clean and polluted. This peculiarity reflects obviously the effect of constantly elevated temperature in this pond.
Mollusks from sites F and N were distinguishing by levels of VtgLP. Only male mol lusks were studied in order to avoid the effect of sex on VtgLP [20]. De Lafontaine et al. (2000) [13] showed a lack of significant effect of mussel size or sex on a response of this biomarker in bivalve Dreissena polymorpha. Therefore, we suggested that environmen tal estrogens were present at site N, despite earlier skepticism concerning the validity of levels of VtgLP in the hemolymph of mollusks as a biomarker [20,36].
In summary, a battery of non-specific markers of stress and markers of exposure and application of differential and classification tree analyses allowed imminent and per sistent danger related to the endocrine disruption and activated apoptosis to be speci fied in the indigenous freshwater mollusks from NPP. The resilience of stress-related responses of cellular lowweight thiols could be classed as critical in these mussels.