**2. Interest for bioaccumulation capacities**

Bioaccumulation is defined as the accumulation of a target compound in an organism relative to its concentration in the surrounding environment. It depends on several intrinsic (e.g., detoxification abilities, lipid content, maturity, or sex of the organism) and environmental (e.g., water temperature or pH) factors that have to be considered in bioaccumulation assessment. The ability of species to bioaccumulate contaminants could represent an indirect evaluation tool of water quality allowing going beyond the limitations associated both to the sporadic nature of water sampling and to potent high dilution of contaminant in the water matrix. In fact, contrary to water matrix, using attached filter-feeding organisms makes contamination measurements representative of their living environment. Measurements in appropriate biological matrices such as mussels will (i) limit the variability (temporal integration) of measurements as compared to measurements on water samples, (ii) convey the degree of contamination of water bodies more reliably, and thus facilitate comparisons. Zebra mussel, *D. polymorpha*, has a high filtration activity and filters between 0.018 and 0.402 L/mussel/h [8] that may give them a capacity to accumulate environmental contaminants.

#### **2.1. Pharmaceuticals**

**1. Introduction**

44 Organismal and Molecular Malacology

The continental ecosystems are subjected to many stresses related to the human activities through the emission of a large number of molecules whose toxicological and ecological effects are currently poorly studied. There needs to be further development of tools not only for highlighting the exposure of nontarget organisms to these molecules but also for assessing possible risk to the stability of their population and, by domino effect, to the biodiversity and ecological functioning of the ecosystems. Water quality evaluation, management, and protection represent a strong challenge of our society to maintain the sustainability of freshwater ecosystems and protect the biodiversity and ecosystem services for human benefits such as the potabilization or recreational uses. This challenge results in particular in a better management of all the new chemical products likely to be introduced into the environment (program recording of chemical substances REACH) as well as the evaluation of its ecological state (European framework directive on water 2000/60/CE). Currently, the texts of the European Community dated October 23, 2000, set a number of environmental objectives in order to reach the "chemical and ecological good state" of freshwaters. Although environmental quality standards appear in European texts, this monitoring program is primarily based on the evaluation of the level of contaminants in water and on the biological diversity of the aquatic communities. In addition to the biocenotic approaches, many researches focus on biological responses evaluated at sub-individual and individual levels, gathered under the generic term of "biomarkers." Among these biomarkers, a very close attention is paid to the processes whose disturbances may induce a chain reaction of damage on the population and community levels. These biomarkers include physiological functions such as the energy metabolism, the reproduction, or the immune system. The development of biomarkers is conducted on a large number of species known as "environmental sentinels" with respect to criteria related to their sedentary lifestyle, their large distribution, knowledge on their biology and ecology, and the feasibility of using them for experimental exposures. Among these species, bivalves appear to be a genuine tool for the detection of the chemical and biological contamination of the environment [1]. Indeed, the filter feeders have the ability to accumulate and concentrate the contamination in a time- and dose-dependent manner that allows to recall events of pollution prior to the sampling point. The studies are mainly based on the marine species of economic interest as the mussels (*Mytilus* sp.), the clams (*Ruditapes* sp.), and the oysters (*Crassostrea* sp.). However, at the continental level, a model organism, zebra mussel *Dreissena polymorpha* (Pallas, 1771), proves its interest and could represent the freshwater counterpart of *Mytilus* sp. [2]. Zebra mussel is an invasive species originated from Ponto-Caspian region in central Europe. It has spread around the world and became invasive in Europe and North-America freshwater ecosystem, where it has been observed in the great lakes at the end of the 1980s. *D. polymorpha* is characterized by a large abundance and widespread distribution, great filtration capacities leading to high levels of xenobiotic accumulation, and a good tolerance of environmental stressors. Being sedentary, their individual responses may be correlated to the quality of the site. Moreover, dreissenids present a lifetime of several years, can be sampled all year around, and their biology and ecology are well known [3]. Therefore, zebra mussel constitutes a valuable bioindicator species, largely used as freshwater biomonitoring tool [4–7].

There are only a few studies dealing with the uptake and bioaccumulation of pharmaceuticals in freshwater mussels; more studies are available for fish. Quantification of pharmaceuticals at trace levels in mollusks is often analytically challenging, due to the complexity of the biological matrix, requiring suitable extraction and purification steps combined with selective and sensitive detection techniques. Various methods are employed for the extraction of pharmaceuticals in bivalves: pressurized liquid extraction [9, 10], microwave-assisted extraction [11, 12], sonication [13], or QuEChERS (Quick, Easy, Cheap, Effective, Rugged, and Safe) extraction [14, 15]. The last one has the advantages of a low solvent consumption combined with rapid extraction and clean-up steps. Most current analytical methods for the separation and detection of pharmaceuticals are based on gas chromatography-mass spectrometry (GC-MS) or more frequently on liquid chromatography (LC) coupled to ultraviolet (UV), diode array (DAD), fluorescence (FL) detectors, or mass spectrometry (MS) and tandem mass spectrometry (MS/MS). Due to large differences in their physicochemical properties (e.g., polarity and solubility), efficient extractions of all the targeted pharmaceuticals are difficult to obtain, especially in multiresidue methods. As a consequence, the few studies related to the bioaccumulation of pharmaceuticals in freshwater mussels generally focused on a very small number of target compounds [15–17]. Li et al. [18] studied the distribution of 22 antibiotics in 11 mollusk species sampled from the Bohai Sea of China, with concentrations up to 1575.10 μg/kg for quinolones, 76.75 μg/kg for sulfonamides, and 36.21 μg/kg for macrolides. Du et al. [19] studied the bioaccumulation of several selected pharmaceuticals in two mussel species (*Pondhorn mussel* and *Paper pondshell* mussel) and fish, among other matrices, collected from a lower-order effluent-dependent stream in Texas, USA. They concluded that pharmaceuticals accumulated to higher concentrations in invertebrates in comparison to fish. Fluoxetine, carbamazepine, and diphenhydramine were quantified at concentrations inferior to 22 μg/kg, whereas elevated concentrations of the antidepressant sertraline and its primary metabolite desmethylsertraline were observed in both mussel species at levels higher than 130 μg/kg. Indeed, as they could exhibit equal or higher bioaccumulation capacities, it is relevant to include metabolites and transformation products in bioaccumulation studies [15, 19]. Bioconcentration factors (BCFs) have been calculated for the anti-inflammatory drug diclofenac in different studies. BCF of 10 have been displayed for blue mussels exposed during 8 days in aquarium to diclofenac at 1 μg/L [20], and comprised between 4 and 13 in another study where diclofenac was introduced in mesocosms at concentrations between 0.05 and 5 μg/L [15]. This study shows the accumulation in *D. polymorpha* as a function of exposure time, and highlights the presence of one of its transformation products in this bivalve species. The bioaccumulation of the contraceptive hormone levonorgestrel was also examined in *D. polymorpha* [21]. The lowest concentration (0.312 μg/L) was 100-fold bioconcentrated within 4 days. A decrease of the BCF was observed within 1 week for the highest tested concentrations (3.12 and 6.24 μg/L).

#### **2.2. Protozoa pathogens**

Protozoa such as *Toxoplasma gondii, Cryptosporidium parvum*, and *Giardia duodenalis* represent a threat to human health because they are infectious at low doses and their (oo)cysts can remain infectious in the environment for months due to a robust wall [22, 23]. *T. gondii* is the agent of toxoplasmosis and infection is mainly acquired by the ingestion of food or water that is contaminated with oocysts. *C. parvum* and *G. duodenalis* are intestinal protozoa found in mammals, fish, and birds. (Oo)cysts can be transmitted via drinking and recreational waters contaminated by agricultural and urban runoff [24]. Protozoa are difficult to detect in environmental samples. Actually, no standardized method has been developed for the detection of *Toxoplasma* oocysts in water samples. On the contrary, *Cryptosporidium* oocysts and *Giardia* cysts are currently detected in samples from 100 L of filtered water, using indirect immunofluorescence after immunomagnetic separation (ISO 15553:2006). Limitations have been identified in this method: it requires large volumes of filtered water and high parasite concentrations, it is costly and time-consuming, and thus does not allow for rapid routine detection. Moreover, filtration and purification techniques from water supplies can yield variable results depending on water quality, sampling period, place, and quantity [24].

Several studies have demonstrated the interest of using aquatic organisms to assess and monitor the quality of water bodies [25] and *D. polymorpha* can represent a tool to reveal protozoan contamination. Indeed, during a laboratory experiment, zebra mussels were exposed to various *C. parvum, G. duodenalis*, and *T. gondii* (oo)cyst concentrations for 1 week, and bioaccumulation by mussels proportionally to ambient contamination was observed. This study showed for the first time *T. gondii* accumulation by zebra mussel [26]. In a second in vivo experiment, *D. polymorpha* were exposed to 1000 *T. gondii* oocysts per individual and per day for 21 consecutive days, followed by 14 days of depuration time in protozoa-free water. *T. gondii* was detected in all organs in a time-dependent manner, but oocysts were found in greater amount in the hemolymph and mantle tissues compared to the other organs (gills, gonads, and foot). This study also shows that mussels accumulate protozoan proportionally to water contamination. Moreover, *T. gondii* was still present in the mussel tissues for up to 14 days postinoculation, reflecting the integrative character of *D. polymorpha* [27]. A field study was also carried out in order to test protozoa detection under environmental conditions. *D. polymorpha* were caged for two seasons (autumn and spring) for 1 month upstream and downstream of wastewater treatment plants (WWTPs). Concerning the study realized in autumn, *T. gondii* was detected in mussels caged downstream of two WWTPs. In spring, *T. gondii* was detected upstream of one WWTP. These results highlight the interest to use *D. polymorpha* caged as a new effective tool in sanitary biomonitoring of water bodies [28].

#### **2.3. Cyanobacteria**

inferior to 22 μg/kg, whereas elevated concentrations of the antidepressant sertraline and its primary metabolite desmethylsertraline were observed in both mussel species at levels higher than 130 μg/kg. Indeed, as they could exhibit equal or higher bioaccumulation capacities, it is relevant to include metabolites and transformation products in bioaccumulation studies [15, 19]. Bioconcentration factors (BCFs) have been calculated for the anti-inflammatory drug diclofenac in different studies. BCF of 10 have been displayed for blue mussels exposed during 8 days in aquarium to diclofenac at 1 μg/L [20], and comprised between 4 and 13 in another study where diclofenac was introduced in mesocosms at concentrations between 0.05 and 5 μg/L [15]. This study shows the accumulation in *D. polymorpha* as a function of exposure time, and highlights the presence of one of its transformation products in this bivalve species. The bioaccumulation of the contraceptive hormone levonorgestrel was also examined in *D. polymorpha* [21]. The lowest concentration (0.312 μg/L) was 100-fold bioconcentrated within 4 days. A decrease of the BCF was observed within 1 week for the highest tested concentrations

Protozoa such as *Toxoplasma gondii, Cryptosporidium parvum*, and *Giardia duodenalis* represent a threat to human health because they are infectious at low doses and their (oo)cysts can remain infectious in the environment for months due to a robust wall [22, 23]. *T. gondii* is the agent of toxoplasmosis and infection is mainly acquired by the ingestion of food or water that is contaminated with oocysts. *C. parvum* and *G. duodenalis* are intestinal protozoa found in mammals, fish, and birds. (Oo)cysts can be transmitted via drinking and recreational waters contaminated by agricultural and urban runoff [24]. Protozoa are difficult to detect in environmental samples. Actually, no standardized method has been developed for the detection of *Toxoplasma* oocysts in water samples. On the contrary, *Cryptosporidium* oocysts and *Giardia* cysts are currently detected in samples from 100 L of filtered water, using indirect immunofluorescence after immunomagnetic separation (ISO 15553:2006). Limitations have been identified in this method: it requires large volumes of filtered water and high parasite concentrations, it is costly and time-consuming, and thus does not allow for rapid routine detection. Moreover, filtration and purification techniques from water supplies can yield variable results depending on water quality, sampling period, place, and

Several studies have demonstrated the interest of using aquatic organisms to assess and monitor the quality of water bodies [25] and *D. polymorpha* can represent a tool to reveal protozoan contamination. Indeed, during a laboratory experiment, zebra mussels were exposed to various *C. parvum, G. duodenalis*, and *T. gondii* (oo)cyst concentrations for 1 week, and bioaccumulation by mussels proportionally to ambient contamination was observed. This study showed for the first time *T. gondii* accumulation by zebra mussel [26]. In a second in vivo experiment, *D. polymorpha* were exposed to 1000 *T. gondii* oocysts per individual and per day for 21 consecutive days, followed by 14 days of depuration time in protozoa-free water. *T. gondii* was detected in all organs in a time-dependent manner, but oocysts were found in greater amount in the hemolymph and mantle tissues compared to the other organs (gills, gonads, and foot).

(3.12 and 6.24 μg/L).

quantity [24].

**2.2. Protozoa pathogens**

46 Organismal and Molecular Malacology

The eutrophication of aquatic ecosystems, associated to the climate change, enhances the frequency and the severity of cyanobacterial proliferations. Cyanobacteria are photosynthetic organisms producing endotoxins such as neurotoxins, hepatotoxins, dermatotoxins, and cytotoxins, threatening target organisms and humans [29]. The World Health Organization derived a guideline value of 1 μg/L in tap water for the hepatotoxins microcystins (MCs), the most reported cyanotoxins in fresh waters worldwide. In addition to being exposed to cyanotoxins through drinking, humans can also be exposed by ingestion or inhalation of contaminated water during recreational activities [30, 31], or by trophic transfer of MCs as demonstrated between invertebrates and vertebrates [32]. The bioaccumulation of MCs and their effects on organisms are overall quite well documented. However, some cyanobacteria and diatom strains may also produce the neurotoxin BMAA (β-N-methylamino-L-alanine) [33–35], a nonproteinogenic amino acid that has been associated with neurodegenerative diseases like amyotrophic lateral sclerosis/Parkinsonism dementia complex (ALS-PDC) that occurred on the Guam Island [36]. The bioaccumulation of BMAA in marine organisms is little documented [37–39]. However, BMAA accumulation and impact on freshwater organisms during laboratory exposure or field investigations remains poorly investigated [35, 37].

Even though monitoring programs are in place to evaluate the concentration of cyanotoxins in fresh waters, prompt management of ecological and public health risk remains difficult due to the high spatiotemporal variability of the cyanobacterial proliferations. Freshwater mussels, and particularly *D. polymorpha*, are known to ingest phytoplankton and cyanobacteria and accumulate cyanotoxins at the laboratory [40, 41] and in the field [42–44]. Therefore, the use of sentinel species to integrate and reveal early MC and BMAA contamination at low cyanobacterial or phytoplankton densities may allow a preventative management of contaminated sites. As MC accumulation by *D. polymorpha* has been largely demonstrated, we primarily focused on the capacity of the mussel to accumulate BMAA using a laboratory approach. The mussel *D. polymorpha* was exposed to exogenous BMAA at 500 μg/L for 3 weeks, and we followed up both free and protein-bound BMAA in tissues. The free amino acid BMAA was detected from day 2 to day 21 and really high concentrations (reaching 283 μg/g dry weight) were quantified in mussels at the end of the contamination experiment. BMAA was therefore taken up by *D. polymorpha* and the internalized concentration of the toxin, detected as its free form, increased linearly over time (*r*² = 0.9635) while no steady state was reached (**Figure 1**).

**Figure 1.** Concentration of free BMAA in *D. polymorpha* with fitted linear regression after free exogenous exposure for 21 days at 500 μg/L. Error bars denote standard deviation of *n* = 5 (except for D21, *n* = 2).

It is conceivable that this freshwater species would have continued to accumulate BMAA from the medium after a longer exposure time. Concentrations of total BMAA in mussels of day 2 and day 21 were not significantly different from the concentrations of free BMAA in the same individuals (data not shown). Thus, the incorporation into or in association of BMAA to proteins was not evident in this study.

Overall current projects also include other freshwater bivalves as potent predictive sentinel species of the presence of MCs and BMAA in freshwater environments through laboratory and *in situ* approaches. The results that will be achieved in these projects will facilitate the long-term tracking of the contamination of ecosystems by cyanotoxins, which will provide an advance in the knowledge about the ecodynamic of cyanotoxins and the main conditions of human exposure. This step forward could constitute a decision-making tool for public health authority in charge with the management of food safety risks.

#### **3. Interest for biomarker measurement**

Different biological responses have been developed as biomarkers on freshwater mussels. Biomarkers related to energy metabolism could be an interesting prognostic tool bearing in mind the close relationships existing between the energy balance and health of individuals. As well, the reproduction function is a key element for the preservation of the species in their environment and can be disrupted by numerous stresses of anthropogenic origin. Furthermore, organisms are subjected to various pathogens. Thus, interaction with hemocytes, a key component of the immune system, undoubtedly occurs. Exploring these three relevant functions will permit to develop early warning and sensitive diagnosis tools with an ecological relevance which seems to be promising for the early assessment of the ecosystem degradation resulting from damaging contaminants.

#### **3.1. Energy and metabolism**

It is conceivable that this freshwater species would have continued to accumulate BMAA from the medium after a longer exposure time. Concentrations of total BMAA in mussels of day 2 and day 21 were not significantly different from the concentrations of free BMAA in the same individuals (data not shown). Thus, the incorporation into or in association of BMAA to

**Figure 1.** Concentration of free BMAA in *D. polymorpha* with fitted linear regression after free exogenous exposure for 21

Overall current projects also include other freshwater bivalves as potent predictive sentinel species of the presence of MCs and BMAA in freshwater environments through laboratory and *in situ* approaches. The results that will be achieved in these projects will facilitate the long-term tracking of the contamination of ecosystems by cyanotoxins, which will provide an advance in the knowledge about the ecodynamic of cyanotoxins and the main conditions of human exposure. This step forward could constitute a decision-making tool for public health

Different biological responses have been developed as biomarkers on freshwater mussels. Biomarkers related to energy metabolism could be an interesting prognostic tool bearing in mind the close relationships existing between the energy balance and health of individuals. As well, the reproduction function is a key element for the preservation of the species in their environment and can be disrupted by numerous stresses of anthropogenic origin. Furthermore, organisms are subjected to various pathogens. Thus, interaction with hemocytes, a key component of the immune system, undoubtedly occurs. Exploring these three relevant functions will permit to develop early warning and sensitive diagnosis tools with an ecological relevance which seems to be promising for the early assessment of the ecosystem

proteins was not evident in this study.

48 Organismal and Molecular Malacology

authority in charge with the management of food safety risks.

days at 500 μg/L. Error bars denote standard deviation of *n* = 5 (except for D21, *n* = 2).

**3. Interest for biomarker measurement**

degradation resulting from damaging contaminants.

All the biochemical and physiological processes involved in the vital cycle of the organisms depend strictly on the energy metabolism. Besides, within the framework of an evaluation of the quality of the environment, the variations at the molecular level (e.g., enzymatic activities) are one of often first answers to the stress; they thus represent useful points in the confirmation of the toxic effects before these last ones are perceptible at superior biological levels of organization (cellular, tissular, and physiological). Biomarkers related to energy metabolism could then constitute interesting prognostic tools considering the casual relationships linking the energy balance and the health status of individuals. A very large number of biological (adenylate energy charge, energy reserves, cellular energy allocation, scope for growth, etc.) responses to stress linked to the metabolism and energy of living organisms may be investigated. While the concept is clear, linking cellular or biochemical responses to the individual and population level remains relatively scarce. However, several studies underline the relevance of cellular responses to extrapolate effects at higher levels of biological organization in *D. polymorpha* under laboratory [45] and environmental conditions [46]. Zebra mussel and all other animals obtain the energy necessary for life from an external source. So, in heterotrophic organisms, the acquisition of energy is controlled by feeding and the subsequent breakdown of food to release the energy contained (assimilation). Feeding rate/clearance was studied in different species of mussels (particularly blue mussel) exposed to various contaminants [47–49]. Generally, feeding appears unaffected and/or less sensitive compared to biochemical responses. Inversely, during the study of different zebra mussel populations, among several genes involved in cellular metabolic activities, detoxification process, oxidative stress, and digestive functions, digestive enzyme genes expression appears particularly relevant to discriminate the sites according to the contamination level [50]. This study has also pointed out different sources of variability in gene expression (individual size, season, trophic resources, and origin of mussels) which are inevitable in natural fluctuant environment. Accordingly, in *D. polymorpha*, we decided to investigate the breakdown of food's mechanisms and particularly the digestive capacity. Digestion in bivalves occurs through a biphasic process involving an extracellular digestion phase of ingested food particles followed by an intracellular digestion phase. The first step occurs within the stomach under the mechanical and chemical actions of the crystalline style (CS), a gelatinous rod saturated with digestive enzymes. The CS protrudes from the style sac into the stomach and revolves against a cuticular structure, the gastric shield, promoting the grinding of food materials and their mixing with the enzymes released by its former dissolution. This extracellular digestion phase allows a preliminary breakdown of food particles into molecules small enough to enter the digestive diverticulae where digestion is completed intracellularly within phagocytic digestive cells. In this way, amylolytic and cellulolytic activities in the crystalline style and the digestive diverticulae of the freshwater bivalve *D. polymorpha* were characterized [51]. In the same time, to avoid the influence of biotic parameters (age, size, etc.), we proposed to use individual from the same population transplanted to the study sites (see part 5). Thus, during active biomonitoring carried out with *D. polymorpha*, in the catchment of the River Vesle, upregulation of digestive activities in digestive gland was observed in mussels exposed to various chemical contaminants and thus potentially facing an increase in their energy expenditure (general adaptation syndrome) [52]. However, this upregulation has not been recorded at all contaminated sites, lower levels of pollution or higher levels of food availability probably explaining the low need for organisms to "invest" in their digestive activities (least energy requirement). These results highlight adjustment of digestive enzymatic activities according to intake and energy requirements of the organism, factors which, if not taken into account, constitute potential confounding sources in the interpretation of response measured. According to literature data, digestive carbohydrate activity of bivalve species shows a great adaptability to variations in food availability, quantity, and quality. However, if food availability exerts a primary control on digestive activities, the latter could also vary in response to other factors, both external (e.g., temperature) and internal (e.g., reproductive state). These "regulation" factors exhibit seasonal variations, which, logically, should be reflected on digestive enzyme activities [52]. Nevertheless, if seasonal amylase and cellulase activities were well recorded in the digestive gland of zebra mussels, no seasonal activities were recorded in their crystalline style. The absence of seasonality in the crystalline-style activities could be related to the innate nature of the style, for example, a secretion organ whose weight does not vary according to the seasons. Inversely to digestive gland, the crystalline-style activity (downregulated) appeared to be more associated with the location and metal exposure level, allowing the discrimination of the most chemically impacted site from the others. Thus, the high discrimination potential of crystalline-style enzymes activities, in association with their low seasonal dependency, makes these parameters very promising biomarkers to develop in biomonitoring studies using *D. polymorpha*. In this way, zebra mussels were transplanted during 2 months along a metal and organic pollution gradient in spring 2008 (Seine River, France; PIREN-Seine program) [53]. Amylolytic and cellulolytic activities measured in the CS displayed similar patterns of response with lower activities at the downstream site compared to the upstream site (20–30% lower). By contrast, digestive enzyme activities in the digestive gland displayed only slight variations during the whole period of exposure, confirming the interest of the crystalline style as matrix in biomonitoring program.

#### **3.2. Reproductive process**

There is a large body of literature on ecophysiology and reproductive biology of dreissenid species [54]. The reproductive function is a key element for the preservation of species, implying a lot of responses which may be studied in reproductive ecotoxicology at different levels of biological organization, for example, from toxic effects on biological macromolecules to the consequences on the reproductive capacity of organisms and the maintenance of their population in the environment. Until today, literature in dreissenids mainly focused on disturbances of gametogenesis and on biomarkers related to the integrity and functionality of gametes. Numerous stressors of anthropogenic origin are susceptible to modify the reproductive physiology of mussels, either by direct or by indirect effects. In dreissenid species, existing data on the disturbances of the reproductive cycle by environmental factors, such as water temperature, food availability, or chemical contamination, highlighted early or delayed sexual maturation as well as asynchronous development between congeners in exposed mussels [55–58]. Many studies also revealed that gonads were the primary organs infected by trematodes of the family *Bucephalidae* in dreissenid mussels, known as the first intermediate host. The development of cercariae in sporocysts coincides with the host sexual maturation and causes in infected mussels an energetic depletion allowed to the reproductive tissue, leading to disturbances of their reproductive cycle as well as in the most severe cases to complete castration and host sterility [59, 60].

The gametogenesis and reproductive effort in dreissenid species has mainly been investigated by either nonhistological examinations, with the measure of the reproductive tissue biomass (gonad) compared to the rest of soft tissues (somatic) and the definition of a *gonado-somatic index*: GSI (the difficulty for discriminating the reproductive tissue and the digestive gland renders this methodology quite imprecise); either histological examination of gonads with the definition of various reproductive stages (*resting, development, prespawning, or postspawning stages*) and the possible measurements of morphological parameters by computer-assisted image analysis systems (e.g., oocyte number in follicles, oocyte diameter, etc.). Although informative, histological examination of the gonads is both time-consuming and experimenter-dependent. Recently, a novel strategy has been developed for a rapid and sensitive determination of an *index of sexual maturity* in male zebra mussels. This index considers the proportion of germ cells in male gonad according to their DNA content ((sub)-haploid, diploid, or tetraploid cells) measured by flow cytometry. This *index of sexual maturity* was used to describe the reproductive cycle in a control population of zebra mussels over a year. This biomarker also highlighted the reprotoxicity of carbamazepine, an anti-epileptic compound which was experimented in a mesocosm study in caged zebra mussels [61]. Initiation of gametogenesis, proliferation, and differentiation of immature germ cells determines both the quantity and quality of mature germ cells or gametes implied in fertilization and *de facto* in the reproductive success of zebra mussels [62]. Gametes represent "cells of choice" for the development of biomarkers in ecotoxicology, by relating the exposure of adult mussels with the long-term consequences in offsprings [63]. Until today, spermatozoa of dreissenid mussels have mainly been studied in comparison with oocytes for the development of biomarkers like sperm mortality (or membrane permeability), sperm motility (duration, velocity, etc.) with computer-assisted sperm analysis (CASA) system or fecundity [64]. Other authors investigated the integrity or the functionality of specific organites, like acrosomal integrity or mitochondrial membrane potential, and demonstrated the sensitivity of zebra mussel's spermatozoa to the toxicity of commercial formulations of pesticides (Bayluscide, Round-Up Ready-to Use Plus®) [65]. In their study, Seaver *et al*. [66] observed in spermatozoa of zebra mussel that *in vitro* irradiated with a physical agent (UV-B) decreased acrosomal integrity, incorporation of sperm into egg cytoplasm, and first zygotic cleavage. These defects in first zygotic cleavages, as also observed by these authors with the irradiation of eggs to UV-B, could be related to genomic damage in gametes. To our knowledge, the use of gametes (spermatozoa and oocyte) for evaluating the genotoxicity of water contaminants has not yet been tested in dreissenid mussels. Other authors also highlighted the sensitivity of early-life stages in zebra mussels to the toxicity of various water contaminants and proposed their use for the development of biomarkers [67].

#### **3.3. Hemocytes**

energy expenditure (general adaptation syndrome) [52]. However, this upregulation has not been recorded at all contaminated sites, lower levels of pollution or higher levels of food availability probably explaining the low need for organisms to "invest" in their digestive activities (least energy requirement). These results highlight adjustment of digestive enzymatic activities according to intake and energy requirements of the organism, factors which, if not taken into account, constitute potential confounding sources in the interpretation of response measured. According to literature data, digestive carbohydrate activity of bivalve species shows a great adaptability to variations in food availability, quantity, and quality. However, if food availability exerts a primary control on digestive activities, the latter could also vary in response to other factors, both external (e.g., temperature) and internal (e.g., reproductive state). These "regulation" factors exhibit seasonal variations, which, logically, should be reflected on digestive enzyme activities [52]. Nevertheless, if seasonal amylase and cellulase activities were well recorded in the digestive gland of zebra mussels, no seasonal activities were recorded in their crystalline style. The absence of seasonality in the crystalline-style activities could be related to the innate nature of the style, for example, a secretion organ whose weight does not vary according to the seasons. Inversely to digestive gland, the crystalline-style activity (downregulated) appeared to be more associated with the location and metal exposure level, allowing the discrimination of the most chemically impacted site from the others. Thus, the high discrimination potential of crystalline-style enzymes activities, in association with their low seasonal dependency, makes these parameters very promising biomarkers to develop in biomonitoring studies using *D. polymorpha*. In this way, zebra mussels were transplanted during 2 months along a metal and organic pollution gradient in spring 2008 (Seine River, France; PIREN-Seine program) [53]. Amylolytic and cellulolytic activities measured in the CS displayed similar patterns of response with lower activities at the downstream site compared to the upstream site (20–30% lower). By contrast, digestive enzyme activities in the digestive gland displayed only slight variations during the whole period of exposure, confirming the interest of the crystalline

There is a large body of literature on ecophysiology and reproductive biology of dreissenid species [54]. The reproductive function is a key element for the preservation of species, implying a lot of responses which may be studied in reproductive ecotoxicology at different levels of biological organization, for example, from toxic effects on biological macromolecules to the consequences on the reproductive capacity of organisms and the maintenance of their population in the environment. Until today, literature in dreissenids mainly focused on disturbances of gametogenesis and on biomarkers related to the integrity and functionality of gametes. Numerous stressors of anthropogenic origin are susceptible to modify the reproductive physiology of mussels, either by direct or by indirect effects. In dreissenid species, existing data on the disturbances of the reproductive cycle by environmental factors, such as water temperature, food availability, or chemical contamination, highlighted early or delayed sexual maturation as well as asynchronous development between congeners in exposed mussels [55–58]. Many studies also revealed that gonads were the primary organs infected by trematodes of the family *Bucephalidae* in dreissenid mussels, known as the first intermediate host. The development of cercariae in sporocysts coincides with the host sexual

style as matrix in biomonitoring program.

**3.2. Reproductive process**

50 Organismal and Molecular Malacology

To be relevant, biomarkers must gather several properties such as being flexible according to the various pressures exerted by the environment and associated with a specific mode of action which is measurable by significant and reproducible techniques. Moreover, the early analyses of these biomarkers have to predict the effects at higher levels of organization, that is, the population. In this context, the hemocytes responses show a particular interest since they are key cells involved in the major physiological functions in invertebrates. Indeed, the hemolymph is a remarkable mussel tissue. With an opened circulatory system, hemolymph circulates through the bodies allowing to bring the nutrients to the various organs. Drawn from the adductor muscle of the bivalves after having carried out a little breach on the shell, hemolymph sampling is not lethal for the individuals and represents a clear advantage for biomonitoring. Circulating cells, the hemocytes, are present in the cellular fraction of the hemolymph and have been recently characterized [68]. These cells, mainly granulocytes and hyalinocytes, provide various functions involved in the essential physiological activities such as the transport and the digestion of the nutrients, breathing, excretion, repairs as well of the tissues and of the shell and, of course, immune defenses [69]. Contrary to vertebrates, which have an *acquired immunity* (or specific), the immunity of the invertebrates is only composed of the *innate immunity* (or not specific) ensured by the hemolymphatic compartment (hemocytes and plasma). In a global context, hemocytes allow the measurement of various markers involved in the physiological processes usually studied in ecotoxicology. Current studies are based mainly on the evaluation of the biomarkers of genotoxicity (measurements of DNA damage in hemocytes [70]), of cellular parameters (oxidative activity, mortality [71]), and immune defense (capacity of phagocytosis [72]). However, all these responses suggest the implication of another component of the immune system and complementary to the cellular responses of the organism, the humoral-mediated responses in plasma [73]. Humoral responses are regrouped in different effectors such as the lysosomal enzymes, lectins, protease inhibitors, or antimicrobial peptides [74]. The phenoloxidase cascade [75] and the fast production of reactive oxygen species (ROS) also take part in organism homeostasis. Both systems interact to give an effective answer relative to the various types of aggressions which the organism undergoes.

A recent study by Juhel et al. [45] highlights the effect of cyanotoxins on the immune functions of *D. polymorpha*. The authors underline a reduction in the full number of hemocytes per milliliter of hemolymph as well as a modulation of the subpopulations of hemocytes, an inhibition of the phagocytosis capacity and an increase in the concentration of lysozyme after 21 days of *in vivo* exposition to various strains of cyanobacteria. However, this study raises the problematic of the complexity to define the real immune effects based only on cellular approaches. Currently, very few studies take into account the entire hemolymphatic compartment, for example, cells and plasma, in the analysis of the biomarkers. However, the evaluation of the interactions and the relations between these two types of responses would permit to understand the functioning of this complex tissue which maintains actively the homeostasis of organism (for review see [76]). Indeed, considering that a panel of enzymes and hydrolytic molecules can influence the cellular responses, it seems paramount to study the humoral compounds in parallel. Considering the central role of hemolymph in the physiology of the organisms, the comprehension of the different functions it provides in relation to other major biological functions is also necessary to ensure relevant conclusions in ecotoxicological studies. However, to date, very little information is available on the hemocyte implications on the physiological responses other than those of immune functions and defenses. This lack of knowledge is certainly due to difficulties in studying simultaneously both cellular and acellular compartment, because it necessarily implies a global and multi-markers approach which is difficult to perform with the techniques commonly used.
