Effects of Cd on the Ca metabolism of freshwater mussels

Danksagung / agradecimentos / acknowledgements

Ich möchte allen, die mich bei der Erstellung meiner Diplomarbeit unterstützt haben, danken, insbesondere:

Herrn Prof. Dr. Hartmut Frank für die Bereitstellung des Themas und die Möglichkeit eines Studienaufenthaltes in Portugal im Rahmen meiner Diplomarbeit.

Frau Dr. Silke Gerstmann für die allzeit gute Betreuung, vor Ort in Bayreuth, sowie aus der Ferne während meines Aufenthaltes in Portugal.

Elke Bär, Huong Ngo, Katja Poxleitner und allen anderen Mitarbeitern des Lehrstuhls für Umweltchemie und Ökotoxikologie für die gute und nette Zusammenarbeit.

Yvonne Hofmann und Barbara Scheitler für ihre Unterstützung bei den Analysearbeiten in Bayreuth.

Também queria exprimir o meu agradecimento a:

O Prof. Dr. Jorge Machado pela oportunidade de fazer as minhas experiências no seu laboratório e o seu interesse pelo meu trabalho.

O Dr. José Andrade, á Dra. Stella Carraça, á D. Maria das Dores, á Dra. Paula Lima e ao Sr. António Rocha pelo seu apoio durante o meu estágio.

O Eng. Pedro Reis pelas determinações de metais feitas no Porto.

A Dra. Denisa Mendonça pela sua ajuda com a avaliação estatística.

I would also like to thank Dr Toshio Matsuda for kindly putting a sample of SEA0400 at our disposal.

Index

1  Introduction  1

2  Material and methods  5

2.1  Animals  5

2.2  Materials  5

2.3  Cleaning of equipment for sample storage  6

2.4  Exposure experiments  7

2.5  Analysis of body fluids  11

2.5.1  Sampling  11

2.5.2  Sample analysis  13

2.6  Analysis of soft tissues  13

2.6.1  Sampling  13

2.6.2  Sample preparation  13

2.6.3  Sample analysis  14

2.7  Data preparation  14

3  Results  16

3.1  Element distribution  16

3.2  Treatment effects  17

3.2.1  Treatment with Cd  22

3.2.2  Treatment with vitamin D  22

3.2.3  Treatment with SEA0400  22

3.3  Correlation between Ca and Cd  22

4  Discussion  26

4.1  Quality control  26

4.2  Troubleshooting  26

4.2.1  Tank water pH and temperature  26

4.2.2  Feeding  27

4.2.3  Cadmium exposure concentration  28

4.2.4  Injection of test substances  28

4.2.5  Kidney tissue sampling  29

4.2.6  Tissue digestion  30

4.3  Ion fluxes in a model freshwater mussel  30

4.3.1  Treatment with Cd  32

4.3.2  Treatment with vitamin D  33

4.3.3  Treatment with SEA0400  34

4.4  Literature review  35

5  Conclusion  40

6  Appendix  42

6.1  Experimental protocols  42

6.1.1  Experiment a  42

6.1.2  Experiment b  46

6.1.3  Experiment c  48

6.1.4  Experiment d  50

6.2  Database  52

6.2.1  Experiment a  52

6.2.2  Experiment b  59

6.2.3  Experiment c  62

6.2.4  Experiment d  70

6.3  Calculation results  77

6.3.1  Cadmium in fluid samples  78

6.3.2  Cadmium in tissue samples  79

6.3.3  Cadmium in kidney samples  80

6.3.4  Calcium in fluid samples  81

6.3.5  Calcium in tissues samples  82

6.3.6  Calcium in kidney samples  83

6.3.7  Magnesium in fluid samples  84

6.3.8  Magnesium in tissue samples  85

6.3.9  Sodium in fluid samples  86

6.3.10  Sodium in tissue samples  87

6.3.11  Potassium in fluid samples  88

6.3.12  Potassium in tissue samples  89

6.3.13  Chlorine in fluid samples  90

6.3.14  Osmolarity of fluid samples  91

6.4  References  92

6.5  Index of tables  95

6.6  Index of figures  97

6.7  Index of abbreviations  100

6.8  Index of symbols  100

1 Introduction

Freshwater mussels live as filter-feeders (Watters, 1998a). They filter plankton, suspended and dissolved organic matter out the aqueous medium to use it as a food source (Russell-Pinto, 1998). For this they take up water from the exterior into their mantle cavity through their inhalant aperture. The inhalant water current is produced by cilia on the gills. Due to the ciliary movement the water streams by the ctenidia, which secrete mucus to bind the food particles in the water. The mucus is then transported with the water current to the mouth at the end of the mantle cavity (Moore, 2001). From there the water returns through the suprabranchial chamber, where it takes up the faeces and excretion products, just before leaving the mussel again through the exhalant aperture (Morton, 1983). As the water circulates through the whole mantle cavity, all body parts of the mussel are exposed to pollutants possibly carried by the water. These pollutants can be taken up by the mussel, and if not metabolised and excreted, they accumulate. In fact, the elimination rates of these organisms are very low (Fent, 1998). Pollutant concentrations in mussels reflect the pollution accumulated over a certain period, rather than momentary environmental concentrations (Gundacker, 1999). This property of mussels makes them useful for the biomonitoring of chronic pollution (Franco, 2002).

One pollutant freshwater mussels are exposed to is cadmium. This heavy metal is used in dry batteries, alloys and anticorrosive coatings. Cadmium compounds are also used as heat resistant colour pigments and stabilizers for PVC. It is released into the environment mainly from smelteries and waste incinerating plants. World-wide emissions are around 8000 t/a, 90% of which are of anthropogenic origin (Bliefert, 1995). In the atmosphere cadmium is transported bound to small aerosol particles. The atmospheric deposition of these particles onto rivers and lakes is only one way cadmium can enter freshwater systems. Run-off from fields fertilized with sewage sludge or other fertilizers containing cadmium, precipitation water from galvanised gutters as well as leachate from landfills are other possible sources (Bliefert, 1995).

Cadmium is a non-essential metal that is accumulated by animals and plants. Its biological half-life in humans can be of up to 30 years (Timbrell, 1993), where it accumulates especially in the kidney, leading to renal failure. Cadmium is also known to disturb the calcium metabolism of humans, causing osteoporosis and making the bones brittle, symptoms of the

1

Itai-Itai disease. This disease appeared during the 1940s in Japan, where people were eating rice contaminated with cadmium (Bliefert, 1995).

In a similar way, cadmium could interfere with the calcium metabolism of freshwater mussels. These animals need calcium primarily to form their shells, which are made of calcium carbonate. Besides calcium is also stored as extracellular concretion spherules (Dube, 1997), 75% of which are found in the gills (Pynnönen, 1987). These concretions are made of calcium phosphate, which is less soluble than calcium carbonate (Dube, 1997). Hence, the shell material dissolves first when body fluid pH goes down, releasing carbonate ions, which act as a pH-buffer (Greenaway, 1971). In fact the carbonate buffer represents the most important pH-buffer in mussels, since the haemolymph contains very low amounts of proteins, which have a buffering capacity as well. To conserve the calcium ions released, they are immobilized in the phosphate concretions (Dube, 1997). Under normal pH conditions the shell is then again regenerated using the calcium ions from the concretion spherules. These have the enzyme carbonic anhydrase on their surface (Istin, 1970), which catalyses the formation of carbonic acid out of water and carbon dioxide (Wilbur, 1983). The protons resulting out of the dissociation of the carbonic acid dissolve material from the phosphate concretions thereby releasing calcium, which becomes available again for shell regeneration (Istin, 1970). Tracer experiments with radio-calcium revealed a half-life of calcium in the phosphate concretions of Velesunio angasi of 106 days (Brown, 1996). The calcium from the concretion spherules is also used to form the glochidial shells (Watters, 1998b). To maintain all these functions, freshwater mussels have to take up calcium from their environment. Even though they live in a medium with low calcium concentrations, it has been estimated that freshwater molluscs take up 80% of the calcium they need from the water (van der Borght, 1966); freshwater Ca concentrations are below 3 mmol/L (Bliefert, 1995). The remaining 20% are taken up with the food. There are three sites in their body, where freshwater mussels take up calcium, namely the gills, the mantle and the intestine, the mantle, however, not being the main uptake route (Coimbra, 1993). The mechanism, however, by which calcium is taken up is still unclear. Due to the different Ca concentrations in the external medium and the body fluids, the latter being around 6 mmol/L in Anodonta cygnea (Coimbra, 1988), it must be an energy-driven mechanism. Little is also known about the transport of calcium within the body. The transfer of calcium from one compartment to another is linked to the transport across the epithelia that border the different organs. All these uptake and transport mechanisms can potentially be disturbed by cadmium, that has been reported to compete with calcium for the transport sites of Ca-ATPases (Long, 1997) and to block calcium channels

2

(Kochegarov, 2003). Likewise, the calcium channel blockers verapamil and lanthanum have been shown to inhibit the Cd uptake in mussels (Sidoumou, 1997; Qiu, 2005), showing the similarity of both ions’ uptake pathways. The Antagonism between calcium and cadmium was also observed by Mungkung (2001), who found that the cadmium toxicity and accumulation in freshwater fish decreased with increasing water hardness, which is primarily controlled by calcium. He traced this finding back to the competition between calcium and cadmium for the binding sites of the same uptake systems. It was found green mussels collected from low salinity sites accumulated six time more cadmium than individuals from high salinity sites. This observation was also attributed to the competition of cadmium with major ions like Ca ²⁺ for the uptake through calcium channels (Blackmore, 2002). These findings have been confirmed recently as well by Qiu (2005), who found the cadmium uptake of Asiatic clams to be inhibited by high calcium concentrations. He describes this competition between calcium and cadmium using a Michaelis-Menten uptake inhibition model. Brown (1996) suggested that ¹⁰⁹ Cd, among other radio-nuclides, is absorbed by freshwater bivalves in a similar manner to Ca with subsequent deposition into the concretion spherules. Another group of membrane proteins that transport calcium are the sodium-calcium exchangers (NCX). Under normal physiological conditions the sodium concentration outside the cell is higher than inside of it. Driven by this concentration gradient across the cell membrane the NCX in the plasma membrane pump calcium to the outside of the cell in exchange for sodium, which is transferred to the inside of the cell (Ca efflux mode). This is, for instance, the case at the basolateral membrane of intestinal cells, where NCX extrude calcium to the extracellular medium (Larsson, 2002). Under special conditions in which sodium accumulates inside the cell, the NCX work inversely (Ca influx mode), pumping calcium from the outside to the inside of the cell (Iwamoto, 2003). NCX are inhibited selectively by the novel compound 2-[4-[(2,5-difluorophenyl)methoxy]phenoxy]-5ethoxyaniline (Fig. 1.1) also known as SEA0400 (Matsuda, 2001). The inhibitory potency of SEA0400 is strongly dependent on the intracellular sodium concentration. Only when high sodium concentrations are present within the cell and the NCX work in the Ca influx mode, SEA0400 develops its inhibitory potential (Lee, 2004). SEA0400 scarcely inhibits other receptors, channels and transporters at the concentrations it inhibits NCX (Iwamoto, 2003). The IC 50 values of SEA0400 for the sodium-dependent radio-calcium uptake in cultured rat neurons, astrocytes, and microglia are between 5 and 33 nmol/L (Matsuda, 2001); the value obtained for ventricular cardiomyocytes of guinea-pigs is between 30 and 40 nmol/L (Tanaka, 2002).

3

A third compound known to have an effect on the uptake of calcium is vitamin D 3 . In the human liver, vitamin D 3 is converted to 25-hydroxyvitamin D 3 , which is transported to the kidney, where it is again converted to 1,25-dihydroxy-vitamin D 3 (1,25(OH)2D3). 1,25(OH)2D3 can then be transported to different tissues. In the intestinal epithelial cells, for instance, 1,25(OH)2D3 promotes the expression of genes encoding calcium binding proteins (calbindin). Calbindin binds calcium in the microvillae region and facilitates its transport to the basolateral membrane, where calcium is finally extruded from the cell. The decrease of the calcium concentration near the apical membrane, however, favors the movement of calcium into the cell and thereby the transport across the epithelium (McCarthy, 1999). It is conceivable that vitamin D 3 regulates the intestinal absorption of calcium in mussels in a similar way. This mechanism of transcellular calcium transport by calbindin is relatively slow and therefore relevant only for long-term calcium regulation. For the short-term regulation the vesicular transport and the tunnelling through the endoplasmic reticulum are more important (Larsson, 2002).

To achieve a better understanding of how cadmium interferes with the calcium metabolism of freshwater mussels, four cadmium exposure experiments were performed with the freshwater clam A. cygnea, which belongs to the family of the Unionidae (Kaestner, 1993). Separate experimental groups were treated with vitamin D 3 and SEA0400 to compare the effects cadmium has with those of other substances influencing the calcium metabolism. The tissue and body fluid samples taken from the animals were not only analysed for cadmium and calcium, but also for other ions of physiological importance, that might elucidate the calcium transport processes further. These additional ions were magnesium, sodium, potassium and chlorine. The osmolarity of the body fluid samples was also measured.

4

2 Material and methods

2.1 Animals

Freshwater clams (A. cygnea) were collected from the Lagoon of Mira near Aveiro in northern Portugal. An analysis of the lagoon water showed that the Cd concentration is below 1 µg/L, and consequently normal for unpolluted freshwater systems (AMAP, 2002).

Table 2.1: Ionic composition of the lagoon water.

As long as they were not needed for experiments, the mussels were kept in darkness in aerated tanks filled with sediment from the lagoon and dechlorinated tap water. Under these conditions the animals survive as well as in water collected from their natural environment (Coimbra, 1988). Animals were considered healthy if they closed their valves when lightly touched. Dead animals found occasionally were immediately removed.

2.2 Materials

The following list gives an overview of the used materials; all chemical compounds used were per analysis grade.

glass tanks (50x25x35cm ³ ) used as experimental tanks and waiting tank

^- plastictanks (45x20x30cm ³ ) used as feeding tanks

^- bleachinglye from Moderna (sodium hypochlorite solution containing 4.5% active Cl) ^- CaCl 2 *2H 2 O ^- MgCl 2 *6H 2 O ^- NaCl ^- NaHCO ^{3 -} K 2 SO ^{4 -} CdCl 2 *H 2 O -

5

VIGANTOL ^® (vitamin D 3 preparation) from MERCK

^- SEA0400from Taisho Pharmaceutical Co. Ltd. ^- single-usesyringes from B. Braun Melsungen AG, 2mL ^- single-useneedles from B. Braun Melsungen AG, 0,8 x 40 mm and 1,20 x 40 mm ^- EppendorfMicrocentrifuge tubes, 1,5mL ^- conc.HNO ^{3 -} centrifugetubes from ORANGE Scientific, 14mL ^- centrifugefrom SIGMA 3K12 ^- PCLMdigital chloride meter from Jenway Ltd. ^- DigitalMicro-osmometer type 13 from ROEBLING ^- AtomicAbsorption Spectrometer SpectrAA 220-FS from Varian ^- Mark7 N 2 O/acetylene burner from Varian ^- GTA-110Graphite Tube Atomizer ^- PEbottles from VWR International, 50mL ^- distilledHNO ^{3 -} HCl(32%) ^- MixerMill type MM2 from Retsch ^- PTFEThread Seal Tape BF-12/10 from VWR International ^- HighPressure Asher HPA-S from Anton Paar Physica ^- single-usesyringes from B. Braun Melsungen AG, 20mL ^- whiterim filters from Schleicher & Schuell MicroScience GmbH, 0,45µm ^- AtomicEmission Spectrometer Vista-Pro CCD-Simultaneous ICP-OES from Varian ^- Mass-Spectrometer7500 c from Agilent ^- certifiedreference material (mussel tissue) ^- fromEuropean Reference Materials CE278, Sample No: 0806

2.3 Cleaning of equipment for sample storage

To make sure the centrifuge tubes and the PE bottles used for sample storage were Cd free,

they were acid washed according to the procedure described by Matter, L. (1997). Around 1

or 2 mL of conc. nitric acid were put into each container and thereafter the same amount of

distilled or deionized water was added. After this, the capped containers were shaken

repeatedly and left standing for at least half an hour. Thereafter the acid solution was disposed

and the container rinsed with distilled or deionized water.

6

2.4 Exposure experiments

Four similar exposure experiments were performed, these shall be refered to as experiment a, b, c and d. Before use the tanks for each experiment were washed in five steps. First the tanks were left standing overnight filled to the brim with tap water with a 100mL of bleaching lye in each. On the following day the tanks were left standing for at least five hours filled with tap water only. Next the tanks were rinsed first with tap water and then with distilled water. Finally they were wiped out with ethanol and cellulose paper. The so cleaned tanks were equipped with air pumps. To better diffuse the outpouring air without creating too large surfaces Cd might adsorb to, pierced plastic pipettes were put at the end of the air tubes. Metallic items were not present in the tank to prevent the leaching of metals from the equipment into the water. To prepare the water for the experimental tanks three different stock solutions were prepared (Table 2.2).

Table 2.2: Stock solution composition.

The experimental tanks were filled with distilled water, to which defined amounts of each stock solution were added giving a total of 10L of tank water in each tank. When no distilled water was available (Table 6.2), they were filled with bottled mineral water and adequate amounts of stock solution to approximate the actual tank water concentrations to the target concentrations (Table 6.3). The ionic composition of the tank water was varied between the experiments to adjust the pH (Fig. 6.2, Fig. 6.4, Fig. 6.6, Fig. 6.8) and the osmolarity (Table 2.3).

Table 2.3: Ionic composition of the tank water.

In each experimental tank there were put four freshwater clams (A. cygnea), that had been brushed with a swab under tap water and then rinsed with distilled water.

7

After at least one day of acclimatisation (Table 6.1, Table 6.4, Table 6.5, Table 6.6) the mussels were exposed to the test substances. Up to four different treatments were compared in each experiment (Table 2.4).

One experimental tank (tank 0) remained the control tank, to which no compound was added. To another experimental tank (tank 1) 1000µL of a 10mM CdCl 2 test solution were added to set the Cd concentration at 1 µmol/L. The mussels in the other two experimental tanks (tanks 2 and 3) were administered vitamin D and SEA0400 respectively, both by injection into the posterior end of the foot. The way the solutions for injection were prepared, the volume of test solution and the amount of test substance injected per mussel varied between the experiments (Table 2.5).

Table 2.5: Injection of vitamin D and SEA0400.

The blue mussel, Mytilus edulis, contains 95 IU of vitamin D per 180 g of ww (Blanchet, 2000), which is in the range of the soft body weights of the experimental animals used (Table 6.7, Table 6.15, Table 6.27). To determine the amount of vitamin D to be injected per mussel (Table 2.5), it was assumed A. cygnea contains a similar amount of vitamin D and that this amount would have to be at least doubled to have a noticeable effect on the Ca-uptake. In experiments on the effects of SEA0400 on cerebral ischemic injury in Sprague-Dawley rats 3 mg per kg body weight were injected (Matsuda, 2001). Based on the same ratio, the amount of SEA0400 to be injected per mussel (Table 2.5) was calculated using the mean soft body weight of A. cygnea (142 g) obtained from the data collected in experiment a (Table 6.7).

8

Because six out of twelve mussels used in the experiments a and b that were injected a test substance dissolved in oil died (Table 6.1, Table 6.2, Table 6.4), distilled water was used to inject vitamin D and SEA0400 in the remaining experiments c and d. To reduce the loss of test substance in these experiments the mixtures for injection were prepared already in the syringes used for the injections. From experiment b onwards, right after the test substance was injected, 0,5 mL of distilled water were injected using the same syringe. In that way the remainder of the test substance in the syringe was to be eluted and the absorbtion of the test substance in the organism to be improved.

To make sure the Cd concentration at inner epithelia (like the pericardium or the ventricle epithelia) not directly in contact with the external medium reached an effective level and was similar to the level at the intestinal epithelium, which is more in contact with the external medium, in the experiments c and d the mussels in tank 1 were injected 500 µL of a 1 µM CdCl 2 solution into the posterior end of the foot. Right after the Cd solution was injected, 0,5 mL of distilled water were injected using the same syringe.

To take the effect into account the injections themselves might have, in the experiments c and d the control mussels were each injected 1mL of water into the posterior end of the foot. Nearly every day (Table 6.1, Table 6.2, Table 6.4, Table 6.5, Table 6.6) the mussels were fed with an algal suspension (Table 2.6). In experiment a the mussels were put for feeding into separate feeding tanks, to which the algal suspension was added; there was one feeding tank for every experimental tank. The feeding tanks were filled with Cd free water from the experimental tanks. In experiments b, c and d the algal suspension was added to the experimental tanks.

Table 2.6: Feeding of mussels.

Two hours after the algae were added to the water the mussels were put into algae free water. In experiment a the mussels were transferred from the feeding tanks back to the respective

9

experimental tanks, which had been filled with new tank water. In experiments b, c and d the mussels were put into a single waiting tank for the time the water in the experimental tanks was being renewed. Depending on the experiment the waiting tank was filled with different types of water (Table 2.7). Once the water of the experimental tanks had been renewed the mussels were transferred from the waiting tank back to their respective experimental tanks.

Table 2.7: Water in the waiting tank.

Before the emptied experimental tanks were refilled with tank water they were washed in three steps. First the tanks were rinsed with tap water; the biofilms that had grown in the tanks were wiped away with a swab. Then they were rinsed with distilled water and finally wiped out with ethanol and cellulose paper. The so cleaned tanks were equipped again with air pumps and filled with tank water of the original composition (Table 2.3). Before the mussels were put back into their respective experimental tanks, they were first brushed with a swab under tap water and then rinsed with distilled water.

Once there was Cd in tank 1 the water and the animals in this tank were handled with special care not to contaminate the other tanks. For this a separate set of swab, tubes, gloves, etc. was used. Before transferring the mussels from tank 1 to another tank they were rinsed first with tap water and then with distilled water.

For experiments b, c and d a shorter exposure period was chosen (Table 2.8) to make sure the organic test substances, vitamin D and SEA0400, were not completely metabolised before the end of the experiment and their effect becomes more visible.

After the exposure period all mussels were sacrificed for body fluid and tissue sampling. The day before the samples were taken, the living mussels were weighed. This was done right before the mussels were put back into their respective experimental tanks after feeding.

10

2.5 Analysis of body fluids

Before sampling each mussel was was brushed with a swab under tap water and then rinsed with distilled water to remove any ions from the tank water that might have been attached to the shell. Fig. 2.1 gives an overview of the sites samples have been taken from.

Fig. 2.1: Cross section of A. cygnea.

2.5.1 Sampling

Once it had been cleaned the mussel was opened a few millimetres with a knife and a stopper was put between its valves to prevent the mussel from closing them again. First haemolymph was extracted from the interepithelial space of the mantle. Depending on the animal size between 1 and 5 mL of haemolymph were extracted from the interepithelial space using a 2mL syringe with a 1,20 x 40 mm needle.

11

Next the adductors were cut through to open the mussel further and the adnate inner gills were separated. In order to be able to lift the posterior end of the visceral sac from the kidney the pedal retractor muscle was cut through. Then the visceral sac was cut off up to the point where the whole kidney was uncovered. Subsequently the ventral kidney epithelium was carefully removed and the dark kidney tissue was extracted with a pair of tweezers. The tissue sample was put directly into a 1,5 mL Eppendorf Microcentrifuge tube; in some experiments this tube contained a liquid medium to facilitate the tissue being taken off the tweezers (Table 2.9). As this was not indispensable, in the last experiment the tissue sample was put into an empty tube as an attempt to improve the later lyophilisation step. Two tubes containing the same as the ones used for sampling served as blank samples.

Table 2.9: Kidney tissue sampling.

Before and after sampling, each tube was weighed to determine the wet weight of the sample. In experiments b, c and d the kidney tissue samples (as well as the respective blank samples) were freeze-dried to determine their dry weight. Afterwards 500 µL of conc. HNO 3 were added to each tube to digest the tissue.

Once the kidney tissue had been removed, a fluid sample from the pericardium was taken. For this the pericardium was penetrated at one side, where the nacre from the shell below could be seen, with the 0,8 x 40 mm needle of a 2mL syringe.

After the pericardium had been exhausted, haemolymph was sampled from the heart. For this the ventricle was penetrated, again using a 2 mL syringe with a 0,8 x 40 mm needle. To make sure the sample was taken from the ventricle, the heart was penetrated as close as possible to the readily identifiable intestine. Haemolymph was drawn until the heart was shrivelled around the intestine.

All the fluid samples were put into acid washed 14 mL centrifuge tubes, that were kept on ice. After sampling they were centrifuged at 4500 rpm for 7 minutes. The supernatant of each sample was finally transferred into a second acid washed tube. Until analysis the samples were kept in the freezer at -18 ºC ; the kidney tissue samples in acid were stored at room temperature.

12

2.5.2  Sample analysis  

The Cl content of all fluid samples was determined using a PCLM digital chloride meter.

  Between three and seven replicate measurements (depending on the variability of the  

  measurements) were made of each sample and averaged out. Every twenty measurements a  

  chloride standard solution (100 mmol/L) was analysed. If a value differing from the known  

  standard concentration by more than 5 was obtained, the silver electrodes were cleaned and  

  the buffer solution was renewed. The same procedure was followed as well regularly every  

  sixty measurements  

  The osmolarity of these samples was determined using an osmometer. This work was done by  

  the clinical chemistry laboratory of the Santo Antonio hospital in Porto  

  Na, ,K Ca, Mg and Cd analyses of the fluid samples as well as Ca and Cd analyses of the  

  kidney tissue samples were performed by the CIIMAR analytical laboratory in Porto. At the  

  analytical laboratory the fluid samples were acidified with HNO 3 to pH 2. Both, fluid and  

  kidney tissue samples, were stored in the fridge at 4 C. Na and K analyses were determined  

  by AES, Ca and Mg by flame AAS and Cd by graphite furnace AAS. For each calibration  

  curve three or more standard solutions were analysed, the correlation coefficient of every  

  curve being above 0 ,99. The fluid samples were diluted as necessary with 0 ,6 N HNO 3 and  

  the kidney tissue samples with ultrapure water  

2.6  Analysis of soft tissues  

2.6.1  Sampling  

  After the fluid samples had been taken, the part of the intestine running through the heart was  

  excised. Next samples from the mantle and from the gills were taken. These tissue samples  

  were taken only from the center of the organ, avoiding tissue that was close to other tissues  

  like adductor mussel tissue or foot tissue. The border of the mantle including the pallial  

  muscles was sampled separately. The metal loss from tissues through the loss of body fluids is  

  negligible (Inza, 1997 ) All tissue samples were put into aluminium cups, that had been  

  weighed before. To determine the wet weight of the tissues the cups were weighed again right  

  after sampling. Finally the tissues were frozen at 18 C  

  The length, the height and the weight of each single valve were also measured  

2.6.2  Sample preparation  

  The frozen soft tissues were freeze-dried for 24 h at 0 ,2 atm and 45 ºC and then weighed  

  again to determine the dry weight. For transport the freeze-dried samples were stored in acid  

13

washed PE bottles with screw caps. The samples of which more than 250 mg dw had been collected were ground with a Mixer Mill for approx. 8 min at 80 swings per minute to homogenize the tissue. Around 250 mg of each sample were weighed out into a quartz glass vessel; if less sample material was available the whole tissue sample was put into the vessel. Next 3 ml distilled HNO 3 and 1 ml HCl (32%) were added to each vessel. The vessels were then sealed with PTFE tape and covered with a glass lid, that was additionally tied to the vessel with PTFE tape. The capped vessels were then placed in a High Pressure Asher, which was programmed according to the following table:

Table 2.10: Temperature programme of the High Pressure Asher.

Once the vessels had cooled down to 40 °C the sample solutions were transferred into 25 mL PE volumetric flasks and filled with bidest water up to the gage mark. Using 20 mL singleuse syringes and 0,45 µm white rim filters the samples were filtered into acid washed PE bottles; the first drops filtered served to condition the filter and were disposed. Every 20 samples 4 mL of acid (3 ml HNO 3 and 1 ml HCl) without sample material were heated, diluted and filtered like the normal samples to serve as a blank. To validate the method every 40 samples 250 mg of certified reference material of a known Cd concentration were prepared for analysis like the normal samples.

2.6.3 Sample analysis

Na, K, Ca, Mg and Cd contents of all normal and blank samples were analysed using ICP-AES. If the Cd concentration was lower than 100 µg/L it was measured with ICP-MS. The samples from reference material were just analysed for Cd as this was the only element the material was certified for. To further dilute the samples a 3:1 distilled HNO 3 : HCl (32%) solution was used. This work was done by the BayCEER analytical laboratory in Bayreuth.

2.7 Data preparation

Outliers were determined using the Grubbs test with P < 0,01 (Faes, 2005). The subsequent calculations were performed without consideration of these outliers. To take the different body sizes of the test animals into account, the remaining metal concentrations were standardised by the following formula (Gundacker, 1999):

14

Fig. 2.2: Standardisation formula

c std = standardised concentration

= measured concentration c

(of one element and from mussels from the same experiment and treatment group) = gradient of straight line from linear regression between c and u a

(of one element and from mussels from the same experiment and treatment group) = mean of shell lengths Ì

(from mussels from the same experiment and treatment group) = individual shell length u

(in case the samples from two mussels were pooled the mean of the two individual shell lengths was used)

With these standardised concentrations (chapter 6.2) the mean values and standard deviations of each element and from mussels from the same experiment and treatment group were calculated (chapter 6.3).

The statistical significance of the difference between the means from treatments with one of the test substances and the mean from the control treatment was tested with the t-test (Trochim, 2002).

15