Résumé : PCBs are among the most problematic marine contaminants. Converging towards the oceans via the rivers and the atmosphere, they concentrate in sediments where they become a permanent threat to organisms living at their contact. PCBs are extremely resistant, bioaccumulated and some congeners are considered as highly toxic. The North Sea is considered as a highly contaminated area ; however little information is available regarding the impact of PCBs on key benthic organisms of this region.

Ubiquist, abundant and generally recognized as a good bioindicator species, the common NE Atlantic sea star Asterias rubens (L.) is an ecosystem-structuring species in the North Sea and was chosen as an experimental model. The present study focused on the characterization of PCB bioaccumulation in A. rubens exposed through different routes (seawater, food, sediments) and on subsequent biological responses, at immune and sucellular levels. The considered responses were respectively (i) the production of reactive oxyggen species (ROS) by sea stars amoebocytes, which constitutes the main line of defence of echinoderms against pathogenic challenges and (ii) the induction of a cytochrome P450 immunopositive protein (CYP1A IPP) which, in vertebrates, is involved in PCB detoxification.

Experimental exposures carried out have shown that A. rubens efficiently accumulates PCBs. Exposure concentrations were always adjusted to match those encountered in the field. PCB concentrations reached in sea stars during the experiments matched the values reported in field studies ; therefore our experimental protocol was found to accurately simulate actual field situations. Uptake kinetics were related to the planar conformation of the considered congeners : non-coplanar PCB uptake was described using saturation models, whereas coplanar PCBs (c-PCBs) were bioaccumulated according to bell-shaped kinetics. Non-coplanar congeners generally reached saturation concentrations whithin a few days or a few weeks, which means that sea stars can be used to pinpoint PCB contamination shortly after occurrence. On the other hand, c-PCB concentrations reached a peak followed by a sudden drop, indicating the probable occurrence of c-PCB-targeted metabolization processes in sea stars. Our experimental studies also demonstrated that seawater was by far the most efficient route for PCB uptake in sea stars and that even if PCB levels in seawater are extremely low compared to sediment-associated concentrations, seawater constitutes a non-negligible route for PCB uptake in marine invertebrates. Among the different body compartments, bodywall displayed the highest bioaccumulative potency and can therefore be considered as particularly interesting for field biomonitoring applications. Rectal caeca, which play a central role in digestion and excretion processes in sea stars, have also rised particular interest as results suggest these organs could be involved in the elimination of PCB 77 degradation products.

The field work carried out during the present study showed that PCB concentrations measured in A. rubens tissues reflect environmental levels of certain congeners. As it was the case in experimental conditions, A. rubens differentially accumulated PCB congeners according to their planarity. Strong relationships were found between concentrations measured in sediments and those determined in sea stars body wall for certain non-coplanar congeners (e.g. 118 and 138), thus allowing to consider A. rubens as a suitable bioindicator species for medium-chlorinated PCB congeners. On the other hand, sea stars appeared to be able to regulate -to a certain extent- their content in coplanar PCBs. This implies that (i) A. rubens cannot be strictly considered as an indicator organism for c-PCBs and (ii) c-PCBs probably affect essential aspects of sea star biology, potentially leading to deleterious effects.

The present study addressed effects of PCB exposure on A. rubens biology, in both experimental and field conditions. In experimental conditions, PCBs were found to significantly alter ROS production by sea stars amoebocytes. This alteration also occurred in a congener-specific way : c-PCBs were found to significantly affect, and probably impair sea stars immune system, whereas non-coplanar congeners had no effect. In the field, the PCB contribution to immunotoxicity could not be determined because none of our studies considered ROS production along with c-PCB concentration measurements. However, the levels of ROS production by sea stars amoebocytes measured in field and experimental conditions were found to potentially lead to altered immunity, and therefore to impair sea stars defence against pathogenic agents.

A specially designed ELISA was used to measure CYP1A IPP in experimental and field conditions. Experimental work has shown that the induction of this protein was related to PCB exposure in a congener-specific fashion : c-PCBs alone were found to strongly induce the production of CYP1A IPP according to a dose-dependent relationship. These results have highlighted many similarities between the dioxin-like responsiveness of CYP1A IPP induction in sea stars and that occurring in vertebrates. This strongly suggests similarities in the toxicity-triggering mechanism of dioxins and c-PCBs. In the field, CYP1A IPP induction was found to be significantly related to PCB levels determined in bottom sediments. It can thus be considered as a valuable biomarker. Further research is however needed to better characterize the influence of physico-chemical and physiological parameters on CYP1A induction to refine the interpretation of the information gathered via this biomarker.

Results obtained in our study have lead to questionning international regulations applying to PCB biomonitoring in the marine environment. For instance, we strongly suggest that the selection of congeners to be systematically considered should be revised to include c-PCBs. Indeed, in our experiments PCB toxicity was almost always attributable to the sole c-congeners. Historically, determination of c-PCB concentrations was extremely difficult due to analytical limitations ; however, nowadays, these problems have been overcome and do no more justify their exclusion from monitoring studies.

Although A. rubens appeared to be quite resistant to PCB contamination, levels measured in sea stars from the southern North Sea can possibly affect their immune and endocrine systems in a subtle way, but with relatively low risk for this species at the short-term. However, this does not mean that other species in this region undergo similarly low risks, or that sea star-structured ecosystems may not become affected in the long-term