Genotoxicity biomarkers in Mytilus galloprovincialis: wild versus caged mussels

A biomonitoring programme of wild and caged mussels (Mytilus galloprovincialis) was carried out at four selected sites along the Ligurian coast: Cornigliano, Voltri, Zinola, and Sanremo (Italy). Mussels of a very narrow size range were left in situ for 30 days. Adult specimen of mussels from natural substrates were collected in the same areas. Animals from a mussel farm located in La Spezia were used as controls. Micronucleus frequency and DNA single strand breaks, evaluated by alkaline elution, were used as biomarkers of genotoxicity. Mussels were also analyzed for polycyclic aromatic hydrocarbons (PAH) and heavy metals (Hg and Cd). Different gradients of PAH and metal concentrations were detected in tissues of mussels from different samplings sites. A weak correlation was found between single strand breaks and PAH content while MN frequency correlated with Hg concentration (r = 0.28, P < 0.002). A clear distinction between the sites, allowing classification along a pollution gradient (Sanremo < Zinola < Voltri < Cornigliano) was demonstrated by the analysis of genotoxicity parameters. The obtained results suggested that the micronucleus assay compared with DNA damage determination by alkaline elution allow to better discriminate the selected sites. DNA damage expressed as constant of elution (k ml(-1) x 10(3)) ranges from 30 +/- 9.6 to 89.60 +/- 40.10, and micronuclei frequency from 1.78 +/- 1.04 to 24.4 +/- 12.9, in control animals and in mussels from the most polluted site, respectively. Wild mussels accumulated significant concentrations of chemicals and showed a higher induction of chromosomal damage than caged mussels, expressed as micronuclei frequency. Caged mussels showed higher level of DNA damage than wild mussels, probably as a result of recent exposure. DNA damage was higher in September than in May, as opposed to micronuclei frequency being higher in May than in September. Endogenous and exogenous factors, such as change of pollutant input levels or compositions, could be considered the cause of such variability.