Accumulation and toxicity of cyanobacterial toxins, particularly microcystin-LR (MCLR) have been extensively studied in fish and aquatic invertebrates. However, MCLR excretion mechanisms, which could reduce this toxins effects, have received little attention. The Patagonian silverside, Odontesthes hatcheri, is an omnivorous-planktivorous edible fish, which has been shown to digest cyanobacterial cells absorbing MCLR and eliminating the toxin within 48h without suffering significant toxic effects. We studied the effects of MCLR on glycoconjugate composition and the possible role of multidrug resistance associated proteins (Abcc) in MCLR export from the cells in O. hatcheri intestine. We treated O. hatcheri with 5μg MCLRg1 body mass administered with the food. Twenty four hours later, the intestines of treated and control fish were processed for lectin-histochemistry using concanavalin A (ConA), Triticum vulgaris agglutinin (WGA), and Dolichos biflorus agglutinin (DBA). MCLR affected the distribution of glycoconjugates by augmenting the proportion of ConA-positive at the expense of WGA-positive cells. We studied MCLR effects on the transport of the Abcc-like substrates 2,4-dinitrophenyl-S-glutathione (DNP-SG) and calcein in ex vivo intestine preparations (everted and no-everted sacs and strips). In treated preparations, CDNB together with MCLR (113μg MCLRg1 intestine, equivalent to 1.14μmolL1 when applied in the bath) or the Abcc inhibitor, MK571 was applied for one hour, during which DNP-SG was measured in the bath every 10min in order to calculate mass-specific DNP-SG transport rate. MCLR significantly inhibited DNP-SG transport (p<0.05), especially in middle intestine (47 and 24%, for luminal and serosal transport, respectively). In middle intestine strips, MCLR and MK571inhibited DNP-SG transport in a concentration dependent fashion (IC50 3.3 and 0.6μmolL1, respectively). In middle intestine strips incubated with calcein-AM (0.25μmolL1), calcein efflux was inhibited by MCLR (2.3μmolL1) and MK571 (3μmolL1) by 38 and 27%, respectively (p<0.05). Finally, middle intestine segments were incubated with different concentrations of MCLR applied alone or together with 3μM MK571. After one hour, protein phosphatase 1 (PP1) activity, the main target of MCLR, was measured. 2.5μM MCLR did not produce any significant effect, while the same amount plus MK571 inhibited PP1 activity (p<0.05). This effect was similar to that of 5μM MCLR. Our results suggest that in O. hatcheri enterocytes MCLR is conjugated with GSH via GST and then exported to the intestinal lumen through Abcc-like transporters. This mechanism would protect the cell from MCLR toxicity, limiting toxin transport into the blood, which is probably mediated by basolateral Abccs. From an ecotoxicological point of view, elimination of MCLR through this mechanism would reduce the amount of toxin available for trophic transference.