Few studies have been performed on nontarget effects of Bt toxins, and none with modified Bt toxins extracted from plants, or together with Roundup residues, even in regulatory files. For natural Bt toxins, their mechanisms of action and insect resistance are not fully understood (Singh and Sivaprasad, 2009), and the metabolism of these proteins in mammals is unknown (Séralini et al., 2011; Chowdhury et al., 2003). They may even interact with extrinsic factors (Then, 2010). Billions of people and wildlife could be exposed to modified Bt toxins; therefore understanding their potential side effects is crucial.
On two biomarkers of cell death, Cry1Ab exposure led to respiration inhibition and plasma membrane alterations, by contrast to Cry1Ac. This could be consistent with the fact that the consumption of MON810 maize producing Cry1Ab (in the ppm range) induced signs of hepatorenal alterations in a subchronic feeding study on rats (Spiroux de Vendômois et al., 2009). It is known that both toxins differ significantly in their domain III structure (Karim and Dean, 2000), which is the only one to be involved at the same time in ion channel function, receptor binding and insertion into the membrane (Dean et al., 1996). This occurred at relatively high concentrations (100 ppm) in comparison to the concentrations produced in GM plants (1–20 ppm, Székács et al., 2010). The content can differ greatly according to the GM variety and environmental conditions (Then and Lorch, 2008). The exposure during consumption can appear low enough to avoid side effects, and whether this occurs in vivo remains to be checked. However, the bioaccumulation in tissues, or bioaccumulative or long-term effects, has to be taken into account since Bt residues were recently claimed to be measured in pregnant women's serum at around 0.2 ppb (Aris and Leblanc, 2011). In addition, high quantities of Bt crops can be consumed by mammals. The procedure for GMO market authorizations for crops such as MON810 (EFSA, 2009) does not require in vitro tests on human cells of Bt toxins, nor on its combinatorial effects, thus our results are raising new questions about the safety of these toxins and the Bt crops in general. Although in vitro studies suggest degradation in human gastric secretions, digestion is never a complete process and insecticide toxins cannot be fully degraded in vivo (Paul et al., 2010). This is known to be the case for Cry1Ab (Chowdhury et al., 2003). It must be underlined that the insecticidal proteins produced by the GM plants are in soluble forms, and thus already biochemically activated, unlike those produced by the microorganism Bt, secreted as inactive precursors or protoxins (Hilbeck and Schmidt, 2006). The importance of Bt toxin activation has been demonstrated in relation to in vitro membrane damages of human erythrocytes, by solubilized Bt toxins, but not by the intact form (Rani and Balaraman, 1996). Cellular response to Bt toxins does not elicit apoptosis; it induces necrotic effects via a plasma membrane disruption for Cry1Ab within only 24 h. This may be due to pore formation like in insect cells owing to binding to specific receptors or membrane lipid rafts (Then, 2010; Soberón et al., 2009).
We also demonstrated that Cry1Ab and Cry1Ac exposures slightly reduced caspase 3/7 activations induced by Roundup. This could be related at least in part to the properties of Roundup compounds, especially adjuvants. We observed previously, in our group, that serum delayed the cytotoxic effects induced by Roundup. This was probably due to serum binding proteins (Benachour et al., 2007). Here we can assume that physico-chemical properties of proteins may give them the ability to bind and form complexes with Roundup adjuvants that have tendencies to form vesicles, buffering their bioavailability to cells. Similarly, a nonsignificant tendency of reduction of the cytotoxic effects of Roundup was observed on mitochondrial respiration and membrane degradation when the toxins were added. The apoptosis induction appeared to be the most sensitive impact of combined effects. This does not exclude other intracellular targets such as endocrine disruption, since Roundup is antiandrogenic from 0.5 ppm, below toxic levels and close to human serum levels (0.1–0.2 ppm in Acquavella et al., 2004).
Here we documented that modified Bt toxins are not inert on human cells, but can exert toxicity, at least under certain in vitro conditions. In vivo implications should be now assessed. Our results raise new questions in the risk assessment of food and feed derived from genetically engineered plants.