interact with the c-secretase cleavage site of APP at the plasma membrane. On the other hand, co-immunoprecipitation of calreticulin with immature APP from a synaptosomal fraction indicates an interaction of APP with calreticulin already in the ER during maturation and trafficking of APP. As seen in our study, an association of calreticulin with immature as well as the mature APP has also been shown by co-immunoprecipitation of calreticulin from APP overexpressing cell lines. In this 
previous study, it was shown that APP and calreticulin interact and that this interaction was detectable at pH 7.5, decreased at pH 6.5 and was not detectable at pH 5.5. These results suggest that the interaction between APP and calreticulin takes place under pH conditions prevailing in the ER and early cis-Golgi compartment or at the plasma membrane, but not under the acidic pH condition prevailing in the trans-Golgi or endosomal compartments. Since the generation of the APP intracellular domain from 17149874 a GFP-tagged C99 occurs exclusively at the plasma membrane, it is very likely that cleavage of APP by c-secretase occurs at the cell surface and not in the ER, Golgi or endosomal compartments. This notion is supported by our finding that Sodium laureth sulfate supplier extracellular application of the C- or P-domain of calreticulin or the calreticulin peptide interferes with the generation of Ab40 and Ab42. In addition, calreticulin seems to be tightly associated with or embedded in the plasma membrane, since it remains associated with synaptosomal membranes after treatment with alkaline pH and EDTA and is released only after detergent treatment. Moreover, immunostaining of calreticulin at the cell surface of live hippocampal neurons with antibodies against its N- or C-terminus indicate that calreticulin is accessible from the outside of the cell and is, thus, present in or at the plasma membrane. Furthermore, extracellular application of the recombinant GST/C- and GST/P-domains to APP overexpressing cells affects APP processing, indicating that they compete with the domains of endogenous calreticulin at the plasma membrane and impair the function of endogenous calreticulin in a dominant-negative manner. This dominantnegative impact on APP proteolysis is even more pronounced when the calreticulin peptide was applied to APP overexpressing cells, implying that the small 14mer is most effective in blocking the function of endogenous calreticulin at the plasma membrane. In contrast, extracellularly applied recombinant GST/calreticulin does not affect the APP processing indicating that it is not functional in this configuration, possibly due to improper folding, sterical hindrance by GST or inaccessibility of domains. Calreticulin also binds to Ab in a time- and concentrationdependent manner, being enhanced in the presence of Ca2+ and optimal at pH 5. These observations suggest that calreticulin binds also to Ab and APP in an acidic environment. However, although this Ca2+-dependent binding appears to be different from the Ca2+-independent interaction between APP and calreticulin observed in the present study, the observation that calreticulin binds to Ab agrees with our findings that calreticulin binds to Ab and 7498254 to sequences flanking the c-secretase cleavage site of APP and overlapping with those of Ab. Calreticulin associates with presenilin and nicastrin and modulates c-secretase activity A novel finding of our study is that calreticulin interacts with presenilin and nicastrin, members of the