We recently published our work in this journal on the ef-fectsofrapamycinonpancreaticb-cellsdysfunctioninisletgrafts, including autophagy (1). The comment followed byPallet(2)providesfurtheropportunitytoemphasizeseveralpoints discussed in our paper.Autophagy is a tightly regulated degradation process thatdelivers cytoplasmic material to the lysosomes via theautophagosomes. We reported previously that rapamycinupregulates autophagy in islet b-cells and that this ef-fect results in cell death and impaired insulin potency.Autophagosomes contain microtubule-associated protein-light-chain-3(LC3),whichisusedasamarkerofautophagy.The rapamycin-induced autophagy in pancreatic islets wasconfirmed by two widely used methods. First, the uncon-jugated (LC3-I) and conjugated forms (LC3-II) of LC3 wereseparated by SDS-PAGE; then immunoblotting showed ac-cumulation of LC3-II, which correlates with the numberof autophagosomes relative to the amount of endoge-nous LC3-II protein in islets. Second, LC3 distribution wasmappedinGFP-LC3transgenicmice.ThismethodisbasedondetectionofLC3-positivestructuresbyfluorescencemi-croscopy, although it has certain limitations. Under certainconditions, e.g. LC3 overexpression, LC3 easily incorpo-rates into the protein aggregates, hindering accurate lo-calization. Although this limitation is independent of au-tophagy, any LC3 localization by this method should beinterpreted with caution. In our study (1), we did not per-form transient transfection with GFP-LC3 gene but usedstable GFP-LC3 transgenic mice. The data showed in-creased autophagosome production in islets, though thiswas not confirmed by electron microscopy.A major problem inautophagyresearch isthelack of highlyspecific autophagy inhibitors. PI3-kinase inhibitors such aswortmannin and 3-MA (3) are the most commonly usedautophagy inhibitors, but they inhibit the activities of bothclass I and III PI3-kinase. Indeed, 3-MA, which was used inour study (1), targets other kinases and affects other cel-lular processes, especially at high concentrations. Accord-ingly, we used this regent at 10 mM, which is the recom-mended concentration for inhibition of autophagy. While3-MA and wortmannin block the formation of autophago-somes at early stages of autophagy, other PI3-kinase in-hibitors (e.g. bafilomycin A1), which was recommended byPallet (2), block later stages of autophagy. Bafilomycin A1was reported to inhibit autophagosome-lysosome fusion,and also altered intralysosomal degradation by inhibitingacidification (4).Recent loss-of-function studies of Atg genes have ana-lyzedtheparadoxicalfunctionsofautophagyincellsurvivaland death. We also examined this issue in our rapamycin-treated islets (1). Another study demonstrated that in-duction of autophagy over 48 h resulted in cell death (5)whereas induction over 12 h had no effect on cell viabil-ity. These findings support the more rational belief that au-tophagyfacilitatestheremovalofdamagedproteinsandor-ganelles, thus providing temporary protection during tran-sient stress; but can also lead to “autophagic cell death”under prolonged stimulation. Perhaps this death-inducingfunctionofautophagyistriggeredbythedigestionofacriti-calamountofcytoplasmiccontentsnecessaryforsurvival,with subsequent activation of death-promoting molecules.Further research is required in this field including crosstalkbetween autophagic and apoptotic cell death.