To the Editor: The commentary by Ding and Eskelinen “Do mitochondria donate membrane to form autophagosomes or undergo remodeling to form mitochondrial spheroids?” on our recently published manuscript raises several important points that we wish to address. To do so, we here include several key experiments to clarify further that the mitochondrial vesicles observed in Cook et al. are likely to be autophagosomes [1], rather than “mitochondrial spheroids”. Ding et al. previously showed that when mouse embryonic fibroblasts are treated with the mitochondrial de-coupler agent CCCP, a structure they called a “mitochondrial spheroid” develops. These authors concluded that the spheroids are not autophagosomes because they still form in ATG5 and ATG7 embryonic fibroblasts [2]. However, autophagy is a complex pathway involving multiple mechanisms of activation. While ATG5 and ATG7 can play an important role in autophagosome formation, they are not obligatory in all cases. For example, autophagy can be activated in an ATG5/ATG7 independent manner involving ULK1 and Rab9 [3]. Ding and Eskelinen’s criticism of our study is the lack of electron microscopy (EM) images from cells with inhibited autophagy. We now show EM images from LCC9 cells transfected with ATG7 siRNA (Figure 1). We confirmed that ATG7 knockdown inhibits autophagy as shown by a reduction of LC3-II formation and an accumulation of p62 (Figure 1A). Furthermore, EM images indicate that ATG7 knockdown reduces autophagosome formation approximately by 50% (Figure 1D). Knockdown of ATG7 by RNAi also resulted in the accumulation of mitochondria as measured by COXIV (Figure 1B). We also observed increased mitochondria number (average 13.8 mitochondria per EM image versus 9.1 mitochondria per EM image) in ATG7 siRNA transfected cells when compared with control transfected cells. Taken together, these data imply that autophagy is a major pathway for