Response: NAD targeting efficiently kills hematologic cancer cells

Response

We thank Cea and coauthors for sharing their results on APO866 activity in hematologic malignancies. We reported that primary cells and cell lines from patients with hematologic malignancies are highly sensitive toward APO866.¹  Results were obtained from MTT (thiazolyl blue tetrazolium bromide), trypan blue dye exclusion, annexin V/7AAD stainings, clonogenic assays, and xenochimeric transplantation models. We are aware that NAD (nicotinamide adenine dinucleotide) synthesis inhibitors including APO866 potentially skew MTT assay results, the reason for which we confirmed results on cell lines by determining annexin V-7AAD expressions (Nahimana et al,¹  Table 1) and in vivo cell viability assays (Nahimana et al,¹  Figure 6). In addition, Cea and coworkers found similar results to ours on all cell lines except for Molt-4. Discrepancy concerning Molt-4 cell sensitivity to APO866 may be due to the time that cells are maintained in culture before testing. Indeed, we found that some cell lines (ML-2, Namalwa, and Jurkat) cultured for long period (ie, over 12 months [> 60 splittings]) developed resistance to APO866, possibly due to secondary genetic events (Figure 1A). We are currently dissecting the mechanisms of resistance of such cell lines.

Cea and colleagues found lower specific cell death on primary cells than what was published in our article (Nahimana et al,¹  Table 2) and wondered whether the difference could be due to cell death method calculation. As a note, we determined specific cell killing by APO866 to avoid any bias in apoptosis and autophagy inhibition experiments (Nahimana et al,¹  Figure 4 and Figure S1), and we additionally presented raw data of cell death–associated parameters for primary cells from several hematologic cancers highlighting the broad efficacy of APO866 (Nahimana et al,¹ Figures 1,2). We have now expanded such determination on all primary hematologic cancer cells originally reported. Results are quite similar to fdc (fraction of dead cells; Figure 1B). Several explanations may account for the difference observed between our data and those from Cea et al. First, our patient samples used to investigate the killing effect of APO866 had cell viability of more than 85% after thawing. High spontaneous cell death may skew specific cell death rate determination. Second, we determined annexin V-7AAD expressions on gated tumoral cells using cell lineage specific labeled antibodies, information we do not know from Cea and colleagues who used samples containing up to 30% nontumoral cells. Third, we would be interested to know in which culture conditions Cea and colleagues determined APO866 killing activity, because they can influence killing efficacy of antimetabolite cytotoxic agents such as APO866. As shown in Figure 1C, we observed variability in APO866 cell killing efficacy according to cell culture. This latter point also stresses the need to validate in vitro results with in vivo data, where we found that APO866 exerts potent antitumor activities in vivo xenochimeric models of human acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), and lymphoblastic lymphomas without significant toxicities to the animals.

In summary, the cell-killing effect of APO866 on primary hematologic cancer cells is not dependent on the calculation methods of cell death. The confirmation of the MTT data by other cell cytoxicity readouts demonstrates that APO866 is a potent anticancer agent in numerous hematologic malignancies.