Apoptosis is induced in ATM and TERT doubly deficient HSCs as a result of ROS fragility and is regulated by p38MAPK and BCL-2 family proteins. (A) ROS levels in LSKs of the indicated genotypes. Comparable ROS levels in WT and TERT−/− LSKs, compared with elevated ROS in ATM−/− and ATM−/−TERT−/− LSKs, were observed. Representative FACS analysis profiles derived from 3 experiments (left) and fluorescence intensity relative to WT (mean ± SD; right) are shown (*P < .05, n = 3). (B) Native apoptosis of CD34+/CD34− LSKs or CD150−/CD150+ LSKs. Freshly isolated CD34+/CD34− LSKs or CD150−/CD150+ LSKs from mice of each genotype were stained with annexin V and PI. CD34− LSKs and CD150+ LSKs in ATM−/−TERT−/− mice showed significantly higher apoptosis rates compared with the other genotypes. Data shown represent the mean percentage (± SD) of annexin V–positive cells out of total analyzed cells (n = 3). (C) Apoptosis following ROS induction. LSKs from WT, TERT−/−, ATM−/−, or ATM−/−TERT−/− mice were cultured with or without BSO and/or NAC. After 12 hours of culture, harvested cells were stained with annexin V and PI. Apoptosis was induced in TERT−/−, ATM−/−, and ATM−/−TERT−/− LSKs by BSO, and was rescued by adding NAC in a dose-dependent manner in all groups. Note: apoptosis was already induced in ATM−/−TERT−/− LSKs in culture without BSO to some extent. The data shown represent the mean percentage (± SD) of annexin V–positive cells (n = 4). (D) Native apoptosis of CD34+/CD34− LSKs or CD150−/CD150+ LSKs in NAC-administered mice. Freshly isolated CD34+/CD34− LSKs or CD150−/CD150+ LSKs from mice treated with NAC were stained with annexin V and PI as for panel B. Native apoptosis of CD34− LSKs or CD150+ LSKs seen in ATM−/−TERT−/− mice is completely rescued. (E) Activation of p38MAPK in TERT−/−, ATM−/−, or ATM−/−TERT−/− murine LSKs (top), which is completely rescued in LSKs from NAC-treated mice in vivo (bottom). LSKs from the indicated genotypes were stained with anti–phospho-p38MAPK Ab (red) and DAPI for nuclear stain (blue: magnification ×400). Representative pictures (left) and the mean ratios (± SD) of phosphorylated p38MAPK-positive cells in untreated LSKs (right) are shown. (F) Rescued ROS-induced apoptosis in LSKs by a p38MAPK inhibitor. LSKs from WT, TERT−/−, ATM−/−, or ATM−/−TERT−/− mice were treated with BSO, and then, for rescue, with a p38MAPK inhibitor, SB203580, or a JNK inhibitor, SP600125. Annexin V staining assays were performed as for panel C. While apoptosis induced in ATM−/− LSKs by BSO was completely rescued by adding SB203580, in a dose-dependent manner, rescue was only partial in TERT−/− and ATM−/−TERT−/− LSKs. The addition of SP600125, on the other hand, did not result in rescue. (G) Expression of apoptosis-related genes in aged murine HSCs. qPCR analysis of p53, PUMA, and BCL-2 in LSKs from mice of the indicated genotypes. While p53 expression was comparable among all groups, elevated PUMA expression in ATM−/− and ATM−/−TERT−/− LSKs and decreased BCL-2 expression in TERT−/− and ATM−/−TERT−/− LSKs were seen. Data shown are the mean ratio (± SD) of mRNA to β-actin levels (*P < .05, n = 4). (H) Schematic summarizing the role of ATM and TERT in HSCs.