Page 27 - Read Online

Seno et al. Cancer Drug Resist 2019;2:335-50 I http://dx.doi.org/10.20517/cdr.2019.01 Page 347
[37]
type p53, but the transcriptional activity is suppressed by methylation of lysine residues . Interestingly,
teratocarcinoma cells undergo apoptosis upon DNA damage, suggesting that the suppression mechanism is
[38]
removable . Thus, (re)activation of p53 could be effective in eliminating tumors in various types of cancers.
Similarly, normal embryonic stem cell express the p53 gene, and p53-mediated cell death or differentiation
can be induced [21,26] . Therefore, p53 is also speculated to regulate CSC properties, and the strategies for
[39]
activation of the p53 pathway by stimuli, such as DNA damage, have been adopted to eliminate CSCs .
Daunorubicin was found to induce oligonucleosomal DNA fragmentation, a typical feature of apoptotic
[23]
cell death. CAD/DFF40 is the major endonuclease responsible for this DNA fragmentation . Under non-
apoptotic conditions, CAD is bound to the inhibitor ICAD/DFF45, thereby repressing CAD activity. Once
apoptosis triggered, executor caspase-3, 6, or 7 cleaves ICAD to release CAD, after which the endonuclease
activity leads to DNA fragmentation. However, in the present study, ICAD cleavage was not observed despite
the apparent activation of the p53-caspase pathway, thereby resulting in PARP-1 cleavage. The observed
discrepancy can be explained by the more efficient cleavage of PARP by caspase-7 than ICAD, while the
[25]
catalytic efficiencies of caspase-3 for ICAD and PARP were comparable . The above findings showed that
both caspases were activated by daunorubicin treatment in miPS-LLCcm cells, indicating that dominant
PARP-1 cleaved by caspase-7 is feasible. Post-translational modification of ICAD is a likely mechanism
behind resistance to ICAD degradation. Recently, O-linked beta-N-acetylglucosamine (O-GlcNAc)
modification of proteins has been demonstrated to regulate apoptosis under conditions of various cellular
stresses [40,41] . ICAD could be modified with O-GlcNAc, and this modification protects cleavage by caspases
[43]
[42]
in T cells . Furthermore, O-GlcNAc modification of proteins regulates the differentiation of ES cells .
Similarly, the oligonucleosomal DNA fragmentation without ICAD degradation was described during
[31]
apoptosis of immature B-cell lines induced by B-cell receptor ligation . Moreover, treatment with DNA
topoisomerase II inhibitors, both daunorubicin and doxorubicin, has been reported to promote histone
[44]
eviction and attenuate the DNA damage response . Although epigenetic analysis was not conducted
in the present study, treatment with daunorubicin and doxorubicin could be expected to induce DNA
fragmentation via that mechanisms.

2+
While Ca -independent CAD activity is not likely to be responsible for the DNA fragmentation observed
in miPS-LLCcm cells duringir daunorubicin-induced apoptosis [Figure 5E] . Ca /Mg -dependent
2+
2+
[32]
endonucleases, such as EndoG, are known to contribute to oligonucleosomal DNA fragmentation. EndoG
is a mitochondrial endonuclease that translocates to the nucleus during apoptosis . Cell death of mouse
[27]
embryonic stem cells induced by etoposide, an inhibitor of topoisomerase II, was found to be EndoG-
[26]
dependent and was proposed as a novel mechanism behind programmed cell death, given that significant
activation of caspases was not observed. The contributions of cathepsins for proteolytic observations and
EndoG for DNA fragmentation were suspected in this programed cell death. By contrast, our results
suggested the involvement of caspases in the cell death of miPS-LLCcm cells, since the inhibition of
caspases by pan-caspase inhibitor, Z-VAD-FMK, attenuated DNA fragmentation [Figure 3E]. Therefore,
the involvement of caspases in EndoG-mediated DNA fragmentation remains controversial. EndoG
[27]
translocation was originally reported to occur in a caspase-independent manner , whereas the release of
[45]
this nuclease from mitochondria was described to require caspase activation . Positive feedback loop of
[46]
caspase-3 was proposed to be mediated by the release of AIF/EndoG . However, EndoG is not likely to be
involved in daunorubicin-induced apoptosis of miPS-LLCcm cells, given that EndoG translocation from
mitochondria to nuclei was not detected in the preliminary experiment [Supplementary Figure 2].

Regulation of apoptosis is complex and dependent on the cellular context. Further investigations using
our miPS-CSCs, such as selection of endonuclease in DNA fragmentation in CSCs, could shed light on the
mechanisms associated with apoptosis. These analyses can help elucidate the mechanisms underlying drug-
and apoptosis-resistance in cancer and CSCs and could lead to the identification of novel therapeutic targets.
Taken together, we propose that our miPS-CSCs could be useful as models for the screening of anti-CSC