Biology Reference
In-Depth Information
types.
86-88
The fact that disruption of
ATR
or
Chk1
in mice is embryonically
lethal
63-65
adds to the concerns about the safety of ATR or Chk1 inhibitors as
drugs. Both ATR and Chk1 play a major role during a normal cell cycle by
monitoring the integrity of replication forks.
59
Even in the absence of exoge-
nously caused DNA damage, their function is critical to maintain genomic
stability, which might cause dangerous side effects in healthy tissues exposed to
Chk1 or ATR inhibitors, especially during long-term or repeated treatment.
In addition, efficacy of these therapies might be highly dependent on the genetic
context of the tumors including their p53 status.
B. Synthetic Lethality Between p53 and ATM/Chk2
Our lab and others have described a similar synthetic lethality between p53
and the ATM/Chk2 pathway.
79
Previous studies in
p53
knockout mouse em-
bryonic fibroblasts (MEFs) suggested that a combined loss of ATM and p53
increased sensitivity to topoisomerase inhibitors when compared to ATM-
expressing cells.
89
We observed that loss of either ATM or Chk2 in both
human and murine cell lines strongly sensitized p53-deficient cells to DNA-
damaging treatment with doxorubicin, resulting in increased cell death.
79
The
interaction between ATM/Chk2 and p53 caused the loss of a functional G2/M
cell cycle checkpoint and drove cells into mitotic catastrophe.
Surprisingly, however, loss of ATM or Chk2 in p53-proficient cells caused
exactly the opposite effect, making the cancer cells resistant to genotoxic stress.
Both sets of results—synthetic lethality in p53-defective tumor cells and syn-
thetic viability in p53-proficient tumor cells—were also confirmed
in vivo
.
Knockdown of ATM in H-Ras
V12
-driven, p53-deficient allograft tumors in
nude mice increased their sensitivity to chemotherapy, as it did in syngeneic
transplants of E
-myc-driven p53-defective lymphomas. In contrast, p53-
proficient tumors in which ATM was downregulated were more resistant to
treatment than control tumors with normal ATM levels in both the xenograft
and lymphoma models.
79
This effect could be explained when the expression of
p53 target genes upon genotoxic stress was analyzed. While cell cycle checkpoint
molecules such as p21 or Gadd45
m
were normally induced in the absence or
presence of ATM, we found that loss of ATM in p53-proficient cells selectively
suppressed the upregulation of proapoptotic target genes such as
Noxa
and
Puma
. Notably, the functional cell cycle arrest without the ability to induce cell
death in p53 wild-type animals seemed to specifically protect the ATM-deficient
p53-expressing cells from doxorubicin treatment (
Fig. 3
).
The results could be directly compared with genetic data on the ATM/Chk2
and p53 status in human cancer patients. Analysis of the 10-year survival data
from a large cohort of breast cancer patients after DNA-damaging chemother-
apy supports the observations about the lethal or protective interactions
between ATM/Chk2 and p53, showing a critical dependence on the p53 status
a
