Farmer et al. Nature 434, 917-921 (2005)
based on a presentation by Samuel F. Bakhoum
Genetics 144: Oncogenomics
Dartmouth Medical School
Course Director: Charles Brenner, Ph.D.
February 27, 2006
Please note: Online
subscriptions necessary to view figures
I. Introduction
One of the most
elusive aspects of cancer treatment is the fact that cancer cells are practically identical to normal cells with the exception of aberrant and uncontrolled cell division. This fact
renders selective cancer treatment a goal difficult to attain and explains the wide-spread use of
non-selective anti-proliferative systemic therapies such as radio- and chemo-therapies. Increased
understanding of the molecular pathways leading to neoplastic transformation has allowed for many
advances toward that goal. In this paper [1] Farmer et al. describe how to selectively target a
subtype of hereditary breast-cancer cells (homozygous mutants for BRCA1 or BRCA2 alleles) by exploiting the role that
BRCA genes play in the double-strand DNA repair pathways [2]. By inhibiting Poly(ADP-ribose)
polymerase (PARP) - which is involved in single-strand break repair - they were able to overwhelm
the BRCA-/- cells with un-repaired DNA lesions which ultimately led to apoptosis. In this manner,
the authors were able to show that DNA damage response could be an anti-cancer 'barrier' to
neoplastic transformation, a claim that was also independently and concurrently shown by Bartkova
et al. [3] and Gorgoulis et al. [4] in this same issue of the journal.
II. PAR and PARP and DNA repair
Poly(ADP-ribose) (PAR) has been recently implicated in many basic cellular functions,
including mitotic spindle organization [Ref. 5, Figure 2]
and transcriptional control. PARP on the other hand uses NAD+ as a substrate to add the
ADP-ribose moiety onto a substrate protein releasing Nicotinamide [Ref. 6, Figure 1]. PARP has been implicated in single strand repair processes that, if left un-repaired,
lead to stalled replication forks and ultimately double strand breaks (Figure 4).
These double-strand breaks lead to two categories of downstream pathways: repair and signaling
that stalls the cell-cycle. DNA double-strand breaks could be repaired either through Homologous
Recombination (HR) or, a less faithful, Non-Homologous End-Joining (NHEJ) pathway [Ref. 7 Figure 1]. The
fidelity of the HR pathway stems from the use of the homologous chromosome as a template to fill
the damaged gap. Conversely, NHEJ is the simple and direct joining of the two broken ends
together which sometime leads to complex chromosomal rearrangement as well as aberrations. Since
double strand breaks are known to be grouped in contiguous clusters, that leads the NHEJ pathway
to be more error prone. BRCA1 and BRCA2 have two essential and non-redundant functions in the HR
pathway and their loss means that the only other alternative for the cells is the
error-prone NHEJ.
The accumulation of DNA double strand breaks leads to various signaling
pathways which in turn activate cell-cycle checkpoints and apoptosis if the signal persists long
enough [Ref. 7 Figure
1]. Hence, in the absence of PARP, BRCA1 and BRCA2 become essential for the cell survival. This
allows for a mechanism of targeting BRCA null cells (typically the cancer genotype of
BRCA heterozygous individuals) while significantly unaffecting BRCA heterozygous or wt cells
(the normal genotype of BRCA heterozygous individuals)
III. PARP Inhibition
The authors
tested their hypothesis by halting PARP function either using PARP siRNA or previously
characterized PARP inhibitors (Figure 1,
a-c). Cells with BRCA-null background showed marked survival disadvantage as compared to
wild-type (wt) cells. Compounds KU0058684 and KU0058948 showed differential toxicity to BRCA-/- cells
as compared to wt cells at concentrations in the micro-molar range. This result presents a convincing
case for the possibility of using these small molecule inhibitors for cancer treatments.
Analyzing the DNA content of these BRCA-/- cells showed G2/M arrest as opposed to almost
normal DNA content in wt cells with PARP inhibition. These arrested cells also showed signs of
early and late-apoptosis as detected by FACS analysis using the markers Annexin V and Propidium
Iodide, respectively. Most importantly, the BRCA-/- cells with depleted PARP activity showed
marked complex chromatid rearrangements and breaks.
IV. Un-repaired DNA double-stranded breaks
The chromosome histology from Figure 1 suggested the persistence of DNA double strand
breaks. The authors decided to stain the nuclei for gamma-H2AX foci, which are indicative of
double-stranded lesions. Both wild-type cells and BRCA-null cells showed equal numbers of foci
per nucleus upon PARP inhibition (Figure 2).
A clearly lacking control herein would be to stain for those foci in wt background without PARP
inhibition to test for background lesions. However, it is reasonable to conclude that PARP
inhibition eventually leads to the accumulation of double-stranded lesions in both cell types. It
is also useful to note that the investigators used ten-fold higher concentration (10 micro-molar)
of the small molecule inhibitor than has been shown to induce apoptosis in their previous work. This could potentially
overwhelm the double-strand repair mechanisms in the wt cells as well as in the BRCA mutant cells
and show excessive number of lesions regardless of the genetic background. Hence, it would have
been more appropriate to stain for H2AX in the presence of only 1 micro-molar drug concentration
and stain at a reasonable time point before the repair pathway start to show an effect.
More convincingly however, the authors stained for RAD51 - indicative of double-strand
break HR repair mechanism - and showed a significant decrease in the double-strand break repair
foci in the BRCA mutant cells as compared with wt cells. Hence, by inhibiting PARP in a BRCA null
background, the cells do not repair the double-stranded lesions via HR which only leaves the
possibility for the error-prone NHEJ repair pathway.
V. In Vivo validation of the drug function
The authors then successfully assayed the effect of the small molecule inhibitors in
vivo by subcutaneously injecting athymic mice with teratomas and showed significant inhibition of
tumour progression only for BRCA null tumours (Figure 3).
These results present a strong case for the use of this drug clinical trials of hereditary breast cancer.
VI. Discussion
The authors have
clearly shown the toxic effect of these small molecule inhibitors on the BRCA-/- cells however they
did not address an important question. That is: cancer cells usually carry a multitude of
bystander and non-bystander mutations that could affect the efficacy of the drug in inducing
apoptosis. In one of the supplementary figures the authors present results showing the effect of
the drug on Human breast cancer BRCA null cells (MCF7) and show much lower effect by one of the
drugs (Supplementary
Figure 2b). Based on the work of Farmer et al., it is assumed that the selective toxicity of
these drugs stems from the inhibition of two essential DNA repair pathways, which leads to the
accumulation of DNA lesions and, consequently, the onset of the for the cell-cycle checkpoint arrest. It
is however known that cancer is a disease of evasion, and hence most likely the cells that
achieve and maintain neoplastic transformation have managed to bypass these cell-cycle
checkpoints. This fact could drastically reduce the efficacy of the drug and explain the much
less pronounced influence on the MCF7 human breast cancer cells with BRCA null background. These cells
likely accumulated a variety of bystander mutations - some which have been selected for - that
allow them to ignore the cell-cycle checkpoints and hence replicate despite chromosomal
aberration. On the other hand, since cellular replication with such dramatic chromosomal rearrangement and
breaks could possibly not proceed for too many cell-division cycles despite bypassing the cell-cycle checkpoints, the
drug might show efficacy over more prolonged period of time, an option the authors failed to
address.
The authors did not address the ramification of inhibiting PARP function on the loss
of PAR and the consequences on the many cellular processes in which PAR is involved such as the
assembly of the mitotic spindle [5]. Failure to assemble the mitotic spindle would equally arrest
the cell-cycle and eventually induce apoptosis. The authors, thus, have not definitively proved
that the induction of apoptosis in BRCA null mutants is solely due to DNA double-stranded
lesions.
VII. Summary
In this report the
authors have presented an appealing case for the potential use of two small molecule PARP
inhibitors as potential selective therapeutics against hereditary BRCA breast cancer which is
highly penetrant within families and individuals with heterozygous mutant BRCA genotype. By
inhibiting the PARP-dependent single-strand repair pathway in BRCA-/- (that lack the
HR double-strand repair pathway) they have selectively caused these cells to undergo apoptosis.
If the same scenario occurs in humans, this drug could selectively target BRCA-/- cells which
usually constitute the breast cancer genotype of otherwise BRCA heterozygous individuals.
VIII. References
1) Farmer et al.
Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature 434,
917-921(2005).
2) D'Andrea AD, Grompe M. The Fanconi anaemia/BRCA pathway. Nat. Rev. Cancer 3, 23-34
(2003).
3) Bartkova et al. DNA damage response as a candidate anti-cancer barrier in early human
tumorigenesis. Nature 434, 864-870 (2005).
4) Gorgoulis et al. Activation of the DNA damage checkpoint and genomic instability in
human precancerous lesions. Nature 434, 907-913 (2005).
5) Chang P, Jacobson MK, Mitchison TJ. Poly(ADP-ribose) is required for spindle assembly
and structure. Nature 434, 645-649 (2004).
6) Burkle A. Poly(ADP-ribose). The most elaborate metabolite of NAD+. FEBS J. 272,
4576-4589 (2005).
7) O'Driscoll M, Jeggo PA.The role of double-strand break repair - insights from human
genetics. Nat. Rev. Genet. 7, 45-54 (2006).
