The Role of PARP in DNA Repair and its Therapeutic Exploitation - DNA Repair in Cancer Therapy - page 65

differentiated cancers; 171 Martin et al. , 172 described
increased RAD51 expression in BRCA1 deficient breast
tumors by circumventing its requirement in RAD51
subnuclear assembly. In a study by Soderlund et al. , 173
weak expression of MRN (MRE11/RAD50/NBS1)
complex was correlated with high histologic grade,
estrogen receptor negativity and radioresistance. It
remains to be seen whether a signature biomarker to
predict HRD can be developed by this approach.
Circulating tumor cells are easy to access in many
advance cancers and are minimally invasive providing
an opportunity to explore predictive biomarkers of
response and to conduct longitudinal studies. 174 CTCs
have been a pharmaco-dynamic biomarker used
to study g -H2AX induction after treatment with DNA
damaging agents. 175 Similar studies evaluating
RAD51 foci formation and other DNA repair proteins
in CTCs are ongoing. 176 Study of CTC DNA and circu-
lating tumor DNA in plasma for methylation status of
the BRCA gene promoter using methylation specific
polymerase chain reaction with bisulphite-converted
DNA has been described in melanoma patients. 177
Preclinial and clinical studies involving development
of CTC will be essential to their future role as tumor
surrogates.
Other approaches to assess HR function described in
cell lines, but not many studies involving clinical mate-
rial/tumor samples. DR-GFP assay was first described
by Pierce et al. in 1999. 178 The principle behind these
assays is to transfect a cell with unknown HR status
with a plasmid vector deficient in certain cellular func-
tion. The vector borrows the HR mechanism of the
host cell to restore that specific cellular function which
can be then quantified. In HR-deficient cells, cellular
function will not be restored. Various studies have
used this assay to assess the contribution of several
genes and proteins involved/not involved in the HR
pathway. 179,180 However, the assay may be of limited
use in primary cultures/tumor samples, as they are
difficult to transfect. In a recent publication, 181 it was
shown that HR-defective cells have hyperactivated
PARP activity in an in vitro PARP activity assay.
However, PARP inhibitor resistant BRCA2 defective
cells reverted back to normal PARP activity. It was sug-
gested that PARP activity could be a potential functional
biomarker to identify HRD tumors that will respond to
PARP inhibitor therapy. Also that it might be used
to identify PARP inhibitor-resistant BRCA1/BRCA2
mutated cancers. While these assays works in cell lines,
it remains to be validated in tumor samples or primary
cultures. In a study by Plummer et al. , 118 PARP activity
in melanotic melanoma human tumor biopsies was
highly variable consistent with tumor heterogeneity.
To summarize, although various approaches to
predict HRD continue to be developed, such assays

sensitive leukemic cell lines and primary cultures devel-
oped from AML patients demonstrated increased
g -H2AX phosphorylation and decreased RAD51 foci
formation compared to PARP inhibitor resistant cells
supporting the hypothesis that g -H2AX/ RAD51 foci
can be a potential biomarker for HR defects and PARP
inhibitor sensitivity irrespective of tumor types.
Aberrant DNA methylation is a frequent epigenetic
event in several cancers and has been shown to have
potential as a predictive molecular biomarker. 155 In the
context of HR, FANCF methylation has been reported
to be 20 27% in primary ovarian tumors. 156,157

Promoter methylation resulting in transcriptional inacti-
vation of BRCA1 gene have been demonstrated in about
5

e

31% of
sporadic breast cancers. 158 e 163 DNA microarray anal-
yses indicate that breast cancers arising in the setting
of germ line BRCA1 mutations have unique gene
expression profiles, which are identical to sporadic
tumors with methylated BRCA1 and distinct from other
types of breast cancer. 164 In a recent study by Press
et al. , 165 however, it was noted that ovarian tumors
with loss of BRCA function due to mutations typically
had decreased PTEN mRNA levels, whereas tumors
with epigenetic loss of BRCA1 had copy number gain
of PIK3CA.These studies appear to indicate a potential
for DNA microarrays as a functional assay for HR
defects. The advantages of this approach would be
that DNA is more stable than RNA or proteins and is
readily available from body fluids and tissues including
paraffin blocks.
Immunohistochemistry analysis on formalin fixed
paraffin embedded blocks has been pivotal in the study
of tumor characteristics and predicting chemotherapy
response in several cancers. Tissue microarrays (TMAs)
have been created to study several tumor specimens
together. In an endeavour to develop TMA as a predictive
biomarker for DNA repair, Alexander et al . 166 demon-
strated that in triple negative breast cancers which is
associated with BRCA1 deficiency, expression of several
DNA repair proteins including XPF, FANCD2, PAR,
MLH1, PARP1, pMK2, p53 and Ki67 assessed by
immuno-histochemistry on a TMA can be used to stratify
patients into recurrence risk categories. Several studies
have evaluated the role of HR related proteins in IHC
in correlating clinico-pathological outcomes. Increased
RAD51 expression has been associated with poor
survival and poor tumor differentiation in head
and neck cancers, 167 non-small-cell lung cancers 168 and
radioresistance in soft tissue sarcomas. 169 In prostate
cancer, RAD51 was found to be overexpressed in high
grade prostate cancers irrespective of BRCA mutation
status, thereby suggesting a role as a biomarmarker to
decide type of treatment. 170 In breast cancer loss of
BRCA1 and overexpression of RAD51 was seen in poorly

31% of sporadic ovarian cancers and 11

e e