Style D 42 by 48

1
Molecular differences in triple negative breast
cancer between race/ethnicities Mark M Bouzyk1,
Benjamin G Barwick1, Mark Abramovitz2, Maja
Kodani1, Gabriela Oprea1,3, Weining Tang1, Carlos
S Moreno1, and Brian Leyland-Jones1,2 1Winship
Cancer Institute, Emory University, Atlanta, GA,
United States 2VM Institute of Research,
Montreal, Canada 3Grady Memorial Hospital,
Atlanta, GA, United States
Abstract
Methods
Results
Cell Cycle Signaling AA v. CAU
RNA Isolation DASL Assay RNA was isolated from
FFPE tissues, either three 5 µm sections or one
core 5 mm in length. Tissues were
deparaffinized, extracted and purified using the
commercially available RNA High Pure Kit (Roche,
Mannheim, Germany) modified as previously
described (Abramovitz Kodani et al., 2008). RNA
concentration and Å260/Å280 ratio were determined
using the NanoDrop spectrophotometer (NanoDrop,
Wilmington, DE). Quality was assessed using
commercially available TaqMan RPL13a assays from
20 ng of RNA on a HT7900 real-time PCR instrument
(Applied Biosystems, Foster City, CA) with a
quality threshold of less than 29.5 CT required
for inclusion. Samples with sufficient quantity
(gt0.4 µg) and quality (Ct lt 29.5) were subjected
to the DASL assay, which is essentially a
multiplexed quantitative RT-PCR (real-time
polymerase chain reaction) and hybridized to
sentrix array matrices an 8x12 plate microarray
according to the manufacturers protocol
(Illumina, San Diego, CA). When ample RNA was
available RNA replicates were run to test for
reproducibility of mRNA expression from FFPE
tissues. Data Quality Normalization DASL mRNA
signal intensities were interpreted in BeadStudio
and quantile-normalized signal intensities were
exported for meta-analysis. Samples with a
signal-to-background ratio less than 3 were
determined unusable and excluded from further
analyses. After data QC and normalization a
discrepancy in average signal intensity was
apparent between the 4 sentrix array matrices
utilized for this experiment. To compensate for
this plate-to-plate variation sample signals were
linearly scaled according to the average plate
signal intensity so that the average signal
across all 4 plates is the same. After
normalization replicate correlations averaged r2
0.97.
Background A disparity in prognosis of triple
negative (TN) breast cancer (BC) has been
observed between African American (AA) and
Caucasian (CAU) race/ethnicities afflicted with
this aggressive and invasive BC subtype.
Etiological understanding of these differences
involves accounting for several factors
associated with phenotype and genotype. Here, we
address the former using the Illumina DASL (cDNA
mediated, Annealing, Selection, Extension, and
Ligation) assay to quantify mRNA expression of
512 breast cancer related genes in a cohort of 24
CAU and 56 AA TN BC tissues sourced from
formalin-fixed, paraffin-embedded (FFPE) blocks.
Material and Methods The DASL assay was used
to measure mRNA expression levels from FFPE
sourced tissues in both cohorts of
self-identified patients. CAU BC patients were
obtained from St. Marys Hospital, Montreal,
Quebec and AA BC patients were obtained from
Grady Hospital, Atlanta, Georgia. RNA extraction
used the RNA High Pure Kit (Roche) and was taken
from archival FFPE tissues either 5µm tissue
sections or cores. Differential mRNA regulation
was identified by Significance Analysis of
Microarrays (SAM) software using a false discover
rate (FDR) less than 1 and a two fold-change
criteria to determine differential
regulation. Results In all, 33 genes were
found differentially expressed between AA and CAU
TN BC tumor samples, 32 of which were upregulated
in the AA cohort, only 1 of which was upregulated
in the CAU group. The upregulated gene in CAU TN
BC was TFF1. Upregulated genes in the AA cohort
(order of statistical significance according to
SAM software) were KIF20A, EP300, AURKB, FGF4,
C14orf155, USP22, EPOR, ZNF668, SCNN1G, MAPT,
FLNB, EP400, LTA, ACOT11, RBP3, CSF3, E2F2,
TGFB1, CCNE1, L1CAM, NDP, VWF, RHOB, FEN1, BIN1,
KRT17, CDC42EP4, SERPINF1, CHI3L2, NES, BCL2, and
RERG. Discussion TFF1 upregulated in the CAU
population, has been indicated as biomarker of
favorable prognosis in endocrine therapy in
clinical studies which is consistent with
race/ethnicity disparities. The remaining genes
upregulated in the AA cohort include
transcription factors E2F2 and RBP3/E2F1 both
with cyclin binding domains which may interact
with CCNE1, extracellular and adhesion related
genes KRT17, L1CAM and FGF4, genes associated
with cell cycle AURKB, EP400, and EP300
(activator of HIF-1A). Several RAS related genes
were also found differentially expressed in the
AA cohort including RHOB, RERG, BIN1, and EPOR.
Moreover, it is worth mentioning that BCL2 which
is expressed in the aggressive mammary cancer
cell line MCF-7 was also found upregulated in the
AA cohort. These initial findings suggest that
several differentially regulated genes between AA
and CAU race/ethnicities may account for the
disparity in outcomes resultant in these
populations. These initial data warrant further
investigation which is currently ongoing.
Data Acquisition Concordance
Immunohistochemical (IHC) staining for ER, PR,
and HER2 was obtained from the respective
coordinating pathology office as was race which
was self-identified. Surveillance, epidemiology
and end results (SEER) data was also obtained for
samples acquired from Grady Memorial Hospital.
Race recorded in the SEER registry was compared
with race identified from the pathology records
and was concordant in 132 of the 134
CHEK2
CDC25B
BRCA1
p53
mRNA Fold Change in AA Cohort
Cyclin B1
BCL-XL
gt 2
CDK1
PCNA
1.5 to 2
RB
PLK1
SWI/SNF
E2F
1 to 1.5
GADD45
p21
-1 to -1.5
Survival
-1.5 to -2
Checkpoint Regulation
Chromatin Remodeling
Cyclin D1
gt -2
CDK4
p57
No Data
breast cancer patients (89 TN). Discordant
records were precluded from further analysis. The
final analysis included 113 TN patients composed
of 87 African Americans and 26 Caucasians.
Inconclusive
Mitotic Kinase
G1/S Phase Transition
CDK4
Cyclin D2
Figure 4. Fold change of mRNA transcripts
superimposed over canonical cell cycle signaling
CDK2
Cyclin E1
Results
Discussion
Introduction
Major health inequities exist between AA and CAU
race/ethnicities emphasized by an unadjusted
mortality rate almost 40 higher in African
Americans. A major component of this disparity is
TN disease for which a lack of therapeutic
targets is driving urgent need for increased
pathological understanding and subsequent
clinical translation of treatment modalities.
Admittedly, there are confounding factors in this
study, specifically that the majority of TN CAU
were sourced from one hospital while the majority
of AA from another, introducing influences such
as diet and standard of healthcare. However, it
is encouraging that dysregulation was indicated
in BRCA1 / p53 biology as previous studies have
found p53 mutational spectra to differ among race
(Trock). Increased expression down the BRCA1
signaling cascade contrasted with slight
attenuation of p53 arm canonically activated by
BRCA suggests a dysfunctional BRCA / p53
interaction. Thus, we have highlighted a well
studied pathway, that shows differential
expression between CAU and AA populations and we
further hypothesize that ablation of cyclin and
E2F transcription factors will be an effective
modality for treating TN disease in African
Americans.
Expression Pathway Analysis Pathway analysis
tools GSEA and IPA yielded a similar theme in
results, specifically indicating at least three
interrelated molecular mechanisms 1) DNA damage
and repair, 2) BRCA1 signaling in DNA damage
response, and 3) the subsequent impact on cell
cycle regulation. The major components of cell
cycle signaling showed a divergence in regulation
between race/ethnicities. We observed significant
upregulation of BRCA1 signaling in the AA cohort.
This signaling was apparent in down stream
targets E2F transcription factors needed to
initiate the G1/S phase transition, SWI/SNF which
is involved in chromatin remodeling and polo-like
kinase PLK1 a mitotic kinase highly expressed in
G2/M cell cycle transition. Retinoblastoma 1
(RB1) a negative regulator of cell cycle
progression was downregulated in AA patients. In
the opposing p53 signaling arm of cell cycle
regulation we observed subtle downregulation in
the AA cohort. This included attenuation of the
CDK inhibitor p21/CDKN1A, the growth arrest and
DNA damage-inducible gene GADD45A, proliferating
cell nuclear antigen PCNA, and increased
expression of BCL-XL, a proliferation gene
negatively associated with p53 signaling.
Counterintuitively we observed abrogated
expression of cyclin D1 as well as the cyclin
dependent kinase CDK4. However, cyclin D2 was
upregulated (1.6 fold change) in AA cohort and
acts as a surrogate in the absence of cyclin D1.
Cyclin E1 was also upregulated (1.5 fold change)
which activates CDK2 to drive G1/S transition in
a p21/CDKN1A independent manner (Ukomadu et al).
Finally, we also observed upregulation of CDC25B
a phosphatase which facilitates cell cycle
progression by activating cyclin dependent
kinases, including CDK1 allowing cyclin B1
upregulated in the AA cohort to bind and
activate components needed for G2/M transition
including the mitotic kinase PLK1. Taken
together, these data suggest upregulation of cell
cycle components through BRCA1 and cyclin
signaling in the AA cohort.
Background 182,460 women are estimated to be
diagnosed with breast cancer (BC) in 2008, which
will result in over 40,000 mortalities from
disease, and account for 26 of new cancer cases
among women in the United States. While BC
incidence is higher in Caucasian (CAU) women, the
mortality rate is substantially higher in African
America (AA) women (33.5 vs. 24.4 per 100,000). A
major component of this health disparity is
triple negative (TN) disease, a particularly
aggressive and invasive BC with little to no
targeted therapies composing 15-25 of breast
cancers and defined by the lack of estrogen
receptor (ER), progesterone receptor (PR), and
human epidermal growth factor receptor 2 (HER2).
This disease differentially affects AA women for
which a disparity in outcomes has been shown to
exist (Lund et al). TN breast cancer patients
experience the worst prognoses as compared to the
HER2 and hormone (ER or PR) positive breast
cancer subtypes. Compounding the tragedy of this
disease is the predisposition for it to afflict
younger and pre-menopausal women. Study Aim
Expansion To investigate potential molecular and
etiological differences between ethnicities in TN
breast cancers we have compared mRNA expression
levels of 87 African Americans and 26 Caucasians
expanded from the original cohort of 56 AAs and
24 CAUs with TN breast cancers sourced from
formalin-fixed, paraffin-embedded (FFPE) tissues.
This expanded cohort yielded expanded yet highly
overlapping results to those identified in the
original abstract. More robust normalization
yielded a higher replicate correlation (r2
0.97) and more balanced results in terms of up
and down regulated genes. Criteria for inclusion
was relaxed to a differential fold change of 1.5
but still imposed a false discovery rate (FDR) of
less than 1. This yielded an expanded set of 113
genes differentially regulated between AA and CAU
cohorts.

References

• Lund MJ, Trivers KF, Porter PL, Coates RJ,
  Leyland-Jones B, Brawley OW, Flagg EW, O'Regan
  RM, Gabram SG, Eley JW. Race and triple negative
  threats to breast cancer survival a
  population-based study in Atlanta, GA. Breast
  Cancer Res Treat 2008 Mar 7
• Bibikova M, Yeakley JM, Wang-Rodriguez J, and Fan
  JB. Quantitative Expression Profiling of RNA from
  Formalin-Fixed, Paraffin-Embedded Tissues Using
  Randomly Assembled Bead Arrays. Methods Mol Biol.
  2008439159-77.
• Abramovitz, M., M. Ordanic-Kodani, Y. Wang, Z.
  Li, C. Catzavelos, M. Bouzyk, G. W. Sledge Jr.,
  C. S. Moreno, and B. Leyland-Jones.
  Biotechniques44417-423, 2008.
• Tusher VG, Tibshirani R, Chu G. Significance
  analysis of microarrays applied to the ionizing
  radiation response. Proc Natl Acad Sci U S A.
  2001 Apr 2498(9)5116-21.
• Ukomadu C, Dutta A. p21-dependent inhibition of
  colon cancer cell growth by mevastatin is
  independent of inhibition of G1 cyclin-dependent
  kinases. J Biol Chem. 2003 Oct 31278(44)43586-94
  .
• Trock BJ. Breast cancer in African American
  women epidemiology and tumor biology. Breast
  Cancer Res Treat. 199640(1)11-24.

Data Analysis Differential mRNA expression was
determined using significance analysis of
microarrays (SAM) (Tusher et al., 2001) for probe
and gene-level data, for which we imposed a false
discovery rate (FDR) of less than 1. Heatmaps
were generated using Gene Cluster 3.0, which
clustered patients and genes using a complete
linkage method (Eisen et al., 1998), heatmaps
were generated with Java Treeview (Saldanha,
2004). Pathway analysis used Ingenuity Pathway
Analysis (IPA), Gene Set Enrichment Analysis
(GSEA), and KEGG.