SRI Supplement to Reproductive Sciences - Volume 25 Number 1 - March 2018 - 92A

92A

Reproductive Sciences Vol. 25, Supplement 1, March 2018

in in vivo subcutaneous models, utilizing the uterine serous cancer (USC)
cell line ARK1, ovarian high-grade serous cell line OVCAR5, and patientderived high-grade serous ovarian cancer xenografts (PDX). Statistical
significance (p<0.05) and IC50 determination was assessed using Prism7.
RESULTS: AXL was found to be over-expressed in tumor cells derived
from ascites of ovarian cancer patients as well as in ovarian and uterine
serous cell lines. Additionally, increased resistance to carboplatin and
paclitaxel was demonstrated in AXL-expressing ovarian and uterine
serous cell lines. Upon AXL inhibition with BGB324, the USC cell line
ARK1 demonstrated a dose-dependent sensitization to paclitaxel therapy
(39-72% decrease in IC50). While BGB324 therapy did not decrease tumor
burden alone in xenograft cell-line mouse models, combination paclitaxel
and BGB324 therapy did decrease tumor volume by 51-67% (p<0.05).
Tumor proliferation rate was also significantly halted by combined AXL
inhibition and paclitaxel therapy. In PDX models, BGB324 therapy
improved tumor response to combined carboplatin and paclitaxel therapy
when compared to chemotherapy alone, inhibitor alone, or vehicle control
(77%, 87%, and 88% decrease in tumor volume respectively p<0.0001).
The mechanism of AXL regulation of chemoresistance appears to be
through the epithelial-mesenchymal transition.
CONCLUSION: In addition to the association of AXL with more
advanced stage and worse survival in ovarian and uterine serous cancers,
AXL contributes to platinum and taxane chemoresistance. Therapeutic
inhibition of AXL with BGB324 restores chemosensitivity in ovarian and
uterine serous cell lines and patient derived xenograft models.

O-105
Deep Immunophenotyping Reveals Unique Immune Signature of
Ovarian Cancer Ascites. Jessica Vazquez†, Gladys Lopez, Melina
Chavarria, Mildred Felder, Arvinder Kapur, Aleksandar K Stanic*, Manish
Patankar*. University of Wisconsin-Madison, Madison, WI, United States.
INTRODUCTION: High grade serous ovarian cancer (HGSOC)
represents a clinical challenge, with late diagnosis and often ineffective
treatment. A lack of diagnostic and stratifying cellular/molecular
signatures available for early diagnosis and to direct effective treatment
presents key research and clinical challenges. As immune cells are now
known to acquire specific phenotypic features resulting from exposure
to disease, they may be a promising resource for the development of
cellular-diagnostic bioassays. To test this hypothesis, we present a proof
of concept machine learning-aided high dimensional flow cytometry
analysis of ascites immune cells.
METHODS: Ascites fluid from 15 HGSOC patients and peripheral blood
from six healthy donors were collected. Mononuclear cells (MCs) were
isolated via centrifugation and frozen until processing. MCs were labeled
with fluorochrome conjugated antibodies against CD1c, 3, 4, 8, 11b, 11c,
14, 16, 19, 25, 27, 33, 34, 45, 45RA, 45RO, 49a, 56, 62L, 80, 94, 117, 123,
127, 123, 141, 161, 183, 184, 194, 196, 197, 209, 335 and intracellular
Eomes, RORγt, and T-bet. Data were acquired with a BD Fortessa flow
cytometer in a 5 laser (355nm, 405nm, 488nm, 562nm, 633nm), 18
parameter configuration. Data was manually analyzed using FlowJo 10.3.
Dimensionality reduction by Barnes Hut-modified t-distributed Stochastic
Neighbor Embedding (t-SNE) and machine-learning aided density-based
clustering (DensVM) was performed using the R Cytofkit package.
RESULTS: Conventional flow cytometry analysis revealed a complex
immune milieu (T-, dendritic-, and innate lymphoid-cells and their
myriad subsets). tSNE-DenseVM analysis mapped these subsets onto
2-dimensional scaffolds, obviating complex expert manual gating. Cluster
distribution revealed a unique "signature" indicative of ascites samples,
highlighting systematic and subtle differences between HGSOC samples
and healthy donors.
CONCLUSION: Dimensionality reduction with machine-learning based
clustering of highly polychromatic flow cytometry data reveals the unique
immune composition of HGSOC ascites fluid and provides the basis for
this analytical pipeline in the development of early diagnosis tools.

Scientific Abstracts

O-106
Identification of Somatic Genetic Alterations in Ovarian Clear Cell
Carcinoma with Next Generation Sequencing. Yusuke Shibuya†,2
Hideki Tokunaga*,2 Bin Li*,2 Jun Yasuda*,1 Nobuo Yaegashi*.2 1Tohoku
Medical Megabank Organization, Tohoku University, Sendai, Miyagi,
Japan; 2Tohoku University School of Medicine, Sendai, Miyagi, Japan.
INTRODUCTION: Ovarian clear cell carcinoma (OCCC) is the most
refractory subtype of ovarian cancer and more prevalent in Japanese than
Caucasians (25% and 5% of all ovarian cancer, respectively). The aim of
this study is to discover the genomic alterations that may cause OCCC
and effective molecular targets for chemotherapy.
METHODS: With institutional approval, paired genomic DNAs of 48
OCCC tissues and corresponding non-cancerous tissues were extracted
from formalin-fixed, paraffin embedded (FFPE) specimens collected
between 2007 and 2015 at Tohoku University Hospital. All specimens
underwent exome sequencing and the somatic genetic alterations were
identified.
RESULTS: We divided the cases into three clusters based on the mutation
spectra. Clinical characteristics such as age of onset and endometriosis
are similar among the clusters but one cluster shows mutations related to
APOBEC activation, indicating its contribution to subset of OCCC cases.
There are three hypermutated cases (showing 12-fold or higher somatic
mutations than the other 45 cases) and they have germline and somatic
mismatch repair gene alterations. The frequently mutated genes are
ARID1A (66.7%), PIK3CA (50%), PPP2R1A (18.8%) and KRAS (16.7%).
Somatic mutations important for selection of chemotherapeutic agents,
such as BRAF, ERBB2, PDGFRB, PGR, and KRAS are found in 27.1% of
OCCC cases, indicating clinical importance of exome analysis for OCCC.
CONCLUSION: We found several somatic genetic alterations, either
novel or reported, in OCCC. FFPE specimens are applicable for exome
analyses. Our study suggests that the genetic instability caused by either
mismatch repair defect or activation of APOBEC, and dysfunction of
chromatin remodeling mediated by ARID1A play critical roles in OCCC
carcinogenesis.

O-107
PLIN2 Functions as a Novel Linker between Progesterone Signaling
and Metabolism in Uterine Leiomyoma Cells. Ijeoma Okeigwe†, Serdar
Bulun, Ping Yin*. Northwestern University, Chicago, IL, United States.
INTRODUCTION: Progesterone and its receptor (PR) exert a
tumorigenic effect on leiomyoma (LM) cells. Using an unbiased ChIPsequencing technique, we previously uncovered perilipin-2 (PLIN2), an
adipose differentiation related protein important in lipid metabolism, as
a novel PR target gene in primary LM cells. The role of PLIN2 in the
pathogenesis of LM is unknown. The present study aimed to determine
the role of PLIN2 in LM pathogenesis.
METHODS: LM and matched adjacent normal myometrial (MM)
tissues were collected from premenopausal women undergoing uterine
surgery. PLIN2 expression was knocked down using small interfering
RNA (siRNA). RNA-seq was performed (N=4) to identify differentially
expressed genes between cultured LM cells transfected with PLIN2 or
non-targeting siRNA. Significance was determined by an FDR-adjusted
p-value < .05. Differentially expressed genes were validated by realtime PCR. Oxygen consumption rates (OCR, mitochondrial oxidation)
and extracellular acidification rates (ECAR, glycolytic respiration) were
determined by measuring extracellular flux (Seahorse Bioscience XF96).
RESULTS: PLIN2 expression was significantly lower in LM vs. adjacent
MM, suggesting that PLIN2 plays a physiological role in MM function.
To further elucidate the role of PLIN2 in LM cells, PLIN2 mRNA were
knocked down 85% via siRNA. Transcriptome profiling demonstrated
3877 significantly differentially expressed genes. A total of 2557 genes
(66%) were upregulated and 1320 (34%) downregulated. Gene Ontology
analysis revealed metabolic processes as the second highest biological
process affected by PLIN2 depletion. Specifically, genes important in
pyruvate metabolism (PKM, LDHA, ENO1, ACACB, ACSS1) and fatty
acid metabolism (ACAT1, ACADM, ACSBG1) were all upregulated
compared to cells treated with non-targeting siRNA, suggesting that
PLIN2 may play a critical role in regulating LM cell metabolism.

Table of Contents for the Digital Edition of SRI Supplement to Reproductive Sciences - Volume 25 Number 1 - March 2018