Page 73 - 2014 Printable Abstract Book
P. 73
(S1803) Identification of biomarkers of therapeutic response in high risk head and neck squamous cell
1
1
carcinoma treated with post-operative radiotherapy (PORT). Michael Story ; John S. Yordy, MD, PhD ;
1
1
2
2
Hao Tang, PhD ; Uma Giri, PhD ; John V. Heymach, MD, PhD ; Liang-hao Ding, MD, PhD ; Narasimha
1
1
1
Karanam, PhD ; and Tae Hyun Hwang, PhD, UT Southwestern Medical Center, Dallas, TX and MD
2
Anderson Cancer Center, Houston, TX
We have chosen a comprehensive approach to develop biomarkers of therapeutic response that includes
not only gene expression but miRNA, protein, methylation and mutation analysis following the notion that
no one biomarker type will be completely informative. Furthermore, we have taken the approach that
biomarkers should have a biological basis. Samples for this study include 102 tumor specimens from
patients considered high risk for local recurrence (LR) or distant metastasis (DM). A 46 gene signature was
able to stratify patients into risk groups for local recurrence or overall survival and was validated in 4 other
data sets. However, no signature could segregate specimen’s specific to recurrence or distant metastasis.
The same approach was taken using miRNA expression and miRNA specific to the LR (60) or DM (94)
groups were identified. Combining the gene expression and miRNA datasets, filtering through the putative
miRNA target identification databases and examining negative correlations of miRNA and gene
expression, led to the isolation of several miRNA for biological validation in a battery of 39 HNSCC cell
lines. Overexpressed in DM specimens, miR-551a and miR551b-3p were shown to enhance cell
proliferation, migration and invasion when up-regulated while all were reduced when knocked down by
inhibitor. Overexpression of these miR’s resulted in increased phosphorylation of AKT, down-regulation
of the tumor suppressor p27Kip1 and down-regulation of GLIPR2, a protein that sequesters the autophagy
regulator BECN1. These miR’s may also down-regulate MUC15 allowing EGFR dimerization and PI3
signaling and which is associated with shorter survival times in HCC. Under-represented in the LR group
was the miRNA miR-125a, a miRNA already known as a negative prognostic indicator and which targets
ERBB2. Indeed ERBB2 gene expression is anti-correlated with miR125a expression in tumor specimens,
and ERBB2 protein is down-regulated when miR-125a is up-regulated in HNSCC cell lines. In addition, the
up-regulation of miR125a via mimic radiosensitized cells while the down-regulation of miR-125a by
inhibitor increased the radioresistance of HN5 cells. Identification of other targets, verification of physical
interaction between miRNA and target and validation of miRNA signatures in independent data sets is
ongoing.
1
1
(S1804) Ongoing developments in radiation metabolomics. Maryam Goudarzi ; Tytus D. Mak ; Waylon
3
4
2
1
2
3
Weber ; Lubomir Smilenov ; Melanie Doyle-Eisele ; Dunstana Melo ; Congju Chen ; Evagelia C. Laiakis ;
1
3
1
2
Raymond A. Guilmette ; David J. Brenner ; Albert J. Fornace, Jr., Georgetown Univ, Washington, DC ;
3
2
Lovelace Respiratory Research Institute, Albuquerque, NM ; Columbia Univ, NY, NY ; and Lovelace,
Albuquerque, NM
4
The use of modern liquid chromatography coupled with time-of-flight mass spectrometry (QTOF) has
demonstrated utility in characterizing responses to radiation exposure at the metabolomic level, and
results are promising for its implementation in high-throughput biodosimetry after a radiologic or nuclear
event. In addition to untargeted profiling by QTOF approach, a number of targeted metabolomics
approaches are being developed to assess a more limited number of specific metabolites of interest. Most
radiation metabolomics studies to date have used external beam [[Unsupported Character - Symbol Font
γ]] irradiation delivered at a standard high dose rate (typically ~1 Gy/m or higher), but many exposure
71 | P a g e
1
1
carcinoma treated with post-operative radiotherapy (PORT). Michael Story ; John S. Yordy, MD, PhD ;
1
1
2
2
Hao Tang, PhD ; Uma Giri, PhD ; John V. Heymach, MD, PhD ; Liang-hao Ding, MD, PhD ; Narasimha
1
1
1
Karanam, PhD ; and Tae Hyun Hwang, PhD, UT Southwestern Medical Center, Dallas, TX and MD
2
Anderson Cancer Center, Houston, TX
We have chosen a comprehensive approach to develop biomarkers of therapeutic response that includes
not only gene expression but miRNA, protein, methylation and mutation analysis following the notion that
no one biomarker type will be completely informative. Furthermore, we have taken the approach that
biomarkers should have a biological basis. Samples for this study include 102 tumor specimens from
patients considered high risk for local recurrence (LR) or distant metastasis (DM). A 46 gene signature was
able to stratify patients into risk groups for local recurrence or overall survival and was validated in 4 other
data sets. However, no signature could segregate specimen’s specific to recurrence or distant metastasis.
The same approach was taken using miRNA expression and miRNA specific to the LR (60) or DM (94)
groups were identified. Combining the gene expression and miRNA datasets, filtering through the putative
miRNA target identification databases and examining negative correlations of miRNA and gene
expression, led to the isolation of several miRNA for biological validation in a battery of 39 HNSCC cell
lines. Overexpressed in DM specimens, miR-551a and miR551b-3p were shown to enhance cell
proliferation, migration and invasion when up-regulated while all were reduced when knocked down by
inhibitor. Overexpression of these miR’s resulted in increased phosphorylation of AKT, down-regulation
of the tumor suppressor p27Kip1 and down-regulation of GLIPR2, a protein that sequesters the autophagy
regulator BECN1. These miR’s may also down-regulate MUC15 allowing EGFR dimerization and PI3
signaling and which is associated with shorter survival times in HCC. Under-represented in the LR group
was the miRNA miR-125a, a miRNA already known as a negative prognostic indicator and which targets
ERBB2. Indeed ERBB2 gene expression is anti-correlated with miR125a expression in tumor specimens,
and ERBB2 protein is down-regulated when miR-125a is up-regulated in HNSCC cell lines. In addition, the
up-regulation of miR125a via mimic radiosensitized cells while the down-regulation of miR-125a by
inhibitor increased the radioresistance of HN5 cells. Identification of other targets, verification of physical
interaction between miRNA and target and validation of miRNA signatures in independent data sets is
ongoing.
1
1
(S1804) Ongoing developments in radiation metabolomics. Maryam Goudarzi ; Tytus D. Mak ; Waylon
3
4
2
1
2
3
Weber ; Lubomir Smilenov ; Melanie Doyle-Eisele ; Dunstana Melo ; Congju Chen ; Evagelia C. Laiakis ;
1
3
1
2
Raymond A. Guilmette ; David J. Brenner ; Albert J. Fornace, Jr., Georgetown Univ, Washington, DC ;
3
2
Lovelace Respiratory Research Institute, Albuquerque, NM ; Columbia Univ, NY, NY ; and Lovelace,
Albuquerque, NM
4
The use of modern liquid chromatography coupled with time-of-flight mass spectrometry (QTOF) has
demonstrated utility in characterizing responses to radiation exposure at the metabolomic level, and
results are promising for its implementation in high-throughput biodosimetry after a radiologic or nuclear
event. In addition to untargeted profiling by QTOF approach, a number of targeted metabolomics
approaches are being developed to assess a more limited number of specific metabolites of interest. Most
radiation metabolomics studies to date have used external beam [[Unsupported Character - Symbol Font
γ]] irradiation delivered at a standard high dose rate (typically ~1 Gy/m or higher), but many exposure
71 | P a g e
