Transcript NKX2.2 - LabRoots

Functional Analysis of EWS/FLI Target Genes
Via an RNAi Mini-Library
Richard Smith, Jeffrey D. Hancock, Michelle Kinsey, Leah A. Owen, and Stephen L. Lessnick
Center for Children at Huntsman Cancer Institute, Department of Oncological Sciences, and Division of Pediatric Hematology/Oncology
University of Utah, Salt Lake City, UT
Background:
Ewing’s sarcoma is a highly malignant small round cell tumor of childhood.
Most cases of Ewing’s sarcoma are associated with a recurrent chromosomal
rearrangement, t(11;22)(q24;q12). This translocation encodes the EWS/FLI
protein. Most prior work supports the hypothesis that EWS/FLI functions as
an aberrant transcription factor to dysregulate genes involved in oncogenic
transformation. The biggest questions in the field are (1) which genes are
dysregulated by EWS/FLI, and (2) what are their roles in the oncogenic
phenotype of Ewing’s sarcoma?
iNKX2.2
Percent Reduction
in NKX2.2
transcript
p-Value
iPPP1R1A
Percent Reduction
in PPP1R1A
transcript
p-Value
73%
0.041
qRT-PCR reveals knockdown of target transcript
Knockdown EWS/FLI and measure
transcript levels of all 33 genes
via qRT-PCR
50%
0.03
NKX2.2
NR0B1
PPP1R1A
CDH12
ARHN
PEG3
qRT-PCR reveals knockdown of target transcript
14
14
A673-iLuc/PQXIN
12
A673-iLuc/NKX2.2 cDNA
10
A673-iNKX2.2-1/PQXIN
8
A673-iNKX2.2-1/NKX2.2 cDNA
6
4
2
0
0
3
6
9
12
ELN
EPHB3
FCGRT
GLI1
SLC15A2
C7
HOOK1
SH2D1A
CNTNAP2
GSTM4
SCGB1A
SSX2
SSX3
CRLF
GYG2
NPY1R
RET
WWOX
A673 cells with PPP1R1A cDNA rescue
A673 cells with NKX2.2 cDNA rescue
Cumulative
population doubling
Genes that are dysregulated by EWS/FLI are likely to include those that are
involved in the oncogenic phenotype of Ewing’s sarcoma, and those that are
not. To identify which of the upregulated genes are important for Ewing’s
sarcoma transformation, we took a stepwise approach that culminated in
functional analysis using a “mini-RNAi library” approach.
PPP1R1A
33 genes upregulated by EWS/FLI
Cumulative
population doubling
To address these questions, we developed a system to study the effects of
EWS/FLI in Ewing’s sarcoma cells. We used retroviral-mediated RNA
interference (RNAi) to “knock-down” endogenous EWS/FLI expression in
A673 Ewing’s sarcoma cells. We compared control cells to cells with
EWS/FLI knocked-down using oligonucleotide microarray to perform gene
expression analysis. To determine which genes were the most reproducibly
dysregulated, we used a 2.5 fold change cutoff value and identified those that
passed this filter in each of four experimental replicates. Using this approach,
we identified 33 genes that were upregulated by EWS/FLI, and 180 genes
that were downregulated. This result was somewhat surprising, since most
prior data suggested that EWS/FLI functions as a transcriptional activator, and
not as a repressor.
NKX2.2
Summary:
A673-iLuc-lacz
12
A673-iluc-PPP1R1A
10
A673-iPPP1R1A-2-lacz
A673-iPPP1R1A-2-PPP1R1A
8
NKX2.2
NR0B1
PPP1R1A
CDH12
6
4
ARHN
C7
CNTNAP2
GSTM4
ELN
SH2D1A
WWOX
2
0
0
15
3
6
9
12
15
Days
Days
Growth curve shows little effect of NKX2.2 knock-down
Knockdown each validated
gene and test for
transformation
Rescue
knockdown
via cDNA
Growth curve shows little effect of PPP1R1A knock-down
NKX2.2 NR0B1
2500
1600
To confirm that the loss in growth or transformation was specific to the
targeted transcript, and not an “off-target effect,” we performed rescue
experiments. Following knock-down of the particular transcript, a cDNA for
that transcript was introduced that was resistant to the RNAi effect (because
the endogenous 3’ UTR was absent from the construct). Phenotypic analysis
was again performed. Transformation was rescued with two targets: NR0B1
and NKX2.2 (see next column for NKX2.2 data).
In some instances, we observed that knockdown of a given target resulted in
a loss of transformation, but that the effect could not be reversed by
introduction of a cDNA expressing that protein. In some of these cases, this
was true even though two different RNAi constructs gave a similar loss-oftransformation phenotype. Indeed, further analysis of some of these (such as
PPP1R1A) revealed that the endogenous protein was not expressed at
detectable levels (e.g., see column 3 for PPP1R1A data). This suggests that
the RNAi effect observed was an “off-target” or other nonspecific effect.
1200
1000
800
600
400
Soft Agar Colonies
We next developed a “mini-RNAi library” consisting of 3 different retroviralRNAi constructs targeting the 3’ UTR of each of the 24 confirmed EWS/FLI
upregulated transcripts. Each retroviral-RNAi construct was introduced into
A673 Ewing’s sarcoma cells, and an analysis was performed, including qRTPCR to evaluate knock-down of the target transcript, growth assays, and soft
agar assays to evaluate oncogenic transformation. Of the 24 targets
evaluated, 10 of these appeared to be necessary for the transformed
phenotype, because knockdown of these caused alterations in growth in
tissue culture or soft agar.
1400
Soft Agar Colonies
Approach and Results:
We first validated the microarray data using independently-derived biologic
replicates. Of the 33 transcripts analyzed, 24 could be confirmed as
upregulated by EWS/FLI in Ewing’s sarcoma cells using quantitative RT-PCR
(qRT-PCR).
2000
1500
1000
500
200
0
Discussion:
-Two EWS/FLI target genes were identified that play a critical
role in the transformed phenotype of Ewing’s sarcoma:
NR0B1 and NKX2.2.
0
-The “RNAi mini-library” screening approach can be used to
identify functionally-relevant gene targets of oncogenic
transcription factors.
-The “RNAi mini-library” screening approach does suffer from
a high rate of “off-target” or other nonspecific growth effects.
-Two RNAi constructs targeting the same transcript is an
inadequate control in these experiments.
Soft agars show loss and rescue of transformation
Western shows knockdown and rescue of protein levels
Soft agars show loss of transformation but no rescue
Western shows no endogenous PPP1R1A in A673 cells!
-Rescue experiments (in which the targeted protein is reexpressed from a cDNA that is immune to the RNAi effect)
are critical to unveil nonspecific effects.
Acknowledgments:
S.L.L. is supported by grant K08 CA96755, American Cancer
Society grant MGO-111812, the Terri Anna Perine Sarcoma Fund,
a Primary Children’s Medical Center Foundation Innovative
Research Grant, a Hope Street Kids grant, and a Catalyst Grant
from the University of Utah School of Medicine. J.D.H. is
supported by the Children’s Health Research Center at the
University of Utah, and an Amgen Fellow’s grant.