Download 113-Duan Z-Clin Cancer Res-2012

Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
Systematic Kinome shRNA Screening Identifies CDK11 (PITSLRE)
Kinase Expression Is Critical for Osteosarcoma Cell Growth and
Proliferation
Zhenfeng Duan, Jianming Zhang, Edwin Choy, et al.
Clin Cancer Res 2012;18:4580-4588. Published OnlineFirst July 12, 2012.
Updated version
Supplementary
Material
Cited Articles
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
doi:10.1158/1078-0432.CCR-12-1157
Access the most recent supplemental material at:
http://clincancerres.aacrjournals.org/content/suppl/2012/07/12/1078-0432.CCR-12-1157.DC1.html
This article cites by 38 articles, 13 of which you can access for free at:
http://clincancerres.aacrjournals.org/content/18/17/4580.full.html#ref-list-1
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
[email protected].
To request permission to re-use all or part of this article, contact the AACR Publications Department at
[email protected].
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
Clinical
Cancer
Research
Human Cancer Biology
Systematic Kinome shRNA Screening Identifies CDK11
(PITSLRE) Kinase Expression Is Critical for Osteosarcoma
Cell Growth and Proliferation
Zhenfeng Duan1, Jianming Zhang3, Edwin Choy1, David Harmon1, Xianzhe Liu1, Petur Nielsen2, Henry Mankin1,
Nathanael S. Gray3, and Francis J. Hornicek1
Abstract
Purpose: Identification of new targeted therapies is critical to improving the survival rate of patients
with osteosarcoma. The goal of this study is to identify kinase based potential therapeutic target in
osteosarcomas.
Experimental Design: We used a lentiviral-based shRNA kinase library to screen for kinases which play a
role in osteosarcoma cell survival. The cell proliferation assay was used to evaluate cell growth and survival.
siRNA assays were applied to confirm the observed phenotypic changes resulting from the loss of kinase gene
expression. CDK11 (PITSLRE) was identified as essential for the survival of osteosarcoma cells, and its
expression was confirmed by Western blot analysis and immunohistochemistry. Overall patient survival was
correlated with the CDK11 expression and its prognosis. The role of CDK11 expression in sustaining
osteosarcoma growth was further evaluated in an osteosarcoma xenograft model in vivo.
Results: Osteosarcoma cells display high levels of CDK11 expression. CDK11 expression knocked down
by either lentiviral shRNA or siRNA inhibit cell growth and induce apoptosis in osteosarcoma cells.
Immunohistochemical analysis showed that patients with osteosarcoma with high CDK11 tumor expression levels were associated with significantly shorter survival than patients with osteosarcoma with low level
of tumor CDK11 expression. Systemic in vivo administration of in vivo ready siRNA of CDK11 reduced the
tumor growth in an osteosarcoma subcutaneous xenograft model.
Conclusions: We show that CDK11 signaling is essential in osteosarcoma cell growth and survival,
further elucidating the regulatory mechanisms controlling the expression of CDK11 and ultimately develop
a CDK11 inhibitor that may provide therapeutic benefit against osteosarcoma. Clin Cancer Res; 18(17);
4580–8. 2012 AACR.
Introduction
Osteosarcoma is the most common primary malignant
tumor of bone. The standard treatment for osteosarcoma
incorporates surgery and chemotherapy involving several
chemotherapeutic agents which include doxorubicin, cisplatin, ifosfamide, and methotrexate (1, 2). If these agents
are unable to lead to favorable tumor response, further
chemotherapeutic options are very limited. Despite aggressive chemotherapy, more than 30% of patients with localAuthors' Affiliations: 1Center for Sarcoma and Connective Tissue Oncology and 2Department of Pathology, Massachusetts General Hospital; and
3
Dana Farber Cancer Institute, Harvard Medical School, Boston,
Massachusetts
Note: Supplementary data for this article are available at Clinical Cancer
Research Online (http://clincancerres.aacrjournals.org/).
Corresponding Author: Zhenfeng Duan, Center for Sarcoma and Connective Tissue Oncology, Massachusetts General Hospital, 100 Blossom
St. Jackson 1115, Boston, MA 02114. Phone: 617-724-3144; Fax: 617726-3883; E-mail: [email protected]
doi: 10.1158/1078-0432.CCR-12-1157
2012 American Association for Cancer Research.
4580
ized osteosarcoma experience metastatic disease. Most of
these patients will eventually develop multidrug resistance
in late stages of osteosarcoma. The average survival period
after metastases is less than one year (1–5). Therefore, there
is urgent need to improve the general condition and the
overall survival rate of patients with metastatic osteosarcoma by identifying novel therapeutic strategies.
The discovery of oncogenic kinases and target-specific
small-molecule inhibitors has revolutionized the treatment
of a selective group of cancers, such as chronic myeloid
leukemia (CML) and gastrointestinal stromal tumors
(GIST). Protein kinases play important roles in regulating
tumor cellular functions—proliferation/cell cycle, cell
metabolism, survival/apoptosis, DNA damage repair, cell
motility, and drug resistance—so it is not surprising that
protein kinases are often oncogenic genes. Kinases such as cSrc, c-ABL, PI3K/AKT, EGFR, MAP, IGF-1R, and JAK are
commonly activated and highly expressed in cancer cells
and are known to contribute to cancer progression (6, 7).
Kinases are now firmly established as a major class of anticancer drug targets. Significant progress has been made in
understanding kinases and their functions. There has been
Clin Cancer Res; 18(17) September 1, 2012
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
CDK11 in Osteosarcoma
Translational Relevance
Kinases play an essential role in cancer cell growth and
survival; however, the roles of most kinases in osteosarcoma cell growth are largely uncharacterized. In the
search for kinases required for osteosarcoma cell growth,
we identified CDK11 (PITSLRE) as a potential target by a
comprehensive human kinome-wide shRNA screening
in osteosarcoma cell lines. Furthermore, knockdown of
CDK11 either by lentiviral shRNA, or by synthetic
siRNA-independent confirmation, can inhibit cell
growth or induces apoptosis in osteosarcoma cells.
Immunohistochemical analysis indicated that osteosarcoma patients with high CDK11 tumor expression levels
were associated with significantly shorter survival than
patients with osteosarcoma low level of CDK11 expression. Systemic in vivo administration of in vivo ready
siRNA of CDK11 reduced tumor growth in an osteosarcoma s.c. xenograft model. These observations show that
CDK11 signaling is essential in osteosarcoma cell growth
and survival, CDK11 may become a promising therapeutic target in the management of osteosarcoma.
mary osteosarcoma tissue. Human osteoblast cells HOB-c
were purchased from PromoCell GmbH, osteoblast cells
NHOst were purchased from Lonza Wallkersville Inc., and
osteoblast cells hFOB were purchased from ATCC. Osteoblast cells were cultured in osteoblast growth medium
(PomoCell) with 10% FBS. All other cell lines were cultured
in RPMI-1640 (Invitrogen) supplemented with 10% FBS,
100 U/mL penicillin and 100 mg/mL streptomycin
(Invitrogen).
Lentiviral human kinase shRNA library screen
The roles of protein kinases in maintaining osteosarcoma
cell growth were examined using MISSION LentiExpress
Human Kinases shRNA library (Sigma). This library contains 3109 lentiviruses carrying shRNA sequences targeting
673 human kinase genes. Screening was carried out by
following the manufacturer’s protocol as previously
described (11, 12).
Proliferation assay
The initial cellular proliferation after lentiviral shRNA
infection was assessed using the CellTiter 96 AQueous One
Solution Cell Assay (Promega) by following the manufacturer’s protocol as previously described (11).
an explosion in the number of kinase inhibitors entering the
clinic, and many more are in preclinical development. The
best known kinase inhibitor is Gleevec used in the treatment
of patients with CML and GIST. Gleevec inhibits the kinase
activity of BCR-ABL in CML and c-Kit in GIST (6, 8).
However, the question remains as to whether other highly
expressed and activated kinases in tumors, including osteosarcoma, can be targeted in a similar manner.
To identify kinases based potential therapeutic target in
osteosarcoma, we conducted a comprehensive kinomewide shRNA screening and found that CDK11 (cyclindependent kinase 11, also known as CDC2L for cell division
cycle 2-like or PITSLRE) is an essential kinase for osteosarcoma cell line KHOS growth and proliferation. Cyclindependent kinases are protein kinases that are critical for
cellular processes such as transcription (9, 10). However, to
our knowledge, the oncogenic role of CDK11 has not been
reported in osteosarcoma previously. Further CDK11 protein expression profiling indicated that the expression levels
of CDK11 in tumor tissues is closely correlated with clinical
outcome.
Synthetic CDK11 siRNA and trasfection
Further validation of CDK11 knockdown phenotype in
osteosarcoma cell lines was carried out with synthetic
human CDK11 siRNA purchased from Ambion at Applied
Biosystems. The siRNA sequence targeting CDK11 corresponded to coding regions (50 - AGAUCUACAUCGUGAUGAAtt-30 , antisense 50 -UUCAUCACGAUGUAGAUCUtg-30 )
of the CDK11 gene. The siRNA oligonucleotides were dissolved in nuclease-free water at a concentration of 100
mmol/L and kept at 20 C until the following transfection
experiment. The nonspecific siRNA oligonucleotides
(Applied Biosystems) were used as negative controls. Osteosarcoma KHOS or U-2OS cells were either plated on 96well plates for cell proliferation assays or plated on dishes
for Western blot protein isolation. Transfections were conducted with Lipofectamin RNAiMax reagent (Invitrogen)
according to the manufacturer’s instruction. Medium was
replaced with RPMI-1640 supplemented with 10% FBS 24
hours after transfection. Total protein was isolated with
RIPA Lysis Buffer (Upstate Biotechnology) 48 hours after
CDK11 siRNA transfection.
Materials and Methods
MTT assay
Effects of CDK11 siRNA on cellular growth and proliferation were assessed in vitro using the MTT assay as described
previously. KHOS or U-2OS were transfected with CDK11
siRNA as described above. After 72 hours of culture, 20 mL of
MTT (5 mg/mL in PBS, obtained from Sigma-Aldrich) was
added to each well and the plates were incubated for 3
hours. The resulting formazan product was dissolved with
acid isopropanol and the absorbance at a wavelength of 490
nm (A490) was read on a SPECTRAmax Microplate Spectrophotometer (Molecular Devices).
Cell lines and cell culture
Human osteosarcoma cell line KHOS was kindly provided by Dr. Efstathios Gonos (Institute of Biological
Research & Biotechnology, Athens, Greece). Ewing sarcoma
cell line TC-71 was provided by Dr Katia Scotlandi (Institute
Orthopedics Rizzoli, Italy). The human osteosarcoma cell
lines, U-2OS and Saos, human ovarian cancer cell line
SKOV-3, and uterine sarcoma cell line MES-SA were purchased from the American Type Culture Collection (ATCC).
Osteosarcoma cell line OSA344 was established from pri-
www.aacrjournals.org
Clin Cancer Res; 18(17) September 1, 2012
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
4581
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
Duan et al.
Apoptosis assay
Caspase-cleaved keratin 18-based quantification of apoptosis was evaluated using the M30-Apoptosense ELISA
assay kit, as per manufacturer’s instructions (Peviva AB).
The ELISA apoptosis detects a 21-kDa fragment of cytokeratin 18 that is only revealed after caspase cleavage of the
protein. KHOS or U-2OS cell lines reverse transfected with
synthetic CDK11 siRNA were seeded at 2,000 cells per well
in a 96-well plate. After 48 hours incubation, the cells were
then lysed by adding 10 mL of 10% NP-40 per well, and the
manufacturer’s instructions for the apoptosis assay were
then followed. Apoptosis was also evaluated by Western
blot analysis using whole-cell lysates immunoblotted with
specific antibodies to PARP (Cell Signaling Technologies)
and its cleavage products.
Western blotting
The concentration of the protein was determined by
Protein Assay Reagents (Bio-Rad) with a spectrophotometer
(Beckman Du-640, Beckman Instruments, Inc.). The rabbit
polyclonal antibody (sc-928) to human CDK11 (PITSLRE)
was purchased from Santa Cruz Biotechnology. The mouse
monoclonal antibody to human actin was purchased from
Sigma-Aldrich. Goat anti-rabbit horseradish peroxidase
(HRP) and goat anti-mouse antibodies were purchased
from Bio-Rad. All other antibodies used in this study were
purchased from Cell Signaling Technologies. Western blot
analysis was conducted as previously reported (13).
Immunofluorescence assay
For immunostainings of cultured osteosarcoma cells,
KHOS or U-2OS cells were grown in 8-well chambers for
24 hours and fixed in 3.7% buffered paraformaldehyde.
Immunostainings were carried out using antibodies against
CDK11 and b-actin. After washing the cells, they were
incubated with Alexa Fluor secondary antibodies. Specifically, the slides were stained with Alexa Fluor 488 (Green)
conjugated goat anti-rabbit antibody (Invitrogen) for
CDK11, and Alexa Fluor 594 (Red) conjugated for goat
anti-mouse antibody for b-actin.
Human sarcoma tumor tissues
Six of the osteosarcoma tissue samples (OST1–OST6)
were obtained from Massachusetts General Hospital sarcoma tissue bank (Boston, MA) and were used in accordance
with the policies of the institutional review board of the
hospital. All diagnoses were confirmed histologically.
Osteosarcoma tissue microarray and
immunohistochemistry
Osteosarcoma tissue microarray was purchased from
Imgenex Corp, which contains 57 tumor tissues. Immunohistochemistry was conducted by following the manufacturer’s instructions with HRP-DAB System Cell and Tissue
Staining Kit (R&D Systems). In brief, primary antibody of
CDK 11 (1:50 dilution) in 1% bovine serum albumin (BSA)
was applied to the deparaffinized slide overnight at 4 C.
After incubation with the HRP-conjugated goat anti-rabbit
4582
Clin Cancer Res; 18(17) September 1, 2012
antibody, and rinses in PBS thrice, bound antibody was
detected with the substrate reagents from HRP-DAB System
Cell and Tissue Staining Kit. Finally, slides were counterstained with Hematoxylin QS (Vector Laboratories) and
mounted with VectaMount AQ (Vector Laboratories).
Evaluation of immunohistochemical staining
CDK11-positive samples were defined as those showing
nuclear staining pattern of tumor tissue. CDK11 staining
patterns were categorized into 6 groups: 0, no nuclear
staining; 1þ, <10% of cells nuclear stained–positive; 2þ,
10%–25% positive cells; 3þ, 26%–50% positive cells; 4þ,
51%–75% positive cells; and 5þ, >75% positive cells. The
percentage of cells showing positive nuclear staining for
CDK11 was calculated by reviewing the entire spot. Categorizing of CDK11 staining was completed by 2 independent investigators. Discrepant scores between the 2 investigators were rescored to get a single final score. CDK11
images were obtained by using a Nikon Eclipse Ti-U fluorescence microscope (Nikon Corp) with a SPORT RT digital
camera (Diagnostic Instruments Inc.).
Statistical analysis
Kaplan–Meier survival analysis (GraphPad PRISM Software; GrahPad Software) was used to analyze the correlation between the level of CDK11 expression and prognosis.
The Student t test was used to compare the differences
between groups. Results are given as mean SD and values
with P < 0.05 were considered as statistically significant.
CDK11 siRNA osteosarcoma tumor therapy
Ambion In Vivo Ready CDK11 siRNA and nonspecific
siRNA were purchased from Applied Biosystems. These In
Vivo Ready, validated siRNA were designed using the Silencer Select algorithm and incorporated with additional chemical modifications for superior serum stability with in vivo
applications (14–17). The Crl:SHO-PrkdcSCIDHrhr nude
female mice at approximately 3 to 4 weeks of age were
purchased from The Charles River Laboratories. To determine the effect of CDK11 siRNA on osteosarcoma cell
growth in xenograft model, KHOS cells (1 106) were
inoculated subcutaneously with Matrigel from BD Biosciences into the right flank of the nude mouse. Two weeks
after injection, the mice were randomized into 3 groups (6
mice/group). Group 1 received injection with sterile saline
(0.9% NaCl), group 2 with In Vivo Ready nonspecific siRNA,
and group 3 with In Vivo Ready CDK11 siRNA. For intratumoral injections, each animal was injected with 20 mL of
PBS containing 10 nmol/L of siRNA. All 3 groups were
treated twice a week for 2 weeks. The health of the mice and
evidence of tumor growth were examined daily. Tumor
volumes were measured at a regular interval of for up to
4 weeks with a digital calipers, Tumor volume (mm3) was
calculated as (width)2 length/2 where W is width and L is
length. Data are presented as mean SD. Tumor tissues
from the above treated animals were collected and placed in
10% formalin and embedded in paraffin for histology
analysis. The silence efficiency CDK11 siRNA on CDK
Clinical Cancer Research
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
CDK11 in Osteosarcoma
shown by cellular proliferation assay (Fig. 1A). To further
characterize the functional role of CDK11 in osteosarcoma,
we rigorously validated the results using multiple independent experiments, including multiple shRNA per CDK11
gene and extensively tested with control nonspecific shRNA
(The sequence of 5 shRNA target different sites of CDK11 in
Supplementary Table S1). The results revealed that 4 out of
5 CDK11 shRNA inhibit osteosarcoma cell growth. We
further validated these results with synthetic human CDK11
siRNA. Consistent observations of the dose-dependent
CDK siRNA inhibition of osteosarcoma cell growth and
survival was established by MTT assay (Fig. 1B). We subsequently measured the expression of CDK11 in siRNA
transfected cells. Western blot analysis suggested that
down-regulated expression of CDK11 protein by CDK11
siRNA associated with the inhibition of cell growth (Fig. 1B
and C).
proteins were determined by immunohistochemical staining, as described above. Animal experiments in the present
study were carried out in compliance with the protocol,
which was approved by the Massachusetts General Hospital
Subcommittee on Research Animal Care (SRAC) under the
protocol number 2009N000229.
Results
CDK11 expression is critical for osteosarcoma cell
growth and survival
To identify the potential therapeutic kinase targets in
osteosarcoma cells, we conducted a comprehensive
kinome-wide screening of lentiviral shRNA kinase library
in osteosarcoma cell line KHOS. Among the targeted kinase
genes, we found that knocking down the expression of
CDK11, PLK1, DYRK1B, and ROCK1 led to inhibitory
growth effects. Of those kinases that we found critical to
osteosarcoma cell growth, the elevated expression of PLK1,
DYRK1B, and ROCK1 have been previously reported in
various human tumors (6, 12, 18–24). However, direct
evidence regarding the relationship between expression of
CDK11 and cancer cell growth and survival is lacking. When
Lentiviral shRNA targeting CDK11 was transduced into
osteosarcoma cell lines KHOS and U-2OS, it led to significantly reduced tumor cell growth and eventual cell death as
1.0
1.0
0.8
0.8
Cell viability (%)
Cell viability (%)
A
0.6
0.4
0.2
CDK11 shRNA (TRCN00000062067)
CDK11 shRNA (TRCN0000006208)
CDK11 shRNA (TRCN00000062069)
0.6
CDK11 shRNA (TRCN000000620610)
pLKO.1 vector shRNA
0.4
Nontarget shRNA
Media control
0.2
KHOS
U-2OS
1.2
Cell viability (absorbance)
Cell viability (absorbance)
CDK11 shRNA (TRCN0000006206)
0.0
0.0
B
1.0
0.8
0.6
0.4
0.2
1.2
Cells only
1.0
Cells + nonspecific siRNA
0.8
Cells + CDK11 siRNA 20 nmol/L
Cells + CDK11 siRNA 40 nmol/L
0.6
Cells + CDK11 siRNA 80 nmol/L
0.4
0.2
0.0
0.0
C
CDK11 knockdown induces apoptosis in osteosarcoma
cell lines
To investigate how CDK11 sustains tumor cell growth
and survival, we investigated cellular events during the cell
death caused by CDK11 knockdown in osteosarcoma cell
lines KHOS and U-2OS. We investigated the potential for
the induction of apoptosis using the M-30-Apoptosense
KHOS
1
2
3
U-2OS
4
5
1
2
3
4
5
1:Cells only
2:Cells + nonspecific siRNA
CDK11
110 kD
3:Cells + CDK11 siRNA 20 nmol/L
Actin
4:Cells + CDK11 siRNA 40 nmol/L
42 kD
KHOS
U-2OS
5:Cells + CDK11 siRNA 80 nmol/L
Figure 1. Effects of CDK11 inhibition by shRNA and siRNA in osteosarcoma cell lines. A, results of lentiviral shRNA targeting CDK11 in KHOS and U-2OS cell
lines. Proliferation was determined by the CellTiter 96 Aqueous One Solution Cell Assay. The data represent a single well of the 96-well plate of MISSION
LentiExpress Human kinases shRNA library as described. B, results of synthetic siRNA against CDK11 in KHOS and U-2OS cell lines. Proliferation was
assessed by MTT as described in the Materials and Methods. The data represent the mean SE of 2 experiments carried out in triplicate. C, dose-dependent
CDK siRNA downregulated the expression of CDK11. KHOS or U-2OS cells were transfected with CDK siRNA in a dose-dependent manner. For Western blot
analysis, 25 mg of total cellular proteins was used for immunoblotting with specific antibody to CDK11. The results were detected by a chemiluminescence
detection system as described in Materials and Methods.
www.aacrjournals.org
Clin Cancer Res; 18(17) September 1, 2012
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
4583
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
Duan et al.
1.4
0.4
CK18 (relative levels)
CK18 (relative levels)
A
0.3
0.2
0.1
Cells only
Cells + nonspecific siRNA
Cells + CDK11 siRNA 20 nmol/L
Cells + CDK11 siRNA 40 nmol/L
Cells + CDK11 siRNA 80 nmol/L
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.0
KHOS
B
1
2
U-2OS
3
4
5
1
2
3
4
5
PARP
118 kd PARP
85 kd cleavage
U-2OS
KHOS
C
1
2
3
4
5
1
2
3
CDK11
4
5
1:Cells only
2:Cells + nonspecific siRNA
MCL-1
3:Cells + CDK11 siRNA 20 nmol/L
Bcl-X L
4:Cells + CDK11 siRNA 40 nmol/L
Survivin
5:Cells + CDK11 siRNA 80 nmol/L
Cyto c
Cyclin D1
Actin
KHOS
U-2OS
Figure 2. Synthetic siRNA targeting CDK11 induces apoptosis in osteosarcoma cells. A, KHOS or U-2OS cells were transfected with CDK siRNA as dosedependent manner. The cells were lysed with 10% NP40 after 48 hours transfection and the apoptosis was determined by M30-Apoptosense ELISA assay as
described in Materials and Methods. B, confirmation of CDK11 siRNA induced apoptosis of PARP cleavage by Western blot analysis. Total cellular proteins
were subjected to immunoblotting with specific antibody to PARP as described in Materials and Methods. C, CDK11 siRNA decreases CDK11 expression and
downregulated antiapoptotic proteins expression. For Western blot analysis, 25 mg of total cellular proteins was subjected to immunoblotting with specific
antibody to CDK11, MCL-1, BcL-XL, survivin, cytochrome C, cyclin D1, and b-actin. The results were detected by a chemiluminescence detection system as
described in Materials and Methods.
ELISA and by Western blotting for the cleavage of PARP.
Depletion of CDK11 by siRNA resulted in a dose-dependent
cell death in KHOS and U-2OS osteosarcoma cell lines,
which was not observed with the nonspecific siRNA transfection (Fig. 2A). Consistent with these results, there was
also a dose-dependent cleavage of PARP for siRNA-mediated depletion of CDK11 (Fig. 2B). CDK11 has been
reported to play an important role in the regulation of gene
transcription and pre-mRNA splicing. We hypothesized
that inhibition of apoptosis induced by CDK11 may be
associated with the deregulation of RNA processing associated with a subsequent decrease of antiapoptotic-dependent proteins. Therefore, we examined whether the inhibition of CDK11 by siRNA could result in decreased
expression of antiapoptotic proteins MCL-1, Bcl-XL, survivin, cytochrome C, and cyclin D1. Most of these antiapoptotic proteins such as MCL-1, BcL-XL, and survivin are
highly expressed in osteosarcoma, and downregulation by
siRNA can inhibit cell growth and induce apoptosis in
osteosarcoma cells. Western blot confirmed that transfection of CDK11 siRNA significantly downregulated the
expression of several of these antiapoptotic proteins in
both in KHOS and U-2OS cells (Fig. 2C), whereas CDK11
knockdown did not alter actin expression. Thus, these
results indicate that CDK11 can control several aspects of
apoptosis signaling.
4584
Clin Cancer Res; 18(17) September 1, 2012
CDK11 is highly expressed in osteosarcoma cell lines
and in tumor tissues
To further confirm the expression of CDK11 and determine CDK11 protein subcellular localization in osteosarcoma cell lines,immunofluorescence assay was used in
KHOS and U-2OS cell lines. Previous studies have reported
that the CDK11 protein localizes both to the nucleus and
cytoplasm. Our results showed that the CDK11 protein is
mainly localized in the nucleus of osteosarcoma cells (Fig.
3A). We also extended our evaluation of the expression of
CDK11 in other types of human cancer cell lines including uterine sarcoma (MES-SA), chondrosarcoma (CS-1),
synovial sarcoma (SS-1), Ewing sarcoma (TC-71), and
ovarian cancer (SKOV-3, 3A, 2008). These tumor cell
lines exhibited a variety of expression levels of CDK11
protein as evaluated by Western blot analysis (Fig. 3B). To
confirm these data in primary cancer, 6 freshly isolated
primary osteosarcoma specimens were also examined by
Western blot analysis to exclude the possibility of CDK11
expression being an artifact induced by in vitro propagation. Different levels of CDK11 expression were once
again observed in these osteosarcoma patient samples,
indicating the endogenous expression of CDK11 in tumor
cells. In the normal human cells, the expression of CDK11
is tightly regulated, for example, in normal human osteoblast cell lines, HOB-c, NHOst, and hFOB there are
Clinical Cancer Research
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
CDK11 in Osteosarcoma
Figure 3. Expression of CDK11 in
osteosarcoma cell lines and
osteosarcoma tissues. A, expression
of CDK11 in KHOS or U-2OS
cells was assessed by
immunofluresecence with antibodies
to CDK11 and b-actin. Cells were
visualized under a fluorescence
microscope after incubated with
secondary fluorescent conjugated
antibodies Alexa Fluor 488 goat antirabbit IgG (green) and Alexa Fluor
594 goat anti-mouse IgG (red) as
described in the Materials and
Methods. B, levels of CDK11
expression were determined by
Western blot analysis in
osteosarcoma (U-2OS and KHOS),
and other types of human cancer cell
lines, including uterine sarcoma
(MES-SA), chondrosarcoma (CS-1),
synovial sarcoma (SS-1), Ewing
sarcoma (TC-71), and ovarian cancer
(SKOV-3, 3A, 2008). C, levels of
CDK11 expression were determined
by Western blot analysis in
osteosarcoma tissues [OST1 to
OST6 and in normal osteoblast cell
lines (HOB-c, NHOst, and hFOB).
OST1 to OST6 represents tissues
samples from 6 patients.
A
CDK11
Merge
KHOS
U-2OS
B
CDK11
110 kD
Actin
42 kD
C
CDK11
110 kD
Actin
42 kD
extremely low and almost undetectable levels of CDK11
(Fig. 3C).
CDK11 expression levels correlates with clinical
prognosis in patients with osteosarcoma
To further validate the clinical relevant of CDK11 expression in patients with osteosarcoma, we analyzed CDK11
protein levels by using osteosarcoma tissue microarray. The
results showed the majority of tumors present on the tissue
microarray had positive staining for CDK11. CDK11 nuclear staining percentage was graded into 6 groups. By comparing the clinical characteristics of low-staining (< ¼ 3) and
high-staining (> ¼ 4) osteosarcomas, no correlation existed
between CDK11 expression and age or tumor location (P >
0.05, Supplementary Table S2). Kaplan–Meier survival
analysis showed that the outcome for patients in the CDK11
high-staining group was significantly worse than for those
in the CDK11 low-staining group (Fig. 4A). On the basis of
60 months survival rates, patients were grouped into survivors (survived up to 60 months post follow up) and nonsurvivors (deceased within 60 months of follow-up). A total
of 30 (67%) samples from survivors and 15 (33%) samples
from nonsurvivors were collected. Comparison of CDK11
staining intensity between 2 group patients revealed that
CDK11 staining for samples from nonsurvivors were significantly higher than that of survivors. The average CDK11
expression levels for survivors and nonsurvivors were 2.8 to
www.aacrjournals.org
Actin
4.3, respectively (Fig. 4B). The immunohistochemical staining of CDK11 protein indicated its location in the nucleus
(Fig. 4C) which is consistent with that of osteosarcoma cells,
as measured by immunofluorescence assay (Fig. 3A).
CDK11 siRNA inhibit tumor growth
The significant association of CDK11 expression with
clinical outcome led us to further verify the essential role
of CDK11 in sustaining osteosarcoma growth in vivo.
KHOS osteosarcoma cells (1 106) were injected subcutaneously into the flank of nude mice. By 2 weeks, visible
tumors had developed at injection sites (mean tumor volume ¼ 52 mm3). CDK11 siRNA was then intratumorally
injected, twice a week for 2 weeks. As shown in Fig. 5A,
CDK11 In Vivo Ready siRNA significantly suppressed tumor
growth as compared with vehicle (saline) and nonspecific
siRNA treatment. The immunohistochemical staining indicated a significant decrease of CDK11 expression in tumor
treated with CDK11 In Vivo Ready siRNA (Fig. 5B). No gross
adverse effects, i.e.the loss of body weight, were observed
during the experimental period.
Discussion
To identify essential kinases which are responsible for
osteosarcoma growth, we first conducted a kinome-wide
shRNA screen. CDK11, PLK1, DYRK1B, and ROCK1 were
the primary hits with loss of expression of these kinases
Clin Cancer Res; 18(17) September 1, 2012
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
4585
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
Duan et al.
Survival proportions
100
≤3
≥4
80
B
CDK11 staining
Percent survival
A
60
40
20
0
P = 0.0006
0
50
100
150
Months
C
5
4.3
4
3
2.8
2
1
P = 0.0016
0
200
Survivor
HE (×40)
Nonsurvivor
CDK11 (×40)
Score 2+
Score 5+
A
Tumor volume (mm3)
significantly reduced cell growth and survival. We and
others had previously found that targeting PLK1, DYRK1B,
and ROCK1 kinases inhibits osteosarcoma cell growth and
250
survival by using similar shRNA or siRNA kinase library
screenings (11, 12, 18, 25–27). However, the relationship
between the expression of CDK11 and osteosarcoma
Vehicle controls
Nonspecific siRNA
In vivo ready CDK11 siRNA
200
150
100
50
0
1
5
9
13
17
21
25
29
Times after treatment (d)
B
Vehicle control
Nonspecific siRNA
CDK11
Figure 4. Association of CDK11
expression with clinical outcome in
osteosarcoma. A, Kaplan–Meier
survival curve of patients with
osteosarcoma are subgrouped as
either CDK11 low staining (CDK11
staining 3) or high staining
(CDK11 staining 4). B,
distribution of CDK11 staining
scores among the survivors and
nonsurvivors. C, representative
images of different
immunohistochemical staining
intensities of CDK11 are shown in
osteosarcoma tissues. For CDK11
immunohistochemical staining, the
percentage of cells showing
positive nuclear staining for CDK11
was calculated by reviewing the
entire spot. On the basis of the
percentage of cells with positive
nuclear staining, the staining
patterns were categorized into 6
groups: 0, no nuclear staining; 1þ,
<10% of cells stained positive; 2þ,
10% to 25% positive cells; 3þ,
26% to 50% positive cells; 4þ,
51% to 75% positive cells; 5þ,
>75% positive cells.
CDK11 siRNA
Figure 5. Inhibition oftumor growth
by In Vivo Ready CDK11 siRNA in a
xenograft mouse model. A, CDK11
In Vivo Ready siRNA, vehicle
(saline) control, and nonspecific
siRNA were injected into the tumor
region. Day 1 corresponds to 2
weeks after inoculation of KHOS
cells when tumor volume was 50 to
3
60 mm . Tumor diameters were
measured at a regular interval of 4
days for up to 4 weeks with a digital
caliper, and the tumor volume was
calculated. B, Histologic analysis
of effect of CDK11 siRNA on
CDK11 staining in osteosarcoma
tumor tissues show
downregulation of CDK11
compared with vehicle or
nonspecific siRNA.
HE
4586
Clin Cancer Res; 18(17) September 1, 2012
Clinical Cancer Research
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
CDK11 in Osteosarcoma
growth and survival was lacking. The important functional
role of CDK11 in osteosarcoma cells was further validated
by using CDK11 gene-specific synthetic siRNA knockdown
endogenous CDK11. Both initial shRNA screenings and
follow-up siRNA validation assay results highlighted the
importance of CDK11 in osteosarcoma.
CDK11 is a serine/threonine protein kinase and encoded
by the CDK11 gene on chromosome 1p36.3 (9, 10). The
function of CDK11 has not been described in osteosarcoma.
There is only one CDK11 gene in mouse, whereas in
humans, there are 2 CDK11 genes that encode 2 almost
identical protein kinases. There are 3 CDK11 protein isoforms, p110, p58, and p46 (28). CDK11 p58 protein is
specifically translated from an internal ribosome entry site
and expressed only in the G2–M phase of the cell cycle.
These different CDK11 isoforms seem to play multiple roles
in transcription, RNA processing, regulating cell-cycle progression, cytokinesis, and apoptosis (10, 28). CDK11
knockout mice display an earlier phenotype and death
during the blastocyst stage of embryonic development
(29). The CDK11 null cells exhibit proliferative defects,
mitotic arrest, and apoptosis, thus suggesting that CDK11
kinase is critical for embryonic development and cellular
viability (29). By kinome-wide siRNA screen for Hedgehog
(Hh) regulators, CDK11 has been shown to directly participate in the Hh pathway. CDK11 is necessary and sufficient
for the activation of the Hh pathway, functioning downstream of Smo and upstream of the glioma-associated (Gli)
transcription factors (19). CDK11 is also a modulator of
autophagy in human cells (30). Although the function of
CDK11 has been examined in different model systems, its
potential role in tumors has not been fully investigated due
to a lack of data regarding CDK11 expression and tumor cell
growth. CDK11 was initially proposed as a tumor suppressor candidate gene, as the CDK11 chromosomal location
region 1p36.3 is frequently deleted or translocated in a
number of different human tumors, including neuroblastoma, breast cancer, and melanoma (31–33). However, a
study of neuroblastoma by FISH analysis excludes the
CDK11 genes as a tumor suppressor gene (34). On the
contrary, CDK11 knockdown by RNAi in HeLa cells
induces abnormal spindle assembly, mitotic arrest by
checkpoint activation, and cell death (20, 35). An unbiased high-throughput RNAi screening showed that
CDK11 is a positive modulator of the Wnt/b-catenin
pathway in colon cancer (36). Consistent with these
studies, a more recent kinome-wide siRNA study in
human multiple myeloma identified CDK11 as 1 out of
15 survival kinases (including PLK1, AKT1, and GRK6)
that was repeatedly vulnerable in myeloma cells (37).
Interestingly, a more recent RNAi lethality screening of
the druggable genome also found CDK11 is the survival
gene in multiple myeloma (21). Our results in osteosarcoma are compatible with HeLa and myeloma cells in
which CDK11 inhibition leads to decreased cell growth
and induces apoptosis.
It has been reported that treatment of the Fas-activated T
cells with a serine protease inhibitor prevented apoptotic
www.aacrjournals.org
death and led to the accumulation of CDK11 p110 isoform,
but not the CDK11 p58 isoform (22). To establish the
mechanisms of CDK11 knockdown-induced cell growth
inhibition and apoptosis in osteosarcoma cells, we examined antiapoptotic protein expression. Several antiapoptotic proteins were reduced by CDK11 knockdown suggesting
that CDK11 can control several aspects of apoptosis
signaling
Many studies have found survival-promoting kinase
genes to be highly expressed in human cancer, especially
in high-grade tumors (10, 38). For CDK11, we show that
osteosarcoma cell lines and tumors express high level of
CDK11 protein in comparison with normal osteoblasts.
Most importantly, the levels of CDK11 expression are
significantly associated with clinical outcome in osteosarcoma. Overexpression of CDK11 was correlated with poor
prognosis. Furthermore, silencing of CDK11 reduced tumor
volume in an osteosarcoma xenograft mouse model. These
in vivo studies have confirmed the anticancer effects of
CDK11 inhibition in vitro and provide a rationale for
pharmacologic investigation of CDK11 as a novel therapy
target.
Taken together, this study identified that CDK11 as
essential for osteosarcoma cell growth and survival.
Experiments are underway to understand the mechanisms
behind CDK11 signaling in human cancer. In turn, these
findings may lead to targeting CDK11 through gene
therapy or kinase-specific inhibitors in the treatment of
osteosarcoma.
Disclosure of Potential Conflicts of Interest
E. Choy is a consultant/advisory board for Amgen, Sanofi Aventis,
and BioMed Valley Discoveries, Inc. No potential conflicts of interest were
disclosed by other authors.
Authors' Contributions
Conception and design: Z. Duan, J. Zhang, X. Liu, N.S. Gray, F.J. Hornicek
Development of methodology: Z. Duan, J. Zhang, F.J. Hornicek
Acquisition of data (provided animals, acquired and managed patients,
provided facilities, etc.): Z. Duan, J. Zhang, D. Harmon, F.J. Hornicek
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Z. Duan, J. Zhang, X. Liu, H. Mankin, F.J.
Hornicek
Writing, review, and/or revision of the manuscript: Z. Duan, J. Zhang, E.
Choy, D. Harmon, P. Nielsen, H. Mankin, N.S. Gray, F.J. Hornicek
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Z. Duan, J. Zhang, F.J. Hornicek
Study supervision: Z. Duan, H. Mankin, F.J. Hornicek
Grant Support
This project was supported, in part, by grants from the Gattegno and
Wechsler funds. Support has also been provided by the Jeff Guyer Fund
and the Kenneth Stanton Fund. Dr. Duan is supported, in part, through a
grant from Sarcoma Foundation of America (SFA), a grant from National
Cancer Institute (NCI)/National Institutes of Health (NIH), UO1, CA
151452-01, and a grant from an Academic Enrichment Fund of MGH
Orthopaedics.
The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate
this fact.
Received April 7, 2012; revised June 29, 2012; accepted June 30, 2012;
published OnlineFirst July 12, 2012.
Clin Cancer Res; 18(17) September 1, 2012
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.
4587
Published OnlineFirst July 12, 2012; DOI: 10.1158/1078-0432.CCR-12-1157
Duan et al.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
4588
Geller DS, Gorlick R. Osteosarcoma: a review of diagnosis, management, and treatment strategies. Clin Adv Hematol Oncol 2010;8:
705–18.
Chou AJ, Geller DS, Gorlick R. Therapy for osteosarcoma: where do we
go from here? Paediatr Drugs 2008;10:315–27.
Bielack SS, Marina N, Ferrari S, Helman LJ, Smeland S, Whelan JS,
et al. Osteosarcoma: the same old drugs or more? J Clin Oncol
2008;26:3102–3; author reply 3104–5.
O'Day K, Gorlick R. Novel therapeutic agents for osteosarcoma. Expert
Rev Anticancer Ther 2009;9:511–23.
Gorlick R. Current concepts on the molecular biology of osteosarcoma. Cancer Treat Res 2009;152:467–78.
Zhang J, Yang PL, Gray NS. Targeting cancer with small molecule
kinase inhibitors. Nat Rev Cancer 2009;9:28–39.
Lapenna S, Giordano A. Cell cycle kinases as therapeutic targets for
cancer. Nat Rev Drug Discov 2009;8:547–66.
Krause DS, Van Etten RA. Tyrosine kinases as targets for cancer
therapy. N Engl J Med 2005;353:172–87.
Shapiro GI. Cyclin-dependent kinase pathways as targets for cancer
treatment. J Clinical Oncol 2006;24:1770–83.
Malumbres M, Harlow E, Hunt T, Hunter T, Lahti JM, Manning G, et al.
Cyclin-dependent kinases: a family portrait. Nat Cell Biol 2009;11:
1275–6.
Duan Z, Ji D, Weinstein EJ, Liu X, Susa M, Choy E, et al. Lentiviral
shRNA screen of human kinases identifies PLK1 as a potential therapeutic target for osteosarcoma. Cancer Lett 2010;293:220–9.
Yang C, Ji D, Weinstein EJ, Choy E, Hornicek FJ, Wood KB, et al. The
kinase Mirk is a potential therapeutic target in osteosarcoma. Carcinogenesis 2010;31:552–8.
Duan Z, Weinstein EJ, Ji D, Ames RY, Choy E, Mankin H, et al. Lentiviral
short hairpin RNA screen of genes associated with multidrug resistance identifies PRP-4 as a new regulator of chemoresistance in
human ovarian cancer. Mol Cancer Ther 2008;7:2377–85.
Takei Y, Kadomatsu K, Yuzawa Y, Matsuo S, Muramatsu T. A small
interfering RNA targeting vascular endothelial growth factor as cancer
therapeutics. Cancer Res 2004;64:3365–70.
Filleur S, Courtin A, Ait-Si-Ali S, Guglielmi J, Merle C, Harel-Bellan A,
et al. SiRNA-mediated inhibition of vascular endothelial growth factor
severely limits tumor resistance to antiangiogenic thrombospondin-1
and slows tumor vascularization and growth. Cancer Res 2003;63:
3919–22.
Fujii T, Saito M, Iwasaki E, Ochiya T, Takei Y, Hayashi S, et al.
Intratumor injection of small interfering RNA-targeting human papillomavirus 18 E6 and E7 successfully inhibits the growth of cervical
cancer. Int J Oncol 2006;29:541–8.
Dormoy V, Beraud C, Lindner V, Thomas L, Coquard C, Barthelmebs
M, et al. LIM-class homeobox gene Lim1, a novel oncogene in human
renal cell carcinoma. Oncogene 2011;30:1753–63.
Friedman E. The role of mirk kinase in sarcomas. Sarcoma
2011;2011:260757.
Evangelista M, Lim TY, Lee J, Parker L, Ashique A, Peterson AS, et al.
Kinome siRNA screen identifies regulators of ciliogenesis and hedgehog signal transduction. Sci Signal 2008;1:ra7.
Franck N, Montembault E, Rome P, Pascal A, Cremet JY, Giet R.
CDK11(p58) is required for centriole duplication and Plk4 recruitment
to mitotic centrosomes. PLoS One 2011;6:e14600.
Tiedemann RE, Zhu YX, Schmidt J, Shi CX, Sereduk C, Yin H, et al.
Identification of molecular vulnerabilities in human multiple myeloma
Clin Cancer Res; 18(17) September 1, 2012
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
cells by RNAi lethality screening of the druggable genome. Cancer Res
2012;72:757–68.
Feng Y, Qi W, Martinez J, Nelson MA. The cyclin-dependent kinase 11
interacts with 14–3-3 proteins. Biochem Biophys Res Commun
2005;331:1503–9.
Degenhardt Y, Lampkin T. Targeting Polo-like kinase in cancer therapy. Clin Cancer Res 2010;16:384–9.
Friedman E. Mirk/Dyrk1B in cancer. J Cell Biochem 2007;102:274–9.
Yamaguchi U, Honda K, Satow R, Kobayashi E, Nakayama R, Ichikawa
H, et al. Functional genome screen for therapeutic targets of osteosarcoma. Cancer Sci 2009;100:2268–74.
Morales AG, Brassesco MS, Pezuk JA, Oliveira JC, Montaldi AP,
Sakamoto-Hojo ET, et al. BI 2536-mediated PLK1 inhibition suppresses HOS and MG-63 osteosarcoma cell line growth and clonogenicity. Anticancer Drugs 2011;22:995–1001.
Liu X, Choy E, Hornicek FJ, Yang S, Yang C, Harmon D, et al. ROCK1 as
a potential therapeutic target in osteosarcoma. J Orthop Res
2011;29:1259–66.
Loyer P, Busson A, Trembley JH, Hyle J, Grenet J, Zhao W, et al. The
RNA binding motif protein 15B (RBM15B/OTT3) is a functional competitor of serine-arginine (SR) proteins and antagonizes the positive
effect of the CDK11p110-cyclin L2alpha complex on splicing. J Biol
Chem 2011;286:147–59.
Li T, Inoue A, Lahti JM, Kidd VJ. Failure to proliferate and mitotic arrest
of CDK11(p110/p58)-null mutant mice at the blastocyst stage of
embryonic cell development. Mol Cell Biol 2004;24:3188–97.
Wilkinson S, Croft DR, O'Prey J, Meedendorp A, O'Prey M, Dufes C,
et al. The cyclin-dependent kinase PITSLRE/CDK11 is required for
successful autophagy. Autophagy 2011;7:1295–301.
Feng Y, Shi J, Goldstein AM, Tucker MA, Nelson MA. Analysis of
mutations and identification of several polymorphisms in the putative
promoter region of the P34CDC2-related CDC2L1 gene located at
1P36 in melanoma cell lines and melanoma families. Int J Cancer
2002;99:834–8.
Martinsson T, Sjoberg RM, Hallstensson K, Nordling M, Hedborg F,
Kogner P. Delimitation of a critical tumour suppressor region at distal
1p in neuroblastoma tumours. Eur J Cancer 1997;33:1997–2001.
Dave BJ, Pickering DL, Hess MM, Weisenburger DD, Armitage JO,
Sanger WG. Deletion of cell division cycle 2-like 1 gene locus on 1p36
in non-Hodgkin lymphoma. Cancer Genet Cytogenet 1999;108:
120–6.
White PS, Fujimori M, Marshall HN, Kaufman BA, Brodeur GM. Characterization of the region of consistent deletion within 1p36 in neuroblastomas. Prog Clin Biol Res 1994;385:3–9.
Petretti C, Savoian M, Montembault E, Glover DM, Prigent C, Giet
R. The PITSLRE/CDK11p58 protein kinase promotes centrosome
maturation and bipolar spindle formation. EMBO Rep 2006;7:
418–24.
Naik S, Dothager RS, Marasa J, Lewis CL, Piwnica-Worms D. Vascular
endothelial growth factor receptor-1 is synthetic lethal to aberrant
{beta}-catenin activation in colon cancer. Clin Cancer Res 2009;15:
7529–37.
Tiedemann RE, Zhu YX, Schmidt J, Yin H, Shi CX, Que Q, et al. Kinomewide RNAi studies in human multiple myeloma identify vulnerable
kinase targets, including a lymphoid-restricted kinase, GRK6. Blood
2010;115:1594–604.
Cicenas J, Valius M. The CDK inhibitors in cancer research and
therapy. J Cancer Res Clin Oncol 2011;137:1409–18.
Clinical Cancer Research
Downloaded from clincancerres.aacrjournals.org on May 23, 2014. © 2012 American Association for Cancer Research.