Cancer

Molecular and Cellular Pathobiology Research

miRNA-708 Control of CD44þ Prostate Cancer–Initiating

Cells
Sharanjot Saini, Shahana Majid, Varahram Shahryari, Sumit Arora, Soichiro Yamamura, Inik Chang,
Mohd Saif Zaman, Guoren Deng, Yuichiro Tanaka, and Rajvir Dahiya

Abstract
Tumor recurrence in prostate cancer has been attributed to the presence of CD44-expressing tumor-
initiating cells. In this study, we report that miR-708 is a key negative regulator of this CD44þ subpopulation
of prostate cancer cells, with important implications for diagnosis and prognosis of this disease. miR-708 was
underexpressed in CD44þ cells from prostate cancer xenografts. Reconstitution of miR-708 in prostate cancer
cell lines or CD44þ prostate cancer cells led to decreased tumorigenicity in vitro. Intratumoral delivery of
synthetic miR-708 oligonucleotides triggered regression of established tumors in a murine xenograft model of

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

human prostate cancer. Conversely, miR-708 silencing in a purified CD44 population of prostate cancer cells
promoted tumor growth. Functional studies validated CD44 to be a direct target of miR-708 and also
identified the serine/threonine kinase AKT2 as an additional target. Clinically, low miR-708 expression was
associated significantly with poor survival outcome, tumor progression, and recurrence in patients with
prostate cancer. Together, our findings suggest that reduced miR-708 expression leads to prostate cancer
initiation, progression, and development by regulating the expression of CD44 as well as AKT2. miR-708
therefore may represent a novel therapeutic target or diagnostic and prognostic biomarker in prostate
cancer. Cancer Res; 72(14); 3618–30.
2012 AACR.

Introduction suffer from disease recurrence, as monitored by rising PSA

Despite many advances, clinical management of prostate
cancer remains a major challenge and this disease continues
to represent a leading cause of cancer-related morbidity and
mortality (1). A major challenge in prostate cancer diagnosis
is posed by the inability of current diagnostic methods,
including prostate-specific antigen (PSA) screening and
histopathologic grading, to distinguish between indolent
and aggressive tumors. Recent changes in recommendations
suggest later and less frequent PSA screenings (2, 3). In view
of this, there is an urgent need to identify alternate prostate
disease biomarkers with better prognostic and diagnostic
potential. Another challenging aspect of prostate cancer is
the high rates of recurrence associated with the disease.
About 40% of men with localized prostate cancer suffer from
relapse after initial therapy (4). In these cases, androgen
ablation therapy is used to reduce tumor burden/circulating
PSA to low or undetectable limits. However, most patients
Authors' Affiliation: Department of Urology, Veterans Affairs Medical
Center and University of California, San Francisco, California
Note: Supplementary data for this article are available at Cancer Research
Online (http://cancerres.aacrjournals.org/).
Corresponding Author: Rajvir Dahiya, Urology Research Center (112F),
Veteran Affairs Medical Center and University of California, San Francisco,
4150 Clement Street, San Francisco, CA 94121. Phone: 415-750-6964;
Fax: 415-750-6639; E-mail: rdahiya@urology.ucsf.edu
doi: 10.1158/0008-5472.CAN-12-0540

2012 American Association for Cancer Research.
and tumor progression to hormone-refractory/castration-
resistant prostate cancer (1, 5) that is essentially untreatable
(1, 6). Therefore, a second major clinical challenge is the
elucidation of pathways of tumor recurrence and metastasis
of prostate cancer, which could lead to the design of better
therapeutic strategies.
Tumor recurrence and progression in prostate cancer has
been associated with the existence of tumor-initiating cells
(TIC) or prostate cancer stem cells within the bulk of the tumor
that are refractory to current therapies (7). Several putative
TICs have been identified in human prostate cancer and are
characterized by various cell surface markers (8–12). Cell
adhesion molecule CD44 has been identified as a predominant
cell surface marker of prostate cancer TICs (7–12). It has been
reported that CD44þ prostate cancer cells are more prolifer-
ative, clonogenic, tumorigenic, and metastatic than the iso-
genic CD44 prostate cancer cells (10). Here, we report that
CD44þ prostate cancer cells are regulated by a miRNA (miR),
miR-708. Analysis of an array of human prostate cancer speci-
mens showed the widespread attenuation of miR-708 expres-
sion, suggesting that miR-708 downregulation is a common
event in prostate cancer. We also found that this molecular
alteration has significant diagnostic and prognostic potential.
Consistent with the functional role of miR-708 in regulating
prostate cancer TIC marker CD44, a positive correlation was
found between reduced miR-708 expression and tumor pro-
gression, biochemical recurrence, and poor survival outcome
in prostate cancer.

3618 Cancer Res; 72(14) July 15, 2012


Materials and Methods
Cell lines and cell culture
Nonmalignant epithelial prostate cell lines (RWPE-1 and
PWR-1E) and prostate carcinoma cell lines (LNCaP, Du145,
PC3, VCaP, MDA-PCa-2b) were obtained from the American
Type Culture Collection (ATCC) and cultured under recom-
mended conditions as described previously (ref. 13; also
described in Supplementary Methods). These human-derived
cell lines were authenticated by DNA short tandem repeat
analysis by ATCC. The experiments with cell lines were con-
ducted within 6 months of their procurement/resuscitation.
Prostate cancer xenografts
Xenograft human prostate tumors LAPC-4 and LAPC-9 were
courtesy of R. Reiter (14, 15), PC3 xenograft tumors were
established using early-passage cells and maintained in non-
obese diabetic/severe combined immunodeficient (NOD/
SCID) mice. NOD/SCID mice (Charles River Laboratories) were
maintained under standard conditions. All animal care was in
accordance with the guidelines of the San Francisco Veterans
Affairs Medical Center (VAMC) and the study was approved by
the San Francisco Veterans Affairs Institutional Animal Care
and Use Committee.
Purification of CD44þ subpopulations of xenograft
tumors
Basic procedures previously described were followed
(10, 16, 17). Single-cell suspensions obtained from xenograft
tumors were stained for CD44 with human-specific fluorescein
isothiocyanate (FITC)-conjugated anti-CD44 antibody (Milte-
nyi Biotech, 130-095-195) as per manufacturer's protocol.
Stained cells were purified into CD44þ and CD44 cells with
fluorescence-activated cell sorting (FACS). Post-sort analysis
revealed purities of both populations being more than 95%.
miRNA/siRNA transfections
Cells were plated in growth medium without antibiotics for
approximately 24 hours before transfections. Transient trans-
fections of miRNA precursor (Ambion)/siRNA (Origene) were
carried out using Lipofectamine 2000 (Invitrogen) according to
the manufacturer's protocol. All miRNA/siRNA transfections
were done for 72 hours.
Tissue samples
Formalin-fixed, paraffin-embedded prostate cancer sam-
ples were obtained from the San Francisco VAMC. Written
informed consent was obtained from all patients and the
study was approved by the University of California San
Francisco Committee on Human Research. All slides were
reviewed by a board-certified pathologist for the identifica-
tion of prostate cancer foci as well as adjacent normal
glandular epithelium.
Migration, invasion, and clonogenicity assays
CytoSelect Cell Migration and Invasion Assay Kit (Cell
Biolabs, Inc.) was used for migration and invasion assays,
according to the manufacturer's protocol. Briefly, 48 hours
www.aacrjournals.org
miR-708 Regulates CD44 in Prostate Cancer
after transfection, cells were counted and placed on control
inserts or Matrigel inserts at 1  105 cells/mL in serum-free
medium and were allowed to migrate for 20 hours at 37 C. Cells
were removed from the top of the inserts and cells that
migrated/invaded though the polycarbonate basement mem-
brane were fixed, stained, and quantified at optical density of
560 nm after extraction. For clonogenicity assay, cells were
counted, seeded at low density (1,000 cells/plate) and allowed
to grow until visible colonies appeared. Then, cells were stained
with Giemsa, and colonies were counted.
Apoptosis assays
FACS analysis for apoptosis was done 72 hours after trans-
fection, using Annexin V-FITC/7-AAD Kit (Beckman Coulter,
Inc.), according to the manufacturer's protocol. Stained cells
were immediately analyzed with a flow cytometer (Cell Lab
Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022
Quanta SC; Beckman Coulter, Inc).
Luciferase assays
For CD44 and AKT2, the corresponding 30 -untranslated
region (UTR) reporter constructs (catalog no., HmiT022972,
HmiT004350-MT01) were obtained from GeneCopoeia along
with control construct (CmiT000001-MT01). All these target
clones were in SV-40 promoter–based vector pEZX-MT01
(GeneCopoeia). Control constructs and various 30 -UTR report-
er constructs (0.2 ug) were cotransfected into PC3/LNCaP cells
cultured in 24-well plates along with 50 nmol/L miR-708 or
miR-CON (Ambion) using Lipofectamine 2000 (Invitrogen). 48
hours after transfection, firefly and Renilla luciferase activities
were measured by using the Dual Luciferase Reporter Assay
System (Promega) according to the manufacturer's protocol.
Firefly luciferase was normalized to Renilla luciferase activity.
In vivo intratumoral delivery of miR-708
We examined the antitumor effects of miR-708 by local
administration in established tumors. Nude mice (4- to 5-
week-old, Charles River Laboratories; n ¼ 12) received
subcutaneous injections of 3  106 PC3 cells in the right
flank area in a volume of 100 mL. Once palpable tumors
developed, caliper measurements were taken twice a week
and tumor volume was calculated on the basis of width (x)
and length (y): x2y/2, where x < y. When tumors reached an
average volume of 100 to 150 mm3, 6.25 mg of synthetic
miRNA (miR-708/miR-CON) complexed with 1.6 mL siPORT
Amine transfection reagent (Ambion) in 50 mL PBS was
delivered intratumorally in 3-day intervals. The dosage was
selected on the basis of previous studies done by our group
and others (18, 19). Synthetic miRNAs are double-stranded,
ready-to-use miRNA mimics and were purchased from
Ambion (pre-miR, cat. no. AM17100). Mice were killed 2
days after the last treatment (day 58), tumors were collected,
and total RNA and protein was extracted.
Statistics
All quantified data represent an average of at least triplicate
samples or as indicated. Data are represented as mean  SEM.
All statistical analyses were conducted using StatView (version
5; SAS Institute Inc.) and Medcalc version 10.3.2. Two-tailed
Cancer Res; 72(14) July 15, 2012 3619
 Saini et al.

A CD44 3′ UTR
1 3066
Site 1 Site 2

CD44 3′ UTR 1521: 5′… UAUCUCCUUUCUGAGGCUCCUAC … 3′ 1606: 5′… AGAAAGAAGAAGAAAAGCUCCUG… 3′

:
hsa-miR-708 3′ GGGUCGAUCUAACA UUCGAGGAA 5′ 3′ GGGUCGAUCUAACA UUCGAGGAA 5′

B 1.2
Figure 1. CD44 is a functional target
of miR-708 in prostate cancer. A,
Relative miR-708 expression

1
schematic representation of CD44
0.8 30 -UTR showing the relative
P = 0.0098 positions of putative miR-708
0.6 CD44–
target sites. B, relative miR-708
P = 0.0009 CD44+ 
P = 0.0162 levels in purified CD44 and
0.4 CD44þ subpopulations of prostate
cancer xenografts (PC3 and
0.2 P = 0.0016
P = 0.0016 P = 0.0007 LAPC9). Data were normalized to

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

RNU48 control and are
0
1 2 3 4 5 6 represented as mean  SEM. C,
PC3 xenografts LAPC9 xenografts relative miR-708 expression in
C PC3 LNCaP
PC3/LNCaP cells transfected with
control miR/miR-708/mock-
P = 0.0003 P = 0.00009 PC3 LNCaP transfected cells as assessed by
300 P = 0.0002 P = 0.0001
18,000

miR-CON

miR-CON
RT-PCR (left). Immunoblots for
Relative miR-708 expression

miR-708

miR-708
16,000
250 endogenous CD44 protein in
Mock

Mock
14,000
transfected PC3/LNCaP cells
200 12,000
(right). Glyceraldehyde-3-
10,000
150 phosphate dehydrogenase
8,000 CD44
100
(GAPDH) was used a loading
6,000
control. D, CD44 30 -UTR construct
50 4,000
GAPDH or the control construct was
2,000
0 cotransfected into PC3/LNCaP
0
Mock miR-CON miR-708 Mock miR-CON miR-708 cells along with miR-708/miR-
CON/mock-treated cells and
D PC3 LNCaP assayed for luciferase activity.
Control CD44 3′UTR Firefly luciferase values were
1.4 1.4 normalized to Renilla luciferase
activity and plotted as relative
1.2 1.2
Relative luciferase activity

luciferase activity
1 1 ( , P < 0.05).

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
Mock miR-CON miR-708 Mock miR-CON miR-708

Student t test was used for comparisons between groups. algorithms that predict the mRNA targets of a miRNA, miRAN-
Results were considered statistically significant at P  0.05. DA (21) and TargetScan (22). From these analyses, miR-708
Full Methods are available in the Supplementary Materials stood out for the presence of 2 binding sites, suggesting
and Methods. cooperative binding and interaction (Fig. 1A). To test the
potential regulation of CD44 by miR-708, we purified CD44þ
and CD44 subpopulations of prostate cancer cells from 2
Results prostate cancer xenograft mouse models (PC3 and LAPC9)
CD44 is a functional target of miR-708 in prostate cancer followed by miR-708 expression profiling (Fig. 1B). miR-708
Because CD44 is an important marker that determines expression was significantly downregulated in CD44þ versus
tumorigenic behavior of prostate cancer cells, it is essential CD44 cells suggesting that miR-708 regulates the CD44þ
to elucidate regulatory mechanisms that converge on this subpopulation in prostate cancer. To validate this further, we
molecule. miRNAs constrain gene expression by binding to overexpressed miR-708 in prostate cancer cell lines (PC3 and
the 30 -UTRs of cognate mRNA targets (20). To explore the LNCaP). Transient transfection of miR-708 precursor led to
potential regulation of this TIC marker by miRNAs, we used 2 overexpression of miR-708 as determined by real-time PCR

3620 Cancer Res; 72(14) July 15, 2012 Cancer Research

 miR-708 Regulates CD44 in Prostate Cancer

(RT-PCR; Fig. 1C, left). In both the cell lines, CD44 protein was miR-708 expression is widely attenuated in prostate
specifically downregulated in miR-708–expressing cells (Fig. cancer
1C, right) suggesting that miR-708 regulates CD44. To validate To evaluate the role of miR-708 in prostate cancer, miR-708
CD44 as a miR-708 target, we conducted luciferase reporter expression was assayed in human prostate cell lines, which
assays with CD44 30 -UTR in miR-708/miR-CON/mock-trans- included normal epithelial prostate cell lines (RWPE-1 and
fected PC3/LNCaP cells (Fig. 1D). We observed a significant PWR-1E) and prostate carcinoma cell lines (LNCaP, LAPC4,
decrease in luciferase activity upon miR-708 transfection, Du145, PC3, VCaP, MDA-PCa-2b, LAPC9). Relative miR-708
suggesting that miR-708 represses CD44 directly. Collectively, expression was specifically attenuated in all the examined
these data indicate that CD44 is a direct target of miR-708 in prostate carcinoma cell lines compared with normal epithelial
prostate cancer. prostate cell lines (Fig. 2A). To examine the clinical relevance of

A C P = 0.0002

1.2
Relative miR-708 expression

1.6
1
P = 0.0623

Relative miR-708 expression score

1.4

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

0.8

1.2
0.6
Figure 2. miR-708 expression is 1
0.4
widely attenuated in prostate cancer.
0.8
A, qRT-PCR analysis of relative miR- 0.2 P = 0.0003
P = 0.0001 P = 0.0001 P = 0.0002 P = 0.0001 P = 0.0001 P = 0.0005
708 expression levels in normal 0
0.6
epithelial prostate cell lines (RWPE-1 0.4
and PWR-1E) and prostate
100

carcinoma cell lines (LNCaP, LAPC4, 0.2

Du145, PC3, VCaP, MDA-PCa-2b,

LAPC9). Data were normalized to
B 0
Normal Tumor
10 (n = 8) (n = 40)
RNU48 control and are represented
as mean  SEM. B, relative
miR-708 expression levels in Patient
microdissected prostate cancer no. 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

tissues and patient-matched

Relative miR-708 expression

adjacent normal regions as assessed

0.1
by RT-PCR. Table below
summarizes the relative miR-708
expression levels in these
Relative miR-708 expression N (%)
specimens. C, in situ hybridization
0.01
(compared with matched normal)

analysis of relative miR-708 Low (<0.75) 67/96 (70)

expression levels in prostate cancer
0.001 No change (0.75–1.25) 13/96 (13)
tissues and normal prostate tissues.
High (>1.25) 16/96 (17)
Tissues were hybridized with
digoxigenin-labeled locked nucleic
acid–based probes for miR-708 and D E miR-708

U6 (control). miR-708 expression Relative miR-708 expression

Total
Low No change High P-value
was scored and normalized to U6 n (%)
AUC: 0.937
Age, y
scores. The average value for each 40–49 5/134 (4) 3/5 (60) 1/5 (20) 1/5 (20)
95% CI: 0.901– 0.963
P value < 0.0001*
group is represented by the
Sensitivity

50–59 39/134 (29) 29/39 (74) 5/39 (13) 5/39 (13)

0.6557
horizontal line. D, correlation of miR- 60–69 64/134 (48) 51/64 (80) 6/64 (9) 7/64 (11)

708 expression with 70-79 25/134 (19) 18/25 (72) 4/25 (16) 3/25 (12)
80–89 1/134 (1) 1/1 (100) - -
clinicopathologic characteristics of
Gleason score
patients with prostate cancer. E, 4-6 46/129 (36) 29/46 (63) 7/46 (15) 10/46 (22)
ROC curve analysis showing the 7 53/129 (41) 39/53 (74) 9/53 (17) 5/53 (9) 0.0098*

ability of miR-708 expression to 8-10 30/129 (23) 27/30 (90) 2/30 (7) 1/30 (3)
Pathologic stage
discriminate between malignant and pT2 71/125 (57) 51/71 (72) 8/71 (11) 12/71 (17)
100-Specificity

nonmalignant tissue samples. pT3 52/125 (42) 43/52 (83) 5/52 (10) 4/52 (8)
0.1095 F Remission
F, Kaplan–Meier survival curves for pT4 2/125 (2) 2/2 (100) - -
patients with prostate cancer, Biochemical recurrence
HR: 6
stratified on the basis of miR-708 Yes 22/92 (24) 18/22 (82) 3/22 (14) 1/22 (5) 0.0138*
Survival probability (%)

95% CI: 2.2 –16.4

levels. P value is based on a No 70/92 (76) 49/70 (70) 9/70 (13) 12/70 (17) P value: 0.0223*

log-rank test. ( , P < 0.05). Group

Time (mo)

www.aacrjournals.org Cancer Res; 72(14) July 15, 2012 3621

 Saini et al.
this finding, we extended our analysis to human clinical
prostate samples (N ¼ 136; Fig. 2B and C, Supplementary Fig.
S1). Clinicopathologic characteristics of the patients are sum-
marized in Supplementary Table S1. We examined the expres-
sion levels of miR-708 in laser capture microdissected (LCM)
prostate cancer tissues (n ¼ 96) and matched adjacent normal
regions by RT-PCR (Fig. 2B). While the expression of miR-708
was unaltered in 13 of 96 cases (13%) and higher in 16 of 96
cases (17%), a major fraction of tissue samples (67 of 96, 70%)
showed lower miR-708 levels relative to matched normal
tissues. The differences were statistically significant with the
Wilcoxon signed rank test (P < 0.0001). We extended our
analysis of clinical specimens by assessing miR-708 levels in
40 additional cases of prostate adenocarcinoma, 8 of which had
matched normal adjacent tissue by in situ hybridization anal-
ysis and again observed attenuated expression of miR-708 in
cancer tissues compared with normal tissues (Fig. 2C, Supple-
mentary Fig. S1). This confirms that miR-708 is widely down-
regulated in prostate cancer and suggests that miR-708 is a
potential tumor suppressor.
Downregulation of miR-708 expression is associated
with tumor progression and biochemical recurrence in
prostate cancer
We also looked to see whether miR-708 expression in clinical
tissues correlated with clinicopathologic characteristics such
as age, Gleason score, pathologic stage, and biochemical
recurrence of prostate cancer (Fig. 2D). While there was no
significant correlation with age, decreased miR-708 expression
was observed in 63% of cases of low Gleason score (4–6), 74% of
cases of Gleason score 7, and in 90% of cases of higher Gleason
score (8–10). This trend suggests that miR-708 expression tends
to attenuate in higher grade prostate cancer (P ¼ 0.0098).
Similarly, decreased miR-708 expression was observed in
72% of cases of pathologic stage pT2, 83% of cases of pT3, and
100% of pT4 cases, though no statistically significant correla-
tion was observed between miR-708 expression and pathologic
stage. Importantly, miR-708 was specifically attenuated in 18 of
22 cases (82%) of PSA failure within this cohort of samples
suggesting that downregulation of miR-708 is associated with
biochemical recurrence.
Furthermore, we stratified the tissue cohort, based on age
and PSA and used age-adjusted PSA ranges to examine the
correlation of miR-708 expression with serum PSA levels
(Supplementary Table S2). Within our dataset, 23 of 85
(21%) samples had age-adjusted PSA within the normal range.
Analysis of miR-708 expression in these cases showed that 16 of
23 cases (70%) exhibited attenuated miR-708 expression
suggesting that miR-708 expression has a better disease pre-
dictive value than PSA in these cases.
miR-708 as a diagnostic and prognostic marker in
prostate cancer
In view of the observed widespread downregulation of miR-
708 in prostate cancer cell lines and clinical malignancies, we
evaluated the potential clinical significance of miR-708 expres-
sion. To evaluate the potential capability of miR-708 as a
diagnostic biomarker for prostate cancer, we conducted
3622 Cancer Res; 72(14) July 15, 2012
receiver operating characteristic (ROC) analyses on the cohort
of clinical samples (Fig. 2E). ROC analyses showed that miR-
708 expression can be a single significant parameter to dis-
criminate between normal and tumor tissues with an area
under the ROC curve (AUC) of 0.937 [95% confidence interval
(CI), 0.901–0.963; P < 0.0001]. Furthermore, Kaplan–Meier
survival analysis for patients with prostate cancer, stratified
on the basis of miR-708 levels, showed that survival was
significantly reduced in patients with low miR-708 expression
(Fig. 2F). The HR between the cases with low or high miR-708
expression was 6 with a 95% CI from 2.2 to 16.4 and with an
associated P value of 0.0223. This analysis suggests that miR-
708 has significant potential to be used as a diagnostic and
prognostic marker for prostate cancer.
miR-708 reexpression suppresses tumorigenicity in vitro
Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022
To assess the tumor-suppressive role of miR-708, we reex-
pressed miR-708 in both prostate cancer cell lines (PC3 and
LNCaP; Fig. 1C, left) followed by functional assays (Fig. 3). A
significant decrease in cell viability was observed over time in
PC3/LNCaP cells overexpressing miR-708 (Fig. 3A) as com-
pared with cells expressing control miR (miR-CON). miR-708
reexpression decreased the clonogenicity (Fig. 3B), migration
(Fig. 3C), and invasiveness (Fig. 3D) of PC3/LNCaP cells.
Reexpression of miR-708 also led to marked morphologic
changes in both the cell lines (Supplementary Fig. S2). Specif-
ically, there was a decrease in the fraction of elongated, spindle-
shaped cells paralleled by an increase in rounded, apoptotic
cells. Apoptosis assay (Fig. 3E–G) showed that the average
apoptotic cell fractions were significantly increased upon miR-
708 reexpression compared with miR-CON/mock-transfected
cells with a concomitant decrease in the viable cell population.
As shown in Fig. 3G, miR-708 introduction led to a 3.15-fold and
3.73-fold increase in apoptosis in PC3 and LNCaP cells, respec-
tively, relative to mock/miR-CON controls. These observations
suggest that miR-708 reexpression suppresses the tumorige-
nicity of prostate cancer cells in vitro.
Intratumoral delivery of miR-708 leads to tumor
regression in prostate cancer xenografts and attenuates
CD44 expression
We further examined the therapeutic potential of synthetic
miR-708 mimics in vivo in a mouse prostate cancer xenograft
model. Toward this, we subcutaneously injected PC3 cells into
nude mice and maintained them until the animals developed
solid, palpable tumors, following which miR-708 or miR-CON
was administered intratumorally every 3 days. Intratumoral
delivery of synthetic miR-708 induced a specific inhibitory
response and significantly inhibited tumor growth compared
with control mice (Fig. 4A and B). To correlate the therapeutic
response with delivery of miR-708, RNA was extracted from
harvested control (n ¼ 2) or miR-708 tumors (n ¼ 5), and miR-
708 expression was assessed by quantitative RT-PCR (qRT-
PCR; Fig. 4C). Tumors injected with miR-708 mimic contained
significantly more miR-708 than control tumors. Furthermore,
as our in vitro data suggested that CD44 is a direct functional
target of miR-708, we surmised that if biologically meaningful,
this mechanism might also be found in these xenograft tumor
Cancer Research
 miR-708 Regulates CD44 in Prostate Cancer

A B PC3 LNCaP
miR-CON miR-708 miR-CON miR-708
PC3 LNCaP
140
Mock
Relative cell viability (%)

120
120 miR-CON
miR-708 100
100

80 80
300 P = 0.0368 P = 0.0025
60

Average colony number

60 300
250 P = 0.0169 P = 0.0027
40 40 250
200
20 20 200
150
0 0
150
100
Days 0 1 2 3 0 1 2 3 100
50
50

C miR-CON
PC3
miR-708 miR-CON
LNCaP
miR-708
0
Mock miR-CON miR-708
0
Mock miR-CON miR-708

D PC3 LNCaP
miR-CON miR-708 miR-CON miR-708

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

P = 0.0399 P = 0.0110
1.4
P = 0.0011 0.9 P = 0.0209
Migration (OD 560 nm)

1.2
0.8
1 0.7 P = 0.0039 P = 0.0183
1.2
P = 0.0099 P = 0.0046
Invasion (OD 560 nm)
0.6 0.6
0.8
0.5 1
0.5
0.6 0.4
0.8
0.4 0.3 0.4
0.2 0.6
0.2 0.3
0.1
0 0.4
0 0.2
Mock miR-CON miR-708 Mock miR-CON miR-708
0.2 0.1

E miR-CON miR-708 0
Mock miR-CON miR-708
0
Mock miR-CON miR-708
% Apoptotic 5±2 % Apoptotic 23 ± 2
% Early apoptotic 4±1 % Early apoptotic 8±3
% Viables 90 ± 3 % Viables 66 ± 5
% Dead 2±1 % Dead 3±1
G
Mock miR-CON miR-708
P = 0.0170
P = 0.0028 P = 0.0140
Apoptosis (relative values)

4
P = 0.0009
3.5
3
2.5
2

F miR-CON miR-708
1.5
1
% Apoptotic 4±2 % Apoptotic 13 ± 2 0.5
% Early apoptotic 4±1 % Early apoptotic 16 ± 4 0
% Viables 91 ± 4 % Viables 68 ± 2 PC3 LNCaP
% Dead 2±1 % Dead 3±1
Total cells Total cells

Figure 3. Restoration of miR-708 expression suppresses tumorigenicity in vitro. To evaluate the role of miR-708 in prostate cancer, miR-708 precursor was
reintroduced in PC3/LNCaP cell lines by transient transfections followed by functional assays after 72 hours. Cell viability assay (A), colony formation assay (B),
Transwell migration assay (C), and invasion assay (D) with PC3/LNCaP cells mock-transfected/transfected with miR-CON or miR-708. E, apoptosis assay in
PC3 and LNCaP (F) cells. Seventy-two hours after transfection, cells were stained with Annexin V-FITC/7-AAD and immediately analyzed with a flow
cytometer. G, representation of relative apoptotic fractions (early þ late apoptotic) in each group for PC3/LNCaP cells. Apoptosis induced by miR-CON/miR-
708 was normalized to mock control ( , P < 0.05).

www.aacrjournals.org Cancer Res; 72(14) July 15, 2012 3623

 Saini et al.

A B
miR-CON miR-708 miR-CON miR-708

1,200

1,000
Tumor volume (mm3)

800

600

400

200

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

Day 30 34 38 42 46 50 54 58

C P = 0.0001 D
2,000 P = 0.0003
Relative miR-708 expression

1,800
P = 0.0006

Control
1,600
miR-708
P = 0.0002
1,400
T1 T2 T3 T4 T5
1,200
1,000
800 CD44
P = 0.0008
600
400
200
P = 0.1000 GAPDH
0
Sample C1 C2 T1 T2 T3 T4 T5
miR-CON miR-708

Figure 4. Intratumoral delivery of miR-708 leads to regression of tumors in a prostate cancer xenograft mouse model and attenuates CD44 expression. A, PC3
cells were subcutaneously injected into nude mice to form solid, palpable tumors (day 30), following which synthetic miR-708/control miR were intratumorally
delivered every 3 days for 4 weeks. Tumor volumes following miR-708 administration were significantly reduced ( , P < 0.05). B, representative
images of mice from the 2 groups on day 58 are shown. C, relative miR-708 expression in prostate cancer xenografts as assessed by RT-PCR. D, CD44 protein
levels in prostate cancer xenografts as assessed by immunoblot analysis.

models. To validate this, we extracted protein from these assay showed that apoptotic cell fractions were significantly
xenografts and gauged whether intratumoral administration increased upon CD44 knockdown compared with control, an
of miR-708 altered CD44 expression (Fig. 4D). Indeed, CD44 effect similar to that observed upon miR-708 reintroduction in
protein levels were repressed in miR-708 injected xenografts PC3 cells (Fig. 5E and F). These results suggest that CD44 is an
compared with the control tumors. This finding reinforces the important determinant of the tumorigenic properties of pros-
idea that CD44 is a direct target of miR-708. tate cancer cells and that CD44 inhibition partially pheno-
copies the antitumorigenic effects of miR-708 in prostate
CD44 knockdown partially phenocopies miR-708 cancer. Furthermore, in view of our present results showing
reexpression in prostate cancer cells that CD44þ cells from prostate cancer xenografts presented
To determine whether CD44 is a functionally relevant target lower miR-708 expression, we also examined whether CD44
of miR-708 in prostate cancer, we inhibited CD44 expression knockdown in PC3 cells altered miR-708 expression (Fig. 5G).
using siRNA to see whether CD44 knockdown functionally Interestingly, siRNA-mediated CD44 knockdown led to aug-
mimics the effects of miR-708 overexpression in prostate mentation of miR-708 expression. This observation reinforces
cancer. We treated PC3 cells with CD44 siRNA followed by in the regulatory interplay between miR-708 and CD44.
vitro functional assays. We initially tested 3 sets of siRNA
against CD44 to achieve efficient knockdown, as assessed by miR-708 inhibits tumor-initiating capacity of prostate
RT-PCR and immunoblot analysis (Fig. 5A). si1 produced the cancer cells in vitro
most efficient knockdown and was used in subsequent experi- Because CD44 has been identified as a major molecular
ments. siRNA inhibition of CD44 led to decreased cell viability, determinant of prostate cancer TICs (7–12), and in view of our
migration, and invasion of PC3 cells (Fig. 5B–D). Apoptosis present results suggesting that miR-708 regulates CD44, we

3624 Cancer Res; 72(14) July 15, 2012 Cancer Research

 miR-708 Regulates CD44 in Prostate Cancer

A B

NS siRNA

siRNA
CD44
NS siRNA CD44 siRNA
120

Relative cell viability (%)

1 2 3
100
1.2
CD44 80
Relative CD44 expression

1
60

0.8 GAPDH 40

0.6 20

0
0.4
Days 0 1 2 3
0.2

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

NS siRNA 1 2 3
CD44 siRNA
C D NS siRNA CD44 siRNA
NS siRNA CD44 siRNA

F
1.2 1.2 25

1 1
P = 0.0051
Relative migration

20
P = 0.011
Relative invasion

0.8 0.8 % Apoptosis

0.6 0.6 15
P = 0.014
0.4 0.4
10
0.2 0.2

0 0
5
NS siRNA CD44 siRNA NS siRNA CD44 siRNA
0
E NS siRNA CD44 siRNA
NS siRNA CD44 siRNA
% Apoptotic 3±1
G
% Apoptotic 14 ± 1 P = 0.0009
% Early apoptotic 4±1 45
% Early apoptotic 5±1
Relative miR-708 expression

% Viables 93 ± 2 % Viables 81 ± 2 40
% Dead 0.2 ± 0.1 % Dead 0.3 ± 0.1 35
Total cells Total cells
30
25
20
15
10
5
0
NS siRNA CD44 siRNA

Figure 5. CD44 knockdown partially phenocopies miR-708 reexpression in prostate cancer cells. PC3 cells were transfected with CD44 siRNA/nonspecific
(NS) control siRNA for 72 hours, followed by various assays. A, relative CD44 mRNA expression and protein expression after siRNA transfections as
assessed by RT-PCR and immunoblotting, respectively. Cell viability assay (B), Transwell migration assay (C), invasion assay (D), and apoptosis assay (E) in
PC3 cells after NS siRNA (left) or CD44 siRNA (right) treatment. F, representation of average apoptotic fractions in each group ( , P < 0.05). G, relative miR-708
levels after siRNA transfections as assessed by RT-PCR. Data were normalized to RNU48 control and are represented as mean  SEM.

www.aacrjournals.org Cancer Res; 72(14) July 15, 2012 3625

 Saini et al.

A CD44+ cells B CD44+ cells

3,000 P = 0.0003 miR-CON miR-708
Relative miR-708 expression

2,500

2,000

1,500

1,000

500 D
miR-CON miR-708
0
miR-CON miR-708
C 35
Figure 6. miR-708 inhibits tumor-
initiating capacity of prostate
18 Colonies (average no.) 30 þ
cancer cells in vitro. CD44 and
16 CD44 subpopulations were
Spheres (average no.)

25

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

14 purified from PC3 xenografts. miR-
12 20 708 was overexpressed in CD44þ
10 P = 0.0102 cells (A–D) and inhibited in CD44
15
cells (E–G), followed by in vitro
8 P = 0.0359
10 assays. A, validation of miR-708
6 overexpression in CD44þ cells by
4 5 qRT-PCR. B, phase contrast
2 0 images of CD44þ cells transfected
miR-CON miR-708 with miR-CON/miR-708. Sphere
0
miR-CON miR-708 formation assay (C) and
clonogenicity assays (D) in miR-
E F CON/miR-708–transfected CD44þ
CD44– cells CD44– cells
cells. E, assessment of miR-708
Anti-miR-CON Anti- miR-708 inhibition in purified CD44 PC3
1.2
Relative miR-708 expression

cells by qRT-PCR. F, phase

1 contrast images of CD44 cells
transfected with anti-miR-CON/
0.8 anti-miR-708 oligos. G, sphere
formation assay in anti-miR-CON/
0.6
anti-miR-708–transfected CD44
cells ( , P < 0.05).
0.4

P = 0.0024
0.2 G 30 P = 0.0424
0
Anti- Anti- 25
Spheres (average no.)

miR-CON miR-708
20

15

10

0
Anti- Anti-
miR-CON miR-708

asked whether miR-708 has the potential to alter the tumor- miR-708 on TICs, we conducted sphere formation and clono-
initiating capacity of prostate cancer cells. To this end, we genicity assays, which have been widely used to measure the
purified CD44þ and CD44 cells from PC3 xenografts, manip- activity of TICs (10–12, 16). miR-708 reconstitution significant-
ulated miR-708 in these subpopulations followed by in vitro ly inhibited sphere formation (Fig. 6C) and clonogenic poten-
assays for determining their tumor-initiating capacity (Fig. 6). tial (Fig. 6D) of the CD44þ subpopulation. We also inhibited
CD44þ prostate cancer cells are more tumorigenic than the miR-708 in the purified CD44 population from PC3 xenografts
isogenic CD44 prostate cancer cells (10, 12). CD44þ cells were (Fig. 6E–G) using anti-miR-708 oligos and examined its effect
transfected with control miR or miR-708 precursor (Fig. 6A–D). on the tumorigenic potential of CD44 prostate cancer cells.
We validated overexpression of miR-708 in CD44þ cells by qRT- Treatment with anti-miR-708 led to miR-708 inhibition com-
PCR (Fig. 6A). As shown in Fig. 6B, miR-708 overexpression led pared with control oligo (anti-miR-CON) as assessed by qRT-
to growth inhibition of CD44þ cells. To evaluate the effect of PCR (Fig. 6E). Inhibition of miR-708 led to increased growth

3626 Cancer Res; 72(14) July 15, 2012 Cancer Research


(Fig. 6F) and sphere formation ability (Fig. 6G) in the CD44
subpopulation. Taken together, these observations suggest
that miR-708 negatively regulates the tumor-initiating capacity
of prostate cancer. These results suggest that the observed
inhibitory effects of miR-708 on the tumorigenicity of bulk
prostate cancer cells in vitro and in vivo might be due to its
effects on TICs in prostate cancer.
AKT2 is also a functional target of miR-708 in prostate
cancer
Using miRNA target prediction algorithms miRANDA (21)
and TargetScan (22), we found that AKT2 is another putative
miR-708 target as it possesses 3 potential miR-708–binding
sites within its 30 -UTR (Fig. 7A). miR-708 reconstitution in PC3/
LNCaP cells led to reduced endogenous protein levels of AKT2
(Fig. 7B). Luciferase reporter assays with control/AKT2 30 -UTR
reporter constructs in miR-708/miR-CON–expressing or
mock-transfected PC3/LNCaP cells showed that miR-708
represses AKT2 directly (Fig. 7C). To validate this further, we
examined AKT2 levels from xenograft mouse model study and
found that intratumoral administration of miR-708 led to
reduced AKT2 expression compared with control (Fig. 7D).
We also examined AKT2 regulation in CD44 and CD44þ
subpopulations of PC3 xenografts (Supplementary Fig. S3) and
found increased AKT2 expression in CD44þ subpopulation.
This observation lends support to our hypothesis that AKT2
levels are directly regulated by miR-708.
To explore whether AKT2 is a biologically relevant mediator
of antitumorigenic effects of miR-708 in prostate cancer, we
conducted phenocopy experiments, where we inhibited AKT2
by siRNA and determined its effects on the tumorigenicity of
prostate cancer cell line PC3 (Fig. 7E–I). Out of 3 sets of siRNAs
tested, si2 produced the most efficient knockdown of AKT2 as
assessed by RT-PCR and immunoblot analysis (Fig. 7E). Sub-
sequent experiments using this siRNA showed that inhibition
of AKT2 led to decreases in cellular viability (Fig. 7F), migration
(Fig. 7G), and invasion (Fig. 7H) of PC3 cells and increased
apoptosis compared with control siRNA–treated cells (Fig. 7I).
This suggests that AKT2 is another biologically relevant target
of miR-708 in prostate cancer and miR-708–mediated repres-
sion of AKT2 may partially determine its effects on prolifer-
ation, survival, apoptosis, and migration in prostate cancer.
Discussion
Here, we identify a novel miRNA alteration that is widely
represented in prostate cancer. miR-708 was consistently
downregulated in all prostate cancer cell lines tested suggest-
ing that miR-708 may have tumor-suppressive activity in
prostate cancer. The widespread attenuation of miR-708 in
prostate cancer was further validated in a cohort of human
prostate cancer tissue specimens. We evaluated the potential
for low miR-708 expression as a diagnostic biomarker for
prostate cancer, and our analyses suggest that miR-708 expres-
sion can be a significant parameter to discriminate between
normal and tumor tissues in prostate cancer. Furthermore, low
miR-708 expression was significantly correlated with poor
survival outcome, tumor progression, and recurrence in clin-
www.aacrjournals.org
miR-708 Regulates CD44 in Prostate Cancer
ical specimens. Taken together, these findings suggest that this
widely observed miR alteration has significant potential as a
disease biomarker for diagnosis and prognosis in prostate
cancer. Furthermore, our in vitro and in vivo data suggest that
reconstitution of miR-708 in prostate cancer suppresses
tumorigenicity, confirming the tumor-suppressive role of
miR-708 in prostate cancer. Moreover, intratumoral delivery
of synthetic miR-708 oligonucleotides was sufficient to trigger
in vivo regression of established tumors in a murine xenograft
model of human prostate cancer, supporting the therapeutic
potential of this novel miRNA in prostate cancer.
The existence of tumor-initiating subpopulations within the
bulk of tumor has been described in murine and human
prostate cancer. These cells are endowed with high tumori-
genic capacity and drive tumor growth, metastasis, cause
treatment resistance, and recurrence (23, 24). Several studies
Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022
have identified CD44 as the predominant marker of prostate
cancer TICs (7–12) in addition to other cell surface markers
(7, 9–12, 23). CD44, a transmembrane receptor for hyaluronan,
plays critical roles in cell–cell and cell–matrix adhesion,
migration, signaling, and tumor metastasis (25). Significantly,
this study documents an intricate interplay between miR-708
and CD44. Several lines of evidence support this regulatory
interplay: (i) We isolated CD44þ prostate cancer cells from
prostate cancer xenografts and found that miR-708 expression
is consistently downregulated in this subpopoulation com-
pared with CD44 subset purified from xenograft tumors. This
observation suggests that miR-708 expression is specifically
attenuated in prostate cancer TICs and also points to a possible
miR-708–mediated regulation of CD44. (ii) Furthermore, miR-
708 reexpression in prostate cancer cell lines led to decreased
endogenous CD44 protein expression, and (iii) reporter assays
suggest that miR-708 directly regulates CD44 expression. (iv)
Intratumoral administration of miR-708 in prostate cancer
xenografts led to reduced CD44 protein levels suggesting that
CD44 is a direct target of miR-708. (v) CD44 inhibition partially
phenocopies the antitumorigenic effects of miR-708. (vi) Also,
siRNA-mediated CD44 knockdown led to increased miR-708
expression. In addition, we also introduced miR-708 in CD44þ
subpopulation of prostate cancer xenografts and observed
growth inhibitory effects in vitro. Conversely, miR-708 inhibi-
tion in the purified CD44 population from PC3 xenografts led
to increased growth. Collectively, this data suggests that miR-
708 is an important regulator of CD44 and the tumor-initiating
capacity of prostate cancer. The positive correlation of miR-708
downregulation with prostate cancer recurrence lends further
support to this interplay. In view of these results, we suggest
that the tumor-inhibitory effects of miR-708 observed in bulk
prostate cancer cells in vitro and in vivo might be due to its
effects on prostate cancer TICs.
Recent evidence suggests that TICs are intricately regulated
by networks of miRNAs (26). Bioinformatical analysis and
experimental evidence suggest that the 30 -UTR of a single gene
is frequently targeted by several different miRNAs (27, 28),
suggesting that miRNAs might cooperate to regulate specific
gene expression (29, 30). CD44 seems to exemplify such a
combinatorial control as it has been reported that miR-34a
(17), miR-373, and miR-520c (31) regulate CD44 expression in
Cancer Res; 72(14) July 15, 2012 3627
 Saini et al.

A AKT2 3′ UTR B PC3 LNCaP

miR-CON

miR-CON
1 3555

miR-708

miR-708
Site 1 Site 2 Site 3

Mock

Mock
Site 1 969: 5′… CCUCUGAAAAGGGAGGCUCCUA G … 3′ AKT2 3′ UTR
:
3′ GGGUCGAUCUAACA UUCGAGGAA 5′ hsa-miR-708
AKT2
Site 2 1602: 5 ...AGACCCCUGGGCCCCAGCUCCUC... 3′ AKT2 3′ UTR

3′ GGGUCGAUCUAACA UUCGAGGAA 5′ hsa-miR-708 GAPDH

Control
Site 3 2043: 5 ...UUGCUGUCACCUGGGAGCUCCUG...3′ AKT2 3′ UTR D
: miR-708
3′ GGGUCGAUCUAACA UUCGAGGAA 5′ hsa-miR-708
C PC3 LNCaP
T1 T2 T3 T4 T5

1.2 1.4 Control 3′ UTR AKT2 3′ UTR AKT2

1.2
Relative luciferase activity

1
GAPDH
0.8

Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022

0.8
0.6 F NS siRNA AKT2 siRNA

Relative cell viability (%)

120
0.6
100
0.4
0.4
80
0.2
0.2 60

0 0 40
Mock miR-CON miR-708 Mock miR-CON miR-708
NS siRNA

20
siRNA
AKT2

0
E Days 0 1 2 3
1 2 3 G NS siRNA AKT2 siRNA
H NS siRNA AKT2 siRNA
1.2 AKT2
Relative AKT2 expression

0.8
GAPDH
1.2
0.6 1.2
Relative migration

1
Relative invasion

1
0.4
0.8
0.8
0.2 P = 0.0001 P = 0.0314
0.6
0.6
0.4
0
0.4
NS siRNA 1 2 3
0.2
AKT2 siRNA 0.2
0
NS siRNA AKT2 siRNA 0
NS siRNA AKT2 siRNA
I NS siRNA
% Apoptotic 3±1
AKT2 siRNA
% Apoptotic 9±1
% Early apoptotic
% Viables
3±1
93 ± 2
% Early apoptotic 5±1 J
% Viables 86 ± 1
% Dead 0.2 ± 0.1 P = 0.0081
% Dead 1 ± 0.4 16

Total cells Total cells 14

12
% Apoptosis

10

0
NS siRNA AKT2 siRNA

Figure 7. AKT2 is also a functional target of miR-708 in prostate cancer. A, schematic representation of the AKT2 30 -UTR showing the relative positions of 3
putative miR-708 target sites. B, immunoblots for endogenous AKT2 protein in PC3/LNCaP cells transfected as indicated. GAPDH was used a loading control.
C, luciferase activity assay with the AKT2 30 UTR construct/control construct contransfected with miR-CON/miR-708. Firefly luciferase values were
normalized to Renilla luciferase activity and plotted as relative luciferase activity ( , P < 0.05). D, AKT2 expression in harvested tumors from mouse prostate
cancer xenografts intratumorally injected with either miR-CON or miR-708 as assessed by immunoblot analysis (T1–T5 are individual tumors). E, PC3 cells
were transfected with AKT2/NS siRNA for 72 hours. Relative AKT2 mRNA expression and protein expression after siRNA transfections as assessed
by RT-PCR and immunoblotting, respectively. Cell viability assay (F), Transwell migration assay (G), invasion assay (H), and apoptosis assay (I) in PC3 cells
after NS siRNA (left) or AKT2 siRNA (right) treatment. J, representation of average apoptotic fractions in each group ( , P < 0.05).

3628 Cancer Res; 72(14) July 15, 2012 Cancer Research


prostate cancer. miR-34a directly represses CD44 thereby
inhibiting prostate cancer stem cells and metastasis
(17, 24, 26). Here, we report that CD44 is a direct and functional
target of miR-708 in prostate cancer.
miRNAs suppress gene expression posttranscriptionally via
sequence-specific interactions with the 30 -UTRs of cognate
mRNA targets (20). An individual miRNA is capable of regu-
lating dozens of distinct mRNAs (32, 33), and it is known that
pleiotropic suppression of multiple downstream effectors may
underlie the phenotypic changes observed upon perturbing the
levels of certain miRNAs (13, 34–37). We found that miR-708
targets AKT2 as well. Luciferase reporter assays showed that
miR-708 represses its dependent genes (CD44 and AKT2) by
direct binding to the complementary sites within their 30 -
UTRs. The serine/threonine kinase AKT2/PKBb is a core
member of the phosphoinositide 3-kinase (PI3K)/AKT signal-
ing pathway, which plays a key role in cellular proliferation,
survival, migration, invasion, and metastasis. Aberrant activa-
tion of the PI3K/AKT pathway contributes to tumorigenesis,
which is associated with poor outcome in prostate cancer (38,
39). AKT2 inhibition suppresses the growth of human prostate
cancer cells in vitro and in vivo (40). PI3K/AKT pathways also
play a role in the maintenance and viability of stem-like cell
populations in prostate cancer (41). In view of this and our
present results, we propose that miR-708–mediated repression
of CD44 and AKT2 may underlie its role in potentially affecting
the tumor-initiating subpopulation of prostate cancer cells.
In conclusion, our study establishes miR-708 attenuation as
a major alteration in prostate cancer that is associated with
poor survival, recurrence, and progression in prostate cancer.
Importantly, we show, to our knowledge, for the first time that
miR-708 is a key negative regulator of CD44þ subpopulation of
prostate cancer cells and that CD44 is a direct functional target
of miR-708. miR-708 reconstitution in prostate cancer cell lines
or CD44þ prostate cancer cells led to reduced tumorigenicity,
supporting the tumor suppressive effects of miR-708 in pros-
tate cancer. In view of these results, we propose that reduced
References
1. Shen MM, Abate-Shen C. Molecular genetics of prostate cancer: new
prospects for old challenges. Genes Dev 2010;24:1967–2000.
2. Brooks DD, Wolf A, Smith RA, Dash C, Guessous I. Prostate cancer
screening 2010: updated recommendations from the American Can-
cer Society. J Natl Med Assoc 2010;102:423–9.
3. Wolf AM, Wender RC, Etzioni RB, Thompson IM, D'Amico AV, Volk
RJ, et al. American Cancer Society guideline for the early detection
of prostate cancer: update 2010. CA Cancer J Clin 2010;60:
70–98.
4. Grubb RL III, Kibel AS. Prostate cancer: screening, diagnosis and
management in 2007. Mol Med 2007;104:408–13.
5. Kasper S, Cookson MS. Mechanisms leading to the development of
hormone-resistant prostate cancer. Urol Clin North Am 2006;33:
201–10.
6. Tannock IF, de Wit R, Berry WR, Horti J, Pluzanska A, Chi KN, et al.
Docetaxel plus prednisone or mitoxantrone plus prednisone for
advanced prostate cancer. N Engl J Med 2004;351:1502–12.
7. Tang DG, Patrawala L, Calhoun T, Bhatia B, Choy G, Schneider-
Broussard R, et al. Prostate cancer stem/progenitor cells: identifi-
cation, characterization, and implications. Mol Carcinog 2007;
46:1–14.
www.aacrjournals.org
miR-708 Regulates CD44 in Prostate Cancer
miR-708 expression leads to prostate cancer initiation, pro-
gression, and development by regulating expression of CD44 as
well as AKT2, that in turn regulates proliferative, invasive, and
migratory abilities of prostate cancer TICs. In view of the
widespread attenuation of miR-708 observed in our patient
cohort, its antitumorigenic effects on bulk or CD44þ subpop-
ulation of prostate cancer cells and in vivo inhibition of
prostate tumor xenografts, we envision that restoring miR-
708 may provide a novel therapeutic modality for prostate
cancer.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: S. Saini, R. Dahiya
Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022
Development of methodology: S. Saini, S. Majid, G. Deng, R. Dahiya
Acquisition of data (provided animals, acquired and managed patients,
provided facilities, etc.): S. Saini, S. Majid, V. Shahryari, S. Yamamura, M.S.
Zaman
Analysis and interpretation of data (e.g., statistical analysis, biostatistics,
computational analysis): S. Saini, S. Majid, G. Deng, Y. Tanaka, R. Dahiya
Writing, review, and/or revision of the manuscript: S. Saini, Y. Tanaka, R.
Dahiya
Administrative, technical, or material support (i.e., reporting or orga-
nizing data, constructing databases): S. Saini, V. Shahryari, S. Arora, I. Chang,
G. Deng, Y. Tanaka
Study supervision: Y. Tanaka, R. Dahiya
Acknowledgments
The authors thank Dr. Roger Erickson for his support and assistance with
preparation of the manuscript.
Grant Support
This research was supported by the National Center for Research Resources of
the NIH through Grant Number RO1CA 138642, RO1CA160079, T32DK007790,
and VA Merit Review and VA Program Project.
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 February 9, 2012; revised April 16, 2012; accepted April 24, 2012;
published OnlineFirst May 2, 2012.
8. Collins AT, Berry PA, Hyde C, Stower MJ, Maitland NJ. Prospective
identification of tumorigenic prostate cancer stem cells. Cancer Res
2005;65:10946–51.
9. Li H, Chen X, Calhoun-Davis T, Claypool K, Tang DG. PC3 human
prostate carcinoma cell holoclones contain self-renewing tumor-initi-
ating cells. Cancer Res 2008;68:1820–5.
10. Patrawala L, Calhoun T, Schneider-Broussard R, Li H, Bhatia B, Tang
S, et al. Highly purified CD44 þprostate cancer cells from xenograft
human tumors are enriched in tumorigenic and metastatic progenitor
cells. Oncogene 2006;25:1696–708.
11. Patrawala L, Calhoun T, Schneider-Broussard R, Zhou J, Claypool K,
Tang DG. Side population is enriched in tumorigenic, stem-like cancer
cells, whereas ABCG2þ and ABCG2- cancer cells are similarly tumor-
igenic. Cancer Res 2005;65:6207–19.
12. Patrawala L, Calhoun-Davis T, Schneider-Broussard R, Tang DG.
Hierarchical organization of prostate cancer cells in xenograft tumors:
the CD44þalpha2beta1þ cell population is enriched in tumor-initiating
cells. Cancer Res 2007;67:6796–805.
13. Saini S, Majid S, Yamamura S, Tabatabai L, Suh SO, Shahryari V, et al.
Regulatory role of mir-203 in prostate cancer progression and metas-
tasis. Clin Cancer Res 2011;17:5287–98.
Cancer Res; 72(14) July 15, 2012 3629
 Saini et al.
14. Craft N, Chhor C, Tran C, Belldegrun A, DeKernion J, Witte ON, et al.
Evidence for clonal outgrowth of androgen-independent prostate
cancer cells from androgen-dependent tumors through a two-step
process. Cancer Res 1999;59:5030–6.
15. Klein KA, Reiter RE, Redula J, Moradi H, Zhu XL, Brothman AR, et al.
Progression of metastatic human prostate cancer to androgen inde-
pendence in immunodeficient SCID mice. Nat Med 1997;3:402–8.
16. Li H, Jiang M, Honorio S, Patrawala L, Jeter CR, Calhoun-Davis T, et al.
Methodologies in assaying prostate cancer stem cells. Methods Mol
Biol 2009;568:85–138.
17. Liu C, Kelnar K, Liu B, Chen X, Calhoun-Davis T, Li H, et al. The
microRNA miR-34a inhibits prostate cancer stem cells and metastasis
by directly repressing CD44. Nat Med 2011;17:211–5.
18. Majid S, Saini S, Dar AA, Hirata H, Shahryari V, Tanaka Y, et al.
MicroRNA-205 inhibits Src-mediated oncogenic pathways in renal
cancer. Cancer Res 2011;71:2611–21.
19. Trang P, Medina PP, Wiggins JF, Ruffino L, Kelnar K, Omotola M, et al.
Regression of murine lung tumors by the let-7 microRNA. Oncogene
2010;29:1580–7.
20. Bartel DP. MicroRNAs: target recognition and regulatory functions.
Cell 2009;136:215–33.
21. Betel D, Wilson M, Gabow A, Marks DS, Sander C. The microRNA.org
resource: targets and expression. Nucleic Acids Res 2008;36:
D149–53.
22. Grimson A, Farh KK, Johnston WK, Garrett-Engele P, Lim LP, Bartel
DP. MicroRNA targeting specificity in mammals: determinants beyond
seed pairing. Mol Cell 2007;27:91–105.
23. Li H, Tang DG. Prostate cancer stem cells and their potential roles in
metastasis. J Surg Oncol 2011;103:558–62.
24. Wicha MS. Stemming a tumor with a little miR. Nat Med 2011;17:
162–4.
25. Iczkowski KA. Cell adhesion molecule CD44: its functional roles in
prostate cancer. Am J Transl Res 2010;3:1–7.
26. Liu C, Tang DG. MicroRNA regulation of cancer stem cells. Cancer Res
2011;71:5950–4.
27. Krek A, Grun D, Poy MN, Wolf R, Rosenberg L, Epstein EJ, et al.
Combinatorial microRNA target predictions. Nat Genet 2005;37:
495–500.
28. Lewis BP, Shih IH, Jones-Rhoades MW, Bartel DP, Burge CB. Pre-
diction of mammalian microRNA targets. Cell 2003;115:787–98.
3630 Cancer Res; 72(14) July 15, 2012
29. Garzon R, Marcucci G, Croce CM. Targeting microRNAs in cancer:
rationale, strategies and challenges. Nat Rev Drug Discov 2010;9:
775–89.
30. Ivanovska I, Cleary MA. Combinatorial microRNAs: working together
to make a difference. Cell Cycle 2008;7:3137–42.
31. Yang K, Handorean AM, Iczkowski KA. MicroRNAs 373 and 520c are
downregulated in prostate cancer, suppress CD44 translation and
enhance invasion of prostate cancer cells in vitro. Int J Clin Exp Pathol
2009;2:361–9.
32. Baek D, Villen J, Shin C, Camargo FD, Gygi SP, Bartel DP. The impact
of microRNAs on protein output. Nature 2008;455:64–71.
33. Selbach M, Schwanhausser B, Thierfelder N, Fang Z, Khanin R,
Rajewsky N. Widespread changes in protein synthesis induced by
microRNAs. Nature 2008;455:58–63.
34. Saini S, Arora S, Majid S, Shahryari V, Chen Y, Deng G, et al. Curcumin
modulates microRNA-203-mediated regulation of the Src-Akt axis in
bladder cancer. Cancer Prev Res (Phila) 2011;4:1698–709.
35. Valastyan S, Benaich N, Chang A, Reinhardt F, Weinberg RA. Con-
comitant suppression of three target genes can explain the impact of a
Downloaded from http://aacrjournals.org/cancerres/article-pdf/72/14/3618/2669162/3618.pdf by guest on 31 May 2022
microRNA on metastasis. Genes Dev 2009;23:2592–7.
36. Valastyan S, Chang A, Benaich N, Reinhardt F, Weinberg RA. Con-
current suppression of integrin alpha5, radixin, and RhoA phenocopies
the effects of miR-31 on metastasis. Cancer Res 2010;70:5147–54.
37. Valastyan S, Weinberg RA. MicroRNAs: crucial multi-tasking compo-
nents in the complex circuitry of tumor metastasis. Cell Cycle
2009;8:3506–12.
38. Kreisberg JI, Malik SN, Prihoda TJ, Bedolla RG, Troyer DA, Kreisberg
S, et al. Phosphorylation of Akt (Ser473) is an excellent predictor of
poor clinical outcome in prostate cancer. Cancer Res 2004;64:5232–6.
39. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS,
et al. Integrative genomic profiling of human prostate cancer. Cancer
Cell 2010;18:11–22.
40. Sasaki T, Nakashiro K, Tanaka H, Azuma K, Goda H, Hara S, et al.
Knockdown of Akt isoforms by RNA silencing suppresses the growth
of human prostate cancer cells in vitro and in vivo. Biochem Biophys
Res Commun 2010;399:79–83.
41. Dubrovska A, Kim S, Salamone RJ, Walker JR, Maira SM, Garcia-
Echeverria C, et al. The role of PTEN/Akt/PI3K signaling in the main-
tenance and viability of prostate cancer stem-like cell populations.
Proc Natl Acad Sci U S A 2009;106:268–73.
Cancer Research