Metastatic heterogeneity of breast cancer cells is...

1

Metastatic heterogeneity of breast cancer cells is associated with expression of a

heterogeneous TGF- -activating miR-424-503 gene cluster

Yun Li1, 2, Wei Li1, 2, Zhe Ying1, 2, Han Tian1, 2, Xun Zhu1, 2, Jun Li2, 3, Mengfeng Li1, 2, *

1 Department of Microbiology, Zhongshan School of Medicine, Sun Yat-sen University,

Guangzhou, Guangdong 510080, China

2 Key Laboratory of Tropical Disease Control (Sun Yat-sen University), Chinese Ministry of

Education, Guangzhou, Guangdong 510080, China

3 Department of Biochemistry, Zhongshan School of Medicine, Sun Yat-sen University,

Guangzhou, Guangdong 510080, China

*To whom all correspondence should be addressed:

Mengfeng Li, MD, Ph.D. 74 Zhongshan Road II, Guangzhou, Guangdong 510080, China.

E-mail: [email protected]

Running Title: miR-424-503 activates TGF- and promotes breast cancer metastasis.

Key Words: miR-424-503 cluster; breast cancer; metastasis; heterogeneous TGF- activity.

Conflict of interest: The authors declare no conflict of interest.

Research. on August 19, 2018. © 2014 American Association for Cancercancerres.aacrjournals.org Downloaded from

Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on August 27, 2014; DOI: 10.1158/0008-5472.CAN-14-0389

http://cancerres.aacrjournals.org/

2

Abstract

Transforming growth factor beta (TGF- ) signaling is known to drive metastasis in human

cancer. Under physiological conditions, the level of TGF- activity is tightly controlled by

a regulatory network involving multiple negative regulators. At metastasis, however, these

inhibitory mechanisms are usually overridden so that oncogenic TGF- signaling can be

overactivated and sustained. To better understand how the TGF- inhibitors are suppressed

in metastatic breast cancer cells, we compared miRNA expression profiles between breast

cancers with or without metastasis and found that the miR-424-503 cluster was markedly

overexpressed in metastatic breast cancer. Mechanistic studies revealed that miR-424 and

miR-503 simultaneously suppressed Smad7 and Smurf2, two key inhibitory factors of TGF-

signaling, leading to enhanced TGF- signaling and metastatic capability of breast cancer

cells. Moreover, antagonizing miR-424-503 in breast cancer cells suppressed metastasis in

vivo and increased overall host survival. Interestingly, our study also found that

heterogeneous expression of the miR-424-503 cluster contributed to the heterogeneity of

TGF- activity levels in, and metastatic potential of, breast cancer cell subsets. Overall, our

findings demonstrate a novel mechanism, mediated by elevated expression of miR-424-503

cluster, underlying TGF- activation and metastasis of human breast cancer.




3

Introduction

Breast cancer is second only to lung cancer as the cause of cancer-related deaths in

women, and the vast majority of breast cancer-related deaths are due to metastatic diseases

(1). It is well established that cancer metastasis, a complex, multi-step process, is driven,

promoted and modulated by aberrantly deregulated cellular signals (2). Numerous signaling

pathways, such as the Myc, -catenin and TGF- pathways, have been identified to play key

roles in breast cancer metastasis (3-5). Among these metastasis-associated cellular signal

transduction pathways, the TGF- cascade is well documented in mediating many of the

pro-metastatic steps (5). High levels of TGF- or overactivated TGF- receptors have been

linked to the invasiveness and metastasis of breast cancer cells (6,7). In contrast, low

expression of TGF- receptors in breast tumors correlates with a favorable disease outcome

(8). Notably, several studies using inhibitors of TGF- have shown repression of metastasis

of mammary carcinomas (9), further highlighting a critical role of TGF- signaling in the

development and progression of breast cancer metastasis.

TGF- signaling is triggered by binding of TGF- receptor ligands to a complex of

transmembrane receptor serine/threonine kinases (types I and II), followed by

phosphorylation of receptor-activated Smads (R-Smads). Once activated, R-Smads form a

complex with a common Smad4 and translocate into the nucleus, where they regulate the

transcription of target genes along with various cofactors (10). Every step of the TGF-

signaling cascade is tightly controlled by specialized factors. For example, Smad7 is key to

regulating the activity of TGF- signaling via a negative feedback mechanism (11). Smad7

was found to block the phosphorylation of Smad2/3 upon TGF- stimulation through binding

to the TGF- receptor complex (12) and inhibit the hetero-complex formation between

R-Smads and Co-Smad (13). Importantly, Smad7 binds to the Smad ubiquitination

regulatory factor 2 (Smurf2), a member of the HECT E3 ligases family, resulting in




4

ubiquitination of TGF- type I receptor and consequent inhibition of R-Smad activation (14).

Furthermore, Smurf2 can target R-Smads for ubiquitin-mediated degradation and hence

terminate R-Smads mediated signaling (15).

Deregulation of TGF- activation associated with decreases of its key suppressors has

been found in various human cancer types. Indeed, recent evidence has shown that

low-level expression of Smad7 correlates with lymph node metastasis in pancreatic cancer

(16), and knockdown of Smurf2 in human breast cancer cells results in enhanced cell

migration in vitro and bone metastasis in vivo (17), suggesting that the quantities of Smad7

and Smurf2 play an important role in modulating the level of TGF- signaling activity during

the progression of tumor metastasis. How these negative regulators of TGF- signaling are

decreased in cancers, however, remains poorly understood, but it remains an important area

of research for future development of anti-TGF- strategies.

miRNAs are small regulatory RNA molecules that posttranscriptionally downregulate

target mRNAs by interacting with their 3’ untranslated region (18). Differential expression

of miRNA in normal and tumor tissues, or between biological scenarios with or without

metastasis, has been analyzed in a variety of cancer types (19,20). Our present work reports

that the miR-424-503 miRNA cluster might contribute to TGF- hyperactivation and thus

might represent a new mechanism underlying breast cancer metastasis.




5

Materials and Methods

Cell Culture. Breast cancer cell lines MCF7, T47D, BT474, MDA-MB-231, SKBR3 and

MDA-MB-435 were purchased from ATCC in 2008, and according to the provider, the cell

lines were authenticated using short tandem repeat (STR) DNA fingerprinting (21). These

cell lines were grown in Dulbecco's Modified Eagle Medium (DMEM) supplemented with

10% fetal bovine serum (FBS) (HyClone, Logan, UT) in our laboratory and re-authenticated

by STR fingerprinting profiles at the Forensic Medicine Department of Sun Yat-Sen

University in May 2014, which revealed fingerprinting profiles identical to those given in the

ATCC STR database.

Clinical specimens and patient information. Paraffin-embedded human breast cancer

specimens were histopathologically diagnosed at the First Affiliated Hospital of SYSU, and

relevant clinical information is presented in Table S1. Prior donors’ consents and approvals

from the Institutional Research Ethics Committee were obtained. TCGA dataset was

accessed at https://tcga-data.nci.nih.gov/tcga/dataAccessMatrix.htm.

Plasmids, infection and transfection. Plasmids were constructed using standard methods

(22). Recombinant retrovirus production and infection were performed as previously

described (23). Transfection of plasmids or oligonucleotides was performed using the

Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s

instruction. For detailed description of each plasmid and sequences of synthetic

oligonucleotides, see Supplementary Materials and Methods.

miRNA extraction and qPCR. Total miRNA isolation and cDNA synthesis were

performed using commercial kits according to the manufacturer’s instructions. miRNA

expression was quantified with qPCR using miRNA-specific primers in a real-time PCR

system. For details, see Supplementary Materials and Methods.




6

Wound healing assay. Cells were seeded in 6-well plates and grown to 90% confluence,

followed by serum starvation for 24 h. A linear wound was created in the confluent

monolayer using a pipette tip, and wounds were observed and photographed at various time

points. Wound size was measured randomly at five sites perpendicular to the wound.

Transwell matrix penetration assay. Cells (2×104) were plated into the top side of

polycarbonate Transwell coated with Matrigel (10%, BD, Franklin Lakes, NJ) and incubated

at 37°C for 22 hours, followed by removal of cells inside the upper chamber with cotton

swabs. Migratory and invasive cells on the lower membrane surface were fixed in 4%

paraformaldehyde, stained with hematoxylin, and counted (Ten random 200× fields per well).

Cell counts are expressed as the mean number of cells per field of view.

Three-dimension spheroid invasion assay. Cells (1×104) were mixed with 20% Matrigel

and seeded in 24-well plates coated with 100% Matrigel (BD, Franklin Lakes, NJ), and

medium was changed every other day. Pictures were taken under microscope at various

time points.

Western blotting and immunofluorescence assays. Western blotting and

immunofluorescence assays were performed according to corresponding standard methods

(24,25). Sources of antibodies used are given in Supplementary Materials and Methods.

Fluorescence images were captured using the LSM710 con-focal microscopy system (Carl

Zeiss, Oberkochen, Germany).

Luciferase assay. Cells (3.5×104) were seeded and settled in 48-well plates for 24 hours.

Luciferase plasmid pGL3-control-Smad7-3’UTR, pGL3-control-Smurf2-3’UTR or

pTAL-basic-3TP (100ng each), plus 1 ng of pRL-TK plasmid (Promega, Madison, WI) were

transfected into cells. Luciferase and Renilla signals were measured 48 hours after

transfection using the Dual-Luciferase Reporter Assay kit (Promega, Madison, WI) according




7

to a protocol provided by the manufacturer.

RNA immunoprecipitation (RIP). Co-immunoprecipitation (Co-IP) of miRNP with

anti-Ago1 (Abcam, Cambridge, MA) was performed as previously described (26). Cell

were lysed in buffer containing 100mM KCl, 5mM MgCl2, 10mM HEPES (pH7.4) and

0.5%NP-40; and the immune complex captured by protein A agarose was washed in buffer

containing 150mM KCl, 5mM MgCl2, 10mM HEPES (pH7.4) and 0.1%NP-40 for 6 times.

RNA extraction was performed using the RNAeasy Kit (Qiagen, Valencia, CA).

Immunohistochemical analysis. After sections were stained using anti-Smad7

(Sigma-Aldrich, Germany), anti-Smurf2 (Epitomics, Burlingame, CA), and anti-p-Smad2

antibodies (Abcam, Cambridge, MA), images were captured using the AxioVision Rel.4.6

computerized image analysis system (Carl Zeiss, Oberkochen, Germany). The degree of

immunostaining of indicated proteins was evaluated and scored as previously described (4).

Immuno-laser capture microdissection. All Immuno-LCM tissues were prepared as

previously described (27). Samples were inspected and microdissected using Arcturus® XT

LCM microscope (Life Technologies Corporation, Carlsbad, CA, USA). Approximately

600 cancer cells were microdissected from each slide onto a CapSure® LCM cap. Total

RNA and microRNA from LCM CapSure® caps were extracted by QIAamp® miRNeasy

Mini Kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer's instructions.

Tumor metastasis model. Breast cancer cell lines MCF7 (1.5×106), MDA-MB-231

(1.0×106) and MDA-MB-435 cells (0.8×106) stably expressing firefly luciferase, alone or

together with miR-204-503, or corresponding control vectors, were injected intravenously via

the caudal vein of BALB/c nude mice (n=5). In antagnomir systemic treatment experiment,

100 μl miR-424 antagomir, miR-503 antagomir or antagomir control (diluted in PBS at 2

mg/ml) (Ribo Biotech, Guangzhou, China) was administrated intravenously 3 times per week




8

for 2 weeks, starting day 10 after tumor cells inoculation. For spontaneous metastasis

assays, MDA-MB-435-luciferase cells (3.0×106) were injected into the mammary fat pads

(n=5), and antagomir was administrated intratumorally as previously described when the

average volume of grown tumors reached approximately 50 mm3. Bioluminescence

imaging was performed using the IVIS Spectrum Imaging System (Caliper Life Sciences,

Mountain View, CA). Image calibration and visualization were performed using the Living

Image 4.2 software (Caliper Life Sciences, Mountain View, CA).

Flow cytometry and single cell cloning. Cells were dissociated into single-cell suspension,

and dead cells were excluded by propidium iodide staining. Generation of

single-cell-derived cultures was performed by BD Influx (BD Biosciences) single-cell plating

in 96-well plates (Corning, NY). Single, RFP-positive/GFP-positive cells were gated

stringently. When visible cell aggregates appeared, they were transferred to flasks (Corning,

NY) and expanded.

Statistical analysis. All statistical analyses were carried out using the SPSS 11.0 statistical

software package. All error bars represent mean ± SD derived from three independent

experiments. In all cases, P < 0.05 was considered statistically significant.




9

Results

miR-424 and miR-503 are upregulated in metastatic breast cancer

To identify miRNAs differentially expressed in breast cancer metastases, we retrieved and

comparatively analyzed miRNA expression profiles of metastatic versus non-metastatic

subsets from The Cancer Genome Atlas (TCGA) datasets. Two miRNAs, miR-424 and

miR-503, emerged as highly upregulated miRNAs in metastatic breast cancer specimens as

compared to non-metastatic ones (Fig. 1A). To validate these data, quantitative reverse

transcription PCR (qRT-PCR) was conducted with 24 grade IV breast cancer specimens and

93 low-grade breast cancer specimens (WHO tumor grades I, II and III). Consistently,

expression of miR-424 or miR-503 was markedly higher in grade IV, but significantly lower

in low-grade, breast cancer samples (Fig. 1B). Furthermore, breast cancer cell lines known

to be highly metastatic (MDA-MB-231, SKBR3 and MDA-MB-435) displayed increases of

miR-424 and miR-503 expression relative to non- or low-metastatic cell lines (MCF7, T47D

and BT474) (Fig. 1C), indicating an association between miR-424/miR-503 high expression

and breast cancer metastasis.

Notably, miR-424 and miR-503 levels linearly correlated with each other in breast cancer

cell lines and tissue specimens (Fig. S1, A and B). Furthermore, miR-424 and miR-503

share substantial sequence identity in their seed sequences, and their coding genes are

separated by only 216 bp on the X chromosome (28), suggesting that the miR-424-503

cluster might play cooperative roles in breast cancer.

The miR-424-503 cluster enhances metastasis-associated properties of breast cancer in

vitro

To understand whether the miR-424-503 cluster is involved in metastasis of breast cancer

cells, in vitro study was performed to observe the effect of stable overexpression of the

miRNAs on cellular migration and invasions (Fig. S2). As shown in Fig. 2A, ectopic




10

miR-424, miR-503 or miR-424-503 enhanced the migratory ability of MCF7 and

MDA-MB-231cells in wound-healing assay, and Transwell matrix penetration assay showed

that miR-424, miR-503 or miR-424-503 overexpression drastically increased the

invasiveness of both cell lines (Fig. 2B). Consistently, compared with the vector control

cells, miR-424-, miR-503-, miR-424-503 cluster-transduced cells displayed a highly

aggressive penetrating growth in 3-dimensional (3D) culture (Fig. 2C), suggesting that

miR-424 and miR-503 are pro-metastatic in breast cancer cells.

Next we examined the effect of suppressing endogenous miR-424 or miR-503 on the

phenotype of MDA-MB-231 (moderately metastatic) and MDA-MB-435 (highly metastatic)

cells, and found that inhibiting miR-424 or miR-503 markedly weakened the metastasis

capability of the tested cells (Fig. 2A-C), further confirming the pro-metastatic effects of

miR-424 and miR-503.

The miR-424-503 cluster promotes metastasis of breast cancer cells in vivo

To further examine whether miR-424 and miR-503 promotes metastasis in vivo,

luciferase-expressing MCF7 and MDA-MB-231 cells transduced with the miR-424-503

cluster or control vector were injected into the caudal vein of nude mice. Strikingly,

bioluminescence imaging showed that mice injected with MCF7/miR-424-503 cells

displayed prominent lung metastasis, whereas no visible metastasis was found in mice

injected with control MCF7 cells (Fig. 3A). Consistently, MDA-MB-231/miR-424-503

mice also generated a significantly larger number of lung metastases than those in

vector-control cells (Fig. 3A). To further validate whether the endogenous miR-424-503

cluster was required for the observed enhanced metastasis in vivo, antagomir-424 and

antagomir-503 were applied to inhibit the endogenous expression of miR-424 or miR-503 in

the experimental metastasis assay. When the endogenous mir-424 or mir-503 in

MDA-MB-231 and MDA-MB-435 cells was reduced by the antagomirs (Fig. S3), lung




11

metastasis of both cell lines were significantly abrogated (Figs. 3B). Moreover, 4 out of 5

mice bearing miR-424-503-transduced MCF7 cells died before 50 days after inoculation,

while only 2 mice in the vector-control group died by day 56 after implantation (Fig. 3C).

Similarly, mice injected with miR-424-503-transduced MDA-MB-231 cells survived shorter

as compared with the vector-control cells injected mice (Fig. 3C). When using antagomir,

inhibition of miR-424 or miR-503 extended mice survival (Fig. 3C).

To investigate whether miR-424-503 promotes dissemination of breast cancer cells from

primary tumors, MDA-MB-435 cells were injected into the mammary gland pads of nude

mice and examined for lung metastasis. As Fig. S4A shows, while the control animals

exhibited marked pulmonary metastases, antagomir treatment inhibited metastasis and

prolonged the life of mice without affecting the weights of the primary tumors (Fig. S4, B

and C). Together, these data indicated that the miR-424-503 cluster was a strong promoter

for breast cancer metastasis in vivo.

The miR-424-503 cluster directly targets and suppresses multiple negative regulators of

TGF- signaling

In the light that the TGF- signaling is important for breast cancer metastasis, we then

examined the effect of miR-424-503 on TGF- activation. In cells containing the TGF-

reporter gene 3TP-Lux (29), overexpression of miR-424, miR-503 or miR-424-503

significantly elevated, while inhibition of miR-424 or miR-503 dramatically reduced, the

transactivating activity of TGF- (Fig. S5A). Concordantly, immunofluorescence staining

showed that upon TGF- treatment, overexpression of miR-424, miR-503 or miR-424-503

elevated the nuclear enrichment of Smad3 (Fig. S5B). Conversely, inhibition of miR-424 or

miR-503 decreased the quantity of nuclear Smad3 (Fig. S5B). These data suggest that the

miR-424-503 cluster overexpression is able to enhance activity of TGF- signaling.




12

Next, we screened for targets of miR-424 and miR-503 using the TargetScan Program and

found that Smad7 and Smurf2, known as negative regulators of TGF- signaling, were

potential target genes of miR-424 and miR-503 (Fig. 4A). To validate this prediction, we

found that both Smad7 and Smurf2 proteins were reduced in miR-424 and/or miR-503

overexpressing cells but increased by specific antagomir(s) (Fig. 4B), whereas no detectable

alterations in Smad7 or Smurf2 mRNA was seen (Fig. S6, A and B). When cells were

transduced with a luciferase reporter containing Smad7 or Smurf2 3’UTR, ectopic expression

of miR-424 or miR-503 decreased, while inhibition of miR-424 or miR-503 increased the

activity of reporter luciferase (Fig. 4C). Furthermore, mutations introduced to miR-424 and

miR-503 target sites of the Smad7 or Smurf2 3’UTR abrogated the suppressive effects (Fig.

4D). Moreover, miRNP IP assay revealed high-level enrichment of Smad7 and Smurf2

transcripts in the miRNP of miR-424 mimic or miR-503 mimic transfected cells (Fig. 4E).

Collectively, these results establish Smad7 and Smurf2 as targets of the miR-424-503 cluster.

Smad7 and Smurf2 repression is essential for miR-424-503 mediated metastasis

To determine the functional significance of Smad7 and Smurf2 genes in promoting breast

cancer metastasis induced by the miR-424-503 cluster, we further stably expressed Smad7

open reading frame (ORF), or Smurf2 ORF, free of 3’UTR in miR-424-503-transduced cells

(Fig. 5A). As shown in Fig. 5B, restoration of Smad7 or Smurf2 protein expression

decreased invasiveness of miR-424-503 transduced cells in the Transwell matrix penetration

assay.

To determine the effects of Smad7 and Smurf2 on tumor metastasis in vivo, we injected

mice intravenously with miR-424-503/Smad7 ORF- or miR-424-503/Smurf2

ORF-transduced cells, as well as their corresponding vector-control cells. As shown in Fig.

5C, Smad7 or Smurf2 abrogated the metastasis-promoting effect of miR-424-503, suggesting

an essential role of Smad7 and Smurf2 in miR-424-503 conferred breast cancer metastasis.




13

We then examine the clinical relevance of the miR-424-503/Smad7-Smurf2/TGF- axis in

human breast cancer specimens. As shown in Fig. 5D, 78.1% (57 cases) and 67.1% (49

cases) of specimens expressing low-level miR-424-503 (73 cases), respectively, exhibited

high Smad7 and Smurf2 expression, whereas 56.8% (25 cases) and 50% (22 cases) of

samples with high miR-424-503 expression (44 cases) showed low Smad7 and Smurf2,

respectively (both P < 0.05). Moreover, those cases expressing high miR-424-503

displayed higher p-Smad3 levels (38 of 44 samples; 86.4%) than the low miR-424-503 cases

(29 of 73 samples; 39.7%) (P < 0.05), suggesting that the miR-424-503 cluster

overexpression in clinical breast cancer lesions was associated with downregulation of

Smad7 and Smurf2 and activation of TGF- .

Expression of the miR-424-503 cluster correlates with TGF- activity in breast cancer

Previous evidence has suggested that the intensity of TGF- signaling is often

heterogeneous in the cell population of a breast tumor, and such heterogeneity can be

associated with different metastatic behaviors of breast cancer cells (30,31). Our current

immunohistochemical study also showed that nuclear p-Smad3 staining in a breast tumor was

not homogeneous (Fig. 5D). To assess whether the differential activities of TGF- signaling

were relevant to the expression levels of miR-424-503, we employed laser capture

microdissection (LCM) method to obtain p-Smad3high and p-Smad3low loci, from 10 clinical

breast cancer specimens. Subsequent qRT-PCR analysis demonstrated that miR-424 and

miR-503 levelswere dramatically higher in p-Smad3high tissues than those in p-Smad3low

tissues (Fig. S7A). Furthermore, we determined the intensities of TGF- signaling in

various cell subsets of the MCF7, MDA-MB-231 and MDA-MB-435 breast cancer cell lines

by using a lentivirally transduced T RE-GFP reporter system (Fig. 6A). To exclude

variations in lentiviral integration sites and in copy numbers among cells, the transduced

culture was single-cell cloned. As shown in Fig. 6A, upon TGF- treatment, cells in these




14

single-cell-derived T RE-GFP cultures showed considerable heterogeneity in TGF-

signaling levels, which were not present in control cells transduced with a GFP construct

driven by a constitutive CMV promoter, indicating that the heterogeneity was due to

differences in TGF- -mediated transcription. Moreover, the highest and lowest 10% of

T RE-GFP-expressing cells expressed differential levels of TGF- targeted gene PAI1 (Fig.

6B) and Smad2 and Smad3 activation levels (Fig. 6C). These results are consistent with the

previous finding that cells derived from the same breast tumor are heterogeneous in their

TGF- activity.

Most notably, the protein levels of Smad7 and Smurf2 significantly decreased in

T RE-GFPhigh cells, as compared with those in T RE-GFPlow cells (Fig. 6C), while their

mRNA levels showed no significant difference between T RE-GFPlow and T RE-GFPhigh

cells (Fig. S7B). Further assessment of miR-424 and miR-503 quantities showed that levels

of both miRNAs were dramatically higher in T RE-GFPhigh cells than in T RE-GFPlow cells

(Fig. 6D).

In parallel, the T RE-GFPhigh cells displayed markedly higher invasiveness than the

T RE-GFPlow cells in our Transwell matrix penetration assay (Fig. 6E), strongly suggesting

close correlations among miR-424-503 level, TGF- activity and cancer metastasis.

Furthermore, when we transiently expressed miR-424 ormiR-503 through transfection of

their mimic oligonucleotides in T RE-GFPlow cells, increased invasiveness was observed

(Fig. 6E). In contrast, inhibition of miR-424 or miR-503 with inhibitory oligonucleotides

decreased invasiveness of T RE-GFPhigh cells (Fig. 6E). Taken together, our data suggested

that the heterogeneous TGF- activities in breast tumors might be attributable to a differential

expression of the miR-424-503 cluster.




15

Discussion

The functional significance of TGF- signaling has been well documented in tumor

metastasis (5). TGF- activation-associated accumulation of phosphorylated Smad2 in the

nucleus drives formation of bone metastasis of human breast cancer cells (31). Clinical

correlations between plasma levels of TGF- and metastatic disease have been reported in

breast cancers, and low expression of TGF- receptors correlates with a favorable disease

outcome (9). The regulatory network that controls the activation level of TGF- signaling,

however, remains to be understood. The finding by this current study demonstrates a new

molecular mechanism via which TGF- is overactivated and thereby provides novel insights

in better understanding the regulatory network of TGF- signaling.

Negative regulators of TGF- activation, including Smad7 and Smurf2, contribute to

maintenance of an appropriate level of the signaling pathway under physiological conditions

(12-15). The level of Smad7 or Smurf2 has been found to be crucial in determining the

activation level of TGF- signaling (17,32), and thus understanding what supervise their

quantities in the context of cancer may reveal key mechanisms contributing to metastasis.

While previous studies identified loss-of-function mutations or deletions in human cancer

(33), Smad7 or Smurf2 expression was found to be regulated also at transcriptional or

posttranscriptional levels (34,35). Our current finding that Smad7 and Smurf2 protein

levels are simultaneously and directly suppressed by overexpressed miR-424-503 cluster in

metastatic breast cancer cells without alterations in their mRNA levels therefore represents a

novel posttranscriptional model for a coordinated regulation of these two factors. Although

miRNAs have been reported to target Smad7 (36,37), and miR-424 or miR-503

overexpression has been separately found to be associated with poor prognosis in breast

cancer (38,39), our results provides the first demonstration for miRNA modulation of Smurf2

level and for a simultaneous, efficient downregulation of both Smad7 and Smurf2 in breast




16

cancer. Importantly, this newly identified mode of action is clinically relevant, as shown by

our human specimens data. Thus, a coordinated miRNA modulation of both Smad7 and

Smurf2 appears to be an importantly new layer of TGF- signaling in breast cancer

metastasis. Further evaluation of anti-miR-424-503 strategies for their potential efficacies

in treating metastatic breast cancer is therefore warranted.

Over half of the miRNAs identified in mammalian cells are clustered in genomes and

transcribed as polycistronic primary transcripts, and distinct miRNAs within one cluster

could work in concert to regulate a cellular process (40). For example, two miRNAs

co-expressed from the miR-144/451 cluster both targeted CUGBP2 gene, and their

overexpression resulted in cardioprotective effect more potently than either one of them to a

synergistic extent (41). As found by the present study, both miR-424 and miR-503 targeted

the 3'UTRs of Smad7 and Smurf2 mRNA and overexpression of either one miRNA of the

cluster individually decreased the protein levels of Smad7 and Smurf2, but a combined

overexpression of miR-424 and miR-503 displayed more potent inhibitory effects.

Apparently, it would be of interest to elucidate whether they have overlapping biological

functions.

Interestingly, Oneyama reported that downregulation of miR-424-503 promotes mTORC2

formation and enhances growth and invasion of colon cancer through upregulating Rictor

(42). To test whether miR-424-503 changes Rictor expression in breast cancer, we analyzed

Rictor levels in MCF-7 and MDA-MB-231 cells overexpressing miR-424 and miR-503 and

found no detectable alteration in Rictor expression (data not shown), suggesting that these

miRNAs may have distinct functions in different cell types. Such discrepant effects of a

miRNA in different cell types are not uncommon (43-45), and the underlying molecular

mechanisms are still unclear. Particularly, it remains to be studied whether, and what, other

cell-specific genetic/epigenetic events are involved in determining the role of a miRNA in




17

modulating its target genes.

Our study also investigated the heterogeneity of TGF- activity in human breast cancer

and found biologically as well as clinically relevant correlations among TGF- activity,

miR-424-503 expression and metastatic potential in various subsets of tumor cells in a

population. As metastasis is a highly inefficient process, and usually only a small fraction

of cancer cells derived from the primary tumors can eventually form metastases in distant

organ sites (46), identifying cells or tumor tissues with high miR-424-503 expression might

be useful for predicting metastatic potential and for identifying the cell population with the

highest probability to form metastases in distant organ sites. Whether these tumors and cell

populations should represent the major targets for anti-metastasis therapies remains to be

further studied.

Apparently, additional work is required to elucidate the mechanisms underlying the

heterogeneous expression of miR-424-503 in breast cancer lesions. Interestingly, it was

reported that miR-424 could be regulated by PU.1, a member of Ets transcription factor

family (47), and hypoxia could regulate PU.1 to drive miR-424 expression (48). As breast

tumor cells are exposed to heterogeneous oxygen pressure (49,50), whether such

microenvironmental factors contribute to the heterogeneity of PU.1 levels, and consequently

to the heterogeneous TGF- activation, needs to be investigated in the context of breast

cancer metastasis.




18

Acknowledgements

This work was supported by the Ministry of Science and Technology of China Grant (No.

973-2011CB11305); The Natural Science Foundation of China (No. 81330058, 81103370,

41176128); Guangdong Recruitment Program of Creative Research Groups (No.

2009010058); National Science and Technique Major Project (No. 201305017,

2012ZX10004213, 311030).




19

Reference

1. Guttilla IK, Adams BD, White BA. ERalpha, microRNAs, and the epithelial-mesenchymal transition in

breast cancer. Trends Endocrinol Metab 2012;23:73-82.

2. Steeg PS. Tumor metastasis: mechanistic insights and clinical challenges. Nat Med 2006;12:895-904.

3. Wolfer A, Wittner BS, Irimia D, Flavin RJ, Lupien M, Gunawardane RN, et al. MYC regulation of a

"poor-prognosis" metastatic cancer cell state. Proc Natl Acad Sci U S A 2010;107:3698-703.

4. Cai J, Guan H, Fang L, Yang Y, Zhu X, Yuan J, et al. MicroRNA-374a activates Wnt/beta-catenin

signaling to promote breast cancer metastasis. J Clin Invest 2013;123:566-79.

5. Padua D, Massague J. Roles of TGFbeta in metastasis. Cell Res 2009;19:89-102.

6. Dalal BI, Keown PA, Greenberg AH. Immunocytochemical localization of secreted transforming

growth factor-beta 1 to the advancing edges of primary tumors and to lymph node metastases of

human mammary carcinoma. Am J Pathol 1993;143:381-9.

7. Muraoka RS, Koh Y, Roebuck LR, Sanders ME, Brantley-Sieders D, Gorska AE, et al. Increased

malignancy of Neu-induced mammary tumors overexpressing active transforming growth factor

beta1. Mol Cell Biol 2003;23:8691-703.

8. Buck MB, Fritz P, Dippon J, Zugmaier G, Knabbe C. Prognostic significance of transforming growth

factor beta receptor II in estrogen receptor-negative breast cancer patients. Clin Cancer Res

2004;10:491-8.

9. Biswas S, Guix M, Rinehart C, Dugger TC, Chytil A, Moses HL, et al. Inhibition of TGF-beta with

neutralizing antibodies prevents radiation-induced acceleration of metastatic cancer progression. J

Clin Invest 2007;117:1305-13.

10. Shi Y, Massague J. Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell

2003;113:685-700.

11. Yan X, Liu Z, Chen Y. Regulation of TGF-beta signaling by Smad7. Acta Biochim Biophys Sin (Shanghai)

2009;41:263-72.

12. Nakao A, Okumura K, Ogawa H. Smad7: a new key player in TGF-beta-associated disease. Trends Mol

Med 2002;8:361-3.

13. Hayashi H, Abdollah S, Qiu Y, Cai J, Xu YY, Grinnell BW, et al. The MAD-related protein Smad7

associates with the TGFbeta receptor and functions as an antagonist of TGFbeta signaling. Cell

1997;89:1165-73.

14. Kavsak P, Rasmussen RK, Causing CG, Bonni S, Zhu H, Thomsen GH, et al. Smad7 binds to Smurf2 to

form an E3 ubiquitin ligase that targets the TGF beta receptor for degradation. Molecular Cell

2000;6:1365-75.

15. Izzi L, Attisano L. Regulation of the TGFbeta signalling pathway by ubiquitin-mediated degradation.

Oncogene 2004;23:2071-8.

16. Wang P, Fan J, Chen Z, Meng ZQ, Luo JM, Lin JH, et al. Low-level expression of Smad7 correlates with

lymph node metastasis and poor prognosis in patients with pancreatic cancer. Ann Surg Oncol

2009;16:826-35.

17. Fukunaga E, Inoue Y, Komiya S, Horiguchi K, Goto K, Saitoh M, et al. Smurf2 induces

ubiquitin-dependent degradation of Smurf1 to prevent migration of breast cancer cells. The Journal

of Biological Chemistry 2008;283:35660-7.

18. Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell 2009;136:215-33.

19. Sachdeva M, Mo YY. MicroRNA-145 suppresses cell invasion and metastasis by directly targeting

mucin 1. Cancer Res 2010;70:378-87.




20

20. Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated by microRNA-10b in

breast cancer. Nature 2007;449:682-8.

21. Dunham JH, Guthmiller P. Doing good science: Authenticating cell line identity. Cell Notes

2008:15-17.

22. Peng Z, Omaruddin R, Bateman E. Stable transfection of Acanthamoeba castellanii. Biochim Biophys

Acta 2005;1743:93-100.

23. Jiang L, Lin C, Song L, Wu J, Chen B, Ying Z, et al. MicroRNA-30e* promotes human glioma cell

invasiveness in an orthotopic xenotransplantation model by disrupting the NF-kappaB/IkappaBalpha

negative feedback loop. J Clin Invest 2012;122:33-47.

24. Matsudaira P. Sequence from picomole quantities of proteins electroblotted onto polyvinylidene

difluoride membranes. J Biol Chem 1987;262:10035-8.

25. Tagliaferro P, Tandler CJ, Ramos AJ, Pecci Saavedra J, Brusco A. Immunofluorescence and

glutaraldehyde fixation. A new procedure based on the Schiff-quenching method. Journal of

Neuroscience Methods 1997;77:191-7.

26. Wang WX, Wilfred BR, Hu YL, Stromberg AJ, Nelson PT. Anti-Argonaute RIP-Chip shows that miRNA

transfections alter global patterns of mRNA recruitment to microribonucleoprotein complexes. RNA

2010;16:394-404.

27. Fend F, Emmert-Buck MR, Chuaqui R, Cole K, Lee J, Liotta LA, et al. Immuno-LCM: laser capture

microdissection of immunostained frozen sections for mRNA analysis. Am J Pathol 1999;154:61-6.

28. Kim J, Kang Y, Kojima Y, Lighthouse JK, Hu X, Aldred MA, et al. An endothelial apelin-FGF link

mediated by miR-424 and miR-503 is disrupted in pulmonary arterial hypertension. Nat Med

2013;19:74-82.

29. Lagna G, Hata A, Hemmati-Brivanlou A, Massague J. Partnership between DPC4 and SMAD proteins

in TGF-beta signalling pathways. Nature 1996;383:832-6.

30. Massague J. TGFbeta in Cancer. Cell 2008;134:215-30.

31. Kang Y, He W, Tulley S, Gupta GP, Serganova I, Chen CR, et al. Breast cancer bone metastasis

mediated by the Smad tumor suppressor pathway. Proc Natl Acad Sci U S A 2005;102:13909-14.

32. Javelaud D, Delmas V, Moller M, Sextius P, Andre J, Menashi S, et al. Stable overexpression of Smad7

in human melanoma cells inhibits their tumorigenicity in vitro and in vivo. Oncogene

2005;24:7624-9.

33. Levy L, Hill CS. Alterations in components of the TGF-beta superfamily signaling pathways in human

cancer. Cytokine Growth Factor Rev 2006;17:41-58.

34. Itoh S, Itoh F, Goumans MJ, Ten Dijke P. Signaling of transforming growth factor-beta family members

through Smad proteins. Eur J Biochem 2000;267:6954-67.

35. Ogunjimi AA, Briant DJ, Pece-Barbara N, Le Roy C, Di Guglielmo GM, Kavsak P, et al. Regulation of

Smurf2 ubiquitin ligase activity by anchoring the E2 to the HECT domain. Mol Cell 2005;19:297-308.

36. Bhagat TD, Zhou L, Sokol L, Kessel R, Caceres G, Gundabolu K, et al. miR-21 mediates hematopoietic

suppression in MDS by activating TGF-beta signaling. Blood 2013;121:2875-81.

37. Smith AL, Iwanaga R, Drasin DJ, Micalizzi DS, Vartuli RL, Tan AC, et al. The miR-106b-25 cluster

targets Smad7, activates TGF-beta signaling, and induces EMT and tumor initiating cell

characteristics downstream of Six1 in human breast cancer. Oncogene 2012;31:5162-71.

38. Pincini A, Tornillo G, Orso F, Sciortino M, Bisaro B, Leal Mdel P, et al. Identification of

p130Cas/ErbB2-dependent invasive signatures in transformed mammary epithelial cells. Cell Cycle

2013;12:2409-22.




21

39. Lerebours F, Cizeron-Clairac G, Susini A, Vacher S, Mouret-Fourme E, Belichard C, et al. miRNA

expression profiling of inflammatory breast cancer identifies a 5-miRNA signature predictive of

breast tumor aggressiveness. Int J Cancer 2013;133:1614-23.

40. Kim VN, Han J, Siomi MC. Biogenesis of small RNAs in animals. Nat Rev Mol Cell Biol 2009;10:126-39.

41. Zhang X, Wang X, Zhu H, Zhu C, Wang Y, Pu WT, et al. Synergistic effects of the GATA-4-mediated

miR-144/451 cluster in protection against simulated ischemia/reperfusion-induced cardiomyocyte

death. J Mol Cell Cardiol 2010;49:841-50.

42. Oneyama C, Kito Y, Asai R, Ikeda J, Yoshida T, Okuzaki D, et al. MiR-424/503-mediated Rictor

upregulation promotes tumor progression. PLoS One 2013;8:e80300.

43. Betel D, Koppal A, Agius P, Sander C, Leslie C. Comprehensive modeling of microRNA targets predicts

functional non-conserved and non-canonical sites. Genome Biol 2010;11:R90.

44. Li J, Donath S, Li Y, Qin D, Prabhakar BS, Li P. miR-30 regulates mitochondrial fission through targeting

p53 and the dynamin-related protein-1 pathway. PLoS Genet 2010;6:e1000795.

45. Yu F, Deng H, Yao H, Liu Q, Su F, Song E. Mir-30 reduction maintains self-renewal and inhibits

apoptosis in breast tumor-initiating cells. Oncogene 2010;29:4194-204.

46. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving paradigms. Cell

2011;147:275-92.

47. Rosa A, Ballarino M, Sorrentino A, Sthandier O, De Angelis FG, Marchioni M, et al. The interplay

between the master transcription factor PU.1 and miR-424 regulates human monocyte/macrophage

differentiation. Proc Natl Acad Sci U S A 2007;104:19849-54.

48. Ghosh G, Subramanian IV, Adhikari N, Zhang X, Joshi HP, Basi D, et al. Hypoxia-induced

microRNA-424 expression in human endothelial cells regulates HIF-alpha isoforms and promotes

angiogenesis. J Clin Invest 2010;120:4141-54.

49. Vaupel P. Tumor microenvironmental physiology and its implications for radiation oncology. Semin

Radiat Oncol 2004;14:198-206.

50. Helmlinger G, Yuan F, Dellian M, Jain RK. Interstitial pH and pO2 gradients in solid tumors in vivo:

high-resolution measurements reveal a lack of correlation. Nat Med 1997;3:177-82.




22

Figure Legends

Figure 1. The miR-424-503 cluster expression is upregulated in metastatic breast cancer.

A. Expression profiles of miRNA were obtained from TCGA database and the expression

levels of miR-424 and miR-503 were analyzed in patients with (n = 31) and without

metastasis (n = 372). B. qPCR analysis of the expression levels of miR-424 and miR-503 in

tumors with (n = 24) and without metastatic relapse (n = 93). C. Relative miR-424 and

miR-503 expression in breast cancer cell lines were analyzed by qPCR. Relative expression

levels were normalized by U6 expression. **: P < 0.01. *: P < 0.05.

Figure 2. The miR-424-503 cluster enhances breast cancer metastasis in vitro.

A. Wound healing assay was performed with indicated cells. B. Representative images (top)

and quantification (bottom) of penetrated cells were analyzed using the Transwell matrix

penetration assay. C. Representative micrographs of indicated cultured cells at day 7 of

culture in three-dimensional spheroid invasion assay. **: P < 0.01.

Figure 3. The miR-424-503 cluster promotes metastasis of breast cancer cells in vivo.

A. and B. Nude mice injected intravenously via the caudal vein with indicated cells were

monitored with luciferase live-imaging system at indicated time points (top) and the photon

quantity in each mice group were analyzed (bottom). Heat map scale bar represents photon

emission. C. Survival curves of mice injected intravenously with indicated cells. **: P <

0.01.

Figure 4. The miR-424-503 cluster targets Smad7 and Smurf2 expression.

A. Conserved miR-424 and miR-503 binding sites in Smad7 and Smurf2 generated by

TargetScan analysis. B. Western blotting analysis of Smad7 and Smurf2 expression in

indicated cells. -Tubulin was used as a loading control. C. Relative activity of reporter




23

luciferase linked to Smad7 or Smurf2 3’UTR measured following miR-424 mimic, miR-503

mimic, miR-424 inhibitor or miR-503 inhibitor transfection in indicated cells. D. Relative

activity of reporter luciferase linked to mutant Smad7 3’UTR or mutant Smurf2 3’UTR

measured following miR-424 mimic or miR-503 mimic in indicated cells. E. Enrichment of

Smad7 and Smurf2 3’UTRs in miRNP detected by qPCR following immunoprecipitation

against Ago1. **: P < 0.01. *: P < 0.05.

Figure 5. Smad7 and Smurf2 are functionally involved in miR-424-503 cluster induced

metastasis.

A. Western blotting analysis of Smad7 and Smurf2 expression in indicated cells. -Tubulin

was used as a loading control. B. Representative images (top) and quantification (bottom)

of penetrated cells were analyzed using the Transwell matrix penetration assay. C. Mice

injected intravenously with indicated cells were monitored by luciferase live imaging system

at indicated time points (top) and the photon quantity in each mice group were analyzed

(bottom). Heat map scale bar represents photon emission. D. Expression of miR-424-503

cluster was associated with p-Smad3, Smad7 and Smurf2 expression levels in clinical breast

cancer specimens. Two representative cases were shown (left panel). Patient specimens

stained for p-Smad3 showed nuclear positive (red arrow) and negative (green arrow) signals.

Percentage of specimens showing low or high miR-424-503 cluster expression in relation to

the expression levels of p-Smad3, Smad7 and Smurf2 (right panel). **: P < 0.01. *: P <

0.05.

Figure 6. The miR-424-503 cluster contributes to heterogeneous TGF- activity levels.

A. Top panels: Schematic representation of the lentiviral vectors CMV-GFP-SV40-RFP

(constitutively expressing RFP and GFP) and TGF- -Smads-GFP-SV40-RFP (expressing

GFP when stimulated by TGF- ). Bottom panels: fluorescence-activated cell sorting (FACS)




24

profile of GFP intensity range for single-cell-cloned T RE-GFP-carrying cells (green line,

middle) compared with the sharp range of GFP in the CMV-GFP (green line, right) cells.

Non-transduced parental cell culture was used as a control for GFP background intensity

(black lines). The lowest and highest 10% of T RE-GFP cell fractions were collected,

labeled as T RE-GFPlow and T RE-GFPhigh, respectively (left). B. qPCR analysis of PAI1

expression in T RE-GFPlow and T RE-GFPhigh cells. Relative expression levels were

normalized by GAPDH expression. C. Western blotting analysis of p-Smad2, Smad2, p-

Smad3, Smad3, Smad7 and Smurf2 levels in indicated cells. -Tubulin was used as a

loading control. D. qPCR analysis of miR-424 and miR-503 expression in T RE-GFPlow

and T RE- GFPhigh cells. Relative expression levels were normalized by U6 expression. E.

Representative images of penetrated cells analyzed using the Transwell matrix penetration

assay. **: P < 0.01.




Published OnlineFirst August 27, 2014.Cancer Res Yun Li, Wei Li, Zhe Ying, et al. miR-424-503 gene cluster

-activatingβwith expression of a heterogeneous TGF-Metastatic heterogeneity of breast cancer cells is associated

Updated version

10.1158/0008-5472.CAN-14-0389doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2014/08/27/0008-5472.CAN-14-0389.DC1

Access the most recent supplemental material at:

Manuscript

Authoredited. Author manuscripts have been peer reviewed and accepted for publication but have not yet been

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/early/2014/08/23/0008-5472.CAN-14-0389To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-14-0389

http://cancerres.aacrjournals.org/content/suppl/2014/08/27/0008-5472.CAN-14-0389.DC1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2014/08/23/0008-5472.CAN-14-0389


Metastatic heterogeneity of breast cancer cells is...

Documents

Transcript of Metastatic heterogeneity of breast cancer cells is...