miR-17 Family miRNAs are expressed during early mammalian development
and regulate stem cell differentiation
Kara M. Foshay, G. Ian Gallicano
To appear in:
9 May 2008
14 November 2008
17 November 2008
Please cite this article as: Foshay, Kara M., Gallicano, G. Ian, miR-17 Family miRNAs
are expressed during early mammalian development and regulate stem cell differentiation,
Developmental Biology (2008), doi:10.1016/j.ydbio.2008.11.016
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
miR-17 Family miRNAs are Expressed During Early Mammalian
Development and Regulate Stem Cell Differentiation
Kara M. Foshay, G. Ian Gallicano
Department of Biochemistry & Molecular & Cellular Biology, Georgetown University
Medical Center, Washington, DC 20007
Running Title: MicroRNAs in Embryonic Development and Stem Cells
G. Ian Gallicano
Department of Biochemistry & Molecular & Cellular Biology
Georgetown University Medical Center
Medical Dental Building, Rm NE205
Washington, DC 20007
MicroRNAs are small non-coding RNAs that regulate protein expression by
binding 3’UTRs of target mRNAs, thereby inhibiting translation. Similar to siRNAs,
miRNAs are cleaved by Dicer. Mouse and ES cell Dicer mutants demonstrate that
microRNAs are necessary for embryonic development and cellular differentiation.
However, technical obstacles and the relative infancy of this field have resulted in few
data on the functional significance of individual microRNAs. We present evidence that
miR-17 family members, miR-17-5p, miR-20a, miR-93, and miR-106a, are differentially
expressed in developing mouse embryos and function to control differentiation of stem
cells. Specifically, miR-93 localizes to differentiating primitive endoderm and
trophectoderm of the blastocyst. We also observe high miR-93 and miR-17-5p
expression within the mesoderm of gastrulating embryos. Using an ES cell model
system, we demonstrate that modulation of these miRNAs delays or enhances
differentiation into the germ layers. Additionally, we demonstrate these miRNAs
regulate STAT3 mRNA in vitro. We suggest that STAT3, a known ES cell regulator, is
one target mRNA responsible for the effects of these miRNAs on cellular differentiation.
et al., 2003; Kanellopoulou et al., 2005; Krichevsky et al., 2006; Murchison et al., 2005;
MicroRNAs (miRNAs) are short, non-coding RNAs that negatively regulate
target mRNAs via binding to their 3’ untranslated regions (UTRs). Although miRNA
silencing is mediated through the RNA-induced silencing complex (RISC), its binding
does not generally result in mRNA cleavage, as is the case with siRNA. The miRNA-
RISC association results in the formation of a bulge within the miRNA, preventing
cleavage and explaining the relatively low level of target complimentarity necessary for
silencing (Chu and Rana, 2006). Specifically, Watson-Crick base pairing between the
target mRNA and nucleotides 2-8 at the 5’ end of the miRNA, frequently referred to as
the “seed region,” is generally believed to determine miRNA binding potential (Lewis et
al., 2003). Following miRNA binding, mRNAs are silenced either by storage in P-bodies
or degradation via mRNA decay pathways (Gregory et al., 2005; Rana, 2007; Zeng et al.,
Newly published literature has implicated miRNA-mediated silencing as having
important regulatory functions during embryonic development and embryonic stem (ES)
cell proliferation and differentiation (Boyer et al., 2005; Hatfield et al., 2005; Houbaviy
Shcherbata et al., 2006; Wang et al., 2007; Zhao et al., 2005). The miR-17 family of
miRNAs, which is expressed as three polycistronic clusters, is highly conserved
throughout species and is thought to have evolved along with vertebrates (Tanzer and
Stadler, 2004). Through evolution, a series of duplications, deletions and mutations have
given rise to the modern miR-17 family, which consists of 14 mature miRNAs located on
chromosomes 13, X and 7 in humans. Hinting at the importance of these miRNAs, these
Additionally, several members of the miR-17 family, including miR-17-5p, miR-
14 genes also exist in the same order and as three clusters in mouse, rat, and chimp
(Tanzer and Stadler, 2004).
To date few miRNAs have been specifically investigated during embryonic
development (Ventura et al., 2008; Zhao et al., 2007). When deleted via gene targeting,
miR-1-2 null mice showed developmental defects including disrupted cardiac
morphogenesis and failed electrical conduction in cardiomyocytes. Most recently,
deletion of miR-17 family members miR-92, 106a, and 106b also revealed developmental
defects (Ventura et al., 2008). In addition, new reports have shown these miRNAs to be
involved in zebrafish development (Giraldez et al., 2005), cancer progression (Dews et
al., 2006; He et al., 2005), and hematopoietic cell differentiation (Garzon et al., 2006).
Within zebrafish, miR-430, a member of the miR-17 family, was shown to be important
for development of both the heart and hindbrain (Giraldez et al., 2005). A second study
examined the role of c-myc in the regulation of expression of the miR-17 cluster
(O'Donnell et al., 2005). They demonstrated that c-myc directly activated transcription of
miR-17 family microRNAs, which then bound to upstream regulators and downstream
targets of myc.
20a, miR-93, and miR-106a have been identified as microRNAs that are specifically
expressed in undifferentiated or differentiating embryonic stem cells (Houbaviy et al.,
2003; Suh et al., 2004; Tang et al., 2006). Interestingly, these same 4 microRNAs, which
exhibit high sequence similarity (Fig. 1A), were among those identified by at least 3
different algorithms as possible binding partners to the 3’-UTR of the Signal Transducer
and Activator of Transcription 3 (STAT3) (Fig. 1B). Based on the current miRNA
(SOCS) and Protein Inhibitors of Activated STATs (PIAS) proteins (Horvath, 2000;
literature and the knowledge that STAT3 interacts with myc in ES cells (Cartwright et al.,
2005), hematopoiesis (Hirano et al., 2000), and cancer (Barre et al., 2005), we
hypothesize that miR-17 family members are expressed during embryonic development
and regulate cellular differentiation by targeting of mRNAs such as STAT3.
STAT3, a downstream transcription factor in the JAK-STAT signal cascade, is an
important player in both developmental and stem cell biology. Surprisingly, few
investigations exist describing STAT3 prior to implantation (Antczak M, and Van
Blerkom J. 1997; Duncan et al., 1997); however, STAT3 knockout mice are embryonic
lethal just prior to gastrulation at E6.0-6.5, and exhibit functional failure of the visceral
endoderm and a lack of mesoderm formation (Takeda et al., 1997). While this
developmental lesson alludes to the important role STAT3 plays in many tissue types, the
constitutive activation of STAT3 in several cancers demonstrates that this protein must
also be tightly regulated (Catlett-Falcone et al., 1999; Corvinus et al., 2005; Nefedova et
al., 2004). STAT3 is negatively regulated by at least 4 known mechanisms, including
various phosphatases, expression of its endogenous dominant negative isoform, STAT3ß,
and negative feedback loops created by expression of Suppressors of Cytokine Signaling
Inagaki-Ohara et al., 2003). Previous work from our laboratory and others has
demonstrated that several of these mechanisms operate simultaneously at the onset of ES
cell differentiation to induce a rapid and drastic decrease in STAT3 activity (Chan et al.,
2003; Feng, 2007; Foshay and Gallicano 2008; Foshay et al., 2005; Li et al., 2005). Our
previous data suggest that the degree and duration of this period of STAT3 inactivity may
affect cell fate choices over the course of differentiation. Thus, as STAT3 regulation by
induced by removal of LIF and transfer to suspension culture.
several different methods appears to be critical for proper ES cell differentiation, we
hypothesize that miRNAs may exert their effects on differentiation through this key
Although as many as 50 miRNAs are predicted to bind the 3’-UTR of STAT3,
there are only two reports of miRNA-mediated regulation of STAT3 in the current
literature. In both cases, the investigators demonstrate that levels of STAT3
phosphorylation can be indirectly affected by miRNAs in neural cell differentiation and
in cancer (Krichevsky et al., 2006; Meng et al., 2007). However, our data is the first to
elucidate a STAT3/miR-17 family member(s) interaction, which results in a functional
regulation of ES cell differentiation.
Materials and Methods
ES cell culture and differentiation
CCE ES cells (Keller et al., 1993; Robertson et al., 1986), obtained from Stem
Cell Technologies, Inc., were grown in feeder free conditions with media containing 15%
ES cell qualified FBS (Invitrogen, Carlsbad, CA). Differentiation of ES cells was
Blastocysts and implanted embryos were obtained as described in Gallicano and
Capco (1995) and Gallicano et al., (1998). Briefly, one female is placed into a cage with
one male to allow copulation. Females are checked the next day for vaginal plugs, which
represents 0.5 days post coitum. Eighty four hrs later, the mouse is sacrificed by CO2
is simply the opposite of the technique used to inject blastocysts with ES cells. Once
asphyxiation, the female reproductive tract is removed, and blastocysts are flushed from
the uterus using a 26 ¾ gauge needle attached to a 10cc syringe filled with medium. For
in situ hybridization and confocal microscopy, blastocysts were transferred to 4.0%
paraformaldehyde fixative for 1 hr.
Implanted embryos were obtained from fertilized dames posthumously 5.5 days,
6.0 days, and 6.5 days post coitum by dissection from decidua and using Dumont #5
forceps. Isolated embryos were then placed in 4.0% para-formaldehyde and processed
for either in situ hybridization or immunohistochemistry.
Removal of cells from blastocysts
Blastocysts were used to extract and isolate cells from three areas, trophectoderm,
ICM, and presumptive primitive endoderm. To do so, blastocysts were placed in a watch
glass containing 0.25% trypsin for 5-10 minutes until the cells comprising the blastocyst
began dissociating. A pulled glass needle similar to those used for ES cell injection into
blastocysts was used to extract three cells from each region of the blastocyst using an
Eppendorf TransferMan NK2 micromanipulator set on negative pressure. This technique
three cells were removed they were then placed into a microcentrifuge tube containing
the qRT-PCR buffer supplied by the mirVana miRNA Isolation Kit (Ambion, Austin,
miRNA Binding Predictions
Immunofluorescence and Confocal Microscopy
Embryos were fixed in 4% fresh paraformaldehyde for 30 minutes followed by
permeabilization overnight at 4°. Following blocking for 1 hour, embryos were incubated
in primary antibody at a 1:50 dilution overnight at 4°. Incubation with secondary
antibodies was conducted for 3 hours at room temperature. Antibodies: STAT3 #9132
from Cell Signaling (Danvers, MA); CK8/18 #03-GP11 from ARP (Belmont, MA).
Slides were viewed using an Olympus Fluoview 500 Laser Scanning Microscope
(Olympus America Inc, Melville, NY) using 1.4 numerical aperture. Images were
acquired and analyzed using the accompanying Fluoview software (version 4.3).
In Situ Hybridization
In situ hybridization probes were obtained from Exiqon (Woburn, MA) and
experiments were performed according to the GEISHA protocol (from the University of
Arizona) available on the Exiqon website. All reagents and apparatus used were DEPC
treated. Hybridizations were conducted overnight at 58° C.
Binding predictions were made by comparing results of three different miRNA
prediction programs, each of which utilizes a different prediction algorithm. The
miRBase website (http://microrna.sanger.ac.uk/targets/v4/), run by the Sanger Institute,
employs the miRanda algorithm to predict miRNA:mRNA pairs. MicroInspector,
another web-based miRNA search program (http://mirna.imbb.forth.gr/microinspector/),
run by the Institute for Molecular Biology and Biotechnology, Heraklion, Greece, uses a
Biosystems ABI Prism 7700 system using SDS 2.1 software. Relative gene expression
different algorithm that allows for G:U wobbles in seed matches. Finally, we used a
custom made miRNA prediction program, written by Bill Foshay, which utilizes a
variation of the TargetScanS algorithm (Lewis et al., 2005) and components of the
Vienna RNA secondary
structure programming library (RNAlib)
which obtains RNA energy
parameters from the Turner laboratory (http://rna.chem.rochester.edu/). The PicTar
miRNA website (http://pictar.bio.nyu.edu/) was also used for generating Table 2.
All RNA samples were prepared by extraction with the mirVana miRNA Isolation
Kit (Ambion, Austin, TX). Samples were treated with DNaseI to digest any
contaminating genomic DNA and diluted to 10ng/uL. All primers were designed by and
purchased from Applied Biosystems (Foster City, CA). All RT and qPCR reagents were
also purchased from Applied Biolosystems. No-RT controls were performed for both
miRNA and mRNA qRT-PCRs. Individual samples were run in triplicate, and each
experiment was repeated at least 3 times. All samples were run on the Applied
was calculated using the 2-∆∆CT method (Livak and Schmittgen, 2001).
Western blots were performed using precast 7.5% Tris-HCL polyacrylimide gels
and run on the BioRad mini gel system (Hercules, CA). Proteins were transferred to
PVDF membranes and blocked in a solution of 2% BSA and 5% milk in PBS-T. The
are as follows: pMIR-miR20 contains two synthesized binding sites that are
STAT3 antibody, available from Cell Signaling (Danvers, MA), was used at a 1:1000
dilution. Bands were visualized using colorimetric detection and exposure to
autoradiography film. Similar exposure times were used for STAT3 and Tubulin blots to
ensure appropriate normalization. Quantification of bands was performed using Adobe
Photoshop CS to calculate the pixel intensity of each band. Pixel intensities for STAT3
and Tubulin were expressed as a ratio to generate a normalized value for STAT3
DNA Constructs and Transient Transfections
All transient transfections were performed using the TransIT-LT1 or TransIT-
TKO reagents from Mirus Bio Corporation (Madison, WI). Transfection reactions were
prepared according to protocols provided with the reagent. For differentiation
experiments, undifferentiated ES cells were transfected and 24 hrs. post-transfection (d0)
cells were harvested or placed into suspension culture for differentiation. The pMIR-
REPORT and pRL-CMV constructs are commercially available from Ambion
(Austin,TX) and Promega (Madison, WI), respectively. pMIR-REPORT 3’UTR inserts
complimentary to miR-20, pMIR-Hoxa11 contains the Hoxa11 3’UTR, pMIR-S3
contains part of the STAT3 3’UTR including the two putative miRNA binding sites,
pMIR-S3st1mt contains the STAT3 3’UTR with 4 point mutations in the first miRNA
binding site, pMIR-S3st2mt contains 4 point mutations in the second miRNA binding
site, pMIR-S32xmt contains 4 point mutations in each of the two miRNA binding sites.
All miRNA mimics and inhibitors were designed by and purchased from Dharmacon
miR-17 family miRNAs are differentially expressed in developing mouse embryos
(Lafayette, CO). The mimics are duplexes and therefore the concentration of active
miRNA mimic (strand loaded into RISC) within the cell is about 50nM or half of the total
transfected concentration (100nM). Inhibitors are single stranded.
Luciferase assays were conducted following transient transfections of pMIR and
pRL-CMV constructs at a 15:1 ratio. Cells were harvested according to specifications of
the Dual Luciferase Reporter Assay Kit (Promega) and 20uL of the lysate was transferred
to each well of a 96 well plate for analysis. For experiments in undifferentiated ES cells,
lysates were prepared 48 hrs. post-transfection. For differentiation experiments,
undifferentiated cells were transfected and placed into suspension culture without LIF 24
hrs. post-transfection (d0). The luciferase reactions were conducted using the Wallac
VICTOR2 96-well plate reader at the Lombardi Cancer Center Shared Resource. All
luciferase data are presented as a normalized ratio of luciferase/renilla.
Based on our hypothesis that miR-17 family miRNAs are important for
embryonic development we sought to determine their expression patterns in early mouse
embryos. Because this family of miRNAs is transcribed from the genome as three
polycistronic clusters, which are regulated by c-myc, we expected to find the mature
miR-17 family miRNAs expressed at similar levels and within the same regions of the
miRNA expression patterns between family members, our observations fit well with our
In our first set of experiments, we identified the expression patterns of miR-17-5p,
miR-20a, miR-93 and miR-106 in E4.0 blastocysts by in situ hybridization (ISH) using
locked nucleic acid (LNA) miRNA specific probes. Despite their small size and
sequence similarity, the highly stable LNA probes allowed us to use increased
hybridization temperatures relative to conventional in situ protocols (Kloosterman et al.,
2006; Nelson et al., 2006). As a result, we determined distinct embryonic localization
patterns for each miRNA. Specifically, miR-17-5p and miR-20 were seen throughout the
blastocyst, with a slight increase in hybridization seen within cells of the trophectoderm
(Fig. 2A). Low levels of miR-106a signal were also observed throughout blastocysts;
however, elevated levels of miR-106a hybridization were clearly evident in the inner cell
mass (ICM) (Fig. 2A). The most striking expression pattern was that of miR-93. This
miRNA was restricted to the trophectoderm and the future primitive endoderm.
Additionally, miR-93 expression was barely detectable within the ICM (Fig. 2A). We
also noticed that the signals for miR17-5 and miR-20 were scattered throughout the
cytoplasm. However the signal for miR-93 was enriched to one side of the cytoplasm in
select cells. Although we were surprised to see both subtle as well as distinctly different
original hypothesis; that the highest miRNA expression would be found in differentiating
tissues. These data suggest that miR-17 family miRNAs may play a role in embryonic
development and differentiation.
Because we hypothesized that these miRNAs may exert their effects via STAT3,
we next looked at STAT3 expression in the blastocysts. As expected, STAT3 expression
was specific to the ICM with little or no staining seen in the trophectoderm and the
showed high levels of specific nuclear STAT3 staining throughout the epiblast,
presumptive primitive endodermal ICM cells that line the blastocoelic cavity (Fig. 2B).
This expression pattern was opposite to that of miR-93, suggesting that endogenous miR-
93 may downregulate STAT3 in specific cells of the blastocysts.
We continued our investigation of miR-17 family expression patterns, conducting
in situ hybridization in E5.5 and E6.5 embryos (Figs. 3 and 4). At E5.5 we saw strong
hybridization of all probes in the extraembryonic cells (Fig. 3A-D). These cell types
include the extraembryonic ectoderm and the visceral endoderm, which, at E5.5, are
differentiating into specialized tissues that will form the yolk sac and placenta, and
facilitate nutrient and gas exchange for the embryo. In contrast, the primitive embryonic
ectoderm, or epiblast, is still undifferentiated at this stage. These cells, which are rapidly
proliferating in preparation for gastrulation, showed little or no expression of miR-17-5p,
miR-20a, miR-93, or miR-106a. These data correlate with the blastocyst ISH data (Fig.
2) and demonstrate the idea that miR-17 family miRNAs are upregulated in
differentiating cell types of the developing embryo.
In addition to the in situ studies, we performed immunohistochemistry for STAT3
expression on paraffin embedded sections of E5.5 embryos (Fig. 3E). These sections
suggesting that STAT3 is active in this area. We detected little or no STAT3 in the
visceral endoderm and low levels in the extraembryonic ectoderm. Taken together, these
data present an expression pattern for STAT3 that is opposite of miR-17 family
expression in E5.5 embryos (Fig. 3A-D, E).
At E6.5, embryos showed highly specific miRNA expression patterns. All four of
the miRNA investigated were localized to the posterior end of the primitive streak (Fig.
has low expression of all tested miRNAs, miR-20a and miR-106a are expressed at
4A-D). Interestingly, the hybridization was restricted to ectoderm cells at least one cell
layer removed from the primitive streak. Thus, the posterior primitive streak ectodermal
cells bordering the mesoderm did not express any of the miRNAs. However, gastrulating
cells, which were in the process of differentiating into mesoderm or mesendoderm
expressed high levels of miR-93 and miR-17-5p (Fig. 4A-B). Although, in some cases
brightly positive cells were observed within the endoderm, generally the visceral
endoderm had low miRNA expression. These cells may be intercalating mesendoderm
cells that will form the definitive endoderm. From these data we suggest that miR-17
family miRNAs are expressed in cells that are approaching the primitive streak, followed
by virtually complete downregulation as they prepare to pass through the primitive
streak. As these cells engress and differentiate, miR-17-5p and miR-93 are upregulated,
most likely downregulating translation of mRNAs that are specific to the ectoderm or are
involved in proliferation. The patterns of ISH and STAT3 expression from blastocysts,
E5.5 and E6.5 embryos are summarized in Table 1.
Of note is the drastic difference in miRNA expression within extraembryonic and
embryonic visceral endoderm of E6.5 embryos. While the embryonic visceral endoderm
relatively high levels in the extraembryonic visceral endoderm (Fig. 4C-D, eve). The
hybridization patterns are extremely specific with miRNAs being detected specifically in
the cytoplasm and at the periphery of cells. In fact, the lack of signal in the nucleus for
any of the miRNA probes would suggest that they are relatively specific for the
processed, mature miRNA and not the unprocessed Pri-mRNA.
miRNA expression as cells of the blastocyst differentiated into primitive endoderm and
In addition to the in situ studies, we performed immunohistochemistry for STAT3
expression on paraffin embedded sections of E6.5 embryos (Fig. 4F). While these
sections showed high levels of specific nuclear STAT3 staining in the ectoderm (ect),
cells near the posterior primitive streak appear devoid of STAT3, suggesting that STAT3
is inactive in this area. We detected little or no STAT3 in nuclei of visceral endoderm
and low levels in the extraembryonic ectoderm.
miRNAs -17, 20a, -93 and 106a are expressed in blastomere and ES cells
To test our hypothesis that miR-17-5p, miR-20a, miR-93 and miR-106a are
involved in regulation of differentiation during embryonic development we used two
model systems; isolated cell-types from blastocysts and ES cells. By conducting our
studies in vitro we could easily manipulate individual miRNAs and assess their effects on
cell fate. We first employed qRT-PCR to quantify mature miRNA expression in
blastocysts from isolated trophoblast cells, pleuripotent ICM, and ICM destined to
become primitive endoderm. We also used qRT-PCR to analyze miRNA expression in
ES cells (Fig. 5). Using this approach, we were able to quantitatively compare changes in
It is important to note how these data were controlled for this set of experiments.
All qRT-PCR samples were normalized to qRT-PCR of U6 snRNA using the delta-delta
Ct method. Then, to make the data easier to interpret, one sample (generally the one with
the lowest overall value) was picked and set to 1 serving as the baseline. The other
samples were then compared to the baseline sample for each graph in figure 5. In Figure
was the lowest (Fig. 5B). Despite the differences between baseline miRNA expression in
5A, miR-20 was set as the baseline and expression of the other miRNAs is compared to
miR-20. In figure 5B, miRNA-106a was set as the baseline and expression of the other
miRNAs is compared to miRNA-106a. The baselines were set differently because
miRNA expression was different when comparing ES cells to blastocysts (see below).
Within the blastocyst, miR-93 was expressed at levels higher than any other tested
miRNA (Fig. 5A). Interestingly upon extraction and analyzing three cells per experiment
from the three areas within the blastocyst, miR-93 was the only miRNA that was
significantly upregulated in the putative primitive endoderm and trophectoderm. Thus,
these qRT-PCR data supported the results previously found using in situ hybridization
(Fig. 2). To this end, miR-17 and miR-20 expression levels were relatively consistent
across cell types, while miR-106a was highest in the ICM. Since these data matched our
in situ hybridization data (Fig. 2), we determined that qRT-PCR would be the best
method for measuring miRNA expression in ES cells.
Surprisingly the miRNA expression profile of ES cells was different than that of
blastocyst ICM cells. In cultured ES cells, miR-20 expression was higher than any other
tested miRNA, miR-17-5p and miR-93 displayed equal expression levels, and miR-106a
ES cells and ICM cells, the upregulation of these miRNAs during differentiation was
consistent (Fig. 5C). Specifically, when compared to miR-93 expression in
undifferentiated ES cells, miR-93 expression nearly doubled during the first three days of
ES cell differentiation. While miR-106a expression also increased, this occurred
approximately 2 days later than the miR-93 increase, suggesting miR-106 may play a
both cultured ES cells and in developing blastocysts. These data suggest that miR-93 is
different role in differentiation. Both miR-20 and miR-17 exhibited slight increases over
the course of differentiation, thus supporting the blastocyst qRT-PCR and in situ data.
Differentiation of individual blastomeres within blastocysts occurs more
synchronously than differentiation of individual cells within an EB. Therefore, we
surmise that the lower fold increase in miRNA expression in ES cells when compared to
blastomeres is due to decreased temporal regulation of differentiation. Based on this
comparison of ES and blastocyst differentiation, we have demonstrated that differences
between the in vivo niche and cell culture conditions may alter basal miRNA expression
levels. This could be due to the mix of growth factors present in the fetal bovine serum
used to culture ES cells. It is also possible that in ES cell culture Drosha functions
differently than in embryos, and thus changes in miRNA expression may be due to
differential post-transcriptional processing (Thomson et al., 2006). Alternatively,
differences in basal miRNA expression could be attributed to the fact that ES cells in
culture do not interact with a supportive trophectodermal niche.
Although we found differences between miRNA expression in vivo and in vitro,
we demonstrated a consistent upregulation of miR-93 upon initiation of differentiation in
an important miR-17 family member for regulating cellular differentiation.
Changes in miRNA expression can alter cell fate during ES cell differentiation
To test the hypothesis that miR-17 family miRNAs affect cell fate during
differentiation, we transfected mimics or inhibitors of miR-93 and miR-20 into
undifferentiated ES cells (a non-specific inhibitor/mimic was used as a negative control).
remained low even after several days of culture in differentiating conditions (Fig. 6B).
This transient transfection was followed by a second transfection once differentiation had
begun (day 2; 72 hrs after first transfection). The cells were allowed to differentiate as
EBs in suspension culture for up to 7 days (day 7). RNA was harvested at days 0, 3 and 7
and was then used for qRT-PCR analysis of germ layer specific differentiation markers
(Fig. 6). Levels of mature miRNA within the cell were also measured by qRT-PCR in an
effort to demonstrate the effectiveness of the transfection (Fig. S1). Because of the large
scale of this experiment, we chose only to use inhibitors to miR-20 and miR-93. Our
rationale stemmed from the strong in vivo data suggesting that miR-93 was the most
likely candidate regulator of stem cell differentiation. As expression of the other
miRNAs seemed less specific to differentiating cells, we chose to examine only miR-20,
which we felt was representative of the other 3 miRNAs.
Upon transfection of inhibitors we saw a statistically significant decrease in Fgf-5
expression when compared to controls only at d0. These data suggest that ectodermal
differentiation was not affected by the miRNA inhibitors (Fig. 6A). However, miR-93
inhibitors disrupted endodermal differentiation. In these cells Hnf4a levels were
significantly lower at the onset of differentiation (24 hours post transfection) and
Interestingly, the inhibitors also delayed the expression of the mesodermal differentiation
marker, Brachyury, (Fig. 6C, d3). Based on these data we suggest that miRNAs exhibit
different effects on different germ layers. As evidenced from our in vivo ISH data, these
miRNAs are quite low in the ectoderm or epiblast (Fig. 3) and therefore their inhibition
may not drastically effect differentiation. Additionally, our in vitro data support our in
outer layer of visceral endoderm that is commonly found in differentiating EBs (Fig. 6,
vivo data and suggest that miR-93 is, in fact, a more potent inducer of differentiation than
miR-20 or other miR-17 family members.
When the opposite experiment was conducted, and miR-93 mimics were
transfected, we saw a significant increase in Brachyury expression (Fig. 6F). However,
this increase was not observed upon transfection of miR-20 mimics. Again, these data
support a functional role for miR-93 in regulation of ES cell differentiation.
Additionally, these data suggest that upregulation of miR-93 may promote mesodermal
differentiation. This idea is supported by the strong mesodermal expression of miR-93 in
gastrulating embryos (Fig. 4B). Taken together, our data demonstrate the first
functionally significant role for miRNAs in regulation of ES cell differentiation in vitro.
In addition to qRT-PCR analysis, we observed the effects of miR-93 mimics and
inhibitors on EBs using H&E staining (Fig. 6G-I). Following the same procedure
outlined above, transfected cells were allowed to differentiate for up to 10 days in culture.
EBs were harvested, fixed, and sectioned for staining. EBs transfected with miR-93
inhibitors were comprised of homogeneous cells, as compared to controls or miR-93
mimic transfected cells. In addition, EBs transfected with miR-93 inhibitors lacked the
black arrows). These histological data agree with our qRT-PCR data and support the
conclusion that inhibition of miR-93 prevents differentiation of ES cells.
miR-20 and miR-93 can bind the STAT3 3’UTR and decrease STAT3 expression
Data from our previous studies demonstrated that STAT3 is important not only
for pluripotency, but also for differentiation of mesodermally and ectodermally derived
cells (Foshay, In Press; Foshay et al., 2005). Additionally, miR-93 and other miR-17
family members are predicted to target the 3’UTR of STAT3 (Fig. 1B). The results of
the mimic and inhibitor transfection experiments could be explained by miRNA mediated
regulation of STAT3, and therefore lend support to our hypothesis that STAT3 is a
functional miR-17 family target.
To further test our hypothesis that miR-17 family miRNAs can target STAT3, we
constructed reporter vectors containing a portion of the STAT3 3’UTR downstream of a
CMV promoter-driven luciferase gene (Fig. 7A). When the miRNAs of interest are
present, they bind to the cloned 3’UTR and silence luciferase expression. Our positive
control vector, pMIR-miR20, contains two synthesized binding sites that are
complementary to the mature miR-20 sequences. The negative control vector, pMIR-
Hoxa11, contains the 3’UTR of the Hoxa11 gene, which is not predicted to bind miR-17
family miRNAs and is specifically silenced by miR-181 during mammalian myoblast
differentiation (Naguibneva et al., 2006). The experimental vectors used in our studies
contain either wild type (pMIR-S3) or mutant versions of one (pMIR-S3st1mt and pMIR-
S3st2mt) or both (pMIR-S32xmt) of the two putative miRNA binding sites within the
We assayed the luciferase activity of our reporter constructs in response to
endogenous miRNAs over the course of ES cell differentiation. As hypothesized, the
pMIR-S3 construct was silenced between days 1 and 3 at the onset of ES cell
differentiation (Fig. 7B). The pattern of luciferase activity corresponds to the period of
differentiation during which STAT3 activity is normally downregulated and miR-93 is
upregulated (Fig. 5C). Moreover, in figure 7B, pMIR-S3 is highest at d4, suggesting that
type construct shows dramatic changes in luc expression that virtually mirror when
once ES cells have initiated differentiation, miRNAs are no longer regulating the STAT3
UTR. This is supported by figure 5 which shows miRNA levels peaking between days 2
and 3 and beginning to show a decrease by d4. To ensure that this pattern of luciferase
activity was actually due to endogenous miRNAs binding the STAT3 3’UTR and not just
an artifact of our system, we performed the same timecourse experiment using STAT3
3’UTR mutant constructs (Fig. 7B and S2). The construct containing mutations in the
second miRNA binding site (pMIR-S3st2mt) displayed some responsiveness to miRNAs
at the onset of differentiation but not later in the process (Fig. S2A). In contrast, the site
one mutant (pMIR-S3st1mt) and double mutant (pMIR-S32xmt) constructs had an
entirely different pattern of luciferase activity when compared to the wild type STAT3
construct over the course of differentiation (Fig. 7B and S2A). These two mutant
constructs showed no decrease in luciferase activity at the onset of differentiation.
Together these data suggest that while both miRNA binding sites on the STAT3 3’UTR
are functional, the first site is more important for miRNA binding at the onset of
differentiation. Moreover, another important aspect of these data that must be stressed is
that there is very little change in luc expression in the mutant construct while the wild-
STAT3 protein is reduced during the initial stages of ES cell differentiation.
We also tested the effectiveness of these constructs by co-transfecting them along
with mimics or inhibitors to miR-20 or miR-93. As expected, miR-20 mimics could
successfully decrease the luciferase activity of our positive control vector (p<.02) but not
our negative control vector (Fig. S2B). The miR-20 mimic also induced a modest but
statistically significant 25% decrease in luciferase activity of the pMIR-S3 construct.
miRNAs) but not when we transfect the pMIR-S3 construct. This suggests that binding
These data suggest that miR-20 can both bind to and silence the STAT3 3’UTR. More
surprising results were seen when co-transfecting these same constructs with miR-93
mimics. Specifically, the miR-93 mimics could not silence the pMIR-miR20 positive
control vector. However, as expected, miR-93 mimics induced a drastic 50% reduction
in luciferase activity from the pMIR-S3 construct (Fig. S2C). This interesting result
suggests, as previously hypothesized, that miR-93 can more efficiently bind to and
silence STAT3 expression. Additionally, these novel data suggest that factors other than
the seed sequence, which is identical in miR-20 and miR-93, may regulate binding of
miRNAs to target UTRs.
Of note is the increase in pMIR-miR20 luciferase activity in response to the miR-
93 inhibitor (Fig. S2C). While the miRNA mimics work in a physiological manner
(being loaded into the miRNA RISC silencing complex) the inhibitors simply bind
complimentary miRNAs. We suggest that the miR-93 inhibitor is capable of binding to
and blocking endogenous miR-20 or other miR-17 family members, resulting in
increased luciferase expression of pMIR-miR20. We see this effect only when using the
positive control vector (which was designed to bind multiple endogenous miR-17
of endogenous miR-17 miRNAs to the STAT3 3’UTR is more restricted than binding to
the control sequence.
To test that the decreases in luciferase expression correlated to a decrease in
STAT3 protein, mimics and inhibitors of miR-20 and miR-93 were transfected into
undifferentiated ES cells. After 48 hours, lysates were made and analyzed by Western
blot for STAT3 expression. We constructed a dose response graph and demonstrated a
when compared to reagent only or negative mimic controls. As STAT3 has no
significant decrease in STAT3 when ES cells were transfected with 80-160nM of miR-93
mimic (Fig. 7C). Based on these results we chose to use 20nM and 100nM
concentrations of miRNA mimics and inhibitors in a larger scale experiment. We
demonstrated that both miR-20 and miR-93 mimics, transfected at the 100nM
concentration, could significantly reduce STAT3 expression (Fig. S3A). Additionally,
this decrease could be rescued by co-transfection of inhibitors along with mimics,
suggesting that the effects on STAT3 were specifically caused by the activity of the
miRNA mimics. The co-transfection of mimics to both miR-20a and miR-93 (each at
50M) also resulted in a significant decrease in STAT3 expression. However, in this case
the decrease was less than that seen when the miR-93 mimic was transfected alone.
These data suggest that miR-93 may be more efficient at silencing STAT3 and that
perhaps miR-93 and miR-20 do not bind the STAT3 3’UTR with the same affinity.
To confirm that the decrease in STAT3 expression was due to downregulation by
miRNAs and not a secondary effect of ES cell differentiation, we transfected negative
control or miR-93 mimics (100nM) into MDA-MB-231 breast cancer cells (Fig. S3B). In
accordance with our expectations, the miR-93 mimic led to a decrease in STAT3 protein
pluripotency function in these cells, any decrease in STAT3 protein is likely due to
miRNA downregulation and not indirect effects on cellular differentiation.
Finally, to determine which endogenous miRNAs were binding the STAT3
3’UTR during ES cell differentiation, we repeated our timecourse experiments, this time
co-transfecting the inhibitors to either miR-20 or miR-93 along with the pMIR-S3
construct (Fig. 7D). While transfection of the miR-20 inhibitor had a slight effect on the
confounding factors decreased the usefulness of a knockout mouse model in this study.
pattern of pMIR-S3 luciferase expression, a decrease in luciferase activity was still
observed at the onset of differentiation. Only transfection of miR-93 inhibitors
completely prevented the decrease of luciferase activity at the onset of differentiation,
closely resembling the pattern of luciferase activity seen with the pMIR-S32xmt
construct (Fig. 7B). These data strongly suggest that endogenous miR-93 binds to and
silences the STAT3 during ES cell differentiation.
While the current literature has demonstrated that ES cells express a unique set of
miRNAs, few studies clearly define a functional consequence of miRNA expression in
this model system. In this manuscript we have not only identified a family of miRNAs
that are present in ES cells, but we have confirmed their existence in developing
mammalian embryos and linked their expression to a functional role in the regulation of
ES cell differentiation.
Although the knockout mouse has become the gold standard for identifying the
effects of specific genes on embryonic development and cellular differentiation, several
First, as the miR-17 family of miRNAs has been duplicated throughout evolution, many
of these miRNAs exhibit high sequence homology and most likely overlapping functions.
Due to this redundancy, the knockout of any single miR-17 family miRNA most likely
would not provide clear insight into its role in development. A second option would be a
knockout of the entire miR-17 family. Because this family exists as three clusters on
three separate chromosomes, this endeavor would be technically challenging and may
that investigation into the varying mechanism that may control splicing of polycistronic
disguise any divergent miRNA functions. Thus, to elucidate both unique and redundant
miRNA functions, knockouts of each miR-17 family member would have to be generated
and screened for mutants that phenocopy one another. However, the creation of 14
individual knockouts would be an overly ambitious project. Taking all of these issues
into consideration, we believe that the ES model system is currently the best system for
examining the functions of individual miRNAs within miRNA families during embryonic
As the miR-17 family of miRNAs is expressed as polycistronic clusters, we were
not surprised to find several members of this family expressed in both ES cells and
embryos. However, within this family, individual miRNAs were clearly expressed at
different levels in different cell types both in vivo and in vitro (Fig. 2, 3, 4, 5). In
addition, these miRNAs seemed to function differently in different germ layers (Fig. 6).
These data suggest that miRNA expression is tightly regulated and simple transcription of
the miRNA genes does not necessarily correlate to or predict expression and function of
the mature form. This idea is supported by current investigations on post-transcriptional
miRNA processing by Drosha in cancer cells (Thomson et al., 2006). Thus, we suggest
transcripts and processing of precursor forms is necessary to completely understand the
actions of miRNAs.
In addition to highlighting the tight regulation of miRNA expression during
development, our data also allude to the fact that “seed sequence” recognition is not
sufficient to predict miRNAs binding. Although several algorithms predicted binding of
miR-17 family members to STAT3, differences in binding affinity were clear.
probable that downregulation of STAT3 is necessary at the onset of both mouse and
Specifically, in light of the identical core and other sequence similarities between miR-93
and miR-20, it was surprising that miR-93 could not silence the pMIR-miR20 positive
control vector. And yet, miR-93 was able to silence the STAT3 3’UTR more efficiently
than miR-20. Based on these interesting data, it is apparent that better algorithms and a
more detailed understanding of binding mechanisms are needed to predict functional
Our data demonstrate the functionality of miR-93 and STAT3 binding as a
mechanism for inducing murine ES cell differentiation. However, as the role of STAT3
in human and mouse ES cells varies, the question of the importance and conservation of
this interaction in human development begs to be answered. Although the human ES cell
experiments have not and cannot be conducted in our laboratory due to ethical policies at
Georgetown University, we predict that the role of miRNA regulation of STAT3 in ES
cell differentiation is conserved between species. First, it is important to note that we are
not referring to promotion of self-renewal but to inhibition of differentiation. Although
STAT3 is not necessary for self-renewal in mouse or human ES cells, it is still expressed
at high levels (Darr and Benvenisty, 2006; Kristensen et al., 2005). Thus, it is completely
human ES cell differentiation. This idea is supported by the fact that STAT3 regulates
differentiation of several different cell types, including hematopoietic stem cells and
neural stem cells (Chung et al., 2006; Foshay and Gallicano, 2008; Foshay et al., 2005;
Hevehan et al., 2002; Hirabayashi and Gotoh, 2005; Krichevsky et al., 2006; Smithgall et
al., 2000). In addition, at least one member of the miR-17 family of miRNAs, miR-17-
5p, is expressed in human ES cells (Suh et al., 2004) and several miR-17 family miRNAs
binding, it is likely that other stem cell or differentiation associated mRNAs are also
are predicted to bind the human STAT3 3’UTR. Thus, as all of the players necessary for
the STAT3-miRNA interaction are present, we believe that within human ES cells and
developing human embryos, miRNAs bind to and downregulate STAT3 at the onset of
differentiation. Further support of this hypothesis comes from Ventura et al., (2008) who
demonstrated that mice deficient for miR-17~92 die shortly after birth with heart and
lung defects. More importantly, double and triple knockouts (DKO or TKO) of paralogs
die prior to E15, suggesting that mature miRNAs from these paralogs can compensate for
one another. Since the major defects in the DKO and TKO embryos as shown by
Ventura et al (2008) were cardiac defects and our previous papers as well as others within
the literature have established a role for STAT3 in heart development, we believe the
phenotype observed in the Ventura paper supports the idea that this family of miRNAs
maintains appropriate levels of STAT3 during differentiation.
Although our novel data clearly demonstrate that modulation of miR-93 and miR-
20 expression can alter fate commitment during ES cell differentiation, several questions
remain. One such question is whether STAT3 is the only target mRNA responsible for
mediating the effects of miR-17 family miRNAs. As miRNAs exhibit promiscuous
functionally downregulated by these same miRNAs. Using a bioinformatics based
approach, we generated a table of miR-17 family target mRNAs that could also be
responsible for regulating the onset of ES cell differentiation (Table 2). Interestingly, this
family of miRNAs seems to target players in all of the major stem cell self-renewal
pathways, including c-Myc, Wnt, BMP, and STAT3 signaling.
equally as important and interesting. Our in vivo mouse embryo studies suggest that
The fact that c-Myc is a predicted target of this family of miRNAs may be highly
important as c-Myc has recently been shown to be a factor controlling pluripotency and
early differentiation. C-Myc was one of four genes capable of generating induced
pluripotent stem cells (iPS cells) from various somatic cell types (Wernig et al., 2007;
Takahashi et al., 2006). We would expect that members of the miR-17 family upon
increase in expression would quickly down regulate c-Myc and Wnt mRNAs resulting in
inhibition of self-renewal (Kristensen et al., 2005). C-myc is also involved in many
cancers (Soucek et al., 2008) suggesting that mis-regulation of miR-17 family of
miRNAs may be involved in the etiology of c-Myc induced cancers. It is enticing to
speculate that these small miRNA molecules could be used to down-regulate c-Myc in
these cancers. In any event, this level of understanding miRNA function within each
pathway will most likely be necessary for determining the control mechanism(s) that
differentiate ES cells and/or iPS cells into desired cell types (i.e., cardiomyocytes,
neurons, b-islet cells, etc.).
While studies of miRNA function in ES cell differentiation may yield interesting
new ideas for stem cell therapeutics, the role of miRNAs in embryonic development is
miR-93 could play a role in STAT3 regulation during gastrulation. Recent studies in the
zebrafish have revealed that STAT3 can regulate cell migration, polarity, and anterior-
posterior axis formation during gastrulation of the zebrafish embryo (Miyagi et al., 2004;
Sepich et al., 2005; Yamashita et al., 2002). Based on our in situ hybridization
experiments and this current literature, we plan to further assess the function of miR-93
during formation of the primitive streak, gastrulation, and axis formation.
We would like to acknowledge Angel Miera and Caitlin MacCarthy for their
technical expertise and advice. We thank Bill Foshay for writing a miRNA binding
prediction program for our use. In addition, we thank Carlos Benitez for his careful
harvesting of blastocysts and Tammy Gallicano for her critical reading of the manuscript.
This work was supported by NIH grant # HL70204-01, awarded to G.I.G., and by the
Transgenic Shared Resource and Flow Cytometry and Cell Sorting Shared Resources,
both supported by the NIH Cancer Center Support Grant CA51008-13.
EB – Embryoid Body
ES(C) – Embryonic Stem (Cell)
ICM – Inner Cell Mass
miRNA – microRNA
RISC – RNA-induced Silencing Complex
STAT3 – Signal Transducer and Activator of Transcription 3
UTR – Untranslated Region
Antczak M, and Van Blerkom J. (1997) Oocyte influences on early development: the
regulatory proteins leptin and STAT3 are polarized in mouse and human oocytes and
differentially distributed within the cells of the preimplantation stage embryo. Mol Hum
Barre, B., Vigneron, A. and Coqueret, O. (2005). The STAT3 transcription factor is a
target for the Myc and riboblastoma proteins on the Cdc25A promoter. J Biol Chem 280,
Boyer, L. A., Lee, T. I., Cole, M. F., Johnstone, S. E., Levine, S. S., Zucker, J. P.,
Guenther, M. G., Kumar, R. M., Murray, H. L., Jenner, R. G. et al. (2005). Core
transcriptional regulatory circuitry in human embryonic stem cells. Cell 122, 947-56.
Stem Cells 23, 530-43.
Garzon, R., Pichiorri, F., Palumbo, T., Iuliano, R., Cimmino, A., Aqeilan, R.,
Volinia, S., Bhatt, D., Alder, H., Marcucci, G. et al. (2006). MicroRNA fingerprints
during human megakaryocytopoiesis. Proc Natl Acad Sci U S A 103, 5078-83.
Giraldez, A. J., Cinalli, R. M., Glasner, M. E., Enright, A. J., Thomson, J. M.,
Baskerville, S., Hammond, S. M., Bartel, D. P. and Schier, A. F. (2005). MicroRNAs
regulate brain morphogenesis in zebrafish. Science 308, 833-38.
Gossrau, G., Thiele, J., Konang, R., Schmandt, T. and Brustle, O. (2007). Bone
morphogenetic protein-mediated modulation of lineage diversification during neural
differentiation of embryonic stem cells. Stem Cells 25, 939-49.
Gregory, R. I., Chendrimada, T. P., Cooch, N. and Shiekhattar, R. (2005). Human
RISC couples microRNA biogenesis and posttranscriptional gene silencing. Cell 123,
Cartwright, P., McLean, C., Sheppard, A., Rivett, D., Jones, K. and Dalton, S.
(2005). LIF/STAT3 controls ES cell self-renewal and pluripotency by a Myc-dependent
mechanism. Development 132, 885-96.
Catlett-Falcone, R., Landowski, T. H., Oshiro, M. M., Turkson, J., Levitzki, A.,
Savino, R., Ciliberto, G., Moscinski, L., Fernandez-Luna, J. L., Nunez, G. et al.
(1999). Constitutive activation of Stat3 signaling confers resistance to apoptosis in human
U266 myeloma cells. Immunity 10, 105-15.
Chan, R. J., Johnson, S. A., Li, Y., Yoder, M. C. and Feng, G. S. (2003). A definitive
role of Shp-2 tyrosine phosphatase in mediating embryonic stem cell differentiation and
hematopoiesis. Blood 102, 2074-80.
Chu, C. Y. and Rana, T. M. (2006). Translation repression in human cells by
microRNA-induced gene silencing requires RCK/p54. PLoS Biol 4, e210.
Chung, Y. J., Park, B. B., Kang, Y. J., Kim, T. M., Eaves, C. J. and Oh, I. H. (2006).
Unique effects of STAT3 on the early phase of hematopoietic stem cell regeneration.
Blood 108, 1208-15.
Corvinus, F. M., Orth, C., Moriggl, R., Tsareva, S. A., Wagner, S., Pfitzner, E. B.,
Baus, D., Kaufmann, R., Huber, L. A., Zatloukal, K. et al. (2005). Persistent STAT3
activation in colon cancer is associated with enhanced cell proliferation and tumor
growth. Neoplasia 7, 545-55.
Darr, H. and Benvenisty, N. (2006). Factors involved in self-renewal and pluripotency
of embryonic stem cells. Handb Exp Pharmacol, 1-19.
Dews, M., Homayouni, A., Yu, D., Murphy, D., Sevignani, C., Wentzel, E., Furth, E.
E., Lee, W. M., Enders, G. H., Mendell, J. T. et al. (2006). Augmentation of tumor
angiogenesis by a Myc-activated microRNA cluster. Nat Genet 38, 1060-65.
Duncan SA, Zhong Z, Wen Z, Darnell JE Jr. STAT signaling is active during early
mammalian development. Dev Dyn. (1997) 208(2):190-8.
Feng, G. S. (2007). Shp2-mediated molecular signaling in control of embryonic stem
cell self-renewal and differentiation. Cell Res 17, 37-41.
Foshay, K., Gallicano, G.I. (2008). Regulation of Sox2 by STAT3 Initiates
Commitment to the Neural Stem Cell Fate. Stem Cells and Development 17(2):269-278.
Foshay, K., Rodriguez, G., Hoel, B., Narayan, J. and Gallicano, G. I. (2005).
JAK2/STAT3 directs cardiomyogenesis within murine embryonic stem cells in vitro.
Krichevsky, A. M., Sonntag, K. C., Isacson, O. and Kosik, K. S. (2006). Specific
microRNAs modulate embryonic stem cell-derived neurogenesis. Stem Cells 24, 857-64.
Kristensen, D.D., Kalisz, M., and Nielsen, J.H., (2005). Cytokine signaling in
embryonic stem cells. APMIS 113:756-772.
Lewis, B. P., Burge, C. B. and Bartel, D. P. (2005). Conserved seed pairing, often
flanked by adenosines, indicates that thousands of human genes are microRNA targets.
Cell 120, 15-20.
Lewis, B. P., Shih, I. H., Jones-Rhoades, M. W., Bartel, D. P. and Burge, C. B.
(2003). Prediction of mammalian microRNA targets. Cell 115, 787-98.
Li, Y., McClintick, J., Zhong, L., Edenberg, H. J., Yoder, M. C. and Chan, R. J.
(2005). Murine embryonic stem cell differentiation is promoted by SOCS-3 and inhibited
by the zinc finger transcription factor Klf4. Blood 105, 635-37.
Hatfield, S. D., Shcherbata, H. R., Fischer, K. A., Nakahara, K., Carthew, R. W. and
Ruohola-Baker, H. (2005). Stem cell division is regulated by the microRNA pathway.
Nature 435, 974-78.
He, L., Thomson, J. M., Hemann, M. T., Hernando-Monge, E., Mu, D., Goodson, S.,
Powers, S., Cordon-Cardo, C., Lowe, S. W., Hannon, G. J. et al. (2005). A microRNA
polycistron as a potential human oncogene. Nature 435, 828-33.
Gallicano GI, Kouklis P, Bauer C, Yin M, Vasioukhin V, Degenstein L, Fuchs E.
(1998) Desmoplakin is required early in development for assembly of desmosomes and
cytoskeletal linkage. J Cell Biol. 143(7):2009-22.
Hevehan, D. L., Miller, W. M. and Papoutsakis, E. T. (2002). Differential expression
and phosphorylation of distinct STAT3 proteins during granulocytic differentiation.
Blood 99, 1627-37.
Hirabayashi, Y. and Gotoh, Y. (2005). Stage-dependent fate determination of neural
precursor cells in mouse forebrain. Neurosci Res 51, 331-36.
Hirano, T., Ishihara, K. and Hibi, M. (2000). Roles of STAT3 in mediating the cell
growth, differentiation and survival signals relayed through the IL-6 family of cytokine
receptors. Oncogene 19, 2548-56.
Horvath, C. M. (2000). STAT proteins and transcriptional responses to extracellular
signals. Trends Biochem Sci 25, 496-502.
Houbaviy, H. B., Murray, M. F. and Sharp, P. A. (2003). Embryonic stem cell-specific
MicroRNAs. Dev Cell 5, 351-58.
Inagaki-Ohara, K., Hanada, T. and Yoshimura, A. (2003). Negative regulation of
cytokine signaling and inflammatory diseases. Curr Opin Pharmacol 3, 435-42.
Kanellopoulou, C., Muljo, S. A., Kung, A. L., Ganesan, S., Drapkin, R., Jenuwein,
T., Livingston, D. M. and Rajewsky, K. (2005). Dicer-deficient mouse embryonic stem
cells are defective in differentiation and centromeric silencing. Genes Dev 19, 489-501.
Keller, G., Kennedy, M., Papayannopoulou, T. and Wiles, M. V. (1993).
Hematopoietic commitment during embryonic stem cell differentiation in culture. Mol
Cell Biol 13, 473-86.
Kloosterman, W. P., Wienholds, E., de Bruijn, E., Kauppinen, S. and Plasterk, R. H.
(2006). In situ detection of miRNAs in animal embryos using LNA-modified
oligonucleotide probes. Nat Methods 3, 27-29.
Rana, T. M. (2007). Illuminating the silence: understanding the structure and function of
small RNAs. Nat Rev Mol Cell Biol 8, 23-36.
Raz, R., Lee, C. K., Cannizzaro, L. A., d'Eustachio, P. and Levy, D. E. (1999).
Essential role of STAT3 for embryonic stem cell pluripotency. Proc Natl Acad Sci U S A
Robertson, E., Bradley, A., Kuehn, M. and Evans, M. (1986). Germ-line transmission
of genes introduced into cultured pluripotential cells by retroviral vector. Nature 323,
Sepich, D. S., Calmelet, C., Kiskowski, M. and Solnica-Krezel, L. (2005). Initiation of
convergence and extension movements of lateral mesoderm during zebrafish gastrulation.
Dev Dyn 234, 279-92.
Shcherbata, H. R., Hatfield, S., Ward, E. J., Reynolds, S., Fischer, K. A. and
Ruohola-Baker, H. (2006). The MicroRNA pathway plays a regulatory role in stem cell
division. Cell Cycle 5, 172-75.
Livak, K. J. and Schmittgen, T. D. (2001). Analysis of relative gene expression data
using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402-
Meng, F., Henson, R., Wehbe-Janek, H., Smith, H., Ueno, Y. and Patel, T. (2007).
The MicroRNA let-7a modulates interleukin-6-dependent STAT-3 survival signaling in
malignant human cholangiocytes. J Biol Chem 282, 8256-64.
Miyagi, C., Yamashita, S., Ohba, Y., Yoshizaki, H., Matsuda, M. and Hirano, T.
(2004). STAT3 noncell-autonomously controls planar cell polarity during zebrafish
convergence and extension. J Cell Biol 166, 975-81.
Murchison, E. P., Partridge, J. F., Tam, O. H., Cheloufi, S. and Hannon, G. J.
(2005). Characterization of Dicer-deficient murine embryonic stem cells. Proc Natl Acad
Sci U S A 102, 12135-40.
Naguibneva, I., Ameyar-Zazoua, M., Polesskaya, A., Ait-Si-Ali, S., Groisman, R.,
Souidi, M., Cuvellier, S. and Harel-Bellan, A. (2006). The microRNA miR-181 targets
the homeobox protein Hox-A11 during mammalian myoblast differentiation. Nat Cell
Biol 8, 278-84.
Nefedova, Y., Huang, M., Kusmartsev, S., Bhattacharya, R., Cheng, P., Salup, R.,
Jove, R. and Gabrilovich, D. (2004). Hyperactivation of STAT3 is involved in
abnormal differentiation of dendritic cells in cancer. J Immunol 172, 464-74.
Nelson, P. T., Baldwin, D. A., Kloosterman, W. P., Kauppinen, S., Plasterk, R. H.
and Mourelatos, Z. (2006). RAKE and LNA-ISH reveal microRNA expression and
localization in archival human brain. Rna 12, 187-91.
Niwa, H., Burdon, T., Chambers, I. and Smith, A. (1998). Self-renewal of pluripotent
embryonic stem cells is mediated via activation of STAT3. Genes Dev 12, 2048-60.
O'Donnell, K. A., Wentzel, E. A., Zeller, K. I., Dang, C. V. and Mendell, J. T. (2005).
c-Myc-regulated microRNAs modulate E2F1 expression. Nature 435, 839-43.
Peerani, R., Rao, B. M., Bauwens, C., Yin, T., Wood, G. A., Nagy, A., Kumacheva,
E. and Zandstra, P. W. (2007). Niche-mediated control of human embryonic stem cell
self-renewal and differentiation. Embo J 26, 4744-55.
Pera, M. F., Andrade, J., Houssami, S., Reubinoff, B., Trounson, A., Stanley, E. G.,
Ward-van Oostwaard, D. and Mummery, C. (2004). Regulation of human embryonic
stem cell differentiation by BMP-2 and its antagonist noggin. J Cell Sci 117, 1269-80.
Nat Genet 39, 380-5.
Yamashita, S., Miyagi, C., Carmany-Rampey, A., Shimizu, T., Fujii, R., Schier, A.
F. and Hirano, T. (2002). Stat3 Controls Cell Movements during Zebrafish Gastrulation.
Dev Cell 2, 363-75.
Ying, Q. L., Nichols, J., Chambers, I. and Smith, A. (2003). BMP induction of Id
proteins suppresses differentiation and sustains embryonic stem cell self-renewal in
collaboration with STAT3. Cell 115, 281-92.
Zeng, Y., Yi, R. and Cullen, B. R. (2003). MicroRNAs and small interfering RNAs can
inhibit mRNA expression by similar mechanisms. Proc Natl Acad Sci U S A 100, 9779-
Zhao, Y., Ransom, J. F., Li, A., Vedantham, V., von Drehle, M., Muth, A. N.,
Tsuchihashi, T., McManus, M. T., Schwartz, R. J. and Srivastava, D. (2007).
Dysregulation of cardiogenesis, cardiac conduction, and cell cycle in mice lacking
miRNA-1-2. Cell 129, 303-17.
Smithgall, T. E., Briggs, S. D., Schreiner, S., Lerner, E. C., Cheng, H. and Wilson,
M. B. (2000). Control of myeloid differentiation and survival by Stats. Oncogene 19,
Soucek, L., Whitfield, J., Martins, C.P., Finch, A.J., Murphy, D.J., Sodir, N.M.,
Karnezis, A.N., Swigart, L.B., Nasi, S., & Evan, G.I. (2008). Modelling Myc inhibition
as a cancer therapy. Nature 455: 679-683.
Suh, M. R., Lee, Y., Kim, J. Y., Kim, S. K., Moon, S. H., Lee, J. Y., Cha, K. Y.,
Chung, H. M., Yoon, H. S., Moon, S. Y. et al. (2004). Human embryonic stem cells
express a unique set of microRNAs. Dev Biol 270, 488-98.
Suzuki, A., Raya, A., Kawakami, Y., Morita, M., Matsui, T., Nakashima, K., Gage,
F. H., Rodriguez-Esteban, C. and Izpisua Belmonte, J. C. (2006). Nanog binds to
Smad1 and blocks bone morphogenetic protein-induced differentiation of embryonic
stem cells. Proc Natl Acad Sci U S A 103, 10294-99.
Takeda, K., Noguchi, K., Shi, W., Tanaka, T., Matsumoto, M., Yoshida, N.,
Kishimoto, T. and Akira, S. (1997). Targeted disruption of the mouse Stat3 gene leads
to early embryonic lethality. Proc Natl Acad Sci U S A 94, 3801-04.
Tang, F., Hajkova, P., Barton, S. C., Lao, K. and Surani, M. A. (2006). MicroRNA
expression profiling of single whole embryonic stem cells. Nucleic Acids Res 34, e9.
Tanzer, A. and Stadler, P. F. (2004). Molecular evolution of a microRNA cluster. J Mol
Biol 339, 327-35.
Thomson, J. M., Newman, M., Parker, J. S., Morin-Kensicki, E. M., Wright, T. and
Hammond, S. M. (2006). Extensive post-transcriptional regulation of microRNAs and
its implications for cancer. Genes Dev 20, 2202-07.
Ventura, A., Young, A. G., Winslow, M. M., Lintault, L., Meissner, A., Erkeland, S.
J., Newman, J., Bronson, R. T., Crowley, D., Stone, J. R. et al. (2008). Targeted
deletion reveals essential and overlapping functions of the miR-17 through 92 family of
miRNA clusters. Cell 132, 875-86.
Walsh, J. and Andrews, P. W. (2003). Expression of Wnt and Notch pathway genes in a
pluripotent human embryonal carcinoma cell line and embryonic stem cell. Apmis 111,
197-210; discussion 210-11.
Wang, Y., Medvid, R., Melton, C., Jaenisch, R. and Blelloch, R. (2007). DGCR8 is
essential for microRNA biogenesis and silencing of embryonic stem cell self-renewal.
Zhao, Y., Samal, E. and Srivastava, D. (2005). Serum response factor regulates a
muscle-specific microRNA that targets Hand2 during cardiogenesis. Nature 436, 214-20.
Figure 1. A. The mature miRNA sequences for miR-17-5p, miR-20a, miR-93, and miR-
106a. These 4 miRNAs are predicted to bind STAT3. B. The 2 predicted miR-17 family
binding sites on the STAT3 3’UTR. Binding sites and miRNA nucleotides are aligned to
shown conservation (black - both binding sites; blue – site 1; green site 2). All 4 miRNAs
share the same seed region, which is perfectly complementary to both STAT3 binding
Figure 2. A. In situ hybridization of mature miRNAs in blastocysts stage embryos.
miR-17-5p and miR-20a showed generally even expression through the cells of the
blastocyst. A mild increase in hybridization of both miRNAs was detected in the
trophectoderm (yellow arrow). miR-93 showed the most distinctive expression pattern
with very little signal in the cells of the ICM and a drastic upregulation in both the
putative primitive endoderm (white arrowhead) and trophectoderm (yellow arrow). miR-
106a showed very specific staining in the ICM (blue arrowhead), although levels
appeared relatively equal in all cell types. Cells with the highest miRNA expression were
those undergoing differentiation (i.e. putative primitive endoderm and trophectoderm). A
probe to cel-miR-159 (a C. elegans miRNA not found in mouse) is used as a negative
control. The embryos shown are representative of 6-10 embryos for each miRNA tested.
B. Immunofluorescence of blastocyst stage embryos showed that STAT3 is localized to
the ICM, an expression pattern opposite of miR-93. Arrowheads point to putative
primitive endoderm that has little or no STAT3 staining. Anti CK8/18 staining clearly
labeled the trophectoderm, which shows low levels of STAT3 staining. The embryo
shown is a representative of 10 embryos from three experiments.. C. A schematic
drawing of a blastocyst stage embryo to demonstrate the orientation of the blastocysts
shown in A and B.
Figure 3. A-D. Whole mount in situ hybridization of E5.5 mouse embryos. At this
stage in development, expression of the miR-17-5p, miR-20a, miR-93 and miR-106a is
high in the differentiating extraembryonic cells that will form the tissues that support
growth of the epiblast. The embryonic ectoderm, or epiblast, is relatively
undifferentiated at this stage of development. These cells exhibit little or no miR-17
family miRNA expression. E. Immunohistochemistry for STAT3 expression in an E5.5
embryo section. STAT3 (brown) is localized to the nuclei within the epiblast or primitive
ectoderm. Little if any staining is observed in the visceral endoderm or extraembryonic
ectoderm. A-E. cone – ectoplacental cone; eve – extraembryonic visceral endoderm; ve
– visceral endoderm; p-ect – primitive ectoderm; p-exe – primitive extraembryonic
ectoderm; pa – proamniotic cavity.
Figure 4. A-E. In gastrulation stage embryos (E6.5-E6.75) miR-17-5p (A), miR-20a
(C), miR-93 (B) and miR-106a (D) are all localized to ectoderm cells one cell layer away
from primitive streak ectoderm (white arrows). In addition, miR-17-5p and miR-93
showed specific upregulation in differentiating mesoderm cells (A, small arrowheads). In
most cases miRNA expression was low in the visceral endoderm (A-D, large
arrowheads). A few bright cells within the visceral endoderm are thought to be
intercalating mesendoderm cells. Note: miR-93 is specifically found in differentiating
cells of both blastocyst and gastrulation stage embryos (B and Fig. 2). F.
Immunohistochemistry of an E6.75, gastrulating embryo demonstrated elevated levels of
STAT3 protein in ect (black arrowheads) not associated with the primitive streak. eve –
extraembryonic visceral endoderm; ve – visceral endoderm; me – mesoderm; ect –
ectoderm; amn – amniotic cavity; dec – deciduas; cone – ectoplacental cone.
Figure 5. Change in miRNA expression after differentiation. A. Mature miRNA
expression in E4.0 blastocyst cell types was examined by qPCR. miR-93 expression was
highest throughout the blastocyst, but showed a 10-fold increase when cells differentiated
from the ICM to the primitive endoderm or trophectoderm. In general these data
correlate to expression profiles determined by in situ hybridization. B. miRNA
expression differences between ES cells and ICM. In ES cells miR-20 was expressed at
higher levels than the other miR-17 family miRNAs. C. Expression analysis of the miR-
17 family members over the course of ES cell differentiation revealed distinct patterns
directly related to differentiation. In most cases, (e.g., miR-93, miR-106a) expression
virtually doubled within the first 3 days of ES cell differentiation. A-C. Each RT
reaction was performed three times, and each qPCR sample was run in triplicate.
Relative gene expression was calculated using the 2-∆∆CT method using U6 snRNA for
normalization (Livak and Schmittgen, 2001). Data are presented as the mean +/- the
Figure 6. A-F. Transfection of miR-20 or miR-93 inhibitors or mimics affects fate
commitment in differentiating ES cells. Cells were transfected in the undifferentiated
state and put into suspension culture without LIF 24 hrs. post-transfection (d0). RNA
was collected for qRT-PCR on days 1, 3, and 7. Transfection of either inhibitor caused a
delay in ectodermal differentiation, as evidenced by significantly lower levels of Fgf-5
expression. The miR-93 inhibitor also blocked endodermal differentiation, resulting in
decreased Hnf4a expression both at the onset of differentiation and after 7 days under
differentiating conditions. Neither inhibitor seemed to affect mesodermal differentiation.
However, transfection of the mimic to miR-93 did cause a significant upregulation in
brachyury, a mesoderm marker. G-I. Using the same transfection procedure as above,
EBs were cultured for 10 days and then embedded in paraffin for H&E staining. In
controls or EBs transfected with miR-93 mimics (G, I) a clear outer layer of visceral
endoderm, as distinguished by the large vacuoles, is apparent. In EBs transfected with
miR-93 inhibitors, these cells are absent.
ES cells results in a rapid decrease in luciferase activity once differentiation begins. The
Figure 7. A. A schematic representation of the pMIR-luciferase reporter constructs used
in this study. B. Transfection of pMIR-S3 (vector containing the STAT3 3’UTR) into
decrease in luciferase lasts approximately 2 days and correlates to a period in which
STAT3 activity is known to be reduced (Foshay and Gallicano 2008; Foshay et al.,
2005). Mutation of the miR-17 family binding sites on the STAT3 3’UTR completely
changes the pattern of luciferase expression, suggesting that these are functional miRNA
binding sites. All luciferase values are normalized to Renilla luciferase. C. Dose
response curve of STAT3 expression 48 hrs. after transfection with miR-93 mimic. Inset
Western blot shows representative results. Values for STAT3 expression were generated
by calculation of the pixel intensity of each band using Adobe Photoshop CS, and were
normalized to tubulin expression. Data are represented the mean +/- the SEM (n=3). A
reduction in STAT3 expression is seen when concentrations of greater than 80nM are
used. D. Luciferase activity of the pMIR-S3 construct alone or co-transfected with
inhibitors to miR-20 or miR-93 was analyzed during ES cell differentiation. Only
inhibitors to miR-93 could prevent the decrease in pMIR-S3 luciferase activity at the
onset of differentiation. Neg. Inhib. is a non-specific inhibitor used as a control. All
values are normalized to Renilla luciferase and presented as the mean +/- SEM.
95% reduction in the amount of miR-93 detected within the cells. These data confirm the
Figure S1. ES cells were transfected with 100nM miR-93 mimic or inhibitor. 48 hours
post transfection, RNA was extracted and used for qRT-PCR to determine levels of
mature miRNA post-transfection. Transfection of mimics resulted in a 4000 fold
increase in the amount of detected mature miR-93. Transfection of inhibitors caused a
efficacy of our transient transfection and demonstrate that the mimics and inhibitors
effect levels of mature miRNA as expected.
ACCEPTED MANUSCRIPT Download full-text
Figure S2. A. Mutations in individual miRNA binding sites change the pattern of
luciferase activity over the timecourse of differentiation when compared to the non-