Mechanism of translational regulation by miR-2 from sites
in the 59 untranslated region or the open reading frame
FRANCESCA MORETTI, ROLF THERMANN, and MATTHIAS W. HENTZE
European Molecular Biology Laboratory, Meyerhofstrasse 1, 69117 Heidelberg, Germany
MicroRNAs (miRs) commonly regulate translation from target mRNA 39 untranslated regions (UTRs). While effective
miR-binding sites have also been identified in 59 untranslated regions (UTRs) or open reading frames (ORFs), the mechanism(s)
of miR-mediated regulation from these sites has not been defined. Here, we systematically investigate how the position of
miR-binding sites influences translational regulation and characterize their mechanistic basis. We show that specific trans-
lational regulation is elicited in vitro and in vivo not only from the 39UTR, but equally effectively from six Drosophila miR-
2-binding sites in the 59UTR or the ORF. In all cases, miR-2 triggers mRNA deadenylation and inhibits translation initiation in
a cap-dependent fashion. In contrast, single or dual miR-2-binding sites in the 59UTR or the ORF yield rather inefficient or no
regulation. This work represents the first demonstration that 59UTR and ORF miR-binding sites can function mechanistically
similarly to the intensively investigated 39UTR sites. Using single or dual binding sites, it also reveals a biological rationale for
the high prevalence of miR regulatory sites in the 39UTR.
Keywords: microRNA; translational control; 59UTR; open reading frame; 39UTR
MicroRNAs (miRs) are key post-transcriptional regulators
of a broad range of physiological and pathological processes
in animals and plants (Bushati and Cohen 2007; Voinnet
2009). Animal miRs trigger regulation in the context of the
miR-induced silencing complex (miRISC), core components
of which are the Argonaute (AGO) and GW182 proteins
(Peters and Meister 2007; Eulalio et al. 2009). miRs have
been shown to regulate gene expression by multiple mech-
anisms, including deadenylation and degradation of target
mRNAs, as well as inhibition of translation. In this latter
case, a growing body of evidence suggests that translation is
blocked at the initiation step, although alternative mecha-
nisms have been reported as well (Filipowicz et al. 2008;
Carthew and Sontheimer 2009).
The vast majority of animal miRs regulates gene expres-
sion by binding to sites located in the 39 untranslated re-
gions (UTRs) of their mRNA targets. This striking posi-
tional bias has led to the development of miR target search
algorithms that focus on 39UTRs and, consequently, to
a further amplification of the bias for functional 39UTR
sites (Bartel 2009). However, effective miR-binding sites
have also been identified in the 59UTR or the open reading
frame (ORF) of target mRNAs (e.g., Jopling et al. 2005;
Duursma et al. 2008; Forman et al. 2008; Henke et al. 2008;
Lal et al. 2008; Ørom et al. 2008; Tay et al. 2008; Elcheva
et al. 2009; Tsai et al. 2009). Non-39UTR miR targeting has
also been studied in silico and in reporter gene assays (e.g.,
Kloostermann et al. 2004; Lytle et al. 2007; Stark et al.
2007). Recently, pervasive conserved miR targeting was
demonstrated computationally in Drosophila ORFs; it was
also suggested that the extent of physiological 59UTR and
ORF miR targeting may be greater in flies than in mammals
(Schnall-Levin et al. 2010). Despite mounting evidence
indicating the biological relevance of this phenomenon, the
mechanisms underlying 59UTR and ORF miR-mediated
regulation are still unknown. Here we address this un-
resolved issue, and show that Drosophila miR-2 represses
translation initiation in a cap-dependent manner and in-
duces mRNA deadenylation with similar efficiencies from
binding sites in the 59UTR or the ORF as from standard
39UTR positions. We also find that single or dual miR-
binding sites are more effective from the 39UTR, which
may provide an explanation for the observed physiological
bias for 39UTR miR regulatory sites.
Reprint requests to: Matthias W. Hentze, European Molecular Biology
Laboratory, Meyerhofstrasse 1, 69117 Heidelberg, Germany; e-mail: hentze@
embl.de; fax: 49-6221-387-8211.
Article published online ahead of print. Article and publication date are
RNA (2010), 16:2493–2502. Published by Cold Spring Harbor Laboratory Press. Copyright ? 2010 RNA Society.
Effective miR-2-mediated silencing from all mRNA
positions in vitro
To explore whether and how miR-mediated regulation
can be confered by 59UTR- and ORF-binding sites, we
initially used the in vitro system derived from Drosophila
embryos, which had previously been used to dissect the
mechanisms of miR-mediated translational control via six
miR-2-binding sites in the 39UTR (Thermann and Hentze
2007; Zdanowicz et al. 2009). We generated reporter con-
structs based on either the firefly luciferase (FL) ORF (for
translation assays) or a shortened derivative (sORF; for
ribosome binding assays), bearing six copies of the reaper
miR-2-binding site (wild type, wt) and 24 nuclceotides
(nt) of flanking sequence alternatively in the 39UTR, the
59UTR, or the ORF (Figs. 1A, 2A–D). The sequence of the
miR-2-binding cassette was optimized to remove all start
or in-frame stop codons. As controls, we used reporters
bearing mutated seed regions (mut), which are therefore
not bound by miR-2. The human b-globin 59UTR (GL,
z60 nt) was inserted into all reporters to widen the spac-
ing between the cap structure and the miR-2-binding sites
when these are placed at the 59 end. The resulting con-
structs are thus highly similar to each other and vary the
miR-binding site positions with minimal changes to the
Upon incubation of the capped and polyadenylated
GL-FL-reporter mRNAs in the Drosophila cell-free system,
we observe that miR-2 triggers effective repression of the
GL-FL wt reporters from the 59UTR and both ORF sites,
as well as from the 39UTR-binding sites (Fig. 1B). This
effect is confirmed by kinetic analyses (Fig. 1C) and is
specific because it is relieved by sequestration with a com-
plementary locked nucleic acid (LNA) anti-miR but not
an unrelated anti-miR (Fig. 1D). We also notice that the
luciferase activities from the mut reporters are not greatly
affected by the position of the miR-2-binding sites, in-
dicating that the effects on translational repression of the
wt reporters are likely to be comparable between the dif-
ferent constructs (Fig. 1E). The lower (approximately
sixfold) luciferase counts from the ORF(39) construct are
likely due to a negative effect of C-terminal fusions on the
catalytic activity of luciferase (Waud et al. 1996; see below).
To check whether the stability of the reporter mRNAs is
affected by miR-2, we performed RT–qPCR analyses of
RNA recovered at the end of the translation reactions (t2).
As previously shown for the 39UTR reporters (Thermann
and Hentze 2007), all GL-FL wt reporters are as stable as
their mut counterparts, indicating that miR-2 does not
alter mRNA stability and that repression occurs at the level
of translation (Fig. 1F). We conclude that Drosophila miR-2
can principally inhibit translation from the 59UTR or the
ORF as efficiently as from the 39UTR.
miR-2 inhibits translation initiation and triggers
deadenylation from all sites
miR-2 binding to 39UTR sites induces deadenylation and
repression of translation initiation (Thermann and Hentze
2007; Zdanowicz et al. 2009). We next wanted to explore
whether this is also the case for the non-39UTR-binding
sites, or whether miR-2 acts in a different way from either
of the alternative sites. We performed sucrose gradient
analyses of translation initiation complex formation in
reactions with cycloheximide to monitor 80S ribosomal
complex formation. As shown for miR-2 regulation via
39UTR-binding sites (Thermann and Hentze 2007), we de-
tect the formation of pseudopolysomes and a reduction of
80S complex formation on all wt reporters, indicative of a
miR-2-mediated block of translation initiation (Fig. 2A–D).
We next analyzed the adenylation status of the reporter
mRNAs using the described RNase H cleavage protocol
(Zdanowicz et al. 2009) with the oligonucleotides depicted
in Figure 2E. These analyses reveal effective and specific
deadenylation of all wt reporter mRNAs regardless of the
location of the miR-2-binding sites (Fig. 2F,G).
Another defining feature of ‘‘classical’’ miR regulation is
that it relies on the interaction between the AGO1 (but not
AGO2) and GW182 proteins in the context of the miRISC
(Eulalio et al. 2008); this requirement is maintained in our
Drosophila embryo cell-free system (Till et al. 2007). To in-
vestigate whether repression from non-39UTR sites also
requires this interaction, we made use of a dominant-
negative derivative of Drosophila GW182 [GW182(1-592)].
This polypeptide contains the AGO1-binding domain but
lacks the so-called silencing domain, and is thought to act
by sequestering AGO1 into inactive complexes (Eulalio et al.
2008). The addition of this protein strongly and specifically
derepresses all reporter mRNAs, indicating that full repres-
sion requires the interaction between AGO1 and GW182 in
all cases (Fig. 3A).
To complete our mechanistic analyses, we wanted to
probe miR regulation of mRNAs bearing different cap
structure analogs. This approach recently uncovered a spe-
cific hallmark of miR regulation via the 39UTR, identifying
two chemical analogs of the cap structure (cap 16 and cap
21) that augment miR regulation. Moreover, these exper-
iments also suggested that translational repression from the
39UTR may occur via a mechanism related to that of the
4E-binding proteins (4E-BPs) (Zdanowicz et al. 2009). We
therefore in vitro synthesized GL-FL-reporter mRNAs bear-
ing either the physiological cap structure (m7G anti-reverse
cap analog, ARCA) or the cap analog 21-ARCA, which
contains a hexaphosphate instead of a triphosphate linker
(Zdanowicz et al. 2009). Reproducing our earlier results for
39UTR miR-2 sites (Zdanowicz et al. 2009), the incorporation
of cap 21-ARCA did not affect general translation of reporter
mRNAs, whereas it augmented miR repression from the
39UTR (Fig. 3B). This feature also characterizes repression
Moretti et al.
RNA, Vol. 16, No. 12
FIGURE 1. miR-2 represses translation from 59UTR- and ORF-binding sites in vitro. (A) Schematic representation of the reporters used in this
study. The firefly luciferase open reading frame is indicated by a white box (GL-FL); six binding sites for miR-2 are shown as black bars. (B) miR-2
represses translation from both UTRs and from ORF-binding sites in the Drosophila cell-free extracts. (C) Kinetic analysis of translational
repression. Samples are analyzed at 10-min intervals during the repression assay. One-hundred percent repression for each construct corresponds
to the values indicated in B. (D) The repression observed requires miR-2, since it is specifically relieved by anti miR-2 LNAs. (E) Luciferase activity
from the different mut constructs is indicated to show the effect of the position of the miR-2-binding sites on enzymatic activity. (F) The RT–
qPCR analyses show no significant destabilization of the reporter mRNAs at the end (t2) of the repression assay. Shown are averages and standard
deviations of five (B,D) or three (C,F) independent experiments. (E) One experiment performed in duplicate from B.
microRNA regulation from the 59 UTR and the ORF
from the other miR positions (Fig. 3B), implicating cap-
mediated translation initiation as the primary target for
repression also for 59UTR and ORF miR-binding sites.
We conclude that miR-2-mediated translational control
from 39UTR sites and non-39UTR sites displays striking
mechanistic similarities with regard to (1) controlling
translation initiation, (2) leading to mRNA deadenylation,
(3) requiring the interaction between AGO1 and GW182,
and (4) targeting cap-dependent translation, possibly via
a mechanism related to the 4E-BPs.
Analyzing miR-2-mediated silencing in vivo
To evaluate the significance of our findings in vivo, we
tested corresponding reporter constructs in Drosophila S2
cells. First, we transfected the capped and polyadenylated
GL-FL-reporter mRNAs. Confirming our findings in the
Drosophila embryo extracts (Fig. 1B,D), we observe that the
endogenous miR-2 specifically represses the expression of
the wt reporters bearing binding sites in the 39UTR, the
59UTR, or the 59 end of the ORF (Fig. 4A,B). Whereas
FIGURE 2. miR-2 controls translation initiation and triggers deadenylation from both UTR- and ORF-binding sites. (A–D) The radiolabeled
GL-sORF reporters were incubated in the Drosophila cell-free extracts in the presence of cycloheximide. At the end of the reaction, they were
resolved through 15%–45% sucrose gradients, fractionated, and analyzed by scintillation counting. On all wt reporters, inhibition of 80S complex
assembly (fractions 10–15) and the formation of pseudopolysomes (fractions 1–8) can be detected. Shown are averages and standard deviations
from three independent experiments. (E) Schematic representation of the oligonucleotides used in the deadenylation assays. The dT 39 and dT
59 fragments were obtained by annealing the reporter mRNAs to both an RnaseH and a dT oligo and subsequent digestion with RNase H.
Therefore, they serve as size markers for the identification of deadenylated reporter mRNAs. Both the wt reporters bearing the miR-2-binding sites
at the 39 end (F) and at the 59 end (G) are deadenylated at the end of the repression assay (t2), while mut RNAs maintain full poly(A) tail lengths.
No change in the adenylation status is observed after the preincubation (p.i.) step. t0 represents input samples at the beginning of the repression
assay. The data shown are representative of three independent experiments.
Moretti et al.
RNA, Vol. 16, No. 12
luciferase activities encoded by the 39UTR, 59UTR, and
ORF(59) reporters are similar (data not shown), the mea-
sured luciferase activities from the reporters bearing bind-
ing sites at the 39 end of the ORF (which corresponds to
the C terminus of the encoded protein) are not significantly
above background, and these constructs had to be excluded
from further in vivo analyses. The mRNA levels of these
ORF(39) constructs were found to be simlar to the other
reporters (data not shown), indicating that the lack of
luciferase activity is caused by misfolding or degradation
of the protein product. C-terminal fusions of firefly lucif-
erase have indeed been reported to cause misfolding and deg-
radation of such fusion proteins in vivo (Waud et al. 1996).
To overcome limitations of the RNA transfection strat-
egy (i.e., the absence of a nuclear history of the transfected
mRNA) and technical difficulties to accurately determine
mRNA levels (Barreau et al. 2006), we also performed
plasmid transfections. As seen before with the RNA trans-
fections, we observe miR-2-specific repression of the three
wt reporter constructs that can be analyzed in vivo (Fig.
4C,D). Interestingly, in plasmid transfection experiments,
the repression of the 59UTR and ORF(59) constructs is
about half of the repression via the 39UTR. Although we do
not know the reason for this difference, it is tempting to
speculate that aspects of the more physiological mRNP
biogenesis via nuclear transcription and export introduce
a bias in favor of 39UTR sites. We then performed RT–
qPCR analyses on total RNA from the S2 cells transfected
with the reporter plasmids. Confirming the results obtained
in vitro (Fig. 1F), we detected no significant changes in
reporter mRNA levels (Fig. 4E), indicating that miR-
2-mediated regulation is exerted at the translational level
also in vivo. We conclude that, both in vivo and in vitro,
miR-2 can silence translation also from 59UTR and ORF-
miR-2-mediated regulation in vitro and in vivo
from single or dual binding sites
Most cellular mRNAs that are regulated by miRs have sin-
gle or dual miR-binding sites; as a consequence, the extent
to which the expression of these mRNAs is affected by miRs
rarely exceeds a factor of two (Grimson et al. 2007; Selbach
et al. 2008; Bartel 2009). The correlation between the num-
ber of 39UTR miR-binding sites and the degree of repres-
sion is recapitulated for miR-2-mediated regulation in the
Drosophila cell-free system, and significant (twofold or less)
repression is observed from dual or single 39UTR miR-2
sites in vitro (Thermann and Hentze 2007). To assess the
function of single or dual 59UTR and ORF miR-binding
sites, we first analyzed appropriate GL-FL-reporter con-
structs in vitro.
Upon incubation in the Drosophila cell-free system, we
observe that two binding sites mediate just over twofold
repression from the 39UTR, whereas repression is less effi-
cient from the 39 end of the ORF and completely ineffective
from the 59UTR and the 59 end of the ORF (Fig. 5A).
Regulation from a single binding site in vitro is only ob-
served from the 39UTR (Fig. 5A). We next transfected the
reporter mRNAs into S2 cells. Two binding sites trigger
robust repression when located in the 39UTR, but are much
less effective in the 59UTR and at the 59 end of the ORF
(Fig. 5B). A single binding site is ineffective from the
59UTR, while it mediates weak repression from the 39UTR
and the ORF (Fig. 5B). Taken together, these findings show
FIGURE 3. Effective repression of all of the reporters requires the
interaction between AGO1 and GW182 and targets cap-dependent
translation. (A) In vitro translation reactions were incubated with
either a dominant-negative version of the Drosophila GW182 protein
[GW182(1-592)] or with protein buffer. The addition of 200 nM
dominant-negative protein significantly derepresses all wt reporters,
while no effect is observed with buffer alone. The graph shows
averages and standard deviations from three independent experi-
ments. (B) GL-FL-reporter mRNAs bearing either the m7G-ARCA cap
or the 21-ARCA cap analog (Zdanowicz et al. 2009) were synthesized
in vitro and incubated in the Drosophila embryo extract. The lucif-
erase counts of the mut constructs (dark-gray bars) and the fold
repression of each wt-mut pair (light-gray bars) bearing cap 21-ARCA
were compared with their m7G-ARCA-capped counterparts. In all cases,
cap 21-ARCA displays no effect on general translation, while augment-
ing miR-mediated repression. The graph shows averages and standard
deviations from six independent experiments. Statistical significance
was evaluated through an unpaired two-tailed t-test: (*) P < 0.05.
microRNA regulation from the 59 UTR and the ORF
that single or dual miR-binding sites in the 59UTR or the
ORF are generally less functional than in the 39UTR or even
Although the vast majority of identified endogenous tar-
gets for miRs bear their miRISC interaction site(s) in the
39UTR, physiologically relevant examples of miR regulation
from binding sites in the 59UTR or the ORF have been
described (Jopling et al. 2005; Duursma et al. 2008; Forman
et al. 2008; Henke et al. 2008; Lal et al. 2008; Ørom et al.
2008; Tay et al. 2008; Elcheva et al. 2009; Tsai et al. 2009).
Our work provides, to our knowledge, the first biochemical
mechanistic analyses of miR regulation from positions
59 of the 39UTR. We show that miRs can principally regulate
FIGURE 4. miR-2 represses translation from 59UTR- and ORF-binding sites in vivo. (A) In vitro transcribed GL-FL-reporter mRNAs are
repressed by endogenous miR-2 following transfection into Drosophila S2 cells. Luciferase activity of the ORF(39) constructs is not significantly
over background, possibly due to misfolding and degradation of luciferase C-terminal fusion proteins in S2 cells. (B) The repression is mediated
by miR-2, since it is specifically relieved by cotransfection of 100 nM anti-miR-2 LNAs. (C) miR-2 represses luciferase expression from transfected
reporter plasmids. (D) This repression is also specifically relieved upon cotransfection of 100 nM anti-miR-2 LNAs. (E) Total RNA was extracted
from S2 cells transfected with the reporter plasmids and subjected to RT–qPCR analyses. The wt reporter mRNAs are equally stable throughout
the experiment. mRNA levels could also be measured for the ORF(39) constructs, as expected. Shown are averages and standard deviations from
five independent experiments. Statistical significance was evaluated through an unpaired two-tailed t-test: (*) P < 0.05.
Moretti et al.
RNA, Vol. 16, No. 12
translation initiation, in a cap-dependent manner, from all
(tested) sites of an mRNA, and that the establishment of
repression is accompanied by miR-2-induced deadenyla-
tion (Figs. 2A–G, 3B), implying that the deadenylation
machinery is recruited to the mRNA irrespectively of the
position of the miRISC.
For miR-binding sites in the 59UTR (and the ORF), it is
conceivable that the miRISC acts by a dual mechanism: by
interfering with translation initiation, and by causing steric
hindrance to scanning/translating ribosomes that escape
a cap-dependent initiation block (Zdanowicz et al. 2009).
Several regulators binding to the 59UTR have, indeed,
already been shown to act sterically (Stripecke et al. 1994;
Gebauer and Hentze 2004). For miR regulation from the
59UTR, this possibility has been raised (Lee et al. 2009), and
it will be interesting to study its potential contribution to
repression. We note that repression from all sites, including
the 59UTR, is relieved equally strongly by the dominant-
negative GW182(1-592) polypeptide (Fig. 3A). If we as-
sume that these inhibited miRISC complexes remain bound
to the target mRNAs by virtue of the miR–mRNA inter-
action, this result suggests that at least these miRISCs do
not act primarily by simple steric hindrance.
miRs targeting the 59UTRs have been reported to trigger
both positive (Jopling et al. 2005; Ørom et al. 2008; Tsai
et al. 2009) and negative (Lytle et al. 2007; Grey et al. 2010;
this study) responses. Although these different regulatory
effects are not yet mechanistically explained, it is tempting
to speculate that they may be affected by specific structural
features of the miR–mRNA interaction (e.g., seed-mediated
recognition vs. noncanonical miR target sites) (Ørom et al.
2008), by the miRISC-mediated recruitment of auxiliary
repressor/activator complexes, or even by the nature of the
targeted RNA itself (e.g., the HCV genome) (Jopling et al.
2005). The availability of in vitro systems may, in the
future, help to shed light on this unresolved issue.
We also validated the key findings of this study by in
vivo transfection analyses. In line with previous findings
(Lytle et al. 2007), these experiments uncover significant
differences in quantitative aspects of the experimental re-
sults, depending on the transfection method used. Whereas
we observe similar repression upon reporter mRNA trans-
fection, regulation from the 59UTR and the 59 end of the
ORF is weaker than from the 39UTR when plasmid DNA is
transfected (Fig. 4, cf. A and C). Although these results do
not challenge the major mechanistic conclusion, they may
illuminate physiological aspects of miR regulation, espe-
cially regarding single or dual binding sites. Furthermore,
they may provide mechanistic clues regarding the relation-
ship between the miRISC, the mRNP, and the translation
(initiation) process. We envisage a scenario in which the
59UTR and ORF miRISCs are challenged by ribosomes that
have escaped the initiation block. The ribosomal challenge
that the miRISC has to face is proportional, on the one
hand, to the efficiency of the translation initiation process
(and therefore the percentage of ribosomes that manage to
escape the miRISC block) and, on the other hand, to
ribosomal processivity (Bartel 2009). In our case, these
parameters may differ between the Drosophila embryo in
vitro system, transfection experiments with in vitro tran-
scribed mRNAs or with mRNAs that are transcribed in
vivo. Especially in the latter case, the nuclear history of the
mRNA is expected to endow it with a different mRNP.
Previous reporter studies in mammals and zebrafish have
used a fixed number of miR-binding sites (either two or
four) to evaluate regulation from the 59UTR or the ORF.
This could possibly explain the discordant conclusions
reached regarding the effectiveness of 59UTR or ORF miR
FIGURE 5. miR-2-mediated regulation from single or dual binding
sites in vitro and in vivo. (A) The GL-FL-reporter mRNAs containing
one, two, or six miR-2-binding sites were incubated in the Drosophila
cell-free extract. A single binding site is only effective from the 39UTR.
Two binding sites yield over twofold repression from the 39UTR and
mediate z1.5-fold repression from the 39 end of the ORF, but are
ineffective from the 59UTR and the 59 end of the ORF. (B) The GL-
FL-reporter mRNAs containing one, two, or six miR-2-binding sites
were transfected into Drosophila S2 cells. A single binding site triggers
modest repression from the 39UTR and the 59 end of the ORF. Two
binding sites trigger robust repression from the 39UTR, but are only
moderately effective from the 59UTR and the 59 end of the ORF.
Shown are averages and standard deviations of five independent
microRNA regulation from the 59 UTR and the ORF
targeting (Kloostermann et al. 2004; Lytle et al. 2007; Gu
et al. 2009). Therefore, we also systematically compared
miR-mediated regulation by different numbers of miR-
2-binding sites in the 39UTR, the 59UTR, or the ORF. We
observe that single or dual miR-binding sites are less
efficient from the 59UTR and ORF than from the 39UTR
(Fig. 5A,B). Consistent with recent findings in mammalian
cells (Gu et al. 2009), one or two miRISCs are expected to
impose a weaker initiation block than six, and are therefore
more likely to be displaced by ‘‘escaping ribosomes,’’ hence
In summary, our results establish a mechanistic basis for
miR function from the 59UTR and the ORF. We demon-
strate that the translation initation block by miR-2 can also
be imposed from the 59UTR and the ORF. Together with
the finding that a higher number of binding sites is re-
quired to establish effective regulation from the 59UTR and
the ORF, our observations help to resolve the conundrum
resulting from mechanistic predictions made for a (cap-
dependent) block of translation initiation and the seem-
ingly inconsistent observation that the vast majority of
physiological miR regulatory sites has been found in
39UTRs. The data may also contribute to the definition
of rules for 59UTR and ORF miR targeting, thereby fa-
cilitating the (computational) identification of new phys-
iological miR targets.
MATERIALS AND METHODS
The GL-FL-reporter plasmids were obtained by insertion of the
human b-globin 59UTR (amplified with primers gl-fw, 59-TTTTG
GTACCACATTTGCTTCTG-39; and gl-rv, 59-TTTTGGGCCCGG
TGTCTGTTTTGAGG-39) between the KpnI and ApaI sites of the
FL-mut vector (Thermann and Hentze 2007). The wt and mut
miR-2-binding sites of Drosophila reaper mRNA were obtained by
PCR amplification using the primers 39box-wt-fw (59-TTTTACT
ATTGG-39), 39box-wt-rv (59-TTTTAAGTTTTAAGATCTAAACA
AAAGCGAGGTATTATTTGG-39), and 39box-mut-rv (59-TTTTA
ATGAGTAAAC-39). The cloning of one, two, or six binding sites
in the 39UTR was done as previously described (Thermann and
Hentze 2007). The ORF(39)constructs were obtained by replacing
a C-terminal fragment of the luciferase ORF, to eliminate the stop
codon, using the primers deltastop-fw (59-TTTTATCGATATTG
TTACAACACCCCAACATCTTCG-39) and deltastop-rv (59-TTT
TACTAGTGGATCTCAATTTGGACTTTCCG-39) and the restric-
tion sites ClaI and SpeI of the GL-FL-39UTR wt and mut re-
porters. The 59UTR and ORF(59) reporters were obtained by am-
plifying the binding sites with the primers ORF(59)box-fw (59-TTT
(59-TTTGGGCCCTCCACTAGTGGATCC-39), and 59box-rv (59-TTT
TCTCGAGGCGGAAGCTTAGATCT-39) and inserting them us-
ing the ApaI and XhoI restricion sites of a GL-FL-39UTR vector
that was previously cut with BamHI and BglII and religated so as
to eliminate the miR-2-binding sites from the 39UTR. The RL-
control plasmid was previously described (Thermann and Hentze
2007). The GL-sORF-reporter plasmids were generated by replac-
ing the firefly luciferase ORF in the GL-FL reporters with the
sORF, using the primers sORF-fw (59-TTTTCTCGAGTCATGGA
CTACAAAGACGACG-39), sORF-deltastop-rv (59-TTTTACTAG
TTAACAATTTGGACTTTCCG-39), and sORF-rv (59-TTTTACT
AGTTTATAACAATTTGGACTTTCCG-39) and the XhoI and
SpeI restriction sites. The HS-GL-FL-reporter plasmids were gen-
erated from the pCaSpeR-hs-pA vector. The GL-FL-39UTR and
-ORF(39) fragments were cloned into the pCaSpeR-hs-pA using
the KpnI and BglII restriction sites. The GL-FL-59UTR and
-ORF(59) fragments were PCR-amplified with the primers gl-fw
and fluc-rv (59-TTTTGTTAACGGATCTATTACAATTTGGACTT
TGCG-39) and cloned via the KpnI and HpaI sites of the pCaSpeR-
hs-pA vector. The HS-RL-control plasmid was previously described
(Duncan et al. 2006).
In vitro transcription
The synthesis of the Gl-FL-reporter, the GL-sORF-reporter, and
the RL-control mRNAs was performed as previously described
(Thermann and Hentze 2007).
Preparation of Drosophila melanogaster embryo
extract and translation experiments
Drosophila embryo extracts were prepared and used for translation
assays as described (Thermann and Hentze 2007). The preincu-
bation step was carried out for 1 h or 30 min at 25°C for luciferase
mRNA and protein quantification and for deadenylation analyses,
respectively, and the amounts of reporter mRNAs were adjusted
to maintain constant RNA concentrations. Where indicated,
30 nM anti-miR LNA (dme-miR-2 or hsa-let-7 miRCURY LNA
MicroRNA Knockdown Probes from Exiqon) was added to the
translation reaction. For the experiment shown in Figure 3A,
200 nM GW182(1-592) protein or an equal volume of buffer was
added to the translation reaction.
Total RNA from translation reactions or from transfected cells was
isolated with Trizol (Invitrogen) according to the manufacturer’s
protocol. Total RNA isolated from cells was treated with Turbo
DNA-free kit (Ambion) and 1 mg was reverse-transcribed with
random primers and SuperScript II Reverse Transcriptase (Invi-
trogen) following the manufacturer’s protocol. RT–qPCR was
then performed as described (Zdanowicz et al. 2009).
Cell culture and S2 cell transfection
Schneider 2 cells were maintained at 25°C in Schneider’s Medium
containing L-glutamine (Gibco) and supplemented with penicil-
lin/streptomycin (Gibco) and 10% fetal bovine serum (Gibco).
RNA transfections were performed with the TransMessenger
Transfection Reagent (Qiagen) according to the manufacturer’s
protocol. Cells were seeded at 106cells per well in 24-well plates
24 h before transfection. Cells were transfected with 30 fmol of
Moretti et al.
RNA, Vol. 16, No. 12
each of the Gl-FL-reporter mRNAs and 6 fmol of the RL-control
mRNA, plus 100 nM anti-miR LNA where indicated. Cells were
lysed after 24 h in 13 Passive Lysis Buffer (Promega), and firefly
and Renilla luciferase activity was determined with the Dual-
Luciferase Assay System (Promega). DNA transfections were per-
formed with the Effectene Transfection Reagent (Qiagen) accord-
ing to the manufacturer’s protocol. Cells were seeded at 1.5 3 106
cells per well in six-well plates 24 h before transfection. Cells were
transfected with 3 fmol of each of the HS-GL-FL-reporter plas-
mids and 0.3 fmol of the HS-RL-control plasmid, plus 100 nM
of anti-miR LNA where indicated. Cells were lysed after 72 h in
13 Passive Lysis Buffer, and firefly and Renilla luciferase activity
was determined with the Dual-Luciferase Assay System.
The deadenylation assay was carried on as previously described
(Zdanowicz et al. 2009). The primers RNaseH3 (59-CGCCCTTC
TTGGCCTTTATG-39) and RNaseH4 (59-GCGGTCAACTATGAA
GAAGTGTTCG-39), complementary to the firefly luciferase ORF,
were used as specified.
Sucrose density-gradient analyses
The analyses of 80S complex assembly were performed through
15%–45% linear sucrose density gradients in the presence of
1 mM cycloheximide as previously described (Thermann and
We thank Elisa Izaurralde (Tu ¨bingen) for the GW182(1-592)
expression plasmid, Agnieszka Zdanowicz for the GW182(1-592)
polypeptide, Edward Darzynkiewicz (Warsaw) for the cap 21-
ARCA, Jan Medenbach for the pCaSpeR-hs-pA vector, and all
members of the Hentze laboratory for helpful discussions. This
work was funded by a grant (HE 1442/12-1) from the Deutsche
Forschungsgemeinschaft to M.W.H.
Received July 22, 2010; accepted September 8, 2010.
Barreau C, Dutertre S, Paillard L, Osborne HB. 2006. Liposome-
mediated RNA transfection should be used with caution. RNA 12:
Bartel DP. 2009. MicroRNAs: Target recognition and regulatory
functions. Cell 136: 215–233.
Bushati N, Cohen SM. 2007. microRNA functions. Annu Rev Cell Dev
Biol 23: 175–205.
Carthew RW, Sontheimer EJ. 2009. Origins and mechanisms of miRs
and siRNAs. Cell 136: 642–655.
Duncan K, Grskovic M, Strein C, Beckmann K, Niggeweg R, Abaza I,
Gebauer F, Wilm M, Hentze MW. 2006. Sex-lethal imparts a sex-
specific function to UNR by recruting it to the msl-2 mRNA
39UTR: Translational repression for dosage compensation. Genes
Dev 20: 368–379.
Duursma AM, Kedde M, Schrier M, le Sage C, Agami R. 2008.
miR-148 targets human DNMT3b protein coding region. RNA 14:
Elcheva I, Goswami S, Noubissi FK, Spiegelman VS. 2009. CRD-BP
protects the coding region of bTrCP1 mRNA from miR-183-
mediated degradation. Mol Cell 35: 240–246.
Eulalio A, Huntzinger E, Izaurralde E. 2008. GW182 interaction with
Argonaute is essential for miR-mediated translational repression
and mRNA decay. Nat Struct Mol Biol 15: 346–353.
Eulalio A, Tritschler F, Izaurralde E. 2009. The GW182 protein family
in animal cells: New insights into domains required for miR-
mediated gene silencing. RNA 15: 1433–1442.
Filipowicz W, Bhattacharyya SN, Sonenberg N. 2008. Mechanisms of
post-transcriptional regulation by microRNAs: are the answers in
sight? Nat Rev Genet 9: 102–114.
Forman JJ, Legesse-Miller A, Coller HA. 2008. A search for conserved
sequences in coding regions reveals that the let-7 microRNA
targets Dicer within its coding sequence. Proc Natl Acad Sci 105:
Gebauer F, Hentze MW. 2004. Molecular mechanisms of translational
control. Nat Rev Mol Cell Biol 5: 827–835.
Grey F, Tirabassi R, Meyers H, Wu G, McWeeney S, Hook L, Nelson
JA. 2010. A viral microRNA down-regulates multiple cell cycle
genes through mRNA 59UTRs. PLoS Pathog 6: e100967. doi:
Grimson A, Farh KK, Johnston WK, Garrett-Engele P, Lim LP, Bartel
DP. 2007. MicroRNA targeting specificity in mammals: Determi-
nants beyond seed pairing. Mol Cell 27: 91–105.
Gu S, Jin L, Zhang F, Sarnow P, Kay MA. 2009. Biological basis for
restriction of microRNA targets to the 39 untranslated region in
mammalian mRNAs. Nat Struct Mol Biol 16: 144–150.
Henke JI, Goergen D, Zheng J, Song Y, Schu ¨ttler CG, Fehr C,
Ju ¨nemann C, Niepmann M. 2008. microRNA-122 stimulates
translation of hepatitis C virus RNA. EMBO J 27: 3300–3310.
Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. 2005.
Modulation of hepatitis C virus RNA abundance by a liver-specific
microRNA. Science 309: 1577–1581.
Kloostermann WP, Wienholds E, Ketting RF, Plasterk RH. 2004.
Substrate requirements for let-7 function in the developing zebra-
fish embryo. Nucleic Acids Res 32: 6284–6291.
Lal A, Kim HH, Abdelmohsen K, Kuwano Y, Pullman Jr. R, Srikatan
S, Subrahmanyam R, Martindale JL, Yang X, Ahmed F, et al. 2008.
p16INK4a translation suppressed by miR-24. Plos One 3: e1864.
Lee I, Ajay SS, Yook JI, Kim HS, Hong SH, Kim NH, Dhanasekaran
SM, Chinnaiyan AM, Athey BD. 2009. New class of microRNA
targets containing simultaneous 59-UTR and 39-UTR interaction
sites. Genome Res 19: 1175–1183.
Lytle JR, Yario TA, Steitz JA. 2007. Target mRNAs are repressed as
efficiently by microRNA-binding sites in the 59UTR as in the
39UTR. Proc Natl Acad Sci 104: 9667–9672.
Ørom UA, Nielsen FC, Lund AH. 2008. MicroRNA-10a binds the
59UTR of ribosomal protein mRNAs and enhances their trans-
lation. Mol. Cell 30: 460–471.
Peters L, Meister G. 2007. Argonaute proteins: Mediators of RNA
silencing. Mol Cell 26: 611–623.
Schnall-Levin M, Zhao Y, Perrimon N, Berger B. 2010. Conserved
microRNA targeting in Drosophila is as widespread in coding
regions as in 39UTRs. Proc Nat Acad Sci 107: 15751–15756.
Selbach M, Schwanha ¨usser B, Thierfelder N, Fang Z, Khanin R,
Rajewsky N. 2008. Widespread changes in protein synthesis
induced by microRNAs. Nature 455: 58–63.
Stark A, Lin MF, Kheradopour P, Pedersen JS, Parts L, Carlson JW,
Crosby MA, Rasmussen MD, Roy S, Deoras AN, et al. 2007.
Discovery of functional elements in 12 Drosophila genomes using
evolutionary signatures. Nature 450: 219–232.
Stripecke R, Oliveira CC, McCarthy JE, Hentze MW. 1994. Proteins
binding to 59 untranslated region sites: A general mechanism for
translational regulation of mRNAs in human and yeast cells. Mol
Cell Biol 14: 5898–5909.
Tay Y, Zhang J, Thomson AM, Lim B, Rigoutsos I. 2008. MicroRNAs
to Nanog, Oct4 and Sox2 coding regions modulate embryonic stem
cell differentiation. Nature 455: 1124–1128.
Thermann R, Hentze MW. 2007. Drosophila miR2 induces pseudo-
polysomes and inhibits translation initiation. Nature 447: 875–878.
microRNA regulation from the 59 UTR and the ORF
Till S, Lejeune E, Thermann R, Bortfeld M, Hothorn M, Enderle D, Download full-text
Heinrich C, Hentze MW, Ladurner AG. 2007. A conserved motif
in Argonaute-interacting proteins mediates functional interactions
through the Argonaute PIWI domain. Nat Struct Mol Biol 14: 897–
Tsai NP, Lin YL, Wei LN. 2009. MicroRNA mir-346 targets the 59-
untranslated region of receptor-interacting protein 140 (RIP140)
mRNA and upregulates its protein expression. Biochem J 424: 411–
Voinnet O. 2009. Origin, biogenesis, and activity of plant microRNAs.
Cell 136: 669–687.
Waud JP, Sala-Newby GB, Matthews SB, Campbell AK. 1996. Engi-
neering the C-terminus of firefly luciferase as an indicator of covalent
modification of proteins. Biochim Biophys Acta 1292: 89–98.
Zdanowicz A, Thermann R, Kowalska J, Jemielity J, Duncan K, Preiss
T, Darsynkiewicz E, Hentze MW. 2009. Drosophila miR2 primarily
targets the m7GpppN cap structure for translational repression.
Mol Cell 35: 881–888.
Moretti et al.
RNA, Vol. 16, No. 12