Analysis of microRNA turnover in mammalian cells following Dicer1 ablation.

Analysis of microRNA turnover in mammalian cells
following Dicer1 ablation
Michael P. Gantier1, Claire E. McCoy1, Irina Rusinova2, Damien Saulep3, Die Wang1,
Dakang Xu1, Aaron T. Irving1, Mark A. Behlke4, Paul J. Hertzog2, Fabienne Mackay3
and Bryan R. G. Williams1,*
1Centre for Cancer Research, 2Centre for Innate Immunity and Infectious Diseases, Monash Institute of Medical
Research, Monash University, Clayton, Victoria 3168, 3Department of Immunology, Monash University,
Alfred Medical Research and Education Precinct, Melbourne, Victoria 3004, Australia and 4Integrated DNA
Technologies Inc., Coralville, IA 52241, USA
Received January 10, 2011; Revised February 27, 2011; Accepted February 28, 2011
ABSTRACT
Although microRNAs (miRNAs) are key regulators
of gene expression, little is known of their overall
persistence in the cell following processing.
Characterization of such persistence is key to the
full appreciation of their regulatory roles.
Accordingly, we measured miRNA decay rates in
mouse embryonic fibroblasts following loss of
Dicer1 enzymatic activity. The results confirm the
inherent stability of miRNAs, the intracellular levels
of which were mostly affected by cell division. Using
the decay rates of a panel of six miRNAs represen-
tative of the global trend of miRNA decay, we estab-
lish a mathematical model of miRNA turnover and
determine an average miRNA half-life of 119 h (i.e.
�5 days). In addition, we demonstrate that select
miRNAs turnover more rapidly than others. This
study constitutes, to our knowledge, the first
in-depth characterization of miRNA decay in mam-
malian cells. Our findings indicate that miRNAs are
up to 10� more stable than messenger RNA and
support the existence of novel mechanism(s)
controlling selective miRNA cellular concentration
and function.
INTRODUCTION
It is now well-established that microRNAs (miRNAs) are
master regulators of most cellular processes. Many viruses
utilize viral-encoded miRNAs during their infectious cycle
(1), mice lacking miRNAs are not viable (2,3) and miRNA
levels are altered in most cancers (4). miRNAs are short
single-stranded RNAs of �22 nt processed from longer
RNA primary transcripts (pri-miRNAs) with high second-
ary structure. Canonical processing of pri-miRNA into
mature miRNA requires sequential cleavage of the pri-
miRNA into a �70-nt miRNA precursor by the endo-
nuclease Drosha (5), and subsequent cleavage into a
�20-bp miRNA duplex by the endonuclease Dicer1 (6).
One strand of this duplex is loaded onto the RNA-induced
silencing complex (RISC) forming the miRISC regulat-
ing cognate messenger RNA (mRNA) stability in GW
bodies (7).
The canonical biogenesis of miRNAs is regulated by
several mechanisms that directly impact on the overall
production of mature miRNAs (8). For instance,
LIN-28 affects the processing of let-7 precursors by
Drosha and Dicer1, allowing it to specifically switch-off
the production of mature forms of let-7 in undifferentiated
embryonic stem cells (8,9). Such regulation of miRNA
processing fine-tunes their intracellular levels, and modu-
lates their biological activity. Indeed, miRNA intracellular
concentration directly relates to their ability to affect
mRNA translation with a suggested threshold of about
100 molecules per cell required for function (10).
Because the intracellular miRNA steady-state levels
result from not only the synthesis of new miRNAs but
also the degradation of previously synthesized miRNAs,
characterization of miRNA persistence following process-
ing is crucial to the understanding of their biological
function.
However, what happens to mature miRNAs is currently
poorly understood, and their overall persistence following
Dicer1 processing is inferred to be ‘highly stable’ from
studies of select miRNAs (8,11). For instance, miR-208
was found to persist in the absence of its precursor for
*To whom correspondence should be addressed. Tel: +613 9594 7166; Fax: +613 9594 7167; Email: bryan.williams@monash.edu
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors.
5692–5703 Nucleic Acids Research, 2011, Vol. 39, No. 13 Published online 29 March 2011
doi:10.1093/nar/gkr148
� The Author(s) 2011. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/
by-nc/2.5), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

>12 days in heart tissue (12) and miR-122 levels remained
unchanged following rapid decrease of pri-miR-122 in
liver tissue (13). Conversely, a rapid decrease of
miRNAs has been observed in neuronal cells, following
blocking of pri-miRNA transcription (14). Furthermore,
selective miRNA stability has been proposed to be
impacted on by various factors, including 30 base modifi-
cations (15), the degree of complementarity to the target
(16) or target abundance (17). With the exception of
miR-451 (18), miRNAs are dependent on Dicer1 matur-
ation to be able to exert their regulatory function (19,20).
Consequently, different approaches have been developed
to disrupt Dicer1 function and characterize the regulatory
roles of miRNAs. To our knowledge, the use of an indu-
cible deletion of Dicer1 to address miRNAs decay has not
been previously adopted.
Here, we investigate the stability of miRNAs following
a global shutdown of miRNA synthesis. Relying on the
inducible genetic ablation of Dicer1 in immortalized em-
bryonic fibroblast cells, we modelled miRNA decay in a
theoretical non-dividing cell and established that the
average miRNA half-life is about 10� that of mRNA,
i.e. about 5 days. In addition, we observed significant vari-
ations in select miRNA half-lives, thereby supporting
the existence of novel mechanism(s) regulating miRNA
function through fine-tuning of steady-state miRNA levels.
METHODS
Ethics statement
The use of animals and experimental procedures were
approved by Monash Medical Centre Ethics Committee
under references MMCA/2008/26/BC and MMCA
2007/07.
Cell culture
Dicer1flox/flox mice (a kind gift from M. McManus,
University of California, San Francisco, CA, USA) (21)
were bred to R26CreER mice expressing the Cre/Esr1
fusion protein from the ROSA26 locus (22). Mouse em-
bryonic fibroblasts (MEFs) from Day 14 embryos were
immortalized following transfection of pSG5-SV40-LT-
Ag (a kind gift from D. Huang, Walter and Eliza Hall
Institute of Medical Research, Melbourne, Australia)
and six successive 1/10 passages. For stable enhanced
green fluorescent protein (EGFP) expression, the MEFs
were transduced with an EGPF-expressing lentiviral con-
struct (pLV-CMV-GFP). Bone marrow was isolated from
the femurs of Dicer1flox/flox xCre/Esr1 mice, and primary
bone marrow derived macrophages (BMDMs) were
differentiated in 20% L929-cell-conditioned medium for
7 days at 37�C in a 5% CO2 atmosphere. BMDMs,
MEFs and HEK293T cells stably expressing GFP (23)
were cultured in complete Dulbecco’s modified Eagle’s
medium (DMEM) (cat. no. 11965-092; Invitrogen
Corporation, Carlsbad, CA, USA) supplemented with
10% sterile foetal bovine serum (#FBS-500, ICPBio Ltd,
Auckland, New Zealand) and 1� antibiotic/anti-mycotic
(Invitrogen) (referred to as complete DMEM). Splenic B
cells were purified using a B Cell Isolation Kit (depletion
of non-B cells) from Miltenyi Biotech. Purity was >95%.
Activation of splenic B lymphocytes was performed in
Roswell Park Memorial Institute (RPMI) 1640+L-glu-
tamine medium (#11875085, Invitrogen) supplemented
with 10% FBS, 55mM b–mercaptoethanol, 2mM L-glu-
tamine, 10mM HEPES and 100U/ml of penicillin–
streptomycin, and cells were stimulated with LPS (25mg/
ml, InvivoGen, San Diego, CA, USA) and murine inter-
leukin 4 (10 ng/ml, eBioscience, San Diego, CA, USA) for
the indicated amount of time.
OHT treatment of the cells
Hydroxy-tamoxifen (OHT) (H7904-5mg, Sigma Aldrich,
St Louis, MO, USA) was resuspended in 0.5ml of 100%
ethanol (resulting in stock solution at �25mM) kept at
�80�C. Prior to cell treatment, the stock solution was first
diluted to 2.5mM in 100% ethanol before being diluted
further to the final concentration in complete DMEM.
For OHT treatment of BMDMs, the cells were differ-
entiated in 20% L929-cell-conditioned medium, and
were washed on Day 3 with fresh medium supplemented
with 20% L929-cell-conditioned medium and 500 nM
OHT. The cells were collected on Day 5 and plated in
24-well plates in 20% L929-cell-conditioned medium,
and collected at Days 6–9 for RNA extraction.
Reverse transcription quantitative real-time PCR
Total RNA containing small RNAs was purified from
cultured cells using an adapted RNeasy protocol
(Qiagen, Valencia, CA, USA) or the mirVana miRNA
isolation kit (Ambion, Austin, TX, USA). For mRNA
quantification, cDNA was synthesized from isolated
RNA using the High-capacity cDNA archive kit
(Applied Biosystems, Foster City, CA, USA) according
to the manufacturer’s instructions. Reverse transcription
quantitative real-time polymerase chain reaction
(RT-qPCR) was carried out with the SYBR GreenERTM
qPCR SuperMix for iCycler� instrument (#11761-500;
Invitrogen Corporation). Murine glyceraldehyde 3-
phosphate dehydrogenase (mGAPDH) (NM_008084)
was used as reference gene and amplified with the follow-
ing primer pair: mGAPDH-FWD: TTCACCACCATGGA
GAAGGC; mGAPDH-REV: GGCATGGACTGTGGTCA
TGA. Detection of the primary miRNAs was carried out
with the following primer pairs: Mir155 (NR_029565.1)—
pri-miR-155-FWD: AAACCAGGAAGGGGAAGTGT
(24); pri-miR-155-REV: ATCCAGCAGGGTGACTCTTG
(24); Mir146a (NR_029558.1)—pri-miR-146a-FWD: GTG
TGTATCCCCAGCTCTGA; pri-miR-146a-REV: CTTCA
CCCCACTCTCTCCAC; Mir21 (NR_029738.1)—
pri-miR-21-FWD: TGTACCACCTTGTCGGATAGC;
pri-miR-21-REV: AAGGGCTCCAAGTCTCACAA. Wild-
type Dicer1 mRNA (NM_148948.2) was detected with
mDicer1-RT-FWD: TCTGCAGGCTTTTACACACG and
mDicer1-Exon-REV: CCAATGATGCAAAGATGGTG.
Each amplicon was sequence verified and used to generate
a standard curve for the quantification of gene expression.
For mature miRNA detection, miRNA TaqMan� assays
(Applied Biosystems) for the indicated miRNAs were
used according to the manufacturer’s instructions, where
Nucleic Acids Research, 2011, Vol. 39, No. 13 5693

10–30 ng of total RNA was reverse transcribed with pools
of up to eight miRNA-specific reverse transcription
primers (with the TaqMan� MicroRNA Reverse
Transcription Kit). With the exception of mmu-miR-155
and snoRNA202, all the individual Taqman� assays used
were compatible between human and mouse. miRNA
levels were determined by RT-qPCR with the TaqMan�
Universal PCR Master Mix on a 7900 RT-PCR system
(Applied Biosystems), and fold changes in expression were
calculated by the 2���Cq (Cq=quantification cycle)
method using snoRNA202 (for all mouse experiments),
or miR-16 (for HEK293T cells) as reference genes.
Low density miRNA arrays
TaqMan� Array Rodent MicroRNA (A Cards v2.0,
Applied Biosystems) were used for the global detection
of miRNAs from one biological sample set. Briefly,
900 ng of total RNA containing small RNAs was reverse
transcribed using the MegaplexTM RT Primers, Rodent
Pool A (Applied Biosystems) with the TaqMan�
MicroRNA Reverse Transcription Kit and each plate
was consecutively ran using the TaqMan� Universal
Master Mix II on the 7900 RT-PCR system, according
to the manufacturer’s instructions. Simultaneous analysis
of the three different plates (one plate at Days 3–5) was
carried out using the RQ Manager software. To determine
significant miRNA expression, we performed the follow-
ing filtering. (i) We only considered miRNAs that had Cq
values lower than 37 for all three time-points. (ii) We used
the average of four RNAU6 probes present on each array
as reference gene (giving �Cq values), and calculated
��Cq values using Day 3 as control condition. Fold
decrease of miRNA levels relative to Day 3 was inferred
using 2(���Cq) values. (iii) We limited our analysis to
miRNAs levels that consistently decreased with time (i.e.
Day 3>Day 4>Day 5)—hereby identifying 104
miRNAs. Importantly, out of the 158 miRNAs detected
by PCR-array with Cqs inferior to 37 at each time-point,
17 were such that the concentrations of miRNAs at Day
3>Day 4 and Day 3>Day 5 but not Day 4>Day 5, 21
had increased miRNA expression at either Day 4 or Day 5
compared with Day 3 (most likely from technical PCR
outlier) and 17 did not decrease with time (i.e. Day
3�Day 4�Day 5) (Supplementary Table S1). Recent
evidence suggests that five of these non-decreasing
RNAs (miR-509-3 p, miR-658, miR-207, miR-680 and
miR-687—see Supplementary Table S1 in bold) would in
fact not be miRNAs (25), hence validating our method of
discrimination and the robustness of the RT-qPCR array
platform. Our data on 104 miRNAs, therefore, rely on 3/4
of the miRNA possibly expressed by the MEFs, and a less
conservative approach including the 17 miRNAs with
decreased expression at Days 4 and 5 compared with
Day 3 gave very similar averaged decrease (data not
shown). Similar results were obtained using normalization
to other controls such as snoRNA202 or snoRNA135.
To account for possibly rapidly decaying miRNAs that
would not be detected after Day 3, we also analysed the
global trend of miRNA decay between Days 3 and 4
(Supplementary Table S2), and Days 4 and 5
(Supplementary Table S3), including miRNAs that were
not detected at one of the time-points (Cq=40). (i) We
only considered miRNAs that had Cq values lower than
40 at Day 3, and (ii) that had a lower Cq value at Day 3
than Day 4. One hundred and eighty-three miRNAs
matched these criteria, with an averaged 64% decrease
of miRNA expression between Days 3 and 4 (when
normalized to RNAU6 as detailed above). (iii) We next
repeated this analysis and assessed the expression of
miRNAs between Days 4 and 5, filtering miRNAs that
had Cq values lower than 40 at Day 4, and a lower Cq
value at Day 4 than Day 5. One hundred and fifty-four
miRNAs matched these criteria, with an averaged 68%
decrease of miRNA expression between Days 4 and 5.
These data, although approximate given that the
non-detected miRNAs were attributed an arbitrary Cq
value of 40, support our claim that the global trend of
miRNA decay is constant between Days 3 and 4, and 4
and 5, even when including miRNAs that are possibly
rapidly degraded.
Synthetic RNAs
All small interfering RNA (siRNA) and si-miRNA
duplexes were synthesized as single-stranded RNAs by
Integrated DNA Technologies (IDT) with HPLC purifi-
cation, and resuspended in duplex buffer (100mM potas-
sium acetate, 30mM HEPES, pH 7.5, DNase–RNase free
H2O) to a concentration of 80 mM. Annealing was per-
formed by incubating the complementary single-stranded
RNAs at 92�C for 2min and leaving them for 30min at
room temperature (giving a final concentration of 40 mM).
miR-155-S: 50 CCCUAUCACAAUUAGCAUUAAUU;
miR-155-AS: 50 UUAAUGCUAAUUGUGAUAGGGG
U. miR-21-S: 50 AACAUCAGUCUGAUAAGCUAUU;
miR-21-AS: 50 UAGCUUAUCAGACUGAUGUUGA.
Reverse transfection of small RNAs
For Figure 1D, 4.5ml of Lipofectamine 2000 was diluted
in 300 ml of Opti-MEM and 2.25 ml of mi-siRNA mol-
ecules (diluted to 4 mM in duplex buffer) was added such
that the final concentration of each siRNA was 5 nM per
well (the volumes indicated are for biological triplicate).
After 20min of incubation, 100 ml of the mix was added
directly into each well of a 24-well plate. HEK293T-GFP
cells (100 000) resuspended in 500 ml of antibiotic-free
DMEM (supplemented with 10% FBS) were added to
each well giving a final volume of 600 ml/well. The cells
were incubated at 37�C for 7 h before being washed with
fresh medium. For Figure 2B, 4.8 ml of Lipofectamine
2000 was diluted in 300ml of Opti-MEM and 3 ml of
siRNA molecules (diluted to 4 mM in duplex buffer) was
added such that the final concentration of siRNA in each
well was 10 nM (the volumes indicated are for biological
triplicate in two cell types). After 20min of incubation,
50 ml of the mix was added directly into each well of a
96-well plate (in triplicate for each cell type). MEFs (20
000; previously treated with OHT or not) in 150 ml of
antibiotic-free DMEM (supplemented with 10% FBS)
were added to each well giving a final volume of 200 ml/
well.
5694 Nucleic Acids Research, 2011, Vol. 39, No. 13

Fluorescent-based measure of EGFP knockdown
EGFP down-regulation in MEFs stably expressing EGFP
cells was measured using a Fluostar OPTIMA
plate-reader as previously reported (23). After 24 h
siRNA treatment, the supernatants were discarded and
50 ml of PBS was added to the cells. A standard curve
was generated by serially diluting a recombinant
EGFP-fusion protein to cover a range from 150 ng/ml to
4.68 ng/ml in 50 ml PBS. EGFP concentration in each well
was then inferred from the fluorescence at ex485/em520
correlated to the standard curve, after background
autofluorescence (from PBS only) correction.
High-content screening assay
Cell proliferation was assayed using high-content screen-
ing assay (HCSA) with a Cellomics ArrayScan HCS
Reader (Thermo Fisher Scientific). EGFP cells were
counted per captured field according to fluorescent inten-
sity and morphology (with the HCS Reader V-6.6.1.2
software, Thermo Fisher Scientific), up to 3000+ cells
per well. The average number of cells per field was used
as a quantification of cell number at each time-point.
Mathematical modelling
To build a mathematical model of miRNA degradation
independent of recombination efficiency and cell prolifer-
ation, we decomposed our analysis in three independent
steps. We first established a model of the intracellular
miRNA concentration per cell (yi= y1,. . ., yn) following
entire blockage of de novo miRNA synthesis, function of
time (ti= t1,. . ., tn) and assuming that it was solely
impacted by the amount of cells that had divided (and
that each cell division halved the initial concentration of
miRNA). We obtained experimental values from the
averaged cell proliferations at different time-points
(Supplementary Figure S3B, see ‘measured’ curve).
Relying on this data set, we modelled miRNA decay due
to cell proliferation under the assumption that this
function is an exponential decay: FcellP (t)= a exp (–cti),
where a is the initial quantity at time t1, and c is the
constant of decay. To identify the co-efficients a and c,
we used a non-linear least-square fitting algorithm. The
method is based on the minimization of the distance (or
error) D between the function FcellP (ti) and the data at all
points yi. The distance D is defined as:
D ðrÞ ¼
Xn
i¼1 jFcellPðti,rÞ � yij
2
where r(a, c) is the vector with adjustable parameters of
function FcellP. The minimization of the function D (r)
0 10 20 30 40 50 60 70
0
20
40
60
80
100
0
5
10
15
20
pri-miR-155
Time (h) following LPS treatment
Re
la
tiv
e
m
iR
-
15
5
le
ve
l
R
elative
pri
-m
iR
-155 le
vel
0 10 20 30 40 50 60 70
0
1
2
3
4
5
0
1
2
3
4
5miR-21 pri-miR-21
Time (h) following LPS treatment
Re
la
tiv
e
m
iR
-2
1
le
ve
l
R
elative
pri
-m
iR
-21 le
vel
miR-155A
B
0 10 20 30 40 50 60 70
0
2
4
6
8
10
0
3
6
9
12
15
pri-miR-146a
Time (h) following LPS treatment
Re
la
tiv
e
m
iR
-1
46
a
le
ve
l
R
elative
pri
-m
iR
-146a le
vel
miR-146aC
0 10 20 30 40 50
0
100
200
300
400
0
1000
2000
3000
4000
5000
miR-21 miR-155
Time (h) following transfection
Re
la
tiv
e
m
iR
-
21

le
ve
l
Relative
m
iR-155
le
vel
D
Figure 1. Sustained persistence of innate immune miRNAs in the
absence of their precursors. (A–C) BMDMs were treated with 10 ng/
ml of LPS and incubated for the indicated time before being lysed in
RNA extraction buffer. Expression levels of pri- and mature miRNAs
were determined by RT-qPCR, reported to that of the internal control
(GAPDH and snoRNA-202, for the pri-miRNA and the mature
miRNA, respectively) and are shown relative to the values at the
initial time-point. The values are averaged from three biological repli-
cates, and are representative of two independent experiments.
(D) HEK293T-GFP cells were reverse-transfected with 5 nM of both
si-miR-21 and si-miR-155 synthetic miRNAs, washed with fresh
medium after 7 h and collected in RNA extraction buffer at the
indicated times. miRNA levels were determined by RT-qPCR
reported to the expression of the house-keeping control RNA
miR-16. The values are presented relative to that at the initial
time-point, averaged from three biological replicates and are represen-
tative of three independent experiments. (A–D) Standard error of the
mean (SEM) is shown.
Nucleic Acids Research, 2011, Vol. 39, No. 13 5695

with respect to the parameter vector r required the zero
gradient condition:
rDðrÞ ¼ JTðFcellPðrÞ � yÞ ¼ 0,
where J is Jacobian. The values of the vector r are
obtained using the Newton–Raphson iteration method.
The process starts with an initial parameter vector r0
(a0, c0) and is corrected to be ri= r0+�ri in the next it-
eration. The increment parameter vector �r therefore
satisfies the system of linear equations:
�r ¼ �H�1JTðFcellPðrÞ � yÞ
and H can be approximated by H= JTJ for a system with
moderate non-linearity.
The second step of modelling defines a function of
miRNA intracellular concentration averaged from two
distinct cell populations; one population lacking de novo
miRNA synthesis (ensuing Dicer1 genetic ablation)
f1 (t)= b exp (�k1t), and the other that escaped Dicer1
recombination and producing normal levels of miRNAs
f2 (t)= b (1� k2t), where b is the initial quantity at the
start time t1. The sum of these two functions models the
experimental measurements of the average miRNA con-
centration per cell yi at a time ti. As a result we obtain the
function F (t)= b (m exp (�k1t)+(1�m) (1 – k2t)), where
m is the relative proportion of the two cell populations
(0�m� 1), and was experimentally determined by the
quantification of wild-type Dicer1 mRNA by RT-PCR
(Figure 3A). We estimated parameters k1, k2 and b with
m=0.995 of recombined cells (Figure 3A and C; 500 nM),
applying the non-linear least-square fitting algorithm ex-
plained above (the resulting fitted model curve is shown in
Figure 3D, FM-99.5%). Next, we validated the relevance
of our model by changing the relative proportions of the
mixed cell population to m=0.5 (i.e. 50% of recombined
cells—Figure 3D, see curve M-50%) and also to m=1
(i.e. where 100% of the cells would lack de novo miRNA
synthesis and consequently where F (t)= f1 (t), Figure 3D,
see curve M-100%).
For the last step of our analysis, we assessed the intra-
cellular miRNA decrease that is independent of that due
siEGFP-N
0.0
0.2
0.4
0.6
0.8
1.0
NT OHT
R
el
at
ive

G
FP

pr
ot
ei
n

le
ve
l
siGFP19+2 siEGFP-N-MIS
B
500 100 10 2 0
0.0
0.2
0.4
0.6
0.8
1.0
OHT (nM)
Re
la
tiv
e
D
ic
er

m
R
N
A








ex
pr
es
si
on
A
DC
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Let-7i
miR-125b
miR-21
miR-155
miR-143
miR-19b
Re
la
tiv
e
m
iR
NA

le
ve
l
72 82 92 102 112 122
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Re
la
tiv
e
m
iR
NA
le
ve
l
miR-155
miR-125b
93%
72 82 92 102 112 122
Time (h) following OHT treatment Time (h) following OHT treatment
Figure 2. Determination of global miRNA decay. (A) Actively dividing Dicer1flox/floxxCre/Esr1 MEFs were treated overnight with the indicated
concentration of OHT, washed with fresh medium and further incubated for 48 h before being harvested for RNA extraction. Wild-type Dicer1
mRNA was quantified by RT-qPCR and reported to the expression of GAPDH. The data are presented relative to that obtained without OHT
treatment (condition ‘0’), are averaged from biological duplicate and are representative of three independent experiments. (B) Actively dividing
Dicer1flox/floxxCre/Esr1 GFP MEFs were treated (OHT) or not (NT) overnight with 500 nM OHT, and expanded for a further 48 h before being
collected and reverse transfected (on Day 3 following initial OHT addition) with 10 nM of the indicated siRNAs. The cells were further incubated for
24 h and EGFP expression was quantified by fluorescent spectrophotometer [see ‘Methods’ section and (23)]. The data are averaged from two
independent experiments in biological triplicate and normalized to the averaged values obtained for the non-targeting siEGFP-N-MIS sequence for
each cell type. siEGFP-N-MIS contains two mismatches in the targeting strand of the siRNA compared to siEGFP-N and both siRNAs are Dicer1
substrates (Supplementary Figure S1C); siGFP19+2 is a Dicer1 product (23). (C and D) Actively dividing Dicer1flox/floxxCre/Esr1 MEFs treated with
500 nM OHT were expanded and collected at Days 3–5 following initial OHT treatment. For each miRNA measured, the levels were reported to that
obtained at Day 3 (T=72h following OHT treatment). (C) The average level of 104 miRNAs significantly detected by low-density PCR microarray
(‘Methods’ section and Supplementary Table S1) is shown with fitted exponential regression. (D) Individual miRNA PCRs were carried out on the
samples analysed with low-density PCR microarray and biological replicates. Highlighted in grey is the variation of miRNA decay observed between
the six miRNAs analysed, and fitted exponential regression is shown for miR-125 b and miR-155. The data are averaged from biological triplicate
and are representative of three independent experiments. SEM (B and D) and standard deviation (SD) (C) are shown.
5696 Nucleic Acids Research, 2011, Vol. 39, No. 13