Regulation of ABCA1 Protein Expression and Function in
Hepatic and Pancreatic Islet Cells by miR-145
Martin H. Kang,* Lin-Hua Zhang,* Nadeeja Wijesekara, Willeke de Haan, Stefanie Butland,
Alpana Bhattacharjee, Michael R. Hayden
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
Objective—The ATP-binding cassette transporter A1 (ABCA1) protein maintains cellular cholesterol homeostasis in several
different tissues. In the liver, ABCA1 is crucial for high-density lipoprotein biogenesis, and in the pancreas ABCA1 can
regulate insulin secretion. In this study, our aim was to identify novel microRNAs that regulate ABCA1 expression in
these tissues.
Approach and Results—We combined multiple microRNA prediction programs to identify 8 microRNAs that potentially
regulate ABCA1. A luciferase reporter assay demonstrated that 5 of these microRNAs (miR-148, miR-27, miR-144, miR145, and miR-33a/33b) significantly repressed ABCA1 3′-untranslated region activity with miR-145 resulting in one of
the larger decreases. In hepatic HepG2 cells, miR-145 can regulate both ABCA1 protein expression levels and cholesterol
efflux function. In murine islets, an increase in miR-145 expression decreased ABCA1 protein expression, increased total
islet cholesterol levels, and decreased glucose-stimulated insulin secretion. Inhibiting miR-145 produced the opposite
effect of increasing ABCA1 protein levels and improving glucose-stimulated insulin secretion. Finally, increased glucose
levels in media significantly decreased miR-145 levels in cultured pancreatic beta cells. These findings suggest that miR145 is involved in glucose homeostasis and is regulated by glucose concentration.
Conclusions—Our studies demonstrate that miR-145 regulates ABCA1 expression and function, and inhibiting this microRNA
represents a novel strategy for increasing ABCA1 expression, promoting high-density lipoprotein biogenesis in the liver,
and improving glucose-stimulated insulin secretion in islets. (Arterioscler Thromb Vasc Biol. 2013;33:2724-2732.)
Key Words: ABCA1 protein ◼ cholesterol efflux regulatory protein ◼ lipid metabolism ◼ microRNAs
M
utations in both alleles of the ATP-binding cassette
transporter A1 (ABCA1) gene results in Tangier disease, an autosomal recessive disorder whose clinical hallmark
is the almost complete absence of high-density lipoprotein
(HDL)-cholesterol in the plasma of affected individuals.1–3
The ABCA1 protein is ubiquitously expressed in murine4 and
human5 tissues. The function of ABCA1 is to remove cholesterol from cells and does so by effluxing free cholesterol to
apolipoprotein A-I (apoA-I) protein acceptors.6
Studies in animal models demonstrate the critical role of
ABCA1 in cholesterol metabolism for different tissues. In
the liver, ABCA1-mediated efflux of cholesterol to apoAI generates nascent or immature HDL particles.7 Targeted
loss of hepatic ABCA1 in mice results in reduced levels of
total and HDL-cholesterol by >80% compared with control
mice.8 Disruption of ABCA1 in macrophages increases the
proinflammatory response to lipopolysaccharides, as well as
increasing the development of atherosclerosis in hypercholesterolemic mouse models.9,10 In mice containing inactivation of
ABCA1 in the central nervous system, structural and functional deficits are observed, including a reduction in the number of synaptic vesicles, attenuated locomotor activity, and a
reduction in acoustic startle response.11 ABCA1 in pancreatic
β-cells functions to maintain islet cholesterol homeostasis and
proper insulin secretion after glucose stimulation.12 Human
studies demonstrate that some carriers of heterozygous mutations in ABCA1 have impaired insulin secretion,13 and individuals diagnosed with Tangier disease can manifest with
diabetes mellitus.14
ABCA1 expression is regulated by epigenetic,15 transcriptional,16 post-transcriptional,17–24 and post-translational25
mechanisms. Potent transcriptional liver X receptor (LXR)
agonists of ABCA1 exist; however, these agonists also activate genes involved in hepatic triglyceride synthesis, fatty acid
synthesis, hypertriglyceridemia, and hepatic steatosis.16
Proteins undergo post-transcriptional regulation through
multiple mechanisms, including protein binding,26 alternative polyadenylation sites,27 and the binding of small noncoding microRNAs (miRNAs) to the 3′-untranslated regions
(3′UTRs) of transcripts.26 Protein translation is partly dependent on the length of the 3′UTR in mRNA. Regions longer
than 1 kb decay much faster than shorter 3′UTRs, partly
because of a greater number of miRNA binding sites present
in long 3′UTRs.28 ABCA1 has an exceptionally long 3′UTR
Received on: December 5, 2012; final version accepted on: October 4, 2013.
From the Centre for Molecular Medicine and Therapeutics, University of British Columbia, Vancouver, BC, Canada.
*These authors contributed equally to this article.
The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.113.302004/-/DC1.
Correspondence to Michael R Hayden, 950 W 28th Ave, Vancouver, BC, V5Z 4H4, Canada. E-mail [email protected]
© 2013 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org
2724
DOI: 10.1161/ATVBAHA.113.302004
Kang et al Regulation of ABCA1 by miR-145 2725
Nonstandard Abbreviations and Acronyms
3′UTR
ABCA1
apoA-I
HDL
miRNA
3′-untranslated region
ATP-binding cassette transporter A1
apolipoprotein A-I
high-density lipoprotein
microRNA
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
at 3309 base pairs (bps). In comparison, the average length
of a 3′UTR in a study of 1500 human genes was only 755
bp.29 This suggests that multiple miRNAs regulate ABCA1.
Indeed, several miRNAs have been validated recently to regulate ABCA1 expression and function.17–24
MiR-33a was the first miRNA demonstrated to regulate
ABCA1 expression in hepatic cells and macrophages.17,19 Few
miRNAs are exclusive to a given tissue, and only one third
of all miRNAs demonstrate preferential tissue expression.30
MiR-33a is ubiquitously expressed17,18 and it can regulate
ABCA1 expression and function in multiple cell types. We
recently demonstrated that miR-33a regulates ABCA1 expression, cholesterol levels, and glucose-stimulated insulin secretion in islets.31
In this study, we identify miR-145 as a novel miRNA regulator of ABCA1 in both hepatic and pancreatic cell lines, as
well as in pancreatic islets.
Materials and Methods
Materials and Methods are available in the online-only Supplement.
Results
ABCA1 Transcript Has Numerous
Binding Sites for miRNAs
A luciferase reporter construct containing the 3′UTR (3309
bp) of human ABCA1 was transfected into murine (MIN6 pancreatic cells, 3T3L1 adipocytes) and human (HEK293 embryonic kidney, skin fibroblasts, HepG2 hepatocytes) cell lines,
with activity measured 24 hours after transfection. Decreased
activity was observed in all cell lines (Figure I in the onlineonly Data Supplement) and suggests that the 3′UTR is a critical component in controlling ABCA1 expression levels.
We used an in silico approach to identify potential miRNAs that could bind the 3′UTR of ABCA1.32 The lack of
definitive miRNA target recognition patterns means most
bioinformatic prediction programs have a high proportion
of false-positives.33 Commonly used programs, such as
TargetScan and PicTar, have a precision of ≈50% and sensitivity between 6% and 12%.32 Prediction programs are
composed of unique algorithms that identify miRNA targets
based on different factors such as site conservation, nucleotide composition, and location of miRNA binding sites.32 We
used 5 programs (TargetScanHuman 5.2, PicTar, microRNA.
org, miRDB, FindTar) to identify miRNA binding sites in
the 3′UTR of ABCA1 (Table I in the online-only Data
Supplement). To narrow down the total number of potential
miRNAs targeting ABCA1, we combined the results from
these 5 programs. This approach can increase specificity but
does so at a cost of decreased sensitivity.32 We also limited
the miRNAs we investigated to those known to be expressed
in the liver because of the critical role of hepatic ABCA1
in HDL biogenesis.8 By selecting miRNAs chosen by 4 of
5 programs, and those miRNAs known to be expressed in
the liver, we identified 8 miRNAs that potentially regulate
ABCA1: miR-17, miR-145, miR-19a, miR-96, miR-33a/33b,
miR-27a, miR-148, and miR-144 (Table II in the online-only
Data Supplement).
Five miRNAs Decrease ABCA1
Activity in a Reporter Assay
To identify the miRNAs that can regulate ABCA1, oligonucleotide mimics representing each of the 8 miRNAs were
cotransfected with the ABCA1 3′UTR reporter construct into
U343 cell lines. Five miRNAs demonstrated significant inhibition of luciferase activity: miR-148 by 16%, miR-27a by
20%, miR-144 by 25%, miR-145 by 28%, and miR-33a/33b
by 52%/48% (Figure 1A). Studies by other groups have since
validated miR-33a/33b,17–19 miR-144,22,23 and miR-27a24 as
regulating ABCA1. Expression of miR-145 resulted in the
second largest repression of activity after miR-33a. Two
miR-145 binding sites are predicted in the 3′UTR of ABCA1,
a site downstream of the stop codon (site 1, 109–115), and
a site upstream of the poly A tail (site 2, 3114–3120; Figure
II in the online-only Data Supplement). Site-directed mutagenesis was performed on the 2 binding sites to identify
which site(s) miR-145 binds to regulate ABCA1 (Figure 1B).
Reporter constructs containing mutations to site 1 eliminated
the inhibitory effect of miR-145 to the same extent as mutations to both sites (mutations to site 1 and mutations to site
2; Figure 1C), suggesting that site 1 is the primary binding
site for miR-145.
MiR-145 Regulates ABCA1 in HepG2 Cells
We focused on the regulation of ABCA1 by miR-145 for 2
reasons: (1) the overexpression of miR-145 resulted in the
second largest repression of 3′UTR activity after miR-33a;
and (2) the primary miR-145 binding site possesses features
known to increase the efficacy of miRNA/mRNA target interactions. These features include a binding site ≥15 nucleotides
away from the stop codon while also being located as far from
the center of UTRs as possible.34 Of all the binding sites for
the 5 miRNAs identified in our reporter assay as regulating
ABCA1, miR-145’s primary binding site (site 1, 109–115) has
a distance of >15 nucleotides from the stop codon while also
being located furthest from the center of the ABCA1 3′UTR
(Figure II in the online-only Data Supplement).
To demonstrate that miR-145 can regulate ABCA1 in
hepatic cells, we manipulated miR-145 expression in the
HepG2 cell-line and assessed ABCA1 protein expression through immunoblotting studies and ABCA1 function
through cholesterol efflux activity assays. Transfection of
pre–miR-145 oligonucleotide mimics decreased ABCA1
protein levels (Figure 2A) and significantly reduced ABCA1
cholesterol efflux to apo A-I in HepG2 cells (Figure 2B). In
all our experiments, scrambled control miRNA was used as a
2726 Arterioscler Thromb Vasc Biol December 2013
A
1.2
Relative luciferase activity
(Normalized to control)
1
p<0.01
0.8
p<0.01
0.6
-28%
p<0.01
p<0.01
p<0.01
0.4
-52%
p<0.01
-20%
-16%
-25%
-48%
0.2
0
SC
miR-17 miR-145 miR-19a miR-96 miR-33a miR-33b miR-27a miR-148 miR-144
B
MU1
(109-115)
MU2
(3114-3120)
1.1
Relative luciferase activity
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
ACUGGAU
TCUAGAU
WT 5’ AAGAAGUAA
MU1 5’ AAGAAGUAA
C
2K
1K
Luciferase
3K
AAAAAAA
WT 5’ CUAUUUUUGCACUGGAAU
MU2 5’ CUAUUUUUGCGCUGTAAU
p<0.01
1
0.9
SC
0.8
Figure 1. A reporter assay validated 5 microRNAs
(miRNAs) as regulating ATP-binding cassette transporter A1 (ABCA1) 3′-untranslated regions (3′UTR)
activity. A, The ABCA1 3′UTR reporter construct
was cotransfected into U343 cells along with 50
nmol/L of miR-17, miR-145, miR-19, miR-96,
miR-33a/33b, miR-27a, miR-148, and miR-144.
Dual-luciferase activity was measured with the error
bars representing SD from measurements made in
triplicate. The data are expressed as the percentage loss in activity with miRNA transfected cells
to cells transfected with control miRNA. SC indicates scrambled control (Ambion). B, The human
ABCA1 3′UTR contains 2 binding sites for miR145 at nucleotides 109 to 115 and 3114 to 3120
downstream of the ABCA1 stop codon (predicted
binding sites to the seed sequence of miR-145 by
TargetScanHuman 5.2). Two nucleotides in both
seed sequences were mutated using site-directed
mutagenesis: mutations to site 1 (MU1)=109 to
115 (110:A→T, 113:G→A) and MU2=3114 to 3120
(3114:A→G, 3118:G→T). C, The luciferase reporter
constructs containing either the wild-type (WT)
or miR-145 binding site mutations (MU1 only or
MU1+MU2) were cotransfected into U343 cells
along with control (SC) or miR-145 miRNA mimics.
The data are expressed as a change in luciferase
activity after miR-145 transfection, compared with
cells transfected with control (SC) miRNA. The error
bars represent SD.
miR-145
0.7
0.6
0.5
WT
MU1
MU1+MU2
negative control, and miR-33a was used as a positive control.
Measurement of miRNA expression levels after transfection of the mimics demonstrated that the effect of miR-145
on ABCA1 expression and function is specific to increased
miR-145 expression, and not a result of changes in miR-33a
expression (Figure 2C and 2D). Previous studies have suggested that miR-145 is cotranscribed with another miRNA,
miR-143 that lies within 1.7 kilobases on chromosome 18.35
We examined whether miR-143 could regulate ABCA1, but
observed no changes to ABCA1 protein expression or cholesterol efflux function after overexpression of miR-143 in
HepG2 cells (Figure 2E and 2F). The specificity of miR-145
regulation on ABCA1 function was further demonstrated
when we observed that miR-145* (passenger strand) does
not decrease ABCA1 function (Figure IIIA and IIIB in the
online-only Data Supplement). To validate the findings from
our luciferase reporter assay, we analyzed the effect of miR17 on ABCA1 expression and function. Although miR-17
was 1 of the 8 miRNAs predicted to target ABCA1 in silico
(Table II in the online-only Data Supplement), our reporter
assay revealed no significant changes to activity after the
expression of miR-17. This was confirmed when our immunoblotting studies and cholesterol efflux assays demonstrated
no changes to ABCA1 expression and function after miR17 expression (Figure IVA and IVB in the online-only Data
Supplement). Our findings validate miR-145 as a regulator of
ABCA1 in HepG2 hepatic cells, a cell-type in which ABCA1
function is critical for HDL biogenesis.
Combination of miR-33a and miR-145 Does Not
Result in Additive or Synergistic Effects on ABCA1
The majority of miRNAs have subtle effects on their predicted targets.36 A combination of miRNAs may additively
or synergistically enhance the repressive effects over a single
miRNA acting alone.37 However, we found that the combination of miR-33a and miR-145 did not enhance repression
of ABCA1 protein levels (Figure 3A and 3B) or cholesterol
efflux activity (Figure 3C) compared with either miRNA
expressed singly.
Inhibiting miR-145 using single-stranded RNA inhibitor
sequences in HepG2 cells increased ABCA1 protein levels
(Figure 3D) and increased cholesterol efflux activity by 29%
(Figure 3E). Again we did not observe an enhanced effect on
ABCA1 protein expression or efflux function when combining the 2 miRNA inhibitors compared with each inhibitor
expressed independently.
B
A
Control
miR-33a
miR-145
ABCA1
250kDa
37kDa
GAPDH
C
% Cholesterol Efflux to ApoA-I
Kang et al Regulation of ABCA1 by miR-145 2727
3.5
3
2.5
1
0.5
Control
miR-33a
miR-145
miR-145 relative expression
levels (Normalized to U6)
60
20
15
10
5
50
40
30
20
10
0
0
miR-33a
Cont
miR-145
F
Control
miR-143
miR-145
ABCA1
GAPDH
% Cholesterol Efflux to ApoA-I
miR-33a relative expression
levels (Normalized to U6)x100
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
37kDa
0
D
Cont
250kDa
p<0.017
1.5
25
E
p=0.01
2
miR-33a
miR-145
2.5
p=0.55 (NS)
2
p=0.001
1.5
Figure 2. MiR-145 regulates ATP-binding cassette
transporter A1 (ABCA1). A, Western blot analysis of
ABCA1 in HepG2 cells transfected with 50 nmol/L
of control miR, miR-33a, or miR-145 for 48 hours.
The loading control is GAPDH. B, Cholesterol efflux
activity in HepG2 cells transfected with 50 nmol/L
of control miR, miR-33a, or miR-145 for 48 hours
(n=7). C, Relative expression levels of miR-33a in
HepG2 cells transfected with 50 nmol/L of control
(Cont) miR, miR-33a, or miR-145 for 48 hours. All
expression levels are normalized to U6 RNA. D,
Relative expression levels of miR-145 in HepG2
cells transfected with 50 nmol/L of control (Cont)
miR, miR-33a, or miR-145 for 48 hours. All expression levels are normalized to U6 RNA. E, Western
blot analysis of ABCA1 in HepG2 cells transfected
with 50 nmol/L of control miR, miR-143, or miR145 for 48 hours. The loading control is GAPDH.
F, Cholesterol efflux activity in HepG2 cells transfected with 50 nmol/L of control miR, miR-143, or
miR-145 for 48 hours (n=6). All data are expressed
as the mean±SD with the number of experiments
performed indicated in parentheses. Significance is
set at P≤0.017 for 3 comparisons.
1
0.5
0
Control
miR-143
MiR-145 Regulates ABCA1
Expression in β-Cells and Islets
We generated adenoviruses expressing miR-33a, miR-145,
along with their respective inhibitors. Transduction of HepG2
cells with adenoviruses resulted in similar effects on ABCA1
protein expression and cholesterol efflux activity as transfecting with their respective oligonucleotide counterparts (Figure
VA and VB in the online-only Data Supplement).
We previously demonstrated that miR-33a influences islet
cell cholesterol content and islet function through ABCA1.31
The identification of miR-145 expression in islets (Figure VI
in the online-only Data Supplement) prompted us to investigate whether miR-145 regulates ABCA1 in MIN6 pancreatic β-cells. Because of low basal expression of ABCA1 in
MIN6 cells, the cells were treated with 10 μmol/L of liver
X receptor agonists (TO-901317) before expressing miR-145
(Figure VIIA and VIIB in the online-only Data Supplement).
Adenoviral-miR-145 expression in liver X receptor-treated
MIN6 cells significantly decreased ABCA1 protein levels
(Figure VIIC and VIID in the online-only Data Supplement),
while significantly elevating total cellular cholesterol content
(Figure VIIE in the online-only Data Supplement). These findings suggest that miR-145 regulates ABCA1 protein expression and cholesterol content in β-cells.
To determine the effects of miR-145 on islet function, we
used primary pancreatic islets isolated from mice. Murine primary islets were used rather than MIN6 cells because of the
low levels of endogenous miR-145 in MIN6 cells (Figure 4A),
miR-145
the low basal levels of ABCA1 in MIN6 cells (Figure VIIA
in the online-only Data Supplement), and primary islets representing a more physiologically relevant assessment of islet
cell function than the MIN6 cell line.
In primary islets, adenoviral-miR-145 expression decreased
ABCA1 protein levels, whereas adenoviral-inhibitors to miR145 increased ABCA1 expression (Figure 4B and 4C). In a cell,
miRNAs range in expression from <1 copy to >10 000 copies.38 Expression analysis by reverse transcription-quantitative
polymerase chain reaction (TaqMan) revealed that miR-145 is
expressed in multiple human (Figure VIA in the online-only
Data Supplement) and murine (Figure VIB in the online-only
Data Supplement) tissues. The regulation of miRNA targets
is partially dependent on the expression level of miRNA in
cells and tissues, and high miRNA expression is expected to
be associated with enhanced suppression of target transcripts.
We compared the concentrations of miR-145 with miR-33a
in murine liver and islets, as well as in MIN6 cells, by absolute
quantification reverse transcription-quantitative polymerase
chain reaction using the standard curve method (Figure
VIII in the online-only Data Supplement). After correcting
for primer efficiencies, we found miR-145 and miR-33a to
be equally expressed in murine liver and islets (Figure 4A).
Based on their expression profiles, we expect that inhibiting
miR-145 in islets could lead to similar increases in ABCA1
protein expression as inhibiting miR-33a. Indeed, we found
that in primary murine islets, inhibiting miR-145 resulted in
equivalent increases to ABCA1 protein levels as inhibiting
2728 Arterioscler Thromb Vasc Biol December 2013
A
Control
miR33a
miR145
+
+
+
+
+
+
+
+
+
-
-
-
-
-
-
+
+
+
-
-
-
+
+
+
-
-
-
-
-
-
+
+
+
+
+
+
250kDa
ABCA1
37kDa
C
80
70
2.5
p<0.001 p<0.001
60
p<0.001
50
40
20
-38%
-45%
-44%
Cont
1
-26%
0
miR-33a miR-145 miR-33a+
+Cont
+Cont
miR-145
Cont
+
+
+
+
+
+
+
+
+
-
-
-
-
-
-
+
+
+
-
-
-
+
+
+
-
-
-
-
-
-
+
+
+
+
+
+
p<0.01
-25%
-28%
miR-33a miR-145 miR-33a+
+Cont
+Cont miR-145
ABCA1
250kDa
37kDa
GAPDH
E
3.5
% Cholesterol Efflux
to Apo A-I
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
Control
miR33a Inh
miR145 Inh
p<0.0125
0.5
10
D
p=0.002
2
1.5
30
0
3
% Cholesterol Efflux
to ApoA-I
B
ABCA1 Protein Quantification
GAPDH
p=0.006
3
p=0.005
p=0.002
2.5
2
1.5
1
+28%
+29%
+28%
+29%
+31%
+31%
Figure 3. MiR-145 and miR-33a do not
regulate ATP-binding cassette transporter A1 (ABCA1) additively or synergistically. A, Western blot analysis of
ABCA1 in HepG2 cells transfected with
control miR (100 nmol/L), miR-33a (50
nmol/L)+control miR (50 nmol/L), miR145 (50 nmol/L)+control miR (50 nmol/L),
or miR-33a (50 nmol/L)+miR-145 (50
nmol/L) for 48 hours. The loading control
is GAPDH. B, Protein quantification of
Western blot in Figure 3E using BioRad
Quantity1 (n=10). C, Cholesterol efflux
activity in HepG2 cells transfected with
control miR (100 nmol/L), miR-33a (50
nmol/L)+control miR (50 nmol/L), miR-145
(50 nmol/L)+control miR (50 nmol/L), or
miR-33a (50 nmol/L)+miR-145 (50 nmol/L)
for 48 hours (n=8). D, Western blot analysis of ABCA1 in HepG2 cells transfected
with inhibitor control miR (200 nmol/L),
anti–miR-33a (100 nmol/L)+inhibitor
control miR (100 nmol/L), anti–miR-145
(100 nmol/L)+inhibitor control miR
(100 nmol/L), or anti–miR-33a (100
nmol/L)+anti–miR-145 (100 nmol/L) for 48
hours. The loading control is GAPDH.
E, Cholesterol efflux activity in HepG2
cells transfected with inhibitor control miR
(200 nmol/L), anti–miR-33a (100 nmol/
L)+inhibitor control miR (100 nmol/L),
anti–miR-145 (100 nmol/L)+inhibitor
control miR (100 nmol/L), or anti–miR-33a
(100 nmol/L)+anti–miR-145 (100 nmol/L)
for 48 hours. Error bars represent SD.
Significance is set at P≤0.0125 (n=10).
ApoA-I indicates apolipoprotein A-I.
0.5
0
Inh
Cont
miR-33a Inh miR-145 Inh miR-33a Inh+
+Inh Cont
+Inh Cont
miR-145 Inh
miR-33a (Figure 4D and 4E). Our results identify miR-145
as a novel regulator of ABCA1 protein expression in islets.
MiR-145 Regulates Total Cholesterol Content and
Insulin Secretion in Murine Pancreatic Islets
We next investigated the effects of miR-145 on islet cell
function. Overexpressing miR-145 resulted in a significant increase to total islet cholesterol content (Figure 5A).
However, changes in cholesterol were not observed after
miR-145 inhibition (Figure 5A). This is similar to what we
observed after miR-33a inhibition in islets in a previous
study.31 In both cases, we speculate that inhibiting miR-145
or miR-33a results in localized reductions in cholesterol levels; however, these localized decreases are not detected in our
total islet cholesterol measurements. In islets from mice lacking ABCA1 in the β-cell (ABCA1 β-cell−/−), no changes were
observed in total cholesterol after miR-145 inhibition (Figure
5B). However, in a mouse model of hypercholesterolemia and
reduced islet ABCA1 expression (ApoE−/−),39 primary islets
treated with miR-145 inhibitors showed a significant reduction in total cholesterol content (Figure 5B). MiR-145 also
influenced islet function, as adenoviral overexpression of miR145 significantly decreased glucose-stimulated (20 mmol/L)
insulin secretion (Figure 5C), whereas adenoviral inhibition
of miR-145 significantly increased glucose-stimulated (20
mmol/L) insulin secretion (Figure 5D). These findings support a critical role for the regulation of insulin secretion and
islet function by miR-145.
Expression of miR-145 Is Regulated
by Glucose Levels
We next investigated the regulation of miR-145 in the context
of islet β-cell function. Previous studies demonstrated that
miR-145 expression is induced after serum starvation in human
colorectal carcinoma-116 and -8 cell lines.40 We verified this in
cultured pancreatic β-cells where we found that reduced serum
levels in cell culture media increased miR-145 expression levels by 38% in MIN6 cells (Figure IX in the online-only Data
Supplement). We next investigated whether glucose levels
could affect miR-145 expression. High glucose levels in cell
culture media (25 mmol/L) resulted in a 35% decrease in miR145 levels in MIN6 cells (Figure 6A). The effect of glucose
Kang et al Regulation of ABCA1 by miR-145 2729
A
miR-145 and miR-33a expression
levels (fmole)
0.07
Liver
0.06
Islet
0.05
Min-6
0.04
0.03
0.02
0.01
0
miR-145
Cont miR145
C
miR145
Cont Inh
ABCA1
250kDa
37kDa
37kDa
ABCA1
GAPDH
ABCA1 Protein Quantification
250kDa
E
miR-145 Inh
Control
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
D
miR-33a Inh
GAPDH
ABCA1 Protein Quantification
(fold over Cont)
B
miR-33a
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
p=0.038
p=0.009
Cont miR145
70
60
Cont miR145
Inh
p=0.008
p=0.013
miR-33a
Inhibitor
miR-145
Inhibitor
50
40
30
Figure 4. MiR-145 and miR-33a are expressed
equally in murine liver and islets, and both microRNAs regulate ATP-binding cassette transporter A1
(ABCA1) islet expression to similar levels. A, Absolute quantification of endogenous miR-33a and
miR-145 levels using the standard curve method
(reverse transcription-quantitative polymerase chain
reaction [RT-qPCR]) was measured in murine liver,
pancreatic islets, and MIN6 pancreatic β-cells.
Error bars represent SD. B, Western blot analysis of
ABCA1 in murine pancreatic islets transduced with
adenovirus control (Cont), adenovirus miR-145, and
adenovirus miR-145 inhibitor (Inh) at a multiplicity
of infection (MOI) of 100 for 48 hours. The loading control is GAPDH. C, Protein quantification of
Western blots represented in Figure 4B using NIH
ImageJ software (n=6). D, Western blot analysis of
ABCA1 in murine pancreatic islets transduced with
adenovirus control, adenovirus miR-33a inhibitor
(Inh), and adenovirus miR-145 inhibitor (Inh) at a
multiplicity of infection (MOI) of 100 for 48 hours.
The loading control is GAPDH. E, Protein quantification of Western blots represented in Figure 4B
using BioRad Quantity1 software (n=7). Data are
expressed as the mean±SD with the number of
experiments performed indicated in parentheses.
Significance is set at P≤0.017.
20
10
0
Control
on miR-33a was not as significant as the effects of glucose on
miR-145 levels (Figure 6B). This finding suggests that miR145 has a role in the glucose homeostatic pathway.
Discussion
Post-transcriptional regulation of protein expression occurs
when miRNAs bind sites in the 3′UTR to inhibit translation
or initiate mRNA degradation. ABCA1 contains a longer than
average 3′UTR with multiple predicted and validated binding sites for miRNAs. In this study, we identify miR-145 as a
novel regulator of ABCA1 in hepatic and pancreatic cell lines,
and in pancreatic islets.
By combining the results from 5 bioinformatics programs
and identifying those miRNAs expressed in the liver, 8 potential
miRNA regulators of ABCA1 were found. This included several
miRNAs recently validated as regulating ABCA1 expression
and function.17–19,22–24 A luciferase reporter assay demonstrated
that 5 of the 8 miRNAs decreased ABCA1 activity including
miR-145. The features of the primary miR-145 binding site
(site 1, 109–115; Figure II in the online-only Data Supplement)
support a regulatory role for ABCA1 expression. These features
include a location rich in AU nucleotides, being situated away
from the middle of UTRs where occlusive interactions can
form, being ≥15 nucleotides away from the stop codon where
interference from bound ribosomes arises, and a location rich
with binding sites for a variety of different miRNAs.34,41
In HepG2 cells, a cell-line relevant for HDL biogenesis,
miR-145 was found to regulate ABCA1 expression and
function. In contrast, miR-143 and miR-145* did not affect
ABCA1 activity, which demonstrates the specific regulation
of ABCA1 by miR-145. MiR-143 was previously thought to
be cotranscribed with miR-145 as part of a cardiovascularspecific miRNA cluster.35 MiR-145* is the complementary
sequence (passenger strand) to miR-145 in the miRNA/
miRNA* duplex that is generated when the RNAseIII enzyme
Dicer processes pre-miRNAs.42 The addition of exogenous
miR-145* may even inhibit miR-145 regulation of ABCA1
by binding endogenous miR-145 and inhibiting its ability to
target and reduce ABCA1 expression and function.
Our studies demonstrated that miR-33a mimics decreased
ABCA1 protein levels and efflux activity more than miR145 mimics in HepG2 cells. Their effect may be explained
by a unique feature of the miR-33a binding sites found in
ABCA1. Multiple binding sites increase the likelihood that
a miRNA regulates a target mRNA. Although miR-145 has
2 binding sites in the 3′UTR of ABCA1 at position 109 to
115 and 3114 to 3120 (TargetScanHuman 5.2), miR-33a has
3 binding sites at positions 134 to 140, 139 to 145, and 149 to
155 (TargetScanHuman 5.2) downstream of the ABCA1 stop
codon. Binding sites can produce a synergistic effect when
neighboring one another compared with 2 distant sites.34
Although miR-33a has 3 binding sites within 20 bp of one
2730 Arterioscler Thromb Vasc Biol December 2013
35
35
30
30
25
25
20
20
15
15
10
10
5
5
0
Cont miR145
C
0
WT Mouse
Islets
p=NS
60
40
30
20
p=0.003
3
2
1
Cont miR145
A
10
0
Cont miR145
Inh
Cont miR145
Inh
4.5
4
3.5
3
p=0.02
2.5
2
-35%
1.5
1
0.5
D
p=0.01
6
5
4
3
2
2.8mM
25mM
Glucose
B
4.5
4
p=NS
3.5
3
2.5
-15%
2
1.5
1
0.5
1
0
0
Cont miR145
Inh
WT Mouse
Islets
7
4
0
p=NS
50
Glucose-Stimulated Insulin Secretion
(fold over basal)
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
Glucose-Stimulated Insulin Secretion
(fold over basal)
5
p=0.04
70
7
6
ApoE -/Mouse Islets
miR-145 expression
(Normalized to U6)x1000
40
p=0.01
ABCA1 β-cell -/Mouse Islets
miR-33a expression
(Normalized to U6)x100
Islet Cholesterol (µg/mg protein)
40
Islet Cholesterol (µg/mg protein)
B
A
0
Cont miR145
Inh
Figure 5. MiR-145 regulates total cholesterol levels and glucosestimulated insulin secretion in murine pancreatic islets. A, Total
cholesterol levels in primary islets transduced with adenovirus
control (Cont) and adenovirus miR-145 or adenovirus control
(Cont) and adenovirus miR-145 inhibitor at a MOI=100 for 48
hours. All cholesterol values are normalized to protein concentration (n=5). B, Total cholesterol levels in pancreatic islets isolated
from mice with β-cell–specific deletion of ATP-binding cassette
transporter A1 (ABCA1; ABCA1 β-cell−/−), and mice with a global
deletion of apolipoprotein E (ApoE−/−) transduced with adenovirus
control (Cont) and adenovirus miR-145 inhibitor at a MOI=100
for 48 hours. All cholesterol values are normalized to protein
concentration (n=4–6). C, Glucose-stimulated (20 mmol/L) insulin
secretion in pancreatic islets isolated from wild-type mice after
transduction with adenovirus control (Cont) and adenovirus miR145 at a MOI=100 for 48 hours. All insulin secretion values are
normalized to total islet DNA concentration (n=6). D, Glucosestimulated (20 mmol/L) insulin secretion in primary pancreatic
islet cells isolated from wild-type mice transduced with adenovirus control (Cont) and adenovirus miR-145 inhibitor at a MOI=100
for 48 hours. All insulin secretion values are normalized to total
islet cell DNA concentration (n=7). All data are expressed as the
mean±SD with the number of experiments performed indicated in
parentheses. Significance is set at P≤0.05.
another, miR-145 has 2 binding sites separated by 3000 bp
(Figure II in the online-only Data Supplement).
Most biological processes are thought to involve multiple
members of a miRNA family, unrelated miRNAs, or a combination of miRNAs and transcription factors.36 A slight additive
effect between miR-758 and miR-33a on ABCA1 regulation
has been reported previously.20 We investigated the possibility that miR-145 and miR-33a could cooperate to enhance the
regulation of ABCA1 in HepG2 cells. No significant difference in ABCA1 protein levels or function was observed after
the combined expression of both miRNAs compared with
either miRNA expressed alone. The lack of a combined effect
by miR-145 and miR-33a may be because of an upper threshold in the regulation of transcripts by miRNAs or mechanisms
that limit the effect of miRNAs on a gene.
MiR-145 is expressed in several different tissues and cell
lines,35,43–45 including in murine pancreatic islets (Figure 4A).
2.8mM
25mM
Glucose
Figure 6. Glucose regulates miR-145 in MIN6 cells. Relative
expression levels (reverse transcription-quantitative polymerase
chain reaction) of (A) miR-145 and (B) miR-33a in MIN6 pancreatic islet cells grown in medium containing 2.8 or 25 mmol/L
glucose after 48 hours. All expression levels are normalized to U6
RNA (n=3). Significance is set at P≤0.05.
ABCA1 plays a key role in maintaining islet cholesterol
homeostasis and ensuring proper β-cell function,12 and we
previously demonstrated that miR-33a can affect islet cholesterol homeostasis and mediate β-cell insulin secretion through
the regulation of ABCA1.31 In this study, we reveal a similar
regulatory function for miR-145, demonstrating that in primary islets miR-145 regulates ABCA1 expression and insulin secretion. Inhibiting miR-145 in islets resulted in similar
increases to ABCA1 protein expression as inhibiting miR33a. Our studies demonstrate inhibiting miR-145 as a novel
strategy to increase ABCA1 expression and function in islets.
The physiological role of miR-145 regulation of ABCA1
may be a response to the presence of nutrients. Human
colorectal carcinoma cells grown without serum show a significant elevation of miR-145 expression.40 We found a similar increase in miR-145 expression in MIN6 β-cells grown
in media containing low serum. MiR-145 plays a major role
in the growth and differentiation of cells, and studies have
validated several transcription factors and cell cycle genes as
targets of miR-145. Increased levels of miR-145 decreases
cell growth, induces cell cycle arrest, increases apoptosis,
and decreases cell migration.43,45 Because of our demonstration that miR-145 has a regulatory role over islet function,
we investigated the effects of glucose on miR-145 expression
levels in cultured pancreatic β-cells. We observed a decrease
in miR-145 expression when MIN6 cells were exposed to
high glucose levels in the media after 48 hours. We propose a
mechanism where increased glucose levels leads to decreased
miR-145 levels in β-cells. This decreased expression of miR145 can then increase the expression of miR-145 targeted
genes involved in insulin secretion such as ABCA1.
Kang et al Regulation of ABCA1 by miR-145 2731
These studies identify miR-145 as a novel post-transcriptional regulator of ABCA1 in both hepatic and pancreatic
cells. ABCA1 in the liver is a critical determinant of plasma
HDL levels. Future in vivo studies will be required to assess
the impact of miR-145 on plasma HDL levels and atherosclerosis. Our studies also show that miR-145 is expressed
in islets and that miR-145 regulates glucose-stimulated
insulin secretion. Glucose levels also regulate miR-145
expression in β-cells. In conclusion, this study identifies
inhibiting miR-145 as a novel strategy to increase ABCA1
expression, cholesterol efflux activity, and islet function.
Acknowledgments
We thank Dr Angel Baldan and Ryan Allen for their helpful
discussions.
Sources of Funding
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
This work was supported by grants from the Canadian Institutes of
Health Research (MOP-106684 and MOP-84437 to M.R. Hayden).
N. Wijesekara was supported by postdoctoral fellowships from
the Canadian Institutes of Health Research and the Michael Smith
Foundation of Health Research (MSFHR). W. de Haan was supported
by postdoctoral fellowships from the Canadian Diabetes Association
and the MSFHR.
Disclosures
M.R. Hayden is a University Killam Professor and holds a Canada
Research Chair in Human Genetics. The other authors report no
conflicts.
References
1.Brooks-Wilson A, Marcil M, Clee SM, et al. Mutations in ABC1 in
Tangier disease and familial high-density lipoprotein deficiency. Nat
Genet. 1999;22:336–345.
2.Bodzioch M, Orsó E, Klucken J, et al. The gene encoding ATPbinding cassette transporter 1 is mutated in Tangier disease. Nat Genet.
1999;22:347–351.
3. Rust S, Rosier M, Funke H, Real J, Amoura Z, Piette JC, Deleuze JF,
Brewer HB, Duverger N, Denèfle P, Assmann G. Tangier disease is caused
by mutations in the gene encoding ATP-binding cassette transporter 1. Nat
Genet. 1999;22:352–355.
4. Wellington CL, Walker EK, Suarez A, Kwok A, Bissada N, Singaraja R,
Yang YZ, Zhang LH, James E, Wilson JE, Francone O, McManus BM,
Hayden MR. ABCA1 mRNA and protein distribution patterns predict multiple different roles and levels of regulation. Lab Invest. 2002;82:273–283.
5. Kielar D, Dietmaier W, Langmann T, Aslanidis C, Probst M, Naruszewicz
M, Schmitz G. Rapid quantification of human ABCA1 mRNA in various
cell types and tissues by real-time reverse transcription-PCR. Clin Chem.
2001;47:2089–2097.
6. Oram JF. Molecular basis of cholesterol homeostasis: lessons from Tangier
disease and ABCA1. Trends Mol Med. 2002;8:168–173.
7. Yokoyama S. ABCA1 and biogenesis of HDL. J Atheroscler Thromb.
2006;13:1–15.
8. Timmins JM, Lee JY, Boudyguina E, Kluckman KD, Brunham LR, Mulya
A, Gebre AK, Coutinho JM, Colvin PL, Smith TL, Hayden MR, Maeda N,
Parks JS. Targeted inactivation of hepatic Abca1 causes profound hypoalphalipoproteinemia and kidney hypercatabolism of apoA-I. J Clin Invest.
2005;115:1333–1342.
9. van Eck M, Bos IS, Kaminski WE, Orsó E, Rothe G, Twisk J, Böttcher A,
Van Amersfoort ES, Christiansen-Weber TA, Fung-Leung WP, Van Berkel
TJ, Schmitz G. Leukocyte ABCA1 controls susceptibility to atherosclerosis and macrophage recruitment into tissues. Proc Natl Acad Sci U S A.
2002;99:6298–6303.
10.Zhu X, Lee JY, Timmins JM, Brown JM, Boudyguina E, Mulya A,
Gebre AK, Willingham MC, Hiltbold EM, Mishra N, Maeda N, Parks
JS. Increased cellular free cholesterol in macrophage-specific Abca1
knock-out mice enhances pro-inflammatory response of macrophages. J
Biol Chem. 2008;283:22930–22941.
11. Karasinska JM, Rinninger F, Lütjohann D, Ruddle P, Franciosi S, Kruit
JK, Singaraja RR, Hirsch-Reinshagen V, Fan J, Brunham LR, Bissada N,
Ramakrishnan R, Wellington CL, Parks JS, Hayden MR. Specific loss of
brain ABCA1 increases brain cholesterol uptake and influences neuronal
structure and function. J Neurosci. 2009;29:3579–3589.
12. Brunham LR, Kruit JK, Pape TD, Timmins JM, Reuwer AQ, Vasanji Z,
Marsh BJ, Rodrigues B, Johnson JD, Parks JS, Verchere CB, Hayden MR.
Beta-cell ABCA1 influences insulin secretion, glucose homeostasis and
response to thiazolidinedione treatment. Nat Med. 2007;13:340–347.
13. Vergeer M, Brunham LR, Koetsveld J, Kruit JK, Verchere CB, Kastelein
JJ, Hayden MR, Stroes ES. Carriers of loss-of-function mutations
in ABCA1 display pancreatic beta-cell dysfunction. Diabetes Care.
2010;33:869–874.
14. Koseki M, Matsuyama A, Nakatani K, et al. Impaired insulin secretion
in four Tangier disease patients with ABCA1 mutations. J Atheroscler
Thromb. 2009;16:292–296.
15.Guay SP, Brisson D, Munger J, Lamarche B, Gaudet D, Bouchard L.
ABCA1 gene promoter DNA methylation is associated with HDL particle profile and coronary artery disease in familial hypercholesterolemia.
Epigenetics. 2012;7:464–472.
16. Duffy D, Rader DJ. Update on strategies to increase HDL quantity and
function. Nat Rev Cardiol. 2009;6:455–463.
17. Najafi-Shoushtari SH, Kristo F, Li Y, Shioda T, Cohen DE, Gerszten RE,
Näär AM. MicroRNA-33 and the SREBP host genes cooperate to control
cholesterol homeostasis. Science. 2010;328:1566–1569.
18. Rayner KJ, Suárez Y, Dávalos A, Parathath S, Fitzgerald ML, Tamehiro N,
Fisher EA, Moore KJ, Fernández-Hernando C. MiR-33 contributes to the
regulation of cholesterol homeostasis. Science. 2010;328:1570–1573.
19.Marquart TJ, Allen RM, Ory DS, Baldán A. miR-33 links SREBP-2
induction to repression of sterol transporters. Proc Natl Acad Sci U S A.
2010;107:12228–12232.
20.Ramirez CM, Dávalos A, Goedeke L, Salerno AG, Warrier N, CireraSalinas D, Suárez Y, Fernández-Hernando C. MicroRNA-758 regulates cholesterol efflux through posttranscriptional repression of
ATP-binding cassette transporter A1. Arterioscler Thromb Vasc Biol.
2011;31:2707–2714.
21. Kim J, Yoon H, Ramírez CM, Lee SM, Hoe HS, Fernández-Hernando C,
Kim J. MiR-106b impairs cholesterol efflux and increases Aβ levels by
repressing ABCA1 expression. Exp Neurol. 2012;235:476–483.
22.de Aguiar Vallim T, Tarling E, Kim T, Civelek M, Baldan A, Esau C,
Edwards P. MicroRNA-144 regulates hepatic ABCA1 and plasma
HDL following activation of the nuclear receptor FXR. Circ Res.
2013;112:1602–1612.
23. Ramírez CM, Rotllan N, Vlassov AV, et al. Control of cholesterol metabolism and plasma high-density lipoprotein levels by microRNA-144. Circ
Res. 2013;112:1592–1601.
24. Shirasaki T, Honda M, Shimakami T, Horii R, Yamashita T, Sakai Y, Sakai
A, Okada H, Watanabe R, Murakami S, Yi M, Lemon SM, Kaneko S.
MicroRNA-27a Regulates Lipid Metabolism and Inhibits Hepatitis C
Virus Replication in Human Hepatoma Cells. J Virol. 2013;87:5270–5286.
25.Kang MH, Singaraja R, Hayden MR. Adenosine-triphosphate-binding
cassette transporter-1 trafficking and function. Trends Cardiovasc Med.
2010;20:41–49.
26. Adeli K. Translational control mechanisms in metabolic regulation: critical role of RNA binding proteins, microRNAs, and cytoplasmic RNA
granules. Am J Physiol Endocrinol Metab. 2011;301:E1051–E1064.
27. Andreassi C, Riccio A. To localize or not to localize: mRNA fate is in
3’UTR ends. Trends Cell Biol. 2009;19:465–474.
28.Yang E, van Nimwegen E, Zavolan M, Rajewsky N, Schroeder M,
Magnasco M, Darnell JE Jr. Decay rates of human mRNAs: correlation
with functional characteristics and sequence attributes. Genome Res.
2003;13:1863–1872.
29. Sharova LV, Sharov AA, Nedorezov T, Piao Y, Shaik N, Ko MS. Database
for mRNA half-life of 19 977 genes obtained by DNA microarray analysis
of pluripotent and differentiating mouse embryonic stem cells. DNA Res.
2009;16:45–58.
30. Landgraf P, Rusu M, Sheridan R, et al. A mammalian microRNA expression
atlas based on small RNA library sequencing. Cell. 2007;129:1401–1414.
31.Wijesekara N, Zhang LH, Kang MH, Abraham T, Bhattacharjee A,
Warnock GL, Verchere CB, Hayden MR. miR-33a modulates ABCA1
expression, cholesterol accumulation, and insulin secretion in pancreatic
islets. Diabetes. 2012;61:653–658.
2732 Arterioscler Thromb Vasc Biol December 2013
32. Alexiou P, Maragkakis M, Papadopoulos GL, Reczko M, Hatzigeorgiou
AG. Lost in translation: an assessment and perspective for computational
microRNA target identification. Bioinformatics. 2009;25:3049–3055.
33. Gommans WM, Berezikov E. Controlling miRNA regulation in disease.
Methods Mol Biol. 2012;822:1–18.
34. Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell.
2009;136:215–233.
35. Cordes KR, Sheehy NT, White MP, Berry EC, Morton SU, Muth AN, Lee
TH, Miano JM, Ivey KN, Srivastava D. miR-145 and miR-143 regulate
smooth muscle cell fate and plasticity. Nature. 2009;460:705–710.
36.Small EM, Olson EN. Pervasive roles of microRNAs in cardiovascular
biology. Nature. 2011;469:336–342.
37. Rottiers V, Näär AM. MicroRNAs in metabolism and metabolic disorders.
Nat Rev Mol Cell Biol. 2012;13:239–250.
38.Brown BD, Naldini L. Exploiting and antagonizing microRNA regulation for therapeutic and experimental applications. Nat Rev Genet.
2009;10:578–585.
39. Kruit JK, Kremer PH, Dai L, Tang R, Ruddle P, de Haan W, Brunham LR,
Verchere CB, Hayden MR. Cholesterol efflux via ATP-binding cassette
transporter A1 (ABCA1) and cholesterol uptake via the LDL receptor
influences cholesterol-induced impairment of beta cell function in mice.
Diabetologia. 2010;53:1110–1119.
40. Sachdeva M, Zhu S, Wu F, Wu H, Walia V, Kumar S, Elble R, Watabe K,
Mo YY. p53 represses c-Myc through induction of the tumor suppressor
miR-145. Proc Natl Acad Sci U S A. 2009;106:3207–3212.
41. Pasquinelli AE. MicroRNAs and their targets: recognition, regulation and
an emerging reciprocal relationship. Nat Rev Genet. 2012;13:271–282.
42. Berezikov E. Evolution of microRNA diversity and regulation in animals.
Nat Rev Genet. 2011;12:846–860.
43. Sachdeva M, Mo YY. miR-145-mediated suppression of cell growth, invasion and metastasis. Am J Transl Res. 2010;2:170–180.
44. La Rocca G, Shi B, Sepp-Lorenzino L, Baserga R. Expression of microRNA-145 is regulated by a highly conserved genomic sequence 3’ to the
pre-miR. J Cell Physiol. 2011;226:602–607.
45.Law PT, Ching AK, Chan AW, Wong QW, Wong CK, To KF, Wong
N. MiR-145 modulates multiple components of the insulin-like
growth factor pathway in hepatocellular carcinoma. Carcinogenesis.
2012;33:1134–1141.
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
Significance
ABCA1 is an ATP-transporter protein that functions in the efflux of cholesterol from several different tissues. In the liver, ABCA1 cholesterol
efflux to apolipoprotein A-1 generates high-density lipoprotein or good cholesterol. In the pancreas, ABCA1 maintains cholesterol homeostasis and proper insulin secretion function in β-cells. This work identifies miR-145 as a novel microRNA regulator of ABCA1 in both hepatic and
pancreatic cells. Although transcriptional regulators can significantly increase ABCA1 levels, they also increase the expression of deleterious
genes. Thus, there has been a focus on identifying novel post-transcriptional regulators of ABCA1. We demonstrate that inhibiting miR-145 in
hepatic cells increases ABCA1 protein expression and cholesterol efflux function. The inhibition of miR-145 in islets increases ABCA1 protein
expression and glucose-stimulated insulin secretion. Furthermore, we demonstrate that glucose regulates miR-145 expression. Inhibiting
miR-145 represents a novel strategy for treating cardiovascular disease and diabetes mellitus.
Downloaded from http://atvb.ahajournals.org/ by guest on June 18, 2017
Regulation of ABCA1 Protein Expression and Function in Hepatic and Pancreatic Islet
Cells by miR-145
Martin H. Kang, Lin-Hua Zhang, Nadeeja Wijesekara, Willeke de Haan, Stefanie Butland,
Alpana Bhattacharjee and Michael R. Hayden
Arterioscler Thromb Vasc Biol. 2013;33:2724-2732; originally published online October 17,
2013;
doi: 10.1161/ATVBAHA.113.302004
Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272
Greenville Avenue, Dallas, TX 75231
Copyright © 2013 American Heart Association, Inc. All rights reserved.
Print ISSN: 1079-5642. Online ISSN: 1524-4636
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://atvb.ahajournals.org/content/33/12/2724
Data Supplement (unedited) at:
http://atvb.ahajournals.org/content/suppl/2013/10/17/ATVBAHA.113.302004.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the
Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of the Web
page under Services. Further information about this process is available in the Permissions and Rights
Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online
at:
http://atvb.ahajournals.org//subscriptions/
Materials and Methods
Cell Culture and Islet Isolation
HepG2 (ATCC), MIN6, and U343 cells (a kind gift from Dr Burton Yang) were
maintained in DMEM (Gibco) media containing 25mM Glucose/10% FBS/4mM LGlutamine. All cell lines were maintained at 37°C with a humidified 5% CO2
atmosphere. Islets were isolated from two to three month old C57BL6/J male mice and
cultured as previously described (1). All experiments were approved by the animal care
committee at the University of British Columbia.
3’ UTR Luciferase Reporter Generation and Measurements
The 3’ untranslated region (3’UTR) of human ABCA1 (length: 3309 nucleotides,
GenBank accession number NM_005502, position 17099-10412) was PCR amplified
from human genomic DNA and cloned into the dual-luciferase expression vector
pmirGLO (containing both Firefly and Renilla luciferases) at the XbaI and SalI restriction
sites. All constructs were confirmed by sequencing. ABCA1 3’UTR-luciferase plasmid
(0.5µg) was transfected into the U343 cell-line using Lipofectamine (Invitrogen) for 2448 hrs. Dual luciferase (Firefly luciferase and Renilla luciferase as an internal control)
protein/light was measured according to the manufacturer’s instructions (Promega).
Firefly luciferase activity was normalized to Renilla luciferase activity and expressed as
relative light units (RLU). Measurements were plotted as the relative fold change in
luciferase activity between the ABCA1 3’UTR construct compared to empty vector
alone. Site-directed mutagenesis of the miRNA binding sites was performed using the
Quick Change Mutagenesis kit according to the manufacturer’s instructions
(Stratagene).
MiRNA Mimic and miRNA Inhibitor Transfections
HepG2 cells were transfected with 50nM of Pre-miR™ miRNA precursor
(Ambion) mimics, or 100nM of Anti-miR™ miRNA (Ambion) inhibitors using
Lipofectamine RNAiMAX for 48 hrs (Invitrogen). Control miRNA oligonucleotides
(Ambion) were simultaneously transfected in each experiment.
Adenovirus Generation
Adenoviral expression vectors (pHM10) containing miRNA precursor mimics or
miRNA inhibitors were generated downstream of a CMV promoter (Applied Biological
Materials and System Biosciences Inc). MiRNA-expressing adenoviruses were
designed with a single copy of murine miR-33a or miR-145 sequence along with 150
bps of respective upstream sequence, and 150bps of respective downstream sequence.
The inhibitor sequences to miR-33a and miR-145 (System Biosciences Inc), were subcloned into pHM10 vectors. All viruses were generated and amplified in HEK293A cells,
and then purified by double cesium chloride gradient centrifugation (Applied Biological
Materials).
Adenovirus Transduction
HepG2 and MIN6 cell-lines, and murine islets were transduced at a multiplicity of
infection (MOI) of 100 in DMEM/10%FBS/4mM L-Glutamine, or RPMI
1640/10%FBS/antibiotics respectively. After 24 hrs the media was changed to fresh
1
DMEM/10%FBS/4mM L-Glutamine or RPMI 1640/10%FBS/antibiotics. After another 24
hrs (total = 48 hrs), the cells or islets were harvested and processed for experiments. All
treatments were compared to a control virus expressing the pHM10 vector alone, or
adenovirus expressing alkaline phosphatase (AP).
Protein Isolation
HepG2 cells, MIN6 cells, and murine islets were lysed in lysis buffer (10mM Tris
pH8.0, 1% Triton X-100, Complete Protease Inhibitors [Roche]) for 30 mins on ice while
vortexing every 10 mins. Five to one hundred micrograms of protein were loaded into
separate lanes and separated on 7.5% acrylamide gels. Protein was transferred onto
polyvinylidene difluoride (PVDF) membranes (Millipore).
Immunoblotting and Antibodies
ABCA1 was immunoblotted using an anti-ABCA1 monoclonal antibody generated
to its C-terminus (2). Anti-GAPDH primary antibody (Millipore), and anti-mouse-HRP
conjugated secondary antibody (Jackson IR Laboratories) were also used in
immunoblotting analyses.
Cholesterol Efflux Assays
HepG2 cells were plated to near confluency in 24 well dishes in DMEM/10%
FBS/4mM L-Glutamine. Cells were either transiently transfected with oligonucleotides or
3
transduced with adenovirus. After 24 hrs, 1µCi/ml of [ H] cholesterol (Perkin-Elmer)
3
were added to the cells. Sixteen hours following loading with [ H] cholesterol, cells were
equilibrated in DMEM (no supplements) for 1 hr, followed by efflux to 10µg/ml of apo A-I
(Lee Biosystems) for 4 hrs. Radioactivity was measured from the collected supernatant,
and from cells lysed with 0.1N NaOH. Both the supernatant and cell lysates were diluted
and measured in high flash-point scintillation cocktail fluid (Perkin-Elmer).
Relative and Absolute Quantification of miRNA Expression Levels
MiRNAs were isolated from cell lines or tissue samples using a MicroRNA
Isolation kit (Qiagen). Reverse transcription – quantitative polymerase chain reaction
(RT-qPCR) was performed with a TaqMan™ MicroRNA Assay kit (ABI) to detect miR145 or miR-33a expression levels. Briefly, relative quantification of miRNA expression
levels was performed with primers specific for mature miRNAs to generate cDNA by
reverse transcription. This was followed by cDNA amplification using Real Time PCR
with TaqMan™ probes specific for miR-145 or miR-33a. All miRNA expression levels
were normalized to the expression of internal U6 RNA.
The comparison of miR-145 to miR-33a expression levels in murine tissues was
measured by absolute quantification using the Standard Curve Method. Briefly, RT
products (cDNA) from miRNA precursor (Ambion) mimics to miR-33a or miR-145 were
serially diluted and qPCR was performed to generate the standard curves. Expression
levels of miR-145 and miR-33a in RNA samples isolated from murine liver and islets
were simultaneously measured. Actual miRNA expression levels of miR-145 and miR33a were determined following correction for primer efficiencies.
2
Cholesterol Measurements
Islet cellular cholesterol levels were obtained and measured as previously
described (3). Total protein was measured using a protein assay kit according to the
manufacturer’s instructions (Bio-Rad).
Insulin Secretion
Insulin secretion measurements from islet cells were carried out as previously
described (3). Islets were stimulated with 20mmol/l of glucose.
Statistical Analysis
Statistical significance was measured using the student’s t-test and corrected for
using the Bonferroni method when there were more than two treatment groups.
Significance is set at p≤0.05 for two comparisons, p≤0.017 for three comparisons, and
p≤0.0125 for four comparisons.
3
References
1. Kruit JK, Kremer PH, Dai L, Tang R, Ruddle P, de Haan W, Brunham LR,
Verchere CB, Hayden MR. Cholesterol efflux via ATP-binding cassette
transporter A1 (ABCA1) and cholesterol uptake via the LDL receptor influences
cholesterol-induced impairment of beta cell function in mice. Diabetologia.
2010;53:1110-1119.
2. Wellington CL, Walker EK, Suarez A, Kwok A, Bissada N, Singaraja R, Yang YZ,
Zhang LH, James E, Wilson JE, Francone O, McManus BM, Hayden MR.
ABCA1 mRNA and protein distribution patterns predict multiple different roles
and levels of regulation. Lab Invest. 2002;82:273-823.
3. Wijesekara N, Zhang LH, Kang MH, Abraham T, Bhattacharjee A, Warnock GL,
Verchere CB, Hayden MR. miR-33a modulates ABCA1 expression, cholesterol
accumulation, and insulin secretion in pancreatic islets. Diabetes. 2012;61:653658.
4
Supplemental Material
Supplemental Figure I
Luciferase activity (% of Vector)
1.2
Min 6 mouse
pancreatic
beta cells
3T3L1 mouse
adipocytes
Human Embryonic
Kidney 293 cells
Human skin
fibroblasts
Human liver
HepG2 cells
1
p<0.01
p<0.01
0.8
-25%
0.6
p<0.01
0.4
-55%
0.2
-32%
p<0.01
-69%
p<0.01
-74%
0
Supplemental Figure I. The 3’ untranslated region (3’UTR) of ABCA1 may contain
binding sites for microRNAs (miRNAs). The human ABCA1 3’UTR (3309bp) was PCR
amplified from genomic DNA and cloned into a luciferase reporter plasmid. This reporter
construct was transiently transfected into MIN6 beta cells, 3T3L1 adipocytes, 293 human
embryonic kidney cells, primary human skin fibroblasts, and HepG2 hepatocytes. Following
transfection, dual-luciferase (fire-fly luciferase, and renilla luciferase as an internal control)
activity was measured. The relative fold change in luciferase activity for each cell line was
plotted compared to pmirGLO plasmid lacking the ABCA1 3’UTR. The error bars represent
standard deviation from measurements made in triplicate.
I
miRNA Prediction Program
(website address)
Number of miRNA binding sites predicted to reside in the
3’UTR of ABCA1 or the number of miRNAs predicted to
target ABCA1
TargetScan Human 5.2
(www.targetscan.org)
40 miRNA family target sites broadly conserved among
vertebrates
PicTar
(pictar.mdc-berlin.de)
37 miRNAs predicted to target ABCA1 in vertebrates based on
conservation in mammals
microRNA.org
(www.microrna.org)
76 miRNAs predicted to target human ABCA1
miRDB
(mirdb.org/miRDB/)
115 miRNAs predicted to target human ABCA1
FindTar
(bio.sz.tsinghua.edu.cn/)
2889 predicted miRNA target sites in ABCA1. Many targets are
redundant for a single miRNA. (miR33a predicted with 2 binding
sites; miR-145 predicted with 4 binding sites; miR-17 predicted
with 9 binding sites)
Supplemental Table I. The number of predicted microRNA (miRNA) binding sites in the
3’UTR of ABCA1, or the number of miRNAs predicted to target ABCA1 using five different
bioinformatic programs: TargetScan Human 5.2; PicTar; microRNA.org; miRDB; and
FindTar.
II
TargetScan
5.2
PicTar
microRNA.org
miRDB
FindTar
miR-17
1 conserved
site
2 sites
1 site
Not
predicted
9 sites
miR-145
2 conserved
sites
1 site
1 site
1 site
4 sites
miR-19a
1 conserved
site; 2 poorly
conserved
sites
3 sites
1 site
2 sites
4 sites
miR-96
1 conserved
site
1 site
1 site
1 site
5 sites
miR-33a/
33b
2 conserved
sites; 1
poorly
conserved
site
Not
predicted
2 sites
3 sites
2 sites
miR-27a
1 conserved
site; 1 poorly
conserved
site
2 sites
2 sites
2 sites
4 sites
miR-148
1 conserved
site
1 site
1 site
1 site
3 sites
miR-144
2 conserved
sites; 3
poorly
conserved
sites
5 sites
4 sites
4 sites
3 sites
Supplemental Table II. Predicted number of miRNA binding sites in the 3’UTR of ABCA1
for eight different miRNAs. A total of eight different miRNAs (miR-17; miR-145; miR-19a;
miR-96; miR-33a/33b; miR-27a; miR-148; miR-144) were identified as potential regulators of
ABCA1 by at least four out of the five programs. All eight miRNAs have been previously
shown to be expressed in the liver or in hepatic cells. The number of binding sites predicted in
the 3’UTR of ABCA1 for each of the eight different miRNAs are listed for each program in the
Table above.
III Supplemental Figure II
Human ABCA1 3’UTR (Length = 3309 base pairs)
ABCA1
ORF
1K
2K
3K
miR-27a
miR-145
109-115
miR-144
115-121
miR-33a
2592-2598
miR-144
2686-2692
miR-145
AAAAAAA
3114-3120
miR-148
3111-3117
134-140
miR-33a
139-145
miR-33a
149-155
Supplemental Figure II. The predicted binding sites (TargetScanHuman 5.2) for the five
miRNAs that repressed 3’UTR activity in our luciferase assay. ORF = open reading frame;
AAAAAAA = poly adenylation tail.
IV
Supplemental Figure III
% Cholesterol Efflux to ApoA-I
3 p=0.0173
p=0.0013
2.5 p=0.064
2 1.5 1 0.5 0 Control
miR-33a* + Control miR-145* + Control
miR-33a* +
miR-145*
Supplemental Figure III. The expression of the miRNA* strand (miR-33a* or miR-145*)
does not repress ABCA1 cholesterol efflux activity in HepG2 cells. Cholesterol efflux
activity to 10µg/ml ApoA-I in HepG2 cells transfected with control miR(100nM),
miR-33a*(50nM) + control miR(50nM), miR-145*(50nM) + control miR(50nM), or
miR-33a*(50nM) + miR-145*(50nM) for 48 hrs. Error bars represent standard deviation.
Significance is set at p≤0.0125. (n=6).
V
Supplemental Figure IV
A.
Control
miR-17
miR-33a
ABCA1
250kDa
37kDa
GAPDH
% Cholesterol Efflux to ApoA-I
B.
3.5
p=0.5 (NS)
3
2.5
p=0.0104
2
1.5
1
0.5
0
Control
miR-17
miR-33a
Supplemental Figure IV. The expression of miR-17 does not repress ABCA1 protein expression
or (A) Western blot analysis of ABCA1 in HepG2 cells transfected with 50nM of control miR,
miR-17, or miR-33a for 48 hrs. The loading control is GAPDH. (B) Cholesterol efflux activity in
HepG2 cells transfected with 50nM of control miR, miR-17, or miR-33a for 48 hrs. (n=6). All data are
expressed as the mean ± standard deviation with the number of experiments performed indicated in
parentheses. Significance is set at p≤0.017 for three comparisons or p≤0.0125 for four comparisons.
VI
Supplemental Figure V
37kDa
Ad-anti-miR145
Ad-anti-miR33a
No Treatment
Ad-miR145
Ad-miR33a
Ad-control
250kDa
B.
HepG2 cells
HepG2 cells
ABCA1
GAPDH
% Cholesterol Efflux to ApoA-I
A.
7
6
5
4
p<0.01
3
p<0.01
p<0.01
2
1
0
Ad-AP
Ad-miR-33a
Ad-miR-145 Ad-miR-33a +
Ad-miR-145
Supplemental Figure V. Adenovirus-mediated expression of miR-145 results in similar
repressive effects on HepG2 ABCA1 protein expression and cholesterol efflux activity as seen
using oligonucleotide precursor miRNA mimics in HepG2 cells. (A) Western blot analysis of
ABCA1 in HepG2 cells transduced with adenovirus control, adenovirus miR-33a, adenovirus
miR-145, adenovirus anti-miR-33a, and adenovirus anti-miR-145 at a MOI=100 for 48 hrs. (B)
Cholesterol efflux activity to 10µg/ml Apo A-1 in HepG2 cells transduced with adenovirus alkaline
phosphatase (control), adenovirus miR-33a, adenovirus miR-145, and adenovirus miR-33a+miR-145
in HepG2 cells at a MOI=200 for 48 hrs. Error bars represent standard deviation. Statistical
significance is set at p≤0.0125. (n=10).
VII
Supplemental Figure VI
Relative miR-145 expression levels
in human tissues
A.
18
16
14
12
10
8
6
4
2
0
B.
Relative miR-145 expression levels
in mouse tissues
18
16
14
12
10
8
6
4
2
0
Supplemental Figure VI. Mature miR-145 is expressed ubiquitously in human and
murine tissues. (A) Expression of miR-145 in human tissues using reverse transcription–
quantitative polymerase chain reaction (RT-qPCR; TaqMan™) analysis. All results are
normalized to U6 control RNA. (B) Expression of miR-145 in murine tissues using RT-qPCR
(TaqMan™ ) analysis. All results are normalized to U6 control RNA. Error bars represent
standard deviation. (Sk=Skeletal, S=Small)
VIII
Supplemental Figure VII
ABCA1
250kDa
37kDa
C.
GAPDH
Ad-miR145
Ad-miR33a
Ad-control
10µM LXR treated MIN6
ABCA1
250kDa
37kDa
GAPDH
ABCA1 Protein Quantification
Untreated
MIN6 cells
B.
10µM LXR
treated MIN6
cells
70
p=0.01
60
50
40
30
20
10
0
MIN6 ABCA1 (-LXR)
MIN6 ABCA1 (+LXR)
D.
ABCA1 Protein Quantification
A.
70 60 p=0.003
50 p<0.001
40 30 20 10 0 Ad-­‐null Ad-­‐miR-­‐33a Ad-­‐miR-­‐145 MIN6 Islet Cholesterol
(µg/mg protein)
E.
25
p<0.017
20
p<0.008
15
10
5
0
Control
miR-33a
miR-145
IX
Supplemental Figure VII. The expression of miR-145 reduces ABCA1 protein levels and
increases total cholesterol content in pancreatic MIN6 cells. (A). Western blot analysis of ABCA1
in MIN6 cells treated with and without 10µM LXR agonists (TO-901317). (B). Protein quantification
of Western blot in Supplemental Figure 5A using BioRad Quantity1™. Error bars represent standard
deviation. Significance is set at p≤0.05. (n=3). (C). Western blot analysis of ABCA1 in LXR-treated
MIN6 cells transduced with adenovirus control, adenovirus miR-33a, and adenovirus miR-145 at a
MOI=100 for 48 hrs. (D). Protein quantification of Western blot in Supplemental Figure 5C using
BioRad Quantity1™. Error bars represent standard deviation. Significance is set at p≤0.017. (n=8).
(E). Total cholesterol levels in LXR-treated MIN6 cells transduced with adenovirus control,
adenovirus miR-33a, and adenovirus miR-145 at a MOI=100 for 48 hrs. Cholesterol values are
normalized to protein. Error bars represent standard deviation. Significance is set at p≤0.017. (n=12).
X
Supplemental Figure VIII
Standards
Samples
miR-145
miR-33a
Mature miR (x10-6)
Supplemental Figure VIII. Generation of Standard Curves. The standard curves were
generated by performing reverse transcription (RT) on known concentrations of mature miR-145
or miR-33a oligonucleotides (Ambion). RT products (cDNA) were serially diluted and probebased (TaqMan) qRT-PCR analysis was performed on triplicate cDNA serial dilutions. For the
curves, the dark points represent the known concentrations of miR-145 or miR-33a. The light
points are the samples including murine liver, murine islets, and MIN6 cells. All our samples are
within the known concentration range used to generate the standard curve. For the miR-145
standard curve (top) the slope = -3.403, r2 = 0.996, and efficiency = 96.721%. For the miR-33a
standard curve (bottom) the slope = -3.591, r2 = 0.987, and efficiency = 89.864%. The actual
expression levels of miR-145 and miR-33a were calculated (Figure 4A) using these standard
curves.
XI
Supplemental Figure IX
miR-145 expression
(Normalized to U6)X1000
p<0.01
Supplemental Figure IX. Regulation of miR-145 by low serum levels in MIN6 cells. Relative
expression levels (qRT-PCR) of miR-145 in MIN6 pancreatic islet cells grown in media
containing 10% FBS or 0.2% FBS after 48 hrs. All expression levels are normalized to U6 RNA.
(n=3). Significance is set at p≤0.05.
XII