ORIGINAL ARTICLE

Metformin ameliorates the proinflammatory state

in patients with carotid artery atherosclerosis
through sirtuin 1 induction

WEI XU, YANG-YANG DENG, LIN YANG, SIJIA ZHAO, JUNHUI LIU, ZHAO ZHAO, LIJUN WANG,
PRABINDRA MAHARJAN, SHANSHAN GAO, YULING TIAN, XIAOZHEN ZHUO, YAN ZHAO,
JUAN ZHOU, ZUYI YUAN, and YUE WU
XI’AN, SHAANXI, CHINA

Metformin is a widely used classic antidiabetic drug. However, its clinical pharmaco-
logic mechanism remains poorly understood. In the present study, we investigated
the anti-inflammatory effects of metformin on circulating peripheral blood mononu-
clear cells (MNCs) of patients with carotid artery atherosclerosis (AS). A total of 42
patients with carotid artery AS were randomly assigned to metformin (500 mg twice
a day; Met; n 5 21) or placebo control (Con; n 5 21) groups. After 12 weeks of treat-
ment, plasma concentrations of high-sensitivity C-reactive protein (hs-CRP), inter-
leukin 6 (IL-6), and tumor necrosis factor a (TNF-a) significantly decreased in the
Met group compared with the Con group. In addition, treatment with metformin
significantly reduced the expression of IL-6 and TNF-a at the messenger RNA level
and attenuated nuclear factor kappa B (NF-kB) DNA binding activity in MNCs. Intrigu-
ingly, metformin did not alter the expression of NF-kB p65 subunit, but markedly
inhibited its acetylation. Furthermore, metformin significantly induced sirtuin 1
(SIRT1) expression in MNCs. Moreover, we found that metformin treatment dramati-
cally induced SIRT1 expression, blocked p65 acetylation, and inhibited NF-kB activity
and the expression of inflammatory factors in MNCs in vitro. We conclude that
metformin has a novel direct protective role to ameliorate the proinflammatory
response through SIRT1 induction, p65 acetylation reduction, NF-kB inactivation,
and inflammatory inhibition in peripheral blood MNCs of patients with carotid artery
AS. (Translational Research 2015;-:1–8)

Abbreviations: AS ¼ atherosclerosis; IL-6 ¼ interleukin 6; IMT ¼ intima-media thickness; Met ¼

metformin group; MNCs ¼ mononuclear cells; SIRT1 ¼ sirtuin 1; TNF-a ¼ tumor necrosis factor a

From the Department of Cardiovascular Medicine, First Affiliated
Hospital of the Medical School, Xi’an Jiaotong University, Xi’an,
Shaanxi, China; Cardiovascular Department of Key Laboratory of
Environment and Genes Related to Diseases, Xi’an Jiaotong
University, Ministry of Education, Xi’an, Shaanxi, China;
Department of Vascular Surgery, First Affiliated Hospital of the
Medical School, Xi’an Jiaotong University, Xi’an, Shaanxi, China.
Wei Xu and Yang-Yang Deng contributed equally to this study.
Submitted for publication January 21, 2015; revision submitted May
30, 2015; accepted for publication June 2, 2015.
Reprint requests: Yue Wu or Zuyi Yuan; e-mail: zuyiyuan@mail.xjtu.
edu.cn or imyuewu@gmail.com.
1931-5244/$ - see front matter
Ó 2015 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.trsl.2015.06.002

1

2 Xu et al
AT A GLANCE COMMENTARY
Xu W, et al.
Background
The clinical pharmacologic mechanism of metfor-
min, a classic antidiabetic drug, is still poorly
understood. Here, we investigated the anti-
inflammatory effects of metformin on circulating
peripheral blood mononuclear cells (MNCs) in pa-
tients with carotid artery atherosclerosis (AS).
Translational Significance
In the present study, we found that metformin has a
novel direct protective role to ameliorate the proin-
flammatory response through sirtuin 1 induction,
p65 acetylation reduction, NF-kB inactivation,
and inflammatory inhibition in peripheral blood
MNCs of patients with carotid artery AS.
INTRODUCTION
Nowadays, atherosclerosis (AS) is one of the leading
causes of morbidity and mortality in developing and
developed countries. Furthermore, carotid artery AS
has become a strong predictor for future stroke. It is
known that AS is a chronic low-grade inflammatory dis-
ease within the arterial wall.1,2 Previous pathologic data
have shown that lymphocytes, macrophages, and foam
cells are involved in the pervasive inflammation
occurring in atherosclerotic arterial walls.3 Activation
of nuclear factor kappa B (NF-kB), a multiprotein com-
plex of transcriptional factors, induces a variety of
proinflammatory mediators and cytokines, such as inter-
leukin 6 (IL-6) and tumor necrosis factor a (TNF-a).1,4
Metformin is a safe, inexpensive, and widely used
glucose-lowering drug for diabetes treatment in clinical
practice as well as a potential anti-inflammatory agent.5
In several randomized clinical studies, metformin has
been shown to confer atheroprotective effects by
decreasing the progression of carotid intimal medial
thickness and coronary AS.6,7 Remarkably, in the UK
Prospective Diabetes Study, patients taking metformin
demonstrated a lower risk of both microvascular and
macrovascular events, including cardiovascular death,
nonfatal myocardial infarction, and nonfatal stroke.8,9
In another study, metformin was shown to have anti-
inflammatory and antiatherogenic effects in vascular
cells in vitro and to limit atherosclerotic lesion develop-
ment in vivo10 Moreover, it has been shown in animal
models that metformin can improve endothelium func-
tion.11 Our previous data demonstrated that metformin
Translational Research
- 2015
attenuates cardiac remodeling in failing rat hearts by
ameliorating overactivated endoplasmic reticulum
stress.12 However, it remains unknown how and to
what extent metformin exerts an atheroprotective and
anti-inflammatory role in patients.
Peripheral blood mononuclear cells (MNCs) provide
a representative view of the overall inflammatory status
in the body.13,14 Proinflammatory transcription factor
NF-kB activated in MNCs induces expression of
proinflammatory genes and correlates with circulating
systemic markers of inflammation.13-15 Our previous
data have also shown that MNCs are involved in the
inflammatory process and may potentially reflect the
degree of AS and proinflammatory state.15-17 In the
present study, we will investigate whether metformin
exerts anti-inflammatory effects on circulating periph-
eral blood MNCs of patients with carotid artery AS.
We examined the plasma concentration of proinflamma-
tory mediators and their gene expression in MNCs. In
this study, we also evaluated NF-kB binding in the nu-
cleus of MNCs. Moreover, we assessed the expression
of NF-kB subunit p65 and its regulator as a measure
of NF-kB activity and inflammation.
MATERIALS AND METHODS
Patients and study design. A double-blind placebo-
controlled study was performed at the First Affiliated
Hospital of Medical College, Xi’an Jiaotong
University. A total of 45 Chinese real-world Han
patients (31 males and 14 females) with carotid artery
AS were consecutively recruited from a real-world
clinical practice. Metformin and placebo were
purchased from Beijing Shengyong Pharmaceutical
Co, Ltd (Beijing, China).
To minimize potential confounding factors, patients
with histories of stroke, recent acute coronary syn-
drome, acute inflammation, and treatment with steroids
or immunosuppressive drugs were excluded from the
study15,18 Patients contraindicated to metformin
treatment or already under metformin administration
were also excluded from the study. Three patients met
our exclusion criteria. The remaining 42 patients were
randomly assigned to the following 2 groups: placebo
control group (Con; n 5 21) and metformin group
(Met, n 5 21). All patients and physicians were
blinded to the patient assignment. Table I summarizes
the demographic data of all subjects.
Day 2 after randomization, either study medication
(500 mg twice a day) or placebo was administered to pa-
tients for 12 weeks. Fasting blood samples were collected
before administration. Patients were followed twice
(4 and 12 weeks after discharge) and fasting blood sam-
ples were collected again at the end of the 12-week study
Translational Research
Volume -, Number - Xu et al 3

Table I. Baseline characteristics of the study population

Characteristics Placebo (n 5 21) Metformin (n 5 21) P

Sex, male/female, n/n 14/7 15/6 1.00

Age, y 55.1 6 10.9 55.2 6 10.4 0.88
Systolic BP, mm Hg 128 6 18 129 6 17 0.89
Diastolic BP, mm Hg 79 6 10 76 6 10 0.55
Smoking 7 9 0.56
Disease
Hypertension 7 8 1.00
Diabetes 2 3 1.00
Stroke 5 4 1.00
Fasting glucose, mmol/L 5.28 6 0.57 5.32 6 0.37 0.86
Total cholesterol, mmol/L 3.87 6 0.65 3.71 6 0.47 0.48
Triglycerides, mmol/L 1.75 6 1.03 1.71 6 1.01 0.89
HDL cholesterol, mmol/L 0.99 6 0.16 0.95 6 0.18 0.74
LDL cholesterol, mmol/L 2.36 6 0.57 2.21 6 0.39 0.34
hs-CRP, mg/L 3.1 (0.9, 7.8) 3.5 (1.3, 9.9) 0.58
Treatment
Aspirin 12 14 0.75
b-Blocker 1 1 1.00
ACE inhibitors/ARBs 2 3 1.00
Insulin 1 2 1.00
Sulfonylureas 2 1 1.00
CCB 8 11 0.54
Statins 5 7 0.74

Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin II type 1 receptor blocker; BP, blood pressure; CCB, calcium antagonists;
HDL, high-density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein.
Data are reported as the mean 6 standard deviation, median (interquartile range), or n.

period. The study protocol was explained and all partici-
pating patients gave written informed consents. This
study was conducted in compliance with the Declaration
of Helsinki, and the research protocol was approved by
the Ethics Committee of Xi’an Jiaotong University.
Carotid AS judgment. The extent of carotid AS was
examined by ultrasound based on carotid plaque or
intima-media thickness (IMT). Ultrasonographic
scanning of the carotid artery was performed by an
ultrasonic phase-locked echo-tracking system, which
was equipped with a high-resolution real-time 10-
MHz linear scanner (HP Sonos 5500 Color Doppler
ultrasonic system). Measurements were determined by
a reader blinded to all clinical information.
Briefly, longitudinal images of the bilateral carotid
arteries, including the proximal (.10 mm proximal to
bulb bifurcation) and distal common carotid artery,
bulb as well as internal and external carotid arteries (10
segments), were acquired. Carotid IMT was determined
in 3 segments as follows: the distal common carotid
artery (1 cm proximal to the carotid bulb); the carotid
artery bifurcation (1 cm proximal to the flow divider);
and the proximal internal carotid artery (1 cm of length).
The maximum IMT was recorded in the longitudinal and
transverse projections at the site of the most advanced
atherosclerotic lesion. Mean values of all IMT measure-
ments from the right and left sides were calculated for
each patient. Plaque was defined as carotid IMT
(cIMT) of 1.5 mm or more, or focal encroachment of
0.5 mm or more into the arterial lumen7,18,19
Isolation of MNCs. Peripheral blood MNCs were
isolated by Ficoll standard density gradient centrifuga-
tion. The upper layer containing MNCs was harvested
and washed with Hanks’ balanced salt solution and
then with phosphate-buffered saline.
Plasma concentrations of proinflammatory media-
tors. Concentrations of plasma IL-6, TNF-a, and IL-4
were assayed with enzyme-linked immunosorbent
assay kits according to the manufacturers’ instructions.
The lower limits of detection (or lowest concentration
of standard sample) were 0.2 pg/mL for both IL-6 and
TNF-a and 0.4 pg/mL for IL-4. All antibodies were
purchased from Excell. High-sensitivity C-reactive
protein (hs-CRP) assays were performed in the clinical
laboratory in our hospital.
Western blot analysis and quantitative polymerase
chain reaction. Cytoplasmic protein levels of p65, sub-
units of MNCs were detected by Western blotting as pre-
viously described.12 Densitometry was performed using
the Bio-Rad molecular analyst software and all values
were corrected by loading b-actin. Total RNA isolation
and real-time quantitative reverse transcription–
polymerase chain reaction were performed as described
previously.20

4 Xu et al
NF-kB DNA binding activity. Nuclear proteins were
extracted according to the manufacturer’s instructions
(Pierce). NF-kB DNA binding activity was measured
with NF-kB transcription factor assay kit (Abcam)
according to the manufacturers’ instructions.20
Short interfering RNA transfection in MNCs. Human
short interfering-sirtuin 1 (si-SIRT1) and mock short
interfering RNA (siRNA) were obtained from Santz
Cruz Biotechnology, and transient siRNA transfection
was carried out according to the protocol of the
manufacturer as described previously.12 Briefly,
siRNAs were dissolved in double distilled water to
prepare a 10 mM of stock solution. Isolated MNCs
grown in 6-well plates were transfected with siRNA in
OptiMEM (Invitrogen) containing RNAiMax
(Invitrogen). For each transfection, 250 mL of
transfection medium containing 5 mg of siRNA was
gently mixed with 250 mL of transfection medium
containing 5 mL of transfection reagent. After
20 minutes of incubation at room temperature, the
mixture was added to cells in 2 mL of culture medium
and cultured for 24 hours before treatment.
Statistical analysis. Statistical analyses were per-
formed using Statistical Product and Service Solutions
(SPSS) for Windows (version 13.0, SPSS Inc, Chicago,
IL). Discrete variables were expressed as numbers and
percentages and analyzed using c2 test. Summary
values were expressed as the mean 6 standard error.
Skewed data were reported as medians (interquartile
range). Paired t test was used to analyze changes in
baseline concentrations. All multiple comparisons
between the Con and Met groups were carried out using
the Holm-Sidak 2-way repeated-measures analysis
of variance (TWRMANOVA) method. All P values
,0.05 were considered statistically significant.
RESULTS
Clinical data. All 42 patients fulfilled the 12-week
follow-up without any drug-related adverse effects or
cardiovascular events. Table I shows that there were no
differences between the Met and Con groups in terms of
baseline characteristics. Table II demonstrates that no
significant changes in body mass index, body weight,
systolic blood pressure, diastolic blood pressure, fasting
glucose, and insulin levels were observed after
12 weeks of treatment. Metformin significantly
decreased low-density lipoprotein cholesterol (LDL-C)
after 12 weeks of treatment, although it did not alter
levels of total cholesterol, triglycerides, and high-
density lipoprotein cholesterol (Table II).
CRP is a well-established inflammatory marker that
has been demonstrated to be increased in patients with
clinical AS. We found that placebo did not alter plasma
CRP level in the Con group. However, the plasma CRP
Translational Research
- 2015
level was dramatically decreased after 12 weeks of
metformin administration, indicating the anti-
inflammatory effect of metformin.
Metformin reduces plasma concentrations of
proinflammatory mediators. To determine whether
metformin attenuated the proinflammatory state in AS pa-
tients, we assessed plasma levels of IL-6, TNF-a, and IL-
4 before and after treatment. After 12 weeks of metformin
treatment, plasma concentrations of IL-6 and TNF-a
were decreased to 71 6 9% (P 5 0.02) and 64 6 9%
(P 5 0.03) of the baseline, respectively, whereas in the
Con group, plasma concentrations of these factors were
increased to 107 6 9% (P 5 0.66) and 109 6 11%
(P 5 0.59) of the baseline, respectively. Therefore,
placebo did not alter the levels of proinflammatory
factors. Furthermore, compared with the Con group,
plasma levels of IL-6 and TNF-a were significantly
decreased after 12 weeks of metformin administration
(TWRMANOVA, P 5 0.001 and 0.015, respectively;
Fig 1), suggesting that metformin attenuated
proinflammatory mediators in AS patients. In addition,
both metformin and placebo did not alter IL-4 level
(TWRMANOVA, P 5 0.690; Fig 1, C).
Metformin suppresses gene transcription of
proinflammatory factors. We performed quantitative
real-time polymerase chain reaction to examine
whether the effect of metformin on plasma
proinflammatory mediators was attributable to
messenger RNA (mRNA) expression of MNCs. In the
Met group, the mRNA levels of IL-6 and TNF-a were
decreased compared with the baseline (Fig 2),
whereas in the Con group, the mRNA levels of these 2
factors were not significantly changed. The mRNA
levels of IL-6 and TNF-a were significantly decreased
in the Met group compared with the Con group
(TWRWANOVA, P 5 0.001 and 0.002, respectively;
Fig 2). In addition, the mRNA level of IL-4 remained
unchanged in either the Met group or the Con group.
Metformin inhibits NF-kB DNA binding activity in
MNCs. We examined the NF-kB DNA binding activity
to explore whether metformin could affect the proin-
flammatory transcription factor NF-kB in MNCs. We
found that the NF-kB DNA binding activity was signif-
icantly decreased to 46 6 10% of the baseline after
metformin administration (P 5 0.006; Fig 2, D),
suggesting that metformin could inhibit NF-kB DNA
binding activity in vivo.
Metformin inhibits p65 acetylation in vivo. Furthermore,
we investigated the NF-kB signaling pathway in MNCs.
The protein levels of NF-kB subunit p65 and acetylated
p65 were detected. Interestingly, the p65 acetylation
was significantly blocked by metformin administration,
whereas the total protein level of p65 showed no
significant changes compared with baseline in either
Translational Research
Volume -, Number - Xu et al 5

Table II. Metabolic and other parameters at baseline and after 12 wk

Placebo controls (n 5 21) Metformin (n 5 21)

Marker 0 wk 12 wk 0 wk 12 wk

Weight, kg 65.4 6 12.7 65.6 6 11.3 66.9 6 9.1 66.3 6 8.9

Systolic BP, mm Hg 128 6 18 127 6 17 129 6 17 126 6 17
Diastolic BP, mm Hg 79 6 10 78 6 8 76 6 10 74 6 9
Fasting glucose, mmol/L 5.28 6 0.57 5.30 6 0.49 5.32 6 0.37 5.31 6 0.45
Fasting insulin, U/L 6.13 6 2.98 6.10 6 3.02 6.09 6 2.44 5.89 6 2.62
Total cholesterol, mmol/L 3.87 6 0.65 3.85 6 0.71 3.71 6 0.47 3.59 6 0.43
Triglycerides, mmol/L 1.75 6 1.03 1.68 6 0.91 1.71 6 0.61 1.73 6 0.51
HDL cholesterol, mmol/L 0.99 6 0.16 1.01 6 0.41 0.95 6 0.18 0.92 6 0.41
LDL cholesterol, mmol/L 2.36 6 0.57 2.33 6 0.43 2.21 6 0.39 1.99 6 0.30*‡
hs-CRP, mg/L 3.1 (0.9, 7.8) 2.9 (0.8, 7.2) 3.5 (1.3, 9.9) 1.3 (0.5, 4.8)†‡

Abbreviations: BP, blood pressure; HDL, high-density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein.
n 5 21 for each group, values are reported as the mean 6 standard deviation or median (interquartile range).
*P , 0.05 compared with baseline.
†
P , 0.01 compared with baseline.
‡
P , 0.01 compared with the placebo group.

A B C then treated with metformin or vehicle. Fig 3, A shows

15 P=0.001 6 P=0.015 6
* *
P=0.690 that metformin markedly induced SIRT1 and blocked
Plasma TNF-α (pg/ml)

Plasma IL-4 (pg/ml)

Plasma IL-6 (pg/ml)

10 4 4
5 2
0 0
0wk 12wk 0wk 12wk
Con Met
Fig 1. Metformin significantly reduced plasma concentrations of IL-6
and TNF-a in patients with carotid artery AS. Changes in plasma con-
centrations of proinflammatory mediators are presented as raw data.
Plasma IL-6 and TNF-a levels were significantly decreased in the
metformin group (Met) compared with the control group (Con) after
12 weeks (12 wk) of treatment (2-way repeated-measures analysis
of variance, P 5 0.001 and 0.015, respectively). *P , 0.05, compared
with baseline (0 wk). Both metformin and placebo did not alter IL-4
concentrations (2-way repeated-measures analysis of variance,
P 5 0.690). AS, atherosclerosis; IL-6, interleukin 6; TNF-a, tumor ne-
crosis factor alpha.
the Con or Met groups after 12 weeks of treatment
(Fig 2, E).
Metformin induces SIRT1 expression in vivo. As a novel
and most important deacetylase in cell, SIRT1 has
been known to play an important role in p65 acetylation.
Therefore, we further detected the level of SIRT1 in
MNCs. Fig 2, F shows that metformin, but not
placebo, significantly increased the SIRT1 expression
in MNCs, suggesting a novel mechanism of NF-kB
regulation in vivo.
Metformin inhibits NF-kB activity through SIRT1 induction
in vitro. We further verified whether the anti-
inflammatory effect of metformin was mediated
through acetyl-p65–NF-kB inactivation at the cellular
level. MNCs isolated from AS patients were
transfected with either si-SIRT1 or mock siRNA, and
p65 acetylation in isolated MNCs. SIRT1 knockdown,
but not the mock siRNA, significantly abolished
2 metformin-inhibited p65 acetylation, NF-kB
activation, and IL-6 and TNF-a induction, revealing
0
0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk that metformin-inhibited NF-kB activity through
Con Met Con Met
SIRT1 induction in vitro.
DISCUSSION
In the present study, we provided solid evidence that
metformin could lessen the circulating inflammatory
responses through the suppression of NF-kB activity
and inhibition of proinflammatory mediators in
MNCs. Furthermore, inhibition of p65 acetylation by
SIRT1 in MNCs plays an important role in metformin
treatment, which could serve a novel mechanism in clin-
ical treatment for AS patients.
Vascular inflammation is recognized as the funda-
mental mechanism for AS, and proinflammatory
mediators play pivotal roles in AS.21 IL-6 and TNF-a
are classic proinflammatory cytokines to promote AS,
whereas IL-4 is an anti-inflammatory cytokine.22 It
appears to be a promising strategy for AS treatment
by reducing circulating IL-6 and TNF-a. Our study
demonstrates that metformin administration signifi-
cantly reduced plasma concentrations and the expres-
sion of IL-6 and TNF-a at mRNA level, indicating
that metformin could directly attenuate the proinflam-
matory state of circulating MNCs. Inhibition of proin-
flammatory factors can further result in the reduction
of neointima volume in metformin treatment.23
However, neither IL-4 protein nor mRNA level was
altered by metformin. The main reason might be that
 Translational Research
6 Xu et al - 2015

A8 P=0.001
B P=0.002 C
P=0.736
8 ** 4

TNF-α mRNA (au)

**

IL-6 mRNA (au)

IL-4 mRNA (au)

6
6 3
4
4 2

2 1
2

0 0 0
0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk
Con Met Con Met Con Met
D E F
150
Veh Ac-p65 Sirt1
Met
p65 Gapdh
NF-κB activity (%)

100
6 **
** 100

Sirt1 (fold)
Ac-p65 (%)
4
50
50 **
2

0 0 0
0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk 0wk 12wk

Fig 2. Metformin markedly attenuated NF-kB–related proinflammation in vivo. (A–C) Changes in mRNA
expression of MNCs in baseline (0 wk) and after 12 weeks (12 wk) of treatment, including IL-6, TNF-a, and
IL-4. (D) Changes in NF-kB binding activity of MNCs in baseline (0 wk) and after 12 weeks (12 wk) of treatment.
(E and F) The protein level of p65, acetyl-p65, and SIRT1 in MNCs of baseline (0 wk) after treatment for 12 weeks
(12 wk). **P , 0.01, compared with baseline. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; IL-6, inter-
leukin 6; MNCs, mononuclear cells; mRNA, messenger RNA; NF-kB, nuclear factor kappa B; SIRT1, sirtuin 1;
TNF-a, tumor necrosis factor alpha.

Fig 3. SIRT1 induction is required for metformin-blocked p65 acetylation and NF-kB activation in vitro. MNCs
isolated from patients with carotid artery AS were transfected with siRNA (mock or si-SIRT1) and then treated
with metformin (Met). (A) Western blots were performed for SIRT1, acetyl-p65, and loading control (GAPDH).
(B) NF-kB binding activity in MNCs. (C) The mRNA levels of proinflammatory genes, including IL-6 and TNF-a.
*P , 0.05 compared with the Met group, #P , 0.05 compared with mock group. GAPDH, glyceraldehyde-3-phos-
phate dehydrogenase; IL-6, interleukin 6; MNCs, mononuclear cells; mRNA, messenger RNA; NF-kB, nuclear
factor kappa B; SIRT1, sirtuin 1; TNF-a, tumor necrosis factor alpha.

different cytokines may have different characteristics. cell types. The cell source of IL-4 is T cell, which might
First, IL-6 and IL-4 have different roles in the inflamma- react differently to metformin from MNCs where IL-6
tory pathway. IL-6 mainly promotes inflammation, and TNF-a are derived from mcarophage.24,25
whereas IL-4 mainly plays anti-inflammatory role. Sec- Several randomized clinical studies have confirmed
ond, the promoter of IL-4 does not contain NF-kB the benefits of metformin treatment,6,8 including
sequence, so IL-4 is not regulated by NF-kB. More cardiovascular protection, lipid control, nonalcoholic
importantly, IL-4 and IL-6 are derived from different simple fatty liver alleviation, and tumor inhibition.
Translational Research
Volume -, Number -
Metformin has exhibited several antiatherosclerotic
effects, including modulation of blood pressure and
lipid concentrations, matrix remodeling activation of
matrix proteases, and inhibition of inflammation.10,26
In the present study, we investigated the effect of
metformin on each of these mechanisms. We did not
observe a significant reduction in blood pressure,
which was in accordance with other reports and might
be attributable to the fact that the blood pressure of all
subjects was already under control by optimal
treatment. Consistent with other reports, metformin
treatment did not alter total cholesterol, high-density
lipoprotein, and triglyceride (TG), but it markedly low-
ered LDL. We also did not observe any changes in levels
of blood glucose and insulin. The possible explanation
could be that not all patients included in this study
were diabetics, and metformin did not exert significant
influence on the blood glucose of patients without dia-
betes.
NF-kB, a key nuclear factor in inflammatory
signaling, initiates the transcription of various cyto-
kines for AS pathogenesis, including TNF-a and
IL-6.27 Previous study28 has demonstrated that metfor-
min inhibits NF-kB activity; however, its exact under-
lying mechanism remains unclear. In the present study,
we found that oral administration of metformin signif-
icantly suppressed NF-kB binding activity in MNCs,
thus resulting in a decrease in proinflammatory factors.
As one major component of NF-kB, p65 has been
known for some modifications.28 When it is acetylated,
the DNA binding function is enhanced, which may
aggregate AS progression. Interestingly, we found
that acetyl-p65, instead of the total level of p65, played
an important role in regulating NF-kB binding activity
in vivo.
SIRT1 is an nicotinamide adenine dinucleotide
(NAD1)-dependent protein deacetylase. It has been
shown to suppress NF-kB signaling through deacetyla-
tion of the p65 subunit of NF-kB, leading to the reduc-
tion of the inflammatory responses mediated by this
transcription factor.27,29,30 Herein, we describe, for the
first time, metformin-mediated SIRT1 induction in
MNCs in vivo, which might be an important and novel
finding of the mechanism, by which metformin
alleviates inflammatory state.
The major limitation of this study is that we could not
fully exclude the potential effect of comedication, such
as statin, a well-established anti-inflammatory drug. As
shown in Table I, statins were also administrated to
patients in both groups, which may be a potential con-
founding factor in the present study. Furthermore, met-
formin, but not placebo, significantly decreased LDL-C
after 12 weeks of treatment as depicted in Table II.
Therefore, we could not exclude the potential contribu-
Xu et al 7
tion of the LDL-lowering effect of metformin, which re-
quires further investigation.
CONCLUSIONS
Our study demonstrated that 3 months of metformin
treatment attenuated proinflammation state in MNCs
through induction of SIRT1, blockade of NF-kB activa-
tion, and inhibition of inflammatory mediators. These
findings suggest that metformin possessed a novel,
direct atheroprotective role by modulating the circu-
lating inflammatory responses in AS patients.
ACKNOWLEDGMENTS
Conflicts of Interest: All authors have read the jour-
nal’s policy on conflicts of interest and have none to
declare.
This study was supported by the National Natural
Science Foundation of China (grant no. 81300226,
81470550, 81400302, 81400660, 91339116, and
81100209), the National Basic Research Program of
China (‘‘973 Project’’, no. 2012CB517804), and the
Clinical Innovation Funds of the First Affiliated Hospi-
tal of XJTU (grant no. 14ZD20). Dr Yuan is a recipient
of the National Natural Fund for Distinguished Young
Scholars of China (81025002).
All authors have read the journal’s authorship
agreement.
REFERENCES
1. Lind L. Circulating markers of inflammation and atherosclerosis.
Atherosclerosis 2003;169:203–14.
2. Del Rincon I, Polak JF, O’Leary DH, et al. Systemic inflammation
and cardiovascular risk factors predict rapid progression of
atherosclerosis in rheumatoid arthritis. Ann Rheum Dis 2014;
74:1118–23.
3. van Diepen JA, Berbee JF, Havekes LM, Rensen PC. Interactions
between inflammation and lipid metabolism: relevance for
efficacy of anti-inflammatory drugs in the treatment of atheroscle-
rosis. Atherosclerosis 2013;228:306–15.
4. Schreck R, Albermann K, Baeuerle PA. Nuclear factor kappa B:
an oxidative stress-responsive transcription factor of eukaryotic
cells (a review). Free Radic Res Commun 1992;17:221–37.
5. Woo SL, Xu H, Li H, et al. Metformin ameliorates hepatic steato-
sis and inflammation without altering adipose phenotype in diet-
induced obesity. PLoS One 2014;9:e91111.
6. Riksen NP, el Messaoudi S, Rongen GA. It takes more than one
CAMERA to study cardiovascular protection by metformin. Lan-
cet Diabetes Endocrinol 2014;2:105–6.
7. Preiss D, Lloyd SM, Ford I, et al. Metformin for non-diabetic
patients with coronary heart disease (the CAMERA study): a
randomised controlled trial. Lancet Diabetes Endocrinol 2014;2:
116–24.
8. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-
year follow-up of intensive glucose control in type 2 diabetes. N
Engl J Med 2008;359:1577–89.
9. Effect of intensive blood-glucose control with metformin on com-
plications in overweight patients with type 2 diabetes (UKPDS

8 Xu et al
34). UK Prospective Diabetes Study (UKPDS) Group. Lancet
1998;352:854–65.
10. Oliveira PW, de Sousa GJ, Caliman IF, et al. Metformin amelio-
rates ovariectomy-induced vascular dysfunction in non-diabetic
Wistar rats. Clin Sci 2014;127:265–75.
11. Arunachalam G, Samuel SM, Marei I, Ding H, Triggle CR.
Metformin modulates hyperglycaemia-induced endothelial senes-
cence and apoptosis through SIRT1. Br J Pharmacol 2014;171:
523–35.
12. Zhuo XZ, Wu Y, Ni YJ, et al. Isoproterenol instigates cardiomyo-
cyte apoptosis and heart failure via AMPK inactivation-mediated
endoplasmic reticulum stress. Apoptosis 2013;18:800–10.
13. Ghanim H, Aljada A, Hofmeyer D, Syed T, Mohanty P,
Dandona P. Circulating mononuclear cells in the obese are in a
proinflammatory state. Circulation 2004;110:1564–71.
14. Satoh M, Tabuchi T, Itoh T, Nakamura M. NLRP3 inflammasome
activation in coronary artery disease: results from prospective and
randomized study of treatment with atorvastatin or rosuvastatin.
Clin Sci 2014;126:233–41.
15. Wu Y, Zhang W, Liu W, Zhuo X, Zhao Z, Yuan Z. The double-
faced metabolic and inflammatory effects of standard drug
therapy in patients after percutaneous treatment with drug-
eluting stent. Atherosclerosis 2011;215:170–5.
16. Zhao Z, Wu Y, Cheng M, et al. Activation of Th17/Th1 and Th1,
but not Th17, is associated with the acute cardiac event in patients
with acute coronary syndrome. Atherosclerosis 2011;217:
518–24.
17. Gao S, Liu W, Zhuo X, et al. The activation of mTOR is required
for monocyte proinflammatory response in patients with coronary
artery disease. Clin Sci 2014;128:517–26.
18. Tan C, Liu Y, Li W, et al. Associations of matrix
metalloproteinase-9 and monocyte chemoattractant protein-1
concentrations with carotid atherosclerosis, based on measure-
ments of plaque and intima-media thickness. Atherosclerosis
2014;232:199–203.
19. Koyama H, Maeno T, Fukumoto S, et al. Platelet P-selectin
expression is associated with atherosclerotic wall thickness in
carotid artery in humans. Circulation 2003;108:524–9.
Translational Research
- 2015
20. Liu J, Zhuo XZ, Liu W, et al. Drug-eluting stent, but not bare
metal stent, accentuates the systematic inflammatory response
in patients. Cardiology 2014;128:259–65.
21. Lang JK, Cimato TR. Cholesterol and hematopoietic stem cells:
inflammatory mediators of atherosclerosis. Stem Cells Transl
Med 2014;3:549–52.
22. Popko K, Gorska E, Stelmaszczyk-Emmel A, et al. Proinflam-
matory cytokines Il-6 and TNF-alpha and the development
of inflammation in obese subjects. Eur J Med Res 2010;
15(suppl 2):120–2.
23. Lu J, Ji J, Meng H, et al. The protective effect and underly-
ing mechanism of metformin on neointima formation in
fructose-induced insulin resistant rats. Cardiovasc Diabetol
2013;12:58.
24. Bruun JM, Lihn AS, Pedersen SB, Richelsen B. Monocyte chemo-
attractant protein-1 release is higher in visceral than subcutaneous
human adipose tissue (AT): implication of macrophages resident
in the AT. J Clin Endocrinol Metab 2005;90:2282–9.
25. Hyun B, Shin S, Lee A, et al. Metformin down-regulates TNF-
alpha secretion via suppression of scavenger receptors in macro-
phages. Immune Netw 2013;13:123–32.
26. Madiraju AK, Erion DM, Rahimi Y, et al. Metformin suppresses
gluconeogenesis by inhibiting mitochondrial glycerophosphate
dehydrogenase. Nature 2014;510:542–6.
27. Yang H, Zhang W, Pan H, et al. SIRT1 activators suppress inflam-
matory responses through promotion of p65 deacetylation and
inhibition of NF-kappaB activity. PLoS One 2012;7:e46364.
28. Rothgiesser KM, Fey M, Hottiger MO. Acetylation of p65 at
lysine 314 is important for late NF-kappaB-dependent gene
expression. BMC Genomics 2010;11:22.
29. Kauppinen A, Suuronen T, Ojala J, Kaarniranta K, Salminen A.
Antagonistic crosstalk between NF-kappaB and SIRT1 in the
regulation of inflammation and metabolic disorders. Cell Signal
2013;25:1939–48.
30. Clarke NE, Belyaev ND, Lambert DW, Turner AJ. Epigenetic
regulation of angiotensin-converting enzyme 2 (ACE2) by
SIRT1 under conditions of cell energy stress. Clin Sci 2014;
126:507–16.