See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269876358

Increased resistin may suppress reactive oxygen species production and

inflammasome activation in type 2 diabetic patients with pulmonary
tuberculosis infection

Article  in  Microbes and Infection · December 2014

DOI: 10.1016/j.micinf.2014.11.009 · Source: PubMed

CITATIONS READS

16 97

8 authors, including:

Wen-Cheng Chao Chia-Liang Yen

Taichung Veterans General Hospital National Cheng Kung University
53 PUBLICATIONS   182 CITATIONS    17 PUBLICATIONS   163 CITATIONS   

SEE PROFILE SEE PROFILE

Horng-Yih Ou Chi-Chang Shieh

National Cheng Kung University National Cheng Kung University
92 PUBLICATIONS   1,020 CITATIONS    96 PUBLICATIONS   2,021 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Infection and IMID risk View project

The role of peroxisome proliferator-activated receptor delta(PPARdelta) in the action of cardiotonic agents View project

All content following this page was uploaded by Wen-Cheng Chao on 13 August 2018.

The user has requested enhancement of the downloaded file.

 + MODEL

Microbes and Infection xx (2014) 1e10

www.elsevier.com/locate/micinf

Original article

Increased resistin may suppress reactive oxygen species production and

inflammasome activation in type 2 diabetic patients with pulmonary
tuberculosis infection
Wen-Cheng Chao a,b, Chia-Liang Yen c, Ying-Hsun Wu d, Shin-Yi Chen a, Cheng-Yuan Hsieh a,
Tsung-Chain Chang e, Horng-Yih Ou f, Chi-Chang Shieh a,g,*
a
Institute of Clinical Medicine, National Cheng Kung University Medical College, Tainan, Taiwan
b
Department of Internal Medicine, Taichung Veteran General Hospital, Chiayi Branch, Chiayi, Taiwan
c
Institute of Basic Medical Sciences, National Cheng Kung University Medical College, Tainan, Taiwan
d
Chest Hospital, Department of Health, Tainan, Taiwan
e
Department of Medical Laboratory Science and Biotechnology, National Cheng Kung University Medical College, Tainan, Taiwan
f
Department of Internal Medicine, National Cheng-Kung University Hospital, College of Medicine, National Cheng-Kung University, Tainan, Taiwan
g
Department of Pediatrics, National Cheng Kung University Hospital, Tainan, Taiwan

Received 5 August 2014; accepted 27 November 2014

Abstract

Although it has been known for decades that patients with type 2 diabetes mellitus (DM) are more susceptible to severe tuberculosis (TB)
infection, the underlying immunological mechanisms remain unclear. Resistin, a protein produced by immune cells in humans, causes insulin
resistance and has been implicated in inhibiting reactive oxygen species (ROS) production in leukocytes. Recent studies suggested that IL-1b
production in patients with Mycobacteria tuberculosis infection correlates with inflammasome activation which may be regulated by ROS
production in the immune cells. By investigating the level of resistin in different patient groups, we found that serum resistin levels were
significantly higher in severe TB and DM-only groups when compared with mild TB cases and healthy controls. Moreover, elevation of serum
resistin correlated with impairment of ROS production of neutrophils in patients with both DM and TB. In human macrophages, exogenous
resistin inhibits the production of ROS which are important in the mycobacterium-induced inflammasome activation. Moreover, macrophages
with defective ROS production had poor IL-1b production and ineffective control of mycobacteria growth. Our results suggest that increased
resistin in severe TB and DM patients may suppress the mycobacterium-induced inflammasome activation through inhibiting ROS production by
leukocytes.
© 2014 Institut Pasteur. Published by Elsevier Masson SAS. All rights reserved.

Keywords: Resistin; Tuberculosis; Diabetes mellitus; ROS; Inflammasome

1. Introduction 8.6 million people developed active pulmonary TB and 1.3

Mycobacterium tuberculosis (TB) infection remains an
important disease in the modern world. In the year 2012 alone,
* Corresponding author. Institute of Clinical Medicine, National Cheng-
Kung University Medical College, 138 Sheng-Li Road, Tainan 704, Taiwan.
Tel.: þ886 6 2353535x5616; fax: þ886 6 275 3083.
E-mail address: cshieh@mail.ncku.edu.tw (C.-C. Shieh).
million died from this disease [1]. The emergence of the com-
bination of TB and type 2 diabetes mellitus (DM) has become a
new global challenge not only because of the rising prevalence
of type 2 DM in populations with higher TB infection rates, but
also due to the fact that type 2 DM patients tend to have higher
TB incidence and increased disease severity [2,3]. The immu-
nological mechanisms underlying this TB susceptibility in
diabetic patients, however, still remain unclear [4].

http://dx.doi.org/10.1016/j.micinf.2014.11.009
1286-4579/© 2014 Institut Pasteur. Published by Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
2 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10
Resistin, a 12-kDa soluble serum protein, was initially
considered an adipokine which can cause insulin resistance
and mediates the progression from obesity to type 2 DM [5].
Further investigations revealed that activated immune cells,
rather than adipocytes, are the main source of resistin in
humans [6]. Serum resistin level hence has been used as an
inflammatory biomarker in various diseases including sepsis,
rheumatoid arthritis, and atherosclerosis [7e9]. Furthermore,
resistin was recently reported to impair the chemotaxis and
production of reactive oxygen species (ROS) in neutrophils
[10].
ROS not only are key weapons used by the phagocytic
leukocytes to kill microorganisms but also serve as signaling
mediators to coordinate innate and adaptive immune re-
sponses [11]. Recent studies have shown that ROS regulate
important macrophage functions including apoptosis, auto-
phagy, and chemokine production during mycobacterial
infection and may affect the development of T cell immunity
against mycobacteria [12]. Since neutrophils, which produce
abundant ROS when stimulated, have been shown to be the
predominant phagocytic cells in the airways of patients with
active pulmonary TB [13], ROS production by leukocytes
may be quite abundant and play a central role in defense
against tuberculosis in the infected tissues. Inflammasomes
are multi-protein complexes responsible for caspase-1 acti-
vation and subsequent proteolytic processing and secretion of
interleukin-1b (IL-1b), which is a pivotal cytokine for anti-
TB immune response [14]. Based on recent studies impli-
cating the role of ROS in activating inflammasome in
mycobacterial infection [15], we postulated that resistin may
weaken the immune defense against TB infection in type 2
DM through affecting ROS production by immune cells.
Here, we analyzed the resistin level in TB patients with or
without type 2 DM and investigated its relationship with ROS
production and IL-1b secretion in defense against
mycobacteria.
2. Materials and methods
2.1. Patients
We prospectively enroll tuberculosis patients at Chest
Hospital, a TB referral center in southern Taiwan with 51
negative-pressure isolation beds. We enrolled culture-proven
pulmonary TB subjects in this study. DM is defined by gly-
cated hemoglobin (HbA1c) S6.5%, fasting blood glucose
S126 mg/dl, or a random glucose S200 mg/dL. Venous
blood of 10 ml was sampled from each subject for laboratory
analysis. The clinical data including Chest X-ray, sputum
culture results, HbAlC, and other important clinical data of the
study subjects were recorded. Our previous study and other
studies showed that DMTB patients tend to have more severe
disease severity when compared with non-DM TB patients
[2,3]. We hence used sputum acid-fast stain (AFS) grades to
classify the severity of TB, and the severity classification by
sputum AFS correlated well with the cavity formation on chest
X-ray in this study. This study was approved by the
Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
Institutional Review Board of Taichung Veteran General
Hospital (C09194).
2.2. Cytokine and prostaglandin E2 measurement
Serum resistin, CRP, IFN-g, and IL-10 of all cases were
measured with ELISA, using Duoset human resistin and
Duoset human CRP kits (R&D systems, Minneapolis, MN).
IL-1b concentrations in sera were measure by high sensitivity
ELISA kits with the detection ranged from 0.16 to 10.0 pg/mL
(eBioscience, USA). Prostaglandin E2 (PGE2) concentration
of macrophage supernatants was determined by an ELISA kit
(Cayman Chemical, USA for PGE2).
2.3. Mycobacterium marinum preparation
Given that RD1 locus plays critical role in TB virulence,
we used M. marinum (Mycobacterium marinum) in this study,
which is an RD-1 containing non-tuberculosis mycobacterium.
M. marinum, obtained from American Type Culture Collection
(ATCC), was further confirmed using chip hybridization and
16S rRNA sequencing.
2.4. Monocytic cell culture and mycobacterial
stimulation
THP-1 cells were grown in RPMI 1640 (Gibco BRL,
Gaithersburg, MD) supplemented with 10% fetal bovine
serum and 1% penicillin and streptomycin. Phorbol-myristate-
acetate (PMA, 100 nM) was added for 24 h to induce differ-
entiation of THP-1 cells into macrophages. Macrophages were
then infected with M. marinum at an MOI of 1, and cell free
supernatants were harvested, double filtered with 0.2 micron
filters, and assayed for cytokines by ELISA and Western blot.
2.5. Preparation and mycobacterial stimulation of
monocyte-derived macrophages (MDMs)
MDMs were prepared from peripheral venous blood from
healthy donors and a chronic granulomatous disease (CGD)
patient with a dinucleotide deletion (711-712 AG) in exon 7
[16]. CGD patients have defect in ROS generation from
granulocytes and are extremely susceptible to mycobacteria
infection [17]. Freshly sampled venous blood was mixed with
6% dextran for 2 h for RBCs sedimentation to get leukocyte-
enriched supernatant. The supernatants were then overlaid on
Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) and centri-
fuged at 400 g for 20 min to separate PBMC and neutrophils.
Monocytes were then isolated from PBMCs by positive se-
lection using CD14þ magnetic beads (Miltenyi Biotech). The
purity of monocytes, determined by flowcytometry, was
consistently more than 95%. The monocytes were then
cultured for 7 days with GM-CSF (2 ng/ml) for differentiation
and activated by human IFN-g (5 ng/ml) for 2 days before
infected with M. marinum at an MOI of 0.1. After being
infected for 3 days, macrophages were lyzed with 0.1%
Triton-X and the amount of intracellular bacteria was
 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10
determined by serial dilution and plating on 7H11 plates. In
this in vitro model, healthy human macrophages survived for
more than 7 days after infection, which allowed accurate
intracellular bacteria growth analysis [18].
2.6. Detection of reactive oxygen species
H2DCFDA (20 ,70 -dichlorodihydrofluorescein diacetate,
Invitrogen) was used to measure ROS production after PMA
stimulation from neutrophils isolated from patients. Briefly,
after 15 min of incubation with H2DCFDA, neutrophils were
stimulated with PMA (1 mM) for 20 min before fluorescence
measurement with flowcytometry. ROS production index was
calculated as the ratio of mean fluorescent intensity of PMA-
stimulated cells and mean fluorescent intensity of control cells.
To detect the immediate ROS production from peripheral
blood mononuclear cells (PBMCs) and macrophages after
PMA and mycobacteria stimulation, we used a luminol-
enhanced chemiluminescence method. PBMCs isolated from
healthy donors were cultured with M. marinum with and
without resistin. The luminescence for the next 20 min was
measured. Macrophages were cultured in white plates, and
resistin with or without NADPH oxidase inhibitor, dipheny-
leneiodonium chloride (DPI, SigmaeAldrich, USA), were
added. Luminol was added one hour later and the kinetic
luminescence for the next 10 min was then continually
measured and recorded with Luminoskan (Thermo).
2.7. Protein analysis and Western blotting
Macrophages were plated in six-well tissue culture dishes
and infected with M. marinum as described above. The su-
pernatants were then collected and proteins were precipitated
by methanolechloroform extraction. The cell lysates were
also collected. The lysates or supernatants were resolved with
10% SDS-PAGE and electroblotted to a polyvinylidene
difluoride (PVDF) membrane (Millipore, Bedford, MA).
Immunoblot analysis was done with anti-murine and human
caspase-1 p10 (m315; Santa Cruz Biotechnology, California).
For the detection of pro-IL1b and mature IL-1b (p17), the blot
was probed with 1:1000 rabbit anti-human IL-1b antibody
(Santa Cruz Biotechnology) and cleaved IL-1b antibody (Cell
Signaling), respectively.
2.8. Statistical analysis
Data were presented as percentages (%) for categorical
factors and as means ± standard deviations for continuous
factors. Differences between patient subgroups were tested by
Student's t test for comparison between two distinct groups
whereas One-way ANOVA test, followed by Tukey's post hoc
test were used for more than two groups. ManneWhitney U
test was used for comparison non-normally distributed data
between two distinct groups, and KruskaleWallis analysis was
used for more than two groups. Statistical significance was set
at p < 0.05, two-sided. All data were analyzed using SPSS
version 16.0 (SPSS Inc., Chicago, IL, USA).
Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
3
3. Results
3.1. Characteristics of study subject
A total of 297 subjects were enrolled. Among them, 151
were TB patients with or without type 2 DM while 71 were
diabetes-only and 75 were healthy controls (Table 1). DMTB
patients tend to have severe TB infection, TB severity classi-
fication is thus critical to avoid the selection bias when
comparing immune responses between DM patients and non-
DM patients. In this study, sputum AFS grades were used to
classify the severity of TB. Among the 151 TB patients, 49
(32%) were severe cases, defined as sputum AFS above 1þ,
while the other 102 (68%) TB patients were mild cases with
AFS below or equal to 1þ. The percentage of DM was higher
in severe TB group (41%, 20/49) than in the mild TB group
(13%, 13/102). As expected, severe TB cases had higher fre-
quency of cavitation in chest X-ray when compared with mild
TB cases (80% and 8% respectively, p < 0.05). In clinical
presentation, severe TB patients had more body weight loss
and leukocytosis, higher serum CRP, and lower serum albumin
levels. These clinical findings indicated that severe TB cases
were in a higher inflammatory state.
3.2. Serum resistin levels were elevated in patients with
severe pulmonary tuberculosis and diabetic patients
We compared the serum levels of resistin in patient groups
of mild TB with or without diabetes, severe TB with or
without diabetes, DM controls and healthy controls (Fig. 1A).
We found that serum resistin levels were significantly higher
in severe TB groups with and without DM ( p < 0.05) when
compared with mild TB cases and non-DM healthy control
groups. Furthermore, the resistin levels were also higher in
DM controls when compared with healthy controls. We further
compared DM and non-DM groups in both severe and mild
TB cases. The DM groups had higher resistin levels than their
non-DM counterpart. The difference between the DM and
non-DM groups with TB, however, did not reach statistical
significance. Since resistin elevation in DM controls can be
attributed to unidentified minor infections, we went on to
measure the well-known infection biomarker, CRP in all cases.
Serum CRP levels were similar between the DM controls and
the healthy controls. It is hence unlikely that the elevation of
resistin in DM controls was caused by unidentified infection
(Fig. 1B). These results showed that diabetic patients tended to
have higher resistin levels and severe TB cases also had higher
resistin levels in comparison with healthy controls.
3.3. Resistin negatively correlated with ROS production
in DMTB patients and inhibited ROS production in
human phagocytes
To investigate the correlation of ROS production ability and
serum resistin in TB patients, we used neutrophils isolated
from TB patients with DM (n ¼ 13), TB patients without DM
(n ¼ 21), DM controls (n ¼ 11), and healthy controls (n ¼ 20)
4 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10

Table 1
Characteristics and laboratory data of patients and controls.a
Characteristic Tuberculosis patients Control subjects p value
Mild TB b
Severe TB c
DM controls (n ¼ 71) Healthy controls (n ¼ 75)
(n ¼ 102) (n ¼ 49)
Age (years) 54 ± 23 59 ± 19 61 ± 11 58 ± 13 NS
Male % 57% (58/102) 74% (36/49) 66% (47/71) 56% (42/75) NS
BW (kgs) 54.8 ± 9.4 53.9 ± 12.7 68.1 ± 13.6 66.1 ± 13.1 <0.05d
DM/non-DM 13% (13/102) 41% (20/49) NA NA <0.05e
DM HbAlC(%) 8.9 ± 2.6 10.2 ± 2.9 8.8 ± 2.0 NA
DM HbA1C (mmol/mol) 74.1 ± 28.7 88.3 ± 32.1 72.5 ± 21.8 NA
Clinical symptoms
Cough % 63% (64/102) 98% (48/49) NA NA <0.05e
Body weight loss % 12% (12/102) 45% (22/49) NA NA <0.05e
Fever % 11% (11/102) 22% (11/49) NA NA NS
Haemoptysis % 3% (3/102) 8% (4/49) NA NA NS
Laboratory data
White blood cell (103/ml) 7521 ± 3281 8211 ± 2489 6775 ± 1825 6768 ± 1702 <0.05d
Hemoglobin (g/dl) 12.9 ± 2.0 12.4 ± 2.1 13.8 ± 1.6 13.9 ± 1.6 <0.05d
Albumin (mg/dl) 3.9 ± 0.6 3.7 ± 0.7 4.4 ± 0.4 4.3 ± 0.3 <0.05d
CXR cavitation % 8% (8/102) 80% (39/49) NA NA <0.05e
Sputum acid-fast stain titer
Negative 76 0 NA NA
1þ 26 0 NA NA
2þ 0 11 NA NA
3þ 0 15 NA NA
4þ 0 23 NA NA
a
Data represent mean ± SD. BW, body weight; DM, diabetes mellitus; NA, not applicable; NS, non-significant; TB, tuberculosis; C, controls; S, severe TB; M,
mild TB.
b
Positive tuberculosis culture with acid fast stain negative or 1þ.
c
Positive tuberculosis culture with acid fast stain 2þ, 3þ or 4þ.
d
Difference between two control groups and two tuberculosis patients.
e
Difference between Mild TB and severe TB.

to measure their PMA-induced ROS production. We found
that neutrophils isolated from DMTB patient had lower ROS
production (Fig. 2A). Moreover, the subjects who have higher
resistin serum levels tend to have lower ROS production.
Pearson's correlation analysis showed significant negative
correlation between serum resistin concentration and neutro-
phil ROS production (r ¼ 0.5, p ¼ 0.03) (Fig. 2B). We then
went on to demonstrate the inhibitory effect of resistin on ROS
production in human phagocytes. We found that resistin
inhibited both PMA-induced ROS production in human neu-
trophils (Fig. 2C) and M. marinum-induced ROS production in
human PBMCs (Fig. 2D). These data suggested that elevation
of serum resistin levels may suppress ROS production of
neutrophils in DM and severe TB patients.
difference between serum IFN-g and IL-10 levels in DM or
non-DM TB patients and in groups with different severities
(Fig. 3B and C). We further tested the IL-1b production by
LPS-treated PBMCs from DM-only groups and healthy con-
trols. Similar to the results from serum samples, we found that
IL-1b production was lower in LPS-treated PBMCs from DM
subjects when compared with that of healthy controls
(Fig. 3D). We hence concluded that immune-mechanisms
leading to decreased IL-1b levels may contribute to the im-
munodeficiency in patients with severe TB with DM.
3.5. Resistin suppressed ROS production which is
critical for controlling intracellular mycobacterial
growth in human macrophages

3.4. Decreased IL-1b level in severe TB patients
with DM
Macrophages with different phenotypes have been reported
to secrete different cytokines and play a key role in the change
of immune responses in metabolic syndrome and type 2 DM
[19]. We hence measured key cytokines including IL-1b, IFN-
g, and IL-10, in serum samples from TB patients with and
without DM and healthy controls. We found that IL-1b level in
DM patients with severe TB was lower than the level in non-
DM severe TB patients (Fig. 3A). There was no significant
Previous studies have showed the anti-oxidant effect of
resistin on neutrophils through PI3K signaling pathway [10].
We hence went on to examine whether resistin can directly
inhibit the ROS production by macrophages. PMA and M.
marinum, the closest genetic relative of M. tb, were used to
infect and stimulate macrophages derived from THP-1 cells in
the presence or absence of resistin [20]. We found that the
presence of resistin inhibited ROS production of THP-1 cells
by 40% after PMA and M. marinum stimulation (Fig. 4A and
Fig. 4B). The suppressive activity of resistin for ROS pro-
duction was close to that of the NADPH oxidase inhibitor,

Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10
Fig. 1. Serum resistin levels were significantly higher in severe TB patients
and diabetes-only group, while serum C-reactive protein levels were not
elevated in diabetes-only group. Serum resistin (A) and CRP (B) levels among
DM and non-DM TB patients with different severities and different control
cases. For box-and-whisker plots, the box outline represented the 25th and
75th percentiles, the central line represented the mean value, and the whiskers
represent minimum and maximum values. One-way ANOVA test, followed by
Tukey's post hoc test were used to assess differences between resistin in
different patient subgroups. For the non-normally distributed data in CRP,
KruskaleWallis analysis was used to assess differences between patient sub-
groups. TB, tuberculosis; DM, diabetes mellitus; HbAlC, Hemoglobin A1c;
ns, not significant; *p < 0.05; **p < 0.005.
diphenyleneiodonium chloride (DPI). To determine the anti-
mycobacterial role of ROS in macrophages, we compared the
intracellular mycobacterial growth between monocyte-derived
macrophages (MDM) of health controls with or without
resistin pretreatment, and MDM from a CGD patient. We
found that intracellular mycobacterial growth strikingly
increased in macrophages from CGD patient when compared
with health controls infected by M. marinum ( p ¼ 0.002).
Moreover, the intracellular mycobacterial growth in MDM
from normal subjects also tended to increase after resistin
treatment ( p ¼ 0.069) (Fig. 4C). Taken together, these data
showed the inhibitory effect of resistin on ROS production and
the critical antimycobacterial role of ROS in human
macrophages.
Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
5
3.6. Decreased ROS may lead to lower inflammasome
activation and IL-1b production in macrophages
To explore the connection between ROS and inflammasome
activation, we measured the caspase-1 activation and the
downstream production of IL-1b after M. marinum stimulation
in macrophages. Caspase-1 is synthesized as 40 kDa precursor
(p40) that is cleaved into 20 kDa (p20) and 10 kDa (p10)
mature proteins to form an active caspase-1 enzyme. Hence,
the appearance of p20 and p10 reflexes caspase-1 activation.
We found increased expression of p10 and processing of IL-1b
after M. marinum stimulation, and both p10 and IL-1b pro-
cessing were suppressed by resistin and DPI (Fig 5A and B).
To further examine the effect of ROS on the activation of
Nod-like receptor pyrin domain e containing-3 (NLRP3), a
key inflammasome activator in mycobacterial infection, MDM
of healthy controls and CGD patient were treated with PBS
and ESAT-6, a potent mycobacterium tuberculosis protein to
activate NLRP3 inflammasome [21]. Lower IL-1b production
was also found in CGD patients when compared with health
controls (Fig 5C). These data strongly suggest that resistin
inhibits M. marinum-induced ROS production and the down-
stream inflammasome activation and IL-1b production in
human macrophages and leads to weakened defense to
mycobacterial infection on both ROS and IL-1b levels.
4. Discussion
In this investigation aimed to unravel the role of resistin in
immune defense against mycobacteria in type 2 DM patients,
we found that serum resistin levels are higher in severe TB
patients and in diabetic patients. The results of our experi-
ments suggest that resistin may suppress the ROS production
and hence compromise the mycobacterium-induced inflam-
masome activation in leukocytes. These data revealed a novel
mechanistic pathway through which type 2 DM patients
become susceptible to severe TB infection.
Adipokines have been postulated to be among the main
factors mediating the linkage between adipocytes and
abnormal metabolism in conditions including obesity and type
2 DM. Human resistin, however, is not only an adipokine
leading to insulin resistance, but also an inflammatory medi-
ator. Recent studies showed that resistin is an inflammatory
marker for the development of atherosclerosis in diabetic pa-
tients [7]. These findings strongly suggested that resistin may
be one of the linkages between type 2 DM and chronic
inflammation. Our finding that serum resistin levels were
higher in both diabetic patients and severe TB cases than in
non-DM healthy controls and mild TB cases confirmed the
dual role of resistin as an adipokine and an inflammatory
marker. However, robust inflammatory responses do not al-
ways lead to good immune defense against microorganisms in
patients with metabolic diseases, including obesity and type 2
DM. On the contrary, leukocytes from patients with type 2 DM
have been known to have weakened cellular response against
infections by some prevalent pathogens including mycobac-
teria and Burkholderia pseudomallei [22]. The weakened
6 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10

Fig. 2. Resistin level negatively correlated with ROS production in DMTB patients and in vitro treatment with resistin inhibited ROS production in human
phagocytes. (A) Neutrophils from TB, DMTB, DM patients, and healthy donors were labeled with H2DCFDA (4 mM) and stimulated by PMA (1 mM). Then, ROS
production was measured with the fluorescence intensity by flowcytometry, and presented as ROS production index. DMTB patients have significantly lower ROS
production in their neutrophils compared with TB and healthy controls. (B) Correlations between serum resistin and ROS production index were calculated
respectively by Pearson's correlation coefficients (r). (C) Neutrophils from healthy donors (n ¼ 6) were treated with the same conditions in (A) with and without
resistin (100 ng/ml) and the ROS production was measured with the fluorescence intensity by flowcytometry. (D) PBMCs from healthy donors (n ¼ 4) were
infected by M. marinum (MOI: 10) with and without resistin and the ROS production was determined by chemiluminescence. Data represent mean ± s.d.
KruskaleWallis analysis. *p < 0.05.

initial response and then dysregulated hyper-inflammation
may result in different features clinically. In previous
studies, DMTB patients were reported to have higher disease
severity than non-DM TB patients, suggesting a dysregulated
hyper-inflammation state in the DMTB patient group [23,24].
In this study, we compared DMTB and non-DM TB with the
same TB severity, and aimed to pinpoint the DM-related
defect in response to TB infection. In our results, we identi-
fied a defect in IL-1b production, which negatively correlates
with the elevated resistin level, as the likely basis for the
paradox of dysregulated hyper-inflammatory state and poor
mycobacterial defense in DMTB.
IL-1b has been shown to be important in defense against
TB as IL-1 receptor-deficient mice showed greatly increased
mortality to M. tb infection [25]. NLRP3 inflammasome
activation has been known to be the key regulator of IL-1b
secretion in mycobacterial infection. In our in vitro experi-
ments, we used M. marinum to study the mycobacteria-
induced inflammasome activation. Although not a common
pathogen in the human lung, M. marinum, which secrets
ESAT-6 like M. tb, induces cutaneous suppurative granulo-
matous inflammation [26,27] with quite similar histological
characteristics to granulomas induced by M. tb in the lung and
has been increasingly used in mechanistic studies of
mycobacteria-induced early inflammation [28e30].
Consistent with our results in M. marinum (shown in Fig
3A and B), Chen et al. found that ROS production plays a
role in Mycobacterium kansasii-induced inflammasome acti-
vation [15]. In addition, Cohen et al. reported that resistin
inhibits the ROS production of neutrophils by 40% [10]. Our
results hence further proved the pivotal role of NADPH
oxidase-produced ROS in mycobacterium-induced inflamma-
some activation in macrophages. Importantly, as resistin is
secreted by inflammatory cells [6], the concentration of
resistin in the lung can be significantly higher than the resistin
level in the serum. Therefore, the relatively small differences
of serum resistin levels in different patients groups are likely
to reflect bigger differences in tissue resistin levels and hence
ROS production and the downstream IL-1b production.
Inflammasome can be activated by both metabolic stress
and various microbial infections [31,32]. However, in patients
with type 2 DM, inflammasome activation state induced by
chronic metabolic stress and different microbial infections can
vary in different clinical situations. Our data revealed the
decreased IL-1b production in diabetic tuberculosis patients
(Fig. 2A). However, Lee, et al. showed the up-regulation of

Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10 7

Fig. 3. Decreased IL-1b level in severe TB patients with type 2 DM. (A) Serum IL-1b levels were measured using high sensitivity ELISA kit. Scatter plot was used
and each dot meant a single data while central line represented mean value. (B) and (C) Serum IL-10 IFN-g levels among DMTB or non-DM TB patients with
different severities were measured using luminex. (D) IL-1b production from PBMCs in DM subjects and healthy control after TLR4 stimulation for 18 h were
measure with ELISA. For box-and-whisker plots, the box outline represented the 25th and 75th percentiles, the central line represented the mean value, and the
whiskers represent minimum and maximum values. IL-1b, interleukin-1b; IFN-g, interferon-g; IL-10, interleukin-10; TB, tuberculosis; DM, diabetes mellitus;
PBMCs, peripheral blood mononuclear cells; TLR, toll-like receptor; *p < 0.05; **p < 0.001; n.s., not significant.

NLRP3 inflammasome activation and IL-1b synthesis in LPS-
stimulated macrophages in patients with type 2 diabetes [33].
The differences in inflammasome activation and IL-1b pro-
duction are likely due to the different stimulations used to
activate the leukocytes. Our data using M. marinum- and
ESAT-6-induced activation of caspase-1 and IL-1b secretion
might be more relevant conditions to show the upstream role
of ROS in mycobacteria-activated inflammasome in type 2
DM.
Innate immune responses have been found to be important
both in controlling the initial infection and in promoting
adaptive responses that mediate host resistance or immuno-
pathology in the course of TB [34,35]. A recent study showed
that neutrophils kill the internalized mycobacteria through the
NADPH oxidase-dependent mechanisms [30]. The decreased
IL-1b production after ESAT-6 stimulation in monocyte-
derived macrophages from our patient with CGD (Fig. 5C)
provides direct evidence for the regulatory role of ROS in
mycobacterium-induced inflammasome activation. We also
found that intracellular mycobacterial growth strikingly
increased in macrophages from the CGD patient in compari-
son with that in healthy controls after infected by M. marinum.
The intracellular mycobacterial growth in macrophages also
tended to increase after resistin treatment, although to a lesser
extent. This may be due the fact that resistin only inhibited
ROS production by 40% (see Fig. 4A and B), not like the
nearly complete deficiency of ROS in cells from CGD pa-
tients. Rather than acting solely as bactericidal effectors, ROS
have been known to be key signaling molecules to co-ordinate
the antimycobacterial host defense [36]. We showed in this
study that resistin inhibits ROS production in phagocytes and
affects IL-1b processing (Fig. 5A and B). These data showed
the key role of ROS in mycobacterial infection. Interestingly, a
recent study further revealed that IL-1b directly enhanced the
antimycobacterial ability of macrophages from M. tb-infected
humans and mice through the augmentation of TNF-a
signaling [14]. The elevated resistin levels hence may weaken
innate defense against TB through both directly ROS-
mediated and IL-1b-mediated anti-mycobacterial
mechanisms.
The compromised ROS and IL-1b production in type 2 DM
patients may also affect the granulomatous inflammation in the
later stage of mycobacterial infection. Our data (Fig. 2A) and a
recent study identified the defect in ROS generation from both
neutrophils and PBMCs in type 2 DM patients [22]. In M. tb-
infected animal models, oxidative stress was found to be

Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
8 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10

important for granulomatous inflammation, which is weakened

by hyperglycemia in guinea pigs [37,38]. Furthermore,
delayed adaptive immune response to TB was recently iden-
tified in an M. tb infected diabetic mice model [39]. The
interaction between cytokine networks and cell death patterns
in innate response is critical to orchestrate the induction of
adaptive immunity against mycobacteria. Innate cytokines
networks, including the important mediators of IL-1b and
TNF-a, are believed to regulate adaptive immunity against
mycobacteria through modulating eicosanoid pathway [35].
Recent studies showed that prostaglandin E2-mediated
apoptotic death of macrophage induces effective priming and
activation of adaptive immunity in TB infection [40]. We also
found that while M. marinum infection induced PGE2 pro-
duction in macrophages, resistin suppressed the increased
PGE2 production to an uninfected level (Supplementary
Fig. 1). The ability of resistin to suppress PGE2 production
in M. marinum-infected macrophages thus implicate that ROS
may also be involved in the linkage between innate and
adaptive immunity in mycobacterial infection. The suppres-
sion of ROS production by elevated resistin in patients with
type 2 DM hence may contribute to the high susceptibility of
TB in DM through weakening adaptive immune responses.
The IL-1b production by human leukocytes after resistin
treatment has been previously investigated with different re-
sults. Bokarewa et al. used resistin to stimulate human PBMCs
and showed that TNF-a and IL-6 was both strongly up-
regulated, but IL-1b remained unchanged [41]. Given the
recent understanding that pro-IL-1b may be processed by
proteases other than caspase-1 in infections and other sterile
inflammatory diseases [42], we believe the different relation-
ship between resistin and IL-1b production may result from
the different activation states of inflammasome and other
proteases in vivo. Our findings suggest that mycobacteria
activate IL-1b mainly through an ROS- and inflammasome-
dependent pathway and are particularly susceptible to the
regulatory effect of resistin in type 2 DM. The role of resistin
in other infections and inflammatory conditions hence should
be further investigated in humans in the future.
Similar to our findings, elevated resistin levels in TB pa-
tients with severe disease or with cachexia have been reported
by other study groups [43,44]. The anti-oxidant effects of
resistin in leukocytes revealed in this study hence provide a
mechanistic explanation for the compromised innate immune

Fig. 4. Resistin suppressed ROS production which is critical for controlling

intracellular mycobacterial growth in human macrophages. Differentiated
THP-1 cells (1  106) were left unstimulated, stimulated by PMA (1 mM) (A)
or M. marinum at MOI:1 (B) in the absence or presence of resistin with
different concentrations and DPI (10 mM), 10 min later, luminol-amplified
chemiluminescence was used to measure ROS production. Data represent
mean ± s.d., n ¼ 6. (C) Monocytes-derived macrophages (8  104) from 6
healthy controls with or without resistin (100 ng/ml) treatment and one CGD
patient were infected with M. marinum at MOI:1, and intracellular bacterial
loads were determined 72 h after infection. The experiments were analyzed
with KruskaleWallis analysis and repeated 3 times with similar results.
*p < 0.05; **p < 0.005.

Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10 9

In conclusion, we found that resistin, which is elevated in

diabetic patients and severe pulmonary TB patients, inhibits
ROS production by NADPH oxidase and may suppress the
inflammasome activity in human macrophages. These findings
provide a novel mechanism for the weakened immune defense
against mycobacteria in diabetic patients and suggest that
therapeutic approaches to lower the level of resistin may
reverse the immunocompromized state in type 2 DM patients.

Conflict of interest

All authors declare no financial or commercial conflict of

interest.

Acknowledgments

Authors thank Dr. Shen, Gwan-Han at Taichung Veteran

General Hospital for kindly providing M. marinum. The
research project was approved by the Institutional Review
Board of Taichung Veteran General Hospital (C09194), and
was supported by Grant 10024 from Central and Southern
Region Hospital Alliance, Department of Health, Taiwan. CCS
and WCC designed the study; WCC, CLY, CYS and SYC
performed the experiments; WCC, YHW, HYO performed
patient study; CCS and WCC wrote the paper. Total word
count is 3776.

Appendix A. Supplementary data

Supplementary data related to this article can be found at

http://dx.doi.org/10.1016/j.micinf.2014.11.009.

References

Fig. 5. Resistin-suppressed ROS production reduced mycobacterium-induced
inflammasome activation. (A) Differentiated THP-1 cells were left unin-
fected (control) or pretreated by PBS, resistin (100 ng/ml) and DPI (10 mM)
then infected for 4 h with M. marinum at MOI:1, 18 h later, pro-IL-1b (p31),
mature-IL-1b (p17), and active caspase-1 (Casp1 p10) were analyzed by
immunoblot. (B) IL-1b release was determined by ELISA. (C) Monocyte-
derived macrophages (8  104) from 4 healthy controls and one CGD pa-
tient were treated with PBS and early secreted antigenic target protein 6
(ESAT-6) (20 mg/mL), a potent mycobacterium tuberculosis protein to activate
inflammasome, and IL1-b in the supernatants were measured 24 h after
stimulation. Data represent mean ± s.d. The experiments were analyzed with
KruskaleWallis analysis and repeated 3 times with similar results. *p < 0.05;
**p < 0.005.
responses in TB patients. Resistin inhibitors are being devel-
oped to be used in patients with metabolic syndrome to
ameliorate elevated serum LDL and atherosclerotic cardio-
vascular diseases [45]. Our results that resistin compromises
the anti-mycobacterial responses in type 2 DM patients
through inhibiting ROS production and inflammasome acti-
vation therefore implicate that the use of resistin inhibitors in
type 2 DM patients may have the additional benefit of
improving the immunity against mycobacteria.
[1] WHO. Global tuberculosis control report. Geneva: World Health Orga-
nization; 2013.
[2] Jeon CY, Murray MB. Diabetes mellitus increases the risk of active
tuberculosis: a systematic review of 13 observational studies. PLoS Med
2008;5:e152.
[3] Chang JT, Dou HY, Yen CL, Wu YH, Huang RM, Lin HJ, et al. Effect of
type 2 diabetes mellitus on the clinical severity and treatment outcome in
patients with pulmonary tuberculosis: a potential role in the emergence
of multidrug-resistance. J Formos Med Assoc 2011;110:372e81.
[4] Restrepo BI, Schlesinger LS. Host-pathogen interactions in tuberculosis
patients with type 2 diabetes mellitus. Tuberc (Edinb) 2013;(93
Suppl):S10e4.
[5] Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM,
et al. The hormone resistin links obesity to diabetes. Nature
2001;409:307e12.
[6] Bostrom EA, Tarkowski A, Bokarewa M. Resistin is stored in neutrophil
granules being released upon challenge with inflammatory stimuli.
Biochim Biophys Acta 2009;1793:1894e900.
[7] Menzaghi C, Bacci S, Salvemini L, Mendonca C, Palladino G,
Fontana A, et al. Serum resistin, cardiovascular disease and all-cause
mortality in patients with type 2 diabetes. PLoS One 2013;8:e64729.
[8] Senolt L, Housa D, Vernerova Z, Jirasek T, Svobodova R, Veigl D, et al.
Resistin in rheumatoid arthritis synovial tissue, synovial fluid and serum.
Ann Rheum Dis 2007;66:458e63.
[9] Koch A, Gressner OA, Sanson E, Tacke F, Trautwein C. Serum resistin
levels in critically ill patients are associated with inflammation, organ

Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
10 W.-C. Chao et al. / Microbes and Infection xx (2014) 1e10
dysfunction and metabolism and may predict survival of non-septic pa-
tients. Crit Care 2009;13:R95.
[10] Cohen G, Ilic D, Raupachova J, Horl WH. Resistin inhibits essential
functions of polymorphonuclear leukocytes. J Immunol
2008;181:3761e8.
[11] Bylund J, Brown KL, Movitz C, Dahlgren C, Karlsson A. Intracellular
generation of superoxide by the phagocyte NADPH oxidase: how, where,
and what for? Free Radic Biol Med 2010;49:1834e45.
[12] Miller JL, Velmurugan K, Cowan MJ, Briken V. The type I NADH de-
hydrogenase of Mycobacterium tuberculosis counters phagosomal NOX2
activity to inhibit TNF-alpha-mediated host cell apoptosis. PLoS Pathog
2010;6:e1000864.
[13] Eum SY, Kong JH, Hong MS, Lee YJ, Kim JH, Hwang SH, et al.
Neutrophils are the predominant infected phagocytic cells in the airways
of patients with active pulmonary TB. Chest 2010;137:122e8.
[14.] Jayaraman P, Sada-Ovalle I, Nishimura T, Anderson AC, Kuchroo VK,
Remold HG, et al. IL-1beta promotes antimicrobial immunity in mac-
rophages by regulating TNFR signaling and caspase-3 activation. J
Immunol 2013;190:4196e204.
[15] Chen CC, Tsai SH, Lu CC, Hu ST, Wu TS, Huang TT, et al. Activation of
an NLRP3 inflammasome restricts Mycobacterium kansasii infection.
PLoS One 2012;7:e36292.
[16] Huang YF, Liu SY, Yen CL, Yang PW, Shieh CC. Thapsigargin and flavin
adenine dinucleotide ex vivo treatment rescues trafficking-defective
gp91phox in chronic granulomatous disease leukocytes. Free Radic
Biol Med 2009;47:932e40.
[17] Xu H, Tian W, Li SJ, Zhang LY, Liu W, Zhao Y, et al. Clinical and
molecular features of 38 children with chronic granulomatous disease in
mainland china. J Clin Immunol 2014;34:633e41.
[18] Vogt G, Nathan C. In vitro differentiation of human macrophages with
enhanced antimycobacterial activity. J Clin Investig
2011;121:3889e901.
[19] Odegaard JI, Chawla A. Pleiotropic actions of insulin resistance and
inflammation in metabolic homeostasis. Science 2013;339:172e7.
[20] Tobin DM, Ramakrishnan L. Comparative pathogenesis of Mycobacte-
rium marinum and Mycobacterium tuberculosis. Cell Microbiol
2008;10:1027e39.
[21] Mishra BB, Moura-Alves P, Sonawane A, Hacohen N, Griffiths G,
Moita LF, et al. Mycobacterium tuberculosis protein ESAT-6 is a potent
activator of the NLRP3/ASC inflammasome. Cell Microbiol
2010;12:1046e63.
[22] Tan KS, Lee KO, Low KC, Gamage AM, Liu Y, Tan GY, et al. Gluta-
thione deficiency in type 2 diabetes impairs cytokine responses and
control of intracellular bacteria. J Clin Investig 2012;122:2289e300.
[23] Restrepo BI, Fisher-Hoch SP, Pino PA, Salinas A, Rahbar MH, Mora F,
et al. Tuberculosis in poorly controlled type 2 diabetes: altered cytokine
expression in peripheral white blood cells. Clin Infect Dis
2008;47:634e41.
[24] Kumar NP, Sridhar R, Banurekha VV, Jawahar MS, Fay MP, Nutman TB,
et al. Type 2 diabetes mellitus coincident with pulmonary tuberculosis is
associated with heightened systemic type 1, type 17, and other proin-
flammatory cytokines. Ann Am Thorac Soc 2013;10:441e9.
[25] Fremond CM, Togbe D, Doz E, Rose S, Vasseur V, Maillet I, et al. IL-1
receptor-mediated signal is an essential component of MyD88-dependent
innate response to Mycobacterium tuberculosis infection. J Immunol
2007;179:1178e89.
Please cite this article in press as: Chao W-C, et al., Increased resistin may suppress reactive oxygen species production and inflammasome activation in type 2
diabetic patients with pulmonary tuberculosis infection, Microbes and Infection (2014), http://dx.doi.org/10.1016/j.micinf.2014.11.009
View publication stats
[26] Travis WD, Travis LB, Roberts GD, Su DW, Weiland LW. The histo-
pathologic spectrum in Mycobacterium marinum infection. Arch Pathol
Lab Med 1985;109:1109e13.
[27] Chandra Rai R, Dwivedi VP, Chatterjee S, Raghava Prasad DV, Das G.
Early secretory antigenic target-6 of Mycobacterium tuberculosis: enig-
matic factor in pathogen-host interactions. Microbes Infect
2012;14:1220e6.
[28] Stamm LM, Brown EJ. Mycobacterium marinum: the generalization and
specialization of a pathogenic mycobacterium. Microbes Infect
2004;6:1418e28.
[29] Davis JM, Ramakrishnan L. The role of the granuloma in expansion and
dissemination of early tuberculous infection. Cell 2009;136:37e49.
[30] Yang CT, Cambier CJ, Davis JM, Hall CJ, Crosier PS, Ramakrishnan L.
Neutrophils exert protection in the early tuberculous granuloma by
oxidative killing of mycobacteria phagocytosed from infected macro-
phages. Cell Host Microbe 2012;12:301e12.
[31] Jin C, Flavell RA. Molecular mechanism of NLRP3 inflammasome
activation. J Clin Immunol 2010;30:628e31.
[32] Schroder K, Zhou R, Tschopp J. The NLRP3 inflammasome: a sensor for
metabolic danger? Science 2010;327:296e300.
[33] Lee HM, Kim JJ, Kim HJ, Shong M, Ku BJ, Jo EK. Upregulated NLRP3
inflammasome activation in patients with type 2 diabetes. Diabetes
2013;62:194e204.
[34] Korbel DS, Schneider BE, Schaible UE. Innate immunity in tuberculosis:
myths and truth. Microbes Infect 2008;10:995e1004.
[35] Kaufmann SH, Dorhoi A. Inflammation in tuberculosis: interactions,
imbalances and interventions. Curr Opin Immunol 2013;25:441e9.
[36] Deffert C, Cachat J, Krause KH. Phagocyte NADPH oxidase, chronic
granulomatous disease and mycobacterial infections. Cell Microbiol
2014;16:1168e78.
[37] Palanisamy GS, Kirk NM, Ackart DF, Shanley CA, Orme IM,
Basaraba RJ. Evidence for oxidative stress and defective antioxidant
response in guinea pigs with tuberculosis. PLoS One 2011;6:e26254.
[38] Podell BK, Ackart DF, Kirk NM, Eck SP, Bell C, Basaraba RJ. Non-
diabetic hyperglycemia exacerbates disease severity in Mycobacterium
tuberculosis infected guinea pigs. PLoS One 2012;7:e46824.
[39] Vallerskog T, Martens GW, Kornfeld H. Diabetic mice display a delayed
adaptive immune response to Mycobacterium tuberculosis. J Immunol
2010;184:6275e82.
[40] Divangahi M, Behar SM, Remold H. Dying to live: how the death mo-
dality of the infected macrophage affects immunity to tuberculosis. Adv
Exp Med Biol 2013;783:103e20.
[41] Bokarewa M, Nagaev I, Dahlberg L, Smith U, Tarkowski A. Resistin, an
adipokine with potent proinflammatory properties. J Immunol
2005;174:5789e95.
[42] Stehlik C. Multiple interleukin-1beta-converting enzymes contribute to
inflammatory arthritis. Arthritis Rheum 2009;60:3524e30.
[43] Ehtesham NZ, Nasiruddin M, Alvi A, Kumar BK, Ahmed N, Peri S, et al.
Treatment end point determinants for pulmonary tuberculosis: human
resistin as a surrogate biomarker. Tuberc (Edinb) 2011;91:293e9.
[44] Chang SW, Pan WS, Lozano Beltran D, Oleyda Baldelomar L,
Solano MA, Tuero I, et al. Gut hormones, appetite suppression and
cachexia in patients with pulmonary TB. PLoS One 2013;8:e54564.
[45] Sahebkar A. Beyond anti-PCSK9 therapies: the potential role of resistin
inhibitors. Nat Rev Cardiol 2014;11:12.