Skip to main content
NCBI home page
Cellular and Molecular Immunology logoLink to Cellular and Molecular Immunology
. 2018 May 30;16(5):495–507. doi: 10.1038/s41423-018-0038-7

MicroRNA-99a mimics inhibit M1 macrophage phenotype and adipose tissue inflammation by targeting TNFα

Anant Jaiswal 1, Sukka Santosh Reddy 2, Mohita Maurya 1, Preeti Maurya 2, Manoj Kumar Barthwal 1,✉
PMCID: PMC6474300  PMID: 29849090

Abstract

In human adipose tissue and obesity, miR-99a expression is negatively correlated with inflammation. Therefore, the present study investigated the role of miR-99a in macrophage phenotype activation and adipose tissue inflammation. M2 BMDMs showed a significant increase in miR-99a expression when compared to the M0 and M1 phenotypes. Phenotype-switching experiments established an association between upregulated miR-99a expression and the M2 phenotype. Overexpression of miR-99a prevented M1 phenotype activation and attenuated bactericidal activity. Likewise, knockdown of miR-99a abolished M2 phenotype activation. By means of in silico target prediction tools and a luciferase reporter assay, TNFα was identified as a direct target of miR-99a. Knockdown of TNFα recapitulated the effect of miR-99a overexpression in M1 BMDMs. In a db/db mice model, miR-99a expression was reduced in eWAT and F4/80+ ATMs. Systemic overexpression of miR-99a in db/db mice attenuated adipocyte hypertrophy with increased CD301 and reduced CD86 immunostaining. Flow cytometry analysis also showed an increased M2 and a reduced M1 macrophage population. Mimics of miR-99a also improved the diabetic dyslipidemia and insulin signaling in eWAT and liver, with an attenuated expression of gluconeogenesis and cholesterol metabolism genes in the liver. Furthermore, adoptive transfer of miR-99a-overexpressing macrophages in the db/db mice recapitulated in vivo miR-99a mimic effects with increased M2 and reduced M1 macrophage populations and improved systemic glucose, insulin sensitivity, and insulin signaling in the eWAT and liver. The present study demonstrates that miR-99a mimics can regulate macrophage M1 phenotype activation by targeting TNFα. miR-99a therapeutics in diabetic mice reduces the adipose tissue inflammation and improves insulin sensitivity.

Introduction

The macrophage phenotype can regulate adipose tissue inflammation and insulin resistance. Macrophage-specific deletion of proinflammatory mediators inhibits M1 polarization and reduces adipose tissue inflammation.1,2 Similarly, M2 phenotype regulators, such as IRF4 and PPARγ, promote an anti-inflammatory response and decrease adipose tissue inflammation.3,4 The genetic deletion of cells such as CD8+ T cells also reduces macrophage recruitment and inflammation in adipose tissue.5 Conversely, the activation of NKT cells enhances adipose tissue M2 polarization and improves insulin sensitivity.6

Whole-genome studies have theorized that there is an essential role for noncoding RNAs in the regulation of insulin resistance and T2DM.7 Among the noncoding RNAs, microRNAs (miRNAs) are a group of small endogenous noncoding RNAs that regulate gene expression at the posttranscriptional level by binding to the 3ʹ-untranslated region of target mRNAs.8 Moreover, several miRNAs have been identified that play an essential role in macrophage function and adipose tissue inflammation.9,10 In therapeutic interventions, a number of miRNAs have also progressed from preclinical to phase-I and phase-II trials.11

Recent studies have demonstrated the role of miR-99a as a tumor suppressor in some types of cancer, as well as a negative regulator of cardiac hypertrophy.12,13 Additionally, miR-99a expression negatively correlates with fatty acids, inflammatory mediators, and obesity in human adipose tissue.14,15 This suggests that miR-99a might be a potential regulator of adipose tissue inflammation.

Emerging evidence also suggests that miRNAs associated in a cluster often exert similar functions in several cell types and diseases.16 miR-99a is a known clustered miRNA and localizes in a cluster with let-7c (on Chr-21 in human and Chr-16 in murine). Let-7c is known to regulate macrophage polarization by inhibiting M1 and promoting M2 phenotype activation.17 However, the role of miR-99a in macrophage phenotype activation is not yet known. Therefore, the current study evaluates the role of miR-99a in macrophage phenotype activation and adipose tissue inflammation.

Materials and methods

Cell culture and treatments

Bone marrow-derived macrophages (BMDMs) were isolated from the femur and tibia of C57BL/6J mice. Briefly, the bones were flushed with ice-cold PBS containing 20 U/ml heparin, followed by incubation with sterile 0.843% ammonium chloride solution for 10 min to lyse the erythrocytes. A single-cell suspension was obtained by straining through a 70-µm cell strainer. The cells were then plated in RPMI medium (Sigma Aldrich) supplemented with 10% FBS (Gibco) and 10 ng/ml MCSF (Peprotech) for 7 days. On the seventh day, cells were starved from MCSF for 12 h and stimulated for 24 h with 1 μg/ml LPS (Sigma-Aldrich) and either 20 ng/ml IFN-γ (Peprotech) to obtain the M1 phenotype or 20 ng/ml IL-4 (Peprotech) to obtain the M2 phenotype. In the reswitching experiments, the M1 macrophages were treated with either LPS/IFN-γ (M1 control) or IL-4 (reswitch) for 24 h. Similarly, M2 macrophages were treated with either IL-4 (M2 control) or LPS/IFN-γ (reswitch) for 24 h.18

Lentivirus production and transduction in BMDMs

Lentiviral particles were produced by cotransfection of lentiviral helping plasmids and a plasmid encoding the gene of interest, miR-99a (MmiR3457-MR03, GeneCopoeia), miR-scramble control (CmiR0001-MR03, GeneCopoeia), TNFα-shRNA (TRCN0000362936, Sigma-Aldrich) or scramble shRNA (SHC001) or lentiviral-miR-99a inhibitor (1020171SMN, Sigma-Aldrich) in HEK293T cells using Lipofectamine 2000 (Invitrogen). The lentiviral particles were titrated and concentrated as previously described.19 On the sixth day, BMDMs were treated with 1.5 × 106 PFU lentiviral particles in the presence of polybrene (8 mg/ml). After 48 h, the cells were starved from MCSF and treated for 24 h with 1 μg/ml LPS and either 20 ng/ml IFN-γ for the M1 phenotype or 20 ng/ml IL-4 for the M2 phenotype.20

Adoptive transfer miR-99 transduced BMDMs in db/db mice

BMDMs were isolated from db/db animals and cultured for 7 days in RPMI medium in the presence of MCSF. For miR-99a overexpression, BMDMs were transduced with either miR-99a or miR-scramble lentiviral particles, as described above. The transduction efficiency was regularly assessed by qRT-PCR and microscopy. For the in vivo adoptive transfer experiments, transduced BMDMs were smoothly transferred to a culture tube and maintained in PBS until injection into the animal. A total of 1 × 106 cells per mice were administered for six consecutive days through tail-vein injection.21 Random and fasting glucose levels were measured after 3 days and 10 days from the first injection, respectively. An intraperitoneal glucose tolerance test (ipGTT) and an insulin tolerance test (ITT) were performed before the first injection and at 4 days after the last injection in the group.

ATM quantification and sorting by flow cytometry

After the desired treatment, the epididymal white adipose tissue (eWAT) from 10- to 12-week-old male db/db mice and control db/ + mice was excised, rinsed in 1×DPBS (PBS with CaCl2 and 0.5% BSA), and minced and digested with collagenase solution (collagenase type-II, Sigma-Aldrich, 1 mg/ml in DPBS) at 37 °C for 20 min with shaking. The digested solution was strained through a 100-μm filter to obtain a single-cell suspension. Floating adipocytes and the SVF pellet were separated by centrifugation at 500 × g for 5 min. The SVF pellet was suspended in ACK buffer to lyse the erythrocytes, followed by washing with DPBS and centrifugation. For flow cytometry, the SVF pellet was resuspended in FACS buffer (DPBS, 2 mM EDTA, and 1% FBS) and stained with the following conjugated antibodies for 15 min at 4 °C: F4/80-PE (12-4801, eBioscience), CD11b-PE-cyanine7 (25-0112, eBioscience), CD11c-eFluor450 (48-0114, eBioscience), CD206−AlexaFluor®647 (565250, BD-Bioscience), and Ly6C-APC (17-5932, eBioscience). After staining, the samples were washed with FACS buffer and analyzed by FACS Diva software on a FACS Aria (Becton Dickinson, USA). For cell sorting, equal numbers of F4/80+CD11b+CD11c+CD206− (M1) and F4/80+CD11b+ CD11c−CD206+ (M2) cells were quantified and sorted for RNA extraction.22

Bactericidal activity of BMDMs

The bacteria-killing capacity of BMDMs was performed as previously described.17 miR-99a or miR-scramble-overexpressed BMDMs were stimulated with 1 μg/ml LPS and 20 ng/ml IFN-γ for the M1 phenotype. After 24 h, the cells were incubated with 1 × 106 CFU/ml Escherichia coli (pCMV6−vector) in 96-well plates. The plates were centrifuged at 400 × g for 5 min and incubated for 60 min at 37 °C. The supernatant from each well was collected and subjected to 100-fold dilutions. A total of 100 μl of the diluted supernatant was plated on Luria broth agar plates. The plates were incubated at 37 °C overnight. Bacterial colonies were counted and expressed as CFU/ml. CFU/ml = number of colonies × dilution factor/volume of the plated diluted supernatant.

microRNA isolation and TaqMan RT-PCR assay

Total RNA, including miRNA, was isolated from cells or tissues utilizing the miRVanaTM miRNA isolation kit (AM1561, Ambion, Lifetech) under RNase-free conditions, according to the manufacturer’s protocol. Total RNA was eluted with preheated nuclease-free water. The purity of the RNA was assessed by a NanoDrop™2000c spectrophotometer (Thermo Scientific), and high-quality RNA (A260/280 ratio > 1.8) preparations were used for further processing. miRNA expression was measured using the Taqman® Single miRNA assay (assay ID; 000435, LifeTech) as per the manufacturer’s protocol. Briefly, 500 ng of total RNA was reverse transcribed using the TaqMan miRNA Reverse Transcription Kit (4366596, LifeTech) with 3 μl of miRNA-specific RT primers on a thermal cycler (Applied Biosystems). Real-time PCR was performed on QuantStudio™12K Flex Real-Time PCR System (Applied Biosystems®, LifeTech) using TaqMan® Universal PCR Master Mix (4304437, LifeTech) with miRNA-specific TM primers. miRNA expression was normalized to U6snRNA. The fold change of miRNA expression was calculated by QuantStudio™12K Flex Real-Time software v1.2.2 (Applied Biosystems®, LifeTech).23

mRNA isolation and qPCR

mRNA isolation and qPCR were performed as previously described.24 Briefly, total RNA was isolated from BMDMs or tissues with TRIzol™ Reagent (Thermo Fisher Scientific) or for sorted ATMs, with an RNeasy Mini Kit (74104-QIAGEN) and reverse transcribed with a high-capacity cDNA reverse transcription kit (4368814, LifeTech). Real-time qPCR was performed in a LightCycler® 480II system (Roche Applied Science, USA) using gene-specific primers and SYBR Green reagent. Gene expression was normalized to 18SrRNA. The primer sequences are listed in Supplementary Table I.

In silico target prediction analysis

The prediction analysis of miRNA binding to the TNF 3′-UTR was performed by using the miRwalk databases utilizing different algorithms, including, miRwalk, miRanda, DIANA-mT, RNA22, TargetScanv6.2, and miRDB.25 Of the six algorithms, four considered TNF to be a potential target for miR-99a. To determine the individual score of identity, conserved miRNAs with good mirSVR scores were selected using the miRanda algorithm.

Luciferase reporter assay

The 3′-UTR of TNFα (MmiT077890-MT06, GeneCopoeia) and miR-99a (MmiR3457-MR03, GeneCopoeia) or miR-scramble control (CmiR0001-MR03, GeneCopoeia) were cotransfected by Lipofectamine 2000 (Invitrogen) into HEK293T cells. The firefly and renilla luciferase activity in the cell lysates was measured at 72 h post transfection using a dual-luciferase assay system (E1910; Promega) according to the manufacturer’s protocol. The data were normalized as the ratio of renilla/firefly luciferase activity.26

Western blot assay

After the desired treatment in db/db animals, liver and adipose tissues were collected and lysed in RIPA lysis buffer. The membranes were probed overnight at 4 °C with primary antibodies including p-IRS-1Ser307 (sc-33956), IRS-1 (C-20, sc-559), p-AktSer473 (D9E, 4060; CST), Akt (11E7, 4685; CST), and GAPDH (ABS16; MERK), followed by incubation with specific secondary HRP-conjugated antibody at RT for 2 h. Phosphorylation was normalized to the respective total protein expression. The bands were detected by ECL, and the intensity was quantified by Image Quant, LAS 4000 (Amersham, MA).27

Animal experiments

All animal experiments were approved by the Institutional Animal Ethical Committee, Council for Scientific and Industrial Research-Central Drug Research Institute (CSIR-CDRI), as per guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA). The 10–12-week-old male diabetic (C57BLKs-db/db) and nondiabetic control (C57BLKs-db/+) mice were maintained on a 12-h light–dark cycle with ad libitum access to food and water at the Council of Scientific and Industrial Research-Central Drug Research Institute (CSIR-CDRI). A total of 1 × 107–108 lentiviral particles containing miR-99a or miR-scramble control were injected intravenously (once per day) for three consecutive days in 6–8 male db/db mice per group using Lipofectamine 2000 (Life Technologies). Random and fasting glucose levels were measured after 3 days and 10 days from the first injection, respectively. An intraperitoneal glucose tolerance test (ipGTT) and an insulin tolerance test (ITT) were performed before the first injection and at 7 days after the last lentiviral injection in the group.28

ipGTT and ITT

ipGTT and ITT were performed after 12 h or 6 h of fasting, respectively. The mice were intraperitoneally administered glucose (2 g/kg for ipGTT) or human insulin (0.75 IU/kg for ITT, Eli Lilly, Indianapolis, IN) for the respective tests. Blood glucose was monitored at the indicated time points with an Accu-Chek Active blood glucometer. The area under the curve (AUC) for ipGTT and ITT was calculated.29

Plasma measurements

Mice fasted for 6 h, and plasma was collected prior killing. Circulatory lipids, total cholesterol (TC), triglycerides (TG), and low- and high-density lipoprotein (LDL, HDL) were estimated using commercial kits (Randox, UK). The insulin levels were measured using ELISA kits (Millipore, USA). The HOMA-IR functionality was calculated as per the previously used formulae.30

Hematoxylin and eosin (H&E) stain and immunohistochemistry

Formalin-fixed tissues were processed in graded concentrations of ethanol followed by xylene prior to liquid paraffin infiltration. The paraffin-embedded tissue blocks were sectioned (4–5-μm slices), and HE-stained images were captured in the bright-field mode using an upright microscope (DM5000, Leica Microsystems, Germany). Fifty fields were scanned for each sample, and ten random fields were analyzed. The number of adipocytes was counted, and their area was obtained. The morphometric results were quantified using the ImageJ software (Image Processing and Analysis in Java, public domains).31 The immunofluorescence analyses of CD86 (1:200, sc-28347; Santa Cruz) and CD301 (1:400, MCA2392; Bio-Rad) were conducted after deparaffinization, as previously described.32 Secondary antibodies were then applied (AlexaFluor 594, A1105; Invitrogen), followed by DAPI (D9542, Sigma). The sections were mounted and visualized using a fluorescence microscope (DM6000, Leica Microsystems, Germany). Macrophages were counted in a randomly selected area, and cell types showing cytoplasmic staining were considered positive for quantitative analysis. The number of macrophages per area was calculated using the following formula: number of positive cells per field/adipocyte area per field.33 The images were quantified using ImageJ.

Statistical analysis

The experimental values in the results are expressed as the mean ± SEM. All experiments were performed at least three times. The significance of the difference between groups was analyzed by performing one-way analysis of variance (ANOVA), followed by the Bonferroni post hoc test. The significance level for the Bonferroni multiple-comparison test was set to 0.05 for three or more groups. The difference between two groups was compared using the unpaired Student’s t-test, and a p-value < 0.05 was considered to be statistically significant.

Results

miR-99a regulates M1 macrophage phenotype activation

To understand the correlation between miR-99a expression and macrophage phenotype, we first estimated the levels of miR-99a using a Taqman RT-PCR assay in unstimulated (M0), LPS/IFN-γ-stimulated classically activated (M1), and IL-4-treated alternatively activated (M2) bone marrow-derived macrophages (BMDMs). Expression of miR-99a was significantly higher in the M2 macrophages (~6-fold) compared to M0 and M1 (Fig. 1a). LPS/IFN-γ stimulation induced a significant increase in the mRNA expression of iNOS (~10-fold) when compared to M2 and M0 macrophages, indicating activation of the M1 phenotype (Fig. 1b). Concurrently, IL-4-induced macrophages showed a significant increase in YM-1 mRNA expression (~8-fold) when compared to M1 and M0 macrophages (Fig. 1c), indicating M2 phenotype activation. To assess whether miR-99a expression is regulated during macrophage phenotype activation, macrophage phenotype reswitch experiments were performed. M1 or M2 BMDMs were reconverted into the M2 or M1 type by treatment with IL-4 or LPS/IFN-γ, respectively. Successful reswitching of M1 macrophages into M2 by IL-4 was confirmed since a significant decrease in iNOS, MCP-1, and IL-1β mRNA expression (Fig. 1d), and a significant increase in YM-1, ARG-1, and PPARγ mRNA expression (Fig. 1e) was observed when compared to control M1 macrophages. Similarly, the reswitching of M2 macrophages into M1 after LPS/IFN-γ stimulation induced a significant increase in iNOS, MCP-1, and IL-1β mRNA expression (Fig. 1f) and a significant decrease in YM-1, ARG-1, and PPARγ mRNA expression (Fig. 1g) when compared to control M2 macrophages. Furthermore, reswitching of the M1 macrophage phenotype into M2 significantly increased miR-99a expression (3.5-fold, Fig. 1h), and reswitching of the M2 macrophage phenotype to M1 significantly reduced miR-99a expression (2.5-fold, Fig. 1i), suggesting the modulation of miR-99a expression with macrophage phenotype. The enhanced expression of miR-99a in M2- and M1-derived M2 macrophages suggested the role of this miRNA in M2 phenotype activation. To validate this, we infected BMDMs with lentivirus containing miR-99a or miR-scramble, followed by stimulation with LPS/IFN-γ for M1 phenotype activation. The overexpression of miR-99a was confirmed by qPCR (Fig. S1A). Overexpression of miR-99a significantly reduced the M1 markers iNOS, MCP-1, and IL-1β (Fig. 2a) and enhanced the expression of the M2 markers, YM-1, ARG-1, and PPARγ (Fig. 2b), in LPS/IFN-γ-stimulated BMDMs when compared to miR-scramble-treated BMDMs. Similar results were obtained at the protein level where a significant decrease in CD86 expression (Fig. 2c) and an enhanced CD206 expression were observed after infecting with miR-99a mimics when compared to miR-scramble (Fig. 2d). These data suggest that miR-99a plays a critical role in inhibiting M1 phenotype activation. Next, to test whether changes in the phenotype also affected the macrophage function, the bactericidal activity of LPS/IFN-γ-stimulated BMDMs was assessed after treatment with miR-99a or miR-scramble mimics. Overexpression of miR-99a markedly diminished LPS/IFN-γ-induced macrophage bactericidal activity when compared to miR-scramble control (Fig. 2e).

Fig. 1.

Fig. 1

Open in a new tab

miR-99a expression and macrophage phenotype activation. a miR-99a expression levels were measured in control (M0) or LPS/IFN-γ (M1) and IL-4 (M2)-stimulated BMDMs by Taqman-qPCR (n = 6). b iNOS mRNA expression in control (M0), LPS/IFN-γ- (M1), and IL-4 (M2)-stimulated BMDMs by qPCR (n = 4). c YM-1 mRNA expression in control (M0), LPS/IFN-γ (M1)-, and IL-4 (M2)-stimulated BMDMs by qPCR (n = 4). d iNOS, MCP-1, and IL-1β mRNA expression in M1 to M1 (control) and M1 to M2 (reswitch) BMDMs by qPCR (n = 6). e YM-1, ARG-1, and PPARγ mRNA expression in M1 to M1 (control) and M1 to M2 (reswitch) BMDMs by qPCR (n = 6). f iNOS, MCP-1, and IL-1β mRNA expression in M2 to M2 (control) and M2 to M1 (reswitch) BMDMs by qPCR (n = 6). g YM-1, ARG-1, and PPARγ mRNA expression in M2 to M2 (control) and M2 to M1 (reswitch) BMDMs by qPCR (n = 6). h miR-99a expression levels in M1 to M1 (control) and M1 to M2 (reswitch) BMDMs by Taqman-qPCR (n = 6). i miR-99a expression levels in M2 to M2 (control) and M2 to M1 (reswitch) BMDMs by Taqman-qPCR (n = 6). Data are presented as the mean ± SEM of 4–6 independent experiments. **P < 0.01, ***P < 0.001, ###P < 0.001 vs. M1 and $$$P < 0.001 vs. M0

Fig. 2.

Fig. 2

Open in a new tab

Effect of miR-99a modulation on M1 and M2 macrophage phenotype and function. BMDMs were infected with 1.5 × 106 PFU lentivirus containing miR-99a/miR- scramble or miR-99a antagomir/miR-scramble antagomir, and after 72 h, the cells were stimulated with vehicle (M0, control) or LPS/IFN-γ (M1) or IL-4 (M2) for 24 h. a iNOS, MCP-1, and IL-1β mRNA expression. b YM-1, ARG-1, and PPARγ mRNA expression (n = 6). c CD86 protein expression. d CD206 protein expression (n = 3). e Bacterial colonies were counted and CFUs of E. coli were determined in miR-99a- or miR-scramble-treated M1 and control M0 (n = 6) cells. f YM-1, ARG-1, and PPARγ mRNA expression in miR-99a antagomir- or miR-scramble antagomir-treated M2 and control M0 cells by qPCR (n = 6). g CD206 protein expression in miR-99a antagomir- or miR-scramble antagomir-treated M2 and control M0 cells by western blot (n = 3). Data are presented as the mean ± SEM of 3–6 independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001 vs. miR-scramble. Blots represent at least one of three similar experiments

miR-99a silencing experiments were performed with miR-99a or miR-scramble antagomir. BMDMs were infected with lentivirus containing miR-99a or miR-scramble antagomir, followed by stimulation with IL-4 for M2 phenotype activation. The inhibition of miR-99a was confirmed by qPCR (Fig. S1B). miR-99a antagomir treatment significantly inhibited the IL-4-induced increase in the mRNA expression of YM-1, ARG-1, and PPARγ when compared to scramble-treated control (Fig. 2f). Similar results were obtained at the protein level, with a significant decrease in CD206 expression after miR-99a antagomir treatment (Fig. 2g).

miR-99a mimics prevent M1 macrophage activation by targeting TNFα

To address the mechanism by which miR-99a affects macrophage phenotype activation, we performed in silico analysis of mRNA target prediction.34 Of the six prediction algorithms, four (miRWalk, RNAhybrid, miRanda, and TargetScan) identified a miR-99a-putative target site within the 3ʹ-untranslated regions (3ʹ-UTR) of TNFα (Fig. 3a, b). TNFα is a known inducer of M1 phenotype activation.35 Furthermore, overexpression of miR-99a in LPS/IFN-γ-stimulated BMDMs significantly reduced TNFα expression at both the mRNA and protein levels when compared to scramble miRNA control (Fig. 3c, d). To further ascertain the direct effect of miR-99a on TNFα, we performed a luciferase reporter assay, in which binding of miR-99a at the 3ʹ-UTR of TNFα was monitored. miR-99a overexpression significantly attenuated TNFα 3ʹ-UTR (~2.5-fold) luciferase activity when compared to miR-scramble control (Fig. 3e), indicating that miR-99a directly binds to the 3ʹ-UTR of TNFα. Lentiviral-mediated knockdown of TNFα by shRNA (shTNF) significantly reduced LPS/IFN-γ-induced TNFα mRNA expression in the BMDMs when compared to scramble shRNA-treated cells (Fig. S1C). Knockdown of TNFα significantly reduced the mRNA expression of proinflammatory M1 macrophage signature genes (iNOS, MCP-1, and IL-1β) (Fig. S1D) and significantly increased the mRNA expression of M2 markers (YM-1, ARG-1, and PPARγ) (Fig. S1E) in LPS/IFN-γ-stimulated BMDMs when compared to scrambled shRNA control. Additionally, the bactericidal activity of LPS/IFN-γ-stimulated cells that were pretreated with TNFα-shRNA was significantly reduced when compared to scramble shRNA control (Fig. S1F).

Fig. 3.

Fig. 3

Open in a new tab

miR-99a targets TNFα. a In silico, four different target prediction software packages predicted TNFα to be the target of miR-99a. b Predicted miR-99a binding site in the 3ʹ untranslated region (UTR) of TNFα seed sequences. c TNFα mRNA expression in control (M0) or miR-99a- and miR-scramble-pretreated and LPS/IFN-γ (M1)-stimulated BMDMs by qPCR (n = 6). d TNFα protein expression in miR-99a- and miR-scramble-pretreated and LPS/IFN-γ (M1)-stimulated BMDMs, as assessed by western blotting (n = 3). e Reporter constructs with a 3ʹ-UTR region containing the binding site of TNFα were cotransfected with miR-scramble or miR-99a in HEK293T cells. Luciferase activity was analyzed after 72 h (n = 6). Data are presented as the mean ± SEM of six independent experiments. **P < 0.01; ***P < 0.001 vs. miR-scramble

miR-99a mimics reduce M1 macrophage phenotype and adipose tissue inflammation in diabetic mice

Since the macrophage phenotype affects adipose tissue inflammation and insulin resistance, miR-99a expression was evaluated in eWAT of control (db/+) and diabetic (db/db) mice. miR-99a expression was significantly reduced in eWAT (~2.5-fold) and F4/80+sorted macrophages (~5-fold) from db/db mice when compared to db/+ control mice (Fig. 4a, b). To further study the association of miR-99a with the macrophage phenotype in vivo, miR-99a expression was analyzed in F4/80+CD11b+CD11c+CD206− M1 and F4/80+CD11b+CD11c−CD206+ M2-sorted ATMs from db/+ animals. M2 ATMs from db/+ mice showed a significantly increased expression of miR-99a when compared to M1 ATMs from db/+ mice (Fig. 4c, d). On the basis of the decreased expression of miR-99a in ATMs from diabetic mice and its positive association with M2 ATMs, we hypothesized that miR-99a mimics may modulate the adipose tissue macrophage phenotype and inflammation. To confirm this, diabetic mice were treated with lentivirus containing miR-99a or miR-scramble mimics for three consecutive days (Fig. 4e). Overexpression of miR-99a was confirmed in the F4/80+CD11b+ ATMs and total eWAT (Fig. S2.A–C), as well as in F4/80+CD11b+ liver macrophages and total liver by qPCR (Fig. S2.D–F). Significantly upregulated mRNA expression of M2 and reduced M1 macrophage markers were found in total eWAT of miR-99a mimic-treated db/db mice when compared to miR-scramble-treated mice (Fig. 4f). To further assess the changes in macrophage phenotype, F4/80+CD11b+CD11c+CD206−M1 and F4/80+CD11b+CD11c−CD206+M2 ATMs from eWAT of db/db mice treated with miR-99a were quantified and evaluated by flow cytometry. miR-99a mimic-treated diabetic mice showed a significant increase in F4/80+CD11b+CD11c−CD206+ M2 ATMs and a decrease in F4/80+CD11b+CD11c+CD206−M1 ATMs when compared to miR-scramble-treated diabetic mice (Fig. 4g, h). Furthermore, eWAT of miR-99a-treated db/db mice showed a significant decrease in CD86 (Fig. 5a) and a significant increase in CD301 (Fig. 5b) staining when compared to miR-scramble-treated db/db mice. Moreover, in miR-99a-overexpressed db/db mice, there was a significant decrease in average adipocyte area when compared to miR-scramble-treated db/db mice (Fig. 5c). However, no significant difference in the number of F4/80+Ly6c+ ATMs was observed between the miR-99a- and miR-scramble-treated db/db mice (Fig. 5d).

Fig. 4.

Fig. 4

Open in a new tab

miR-99a overexpression modulates ATM phenotype. miR-99a expression was measured in a eWAT b F4/80+-sorted ATMs by Taqman-qPCR (n = 6). c The dot plot represents labeling of F4/80+ CD11b+CD11c+ CD206-(M1) and F4/80+CD11b+CD11c−CD206+(M2) ATMs sorted from db/ + mice. d miR-99a expression in M1- and M2-sorted ATMs from db/+ mice by Taqman-qPCR (n = 6). e Treatment regime of miR-99a or miR-scramble lentivirus and timeline for the measurement of parameters. f db/db mice were treated with miR-99a or miR-scramble, and mRNA expression of M1 (iNOS, MCP-1, and TNFα) and M2 (YM-1, ARG-1, and PPARγ) macrophage signature genes was determined in total eWAT by qPCR (n = 6). g Dot plot represents labeling of F4/80+CD11b+CD11c+CD206−(M1) and F4/80+CD11b+CD11c−CD206+(M2) ATMs from miR-99a- or miR-scramble-treated db/db mice by flow cytometry (n = 6). h Bar diagram represents quantification of figure g. Data are presented as the mean ± SEM of at least six mice per group from three independent experiments, *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 5.

Fig. 5

Open in a new tab

Overexpression of miR-99a regulates macrophage phenotype activation. Representative microscopic images of adipose tissue sections derived from db/+ and miR-99a- or miR-scramble-treated db/db mice. A CD86 staining depicting M1 macrophage and B CD301 staining depicting M2 macrophages, magnification; ×40, C hematoxylin and eosin (H&E) staining depicting adipocyte area, magnification; ×20, scale bars = 50 μm. The results are representative of adipose tissue sections from at least six animals from each group. D Dot plot represents labeling of F4/80+Ly6c+ ATMs by flow cytometry from miR-99a- or miR-scramble mimic-treated db/db mice (n = 6). Bar diagram showing the number of F4/80+Ly6c+ ATMs/gram eWAT shown in  D. Bar diagrams a, b, and c are quantifications of images A, B, and C, respectively, as described in the methods. Data are presented as the mean ± SEM of at least six mice per group from three independent experiments. **P < 0.01; ***P < 0.001 db/+ vs. miR-scramble db/db, and #P < 0.05; ###P < 0.001 miR-scramble- vs. miR-99a-treated db/db

miR-99a mimics improve systemic glucose homeostasis, insulin sensitivity, and dyslipidemia in diabetic mice

To assess whether miR-99a mimics also affect disease pathology, systemic glucose tolerance and insulin sensitivity were monitored in miR-scramble- or miR-99a-treated db/db mice (Fig. 6a–d). Administration of miR-99a in db/db mice significantly improved glucose tolerance (p < 0.01, Fig. 6a, b) and insulin sensitivity (p < 0.01, Fig. 6c, d) when compared to miR-scramble-treated db/db mice. The HOMA-IR index also decreased in miR-99a-treated db/db mice when compared to miR-scramble control (Fig. 6e). Moreover, random blood glucose levels were significantly lower in miR-99a-treated db/db mice than in miR-scramble control (Fig. S3A). At day 7, a significant decrease in the fasting glucose (Fig. S3B) and insulin levels (Fig. S3C) was observed in miR-99a-treated mice when compared to miR-scramble control. To check the effect of miR-99a on lipid parameters, a plasma lipid profile was generated. It was shown that miR-99a induced a significant decrease in the circulatory total cholesterol, HDL-cholesterol, and LDL-cholesterol levels (Fig. 6f, h) when compared to miR-scramble-treated db/db mice. However, the circulatory triglyceride levels and body weight of db/db animals were unaffected by miR-99a treatment (Fig. S3D and E).

Fig. 6.

Fig. 6

Open in a new tab

Overexpression of miR-99a regulates systemic glucose, insulin tolerance, and dyslipidemia. Post 7 days from the last miR-99a or miR-scramble injection, a intraperitoneal glucose tolerance test (ipGTT) and b the area under the curve (AUC) during ipGTT, c insulin tolerance test (ITT), d the area under the curve (AUC) during ITT, e HOMA-IR, f total cholesterol, g high-density lipoprotein (HDL), and h low-density lipoprotein (LDL). Data are presented as the mean ± SEM of 9–12 mice per group from three independent experiments. **P < 0.01, ***P < 0.001 db/+ vs. miR-scramble-treated db/db, and #P < 0.05; ##P < 0.01; ###P < 0.001; miR-scramble- vs. miR-99a-treated db/db

miR-99a mimics improve insulin signaling and regulate metabolism in diabetic mice

Inactivation of the insulin pathway is accompanied by an increased phosphorylation of IRS-1 at serine 307 (Ser-307)36 and a decreased phosphorylation of Akt at serine 473 (Ser-473). miR-99a mimics significantly reduced the phosphorylation of IRS1 at Ser-307 (Fig. 7a, b) and concomitantly enhanced the phosphorylation of AKT at Ser-473 (Fig. 7c, d) in the eWAT and liver of db/db mice when compared to miR-scramble control. In the liver of miR-99a-treated mice, the mRNA expression of gluconeogenesis (PEPCK, G6PC, and FOXO1) and cholesterol metabolism-associated genes (SREBP1 and LXRα) was significantly downregulated when compared to miR-scramble-treated mice (Fig. 7e, f). Similarly, in the eWAT of miR-99a-treated db/db mice, a significant decrease in the mRNA expression of the lipolysis gene HSL (~2.5-fold) was observed when compared to miR-scramble-treated db/db mice (Fig. 7g). However, genes associated with lipogenesis (ACC1 and FAS) were unaffected (Fig. 7g). Additionally, miR-99a mimics significantly enhanced adiponectin levels (~2.5-fold) in the eWAT when compared to miR-scramble-treated db/db mice (Fig. 7h).

Fig. 7.

Fig. 7

Open in a new tab

Overexpression of miR-99a improves insulin signaling and glucose and lipid metabolism in db/db mice. Adipose tissue and liver samples were collected from db/db mice treated with miR-99a or miR-scramble and western blotting was performed to detect the levels of a p-IRS-1 (Ser-307) and IRS-1 in eWAT. b p-IRS-1 (Ser-307) and IRS-1 in liver. c p-AKT (Ser-473) and AKT in eWAT. d p-AKT (Ser-473) and AKT in liver. Phosphorylation of IRS-1 and AKT was normalized to their respective total proteins. The bar diagram shows the quantification of the blots from at least three independent experiments. GAPDH was used as a loading control. mRNA expression of metabolic genes associated with e gluconeogenesis, f cholesterol metabolism, g lipogenesis and lipolysis, and h adiponectin level as assessed by qPCR (n = 6). Data are presented as the mean ± SEM of six mice per group from three independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001 vs. miR-scramble

Adoptive transfer of miR-99a-overexpressed macrophages reduces the M1 macrophage phenotype and adipose tissue inflammation in diabetic mice

To assess the role of macrophage specific miR-99a effect on macrophage phenotype, adipose tissue inflammation, and insulin resistance, adoptive transfer experiments were performed. db/db mice were administered miR-99a- or miR-scramble-overexpressing macrophages for six consecutive days (Fig. 8a). The adoptive transfer of macrophages containing overexpressed miR-99a into db/db mice induced a significant increase in the mRNA expression of YM-1, ARG-1, and PPARγ genes, and a significant decrease in the mRNA expression of iNOS, MCP-1, and IL-1β genes in the eWAT when compared to miR-scramble (Fig. 8b), indicating M2 macrophage phenotype activation. To further assess macrophage phenotype changes, quantification of F4/80+CD11b+CD11c+ CD206−(M1) and F4/80+CD11b+CD11c−CD206+(M2) ATMs from eWAT was assessed by flow cytometry. Adoptive transfer of macrophages containing overexpressed miR-99a to db/db mice induced a significant increase in the F4/80+CD11b+ CD11c-CD206+(M2) cell population, while a significant decrease in the F4/80+CD11b+ CD11c+CD206−(M1) ATMs was observed when compared to miR-scramble control (Fig. 8c, d).

Fig. 8.

Fig. 8

Open in a new tab

miR-99a-overexpressed macrophages modulate ATM phenotype. a Treatment regime of miR-99a- or miR-scramble-overexpressed macrophages and timeline for the measurement of parameters. b db/db mice were treated with miR-99a- or miR-scramble-overexpressed macrophages, and mRNA expression of M1 (iNOS, MCP-1, and TNF) and M2 (YM-1, ARG-1, and PPARγ) macrophage signature genes was determined in total eWAT by qPCR (n = 6). c Dot plot represents labeling of F4/80+CD11b+CD11c+CD206−(M1) and F4/80+CD11b+ CD11c−CD206+(M2) ATMs by flow cytometry in db/db mice treated with miR-99a- or miR-scramble-overexpressed macrophages (n = 6). d Quantification of figure c. Data are presented as the mean ± SEM of six mice per group from three independent experiments. **P < 0.01; ***P < 0.001

Adoptive transfer of miR-99a-overexpressed macrophages improves systemic glucose tolerance, insulin sensitivity, and dyslipidemia in diabetic mice

To assess whether adoptive transfer of miR-99a-overexpressing macrophages also affects disease pathology, the systemic glucose tolerance and insulin sensitivity were monitored in diabetic mice treated with miR-99a- or miR-scramble-overexpressed macrophages (Fig. 9a–d). Delivery of miR-99a-overexpressed macrophages to db/db mice significantly improved the glucose tolerance (p < 0.001, Fig. 9a, b), insulin sensitivity (p < 0.01, Fig. 9c, d), HOMA-IR index (p < 0.05, Fig. 9e), and random glucose (p < 0.01, Fig. S4A) when compared to miR-scramble. A significant decrease in the fasting blood glucose (p < 0.001, Fig. S4B) and insulin levels (p < 0.05, Fig. S4C) was observed in db/db mice treated with miR-99a-overexpressing macrophages when compared to miR-scramble control. Additionally, a significant decrease in the circulatory total cholesterol (p < 0.05, Fig. 9f), HDL-cholesterol (p < 0.05, Fig. 9g), and LDL-cholesterol (p < 0.05, Fig. 9h) was observed in db/db mice treated with miR-99a-overexpressed macrophages when compared to miR-scramble control. However, circulating triglycerides and the body weight (Fig. S4.D and E) of db/db animals treated with miR-99a-overexpressing macrophages were unaffected when compared to miR-scramble control.

Fig. 9.

Fig. 9

Open in a new tab

miR-99a-overexpressed macrophages regulate systemic glucose, insulin tolerance, and lipid metabolism. Analysis was performed 4 days after the adoptive transfer of miR-99a- or miR-scramble-overexpressed macrophages. a Intraperitoneal glucose tolerance test (ipGTT), b the area under the curve (AUC) during ipGTT. c Insulin tolerance test (ITT). d The area under the curve (AUC) during ITT. e HOMA-IR. f Total cholesterol. g High-density lipoprotein (HDL). h Low-density lipoprotein (LDL). Data are presented as the mean ± SEM of 9–12 mice per group from three independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001 db/+ vs. miR-scramble macrophage-treated db/db, and #P < 0.05; ##P < 0.01; ###P < 0.001; miR-scramble- vs. miR-99a macrophage-treated db/db

Adoptive transfer of miR-99a-overexpressed macrophages improves insulin signaling and regulates metabolism in diabetic mice

To evaluate the effect of miR-99a-overexpressing macrophages on insulin sensitivity, the insulin signaling cascade was assessed. Adoptive transfer of miR-99a-overexpressing macrophages in db/db mice significantly reduced the phosphorylation of IRS1 at Ser-307 (Fig. 10a, b) and enhanced AKT Ser-473 phosphorylation (Fig. 10c, d) in the eWAT and liver when compared to miR-scramble control. Hepatic mRNA expression of gluconeogenesis (PEPCK, G6PC, and FOXO1) and cholesterol metabolism-associated genes (SREBP1 and LXRα) was significantly downregulated in db/db mice that received miR-99a-overexpressing macrophages when compared to miR-scramble control (Fig. 10e, f). miR-99a-overexpressing macrophages induced a significant decrease in the eWAT mRNA expression of lipolysis gene HSL when compared to miR-scramble control (Fig. 10g). However, genes associated with lipogenesis (ACC1 and FAS) were unaffected (Fig. 10g). In the diabetic mice, miR-99a-overexpressing macrophages significantly enhanced adiponectin levels (~3-fold) in the eWAT when compared to miR-scramble control (Fig. 10h).

Fig. 10.

Fig. 10

Open in a new tab

miR-99a-overexpressed macrophages improve insulin signaling, glucose, and lipid metabolism in db/db mice. Adipose tissue and liver samples were collected from db/db mice treated with miR-99a- or miR-scramble-overexpressed macrophages, and western blotting was performed to detect the levels of a p-IRS-1 (Ser-307) and IRS-1 in eWAT. b p-IRS-1 (Ser-307) and IRS-1 in liver. c p-AKT (Ser-473) and AKT in eWAT. d p-AKT (Ser-473) and AKT in liver. Phosphorylation of IRS-1 and AKT was normalized to their respective total proteins. The bar diagram shows the quantification of the blots from at least three independent experiments. GAPDH was used as a loading control. mRNA expression of metabolic genes associated with e gluconeogenesis, f cholesterol metabolism, g lipogenesis and lipolysis, and h adiponectin levels, as assessed by qPCR (n = 6). Data are presented as the mean ± SEM of at least six mice per group from three independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001 vs. miR-scramble macrophage

Discussion

Several studies have established the importance of macrophages and their phenotype in the maintenance of metabolic homeostasis.37 The present study demonstrates the role of miR-99a in M1 macrophage phenotype activation by targeting TNFα. miR-99a mimics regulate adipose tissue inflammation and insulin resistance. A differential expression pattern of miR-99a was observed during M1 and M2 macrophage phenotype activation by LPS/IFN-γ and IL-4, respectively. The enhanced expression of miR-99a in M2- (IL-4) and M1 (LPS/IFN-γ)-derived M2 macrophages suggests a role of this miRNA in preventing M1 phenotype activation. Since miR-99a mimics in LPS/IFN-γ-stimulated macrophages increased M2 and decreased M1 markers, it can be speculated that miR-99a can inhibit M1 phenotype activation. miR-99a mimics also diminished the bactericidal activity of LPS/IFN-γ-stimulated BMDMs (Fig. 2d), indicating an inhibition of M1 macrophage function. Furthermore, silencing of miR-99a by miR-99 antagomir reduced the M2 phenotype, suggesting the specific role of this miRNA in M2 phenotype activation.

miRNAs act as a critical regulator of several biological processes by regulating the expression of their target genes.38,39 Four of the six computational algorithms used in the present study identified the 3’-UTR of TNFα as a probable target site of miR-99a. A decrease in the TNFα expression in miR-99a-overexpressing and LPS/IFN-γ-stimulated BMDMs suggests that TNFα is a probable target of miR-99a in macrophages. Furthermore, the reduced luciferase activity of TNFα 3ʹ-UTR in miR-99a-overexpressing cells confirmed that TNFα is a direct target of miR-99a. TNFα silencing in LPS/IFN-γ-stimulated BMDMs promoted the M2 phenotype, probably by reducing M1 phenotype activation since a decrease in mRNA expression of iNOS, MCP-1, and IL-1β was observed along with an enhanced expression of YM-1, ARG-1, and PPARγ. TNFα negatively regulates PPARγ expression and activity, which is a positive modulator of M2 phenotype activation.40,41 Therefore, silencing TNFα in M1 macrophages may indirectly promote M2 phenotype activation. Furthermore, the reduced bactericidal activity observed in terms of enhanced CFU formation in TNFα-silenced and LPS/IFN-γ-stimulated BMDMs demonstrated the significance of TNFα in attenuating M1 macrophage function. It can therefore be speculated that in BMDMs, miR-99a modulates M1 phenotype activation and function by targeting TNFα.

Previous studies have revealed that under diabetic conditions, ATMs preferentially undergo proinflammatory M1 phenotype activation, which regulates adipose tissue inflammatory responses.42 In concomitance with this notion, a reduced expression of miR-99a in eWAT and F4/80+ ATMs of diabetic mice established an inverse relation between miR-99a expression and adipose tissue inflammation (Fig. 5a, b). This correlated with the in vitro findings in the BMDMs.

Moreover, phenotypic switching of ATMs into the alternatively activated M2 is sufficient to modulate adipose tissue function.22,42,43 In the eWAT of diabetic mice, miR-99a mimics significantly attenuated F4/80+CD11b+CD11c+CD206−M1 and enhanced the F4/80+CD11b+CD11c−CD206+M2 macrophage populations, while significantly decreasing iNOS, MCP-1, and TNFα and significantly increasing ARG-1, YM-1, and PPARγ mRNA expression. In miR-99a-treated db/db mice, an attenuated CD86 and enhanced CD301 staining in eWAT again suggested the suppression of M1 macrophage and activation of the M2 phenotype. The beneficial effect of miR-99a mimics was also evident at the morphological level, since significant attenuation in adipocyte hypertrophy was observed when compared to miR-scramble mimics.

Macrophage recruitment in adipose tissue is a key feature of the diabetic condition.22 However, the number of F4/80+Ly6c+ ATMs did not differ significantly between miR-99a- and miR-scramble-overexpressed diabetic mice, indicating that macrophage infiltration was unaffected.

The therapeutic potential of miR-99a in diabetic mice was evident due to an improved glucose tolerance and insulin sensitivity and the alleviation of hyperglycemia, hyperinsulinemia, and hyperlipidemia after miR-99a mimics were administered when compared to miR-scramble mimics.

Inflammatory stimuli such as TNFα inhibits the action of insulin by facilitating the inhibitory phosphorylation of insulin receptor substrate-1 (IRS1) on Ser-307 in the diabetic state.36 Therefore, in the present study, it is quite possible that miR-99a improves insulin signaling by inhibiting TNFα. This hypothesis gains support from the observation that miR-99a mimics reduced inhibitory IRS-1 phosphorylation and induced hyperphosphorylation of AKT in the liver and eWAT when compared to miR-scramble control. In the adipose tissue, miR-99a mimics enhanced anti-inflammatory adiponectin, which may also complement M2 macrophage phenotype activation.

Impaired gluconeogenesis in the liver leads to hyperglycemia in type-2 diabetes.44,45 In diabetic mice, miR-99a mimics significantly reduced the hepatic mRNA expression of genes involved in glucose (PEPCK, G6PC, and FOXO1) and cholesterol (SREBP1 and LXRα) metabolism, thereby explaining the beneficial effects of this therapy on circulatory glucose and cholesterol levels. In the adipose tissue, miR-99a mimics can also inhibit lipolysis, since a significant decrease in HSL mRNA expression was observed. However, lipogenesis seems to be unaffected, since mRNA expression of ACC1 and FAS was unaffected in miR-99a mice.

Adoptive transfer of miR-99a-overexpressed macrophages in diabetic animals established an essential role of macrophage specific miR-99a in the disease pathology. Adoptive transfer of miR-99a-overexpressed macrophages in diabetic animals promoted M2 phenotype activation and improved systemic glucose levels, insulin sensitivity, diabetes-associated dyslipidemia, and insulin signaling.

The present study demonstrates the role of miR-99a in preventing M1 macrophage phenotype activation and adipose tissue inflammation. Mechanistically, miR-99a targets TNFα and inhibits M1 macrophage phenotype and function. The present study also demonstrates the modulation of M1 macrophage phenotype activation as one of the possible mechanisms by which miR-99a mimics alleviate adipose tissue inflammation and insulin resistance. Our study speculates the therapeutic potential of miR-99a in diabetes-associated metabolic dysfunction.

Electronic supplementary material

Supplementary Figure S1 (93.9KB, gif)
Supplementary Figure S2 (93.9KB, gif)
Supplementary Figure S3 (79KB, gif)
Supplementary Figure S4 (25.5KB, gif)
Supplementary Figure S5 (39.4KB, gif)
PRIMER TABLE (14.7KB, docx)

Acknowledgements

The authors gratefully acknowledge the technical help provided by Mr. A.L. Vishwakarma and Mr. C.P. Pandey. This work was supported by the CSIR-Network project: “Towards a holistic understanding of complex diseases: Unraveling the threads of complex disease (THUNDER)”. A.J. and M.M. are supported by UGC, New Delhi, P.M. is supported by CSIR, New Delhi, and S.S.R. is supported by THUNDER. CDRI Communication number: 9683

Competing interests

The authors declare no competing interests.

Electronic supplementary material

Supplementary Information accompanies for this paper at 10.1038/s41423-018-0038-7.

References

  • 1.Han MS, et al. JNK expression by macrophages promotes obesity-induced insulin resistance and inflammation. Science. 2013;339:218–222. doi: 10.1126/science.1227568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Orr JS, et al. Toll-like receptor 4 deficiency promotes the alternative activation of adipose tissue macrophages. Diabetes. 2012;61:2718–2727. doi: 10.2337/db11-1595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Eguchi J, et al. Interferon regulatory factor 4 regulates obesity-induced inflammation through regulation of adipose tissue macrophage polarization. Diabetes. 2013;62:3394–3403. doi: 10.2337/db12-1327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Odegaard JI, et al. Macrophage-specific PPARgamma controls alternative activation and improves insulin resistance. Nature. 2007;447:1116–1120. doi: 10.1038/nature05894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Nishimura S, et al. CD8+ effector T cells contribute to macrophage recruitment and adipose tissue inflammation in obesity. Nat. Med. 2009;15:914–920. doi: 10.1038/nm.1964. [DOI] [PubMed] [Google Scholar]
  • 6.Ji Y, et al. Activation of natural killer T cells promotes M2 Macrophage polarization in adipose tissue and improves systemic glucose tolerance via interleukin-4 (IL-4)/STAT6 protein signaling axis in obesity. J. Biol. Chem. 2012;287:13561–13571. doi: 10.1074/jbc.M112.350066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Scott LJ, et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science. 2007;316:1341–1345. doi: 10.1126/science.1142382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.O’Neill LA, Sheedy FJ, McCoy CE. MicroRNAs: the fine-tuners of Toll-like receptor signalling. Nat. Rev. Immunol. 2011;11:163–175. doi: 10.1038/nri2957. [DOI] [PubMed] [Google Scholar]
  • 9.Zhuang G, et al. A novel regulator of macrophage activation: miR-223 in obesity-associated adipose tissue inflammation. Circulation. 2012;125:2892–2903. doi: 10.1161/CIRCULATIONAHA.111.087817. [DOI] [PubMed] [Google Scholar]
  • 10.Yao F, et al. Adipogenic miR-27a in adipose tissue upregulates macrophage activation via inhibiting PPARgamma of insulin resistance induced by high-fat diet-associated obesity. Exp. Cell Res. 2017;355:105–112. doi: 10.1016/j.yexcr.2017.03.060. [DOI] [PubMed] [Google Scholar]
  • 11.Rupaimoole R, Slack FJ. MicroRNA therapeutics: towards a new era for the management of cancer and other diseases. Nat. Rev. Drug Discov. 2017;16:203–222. doi: 10.1038/nrd.2016.246. [DOI] [PubMed] [Google Scholar]
  • 12.Sun D, et al. miR-99 family of MicroRNAs suppresses the expression of prostate-specific antigen and prostate cancer cell proliferation. Cancer Res. 2011;71:1313–1324. doi: 10.1158/0008-5472.CAN-10-1031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Li Q, et al. Overexpression of microRNA-99a attenuates cardiac hypertrophy. PLOS One. 2016;11:e0148480. doi: 10.1371/journal.pone.0148480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Arner P, Kulyte A. MicroRNA regulatory networks in human adipose tissue and obesity. Nat. Rev. Endocrinol. 2015;11:276–288. doi: 10.1038/nrendo.2015.25. [DOI] [PubMed] [Google Scholar]
  • 15.Heneghan HM, Miller N, McAnena OJ, O’Brien T, Kerin MJ. Differential miRNA expression in omental adipose tissue and in the circulation of obese patients identifies novel metabolic biomarkers. J. Clin. Endocrinol. Metab. 2011;96:E846–E850. doi: 10.1210/jc.2010-2701. [DOI] [PubMed] [Google Scholar]
  • 16.Liao B, et al. MicroRNA cluster 302–367 enhances somatic cell reprogramming by accelerating a mesenchymal-to-epithelial transition. J. Biol. Chem. 2011;286:17359–17364. doi: 10.1074/jbc.C111.235960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Banerjee S, et al. MicroRNA let-7c regulates macrophage polarization. J. Immunol. 2013;190:6542–6549. doi: 10.4049/jimmunol.1202496. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Kadl A, et al. Identification of a novel macrophage phenotype that develops in response to atherogenic phospholipids via Nrf2. Circ. Res. 2010;107:737–746. doi: 10.1161/CIRCRESAHA.109.215715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Nasri M, Karimi A, Allahbakhshian Farsani M. Production, purification and titration of a lentivirus-based vector for gene delivery purposes. Cytotechnology. 2014;66:1031–1038. doi: 10.1007/s10616-013-9652-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Miller MR, Blystone SD. Reliable and inexpensive expression of large, tagged, exogenous proteins in murine bone marrow-derived macrophages using a second generation lentiviral system. J. Biol. Methods. 2015;2:e23. doi: 10.14440/jbm.2015.66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Li J, et al. Exosomes mediate the cell-to-cell transmission of IFN-alpha-induced antiviral activity. Nat. Immunol. 2013;14:793–803. doi: 10.1038/ni.2647. [DOI] [PubMed] [Google Scholar]
  • 22.Lumeng CN, Bodzin JL, Saltiel AR. Obesity induces a phenotypic switch in adipose tissue macrophage polarization. J. Clin. Invest. 2007;117:175–184. doi: 10.1172/JCI29881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Xiang Y, et al. MicroRNA-487b is a negative regulator of macrophage activation by targeting IL-33 production. J. Immunol. 2016;196:3421–3428. doi: 10.4049/jimmunol.1502081. [DOI] [PubMed] [Google Scholar]
  • 24.Jain M, Singh A, Singh V, Barthwal MK. Involvement of interleukin-1 receptor-associated kinase-1 in vascular smooth muscle cell proliferation and neointimal formation after rat carotid injury. Arterioscler. Thromb. Vasc. Biol. 2015;35:1445–1455. doi: 10.1161/ATVBAHA.114.305028. [DOI] [PubMed] [Google Scholar]
  • 25.Zheng L, et al. Differential microRNA expression in human macrophages with mycobacterium tuberculosis infection of beijing/W and non-Beijing/W strain types. PLOS One. 2015;10:e0126018. doi: 10.1371/journal.pone.0126018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Sun X, et al. MicroRNA-181b improves glucose homeostasis and insulin sensitivity by regulating endothelial function in white adipose Tissue. Circ. Res. 2016;118:810–821. doi: 10.1161/CIRCRESAHA.115.308166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Tiwari RL, et al. PKCdelta-IRAK1 axis regulates oxidized LDL-induced IL-1beta production in monocytes. J. Lipid Res. 2014;55:1226–1244. doi: 10.1194/jlr.M045658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Kaur K, et al. Elevated hepatic miR-22-3p expression impairs gluconeogenesis by silencing the wnt-responsive transcription factor Tcf7. Diabetes. 2015;64:3659–3669. doi: 10.2337/db14-1924. [DOI] [PubMed] [Google Scholar]
  • 29.Sun XJ, et al. Deletion of interleukin 1 receptor-associated kinase 1 (Irak1) improves glucose tolerance primarily by increasing insulin sensitivity in skeletal muscle. J. Biol. Chem. 2017;292:12339–12350. doi: 10.1074/jbc.M117.779108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Yokoyama H, et al. Quantitative insulin sensitivity check index and the reciprocal index of homeostasis model assessment in normal range weight and moderately obese type 2 diabetic patients. Diabetes Care. 2003;26:2426–2432. doi: 10.2337/diacare.26.8.2426. [DOI] [PubMed] [Google Scholar]
  • 31.Ramkhelawon B, et al. Netrin-1 promotes adipose tissue macrophage retention and insulin resistance in obesity. Nat. Med. 2014;20:377–384. doi: 10.1038/nm.3467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Kanuri BN, et al. Altered glucose and lipid homeostasis in liver and adipose tissue pre-dispose inducible NOS knockout mice to insulin resistance. Sci. Rep. 2017;7:41009. doi: 10.1038/srep41009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Kanematsu Y, et al. Critical roles of macrophages in the formation of intracranial aneurysm. Stroke. 2011;42:173–178. doi: 10.1161/STROKEAHA.110.590976. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Dweep H, Sticht C, Pandey P, Gretz N. miRWalk—database: prediction of possible miRNA binding sites by “walking” the genes of three genomes. J. Biomed. Inform. 2011;44:839–847. doi: 10.1016/j.jbi.2011.05.002. [DOI] [PubMed] [Google Scholar]
  • 35.Kroner A, et al. TNF and increased intracellular iron alter macrophage polarization to a detrimental M1 phenotype in the injured spinal cord. Neuron. 2014;83:1098–1116. doi: 10.1016/j.neuron.2014.07.027. [DOI] [PubMed] [Google Scholar]
  • 36.Lavin DP, White MF, Brazil DP. IRS proteins and diabetic complications. Diabetologia. 2016;59:2280–2291. doi: 10.1007/s00125-016-4072-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Liu YC, Zou XB, Chai YF, Yao YM. Macrophage polarization in inflammatory diseases. Int. J. Biol. Sci. 2014;10:520–529. doi: 10.7150/ijbs.8879. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Alvarez-Garcia I, Miska EA. MicroRNA functions in animal development and human disease. Development. 2005;132:4653–4662. doi: 10.1242/dev.02073. [DOI] [PubMed] [Google Scholar]
  • 39.Rayner KJ, et al. MiR-33 contributes to the regulation of cholesterol homeostasis. Science. 2010;328:1570–1573. doi: 10.1126/science.1189862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Ye J. Regulation of PPARgamma function by TNF-alpha. Biochem. Biophys. Res. Commun. 2008;374:405–408. doi: 10.1016/j.bbrc.2008.07.068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Zhang B, et al. Negative regulation of peroxisome proliferator-activated receptor-gamma gene expression contributes to the antiadipogenic effects of tumor necrosis factor-alpha. Mol. Endocrinol. 1996;10:1457–1466. doi: 10.1210/mend.10.11.8923470. [DOI] [PubMed] [Google Scholar]
  • 42.McNelis JC, Olefsky JM. Macrophages, immunity, and metabolic disease. Immunity. 2014;41:36–48. doi: 10.1016/j.immuni.2014.05.010. [DOI] [PubMed] [Google Scholar]
  • 43.Zou J, et al. CD4+ T cells memorize obesity and promote weight regain. Cell Mol Immunol. 2017;14:1–10. doi: 10.1038/cmi.2016.45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Basu R, Chandramouli V, Dicke B, Landau B, Rizza R. Obesity and type 2 diabetes impair insulin-induced suppression of glycogenolysis as well as gluconeogenesis. Diabetes. 2005;54:1942–1948. doi: 10.2337/diabetes.54.7.1942. [DOI] [PubMed] [Google Scholar]
  • 45.Kabir M, et al. Molecular evidence supporting the portal theory: a causative link between visceral adiposity and hepatic insulin resistance. Am. J. Physiol. Endocrinol. Metab. 2005;288:E454–E461. doi: 10.1152/ajpendo.00203.2004. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figure S1 (93.9KB, gif)
Supplementary Figure S2 (93.9KB, gif)
Supplementary Figure S3 (79KB, gif)
Supplementary Figure S4 (25.5KB, gif)
Supplementary Figure S5 (39.4KB, gif)
PRIMER TABLE (14.7KB, docx)

Articles from Cellular and Molecular Immunology are provided here courtesy of Nature Publishing Group

RESOURCES