miR-181a Induces Macrophage Polarized to M2 Phenotype and Promotes M2 Macrophage-mediated Tumor Cell Metastasis by Targeting KLF6 and C/EBPα.

Macrophages can acquire a variety of polarization status and functions: classically activated macrophages (M1 macrophages); alternatively activated macrophages (M2 macrophages). However, the molecular basis of the process is still unclear. Here, this study addresses that microRNA-181a (miR-181a) is a key molecule controlling macrophage polarization. We found that miR-181a is overexpressed in M2 macrophages than in M1 macrophages. miR-181a expression was decreased when M2 phenotype converted to M1, whereas it increased when M1 phenotype converted to M2. Overexpression of miR-181a in M1 macrophages diminished M1 phenotype expression while promoting polarization to the M2 phenotype. In contrast, knockdown of miR-181a in M2 macrophages promoted M1 polarization and diminished M2 phenotype expression. Mechanistically, Bioinformatic analysis revealed that Kruppel-like factor 6 (KLF6) and CCAAT/enhancer binding protein-α (C/EBPα) is a potential target of miR-181a and luciferase assay confirmed that KLF6 and C/EBPα translation is suppressed by miR-181a through interaction with the 3'UTR of KLF6 and C/EBPα mRNA. Further analysis showed that induction of miR-181a suppressed KLF6 and C/EBPα protein expression. Importantly, miR-181a also diminishes M2 macrophages-mediated migration and invasion capacity of tumor cells. Collectively, our results suggest that miR-181a plays a significant role in regulating macrophage polarization through directly target KLF6 and C/EBPα.

Introduction

Macrophages are derived from tissue-residing precursors or circulating monocytes. Macrophages residing in distinct tissue microenvironments can display divergent phenotypes and functions.[1] In response to various signals macrophages undergo polarization into classically activated (M1 or proinflammatory) or alternatively activated (M2 or anti-inflammatory) phenotypes. M1 macrophages can be generated in the presence of Toll like receptor ligands, such as IFN-γ and/or Lipopolysaccharides (LPS) and they exhibit potent antimicrobial properties and promote inflammation responses. Th2 cytokines such as IL-4 and IL-13 activated M2 macrophages which can suppress immune response and induce angiogenesis.[2,3] Transcriptional factors nuclear factor κB (NF-κB), CCAAT/enhancer binding protein-α (C/EBPα), PU.1, and IFN regulatory factor 5 (IRF5) participate in M1 activation, while STAT6, peroxisome proliferator–activated receptor-γ (PPARγ), CCAAT/enhancer binding protein-β (C/EBPβ), and Kruppel-like factor 4 (KLF4) are involved in the polarization of M2 phenotype.[4,5,6]

Macrophages are recruited into tumor site by multiple cytokines expressed in the tumor microenvironment and become tumor-associated macrophages (TAMs).[7] TAMs coexist in tumors and function as an accomplice to promote tumor progression and metastasis, especially once polarized into M2 phenotype by the tumor microenvironment.[8,9,10] TAMs stimulate cancer cell invasion, motility, and migration, and that these effects can be impaired by inhibiting expression of epidermal growth factor.[11] The shift of the TAMs phenotype from M2 toward M1 lead to a shift of plasma chemokine profiles toward tumor-attacking chemokines.[12] The histidine-rich glycoprotein inhibits tumor growth and metastasis, while improving chemotherapy by skewing TAM polarization away from the M2- to a tumor-inhibiting M1-like phenotype.[13] Thus regulating TAM polarization is a new perspectives for the design of new and more efficient therapeutic strategies to counteract cancer cell invasion. However, the molecular mechanisms underlying TAM polarization are unclear.

miRNAs are a class of noncoding small RNAs, which induce gene silencing by modulating gene expression at the post-transcriptional level. miRNAs processing machinery is expressed in most eukaryotic cells, including ~30–90% human genes, thereby indicating that miRNA regulation of gene expression is a widespread phenomenon. It plays essential roles in many cellular and developmental processes, including cell proliferation, apoptosis, and differentiation, as well as organ morphogenesis. Increasing evidence suggested that miRNAs are involved in inflammatory responses and regulate differential activation of macrophages.[14,15,16] miR-155 was reported to promote inflammatory reaction and induce M1 phenotype by targeting C/EBPβ and interleukin 13 receptor-α1.[17,18,19,20] miR-181a has been found to suppress inflammatory reactions by down-regulating IL-1α.[21] Recent studies have found that miR-181a differently expressed between M1 and M2.[22] However, the function of miR-181a in macrophage polarization is unknown.

In this study, we found that miR-181a is expressed at a higher level in M2 than in M1 macrophages. miR-181a suppresses polarization of macrophages to M1 phenotype and enhances M2 polarization through targeting the KLF6 and C/EBPα. Further, miR-181a could inhibit cancer metastasis and invasion, which induced by M2-type macrophages. Taken together, these findings suggest that the miR-181a plays a role in regulating macrophage plasticity and aided us in exploring the potential of therapy targets on miRNA capable of regulating the switching of the macrophages phenotype, resulting in remodeling of the tumor microenvironment.

Results

Establishment of human M1 and M2

To study macrophage polarization and plasticity, we chose human macrophage stimulated by LPS and IFN-γ (H-M1) or IL-4 (H-M2) as representative of the two opposite polarized states (M1 phenotype versus M2 phenotype). As expected, the mRNA expressions of CCL17, CCL22, and CD163, which are M2 macrophage markers, were significantly higher in M2 macrophages polarized by IL-4, and the mRNA expressions of M1 macrophage marker, TNFα, IL-1β, and HLA-DR were higher in M1 macrophages induced by LPS and IFN-γ (). At the protein level, CCL17 and CD163 were previously shown to have higher levels in M2 than in M1, and IL-1β and HLA-DR were higher in M1 than in M2 (). Besides, as shown in , M1 enhanced levels of CD86, a surface marker of M1 phenotype and reduced levels of CD206, a surface marker of M2 phenotype. Data from immunofluorescence experiments reveal that the fluorescence intensity of CD86 was decreased in M2 and CD206 was decreased in M1 (). Furthermore, M2 also released significantly higher levels of CCL17 and CCL22, and lower levels of TNFα and IL-1β ().

miR-181a is expressed at a higher level in M2 than in M1

Firstly, we detected the differential expression level of miR-181a between M1 and M2 by real-time polymerase chain reaction (PCR). We found that miR-181a levels are significantly up-regulated in H-M2 than in H-M1 (). This differential expression suggested that miR-181a may be a negative regulator of classical activation of M1 macrophages. To examine whether miR-181a contributes to the plasticity of macrophage polarization, we attempted to convert one population into another by culturing H-M1 macrophages with IL-4 and H-M2 macrophages with LPS and IFN-γ. As expected, the M2 macrophage marker, CCL17, CCL22, and CD163 were increased in M1–to–M2 conversion (), however, decreased in M2–to–M1 conversion (). M1–to–M2 conversion resulted in increased miR-181a (), whereas M2–to–M1 conversion led to miR-181a expression decreased (). These results explain that miR-181a was dynamically changing with macrophage phenotype: miR-181a decrease when M2 phenotype transform to M1, however, increase in higher M2 polarization degree. These data suggest that miR-181a may participate in macrophage polarization.

Knockdown of miR-181a in M2 diminishes the expression of M2 phenotypes

To further determine the role of miR-181a in macrophage polarization, H-M2 was transfected with miR-181a inhibitors or control inhibitors. miR-181a knockdown by miR-181a inhibitors enhanced the mRNA expression of TNFα, IL-1β, and HLA-DR, diminished expression of CCL17, CCL22, and CD163 in human macrophages (). Western blot analysis also revealed the higher level of IL-1β and HLA-DR and the lower level of CCL17 and CD163 after knockdowning miR-181a in M2 (). Besides, miR-181a knockdown in M2 substantially enhanced the percentage of CD86 positive macrophages and diminished the percentage of CD206 positive macrophages (). miR-181a inhibition in M2 resulted in the fluorescence intensity of CD86 significantly increased, however, fluorescence intensity of CD206 was declined (). Knockdown of miR-181a in M2 also diminished the release of TNFα and IL-1β, simultaneously, induced the release of CCL17 and CCL22 ().

As miR-181a is broadly conserved among vertebrates, an investigation was conducted to determine whether its functions are similar to that of murine macrophage cells. This was done by detecting the expression of M1/M2 markers in RAW246.7 derived M2 macrophage (M-M2) and mouse bone marrow-derived M2 macrophages (B-M2). As expected, the expressions of Arg-1, YM-1, TGFβ, and IL-10, which are M2 macrophage markers, were significantly lower after miR-181a knockdown, and the expressions of M1 macrophage marker, TNFα, IL-1β, and inducible nitric oxide synthase (iNOS) were higher (–).

These data further suggest that miR-181a participates in sustaining the M2 macrophage phenotype. Inhibition of miR-181a in M2 reduced the expression of M2 marker, transformed macrophages phenotype to M1.

Overexpression of miR-181a promotes M1 transition to the M2 phenotype

Because our experiments showed that knockdown of miR-181a in M2 reduces the expression of M2 phenotypes, we next determined if miR-181a overexpression in M1 also participates in macrophage plasticity by promoting the transition of M1 to M2 phenotype. In these experiments, M1 (H-M1, M-M1, and B-M1) was transfected with miR-181a mimics to induce overexpression of miR-181a. We found that expressions of TNFβ, IL-1β, HLA-DR, CD86, and iNOS in M1 transfected with miR-181a mimics was significantly less than that in M1 cells transfected with control mimics (–). Meanwhile, miR-181a overexpression in M1 reduced the expressions M2 markers: CCL17, CCL22, CD163, CD206, Arg-1, YM-1, TGFβ, and IL-10 (–). Collectively, these results indicate that miR-181a is a negative regulator of M1 macrophage phenotypes. Overexpression of miR-181a increased M2 markers, drive the transition of M1 toward the M2 phenotype.

miR-181a downregulates C/EBP-α and KLF6 and their downstream pathway genes

To delineate the mechanism by which miR-181a regulates macrophage polarization, we searched predicted targets of human miR-181a through miRBase and TargetScan. We found that the 3′UTR of KLF6 and C/EBPα gene containing potential miR-181a binding sites, the two genes participates in macrophages polarization (). Through RNAfold software, we found that the minimum free energy between human miR-181a and the putative binding sites in the 3′-UTR of KLF6 and C/EBPα mRNA were <−20, which suggested that miR-181a may target KLF6 and C/EBPα via binding these putative sites (). KLF6 has been shown to promote inflammatory in macrophage polarization by inhibiting PPARγ, an important transcriptional factor in M2 polarization.[23] C/EBPα has been reported to induce M1 macrophage activation.[24,25] To determine whether KLF6 and C/EBPα could participate in macrophage polarization, we evaluate KLF6 and C/EBPα mRNA expression in M1 and M2. We found that KLF6 and C/EBPα both have a higher expression in M1 (). To confirm whether miR-181a targets KLF6 and C/EBPα, we respectively cloned the 3′UTR of KLF6 including the positions 314–938 bp (KLF6-1) and 2,288–3,080 bp (KLF6-2), meanwhile, we cloned the 3′UTR of C/EBPα into luciferase reporter and cotransfected it with control or miR-181a mimics. As shown in , compared with the negative control, miR-181a downregulated luciferase activity of the reporter that contained the 3′UTR of KLF6 or C/EBPα. Conversely, luciferase activity has little change in cells transfected with miR-181a and mutant 3′UTR reporter plasmids. These results suggest that miR-181a directly targets KLF6 and C/EBPα.

To confirm whether miR-181a represses the expression of KLF6 and C/EBPα in macrophages, we transfected H-M1 with miR-181a mimics and transfected H-M2 with miR-181a inhibitors and found that KLF6 and C/EBPα expressions were significantly suppressed by the miR-181a mimic at 24 hours (), while the transfection of miR-181a inhibitors led to a significant increase in expression of KLF6 and C/EBPα (). We next explore whether miR-181a could regulate the downstream pathway genes of KLF6 and C/EBPα. PPARγ was downstream pathway gene of KLF6, which promote H-M2 polarization. KLF6 was found to reduce PPARγ expression and activity.[23] PU.1 was downstream pathway gene of C/EBPα. C/EBPα was reported to cooperate with PU.1 to promote M1 polarization.[24] We transfected miR-181a mimic into H-M1, and found that the protein expression of PPARγ was increased and PU.1 was suppressed (). In agreement with the above results, knockdown of miR-181a in M2 resulted in a significantly higher level of PU.1 and lower level of PPARγ ().

To gain further evidence that miR-181a regulates macrophage polarization through targeting KLF6 and C/EBPα, we used siRNA against KLF6 and C/EBPα to assay its function. As shown in , transfection of siRNA KLF6 into H-M1 effectively downregulated KLF6 in protein levels. Meanwhile, the protein expression of IL-1β was significantly suppressed and CD163 was upregulated. When we transfected siRNA C/EBPα into H-M1, the protein expression of C/EBPα and IL-1β was significantly downregulated and CD163 was increased. When the two entioned siRNAs were used to transfect H-M1, we found that the inhibition of KLF6 and C/EBPα significantly blocked the IL-1β expression and induced the CD163 production. These results indicate that knockdown of KLF6 and C/EBPα may trigger M2 polarization.

Together, these findings indicate that miR-181a can bind to the 3′-UTR of KLF6 and C/EBPα to decrease the expression and regulate their downstream pathway. Furthermore, knockdown of KLF6 and C/EBPα can drive the transition of M1 toward the M2 phenotype.

miR-181a regulates M2 mediated tumor cell migration and invasion

We have shown that miR-181a suppresses M1 macrophage polarization and promotes M2 macrophage activation. Next, we investigated whether miR-181a regulates cellular functions associated with the M1 and M2 phenotypes. It was previously shown that M2 macrophages possess greater activity to promote cancer cells invasion, metastasis and Epithelial–Mesenchymal Transition (EMT) than do M1 macrophages.[26,27,28] We assessed cancer cells invasion and metastasis abilities and expression of EMT markers when conditional culture with M2 transfected with miR-181a inhibitor. To this end, three human tumor cell lines, colorectal carcinoma cell line (HCT116), breast adenocarcinoma cell line (MCF-7), and ovarian carcinoma cell line (OVCAR3), were used. Transwell migration assays indicated that miR-181a promote tumor cell migration by inducing M2 polarization. The migration property of MCF-7, OVCAR3, and HCT116 cells was suppressed 31.5 ± 6.5, 37.0 ± 4.5, and 29.5 ± 6.3% when conditional cultured with M2 after being transfected with miR-181a inhibitor (). Meanwhile, transwell invasion assays clearly showed that the invasion capacity of HCT116, MCF-7, and OVCAR3 was also attenuated when inhibiting miR-181a expression in M2 (). The invasion property of MCF-7, OVCAR3, and HCT116 cells was suppressed 25.9 ± 8.2, 30.0 ± 2.4, and 26.0 ± 8.6%. It indicates that transfection of miR-181a inhibitor into M2 inhibits the invasion and metastatic properties of tumor cells. showed that knockdown of miR-181a in M2 enhanced E-cadherin protein expression, whereas decreased vimentin protein expression in three tumor cell lines, compared with transfecting with NC-i in M2.

TAMs mediated production of matrix metalloproteinase (MMP)2, MMP9, and uPA is a major phenotype of invasion of tumor cells. To examine whether those recognized soluble factors were regulated by miR-181a in the conditional media of M2 macrophages, we examine MMP2, MMP9, and uPA expression by enzyme-linked immunosorbent assay (ELISA). We found that MMP2, MMP9, and uPA levels are significantly down-regulated in M2 which knockdowned miR-181a ().

These data are consistent with the findings that miR-181a attenuated M2 activation of macrophages. Furthermore, knockdown miR-181a inhibits M2 mediated production of invasion associated cytokines.

Discussion

The biological role of miR-181a in the macrophage has not been clearly described. In these experiments, we for the first time found that miR-181a promotes M2 macrophage polarization and suppresses M1 polarization. miR-181a is at a higher level in M2 than in M1. Furthermore, when M2 were converted to M1, the levels of miR-181a were significantly decreased. In contrast, when M1 were converted to M2, the levels of miR-181a were significantly increased. Transfection of macrophages with miR-181a mimic resulted in up-regulation of markers and cytokines associated with the phenotype of the M2 macrophages, whereas cytokines and markers associated with the phenotype of M1 macrophages were downregulated. These data suggest that miR-181a skews their polarization from an M1 toward an M2 phenotype. Although many miRNAs can regulate inflammatory response in macrophages, only a few miRNAs are shown to participate in both M1 and M2 macrophage polarization. Previously, there was an effort to systemically identify miRNAs that change of expression in differentially polarized macrophages (miR-181a, miR-155-5p, miR-204-5p, miR-451, miR-125-5p, miR-146a-3p, miR-143-3p, and miR-145-5p).[29] Among these miRNAs, some of them have been identified to have regulatory roles on macrophages polarization. miR-155 was found to promote M1 phenotype by targeted suppressing key genes of M2 phenotype, C/EBPα and interleukin 13 receptor alpha1.[19,20] miR-125-5p and miR-146a were found to promote M2 polarization.[30,31] This study also showed that miR-181a promotes macrophage conversion from M1 to M2. Therefore, miRNAs play an important role in regulating macrophage polarization.

Previous studies have reported that miR-181a regulate inflammation responses through inhibiting inflammatory factors levels. miR-181a directly targeted the 3′UTR of IL-1α and down-regulated IL-1α levels in monocytes and macrophages.[21] In addition, miR-181a targeted inhibits TNF expression.[32] It has been reported to regulate TLR-NF-κB signaling in monocytes.[33] Those above-mentioned inflammatory cytokines are all M1 markers.

Transcriptional regulation is an important approach to regulate polarization of macrophages. PU.1 is a transcription factor and activated by TLRs to induce inflammatory.[34] C/EBPα is the upstream of PU.1, which directly activates PU.1 gene transcription. C/EBPα and NF-κB cooperatively induce numerous genes during the inflammatory response.[35] At the same time, C/EBPα and PU.1 have been found to participate in TLR ligand induced M1 activation.[24] PPARγ is a key transcription factor involved in the polarization of M2 macrophages. PPARγ activation significantly increases expression of the M2 marker ARG-1, IL-10, and MR.[36] PPARγ also inhibits inflammatory by suppressing the activity of NF-κB.[37] KLF6 has been shown to promote inflammatory macrophage polarization by inhibiting PPARγ.[23] Mechanistically, we further demonstrate that miR-181a regulates macrophage phenotype through combinatorial targeting of C/EBP-α and KLF6 genes. Several lines of evidence in our study support a direct regulation of C/EBP-α and KLF6 by miR-181a: first, overexpression of miR-181a downregulates C/EBP-α and KLF6 in macrophages at both mRNA and protein levels; second, the 3′UTR in C/EBP-α and KLF6 transcripts contains a miR-181a binding site; and third, the C/EBP-α and KLF6 3′UTR is responsive to miR-181a regulation. In our data, knockdown of miR-181a was found to induce KLF6 expression while PPARγ was reduced. Furthermore, miR-181a knockdown promotes both C/EBPα and PU.1 expression. Collectively, these findings raise the possibility that miR-181a regulates macrophage polarization by targeting KLF6 and C/EBPα 3′UTR and regulating PPARγ and PU.1 expression. miR-181a has stronger binding capacity with C/EBPα than KLF6 and has stronger inhibitory effect on C/EBPα. However, which one has a stronger ability on macrophage polarization needs further research.

EMT plays an important role in tumor invasion and metastasis.[38] In addition to EMT, MMP2, MMP9, and uPA are perceived to play a key role in cancer cell invasion and metastasis.[39,40] A high level of infiltration of TAMs was associated with EMT-related proteins in human GC tissues.[41] TAMs mediated production of MMP2, MMP9, and uPA are major phenotype of invasion of tumor cells.[42,43] These studies suggest that TAMs are recruited to the tumor site, promoting tumor metastasis and invasion. There is increasing evidence that TAMs show M2-like macrophage and the transformation from M1 to the M2 phenotype in TAMs is a critical event for tumor promotion.[9] Therefore, the phenotypical reversion of M2-like TAMs may be used as a therapeutic target in tumors.

Our experiments demonstrated that miR-181a not only regulates the expression of M2 macrophage markers, but also controls macrophage functions associated with the M2 states. When miR-181a was inhibited in M2, the facilitated capacity of the migration and invasion of cancer cells (HCT116, MCF-7, and OVCAR3) was attenuated to varying degrees. In addition, knockdown miR-181a in M2 suppressed the EMT and the expressions of MMP2, MMP9, and uPA of tumor cells. All of these data suggest that regulation of miR-181a could impair the ability of M2 macrophages to induce metastasis and invasion of tumor cells. The aberrant expression of human miR-181a has been implicated in the pathogenesis of various cancers, serving as an oncogene or tumor suppressor. miR-181a is associated with tumor progression and metastasis. High miR-181a expression is correlated to shorter time to recurrence and poor outcome in many cancers, such as epithelial ovarian cancer, breast cancer, and colorectal cancer.[44,45,46] MiR-181a as a TGF-β–regulated target that enhanced the metastatic potential of cancers by promoting EMT, migratory, and invasive phenotypes.[47] The results obtained here suggest that, at least in part, miR-181a exerted promoting migratory, and invasive phenotypes effect by regulating macrophages polarization. Therefore, targeting miR-181a in M2 or cancer cell may have potential utility in the treatment of cancer.

In summary, we provide a new understanding of the roles miR-181a has in macrophages by regulating the phenotype of macrophages through targeting the KLF6 and C/EBPα and other genes in their downstream signaling pathway. The suppression of miR-181a promotes transformation of M2 to M1 and results in the inhibition of migration and invasion of tumor cells. As polarization has been shown to be important in macrophages' functions, any dysregulation of macrophages differentiation state will lead to pathologic conditions. Thus, miR-181a can potentially be used in cancer immunotherapy and for other threatening diseases through modulation of macrophage activity.

Materials and methods

Cell lines. Human monocyte cell lines (THP-1), mouse macrophage cell lines (RAW264.7), human breast cancer cell lines (MCF-7), human ovarian cancer cell lines (OVCAR3), and human colorectal cancer cell lines HCT116 were purchased from Cell Resource Center of the Shanghai Institute for Biological Sciences, Chinese Academy of Sciences, China. THP-1 was grew in suspension cultures and cultured in RPMI (Roswell Park Memorial Institute) 1640 (HyClone, Logan, UT) supplemented with 10% fetal bovine serum (FBS; HyClone), antibiotics (100 U/ml penicillin and 0.1 mg/ml streptomycin), and 1% N-2-hydroxyethylpiperazine-N9-2-ethanesufonic acid (HEPES) . RAW264.7 was adherent cultured in Dulbecco's modified essential medium (DMEM) media (HyClone) containing 10% FBS and antibiotics. MCF-7, OVCAR3, and HCT116 were adherent cultured in RPMI 1640 medium with 10% FBS and antibiotics. All above cells were incubated in a humidified atmosphere 5% CO2 at 37°C.

Primary macrophages were derived from bone marrow cells of C57BL/6 mice. In brief, bone marrow cells were cultured in Dulbecco's modified essential medium media containing 10% FBS and 50 ng/ml M-CSF (Peprotech, Rocky Hill, NJ) for 5 days to establish mouse bone marrow-derived macrophages. Bone marrow-derived macrophages were incubated in a humidified atmosphere 5% CO2 at 37°C.

Generation of M1 and M2 macrophage. THP-1 cells attached to the plate bottom differentiated from human monocytes to human macrophages (H-MÖ) after Phorbol-12-myristate-13-acetate (Sigma-Aldrich, Santa Clara, CA) stimulation at a final concentration of 200 ng/ml for 24 hours. Then H-MÖ, RAW264.7, and bone marrow-derived macrophages were cultured in respective media supplemented with 10% FBS, and containing LPS (100 ng/ml) (Sigma-Aldrich)+IFN-γ (100 ng/ml) (Peprotech) or IL-4 (20 ng/ml)(Peprotech), to generate M1-polarized macrophages (H-M1, M-M1, and B-M1) or M2-polarized macrophages (H-M2, M-M2, and B-M2), respectively. Polarized macrophages were stimulated for 24 hours.

Cell transfection. Macrophages were transfected with miRNA mimics, inhibitors or small interfering RNAs (siRNAs) using Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instruction. Human and murine miR-181a mimics, miR-181a inhibitors and their negative control miRNAs (scrambled control oligonucleotides) were purchased from RIBOBIO (Guangzhou, China). KLF6 siRNA, C/EBPα siRNA and Control siRNA were purchased from RIBOBIO (Guangzhou).

M1 were transfected with 2 nmol/l (final concentration) miR-181a mimics (181a-m) or negative control mimics (NC-m), M2 were transfected with 20 nmol/l miR-181a inhibitors (181a-i) or control inhibitors (NC-i) for 24 hours. KLF6 siRNA, C/EBPα siRNA or control siRNA were transfected into H-M1 with 50 nmol/l (final concentration) for 48 hours.

Preparation of conditional medium. After transfected with miR-181a inhibitors (181a-i) or control inhibitors (NC-i) for 24 hours, the H-M2 macrophages were further cultured in RPMI 1640 without FBS for 24 hours. H-M2 macrophages without transfection were cultured for 24 hours in RPMI 1640 with 10% FBS, and then cultured in RPMI 1640 without FBS for another 24 hours to be blank control. The supernatant was centrifuged at 2,000×g for 15 minutes and collected to be used as conditional medium in the following experiments.

Quantitative real-time PCR. Total RNA was extracted using the Trizol reagent kit (CWBIO, Beijing, China) according to the manufacturer's directions and converted to complementary DNA (cDNA) using the PrimeScript RT-PCR Kit (Takara, Dalian, China) primed with oligo (dT). The reaction condition was as follows: 37°C, 15 minutes; 85°C, 5 seconds; and 95°C, 5 minutes. miRNA was extracted from cells with miRNA Purification Kit (CWBIO). miRNA was reverse transcribed into cDNA using M-MLV reverse transcriptase (Takara). The reaction condition was as follows: 16°C, 30 minutes, 42°C, 1 hour, and 85°C, 5 minutes. The PCR primers of U6 and miR-181a were purchased from RIBOBIO. Quantitative real-time PCR was carried out using the SYBR Green PCR Mix Kit (Takara). The reaction condition was followed by 95°C, 10 seconds, 56°C, 45 seconds, and 72°C, 20 seconds for 40 cycles of amplification. Primer sequences are detailed in and . Assays were made in triplicates, and results were normalized according to the expression levels of β-actin mRNA or U6 miRNA. Results were expressed using the ▵▵CT (cycle threshold) method for quantification.

Western blotting. Cells were lysed in Radio Immunoprecipitation Assay (RIPA) lysate buffer, and the cell lysates were incubated on ice for 30 minutes and centrifuged at 13,000×g at 4°C for 15 minutes before the supernatant was collected. All above cell supernatant was collected and protein concentrations were determined using a bicinchonininc acid (BCA) protein quantitation kit (Beyotime, Jiangsu, China). Proteins (50 μg) from each cell lysate were subjected to SDS-PAGE electrophoresis and transferred to PVDF membranes. Membranes were blocked with 1% BCA Fraction V for 1 hour, and incubated with primary antibody overnight at 4°C. Antibody against IL-1β (rabbit antihuman polyclonal antibody, 1:400 dilution), β-actin (mouse antihuman and antimouse polyclonal antibody, 1:1,000 dilution) were purchased from BOSTER (Wuhan, China). Antibody against CD163, (mouse antihuman polyclonal antibody, 1:1,000 dilution), HLA-DR, CCL17, KLF6, and C/EBPα (rabbit antihuman polyclonal antibody, 1:1,000 dilution) were purchased from abcam (Cambridge, UK). Antibody against PPARγ (rabbit antihuman monoclonal antibody, 1:800 dilution) and PU.1 (rabbit antihuman monoclonal antibody, 1:1,000 dilution) were purchased from Cell Signaling Technology (Boston, USA). TGFβ (rabbit antimouse polyclonal antibody, 1:1,000 dilution) and TNFα (rabbit antimouse polyclonal antibody, 1:1,000 dilution) were purchased from BIOSS (Beijing, china). Membranes were then incubated with horseradish peroxidase-linked goat anti-rabbit secondary antibodies or goat antimouse secondary antibodies (1:4,000 dilution, Santa Cruz, CA) at room temperature for 2 hours, and detected with a chemiluminescent detecting system (Amersham, Freiburg, Germany).

ELISA. Macrophage supernatants were tested for the presence of cytokines and chemokines using commercially available ELISA for human IL-1β, TNFα, CCL22, CCL17, MMP2, MMP9, and uPA, for mouse IL-10 and TNFα (all from BOSTER, Wuhan, China) following the protocols supplied by the manufacturers.

Flow cytometry assay. Macrophages were trypsinized and suspended in phosphate-buffered saline, followed by incubation with 1 μg/ml fluorescein isothiocyanate-conjugated human anti-CD206 (BD Bioscience, San Jose, CA) or phycoerythrin-conjugated human anti-CD86 (BD Bioscience) monoclonal antibody for 30 minutes at 4°C in the dark. The samples were subjected to flow cytometry analysis within 1 hour.

Immunofluorescence. Macrophages were 4% formaldehyde fixed (10 minutes) and then incubated in 10% normal goat serum for 1 hour. The cells were then incubated with the primary antibodies: rabbit antihuman CD206 (1:1,000 dilution, abcam, Cambridge, UK) and mouse antihuman CD86 (1:100 dilution, Santa Cruz, CA) overnight at 4°C. The secondary antibodies DyLight 488 goat antirabbit IgG (H + L) and Alexa Fluor 594 WGA were used at a 1:200 dilution for 1 hour. 4',6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI) was used to stain the nucleus at a concentration of 100 ng/ml.

Luciferase reporter assay. The 3′-UTR of human KLF6 gene contains two potential miR-181a-targeting sites at the positions 403–409 bp and 2,955–2,962 bp. Luciferase reporter vector pmiR-RB-report plasmid (RIBOBIO) contains the positions 314–938 bp and 2,288–3,080 bp respectively. The resulting construct was designated as pmiR-RB-report-KLF6-1 and pmiR-RB-report-KLF6-2. The 3′UTR of human C/EBPα gene (997–1,003 bp) was cloned into pmiR-RB-report plasmid (RIBOBIO). Site-directed mutagenesis plasmid of the miR-181a target-sites in the 3′-UTR of KLF6-1 (314–938 bp), KLF6-2 (2,288–3,080 bp) and C/EBPα (703–1,331 bp) were purchased from Genechem (Shanghai, China). 293 T-cells were cotransfected with 100 ng (final concentration) wt or mutant reporter plasmid and 20 nmol/l miR-181a mimic (final concentration) using Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA). Firefly and Renilla luciferase activities were measured using Dual Luciferase Reporter Gene Assay kit (Promega, WI). Relative luciferase activity normalized to the negative control was used for comparison among groups.

In vitro migration assays and invasion assays. Assays were performed using 8-μm pore size FalconTM cell culture inserts (BD Biosciences) in a 24-well format according to the vendor's instructions. In the migration assay, MCF-7, OVCAR3, and HCT-116 cells (104 cells/well) were seeded onto membranes of the upper chambers with 0.5 ml serum-free medium, which had been inserted into wells of 24-well plates containing 10% FBS-supplemented conditional medium. After 24 hours, cells were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet for 30 minutes. Un-migrated cells that remained in the upper chambers were removed by wiping the top of the insert membranes with a damp cotton swab, which left only those cells that had migrated to the underside of the membranes. The membranes were mounted on glass slides, and numbers of cells in three randomly chosen high-power fields were counted. For the invasion assay, MCF-7, OVCAR3, and HCT-116 (105 cells/well) were seeded onto Matrigel-coated membranes (BD Biosciences) of the upper chambers and incubated at 37°C. The lower chambers contained the same amount of berberine in 10% FBS conditional medium. After 48 hours, noninvasive cells remaining on the upper surface of the membranes were removed with a cotton swab. Cells on the lower surface of the membrane were fixed in 4% paraformaldehyde and stained with 0.5% crystal violet for 30 minutes. Membranes were mounted on glass slides, and numbers of cells in three randomly chosen high-power fields were counted. All experiments were performed three times and photographed under a phase-contrast microscope.

Statistical analysis. Statistical analysis was performed using SPSS16.0 statistical software (SPSS, Chicago). Values were expressed as mean ± SD. One-way analysis of variance was performed for multiple group comparisons. A value of P < 0.05 was used as the criterion for statistical significance.

Table 1

List of polymerase chain reaction (PCR) primers of human used in the study

Table 2

List of polymerase chain reaction (PCR) primers of murine used in the study