Background
Immune cells infiltrate all neoplastic lesions, and together, the immune cells and tumor cells constitute the tumor microenvironment. Such immune cells were previously thought to function in the defense response against the tumor [
1]. However, recently, increasing evidence indicates that tumor-associated inflammatory cells may enhance tumor progression, and among these cells, macrophages play the most important role [
2]. Macrophages can alter their profiles such that they become M1 or M2 macrophages according to the tumor microenvironment. It has been shown that M2 macrophages are a type of tumor-associated macrophage (TAM) [
3]. TAMs can promote tumor cell growth and metastasis, and recent research has indicated that TAMs can stimulate colorectal cancer cell invasion by upregulating matrix metalloproteinase (MMP) expression and activating epidermal growth factor receptor (EGFR) [
4].
Phosphatase of regenerating liver (PRL-3), a protein tyrosine phosphatase, has been demonstrated to play an important role in colorectal cancer progression and metastasis [
5]. PRL-3 is significantly elevated (>90%) in metastases and moderately elevated (25–45%) in primary colorectal cancer tumor. Moreover, the expression of PRL-3 in primary tumors indicated their tendency toward liver metastasis [
6]. Our previous studies have demonstrated that PRL-3 can promote the proliferation and metastasis of tumor cells through the autocrine secretion of tumor necrosis factor-alpha (TNF-α), which induces intermediate-conductance Ca
2+-activated K
+ (KCNN4) channel expression by activating the NF-κB signaling pathway [
7]. Previous studies also revealed that TNF-α contributed to tumor progression in a paracrine manner. Because TNF-α secreted from tumor cells and/or macrophages can affect the phenotype of these cells in a paracrine and/or autocrine manner, we hypothesize that colorectal cancer cells may interact with TAMs in the microenvironment and alter the cytokine profile of TAMs to promote tumor progression and metastasis through TNF-α, the secretion of which is stimulated by PRL-3 in a paracrine manner.
Interleukin-6 (IL-6) is a potent pleiotropic cytokine that is predominantly produced by monocytes and macrophages during chronic inflammation [
8]. IL-6 has been shown to be involved in tumor progression and metastasis through STAT3 signaling pathways [
9]. Moreover, the level of IL-6 is positively correlated with poor prognosis in different cancers. In addition to IL-6, interleukin-8 (IL-8) is also a known proinflammatory cytokine [
10]. Extensive studies have demonstrated that the levels of IL-8 and its receptor CXCR2 are significantly increased in colorectal cancer (CRC) cells, and that these proteins play an important role in tumor development [
11]. Similarly, the expression of IL-8 is correlated with tumor size and tumor stage. However, the question of whether IL-6 and IL-8 are involved in the metastasis of CRC induced by PRL-3 remains unclear.
In this study, we aimed to investigate whether TAMs participate in the metastasis of CRC, which is induced by PRL-3 in the tumor microenvironment. Our study revealed that PRL-3 could induce the expression of IL-6 and IL-8 secreted by TAMs through TNF-α released by CRC cells in a paracrine manner. Further study also revealed that such regulation could be inhibited by blocking KCNN4 channels expressed by TAMs.
Methods
Reagents and antibodies
G418, PMA, and Lipofectamine2000 were purchased from Sigma (St Louis, Missouri, USA). Fetal bovine serum (FBS) was purchased from BioInd (Kibbutz Beit Haemek, Israel). RPMI was purchased from Invitrogen (Carlsbad, CA, USA). Trizol and Prime Script RT were purchased from Takara (Dalian, China). Matrigel matrix was purchased from BD Biosciences (Biosciences, Bedford, MD,USA). siRNA was purchased from GenePharma (Shanghai, China). Antibodies against GAPDH (cat: ab8245), KCNN4 (cat: ab83740), p50 (cat: ab7971), and p65 (cat: ab7970) were purchased from Abcam(Cambridge, MA, USA). Antibodies against CD68 (cat: sc-393951) and CD206 (cat: sc-376108) were purchased from Santa Cruz (Santa Cruz, CA, USA). Antibodies against IL-6 (cat: 1457–1) and IL-8 (cat: 3518–1) were purchased from Epitomics (Burlingame, CA, USA).
Cell cultures and treatment
LoVo cells were purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences and then transfected with PAcGFP-PRL-3 (LoVo-P) or PAcGFP (LoVo-C) using Lipofectamine2000. Stable clones were selected by culturing with 600 ug/ml G418 for 3 weeks. THP-1 cells were obtained from Shanghai Cell Bank of the Chinese Academy of Sciences. Both LoVo and THP-1 cells were cultured in RPMI 1640, supplemented with 10% fetal bovine serum (FBS), 100 mg/ml penicillin, and 100 mg/ml streptomycin. The cells were incubated at 37°C, 5% CO2 in a humidified atmosphere. M2-polarized THP-1 cells were generated by phorbol myristate acetate (PMA) treatment (320 nM/106 cells) for 6 h followed by incubation with IL-4 (20 ng/ml) for 18 h.
Samples and patients
CRC samples were obtained from 71 patients, who were admitted to the Department of Gastroenteropancreatic Surgery of Sun Yat-sen Memorial Hospital, Sun Yat-sen University from 2005 to 2007. Surgically resected specimens were collected immediately after tumor removal. All samples were collected with informed consent according to the Internal Review and the Ethics Boards of the Sun-Yat-Sen Memorial Hospital of Sun-Yat-Sen University. The protocol was approved by the Ethics Committee of Sun Yat-Sen Memorial Hospital.
Flow cytometry
Cells were washed in phosphate buffered saline (PBS), resuspended and then stained with murine anti-human CD68 or CD206 for 30 min, then washed and incubated with PE-conjugated goat anti-mouse secondary antibody. Cells were analyzed by flow cytometry (BD FACS Cabiler).
M2 macrophage and LoVo cell coculture
1 × 106 M2 macrophage cells were seeded into six-well plates, while LoVo cells were cocultured with M2 macrophages in upper transwell inserts. After coculturing, LoVo and M2 macrophage cells were washed and used for later experiments.
Enzyme linked immunosorbent assay (ELISA)
TNF-α was assayed in the culture supernatant of LoVo-P or LoVo-C cells using the Quantikine Kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s protocol.
Western blot assay
Cells were lysed on ice with RIPA buffer containing 1% PMSF. Sample protein concentration was determined by the Bradford assay. Denatured proteins were separated by 10% or 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to polyvinylidene fluoride membranes and then blocked in 5% non-fat milk. Membranes were washed 3 times with Tris-buffered saline + 0.1% Tween-20 (TBST), incubated with relevant primary antibodies overnight at 4°C, washed and incubated for 1 h at room temperature with horseradish peroxidase-conjugated secondary antibodies. Labeled proteins were visualized by chemiluminescence.
siRNAs targeting human p50 cDNA and p65 cDNA were purchased from Shanghai GenePharma. The siRNA sequences, and their non-inhibitory controls, were as follows: p50: 5′-CGCCAUCUAUGACAGUAAATT-3′; control: 5′-UUCUCCGAACGUGUCACGUTT-3′; p65: 5′-GGACAUAUGAGACCUUCAATT-3′; control: 5′-ACGUGACACGUUCGGAGAATT-3′; KCNN4 sense: 5′-GCCGUGCGUGCAGGAUUUA-3′; anti-sense: 5′-UAAAUCCUGCACGCACGGC-3′; Lipofectamine 2000 was used to transfect siRNA into M2 macrophage according to the manufacturer’s protocol.
Cell invasion assays
Transwell inserts were used to perform cell invasion assays. After coating the upper chamber with Matrigel, 1 × 105 cells in 0.2 ml serum-free RPMI 1640 medium were added. The lower chamber contained 0.8 ml medium with 10% FBS. After incubating at 37°C, 5% CO2 for 24 h, cells that had migrated to the lower chamber were fixed with 4% paraformaldehyde, and stained with 0.1% crystal violet in methanol, then counted under a microscope.
mRNA extraction and real time quantitative RT-PCR
Total RNA was extracted using Trizol, and reverse transcribed using PrimeScript RT from 500 ng RNA according to the manufacturer’s protocol. Quantitative real-time RT-PCR was performed using the LightCycler 480 (Roche, Basel, Switzerland) and SYBR Assays (Takara, Dalian, China). Primers were designed to detect
CCL2, CXCL12, CCL17, CCL18, CCL22, EGF, IL-1, IL-6, IL-8, IL-10, VEGFA, TGF-β and
GAPDH. Oligonucleotide sequences of qRT-PCR primers are shown in Table
1. Each sample contained 1× SYBR Premix Ex TaqTM, 0.2 μM of each forward and reverse primers and 500 ng template cDNA in a final volume of 20 μl. Cycling parameters were set as follows: denaturation at 95°C for 30 s, followed by 40 amplification cycles (95°C for 5 s and 60°C for 20 s). For relative quantification, 2
-ΔΔCt was used to calculate the fold change in gene expression. All of the experiments were performed in triplicate.
Table 1
Oligonucleotide sequence of qRT-PCR primers
CXCL12 | 5′-CCCGAAGCTAAAGTGGATTC-3′ | 5′-TTCAGAGCTGGGCTCCTACT-3′ | 112 |
CCL18 | 5′-CTCTGCTGCCTCGTCTATACCT-3′ | 5′-CTTGGTTAGGAGGATGACACCT-3′ | 108 |
CCL17 | 5′-AGGGACCTGCACACAGAGAC-3′ | 5′-CTCGAGCTGCGTGGATGTGC-3′ | 133 |
CCL22 | 5′-ATGGCTCGCCTACAGACTGCACTC-3′ | 5′-CACGGCAGCAGACGCTGTCTTCCA-3′ | 114 |
IL-6 | 5′-AATAACCACCCCTGACCCAAC-3′ | 5′-ACATTTGCCGAAGAGCCCT-3′ | 149 |
IL10 | 5′-AACAAGAGCAAGGCCGTGG-3′ | 5′-GAAGATGTCAAACTCACTCATGGC-3′ | 93 |
IL-1 | 5′-TCTGTTCTTGGGAATCCATGG-3′ | 5′-TCAGTGATGTTAACTGCCTCCAG-3′ | 96 |
IL-8 | 5′-AAACCACCGGAAGGAACCAT-3′ | 5′-CCTTCACACAGAGCTGCAGAAA-3′ | 101 |
CCL2 | 5′-AAGATCTCAGTGCAGAGGCTCG-3′ | 5′-CACAGATCTCCTTGGCCACAA-3′ | 103 |
EGF | 5′-CTTGTCATGCTGCTCCTCCTG-3′ | 5′-TGCGACTCCTCACATCTCTGC-3′ | 118 |
VEGFA | 5′-ATGACGAGGGCCTGGAGTGTG-3′ | 5′-CCTATGTGCTGGCCTTGGTGAG-3′ | 91 |
TGF-β | 5′-AAGGACCTCGGCTGGAAGTGC-3′ | 5′-CCGGGTTATGCTGGTTGTA-3′ | 137 |
GAPDH | 5′-ATCACCATCTTCCAGGAGCGA-3′ | 5′-CCTTCTCCATGGTGGTGAAGAC-3′ | 112 |
ChIP-qPCR
Chromatin immunoprecipitation (ChIP) assays were performed according to manufacturer’s instructions using the ChIP assay kit from Thermo Scientific and the NF-κB antibody from Abcam. Briefly, DNA and proteins were cross-linked by the addition of formaldehyde (1% final concentration) 10 min before harvesting, and crosslinking was terminated by the addition of glycine solution for 5 min at room temperature. After that, the cells were scraped off the plates, and resuspended in PBS with lysis cocktail (1% final concentration). The DNA was then sheared into 0.5–1 kbp fragments using sonication at 20% amplitude, seven times, each for 30 s. After centrifugation, the supernatant was precleared by incubation with Protein A/G beads, adsorbed with salmon sperm DNA at 4°C. The cleared lysates were then incubated overnight with NF-κB antibody. Immune complexes were precipitated with protein A/G beads. Pre-immunized rabbit serum was used as a negative control, and the supernatant of this reaction was used as an input control. Immunoprecipitated samples were incubated at 65°C for 12 h to reverse the crosslink. DNA was extracted using a DNA extraction kit (CoWin Biotech, Beijing, China) and qPCR was performed with the following primers using the SYBR Premix Ex Taq II kit (Takara, Dalian, China): KCNN4-F: 5′-TCATCACTGCGAGCACTTGT-3′ KCNN4-R: 5′-CGAAACCCAATACGTGTAGACA-3′. A melting curve analysis was performed at the end of target gene amplification. For relative quantification, 2-ΔΔCt was used to calculate the fold change in gene expression. All of the experiments were performed in triplicate.
Plasmids and recombinants
The plasmids including pGL3-basic, pRL-TK and pGL3-promoter used for luciferase reporter gene expression analysis were purchased from Promega Ltd. A 127 bp fragment comprising -585 to -459 bp upstream of the KCNN4 transcription start site (TSS), which contained the predicted NF-κB binding site (CCATACAGGG), was amplified and inserted into the pGL3-promoter vector to construct pGL3 - 585/-459 vector. Additionally, pGL3 - 585/-459-M vector with a mutated NF-κB binding site (CCCCGGAGGG) in the KCNN4 regulatory region was constructed. Key regions in all constructs were verified by DNA sequencing.
Reporter gene assays
TAMs with high endogenous expression of NF-κB were allowed to grow to 60% confluency in 24-well dishes. After 24 h, pGL3 - 585/-459, pGL3 - 585/-459-M and pGL3-basic were transfected into TAM cells using Lipofectamine™ 2000 reagent and incubated for 24 h. Cells were washed twice, suspended in 100 μl reporter lysis buffer (Promega) and luciferase activity measured using the dual luciferase reporter assay system and a GloMax 20/20 luminometer (Promega, Madison, Wisconsin, USA) according to the manufacturer’s protocol. The Renilla luciferase vector pRL-TK (Promega, Madison, Wisconsin, USA) was co-transfected to standardize transfection efficiency in each experiment.
Immunofluorescence staining
For immunofluorescence staining, the specimens were incubated with mouse anti-hCD68 mAb (diluted 1:100), rabbit anti-hIL-6 Ab (diluted 1:100) and rabbit anti-hIL-8Ab (diluted 1:100) at 4°C overnight. Secondary staining with Alexa-Fluor-555 conjugated donkey anti-rabbit and Alexa-Fluor-488 conjugated goat anti-mouse secondary antibodies was carried out at room temperature for 60 min, followed by DAPI nuclear counterstaining for 10 min. Images were taken with a Zeiss LSM 700 laser scanning microscope (Carl Zeiss) with a core data acquisition system (Applied Precision). For control experiments, primary antibody was substituted with normal rabbit serum.
Statistics
Statistical analyses were performed using SPSS 13.0 (SPSS Inc, USA). All data are present as the mean ± S.D. Unpaired Student’s t test and one-way ANOVA were used, as appropriate, to assess the statistical significant of differences between two groups and three or more groups respectively. χ2 test was applied to analyze the relationship between IL-6 and IL-8 double-positive TAMs counts and clinicopathologic features. Kaplan–Meier survival curves were plotted and log-rank test was carried out. In all cases, a value of p < 0.05 was accepted as significant.
Discussion
The tumor microenvironment has been shown to be composed of both of tumor cells and mesenchymal cells, and that the microenvironment itself is involved in tumorigenesis [
12]. Tumor-associated macrophages, or TAMs, are macrophages that are located in the tumor environment. There are two types of macrophages, M1 and M2. M1 macrophages have antitumor activities and can produce TNF-α, whereas M2 macrophages are a type of TAM that can produce TGF-β and express CD68 and CD206 surface markers. We used PMA to induce the transformation of THP-1 monocytes into M2 macrophages according to described methods [
13]. TAMs contribute to tumor progression by releasing a variety of cytokines, such as VEGF, PDGF and IL-10 [
14]. In the tumor microenvironment, autocrine and paracrine loops, controlled by cytokines and receptors, have been observed between tumor-associated macrophages and tumor cells during tumor initiation, promotion, and metastasis [
15,
16]. Our previous studies have demonstrated that PRL-3 can promote the proliferation and metastasis of CRC cells through the autocrine secretion of TNF-α, which induces KCNN4 channel expression by activating the NF-κB signaling pathway. Considering that TNF-α could act as an autocrine and paracrine cytokine to promote the proliferation and metastasis of tumor cells [
17], we speculated that there might be a paracrine loop between CRC cell and TAMs in the tumor microenvironment that is controlled via TNF-α. In this study, we showed that TAMs participate in the progression of CRC induced by PRL-3 through the TNF-α mediated-secretion of IL-6 and IL-8 in a paracrine manner. Moreover, such regulation could be inhibited by TRAM-34, a KCNN4 channel-specific inhibitor.
To our knowledge, this is the first report indicating that KCNN4 channels participate in PRL-3 induced secretion of IL-6 and IL-8 by TAMs. Previous studies have demonstrated that KCNN4 channels belong to the Ca
2+-activated potassium channel superfamily, and the activation of these channels is dependent on conformational changes in calcium calmodulin [
18]. KCNN4 channels are mainly expressed in peripheral tissues, including the hematopoietic system, colon, lung and pancreatic tissue, and play an important role in the transport of substances [
19]. Previous research has also revealed that KCNN4 channels regulate cell cycle progression and cell growth in human endometrial cancer and prostate cancer cells [
20,
21]. Our data showed that when LoVo-P cells were cocultured with TAMs, the expression of KCNN4 channels was significantly increased, indicating that transcriptional mechanisms are likely to be responsible for the increased KCNN4 expression. A previous study revealed that activation protein-1 (AP-1) could regulate KCNN4 channel expression in T-cell activation [
22]. Our research demonstrated that LoVo-P cells could release TNF-α and subsequently regulate the KCNN4 expression of TAMs in a paracrine manner. Consistent with our hypothesis, ChIP-qPCR and reporter gene assays indicated that NF-κB was required for transcription of the
KCNN4 gene and mediated the transcriptional activation of KCNN4 channel expression in TAMs when they were cocultured with LoVo-P cells.
Although many studies have highlighted the role of TAMs in tumor metastasis, we still questioned whether TAMs could enhance the metastasis of tumor cells induced by PRL-3. In this study, when LoVo-P cells were cocultured with TAMs, invasion was significantly enhanced. Moreover, we demonstrated that PRL-3 is important for the ability of TAMs to enhance LoVo cell invasiveness, as LoVo-P cells cocultured with TAMs exhibited increased invasiveness compared with LoVo-P cells that were not cocultured with TAMs. LoVo-C cells did not exhibit the same features. Additionally, when TAMs were pretreated with p50/p65-siRNA or BAY11-7082, the invasive ability of LoVo-P cells was inhibited and the expression of KCNN4 channels was decreased. It is possible that the KCNN4 channels of TAMs enhanced the PRL-3-induced metastasis of CRC cells.
Previous studies have shown that potassium channels can regulate the cytokine secretion of human activated macrophages [
23]. We therefore tested whether KCNN4 channels could regulate the cytokine secretion of TAMs. Our data showed that once LoVo-P cells were cocultured with TAMs, the secretion of IL-6 and IL-8 was significantly increased. Moreover, it is notable that the secretion of IL-10 was also increased significantly. IL-10 is produced mainly by T cells and macrophages and has a role in immunoregulation. It has been shown that IL-10 can inhibit the expression of MHC molecules and costimulatory molecules, and can reduce antigen-presentation [
24]. Besides, IL-10 can also impair secondary CD8
+ T cell responses and thus inhibit tumor immunity [
25]. In addition, it has been shown that TAMs in the tumor microenvironment not only secret inflammatory mediators but also have immunoregulatory effects [
26]. We are therefore curious as to whether IL-10 produced by TAMs may contribute to the immunosuppressive tumor environment, and this feature of TAMs will be studied in our future research. Next, we explored whether KCNN4 channels could regulate IL-6 and IL-8 expression. When TAMs were pretreated with TRAM-34, the expression of IL-6 and IL-8 was significantly decreased, compared with controls, when the TAMs were cocultured with LoVo-P cells. These observations suggest that KCNN4 channels have the ability to regulate TAM secretion of IL-6 and IL-8.
The mechanism by which KCNN4 channels regulate IL-6 and IL-8 expression of TAMs remains to be investigated. It is known that KCNN4 channels in lymphocytes can maintain a hyperpolarized membrane potential, thereby facilitating and maintaining the intracellular Ca
2+ levels required for cell proliferation and gene expression [
27,
28]. Recent studies have revealed that intracellular Ca
2+ levels regulate the secretion of TNF-α and IL-6, which play important roles in inflammation [
29]. However, another previous study revealed that a calcium ionophore can inhibit IL-6 and IL-8 expression in Jurkat T-cells [
30]. Further studies are needed to verify whether the intracellular Ca
2+ levels of TAMs can regulate the expression of IL-6 and IL-8 and to determine the underlying mechanism.
Cancer-associated inflammation is known to affect the proliferation, angiogenesis, and metastasis of tumor cells [
31]. For example, the activation of NF-κB triggers the production of the inflammatory chemokine IL-8 by tumor cells [
32]. Inhibition of the IL-6 signaling pathway inhibits tumor development in a colitis-associated carcinogenesis model [
33]. In our research, the levels of TAM-derived IL-6 and IL-8 were significantly increased upon coculture with LoVo-P cells. Using the transwell-invasion system, IL-6 and IL-8 significantly enhanced the ability of LoVo-P cells to metastasize, and this response was attenuated by IL-6 and IL-8 antibodies. These results suggest that TAM-derived IL-6 and IL-8 induced by PRL-3 affect the metastasis of tumor cells in a paracrine manner.
Further confirming this result, and extending it to human disease, we demonstrated that IL-6 and IL-8 were mainly expressed by TAMs in CRC tissues. Studies have shown that serum levels of IL-6 and IL-8 were higher in CRC patients versus controls, indicating that IL-6 and IL-8 may be potential targets for CRC [
34]. Furthermore, our data also demonstrated that a high density of TAMs expressing IL-6 and IL-8 was positively correlated with tumor stage. Therefore, targeting therapies against IL-6, IL-8, and KCNN4 channels of TAMs may prevent CRC liver metastasis through a number of mechanisms, including both metastasis prevention and anti-inflammatory effects.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ZHC designed the research; HYX, WL carried out the vector construction and stable cell line generation; HYX, WL, YZ, carried out the flow cytometry, cell coculture, western blotting, RT-PCR, ELISA, ChIP-qPCR, reporter gene assays; LL collected the patients’ data and carried out immunofluorescence staining. XXL carried out the invasion assay; HW and QSL carried out the statistical analysis; HYX and WL wrote the paper. All of the authors have been involved in revising the manuscript and have given final approval of the version to be published.