MCPIP1 Downregulation in Clear Cell Renal Cell Carcinoma Promotes Vascularization and Metastatic Progression

Over the last 20 years, the incidence of renal cancer has increased by 30%, and more than 270,000 of new cases are diagnosed worldwide each year (1). The highest incidences of renal cell carcinoma (RCC) tend to occur in Western Countries (2). Clear cell RCC (ccRCC) is the most common type among renal carcinomas, constituting 80% of all cases (3). Over 30% of patients present with metastasis at the time of diagnosis (4). Despite the use of targeted therapy, the survival time of patients with metastatic ccRCC ranges from three months to five years (4).

Tumor vascularization is a critical step during ccRCC development because it promotes local tumor progression and metastatic spread (5). The tumor production of proangiogenic proteins such as VEGF promotes the recruitment of endothelial precursor cells from the bone marrow to form new circulatory networks throughout the tumor (6, 7). Angiogenic processes can also be induced by the downregulation of antiangiogenic agents, activation of oncogenes, inactivation of tumor suppressor genes, and hypoxia (8). One of the major inducers of angiogenesis during tumor progression is inflammation (9). It has been shown that proinflammatory cytokines, IL1, IL6, or monocyte chemotactic protein 1 (MCP-1) are required for angiogenesis and tumor growth, invasion, and metastasis in animal models. Moreover, these factors promote the expression of certain angiogenic factors, for example, VEGF, whereby they indirectly stimulate the process of tumor vasculature (9, 10).

Monocyte chemotactic protein 1–induced protein 1 (MCPIP1), also known as Regnase-1 encoded by the ZC3H12A gene, is a newly discovered zinc finger protein that is upregulated in macrophages during inflammation in response to certain molecules, such as TNF, IL1β, and lipopolysaccharide (LPS; refs. 11–13). MCPIP1 acts mainly as an endonuclease that degrades the mRNA of proinflammatory cytokines, such as IL6, IL1β, IL12, and IL2 (13–15). In addition, MCPIP1 is a negative regulator of NFκB activity (11, 16, 17) and may suppress miRNA biosynthesis (18). Recently, it has been shown that MCPIP1 is involved in the regulation of viability and proliferation in numerous cancer cell lines (19–21). Moreover, MCPIP1 overexpression in breast cancer cells induces apoptosis in vitro and reduces tumor growth and metastatic disease in vivo (22). Interestingly, some studies have suggested that MCPIP1 may promote endothelial cell angiogenesis by increasing the secretion of angiogenic factors, such as VEGF and hypoxia-inducible factor 1-alpha (HIF1α; refs. 23, 24). Our recent findings have suggested that MCPIP1 may decrease the level of HIFs in tumor cells, and we showed that MCPIP1 is downregulated in ccRCC samples compared with normal tissue (25). In this study, we demonstrate that decreasing the MCPIP1 protein level correlates with renal cancer progression and better tumor vascularity. We found that the lack of RNase activity of MCPIP1 is responsible for the secretion of proangiogenic factors, the acquisition of the mesenchymal phenotype, and the metastatic spread of ccRCC cells. Our findings suggest that ccRCC growth, vascularization, and metastasis depend on the level of MCPIP1 and its endonuclease activity.

Renal tumor tissue samples were obtained from patients surgically treated for renal cancer in the Centre of Oncology, Maria Sklodowska-Curie Memorial Institute (Cracow Branch, Poland). The study was approved by Institutional Review Board and informed consent was obtained from each patient. All human tissue was collected using protocols approved by the Local Ethics Committee (approval no. 68/KBL/OIL/2011). All samples were subjected to histologic evaluation by pathologists, diagnosed according to World Health Organization classification, and divided into four groups (I–IV) according to the histologic grading Fuhrman system. In each case, the sample was divided and frozen in liquid nitrogen and stored at −80°C for protein isolation or incubated overnight in RNA-later (Invitrogen) at 4°C and stored in −80°C for RNA isolation. The analysis of the MCPIP1 protein level included 60 samples: 15 samples from each group (I–IV). The microarray analysis included 16 samples.

Experiment was performed using Affymetrix HuGene ST 2.1 microarrays on 16 RNA samples isolated from ccRCC patient tissues from grades I–IV (8 samples for I+II and 8 samples for III+IV). Total RNA was extracted using Eurx Universal RNA Purification Kit (EURx) according to the manufacturer's protocol. The quantity of ribosomal RNA and DNA contamination was examined using electrophoresis in 1% denaturing formaldehyde gel. Concentration of total RNA was assessed using NanoDrop 1000 Spectrophotometer (Thermo Fisher Scientific). One-hundred ng of total RNA were used to synthesis and amplification of ss-cDNA, and then the samples were fragmented and labeled with biotin. Each step was performed according to Affymetrix GeneChip WT PLUS Reagent Kit Manual. Following fragmentation, 10 μg of cRNA were hybridized for 16 hours at 48°C on Affymetrix HuGene2.1 ST Array Strip. Arrays were washed and stained in the Affymetrix Gene Atlas Fluidics Station. Affymetric HuGene2.1 ST Array Strips were scanned using the GeneAtlas Imaging Station. The data were normalized with Expression Console Software 1.4.1 with RMA algorithm and analyze using Affymetrix Transcriptome Analysis Console (TAC) Software 3.1 with ANOVA one-way between subjects (unpaired). We have followed the Minimum Information About a Microarray Gene Experiment (MIAME) guidelines and deposited raw and processed data in the Gene Expression Omnibus (GEO) repository with the accession number GSE89563.

Human clear cell renal cell carcinoma cell line of primary tumor (Caki-2) was obtained from Sigma in 2015. Caki-1 cells were obtained from ATCC in 2011. HEK-293 cell line was obtained from ATCC in 2010. Human microvascular endothelial cells (HMEC-1) were obtained in July 2016 from ATCC and human umbilical vein endothelial cells (HUVEC) were kindly provided by the Department of Transplantology, Medical College (Krakow, Poland) in 2016. The initial vials were expanded and cryopreserved in the laboratory and propagated no longer than fifteen consecutive passages. HUVECs were used at second or third passage. All cell lines were routinely validated to be mycoplasma free by PCR every three months.

Caki-2 cells were cultured in McCoy 5A medium (Lonza) supplemented with 10% FBS (BioWest). Caki-1 was cultured in Eagle minimal essential medium (EMEM; Lonza) with 10% FBS. HEK-293 was cultured in DMEM high glucose medium (Lonza) supplemented with 10% FBS. HMEC-1 cell line was cultured in MCDB 131 medium (Lonza) supplemented with 10% FBS, l-glutamine (2 mmol/L), EGF 10 ng/mL, hydrocortisone 1 μg/mL, and HUVEC was maintained in EBM-2 medium with supplements for HUVECs (Clonetics). All cell lines were cultured at 37°C in a 5% CO₂ atmosphere. To collect conditioned media from ccRCC cell lines, cells were cultured for 48 hours in normoxic (21% oxygen) or hypoxic (1% oxygen) conditions in media supplemented with 0.5% BSA (BioShop).

All cell lines were authenticated by GenMed, using STR DNA profiling methods within loci of D8S1179, D21S11, D7S820, CSF1PO, D3S1358, TH01, D13S317, D16S539, D2S1338, D19S433, vWA, TPOX, D18S51, D5S818, FGA, and AMEL. Caki-1, Caki-2, and HUVECs were analyzed in August 2016 and HEK293 cell line was analyzed in April 2017.

Raw and processed data from microarray analysis are deposited in the Gene Expression Omnibus repository: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE89563 (accession number: GSE89563).

To examine the effect of MCPIP1 in ccRCC three different viral vectors were used. To stably inhibit MCPIP1, lentiviral vectors with control (shCtrl) were used (Sigma). Two different shRNAs were tested (sh4 and sh5) and most efficient was chosen for further experiments. For stable overexpression of MCPIP1, doxycycline-dependent TetON system was used (pLIX MCPIP1, pLIX PURO and mutant form pLIX D141N) or retroviral vectors with control (pMX MCPIP1, pMX PURO). Briefly, ccRCC cells were plated at 50% of confluency. Viral vectors were added at multiplicity of infection (MOI) of 50 with 6 μg/mL polybrene (Millipore). To increase transduction efficiency after 24 hours, the process was repeated. Cells were incubated with viruses for 24 hours, and then the medium was changed. After additional 24 hours, 1 μg/mL of puromycin (InvivoGen) was added to start selection. To induce MCPIP1 overexpression in TetON system, 1 μg/mL of doxycycline (BioShop) was added.

The experiments have been conducted in accordance with an Institutional Animal Care and Use Committee: the Local Ethics Committee of the Collegium Medicum of the Jagiellonian University in Krakow, Poland (approval no 117/2014) and II Local Ethics Committee of Institute of Pharmacology Polish Academy of Sciences (approval no. 20/2017 and 21/2017). Mice were handled according to the regulations of national and local animal welfare bodies. Foxn1^nu/Foxn1^nu (Envigo or Charles River Laboratory) and NOD-SCID (Charles River Laboratory) 6 weeks old female mice were kept under SPF conditions, with water and food ad libitum. Caki-1 and Caki-2 cells with stable overexpression, mutation (pMX PURO, pMX MCPIP1, pLIX PURO, pLIX D141N, pLIX MCPIP1), or downregulation (shCtrl, shMCPIP1, shCtrl-GFP, shMCPIP1-GFP) of MCPIP1, were injected subcutaneously as a cell suspension (2.5 × 10⁶ cells in PBS-Caki-1, 5 × 10⁶ cells in PBS-Caki-2). Tumor growth was monitored for 6 (shCtrl, shMCPIP1, pLIX PURO, pLIX D141N, pLIX MCPIP1) or 8 (pMX PURO, pMX MCPIP1) weeks. Tumor volume was estimated using caliper measurements, according to the formula: volume = width × depth² of tumor or 3D ultrasonography. After tumor and lung excision, RNA isolation and histologic analysis were performed.

Tumor volume and vascularity in nude mice were measured every week after subcutaneous (s.c.) injections using 3D ultrasonography VEVO 2100 with MS550D transducer (VisualSonics). Mice were anesthetized with 2% isoflurane with air flow 0.5 L/minute (Baxter). Vasculature measurements were performed using Power Doppler mode. All results were analyzed in 3D mode using VEVO 2100 Software.

Total RNA from cultured cells and tissues was isolated using the Universal RNA Purification Kit (EURx). RNA from total mice lungs were isolated using Fenozol (A&A Biotechnology). The quantity of ribosomal RNA and DNA contamination was examined using electrophoresis in 1% denaturing formaldehyde gel. Concentration of total RNA was assessed using NanoDrop 1000 Spectrophotometer (Thermo Fisher Scientific). Reverse transcription was performed using 1 μg of total RNA, oligo(dT) 15 primer (Promega) and M-MLV reverse transcriptase (Promega). Real-time PCR was carried out using SybrGreen Master Mix (A&A Biotechnology) and Eco Real-Time PCR System (Illumina). For the examination of mice lung metastasis, specific probes for human GAPDH and mouse GAPDH (Life Technologies) were used with Taq PCR Master Mix (EURx). Gene expression was normalized to elongation factor-2 (EF2). mRNA level in each sample was analyzed in duplicates. The relative level of transcripts was quantified by the ΔΔC_t method. Sequences of primers (Genomed) and annealing temperatures are listed in Table S1 in Supplementary Material.

Cultured cells were washed with ice-cold PBS (Lonza), harvested, and lysed with Mammalian Protein Extraction Reagent (M-PER) (Thermo Scientific) supplemented with proteinase and phosphatase inhibitors, for 10 minutes. Lysates were spun for 20 minutes, 11,000 × g at 4°C.

SDS-PAGE electrophoresis was conducted in 10% polyacrylamide gel. After wet transfer onto polyvinylidene difluoride membrane (Millipore), membranes were blocked in appropriate buffer [5% BSA or 5% non-fat powdered milk in Tris-buffered saline with 0.1% Tween 20 (Sigma)]. Next, membranes were incubated with primary antibodies at 4°C overnight, with gentle agitation. On the following day, membranes were washed three times for 10 minutes with TBS 0.1% Tween 20 and incubated with secondary antibody for 1 hour at room temperature with gentle rocking. Chemiluminescence was detected after 5-minute incubation with Immobilon Western HRP substrate (Millipore) using ChemiDoc system (Bio-Rad). All antibodies and dilutions are listed in Table S2 in Supplementary Material.

Five-thousand cells per well were seeded in 96-well plates. Then 0.5 mg/mL of MTT solution (Sigma) was added for 2 hours. Reaction was stopped by adding 5 mmol/L HCl (POCH) in isopropanol (Chempur) and stirred for 30 minutes to dissolve the formazan crystals. Absorbance was measured at 570 nm with the reference wavelength of 500 nm using Tecan Spectra Fluor Plus Microplate Reader (Tecan Group Ltd.). Measurements were performed 24, 48, 72, and 96 hours after seeding. Three independent experiments were performed in triplicate.

Cells were cultured on 24-well plates (20,000 cells per well). After 24, 48, 72, and 96 hours, cells were trypsinized and counted in Bürker chamber. Three independent experiments were performed in triplicate. Results are expressed as number of cells per 1 mL.

Cells were grown on 24-well plates to full confluence. Right before experiment, 10 mmol/L of hydroxyurea (Sigma) was added to media to block cell proliferation. A scratch was generated with a small pipette tip. Then the plate was transferred to microscope culture chamber with 37°C, 5% CO₂. Images were acquired every 10 minutes for 16 hours using Leica DMI6000B inverted widefield fluorescence microscope (Leica Microsystems). All images were recorded using 10× dry objective with Leica LAS X image acquisition software and analyzed with Hiro v. 1.0.0.4 software.

The directional movement of HMEC-1 cells toward conditioned medium was evaluated using modified Boyden chamber with 8-μm pore polycarbonate membrane inserts (Transwell; Costar-Corning). Cells were seeded into the top chamber of an insert at a density of 20,000 cells in 100 μL, in duplicates. The bottom chamber was filled with prewarmed conditioned medium mixed with MCDB131 medium with 0.5% of BSA (1:1) or with 0.5% BSA DMEM as a negative control. After 24 hours, cells that did not migrate were scraped off with cotton wool from the top membrane while cells that had transmigrated to the bottom side of the membrane were fixed with methanol, stained with Wright solution, and counted under inverted microscope with high power field (HPF). At least five fields were counted each time and the mean number of cells per HPF was calculated. Three independent experiments in duplicates were performed.

Fifty μL of growth factor–reduced Matrigel (BD Biosciences) was plated in a 96-well plate and incubated at 37°C for 30 minutes. HMEC-1 cells were detached, counted, and single-cell suspensions at a density of 10,000 cells per well in 200 μL of conditioned media and proper controls were plated on the Matrigel. Four hours after seeding endothelial tube formation was stained with Calcein AM (1 μmol/L). Images were captured using Olympus IX50 inverted microscope (Olympus) with 4× dry objective. Formation of tubule-like structures was analyzed with Angiogenesis Analyzer for ImageJ (Carpentier G., Angiogenesis Analyzer for ImageJ; available online: http://imagej.nih.gov/ij/macros/toolsets/Angiogenesis%20Analyzer.txt).

To study angiogenesis, fibrin beads test was performed. Shortly, latex beads (Sigma) were incubated with endothelial cells for 6 hours (approximately 400 cells per bead) to coat the beads. After 24 hours, beads were resuspended in 2 mg/mL fibrinogen (Sigma) solution in medium with 0.5% BSA. To generate gel, 0.625 U/mL of thrombin (Sigma) was added, and then the beads were embedded in 24-well plate. After polymerization, on the top of the gel, ccRCC cells were seeded (30,000 cells per well) in medium supplemented with 0.5% BSA. Medium was changed every second day. To evaluate migration and formation tubular structures by endothelial cells, images were taken every day for six days using Nikon Eclipse Ti-S (Nikon) microscope with 10× objective. For migration analysis, ImageJ (NIH, Bethesda, MD) software was used.

To evaluate levels of secreted proteins in conditioned media (IL8, VEGF, and IL6) or in mice plasma (SDF-1) DuoSet ELISA assays were performed (R&D Systems) according to the manufacturer's instructions. Absorbance was measured at 450 nm with the reference wavelength of 540 nm using Tecan Spectra Fluor Plus Microplate Reader. Three independent experiments were performed in triplicate.

Mouse blood were collected at the end of experiment, and next, the erythrocytes were lysed with High-Yield Lyse Fixative-Free Lysing Solution (Life Technologies) for 30 minutes. Lysed blood samples were diluted 100× in PBS, and 10 mL of each sample was taken for analysis. Measurement was performed using Attune NxT Acousting Focusing Cytometer and analyzed with Attune software v2.4.

Cells were plated on glass coverslips in 12-well culture plates at density of 100,000 cells per well and grown to full confluence. Then, cells were stimulated with conditioned media for 3 hours and fixed for 15 minutes in 4% methanol-free formaldehyde (Thermo Scientific) in PBS at room temperature. Next, slides were permeabilized and blocked in blocking buffer [1% BSA in PBS with 0.3% Triton X-100 (Sigma)] at room temperature for 1 hour. Incubation with primary antibody for VE-Cadherin (Abcam) diluted 1:100 in PBS with 1% BSA was performed overnight at 4°C in a humidifying chamber. On the following day, secondary antibody AlexaFluor 488–conjugated goat anti-rabbit (1:1,000; Thermo Fisher Scientific) was added for 1 hour in the dark at room temperature in a humidifying chamber. All samples were mounted with prolong gold antifade reagent with DAPI (Invitrogen) and sealed with nail polish. Images were taken with Leica DMI6000B inverted widefield fluorescence microscope with 100× or 63× oil immersion objective with Leica LAS X image acquisition software.

Tumor sections were stained with hematoxylin–eosin to visualize tumor morphology. IHC evaluation was performed using primary rabbit polyclonal anti-CD31 antibody (1:50, Abcam) and EnVision Detection Systems Peroxidase/DAB, Rabbit/Mouse (DakoCytomation) to visualize tumor vascularization. All images were taken using Nikon Eclipse Ti-S microscope with 10× objective. For analysis, ImageJ software was used. Slides for CD31 on frozen sections: tissues were prefixed in pure buffered formaldehyde (Chempur), washed in PBS, incubated for 12 hours in 30% sucrose at 4°C and embedded in OCT (VWR chemicals). Nine micron–thick sections were stained immunohistochemically with primary rat anti-mouse CD31 antibody or isotype control (1:150, BD Pharmingen) and subsequently with secondary goat anti-rat AlexaFluor 594–conjugated antibody (1:1,000; Thermo Fisher Scientific). Slides were mounted with Dako Fluorescent Mounting Medium (Dako). Images were taken with Leica DMI6000B inverted widefield fluorescence microscope with 40× or 63× oil immersion objective with Leica LAS X image acquisition software.

In vitro experiments were conducted in three independent repetitions. The number of animals or patient samples is indicated in the figure legends. All results are shown as the mean ± SD. For comparison of two groups, two-tailed unpaired Student t test, two-way, or one-way ANOVA with post hoc Bonferroni test were used. The P values are marked with the asterisks in the charts (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 versus control).

The animal experiments were approved by the Local Ethics Committee of the Collegium Medicum of the Jagiellonian University in Krakow, Poland (approval no. 117/2014) and II Local Ethics Committee of Institute of Pharmacology Polish Academy of Sciences (approval no. 20/2017 and 21/2017). Biopsies of renal tumors were obtained from patients surgically treated for renal cancer in the Centre of Oncology, Maria Sklodowska-Curie Memorial Institute, Cracow Branch under the supervision of the Local Ethics Committee (approval no. 68/KBL/OIL/2011).

To assess MCPIP1 expression during ccRCC progression, we analyzed 60 tumor samples. Western blot analysis revealed that MCPIP1 expression varied depending on the tumor grade, with MCPIP1 expression decreasing significantly with tumor progression (Fig. 1A).

Microarray analysis showed a high level of proangiogenic factors in tumor tissue during ccRCC progression (Fig. 1B). We observed a high expression of VEGFA, FLT1, CDH5, or EPAS1 transcripts for all grades (I–IV). The expression was slightly decreased in grades III and IV compared with grades I and II (Fig. 1B). We also observed that the transcript levels of HGF, MET, SRC, ZEB2, IL6, and LIF, which are related to tumor invasiveness and metastasis, were increased during tumor progression (Fig. 1B), together with a decrease of MCPIP1 protein (Fig. 1A). Transcript of MCPIP1 (ZC3H12A) level in tumor samples was generally low (Fig. 1B).

The observation that the MCPIP1 protein level decreases as the tumor grade increases led us to examine the role of MCPIP1 in ccRCC. In the first step, we downregulated MCPIP1 at the mRNA and protein levels using six different siRNAs together with the proper controls (Supplementary Fig. S1A) and two lentiviral vectors expressing shMCPIP1#4 and #5 sequences (Supplementary Fig. S1B). The shMCPIP1#5 lentiviral vector, efficiently downregulated MCPIP1, was selected for further experiments and was called shMCPIP1 (Supplementary Fig. S1C). To determine whether MCPIP1 silencing influences ccRCC viability and proliferation, cells were cultured for 96 hours, and MTT and proliferation assays were performed. We found that MCPIP1 depletion in ccRCC cells significantly enhanced tumor cell viability and proliferation in both investigated cell lines, Caki-1 and Caki-2 (Fig. 1C and D; Supplementary Fig. S1D). The largest difference between the control and MCPIP1-deficient cells was observed after 96 hours of culture. To examine the growth of human ccRCC in vivo, we established a xenotransplantation model in NOD-SCID (Caki-1) mice. We observed that the inhibition of MCPIP1 affected both tumor growth and weight (Fig. 1E and F; Supplementary Fig. S2A–S2C). We obtained opposite results after the stable overexpression of MCPIP1 using retroviral vectors (Supplementary Fig. S3A and S3B), where the tumors formed by Caki-1 pMX MCPIP1 were significantly smaller than those in the controls (Supplementary Fig. S3C). The main function of MCPIP1 is the negative regulation of inflammation by acting as an endoribonuclease. In the next step, we checked whether the RNase activity of MCPIP1 is important for tumor growth in vivo (Fig. 1G and H). We expanded our in vivo model with cells overexpressing MCPIP1 (pLIX MCPIP1) and cells carrying a point mutation of D141N that completely abolishes MCPIP1 RNase activity (pLIX D141N; Supplementary Fig. S3A and S3B; ref. 13). Our findings show that for both investigated xenograft models, Caki-1 and Caki-2 cells overexpressing MCPIP1 formed the smallest, slowly growing tumors, whereas control cells and cells lacking RNase activity produced similar tumors (Fig. 1G and H).

Staining for CD31 (platelet endothelial cell adhesion molecule, PECAM-1), a vascular marker, showed that more advanced grading (III and IV) was correlated with significantly increased lumen areas of blood vessels compared with those seen in grade I and II tumors (Fig. 2A).

To study the role of MCPIP1 in tumor vascularity development, we analyzed tumors that were formed by cells with MCPIP1 downregulation. CD31 and hematoxylin and eosin staining of tumor sections showed an increase in the number of functional blood vessels with larger lumen areas (Fig. 2B; Supplementary Fig. S4A). Furthermore, the analysis of tumor by 3D ultrasonography revealed increased vascularization of tumors generated by ccRCC cells with MCPIP1 depletion (Fig. 2C). We also observed more CD31-positive cells during tumor development in tumors formed by MCPIP1-deficient cells (Fig. 2D).

As MCPIP1 acts as an endonuclease, we analyzed the vascularity of tumors generated by MCPIP1 overexpressing and lacking RNase activity (D141N) cells. We observed increased vascularization of tumors formed by ccRCC cells carrying a D141N mutation (Fig. 2E and F). Further analysis of CD31 staining revealed that the lack of RNase activity of MCPIP1 led to an increase in the number of blood vessels with larger diameters in Caki-1–derived tumors (Fig. 2E). Similar results were obtained in Caki-2 tumors using 3D ultrasonography (Fig. 2F).

In response to stimulation with conditioned media isolated from ccRCC cell lines, we first observed an increased chemotactic response of microvascular endothelial cells migrating toward the media isolated from shMCPIP1 Caki-1 and Caki-2 cells compared with that of cells that were chemoattracted by the media isolated from control Caki cells (Fig. 3A). Moreover, HMEC-1 cells suspended in conditioned media isolated from ccRCC cells with MCPIP1 downregulation formed higher numbers of branch points in the angiogenesis assay (Fig. 3B). To show the direct influence of MCPIP1 silencing in ccRCC cells on the activity of endothelial cells, we performed a coculture of HMEC-1 cells growing on beads in fibrin gel with ccRCC cells growing on top. We observed that endothelial cells under the influence of tumor cells lacking MCPIP1 showed an increased number of tubular structures (Fig. 3C, left) and higher migration rates of endothelial cells (Fig. 3C, right).

Vascular permeability and angiogenesis are regulated by VE-cadherin, a vascular endothelial adhesion molecule, specifically and exclusively expressed by endothelial cells and important in the maintenance of vascular integrity (26). Tyrosine phosphorylation of VE-cadherin is associated with weak junctions and impaired barrier function (27). To determine whether the silencing of MCPIP1 in ccRCC cells influences the level, localization, and phosphorylation of VE-cadherin in endothelial cells, we stimulated two types of endothelial cells, HUVECs (Fig. 3D) and HMEC-1s (Supplementary Fig. S4B), with conditioned media isolated from ccRCC cells. Western blot analysis showed that endothelial cells exhibited VE-cadherin phosphorylation in response to the media (Fig. 3D; Supplementary Fig. S4B). The phosphorylation level was very strong after stimulation with media from Caki-1 shMCPIP1 cells (Fig. 3D; Supplementary Fig. S4B). We observed a similar pattern after stimulation with media isolated from Caki-2 cells with MCPIP1 downregulation (Fig. 3D; Supplementary Fig. S4B). The phosphorylation of VE-cadherin in endothelial cells after stimulation with medium from ccRCC-overexpressing MCPIP1 (pMX MCPIP1) was significantly impaired (Fig. 3D; Supplementary Fig. S4B).

Because VE-cadherin is a component of endothelial cell-to-cell adherens junctions and because loosening cell–cell contacts is the first step in inducing angiogenesis, we evaluated the influence of ccRCC conditioned media on endothelial cell organization, vascular integrity, and cellular localization using VE-cadherin immunofluorescent staining (Fig. 3E). We observed that media isolated from ccRCC with silenced MCPIP1 blocked the adhesive function of VE-cadherin, disrupted the monolayer integrity of HUVECs, and increased endothelial cell dissociation (Fig. 3E). In contrast, the influence of the media isolated from ccRCC with MCPIP1 overexpression (pMX MCPIP1) and control cells (pMX PURO) was similar and did not change the cell organization effectively or in a visible manner (Fig. 3E).

Among angiogenic stimuli, the best characterized are VEGF (6, 7, 26), IL6 (28), and a member of the CXC chemokine family, IL8 (29). In the next step of our study, we examined whether a change in the level of MCPIP1 protein in ccRCC cells affects the secretion of IL8, VEGF, and IL6 (Fig. 4A–C). Using real-time RT-PCR, we observed that MCPIP1 downregulation increased the transcript level of IL8, VEGF, and IL6 both under normoxia and hypoxia conditions (Fig. 4A–C). Next, conditioned media from ccRCC cells were evaluated using ELISA, and the analysis revealed that low MCPIP1 levels significantly promoted the secretion of IL8, VEGF, and IL6 under hypoxic conditions (Fig. 4A–C).

The RNase activity of MCPIP1 was crucial for maintaining the low level of proangiogenic factors in ccRCC cells because a point mutation of D141 significantly increased the level of IL8, VEGF, and IL6 transcripts (Fig. 4D) and protein (Fig. 4E) under hypoxia conditions compared with the control and MCPIP1-overexpressing cells.

It has been previously shown that VEGF, together with stromal cell–derived factor (SDF)-1, synergistically induces neoangiogenesis in tumors (30). Moreover, it has been demonstrated that SDF-1, together with its receptor, CXCR4, promotes VEGF expression (31). We noticed that the downregulation of MCPIP1 in ccRCC cells significantly increased SDF-1 and CXCR4 expression in both investigated cell lines (Fig. 5A). Moreover, we observed a significant increase in SDF-1 expression in tumors isolated from nude mice injected with ccRCC cells with silenced MCPIP1 (Fig. 5B, left). Furthermore, analysis of blood serum from mice injected with MCPIP1-deficient cells showed higher levels of secreted human SDF-1 (Fig. 5B, right).

Interestingly, we observed decreased SDF-1 expression in ccRCC cells with MCPIP1 overexpression, whereas control and D141N cells had a similar level of SDF-1 (Fig. 5C). We found that the mRNA level of SDF-1 and CXCR4 was significantly decreased after MCPIP1 upregulation in both the Caki-1 and Caki-2 cell lines (Supplementary Fig. S5A and S5B).

MCPIP1 was shown to degrade the transcript of CCAAT/enhancer-binding protein beta (c/EBPβ; ref. 32), a transcription factor regulating the transcription of SDF-1 (33). Our findings show that the downregulation of MCPIP1 led to a significant increase in c/EBPβ and that the RNase activity of MCPIP1 regulates its expression in ccRCC cells (Fig. 5D).

SDF-1, together with its receptor CXCR4, is responsible for the invasive phenotype and metastatic behavior of many tumor types (34). Our studies have shown that silencing MCPIP1 increased the metastatic ability in vivo (Fig. 5E). We observed that nude mice injected s.c. with Caki-1 shMCPIP1 had considerably more lung metastases than did mice injected with control cells both during tumor development and after 6 weeks of tumor growth (Fig. 5E). Stable upregulation of MCPIP1 resulted in a decrease in the number of tumor cells metastasizing to the lungs (Fig. 5F; Supplementary Fig. S5C). Furthermore, we observed an increase in the number of circulating shMCPIP1-GFP cells in the mouse blood (Fig. 5G; Supplementary Fig. S5D and S5E). These findings were in concordance with lung metastasis (Fig. 5E).

The process of metastasis requires the acquisition of the ability to perform active migration, and we noticed that motile activity was strongly increased in ccRCC shMCPIP1 cells compared with control cells (Fig. 6A). Analysis of the individual tracks of both cell types revealed that the motile activity of MCPIP1-deficient cells was higher than that of control cells (Fig. 6B). A 16-hour time-lapse recording revealed that parameters for distance and speed were significantly higher for shMCPIP1 cells (Fig. 6C).

Because MCPIP1 downregulation in ccRCC led to increased migration, we assumed that MCPIP1 silencing at least partially modified the cell phenotype into a more aggressive form. To expand on these observations, we evaluated potential changes in the level of various epithelial-to-mesenchymal transition (EMT) markers and the c-Met receptor known to regulate cell motility (35). Western blot analysis confirmed by densitometry revealed that MCPIP1 silencing suppressed E-cadherin and upregulated β-catenin and vimentin. Moreover, c-Met and Src protein levels were increased. We also observed the increased phosphorylation of c-Met receptors in shMCPIP1 cells (Fig. 6D; Supplementary Fig. S6). Interestingly, similar results were obtained in the normal human renal cell line HEK293 after the downregulation of MCPIP1 (Fig. 6D). Further analysis of E-cadherin repressors revealed that MCPIP1 downregulation induced the expression of the transcription factors Snail and ZEB2 (Fig. 6E).

Subsequent studies have shown that MCPIP1 overexpression in ccRCC cells caused a significant decrease in motile activity (Fig. 7A–C). Activation of MCPIP1 expression resulted in a decrease in the distance and speed of ccRCC cells (Fig. 7A and B). Moreover, when we evaluated the levels of EMT markers, we observed a strong increase in E-cadherin and a decrease in vimentin and β-catenin. Furthermore, the levels of the activated and total form of the c-Met receptor, as well as those of Src kinase, were significantly reduced (Fig. 7D). The analysis of the influence of MCPIP1 overexpression and activation on Snail and ZEB2 transcription factors revealed that a high level of MCPIP1 significantly reduced the expression of both factors (Fig. 7E). The lack of RNase activity in pLIX D141N cells did not change the expression of both transcription factors compared with that in control cells (Fig. 7E), indicating the direct influence of the MCPIP1 protein.

In this study, we revealed for the first time that MCPIP1 plays an important role in the regulation of angiogenesis and metastasis of ccRCC. We present evidence that the abolished RNase activity of MCPIP1 is responsible for the secretion of proangiogenic factors and, consequently, the better vascularization of emerging tumors. Furthermore, we show that a low MCPIP1 level is crucial for the acquisition of the mesenchymal phenotype and that the downregulation of MCPIP1 is associated with the progression of ccRCC.

MCPIP1 has been shown to suppress chronic inflammation by promoting the mRNA decay of proinflammatory cytokines (13, 14), as well as regulating NFκB and AP1 activity (11, 16, 17). There is growing evidence that MCPIP1 plays a role during the process of tumorigenesis. It has been already demonstrated that MCPIP1 regulates the viability and proliferation of neuroblastoma (20), HeLa, HepG2 (19), ccRCC (25), and breast cancer cells (21). Previous findings have demonstrated that MCPIP1 degrades the mRNA of antiapoptotic gene transcripts in breast cancer cells, leading to apoptosis and tumor regression (22). Moreover, the same study showed that MCPIP1 expression in breast cancer samples was reduced compared with that in healthy tissue, and low MCPIP1 expression was associated with tumor progression and poor survival (22). Our group has already demonstrated that ccRCC tumors have a lower level of MCPIP1 than normal renal tissues, and MCPIP1 overexpression decreases viability and proliferation of ccRCC cells (25). These findings led us to evaluate the level of MCPIP1 during tumor progression and the influence of MCPIP1 downregulation on ccRCC.

The current study showed that the MCPIP1 protein level was decreased together with the increased grading of the tumor; however, the MCPIP1 (ZC3H12A) transcript level is generally low in tumor patient samples, regardless of the tumor grade. In our models, low MCPIP1 levels significantly increased the cell proliferation, tumor growth, and tumor weight of ccRCC. In contrast, MCPIP1 overexpression decreases the proliferation in vitro and reduced tumor growth in vivo as it was demonstrated for breast cancer (22). Interestingly, the analysis of tumors generated by cells carrying a point mutation of D141 (D141N) that completely abolishes MCPIP1 RNase activity (13) demonstrated that the loss of endoribonuclase activity of MCPIP1 significantly increased tumor growth in both investigated models. This observation suggests that, at least in this model, the RNase activity of MCPIP1 is crucial in the process of tumor growth.

ccRCC is a highly vascularized cancer, and dysregulation of angiogenesis is implicated in its development. It has been already presented that high-grade ccRCC had a higher endothelial cell proliferation fraction, larger vessel area, and higher VEGF protein expression (36). Consistent with these findings, the analysis of ccRCC patient–derived tumors revealed that low- and high-grade tumors differ in their blood vessel structures. Grade III and IV tumors were characterized by an increased area of vessels expressing CD31 antigen.

In our study, for the first time, we found that MCPIP1 downregulation in ccRCC led to an increased tumor vasculature. Moreover, our findings demonstrated that tumors formed by ccRCC lacking RNase activity have better vascularization than control and MCPIP1-overexpressing tumors. This observation proves that MCPIP1 might regulate the process of ccRCC vascularization via its RNase activity.

The blood vessel formation process depends on endothelial cells forming blood vessels, tumor cells, and secreted proangiogenic factors. One of the initial responses of endothelial cells to induce angiogenesis is endothelial cell migration to form capillary tube networks (7).

It has been shown that MCPIP1 may promote angiogenesis by increasing the secretion of angiogenic factors, such as VEGF and HIF1α, in HUVECs (23, 24). It has also been demonstrated that MCPIP1 enhances the angiogenic potential of murine mesenchymal stem cells (37). The authors showed that MCPIP1-overexpressing MSCs expressed increased levels of proteins involved in angiogenesis and induction of differentiation (37).

In this study, we provide evidence that MCPIP1 downregulation in ccRCC cells induces the secretion of proangiogenic factors by tumor cells, including VEGF, a key regulator of physiologic angiogenesis (6), IL8, an inducer of endothelial cell proliferation, survival, and capillary tube organization (38) and IL6, a major tumor-promoting cytokine and stimulator of angiogenesis (28). In addition, our results show that low MCPIP1 levels in ccRCC cells influenced the biology of endothelial cells by promoting its migration, capillary network formation and phosphorylation of VE-cadherin, which was required for the control of vascular integrity (27). Our study demonstrated that the secretion of proangiogenic factors is controlled by the RNase activity of MCPIP1, and MCPIP1 overexpression decreases the level of proangiogenic factors. It is in concordance with previous findings showing that MCPIP1 regulates the level of IL6 and IL8 transcripts by binding to stem-loop structures present in 3′UTRs (13, 39). VEGF expression is induced by HIF (hypoxia-inducible factors), and we have previously demonstrated that MCPIP1 overexpression reduces the level of HIFs and transcripts for VEGF (25). However, it was suggested that HIF1-independent mechanisms regulating VEGF expression may also exist. IL8 can rescue the proangiogenic phenotype in HIF1α-deficient cancer cells (40).

It has been reported that the overexpression of VEGF stimulates angiogenesis by upregulating SDF-1 and, thereby, recruits CXCR4-positive proangiogenic myeloid cells (41). SDF-1 is a chemotactic chemokine that promotes the recruitment of endothelial progenitor cells from the bone marrow (42). It has also been documented that SDF-1, together with its receptor, CXCR4, promotes VEGF expression (43) and that VEGF together with SDF-1 synergistically induces neoangiogenesis in tumors (30). Our results revealed that MCPIP1 silencing promotes SDF-1 expression in vitro and in vivo and RNase activity of MCPIP1 is responsible for the level of SDF-1 transcript (Fig. 5). Moreover, MCPIP1 degrades CCAAT/enhancer-binding protein beta (c/EBPβ; ref. 32), a transcription factor regulating the transcription of SDF-1 (33). These observations not only demonstrate the impact of MCPIP1 on tumor development, such as promoting angiogenesis, but also highlight the role of MCPIP1 downregulation in potentiating SDF-1-CXCR4 signaling. The CXCR4/SDF-1 axis can coordinate the metastasis of various cancers (34), and because of the increased expression of CXCR4 in MCPIP1-deficient cells, we examined the influence of MCPIP1 downregulation on the metastatic potential of ccRCC.

We observed that the silencing of MCPIP1 increased the number of circulating shMCPIP1 cells in mouse blood and increased lung metastases, whereas the enhanced MCPIP1 expression in ccRCC cells reduced tumor growth and metastasis in immunodeficient mice (Figs. 1 and 5; Supplementary Fig. S3). Our observation revealed a significantly increased motile activity of ccRCC cells, decreased E-cadherin, and increased level of vimentin and β-catenin. The reduction of E-cadherin correlated with the expression of the Snail and ZEB-2 transcription factors.

Snail and ZEB2 were shown to repress E-cadherin expression and promote invasive properties (44, 45). Overexpression of MCPIP1 decreased the migratory potential and increased E-cadherin levels together with decreased vimentin, β-catenin, Snail, and ZEB-2 expression (Fig. 7), indicating the role of MCPIP1 in acquiring the mesenchymal phenotype of ccRCC cells. It was previously shown that IL1, IL6, and IL8 induce tumor cell migration, invasion, and epithelial–mesenchymal transition and MCPIP1 regulating half time of IL6 and IL1, IL8 transcripts, controlling the level of transcription factors (NFκB, C/EBPβ) and signaling proteins (JNK, Akt) may influence on the appearance of EMT signs (13, 16, 17, 39).

We previously found that E-cadherin expression and the invasive properties of cervical carcinoma depend on the c-Met receptor (46). In this study, we observed an increased level of c-Met and strong phosphorylation of the receptor together with the elevated level of kinase Src after MCPIP1 downregulation (Fig. 6). MCPIP1 overexpression reduced the level of c-Met, its endogenous phosphorylation, and decreased the level of kinase Src in ccRCC (Fig. 7). It has been previously shown that the c-Met gene is a direct target of NFκB, and MET participates in NFκB-mediated cell survival (47). We suspect that the MCPIP1 regulation of NFκB can indirectly influence the level and function of the c-Met receptor.

In conclusion, we provide evidence that the endoribonuclease activity of MCPIP1 regulates tumor angiogenesis and the metastatic cascade and may contribute to ccRCC progression.

Our data highlight the importance of MCPIP1 as a novel tumor suppressor and enable us to postulate that potential MCPIP1 inducers can be used therapeutically for ccRCC in the future.

No potential conflicts of interest were disclosed.

Conception and design: P. Marona, K. Miekus

Development of methodology: P. Marona, K. Miekus

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): P. Marona, J. Górka, Z. Mazurek, W. Wilk, J. Rys, M. Majka, K. Miekus

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P. Marona, J. Jura, K. Miekus

Writing, review, and/or revision of the manuscript: P. Marona, J. Jura, K. Miekus

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): P. Marona, K. Miekus

Study supervision: K. Miekus

Other (histological evaluation of RCC cases studied (senior pathologist)): J. Rys

The authors would like to thank Mr. Tomasz Uśpieński for preparing lentiviral construct with Tet-On system, Prof. Zbigniew Madeja from Department of Cell Biology for permission to use the system to analyze cell motility, and Mrs. Malgorzata Hajduk and Mrs. Katarzyna Dylanowska from the Animal House.

This work was supported in part by research grants from National Science Centre no. 2013/09/D/NZ5/00249 and 2015/19/B/NZ5/01405 (to K. Miekus) and 35p/3/2015 (to K. Miekus) and KNOW BMN 7/2015 (to P. Marona). The Faculty of Biochemistry, Biophysics and Biotechnology of the Jagiellonian University is a beneficiary of the structural funds from the European Union and the Polish Ministry of Science and Higher Education (grants no: POIG.02.01.00-12-064/08 and 02.02.00-00-014/08) and is a partner of the Leading National Research Center (KNOW) supported by the Ministry of Science and Higher Education.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.