The mechanism of lncRNA‐CRNDE in regulating tumour‐associated macrophage M2 polarization and promoting tumour angiogenesis

Abstract M2 macrophages can promote liver cancer metastasis by promoting tumour angiogenesis; however, the mechanism underlying macrophage polarization has not been completely revealed. In this study, we mainly explored the mechanism underlying long non‐coding RNA‐CRNDE (lncRNA‐CRNDE) in regulating M2 macrophage polarization and promoting liver cancer angiogenesis. The expression of CRNDE was up‐regulated or down‐regulated in THP‐1 cells (CRNDE‐/‐‐THP‐1 cells and pcDNA3.1‐CRNDE‐THP‐1). THP‐1 cells were co‐cultured with liver cancer cell line H22, and M2 polarization was induced in THP‐1 by IL‐4/13 to simulate tumour‐induced macrophage polarization. As a result, after CRNDE overexpression, THP‐1 cell viability was up‐regulated, the expression of M2 membrane marker CD163 was up‐regulated, and the proportion of F4/80 + CD163+ cells was also up‐regulated. ELISA assay showed that the expression of M2 markers (including TGF‐β1 and IL‐10) and chemokines (including CCl22 and CCL22) was up‐regulated, and the expression of key signals (including STAT6, JAK‐1, p‐AKT1, and Arg‐1) was also up‐regulated, which were significantly different compared with the control group (Con). In addition, the intervention effect of CRNDE on THP‐1 was consistent between co‐culture with H22 cells and IL‐4/13 induction assay. The induced M2 THP‐1 cells were co‐cultured with HUVEC. As a result, THP‐1 cells with CRNDE overexpression can promote the migration and angiogenesis of HUVEC cells in vitro and simultaneously up‐regulate the expression of Notch1, Dll4 and VEGFR2, indicating that THP‐1 M2 polarization induced by CRNDE could further promote angiogenesis. The H22 cell tumour‐bearing mouse model was constructed, followed by injection of CRNDE anti‐oligosense nucleotides and overexpression plasmids to interfere CRNDE expression in tumour‐bearing tissues. Consequently, down‐regulation of CRNDE could down‐regulate tumour volume, simultaneously down‐regulate the expression of CD163 and CD31 in tissues, decrease the expression of key proteins (including JAK‐1, STAT‐6, p‐STAT6 and p‐AKT1), and down‐regulate the expression of key angiogenesis‐related proteins (including VEGF, Notch1, Dll4 and VEGFR2). In this study, we found that CENDE could indirectly regulate tumour angiogenesis by promoting M2 polarization of macrophages, which is also one of the mechanisms of microenvironmental immune regulation in liver cancer.


| INTRODUC TI ON
to secrete cytokines to improve the tissue microenvironment, thereby providing suitable microenvironment for the distant metastasis and survival of tumour cells. 4 In addition, M2-TAM can also secrete TGFβ, IL-10, etc to inhibit tumour immunity. 5,6 In multiple stages of tumour progression, macrophages can exhibit different phenotypes. The formation of tumour microenvironment can promote the transformation of M1 to M2. M1 macrophage is dominant in early tumour tissues, which can inhibit angiogenesis and activate tumour immunity. After tumour progression, M2 macrophage becomes the dominant type in the microenvironment infiltration, and the increased M2 proportion is parallel to the tumour vessel density. 7,8 Recent studies have demonstrated that the correlation between tumour cells and the microenvironment can be achieved through non-coding RNA. Liver cancer is a type of malignant tumour with relatively poor prognosis. Similarly, exosomes derived from liver cancer cells can affect tumour microenvironment, supporting for tumour progression. 9,10 CRNDE currently has few reports on the regulation of tumour microenvironment, so it is necessary to further explore its expression and mechanism in liver cancer, which can provide a new reference for immunotherapy of liver cancer. Therefore, this research starts with the polarization of macrophages, acting CRNDE promotes the M2 type polarization of macrophages and promotes the role and mechanism of tumour angiogenesis.
At present, CRNDE has been rarely investigated in regulating tumour microenvironment. Therefore, it is necessary to further explore its expression and mechanism in liver cancer, which can provide new reference for immunotherapy in liver cancer. To this end, in this study, we investigated the role of mechanism of CRNDE in promoting tumour angiogenesis by promoting macrophage M2 polarization.

| Effects of CRNDE on THP-1-induced angiogenesis
After inducing by IL4/13 or co-culturing with H22 cells for 72h, THP-1 cells were co-cultured with HUVEC cells to observe the angiogenesis of HUVEC cells.

| Change of HUVEC migration ability by wound-healing assay
After induction and co-culture, THP-1 cells were co-cultured with HUVEC cells in Transwell chambers. HUVEC cells were inoculated to reach 80% confluency. Afterwards, a straight line was drawn in the middle, followed by co-culture of THP-1 cells and HUVEC cells.
After 48 hours, the migration of HUVEC cells was observed. The results were processed by Image J software and shown as migration rate (%).

| Viability of HUVEC cells by CCK-8 assay
THP-1 cells and HUVEC cells were co-cultured at 37°C in an incubator containing 5% CO 2 . After co-culture for 0h, 12h, 24h, 36h and 48 h, cell viability was tested. After replacing the medium without culture medium, 10μl of CCK-8 reagent was added for additional incubation for 4 h. The absorbance was subsequently measured using a microplate reader at a wavelength of 450nm. The blank medium was used as the control to calculate cell viability (results were shown as %).

| Blood formation capacity of HUVEC cells by in vitro tube formation assay
After pre-coating Matrigel into the lower layer of Transwell chambers, HUVEC cells were inoculated and further co-cultured with THP-1 after cell adherence, followed by observation of formation of in vitro blood vessels in HUVEC cells after 48h.

| Chick embryo chorioallantoic membrane assay
The chick embryo was incubated in a sterile incubator for 72 h at 37.8°C with humidity of 60%. Sterile surgical scissors were used to punch holes in chick embryos, and 6ml of egg white was sucked by using a 10-ml syringe, followed by injection of THP-1 cells into the hole. Five chick embryos were set in each group, and the opening was sealed with scotch tape, followed by incubation for 48h.
Image-Pro Plus 6.1 software was used to analyse the area of blood vessels.

| Effects of CRNDE on tumour-bearing mice
H22 cells in the logarithmic phase were collected and washed with PBS for two times. Cell density was adjusted to 5x10 7 /ml and followed by injection of 0.2ml of cell suspension into the armpits of the forelegs of nude mice. Nude mice were fed in a clean environment, followed by observation of growth of nude mice and

| Detects key protein expression by Western blot
After carbon dioxide asphyxiation, the tumour tissues were taken from mice and ground with liquid nitrogen, followed by cell lysis and protein extraction using RIPA lysate. The BCA kit was used for protein quantification, and protein expression detection was the same as described above. The expression of M2 polarization key proteins (including STAT6, p-STAT6, JAK-1, AKT1 and p-AKT1) and angiogenesis regulatory proteins (including Notch1, Dll-4 and VEGFR-2) was detected.

| Statistical analysis
SPSS19.0 software was used for statistical analysis. The measurement data were expressed as mean ± standard deviation (x±s). Oneway ANOVA was used for comparison between multiple groups, and the SNK test was used for comparison between groups. A P <.05 indicated statistical significance.
In the study, CRNDE silencing decreased the viability of THP-1 cells, which was significantly different from the THP-1-Con group (P <.05). However, the overexpression of CRNDE significantly increased cell vitality, which was significantly higher than the THP-1-Con group (P <.05), indicating that CENDE was significantly correlated with the vitality of THP-1 cells ( Figure 1A). CD163 is a main marker of M2 macrophages. Flow cytometry was used to F I G U R E 1 Effects of CRNDE on M2 polarization of THP-1 cells. A, Cell viability results (n = 5): The cell viability of the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group, while the cell viability of the THP-1-shRNA group was significantly lower than that of the THP-1-Con group. The expression of CRNDE was associated with the viability of THP-1 cells. Comparison with THP-1-Con, *P <.05; **P <.01. B and C, The proportion of F4/80 + CD163+M2 macrophages by flow cytometry (n = 5): The proportion of F4/80 + CD163+M2 macrophages in the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group, while the proportion of F4/80 + CD163+M2 macrophages in the THP-1-shRNA group was significantly lower than that of the THP-1-Con group. Comparison between groups, *P <.05; **P <.01. D, CD163 expression by immunofluorescence staining (n = 3): The expression of CD163 in the THP-1-pcDNA3.1 group was significantly higher than that in the THP-1-Con group, while the expression of CD163 in the THP-1-shRNA group was lower than that of the THP-1-Con group detect the proportion of F4/80 + CD163+ cells. As a result, the proportion of F4/80 + CD163+ cells in the THP-1-pcDNA3.1 group was significantly higher than that of THP-1-Con group, while the proportion of F4/80 + CD163+ cells in the THP-1-shRNA group was significantly lower than that of THP-1-Con group, indicating that CRNDE affected M2 polarization ( Figure 1B,C). IF staining also showed that the expression of CD163 in the THP-1-pcDNA3.1 group was significantly higher than THP-1-Con group, suggesting that CRNDE affected the expression of CD163. And high expression of CRNDE could promote the expression of CD163, thereby promoting M2 polarization ( Figure 1D). In the detection of markers, the expression levels of M2 macrophage markers (including VEGF, IL-10, TGF-β1, CCl22 and CCl24) in the THP-1-pcDNA3.1 group were significantly up-regulated, which was higher than the THP-1-Con group, while the expression level in the THP-1-shRNA group was lower than THP-1-Con group, indicating that CRNDE was associated with the expression of markers of M2 macrophages (Figure 2A-E

| Effects of CRNDE on M2 polarization of macrophages induced by liver cancer cells
To further verify the tumour-induced M2 polarization of macrophages, H22 cells were co-cultured with THP-1 to investigate the effects of CRNDE on tumour-induced macrophage polarization.

F I G U R E 2 Effects of CRNDE on M2 polarization of THP-1 cells and the mechanism.
A-E, Detection of M2 macrophage marker expression by CRNDE (n = 5): The expression of M2 macrophage markers in the THP-1-pcDNA3.1 group was significantly increased, which was significantly different from the THP-1-Con group. Meanwhile, the expression of M2 macrophage markers in THP-1-shRNA group was lower than that of the THP-1-Con group. Comparison between groups, *P <.05; **P <.01. F-I, Effects of CRNDE on JAK-STAT6 and AKT1 signal expression (n = 3): The expression of JAK1, STAT6, p-STAT6, AKT1 and p-AKT1 in the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group. Meanwhile, the expression level in the THP-1-shRNA group was lower than that of the THP-1-Con group. Comparison between groups, *P <.05; **P <.01 Similar with M2 polarization induced by IL4/13, cell viability was significantly increased after CRNDE overexpression, while cell viability was decreased after CRNDE silencing, which was significantly different from the THP-1-Con group ( Figure 3A). In the detection of F4/80 + CD163+ cells, the proportion of M2 macrophages in the THP-1pcDNA3.1 group was significantly higher than that of the THP-1-Con, and the proportion of the THP-1-shRNA group was lower than that of the THP-1-Con ( Figure 3B,C). The IF staining of CD163 was significant. To be specific, CRNDE overexpression caused the up-regulated expression of CD163, while down-regulation of CRNDE led to decreased expression of CD163 ( Figure 3D). In the detection of M2 cell markers, the expression of M2 macrophage markers (including VEGF, IL-10, TGF-β1, CCl22 and CCl24) in the THP-1-pcDNA3.1 group was significantly higher than that of THP-1-Con group, while the expression of M2 macrophage markers in the THP-1-shRNA group was lower than that of the THP-1-Con group, indicating that CRNDE was associated with M2 macrophage marker expression ( Figure 4A-E). In the detection of the mechanism, the protein expression of JAK1, STAT6 and p-STAT6 in the THP-1-pcDNA3.1 group was significantly higher than that in the THP-1-Con group ( Figure 4F-I). The cell viability of the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group. The expression of CRNDE was associated with THP-1 cell viability. Comparison with THP-1-Con, *P <.05; **P <.01. B and C, Proportion of F4/80 + CD163+M2 macrophages by flow cytometry (n = 5): The proportion of F4/80 + CD163+M2 macrophages in the THP-1-pcDNA3.1 cells was significantly higher than that of THP-1-Con group, while the proportion in the THP-1-shRNA group was significantly lower than that of THP-1-Con group. Comparison between groups, *P <.05; **P <.01. D, CD163 expression by immunofluorescence staining (n = 3): The expression of CD163 in the THP-1-pcDNA3.1 group was significantly higher than that in the THP-1-Con group, while the expression of CD163 in the THP-1-shRNA group was lower than that of the THP-1-Con group cell viability induced by THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group, while HUVEC cell viability induced by THP-1-shRNA group was lower than that of THP-1-Con group ( Figure 5A). In the wound-healing assay, the migration rate of HUVEC cells induced by THP-1-pcDNA3.1 group was 79.87 ± 4.55%, while the migration rate of HUVEC cells induced by THP-1-Con group was 67.76 ± 5.44%, which was 41.65 ± 8.65% in the THP-1-shRNA group ( Figure 5B). In vitro vessel formation assay showed that the lumen formation of HUVEC cells induced by the THP-1-pcDNA3.1 group, the cell nodes, loops and loop areas were significantly higher than those in the THP-1-Con group, while the THP-1-shRNA group was lower than that of the THP-1-Con group.

| Effects of IL4/13-induced M2 macrophages on angiogenesis in HUVEC
Chick embryo chorioallantoic membrane assay also showed that THP-1 cells in the THP-1-pcDNA3.1 group could promote blood vessel formation, and the number of capillaries was significantly higher than that in the THP-1-Con group. The number of formed blood vessels in THP-1-shRNA was significantly lower than that of the THP-1-Con group ( Figure 5C). The expression of key angiogenesis proteins (including VEGFR2, Notch1 and Dll4) was significantly up-regulated in the HUVEC induced by THP-1-pcDNA3.1, while the expression in the THP-1-shRNA group was lower than that in the THP-1-Con group ( Figure 5D,E).

| Effects of M2 macrophages induced by liver cancer cell H22 on angiogenesis in HUVEC
Angiogenesis induction assay was performed on M2-THP-1 after co-culture with H22 cells. The results were similar to IL-4/13induced M2 macrophages. The HUVEC cell viability induced by THP-1-pcDNA3.1 group was significantly higher than that of THP-1-Con group ( Figure 6A). The migration rate of HUVEC cells

F I G U R E 4
Effects of CENDE on the expression of M2 macrophage markers and key protein expression. A-E, Detection of M2 macrophage marker expression (n = 5): The expression of M2 macrophage markers was significantly increased in the THP-1-pcDNA3.1 group, which was significantly different from the THP-1-Con group. Meanwhile, the expression in the THP-1-shRNA group was lower than that of THP-1-Con group. Comparison between groups, *P <.05; **P <.01. F-I, Effects of CRNDE on JAK-STAT6 and AKT1 signal expression (n = 3): The expression of JAK1, STAT6, p-STAT6, AKT1 and p-AKT1 in the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group. Meanwhile, the expression level in the THP-1-shRNA group was lower than that of the THP-1-Con group. Comparison between groups, *P <.05; **P <.01 induced by THP-1-pcDNA3.1 group was 74.87 ± 8.56%, while the migration rate of HUVEC cells induced by THP-1-Con group was 60.16 ± 5.91%, which was 38.34 ± 6.12% in the THP-1-shRNA group ( Figure 6B). In vitro vessel formation assay showed that the tube formation of HUVEC cells induced by the THP-1-pcDNA3.1 group, the cell nodes, loops and loop areas were significantly higher than those in the THP-1-Con group. Chick embryo chorioallantoic membrane assay also showed that THP-1 cells in the THP-1-pcDNA3.1 group could promote blood vessel formation, and the number of capillaries was significantly higher than that in the THP-1-Con group ( Figure 6C). The levels of key angiogenesis protein (including VEGFR2, Notch1 and Dll4) were significantly up-regulated in HUVEC induced by THP-1-pcDNA3.1, while the expression of was lower in THP-1-shRNA group than that in THP-1-Con group ( Figure 6D,E).

| Effects of CRNDE on angiogenesis in tumourbearing mice
To further confirm the effects of CRNDE on angiogenesis in mice, liver cancer cell line H22 was used to construct tumour-bearing mouse model and CRNDE anti-oligosense was used to inhibit the expression of CRNDE. As a result, the expression of CD163 and CD31 in the tissues of the anti-CRNDE group was significantly lower than that in the Con group. CD163 is a marker of M2 macrophages, and CD31 is a marker of tumour blood vessels. The down-regulated expression indicates weaker angiogenesis, which is positively correlated with CD163 ( Figure 7A) The cell viability of THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group. CRNDE expression was associated with THP-1 cell viability. Comparison with THP-1-Con group, *P <.05; **P <.01. B, Results of HUVEC cell migration ability (n = 5): The cell migration ability of the THP-1-pcDNA3.1 group was significantly increased than that of the THP-1-Con group. C, In vitro tube formation assay of HUVEC and chick embryo chorioallantoic membrane assay (n = 5): the vessel formation of the THP-1-pcDNA3.1 group was significantly increased than that of THP-1-Con group. D and E, Expression of key signal proteins of angiogenesis (n = 3): The expression of Notch1, Dll4 and VEGFR2 proteins in the THP-1-pcDNA3.1 group was significantly higher than that in the THP-1-Con group, and the expression in the THP-1-shRNA group was lower than the THP-1-Con group. Comparison with THP-1-Con, *P <.05; **P <.01 p-AKT1) was significantly higher in mouse tissues in the pcDNA3.1-CRNDE group than that in the Con group ( Figure 7B,C). Meanwhile, the expression of angiogenesis-related protein (including VEGFR2, Notch1 and Dll4) was also up-regulated ( Figure 7D,E). The cell viability of the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con. Comparison with THP-1-Con, *P <.05; **P <.01. B, HUVEC cell migration assay (n = 5): The cell migration capacity of THP-1-pcDNA3.1 group was significantly enhanced than that of THP-1-Con group. C, In vitro tube formation assay of HUVEC and chick embryo chorioallantoic membrane assay (n = 5): The vessel formation ability of THP-1-pcDNA3.1 group was significantly enhanced than THP-1-Con group. D and E, Expression of key signal proteins for angiogenesis (n = 3): The protein expression of Notch1, Dll4 and VEGFR2 in the THP-1-pcDNA3.1 group was significantly higher than that of the THP-1-Con group, and the expression in the HP-1-shRNA group was lower than that of the THP-1-Con group. Comparison with THP-1-Con, *P <.05; **P <.01  : H&E staining showed no obvious tissue lesions among the three groups of mice. The level of CD163 and CD31 in the anti-CRNDE group was lower than Con group, while the expression of CD163 and CD31 was higher in the pcDNA3.1-CRNDE group than the Con group. B and C, Expression of key protein of M2 macrophage polarization (n = 3): The protein expression in the pcDNA3.1-CRNDE group was significantly higher than that in the Con group, while the expression of the anti-CRNDE group was lower than that of the Con group. Comparison with Con, * P <.05; **P <.01. D and E, Expression of key proteins in angiogenesis (n = 3): The protein expression in the pcDNA3.1-CRNDE group was significantly higher than that in the Con group, while the expression of the anti-CRNDE group was lower than that of the Con group. Comparison with Con, *P <.05; **P <.01 the viability of THP-1, demonstrating that CRNDE plays a vital role on the viability of THP-1. CRNDE also exerted an effect on the proportion of F4/80 + CD163+M2 macrophages after induction. When CRNDE was overexpressed, the cell proportion was significantly upregulated, which was associated with the high expression of CD163.

| D ISCUSS I ON
IF staining also showed that CD163 was regulated by CRNDE. In tumour immune research, JNK signalling has been reported to play an important role in M2 polarization. When JNK1 inhibitors are used, the proportion of M2 cells is significantly down-regulated, and the expression of Arg-1 and Mrc1 is also down-regulated. JNK1 plays an important role. 22 In its downstream STAT receptor, the expression of STAT6 is associated with M2 polarization. 23

| CON CLUS ION
In this study, we have found that lncRNA-CRNDE can promote the M2 polarization of macrophages through the high expression of CD163 to further promote tumour angiogenesis, which is one of the mechanisms by which CRNDE promotes liver cancer progression.

E TH I C A L A PPROVA L
The study approvaled with Ethics Committee.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.