Activin A, a Novel Chemokine, Induces Mouse NK Cell Migration via AKT and Calcium Signaling

Abstract

Natural killer (NK) cells can migrate quickly to the tumor site to exert cytotoxic effects on tumors, and some chemokines, including CXCL8, CXCL10 or and CXCL12, can regulate the migration of NK cells. Activin A, a member of the transforming growth factor β (TGF-β) superfamily, is highly expressed in tumor tissues and involved in tumor development and immune cell activation. In this study, we focus on the effects of activin A on NK cell migration. In vitro, activin A induced NK cell migration and invasion, promoted cell polarization and inhibited cell adhesion. Moreover, activin A increased Ca²⁺, p-SMAD3 and p-AKT levels in NK cells. An AKT inhibitor and Ca²⁺ chelator partially blocked activin A-induced NK cell migration. In vivo, exogenous activin A increased tumor-infiltrating NK cells in NS-1 cell solid tumors and inhibited tumor growth, and blocking endogenous activin A with anti-activin A antibody reduced tumor-infiltrating NK cells in 4T-1 cell solid tumors. These results suggest that activin A induces NK cell migration through AKT signaling and calcium signaling and may enhance the antitumor effect of NK cells by increasing tumor-infiltrating NK cells.

Keywords: AKT signaling; activin A; calcium signaling; migration; natural killer cells.

Figure 1

Effect of activin A on viability and cytotoxicity of mouse NK cells. (A) Cells sorted by magnetic beads were stained with Giemsa. Scale bar = 10 µm. (B) The cells sorted by magnetic beads were identified by flow cytometry with anti-CD49b and NKp46 antibodies. (C) The viability of mouse NK cells was examined by CCK-8 assay (n = 3). (D) The killing rate of NK cells was assayed by flow cytometry, after treatment with the culture medium (Control), 20 ng/mL IL-2, 10 ng/mL activin A or 20 ng/mL IL-2 + 10 ng/mL activin A (n = 3). * p < 0.05, compared with the control group (mean ± S.D., one-way ANOVA).

Figure 2

Effect of activin A on the migration of mouse NK cells. (A) Transwell assay was used to detect migration of NK cells treated with culture medium (Control), 10 ng/mL activin A, 100 ng/mL CXCL12 or 10 ng/mL activin A + 100 ng/mL CXCL12 (n = 3). Scale bar = 30 µm. * p < 0.05, ** p < 0.01, compared with the control group; # p < 0.05, compared with the activin A group (one-way ANOVA; mean ± S.D.). (B) NK cell migration was examined by microfluidic chips after treatment with culture medium (Control), 20 ng/mL activin A, 200 ng/mL CXCL12 or 20 ng/mL activin A + 200 ng/mL CXCL12. Data related to cell migration were analyzed by NIH ImageJ software (n = 20). Scale bar = 100 µm. ** p < 0.01, compared with the control group; # p < 0.05, ## p < 0.01, compared with the activin A group; §§ p < 0.01, compared with the CXCL12 group (one-way ANOVA; mean ± SEM).

Figure 3

Effects of activin A on the invasion and adhesion of mouse NK cells. (A) NK cell invasion was determined by Matrigel-coated transwell assay after treatment with culture medium (Control), 10 ng/mL activin A, 100 ng/mL CXCL12 or 10 ng/mL activin A + 100 ng/mL CXCL12 (n = 3). Scale bar = 50 µm. * p < 0.05; ** p < 0.01, compared with the control group; ## p < 0.01, compared with the activin A group; §§ p < 0.01, compared with the CXCL12 group (one-way ANOVA; mean ± S.D.). (B) Real-time cell adhesion was assessed by RTCA. NK cells were treated with culture medium (Control), 10 ng/mL activin A, 100 ng/mL CXCL12 or 10 ng/mL activin A + 100 ng/mL CXCL12 (n = 4). * p < 0.05; ** p < 0.01, compared with the control group (one-way ANOVA; mean ± S.D.).

Figure 4

Effects of activin A on F-actin polarization and expression of migration-related proteins of mouse NK cells. (A) The nuclei were stained with DAPI (blue) and the cell was stained with FITC-labeled phalloidin that preferentially labels filamentous actin (F-actin) (green). NK cells (white arrows) were treated with culture medium (Control), 20 ng/mL activin A, 200 ng/mL CXCL12 or 20 ng/mL activin A + 200 ng/mL CXCL12. Scale bar = 10 µm. (B) Levels of β-catenin, vimentin and MMP2 protein in mouse NK cells were examined by Western blotting, after treating the cells with culture medium (lane 1), 10 ng/mL activin A (lane 2), 100 ng/mL CXCL12 (lane 3) or 10 ng/mL activin A + 100 ng/mL CXCL12 (lane 4) for 2 h. The levels of protein expression were normalized against those of GAPDH (n = 3). * p < 0.05; ** p < 0.01, compared with the control group (one-way ANOVA; mean ± S.D.).

Figure 5

Effect of activin A on the expression of signaling proteins in mouse NK cells. (A) Levels of p-Smad3, Smad3, p-AKT, and AKT proteins in mouse NK cells were examined by Western blotting, after treating the cells with culture medium (lane 1), 5 ng/mL activin A (lane 2) or 10 ng/mL activin A (lane 3) for 1 h. The levels of these proteins’ expression were normalized against those of GAPDH (n = 3). * p < 0.05; ** p < 0.01, compared with the control group (one-way ANOVA; mean ± S.D.). (B) After pretreating the cells with DMSO or DMSO-dissolved GSK690693 (10 μM) for 1 h, the migration of mouse NK cells was detected by the transwell assay (n = 3). Scale bar = 30 µm. * p < 0.05, compared with the DMSO group (Student’s t test; mean ± S.D.).

Figure 6

Effect of activin A on calcium flux in mouse NK cells. (A) Kinetics of calcium mobilization was assessed using mouse NK cells loaded with Fluo-4 AM, after treating the cells with medium (Control), 10 ng/mL activin A, 100 ng/mL CXCL12 or 10 ng/mL activin A + 100 ng/mL CXCL12. The Ca²⁺ level is represented by the Fluo-4 signal intensity normalized to the baseline (F/F0). The graph shows the peak values of calcium signal upon stimulation under the different treatments (n = 3). ** p < 0.01, compared with the control group (one-way ANOVA; mean ± S.D.). (B). Migration of mouse NK cells pretreated with BAPTA-AM (10 μM) was examined using a transwell assay (n = 3). Scale bar = 30 µm. ** p < 0.01, compared with the DMSO group (Student’s t test; mean ± S.D.).

Figure 7

Effects of exogenous activin A on the infiltration of NK cells in NS-1 cell solid tumors. (A) The expression of activin-βA mRNA in NS-1 and 4T-1 cells was detected by RT-PCR (n = 3). ** p < 0.01, compared with the NS-l group (Student’s t test; mean ± S.D.). (B) The level of activin A protein in the culture supernatants of NS-1 and 4T-1 cells was determined by ELISA (n = 3). ** p < 0.01, compared with the NS-l group (Student’s t test; mean ± S.D.). (C–E) The NS-1 cell solid tumors were injected with 1 μL normal saline (Control) or 20 ng activin A in 1 μL normal saline every 24 h, three times continuously, and then (C) the body weight of mice was recorded, (D) tumor volume was measured, (E) and the percentage of NK cells infiltrating the tumor was examined by flow cytometry (n = 4). * p < 0.05, compared with the control group (n = 4) (Student’s t test; mean ± S.D.).

Figure 8

Effect of blocking endogenous activin A on the infiltration of NK cells in 4T-1 cell solid tumors. (A–C) The 4T-1 cell solid tumors were injected with 200 ng IgG or anti-activin A antibody in 1 μL normal saline every 24 h, three times continuously, and then (A) the body weight of mice was recorded, (B) tumor volume was measured, (C) and the percentage of NK cells infiltrating the tumor was examined by flow cytometry (n = 4). * p < 0.05, compared with the IgG group (Student’s t test; mean ± S.D.). (D) The infiltrating NK cells (white arrows) in the tumor tissue were stained with FITC-labeled anti-CD49b antibody (green) and DAPI (blue). Scale bar = 50 µm.