E-M, an Engineered Endostatin with High ATPase Activity, Inhibits the Recruitment and Alternative Activation of Macrophages in Non-small Cell Lung Cancer

Abstract

Endostatin recently was reported by our laboratory to possess ATPase activity that is indispensable for its anti-angiogenesis and anti-tumor effects. An engineered endostatin, E-M, which owns higher ATPase activity exhibits stronger inhibitory effects on angiogenesis. Tumor-associated macrophages (TAMs), especially M2-polarized TAMs, contribute to tumor progression by promoting tumor cell proliferation, metastasis, angiogenesis, and immunosuppression, thus emerging as crucial targets for therapeutic intervention. Endostatin reportedly modulated functions of TAMs, but the detailed mechanisms remain unclear. Here, in our study, we demonstrated that E-M exhibited stronger inhibitory effects on macrophages than endostatin and other low ATPase mutants, which indicates that the ATPase activity is required for the inhibitory effects of endostatin on TAMs. Moreover, we elucidated that endostatin co-receptor, nucleolin and integrin α5β1, overexpressed on the surface of M2 macrophages, facilitated the internalization of E-M via the caveolae/lipid raft- and clathrin-dependent pathways. E-M inhibited the migration of TAMs through blockade of p38 MAP kinase and Erk1/2 signaling pathways, and prevented the alternative activation of TAMs. As a result, TAM-induced tumor cell proliferation and angiogenic activities in vitro were dramatically suppressed by E-M. In a transplanted non-small cell lung cancer model, E-M remarkably decreased the density of intratumoral macrophages and blood vessels, leading to tumor regression. This study unravels a novel mechanism of endostatin on regulating TAM recruitment and polarization, and suggests that E-M is a remarkably promising and multifunctional anti-tumor agent.

Keywords: ATPase activity; alternative activation; cell recruitment; endostatin; tumor-associated macrophage.

FIGURE 1

E-M exhibits strong inhibitory effects on macrophages and interacts with both nucleolin and integrin α5β1. (A) Immunofluorescent images showing that uPAR, nucleolin and integrin α5 were co-localized on BMDM surface. (B) Flow cytometric analysis showing the presence of nucleolin and integrin α5 on BMDM surface. (C) Representative images of effects of endostatin, E-M and low ATPase mutants (E176A, K96A, and K96R) on BMDM migration determined by modified Boyden chamber assay; Scale bar = 100 μm. BMDMs were pre-treated with 40 μg/mL different recombinant proteins for 1 h. Then 2% FBS was added to the lower chamber to induce BMDM migration. (D) Quantified result of (C). (E) Immunoprecipitation showing both nucleolin and integrin α5 were able to interact with E-M. (F) Pull-down assay showing the interaction between E-M and nucleolin or integrin α5. E-M and BSA were linked to CNBr-activated sepharose, respectively. BMDM cell lysates were incubated with these sepharose. Then samples were immunoblotted with antibodies against nucleolin and integrin α5. Data were representative of mean ± SD from at least three independent experiments. P-value: One-way ANOVA; ^∗∗∗P < 0.001; ns, not significant.

FIGURE 2

E-M can be internalized into macrophages via cell surface nucleolin and integrin α5β1 in caveolae/lipid raft- and clathrin-dependent pathways. (A) Western blot indicating the internalization of E-M into BMDMs and Raw 264.7 cells in a time-dependent manner. The culture medium was replaced with fresh DMEM without FBS. Then 5 μg/mL E-M was added to the medium. After 30 or 60 min, cells were washed with ice cold acid buffer (pH = 3.5) twice and PBS once. Western blot was used to detect the internalization of E-M. (B) Flow cytometric analysis showing the internalization of E-M into BMDMs. (C) Representative images of immunofluorescence showing the localization of Rh-E-M in BMDMs. Blue: DAPI, green: DiO, red: Rh-E-M; Scale bar = 50 μm. (D) Blocking nucleolin and integrin α5 with respective antibodies and detecting the internalization of E-M into BMDMs. (E) Flow cytometric analysis showing the internalization of E-M in BMDMs after blockade of nucleolin and integrin α5 with respective antibodies. (F) Western blot indicating effects of caveolin inhibitor nystatin (50 μg/mL) and clathrin inhibitor chlorpromazine (6 μg/mL) on E-M internalization. (G) Flow cytometric analysis showing the effects of nystatin and chlorpromazine on E-M internalization. (H) Flow cytometric analysis showing expression levels of nucleolin and integrin α5 on BMDM surface after the treatment with A549 CM. (I) Quantified result of MFI in (H). (J) Flow cytometric result displaying the internalization of Rh-E-M into BMDMs after the treatment of A549 CM. (K) Flow cytometric analysis showing expression levels of nucleolin and integrin α5 on BMDM surface after the treatment with IL-4 (20 ng/mL) and IL-13 (20 ng/mL) for 6, 12, and 24 h. Data were representative of mean ± SD from at least three independent experiments. P-value: Student’s t-test for two groups and One-way ANOVA for more than two groups; ^∗∗P < 0.01, ^∗∗∗P < 0.001; ns, not significant.

FIGURE 3

E-M exhibits its inhibitory effects on TAM motility both in vitro and in vivo. (A) Representative images of effects of E-M on BMDM migration determined by modified Boyden chamber assay; Scale bar = 100 μm. BMDMs were pre-treated with 40 μg/mL recombinant E-M for 1 h. Then concentrated A549 CM with or without E-M was added to the lower chamber to induce BMDM migration. (B) Quantified result of (A). (C) Representative images of the density of F4/80⁺ macrophages recruited by A549 CM in Matrigel plugs; Scale bar = 100 μm. (D) Quantitation of recruited F4/80⁺ macrophages in (C). (E) Quantified result of recruited macrophages by B16-F10 CM. (F) Western blot showing the effect of E-M on the activation of p38 and Erk1/2 induced by A549 CM. BMDMs were starved overnight in DMEM without FBS. Before stimulating BMDMs with A549 CM for 10 min, BMDMs were pre-treated with 20 and 40 μg/mL E-M for 1 h. Then cell lysates were immunoblotted for detecting the activation of p38 and Erk1/2. (G) The effects of p38 inhibitor SB203580 (10 μM) and Erk1/2 inhibitor U0126 (10 μM) on BMDM migration induced by A549 CM. BMDMs were pre-treated with SB203580 and U0126 for 30 min. Then concentrated A549 CM with or without inhibitors was added to the lower chamber to induce BMDM migration. Date were representative of mean ± SD from at least three independent experiments. P-value: One-way ANOVA; ^∗P < 0.05, ^∗∗∗P < 0.001; ns, not significant.

FIGURE 4

E-M inhibits the switch of macrophage polarization toward M2 phenotype. (A) qRT-PCR detecting genes of M2 markers in BMDMs after the treatment of A549 CM and E-M. (B) Western blot showing the activation of STAT3, STAT6 and the expression of Arg-1 in BMDMs after the treatment of A549 CM either alone or in the presence of E-M. (C) Western blot displaying the effect of E-M on the expression of Arg-1 in BMDMs in A549 cell-BMDM co-culture system. (D) Representative fluorescent images showing the number of A549-GFP cells in the A549 cell-BMDM co-culture system; Scale bar = 100 μm. A549-GFP cells were co-cultured with BMDMs for 12 h either alone or in the presence of E-M. (E) Quantitative result of A549-GFP cell number in (D). (F) BMDMs were pre-treated with E-M for 1 h. Then IL-4 (40 ng/mL) was added to induce M2 macrophages. qRT-PCR was used to detect genes of M2 markers. (G) Flow cytometric result showing the effects of B16-F10 CM either alone or in the presence of E-M on alternative activation of BMDMs. (H) Western blotting result showing the effects of B16-F10 CM either alone or in the presence of E-M on the expression of Arg-1 in BMDMs. Data were representative of mean ± SD from at least three independent experiments. P-value: One-way ANOVA; ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001, ^##P < 0.01, ^###P < 0.001; ns, not significant.

FIGURE 5

E-M suppresses the pro-angiogenic effects of TAMs. (A) qRT-PCR result showing the effect of E-M on the expression of pro-angiogenic cytokines in BMDMs which were treated with A549 CM. BMDMs were pre-treated with E-M for 1 h and then treated with A549 CM either alone or in the presence of E-M for 12 h. (B) Representative images showing BMDM CM from different groups on tubule formation. BMDMs were treated with A549 CM either alone or in the presence of E-M for 12 h. Then BMDMs were starved overnight and BMDM CM from each group was collected. These CM was used to treat MS1 cells for detection of tubule formation ability; Scale bar = 100 μm. (C) Quantified result of (B). (D) Representative images showing the effect of BMDM CM from different groups on SVEC4-10 migration determined by Boyden chamber assay; Scale bar = 100 μm. (E) Quantified result of (D). (F) Representative images displaying the effect of BMDM CM from different groups on SVEC4-10 motility determined by wound healing assay; Scale bar = 100 μm. (G) Quantified result of (F). (H) qRT-PCR result indicating the effect of A549 CM either alone or in the presence of E-M on HIF-1α expression. (I) Western blot showing the effect of A549 CM either alone or in the presence of E-M on HIF-1α expression. (J) Western blot showing the effect of A549 CM either alone or in the presence of E-M on activation of Akt and 4E-BP1 in BMDMs. (K) Western blot showing the effect of A549 CM either alone or in the presence of PI3K inhibitor LY294002 (10 μM) or mTOR inhibitor KU0063794 (10 μM) on HIF-1α expression in BMDMs. Data were representative of mean ± SD from at least three independent experiments. P-value: One-way ANOVA; ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001; ns, not significant.

FIGURE 6

E-M inhibits the recruitment of macrophages and tumor angiogenesis in vivo. (A) Tumor volumes of A549 after treatment of PBS or E-M for 6 days (n = 5 mice/group). (B) Representative images of immunofluorescence showing the density of F4/80⁺ macrophages in A549 tumor tissues; Blue: DAPI, green: F4/80; Scare bar = 100 μm. (C) Quantitation of the density of macrophages in (B). (D) Representative images of immunofluorescence displaying the density of blood vessels in A549 tumor tissues via detecting CD31⁺ blood vessels; Blue: DAPI, green: CD31; Scare bar = 100 μm. (E) Quantified result of (D). (F) A549 tumor growth in mice treated with PBS and E-M (n = 5 mice/group). When the tumor volume reached 100 mm³, PBS and E-M (12 mg/kg) were i.v. administered every other day. (G) A549 tumor weight in mice treated with PBS and E-M. (H) Representative images of immunofluorescence showing the density of F4/80⁺ macrophages in A549 tumor tissues; Blue: DAPI, green: F4/80; Scare bar = 100 μm. (I) Quantitation of the density of F4/80⁺ macrophages in (H). (J) Representative images of immunofluorescence displaying the tumor angiogenesis in A549 tumor tissues via detecting CD31⁺ blood vessels; Blue: DAPI, green: CD31; Scare bar = 100 μm. (K) Quantified result of (J). Data were representative of mean ± SD or SEM for animal experiment. P-value: Student’s t-test; ^∗∗P < 0.01, ^∗∗∗P < 0.001; ns, not significant.