Control of metastatic niche formation by targeting APBA3/Mint3 in inflammatory monocytes

Abstract

Cancer metastasis is intricately orchestrated by both cancer and normal cells, such as endothelial cells and macrophages. Monocytes/macrophages, which are often co-opted by cancer cells and promote tumor malignancy, acquire more than half of their energy from glycolysis even during normoxic conditions. This glycolytic activity is maintained during normoxia by the functions of hypoxia inducible factor 1 (HIF-1) and its activator APBA3. The mechanism by which APBA3 inhibition partially suppresses macrophage function and affects cancer metastasis is of interest in view of avoidance of the adverse effects of complete suppression of macrophage function during therapy. Here, we report that APBA3-deficient mice show reduced metastasis, with no apparent effect on primary tumor growth. APBA3 deficiency in inflammatory monocytes, which strongly express the chemokine receptor CCR2 and are recruited toward chemokine CCL2 from metastatic sites, hampers glycolysis-dependent chemotaxis of cells toward metastatic sites and inhibits VEGFA expression, similar to the effects observed with HIF-1 deficiency. Host APBA3 induces VEGFA-mediated E-selectin expression in the endothelial cells of target organs, thereby promoting extravasation of cancer cells and micrometastasis formation. Administration of E-selectin-neutralizing antibody also abolished host APBA3-mediated metastatic formation. Thus, targeting APBA3 is useful for controlling metastatic niche formation by inflammatory monocytes.

Keywords: APBA3/Mint3; macrophage; metastasis.

Fig. 1.

Host APBA3 deficiency affects metastasis. (A) Lung metastatic burden of i.v.-injected B16F10 cells 14 d after inoculation in WT (n = 11) or Apba3^−/− mice (n = 10). (B) Survival of WT or Apba3^−/− mice after i.v. injection with B16F10 cells. n = 7 per group. (C) Lung metastatic burden of i.v.-injected LLC cells at 14 d after inoculation in WT or Apba3^−/− mice. n = 10 per group. (D) Survival of WT or Apba3^−/− mice after i.v. injection with LLC cells. n = 9 per group. (E) Lung metastatic burden of i.v.-injected 4T1 cells at 14 d after inoculation in WT (n = 11) or Apba3^−/− mice (n = 9). (F) Survival of WT or Apba3^−/− mice after i.v. injection with 4T1 cells. n = 6 per group. Data represent mean ± SEM. *P < 0.05, **P < 0.01, as determined by the Mann–Whitney U test.

Fig. 2.

Host APBA3 deficiency decreases metastatic colonization and extravasation. (A and B) Macroscopic analysis of metastatic growth of B16F10 cells in the lungs of WT and Apba3^−/− mice. (A) Representative photo of metastatic lungs from WT and Apba3^−/− mice. (B) Metastatic growth was quantified and categorized by measuring the diameter of foci. n = 3 (18 to 74 foci) per group. (C and D) Colonization of B16F10 cells in the lungs of WT or Apba3^−/− mice 1 or 24 h after i.v. inoculation. (C) Representative photos of B16F10 cell colonies (red) in the lungs. (Scale bars, 100 μm.) (D) The number of tumor foci was counted. n = 7 per group. *P < 0.05, as determined by the Mann–Whitney U test. (E and F) Confocal microscopic analysis of B16F10 cells (green) and endothelial cells (CD31; red) in the lungs of WT or Apba3^−/− mice. (E) Representative photos. (Scale bars, 20 μm.) (F) The ratio of intravascular and extravascular B16F10 cells was analyzed. n = 45 to 68 cells. P < 0.05, as determined by Fisher’s exact test. Data are represented as mean ± SEM.

Fig. S1.

Host APBA3 deficiency decreases metastatic colonization and extravasation of LLC and 4T1 cells. (A) Representative photos of lung sections with B16F10 cell colonies (red). Colonization of B16F10 cells in the lungs of WT or Apba3^−/− mice 1 or 24 h after i.v. inoculation. (Scale bars, 100 μm.) (B) Representative 3D images of B16F10 cells (green) and CD31-positive endothelial cells (red) in the lungs of WT or Apba3^−/− mice 1 or 24 h after i.v. inoculation. (Scale bars, 20 μm.) (C) Colonization of LLC cells in the lungs of WT or Apba3^−/− mice 1 or 24 h after i.v. inoculation. The numbers of tumor foci were counted. n = 6 or 7 per group. *P < 0.05, as determined by the Mann–Whitney U test. (D) Confocal microscopy-based analysis of LLC and endothelial cells in the lungs of WT or Apba3^−/− mice. The ratio of intravascular and extravascular LLC cells was analyzed. n = 52 or 53 cells. P < 0.05, as determined by Fisher’s exact test. (E) Colonization of 4T1 cells in the lungs of WT or Apba3^−/− mice 1 or 24 h after i.v. inoculation. The numbers of tumor foci were counted. n = 6 or 7 per group. **P < 0.01, as determined by the Mann–Whitney U test. (F) Confocal microscopy-based analysis of 4T1 and endothelial cells in the lungs of WT or Apba3^−/− mice. The ratio of intravascular and extravascular 4T1 cells was analyzed. n = 52 to 65 cells. P < 0.05, as determined by Fisher’s exact test.

Fig. 3.

Host APBA3 deficiency decreases tumor-induced E-selectin expression in endothelial cells. (A and B) Immunostaining of E-selectin in the lungs at 6 h after i.v. injection with B16F10 cells (green). (A) Representative photo of E-selectin (red) and B16F10 cells (green) in the lungs of WT mice. (Scale bar, 20 μm.) (B) Tumor-induced E-selectin–positive cells were counted. n = 6. (C) Immunostaining of E-selectin (red) and CD31 (green) in the lungs of WT mice 6 h after i.v. injection with B16F10 cells. (Scale bars, 10 μm.) (D and E) Analysis of lung metastatic burden at 14 d after i.v. injection with E-selectin–neutralizing antibody or rat IgG followed by B16F10 cell inoculation in WT or Apba3^−/− mice. (D) Representative photos of metastatic lungs. (E) The number of metastatic foci on the lungs was counted. n = 5 or 6 per group. Data are represented as mean ± SEM. *P < 0.05, **P < 0.01, as determined by the Mann–Whitney U test. NS, not significant.

Fig. S2.

Host APBA3 deficiency decreases tumor-induced E-selectin mRNA expression in the lungs. Quantitative real-time PCR analysis of E-selectin (A), P-selectin (B), VCAM (C), and ICAM (D) mRNA expression in the lungs of WT or Apba3^−/− mice at the indicated times after i.v. injection with B16F10 cells (n = 3). Data are represented as mean ± SEM. *P < 0.05.

Fig. 4.

VEGFA inhibition and APBA3 deficiency in myeloid cells hamper E-selectin induction. (A) Immunostaining analysis of E-selectin in the lungs at 6 h after i.v. injection with VEGFA-neutralizing antibody or goat IgG followed by B16F10 cell inoculation in WT mice. Tumor-induced E-selectin–positive cells were counted. n = 5 per group. (B and C) Immunostaining analysis of VEGFA and cellular markers in the lungs of WT mice 6 h after i.v. injection with B16F10 cells. (B) Representative photos of VEGFA (green) and cellular markers (red) in the lungs. (Scale bars, 10 μm.) (C) The ratios of cellular marker-positive cells and VEGFA-positive cells were analyzed. n = 3 (21 to 130 cells) per group. (D) Immunostaining analysis of CD68 (red) and E-selectin (green) expression in the lungs of WT mice 6 h after i.v. injection with fluorescein-labeled B16F10 cells (blue). (Scale bar, 10 μm.) (E) Quantified analysis of E-selectin–positive cells in the lungs of LysM-cre (control) or LysM-cre; Apba3^fl/fl mice 6 h after i.v. injection with B16F10 cells. n = 5. (F) Lung metastatic burden in LysM-cre (control) or LysM-cre; Apba3^fl/fl mice 14 d after i.v. injection with B16F10 cells. n = 11 per group. Data represent mean ± SEM. *P < 0.05, **P < 0.01, as determined by the Mann–Whitney U test.

Fig. S3.

VEGFA depletion in B16F10 cells did not affect tumor-induced E-selectin expression in metastatic lungs. (A) VEGFA mRNA levels in control (shLacZ-treated) and VEGFA-depleted (shVEGFA-treated) B16F10 cells. n = 3 per group. Data shown represent mean ± SD. (B) Quantified analysis of E-selectin–positive cells in the lungs of mice 6 h after i.v. injection with control and VEGFA-depleted B16F10 cells. n = 5 per group. Data shown represent mean ± SEM.

Fig. S4.

Depletion of monocyte/macrophage-suppressed metastasis. Lung metastatic burden in WT or Apba3^−/− mice 14 d after i.v. injection with liposome–clodronate or control liposome (L-PBS) followed by B16F10 cell inoculation. (A) Representative photos of metastatic lungs. (B) The number of metastatic foci on the lungs was counted. n = 6 per group. Data are represented as mean ± SEM. *P < 0.05, **P < 0.01. NS, not significant.

Fig. 5.

APBA3 deficiency attenuates chemotaxis of inflammatory monocytes to metastatic sites. (A) Flow cytometric analysis of IMs in the bone marrow and spleen of WT or Apba3^−/− mice. n = 6 per group. (B and C) Flow cytometric analysis of IMs in the peripheral blood (B) and lungs (C) of WT or Apba3^−/− mice 4 h after i.v. injection with B16F10 cells or PBS. n = 4 or 5 per group. (D and E) Flow cytometric analysis of resident monocytes (D) and neutrophils (E) in the lungs at 4 h after i.v. injection with B16F10 cells or PBS in WT or Apba3^−/− mice. n = 4 or 5 per group. (F) Flow cytometric analysis of IMs in the lungs at 4 h after i.v. injection with B16F10 cells in WT or LysM-cre; Apba3^fl/fl mice. n = 5 per group. Data represent mean ± SEM. *P < 0.05, **P < 0.01, as determined by the Mann–Whitney U test.

Fig. S5.

Flow cytometric strategy for definition of myeloid cells. Inflammatory monocytes, resident monocytes, and neutrophils were characterized by the cell-surface markers CD45, CD11b, CD115, Gr-1, Ly-6C, Ly-6G, and CCR2.

Fig. S6.

APBA3 deficiency does not affect expression of CCL2 and CCR2. (A and B) Quantitative analysis of CCL2 in the lungs (A) or peripheral blood (B) of WT or Apba3^−/− mice at 4 h after i.v. injection with B16F10 cells or PBS. n = 4 per group. Error bars indicate SEM. (C) Flow cytometric analysis of CCR2 expression on IMs from WT or Apba3^−/− mice. APC, allophycocyanin.

Fig. 6.

APBA3 deficiency in IMs attenuates glycolysis-dependent chemotaxis, VEGF expression, and metastasis. (A) Flow cytometric analysis of peritoneal IMs at 6 h after i.p. injection with recombinant CCL2 in LysM-cre or LysM-cre; Apba3^fl/fl mice. n = 5 per group. (B) Flow cytometric analysis of peritoneal IMs at 6 h after i.p. injection with recombinant CCL2, following administration with a glycolytic inhibitor, 2-deoxyglucose, in WT mice. n = 4 or 5 per group. (C) Flow cytometric analysis of IMs in the lungs at 4 h after i.v. injection with B16F10 cells in WT or Apba3^−/− mice treated with or without 2-DG. n = 5 per group. (D) Quantitative analysis of VEGFA-positive cells in the lungs of vehicle- or 2-DG–administered WT or Apba3^−/− mice 6 h after i.v. injection with B16F10 cells. n = 6 per group. (E) Quantitative analysis of E-selectin–positive cells in the lungs of vehicle- or 2-DG–administered WT or Apba3^−/− mice 6 h after i.v. injection with B16F10 cells. n = 6 per group. (F) The number of metastatic foci and the representative photos of lungs in vehicle- or 2-DG–administered WT or Apba3^−/− mice 14 d after i.v. injection with B16F10 cells. n = 6 per group. (G) ATP levels in WT, Apba3^−/−, and LysM-cre; Hif1a^fl/fl IMs with or without 2-DG. n = 3 per group. (H) mRNA levels of glycolysis-related genes in WT, Apba3^−/−, and LysM-cre; Hif1a^fl/fl IMs. n = 3 per group. (I) Vegfa mRNA levels in WT, Apba3^−/−, and LysM-cre; Hif1a^fl/fl IMs. n = 3 per group. (J) Lung metastatic burden in WT or Apba3^−/− mice 14 d after tail vein injection with B16F10 cells with or without WT IMs. n = 9 per group. (A–F and J) Error bars indicate the SEM. (G–I) Error bars indicate the SD. *P < 0.05, **P < 0.01, as determined by the Mann–Whitney U test (A–F and J) or Student’s t test (G–I).

Fig. S7.

2-DG treatment of B16F10 cells in vitro did not affect subsequent metastasis formation. Lung metastatic burden with B16F10 cells treated with vehicle or 2-DG for 1 h in vitro before inoculation. (A) Representative photos of metastatic lungs. (B) The numbers of metastatic foci in the lungs were counted. n = 6 per group. Data shown represent mean ± SEM.

Fig. 7.

Host APBA3 governs a metastatic microenvironment in spontaneous lung metastasis. (A) Subcutaneous tumor growth of Lewis lung carcinoma cells in WT (n = 6) or Apba3^−/− mice (n = 7). (B) Tumor growth of 4T1 cells in the fat pad of WT or Apba3^−/− mice. n = 6 per group. (C) Spontaneous lung metastatic burden of LLC cells 21 d after inoculation in WT (n = 14) or Apba3^−/− mice (n = 12). (D) Spontaneous lung metastatic burden of 4T1 cells 28 d after inoculation in WT or Apba3^−/− mice. n = 6 per group. (E) Flow cytometric analysis of IMs in the lungs at 14 d after s.c. injection with LLC cells in WT or Apba3^−/− mice. n = 5 per group. (F) Flow cytometric analysis of IMs in the lungs at 7 d after fat pad injection with 4T1 cells in WT or Apba3^−/− mice. n = 5 per group. (G) Immunostaining analysis of VEGFA in the lungs of LLC tumor-bearing WT or Apba3^−/− mice at 14 d after tumor inoculation. n = 6 per group. (H) Immunostaining analysis of VEGFA in the lungs of 4T1 tumor-bearing WT or Apba3^−/− mice at 7 d after tumor inoculation. n = 5 per group. (I) Immunostaining analysis of E-selectin in the lungs of LLC tumor-bearing WT or Apba3^−/− mice at 14 d after tumor inoculation. n = 5 per group. (J) Immunostaining analysis of E-selectin in the lungs of 4T1 tumor-bearing WT or Apba3^−/− mice at 7 d after tumor inoculation. n = 5 per group. Data represent mean ± SEM. *P < 0.05, **P < 0.01, as determined by the Mann–Whitney U test.

Fig. 8.

Schematic illustration of APBA3-mediated metastasis. Cancer and/or stromal cells at the metastatic sites secrete chemokines (e.g., CCL2). Inflammatory cells, including inflammatory monocytes, migrate toward chemokines. APBA3 in IMs promotes glycolysis-dependent chemotaxis to the metastatic sites. APBA3 also promotes VEGFA transcription in IMs. IMs at metastatic sites secrete VEGFA, which induces E-selectin expression in endothelial cells and vascular permeability. Cancer cells take advantage of E-selectin in endothelial cells, facilitating efficient extravasation and metastasis.

Fig. S8.

Host APBA3 deficiency decreased vascular permeability in metastatic lungs. (A) Representative photos of the lungs of WT or Apba3^−/− mice at the indicated times after i.v. injection with B16F10 cells. Mice were injected i.v. with fluorescein-labeled dextran (red) 30 min before they were killed. (B) Quantitative analysis of dextran permeability in the lungs of WT or Apba3^−/− mice at the indicated times after i.v. injection with B16F10 cells. n = 6 per group. (C) Analysis of dextran permeability in the lungs of WT mice 6 h after i.v. injection with a VEGFA-neutralizing antibody or goat IgG, followed by inoculation with B16F10 cells. n = 6 per group. Data shown represent mean ± SEM. **P < 0.01. (D) Analysis of dextran permeability in the lungs of WT mice 6 h after i.v. injection with an E-selectin–neutralizing antibody or rat IgG, followed by B16F10 cell inoculation. n = 6 per group. Data shown represent mean ± SEM.