Host Cxcr2-Dependent Regulation of Pancreatic Cancer Growth, Angiogenesis, and Metastasis

Abstract

Pancreatic ductal adenocarcinoma (PDAC) manifests aggressive tumor growth and early metastasis. Crucial steps in tumor growth and metastasis are survival, angiogenesis, invasion, and immunosuppression. Our prior research showed that chemokine CXC- receptor-2 (CXCR2) is expressed on endothelial cells, innate immune cells, and fibroblasts, and regulates angiogenesis and immune responses. Here, we examined whether tumor angiogenesis, growth, and metastasis of CXCR2 ligands expressing PDAC cells are regulated in vivo by a host CXCR2-dependent mechanism. C57BL6 Cxcr2^-/- mice were generated following crosses between Cxcr2^-/+ female and Cxcr2^-/- male. Cxcr2 ligands expressing Kirsten rat sarcoma (KRAS-PDAC) cells were orthotopically implanted in the pancreas of wild-type or Cxcr2^-/- C57BL6 mice. No significant difference in PDAC tumor growth was observed. Host Cxcr2 loss led to an inhibition in microvessel density in PDAC tumors. Interestingly, an enhanced spontaneous and experimental liver metastasis was observed in Cxcr2^-/- mice compared with wild-type mice. Increased metastasis in Cxcr2^-/- mice was associated with an increase in extramedullary hematopoiesis and expansion of neutrophils and immature myeloid precursor cells in the spleen of tumor-bearing mice. These data suggest a dynamic role of host CXCR2 axis in regulating tumor immune suppression, tumor growth, and metastasis.

Figure 1

The host Cxcr2 deletion does not affect tumor growth but enhances metastasis. A: Intravital luciferase images demonstrating the development of orthotopic tumors in wild-type (WT) and Cxcr2^−/− mice inoculated with Kirsten rat sarcoma (KRAS)–pancreatic ductal adenocarcinoma–green fluorescent protein cells. B: Graph showing the weight of tumors harvested from WT and Cxcr2^−/− mice. Each dot on the graph represents data point from an individual animal. C: Representative images of hematoxylin and eosin (H&E) staining, showing the histopathology of tumors. D: Representative images of H&E of the liver, peritoneum wall, and diaphragm in WT and Cxcr2^−/− mice, demonstrating metastasis. Arrows indicate metastatic cells. E: Graphs showing quantitation of spontaneous micrometastasis in the liver of WT and Cxcr2^−/− mice. Cxcr2^−/− mice have higher metastasis in the liver in comparison with those in WT mice. Error bars represent SD. Statistical significance determined by U-test. n = 9 WT mice (A); n = 8 Cxcr2^−/− mice (A). ∗P < 0.05. Scale bars = 100 μm (C and D).

Figure 2

Host Cxcr2 deletion inhibits microvessel density, in situ cell proliferation, and enhanced apoptosis. Representative photomicrographs of hematoxylin and eosin (H&E) staining, along with corresponding CD31 (A), Ki-67 (B), and cleaved caspase 3 (CC3; C) immunohistochemistry (IHC) staining, and graphs representing quantitation of each IHC stain. Each dot on the graph represents data point from an individual animal. Error bars represent SD. Statistical significance determined by U-test. ∗P < 0.05. Scale bars = 100 μm (A–C). WT, wild type.

Figure 3

The frequency of myeloid cells in the pancreas of tumor-bearing wild-type (WT) and Cxcr2^−/− mice. A and B: Representative photomicrographs demonstrating hematoxylin and eosin (H&E) staining, immunohistochemistry (IHC) staining, and their quantitation for Ly6 (G + C)–positive myeloid cells (A) and myeloperoxidase (MPO)–positive neutrophils (B) in the tumors of WT and Cxcr2^−/− tumor-bearing mice. Each dot on the graph represents data point from an individual animal. Error bars represent SD. Statistical significance determined by U-test. C–E: Effect of Cxcr2 deletion on the percentage of neutrophils [CD11b⁺Ly6C⁻Ly6G^hi (%CD11b⁺) as determined by flow cytometry; C], myeloid-derived suppressor cells [MDSCs; CD11b⁺Ly6C⁺Ly6G⁺ (%CD11b⁺); D], and T-regulatory (Treg) cells [CD3⁺CD4⁺CD25⁺ (%CD3⁺); E] in the pancreas of WT and Cxcr2^−/− mice. F: Bar graph comparing the ratios of cytotoxic T cells/MDSCs in the pancreas of tumor-bearing WT and Cxcr2^−/− mice. Each dot on the graph represents a data point from an individual animal. Error bars represent SD. Statistical significance determined by U-test. ∗P < 0.05. Scale bars = 100 μm (A and B).

Figure 4

Splenomegaly and extramedullary hematopoiesis in the spleens of tumor-bearing Cxcr2^−/− mice. A: Representative images of spleens resected from tumor-bearing wild-type (WT) and Cxcr2^−/− mice. B: Graph demonstrating increased spleen weight in the tumor-bearing Cxcr2^−/− mice versus the WT group. Each dot on the graph represents a data point from an individual animal. C: Representative images of histopathologic evaluation of hematoxylin and eosin–stained spleens of non–tumor-bearing and tumor-bearing mice showing higher extramedullary hematopoiesis (EMH) in the spleens of Cxcr2^−/− mice versus the WT genotype. Follicular zones are marked with white dashed lines. Arrows indicate EMH. Insets show EMH at higher magnification. D: Representative images of Hema 3 staining of cytospins prepared from the splenocytes of WT and Cxcr2^−/− non–tumor-bearing and tumor-bearing mice. Scale bars = 100 μm (C and D).

Figure 5

Increased accumulation of neutrophils and immature myeloid precursor cells in the spleens of tumor-bearing Cxcr2^−/− mice. A: Representative photomicrographs demonstrating immunohistochemistry (IHC) staining for Ly6 in cytospins prepared from splenocytes of wild-type (WT) and Cxcr2^−/− non–tumor-bearing and tumor-bearing mice. B: Ly6 in the red pulp of spleens of WT and Cxcr2^−/− tumor-bearing mice. C: Representative photomicrographs of immunohistochemical staining for myeloperoxidase (MPO), showing marked neutrophils in the spleens of Cxcr2^−/− tumor-bearing mice, and graphs depicting the quantitation of myeloperoxidase IHC stain. Each dot on the graph represents data points from an individual animal. Error bars represent SD. Statistical significance determined by U-test. D and E: Flow cytometry analysis for the evaluation of the percentage of neutrophils (D) and myeloid-derived suppressor cells (MDSCs; E) inside the spleens of WT and Cxcr2^−/− tumor-bearing mice. F: Bar graph showing the ratios of cytotoxic T cells/MDSCs in the spleen of tumor-bearing WT and Cxcr2^−/− mice. Each dot on the graph represents data point from an individual animal. Error bars represent SD. ∗P < 0.05. Scale bars: 100 μm (A–C). H&E, hematoxylin and eosin.

Figure 6

Host Cxcr2 deletion enhanced experimental metastasis by generating a premetastatic niche inside the livers with the accumulation of neutrophils. A: Representative images of hematoxylin and eosin (H&E) staining of livers, demonstrating metastatic lesions in the livers. Arrows indicate metastatic lesions. B and C: Graphs showing quantitation of macrometastasis and micrometastasis in wild-type (WT) and Cxcr2^−/− mice. Error bars represent SD. Statistical significance determined by U-test. D and E: Representative images showing immunohistochemistry of myeloperoxidase (MPO) and graph showing quantitation of increased accumulation of MPO-positive cells inside the livers of Cxcr2^−/− hosts. Statistical significance determined by U-test. ∗P < 0.05. Scale bars: 500 μm (A); 100 μm (D).

Figure 7

Diagrammatic representation of the effect of depletion of Cxcr2 signaling on pancreatic cancer. A: Host Cxcr2 deletion does not effect the overall tumor burden despite enhanced apoptosis in the tumors. Angiogenesis is significantly inhibited in the tumors of Cxcr2^−/− hosts. B: Host Cxcr2 depletion results in enhanced metastasis to livers. Peripheral accumulation of immature myeloid cells inside the enlarged spleens of Cxcr2^−/− tumor-bearing hosts may result in immune evasion. MDSC, myeloid-derived suppressor cell; PDAC, pancreatic ductal adenocarcinoma; Treg, T regulatory.