Cancer-Associated Stromal Cells Promote the Contribution of MMP2-Positive Bone Marrow-Derived Cells to Oral Squamous Cell Carcinoma Invasion

Abstract

Tumor stromal components contribute to tumor development and invasion. However, the role of stromal cells in the contribution of bone marrow-derived cells (BMDCs) in oral squamous cell carcinoma (OSCC) invasion is unclear. In the present study, we created two different invasive OSCC patient-derived stroma xenografts (PDSXs) and analyzed and compared the effects of stromal cells on the relation of BMDCs and tumor invasion. We isolated stromal cells from two OSCC patients: less invasive verrucous OSCC (VSCC) and highly invasive conventional OSCC (SCC) and co-xenografted with the OSCC cell line (HSC-2) on green fluorescent protein (GFP)-positive bone marrow (BM) cells transplanted mice. We traced the GFP-positive BM cells by immunohistochemistry (IHC) and detected matrix metalloproteinase 2 (MMP2) expression on BM cells by double fluorescent IHC. The results indicated that the SCC-PDSX promotes MMP2-positive BMDCs recruitment to the invasive front line of the tumor. Furthermore, microarray analysis revealed that the expressions of interleukin 6; IL-6 mRNA and interleukin 1 beta; IL1B mRNA were higher in SCC stromal cells than in VSCC stromal cells. Thus, our study first reports that IL-6 and IL1B might be the potential stromal factors promoting the contribution of MMP2-positive BMDCs to OSCC invasion.

Keywords: MMP2; bone marrow-derived cells (BMDCs); oral squamous cell carcinoma invasion; patient-derived stromal cell xenograft (PDSX); stromal factor IL-6; stromal factor IL1B.

Figure 1

Histological characterization of VSCC-PDSX and SCC-PDSX: (A) Illustration of the establishment of OSCC-PDSX models. Bone marrow transplantation (BMT) with GFP-positive bone marrow cells is performed on lethal whole-body radiated nude mice. After 4 weeks, tumor transplantation is performed which allows tumors to grow for 4 weeks. Then, mice are sacrificed, and tumor tissues are harvested. (B) Representative images of hematoxylin and eosin (H&E) staining of VSCC-PDSX; (C,D) High magnified images of VSCC-PDSX (C) Skin side, (D) Bone; (E) Representative images of hematoxylin and eosin (H&E) staining of SCC-PDSX; (F,G) High magnified images of VSCC-PDSX (F) Skin side, (G) Bone side. Dotted lines represent the boundary of the tumor (Tu) and the stroma (St). VSCC, verrucous oral squamous cells carcinoma; SCC, conventional oral squamous cell carcinoma, PDSX, patient-derived stromal cells xenograft; OSCC, oral squamous cell carcinoma.

Figure 2

SCC-PDSX promotes GFP-positive BMDCs infiltration in the periphery of tumor: (A) Representative images of GFP IHC: upper panel: skin side; lower panel: bone side (left panel: VSCC-PDSX; right panel: SCC-PDSX); (B) The rate of infiltrating BMDCs into the skin side of tumors; (C) The rate of infiltrating BMDCs into the bone side of tumors. Dotted lines represent the boundary of the tumor (Tu) and the stroma (St). VSCC, verrucous oral squamous cells carcinoma; SCC, conventional oral squamous cell carcinoma, PDSX, patient-derived stromal cells xenograft. Statistical analyses were performed using Student’s t-test, 2-tailed; **** p < 0.0001. Dots in the plots represent the mean value of each mouse. Data are presented as the mean ± SD, n = 5.

Figure 3

SCC-PDSX shows MMP2-positive cells’ infiltration in the periphery of tumor: (A) Representative images of MMP2 IHC: upper panel: skin side; lower panel: bone side (left panel: VSCC-PDSX; right panel: SCC-PDSX); (B) The rate of infiltrating MMP2-positive cells into the skin side of tumors; (C) The rate of infiltrating MMP2-positive cells into the bone side of tumors. Dotted lines represent the boundary of the tumor (Tu) and the stroma (St). VSCC, verrucous oral squamous cells carcinoma; SCC, conventional oral squamous cell carcinoma, PDSX, patient-derived stromal cells xenograft. Statistical analyses were performed using Student’s t-test, 2-tailed; ** p < 0.01, **** p < 0.0001. Dots in the plots represent the mean value of each mouse. Data are presented as the mean ± SD, n = 5.

Figure 4

GFP/MMP2 double-positive BMDCs localized in the periphery of SCC-PDSX: (A) Representative images of GFP/MMP2 double-IHC staining on the skin side of tumors; (B) Representative images of GFP/MMP2 double-IHC staining at the skin side of tumors; (C) Comparison of GFP/MMP2 double-positive cells per one field (×400) in skin side of tumors; (D) Comparison of GFP/MMP2 double-positive cells per one field (×400) in bone side of tumors. Dotted lines represent the boundary of the tumor (Tu) and the stroma (St). VSCC, verrucous oral squamous cells carcinoma; SCC, conventional oral squamous cell carcinoma, PDSX, patient-derived stromal cells xenograft. Statistical analyses were performed using Student’s t-test, 2-tailed; **** p < 0.0001. Dots in the plots represent the mean value of each mouse. Data are presented as the mean ± SD, n = 5.

Figure 5

Identification of potential stromal factors involved in the relation of BMDCs and tumor invasion: (A) Flow chart of genes screening related to bone marrow cells migration; (B) The biological process enrichment analysis of common differentially expressed genes (DEGs) of VSCC stromal cells and SCC stromal cells; (C) Top 10 hub genes in bone marrow cells migration analyzed using protein-to-protein (PPI) network and Cytoscape software. (D) Heatmap presentation of the expression level of top 10 hub genes. VSCC, verrucous oral squamous cells carcinoma; SCC, conventional oral squamous cell carcinoma.