Figure 4

Disruption of the CXCL12/CXCR4 axis impaired the neovascularization of rat RG2 glioblastoma in vivo. (A, B) H&E staining of sections collected from orthografts showed more cells sprouting from the tumor core from the control shGFP RG2 (A) than from shrCXCR4 RG2 (B) (red arrow: sprouting RG2 cells). (C, D) Immunohistochemistry using anti-PCNA staining showed more proliferating cells spreading in the sections collected from grafts of shGFP RG2(C) than those from shrCXCR4 (D). Immunohistochemistry using anti-CD31, an endothelium cell marker, showed more CD31-positive microvessels in the grafts derived from shGFP RG2 (E, G) than in those derived from shrCXCR4 (F, H). The CD31 positive tubes were sprouting from the tumor core in the sections from the grafts of shGFPRG2. (red arrows, CD31 positive vessels surrounded by cancer cells). (I, J) The levels of VEGF were reduced in the grafts derived from shrCXCR4-1 RG2 (I), compared to those from shGFP RG2 (J). (K, L) PAS, a marker of basement membrane staining, indicated that more PAS-positive microvessels existed in tumors derived from shGFP (K) than in those derived from shrCXCR4-1 (L) (red arrow, PAS positive tubes). (M, N) Quantitative analysis showed that the density of CD31 positive (M) and PAS-positive (N) microvessels was higher in the tumor derived from shGFP. (O) RT-PCR indicated that disruption of CXCR4 reduced the expression of genes regulating angiogenesis, including VEGF, ANGT1, MMP2, and MMP9, but not VE-cadherin. Arrowheads, the boundary of the tumor core, *result from zymography. *t test, P < 0.05.