Malignant brain tumors are highly aggressive and their prognosis remains extremely poor. Among them, glioblastoma multiforme (GBM) is the most common subtype. Glioma progression involves complex interactions within the tumor microenvironment including the tumor vasculature and infiltrating anti-tumor immune cells. It is known that poor prognosis of GBM patients is associated with robust angiogenesis. We have previously reported that deletion of one of the important MAP Kinase family-extracellular signal-regulated kinase 1 (ERK1) triggers immune cell migration from peripheral into tumor site which significantly reduces GBM progression ^{[1] [2]}. Prior to brain infiltration, immune cells interact with the brain vasculature. As we found that mouse lacking ERK1 formed significantly smaller tumors, it was still not clear whether ERK1 deletion in the tumor vascular architecture might influence the outcome of glioma growth. We hypothesized that ERK1 is required for tumor angiogenesis as one mechanism of promoting tumor growth. To further study the impact of ERK1 on the tumor vasculature in a mouse GBM model, we examined the relative cerebral blood volume (rCBV) in the tumor region. For this, R₂* changes induced by an i.v. administered contrast agent, ferumoxytol, were quantified. Ferumoxytol is clinically applicable and has successfully been used for CBV estimation in the CNS ^{[3] [4]}. Thus, we established a model system to noninvasively monitor brain tumor angiogenesis, and found evidence supporting the role of ERK1 in regulating glioma growth via angiogenesis.Wild-type (WT) and ERK1^-/- C57BL/6 female mice were induced GBM by a stereotactic injection of 2x10⁴ WT-GL261 cells into brain. Experiments were conducted in-vivo after 2 weeks of tumor inoculation. A 9.4T animal MR scanner was used (Biospec 94/20). Following baseline T₂* mapping, 10mg of Fe/kg ferumoxytol was administered using a power injector (0.2ml/min) via a tail vein catheter. T₂* mapping was repeated after 3mins. For the T₂* mapping, images were acquired using a volume birdcage resonator (Rapid Biomed, Germany) and multi-echo gradient-echo (MGE, TR=620ms, TE=2.14-12.14ms, FA=40°, thickness=0.5mm, FOV=17.5x15mm², matrix=105x90) sequence. R₂* maps were calculated offline using custom made Matlab^® scripts. Subtraction of pre-contrast from post-contrast maps yielded ΔR₂* maps (Fig. 1A) from which the tumor was extracted by manual segmentation. After MRI studies, mice were perfused with 4% PFA. Fixed brains were cryo-sectioned at 15µm and stained with anti-CD31 antibody.The ΔR₂* maps from WT (n=2) and ERK1^-/- (n=2) GBM bearing mice clearly show that the ΔR₂* of tumor area from WT group is larger than in ERK1^-/- group (Fig. 1A). Histogram analysis was carried out by manually segmenting the tumor area (Fig. 1A, where the white dotted line depicts tumor) for both groups. It showed that ΔR₂* in ERK1^-/- mice brain tumors is significantly reduced and more homogeneous compared to the WT mice brain tumors (Fig. 1B). Since ΔR₂* is a surrogate for rCBV, the ΔR₂* reduction indicates a decrease in rCBV. When further analyzing the brain microvessels by labeling vessel endothelial layer with anti-CD31 antibodies, we found that intra-tumor vessel density is higher in the WT mice. In contrast, although the vessel density is not as high as in the WT mice, the intra-tumor vessel lumen is larger in the ERK1^-/- mice brain (Fig. 2).Estimation of rCBV by means of ΔR₂* mapping in conjunction with an intravascular contrast agent clearly demonstrated the influence of ERK1 on glioma vasculature. A reduction in rCBV was found in WT-GBM residing in ERK1^-/- hosts, which supports our hypothesis that ERK1 plays pivotal role in the tumor microenvironment. The tumor vessels in GBMs are different from normal blood vessels morphologically and functionally. They are usually disorganized, highly permeable and abnormal in the endothelial walls ^[5]. Both our rCBV and fluorescent staining data suggested that in the WT mice, tumor vessel density increased dramatically while in the absence of ERK1 tumor vessels increased less and had impaired vascular maturation. These findings suggest that the angiogenic vasculature in tumors devoid of ERK1 may be functionally impaired. WT-GBMs in ERK1^-/- hosts form smaller tumors and the tumors were much less perfused, which was shown by a low degree of CD31⁺ tumor vessels. In conclusion, brain tumor-associated neovascularization detected on blood-pool-contrast-enhanced MRI may provide as a useful biomarker that correlates with GBM progression.