VEGF is an important mediator of tumor angiogenesis in malignant lesions in a genetically engineered mouse model of lung adenocarcinoma

Kras^G12D-LSL heterozygous mice were obtained from Jackson Laboratories (Jax West, CA) at approximately 3–4 weeks of age and were maintained by Pfizer La Jolla comparative medicine under guidelines provided by IACUC (Institutional Animal Care and Use Committee). Lung tumors were generated in Kras^G12D-LSL mice, using a recently published protocol [15]. Briefly, adenovirus expressing Cre recombinase (Adeno-Cre;the University of Iowa Gene Transfer Vector Core, Iow, IA) were titrated by Adenoviral Titration Kit (Clontech, CA) using instruction provided by the manufacturer. Prior to administration, Adeno-Cre virus was prepared in 50 ul of plain MEM (minimal essential media; Gibco BRL; life Sciences, CA) supplemented with CaCl2 (10 mM) followed by incubation at room temperature (RT) for 20 minutes. The recipients (n = 10) were anesthetized using Ketamine (Baxter) and Xylazine (Vedco) and the adeno-Cre preparation (2.5 × 10^7 infectious units; IU) was administered intra-nasally. To monitor tumor formation and progression, lung tissue was isolated (n = 1-3) at several time points (4, 8 and 12 wks) post inhalation and were stained with H&E (Hematoxylin and Eosin) using standard protocols in the laboratory [15]. The inhaled mice were randomized at 14 wks post-inhalation and were treated with vehicle, sunitinib (40 mg/kg qd), axitinib (15 mg/kg bid) and PF-210 (40 mg/kg qd) using oral route of administration and formulation protocols as described previously [8]. All the animal study procedures were monitored by the veterinary personnel to comply with guidelines provided by IACUC.

To assess therapeutic response to angiogenic inhibitors, lung lesions were quantified in the recipients by a certified pathologist. As previously described, lesions were categorized as hyperplastic, benign adenoma and adenocarcinoma [15]. Lesion quantification provided two types of analyses in the recipients: 1) percentage of each type of lesion in the recipient lung; 2) percentage of mice carrying these lesions in each treatment. To provide statistical analyses, we applied student’s t test (p < 0.05 considered significant) to compare data between the vehicle vs. each treatment.

Formalin fixed paraffin embedded lung tissues were cut into 5 μm sections and were stained for CD31, desmin, and F4-80 separately. Immunohistochemical staining was performed on Leica Bond III automated machine. Bond polymer refine detection kit was used for desmin and CD31 staining and bond intense R detection was used for F4-80 staining. For CD31 staining, lung sections were incubated for 45 minutes with rabbit anti-CD31 monoclonal antibody (clone SP38, Spring Bioscience, cat # M3384, 1:100 dilution). Desmin was stained by incubating lung section with mouse anti-huDesmin antibody (Dako Cytomation, cat# M0760, 1:1500 dilution) for 15 minutes. VEGFR1 and VEGFR2 was stained using anti-VEGFR1 antibody (abcam, cat# ab2350, 1:400 dilution) and anti-VEGFR2 antibody (cell signaling, cat# 2479, 1:200 dilution) respectively. Finally, F4-80 was stained with biotin anti-mouse F4-80 antibody (eBioscience, Cat # 13-4801-82, 1:75 dilution and 45 minutes incubation at RT). Images of stained-slides were captured using a Nanozoomer instrument (Hamamatsu, Japan) and the data was analyzed using Aperio Imagescope software.

Our strategy to investigate anti-tumor efficacy of AIs in Kras^G12D-LSL mice is depicted in Figure 1A. Kras^G12D-LSL mice were inhaled intranasally with Adeno-Cre at 6–8 weeks of age and were maintained without any further intervention. At 8–10 weeks post inhalation, few mice (n = 1-3) were randomly euthanized to assess tumor formation and progression in the lung (Figure 1B). All the remaining mice were randomized into treatment groups and treated with vehicle, PF-210, axitinib and sunitinib for about 8 weeks. Upon termination of the study, lung tissues were analyzed for tumor lesions using H&E staining (Figure 2). Compared to vehicle-treated group, there was a significant reduction in lung lesion in all the three drugs treated groups. To further understand mechanism of action of AIs, we classified lung lesions into three categories including hyperplasia, benign neoplasia and malignant (adenocarcinoma). Detailed pathology analyses of lesions revealed that hyperplastic lesions were not significantly affected by AIs compared to control-treated animals (Figure 3A). However, the percentage of benign neoplastic lesions was significantly (p < 0.01) inhibited by PF-210 (more than 90%) and also axitinib or sunitinib (more than 50%) compared to vehicle-treated mice. Finally malignant lesions were significantly (p < 0.01) inhibited by all the AIs. Additionally we investigated percentage of mice carrying the above-mentioned lesions (Figure 3B). Irrespective of the type of treatment, all mice carried hyperplastic lesions. While all mice treated with axitinib or sunitinib carried benign neoplasia, only 40% of PF-210 treated animals carried these lesions indicating the potency of this compound. Finally all three AIs reduced frequency of malignant lesions by at least 50% in treated mice (Figure 3B). Overall, two types of analyses (percentage of lesions in the lung and percentage of mice carrying specific lesion) indicate that AIs specifically target advanced lesions (malignant or adenocarcinoma).

To further investigate tumor vasculature, we stained lung tissues with different markers such as CD31 and desmin to stain endothelial cells and smooth muscle cells respectively [16]. Vasculature analysis by CD31 staining showed high density of tumor blood vessels in adenoma and adenomacarcinoma lesions in the vehicle group (Figure 4). Moreover, these vessels were desmin positive indicative of a mature vasculature in these lesions. In contrast, tumor lesions in AI treated groups had less number of blood vessels further suggesting that vasculature is the main target of these AIs. Additionally, vasculature was found to be more fragmented compared to the blood vessels in vehicle treated mice. Similar to CD31 staining, all three AIs (especially PF-210 and axitinib) targeted smooth muscle cells suggesting that not only blood vessels but also other components of vasculature are affected. . We also investigated the effects of AIs on the expression of VEGFR1 and VEGFR2 which play an important role in angiogenesis and tumor progression [1]. High levels of VEGFR1 was observed on tumor cells in vehicle treated mice (Figure 4) that is consistent with the expression of VEGFR1 on tumor cells isolated from Kras mutant NSCLC tumors in an earlier report [17]. Interestingly, sunitinib and PF-210, but not axitinib, inhibited VEGFR1 expression on tumor cells (Figure 4). Compared to vehicle treated tumors that expressed abundant levels of VEGFR2 on blood vessels, all three AIs inhibited VEGFR2 expression on the tumor vasculature further providing a mechanism for the anti-angiogenic activity of these compounds. Overall, these results suggest that inhibition of angiogenesis is the main mechanism by which AIs suppress growth of benign and malignant lesions in this model of NSCLC.

Tumor associated macrophages (TAMs) are a key component of tumor microenvironment and have been implicated in tumor progression and angiogenesis. It has been shown that NSCLC patients with higher density of TAMs have lower median relapse-free survival (7 months) compared to patients whose tumors had lower density of TIMs (26 months survival) [18]. Macrophage staining (using F4-80) indicated infiltration of these TAMs in the lung in vehicle treated mice (Figure 4). Treatment with AIs particularly sunitinib and axitinib was associated with lower density of TAMs further suggesting an additional mechanism for anti-tumor efficacy of AIs in Kras^G12D-LSL lung tumors.