WO2015157655A2

WO2015157655A2 - Method for predicting response to vegf targeted therapeutics

Info

Publication number: WO2015157655A2
Application number: PCT/US2015/025341
Authority: WO
Inventors: Lawrence M. Pfeffer; Jo Meagan GARNER
Original assignee: University Of Tennessee Research Foundation
Priority date: 2014-04-10
Filing date: 2015-04-10
Publication date: 2015-10-15
Also published as: WO2015157655A3

Abstract

The presently disclosed subject matter involves methods of predicting response in a subject to treatment for a disease with a VEGF targeted therapeutic, and for evaluating and/or monitoring treatment for a disease involving use of a VEGF targeted therapeutic in a subject. Such methods involve determining a level of ANGPTL4 expression in a biological sample from the subject; and comparing the level of ANGPTL4 expression in the sample with a reference, wherein the subject is predicted to be a likely responder or non-responder based on the ANGPTL4 expression in the sample relative to the reference. Administering a VEGF targeted therapeutic to the subject is initiated or continued when the subject is predicted to be a likely responder.

Description

METHOD FOR PREDICTING RESPONSE TO VEGF TARGETED THERAPEUTICS

RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application Serial No.

61/977,724, filed April 10, 2015, the entire disclosures of which are incorporated herein by this reference.

STATEMENT OF GOVERNMENT INTEREST

[0002] The presently disclosed subject matter was made with U.S. Government support contract no. W81XWH-11-1-0533 awarded by the Department of Defense. The U.S. Government has certain rights in this presently disclosed subject matter.

TECHNICAL FIELD

[0003] The presently disclosed subject matter relates to a method for predicting responses to VEGF targeted therapeutics. More particularly, the presently disclosed subject matter relates to a method of predicting response in a subject to treatment for a disease with a VEGF targeted therapeutic by determining a level of ANGPTL4 expression in a biological sample from the subject and comparing the level of ANGPTL4 expression in the sample with a reference.

INTRODUCTION

[0004] Renal cell carcinoma (RCC) is the most common type of kidney cancer in adults, accounting for approximately 3% of adult malignancies and 90-95% of neoplasms arising from the kidney. RCC is not a single entity; it includes a group of tumors that arise from the epithelium of renal tubules. Over the past 30 years, kidney cancer incidence has continued to rise ~2 % annually. The American Cancer Society estimates that in the upcoming year there will be around 64,000 new cases and over 13,000 deaths due to the disease. Different treatment options are available for patients with renal cell carcinoma including surgery, radiation, chemotherapy, immunotherapy, and targeted therapy However, surgical resection remains the only known effective treatment for localized cases and many patients with advanced disease can be offered only palliative therapy. The past five years have witnessed a major change in the treatment of advanced RCC with the introduction of targeted therapies that derive their efficacy through affecting angiogenesis. [0005] Current therapies for renal cell carcinoma rely on a combination of molecular approaches, using multidrug regimens consisting of small-molecule kinase inhibitors with biologic therapies, immunomodulatory therapies, or both. While treatment strategies that include VEGF- targeted inhibitors have greatly improved management of patients with renal cell carcinoma, clinical trials reveal many cases where patients did not exhibit a therapeutic response by standard clinical criteria. Thus, new approaches in personalized medicine are needed.

SUMMARY

[0006] The presently-disclosed subject matter meets some or all of the above -identified needs, as will become evident to those of ordinary skill in the art after a study of information provided in this document.

[0007] This Summary describes several embodiments of the presently-disclosed subject matter, and in many cases lists variations and permutations of these embodiments. This Summary is merely exemplary of the numerous and varied embodiments. Mention of one or more representative features of a given embodiment is likewise exemplary. Such an embodiment can typically exist with or without the feature(s) mentioned; likewise, those features can be applied to other embodiments of the presently-disclosed subject matter, whether listed in this Summary or not. To avoid excessive repetition, this Summary does not list or suggest all possible combinations of such features.

[0008] In some embodiments of the presently disclosed subject matter, a method of predicting response of a subject to treatment with a VEGF -targeted therapeutic is provided. The method includes the steps of (a) providing a biological sample from the subject; (b) determining a level of ANGPTL4 in the sample from the subject; (c) comparing the level of ANGPTL4 in the biological sample with a reference, wherein the subject is predicted to be a likely responder or a likely non-responder based on the level of ANGPTL4 in the sample relative to the reference. In some embodiments, the subject is predicted to be a likely non-responder to VEGF targeted therapeutic when there is a measurable decrease in the level of ANGPTL4 in the sample relative to the reference, and the subject is identified as a likely responder to the VEGF -targeted therapeutic when there is a measurable increase in the level of ANGPTL4 in the sample relative to the reference. In some embodiments, the method further includes the step of administering a VEGF targeted therapeutic to the subject when the subject is predicted to be a likely responder.

[0009] Further provided in the presently disclosed subject matter, is a method of diagnosing a disease treatable with a VEGF targeted therapeutic in a subject. The method includes the steps of (a) providing a biological sample including one or more cells from the subject; (b) determining the level of ANGPTL4 in the sample from the subject; (c) comparing the level of

ANGPTL4 in the sample with a reference; (d) diagnosing the subject as having a disease treatable with VEGF targeted therapeutic if there is a measurable difference in the level of ANGPTL4 in the one or more cells in the sample as compared to the reference; and (e)administering an effective amount of a VEGF targeted therapeutic to the subject having a disease treatable with VEGF targeted therapeutic. Non-limiting examples of the biological sample are serum and cancer tissue. Non- limiting examples of the disease are brain cancer, breast cancer, type 2 diabetes, macular degeneration, kidney cancer. In some embodiments, the disease is renal cell carcinoma.

[00010] In some embodiments of the presently disclosed subject matter, the VEGF targeted therapeutic is a biological molecule or small molecule that directly or indirectly modulates the VEGF signal transduction pathway. In some embodiments, the VEGF targeted therapeutic is a VEGF inhibitor. In some embodiments, the VEGF inhibitor is Sorafenib.

[00011] In some embodiments, the reference comprises a level of the ANGPTL4 in a sample from the subject taken over a time course. In some embodiments, the reference comprises a sample from the subject collected prior to initiation of a treatment program including use of a VEGF targeted therapeutic and the biological sample is collected after initiation of the treatment program. In some embodiments, the reference comprises a standard sample. In some embodiments, the reference comprises control data. In some embodiments, the reference comprises a level of ANGPTL4 in a sample in one or more samples from one or more individuals who are known responders or who are known non-responders to treatment of the disease with a VEGF targeted therapeutic.

[00012] Further provided in the presently disclosed subject matter, the steps of determining the a level of ANGPTL4 comprises at least one of determining the level of ANGPLT4 protein, determining the level of mRNA encoding ANGPLT4; determining the copy number of ANGPLT4, and determining the expression level of any non-coding or coding polymorphism within the nucleotide sequence ANGPLT4.

[00013] In some embodiments, a kit is provided in the presently disclosed subject matter.

The kit comprises an agent that selectively binds to ANGPTL4. In some embodiments, the agent comprises probes or primers to detect ANGPTL4. In some embodiments, the agent is an antibody.

[00014] Further advantages of the presently-disclosed subject matter will become evident to those of ordinary skill in the art after a study of the description, Figures, and non-limiting Examples in this document.

BRIEF DESCRIPTION OF THE DRAWINGS

[00015] Figure 1 shows illumina array analysis of GBM6 cells and tumor tissue. RNA was prepared from GBM6 bulk tumor cells, Ad-GSC, and Sp-GSC cultures as well as from

subcutaneous tumors that arose from these injected cells. Biological duplicates were run for each sample and data were collected and analyzed. A. Gene expression was measured by Illumina array and genes reported in the TCGA database were analyzed. B. The PCA plot represents the comparison of gene signatures from each condition.

[00016] Figure 2 shows molecular classification of GBM6 cells and tissue. A. Array analysis was performed and compared to glioblastoma molecular subclasses; PCA map depicts the classical, mesenchymal, neural, and proneural gene signatures and the gene signature of GBM6 cells and tumor tissue derived bulk tumor cells, Ad-GSCs and Sp-GSCs. B. RNA was pooled from three individual tumor tissues derived from GBM6 bulk tumor cells, Ad-GSCs and Sp-GSCs to determine gene expression of molecular markers of the Mesenchymal (CHI3L1, TRADD, NF1, RelB & CASP4) and Classical (FGFR3, PDGFA, EGFR, AKT2 & Nestin) subclass of glioblastoma by qPCR and normalized to actin expression (n=3). Error bars, S.D. * p < 0.05, ** p < 0.01, *** p < 0.001.

[00017] Figure 3 shows pathology of GBM6 tumor xenografts. GBM6 bulk tumor cells,

Ad-GSC, and Sp-GSC were injected subcutaneously and allowed to grow to a diameter of

~400mm³. Tumors were harvested and paraffin embedded for histology. A. H&E staining and B. Immunoreactivity for GFAP, S 100, OLIG2, MAP2, and SYN. (All photomicrographs taken at 200x).

[00018] Figure 4 shows characterization of orthotopic GBM6 tumors A. CB17 SCID mice

6

were orthotopically injected with 1x10 luciferase-tagged GBM6 cells grown as bulk tumor cells, Ad-GSCs, and Sp-GSCs, and tumor burden was measured by bioluminescence twice a week (n=10 per group). B. Tumor-bearing brains were extracted, paraffin embedded, and analyzed by H&E staining (All photomicrographs taken at 200x). C. RNA extracted from intracranial tumors was extracted, and subjected to qPCR as indicated and normalized to actin expression (n=3). Error bars, S.D. * p < 0.05, ** p < 0.01, *** p < 0.001.

[00019] Figure 5 shows enrichment of pro-survival and pro-angiogenic genes in glioblastoma tumor tissue derived from Ad-GSCs. A. RNA was prepared from three separate tumors generated from GBM6 bulk tumor cells and Ad-GSCs, and Nanostring analysis was performed. B. Volcano plot depicting the genes of interest within the GSC tumors that passed the Bonferroni test. C. Verification of significantly deregulated genes found in Nanostring analysis. The RNA used for Nanostring analysis was prepared and the gene expression of SPP1, ETV1, CCND2, CDH1, NQOl, and LYN was quantified by qPCR and normalized to actin expression. D. RNA was prepared from tumor tissue derived from GBM6 bulk tumor cells and Ad-GSCs to perform microarray analysis; schematic representation of Ingenuity pathway analysis showing the upregulation of genes involved in angiogenesis. E. qPCR validation of pro-angiogenic genes (ANGPTL4, IL8, CDKN2A and CXCL1) upregulated in tumors that arose from injected GBM6 Ad-GSCs normalized to actin expression (n=3). Error bars, S.D. * p < 0.05, ** p < 0.01, *** p < 0.001. [00020] Figure 6 shows expression of STAT3 and ANGPTL4 in human GBM samples. A.

Gene expression of STAT3 and ANGPTL4 in the TCGA database for normal brain tissue (n=10), low grade glioma samples (n=265) and GBM samples (n=328). B. Gene expression of STAT3 and ANGPTL4 in the TCGA database for normal brain tissue (n=10) and GBM samples (n=328) was related to Karnosky performance status (the number of patients in Group I, II and III were 247, 64, and 17, respectively). C and D. RNA was extracted from 36 GBM patient biopsies, and STAT3 and ANGPTL4 expression was measured by qPCR and normalized to actin expression (n=3), and expression was correlated to short term and long term survival (C) and (D) as well as to each other. Error bars, S.D. * p < 0.05, ** p < 0.01, *** p < 0.001.

[00021] Figure 7 shows the antiglioma effects of a STAT3 inhibitor in vitro and in vivo. A.

RNA was prepared from GBM6 Ad-GSCs treated with WP1066 (50 μΜ) for 6 hrs, and the expression of ANGPTL4, VEGF-Rl, VEGF-R2, CD 133, SOX2 and Nestin was quantified by qPCR and normalized to actin expression (n=3). B. GBM6 bulk tumor cells and Ad-GSC were quantified by MTT assays after treatment (72 hr) with WP1066 at the indicated concentrations. C. Apoptosis of GBM6 bulk tumor cells and Ad-GSCs was determined by flow cytometry of Annexin V stained cells at 48 hrs after treatment with WP1066 (10 μΜ) and/or Sorafenib (5 μΜ) (n=3). D. ChlP- enriched STAT3 binding to the ANPTL4 promoter was analyzed by qPCR and normalized to input DNA, followed by subtraction of nonspecific binding determined by control IgG. E and F. Mice were subcutaneously injected with lxlO⁶ luciferase-tagged bulk tumor cells and Ad-GSCs, and upon initial detection of tumor, mice were treated with WP1066 (40 mg/kg) every other day. E. Tumor volume was determined with calipers twice weekly (n=5). F. Representative bioluminescence images at Day 31. G. Lysates of tumor tissue at one week following WP1066 treatment were immunoblotted for STAT3 and phospho-STAT3 as indicated. Error bars, S.D. * p < 0.05, ** p < 0.01, *** p < 0.001.

[00022] Figure 8 includes bar graphs showing the expression of cancer stem cell (CSC) markers in adherent cancer stem cells. ANGPTL4 is identified as a novel marker of renal cell carcinoma stem cells. RNA was prepared from bulk monolayer, and adherent (Ad) cancer stem cell cultures and spheroid (Sp) cancer stem cell cultures of Caki, RC-17 & RC-04 renal cell carcinoma cell lines. Relative gene expression of CD 133, CXCR4 and ANGPTL4 was quantified by quantitative PCT (qPCR) and normalized to actin expression (n=3).

[00023] Figure 9 includes bar graphs showing knockdown of ANGPTL4 leads to a decrease in stem cell and earlier progenitor markers of renal cell carcinoma adherent cancer stem cells. As shown in Figure 9A, ANGPTL4 was knocked down in Caki adherent cancer stem cells by shRNA and verified by ELISA assay. In Figure 9B, following knockdown, RNA was prepared from monolayer and adherent CSC +/- ANGPTL4 and the expression of CD 133 , CXCR4, Pax-2, Wnt-4, Bmp-7, and gp-160 was quantified by qPCR and normalized to actin expression (n=3). [00024] Figure 10 includes images and bar graphs illustrating tumorigenicity of adherent renal cell carcinoma cancer stem cells in the presence or absence of ANGPTL4 in vivo. Mice were injected in the renal cavity with 100,000 luciferase-tagged Caki adherent CSCs transduced with empty vector as a control or with ANGPTL4 sh RNa (ANGPTL4 KD) and (A) tumor growth was measured by live animal imaging twice a week (n=10 per group). (B) Tumors were weighed when animals were sacrificed 1 month post-injection.

[00025] Figure 11 is a graph showing the effects of Angptl4 knockdown on the sensitivity of RCC cancer stem cells to a VEGF pathway inhibitor. Proliferation of Caki Ad-CSCs with ANGPTL4KD (Caki Ad.CSC-ANGPTL4) or transduced with empty vector (Caki Ad.CSC) was measured by MTT assay after treatment (72 hr) with Sorafenib (VEGF pathway inhibitor (at the indicated concentrations (n=3). This result shows that ANGPTL4 knockdown cells were resistant to a VEGF pathway inhibitor.

[00026] Figure 12 includes graphs showing the effects of Angptl4 knockdown on the expression VEGF receptors on RCC cancer stem cell. Following knockdown in CaKi, RNA was prepared from bulk monolayer and adherent cancer stem cells with or without AngptWKD and the expression of VEGF receptors was quantified by qPCR and normalized to actin expression (n=3).

[00027] Figure 13 is graphs showing that ANGPTL4 expression is inversely related to patient survival in 187 samples from renal cell carcinoma in The Cancer Genome Atlas database. Patients that survived less than 2 years had a significantly greater level of ANGPTL4 expression when compared to patients that survived more than 2 years.

DESCRIPTION OF EXEMPLARY EMBODIMENTS

[00028] The details of one or more embodiments of the presently-disclosed subject matter are set forth in this document. Modifications to embodiments described in this document, and other embodiments, will be evident to those of ordinary skill in the art after a study of the information provided in this document. The information provided in this document, and particularly the specific details of the described exemplary embodiments, is provided primarily for clearness of understanding and no unnecessary limitations are to be understood therefrom. In case of conflict, the specification of this document, including definitions, will control.

[00029] Some of the polynucleotide and polypeptide sequences disclosed herein are cross- referenced to GENBANK^® / GENPEPT® accession numbers. The sequences cross-referenced in the GENBANK^® / GENPEPT® database are expressly incorporated by reference as are equivalent and related sequences present in GENBANK^® / GENPEPT® or other public databases. Also expressly incorporated herein by reference are all annotations present in the GENBANK^® / GENPEPT® database associated with the sequences disclosed herein. Unless otherwise indicated or apparent, the references to the GENBANK / GENPEPT® database are references to the most recent version of the database as of the filing date of this Application.

[00030] While the terms used herein are believed to be well understood by one of ordinary skill in the art, definitions are set forth to facilitate explanation of the presently-disclosed subject matter.

[00031] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the presently-disclosed subject matter belongs. Although any methods, devices, and materials similar or equivalent to those described herein can be used in the practice or testing of the presently-disclosed subject matter, representative methods, devices, and materials are now described.

[00032] Following long-standing patent law convention, the terms "a," "an," and "the" refer to "one or more" when used in this application, including the claims. Thus, for example, reference to "a cell" includes a plurality of such cells, and so forth.

[00033] Unless otherwise indicated, all numbers expressing quantities of ingredients, properties such as reaction conditions, and so forth used in the specification and claims are to be understood as being modified in all instances by the term "about". Accordingly, unless indicated to the contrary, the numerical parameters set forth in this specification and claims are approximations that can vary depending upon the desired properties sought to be obtained by the presently-disclosed subject matter.

[00034] As used herein, the term "about," when referring to a value or to an amount of mass, weight, time, volume, concentration or percentage is meant to encompass variations of in some embodiments ±20%, in some embodiments ±10%, in some embodiments ±5%, in some embodiments ±1%), in some embodiments ±0.5%, and in some embodiments ±0.1% from the specified amount, as such variations are appropriate to perform the disclosed method.

[00035] As used herein, ranges can be expressed as from "about" one particular value, and/or to "about" another particular value. It is also understood that there are a number of values disclosed herein, and that each value is also herein disclosed as "about" that particular value in addition to the value itself. For example, if the value "10" is disclosed, then "about 10" is also disclosed. It is also understood that each unit between two particular units are also disclosed. For example, if 10 and 15 are disclosed, then 11, 12, 13, and 14 are also disclosed.

[00036] The presently disclosed subject matter relates to a method for predicting responses to

VEGF targeted therapeutics. More particularly, the presently disclosed subject matter relates to a method of predicting response in a subject to treatment for a disease with a VEGF targeted therapeutic by determining a level of ANGPTL4 expression in a biological sample from the subject and comparing the level of ANGPTL4 expression in the sample with a reference. [00037] The presently disclosed subject matter involves methods of predicting response in a subject to treatment for a disease with a VEGF targeted therapeutic, and for evaluating and/or monitoring treatment for a disease involving use of a VEGF targeted therapeutic in a subject. Such methods involve determining a level of ANGPTL4 expression in a biological sample from the subject; and comparing the level of ANGPTL4 expression in the sample with a reference, wherein the subject is predicted to be a likely responder or non-responder based on the ANGPTL4 expression in the sample relative to the reference. Administering a VEGF targeted therapeutic to the subject is initiated or continued when the subject is predicted to be a likely responder. Non- limiting examples of the biological sample is serum and cancer tissues. The non-limiting examples of diseases that can be treated with a VEGF targeted therapeutic include, but not limited to brain cancer, breast cancer, type 2 diabetes, macular degeneration and kidney cancer. In some embodiments, the kidney cancer is renal cell carcinoma (RCC). In some embodiment, brain cancer is glioblastoma

[00038] In some embodiments, the presently disclosed subject matter relates to the specific role of ANGPTL4 in renal cell carcinoma cancer stem-like cells (CSCs) and ANGPTL4's association with cancer progression. In some embodiments of the presently disclosed subject matter, ANGPTL4 mRNA is found to be up-regulated in a subset of renal cell carcinomas. In some embodiments as shown in the examples below, ANGPTL4 expression is associated with renal cell carcinoma CSCs and the efficacy of VEGF inhibitor treatment of a patient. In some embodiment, ANGPTL4 can be used as a biomarker in a method of detecting and prognosing RCC, and predicting the response to VEGF inhibitor treatment.

[00039] In some embodiments of presently disclosed subject matter, the kidney cancer is renal cell carcinoma treated with VEGF targeted therapeutic. The past five years have witnessed a major change in the treatment of advanced renal cell carcinoma with the introduction of targeted therapies that derive their efficacy through affecting angiogenesis. The main class of agents involves drugs that target the vascular endothelial growth factor (VEGF).

[00040] VEGF is a family of glycoproteins that promotes tumor angiogenesis through tyrosine kinase signaling. As used herein, "VEGF targeted therapeutic" refers to a biological molecule or small molecule that directly or indirectly modulates the VEGF signal transduction pathway. In some embodiments, VEGF targeted therapeutics are inhibitors of the VEGF receptor mediated signaling pathway that are presently used in the clinic or in clinical trials in cancer patients.

[00041] The term "VEGF inhibitor" as used herein refers to inhibitors of the VEGF receptor mediated signaling pathway. Examples of VEGF inhibitor include, but not limited to, Sunitinib, Sorafenib, Pazopanib, Axitinib, Cabozantinib, Vandetanib, Regorafenib, Nintedanib, Brivanib, Tivozanib, Dovitinib, Bevacizumab, Ramucirumab, Ziv-aflibercept. For example, Sunitinib has been used in clinic or clinical trials for renal cell carcinoma (RCC) and gastrointestinal stromal tumor (GIST), Sorafenib has been used in clinic or clinical trials for RCC and hepatocellular carcinoma; Pazopanib has been used in clinic or clinical trials for RCC; Brivanib, Tivozanib, Dovitinib, and Bevacizumab have been used in clinic or clinical trials for lung, brain, and colon cancer;

Ramucirumab has been used in clinic or clinical trials for lung and gastric cancer, and Ziv-aflibercept has been used in clinic or clinical trials for colon cancer.

[00042] Several VEGF inhibitors are now approved for the treatment of metastatic renal cell carcinoma, but not all patients benefit from VEGF-targeted therapy. Therefore, it is important to uncover the mechanisms associated with VEGF inhibition and identify reliable markers that predict which patients are more likely to respond to anti-VEGF therapy. Establishing predictive markers for VEGF-targeted therapy is important for the development of effective treatment regimens for renal cell carcinoma.

[00043] In additional to treatment of renal cell carcinoma, in some embodiments, VEGF- targeted therapies are being tested in clinical trials to treat brain and breast cancer as well as to treat type 2 diabetes and macular degeneration.

[00044] An exemplary object of the presently disclosed subject matter is to predict responsiveness of renal cell carcinoma patients to treatment with VEGF inhibitors by detecting the expression of the potential biomarker Angiopoietin-like 4 (ANGPTL4) in tissue or sera samples. In some embodiments, high expression of ANGPTL4 is identified by microarray analysis of a number of renal cell carcinoma cell lines and confirmed by RT-PCR.

[00045] ANGPTL4 is a matricellular protein that provides signals to support tumorigenic activities characteristic of the metastatic cascade, such as cell proliferation, migration, survival, angiogenesis, epithelial-to-mesenchymal transition, and the maintenance of stem cell niches.

[00046] ANGPTL4 is identified as a prominent gene in in vivo hypoxia gene sets that predict poor outcome in multiple tumor types. This gene is induced under hypoxic (low oxygen) conditions in various cell types and is the target of Peroxisome proliferator-activated receptors. The encoded protein is a serum hormone directly involved in regulating lipid metabolism and also acts as an apoptosis survival factor for vascular endothelial cells.

[00047] Further, in a study of several epithelial tumor types, ANGPTL4 levels increased as tumors progress from local to metastatic disease. Therefore, it is believed to play a role in metastasis and lymphovascular invasion. It is anticipated that defining the specific role of ANGPTL4 in renal cell carcinoma tumorigenesis and its overlapping roles with VEGF in promoting angiogenesis lead to improved therapeutic design.

[00048] In accordance with the presently-disclosed subject matter, in some embodiments, the

ANGPTL-4 biomarkers are human ANGPTL-4 protein or rriRNA biomarkers. However, these exemplary human biomarkers are not intended to limit the present subject matter to human polypeptide biomarkers or mRNA biomarkers only. Rather, the present subject matter encompasses biomarkers across animal species that are associated with diseases that are treated with a VEGF targeted therapeutic. In addition, standard gene/protein nomenclature guidelines generally stipulate human gene name abbreviations are capitalized and italicized and protein name abbreviations are capitalized, but not italicized. Further, standard gene/protein nomenclature guidelines generally stipulate mouse, rat, and chicken gene name abbreviations italicized with the first letter only capitalized and protein name abbreviations capitalized, but not italicized. In contrast, the gene/protein nomenclature used herein when referencing specific biomarkers uses all capital letters for the biomarker abbreviation, but is intended to be inclusive of genes (including mRNAs and cDNAs) and proteins across animal species.

[00049] A "biomarker" is a molecule useful as an indicator of a biologic state in a subject. With reference to the present subject matter, the biomarkers disclosed herein can be polypeptides that exhibit a change in expression or state, which can be correlated with the response to treatment with VEGF targeted therapeutic for a disease in a subject. In addition, the biomarkers disclosed herein are inclusive of messenger RNAs (mRNAs) encoding the biomarker polypeptides, as measurement of a change in expression of an mRNA can be correlated with changes in expression of the polypeptide encoded by the mRNA. As such, determining an amount of a biomarker in a biological sample is inclusive of determining an amount of a polypeptide biomarker and/or an amount of an mRNA encoding the polypeptide biomarker either by direct or indirect (e.g., by measure of a complementary DNA (cDNA) synthesized from the mRNA) measure of the mRNA.

[00050] In some embodiments of the presently-disclosed subject matter, a method for diagnosing a disease using expression levels of ANGPTL4 is provided. In some embodiments, the method relates to diagnosing or prognosing a disease in need of VEGF therapeutics, or the risk thereof, in a subject. The method includes the steps of providing a biological sample from the subject; determining the level of ANGPTL4 expression in a biological sample including a biological fluid or a biological tissue from the subject, and comparing the level of ANGPTL4 expression in the sample with a reference, wherein if increased expression level of

ANGPTL4 is indicative that the subject is diagnosed as having, or at an increased risk of developing, the disease. In some embodiments, the disease is kidney cancer.

[00051] In some embodiments of the presently-disclosed subject matter, a method to identify a subject that has poor survival prognosis kidney cancer is provided. The method includes obtaining a kidney cancer cell containing biological sample from the subject, determining the level of ANGPTL4 expression in the biological sample, and identifying the subject as one who has increased risk of poor prognosis kidney cancer if the expression level of ANGPTL4 is increased in the biological sample than the expression level of ANGPTL4 in the reference. In some embodiments, the disease is a cancer. In some embodiments, the cancer is renal cell carcinoma.

[00052] The terms "diagnosing" and "diagnosis" as used herein refer to methods by which the skilled artisan can estimate and even determine whether or not a subject is suffering from a given disease or condition. The skilled artisan often makes a diagnosis on the basis of one or more diagnostic indicators, such as for example a marker, the amount (including presence or absence) of which is indicative of the presence, severity, or absence of the condition.

[00053] Along with diagnosis, clinical cancer prognosis is also an area of great concern and interest. It is important to know the aggressiveness of the cancer cells and the likelihood of tumor recurrence in order to plan the most effective therapy. If a more accurate prognosis can be made, appropriate therapy, and in some instances less severe therapy for the patient can be chosen.

Measurement of biomarker levels disclosed herein (e.g., ANGPTL4) can be useful in order to categorize subjects according to advancement of a cancer who will benefit from particular therapies and differentiate from other subjects where alternative or additional therapies can be more appropriate.

[00054] As such, "making a diagnosis" or "diagnosing", as used herein, is further inclusive of determining a prognosis, which can provide for predicting a clinical outcome (with or without medical treatment), selecting an appropriate treatment (or whether treatment would be effective), or monitoring a current treatment and potentially changing the treatment, based on the measure of diagnostic biomarker levels disclosed herein.

[00055] The phrase "predicting response" as used herein refers to methods by which the skilled artisan can predict the course or outcome of the treatment for a condition in a subject. The term "response" does not refer to the ability to predict the course or outcome of a condition with 100% accuracy, or even that a given course or outcome is predictably more or less likely to occur based on the presence, absence or levels of test biomarkers. Instead, the skilled artisan will understand that the term "response" refers to an increased probability that a certain course or outcome will occur; that is, that a course or outcome is more likely to occur in a subject exhibiting a given condition, when compared to those individuals not exhibiting the condition. For example, in individuals not exhibiting the condition (e.g., not expressing the biomarker or expressing it at a reduced level), the chance of a given outcome may be about 3%. In certain embodiments, a prognosis is about a 5% chance of a given outcome, about a 7% chance, about a 10%> chance, about a 12% chance, about a 15% chance, about a 20% chance, about a 25% chance, about a 30% chance, about a 40% chance, about a 50% chance, about a 60% chance, about a 75% chance, about a 90% chance, or about a 95% chance.

[00056] The skilled artisan will understand that associating a prognostic indicator with a predisposition to an adverse outcome is a statistical analysis. For example, a biomarker level (e.g., quantity of expression in a sample) of less than a control level in some embodiments can signal that a subject is less likely to respond to VEGF targeted therapeutics than subjects with a level greater than or equal to the control level, as determined by a level of statistical significance. Additionally, a change in marker concentration from baseline levels can be reflective of subject prognosis, and the degree of change in marker level can be related to the severity of adverse events. Statistical significance is often determined by comparing two or more populations, and determining a confidence interval and/or a p value. See, e.g., Dowdy and Wearden, Statistics for Research, John Wiley & Sons, New York, 1983, incorporated herein by reference in its entirety. Preferred confidence intervals of the present subject matter are 90%, 95%, 97.5%, 98%, 99%, 99.5%, 99.9% and 99.99%, while preferred p values are 0.1, 0.05, 0.025, 0.02, 0.01, 0.005, 0.001, and 0.0001.

[00057] In other embodiments of the presently-disclosed subject matter, a threshold degree of change in the level of a prognostic or diagnostic biomarker can be established, and the degree of change in the level of the indicator in a biological sample can simply be compared to the threshold degree of change in the level. A preferred threshold change in the level for markers of the presently- disclosed subject matter is about 5%, about 10%>, about 15%, about 20%, about 25%, about 30%, about 50%, about 75%, about 100%, and about 150%.

[00058] In some embodiments of the presently disclosed subject matter, multiple

determinations of one or more diagnostic or prognostic peptide biomarkers can be made, and a temporal change in the biomarker can be used to monitor the progression of disease and/or efficacy of appropriate therapies directed against the disease. In such an embodiment, for example, one might expect to see a decrease or an increase in the biomarker(s) over time during the course of effective therapy. Thus, the presently-disclosed subject matter provides in some embodiments a method for predicting and determining treatment efficacy of VEGF therapeutics in a subject.

[00059] In certain embodiments, a diagnostic or prognostic biomarker is correlated to a condition or disease by merely its presence or absence. In other embodiments, a threshold level of a diagnostic or prognostic biomarker can be established, and the level of the indicator in a subject sample can simply be compared to the threshold level.

[00060] The "reference" to which the sample is compared can include, for example, a level of

ANGPTL4 in one or more samples from one or more individuals without the disease. In some embodiments, the reference includes a level of ANGPTL4 in a sample from the subject taken over a time course. In some embodiments, the reference includes a sample from the subject collected prior to initiation of treatment for the disease and/or onset of the disease and the biological sample is collected after initiation of the treatment or onset of the disease.

[00061] In some embodiments, the reference can include a standard sample. Such a standard sample can be a reference that provides amounts of ANGPTL4 at levels considered to be control levels. For example, a standard sample can be prepared with to mimic the amounts or levels of ANGPTL4 in one or more samples (e.g., an average of amounts or levels from multiple samples) from one or more individuals. In some embodiments the standard sample can be a reference that provides amounts of ANGPTL4 at levels considered to associated with a responder or non-responder to treatment. [00062] In some embodiments, the reference can include control data. Control data, when used as a reference, can comprise compilations of data, such as may be contained in a table, chart, graph, e.g., standard curve, or database, which provides amounts or levels of ANGPTL4 considered to be control levels, responder levels, and/or non-responder levels. Such data can be compiled, for example, by obtaining amounts or levels of ANGPTL4 in one or more samples (e.g., an average of amounts or levels from multiple samples) from one or more individuals.

[00063] As used herein, the term "subject" refers to an animal, including a mammal, such as a human being.

[00064] With regard to the step of providing a biological sample from the subject, the term "biological sample" as used herein refers to a sample that comprises a biomolecule and/or is derived from a subject. As such, a biological sample includes one or more cells from the subject. In some embodiments, for example, the biological sample is a body fluid or tissue potentially comprising the presently-disclosed biomarkers, including ANGPTL4. In some embodiments, for example, the biological sample can be a blood sample, a serum sample, a plasma sample, or sub-fractions thereof. In some embodiments, the biological sample is a tissue biopsy, such as a tissue biopsy containing cancerous cells or cells suspected to be cancerous.

[00065] Turning now to the step of identifying one or more markers in the biological sample, various methods known to those skilled in the art can be used to identify the one or more markers in the provided biological sample. In some embodiments, determining the amount of biomarkers in samples comprises using an RNA measuring assay to measure mRNA encoding biomarker polypeptides in the sample and/or using a protein measuring assay to measure amounts of biomarker polypeptides in the sample.

[00066] In certain embodiments, the amounts of biomarkers can be determined by probing for mRNA of the biomarker in the sample using any RNA identification assay known to those skilled in the art. Briefly, RNA can be extracted from the sample, amplified, converted to cDNA, labeled, and allowed to hybridize with probes of a known sequence, such as known RNA hybridization probes (selective for mRNAs encoding biomarker polypeptides) immobilized on a substrate, e.g., array, or microarray, or quantitated by real time PCR (e.g., quantitative real-time PCR). Because the probes to which the nucleic acid molecules of the sample are bound are known, the molecules in the sample can be identified. In this regard, DNA probes for ANGPTL4 can be immobilized on a substrate and provided for use in practicing a method in accordance with the present subject matter.

[00067] With regard to determining amounts of biomarker polypeptides in samples, in some embodiments, mass spectrometry and/or immunoassay devices and methods can be used to measure biomarker polypeptides in samples, although other methods are well known to those skilled in the art as well. See, e.g., U.S. Pat. Nos. 6,143,576; 6,113,855; 6,019,944; 5,985,579; 5,947,124; 5,939,272; 5,922,615; 5,885,527; 5,851,776; 5,824,799; 5,679,526; 5,525,524; and 5,480,792, each of which is hereby incorporated by reference in its entirety. Immunoassay devices and methods can utilize labeled molecules in various sandwich, competitive, or non-competitive assay formats, to generate a signal that is related to the presence or amount of an analyte of interest. Additionally, certain methods and devices, such as biosensors and optical immunoassays, can be employed to determine the presence or amount of analytes without the need for a labeled molecule. See, e.g., U.S. Pat. Nos. 5,631,171; and 5,955,377, each of which is hereby incorporated by reference in its entirety.

[00068] Thus, in certain embodiments of the presently-disclosed subject matter, the marker peptides are analyzed using an immunoassay. The presence or amount of a marker (e.g., ANGPTL4) can be determined using antibodies or fragments thereof specific for each marker and detecting specific binding. For example, in some embodiments, the antibody specifically binds ANGPTL4, which is inclusive of antibodies that bind the full-length peptide, fragments thereof, or

phosphorylated forms thereof. In some embodiments, the antibody is a monoclonal antibody.

[00069] Any suitable immunoassay can be utilized, for example, enzyme-linked

immunoassays (ELISA), radioimmunoassays (RIAs), competitive binding assays, and the like.

Specific immunological binding of the antibody to the marker can be detected directly or indirectly. Direct labels include fluorescent or luminescent tags, metals, dyes, radionuclides, and the like, attached to the antibody. Indirect labels include various enzymes well known in the art, such as alkaline phosphatase, horseradish peroxidase and the like.

[00070] The use of immobilized antibodies or fragments thereof specific for the markers is also contemplated by the presently-disclosed subject matter. The antibodies can be immobilized onto a variety of solid supports, such as magnetic or chromatographic matrix particles, the surface of an assay plate (such as microtiter wells), pieces of a solid substrate material (such as plastic, nylon, paper), and the like. An assay strip can be prepared by coating the antibody or a plurality of antibodies in an array on a solid support. This strip can then be dipped into the test biological sample and then processed quickly through washes and detection steps to generate a measurable signal, such as for example a colored spot.

[00071] In certain embodiments of the method, it can be desirable to include a control sample that is analyzed concurrently with the biological sample, such that the results obtained from the biological sample can be compared to the results obtained from the control sample. Additionally, it is contemplated that standard curves can be provided, with which assay results for the biological sample can be compared. Such standard curves present levels of protein marker as a function of assay units, i.e., fluorescent signal intensity, if a fluorescent signal is used. Using samples taken from multiple donors, standard curves can be provided for control levels of the one or more markers in normal tissue.

[00072] The analysis of markers can be carried out separately or simultaneously with additional markers within one test sample. For example, several markers can be combined into one test for efficient processing of a multiple of samples and for potentially providing greater diagnostic and/or prognostic accuracy. In addition, one skilled in the art would recognize the value of testing multiple samples (for example, at successive time points) from the same subject. Such testing of serial samples can allow the identification of changes in marker levels over time. Increases or decreases in marker levels, as well as the absence of change in marker levels, can provide useful information about the disease status that includes, but is not limited to, identifying the approximate time from onset of the event, the presence and amount of salvageable tissue, the appropriateness of drug therapies, the effectiveness of various therapies, and identification of the subject's outcome, including risk of future events.

[00073] The analysis of markers can be carried out in a variety of physical formats as well.

For example, the use of microtiter plates or automation can be used to facilitate the processing of large numbers of test samples. Alternatively, single sample formats could be developed to facilitate immediate treatment and diagnosis in a timely fashion, for example, in ambulatory transport or emergency room settings.

[00074] Further provided, in some embodiments, is that the VEGF targeted therapeutic is a biological molecule or small molecule that directly or indirectly modulates the VEGF signal transduction pathway. In some embodiments, the VEGF targeted therapeutic is a VEGF inhibitor. For example, the VEGF inhibitor is Sorafenib.

[00075] As used herein, the terms "treatment" or "treating" relate to any treatment of a condition of interest, including but not limited to prophylactic treatment and therapeutic treatment.

[00076] The terms relate to medical management of a subject with the intent to substantially cure, ameliorate, stabilize, or substantially prevent a condition of interest (e.g., disease, pathological condition, or disorder), including but not limited to prophylactic treatment to preclude, avert, obviate, forestall, stop, or hinder something from happening, or reduce the severity of something happening, especially by advance action.

[00077] As such, the terms treatment or treating include, but are not limited to: inhibiting the progression of a condition of interest; arresting or preventing the development of a condition of interest; reducing the severity of a condition of interest; ameliorating or relieving symptoms associated with a condition of interest; causing a regression of the condition of interest or one or more of the symptoms associated with the condition of interest; and preventing a condition of interest or the development of a condition of interest.

[00078] The terms includes active treatment, that is, treatment directed specifically toward the improvement of a condition of interest, and also includes causal treatment, that is, treatment directed toward removal of the cause of the condition of interest. In addition, the terms includes palliative treatment, that is, treatment designed for the relief of symptoms rather than the curing of the condition of interest; preventative treatment, that is, treatment directed to minimizing or partially or completely inhibiting the development of the associated condition of interest; and supportive treatment, that is, treatment employed to supplement another specific therapy directed toward the improvement of the associated condition of interest.

[00079] Further still, in some embodiments of the presently disclosed subject matter, the step of determining the expression level is provided to include at least one of the following steps:

determining the level of ANGPLT4 protein; determining the level of mRNA encoding ANGPLT4; determining the copy number of ANGPLT4; and determining the expression level of any non-coding or coding polymorphism within the nucleotide sequence ANGPLT4.

[00080] In some embodiments, the presently disclosed subject matter provides a kit that includes an agent that selectively binds to ANGPTL4. In some embodiments, the agent includes probes or primers to detect expression of ANGPTL4. An non-limiting example of the agent is an antibody.

[00081] The presently-disclosed subject matter is further illustrated by the following specific but non-limiting examples. The following examples may include compilations of data that are representative of data gathered at various times during the course of development and experimentation related to the presently disclosed subject matter.

[00082] EXAMPLES

[00083] Example 1

[00084] Brain tumors represent an important cause of cancer-related morbidity and mortality in the United States, with malignant gliomas being among the most aggressive and difficult to treat (1). Although they rarely metastasize, malignant gliomas are locally invasive, highly vascular tumors with extensive areas of necrosis and hypoxia. The prognosis for patients with glioma is poor. Most patients with glioblastoma multiforme (GBM), the most severe grade of glioma (WHO grade IV) and the most common glioma subtype in adults, die within 1 -2 years of diagnosis, and patient survival has remained dismal for decades (1). Surgical resection of GBM remains the primary treatment modality with present adjuvant therapies, including chemotherapy and radiation therapy, only providing slight improvement in the disease course and outcome (2). Patients with recurrent GBM have an even bleaker prognosis (3).

[00085] Glioblastoma tumors are a heterogeneous mixture of cellular and molecular subtypes, which may underlie the inability of conventional and targeted therapies to significantly impact patient outcomes. Furthermore, there is potential that the heterogeneity of glioblastomas may arise from different tumor-initiating subpopulations (4). Genomic profiling has identified four molecular subtypes of glioblastoma: Proneural, Neural, Classical and Mesenchymal. The Proneural subtype has been associated with PDGFRA abnormalities, IDH1 and TP53 mutations and is associated with younger patients. Most gliomas are classified as Proneural due to their oligodrendrocytic signature (5- 7). The gene expression of the Neural class of glioma most closely resembles normal brain tissue, and has a strong enrichment for genes differentially expressed by neurons. The Classical glioma subtype has an astrocytic signature; EGFR amplification is commonly observed in this tumor type along with low levels of CDK 2A and TP53. The neural precursor and stem cell marker Nestin as well as NOTCH3 and Shh signaling pathways are associated with the Classical subtype. Gliomas classified as Mesenchymal exhibit higher activity of mesenchymal and astrocytic markers, which mimics cells undergoing the epithelial-to-mesenchymal transition (8). The subtype is characterized by the high expression of the mesenchymal markers, MET and CHI3L1, as well as deletion of NF1 (9,10).

Markers of proliferation and angiogenesis also distinguish the Mesenchymal phenotype, suggesting an aggressive glioblastoma tumor subtype. Tumors with a Mesenchymal gene signature tend to be more aggressive, highly resistant to therapy, lead to a higher rate of relapse and have worse overall outcomes than tumors of the Classical, Proneural and Neural subtype (9). Therefore, a more detailed molecular understanding of this GBM subtype is crucial to improve therapeutic design and patient outcome.

[00086] The tumorigenic process in GBM is apparently initiated and sustained by a rare subpopulation of GBM stemlike cells (GSCs) (11-13). As is the case with normal stem cells, GSCs can self-renew and undergo differentiation, but they have high tumor-initiating capacity and therapeutic resistance (14). These stemlike cells are often isolated based on their ability to grow as multicellular, nonadherent spheres from single cell suspensions (15,16), which was denoted as spheroid GBM stemlike cells (Sp-GSCs). However, expansion of Sp-GSCs is technically challenging. Moreover, as spheres enlarge differentiated progeny appear and dying cells accumulate within the sphere's core. As an alternative approach, adherent GSCs isolated from GBM (Ad-GSCs) and grown in chemically defined medium in the absence of serum on laminin-coated tissue culture flasks display stem cell properties and initiate high-grade gliomas following xenotransplantation (17).

[00087] In the present report, Ad-GSCs and Sp-GSCs were isolated from a GBM patient- derived xenoline (PDX) that has the Classical gene signature. Both of these GSC populations were found to have markedly enhanced tumor-initiating activity (TIA) when compared to more differentiated, bulk tumor cells grown as adherent monolayers in serum. When injected

subcutaneously into the flanks or orthotopically into the brains of immunocompromised mice, Ad- GSCs and Sp-GSCs have very enhanced tumor-initiating capacities as compared to bulk tumor cells, but form tumors that display identical histological features. Of note, while both GSC populations in vitro have a mesenchymal gene signature, Sp-GSCs form tumors with a Classical gene signature that recapitulate the molecular properties of the original GBM PDX. In contrast, Ad-GSCs form tumors with a mesenchymal gene signature. Besides upregulated expression of many genes typical of the mesenchymal subclass in tumors formed from Αά-GSCs, upregulated expression ofSTAT3 and angiopoietin like-4 (ANGPTL4) was also observed. STAT3 is an important transcriptional factor that plays a significant role in oncogenesis. ANGPTL4 has been reported to act not only as a tumor suppressor (18), but also as an enhancer of tumor metastasis and angiogenesis (19). A pharmacological STAT3 inhibitor blocked STAT3 binding to the ANGPTL4 promoter, and promoted apoptosis of Ad-GSCs in vitro as compared to bulk tumor cells, and had in viv antitumor activity on Ad-GSC xenografts .

[00088] MATERIALS AND METHODS

[00089] Cell Culture. The human GBM6 patient-derived xenograft (PDX) of adult glioblastoma tissue was provided by Dr. C. David James, (Department of Neurological Surgery, University of California, San Francisco) and continuously maintained as subcutaneous xenografts in five-week-old male NOD.Cg Prkdc ^cid Il2rg^{ml Wil}I zi (NSG) mice (Jackson Laboratory, Bar Harbor, ME). Cell cultures of the GBM6 PDX were derived from minced, freshly harvested tumor tissue. Short-term GBM6 cultures of differentiated bulk tumor cells were grown as adherent monolayers in DMEM (Cellgro, Herndon, VA) supplemented with 10% heat-inactivated fetal bovine serum

(Hyclone Labs, Thermo Scientific, Rockford, IL), 100 units/mL penicillin and 100 μg/mL streptomycin. Ad-GSCs and Sp-GSCs were maintained in NeuroBasal-A medium (Invitrogen, Carlsbad, CA) containing 2% B27 supplement, 2 mM L-glutamine, 100 units/mL penicillin, 100 μg/mL streptomycin, EGF (20 ng/ml), and basic FGF (40 ng/ml). For isolation of Ad-GSCs, culture flasks were coated with 100 μg/mL poly D-lysine (Sigma- Aldrich, St. Louis, MO) for 1 hr followed by coating with 10 μg/mL laminin (Gibco, Life Technologies Inc., Grand Island NY) for 2 hr prior to use. Ad-GSCs were plated at 1 X 10⁵ cells per 75 cm² flask, grown to confluence, dissociated with HyQTase (Thermo Scientific, Scientific, Rockford, IL), and split at a 1 :3 ratio. For isolation of Sp- GSCs, glioma cells were dissociated with HyQtase and plated at ~ 1 x 10⁵ cells/mL in ultra-low adhesion flasks.

[00090] Subcutaneous Xenografts. Animal experiments were performed in accordance with a study protocol approved by the Institutional Animal Care and Use Committee of the University of Tennessee Health Science Center. Glioma cancer xenografts were established in five -week-old male NOD.Cg- rfefc"** Il2rg^mlWilI zi (NSG) mice (Jackson Laboratory, Bar Harbor, ME) by direct flank injection of 1 xlO⁶ GBM6 cells transduced with lucif erase lentivirus constructs that mediate stable transfer of the cDNA for the enzyme luciferase. GBM xenografts were established in five -week-old male NOD.Cg- rfefc^*** Il2rg^{ml WJl}/SzJ (NSG) mice (Jackson Laboratory) by injection of cells (lxlO⁶) directly into the flanks (20). Tumors were measured twice weekly with a handheld caliper. For bioluminescence imaging, mice were injected intraperitoneally with d-luciferin (the luciferase substrate), imaged on the IVIS in vivo imaging system (Caliper Life Sciences, Hopkinton, MA), and photonic emissions assessed using Living image® software. To determine the effect of STAT3 inhibition, once detectable tumors were determined by caliper measurement, WP1066 (40 mg/kg) in DMSO/Polyethylene glycol were delivered every other day by intratumoral injection. The dose of WP1066 is consistent with previous in vivo studies (REF). [00091] Orthotopic Injections. Animal studies were performed under established guidelines and supervision by the St. Jude Children's Research Hospital's Institutional Animal Care and Use Committee, as required by the United States Animal Welfare Act and the National Institutes of Health's policy to ensure proper care and use of laboratory animals for research. Anesthetized (ketamine/xylazine) CB17 SCID mice were placed on stereotactic equipment where the scalp was prepped using alcohol and iodine swabs and artificial tear gel applied to the eyes. Following scalp excision, a rectangular window was carved out and the dura was completely removed from the surface of the brain, and lxlO⁶ cells suspended in 10 uL of medium were injected approximately 2.5 mm deep in the right motor cortex. The excision site was closed with skin glue, and all animals were monitored closely 24 hrs post-operative ly.

[00092] Gene Expression Analysis. Total RNA was isolated by treating tissue homogenates with Trizol followed by isolation with the RNeasy Mini kit (Qiagen Inc., Valencia, CA). Samples were submitted for complete mRNA expression profiling to the UTHSC Center of Genomics and Bioinformatics (Memphis, TN) for labeling and hybridization to Human-HT12 BeadChips (Illumina Inc.). Gene expression was also measured on the nCounter Analysis System (Nanostring

Technologies, Seattle, WA) using a panel of 230 human cancer-related genes. In brief, total RNA was mixed with pairs of capture and reporter probes, hybridized on the nCounter Prep Station, and purified complexes were measured on the nCounter digital analyzer. To account for differences in

hybridization and purification, data were normalized to the average counts for all control spikes in each sample and analyzed with nSolver software. Gene expression patterns were quality controlled by Principal Component Analysis (PCA). Genes were statistically tested by unequal variance t tests. The false discovery rate (FDR) was calculated to control for multiple comparisons using Partek Genomics Suite 6.6. Results were visualized using STATA/MP 11.2.

[00093] Ingenuity Pathway Analysis (IP A). IPA (Qiagen Inc., Valencia, CA) was used to identify canonical signaling pathways and functional pathways as well as to produce networks of related genes derived from genes changed in the analyzed comparisons. Here, the

rank-product-generated gene lists using a 5% FDR were uploaded into the IPA server as input data. IPA uses pathway libraries derived from the scientific literature. Statistics for functional analysis were carried out by Fischer's exact test.

[00094] Histopathology. Tumor tissue derived from non-GSCs, Ad-GSCs and Sp-GSCs

(four separate tumors for each condition) were fixed in 10% neutral buffered formalin for 24 hours, embedded in paraffin wax, and sectioned at 5 μιη thickness. For each sample, sections were stained using a standard hematoxylin and eosin (H&E) method, or by immunohistochemistry using antibodies to neural markers: GFAP, SI 00, OLIG2, MAP2 and synaptophysin. Representative images of each sample/stain combination were captured at 20x original magnification on a Nikon SI digital camera. [00095] Quantitative RT-PCR. Gene expression of RNA used for microarray analysis was measured by q-PCR on an iCyclerlQ (Bio-Rad Laboratories, Richmond, CA) using an iScript One- Step RT-PCR kit with SYBR Green (Bio-Rad Laboratories, Richmond, CA). Reaction parameters were as follows: cDNA synthesis at 50°C for 20 min, transcriptase inactivation at 95°C for 5 min, PCR cycling at 95°C for 10 sec, and 60°C for 30 sec for 40 cycles. The following primers were used for RT-PCR:

P-actim 5'-AGAAGGAGATCACTGCCCTG-3' (SEQ ID NO: 1) (forward),

5 ' -CACATCTGCTGGAAGGTGGA-3 ' (SEQ ID NO: 2) (reverse);

CHI31 5 ' -GTGAAGGCGTCTCAAACAGG-3 ' (SEQ ID NO: 3) (forward),

5 ' -GAAGCGGTCAAGGGCATCT-3 ' (SEQ ID NO: 4) (reverse);

TRADD 5'-GCTGTTTGAGTTGCATCCTAGC-3' (SEQ ID NO: 5) (forward),

5'-CCGCACTTCAGATTTCGCA-3' (SEQ ID NO: 6) (reverse);

NF1 5 ' -AGATGAAACGATGCTGGTCAAA-3 ' (SEQ ID NO: 7) (forward),

5 ' -CCTGTAACCTGGTAGAAATGCGA-3 ' (SEQ ID NO: 8) (reverse);

RelB 5 ' -CAGCCTCGTGGGGAAAGAC-3 ' (SEQ ID NO: 9) (forward),

5 ' -GCCCAGGTTGTTAAAACTGTGC-3 ' (SEQ ID NO: 10) (reverse);

CASP4 5'-TTTCTGCTCTTCAACGCCACA-3' (SEQ ID NO: 11) (forward),

5'-AGCTTTGGCCCTTGGAGTTTC-3' (SEQ ID NO: 12) (reverse);

FGFR3 5 ' -TGCGTCGTGGAGAACAAGTTT-3 ' (SEQ ID NO: 13) (forward),

5 ' -GCACGGTAACGTAGGGTGTG-3 ' (SEQ ID NO: 14) (reverse);

PDGFA 5'-GCAAGACCAGGACGGTCATTT-3' (SEQ ID NO: 15) (forward),

5 ' -GGCACTTGACACTGCTCGT-3 ' (SEQ ID NO: 16) (reverse);

EGFR 5'-CTACGGGCCAGGAAATGAGAG-3' (SEQ ID NO: 17) (forward),

5 ' -TGACGGCAGAAGAGAAGGGA-3 ' (SEQ ID NO: 18) (reverse);

AKT2 5 ' -ACCACAGTCATCGAGAGGACC-3 ' (SEQ ID NO: 19) (forward),

5 ' -GGAGCCACACTTGTAGTCCA-3 ' (SEQ ID NO: 20) (reverse);

Nestin 5'-GGCGCACCTCAAGATGTCC-3' (SEQ ID NO: 21) (forward),

5'- CTTGGGGTCCTGAAAGCTG-3' (SEQ ID NO: 22) (reverse).

[00096] TCGA Data Analysis. To examine the relationship between STAT3 and ANGPTL4 expression in human GBM brain tissue, the TCGA data portal was queried for all low-grade glioma and GBM samples with gene expression (BI HT HG-Ul 13A Array Data Set) data available as well as accompanying clinical data. The data set was filtered for samples having expression data for STAT3, ANGPTL4 and clinical data, yielding a final set of 466 individual low-grade glioma samples and 328 independent GBM patient samples. The samples were then grouped according to glioma grade and Karnofsky Performance Status, which takes into account the patient performance in general daily life activities. The Karnofsky score runs from 100 to 0, where 100 is defined as "perfect" health and 0 as death. Statistical analysis was performed using Graphpad Prism. [00097] Apoptosis Assay. The induction of apoptosis was monitored by flow cytometry

(Accuri Model 6C) using the Annexin V-FITC apoptosis detection kit (BD Pharmingen, San Diego, CA), according to the manufacturer's instructions.

[00098] Chromatin Immunoprecipitation. Chromatin Immunoprecipitation (ChIP) was carried out using the ChIP-IT™ Express Enzymatic kit (Active Motif, Carlsbad, CA) according to the manufacturer's instructions. In brief, chromatin from cells was cross-linked with 1% formaldehyde (10 min at 22°C), sheared to an average size of ^"¾200 bp, and then immunoprecipitated with anti-STAT3 (Santa Cruz Biotechnology). ChlP-PCR primers were designed to amplify a proximal promoter region containing a putative STAT3 (-1369 to -1348) binding site in the ANPTL4 promoter. The primers used were 5'- CATTAAAGACCCTGGCGGTA -3' (SEQ ID NO: 23) (forward), 5'- GGATCACAGTCGTGTGAGGA -3' (SEQ ID NO: 24) (reverse).

[00099] Statistical Analysis. At least three independent experiments were performed in duplicate, and data are presented as means ± sd. ANOVA and post-hoc least significant difference analysis or Student t tests were performed, p values < 0.05 (*) were considered statistically significant.

[000100] RESULTS

[000101] Differences in the molecular signatures of GSCs grown in vitro and as subcutaneous xenografts. Glioblastoma is characterized by extensive heterogeneity at the cellular and molecular levels (21), which may reflect the presence of different cancer stem cell populations. In a previous study, Ad-GSCs and Sp-GSCs were isolated from the GBM6 PDX, and identified that constitutive STAT3 and NF-κΒ activation in both GSC subpopulations upregulates the Notch pathway in GSCs (22). Moreover, consistent with both GSC population being stem-like cells both Ad- and Sp-GSCs were found to have high expression of stem cell markers, such as CD133, Sox2 and Nestin, but low expression of differentiation markers such as βΙΙΙ-tubulin and glial fibrillary acid protein as compared to bulk tumor cells. To determine the molecular signatures of these GSC populations, RNA was prepared from biological replicates of GBM6 cells grown as either short term cultures of differentiated bulk tumor cells, Ad-GSCs, or Sp-GSCs, and whole genome expression profiling was performed on HT-12 expression Bead-Chips by the UTHSC Center of Genomics and Bioinformatics. Analysis of the TCGA database of gene expression from 200 GBM and 2 normal samples has previously identified -840 GBM subtype predictor genes (4). Thus, hierarchical clustering was performed to assess the differential expression of these predictor genes in bulk tumor cells, Ad-GSCs and Sp-GSCs. As shown in Figure 1 A, a heatmap of the variation in expression of these predictor genes showed that cells grown in vitro under both GSC conditions exhibited very similar expression profiles, which was markedly different from the expression profile of that highly differentiated bulk tumor cells grown in serum-containing medium. Consistent with previous studies, Nestin, Sox2 and CD133 expression was expressed at relatively high levels in both GSC populations, while βΠΙ-tubulin and glial fibrillary acid protein was expressed at relatively low levels. Thus, although the two GSC populations were grown under somewhat different culture conditions (adherent versus suspension culture), their gene expression profiles were very similar. Then the mathematical variance was compared among the data samples by PCA, which is an unsupervised analytical method similar to factor analysis that is sensitive to all causes of variability within the data. Visualization of the first three components allows for quality control and the relative assessment of the variability of replicates. This PCA visualization also allows the assessment of the categorical factors of interest by demonstrating whether the data naturally aggregates by these factors or by an unknown or systematic factor like batch. When gene expression of the GSC populations and bulk tumor cells grown in vitro were subjected to PCA analysis, the gene expression profiles in the two different GSC cultures were found to be relatively similar, while that of bulk tumor cells grown in vitro was markedly different (Fig. IB).

[000102] Whole genome expression was also performed on tumors that arose from each of these GBM cell populations when injected into immunocompromised mice. In brief, bulk tumor cells, Ad-GSCs and Sp-GSCs (1 x 10⁶ cells) were injected into the flanks of NSG mice, and once tumors reached a volume of -200 mm³, tumors were excised, RNA was prepared and subjected to microarray analysis. In contrast to the findings with cells grown in vitro, tumors that arose from Ad-GSCs and Sp-GSCs had markedly different expression profiles as evidenced by the heatmaps of the gene expression profiles (Fig. 1A) as well as PCA of the average gene expression (Fig. IB). Moreover, the tumors that arose from the differentiated bulk tumor cells and Sp-GSCs were extremely similar. This finding is consistent with previous studies that showed that Sp-GSCs repopulate the GBM tumors with gene expression profiles nearly identical to that of bulk tumor cells (23). Therefore, while the gene expression of Sp-GSCs and Ad-GSCs grown in vitro are highly similar, the molecular signature of the tumor tissue that arose from these GSCs is quite different. These results suggest that distinct stem cell populations exist and promote tumor heterogeneity in GBM.

[000103] Ad-GSC and Sp-GSC xenografts have distinct molecular profiles. As described, genomic profiling of GBM samples in the TCGA database has identified four subtypes of glioblastoma: Proneural, Neural, Classical and Mesenchymal. To determine if the molecular signatures in glioblastoma stem cells grown in vitro and in vivo as Ad-GSCs and Sp-GSCs corresponded to any of the known molecular subtypes of glioblastoma, the expression of the predictor genes in bulk tumor cells was analyzed, Ad-GSCs and Sp-GSCs grown in vitro and in vivo, as compared to that of GBM tumors in the TCGA database. The data were subjected to PCA as shown in Figure 2A, which reveal the relationship between the gene expression patterns of glioblastoma bulk tumor cells, GSCs and the subsequent tumor tissue to that of the four molecular subtypes of glioblastoma. The overall predictor gene expression signatures among glioblastoma tumor tissue (202 samples) from Classical (white spheres), Neural (black spheres), Proneural (blue spheres) and Mesenchymal (grey spheres) subclasses and that of in vitro and in vivo samples are presented. As previously shown (4), the GBM tumor samples in the TCGA database form four distinct groups, but it is important to note that there is also some overlap in the expression patterns among these four subclasses. The gene expression profiles of Ad-GSCs (red spheres) and Sp-GSCs (yellow spheres) grown in vitro resemble that of the Mesenchymal glioblastoma subclass. In contrast the expression profile of bulk GBM6 tumor cells grown in vitro (orange spheres) is associated with the classical GBM subtype, which is consistent with what was previously found (Y. Gillespie, personal communication). Next the pattern of gene expression was examined in flank tumors that arose when immunocompromised mice were injected with bulk tumor cells, Ad-GSCs and Sp-GSCs. Most interestingly, tumors that arose from Ad-GSCs expressed a Mesenchymal gene signature, which is very similar to that of Ad-GSCs grown in vitro. In contrast, tumors that arose in mice injected with Sp-GSCs had a Classical subtype signature, which closely resemble the gene signature of tumors derived from bulk tumor cells. Thus, the two GSC populations differ at the molecular level. Sp-GSCs and Ad-GSCs have similar molecular properties (Mesenchymal subclass) in vitro, which are distinct from that of bulk tumor cells grown (classical) in vitro. When these GSCs were injected into mice, they form nio!ecularly distinct tumors. Ad-GSCs maintain a. Mesenchymal gene expression pattern in vitro and in vivo, while Sp-GSCs repopukte the tumor with a Classical gene signature, similar to that of bulk tumor cells.

[000104] To further characterize the glioblastoma xenografts, the expression of genes used that are typical of the Classical (FGFR3, PDFA, EGFR, AKT-2 and Nestin) and Mesenchymal (CHI3L1 , TRADD, NF1, RelB and CASP4) glioblastoma subtypes were examined. RNA was extracted and pooled from 3 individual subcutaneous xenografts derived from bulk tumor cells, Ad-GSCs or Sp- GSCs, and the expression of these marker genes was determined by qPCR. As shown in Figure 2B, the expression of Mesenchymal markers CHI3L1, TRADD and RelB are significantly elevated in xenografts derived from Ad-GSCs as compared to tumor tissue derived from bulk tumor cells and Sp- GSCs. In contrast, NF1, which is a tumor suppressor commonly downregulated in Mesenchymal glioblastoma, is expressed at relatively low levels in tumor tissue from Ad-GSCs as compared to tumors derived from Sp-GSCs and bulk tumor cells. The Mesenchymal marker CHI3L1 in combination with astrocytic markers is indicative of an epithelial-to-mesenchymal transition that has been linked to aggressive, dedifferentiated tumors (8). Genes in the TNF and NF-κΒ pathways, such as TRADD and RelB, are highly expressed in this subtype as well, potentially as a consequence of increased necrosis and associated inflammatory infiltrates (24). In addition, the Classical marker genes FGFR3, PDFA, EGFR and Nestin are all expressed at relatively high levels in tumors derived from bulk tumor cells and Sp-GSCs, as compared to Ad-GSCs, which is consistent with the PCA analysis classification of the tumors derived from these cells as the Classical subtype. In contrast,

EGFR and Nestin are expressed at extremely low levels in tumors derived from Ad-GSCs. Significant EGFR amplification is observed in 97% of Classical glioblastomas in the TCGA database and infrequently in other subtypes along with the neural precursor and stem cell marker, Nestin. The GBM6 xenograft is derived from a patient with overexpression of the VIII mutant of EGFR, and the finding of high EGFR expression in bulk tumor cells and in tumors derived from them is consistent with EGFR overexpression. However, it is extremely interesting that the tumors from Ad-GSCs express relatively low EGFR levels providing additional evidence that Ad-GSCs are a distinct GSC population. Overall the studies indicate that Ad-GSCs are a molecularly distinct GSC subpopulation from that of Sp-GSCs. The Ad-GSCs exhibit a Mesenchymal gene signature in vitro, and promote the formation of Mesenchymal tumors in vivo.

[000105] Histopathology of Sp-GSC and Ad-GSC xenografts. An important characteristic of GBM is marked morphologic and genetic heterogeneity within the tumor itself, as well as among different GBM tumors. The molecular heterogeneity observed in the glioblastoma xenografts derived from bulk tumor cells, Ad-GSCs and Sp-GSCs lead us to examine the histopathology of these tumors. Four individual subcutaneous tumors derived from each cell culture condition were formalin-fixed, paraffin-embedded, and sectioned. Each sample was stained for H&E analysis, and

immunohistochemistry was performed to measure immunoreactivity with antibodies for the following neural markers: GFAP, SI 00, OLIG2, MAP2 and synaptophysin. As shown in Figure 3 A, the Classical glioblastoma tumors of bulk tumor cells and Sp-GSCs xenografts, and the Mesenchymal tumors derived from Ad-GSCs were all high-grade gliomas and morphologically indistinguishable. Tumor cells with a high nuclear:cytoplasmic ratio and little nuclear pleomorphism showed the morphology of relatively undifferentiated high-grade gliomas. Immunohistochemistry of the tumor tissue demonstrated a uniform phenotype among the diverse glioblastoma tumors (Figure 3B). There was strong immunoreactivity for GFAP and OLIG2 in many tumor cells, while all tumor cells expressed S-100 and MAP-2. There was no immunoreactivity for synaptophysin. Across the range of neural tumors, GFAP, OLIG-2 and MAP-2 are generally expressed by gliomas, while the neuronal marker synaptophysin is not (25-28). These findings reveal the similarity at the microscopic level between the molecularly distinct xenografts of the bulk tumor cells, Sp-GSCs and Ad-GSCs.

[000106] Adherent GSCs promote the formation of intracranial tumors with a

Mesenchymal signature. In order to understand the role of the tumor microenvironment on gene expression in glioblastoma, intracranial injection of tumor cells was performed. The human glioblastoma orthotopic mouse model results in invasive growth in mice and allows quantitation of intracranial tumor growth (29). To determine if the findings on molecular heterogeneity of GBM tumors were also observed in the orthotopic microenvironment of glioblastoma, intracranial injections of 1 x 10⁶ luciferase-expressing GBM6 cells grown as short-term bulk tumor cells, Ad-GSCs or Sp- GSCs were performed. First the tumor initiating capacity of the glioblastoma bulk tumor cells was compared to both GSC conditions by bioluminescence imaging. As shown in Figure 4A, both Ad- GSCs and Sp-GSCs formed tumors more rapidly than bulk tumor cells, within 25 days as compared to 35 days, respectively. Moreover, animal survival was shorter with the more aggressive tumors (data not shown). This finding is also consistent with the previous studies that revealed Ad-GSCs were more potent in inducing tumors by subcutaneous injection when compared to bulk GBM tumor cells (22).

[000107] Next the histological phenotypes and molecular heterogeneity of the different intracranial glioblastoma xenografts were examined. As shown in Figure 4B, the intracranial glioblastoma tumors derived from bulk tumor cells, Ad-GSCs and Sp-GSCs were all determined to be high-grade glioma and exhibited no distinction in H&E staining, as determined with subcutaneous tumors. All of the tissue displayed characteristics of glioblastoma, such as hypercellularity, atypical nuclei, pseudopalisading necrosis and microvascular proliferation (30). The traditional glioblastoma prognostic markers GFAP, S 100, OLIG2, MAP2 and SYN were also analyzed and found to be indistinguishable among the glioblastoma xenografts (data not shown).

[000108] To further molecularly characterize the intracranial glioblastoma xenografts, the expression of genes used to characterize subcutaneous tumor xenografts that are typical of the Classical and Mesenchymal glioblastoma subtypes was also examined. RNA was extracted from 3 individual intracranial tumors derived from bulk tumor cells, Ad-GSCs or Sp-GSCs. As shown in Figure 4C, expression of the Mesenchymal markers CHI3L1, TRADD and RelB was significantly elevated in intracranial tumors derived from Ad-GSCs compared to the other xenografts, while NF1 was decreased. These results are very similar to the findings with subcutaneous tumors (Fig. 2B). In addition, the Classical markers FGFR3, PDGFA, EGFR, AKT-2 and Nestin were found to be highly expressed in intracranial tumor tissue derived from bulk tumor cells and Sp-GSCs, while there is lower expression in Ad-GSC tumors. These orthotopic studies support the previous finding that Ad- GSCs are a different tumor initiating subpopulation than Sp-GSCs. The adherent glioblastoma GSCs exhibit a Mesenchymal gene signature and promote the initiation and progression of the Mesenchymal glioblastoma subtype in vivo.

[000109] Upregulated expression of STAT3 and ANGPTL4 in Ad-GSC xenografts. To identify genes in oncogenic pathways differentially expressed in GBM subcutaneous xenografts, expression profiling was performed using the human cancer-related panel on the nCounter Analysis System (Nanostring Technologies, Seattle, WA). RNA was prepared from three individual tumors from NSG mice injected subcutaneously with bulk tumor cells and Ad-GSCs. Figure 5A reveals the different gene expression profiles of the individual subcutaneous GBM xenografts. While individual biological replicates of tumor tissue showed little variation in gene expression, there were clear differences in the gene expression pattern in tumors derived from short-term monolayer cultures compared to Ad-GSCs. Genes upregulated (SPP1, ETV1, CCND2) or downregulated (CDH1, NQOl, STAT3 and LYN) in the Ad-GSC-derived tumors that passed the Bonferroni threshold are shown in Figure 5B by Volcano Plots. The differential expression of these genes was determined by qPCR in three individual subcutaneous tumors derived from bulk tumor cells and Ad-GSCs, which is shown in Figure 5C. The enhanced expression of STAT3 in Ad-GSC derived tumors is in agreement with an upregulation of STAT3 activity in GSCs in vitro that was previously published (22). STAT3 has been reported to be an initiator and master regulator of mesenchymal transformation in glioblastoma, which is consistent with it being upregulated in GBM tumors derived from Ad-GSC that have a

Mesenchymal expression pattern (31). Elevated expression of CDH1, NQOl and LYN has also been previously identified in glioblastoma, with each of them contributing to the growth and invasion of this aggressive brain tumor subtype in different ways (32-34).

[000110] Functional annotation of the genes differentially expressed in glioblastoma tumors derived from Ad-GSCs also reveals that genes involved in angiogenesis are significantly enriched (P<0.05, Ingenuity Pathway Analysis). The schematic expression of these pro-angiogenic genes is shown in Figure 5D, and the qPCR validation of ANGPTL4, IL8, CDKN2A and CXCL1 mRNA levels in adherent GSC-derived tumors is shown in Figure 5E. ANGPTL4 upregulation has also previously been identified in an angiogenic gene signature that correlates with the Mesenchymal subtype of glioblastoma (9). The chemokine IL-8 has been found to be expressed and secreted at high levels in glioblastoma both in vitro and in vivo, and recent experiments suggest it is critical to glial tumor neovascularity and progression (35). CXCL1 is a chemokine that has been implicated as an oncogenic factor in glioma and directly related to attenuated angiogenic activity through NF-KB regulation (36). These pro-angiogenic genes have been shown to contribute to the vascularization of highly aggressive glioblastoma tumors and therefore represent therapeutic interests.

[000111] The Relationship between STAT3 and ANGPTL4 Expression to Glioma Grade and Patient Performance Status in Clinical Specimens. To examine STAT3 and ANGPTL4 expression in human glioma, information in the TCGA database was analyzed for the correlation between glioma grade and the expression of STAT3 and ANGPTL4. Gene expression of STAT3 and ANGPTL4 was examined in 10 normal brain samples, 466 individual low grade glioma patients and 328 individual GBM patients for comparison. As shown in Figure 6A, both STAT3 and ANGPTL4 levels are significantly increased in GBM samples when compared to normal tissue and low-grade gliomas. GBM specimens in the TCGA database were also analyzed to investigate the relationship between patient performance status and STAT3 and ANGPTL4 levels. The 328 GBM patients were placed into groups according to their Kamofsky performance status, where 100 is perfect health and 0 is death (Group I Kamofsky score 80-100, Group II score 50-70, and Group III score 20-40), and their gene expression was compared to the 10 normal brain samples. As shown in Figure 6B, the levels of ANGPTL4 are similar among groups, while a significant increase in STAT3 expression was found in lower performing patients when compared to those with good performance. Moreover, GBM patients had significantly higher STAT3 and ANGPTL4 expression when compared to normal brain samples.

[000112] Whether the expression of STAT3 and ANGPTL4 correlates with patient survival was independently determined. Brain biopsy specimens from 24 GBM patients that represented long- term and short-term survival after diagnosis were obtained from the UTHSC tissue core, and RNA was isolated from FFPE tissue blocks to determine expression of STAT3 and ANGPTL4. As shown in Figure 6C, although there was significant patient-to-patient variability as expected, there was a direct relationship between STAT3 and ANGPTL4 expression and patient survival. A statistically significant increase in STAT3 and ANGPTL4 gene expression was observed in GBM patients that survived less than one year when compared to long-term survivors. Interestingly, several patients exhibited elevated expression of both STAT3 and ANGPTL4, while others were shown to have low levels of both genes. The direct correlation of the expression of these genes in GBM is shown in Figure 6D.

[000113] The antiglioma effects of a STAT3 inhibitor in vitro and in vivo. To characterize the functional importance of upregulated expression of STAT3 and ANGPTL4 expression in GSCs, the effects of treatment with WP1066 was examined, which is a STAT3 inhibitor (22), on the expression of genes that are involved in angiogenesis as well as on stem cell marker genes in Ad- GSCs. As shown in Figure 7A, treatment with WP1066 decreased the expression of angiogenic genes (ANGPTL4, VEGFR-1 and VEGFR-2. In addition, WP1066 inhibited the expression of stem cell marker genes (CD133, SOX2 and Nestin). Then the effects of WP1066 on cell proliferation in cultures of bulk tumor cells and Ad-GSCs were examined by MTT assays. As shown in Fig. 7B, although WP1066 induced a dose-dependent reduction in cell number in both bulk tumor cells and Ad-GSCs, there was a markedly greater inhibitory effect on Ad-GSCs. Microscopic examination of the cultures indicated that this effect on Ad-GSCs may in part be due to cell apoptosis. Thus, the induction of apoptosis in Ad-GSCs was directly determined by flow cytometry of Annexin V-stained cells. Cells in 6-well plates were treated with 10 μΜ WP1066 for 48 hrs. As shown in Fig. 7C, WP1066 induced apoptosis of Ad-GSCs. To further investigate the relationship between STAT3 and angiogenesis, ChIP analysis was performed on ANGPTL4. Examination of putative transcription binding sites revealed a STAT3 binding site proximal (-1369 to -1348) to the ANGPTL4 promoter. To determine STAT3 binding to this site in the ANGPTL4 promoter, protein-DNA complexes were crosslinked with formaldehyde, chromatin was sheared to average of 200bp and immunoprecipitated with anti-STAT3, crosslinking reversed and the resulting DNA sequences detected by qPCR using specific primers designed and synthesized for this potential binding site. As shown in 7D, significantly increased binding of STAT3 to the ANGPTL4 promoter was observed in Ad-GSCs as compared to bulk tumor cells. In addition, the STAT3 inhibitor WP1066 decreased STAT3 binding to ANGPTL4 in GSCs. Taken together, these results show that STAT3 regulates ANGPTL4 expression in Ad-GSCs by directly binding to its promoter, and this interaction can be blocked with targeted STAT3 inhibitors.

[000114] To determine the anticancer activity of this STAT3 inhibitor in vivo, NSG mice were flank injected with Ad-GSCs (1 xlO⁶ cells) that express luciferase for noninvasive bioluminescence live animal imaging, and imaged twice a week. Once palpable tumors were detected (at ~ two weeks after tumor cell injection), WP1066 (40 mg/kg) was delivered intraperitoneally every other day for two weeks. As shown in Fig. 7D, the tumorigenicity of Ad-GSCs was markedly suppressed by treatment with WP1066. The apparent increase in tumor size upon treatment with WP1066 at 24 and 28 days is apparently due to tumor necrosis, because live animal imaging showed a marked decrease in bio luminescence during this time after treatment (Fig. 7D and E). In addition, immunoblotting of tumor tissue showed that high STAT3 activity as demonstrated by Y705-STAT phosphorylation was evident, and that treatment of mice nearly abolished STAT3 activity (Fig 7F). These results demonstrate the anticancer activity of this STAT3 inhibitor against the Ad-GSC subpopulation in vivo.

[000115] Example 2

[000116] Sorafenib is being used to treat kidney cancer and hepatocellular carcinoma, and is in clinical trials in glioblastoma. However, it is presently not possible to define which patients would most likely respond to sorafenib.

[000117] This study identifies that Angiopoietin-like 4 (ANGPTL4) controls the expression of vascular endothelial growth factor (VEGF) receptor expression in kidney cancer cells and thereby controls cellular sensitivity to sorafenib. Additionally, it has been found that Angiopoietin-like 4 can be readily measured in plasma by ELISA. Thus, Angiopoietin-like 4 can be used to predict patient response to Sorafenib, and other VEGF receptor antagonistics.

[000118] To understand the potential of ANGPTL4 as a biomarker for aggressive renal cell carcinoma, this study first obtains fresh human renal samples from patients diagnosed with renal cell carcinoma. The adherent stem cell culture used for gliomas was established for renal cell carcinoma cell lines and tested for sternness by looking at some key traits given to cancer stem cells (CSCs). Normal and cancerous tissue is collected and analyzed by RT-PCR and immunohistochemistry. RT- PCR is carried out to determine expression of ANGPTL4, as well as Actin, CD133, and CXCR4.

Actin is used as a reference gene to normalize the expression levels of other genes. Protein expression is also assessed by immunohistochemistry to distinguish high expression of ANGPTL4 in tumor tissue compared to normal kidney samples. Tissue samples are also counter-stained with CD133 or CXCR4 to determine if ANGPTL4 co-localizes with these proteins within the tumor-initiating cell subpopulations of the renal cell carcinoma (RCC) samples.

[000119] The cell lines used in this study are validated as following. First, the ability to self- renew of the RCC cell lines is confirmed by colony formation assays in cell culture and limiting dilution assays for tumor-initiation in animals. For colony formation assays, single cell suspensions of 5,000 cells in 1 ml of 0.4% agarose in tissue culture medium are added to triplicate wells of ultralow adhesion 6-well plates. Cells are fed twice a week with an additional 0.5 ml media. At day 14, plates are stained with MTT (10 μg/ml) for 3 hours and colonies are counted on a light microscope. For limiting dilution assays cells (between 10 and 10⁶) transduced with luciferase lentivirus constructs were injected directly into the flanks of five -week-old male NOD.Cg- Prkdc ^cid Il2rg^{ml WJl}/Szi (NSG) mice (Jackson Laboratory, Bar Harbor, ME) For bioluminescence imaging, mice were injected intraperitoneally with d-luciferin, imaged on the IVIS in vivo imaging system (Caliper Life Sciences, Hopkinton, MA), and photonic emissions assessed using Living image® software. Next adherent cancer stem cells (CSCs) also exhibit therapeutic resistance as compared to monolayers which was shown by Annexin-V staining. In brief, the induction of apoptosis was monitored by flow cytometry (Accuri Model 6C) using the Annexin V-FITC apoptosis detection kit (BD Pharmingen, San Diego, CA), according to the manufacturer's instructions. Among the 3 cell lines (CaKi, SK-RC-17, and SK-RC-04), significant increases in tumor initiation are detected in vivo in all 3 cell lines.

[000120] Now referring to Figure 8, which shows relative gene expression for tumor initiating genes CD 133, CXCR4 and ANGPTL4 in three cell culture conditions, monolayer, adherent CSC, and tumorsphere. The gene expression levels of these tumor-initiating cell markers are detected within the three culture conditions by RT-PCR. Two recognized RCC stem cell markers, CD133 and CXCR4, are found to be up-regulated in stem cell conditions. Adherent and spheroid CSCs are maintained in NeuroBasal-A medium (Invitrogen) containing 2% B27 supplement, 2 mM L-glutamine, 100 unite/mL penicillin, 100 pg/aiL streptomycin, EGF (20 ng/ml), and basic FGF (40 ng/ml). For isolation of adherent CSCs, culture flasks are coated with 100 μg/mL poly D-lysine (Sigma) for 1 hr followed by coating with 10 μg/mL laminin (Gibco) for 2 hr prior to use. Adherent CSCs are plated at 1 X 10⁵ cells per 75 cm² flask, grown to confluence, dissociated with HyQTase (Thermo Scientific), and split 1 :3. For isolation of spheroid CSCs, cells are dissociated with HyQtase and plated at ~ 1 x 10⁵ cells/mL in ultra-low adhesion flasks. Gene expression was measured b RT-PCR performed on an iCycierlQ (Bio-Rad Laboratories,

Richmond, CA) using an i Script Οηε-Step RT-PCR kit with SYBR Green (Bio -Rad Laboratories, Richmond, CA). Reaction parameters were as follows: cDN synthesis at 50 °C for 20 min, transcriptase inactivation at 95°C for 5 min, PGR cycling at 95°C for 10 sec, and 60°C for 30 sec for 40 cycles. The following primers are used for RT-PCR:

CD133 5 ' -CATCCACAGATGCTCCTAAGG-3 ' (SEQ ID NO: 25) (forward), 5'- AAGAGAATGCCAATGGGTCCA-3' (SEQ ID NO: 26) (reverse);

ANGPTL4 5'-GGCTCAGTGGACTTCAACCG-3' (SEQ ID NO: 27) (forward), 5'- CCGTGATGCTATGCACCTTCT-3 '(SEQ ID NO: 28) (reverse);

/3 -actin 5 ^" -AGAAGGAGATCACTGCCCTG-3 ^" (SEQ ID NO: 1) (forward), 5 ^" - CACATCTGCTGGAAGGTGGA-3 ^" (SEQ ID NO: 2) (reverse);

[000121] Tumor initiating cell marker CD133 is a five -transmembrane domain glycoprotein originally found on the cell surface of neuroepithelial stem cells in mice. CD133 has been used to identify normal and cancer stem cells from several different tissues, including renal epithelia and kidney cancer cells. In Figure 1, the expression of CD133 is high in adherent conditions. Adherent CSCs are plated at 1 X 10⁵ cells per 75 cm² flask, grown to confluence, dissociated with HyQTase (Thermo Scientific), and split 1 :3. For isolation of spheroid CSCs, cells are dissociated with HyQtase and plated at ~ 1 x 10⁵ cells/mL in ultra-low adhesion flasks.

[000122] Another tumor initiating cell marker CXCR4 is a chemokine receptor known for its role in metastasis of several solid tumors. Increased expression of CXCR4 and its ligand, stromal- derived factor (SDF-1), have been found in RCC cells and their target metastatic organs respectively. Figure 1 shows that the expression of CXCR4 is high in tumorsphere culture.

Tumorspheres are maintained in NeuroBasal-A medium (Invitrogen) containing 2% B27 supplement, 2 mM L-glutamine, 100 uniis/mL penicillin, 100 :.·;.· η ·! streptomycin, EGF (20 ng/ml), and basic FGF (40 ng/ml) in low adhesion plates.

[000123] Figure 1 further shows microarray analysis performed to determine genes of interest within the stem cell population. The microarray data shows the high expression of ANGPTL4 in the stem cell conditions, and high expression is confirmed by RT-PCR analysis as shown in Figure 1.

[000124] Further, to understand the potential role of ANGPTL4 in RCC tumorigenesis, several RCC cell lines (CaKi, SK-RC-17, and SK-RC-04) is transduced with a lentivirus-delivered shRNA against ANGPTL4, which resulted in -80% knockdown of ANGPTL4 expression. Although similar results were obtained with all RCC cell lines, only results with CaKi cells for brevity are shown here.

[000125] As such, ANGPTL4 expression is knocked down to study effects of ANGPTL4's absence on RCC stem cells. To knock down the human ANGPTL4 gene, Caki, SK-RC-17, and SK- RC-04 cell lines are transduced with a lentivirus-delivered shRNA against ANGPTL4. A human pLKO.l lentiviral shRNA target gene set that contains 5 clones is tested, and the clone

TRCN0000151318 with the primer sequence 5' - AAACCCAGGGCTGCCTTGGAAAAG - 3'(SEQ ID NO: 31) is proved to be most effective. Upon ANGPTL4 knockdown of at least 70%, the effects of its absence on renal CSCs and tumor progression is examined by testing cell proliferation, in vivo tumorigenicity, stem cell marker expression, differentiation capability, and therapeutic- resistance.

[000126] Now referring to Figure 9 which shows knockdown of ANGPTL4 leads to a decrease in stem cell and earlier progenitor markers of RCC adherent CSCs. Figure 9A shows that ANGPTL4 is knocked down in Caki adherent CSCs by shRNA and verified by ELISA assay.

Moreover, Figure 9B shows, following knockdown, the decreased expression of the earlier progenitor markers of RCC. In this experiment, RNA is prepared from monolayer and adherent CSC +/- ANGPTL4 and the expression of CD133, CXCR4, Pax-2, Wnt-4, Bmp-7, and gp-160 quantified by qPCR and normalized to actin expression (n=3). [000127] Further study in Figure 10 shows tumorigenicity of adherent CSCs in the presence or absence of ANGPTL4. Mice are injected in the renal cavity by Renal Capsule Injections with 100,000 luciferase-tagged Caki adherent CSCs with or without ANGPTL4. In brief, mice are anesthetized with isoflurane and a small incision is made along the left flank. The kidney is exposed and injected with 20 μΐ dPBS containing Caki cells, and tumor burden is measured by Xenogen imaging twice a week (n=10 per group). Tumors are weighed when animals were sacrificed 1 month post-injection. As shown in Figure 3 A, RCC cell lines that ANGPTL4 expression was knocked down are significantly less tumorigenic when injected into the kidney capsule of

immunocompromised mice as compared to EV-transduced RCC cells. Additionally, the tumor volume is smaller in mice that are injected with RCC cells that lacks ANGPLT4 expression.

Therefore, these results suggested a vital role of ANGPTL4 in RCC tumorigenesis.

[000128] Next, the effects of ANGPTL4 knockdown on sensitivity to the VEGF pathway inhibitor, Sorafenib, is examined. Sorafenib is an FDA approved drug used to treat kidney cancer. CaKi cells with either normal or knocked down ANGPTL4 expression were treated for 72 hours with varying concentrations of Sorafenib, and cell viability was determined by MTT assays. CaKi cells with knockdown of ANGPTL4 expression were relatively resistant to Sorafenib treatment.

[000129] Now referring to Figure 11 which shows the absence of ANGPTL4 leads to therapeutic resistance in renal cell carcinoma adherent cancer stem cells. Proliferation of Caki adherent CSCs +/- ANGPTL4 was measured by MTT assay after treatment (72 hr) with AM3100 (data not shown) and Sorafenib at the indicated concentrations. Figure 11 shows that, following knockdown If ANGPTL4, RNA of the VEGF receptors was prepared from monolayer and adherent CSC +/- ANGPTL4 and the expression of VEGF receptors was quantified by qPCR and normalized to actin expression (n=3).

[000130] As shown in Figure 11, -80% of ANGPTL4 knockdown CaKi cells remained viable at the highest dose of Sorafenib tested as compared to -15% cell viability of EV-transduced cells. The relationship of ANGPTL4 expression and sensitivity to the effect of Sorafenib on cell viability was confirmed when the induction of apoptosis by Sorafenib treatment is examined as determined by flow cytometry of Annexin V and Propidium Iodide stained cells. CaKi cells with ANGPTL4 expression knocked down were much more resistant to apoptosis induced by Sorafenib as compared to EV- transduced cells.

[000131] Additionally, to examine the underlying mechanism of VEGF-targeted therapeutic resistance in ANGPTL4 knockdown RCC cells, the expression of various genes involved in the VEGF signaling pathway are determined. As shown in Figure 4B, knockdown of ANGPTL4 not only reduced ANGPTL4 expression as expected, but also resulted in a decrease in VEGF receptor (VEGFR-1 and VEGFR-2) expression. [000132] In addition, ANGPTL4 knockdown resulted in reduced levels of ANGPTL4 expression in CaKi cells is validated by ELISA of conditioned media collected from CaKi cells.

Taken together the results show that ANGPTL4 regulates VEGF signaling in RCC cell lines and thereby cellular sensitivity to the clinically used VEGF pathway inhibitor, such as Sorafenib.

Therefore, this study shows that ANGPTL4 serves as a predictive marker for clinical sensitivity to VEGF-targeted therapy.

[000133] Next, to determine the importance of CXCR4 and ANGPTL4-targeted therapy in renal cell carcinoma, the CXCR4 inhibitor, AMD3100, and the anti-ANGPTL4 mAb, 14D12, is used in mono- and combination therapy to study effects on RCC tumor progression.

[000134] Renal cell lines under monolayer and stem cell conditions is plated, and various concentrations of AMD3100 and 14D12, and lengths of treatment is used to define the optimal parameters to inhibit cell proliferation by MTT assays. Figure 12 shows that Caki cell line with or without ANGPTL4 is treated with various concentrations of AMD3100. The knockdown of

ANGPTL4 results in a decrease in cell proliferation in AMD3100 treated Caki cell line.

[000135] Following CXCR4 and ANGPLT4-targeted therapy of RCC adherent stem cells, biomarkers can be assessed by RT-PCR, apoptosis measured by Annexin V staining and flow cytometry analysis, and invasion and angiogenesis analyzed in vitro.

[000136] Now referring to Figure 13 which shows that high ANGPTL4 expression is associated with poor patient survival. In this study, ANGPTL4 expression in the TCGA database for 187 RCC patients was plotted against long-term (> 2 yr) and short-term (< 2 yr) survival after diagnosis. P=0.0057. Therefore, Figure 13 shows that ANGPTL4 expression is inversely related to patient survival in renal cell carcinoma.

[000137] Therapeutic effects can also be evaluated in vivo following intra-cardiac mouse injections. Anti-cancer efficacy of systemically delivered AMD3100 and mAb 14D12 will be measured by bioluminescence imaging of live animals. In addition, the expression of RCC stem cell biomarkers within the extracted tumors will be measured by qPCR. Tissue will also be sectioned and stained for Ki-67, a nuclear protein associated with cell proliferation, to determine the effects of targeted CXCR4 and ANGPTL4 therapy on proliferation.

[000138] REFERENCES

1. Surawicz, T. S., Davis, F., Freels, S., Laws, E. R., Jr., and Menck, H. R. (1998) Brain tumor survival: results from the National Cancer Data Base. Journal of neuro-oncology 40, 151-160 2. Stupp, R., Mason, W. P., van den Bent, M. J., Weller, M., Fisher, B., Taphoorn, M. J., Belanger, K., Brandes, A. A., Marosi, C, Bogdahn, U., Curschmann, J., Janzer, R. C, Ludwin, S. K., Gorlia, T., Allgeier, A., Lacombe, D., Cairncross, J. G., Eisenhauer, E., and Mirimanoff, R. O. (2005) Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352, 987- 996 3. Wong, E. T., Hess, K. R., Gleason, M. J., Jaeckle, K. A., Kyritsis, A. P., Prados, M. D., Levin, V. A., and Yung, W. K. (1999) Outcomes and prognostic factors in recurrent glioma patients enrolled onto phase II clinical trials. J Clin Oncol 17, 2572-2578

4. Verhaak, R. G., Hoadley, K. A., Purdom, E., Wang, V., Qi, Y., Wilkerson, M. D., Miller, C. R., Ding, L., Golub, T., Mesirov, J. P., Alexe, G., Lawrence, M., O'Kelly, M., Tamayo, P., Weir, B.

A., Gabriel, S., Winckler, W., Gupta, S., Jakkula, L., Feiler, H. S., Hodgson, J. G., James, C. D., Sarkaria, J. N., Brennan, C, Kahn, A., Spellman, P. T., Wilson, R. K., Speed, T. P., Gray, J. W., Meyerson, M., Getz, G., Perou, C. M., Hayes, D. N., and Cancer Genome Atlas Research, N. (2010) Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell 17, 98-110

5. Furnari, F. B., Fenton, T., Bachoo, R. M., Mukasa, A., Stommel, J. M., Stegh, A., Hahn, W. C, Ligon, K. L., Louis, D. N., Brennan, C, Chin, L., DePinho, R. A., and Cavenee, W. K. (2007) Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes Dev 21, 2683-2710

6. Watanabe, K., Tachibana, O., Sata, K., Yonekawa, Y., Kleihues, P., and Ohgaki, H. (1996) Overexpression of the EGF receptor and p53 mutations are mutually exclusive in the evolution of primary and secondary glioblastomas. Brain pathology 6, 217-223; discussion 223-214

7. Yan, H., Parsons, D. W., Jin, G., McLendon, R., Rasheed, B. A., Yuan, W., Kos, I., Batinic- Haberle, I., Jones, S., Riggins, G. J., Friedman, H., Friedman, A., Reardon, D., Herndon, J., Kinzler, K. W., Velculescu, V. E., Vogelstein, B., and Bigner, D. D. (2009) IDH1 and IDH2 mutations in gliomas. The New England journal of medicine 360, 765-773

8. Thiery, J. P. (2002) Epithelial -mesenchymal transitions in tumour progression. Nat Rev Cancer 2, 442-454

9. Phillips, H. S., Kharbanda, S., Chen, R., Forrest, W. F., Soriano, R. H., Wu, T. D., Misra, A., Nigro, J. M., Colman, H., Soroceanu, L., Williams, P. M., Modrusan, Z., Feuerstein, B. G., and Aldape, K. (2006) Molecular subclasses of high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell 9, 157-173

10. Zhu, Y., Ghosh, P., Charnay, P., Burns, D. K., and Parada, L. F. (2002) Neurofibromas in NF1 : Schwann cell origin and role of tumor environment. Science 296, 920-922

11. Clarke, M. F., Dick, J. E., Dirks, P. B., Eaves, C. J., Jamieson, C. H., Jones, D. L., Visvader, J., Weissman, I. L., and Wahl, G. M. (2006) Cancer stem cells-perspectives on current status and future directions: AACR Workshop on cancer stem cells. Cancer Res 66, 9339-9344

12. Mani, S. A., Guo, W., Liao, M. J., Eaton, E. N., Ayyanan, A., Zhou, A. Y., Brooks, M., Reinhard, F., Zhang, C. C, Shipitsin, M., Campbell, L. L., Polyak, K., Brisken, C, Yang, J., and Weinberg, R. A. (2008) The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 133, 704-715

13. Marotta, L. L., and Polyak, K. (2009) Cancer stem cells: a model in the making. Current opinion in genetics & development 19, 44-50 14. Huntly, B. J., and Gilliland, D. G. (2005) Leukaemia stem cells and the evolution of cancer- stem-cell research. Nat Rev Cancer 5, 31 1 -321

15. Dontu, G., Abdallah, W. M., Foley, J. M., Jackson, K. W., Clarke, M. F., Kawamura, M. J., and Wicha, M. S. (2003) In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev 17, 1253-1270

16. Lee, A., Kessler, J. D., Read, T. A., Kaiser, C, Corbeil, D., Huttner, W. B., Johnson, J. E., and Wechsler-Reya, R. J. (2005) Isolation of neural stem cells from the postnatal cerebellum. Nat Neurosci 8, 723-729

17. Pollard, S. M., Yoshikawa, K., Clarke, I. D., Danovi, D., Strieker, S., Russell, R., Bayani, J., Head, R., Lee, M., Bernstein, M., Squire, J. A., Smith, A., and Dirks, P. (2009) Glioma stem cell lines expanded in adherent culture have tumor-specific phenotypes and are suitable for chemical and genetic screens. Cell Stem Cell 4, 568-580

18. Okochi-Takada, E., Hattori, N., Tsukamoto, T., Miyamoto, K., Ando, T., Ito, S., Yamamura, Y., Wakabayashi, M., Nobeyama, Y., and Ushijima, T. (2014) ANGPTL4 is a secreted tumor suppressor that inhibits angiogenesis. Oncogene 33, 2273-2278

19. Tan, M. J., Teo, Z., Sng, M. K., Zhu, P., and Tan, N. S. (2012) Emerging roles of angiopoietin-like 4 in human cancer. Mol Cancer Res 10, 677-688

20. Yang, C. FL, Yue, J., Pfeffer, S. R., Handorf, C. R., and Pfeffer, L. M. (201 1) MicroRNA miR-21 Regulates the Metastatic Behavior of B16 Melanoma Cells. J Biol Chem 286, 39172-39178 21. Liang, Y., Diehn, M., Watson, N., Bollen, A. W., Aldape, K. D., Nicholas, M. K., Lamborn, K. R., Berger, M. S., Botstein, D., Brown, P. O., and Israel, M. A. (2005) Gene expression profiling reveals molecularly and clinically distinct subtypes of glioblastoma multiforme. Proc Natl Acad Sci U S A 102, 5814-5819

22. Garner, J. M., Fan, M., Yang, C. FL, Du, Z., Sims, M., Davidoff, A. M., and Pfeffer, L. M. (2013) Constitutive Activation of Signal Transducer and Activator of Transcription 3 (STAT3) and

Nuclear Factor kappaB Signaling in Glioblastoma Cancer Stem Cells Regulates the Notch Pathway. J Biol Chem 288, 26167-26176

23. Lee, J., Kotliarova, S., Kotliarov, Y., Li, A., Su, Q., Donin, N. M., Pastorino, S., Purow, B. W., Christopher, N., Zhang, W., Park, J. K., and Fine, H. A. (2006) Tumor stem cells derived from glioblastomas cultured in bFGF and EGF more closely mirror the phenotype and genotype of primary tumors than do serum-cultured cell lines. Cancer Cell 9, 391 -403

24. Bhat, K. P., Balasubramaniyan, V., Vaillant, B., Ezhilarasan, R., Hummelink, K., Hollingsworth, F., Wani, K., Heathcock, L., James, J. D., Goodman, L. D., Conroy, S., Long, L., Lelic, N., Wang, S., Gumin, J., Raj, D., Kodama, Y., Raghunathan, A., Olar, A., Joshi, K., Pelloski, C. E., Heimberger, A., Kim, S. FL, Cahill, D. P., Rao, G., Den Dunnen, W. F., Boddeke, H. W., Phillips, H. S., Nakano, I., Lang, F. F., Colman, FL, Sulman, E. P., and Aldape, K. (2013) Mesenchymal differentiation mediated by NF-kappaB promotes radiation resistance in glioblastoma. Cancer Cell 24, 331-346

25. Eng, L. F., Ghirnikar, R. S., and Lee, Y. L. (2000) Glial fibrillary acidic protein: GFAP- thirty-one years (1969-2000). Neurochemical research 25, 1439-1451

26. Kahn, H. J., Marks, A., Thorn, H., and Baumal, R. (1983) Role of antibody to S100 protein in diagnostic pathology. American journal of clinical pathology 79, 341-347

27. Hirokawa, N., Hisanaga, S., and Shiomura, Y. (1988) MAP2 is a component of crossbridges between microtubules and neurofilaments in the neuronal cytoskeleton: quick-freeze, deep-etch immunoelectron microscopy and reconstitution studies. The Journal of neuroscience : the official journal of the Society for Neuroscience 8, 2769-2779

28. Marie, Y., Sanson, M., Mokhtari, K., Leuraud, P., Kujas, M., Delattre, J. Y., Poirier, J., Zalc,

B. , and Hoang-Xuan, K. (2001) OLIG2 as a specific marker of oligodendroglial tumour cells. Lancet 358, 298-300

29. Xie, Q., Thompson, R., Hardy, K., DeCamp, L., Berghuis, B., Sigler, R., Knudsen, B., Cottingham, S., Zhao, P., Dykema, K., Cao, B., Resau, J., Hay, R., and Vande Woude, G. F. (2008) A highly invasive human glioblastoma pre-clinical model for testing therapeutics. Journal of translational medicine 6, 77

30. Abdelzaher, E. (2013) CNS Tumors Pathology - Glioblastoma Multiforme. PathologyOutlines.com

31. Carro, M. S., Lim, W. K., Alvarez, M. J., Bollo, R. J., Zhao, X., Snyder, E. Y., Sulman, E. P., Anne, S. L., Doetsch, F., Colman, H., Lasorella, A., Aldape, K., Califano, A., and lavarone, A. (2010) The transcriptional network for mesenchymal transformation of brain tumours. Nature 463, 318-325

32. Lewis-Tuffin, L. J., Rodriguez, F., Giannini, C, Scheithauer, B., Necela, B. M., Sarkaria, J. N., and Anastasiadis, P. Z. (2010) Misregulated E-cadherin expression associated with an aggressive brain tumor phenotype. PLoS One 5, el 3665

33. Liu, W. M., Huang, P., Kar, N., Burgett, M., Muller-Greven, G., Nowacki, A. S., Distelhorst,

C. W., Lathia, J. D., Rich, J. N., Kappes, J. C, and Gladson, C. L. (2013) Lyn facilitates glioblastoma cell survival under conditions of nutrient deprivation by promoting autophagy. PLoS One 8, e70804

34. Okamura, T., Kurisu, K., Yamamoto, W., Takano, H., and Nishiyama, M. (2000) NADPH/quinone oxidoreductase is a priority target of glioblastoma chemotherapy. Int J Oncol 16,

295-303

35. Brat, D. J., Bellail, A. C, and Van Meir, E. G. (2005) The role of interleukin-8 and its receptors in gliomagenesis and tumoral angiogenesis. Neuro-oncology 7, 122-133

36. Strieter, R. M., Burdick, M. D., Mestas, J., Gomperts, B., Keane, M. P., and Belperio, J. A. (2006) Cancer CXC chemokine networks and tumour angiogenesis. Eur J Cancer 42, 768-778

37. Kleihues, P., and Sobin, L. H. (2000) World Health Organization classification of tumors. Cancer 88, 2887 38. Bonavia, R., Inda, M. M., Cavenee, W. K., and Fumari, F. B. (2011) Heterogeneity maintenance in glioblastoma: a social network. Cancer Res 71, 4055-4060

39. Rong, Y., Durden, D. L., Van Meir, E. G., and Brat, D. J. (2006) 'Pseudopalisading' necrosis in glioblastoma: a familiar morphologic feature that links vascular pathology, hypoxia, and angiogenesis. Journal of neuropathology and experimental neurology 65, 529-539

40. Katanasaka, Y., Kodera, Y., Kitamura, Y., Morimoto, T., Tamura, T., and Koizumi, F. (2013) Epidermal growth factor receptor variant type III markedly accelerates angiogenesis and tumor growth via inducing c-myc mediated angiopoietin-like 4 expression in malignant glioma. Molecular cancer 12, 31

41. Niu, G., Wright, K. L., Huang, M., Song, L., Haura, E., Turkson, J., Zhang, S., Wang, T., Sinibaldi, D., Coppola, D., Heller, R., Ellis, L. M., Karras, J., Bromberg, J., Pardoll, D., Jove, R., and Yu, H. (2002) Constitutive Stat3 activity up-regulates VEGF expression and tumor angiogenesis. Oncogene 21, 2000-2008

42. Semenza, G. L. (2003) Targeting HIF-1 for cancer therapy. Nat Rev Cancer 3, 721-732 43. Abou-Ghazal, M., Yang, D. S., Qiao, W., Reina-Ortiz, C, Wei, J., Kong, L. Y., Fuller, G. N., Hiraoka, N., Priebe, W., Sawaya, R., and Heimberger, A. B. (2008) The incidence, correlation with tumor-infiltrating inflammation, and prognosis of phosphorylated STAT3 expression in human gliomas. Clin Cancer Res 14, 8228-8235

44. Verine, J., Lehmann-Che, J., Soliman, H., Feugeas, J. P., Vidal, J. S., Mongiat-Artus, P., Belhadj, S., Philippe, J., Lesage, M., Wittmer, E., Chanel, S., Couvelard, A., Ferlicot, S., Rioux-

Leclercq, N., Vignaud, J. M., Janin, A., and Germain, S. (2010) Determination of angptl4 mRNA as a diagnostic marker of primary and metastatic clear cell renal-cell carcinoma. PLoS One 5, el 0421

45. Zhu, P., Goh, Y. Y., Chin, H. F., Kersten, S., and Tan, N. S. (2012) Angiopoietin-like 4: a decade of research. Biosci Rep 32, 211-219

46. Hussain, S. F., Kong, L. Y., Jordan, J., Conrad, C, Madden, T., Fokt, I., Priebe, W., and Heimberger, A. B. (2007) A novel small molecule inhibitor of signal transducers and activators of transcription 3 reverses immune tolerance in malignant glioma patients. Cancer Res 67, 9630-9636 47. Iwamaru, A., Szymanski, S., Iwado, E., Aoki, H., Yokoyama, T., Fokt, I., Hess, K., Conrad, C, Madden, T., Sawaya, R., Kondo, S., Priebe, W., and Kondo, Y. (2007) A novel inhibitor of the STAT3 pathway induces apoptosis in malignant glioma cells both in vitro and in vivo. Oncogene 26, 2435-2444

Claims

1. A method of predicting response of a subject to treatment with a VEGF -targeted therapeutic, comprising:

(a) providing a biological sample from the subject;

(b) determining a level of ANGPTL4 in the sample from the subject;

(c) comparing the level of ANGPTL4 in the biological sample with a reference, wherein the subject is predicted to be a likely responder or a likely non-responder based on the level of ANGPTL4 in the sample relative to the reference.

2. The method of claim 1, wherein the subject is predicted to be a likely non-responder to VEGF targeted therapeutic when there is a measurable decrease in the level of ANGPTL4 in the sample relative to the reference, and the subject is identified as a likely responder to the VEGF -targeted therapeutic when there is a measurable increase in the level of ANGPTL4 in the sample relative to the reference.

3. The method of claim 2, and further comprising administering a VEGF targeted therapeutic to the subject when the subject is predicted to be a likely responder.

4. A method of diagnosing a disease treatable with a VEGF targeted therapeutic in a subject, comprising:

(a) providing a biological sample including one or more cells from the subject;

(b) determining the level of ANGPTL4 in the sample from the subject;

(c) comparing the level of ANGPTL4 in the sample with a reference;

(d) diagnosing the subject as having a disease treatable with VEGF targeted therapeutic if there is a measurable difference in the level of ANGPTL4 in the one or more cells in the sample as compared to the reference; and

(e) administering an effective amount of a VEGF targeted therapeutic to the subject having a disease treatable with VEGF targeted therapeutic. .

5. The method of any one of the prior claims, wherein the biological sample is serum.

6. The method of any one of the prior claims, wherein the biological sample is cancer tissue.

7. The method of any one of the prior claims, wherein the disease is brain cancer.

8. The method of any one of claims 1-6, wherein the disease is breast cancer.

9. The method of any one of claims 1-6, wherein the disease is type 2 diabetes.

10. The method of any one of claims 1 -6, wherein the disease is macular degeneration.

11. The method of any one of claims 1 -6, wherein the disease is kidney cancer.

12. The method of claim 11 , wherein the disease is renal cell carcinoma.

13. The method of any one of the prior claims, wherein the VEGF targeted therapeutic is a biological molecule or small molecule that directly or indirectly modulates the VEGF signal transduction pathway.

14. The method of any one of the prior claims, wherein the VEGF targeted therapeutic is a VEGF inhibitor.

15. The method of claim 14, wherein the VEGF inhibitor is Sorafenib.

16. The method of any one of the prior claims, wherein the reference comprises a level of the ANGPTL4 in a sample from the subject taken over a time course.

17. The method of any one of the prior claims, wherein the reference comprises a sample from the subject collected prior to initiation of a treatment program including use of a VEGF targeted therapeutic and the biological sample is collected after initiation of the treatment program.

18. The method of any one of the prior claims, wherein the reference comprises a standard sample.

19. The method of any one of the prior claims, wherein the reference comprises control data.

20. The method of any one of the prior claims, wherein the reference comprises a level of ANGPTL4 in a sample in one or more samples from one or more individuals who are known responders or who are known non-responders to treatment of the disease with a VEGF targeted therapeutic.

21. The method of any one of the previous claims, wherein the step of determining the a level of ANGPTL4 comprises at least one of determining the level of ANGPLT4 protein; determining the level of mRNA encoding ANGPLT4; determining the copy number of ANGPLT4; and determining the expression level of any non-coding or coding polymorphism within the nucleotide sequence ANGPLT4.

22. A kit, comprising an agent that selectively binds to ANGPTL4.

23. The kit of claim 22, wherein the agent comprises probes or primers to detect ANGPTL4.

24. The kit of claim 23, wherein the agent is an antibody.