[CANCER RESEARCH 58. 4185-4192, September 15, 1998]

Tumor Growth Modulation by Sense and Antisense Vascular Endothelial GrowthFactor Gene Expression: Effects on Angiogenesis, Vascular Permeability, BloodVolume, Blood Flow, Fluorodeoxyglucose Uptake, and Proliferation ofHuman Melanoma Intracerebral Xenografts1

Takamitsu Oku, Juri G. Tjuvajev, Tadashi Miyagawa, Toshio Sasajima, Arjun Joshi, Revathi Joshi, Ronald Finn,Kevin P. Claffey, and Ronald G. Blasberg2

Departments of Neurology [T. O.. J. G. T., T. M., T. S.. A. J., R. J., R. G. B.j and Radiology ¡R.FJ, Memorial Sloan-Kettering Cancer Center, ¡275 York Avenue, New York, NewYork 10021; Department of Nenrosnrgery, Miya~aki MédicalCollège 5200 Kihara, Kivotake, Miyazaki 889-16 Japan IT. O¡: ami Departmeni of Pa(lto!oi>\, Beth Israel

Deaconess Medical Center and Howard Medical School, Boston. Massachusetts 02215 [K. P. Cl

ABSTRACT

Vascular endothelial growth factor (VEGF), also known as vascularpermeability factor, has been investigated as a potent mediator of braintumor angiogenesis and tumor growth. We evaluated the effect of VEGFexpression on the pathophysiology of tumor growth in the brain. HumanSK-MEL-2 melanoma cells, with minimal VEGF expression, were stably

transfected with either sense or antisense mouse VEGF cDNA and used toproduce intracerebral xenografts. Vascular permeability, blood volume,blood flow, and tumor fluorodeoxyglucose metabolism were assessed usingtissue sampling and quantitative autoradiography. Tumor proliferationwas assessed by measuring bromodeoxyuridine labeling indices. Tumorvascular density and morphological status of the blood-brain barrier wereevaluated by immunohistochemistry. SK-MEL-2 cells transfected with

sense VEGF (V+) expressed large amounts of mouse and human VEGFprotein; V+ cells formed well-vascularized, rapidly growing tumors with

minimal tumor necrosis. V+ tumors had substantial and significant increases in blood volume, blood flow, vascular permeability, and fluorodeoxyglucose metabolism compared to wild-type and/or V- (antisenseVEGF) tumors. VEGF antisense transfected V- expressed no detectableVEGF protein and formed minimally vascularized tumors. V- tumors

had a very low initial growth rate with central necrosis; blood volume,blood flow, vascular permeability, and glucose metabolism levels were lowcompared to wild-type and V+ tumors. A substantial inhibition of intrac

erebral tumor growth, as well as a decrease in tumor vascularity, bloodflow, and vascular permeability may be achieved by down-regulation ofendogenous VEGF expression in tumor tissue. VEGF-targeted antiangio-

genic gene therapy could be an effective component of a combined strategy to treat VEGF-producing brain tumors.

INTRODUCTION

Antiangiogenic therapy of solid tumors was proposed by Folkmanet al. (1) more than 20 years ago, and a number of targets forantiangiogenic therapy have been investigated. These included severalangiogenic factors, such as bFGF3 (2), angiogenin (3). and VEGF

Received 2/18/98; accepted 7/10/98.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by the departmental research fund (Department of Neurology.

Memorial Sloan Kettering Cancer Center. New York; to J. G. T.); the RassmussenFoundation (to K. P. C.); NIH Grants CA-64436 (to K. P. C.) and CA-60706 and CA-69769 (to R. G. B.); and Grant DOE 86ER60407.

2 To whom requests for reprints should be addressed, at Department of Neurology.

K923. Memorial Sloan-Kettering Cancer Center, 1275 York Avenue. New York. NY

10021.3 The abbreviations used are: bFGF. basic fibroblasl growth factor; VEGF. vascular

endothelial growth factor; DTPA, diethylenetriaminepentaacetic acid chelate; FDG. fluorodeoxyglucose; IAP, iodoantipyrine; CMV. cytomegalovirus; BBB. blood-brain barrier;BrdUrd. 5-bromodeoxyuridine; EC, endothelial cell; i.e.. intracerebral(ly); EBA. endothelial barrier antigen: FVIII, factor Vili; TGF. transforming growth factor; R^, relativeglucose utilization; F, tissue perfusion; kt, lower limit estimate of vascular permeability;GLUT-1. glucosetransporter type one; KGF. keratinocyte growth factor; PDGF. platelet-derived growth factor; PAF, platelet-activating growth factor.

(4-9), as well as the use of angiostatic peptides such as endostatin

(10), angiostatin (11), and thrombospondin (12).VEGF was originally described as a permeability factor, a tumor-

secreted protein that increases microvascular permeability in vivo witha potency 50,000 times that of histamine ( 13). In vitro, VEGF hasbeen shown to promote EC-specific mitosis, induction of intracellularCa2 +, and EC migration. VEGF treatment of ECs initiates VEGF-

receptor (tyrosine kinase receptor) autophosphorylation, and its downstream signal transduction cascades include phospholipase C-y, phos-phatidylinositol 3-kinase, ras GTPase, and the oncogenic adapter

NcK. This cascade results in the induction of several proteins that areimportant in the angiogenic process. These proteins include activatorsof plasminogen that play a critical role in hydrolyzing the endothelialbasement membrane and the integrin subunit proteins that are important for cell migration (14-17: for review, see Ref. 18).

Brain is an attractive organ for the investigation of the role of VEGFin tumor angiogenesis. As a host tissue, brain has a unique and clinicallyimportant characteristic, namely, the BBB and blood-tumor barrier. Olio-

blastoma is the most common malignant brain tumor and is one of themore highly vascularized tumors, often with a disrupted blood-tumor

barrier. Similarly, metastatic melanoma is a rapidly growing tumor inbrain and results in considerable morbidity and is frequently the directcause of death. Both glioblastoma and metastatic melanoma are resistantto current modes of therapy. Recent studies on specimens of surgicallyresected tumors have revealed that the expression of VEGF and itstyrosine kinase receptor KDR/flk-1 is up-regulated in a variety of brain

tumors including glioblastoma and melanoma (19).It was shown previously (5, 9) that suppression of endogenous

VEGF gene expression using antisense gene transfer resulted in theinhibition of s.c. tumor growth and metastasis. Only one previousstudy (20) used antisense VEGF to suppress i.e. tumor growth in mice,but none of these studies have investigated the effects of sense andantisense VEGF gene transfer on tumor vascular physiology andglucose metabolism.

In this report, we applied sense and antisense gene transfer technology to modulate the expression of VEGF in stably transduced V+and V—SK-MEL-2 human melanoma cells. We continued studies

initiated by Claffey et al. (9) and examined the role of VEGF expression in the progression of i.e. V+ and V- xenograft growth. We

assessed the effects of up- or down-regulation of VEGF expression on

microvascular density, blood volume, blood flow, and FDG metabolism of i.e. growing melanoma xenografts.

MATERIALS AND METHODS

Cell Culture, VEGF cDNA, Transfection, and Selection

Three stably transfected cell lines derived from the human melanoma cellline SK-MEL-2 were used in these studies. Briefly, the full-length mouse

j-amino acids cDNA in sense or antisense orientation was cloned into

4185

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANGIOGENESIS BY V- EXPRESSION

an expression vector (pCMV-NEO) in which the transcription of transgene isdriven by a CMV enhancer-promoter, and a neomycin phosphotransferase

gene (G418 resistance) is driven by the SV40 promoter as described previously(9). As a control, other SK-MEL-2 cells were transfected with vector alonewithout a VEGF insert. Constitutive expression of vector-derived VEGF

mRNA (sense, antisense) was confirmed by Northern blot. Two stably transfected cell lines containing the murine VEGF insert with sense (SK-MEL-2V+) or antisense (SK-MEL-2 V-) expressed high levels of vector-derived

mRNAs. SK-MEL-2 cells transfected with vector alone (SK-MEL-2 Nl)

expressed no detectable murine VEGF mRNA (9). Growth of all of the threetransfectant cell lines was tested by counting cells growing in 150-mm culture

dishes daily for 8 days (duplicate studies).

Capacity of VEGF Expression

Cells were grown in 150-mm culture dishes, and the experiments were

initiated when the cells approached near confluence. The medium was changed3 h before cobalt chloride (CoCl,) and hypoxic atmosphere studies. Cobaltchloride experiments were performed in a conventional incubator (20% oxygen and 5% CO2). In CoCl2 experiments, 250 /¿Ior 500 /*! of 10 mM CoCI2were added to 25 ml of serum free medium to produce the final concentrationsof 100 /J.Mor 200 /¿MCoCl2, respectively. The hypoxic chamber studies wereconducted in a nitrogen controlled incubator NU3500 (NuAire, Plymouth,MN) at 2 or 5% oxygen and 5%CO2 in a nitrogen atmosphere. Samples ofculture medium were obtained at 0, 3, 6, 12, and 24 h after exposure to CoCl2or a hypoxic atmosphere, and assayed for human and murine VEGF concentration using the human and mouse VEGF ELISA kits (R&D Systems, Minneapolis, MN). Cells were trypsinized and counted in each culture dish. TheVEGF protein concentration in the conditioned media was normalized by cellnumber and medium volume in the corresponding culture dish.

Radiopharmaceutical Preparation

[18F]-FDG was synthesized by the cyclotron/chemistry core at Memorial

Sloan-Kettering Cancer Center (New York) by a nucleophilic reaction usingthe modification described by Harnacher et al. (21, 22). [67Ga]-DTPA solutionwas prepared mixing [67Ga]-citrate (42 Ci/mmol) solution (Mallinckrodt Médical. Hicksville, NY) with 0.1 volume of a 1 mM cold DTPA and 3.0 X 10~3

mol/1 CaCl2 saline solution and adjusted to pH 7.4. 4-[Aí-methyl-uC]-IAP

(40-60 mCi/mmol) was purchased from American Radiolabeled Chemicals(St. Louis, MO). [67Ga]-DTPA and [I4C]-IAP were assayed by high-pressureliquid chromatography before use to confirm purity >98%. [99mTc]-labeledRBCs were prepared by mixing ["'"Tel-sodium pertechneate with autologous

RBCs using UltraTag RBC Kit (Mallinckrodt Medical) according to themanufacturer's protocol.

i.e. Xenograft Production and Study Protocol

The experimental protocol involving animals was approved by the Institutional Animal Care and Use Committee of the Memorial Sloan-KetteringCancer Center. RNU/rnu rats weighing 300-350 g (HaríanSprague Dawley,

Indianapolis, IN) were anesthetized with a gas mixture containing Isoflurane2%, 68% nitrous oxide, and 30% oxygen. A 2% lidocaine gel was applied tothe ears, and the head was fixed in a stereotaxic device. After a midline scalpincision, bilateral burr holes were made through the skull 1 mm posterior and3 mm lateral to the bregma. The cell suspensions were aspirated in a 100-/xlgas tight Hamilton syringe with a 25-gauge needle and attached to a stereotaxicdevice. The needle was then inserted 6 mm deep and the V + , V-, or Nl tumorcells (5 x 10s cells in 10 jul) were stereotactically injected into opposite

hemispheres.In experimental sets 1-3,n = 6 for each i.e. xenograft pair: (a) V- andV + ;

(b) V —and N1. In set 1, V + and V —or N l and V —tumor cells were injected

i.e. into opposite hemispheres, and the animals were studied 8 days later toassess xenograft growth and vascular physiology. In set 2, V- tumor cells

were injected i.e. 2 weeks before contralateral i.e. injection of V+ orNl tumorcells to obtain xenografts of comparable size for assessment of vascularphysiology. In set 3, V- tumor cells were injected i.e. 2 weeks before

contralateral i.e. injection of V+ or Nl tumor cells to obtain xenografts ofcomparable size for blood volume measurements.

Femoral artery and vein catheters were placed under gas anesthesia. The

hind quarters were wrapped in plaster bandage up to the mid-thorax for

restraint, and the rats were allowed to recover from anesthesia for at least 2 h.Arterial blood pressure was monitored, and body temperature was maintainedat 37°C with a heat lamp that was automatically controlled by a rectal

temperature probe.For experimental sets 1 and 2 (blood flow, vascular permeability, and

glucose metabolism measurements), BrdUrd dissolved in PBS (15 mg/ml)

was injected i.v. 3 h before the animals were killed. Continuous withdrawalof arterial blood at the rate of 0.025 ml/min (Harvard Apparatus, South-nawtick, MA) was initiated 1-2 min before i.v. injection of 5.0 mCi of[18F]-FDG. At 50 min after FDG injection 5 mCi of [67Ga]-DTPA was

injected. At 59 min 30 s after FDG injection a constant infusion (2 cc/min)of [14C]-IAP (25 fiCi in 1 cc saline) was initiated (30 s before decapitation)

and arterial blood was sampled every 5 s. At 60 min after FDG injection theexperiment was terminated by decapitation, and a final arterial bloodsample was obtained.

For experimental set 3 (blood volume measurements), ["'"TcJ-labeled

RBCs were injected i.v. over 1 min. The animal was killed by decapitation 10min later, and arterial blood was collected. The brain was rapidly removed, andsamples of tumor and normal brain were dissected, weighed, and processed forgamma spectroscopy.

Radioactivity Assay

All tissue, blood, and plasma samples were solubilized with Soluene-350®

(Packard Instrument Co., Meriden, CT) and assayed for radioactivity. For sets1 and 2 (triple-label experiments), [18F] and |67Ga] radioactivity was measuredfirst in a 3-channel gamma counter (AutoGamma 5550 Spectrometer®: Pack

ard Instrument Co.) using narrow windows settings, splash correction, anddecay correction. After [I8F] decay (24 h, 13 half-lives), the samples wererecounted to determine [67Ga] radioactivity alone. The samples were thenstored at 4°Cfor 45 days to allow for [67Ga] decay (14 half-lives). A scintillantInsta-Fluor®(Packard Instrument Co.) was added and [I4C] radioactivity wasdetermined by liquid scintillation counting (Tri-Carb® Liquid Scintillation

Analyzer, Model 1600TR; Packard Instrument Co.) using external standardquench corrections. For set 3 (single-label experiment), [99mTc] radioactivity

was determined by the gamma counter.Triple-label Quantitative Autoradiography. Tissue processing and tri

ple-label autoradiographic techniques were described previously (23, 24).Briefly, 22 serial coronal sections were cut at 20-jU.mthickness at intervals of400-600 /¿min a cryomicrotome at —13°C.and the sections were used forautoradiography and histology. [I8F] and [67Ga] autoradiographic standards

with different radioactivity concentrations were prepared, frozen, and cut 20firn thick on the cryomicrotome, and the radioactivity of each standard wasmeasured in the gamma counter. Sixteen ('4C]-methylmethacrylate standards

(Amersham Corp., Arlington Heights, IL), previously calibrated to brain tissueradioactivity for the range of 4.4-2354 nCi/g brain, were also used. The tissuesections and autoradiographic standards were exposed to X-ray film (SB-5;

Kodak, Rochester, NY) for three different time intervals to generate threeseparate autoradiograms. The resulting autoradiographic images have previously been shown to represent the tissue distribution of [I8F], [67Ga], and [I4C]

radioactivity, respectively (24). All of the sections used for autoradiographicexposure were subsequently fixed in acetone and stained with H&E or tolui-

dine blue.

Autoradiographic Image Analysis. Digitization and registration of corresponding images from the three autoradiograms and histology was performedusing a CCD-72 series video camera (Dage-MTI Inc., Michigan City, IN), a

microcomputer imaging system, and MCID software (Imaging Research Inc.,Ontario, Canada). The optical density of the standard images on X-ray film

was measured and a standard curve that relates mean optical density to

radioactivity, expressed as % administered dose/g tissue weight, was generatedfor each film. The same tissue section was used to generate each of the threecorresponding autoradiographic images as well as the histological image. TheMCID software allowed us to draw regions of interest on the histologicalimage based on morphological criteria, to transpose those regions to thecoregistered autoradiographic images, and to obtain measurements of meantissue radioactivity, ±SD. Parametric images of tissue perfusion (F), thelower-limit plasma clearance (influx) constant (Ai,), and relative glucose utili-

4186

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANOIOGENESIS BY V- EXPRESSION

zation (Rg|U) were also generated using MCID software and color-coded to a

range of values.

Calculations

All radioactivity measurements were converted to % injected dose/g tissueor blood (plasma) to facilitate comparisons between different animals. Bloodvolume (Vh, ml/hg) was calculated from the ["mTc]-labeled RBC data by:

Vb = (AT)/CbT) X 100 (A)

where AT is the measured radioactivity (% administered dose/g tissue) at timeT (10 min) and CbT is the arterial blood concentration (% administereddose/ml).

Tissue perfusion (F, ml/min/g) was calculated from the IAP data (25) anddepends on the tracer distribution relationships developed by Kety (26, 27):

..,/, (B)

where /4T is the tissue radioactivity measured autoradiographically at 30 s,Cb(t) is the concentration of IAP in arterial blood over the 30-s experimentalperiod, A is the tissue-blood partition coefficient of IAP, and k is a term that

incorporates F. F is calculated from:

F = U/m (C)

where m is a constant which defines the extent of tracer (IAP) equilibriumbetween tissue and blood. Based on the work of Sakurada et al. (25), m wastaken to be 1.0 and Àwas found to be equal to 0.8 ml/g.

Vascular permeability (AT,,/xl/min/g) was calculated from the [67Ga]-DTPA

data:

K, = -vhx cbr)/ c,(t)dt (D)

where }Cf(t) dt is the plasma concentration-time integral (input function).

/Cp(f) dt was calculated from the measurement of total radioactivity withdrawn during the experiment and knowledge of the rate of withdrawal and timeof exposure (28). The values of Vb for tumor and brain were obtained from the[""Tel-labeled RBC data (Equation A above). Equation D assumes unidirec

tional movement of tracer, blood-to-tumor, and is necessary for accuratemeasurements of vascular permeability from single-time experiments. Toavoid significant tumor-to-blood backflux of the tracer during the experimental

period, the experimental time was kept short (10 min; Ref. 29). AT,is alower-limit estimate of vascular permeability.

Rgiu (/¿moles/min/g) of brain and tumor (30, 31) was calculated from theFDG data because the "lumped constant" for tumor tissue in the operational

equation of the deoxyglucose method (32) is not known:

= A-tl \ [Cp*(f)/C.t(0]</f (E)

where Cp°is plasma FDG and Cp+ is the plasma glucose concentration.

Histology and Immunohistochemistry Procedures

Frozen tissue sections ( 10 ¿im)adjacent to the sections used for autoradiogra-phy were used for morphological characterization of V—, V+, and Nl i.e.

xenografts. For BrdUrd staining, sections were incubated in 2 N HC1. After threewashes in PBS, sections were incubated in 1% hydrogen peroxide in methanol for30 min, washed again, and incubated with the blocking serum. Then sections wereincubated with primary antibody at 4°Covernight. Subsequent steps were accom

plished using the Vectastain Elite kit for mouse and rabbit primary antibodies(Vector Laboratories, Burlingame, CA) according to the manufacturer's protocol:

3,3-diaminobenzidine tetrahydrocloride was used as a chromagen that provided

brown stain. Sections were counterstained with methyl green (Vector Laboratories), which provided a light-green nuclear stain. Primary antibodies, dilution rates,and fixations were as follows: (a) VEGF staining, fixed in 0.2% 2-mercaptoeth-

anol and 4% paraformaldehyde in PBS (pH 7.6) and rabbit polyclonal antihumanrecombinant VEGF|65 at 1:100 dilution (Neo Markers, Fremont, CA); (b) FVIII-related antigen staining, fixed in acetone and rabbit antihuman FVIII-related

antigen at 1:100 dilution (DAKO, Carpintería,CA); (c) EBA staining, fixed inacetone and mouse monoclonal anti-EBA of rat SMI 71 at 1:100 dilution (Stern-

berger Monoclonals, Baltimore. MD); and (d) BrdUrd staining, fixed in acetonefor 10 min and mouse monoclonal anti-BrdUrd IU-4 at 1:100 dilution (Caltag

Laboratories, San Francisco. CA). For the BrdUrd labeling indices assay, imagesof BrdUrd-stained tissues were obtained randomly using a digital video micros

copy system. The labeling indices were determined by counting more than 200cells in a field of view (X40).

Statistics

Comparisons between region of interest mean values were analyzed by twostatistical methods. A paired / test was used for comparisons within individualanimals (e.g., between two different tumors in opposite hemispheres), and anonpaired t test was applied for comparisons between different sets of animals.Significant differences were determined at P < 0.05.

RESULTS

Capacity of VEGF Expression in Vitro. All of the three transducedcell lines (V—,Nl, and V+ cell lines) grew at equivalent rates in vitro

(data not shown). We tested for cross-reactivity of the human and mouse

VEGF ELISA kits (R&D Systems) to mouse recombinant VEGFI64 orhuman recombinant VEGF)65 protein, respectively. The human VEGFELISA kit recognized only human VEGF (lower assay limit < 5 pg/ml),and the mouse VEGF ELISA kit recognized only mouse VEGF (lowerassay limit < 3 pg/ml). These ELISA kits provided selective measurements of human VEGF protein and mouse VEGF protein concentrationsin the conditioned culture medium (Table 1). Both mouse and humanVEGF proteins were expressed in a linear fashion for 24 h.

Mouse VEGF protein was constitutively overexpressed only in theV+ cells (320 pg/106 cells/h under 20% oxygen), and exposure of V +

cells to CoCl2 did not increase mouse VEGF production. However,mouse VEGF protein expression was somewhat depressed in a low-oxygen atmosphere (165 pg/106 cells/h under 5% oxygen and 119pg/106 cells/h under 2% oxygen). Neither V—nor Nl cells expressed

measurable levels (<3 pg/ml) of mouse VEGF.

Table 1 Human and mouse VEGF protein expression over time in vitro after exposure to CoC/i or Hypoxie atmosphere <pg/Iu cells/hi

CoCl2 concentration (20% O2)0mM100mM200

mMHypoxieatmosphere

5% O,2%02SK-MEL

V - cells" SK-MEL Nlcells*human

VEGF humanVEGF(-)

123±4(-)138±5(-)271±9(-)

103 ±8(-) 114 ±11SK-MEL

V+human

VEGF195

4250333014138

9181 13cells*mouse

VEGF320

133251532017165

10119 11

" Below the lower limit of human VEGF ELISA kit sensitivity (<5 pg/ml).h Values are mean ±SD.

4187

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANGIOGENESIS BY V- EXPRESSION

Table 2 i.e. tumor growth and proliferative activity (BrdUrd labeling index)"

Tumor volume (mm )BrdUrd labeling index (%)SKMEL

V-tumorsset

1 (8 days) set 2 (3wk)1.6±1.4* 24.8 ±8.0

<2C 2.1 ±3.5SKMEL

Nl tumors(8days)12.2

±6.526.6 ±6.2SKMEL

V+tumorsset

1 (8 days) set 2 (8days)34.0±11.9* 29.5 ±4.8*

3 1.5 ±5.5 30.7 ±3.4a Values are mean ±SD, n = 6 for each set of tumors.* Significantly different from control Nl tumors at P < 0.0001 by ANO VA test.c A reliable estimate of the BrdUrd labeling index was not possible because of the gradient and heterogeneity of labeled cell distribution.

Human VEGF protein was expressed by V+ and Nl cells, whereasno detectable human VEGF protein (<5 pg/ml) was present in the V—

cell culture medium. The baseline level of human VEGF proteinexpression by V+ tumor cells was high (195 pg/106 cells/h) under the

normoxic conditions and increased more when exposed to CoCl2 (upto 330 pg/106 cells/h). Of interest, the level of human VEGF protein

expression was 1.8-fold higher in V+ cells compared with Nl cells

under both normoxic and hypoxic conditions. Human VEGF proteinexpression in V+ cells was lower (138 pg/106 cells/h) under modest

hypoxia (5% oxygen) compared with that under normoxic conditions,but increased (181 pg/106 cells/h) under more severe hypoxic condi

tions (2% oxygen). As expected, V- tumor cells expressed no de

tectable levels of either human or murine VEGF proteins under eithernormoxic (20% oxygen) or hypoxic conditions (low-oxygen atmo

sphere or CoCl2). This observation indicates a near complete blockadeof human VEGF expression and absence of human VEGF inductionby hypoxia in transduced V- tumor cells (Table 1).

Growth of i.e. Xenografts. i.e. growth of the three tumor cell lines(V+, V-, and Nl) was assessed at 8 days after i.e. implantation into

immunodeficient RNU/rnu rats. Tumor volume measurements wereobtained and morphological characteristics were assessed in H&E-

stained sections. V+ xenografts were significantly larger than Nlxenografts, and V- xenografts were significantly smaller than Nlxenografts (Table 2; Fig. \A). Although V- xenografts were small, all

of the xenografts had a relatively large area of central necrosis (Fig.1, C and D). V- xenografts initially grew very slowly but reached a

modest size 3 weeks after i.e. inoculation (Table 2; Fig. IE). Thenecrotic area of 3 week-old V—xenografts was always larger than

that in V+ or Nl xenografts of corresponding size.Tumor cell proliferation in the brain was assessed by BrdUrd

labeling index. In the V+ xenografts, a homogeneous pattern of activeproliferation was observed 8 days after i.e. inoculation, which wasconsistent with other results in set 1 and set 2 studies (Table 2; Fig.Iß).In contrast, low proliferative activity was observed in the V-

xenografts 8 days after i.e. inoculation (set 1 study); the BrdUrd-

positive cells were observed mainly in the periphery of the xenografts(Fig. 1, C and D). Three weeks after i.e. injection of V—cells (set 2

study), the BrdUrd labeling index was more homogeneous (Fig. IK)but significantly lower than that in V+ (Fig. \L) or Nl xenograftsgrowing in the opposite hemisphere (Table 2).

Tumor Blood Volume, Blood Flow, Vascular Permeability, andRelative Glucose Utilization. Measurements were performed 8 daysafter i. c. injection of V+ and Nl cells and 3 weeks after V- cell

injection to achieve comparable xenograft size (Fig. \E). Blood volume of V+ xenografts was very high (4-fold more than that of Nlxenografts); blood volume of the V- xenografts was lower than that

of N l xenografts but slightly higher than normal brain cortex (Table3). Blood flow was homogeneously high in V+ xenografts and verylow in V—xenografts (Table 3; Fig. Iff). Blood flow in Nl xenograftswas intermediate between that in V+ and V- xenografts (Table 3).The [67Ga]-DTPA images showed a clear difference in vascular

permeability between the V+ and V—xenografts (Fig. IF). Vascularpermeability (Kt) measured with [67Ga]-DTPA was 45-fold greater in

V+ xenografts compared to that in V- xenografts, and 32-fold

greater than that of Nl xenografts (Table 3). Mean FDG uptake andRg)u were lower in the V—xenografts compared to the V+ (Fig. 1C)and Nl xenografts (Table 3). The pattern of Rglu in V- xenografts

was heterogeneous. A rim of relative hyper-metabolism was observedin the margin of V- xenografts (Fig. 1C) that consisted of sparsely

distributed islets of invading tumor cells in brain parenchymal tissueadjacent to the tumor (Fig. 1, C, D, N, and P).

Tumor Vasculature. An expansive pattern of tumor growth in thebrain was observed with all of the three cell lines. However, themargin of V- xenografts contained sparse islets of tumor cells that

appeared to be invading adjacent brain tissue (Fig. l,C, D, N, and P).Immunohistochemical staining for EBA and FVIII was performed toevaluate tumor and brain vasculature. In V+ xenografts, the absenceof EBA-positive vessels was observed not only within the tumor but

also along the tumor margin and in the adjacent brain beyond thetumor margin (Fig. \M). The distance between the tumor margin andEBA-positive vessels was 0.46 ±0.18 mm and 0.15 ±0.06 mm inV+ and Nl xenografts, respectively. In V- xenografts, EBA-positivevessels were observed in adjacent brain as well as within the V-

xenograft mass (Fig. \N). The V+ xenografts were better vascular-ized than the Nl or V- xenografts. Large FVIII-positive vessels were

noted to branch and penetrate directly into the xenografts from adjacent brain (Fig. \O). This branching and penetration pattern wasuncharacteristic of the V—xenografts (Fig. IP).

DISCUSSION

Neovascularization plays an important role in tumor progressionand metastasis. One of the key mediators of angiogenesis is VEGF, amultifunctional growth factor that is overexpressed and secreted by amajority of human and animal tumors (33-42). To further elucidate

the pathophysiological effects of VEGF induced during tumor growth,we used tumor cells in which the capacity of VEGF expression hasbeen modulated by sense or antisense VEGF cDNA transfection, asdescribed earlier by Claffey et al. (9). The rationale for using senseand antisense VEGF cDNA transfected cells is that this approachallows for the selection of tumor cell clones with defined constitutiveand inducible levels of VEGF expression. Our results complement theprevious report by Claffey et al. (9) and provide additional evidencethat VEGF plays an important role in facilitating tumor growth inbrain by stimulating tumor neoangiogenesis and by altering bothtumor and adjacent host tissue vascular physiology.

SK-MEL-2 human melanoma cells with low levels of human

VEGF expression were stably transfected with murine sense VEGFcDNA (V+ cells) and with murine antisense VEGF cDNA (V-

cells); a vector-only transfected cell line was used as a control (Nl

cells). The V+ cell line expresses both human and mouse VEGF inculture. Human VEGF expression by V+ cells was up-regulated by

exposure to CoCl2, whereas mouse VEGF was constitutively high andwas not further up-regulated. This observation is consistent with theabsence of an effective promoter element for up-regulating the expression of the transduced mouse VEGF gene under hypoxic condi-

4188

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANGIOGENESIS BY V- EXPRESSION

B «¿'^r .-C* V * "* !f¿- ' ** '••v'-*. «¿«v*-'S*-»

'l * V- '.'i? ¿-'r*-. ••

:"•;••''•XrV.'V*/-' •

DN

* • N

tumor •••'.»'

V+ V-tumor Histology tumor(day 8) (3 weeks)

K

Fig. 1. Panels A-D, i.e. V+ and V—human melanoma xenografts 8 days after i.e. injection of 5 X IO5 V+ and V—tumor cells into the left and right hemisphere, respectively,

of a RNU/rnu rat. The H&E-stained macrohistology of the brain xenografts (A) and BrdUrd immunohistochemically stained tumor regions (B-D) inelude a high-power view of the V +tumor (B) showing a high labeling index, a low-power view of the V—tumor (C) showing central necrosis (letter AOand proliferation along the tumor periphery, and a high-powerview of the same V- tumor (D\ letter A' indicates necrosis). Panels E-P. a RNU/rnu rat with a left hemisphere V+ human melanoma xenograft (8 days after i.e. injection of 5 X 10sV+ tumor cells) and a right hemisphere V- human melanoma xenograft (3 weeks after i.e. injection of 5 X IO5 V—tumor cells). The H&E-stained macrohistology of the xenografts(E) and triple-label quantitative autoradiographic images of vascular permeability (K, for [ft7Ga]-DTPA: panel F), relative glucose metabolism {R¥,ufor [18F1-FDG; panel G). and bloodflow (tissue perfusion measured with [I4C]-IAP: panel H) are shown. The histology and autoradiographie images were generated from the same tissue section; the autoradiograms arecolor-coded to a range of values shown in the scale bar at the right of each panel. Note the high vascular permeability of the V + tumor compared with the V - tumor and normal braintissue (panel /O; the lower relative glucose metabolism for the V—tumor compared with the V+ tumor and the hypermetabolism in brain adjacent to the V—tumor (G'); and the lowblood flow in the V—tumor (particularly at the center of the tumor) compared with the V+ tumor (//). / and J, magnification views of the V4- xenograft shown in panels E and /•",

respectively. Note the increase in vascular permeablily of brain tissue adjacent to the V+ tumor (7); brain adjacent to the V+ tumor shows no macroscopic infiltration of tumor cellsbeyond the tumor margin (/). K and L, BrdUrd immunohistochemically stained V- and V+ tumor regions, respectively; note the difference in labeling index between the V- andV+ xenografts. M and N, EBA staining in V+ and V- tumors, respectively (letter T indicates tumor). EBA staining of blood vessels is absent in V+ tumors and in brain adjacentto the V+ tumor (Af), whereas EBA-stained blood vessels are seen in brain immediately adjacent to the V—tumor margin and in the periphery of the V—tumor itself (AO.O und P,FVIIl-related antigen staining in V+ and V- tumors, respectively. The FVlll-positive vessels (arrows) branch and penetrate directly into the V+ tumor from the adjacent brain (panelO). This vascular branching and penetration pattern is not seen in the V—tumor (pane! P).

4189

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANGIOGENESIS BY V- EXPRESSION

Table 3 Blood volume, blood flow, vascular permeability, and glucose uptake of different VEGF-expressing tumors and normal tissues"

SK-MEL V + tumorsSK-MEL V- tumorsSK-MEL Ml tumorsFrontal cortexCorpus callosumBlood

volume ml/100 g(n =3)7.29

±0.96''1.26 ±0.16e

1.99 ±1.030.96 ±0.050.50 ±0.05dBlood

flow ml/min/100 g(n =6)160

±29h81 ±28r

105 ±22200 ±14

47 ±12Vascular

permeability fil/min/g(«=6)3.17

±Q.61h0.07 ±0.02C

0.10 ±0.050.04 ±0.020.08 ±0.03Glucose

uptake /¿mol/min 100 g(n =6)40.3

±17.429.3 ±17.535.1 ±10.143.6 ±4.622.1 ±2.9

a Values are mean ±SD.''Significantly different from control Nl tumors at P < 0.001 by ANOVA test.' Significantly different from V + tumors at P < 0.0001 by ANOVA test.

Blood volume value of corpus callosum was quoted from Nakagawa et al. (48).

lions. Mouse VEGF expression in V+ tumor cells is driven by theCMV promoter and is constitutive. Thus, V+ xenografts (that developed after i.e. injection of V+ cells) grew much more rapidly in ratbrain and had higher blood volume, perfusion, vascular permeability,and FDG uptake than Nl or V—xenografts. This is not surprising,given the level of mouse VEGF expression by V-t- cells and the fact

that mouse and rat VEGF proteins are nearly homologous with only atwo-amino-acid difference (43).

The observed differences in VEGF expression by the V + , V-, and N l

tumor cell lines in culture corresponded to measured differences ingrowth rate, blood volume, perfusion, vascular permeability, and Rglu incorresponding i.e. xenografts in animals. The morphological differencesbetween V+, V-, and Nl xenografts with respect to neovascularization

(FVIII staining) and integrity of BBB (EBA staining: Refs. 44, 45) werealso related to VEGF expression and to local vascular physiology (tumorblood volume, perfusion, vascular permeability). Several interesting morphological correlates with local vascular physiology were noted. There isgood correspondence between the absence of EBA-stained vessels in thebrain adjacent to V+ xenografts (approximately 0.4-0.5 mm from the

tumor margin) and a peritumoral zone of increased vascular permeability(Fig. 1, 7 and J). These observations suggest that mouse VEGF that isproduced by V+ xenografts penetrates adjacent brain tissue and can alterthe permeability of the adjacent brain-tissue vasculature. The corollaryobservation is the presence of EBA-stained vessels in the periphery ofV—xenografts and that these tumors are associated with very low levels

of vascular permeability.Blood volume of the V+ xenografts was 7-fold higher than normal

cortex and significantly higher than that of other experimental brainxenografts including RG2 (46) and C6 gliomas (47). Blood volume ofV+ xenografts was 6-fold higher than that of V- xenografts. This

difference in vascular volume was associated with marked vesselhyperplasia and branching in V+ xenografts; there were no similarvascular features in V—or N l xenografts. In comparison with the

reference Nl xenografts, the difference in vascular volume of V+xenografts was considerably greater than the difference in xenograftperfusion (3.7- and 1.5-fold, respectively). This indicates that an

increase in tumor vascularity (vascular volume) does not necessarilyresult in a similar increase in tumor perfusion. The vascular transittime can be calculated from the ratio of vascular volume/tissue perfusion (Vfc/F); for V + , V—,and Nl xenografts, the transit times were

2.7, 0.9, and 1.1 s, respectively. Compared with that of normal ratbrain (0.5 s; Ref. 48), the long transit time through V+ xenograftblood vessels suggests "sluggish" flow.

Another issue related to the increase in vascularity of V+ xenografts is the interpretation of the 32-fold difference in K, for[67Ga]-DTPA measured in V+ and Nl xenografts. Does this differ

ence in K, reflect an actual change in vascular permeability of V+xenografts or does it reflect a substantial difference in capillarysurface area? Because K, is relatively low compared with F, the valueof K, will approximate the permeability surface area product (PS) ofthe vasculature, and vascular permeability can be calculated if S is

known. When vascular density, vascular volume, and capillary surfacehave been measured in tumors and normal tissue, the ratio of capillarysurface area/vascular volume is usually lower in tumors. This isconsistent with morphometric data demonstrating larger diametervessels in brain tumors compared with normal tissue (49). Also, it isconsistent with findings reported elsewhere that neovasculature induced by VEGF is fenestrated (50).

Thus, the K, data reported here are likely to reflect a greaterdifference in vascular permeability than vascular surface area betweenV+ and Nl or V—xenografts. The continuous stimulation of ECs by

VEGF in the V+ xenografts may also result in vasodilatation (51) oftumor vessels along with the BBB disruption. These results provideadditional evidence (see Ref.40) that an increase in microvascularpermeability represents the first step in a cascade of pathophysiolog-

ical events involved in tumor neoangiogenesis and progression.SK-MEL-2 human melanoma cells stably transfected with mouse

antisense VEGF cDNA (V- cells) expressed no detectable level ofeither human or mouse VEGF protein in vitro. The growth of i.e. V—

xenografts was quite different from i.e. V+ xenografts. Eight daysafter i.e. inoculation, the proliferation of V- tumor cells was pre

dominantly localized to the tumor margin. A reliable estimate of theBrdUrd labeling index for 8-day V— xenografts was not possible

because of the periphery-to-center gradient in these small tumors. Allof the V—xenografts (n = 6) had a proportionately large area ofcentral necrosis. The BrdUrd labeling index data suggests that V—

tumor cell proliferation is limited to the margin of these small tumors,whereas central tumor regions undergo necrosis because of inadequatevascularization and shortage of nutrient and oxygen supply. Nevertheless, 3 weeks after i.e. inoculation of the V- cells, V- xenografts

reached a moderate size. The overall BrdUrd labeling index in theseV— xenografts was more homogeneous, but still quite low

(2.1 ±3.5%), even along the tumor margin. The retardation of i.e.V—xenograft growth was also evident from the results of a prelim

inary survival study. The median survival was more than 4 weeks afterunilateral i.e. inoculation of V—xenograft cells (n = 4), whereas the

median survival after unilateral i.e. inoculation of V+ xenograft cellswas only 13 days (n = 4).4

Ischemie brain tissue has been shown to be associated with highFDG accumulation (high glucose utilization). This observation hasbeen interpreted to reflect an uncoupling of cerebral blood flow andoxidative glucose metabolism, namely, a shift to anaerobic glucosemetabolism to meet the energy requirements of ischemic/hypoxicbrain tissue (52). A similar pattern of low blood flow and high FDGaccumulation has been reported in prenecrotic tumor regions (53).Recently, we reported a similar ischemic/hypoxic stimulation of glucose utilization (FDG accumulation) in focal regions of three differenti.e. rat tumors (24). In vitro studies in human cancer cell lines havealso shown that acute and chronic hypoxia results in an increase of

4 Unpublished data.

4190

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANCIOOENESIS BY V- EXPRESSION

glucose (FDG) utilization (54), which is coupled with up-regulation ofGLUT-1 transporter and VEGF expression. The up-regulation ofGLUT-1 transporter and VEGF expression during hypoxia appears to

be mediated mainly through mRNA stabilization rather than increasedmRNA transcription (55).

A paradoxical pattern of glucose metabolism was observed in V—i.e. xenografts. One would expect V—xenografts to be more hypoxic

and to have a higher level of anaerobic glucose metabolism (higherFDG accumulation) than corresponding Nl and V+ xenografts. Surprisingly, Rglu was low in V—xenografts compared to that in V+ orNl xenografts (P < 0.001; paired / test). Viable areas of V- xe

nografts with densely packed tumor cells are likely to be hypoxic, andthis was reflected by low blood flow and central necrosis. Moreperipheral regions of V- xenografts also had low blood flow, but they

were not necrotic. Surprisingly, Rglu in the viable-appearing, moreperipheral zone of V- xenografts was low. In contrast, a rim of

relatively high glucose utilization was observed in adjacent brainparenchyma along the margin of the V- xenografts. As indicatedabove, this marginal zone of brain parenchyma adjacent to V-

xenografts contained microscopic islets of invading tumor cells. It islikely that the brain parenchyma between the islets of infiltrating V—

xenograft cells was hypoxic (as reflected by a substantially lowerblood flow in this marginal zone) and contributed to the measuredincrease in Rg]u. This observation suggests that V- cDNA not only

down-regulates VEGF expression but also suppresses up-regulation of

the glycolytic pathway in tumor tissue or may inhibit the transitionfrom aerobic to anaerobic glycolysis that occurs in hypoxia.

Our results are in agreement with the results of other investigators whotargeted VEGF for developing antiangiogenic therapeutic approaches. Asignificant inhibition of tumor growth was reported using V—and monoclonal antibody to VEGF (4-9,56). The vascular permeability of i.e. V—

xenografts was found to be similar to that of normal brain tissue in ourstudy. This result demonstrates that the induction of neovascularizationand vascular permeability in i.e. SK-MEL-2 tumors was effectively

blocked by inhibition of endogenous expression of VEGF with antisensecDNA in V- xenografts. In contrast, overexpression of VEGF results in

enhanced tumor growth (9, 56-58), and the vascular permeability of i.e.V+ xenografts was found to be 45-fold greater than V- xenografts and

79-fold greater than normal brain (Table 3). In a recent paper by Yoshiji

et al. (56), an important question relating to the role of VEGF at differentstages of tumor growth was raised. This study demonstrated that thesuppression of VEGF production had no effect on the growth of largetumors (late stage), although it resulted in a marked inhibition of the smalltumor growth (early stage). It was suggested that the production of bFGFand TGF-a or other as-yet-unidentified angiogenic factors in the larger

tumors could adequately compensate for the loss of VEGF and promoteangiogenesis and tumor growth (56). Also, a number of putative angiogenic factors including small molecules (e.g., prostaglandins; Ref. 59) aswell as many cytokines (e.g., TGF-a, TGF-/3, tumor necrosis factor a,

bFGF, KGF, PDGF, PAF) secreted by tumor cells have all been shownto up-regulate VEGF expression in other (nontumor) cells (60-69).

Under these conditions, parenchymal cells adjacent to or within themargin of the tumor could be stimulated to produce VEGF and in turninduce neoangiogenesis and promote tumor growth.

These relationships may also explain the slow progression of i.e.V— tumors observed in our studies, despite an almost completeinhibition of VEGF expression by V—tumor cells. A likely reason forthe progression of V—xenograft growth is the retained ability of V—

cells to invade surrounding brain tissue, which is probably caused bythe expression of invasion and molility factors and the ability of V—

cells to proliferate (at least initially) in the tumor margin. This wouldalso explain the pattern of tumor proliferation and growth in themarginal zone of small V—xenografts in contrast to the nonprolifer-

ating, ischemie central zone, which undergoes necrosis. In addition,large central necrosis would likely contribute growth-promoting cy

tokines, especially bFGF, which would have pleitropic effects on theperiphery of the tumor. Xenograft growth would continue as tumorcells fill the space between invading islets resulting in the formationof a more compact peripheral tumor layer as the deeper, less wellvascularized portions of the tumor eventually undergo necrosis. Thishypothetical sequence of tumor proliferation and growth is consistentwith the much larger area of central necrosis observed in V—xe

nografts as compared to Nl and V+ xenografts.In conclusion, i.e. V- xenografts derived from antisense-trans-

fected human SK-MEL-2 melanoma cells were found to have sub

stantially different growth, vascular physiology, and glucose metabolism in comparison with control and constitutive VEGF secretingV+ xenografts. These findings indicate that VEGF should be considered as a target for antiangiogenic therapy of malignant brain tumors.Effective V- gene therapy of malignant brain tumors may require the

development of new and more efficient vector systems for optimizinggene delivery and gene expression.

REFERENCES

1. Folkman. J. Anti-angiogenesis: new concept for therapy of solid tumors. Ann. Surg..175: 409-416, 1972.

2. Wang Y., and Becker, D. Antisense targeting of basic fibroblast growth factor andfibroblast growth factor receptor-1 in human melanomas blocks intratumoral angiogenesis and tumor growth. Nat. Med.. 3: 887-893. 1997.

3. Oho. Y. S., Lee J. E., Oh K. S.. Bae, D. G.. and Chae. C. B. Development ofantiangiogenin peptide using a phage-displayed peptide library. Cancer Res.. 57: 17,3733-3740, 1997.

4. Kim, K. J., Li, B., Winer. J.. Armanini, M., Gillett. N.. Phillips. H. S., and Ferrara, N.Inhibition of vascular endothelial growth factor-induced angiogenesis suppressestumor growth in vivo. Nature (Lond.), 362: 841-844. 1993.

5. Saleh, M. Stacker, S. A., and Wilks. A. F. Inhibition of growth of C6 glioma cells invivo by expression of antisense vascular endothelial growth factor sequence. CancerRes., 56: 393-401, 1996.

6. Asano. M., Yukita, A., Matsumoto, T., Kondo. S.. and Suzuki, H. Inhibition of tumorgrowth and metastasis by an ¡mmunoneutralizing monoclonal antibody to humanvascular endothelial growth factor/vascular permeability factor]-,]. Cancer Res.. 55;5296-5301, 1995.

7. Borgström, P.. Hillan, K. J., Sriramarao. P.. and Ferrara. N. Complete inhibition ofangiogenesis and growth of microtumors by anti-vascular cinloihdi.il growth factorneutralizing antibody: novel concepts of angiostatic therapy from intravital videomi-croscopy. Cancer Res.. 56.- 4032-4039. 1996.

8. Cheng. S. Y., Huang, H. J., Nagane, M., Ji. X. D., Wang, D., Shih, C. C., Arap, W.,Huang. C. M., and Cavenee. W. K. Suppression of glioblastoma angiogenicity andtumorigenicity by inhibition of endogenous expression of vascular endothelial growthfactor. Proc. Nati. Acad. Sci. USA, 93: 8502-8507, 1996.

9. Claffey, K. P., Brown, L. F.. del Õguila, L. F., Tognazzi, K., Yeo. K-T., Manseau.E. J., and Dvorak, H. F. Expression of vascular permeability factor/vascular endothelial growth factor by melanoma cells increases tumor growth, angiogenesis. andexperimental metastasis. Cancer Res., 56: 172-181, 1996.

10. O'Reilly, M. S., Boehm, T., Shing, Y., Fukai, N.. Vasios, G., Lane, W. S., Flynn, E.,

Birkhead, J. R., Olsen, B. R., and Folkman, J. Endostatin: an endogenous inhibitor ofangiogenesis and tumor growth. Cell, 88: 277-285, 1997.

11. O'Reilly. M. S. Angiostatin: an endogenous inhibitor of angiogenesis and of tumor

growth. Exper. Supp. (Basel), 79: 273-294, 1997.12. Tolsma, S. S., Volpert, O. V., Good, D. J., Frazier. W. A., Polverini, P. J., and Bouck,

N. Peptides from two separate domains of the matrix protein thrombospondin-1 haveanti-angiogenic activity. J. Cell Biol., /22: 497-511, 1993.

13. Senger, D. R., Connolly. D. T., Van de Water, L, Feder, J., and Dvorak, H. F.Purification and NH2-terminal amino acid sequence of guinea pig tumor-secretedvascular permeability factor. Cancer Res., 50: 1774-1778, 1990.

14. Seetharam, L., Gotoh. N.. Mam, Y., Neufeld. G.. Yamaguchi. S.. and Shibuya, M. Aunique signal transduction from FLT tyrosine kinase, a receptor for vascular endothelial growth factor VEGF. Oncogene, 10: 135-147, 1995.

15. Waltenberger, J.. Claesson-Welsh, L.. Siegbahn, A.. Shibuya, M.. and Heldin, C. H.

Different signal transduction properties of KDR and Fit I. two receptors for vascularendothelial growth factor. J. Biol. Chem., 269: 26988-26995, 1994.

16. Takahashi, T., and Shibuya, M. The 230 kDa mature form of KDR/Flk-1 (VEGFreceptor-2) activates the PLC-gamma pathway and partially induces mitotic signals inN1H3T3 fibroblasts. Oncogene. 14: 2079-2089, 1997.

17. Guo, D., Jia, Q., Song, H. Y., Warren, R. S., and Donner, D. B. Vascular endothelialcell growth factor promotes tyrosine phosphorylation of mediators of signal transduction that contain SH2 domains. Association with endothelial cell proliferation.J. Biol. Chem., 270.- 6729-6733, 1995.

18. Senger, D. R., Van de Water. L., Brown, L. F.. Nagy, J. A., Yeo, K. T., Yeo. T., Berse,B.. Jackman, R. W.. Dvorak. A. M.. and Dvorak. H. F. Vascular permeability factor(VPF, VEGF) in tumor biology. Cancer Metastasis Rev., 12: 303-324. 1993.

4191

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

SUPPRESSION OF BRAIN TUMOR ANGIOGENESIS BY V- EXPRESSION

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Samolo, K„Ikezaki, K„Ono, M., Shono. T., Kohno, K., Kuwano, M., and Fukui, M.Expression of vascular endothelial growth factor and its possible relation withneovascularization in human brain tumors. Cancer Res., 55: 1189-1193, 1995.Cheng, S. Y.. Huang, H. J., Nagane. M., Ji, X. D., Wang, D., Shih, C. C, Arap. W.,Huang. C. M., and Cavenee, W. K. Suppression of glioblastoma angiogenicity andtumorigenicity by inhibition of endogenous expression of vascular endothelial growthfactor. Proc. Nati. Acad. Sci. USA, 93: 8502-8507, 1996.

Harnacher, K., Coenen, H. H., and Stoklin, G. Efficient stereospecific synthesis ofno-carrie-added 2-[ 18F]-fluoro-deoxy-D-glucose using aminopolyether supported nu-cleophilic substitution. J. Nucí.Med., 27: 235-238, 1986.

Kothari, P., Ginos, J., Finn, R., Larson, S., Link. J., Krohn, K., and Garmestani, K.Cryptand [2.2.2] quantitation in the synthesis of 2-[ 18F]-fluoro-deoxy-D-glucose.

Presented at the Fourth International Workshop on Targetry and Target Chemistry.Villigen, Switzerland. September 9-12, 1991.Blasberg, R. G., Kobayashi, T., Patlak, C. S., Shinohara, M., Miyoaka. M., Rice,J. M., and Shapiro. W. R. Regional blood flow, capillary permeability, and glucoseutilization in two brain tumor models: preliminary observations and pharmacokineticimplications. Cancer Treat. Rep., /: 3-12, 1981.Uehara, H., Miyagawa, T., Tjuvajev, J., Joshi, R.. Beattie, B., Oku, T., Finn. R. andBlasberg, R. Imaging experimental brain tumors with 1-aminocyclopentane carbox-ylic acid and a-aminoisobutyric acid: comparison to fluorodeoxyglucose and dieth-

ylenetriaminepentaacetic acid in morphologically defined tumor regions. J. Cereb.Blood Flow Metab., 17: 1239-1253, 1997.

Sakurada, O., Kennedy, C., Jehle. J., Brown, J. D., Carbin, G. L.. and Sokoloff, L.Measurement of local cerebral blood flow with iodo [14C] antipyrine. Am. J. Physiol.,

234: 59-66, 1978.

Kety, S. S. The theory and applications of the exchange of inert gas at the lungs andtissue. Pharmacol. Rev., 3: 1-41, 1959.

Kety, S. S. Measurement of local flow by the exchange of an inert diffusiblesubstance. Methods Med. Res., 8: 228-236, 1960.

Gjedde, A. Calculation of cerebral glucose phosphorylation from brain uptake ofglucose analogs in vivo: a re-examination. Brain Res. Rev., 14: 237-274, 1982.

Blasberg, R. G.. Patlak, C. S., and Fenstermacher. J. D. The selection of experimentalconditions for the accurate determination of blood-brain transfer constants from single-time experiments: a theoretical analysis. J. Cereb. Blood Row Metab., 3: 215-225, 1983.Blasberg. R. G., Groothuis. D., and Molnar, P. Application of quantitative autora-diographic measurements in experimental brain tumor models. Semin. Neuroi., 1:203-224, 1981.

Blasberg, R. G., Shinohara, M., Shapiro, W. R., Patlak, C. S., and Fenstermacher, J. D.Apparent glucose utilizations in Walker 256 brain tumors. J. Neuro-oncol., 4: 5-16, 1986.

Sokoloff. L., Reivich, M., Kennedy. C., Des Rosiers, M. H., Patlak, C. S., Pettigrew,K. D., Sakurada. O.. and Shinohara, M. The [l4C]deoxyglucose method for the

measurement of local cerebral glucose utilization: theory, procedure, and normalvalues in the conscious and anesthetized albino rat. J Neurochem.. 28: 897-916, 1977.

Dvorak, H. F., Orenstein, N. S., Carvalho, A. C., Churchill. W. H., Dvorak, A. M.,Galli, S. J.. Feder, J., Bitzer, A. M., Rypysc, J., and Giovinco, P. Induction of afibrin-gel investment: an early event in line 10 hepatocarcinoma growth mediated bytumor-secreted products. J. Immunol., 122: 166-174. 1979.

Senger, D. R.. Galli. S. J.. Dvorak, A. M., Perruzzi, C. A., Harvey, V. S., and Dvorak,H. F. Tumor cells secrete a vascular permeability factor that promotes accumulationof ascites Huid. Science (Washington DC), 279: 983-985. 1983.Senger, D. R., Perruzzi, C. A., Feder, J., and Dvorak, H. F. A highly conservedvascular permeability factor secreted by a variety of human and rodent tumor cellslines. Cancer Res., 46: 5629-5632, 1986.

Brown. L. F., Berse, B., Jackman, R. W.. Tognazzi. K.. Manseau, E. J., Senger, D. R.,and Dvorak, H. F. Expression of vascular permeability factor (vascular endothelialgrowth factor) and its receptors in adenocarcinomas of the gastrointestinal tract.Cancer Res., 53: 4727-4735, 1993.Brown, L. F., Berse. B.. Jackman, R. W.. Tognazzi, K., Manseau, E. J.. Dvorak. H. F..and Senger. D. R. Increased expression of vascular permeability factor (vascularendothelial growth factor)) and its receptors in kidney and bladder carcinomas. Am. J.Pathol., 143: 1255-1262. 1993.

Brown. L. F.. Berse, B., Jackman. R. W., Tognazzi. K., Guidi, A.. Dvorak, H., Senger.D., Connolly, J., and Schnitt, S. Expression expression of vascular permeability factor(vascular endothelial growth factor) and its receptors in breast cancer. Hum. Pathol.,26: 86-91. 1995.

Dvorak, H. F.. Sioussat, T. M., Brown, L. F.. Nagy, J. A., Sotrel. A., Manseau, E.,Van De Water, L.. and Senger, D. R. Distribution of vascular permeability factor(vascular endothelial growth factor) in tumors: concentration in tumor blood vessels.J. Exp. Med., 174: 1275-1278, 1991.

Dvorak, H., Brown, L. F., Detmar, M., and Dvorak. A. Vascular permeabilityfactor/vascular endothelial growth factor, microvascular hyperpermeability, and an-giogenesis. Am. J. Pathol., 146: 1029-1039. 1995.

Plate, K. H.. Breier, G., Weich, H. A., and Risau, W. Vascular endothelial growthfactor is a potential tumor angiogenesis factor in human gliomas in vivo. Nature(Lond.), 359: 845-848. 1992.

Berkman, R. A., Merrill, M. J., Reinhold, W. C.. Monacci. W. T., Saxena, A., Clark,W. C., Robertson, J. T., Ali, I. U., and Oldfield, E. H. Expression of the vascularpermeability factor/vascular endothelial growth factor gene in central nervous systemneoplasms. J. Clin. Invest., 91: 153-159, 1993.Claffey, K. P., Wilkison, W. O., and Spiegelman, B. M. Vascular endothelial growthfactor. Regulation by cell differentiation and activated second messenger pathways.J. Biol. Chem., 267: 16317-16322, 1992.Stemberger. N. H., Sternberger, L. A., Kies, M. W., and Shear, C. R. Cell surfaceendothelial proteins altered in experimental allergic encephalomyelitis. J. Neuroim-munol.. 21: 241-248, 1989.

45. Sternberger, N. H., and Sternberger, L. A. Blood-brain barrier protein recognized bymonoclonal antibody. Proc. Nati. Acad. Sci. USA, 84: 8169-8173, 1987.

46. Nakagawa H., Groothuis D. R., Owens E. S., Patlak C. S.. Pettigrew K. D., andBlasberg R. G. Dexamethasone effects on vascular volume and tissue hematocrit inexperimental RG-2 gliomas and adjacent brain. J. Neuro-oncol., 6: 157-168, 1988.

47. Nomura, T.. Inamura, T.. and Black, K. L. Intracarotid infusion of bradykininselectively increases blood-tumor permeability in 9L and C6 brain tumors. Brain Res.,659: 62-66, 1994.

48. Nakagawa, H., Lin, S-Z., Bereczki, D., Gesztelyi. G., Otsuka, T., Wei, L., Hans, F-J.,Acuff. V. R., Chen, J-L.. Pettigrew, K. D., Patlak, C. S., Blasberg, R. G., andFenstermacher, J. D. Blood volumes, hematocrits, and transit-times in parenchymal

microvascular systems of the rat brain. In: D. Le Bihan (ed.). Diffusion and PerfusionMagnetic Resonance Imaging, pp 193-200. New York: Raven Press, Ltd., 1995.

49. Wesseling, P., van der Laak. J. A. W. M., de Leeuw, H.. Ruiter, D. J., and Burger.P. C. Quantitative immunohistological analysis of the microvasculature in untreatedhuman glioblastoma multiforme. Computer assisted analysis of whole-tumor sections.J. Neurosurg., 81: 902-909. 1994.

50. Roberts, W. G., and Palade, G. E. Neovasculature induced by vascular endothelialgrowth factor is fenestrated. Cancer Res., 57: 765-772, 1997.

51. Sellke. F. W., Wang, S. Y., Stamler, A., Lopez, J. J., Li. J., Li, J.. and Simons, M.Enhanced microvascular relaxations to VEGF and bFGF in chronically ischemieporcine myocardium. Am. J. Physiol.. 27/: 713-720, 1996.

52. Ginsberg. M. D., Reivich, M., Giandomenico. A., and Greenberg, J. H. Local glucoseutilization in acute focal cerebral ischemia: local dysmetabolism and diaschisis.Neurology, 27: 1042-1048. 1997.

53. Kubota. R.. Kubota, K., Yamada, S., Tada. M., Ido. T.. and Tamahashi, N. Micro-autoradio-graphic study for the differentiation of intratumoral macrophages, granulation tissues and cancer cells by the dynamics of fluorine-18-fluorodeoxyglucoseuptake. J. Nucí.Med., 35: 104-112, 1994.

54. Clavo, A. C., and Wahl, R. L. Effects of hypoxia on the uptake of tritiated thymidine,L-leucine, L-methionine and FDG in cultured cancer cells. J. Nucí.Med., 37: 502-506,

1996.55. Stein, I.. Neeman, M.. Shweiki, D., Itin, A., and Keshet, E. Stabilization of vascular

endothelial growth factor mRN A by hypoxia and hypoglycemia and coregulation withother ischemia-induced genes. Mol. Cell. Biol., 15: 5363-5368. 1995.

56. Yoshiji, H.. Harris, S. R., and Thorgeirsson, U. P. Vascular endothelial growth factoris essential for initial but not continued in vivo growth of human breast carcinomacells. Cancer Res., 57: 3924-3928. 1997.

57. Zhang. H. T.. Craft, P.. Scott, P. A., Ziehe, M., Weich, H. A., Harris, A. L., andBicknell, R. Enhancement of tumor growth and vascular density by transfection ofvascular endothelial cell growth factor into MCF-7 human breast carcinoma cells.J. Nati. Cancer Inst., 87: 213-219, 1995.

58. Pötgens,A. J.. van Allena, M. C., Lubsen, N. H., Ruiter, D. J., and de Waal, R. M.Analysis of the tumor vasculature and metastatic behavior of xenografls of humanmelanoma cell lines transfected with vascular permeability factor. Am. J. Pathol.,148: 1203-1217, 1996.

59. Harada, S., Nagy, J. A., Sullivan, K. A., Thomas, K. A., Endo, N., Rodan, G. A., andRodan, S. B. Induction of vascular endothelial growth factor expression by prostag-landin E2 and El in osteoblasts. J. Clin. Invest., 93: 2490-2496, 1994.

60. Brogi, E., Wu, T., Namiki, A., and Isner. J. M. Indirect angiogenic cytokines upregulateVEGF and bFGF gene expression in vascular smooth muscle cells, whereas hypoxiaupregulates VEGF expression only. Circulation, 90: 649-652. 1994.

61. Dolecki, G. J., and Connolly, D. T. Effects of variety of cytokines and inducing agentson vascular permeability factor mRNA levels in U937 cells. Biochem. Res. Commun., 180: 572-578, 1991.

62. Detmar, M., Brown, L., Claffey, K., Yeo, K-T., Kocher, O., Jackman, R., Berse, B.,

and Dvorak, H. Overexpression of vascular permeability factor and its receptors inpsoriasis. J. Exp. Med., 180: 1141-1146, 1994.

63. Finkenzeller, G.. Marme, D., Weich, H. A., and Hug. H. Platelet-derived growthfactor-induced transcription of the vascular endothelial growth factor gene is mediated by protein kinase C. Cancer Res.. 52: 4821-4823. 1992.

64. Frank, S., Hubner, G., Breier. G.. Longaker, M. T., Greenhalgh, D. G., and Werner,S. Regulation of vascular endothelial growth factor expression in cultured keratino-

cytes. Implications for normal and impaired wound healing. J. Biol. Chem., 270:12607-12613. 1995.

65. Li, J., Perrella, M. A., Tsai, J. C.. Yet, S. F., Hsieh, C. M., Yoshizumi, M. Patterson,C. Endege, W. O.. Zhou, F., and Lee. M. E. Induction of vascular endothelial growthfactor fene expression by interleukin-1 beta in rat aortic smooth muscle cells. J. Biol.Chem., 270: 308-312. 1995.

66. Petrovaara, L., Kaipainen, A., Mustonen. T., Orpana. A., Ferrara, N.. Saksela, O., andAlitalo, K. Vascular endothelial growth factor is induced in response to transforminggrowth factor-b in fibroblastic and epithelial cells. J. Biol. Chem., 269: 6271-6274, 1994.

67. Starvi, G. T.. Hong, Y., Zachary, I. C., Breier, G., Baskerville, P. A., Yla. H. S.,Risau, W.. Martin. J. F., and Erusalimsky, J. D. Hypoxia and platelet-derived growthfactor-BB synergistically upregulate théexpression of vascular endothelial growthfactor in vascular smooth muscle cells. FEBS Lett., 358: 311-315, 1995.

68. Starvi, G. T., Zachary. I. C., Baskerville, P. A., Martin, J. F., and Erusalimsky, J. D.Basic fibroblast growth factor upregulates the expression of vascular endothelialgrowth factor in vascular smooth muscle cells. Synergistic interaction with hypoxia.Circulation. 92: 11-14, 1995.

69. Williams, B.. Quinn, B. A., and Gallacher. B. Serum and platelet-derived growthfactor-induced expression of vascular permeability factor mRNA by human vascularsmooth muscle cells in vitro. Clin. Sci. (Colch), 88: 141-147. 1995.

4192

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1998;58:4185-4192. Cancer Res   Takamitsu Oku, Juri G. Tjuvajev, Tadashi Miyagawa, et al.   Melanoma Intracerebral XenograftsFlow, Fluorodeoxyglucose Uptake, and Proliferation of Human Angiogenesis, Vascular Permeability, Blood Volume, BloodEndothelial Growth Factor Gene Expression: Effects on Tumor Growth Modulation by Sense and Antisense Vascular

  Updated version

  http://cancerres.aacrjournals.org/content/58/18/4185

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/18/4185To request permission to re-use all or part of this article, use this link

on June 8, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from