Increased gene expression for VEGF and the VEGF receptors KDR/Flk and Flt in lungs exposed to acute or to chronic hypoxia Modulation of gene expression by nitric oxide

(1)

Increased gene expression for VEGF and the

VEGF receptors KDR/Flk and Flt in lungs

exposed to acute or to chronic hypoxia.

Modulation of gene expression by nitric oxide.

R M Tuder, … , B E Flook, N F Voelkel

J Clin Invest.

1995;

95(4)

:1798-1807.

https://doi.org/10.1172/JCI117858

.

Endothelial cells constitute an essential integrator of factors that effect blood vessel

remodeling induced by chronic hypoxia. We hypothesized that vascular endothelial growth

factor (VEGF) may participate in the lung response to acute and to chronic hypoxia. We

found that ex vivo perfusion of isolated lungs under hypoxic conditions (when compared

with normoxia) caused an increase in lung tissue mRNA of VEGF and of the VEGF

receptors KDR/Flk and Flt. Chronic hypobaric hypoxia also increased lung tissue mRNA

levels of VEGF, KDR/Flk, and Flt and the amount of VEGF protein. In situ hybridization

studies demonstrated increased VEGF and KDR/flk hybridization signals in lungs from

chronically hypoxic rats. Since endotoxin treatment of rats decreased lung VEGF mRNA,

we postulated that nitric oxide (NO) or an NO-related metabolite might be involved in lung

VEGF gene expression. Indeed, sodium nitroprusside, a NO donor, decreased and L-NAME

(N-nitro-L-arginine methyl ester), an inhibitor of NO-synthesis, increased both VEGF and

VEGF receptor transcripts. We conclude that VEGF in the isolated perfused lung acts as an

early gene in response to hypoxia and that lung VEGF and VEGF receptor mRNA levels

are influenced by hypoxia and NO-dependent mechanisms.

Research Article

(2)

Increased Gene Expression for VEGF

and the

VEGF

Receptors

KDRIFIk

and

Fit

in

Lungs Exposed

to

Acute

or

to

Chronic

Hypoxia

Modulation of Gene Expression by

Nitric

Oxide

Rubin M.Tuder, Barbara E. Flook, andNorbert F. Voelkel

Divisionof Pulmonary Sciencesand Critical Care Medicine, PulmonaryHypertension Center and Department ofPathology, University

of Colorado, HealthSciences Center, Denver, Colorado 80262

Abstract

Endothelial cells constitute an essential integrator of factors that affect blood vessel remodeling induced by chronic hyp-oxia. Wehypothesized that vascular endothelial growth fac-tor (VEGF) may participate in the lung response to acute and to chronic hypoxia. Wefoundthat ex vivo perfusion of isolated lungs under hypoxic conditions (when compared withnormoxia) causedanincrease in lung tissue mRNA of VEGF and of the VEGF receptors KDR/Flk and Flt. Chronic hypobaric hypoxia also increased lung tissue mRNA levelsofVEGF, KDR/Flk, and Fltand the amount of VEGFprotein. In situ hybridization studies demonstrated increased VEGF and KDR/flk hybridization signals in lungs from chronically hypoxic rats. Since endotoxin treatment of ratsdecreased lungVEGFmRNA, we postulated that nitric oxide(NO)oranNO-related metabolite might be involved in lung VEGF gene expression. Indeed, sodium

nitroprus-side,aNO donor,decreasedand L-NAME

(N-nitro-L-argi-nine methyl ester), aninhibitor ofNO-synthesis,increased both VEGF and VEGF receptor transcripts. We conclude that VEGF in the isolated perfused lung acts as an early gene in responsetohypoxia and that lung VEGF and VEGF receptor mRNA levelsareinfluenced byhypoxia and

NO-dependent mechanisms. (J. Clin. Invest. 1995.

95:1798-1807.) Key words: vascular endothelial growth factor (VEGF) * VEGF receptors * acuteand chronic

hypoxia.

pulmonary hypertension* nitric oxide

Introduction

Chronichypoxiarepresentsoneof the stimuli thataccountsfor

pulmonary vascularremodeling in chronic lung diseases. This remodeling process involves hypertrophy of the precapillary

vascular smooth musclecells,proliferation of endothelial cells andpericytes,andenhanced vascular

permeability

(1-3).Since hypoxia persedoes not causehyperplasiaof culturedpulmonary

vascular smooth muscle cells, it is reasonable to assume that,

invivo, hypoxia and vascular cellgrowthfactorsact

synergis-Address correspondence to Norbert F. Voelkel, M.D., University of

Colorado Health Sciences Center, Box C 272, 4200 E. Ninth Ave.,

Denver,CO 80262. Phone: 303-270-4211;FAX: 303-270-5632.

Receivedforpublication 12 August 1994and in revisedform 13

October 1994.

tically on growth and differentiation of pulmonary vascular cells (4).

Endothelial cells have a central role in integrating signals that may affect pulmonary vessel structure and function. In-creased shear stress, as seen during hypoxic vasoconstriction, results in endothelial cell expression of vascular cell growth factors, such as platelet derived growth factor (PDGF) or

trans-forming growth factor

_(TGF-P3)

orproinflammatory molecules like ICAM-1 (5).

Vascular endothelial cell growth factor

(VEGF),'

apeptide

mitogenspecific for endothelial cells (6-14), is abundantly ex-pressed in lung tissue, including pneumocytes and activated alveolarmacrophages (8, 10). Although the function of VEGF in normal tissue remains unknown, brain tumor cells, cancer cells, muscle myoblasts and fibroblasts upregulate VEGF ex-pression under conditions of hypoxia (15). Indeed, VEGF repre-sentsoneof the mainendothelial cell growth factors produced by malignant cells, and may be involved in endothelial cell proliferation observed in pulmonary vascular remodeling.

VEGF bindstotwohighly homologous tyrosine kinase re-ceptorsexpressedexclusively by endothelial cells: thetyrosine

kinase receptorKDR (or the murine homologue fetal liver ki-nase-flk) (16-18), and the fins-like tyrosine kinase receptor

Flt (19, 20). Although VEGF binding to Flt triggers receptor autophosphorylation and

Ca2"

influx, Flt activation does not sufficeto causeendothelial cellproliferationinboth in vitroas well as in vivo models of tumor-mediated angiogenesis (21). On the otherhand,transfection withKDR/Flkcausescell

prolif-eration, chemotaxis and actin rearrangement upon VEGF expo-sure of porcine endothelial cells (21). Furthermore, mutant forms of Flkcan abrogate VEGF induced signal transduction in vitro andcanalsoreduceblood vesselproliferationin in vivo models of brain cancers (22). These studies suggest that the biologic function of VEGF in disease states,including hypoxic pulmonary hypertension, may rely on increased amounts of

VEGFproteinaswellasincreased VEGF receptor number and affinity.

In the present study, wepursue thehypothesis that VEGF

mRNA

andprotein andtheVEGF receptors Flt andKDR/Flk

mRNAare increased in lungssubjected to acute or tochronic

hypoxia.Tocompare the effects ofacute

hypoxia

withanagent thatalso alterspulmonary vascular tone,wealso evaluated the acuteeffect of endotoxinonVEGF geneexpression.Endotoxin inhibits acute hypoxic vasoconstriction and may alter

growth

factor expression in the

hypertensive lungs.

Since endotoxin causeslargeincreasesinnitric oxide(NO)

production,

wealso

1. Abbreviations used in thispaper:GAPDH,glyceraldehyde phosphate

dehydrogenase; NO, nitric oxide; SNP, sodium-nitroprusside; VEGF,

vascularendothelialgrowthfactor.

J. Clin. Invest.

0021-9738/95/04/1798/10 $2.00

(3)

tested the effect of the NO-donorsodium-nitroprusside (SNP)

onlungtissue VEGF and VEGFreceptor gene expression. Our

data in the aggregateindicate that both VEGF and the VEGF

receptors in thelungareregulatedbyacuteand chronichypoxia, and that NO modifies the VEGFand VEGFreceptor gene

ex-pression inthelung during hypoxic and normoxic conditions.

Methods

Chronic exposureto

hypoxia

Adult male Sprague-Dawley rats were purchased froma commercial

vendor andkeptfor 7d in the Animal CareFacility of theUniversity

of Colorado Health Sciences Center.Thereafter,therats wereweighed

andrandomized to acontrol (Denveraltitude, 5600 ft) andachronic

hypoxia (16,000 ft simulatedhigh altitude) group. The rats weighed

300-330 grams and weredivided into different groups: control,and

variousperiodsofhypoxic exposure rangingfrom 1 to32 d. Control

(n=3)andhypoxia-exposedrats(n= 15)hadfreeaccess tofood and

waterand weresubjectedtothesamedayandnight light cycle.After

conclusion of the predetermined hypoxiaexposure, the animals were

removed from thehypobaric chamber andanesthetized with 100mg/

kg pentobarbital (Nembutal). Then a midline thoracotomy was

per-formed,and thelungswerequicklyisolated and excised. The left main

bronchuswas crossclampedand therightlungwasimmediatelyinstilled

with melted agarose andsuspendedinaphosphate bufferand storedat

40Covernight (23);theselungswerethenfrozen,sectioned andprepared

for in situhybridization.The leftlungwasflashfrozeninliquid nitrogen

and storedat-720Cuntil usedforRNAextractionorthe VEGF ELISA

assay.

Acute exposure to

hypoxia

Lungswereisolated from adultratsandperfusedat constantflow(0.03 ml/gBW/min)withacell-freephysiologicalsalt solution and the

perfu-sion pressure was measuredcontinually aspreviouslydescribed (24).

Thelungs werekept moist inaplexiglass chamberat370C; someof thelungs wereventilated with aroom air gas mixture(21% 02, and

5% CO2and 74% N2), otherlungswere ventilated witha "hypoxic"

gas mixture (0% 02, 5% CO2balance N2). Thismode of ventilation

results inalungeffluentP02of 30-35mmHg.After 2 hof perfusion

andventilation the lungs were rapidly frozenin liquid nitrogenand

storedat-720Cuntilthe time of RNA extraction.

Endotoxin treatment

12ratsreceivedi.p. injectionof S. enteritidis endotoxin(Sigma

Chemi-calCo,St. Louis, MO; 2mg/kg).At 2, 4,6,or 12 h(n = _{3 per time}

point) after endotoxin injection, the animals were sacrificed and the

lungs wereremovedafter thoracotomyand flashfrozenuntil the time

of RNA extraction. Threerats wereinjectedwith saline and served as

controls.

Nitric oxide studies

Isolated lungs wereperfused, under normoxic orhypoxic conditions,

witha salt solution that contained either sodium nitroprusside at 10'

M(n=₄₎orN-nitro-L-arginine methylesterhydrochloride (L-NAME)

at 10-4M(n =_{6) (Research}_{Biochemicals International, Natick, MA).}

Thesestudiesweredesignedtotestwhether the increaseorthedecrease

oflung nitric oxide (NO) production alters the expression ofVEGF,

KDR/Flk,and Fltbythelungtissue.

Specific

methods

[3Hfthymidine

labelingandautoradiography. Threeratshad been

ex-posedtochronichypoxiafor28d, (threelow altitude rat lungs served ascontrols);1

_ICi

[3H]thymidine(NewEngland Nuclear, Boston,MA) wasinjected i.p. 1 h after thisinjectiontheanimals received

pentobarbi-tal sodium_{(100 mg/kg) i.p.,}thelungswereexcised,fixed informalin,

cutintosections,covered withphotoemulsionandexposedfor 4 wk (2). The

_{autoradiograms}

ofthe

lung

sliceswere_inspectedand_{photographed.}

Northern blot studies. Total RNA wasextracted fromrat lungas

describedby Chomcynsky and Sacchi (25), fractionated by

formalde-hyde gel electrophoresis, and transferred onto Nylon membranes

(Schleicher andSchell, Keene, NH). 20jigof total RNAwereloaded

per lane. Theprobeswereprepared by random-priming labeling using

[32P]dCTP, 3000 Ci/nMol, 10 pCi, il (NewEngland Nuclear). The

mouse cDNAprobe for/i-actinand the mouseglyceraldehydephosphate

dehydrogenase (GAPDH) cDNA probes were a gift from Dr. J. Fisher (University of Colorado Health Sciences Center). The 28 S ribosomal

cDNA was used for correction of RNA loading. The VEGF human

cDNAprobe (687 bp) was provided by Irene Smith (Genentech, South

SanFrancisco, CA). The human Flt cDNA was a 458-bp sac- I fragment

subcloned from the 3-9 cDNA (gift from Dr. Masabumi Shibuya) (19)

into the 4Zplasmid(Promega, Milwaukee, WI). The KDR cDNAwas

provided by Dr. Bruce Terman (16, 17). Development of blots

hybrid-ized with the VEGF and VEGF receptor probeswereperformed for - 8

hand3 d,respectivelyandsubsequentlyquantified by phosphorimaging andscanningdensitometry.

In situhybridization, VEGF mRNA, KDR mRNA, Fit mRNA. Slides

withsections of rat lungs were vacuum-dried for 2 h for attachment of

sections to Superfrost slides (Fisher Scientific Co., Pittsburgh, PA),

followed by treatment with 0.2 N HCl for 20min, washing in

depc-PBS, proteinase K digestion at 10

_Ag/ml

at370Cfor 20 min, fixation

in 4%paraformaldehydefor 5min, acetylationwithethanolamine with

acetic-anhydrade (1:400 ratio) for 5 min. Blocking was performed in a

premade solution containing 50% formamide (Hybridization cocktail,

Ameresco, Solon, OH) for 2 hat roomtemperature.

VEGF transcripts weredetected with the subclone rat-VEGF 2 (316

basepairs)cloned intoaBluescriptKS+ vector(Stratagene,La Jolla, CA), kindly provided by Drs. G. Breier and W. Risau (11). Sense

andanti-sense cRNA probes for VEGF and KDR were labeled with

digoxigenin-dUTP with the Genius IV kit as recommendedbythe

manu-facturer (Boehringer Mannheim, Indianapolis, IN). The probes were

quantifiedby dot blot analysisusingknown concentrations of digoxi-genin-labeled plasmidDNAasstandard. Theefficiencyoflabelingwas

furtherconfirmed by agarosegelelectrophoresis. Probes were used at

aconcentration of 25ng/section.

Hybridization was performed overnight at 42°C. After extensive

sequential washingsin2x, lx,and0.5xSSC, the unhybridized probe

wasdigested with ribonuclease (Promega, Madison, WI) in 0.5X SCC.

Thehybridization product was detected after incubation with a sheep

anti-digoxigenin antibody (1/2000dilution; Boehringer Mannheim,

Indi-anapolis, IN) for 90 min followed by development in 4-tetrazolium

chloride 4.5

_/sg/ml

andx-phosphate3.5

Al

for4-12 h.

ELISAfor VEGF. Ratlungs (- 1 g of tissue/lung) were

homoge-nized inlysis buffer(0.1 Msodium carbonate pH 9.6,0.1% NP-40, 1

AM

leupeptin, 1 mM PMSF, and 10

Al/ml

ethylenediaminetetraacetic

acid fromasaturated stock), followedby sonication (Fisher Sonic

Des-membrator; Fischer Scientific). The homogenate was then cleared by

centrifugationat750 g,at4°C for 6 min. The supernatant was stored at-20°Cuntil assayed for VEGF protein levels. The experimental

sam-plesand recombinantVEGFI65standards (R & D Systems, Inc.,

Minne-apolis, MN)weredilutedin Carbonatebufferand added in triplicate to Immulon IV 96-well plates (Dinatech, Chantilly, VA). The samples

wereincubated at 37°C for 8 h. Standards consisted of 2-fold dilutions

ofrVEGF165, starting at 25 ng/well to 390 pg/well. The samples were

removedand the wells were briefly washed in washing buffer (PBS pH

7.4,0.05% Tween20). The wells were blocked withblocking buffer

(washing buffer with1%BSA)at 37°C for 4 h. After two brief washes,

10IL of rabbit anti-VEGF antibody (kindly provided by Dr. Napoleone

Ferrara,Genentech, CA) was added to each well for I h at 37°C. After

washing twice, goat anti-rabbit peroxidase (Tago Inc., Burlingame, CA)

wasadded for 1 h. O-phenylenediamine dihydrochloride was used as

substrate and theopticaldensitywasdeterminedat490 nm on a Syva

Microtrak spectrophotometer. The program Statview 512 was used to

plot optical densityaverages. The concentrationof VEGF in the

experi-mentalsampleswascalculatedwith aformuladerived by linear

(4)

con-Figure 1. Rat lung[3H]thymidine

autoradiogram.Hypertrophied small pulmonary artery from a rat

exposed to hypobarichypoxiafor

4wk. Two endothelial cells

dem-onstrate[3H]thymidine

incorpora-tion(arrows).

centrations of VEGF. The optical density curve for VEGF was linear

in concentrations ranging between 390 and 6.25pg/SO ILI (correlation

coefficient =.992, see Fig. 4 B).

Results

In vivostudies

CHRONIC HYPOXIA

[3H]Thymidine autoradiography. To determine which of the vascular cell types (endothelium, smooth muscle cells or fibro-blasts) undergo proliferation during hypoxia-induced lung vas-cularremodeling, sections ofratlung from animals exposed for 28 d to hypobaric pressure/hypoxia were inspected for [3H]thymidine labeling and autoradiography. The remodeled, muscularized arteries demonstrated multiple endothelial cells with 3H-incorporation consistent with endothelial cell prolifera-tion (Fig. 1), whilecontrol lung vesselswerecharacterized by rare cells with[3H]thymidineuptake. These results agreed with our earlier studies where differences of 3H-incorporation by pulmonary vascular cells under hypoxic conditions had been moreprecisely quantified (2).

VEGFmRNAand VEGF protein. VEGF mRNAwas abun-dantly present in control lungs, as described in prior studies (10). At28-32 days of hypoxic exposure, there wasatwo- to threefold increaseinlung VEGFmRNA overcontrols.(Fig. 2). We recognized that the levels of transcripts for the so-called housekeeping genes

,(-actin

and GAPDH varied during hypoxia and thuscould notbe usedfor densitometric quantification of mRNAfromchronically hypoxic lungs. This variability of the GAPDHgene expressioninconditions ofhypoxia has recently beenrecognized by others (26, 27). Since theribosomalRNA (rRNA) proved to be unchanged by hypoxia or hypobaric expo-sure,weelectedto use 28 S rRNAas areference. (Fig. 2).

Comparisonof normoxic andhypoxiclung tissueprocessed

by in situ hybridization technique, using the VEGF antisense probe labeled with digoxigenin, showedanincrease inthe

hy-bridizationsignalinalveolar,intraalveolar and

bronchio-alveo-lar cells when compared with control lungs. The signal however wasmostly concentrated in epithelial alveolar cells in both the experimental and the control group of animals. (Fig. 3). VEGF protein measured in lung homogenates was likewise increased fromday 3 uptoday 32 of hypoxic exposure when compared withlungs of control rats (Fig. 4).

VEGF receptor,KDRIFlk,and Flt mRNA. The VEGF func-tion as amitogen and as a permeability factor for endothelial cells relies on its interaction with tyrosine-kinase receptors. Northern blot technique and in situ hybridization were applied to assess changes induced by chronic hypoxia in lung tissue expression of VEGF receptor transcripts. Both KDR/Flk and Flt mRNA increased with prolonged (7-32 d) exposure to hypo-barichypoxia (Fig. 3). The increase of VEGF receptor expres-sionby day7ofhypoxia coincided with the first clear histologi-cal evidence of vascular remodelingasdescribed inprior studies (28). It was apparent that, by in situ hybridization, the KDR/ Flkand Flt gene expression wasconcentrated in the pre-capil-lary vessels and in intraseptal cells throughout the period of the chronichypoxia (Fig.5). Alveolarmacrophages,which didnot stain in sections from control lungs, exhibitedaweak

hybridiza-tion signal for KDR/Flk transcriptsinspecimens from chroni-cally hypoxic lungs.

Effect of endotoxinon VEGFtranscripts. To put the effect of acute hypoxia in perspective, we examined the effect of a singleintraperitonealendotoxininjectiononlungVEGF mRNA andfoundatime-dependent decrease in the abundance of VEGF transcripts in lungs fromendotoxin-injectedrats(Fig. 6).

Ex vivostudies

Acute hypoxia. To examine whetheracutehypoxiaofthe iso-latedlungpreparation (underconstantflow and pressure condi-tions) affects VEGF and VEGF receptor gene expression, we

(5)

VEGFmRNA

-'

FlimRNA

IQ=-It-actin - 2.4kb

GALPDH _404. hI - 1.4kb

M~o-RNA\ - 4.4kh

--

A

3.7kb

-6.8cb

-7 5 kb

-C

Figure2. (A) Ratlung/3-Actinand GADPH,but not a 28 S ribosomal gene expression increases during chronic hypoxia. (B and C) Northern blot forVEGF, KDR/Flk,Flt mRNA. The bargraphsrepresentscanning densitometry ofthehybridization productsfor the VEGF,KDR/Flk, andFPt signals(B).28Sribosomal RNAwasused tocontrol forRNAloading. Comparedaresamples from control(normoxic)andchronicallyhypoxic

animals.

KDR/Flk and Flt hybridization signals when compared with perfused, room-air ventilated lungs (Fig. 7). These lungs were not subject to increased vascular shear stress since, in lungs

perfused with a salt solution, hypoxic vasoconstriction is not maintained.

Effect ofNOonVEGFTranscripts. Since endotoxin induces the synthesis of NO (29), we wondered whether the effect of endotoxinonlung VEGF gene expression could be relatedto NO. Isolatedlungswereperfusedandventilated under normoxic orhypoxic conditions, and SNPorL-NAMEwasaddedtothe

perfusate. SNP(104 M) decreased VEGF and VEGF receptor mRNAlevels in the isolatedlungs (Fig. 7), whereas L-NAME, anitric oxide synthase inhibitor, increased VEGFmRNA ex-pression(Fig. 8).

Discussion

It has been appreciated for many years that chronic hypoxia leadstopulmonaryhypertensionandremodeling ofpulmonary

arteries (1-3, 30, 31). Since the muscularization of the pulmo-naryprecapillaryvessels isaprominent feature in animal mod-els ofhypoxic pulmonary hypertension, traditionally hypoxic vasoconstriction and "work hypertrophy" of the constricted vascular smooth muscle cells have been used as concepts to

explain the mechanism of vascularremodeling.However, evi-dence for vascular cell proliferation (as in Fig. 1) has been presented by many investigators (1, 30, 31).

Recent data supports the concept that

hypoxia-unaccom-panied byshearstress -canalter theexpressionof genes

encod-ingvascular cellgrowth andangiogenesisfactors (32).For ex-ample, activated protein kinase C (PKC) and activated PKC acting synergistically with added insulin-like growth factor stimulateproliferationof cultured pulmonary smooth cells (33).

Clearly,hypoxiaincreases VEGF mRNA in culturedtumorcells (15), and angiogenesis associated with necrotic hypoxic foci in tumors and with woundhealingmay share certain aspects with thechronic hypoxia-inducedlung vessel remodeling(14). We

recently reported increased VEGFtranscripts in thelungfrom

patients with primary (unexplained) pulmonary hypertension

and localized both VEGF receptortranscripts and VEGFprotein

in the characteristic proliferatedblood channels present in the

plexiformlesions (34).

The major findings presented in this report are that the VEGF gene and also the genes encoding the VEGF receptors

KDRIFlkand Fltwereincreasingly expressedduringthe devel-opment ofpulmonary hypertension in rat lungs in the course of prolonged exposure to hypoxia. An important additional

findingofourstudyisthatVEGFmRNAinthelungsincreased I

Jj.

_B

.4,

r- - -, t- ;:4

,;: ;, ;I. ;;,-. >. E = = R m

(6)

""I-0'4

U

'.

:t,

I..

~,T0

)+

>

v_ts

*; t-e: @ i,._

...

.t

.:.- ,-' * .\ '

.j; t.

.8

,r _,- _}

*vW

W

.

d:

*t s

b...e'.

i'+ F:

: *''' "' 4 1

b

ivI

V:

;EI;.:

49~~~~~.

4 .

A. V

2.Q.j~

p~~~~~~~~~~~~~~~~~~

b

ae ..

2

4,,

,a:. 3' 4r

S f ;' f}*~~~~~~04

A.r

4.

*

t,

x

-I.h .

AI

. 4

<, .l;

250um

D

I.

D

.el., I

-,i

%.It iii

,'. _-..M -t "; A1.

'04.

V, .

(7)

A

Figure 4. (A) Measurement oflungtissue

VEGFby ELISA. (B)Astandardcurve was

generated usingknown concentrations of VEGF.

within 2 h of hypoxia and further, that endotoxin and SNP decreasedVEGFmRNA, indicatingthat VEGFgeneexpression

is alsomodulatedbyNO. Thepotentialeffect of NOongrowth

factor geneexpression was recentlydemonstrated bythe

NO-induced downregulation of platelet derived growth factor (PDGF-B) transcripts incultured endothelial cells during hyp-oxia(32).

Theincreased VEGFgene expression duringchronic hyp-oxia correlated with the levels of VEGF protein. Our in situ hybridization studies using lungs from control rats and from

rats afterprogressively longer hypoxia exposures confirm the

studies by Monacci et al. (10) whoreported abundant VEGF transcriptsinratalveolarepithelialcells. However, inourstudy,

at28-32 d of chronichypoxia,theVEGFtranscript expression

wasincreased in alveolar cellsand,albeittolesserdegree,also invascularsmooth muscle cells and alveolarmacrophages.

Perhapsofgreaterimportance forthe roleof VEGF in the pathophysiology of vascularremodelingis thefactthat, during chronichypoxia, there isaparallelandevengreater upregula-tion in the expression of the KDR/Flk and Flt transcripts in lungs from animals exposed to hypoxia for up to 28-32 d. Notably, short-termexposure of isolated lungstohypoxiaalso increased the Flt and KDR/Flk mRNA. Whereas it has been reported that hypoxia is a strong stimulator of VEGF gene

expressioninculturedcells,it isanovelobservation that chronic hypoxia increases gene expression of VEGF receptors. Shear

stress mayaccount for theupregulation of the VEGFreceptor

genes inthe in vivo chronic hypoxia experiments since both

Figure3. In situhybridization for VEGF mRNA in lungs ofratsexposedtohypoxia for 21 and for 28 d. (A) Control lung: VEGF mRNA is

expressedin cellsof thealveolarsepta(arrow) and in intraalveolar cells, possibly macrophages (arrowhead). (B) Day 21 of hypoxia: notethe

increase inVEGF mRNAsignalasassessed semiquantitatively byanincrease in fast blue grain-like deposits in both alveolar septal, intra-alveolar

andbronchiolarcells(b, arrowhead).Asmallpulmonary vessel (v) exhibits low levelsofVEGFexpressionwhencomparedwiththeparenchymal

cells.(C) Day28ofhypoxia: high VEGF transcript levelsare seenin thesamecellularcompartments as seeninday21 ofhypoxia. D. Control

VEGFin situhybridization withsenseVEGF cRNA probe. (A-D) Digoxigenin-labeled probes, and immunoalkaline phosphatase; X400).

5-

4-z

a

943-a

1-1

U U

.

U

.

.IL-I

I

U

Day 3 Day 24 Hypoxia

Day28 Day 32

y a .291x + .003, R-squared: .002

0 0

1.5 2 2.5 3 3.5

VEGFinngfti B

(8)

2-50-urn

A

Figure5. In situhybridization forKDR/FlkmRNA inlungsofratsexposedtohypoxiafor 7 d.(A)Controllung: the alveolar

precapillary

blood

vessels (v) exhibitKDR/FlkmRNA,mostlyinthe endotheliallayer(arrows).Note weaksignalin the alveolar_capillaries_{(arrowhead). (B)}Control

senseKDR/Flk cRNA in normoxiclung. (C) Day7 ofhypoxia:KDR/FIkmRNAis present inmultipleprecapillary vessels,inadistribution similar

tothatseenin control animals.(D)Sense KDR/Flk cRNA inday7hypoxic lung. (A-D)Digoxigenin-labeledcRNAprobes,and immunoalkaline

(9)

I l

CONTROL ETX 2 NRS

-

-ETX 4 HiS ETX 6HiS ETX 12 NS

Figure6.Effect ofendotoxin on lung tissue: VEGF mRNA. Rats wereinjected withS.enteritidis

endotoxin (2mg/kg i.p.). Lung

RNA wasextracted andsubjected

to Northernanalysis. Bargraphs

representscanningdensitometry

ofhybridization products for

VEGF mRNA. VEGFtranscript

levelsin thelung decreasetime H dependentlyfollowingendotoxin

injection.

2 3 4 5 7 S

fl0SflwaN.4S

e C.

Figure7.(AandB)Effects of shorttermhypoxiaeffectsonlungVEGF,KDR/Flk andFltmRNA.Lungswereisolated andperfusedwitha

cell-free salt solution and either ventilatedwithanormoxic(21%02,5% CO2 balance N2)or ahypoxicgasmixture(0%02,5%CO2balanceN2)for

2 h. Insomeexperiments, either SNP(IO'M)orNAME(10-4 M)wasaddedtoperfusate.Each bargraphrepresents themeanoftwoorthree

independent experiments (A). 1600 r

1400 ±

1200 +

1000 +

800 +

600 +

400 ±

200 +

0

o Pr.oe9

I' -_ lay&

VEGF mRNA

.Wo

-KDR mRNA 'kt.0

M1 mRNA s00]

2mjj

10s

685kb- .

'Ka

To 't

K-1o 0o

Z

A

U_

(10)

0%02

21%

02

-N

CTL

IL-NAME~~~~~~

2 1 r;7 ( )2 () (7( 02

'1I:

I 2

)20

+ ( 2 +

l-NANMT L-NAM1

'I _1% ₄ _i _f ₇ ₈

I)10 11 1I

VEGF rnRNA

cTL

IL-NAME

A

Figure 8. (A and B) Effect of nitric oxide synthase inhibitiononVEGFgeneexpression inisolated perfused lungs. Lungswereventilated either

withroomairorunder hypoxic conditions (0%02)for2 h with andwithoutL-NAME (CTL)additiontothecell-free perfusate. The bar graphs

representthe densitometric signals. Each barrepresentsthemeanofthreeseparateexperiments.

KDR/Flk and Flt contain in theirsequencetheshearstresscore

element (5). This element is also present in the shear stress

sensitivegrowthfactorsPDGF-B and TGF-/3. However, inour

acutehypoxic experiments, the isolated lungswereperfusedat

constant flow andpressure throughout theentire experimental

period; sincenoagentwasusedto"prime" the hypoxicpressor response, nohypoxic vasoconstriction and thereforenoincrease

in shearstressoccurred. ThereforewesuggestthatlungVEGF

receptor gene expression can be modulatedby hypoxia alone

orbyhypoxiaand shear stress.

Hypoxia (either acutely or chronically) increased VEGF gene expression, whereas endotoxin time-dependently de-creased thelung tissue VEGF mRNA(Fig. 6). Ourstudies did

not address whether endotoxinsuppressesVEGFtranscription

ordecreases"messagestability."Sinceendotoxin isapowerful activator of NOsynthesis(29),wehypothesizedthatendotoxin

decreasedlung tissue VEGFgene expression via induction of NO synthesis in thelung. In agreement with this hypothesis,

wefound that the NO donor SNP inhibited thehypoxia-induced

upregulationoflungVEGFgeneexpression.Moreover, inhibi-tion of NOsynthesis byL-NAMEduringnormoxicorhypoxic

ventilation of isolatedlungs,increasedVEGFgeneexpression. Endogenously producedNO maymodulate both the extentof hypoxicvasoconstriction (35, 36)and also the VEGFsynthesis in the normal lung. Furthermore, during chronichypoxia,

en-dogenous lung NO production is likely increased (37) and, if so,duringchronichypoxiaVEGFgeneexpressionmaybe

mod-ulated(downregulated) by suchaNOdependentmechanism.

The role, if any, that VEGF plays in chronic pulmonary hypertensivestatesispresentlyunclear. On theoretical grounds

one mightbe attracted to the idea that VEGF, because of its

permeability/angiogenesisandpro-inflammatory properties (12) and because of its specificity for endothelial cells, might be involved in acute lung vascular injury and in endothelial cell proliferation.Inseverepulmonary hypertensioninman, particu-larly in theunexplainedform ofprimarypulmonary hyperten-sion,there is evidence forvigorousendothelialcellproliferation in the so-calledplexiform and intimal concentric proliferative lesions(38). BothVEGF protein andtranscripts arepresent in these lesions (34).We wondered whether VEGF couldplaya

role in thepulmonary vascularremodelingin the chronic

hyp-oxia rat model. We treated rats with suramin during chronic hypoxicexposure since suramin has been shownto block the interaction of VEGF with itsreceptors(39). Suramindecreased the degree of pulmonary hypertension and the development ofright ventricular hypertrophy, without affecting lung VEGF mRNAand protein levels (40). One possibility is that the

sura-min-induced reduction in pulmonary hypertensionwas

associ-ated with inhibition of the VEGF-receptor interaction. More specific therapeutic strategies (including theuseof anti-VEGF antibodies) will be requiredto testthishypothesis more

rigor-ously since suramin also blocks the actions of other growth factors suchasPDGF and fibroblast growthfactor(41, 42).

Inconclusion,bothacuteandchronic hypoxia increase lung tissue gene expression for VEGF and its receptors. Chronic hypoxia leadstoincreased synthesis (or decreased breakdown) of the VEGFprotein. The factthatlungVEGFtranscripts are

increased after only 2 hofhypoxic challenge in isolated

per-fusedlungs is consistent with theconceptthat VEGF behaves

as an earlygene. Its expressionin thelungappears tobe also under the control of NOor aNO-dependentmetabolite. Whether VEGF and VEGFreceptorsareinvolved in the vascular

remod-elingassociated withhypoxia-induced pulmonary hypertension remainsto beseen.

Acknowledgments

The authors wish tothank RebeccaKendigfor herhelpintypingthe

manuscriptand Mr. Thomas Bates andKellyWade for technicalhelp.

This workwassupported byaGrant in Aidto Norbert F. Voelkelby

the American Heart Association andbythe Vascular Academic Award

NIH K07 HLB.

References

1. Voelkel, N.F., U.Wiegers, V.Sill, and J. Trautmann. 1977.A kinetic study oflung DNA-synthesis duringsimulated chronic high-altitude hypoxia. Thorax. 32:578-581.

2. Niedenzu, C., K.Grasedyck,N. F.Voelkel, S. Bittman,andJ. Lindner. 1981. Proliferation oflungcells inchronically hypoxicrats.Int.Arch. Occup. Environ. Health. 4:185-193.

3.Rabinovitch,M., W. Gamble,A. S. Nades,0.S.Miettinen,and L. Reid.

1979.Ratpulmonarycirculation after chronichypoxia: hemodynamicsand

struc-turalfeatures.Am. J.Physiol.236:H818-H827. 800

-400

-200

-01

(11)

4.Dempsey, E. C., I. F. McMurtry, and R. F. O'Brien. 1989. Protein kinase C activation allows bovine pulmonary artery smooth muscle cells to proliferate inresponse to hypoxia. Am. Rev. Respir. Dis. 139:A171.

5. Resnick, N., T. Collins, W. Atkinson, D. T. Bonthron, C. Forbes Dewey Jr.,and M. A. Gimbrone, Jr. 1993. Platelet-derived growth factor B chain promoter contains a cis-acting fluid shear-stress-responsive element. Proc. Natl. Acad. Sci. USA.90:4591-4595.

6. Senger, D. R., S. J. Galli, A. M. Dvorak, C. A. Perruzzi, V. S. Harvey, and H.F. Dvorak. 1983. Tumor cells secrete a vascular permeability factor that pro-motes accumulation of ascites fluid. Science (Wash. DC). 219:983-985.

7. Ferrara, N., and W. J. Henzel. 1989. Pituitary follicular cellssecrete anovel heparin-binding growth factor specific for vascular endothelial cells. Biochem. Biophys. Res. Commun. 161:851-858.

8. Berse, B., L. F. Brown, L. van de Water, H. F. Dvorak, and D. R. Senger. 1992. Vascular permeability factor (vascular endothelial growth factor) gene is expressed differentially in normal tissues, macrophages and tumors. Mol. Biol. Cell.3:211-220.

9.Shweiki, D., A. Itin, G. Neufeld, H. Gitay-Goren,and E. Keshet. 1993. Patterns of expression of vascular endothelial growth factor (VEGF) and VEGF receptors in mice suggest a role in hormonally regulated angiogenesis. J. Clin. Invest.91:2235-2243.

10. Monacci, W. T., M. J. Merrill, and E. H. Oldfield. 1993. Expression of vascular permeability factor/vascular endothelial growth factor in normal rat tis-sues.Am. J. Physiol. 264:C994-C1002.

11. Breier, B., U. Albrecht, S. Sterrer, and W. Risau. 1992. Expression of vascular endothelial growth factor during embryonic angiogenesis and endothelial celldifferentiation. Development (Camb.). 114:521-532.

12.Clauss, M., M. Gerlach, H. Gerlach, J. Brett, F. Wang, P. C. Familletti, Y.-C. E. Pan, J. V. Olander, D. T. Connolly, and D. Stern. 1990. Vascular

permeabilityfactor:atumor-derived polypeptide that induces endothelial cell and

monocyte procoagulant activity, and promotes monocyte migration. J. Exp. Med. 172:1535-1545.

13. Connolly, D. T., J. V. Olander, D. Heuvelman, R. Nelson, R. Monsell, N. Siegel, B. L. Haymore, R. Leimgruber, and J. Feder. 1989. Human vascular permeability factor. J. Biol. Chem. 264:20017-20024.

14.Senger, D. R., L. Van De Water, L. F. Brown, J. A. Nagy, K.-T. Yeo, T.-K. Yeo, B. Berse, R. W. Jackman, A. M. Dvorak, and H. F. Dvorak, H.F. 1993. Vascularpermeability factor (VPF, VEGF) in tumor biology. Cancer Metas-tasisRev. 12:303-324.

15.Shweiki, D., A. Itin, D. Soffer, and E. Keshet. 1992. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature (Lond.). 359:843-845.

16. Terman, B. I., M. E. Carrion, E. Kovacs, B. A. Rasmussen, R. L. Eddy, and T. B.Shows. 1991. Identification of a new endothelial cell growth factor receptortyrosine kinases. Oncogene. 6:1677-1683.

17. Terman, B. I., M. Dougher-Vermazen, M. E. Carrion, D. Dimitrov, D. C. Armellino, D. Gospodarowicz,and P. Bohlen. 1992. Identification of the KDR

tyrosinekinaseas areceptor for vascular endothelial cell growth factor. Biochem.

Biophys.Res. Commun. 187:1579-1586.

18.Quinn,T.P., K. G. Peters, C. De Vries, N. Ferrara, and L. T. Williams.

1993. Fetal liver kinase 1 is a receptor for vascular endothelial growth factor and is selectively expressed in vascular endothelium. Proc. Natl. Acad. Sci. USA. 90:7533-7537.

19. Shibuya, M., S. Yamaguchi, A. Yamane, T. Ikeda, A. Tojo, H. Matushime, and M.Sato. 1990. Nucleotide sequence and expression of a novel human recep-tor-type tyrosine kinase gene(flt) closely related to thefinsfamily. Oncogene. 5:519-524.

20. De Vries, C.,J.A. Escobedo, H. Ueno, H., K. Houck, N. Ferrara, and L.T. Williams. 1991. Thefins-liketyrosine kinase, a receptor for vascular endo-thelialgrowth factor. Science (Wash. DC). 255:989-991.

21. Waltenberger, J., L. Claesson-Welsh, A. Siegbahn, and C.-H. Heldin. 1994. Functionalcomparison of KDR and Flt receptors for the vascular endothelial

growthfactor. J. Cell Biochem. Suppl. 18A:322.

22. Millauer, B., L. K. Shawver, K. H. Plate, W. Risau, and A. Ulrich, A.

1994.Glioblastoma growthinhibited in vivoby dominant-negativeFlk- 1mutant.

Nature(Lond.).367:576.

23. Halbower, A. C., R. J. Mason, S. H. Abman, and R. M. Tuder. 1994. Agarose infiltration improves morphology of cryostat sections oflung.Lab. Invest. 71:149-153.

24. Ono, S., J. Y. Westcott, and N. F. Voelkel. 1992. PAF antagonists inhibit pulmonary vascular remodeling inducedbyhypobaric hypoxiainrats.J.Appl. Physiol 73:1084-1092.

25. Chomcynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation

byacidguanidinium-thiocyanate phenol chloroform extraction. Anal. Biochem.

162:156-159.

26. Webster, K. A., P. Gunning, E. Hardeman, D. C. Wallace, and L. Kedes. 1990. Coordinate reciptrocal trends in glycolytic and mitochondrialtranscript accumulations during the in vitro differentiation of human myoblasts. J. Cell. Physiol. 142:566-573.

27. Kletzien, R. F., P. K. Harris, and L. A. Foellmi. 1994.Glucose-6-phosphate dehydrogenase: a "housekeeping" enzyme subjecttotissue-specific regulation by 7-21 hormones,nutrients, and oxidant stress. FASEB(Fed. Am. Soc. Exp. Biol.) J.8:174-181.

28.Reid, L. M., and P. Davies. 1989. Control of cellproliferationinpulmonary

hypertension. InPulmonary Vascular Physiology and Pathophysiology. E. K.

Weir,and J. T. Reeves, editors. Marcel Dekker, New York. 541-611.

29.Kilbourn,R.G.,and P. Belloni. 1990. Endothelial cellsproduce nitrogen

oxides in response to interferon-y, tumor necrosis factor and endotoxin. In Nitric Oxidefrom L-Arginine: A Bioregulatory System. S. Moncada, and E. A.Higgs, editors. 61-67

30.Meyrick,B. O., and L. Reid. 1987. The effect of continuedhypoxiaon

ratpulmonary arterial circulation: anultrastructuralstudy. Lab. Invest. 38:188-200.

31. Jones, R. C. 1993. Role of interstitial fibroblasts andintermediatecellsin microvascular wall remodelling in pulmonary hypertension. Eur. Respir. Rev. 3:569-575.

32. Kourembanas, S., L. P.McQuillan,G. K. Leung, and D. V. Faller. 1993. Nitric oxideregulates the expression of vasoconstrictors and growth factors by vascular endothelium under both normoxia and hypoxia. J. Clin. Invest. 92:99-104.

33. Dempsey, E.C., K. R. Stenmark, I.F. McMutry, R. F. O'Brien, N. F. Voelkel, and D.B. Badesch. 1990. Insulin-Like growth factor I and protein kinase C activation stimulatepulmonaryartery smooth muscle cellproliferationthrough separate butsynergisticpathways. J. Cell. Physiol. 144:159-165.

34.Tuder, R. M., D. B.Badesch,B.Groves,D. A.Lynch, and N. F. Voelkel. 1994. Vascular endothelialpermneability/growthfactorexpression in plexogenic pulmonary hypertension. J. Cell. Biochem. Suppl 18A:330.

35.Hasunuma, K., T. Yamaguchi, D. M. Rodman, R. F. O'Brien, andI. F.

McMurtry. 1991. Effects of inhibitors of EDRF and EDHF on vasoreactivity of

perfusedratlungs. Am. J.Physiol. 260:L97-L104.

36. Eddahibi, S., S. Adnot,Cl Carville,Y.Blouquit,and B.Raffestin. 1992.

L-argininerestoresendothelium-dependent relaxation in pulmonary circulation of

chronicallyhypoxicrats. Am.J.Physiol. 263:L194-L200.

37. Barer, G., C. Emery, A. Stewart, D. Bee, and P. Howard. 1993.Endothelial controlof thepulmonarycirculation in normal and chronically hypoxic rats. J.

Physiol.463:1-16.

38.Tuder, R. M., B. Groves, D. B. Badesch, and N. F. Voelkel. 1994.

Exuber-antendothelial cellgrowth and elements of inflammation are present in plexiform lesions of pulmonary hypertension. Am. J. Pathol. 144:275-285.

39.Vaisman, N., D. Gospodarowicz, and G. Neufeld. 1990. Characterization of the receptors for vascular endothelial growth factor. J. Biol. Chem.

265:19461-19466.

40.Voelkel,N.F., M. A. Voelkel, and B. Flook, B. 1993. Suramin inhibits

monocrotaline(MCT)-induced pulmonary hypertension in rats. Circulation. Suppl

88:1-285.(Abstr.)

41. Pesenti, E., F. Sola, N. Mongelli, M. Grandi, and F. Spreafico. 1992.

Suramin prevents neovascularisation and tumour growth through blocking of basic fibroblast growth factor activity. Br. J. Cancer. 66:367-372.