An autocrine role for urokinase in phorbol
ester-mediated differentiation of myeloid cell
lines.
A R Nusrat, H A Chapman Jr
J Clin Invest.
1991;
87(3)
:1091-1097.
https://doi.org/10.1172/JCI115070
.
The human myeloid cell line HL60 secretes urokinase-type plasminogen activator (uPA)
and expresses its receptor. When stimulated with phorbol myristate acetate (PMA), both
secretion of uPA and the expression of its receptor are up-regulated, and these cells
differentiate to an adherent phenotype. This adhesive response is markedly reduced in the
presence of uPA antibodies. The PMA response is restored by the addition of native uPA,
an amino-terminal fragment of uPA (residues 1-143) devoid of proteolytic activity, or a
synthetic peptide (residues 12-32) from the uPA growth factor domain known to mediate
receptor binding. In contrast, the addition of catalytically active low molecular weight uPA,
which is missing the growth factor domain, or a peptide from the catalytic domain (residues
247-266) is ineffective. The influence of uPA antibodies on a second marker of macrophage
differentiation, cysteine proteinase activity, was also examined. Cysteine proteinase activity
of HL60 cells is increased in PMA-treated cells after 24 h but it fails to increase in the
presence of anti-uPA. This increase in cathepsin B-like activity is also restored by
exogenous uPA. These experiments indicate that an autocrine interaction of the growth
factor domain of uPA with its receptor mediates an essential step in PMA-mediated myeloid
cell differentiation.
Research Article
An
Autocrine Role for Urokinase in Phorbol
Ester-mediated Differentiation of Myeloid Cell Lines
A. RoomiNusrat and Harold A. Chapman, Jr.
Department ofMedicine, Brigham andWomen'sHospital andHarvard Medical School,Boston, Massachusetts02115
Abstract
The human myeloid cell line HL60 secretes urokinase-type plasminogen activator (uPA) and expresses its receptor. When stimulated with phorbol myristate acetate (PMA), both secre-tionofuPAand theexpression ofits receptor are up-regulated, and these cells differentiate to an adherent phenotype. This adhesive responseismarkedly reduced in the presenceofuPA antibodies. The PMA response is restored by the addition of native uPA, an amino-terminal fragment of uPA (residues 1-143) devoid of proteolytic activity, or a synthetic peptide
(residues 12-32)fromthe uPAgrowthfactor domainknownto
mediatereceptorbinding.In contrast, theaddition of catalyti-cally active low molecular weight uPA, which is
missing
the growthfactor domain, or apeptidefrom thecatalytic domain (residues247-266) isineffective.Theinfluence ofuPAantibod-ies onasecond marker ofmacrophagedifferentiation, cysteine proteinase activity,wasalso examined.Cysteine proteinase
ac-tivityof HL60 cells is increased in PMA-treated cells after 24h
butit fails to increase in the presence of anti-uPA. This
in-crease in cathepsin B-like activityis also restored by exoge-nousuPA. Theseexperimentsindicate thatanautocrine
inter-action of the growth factordomain ofuPA with its receptor
mediatesanessentialstep in PMA-mediatedmyeloidcell dif-ferentiation.(J. Clin.Invest.1991.87:1091-1097)Key words:
cathepsin B *celladhesion* HL6O cells-phorbol myristate
acetate*urokinasereceptor
Introduction
Manymammaliancellssynthesizeand release
urokinase-type
plasminogen activator (uPA)
(1).'
Expression of thisenzyme has been implicated in awide variety of biological processesincluding cellular migration, connective tissue matrix
turn-over, trophoblastic implantation, and tumortransformation (2-5). Many if not allofthecellswhichsecreteuPA,including
This workwaspresented inpartatthe AnnualMeetingof the
Ameri-canFederation for ClinicalResearch,Washington, DC, May1990. Address reprint requests to Dr.Nusrat, RespiratoryDivision,
Brigham and Women's Hospital, 75 FrancisStreet, Boston, MA
02115.
Receivedfor publication6August1990andinrevisedform22 Oc-tober1990.
1. Abbreviations usedinthispaper:ATF, amino-terminalfragment (of
urokinase); CD, catalytic domain; DFP, diisopropyl-fluorophosphate; GFD, growthfactordomain; LMW,lowmolecularweight; uPA,
uroki-nase-typeplasminogen activator.
myeloid cell lines, monocytes, and macrophages, also express a high-affinity uPA receptor (1, 6-10). The amino-terminal re-gion of the uPA molecule, termed the growth factor domain (residues 1-32), mediates binding of uPA to this receptor (1 1). Receptor-bound uPA on mononuclear phagocytes has a rela-tively long half-life (hours) and may be protected from inhibi-tion by soluble plasminogen activator inhibitors, facilitating cellular migration and cell-mediated connective tissue
turn-over(12-15). Thus, one established role for cellular expression of uPAand its receptor is to focus plasminogen activator activ-ity to the cell surface.
When stimulated with PMA, the anchorage-independent myeloid cell lines, THP- 1, U937 and HL60, differentiateto an
adherentphenotype that have manycharacteristics of
macro-phages (16-19). Among the known changes which followPMA
stimulationareincreased secretion of uPAaswellasincreased expression of uPA receptors (8, 9, 20-22). We now show that
binding of the growth factor domain of uPA to its receptor playsanessential role in PMA-mediated myeloid cell adhesion and maturation. These data extend the functionsascribed to
the uPA receptorto onethatincludesarole insignal transduc-tion. The manuscript details work done with the HL60 cell
line. However, the observations are similar for the PMA-re-sponsive human leukemic THP-1 and U937 cells, indicating
thegenerality ofthisphenomenon.
Methods
Reagents. RPMI 1640wasfrom GibcoLaboratories, Grand Island, NY; bovine fetalserumfrom Hyclone,Logan,UT;andtissue
culture-warefrom Falcon Labware, LincolnPark, NJ. CelllinesHL60, U937, and THP-1 cellswereobtained from American Type Culture
Collec-tion, Rockville,MD. [3H]thymidine(2 Ci/mmol)wasfrom New
En-gland Nuclear, Boston,MA.Two-chainhigh molecular weight
uroki-nase was agift oftheAmericanRedCross,and 2 99% pure
recombi-nanttwo-chainhigh molecular weight urokinase (103,900 IU/mg) from Chinese hamsterovary cellswas a gift ofDr. Robert Broeze,
Collaborative Research, Bedford,MA.33-kDlowmolecularmassuPA
(no detectable 54-kDuPAby zymography [23]), murine monoclonal anti-uPA (No. 377) and (No. 394),andprotein-Aaffinity-purified rab-bitanti-uPA (No. 389)were obtained from AmericanDiagnostica,
Greenwich,CT.Polyclonal rabbit anti-uPA (No. 6200)waspurchased from Alpha Therapeutics, LosAngeles, CA.Theamino-terminal
frag-mentof uPA, residues 1-143,was agiftofDr.JackHenkin, Abbott Laboratories,AbbottPark, IL (24). Alkaline-phosphatase labeled goat
anti-rabbit IgG (051506) wasfrom Kirkegaard Perry, Gaithersburg,
MD;fluoresceinatedFab2 goatanti-mouseIgGfromZymed, San
Fran-cisco,CA;andsupernatants of murine monoclonal hybridomaclone TS1/18 to theinvariant chain ofafamilyleucocyte celladhesion mole-cules(25)were agift ofDr. L. Kobzik, HarvardSchool ofPublic
Health. Synthetic peptides comprisinguPAresidues 12-32and
247-266weresynthesizedon apeptide synthesizer(model9600, Biosearch/ Milligen,SanRafael, CA)andHPLC-purified. Boc-valine-leucine-ly-sine-aminomethylcoumarinand benzoyloxycarbonyl-phenylanalyl-ar-ginine-4-methyl-7-aminomethylcoumarin (Z-Phe-Arg-NMec) were
J.Clin. Invest.
© TheAmericanSocietyfor ClinicalInvestigation,Inc.
0021-9738/91/03/1091/07 $2.00
fromEnzymeSystems Products, Livermore, CA. Silicone oil and dioc-tylphthalatewasobtained from Aldrich ChemicalCo., Milwaukee,WI.
ProteinA-purified neutralizing rabbit anti-bovine tissue factor, also effective against human tissue factor,was agift ofDr. RonBach(26). Other chemical reagents were from Sigma Chemical Co., St. Louis, MO.
Cell culture. HL60 and U937 cells were cultured in RPM1 1640 supplemented with penicillin (100 U/ml), streptomycin (100 ,ug/ml),
and 10%bovinefetalserum.For THP-1 cellculturemedium, 48 ,uM 2-mercaptoethanolwasaddedtothe mixture described above.Human
alveolar macrophageswere cultured in Dulbecco's modified Eagle's medium supplemented with 20mM Hepes, penicillin (100 U/ml), streptomycin (100 ,ug/ml), and10%bovine fetalserum. Allincubations
were at 37°C, 5% C02/95%airandcells were plated at 1X106/ml.The
celllinesweredifferentiated along the macrophage pathway with 16 nM PMA(16).
Bronchoalveolar lavage. After informed consent, normal
volun-teerswithahistory ofsmoking underwent bronchoalveolar lavage
us-ingafiberoptic bronchoscope ina wedgedpositionaspreviously
de-scribed (27).
Assayfor adherent cells. Adherent cells were assayed by three tech-niques: visualization by phase-contrast microscopy, incorporation of [3H]thymidine intheadherent cellpopulation and bymeasurementof total adherent cellprotein. Similar quantitativeresults wereobtained
when eitheradherent labeled cells or adherent cellprotein was mea-sured.Visualinspection consistently correlated withdataobtained
ei-therbyproteinor[3H]thymidine incorporationassays. Forthymidine incorporation,cells at 1 x106/mlwere labeledwith 1,uCi/ml
[3H]thy-midine andpulsed with 16nMPMA,and 100,u of cellsweredispensed intriplicatesorquadruplicates in 96-well microtiter plates. After 16-24 hof incubation, the plateswasimmersed inabeakerof normalsaline, and nonadherent cellswere flicked out. After three plate rinses the adherent cellswerelysed in 50Al oflysisbuffer(phosphate-buffered
saline pH 7.4/10% glycerol/0.2% sodium dodecyl sulfate/0.2% Triton
X-100) and counted inaliquidscintillationcounter.Thisassay scored
cellsin asimilarphysiologicalstate,thatis,cellstightly attachedtothe microtiter plates. Similarly, afterthenonadherent cellswereremoved
by the saline washes described above, adherent cellproteinwasassayed byhydrolyzing the cellsin 5 NNaOHfor4 hwhile the cellswere on a
microtiter plate shaker. The lysatewasneutralized with equal volume5 NHCL andproteinwasmeasuredusing the Bradfordassay (No. 500-0006, Bio-Rad Laboratories,Richmond,CA).
AssayfortheeffectofuPA onunstimulated HL60 cells. Cells at 1
X 106/mlwerepulsedwith0.1 tC/ml [3H]thymidineinthepresenceof graded doses ofuPA(0-100 nM/ml) and 100
AI
cellsweredispensed in microtiter plates. 16-24hlater, the cellswereharvestedusinganauto-matic cellharvester, and cell associated radioactivityonfilters counted inascintillationcounter.
Diisopropylfluorophosphate (DFP) inactivation ofuPA. uPA was inactivated with 5 mM DFP at room temperature forone hourand
exhaustively dialyzed. Residualcatalytic activitywasassayed fluoro-metrically using the plasmin dependent substrate Boc-val-leu-lys-aminomethylcoumarin, 0.2 mM, in the presence orabsence of plasminogen, 5lAg/ml. DFP treatment resulted in undetectable uPA
activity.
Radiolabeled uPA equilibriumbindingstudies. uPA was radiola-beledbythemethod ofHunter andGreenwood(28),except that the amountchloramine-Twasreducedto 30Aginareactionvolumeof 115
Al.
Unstimulatedorcells stimulatedovernight with 16nMPMA wereusedfor equilibrium binding studies and Scatchard analysisper-formed. PMA-stimulated cellswerescraped offthe dishes witharubber policeman (8, 9). Iodinated recombinant uPAwasincubated in 3
X 10-12to 9 x10-9Mrangewithandwithout100-foldexcesscolduPA
for 2 hon ice. Boundradioactivity wasseparated bycentrifugation
throughanoil cushion of1:1 mixture of silicone oiland
dioctylphthal-ate.Microfugetipswere cutandcounted inagamma counter.Initial studies confirmed previously reported observations ofothers that
HL60cellshavespecific saturable receptorsforuPAin the
subnano-molar range and that the receptor number increases five-toeightfold
afterovernightstimulation withPMA(9).
AssayofprourokinaseantigenbyELISA. HL60cells were main-tained inserumfree mediacontaining0.1% BSAsupplementedwith insulin, transferrin and seleniumasdescribed(21). Supernatantsfrom control or overnightPMA stimulated cellswereconcentrated using
filterswith a I0-kDcutoff from Amicon Corp., Danvers, MA. The sandwich ELISA consisted ofauPAmonoclonal(No.394)forantigen
capture,polyclonal rabbit anti-uPA(No. 6200)asthe secondlayer,
andalkaline-phosphatase labeled goat anti-rabbit IgG for detection (29). This assay hasnocrossreactivitytotissueplasminogenactivator. Fibrin zymography (23) verified that the supernatants from these cells containeda plasminogenactivator that comigratedasasingle clear 54-kD bandidenticaltopurified uPA. Wefound that theamountof
prourokinaseantigeninthesupernatantsofPMAstimulated cells
in-creasesbyabout 15-fold ascomparedtocontrolunstimulated cells,
confirmingworkdoneby otherinvestigatorswith the HL60 and U937 cell lines(21, 22).
Flow cytometry. Control unstimulated, orPMA-stimulated cells
werecultured overnight with anti-uPA (No. 6200), 30 jil/ml. PMA stimulated and anti-uPA treated nonadherent cells, 90% of total,were
aspiratedandtreated with pH 3 buffertoremoveany cell associated uPA(10). TheywerethenexposedsequentiallytouPA(2 nM),
mono-clonal anti-uPA (No. 377) (20jg/ml),andfluoresceinatedFab2goat anti-mouse. Hybridomasupernatants of clone TS 1/18wereusedto
assess CD 18expression(25).Nonspecificfluorescencewasdefinedas
thatmeasuredon cellsexposedto amixtureof murineIgG1,IgG2a,
andIgG2b (each10jg/ml)followedbyfluoresceinatedFab2. Analysis
wasperformed usinganOrtho 2151 cytofluorograph.
Cysteineproteinase assay.Z-Phe-Arg-NMecwasusedas asubstrate forcathepsinB-and L-like cellularcysteine proteinases employingan
assay basedonthatdescribedbyBarrett(30-32).Adherent and nonad-herent cells were combined and centrifuged to separate them from supernatant fluid beforelysisinabuffercomposedof 20mMsodium acetate, 1% TritonX-100,3mMEDTAatpH 5.5, andsubjectedto a
freeze-thaw cyclethreetimes. Thefluorometric assaywasperformedby
incubating10-or
20-;il
aliquotsoflysatesat37°Cinatotalofone mlincubation buffercontaining20mM acetate,10AMZ-Phe-Arg-NMec, 3mMcysteineatpH5.5for 30min.Incubationinthe presence ofthe
cysteine proteinase inhibitor L-3-carboxy-trans-2,3-epoxypropyl-leucylamido[4-guanidino]butane(E-64)wasused toestablish
specific-ity of the assay (31). Each lysatewas incubated in the presence or
absence I ,uME-64 for the incubationperiod.Fluorescencewas
mea-suredusinganexcitationwavelengthof 383 nMandemission wave-length of 460 nM. Cysteine proteinase activitywas definedas E-64 inhibitablefluorescence andexpressedaspercentactivityof unstimu-latedcells.Expressionof fluorescenceas afunctionof totalproteinin
thelysatesdidnotaltertheresults.
Statisticalanalysis. Student'st test wasperformedforcomparison
ofpairedmeanexperimentalvalues(33).
Results
Urokinase antibodies
affect
the adhesionof
PMA stimulatedHL60cells.
Overnight
stimulation withPMAresults in - 50%ofthe cells
differentiating
to anadherentphenotype
(16).
Thisresponseto PMAwastested in the presence of rabbit antibodies
to uPA
(anti-uPA).
Adherence toplastic
wasassayed
asoneindex of differentiation. Cellswerepulsed with
[3H]thymidine
atthe time ofPMAstimulation and the number of adherent cellsat 16-24hdeterminedby
quantitation
ofresidualactivity
in culture wells washed to remove nonadherent cells. Halfa
microgram of
affinity
purified
anti-uPA per microtiter well resulted ina50%reduction,and 20jig
in> 85-90% reductioncellsmaintainedin serum-free mediumaswell aswith U937 and THP- 1 cells (not shown). Nonimmune rabbit IgG and neu-tralizing anti-tissue factor antibody did not attenuate cellular adherence (26). The addition of control nonimmune IgG
con-sistently resulted in more cells adhering than in PMA only wells(Fig. 1). This observation was not explored further.
The anti-uPA effect illustrated in Fig. 1 could be due to
toxicity,neutralization ofuPAenzymaticactivity, or
interfer-encewithbinding ofuPA toitsreceptor.Equilibriumreceptor
binding studies demonstrated that 20 ,ug of affinity-purified anti-uPAper 100
Ail
ofcellsreducedspecific binding of1 nMiodinateduPA to 24-h PMAstimulatedHL60 cells by . 85%. Theviabilityof PMA-stimulated, antibody-treated,
nonadher-ent cells was slightly reduced. These nonadherent cells were 15±5% (n = 6) trypan blue positive whereas control cells in culture were <5% positive. However, the addition of exoge-nousuPA(100 nM) to theseantibody-treatedcells resulted in
anincreaseinthenumberof adherentcells tovalues compara-bletoPMAonlytreated cells. Thisreversibility suggestedthat thereductioninadhesionwasnotduetotoxicity and butrather
involvedaspecificinteraction ofuPAwithuPAantibodies. Determination of the uPA domain(s) necessary for
adhe-sion. Asshown inFig. 2,uPAisamultidomain protein within
theseine proteinase
family.
Atthe amino terminus isagrowthfactor domain followed by aconnecting peptide, a kringle
structure, and carboxy-terminal catalytic domain (1). We testedvarious fragments ofuPAfor their abilityto reversethe
effect of anti-uPA shown in Fig. 1. Previous workhas
estab-lishedthat acatalytically inactive amino-terminalfragment of
uPA(ATF),residues 1-135,aswellassyntheticpeptides from
thegrowthfactor domain(residues 1-32) competeeffectively
with uPA forbindingtothe uPA receptor(11,20). A peptide
comprising
residues 12-32 (21 amino acids) fromthegrowth factordomain (GFD-peptide) known tobeeffective in this competition was synthesized. As acontrol fortheGFD-pep-tide,
apeptide
comprising
residues 247-266 (20 amino acids) from the catalytic domain (CD-peptide) was also made. Theseine proteinase inhibitor, DFP,irreversibly inactivates uPA (DFP-uPA), and DFP-uPA interacts with uPA receptor as
avidlyastheactiveenzyme(34).Catalytically active33-kD low
160-100 l 2
c 0 L) 'O 0
:0
z
10 c
'E T
(D .r-10
7-IL
0
80-
60-40
20
n
{a.
i NonimmunekgG
-tissue tactor
oL-uPA
0.1 0.5 1.0 5.0 10 MicrogramsIgG
2 .io o
Figure 1. ProteinA-purified rabbit anti-uPA blocks adhesion of PMA-stimulatedHL60 cells. Indicatedamountsof antibodywas
addedatthetime ofPMApulse.Valuesaredisplayedaspercentof
control(no addedantibody). Comparableresults wereobtained with bothaffinity-purifiedornonaffinity-purifiedanti-uPA.
N9\
Growth Factor Domain
12 - 132
Growth Factor DomainPeptide
411
C
Kringle
I
S-S2471 1266 CatalyticDomain
Peptide
ii 1135
Amino-TerminalFragment
Catalytic Domain
(LMWuPA)
Figure2.Schematicrepresentation oftheurokinasemolecule andits fragmentsused in theexperiments. The ATF used here consisted of
residues 1-143(eightresidueslonger than the ATForiginally
described by Stoppellietal.[20]) and doesnothave thecatalytic domain. LMW-uPA is the 33-kD form of the molecule comprising thecatalytic domainand has no receptorbindingactivity.
GFD-peptide is residues 12-32 from the growth factor domain,and
CD-peptide is residues 247-266 fromthecatalytic domain. Diagram modified from FrancisandMarder [54].
molecular weight form of uPA (LMW-uPA),aproteolytically
cleavedproduct of thenative molecule devoid of the
amino-terminalfragment, doesnotbindtothe uPAreceptor(34). All
of thesefragments andforms of uPAweretested for their abil-ity to overcome the effect of anti-uPA on PMA-stimulated
cells.
Fig. 3 presents data from these experiments which have
been reproduced at least three times. When PMA stimulated
cells are cultured in the presenceof anti-uPAsufficientto
de-creasethe numberof adherent cells by atleast85%, the addi-tion ofuPA, DFP-uPA, ATF, and theGFD-peptide is asso-ciated with adose-dependent increase in the number of
adher-entcells. In addition, segments of uPA which include the
200
DFP-uPA
0
l/ATF GFD-peptide
0
100
~~~~~uPA
C ~~~~~~~~~~~CD-peptide
< UA~~~~~~~~LW-LPA
0
0.1 1 10 100 1000 10000
Concentration (nM)
Figure3. Inhibitionof adhesion by anti-uPA antibodies is reversible. ExperimentswereperformedasdescribedinFig. 1with adherence
measured as counts perminute [3H]thymidine. The various forms of
uPAwereaddedatthetime ofPMApulseto testfor theirabilityto
overcometheeffect of anti-uPA. TheuPAfragments testedare
described inFig.2 and thetext. Valuesaredisplayedaspercentof control (noaddedantibody).In separateexperiments, uPAitselfwas
also shown to beeffectiveinincreasingthe numberofadherent cells
to > 100%ofcontrol values. 100 nM LMW-uPAdidnot reversethe
effect of anti-uPA.
growth factor domain arecapable of increasing the adherent
cell number to values greater than thatseenwith PMA alone
(100%). Fragments of the uPA molecule which have only the
catalyticdomain (LMW-uPA),orjustapeptide fragment from
thisdomain (CD-peptide) are ineffective inreversingthe anti-uPAeffect.Experiments with U937 and THP- 1 cells also
dem-onstratedthattheanti-uPA mediated block in adhesion could bereversed with either uPAorGFD-peptide(not shown).
Photomicrographs ofthe effect of anti-uPA on the PMA responseofHL60 cells and the reversal of this phenomenon by ATF are shown inFig. 4. PMA stimulated cells are adherent andspread(Fig. 4 A). In the presence ofanti-uPA very few cells
adhere(Fig.4B), but withthe addition ofATF, 25 nM, thecells
areagainadherent and spread (Fig. 4 C). The presence of more cells inFig.4Cis relatedtothe amount of ATF usedas illus-tratedinFig. 3. Higherconcentrations of ATF result in almost twotimesmorecellsadhering than in PMA only controlwells.
Thesefindings indicate thatuPAfragmentswhich include the
growth factor domain,regardlessofthecatalytic domain, will
promote thePMA-inducedadherent phenotype.
Effect ofthe GFD
peptide
on PMA response. Data in Fig. S showthat for PMA stimulated cells in the presence ofanti-uPA, portions ofthe uPA molecule possessingthe GFD in-crease the number ofadherent cells above control values.
Therefore, the effect ofthe GFD-peptide itselfon the PMA response wastested inthe absenceofuPAantibodies.The
ad-dition ofthe GFD-peptide at the time ofPMA stimulation resulted inadose-dependent increaseinthe numberof
adher-ing cells (Fig. 5).Whenvisualized usingphase-contrast
micros-copy, the PMA and GFD-peptide stimulated cells appeared
morewell spread anddensecomparedwith PMA onlytreated
counterparts. This observation was additional support for a
roleofthe growth factor domain in augmenting PMA-triggered maturation.Incontrast, thecontrol CD-peptide fromthe
cata-lytic domain didnot alter the numberofadheringcells.
Effect
ofexogenous
urokinase on HL60proliferation.
It hasbeenreportedthatfortheCCL20.2 human epidermal cellline that uPA can serve as an autocrine mitogen and stimulates
thymidine incorporation
(42). However, thisactivityrequires
the presenceof both theGFD andanactivecatalytic domain.
Wetested to seeifexogenous uPA has anyeffectonHL60 cells. When uPA (0-100 nM) is added to cells in the absence of
PMA, we observed no induction of adherence and the cells remain viable.Inaddition,therewas noincrease inthe prolifer-ation ofcellsasindicatedby
[3H]thymidine
incorporation stud-ies. After overnight incubation, control cellsincorporated5,610±171
cpm. Thethymidine
uptakein the presenceofuPA was as follows: 0.8 nM (5,830±86 cpm), 4 nM (5,593±431 cpm), 20 nM (5,799±194 cpm), and 100 nM (5,578±244 cpm). Thisdata isrepresentative
ofoneofthreeexperiments,each done intriplicate. Thus,exogenous uPA doesnothavea
proliferative effectonthese cells.
Lack of effect
of
uPA antibodies on adhesion ofdifferen-tiated
HL60 cells and human alveolarmacrophages.
Human alveolar macrophages, obtained by bronchoalveolar lavage, were tested to assess theeffect of anti-uPAon maturetissue-type macrophages. These cells are known to synthesize and
binduPA(7, 27).An
overnight
adherence assay similartothat I describedinFig. 1, butwithoutPMAwasused.Therewasno differencein adherence betweenanti-uPA-treatedoruntreated macrophages (Table I). Moreover, uPA antibodies added toHL60cells 72hafterPMAstimulation didnot result in
appre-Figure4. Inhibition of PMA induced adhesion of HL60 cells by anti-uPA antibodies is reversed by the amino-terminal fragment of uPA. Cellswereculturedasdescribed in Methods and Figs. 1 and 3. (A) Control PMA stimulated cells. (B) Cells stimulated with PMA in the presence of rabbit anti-uPA to reduce adherence by>85%. (C) CellsasinBwith the addition of 25nMATF. Bar, 200 ,um.
ciabledetachmentofadherent cells (notshown).These obser-vations suggested that anti-uPA may interfere with the process
ofmyeloid cell differentiation and not directly with attach-ment.
Anti-uPAantibodies blockPMA inducedincreasein cellu-larcysteine proteinase
activity.
Macrophagesarecharacterizedby high levels of acidic lysosomal proteinases, including
membersofthecysteine proteinase familysuch ascathepsinB
andL(35,
36).
PMA-induced differentiation ofHL60 cellsis0
c-z
0
0
C.) <. z
0
ILl
0 z w w
z a
165
146
127
108
89
70
-20
GFD-peptide
100 1000
I I
> 14
1;
z
a-4 1
<L.
10000
Concentration (nM)
Figure 5.GFD-peptide augmentsthePMAadhesiveresponseina
dose-dependent fashion. The addition of thegrowthfactor domain peptidealoneatthetime ofPMAstimulationsignificantlyincreases the number of adherent HL60 cells. ControlCD-peptidefrom the carboxy terminus doesnothavethis effect.Datafromoneof three comparableexperiments. Adherencewasmeasuredas countsper
minute[3H]thymidine.
synthetic substrateZ-phe-arg-NMec at pH 5.5(30-32). Fig. 6 illustrates that the lysatesofPMA-treated cells have analmost twofoldincrease incathepsin B-like activitywhencompared with untreated HL60 cells (P<0.05, n =4). The classspecific cysteine proteinaseinhibitor E-64 reduced thecleavageof this substrate by2 80% indicatingthat the increase inactivity is dueprincipally tocysteine proteinases. PMAtreatmentin the presenceof anti-uPApreventscellsfromadheringandresults in a significant reduction in cysteine proteinase activity(P =0.012,n=4).AsindicatedinFig.6, theaddition ofuPA(100
nM) atthe timeof addition ofthe antibody overcomesthis effect,the cellsadhere,and theinduction ofcysteineproteinase
activityis restored.
PMA-stimulated and uPA antibody-treated cells express uPA receptors. To determine thestatus ofuPA receptors on
nonadheringHL60 cellsstimulatedby PMAinthe presenceof anti-uPA, analyses by flow cytometry were performed. The
expression ofuPA receptorsas indicatedby cellsurface uPA
staining is increased - 30-foldonthesenonadheringcells
(Ta-Table I. Anti-uPA Does Not
Efect
AdherenceofHumanAlveolarMacrophages
Treatment Cells adherent
,ulanti-APA ;Lgadherent cell protein
None 6.1±0.5 0.63 8.0±1.2 1.25 9.4±0.8 2.5 7.5±1 .0 5.0 8.4±1.3
Humanalveolar macrophages
(100-Mu
aliquots of 1 X 106/ml)wereincubated in microtiter plates with anti-uPAin theindicated quanti-tiesatthetime of initiation ofcell culture.After overnight culture,the wells were washed and adherent cell protein measured. PMA-stimu-latedHL60 cellsin the presenceof2.5,lofanti-uPA have a>85%
reduction in adherence.
CONTROL PMA PMA PMA Anti-uPA Anti-uPA +
uPA
Figure6. Anti-uPA (25jul/ml)inhibits accumulation of cellular cysteine proteinase activity inresponse toPMA, andexogenousuPA
overcomesthis effect.Activity of total HL60 celllysates,adherent andnonadherent, is shown. 100nMuPAwasusedto reversethe effectof anti-uPA.Data representE64 inhibitablecysteine proteinase activity.
ble I1). Adherence of mononuclear phagocytes is associated
with cell surface expression of molecules from the integrin fam-ily (37, 38). The effect of uPA antibodies on adhesion could be duetointerference with the expression of these molecules. Therefore staining for CD 18, the invariant chain of a family of cell adhesion molecules, was alsoperformed (25). The ob-served increaseinCD 18 after thePMApulsewasabout
two-fold(Table II),similartopreviously reportedvaluesusing
com-parable assay conditions(39). Thus uPAantibodies, while
blockingadhesion and induction ofmacrophage cysteine pro-teinase activity, do not eliminate PMA-induced increasesin
uPA receptorand CD 18antigen.
Discussion
Datapresented here demonstrateforthefirst timethat macro-phage differentiation ofmyeloidcells by phorbol estersis me-diated in partbyanautocrine interaction of thegrowthfactor domain of uPA with its receptor. Several lines of evidence
sup-TableII.HL60 CellsExposedtoPMAandAnti-uPAAntibodies
IncreaseExpressionofuPAReceptors
Mean fluorescence
Surfaceantigen Unstimulated PMA+anti-uPA
arbitrary units
uPA receptor ExpA. 10.2 321.4(31X) ExpB. 11.6 404.4(35X)
CD 18 ExpA. 190.4 278.2(1.5X)
Exp B. 175.0 354.2(2X)
Control unstimulated HL60 cells,orPMAstimulated cellsin the presenceofuPAantibodywerepreparedfor flowcytometry as
de-scribedin Methods. Thequantity of anti-uPAused(3
gl/
100glcells)resulted in 290%of PMA-stimulatedcells notadhering. The mean
nonspecific fluorescencein two separateexperimentswas15.7±7.6
arbitraryunitswithnodifference between unstimulatedand
stimu-lated cells. The values inparentheses indicate magnitudechangewith
respect tounstimulatedcontrols.
portthis conclusion. PMAstimulation is known to increase both the secretion of uPA and theexpressionof uPA receptors onmyeloidcell lines(8, 9, 21, 22).We have shown that anti-uPA will inhibit the binding of uPA to its receptors. When HL60 cells are treated with anti-uPA at thetimeofPMA
stimu-lation there is increasedcellsurfaceexpression ofuPA recep-torsand the CD 18antigen;however, there is aninhibition of other markersofdifferentiationsuch as adherence and cathep-sinB-like activity. Theaddition of uPA, ATF,or a 21 residue peptide fromtheGFDreversesthis inhibitionofdifferentiation byuPAantibodies.In contrast,thecatalyticdomainofuPA or acontrolpeptide from this domain failstoreversethis effect. Thus, thecatalytic domain doesnotappear toplayamajor role in this process. These data indicate that uPA mediatestwo
different functions in macrophage physiology: an autocrine
function(GFDdominated) independent ofenzymatic activity,
and acatalyticfunction (serine proteinase domain dominated) involvingplasminogen activation.Theformer being important in macrophage ontogeny and the latterimportant in cellular
movement and extracellular matrix metabolism. Both func-tions are mediated, at least in part, by interactions between uPAanditsreceptor.
Itis of interestthat the growthfactor domainaugments the
PMA-induced differentiation and adhesion ofcells when
addedby itself(Fig. 5). This observation raises the possibility thatin additionto anautocrine mechanism, uPA may
influ-ence myeloid cell differentiation in a paracrine fashion. Be-cause
fibroblasts,
endothelial cells,neutrophils,
andmonocytes are able to secrete uPA (1), stromal and inflammatory cells mayhaveaprominent roleinmodulating macrophage differ-entiation in vivo.The secretion ofplasminogen activators byhuman bone marrow cells has previously been linked to their differentia-tion. Earlyhumangranulocyte/macrophage progenitorcells secrete
tissue-type
plasminogen activator (tPA) whereasmore mature granulocyte and macrophage progenitor cells secrete uPA(40, 41). Althoughthe reasonfor this changeinpatternin enzymesecretion isnotknown,ourdatasuggest amechanism by whichuPAsecretion andmyeloid
celldifferentiation maybe
mechanistically
linked.Evidence that the uPA receptor may transducesignalling
has been suggested in otherexperimentalsystems. Forcertain
epidermal
cells,uPAhasbeenfoundtobe a weakmitogen,
and thedatahaveindicatedthat both thegrowth factor domainand thecatalytic domainarenecessary(42).Inview oftheseobser-vations,wemeasured theproliferativeresponseofHL60 cells
after stimulation with exogenous uPA (0.8-100 nM range),
andfound thatthere was nochange in
3H-thymidine
uptakeorinduction ofadherence. Thusforthese cells, uPAitselfdoes not
inducedifferentiationandthe secondsignal,PMA,whichacts
viaprotein kinaseCisalsorequired.Furthermore, ithas been shownthatuPAinthe absenceofits catalytic activitycan
medi-ate achemotactic role forneutrophils andaortic endothelial
cells
indicating
that the uPA receptorcanaltercellactivity (43, 44). Murine erythroleukemia (Friend) cells differentiated invitrohave asignificant reduction intime forcommitment to
terminaldifferentiationwhencultured in the presenceofATF
(45), consistent with our observations with the effect ofthe
GFD-peptide
and ATF onPMA-stimulated myeloidcells. These and ourobservationscontrastwiththose where thecatalytic domain ofuPA bound toitsreceptor appears toplaya
role inmatrix metabolism and detachmentfromculture
sur-faces. For Rous sarcoma virus-transformed chicken embryo
fibroblasts and mesangial epithelial cells, treatment with anti-uPA directedagainstthecatalyticuPA domain promotes adhe-sion and prevents themorphologicaleffects inducedbyPMA orcyclic-AMP mediated pathways (46, 47). Moreover, in ma-turing Frienderythroleukemia cells,where evidence exists for a role of ATF in inducing differentiation, monoclonal antibodies directed against the catalytic domain promote adhesion on fi-bronectin coated surfacespresumablybyinterferingwith me-tabolism of the matrix (48).
Induction of cathepsin B-like lysosomal proteinase activity is a reliable marker of macrophage differentiation (32, 35, 36). Our data show that the addition of uPA antibodies to PMA-stimulatedHL60 cells not only blocked induction of this
pro-teinaseactivity, but alsoresultedinlowerfluorometric activity than that of unstimulated cells (Fig. 6). Thereasonfor the latter is not clear. Itispossible that uPA has a"trophic" effect for
these cellsalongthe linesreported for variouscolony stimulat-ing factors(49). This would beconsistent with the small
de-crease inviability inantibody-treatedcells ascomparedwith controlunstimulated cells.Itis alsopossible that the
fluoromet-ric activities in control and PMA-stimulated and uPA anti-body-treatedcells are due to different enzymes. Thesynthetic substrateemployed couldbehydrolyzed bymorethanone
cys-teineproteinase. Sinceourdatashowthat notallattributes of PMA-induced differentiation are blocked by uPA antibodies
(Table II), the loweractivityinPMA-stimulated and antibody treated cellscomparedtounstimulated cells could reflect a
dif-ference inenzymeratioduringa stateofstimulationwhencells
cannotadhere.
Accumulating evidenceindicates that cellshapeand
cellu-larinteractions withextracellularmatricesmodulate gene
ex-pressionanddifferentiation (50-52).Forexample, adhesion of
monocytes toplastic surfaces has beenreported toregulate c-fos and m-CSF mRNA (53). Adhesion appears to be botha
markerforand anintrinsicpartof myeloid cell differentiation. Ourdataindicatethat uPA receptorbindingmodulates both
cathepsin
B-likeenzymeexpression
aswellascelladhesiveness.Itremainstobedefined whether the observed block in cathep-sin B-like
activity (Fig.
6)resultsfrom attenuated adhesivenessor from a more proximate block in differentiation affecting both markers. Ausefulaspectof the
experimental
system pre-sented here isthatitoffersaconvenient modelwhere thecon-sequences of interaction ofuPA with its receptorcan be
ex-plored. Further structural andfunctional studiesshould
delin-eatetheimportance of thispathwayin
understanding
another setof mechanisms whichregulatemyeloid
cellbehavior.Acknowledgments
Theauthors thank Drs.J.Drazen,L.Kobzik, R.J.Broeze,D.O'Hara, N. Stimler, and J. Swanson, for helpful discussions and advice. M.
Zeidelkindly permitteduseof his laboratory andreagents. L. Z.Sailor andA.Colbyprovidedexcellenttechnicalsupport. Dr. Nusratis
recipi-entof NRSA HL8002. Dr.Chapman isa CareerInvestigatorof the American Lung Association. This studywassupportedbyHL-35505,
NIH ES0002, and agrant(2289) from the Council for Tobacco Re-search.
References
2. Liotta,L., R. H.Goldfarb,R.Brundage,G. P.Siegal,V.Terranova,and S. Garbisa. 1981.Effect ofplasminogenactivator(urokinase), plasmin,and throm-binonglycoprotein and collagenouscomponents of basement membrane. CancerRes. 41:4629-4636.
3. Saksela, 0. 1985. Plasminogen activation and regulation of pericellular proteolysis.Biochim. Biophys.Acta.832:35-65.
4.Valinsky,J. E., E.Reich,and N. M. LeDouarin. 1981. Plasminogen activa-torin theBursaof Fabricius: correlations with morphogenetic remodelingand cellmigrations.Cell.25:471-476.
5.Ossowski,L.1988.Invivo invasion of modifiedchorioallantoic membrane bytumorcells:the role of cellsurface-bound urokinase. J. Cell Biol. 107:2437-2445.
6.Kirchheimer,J.C.,Y.-H. Nong, andH. G. Remold. 1988. Interferon-gamma, tumornecrosisfactor-alpha, and urokinase regulate the expression of urokinasereceptorsof humanmonocytes. J.Immunol. 141:4229-4234.
7.Chapman,H. A., P.Bertozzi,L.S.Zenzius,and R. Nusrat.1990.Human alveolarmacrophage urokinasereceptor:roleinlocalizingenzymeactivitytothe cellsurface.Am.J.Physiol.In press.
8.Picone, R.,E. L.Kajtaniak,L. S.Nielsen,N.Behrendt,M. R. Mastroni-cola,M. V.Cubellis, M. P. Stoppelli, S. Pedersen,K.Dano,and F. Blasi. 1989. Regulation of urokinasereceptorsinmonocytelike U937 cells by phorbolester, phorbol myristateacetate.J.CellBiol. 108:693-702.
9.Nielsen, L. S., G. M. Kellerman,N.Behrendt,R.Picone,K.Dano, andF. Blasi. 1988.A55,000-60,000M,receptorproteinfor urokinase-type plasmino-genactivator.J.Biol. Chem. 263:2358-2363.
10.Stoppelli,M.P., C.Tacchetti,M. V.Cubellis,A.Corti,V.J.Hearing,G.
Cassani,E.Appella, and F. Blasi. 1986. Autocrine saturation ofprourokinase
receptors onhuman A431cells.Cell. 45:675-684.
11.Appella, E.,E. A.Robinson,S.J.Ullrich, M. P.Stoppelli,A.Corti,G.
Cassani,and F.Blasi. 1987. Thereceptor-bindingsequenceof urokinase.J.Biol. Chem. 262:4437-4440.
12. Chapman,H. A., Z.Vavrin,and J. B.Hibbs. 1982. Macrophage fibrino-lyticactivity. Identification oftwopathways of plasmin formation by intact cells andofaplasminogen activator inhibitor. Cell. 28:653-662.
13.Kirchheimer,J.C., andH.G.Remold.1989.Functional characteristics of receptor-boundurokinaseon human monocytes:catalyticefficiency and
suscepti-bilitytoinactivation byplasminogenactivator inhibitors. Blood. 74:1396-1402. 14.Cubellis,M.V., P. Andreasen, P. Ragno, M. Mayer, K. Dano, andF. Blasi. 1989. Accessibilityof receptor-bound urokinasetotype- 1 plasminogen
activator inhibitor.Proc.Natl. Acad.Sci. USA. 86:4828-4832.
15.Ellis, V., T.-C. Wun, N. Behrendt,E.Ronne, andK.Dano.1990. Inhibi-tion of receptor-bound urokinasebyplasminogen-activatorinhibitors. J. Biol. Chem. 265:9904-9908.
16. Rovera,G.,D.Santoli, and C. Damsky. 1979. Humanpromyelocytic leukemia cells inculturedifferentiate intomacrophage-likecellswhentreated withaphorboldiester.Proc.Nail.Acad. Sci. USA. 76:2779-2783.
17.Harris, P., andP.Ralph. 1985.Humanleukemic models of myelomono-cytic development:areviewof theHL-60 andU937celllines. J.LeukocyteBiol. 37:407-422.
18.Collins, S.J.1987.TheHL60promyelocytic leukemiacellline:
prolifera-tion,differentiation,and cellular oncogeneexpression.Blood. 70:1233-1244. 19.Tsuchiya, S.,Y.Kobayashi,Y.Goto,H.Okumura, S. Nakae,T.Konno, and K. Tada.1982. Inductionof maturation ofcultured humanmonocytic leuke-mia cells byaphorbol diester. CancerRes.42:1530-1536.
20.Stoppelli,M.P.,A.Corti,A.Soffientini,G.Cassani,F.Blasi,and R.K.
Assoian. 1985.Differentiation-enhanced binding oftheamino-terminal
frag-mentof the human urokinaseplasminogenactivatorto aspecificreceptor on U937monocytes.Proc.Natl.Acad. Sci. USA. 82:4939-4943.
21.Alving,B.M., C.Krishnamurti,Y.-P.Liu,D. L.Lucas, and D. G.Wright. 1988. Stimulatedproduction of urokinaseandplasminogen activator inhibitor-2 byhumanpromyelocyticleukemia cell lineHL60. Thromb. Res. 51:175-185.
22.Genton, C.,E. K.0.Kruithoff,and W.-D.Schleuning. 1987. Phorbol esterinduces the biosynthesis of glycosylated and nonglycosylated plasminogen activatorinhibitor 2 in highexcess overurokinase-type plasminogen activator in humanU-937lymphomacells.J.CellBiol. 104:705-712.
23.Granelli-Piperno,A., and E.Reich. 1978.Astudy ofproteasesand pro-tease-inhibitor complexesinbiological fluids. J. Exp. Med. 148:223-234.
24.Barnathan,E.S.,A.Kuo, L.Rosenfeld,K.Kariko,M.Leskit,F.Robbiati, M. L.Nolli,J.Henkin,and D. B.Cines.Interaction of single-chain urokinase-typeplasminogen activator withhumanendothelialcells.J.Biol. Chem.In press. 25.Sanchez-Madrid, F.,J.Nagy,E.Robbins,P.Simon,and T. A. Springer. 1983.Ahumanleukocyte differentiation antigen familywith distinct alpha-sub-unitsandacommonbeta-subunit.J.Exp. Med. 158:1785-1803.
26.Bach, R.,Y.Nemerson,and W.Konigsberg. 1981. Purification and char-acterizationof bovine tissue factor.J.Biol. Chem.256:8324-8331.
27.Chapman,H.A.,0.L.Stone, and Z. Vavrin. 1984. Degradation of fibrin andelastin by intacthumanalveolarmacrophages in vitro: characterization of a plasminogen activatoranditsrole inmatrixdegradation.J.Clin. Invest. 73:806-815.
28. Hunter, W. M., and F. C. Greenwood. 1962. Preparation of iodine la-belledhuman growth hormoneof high specific activity. Nature(Lond.). 194:495-496.
29. Harlow, E., and D. Lane. 1988.Antibodies:a Laboratory Manual. Cold Spring Harbor Press, Cold Spring Harbor, NY. 553-612.
30. Barrett, A. J. 1980. Fluorimetric assaysforcathepsinB andcathepsin H with methylcoumarylamide substrates.Biochem.J. 187:909-912.
31. Barrett, A. J., and H. Kirschke. 1981.CathepsinB, cathepsin H, and cathepsinL.Methods Enzymol. 80:535-561.
32. Burnett, D., J. Crocker, and A. T. M. Vaughan. 1983. Synthesis of cathep-sinB by cellsderived fromtheHL60 promyelocytic leukemiacellline.J. Cell. Physiol.115:249-254.
33. Colton,T. 1974.Statistics in Medicine. Little,Brown & Company, Bos-ton, MA. 128-143.
34. Vassalli, J-D., D. Baccino, and D. Belin. 1985. A cellular binding site for theM,55,000form ofthe human plasminogenactivator,urokinase. J. CellBiol.
100:86-92.
35. Chang,J. C., M. Lesser,0.H. Yoo,andM.Orlowski. 1986.Increased cathepsinB-like activity in alveolar macrophages and bronchoalveolar lavage fluid fromsmokers. Am. Rev. Respir. Dis. 134:538-541.
36. Reilly, J.R., R. W.Mason,P.Chen,L. J.Joseph,V. P.Sukhatme, R. Yee, andH. A.Chapman. 1989.Synthesis andprocessing of cathepsin L,anelastase, byhumanalveolarmacrophages. Biochem. J. 257:493-498.
37. Anderson, D. C., L. J. Miller, F. C. Schmalsteig, R. Rothlein, and T.A. Springer. 1986. Contributions of theMac- Iglycoprotein familyto adherence-de-pendent granulocyte functions: Structure-functionassessmentsemploying sub-unit-specific monoclonal antibodies. J. Immunology. 137:15-27.
38. Miller, L. J., D. F.Bainton,N. Borregaard, andT. A.Springer. 1987. Stimulated mobilization ofmonocyte Mac- 1 andp50,95 adhesion proteins from anintracellularvesicular compartment to thecell surface.J.Clin.Invest. 80:535-544.
39. Rosmarin, A. G., S. C. Weil, G. L. Rosner, J. D. Griffin, M. A.Arnaout, and D. G.Tenen. 1989. Differential expression of CD1 lb/CD18 (Mol), and myeloperoxidasegenesduring myeloiddifferentiation. Blood.73:131-136.
40. Wilson,E. L., G.E. Francis. 1987. Differentiation-linked secretion of urokinase and tissue plasminogen activators by normal human hemopoetic cells. J. Exp. Med. 165:1609-1623.
41.Wilson,E. L., P. Jacobs, and E. B. Dowdle. 1983.The secretionof plas-minogen activatorsby humanmyeloid leukemiccellsin vitro.Blood. 61:568-574.
42.Kirchheimer,J.C.,J.Wojta, G. Christ, and B. R. Binder. 1987. Prolifera-tionofahumanepidermaltumorcellline stimulated by urokinase. FASEB(Fed.
Am.Soc.Exp.Biol.)J. 1:125-128.
43.Gudewicz,P.W., andN.Gilboa. 1987. Human urokinase-type plasmino-genactivator stimulates chemotaxis ofhumanneutrophils. Biochem.Biophys. Res.Commun. 147:1176-1181.
44.Fibbi, G.,M.Ziche,L.Morbidelli,L.Magnelli, andM.Del Rosso. 1988. Interaction of urokinase with specificreceptorsstimulates mobilization ofbovine adrenal capillary endothelial cells. Exp. Cell. Res. 179:385-395.
45. Del Rosso, M., M. Pucci, G. Fibbi,andG. Dini. 1987. Interaction of urokinase andspecificreceptorsabolishes the timeof commitmenttoterminal differentiation of murine erythroleukaemia (Friend) cells. Br. J. Haematol. 66:289-294.
46. Sullivan,L. M., andJ. P.Quigley. 1986. Ananticatalytic monoclonal antibodytoavian plasminogen activator: itseffectonbehaviorof RSV-trans-formed chick fibroblasts.Cell.45:905-915.
47.Glass,W. F. II., R. A.Radnik, J.A.Garoni, andJ.I.Kreisberg. 1988. Urokinase-dependent adhesion loss and shape change after cyclic adenosine monophosphate elevation in culturedratmesangialcells. J. Clin. Invest. 82:1992-2000.
48.Amici, C.,A.Benedetto, 0.Saksela,E.-M. Salonen, and A. Vaheri. 1989. Plasminogen activatorandits enhancement indifferentiatingmouse Friend erythroleukemiacells. Int.J. Cancer.43:171-176.
49.Metcalf,D. 1985.The granulocyte-macrophage colony-stimulating fac-tors.Science(Wash. DC.). 229:16-22.
50.Haskill, S., C. Johnson,D.Eierman, S. Becker,and K. Warren. 1988. Adherence inducesselective mRNA expression of monocyte mediators and proto-oncogenes.J. Immunol. 140:1690-1694.
51. Kaplan, G., and G. Gaudernack. 1982. In vitro differentiation of human monocytes:differences inmonocytic phenotypeinduced by cultivation on glass orcollagen.J.Exp.Med. 156:1101-1114.
52.Fuhlbrigge,R.C.,D.D.Chaplin,J.-M.Kiely,and E. R. Unanue. 1987. Regulation of interleukinIgeneexpression byadherence andlipopolysaccharide. J. Immunol. 138:3799-3802.
53.Eierman,D.F.,C. E.Johnson,and J.S. Haskill. 1989.Human monocyte inflammatory mediatorgeneexpression isselectively regulated by adherence sub-strates.J.Immunol. 142:1970-1976.
54.Francis, C. W., andV. J. Marder. 1987.Physiologic regulationand
patho-logicdisorders offibrinolysis.Hum.Pathol. 18:263-274.
