with a monoclonal antibody Identification and ...

4 downloads 0 Views 1MB Size Report
Nov 4, 1982 - M. Parks, Roger D. Gingrich, ...... heman- gioendothelioma from the mouse did not. It therefore seems that the. HEC-I .... JR: An enzyme. Semin.
From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

1985 66: 816-823

Identification and characterization of an endothelial, cell-specific antigen with a monoclonal antibody WM Parks, RD Gingrich, CE Dahle and JC Hoak

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

Identification

and

Characterization

Antigen By William The

purpose

antibodies

of

to

components

assumption the

these and

unique

to

was

to

characterize the

With

M. Parks,

studies

identify

use

membrane

endothelium.

Our

such components may perform functions of the endothelium and,

specialized

with

antibody

study

their

function

further

probes. and

D. Gingrich,

monoclonal

is that

identification

a Monoclonal

Roger

plasma

vascular

of an Endothelial,

we

may

structure.

some of by their

be

able

Thus,

to

primary

cultures of human umbilical vein endothelium were used to immunize mice whose spleen cells were fused with the mouse myeloma cell NS-1 . HEC-1 #{149}is a monoclonal antibody derived from such a fusion that appears to react with an antigen located only on endothelial cells. The antigen has been characterized by immunoprecipitation and poly-

I

T

IS

NOW

accepted

important rosis.2

role

Initiated

by

the

prostaglandin

pathways

have

largely

focused

products

of

that

the

in hemostasis,

this

discovery

or suspected Important

special in this

to

platelet

function,6 Thus,

and

endo-

pathological

to endothelial

humonal very

cell

has been shown as, for example,

motility,7

primary

interfacial

or tumor

to in

elements from the interstitium, likely makes important contributions

the

function.

monoclonal antibodies. and characterization

to

be a receptor

by Grune

a 1985

cell

membrane

Cells

and

5 x

Supported for Research

Heart, the

This paper describes the of a human endothelial

Presented

Institutes

in part Federation

November

4, 1982,

ence

American

Submitted

Address Medicine, ©

I 985

March

reprint

Medicine,

No. HL-14320 and HL-27561

R.D.G.

Physicians.

ofHealth

and

Clinical at the

Heart 12,

is a Fellow

1984;

requests

and Grant

Research

Fifth

National

accepted

to Dr Roger

University

oflowa

Hospitals.

by Grune

& Stratton,

Inc.

29, City.

A.

of by

HL-07344.

Section Chicago. Nov

D. Gingrich, Iowa

John

of the Illinois.

Thrombosis

in Dallas, March

Center National

supported

T32

in

the

Scholar

was No.

and

ofthe

Research

of the Midwest

Association

City,

Burlington.

(Specialized from the

W.M.P.

training

at the meetings for

Iowa

of Vermont,

is a Teaching

of

0006-4971/85/6604-0012$03.OO/O

816

of

Institute. and

College

American ofthe

College

The University

and Blood

Foundation

American

National

Iowa

in part by Grants in Atherosclerosis),

Lung

Hartford

of

of Medicine,

Confer16, 1982.

1985.

Department IA

& Stratton.

antigen

52242.

lines

of evidence

Inc.

detected

by a monoclonal thus

far

antibody-

appears

specific

for

The mouse myeloma cell line P3/N5I/ was used for fusion and was maintained in culture at

cell

(NS-1) l0

cells

of

culture.

per

milliliter

in

RPMI

1640

containing

bovine serum (K. C. Biological, Lenexa, L) penicillin and streptomycin (GIBCO,

Kan), Grand

ration

vein

and

culture

of human

umbilical

10%

fetal

glutamine (4 mmol/ Island, NY). Prepaendothelial

cells

were

performed as described previously.’0 For most studies, endothelial cells were primary cultures of confluent monolayers at 5 to 7 x 106 cm2

tissue

culture

flask.

The

monolayer

was

detached

by

two washes in phosphate buffered saline (PBS) followed by a brief incubation at room temperature with PBS containing 0.1% EDTA and 0.2 mol/L of urea.” The cells were collected on ice, centrifuged, and washed twice with PBS. Human umbilical vein fibroblasts, umbilical artery smooth musdc and adult endothelial cells’2 were prepared as described previously

The HEC-l antibody was previously referred to as A23.5 in an abstract: Clin Res 30:730A, 1982 (abstr). From the Cardiovascular Center and the Department of MediUniversity

several

endothelium.

and

grown

in M-199

with

20%

fetal

bovine

serum

(FBS).

The fibroblasts and smooth muscle cells were used in the seventh to the twentieth passage at confluence, whereas the adult endothelium was used in the third to fifth passage. HM I 1-22-36 is a mouse hemangioma

carcinoma

Department

transferrin.

HEC-1-which

that

promyelocytic

cine,

for

reported in this paper indicate that this is not the function of the HEC-1 antigen. These data show that monoclonal antibodies can be used to identify and characterize membrane components of the vascular endothelium. Moreover, these probes can be used to inquire about the structure and function of the antigen with which they react.

cells/75

separating

In this study, we made the assumption that membrane components, possibly unique to endothelium, might provide these special functions. Thus, the initial task was to identify such components, and to do this we used hybnidomagenerated identification

acrylamide gel electrophoresis as a glycoprotein with a mol wt of 1 80.000 daltons under nonreducing conditions and 90,000 daltons under reducing conditions. Despite a close resemblance to a membrane component shown by others

l-Ag4

metasta-

structure

and John C. Hoak

METHODS

of the

normal

Chris E. Dahie,

studies

functions of the endothelial cell.4’5 regard are observations in other biologi-

the

the cellular and plasma membrane

recent

the

It is becoming clear, however, that may not fully account for the observed

granulocyte

as

of

Antibody

designated

the contributions

cal systems in which the cell membrane play a critical role in cell-cell interactions sis.8’9

ago

cell,

an

atheroscle-

a decade

in the endothelial

pathway

function.3 processes

plays and

almost

on elucidating

thelial cell intracellular

endothelium

thrombosis’

Cell-Specific

minimal

grows

either

leukemia

line.’5 They essential

in vitro

line,’4

were grown

medium,

or in vivo.’3

and

H.Ep

in RPMI

respectively.

with 10% FBS. Bovine pulmonary obtained from the American Type

is a human

1640 and

Each

artery Culture

HL-60

2 is a human was

endothelial Collection,

laryngeal

Dulbecco’s

supplemented

cells were Rockville,

Md.

Peripheral blood cells were obtained from freshly drawn blood anticoagulated with 3.8% sodium citrate. Red cells were sedimented with 3% dextran (mol wt 500,000; Sigma, St Louis), and the supernatant was layered onto a density gradient (Ficoll-Paque, Pharmacia, Piscataway, Ni) and centrifuged at 400 g for 45 minutes. Granulocytes with a purity of 95% (by morphology) were removed from beneath the gradient, and mononuclear cells were taken from the interface of the gradient and the plasma. In some studies, the mononuclear cell fraction was used in its entirety whereas in other studies the monocyte-macrophage subfraction was obtained by incubating the entire cell fraction in a plastic Petri dish at 37 #{176}C for 60 minutes. The nonadherent cells were washed away by pipetting PBS over the dish surface, and the adherent monocytemacrophage cells were scraped up by a rubber spatula. Platelets were obtained as previously described.’6 After the final wash, the platelets were suspended in Tyrode’s solution with 10 mmol/L of

Blood, Vol 66. No 4 (October). 1985: pp 8 16-823

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

ENDOTHELIAL

EDTA

CELL-SPECIFIC

at a concentration

blood

cells

were

817

ANTIGEN

of 2 x i0

collected

from

the

platelets

cell

per milliliter.

pellet

after

Red

removal

V8 protease

of the

platelets. They were washed twice with buffered saline and brought to a final concentration of 2 x 10 cells per milliliter. Monoclonal antibody production. BaIb/c mice were hyperimmunized with primary culture human umbilical vein endothelial cells. Their spleen cells were fused with the NS- 1 cells at a ratio of ten spleen cells to one NS-1 cell using 50% polyethylene glycol I ,500 (Aldrich Chemical Co, Milwaukee, Wis). The fused cells were dispensed into 96-well plates, and the culture supernatants were tested for antiendothelial cell activity eight to I 2 days later by the indirect,

whole

cell

radioimmunoassay

with

7 x

I 0 viable

0.5%

bovine

serum

albumin,

0.2%

sodium

and

point

were

measured

and

the

mean

and

was

adsorption

to the cell

vein

endothelial

cells

were

labeled

cultures after

four

days

with

minutes.

The

soluble

fraction

Approximately

was 100

clarified tL

by addition

of cell

lysate

of 0.5 was

incubation,

five

minutes

25 mmol/L

and

ofTris

washed HC1

(ph

three 7.6),

times 0.3

with

mol/L

and

cocktail

(Budget-Solve,

Research

Products

was

in

sodium

electrophoresed running

restarted,

the

transfernin

at 25 zCi/

presence

(Sigma)

I zg of protein.

incubated

with

presence and

and

unlabeled absence

of the

polyacrylamide

were analyzed

mL

until

gel.

the

After

and

the

of

‘25I-human

a 90-

electrophoresis

This

was

iodinated diluted

g/l mL or of the diluted vein

HEC-I

antibody. were

trans-

using with

106

chloram-

PBS

cpm

and

per

used

milliliter.

‘251-transfernin

were

cell

in the

endothelial

gel electrophoresis

lysate

Immunoprecipitation then

done,

and

the

gels

by autoradiography.

Immunohistological sectioned,

was

umbilical

and

Fresh

technique. allowed

to dry

on glass

tissues

slides

were frozen,

overnight

cryo-

at 4 #{176}C. They

were then fixed in acetone for five minutes at room temperature followed by two washes in PBS containing 0.5% BSA. The slide was covered

with

antibody

for

30 minutes

PBS containing 0.1% BSA, covered (FITC) conjugated goat antimouse 4 #{176}C, and

then

30% glycerol fluorescence

washed

three

in PBS beneath

at 4 #{176}C, washed

Ig (Cappel)

times.

The

glass

two

fluorescein

with

in

for 30 minutes

sections

coverslips

times

isothiocyanate were

covered

and examined

at with

with a

microscope.

Cell fusion five monoclonal human of

,

results. antibodies

umbilical

these

vein

antibodies

cell types, monoclonal

lony The

immunodiffusion, specificity of

indirect

From one cell fusion showing a high level endothelial

also

other This

whereas antibody,

cells

showed

were

experiment, of binding

to

obtained.

Four

reaction

with

significant

one, designated HEC-1 , did not. shown to be an 1gM by Ouchten-

reacted only with endothelial cells. binding was tested by the whole cell

nadioimmunoassay

and

microscopy.

% of Control

The

was

reactivity

confirmed

by

of HEC-l

indirect

was

com-

Binding

(xlO2)

3

9

40

30

20

vol 2%

10

in

I

27 Reciprocal

2 mmol/L

International,

Ill) was added, and the amount of radioactivity slice was determined by scintillation counting. Prospect,

acrylamide

2%

and 50 tg/I

251

of PMSF. The precipitates were solubilized with or without 5% 2-mercaptoethanol and analyzed by electrophoresis on 8% polyacrylamide gels. For analysis of precipitated antigen labeled with tnitiated-glucosamine, 2-mm slices of the gel were incubated in 1 mL of PBS at 37 #{176}C overnight. Nine and one-half milliliters of scintillation

the

I 5%

containing

and

was

in a lane slot of

incubated

0. 1% deoxycholate ofNaCl

(Miles)

at a final concentration of0.02 Twenty-five microliter-aliquots

with 250 .tL of tenfold concentrated monoclonal antibody culture supernatant for four hours at 4 #{176}C. Fifty micrograms of rabbit antimouse 1gM (Miles) was added and further incubated at I 8 hours at 4 #{176}C. The precipitate was collected by centnifugation at 9000 G for

buffer

dithiothreitol

protease the current

Human

fluorescence

by lactoperoxidase-glucose oxidase or by incubation for 48 hours with D-[1, 6-H3 (N)] glucosamine HC1 (39.6 Ci per millimole, New England Nuclear, Boston) at 20 zCi/mL of culture medium. Labeled cells were solubilized with 1% Triton X-100 (Sigma) in 25 mmol/L of Tnis HC1 (pH 7.6), 2 mmol/L of iodoacetamide and 2 mmol/L of phenylmethylsulfonyl fluoride (PMSF) at 4 #{176}C for 30 deoxycholate.

reached

with

meth-

RESULTS

umbilical

ofculture

V8

published of the antigen

gel. This was placed

20 mmol/L

Immunoprecipitation

standard

of human

(SDS),

f errin.

15

surface.

Primary

Immunoprecipitation.

overlaid

to

completed.

ine-T

error calculated. Cells that reacted with the fluorescent second antibody were suspended in PBS with 30% glycerol and 0.1% paraphenylenediamine and examined with a Leitz Orthoplan Fluorescent microscope.’7 The mouse monoclonal 1gM proteins W6/ 18 and TEPC 183 IgM.K (Bionetics Laboratory Products, Kensington, Md) were used in each experiment as the negative control. Both of these antibodies in the radioimmunoassay consistently reacted at two to three times the PBS control, indicating a low level of nonspecific

gel,

sulfate

front

minute

izCi/ig protein) or fluorescein isothiocyanate (Cappell Laboratories, West Chester, Pa). In the case of the nadioimmunoassay, of each

sample

dye

U/I mL of hepanin and dispensed into wells of a U-bottom 96-well plate (GIBCO). Monoclonal antibody was added and incubated for 60 minutes with agitation at 4 #{176}C. The cells were then washed twice with PBS containing 0.1% bovine serum albumin, 0.2% sodium azide, and 1 5 U per milliliter of heparin. Cell-bound antibody was detected by reaction with a second antibody, rabbit anti-mouse IgG (heavy and light chains) labeled with either 251 (iodinated at 25

tniplicates

According

a stacking staphylococcus

human

azide,

mapping.

immunoprecipitate

dodecyl

endothelial cells per well as targets. Cells from wells showing activity were expanded, cloned twice by limiting dilution, and frozen. Antibody isotype determination. Immunodiffusion in 1% agarose against rabbit antimouse heavy chain-specific antisera (Miles Laboratories, mc, Elkhart, Ind) was used to determine the isotype of the monoclonal antibodies. Whole cell, indirect, antibody binding assay. Fresh, viable cells were tested for reactivity with monoclonal antibodies by immunofluorescence and radioimmunoassay. Cells were suspended in PBS containing

peptide

ods,’9 the ‘25I-labeled, reduced cut from a dried polyacrylamide

Mt

in each

Fig

1

.

Titration

of

antibodies

81 Dilution against

243

729

endothelial

2187

cells.

The

monoclonal antibody HEC-1 and the two antibodies used as negative controls (TEPC 183 IgM.K, W6/1 ) were reacted with human umbilical vein endothelial cells in the indirect, radioimmunoassay. Binding is expressed as a multiple of the value obtained when phosphate-buffered saline with 0.5% bovine serum albumin was used in place of antibody. In each assay. the number of target cells was the same for each antibody.

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

PARKS

818

pared

in all experiments

proteins.

to two other

TEPC-183

known

antibody

clonal

antibody

A antigen

is an

specificity, with

and

very

described

in this

rapidly

with

paper,

endothelial

from

vascular

cells tissue,

findings showed

except

monoclonal

toward

the

group

I shows

that

these

concentrations cells, with

dilution.

For

with blood,

of the

studies

antibodies

were

was used to human

undiluted. umbilical

HEC-1 of HEC-I

compared

reacted nontheir binding

most

control

1gM

protein without a is an 1gM mono-

Fig

peripheral

a variety and

of cells

used

derived

cell culture.

All of

were confirmed by fluorescence microscopy, no specific reaction with any of the cell types

for the endothelial

showed

cells.’8

the two

at a dilution of 1:50, whereas Figure 2 shows the reactivity vein

mouse

myeloma the W6/1 directed

red blood

antibodies at high with endothelial

dropping

these which

ic

a specificity

on human

two control specifically

1gM

that

cells.

HEC-l

Fluorescence

reacted

with

microscopy

endothelial

also

cells

from

bovine pulmonary artery and adult human vena cava aorta’2 (Fig 3), but did not react with the subfnaction mononuclear cells adherent to plastic; neither was there reaction

to the cultured

Frozen

section

and

mining

bovine

in vivo.

was

obtained

hemangioma

cell Having

with an antigen expressed culture, as well as passaged sources,

whether

cells

mouse

immunofluorescence.

HEC-1 reacted cells in primary adult

ET AL

the

A grossly at

we were

antigen

normal

surgery

then

was

line. shown

that

on endothelial cells from both

interested

expressed

portion

from

and of any

in deter-

on

endothelial

ofa

superficial

a 23-year-old

man

leg vein who

had

undergone vein stripping for varicose disease. A temporal artery biopsied for diagnostic reasons but showing no inflammation

was

with

also

HEC-1

cence reacted both

These

control

antibodies

microscopy. specifically the

vein

tissue (4A no specific internal

and

artery

but

sectioned

and

not

and

examined

reacted

by fluores-

shown in Fig 4; HEC-1 cells lining the lumen of with

the

elastic

of

lamina.

enucleated

autofluonescence

Sections

because 01

l9M Control

of

taken

trauma,

Antibody

W6/1 show a

produced

from a

kidney

the

by

retina

with

Fig 3. thelial

subendothelial

and B), whereas with the control antibody tissue reaction was seen (4C). All sections band

% Binding

were

The results are with endothelial

subendothelial

eye

obtained.

and

lmmunofluorescence

cells.

Single

cell

of bovine

suspensions

artery (A) or human adult aortic with HEC-1 and a fluorescein anti-mouse

1G. (Original

of an

minimal

change

disease,

(x1O2)

tissue

studies

antigen endothelial

5

their

have

not

of large,

reacted cells

flasks,

been

Figure

#{149}i.iI. Mono.

Plat.

RBC

HL-60

cipitation

H Ep -2

Fig 2. Reaction of HEC-1 to various types of cells. HEC-1 was reacted with various target cells. and the extent of binding was determined by reaction with lSSlrabbit anti-mouse lgG. Each target cell served as its own control, and the amount of HEC-1 binding

is expressed

control

antibody

target

cells was

as a multiple

TEPC

183

the same

lgM.K.

for both

of the

binding

In each antibodies.

assay.

of

the

the

negative

number

of

5 shows and

it is apparent

and

small

that

vessels,

and

the

an autoradiograph

labeled

These lysate from

that

the

cells reacted such HEC-1

with

were

of the

as well the

human umbilical days were detached and

oxidase.

demonstrates

melanoma

the HEC-1 antigen. In order to characterize

thoroughly,

detergent,

a vascular

of the endothelium Although exhaustive

completed,

with by HEC-1, in culture for four

washed

in nonionic

and

staining channels.

middle-sized,

lactoperoxidase-glucose

Gran.

reacted goat

x400.)

placenta,

specific vascular

as ofcapillaries, expresses Antigen characterization.

S M.

endo-

pulmonary

endothelial cells (B) were isothiocyanate-conjugated

magnification

human

tumor also showed capillaries and other endothelium

Fib

adult

bovine

the

10

H U V E

and human

of cultured

vein from

1251 using

then

with

lysed

HEC-1.

an immunopreantigen

had

a

mol wt of -180,000 daltons under nonreducing conditions. However, under reducing conditions, it had an apparent mol wt of 90,000 daltons. It therefore appeared that the antigen in its

native

state

was

composed

of two

disulfide

bonded

subunits of equal mol wt at 90,000 daltons. The smaller size of the band seen under nonreducing conditions is likely secondary to a lower solubility of the unreduced antigen in

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

ENDOTHELIAL

Fig

human normal

4.

CELL-SPECIFIC

Indirect

ANTIGEN

immunofluorescence

vein and artery. adult leg vein

819

of

adult

Frozen sections of grossly or temporal artery were

reacted with HEC-1 (A and B) or W6/1 (C. vein only) and a fluorescein isothiocyanate-conjugated goat anti-mouse Ig. (Original magnification lumen; M. endothelial cell membrane; elastic lamina.

x4.00.) 1, EL, internal

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

PARKS

820

ET AL

Origin

I

92

200

68

BPS

43

. 4

#{149} Reduced .

Unreduced

3 Percent of Background (xlO )

0

5

10

20

15

Gel

Fig

6.

HEC-1 HEC-1

Electrophoretic

25

Slice

analysis

30

40

35

50

45

Number

of

3H-o-glucosamine-labeled

antigen. Immune precipitates with 3H-D-glucosamine-Iabeled

formed endothelial

by incubation of cell lysate were

analyzed with (#{149}) and without (#{149}) reduction on an 8% acrylamide gel. Two-millimeter gel slices were incubated in 1 ml of phosphate-buffered saline overnight at 37 ‘C, mixed with 9.5 mL of scintillation fluid. and counted in a scintillation spectrometer. Activity

is

expressed

as

a

percentage

of

background

counts

obtained

from lanes in which precipitates formed by reaction with TEPC 1 83 IgM.K were analyzed. Molecular weight markers were run in parallel lanes and their positions are indicated at the top.

BPB. bromophenol

blue.

A number of laboratories molecule with very similar behaves the fact

B Fig 5. Electrophoretic analysis of iodinated HEC-1 antigen. HEC-1 immunoprecipitates of iodinated endothelial cell membrane Iysate were analyzed on an 8% polyacrylamide gel in the presence of sodium dodecyl sulfate. Lane A. HEC-1 precipitate without reduction; lane B. HEC-1 precipitate run in the presence of 5% 2-mercaptoethanol. Molecular weight standards were run in a parallel lane, and their positions are indicated on the left: a. 200,000; b. 1 1 6.000; c, 92.000; d. 68,000; and e, 43.000 daltons, respectively.

loading

buffer

with

some

of the antigen

not entering

the

top

on the gel. Because many functionally tides are glycosylated, we whether or not the HEC-l Primary were

cultures

of human

metabolically

hours,

then

umbilical

labeled

detached

important membrane polypepwere interested in determining antigen was a glycoprotein. with

from

their

and lysed with nonionic detergent. analysis of the precipitates on carried out as before, except that counting

2-mm

slices

spectrometer. The under nonreducing appeared in

the

at presence

that

the

show dimenic

the that

daltons. 2-mercaptoethanol, at

90,000

a

dimenic

has

previous the

polypeptide

study

molecule in the

with

lane

the

in

thoroughly,

however, again

structure. the state.

iodinated

was the

6;

peak

of

antigen

we sought possibly

to determine be

an

and

whether

endothelial

the

cell

HEC-1

and the rapidly,

antigen

receptor

for

could

transfernin.

Several experiments were done to determine this possibility. Whole cell radioimmunoassays and fluorescence microscopy were performed on primary cultures of human umbilical

vein

endothelium

using

HEC-I,

and

that

binding

was

compared with the reaction of two monoclonal antibodies the transferrin receptor, B3/252’ and OKT9.#{176} In contrast the strong reaction of HEC-1 with endothelial cells,

to to the

B3/25 shown).

not not

react high with

and OKT9 antibodies showed In addition, as previously

no reaction noted, HEC-1

(data did

with HL-60 cells, and this cell line is known to express levels of the transfernin receptor and reacts strongly B3/252’ and OKT9.2#{176}

A second

set

HEC-1

of experiments

labeled antigen.

demonstrate ton indeed

the

methodology

nonprecipitated

was

lanes

D and

the

used

and

E,

to

unlabeled initially

to

transfennin recepFigure 7 shows an

a number ofcontrols. cell extract, whereas

respectively. These and show the HEC-1 In

to attempt

with

the putative membrane with that molecule.20’2’

tated cell extracts, reducing conditions daltons.

designed

transfernin

study including iodinated whole

the

C show

was

human

This

that reacted

autoradiographic A represents

-90,000

results

disulfide-bonded

cell surface is associated with cell proliferation primary cultures of endothelial cells do not replicate

B and run

suggesting

These

as membrane receptor for transferrin.#{176}22 Despite that the expression of the transfernin receptor at the

coprecipitate

a scintillation

precipitate

daltons,

is a glycosylated, native

washed

study are shown in Fig a peak of radioactivity

When

of

cells for 48

Immunoprecipitation and polyacrylamide gels were the gels were analyzed by gel

appeared

antigen

confirm

each

endothelial

glucosamine

flasks,

results of this conditions,

180,000

radioactivity

of

vein

tnitiated

have described a cell membrane biochemical characteristics that

the

HEC-1

Lane lanes precipi-

lanes are run under antigen in lane C at immunoprecipitation

conditions were identical to those of lanes B and C except that the precipitation took place in the presence of unlabeled transfernin fernin

could

(2 mg/mL) block

the

to determine precipitation

whether of the

excess HEC-1

transantigen.

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

ENDOTHELIAL

CELL-SPECIFIC

821

ANTIGEN

a-

-.

b-

I

C.-

d-

e-

#{149}1 o

ABCDE

I

FGH

Fig 7. Reaction of the HEC-1 antigen with human transferrin. Lanes A through I show the effect of the presence of transferrin on HEC-1 -induced immunoprecipitate formation from either iodinelabeled or nonlabeled endothelial cell lysate. The precipitates were analyzed on an 8% polyacrylamide gel in the presence of 2% sodium

dodecyl

sulfate

graphic

results

are

membrane

lysate;

precipitated

and

5%

shown.

lanes

fractions.

2-mercaptoethanol.

Lane

B and

A.

C.

respectively.

Autoradio-

iodinated

endothelial

and

of

endothelial

the

iodinated

I

cell

nonprecipitated

HEC-1

cell lysate; lanes D and E. results of the same reaction except that transferrin at 2 mg/mL was present in the reaction mixture; lanes F and G nonprecipitated and HEC-1 precipitated fractions, respectively, of nonlabeled endothelial cell membrane lysate occurring in the presence of iodinated transferrin 1 0 cpm per milliliter; lanes H

and I, same as lanes F and G except that TEPC 1 83 IgM.K was used in place of HEC-1 in immunoprecipitation. ers were run in a parallel lane and their

the left and correspond

No

blocking

Lane

could

E shows

demonstrated

antigen

under

at 90,000

these

daltons,

indicating

no significant inhibition has taken place. Lanes the nonprecipitated and the HEC-l precipitated cell extracts

with

the precipitation

of iodinated transfernin. fennin present in the whereas failed

lane

G shows

to copnecipitate

I show

that

to lanes

in place

of the

the

HEC-1

done

G except antibody.

that Not

under

TEPC-183 shown

are

H and

gent,

and

obtained

The

HEC-l from

membranes and

endothelial

were

OKT9 and

gel as shown

cells,

digested were

in Fig 8. The

dithiothreitol. of the

The

used addi-

the amount judgment indirect,

analyzed

antigen

The

with staphylococanalyzed on a I 5% profile

of

cells

of the being

per endothelial

of

autora-

blue.

immunoprecipitate) profile of HEC-l

in each

of OKT9 antigen is based on relative

mmol/L gel;

bromophenol

endothelial

staphylococcal

of 20

polyacrylamide

(OKT9 than the

immunofluonescence

cell

per HL-60 fluorescence

from antigen.

two

profiles

iodinated as compared cell.

is

as well with

The latter intensity on

microscopy. DISCUSSION

After were

BPB.

with

presence

a 1 5%

in radioactivity to fewer

digested

in the on

and deter-

gel.

zg/mL

gel is shown.

difference HEC-l

with

50

and

as less

two

polyacrylamide

at

identical

respectively.

polypeptide

8%

was

immunoprecipitates HL-60

reduced form of each antigen was cal V8 protease and the polypeptides acrylamide

solubilized

an

protease

due

Finally, both human umbilical vein endothelial cells HL-60 cells were iodinated by the lactoperoxidase/glucose method.

from

V-8

partly

tional lanes in which ‘25I-transfernin and unlabeled transfernin were shown to migrate to the identical position in the gel as the tnansfenrin molecules noted in lanes F and H.

oxidase

cut

the transfernin receptor HL-60 cells is different

antigen

Lanes

Fig 8. Staphylococcus V-B protease peptide maps of the HEC-1 and OKT9 antigens. The reduced. iodinated HEC-1 antigen from endothelial cells and the OKT9 antigen from HL-60 cells were

diograph

transdaltons,

HEC-1

transfennin.

analyses

F and

in the presence

the iodinated at -80,000

unlabeled

the iodinated

immunopnecipitate

conditions

F shows mixture

that

F and G show nonlabeled

performed

Lane reaction

conditions.

B

A

to those in Fig 5.

be

the

Molecular weight markpositions are indicated on

nearly

a decade

laboratories,

there

lism

the

within

when

released

intravascular evidence, products

is no

of intensive question

endothelial

from thrombosis

while not in hemostatic

the

study

that cell

cell, and

diminishing regulation,

produces

have

in a number

anachidonate substances

that,

a significant

hemostasis.”23

of

metabo-

effect More

the importance suggests that

on

recent

of these additional

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

822

PARKS

factors

may

contribute

to the

endothelium.4’5’24 Indeed, cell membrane-associated sin-converting

enzyme,25

Ia-like

determinants regulating that the plasma membrane play

a greaten

of that

than

tissue.

suspected

been

used

antigen.2t’28 In this monoclonal thus fan

HEC-

hum. from

of the

tially

antigen

a way

delineate

have

that reacts on vascular

in some

the

described

possible

of identifying

have,

mdi-

the

of

the

derivation

of

with a membrane endothelium. The

after

careful

1 appeared

Furthermore, human umbilical

aorta and vena pulmonary artery gioendothelioma

screening

to react

on other

specifically

a

antigen HEC-1’

primary veins,

types

with

cultures passaged

from at extent

tissue

that

of fresh

antigen did not arise the HEC-l antigen endothelium the

human

the

Our goal in these studies

topoietic behaved

by

cule has expressed

polyacrylamide

been shown exclusively

transferrin receptor sylated membrane virtually identical the

to be a receptor on proliferating

of

antibodies. on hema-

molecule antigen That

a receptor between

that when mole-

for transfernin and is cells.2#{176}22Because the

of the transfernin

for transfernin. cell proliferation

receptor

would

make

and

results

present block

in the reaction the precipitation

These

data

possess antigen

suggest

antigen

in

demonstrate

immunoprecipitated did not bring

from down the

mixture, of the that

shown

and

the

that

nor did iodinated cells

membrane

receptor

the data

presented

wt

of 92,000

structures

were

the same

in this

the

it unlikely,

as

both

paper

observation

on

the

mol

and wt

can steric HEC-1

molecule

with that antibody

of on

structural transferrin

indicate

defined previously

estimated

with

1

the protease

there are some antigen and the

the antigen has not

daltons

HEC-

obtained in an identical suggest that the molecules

although the HEC-1

are distinct molecules. To our knowledge, vascular endothelium

do not

the

and/or that the

of the iodinated HEC-1 antigen receptor as defined by the OKT9

are different. Thus, similarities between receptor,

that

by comparing

cells. The two profiles, are quite different and

anti-

endothelial transferrin

in culture

and

transport

structure

HEC-1

excess transferrin HEC-1 antigen.

due to glycosylation we tested the possibility

in tertiary

peptide profile the transferrin

coprecipitation

the

unlabeled iodinated

endothelial

the transfernin

only

transferrin

HL-60 shows a inhibited with

that

a functional transfernin receptor does not react with transfernin.

electrophoresis

and the HEC-1 antigen are both glycopolypeptides with a dimeric structure and mol wt, we were led to consider whether

HEC-1 antigen was The known relationship

expression

or in

the function

electrophoresis.

our

the

that

by

without

without

they

by HEC-1 on the been described.

Kaplan, et al3#{176} have reported on two monoclonal that define an endothelial cell polypeptide antigen

study.

by the primarily

a membrane like the HEC-l gel

vessels,

under

is to characterize

delineated very much

the

culture. Whether on the vascular

on diseased

identified laboratories

from

indicated

is currently

components done in several

cells had in a fashion

analyzed

body,

of all tumors

the membrane Earlier work

tissue

from conditions ofcell is distributed uniformly

throughout

vasculature

adult

transfernin.

when lysate,

HL-60 fashion,

bovine hemanseems

that the HEC-I antigen is expressed on endothelium least one other species and is not lost to any great when cells are passed in culture. The positive reaction of HEC-1 with the endothelium sections

Furthermore,

differing

cells adult

to express

unlabeled

It has been

of cells,

cava, and passaged cultures from bound HEC- 1 , whereas a cultured from the mouse did not. It therefore

known

experiments, we have also shown only a low level of noncompetitive

transferrin, whereas that is competitively

be nonfunctional changes.29 Thus,

the endothe-

of endothelial cultures from

cells

binding of ‘251-human high level of binding

gen, cell

instances,

function

HL-60

experiments,

functions

made

growing

receptor.2’ In preliminary that endothelial cells show

one of five antibodies from a single fusion all of which reacted strongly with endothelial

However,

only

probes

and

to

we

antibody found only was

provide

tool

paper,

antibody experiment, cells.

a

and

in the specialized

antibody

components

as

antigens,26

role

technology

membrane

role

lymphocyte migration27 indicate of the vascular endothelium may

Monospecific

by hybridoma vidual

thromboregulatory

the demonstration of endothelial anticoagulant activity,5 angioten-

ET AL

antibodies with a mol

polyacrylamide reduction.

reduction

The as

well

gel latter as

the

description of significant antibody reaction with fibroblasts indicate that the E92 antigen described by Kaplan et al is not the same as the HEC-I antigen. No other well-defined antigen of the vascular endothelium modulin (mol wt 74,#{216}#{216}#{216}),hIangiotensin (mol wt 140,000),32 antithrombin III thrombospondin actenistics can

(mol of the

wt I 7S,000)

HEC-1

including thromboconverting enzyme (mol wt 62,300),” or

corresponds

to the char-

antigen.

In conclusion, we have shown that monoclonal antibodies be used to identify and study membrane components on

the endothelium. The antibody HEC-1 identifies an antigen apparently specific for endothelial cells. This antibody should be useful in further studying the thnomboregulatory

a priori, that confluent, primary cultures of very slowly dividing endothelial cells would express a receptor for transferrin. This proposition was supported in antibody binding

role of the vascular endothelial membrane and as a marker for endothelial cells. As well as examining the function of this antigen, we are pursuing the possibility that the antibody can

experiments react with

be used to assess may be occurring.

which showed OKT9 or B3/25,

ies,20’2’ and,

conversely,

that endothelial cells antitransfernin receptor

HEC-1

does

not react

with

did not antibod-

clinical

states

in which

endothelial

injury

exponenREFERENCES

I . Gingrich Semin 2.

Hematol Ross

R, Hoak

iC:

I 6:208,

1979

R, Glomset

EngI J Med 295:369,420,

JA:

The

1976

Platelet-endothelial pathogenesis

cell interactions. of atherosclerosis.

New

3. Moncada

5, Gryglewski

isolated

from arteries

unstable

substance

263:663,

1976

R, Bunting

transforms that

inhibits

5, Vane

prostaglandin platelet

JR: An enzyme

endoperoxides aggregation.

to an Nature

From bloodjournal.hematologylibrary.org by guest on July 10, 2011. For personal use only.

ENDOTHELIAL

4.

CELL-SPECIFIC

Czervionke

Inhibition

RL,

Smith

of prostacyclin

Correlation

with

ANTIGEN

JB,

823

Fry

GL,

by treatment

platelet

Hoak

JC,

adherence.

i Clin

Invest

DL:

Haycraft

of endothelium

with 63:1089,

aspirin. 1979

Owen WG, Esmon CT: Functional properties ofan endothelial cell cofactor for thrombin-catalyzed activation of protein C. i Biol Chem 257:859, 1982 6. McEver RP, Baenziger NL, Majerus PW: Isolation and quantitation ofthe platelet membrane glycoprotein deficient in thrombas5.

thenia

using

66:1311,

a

monoclonal

proliferation-associated

Hanninen

A, Vuopio

molecular

USA78:4515,

Ruutu

P. Timouen

TTT,

A: Decrease

of the

high

related to cell proliferation is the receptor Acad Sci USA 78:3039, 1981 22. Goding JW, Burns GF: Monoclonal

LC,

P, de Ia Chapelle

weight

monosomy-7

Andersson

surface

glycoprotein

associated

with

i

of human

defective

major

granulocytes

chemotaxis.

Blood

in

54:401,

nizes

the

leukemia

Nature

din

286:267,

1980

and metastasis of blood-borne of plasma membrane vesicles Natl Acad Sci USA 77:399,

23.

on the

confluent

cell

vascular

protein

endothelial

(CSP-60)

hal

cells

or in cell

Sci

USA

76:2306,

12.

Fry

Nghiem

types

G, Parsons

human

monolayers

growing

of

in actively

in multiple

a specific

growing

layers.

Proc

endotheNatI

T, Hoak

vessels.

J, Sage

R: Properties

H,

Gingrich

R, Ercolani

of cultured

Arteriosclerosis

endothelium

L, from

77:941, 14.

Collins

Si,

I 5.

RC,

of human

Nature

istics

Gallo

270:347,

Moore of four

AE,

Gallagher

myeloid

RE:

leukemic

Continuous cells

growth

and

in suspension

cul-

1977

Sabachewsky

permanent

L, Toolan

lines

HW:

of cancer

cells.

Culture Cancer

R: Comparative effect prostaglandin synthesis.

of i

GM: A simple method during microscopy.

of J

18. Williams

Barnstable AF,

by

monoclonal

Natl

Acad

Ci, Zeigler

Bodmer A:

WF,

Brown

Production

of

monoclonal

Smith

cells:

iB,

Haycraft

BC,

Hsu

KC,

Vascular

from circulation J Clin

Das

MA

cells is inducible

ir:

binding

79:6641,

Expression and

of Ia-like

antigens

immunoprecipitation.

determinants

10:556,

K, Pratt

RM,

Proc

cell

Jaworski Proc

surface

G, Milstein

C,

antibodies

to

Chan

V. Chan Thromb

McPherson

endothelial

C, Yamada

KM:

NatI

Acad

Sci

Evidence

transport: USA

antigen.

Arteriosclerosis

76:791,

RL:

Res J, Sage

15:209,

1983

angiotensin-converting

1975

Antithrombin

cells: A natural

1979

L, Sepe E92, an

CT: Isolation of a membraneactivation of protein C. i Biol

Pulmonary

250:6762, PK:

3:403,

for

Specific

D, Cook P, Dalecki M, Rogozinski VP: Monoclonal antibodies to

Soffer

endothelial

endothelium.

ization culture.

M,

J Biol Chem

high

Inhibition of mixed specific for a rat T

in membrane

N, Owen WG, Esmon cofactor for thrombin-catalyzed 257:859, 1982 Das

on

specificity of lymphocyte

1980

carbohydrate

by tunicamycin.

by

Demonstration

in vitro.

M, Mason DW, Williams AF: response by monoclonal antibody subset. Nature 282:841, 1979

Olden

RL:

localization.

1982

organ-specific

J Immunol

Webb

enzyme. human

in

Invest

M, Soffer

endothelial

endothelial

30. Kaplan K, Weber 0, Knowles D, Butler

32.

DL:

of prostacy-

on endothelium.

Gimbrone

USA

with Eur

inhibition

bound Chem

recog-

lymphoblastic

Influence

of thrombin

sites

Seegal

antibody Sci

lymphocyte lymphocyte

34.

G, Galfre

OKT-9

acute

1976

iS,

interaction venules.

33.

Johnson GD, de C Nogueira Araujo reducing the fading of immunofluorescence Immunol Methods 43:349, 1981 17.

Proc Natl

1980

3 1 . Esmon

15:598,

1955

16. Parks W, Hoak iC, Czervionke ibuprofen on endothelial and platelet Pharmacol ExpTher2l9:415, 1981

glycoprotein

antibody

JC,

vascular

enzyme:

vascular

29.

Sci

1981

W: Clearance

PRB,

191:1050, Pober

endothelial

characterRes

cell surface

human

Hoak

binding

role of glycoprotein

1971

differentiation

ture.

Science

28.

4:4, 1984

13. Hoak iC, Warner ED, Cheng HF, Fry GL, Hankenson RR: Hemangioma with thrombocytopenia and microangiopathic anemia (Kasabach-Mernitt syndrome): An animal model. J Lab Clin Med

Acad

for transferrin.

27. Butcher EC, Scollay RG, Weissman IL: Organ lymphocyte migration: Mediation by highly selective

Acad

1979

D, Czervionke

adult

cell

not detected

55:271,

Angiotensin-converting

human

NatI

1980

Caldwell

26. in

to cultured

P, Owen

on

RL,

by high-affinity

25.

Human

127:1256,

Czervionke

Blood

LolIar

MB:

transferrin

J Immunol

GL,

(PGI2).

66:1222,

10. Czervionke RL, Hoak iC, Fry GL: Effect of aspirin thrombin-induced adherence of platelets to cultured cells from blood vessel walls. i Clin Invest 62:847, 1978 I I . Vlodavsky I, Johnson LK, Gospodarowicz D: Appearance

for

adherence

rabbits

I 980

surface

cells. Fry

24.

IS, Omary

receptor

Platelet

Proc

1981

Trowbridge

1979 8. Shearman Pi, Gallatin WM, Longnecker BM: Detection of a cell surface antigen correlated with organ-specific metastasis. 9. Poste G, Nicolson GL: Arrest tumor cells are modified by fusion from highly metastatic cells. Proc

for transferrin.

Invest

antibody.

21.

CG,

receptor

Clin

hybnidoma

1980

Gahmberg

7.

group A erythrocytes, HLA and other human cell surface antigens. Cell 14:9, 1978 19. Cleveland DW, Fischer 5G. Kirschner MW, Laemmli UK: Peptide mapping by limited proteolysis in sodium dodecyl sulfate and analysis by gel electrophoresis. i Biol Chem 252:1 102, 1977 20. Sutherland R, Delia D, Schneider C, Newman R, Kemshead J, Greaves M: Ubiquitous cell-surface glycoprotein on tumor cells is

III

in fresh

anticoagulant

and

from

cultured

the vascular

1979

J, Bornstein

P: Isolation

of a glycoprotein secreted by aortic i Biol Chem 256:1 1330, 1981

and

character-

endothelial

cells

in