Adhesion of endothelial cells and adsorption of serum proteins on gas plasma-treated polytetrafluoroethylene

Adhesionof endothelialcells and adsorptionof serum proteins on gas plasma-treatedpolytetrafluoroethylene A. Dekker,K. Reitsma,T. Beugeling,A. Bantjes,J. Feijenand W.G.van Aken University of Twente, Faculty of Chemical Technology, PO Box 2 17, 7500 AE Enschede, The Netherlands Presented at Biointeractions ‘90, Oxford, UK 2 1-23 August 1990

From in vitro polymers.

experiments

it is known that human endothelial

The hydrophobicity

of vascular

reason why endothelialization

prostheses

cells show poor adhesion

manufactured

to hydrophobic

from Teflon@ or Dacron@ may be the

of these grafts does not occur after implantation

in humans. We modified

films of polytetrafluoroethylene (Teflon@) by nitrogen plasma and oxygen plasma treatments to make the surfaces more hydrophilic. Depending on the plasma exposure time, modified polytetrafluoroethylene surfaces

showed

water-contact

ESCA measurements polytetrafluoroethylene

of 15-56’.

incorporation

versus

96’

for unmodified

of both nitrogen-

polytetrafluoroethylene.

and oxygen-containing

groups

into the

surfaces,

dependent on the plasma composition and exposure time. The thickness layer was -1 nm. The adhesion of cultured human endothelial cells from 20% culture medium to modified polytetrafluoroethylene surfaces with contact angles

of the modified surface human serum-containing of 20-45”

angles

revealed

led to the formation

of a monolayer

of cells, which

was similar

to the one formed

on tissue

culture polystyrene, the reference surface. This was not the case when endothelial cells were seeded upon unmodified polytetrafluoroethylene. Surface-modified expanded polytetrafluoroethylene prosthesis material (GORE TEX@ soft tissue) also showed adhesion of endothelial cells comparable to cell adhesion to the reference surface. The amounts of serum proteins, including fibronectin, adsorbed from serumcontaining unmodified adsorbed

medium to modified polytetrafluoroethylene surfaces were larger than those adsorbed polytetrafluoroethylene. Moreover, the modified surfaces probably allow the exchange serum proteins

with

Keywords: Polytetrafluoroethylene,

Synthetic diameter ment

vascular (>5

grafts

mm)

However, because

relatively

large

employed

for replace-

of stenosed

or obstructed

grafts

rapidly

The ideal blood-contacting an endothelial

lining,

non-thrombogenic endothelium

surface

because

has been shown and improves

the patency.

by an increased

coverage

in dogs

terephthalate)

(PET)

using

of vascular

platelet Improved with

prostheses4,

Dacron@ 7. ’

Teflon@ (polytetrafluoroethylene;

it

laboratory

adhesion

with

accompanied

also preclotted

leakage

cell-graft

covered

with

surface

Biomaterials

199 1, Vol 12 March

and

provides

ePTFE prostheses

has been

a are

a fibronectin

used to improve

‘*. However,

this coating

onto those parts of the surface cells12-15.

cell overgrowth

An

alternative more

is to increase

the

of the surface”.

In the

present

PTFE surfaces

study,

this

(PTFE)

concept

films

were modified

0 1991 Butterworth-Heinemann 130

grafts

side of a prosthesis

for endothelial

material.

the

endothelial

poly-tetrafluoroethylene Correspondence to Dr T. Beugeling.

cells

polymers”.

the luminal

suitable wettability

human

(e.g. PET) and

water-wettable

interaction”,

(poly(ethylene

(ePTFE)) grafts7“.

of

platelet deposition

for making

seeded

materials

for the latter reason. Alternatively,

method

as

of in vitro of

revealed that endothelial

to moderately

cells, has been

as well

Results

interactions

to hydrophobic

of the luminal

endothelial not

on the

suitable for cell adhesion7-‘.

stimulates

cells with the graft

grafts before cell seeding are normally preclotted,

prevents

coating

of endothelial

for cell overgrowth.

cells with polymers

PETvascular substrate

onto the graft

patency,

optimal which

cells during

prostheses

endothelial

seeded

be

natural

humans5.6,

deposition

in our

show poor adhesion

on the luminal

endothelial

interaction

is necessary

endothelial

Since in humans

In bothdogs4and

that seeding

cells decreases

would

is the

spontaneously

of autologous

be necessary’-3.

endothelial

of a prosthesis

surface studies

arteries’.

due to thrombosis’.

endothelium

Optimal

inner

grafts is limited,

lining of blood vessels*.

side of a graft, seeding

expanded

a

occlude

does not develop

surgerywill

observed

human endothelial cells, gas plasma

the use of small-diameter vascular these

fibronectin.

are successfully

or bypassing

with

cellular

to of

and

was

applied

to

to ePTFE

graft

by gas plasma

(glow

Ltd. 0142-961

Z/91/0201

30-09

Adhesion

discharge)

treatment

which

effect of this treatment charge, adsorption endothelial

made them

on the surface

more wettable. composition,

of serum proteins and adhesion

cells was

The

surface of human

investigated.

Plasma

AND

ESCA measurements 800

(Kratos

surfaces,

METHODS

the

(13 cm X 13 cm X 100 pm;

ultrasonically (RBS

25,

extensively (Merck,

The Netherlands)

Oud-Beierland,

rinsed with distilled

Darmstadt,

used as received.

The

gas

performed

in

(Elecrotech, 25 mm

plasma a

oxygen

(>99.5%

UK). The

PTFE

Pa for

the

The

Hoekloos,

and 300

oxygen

plasma. was

gases was chosen plasma

were

The plasma the

PTFE

Second, Third,

plasma

the plasma chamber

samples

steps.

1 min

for

contact

angle (0) was calculated

at

First,
Pa. time.

determinations

are

nm/min

Protein

with

plasma

at the polymer

adsorption

high-density PTFE

culture

medium

enzyme

immunoassay

system17,18. solution KH,PO,; from:

The

(h) and the

h/b)

of the polymer

0.01

pH 7.4.

M KCI,

The

surfaces

of three

8.10m4

zeta

in a flat

(vy) was

plasma,

were

albumin

in

treated

etched

for to

(HDL)

(HSA),

to unmodified

20%

human

detected

human

G (IgG) and human and plasma-

serum-containing

by means

of a two

step

by van Wachem

(EIA) as described directed

from Behringwerke

with

phosphate

plate

electrolyte M

2.10m4 calculated

against

Service

and the antibody

twice

et

HSA, Fn and IgG

(CLB, Amsterdam,

directed

against

AG (Marburg,

in a 24 buffered

12H,O,0.2

saline

Germany).

(PBS)

g/l NaH*PO,.

obtained from NPBI, Emmer-Compascuum, 20%

human

wells

were

serum-containing

against

with

PTFE

(8.2

g/l

2H,O;

NaCI,

pH 7.4,

The Netherlands). with

for 1 h, the

medium

PBS containing

0.005%

(Sigma, St. Louis, MO, USA) and subsequently

for 1 h with a solution of the first antibody

directed

HSA, IgG, HDL or Fn, After rinsing with PBS/Tween-

the wells

were

incubated

peroxidase-labelled IgG)

and

was

started

(H202)

and

leuko

solution

and after

rinsed after dye

for

1 h with

second

again.

The

incubation

are mean values

obtained

of

(goat-

enzymatic

colour

the

substrate

(3,3’,5,5’-tetramethylbenzidine)

30 min this reaction from

nm) of dye solutions

a solution

antibody

with

adding 2 M H, SO,. The data, presented (450

was

and rinsed

of the PTFE surfaces culture

rinsed four times

The

HDL

well test device”

The EIA in brief: after incubation

reaction

were determined

M NazHP04,

potential

were

was terminated

absorbance

generated

by

as protein adsorption, measurements

in four wells

of a test

device.

vi = 8.4922.

1 O-8

(AE,,,/AP)

. (KB . q/s)

AE,,, _..is the streaming potential, difference, K, is the specific conductivity

AP is the pressure

17 (1 .O mPa.s)

and e (=80.14)

wherein

the dielectric

was

mounted

20,

deviation).

of the streaming

s. PTFE samples,

serum

from

were

incubated

measurements

The composition

was:

films

films

anti-rabbit

potential

films

of these elements

immunoglobulin

lipoprotein

treated

obtained

Zeta potential streaming

of human

(Fn), human

horseradish

from

PTFE

of

ions at a rate corresponding

Red Cross Blood Transfusion

of the captive

mean

in plasma-treated

or oxygen

at

concentrations

adsorption

Netherlands)

surface:

as the

different

are made

for Ta,05.

(v/v) Tween-20

expressed

surface

for 600

the

were obtained from the Central Laboratory of The Netherlands

treatment.

from the height

of spectra

argon

in hyperfiltrated

by means

h after

it is not possible to compare

the

Ar+

with

by the manufacturer.

relative to the concentrations

with

of the various

a/.“. Polyclonal antibodies

The Netherlands).

of the same film (*standard

The zeta potentials

supplied

only

and

of the

in the

pressure

flushing

stored

determined

2 arctan(2

angles

0.5

Integration

were performed

if the detail

nitrogen

detail

(20 eV pass energy)

each other

s with

the polymer

analysis,

cm2 spot size).

were treated

source

made of the Cl s, Fl s, 01 s and

concentrations

nitrogen

PTFE

Mg

between

surface

profiling

a

For quantitative

(DS 800)

3.1 g/l NapHP04.

angles

contact

and

calculated

The

for 5 min with the

Schiedam,

method16,

-

samples

of both

for the desired

by

were

bubble

All

1

and at 9 t

of three

Atmospheric

Contact

0 = 180”

oxygen

depth

angle measurement

(b) of the air bubble

calculated

for 600

before being used for experiments.

20-24

the

fibronectin

restored

Hoekloos,

were

software

films, which

chamber

The gas flow

was flushed

PTFE

width

the

were

unless

degassed

(v/v);

Contact

used

plasma

generated

(>99.997%

etching

of the

from

of the surface concentration

resolution;

power

was

water at room temperature

X-rays

and X-ray satellite subtraction

medium

The Nether-

plasma

consisted

The

treated

peaks, calculation

measurements

resolution

(0.75

placed

above

gases

electrode

were

was

low magnification

with

realized.

chamber

For

a differential

All polymer

angle of 30”

eV) were

elements

W for the nitrogen

plasma.

same gas at the same pressure. plasma

with

reactor

were

Schiedam.

The

treatment

surfaces

the

ESCA

eV) at a take-off

AZ,

in such a way that the above mentioned

pressures

pump.

and the analyser.

standard

was

plasma

inside the plasma

250

W for the oxygen

UK). with

ion gun.

performed

(1253.6

elements,

(v/v); Hoekloos, Schiedom,

The pressure

mentioned

treatment

immediately

was kept at 20 k 1 Pa for the nitrogen otherwise

argon

surface

(soft

Flagstaff,

surfaces

plasma.

lands) and nitrogen (>99.9990% The Netherlands).

ethanol

PTFE patch

barrel

electrode,

(v/v);

and

and absolute

discharge) 505

generated

with a Kratos XSAM

equipped

Using this ESCA spectrometer,

(glow

the inner

of the

performed Manchester,

was

N 1 s peaks, at medium

solution

Netherlands)

Expanded

Plasmafab

Bristol,

above

position

vacuum were

scans (steps 0.05

were cleaned

W.L. Gore and Associates,

was

Neder-

detergent

The water

Germany).

tissue, GORETEX@, USA)

Fluorplast

for 30 min in a 1% (v/v) Hicol,

system

microbeam

zeolite films

land BV., Raamsdonksveer,

were

Analytical,

were dried for at least 1 wk in a bell jar, kept in vacuum with a

treatment

PTFE films

of PTFE: A. Dekker et a/.

ESCA

pumped

MATERIALS

and adsorption

is the electrolyte constant18.

viscosity

of bulk electrolyte,

Cell adhesion Human

is

endothelial

cells were

vein according

to the method

were

cultured

routinely

isolated of Willems

for up to three

Biomaterials

from

the

umbilical

et a?‘.

The cells

passages

in tissue

199 1, Vol 12 March

131

Adhesion and adsorption of PTFE: A. Dekker et al.

culture flasks (Costar Europe, Badhoevedorp, The Netherlands) precoated with partially purified human fibronectin (coproduct obtained during the preparation of Factor VIII concentrate from cryoprecipitate; CLB) as described by van Wachem et a1.22. The culture medium consisted of a 1 : 1 mixture of Medium 199 and RPM1 1640 (Gibco Europe, Breda, The Netherlands) containing 20% pooled human serum derived from 20 healthy male donors, 2 mM L-glutamine (Merck, Darmstadt, Germany), 100 units/ml penicillin, lOO~g/mi streptomycin (both Flow Labs, Irvine, UK) and 4 ,ug/ml fungizone (Gibco). Endothelial cells were harvested by trypsin treatment (0.05% tt-ypsin/0.02% EDTA in PBS; Gibco), after which trypsin was inactivated by adding serumcontaining culture medium to the cell suspension. Cell adhesion experiments were performed with unmodified and plasma-treated PTFE films and ePTFE, which were mounted in a test device with 12 wells having test surfaces of 1.5 cm2, essentially as described by van Wachem et a/.“. The test surfaces were kept under hype~iltrated water until the beginning of the experiment. After rinsing the wells twice with PBS, endothelial cells, resuspended in 20% human serum-containing culture medium, were seeded into the wells at a density of 60 000/cm2. After 1, 2 and 6 h, the numbers of adherent cells were determined. Before detaching the cells, the wells were washed with serum~ontaining culture medium followed by rinsing with culture medium lacking serum. The adherent cells were detached by adding a known volume of trypsin solution and the suspended cells were counted in a Biirker chamber. In all experiments, adhesion of endothelial cells to the reference surface, tissue culture polystyrene (TCPS) (Costar Europe), was also determined.

I

0

t

200

I

400

Treatment

time

*

600

(s)

a

Cell spreading Samples for electron microscopy were fixed with a mixture of 1% (v/v) glutaraldehyde (Merck-Succhard, Hohenbrunn, Germany) and 1% (v/v) formaldehyde (J.T. Baker, Deventer, The Netherlands) in PBS and post-fixed in 3% (v/v) glutaraldehyde. The samples were dehydrated through a graded series of ethanol solutions to absolute ethanol, then treated for 5 min with hexamethyldisilazane (Polysciences, Warrington, PA, USA), dried, and stored desiccated23. Samples were sputter-coated with approximately 10 nm gold (cathode sputtering unit 07.120, Balzers Union Ltd, Liechtenstein). The samples were examined by means of a JSM-35 CF scanning electron microscope (Japan Electron Optics Laboratory, Tokyo, Japan) at 15 kV accelerating voltage and cell spreading was qualitatively interpreted.

RESULTS Plasma treatment

and contact angles

PTFE surfaces were modified by plasma treatment using nitrogen or oxygen. Dependent on the treatment time, modified PTFE surfaces were prepared with contact angles 1557”, versus 94-99” for unmodified PTFE. Figures la and b show a rapid, asymptotic decrease of the contact angle of PTFE surfaces as a function of the treatment time. No substantial differences between the decrease of the contact angles of PTFE films treated either with nitrogen or oxygen plasma were observed. The standard deviations of the mean values of the contact angles in Figure 7 are relatively small, although the means were calculated from contact angles of surfaces from three series of PTFE films which were treated independently.

132

Biomaterials

1991, Vol 12 March

I 0

t 200

1 400

I 600

,

Treatment time ( s)

b Figure 1 Relationship betwzen contact angle and treatment time of PTFF films modified by means of nitrogen (a] and oxygen (b) plasma. The values are the means of contact angles of three separately modified PTFE films (i- standard deviation).

Contact angle measurements at different sites of the PTFE films treated either with nitrogen or oxygen plasma, showed that the contact angles at various places in the central part (10 x 10 cm) of the films hardly differed from each other; this was the case for each treatment time (data not shown). These central parts of the PTFE films were used for further experiments. To obtain a flat surface of ePTFE graft material for contact angle measurements, patches of this material were compressed (543 K; 20 MPa). The mean contact angle of compressed ePTFE material was 104”.

Zeta potential The zeta potentials of unmodified and plasma-treated PTFE films are listed in Tab/e 1. Although the plasma treatment of hydrophobic PTFE films resulted in a drastic increase of the wettability, the zeta potential was hardly influenced. Even after a plasma treatment of 600 s the zeta potential was not much different compared to that of unm~ifjed PTFE.

Adhesion

Table

1

PTFE

films

Conracr

angles

and zera porenrials

of unrreared

Analysis

and plasma-rreared

revealed

Plasma treatment

Treatment time

Contact angle

Zeta potential

(s)

(“I

(mV)

unmodified

oxygen-containing

such groups.

94.1 -e 1.5 54.1 f 1.1 19.5t 1.3

-27.0 -24.3 -29.9

i 2.5 f 0.7 f 3.9

20

40.7

+ 0.8

~24.0

+ 0.3

600

23.2

+ 3.7

-22.6

+

0 20 600

increasing PTFE

dependent

ESCA

surface

Upon

treated

PTFE films

plasma treatment, gas (Figures

of unmodified

showed

that the spectra

depending

on treatment

2a and b). Compared

upon

time and plasma

to the Cl s spectrum

of

unmodified

PTFE, the spectra of the nitrogen

plasma-treated

PTFE films

showed

in the binding

energy

region

an increased

of 285-289

oxygen plasma-treated the region energy

eV. The intensity

increased

with

increasing

The Cl s spectra of surfaces

treated

(nitrogen

qualitatively

various

or oxygen) treatment

the spectra

PTFE films demonstrated

of 285-286

regions

intensity

eV, whereas

were

in these

in

binding

treatment same

could

plasma

oxygen

of oxygen

of PTFE films increased

resulted

with

occurred.

of

in a time-

nrtrogen

plasma

of nitrogen

analysis

of nitrogen-

3a and 6). Treatment

in incorporated

with

concentration

of the treatment medium

and oxygen.

a time-dependent

However,

the

in the surfaces

time. Since the spectra

resolution,

elements

could

indicated

and fluorine. the

not be correctly

After

surface

-3%.

relative of PTFE

time.

600

1.2

the

measured

but our data

concentratrons

to the concentrations

s treatment

with

of oxygen

concentration

were

of carbon

nitrogen

and

of

at

of thevarious

calculated,

and nitrogen

concentrations

whilst

were

the surface concentratrons

that the oxygen

with one type of plasma the

nitrogen

increase

generally

(Figures

in general very low compared

of the

changes

did

films treated with an oxygen plasma was almost independent

and plasmachanged

plasma,

time

with

incorporation

of the Cl s spectra

in the Cl s spectra

groups, after treatment

treatment

treatment

or

surfaces

The results of quantitative

or oxygen

surfaces

Contact angles were determined by means of the captive bubble method (+ standard deviation; n = 3). Zeta potentials were calculated from streaming potential measurements (k standard deviation; n = 3).

Comparison

did not have nitrogen-

that the relative surface concentrations

nitrogen

et al.

in the detail scans of the 01 s and N 1 s

and oxygen-containing with

1.3

A. Dekker

but plasma-treated

The changes

region (data not shown). showed

of PTFE:

of the 01 s and N 1 s region

PTFE

groups,

be related to changes Nitrogen Nitrogen Oxygen Oxygen

adsorption

of detail spectra

that

contain

and

plasma,

nitrogen

oxygen

after

were 600

s

1

for the

times. 0.8

0.6

0

30 Treatment

a

600

120 time

(s )

Bindlng energy (eV)

a

0.8

0.6

600

30 295

290

b Figure

( s)

Treatment

280 Figure

2

Cls

modified wirh

285

Binding energy (eV)

PTFE

nirrogen

spectra, films

obtained

(a) nitrogen

and oxygen

2 1 and

14 Pa, and

(4) 600

s.

180

plasma

and 220

by

ESCA

plasma: were

of

unmodified

(bJ oxygen

performed

W. respectively.

plasma.

and

after

plasma-

various

rimes.

at

treated

PTFE

s;

elements

Treatments

for the indicated

rimes

f 1) 0 s; (2) 30 s: (3)

120

The relative

3

rrearmenr

conditions

films

Surface films

in films see

surface

of PTFE

nirrogen

concentrations

were which

legend

concenrrartons with

of

calculated were

of oxygen /a) and

of oxygen relative

treated

and

(b) plasma

mrrogen

s. For further

for

In plasma-

ro rhe concentrations

for 600

q

m and nirrogen

oxygen

of rhese

experimental

Figure 2.

Biomarerfals

199 1, Vol

12 March

133

Adhesion

and

adsorption

Table

The

relative

2

sum ofboth with

(0

argons

of PTFE:

concentrations

+ N) in the surface

ions

A. Dekker

for various

of atomic

oxygen,

of plasma-treated

PTFE

nitrogen films

and

the

after etching

times

Etching

Plasma treatment

et al.

Oxygen

Nitrogen

O+N

w

W)

(%I

0.8

Time (min)

Layer’ Mm)

Nitrogen Nitrogen Nitrogen

0 1 2

0 0.5 1

100 8 3

100 16 8

100 11 5

Oxygen Oxygen Oxygen

0 1 2

0 0.5 1

100 18 14

100 56 76

100 22 21

0.6

18’=

All surfaces were treated wtth either nitrogen or oxygen plasma for 600 s. The atomic concentrations were determined by means of ESCA and were expressed as percentages ofthe respective oxygen or nitrogen concentrations before etchtng. For further experimental details see legend of Figure 2. is assumed the etching for PTFE about the as forTa,O,.

treatment

with

treatment

with oxygen

oxygen

plasma

was

plasma

about

resulted

5%.

24“

290

Nitrogen-plasma (contact angles)

43O

treated PTFE

Untreated PTFE

a

However,

in the incorporation

of < 1% nitrogen. The thickness treated

of the modified

at a rate corresponding assumed for

with

that the etching

Ta205.

removed,

When most

0.5

induced

nm

of the

restricted

oxygen

plasma.

plasma-treated

nitrogen

and oxygen

whilst

after

nm/min

of the

It is

surface

nitrogen

nitrogen

and

was

etching

PTFE

plasma

from

more

a layer

of 1 nm from

films,

20%

17O

of incorporated

atoms was still present 1 nm

were

by treatment

nitrogen

plasma-treated

nitrogen

Figure4

Adsorption

lipoprotein

(HDL)

and

oxygen

amounts

human

of HSA,

serum-containing films were

much

PTFE films

HDL

culture

and oxygen

independent

with

treated

from

to plasma-treated

of

human

serum

serum-containing

PTFE

culture

of dye solutions

to nitrogen

contact

or oxygen

large amounts Irrespective

with nitrogen

films

after

medium.

generated

(HSA)

8,

high-density

KI to untreated,

nitrogen

1 h incubation

The

data

in four wells

are

with

mean

(&standard

(a) 20%

values

of

deviation).

20% PTFE

PTFE

plasma were of IgG

angle. culture

plasma-treated

compared

with

medium,

decreasing

to treated

unmodified

contact

PTFE films

Fn adsorbed

PTFE films in relatively PTFE (Figure

of the type of plasma treatment,

Fn adsorbed

5).

the amounts

increased

slightly

of

with

angle. 0

Cell adhesion

20

60

40 Contact

In Figures endothelial

6a

and b, the

cells from 20%

to unmodified

results

and plasma-treated

period. After

to plasma-treated

PTFE films relatively

on plasma-treated

6 h of incubation, surfaces

1991,

Vol

are shown. of

Figure

5

Adsorption

oxygen-plasma,

A,

serum-containing surfaces

before

solutions

of fibronectin treated

culture

PTFE medium,

incubation.

generated

in four

to untreated, films

after

100

80

(“)

angle

nitrogen-plasma,

incubation

as a function

with

of the contact

The data are mean

values

wells

(+ standard

deviation).

20

45”.

20%

U

and

human

angle

of absorbances

of the of dye

PTFE films during this

approximated

12 March

to the

large numbers

the number

to TCPS. This was most evident

Biomaterials

of human

culture medium

cells hardly adhered

PTFE film, whereas

cells were detected

of adhesion

serum-containing

the first 6 h, endothelial

unmodified

134

albumin G (IgG)

of adsorbed

plasma-treated

of the contact angle. The amount

From serum-containing

adhered

Untreated PTFE

to PTFE films treated with oxygen plasma increased

decreasing

During

PTFE

to unmodified

#a and 6). The amounts

films and IgG to surfaces adsorbed

IgG adsorbed

larger than those adsorbed

(Figures

HSA and HDL to nitrogen almost

and

medium

treated

B andimmunoglobolin

(b) plasma-treated

absorbances

The

42”

34O

b

and oxygen

adsorption

27’=

Oxygen-plasma (contact angles)

in the surface,

remained.

Protein

0.6

oxygen

induced

PTFE films, only 5% of the incorporated atoms

for TazOs.

modified

than those

After

etching

0.8

with Ar+ ions

(Table 2). The surface modifications with

to the surface

oxygen

0.5

incorporated

by treatment

layer of plasma-

by etching

rate for PTFE is about the same as

atoms were also removed

with

surface

PTFE films was determined

the

of cells adhered

number of cells

for PTFE films with

contact adherent

angles

between

and

cells on these films was 90-l

of cells found on TCPS. For plasma-treated

The

10%

number

of

of the number

PTFE films with a

Adhesion

60000

39’=

41°’

49’=

17O

Nitrogen-plasma treated PTFE (contact angles)

Untreated PTFE

TCPS

for

contact

1 I

of human

from

or >45” the

treatment

larger than the number

(a) and oxygen

is expressed

percentage treatment

after

(Table

material,

or oxygen

3).

The

(b) plasma-treated

as cells/cm2

plasma

number

Since

the

incubation

with cell suspensions was

7a),

observed

whilst

cells

PTFE

endothelial

observed were

on

were

Incubation

a suspension

Occasionally,

adherent

this surface.

A monolayer

was formed

ePTFE

PTFE spread

7b).

(Figure

structure

films on all

Detachment PTFE

ePTFE

of

films

was

lining of a normal

material

of cells were observed endothelial

monolayer

on

cells

ePTFE after 6 h exposure

of the ePTFE

in

cells (Table 3: Figure 7~).

7d). This

Table 3

Adhesjon

Plasma

to a

covered

the

to investigate

and spreadmg

surface.

of human

endothelial

cells

on untreated

ePTFE

(s)

been

Several

free energy’0,26 concerning

unknown

factors

proliferation

of 15-55”

was

adsorption

of serum

Cell adhesion

Cell spreadmg

Absolute

Relative

(x 1 03/cm2)

TCPS

Surface Treatment

(%)

PTFE surfaces,

Nitrogen

180

54.4

+ 1.9

93.6

+ 3.2

+**

wettability

9.8

Oxygen

210

53.9

+ 1.6

92.7

t

2.7

+

Oxygen

360

53.4

+ 0.8

92.0

k 1.3

t

plasma,

were

7c.

values

spreading

nitrogen

are means

of almost

and oxygen

1 5 Pa and were

180

determmed

of three

all cells

plasma and after

determinations

was

(+)

or hardly

performed

any

for the

2 10 W.

respectively.

Cell

6 h

incubation.

Cell

(t

of

standard

dewatlon).

which

and on

to stimulate

of gas plasma

and adhesion

(glow on the

of endothelial by physical

the adhesion

and

of endothelial

PTFE graft material

of plasma-treated

was

PTFE films

plasma was shown

to

after a short exposure

create

surfaces

only the treatment

the clear effect

groups

of plasma

by modification.

incorporated

to be

of hydrophobic of

varying

time. treatment

of PTFE films, the zeta potentials

plasma treatment endothelial

us

by changing

hardly affected

chemical

or oxygen

even increased

allowing

Despite wettability

adhesion

the

adhesion

the wettability

method to improve the wettability

to

2 1 and

Following

cells are promoted

characterized

expanded

with nitrogen

p*

at

that

we attempted

Moreover,

characterization

t

spreading

role.

on cell that yet

to

t

and

a al.”

by means

were

cells to plasma-treated,

+ 2.8

times

or suggest

cells by improving

proteins

parameters.

57.9

mdtcated

Results

to study the effect of wettability

surfaces

97.4

adhesion

charge27,28.

of these factors

endothelial

prepared

treatment,

+ 1.6

with

have

groupsz4, 25,

A series of PTFE films with contact angles

discharge)

+ 5.7

Treatment

er

determine

characteristics chemical

play

polymers,

of endothelial

of PTFE surfaces.

56.6

(-).

also

Wachem

wettable

the adhesion

surface

specific

and surface

may

cells

of vascular grafts.

of them are of primary

contradictory

of human

moderately

of endothelial

characteristics

the influence

of van

33.6

0

cells

surface

are sometimes

90

7d. Complete

cells

study was undertaken

and which

polymer

interfacial adhesion

adhesion

the adhesion

including

of studies

Nitrogen

Figure

is not sponis the natural

of endothelial

used for the production

to polymers

proposed,

an efficient

Figure

a cell

studied.

Treatment

treatment

grafts which

seeding

is optimal

and optimize

cells. These

t,me

**See

blood vessel,

to PTFE, commonly

chemical and plasma-treated

of vascular

by endothelium,

cell seeding

suggestion

or >45”

graft

cells for 6 h resulted

of well-spread

on plasma-treated

cell suspension porous

aggregates

with

of endothelial

for 1, 2 and 6 h, no

of unmodified

of endothelial

incubation

cells to the graft surface. The present

of cells found on

modified

adhesion of rounded, notwell-spread

after

for successful

completely

from

to TCPS,

ePTFE surfaces,

unmodified

(figure

films

cells

and

was considerably

when surfaces with a contact angle <20”

used.

TCPS

n = 3).

surface

overgrown

importance.

plasma-treated

*See

films

deviation:

luminal

taneously

cell adhesion

spreading

spread

Untreated PTFE

A prerequisite

to unmodified

to the number

of

Cell spreading

with

PTFE

(Y? standard

It is not clear which

(Figure

5o” treated angles)

DISCUSSION

TCPS.

cell

et a/.

has been proposed to obtain endothelialization.

to plasma-treated

of cells adhered

on the

graft

after 6 h of incubation with a cell suspension,

After

A. Dekker

to TCPS.

with nitrogen

and almost equal (92-97%)

nitrogen

adhesion

to vascular

times

cells adhered

Cell

of plasma

cells

ePTFE patches were treated endothelial

to untreated,

compared

effect

of endothelial

various

cells

the adhesion

65 to 85%

To evaluate adhesion

endothelial

q and 6 h 0. respectively.

2

angle <20%

6 h varied

for

of PTFE:

b

Adhesion

suspension

440

Oxygen-plasma PTFE (contact

a 6

adsorotion

60000

23O

Figure

and

on the

of the surfaces

This indicates

that the

into the PTFE surface

upon

are not ionized at neutral pH. The results of

cell adhesion

neutral pH, revealed

experiments,

also performed

a clear effect of plasma treatment

B/omatenals

199 1, Vol

12 March

at

on cell

135

Adhesion and adsorption of PTFE: A. Dekker et al.

Figure 7 Scanning electron micrographs of endothelial cells adherent to: (a) untreated PTFE; (b) nitrogen plasma-treated (240 s) PTFE; (c) untreated ePTFE; (dJ nitrogen plasma-treated (180 sj ePTFE. All surfaces were incubated with cell suspensions for 6 h. Note the rounded cells in Figures 7a and c and the flattened, fully spread cells in Figures 7b and d.

adhesion.

Since the zeta potential

affected

by

surface

charge

unmodified although

plasma

probably

PTFE

and poly(vinyl

specific

adsorption

Neither detected means

within

resulted

for uncharged

cell adhesion. zeta

polymers

nor oxygen-containing

whereas

groups

of unmodified

plasma

incorporation

of

for

the relative

surface

of treated

there

and

oxygen-containing PTFE treated

groups

were

either

by nitrogen

to

a subsequent

nitrogen

and

oxygen

with

detected

a/.25 demonstrated

reactive toward This

increased were

why

as a function

treated

content with

polymer

may explain with

content

of treatment plasma,

is

radicals,

time

Comparison

of the Cls

Biomaterials

treated

with

suggest

that,

qualitative

and

irrespective

modification

processes

nitrogen

and oxygen

Etching showed

ePTFE

of the treatment

occur when

experiments

grafts such

time,

a particular

during

ESCA

nitrogen.

the same plasma

gas

different

for

of the surface

resulted (-1

in a

nm). This

with regard to modification

to improve

adhesion

a modification

and the mechanical

should

properties

of

of endothelial not affect

the

of the graft.

Adsorption of serum proteins and endothelial adhesion to plasma-treated PTFE The

adsorption

of HSA,

serum-containing films

was

HDL

culture

increased

compared

These results are comparable who demonstrated large amounts

and

medium

IgG from

20%

human

to plasma-treated

to unmodified

PTFE

PTFE films.

to those of van Wachem

that these proteins

from 20%

cell

adsorbed

serum-containing

et&l9

in relatively

medium

toTCPS

surfaces),

the films

the nitrogen treatment

of unmodified

films results

measurements

of PTFE films

is of importance

because

porosity

in

These

are somewhat

treatment

modification

vascular

plasma.

plasma.

that plasma

very superficial

differences

oxygen

when during

were treated

199 1, Vol 12 March

nitrogen

(contact angle within the range of those of the modified

whereas

on the

showed

surfaces

compared

ethylenepropylene similar

to

fibronectin spectra

depended

qualitatively

of the surfaces

plasma.

with those of PTFE films which

136

which

oxygen is more

radicals than is molecular

was very low and did not change

oxygen

of

have a purity of >99.5%.

the oxygen

nitrogen

and

plasma treatment,

that molecular

were

angle

of atmospheric polymer

in the surface during

the changes

ESCA spectra of the C 1s region of modified

cells,

and

Nitrogen-

plasma,

reaction

since the used gases themselves Ramsayet

relationship

in the surfaces

or oxygen

long-living

that the spectra alteration

whilst

PTFE

superficiality

and

of nitrogen

(data not shown).

due

are generated

linear

concentration

probably which

wettability

to this nitrogen

is a

time,

same. However,

by

oxygen-

PTFE films and the contact

at these surfaces

were

of PTFE films

nitrogen-

can be ascribed

incorporation atoms

by

PTFE films

treatment

groups into the surface. The increased

measured

caused

plasma gas revealed treatment

is used, but that these processes

PTFE films

oxygem

like poly-

and is probably

of treated between

The

potential,

groups. This

containing oxygen

in

of ions to the surface.

the surface

in the

is hardly

differences

had a negative

chloride)‘*

nitrogen-

of ESCA,

small

do not contain charged

reported

styrene

the

do not influence

surfaces

these surfaces

has also been

of PTFE surfaces

treatment,

PTFE

with the same

larger Similar

than

that

(FEP) of

adsorbed the

to the

copolymer

material

with

PTFE.

to modified

PTFE surfaces

have

adsorbed been

to

The

angle

amounts were

unmodified

reported

PTFE fluoro-

a contact

unmodified

amount

observations

hydrophobic

by others

of also

PTFE. who

Adhesion

determined fibronectin adsorption to hydrophilic and hydrophobic substrates by means of immunological and radiolabelling techniqueslg. “. 30. The relaively large amounts of fibronectin adsorbed to plasma-treated PTFE surfaces from culture medium containing 20% human serum compared to untreated PTFE strongly suggests that these modified surfaces have a high affinity for cellular fibronectinlg. Adhesion of human endothelial cells to plasma-treated PTFE films was strongly enhanced compared to unmodified films. This effect was irrespective of the gas used to generate a plasma. These results are in agreement with those of other investigators, who also found an increased adhesion of cells to more hydrophilic surfaces, obtained by plasma treatment of polymers, compared to unmodified polymers” 31-33. The adhesion of endothelial cells after 6 h incubation to modified PTFE surfaces showing a contact angle between 20 and 45” was comparable with cell adhesion to TCPS, which is known for its excellent cell adhesion properties”. Modified PTFE surfaces with a contact angle smaller or larger than those of this range also showed increased cell adhesion, compared to unmodified PTFE surfaces, but the numbers of adherent ceils were fewer than found on TCPS after 6 h incubation. Plasma treatment of PTFE does not only lead to an increased number of adherent endothelial cells, but also to a pronounced morphological change of these ceils. Endothelial cells were well spread on plasma-treated PTFE films after all cell incubation periods, but cells adherent to unmodified PTFE were hardly spread. Modified PTFE surfaces outside the contact angle range of 20-45” sometimes showed detachment of cells, especially after 6 h of incubation. The optimal spreading of endothelial cells on modified PTFE surfaces showing contact angles within the range of 20-45”, strongly suggests that endothelial cells are able to proliferate on these surfaces”. Adhesion and spreading on plasma-treated PTFE films was promoted, although a relatively large amount of proteins (IgG, HDL, HSA) adsorbed to these surfaces, which are known to inhibit cell adhesionlg. On the contrary, cell adhesion was absent on untreated PTFE, whilst fewer adhesion-inhibiting proteins adsorbed to untreated PTFE. An explanation of these phenomena may be given in terms of displacement of adsorbed serum proteins by cellular fibronectin. Endothelial cells require extracellular matrixcompounds such as fibronective for optimal adhesion and spreading. Since relatively small amounts of fibronectin adsorb from 20% serum-containing medium to plasma-treated PTFE and almost no fibronectin adsorbs to unmodified F’TFE (Figure 5), endothelial cells have to deposit their own cellular fibronectin for adhesion and spreading34. However, when endothelial cells are seeded from serum-containing medium, serum proteins such as IgG, HDL and HSA will irreversibly adsorb to the hydrophobic unmodified PTFE35.36. These irreversibly adsorbed proteins will not be displaced by cellular fibronectin, preventing optimal adhesion and spreading of endothelial cells3’. Exchange of adsorbed serum proteins with cellular fibronectin will be possible on more hydrophilic surfaces35, 36, Plasma-treated surfaces with contact angles of 20-45” most probably allow the displacement of adsorbed serum proteins by fibronectin secreted by the cells3’. PTFE surfaces with contact angles <20” probably show completely reversible protein adsorption and the interaction of adsorbed fibronectin with the surface is too weak to support optimal cell adhesion and spreading. According to van Wachem et a/.” endothelial cell adhesion is optimal to moderately water-wettale polymers, which show a contact angle of about 40”. From the results of

and adsorption

of PTFE: A. Dekker et al.

the present study, we assume that the adhesion process is not exclusively governed by the wettability of the surfaces, but also by chemical characteristics. None of the unmodified polymers studied by van Wachem ef a/.“, including those which are moderately wettable, supported complete spreading of endothelial cells. Fully spread cells were only seen on TCPS and tissue culture poly(ethylene terephthalate) (TCPET). Cellulose-2.5-acetate differs by only 4” in contact angle from TCPS; nevertheless, the number of adherent cells was <30% compared to the number of cells on TCPS. In our study, and in that of van Wachem et a/.” optimal adhesion and spreading was only observed on gas plasma-treated surfaces like TCPS, TCPET and modified PTFE. Recently Pratt eta/.32 showed thattreatment of TCPETwith air plasma also enhanced the adhesion of human adult endothelial cells. The foregoing suggests that plasma treatment of polymers introduces specific chemical groups into the surface, which not only increase the wettability of the surface, but which also have a specific effect on the interaction of endothelial cells with the surface. The surfaces of plasma-treated polymers, including TCPS2g,32 (and unpublished ESCA data, Costar Europe) as well as our treated PTFE surfaces, are enriched in oxygen and nitrogen (Figure 3). It is likely that oxygen- and/or nitrogen-containing groups are involved in making the surface optimal for endothelial cell adhesion and spreading. In view of the beneficial effects described of plasma treatment of PTFE films on the adhesion and spreading of human endothelial cells, it is logical to study the possibility of modifying Teflon vascular grafts (ePTFE; GORE TEX) in the same way. The contact angle of compressed ePTFE graft material, which was about loo”, indicated that this material had not been surface-treated by the manufacturer. Treatment of expanded PTFE patches with oxygen or nitrogen plasma improved adhesion and spreading of endothelial cells, compared to unmodified patches. Adhesion of endothelial cells to the modified graft material was comparable with adhesion to TCPS, which indicates the feasibility of plasma modification of ePTFE to improve endothelial cell adhesion. The difference in cell adhesion between modified and unmodified ePTFE patches was not as large as that observed in adhesion experiments with modified and unmodified PTFE films. This is probably due to the porous structure of the surface of ePTFE patches, which allows the cells to attach to the unmodified material. Scanning electron micrographs confirmed this, since attachment of aggregates of endothelial cells to the unmodified graft surface was observed. This phenomenon has never been observed on unmodified PTFE films. Though the surface of the ePTFE graft material is porous (internodal distance 22 pm), endothelial cells spread completely on the plasma-treated graft material and covered the pores. Cell spreading on the modified ePTFE patches was much like the spreading of endothelial cells on the luminal surface of a preclotted ePTFE graft, used in canine experiments38,3g. In conclusion, the present study shows that treatment of PTFE with nitrogen or oxygen plasma improves the wettability of the surface by introducing nitrogen- and oxygen-containing groups into the surface. Adhesion of human endothelial cells from serum-containing culture medium to these plasma-treated surfaces is comparable to cell adhesion to TCPS. Plasma treatment of ePTFE vascular graft material makes the surface optimal for the in v&o adhesion of human endothelial cells. This is promising in view of endothelial cell seeding into ePTFE vascular grafts for use in humans.

Biomatenals

199 1. Vol 12 March

137

Adhesion and adsorption of PTFE A Dekker et al.

ACKNOWLEDGEMENTS We thank Mr L. Terlingen for his help during the ESCA measurements, and Drs G. van der Sluijs and I. Vermes from the Medisch Spectrum Twente Hospital, Enschede, The Netherlands, for their support, and the obstetric staff of this hospital for the supply of umbilical cords.

2

3 4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

138

Callow, A.D., Historical overview of experimental and clinical development of vascular grafts, in Biological and Synthetic Vascular Prostheses (Eds J.C. Stanley, W.E. Burkel, S.M. Lindenauer. R.H. Bartlett and J.G. Turcotte), Grune and Stratton, New York, USA, 1982, p. 1 1 Herring, M.B., Endothelial seeding of blood flow surfaces, in Vascular Grafting, Clinical Applications and Techniques (Ed. C.B. Wright), John Wright PSG Inc., Boston, USA, 1983, p. 275 Berger, K., Sauvage, L.R., Rao. A.M. and Wood, S.J., Healing of arterial prostheses in man: its incompleteness, Ann. Surg. 1972,175,l 18 Stanley, J.C., Burkel, W.E., Graham, L.M. and Lindblad, B.. Endothelial cell seeding of syntheticvascular prostheses,Acta Chir Stand. Suppl. 1985,529, 17 Herring, M., Gardner, A. and Glover, J., Dacron” femoral-popliteal bypass grafts seeded with mechanically derived endothelium: an update, ASAIO J. 1985, 8 (2). 74 &tenwall, P., Wadenvik, H., Kutti, J. and Risberg, B., Reduction in deposition of lndium 1 1 1 -labelad platelets after autologous endothelial cell seeding of dacron aortic bifurcation grabs in humans: a preliminary report, .I. Vast. Surg. 1987, 6 (1). 17 Graham, L.M., Stanley, J.C. and BurkeI, W.E., Improved patency of endothelial-cell-seeded, long, knitted Dacron” and ePTFE vascular prostheses, ASAIO J. 1985,8 (2). 65 Burke/, W.E., Ford, J.W.. Vinter, D.W., Kahn, R.H., Graham, L.M. and Stanley, J.C., Endothelial seeding of enzymatically derived and cultured cells on prosthetic grafts, in Biological and Synthetic VascularProstheses (Eds J.C. Stanley, W.E. Burkel, S.M. Lindenauer. R.H. Bartlett and J.G. Turcotte), Grune and Stratton, New York, USA, 1982, p. 631 Graham, L.M., Burkel, W.E., Ford, J.W.,Vinter, D.W., Kahn, R.H. and Stanley, J.C.. Expanded polytetrafluoroethylene vascular prostheses seeded with enzymatically derived and cultured canine endothelial cells, Surgery 1982, 91 (5) 550 van Wacham, P.B.. Beugeling. T., Feijen, J., Santjes, A., Detmers. J.P. and van Aken, W.G., Interaction of cultured human endothelial cells with polymeric surfaces of different wettabilities, Nomaterials 1985, 8,403 Ramalanjaona, G., Kempczinski, R.F., Rosenman, J.E., Douville, EC. and Silberstein, E.B., The effect of fibronectin coating on endothelial cell kinetics in polytetrafluoroethylene grafts, J. Vast. Surg. 1986, 3 (2). 264 Seeger, J.M. and Klingman, N., Improved in vivo endothelialization of prosthetic grafts by surface modification with fibronectin, J. Vast. Surg. 1988, 8 (4). 476 Kempczinski, R.F.. Douville, E.C., Ramalanjaonia, G., Ogle, J.D. and Silberstein, E.B., Endothelial cell seeding on a fibronectin-coated substrate, in Endothelial Seeding in Vascular Surgery, (Eds M. Herring and J.L. Glover). Grune end Stratton, Orlando, Florida, USA, 1987, p. 57 Allen, B.T.. Long, J.A., Clark, R.E., Sicard, G.A., Hopkins, K.T. and Welch, M.J., Influence of endothelial cell seeding on platelet deposition and patency in small-diameter Dacron arterial grafts, J. Vast. Surg. 1984, 1 (1). 224 Dekker, A., Poot, A., Beugeling,T., Bantjes, A. and van Aken, W.G.,The effect of vascular cell seeding on platelet deposition in an in vitro capillary perfusion model, Thromb. Haemost 1989, 81 (3). 402 Andrade, J.D., Smith, L.M. and Gregonis, D.E., The contact angle and interface energetics, in Surface and InterfacialAspects of Biomedical Polymers Volume 1, Surface chemistry and physics, (Ed. J.D. Andrade), Plenum Press, New York, USA, 1985, p. 249 van Wagenen. R.A. and Andrade, J.D., Flat plate streaming potential investigations: Hydrodynamics and electrokinetic equivalency, J. Colloid. Interface Sci. 1980, 78 (2). 305 van Wagenen. R.A., Coleman, D.L.. King, R.N., Triolo, P., Brostrom, L., Smith, L.M.. Gregonis, D.E. and Andrade, J.D., Streaming potential

Biomaterials

20

21

REFERENCES 1

19

199 1, Vol 12 March

22

23

24 25 26 27

28

29

30

31

32

33

34

35

36

37

38 39

investigations: Polymer thin films, J. Colloid. lnrerface Sci. 198 1, 84 (1) 155 van Wachem, P.B., Vreriks, C.M., Beugeling, T., Feijen, J., Bantjes, A., Detmers, J.P. and van Aken, W.G.,The influence of protein adsorption on interactions of cultured endothelial cells with polymers, J. Biomed. Mater. Res. 1987,21, 701 Poot, A., Beugeling, T., van Aken, W.G. and Bantjes. A., Detection of surface-adsorbed (lipo)proteins by means of a two step enzymeimmunoassay: A study on the Vroman effect, J. Biomed. Mater. Res. 1990,24,102 1 Willems.Ch.. Astaldi,G.C.B.,deGroot, Ph.D., Jansen, M.C.,Gonsalves, M.D.. Zeijlemaker, W.P., van Mourik, J.A. and van Aken, W.G., Media conditioned by cultured vascular endothelial cells inhibit thegrowth of vascular smooth muscle cells, Exp. Cell Res. 1982, 139, 191 van Wachem, P.8.. Reinders, J.H., van Buul-Wortelboer, M.F., de Groot. Ph.G., van Aken, W.G. and van Mourik, J.A., Von Willebrand factor in cultured human vascular endothelial cells from adult and umbilical cord arteries and veins, Thromb. Haemost 1986,58, 189 Nation, J.L.. A new method using hexamethyldisilazane for preparation of soft insect tissues for scanning electron microscopy, Stain Technology 1983,58 (8). 347 Lydon. M.J., Minett, T.W. and Tighe, B.J., Cellular interaction with synthetic polymer surfaces in culture, Biomaterials 1985, 6, 396 Ramsay. W.S., Hertl, W., Nowlan, E.D. and Binkowski, N.J., Surface treatments and cell attachment, ln Vitro 1984, 20 (10). 802 Absolom, D.R., Hawthorn, L.A. and Chang, G., Endothelialization of polymer surfaces, J. Biomed. Mater. Res. 1988, 22, 271 Hatton,S.,Andrade, J.D., Hibbs Jr, J.B., Gregonis, D.E. and King, R.N., Fibroblast cell prolieration on charged hydroxyethyl methacn/late copolymers. J. Colloid, Interface. Sci. 1985, 104. 73 van Wacham. P.B., Hogt, A.H., Beugeling, T., Feijen, J., Bantjes, A., Detmers. J.P. and van Aken, W.G., Adhesion of cultured human endothelial cells onto methacrylate polymers with varying surface wettability and charge, Eiomarerials 1988, 8, 323 Chinn, J.A.. Horbett, T.A., Ratner, B.D., Schway, M.B., Haque, Y. and Hauschka. SD.. Enhancement of serum fibronectin adsorption and the clonal plating efficiencies of Swiss mouse 3T3 fibroblast and MM14 mouse myoblast cells on polymer substrates modified by radiofrequency plasma deposition, J. Colloid. interface Sci. 1989, 127,67 Chilkoti, A., Ertel, S.I., Ratner, B.D., Horbett, T.A. and Briggs, D., Acetone-N2 glow discharge modified surfaces: investigation of surface chemistn/, protein adsorption, and endothelial cell growth, Transactions Fifteenth Annual Meeting of the Society for Biomaterials 28/4-2/5 1989, Vol. XII. Lake Buena Vista, Florida, USA, p. 99 van Wachem, P.B., Interactions of cultured human endothelial cells with polymeric surfaces, PhD Thesis, University of Twente, Enschede, The Netherlands, 1987, p. 39 Pratt, K.J., Williams, S.K. and Jarrell, B.E., Enhanced adherence of human adult endothelial cells to plasma discharge modified polyethylene terephthalate, J. Biomed Mater. Res. 1989, 23, 1 131 Klee, D.. Breuers, W., Bilo-Jung, M., Mittermayer, C. and Hmker, H., Modifizierung von polymeroberflschen zur erhohung der zelladhasion, Die Angewandte Makromolekulare Chemie 1989, 168/l 87, 179 van Wachem. P.B., Mallens, B.W.L., Dekker,A., Beuge1ing.T.. Feijen. J., Bantjes. A., Detmers. J.P. and van Aken, W.G., Adsorption of fibronectin derived from serum and from human endothelial cells onto tissue culture polystyrene, J. Biomed. Mater. Res. 1987, 21, 1317 Feijen, J.. Beugeling, T., Bantjes. A. and Smit Sibinga, C.Th., Biomaterials and interfacial phenomena, in Advances in Cardiovascular Physics 3 (Ed. D.N. Ghista), S. Karger, Easel, Switzerland, 1979, p. 100 Brash, J.L., Mechanism of adsorption of proteins to solid surfaces and its relationship to blood compatibility, in Biocompatible Polymers, Metals and Composires (Ed. M. Szycher), Technomic, Lancaster, Pennsylvania, USA, 1983, p. 35 Dekker, A., Beugeling, T., Wind, H., Poot,A., Bantjes, A., Feijen, J. and van Aken, W.G., Deposition of cellular fibronectin and desorption of human serum albumin during adhesion and spreading of human endothelial cells on polymers, J. Mater. Sci: Materials in Medicine (in press) Plate, G., Hollier, L.H., Fowl, R.J., Sande, J.R. and Kaye, M.P.. Endothelial seeding of venous prostheses, Surgery 1984,98 (5). 929 Pearce, W.H., Rutherford, R.B., Whitehill. T.A.. Rosales, C., Bell, K.P., Patt, A. and Ramalanjaona, G., Successful endothelial seeding with omentally derived microvascular endothelial cells, J. Vast. Surg. 1987, 5 (1). 203