Requirements for cell spreading on polyHEMA coated culture substrates

Cell Biology

International

Reports,

Vol. 8, No. 2, February

1984

151

REQUIREMENTS FOR CELL SPREADING ON PolyHEMA COATED CULTURE SUBSTRATES T.W. Minett and B.J. *M.J. Lydon, Unilever D.A. Rees, N.I.M.R., *Author

Tighe, University of Aston, Birmingham U.K. Research, Sharnbrook, Bedfordshire, U.K. The Ridgeway, Mill Hill, London, U.K.

to whom correspondence

should

be addressed.

ABSTRACT We have investigated the topography of polyHEMA coated culture substrates by scanning electron microscopy, and quantitatively assessed their effect upon the spreading activity of mammalian cells. Results indicate a clear correlation between cell Preparation of spreading activity and pblymer film discontinuity. polyHEMA films on modified tissue culture substrates has allowed direct investigation of the role of the underlying substrate in regulating cell spreading, and confirms that apparent modulation of cell spreading by polyHEMA reflects increasing expression of We have further employed a spinning technique the coated surface. by which films of precise thickness, down to 0.01 urn, may be produced on coverslips. All polyHEMA coatings prepared in this way are smooth and complete. They do not allow cell attachment We conclude that polyHEMA is non-adhesive for at any thickness. mammalian cells. INTRODUCTION The development of model substrates for anchorage-dependent cell growth in culture is an important experimental approach to understanding the significance of cell/matrix and cell/cell interactions in vivo. Our attention was drawn to the potential value of the biocompatible hydrogel poly-2-hydroxyethyl methacrylate (polyHEMA) in modulating cell spreading and metabolism (Folkman and Moscona, 1978; Ben Ze'ev et al., Polymer films 1980). of various "thicknesses" are prepared by a solvent evaporation method and cell spreading behaviour on these films appears to correlate with increasing polymer dilution (decreasing film "thickness"). At first sight, however, such continuous modulation is difficult to reconcile with expectations from physical principles that polyHEMA gels should always present a surface of identical physicochemical characteristics irrespective of "thickness". With this in mind, we have investigated the topography of polyHEMA gels by scanning electron microscopy, and quantitatively assessed their effect upon the spreading activity of BHK fibroblasts. We have further employed a spinning technique by which intact polyHEMA films of precisely measured thickness (down to 0.01 urn) are produced on coverslips. Our 03OS-1651/84/020151-09/$03.0010

@ 1984 Academic

Press Inc. (London)Ltd.

Cell Biology

152

International

Reports,

Vol. 8, No. 2, February

results demonstrate that "modulation" of cell polyHEMA derives from the coating technique.

spreading

1984

by

RESULTS AND DISCUSSION Our initial experiments were designed to prepare, by a spincoating technique (see Legend, Table I) polyHEMA films of uniform thicknesses down to 0.01 urn. These films would allow us to study the formation of specialised areas of cell/substrate interaction ("focal adhesions") using interference reflection microscopy (Curtis, 1964; Abercrombie and Dunn, 1975). To our surprise, however, we observed only limited cell adhesion and no cell spreading on these surfaces at any polymer thickness (Table I). Nevertheless, polyHEMA films prepared by the solvent evaporation method of Folkman and Moscona (1978) did show some cell attachment and spreading at lower polymer concentrations as previously reported (Table I). Table

I

BHK Cell Spreading Activity on PolyHEMA-Coated Substrata BHK fibroblasts were routinely maintained at 37’C in Dulbecco's MEM (Gibco, Europe) supplemented with L-glutamine (Gibco, Europe) For attachment and and 10% foetal bovine serum (Flow Labs.). spreading assays, ccl 8 were detached by trypsinization, resuspended to 5 x 10 4 /ml and 2ml of cell suspension seeded onto Cultures were maintained for up to 6 hours dishes or coverslips. at 37’C and examined by light microscopy (x400 magnification). 1. Spun polyHEMA coatings were prepared using a spin coater (DAGE PRECIMA model PRS 14E) usually employed for the Glass coverslips production of very smooth photoresist films. were centrifugally coated with solutions of polyHEMA in 2Preliminary experiments were carried out to ethoxy ethanol. determine the relationship between polymer concentration, spin Thickness measurements were by speed and final film thickness. method using a Zeiss interference microscope. the "Interphaco" polyHEMA coatings were prepared by the method of 2. Evaporation Dilutions as shown of a 12% Folkman and Moscona (1978). stock solution of polyHEMA in ethanol were pipetted onto 35mm petri dishes or 22 x 22mm coverslips contained &n 35mm petri dishes and allowed to evaporate overnight at 37 C. 35mm petri dishes were pre-coated with 1OOug bovine plasma 3. fibronectin in lml phosphate buffered saline and coated with polyHEMA as described. Polymer surface topography was assessed by scanning electron microscopy as described in the legend to Fig. 1. N.D. = Not Determined

1 to 1:lO 1:50 to 1:200 1:500 to 1:88g 1:lOOO to 1:1667 Ethanol-treated fibronectin.

3 Evaporation

coatings on fibronectin coated substrates. Dilutions of 12% stock solution.

Complete 4, ,t t,

1 to 1:lO to 1:200 1:500 to 1:88g 1:lOOO to 1:1667

2 Evaporation coatings on bacteriological grade plastic. Dilutions of 12% stock solution.

1:50

Complete

dilutions

All

2 Evaporation coatings on siliconized glass.

Complete (1 II f!

1 to 1:lO 1:50 to 1:200 1:500 to 1:889 1:lOOO to 1:1667

2 Evaporation coatings on tissue culture plastic. Dilutions of 12% stock solution.

Medium II II 11

Medium

Medium 1, !I 11

Medium

Conditions

Complete

Culture

0.01 urn

Surface

1 Spun coatings

Culture

TABLE I Morphology

No Adherence Rounded Partially Spread Fully Spread Fully Spread

No Adherence Rounded Aberrant spreading Aberrant spreading

No Adherence

No Adherence Rounded Partially Spread Fully Spread

No Adherence

Cell

Broken

Complete

N.D.

N.D.

Discontinuous at dilutions h 1:50

Smooth, Rugose Rugose Clearly

Polymer Surface Topography Smooth, Complete

Cell Biology

Figure

International

Reports,

Vol. 8, No. 2, February

1984

1: PolyHEMAfilm

Scanning electron micrographs of polyHEMA spun coatings and evaporation films (prepared as described in the legend to Table I). (a-f): evaporation films, (g, h): spun coatings. Coatings were prepared on coverslips, incubated in Dulbecco's MEM supplemented with L-glutamine and 10% foetal bovine serum for 6h at 37'C, and fixed in 2% glutaraldehyde/O.lM sodium cacodylate buffer at pH 7.4 for 30 mins. Following a further two washes in cacodylate buffer, films were dehydrated through a graded series of ethanol:water solutions (50%, 70%, 90% and 100% ethanol) for 3 x 15 minutes for each dilution. Samples were then critical-point dried via liquid CO2 in a Polaron E3100 II critical-point bomb. The coverslips were mounted on brass specimen stubs using colloidal silver and coated with 1008 of gold/palladium in a Polaron E5100 'Cool' Spatter Coating Unit at 800~ and 20mA. The samples were examined in a JOEL-35X at 25kV accelerating voltage. a. b. c. d. e. f.

Dilution Dilution Dilution Dilution Dilution Dilution

= = = = = =

1. 1:50. 1:lOO. 1:200. 1:500. 1:lOOO.

Estimated Estimated Estimated Estimated Estimated Estimated

thickness thickness thickness thickness thickness thickness

= = = = = =

35 0.7 0.35 0.18 0.07 0.035

urn urn urn urn urn urn

Cell Biology

International

Reports,

Vol. 8, No. 2, February

1984

155

g. Stock solution = 556 w/v polyHEMA. Spin speed = 1150 rpm. Measured thickness = 0.2 urn. h. Stock solution = 3.5% w/v polyHEMA. Spin speed = 2000 rpm. Measured thickness = 0.07 ?im. Bar = 1 urn. Investigations of the surface topography of evaporation polyHEMA films by scanning electron microscopy showed clear evidence of 1: the polymer appears discontinuity of the thinner coats (Fig. to break up at dilutions of 1:50 (estimated thickness = 0.7um [Folkman and Moscona, 1978])or more. Moreover, there is a correlation between polymer film discontinuity and cell spreading polyHEMA films In contrast, behaviour (Table I, Figs. 1 and 2). prepared by our spinning technique (e.g. see Fig. 1) were smooth They do not support cell spreading at any thickness. and complete. It is unlikely that the rugose surface topography of thinner evaporation films is an artefact of ethanol dehydration for spun coatings are never presented scanning electron microscopy: as rugose and evaporation films dehydrated in acetone (which has a different swelling constant for polyHEMA) also show surface rugosity of the thinner films. It is also interesting to note that Toselli et al (1983) have recently described a "crater-like" topography of collagen-HEMA and elastin-HEMA gels prepared for transmission electron microscopy by dehydration in isotonic saline/ethanol solutions containing polyethylene glycol. Quantitation of cell spreading behaviour on evaporation polyHEMA films (Fig. 2) suggests that the transition from a fully rounded to a fully spread state is not a gradual progression. Although values for mean cell diameter increase gradually from 1:50 to I:800 dilution, there is a dramatic transition between 1:800 and l:l,OOO polyHEMA. On the other hand, modal cell diameters remain at roughly the same level between 1:50 and 1:800. Analysis of distribution of largest Feret's diameter indicates significant positive skew for each substrate except uncoated glass (Table II). Only discontinuity in the polymer surface can explain this statistic. The above spreading substrate Preparation has allowed instance, coverslips. adherence

data are consistent with the observation that cell is achieved only when the underlying tissue culture is "unmasked" by disintegration of the polymer film. of solvent evaporation films on modified substrates us to confirm this view (Table I). In the first polyHEMA films were prepared on siliconized glass BHK cells seeded onto these films showed no irrespective of polymer concentration (Table I).

Cell Biology

156

1:lOO

E =

International

1:200

1:300

1:500

Reports,

1:600

Vol. 8, No. 2, February

1:800

1984

1:lOOO Glass

200-

3 zi .-s a =

. I

100 -

8

I

.

I 50

I 50

I 50

I 50

I 50

r

I

I 50

0

0

0

0

0

0

0

r

r

r

t

I

0

50 0

50 0

0

% Frequency Figure

2: Quantitation polyHEMA

of BHK spreading

response

on evaporation

films

Cells in exponential phase of growth were detached by trypsinisresuspended in Dulbecco's MEM suppkemented with L-glutamine ation, seeded onto 35mm and 10% foetal bovine serum and 0.35 x 10 cells petri dishes coated with the dilutions shown of a 12% stock After 6 hours at 37'C, cells solution of polyHEMA in ethanol. were fixed and stained in a 2% solution of amido black in water: Measurement of cell spreading acetic acid: ethanol (5:1:5; v/v). was performed. on a Quantimet image analyser (Cambridge Instruments) For each sample, largest Feret's using the "rawdat" programme. diameter (longest measured cell axis) of approximately 150 cells this required 6 separate fields per dish Typically, was measured. These were selected"blind" by the microscope to be analysed. ensuring random collection of data. operator, A = modal

diameter

0 = mean diameter. Histograms represent distribution polymer dilutions 1:50 to l:l,OOO.

of largest

Feret's

diameter

for

Cell Biology Table

II

International

Reports,

Distribution : Analysis

Vol. 8, No. 2, February

of Largest of Skewness

Feret's

157

1984

Diameter

* Jb

Significance

Glass

0.0374

Not Significant

1:lOOO

0.5577

P < 0.05

I : 800

1.2625

P< 0.01

1:600

1.4049

PCO.01

1:500

0.8610

PC 0.01

1:300

0.8902

P< 0.01

1:200

1.0675

P< 0.01

1:lOO

1.6541

P< 0.01

1:50

0.8612

PolyHEMA Dilution

P< 0.01 3/2

* db

1

m3

m2 *Departure frequency

of Jb1 function

= m /m 3

2

= c (x-xl3 =variance

f rom zero is an indication of the sample population

of skewness

in the

A second approach was to seed cells onto evaporation polyHEMA films prepared on bacteriological plastic. Although cells attached and showed some spreading at similar polymer concentrations to controls on tissue culture plastic, spreading was aberrant, as might be expected on bacteriological grade surfaces (e.g. Grinnell and Feld, 1982, Curtis et al.1983) (Table I). Finally, polyHEMA films were prepared on substrates pre-coated with the adhesive glycoprotein fibronectin (see Hynes and Yamada, 1982), which permits full cell spreading in serum-free culture conditions (Hughes et al Cell attachment and spreading 1979). behaviour on these films: in serum-free medium, was essentially identical to that in control experiments (Table I). From our observations, it is clear that BHK cell attachment and spreading on polyHEMA coated substrates is entirely dependent upon polymer film disintegration. Our results have been independently confirmed for 3T3 fibroblasts, which fail to spread onto spun polyHEMA films except within breaks in the polymer or on palladium shadowed areas (Dr.G.Ireland; personal communication).

158

Cell Biology

International

Reports,

Vol. 8, No. 2, February

1984

We have now extended our observations to a series of mammalian cell lines, both fibroblastic (MRCS, 16~, REF, CEF, Vera) and epithelial (HEp-2). We find no cell attachment to complete polyHEMA films. To our knowledge, only cultured neurones have so far been reported to attach to "thick" (and presumably smooth) polyHEMA gels (Carbonetto et al, 1982). Even so, neurite outgrowth from these cells requires incorporation of fibronectin or some other adhesionpromoting species into the hydrogel. Fibronectin itself shows very little adsorption (a pre-requisite for cell spreading) to 1981). We conclude, therefore, polyHEMA surfaces (Klebe et al that polyHEMA is non-adhezior mammalian cells. PolyHEMA controls spreading only through extent of substratum "masking". defined chemical modifications of polyHEMA'may Nevertheless, provide interesting polymers capable of holding cells in a given spread state. We are currently pursuing this approach. ACKNOWLEDGEMENTS We thank Chris Clay (Unilever) for electron microscopy, Paul Robinson (Unilever) for advice and assistance with Quantimet analysis and Christine Foulger and Amanda Kelland (Unilever) for routine cell culture. Dr. Peter J. Skelly and Dr. Alan M. Jolly (University of Aston) advised on polymer synthesis and coatings. We are also grateful to Tom Kirkwood (N.I.M.R.) for discussions on statistical analysis. T.W.M. is supported by a S.E.R.C. C.A.S.E. award.

REFERENCES Abercrombie, M. and Dunn, G.A.; Adhesions of fibroblasts to substratum during contact inhibition observed by interference reflection microscopy. Experimental Cell Research, -92 57-62 (1975). Ben-Ze'ev, A., Farmer, S.R. and Penman, S.; requires cell-surface contact while nuclear cell shape in anchorage-dependent fibroblasts. (1980). Carbonetto, S.C., growth on defined

Gruver, hydrogel

Protein synthesis events respond to Cell &I 365-372

M.M. and Turner, D.C.; Nerve fibre substrates. Science 216 897-899

(1982).

The mechanism of adhesion of cells to glass: a Curtis, A.S.G.; Journal of Cell study by interference reflection microscopy. Biology, -20 199-215 (1964). Curtis, A.S.G., Forrester, J.V., Adhesion of Cells to Polystyrene Biology, 97 1500-1506 (1983).

McInnes, Surfaces.

C. and Lawrie, F.; Journal of Cell

Ceil Biology Folkman, control.

International

Reports,

J. and Moscona, A.; Nature 273 345-349

Vol. 8, No. 2, February Role of cell (1978).

159

1984

shape in growth

adsorption on Grinnell, F. and Feld, M.K.; Fibronectin hydrophilic and hydrophobic surfaces detected by antibody binding and analyzed during cell adhesion in serum-containing medium. Journal of Biological Chemistry. 257 4888-4893 (1982). Hughes, R.C., Pena, S.D.J., Molecular requirements for fibroblasts. Experimental

Clark, J. and Dourmashkin, the adhesion and spreading Cell Research, 121 307-314

R.R.; of hamster (1979).

Hynes, R.O. and Yamada, K.M.; Fibronectins: multi-functional modular glycoproteins. Journal of Cell Biology, s 369-377

(1982).

Klebe, R.J., Bentley, K.L. and Schoen, R.C.; Adhesive substrates for fibronectin. Journal of Cellular Physiology, 109 481-488 (1981). Toselli, P., Faris, B., Oliver, P., Wedel, N. and Franzblau, Preservation of ultrastructure of cells cultured on proteinhydroxyethylmethacrylate hydrogels. Journal of Ultrastructure Research, & 220-231 (1983).

Received:

19th December 1983.

Accepted:

5th January

C.;

1984