In vitro studies on biocompatibilitg of hyaluronic acid esters

R. Cortivo, P. Brun, A. Rastrelli* and G. Abatangelo Institute of Histology, University of Padova, 75 Via Trieste, 35 12 1 Padova, ‘FIDIA Bioskin Division, Via Ponte de//a Fabbnca 3/A

Abano T., Italy

(Received 30 November 1990; revised 14 January 199 1; accepted 15 February 199 11

The biocompatibility of semisynthetic polymers formed from hyaluronic acid esters has been studied using fibroblast cultures. The polymers, added to the culture medium, used either in powdered form or as thin membranes, behave as inert materials. The cells used in the experiments grow normally in the culture dishes. With regard to adhesiveness the cells were not able to spread on the biomembranes and tended to form isolated clusters of round cells. Human fibronectin, placental collagen (type I-IV) and fibrin could be stratified on biomembranes. When these molecules reacted with the biomaterial the film became suitable for fibroblasts spreading and growth.

Keywords: Hyaluronic acid esters, biocompatibility, cell adhesion, cell cuhwation

Hyaluronic acid (HA) is a naturally occurring non-sulphated

glycosaminoglycan consisting of a linear sequence of

D-glucuronic acid and A/-acetyl-D-glucosamine. It is present

in connective tissue, in the synovial fluid of articular joints

and in the vitreous humor of the eye. HA is important in many

biological processes such as tissue hydration, proteoglycan

organization in the extracellular matrix” *, cell differentiation3, 4,

cell behaviour5-lo and tissue repair”-‘3. In recent years HA

has been used for clinical purposes in many different

pathological conditions, such as eye surgery, in osteo-

arthritis’*, 14, wound repair’ 5-1 7 and for cosmetological use.

Recently FIDIA S.p.A. used HA in the preparation of

semisynthetic insoluble polymers through the esterification

of carboxyl groups with alcohols’8. The products obtained

with such procedure (Hyaff) are soluble in DMSO and their

molecular weight and polydispersity, measured by GPC-

LALLS, are respectively 145.000 Da and 2.0. The polymers

exhibit good stability to hydrolysis at acidic pH and from

contact angle measurements, they show high surface

energy (I,, = 92.5) and strong ability to interact with polar

molecules’g. With the phase inversion technique, the totally

esterified hyaluronic acid derivatives have been transformed

in membranes 1 7-25 pm thick. As these materials are of

great interest in the pharmaceutical and clinical practice,

expecially for wound dressing or drug release, we studied

the biocompatibility of these materials in vitro, by cultivating

different cell types in the presence of either the powdered

material (raw material), or films of Hyaff 7 (ethyl ester of HA)

and Hyaff 9 (prop#e&erof HA), with particular attention to

Correspondence to Dr G. Abatangelo.

0 1991 Butterworth-Hmnemann Ltd 0142-9612/91/080727-04

adhesiveness or cell attachment. In additron, we tried to

study the interaction of some naturally occurring proteins

(collagen, fibronectin and fibrin) with the Hyaff biopolymers.

This was done in consideration of the fact that these

semisynthetic biomaterials can be used in wound dressing.

It is well known that in wound healing some proteins, such

as collagen, fibrin, fibronectin, exert a crucial role in cell

behaviour, facilitating the interactions between cells and the

extracellular matrix.

MATERIALS AND METHODS

Materials

Thromboplastin/Ca*+ solution was purchased from

Boehringer (Mannheim GmbH, Germany). Affi-Gel Gelatin

Fibronectin purification gel was obtained from BioRad (CA,

USA). Collagen was obtained from human placenta through

guanidinium-HCVDTE extraction and DEAE-Sephacel

column chromatography. The human fibronectin was purified

from 30 ml of fresh titrated plasma, using a column of Affi-

Gel Filtration Fibronectin purification (BioRad). 4-5%mg of

pure fibronectin were obtained, the purity of which was

confirmed by gel electrophoresis analysis. Films formed

from the ethyl (Hyaff 7) and n-propyl (Hyaff 9) esters were

supplied in sterile form by FIDIA S.p.A. (Abano T., Italy).

Polymer in powdered form (raw material) was also supplied

by FIDIA. The degree of esterification of the powder and the

films of Hyaff was 100% and in this form the biopolymers

were insoluble in water.

Biomatenals 199 1. Vol 12 October 727

Biocompatibility of hyaiomnic acid esters: R. Cortivo et al.

Cell cultures

In control culture dishes and in dishes containing either powdered material or adherent strips of Hyaff, five different types of fibroblasts were seeded: 15 d chick embryo fibroblasts, dermal fibroblasts of adult rat, CHO, NIH 3T3 and,RAT 2 fibroblasts. Cells were seeded at a concentration of 5 X 1 05/6 cm dish and maintained in DMEM containing antibiotics and 5% FCS. Every 24 h each dish was observed by phase contrast microscopy until confluency. The 15 d chick embryo fibroblasts reached confluence in the shortest period (3 d), the dermal fibroblasts of adult rats showed the slowest growth rate (5 d to reach confluence). All other fibroblasts required 4 d to reach confluence.

The polymer in powdered form was suspended in the culture medium at concentrations ranging from 100 to 800 mg/dish. The films of both Hyaff 7 and Hyaff 9 were attached to the dishes in the following way: to a drop of human titrated plasma, applied on the culture dish, 10 ,uI of a thromboplastin/Ca2+ solution was added. After a few seconds, a strip (1 X 2 cm) of Hyaff film was gently applied on the plasma drop taking care to avoid ruffling. After a few minutes the film stuck to the culture plate and then cells could be added. In all the culture plates, containing two adherent strips of Hyaff film, the cells were seeded in DMEM. The membranes remained attached to the dishes until the cells reached the confluence, that is for 3-5 d.

In a parallel set of experiments, collagen, fibronectin and fibrin were separately coated on the membranes of both Hyaff 7 and Hyaff 9 before the cell inoculum. In a defined area of the membrane adherent to the culture dish, 100 ,uI of a sterile solution containing collagen (type I-III; 2OO~glml) or fibronectin (1 OO~g/ml) were applied. The solution was taken to dryness at room temperature in sterile conditions, avoiding shaking. Fibrin was coated on the biomaterial in the following way: to a drop (100 pl) of human titrated plasma applied on the Hyaff strip, 10~1 of thromboplastin/Ca2+ solution was added, and after gentle mixing with a thin sterile glass rod, the drop was dessiccated at room temperature.

RESULTS

All thefibroblastsseeded in theculturedishescontaining the suspended powdered biomaterials grew normally and reached confluence after a period of time similar to the control cultures. Microscopicallythe biomaterial appeared as highly hydrated floating granules or rods, sometimes giving rise to small aggregates. The cells attached, spread and grew on the piastic surface of the dishes as seen in the control cultures (data not shown).

Cells seeded in culture dishes containing the Hyaff strips adherent to the dish (Figure f ) spread and grew in all the areas surrounding the biomembranes and reached a state of confluence similar to the control cultures. With regard to adherence, the cells did not spread easily on the Hyaff biomembranes, tending to form only small clusters of rounded cells (Figure 1 B). At a time when in the surrounding areas of the culture plate the cells had reached confluence (3-5 d), only a few scattered cell clusters were formed on the biomaterial. On the contrary, Figure 2 shows that fibroblasts cultivated on the Hyaff strip containing ‘a defined area (A) coated with fibronectin (see Methods) were able to attach, spread and proliferate. As expected, in the zone without fibronectin (figure 2 B) the cells did not spread and remained

728 Biomaterials t99 1, Vol 12 October

Figure 7 15 d chick embryo fibroblasts 72 h after seeding. R the plastic surface of the culture dish on which the fibmblasts are near confluency. 8, the Hyaff membrane on which the fibroblasts form only smaflctusters (see text).

Figure 2 15 d chick embryo fibrobtasrs 72 h after seeding. A the area of rhe Hvaff membrane on which human ~~roffecti~ has been coatedbefore the cell inoco/um; in this zone the cefls are c~ffue~t. 8, the membrane surface without fibmnectin on which the cells form only small clusters.

in small clusters. In Figure 3 it can be seen that the fibroblasts grew normally on the plastic surface of the tissue culture dish (A) while on the Hyaff strip they were able to grow only in the zone where collagen had been previously coated (upper corner of 6). The zone of the Hyaff strip devoid of collagen (lower corner of B) showed only a few scattered cells. Fibrin stratified on the membranes tended to form a delicate, thin veil on which thecells easilya~ached and grew, as clearly seen in Figure 4.

The characteristics of cell spreading and growth in the culture medium containing the biomaterials were similar for all the fibroblasts we used, except the already mentioned difference in the growth rate.

DlSCUSSlON

Our results can be summarized as follows:

The powdered form of biomateriai Hyaff 7 and Hyaff 9 (raw material), the ethyl and n-propyl esters of hyaluronic acid, respectively, added in suspension in the culture

Biocompatibilitv of hyaluronic acid esters: R. Cortivo et al.

Figure 3 15 d chick embryo fibrobiasts 72 h after seeding. A, the plastic surface of the culture dish with confluent cells. B, the Hyaff membrane on which the cells grow normally only in the zone where collagen has been coated (upperpartJ, while the zone deprived of collagen flower part) shows onlv few scattered cells.

Figure 4 15 d chick embryo fibroblasts 72 h after seeding. The upperpart of the figure represents the zone of the Hyaff membrane coated with human fibrin (see text) on which the cells are near con fluency. The lower part of the figure represents the zone of the Hyaff membrane without fibrin on which small cell clusters are present.

2.

3.

4.

5.

medium, did not influence the cellular growth and behaviour in v&o, and appeared to be inert materials without any toxicity. Cells grown in the presence of Hyaff powder spread and grew on the plastic surface of the dish, reaching confluence in a period of time comparable to the control cultures. The Hyaff 7 and Hyaff 9 membranes were easily attached to the plastic surface of the culture plates by means of a fibrin clot. Cells seeded in the dishes containing the adherent strips of Hyaff spread and grew in all the areas surrounding the biomaterial, reaching confluence after the same time as the control cultures. With regard to adhesiveness, the cells were not facilitated to spread on Hyaff biomembranes and tended to form only a few isolated clustels of round cells. Human fibronectin, placental collagen (type f-111) and fibrin can be stratified on the Hyaff membranes. When

these molecules reacted with the biomaterial the adhesive properties of the film became suitable to allow spreading and growth of the fibroblasts.

From these data, it is reasonable to conclude that the esters of hyaluronic acid, Hyaff 7 and Hyaff 9, represent a non-toxic material when in contact.with the cells cultivatedin vitro, and it is presumed that these biomaterials can be easily used in contactwith living tissue. It appears that Hyaff esters, at least in short-term cell cultures, are biologically inert. ln viva, one would presume that the biopolymer would be rapidly opsonized by attachment factors such as fibrinogen and fibronectin. In fact, fibronectin, collagen and fibrin reacted readily with Hyaff membranes, giving rise to a modified surface suitable for cell spreading and growth. Withrn the perspective that these hyaluroni~ acid esters will be used for pharmaceutical and clinical purpose’8,‘g our data may be of interest in confirming the absence of toxicity and the biocompatibilityof the polymer. As matter of fact, concerning the pharmaceutical use of these biopolymers, they have been demonstrated to be suitable in applications where rapid drug release is desired”.

REFERENCES

13

14

15

16

17

Wnght, T.C., Underhtll, C.B., Toole, B.P., Karnovsky. M.J., Dtvalent canon lnde~ndent aggregation of RAT 1 flbroblasts Infected with a temperature aensitwe mutant of Row Sarcoma Virus, Cancer ffes. 1981,41,5107-5113 Laurent, T.C. and Fraser, J.R.E.. The properttes and turnover of Hyaluronan, tn Ciba Foundation Symposium: Functions of the Proteoglycans. Wtiey. Chrchester, UK, 1986. pp S-29 Toole, BP., GAGS tn morphogenests. tn Cell Biology of the Extracellular Matrix (Ed. ED. Hay). Plenum. NY, USA, 1981, pp 259-294 Toole, B.P. and UnderhIll, C.B. Regulatton of morphogenests by the pericellular matnx, tn Cell Interactions and Development: Molecular Mechanism (Ed. K. Yamada), Wiley, NY, USA, 1983, pp 203-230 Underhtll. C.B., Dorfman. A., The role of HA tn vttercellular adheston of cultured mouse cells, Exp. Cell. Res. 1978, 117, 155-l 64 Hakansson. L., Hallgren. R.. Venge. P., Regulatton of granulocyte function by HA. fn vitro and in viva effects on phagocytoscs, locomotion and meta~iism, J. C/in. Invest. 1980, 66, 298-305 Forrester. J.V., Lackie, J.M., Effect of HA on neutrophii adheston, .I. Cell SC. 1981, 50. 329-344 Rolitns, B.J.. Gulp, L.A., Givcosaminoglycans tn the substrateadhesion sites of normal and wrus-transformed murtne cells, Biochemistry 1979,18, 141-148 Abatangelo, G., Coruvo. R., Martellt. M., Vecchta, P.. Cell detachment medtated by hyalurontc actd. Exp. Ce//. Res. 1982, 137. 73-78 Cort~vo. R.. De Galateo, A., Castellanr, I., Brun, P.. Gtro, M.G. and Abatangelo, G., Hyaluromc acrd promotes chtck embryo ftbroblast and chondroblast expresston, Cell Biol. fnt. Rep 1990,14(2), 1 1 1 - 122 Abatangelo, G., Born. P., Del Bue, M.. Get, G., Samson, J.C.. Contvo. Ft., De Galateo. A. and Martelli. M., Intraarttcular sodium-hyaluronate qectron in the pond-nuki experimental model of osteoarthntts tn dogs. I. Biochemccal results, Cl/n. Orthop. &?I. Res. 1989. 241, 278-285 Ashetm, A., Lindblad. G., lntraa~lcular treatment of arthntrs in race- horses with sodium hyaiuronate~~~a Vet. Stand 1976. f 7.379-394

Abatangeio, G., Martelk, M. and Vecchta, P., Healtng of hyalurontc acrd-enrtched wounds: histologtcal observatrons,J Surg. Res 1983, 35,410-416 Bragantim. A., Cassiw M.. De Basttani, G.. Perbellint. A., Controlled single blrnd tnal of tntraarttcularly Injected hyaluronic acid (Hyalgan”) tn osteoarthntts of the knee, Clin. Trials J. 1987. 24, 333-340 Dotilon. C.J.. Sliver, F.H., Collagen based wound dressmg: effects of hyaluromc acid and fibronectrn on wound healing, Biomatena’ials 1 986. 7. 3-8 Wetgel, P.H., Fuller, G.M.. Le Boeng. R., A model for the role of hyaluronic acid and fibnn tn the early events during the tnflammatory response and wound heaimg, J Theor. Biol. 1986, 119, 229-234 Rafadellt. A., Zinna. G., Studio clinico sull’attcvita dell’acido ialuronrco nella terapia delle ulcerazioni degii am tnfenori, Chron. Dermafof. 1986, XVII(Z), 251-258

Bromatenals 199 1, Vol 12 October 729

Biocompatibiiity of hyaluronic acid esters: R. Cortivo et al.

18 Della Valle. F. and Romeo, A., Hyaluronrc acid esters and their medical Advances in Biomaterials, Vol. 9, (Eds G. Heimke, U. Soltesz and and cosmetic uses and formulations, Eur. Patent Appl. EP 216453, A.J.C. Lee) Elsevier Science Publishers, Amsterdam.The Netherlands, 1987, April 1, 129 pp 1990, pp 199-205

19 Rastrelli. A., Beccaro, M., Biviano, F., Calderint, G. and Pastorello, A., 20 Hunt, J.A.. Joshi, H.N., Stella, V.J. andTopp, E.M., Diffusion and drug Hyaluronic acid esters, a new class of semisynthetic biopolymers: release in polymer films prepared from ester derivatives of hyaluronic chemical and physico-chemrcal properties, in C/in. implant Materials. acid, .I. Cont. Rel. 1990, 12, 159-l 69

730 Biomaterials 199 1. Vol 12 October