Molecular design of three-dimensional artificial extracellular matrix: Photosensitive polymers containing cell adhesive peptide

Molecular Design of Three-Dimensional Artificial Extracellular Matrix: Photosensitive Polymers Containing Cell Adhesive Peptide MlNOO JALILI M O G H A D D A M and TAKEHISA MATSUDA"

Department of Bioengineering, National Cardiovascular Center Research Institute, 5-7-1 Fujishirodai, Suita, Osaka 565, Japan

SYNOPSIS

A photodimerizable monomer, methacrylic acid- ( 7-coumaroxy ) ethyl ester, was synthesized and was copolymerized with a hydrophilic monomer ( N ,N-dimethylacrylamide) to obtain a water-soluble photosensitive polymer. Irradiation of the copolymer film and aqueous solution with a high-pressure mercury lamp resulted in highly hydrated gel. The gel yield was enhanced with the content of the photodimerizable group in the copolymer and the irradiation time. The degree of swelling of the gels decreased concomitantly. Incorporation of the well-known cell adhesive peptidyl ligand Arg-Gly-Asp-Ser ( RGDS) into photosensitive copolymers attained a biologically active hydrophilic gel matrix upon UV light irradiation. Irradiation of a buffer solution of the latter copolymer premixed with smooth muscle cells entrapped the cells throughout the gel matrix. This indicates that the designed polymer and the resulting cell-incorporated hydrogel are biomimic to a n extracellular matrix and to the media of the vascular wall, respectively. 0 1993 John Wiley & Sons, Inc. Keywords: extracellular matrix coumarin photodimerization photocrosslinking hydrogel adhesive ligand

INTRODUCTION Photosensitive polymers have received increased attention due to their broad applications, 1-6 particularly in coatings and photoresist technologies. Recent interests have been expanded to biotechnological field.7-13Previous research has documented the ability of photocrosslinked hydrogels to immobilize enzymes and microbial cells.'2913However, few studies have been conducted on the entrapment of mammalian cells within synthetic hydrogels. In this article, we utilized photochemical techniques for entrapping the cells via photogelation of a watersoluble photosensitive copolymer with cell adhesive sites, aiming for the development of an artificial three-dimensional ( 3-D ) extracellular matrix (ECM) which is able to entrap the cells and provide an artificial hybrid tissue. * To whom all correspondence should be addressed. Journalof Polymer Science:Part A Polymer Chemistry,Vol.31,1589-1597 (1993) 0 1993John Wiley & Sons, Inc. CCC 08S7-624X/93/061589-09

Living cells in tissues exist in 3-D hydrated ECMs, which are composed of an insoluble network of proteins and proteoglycans filling the intercellular spaces. Hydrogel-like structuring is attained by proteoglycans which function to hold water and generate osmotic pressure, which resists the compressive strength rendered by internal or external stresses. The major function of proteins is to serve as a cellular adhesive matrix on which cells adhere and grow. The adhesive matrices for cellular components include the so-called adhesive proteins, typified by collagen and fibronectin. Recent studies have shown that the cell adhesion proceeds via the binding of ligands of adhesive proteins to cell membrane receptors. The minimal active amino acid sequence of adhesive ligands common to adhesive proteins has recently been identified as Arg-GlyAsp (RGD; one-letter description of amino acid) tripeptide.14-16These are the fundamental requirements for the extracellular environment of tissues. In our previous studies, we have examined the hypothesis that the surface modification with a 1589

1590

MOGHADDAM AND MATSUDA

RGD-containing peptide enhanced cell adhesion and growth in the absence of adhesive proteins.l7-l9Such a cell-adhesive ligand-immobilized surface serves as a 2-dimensional (2-D) artificial ECM. The object of this study is the molecular design of a 3-D artificial ECM and the preparation of hybrid tissue in which cells are entrapped in a 3-D ECM. The design concept of the 3-D ECM under study include the incorporation of three functions: first, water solubility, i.e., artificial ECM should be watersoluble when mixed with cells; second, photocrosslinkability, i.e., ECM should be gelled via the solgel transformation after being mixed with cells; and third, cell adhesivity, i.e., the ECM should serve as an adhesive bed for cells. These functions are realized by preparing photocrosslinkable water-soluble copolymers with adhesive peptidyl ligand in the side chains.

EXPERIMENTAL Materials

Methacrylic Acid- (7-Coumaroxy) ethyl ester (MACEE)

To a 50-mL acetone solution of 7-hydroxy coumarin ( 1.62 g) , a saturated aqueous solution of potassium carbonate (2.8 g) was added and the reaction mixture was stirred overnight in the dark at room temperature. The solvent was removed under reduced pressure, and the residue was dissolved in dimethylformamide (DMF). The excess potassium carbonate was removed by filtration. Then, 2-bromoethyl methacrylate ( 3.9 g) , purchased from Polyscience Inc., was added to the filtrate, and the reaction mixture was stirred for 2 days a t room temperature. The solvent was removed under reduced pressure and the precipitate was thoroughly washed with diethyl ether. The precipitate was redissolved in chloroform, and washed with water three times, and the organic layer was dried over sodium sulfate. The monomer was characterized by TLC, but only one single spot without any trace of the dimer was observed. The product was crystallized from cyclohexane to give white needle crystals of MACEE in 80%yield (2.2 g) . IR ( KBr, cm-l): 3431,2978,2787, 1736,1716,1617,1508,1473,1450,1400,1354,1324, 1290,1230,1181,1161,1131, 1096,1052,1002,952, 943,918,893,820,769,754,650, and 635. 'H-NMR ( 270 MHz, in dimethylsulfoxide (DMSO) -de a t 30°C, ppm): 2.0 (s, 3, CH3), 4.4 (t, 2, CHZ), 4.45 (t,2, CHZ), 5.7 ( s , 1, CH2=),6.02 ( s , 1,CH2=),

6.35 (d, 1,5 H ) , 7.0 (d, 1, 6H), 7.05 (s, 1, 8 H ) , 7.65 (d, 1,4H),8.05 (d, 1,3H). ANAL.Calcd for C15H1405: C, 65.68%;H, 5.14%. Found C, 65.25%;H , 5.18%.

Octapeptide (Gly-Gly-Gly-A rg-Gly-Asp-SerPro), GGGRGDSP

The octapeptide was synthesized with t-Boc chemistry in the solid phase using the Applied Biosystem Model 430A peptide synthesizer (San Francisco, CA) . Trifluoromethane sulfonic acid was used for cleavage and removal of the peptide from solid resin and deprotection of the side chains. The peptide was purified by gel filtration (Sephadex LH20, Pharmacia, Uppsala, Sweden; 2 X 180 cm in methanol). The octapeptide GGGRGDSP showed only one single spot on thin layer chromatography (silica gel, eluent; chloroform-methanol-acetic acid; 95/ 15/ 3 by volume) after purification on a Sephadex column.22In addition, 24 h hydrolyzate of the peptide was subjected to amino acid analysis on an automated Beckman amino acid analyzer.22The purify of the peptide was > 95%. Copoly(DMAAm, MACEE), I

A DMF solution of N, N-dimethylacrylamide (DMAAm; 1g) and MACEE (0.145 g) and azobisisobutyronitrile ( AIBN, 11 mg) were put in a pyrex tube. After three freeze/thaw cycles, the tube was sealed under vacuum. Polymerization was conducted in a thermostated bath at 60°C. The obtained copolymer was precipitated in 500 mL diethyl ether. After three reprecipitation operations the product was filtered off and dried under vacuum. At 95/5 feed molar ratio of DMAAmlMACEE, 0.6 g of copolymer I was obtained (52 wt % yield). IR ( KBr, cm-'): 3447, 2934, 1731, 1645, 1621, 1560, 1508, 1457,1402,1354,1283,1257,1235,1199,1138,1068, 996, 841, 635, 616. Copoly(DMAAm, MACEE, NASI), II

The copolymerization of DMAAm, MACEE, and N acryloxy succinimide (NASI),which was purchased from Kokusan Chem. Ind. (Tokyo),was conducted according to the same method as mentioned above. The polymerization conditions are given in Table I. IR (KBR, cm-l): 3451,2943,1813,1783,1742,1626, 1558,1506,1453,1435,1405,1360,1258,1206,1142, 1067, 996,842, 812, 719,640, 621.

MOLECULAR DESIGN OF 3-D ECM

1591

Copoly (DMAAm, MACEE, AcGGGRGDSP), 111

Photogelation Characteristics

To 5 mL DMF solution of the copolymer I1 (0.55 g ) and octapeptide, H-Gly-Gly-Gly-Arg-Gly-AspSer-Pro-OH (GGGRGDSP ) (0.33 g) and triethylamine (0.13 ml) were added, and the reaction mixture was stirred at room temperature for 7 days. After the solvent was removed under reduced pressure, 30% aqueous ethylamine (20 mL) was added to the precipitate and stirred overnight. It was dialyzed on the cellulose membrane (SPECTRAPOR molecular tubing; MWCO 1000-Wet Tubing) for 2 days. After removal of the solvent under reduced pressure, the obtained copolymer was redissolved in DMF and was precipitated in diethyl ether. The reprecipitation was repeated three times. The product was filtered off and dried in vacuum. The resulting yield was 0.326 g of the peptidyl copolymer ( 111).IR (KBr, cm-'): 3451, 3082, 2943, 1731, 1627, 1565, 1540,1509,1459,1402,1359,1258,1145,1058,1032, 995, 844, 638, 619.

The photodimerization kinetics of polymer films were measured as follows: Thin films of copolymer I with different contents of the coumarin group were obtained by casting their methanol solutions on quartz plates. The solutions were adjusted to produce films with an optical density of 1.2 at their absorption maximum. After being dried at ambient temperature, the films were exposed to the cold UV light from a high pressure mercury lamp (Hitachi, 400 W ) , which was passed through a water filter in a Pyrex vessel ( I = 0.41 mW/cm2). The UV spectra of the films were immediately recorded with a UV spectrophotometer. Measurements of the degree of gelation and swelling of the polymer films were conducted as follows: Films of the same thickness, ca. 100 pm, were cast on glass slides (diam: 15 mm) from their methanol solutions, and were dried at ambient temperature. They were subsequently exposed for different intervals to the cold UV light from a high-pressure mercury lamp as mentioned above. After irradiation, the films were immersed in 50 mL methanol for 24 h. The insoluble photogels were filtered with a glass filter under reduced pressure, and washed with 30 mL fresh methanol. The photogels were then dried under vacuum, and the methanol solution was pooled and its UV absorbance a t 320 nm was measured. The weight of the polymer which was soluble in methanol was determined from UV absorbance of the methanol solution a t 320 nm. The degree of gelation was defined as the weight ratio of the dried, water-insoluble photogel to the initial photocured film. The degree of gelation was alternatively calculated from the insoluble gel weight which was obtained by subtracting the weight of the methanol-

Instrumentation

IR spectra were measured with a Nicolet FTIR 5DX spectrophotometer (Madison, USA). 'H-NMR spectra were obtained on a Jeol JNM-GX270 FTNMR spectrometer (Tokyo, Japan). The UV spectra were recorded with a Jasco Ubest-30 UV/ VS spectrophotometer (Tokyo, Japan). The molecular weight of the polymers were measured by gel permiation chromatography (GPC) using a Toyo Soda HPLC (8012 Series) with a thermostated column (TSK gel G6000PW XL and TSK gel G3000PW X L ) ,UV and RI detectors. Polyethylene glycol was used as the standard polymer for calibration of the column. Table I.

Preparation of Hydrophilic Photosensitive Polymer with Pendant Coumaryl Group [DMAAm]"

(mol/L)

[MACEE]' (mol/L)

[NASI]" (mol/L)

[AIBN] (mol/L)

Time (h)

T ("C)

Yield (%)

M,"

-

0.0055 0.0055 0.0055 0.0055

2 2 2 2

60 60 60 60

52.5 53.6 52 52

28,200 19,300 8500 8300

0.09

0.0055

2

60

52

16,200

Copolymer I1 (DMAAm/MACEE) 99/1 97/3 95/5 90/10

0.84 0.84 0.84 0.84

0.0084 0.0259 0.042 0.093

Copolymer I1 (DMAAm/MACEE/NASI) 85/5/10 a

0.75

0.042

Monomer concentration in feed. Determined by GPC measurement.

1592

MOGHADDAM AND MATSUDA

I

II

CH2= C-C- OCH2 CH2Br Scheme 1.

soluble fraction from the initial weight of the photocured films. The dried photogels were then immersed in deionized water for 24 h. The degree of swelling was defined as the weight ratio of the water in the equilibrated swollen gel to the dried photogel. Entrapment of Smooth Muscle Cells

A volume of 0.5 mL of 20% phosphate buffer solution of peptidyl terpolymer I11 was mixed with 0.5 mL of Dulbeco's modified Eagle's medium (DMEM) containing bovine smooth muscle cells ( SMCs) at a concentration of 2.5 X lo4 cells/mL, which were derived from a bovine aortae. The mixed solution was exposed to irradiation of a high-pressure mercury lamp passed through UV-D-36C, UV-29, and water filters. Every 24 h, 0.5 mL of DMEM supplemented with 10% calf serum was added to the gel, and the culture was allowed to proceed up to 10 days.

RESULTS AND DISCUSSION The molecular design of 3-D ECM in this study is based on a bioactive hydrogel obtained by photoirradiation of a copolymer with three different func-

Feed Mole% of Photosensitive Monomer in Copolymerization

Figure 1. Relationship between content of photosensitive group in copolymer I and photosensitive monomer ratios in the feed.

tions: water solubility, cell adhesivity, and photocrosslinkability. In the first part of this study, the synthesis of hydrophilic photocurable copolymer, its photogelation, and degree of swelling are described. In the latter part, the synthesis of a terpolymer with photocurable and cell adhesive functions, and the entrapment of cells are described. Table I lists the reaction conditions and compositions of the copolymers obtained. Preparation of Hydrophilic Photosensitive Copolymer

The chemistry utilized for photocrosslinking of hydrophilic copolymers is based on photodimerization of reactive functional groups which undergo the [ 2 21 cycloaddition reaction between paired functional groups such as cinnamoyl and coumarin residues." The photosensitive monomer having a coumarin residue, MACEE, was synthesized according to Scheme 1.This monomer prepared from 7-hydroxy coumarin and 2-bromoethyl methacrylate was obtained in high yield. To assess photoreactivity and factors controlling the gelation via photodimerization, two-component copolymerswere prepared from

+

Wavdength(m)

Figure 2. Changes in the UV spectra of a thin film of copolymer I during photoirradiation with a high-pressure mercury lamp.

1593

MOLECULAR DESIGN OF 3-D ECM yn3

+ c n z 1- c n ~ c t 4 I2 - c ~ c=o N b c n 2 - c n 2 - o ~ c=o cH 3 +cHz-cn*Cn2-c* c=o c=o N & n 2 - c n0 2-o~~o cn: \CH,

hv

cn;

\CH3

cn2-CH2-0 c=o fCnz-Cn*Cn2-h* I c=o cn3 cn:

A

\cn3

Scheme 2.

the photoreactive monomer (MACEE ) and hydrophilic monomer (DMAAm) as follows. Radical copolymerization of MACEE with DMAAm in DMF solution resulted in a viscous solution a t ca. 50% conversion after 2 h polymerization, but yielded a gel after 3 h. Therefore, the polymerization time was restricted to 2 h to obtain water-soluble photosensitive copolymers. The gelation occurring a t longer polymerization times might be due to contamination of a very small amount of the dimer of MACEE, which could not be detected by IR or NMR spectroscopy. The content of the photosensitive group in the copolymers was determined from the UV absorbance at 320 nm ( c = 16200), which is ascribed to the coumarin group, and was plotted against the photosensitive monomer ratio in the feed (Fig. 1). A fairly good linear relationship between the copolymer composition and feed monomer ratio was obtained, indicating that the content of photosensitive moieties in the copolymer was approximately the same as that in the feed.

v

Photochemical Gelation of Cast Film

A typical UV spectral change of a thin film of copolymer I during irradiation is shown in Figure 2. The maximum absorbance at 320 nm decreased with an increase in the irradiation time, indicating that photodimerization of the coumarin groups via concerted [ 2 2 ] cycloaddition of the C =C double bond proceeded time-dependently to form a cyclobutane-type ring (Scheme 2 ) . The amount of remaining unreacted coumarin group was calculated from the absorbance of irradiated films at 320 nm. To determine quantitatively the dependency of the coumarin content on the degree of photodimerization, the maximum absorbance of the copolymer films was kept constant, regardless of the content of the coumarin group in the copolymers. This was achieved by adjusting film thickness. As shown in Figure 3, the dimerization reaction occurred a t a high rate in the early stages of irradiation, followed by a low rate. The photodimerization rate was enhanced with an increase in the content of the coumarin group in the copolymers.

+

254nm

0

10

20 30 40 50 Irradiation Time (min)

60

7

Figure 4. Degree of gelation of copolymer I films upon UV irradiation (degree of irradiation was determined from the weight of insoluble gel). DMAAm/MACEE: ( A )991 1, ( W ) 97/3, ( 0 )95/5, (V)90/10.

1594

MOGHADDAM AND MATSUDA

This indicates that the degree of photodimerization is enhanced with the concentration of the coumarin group, probably due to increased density of pairs of hydrophobically associated coumarin groups. It has been reported that the coumarin photodimer is cleaved by UV irradiation at lower wavelength.'l The subsequent exposure of copolymer films to a low-pressure mercury lamp (254 nm) showed a time-dependent increase in absorbance at 320 nm, indicating that the coumarin group is regenerated due to photocleavage of the cyclobutane dimer (Figure 3). However, the dissociation reaction reached a photoequilibrium state since the absorption of coumarin overlaps with that of dimer on the wavelength of the light used. The polymer films cast from methanol solutions were irradiated a t different periods, and then the irradiated films were immersed in methanol. The insoluble polymer was filtered off and dried in vacuum. The degree of gelation (DG ) determined from the weight of the insoluble polymer was plotted against the irradiation time (Fig. 4). The DG was also determined by measuring the methanol-soluble fraction of polymers, which was determined by measuring the UV absorbance of the methanol solution at 320 nm (Fig. 5 ) . DGs calculated from UV-absorbance of soluble parts were found to be higher than those obtained from the insoluble gels. This may be attributed to the occurrence of intramolecular dimerization of soluble parts. Higher content of coumarin groups in copolymers provided higher DG. This was in accordance with the enhanced photodimerization rate determined from the UV absorbance, as previously shown in Figure 3. The degree of swelling (DS) of gels in water was plotted against the irradiation time (Fig. 6 ) . Higher content of the

v

t

I

10

0

20

30

40

50

60

Irradiation Time (min) Figure 6. Degree of swelling of the gelled copolymer I films as a function of the irradiation time. DMAAm/MACEE: ( A )99/1, (B) 97/3, ( 0 )95/5, (V)90/10.

coumarin group and longer irradiation time resulted in less swelling.

Photogelation in Aqueous Solution The aqueous-phase-photogelation was conducted at various concentrations of polymer under 30 min irradiation. Figure 7 shows DG dependency on initial water content, in which the horizontal axis was expressed as the water /polymer weight ratio ( w t % ). The increase in the coumarin content in copplymers resulted in a higher degree of gelation. Irrespective of the content of the coumarin group, the maximum yield of gels was obtained at ca. 10% of the polymer concentration or 10-fold of water content by weight.

P

10 0

10

20 30 40 50 irradiation Time (min)

60

70

Figure 5. Degree of gelation of copolymer I films upon UV irradiation (degree of gelation was determined by UV absorption of the soluble part). DMAAm/MACEE: ( A ) 99/1, (B) 97/3, ( 0 )95/5, (V)90/10.

0

1000 3000 Water Content (%)

5000

Figure 7. Degree of gelation upon 30 min UV irradiation as a function of water content in aqueous mixture. DMAAm/MACEE: (A)99/1, (B) 97/3, ( 0 )95/5.

MOLECULAR DESIGN OF 3-D ECM

i 5000

Water Content (%)

Figure 8. Degree of swelling of photogels obtained upon UV irradiation of the aqueous solution of copolymer I DMAAm/MACEE: ( A )99/1, ( W ) 97/3, ( 0 )95/5.

Moreover, the polymer containing 5 mol % of the photosensitive group attained gel form at 100%yield at this concentration. Irrespective of the coumarin content, the higher DG found at 10% of the polymer concentration may indicate that intermolecular hydrophobic association is facilitated in a more polar environment such as water, as compared with those in solid films. Further dilution seems to reduce the degree of intermolecular hydrophobic association, resulting in a sharp drop in DG, as found in Figure 7. The increase in the content of the coumarin group above 5 mol % in the copolymer drastically reduced the water solubility due to the hydrophobicity of the coumarin groups. The DS of photogels were plotted against the initial water content when irradiated (Fig. 8). As expected, smaller DG resulted in higher DS. Thus, it is concluded that DS is controlled either by the content of the coumarin group, by the polymer solution concentration, or by irradiation time. These are all controlling factors of DS and gelation rate. Bioactive Hydrogel and Entrapment of Smooth Muscle Cells

The water-soluble photocrosslinkable polymer derivatized with bioactive peptide was synthesized by the terpolymerization of DMAAm, MACEE, and

1595

NASI, and by subsequent incorporation of the cell adhesive peptidyl sequence into the polymer side chains (Scheme 3 ) . The incorporation of the octapeptidyl ligand GGGRGDSP where G, R, D, S, and P denote glycine, arginine, aspartic acid, serine, and proline residues, respectively, was achieved through the activated ester, succinimidyl group, of the NASI unit. The incorporation of the peptidyl ligand was confirmed by ‘H-NMR spectroscopy. No appreciable peak at low molecular weight region was observed by GPC, although a slight increase in molecular weight was appreciated. This indicates that nonbound peptide was removed by dialysis. In this experiment, a quite high feed molar ratio of the octapeptide to the activated ester was employed to assure the coupling of the peptide onto the photoreactive copolymer. Due to high cost of the peptide, an optimal reaction condition, which we did not attempt, could reduce the cost. A buffer solution of peptidyl terpolymer I11 resulted in a hydrogel upon UV irradiation similar to copolymer 11. To incorporate the smooth muscle cells into the above hydrogel, a phosphate buffer solution of the bioactive terpolymer, which was mixed with bovine smooth muscle cells ( SMCs) in DMEM, was exposed to the irradiation of UV light. Phase contrast microscopic photographs of the hydrogel produced by 30 min irradiation showed that SMCs were entrapped in the transparent gel. No appreciable shape change was observed within several hours after gelation. Entrapped SMCs spread and elongated in the gel in a few days. Further culturing up to 10 days resulted in very slow proliferation of SMCs (Fig. 9). It has been well known that irradiation of UV light on living tissues or cells results in metabolic disorder or death, depending on the irradiation power and time. It is assumed that a slow proliferation observed is derived from reduced viability due to UV irradiation. In addition, 3-D culture environment where cells exhibit usually slow growth rate, as exemplified by cell culture in collagen gels, might be also responsible for slow growth rate observed here. The result shows that the resultant hydrogel with SMCs can be considered to be a hybrid

c n3

cnl

(0;GIy. R;Arg, D;Asp, S;Ser, &Pro)

Scheme 3.

1596

MOGHADDAM AND MATSUDA

Figure 9. Three-dimensionally trapped SMC in the synthetic extracellular matrix: ( a ) SMC in the buffer solution of copolymer 111, ( b ) 3 h after gelation, ( c ) 2 days culture, ( d ) 10 days culture.

tissue. Further studies on quantitative analysis of peptide content, crosslinking density, and hybrid tissue formation are now in progress, which will be reported in a forthcoming article.

CONCLUSIONS Hydrophilic copolymers with photodimerizable coumarin groups were synthesized. UV irradiation resulted in the formation of a transparent hydrogel. Gel yield, gelation rate, and water swelling were found to be controlled by the coumarin content, irradiation time, and polymer concentration. As an artificial model of an extracellular matrix, the adhesive peptidyl sequence ( Arg-Gly-Asp-Ser ) was incorporated into the side chain of photocrosslinkable copolymers. Vascular smooth muscle cells were entrapped in the three-dimensional bioactive hydrogel.

REFERENCES AND NOTES 1. L. M. Minsk, J. G. Smith, van W. P. Deusen, and J. F. Wright, J. Appl. Polym. Sci., 2 , 302 (1959). 2. M. Kato, T. Ichijo, K. Ishii, and M. Hasegawa, J. Polym. Sci. A-1, 9 , 2109 (1971). 3. S. Reiser and P. L. Egerton, Photo. Sci.' Eng., 23,144 ( 1979). 4. P. L. Egerton, E. Pitts, and A. Reiser, Macromolecules, 1 4 , 9 5 (1981). 5. A. Reiser and E. Pitts, J. Photo. Sci., 29,187 (1981). 6. C. Azuma, K. Sanui, and N. Ogata, J. Appl. Polym. Sci., 27,2065 (1982). 7. K. Ichimura, J. Polym. Sci. Polym. Chem. Ed., 20, 1411 (1982). 8. K. Ichimura and S. Watanabe, J. Polym. Sci. Polym. Chem. Ed., 2 0 , 1419 (1982). 9. K. Ichimura and S. Watanabe, J. Polym. Sci. Polym. Chem. Ed., 18,891 (1980). 10. K. Ichimura, J. Polym. Sci. Polym. Chem. Ed., 2 2 , 2817 (1984).

MOLECULAR DESIGN OF 3-D ECM

11. K. Ichimura, J. Polym. Sci. Part A: Polym. Chem., 25,1475 (1987). 12. T. Omata, T. Ida, A. Tanaka, and S. Fukui, Eur. J. Appl. Microbiol. Bwtechnol., 1 8 , 143 (1979). 13. K. Sonomoto, A. Tanaka, T. Omata, T. Yamane, and S. Fukui, Biotechnol. Bioeng., 2 1 , 2133 (1979). 14. M. D. Pierschbacher and E. Ruoslahti, Nature, 3 0 9 , 30 (1984). 15. E. Rouslahti and M. D. Pierschbacher, Cell, 4 4 , 517 (1986). 16. E. Rouslahti and M. D. Pierschbacher, Science, 2 3 8 , 419 (1987). 17. T. Matsuda, A. Kondo, T. Akutsu, and K. Makino, ASAIO Trans., 3 5 , 6 7 7 ( 1989).

1597

18. T. Matsuda, E. Ozeki, and T. Akutsu, Znt. J. Artif. Organs, 1 4 , 1 1 0 (1990). 19. E. Ozeki and T. Matsuda, ASAIO Trans., 3 6 , 195 (1990). 20. Y. Chujo, K. Sada, and T. Saegusa, Macromolecules, 2 3 , 2693 (1990). 21. M. Hasegawa, Y. Suzuki, and N. Kita, Chem. Lett., 317 (1972). 22. S. Ito, T. Matsuda, Y. Takemoto, K. Yamamoto, T.

Kishimoto, and M. Maekawa, Int. J. Artif. Organs, 1 5 , 737 (1992).

Received February 24, 1992 Accepted October 29, 1992