Advanced Drug Delivery Reviews 31 (1998) 287301

Protein release from gelatin matrices

Yasuhiko Tabata, Yoshito Ikada*
Research Center for Biomedical Engineering, Kyoto University, 53 Kawahara-cho Shogoin, Sakyo-ku, Kyoto 606, Japan

Abstract Gelatin is a denatured, biodegradable protein obtained by acid and alkaline processing of collagen. This processing affects the electrical nature of collagen, yielding gelatin with different isoelectric points (IEPs). When mixed with positively or negatively charged gelatin, an oppositely charged protein will ionically interact to form a polyion complex. This review article describes protein release from charged gelatin matrices on the basis of this polyion complexation. The biodegradable hydrogel matrices are prepared by chemical crosslinking of acidic or basic gelatin and are enzymatically degraded in the body with time. The degradation is controllable by changing the extent of crosslinking, which, in turn, produces hydrogels with different water contents. The time course of protein release is in good accordance with the rate of hydrogel degradation. It is very likely that the protein drug complexed with gelatin hydrogel is released as a result of its biodegradation. This gelatin hydrogel system releases the protein drug under maintenance of biological activity. This article will focus on experimental data that sustained release of growth factor from the gelatin hydrogels is very effective in exerting the biological functions of the growth factor. 1998 Elsevier Science B.V. Keywords: Gelatin; Growth factor; Sustained release; Degradation; Polyion complexation; Neovascularization; Bone formation

Contents 1. Introduction ............................................................................................................................................................................ 2. Preparation of gelatin hydrogels ............................................................................................................................................... 2.1. Block matrices ................................................................................................................................................................. 2.2. Injectable matrices ........................................................................................................................................................... 3. Complexation of protein with gelatin ........................................................................................................................................ 3.1. Polyion complexation in aqueous solution.......................................................................................................................... 3.2. Interaction of protein with gelatin hydrogels ...................................................................................................................... 4. Degradation of gelatin hydrogels .............................................................................................................................................. 5. bFGF release from gelatin hydrogels ........................................................................................................................................ 5.1. In vitro release ................................................................................................................................................................. 5.2. In vivo release ................................................................................................................................................................. 6. Biological activity ................................................................................................................................................................... 6.1. Neovasuclarization ........................................................................................................................................................... 6.2. Bone formation ................................................................................................................................................................ 7. Conclusions ............................................................................................................................................................................ References .................................................................................................................................................................................. *Corresponding author. Tel.: 075-751-4115; fax: 075-751-4144; e-mail: yyikada@medeng.kyoto-u.ac.jp 288 290 290 290 291 291 292 293 293 293 294 295 295 296 297 298

0169-409X / 98 / $19.00 1998 Elsevier Science B.V. All rights reserved. PII S0169-409X( 97 )00125-7

288

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

1. Introduction Recent advances in biotechnology has made it possible to produce various clinically useful peptides and proteins. While this technology has brought about the discovery and mass production of these bioactive macromolecules, several challenges need to be addressed with regard to their sustained delivery in a convenient, controlled manner, and targeting formulations. In contrast to conventional synthetic pharmaceuticals, proteins are susceptible to proteolysis, chemical change and denaturation during storage and administration in the body [1,2]. Signicant efforts have been made to improve formulations for better stabilization of proteins over a sufciently long storage time. Additional research has focused on the development of dosage forms that either prolong the biological activity of protein in the body or assist in targeting the protein to a specic tissue. One possible way to prolong activity is to incorporate a protein drug into an appropriate matrix for achieving sustained release of the drug at the site of action over a long period of time. It is highly possible that protein is protected against proteolysis and antibody neutralization, as far as it is, at least, incorporated in a release matrix for prolonged retention of the protein activity in vivo. There have been a number of research reports on protein release from polymer matrices: poly( L-lactic acid) (PLLA) and its copolymers with glycolic acid (PLGA) [3 31], PLGA polymer blends [18,32,33], PLLApolyethylene glycol (PEG) copolymers [34,35], poly(cyanoacrylates) [36,37], poly(anhydrides) [3840], poly(ortho esters) [41,42], polyphosphazene [43], poly(vinyl alcohol) [44], poly(vinyl pyrrolidone) [45], poly(acrylic acid) [46], poly(ethyleneco-vinyl acetate) [47,48], cellulose derivatives [4951], hyaluronic acid derivatives [52,53], alginate [5458], collagen [5961], gelatin [6067], starch [68,69], dextran [70] and brin [71]. As is stated in other chapters of this special issue, the largest problem of protein release technology is the loss of biological activity of the protein released from a proteinpolymer formulation. Thus, unless this problem is solved by a breakthrough, it seems difcult to expect a further research development in the area of protein release. It has been demonstrated that this activity loss results from denaturation and deactivation of

protein during the formulation process with a polymer matrix. When exposed to harsh environmental changes, such as heating and exposure to sonication and organic solutions, protein is generally denatured, losing its biological activity [9,14,20]. Therefore, it is important to exploit a new formulation method of protein carrier with polymers under mild conditions to minimize protein denaturation. From this viewpoint, polymer hydrogel may be a preferable candidate as a protein release matrix because of its biosafety and its high inertness towards protein drugs [72]. However, sustained release of protein over a long time period will not be expected from hydrogels, since the release rate of protein from hydrogels is generally diffusion-controlled through aqueous channels in the hydrogels. Thus, for achieving effective protein release, it will be a key strategy to immobilize the protein drug to polymer carrier molecules constituting the hydrogel through some molecular interactions. For one trial, we have been attempting to take advantage of electrostatic interactions between protein and polymer molecules for the sustained release of protein from the polymer hydrogel. It has been well recognized in polymer science that a positively or negatively charged polyelectrolyte electrostatically interacts with an oppositely charged partner to form a polyion complex [73,74]. It seems unlikely that all of the ionic interactions between the two polyelectrolytes with many charged groups are dissociated at the same time. As a result, in contrast to low-molecular-weight electrolytes, stable bonding will occur between the oppositely charged polyelectrolytes, which will not be dissociated easily. In the research eld of pharmaceutical science, this polyion complexation is not a new technology but has been extensively explored for drug coating and encapsulation. The application of this polyion complexation, which we will describe here, is Drug complexation with polymer carriers. This is a new trial that will allow us to pharmaceutically modify a charged polymeric drug to increase its stability, targeting and sustained release, leading to enhanced therapeutic efcacy. Charged drugs available for this trial include proteins and oligo- and polynucleotides, while biodegradable polymers, such as proteins, polysaccharides and poly(amino acid)s, are applicable as the polymer

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

289

carriers. Another representative research eld of Drug complexation with polymer carriers that has been reported is gene therapy. It has been demonstrated that complexation with positively charged polymers enabled negatively charged DNA to have an enhanced stability and transfection efciency to cells [7577]. However, it is unclear whether or not such a formulation also functions as a matrix for sustained release of polynucleotides. On the other hand, few applications have been reported on polyion complexation for sustained release of macromolecular drugs from polymer matrices. Although low-molecular-weight pharmaceuticals have been shown to release from polymer matrices on the basis of their ionic interaction [7880], this is, however, different from polyion complexation. Fig. 1 shows a conceptual scheme of protein drug release from a biodegradable polymer carrier on the basis of polyion complexation. A positively charged protein drug is electrostatically complexed with negatively charged polymer chains, constituting a carrier matrix. If an environmental change, such as increased ionic strength, occurs, the complexed drug will be released from the drugcarrier complex. Even if such an environmental change does not take place, degradation of the polymer carrier itself will also lead to drug release. Because the latter is more likely to happen in vivo than the former, it is preferable that the drug carrier is prepared from biodegradable polymers. The prole of drug release in this drugcarrier system is regulated by the change of carrier biodegradation. When we make use of polyion complexation for sustained release of a protein drug, it is absolutely necessary to employ a highly bio-safe polyelectrolyte as the carrier matrix. In addition, if biodegradability

is required for the carrier, the material to be used will be restricted to natural polymers with charged groups, such as proteins and polysaccharides. Therefore, as the carrier polymer, we have selected biodegradable gelatin, which is extensively used for industrial, pharmaceutical and medical purposes. The biosafety of gelatin has been proved through its long clinical usage as a plasma expander, in surgical biomaterials and as an ingredient in drugs [81]. Another unique advantage of gelatin as a drug carrier is the electrical nature of gelatin, which can be changed by the collagen processing method [82]. For example, the alkaline process, through hydrolysis of amide groups of collagen, yields gelatin with a high density of carboxyl groups, which makes the gelatin negatively charged. This reduces the isoelectric point (IEP) of gelatin. In contrast, the electrostatic nature of collagen is hardly modied through the acid process because of a less invasive reaction to amide groups of collagen. As a result, the IEP of the gelatin that is obtained will remain similar to that of collagen. In other words, a variety of gelatin samples with different IEP values are available (Fig. 2). If a protein to be released is acidic, basic gelatin with an IEP of 9.0 is preferable as the carrier material, while acidic gelatin, with an IEP of 5.0, will be applicable to the sustained release of a basic protein. Both gelatins are insolubilized in water to prepare a hydrogel through chemical crosslinking, for instance, with water-soluble carbodiimides and glutaraldehyde. It was reported that a model protein

Fig. 1. Release of protein drug from biodegradable polymer carrier on the basis of polyion complexation.

Fig. 2. Preparative process for acidic and basic gelatins from collagen.

290

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

could be immobilized into albuminheparin microspheres [83] or into a carrier of non-biodegradable synthetic polymer [84], through polyion complexation, and that this protein was released from the carriers upon environmental change. However, these experiments were conducted under in vitro conditions and the biological activity of the protein released was not determined. Edelman et al. [54] reported one trial of sustained release of basic growth factor by using heparin incorporated into alginate beads. The sustained release of various bioactive proteins from a collagen matrix has also been investigated [61,85,86] and it has been shown that protein release was regulated by collagen swelling, but the contribution of ionic interactions between the proteins and collagen was not studied. Protein release from charged polysaccharides is discussed in another chapter in this issue. Since research on protein release based on polyion complexation has just started, this article mainly describes the preparation of biodegradable hydrogels from gelatin with two different IEP values and their efcacy as a sustained release carrier of a bioactive protein, together with our current ndings on hydrogel degradation and protein release.

water to deactivate and remove the unreacted WSC. The resulting hydrogels could be shaped into disks, cubes or strips by punching out or cutting them with a razor. Hydrogel tubes could also be prepared by chemically crosslinking the gelatin in a tube-shaped mold. The hydrogels prepared were thoroughly rinsed with double-distilled water, freeze-dried and sterilized using ethylene oxide gas. No big change in hydrogel shape was observed before and after freezedrying and sterilization. As a measure to evaluate the extent of crosslinking of gelatin hydrogels, their water content was determined from the hydrogel weight before and after swelling in a phosphatebuffered saline solution (PBS, pH 7.4) for 24 h at 378C and expressed as the weight ratio of water in hydrogel to the whole wet hydrogel [65]. The water content of gelatin hydrogels decreases with an increase in the concentration of crosslinking agent and gelatin and with the reaction time. The water content can be changed over the range from 98 to 85 wt%, irrespective of the gelatins IEP and the type of crosslinking agent.

2.2. Injectable matrices

Surgical invasion is required to implant block matrices in the body. However, this invasive operation can be avoided, as gelatin hydrogels can be easily formulated into injectable shapes. For example, hydrogel microspheres can be prepared through GA crosslinking of gelatin after an aqueous solution of gelatin is dispersed in an oil phase [67]. Briey, aqueous gelatin solution is preheated and then homogenized or sonicated at different input powers to yield a water-in-oil emulsion. The temperature of the emulsion is lowered to approximately 208C, followed by further continuous stirring for 30 min to complete gelation of the gelatin solution in the dispersed phase. After gelatin dehydration by the addition of acetone, the resulting microspheres are collected by centrifugation, washed with acetone and isopropyl alcohol, and dried under vacuum. The non-crosslinked, dried gelatin microspheres are placed in an aqueous solution containing various amounts of GA and are stirred to facilitate their crosslinking. The resulting microspheres are treated with glycine, washed with distilled water and, nally, freeze-dried. The size of the microspheres can be

2. Preparation of gelatin hydrogels

2.1. Block matrices

The gelatin samples used were acidic gelatin with an IEP of 5.0 and basic gelatin with an IEP of 9.0, isolated from bovine bone by the alkaline process and from pig skin by the acid process, respectively. Both of them were chemically crosslinked to prepare gelatin hydrogels with different biodegradation rates. Briey, various amounts of glutaraldehyde (GA) [66] or a water-soluble carbodiimide (WSC) [65] were added to aqueous gelatin solutions and the crosslinking reaction was allowed to proceed at 48C for various time periods. Following the crosslinking reaction, GA-crosslinked hydrogels were immersed in an aqueous solution of glycine at 378C for 1 h, to block residual aldehyde groups of GA, and then were rinsed with water. On the other hand, WSC-crosslinked hydrogels were immersed for 1 h in an aqueous HCl solution (pH 3.0) and washed with

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

291

regulated by altering the preparative conditions, e.g. the gelatin concentration and the input power in the emulsication reaction. Optical microscopic observation reveals that GA-crosslinked gelatin hydrogel microspheres are all spherical, with average diameters ranging from 3 to 200 mm, irrespective of the IEP of the gelatin. The water content of gelatin hydrogel microspheres is in the range from 98 to 85 wt%, depending on the concentration of gelatin and GA during the preparation of the microspheres. Even if the gelatin hydrogel is of the block type, the hydrogel with a high water content is soft when swollen in water and, hence, can be extruded from a needle-attached syringe at a clinically acceptable pressure. Such a uid hydrogel will be applicable as an injectable carrier, similar to microspheres.

3. Complexation of protein with gelatin

3.1. Polyion complexation in aqueous solution

As a simple method to evaluate polyion complexation between gelatin and protein, turbidity of the mixed solution is measured at different temperatures [87]. Proteins with different IEP values can be used for polyion complexation with charged macromolecules: Bovine milk lactalbumin (MW 5 14,400;

IEP 5 4.3), soybean trypsin inhibitor (MW 5 21,000; IEP 5 4.4), bovine pancreas trypsinogen (MW 5 24,000; IEP 5 9.3), chicken egg lysozyme (MW 5 11,400; IEP 5 11.0) and basic broblast growth factor (bFGF; MW 5 17,000; IEP 5 9.6). Fig. 3 shows turbidimetric titration curves for basic bFGF at various concentrations of aqueous solution containing acidic and basic gelatins at 258C. When acidic gelatin was used for the titration, the mixed solution exhibited a turbidity maximum and the turbidity increased with complexation time. In contrast, basic gelatin does not cause an increase in the turbidity of a mixed solution, irrespective of the bFGF concentration. A similar trend of increasing solution turbidity with time was observed for basic trypsinogen and lysozyme, but not for acidic lactalbumin and trypsin inhibitor. In contrast, solution mixtures formed from basic gelatin and acidic proteins become turbid, in contrast to those between basic gelatin and basic proteins. This nding indicates that a polyion complex of basic (or acidic) protein is formed with the acidic (or basic) gelatin but not with the basic (acidic) gelatin. High-performance liquid chromatography (HPLC) with a heparin afnity column (HPLC) can be used to evaluate the change in heparin afnity of bFGF accompanied by gelatin complexation. The peak area of intact bFGF decreases with complexation time and

Fig. 3. Turbidimetric titrations of various concentrations of bFGF in 1 / 15 M phosphate-buffered solution (pH 7.0) containing 5.0 mg / ml of gelatin at 258C. (A) Acidic gelatin with an IEP of 5.0, (s) 6, (d) 12, (n) 24, (m) 48 and (h) 72 h after mixing the gelatin and bFGF and (B) basic gelatin with an IEP of 9.0, (s) 6, (d) 12, (n) 24, (m) 48 and (h) 72 h after mixing the gelatin and bFGF.

292

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

a new peak appears at a shorter retention time, when bFGF is mixed with acidic gelatin. In contrast, mixing bFGF with basic gelatin does not affect the HPLC peak of intact bFGF, indicating that the afnity of bFGF for heparin has decreased due to the formation of a polyion complex with acidic gelatin, resulting in a shortened retention time. On the other hand, basic gelatin does not interact ionically with bFGF, which is the reason for the lack of a new peak attributable to complex. The turbidity of the mixed solution from basic bFGF and acidic gelatin increases with complexation time, irrespective of the ionic strength of the solution, however, the higher the ionic strength of the mixed solution, the less signicantly the turbidity increases. A HPLC study has revealed that the peak area of the intact bFGF, 24 h after mixing with gelatin, decreases with a decrease in the ionic strength, and reaches zero when double-distilled water is used to mix. The peak area of the bFGFgelatin complex increases with decreasing ionic strength. Thus, complexation between bFGF and acidic gelatin is weakened by an increase in the ionic strength of the mixed solution. It is obvious that the electrostatic interaction between bFGF and gelatin mainly contributes to the polyion complexation.

Fig. 4. Inuence of the temperature of the solution on bFGF sorption into hydrogels prepared from gelatin with IEP values of 5.0 (open marks) and 9.0 (solid marks) in water at (s, d) 4, (n, m) 25 and (h, j) 378C.

3.2. Interaction of protein with gelatin hydrogels

The interaction of protein with gelatin results in protein sorption to the gelatin hydrogel. Fig. 4 shows the time course of bFGF sorption to hydrogels prepared from gelatin with IEP values of 5.0 and 9.0 at different temperatures. Basic bFGF is sorbed to the hydrogel of acidic gelatin with an IEP of 5.0 at 4 and 378C with time, in contrast to basic gelatin hydrogel. When bFGF sorption is examined in aqueous solution containing different amounts of NaCl, the amount of adsorption of bFGF decreases with an increase in the NaCl concentration, although the addition of NaCl does not affect the time prole of bFGF desorption (bFGF release). No bFGF sorption is observed for basic gelatin hydrogels, irrespective of whether or not NaCl is added. When acidic gelatin hydrogels with water contents ranging from 98 to 85 wt% are used, the time prole of bFGF sorption to every hydrogel is similar, irrespective of the water content. Probably, the bFGF

molecule, with a molecular weight of 17,000, is small enough to freely diffuse into the interior of hydrogels with water contents of more that 87 wt% [88]. It should be noted that bFGF sorption to gelatin hydrogels depends on the temperature of the solution. The amount of bFGF sorbed to acidic gelatin hydrogels tends to increase with temperature. In addition, bFGF sorption to basic gelatin hydrogels was observed at 378C following an initial period when there was no release, although no bFGF was sorbed at lower temperatures. The sorption prole of bFGF to hydrogels of poly(acrylic acid), which is a synthetic polymer that has much more negative charges than gelatin, does not depend on the temperature of the solution. As it is well recognized that electrostatic interaction does not depend theoretically on temperature, this nding indicates that bFGF sorption to poly(acrylic acid) hydrogels is primarily driven by electrostatic interaction. However, the signicant temperature dependence of bFGF sorption to acidic gelatin hydrogels suggests that other factors, such as conformational changes in the gelatin molecules, are also contributing to the interaction. The rate of bFGF sorption to acidic gelatin hydrogels is much lower than that to poly(acrylic acid) hydrogels when compared at a similar water content. However, when the amino groups of gelatin are chemically converted to carboxyl groups, bFGF sorption to the carboxylated gelatin tends to increase

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

293

with an increase in the extent of carboxylation. This indicates that the relatively slow sorption of bFGF is due partially to the small amount of negative charges in one gelatin molecule. Scatchard analysis has demonstrated that the dissociation constant (K d ) of bFGF for the acidic gelatin hydrogel is 6.75 3 10 27 M. It is reported that the K d of bFGF to a low afnity receptor, heparan sulfate, is 2.0 3 10 29 M [89]. It may be concluded that the initial driving force of bFGF sorption to the acidic gelatin hydrogel is electrostatic interaction between the two molecules, although it is not as strong as the bFGFheparan sulfate interaction but is sufcient to form a polyion complex. As is well recognized, FGF has a strong afnity for acidic polysaccharides, such as heparin and heparan sulfate, probably resulting in protection of FGF from denaturation and enzymatic degradation in vivo [90,91]. This proposed hydrogel formulation mimics the natural manner in which bFGF is stored in the extracellular matrix. It is likely that gelatin complexation enhances the in vivo stability of bFGF, similar to the bFGFacidic polysaccharide complexation that occurs in biological systems.

Fig. 5. In vivo degradation proles of bFGF-incorporating acidic gelatin hydrogels with water contents of (s) 98.8 and (d) 96.9 wt% [65].

4. Degradation of gelatin hydrogels Since gelatin hydrogels undergo enzymatic hydrolysis in the body, it is too difcult to evaluate their degradation prole under in vitro conditions without any enzymes. Even if enzyme is present in the test solution, the in vitro result cannot simulate the in vivo prole of hydrogel degradation because the type and concentration of enzymes for collagen hydrolysis are not clear. Thus, gelatin hydrogels were subcutaneously implanted into the backs of mice and the weights of the hydrogels were measured at different time intervals to evaluate the time prole of in vivo hydrogel degradation [92]. The weight of the hydrogel was found to decrease with implantation time and, nally, the mass disappeared from the implantation site, indicating that the hydrogels were degraded in vivo. The degradation period for hydrogels depends on their water content (Fig. 5). The higher the water content of the hydrogels, the faster they degrade. Following the subcutaneous implantation of 125 I-labeled gelatin

hydrogels, the remaining radioactivity of the hydrogels was also measured, to evaluate the in vivo rate of hydrogel degradation. This study has again demonstrated that the radioactivity in hydrogels with higher water contents decreases faster than in those with lower water contents. The time prole of the loss of radioactivity is in good agreement with the weight loss of hydrogels. Maintenance of the degradation period of hydrogels from ve days to ve weeks is possible if the water content is changed, and incorporation of bFGF into gelatin hydrogels does not affect their prole of in vivo degradation. Gelatin hydrogels of the microsphere type are also degraded in vivo with time, irrespective of their size, although the degradation rate is high compared with that of block gelatin hydrogels with a similar water content, because of the high surface area [67].

5. bFGF release from gelatin hydrogels

5.1. In vitro release

The preparation of hydrogels in the presence of proteins will lead to a loss in their activity through chemical crosslinking of gelatin. In contrast, the present method, in which an aqueous solution of bFGF is dropped onto freeze-dried gelatin hydrogels, followed by leaving them under various conditions to allow bFGF to sorb into the gelatin hydrogels, at

294

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

least will prevent protein from being chemically deactivated. This method is also effective in quantitatively incorporating bFGF into gelatin hydrogels with high reproducibility, irrespective of the water content, because the volume of the bFGF solution is much less than that theoretically required to impregnate bFGF into the hydrogel. The following uorescent microscopic study has demonstrated that uorescent-labeled bFGF molecules impregnated by this procedure are homogeneously distributed throughout the interior of the hydrogel [93]. bFGF was incorporated overnight into a gelatin hydrogel through impregnation at 48C. The bFGFincorporating gelatin hydrogel was placed in PBS at 378C and the bFGF concentration of the supernatant was quantitated by HPLC at different time intervals, to estimate the time prole of bFGF release. When an acidic gelatin hydrogel was used to incorporate bFGF, up to about 40% of the initial loading was released within one day, but, thereafter, no substantial release was observed. On the other hand, the hydrogel prepared from basic gelatin exhibited almost complete release of the incorporated bFGF within one day (Fig. 6). This demonstrates that bFGF cannot be released from acidic gelatin hydrogels under in vitro non-degradation conditions if basic bFGF molecules are complexed with acidic gelatin. It is apparent from Fig. 4 that all bFGF

molecules are not ionically complexed with the acidic gelatin constituting the hydrogel, even after overnight incubation at 48C. In addition, prolonged bFGF impregnation reduces the initial burst of bFGF in the in vitro release test. It is probable that the non-complexed bFGF is released from the hydrogel during the initial period of release, followed by no release of complexed bFGF, whereas no formation between basic gelatin and basic bFGF leads to prompt release of all of the bFGF molecules from the basic gelatin hydrogel, in contrast to the situation with an acidic gelatin hydrogel.

5.2. In vivo release

To assess the in vivo prole of bFGF release from hydrogels, acid gelatin hydrogels containing 125 Ilabeled bFGF were implanted subcutaneously into the backs of mice and the residual radioactivity was measured at different time intervals. The 125 I-labeled bFGF-incorporating gelatin hydrogels showed decreased residual radioactivity with implantation time, while no radioactivity was observed in the blood, suggesting that bFGF is released in vivo from the bFGF-incorporating gelatin hydrogel. The decrement pattern of radioactivity depends on the hydrogels degradability, in such a manner that the radioactivity is retained for longer when the water content of the hydrogel is lower. Also, the in vivo degradation prole of hydrogels can be radiotraced through subcutaneous implantation of bFGF-incorporating hydrogels prepared from 125 I-labeled acidic gelatin. The radioactivity of hydrogels decreases with implantation time, while the rate of decrease increased with an increase in the water content of the hydrogel. The decrement pattern of bFGF radioactivity in the hydrogel is in good agreement with that of gelatin radioactivity, irrespective of the hydrogels water content (Fig. 7). In addition, the half-life period for retention of bFGF in gelatin hydrogels of different water contents was found to be linearly related to the amount of hydrogel remaining. These ndings indicate that bFGF is probably released from the gelatin hydrogel together with degraded gelatin fragments in the body as a result of hydrogel degradation. As is apparent in Fig. 7, the amount of gelatin remaining during the initial degradation period is larger than the amount of bFGF remaining, irrespective of the water

Fig. 6. In vitro release proles of bFGF at 378C from bFGFincorporating hydrogels prepared from gelatin with IEP values of 5.0 (s) and 9.0 (d) in 1 / 15 M phosphate-buffered solution (pH 7.4). The water content of the gelatin hydrogels was 95.2 wt%, irrespective of the type of gelatin used [65].

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

295

Fig. 7. Relationship between the radioactivity of bFGF and the gelatin remaining after subcutaneous implantation of 125 I-labeled bFGF-incorporating hydrogel prepared from 125 I-labeled acidic gelatin. The water contents of the hydrogels were (s) 98.8 and (d) 96.9 wt%.

proteins is whether or not the protein released in the body actually retains its biological activity. To evaluate protein activity, in vitro culture techniques are normally employed because of their simplicity and convenience, compared with in vivo animal experiments. However, any in vitro non-degradation system cannot be applied to evaluate the biological activity of released bFGF, since protein release is involved with the in vivo degradation of hydrogel matrices in our release system. Thus, to obtain information on the retention of bFGF activity, we directly assessed vascularization and bone formation after implantation of bFGF-incorporating gelatin hydrogels in animals.

6.1. Neovasuclarization [65]

bFGF-incorporating gelatin hydrogels were subcutaneously implanted into the backs of mice and their effect on neovascularization was evaluated and compared with that of bFGF in solution. Histological examination demonstrated that vascularization was remarkable around the implantation site of bFGFincorporating gelatin hydrogels, in contrast to sites injected with an aqueous solution of bFGF. Injection of bFGF in the form of a solution was not effective in inducing vascularization at all and a bFGF-free gelatin hydrogel alone did not induce any vascularization effect. The amount of tissue hemoglobin, which is a measure of bFGF-induced neovascularization, notably increased within one day of implantation of bFGF-incorporating gelatin hydrogels with a water content of 95.2 wt% and the increased level was retained for one week, followed by a return to the initial level of hemoglobin at day fourteen. On the other hand, injection of an aqueous solution containing the same dose of bFGF, as a bFGF hydrogel, did not increase the amount of hemoglobin at the injection site over the time range studied; the level of tissue hemoglobin remained at approximately the same level as that found on injection of bFGF-free PBS or in untreated mice (Fig. 8). No increase in the amount of hemoglobin was observed even when the amount of bFGF in solution that was injected was increased to 1 mg per mouse. This must be due to a rapid elimination of bFGF from the injection site. In contrast, incorporation of bFGF into gelatin hydrogels enabled us to reduce the dose that

content of the hydrogels. This can be explained in terms of the initial release of free bFGF. As demonstrated in Fig. 6, a certain amount of bFGF is released from the acidic gelatin hydrogel, probably because the impregnation conditions are not sufcient to completely form a polyion complex between bFGF and gelatin. Therefore, it is possible that bFGF molecules that are not complexed with gelatin are released, even in vivo, from gelatin hydrogels during the initial period after implantation. However, we cannot completely rule out the possibility that bFGF is released from hydrogels through in vivo dissociation of bFGFgelatin complexes, as is illustrated in Fig. 1.

6. Biological activity The bFGF used here was originally characterized in vitro as a growth factor for broblasts and capillary endothelial cells and in vivo as a potent mitogen and chemoattractant for a wide range of cells. In addition, bFGF is reported to have a variety of biological activities [90,91,94] and to be effective in enhancing wound healing through induction of neovascularization [95,96] and regeneration of bone [9799], cartilage [100,101] and nerve [102,103], when administrated in the form of a solution. The most important concern regarding the delivery of

296

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301

Fig. 8. Time-course of neovascularization induced by free bFGF and by a bFGF-incorporating acidic gelatin hydrogel. (A) Mice received a subcutaneous injection of a PBS solution containing 100 mg of bFGF (s) and bFGF-free PBS solution (d). (B) Mice received a subcutaneous implantation of hydrogel containing 100 mg of bFGF (s) and bFGF-free hydrogel (d). **indicates signicance at P , 0.01 against the value for each dosage at day zero [65].

was effective in inducing signicant vascularization to 30 mg per mouse. This enhanced vascularization effect was observed also on injection of gelatin hydrogel microspheres that had bFGF incorporated into them [67]. The in vivo degradation prole of bFGF-incorporating gelatin hydrogels can be modied by changing the water content. For example, a gelatin hydrogel with a water content of 95.2 wt% was degraded in the mouse subcutis with time and completely resorbed after fourteen days of implantation. At that time, neovascularization could no longer be detected and the appearance of the tissue returned to that found in untreated mice. This indicates that the retention period of the hydrogel-induced vascularization effect is in good agreement with the degree of degradation of hydrogel. In addition, enhanced neovascularization was observed for all types of implanted hydrogels, irrespective of the water content, but the time prole of vascularization depended on the water content of the hydrogels. The hydrogelinduced vascularization effect was prolonged when the water content was decreased. This phenomenon can be explained in terms of the sustained release of bFGF. As described earlier, bFGF seems to be released from the gelatin hydrogel as a result of hydrogel degradation. It follows that the period of bFGF release can be regulated by changing the rate at which the hydrogel degrades. Hydrogels with lower water contents will be degraded and release bFGF in vivo more slowly than those with higher

water contents, leading to a prolonged neovascularization effect. Thus, these ndings demonstrate that the bFGF released from the gelatin hydrogel system may maintain biological activity, although the percentage of the activity that remained could not be quantitatively evaluated from this in vivo study.

6.2. Bone formation [66]

A rabbit model of a skull bone defect was employed to evaluate the in vivo efcacy of bFGFincorporating gelatin hydrogels in bone formation. When implanted into a skull defect, the bFGF-incorporating gelatin hydrogel accelerated bone regeneration at the skull defect and almost closed the defect after twelve weeks of implantation. In contrast, insignicant bone regeneration and remarkable ingrowth of soft connective tissue were observed at the bone defect when rabbits were treated with a bFGFfree gelatin hydrogel and free bFGF or were left without treatment. The bFGF-free gelatin hydrogel neither induced bone formation nor interfered with bone regeneration at the skull defect. Table 1 summarizes the results of the measurement of bone mineral density (BMD) at the skull defect of rabbits eight and twelve weeks after different treatments. The BMD of intact rabbit skulls was 120610 mg / cm 2 . Clearly, both the bFGF-incorporating gelatin hydrogels with water contents of 85 and 98 wt% enhanced the BMD of the skull defect, but the BMD was signicantly higher for rabbits that were treated

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301 Table 1 Bone mineral densities at the skull defect of rabbits eight and twelve weeks after different treatment a regimens Treatment group Water content (wt%) 85 98 NA 85 98 NA BMD (mg / cm 2 ) 8 weeks 102.066.13 100.8611.3 f 94.0611.4 82.769.11 85.8616.1 82.268.47
b

297

12 weeks 115.966.97 c ,d ,e 101.5610.6 b ,g 82.6614.2 76.669.77 80.567.75 74.169.62

bFGF-containing gelatin hydrogels (100 mg bFGF / rabbit) Free bFGF (100 mg bFGF / rabbit) Empty gelatin hydrogels PBS
a

Signicance levels were calculated for the following comparisons: b P , 0.01; c P , 0.001 versus PBS-treated group; d P , 0.001 versus free bFGF-treated group; e P , 0.01 versus 98 wt% water content bFGF-containing hydrogel-treated group; f P , 0.05 versus PBS-treated group and g P , 0.05 versus free bFGF-treated group. Abbreviation: NA 5 not applicable.

with 85 wt.% hydrogel compared with the 98 wt% hydrogel. On the other hand, free bFGF did not give rise to any signicant bone regeneration, although the BMD tended to be somewhat higher than that of PBS-treated, control rabbits. The BMD at the defect of rabbits treated with bFGF-free, empty gelatin hydrogels was similar to that of control rabbits, indicating that implantation of hydrogels in the defect did not disturb bone regeneration at the site. A similar trend was observed for bone formation after eight weeks of implantation, but the efcacy of bFGF-incorporating hydrogels in enhancing bone regeneration was not as clear as that of hydrogels implanted for twelve weeks. No signicant differences in the BMD were observed between these experimental groups after four weeks of treatment. The time course of the number of osteoblasts residing near the edge of the bone of the skull defect was examined after different treatments. Interestingly, during the initial two weeks, the time course of cell number was similar among the differently treated rabbits. No differences were observed between the rabbit groups treated with free bFGF and bFGF-incorporating gelatin hydrogels in the distribution prole of the cells that were positively stained with alkaline phosphatase during the initial period. The prole of the osteoblast number thereafter depended on the treatment type. The cell number increased for up to two weeks and then decreased for both non-treated rabbit groups and for those treated with bFGF-free, empty gelatin hydrogels, irrespective of their water content. Free bFGF treatment

prolonged the retention period of cells by a few weeks, but the number returned to the basal level in week eight. In contrast, the group of rabbits treated with bFGF-incorporating gelatin hydrogels did not exhibit such a rapid decrease in the cell number as the group that received treatment with free bFGF, and retained a signicantly higher number of cells over the time range studied, irrespective of the type of hydrogel used. However, the retention period of the enhanced cell level tended to increase with a decrease in the water content of the hydrogel. It seems that bFGF was released in the biologically active state from the gelatin hydrogel, activating osteoblasts to induce bone regeneration at the skull defect. This nding again suggests that the gelatin hydrogel system enables bFGF to be released with biological activity remaining.

7. Conclusions The need for sustained release of proteins will increasingly become larger in concert with their production on an industrial scale. However, little has been reported on the technology that can achieve the sustained release of proteins with their biological activity maintained. The main reason for this may be that recombinant bioactive proteins are at present expensive and still difcult to obtain, even if they have been commercialized. Our technology for releasing proteins is based on polyion complexation, which is commonly observed in the body between

298

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301 [2] Y.J. Wang, A. Hanson, Parenteral formulations of proteins and peptides: stability and stabilizers, J. Parenter. Sci. Technol. Suppl. 42 (1988) S3S26. [3] L.M. Sanders, G.I. McRae, K.M. Vitale, B.A. Kell, Controlled delivery of a LHRH analogue from biodegradable injectable microspheres, J. Control. Release 2 (1985) 187 195. [4] M. Asano, M. Yoshida, I. Kaetsu, K. Imai, T. Mashimo, H. Yuasa, H. Yamanaka, K. Suzuki, I. Yamazaki, Biodegradability of hot-pressed poly(lactic acid) formulation with controlled release LHRH agonist and its pharmacological inuence on rat prostate, Makromol. Chem., Rapid Commun. 6 (1985) 509513. [5] F.G. Hutchinson, B.J.A. Furr, Biodegradable polymers for the sustained release of peptides, Biochem. Soc. Trans. 13 (1985) 520523. [6] A.K. Kwong, S. Chou, A.M. Sun, M.V. Sefton, M.F.A. Goosen, In vitro and in vivo release of insulin from poly(lactic acid) microbeads and pellets, J. Control. Release 4 (1986) 4762. [7] H.V. Maulding, Prolonged delivery of peptides by microcapsules, J. Control. Release 6 (1987) 167176. [8] S.-Y. Lin, L.-T. Ho, H.L. Chiou, Insulin controlled release microcapsules to prolong the hypoglycemic effect in diabetic rats, Artif. Cells Artif. Org. 16 (1988) 815828. [9] M.S. Hora, R.K. Rana, J.H. Nunberg, T.R. Tice, R.M. Gilley, M.E. Hudson, Release of human serum albumin from poly(lactideco-glycolide) microspheres, Pharm. Res. 7 (1990) 11901194. [10] P.J. Watts, M.C. Dais, C.D. Melia, Microencapsulation using emulsion / solvent evaporation: An overview of techniques and applications, CRC Crit. Rev. Ther. Drug Carrier Syst. 7 (1990) 235259. [11] M.S. Hora, R.K. Rana, J.H. Nunberg, T.R. Tice, R.M. Gilley, M.E. Hudson, Controlled release of interleukin-2 from biodegradable microspheres, Bio / Technology 8 (1990) 755 758. [12] N. Marcotte, A. Polk, M.F.A. Gosen, Kinetics of protein diffusion from poly( D,L-lactide) reservoir systems, J. Pharm. Sci. 79 (1990) 407410. [13] R.E. Johnson, L.A. Lanaski, V. Gupta, M.J. Grifn, H.T. Gaud, T.E. Neehdam, H. Zia, Stability of atriopeptin III in poly( D,L-lactideco-glycolide) microspheres, J. Control. Release 17 (1991) 6168. [14] S. Cohen, T. Yoshioka, M. Lucarelli, L.H. Hwang, R. Langer, Controlled delivery of systems for proteins based on poly(lactic / glycolic acid) microspheres, Pharm. Res. 8 (1991) 713720. [15] M. Singh, A. Singh, G.P. Talwar, Controlled delivery of diphtheria toxoid using biodegradable poly( D,L-lactide) macrocapsules, Pharm. Res. 8 (1991) 958961. [16] D. Bodmer, T. Kissel, E. Traechslin, Factors inuencing the release of peptides and proteins from biodegradable parenteral depot systems, J. Control. Release 21 (1992) 129138. [17] P.J. Camarata, R. Suryanarayanan, D.A. Turner, R.G. Parke, T.J. Ebner, Sustained release of nerve growth factor from biodegradable polymer microspheres, Neurosurgery 30 (1992) 313319.

different biological substances, such as between growth factors and the extracellular matrix. Therefore, it will be a promising strategy for the sustained release of bioactive proteins so that we might learn more about the way in which bioactive substances are stored in vivo. The gelatin hydrogel system described above is one example of this approach, which prevents proteins from denaturing. The interaction with biological macromolecules existing in the extracellular matrix enables growth factor to regulate its biological functions [104]. Thus, the gelatin hydrogels that form polyion complexes with proteins will facilitate the release of biologically active proteins for a certain period of time. The release of protein is governed by matrix degradation and, hence, the period of protein release can be regulated by changing the rate of hydrogel degradation. Although these conclusions have resulted from our experimental data on bFGF-incorporating gelatin hydrogels, this release technology seems to be applicable for any charged biomacromolecules, such as other proteins and oligo- or polynucleotides. It is theoretically possible for gelatin to form polyion complexes with any type of charged macromolecules, although the strength of the interaction depends on the type of proteins used. The regeneration and substitution of human tissues and organs using growth factors and cells are exciting and important scientic developments. This area is called tissue engineering, which requires sustained release of growth factors as one of the essential technologies. We have demonstrated that the hydrogel system has high in vivo efcacy in releasing growth factors, such as bFGF, and enhancing their biological performance. However, a surgical operation is needed to put the block hydrogel matrix into the body. If this hydrogel system was injectable, it would be very convenient for clinical applications. Indeed, the gelatin hydrogel formulated into an injectable shape could be administered more readily into the body and exhibit similar results in respect of growth factor release and in vivo biological activity [105].

References
[1] M.C. Manning, K. Patel, T. Borchardt, Stability of protein pharmaceuticals, Pharm. Res. 6 (1989) 903918.

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301 [18] T.G. Park, S. Cohen, R. Langer, Controlled protein release from polyethyleneimine-coated poly( L-lactic acid) / pluronic blend matrices, Pharm. Res. 9 (1992) 3739. [19] W.R. Gombotz, S.C. Pankey, L.S. Bouchard, J. Ranchalis, P. Puolakkainen, Controlled release of TGF-b1 from a biodegradable matrix for bone regeneration, J. Biomater. Sci., Polym. Ed. 5 (1993) 4663. [20] Y. Tabata, Y. Takebayashi, T. Ueda, Y. Ikada, A formulation method using D,L-lactic acid oligomer for protein release with reduced initial burst, J. Control. Release 23 (1993) 5564. [21] M. Alonso, S. Cohen, T.G. Park, R.K. Gupta, G.R. Siber, R. Langer, Determinants of release of tetanus vaccine from polyester microspheres, Pharm. Res. 10 (1993) 945953. [22] A. Sopersaxo, J.H. Kou, P. Teitelbaum, R. Maskiewicz, Preformed porous microspheres for controlled and pulsed release of macromolecules, J. Control. Release 23 (1993) 157164. [23] X. Zhang, U.P. Wyss, D. Pichora, B. Amsden, M.F.A. Goosen, Controlled release of albumin from biodegradable poly( DL-lactide) cylinders, J. Control. Release 25 (1993) 6169. [24] H. Okada, Y. Doken, Y. Ogawa, H. Toguchi, Preparation of three-month depot injectable microspheres of leuprolin acetate using biodegradable polymers, Pharm. Res. 11 (1994) 11431147. [25] T. Niwa, H. Takeuchi, T. Hino, N. Kunou, Y. Kawashima, In vitro drug release behavior of D,L-lactide / glycolide copolymer (PLGA) nanospheres with nafarelin acetate prepared by a novel spontaneous emulsication solvent diffusion method, J. Pharm. Sci. 83 (1994) 727732. [26] H. Sah, R. Toddywala, Y.W. Chien, The inuence of biodegradable microcapsule formulations on the controlled release of a protein, J. Control. Release 30 (1994) 201211. [27] C. Yan, J.H. Resau, J. Hewtson, M. West, W.L. Rill, M. Kende, Characterization and morphological analysis of protein-loaded poly(lactideco-glycolide) microparticles prepared by water-in-oil emulsion technique, J. Control. Release 32 (1994) 231241. [28] R. Mehta, R. Jeyanthi, S. Calis, B.C. Thanoo, K.W. Burton, P.P. DeLuca, Biodegradable microspheres as depot system for parenteral delivery of peptide drugs, J. Control. Release 29 (1994) 375384. [29] W. Lu, T.G. Park, Protein release from poly(lacticco-glycolic acid) microspheres: Protein stability problems, J. Pharm. Sci. Technol. 49 (1995) 1319. [30] W. Lambert, K.D. Peck, Development of an in situ forming biodegradable poly-lactideco-glycolide system for the controlled release of proteins, J. Control. Release 33 (1995) 189195. [31] R. Nakaoka, Y. Tabata, Y. Ikada, Adjuvant effect of biodegradable poly( DL-lactic acid) granules capable of antigen release following intraperitoneal injection, Vaccine 14 (1996) 16711676. [32] T.G. Park, S. Cohen, R. Langer, Poly( L-lactic acid) / pluronic blends: Characterization of phase separation behavior, degradation, and morphology and use as protein-releasing matrices, Macromolecules 25 (1992) 116122.

299

[33] C.G. Pitt, Y. Cha, S.S. Shah, K.J. Zhu, Blend of PVA and PGLA: control of permeability and degradability of hydrogels by blending, J. Control. Release 19 (1992) 189200. [34] S. Miyamoto, K. Takaoka, T. Okada, H. Yoshikawa, J. Hashimoto, S. Suzuki, K. Ono, Polylactic acidpolyethylene glycol block copolymer. A new biodegradable synthetic carrier for bone morphogenetic protein, Clin. Orthop. Relat. Res. 294 (1993) 333343. [35] L. Yiuxin, C. Volland, T. Kissel, In vitro degradation and bovine serum albumin release of the ABA triblock copolymer consisting of poly( L( 1 ) lactic acid), or poly( L( 1 ) lactic acidco-glycolic acid) A-blocks attached to central polyoxyethylene B-blocks, J. Control. Release 32 (1994) 121128. [36] J.C. Gautier, J.L. Grainger, A. Barbier, P. Dupont, D. Dussossoy, G. Pastor, P. Couvreur, Biodegradable nanoparticles for subcutaneous administration of growth hormone releasing factor (hGRF), J. Control. Release 20 (1992) 67 78. [37] C. Tasset, N. Barette, S. Thysman, J.M. Ketelslegers, D. Lemoine, V. Preat, Polyisobutylcyanoacrylate nanoparticles as sustained release system for calcitonin, J. Control. Release 33 (1995) 2330. [38] E. Ron, T. Turek, E. Mathiowitz, M. Chasin, M. Hageman, R. Langer, Erosion kinetics of hydrolytically degradable polymer, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 4176 4180. [39] Y. Tabata, S. Gutta, R. Langer, Controlled delivery systems for proteins using polyanhydride microspheres, Pharm. Res. 10 (1993) 487496. [40] A. Gopferich, R. Gref, Y. Minamitake, L. Shieh, M.J. Alonso, Y. Tabata, R. Langer, Drug delivery from bioerodible polymers, in: J.L. Cleland, R. Langer (Eds.), Formulation and Delivery of Proteins and Peptides, ACS Symposium Series 567, Washington DC, 1994, pp. 242277. [41] J. Heller, A.C. Cahng, G. Rodd, G.M. Grodsky, Release of insulin from a pH-sensitive poly(orthoester), Proc. Int. Symp. Control. Release Bioact. Mater. 15 (1989) 155156. [42] J. Heller, K.V. Roskos, S.Y. Ng, P. Wuthrich, R. Duncan, L.W. Seymour, The use of poly(orthoester) in the treatment of cancer and in the pulsed release of proteins, Proc. Int. Symp. Control. Release Bioact. Mater. 19 (1992) 128129. [43] A.K. Andrianov, L.G. Payne, Polyphosphazene hydrogel microspheres for protein delivery, in: S. Cohen, H. Bernstein (Eds.), Microparticulate systems for the delivery of proteins and vaccines. Drugs and the Pharmaceutical Sciences, Marcel Dekker, New York, (1996) pp. 127148. [44] N.A. Peppas, J.E. Scott, Controlled release from poly(vinyl alcohol) gels prepared by freezingthawing processes, J. Control. Release 18 (1992) 95100. [45] J. Heller, R.F. Helwing, R.W. Baker, M.E. Tuttle, Controlled release of water-soluble macromolecules from bioerodible hydrogels, Biomaterials 4 (1983) 2233. [46] N. Celebi, N. Erden, B. Gonul, M. Koz, Effects of epidermal growth factor dosage forms on dermal wound strength in mice, J. Pharm. Pharmacol. 46 (1994) 386387. [47] J. Murray, L. Brwon, R. Langer, M. Klagsburn, A micro

300

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301 sustained release system for epidermal growth factor, In Vitro 19 (1993) 743748. E.B. Christine, W.M. Saltzman, Controlled growth factor delivery induces differential neurite outgrowth in three-dimensional cell cultures, J. Control. Release 24 (1993) 1523. S.L. Beck, L. Deguzman, W.P. Lee, Y. Xu, L.A. McFatridge, E.P. Amento, TGF-b1 accelerates wound healing: reversal of steroid-impaired healing ion tars and rabbits, Growth Factors 5 (1991) 295304. S.L. Beck, L. Deguzman, W.P. Lee, Y. Xu, L.A. McFatridge, N.A. Gillet, E.P. Amento, TGF-b1 induces bone closure in skull defect, J. Bone Miner. Res. 6 (1991) 12571265. B. Matuszewska, M. Keogan, D.M. Fisher, K.A. Soper, C.A. Hoe, J.V. Bondi, Acidic broblast growth factor: Evaluation of topical formulations in a diabetic mouse wound healing model, Pharm. Res. 11 (1994) 6571. L. Illum, N.F. Farraj, A.N. Fisher, I. Gill, M. Miglietta, L.M. Bendetti, Hyaluronic acid microspheres as a nasal delivery system of insulin, J. Control. Release 29 (1994) 133141. E. Ghezzo, L.M. Bendetti, M. Rochira, F. Biviano, L. Callegaro, Hyaluronic acid derivative microspheres as NGF delivery devices: preparation methods and in vitro release characterization, Int. J. Pharm. 87 (1992) 2129. E.R. Edelman, E. Mathiowitz, R. Langer, M. Klagsburn, Controlled and modulated release of basic broblast growth factor, Biomaterials 12 (1991) 619625. E.C. Downs, N.E. Robertson, T.L. Riss, M.L. Plunkett, Calcium alginate beads as a slow-release system for delivering angiogenic molecules in vivo and in vitro, J. Cell Physiol. 152 (1992) 422429. D. Maysinger, I. Jalsenjak, A.C. Cuello, Microencapsulated nerve growth factor: effects on the forebrain neurons following devascularizing cortical lesions, Neurosci. Lett. 140 (1992) 7174. E.R. Edelman, M.A. Nugent, M.J. Karnovsky, Perivascular and intravenous administration of basic broblast growth factor: vascular and solid organ deposition, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 15131517. R.J. Mumper, A.S. Hoffman, P.A. Puolakkainen, L.S. Bouchard, W.R. Gombota, Calciumalginate beads for the oral delivery of transforming growth factor-b1 (TGF-b1): stabilization of TGF-b1 by the addition of polyacrylic acid within acid-treated beads, J. Control. Release 30 (1994) 241245. G.L. Brown, L.J. Curtsinger, M. White, R.O. Mitchell, J. Pietsch, R. Nordquist, A. vonFraunhofeer, G.S. Schultz, Acceleration of tensile strength incisions treated with EGF and TGF-b, Ann. Surg. 208 (1988) 788794. D.L. Gilbert, S.W. Kim, Macromolecular release from collagen monolithic device, J. Biomed. Mater. Res. 24 (1990) 12211239. S. Yamamoto, T. Yoshimone, T. Fujita, R. Kuroda, T. Irie, K. Fujioka, T. Hatakeyama, Protective effect of NGF atelocollagen mini-pellet on the hippocampal delayed neuronal death in gerbils, Neurosci. Lett. 141 (1992) 161165. Y. Tabata, Y. Ikada, Synthesis of gelatin microspheres containing interferon, Pharm. Res. 6 (1989) 422427. B.G. Shinde, S. Erhan, Flexibilized gelatin lm-based articial skin model: II. Release kinetics of incorporated bioactive molecules, Bio-Med. Mater. Eng. 2 (1992) 127131. P.T. Goulmbek, R. Azhari, E.M. Jaffee, H.I. Levitsky, A. Lazenby, K. Leong, D. Paradoll, Controlled release, biodegradable cytokine depots: a new approach in cancer vaccine design, Cancer Res. 53 (1993) 58415844. Y. Tabata, S. Hijikata, Y. Ikada, Enhanced vascularization and tissue granulation by basic broblast growth factor impregnated in gelatin hydrogels, J. Control. Release 31 (1994) 189199. K. Yamada, Y. Tabata, K. Yamamoto, S. Miyamoto, I. Nagata, H. Kikuchi, Y. Ikada, Potential efcacy of basic broblast growth factor in biodegradable hydrogels for skull bone regeneration, J. Neurosurg. 86 (1997) 871875. Y. Tabata, Y. Ikada, Potentiated in vivo biological activity of basic broblast growth factor by incorporation into polymer hydrogel microspheres. Proc. 4th Jpn. Int. SAMPE Symp. 4 (1995) 2528. P. Artursson, P. Edman, T. Laakso, I. Sjoholm, Characterization of polyacryl starch microparticles as carriers for protein drugs, J. Pharm. Sci. 73 (1984) 15071513. L. Degling, P. Stjarnkvist, I. Sjoholm, Interferon-alpha in starch microparticles: Nitric oxide-generating activity in vitro and antileishmanial effect in mice, Pharm. Res. 6 (1993) 783790. P. Edman, B. Ekman, I. Sjoholm, Immobilization of proteins in microspheres of biodegradable polyacryldextran, J. Pharm. Sci. 69 (1980) 838842. H.-S. Ho, C.-C. Hsiao, T.D. Sokoloski, C.-Y. Chen, M.-T. Sheu, Fibrin-based drug delivery system. III: The evaluation of the release of macromolecules from microbeads, J. Control. Release 34 (1995) 6570. K. Park, W.S. Shalaby, H. Park, Biodegradable Hydrogels for Drug Delivery, Technomic Publishing, Lancaster, 1993. M. Hara, Polyelectrolytes, Marcel Dekker, New York, 1993. P. Dubin, J. Beck, R.M. Davies, D.H. Schulz, C. Thies, Macromolecular Complexes in Chemistry and Biology, Springer-Verlag, Berlin, 1994. A.V. Kabanov, V.A. Kabanov, DNA complexes with polycations for the delivery of genetic material into cells, Bioconj. Chem. 6 (1995) 720. F.D. Ledley, Nonviral gene therapy: The promise of genes as pharmaceutical products, Hum. Gene Ther. 6 (1995) 1129 1144. E. Tomlinson, A.P. Rolland, Controllable gene therapy. Pharmaceutics of non-viral gene delivery systems, J. Control. Release 39 (1996) 357372. W.E. Long, H. Iwato, T. Lindheimer, and E.P. Goldberg, Preparation and drug release properties of albumin-polyasparatic acid-adriamycin microspheres. Polym. Prep. 24 (1983) 5659. Y. Kawashima, T. Handa, A. Kasai, H. Takenaka, S.Y. Lin, Y. Ando, Novel method for the preparation of controlledrelease theophylline granules coated with a polyelectrolyte complex of sodium polyphosphatechitosan, J. Pharm. Sci. 74 (1985) 264268. A. Sawaya, J.P. Benoit, S. Benita, Binding mechanism of

[48]

[64]

[49]

[65]

[50]

[66]

[51]

[67]

[52]

[68]

[53]

[69]

[54]

[70]

[55]

[71]

[56]

[72] [73] [74]

[57]

[58]

[75]

[76]

[59]

[77]

[60]

[78]

[61]

[79]

[62] [63]

[80]

Y. Tabata, Y. Ikada / Advanced Drug Delivery Reviews 31 (1998) 287 301 doxorubicin in ion-exchange albumin microcapsules, J. Pharm. Sci. 76 (1987) 475480. H.F.M. Cremers, J. Feijen, G. Kwon, Y.H. Bae, S.W. Kim, H.P.J.M. Noteborn, J.G. McVie, Albuminheparin microspheres as carriers for cytostatic agents, J. Control. Release 11 (1990) 167179. A. Veis, The Macromolecular Chemistry of Gelatin, Academic Press, New York, 1964. G.S. Kwon, Y.H. Bae, H. Cremers, J. Feijen, S.W. Kim, Release of proteins via ion exchange from albuminheparin microspheres, J. Control. Release 22 (1992) 8394. K. Nakamae, T. Nizuka, T. Miyata, M. Furukawa, T. Nishino, Kato, T. Inoue, A.S. Hoffman, Y. Kanzaki, Lysozyme loading and release from hydrogels carrying pendant phosphate groups, J. Biomater. Sci., Polym. Ed. (1997), in press. T. Fujiwara, K. Sakagami, J. Matsuoka, S. Shiozaki, K. Fujioka, Y. Takada, S. Uchida, T. Onoda, K. Otira, Augmentation of antitumor effect on syngeneic murine solid tumors by IL-2 mini-pellet, Biotherapy 3 (1991) 203209. K. Fujioka, Y. Takada, S. Sato, T. Miyata, Long-acting delivery system of interferon: IFN minipellet, J. Control. Release 33 (1995) 317323. M. Munirzzaman, Y. Tabata, Y. Ikada, Complexation of basic broblast growth factor with gelatin for tissue engineering, J. Biomater. Sci., Polym. Ed. (1997) in press. K. Burczak, T. Fujisato, M. Hatada, Y. Ikada, Protein permeation through polymer membranes for hybrid-type articial pancreas, Y. Proc. Jpn. Acad. 67 (1991) 8388. M. Klagbrun, A. Baird, A dual receptor system is required for basic broblast growth factor, Cell 67 (1991) 229231. N. Gospodarowizc, L. Ferrara, L. Schweigerer, G. Neufeld, Structural characterization and biological functions of broblast growth factor, Endocr. Rev. 8 (1987) 95114. D.B. Rifkin, D. Moscatelli, Recent development in the cell biology of basic broblast growth factor, J. Cell Biol. 109 (1989) 16. Y. Tabata, A. Yagano, Y. Ikada, Biodegradation of hydrogel carrier incorporating broblast growth factor, Tissue Engineering, in press. S. Hijikata, Y. Tabata, Y. Ikada, Distribution of basic broblast growth factor throughout ionic hydrogels, Polym. Preprint Jpn. 43 (1995) 3435.

301

[81]

[82] [83]

[84]

[85]

[86]

[87]

[88]

[89] [90]

[91]

[92]

[93]

[94] N. Gospodarowizc, Biological activities of broblast growth factor, Ann. NY Acad. Sci. 638 (1991) 18. [95] P. Buntrock, K.D. Jentzsch, G. Hender, Stimulation of wound healing, using brain extract with broblast growth factor (FGF) activity. II. Histological and morphometric examination of cells and capillaries, Exp. Pathol. 21 (1982) 6267. [96] G.S. McGee, J.M. Cavidson, A. Buckley, A. Sommer, S.C. Woodward, A.M. Aquino, R. Barbour, A.A. Demetrious, Recombinant basic broblast growth factor accelerates wound healing, J. Surg. Res. 45 (1988) 145153. [97] S.B. Rodan, G. Wesolowski, K.A. Thomas, K. Yoon, G.A. Rodan, Effects of acidic and basic broblast growth factors on osteoblastic cells, Connect. Tissue Res. 20 (1989) 283 288. [98] H. Kawaguchi, T. Kurokawa, K. Hanada, Y. Hiyama, M. Tamura, E. Ogata, T. Matsumoto, Stimulation of fracture repair by recombinant human basic broblast growth factor in normal and streptozotocin-diabetic rats, Endocrinology 135 (1994) 774781. [99] T. Nakamura, K. Hanada, M. Tamura, T. Shibanushi, H. Nigi, M. Tagawa, S. Fukumoto, T. Matsumoto, Stimulation of endosteal bone formation by systemic injections of recombinant basic broblast growth factor in rats, Endocrinology 136 (1995) 12761284. [100] P. Cuevas, J. Burgos, A. Baird, Basic broblast growth factor (FGF) promotes cartilage repair in vivo, Biochem. Biophys. Res. Commun. 156 (1988) 611618. [101] S.B. Trippel, Growth factor actions on articular cartilage, J. Rheumatol. 22 (1995) 129132. [102] P. Aebischer, A.N. Salessiotis, S.R. Winn, Basic broblast growth factor released from synthetic guidance channels facilitates peripheral nerve regeneration across long nerve gaps, J. Neurosci. Res. 23 (1989) 282289. [103] A. Laquerriere, P. Peulve, O. Jin, J. Tiollier, M. Tardy, H. Vaudry, J. Hemet, M. Trdie, Effect of basic broblast growth factor and a-melanocytic stimulating hormone on nerve regeneration through a collagen channel, Microsurgery 15 (1994) 203210. [104] J. Taipale, J. Keski-Oja, Growth factors in the extracellular matrix, FASEB J. 11 (1997) 5159. [105] H.-W. Kang, Y. Tabata, Y. Ikada, Preparation of injectable gelatin hydrogels for tissue engineering, Proc. 3rd FarEastern Symp. Biomed. Mater. (1997) 67.