The hydrolytic and enzymatic degradation of newly developed hydrogels, produced by cross- linking purified poly(v-glutamic acid) (TPGA) with dihaloalkane ...
Journal of Environmental Polymer Degradation. Vol. 4, No. 4. 1996
Hydrolytic and Enzymatic Degradation of Poly(v-Glutamic Acid) Hydrogels and Their Application in Slow-Release Systems for Proteins K e s u o F a n , 1 Denis G o n z a l e s , 1 a n d M a r t i n Sevoian 1'2
The hydrolytic and enzymatic degradation of newly developed hydrogels, produced by crosslinking purified poly(v-glutamic acid) (TPGA) with dihaloalkane compounds, was studied and is reported in this paper. Analysis of hydrolysis of the hydrogel as a function of pH indicated that the hydrolysis occurred slowly at neutral pH. but fast in both acidic and alkaline solutions, while the polymer could be hydrolyzed rapidly only in acidic solutions. The ester bonds were more sensitive to hydrolysis than peptide bonds. The biodegradability of the hydrogel and polymer was further confirmed when enzymatic degradation was studied by three enzymes (cathepsin B, pronase E, and trypsin), which were able to cleave both ester and peptide bonds gradually. A slowrelease system for porcine somatotropin (pST) formed by using the hydrogel as matrix to entrap the hormone was evaluated in vitro and in vivo. Results demonstrated that the hydrogel was able to release the hormone for a period of 20-30 days and indicated its potential application in slowrelease systems for bioactive materials, especially macromolecules, such as peptides and proteins. KEY WORDS: Poly(T-glutamic acid); hydrogel; polymer hydrolysis; controlled-release system; porcine somatotropin.
been applied for slow release purposes. However, each of them occasionally had unsatisfactory characteristics, such as interactions with the materials to be released, toxicity caused by the products of polymer degradation, unstable bonds, etc. [5, 6]. Therefore, the search for new biodegradable polymers continues [2]. For applications in slow-release systems, important characteristics of the polymer are stability, hydrophobicity, morphology, molecular weight, and degree of swelling in water of the polymer backbone. Susceptibility of the polymeric backbone toward hydrolytic cleavage is probably the most fundamental parameter [1]. Polymers with various sensitivities to hydrolysis can be used for different purposes of drug release, from short-term to longterm release. Many release systems based on polymers have been developed over the past decade. However, few of them have been capable of slowly releasing drugs of high molecular weight such as proteins [5]. The development of recombinant polypeptides and proteins as commercially
INTRODUCTION During the past two decades, there has been a great interest in biodegradable polymers for the application of slow-release systems because of polymers' compatibility, permeability, biodegradability, and nontoxicity [1, 2]. By using suitable polymeric materials to construct delivery systems, the predictable and reproducible rates of release of the bioactive agents could be achieved over a set period of time, which would result in optimum biological responses, prolonged efficiency, decreased toxicity, and a reduction of the required total dose levels [3, 41. Several kinds of biodegradable polymers, including polyamides, polyesters, and polyanhydrides, have
Department of Veterinary and Animal Sciences, University of Massachusetts, Amherst, Massachusetts 01003. "To whom correspondence should be addressed.
253 1064-7.546/96/1000-0253~)9.50/0 (~) 1996 Plenum Publishing Coq~oration
254 viable therapeutic agents has made it important to develop macromolecular delivery systems [7]. Hydrogels, which are usually cross-linked hydrophilic polymer networks, are useful, flexible formats, because they do not dissolve in water but do swell. This allows an entrapped drug to be released through desorbed [4, 8]. By changing the degree of cross-linking of the hydrogel, it is possible to entrap drugs of various molecular weights for delivery. Therefore, hydrogels seem to be especially promising in the development of delivery systems for nonconventional fragile drugs, such as peptides and proteins [91. Poly(3,-glutamic acid) (~,PGA) is a bacterially synthesized polymer which has been reported to be nontoxic, nonpathogenic, and nonimmunogenic [10-12]. Modification of the polymer by esterification reaction has been carried out by several investigators [13-15]. Its use in drug delivery systems has not been reported, although it is believed to possess significant potential. We are reporting here the preparation and hydrolysis of newly synthesized hydrogels by cross-linking this polymer with dihaloalkanes and their preliminary application for the slow release of recombinant porcine somatotropin (pST). A successful slow-release system for pST becomes economically valuable because the recombinant hormone can effectively improve the growth rate, feed efficiency, and carcass leanness in finishing pigs [16-18].
EXPERIMENTAL
Fan, Gonzales, and Sevoian Table 1. Recipe of Medium E for Synthesis of Poly(~-Glutamic Acid) by Bacillus subtilis" i
Composition
Grams per liter
t=Glutamic Acid Citric Acid Glycerol NH4CI K2HPO4 MgSO4 - 7H,O FeCI~ • 6H20 CaCI2 • 2H20 MnSO.~ I'
20.0 12.0 80.0 7.0 0.5 0,5 0.04 0.15 0.08
"Distilled water to 1 L, pH 7.4 (with NaOH). hVariable from 0.000026 to 0.42 g L -~.
the bacterial culture was then elevated to 65°C for 30 min, allowing the formation of spores. Fermentation of the obtained spores in the medium E (Table I) produced the crude polymer. After fermentation, the medium was centrifuged (10,000g) and the membrane filtered (cellulose membrane, 0.22 #m) to remove the bacteria. Low molecular weight contaminants were cleared from the medium by ultrafiltration with distilled water using hollow fibers [molecular weight cutoff (MWCO): 10,000], and the polymer solution was acidified to pH 2.2 (pK, condition). The polymer was precipitated with the addition of a 3:1 mixture of ethylic ether/isopropanol, redissolved in water, and freeze-dried. The polymer preparations were free of proteins because of their nonreacting with Lowry and Bradford reagents.
Molecular Weight Determination Molecular weights were determined by gel permeation chromatography (GPC) with an Asahipak GSM700 column (7.6-mm i.d., 500-mm length; exclusion limit, 10 million Dal) using a Water ELC chromatogram coupled with a UV detector set at 220 nm. The mobile phase (1 ml/min) was 50 mM NaCI buffered at pH 7.2 with 50 mM phosphate. The number-average molar mass (M,,) was calculated from GPC pattern by using pullulan and polyethylene glycol as standards.
Synthesis and Purification of Poly(y-Glutamic Acid) -yPGA was produced by bacterial fermentation adapted from a report in the literature [19]. Bacteria (Bacillus subtilis ATCC9945a, provided by Prof. C. B. Thome) were grown on agar plates from which the most mucoid colony was selected to inoculate trypticase soy broth and grown for 24 h at 37°C. The temperature of
Cross-Linking of Poly(~/-Glutamic Acid) to Form Hydrogels 3,PGA was dissolved at 20-80 mg/ml in anhydrous DMSO at 65°C. To this solution was added 2 equiv of sodium bicarbonate per glutamate residue as a base and dihaloalkane compounds as cross-linkers (1,2-dihaloethane, C2; 1,5-dihalopentane, C5; or 1,10-dihalodecane, Cl0). The cross-linking reaction was carried out for 48 h (chloro-compounds) or 10 h (bromo-compounds) at 65°C. The resulting hydrogels were washed carefully with a buffered solution (pH 7, 0.1 M), then, several times with distilled water. Finally, gels were freeze-dried. Fractions of cross-links per 100 glutamic residue in the hydrogel were calculated from the tH NMR spectra of the fully hydrolyzed gels in alkaline solutions (pH 10, 40°C) [20]. Various gels with different cross-linker types or fractions were synthesized through this process. Un-
Degradation of Poly(~t-Glutamic Acid) Hydrogeis less otherwise specified, the 3,PGA hydrogel synthesized by 1,5-dichloropentane with a 2% cross-linking ratio and with a swellability of 40 ml per g dried gel in water was used in the rest of the experiment, because it was an injectable gel and still was able to retain its integrity and network properties for a reasonable period.
Hydrolytic and Enzymatic Degradation of the Gels and yPGA Hydrolytic degradations were conducted at various pH levels in buffered solutions (0.1 M carbonate-bicarbonate, pH 10; 0.1 M boric acid-borax, pH 9; 0.1 M sodium phosphate, pH 7; 0.1 M citrate buffer, pH 5 and pH 4). Three enzymes (cathepsin B, pronase E, and trypsin) were used for enzymatic hydrolysis. Cathepsin B (Sigma; from bovine spleen, 12 U/rag solid) was activated immediately before use with 0.02 M L-cysteine for 5 rain at 37°C in 0.1 M PBS (pH 6.0) plus 0.02 M EDTA. Pronase E (Sigma; 5.4 U/rag solid) was used in 0.1 M Tris-HCl (pH 7.4) containing 0.01 M CaCI2. Trypsin (GIBCO; 2.5% solution) was used in a Hank's balanced salt solution (HBSS; pH 7.2). To the hydrogel or "yPGA solutions, one enzyme was added, at a final concentration of 25 #g/ml for cathepsin B, 50 ttg/ml for pronase E, and 250 #g/ml for trypsin. The activity of enzymes was monitored and maintained by adding fresh enzyme solutions when decreased. The pH value in both hydrolytic and enzymatic degradation solutions was maintained by periodic adjustments with 0.1 M HC1 or NaOH. For both hydrolytic and enzymatic hydrolysis experiments, the dried hydrogels or 3,PGA were added to the selected buffer at a concentration of 4 mg/ml. The experiments were conducted at 37°C. The hydrolysis reaction was monitored periodically by GPC analysis of the supematants.
Experiments in Vitro and in Vivo for the Release of Porcine Somatotropin Using ~,PGA Based Hydrogei as Matrix The recombinant porcine somatotropin (a 22-kD growth hormone; Bresatec Limited) was chosen for this experimentation to test the efficiency of the 3,PGA-based hydrogel as a slow delivery matrix. The hormone was dissolved in HBSS at a concentration of 1 mg/ml. The solution was filtered to remove any insoluble materials. Two 30-ml aliquots were prepared for in vitro experiments. One was used as the hormone control; to the other was added 0.42 g of the ",/PGA gel powder. The
255 gel was allowed to swell for 2 days in order to entrap the hormone. After lyophilization, a pellet was made with 500 lb pressure in a pressure machine and inserted into a dialysis bag (MWCO: 100,000) through which the diffusion of the hormone was allowed. The dialysis bags were placed in 50-ml plastic tubes containing 40 ml HBSS. Then the tubes were slowly shaken at 37°C for a 30-day period. The dialysis solution was changed every 2 days and analyzed for hormone released. The hormone level was measured by a sandwich ELISA using mouse anti-pST monoclonal antibody (Ab) as the first Ab, biotinylated rabbit Ab as the second Ab, streptavidin-alkaline phosphatase as the enzyme, and pNPP as the substrate. The pST content was measured by absorbency at 405 nm [21]. The in vivo experiments were conducted on the hypophysectomized (hypox) female rats (from Charles River Laboratories). Release of pST from the matrix injected subcutaneously was monitored by body weight gain (BWG) [22, 23], and compared with that of the controls. Rats were housed one per cage at a controlled temperature of 24-26°C, with lighting regulated on a 14 h light, 10 h dark schedule. Feed was made readily available in the cages" feed bins as well as a 5 % glucose solution for water. Rats were divided into four groups, with three rats in each group. A total dosage of 3 mg pST for each rat was administered for the 30-day period of the experiment (100 #g pST/day) [22]. One group with no treatment was for negative control. The second group was for positive control, with each rat receiving a subcutaneous injection of 100 ~g pST in 0.2 ml solution every day. The third group was injected with a total dosage of pST (3 mg/0.3 ml) on the first day of the experiment. The fourth group was for the slow-release system. For this group, a gel/hormone preparation with the same total dosage of pST was injected in a 0.3-ml volume, Body weight gains (BWG) of the rats were recorded everyday.
RESULTS AND DISCUSSION
Production of ,/PGA and Formation of "/PGA-Based Hydrogels Several fermentations were conducted. The typical yield and number-average molecular weight, M,,, of polymer were usually in the range of 5-10 g/L and 1~ × 105 to 5 × 105, respectively, depending on each fermentation. Representative yield and Mn results during production are shown in Fig. 1. It was found that the selection of bacterial colonies for inoculation was very
256
Fan, Gonzales, and Sevoian
9
160000
8
l~txx~o
7
1ate00
m
E E
v
=,
6
O
100000
c
oeo
=
5 4
@ en
3 a.
40ooo
2
E
Z
1
0
0 2
1
3
4
5
6
7
8
9
10
Fermentation Days Fig. l. Productionof poly(y-glutamicacid) by Bacillus subtilis. Fermentation was carried out at 37°C and by using the formulaof Medium E. The yield and Mn of polymer were determiend by GPC.
important for the polymer formation, because B. s u b t i l i s can easily change to a non-PGA-producing variant, as reported by other authors [15]. Modification of the polymer by esterification reaction resulted in various solubilities of 3,PGA esters, from water-soluble esters to water-insoluble esters [1315]. The water-insoluble 3,PGA esters were unable to entrap bioactive peptides to form the slow-release system, whereas water-soluble 3,PGA esters released materials rapidly. By using dihaloalkanes, hydrogels (or networks), produced by cross-linking the polymer as shown in the following reaction, have been successfully synthesized in this laboratory. The gel formation rates varied with different polymer concentrations, temperatures, and base concentrations [20]. The effects of initial polymer concentrations on the reaction rates at fixed temperature and base concentration are indicated in Table II. Both very low and high polymer concentrations resulted in slower reaction
speeds because of the decreased probability of intramolecular reactions at low concentrations and limited mobility of the reactants at high concentrations. Hydrolytic Degradation
Hydrolysis studies not only can give data on the rate of hydrolysis of the covalent bonds, but also can provide information on the physical properties of hydrogels which are important in slow-release applications. Results of hydrolysis in different pH solutions demonstrated that the hydrogel was hydrolyzed rapidly in either alkaline or acidic solutions (Fig. 2). At neutral pH, however, the hydrolysis was slow. In contrast, the polyamide itself was hydrolyzed readily only in acidic solutions (Fig. 3). From the different hydrolytic patterns, it appears that ester bonds were hydrolyzed far more quickly than peptide bonds. This phenomenon was not so obvious when the hydrolysis of the gel at pH 4
-CO-CH2-CH2-CH-NH-CO-CH2-CH2-CH-NH-COCH2-CH2-CH-NH- ~ ~
0
CH2 • (CI~2)°"C-- Crosslinker
(fr O,, CH2 \ (CHg.
%
0 --
O'Na÷ Ester bond
o ¢./,o c
-CO-CH2-CH2-CH-NH-CO-CH~-CH2-CH-NH-_CO-CH~-CH2-CH-NH-
Peptidebond Scheme 1
yl~A
backbone
Degradation of Poly(T-Glutamic Acid) Hydrogels Table !1. Formation of Hydrogel Based on 3@GA as Polymer and 1,5-Dichloropentane as Cross-Linker: Relationship Between Reaction Speed and Initial Polymer Concentration Reaction h
20 mg 3,PGA/ml
40 mg "),PGA/ml
80 mg 3,PGA/ml
0 7 18 40
100" 94.5 78.5 5.5
100 12.6 0.02 0
I00 56.2 0 0
257 160
A
140 120 g" 100 O 80 C
60
- ~ - p H 10 -,a,- pH 7 - e - pH 4
, "L
40 20
--~
t
i
"Percentage of "rPGA remaining un-cross-linked (detected by GPC).
i
=
1
50
lOO
15o
Hydrolysis Days
was considered. For the study at pH 4, hydrolysis of the ester bonds in the cross-links (the highest cross-link density was approximately 2 for every 100 glutamic repeat units or peptide bonds) was the major phenomenon in the gel degradation up until 50 experimental days, and hydrolysis of the amide function represented an average of fewer than three cleavages per 3,PGA chain. The slow hydrolysis of the gels at neutral pH is very important for application in slow-release systems. In this case, if only hydrolysis is considered (even inside ly-
60 B O
0 DI~
____
