Biomacromolecules 2004, 5, 1662-1666
1662
Influence of Weak and Covalent Bonds on Formation and Hydrolysis of Gelatin Networks Se´ bastien Giraudier,† Dominique Hellio,‡ Madeleine Djabourov,‡ and Ve´ ronique Larreta-Garde*,† Laboratoire Errmece, Universite´ de Cergy Pontoise, 95302 Pontoise Cedex France, and Laboratoire de Physique et Me´ canique des Milieux He´ te´ roge` nes, Ecole Supe´ rieure de Physique et Chimie Industrielles, 75231 Paris Cedex 5, France Received June 7, 2004; Revised Manuscript Received June 30, 2004
The relative influence of physical and chemical bonds to overall gel properties are explored in gelatin gels. Physical, chemical, chemical-physical, and physical-chemical gels are obtained by cooling the protein solution and/or by transglutaminase reaction. Each type of network is characterized by rheology and polarimetry. It is shown that the overall properties as well as the dynamics inside the gels are dependent upon the order of formation and on the relative amount of triple helices and covalent bonds. Enzyme hydrolysis of covalent gels is slower than that of physical gels, as confirmed by the kinetics of helix release and degradation. A scheme is proposed to explain the results at both the physicochemical and the molecular levels. 1. Introduction Many biopolymers display gelling properties in aqueous media. Gelation is a complex operation, and whereas theoretical models have been elaborated to describe the macroscopic sol/gel transition,1 its molecular mechanisms are not yet perfectly elucidated for the various possible cases. In vivo, protein biopolymers are dense networks composed of different proteins linked by various bonds. In both physiological and pathological events, they undergo changes in composition leading to variations in macroscopic properties. This process involves weak interactions but also enzyme-catalyzed cleavage or formation of covalent bonds.2,3 Moreover, the overall feature of a protein network depends on the coexistence of ordered structures and flexible coils inside the gel. Due to the complexity of these systems, the use of model gels made of one single protein is a useful approach to the study of the relationship between molecular characteristics and macroscopic properties of protein gels. Among the proteins able to give rise to a gel, gelatin has the ability to form thermally reversible networks.4 Below the sol-gel transition temperature, part of the protein coils gives rise to triple helices reminiscent of the native collagen5 and protein solution turns into gel. For various gelatins, a common curve expresses the relation between elasticity and helix content.7,8 Physical protein gels may be stabilized by the further addition of covalent bonds due to transglutaminase reaction. This is the case in physiological processes such as hemostasis and wound healing.9,10 This enzyme catalyzes intra- and * To whom correspondence should be addressed. Phone: 33 134 256 605. Fax: 33 134 256 552. E-mail:
[email protected]. † Universite ´ de Cergy Pontoise. ‡ Ecole Supe ´ rieure de Physique et Chimie Industrielles.
intermolecular cross-linking of some proteins, including gelatin, by N-(γ-L-glutamyl)-L-lysine side chain bridges.11 Recently, gelatin gels covalently cross-linked by transglutaminase have been studied.12 Chemical-physical and physical-chemical gels were described, within which two kinds of network could coexist. In this paper, we describe the molecular dynamics inside different types of gel, correlating the macromolecular assemblies with their viscoelastic properties and showing that proteolysis resistance depends on network organization. 2. Materials and Methods 2.1. Materials. Protocol of Gelatin Dissolution. Gelatin used in this study is a kind gift from Rousselot. It was extracted from pig skin by an acidic process, has a pI of 8.74, a bloom of 292 g, an average molecular weight of 168 500 and a polydispersity of 1.91. Gelation properties are similar to those of Sigma Type A1 gelatin (G2500). Concentration used was 5 g.100 mL-1 (5%, W/V) in 50 mM Tris HCl buffer, pH 7.4. Enzymes. Ajinomoto Transglutaminase ActiVa WM from StreptoVerticillium sp. was provided by UNIPEX, France (mTgase). It was used without further purification, except for a filtration through a 0.2 µm membrane. The specific activity was determined according to de Mace´do et al.13 Thermolysin (protease type X from Bacillus thermoproteolyticus rokko, from Sigma P-1512) is a zinc metalloprotease. Enzyme activity was measured at 27 and 40 °C on a model substrate (N-(3-[2-Furyl]Acryloyl)-Gly-Leu-Amide from Sigma F-7383). The calculated Vi40°C/Vi27°C ratio is 1.66. All enzymes were stored at -20 °C and enzyme solutions were made fresh before each reaction in 50 mM Tris-HCl buffer, pH 7.4.
10.1021/bm049670d CCC: $27.50 © 2004 American Chemical Society Published on Web 08/04/2004
Biomacromolecules, Vol. 5, No. 5, 2004 1663
Communications
2.2. Optical Rotation. Optical rotation was measured on a Perkin-Elmer 341 polarimeter. Temperature control was performed by a Julabo FS18 bath. Cooling and heating ramps of 0.5 °C min-1 were applied. The amount of helices is derived from the specific optical rotation.17 The helix amount χ is derived from χ)
[R]helix - [R]coil λ λ [R]collagen - [R]coil λ λ
(1)
where λ is the wavelength (λ ) 436 nm), [R]λ ) R/lc is the specific optical rotation of the protein in solution, c is the concentration (grams per cubic centimeter), l is the optical path (0.1 decimeter), R is the optical rotation angle (degrees) is the specific optical measured experimentally, [R]collagen λ rotation of native soluble collagen (χ ) 1), which contains is the specific optical rotation only triple helices, and [R]coil λ 3 -1 of the coils (χ ) 0). ([R]collagen dm-1 at 436nm ) -800 deg cm g 3 -1 -1 coil 27 °C, [R]436nm ) -274 deg cm g dm at pH ) 7, 4 and 40 °C8). The additivity of the signals due to both transglutaminase and gelatin was checked. The signal from the enzyme being constant, it was subtracted from the apparent angle. 2.3. Rheology. Rheology measurements were performed with an AR 1000 from TA Instruments operating in the oscillatory mode, with a strain of deformation of 1% and a frequency of 1 Hz. Deformation, storage modulus G′ and loss modulus G′′ were recorded as a function of time. The results are expressed as tan δ ) G′/G′′. Temperature was controlled by a Peltier device. A cone/plate geometry with a cone of 6 cm/2° was used. Temperature ramps of 0.5 °C min-1 were applied. 2.4. Preparation of Gels. Gelatin (10%, W/V) was swelled overnight at 4 °C in buffer. Before use, this stock solution was incubated at 40 °C for 30 min, then diluted at 40 °C to obtain a 5% solution. The protocols to obtain the various gels are compared below: Label Treatment Description. A. Physical Gel. Unmodified gelatine incubated at 40 °C for 155 min and then 0.5 °C/min temperature reduction to 27 °C. To evaluate the gel time, time is measured from the beginning of the cooling ramp. B. Chemical Gel. Gelatin, incubated with 1 unit mTgase at 40 °C for 155 min and no temperature reduction. This gel is cross-linked by enzyme- catalyzed covalent bonds. C. Chemical-Physical Gel. Gelatin, incubated with 1 unit mTgase at 40 °C for 155 min and then 0.5 °C/min temperature reduction to 27 °C. This gel is a cooled chemical gel. D. Physical-Chemical Gel. 1 unit mTgase added to gelatin at 40 °C and then instant temperature reduction to 27 °C 2.5. Thermolysin-Catalyzed Degradation of the Gels. Three different types of gel (physical, chemical, physical-chemical, 100 µL) were directly formed in 1.6 mL cuvettes. After 2h30 gelation, transglutaminase, when used, was inactivated by heat shock (10 min at 80 °C) (http://195.68.24.130/UNIPEXINS/FRA/htm/frame_cata_alimentaire.htm).
The gels were covered in the cuvette with 900 µL TrisHCl buffer, pH 7.4 at 40 °C for the “chemical gel” and at 27 °C for the others. Under these conditions, the gels do not spontaneously solubilize. After 30 min of storage at the desired temperature, 100 µL of thermolysin solution (50 µM) was added to the upper liquid phase. Regular mixing was performed. As the gel hydrolysis proceeded, solubilized products diffused from the bottom gel to the upper liquid phase. The solubilization kinetics was monitored at 280 nm with a Uvikon spectrophotometer by measuring the appearance of hydrolysis products in the liquid phase (for a scheme, see ref 14). Gelatin contains only few aromatic residues but the signal is high enough to be significant. 2.6. Helix Release and Degradation. The same hydrolysis protocol was followed and the release of triple helices observed by optical rotation. The helix amount χ in the liquid phase was estimated for both physical and physical-chemical gels. Additivity of signals was checked (only RTgase has an influence on Rmeas, contribution of thermolysin to optical rotation was neglected): Rmeas ) Rhelix + RTgase
(2)
Before total gel hydrolysis: [R] ) R(t)/l V0 t
(3)
Then with eqs 1, 2 & 3:
χ)
t Rmeas(t) - [R]TgaselV Tgase 0 - [R]coil helix lV 0 t [R]collagen - [R]coil
where V0 (protein apparition rate) was estimated through helix OD280nm. V Tgase ) 12.5 10-6 µg min-1; V0,physicalgel ) 80 µg 0 helix -1 -1. min ; V 0,physicalchemicalgel ) 62.5 µg min After total gel hydrolysis: RTgase ) 0.33° (calculated in solution) Gelatin concentration in the solution was 5 × 10-3 g/cm3 3. Results and Discussion 3.1. Physical Gels. When a gelatin solution is cooled, protein coils locally assemble into triple helices and a network is formed giving rise to a gel.4 The emergence of the physical network was followed through the viscoelastic properties of the gel (Figure 1A squares, Table 1). The triple helix amount increased to 15% in 150 min after the beginning of the cooling ramp (Figure 1B, squares). This gel is not an equilibrium state, and it continuously evolves with time. Raising back the temperature leads to a gel-sol transition (Figure 1A). The mechanical and thermal properties of the gel depend on many parameters including the amino acid composition of the biopolymer, the environment conditions, as well as the thermal history.15 When gelatin was used, no physical gel was obtained for temperatures above 34 °C at concentrations up to 10%.
1664
Biomacromolecules, Vol. 5, No. 5, 2004
Communications Scheme 1. Proposed Organization of the Various Gels
Figure 1. (A) Storage (open symbols) and loss modulus (filled symbols) and (B) % helix measured at 40, 27 then 40 °C as a function of time for physical get (squares) and chemical then chemicalphysical gel (circles). The % helix is 100χ. Table 1. Viscoelastic Properties and Helix % in Different Types of Gela gel time G′ G′′ helix content (min) (Pa) (Pa) tan δ (%) physical gel chemical gel chemical-physical gel physical-chemical gel
61 8.5 8.5 5.5
60 527 722 907
4 0.63 1.3 3.6
15 837 555 252
15 0 3 9
a To estimate gel time, t was considered as the beginning of the cooling 0 ramp. For the chemical and physical-chemical gels, t0 corresponds to transglutaminase addition. Gel time was estimated when G′ ) G′′. G′, G′′, and helix % were measured at t150min.
3.2. Chemical and Chemical-Physical Gel. The covalent bonds were formed prior to the helices with a 40-2740 °C thermal protocol. In a first step, a chemical gel was obtained with 1 unit Tgase mL-1 at 40 °C. The time dependence of the storage modulus (Figure 1A, circles) indicates a fast sol-gel transition. Under the same conditions, a 5% gelatin solution without enzyme shows no gelation (Figure 1A, squares). In both samples, no helix was formed (Figure 1B) at 40 °C as shown by optical rotation measurements. Then, the temperature was decreased to 27 °C. This particular temperature was chosen because the thermal stability of gelatin triple helices is optimum. A physical gel was obtained from the gelatin solution without enzyme and a chemical-physical gel (presence of covalent bonds and helices) was obtained from the chemical gel. The storage modulus G′ of the physical gel was much lower than that of the chemical gel (Figure 1A). In both gels, helix formation was observed (Figure 1B). When temperature
increased to 40 °C, the physical gel melt, as opposed to the chemical gel, which was irreversible. With the chemical gel, a very high tan δ value was observed (Table 1) and the obtained gel was very “resonant” (after a choc the gel vibrated for a long period of time). These particular properties are due to a network with a high amount of covalent bonds. However, for gelatin, covalent crosslinkage is limited to 32% of the theoretical possible N-(γL-glutamyl)-L-lysine bonds.16 It implies that, inside the gel, gelatin chains keep a large flexibility which supposes the binding of long protein chains in random coil conformation as illustrated on Scheme 1. When the chemical gel was cooled, no discontinuity in G′ time evolution was observed (Figure 1A open circles), whereas tan δ (Table 1) and the resonant properties decreased. The presence of a covalent network before helix formation strongly limited, but did not totally prevent triple helix propagation (3% helices after 150 min cooling compared to 15% for the physical gel (Table 1)). This indicates that macromolecular dynamics was allowed inside the covalent gel. Using a low concentration of transglutaminase, Babin and Dickinson17 have shown that a low amount of covalent bonds hinders the later formation of the physical gel; here we show that this is directly due to the reduced formation of triple helices and we quantify this effect. It was also assumed that cooling the covalent gel could locally allow the formation of ordered structures.12 We show here that these structures correspond to triple helices that are not thermally stabilized by covalent bonds as they melt at the same temperature as those constituting the physical network. These results indicate that two different networks, with different types of bonds, could coexist, each one keeping its own behavior toward temperature variation. 3.3. Physical-Chemical Gel. The enzymatic reaction was performed at 27 °C where coils undergo conformational transition and form triple helices so that a “physicalchemical” gel due to both weak interactions and covalent bonds was obtained. This physical-chemical gel showed properties and helix content intermediary between those of physical and chemical-physical gels. (Table 1). Gel time was similar to that of the chemical gel at 40 °C although enzyme activity decreased with temperature.
Communications
Biomacromolecules, Vol. 5, No. 5, 2004 1665
Figure 2. Release of degradation products in the liquid phase as a function of time. Hydrolysis was carried out by 5 µM thermolysin on physical (filled squares), physical-chemical gel (open squares), and chemical (open circles, standardized curve) gelatin gels. Arrows indicate total gel degradation.
These results showed that covalent bonding may occur simultaneously to the coil to helix transition. A comparison with the chemical-physical gel confirms the importance of thermal history on gel characteristics. This particular gel was connected by weak interactions and covalent bonds as shown on Scheme 1, both having a significant influence on network viscoelasticity. The dynamics of helix directly depends on covalent association of gelatin chains by transglutaminase. 3.4. Proteolysis of Different Gels. Enzymatic degradation of the gels by Thermolysin was performed. Peptides and helices released from the gel were observed in the solution phase through OD280nm and by polarimetry at 436 nm respectively. After completion of gel hydrolysis, the OD280nm became constant but the degradation of released helices carried on in the solution. Peptide Release. To abolish the difference due to temperature on Thermolysin activity, the theoretical Vi at 27 °C of the chemical gel hydrolysis (performed at 40 °C) was calculated using the Vi40°C/Vi27°C ratio (1.66). The enzymatic hydrolysis of the various gels gave additional information about their structure and organization. Hydrolysis of the physical gel occurred faster than that of the covalently linked gels, confirming that transglutaminase cross-linked products are more resistant to proteolytic degradation.18 Furthermore, covalent gel degradation occurred at the same rate with or without helices. This may be due to the specificity of thermolysin which preferentially recognizes hydrophobic residues with large lateral chains,19 as they are rare in helices, thermolysin should preferentially act on the random coil part of the gel chains. Helix Release. Helix release in the liquid phase due to the gel hydrolysis was followed by polarimetry (Figure 3A). The observed signal is the sum of a generation term due to helices liberated by the degrading gel and a consumption term corresponding to the degradation of helices in the supernatant. When helices are released from the physical gel, the helix amount in solution fluctuates with time ranging from 9.8% to 31.7% with an average value of 16.4%. On the contrary, the amount of helix released from the physicalchemical gel is nearly constant over time and equal to 9% ( 1%. Here again, the total hydrolysis comes later with the physical-chemical gel than with the non covalent gel
Figure 3. (A) Release of triple helices in the liquid phase as a function of time. Hydrolysis was carried out by 5 µM thermolysin on Physical (filled squares) and physical-chemical (open squares) gelatin gels. Arrows indicate total gel degradation. (B) Kinetics of the degradation of released helices after total gel solubilization.
(92 min instead of 75). This indicates a different repartition of the cleavage bonds inside the two gels. After total gel degradation, no more helices are released in the solution, the generation term is null, only the consumption term is measured. To compare the kinetics of helix degradation in solution (Figure 3B), both time and helix content were standardized using the data corresponding to the gel/sol transition as measured on Figure 3A: 100% helix corresponds to 16.4% helix content in the liquid phase for the physical gel and to 9% helix content in the liquid phase for the physical-chemical gel; the beginning of hydrolysis kinetics in solution were considered after 75 min for the physical gel and 92 min for the physical-chemical one. After gel solubilization, the kinetics of helix degradation in solution were superimposable (Figure 3B): when no more networks existed, the same proteolysis of helices was observed. This was a slow phenomenon as only few cleavage sites (Ala) were present in gelatin helices.20 This difference in hydrolysis indicates a different supramolecular organization of the two types of gel leading to different accessibilities for the protease to the cleavage sites. This is consistent with the various ultrastructures proposed on Scheme 1. 4. Conclusion These results offer new insights into the relationship between gel properties and helix content and the effect of transglutaminase-catalyzed cross-linking on the ability of gelatin chains to undergo triple helix formation. They confirm
1666
Biomacromolecules, Vol. 5, No. 5, 2004
that the history of the gel as well as its composition is of crucial importance for its macroscopic properties. The macromolecular dynamics of coexisting helical and covalent networks has a significant influence on viscoelasticity and on hydrolysis by protease. They suggest that a network with both helices and covalent bonds could be protected from the enzymatic hydrolysis which could be of biological or industrial relevance. However, these are complicated systems with non trivial behaviors where enzymatic reactions do not follow linear kinetics. These enzymatic reactions will be described elsewhere in details. Acknowledgment. Se´bastien Giraudier was supported by a grant from “Projet interre´gional: Re´gulation de la matrice extracellulaire et pathologie”. The authors thank Dr. A. Pimenta-Blight (University of Cergy Pontoise) for kindly correcting their paper and Dr. G. Cahiez (CNRS, Cergy Pontoise) for lending them his polarimeter. References and Notes (1) De Gennes, P. G. Scaling concepts in polymer physics; Cornell University Press: Ithaca, NY, 1985 (2) McCawley, L. J.; Matrisian, L. M. Mol. Med. Today 2000, 6, 149156.
Communications (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)
Aeschlimann, D.; Thomazy, V. Connect. Tissue Res. 2000, 41, 1-27. Djabourov, M. Contemp. Phys. 1988, 29-3, 273-297. Hulmes, D. J. S. Essays Biochem. 1992, 27, 49-67 Ross-Murphy, S. B. Polymer. 1992, 33, 2622-2627. Joly-Duhamel, C.; Hellio, D.; Djabourov, M. Langmuir 2002, 18, 7208-7217. Joly-Duhamel, C.; Hellio, D.; Ajdari, A.; Djabourov, M. Langmuir 2002, 18, 7158-7166. Okada, M.; Blomba¨ck, B.; Chang, M.-D.; Horowitz, B. J. Biol. Chem. 1985, 260-3, 1811-1820. Haroon, Z. A.; Hettasch, J. M.; Lai, T.-S.; Dewhirst, M. W.; Greenberg, C. S. FASEB J. 1999, 13, 1787-1795. Nio, N.; Motoki, M.; Takinami, K. Agric. Biol. Chem. 1986, 50-4, 851-855. Crescenzi, V.; Francescangeli, A.; Taglienti, A. Biomacromolecules 2002, 3, 1384-1391. de Mace´do, P.; Marrano, C.; Keilor, J. W. Anal. Biochem. 2000, 285, 16-20. Berry, H.; Pelta, J.; Lairez, D.; Larreta-Garde, V. Biochim. Biophys. Acta 2000, 1524, 110-117. Michon, C.; Cuvelier, G.; Relkin, P.; Launay, B. Int. J. Biol. Macromol. 1997, 20, 259-264. Fuchbauer, H. L.; Gerber, U.; Engelmann, J.; Seeger, T.; Sinks, C.; Hecht, T. Biomaterials 1996, 17, 1481-1488. Babin, H.; Dickinson, E. Food Hydrocolloids 2001, 15, 271-276. Griffin, M.; Casadio, R.; Bergamini, C. M. Biochem. J. 2002, 368, 377-396. Matsubara, H. Biochem. Biophys. Res. Com. 1966, 24, 427-430. Kersteen, E. A.; Raines, R. T. Biopolymers 2001, 59 (1), 24-8.
BM049670D
