8876
J. Phys. Chem. B 2000, 104, 8876-8883
Calorimetric Solution Properties of Simple Saccharides and Their Significance for the Stabilization of Biological Structure and Function Danforth P. Miller† and Juan J. de Pablo* Department of Chemical Engineering, UniVersity of Wisconsin-Madison, 1415 Engineering DriVe, Madison, Wisconsin 53706 ReceiVed: March 1, 2000; In Final Form: June 29, 2000
Thermodynamic and kinetic quantities were measured for 13 common saccharides: cellobiose, gentiobiose, glucose, lactose, lactulose, leucrose, maltose, melibiose, palatinose, raffinose, sucrose, trehalose, and turanose. Thermodynamic data included heats of solution (in water) of the amorphous and crystalline solids and melting temperatures. The glass transition temperature (Tg) and the increase in heat capacity at Tg were measured for each rigorously dried product. Many of these data were previously unavailable. We provided an interpretation of the heat of solution of the amorphous saccharides in terms of their capacity to hydrogen bond to water. A correlation was noted between the heats of solution of the amorphous carbohydrates and their Tg values. Furthermore, a thermodynamic analysis was performed to obtain values of the Gibbs free energy change for the amorphous-to-crystalline transformation at 25 °C.
Introduction In addition to their importance to the food and pharmaceutical industries, the simple saccharides have received considerable attention for their ability to protect biological molecules and structures against the stresses induced by freezing and drying processes and during subsequent storage. For proteins and phospholipid bilayers, drying processes have received particular attention, as the simple saccharides are the only known compounds that protect during extreme dehydration.1-4 Several biophysical mechanisms have been used to explain the protective action of disaccharides. One mechanism of preservation during dehydration was first suggested in 1971 by Crowe and has come to be known as the “water replacement hypothesis”.5 To account for the ability of certain organisms to survive during desiccation, Crowe proposed that certain physiological solutes replace the lost water around polar residues of biological macromolecules. Carpenter and Crowe suggested that this may be the mechanism by which certain solutes maintain native protein structure during dehydration.4 According to those authors, some of the hydroxyl groups of the saccharide form hydrogen bonds with the polar residues of the proteins. This view is further supported by the studies of Tanaka and co-workers,6 in which they freeze-dried catalase in the presence of various protective excipients. They found that the degree of protection of catalase depended on the mass ratio of excipient to protein and not on the bulk concentration of excipient. This result suggests that stabilization during dehydration is a result of direct interactions between the protein and excipient molecules. Constantino et al. provided further evidence of molecular interactions between disaccharides and proteins.7 Gravimetric moisture sorption analysis measurements showed that the calculated water monolayer of colyophilized protein:disaccharide formulations was smaller than that based on the individual * Corresponding author:
[email protected], Fax (608) 262-5434. † Current address: Inhale Therapeutic Systems, 150 Industrial Road, San Carlos, CA 94070.
contributions of the protein and sugar. This led them to conclude that the reduction in water binding sites was a result of interactions between the protein and the amorphous sugar. In addition to confirming the above results for low sucrose:protein ratios, Tzannis and Prestrelski observed destabilization of trypsinogen at mass ratios greater than 1:1 (sucrose:trypsinogen).8 They suggested that, in those formulations, a phase separation occurred prior to spray drying. In that case, sucrosesucrose interactions replaced protein-sucrose interactions, which resulted in an overall destabilization of the protein. Prestrelski et al. demonstrated that excipients that substitute for water are able to preserve the native structure and function of interleukin-2 during dehydration.9 During storage, excipients with high glass transition temperatures were shown to provide the highest level of protein stability. Those results suggest that, at least for this protein, there exists an optimum formulation that has sufficient ability to replace water during dehydration, yet also has a sufficiently high Tg to maintain stability in the dry state. Another important property of a stabilizing agent is its ability to form an amorphous, or glassy matrix.10-12 The glassy “state” severely restricts the translational and relaxational motions of the protein. Good glass formers are those molecules that undergo glass transitions at higher temperatures. Indeed, Green and Angell noted a correlation between the glass transition temperature (Tg) of several mono- and disaccharides and their ability to protect Ca2+-transporting microsomes during lyophilization.12 These results indicate that the ability to form an amorphous solid is important for long-term storage of biologicals. Although the disaccharides are structurally similar, the protective efficacy depends on the particular compound employed.13-16 Some workers have stated that, among the disaccharides, trehalose is superior17-19 or even unique20,12 in terms of its protective properties, whereas others have criticized such claims.10 The special status ascribed to trehalose is partially based on the fact that several plants and organisms that can withstand extreme desiccation synthesize this disaccharide of
10.1021/jp000807d CCC: $19.00 © 2000 American Chemical Society Published on Web 09/14/2000
Calorimetric Solution Properties of Simple Saccharides glucose.21 However, trehalose is not unique in this manner; desiccation-resistant maize embryos (corn seeds), for example, contain raffinose and sucrose.22 Much speculation has been made regarding the protective properties of trehalose and its isomers. In an effort to develop superior protective formulations, it is helpful to be able to determine the molecular structures and interactions that are responsible for stabilization against stresses incurred during freezing, drying, and subsequent storage. Comparison of the kinetic and thermodynamic properties of the disaccharides that have the same molecular weight may provide clues to the structural nuances that influence protection. For example, Crowe et al. reported that the relative effectiveness of saccharides at preserving the structural and functional activity of Ca2+transporting microsomes during lyophilization follows the order trehalose, lactose, maltose, cellobiose, sucrose, and raffinose.15 Although this order is not universal and depends somewhat on the biological system involved, trehalose appears to be consistently more effective than the other disaccharides.12 Miller and co-workers noted an additional correlation between the aforementioned order of protective efficacy and the heat of solution of the crystalline saccharide.23 Aldous and co-workers suggested that the ability of some amorphous carbohydrates to crystallize as stoichiometric hydrates may give them exceptional stabilization properties during storage.11 This could happen in a humid environment, where sorption of water reduces Tg and plasticizes the amorphous matrix.24 According to this theory, a carbohydrate that is able to form a stoichiometric hydrate could sequester sorbed water into its crystal structure, thereby prohibiting the residual moisture from depressing Tg or enabling biochemical reactions in the rest of the amorphous sample. Despite the appeal of identifying a property that correlates well with protective capacity and all that that entails (freezing, drying, storage, etc.), we realize that it is highly unlikely that a single property or feature will be able to explain all observations. In this study, our aim is to carefully measure some fundamental thermodynamic quantities that will help us quantify the interactions between saccharides and water and, eventually, biological macromolecules. Given these data and observations from the literature, we hope to gain some insight into the relationships between the thermophysical properties of the simple saccharides and their relative protective efficacies. In the dried state, the propensity to crystallize may be an indicator of long-term storage stability. At constant temperature and pressure, the relative values of the Gibbs free energies of the amorphous and crystalline states are one measure of the instability of the amorphous solid. That is, the larger the difference in the Gibbs energies, the greater the thermodynamic driving force for crystallization. More than 50 years ago, Rowe and Parks calculated these energies in a work among their pioneering studies on the kinetic and thermodynamic properties of the solid phases of glucose.25,26 Their studies have provided some of the motivation for this work. In an effort to ascertain the relative thermodynamic stabilities of the two states, we have measured the heat of solution of the amorphous and crystalline forms of a monosaccharide (glucose), a trisaccharide (raffinose), and 11 disaccharides: cellobiose, gentiobiose, lactose, lactulose, leucrose, maltose, melibiose, palatinose, sucrose, trehalose, and turanose. We have used heat of solution measurements, along with measurements of the glass transition and melting temperatures of the amorphous and crystalline solids, respectively, to calculate the Gibbs free energy
J. Phys. Chem. B, Vol. 104, No. 37, 2000 8877 TABLE 1: Systematic Names of Saccharides Studied cellobiose gentiobiose glucose lactose lactulose leucrose maltose melibiose palatinose (isomaltulose) raffinose sucrose trehalose turanose
4-O-β-D-glucopyranosyl-D-glucopyranose 6-O-β-D-glucopyranosyl-D-glucopyranose R-D-glucose 4-O-β-D-galactopyranosyl-D-glucopyranose 4-O-β-D-galactopyranosyl-D-fructofuranose 5-O-R-D-glucopyranosyl-D-fructopyranoside 4-O-R-D-glucopyranosyl-D-glucopyranose 6-O-R-D-galactopyranosyl-D-glucopyranose 6-O-R-D-glucopyranosyl-D-fructopyranoside O-R-D-galactopyranosyl-(1 f 6)-O-R-Dglucopyranosyl-(1 f 2)-β-D-fructofuranoside R-D-glucopyranosyl-β-D-fructofuranoside R-D-glucopyranosyl R-D-glucopyranoside 3-O-R-D-glucopyranosyl-D-fructopyranoside
change, ∆G(298.15 K), for the process
SaccharideAmorphous f SaccharideCrystalline This quantity is necessarily negative as amorphous solids are thermodynamically unstable. In the past, solution calorimetry has been championed as a sensitive technique for assessing the degree of disorder in partially crystalline mixtures.27,28 This technique is especially useful for determining the stability of solids at ambient (and lower) temperatures. Stability testing on solids is often performed at elevated temperatures in order to accelerate various chemical and physical processes. It is then assumed that the same processes occur during storage, and the results are extrapolated back to the storage temperature. This can result in a high degree of uncertainty. When coupled with heat capacity data, heat of solution calorimetry can be used to determine enthalpy changes of processes that are often obscured by chemical decomposition. For example, conventional methods, such as differential scanning calorimetry (DSC), are often unable to accurately determine the enthalpy of melting of many organic compounds because of the formation of decomposition products during melting. We also find that solution calorimetry is quite sensitive to small quantities of adsorbed water and can be used to investigate the energetics of absorption and adsorption. Materials and Methods Systematic names of all carbohydrates studied are provided in Table 1. Cellobiose, gentiobiose, maltose monohydrate, melibiose monohydrate, raffinose pentahydrate, sucrose, trehalose dihydrate, and turanose were obtained from Pfanstiehl (Waukegan, IL) at stated purities of greater than 99%, 99.5%, 99.99%, 99.9%, 98.9%, 99.8%, 99.6%, and 99.8%, respectively. Lactose monohydrate and R-D-glucose (99.99% and 99.8% pure, respectively) were obtained from Sigma Chemical Co. (St. Louis, MO). Lactulose, leucrose, and palatinose were obtained from Fluka (Milwaukee, WI) at purities greater than 99%, 98%, and 99.99%, respectively. Ethanol (100%) and phosphorus pentoxide (P4O10) were obtained from Aaper Alcohol and Chemical Co. (Shelbyville, KY) and Sigma Chemical Co., respectively. All water was prepared by a Milli-Q filtration system that yielded water with a measured conductivity of 18 MΩ cm or less. All solutions were prepared gravimetrically on an analytical microbalance; the results of all weighings were adjusted to vacuum by the appropriate buoyancy factor. It is well-known that the processing and handling of the simple saccharides can lead to changes in structure. For example, grinding or milling a sugar imparts a degree of amorphous character.29 Manufacturers typically mill crystalline compounds to obtain a uniform particle size. Also, heating the crystalline
8878 J. Phys. Chem. B, Vol. 104, No. 37, 2000 form (to remove residual water) can lead to the formation of an amorphous surface coating on the crystals. For these reasons, we recrystallized most of the above sugars to obtain compounds that were 100% crystalline. Note, however, that the excessive cost and low yields of some sugars precluded us from recrystallizing them. Recrystallization of most saccharides was achieved by first preparing a solution of each sugar at about one-half of its saturation concentration at 25 °C. This solution was mixed with three volumes of ethanol (25 °C) and poured into a flask that was constantly agitated at 4 °C. A few seed crystals were added to accelerate crystallization. After several hours or days, the resulting crystals were filtered, washed several times with cold ethanol, and dried in a vacuum oven at 50 °C for 3 days. A nitrogen purge was used in the vacuum oven to aid in removing residual solvent. All water contents were measured by coulometric Karl Fischer titration (Brinkmann Metrohm, Model 737). All transfers of product were performed in a dry nitrogen atmosphere (glovebox, relative humidity less than 1%). Details of the recrystallization technique are available in the literature.30 Preparation of each amorphous compound was done by freeze-drying a 10 wt % aqueous solution in a Virtis Genesis tray dryer. Each solution was filtered through a 20-µm syringe filter (Gelman Sciences) and dispensed dropwise into liquid nitrogen. The glass Petri dish with the frozen spherical pellets was then placed directly on the precooled freeze-dryer shelf. The shelf temperature was maintained at -45 °C under vacuum (10-2 mbar) and held for 2 h. The temperature was increased to 25 °C over 3 days. The lyophilized spheres were then dried in a vacuum oven at room temperature for 2 days; they then underwent a gradual increase in drying temperature to a maximum temperature of 40 °C below Tg (of the completely dry saccharide) for 2 days to remove residual water. This procedure typically yielded regular spheres with an average diameter of 2.4 mm. To minimize relaxational effects, all analytical work was performed within 36 h of the completion of secondary drying. Samples were handled in a dry nitrogen atmosphere (