Colloid & Polymer Science

Colloid & Polymer Sci 263:396-398 (1985)

COLLOID SCIENCE Non-freezing water in protein solutions S. All and E A. Bettelheim Department of Chemistry, Adelphi University, Garden City, New York, USA

Abstract: Differential scanning calorimetric studies of protein solution provided the freezable water content as a function of concentration. Total water content was obtained by vacuum dehydration. The difference between the two gave the non-freezing water content in protein solutions. The values obtained by these techniques were 2-3 times larger than those obtained on the basis of NMR measurements on the same solutions.

Key words: Bound water, differential scanning calorimetry, free water, proteins.

The properties of supercooled water are currently generating great interest [1]. Derbyshire [2], upon examining a number of polymeric aqueous solutions, has claimed that, to a first-order approximation, the amount of "non-freezing water is independent of the particular systems examined, it is usually in the range of 0.4 to 1.0 g of water/g of macromolecule." Simple systems such as agarose [3] and iota carrageenan [4] show concentration independent non-freezing water contents of 0.59 and 0.43 g water per g polymer, respectively. Mostly based on NMR studies, others, notably Kuntz [5], have assigned hydration (nonfreezing water) values to amino acid residues and been able to predict reasonably well the hydration of proteins on the basis of their amino acid compositions. Based on our own studies on cataract formation in lenses, we proposed a syneretic mechanism [6] by which water of hydration is released upon the collapse of the protein network; this released water enters the bulk (free water) phase. Such syneresis enhances the refractive index fluctuations and thus contributes to the development of opacities. We have shown that syneresis is involved in aging [7] and in nuclear cataract [8] formation in human lenses (as shown by light scattering analysis). NMR studies of the lenses [9,10] supported our conclusion. Finally, direct measurement of freezing and non-freezing water contents [11] in normal and cataractous lenses proved the loss of bound water and the increase in free water content upon cataractogenesis. However, the numerical values of the non-freezing water content measured by NMR EM909

were lower than those obtained by thermal analysis. Therefore, it was desirable to investigate if such discrepancies between NMR and thermal analysis also exist in simple systems such as protein solutions. Purified casein was acquired from Fisher Scientific Company (Lot 755727). Egg white lysozyme (EC.3.2.1.17), three time crystallized, and ovalbumin (Grade III crystallized and lyophilized) were obtained from Sigma Company. Bovine serum albumin, Fraction V, was purchased from Baker Chemical Company. A mixture of lens crystallins (ce and 3) was obtained from isoelectric precipitation (pH 4.5) of water soluble crystallins from calf cortex homogenates. Extensive dialysis against distilled water yielded a residue that contained a mixture of a and 3 crystallins in 62:38 weight ratio [12]. Homogeneous aqueous protein solutions were obtained with varying concentrations between 120 %; samples ranging from 6 to 12 mg of protein solution were hermetically sealed into pre-weighed aluminum pans. The heat of fusion of each sample was obtained m a differential scanning calorimeter (DuPont 990). First, the sample and the reference pan (empty aluminum pan) was cooled to - 3 0 ~ by external dry ice acetone bath. The heating rate was 3~ During the heating, a flow of nitrogen (50 cm3/min) was provided. The instrument was calibrated with a sapphire disk. The differential scanning calorimeter records endotherms of differential heat flow (Aq) as a function of time. The area under the endotherm provides the heat of fusion. All of the protein

All and Bettelheim, Non-freezing water in protein solutions

397

following values for non-freezing water expressed in g water/g protein: 1.14 for crystallins; 1.14 for lysozyme; 1.08 for casein; 0.87 for serum albumin and 0.72 for ovalbumin. These values are two or three times larger than those obtained or calculated by Kuntz (5) on the 2o basis of N M R measurements. Differences in non-freezing water content measured by DSC and N M R were found also in lens tissues. For example, 24 % of the total water was found to 16 be non-freezing water in normal human lenses by I= N M R measurements [9,10], while DSC and total water measurements gave a value of 60 0/0 [11]. Therefore, the discrepancy between DSC measurements and 12 N M R data is not due to the complexity of the lenses under investigation, but it is a systematic discrepancy. Some people may interpret the non-freezing water content in terms of kinetics, i. e., at low temperature the solution becomes freeze concentrated and due to the viscosity, supersaturation may occur which will allow crystallization only over infinitely long periods. In this concept, the system is metastable and thus the non-freezing water content cannot be identified with a special kind of water (bound water or water of hydration). Without entering into a theoretical discussion on the nature of non-freezing water content, we side with the majoritiy of the papers in the literature which 0 f | t regard non-freezing water as bound water [10 4 8 12 16 3,9,10,13]. % PKOTEnl The purpose of the present paper is to show that difFig. 1. Non-freezingwater content of protein solutions(in terms of percent of total water) as a function of protein concentration. ferent techniques under the same boundary conditions 9 9 crystallins;[] [] lysozyme;O O casein; (lowest freezing temperature) give different non-freez9 9 serum albumin;9 9 ovalbumin ing water content. This cannot be due to the different interpretations of the nature of non-freezing water content. The phenomena under investigation in N M R studies involve different molecular motions from solutions had only small melting point depression (AT those associated with absorption of heat. It seems that -0.3 ~ hence the heat of fusion of water could be while the non-freezing (bound) water molecules are used to convert the experimental heat of fusion to less likely to have translational motions than "free" freezing water content. (freezing) water, they have greater potential for reTotal water content of the protein solutions was orientation due to the protein surface in their vicinity. obtained by vacuum dehydration. The samples, in weighing bottles, were placed in a vacuum desiccator Acknowledgements over P205 and the desiccator placed in an oven at 60 ~ This work has been supportedby a grant from the NationalEye and evacuated (i x 10-3 torr). This procedure was Institute, EY 02571. repeated until constant weight was established. The amount of non-freezing water was obtained by the difReferences ference between total and freezing water content. 1. Franks F (ed) (1982) Water: A comprehensivetreatise Vol 7. Figure 1 shows the variation of non-freezing water Water and aqueoussolutionsat subzero temperatures,Plenum for the five protein solutions as a function of concenPress, New York tration. 2. DerbyshireW (ed) (1982)Franks F, The dynamicsof water in Using only the approximate straight line portion of heterogeneous systems with emphasis on subzero temperathe curves at low concentrations (1-5 %), one finds the tures, Water, Vol 7, Plenum Press, New York 24

b-

398 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Colloid and Polymer Science, VoL 263 9No. 5 (1985)

Derbyshire W, Duff ID (1974) Disc Faradays Soc 57:243-251 Baghdadi SMA (1977) PhD thesis, University of Nottingham Kuntz ID Jr (1971)J Amer Chem Soc 93:214-218 Bettelheim FA (1979) Exp Eye Res 28:189-197 Siew EL, Opalecky D, Bettelheim FA (1981) Exp Eye Res 33:603-614 Bettelheim FA, Siew EL, Chylack LTJr (1981) Invest Ophthalmol 29:348-354 Racz P, Tompa K, Pocsik I (1979) Ezp Eye Res 28:129-135 Racz P, Tompa K, Pocsik I, Banki P (1983) Lens Res 1:199-206 Bettelheim FA, Christian S, Lee LK (1983) Current Eye Res 2:803-808 Bettelheim FA, Wang TJY (1977) Exp Eye Res 25:613-620

13. Ohno H, Shibayama M, Tsuchida E (1983) Makromol Chem 184:1017-1024 Received July 16, 1984; accepted January 9, 1985

Authors' address: E A. Bettelheim, S. All Department of Chemistry Adelphi University Garden City, New York 11530, USA