Role of protein conformation and aggregation in pumping water in and out of a cell

Cell Biology International 30 (2006) 78e85 www.elsevier.com/locate/cellbi I.L. Cameron a, K.M. Kanal b, G.D. Fullerton b,* a Department of Cellular and Structural Biology (ILC), The University of Texas Health Science Center at San Antonio, San Antonio, TX 78284, USA b Department of Radiology (KMK, GDF), The University of Texas Health Science Center at San Antonio, San Antonio, TX 78284, USA Received 30 May 2005; accepted 30 September 2005 Abstract Dialysis cassettes containing BSA solutions were used to simulate passive in vivo conditions to assess the effect of protein conformation and aggregation on cell water content. The cassettes were suspended in dextran solutions to provide a range of fixed osmotic stress values simulating blood plasma. The system was placed on a shaker for 24 h to attain equilibrium. Four manipulation methods; pH, cosolute salt concentration, e.g. NaCl, temperature annealing and urea concentration denaturant were varied to produce well-known manipulations of BSA conformation. It was observed that the cell water content varied from þ14% to about ÿ13% with changes in protein conformation and aggregation. The findings demonstrate that a change in protein conformation and aggregation, pumps water in and out of a cell to maintain equilibrium % water content matching the protein conformational hydration parameter. This concept supplements existing theories on cell volume regulation. Ó 2005 International Federation for Cell Biology. Published by Elsevier Ltd. All rights reserved. Keywords: Osmosis; Interfacial water; Osmotically unresponsive water; Protein conformation; Cell water 1. Introduction 1.1. Background Cells are most frequently considered as simple osmotic sacks that require expenditure of energy via ATP to power a so- dium/potassium pump to maintain higher ion concentration outside the cell and compensate for higher protein concentra- tion inside. These pumps are attributed exclusively to trans- membrane proteins. This study proposes that cytoplasmic proteins contribute significantly to passive changes in colliga- tive equilibrium water content. Recently, we showed that pro- tein unfolding causes significant changes in transmembrane osmotic pressure necessary to maintain water equilibrium (Kanal et al., 1994; Kanal, 1996; Zimmerman et al., 1995; Fullerton et al., 2006). This observation suggests an alternate * Corresponding author. E-mail address: fullerton@uthscsa.edu (G.D. Fullerton). cytoplasmic protein water pump mechanism that may play a significant but as yet unevaluated biological role in cell volu me regulation. Because it is not possible to rule out active response of a living cell to extracellular tonicity, it was decid- ed to circumvent this problem by study of volume changes in a dialysis cassette. Knowledge of protein folding is based on information ex- tracted primarily from x-ray diffraction studies of protein crys- tals. Changes in bovine serum albumin (BSA) folding or unfolding in solution have been studied with NMR (Jeng and Englander, 1991; Englander and Mayne, 1992), viscosity (Yang and Foster, 1954; Tanford et al., 1955), optical rotary dis- persion (Imahori, 1960), Raman spectroscopy (Lin and Koenig, 1976) and circular dichroism (Takeda et al., 1987, 1988, 1989). The CD method of Takeda allows extraction of the relative amounts of a-helix, b-sheet and random coil components of BSA under differing environmental conditions. This data pro- vided good understanding of BSA response to environmental factors such as temperature, co-solute salt, and pH. 1065-6995/$ - see front matter Ó 2005 International Federation for Cell Biology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cellbi.2005.09.010 I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 1.2. Structure and behavior of intracellular water The state of water in the cell has been a topic of controversy. It is believed by some that the water near the surface of the cell constituents behaves as if bound to the surface with dimin- ished activity and solvent properties (House, 1974). Cameron et al. (1988a) have found that in both the intra- and extra- cellular environments of the human erythrocyte, a significant and measurable fraction of water differs from bulk water in motional properties. In a separate report, Cameron et al. (1988b) have also used crystalline lenses to study and measure the properties of cellular water. Their results led them to conclude that most, if not all, of the water in lens cells has motional and osmotic properties that distinguish it from bulk water. These reports give a brief sample of some of the extensive literature regarding the state of the cellular water. However, although a significant fraction of cell water may indeed have altered properties; osmotic experiments provide little evidence to sup- port the idea that the structure of the intracellular water signif- icantly affects the equilibrium thermodynamics or the kinetics of water flow in cells. 1.3. Cellular pumps The pump-and-leak concept for maintenance of cell volume was first formulated by Davson (1940), Ussing (1960), Toste- son and Hoffman (1960) and see Strange (1994). A cellular pump is different from other cellular transport pathways in its requirement for the coupling of metabolic energy to trans- locate ions uphill against their electrochemical gradients. Sev- eral pumps are critical for cellular function, some of which include Kþ/Hþ exchange pump found in parietal cells of the gastric mucosa (Rabon and Reuben, 1990); Hþ-ATPases found in vacuolar and plasma membranes (Swallow et al., 1990); the Ca2þ-ATPase found in mitochondrial and plasma membranes (Carofoli, 1991). But the most important pump for cell volume regulation is the Naþ,Kþ-ATPase found in the plasma mem- brane of nearly all animal cells (De Weer, 1992). In the Naþ,Kþ-ATPase (or sodium pump), three intracellular Naþ ions are exchanged for two extracellular Kþ ions with the hy- drolysis of one molecule of ATP for each full pumping cycle. This 3:2:1 stoichiometry remains constant under different physiological conditions (De Weer, 1992). This sodium pump is a self-stabilizing system in which if the Naþ entry into the cell increases, then the rate of Naþ extrusion increases until Naþ efflux once again balances Naþ influx. The most im- portant function of the sodium pump in cells is to maintain transmembrane ion gradients, which serve as energy sources that drive many other transport processes. It also plays an im- portant role in the maintenance and modulation of the resting membrane potential in all cells (De Weer, 1992). 1.4. Cellular leaks In comparison to the cellular pump, cellular leak pathways are passive and do not require direct coupling to an external 79 energy source. Leak pathways are further classified into mem- brane channels and coupled transporters. (i) Membrane channels are transmembrane proteins, con- sisting functionally of a pore and some selectivity mechanism, that allow the chemically uncoupled passage of small ions, usually in response to various stimuli. The selectivity of the ions is rather low and depends on the size and charge of the ion. Channels do not bind the transported ion tightly and there- fore the maximum turnover rate is very high. Major conforma- tional changes in the channel are unnecessary for the ion transport to occur. The transport rate is a linear function of substrate concentration under physiological conditions. Hodgekin and Huxley (1952) first identified voltage-gated Naþ channels in squid giant axons. The best-characterized ion channels include Naþ, Kþ, Ca2þ and Clÿ channels (Hillie, 1992). Transport of amino acids and other nonionic solutes through simple leak pathways were also identified (Lambert, 1985). In 2003 the Nobel prize was awarded to Agre and MacKinnon for discovery of a water channel (Agre and Koyona, 2003). (ii) Coupled transporters in contrast to channels, are highly selective among substrates, have a lower turnover rate and usu- ally saturate at concentrations not much above the physiolog- ical range. A number of activation mechanisms including external ligands (Raftery et al., 1980; Schofield et al., 1987; Grenningloh et al., 1987; Westbrook and Jahr, 1989), various intracellular signals and second messengers (Hartzell, 1988; Trautwein and Hescheler, 1990; Von Tcharner et al., 1986) and mechanical stretch/deformation of the membrane (Morris, 1990) are shared by coupled transport systems and channels. The research results reported herein demonstrate that changes in endoplasmic protein conformation and aggregation also play a significant role responsible for pumping water into and out of cells. The results further suggest that the global P average protein hydration parameters iI (g water/g pro- tein)  (Mi/MT) for the cell may be an unappreciated factor in measuring cell water content (i is the protein index, Mi is P the mass of protein, Mi 1⁄4 total protein mass 1⁄4 MT). 2. Materials and methods Table 1 gives a summary of the experiments done to determine how mod- ification of protein conformation and protein aggregation can influence the amount of water pumped in and out of a cell (cassette). 2.1. Temperature annealing experiments Five-milliliter solutions of bovine serum albumin (BSA) (Sigma Chemical Co., St. Louis, MO, 98e99% albumin, chemical MW 66,336) were prepared in 150 mmolal NaCl, 15 mmolal NaCl or pure water. The solutions were fil- tered (0.45 mm filters) (Gelman Sciences, Inc. Ann Arbor, MI) and degassed in a vacuum for 15 min at ÿ96 kPa. Cell behavior was modeled using dialysis cassettes containing the BSA solution in osmotic equilibrium with an outer dextran solution, used to simulate mammalian plasma, to provide fixed osmotic stress (Table 2) using the methods of Parsegian et al. (1986). The outer com- partment was sufficiently large so that no changes in the dextran concentration in the outer compartment could be observed. Parsegian et al. measured the os- motic pressure P (dyn/cm2) of various polymers and established the following I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 80 Table 1 Table 2 Pumping method Maximum mass Molecular protein Experimental design: known concentrations of BSA solutions were placed in Simulated osmotic water pumping change (DM/Mc  100%) mechanism dialysis cassettes and exposed to dextran induced osmotic stresses of 62, 400 and 1000 cm water for 24 h Protein Range of change Temperature ÿ7% (Water out) Unfold and cross-link conformation for conformation annealing Decrease AN method method (25e70  C) Decrease ASA Protein variable heating 25 Salt e NaCl ÿ7% (Water out) Fold-lose segments in pure water 25 (1e2000 mM) Decrease AN in 15 mM NaCl 25 Decrease ASA in 150 mM NaCl 25 in 150 mM NaCl 25 in 150 mM NaCl Urea e 150 mmolal ÿ14% (Water out) Cross-link NaCl (0e8 M) Decrease AN Decrease ASA pH e 200e300 +14% (Water in) Disaggregation above mosmol NaCl isoelectric point (pH 1⁄4 5.4e9.0) and then unfold Increase N (near 5.4) Increase ASA Decrease AN (near 9)     Ce75 Ce75  Ce75  Ce75  Ce75 C C  C  C  C Extracellular osmotic pressure (cm water) 62 62 62 400 1000 Protein at 22  C in variable NaCl 62 Protein (22  C, 150 mM NaCl) in variable Urea 62 n, number of particles; ASA, accessible surface area; AN, apparent number of 1e2000 mM 400 0e8 M 400 particles due to segmental motion. 1e2000 mM 1000 0e8 M 1000 1e2000 mM 400 plus heat anneal 0e8 M 1e2000 mM Protein (22  C, 100e200 mM NaCl) in variable pH 62 with solvents consisting of 150 mmolal NaCl and urea ranging in concentra- 5.4e9.0 400 tion from 0 to 8 molal. The solutions were filtered (0.45 mm filters) (Gelman 5.4e9.0 1000 Sciences, Inc. Ann Arbor, MI) and degassed in a vacuum for 15 min at 5.4e9.0 ÿ96 kPa. Five methods were used to manipulate protein conformational changes. Gravi- metric measurement of the wet and dry weights of cassette contents were used to calculate the amount of water pumped in or out of the cassette/cell by change in protein conformation. relationship between the osmotic pressure and the weight percent ‘w’ of dextran: log1⁄2Pdex ðwÞŠ 1⁄4 2:75 þ 1:03w0:383 ð1Þ A portion of each BSA solution was transferred to six dialysis cassettes (Pierce Slide-A-Lyzer dialysis cassette, Pierce Co., Rockford, IL), which were then heated to different temperatures (37, 42, 46, 50, 60, 70  C) for 1 h and then allowed to cool down to room temperature (25  C) before osmotic measurements of equilibrium concentration were made. The dialysis cassettes containing the BSA solutions were then immersed in three separate beakers with dextran 5.9% (wt dextran/wt solvent) salt mixture (0, 15 and 150 mM NaCl) solutions of concentration. This gave constant osmotic stress of 62 cm water calculated using Eq. (1). 2.2. Salt experiments Five-milliliter solutions of bovine serum albumin (BSA) (Sigma Chemical Co., St. Louis, MO, 98e99% albumin, chemical MW 66,360) were prepared in solvents with NaCl concentrations ranging from 1 to 2000 mmolal. The sol- utions were filtered (0.45 m filters) (Gelman Sciences, Inc. Ann Arbor, MI) and degassed in a vacuum for 15 min at ÿ96 kPa. Cell behavior was modeled us- ing dialysis cassette with an outer dextran solution to provide fixed osmotic stress. The dialysis cassettes containing the BSA solutions were then im- mersed in three separate beakers consisting of dextran and salt solutions of dif- ferent concentrations, 5.9%, 14,17% and 20% (wt dextran/wt solvent), respectively. This gave constant osmotic stress of 62, 400, and 1000 cm water, respectively. The procedure was repeated for BSA solutions heated to 80  C for 1 h, cooled to room temperature and then subjected to a fixed osmotic stress of 400 cm water. 2.3. Urea experiments Five-milliliter solutions of bovine serum albumin (BSA) (Sigma Chemical Co., St. Louis, MO, 98e99% albumin, chemical MW 66,336) were prepared 2.4. pH experiments Five-milliliter solutions of bovine serum albumin (BSA) (Sigma Chemical Co., St. Louis, MO, 98e99% albumin, chemical MW 66,336) were prepared in pH in the range from 5.4 to 9 using various buffers. Buffers used to prepare these solutions have been described in detail elsewhere (Kanal et al., 1994). In each instance the osmolality of the sample was measured by freezing point de- pression and maintained between 0.2 and 0.3 osmol/kg by adding NaCl as re- quired to eliminate dependence on ionic strength. The pH of each solution was measure with an Orion research digital pH/millivolt meter 611 (Orion Re- search, Cambridge, MA). The meter was calibrated with pH 7.0 and pH 4.0 buffers for acid solutions and pH 7.0 and pH 10.0 for basic solutions. The sol- utions were filtered (0.45 mm filters) (Gelman Sciences, Inc. Ann Arbor, MI) and degassed in a vacuum for 15 min at ÿ96 kPa. 3. Results 3.1. Temperature annealing experiments The change in % water of a cell subjected to constant os- motic stress (62 cm water) for BSA in 150 mmolal, 15 mmolal NaCl and pure water as a function of annealing temperature (degrees Celsius) is shown in Fig. 1 (top). The plot also dem- onstrates the change in % water for BSA in 150 mmolal NaCl for each annealing temperature as a function of the applied os- motic stress (400 and 1000 cm water). For each salt concentra- tion and annealing temperature, experiments were performed three times to estimate experimental reproducibility. The mean of the three experiments is plotted for each salt concen- tration and annealing temperature. One-way ANOVA was used to determine the statistical significance of the data. The % wa- ter content was decreased significantly for all the BSA solu- tions as a function of annealing temperature above 50  C ( p < 0.0001, n 1⁄4 21). There is no change in % water content I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 81 Fig. 2. Change in % cell water content due to exposure to increasing salt con- centrations at different osmotic pressures and to a BSA annealing temperature of 80  C. Fig. 1. (Top) Change in % cell water content due to exposure to different tem- peratures at different osmotic pressures and salt conditions. (Bottom) Cartoon of change in % cell water content before and after exposure of BSA to anneal- ing temperatures greater than 50  C. up to 50  C and then a decrease thereafter up to 70  C for BSA solutions in both, 150 and 15 mmolal NaCl and also in water with zero NaCl subjected to osmotic stress of 62 cm water. This decrease in % water content was also observed for BSA in 150 mmolal NaCl with increasing annealing temperature as a function of the applied osmotic stresses, 400 and 1000 cm water. The protein is in a native state up to 50  C whereupon de- naturing occurs with an increase in annealing temperature leading to a decrease in % water content and an increase in % BSA content. As the osmotic stress increases, there is a greater decrease in the % water content for the BSA solutions in 150 mmolal NaCl with an increase in annealing tempera- ture. Fig. 1 (bottom) is a cartoon presentation of the change in % water content as a function of annealing temperature. Above 50  C, the protein denatures, the solvent accessible sur- face area increases, leading to an efflux of water from the cell. The decrease in % water content due to annealing ranged from ÿ8% for 150 mmolal NaCl, ÿ7% for 15 mmolal NaCl and ÿ2% for pure water. The decrease in % water content as a func- tion of the annealing temperature for BSA solutions in 150 mmolal, subjected to a constant osmotic stress of 400 and 1000 cm water ranged from ÿ8% to ÿ8.5%, respectively. of the data. The % water content for all the BSA solutions de- creased significantly as a function of the NaCl concentration ( p < 0.0001, n 1⁄4 24). The % water content decreases at first, up to about 150 mmolal NaCl, and then flattens out and re- mains constant with an increase in NaCl concentration up to 2000 mmolal. The plot also shows a decrease in % water con- tent for BSA solutions heated to an annealing temperature of 80  C for 1 h. 3.3. Urea denaturation experiments Fig. 3 shows the change in % solvent content for BSA sol- utions in urea concentrations ranging from 0 to 8 mmolal ex- posed to different osmotic pressures. In these urea experiments the results are reported in terms of solvent instead of water as the high concentration of urea may have influenced the dry mass measurements. One-way ANOVA was used to de- termine the statistical significance of the data. The % solvent content for the BSA solutions exposed to a fixed osmotic stress of 400 cm water decreased initially up to 2 molal urea, then in- creased up to 4 molal urea, before decreasing thereafter with increasing urea concentration ( p < 0.0001, n 1⁄4 12). For the BSA solutions exposed to a fixed osmotic stress of 1000 cm water, there was a significant ( p < 0.0001, n 1⁄4 12) decrease in % solvent content with increasing urea concentration up to 6 molal and then a plateau was reached from 6 to 8 molal urea. 3.2. Salt experiments Fig. 2 shows the change in % water content for BSA solu- tions in salt concentrations ranging from 0 to 2000 mmolal ex- posed to different osmotic pressures. The mean of three experiments is plotted for each salt concentration. A total of eight data points for each osmotic pressure are plotted. One- way ANOVA was used to determine the statistical significance Fig. 3. Change in % cell solvent content due to exposure to the protein dena- turant, urea, at different osmotic pressures. 82 I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 The results can be explained as follows. As the urea con- centration increases, the protein unfolds to a random coil, and then is almost completely unfolded giving rise to cross- linking between BSA molecules, thereby increasing the % sol- vent content and decreasing the % BSA content. The gel-like appearance of the BSA at 6 and 8 M urea concurs with this conclusion. For BSA solutions exposed to pressures of 62, 400, and 1000 cm water, the decrease in % solvent content as a function of increasing urea concentration is ÿ13.6%, ÿ12.2% and ÿ10.8%, respectively. 3.4. pH experiments Fig. 4 shows the changes in % water content as a function of pH for BSA solutions exposed to a fixed osmotic stress of 62, 400 and 1000 cm water. The mean of the three experiments is plotted for each pH. A total of four data points are plotted for each osmotic stress. One-way ANOVA was used to determine the statistical significance of the data. There was no significant change in the % water content for the solutions exposed to a fixed osmotic stress of 62 cm water. For the BSA solutions subjected to a fixed osmotic stress of 400 and 1000 cm water, there was a significant increase in the % water content as the pH increased from 5.4 to 9 ( p < 0.0001, n 1⁄4 12). As the pH shifted above the isoelectric point (pH 5.4) where the BSA is partially aggregated, there is an increase in the negative charges on the surface of the protein causing self-repulsion, leading to a disaggregation of the protein mol- ecules. This in turn increases the solvent-accessible surface area of the protein, thereby increasing the % water content and decreasing the % BSA content. The aggregation at pH 5.4 and the disaggregation effect at higher pH values is depen- dent on the BSA concentration as indicated by the data for the BSA solutions exposed to a osmotic stress of 62 cm water. In this case since the protein concentration is low, there is a re- duced probability of proteins aggregating indicating no change in % water content as a function of pH. The increase in the % water content with increasing pH ranged from 9.1% to 14.3% for the BSA solutions exposed to a osmotic stress of 400 and 1000 cm water, respectively. The four conformational states, cross-linked gel caused by temperature annealing, molten globule effect caused by Fig. 4. Change in % cell water content due to exposure to increase in pH above the isoelectric point of BSA, i.e. pH 5.4, at different osmotic pressures. varying the NaCl concentration, random coil caused by urea and aggregation/deaggregation caused by altering pH, are dis- cussed separately. 4. Discussion 4.1. Temperature annealing experiments (cross-linked gel effect) As seen in Fig. 1, we see a continuous decrease in % water content with an increase in annealing temperature above 50  C for BSA in 150 mmolal NaCl, 15 mmolal NaCl, and zero salt. These are under conditions of fixed osmotic stress (62 cm water). The plot also shows the increase in % BSA content for BSA in 150 mmolal NaCl subjected to osmotic stress of 400 and 1000 cm water. For these BSA solutions too, there is a decrease in the % water content with increasing osmotic stress as well as increasing annealing temperatures. As sum- marized by Lin and Koenig (1976) the protein refolds to its original state up to a threshold annealing temperature (~50  C) beyond which it undergoes a irreversible conforma- tion change. Thus, up to 50  C, no changes in the % water content were observed. Above 50  C, the protein changes from a native state to a denatured (cross-linked) state causing the protein surface-accessible area to increase and the effective number of molecules to decrease. Thus there is more water associated with the protein surface but an even larger contrary change in % water content due to decrease in the effective number of protein particles due to cross-linking. This leads to a net efflux of water from the cell, increasing the equilibri- um % BSA content and decreasing the % water content for BSA solutions as a function of the annealing temperature, subjected to a constant osmotic stress of 62 cm water ranged from ÿ8% for 150 mmolal NaCl, ÿ7.3% for 15 mmolal NaCl and ÿ2% for pure water. The decrease in % water content for BSA solutions in 150 mmolal with increase in the annealing temperature, subjected to a constant osmotic stress of 400 and 1000 cm water ranged from ÿ8% to ÿ8.5%, respectively. 4.2. Salt experiments (molten globule effect) Fig. 2 shows the change in % water content for BSA solu- tions exposed to different osmotic pressures, as a function of NaCl concentration ranging between 0 and 2000 mmolal. Our previous studies (Zimmerman et al., 1995) have indicated the protein to be in a molten globular state in pure water and for low NaCl concentration of up to 50 mmolal. The protein is more stable and in a native configuration in NaCl concentration greater than 100 mmolal. The decrease in % water content with increasing NaCl concentration agrees with our previous obser- vations on the conformational state of the protein. Increase in the salt concentration causes the protein to fold more com- pactly, decreasing not only the solvent-accessible surface area but also the effective number of protein molecules (loss of the segmental motion effect), leading to a decrease in % water content. For the BSA solutions exposed to a pressure I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 of 62 cm water, the decrease in % water content with increas- ing NaCl concentration is ÿ3.1%. For the BSA solutions (not heated) exposed to a fixed osmotic stress of 400 cm water, the decrease in % water content is ÿ7%. The decrease in % water content with increasing NaCl concentrations for the BSA solutions exposed to an osmotic stress of 1000 cm water is about ÿ4%. 4.3. Urea experiments (random coil) Fig. 3 shows the change in % solvent content for BSA sol- utions subjected to a fixed osmotic stress of 62, 400 and 1000 cm water as a function of urea concentration. As indicated by the plot, there is a decrease in the % solvent content with increasing urea concentrations from 0 to 8 molal for BSA sol- utions exposed to fixed osmotic stresses of 62, 400 and 1000 cm water. For the BSA solutions subjected to a fixed os- motic stress of 400 cm water, there is an initial decrease in the % solvent content with increasing urea up to 2 molal urea and then a decrease up to 4 molal urea before increasing thereafter up to 8 molal urea. However, when exposed to a osmotic stress of 1000 cm water, the % solvent content of the BSA solutions 83 increases with increasing urea concentration up to 6 molal urea and thereafter reaches a plateau up to 8 molal urea. Urea is a strong denaturant and is known to denature BSA (Takeda and Miaura, 1981; Takeda et al., 1987, 1988). As the urea concentration increases, the protein unfolds to a random coil, and then is almost completely unfolded giving rise to cross-linking between BSA molecules, thereby decreasing the % solvent content and leading to an increase in the % BSA content. The gel-like appearance of the BSA at 6 and 8 molal urea concurs with this conclusion. However, this does not seem to be the case for the BSA solutions exposed to an osmotic stress of 400 cm water. There is an initial de- crease in % solvent content from 0 to 2 M urea and then an increase at 4 M. This initial decrease was, however, not very large but remains unexplained. For the BSA solutions exposed to a pressure of 62 cm water, the decrease in % solvent content as a function of increasing urea is ÿ13.6%. For the BSA sol- utions exposed to a fixed osmotic stress of 400 cm water, the decrease in % solvent content with increasing urea concentra- tion is ÿ12.2% and for the BSA solutions exposed to an os- motic stress of 1000 cm water, the decrease in % water content is about ÿ10.8%. Fig. 5. Plots of the Mw/Mp ratio as a function of inverse osmotic pressure 1/p as proposed by the Fullerton molecular model (Fullerton et al., 2005) gives the correct molecular weight calculated from the slope (S h the slope of the regression line 1⁄4 RTr/Ae) for BSA in native physiological conditions as demonstrated using BSA data from Scatchard et al. (1946) on the upper left. The large osmotically unresponsive water (I 1⁄4 3.84 g/g) completely describes nonideal osmotic behavior. Com- parison of the Scatchard data plotted as a dotted line in the other panels with BSA in modified conformations from high temperature annealing, pH and Urea in this study show that both effective molecular weight inversely related to the slope S, Ae 1⁄4 Constant/S, and changes in the large I value are present. Evaluation of these plots allows separation of osmotic pressure change due changing effective molecular weight (aggregation, cross-linking or segmental motion effects) from that due to changes in solute solvent interaction parameter, I(g/g), caused by water displacement from protein folding, aggregation and ligand binding. As expected the largest changes in slope or molecular weight are induced by Urea or heating while the largest changes in osmotically unresponsive water are caused charge induced molecular expansion resulting from pH changes away from the isoelectric point. 84 I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 4.4. pH experiments (aggregation) There was no significant change in the % water content for the BSA solutions exposed to an osmotic stress of 62 cm wa- ter. However, for the BSA solutions subjected to 400 and 1000 cm water, there was a significant increase in the % water content with increasing pH ( p < 0.0001, n 1⁄4 12). This can be explained as disaggregation of the proteins as they move away from the isoelectric point (pH 5.4). Aggregation at the isoelec- tric point is a well-known phenomenon (Creighton, 1984; Lehninger, 1975). As the pH is increased beyond 5.4, there is an increase in the negative charges on the surface of the pro- tein causing self-repulsion, leading to a disaggregation of the protein molecules. This repulsion in turn causes an increase in the solvent-accessible surface area of the protein leading to an increase in the % water content and an increase in the % BSA content. However, this aggregation and disaggregation effect is dependent on the BSA concentration as indicated by the data for BSA solutions exposed to a stress of 62 cm water. In this case, since the protein concentration was low, there is a reduced probability of the protein molecules aggregating, thus there is no indication of a change in the % water content as a function of pH. The increase in % water content as a func- tion of increasing pH ranged from 9.1% to 14.3% for the BSA solutions exposed to a stress of 400 cm water and 1000 cm water, respectively. 4.5. Molecular analysis Analysis of the above results is made simpler by use of the molecular analysis procedure discussed in the companion pa- per in this issue (Fullerton et al., 2005). As shown in Fig. 5 plots of Mw/Mp as a function of 1/p allows separation of effects due to the effective molecular weight (measured by the slope) and the solute/solvent interaction parameter (measured by the y-intercept). These analyses can be further elaborated by plotting effective molecule weight Ae and solute solvent in- teraction parameter I as a function of the controlling parameter as we have done for salt concentration in Fig. 6. At very low concentration of NaCl, Ae is 40,000 Da due to electrostatic denaturation effects (see details in companion paper) but approaches the chemical molecular weight at physiological salt concentrations above 100 mM. The open, denatured state of the BSA molecule also causes the large I value (I 1⁄4 3.25) which reduces to I 1⁄4 2.25 at 500 mM NaCl concentration due to more compact protein folding. 4.6. Alternate protein pump mechanism The observations indicate that the % water content is influ- enced strongly by all types of protein conformational change (molten globule, aggregation, random coil and cross-linked gel) irrespective of the manipulation method used to control the conformation. Changes in % water content ranging be- tween þ14.3% and ÿ13.6% were observed due to changes in the protein conformation. These findings show the mass of water per mass BSA pumped in and out of the cell as Fig. 6. Plots of effective molecular weight Ae and solute solvent interaction parameter I(g water/g dry mass) as a function of NaCl concentration shows changes in both values as a result of protein denaturation at low salt concen- trations as described in the text. Osmotic compression from the salt is neces- sary for BSA to maintain the compact globular form of the native protein. The protein in distilled water is unstable causing segmental motion and a decrease in the apparent molecular weight. The open molten globule contains more water in the volume occupied by the molecule causing an increase in I(g/g). a function of the BSA conformation and the applied osmotic stresses of 62, 400 and 1000 cm water for each manipulation method used to change the BSA conformation or aggregation. There was a decrease in the mass water per mass BSA pumped in and out of the cell as the applied osmotic stress increases. The findings indicate that changes in protein conformation and aggregation cause changes in the mass water per mass BSA pumped in and out of the cell. As such protein conforma- tional and aggregation changes occur in biological cells, it is suggested that such changes constitute an endoplasmic mech- anism to pump water in and out of the cell. Cytoplasmic pro- teins thus contribute in a passive colligative sense to the equilibrium water content. The cytoplasm contains a variety of proteins that perform various functions and provide a general framework that is used to help organize the many enzymatic reactions. About 50% of the ‘‘dry’’ weight of cells is protein. 4.7. Conclusions The results of this study suggest an alternate cytoplasmic protein water pump in addition to existing theories mentioned in the introduction that potentially can play a significant role in biological systems. Natural biological phenomena such as changes in cosolute concentrations and pH in the cell are I.L. Cameron et al. / Cell Biology International 30 (2006) 78e85 sources for change in protein conformation and aggregation. The reversible covalent modification of proteins provides an important mechanism for regulating the activity of specific proteins in the cell. For example, the activities of many cellu- lar proteins are controlled by cycles of phosphorylation and dephosphorylation. Thus protein conformation and aggrega- tion are demonstrated to play a significant role in altering the equilibrium water content of a cell. As we show in a related paper in this issue the primary factor maintaining the water content of the cell is the average solute solvent interaction parameter for all macromolecular components of the cell. (Because of the crowded nature of proteins in cells, an even smaller cellular volume change can cause a large change in the chemical activity of the solutes.) References Agre P, Koyona D. Aquaporin water channels: molecular mechanism for human disease. FEBS Lett 2003;555:72e8. Cameron IL, Ord VA, Fullerton GD. Water of hydration in the intra- and extra- cellular environment of human erythrocyte. Biochem Cell Biol 1988a;66: 1186e99. Cameron IL, Contreras E, Fullerton GD, Kellermayer M, Ludany A, Miseta A. Extent and properties of non-bulk bound water in crystalline lens cells. J Cell Physiol 1988b;137:125e32. Carofoli E. The calcium pumping ATPase of the plasma membrane. Annu Rev Physiol 1991;53:531. Creighton TE. Proteins: structures and molecular principles. New York: W.H. Freeman; 1984. Davson H. The permeability of the erythrocyte to cations. Cold Spring Harbor Symp Quant Biol 1940;8:225. De Weer P. Cellular sodium-potassium transport. In: Seldin DW, Giebisch G, editors. The kidney: physiology and pathophysiology. 2nd ed. New York: Raven Press; 1992. Englander SW, Mayne L. Protein folding studied using hydrogen-exchange labeling and two-dimensional NMR. Annu Rev Biophys Biomol Struct 1992;21:243e65. Fullerton GD, Kanal KM, Cameron IL. Osmotically unresponsive water frac- tion on proteins and in cells. Cell Biol Int 2006;30:86e92. Grenningloh G, Rienitz A, Schmitt B, Methfessel C, Zensen M, Beyreuther K, et al. The strychnine-binding subunit of the glycine receptor shows homol- ogy with nicotinic acetylcholine receptors. Nature 1987;328:215e20. Hartzell HC. Regulation of cardiac ion channels by catecholamines, acetylcho- line and second messenger systems. Prog Biophys Mol Biol 1988;52:165. Hillie B. Ionic channels of excitable membranes. 2nd ed. Sunderland, MA: Sinauer Associates; 1992. Hodgekin AL, Huxley AF. Currents carried by sodium and potassium ions through the membrane of the giant axon of Loligo. J Physiol 1952;116: 449e72. House CR. Water transport in cells and tissues. London: Edward Arnold; 1974. Imahori K. Rotary behavior of protein denaturation. Biochim Biophys Acta 1960;37:336e41. Jeng MF, Englander SW. Stable submolecular folding units in a non-compact form of cytochrome c. J Mol Biol 1991;221:1045e61. Kanal KM, Fullerton GD, Cameron IL. A study of the molecular sources of nonideal osmotic pressure of bovine serum albumin solutions as a function of pH. Biophys J 1994;66:153e60. 85 Kanal KM. Influence of globular protein conformation on cellular hydration and NMR contrast. PhD dissertation, University of Texas Health Science Center at San Antonio, Texas, 1996. Lambert IH. Taurine transport in Ehrlich ascites tumor cells. Specificity and chloride dependence. Mol Physiol 1985;7:323. Lehninger AL. Biochemistry. 2nd ed. New York: Worth; 1975. p. 157e82. Lin V, Koenig JL. Raman studies of bovine serum albumin. Biopolymers 1976;15:203e18. Morris CE. Mechanosensitive ion channels. J Membr Biol 1990;113:93. Rabon EC, Reuben MA. The mechanism and structure of the gastric H, K-ATPase. Ann Rev Physiol 1990;52:321. Parsegian VA, Rand RP, Fuller NL, Rau DC. Osmotic stress for the direct measurement of intermolecular forces. Methods Enzymol 1986;127: 400e16. Raftery MA, Hunkapillar MW, Strader CD, Hood LE. Acetylcholine receptor: complex of homologous subunits. Science 1980;208:1454e6. Scatchard G, Batchelder AC, Brown A. Preparation and properties of serum and plasma proteins. VI. Osmotic equilibria in solutions of serum albumin and sodium chloride. J Am Chem Soc 1946;68:2320e9. Schofield PR, Darlison MG, Fagita N, Burt DR, Stephenson FA, Rodriguez H, et al. Sequence and functional expression of the GABAA receptor shows a ligand-gated receptor super-family. Nature 1987;328:221e7. Strange K. Cellular and molecular physiology of cell volume regulation. Boca Raton, FL: CRC Press; 1994. Swallow CJ, Grinstein S, Rotstein OD. A vacuolar type Hþ-ATPase regu- lates cytoplasmic pH in murine macrophages. J Biol Chem 1990;265: 7645. Takeda K, Miaura M. Stepwise formation of complexes between sodium do- decyl sulphate and bovine serum albumin detected by measurements of electrical conductivity, binding isotherms and circular dichroism. J Colloid Interface Sci 1981;82:38e44. Takeda K, Sasa K, Kawamoto K, Wada A, Aoki K. Secondary structure changes of disulfide bridge-cleaved bovine serum albumin in solutions of urea, guanidine hydrochloride, and sodium dodecyl sulfate. J Colloid Interface Sci 1988;124:282e9. Takeda K, Shigeta M, Koichiro A. Secondary structures of bovine serum albu- min in anionic and cationic surfactant solutions. J Colloid Interface Sci 1987;117:120e6. Takeda K, Wada A, Yamamoto K, Moriyama Y, Aoki K. Conformational change of bovine serum albumin by heat treatment. J Protein Chem 1989; 8:653e9. Tanford C, Swanson SA, Shore WS. J Am Chem Soc 1955;77:6414. Tosteson DC, Hoffman JF. Regulation of cell volume by active cation trans- port in high and low potassium sheep red cells. J Gen Physiol 1960;44: 169. Trautwein W, Hescheler J. Regulation of cardiac L-type calcium current by phosphorylation and G proteins. Annu Rev Physiol 1990;52:257. Ussing HH. Active and passive transport of the alkali metal ions. In: Ussing HH, Kruhoffer P, Hess Thaysen J, Thorn NA, editors. The alkali metal ions in biology. Berlin: Springer-Verlag; 1960. Von Tcharner V, Prod’hom B, Baggiolini M, Reuter H. Ion channels in human neutrophils activated by a rise in free cytosolic calcium concentration. Nature 1986;324:369. Westbrook GL, Jahr CE. Glutamate receptors in excitory neurotransmission. Semin Neurosci 1989;1:103. Yang JT, Foster JF. Changes in the intrinsic viscosity and optical rotation of bovine plasma albumin associated with acid binding. J Am Chem Soc 1954;76:1588e95. Zimmerman RJ, Kanal KM, Sanders J, Cameron IL, Fullerton GD. Osmotic pressure measures of salt induced folding/unfolding of bovine serum albu- min. J Biochem Biophys Methods 1995;30:113e31.