Constant Enthalpy Change Value during Pyrophosphate Hydrolysis within the Physiological Limits of NaCl*

Background: Decreased water activity was thought to result in smaller enthalpy change values during pyrophosphate hydrolysis. Results: The enthalpy change in pyrophosphate hydrolysis caused by halophilic and non-halophilic enzymes is constant up to 4.0 m NaCl. Conclusion: Water activity with high NaCl does not affect the pyrophosphate hydrolysis enthalpy change. Significance: Biological energy conversion is enthalpically driven within the physiological limits of NaCl. A decrease in water activity was thought to result in smaller enthalpy change values during PPi hydrolysis, indicating the importance of solvation for the reaction. However, the physiological significance of this phenomenon is unknown. Here, we combined biochemistry and calorimetry to solve this problem using NaCl, a physiologically occurring water activity-reducing reagent. The pyrophosphatase activities of extremely halophilic Haloarcula japonica, which can grow at ∼4 m NaCl, and non-halophilic Escherichia coli and Saccharomyces cerevisiae were maximal at 2.0 and 0.1 m NaCl, respectively. Thus, halophilic and non-halophilic pyrophosphatases exhibit distinct maximal activities at different NaCl concentration ranges. Upon calorimetry, the same exothermic enthalpy change of −35 kJ/mol was obtained for the halophile and non-halophiles at 1.5–4.0 and 0.1–2.0 m NaCl, respectively. These results show that solvation changes caused by up to 4.0 m NaCl (water activity of ∼0.84) do not affect the enthalpy change in PPi hydrolysis. It has been postulated that PPi is an ATP analog, having a so-called high energy phosphate bond, and that the hydrolysis of both compounds is enthalpically driven. Therefore, our results indicate that the hydrolysis of high energy phosphate compounds, which are responsible for biological energy conversion, is enthalpically driven within the physiological limits of NaCl.

Previous studies showed that a decrease in water activity leads to a smaller enthalpy in PP i hydrolysis (1)(2)(3)(4). Although these previous studies suggested the importance of solvation as the energy origin of high energy phosphate compounds, including PP i and ATP, its physiological significance remained unknown. This is because the previous results were obtained by theoretical calculations considering the gas phase with water activity to be zero or from experiments involving non-physiological solvents to reduce water activity, not by biologically relevant methods.
A variety of microbes adapt to natural environments containing up to 4 M NaCl. For example, extremely halophilic archaea require a nearly saturated level of NaCl for their growth. Such halophiles accumulate NaCl or KCl within their cells to a concentration equivalent to that in the external environment (5). An environmental decrease in the salt concentration causes a decrease in the cellular salt concentration in such archaea (6). Therefore, in extremely halophilic archaeal cells, water activity might change in accordance with the amount of salt present. We wondered whether or not naturally occurring physiological salt conditions affect the PP i hydrolysis enthalpy change value.
PP i is an ATP analog, being a so-called high energy phosphate compound, and the hydrolysis of both compounds is enthalpically driven (7). If the enthalpy change in PP i hydrolysis becomes smaller with increasing NaCl concentrations, analogous to the previous suggestion made for non-physiological conditions, the decrease in the energy derived from PP i and ATP would have deleterious effects on the growth of halophilic organisms. However, the effect of salt on PP i hydrolysis enthalpy is unknown.
The aim of this study was to determine the PP i hydrolysis enthalpy change values within the physiological limits of NaCl, which is one of the naturally occurring solutes that decrease water activity. First, we investigated the salt dependence of the biochemical pyrophosphatase (PPase) 3  NaCl dependence of these PPase activities using isothermal titration calorimetry (ITC). The data obtained were compared.

EXPERIMENTAL PROCEDURES
Microorganism, Medium, and Cultivation-H. japonica TR-1 was cultivated in 6 liters of CM medium (8) at 37°C for 4 days. The cells were harvested by centrifugation at 8000 ϫ g for 10 min and suspended in buffer containing 10 mM Tris-HCl (pH 8.0), 3.4 M NaCl, 140 mM KCl, and 10% (v/v) glycerol.
Preparation of a H. japonica Soluble Extract-The procedure for the H. japonica soluble extract preparation was carried out in the presence of 3.4 M NaCl to maintain high salt conditions. H. japonica cells were disrupted with a French pressure cell (Thermo Fisher Scientific) at 138 megapascals, followed by centrifugation at 12,000 ϫ g to obtain a cell-free extract. This extract was then ultracentrifuged at 200,000 ϫ g, and the supernatant was used as the soluble extract and was stored at Ϫ20°C. Protein concentrations were determined by the Lowry method (9) using bovine serum albumin as a standard.
Conventional Biochemical PPase Activity-The conventional biochemical PPase activity of the H. japonica extract (100 g of protein) was measured at 37°C in a mixture (0.6 ml) containing 10 mM Tris-HCl (pH 8.0), 0 -4.0 M NaCl, and 10 mM MgCl 2 . The reaction was initiated by the addition of 4 mM disodium PP i and terminated by the addition of 0.33 N trichloroacetic acid. PPase activity was calculated using a colorimetric assay (10) that determines the amount of P i liberated from PP i . Similarly, the purified PPases from the non-halophiles E. coli and S. cerevisiae (purchased from Sigma) were assayed at 25°C. The incubation time and the protein amount were adjusted so that the PPase activity was linear with time. One unit of PPase activity was defined as the amount of enzyme that liberated 1 mol of P i /min. Hydrolysis Enthalpy Measured by ITC-The calorimetric output during PP i hydrolysis by the H. japonica extract was measured with an isothermal titration calorimeter (VP-ITC, MicroCal, Northampton, MA). The resulting data were analyzed using Origin 7.0 software (OriginLab, Northampton, MA) with a MicroCal ITC data analysis add-on. All experiments were performed at 25°C. The reference cell was filled with distilled deionized water. The stirring speed used was 310 rpm, and the reference offset value was set at 20% full power. The volume of the ITC cell was 1.444 ml.
The calorimeter cell was filled with 10 mM Tris-HCl (pH 8.0 at 25°C), 0.1-4 M NaCl, 10 mM MgCl 2 , and H. japonica extract (0.5 mg of protein/ml). The injection syringe was filled with the same buffer without the extract but with 20 mM PP i . A single injection was carried out sequentially at least six times, with 60 nmol of PP i constantly introduced into the cell per injection and with a time interval between injections of 10 -120 min. This interval was sufficient for completion of the PP i hydrolysis reaction in each titration; thus, the calorimetric output returned to the base level in each case.
The same ITC measurements were carried out using the purified E. coli and S. cerevisiae PPases instead of the H. japonica extract. The time interval between injections for the enzyme assay was 5-60 min, which was sufficient for completion of the PP i hydrolysis reaction with these enzymes.
Determination of Water Activity-The water activities of the NaCl solutions of a range of concentrations used for calorimetric analysis were determined at 25°C using a Rotronic Hygro-Palm water activity machine (GSI Creos). This equipment was calibrated using a saturated NaCl solution of known water activity (0.753).

RESULTS AND DISCUSSION
Conventional Biochemical PPase Activity with NaCl-We first examined the NaCl dependence of the conventional PPase activity in a soluble extract of extremely halophilic H. japonica (8). This was then compared with those of PPases purified from the non-halophiles E. coli and S. cerevisiae.
The PPase activity of the halophilic H. japonica extract increased with NaCl concentration to reach a maximum at 2.0 M (Fig. 1), giving a specific activity of 1.39 units/mg. In the NaCl concentration range of 1.5-4.0 M, at least 40% of the maximal PPase activity value remained in the halophilic H. japonica extract, indicating that a cellular NaCl concentration change, if one occurs, does not have a severe effect on the PPase activity. This might be one of the adaptive mechanisms of this archaeon against hyposaline stress. A similar NaCl dependence was also observed for other enzymes from halophiles (11,12). In contrast, the PPases purified from non-halophilic E. coli and S. cerevisiae exhibited maximal activity at 0 -0.1 M NaCl (Fig.  1), with specific activities of 1120 units/mg for the E. coli enzyme and 438 units/mg for the S. cerevisiae enzyme. These non-halophilic activities were inhibited by higher NaCl concentrations, e.g. they were Ͻ10% of the maximal values with 1.5 M NaCl.
We are now able to examine both halophilic and non-halophilic PPases, which show distinct maximal activities at different NaCl concentration ranges. In particular, a halophilic source was useful for this study because a high salt concentration generally causes problems regarding the stability and structure of non-halophilic proteins; a halophilic source overcomes these problems.

ITC of PPases-The
PPase activities detected at the wide NaCl concentration range (0.1-4.0 M) examined were further verified by biophysical means using ITC. The traces in Fig. 2A show the calorimetric output due to the H. japonica PPase activity determined after the second of sequential injections of a constant amount of PP i at various NaCl concentrations. In these traces, just after the PP i injection (arrow), positive spikes appeared, representing the dilution heat of the PP i solution. After the positive spikes, negative displacement was immediately observed, representing an exothermic feature due to the enzymatic reaction.
The steepest negative displacement was observed at 2.0 -3.0 M NaCl ( Fig. 2A). In addition, the time interval until the negative displacement returned to the base level (dotted line) at 2.0 -3.0 M NaCl was shorter than those at lower and higher NaCl concentrations, indicating the faster reaction with 2.0 -3.0 M NaCl. It was obvious that the levels of negative displacement at all of the NaCl concentrations tested correlated well with the levels of PPase activity determined in the conventional biochemical assay shown in Fig. 1.
Similar to the case with the H. japonica extract, the calorimetric output due to the purified E. coli PPase determined after the second of sequential injections of a constant amount of PP i at various NaCl concentrations was measured (Fig. 2B). Again, the levels of negative displacement at all of the NaCl concentrations tested correlated well with the levels of PPase activity determined in the conventional biochemical assay shown in Fig. 1. Essentially similar results were obtained with the S. cerevisiae PPase (data not shown).
Evaluation of the PP i Hydrolysis Enthalpy Change Value-The calorimetric output ( Fig. 2A) was further analyzed to quantitatively compare the effects of NaCl on the hydrolysis enthalpy of PPase activity as shown in Fig. 3. The integrated area of the positive spike obtained upon PP i injection into the calorimeter cell without the H. japonica extract (Fig. 3A, upper) was deducted from the sum of the areas of the positive spike and the negative displacement driven by the PPase activity (solid trace) with regard to the base level (dotted line) (Fig. 3A, lower). The resulting calculated value was defined as the experimen-tally determined molar enthalpy change (⌬H Exp ) for the PP i hydrolysis reaction.
It has been postulated that the ⌬H Exp value can depend on the ionization heat of the buffer (⌬H Ion ) (13). Therefore, an enthalpy change value that does not depend on ⌬H Ion must be obtained to be quantitatively compared regardless of the reaction conditions used in the study. Such a ⌬H Ion -independent value, i.e. the whole catalytic enthalpy change denoted as ⌬H Whole , cannot be directly measured. To obtain ⌬H Whole , we measured ⌬H Exp as described above in three different buffers (Tris-HCl, HEPES-NaOH, and imidazole HCl) at pH 8.0, with different ⌬H Exp values obtained depending on the buffer used. At the same time, ⌬H Ion was measured using the same buffers (14). A HCl solution (4 mM) was injected sequentially six times, with 12 nmol of HCl constantly introduced into the solutions with or without buffer each time, and calorimetric displacement was exhibited (Fig. 3B). Deduction of the integrated area obtained without buffer (Fig. 3B, upper) from that with buffer (Fig. 3B, lower) gave ⌬H Ion .
Following the scheme described above, the correlation between the resulting ⌬H Exp and ⌬H Ion values was determined (Fig. 4). Linear regression of plots gave ⌬H Whole through extrapolation of the line to zero heat of buffer, as defined previously (15)(16)(17), with the following equation: ⌬H Exp ϭ n ⌬H Ion ϩ ⌬H Whole , where n represents the slope of the line (Fig. 4) giving the experimentally determined number of protons released due to the PP i hydrolysis reaction (and absorbed by the buffer). The same experiments and data analyses as those represented in Figs. 3 and 4 for the H. japonica extract were also carried out in the presence of various NaCl concentrations for the purified E. coli and S. cerevisiae PPases (data not shown).
NaCl Dependence of the PP i Hydrolysis Enthalpy Change Value and Water Activity-Through the studies described above, we finally determined the PP i hydrolysis enthalpy change (⌬H Whole ) values for the three enzyme sources in the presence of various concentrations of NaCl (Fig. 5). At 2.0 M  NaCl, which gave the maximal PPase activity for the halophilic H. japonica extract (Fig. 1), the exothermic ⌬H Whole value was Ϫ36.2 Ϯ 1.0 kJ/mol. Similarly, at 0.1 M NaCl, which gave the maximal non-halophilic PPase activities (Fig. 1), the exothermic ⌬H Whole values obtained were Ϫ34.6 Ϯ 0.3 kJ/mol for the E. coli enzyme and Ϫ33.7 Ϯ 1.7 kJ/mol for the S. cerevisiae enzyme. From the scatter of only these three experimental points, it was difficult to determine whether the values were significantly the same or not. The ⌬H Whole values obtained for the H. japonica extract at 1.5-4.0 M NaCl and those for the non-halophilic PPases at 0.1-2.0 M NaCl were plotted, and linear least-square fitting against all of the data points was carried out (Fig. 5).
As a result, the average ⌬H Whole value at 0.1-4.0 M NaCl obtained for the H. japonica extract and the purified PPases from E. coli and S. cerevisiae was calculated to be Ϫ35.4 Ϯ 1.5 kJ/mol, and this value did not correlate with the NaCl concentration with the yielded R 2 value of 0.31. Therefore, we concluded that the ⌬H Whole values at 0.1-4.0 M NaCl were significantly the same. This constant enthalpy change value is close to that experimentally determined in an aqueous solution (water activity of 1.0): Ϫ32 kJ/mol (1).
We also measured the water activities of the NaCl solutions used in the PPase activity assays by ITC. The solutions containing 0.1 and 4.0 M NaCl exhibited water activities of 1.00 and 0.84, respectively (Fig. 5); therefore, NaCl under these experimental conditions was regarded as a water activity-reducing solute. In contrast to the correlation between the ⌬H Whole values and NaCl concentrations, the water activity values were linearly correlated with the NaCl concentration with the yielded R 2 value of 0.97.
In this study, dealing with halophilic and non-halophilic enzyme activities, we have revealed the relationship between water activity and PP i hydrolysis enthalpy change under conditions in which life forms occur. Although the water activity is dependent on the physiological NaCl concentration, the enthalpy change in PP i hydrolysis is not. Therefore, changes in water activity and solvation within the physiological limits of NaCl do not affect the PP i hydrolysis enthalpy change.
Extended Views-Our results for PP i used as a substrate can be extended to ATP. Because PP i and ATP contain the same high energy phosphate bond and the hydrolysis of both compounds is enthalpically driven (7), their energy origins appear to be the same (18). Therefore, our results indicate that hydrolysis of high energy phosphate compounds, which are responsible for biological energy conversion, is enthalpically driven within  the physiological limits of NaCl. Extremely halophilic archaea such as H. japonica can thrive in high NaCl (up to 4 M) environments by accumulating salts inside their cells without thermodynamic problems.
The significance of the results of this study can also be extended in terms of environmental water activity. There is a microbe that can survive at 1.8 M MgCl 2 (water activity of ϳ0.8) (19). Although the salt concentrations inside such MgCl 2 -surviving cells are unknown, MgCl 2 -rich environments will be of interest because this salt causes further reduced water activity compared with NaCl (20,21). In the future, the microbiology of extreme halophiles in further lower water activity environments thus offers a chance for research on both the biochemistry and biophysics of biological energy metabolism. The findings and methods presented in this study will facilitate future studies.