ATP regeneration by thermostable ATP synthase

ATP Regeneration by Thermostable ATP Synthase

Ki Y. Nam,l Douglas K. Struck: and Mark T. Holtzapple’* Departments of ‘Chemical Engineering, and *Medical Biochemistry, Texas A&M University, College Station, Texas 77843

Received December 18, 1995/Accepted February 8,

We investigated the possibilityof using thermostable ATP synthase (TFoFl) for a new ATP regeneration method. TFoF1 was purified from a thermophilic bacterium, PS3, and reconstituted into liposomes. ATP synthesis experi- ments showed that TFoFl liposomes could synthesize ATP in micromole concentrations by acid-base change. The acid-base change was repeated six times over an 1 l-day period with no detectable loss of activity at the reaction temperature (45°C). Given these encouraging results, we conceptualized and modeled a system to synthesize ATP using ATP synthase with energy supplied by acid-base change. In this system, liposomes containing ATP synthase are immobilized on small glass spheres that facili- tate separation of buffers from the liposomes after the acid-base change. Compared to an alternate system that uses membranes to separate the buffers from the liposomes, the glass spheres reduce inefficient mixing of acidic and basic buffers during the acid-base change. To increase the ATP synthesis yield, this system uses electrodialysis to regenerate a potassium gradient after the acid-base change. It also employs water-splitting electrodialysis to regenerate KOH and HCI required to adjust the pH of acidic and basic buffers. All reagents are recycled, so electrical energy is the only required input. 0 1996 John Wiley & Sons, Inc. Key words: ATP regeneration ATPase ATP synthase electrodialysis

INTRODUCTION

Applications of enzymes to industrial processes have been limited to a few degradative reactions or simple transformations. Many enzymes that perform synthetic reactions require high-energy phosphate esters (e.g., ATP) to overcome unfavorable thermodynamics. These phosphate esters are expensive and cannot be used stoi- chiometrically for most reactions.

There have been many attempts to regenerate ATP from ADP generated as a waste product in synthetic reactions including chemical use of whole cells (dried yeast cells”), organelles (submitochondrial particle^:^ chloroplasts,’ and chromatophores”), and dephosphorylation of chemically synthesized, high- energy phosphate donors by k ina~es .~ However, each of these methods has serious flaws. The chemical synthesis method uses organic solvents that require additional separations before and after ATP synthesis, and it pro-

* To whom all correspondence should be addressed.

1996

duces unwanted by-products that must be removed. Dried yeast cells and organelles have half-lives that are too short to be useful. The phosphate donor-kinase method, which is the most attractive method currently available, uses unstable high-energy phosphate donors or produces undesired products that must be removed to reduce inhibition of the synthesis reaction.

In this research, we investigated the use of ATP synthase (FoF1 ATPase) to regenerate ATP from ADP. ATP synthase has been studied extensively in biochemistry because of its importance in oxidative phosphorylation and has been purified and reconstituted into liposome^.^.^,^' In a few cases, purified ATP synthase reconstituted into liposomes has synthesized ATP by acid-base changeI6 or by light-driven proton pumping.l0 However, its use in an ATP regeneration scheme has not been investigated at all, probably because of its reported instability and difficult purification.

In this study, we purified ATP synthase (TFoFl) from a thermophilic bacterium, PS3, according to the procedures of Sone et al.,23 because this enzyme is the most stable and has the highest activity among the purified ATP synthases reported in the literature. We tested the stability of reconstituted ATP synthase by repeating the acid-base change and storing the enzyme system at the reaction temperature (45°C).

MATERIALS AND METHODS

Complete details of the following procedures appear in the dissertation by Ki Nam.13

Materials

PS3 cells were kindly donated by Dr. Kagawa (Jichi Medical School, Japan). Sodium cholate was prepared as a 10% solution (pH 8.0) by dissolving cholic acid (purified three times by recrystallization as described by Kagawa and Racker’) and adjusting the pH to 8.0. Carrier-free [32P]phosphoric acid was obtained from ICN. Valinomycin, 9-aminoacridine orange (9-AA), sodium salts of ATP and ADP, hexokinase (H-4502), glycerokinase (G-1889), and lysozyme were purchased from Sigma. DNase 1 was purchased from Boehringer.

Biotechnology and Bioengineering, Vol. 51, Pp. 305-316 (1996) 0 1996 John Wiley & Sons, Inc. CCC 0006-3592/96/030305- 12

ATP Synthase Purification and Reconstitution into Liposomes Purification of ATP synthase and reconstitution into liposomes were performed using slightly modified procedures of Sone et al.23 The enzyme purification procedure involves growing PS3 cells at 70°C in a medium (pH 7.0) containing 0.8% Tryptone, 0.4% yeast extract, and 0.3% NaCl with vigorous aeration for 4-5 h until the optical density at 600 nm is about 2.1. Then cells were disrupted and membranes prepared by lysozyme treatment by agitating 50 g of cells in 450 mL of 50 mM Tris-sulfate (pH 8.0) containing 50 mg of lysozyme at 37°C for 30 min and then 2 h at the same temperature after adding 1 mg of DNase 1. Then the membranes were washed three times with 100 mL of 50 mM Tris- sulfate (pH 8.0), homogenized in 50 mM Tris-sulfate (pH 8.0), and then centrifuged at 17,OOOg for 20 min. Membranes were washed with a cholate solution, i.e., incubating the cell membranes in 0.5M Tris-sulfate (pH 8.0) containing 1% sodium cholate and 0.25M Na2S04 at a protein concentration of 9 mg/mL for 1 h and centrifuging at 98,OOOg for 40 min. ATP synthase was extracted from membranes with Triton X-100 by incubating membranes in 50 mM Tris-sulfate (pH 8.0) containing 2% Triton X-100, 0.2M Na2S04, and 0.2M sucrose at a protein concentration of 8 mg/mL for 1 h and then by centrifuging at 140,OOOg for 40 min. ATP synthase was purified from Triton X-100 extract (supernatant) by DE52 anion-exchange chromatography in a 5 X 10 cm column equilibrated with 50 mM Tris-sulfate (pH 8.0) containing 0.5% Triton X-100; loaded with Triton X-100 extract diluted six-fold with water; washed successively with equilibration buffer containing 50,65, and 75 mM Na2S0,; and eluted with equilibration buffer containing 0.2M Na2S04 at a flow rate of 100 mL/h. The last step was purifying by Sepharose 6B gel chromatography in a 1.7 X 75 cm column eluted with 50 mM Tris-sulfate (pH 8.0) containing 0.25% Triton X-100 and 0.2544 Na2S04 at a flow rate of 6 mL/h.

The procedure for reconstituting purified ATP synthase into liposomes includes suspending 50 mg of PS3 lipids (extracted by Folsh et al.3 method7 from the membranes prepared by the lysozyme treatment of PS3 cells as described in the enzyme purification) in 1 mL of 20 mM Tricine-NaOH (pH 8.0) containing 20 mg of sodium cholate, 10 mg of sodium deoxycholate, 5 pmol of dithiothreitol, and 0.2 pmol of EDTA. Then the lipids were solubilized by sonicating the lipid suspension for 15 min at 20% power using a microtip in a Heat Systems- Ultrasonics model W200R sonicator. The solubilized lipids were ultracentrifuged (sonicated lipid suspension) at 104,OOOg for 10 min. The mixture of solubilized lipids (0.4 mL, ultracentrifuged) and purified ATP synthase (1 mg) were dialyzed in dialysis membrane (molecular weight cut off [MWCO]: 12,000-14,000) for 20 h at 45°C against 250 mL of 10 mM Tricine-NaOH (pH 8.0) containing 2 mM MgS04, 0.2 mM EDTA, and 0.25 mM DTT.

.

ATP Synthesis by Acid-Base Change

ATP synthesis using ATP synthase was performed by the so-called acid-base change (see Fig. 1). First, TFoF1 liposomes were incubated in acidic buffer (pH 5.5) containing valinomycin to acidify the interior. Then, basic buffer (pH 8.5) containing KCl was added to produce a pH gradient across the liposomes membrane thus syn- thesizing ATP. Valinomycin “ports” in the membrane wall allowed K+ to flow into the liposome interior, thus maintaining charge balance and increasing ATP synthesis by facilitating proton flow through ATP synthase. To prevent destruction of the ATP product by reverse reactions of ATP synthase, the ATP was “captured” by reacting it with glucose or glycerol.

The detailed procedures follow. The TFoF1 liposomes suspension (20 pL) was added to 100 p L of 50 mM malonate (pH 5.5) containing 5 mM ADP and 0.1 pg valinomycin. After incubating the mixture for 10 min at 45”C, it was pipetted into 250 pL of 160 mM glygly (pH 8.5) that was at 45°C and contained 2 mM MgS04, 20 mM [32P]NaH2-P04 (1 pCi), 0.1M glucose, and 10 units of hexokinase. The resulting mixture was incubated at 45°C for 5 min, and then pipetted into 2 mL of 2% ammonium molybdate solution (in 1.6N HC104). Triethylamine (50 pL) was added, and the mixture was centrifuged at 4000g for 20 min. After extracting the supernatant twice with 2 mL of isobutanol-benzene (1:l by volume and saturated with water), 1 mL of the aque- ous phase was taken and counted in a Beckman liquid scintillation counter. In the ATP synthesis experiment using glycerokinase, the basic buffer contained 10 mM glycerol and 10 units of glycerokinase instead of 0.1M glucose and 10 units of hexokinase. All the other procedures were the same as when hexokinase was used.

9-AA Fluorescence Quenching Assay

The 9-AA fluorescence quenching assay measures the pH difference across a membrane and is widely used to measure ATP synthase activity. The TFoFI suspension

PH 5.5

K+

pH 8.5

Acidic Buffer Basic Buffer

Figure 1. ATP synthesis by acid-base change.

306 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 51, NO. 3, AUGUST 5, 1996

(20 pL) was incubated in 2 mL of 20 mM Tricine-NaOH (pH 8.0) containing 0.2 pg valinomycin, 0.375 mmol KCI, 4 pmol MgS04, and 8 nmol 9-AA at 45°C for 10 min. After adding 20 pL of 50 mM ATP, the resulting fluorescence decrease was measured in an SLM 8000 spectrofluorometer using an excitation wavelength of 365 nm and an emission wavelength of 451 nm.

0

Repeated Acid-Base Change

The mixture produced by an acid-base change was ultracentrifuged to precipitate the liposomes. Then, the recovered liposomes were resuspended in 50 p L of 10 mM Tricine-NaOH (pH 8.0) to allow for repeated acid-base changes. After the last acid-base change, the liposomes were resuspended in a small amount of 10 mM Tricine-NaOH (pH 8.0) and tested by the 9- AA fluorescence quenching assay. After the assay, the liposomes were discarded. The procedure included adding 50 pL of TFoF1 liposome suspension to a microcen- trifuge tube containing 200 pL of 50 mM malonate (pH 5.5) and 20 pL of a valinomycin solution (10 pg valinomycin/mL methanol) and incubating the tubes at 45°C for 10 min (the number of tubes depended on the number of acid-base changes to be repeated). Then 250 p L of 200 mM glygly (pH 8.5) were added that contained 25 mM NaH2S04, 2.5 mM MgS04, and 0.45M KC1 for an acid-base change. The liposome suspension produced by the acid-base change was incubated for 2 days at 45°C. The stored liposome suspension was ultracentrifuged at 200,OOOg for 20 min and the pellet was sus- pended in 50 p L of 10 mM Tricine-NaOH (pH 8.0). The liposome suspension was tested in one of the tubes by the fluorescence quenching assay after adding 2 mL of 20 mM Tricine-NaOH (pH 7.7), 40 p L of 0.1M MgS04, 40 pL of 0.1M 9-AA, 20 pL of a valinomycin solution (10 pg valinomycidml methanol), and 100 pL of 3.75M KC1 (the assay mixture is discarded after the assay). Subsequently, 200 pL of 50 mM malonate (pH 5.5) and 10 pL of a valinomycin solution (10 pg valinomycin/mL methanol) were added to the liposome suspension in the other tubes and the mixture was incubated for 10 min at 45°C. The steps were repeated until all the samples were tested.

z - I

W 3 -

g 80-

w 0 O ,O: z $ 40- v)

RESULTS AND DISCUSSION

i ATP Synthesis by Acid-Base Change

Table I shows that the ATP synthase activity was about 10 nmol ATP/mg TFz1 using hexokinase to capture the ATP. However, when glycerokinase was used to capture ATP as glycerol-3-phosphate, the synthetic activity increased five-to six-fold to 58-68 nmol ATP/mg TFoF1. These results are comparable to those of Sone et al.24 The increased synthetic activities obtained using

Table I. Effect of using glycerokinase in ATP synthesis.

Synthetic activity (nmol ATP/mg TF&)

Experiment Trial 1 Trial 2

Synthesis sample using Hexokinase 10.8 10.4 Glycerokinase 57.5 68.1

~~ ~ _ _ _ _ _ _ _ _

Synthetic activity was calculated using the samples without ADP in the acidic buffer as the blanks.

glycerokinase may have resulted from efficient ATP capturing by glycerokinase because the Michaelis constant for glycerokinase is 9.0 X mol ATPL compared to 2.0 X mol ATP/L for hexokinase.*

Repeated Acid-Base Change

Figure 2 shows that ATP synthase activity, as measured by fluorescence quenching, did not decrease even after six acid-base changes over an 11-day period. During this period, the liposomes were stored at the reaction temperature (45"C), thus showing that the reconstituted ATP synthase is thermally stable. Encouraged by these results, we decided to model an ATP regeneration system that uses ATP synthase.

MODELING

As shown in Figure 3, a simple way to regenerate ATP using ATP synthase would be to first ultrafilter the liposome suspension after the acid-base change, and then adjust the filtrate pH to 8.5 and the filtered suspension pH to 5.5. Unfortunately, this approach is slow, and most of the driving force (i.e., pH gradient) provided by the acidic and basic buffers is not used for ATP synthesis, but is wasted by mixing during the acid-base change. Therefore, we propose a system in which liposomes are immobilized on glass spheres that are packed

W g 2 0 4

307 NAM, STRUCK, AND HOLTZAPPLE: ATP REGENERATION BY THERMOSTABLE ATP SYNTHASE

ATP synthesis by acid-base change

20 mhl mal 80 glygly

o oo pH8.23 O 0 0 1.OL

J. Ultrafitration

0.5 L 0 5 L

4ddHCI A d d H c ~ rKoH AddK01

1 MHCl

bp

1 ATP synthesis by acid-base change

O 0 0 1.OL

1 Ultrafitration

I MKOH

A Water-Sphtting Electrodialysis

a c bp

I I

8 0 --

Figure 3. ATP synthesis by a liposome suspension scheme.

0 1 1 0 %H+I I&

I I O Q +CI-'OH$ --

0 l o

in a reactor (see Fig. 4). By completely draining the column, acid-base mixing is avoided. The proposed system uses electrodialysis to regenerate the K+ gradient after the acid-base change. Also, water-splitting electrodialysis produces KOH and HCl that are required to adjust the pH of the acidic and basic buffers after the acid-base change. In this system, all reagents are recycled; therefore, electricity is the only input.

Liposomes can be immobilized on glass spheres in several ways including dinitrophenyl (DNP) group- antibody interaction as described below2':

fluorodinitrobenzene + lipid (phosphatidylethanolamine, RNH2) -+ dinitrophenyl

lipid + HF, dinitrophenyl lipid + PS3 lipid -+ liposomes with

DNP groups, liposomes with DNP groups + glass spheres with immobilized DNP antibody -+ glass spheres with

immobilized liposomes. Monoclonal antibody to the DNP group is available from Sigma and can be attached to glass spheres using a CNBr technique? Alternatively, liposomes may be attached di- rectly to glass spheres by the CNBr technique4 because PS3 lipids contain phosphatidylethanolamine (FWH2).

As shown in Figure 4, ATP synthesis by acid-base change is performed in the reactor as follows. Glass spheres with immobilized ATP synthase liposomes are first packed in the reactor, and then acidic buffer is

n20

1-1 p Storage Storage

I HzO

ATP Synthase Valinomycin

P,-+ATP I ADP +

Acid

I Electrodialysis (

Desalted Acidic Buffer

1 M KCI

Electrodialysis Unit # i + +

I 0.5 M KCI

Basic Buffer with Purge 0.2 M KCI

Figure 4. ATP synthesis by an immobilized Iiposome scheme.

added. After allowing about 10 min to equilibrate the liposomes with acidic buffer, the acidic buffer is drained from the reactor and collected for the next acid-base change. It should be noted that, because of capillary action, some of the acidic buffer is retained between the glass spheres. Basic buffer containing ADP, phosphate, and KCl is then added to synthesize ATP by acid-base change. The ATP may be used in a chemical reaction, for example, the phosphorylation of a sugar. After the acid-base change, buffer is drained and collected. The cycle is repeated by adding acidic buffer.

Through the buffer retained in the capillary spaces, KCl is transferred from the basic buffer to the acidic buffer, thus decreasing the K+ gradient across the liposome wall. Therefore, it is necessary to regenerate the K+ gradient for the next acid-base change. This is ac- complished by passing the drained acidic and basic buffers through electrodialysis unit 1 (Fig. 4). An electrodialysis unit consists of an alternating series of cation-permeable and anion-permeable membranes that are separated by flow distribution gaskets and bounded by an anode compartment on one end and by a cathode compartment on the other end.I4 During electrodialysis, K+ ions in the acid buffer move toward the cathode and pass through the adjacent cation- permeable membranes, but are blocked by the next anion-permeable membranes. Similarly, C1- ions in the acidic buffer move toward the anode and pass through


the adjacent anion-permeable membranes but are blocked by the next cation-permeable membranes. However, K+ and C1- ions in the basic buffer do not move because they are faced with anion-permeable and cation-permeable membranes in their moving direction, respectively. As a result, the acidic buffer is depleted of KC1 while the basic buffer is enriched with KCl, thus restoring the K+ gradient.

After restoring the K' gradient, the acidic buffer and basic buffer are sent to storage tanks for pH adjustment. To maintain the acidic buffer at pH 5.5, HCI is added. To maintain the basic buffer at pH 8.5, KOH is added that reacts with HC1 to form KC1. If the KC1 concentrations of the basic buffer becomes higher than 0.2M, it is passed through electrodialysis unit 2 before pH adjustment (Fig. 4). This step is necessary because the ATP synthase liposomes are unstable at KC1 concentrations higher than 0.2M.24 Electrodialysis unit 2 is oper- ated so that the outlet KCI concentration of the concentrated stream is about 1M because the next step (regeneration of HC1 and KOH from KC1 by water- splitting electrodialysis) works economically only when the inlet salt concentration is relatively high."

KCI removed from the basic buffer by electrodialysis unit 2 (Fig. 4) is regenerated into HCI (about 1M) and KOH (about 1M) by water-splitting electrodialysis. Water-splitting electrodialysis is similar to conventional desalting electrodialysis, but is has additional bipolar membranes that split water molecules into HC and OH- ions under the influence of an electrical potential (Fig. 5) . (This is an energy-efficient technology. For example, it requires half the energy of the conventional electro- lytic process for producing caustic soda and chlorine from brine.'*) The HCI and KOH produced by water- splitting electrodialysis are used to adjust the pH of the acidic and basic buffers, respectively, for the next acid-base change.

Because it is important to know the energy needed to synthesize ATP, we calculated the ATP synthesis yield, defined as the amount of ATP synthesized per electrical energy input, for the proposed system.

To calculate the ATP synthesis yield, we assumed that the energy required to synthesize ATP by acid-base change consists of three terms: 1. energy required to regenerate KOH and HCI by

water-splitting electrodialysis; 2. energy required to regenerate the KCl gradient by

electrodialysis unit 1; and 3. energy required to remove KC1 from the basic buffer

to the concentrated KCI solution by electrodialysis unit 2 (Fig. 4).

The ATP sythesis yield was determined by sequentially calculating the following quantities.

f bD I a

Holdup Volume Following liposome acidification, acidic buffer is retained in the reactor after drainage. Similarly, after the

f C I bv

II 0 I I II II 0

f I I I

Concentrated KCI Solution

Legend bp bipolar membrane

c cation-permeable membrane

a anion-permeable membrane

Bipolar Membrane Construction

Figure 5. Electrodialysis water splitting system.

acid-base change, basic buffer is retained after drainage. The amount of buffer retained in the reactor (holdup volume) was calculated using the method of Dombrowski and Brownells

Vh = S0V, (1) where Vh is the holdup volume, V, is the void volume, and So (the residual saturation) is the fraction of voids filled with wetting fluid after drainage by gravity. So is a function of the capillary number defined as pK/y cos 8 where K is permeability, y is surface tension, p is density, and 8 is the contact angle. Dombrowski and Brownell' measured permeability and void volume (i.e., void volume/total bed volume) for glass spheres of diameters ranging from 0.0096 to 0.213 mm, and also determined the relationship between So and capillary number. Therefore, the holdup volume of a reactor can be easily calculated given the diameter of the glass spheres packing the reactor.

Amount of ATP Synthase Immobilized

For the system in which the liposomes are immobilized on glass spheres (Fig. 4), the number of liposomes per milliliter, NL, was calculated assuming the glass spheres are completely covered with liposomes

NAM, STRUCK, AND HOLTZAPPLE: ATP REGENERATION BY THERMOSTABLE ATP SYNTHASE 309

N l = (*)(?) 377 g

where E is the void fraction of the packed reactor, rg is the glass-sphere radius, and rl is the liposomes radius.

For the system in which the liposomes are in free suspension (Fig. 3), the following calculation describes the amount of ATP synthase in a given volume. Sone et aLZ5 reported that the liposome radius is 0.05 pm and the liposome volume is about 1-2% of the suspension volume. Assuming the liposome volume is 1.5%,

0.015 N~ = = 2.86 x 10131m~.

3 4 (3)

Because 50 mg of lipids and 1 mg of ATP synthase are generally used to prepare 1 mL of a liposome suspension, it can be estimated that 2.86 X lOI3 0.05-~m liposomes correspond to 1 mg of ATP synthase. Therefore, the concentration of ATP synthase, Csynthase, is

). (4) 1 rng ATP synthase

Csynthase = Nl ( 2.86 X l O I 3 liposomes

Buffer Mixture pH Produced by Liposome Acidification

The pH of the buffer mixture produced after liposome acidification was calculated by solving Equation ( 5 ) nu- merically (see Appendix for its derivation).

( 5 )

Here the added acidic buffer was designated solution 1 and the retained alkaline buffer (from the previous acid-base change) was designated solution 2.

Buffer Mixture pH Produced by Acid-Base Change

The pH of the buffer mixture produced after the acid- base change was also calculated using Equation (5). The added basic buffer was designated solution 2 and the retained buffer (from the previous liposome acidification) was designated solution 1.

Amount of ATP Synthesized Per Acid-Base Change

The amount of ATP synthesized during the acid-base change depends on the initial proton-motive force, ApHi ,

which is determined by both the K+ and H+ gradients.

- - - Zlog- IK+Iin - Z(pHOut - pHin) W + l o u t

= A q i - 2-ApHi. (6)

The relationship between the initial proton-motive force and the amount of ATP synthesized per acid-base change was derived from the data reported by Sone et aLZ4 (see Fig. 6).

Energy Required to Regenerate K+ Gradient by Electrodialysis

The energy required to regenerate the K+ gradient after the acid-base change was calculated using Spiegler's equation."

U = 5.21 X AN - - [Fl "u] a - 1 ' (7)

where AN is Cac,in - Cac,out; a is Cac,inlCac,out; P is Cac,inl Cbs,out; C is potassium concentration in normality; and subscripts ac, bs, in, and out indicate acidic buffer, basic buffer, inlet, and outlet, respectively.

Energy Required to Remove KCI Exceeding 0.2M from Basic Buffer by Electrodialysis

The energy required to remove KCI from the basic buffer when its concentration exceeds 0.2M is neglected because this step is necessary only once for about 20 acid-base changes.

Energy Required to Regenerate KOH and HCI from Recovered KCL by Water-Splitting Electrodialysis

The amounts of 1M KOH and 1M HC1 required to adjust the acid buffer pH to 5.5 and the basic buffer pH to 8.5 were calculated by integrating the buffer capac- ity eq~a t ion , '~

The energy required to regenerate the KOH and HC1 was calculated assuming that 1180 kW h of electricity (DC) is required to produce 1 metric ton of KOH (and a corresponding amount of HCl) by water-splitting electrodialysis."

A FORTRAN program was prepared considering all the steps described above. Using this program, ATP


01 I I I I 0 I00 200 300 400

PROTON-MOTIVE FORCE (mV)

Figure 6. ATP synthesis dependence on proton-motive force. For the Kin curve, the proton-motive force was changed by varying the K' concentration inside the liposomes; for the KO, curve, the K' concentration outside the liposomes was varied.

synthesis yields were calculated for different glass sphere diameters and K+ gradients.

Holdup Volume and Amount of ATP Synthase Immobilized Depending on Glass Sphere Diameter

As shown in Table 11, the reactor holdup volume did not change much with different glass sphere diameters. However, the amount of ATP synthase immobilized per milliliter depended on the size of glass spheres; it was proportional to the inverse of the glass sphere diameter.

Composition and pH Change During Repeated Acid-Base Change

When the synthesis reaction is started using an acidic buffer of pH 5, 40 mM malonate and a basic buffer of pH 8.5,160 mM glygly, the compositions of both buffers change as the acid-base change is repeated, giving steady-state values of 20 mM malonate and 80 mM glygly for both buffers after 30 cycles (Fig. 7). The changes in potassium concentration and pH of the acidic and basic buffers are shown in Figure 8.

ATP Synthesis Yield Depending on K' Gradients

The ATP synthesis yield depends on the K+ gradient available during the acid-base change (Fig. 9). Higher

Table II. Holdup volume and immobilized ATP synthase content.

K' gradients are obtained by removing KCl more thor- oughly from the acidic buffer to the basic buffer by electrodialysis (unit 1 in Fig. 4). When the acidic buffer is desalted to 40 mM KC1, the ATP synthesis yield is about 140 g ATP/kW h with 0.016-mm diameter glass spheres. The yield increases to 190 g ATP/kW h if the acidic buffer is desalted to 10 mM KC1 with glass spheres of the same diameter. When the acidic buffer is lowered to 2 mM KC1 by electrodialysis, the yield increases up to 350 g ATP/kW h. In practice, this low KC1 concentration may be difficult to achieve because buffer ions (malonate anions and glygly cations) can also migrate with the potassium and chloride ions, thus depleting ions from the acidic buffer. However, desalting the acidic buffer to 10 mM KC1 appears to be reasonable because K' and C1- ions migrate much faster than the big buffer ions (malonate anions and glygly cations) at the same concentrations. The yield drops to 60 g ATP/kW h (with 0.016-mm diameter glass spheres) if the synthesis reaction is driven only by a pH gradient (Fig. 9).

The initial decrease in ATP synthesis yield when the acidic buffer is desalted to 40 mM KC1 (see Fig. 9) occurs because of the diminishing K' gradient; electrodialysis unit 1 is not used until the KC1 concentration in the acidic buffer reaches about 40 mM KC1 (after six acid-base changes). On the other hand, there is no initial yield decrease when the acidic buffer is desalted

~~ ~

Glass sphere Holdup volume/reactor Liposome concentration ATP synthase diameter (mm) Capillary number volume (1iposomeslmL) content (mglmL)

0.0096 0.0135 0.0161 0.0226 0.0382 0.0455 0.0764

~~~

0.49 X 0.11 x 10-5 0.28 x 10-5 0.52 x 10-5 0.10 x 10-4

0.54 x 1 0 - 4 0.17 X

0.0028 0.0028 0.0028 0.0027 0.0028 0.0028 0.0026

0.5 x 10'4 0.36 x 1014

0.30 x 1014 0.22 x 1014 0.13 x 1014 0.11 x 1014 0.06 x 1014

1.73 1.25 1.04 0.76 0.44 0.37 0.23

NAM, STRUCK, AND HOLTZAPPLE: ATP REGENERATION BY THERMOSTABLE ATP SYNTHASE 31 1

- BASIC BUFFER

ACIDIC BUFFER ----

0 V

0.05- LA LL 2 m

_ _ _ ___------ - GLYGLY ___- - - - * -

--- , / MALONATE 7- - - - - - - - - _ _ - -

o < , I

I I I

NUMBER OF ACID-BASE CHANGES

Figure 7. Change of buffer compositions during acid-base change for the immobilized system. Acidic buffer is desalted to 40 mM KCl.

BASIC BUFFER - - - - - - - - - - - - - - - - - - - - - - - - - - - -

5

z 0 - KCI CONCENTRATION

- $ 0.10- n z w V

I- - - -pH

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -----------

ACIDIC BUFFER

0' I I I

9

- 8

- 7

6 -

- PH

5

TO 10mM KCI

TO 4 0 m M KCI

"T--- r I I I

0 10 20 30 40 NUMBER OF ACID-BASE CHANGES

Figure 9. Yield of ATF' synthesis depending on K+ gradient. Glass sphere diameter is 0.016 mm.


to 10 mM KCl because the KCl concentration is already higher than 10 mM after the first acid-base change and electrodialysis unit 1 must operate from the beginning.

A TP Synthesis Yield Depending on Glass Sphere Diameter

Higher ATP synthesis yields are obtained with smaller glass spheres as shown in Figure 10. The yield is roughly proportional to the inverse of the glass sphere diameter because the glass sphere surface area per unit volume increases inversely with diameter. When the acidic buffer is desalted to 40 mM KCl, the synthesis yields are 30, 130, and 220 g ATP/kW h for glass spheres of 0.076, 0.016, and 0.009 mm diameter, respectively.

ATP Synthesis Yield of Immobilized System Compared to Suspension System

ATP synthesis yields with the immobilized system are much higher than the suspension system as shown in Table I11 (about five times more at the same liposome concentration). The ATP synthesis yield with liposomes immobilized on 0.076-mm diameter glass spheres is 15 g ATP/kW h at a protein content of 0.22 mg/mL and without a K' gradient; the yield with the suspension system is 2.72 g ATP/kW h at a protein content of 0.2 mg/mL and without a K+ gradient. The higher yields with the immobilized system are due primarily to less mixing of the acidic and basic buffers.

CONCLUSIONS

ATP synthase can synthesize ATP by repeated acid- base changes and the enzyme/liposome system is thermally stable. Given these encouraging results, a system is proposed that regenerates ATP using ATP synthase embedded in liposomes that are immobilized on glass spheres and subjected to acid-base change. This system

2 2 5 0 I-

allows the acid-base change to be repeated much faster than the suspension approach that uses ultrafiltration to separate liposomes after the acid-base change. In addition, there is less acidbase mixing using the immo- bilization system.

The proposed system uses electrodialysis to regenerate the K' gradient after the acid-base change, which increases ATP synthesis yield significantly compared to ATP synthesis driven only by a pH gradient. It also uses water-splitting electrodialysis to regenerate KOH and HCl from the KC1, which is inevitably produced from the reaction of KOH and HC1.

The proposed system is potentially economical because the calculations show that about 130 g of ATP could be synthesized per kW h of electricity (DC). Assuming that electricity can be purchased for $0.07/kW h, 1 kg ATP could be regenerated using $0.54 worth of electricity. This compares very favorably with the purchase price of $370/kg of ATP. Of course, other costs (e.g., captial, labor, enzyme replacement) must be determined before the final cost can be determined.

APPENDIX

There is no equation currently available to determine the pH of a mixture produced by mixing two buffer solutions of different pH; therefore, it is necessary to derive one. First, assume that solution 1 contains buffer a, a weak acid, with concentration of Cla, and buffer b, a weak base, with concentration of Clb, and solution 2 contains buffer a with concentration C2, and buffer b with concentration C2,,. The volume of solution 1 is x and its pH has been adjusted to pH1 with a strong base (NaOH or KOH) or with a strong acid (HC1 or H2S04). The volume of solution 2 is y and its pH has been adjusted to pH2 in the same way as solution 1. Buffers a and b satisfy the following equilibrium relationships:

DIAMETER

(mm)

0.076

0 !z

0 10 20 30 40 N U M B E R OF ACID-BASE C H A N G E S

Figure 10. Yield of ATP synthesis depending on glass sphere diameter. Acidic buffer is desalted to 40 mM KCl.


Table III. Comparison of ATP synthesis yields for immobilized and suspension systems.

Suspension system, 0.2 mg/mL protein

Filtration ATP synthesized KOH required Water-splitting Yield extent (mg) (nmol) energy (W h) (g ATFVkW h)

0.7 0.62 13.25 0.875 0.71 0.9 1.02 5.68 0.375 2.72

Desalting ATP synthesized Desalting energy Water-splitting Yield condition (mg) (Wh) energy (W h) (g ATPlkW h)

Immobilized systems

0.076-mm dia. glass spheres, 0.226 mg/mL protein

wlo KCl 1.36 0.0 0.092 15 40 mM KCI 3.00 0.0026 0.092 32 10 mM KCl 4.38 0.0078 0.092 44

0.016-mm dia. glass spheres, 1 mg/mL protein

wlo KCl 6.2 0.0 0.099 63 40 mM KCl 13.8 0.0027 0.099 136 10 mM KCl 20.0 0.0083 0.099 187

Stewart's26 SID (strong ion difference) concept may be used to derive various equations for buffers. SID is defined as follows:

[SID] = x [(strong cation concentration)

- Z [(strong anion concentration) X (strong cation charge)]

X (strong anion charge)],

where Na+ and K+ are examples of strong cations and C1- and SO;- are examples of strong anions.

Solution 1 satisfies the following relationships:

water dissociation: [H+]1 X [OH-], = K, (A.3)

weak acid equilibrium: [H+]1 X [A-ll = Ka[A]1 (A.4)

weak acid conservation: [A11 + [A-]l = CI, (AS)

weak base equilibrium: [B]] X [H']1 = Kb [BH+]l (A.6)

weak base conservation: [B]1 + [BH+]1 = Clb (A.7)

electrical neutrality: [SID]1 + [H+]1 + [BH+]l - [A-ll - [OH-Il = 0. (A.8)

From Equation (AS),

[A11 = C1a - [A-ll-

[H+]I X [A-], = Ka (C1a - [A-II).

(A.9)

(A.10)

Substituting this expression into Equation (A.4) gives

Rearranging this equation,

(A.ll)

From Equation (A.7),

[B]l = Clb - [BH+]I. (A.12)

Substituting this expression into Equation (A.6) gives

(Clb - [BH+]l) X [H']1 = Kb [BH+]l. (A.13)

Rearranging this equation,

From Equation (A.l),

[OH-Il = [H+11'

(A.14)

(A.15)

Substituting Equations (A.ll), (A.14) and (A.15) into Equation (A.8),

Kw - 0. (A.16) [H+11

Because [H+],, Clar and Clb are known, [SIDIl can be calculated by Equation (A.16).

+- Kw (A.17) [H+11'

An analogous equation for solution 2 may be obtained by applying Equations (A.3)-(A.15) to solution 2,


+- Kw (A.18) [H+12'

When volume x of solution 1 is mixed with volume y of solution 2, the resulting mixture will contain buffers a and b at concentrations of C,, and c m b .

(A.19)

(A.20)

Because strong ions are conserved during mixing, the strong ion difference for the mixture ([SID],) is

For the mixture, an equation analogous to Equations (A.17) and (A.18) may be derived.

+- Kw . (A.22) [H'lm

Because [SID],, C,,, and Cmb are known, [H'],,, can be calculated by solving Equation (A.22) using the New- ton-Raphson method.

This work was supported by Grant BCS-8909333 from the National Science Foundation. We express our thanks to Dr. Kagawa of Jichi Medical School in Japan for.donating the thermophilic bacterium, PS3.

NOMENCLATURE

concentration of buffer a, M concentration of anion in buffer a, M concentration of buffer b, M concentration of cation in buffer b, M total concentration of buffer a, M total concentration of buffer b, M concentration of TFoFl synthase, mg/mL Faraday's constant, 96,485 Umol proton ion concentration, M permeability, cm3/s2 equilibrium constant of buffer a, M equilibrium constant of buffer b, M water ion product, M Z potassium ion concentration inside the liposomes, M potassium ion concentration outside the liposomes, M number of liposomes per unit volume, mL-' concentration of OH-, M pH inside the liposomes pH outside the liposomes gas constant, 8.314 J/(mol . K) glass sphere radius, cm liposome radius, cm amount of strong acid or base required to adjust pH of buffer a, M

s o

[SID] strong ion difference, M T absolute temperature, K U V, holdup volume, mL V V void volume, mL 2 2.303 RT/F, a constant to convert pH to millivolts,

62.1 mV at 40°C

residual saturation, fraction of voids filled with wetting fluid

energy required for electrodialysis, kW h/3785 L

Greek letters

Cac,inlCac,out Cac,inGs,out porosity, volume of void spacekotal volume of bed surface tension, dydcm proton-motive force, mV (i.e., mJ/C) proton-motive force, J/mol membrane potential, mV (i.e., mJ/C) density, g/cm' contact angle

Subscripts

1 solution 1 2 solution 2 m mixture

References

1. Avron, M. 1960. Photophosphorylation by swiss-chard chloroplasts. Biochim. Biophys. Acta 40: 257-272.

2. Bergmeyer, H. U. 1974. Methods of enzymatic analysis, 3rd edition. Academic Press, New York.

3. Crans, D. C., Kazlauskas, R. J., Hirschbein, B. L., Wong, C. H., Abril, O., Whitesides, G. M. 1987. Enzymatic regeneration of ATP. Methods Enzymol. 136: 263-280.

4. Dean, P. D. G., Johnson, W. S., Middle, F. A. 1985. Affinity chromatography. IRL Press, Oxford.

5. Dombrowski, H. S., Brownell, L. E. 1954. Residual equilibrium saturation of porous media. Ind. Eng. Chem. 46: 1207-1219.

6. Fillingame, R. H., Foster, D. H. 1986. Purification of FIFO H+ -ATPase from Escherichia coli. Methods Enzymol. 126: 545-557.

7. Folsh, J., Lees, M., Sloan-Stanley, G. H. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226: 497-509.

8. Friedel, P., Schairer, H. U. 1986. Preparation and reconstitution of F,Fo and F, from Escherichia coli. Methods Enzymol. 126:

9. Kagawa, Y., Racker, E. 1971. Partial resolution of the enzymes catalyzing oxidative phosphorylation. J. Biol. Chem. 246: 5477- 5487.

10. Krupinski, J., Hammes, G. 1986. Steady-state ATP synthesis bac- teriorhodopsin and chloroplast coupling factor co-reconstituted into asolectin vesicles. Proc. Natl. Acad. Sci. USA 83: 4233-4237.

11. Mani, K. N. 1991. Electrodialysis water splitting technology. Allied-Signal Inc., Warren, NJ.

12. Nagasubramanian, K., Chlanda, F. P., Liu, K. J. 1977. Use of bipolar membranes for generation of acid and base-An engineering and economic analysis. J. Membr. Sci. 2 109-124.

13. Nam, K. 1992. The use of H'-ATP synthase for regenerating ATP in synthetic enzymatic reactions. Ph.D. thesis, Texas A&M University, College Station, TX.

14. Oshen, S. 1991. HPD Incorporated technology update. Topic: Electrodialysis systems for water purification and reclamation. HPD Incorporated, Naperville, IL.

579-588.


15. Perrin, D. D., Dempsey, B. 1974. Buffers for pH and metal ion control, 2nd edition. Chapman & Hall, London.

16. Pick, U., Racker, E. 1979. Purification and reconstitution of the N,N’-dicyclohexylcarbodiimide sensitive ATPase complex from spinach chloroplasts. J. Biol. Chem. 254 2793-2799.

17. Pick, U. 1986. Isolation and reconstitution of CF,F1 from chloroplasts. Methods Enzymol. 126: 512-520.

18. Sakata, K., Kitano, H., Ise, N. 1981. Continuous ATP production by enzyme or yeast immobilized in amphoteric gel. J. Appl. Bio- chem. 3 518-525.

19. Shaffer, L. H., Mintz, M. S. 1980. Electrodialysis, pp. 257-267. In: K. S. Spiegler and A. D. K. Laird (eds.), Principles of desalination. Academic Press, New York.

20. Sigma Chemical Company. 1989. Biochemicals and organic com- pounds for research and diagnostic reagents. Sigma Chemical Company, St. Louis.

21. Smeds, A. L., Enfors, S. 0. 1986. Stability of chromatophores during ATP regeneration in an enzyme reactor. Enzyme Microb. Techno]. 8: 755-760.

22. Smith, M., Korana, H. G. 1957. Nucleoside polyphosphates, VI. An improved and general method for the synthesis of ribo- and

deoxyribonucleoside 5’-triphosphate. J. Am. Chem. SOC. 80: 1141- 1145.

23. Sone, N., Yoshida, M., Hirata, H., Kagawa, Y. 1975. Purification and properties of a dicyclohexylcarbodiimide-sensitive adenosine triphosphatase from a thermophilic bacterium. J. Biol. Chem.

24. Sone, N., Yoshida, M., Hirata, H., Kagawa, Y. 1977. Adenosine triphosphate synthesis by electrochemical proton gradient in vesicles reconstituted from purified adenosine triphosphatase and phospholipids of thermophilic bacterium. J. Biol. Chem. 252

25. Sone, N., Yoshida, M., Hirata, H., Kagawa, Y. 1977. Reconstitu- tion of vesicles capable of energy transformation from phospholipids and adenosine triphosphatase of a thermophilic bacterium. J. Biochem. 81 519-528.

26. Stewart, P. A. 1981. How to understand acid-base. Elsevier, New York.

27. Thayer, W. S., Hinkle, P. C. 1975. Synthesis of adenosine triphosphate by an artifically imposed electrochemical proton gradient in bovine heart submitochondrial particles. J. Biol. Chem. 250:

250 7917-7923.

2956-2960.

5330-5335.


Documents

ATP regeneration by thermostable ATP synthase