Eur. J . Biochem. 102, 251 -256 (1979)
Characterization of the Membrane-Bound Inorganic Pyrophosphatase in Rh odosp irillum rubrum Hikan R A N D A H L University of Stockholm, Arrhenius Laboratory, Department of Biochemistry (Received June 27, 1979)
The membrane-bound inorganic pyrophosphatase (EC 3.6.1 .l) from Rhodospirillum rubrum has been investigated with the tools of enzyme kinetics, and with two amino acid reagents, N-ethylmaleimide (MalNET) and 4-chloro-7-nitrobenzofurazan(Nbf-Cl). 1. The concentration of the true substrate, MgPPi, was varied with constant concentrations of free Mg2+ or PPi. It was observed that Mg2+ acted as an activator. 2. Heat inactivation of the enzyme at 62°C was slowed down in the presence of Mg2+. 3. MalNET and Nbf-C1 bind to the enzyme, and inhibit its activity. The effect of both reagents is dependent on the temperature. 4. A model is proposed where the 1 : 1 complex of Mg2+:PPi acts as substrate and Mg2+ interacts directly with the enzyme as an activator. PPi can bind to the enzyme, but is not hydrolyzed in the uncomplexed form. The cytoplasmic pyrophosphatase has been studied
in several different organisms. In Rhodospirillum rubrum Klemme and Gest [1,2] have made a very comprehensive study of this enzyme, which accounts for 80% of the total PPi hydrolysis activity in crude extracts. The membrane-bound pyrophosphatase coupled to the electron transport chain in chromatophores has been found only in certain organisms. This enzyme is unique in that it catalyzes not only the hydrolysis of PPi, but also the synthesis of PPi in the light [3]. The synthasis of PPi is believed to be in equilibrium with its hydrolysis [3,4], although PPi hydrolysis is significantly depressed in the light. It was suggested [4] that this depression results from an alteration in the oxidation/reduction state of the electron transport system. PPi has been shown to act as an energy donor both at the cytochrome level and coupled to electron transport chain [ 5 ] .Most inorganic pyrophosphatases require divalent cations for activity and this is also true for membrane-bound pyrophosphatases. But in contrast to the cytoplasmic pyrophosphatases, which require Zn2+ both for stability and for activity and Mg2+ to complex the substrate [l], membrane-bound -~
Ahhreviurions. MalNEt, N-ethylmaleimide; Nbf-CI, 4-chloro7-nitro benzofurazan. Enzymes. Membrane-bound inorganic pyrophosphatase (EC 3.6.1 , I ) ; cytoplasmic inorganic pyrophosphatase (EC 3.6.1.1).
pyrophosphatases require only Mg2 +. The membranebound enzyme has very poor activity with any other divalent cation: about 50% activity is obtained with Z n 2 + and M n 2 + and none with Ca2+ (unpublished results from this laboratory). However, Cooperman and Mark [6] have shown that Mn 2 + and PPi aggregate in dilute solution, which could account for the lower activity with this cation. This paper presents an analysis of the membranebound pyrophosphatase kinetics, its dependence on Mg2+ and the role of PPi.
MATERIALS AND METHODS Pyrophosphate was purchased from Merck (Darmstadt, F.R.G.). N-Ethylmaleimide (MalNEt) and reduced glutathione were obtained from Sigma Chem. Co. (St Louis, Mo., U.S.A.) and Nbf-CI from Serva FeinbiochemicaGmbHundCo.(Heidelberg1 ,F.R.G.). Preparation Techniques Rhodospirillum rubrum strain S1 was grown anaerobically in light at 30 "C in the medium described by Bose et al. [7]. The bacteria were harvested in the exponential phase, after 36 h of growth. After harvesting and washing of the cells, chromatophores were prepared by mechanical disruption in a Ribi cell
Characterisation of the Membrane-Bound Pyrophosphatase In R. nrhrum
252
fractionator in 0.2 M glycylglycine, pH 7.4, at 138 MPa (20000 Ib/in2).Cell debris was removed by centrifugation 10000 x g for 60 min and the supernatant was further centrifuged at 100000 x g for 90 min. The pellet was washed by resuspensing and centrifuging once with 0.35 M NaCl 1 mM EDTA 200 pM dithioerythritol and once with 0.2 M glycylglycine, pH 7.4, in which the chromatophores also were suspended. The preparation was stored under nitrogen at 0 "C.
+
+
Assuys
The membrane-bound pyrophosphatase activity was assayed in a reaction mixture containing chromatophores corresponding to 0.1 mg protein, MgC12, Na2P207,1 ml 0.1 M Tris-HC1, pH 8.0, and HzO in a total volume of 2 ml. The assay after incubation with the amino acid reagents is as above with 30 pl 50 mM MgClz and 50 pl 10 mM Na2P207. The amino acid reagents were incubated with the chromatophores and 0.15 mg protein was taken for the assay. In this case the assay mixture contained reduced glutathione at a concentration equal to the initial concentrations of reagents to stop the reaction. The assay mixture was incubated at 28°C and the reaction was terminated after 7 min by addition of 10% trichloroacetic acid. No activity could be detected even after prolonged incubation (30 min) without addition of Mg2+. The reaction was linear with time for at least 15 min, though it has been reported [4] that Pi may interfere with the reaction. Every point in the figures represent the mean value of triplicates. Blanks were run where trichloroacetic acid was added
before the chromatophores. Pi was assayed colorimetrically as described by Rathburn et al. [8]. The dissociation constant for MgPPi was taken from Martell and Siltn [9]: Kd = 2 pM. Protein was determined according to Lowry et al. [lo]. RESULTS Dependence of the concentrations of the substrate MgPPi, of Mg2+ and of PPi. In Fig.1 the rate of hydrolysis as a function of MgPPi is studied at different concentrations of free Mg2+. With increasing concentrations of free Mg2 the activity increases until it reaches its maximum at about 50 pM free Mg2+.A further increase in fM$+] does not give higher activity, but the maximum is reached at lower substrate concentrations. This can be explained by assuming that the enzyme becomes more saturated with Mg2 and therefore the possibility for MgPPi to bind to an unactivated enzyme molecule becomes reduced. On the other hand, when the concentration of free PPi increases, it produces a decrease in velocity and the maximum velocity is reached at higher substrate concentrations (Fig.2). This is to be expected if the enzyme binds free PPi which cannot be hydrolyzed and if the enzyme has to compete with PPi for the free Mg2+ ions which activate the enzyme. +
+
Kinetic Analysis The Lineweaver-Burke plot in Fig. 3 also illustrates the dependence of membrane-bound pyrophospha-
H. Randahl
253
Fig.). Dependence of' the activity on the concentration o f MgPPi, at diflerent concentrations ofj?ee PP,. 1 (0).10 (A), 50 (0). 100 ( x ), loo0 pM (0)free PPi
500 (A)and
0
0.01
1I[%PPiI
0.02
(W')
Fig. 3. Lineweaver-Burke plot: iflustrating the dependence of the membrane-bound pyrophosphatase activity on rhe concentruriorr offree Mg2 ', data 1uken.from Fig. I . 1 (O),10 (A), 50 ( x ) . 100 (0),and 500 pM (A)
tase activity on M g + . The determination of apparent K,,, is very difficult, since it is dependent on the free M$" concentration. The apparent K, values for substrate concentrations between 100pM and 1000pM with a free Mgf ion concentration of 10, 50, 100 and 500 pM are 500, 350, 150 and 65 pM respectively. The Hill plot shown in Fig.4 is derived from the data in Fig. 1, substrate concentration 25 - 750 pM. The plot indicates two binding sites. If the activation effect in the presence of excess Mg2+ were due only to complex formation of PPi, the Hill coefficient would not exceed a value of 1 . There is also a positive cooperativity, which is also revealed by the apparent K, values. The higher the concentration of uncomplexed M$+, the better the affinity for the substrate. The inhibition produced by higher substrate concentrations may result From binding of substrates both to the
activation site and to the substrate site, thus giving a sharp decrease in velocity. Derived from the data in Fig. 1 a Dixon plot gives a K = 4 pM. It is very difficult to obtain an accurate figure for substrate concentrations below 250 pM where the complexed form of PPi is so low compared to the free PPi concentration that the velocity approaches zero and Ki also approaches zero. Heat Inactivation Heat inactivation experiments indicate that binding of Mg2+ to the enzyme renders it more stable (Fig. 5). Incubation of chromatophores at 62 "C without Mg2+ or with PPi gives a sharp decrease in activity, while with Mg2+ the activity is almost completely retained.
Characterisation of the Membrane-Bound Pyrophosphatase in R. rubrum
254
100
80 60
40
1.0
-
20
."
10
-5 2
x
.-
c
2. .c
2 5
C
5
1 0
15
20
25
30
Time (rnin)
Fig. 4. Hill plot for the membrane-bound pyrophosphatase activity at various concentrations of free M g z + , data taken from Fig.1. ( @), 10 (A), 50 ( x ) , 100 (0),500 (A)and 1000 pM (U) free Mgz+
0' 0
10
x)
30
Temperature ("c)
Fig. 6 . Inhibition of'the membrane-houndpyrophosphatase by MulNEt and Nbf-Cl. Chromatophores were incubated with 0.9 mM MalNEt ( 0 ) and 1.5 mM Nbf-C1 (0) for 5 min at different temperatures. Activity plotted on a log scale
Amino Acid Reagents MalNEt is an -SH reagent, incubation with Nbf-C1 gives a differential spectra with an absorbance
Fig. 5. Heat inactivution of' the membrane-bound pyrophosphatase at 62 "C. Chromatophores (O), chromatophores incubated with PPi (A), chromatophores incubated with Mgz+ (n).Activity plotted an a log scale
increase with a broad maximum of 442nm. It has been shown [ l l ] that reaction with cysteine gives a maximum at 425nm. So it might be that Nbf-C1 reacts with something other than - SH groups, further investigation has to be done to reveal this. Inactivation of the membrane-bound pyrophosphatase with MalNEt or Nbf-C1 is very similar except for the temperature dependence, MalNEt gives total inhibition after 10 min at O"C, but does not affect the enzyme at all at 28°C. With the Nbf-C1 there is a 30% inhibition at 0 ° C and at 28°C total inhibition after 10 min (Fig. 6). So there seems to be a strong conformational change in the membrane with temperature. There seems also to be a conformational change when PPi is added to the incubation medium together with one of the reagents. At 0 ° C PPi increases the rate of inhibition caused by MalNEt. At 28 *C incubation with both PP, and MalNEt gives an even greater inhibition than is obtained with incubation with MalNEt alone (Fig. 7). The same pattern is seen with Nbf-C1, where the inhibition at 0°C is increased by incubation with PPi (Fig. 8). When Mg2+ is included together with either of the reagents it protects the enzyme, but only to a certain degree depending on the concentration of Mg2+(Fig. 7 and 8). Mg2+ incubated with MalNEt at 28°C gives the same pattern as with MalNEt alone and incubation with Nbf-C1 with Mg2+ at 0°C gives an almost full protection of the enzyme.
H. Randahl
255
80
100 80
60
60
40
40
100
2
I
5
20
c
,-"
-
20
c
c
0 c
8 10
10
.->
c ZI .> ._ c
%
c
._ c
=r"
2 5
5
0 0
5 Time (min)
Time (min) Fig. I. Inhibition of membrane-bound pyrophosphatase by MalNEt in the presence o f P P i and Mg2+.Incubation with 0.9 mM MalNEt and respectively 15 m M M g z + and 0.5 mM PPi at two different temperatures. At 28 "C chromatophores incubated with MalNEt (0),with MalNEt + PPi (A), and with MalNEt + M g 2 + (0) at 0°C as above respectively (0,A,W). Activity plotted on a log scale
Scheme 1. Proposed model for the membrane-boundpyrophosphatase. E = membrane-bound pyrophosphatase, M = Mgz+,P = PPi and C = MgPPi
DISCUSSION The model shown in Scheme 1 is based on a general model made by London and Steck [12] with the following properties. a) Mg2+ has two roles, as an activator and to complex PPi to give the true substrate, MgPPi. b) PPi is able to bind directly to the enzyme, and is not hydrolyzed in the uncomplexed form. c) The enzyme has an activator site and a substrate site. The evidence for these properties is as follows. a) There seems to be an absolute requirement for Mg2+ ions to activate the enzyme. This is demonstrated by the experiments shown in Fig.1 and 2. That Mg2+ ions can bind to the enzyme is shown in
10
Fig. 8. Inhibition of membrane-bound pyrophosphatase by NbJCI in thepresence of PPiandMg". As in Fig.6 but with 1.5 mM Nbf-CI. Incubation at 28 "C, chromatophores with Nbf-CI (O), with Nbf-CI + PPi (A), and with Nbf-CI M g z + (0). At O'C as above respectively (0,A, M). Activity plotted on a log scale
+
Fig. 5 where Mg2+ also protects against the inhibition caused by the reagents MalNEt and Nbf-CI. That the two amino acids are available at different temperatures makes it difficult to see how Mg2+ is binding to the amino acids. However, Mg2+ protects them when it binds to the activator site. That Mg2+ can be competed out by an irreversible reagent means that Mg2+ does not sit very firmly bound but is in equilibrium the enzyme. Therefore with an increased concentration of Mg2+ the equilibrium between enzyme and Mg2+ will be displace towards the complex enzyme -Mg2+. From this is seen that for V the substrate concentration could be lowered. For low concentrations of free Mg2+ a competition between the enzyme and PPi of the available Mg2+ ions will occur and thus a decrease in velocity is observed. That Mg2+ in this way is very easily detached from the enzyme would also facilitate the regulation of the enzyme. b) That PPi is not hydrolyzed without Mg2+ was mentioned earlier in the text. That PPi can bind to the enzyme without Mg2+ is shown by the experiments in Fig.7 and 8. PPi can bind to a side other than that which binds Mg2 and cause a conformational change. This conformational change and that induced by temperature could in part be the same. However, the total percentage inhibition is greater when PPi is incubated with MalNEt at 28 "C and with Nbf-C1 at 0°C. c) To make the model as simple as possible the substrate is proposed to be able to bind to both sites. The substrate inhibits when the concentration is so +
256
H . Randahl : Characterisation of the Membrane-Bound Pyrophosphatase in R. rubrum
high that it binds to both sites and thus prevents Mg2+ from binding and activating. The substrate inhibition is propably not very important in regulation of membrane-bound pyrophosphatase activities, because such high substrate concentrations are very unlikely to be encountered in the cell. What is more likely to regulate the activity is the Mg2+concentration. In Escherichia coli the total Mg2+ concentration is about 20 mM. Most of this Mg2+ is not freely available, since the ribosomes bind about 85 %. The adenine nucleotide pool binds a very large part of what is left and in that way regulates the free Mg2+ ion concentration in the cell. Calculation has shown that about 300 pM free Mg2+ is available for other reactions (see [13]). The cytoplasmic concentration of PPi in E. coli is about 1.3 mM [14]. If these figures for Mg2+ and PPi concentrations are valid for Rhodospirillum rubrum, neither the membrane-bound pyrophosphatase nor the cytoplasmic pyrophosphatase would function very well. However, cytoplasmic pyrophosphatase may be more active, though the K , values for cytoplasmic pyrophosphatase are 60 pM and 30 pM at 1 : 1 and 4: 1 ratios of Mg2+:PPi respectively [ 2 ] . For the membrane-bound pyrophosphatase the K , is 150 pM and 65 pM for Mg2+:PPi ratios respectively 5 : 1 and 125 : 1. This comparison shows that cytoplasmic pyrophosphatase has a much greater affinity for the substrate and thus can work much more efficiently than the membrane-bound pyrophosphatase, and there has to be a very great excess of Mg2+ over PPi in R.rubrum for the membrane-bound pyrophosphatase activity to be of any significance. The cytoplasmic pyrophosphatase is also allosterically regulated by nucleotides: 50 inhibition is found with NADH, NADPH and ATP in the concentration range 0.4-0.7 mM.
In light, as has been mentioned earlier, the membrane-bound pyrophosphatase is significantly depressed when the bacteria are in their exponential growth phase and the synthesis of ATP and PPi is at its maximum, and the hydrolysis of PPi by membranebound pyrophosphatase is probably very small. These arguments suggest that both enzymes are very firmly regulated during rapid growth of the bacteria in the light. I would like to thank Drs T. Bartfai, J . W . DePierrc and S. Nordlund for their helpful suggestions and advice. This work was supported by grant 2905-100 from the Swedish Natural Science Research Council to Margareta Baltscheffsky.
REFERENCES 1. Klemme, J. H. & Gest, H. (1971) Proc. Nut1 Acad. Sci. (I.S.A. 68, 721 -725. 2. Klemme, J. H. & G a t , H. (1971) Eur. J . Biorhem. 22,529-537. 3. Baltscheffsky, H. & von Stedingk (1966) Biochim. Biophys. Res. Commun. 22, 722-728. 4. Nishikawa, K., Hosoi, K., Suzuki, Y., Yoshimusa, S. & Horio, T. (1973) J . Biochem. ( T o k y o ) 73, 537-550. 5. Baltscheffsky, M. (1967) Nature (Lond.) 214, 241 -243. 6. Cooperman, B. S. & Mark, D. H. (1971) Biochim. Biophys. Acta, 252, 221 -234. 7. Bose, S . K., Gest, H. Sr Ormerod, J. G. (1961) J . Biol. Chem. 236, 13-14. 8. Rathburn. W. B. & Betlach, V. (1969) Anal. Biochem. 28. 436 - 445. 9. Martell, A. E. & Sillen, C. G. (1964) Stability Constants ($ Metal Ion Complexes, p. 190, The Chemical Society Burlington House. London. 10. Lowry, 0 . H., Rosebrough, N. J., Farr, A. L. & Randall, K. J. (1975) J . Biol. Chem. IY3, 265-275. 11. Birkett, D. J., Prince, N. C., Radda, G . K. & Salomon, A. G . (1970) FEBS Lett. 6, 346-348. 12. London, W. P. & Steck, T. L. (1969) Biochemistry, 8, 17671779. 13. Klemme, J. H. (1976) Z . Naturforsch. 31c, 544-550. 14. Heinonen, J. (1974) Anal. Biochem. 59, 366- 374.
H. Randahl, Avdelningen for Biokemi, Arrheniuslaboratoriet, Stockholm Universitet, Fack, S-106 91 Stockholm, Sweden
Characterization of the Membrane-Bound Inorganic Pyrophosphatase in Rh odosp irillum rubrum Hikan R A N D A H L University of Stockholm, Arrhenius Laboratory, Department of Biochemistry (Received June 27, 1979)
The membrane-bound inorganic pyrophosphatase (EC 3.6.1 .l) from Rhodospirillum rubrum has been investigated with the tools of enzyme kinetics, and with two amino acid reagents, N-ethylmaleimide (MalNET) and 4-chloro-7-nitrobenzofurazan(Nbf-Cl). 1. The concentration of the true substrate, MgPPi, was varied with constant concentrations of free Mg2+ or PPi. It was observed that Mg2+ acted as an activator. 2. Heat inactivation of the enzyme at 62°C was slowed down in the presence of Mg2+. 3. MalNET and Nbf-C1 bind to the enzyme, and inhibit its activity. The effect of both reagents is dependent on the temperature. 4. A model is proposed where the 1 : 1 complex of Mg2+:PPi acts as substrate and Mg2+ interacts directly with the enzyme as an activator. PPi can bind to the enzyme, but is not hydrolyzed in the uncomplexed form. The cytoplasmic pyrophosphatase has been studied
in several different organisms. In Rhodospirillum rubrum Klemme and Gest [1,2] have made a very comprehensive study of this enzyme, which accounts for 80% of the total PPi hydrolysis activity in crude extracts. The membrane-bound pyrophosphatase coupled to the electron transport chain in chromatophores has been found only in certain organisms. This enzyme is unique in that it catalyzes not only the hydrolysis of PPi, but also the synthesis of PPi in the light [3]. The synthasis of PPi is believed to be in equilibrium with its hydrolysis [3,4], although PPi hydrolysis is significantly depressed in the light. It was suggested [4] that this depression results from an alteration in the oxidation/reduction state of the electron transport system. PPi has been shown to act as an energy donor both at the cytochrome level and coupled to electron transport chain [ 5 ] .Most inorganic pyrophosphatases require divalent cations for activity and this is also true for membrane-bound pyrophosphatases. But in contrast to the cytoplasmic pyrophosphatases, which require Zn2+ both for stability and for activity and Mg2+ to complex the substrate [l], membrane-bound -~
Ahhreviurions. MalNEt, N-ethylmaleimide; Nbf-CI, 4-chloro7-nitro benzofurazan. Enzymes. Membrane-bound inorganic pyrophosphatase (EC 3.6.1 , I ) ; cytoplasmic inorganic pyrophosphatase (EC 3.6.1.1).
pyrophosphatases require only Mg2 +. The membranebound enzyme has very poor activity with any other divalent cation: about 50% activity is obtained with Z n 2 + and M n 2 + and none with Ca2+ (unpublished results from this laboratory). However, Cooperman and Mark [6] have shown that Mn 2 + and PPi aggregate in dilute solution, which could account for the lower activity with this cation. This paper presents an analysis of the membranebound pyrophosphatase kinetics, its dependence on Mg2+ and the role of PPi.
MATERIALS AND METHODS Pyrophosphate was purchased from Merck (Darmstadt, F.R.G.). N-Ethylmaleimide (MalNEt) and reduced glutathione were obtained from Sigma Chem. Co. (St Louis, Mo., U.S.A.) and Nbf-CI from Serva FeinbiochemicaGmbHundCo.(Heidelberg1 ,F.R.G.). Preparation Techniques Rhodospirillum rubrum strain S1 was grown anaerobically in light at 30 "C in the medium described by Bose et al. [7]. The bacteria were harvested in the exponential phase, after 36 h of growth. After harvesting and washing of the cells, chromatophores were prepared by mechanical disruption in a Ribi cell
Characterisation of the Membrane-Bound Pyrophosphatase In R. nrhrum
252
fractionator in 0.2 M glycylglycine, pH 7.4, at 138 MPa (20000 Ib/in2).Cell debris was removed by centrifugation 10000 x g for 60 min and the supernatant was further centrifuged at 100000 x g for 90 min. The pellet was washed by resuspensing and centrifuging once with 0.35 M NaCl 1 mM EDTA 200 pM dithioerythritol and once with 0.2 M glycylglycine, pH 7.4, in which the chromatophores also were suspended. The preparation was stored under nitrogen at 0 "C.
+
+
Assuys
The membrane-bound pyrophosphatase activity was assayed in a reaction mixture containing chromatophores corresponding to 0.1 mg protein, MgC12, Na2P207,1 ml 0.1 M Tris-HC1, pH 8.0, and HzO in a total volume of 2 ml. The assay after incubation with the amino acid reagents is as above with 30 pl 50 mM MgClz and 50 pl 10 mM Na2P207. The amino acid reagents were incubated with the chromatophores and 0.15 mg protein was taken for the assay. In this case the assay mixture contained reduced glutathione at a concentration equal to the initial concentrations of reagents to stop the reaction. The assay mixture was incubated at 28°C and the reaction was terminated after 7 min by addition of 10% trichloroacetic acid. No activity could be detected even after prolonged incubation (30 min) without addition of Mg2+. The reaction was linear with time for at least 15 min, though it has been reported [4] that Pi may interfere with the reaction. Every point in the figures represent the mean value of triplicates. Blanks were run where trichloroacetic acid was added
before the chromatophores. Pi was assayed colorimetrically as described by Rathburn et al. [8]. The dissociation constant for MgPPi was taken from Martell and Siltn [9]: Kd = 2 pM. Protein was determined according to Lowry et al. [lo]. RESULTS Dependence of the concentrations of the substrate MgPPi, of Mg2+ and of PPi. In Fig.1 the rate of hydrolysis as a function of MgPPi is studied at different concentrations of free Mg2+. With increasing concentrations of free Mg2 the activity increases until it reaches its maximum at about 50 pM free Mg2+.A further increase in fM$+] does not give higher activity, but the maximum is reached at lower substrate concentrations. This can be explained by assuming that the enzyme becomes more saturated with Mg2 and therefore the possibility for MgPPi to bind to an unactivated enzyme molecule becomes reduced. On the other hand, when the concentration of free PPi increases, it produces a decrease in velocity and the maximum velocity is reached at higher substrate concentrations (Fig.2). This is to be expected if the enzyme binds free PPi which cannot be hydrolyzed and if the enzyme has to compete with PPi for the free Mg2+ ions which activate the enzyme. +
+
Kinetic Analysis The Lineweaver-Burke plot in Fig. 3 also illustrates the dependence of membrane-bound pyrophospha-
H. Randahl
253
Fig.). Dependence of' the activity on the concentration o f MgPPi, at diflerent concentrations ofj?ee PP,. 1 (0).10 (A), 50 (0). 100 ( x ), loo0 pM (0)free PPi
500 (A)and
0
0.01
1I[%PPiI
0.02
(W')
Fig. 3. Lineweaver-Burke plot: iflustrating the dependence of the membrane-bound pyrophosphatase activity on rhe concentruriorr offree Mg2 ', data 1uken.from Fig. I . 1 (O),10 (A), 50 ( x ) . 100 (0),and 500 pM (A)
tase activity on M g + . The determination of apparent K,,, is very difficult, since it is dependent on the free M$" concentration. The apparent K, values for substrate concentrations between 100pM and 1000pM with a free Mgf ion concentration of 10, 50, 100 and 500 pM are 500, 350, 150 and 65 pM respectively. The Hill plot shown in Fig.4 is derived from the data in Fig. 1, substrate concentration 25 - 750 pM. The plot indicates two binding sites. If the activation effect in the presence of excess Mg2+ were due only to complex formation of PPi, the Hill coefficient would not exceed a value of 1 . There is also a positive cooperativity, which is also revealed by the apparent K, values. The higher the concentration of uncomplexed M$+, the better the affinity for the substrate. The inhibition produced by higher substrate concentrations may result From binding of substrates both to the
activation site and to the substrate site, thus giving a sharp decrease in velocity. Derived from the data in Fig. 1 a Dixon plot gives a K = 4 pM. It is very difficult to obtain an accurate figure for substrate concentrations below 250 pM where the complexed form of PPi is so low compared to the free PPi concentration that the velocity approaches zero and Ki also approaches zero. Heat Inactivation Heat inactivation experiments indicate that binding of Mg2+ to the enzyme renders it more stable (Fig. 5). Incubation of chromatophores at 62 "C without Mg2+ or with PPi gives a sharp decrease in activity, while with Mg2+ the activity is almost completely retained.
Characterisation of the Membrane-Bound Pyrophosphatase in R. rubrum
254
100
80 60
40
1.0
-
20
."
10
-5 2
x
.-
c
2. .c
2 5
C
5
1 0
15
20
25
30
Time (rnin)
Fig. 4. Hill plot for the membrane-bound pyrophosphatase activity at various concentrations of free M g z + , data taken from Fig.1. ( @), 10 (A), 50 ( x ) , 100 (0),500 (A)and 1000 pM (U) free Mgz+
0' 0
10
x)
30
Temperature ("c)
Fig. 6 . Inhibition of'the membrane-houndpyrophosphatase by MulNEt and Nbf-Cl. Chromatophores were incubated with 0.9 mM MalNEt ( 0 ) and 1.5 mM Nbf-C1 (0) for 5 min at different temperatures. Activity plotted on a log scale
Amino Acid Reagents MalNEt is an -SH reagent, incubation with Nbf-C1 gives a differential spectra with an absorbance
Fig. 5. Heat inactivution of' the membrane-bound pyrophosphatase at 62 "C. Chromatophores (O), chromatophores incubated with PPi (A), chromatophores incubated with Mgz+ (n).Activity plotted an a log scale
increase with a broad maximum of 442nm. It has been shown [ l l ] that reaction with cysteine gives a maximum at 425nm. So it might be that Nbf-C1 reacts with something other than - SH groups, further investigation has to be done to reveal this. Inactivation of the membrane-bound pyrophosphatase with MalNEt or Nbf-C1 is very similar except for the temperature dependence, MalNEt gives total inhibition after 10 min at O"C, but does not affect the enzyme at all at 28°C. With the Nbf-C1 there is a 30% inhibition at 0 ° C and at 28°C total inhibition after 10 min (Fig. 6). So there seems to be a strong conformational change in the membrane with temperature. There seems also to be a conformational change when PPi is added to the incubation medium together with one of the reagents. At 0 ° C PPi increases the rate of inhibition caused by MalNEt. At 28 *C incubation with both PP, and MalNEt gives an even greater inhibition than is obtained with incubation with MalNEt alone (Fig. 7). The same pattern is seen with Nbf-C1, where the inhibition at 0°C is increased by incubation with PPi (Fig. 8). When Mg2+ is included together with either of the reagents it protects the enzyme, but only to a certain degree depending on the concentration of Mg2+(Fig. 7 and 8). Mg2+ incubated with MalNEt at 28°C gives the same pattern as with MalNEt alone and incubation with Nbf-C1 with Mg2+ at 0°C gives an almost full protection of the enzyme.
H. Randahl
255
80
100 80
60
60
40
40
100
2
I
5
20
c
,-"
-
20
c
c
0 c
8 10
10
.->
c ZI .> ._ c
%
c
._ c
=r"
2 5
5
0 0
5 Time (min)
Time (min) Fig. I. Inhibition of membrane-bound pyrophosphatase by MalNEt in the presence o f P P i and Mg2+.Incubation with 0.9 mM MalNEt and respectively 15 m M M g z + and 0.5 mM PPi at two different temperatures. At 28 "C chromatophores incubated with MalNEt (0),with MalNEt + PPi (A), and with MalNEt + M g 2 + (0) at 0°C as above respectively (0,A,W). Activity plotted on a log scale
Scheme 1. Proposed model for the membrane-boundpyrophosphatase. E = membrane-bound pyrophosphatase, M = Mgz+,P = PPi and C = MgPPi
DISCUSSION The model shown in Scheme 1 is based on a general model made by London and Steck [12] with the following properties. a) Mg2+ has two roles, as an activator and to complex PPi to give the true substrate, MgPPi. b) PPi is able to bind directly to the enzyme, and is not hydrolyzed in the uncomplexed form. c) The enzyme has an activator site and a substrate site. The evidence for these properties is as follows. a) There seems to be an absolute requirement for Mg2+ ions to activate the enzyme. This is demonstrated by the experiments shown in Fig.1 and 2. That Mg2+ ions can bind to the enzyme is shown in
10
Fig. 8. Inhibition of membrane-bound pyrophosphatase by NbJCI in thepresence of PPiandMg". As in Fig.6 but with 1.5 mM Nbf-CI. Incubation at 28 "C, chromatophores with Nbf-CI (O), with Nbf-CI + PPi (A), and with Nbf-CI M g z + (0). At O'C as above respectively (0,A, M). Activity plotted on a log scale
+
Fig. 5 where Mg2+ also protects against the inhibition caused by the reagents MalNEt and Nbf-CI. That the two amino acids are available at different temperatures makes it difficult to see how Mg2+ is binding to the amino acids. However, Mg2+ protects them when it binds to the activator site. That Mg2+ can be competed out by an irreversible reagent means that Mg2+ does not sit very firmly bound but is in equilibrium the enzyme. Therefore with an increased concentration of Mg2+ the equilibrium between enzyme and Mg2+ will be displace towards the complex enzyme -Mg2+. From this is seen that for V the substrate concentration could be lowered. For low concentrations of free Mg2+ a competition between the enzyme and PPi of the available Mg2+ ions will occur and thus a decrease in velocity is observed. That Mg2+ in this way is very easily detached from the enzyme would also facilitate the regulation of the enzyme. b) That PPi is not hydrolyzed without Mg2+ was mentioned earlier in the text. That PPi can bind to the enzyme without Mg2+ is shown by the experiments in Fig.7 and 8. PPi can bind to a side other than that which binds Mg2 and cause a conformational change. This conformational change and that induced by temperature could in part be the same. However, the total percentage inhibition is greater when PPi is incubated with MalNEt at 28 "C and with Nbf-C1 at 0°C. c) To make the model as simple as possible the substrate is proposed to be able to bind to both sites. The substrate inhibits when the concentration is so +
256
H . Randahl : Characterisation of the Membrane-Bound Pyrophosphatase in R. rubrum
high that it binds to both sites and thus prevents Mg2+ from binding and activating. The substrate inhibition is propably not very important in regulation of membrane-bound pyrophosphatase activities, because such high substrate concentrations are very unlikely to be encountered in the cell. What is more likely to regulate the activity is the Mg2+concentration. In Escherichia coli the total Mg2+ concentration is about 20 mM. Most of this Mg2+ is not freely available, since the ribosomes bind about 85 %. The adenine nucleotide pool binds a very large part of what is left and in that way regulates the free Mg2+ ion concentration in the cell. Calculation has shown that about 300 pM free Mg2+ is available for other reactions (see [13]). The cytoplasmic concentration of PPi in E. coli is about 1.3 mM [14]. If these figures for Mg2+ and PPi concentrations are valid for Rhodospirillum rubrum, neither the membrane-bound pyrophosphatase nor the cytoplasmic pyrophosphatase would function very well. However, cytoplasmic pyrophosphatase may be more active, though the K , values for cytoplasmic pyrophosphatase are 60 pM and 30 pM at 1 : 1 and 4: 1 ratios of Mg2+:PPi respectively [ 2 ] . For the membrane-bound pyrophosphatase the K , is 150 pM and 65 pM for Mg2+:PPi ratios respectively 5 : 1 and 125 : 1. This comparison shows that cytoplasmic pyrophosphatase has a much greater affinity for the substrate and thus can work much more efficiently than the membrane-bound pyrophosphatase, and there has to be a very great excess of Mg2+ over PPi in R.rubrum for the membrane-bound pyrophosphatase activity to be of any significance. The cytoplasmic pyrophosphatase is also allosterically regulated by nucleotides: 50 inhibition is found with NADH, NADPH and ATP in the concentration range 0.4-0.7 mM.
In light, as has been mentioned earlier, the membrane-bound pyrophosphatase is significantly depressed when the bacteria are in their exponential growth phase and the synthesis of ATP and PPi is at its maximum, and the hydrolysis of PPi by membranebound pyrophosphatase is probably very small. These arguments suggest that both enzymes are very firmly regulated during rapid growth of the bacteria in the light. I would like to thank Drs T. Bartfai, J . W . DePierrc and S. Nordlund for their helpful suggestions and advice. This work was supported by grant 2905-100 from the Swedish Natural Science Research Council to Margareta Baltscheffsky.
REFERENCES 1. Klemme, J. H. & Gest, H. (1971) Proc. Nut1 Acad. Sci. (I.S.A. 68, 721 -725. 2. Klemme, J. H. & G a t , H. (1971) Eur. J . Biorhem. 22,529-537. 3. Baltscheffsky, H. & von Stedingk (1966) Biochim. Biophys. Res. Commun. 22, 722-728. 4. Nishikawa, K., Hosoi, K., Suzuki, Y., Yoshimusa, S. & Horio, T. (1973) J . Biochem. ( T o k y o ) 73, 537-550. 5. Baltscheffsky, M. (1967) Nature (Lond.) 214, 241 -243. 6. Cooperman, B. S. & Mark, D. H. (1971) Biochim. Biophys. Acta, 252, 221 -234. 7. Bose, S . K., Gest, H. Sr Ormerod, J. G. (1961) J . Biol. Chem. 236, 13-14. 8. Rathburn. W. B. & Betlach, V. (1969) Anal. Biochem. 28. 436 - 445. 9. Martell, A. E. & Sillen, C. G. (1964) Stability Constants ($ Metal Ion Complexes, p. 190, The Chemical Society Burlington House. London. 10. Lowry, 0 . H., Rosebrough, N. J., Farr, A. L. & Randall, K. J. (1975) J . Biol. Chem. IY3, 265-275. 11. Birkett, D. J., Prince, N. C., Radda, G . K. & Salomon, A. G . (1970) FEBS Lett. 6, 346-348. 12. London, W. P. & Steck, T. L. (1969) Biochemistry, 8, 17671779. 13. Klemme, J. H. (1976) Z . Naturforsch. 31c, 544-550. 14. Heinonen, J. (1974) Anal. Biochem. 59, 366- 374.
H. Randahl, Avdelningen for Biokemi, Arrheniuslaboratoriet, Stockholm Universitet, Fack, S-106 91 Stockholm, Sweden
