respiration."' 12 Contraction is closely linked to electron flow. It cannot X), yielding a burst of respiration, H+ release and Ca++ transport,

Transcription

1 EVIDENCE FOR ACTIVE PHOSPHATE TRANSPORT IN MIAIZE AHITOCHONDRIA* BY J. B. HANSON AND RAYMOND J. MILLER DEPARTMENT OF AGRONOMY, UNIVERSITY OF ILLINOIS, URBANA Communicated by Sterling B. Hendrick, May 26, 1967 The mechanism of Ca++ uptake in animal mitochondria has been proposed to involve a reaction at coupling sites between Ca++ and the nonphosphorylated intermediate (I - X), yielding a burst of respiration, H+ release and Ca++ transport, the latter being dependent on a permeant anion such as phosphate or acetate.1' 2 Alternatively, cation transport at the coupling site might be due to a H+-cation exchange pump.3 Permeation of anions can be as nondissociated acids3 or in exchange for hydroxyl ions.3' 4 Collectively, these events account for the uncoupling and swelling responses to Ca++. The bond energy conserved during electron transport is expended in cation transport, releasing acceptorless respiration or reducing phosphorylation in the presence of acceptor. Swelling and contraction are considered to be simple osmotic adjustments in response to salt transport or precipitation.' 3 Rasmussen5 summarizes, "Cation transport is the primary energy-linked phenomenon in mitochondria; anion movements are passive and depend on the distribution of cations; and H+ may be evolved under circumstances where there is no change in energy reserves." The possibility of active anion transport in animal mitochondria seems not to have been seriously considered except for an early suggestion by Rossi and Lehninger,6 which was dropped upon discovery of calcium binding in the absence of phosphate.' Experiments in our laboratory on Ca++ + Pi transport in corn mitochondria are not consistent with the above views.8-'4 Calcium produces no significant uncoupling responses except in the presence of phosphate; in this case there is phosphate transport which is competitive with adenosine 5'-triphosphate (ATP) formation.8' I We have found no "permeant" anion which will substitute for phosphate in Ca++ accumulation except for the related arsenate ion which is weakly effective in the presence of oligomycin.14 Phosphate does not appear to freely permeate the mitochondrial membrane; rather, it reacts with the energized mitochondria to produce a potent inhibition of substrate-powered contraction. '0 Addition of Ca++ relieves this inhibition, producing Ca++ + Pi transport and accelerated respiration."' 12 Contraction is closely linked to electron flow. It cannot be uncoupled from respiration while ion transport can, and is thus to be distinguished from osmotic volume adjustments accompanying salt transport.0 11, 13 The physical basis for contraction is thought to lie with the matrix or inner membrane. 13 Mitochondria placed in the contracted state possess a potential for binding Ca++ in an exchangeable form, but the ion is not accumulated until phosphate is added.'3 In producing transport it does not matter whether Ca++ or Pi reacts first with the contracted state. 13 From these observations we have postulated a Ca++-activated phosphate transport from X P, with Ca++ (and Mg++ if present8) also serving as the accompanying cation."' 13, 14 In our view, phosphate transport reduces to a diversion -- from 727

2 729 BIOCHEMISTRY: HANSON AND MILLER PPROC. N. A. S mM Mg.412 mm Ca 1 3M 3.~~~~~~~~~~~~~~~0 E TOTAL &P 0.10 TOTAL -1 normal oxidative phosphorylation pathways brought about by substituting Ca++ for Alg++ on the phosphorylated intermediate; that is, an unstable CaX P breaks down to yield Ca++ + Pi transport, whereas the more stable 1\/IgX P exists long enough to be utilized in ATP formation. In this report we describe recent experiments which support this hypothesis, and which indicate something about the nature of permeant anions. Methods.-Mitochondria were isolated as previously described, but with the omission of adenosine 5'-diphosphate (ADP) during washing." Swelling and contraction were followed at room temperature by light-scattering, using nicotinamideadenine dinucleotide (NADH) (Sigma) as substrate."3 Bound '-Ca++ was measured in parallel by centrifuging the mitochondria from 2.6-ml aliquots of the medium down through 10 ml of 0.6 M sucrose at 31,000 X g for two minutes (00C), and collecting the pellet for measurement of radioactivity." Respiration was measured with the Clark oxygen electrode. Result.--Ca++ vs. Mg++ in oxidative phosphorylation and in aubstrate-powered contraction: Figure 1 shows the way in which the Ca/Mg ratio governs the relative amount of phosphate accumulation. The cations were varied in a medium otherwise complete for oxidative phosphorylation, including a hexokinase trap. According to our hypothesis, the addition of Ca++ should lead to a displacement of Mg++ from X - P with a resulting diversion of phosphate from oxidative phosphorylation into uptake. A diversion does occur. As shown, the experiment can be turned around, starting with Ca++-activated phosphate uptake and shifting to phosphorylation on addition of Mg++. A complicating feature is introduced by the depressive effect of Ca++ on electron transport, a phenomenon for which there is no ade- lio- LB Pi~~~~~~~~~~~~~~~~~~~~~~- -~~.ATP PIepo.P-.uptakeI \PimATP I I I 4 COCIR mmmmm Mg $04 FIG. 1.-Competitive induction of phosphate utkbyclimion. Centrifug tubes were prepared to contain in a total volume of 2.6 ml: 0.1 ~,2m Tricine (ph 7.5)f,l mg/ml bovine serum albumin (BSA), 10 mm succinate, 10 mm pyruvate, 40pnmoles CoA, 230 Whmoles NAD+ 170 jsmoles TPP, 4 mm KH2PO4, 1 mm ADP, 50 mm glucose 3 mg hexokinase (Sigma, type III), and MgSO4 and CaCl2 as indicated. Reaction started by addition of 0.1 ml (0.95 mg protein) of mitochondrial preparation to 2.5 ml of prepared solution, and continued for 3.5 min at room temperature (approximately 25 ). Ten ml of ice-cold 0.6 M sucrose were layered beneath the solutions (additional 0.5 min) and the mitochondria rapidly centrifuged through the sucrose. The supernatant was mixed and analyzed for Pi content. The mitochondrial pellet was extracted in 5 ml of ice-cold 10% trichloroacetic acid and similarly analyzed. Respiration was measured in parallel treatments with the oxygen electrode. E

3 VOL..59, 1 96i7 BIOCHEMISTRY: HANSON AND MILLER CPM ADH 45Co c 610 CPM oddk 45Ca or Mg54 + NADH MINUTES FIG. 2.-Swelling and contraction of corn mitochondria in the presence of Ca++ or Mg++. Swelling and contraction were followed by light-scattering as previously described.13 The 2.6 ml of medium contained 0.1 M KCI, 0.02 M Tricine (ph 7.5), 1 mg/ml BSA, 40 Mmoles CoA, 230 iumoles NAD +, 170 Mmoles TPP, and 0.82 mg mitochondrial protein. After 10 min of swelling, additions were made to give 1 mm 45CaC12 (58,400 cpm/tube) or 2 mm MgSO4 with 0.9 Mmoles NADH as indicated. Identical treatments were carried out in parallel test tubes; these were collected in duplicate at the points indicated and the mitochondria recovered for "Ca analysis. Respiration rates as determined in parallel with the oxygen electrode were zero before addition of NADH, and after addition were 70 and 38 mpmoles 02/min for CaCl2 and MgSO4 treatments, respectively, and continued at linear rate until NADH exhaustion. quate explanation. The inhibition is associated with some depression of P/O ratio. However, with allowance for this action of Ca++ it can be seen that the Ca/Mg ratio directs what happens to the phosphate taken up by phosphorylating mitochondria. The postulate that the Mg++ complex (i.e., Mg:X - I and Mg:X -- P) is more stable is supported by contraction and reswelling experiments. In Figure 2, the mitochondria were allowed to swell in buffered KCl, 45Ca++ or Mg++ was added, and the mitochondria were contracted with NADH. Analyses were made for "Ca at the peak of contraction and after the rapid reswelling that occurs on exhaustion of NADH."3 Nonspecific binding was determined on the -NADH control treatment. (The level of nonspecific binding shows no significant change with time.) Extra Ca++ was bound when the mitochondria were contracted and this extra Ca++ was lost on reswelling. In three repetitions of this experiment the extra binding was 47, 57, and 65 msmoles Ca++/mg protein. As discussed elsewhere,'0, 11, 13 contraction seems to be a function of the steady-state concentration of nonphosphorylated intermediate, and it is to this unknown that 46Ca++ must bind. Upon exhaustion of substrate the continued spontaneous hydrolysis of I - X leads to swelling and release of Ca++.

4 730 BIOCHEMISTRY: HANSON AND MILLER PROC. N. A. S. I+X+Ca++ swollen electron transport ' I - spontaneous hydrolysis contracted X:Ca++ When Mg++ was substituted for Ca++, the rate and extent of contraction was lower and the stability of the contracted state was higher (Fig. 2). The oxidation rate of NADH with Mg++ was only 54 per cent of that with Ca++ (legend, Fig. 2). The simplest explanation of this result would be that Mg++ reacts more slowly than Ca++ and produces a more stable intermediate; hence, the rate of intermediate turnover and coupled respiration would be lower. Unfortunately, for want of a suitable isotope it has not been possible to demonstrate directly that extra Mgg++ is bound in an exchangeable form. Acetate as a permeable anion: Past failures to obtain Ca++ uptake with acetate led to a search for passive permeation of acetate in the swelling-contraction phenomenon (Fig. 3). Spontaneous swelling in corn mitochondria is unlike that of animal mitochondria'5 in that it is a passive process that does not require swelling agents or an energy supply. 10 In part we think it is a simple osmotic event governed by the rate of salt permeation, although the volume and light-scattering response are probably also conditioned by relaxation of the contractile mechanism As shown in Figure 3, where 0.1 M KCl and K acetate are compared, acetate was if anything less effective than chloride in the passive swelling. However, upon introduction of NADH a remarkable difference was found; with acetate there was Co BOUND - CPM i\ ACETATE CHLORIDE NADH CHLORIDE 0.5 ACETATE I I l I MINUTES FIG. 3.-Substrate-dependent swelling of mitochondria in acetate. Experiment performed as in Fig. 2, but with the omission of CoA, NAD+, and TPP, and substitution of 0.1 M potassium acetate for 0.1 M KCl in one treatment. Aliquots were taken for 4Ca++ binding at times indicated by arrows. The *CaCl2 (101, 100 cpm/tube) was added initially. Respiration rates after addition of N ADH were 70 and 62 mpumoles 02/min for chloride and acetate, respectively.

5 VOL. 58, 1967 BIOCHEMISTRY: HANSON AND MILLER 731 a rapid energy-linked swelling and no extra Ca++ was bound. With both anions there was rapid respiration (legends, Figs. 3 and 4). If one must assume that the rapid swelling was the osmotic consequence of passive acetate permeation, then acetate can permeate only when electrons are flowing. Other monovalent anions were tested in the same system. Propionate and butyrate were nearly as effective as acetate in producing active swelling (e.g.: A OD of , , after addition of substrate for 0.1 M K acetate, propionate, and butyrate, respectively). A complete report on this organic acid response will be made later when we have adequate details on the flux of anions and K+ in and out of the mitochondria. Current experiments show no net accumulation of Ca++ (see also Figs. 3 and 4), K+, or H3-labeled acetate during active swelling, but the ions may be lost during reisolation of the mitochondria. Nitrate was discovered to give excellent passive swelling (Fig. 4), suggesting that here was a truly permeant anion. However, the qualitative response to addition of NADH was like that with chloride-contraction with labile Ca++ binding, not true accumulation. Note that contraction as an energy-linked process is superimposed on the osmotic adjustment obtained during passive swelling Co BOUND - CPM ACETATE CHLORIDE NITRATE k- 0.D NADH I I I I MINUTES FIG. 4.-Failure of permeant nitrate ion to give active swelling. Experiment performed as in Fig. 2, but using 0.1 M K acetate and 0.1 M KNOa in place of 0.1 M KCl as indicated, and adding the 16CaCl2 (103,400 cpm/tube) initially. Mitochondrial protein was 0.84 mg/tube. Respiration rates after addition of NADH were 85, 77, and 83 mgmoles 02/min for acetate, chloride. and nitrate, respectively.

6 732 BIOCHEMISTRY: HANSON AND MILLERo P)ROC. N. A. S. Succinate was also effective in energizing the active swelling in acetate (Fig. 5). Addition of 1 mm CN-, which produces per cent inhibition of respiration, produced an equivalent inhibition of active swelling. Figure 5 also shows that Ca++ is not needed in the active swelling. Although not shown, addition of 2 mm Mg++ was equally without effect. Coenzyme A, which we sometimes add to maximize substrate acid oxidation, proved not to be a determinant in the active swelling with organic acids (e.g., Fig. 3 vs. Fig. 4). 0.9 SUCCINATE D \ +CN FIG. 5.-Inhibition of substrate-dependent swelling by cyanide. All treatments were in 0.1 M K acetate, 0.02 M Tricine (ph 7.5), and 1 mg/ml BSA, with 1 mm CaC12 and 1 mm KCN as + co++ indicated. Succinate at 10 mm was CONTROL not added determined as substrate. here, but Respiration 1 mm KCN was 0.6 l produces 80-90% inhibition MINUTES Discuss8ion. Tor p)uriposes of this discussion it is assunied that both the 1tonlphosphorylated and phosphorylated intermediates of oxidative phosphorylatioll are real chemical entities possessing a covalent linkage. This holds even for AMitchell's hypothesis,16 in which formation of I -- X is visualized to result from water removal as H+ and OH- during electron flow. So far as one can tell, the bond between the unknowns I and X is an acid anlhydride which depends on the hydrophobic membrane phase for stability. Active Ca++ binding can best be explained by assuming that the reaction of I + X to produce I X creates ligands for metal binding. (As acknowledged previously, this concept evolves from suggestions of Chance2 aand Gregg and Lehningert7 that X can bind cations.) Indeed, metal binding is probably an essential part of I X formation, amid not just - an ancillary event arising froml exchange

7 VOL. 58, 1967 BIOCHEMISTRY: HANSON AND MILLER 733 of cation for dissociable H+. In this view, the actively bound Ca++ in Figures 2 to 4 reflects the steady-state levels of I - X: Ca++ plus any other X: Ca++ -- intermediates that might be subsequently derived. Rasmussen et al. I estimate the concentration of Ca++-binding X to be severalfold that of I X. Since X is -- completely unknown, the mode of binding can only be speculated. With an acid anhydride, the coordination might be through carbonyl oxygen. For our purposes, -X can best be operationally defined as a substance which is created by electron transport (or ATP hydrolysis); reversibly binds Ca++; provides energy for Ca++ + Pi transport; and is somehow related to the contraction (water expulsion from the matrix) which accompanies electron flow. The concentration of --X seems to vary from 45 to 90 mgimoles/mg protein, assuming a 1:1 Ca++ binding ratio. Perhaps X is, or functions through, an acid group (phosphoryl or carboxyl) which may vary in concentration depending upon internal pools of essential metabolites. The best explanation we can find for the apparent instability of the Ca++ complex compared to Mg++ lies with the inner coordination sphere of water of hydration.18 The activation energy for exchange of nearest-neighbor water molecules is 450 and 2610 cal/mole for Ca++ and Mg++, respectively. The rates of water substitution in complex formation are 108/sec for Ca++ and 105/sec for Mg++. Hence, Ca++ could more quickly form and dissociate from the --X: metal complex. In the hydrophobic membrane phase, the more readily available water of hydration of Ca++ could rapidly hydrolyze I - X or P -- X. In the active swelling with acetate (or other short-chain acids) we finally have a reaction in corn mitochondria analogous to the active swelling of animal mitochondria.'5 However, there is no evidence here that acetate is passively permeating to accompany actively accumulated Ca++ or K+. Calcium is neither bound nor accumulated, and is not even needed to produce the reaction. Nitrate passively permeates at a much better rate than chloride or acetate, but it does not support true ion accumulation accompanied by swelling. Only Ca++ binding is found. Hence the important characteristic of the so-called "permeant" anions lies not with passive permeation but with their carboxyl groups. Together with phosphate they possess the capacity to make an acid anhydride by reacting with the nonphosphorylated high-energy intermediate. I - X + acetate I + X acetyl. - The reaction is analogous to that with phosphate and could under certain coiiditioins yield cation-activated anion transport in the same fashion. High concentrations of K+ would have to serve as the activating cation. To date we have not been able to find net K+ or acetate transport during active swelling, but, as indicated, we are dissatisfied with the recovery technique. Swelling need not be wholly in response to ion transport, since the contractile mechanism associated with I X would be dissipated by the attack of acetate. Summary.-Maize mitochondria show greater stability in oxidative phosphorylation and in contraction with Mg++ than with Ca++. This is believed due to necessity for cations in the reaction to form the nonphosphorylated high-energy intermediate (I X: Me++), with Mg++ showing the greater stability because of a -- more stable inner coordination sphere of hydration. Binding and release of extra Ca++ can be demonstrated to accompany the formation and dissipation of the energized and contracted state of the mitochondria.

8 734 BIOCHEMISTRY: HANSON AND MILLER PROC. N. A. S. Acetate and other short-chain monocarboxylic organic acids do not passively permeate corn mitochondria. Rapid swelling only occurs upon introduction of an energy source, and does not lead to Ca++ accumulation. This swelling is believed due to a reaction analogous to that of phosphate, producing X --. acyl which breaks down spontaneously to produce swelling. The swelling may be partially an osmotic adjustment to ion transport, but this has not been demonstrated. * Supported by grants from the Atomic Energy Commission (AT/11-1/790) and the Office of Saline Water, Department of Interior ( ). 1 Rasmussen, H., B. Chance, and E. Ogata, these PROCEEDINGS, 53, 1069 (1965). 2 Chance, B., J. Biol. Chem., 240, 2729 (1965). 3 Chappell, J. B., and A. R. Crofts, in Regulation of Metabolic Processes in Mitochondria, ed. Tager et al. (New York: Elsevier Pub. Co., 1966), vol. 7, p Chance, B., and L. Mela, these PROCEEDINGS, 55, 1243 (1966). 6 Rasmussen, H., Federation Proc., 25, 903 (1966). 6 Rossi, C. S., and A. L. Lehninger, Biochem. Biophys. Res. Commun., 11, 441 (1963). 7 Rossi, C. S., and A. L. Lehninger, J. Biol. Chem., 239, 3971 (1964). 8 Hodges, T. K., and J. B. Hanson, Plant Physiol., 40, 101 (1965). 9 Hanson, J. B., S. S. Malhotra, and C. D. Stoner, Plant Physiol., 40, 1033 (1965). 10 Stoner, C. D., and J. B. Hanson, Plant Physiol., 41, 255 (1966). l1 Truelove, B., and J. B. Hanson, Plant Physiol., 41, 1004 (1966). 12Kenefick, D. G., and J. B. Hanson, Biochem. Biophys. Res. Commun., 24, 899 (1966). 13 Kenefick, D. G., and J. B. Hanson, Plant Physiol., 41, 1601 (1966). 14 Kenefick, D. G., and J. B. Hanson, Symposium on the Use of Isotopes in Plant Physiology, International Atomic Energy Agency, Vienna (1966), in press. 16 Lehninger, A. L., Physiol. Rev., 42, 467 (1962). 16 Mitchell, P., Biol. Rev., 41, 445 (1966). 17 Gregg, C. T., and A. L. Lehninger, Biochim. Biophys. Acta, 78, 27 (1963). 18 Eigen, M., Pure Appl. Chem., 6, 97 (1963); Eigen, M., and G. G. Hammes, Advan. Enzymol., 25, 1 (1963); Samoilov, 0. Y., Discussions Faraday Soc., 24, 141 (1957).