Measurement of the Electrochemical Proton Gradient in Submitochondrial Particles*

(Received for publication, April 5, 1982)

Edward A. Berry$ and Peter C. Hinkle From the Division of Biological Sciences. Section of Biochemistry, Molecular and Cell Biology, Cornel1 Uniuersity, Zthaca, New York 14853

The pH gradient and membrane potential of submi- tochondrial particles from bovine heart were estimated by the uptake of ["C]ethylamine and ["'Cl]perchlorate, using filtration through a glass fiber prefilter and Mil- lipore filter without washing to separate the vesicles from the medium. An external volume probe of ['HI sucrose was also used. Internal volume of the vesicles was measured by the extent of uptake of glucose, which equilibrates slowly across the membrane.

The electrochemical potential gradient of H+ (A,&+) calculated from uptake of ethylamine and perchlorate, assuming the ions taken up were free in solution inside the vesicles, was 23 to 24 kJ/mol of H+ (240-250 mV) during respiration in the absence of ATP. The ratio of the free energy of ATP synthesis (AGAT~) to A,iH+ was 2.2 to 2.3 during oxidative phosphorylation and only slightly higher during ATP hydrolysis indicating that the H+-translocating ATPase is close to equilibrium under both conditions. The nonintegral ratio suggests there is a systematic error in the measurement of

The value of As"+ calculated from ion uptake could be too high if some of the ions taken up are bound to the membrane or concentrated into the electric double layer at the inner membrane-water interface. The ef- fects of vesicle volume (varied osmotically) and per- meant ions (which affect internal ionic strength and pH) on the ratio of AGATp to Aj iH* suggested that ion association with the membrane in fact caused signifi- cant overestimation of A,%"+. Association of ethylam- monium and perchlorate ions with unenergized sub- mitochondrial particles was measured by centrifuga- tion, in the presence of a high concentration of imper- meant salt to minimize association with the external surface. The results were used to estimate the extent of binding during the ion uptake assays, and Ail,+ was recalculated taking this binding into account. The re- sulting values were between 19 and 20 kJ/mol of H+ (197-207 mV) during respiration in the absence of ADP, and the ratio of AGATp to Afilt- was about 3 during oxidative phosphorylation.

AfiH'.

Respiration or ATP hydrolysis by mitochondria or submi- tochondrial particles generates a difference in the electro- chemical potential of hydrogen ions across the mitochondrial

* This work was supported by National Institutes of Health Grant HL 14483. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. + Current address, Department of Biochemistry, Dartmouth Med- ical School, Hanover, NH 03755.

inner membrane (A, i i~+ ). According to the chemiosmotic hy- pothesis of Mitchell (1, 21, this gradient is responsible for the energy coupling between oxidation and phosphorylation. Studies of the energetics of oxidative phosphorylation in whole mitochondria are complicated by transport of nucleo- tides (31, phosphate (4), and respiratory substrates ( 5 ) . These complications can be avoided by studying oxidative phospho- rylation in inverted submitochondrial particles.

The A ~ H + across energy-transducing membranes consists of an electrical potential difference ( A i ) and a pH difference (ApH), which can be measured by their effects on the partition of certain ions across the membrane (6). Ions which cross the membrane electrogenically distribute according to the mem- brane potential, while ions which cross the membrane electro- neutrally by dissociating or biding a proton reflect the pH gradient.

Ion distribution measurements of A , i i ~ + in mitochondria have given values from 140 to 240 mV (7-10). The lowest values (IO) were obtained using centrifugation to measure ion uptake, which may have resulted in loss of isotopes from the pellet after oxygen or substrate was exhausted. The highest values ( 7 , 8) assumed an internal volume of 0.4 pl/mg, which is lower than recently determined values (9).

Use of the ion distribution technique with submitochondrial particles is more difficult because of the difficulty in separating the vesicles from the medium without disturbing the ion distribution. Sorgato and co-workers (11-13) have avoided this problem by using t.he flow dialysis method of Ramos et al. (14) to measure uptake of ions. With succinate as substrate in a HEPES'-KC1 medium, A , i i ~ + was about 240 mV (12). Lower values were obtained with NADH as substrate (12) or in a Tris-phosphate medium (13).

In this paper, we describe a filtration assay for ion uptake by submitochondrial particles and survey possible probes to measure A+ and ApH. In addition, we measured ion binding inside the vesicles which must be taken into account in the determination of A+ and ApH, and compared values ofA,iiH- with the Gibbs free energy of ATP synthesis in equilibrium with the proton gradient.

MATERIALS AND METHODS' J'bcl]I(CIOI WJL prepared by elrcrrolyris of ['bCllNaC1 essenliilly as described by

rroob.,,i (IS,. , O , l , > " ~ d b" prPCIp1Tdcl"" W i t h K L I . l%llerhyi mvrphvline "as prepare*

hy ll'c)merhyl i n d l d e v l r h a larpp p ~ r r m of morpholine. rpacf ing excess morpho-

"" .- .. "

' The abbreviations used are: HEPES, 4-(2-hydroxyethyl)-I-piper- azineethanesulfonic acid ETPH, electron transport particles prepared by sonication of heavy layer bovine heart mitochondria; CCCP, carbonyl cyanide m-chlorophenylhydrazone; MES, 2-(N-morpho- 1ino)ethanesulfonic acid; MOPS, 3-(N-morpholino)propanesulfonic acid; (Me),;DAE, hexamethylethylenediamine; SDS, sodium dodec,vl sulfate.

"Materials and Methods" is presented in miniprint as prepared by the authors. Miniprint is easily read with the aid of a standard

1474

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Electrochemical Proton Gradient 1475

magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Besthesda, MD 20814. Request Document No. 82M-873, cite the authors, and include a check or money order for $1.60 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

eqoariun Fur m e c h u l i n o r p h ~ ~ l i n i . , u r r 5 ~ c l r s p r ~ e of t h e proiondfed l ~ r m "a i rrlrvlaird ~ s l n x the

v c s i l l u s w r (BH*I = (vcsl i le space (sLnl)l (1 + K , I - H ~ ~ ~ I - Y , ~ K ~ I I ~ , , ~ ~ ~ ( 3 ,

where "Vrrlrlr sp*.r (B"+l , , 15 ve,rcic O f the pT.,f"nared r u m . ' 'vcs ir i? Si''," ,",,',," ciltlon ~ ~ n s l d n l uf the amine. end Y,,,r IS the hydrrlgrn i o n a i t l u l l y In th? iiltrarr. Fill Is vrblrli w e i e o f rhr r o r a l amine. r r l < o i a l e d by equaflon I or 6 , Ka LS the i c l d d i s * c , -

all the ammes used i n lh ls work e i c e p l mechylmorphullnr, the pKa 16 so hlgh Lhac ve%rcIc 5p.icr c u f t h e protunared form IS essenf~ally equal c o veslclr space c o l Lhe t o t a l rmlne.

Tn inlcularc '.& m d 5 H from the upc.ike prr ih lurarz ind e c h v l mme, YC ,,sed K J 6 L 1 O ~ and Il~Llrrh"lrmlnr, and L ~ H l r , , c r u s r t', mraaure ti1ier ""lumc. \,'r,lLie s p a c e ne p e i , h L " r i i < . And e t h y l m i n e were ~alculatrd u a i n ~ rquatlons 1 . 1 , and L . L f a l l n f rhc prohe lms Luke" ,up *re free lnb idr the uerrrle, the i u n r r l b u ~ ~ o n c o f ' _ *no *pH L O 4 C d rdn hr l i l C l l l t l i , 4 irun, "PIIrls q p i r c O f P P r c h l " r a i r and rlhylamlnr 7 s

I , , ' . - Kll"l0 1-4 [ " c r i i ' e sp,4ce [L'"?:),

"m I h l

.",, " j" >,'I, = R , 1 ", 1 "i: , %?% ~ 5 E L e J b I e : ( 1 1

V I "

*here I,>,, 1s the, ~ n t r m i l Y U I Y ~ ,I( the veslc,ier m d RTlnlO has the Y B ~ Y L 5 . 6 4 8 kl!Nc,l . % I ? 2 r C . "it+ ,in Lhzn he u h t i l n e d hv add ing lhesr IYC conrrlbuilons.

~~~ ~~

. + ..," = "i'" ~ RT:"lO "pH ( H i

I)LJ~ rr<ulis suggert hat hindlnz u f the ?robL 10ns l # , + l d e t h c \rhlrlcs ~ ~ ~ ~ ~ l f ~ c d n t , cil thdL :zH+ c i l r u l d i r d b, this m e t h o d n v e r ~ ~ : m d L r s chr t r ~ c YJIIIP A mrrhod IY c*,rrect f o r I ," blndlng lnlldc t h e rrslilcL IS d e r r r i b e d I" i l i l e r B F r r l c > , /

-. l r l w ~ I > . I ) C I I 1,, ? x p r c % ';e ~n k,#:Ym?I / /+ r a t h e r lhrn I I w m . 7 ~ , ~ ~ n ~ r r l t l ~ l n l l r l 1 J 1 - ,.,?I:% !or c a i ~ I I . , , m p ~ i l + u n wlth thr (rec encrxy sui ATP hymrhr-l% ( ' ( . * ~ p l i n d w i t h LIIC

( r l l m L~ii.ir~ t u , l L ~ , , b , , ~ t s , m , r l l l , l l , h v 1 0 . ? h . I , , n~ , i i r t I , , &, .~ l I rnc , I d i r i d r h r -i 1 X * I , La1

I / ' L."l.r*i ,,, ,Y,d?,l<>" ro.*l , l , ' l , . - , i l i : i r c d ii lllr rL.p"'.I"' Ch.*,". TI> d,l1i*Il \,llhe-

RESULTS

Internal Volume of Submitochondrial Particles-The in- ternal volume of vesicles can be measured by the extent of uptake of a compound whose transport is passive, electroneu- tral, and not linked to H' ion translocation. Because the internal volume is small relative to the volume of external medium trapped when the vesicles are separated by filtration or centrifugation, the double isotope method used with probes of A# and ApH is not very accurate. Because the internal volume can be assumed constant for long time periods, how- ever, it is possible to use a compound which equilibrates across the membrane slowly and to dilute or wash away the external medium. To choose a compound which permeates fast enough to reach equilibrium in several hours but slowly enough that efflux on dilution or washing will be negligible, we compared equilibration rates of several compounds using light-scattering changes when ETPH were mixed with medium containing the compound. Light scattering a t 90" was measured using an Aminco-Bowman spectrofluorometer and for rapidly equili- brating molecule absorbance was measured with an Aminco- Morrow stopped flow apparatus. The compounds investigated and approximate half-times were: water, 55 ms; glycerol, 450 ms; urea, 500 ms; malonamide, 2.3 s; erythritol, 12.5 s; ribitol, 3 min; mannitol, 230 min; and glucose, 2 3 0 min. Uptake of radioactive ribitol and glucose were measured to estimate internal volume.

Fig. 1 shows the uptake time course of ribitol and glucose in a medium containing 250 mM sucrose. The half-time for ribitol uptake is several minutes, and that for glucose uptake is about half an hour. The extent of ribitol uptake varied from 0.55 to 0.95 pl/mg in seven different preparations of ETPH. Glucose uptake has only been measured in one preparation, in which the extent of uptake was 0.95 pl/mg and of ribitol uptake 0.55 pl/mg. The lower values obtained with ribitol may be due to efflux of part of the ribitol during the 20 s required for dilution and fitration. This was done at 0 "C where the half-time for ribitol efflux was 55 min (not shown). There was a fast early phase, however, which may have resulted in significant loss of isotope from the vesicles in 20 s.

The change in extent of ribitol and glucose uptake with osmolarity of the medium is shown in Fig. 2. The values reached a maximum of about 1.2 pl/mg at low osmolarity, presumably because the small diameter of the vesicles allows the membrane strength to hold a hydrostatic pressure inside when the vesicles are spherical. This interpretation was sup- ported by the observation that ETPH prepared in 250 mM ['4C]sucrose did not lose a significant amount of the label when diluted with water to 5 mM sucrose.

Internal volume calculated from the amount of isotope

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1476 Electrochemical Proton Gradient

T m e , hrs

FIG. 1 (left). Time course of ribitol and glucose uptake. ['HIGlucose or ribitol was added to ETPH at 25 mg/ml in 250 mM sucrose, 10 mM K-MOPS. At the indicated time, 10 pl were taken to a chilled test tube, diluted with 2 ml of chilled dilution mix containing 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 10 pg/ml polylysine HBr, 1 mM ribitol, or 1 mM glucose and filtered through a Millipore filter with a glass fiber prefilter. The amount of the ribitol or glucose inside the vesicles, divided by concentration in the incubation medium, is plotted against the time of dilution. Dilution and filtration took about 20 s. Results with two different preparations are shown for ribitol uptake (X and O), the one denoted by circles being the same preparation as that used for L-glucose uptake (A).

FIG. 2 (center). Effect of medium osmolarity on internal volume. The experiment was as in Fig. 1 except that the concentrated ETPH suspension was diluted I-fold or 2-fold with sucrose solutions of different concentrations before the incubation. Internal volume was measured by diluting and fiitering about 5 (ribitol) or 7 h (glucose) after adding the radioactive compounds. 0.25 mg of ETPH protein was taken to filter each time. The amount of uptake or ribitol (0) and glucose (A) is plotted against the reciprocal of final medium osmolarity.

FIG. 3 (right). Osmotic sensitivity of SCN- and Clod- uptake. The medium contained 0.12 to 1.2 M sucrose, 10 mM K-MOPS (pH 7.0), 20 pg/ml polylysine HBr, 5 mM (NH.,)2SOn, ["Hlsucrose, 0.5 p~ [I4C]KSCN, 17 p~ K"'C10,. Uptake was initiated with 1 mM NADH and the mixture was filtered after 5 min to determine uptake as described under "Materials and Methods." The data aoints are: 0, SCN- uptake; 0, C10; uptake; and A, ratio of SCN- uptake to C104 uptake.

trapped inside the vesicles when ['4C]sucrose was present at the sonication step was 0.76, 0.73, 0.71, and 0.63 pl/mg in four experiments. Sucrose is not completely impermeant, however, and significant efflux may have occurred while washing the pellet (about 1 h on ice).

Internal volume measured as the difference between glyc- erol and sucrose space in centrifuge pellets varied from 0.3 to 1.0 pl/mg at 250 mM osmolarity. The large variation was due to the large volume of interstitial water in the pellet (about 5 pl/mg), which prevented accurate measurement of the small internal volume. We concluded that the glucose uptake method was probably the most accurate and that measure- ment should be made on each preparation. A typical value is 0.95 pl/mg a t 0.25 M sucrose concentration.

Papa and co-workers measured internal volumes of about 1.4 pl/mg for EDTA particles (21) and 2.5 pl/mg for Mg-ATP particles (22) as the difference between water and dextran space in pellets. Branca et al. (13), using sucrose and water, found the internal volume of Mg-ATP particles to be 0.7 p1/ mg. Sorgato and Ferguson (12) suggested the higher values measured by Papa's group were due to exclusion of dextran from a portion of the external pellet volume because of its large radius. Rottenberg et al. (23) found that with chloro- plasts inulin, but not sorbitol, was excluded from part of the external pellet volume.

Indicators of Membrane Potential-We have measured uptake of four anions, SCN-, C104-, I-, and C1-, as possible indicators of A$. Chloride uptake was too slow to be useful as an indicator. Iodide uptake during NADH oxidation by ETPH was much smaller than that of SCN- and C104-, and was only partially reversed by addition of rotenone and CCCP. Only thiocyanate and perchlorate were investigated in detail. Thi- ocyanate has been used to measure A$ in bacterial chroma- taphores (24) and in submitochondrial particles (11-13). Per-

chlorate has not been used before for this purpose, but Gro- met-Elhanan and Leiser (25) found that perchlorate and thi- ocyanate behave as permeant ions and have similar effects on the balance between A$ and ApH in Rhodospirillum rubrum chromatophores. Lehninger (26) showed that valinomycin causes swelling of mitochondria suspended in KC104, indicat- ing that the inner membrane is permeant to perchlorate as the anion. Dilger et al. (27) have measured the permeability of black lipid membranes to thiocyanate and perchlorate.

When we measured simultaneous uptake of [14C]thiocya- nate and [:'"Cl]perchlorate under a wide range of conditions, we found that perchlorate was always taken up to a greater extent than thiocyanate. A possible source of this difference was binding of perchlorate to the ETPH membranes. Binding of either ion to the unenergized vesicles is small relative to the amount taken up in response to a membrane potential, but binding to sites inside the vesicles would be expected to increase with increasing internal concentration, and so may constitute a significant part of the uptake in response to A$. To investigate this possibility, we compared the osmotic sen- sitivity of the two indicators using sucrose to vary the osmo- larity. If the vesicles behaved as ideal osmometers and A$ was the same at each sucrose concentration, a plot of uptake expressed as vesicle space uersus reciprocal osmolarity should give a straight line. Extrapolating to zero (infinite osmolarity) should give the contribution of bound probe to the uptake. Our results are plotted in this fashion in Fig. 3. The lines are curved, probably due to the nonideal osmotic dependence of internal volume seen in Fig. 2. Interpretation is further com- plicated because A# cannot be assumed to be constant. High concentrations of sucrose inhibit respiration and so presum- ably decrease A#. (1 M sucrose inhibits respiration by about 35% relative to 0.25 M sucrose under these conditions; the uncoupled rate is also inhibited.) The effect of sucrose on A$

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Electrochemical Proton Gradient 1477

should change any uptake due to bound probe by approxi- mately the same factor as it changes uptake due to free probe, but the effect of sucrose on internal volume should only affect uptake due to free probe. From the fact that perchlorate and thiocyanate are taken up in essentially the same ratio over a range of sucrose concentrations that results in a large change in vesicle volume (see Fig. 2), we conclude that if either probe is binding they are binding to essentially the same extent. The difference between perchlorate and thiocyanate uptake is thus not due to binding of perchlorate.

Fig. 4 shows the effect of pH on uptake of perchlorate and thiocyanate in the presence and absence of nitrate, a permeant anion. The agreement between the two indicators is fairly good at pH values above 8, but thiocyanate uptake is less than a third of perchlorate uptake below pH 6. A possible expla- nation for this effect of pH on the ratio of anion uptakes is that efflux of HSCN from the acidic interior of the vesicles prevents SCN- from coming to equilibrium with the mem- brane potential. This effect would be greater at low pH because of the greater concentration of HSCN. Perchlorate, being a much stronger acid, would be less susceptible to this problem. Sorgato et al. (11) have concluded that the pK, of HSCN (-1.8) is low enough to ensure that the concentration of the neutral form is insignificant. The transmembraneous distribution of an ion that can cross the membrane in two ionization states depends on the relative permeabilities as well as the relative concentrations of the two forms. Assuming that

I O , , I , I I I ,

perchlorate accurately measures the membrane potential, a permeability coefficient of 9.5 X 10" cm/s for SCN- was calculated from the initial uptake rate. It could be shown that a reasonable permeability coefficient for HSCN (0.5 cm/s) would result in the observed difference between thiocyanate and perchlorate uptake (28). This value is not unreasonably high as the permeability of phosphatidylcholine bilayer mem- branes to HSCN is 2.6 cm/s (29). A significant efflux of HSCN implies that thiocyanate behaves as an uncoupler, increasing the permeability of the membrane to protons, and there have been some reports of uncoupling by thiocyanate at low pH (30, 31, 28). HClO, is a much stronger acid than HSCN. Raman and '"Cl NMR spectroscopic data show that it is probably at least 91% ionized at a concentration of 12 M (32). This concentration has a Hammet acidity function value of -8.12 (estimated from the data of Ref. 33), so the pK,, of HClO, is probably below -8. Earlier reports that the pK,, of HC10, is on the order of -2 were based on proton NMR measurements of the extent of dissociation which were shown to be incorrect (32). The initial rate of uptake of perchlorate is about the same as that of thiocyanate (28), and so the permeability coefficient of HClO, would have to be unreason- ably high for HC10, efflux to significantly affect the extent of uptake.

All of our results are consistent with the difference between perchlorate and thiocyanate uptake being due to HSCN efflux, and so we feel that perchlorate is the better ion for measuring

0- 6 7 8 9

1

pH pH (Osmoiarlty)"

FIG. 4 (left). pH dependence of SCN- and Clod- uptake. The medium contained 250 mM sucrose, 20 pg/ml polylysine, 5 mM KS04 ( to minimize variation in ionic strength with pH), [~'H]sucrose, 0.4 p~ [14C]KSCN, 19 (solid symbols) or 28 (open symbols) p~ K;16 CIO,, and 10 mM K-MES, K-MOPS, or K-HEPES at the indicated pH values. For the open symbols, 1 mM KNOI was also present. ETPH were suspended at 0.5 mg/ml, respiration was initiated with 1 mM NADH, and the suspensions were filtered after 5 min to determine uptake. The data points are: 0, ., SCN uptake; 0, 0, CIOl uptake; and A, A, ratio of SCN uptake to C10, . uptake.

FIG. 5 (center). pH dependence of amine uptake. Uptakes were measured with 0.5 mg/ml ETPH protein in 250 mM sucrose, 10 mM K-MES, K-MOPS, or K-HEPES at various pH values, 10 mM KNOj, 20 pg/ml polylysine HBr, 6.4 p~ [:'H]hexylamine, and either 0.2 p~ ["Clmethylamine (A, A), 0.6 p~ [I4C]ethylamine (0, H), or 0.2 p~ [14C]methylmorpholine (0, 0). Uptake was initiated with 1 mM NADH and the mixture was filtered after 5 min. The data points are: methylamine (A), ethylamine ( 0 , methylmorpholine (O), and hexylamine (A, H, 0). The ratios of methylamine, ethylamine, and methylmorpholine uptake to hexylamine uptake are plotted at the top. Methylmorpholine uptake is expressed as "vesicle space of the protonated form" (see "Materials and Methods") to correct for incomplete ionization.

FIG. 6 (right). Osmotic sensitivity of amine uptake. Uptakes were measured with 0.5 mg/ml ETPH protein in 0.1-1.2 M sucrose as indicated, 10 mM K-MOPS, pH 7.0, 10 mM KNO.1, 20 pg/ml polylysine HBr, 4.5 p~ ['HI hexylamine, and either 0.2 PM [14C]methylamine (A, A), 0.7 p~ ['4C]ethylamine (0, H), or 1.2 PM [14C]methylmor- pholine (0, 0). Uptake was initiated with 1 mM NADH and the mixtures were filtered after 5 min. The data points are: methylamine (A), ethylamine (O), methylmorpholine (O), and hexylamine (A, W, 0). The ratios of methylamine, ethylamine, and methylmorpholine uptake to hexylamine uptake are plotted at the top. Methylmorpholine uptake is expressed as vesicle space of the protonated form (see "Materials and Methods").

J 8

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1478 Electrochemical Proton Gradient

the membrane potential. At pH 7.0 and above, however, the difference in A$ measured by the two probes is less than 20 mV.

Indicators of ApH-We have investigated the use of meth- ylamine, ethylamine, hexylamine, and N-methylmorpholine for measuring ApH in submitochondrial particles. Methyla- mine has been used for measuring pH gradients in several systems since its introduction for this purpose by Rottenberg et al. (23). Ethylamine and hexylamine have been used for measuring ApH in chloroplasts (34,35). N-Methylmorpholine was chosen for its hydrophilic nature and low pK, of 7.41 (36). Ethanolamine was also found to be taken up in response to a pH gradient, but was not studied in detail. When uptake of these amines was measured under the same conditions and expressed as vesicle space of the protonated form, methyla- mine and ethylamine were taken up to about the same extent. Methylmorpholine uptake was slightly greater and hexyla- mine uptake was significantly greater, the magnitude of the discrepancy depending on conditions.

Portis and McCarty (34) found that thylakoids take up more hexylamine than ethylamine, and that decreasing pH inhibits ethylamine uptake more than hexylamine uptake. They concluded that ethylamine fails to reach equilibrium because of efflux of the protonated form, which becomes more significant at acid pH because of the low concentration of the neutral form of the amine. This does not seem to be the cause of the discrepancy in submitochondrial particles, where all the amines show approximately the same pH dependence (Fig. 5). Depression of the uptake due to permeability of the protonated form should vary with the ratio of protonated to neutral form of the amine, and the pH at which it becomes significant should depend on the pK,, of the amine and the ratio of permeabilities of the two forms. The fact that all the uptakes show the same pH dependence thus makes it unlikely that any of the uptakes are significantly affected by efflux of the protonated form. The pH dependence of uptake thus presumably reflects the pH dependence of ApH.

An alternative explanation of the difference beween hex- ylamine and ethylamine uptake in submitochondrial particles is that part of the hexylamine taken up in response to a pH gradient is bound to the membrane. Several investigators (36, 37) have concluded that the high values of ApH calculated from 9-aminoacridine uptake are incorrect because most of the amine taken up is bound to the membrane. Because of its positive charge and hydrocarbon chain, it would not be sur- prising if hexylamine also binds to the membrane. The osmotic dependence of the uptakes is consistent with this interpreta- tion. As Fig. 6 shows, hexylamine uptake was proportionately

I I EtNHv -120 J

-0 2 4 6 IO Time, rnin

-

less sensitive to sucrose than ethylamine, methylamine, or methylmorpholine uptake. The ratio of ethylamine to hex- ylamine vesicle space, for example, decreased from 0.5 at low sucrose to 0.2 at high sucrose concentration. Quantitative analysis of the results is complicated by the inhibition of respiration by high sucrose concentration. (1 M sucrose in- hibits respiration by 50% relative to 0.25 M sucrose under these conditions.) Qualitatively, however, the lesser sensitivity of hexylamine uptake to sucrose may be taken as evidence that a significant fraction of the hexylamine is bound, and so, osmotically insensitive.

The pH gradient is decreased by permeant amines, because influx of the free base form increases internal pH. At a given concentration, hexylamine was more effective than ethyla- mine at inhibiting uptake of either amine, as expected from the greater extent to which hexylamine is taken up. While ethylamine inhibited both uptakes by almost the same frac- tion, hexylamine inhibited its own uptake more than that of ethylamine, so that the ratio of ethylamine uptake to hexy- lamine uptake increased from 0.35 at 6 p~ hexylamine to 0.5 at 0.5 mM hexylamine. This could be accounted for by satur- able binding of hexylamine to the membrane, as such binding would become less significant at high amine concentrations.

As shown later, binding is probably significant with all of the amines. If permeability of the protonated forms is insig- nificant, as indicated by the fact that all the amines show the same pH dependence, it seems reasonable to assume that the differences in uptake are due to binding, and thus, that the amines that give lowest uptakes are most likely to be accurate indicators of ApH. By this criterion, ethylamine and methyl- amine are the most desirable of the indicators we have studied. Methylamine binds to the glass fiber filters more than ethyl- amine does, so we have chosen ethylamine for routine use.

Simultaneous Measurement of ApH and A# Using Ethyl- amine and Perchlorate-From the results of the preceding sections, we chose to use perchlorate to measure A# and ethylamine to measure ApH. The measurements can be made simultaneously using [""Cl]perchlorate and [ ''Clethylamine, with [.'H]sucrose to correct for filter volume. Fig. 7 shows the time course of NADH-driven uptake of these two probes in the absence and presence of the permeant anion nitrate. The steady state extent is reached in 2 to 5 min, the longer time being required when one probe is taken up to a very large concentration ratio. In the sucrose K-MOPS medium without nitrate, perchlorate uptake was about 100-fold greater than ethylamine uptake. Nitrate (10 mM) stimulated ethylamine uptake and inhibited perchlorate uptake. In both cases, the uptakes were reduced nearly to zero by rotenone and CCCP.

Tlrne, rnln

FIG. 7. Time course of uptake of ethylamine and perchlorate. ETPH were suspended a t 0.5 mg/ml in 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 20 pg/ml polylysine HBr, without (A) or with ( B ) 10 mM KN0.r. Respiration was initiated with 1 mM NADH, and after 1.5 min, ["H]sucrose, ["C]ethylamine (0.5 p ~ ) , and K'"CIO1 (11 pM) were added from a stock solution containing the three isotopes, The mixtures were filtered at the indicated times after adding isotopes to determine uptake of ethylamine (0) and Clod- (0). In some cases (dotted lines), rotenone and CCCP (both 1 pM) were added a t 5 min. In A, ethylamine uptake is plotted on the expanded scale at right.

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Electrochemical Proton. Gradient 1479

I " " ' I ' " "

L-"-L" i O 8 6 4 2 0 2 4 6 8 i O

[ NHJSO ) I I , mM [KNO,]. mM * ?

I I 0 20 40 60

Ttme. mm

t 4

FIG. 8 (left). The effects of nitrate and ammonium on uptake of perchlorate and ethylamine. Uptakes of ethylamine and perchlorate were measured with 0.5 mg of ETPH in 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 20 pg/ml polylysine HBr, ["Hlsucrose, ['4C]ethylamine, KJ"C104, and the indicated concentrations of KNO, or NH4(S04)1/1. Respiration was initiated with 1 mM NADH and the mixtures were filtered after 5 min. ApH and A$ were calculated from the uptake values, assuming an internal volume of 0.95 pl/mg and that all of the isotope associated with the vesicles is free in solution inside.

FIG. 9 (center). Time course of A G A T ~ and A j h during oxidative phosphorylation. ETPH were suspended at 0.5 mg/ml in 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 5 mM K-P,, pH 7.5, 2 mM MgS04, 250 p~ ADP, 10 mM K2-succinate, 20 pg/ml polylysine HBr, ["H]sucrose, [14C]ethylamine, and K"'CIOJ. The mixture was aerated by shaking during the incubation. Samples (1 ml) were taken at the indicated times to measure uptake and adenine nucleotide composition. A ~ H + was calculated from the ion uptake values, assuming there is no ion binding.

FIG. 10 (right). Time course of AGATp and A&+ during ATP hydrolysis. ETPH were suspended at 0.5 mg/ ml in 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 2 mM K-P,, pH 7.5, 2 mM MgS04, 20 pg/ml polylysine HBr, ["HI sucrose, ['4C]ethylamine, and K"'CI0,. At zero time, 2 mM ATP was added, and at the indicated time, samples were taken to filter for determination of uptake and adenine nucleotide composition. AbH1 was calculated from the ion uptake values, assuming there is no ion binding.

The gradients calculated from the extent of uptake at 5 min and the internal volume of 0.95 pl/mg are 6.67 and 16.96 kJ / mol for ApH and A+ in the absence of nitrate and 12.68 and 8.54 for ApH and A+ in the presence of nitrate.

Fig. 8 shows the effect of NO:$- and NH1' ions on the gradients measured by perchlorate and ethylamine. KNO:( reduced A+ and increased ApH, while (NH&S04 had the opposite effects. These effects were expected because uptake of nitrate should partially collapse A+, allowing respiration to build u p a larger pH gradient, while NH,' uptake (as NH:,) should partially neutralize the pH gradient, allowing respira- tion to increase A+.

Table I shows the calculated values of ApH, A$, andA,&- with NADH, succinate, or ATP as substrates, in the presence or absence of 10 mM KNOn. With each substrate, A+ was larger than ApH in the absence of KNO:) and ApH was larger in the presence of KNO:I. ATP and succinate gave slightly lower proton gradients than NADH. These values are proba- bly overestimated since, as will be discussed later, part of the ions taken up in response to A+ and ApH are bound to the membrane. The relative values are probably correct, however. It is significant that A j i ~ + was larger with NADH as substrate than with succinate, as Sorgato and Ferguson (12) found the opposite to be the case and concluded that NADH increases the proton permeability of the membrane. The difference between our results and theirs may be due to differences in conditions: we used low protein concentrations with NADH added directly; they used high protein concentration and NADH generated by alcohol dehydrogenase. Because of the relatively positive midpoint potentia1 of the ethanol/acetal- dehyde couple, it is questionable whether alcohol dehydrogen-

TABLE I Uncorrected values of A$, QH, and Ai,,+ with different substrates

The reaction mixtures contained 350 mM sucrose, 10 mM K-MOPS, 20 pg/ml polylysine-HBr, ["H]sucrose, 1.4 p~ ['4C]ethylamine, 22 p~ Kl'CIO.,, and 0 or 10 mM KNO:j. When NADH or ATP was the substrate, ETPH were suspended at 0.5 mg/ml in the reaction mix and 1 mM NADH or 1 mM ATP plus 3 mM MgSO, was added. If succinate was the substrate, ETPH were preincubated with 200 mM mM K2-succinate, 200 mM sucrose, 9 mM K-MOPS (pH 7.5) for 10 min at 20 "C and then suspended in reaction mix at 0.5 mg/ml. In each case, the mixtures were filtered after 5 min to measure uptake of Cl0,- and ethylamine. A+ and ApH were calculated from the extents of uptake assuming no binding. As binding seems to be significant (see below), the actual values of A$, ApH, and A ~ H + are lower. p______

Substrate ~ - .~___ "~

[KNO I] A+ b H A&{- mM k J / m o l

NADH 0 15.5 8.5 23.9 10 8.3 13.3 21.6

Succinate 0 15.4 7.8 23.2 10 8.2 12.4 20.6

ATP 0 14.2 7.9 22.1 10 7.8 12.0 19.9

ase can keep NADH reduced for 10 min under these condi- tions.

Simultaneous Measurement of A&++ and AGKrp-Under conditions of equilibrium of the proton-translocating ATPase, the ratio of the free energy of A T P synthesis (AGA.I.P) to A,i&+ should be equal to the H'/ATP stoichiometry of the ATPase. While it may not be possible to obtain complete equilibrium of the ATPase, the equilibrium ratio can be bracketed between the ratios measured during ATP synthesis

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1480 Electrochemical Proton Gradient

and ATP hydrolysis. Fig. 9 shows the time course of AGATp and A,&+ during oxidative phosphorylation measured by per- chlorate and ethylamine uptake without correction for bind- ing. AGATP and A,&+ increase and then decrease in parallel, so that the ratio remains close to 2.3. In other studies where the initial concentration of ADP was varied from 0.25 to 2 mM and Pi was 5 or 10 mM, the ratio AGATP/APH+ varied only from 2.14 to 2.26.

Branca et al. (13) observed values of AGAT~/A~~H+ close to 2 with succinate as substrate in a medium containing Hepes and KC1. Because the ratio was close to 3 in a medium containing Tris-phosphate, they concluded that the ATPase was not close to equilibrium in the Hepes-KC1 medium. In our me- dium, which is similar to their Hepes-KC1 medium, nearly the same ratio was reached during ATP hydrolysis (Fig. lo), so the ATPase was probably close to equilibrium.

Fig. 11 shows that essentially the same ratios are obtained when the absolute values of AGATP and A , i i ~ + are lowered by inhibiting succinate oxidation with fumarate or by adding ADP during ATP hydrolysis. Fig. 11 also illustrates the fact that the system is close to, but not at, equilibrium. When ATP was the driving force, the values of AGATP were about 2 kJ/ mol higher than when respiration was the driving force at the same value Of&+. The equilibrium ratio of AGATP/APH+ must lie between the values measured during ATPase and during ATP synthesis driven by respiration since the reaction of the

A,,+, kJ/mol FIG. 11. Relation between AGATP and ion uptake near equi-

librium of the F,-ATPase. Triangles: ETPH were suspended at 0.5 mg/ml in 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 20 pg/ml poly- lysine HBr, 2 mM K-P,, 3 mM MgS04, [’H]sucrose, Ci4C]ethylamine, and K”‘CCI0,. ATP and ADP in various ratios were added to give a total concentration of 2 mM nucleotide, and after 5 min, the mixtures were filtered to determine A , L + and AGATE’. Open circles: as above but with succinate and fumarate in different ratios at a total concen- tration of 5 mM, 5 mM K-P,, 2 mM MgS04, 0.25 mM ADP instead of ATP, and the samples were shaken 15 min between adding ETPH and filtering.

proton transport ATPase is going in different directions under these two conditions.

The nonintegral ratio of AGAW to the measured value of A,&+ in these experiments suggests that there is some system- atic error in the measurement OfA,&+. If the actual ratio is 2, then A~*H+ must be underestimated by about 2.7 kJ/mol. If the ratio is 3, then A , i i ~ + must be overestimated by about 5.7 kJ/mol. Systematic errors which could cause underestimation of AFH+ include loss of probes from the vesicles during filtra- tion, permeability of the wrong ionization state of the probes, and presence of vesicles which contribute to the measured internal volume but do not develop a membrane potential and pH gradient.

Filtration of the vesicles is not likely to disturb the ion distribution until after flow through the filter stops, as parti- cles on the filter are continually being washed with medium containing substrates. The agreement between our measure- ments with succinate as substrate and those of Sorgato and Ferguson (12) support this conclusion, as these authors used the flow dialysis method which does not involve separation of the vesicles from the medium. In addition, perchlorate uptake measured by a perchlorate electrode was only slightly greater than uptake measured by the filtration method (not shown).

As discussed previously, the strong acid nature of perchloric acid makes it highly unlikely that permeability of HClO, affects the distribution of perchlorate. The fact that uptake of ethylamine, methylamine, hexylamine, and methylmorpho- line have the same pH dependence makes it unlikely that permeability of the protonated form affects distribution of the amines.

The submitochondrial particles may contain some vesicles which do not develop membrane potentials and pH gradients, or whose membrane potentials and pH gradients are much smaller than those in typical vesicles or which are of opposite polarity. These vesicles would have to contribute 40% of the measured internal volume to cause underestimation of A&+ by 2.7 kJ, however, and this is unlikely. The cytochrome c oxidase activity of the vesicles is stimulated 5- to 8-fold by deoxycholate. This and the fact that greater than 90% of the vesicles have F1-ATPase molecules on the external surface (as seen in electron micrographs) make it unlikely that 40% of the vesicles are inverted. There may be some outer membrane vesicles present, but they would not contribute to the internal volume measured by glucose or ribitol uptake because of their high permeability to small molecules.

A,GH+ could be overestimated because of association of the probe ions with the inner surface of the vesicle membrane, which would result in the internal activity being lower than calculated from the total amount of the ion inside the vesicle and the internal volume. Such association could be either simple binding or accumulation of the ions in the electric double layer adjacent to the membrane surface. As described in the discussion of Figs. 3 and 8, association of an ion with the membrane should make uptake of that ion less sensitive to medium osmolarity. Experiments such as those of Figs. 3 and 8 can only indicate a difference in the degree to which two probes are associated with the membrane, however. The absolute extent of association cannot be determined because there is no way to maintain a constant A+ or ApH while varying osmolarity. Because the ATPase equilibrates readily, however, it should be possible to use the value of AGATP to monitor changes in A&+, and thus, in the sum of A+ and ApH. Fig. 12 shows an experiment in which AGATP and uptake of ethylamine and perchlorate were measured during oxidative phosphorylation at different sucrose concentrations. The ef- fect of osmolarity on internal volume (glucose uptake) has been replotted from Fig. 2 for comparison. Ethylamine and

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Electrochemical Proton Gradient 1481

I20 -

E" I00 -* 80

- .. -

6 0 -

4 4 0 -

20 -

00-

2 44 , 7 *

0

l o I I I 10

[&molorlty I " 0 I 2 5 0 I 2 "

IKCIO,], rnM [ EtNH, ClO, I . rnM

FIG. 12 (left). Effect of osmolarity on uptake of ions and AGATP. AGAW and uptake of ethylamine and perchlorate were measured during ATP hydrolysis (0) and in the steady state after oxidative phosphorylation of a limited amount of ADP (A, A), For 0, each mixture contained 10 mM K-MOPS, pH 7.5, 2 mM K-P,, pH 7.5, 2.5 mM MgSO,, 0.25 mM KClO,, ["H]sucrose, ['4C]ethylamine, K'"'CIO,, 0.5 mg/ml ETPH protein, and 0 to 1 M sucrose. 1 mM ATP was added and the mixture was filtered after 5 min. For A, each mixture contained 5 mM Kz-succinate, 0.25 mM ADP, 2 mM K-P,, 2.5 mM MgS04, and other components as in Fig. 9. The reaction was started by adding the ETPH, incubated 14 min with shaking, and filtered. The bottom shows uptake of C104 (A) and ethylamine (A), and internal volume taken from Fig. 2. The middk and top show AGAw and the ratio of AGAw t o A j i ~ + calculated from the uptake.

FIG. 13 (center). Effect of KCIOl on uptake of ethylamine and perchlorate and on AGATP. The conditions were the same as in Fig. 12 except that sucrose was 0.25 M, KClO, was varied as indicated, and in the ATP hydrolysis experiments, ATP was 2 mM and MgSO, was 3 mM. The time allowed before filtration was 8 min with ATP hydrolysis and 20 min with respiration. A,&* calculated from ion uptake is plotted against KCIOJ concentration in the medium during ATP hydrolysis (A) and after oxidative phosphorylation (0).

FIG. 14 (rcght). Effect of ethylammonium perchlorate on perchlorate and ethylamine uptake and on AGATP. ETPH (0.5 mg of protein) were diluted into 1 ml of 250 mM sucrose, 10 mM K-MOPS, pH 7.5, 2 mM MgS04, 2 or 5 mM K-P,, ["Hlsucrose, ['4C]ethylamine, and K''Cl0,. After 2 min, 1 mM ADP and 5 mM potassium succinate were added. 20 min later, the mixtures were filtered to determine uptake and adenine nucleotide composition. APH* calculated from ion uptake, AGA.,.~, and AGA.I.IJ/A~~H+ are plotted against ethylamine concentration in the medium during ATP hydrolysis (A) and after oxidative phosphorylation (0).

perchlorate uptake are both less sensitive to osmolarity than is uptake of glucose. This could indicate a "bound" component in the uptake of one or both probes, or could indicate an increase in A,ii~+ with increasing osmolarity. The latter is unlikely, as A G A ~ p decreased with increasing osmolarity (mid- dle). A , i i ~ + was calculated from uptake of perchlorate and ethylamine and the appropriate internal volume, and the ratio of A , i i ~ + to A G A T ~ is plotted at the top of Fig. 12. To ensure that the high osmolarity did not prevent equilibration of the ATPase, the experiment was repeated with ATP hydrolysis rather than oxidative phosphorylation, and the results are also plotted at the middle and top of Fig. 12. In both cases, the ratio of AGATP to APH- decreased at high osmolarity. This is probably because association of the probe ions with the membrane becomes more significant as the internal volume, and thus, the amount of probe free in solution inside the vesicle decreases.

Changes in AGATP can reflect changes in A , i i ~ + , but not changes in t,he relative magnitudes of A$ and ApH. The experiment of Fig. 12, thus, cannot indicate whether ethyla- mine or perchlorate is associated with the membrane, but only that at least one is. Changing experimental conditions in a way that should have the opposite effect on association of perchlorate and ethylamine with the membrane may indicate which of the probes is involved. Because ethylammonium and perchlorate have opposite charge, changing the surface charge of the inner membrane surface should increase association of

one ion and decrease association of the other ion. The surface charge can be changed by varying the internal pH. This can be achieved by using a permeant ion such as perchlorate to vary the pH gradient at constant pH. Fig. 13 shows the effect of KClO, on the measured ratio of AGATP to Aj&+ during oxidative phosphorylation and ATP hydrolysis. KC104, like KNOa in Fig. 8, decreased perchlorate uptake while increasing ethylamine uptake by a smaller ratio, so that Aj&+ calculated from the uptakes decreased. AGATP decreased by a propor- tionately smaller amount, so that the ratio of AGA.TP to the measured Aj&- increased with increasing KC104 concentra- tion.

Ethylamine perchlorate also increased the ratio of A G A T ~ to A, i i~+ calculated from ion uptakes (Fig. 14). With ethyla- mine perchlorate, the decrease in the measured value of A , i i ~ + occurs at lower concentrations than with KC10,. Because there is no significant effect on AGATP in this concentration range, it is unlikely that actually decreased, so the decrease presumably reflects a change in some error associ- ated with the measurement of A,&+ by ion distribution, such as in the extent of association of the probe ions with the membrane.

KClO, and ethylamine perchlorate could affect the ionic association of ethylamine and perchlorate and the membrane in three ways: by increasing ApH and thus making the mem- brane inner surface more positive, by increasing internal ionic strength and thus collapsing the electric double layer or local

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1482 Electrochemical Proton Gradient

ionic binding, and by competing for saturable binding sites. The effect on pH would presumably be larger with KClO, than with ethylamine perchlorate, while the effect on internal ionic strength and saturation of binding sites would be greater with ethylamine perchlorate. Because ethylamine perchlorate is more effective than KClO, at low concentrations, it is likely that the increase in ionic strength and competition for ion binding sites are most important. Because these effects would decrease binding of both cations and anions, these experiments do not indicate whether binding of ethylammonium or per- chlorate is more important.

Direct Measurement of Zon Association with the Submito- chondrial Membrane-Association (or binding) of ethylam- monium and perchlorate ions with the vesicle membrane can be measured directly by suspending submitochondrial parti- cles in a medium containing the radioactive ions and glycerol (as a measure of total water volume) and then centrifuging to separate the vesicles from the medium. The filtration assay is not sensitive enough to measure binding to membranes be- cause of the large fiiter volume. With no respiratory substrates or ATP and with an uncoupler added, AFH+ can be assumed to be zero. If the free concentrations (or activities) of permeant ions inside and outside the vesicles are unequal, there will be an electrical (Donnan) potential across the membrane and an equal but opposite pH gradient. The major ions in the prep- aration medium (Cl-, S042-, Mg2+, Mn2+, and Tris’) are lost on washing the vesicles or on resuspending in medium con- taining EDTA, so that little impermeant salt is trapped inside the vesicles. By keeping the external concentration of imper- meant anions and cations equal and greater than 1 mM, the Donnan potential can be minimized. Under these conditions, the ions associated with the vesicles in excess of glycerol must be associated with the membrane either by binding or as part of the electric double layer of charge at the membrane-water interface.

Fig. 15 shows association of perchlorate and ethylamine with the membrane measured in this way as a function of pH and ion concentration to approximate conditions inside ETPH during uptake experiments. As expected, increasing pH de- creases association of perchlorate and increases association of ethylamine with the membrane. The binding in Fig. 15 is reported in units of microliters of external medium which is equivalent to nanomoles bound divided by the millimolar

‘0- ‘ 0 IO 20 30 40 50 1.,...1

[ EtNH, ClO, I , mM [EtNH, CIQ I . mM

FIG. 15. Direct measurement of ethylammonium (A) and perchlorate (B) binding to submitochondrial membranes. ETPe vesicles were suspended a t 0.62 mg/ml in 230 mM sucrose, 2.5 mM choline-MES, 0.1 mM choline-EDTA, 2 pg/ml valinomycin, 2 p~ CCCP, [’H]glycerol, [‘4C]ethylamine, [:“‘C1]KC104, and the indicated concentration of ethylamine perchlorate. Binding was measured after centrifugation as described under “Materials and Methods.” Binding of ethylamine (A) and perchlorate ( B ) is shown at pH 6.6 (O), 6.1 (O), and 5.6 (A).

concentration of the probe. Thus, a horizontal line on the plot represents nonsaturable binding. For both ions, there appear to be two components to the observed binding, a saturable component which is pH dependent and a nonsaturable com- ponent.

These experiments measure association of the ions with both faces of the membrane. Association of ions with the external face does not present a problem for measurement of A# and ApH, as shown by the fact that the amount of ion uptake without substrate is generally insignificant relative to uptake with substrate. The amount of association with the inner face, on the other hand, is a function of the internal concentration and increases on energization. The unenergized control is therefore not relevant.

In order to estimate the amount of association of ethylamine and perchlorate with the internal surface, we used an imper- meant salt a t a high concentration to minimize association with the external surface. Fig. 16 shows the effect of (Me),DAE-mucate on association of ethylamine and perchlo- rate with submitochondrial membranes.

Perchlorate binding at pH 5.6 is only partially inhibited by the impermeant salt. The remaining binding (3.5 pl/mg) was then eliminated by addition of the permeant salt ethylamine Perchlorate. Similarly, ethylamine binding was decreased to about 0.5 pl/mg by the impermeant salt, (Me)6DAE-mucate. The component of ethylamine and perchlorate binding which is insensitive to the impermeant salt probably represents binding to the internal surface of the vesicles.

Fig. 17 shows binding of ethylamine and perchlorate to submitochondrial particles in the presence of 100 mM (Me),;DAE which almost completely prevents binding to the outside of the vesicles. At the pH values chosen, which are similar to those inside energized vesicles, perchlorate binding is greater than ethylamine binding.

Correction of Measurements o f A/&+ for Zon Binding- Calculation of A,&+ requires knowledge of the free concentra- tions of the ethylamine and perchlorate probes inside the vesicles. The values measured in uptake experiments are the total amount inside (free + bound). The relationship between free probe concentration and total probe taken up can be calculated from the data of Fig. 17 and the internal volume of the vesicles. This relation is plotted in Fig. 18, assuming an internal volume of 0.95 pl/mg. This figure can be used to estimate the internal concentration of ethylamine and per- chlorate, and thus, ApH and A# from the total amount of uptake. This is valid, however, only if the internal conditions during the uptake measurement are similar to those during the binding experiment. The amount of ethylamine and per- chlorate bound presumably depends on the internal pH, ionic strength, and concentration of the ion in question. The binding data of Fig. 18 covers the pH range 6.3-6.8, and are fairly pH independent within that range. These data can thus be applied to uptake experiments at pH 7.5 if the pH gradient is 0.7 to 1.2 pH units. The binding data were obtained with ethylam- monium perchlorate constituting most of the internal ionic strength and with ethylammonium and perchlorate at equal concentrations. In general, the internal ionic strength is not known in the uptake experiments, and the internal concentra- tions of ethylammonium and perchlorate are not the same. When the probe ions are present in the external medium at relatively high concentrations, as in the experiment of Fig. 14, massive uptake results in the probes’ contributing most of the internal ionic strength. Because the center of the vesicle must be electrically neutral (except at very low ionic strength when the membrane double layer may extend to the center), the ethylammonium concentration will approximately equal the perchlorate concentration at the center of the vesicles. I t is

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Electrochemical Proton Gradient

, I , I I I f ,-

[EtNH3 Cl0,l

1483

[EtNH, CI0,l lnstde vesicles, mM

FIG. 16 (left). The effect of (Me)eDAE-mucate on ion binding to submitochondrial membranes. Binding of ethylamine at pH 6.6 (0) and perchlorate at pH 5.6 (A) were measured as in Fig. 15, except that the medium contained (Me),,DAE-mucate at varied concentrations. For the points at ionic strength 400 mM and below, 1 mM ethylamine perchlorate was present and (Me)hDAE-mucate was varied between 0 and 100 mM. For ionic strengths greater than 400 mM, 100 mM (Me)tiDAE-mucate was present and ethylamine perchlorate was varied between 1 and 50 mM.

FIG. 17 (center). Binding of ethylammonium and perchlorate ions to submitochondrial membranes in the presence of (Me)6DAE-mucate. Binding of perchlorate ( A ) and ethylammonium ( E ) at pH 6.3 (A), 6.5 (0). and 6.8 (R) were measured as described under “Materials and Methods.” The medium contained (Me),,DAE- mucate to minimize binding to the external surface of the membranes. The incubation mixture contained 14 mg/ml ETPH, 250 mM sucrose, 100 mM (Me)tiDAE-mucate, 5 mM choline-MES, 0.4 mM K-MOPS, 0.2 mM choline-EDTA, 0.35 nmol of CCCP/mg of protein, 0.35 pg valinomycin/mg of protein, [:‘H]glycerol, [14C]ethylamine, KJtiCCIO4, and the indicated concentrations of ethylamine perchlorate. The pH indicated was measured in the supernatant after centrifuging.

FIG. 18 (right). The relation between internal concentration of ethylamine perchlorate and the total amount of ethylamine and perchlorate inside the vesicles. The values were calculated assuming the amount of each ion bound at a given internal concentration is equal to the amount bound at the same (external) concentration in the experiment of Fig. 17. An internal volume of 0.95 pl/mg was used to calculate the amount of ethylamine and perchlorate free in solution inside the vesicles at each concentration at internal pH 6.3 (A), 6.5 (O), and 6.8 (0).

therefore reasonable to apply the binding data to the uptake experiment of Fig. 14 in order to calculate ApH and A+ with allowance for binding. From Fig. 17, ethylamine binding was within experiment error of 0.3 pl/mg under all the conditions tested. When these conditions apply, therefore, internal eth- ylamine concentration can be calculated by dividing the total uptake by the internal volume plus 0.3 pl, that is, by using an “effective internal volume” of 1.25 pl for ethylamine. The resulting values of ApH are lower by 0.12 pH unit (0.68 kJ/ mol) than values calculated without correcting for binding.

Perchlorate binding was pH dependent and partly satura- ble. A# was therefore calculated from the internal perchlorate concentration determined from total uptake by interpolation from Fig. 18. The internal pH for this interpolation was determined from the external pH and the pH gradient calcu- lated as described above.

If ethylamine and perchlorate were the only ions inside the vesicles, their concentration at the center of the vesicles would be equal, ApH and A+ would be equal, and the same internal ethylamine perchlorate concentration could be interpolated from Fig. 18 using either ethylamine or perchlorate uptake. Actually, the amount of ethylamine uptake indicated by Fig. 18 for the concentration determined from perchlorate uptake was higher than the observed ethylamine uptake by 5 to 10 nmol/mg. This is probably due to presence of about 7 mM K’ inside the vesicles from the K-MOPS buffer in which they are stored. This K’ presumably escapes during the binding ex- periment due to the low external concentration and long time allowed for equilibration, but is probably retained in the active uptake experiments due to the short time period, presence of

external K’, and perhaps action of the cation/hydrogen ex- changer (39).

The values of ApH, A+, and A,&+ calculated from the data of Fig. 15 using the procedure outlined above are shown in Fig. 19. The values for A j k + were roughly one-third those of A G A ~ p . While the correction for binding used here is obviously a crude one, the result shows that ion distribution data, which seem to indicate a stoichiometry of 2 for the ATPase if ion binding is neglected, are consistent with a stoichiometry of 3 when ion binding is taken into account.

It would be very useful if we could estimate the error introduced by binding when the probes are at tracer concen- trations, as in most of our experiments. The internal ionic strength in the absence of permeant ions is probably 5 to 10 mM. If binding is unsaturable and depends only on ionic strength and pH, then Fig. 17 can be applied, taking the horizontal axis as ionic strength instead of ethylamine per- chlorate concentration. The pH gradient is usually 7 to 8 kJ/ mol (uncorrected) in the absence of permeant ions. Again taking ethylamine binding as 0.3 pl/mg gives corrected values of ApH of 6.3 to 7.3 kJ/mol, and internal pH values of 6.2 to 6.4. From the pH 6.3 perchlorate binding curve of Fig. 17, perchlorate binding should be between 2.5 and 3 pl/mg. To- gether with the ethylamine binding, this gives a total correc- tion of -3.9 to -4.2 kJ/mol to be applied tOAjiH+. The actual correction may be much higher, however, if there is specific binding of ethylamine or perchlorate.

Applying a correction of -4 kJ/mol to the values in Table I gives A,iiH+ values of 19.7 (NADH) and 19.22 (succinate) kJ/ mol during respiration in the absence of ADP, and 18.1 kJ/

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1484 Electrochemical Proton Gradient

3.4 1 1 I I I 1

I I I 1 ,I I

a o 5 5 ’ I

2 3.65 C EtNH, ClO, I , m M

FIG. 19. Recalculation ofA& from the experiment of Fig. 14 with ion binding taken into account. The procedure used is described in the text. The conditions were oxidative phosphorylation (0) or ATP hydrolysis (A).

mol during ATP hydrolysis. The values during oxidative phosphorylation with succinate as respiratory substrate are about 15 (Fig. 14) to 17 (Figs. 9 and 13) kJ/mol. This is about one-third the value of AGATP (44-48 kJ/mol). It may be that specific, saturable binding of ethylammonium or perchlorate to a small numbet- of sites increases the fraction bound at low concentrations. Another possibility is that a slight permeabil- ity of the (Me)6DAE-mucate used in Fig. 17, or permeability of a minor impurity, results in the internal ionic strength being significantly higher than calculated. This would be most significant at low ethylammonium perchlorate concentrations and may account for the fact that the “corrected” ratio of AGA*p to Aji~+ decreases sharply at low ethylammonium per- chlorate concentrations (Fig. 19).

DISCUSSION

In this study, we have tried to systematically investigate each parameter in the measurement of the electrochemical proton gradient of submitochondrial particles. Several meth- ods were used to measure the internal volume, which was found to decrease at high external osmolarities but to increase only a limited amount at low osmolarities because of the small size of the vesicles. Various radioactive probe ions for A$ and ApH were surveyed. Differences between amine ApH probes were minor or, in the case of hexylamine, were due to binding to the membrane. Of the two likely A+ probes, perchlorate and thiocyanate, the lower uptake of thiocyanate was shown to be caused by permeability of the protonated form. Conven-

tional calculation ofAjiH+ based on perchlorate and ethylamine uptake, however, yielded nonintegral values of the ratio of AGATP to A,ii~+. Looking further for possible errors, we discov- ered significant association (or binding) of the probe ions with the membrane depending upon the pH and ionic strength. This binding was not apparent from earlier studies of the effect of changing internal volume by high sucrose media because of secondary effects of the sucrose and had to be measured directly. The binding of probes is probably more significant in submitochondrial particles than other systems because they are prepared by sonication in a low ionic strength medium and contain few ions. Direct measurement of probe binding and correction for it gave typical values forAjiH+ of 15 to 19 kJ/mol (155 to 197 mV). The ratio of the H+-ATPase (Fo - F,) indicated by the corrected value ofAjiH+ and AGATP was close to 3, the equilibrium position of the reaction being bracketed by measurements with and without respiration. Previous studies of A,&+ in submitochondrial particles, using flow dialysis to measure uptake of thiocyanate and methyla- mine, showed ratios of AGATP to&+ of 1.89, 2.84, and 3.09 in three different media (13). Assuming that ion binding causes overestimation of Aji~+, the ratios obtained seem consistent with a ratio of 3 H+/ATP.

This laboratory previously reported direct measurements of the H+/ATP ratio in ETP“ of 1.7 2 0.09 and 1.22 f 0.07, using mitochondria prepared in the winter and summer, respectively (40). These values were thought to be most consistent with a stoichiometry of 2. Direct pulse methods of ratio determina- tion are susceptible to interference by permeant acids and bases such as phosphate (41). We have not found conditions which increase the values determined directly in submito- chondrial particles, but it is likely that they are too low. Tightly bound permeant acids or bases such as fatty acids or amines are particularly active in rapidly transporting stoichi- ometric amounts of protons across the membrane, and such compounds may be present in the mitochondrial inner mem- brane. Because of possible interference by such phenomena, direct pulse measurements should be taken as minimum es- timates. Direct measurements of the Hf/ATP ratio in mito- chondria by ATP pulse methods originally indicated a value of 2 (42) which was confirmed in the presence of N-ethylmal- eimide to inhibit phosphate efflux (43). More recently, Alex- andre et al. (44) used a rate method in the presence of N- ethylmaleimide and obtained a value of 3.1, although only half of the phosphate efflux was actually inhibited (45) and it is not clear that the value of 3 represented proton transport by the Fo - F, ATPase alone.

Van Dam et al. (46), Azzone et al. (47), and Holian and Wilson (9) reported ratios of AGATP to A,&+ between 3 and 4 in whole mitochondria under conditions favorable for oxida- tive phosphorylation. The slightly lower ratios reported by Nicholls (8) can be attributed to his use of a lower internal volume. Azzone et al. (47) and Holian and Wilson (9) were able to obtain much higher ratios by various treatments that decreased Aji~+ but had a proportionately smaller effect on AGATp, It is uncertain, however, whether the ATPase reaction is close to equilibrium under these conditions. In particular, it is possible that the ATPase is inhibited by the endogenous inhibitor a t low Aj i~+ (48), and AGATP is maintained by sub- strate level phosphorylation and adenylate kinase.

Because the phosphate and adenine nucleotide transport reactions are believed to be coupled to translocation of one proton (3, 4), the H+/ATP ratio for external ATP synthesis by whole mitochondria is expected to be one greater than that in submitochondrial particles. Our results together with those of Branca et al. (13) indicate a ratio of 3 in submitochondrial particles, and so we expect a ratio of 4 in whole mitochondria.

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Electrochemical Proton Gradient 1485

The lower ratios measured in mitochondria could be due to poor equilibration of the ATPase, or to ion binding such as we found in submitochondrial particles. The ATPase and ATP/ ADP exchange may fail to equilibrate with the proton gradient because the ADP concentration becomes very low at the high phosphate potentials achieved during mitochondrial oxidative phosphorylation. Lack of equilibration during oxidative phos- phorylation due to a low level of external ATPase activity would result in AGATP/APH+ ratios lower than the equilibrium ratio. On the other hand, Van Dam et al. (46) reported a ratio of 3.46 in whole mitochondria during ATP hydrolysis. During ATP hydrolysis, poor equilibration would make the ratio higher than the equilibrium ratio, so this is incompatible with a ratio of 4 unless some other error caused underestimation of the ratio.

The internal ionic strength is higher in mitochondria than in ETPH submitochondrial particles, and so ion binding might be expected to be less of a problem. As shown in Fig. 17, however, ion binding in submitochondrial particles is signifi- cant even a t high ionic strength. In addition, the high concen- tration of protein in the mitochondrial matrix is a much larger surface for ion binding. Nicholls (8) showed that ApH and A$ measured by uptake of methylamine and rubidium were equal and opposite, and so claimed that binding was not significant. Because both probes had the same charge, how- ever, this experiment would not detect nonspecific cation or anion binding, such as might result from ionic interaction with the electric double layer at membrane and protein surfaces. The observation of Holian and Wilson (9) that uptake of methylamine indicated essentially the same pH gradient as uptake of acetate or dimethyloxazolidine-2,4-dione provides a stronger argument against ion binding. Because methylamine was excluded from the matrix by the pH gradient, however, a very small error in measuring either methylamine uptake or internal volume would make a large error in the calculated value of ApH. It would be useful to see this experiment repeated using oppositely charged indicators for both the membrane potential and pH gradient.

An H' to ATP ratio of 4 for ATP synthesis in whole mitochondria would result in P/O ratios of l/"site" if the respiratory chain transports 4 H'/pair of electron/site, as some recent measurements have indicated (44, 48-50). Wik- strom et al. (51), however, have provided evidence that during succinate oxidation the H'/2e ratio is only 6. This would result in a P/O ratio of 1.5 with succinate as substrate, slightly higher than values measured in this laboratory (52). Equilib- rium studies of the P/O ratio at the first site (53), which we have conf i i ed , show that the H+/2e ratio of the first site must be greater than the H+/ATP ratio in submitochondrial particles. An H+/ATP ratio of 3 is thus consistent with an H'/2e ratio of 4 for the first site. The H+/2e ratio for NADH- linked respiration is therefore likely to be 10, and the p/O ratio in mitochondria to be 2.5. Such a scheme seems to be most consistent with the data at the present time (54).

REFERENCES

1. Mitchell, P. (1961) Nature (Lond.) 191, 144-148 2. Mitchell, P. (1979) Science (Wash. D.C.) 206, 1148-1159 3. Klingenberg, M., and Rottenberg, H. (1977) Eur. J. Biochem. 73,

4. Chappell, J . B., and Haarhoff, K. N. (1967) in Biochemistry of Mitchondria (Slater, E. C., Kaniuga, A., and Wojtczak, L., e&) pp. 75-92, Academic Press, New York

5. Fanyo, A., Palmieri, F., and Quagliariello, E. (1976) in Horizons in Biochemistry and Biophysics (Quagliariello, E., Palmieri, F., and Singer, T. P., eds) Vol. 2, pp. 60-105, Addison-Wesley Publishing Co., Reading, MA

125-130

6. Rottenberg, H. (1975) Bioenergetics 7, 61-74

7. Mitchell, P., and Moyle, J . (1969) Eur. J . Biochem. 7, 471-484 8. Nicholls, D. G. (1974) Eur. J. Biochem. 50, 305-315 9. Holian, A,, and Wilson, D. F. (1980) Biochemistry 19,4213-4221

10. Padan, E., and Rottenberg, H. (1973) Eur. J. Biochem. 40,431-

11. Sorgato, M. C., Ferguson, S. J., Kell, D. B., and John, P. (1978)

12. Sorgato, M. C., and Ferguson, S. J . (1979) Biochemistry 18,5737-

13. Branca, D., Ferguson, S. J., and Sorgato, M. C. (1981) Eur. J.

14. Ramos, S., Schuldiner, S., and Kaback, H. R. (1976) Proc. Natl.

15. Tromballa, H. W. (1970) Radiochem. Radioanal. Lett. 4,285-292 16. Beyer, R. E. (1967) Methods Enzymol. 10, 186-194 17. Racker, E., and Kandrach, A. (1973) J. Biol. Chem. 248, 5841-

5847 18. Lamprecht, W., and Trautschold, I. (1974) in Methods of Enzy-

matic Analysis (Bergmeyer, H. W., ed) 2nd ed., pp. 2102-2110, Academic Press, New York

19. Jaworek, D., Gruber, W., and Bergmeyer, H. U. (1974) in Methods of Enzymatic Analysis, (Bergmeyer, H. W., ed) 2nd ed., pp. 2127-2129, Academic Press, New York

20. Rosing, J., and Slater, E. C. (1972) Biochim. Biophys. Acta 267, 275-290

21. Papa, S., Storey, B. T., Lorusso, M., Lee, C. P., and Chance, B. (1973) Biochem. Biophys. Res. Commun. 52, 1395-1402

22. Papa, S., Guerrieri, S., Simone, S., Lorusso, M., and Larosa, D. (1973) Biochim. Biophys. Acta 292, 20-38

23. Rottenberg, H., Grunwald, T., and Avron, M. (1972) Eur. J . Biol. 25, 54-63

24. Schuldiner, S., Padan, E., Rottenberg, H., Gromet-Elhanan, Z., and Avron, M. (1974) FEBS Lett. 49, 174-177

25. Gromet-Elhanan, Z., and Leiser, M. (1973) Arch. Biochem. Bio- phys. 159,583-589

26. Lehninger, A. L. (1974) Proc. Natl. Acad. Sci. U. S. A . 71, 1520- 1524

27. Dilger, J . P., McLaughlin, S. G. A., McIntosh, T. J., and Simon, S. A. (1979) Science (Wash. D. C.) 206, 1196-1198

28. Berry, E. A. (1981) Ph.D. thesis, Cornel1 University 29. Gutknecht, J., and Walter, A. (1982) Biochim. Biophys. Acta 685,

30. Cockrell, R. (1973) Fed. Proc. 32, 515 31. Azzone, G. F., Gutweniger, H., Viola, E., Strinna, E., Massari, S.,

32. Akitt, J . W., Covington, A. K., Freeman, J . G., and Lilley, T. H.

33. Yates, K., and Wai, H. (1964) J . Am. Chem. Soc. 86,5408-5413 34. Portis, A. R., Jr., and McCarty, R. E. (1973) Arch. Biochem.

35. Portis, A. R., Jr., and McCarty, R. E. (1976) J . Biol. Chem. 251,

36. Hall, H. K. (1956) J. Phys. Chem. 60,63 37. Fiolet, J . W. T., Bakker, E. P., and van Dam, K. (1974) Biochim.

38. Elema, R. P., Michels, P. A. M., and Konings, W. N. (1978) Eur.

39. Douglas, M. G., and Cockrell, R. S. (1974) J. Biol. Chem. 249,

40. Thayer, W. S., and Hinkle, P. C. (1973) J. Biol. Chem. 248,5395-

41. Brand, M. D., Reynafarje, B., and Lehninger, A. L. (1976) J. Biol.

42. Mitchell, P., and Moyle, J . (1967) Biochem. J. 105, 1147-1162 43. Brand, M. D., and Lehninger, A. L. (1977) Proc. Natl. Acad. Sci.

44. Alexandre, A., Reynafarje, B., and Lehninger, A. L. (1978) Proc.

45. Reynafarje, B., and Lehninger, A. L. (1978) Proc. Natl. Acad. Sci.

46. Van Dam, K., Wiechrnan, A. H. C. A., Hellingwerf, K. J., Arents,

47. Azzone, G. F., Pozzan, T., and Massari, S. (1978) Biochim. Bio-

48. Van de Stadt, R. J., De Boer, B. L., and van Dam, K. (1973)

49. Vercesi, A., Reynafarje, B., and Lehninger, A. L. (1978) J. Biol.

437

Biochem. J. 174, 237-256

5742

Biochem. 116, 341-346

Acad. Sci. U. S. A. 73, 1892-1896

233-240

and Colonna, E. (1976) Eur. J. Biol. 62, 77-86

(1965) Chem. Commun. Long. Chem. Soc. 349-350

Biophys. 156,621-625

1610-1617

Biophys. Acta 368,432-445

J. Biochem. 92, 381-387

5464-5471

5402

Chem. 251, 5670-5679

U. S. A . 74, 1955-1959

Natl. Acad. Sci. U. S. A. 75, 5296-5300

U. S. A. 75,4788-4792

J . C., and Westerhoff, H. V. (1977) FEBS Symp. 45, 121; 132

phys. Acta 501, 307-316

Biochim. Biophys. Acta 292, 338-349

Chem. 253,6379-6385

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1486 Electrochemical Proton Gradient

50. Pozzan, T., Di Virgilio, F., Bragadin, M., Miconi, V., and Azzone, 53. Rottenberg, H., and Gutman, M. (1977) Biochemistry 16, 3220- G. F. (1979) Proc. Natl. Acad. Sci. U. S. A. 76, 2123-2127 3227

51. Wikstrom, M., Krab, K., and Saraste, M. (1981) Annu. Reu. 54. Hinkle, P. C. (1981) in Chemiosmotic Proton Circuits in Biolog- Biochem. 50,623-656 ical Membranes (Skulachev, V. P., and Hinkle, P. C., eds) pp.

52. Hinkle, P. C., and Yu, M. L. (1979) J. B i d . Chem. 254, 2450-2455 49-58, Addison-Wesley Publishing Co., Reading MA

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

E A Berry and P C Hinkleparticles.

Measurement of the electrochemical proton gradient in submitochondrial

1983, 258:1474-1486.J. Biol. Chem. 

  http://www.jbc.org/content/258/3/1474Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/258/3/1474.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from