A Protonmotive Force Drives ATP Synthesis in Bacteria
Peter C. Maloney, E. R. Kashket, and T. Hastings Wilson
PNAS 1974;71;3896-3900
doi:10.1073/pnas.71.10.3896
This information is current as of June 2007.
This article has been cited by other articles:
www.pnas.org#otherarticles
E-mail Alerts
Receive free email alerts when new articles cite this article - sign up in the box at the
top right corner of the article or click here.
Rights & Permissions
To reproduce this article in part (figures, tables) or in entirety, see:
www.pnas.org/misc/rightperm.shtml
Reprints
To order reprints, see:
www.pnas.org/misc/reprints.shtml
Notes:
Proc. Nat. Acad. Sci. USA
Vol. 71, No. 10, pp. 3896-3900, October 1974
A Protonmotive Force Drives ATP Synthesis in Bacteria
(themiosmotic hypothesis/membrane-bound ATPase/membrane potential/valinomycin/ATPase-negative mutants)
PETER C. MALONEY, E. R. KASHKET, AND T. -HASTINGS WILSON
Department of Physiology, Harvard Medical School, Boston, Massachusetts 02115
Communiwed by DeWit Stetten, Jr., July U2, 1974
When cells of Streptococcus lactis or
ABSTRACT
Escherichia coli were suspended in- a potassium-free
medium, a membrane potential (negative inside) could
be artificially generated by the addition of-the potassium
ionophore, valinomycin. In 'response to this inward
directed protonmotive force, ATP synthesis catalyzed by
the mnembrane-bound'ATPase (EC 3.6.1.3) was observed.
The formation of ATP was not found in S. lactis that had
been treated 'with the ATPase inhibitor, NN'dicyclohexylcarbodiimide, nor was it observed in a mutant -of
E. coli lacking the ATPase. Inhibition of ATP synthesis in
S. lactis was also observed when the membrane potential
was reduced by the' presence of external potassium, or
when cells were first incubated with the proton conductor,
carbonylcyanidefluoro-methoxyphenyihydrazone. These
results are in agreement with predictions made by the
chemiosmotic, hypothesis of Mitchell.
In microorganisms the membrane-bound ATPase (EC
3.6.1.3; ATP phosphohydrolase) plays a central role in both
aerobic and anaerobic energy transductions. Studies of
ATPase-deficient mutants of Escherichia coli have led to the
conclusion that one function of the ATPase is to catalyze the
synthesis of ATP during oxidative phosphorylation (1-5). A
second function of this enzvme, distinguished under anaerobic
conditions, is thought to be the coupling of ATP hydrolysis to
essential membrane events that require the expenditure of
metabolic energy. In the absence of respiration, ATPasenegative mutants cannot utilize ATP from substrate level
phosphorylations to drive the ATP-linked transhydrogenase
(6, 7)'or the accumulation of various metabolites (3, 5, 8).
Such anaerobic function of the ATPase is also suggested by the
effects of NN'-dicyclohexylcarbodiimide (DCCD), an inhibitor of this enzyme (9, 10). In E. coli, DCCD blocks both
the ATP-linked transhydrogenase (7) and the accumulation of
proline found under anaerobic conditions (11). DCCD also
inhibits active transport of metabolites in Streptococci, which
thesis. The alternative (anaerobic) function of the ATPase is
required when protons cannot be extruded by the respiratory
chain. Under these conditions, the ATPase couples the hydrolysis of ATP to the electrogenic movement of protons out
of the cell (Fig. 11B). The protonmotive force generated by
ATP hydrolysis is then utilized by energy-dependent reactions
such as the "ATP-linked" transhydrogenase, or the active
transport of metabolites.
Evidence in support of this anaerobic function of the
ATPase has been presented by Harold and his collaborators,
who have studied the anaerobe S. fecalis (faecium). They
showed that glycolyzing cells establish both a pH gradient
(interior alkaline) and a membrane potential (interior negative), and that DCCD inhibits the formation of each of these
components of the protonmotive force (17-19). More recently,
West and Mitchell, studying membrane vesicles from E. coli,
have shown that ATP hydrolysis is associated with the movement of protons across the membrane (20).
In microorganisms, the evidence in support of the chemiosmotic hypothesis remains incomplete without the direct
demonstration of ATP' synthesis driven 'by a protonmotive
force. The experiments reported here show that the membrane-bound ATPase catalyzes the synthesis of ATP when an
inward directed protonmotive force is imposed across the cell
membrane.
MATERIALS AND METHODS
Cultures of Streptococcus lactis (ATCC 7962) were grown to
early stationary phase, by described methods (21). Cells were
harvested by centrifugation, washed twice with 0..1 M sodium
phosphate (pH 6) unless otherwise indicated, and resuspended
in a small volume of this same buffer. Wild-type E. coli strain
1100 and its ATPase-negative derivative,- strain 72, were
obtained from T. H.- Yamamoto. Strains 1100 and 72 were
grown at 370 in medium 63 supplemented with 1% (w/v)
lack oxidative metabolism (9, 12, 13).
These observations are in agreement with predictions made
by the chemiosmotic. hypothesis of Mitchell (14, 15; for a review see ref. 16). According to this view, oxidation of substrates by the electron transport chain leads to the net transfer
of protons (H+) from the inside to the outside of the cell. This
extrusion of protons establishes a gradient of p1I (interior
alkaline) as well as a membrane potential (interior negative).
Mitchell has proposed that ATP synthesis during- oxidative
phosphorylation occurs when protons, moving down their
electrochemical' gradient, reenter the cell via the ATPase
(Fig. 1A). Thus, the electrochemical potential of protons (the
protonmotive force) provides the driving force for ATP syn-
A
B
ATP
ADP + Pi
PROTON ENTRY
PROTON EXTRUSION
(ANAEROBIC)
(AEROBIC)
FIG. 1. The ATPase of bacteria. (A) Proton entry coupled to
ATP synthesis occurs in aerobic organisms or in facultative
anaerobes (e.g., E. coli). (1B) Proton extrusion coupled to ATP
hydrolysis occurs in anaerobes (e.g., S. lactis) or in Jacultative
anaerobes.
Abbreviations: DCCD, NN'-dicyclohexylcarbodiimide; CCFP,
carbonylcyanidefluoromethoxyphenylhydrazone.
3896
Proc. Nat. Acad. Sci. USA 71
2E 3.0
|\
0.
B
IE-0
ATP Synthesis Driven by a Protonmotive Force
(1974)
3897
i
GLUCOSE
-
0.
0.8
j
CONTROL
2.0
0.6
-J
sdu0.4 Zp
l
u
z
0.2
VALINOMMINUE
+
DCCD
n9%etao
a
1
5~~~~~~~~~~
feter9%ehaoCr0.CDCD
00
4
5
FIG. 2.- Comparison of ATP levels in S. latit treated with
valinom'ycin 'or glucose. Cells were washed and resuspended in
0.1 M sodium phosphate (pH 6) and diluted with this same buffer
(final volume of 5 ml) to a cell density of 176 -Mett units. After
samples were removed for measurement of zero-time ATP levels,
either 0.1 ml of 1.25 M glucose (25 mM final concentration) or 5
;1of 10 mM valinomycin (101AuM final concentration) was added.
At the indicated titnes, aliquots were removed for the determination of intracellular ATP concentrations.
Difco-Bacto Tryptone, 0.5% (w/v) glucose, and 1 ;4g/ml of
thiamine.
All experimental procedures were done at 23°_ unless otherwise indicated. ATP was measured by use of firefly extract by
the procedure of Cole et al. (22). Cell density was determined
turbidimetrically with a Kle-tt-Summerson calorimeter (no. 42
filter). The intracellular concentration of ATP was calculated
from the known relationship between intracellular water
volume and cell density. For S. lactis, 1 ml of a cell suspension
of 100 Klett units is equivalent to 0.24 IAI of intracellular water
or 165 ug dry weight (23). The 'corresponding relationship for
E. coli is 0.6 IAI of cell water or 220 ;&g dry weight (24).
Firefly extract (FLE-50) wa's obtained from Sigma Chemical Co. Valinomycin was purchased from Calbiochem. Corp.,
and DCCD was obtained from Baker Chemical Co. Carbonylcyanidefluoromethoxyphenylhydrazone (CCFP) was a gift of
Dr. E. P. Kennedy. Valinomycin, DCCD, and CCFP were
added to cell suspensions as small volumes of stock solutions in
95%O ethanol; final ethanol concentrations were never more
than 0.2%.
~~~MINUTES3
FIG. 3. Effect of DCCD on ATP sy-nthesis in valinomycintreated S. lactig. Cells were washed and resuspended in 0.1 M
sodium phosphate (pH 8). To 0.6 ml of cells (13000 Klett units),
5 ,d of either 95% ethanol or 0.1 M. DCCD in 95% ethanol was
added (final concentration of DCCD was 0.83 mM). After 30
min the cells were centrifuged and resuspended in 0.6 ml of 0.1 M
sodium phosphate (pH 5). Fifteen minutes later they were
diluted with this same buffer to a final cell density of 260 Klett
units. After samples were removed for measurement of zero-time
ATP, each suspension (either DCCD- or ethanol-treated) was
divided into two portions. To one portion, valinomycin (10 pM
final concentration) was added and samples were removed at the
indicated times. (Inset) To the second portion, glucose was added
(25 mM final concentration) and samples were removed after
25 min for determination of ATP levels.
RESULTS
Valinomycin-Induced A TP Synthesis in Streptococcus lacti8.
An inward directed protonmotive force was artificially generated by treatment of cells with valinomycin. This ionophore
makes the cell membrane highly permeable to the potassium
ion, and the efflux of K+ establishes a membrane potential,
was attained after 1 min and represented a 19-fold increase
over the initial level (0.15 mM ATP). Within 5 min, ATP had
returned to about its original concentration. In this experiment the valinomycin-induced synthesis of ATP was compared to that found when cells were presented with glucose.
The ATP generated by glycolytic reactions attained a stable
level of about 2.4 mM ATP after an initial "overshoot." Thus,
the maximum level of ATP observed in valinomycin-treated
cells was comparable to the steady-state level of ATP found in
glycolyzing cells. Other experiments showed that the valinomycin-induced ATP synthesis was dependent on the pH of the
incubation medium. Between pH 4 and pH 6, the kinetics of
formation of ATP were as described in Fig. 2. At pH 7,
maximal levels of ATP were only 3-fold increased over the
basal value; at pH 8, no ATP synthesis was detected.
Inhibition of Valinomycin-I nduced ATP Synthesis in
Streptococcus lacti. To determine whether the valinomycininduced synthesis of ATP required the activity of the ATPase,
the effect of the inhibitor, DCCR, was studied (Fig. 3). In
cells previously exposed to DCCD, no ATP formation was
interior negative (25).
The basic observation is illustrated by the results presented
in Fig. 2. Cells from- the stationary phase of growth were
washed and resuspended in a potassiumn-free medium. After
they were sampled to determine the basal level of ATP,
valinomycin was added. The' addition of valinomycin resulted
in a rapid increase in the intracellular level of ATP, followed
by a somewhat slower decline. The peak level (2.8 mM ATP)
found after addition of valinomycin. However, DCCD-treated
cells retained the capacity to generate ATP from substrate
level phosphorylations. When incubated with glucose, both
control and DCCD-treated cells attained similar levels of ATP
(inset to Fig. 3).
The ratio of intracellular to extracellular potassium determines the size of the membrane potential present in valinomycin-treated cells. If ATP synthesis in such cells depends on
3898
Microbiology: Maloney et al.
Proc. Nat. Acad. Sci. USA 71
(1974)
0.3
PARENT (ATPase )
E
\
0.2
0
mM K
cc
-j
f
rI
2.0
1
0.1
mMK
A__
0
.MUTANT((A
z
0
1.0
0
1
2
3
4
5
MINUTES
FIG. 4. Effect of external potassium pd ATP synthesis in
valinomycin-treated S. lactis. Cells were washed and resuspended
in 0.1 M sodium phosphate (pH 6) and diluted to a final cell
density of 240 Klett units using this same buffer containing 1KC0
at the indicated concentrations. After samples were removed for
measurements of zero-time ATP levels, vallnomycin was added
(10 sM final concentration) and later samples were withdrawn
at the indicated times. No ATP synthesis was observed after
valinomycin was added to cells incubated with 30 mM or 100
mM KCI (not shown).
the size of the membrane potential, one would expect to find
inhibition of ATP formation when the membrane potential is
reduced by the addition of potassium to the external medium.
In the experiment shown in Fig. 4, ATP synthesis was
measured in cells incubated in media to which various amounts
of KC. had been added. With increasing concentrations of
KCI the maximal levels of ATP were lowered; and were
reached at progressively later times. At concentrations of
KCl of 10 mM or above, there was complete inhibition of ATP
synthesis. This effect was specific for the potassium ion, since
the addition of 100 mM NaCl had no significant effect on ATP
synthesis in this experiment (data not shown), and since, in
other experiments, a similar inhibition was observed when
potassium phosphate rather than potassium chloride was
used.
According to the chemiosmotic hypothesis, the synthesis of
ATP catalyzed by the ATPase occurs only when there is a
concomitant inovement of protons into the cell via this pathway (Fig. 1A). ATP synthesis should be blocked by conditions
that provide alternate routes for proton entry. In agreement
with this prediction, it was found that pretreatment of cells
with 0.5MuM CCFP, which renders bacterial membranes highly
permeable to protons (26), resulted in complete inhibition of
valinomycin-induced ATP formation (experimental conditions
as in Fig. 2). Pretreatment of cells with CCFP did not impair
their capacity to make ATP from glycolytic reactions.
Valinomycin-Induced A TP Synthesis in Escherichia coli.
Unlike S. lactis, washed cells of E. coli contain high levels of
2
se)
8
4
MINUTES
8
10
FiGo. 5. ATP synthesis in valinomycin-treated E. coli strains
1100 (parent) and 72 (ATPase-negative mutant). Exponentially
growing .cells'were harvested and wabd once with 0.12 M
TrisEHCi (pH 8) and resuspended in swmall volumie of this
buffer. After incubation for 10 min at 37°, ethylenediaminetetraacetic acid (0.05 mM final concentration) was added. Two
minutes later the cells were diluted 20-fold into the starvation
medium' suggested by Goldberg, Olden, and Prouty (27) in their
modification of the method of Koch (28). This medium (pH 7)
contained the following components at the indicated final concentrations: a-methylglucoside (20 mM), sodium arsenate (20
mM), NaCN (1 mM), Tris.HCl (30 mnM), NaCl (50 mM),
(NH4)2SO4 (20 mM), MgSO4 (7 mM), and KCl (300 mM).
Incubation time in this medium at 37° was 125 min for strain
1100, and 45 miiiin for strain 72. This tieathient reduced ATP
levels by more than 98% in both strains. After this starvation the
cells were returned to 230, washed once with 0.1 M sodium
phosphate (pH 8) tontaining 0.2 M KCI; then washed and resuspended in 0.2 M sodium phosphate (pH 8). For valinomycin
treatment, the cells were diluted 20-fold into 0.2 M sodium
phosphate (pH 5), and valinomycin (10 uM final concentration)
was added 10 sec later. Samples were removed for determination
of ATP levels at the indicated times after the addition of vainomycin.
ATP (about 3 mM). Therefore, before the addition of valinomycin, cells were subjected to a starvation procedure, which
lowered basal levels of ATP and eliminated endogenous
reserves of "energy"-yielding materials. In. starved cells
(Fig. 5) valinomycin-induced synthesis of ATP was observed
in wild-type strain 1100, but not in its derivative, strain 72,
which lapks ATPase activity. However, when these same cells
were incubated with 10 mM glucose in the presence of 1 mM
NaCN at pH 7, ATP levels attained after 20 min were similar
for the two strains (0.11 mM ATP, and 0.08 mM ATP,
respectively, for strains 1100 and 72). It should be noted that
in this experiment an attempt was made to maximize the
inward directed protonmotive force by shifting cells from pH 8
to pH 5at the time of valinomycintreatment (see legend to Fig.
5). A pH shift alohe (no valinomycin present) did not result in
ATP synthesis.
DISCUSSION
In a number of instances it has been possible to demonstrate
the synthesis of ATP coupled to the movement of a specific
cation down its electrochemical gradient. The appropriate
cation gradients have been shown to reverse both Na+,K+-
Proc. Nat. A cad. Sci. USA 71 (197-4)
ATPase activity in the erythrocyte (29) and Ca++-ATPase
activity in membrane vesicles of the sarcoplasmic reticulum
(30). Reid, Moyle, and Mitchell (31) were the first to show
that mitochondrial ATP synthesis may be driven by a protonmotive force. They imposed a large chemical gradient for
protons by suddenly immersing mitochondria in an acid
medium. Subsequently, it was shown by others (32, 33) that in
valinomycin-treated mitochondria, ATP synthesis may be
coupled to the efflux of potassium. As discussed by G'ynn
(25), these observations are compatible with the idea that R.uch
AT1P synthesis is driven by an electrical potential resulting
from potassium efflux mediated by valinomycin. Using an
artificial system, Racker and Stoeckenius (34) have found that
ATP synthesis catalyzed by mitochondrial ATPase may be
coupled to light-induced proton movements. Gradients of pH
alone elicit ATP formation in chloroplasts (35), and, as in
mitochondria (36), yields of ATP are increased when a pH
gradient and a membrane potential complement one another
(37).
The work reported here extends such observations to
bacterial systems. ATP synthesis could be demonstrated in
cells exposed to an inward directed protonmotive force. The
experimental plan was to generate a protonmotive force whose
major component was a membrane potential, by- treating cells
with the potassium ionophore, valinomycin. Under these
conditions, the cell membrane is more permeable to K+ than
to other ions, and the size of the membrane potential is
determined primarily by the ratio of intracellular to extracellular potassium. It is shown here that ATP synthesis occurs
after the addition of valinomycin to bacterial cells suspended
in a potassium-free medium. ATP formation was reduced in
the presence of added potassium, indicating that the yield of
ATP was governed by the size of the membrane potential. This
suggests a partial explanation for the finding that valinomycin-treated cells show only a transient increase in ATP
levels. In such experiments there is a rapid loss of internal
potassium (21), which would reduce the membrane potential
and lower the protonmotive force. This, in turn, would depress
ATP levels, since these are determined by the balance between
synthetic and degradative reactions. The ATPase itself may
contribute to net ATP hydrolysis when intracellular ATP is
high and the protonmotive force is reduced. The valinomycininduced ATP synthesis in S. lactis gave maximal yields of
ATP equivalent to an intracellular concentration of about 3
mm, whereas in starved cells of E. coli, maximal yields of 0.3
mM ATP were obtained. The reason for the lower yields in E.
coli is not clear, but it seems likely that the starvation procedure introduced limitations on the capacity of cells to form
ATP.
Valinomycin-induced ATP synthesis was not observed in
S. lactis pretreated with the ATPase inhibitor, DCCD, nor was
it found in the ATPase-negative mutant of E. coli. We believe
that these results, considered together with the other observations reported here, provide strong evidence in support of the
view that the membrane-bound ATPase catalyzes the synthesis of ATP in response to an inward directed protonmotive
force. The ATP synthesis observed in E. coli reflects the activity of the ATPase found during aerobic growth (Fig. lA). For
S. lactis, however, such ATP synthesis represents a reversal of
the normal function of this enzyme, which is to couple the
hydrolysis of ATP to the extrusion of protons (Fig. 1B).
Data from preliminary experiments allow an estimate of the
size of the protonmotive force required to drive ATP synthesis
ATP Synthesis Driven by a Protonmotive Force
3899
in S. tactis. Under the conditions described for the experiment
in Fig. 2, intracellular potassium was about 300 mM while
extracellular potassium was about 0.15 mM, giving a potassium ratio IN/OUT of roughly 2000. Assuming that cells
treated with valinomycin are far more permeable to K+ than
to other ions, one may use the Nernst equation to calculate
that this potassium distribution would generate a diffusion
potential of 195 mV (interior negative). An additional protonmotive force was present in the form of a chemical gradient of
protons, since measurements have indicated that the internal
pH was about 0.5 pH units more alkaline than that of the
medium in such experiments. Thus, the initial protonmotive
force (electrical + chemical components) would have a
minimum value of about 200 mV. Comparable data are not yet
available for ATP synthesis in E. coli. These approximate
calculations are in agreement with requirements of the chemiosmotic hypothesis. Mitchell (15) has stated that a protonmotive force of 210 mV is needed to maintain an ATP/ADP
ratio of 1.
We thank Dr. Raymond C. Valentine for providing us with
detailed information on the properties of the ATPase-negative
mutant, strain 72. We also thank 1)r. Kenneth Olden for providing us with his procedure for depleting ATP levels in E. coli.
This work was supported in part by a grant from the U.S.
Public Health Service (AM-05736). P.C.M. was the recipient of
a National Institutes of Health Postdoctoral Fellowship (5-F02-
GM52320).
1.
2.
3.
4.
5.
6.
7.
Butlin, J. D., Cox, G. B. & Gibson, F. (1971) Biochem. J.
124, 75-81.
Kanner, B. I. & Gutnick, 1). L. (1972) J. Bacteriol. 111,
287-289.
Schairer, H. U. & Haddock, B. A. (1972) Biochemn. Biophys.
Res. Commun. 48, 544-551.
Simoni, R. D. & Shallenberger, M. K. (1972) Proc. Nat.
Acad. Sci. USA 69, 2663-2667.
Yamamoto, T. H., Mevel-Ninio, M. & Valentine, It. C.
(1973) Biochim. Biophys. Ada 314, 267-275.
Cox, G. B., Newton, N. A., Butlin, J. D. & Gibson, F.
(1971) Biochem. J. 125, 489-493.
Kanner, B. I. & Gutnick, 1). L. (1972) FEBS Lett. 22,
197-199.
Berger, E. A. (1973) Proc. Nat. Acad. Sci. USA 70, 15141518.
9. Harold, F. MI., Baarda, J. I1., Baron, C. & Abrams, A.
(1969) J. Biol. (Chem. 244, 2261-2268.
10. Evans, D. J. (1970) J. Bacteriol. 104, 1203-1212.
11. Klein, W. I.. & Boyer, P. 1). (1972) J. Biol. Chem. 247,
8.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
7257-7265.
Asghar, S. S., Levin, E. & Harold, F. M. (1973) J. Riot.
Chem. 248, 5225-5233.
Kashket, E. R. & Wilson, T. H. (1972) Biochem. Biophys.
Res. Commun. 49, 615-620.
Mitchell, P. (1961) Nature 191, 144-148.
Mitchell, P. (1966) Biol. Rev. (Cambridge) 41, 445-502.
Harold, F. AM. (1972) Bacteriol. Rev. 36, 172-230.
Harold, F. M., Pavlasova, E. & Baarda, J. 1. (1968) Biochim. Biophys. Acta 196, 235-244.
Harold, F. M. & Papineau, D. (1972) J. Mfembrane Riot. 8,
27-44.
Harold, F. M. & Papineau, D. (1972) J. Membrane Biol. 8,
45-62.
West, I. C. & Mitchell, P. (1974) FEBS Lett. 40, 1-4.
Kashket, E. R. & Wilson, T. H. (1973) Proc. Nat. Acad. Sci.
USA 70, 2866-2869.
Cole, H. A., Wimpenny, J. W. T. & Hughes, D. E. (1967)
Biochim. Biophys. Acta 143, 445-453.
Kashket, E. R. & Wilson, T. H. (1972) J. Bacteriol. 109,
784-789.
Winkler, H. H. & Wilson, T. 11. (1966) J. Biot. Chem. 241,
2200-2211.
25. Glynn, I. M. (1967) Nature 216, 1318-1319.
24.
3900
26.
27.
Microbiology:
Maloney
Proc. Nat. Acad. Sci. USA 71 (1974)
et al.
Scholes, P. & Mitchell, P. (1970) J. Bioenerg. 1, 61-72.
Goldberg, A. L., Olden, K. & Prouty, W. F. (1974) in
Intracellular Protein Turnover, eds. Schimke, R. T.
Katunuma, N. (Academic Press, New York), in press.
&
28. Koch, A. L. (1971) J. Mol. Biol. 59, 447-459.
29. Garrahan, P. J. & Glynn, I. M. (1966) Nature 211, 14141415.
30. Makinose, M. & Hasselbach, W. (1971) FEBS Lett. 12,
271-272.
31. Reid, R. A., Moyle, J. & Mitchell, P. (1966) Nature 212,
257-258.
32. Cockrell, R. S., Harris, E. J. & Pressman, B. C. (1967)
Nature 215, 1487-1488.
33. Rossi, E. & Azzone, G. F. (1970) Eur. J. Biochem. 12,
319-327.
34. Racker, E. & Stoeckenius, W. (1974) J. Biol. Chem. 249,
662-663.
35. Jagendorf, A. T. & Uribe, E. (1966) Proc. Nat. Acad. Sci.
USA 55, 170-177.
36. Thayer, W. S. & Hinkle, P. C. (1973) Fed. Proc. 32, 669abstr.
37. Schuldiner, S., Rottenberg, H. & Avron, M. (1972) FEBS
Lett. 28, 173-176.