Published October 1, 1964
R O L E OF E D T A A N D M E T A L S
IN MITOCHONDRIAL
CONTRACTION
W I L L I A M S. L Y N N , JR., M . D . , S Y D N E Y
R O S E H. B R O W N
FORTNEY,
M.D., and
From the Departments of Biochemistry and Medicine, Duke University Medical Center, Durham,
North Carolina
ABSTRACT
Vasington et al. (I), Brierley et al. (2), Rossi et al.
(3), Lehninger et al. (4, 5), and Sallis et al. (6)
have recently demonstrated that mitochondria are
remarkably active in energetically concentrating
Mg ++ and Ca++ phosphates. The energy source
for both of these reactions can be supplied by
ATP. Mitochondria, likewise, are very active in
transporting water (6, 7); i.e., contraction (dehydration) of mitochondria requires chemical
energy (either ATP or electron transport), and
swelling of mitoehondria occurs under conditions
of energy deficiency. For example, when Ca++ or
Mg++ or fatty acids, all of which are substrates
which require the utilization of ATP, are added
to mitochondria, ATP disappears, and swelling
promptly ensues. A preliminary communication of
these data has been published (8). This swelling
can be reversed by addition of sufficient energy in
the form of ATP. Somewhat paradoxically, addition of electron-transport substrates to mitochondria at 23 ° also results in swelling (5), as well
as in a decrease in ATP concentration (8). Data
from this laboratory have indicated that addition
of EDTA to substrate-swollen mitochondria
induces prompt contraction (8, 9), and addition
of EDTA plus serum albumin causes contraction
of the mitochondria back to their original state.
ATP concentration increases after addition of
EDTA, and Ca ++ and free fatty acids are also
removed from the mitochondria.
It is the purpose of this paper to delineate the
competitive aspects of the various energy-using
processes (ATPases) in mitochondria, especially
those related to shrinking and swelling, and by
Downloaded from on June 17, 2017
Studies comparing the state of hydration and dehydration of rat liver mitochondria to
their content of ATP, Ca, and fatty acid, along with the rate of ATP hydrolysis, as well as
microscopic appearance of mitochondria, have led to the following generalizations: 1. The
competition between cationic translocations and water translocation for the available
chemical energy (ATP) determines under many circumstances the water content of mitochondria. 2. Swelling of mitochondria by electron transport substrates is an example of
the activation of the cationic translocations at the expense of water translocation. 3. Electron
micrographic studies are interpreted to indicate that EDTA alone can cause condensation
and dehydration of the mitochondrial matrix. However, both EDTA and substrate are
necessary to remove appreciable quantities of water from mitochondrial intramembranous
spaces. 4. Since the data in the accompanying report indicated that EDTA, in the absence
of energy, decreased the permeability of mitochondrial membranes, it appears likely that
ballooning of intramembranous spaces, following addition of EDTA, represents trapping of
water between two semipermeable membranes following dehydration of mitochondrial
matrix.
Published October 1, 1964
m i c r o s c o p i c studies to d e m o n s t r a t e the g e o g r a p h i c
roles of E D T A , metals, a n d e n e r g y on the d e h y d r a t i o n of m i t o c h o n d r i a .
the sonicated (150 kc for 30 seconds) mitochondrial
pellets, after removal of the T C A by ether extraction,
using hexokinase and glucose-6-phosphate dehydrogenase.
Fatty acids were measured on chloroform-methanol
(1/1 by volume) extracts, using the m e t h o d of Dole
(12).
Ca ++ was measured in buffer and in pellets (the
pellets were digested in boiling concentrated nitric
acid for 10 minutes and then diluted) by flame
photometry using the Zeiss photometer.
INCUBATIONS : M i t o c h o n d r i a were stored in
0.25 M Sucrose at 5 ° before use. Aliquots were then
removed and incubated in 0.125 M KC1 containing
0.03 M Tris, p H 7.4.
Electron micrographs of the mitochondrial pellets
METHODS
M i t o c h o n d r i a were p r e p a r e d from livers obtained
from 200-gram Osborne-Mendel rats, according to
the procedure of H o g e b o o m et al. (10), using 0.25 M
sucrose and 0.02 M Tris (hydroxymethylaminomethane), p H 7.4.
Hydration and dehydration of mitochondria were
followed turbidimetrically a n d / o r gravimetrically,
using the m e t h o d of Amoore and Bartley (11) and
Lynn et al. (9).
A T P was measured in 5 per cent T C A extracts of
0.7-
0.65
Downloaded from on June 17, 2017
0.6
W
o
z
(D
f13
ATP
0.55
0.65.
ED~
Anfimycin
0.6
j}
0.55-
r:
l
0.5"
SUCCINAZ
0.47'
I
I
5
l0
ED TA
15
I
I
20
25
50
I
I
I
35
40
45
MIN.
F m u n E 1 Contraction of mitoehondria by suecinate, EDTA, and ATP. Rat liver mitochondria were
isolated from white rats, weighing 200 grams, according to the procedure of Hogeboom et al. (10). The
mitochondria were washed once in 0.25 M sucrose. Incubations were done in 2.5 ml of 0.09 M KCI, 0.02
M Tris, p H 7.8. Swelling was measured as absorbanee at 520 mu. Oligomycin, 8 t~g/ml, and antimycin A,
2 t~g/ml, were added at 0 time. At 15 minutes, ATP, 2 X 10-3 M, and Tris sueeinate, 2 X 104 u, were
added. At 80 minutes, EDTA, 2 X 10-4 M, was added. Similar experiments, using larger amounts of
mitochondria in a larger volume, were also performed, and the weights of the mitoehondrial pellets were
obtained at the end of the incubations. In all experiments, the gravimetrie measurements checked within
10 per cent the absorbance measurements. - . . . . .
E D T A a d d i t i o n . - - - - = no addition of EDTA.
All incubations were done at 28%
I0
TIlE $OURNAh OF CELL BIOLOGY • VOLIYME 23, 1964
Published October 1, 1964
were prepared after incubation. The pellets were
fixed immediately, after centrifugation, in osmium
tetroxide buffered at p H 7.4 with veronal acetate for
1 hour in the cold. They were dehydrated in ethanol,
embedded in Araldite, and sectioned at about 800 A,
TABLE I
RESULTS
ATP, Ca ++, and Fatty Content of Mitochondria
Additions
ATP
Ca++ Freefatty
content bound acid
content
.ug
/zrnoles
13.5
46
0.32
2. Control
10.1
33
0.56
7.2
39
0.96
4. EDTA 5 X 10-4 M
12.5
21
0.68
5. EDTA~ 5 X 10-4 M +
Succinate 2 X 10 -3
12.1
27
0.65
6. Ca ++ 10-8 ~l
4.9
--
1.25
7. Mg3(PO4)~, 10-8 u
7.5
27
0.82
8. ATP 10-a M
--
40
0.24
9. A T P 10-8 ~! d- Suecinate 2 X 10-a M
--
56
0.14
M
* All contents expressed as total amount recovered
in the mitochondrial pellet (7.8 mg protein).
EDTA was added at the end of 10 min., and this
incubation was allowed to proceed for 2 more min.
(Contraction was complete at the end of this time,
and the turbidity reading of this tube had returned
to the control value.)
Mitochondria were incubated, as in Fig. 1, for 10
minutes at 23 °. A T P was measured in trichloroacetic acid (5 per cent) extracts of the total incubation (TCA was removed by extraction with ether),
enzymatically. Ca ++ was measured in nitric acid
digests of the dried pellets obtained by centrifugation at 15,000 g for 10 minutes. Free fatty acids were
also measured in chloroform-methanol (pH 2) extracts of the Initochondrial pellets (Dole procedure). All experiments were done in triplicate
and repeated 3 times. Similar results were obtained
in each experiment. All additions were made at 0
time, except in incubation No. 5. In this one, EDTA
was added at the end of 10 minutes, and the incubation was allowed to proceed for 2 more minutes.
I. Contraction of Mitochondria by Substrate-EDTA
T h e d a t a in Fig. 1 indicate t h a t succinate
causes swelling of fresh mitochondria, a n d this
swelling is completely reversed by addition of
E D T A . These effects can be observed with m a n y
substrates; i.e., b-hydroxybutyrate, glutamate, or
malate. Antimycin, a known inhibitor of electrontransport, blocks both swelling and contraction.
However, oligomycin, a n inhibitor of the transphosphorylation reaction involving A T P (2, 13),
inhibits A T P contraction, but has no effect on
substrate-EDTA contraction. A l b u m i n is not
necessary for substrate-EDTA contraction in
freshly p r e p a r e d mitochondria. However, if the
mitochondria are aged at 5 °, or if allowed to
incubate in the presence of succinate for longer
t h a n 30 minutes at 23 °, it becomes necessary to
add both E D T A and serum albumin to obtain
p r o m p t and maximal contraction. It is well known
that fatty acids, which are potent swellers of mitochondria, are liberated from aged mitochondria
(5, 14) or from mitochondria during the swelling
phase (Table I) and that these fatty acids can be
removed from mitochondria by serum albumin.
Thus, it is not surprising t h a t albumin potentates
the contractile effects of E D T A on aged mitochondria.
II.
Effect of E D T A on Mitochondrial Content of A T P , C a ++, Fatty Acids, and on
A TPase Activity
Previous d a t a (8) have shown that addition of
substrate alone causes a loss of endogenous ATP.
Addition of E D T A prevents loss of ATP. Addition
of EDTA, after the succinate induced loss of ATP,
restores the concentration of A T P to the control
level. T h a t the utilization of exogenous A T P
(ATPase) is also decreased by E D T A and albumin, as well as oligomycin, is demonstrated in
Table II. Also, both Ca ++ and palmitic acid
increase the hydrolysis of ATP. It has been
demonstrated previously that fatty acids are
W. S. Lrz~N, JR., S. FOaTNEY, AnD R. H. BnowN
Roleof ED TA and Metals
11
Downloaded from on June 17, 2017
/zg*
1. 0 Time
3. Succinate 2 X 10-3 M
sections were stained with uranyl acetate. Multiple
sections were made of each pellet. Electron micrographs were made with a Phillips EM-100B by Mr.
J. M. Price in the laboratory of Dr. Montrose Moses,
Department of Anatomy, Duke University Medical
Center.
Published October 1, 1964
rapidly reesterified to mitochondrial phospholipids (5), and that Ca ++ is concentrated by mitochondria (1), both at the expense of ATP. These
observations have led to the hypothesis that
ATPase was activated in swollen mitochondria.
However, as is shown in Table II, the actual size
of mitochondria has little or no relation to the
a m o u n t of A T P hydrolyzed in the presence of
EDTA. As is shown in Table II, mitochondria,
previously swollen to varying degrees by several
known swelling agents, were all hydrolyzing
exogenous A T P in the presence of E D T A at equal
rates, despite the fact that some of the mito-
chondria contracted and some did not. Also, both
oligomycin and albumin decreased A T P hydrolysis, and this effect was additive to that of EDTA.
Since M g ++ was present in excess of E D T A in all
incubations, E D T A could not be blocking ATPase
activity by removal of M g ++.
I I I . Role of Ca ++ and Mg ++ in Mitochon-
drial Movement of Water
T h e data in Fig. 2 indicate that M g ++ in high
concentration will also block contraction caused
by EDTA. However, if the mitochondria are supplied with outside ATP, contraction results. Both
T A B L E II
Mitochondrial Size and A TPase Activity
0.125 M KC1
0 to 10 min.
Initial
mitochondrial
size
POT .umoles
PO~ .umoles
--EDTA + E D T A
- - E D T A +EDTA
AOD X 100"
None
-- 2
0.54
0.30
3.66
1.91
--21
0.89
0.29
6.61
1.71
Cysteine 2 X 10-4 M
--8
0.49
0.31
3.62
1.49
Albumin 1 m g / m l
--2
0.32
0.26
1.35
0.91
H2Oq3.02 M Tris
--13
0.60
0.35
3.62
2.26
Oligomycin 5 # g / m l
--2
0.39
0.14
1.31
0.92
Gramicidin 1 # g / m l
--18
0.95
0.31
6.61
1.31
0
0.51
0.29
3.01
1.95
--29
0.99
0.35
6.6
2.66
Ca +J- 2 X 10-4 M
Antimycin 3 # g / m l
Palmitic acid 5 X 10-5 ~
ATP, 6 pmoles, added with and without EDTA (2 X 10-4 M) at 0 min.
* &OD = change in absorbance.
All experiments were performed in 0.125 M KC1-0.02 M Tris pH 7.3 containing Mg ++,
5 X 10-4 M. Enough mitochondria were added to 2.5 ml of buffer to give an initial
absorbance of 0.8 at 520 m#. The mitochondria were then incubated at 25 ° in the
presence of the various swelling agents for 10 minutes. ATP, 6 #moles, with and without EDTA (2 X 10-4 M), was then added to each tube. Swelling is expressed as the
change in absorbance at the time period indicated. Total inorganic phosphate was
measured (Gomori, 1942) by placing aliquots of the mixture in 10 per cent trichloroacetic acid at the stated time intervals. Inorganic phosphate is expressed as total
phosphate in the reaction vessel. Each number in the table represents the average of
one experiment done in triplicate. No more than a 15 per cent variation was observed
in any of the triplicate determinations.
12
THE JOURNAL OF CELL BIOLOGY " VOLUME ~8, 1964
Downloaded from on June 17, 2017
Additions
10 to 60 min.
Published October 1, 1964
0.75
o.e0
0.7
"-,-IW#
Albuminf
~ I I ~ "I"
o.so- k
:r
c
2
It
I
< 0.6
Ll
Succinote+
i.++
"~,
~v,
uccin ,"\',
/
l
/I
I
if
i !
0.40-
_
~l
~
CA++ 5 XlO-~'M
/
I
!
i
it
i!ii
0.33
0.30
,
Io
2'o
30
4'0
$ $ 6'o
Minutes
O.~
I
[~) [:ST 20I TV
310 31~ 410
Succinate
EDTA ATPadded
-+.MgAdded A d d e d +and-Albumin
~l
FIGURE ~ Effect of Mg++ on mitochondrial contraction. Mitochondria were incubated as in Fig. 1. Concen-
M g ++ a n d C a H- cause A T P c o n c e n t r a t i o n to decrease, as well as cause the content of fatty acids
to decrease (Table I a n d II). T h a t Ca++ is removed from m i t o c h o n d r i a by E D T A is shown in
T a b l e I. M a x i m u m b i n d i n g of Ca ++ requires b o t h
succinate a n d A T P (1). M g ++ b i n d i n g is also
k n o w n to require large a m o u n t s of energy ( O / P i
ratio of greater t h a n 2) in the form of substrate
or A T P (2, 6). Thus, b o t h Ca++ a n d M g ++ compete for the available high energy intermediate or
intermediates w h i c h cause w a t e r m o v e m e n t in
mitochondria. T h e removal of Ca++ by E D T A
allows the energy source, as indicated by A T P
concentration, to be m a i n t a i n e d , a n d thus to be
more available for w a t e r movement. Esterification
of the released fatty acids d u r i n g swelling also
competes with the m o v e m e n t of w a t e r for the
available chemical energy.
T h e d a t a in Fig. 3 indicate that, even in the
presence of excess A T P , c o n t r a c t i o n c a n n o t occur
unless excess Ca ++ is r e m o v e d by E D T A . Fatty
acids or other detergents likewise block contraction, regardless of the availability of A T P (5, 7).
Thus, b o t h Ca++ a n d fatty acids do compete with
other energy-requiring processes in mitochondria,
b u t also in larger quantities inhibit contraction by
some m e c h a n i s m not related to the availability of
A T P . As d e m o n s t r a t e d in the previous report (9),
Ca ++ or fatty acids, in the presence or absence of
energy, abolish the semipermeable property of
m i t o c h o n d r i a l membranes. I t was also d e m o n strated t h a t either osmotic or metabolic contraction require t h a t mitochondrial m e m b r a n e s be
m a i n t a i n e d in a semipermeable state.
I V . Electron Micrograph Studies of Shrink-
ing and Swelling in Mitochondria
M i t o c h o n d r i a , as isolated from liver b y the
S c h n e i d c r - H o g e b o o m procedure, in 0.25 M sucrose w i t h o u t E D T A , are usually heavily cont a m i n a t e d with endoplasmic reticulum as well as
ribonucleic acid particles (15, 16). T h a t this is
true of the m i t o c h o n d r i a used in these experiments
is d e m o n s t r a t e d in Fig. 4 a. However, the isolated
m i t o c h o n d r i a a p p e a r morphologically homogeneous. After i n c u b a t i o n of these m i t o c h o n d r i a with
a n d w i t h o u t succinate, succinate produced gravimetrically measured swelling, a n d the microscopic
W. S. LYNN, JR., S. FORTNEY,AND R. H. BROWN Role of E D T A and Metals
13
Downloaded from on June 17, 2017
trations of additions were as follows: Tris-succinatc,
X 10-3 M; Mg++CI2, 10 3 M; A T P , ~ X 10-4 M;
albumin, 1 mg/ml, Succinate and Mg ++ were added
at 5 minutes; EDTA, ~ X 10 4 M, at 15 minutes;
and ATP and albumin at ~5 minutes. Change in turbidity was followed as in Fig. 1.
FIGURE ~ Reversibility of mitochondrial contraction by
EDTA and Ca++. Mitochondria were incubated as in
Fig. 1. Additions were added as follows. Mitochondria
were incubated with and without Ca ++ for 30 minutes.
At 30 minutes, EDTA ~ X 10-4 M, ATP 5 X 10-4 M,
serum albumin 1 mg/m], MgC12 5 X 10-4 M, were
added to the Ca ++ incubation. At 45 minutes, Ca ++,
10-3 M, was added. At 50 minutes, ATP, 10-4 M, was
added and again at 55 minutes. Absorbance was
plotted along the ordinate, and time was plotted along
the abscissa. The arrows indicate time of above additions.
Published October 1, 1964
vacuoles are not present (Fig. 6). Although in the
presence of A T P , Mg ++, and albumin, dehydration, as measured turbidimetrically or gravimetrically, does occur, there is very little difference
in the morphological appearance of the mitochondrial pellet between control and contracted
mitochondria (Figs. 4 a and 6). A T P and M g ++
must, therefore, have caused equal dehydration
of both the matrix and the intramembranous
spaces, in contrast to E D T A which causes dehydration of matrix but hydration and vesicle
formation of the intramembranous spaces.
It is also apparent in these micrographs that
E D T A caused the dissolution or disappearance of
the ribonucleic acid particles, but had no effect on
the micrographic appearance of the membranes
of the endoplasmic reticulum. This observation
was confirmed by measuring the large amounts of
R N A (orcinol reaction and 260 m # absorbance)
released into the soluble supernatant buffer after
centrifugation of the EDTA-treated mitochondria.
The dissolution of ribonucleic acid particles by
E D T A has been previously observed (18).
Although no quantitative counts were made of
the numbers of the variously shaped mitochondria
present in each pellet, several sections were made
throughout each pellet, both from the top, bottom,
and middle of the pellet, and the sections shown
in Figs. 4 to 6 are quite representative of the
entire pellet.
V. Reversibility of Mitochondrial Shrinking
and Swelling
The data in Figs. 2 and 3 indicate that, in the
presence of small amounts of A T P (2 X l0 -4 M)
and EDTA, mitochondria quickly contract. A T P
at this concentration, without E D T A , will not
produce contraction (3). The data in Fig. 3 also
indicate that contraction of mitochondria can be
shown to be readily and repeatedly reversible, in
the presence of ATP, by sequential additions of
FIGURE 4 a to d Electron micrographs of contracted and swollen mitachondria. Mitochondria were incubated, as in Fig. 1, for 10 minutes and then centrifuged at 15,000 g for
10 minutes. The pellets were then immediately fixed in buffered osmium tetroxide for 1
hour. Sections were stained with uranyl acetate. Succinate, 10-3 ~, was added at 0 time,
and EDTA at 5 minutes. For Fig. 4 a to d, freshly prepared mitochondria were used. Final
magnification for all micrographs was 55,000. Fig. 4 d was obtained from a pellet which had
been incubated with succinate, 10-2 M, for 5 minutes, then EDTA, ~ )< 10-~ M, was added
for the final 5 minutes. Magnification marks indicate 1 micron.
14
T H E ,~OURNAL OF CELl, BIOLOGY • VOLUME ~3, 1964
Downloaded from on June 17, 2017
appearance of the mitochondria was altered
(Fig. 4 b). The matrix is less dense and appears
swollen. The intracristal spaces appear unchanged.
Incubation with E D T A alone, which does not
cause turbidimetric or gravimetric change, produces a striking morphological change. The
matrix becomes dense (dehydrated), but the
intracristal spaces distend and become vacuolated
(Fig. 4 c). Addition of succinate and E D T A produces apparently even denser matrix and an
apparent decrease in size of the intracristal spaces.
However, these spaces are still somewhat vacuolated (Fig. 4 d). The space between the two outside
membranes remains unchanged. If, on the other
hand, mitochondria are allowed to age at 5 ° for
3 days in 0.25 M sucrose, the mitochondrial matrix
becomes dispersed, apparently fragmented, and
full of non-staining material, presumably water
(Figs. 5 a and b). As measured turbidimetrically,
these mitochondria have swollen to twice their
control weight. Addition of E D T A to these mitochondria, which does not alter water content,
also produces striking morphological changes. The
matrix becomes dense, intracristal spaces distend,
and, in addition, large vacuoles between the outside membranes (i.e., in the perimitochondrial
space) appear (Fig. 5 c). The cristae and mitochondrial matrix occupy a small portion of the
intramitochondrial space and seem to be compressed into one side of the mitochondria. T a n gential sectioning of these large vacuolated mitochondria accounts for the crescent-shaped and
doughnut-shaped appearance of aged mitochondria. Thus, E D T A can, depending on the age
of the mitochondria, cause swelling of both the
intracristal space as well as the perimitochondrial
space, but in both instances an increase in density
of the matrix occurs. Although these aged mitochondria have lost most of their ability to be contracted by suecinate-EDTA, they will contract if
M g ++, albumin, and A T P are added (17). U n d e r
the latter condition, the large intramembranous
Published October 1, 1964
Downloaded from on June 17, 2017
Published October 1, 1964
E D T A vs. Ca ++. These data strongly imply that
E D T A is causing contraction by binding Ca ++.
Lehninger (5) previously has suggested that E D T A
may be exerting its effect on mitochondria by
binding to mitochondrial membranes. T h a t this
is probably not true is shown by the fact that the
E D T A - C 14 space, as measured by the method of
Amoore and Bartley (I1), is only 15 per cent of
the total water. This is in good accord with the
data of Werkheiser and Bartley (19), using polyglucose, and Jackson and Pace (20), using labeled
albumin, who demonstrated that the extramitochondrial space is about 18 per cent. Furthermore,
the data in Table I show that Ca++ is actually
removed from the mitochondrial preparations
by E D T A .
DISCUSSION
I~GURE 5 a to d Electron micrographs of aged mitochondria. The mitochondria were aged
for 3 days at 5° in 0.25 M sucrose before incubation. Final magnification for Figs. 5 a and 5
c was 18,000, and Fig. 5 b, 55,000. Figs. 5 a and b were obtained from pellets which had
been incubated without any additions; Fig. 5 c was obtained after incubation with EDTA,
X 10-4 M. Fig. 5 d is a higher magnification of 5 c (X 55,000). Magnification marks
indicate 1 micron.
16
THE JOURNAL OF CELL BIOLOGY • VOLUME23, 1964
Downloaded from on June 17, 2017
These data show that processes which use chemical energy in mitochondria, i.e., fatty acid esterification and M g ++ or Ca ++ translocation, can effectively compete with water translocation for the
available energy. Reduction of the electrontransport carriers by substrate in Ca++-laden
mitochondria, but not in Ca++-dcpleted mitochondria, causes the loss of ATP, and presumably
other high energy phosphate carriers (13), as well
as an increase in fatty acid content and an uptake
of water. After the removal of Ca ++ by E D T A ,
and fatty acids by albumin, the concentration of
A T P is quickly restored if electron-transport substrates are present, fatty acids are reesterified, and
contraction occurs. It is not clear from these data
what the sequence of events is, nor why reduction
of the electron-transport system in the presence of
Ca ++ or Mg++ causes the loss of available chemical
energy. It is clear, however, that this loss of A T P
requires either M g ++, Ca++, or fatty acids, and
that if these three energy-using agents are re
moved, electron-transport energy can maintain
and increase the concentration of A T P to a level
sufficient to affect water translocation.
The chemical mechanisms involved in translocations of water are not known. However, the
electron microscopic studies imply that at least
two processes are involved: one, the condensation
of the mitochondrial matrix which requires only
the removal of Ca++; and two, the removal of
water from mitochondrial vacuoles and intramembranous spaces, which requires chemical
energy. Tedeschi (21) and Lynn et al. (9) have
previously demonstrated
that mitochondrial
membranes which will contract are also semipermeable and do obey osmotic law. It is, however, highly unlikely that metabolic contraction is
the result of any energetically driven ion movement, since contraction readily occurs in solutions
which contain very small amounts of Tris as the
only cation. The energetic removal of water from
the intramembranous spaces may involve some
type of energetic binding between the inner and
outer membranes. Nothing is known, however,
about the forces which normally hold these two
membranes together.
It is known that the phospholipids of mitochondria are rapidly turning over, both in terms
of fatty acids (5) and glycerophosphate (22), and
also that phospholipids are required for mitochondrial contraction of aged mitochondria (23).
These two facts make it attractive to assume that
the chemical energy which is required to dehydrate the perimitochondrial and intracristal
spaces is needed to reesterify the phospholipids in
such a way that hydrophobic bonds can be established between the membranes, with resultant
exclusion of water. However, since M g ++ as well
as phospholipids is necessary for mitochondrial
contraction (17, 22), it may be that the magnesium salts of the acidic phospholipids are what
is necessary for contraction. Subsequent studies on
the role of M g ++ in contraction will be the subject
of another report.
Ohnishi has reported (24) that mitochondria
contain a contractile protein, and he suggested
that contraction by this protein may account for
translocation of water, i.e., like the squeezing of a
sponge. This may indeed be one factor in contraction of mitochondria. The micrographs (Figs. 4
Published October 1, 1964
Downloaded from on June 17, 2017
Published October 1, 1964
and 5) indicate that E D T A does cause dehydration and contraction of mitochondrial matrix.
However, no net efllux of water from the mitochondria occurs. Water does seem to be translocated from the matrix into the intracristal and
perimitochondrial spaces. Thus, the dehydration
of the matrix may represent protein contraction.
However, if this is the case, no chemical energy
seems to be required for this process. But rather
the removal of Ca ++ from the protein or lipoprorein results in contraction, i.e., superprecipitation
or dehydration. The studies of Vignais et al. (23),
indicating that the acidic phospholipids which
were removed from Ohnishi's "contractile pro-
that alterations in both ionic and hydrophobic
bonds between the acidic phospholipids of the
mitochondrial membranes could be the primary
factors involved in the hydration and dehydration
of the oriented protein-lipid micelles of which
mitochondria are made.
Thus, the actual decrease in size of mitochondria is primarily the result of the movement of
water from the intramembranous spaces rather
than of dehydration (or contraction) of the matrix.
However, since either A T P or E D T A is necessary
for contraction (both A T P and E D T A are good
chelators of Ca ++ , and both cause contraction of
the matrix), it is not possible to state that contrac-
tein" (24) are the factors which are necessary for
contraction of aged mitochondria rather than the
protein moiety of the contractile protein, appear
to indicate that some interaction between phospholipid and structural protein or between phospholipid and phospholipid is what is responsible
for contraction. The recent studies of Abramson
et al. (25) on the effects of various cations, especially Ca ++, on coagulation of acidic phospholipid micelles also lend further support to the idea
]8
tion of the matrix is an obligatory requirement for
mitochondrial contraction. However~ all conditions thus far observed, in which water movement
out of mitochondria occurs, are also accompanied
by dehydration of the matrix. The data in Table I
indicate that contraction of mitochondria is
completely unrelated to the actual mitochondrial
content of Ca ++. At 0 time and in the presence of
A T P and substrate (Table I), conditions which
are associated with maximal contraction are also
THE JOURNALOF CELL BIOLOGY • VOLUME ~3, 1964
Downloaded from on June 17, 2017
FIGURE 6 Effect of A T P on aged-swollen mitochondria. Mitochondria, aged as in Fig. 5, were
incubated, as in Fig. 1: with succinate, 10 3 M,
and serum albumin, ~ mg/ml, for 5 minutes. At
this point, ATP, 8 X 10-4 M, and Mg, 8 X 10-a ~,
were added~ and incubation was continued for
another 5 minutes. Mitochondria] pellets were
then obtained and fixed as in Fig. 4. Contraction
of these mitochondria was observed turbidimetrically before centrifugation. Final magnification, 18,000.
Published October 1, 1964
transport substrates are known to cause the liberation of H + from mitochondria by translocation of
calcium phosphate (2), and since Ca ++ causes
displacement of H + from acidic phospholipids (25),
and since electron transport (e.g., oxidation of reduced flavoprotein by cytochrome b) is, in principle, capable of liberating H +, it seems reasonable
to speculate that the cycling of the electrontransport system makes protons more available
inside the mitochondria, with the resultant solubilization of mitochondrial calcium phosphate by
an exchange of H + for Ca++ in the calcium phosphate crystals. (Similar conjectures about the
separation of charge by electron transport have
been made by Mitchell, 26). The liberated phosphate could accept the protons, and the acidic
phospholipids the Ca++. Swelling could thus
ensue as a consequence of the liberated Ca++.
Whether or not swelling is a result of the exchange
of protons for Ca++ on the acidic phospholipids is,
of course, pure conjecture at this time. Experiments designed to validate this conjecture will be
the subject of another report.
Received for publication, October 24, 1963.
REFERENCES
1. VASIN~TON,F. D., and MURPHY, J. Y.: ft. Biol.
15. PALADE, G. E., in Enzymes: Units of Biological
Chem. 1962, 237, 2670.
Structure and Function, (O. H. Gaebler,
2. BRIERLEY, G. P., MURER, E., BACHMANN, E.,
editor), New York, Academic Press, Inc.,
and GREEN, D. E., J. Biol. Chem., 1963, 238,
1956, 185.
3482.
16. SJOSTRAND, F. S., in Bioenergetics, Radiation
3. RossI, C. S., and LEHNINGER,A. L., Biochem. Z.,
Research, New York, Academic Press, Inc.,
1963, 338,698.
1960, suppl. 2, 349.
4. LEHmNOER, A. L., ROSSL C. S., and GREENA- 17. LYNN, W. S., and BROWN,R. H., J. Clin. Invest.,
WALT, J. W., Biochem. and Biophys. Research
in press.
Com., 1963, 10,444.
18. KARNOVSKY, M. J., J. Biophysic. and Biochem.
5. LEHNINOER, A. L., Physiol. Rev., 1962, 42, 467.
Cytol., 1961, 11,729,
6. SALLIS,J. D., DELucA, H. F., and RASMUSSEN, 19. WER~Ct-mISER,W. C., and BARTLEY,W., Biochem.
H., J. Biol. Chem., 1963, 238, 4098.
J., 1957, 66, 79.
7. LEHNINOER,A. L., and NEUBERT, D., Biochim. et 20. JACKSON,K. L., and PACE, N. J., J. Gen. Physiol.,
Biophysica Acta, 1962, 62, 556.
1956, 40, 47.
8. LYNN, W. S., BROWN, R. H., FORTNEY, S., and 21. TEDESCHI, H., Biochim. et Biophysica Acta, 1961,
CLANCY, T., Fed. Proc., 1963, 22, 526.
46, 159.
9. LYNN, W. S., FORTNEV, S., and BROWN, R. H., 22. WOJTCZAK,L., WLODAWER,P., and ZBOROWSKI,
J. Cell Biol., 1964, 23, 1.
J., Biochim. et Biophysica Acta, 1963, 70,290.
10. HOOEBOOM, G. H., SCHNEIDER, W. S., and
23. VmNmS, P. M., VIONAIS,P. V., and LEHNINOER,
PALADE, G. E., J. Biol. Chem., 1948, 172, 619.
A. L., Biochem. and Biophys. Research Com., 1963,
11. AMOORE,J. E., and BARTLEV, W., Biochem. J.,
11,313.
1958, 69, 223.
24.
Onmsm,
T., and Onmsm, T., J. Biochem., 1962,
12. DOLE, V. P., J. Clin. Invest., 1956, 35, 150.
51, 380.
13. GRIFFITHS,D. E., and CHAPLAIN,R. A., Biochem.
and Biophys. Research Com., 1962, 8, 497, 501. 25. ABRAMSON,M. B., KATZMAN,R., and GREGOR,
H. P., J. Biol. Chem., 1964, 239, 70.
14. BjSRNTORP, P., ELLS, H. A., and BRADFORD,
26. MITCHELL,P., Nature, 1961, 191, 144.
R. H., J. Biol. Chem., 1964, 239, 339.
W. S. LYNN, JR., S. FORTNEY, AND R. H. BaOWN Role of E D T A and Metals
19
Downloaded from on June 17, 2017
the conditions which cause translocation of calcium phosphate into mitochondria (1). This
calcium phosphate, however, is probably not in
solution, but is present as insoluble calcium phosphate crystals, as previously observed microscopically by Lehninger (4), and is apparently not
available to hydrate the matrix.
The question of exactly how electron-transport
substrates cause swelling deserves further consideration. It is clear that Ca++ is the final cause of
this type of swelling, since, when it is removed by
EDTA, contraction ensues. The data in Table I
indicate that succinate is not causing a leak of
Ca++ from the mitochondria, but rather that just
the reverse is occurring. It is also clear that the
actual content of Ca++ in the mitochondria is also
not related to swelling. Since it is known that Ca++
does exist in mitochondria as insoluble precipitates
of calcium phosphate, it appears that substrates
are solubilizing the insoluble calcium phosphate
by the cycling of electrons through the electrontransport carriers. (Inhibition of electron transport by cyanide, Ns, or antimycin blocks the
swelling effect of succinate.) Since electron-