: . i I a
1 I 4 v -V .
. . Law. 3514... x!
inmkkunnazh .1. .m....l.r:..... 5 ‘ .
will.” “5.3..Efll..ln\wu - .. .. any... x u .
r .L :flvtn 11-5? '{V‘III J)..ilfl| ~C'.’
a». o . I.Iv~\vovvo\qi $' {IlIIA-u oo‘jlhuu -
as v to: u . .9 0“! L .Wll r£0.0!v)ss’v"’¢ .-
.oo-‘. 0 on. tP.o.iocc.b v\\bv\v.|"1¢.ulu\.ncl
V: V .9 u: v 1.556 (I .v-‘ulV ‘vn.l1!Olv.o- o
. afi-I- o o-I':.>....l “1“”. .0. .141 it‘lp
uou I c: 04. i :OQA..$I‘-n
I a I: I:

u.
.1. 1.13;. . 2 p ”ti! No pin-
-.x 3.... .. . ... .Mmfidu... win")! 15m.” . .
.. 3.. Int»)? .(r. 12”.! .......lr.u....,.l.Il.2 shift. - n
t v I. o..\..N¢o . g :1. I! v '17 Ol- . . .‘nvlrni. .93.-
. a. - at- 3.2. .lf!&..L.nf..«N\.~\!€-!-:‘ill .
. vii-I‘m u~ 1."!- oudoml‘HLov 5}":th v«\ 9". 1043'? ‘1
1' ' . .1. | -' I 01
v I... Abufzvlpikfimhafiolflgmfwmfieubflxfig . luv.“ . . 7!. {2.1: u.
. I J I .
iflfibwfiukfilt -. . .9...Dm$h.¥)&h .CI.«.( 11¢,“
‘ oI‘I lol‘tlll, ‘II ‘
«sunflofl5...nlflntil L. . $.
. ollltvotflsl .

t. w

5% r. g? .mt
‘o .t a.
.n.» “IO“. 0'0 JfiM-hlvhnflcm.
Nina“. £huh.£ I . 93¢.» Mu: n.
I , iii. t‘ to
IV»

  

      

.
0...!
i.
.. . 1‘ O‘
..v o. . I.”mso.lvnv¢.fl:l H. “Lu o.
.15.}'! ' -.‘. k '.
[3.2. f. . - v \ yum-3k.
. I I . v1 .0
t9

. 2‘ a .w.
(tr l . L
)hM ... u. . l!

 

\
n e 0.
1.! i
. ’5. ’5. 1A .
. ‘ to“ . -
. E38”?! . &. 3..va .51...
L
.. . .4 Aniluuflhflml. 1!:cfllsullngfl—Vr
I . ‘5'. . It“; I:
. V In. til -.|1.'.A,v"lvr|l ‘0 ’5 Q
221$! {iii nun" . m . ‘01-}!
I D I. ' . )D
. 4. ..§xiin§wflfl$ : s
v I 1.1 .lbrlqulfilc -
. .1».th (fined.
I I: ,2. ‘
I. q :5 I
It 3..-.strtoL I91 3. i-
if. I] :iummtnitmmwrflwur
aux n...“
'08.
«(r -; .
- .
.

i I a . If... ‘I
3 I {I} 1. {$331!}: .I(.
III! ll .1! .- 'h”; I No: qulnsvil’ol‘flnli .. a”!
. . n . . Mug". . I .Hll’-. 11.1». .Ifll ‘7 i I» utl tom“ .1." t Novlnnuvlllnnkr't
‘01“. {(1}!!! {INHIIUN H «F v , .l) Ilslvo .rllb}s" ‘3“thUIhn‘ 1. Jbsfll.‘.lllr|! 1b .
(I I ‘ ' \a‘nfi . i I

‘ '11"
n ' C 1
. - U «infill alfli, ‘15

‘l ml: vul’uHNIla ,

 

    
     

 

O ‘1' l. I
I J ( I! :l b(\ ‘ (.1 |u lif‘l.‘hfl
‘ . A . v- o tn o ‘. «MMV'lnln It :IQIII VMI |'. 0’. . “UH-bl”.
. - I n w 9H t [It I ‘ V L\ ’9 ”WI Ir 1
. x ‘ . fir v - l- officoht{.fl‘.f .arl .15.. .Isl’lvb huhllttinuvifl :1 ‘ 1\
J. . I *0; I1": a!” . [1.9.1mn1tW1Uovu . loan}? :)!.-:.U..: i: w \
l Ih ( “July“? r|r. » .13!!!“ JIGIVTDJK)‘ '«ll’ll.11’h.llav¢.. ‘O'Ivnllllu I \ .~ .!
A ‘I l'lll. I. Ill-Ill} . .I ‘vllx 3.)!) .v’fiIPV“ .
“M5! . 'P‘Ovl‘lllll" tutu-If! E III I» IIIIItnllltJAIl‘ ‘1‘! I‘ll“
I H . p D’O- ’JI'II” II} I {of ‘11’5'?! t.‘
-J , I. .Dw.h....;|!z.rn?IH llpvllu {Ill .14 . 1. -
a??? .."oml|;..h‘sll.-lvgl jtlljhflflllau '11 .

. Mlle. IXmJ. .
. ‘ 0! ft)?" . u. .
hflflwntuf q IIIIIII~ -0 o, lo! q 1.
\l

 

I 1'. .
5.
.l

r

- I . u‘ »

ll
‘ I'- I h.
i i :iutdwuwhioiwdmmhfiWLvJ .
’1‘! firm ‘ - HDO‘DI y
z ‘ .. - -

lit .. .13 ’1:
.Ir! {a}.
v q. . n .

All 1". l.

 

LIBRARY
Michigan State

University

 

 

 

This is to certify that the

thesis entitled

“\Q \octaJr'xovx cane. Mfib\\“‘j °Q \\~\omescemk\j

\ can at ck uJos Vw'wwnq W\ mfiochmn 54M 0A Mmbwwe 5

MA, M\\M1\\¢m ?V\o<, \nok ‘\ .
presented be Q 6‘ ‘12 S‘QQY

QM»; o~3\ 05mm, \< may“ emu“

I

has been accepted towards fulfillment
of the requirements for

M' 5' degree in Q7" OWQ’WLS‘t «.3

r 'W

Major professor

 

Date—Sffi‘f Jf/gf

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

MSU

LIBRARIES
_—JI—.

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

 

THE LOCATION AND MOBILITY OF FLUORESCENTLY LABELED
UBIQUINONE IN MITOCHONDRIAL MEMBRANES AND

UNILAMELLAR PHOSPHOLIPID VESICLES

BY

Krishnakumar Rajarathnam

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Biochemistry

1987

ABSTRACT

THE LOCATION AND MOBILITY OF FLUORESCENTLY LABELED
UBIQUINONE IN MITOCHONDRIAL MEMBRANES AND

UNILAMELLAR PHOSPHOLIPID VESICLES

BY
Krishnakumar Rajarathnam

The location and mobility of ubiquinone (Q10) in

mitochondrial membranes are subjects of controversy. A

9

diffusion coefficient of 3x10- cmzsec-1 was measured by

fluorescence redistribution after photobleaching for a
ubiquinone analog whereas a value of of 1x10.6 cmzsec—1 was
determined for native ubiquinone by collisional quenching.
To assess the influence of the isoprene chain on mobility,
we have synthesized a fluorescent derivative of the quinone
moiety of native ubiquinone (NBDCOQ) and have measured its
diffusion and location in membranes of giant mitochondria
and phospholipid vesicles. Lateral diffusion rate for

9 1

NBDCOQ was 3.1 x 10- cmzsec— in mitochondria and

8 1 in vesicles. Similar rates were

1.1 x 10- cmzsec-
observed for head group labeled NBD-phosphatidylethanolamine
(NBDPE). However fluorescence emission and quenching
studies show that the quinone moiety is in a more

hydrophobic environment and is less accessible to quenching

agents than the modified phospholipid head group. These
results indicate that ubiquinone is oriented differently
than a phospholipid in the membrane, but their diffusion

rates are similar.

To Peace and Harmony

iv

ACKNOWLEDGEMENTS

 

It is my great pleasure to sincerely acknowledge my
mentor, Dr. Shelagh Ferguson-Miller for her guidance,
support and inspiration. .

Thanks a many to my lab mates Taha, Linda, Bob, and
Wendy for the exciting science, fun, and friendship. My
thanks also go to Maria, Debra, Jerome, Becky, Bridgette

and Peter for their companionship during some part of my
stay in the lab.

I am also grateful to Dr. Schindler for his guidance
and enthusiasm and to Dr. Holland for the many useful
discussions. I am grateful to the mass spec facility,
especially to Brian Musselman, for their help in getting the
mass spectra. My thanks also go to Mark, Jeff, Vinod and
Usha for their help in one way or the other and to George
for his help in printing out this thesis.

Finally, but not the least, to my family and friends for

their love, support and faith in me.

TABLE OF CONTENTS

LIST OF FIGURES
LIST OF TABLES
LIST OF ABBREVIATIONS
INTRODUCTION
A breif overview of structure, function, and
location of ubiquinone and their relation to
diffusion °
EXPERIMENTAL METHODS
Materials
Preparation of giant mitochondria
Preparation of phospholipid vesicles for FRAP
Synthesis of NBD-Ubiquinone (NBDCOQ)
PRAP experiments:
Preparation of slides for FRAP

Incorporation of the fluorescent probes
into mitoplasts

Bleaching conditions and diffusion
measurements

Fluorescence quenching of the probes in
mitochondria and vesicles

Electron transfer measurements
RESULTS
Synthesis of NBDCOQ
Characterisation of NBDCOQ
Diffusion meaurements of the fluorescent
probes in the membranes
Preparation and characterisation
of giant mitochondria

Preparation and characterisation
of phospholipid vesicles

vi

viii

ix

10

10

11

12

15

15

16

17

20

21

23

34

34

Incorporation of the fluorescent probes
into the membranes

Diffusion coefficients of NBDCOQ and
NBDPE in mitochondria and vesicles

Fluorescence emission of the probes in
the membranes

Quenching of the fluorescent probes in
the membranes

Effect of the fluorescent probes on
electron transfer in mitochondria

DISCUSSION

Diffusion of NBDCOQ in vesicles and mitochondria
Location of Q in the membranes
7 Electron transfer through Q pool

Is diffusion of Q10 rate limiting in electron
transfer?

Heterogenity of mitochondrial membrane and its
implication in electron transfer through Q1°

SUMMARY
REFERENCES

APPENDIX: Publications & Abstracts

vii

37

39

39

44

52

55

59

67

71

75

82

85

91

FIGURE

1

10

11

12

13

W

Space filling model of ubiquinone.
Various possible orientations of short
and long chain ubiquinones in a membrane.

Reaction scheme for the synthesis of
NBD-ubiquinone (NBDCOQ).

A representative FRAP experiment.
Absorbance spectrum of NBDCOQ.

Fluorescence excitation and emission
spectra of NBDCOQ.

NMR spectrum of NBDCOQ.

Blution profiles of the two isomers of
NBDCOQ from HPLC and their mass spectra.

A plot of the diffusion coefficient values
of the fluorescent probes as a function of
the diameter of mitochondria and vesicles.

Fluorescence recovery profiles of NBDCOQ in
mitochondria and vesicles.

Stern-Volmer quenching plots of NBDCOQ and
NBDPE in mitochondria and vesicles.

Lehrer inverse plots of the quenching of

NBDCOQ and NBDPE in mitochondria and vesicles.

The dynamic aggregate model of the
mitochondrial respiratory chain.

viii

PAGE

14

19

25

28

3O

33

36

42

47

50

77

LIST OF TABLES

TABLE PAGE
1 Lateral diffusion coefficients of
‘ NBDCOQ and NBDPE in mitochondria and
vesicles. 40
2 Fluorescence emission maxima of the
fluorescent probes in solvents, vesicles
and mitochondria. 43
3 Stern-Volmer quenching (k ) constants of
NBDCOQ and NBDPB in mitocRXndria and vesicles. 51
4 Effect of the fluorescent probes on electron
transfer in mitochondria 54

ix

LIST OF ABBRggIATIONS

BSA, bovine serum albumin

CCCP, carbonyl cyanide urchlorophenylhydrazine

CMC, critial micelle concentration

DiI, dioctyl- or dihexyldecylindocarbocyanine

DMAP, dimethylaminopyridine

DMPC, dimyrsitoyl phosphatidylethanolamine

DPPC. dipalmitoyl phosphatidylethanolamine

Hepes. N-(2-hydroxyethyl)piperazine-N'-2-ethane-sulfonic
acid

IANBD. 4-[N-(iodoacetoxy)ethyl-N-methyl]amino-7-nitrobenz-2-
oxa-1,3-diazole

NBDCOQ, 4-[N-(acetoxy)ethyl-N-methyl]amino-7-nitrobenz-2-
oxa-i,3-diazole-ubiquinone .

NFDPE, N-(7-nitro-2-1,3-benzoxadiazol-4-yl)dipalmitoyl
phosphatidylethanolamine

G—NBDPE, i-acyl-2-[6-{(7-nitro-2—1,3-benzoxadiazol-4-yl)
amino}caproyl] phosphatidylethanolamine

iZ-NBDPE, 1-acyl-2-[12-{(7-nitro-2-1,3-benzoxadiazol-4-yl)
amino}dodecanoyl] phosphatidylethanolamine

NMR, nuclear magnetic resonance

Tc' lipid transition temperature

TLC, thin layer chromatography

STRUCTUREI FUNCTIONI LOCATION AND MOBILITY

OF OBIQUINONE: AN OVERVIEW

Ubiquinone (010) is an integral component of the
electron transport chain and mediates the transfer of
electrons between the dehydrogenases and the cytochromes in
mitochondria and participates in the translocation of

protons across the membrane. is a hydrophobic molecule

Q1o
with a long isoprenoid side chain of about 50 A0
(Trumpower, 1981) and if it were in an extended form it
would span the bilayer (Figures 1, 28). The isoprenoid
chain of Q10 is fairly rigid (Trumpower, 1981) due to the
presence of the double bonds and the vicinal methyl groups
and it is likely that the rigidity of the side chain
strongly influences the location of the molecule in the
membrane. It has been shown that the orientation of short
chain ubiquinones is parallel to a phospholipid with the
quinone moiety near the membrane surface (Figure 2A) while
long chain ubiquinones appear to be located deeper in the
membrane. Studies by Crane (1977) have shown that 010 in
the mitochondrial membrane is largely inaccessible to
hydrophilic electron donors and acceptors. These and other
studies on reconstitution of mitochondrial function led him
to suggest a model for orientation of Q10 in which the
quinone ring and the prenyl side chain are in the center of

the bilayer (Figure 20). This model differs from the more

traditional model of Q10 in which transmembrane flip-flop

1

Figure 1. Space filling model of ubiquinone. Reproduced
from Trumpower, B.L. (1981) in J. Bioenerg.

Biomembr. 13, 1-24.

 

Figure 1

I07:

 

 

Figure 2. Various possible orientations of short and long
chain ubiquinones in a membrane. Reproduced from
Siedow, J.N., & Stidham, M.A. (1986) in Biomedical
& Clinical Aspects of Coenzyme Q, Volume 5,

(page 70).

 

across the membrane provides the mechanism for electron and
proton transport (Mitchell. 1976) (Figure 2D). Existence of
010 as a micelle spanning the bilayer has also been proposed
(Ondarroa & Quinn. 1986).

The exact mechanism of energy transduction through 010
remains unclear. Many of the hydrogen carrying substrates
of the electron transfer chain such as succinate and NADH
are hydrophilic and their function is localized to the
aqueous domains on either side of the membrane. Since redox
reactions of Q10 involve protons. communication with an
aqueous environment would appear necessary (Rich, 1982).

The membrane bound enzymes that interact with 010 would
require their Q-reactive centers to be accessible to the
surface of the membranes. Rich (1984) proposes Q and 0H2
are in such an environment that they are essentially
unreactive until they collide with an appropriate site on a
donor or acceptor, ensuring the biological specificity of
the reaction route. One reason, therefore, that the
biological quinones possess such long hydrophobic
side-chains might be to ensure that they remain unreactive
in possible damaging chemical side reactions until they
reach their reaction sites.

It has become increasingly evident that diffusion of
Q10 is important for electron transfer between the
dehydrogenases and the cytochromes in mitochondria. Whether

its diffusion is rate limiting in electron transfer is not

clear. To gain insight into this question it is would be

useful to know the diffusion rates of 01°. the Eact for the

diffusion (Hackenbrock g; 31., 1986b). the kn values of the
electron carriers for the native quinone in the mitochondria
Q10 has to
diffuse to accomplish electron transfer. Currently, there

(Fato g3 51., 1986), and the average distance

is controversy regarding all these values.

Gupte g; 5;. (1984) have measured the diffusion of a
fluorescent derivative of a ubiquinone analogue which has an
alkyl side chain corresponding in length to only two
isoprene units. The lateral diffusion coefficients they
determined by fluorescence redistribution after

9 1 in the

1

cmzsec-

mitochondrial membranes and 5.5 x10.8 cmzsec-

photobleaching (raap) was a x10-
in
phospholipid vesicles (Hackenbrock 35 31., 1986a). From
diffusion measurements as a function of temperature
(Hackenbrock 2; 31., 1986b) and as a function of increased
membrane dilution with phospholipids (Schneider 33 51.,
1982). they propose that diffusion of Q1° is rate limiting
in electron transfer. In contrast, Fato gt 5;. (1986) have
measured the diffusion of 010 in small sonicated vesicles
and submitochondrial particles by the technique of
collisional quenching of fluorescently labeled fatty acids

6

and report a value of 1 x10- cmzsec-l. They have

calculated the sect for diffusion by measuring quenching
constants as a function of temperature and the km values for
Q1° with cytochrome bc1 and NADH dehydrogenase from kinetic

measurements. They propose from these studies that

diffusion is not rate limiting in electron transfer.

Q10 is in excess over its redox partners in the
mitochondrial membrane (about 20 fold excess over complex
III; Capaldi, 1982) and it was originally proposed by Green
(1962) that Q10 acts as a mobile electron carrier. The
concept of a Q pool is widely accepted but for a few
investigators (King 8 Suzuki.1984., Yu & Yu, 1980) (see
discussion on pages 66-70). However, there is conflicting
data regarding the homogeneity of the pool (for reviews see
Rich, 1984; Ragan & Cottingham, 1985)'and the concept that
the Q pool is heterogeneous has been postulated by Gutman
(1980, 1985). The basic tenet of the concept is that
diffusion is not rapid enough to sufficiently randomize the
Q pool. It appears that diffusion rate of Q10 would affect
whether the electron transfer through Q10 would follow a
simple Q pool behavior or not (Rich, 1984) since it will
determine the number of protein molecules a diffusing
quinone or quinol molecule could see in its life time.

The major controversies regarding Q1° behavior center
around how fast it diffuses in the native membrane and where
it is located. Two possible experimental problems may have
led to the discrepancy observed in the reported diffusion
rates: 1) the use of a ubiquinone derivative in lateral
diffusion measurements without a native 10 unit isoprene
tail; 2) systematic errors in the technique or calculations
involved in determining diffusion rates by collisional

quenching.

To test the former possibility, we have synthesized a
fluorescent derivative of the native ubiquinone (Q10)
labeled in the head-group moiety so as to retain an
unaltered isoprene tail. we have measured its diffusion in
giant mitochondria and cell size phospholipid vesicles. We
also report here the experiments performed to locate the
position of the fluorescently labeled ubiquinone by
fluorescence emission and fluorescence quenching studies.
Our results show that the NBDCOQ is located deeper in the
membrane than a phospholipid bearing the same fluorescent
moiety in the head group but that its diffusion rate is

similar.

EXPERIMENTAL METHODS

Materials. The following materials were obtained from
the sources indicated. Cuprizone, Aldrich Chemical Co.,
recrystallized from gas ethanol according to Bowman &
Tedeschi (1983); asolectin, Associated Concentrates, Long
Island, NY; NBD-labeled phospholipids, Avanti, Birmingham,
AL; IANBD, Molecular Probes, Eugene, OR; preparative TLC
plates, Analtech, Newark, DE; DMAP, recrystallized from
dichloromethane, ubiquinone(Q10) and CCCP, Sigma Chemical

Co.

Preparation of giant mitochondria. Fifteen to seventeen
day old mice (Swiss Albino ICR) are fed a diet containing 3

grams of cuprizone in 500 grams of rodent chow. Giant
mitochondria from these mice are prepared 6 to 12 days after
initiating the diet, essentially according to the procedure
of Hochman 53 a1. (1985). 3 or 4 livers are taken in 5 ml
of isolation buffer (220 mu mannitol, 70 mM sucrose, 2 mM
Hepes, 0.053 BSA, pH 7.4 (with Tris)}, minced and washed a
few times in the same buffer. The minced livers are
homogenized twice using a loose fitting glass Dounce
homogenizer and the homogenate is centrifuged at 120xg for a
minute. The supernatant is saved and the pellet is
homogenized and centrifuged again at 120xg for one minute.
The supernatants are pooled and layered on 10 ml of 0.5 M

sucrose and centrifuged at 730xg for 5 minutes in a swinging

lO

11
bucket bench top centrifuge. The top layer containing

smaller mitochondria and a loose bloody pellet at the bottom
are discarded and the rest of the 0.5 M sucrose layer and
the interface are gently resuspended. The solution is
diluted to 0.3 M sucrose with cold distilled water and then
spun at 750xg for 5 minutes. The pellet is suspended in 5
ml of isolation buffer, layered on 0.5 M sucrose and
centrifuged at 240xg for 3 minutes in a bench top centrifuge
and the top layer excluding the interface is collected and
pelleted at 750xg for 5 minutes. The giant mitochondria are
depleted of outer membrane by first swelling the
mitochondria in 5 ml of 5 mM Tris-Phosphate (pH 7.5) for 5
minutes and then adding an equal volume of shrinking buffer
(1.8M sucrose, 2mM ATP, 2mM Mg2+) and incubating for another
5 minutes. The mixture is then sonicated for 20 seconds
(setting 3, model W-225, Heat Systems Ultrasonics, Inc) and
the mitoplasts are pelleted at 1900xg for 10 minutes. The
mitoplasts are suspended in 0.1-0.2 ml of 250 mM mannitol,
50 mM Hepes (pa 7.2) buffer and the amount of protein is
determined by a modified Lowry assay using bicinchoninic

acid (BCA) (Smith t 1., 1985).

Preparation of cell size vesicles for FRAP. Unilamellar

phospholipid vesicles are prepared essentially according to
the procedure of Mueller 3; 3;. (1983). Asolectin

(6.25 mg), a-tocopherol (0.04 mg) and NBDCOQ (0.125 mg) are
dissolved in chloroform in a test tube to give a 1:50 w/w

ratio of NBDCOQ to phospholipid. This is made to a final

12
volume of 2 ml in chloroform/ methanol/ 0.5 NaCl

(2:2:1.8 v/v) as to make it biphasic. The mixture is
vortexed and then spun at room temperature for five minutes
in a bench top centrifuge. The upper aqueous layer is
discarded and the chloroform layer is filtered through

anhydrous MgSO into a 25 ml erlenmeyer flask. The

4
filterate is dried under argon and 25 mls of 0.05 mM NaCl
solution is added gently into the flask. The flask is
wrapped with silver foil and is kept at 4° C. Vesicles form

at the bottom of the flask after one to two days and are

stable over a period of one to two weeks.

Synthesis of NBD-ubigginone. The reaction scheme for
synthesis of NBDCOQ is given in Figure 3. Ubiquinone (5 mg)

is dissolved in one ml of chloroform/methanol (1:2v/v) in a
screw top culture test tube and reduced with a few grains of
of sodium borohydride. The solution turns colorless and
there is a brisk effervesence of hydrogen gas. The solution
is stirred until the evolution of hydrogen ceases and then
dried using argon. 1.7 millligrams of DMAP (3:4 mol/mol)
and 4.7 millligrams of IANBD (2:1 mol/mol) is added to the
reduced ubiquinone and the mixture is suspended in 1 ml of
dry tetrahydrofuran, flushed with argon and the reaction is
carried out over night. The reaction mixture is dried

with argon and dissolved in 300 microliters of
chloroform/methanol (2:1) and applied to a preparative TLC
(uniplate-T, Analtech). The plate is chromatographed in

chloroform/hexane/methanol (100:50:2.5 v/v) for fifteen

Figure 3.

13

Reaction scheme for the synthesis of
NED-ubiquinone (NBDCOQ). Ubiquinone is reduced
with borohydride (BH‘-) and reacted with IANBD
using dimethylaminopyridine (DMAP) as the
catalyst. Tetrahydrofuran (THF) is used as the
solvent and the reaction is carried out in the

presence of argon (Ar) under dark.

 

Figure 3

 
 

 

 

3H,: 0 083 10
CHCl3tMeOH(I:2)\l,
' OH
CH30 I CH3
0H 10 \ //
CH3 N
/ /N\
under /0:
darkAr \ \N
DMAP /N\
H30 cuz-CHz-o-fi-cuz-I
AINBD
V
0\\ [CH3
7" O .\
° / \ :"2
N N CH2
.. 0
I
1‘”
"2
I
0
CH30 . CH3
cn3o : H
OH 10

15

minutes. NBDCOQ (Rf-0.55) runs between IANBD (Rf=0.4) and
ubiquinone (Rf=0.85) and is scrapped from the plate and
extracted with chloroform/methanol (2:1). The product is

filtered, dried using argon, and stored at —200 C until use.

ERA? experiments. Glass slides are washed and rinsed
with distilled water and ethanol and then allowed to air
dry. For experiments using liposomes, the clean slides are
coated with gelatin according to Bowman & Tedeschi (1983).
15 to 20 microliters of labeled mitoplasts or liposomes are
applied to a slide and the cover slips are sealed around the

edges with paraffin to prevent evaporation.

incorporation of theflfluorescent probe into the
mitoplasts. To 100 microliters of mitoplasts
(10-15 mg protein/ml) in 250 mM mannitol, 50 mM Hepes buffer
(pH 7.2), 3-5 microliters of NBDCOQ (3-5 mg /ml) or 3
microliters of head group NBDPE (1mg/ml) in ethanol, and
5 microliters of BSA (1mg/ml) are added and incubated for
15 minutes at room temperature. The mitoplasts are then
swollen with 3 volumes of water over a period of 10 minutes
on ice. These conditions, especially the presence of BSA,
were important for achieving sufficient incorporation to
perform the FRAP experiments. To quantitate the amount of
NBDCOQ incorporated into the mitochondrial membranes under
the conditions used for FRAP, 50 microliters of NBDCOQ

labeled mitoplasts (3 mg protein/ml) were diluted to one ml

16

with 62.5 mM Mannitol, 12.5 mM Hepes (pH 7.2) buffer and
centrifuged at 4000xg for 10 minutes. Unincorporated NBDCOQ
'was extracted from the supernatant by partitioning it into
an organic phase by adding chloroform, methanol, and

0.5M NaCl to a ratio 2:2:1.8 (v/v) and spun in a bench top
centrifuge for 5 minutes. The upper aqueous layer was
discarded and the chloroform layer was filtered through
anhydrous MgSO4 and is dried with argon. The pelleted
mitochondria were resuspended in 0.1 ml of the Mannitol-
Hepes buffer mentioned above and the incorporated NBDCOQ was
extracted by adding chloroform, methanol, and 0.5M NaCl to a
final volume of 1:2:0.8(v/v). The mixture was spun as
before and the supernatant was transferred to another test
tube and made biphasic by adding chloroform, methanol, and
0.5M NaCl to a final ratio of 2:2:1.8(v/v) and the
extraction was carried out as described for the supernatant.
From the absorbence at 457 nm of the mitochondrial and
supernatant extracts redissolved in chloroform, the amount
of NBDCOQ incorporated into the membrane was calculated. In
the case of the liposomes, the final concentration of NBDCOQ
or NBDPE is determined from the amount originally added to

the phospholipid mixture.

Bleaching conditions and diffusion measurements.
Diffusion measurements were performed by the technique of
fluorescence redistribution after photobleaching using the

instrumentation and analysis developed by Koppel 33 a1.

17

(Koppel, 1979; Koppel _3 _l., 1980). The low intensity
laser beam is scanned across the fluorescently labeled
membrane and profiles of the fluorescence emission are
recorded. A sample profile for a phospholipid vesicle
labeled with NBDCOQ is shown in Fig 4A. The peaks indicate
that the fluorescent probe is associated with the membrane,
giving a more intense fluorescence at the edges. A higher
intensity pulse is applied to one edge of the vesicle
resulting in bleaching of the probe in this region (chemical
destruction) and giving the profile seen in Figure 4B. The
redistribution of the unbleached fluorescent probe

(Figure 4C) is followed with sequential scans and the data
is analyzed using a normal—mode analysis for diffusion on
spherical surfaces as described by Koppel gt 2;. (1980).

N80 fluorescence is monitored with an incident wavelength of

477 nm and a combination of Leitz dichroic mirror TX 510 and

a barrier filter K 530.

Fluorescence Quenching Measurements. Steady state

fluorescence measurements were obtained in a Perkin~Elmer

 

550-60 spectrofluorimeter. Quenching experiments with Cuso4

was carried out as follows. To a cuvette containing 1 ml of

NBDCOQ or NBDPE loaded mitochondrial solution, a 3 mM CuSO4

was titrated in 5 ul aliquots. For the vesicles, a 1 mM
solution of CuSO4 was added in 3 or 5 ul aliquots. After
each addition, the sample was mixed and then allowed to

equilibrate before measuring the fluorescence intensity.

18

Figure 4. A representative FRAP experiment. Profiles of
fluorescence intensity of NBDCOQ in a phospholipid
vesicle. Fluorescence distribution before
bleaching (t<0 secs) is represented as ( ----- ) and

after bleaching (t=4 secs) is represented as

 

( ). The fluorescence recovery after 5

seconds (10 scans) is shown as ( .......... ). The

diameter of the vesicle is 8.7 microns.

 

Figure 4

Relative fluorescence intensity (x IOOO)

.... N
on A
l T r

 

 

Jaqumu :ugod

 

 

 

 

 

20

NBD moiety was excited at 468nm and the emission was
monitored at 527 nm for NBDCOQ and 536 nm for NBDPB. Slit
width was 5 nm for the excitation beam and 10-13 nm for the

emission beam.

Effect of NBDCOQ on succinate-oxidase activity.

Succinate oxidase activity of mitochondria was measured on a
Gilson polarograph in 1.75 ml of 250 mM mannitol, 50 mM
Repes, (pH 7.2), 13.8 mM succinate, 7.6 uM cytochrome g and
0.55uM CCCP. Turnover numbers (TN) were calculated from
the rates of 02 consumption multiplied by 4 to give the
nanomoles of cytochrome 9 required to reduce 1 nanomole of
02, and divided by the total nanomoles of cytochrome aa

—9
present in the reaction vessel.

Results

Sygthesis of NBD-ubiguinone (NBDCOQ)

We are reporting for the first time the synthesis of a
fluorescent derivative of Q10, NBDCOQ. There were several
considerations involved in deciding on the appropriate
labeling method. 1)The only functional group that is
sufficiently reactive in ubiquinone to allow modification is
the keto group or in reduced form the phenolic group. 2)The
fluorescent tag to be attached should be excitable in the
visible region so that it can be used for FRAP experiments
with our laser. 3)The label should be small and hydrophobic
so as not to perturb the hydrophobic character of the
quinone. NBD (excitation maximum = 477 nm) and rhodamine
(excitation maximum = 531 nm) were two possibilities
considered because rhodamine is available as a sulfonyl
chloride (Lissamine rhodamine B sulfonyl chloride) and as an
alkyl halide derivative (Tetramethyl rhodamine
iodoacetamide), and N80 is available as an alkyl halide
(IANBD). These compounds can undergo nucleophilic
substitution reactions since chloride (in sulfonyl chloride)
and iodide are good leaving groups. They are generally used
for reactions with sulfhydryl and amino compounds which are
fairly strong nucleophiles. Though the hydroxyls (and
phenols) are weak nucleophiles. the above compounds can be
used to derivatize them, but the conditions have to be more

rigorous and the yields are likely to be low. Between

21

22

rhodamine and N80, the latter was preferred as it is
smaller, less hydrophilic, and likely to perturb the
structure of ubiquinone to a lesser extent.

Phenols in general are weak nucleophiles. is a

Q10
tetra-substituted bisphenol (see fig.3), making it even less
reactive. IANBD can be considered as a primary iodide (see

fig.3 and the reaction will probably proceed through a 8N2

(second order nucleophilic substitution) rather than a 8N1
(first order nucleophilic substitution) mechanism. The
formation of the primary carbocation would not be favorable
especially because of the presence of the keto group in the
alpha position.

The initial strategy followed was to increase the
nucleophilicity of reduced Q10 by reacting it with a strong
base, potassium tart-butoxide to generate a phenoxide ion, a
relatively stronger nucleophile. Aprotic polar solvents
like dimethylforamide (DMF) and tetrahydrofuran (THF) had to
be used to increase the solubility of the reactants and as
well as to accommodate the base. THF was favored as it has
a low boiling point (b.p.=66o C) and hence can be removed
easily. DMF has a b.p.=153o C.

When reduced Q10 was reacted with butoxide in THF, the
colorless solution turned a intense brown. IANBD was added
and the reaction was carried out for 4 hours at the end of
which the solution was pale brown in color. NBDCOQ could be

purified by preparative TLC. However, the reaction was

extremely sensitive and unless conditions were carefully

23

controlled, it was difficult to reproduce. Butoxide is a
strong base and hence is sensitive to moisture and the
reaction is affected by humidity. It was also observed that
if excess butoxide was taken, the reaction did not work and
products other than NBDCOQ were obtained.

An alternative strategy of activating IANBD was tried
so that phenol rather than the phenoxide would act as a
nucleophile. Dimethylaminopyridine (DMAP) was used as a
catalyst and we were more successful in obtaining NBDCOQ in
a reproducible manner. DMAP is a widely used, highly active
acylation catalyst (Hofle g; 31., 1978) and has been shown
to be far superior to pyridine. DMAP is not generally used
as an alkylation catalyst. In our case, it turned out to be
very effective since DMAP is a better leaving group than
iodide. The reduced Q10 could act as an effective
nucleophile to displace DMAP and form a phenoxy bond to give

Nancoq.

Characterization of MBQCOQ

NBDCOQ has been characterized by various spectroscopic
techniques. Absorbance spectra of NBDCOQ (A), IANBD (B),
and reduced Q10 (C) taken in chloroform are shown in Figure
5. IANBD has absorbance maxima at 463 nm, 332 nm, and 241
nm and reduced Q10 at 234 nm and 290 nm. The spectrum of
NBDCOQ shows the absorbance characteristics of both IANBD
and reduced Q10.

The excitation and emission fluorescence spectra of

Figure 5.

24

Absorbance spectra of NBDCOQ. The absorbance
spectra were obtained in a Lambda 4B Perkin~Elmer
spectrophotometer using chloroform as the solvent
(conc. 56nmol/ml). The spectra of the starting
materials, IANBD (B) (36 nmol/ml) and reduced

ubiquinone (C) (36 nmol/ml) are also given.

Figure 5

 

 

 

 

 

)fl‘\.
.0"! K"
A f"; .I'I'.
."3' i.
.‘f t"
.” I"
III
l/"I - ‘1‘ I."
-. x; a. a.
\ XI \\ .
\_x :\ I
3 "‘-.._____,../
.2:
g /"’-\‘\E
O x"
3 '5
< 10 .2 a bs "‘
B 1'
m {A If IS}
/// \\ ,/ a
I \\k f/ ‘3‘
\ ,/
w —
‘\
\._
200 300 400 500

 

26

NBDCOQ taken in chloroform are shown in Figure 6. The
excitation maximum is at 468 nm and emission maximum is at
518 nm. These spectra were identical to the excitation and
emission spectra of IANBD (not shown).

NMR spectra of NBDCOQ(a), Q10(b), and IANBD(c) are
shown in Figure 7. Chemical shift values for Q10 were
assigned as given in Ulrich t al. (1985). Some of the

prominent signals are numbered to facilitate comparison.

1.60 ppm, isoprenoid CH3; 1.68 ppm, trans-terminal CH3; 1.74
I
ppm, methyl at 03 ; 1.97 ppm, CH2(CH3)=CH; 2.01 ppm(1),
I
methyl at C2; 2.06 ppm(2), CHch=C; 3.2 ppm(3), 1 —CH2; 3.98

and 3.99 ppms(4), OCH3; 5.1 ppm(5), vinyl CH: 7.24 ppm, from
CHCla. Chemical shift values for IANBD are: 1.50 ppm, from
water impurity; 3.48 ppm(6), N(CH3): 3.60 ppm(7), (CH2)I;
4.47 ppm(8), a multiplet due to -CH2-CH2-: 6.17 ppm(9) and
8.44 ppm(10), aromatic peaks with a coupling constant (J)=9
Hz. Signal at 8.44 ppm is from H ortho to the nitro group.
7.24 ppm, from 03013. It is seen that NBDCOQ has signals
present both in IANBD and Q10. However, some of the signals
are shifted either down or up field [example, the methoxy
resonance (4)] in NBDCOQ implying that there is some
interaction between the N80 moiety and the benzoquinone head
group.

There are potentially two reactive groups in reduced
Q10 and both of them react with the iodoacetoxy group of

IANBD to give a mixture of isomers. The two isomers can be

separated by preparative TLC. HPLC of the purified products

Figure 6.

27

Fluorescence excitation and emission spectra of
NBDCOQ. The spectra of NBDCOQ (come. 44 nmol/ml)
were obtained in a 550-60 Perkin-Elmer
spectrofluorimeter using chloroform as a solvent.
The slit width for excitation and emission beam

was 5 nm.

(um) q15U6|8ADM

ZIQ

0617 0917 0217

1729

929

fluorescence intensity (arbitrary units)

Figure 6

 

l

 

I

T

F

 

—l

uogiouoxa

 

 

 

 

 

 

uogssgwa ' "

 

 

29

Figure 7. NMR spectra of NBDCOQ. The spectra were taken in
a Bruker 250 MHz spectrometer using CDCl3 as a
solvent. The spectra of ubiquinone(B) and IANBD

(C) are also given for comparison.

Figure 7

 

o—N

 

 

 

 

 

 

 

 

 

fi

 

 

 

 

 

 

 

 

 

we so so
9318. mg: m A35

 

31

performed on a reverse phase column is shown in Figure 8.
They had similar absorbance spectra (not shown) and their
isomeric character was unequivocally confirmed by mass
spectrometry (Figure 8). Both the spectra have peaks
corresponding to the molecular weight of NBDCOQ and they
have similar fragmentation patterns. Some of the prominent
peaks are [(NBDCOQ) - 3]“ -114a, [(CoQOCHchO) - HJ‘ -921,
[(CoQ) - H]- sees.

NBDCOQ is purified in a single step by preparative TLC
(see methods). Good separation between the starting
materials (IANBD, Q10) and NBDCOQ is achieved and when
NBDCOQ is rerun on the preparative TLC, the product runs as
a single spot. The yield is around 53. Generally the
mixture of both isomers of NBDCOQ was used for the
experiments. NBDCOQ is fairly stable when kept in the
freezer in the dried form. In solution there is some break
down to the starting materials, about 5-1ox in 24 hours.
NBDCOQ was freshly synthesized whenever needed and once
dissolved in ethanol is used within 6 hours. Any
contribution of small amounts of hydrolyzed NBD probe to the
mobility and quenching measurements would be negligible as
it is more water soluble and would partition less into the
membrane. Furthermore, the emission maximum of IANBD (and
likely of hydrolyzed NBD probe) in membranes is around 540
nm while that of NBDCOQ is 527 nm. So the presence of
significant amounts of contaminating NBD would be detectable

from an altered emission maximum.

Figure 8.

32

Elution profiles of the two isomers of NBDCOQ from
HPLC and their mass spectra. The two isomers of
NBDCOQ were separated by preparative TLC under
identical conditions used during purification
except that the chromatography time was increased
to 50 minutes. HPLC of the purified products was
performed using a Waters C-18 reverse phase column
(3.9 mm x 30 cm). The solvent system used was

methanol containing 0.7% NaClO4.H 0 and 0.1% (v/v)

2
703 H0104 (Katayama gt $1., 1980). The flow rate
was 1 m1/min and elution was monitored at 254 nm.
fast atom bombardment (FAB) mass spectra were
obtained on a JEOL HX—110 HF mass spectrometer,
operating in the negative mode. The molecular ion
is indicated by (M-H)-. The FAB matrix used was
triethanol amine. (Mass spectral data were
obtained from Michigan State University Mass
Spectrometry Facility supported by a grant

RR~00480 from NIH).

Figure 8-

Intensity

Relative

 

.9

H9033

 

 

 

 

 

 

 

woo

- - q - ”-0-
336 :13

 

3\~

 

 

‘ mam
«
8.
e
il m8
mm.
+
mwo

5-5:
:3

12.1...
:3

 

 

.NOO

34

Diffusion measurements of the glpgrescent_pr9bes_in the
membranes

Preppration and characterization of giant mitochondria
Lateral diffusion coefficients of NBDCOQ and NBDPE were
measured in giant mitochondria prepared from cuprizone fed
mice and in very large unilamellar vesicles. But for the
size, the giant mitochondria have the same characteristics
as the normal mitochondria. Their heme content, lipid
content (Hochman t 1., 1985), electron transfer (Hochman

_£,gl., 1982), and respiratory control (Maloff., 1978) are
comparable to those found in normal rat liver mitochondria.
The mitochondria are used within 10-14 hours after
preparation and they are fairly stable in this time frame.
NBDCOQ labeled mitochondrial inner membranes on a slide

(during FRAP experiments) are used within 15—45 minutes

after preparing the slide.

Preparation and characterization oprhospholipid

 

vesicles Very large unilamellar vesicles are prepared
essentially according to Mueller g3 a1. (1983). As the
vesicles were mobile on a glass slide, gelatin coated slides
were used to retard the movement of the vesicles. The size
of the vesicles was heterogeneous, diameters ranging from
3-20 microns. Diffusion coefficients seemed to show a
slight dependence on the diameter of the vesicles

(Figure 9), higher diffusion coefficients being obtained at

larger diameters. However. only the data obtained from

35

Figure 9. A plot of the diffusion coefficient values of the
fluorescent probes as a function of the diameter
of mitochondria (open symbols) and vesicles
(closed symbols). (. 0) represent the diffusion
coefficients obtained for NBDCOQ; (I [3) head
group labeled NBDPE; and (A) acyl chain labeled

NBDPE.

('1’) Jaiawola

Figure 9

D (x l09) cm2 . s'l

 

 

on 5 "
u I l a:
0'" L + 4 ‘
l-——O—-1
Cl
Vi— —
0 l-—O——-l
e
co— ..
is ‘
63- _.
1 ' 1

 

 

 

37

vesicles of similar dimensions to the mitochondria were
tabulated (Table 1). Since the location of Q10 in membranes
may be concentration dependent, diffusion measurements were
also performed as a function of concentration to see whether
it had an affect on diffusion coefficients and nature and
extent of recovery. No discernable differences were
observed (data not shown) when the experiments were
performed over a 8 fold concentration range (1:200 to 1:25
(w/w), NBDCOQ:phospholipid}. The spot bleaching technique
was also applied to the large vesicles. The preliminary
results gave a diffusion coefficient an order slower than
that measured by the edge bleach analysis. The discrepancy
could be because the larger vesicles were multilamellar, or
that the bleaching conditions were not optimized
(overbleaching). Further experiments to determine whether
there is actually a difference between the edge and spot
bleach analysis in this situation have not been carried out,
but extensive analysis in other systems (Swaisgood &
Schindler, unpublished observations) indicate that this is

normally not the case.

Incorporation of the fluorescent probes into the

 

membrane Incorporation of NBDCOQ or NBDPE was carried out
as described in methods. Incorporation of hydrophobic
molecules into a membrane seems to be a complex process
dependent on a number of parameters such as the structure of
the molecule (Struck 8 Pagano, 1980) and the nature of the

membrane system into which the molecule is partitioning

38

(Arvinte pp 51., 1986). Extent of incorporation of
fluorescently labeled phospholipids is dependent upon
whether the fluorescent tag is located in the head or the
acyl region (Struck & Pagano, 1980). Acyl labeled

phospholipids incorporated more easily than the head group

labeled phospholipids. Q10 is a hydrophobic molecule and
has a very low critical micelle concentration (CMC)
(<0.05 x 10-7 M, Lenaz & Degli Bsposti, 1985) and does not

partition into a membrane easily. A model for the
incorporation of Q10 into the membranes has been proposed by
Lenaz & Degli Bsposti (1985) according to which conditions
favoring the formation of monomers of Q10 such as increase
in temperature or sonication results in better
incorporation. NBDCOQ, because of the RED moiety, is
somewhat less hydrophobic and hence its CMC may be slightly
higher. However, under a variety conditions tested the
amount of incorporation was not sufficient to carry out the
FRAP experiments. It was not possible to sonicate the
system or increase the temperature as it would result in
damaging the intactness of the mitochondrial membrane.
Sufficient incorporation could be achieved in presence of
BSA (see methods) which seemed to act as a carrier protein
for delivering NBDCOQ into the mitochondrial membrane.

Final concentration of NBDCOQ in the membrane was around 2—6
mol X with respect to lipid (>10 times in excess of

endogenous Q As the amount of NBDCOQ incorporation was

10)'
low, the levels to be quantified was limited by the

39

sensitivity of the measurement and there is some error in
the measurement as indicated by the range of values obtained
and by the fact that the sum of the amounts found in the
supernatant and pellet was around 70-80% of the added probe.
For the same apparent levels of NBDCOQ in the mitochondria
and vesicles, fluorescence intensity was greater in the
vesicles (3-4 times). A lower fluorescence intensity of
HBDPB in the mitochondria as compared to vesicles was also
seen. This seems to indicate that some quenching of N80 is

occuring by the mitochondrial proteins.

Diffusion coefficients of the fiuorescence_probes in the

membranes
Diffusion coefficients of NBDCOQ were 3.1 (11.0) x10-9
cmzsec-1 in the mitochondria and 1.1 (10.2) xlo-8 cmzsec-1

.in the vesicles. The diffusion of NBDPB was similar in both
the systems (Table 1). The recovery of the fluorescence
after photobleaching of NBDCOQ was monophasic (Figure 10)

and was more than 903 both in the mitochondria and vesicles.

glporescence emiggion of the_probes in the membranes
Table 2 shows the fluorescence emission maxima of
NBDCOQ, NBDPB and IANBD measured in different organic
solvents, in phospholipid vesicles and mitochondrial
membranes. The emission maximum is often a sensitive

indicator of the polarity and structure of the environment

surrounding the fluorophore. The fluorescence emission

40

 

Table 1: Lateral diffusion of NBDCOQ and NBDPE
in mitochondria and phospholipid vesicles

 

 

NBDCOQ

NBDPE

D (cmzsec-l)
Mitochondria Vesicles
-9 -8
3.1:1.0 x10 (n=4) 1.1:0.2 x10 (n=6)

3.0:o.4 x10-9(n=4) 1.3:o.3 x10-8(n=3)

 

 

41

Figure 10. Fluorescence recovery profiles of NBDCOQ in
mitochondria (A) and vesicles (B). Diffusion coefficients
were measured and calculated using the edge bleach analysis
technique as developed by Koppel gt pl. [Biophys. J. 391,
187,(1980)]. 100 microliters of mitochondria (10-15 mg
prot/ml) was incubated with 5 microliters of 1x BSA and 3
microliters of NBDCOQ (3 mg/ml) at room temperature for 15
minutes. 20 microliters of the swollen mitochondria were
placed on a glass slide, sealed with wax and used for
diffusion measurements. For phospholipid vesicles (2.5 mg
lipid/m1) 15 microliters was taken. Each point along the
x-axis represents a 0.5 second scan. The symbol (u)
represents the proportion of fluorescent probe on the
bleached edge of the mitochondria at time t and is defined
as the normalized first moment of the unbleached fluorophore

concentration distribution.

M(mp)

-|.900

-2.300

-2.700

-2.700

-3.300

-3.900

Figure 10

 

 

 

 

 

_

 

2|
Scan Number

3|

43

 

Table 2. Emission maxima of fluorescent probes in organic
solvents, phospholipid vesicles, and mitochondrial

 

 

membranes. (Excitation wavelength :468 nm)
Emission maximum (nm)

Solvent NBDIA NBDCOQ NBDPE
032812 520 521 519
n-Propanol 531 530 530
CHCl3 516 516 519
EtOH 532 531 531
Acetone 530 529 528
MeOH 535 533 ‘ 533

H20 552 560 550
Mitochondria 536-538 527-529 535-537
Vesicles 542-545 526-528 536-537

 

 

44

maxima of the NBD fluorophore attached to these different
molecules showed similar changes in response to different
solvents. In general, there was a blue shift as the
polarity of the solvent decreased. There were exceptions to
this: for example the emission maxima of the fluorophores in
chloroform were more blue shifted than in dichloromethane
and propanol. This implies that the relationship between
the emission maxima and the properties of the organic
solvents is more complex and could be dependent on
parameters such as whether the solvent is protic or aprotic,
and its dipole moment. However, when the fluorescent
molecules were placed in an anisotropic environment, namely
in vesicles and mitochondria, significant differences were
observed in the emission maxima obtained. The emission
maximum for NBDCOQ was 525-527nm in both the vesicles and
mitochondria, significantly blue shifted compared to NBDPE
(535-537 nm) and IANBD (342-544 nm), implying that NBDCOQ is

in a more hydrophobic environment.

Quenching of the fluorescent probes in the_mgmbranes
Quenching of fluorescent molecules has proved useful
for establishing the relative position of the fluorescent.
probes in a membrane. The data is presented in the form of
Stern-Volmer plots where the quenching efficiency is related

to the total quencher concentrations (Stern & Volmer, 1919).

I
-—9 = 1 + k t [Q] (1)
I

45

where Io and I are fluorescence intensities in the absence
and presence of the quencher, [Q] is the quencher
concentration, kq is the bimolecular quenching constant, and
to is the lifetime of the fluorophore in the absence of the
quencher. However, the quenching process in the membrane is
complicated, depending on the extent to which the quencher
binds or partitions into the membrane (Blatt g; 31., 1986).
Also, quenching in the membrane can occur either by dynamic
or static mechanism or both. Static quenching occurs when a
fluorophore within a spherical volume surrounding the
quencher is quenched instantaneously, while fluorophores
located outside the active sphere may be quenched by
collisional interactions (Blatt g; a1., 1986). Cu2+ is a
water soluble paramagnetic quenching agent that has been
shown useful for locating the relative depth of fluorophores
in a membrane by virtue of its ability to bind to the
membrane surface and quench the fluorescence of the probes
located on or near the surface (Thulborn & Sawyer, 1978).

Thus Cu2+

was chosen to investigate the relative position of
NBDPE and NBDCOQ in the mitochondrial membranes and
vesicles. Stern-Volmer quenching plots are given in Figure
11. As the experiments were performed under nearly
identical conditions for both NBDPE and NBDCOQ in the
vesicles and mitochondrial membranes, we can assume that the

2+ will be the same for both the

amount of bound and free Cu
probes.

In vesicles it is observed that quenching of head group

46

FIGURE 11. Stern-Volmer plots for quenching of NBDCOQ (. O)
and head group labeled NBDPE (A A) in giant mouse
liver mitochondria (open symbols) and phospholipid
vesicles (closed symbols). Swollen mitochondria
(0.3 m1, prepared as described in the legend to
Figure 10) incubated with the fluorescent probe
were diluted to 1 ml for the fluorescence measurements.
Vesicles containing the fluorescent probes (1:50, w/w)
were prepared according to Mueller g; 31. (1983) and
were used as such for the fluorescence measurements.
Steady-state fluorescence was measured in a Perkin—
Elmer 550-60 spectrofluorimeter. Excitation was at 468
nm and the emission was at 527 nm for NBDCOQ and at 536
nm for NBDPE. Slit width was 5 nm for the excitation

beam and 10-13 nm for the emission beam.

Figure 11

 

o.ov o.om o.o~ 0.3 0.0
_ _ . _ L ad
0 O .0 O O O. a. 4‘1.
‘x 4
I . 1F;
4
.. . In;
4
.. . .. m;
4 <
I I he
_ _ _

 

 

 

48

labeled NBDPE was much higher than NBDCOQ (Figure 11). The
plots are non-linear and show apparent saturation at higher
Cu2+ levels. This behavior is expected in cases where total
concentration rather than bound concentration is used on the

2+ is assumed to be the most

abscissa and where the bound Cu
effective quencher (Thulborn & Sawyer, 1978). To calculate
apparent Stern-Volmer quenching constants, slopes were
calculated from the initial changes in fluorescence at low
quencher concentrations, since it can be assumed that most
of the quenching is occuring by a dynamic mechanism. In
Table 3, it can be seen that the apparent k8v values are

5 to 6 fold higher for NBDPE and NBDCOQ. In mitochondria,
the difference is even greater (>30x), in fact there is
essentially no quenching of NBDCOQ while NBDPE is totally
quenched by a mainly static mechanism revealed by the upward
slope of the Stern-Volmer plot (Figure 11). (It is possible

2+ to the protein of the mitochondrial

that the binding of Cu
membranes may contribute to the static quenching observed.
Regardless of the different quenching patterns, it is clear
from the overall data that NBDCOQ is much less accessible to
Cu2+ than NBDPE in both the membrane systems.

Extent of quenching can be obtained by using the
modified Stern-Volmer equation proposed by Lehrer (1971).

I 1 1
o

—= 1.-

10-1 [Q].fa.K fa

 

 

where fa is the fraction of the fluorophore accessible

to the quencher. A plot of Io/Io—I vs 1/[Q] yields a

Figure 12.

49

Lehrer inverse plots of the quenching of NBDCOQ (.)
and NBDPE (AA) in giant mitochondria (open symbols)
and phospholipid vesicles (closed symbols). Data
plotted according to the modified Stern-Volmer
equation. The x-intercept gives the fraction of the

fluorophore accessible to the quencher.

Figure 12

mwd

 

 

l 0.0

10.4%

 

 

 

51

 

Table 3: Stern-Volmer (ks quenching constants (uM) 1of NBDCOQ
and NBDPE in the mitoséhondria and phospholipid vesicles.

 

 

 

mitochondria vesicles
NBDCOQ NBDPE NBDCOQ NBDPE
k 0.310.5(ns6) 10.811.9(ns7) 7.0:1.1(n=5) 4114(n=4)

8V

fa‘“) a 100 17:3 50:9

 

 

 

52

straight line and the intercept gives the extent of
quenching at infinite concentrations of the quencher.
Modified S-V plots are given in Figure 12. The intercepts
indicate that about 40—503 of NBDPE is quenched in the
vesicles and all of it (100%) is quenChed in the
mitochondria. For HBDCOQ, about 15-20% can be quenched in
the vesicles but none is quenched in the mitochondria.

In the vesicles, which were formed after the probe and
phospholipid were mixed, about 50% of the probe would be
expected to be located in the outer leaflet of the bilayer
and thus accessible to the quencher (as observed for NBDPE).
In the case of mitochondria, the probe is added exogenously
and would be located in the outer leaflet since the
flip-flop process for phospholipids is very slow in the
natural membranes. In the case of NBDCOQ, it appears to
partition into a hydrophobic environment in either membrane
system and to be relatively inaccessible to Cu2+. Our

results support the idea that it may be predominantly

stationed near the mid~plane of the bilayer.

Effect of the fluorescent probes on electron transfer in the
mitochondria

The effect of NBDPE and NBDCOQ on uncoupled electron
transfer from succinate to oxygen was measured (Table 4).
The fluorescent probes had no affect on electron transfer at
concentrations similar to those used for mobility

measurements. Since the head group of Q10 is modified, it

53

would not be expected to bind at the Q-binding sites of the
electron transfer and hence would not be able to compete

with the endogenous Q10.

54

 

Table 4. Effect of the fluorescent probes on succinate

oxidase activity in mitochondria

 

 

 

Treatment Relative Activity (%)

5,11a lQpl
controlb
(ethanol) 100 100
NBDCOQC
(3 mg/ml in ethanol) 105:5 (n=3) 115:5 (n=2)

NBDPE
(1 mg/ml in ethanol)

 

97:7 (n=4)

 

105113 (n=4)

 

a

b

5 and 10 pl are the volumes of ethanol or NBD solution
added to mitochondria before measuring activity

Activity obtained with ethanol is taken as 100% activity

C

Amount of NBDCOQ in the membrane is similar to the
levels present during FRAP measurements.

 

DISCUSSION

Diffusion of NBDCOQ in vesicles and mitochondria
Diffusion of NBDCOQ was 3.1 i (1.0) x10"9 cmzsec-1 in

the mitochondrial membranes and 1.1 1-(0.2) x10.8 cmzsec”1
in very large unilamellar vesicles. Diffusion of NBDPE was
similar. The diffusion coefficient we report is close to
that obtained by Hackenbrock's group for a fluorescent
derivative of a ubiquinone analog which had an alkyl side
chain corresponding to only two isoprene units. It is well
established that the short chain ubiquinones orient parallel
to the phospholipids near the lipid-water interface.
However, evidence suggests that long chain ubiquinones
predominantly reside deeper in the membrane. One of the
concerns in using the fluorescent derivative of Q10 was
whether the NBD moiety might cause the quinone moiety to be
pulled nearer to the membrane surface. To address this
question, we have compared the fluorescence emission
characteristics of NBDCOQ, NBDPE, and IANBD. They had
similar emission maxima in a variety of organic solvents,
but the emission maximum of NBDCOQ in membranes was
significantly blue shifted compared to the fluorescent
phospholipid, implying it is in a more hydrophobic medium.
The quenching of NBDCOQ and NBDPE by a water soluble
quenching agent, Cu2+, was also compared in both the
membrane systems. The quenching of NBDCOQ was much less

implying it is less accessible. These studies indicate that

55

56

the fluorescent derivative is indeed located deep in the
membrane. Similar studies were also carried out with 6-NBDPE
and 12-NBDPE (results not shown) but their quenching
behavior and fluorescence emission maxima were similar to
head group labeled NBDPE, as also observed by Chattopadhyay
and London (1987). They postulate that the NBD moiety of
6-NBDPE and 12-NBDPE loops back and preferentially locates
itself at the membrane surface, while in their studies with
NED-cholestrol, the RED moiety is deeply embedded in the
membrane, presumably because the rigidity of the cholestrol
structure prevents the fluorescent moiety from looping back
as observed for 60-NBDPE and 12C-NBDPE. For NBDCOQ, it
appears that any tendency for the N80 moiety to draw the
quinone moiety to the membrane surface is counteracted by
the hydrophobic and rigid character of the isoprene side
chain. Nevertheless, the similarity between NBDCOQ and
NBDPE mobilities suggest that the diffusion of Q10 is
limited by the diffusion of the phospholipid. This seems to
be in agreement with the free—volume theory postulated by
Cohen and Turnbull (1959), which predicts that a solute can
diffuse no faster than its solvent since a diffusive step of
the solute is only completed when a solvent molecule moves
into the void left by the solute (Vaz g3 al., 1984). If Q10
diffuses at the same rate as a phospholipid, it would imply
that Q10 does not occupy a completely independent phase but
rather that it is solvated to some degree by the

phospholipid. It is possible that if Q10 is strictly

57

located in the mid-plane of the bilayer, it could diffuse
very rapidly (Millner & Barber, 1984) since the resistance
would be minimal and such diffusion would not be limited by
diffusion of the phospholipids. A model consistent with
these data would have the major part'of the isoprene chain
in the mid-plane while the head group penetrates to some
degree among the phospholipid acyl chains. Alternatively,
even if totally restricted to the mid—plane, there is no
evidence for a fast diffusion rate in such a phase.

Fato g1 p1. (1985, 1986) using fluorescent quenching
techniques report a diffusion coefficient for Q10 as
1 x10"6 cmzsec-l, which is 2-3 orders of magnitude higher
than diffusion values reported by the technique of FRAP.
Fluorescence quenching supposedly measures only short range
diffusion while FRAP measures long range diffusion and it is
argued by Lenaz and co-workers that only short range
diffusion is of physiological significance (Fato 33 $1.,
1986). However long range diffusion measurements in
vesicles should be similar to short range diffusion as there
are no proteins to retard the movement (Vaz 35 31., 1984)
and a diffusion value of 1 x10-6 cmzsec"1 is still two
orders of magnitude higher than what we observe by FRAP in
artificial vesicles. Diffusion rates in the order of

10-5-10.8 cmzsec“1 have been measured by the technique of

5 -7 2

FRAP. Diffusion rates in the order of 10- -10 cm sec"1

have been measured for a fluorescent fatty acid in

phospholipid monolayers (Teissie g1 31., 1978). The authors

58

propose the relative slower diffusion in the bilayer
membranes could be explained by each leaflet having a drag
effect on the other which brings about slowing down of the
translational diffusion of the molecules. In addition,
Fahey 8 Webb (1978) have measured diffusion of
dioctadecylindotricarbocyanine (DiI) to be 10-7 cmzsec-1 in
black lipid membranes and 10-8 cmzsec-1 in large bilayer
vesicles. These results indicate that the diffusion rates
we measure are not limited by the FRAP technique.

With regards to the collisional quenching analysis by
Lenaz and coworkers, numerical analysis by Blackwell g; l.
(1987) has shown diffusion-controlled quenching results in a
non-linear concentration dependence for diffusion

6 cmzsec-1 and

coefficients less than or of the order of 10—
hence lateral diffusion coefficients in membranes are
typically overestimated by an order of magnitude or more.
They present an alternative empirical method to calculate
the diffusion coefficients and have determined the diffusion

of plastoquinone and plastoquinol to be 1.3-3.5x 10-7

cmasec-1 in phospholipid membranes. It appears that the
very high diffusion coefficients reported by Fato 55 a1. are
likely to be overestimates for the reasons pointed out by
Blackwell _1 _1. (1987). They use the modified
Smoluchowski's equation (Lakowicz and Hogan, 1980) which
assumes the quenching and quenched molecules are spherical

and tumbling isotropically in the membrane. The diffusants,

Q10 and anthroyloxy fatty acids are not spherical and

59

membrane diffusion is likely to be 2-dimensional rather than

3—dimensional, and the orientation is unlikely to be

isotropic. In fact, it has been shown that the motion of

the anthracene ring of n—(anthroyloxy) fatty acids when“
incorporated into phospholipids, is hindered and rotational

constraints would make its motion highly anisotropic (Badley
t 1., 1973).

Thus it is reasonable to conclude that the value of
10"6 cmfisec-1 originally calculated by Fato g1 a1. (1986) is
an overestimate by atleast an order of magnitude (as
discussed by Blackwell 31 a1., 1987), but that still leaves
an order of magnitude difference between the FRAP analysis
and the collisional quenching analysis for Q10 diffusion in

artificial membranes, which cannot be accounted for by the

interference of proteins in long range diffusion.

Location of ubiguinone41n the membranes

Fluorescence quenching and emission studies of NBDCOQ
in our lab are consistent with a location for NBDCOQ deep in
the membrane. The emission maximum of NBDCOQ is blue
shifted at least by 10 nm compared to a head group labeled
phospholipid in mitochondria and vesicles. In mitochondria,
no quenching of NBDCOQ was observed while NBDPE was totally
quenched indicating a location for NBDCOQ near the center of
the bilayer. However, this does not rule out some
transverse motion of the head group as our studies are not

capable of locating such motion.

60

Though a variety of physical techniques have been
applied to probe the location of ubiquinone,its location in
the membranes remains unclear. Most of the studies suggest
that ubiquinone is located in a hydrophobic region but there
is little agreement as to orientation, state of aggregation,
and the possible existence of more than one population in
the membrane. NMR has been used by a number of
investigators to locate the position of ubiquinone in
unilamellar and multilamellar vesicles and the chemical
shift values of the methoxy resonances has proved valuable
in these studies. Short chain ubiquinones generally exhibit
a single chemical shift (3.96ppm) for the methoxy resonances
while the long chain ubiquinones (Q9_10) showed a doublet
(3.96,3.80ppm). The chemical shift at 3.96 ppm is similar
to the short chain ubiquinones, while the methoxy resonance
at 3.80 ppm is up field shifted implying it is in a more
hydrophobic medium. The relative intensity of the two peaks
seems to be dependent on a number of parameters including
the composition of the phospholipid, the diameter of the
phospholipid vesicles and whether the phospholipid vesicles
were unilamellar or multilamellar. Kingsley and Feigenson
(1981) have performed NMR experiments using DMPC
perdeuterated unilamellar vesicles. They observe the up
field shift of the methoxy resonance only at concentrations
>1 mol%. They interpret their results to mean that, the
down field peak (3.96ppm) is from ubiquinone dispersed in

the phospholipid layer while the up field peak is from

61

ubiquinone in a separate ubiquinone rich phase. Although
ubiquinone may exist as a separate phase, it seems to
experience considerable local motion since the line widths
were narrow. The authors also propose that two types of
vesicles may exist with different ubiquinone/phospholipid
ratios. Ulrich £1 31. (1985) make similar observations in
their studies using DMPC and DPPC unilamellar vesicles.
They also measured the reduction of entrapped ferricyanide
in vesicles containing Q2 and Q10 by externally added
reductants and they observe that reduction by Q10 containing
vesicles was relatively less and propose that one essential
property of the long chain quinones is their residence in
the hydrophobic core. Michaelis and Moore (1985) have
performed NMR studies of Q10 in egg phosphatidylcholine
(having polyunsaturated fatty acids of differing acyl chain
lengths) to see if the location of Q10 is different than
that seen by Kingsley & Feigenson (1980) and Ulrich £1 21.,
(1985). After short sonication times (20 minutes), they
observe that most of the lipid is in 1000 A0 vesicles and
some in vesicles of size of 500-700 A0 and they also observe
two methoxy resonances, (3.82, 3.98 ppm) with most of the
signal at 3.98 ppm. However, as the sonication time is
increased, the relative number of small vesicles increases
and after 150 minutes only small vesicles are present
(500-700 A0) and the low field signal (3.98 ppm) almost
disappears. They also studied the affect of lanthanide

shift reagents on methoxy resonances and they do not see any

62

effect. They conclude that Q10 is incorporated at least
beyond C-2 of the acyl chain in the membrane and the
appearance of the two methoxy resonances is a curvature
affect. However, the latter observation may just reflect
partitioning of Q10 into the membranes as these studies were
performed at 200 C while incorporation in the other studies
(Kingsley & Feigenson, 1981; Ulrich 31 31., 1985) was
performed at higher temperatures.

Quinn's lab (0ndarroa & Quinn, 1986) has studied the
location of Q10 in multilayer liposomes and they observe
only one signal (at low field, 3.80 ppm) for the methoxy
resonance up to concentrations of 20 mol %. Their studies
(Katsikas & Quinn, 1981) and those of others (Aranda 8
Gomez-Fernandez, 1985), using differential scanning
calorimetry have shown that Q10 up to 10 mol % does not
effect the Tm of the phospholipid and it has been suggested
that Q10 is preferentially located between the monolayers of
the bilayer. Their NMR studies show some interaction
between the Q10 molecules in the membrane but they also and
others, (Kingsley & Feigenson, 1981; Ulrich 31 31., 1985)
observe high resolution of the methoxy resonances which
implies significant local motion. Hence they propose a
alternative to the mid-plane model, where micelles or
aggregates of Q10 can exist, which has a diameter similar to
that of the bilayer. However, they agree there is not
enough evidence in favor of one or the other.

Using X-ray diffraction, electron microscopy, and

63

130 NMR, Stidham 31 31. (1984) propose a

13

especially
different model. Their C NMR spin-lattice relaxation
measurements show a marked effect of the ubiquinone on the
lipid hydrocarbon chain atoms near the polar region and no
effect on the atoms at the end of the acyl chain, which
provides evidence that Q10 head group is situated near the
polar end of the lipid hydrocarbon chain. This may
represent only a fraction of Q10, the remaining Q10 existing
as a separate phase in the mid-plane of the bilayer and it

is conceivable that the perturbation by of the terminal

010
atoms is not sufficient to affect the relaxation times.
However Ulrich 33 31. (1985) observe the perturbation of the
terminal proton resonances of the hydrocarbon chain by Q10.
These data again raise the issue of two populations of
quinone in the membrane. The relaxation measurements of
Stidham 31 31. (1984) may be detecting a small population
that may not be measured by other techniques. Indeed,
surface pressure studies indicate that only a small

percentage of the could intercalate with the

Q10
phospholipids and be present near the membrane surface
(MacDonald, 1987).

MacDonald (1987) discusses partitioning of hydrophobic
molecules into a membrane in terms of thermodynamics. He
proposes that molecules having the same surface pressure as
a phospholipid can freely intercalate with the

phospholipids. However if the surface pressure is low,

intercalation is not favorable and most of the molecules

64

would partition deeper into the membrane. The surface
pressure of Q10 is fairly low (12 dynes/cm) compared to the
surface pressure of a phospholipid (35 dynes/cm) (Quinn &
Esfahani, 1980) and hence it is likely that major portion of
Q10 partitions deep into the membrane. It is likely that
the small fraction of Q10 which is located near the surface
of the membrane may be of functional significance as the
oxidized and reduced quinones seem to have different
locations in the membrane. Aranda 31 31. (1986) observe
that the reduced and oxidized Q10 behave differently in
their studies using DPPC multilamellar vesicles. Using
Fourier transform infra red (FT-IR) spectroscopy, they
observe that reduced Q10 affected the transition temperature
of DPPC (Tm=38.5°C) while Q10 had no effect (Tm=41.5°C).
They (Aranda & Gomez-Fernandez, 1986) also observe that
reduced Q10 increases the anisotropy of diphenylhexatriene

while had no effect. Kingsley & Feigenson (1981) and

Q10
Ulrich 33 31. (1985) also observe that the methoxy
resonances are down field shifted when Q10 is reduced
implying that Q10 in the reduced state lies closer to the

membrane surface.

Schindler (1980) has measured the surface pressure as a
function of the diameter of the vesicles. His studies show
that surface pressure increases parabolically from diameters
of 300 to 800 A0, after which it remains a constant. This
implies that the surface pressure of the sonicated vesicles

and multilamellar liposomes will differ if the former have

65

diameters<800 A0. This may account for the appearance of
two methoxy resonances in the studies of Kingsley &
Feigenson (1980) and Ulrich 31 31., (1985). They use either
DMPC (14:0) or DPPC (16:0) unilamellar vesicles, and these
vesicles would tend to be small (diameters 5300-400 A0) and
have low surface pressures allowing more intercalation of
Q10 between the phospholipids. At a concentration of 1 mol%
of Q10 in the studies of Kingsley & Feigenson (1980), the up
field shift was virtually absent implying that all of the
Q10 was dispersed in the bilayer. At higher concentrations

of Q the intensity of the down field signal increases

10'

suggesting that most of the Q is located deep in the

10
membrane. Similarly, Ulrich 31 31. (1985), observe the high
field signal only at concentrations higher than 2 mol%. The
appearance of only one methoxy resonance in the studies of
Michaelis & Moore (1985) can also be explained. The
diameter of the vesicles used in their studies was in the
order of 500-700 A0 and would have higher surface pressures
and hence would exclude most of Q10 leading to the only
methoxy resonance at 3.82 ppm.

Q10 is present to the extent of 1-2 mol % in
mitochondria. The inner membrane of the mitochondria is
highly convoluted and hence may exhibit curvature effects
and in addition, the inner membrane has a high protein
concentration and the presence of Q-binding proteins may

affect its orientation and location. The surface pressure

of a biological membrane is considered to be the same as

66

that of a planar bilayer but higher values have also been
suggested (Conrad & Singer, 1981). The spectrophotometric
detection of Q10 in native membrane has suggested a location
in the hydrophobic region (Chance, 1965). Chatelier and
Sawyer (1985) have tried to locate the position of Q10 in
the native membranes by fluorescence quenching techniques
using n-anthroyl fatty acids and they conclude that there
are two populations, one near the center and one near the
surface of the bilayer. However, for their studies, they
measure the quenching of exogenously added Q10 and it is
known that Q10 does not easily partition into the membrane
(Degli Esposti 31 31., 1981) and it is possible that the
quenching observed by the short chain anthroyl fatty acids
could be that of Q10 adhering to the surface of the
membrane. It is still possible that two populations of Q10
are present in the mitochondrial membrane, and a more
sophisticated and sensitive analysis should provide a more
definite answer.

Although there appears to be a considerable amount of
conflicting experimental data regarding the location of Q10
when all the potential sources of artifacts are taken into
consideration a reasonable picture emerges. The sources of
experimental difference have been incomplete incorporation,
size of the vesicles and hence the difference in internal
pressure, and the limitation and sensitivity of the
techniques applied. We postulate that most of the evidence

is consistent with majority of Q10 occupying a position in

67

the mid-plane of the membrane but in equilibrium with a
small portion (especially the reduced form) lying close to
the membrane surface and this may be of functional

significance in electron and proton transfer.

_1ectron tr3nsfer through 9 pool.

Rate of electron transfer through Q segment of the
electron transport chain has been observed to follow the

simple relationship

vobsd

where is the overall observed rate, vred is the rate

vobsd
of quinone reduction and Vox is the rate of quinone
oxidation. The above expression is known as the homogeneous
pool equation. Evidence that Q10 acts as a homogeneous
quinone pool has been obtained by a number of investigators
including Kroger and Klingenberg (1973a, 1973b): Ragan &
Heron (1978); Trumpower (1978); Unden & Kroger (i981); and
Schneider 31 31., (1982). A simple Q pool behavior can be
explained as follows. In a system in which Q10 acts as a
freely and rapidly diffusing entity between reductases and
oxidases, which are themselves randomly diffusing in the
membrane, it follows that electrons donated into the Q pool
by any quinone reductase can be withdrawn with equal
probability by any quinol oxidase. The quinone redox state
is therefore a function of the total activity of the quinone

reductases and oxidases, regardless of the numbers of

molecules of Q10 actually present. Any restrictions on

68

quinone and protein diffusion or a non-random distribution
of proteins in the membrane, could lead to deviation from
simple Q pool kinetics (Ragan & Reed, 1986).

Kroger & Klingenberg (1973a, 1973b) have measured the
redox kinetics of Q10 in uncoupled mitochondria and have
correlated this to the overall rate of electron transfer.
They found that more than 80% of Q10 responded to changes in
respiratory activity and the response was kinetically
homogeneous. Ragan and Heron (1978, Heron 31 31., 1978)
have studied the electron transfer between complex I and
complex III in artificial membranes at different
protein:phospholipid ratios and at different complex
I:complex III ratios. At high protein concentrations, the
pool behavior was lost and this was attributed to the
formation of associations between complexes I and III at 1:1
molar ratios which permitted direct electron transfer
through the bound quinone. When the experiments were
performed at lipidzprotein ratios representative of the
native membrane, the pool behavior was observed. They
proposed that electron transfer from NADH to cytochrome 3
takes place only through complex I-complex III units which,
however, are formed and reformed at rates that exceed the
rates of electron transfer from complex I to complex III.
This proposal accounted for Q pool behavior and flexible
stoichiometry of dehydrogenases to cytochromes, but still
assigned a special function to the association of complex I

and complex III. However, in a later review (Ragan & Reed,

69

1986), the importance of the experiments performed at high
protein concentrations was discounted, since these were non-
physiological conditions and might not reflect the
organization of the electron transfer components in the
native membrane. Pool behavior of Q10 is also indicated in
experiments of Schneider 31 31. (1982) where increasing the
phospholipid content of the mitochondrial membranes
decreased the electron transfer rates from NADH and
succinate dehydrogenase to cytochrome 3. This effect was
reversed when Q10 is added along with the phospholipids.

Yu's lab (Yu & Yu, 1980) and King's lab (King & Suzuki,
1984) promote the idea that functionally active ubiquinones
are associated with apoquino proteins and therefore all
quinone mediated reactions occur through protein bound redox
reactions. They rule out any role for freely diffusing Q10
in electron transfer. Quinone binding proteins have been
identified in complexes I, II and III and and it is proposed
that such binding proteins are the actual electron carriers
in ubiquinone dependent electron transfer (Yu & Yu, 1980,
19323, 1982b; Yu 31 31., 1985; and Suzuki & King, 1983).

The mitochondrial membranes seem to be saturated with

the levels of Q normally present in the membrane. is

Q10
present at a total mole fraction of around 1% with respect
to the phospholipid concentration,corresponding to an
average volume concentration of around 20 mM in the

hydrocarbon region of the membrane (Mitchell & Moyle, 1985).

Norling t al. (1974) have systematically studied the effect

70

of depletion and repletion of endogenous Q10 in beef heart
mitochondria on NADH and succinate oxidase activity. They
have devised a procedure for addition of Q10, where all of
the added Q10 seems to incorporate into the membrane. 0n
depletion, the activity measured was almost zero and an
addition of Q10 to levels present in the membrane, the
activity was restored to 100%. Further incorporation did
not result in higher activity though all of the Q10 seemed
to undergo reduction consistent with the Q pool mechanism.
Studies with yeast mutants (Beattie & Clejan, 1986) have
shown that succinate-cytochrome 3 reductase activities can
be restored by addition of Q6 to levels comparable to the
wild type. However, further activity could not be
stimulated by Q6 either in the wild type or in the mutants.
In contrast, short chain Q analogs could stimulate activity
many fold. These studies emphasize the importance of the
structural specificities of biological quinones and suggest
that studies with Q analogs may be too simplistically
interpreted. Zhu _1 _1. (1982) find simple Q-pool behavior
even when a large proportion of Q10 was extracted. Studies
in cultured neuroblastoma cells (Maltese & Aprille, 1985)
where the levels of endogenous Q10 were reduced nearly to
50% by blocking the de novo synthesis of Q10, did not affect
the succinate-cytochrome 3 reductase activity, implying that
Q10 may not be rate limiting. However, they do observe a
decrease in cell cycling and hence they suggest that Q10 may

play a role in the mitochondrial and cell proliferation

71

other than in energy transduction.

In general, the majority of the available data support
a pool behavior of ubiquinone under many conditions,
implying that diffusion is reasonably fast (310"9 cmzsec_1)
in agreement with the numbers measured in our experiments.
However, deviation from homogeneous pool behavior has been
observed in a variety of experimental situations (Gutman,
1985; Ragan & Heron, 1978; Cottingham & Moore, 1983; also
see discussion in pages 74-80). The question remains as to
whether the diffusion of Q10 is always free and random or
whether it may be segregated into more than one pool under
some conditions. It is also a major issue whether the
diffusion of Q10 is ever (or always) rate limiting in

electron transfer.

Is diffusion of 918 rate limiting in electron transfer?

Hackenbrock's group from their diffusion measurements
as a function of membrane dilution with phospholipids
(Schneider 31 31., 1982) and as a function of temperature
(Hackenbrock _1 _1., 1986b), propose that diffusion of Q10
is rate limiting in electron transfer. They measured the
(activation energy (Eact) for diffusion of Q10 by measuring
the temperature dependence of the diffusion of DiI (dioctyl
or dihexyldecylcarbocyanine; a fluorescent lipid analog)
under the assumption that 011 and ubiquinone will have the
same temperature dependence for diffusion. The Eact for D11

and complex III were 12.04 kcals/mole and 10.8 kcals/mole,

respectively. They have measured Ea for uncoupled

ct

72

succinate-cytochrome 3 reductase activity to be 12.9
kcals/mole. Comparing the temperature dependency of the
overall kinetic process to the temperature dependency of the
overall diffusion steps, they infer that diffusion of Q10
and its redox partners is rate limiting in electron
transfer. However, the assumption that temperature
dependence of DiI diffusion will be the same as native
quinone is questionable since Q10 appears to occupy a
separate phase in the membrane. Furthermore, if the protein
distribution is non-random (see discussion on pages 74-80)
quinone may not always be diffusing long distances and hence

the Eac measurements may not reflect the real picture.

t

Minimum Eac for diffusion of a phospholipid is around

t
7 kcals/mole when measured in various DPPC multilayer
liposomes (Vaz & Hellman, 1983). Thus the Eact for
diffusion of Q10 in the membrane may vary from 7-12
kcals/mole.

Another line of evidence presented by Hackenbrock's
group (Schneider 31 31., 1982) for a rate limiting role of
Q10 diffusion is the observation that electron transfer
rates are lowered when mitochondrial membranes are enriched
with phospholipids which is reversed when Q10 is
incorporated into the membrane. Fato 31 31. (1986) present
an alternative explanation for the decrease in electron
transfer with membrane dilution. They propose the decrease

is due to the concentration of Q10 falling below the km for

the partner enzymes. The concentration of Q10 in the

73

mitochondrial membrane is 20 mM (Mitchell & Moyle, 1985).
The km values of complex III and complex I for Q10 in the
mitochondrial membrane has been calculated to be 2.4 mM and
32 mM respectively and the authors argue that the known
turnovers of complex I and complex III are sufficient to
explain the decrease in activity of NADH:cytochrome 3
reductase due to dilution.

A rate limiting role for Qlo (Crofts & Wraight, 1983)
has also been proposed from the studies of chromatophores
from Rhodopseudomonos sphaeroids. They measured the second
order rate constant of the reduction of cytochrome b561 by
the quinol produced by the photoreduction in the reaction
center and calculated a diffusion coefficient of 10.10 cm2
sec-1 assuming that diffusion is rate limiting. They argue
that the quinol release from the reaction center or its
binding to the cytochrome complex are not rate limiting,
suggesting that these steps constitute only a minor
contribution to the rate constant for the reaction.
Consistent with a rate limiting role for diffusion of Q10,
they also observed that there was a lag in cytochrome b
reduction at low quinol concentrations or when the distance
between reaction centers and cytochrome complex was
increased by dilution with phospholipids.

In stacked chloroplasts, there is considerable evidence
that the PS I and PS 11 units may be separated by large

distances, with PS I (along with cytochrome bsf) mainly in

the non-appressed regions and PS II concentrated in the

74

stacked membranes (Anderson, 1981). Plastoquinone mediates
the transfer of electrons between PS II and cytochrome b6f.
These redox complexes appear to be separated over long
distances and the distances to be travelled by quinone
during one redox cycle may be as high as 200 nm each way
(Millner & Barber, 1984). Hence a diffusion value in the

6 cm2sec-1 was postulated as essential for

order of 10-
electron transfer and it was argued that plastoquinone can
accomplish this if it is in the mid-plane of the bilayer.
However, it appears that it may not be necessary to ascribe
all of the electron transfer between PS I in the exposed and
PS II in the appressed region of the membrane to
plastoquinone, as cytochrome bf complex may equally be
distributed between stacked and unstacked membranes
(Cox & Andersson, 1981; Anderson, 1982).

Lenaz's group (Fato 31 31., 1985, 1986) have measured
the diffusion of Q10 by collisional quenching and report a

6 1

value of 1 x10- cmzsec_ They measured diffusion rates as

a function of temperature and calculate the Eact to be
1.5-2.0 kcal/mole in contrast to values of 12 kcals/mole
reported by Hackenbrock 31 31., (1986b). They further
calculate the second order rate constants for the reactions
of Q10 with complex III and complex I, making use of the km
values measured by other investigators. Comparing the
calculated second order rate constant with the bimolecular

collision constants determined by fluorescence quenching,

they conclude that diffusion cannot be rate limiting because

75

the latter constants are 3 orders of magnitude greater than
the former. However, the accuracy and applicability of any
of these numbers to electron transfer in mitochondria remain
to be determined.

Whether the diffusion of Q10 is rate limiting or not is
still an open question. Further insight into the ultra
structure of the mitochondrial membrane and mechanism of

energy transduction should provide a more definite answer.

Heterogeneity of mitochondrial membrane and and its
implica11on 1n e1ec1gon transfer through 913

Hackenbrock's group (Hackenbrock, 1976; Gupta 31 31.,
1984; Hackenbrock 31 31., 1986a) propose a random collision
model for electron transfer from their ultrastructural
observations and diffusion studies of redox components.
According to this model, all redox components are
independent lateral diffusants, diffusion is long range,
electron transfer is diffusion coupled and diffusion of the
electron transfer components including that of Q10 is rate
limiting in electron transfer. From earlier diffusion
measurements of cytochrome 3 and cytochrome-c oxidase, we
have postulated a dynamic aggregate model for electron
transfer in the mitochondrial membrane (Hochman 31 31.,
1983, 1985; Ferguson-Miller 31 31., 1986). According to
this model, electron transfer can occur by random diffusion

(Figure 13, top), but higher rates can be achieved by

Figure 13.

76

The dynamic aggregate model of the mitochondrial
respiratory chain. The major electron transfer
components are represented in various states of

aggregation (aa = cytochrome oxidase:

3
bc1 = cytochrome bcl; c = cytochrome 3;
fl = NADH or succinate dehydrogenase). The model

assumes that all components are diffusing in or
on the membrane, and that aggregates with
significant lifetimes are formed between correct

redox partners.

Figure 13

DYNAMIC AGGR EGATE MODEL

f-O-r

a. ‘5.” '55".
‘_ .-‘ - g 6" Q
I .: b c,

1
u

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a:

 

 

 

 

v‘ .
a.“ .2‘ .

dissociated \\
\ - ",9". ..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

complete aggregate

78

formation of complexes with life times that are significant
compared to electron transfer rates. A crucial distinction
between these models is that the former would predict a
random distribution of components of the respiratory chain
in the mitochondrial inner membrane, while the latter would
imply non-random organization, at least in the time scale
where multiple electron transfer can occur through the same
two components.

It has become increasingly evident that the molecular
constituents of biological membranes are not randomly
organized within the bilayer matrix and many of the physical
and functional properties are sensitive to the particular
way in which lipid and protein molecules are distributed in
the bilayer (Jain, 1983). Various types of ordered
molecular arrangements are possible, ranging from
non-specific aggregation or lateral phase separation to
highly specific molecular interactions leading to the
formation of complex structural patterns. In general, the
lateral distribution of molecules in a lipid bilayer is
dictated by the energetics of the interactions between the
various components.

Non-randomness in the membrane seems to be the rule
rather than the exception. Even a simple system of two
phospholipids show phase separation (Klausner 31 31., 1980).
In the mitochondrial membrane, different types of
phospholipids exist at different concentrations and it has

been shown that they are asymmetrically distributed between

79

the two monolayers of the membrane (Krebs 31 31., 1979) and
can exist in more than one phase in the mitochondrial
membrane (Venetie & de Verkeij, 1982). There is also
evidence for segregation of proteins in a number of systems
(Friend,1982; Goldstein 31 31., 1979; De Petris,1979;
Bretscher, 1983) including chloroplasts (Anderson, 1981;
Staehlin, 1976) and mitochondria (Werner & Neupert, 1972;
Silverstein & Rottenberg, 1987; Srere, 1982; Williams, 1983:
Sjostrand, 1983).

When rat liver mitochondria were fragmented and
separated by density gradient centrifugation, different
fractions of the gradient varied in several aspects,
especially in protein content and associated enzyme
activities (Werner & Neupert, 1972). Time resolved
anisotropy measurements in sub-mitochondrial particles have
shown that nearly two third of the membrane proteins may
exist as dynamic aggregates (Silverstein & Rottenberg,
1987). In the native state, mitochondria have cristae in
which the inner membranes appear to be in close apposition
under many conditions and separated approximately by

100-200 A°

(Srere, 1982). Intrinsic proteins extend into
this space as much as 70 A0. Observations from electron
microscopy (Sjostrand, 1983; Williams, 1983) indicate that
proteins are present in larger aggregates than single
electron transfer complexes, though accurate size estimates

are difficult from freeze-fractured studies. The

reconstitution experiments (Ragan & Heron, 1978) have shown

80

that at sufficiently high concentrations of complex I and
complex III, there is a stoichiometric association between
them. In such a case, no Q pool behavior was observed.
Differential scanning calorimetric studies (Gwak 31 31.,
1986) also have shown that there is-a preferential
interaction between complexes II and III. There is

also circumstantial evidence for preferential

interactions between specific proteins from
co-purification of complexes: succinate dehydrogenase/

cytochrome bc (Yu & Yu, 1980), NADH dehydrogenase/

1

cytochrome bc (Hatefi 31 31., 1962), and cytochrome bf/

1
PS I (Boardman, 1971).

A non-random distribution of enzymes coupled with a
limited path for Q10 implies that the Q pool would not be
expected to behave homogeneously under all conditions since
reduced Q10 produced by one substrate may not mix with that
part of the Q pool associated with the other dehydrogenase
before it is oxidized by the local oxidase (Ragan & Reed,
1986). Indeed, when kinetics of electron transfer through
Q10 is measured in presence of both succinate and NADH, a
simple Q pool behavior is no more observed (Gutman, 1985).
The data is explained by postulating the presence of domains
in which the redox state of the quinones may be different.
Electron transfer in plant mitochondria show strong
deviations from the Q pool behavior as electron flux between

the cyanide sensitive and cyanide insensitive oxidase

pathways can be strongly dependent on the nature of the

81

substrate. In Panicum.Miliaceum bundle sheath mitochondria
oxidize exogenous NADH almost exclusively (95%) via the
cyanide sensitive oxidase, while malate is almost
exclusively (91%) oxidized via the alternative cyanide-
insensitive oxidase (Gardestrom & Edwards, 1983). Similar
trend has been also observed in mitochondria from spinach
(Douce 31 31., 1977),cassava (Huq & Palmer, 1978), and sweet
potato (Cottingham & Moore, 1983). However, in A.macu1atum
mitochondria, exogenous NADH is oxidized primarily by the
alternative oxidase (Cottingham & Moore, 1983., Huq &
Palmer, 1978) and a direct interaction between the external

NADH dehydrogenase and bc complex seems unlikely.

1
Above observations of non-random distribution of
proteins in mitochondria and selective electron transfer
through Q10 can be interpreted to mean that diffusion of Q10
is not random. A dynamic aggregate model and the non-random

distribution of proteins would account for the function and

location of Q10 in the mitochondrial membrane.

82

SUMMARY

It has become increasingly evident that knowing the
correct lateral diffusion rates of ubiquinone is important
for understanding its function as a transporter of reducing
equivalents (electrons and protons) in the mitochondrial
membrane.

The two studies performed so far give very different
values. Gupte 31 31. (1984) have measured a diffusion

9 cmzsecm1

coefficient of 3 x10- for a ubiquinone analog that
has an alkyl side chain corresponding to only two isoprene
units,using the technique of fluorescence redistribution
after photobleaching (FRAP). It is known that short chain
quinones orient themselves in a membrane similarly to a
phospholipid. In fact, the diffusion rates reported for the
ubiquinone analog are about the same as measured for a
phospholipid. However, long chain quinones are located
deeper in the membrane, probably in the mid-plane of the
bilayer, raising the question of whether the quinone analog
is giving a diffusion rate representative of native
ubiquinone.

Fato 31 31. (1986) have measured the diffusion of
native ubiquinone by the technique of collisional quenching
using fluorescently labeled fatty acids. They report a

diffusion coefficient of 1x10-6

cmzsec-1, several orders of
magnitude higher than the values measured by FRAP.

Although it has been proposed that ubiquinone can have such

83

high diffusion rates if located in the mid-plane of the
bilayer, some of the assumptions made in calculating the
diffusion coefficients are questionable, leading to concern
that the high rates may not be valid.

To determine whether the long isoprene chain of native
Q10 has an important influence on the mobility and location
of the molecule, we have synthesized a fluorescent
derivative of the head group moiety of ubiquinone and have
measured its diffusion in giant mitochondria and very large
unilamellar vesicles by the technique of FRAP. The

9

diffusion coefficient of NBDCOQ was 3.1 x10— cmzsec"1 in

8 1 in the vesicles.

mitochondria and 1.1 x10- cmzsec-
Similar diffusion rates were observed for a head group
labeled NBD-phosphatidylethanolamine. Fluorescence emission
studies carried out in organic solvents of different
polarity, and in vesicles and mitochondria indicate that
NBDCOQ is located in a more hydrophobic environment compared
to NBDPE and IANBD. Fluorescence quenching studies carried
out using CuSO4, a water soluble membrane impermeable
quenching agent, also indicate that NBDCOQ is located deeper
in a membrane, compared to head group labeled NBDPE. These
results imply that ubiquinone is oriented differently from a
phospholipid in a membrane, but their diffusion rates are
similar. This diffusion rate appears to be compatible with
a random diffusion (Hackenbrock 31 31., 1986a), or a dynamic

aggregate model of electron transfer (Hochman 31 31., 1985).

Determining whether diffusion of Q10 is actually a rate

84

limiting step in electron transfer will require a more
detailed knowledge of the structural and kinetic properties

of the electron transfer components.

85
REFERENCES

Anderson, J.M. (1981) FEBS Lett. 124, 1-10.
Anderson, J.M. (1982) FEBS Lett. 138, 62-66.

Aranda, F.J., & Gomez-Fernandez, J.C. (1985) Biochim.
Biophys. Acta 820, 19-26. ‘

Aranda, F.J., & Gomez-Fernandez, J.C. (1986) Biochem.
International 12, 137-143.

Aranda, F.J., Villalain, J., & Gomez-Fernandez, J.C. (1986)
Biochim. Biophys. Acta 861, 25-32.

Arvinte, T., Cudd, A., & Hildenbrand, K. (1986) Biochim.
Biophys. Acta 860, 215-228.

Badley, R.A., Martin, W.S. & Schneider, H. (1973)
Biochemistry 12, 268-275.

Beattie, D.S., & Clejan, L. (1986) Biochemistry 25,
1395-1402.

Blackwell, M.F., Gounaris, K., Zara, S.J., & Barber, J.
(1987) Biophys. J. 51, 735-744.

Blatt, E. Chatelier, R.C., & Sawyer, W.H. (1986) Biophys.J.
50, 349-356.

,Boardman, N.K. (1971) Methods Ehzymol. 23,268-276.

Bowman, C.L., & Tedeshi, H. (1983) Biochim. Biophys. Acta
731, 261-266.

Bretscher, M.S. (1983) Proc. Natl. Acad. Sci. U.S.A. 80,
454-458.

Capaldi, R.A. (1982) Biochim. Biophys. Acta 694, 291-306.

Chance, 8. (1965) in Biochemistry of Quinones. (Morton, R.A.
Ed.) pp 459-501, Academic Press, New York.

Chatelier, R.C., & Sawyer, W.H. (1985) Eur. Biophys. J. 11,
179-185.

Chattopadhyay, A., & London, E. (1987) Biochemistry 26,
39-45.

Cohen, M.H., & Turnbull, D. (1959) J. Chem. Phys. 31,
1164-1169.

Conrad, M.J., & Singer, S.J. (1981) Biochemistry 20,

86

808-828.

Cottingham, I.R. 8 Moore, A.L. (1983) Biochim. Biophys. Acta
724, 191-200.

Cox, R.P., 8 Andersson, B. (1981) Biochem. Biophys. Res.
Commun. 103, 1336-1342.

Crane, F.L. (1977) Annu. Rev. Biochem. 46, 439-469.

Crofts, A.R., 8 Wraight, C.A. (1983) Biochim. Biophys. Acta
726, 149-185.

Degli Esposti, M., Bertoli, E., Castelli, G.P., 8 Lenaz, G.
(1981) Arch. Biochem. Biophys. 210, 21-32

De Petris, S. (1978) Nature 272, 66-68.

Douce, R., Moore, A.L., 8 Neuberger, M. (1977) Plant
Physiol. 60, 625-628.

Fahey, P.F., 8 Webb, W.W. (1978) Biochemistry 17, 3046-3053.

Fato, R., Battino,M., Castelli, G.P., 8 Lenaz, G. (1985)
FEBS Lett. 179, 238-242.

Fato, R., Battino,M., Degli Esposti, M., Castelli, G.P.,
8 Lenaz, G. (1986) Biochemistry 25, 3378-3390.

Ferguson-Miller, S., Hochman, J., 8 Schindler, M. (1986)
Biochem. Soc. Trans. 14, 822-824.

Friend, 9.9. (1982) J. Cell. Biol. 93, 243—249.

Gardestrom, P. 8 Edwards, G.E. (1983) Plant Physiol. 71,
24-29.

Goldstein, J.L., Anderson, R.G.W., 8 Brown, M.S. (1979)
Nature 279, 679-685.

Green, D.E. (1962) Comp. Biochem. Physiol. 4, 81-122.
Gupte, S.S., Wu, E.S., Hoechli, L., Hoechli, M., Jacobson,
K., Sowers, A.E., 8 Hackenbrock, C.R. (1984) Proc. Natl.
Acad. Sci. U.S.A. 81, 2606-2610.

Gutman, M. (1980) Biochim. Biophys. Acta 594, 53-84.

Gutman, M. (1985) in Coenzyme Q, (Lenaz, G. Ed) pp 215-234,
Wiley, Chichester, England.

Gwak, S.H., Yu, L., 8 Yu, C.A. (1986) Biochemistry 25,
7675-7682.

87

Hackenbrock, C.R., Chazotte, B., 8 Gupte, S.S. (1986a)
J. Bioenerg. Biomembr. 18, 331-368.

Hackenbrock, C.R., Chazotte, B., 8 Gupte, S.S. (1986b) in
Biomedical and Clinical Aspects of Coenzyme 0, Volume 5
(Folkers, K., 8 Yamamura, Y., Eds ) pp 43-55, Elsevier,
Amsterdam.

Hatefi, Y., Haavik, A.G. & Griffiths, D.E. (1962)
1.8101. Chem. 237, 1676-1680.

Heron, C., Ragan, C.I., 8 Trumpower, B.L. (1978) Biochem. J.
174, 791-800.

Hochman, J., Schindler, M., Lee, J.G., 8 Ferguson-Miller, S.
(1982) Proc. Natl. Acad. Sci. U.S.A 79, 6866-6870.

Hochman, J., Schindler, M., Lee, J.G., 8 Ferguson-Miller, S.
(1983) in Biochemistry of Metabolic Processes (Lennon, D.,
Stratman, F., 8 Zahlten, R., Eds) pp 441-450, Elsevier
Biomedical Press, New York.

Hochman, J., Ferguson-Miller, S., 8 Schindler, M. (1985)
Biochemistry 24, 2509-2516.

Hofle, G., Steglich, W., 8 Vorbruggen, H. (1978) Angew.
Chem. Int. Ed. Engl. 17, 569-583.

Huq, S. 8 Palmer, J.M. (1978) in Plant Physiol. 71, 24-29.

Jain, M. K. (1983) in Membrane Fluidity in Biology,
Volume 1 (Aloia, R. C. Ed. ) pp 1- 37.

Katayama, K., Takada, M., Yuzuriha, T., Abe, K., 8 Ikenoya,
S. (1980) Biochem. Biophys. Res. Commun. 95, 971-977.

Katsikas, H., 8 Quinn, P.J. (1981) FEBS Lett. 132, 19-22.

King, T.E., 8 Suzuki, H. (1984) in Biomedical and Clinical
Aspects of Coenzyme Q, Volume 4 (Folkers, K. 8 Yamamura, Y.,
Eds ) pp 43-55, Elsevier, Amsterdam.

Kingsley, P. B., 8 Feigenson. G. (1981) Biochim. Biophys.
Acta 635, 602-618.

Klausner, R. D. Kleinfeld, A. M., Hoover, R. L., 8
Karnovsky, M. J. (1980) J.Biol. Chem. 255, 1286-1295.

Koppel, D.E. (1979) Biophys. J. 28, 281-292.

Koppel, D.E., Sheetz, M.F., 8 Schindler, M. (1980)
Biophys. J. 30, 187-192.

Krebs, J.J.R., Hauser, H. 8 Carafoli, E. (1979)

J. Biol. Chem. 254, 5308-5316.

Kroger, A., 8 Klingenberg, M. (1973a) Eur. J. Biochem. 34,
358-368.

Kroger, A., 8 Klingenberg, M. (1973b) Eur. J. Biochem. 39,
313-323.

Lakowicz, J.R., 8 Hogen, D. (1980)_Chem. Phys. Lipids 26,
1-40.

Lehrer, 8.8. (1971) Biochemistry 10, 3254-3263.

Lenaz, G., 8 Degli Esposti, M. (1985) in Coenzyme Q,
(Lenaz, G. Ed.) pp 83-105, Wiley, Chichester, England.

Lenaz, G., 8 Fato, R. (1986) J. Bioenerg. Biomembr. 18,
369-401.

MacDonald, R.C. (1987) Biophys. J. 51, Abstr. M-319.

Maloff, B.L., Scordiles, S.P., Reynolds, 0., 8 Tedeschi, H.
(1978) J. Cell. Biol. 78, 199-213.

Maltese, W.A., 8 Aprille, J.R. (1985) J. Biol. Chem. 260,
11524-11529.

Michaelis, L., 8 Moore, M.J. (1985) Biochim. Biophys. Acta
821. 121-129.

Millner, P.A., 8 Barber,J. (1984) FEBS Lett. 169, 1-6.
Mitchell, P. (1976) J. Theor. Biol. 62, 327-367.

Mitchell,P., Moyle, J. (1985) in Coenzyme Q,
(Lenaz, G. Ed.) pp 145-163.

Mueller, P., Chien, T.F., 8 Rudy, B. (1983) Biophys. J. 44,
375-381.

Norling, B., Glazek, E., Nelson, B.D., 8 Ernster, L. (1974)
Eur. J. Biochem. 47, 475-482.

Ondarroa, M., 8 Quinn, P.J. (1986) Eur. J. Biochem. 155,
353-361.

Quinn, P.J., 8 Esfahani, M.A. (1980) Biochem. J. 185,
715-722.

Ragan, C.I., 8 Heron, C. (1978) Biochem. J. 174, 783-790.

Ragan, C.I., 8 Cottingham, I.R. (1985) Biochim. Biophys.
Acta 811, 13-31.

89
Ragan, C.I., 8 Reed, J.S. (1986) J. Bioenerg. Biomembr. 18,
403-418.
Rich, P.R. (1982) Biochem. Soc. Trans. 10, 482-484.
Rich, P.R. (1984) Biochim. Biophys. Acta 768, 53-79.
Schindler, H. (1980) FEBS Lett. 122, 77-79.

Schneider, H., Lemasters, J.J., 8 Hackenbrock, C.R. (1982)
J. Biol. Chem. 257, 10789-10793.

Silverstein, T. P., 8 Rottenberg, H. (1987) Biophys. J. 51,
Abstr. W-192.

Sjostrand, F.S., (1983) in Membrane Fluidity in Biology,
Volume 1 (Aloia, R.C. Ed. ) pp 83-142, Academic Press, New
York.

Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.T.,
Gartner, F.H., Provenzano, M.D., Fujimato, E.K., Goeke,
N.M., Olson, B.J., 8 Klenk, D.C. (1985) Anal. Biochem. 150,
76-85.

Srere, P.A. (1982) Trends Biochem. Sci. 82, 375-378.
Staehelin, L.A. (1976) J. Cell. Biol. 71, 136-158.

Stern, 0., 8 Volmer, M. (1919) Phys. 2. 20, 183-189.

Stidham, M.A., McIntosh, T.J., 8 Siedow, J.N. (1984)
Biochim. Biophys. Acta 767, 423-431.

Struck, D.K., 8 Pagano, R.E. (1980) J. Biol. Chem. 255,
5404-5410.

Suzuki, H., 8 King, T.E. (1983) Fed. Proc., 42, 2060.

Teissie, J., Tocanne, J.F., 8 Baudras, A. (1978) Eur.
J.Biochem. 83, 77-85.

Thulborn, K.R., 8 Sawyer, W.H. (1978) Biochim. Biophys. Acta
511, 125—140.

.Trumpower, B.L. (1978) Biochem. Biophys. Res. Commun. 83,
528-535.

Trumpower, B.L. (1981) J. Bioenerg. Biomembr. 13, 1-24.

Ulrich, E. L., Girvin, M.E., Cramer, W.A., 8. Markley, J.L.
(1985) Biochemistry 24, 2501-2508.

Unden, 6., 8 Kroger, A. (1981) Eur. J. Biochem. 120,
577-584.

90

Vaz, W.L.C., 8 Hallmann, D. (1983) FEBS Lett. 152, 287-290.

Vaz, W.L.C., Goodsaid-Zalduondo,F., 8 Jacobson, K. (1984)
FEBS Lett. 174, 199-207.

Venetie, R.V., 8 Verkleij, A.J. (1982) Biochim. Biophys.
Acta 692, 397-405.

Werner, S. 8 Neupert, W. (1972) Eur. J. Biochem. 25,
379-396.

Williams, R. J. P. (1983) Trends Biochem. Sci. 8, 48.

Yu, C.A., 8 Yu, L. (1980) Biochim. Biophys. Acta 591,
409-420.

Yu, L., & Yu, C.A. (19828) J. Biol. Chem. 257, 2016-2021.
Yu, L., & Yu, C.A. (1982b) J. Biol. Chem. 287, 10215-10221.

Yu, L., Yang, F.D., 8 Yu, C.A. (1985) J. Biol. Chem. 260,
963-973.

Zhu, Q.S., Berden, J.A., De vries, S., 8 Slater, E.C. (1982)
Biochim. Biophys. Acta 680, 69-79.

91
APPENDIX

PUBLICATIONS:

Rajarathnam, K., Hochman, J., Schindler, M.,
8 Ferguson-Miller, S. (1987) Mobility and Location of NBD-
Ubiquinone (NBDCOQ) in mitochondrial membranes and
phospholipid vesicles. (manuscript in preparation, to be
submitted to Biochemistry).

Ferguson-Miller, S., Rajarathnam, K., Hochman, J.,
8 Schindler, M. (1988) Is electron transfer mediated by
random diffusion alone? (manuscript in preparation, to be
published as the proceedings of the ‘International
Conference on Integration of Mitochondrial Function' Plenum,
New York.

ABSTRACTS:

Rajarathnam, K., Hochman, J., Hupp, T., Schindler, M.,
8 Ferguson-Miller, S. (1987) Lateral Diffusion of Ubiquinone
(CoQ) in Mitochondrial Membranes and Phospholipid Vesicles,
Measured by Fluorescence Redistribution After Photobleaching
(FRAP). Fed. Proc. 13, 1930.

Gregory, L., Rajarathnam, K., 8 Ferguson-Miller, S.
(1987) Complete Removal of Subunit III from Cytochrome
Oxidase with Retention of High Electron Transfer and Proton
Translocating Activities. Fed. Proc. 33, 1931.

Rajarathnam, K., Hochman, J., HuPp, T., Schindler, M.,
8 Ferguson-Miller, S. (1987) Location and Diffusion of
Fluorescently Labeled Ubiquinone (CoQ O) in Mitochondrial
Membranes 8 Phospholipid Vesicles Meaéured by Photobleaching
and Fluorescence Quenching Techniques. Presented at the
‘International Conference on Integration of Mitochondrial
Function' Chapel Hill, North Carolina.