Scheffler Immo E. Mitochondria

Подождите немного. Документ загружается.

be further discussed. The F - ATPases typically synthesize ATP driven by a

proton motive force, while the V - type ATPases hydrolyze ATP to generate an

electrochemical gradient across a membrane. The latter two are operating on

similar principles. The discussion will focus on the mitochondrial complex V,

an F - ATPase. In the above model system, ATP hydrolysis by the a/ β subunits

caused the γ subunit to rotate, and one can therefore imagine this rotation to

be the reverse of the rotation caused by a proton ﬂ ux driving ATP synthesis.

Following the original publication describing rotary motion, several more

elegant demonstrations of rotary motion have appeared (288 – 293) . Some of

these model systems have included the c subunits of the rotor, demonstrating

their rotation when attached to the γ subunit with the help of ε . These studies

are beautiful illustrations of the power of nanotechnology applied to the study

of single molecules, but they have not only provided a picture, but have yielded

quantitative data fully reconciling mechanistic and kinetic studies on ATP

synthesis by this enzyme (283, 290, 291) . The reaction is completely reversible.

Hydrolysis of one ATP causes a 120 ° rotation of γ , and the rotation of γ in the

opposite direction can be driven by proton ﬂ ux coupled to ATP synthesis from

ADP and P

. The 120 ° rotation can be decomposed into two sub - steps of 80 °

and 40 ° , where the 80 ° rotation is brought about by the binding of ATP to a

single empty site, and the 40 ° rotation is associated with the release of ADP

from the site. Under physiological conditions the bacterial enzyme is estimated

to rotate at ∼ 100 Hz, and averaged over time all three catalytic sites are occu-

pied, two by ADP and one by ATP. All three sites are involved in driving the

rotation of γ . Pure β subunits do not catalyze hydrolysis at a signiﬁ cant rate,

and even the α

complex shows suboptimal activity and altered afﬁ nities and

kinetics. In the complete structure, “ multisite catalysis ” is now accepted as the

dominant mode of catalysis in vivo , based on the estimated intracellular con-

centrations of substrates and the measured values for the K

of the enzyme.

Figure 5.29 Demonstration of the rotation of the shaft ( γ subunit) relative to the F1

complex. (Adapted from reference 287 .)

Actin filament

Streptavidin

His-tag

α β γ complex

rotation

ATP SYNTHASE (F

-ATPase) 251

252 MITOCHONDRIAL ELECTRON TRANSFER

Multisite catalysis is characterized by an impressive ( ∼ 5 orders of magnitude)

increase in the rate of hydrolysis attributed to the highly cooperative interac-

tion between the subunits and their catalytic sites.

There appears to be a consensus that the binding of ADP and P

at one of

the three active sites of the F

- ATPase provides the energy for combining the

two substrates to make ATP. In other words, ATP formation is spontaneous

after binding and does not require the input of additional energy. The reaction

is a simple reversal of hydrolysis and has no additional covalent intermediates.

A nearby glutamate participates in acid catalysis. The electrostatic repulsion

between the phosphate groups is overcome by interactions with amino acid

side chains in the active site, and the perfect alignment presumably makes the

formation of a covalent bond and the elimination of water entropically favor-

able. The strong binding energy, however, now makes the release of the ATP

unfavorable and slow. It is the release of ATP which is believed to be the step

driven by the proton translocation and rotation. The release must be triggered

by a change in afﬁ nity for ATP, which can be brought about by a change in

the conformation of the protein subunits. The three orientations of a given β

subunit relative to γ can be shown to correspond to the progression from ATP -

occupied ⇒ ADP - occupied ⇒ empty, the reaction sequence expected for the

hydrolysis reaction.

An authoritative and exhaustive review in 1993 by Boyer (278) has provided

a summary of the biochemical observations related to the ATP synthase mech-

anism and provided critical comments on each. An update by the same author

incorporating information from the crystal structure has appeared subsequently

(294) , and highlights in the development of this penetrating and detailed analy-

sis can be found in the acceptance speeches and publications of the 1997 Nobel

Prize winners in Chemistry. Very signiﬁ cant contributions and reﬁ nements have

also been contributed by Senior and colleagues (see reference 282 for an expert

review). They are testimony to the ingenuity and insights of a large number of

investigators keeping pace with or even staying ahead of the structural studies,

until this amazing molecular engine and its inner workings are now revealed

in exquisite detail. Once more it should be emphasized that biochemical studies

were constantly reviewed and tested for consistency with the structural studies.

The clariﬁ cation of the subunit stoichiometry was followed by cross - linking

studies under various conditions, demonstrating conformational changes asso-

ciated with the presence of, for example, Mg

- ADP in the catalytic site, con-

trasted to the presence of the nonhydrolyzable ATP analog AMP - PNP. When

the β and γ subunits were cross - linked, enzyme activity was severely inhibited,

a result now fully explainable in terms of the rotary motion associated with

one cycle of activity. Finally, although most of the biophysical studies were

performed with bacterial enzymes, it is a consensus that the other F - and V - type

ATPases operate on the same principles, with perhaps notable differences in

kinetics and mechanisms of regulation.

The noncatalytic nucleotide binding sites have a structural role and are

likely to play a part in regulating or modulating the stability and/or the

structure of this enzyme (266, 278) . Nucleotide exchange at the noncatalytic

sites is slow. One is reminded of the role of one nonhydrolyzable GTP in the

αβ tubulin dimer. Nevertheless, while a physiologically relevant regulatory

role has not yet been demonstrated, abolition of the noncatalytic nucleotide

binding sites by site - directed mutagenesis has shown that they are not dispens-

able. Such a modiﬁ ed enzyme ( α

γ complex) exhibits an initial ATPase

activity, but this activity decays rapidly to zero. This observation has been

interpreted to indicate an entrapment of MgADP in a catalytic site (see refer-

ence 295 for additional references to investigations of the speciﬁ c role of the

noncatalytic sites).

A schematic representation of the three - site model is shown in Figure 5.30

for the synthesis of ATP (Figure 5.30A ) and for the hydrolysis of ATP (Figure

5.30B ). (see 281, 282 , for details).

5.5.4 The F

Subcomplex and Proton Flow

Attention now must focus on the stator (b

OSCP) in the periphery and on the

subcomplex (ab

) for which complete high - resolution crystal structures

are not yet available. The F

subcomplex provides a path (channel) for the

protons. Proton ﬂ ow through the F

channel must be converted into a torque

on the rotor ( γε c

) in the center. Thus, two questions can be formulated: (1)

What is the path of the protons through the F

subcomplex? (2) How does the

translocation of protons from one side of the membrane to the other produce

a torque? To avoid confusion, the following discussion will refer to the essen-

tial subunits of the bacterial complex, and the reader is referred to Table 5.2

for the corresponding subunits in the mitochondrial complex in various

organisms.

While δ and ε subunits were dispensable in the demonstration of F

- ATPase

(α

γ ) as a rotary motor, the ε subunit is likely to be necessary for the coupling

between the F

and F

subcomplexes. It can be cross - linked via its C - terminal

domain to the α

barrel and via its N - terminal domain to the γ subunit.

Sequence conservation in bacterial and chloroplast ε subunits has focused

attention on an N - terminal domain, and studies employing site - directed muta-

genesis have shown that mutations (e.g., ε H38C) affect coupling between

proton ﬂ ux and ATPase (266) . Cryo - electron microscopy has shown a major

shift of the ε subunit relative to the α and β subunits when viewed down the

threefold symmetry axis upon binding of different nucleotides (266) .

To function as a highly speciﬁ c inhibitor of the ATP synthase in mitochon-

dria, oligomycin (Figure 5.31 ) requires the presence of the OSCP subunit

(oligomycin - sensitivity conferring protein); it was shown to be equivalent to

the bacterial δ subunit. Since it blocks proton ﬂ ow, it had been associated with

the proton translocating mechanism, but it seems clear now that the inhibition

is indirect. The δ subunit is located “ on top ” of the a

barrel, facilitating the

interaction of the tip of the b

complex with the α

barrel (see Figure 5.26 ).

A role in the assembly of the ATP synthase is discussed brieﬂ y below.

ATP SYNTHASE (F

-ATPase) 253

254 MITOCHONDRIAL ELECTRON TRANSFER

ADP

ATP

ADP

ATP

ADP

ATP

ADP - P

ADP

ATP

ADP

ATP

ADP

hydrolysis

Figure 5.30 (A) Multisite catalysis model for the synthesis of ATP. (B) Multisite

catalysis model for the hydrolysis of ATP. In this model all three catalytic sites are

occupied, except for a very transient period when the product has left and before a

new substrate enters the site. (Adapted from reference 281 .)

ATP

ADP-P

ATP

ADP-P

ATP

ADP - P

ATP

ADP

synthesis

ADP - P

ATP

The bovine F

subcomplex had ﬁ rst been prepared in Racker ’ s laboratory

by a combination of sonication and urea treatment. A more homogeneous

preparation suitable for structural analysis has been devised by the Cambridge

group of J. Walker. The ﬁ nal complex from bovine heart contains nine different

proteins (a – g, F6, A6L) (296) . The stoichiometry for the mammalian mitochon-

drial F

complex had also been established independently (e.g., reference 260 ).

All of the subunits have been expressed in bacteria from cloned cDNAs, and

successful re - constitution experiments making various subcomplexes as well

as the entire F

- ATPase have been reported (296) . Complete information

for the yeast mitochondrial F

- ATPase peptides (238) is based primarily on

genetic analyses of respiration - deﬁ cient yeast mutants, followed by biochemi-

cal experiments. Additional gene products were found to be necessary for the

assembly of the complex, but were not present in the puriﬁ ed complex (238,

297) .

From comparisons with the simpler complex in bacteria, one can conclude

that three essential peptides make up the proton - translocating hydrophobic

core (298) . Subunit a (271 residues in bacteria) has multiple transmembrane

regions, most likely an even number (4 – 8). This places both the N - terminal

and the C - terminal on the same side of the membrane. With the help of fusion

proteins, the ends have been placed on the outside (periplasmic side) of the

bacterial membrane (298) . It is not required for the binding of the F

. The b

subunit may be present as a dimer. In bacteria, it has been deduced from the

amino acid sequence that subunit b has a single transmembrane segment near

the N - terminal, with the remaining domain forming a predominantly α - helical,

highly charged (hydrophilic) secondary structure. Early ideas about its interac-

tions with the γ and δ subunits of F

(298) have been abandoned in favor of a

model placing the two b subunits on the outside of the α / β complex where it

acts as the stator (see Figure 5.26 ). A high - resolution structure of this stator

has been published recently (271) . The a and b subunits can be cross - linked.

They are attached to the outside of the c ring (see below). A high - resolution

Figure 5.31 Structure of oligomycin.

CH2CCH

ATP SYNTHASE (F

-ATPase) 255

256 MITOCHONDRIAL ELECTRON TRANSFER

structure for this interaction is not yet available in spite of the intense interest

focused on this interaction. Protons translocated into the matrix are very likely

taking a path that includes the interface between the a subunit and the c

ring.

The b

dimer is not believed to be involved in forming the proton channel.

Thus, understanding this speciﬁ c protein – protein interaction promises to

shed much light on the mechanism of torque production by proton

translocation.

The c subunit is present in multiple copies (10 – 14, depending on the species)

compared to the a subunit; 10 – 14 c - subunits are associated to form the c ring.

Within its short sequence (79 residues in bacteria) two transmembrane α -

helices are predicted, ﬂ anking a polar loop region. Such a structure has been

veriﬁ ed by NMR studies of the monomer in chloroform:methanol:water (277) .

Using this structure for molecular modeling structures for the c ring were

proposed. A high - resolution structure for the c ring of a related F - type Na

ATPase of Ilyobacter tartaricus was ﬁ nally reported in 2005 (275) (Figure 5.28 ),

while Murata et al. (276) published the corresponding structure for the V - type

- ATPase of Enterococcus hirae . Some signiﬁ cant differences between these

structures and the mitochondrial c

ring can still be expected. First, the number

of monomers per ring may be different; and second, the mammalian ATP

synthase is driven by protons.

Even before detailed crystallographic structures were available, several

revealing structure – function studies were carried out with the bacterial c

subunit modiﬁ ed by site - directed mutagenesis. For example, a Q42E mutant

c - peptide allowed normal assembly of the F

- ATPase, but ATP hydrolysis

was uncoupled from proton translocation (the bacterial enzyme is fully revers-

ible depending on physiological conditions) (298) . The well - known inhibition

of the F

ATPase by DCCD is due to a unique reaction of Asp61 (D61) with

dicyclohexylcarbodiimide (DCCD). Even if only a single c - subunit is deriva-

tized by DCCD, the ATPase activity of the whole complex is abolished. Studies

of the isolated subunit c in a chloroform – methanol – water mixture were

encouraged by experiments suggesting that the isolated protein retains fea-

tures of the native protein, including the reactivity of D61 with DCCD, and

the inhibition of this reaction by the I28T mutation. Experiments of this type

and mutational analysis at this Asp residue have been used to argue that the

carboxyl group of D61 is a participant in proton translocation. At this time it

appears to be the best candidate (277, 298) . Similarly, a mutational analysis of

the bacterial a subunit has identiﬁ ed the R210 residue as an absolutely essen-

tial residue for proton translocation. Thus, the single a subunit and 9 – 12 c

subunits form a path for proton ﬂ ux across the membrane.

Proton ﬂ ow in such a molecular motor must produce a torque. Proton

transfer must be associated with a rotation or displacement of an a - subunit

relative to a c subunit. The model must also include the experimental fact that

several protons and several c subunits are involved in a 120 ° displacement

(more than one proton has to be moved across the membrane for one ATP

to be made). Junge and his colleagues as well as Oster and colleagues have

proposed ingenious solutions in which biochemistry meets engineering (280,

299, 300) ; they await further experimental veriﬁ cation. The reader is referred

to the original literature for details on two major models under consideration

that can be labeled as the “ power stroke model ” and the “ Brownian ratchet

model. ” Evidence from available crystal structures and theoretical consider-

ations favor the ratchet model (299) . In this model the c ring executes stochas-

tic rotational motions. There are two access channels for ions on opposite sides

of the stator (a subunit), but they are not collinear. After an ion has entered

from one side, the c ring has to turn through some angle to allow this ion to

escape on the other side. The unequal concentration and potential of the ions

on either side of the membrane drives a ﬂ ux of ions in one direction, and it

also biases the Brownian rotational ﬂ uctuations of the c ring to produce a net

rotation in one direction. The emerging structural details are clearly of great

importance for deﬁ ning the proton channels, the relevant side chains on the a

and c subunits (already partially determined from mutagenesis experiments),

and the possible conformational changes associated with this mechanism.

Finally, two challenges remain that are closely related in some aspects.

Thermodynamic considerations strongly suggest that on the order of three

protons must be translocated across the membrane for every ATP produced,

or, for every 120 ° turn of the rotor. This had led to predictions at an earlier

stage of exploration that the c ring should have a number of subunits that are

a multiple of three (i.e., 9 or 12). The experimental ﬁ nding of 10 or 14 subunits

in the c

ring raises the issue of a “ symmetry mismatch ” : The symmetry of the

c - ring does not match the symmetry of the α

barrel of the F

subcomplex.

How many c subunits pass the single a subunit of the stator in a 120 ° turn?

The number must be non - integer, and hence the number of protons passing

from one side to the other on average must also be non - integer. There are

considerations that this symmetry mismatch is an essential characteristic of

the torque - producing mechanism, but the problem will keep biophysicists

occupied in the future.

Upon reaching the end of this chapter, one should expect the reader to be

left with a sense of awe, without, however, being pushed into the camp of those

who believe in “ intelligent design. ” Francois Jacob compared evolution to the

work of a “ tinkerer, ” and it is indeed amazing what several billion years of

tinkering can achieve. It is also truly “ awesome ” to think of the billions of cells

in our body, each containing tens of thousands of little motors going continu-

ously at 50 – 100 rps, thus sustaining a turnover of an amount of ATP/ADP per

day that is equivalent to or greater than our body weight.

5.5.5 Assembly of Complex V and Dimerization

The biogenesis of ATP synthase raises several questions that have already

come up in the discussion of the other complexes of the ETC. One has to

consider (a) the expression of multiple genes and (b) the assembly of multiple

subunits making up the subcomplexes of the mature enzyme. In eukaryotes,

ATP SYNTHASE (F

-ATPase) 257

258 MITOCHONDRIAL ELECTRON TRANSFER

mitochondrial genes as well as nuclear genes contribute to the biogenesis; and

furthermore, the nuclear genes are typically dispersed on different chromo-

somes (in contrast to E. coli , where they form an operon). The control and

coordinate expression of such genes is brieﬂ y discussed elsewhere (Chapter

4 ). In yeast the nuclear genes of ATP synthase appear to be constitutively

expressed, regardless of the carbon source, in contrast to the nuclear genes for

the electron transport complexes. At the same time, there appears to be no

speciﬁ c mechanism to coordinate the expression of nuclear genes and mito-

chondrial genes in yeast. A more detailed discussion can be found in the review

by Ackerman and Tzagoloff (238) .

It seems intuitive that the assembly of a complex of 14 or more subunits

would occur in an orderly fashion, but the challenge has been to elucidate the

assembly pathway. Work in S. cerevisiae has been pioneering by taking advan-

tage of the powerful methodologies of molecular genetics. Thus, it is possible

to investigate gene knockouts and determine how far assembly can proceed

in the absence of a particular subunit. As a ﬁ rst approximation, one can state

that the core subunits of the F

subcomplex ( α , β , γ ) and the core subunits of

the F

subcomplex (c, a, b) assemble independently, and at some stage they

are joined and combined with other subunits to form the functional complex.

An F

subcomplex with ATPase activity can be made in yeast mutants lacking

mtDNA ( ρ

mutants). This activity is essential in combination with the adenine

nucleotide transporter to generate a membrane potential for protein import

and for the maintenance of mtDNA in ρ

−

mutants. Alternatively, the assembly

(and hence activity) of ATP synthase has been found to be defective in mutants

even though all known structural genes were normal. Thus, it became apparent

that assembly factors or molecular chaperones are involved (a conclusion

equally applicable to the assembly of the other complexes of the ETC). These

assembly factors are not found in the ﬁ nal complex. Two other properties of

such factors are noteworthy: (1) For example, the factors Atp11p and Atp12p

are speciﬁ cally required only for the assembly of the F

subcomplex, in con-

trast to chaperones such as mtHsp70 and Hsp60; (2) because of their restricted

function, their deletion is not necessarily lethal under conditions where gly-

colysis can satisfy the energy requirements of the cells. The most explicit and

detailed description of the ATP synthase assembly in yeast can be found in a

review by the authors who have also contributed much of the experimental

support for the model (238) .

There are two ATP synthase subunits whose signiﬁ cance has been appreci-

ated only recently. The subunits e and g are required for dimerization of

complex V, detected by blue - native gel electrophoresis. In this context, the most

intriguing observation was made that ATP synthase dimers are playing a essen-

tial role in the establishment of the morphology of mitochondrial cristae (37,

38, 301, 302) . Yeast mutants lacking either e or g are devoid of lamellar cristae,

with the inner membrane folded into onion - like structures or highly abnormal

folds. How ATP synthase dimerization can lead to the “ zippering up ” and align-

ment of cristae membranes remains a challenge for the future.

5.6 CONTROL OF RESPIRATION AND

OXIDATIVE PHOSPHORYLATION

5.6.1 General Considerations

It is quite apparent that the energy demands of a given cell or tissue will vary

over time, and the need for regulatory mechanisms becomes obvious. A dis-

tinction to be made at the beginning of this section is to separate mechanisms

involved in long - term control, in the course of development and tissue differ-

entiation, from short - term control mechanisms operating on a time scale of

seconds to hours. A prominent example is muscle, but other tissues and organs

in mammals and other organisms may also exhibit ﬂ uctuations in respiration

and oxidative phosphorylation in response to regulatory signals from hor-

mones, growth factors, or the nervous system. A subtle but very important

aspect of this problem is the balancing between (a) the capture of the free

energy from combustion in the form of ATP and (b) the release of some of

that energy in the form of heat. This is of course the basic mechanism for

maintaining the body temperature in warm - blooded animals, and speciﬁ c

tissues and mechanisms have evolved for this task.

The long - term control in differentiating tissues, or in the adaptation of a

microorganism such as yeast to changing environmental conditions, is likely

to be based on the control of gene expression and the rate of biogenesis of

mitochondria or the components of the inner membrane. In Chapter 3 a dis-

cussion of the morphology of mitochondria and of the density and shape of

cristae has made reference to the likely explanation that the density of cristae

reﬂ ects the capacity for respiration of the particular cells examined. One could

add to this the control of the copy number of mtDNA per cell.

The present discussion will focus on short - term control. A highly readable

and thoughtful review of this subject with many primary references has been

written by Brown (303) . It concludes with the statement that “ there is

no simple answer to the question ‘ what controls respiration? ’ . ” The question

becomes increasingly more complex as one considers an isolated mitochon-

drion, a uniform cell population in the controlled environment of the culture

medium, or an organ. The answer is likely to vary also depending on the origin

of the mitochondria. Even more variations will have to be considered when

one compares mitochondria from different organisms. One may also have to

make a distinction between (a) studies with isolated mitochondria where

extremes of conditions can be explored and (b) in vivo situations where the

parameters cannot be so well - controlled or are not even precisely known. In

the following discussion, therefore, the emphasis will be on the basic principles

that will have to be taken into consideration when the question of the regula-

tion of mitochondrial activity is raised.

A formal analysis of the problem that has led to important insights and

conclusions is based on the theoretical framework now called Metabolic

Control Analysis (304, 305) . The “ Control of Flux ” was a landmark paper by

CONTROL OF RESPIRATION AND OXIDATIVE PHOSPHORYLATION 259

260 MITOCHONDRIAL ELECTRON TRANSFER

Kacser and Burns ﬁ rst published in 1973 and then released as a recent updated

and annotated reprint (304) . It begins with the emphatic statement that “ ﬂ ux

is a systemic property and questions of its control cannot be answered by

looking at one step in isolation — or even each step in isolation ” , and it ulti-

mately challenges our familiar and simplifying concept of a “ rate - controlling

enzyme ” or of a “ bottleneck in the pathway. ” In its explicit formulation it

constitutes a highly mathematical treatment seeking to express how sensitive

a ﬂ ux in a pathway is to an enzyme ’ s concentration or activity. A ﬂ ux control

coefﬁ cient is introduced as a quantitative measure of the system ’ s sensitivity

to modulations of one of its components (enzyme activity/concentration), and

it is a measure of the importance of this enzyme for control of ﬂ ux through

the pathway, regardless of whether and how this enzyme is controllable.

A coefﬁ cient of 1.0 would indicate that an enzyme is controlling the pathway

and that in practice a 1% increase in enzyme concentration/activity would

lead to a 1% increase in ﬂ ux. Such an enzyme would constitute the typical

“ bottleneck ” in a pathway. Conversely, a coefﬁ cient of 0 would indicate

that the enzyme is in excess and that changing its concentration fractionally

would not lead to a signiﬁ cant change in the ﬂ ux. It appears that the control

coefﬁ cient for many enzymes is small, hence heterozygous individuals are

normal and mutations inactivating one allele are classiﬁ ed as recessive. Post -

transcriptional mechanisms may inﬂ uence such simple considerations based

on gene dose. The formal treatment also introduces elasticity coefﬁ cients for

each enzyme, formerly known as the controllability coefﬁ cients, which roughly

deﬁ ne the sensitivity of an enzyme to all the metabolites and effectors that

interact with it. The reader is referred to the original literature for a formal

discussion of the terminology and mathematical treatment.

While the treatment is highly theoretical and lends itself to explicit computer

simulations, it is also highly suitable for the interpretation of experimental data

from well - known pathways such as glycolysis and oxidative phosphorylation.

Input substrate concentrations can be varied, speciﬁ c inhibitors can be used to

modulate the activity of individual enzymes in the pathway, and more recently

it has become possible to manipulate individual enzyme levels by genetic meth-

odologies (e.g., in yeast). On the other hand, the best computer model is useless

if it is too simple to reﬂ ect the reality inside the cell.

Applications of metabolic control analysis to oxidative phosphorylation

have been numerous over the years, and experimental systems have included

isolated mitochondria, intact cells, perfused tissues, and whole living organisms.

Recent examples by Brand and his colleagues describe (a) the metabolic

control analysis of the effects of cadmium on oxidative phosphorylation in

mitochondria isolated from potato tubers (306) and (b) the control of oxida-

tive phosphorylation in liver mitochondria, hepatocytes, and other tissues

(307 – 312) . Before discussing the formalism used by these groups and the con-

clusions derived from this “ top - down approach ” (i.e., “ starting from the top

and analysing the whole system ” ), a more conventional set of considerations

will be presented.