Scheffler Immo E. Mitochondria

Подождите немного. Документ загружается.

TRANSPORT OF SMALL SOLUTES INTO AND OUT OF MITOCHONDRIA 331

voltage gating parameters. For a detailed discussion the reader is referred to

the work by Colombini and co - workers (51) . How are these properties related

to function? It is suggested that the channel can exist in various functional

states differing in selectivity between anions and cations. In the open state the

selectivity for anions is not very large, while cations are favored by the closed

state. It is argued that pore size is one major determinant with regard to non -

electrolytes (e.g., PEG 3400), but the electrostatic proﬁ le within the channel

can have an inﬂ uence on the passage of charged molecules such as ATP or

NAD

. Thus, in the closed state the channel may prevent ATP/ADP exchange

with the cytosol. The question becomes, What controls the open and closed

states of the VDAC channel? Several suggestions are summarized in the

review by Rostovtseva et al. (51) , among them phosphorylation. Another

group (52) has elevated VDAC to the “ governator of global mitochondrial

function both in health and in disease. ” As has been pointed out, the lack of

a speciﬁ c and potent pharmacological inhibitor of this channel has made physi-

ological studies difﬁ cult.

VDAC has also assumed a prominent place in discussions of early events

in apoptosis, where it has been implicated in the formation of an even larger

pore allowing the escape of cytochrome c from the IMS (51) . Either it could

become part of the pore, or its behavior could cause outer membrane permea-

bilization by an unknown mechanism. Alternatively, VDAC has been hypoth-

esized to stack on top of the adenine nucleotide transporter (ANT) to form a

contact site and a continuous pore through the outer and inner membrane.

Such a structure could be responsible for the permeability transition and

breakdown of ΔΨ at some stage during apoptosis. This topic will be expanded

on in the section on apoptosis.

6.9.3 The ADP/ATP Translocator

The ADP/ATP translocator (adenine nucleotide translocase, ANT) is an inter-

esting integral membrane protein with a rather unique function in mitochon-

dria. Its properties, shared with the uncoupling protein, UCP - 1 and with what

is believed to be a large family of mitochondrial carriers, have already been

emphasized (53) . To repeat: The protein is an integral membrane protein with

a molecular mass of ∼ 32 kDA (297 residues in the bovine protein). There are

three repeats of about 100 amino acids, each of which contains two shorter

motifs representing transmembrane helices. The protein therefore is localized

in the membrane with six transmembrane helices, and both the N terminal and

the C terminal are on the same side, facing the intermembrane space. The

loops on the matrix side contain an exceptionally polar combination of about

40 side chains. There are also signature motifs with short sequences conserved

in all members of the family. These have been exploited very successfully in

blast searches of whole genomes to identify novel family members (54) . Bio-

chemical and biophysical experiments had suggested that a dimeric structure

makes up the functional carrier, with a twofold axis of rotational symmetry

332 METABOLIC PATHWAYS INSIDE MITOCHONDRIA

that has also been proposed to form the channel through which ADP/ATP

translocation occurs. A structure at a resolution of 2.2 Å by X - ray crystallo-

graphy in complex with the inhibitor carboxyatractyloside (see below) was

published in 2003, representing the ﬁ rst high - resolution structure of a member

of this large carrier family (Figure 6.9 ) (55 – 57) . It has since become a model

structure for predicting and understanding other members of this mitochon-

drial carrier family (see page 334) . Six transmembrane α - helices are clearly

seen, but they are severely tilted relative to each other and relative to the

orthogonal axis through the membrane. In the crystal the unit cell contained

only the monomer. Furthermore, the six helices form a barrel - shaped structure

with the postulated nucleotide binding site. These results clearly contradict the

idea of a dimeric structure with 12 helices required for the formation of the

pore through the membrane. It is therefore possible that the proteins are

aggregated as dimers (or even higher - order oligomers) while each monomer

still functions independently (55, 57) . The crystal structure suggests a cavity

on one side lined with side chains that are presumably responsible for the

speciﬁ city of the carrier. It is thought to represent one conformation of the

carrier, as stabilized by the inhibitor.

A distinguishing feature of the translocator/carrier and UCP family is the

interaction with purine nucleotides. In the UCP - 1 the nucleotide is a regula-

tory ligand controlling unidirectional transport of other small ions (H

/OH

–

or Cl

–

), while in the ADP/ATP translocator the purine nucleotide is in fact

the substrate and activator of translocation. UCP - 1 binds the regulatory ligand

only on the outside (cytoplasmic side) of the inner membrane. The translocator

must bind nucleotides on either side, and the net reaction (ADP imported, or

ATP exported) is driven by the membrane potential, ΔΨ , since ATP has one

more negative charge than ADP. The function of this protein is therefore inti-

mately linked to the overall process of oxidative phosphorylation, although its

control coefﬁ cient may be modest (58) . Nevertheless, a more severe dysfunc-

Figure 6.9 X - ray structure of ANT (from reference 57 with permission).

TRANSPORT OF SMALL SOLUTES INTO AND OUT OF MITOCHONDRIA 333

tion may contribute to pathological effects that may be implicated in neuro-

muscular diseases (59) .

Two highly speciﬁ c inhibitors have been employed in the study and in the

puriﬁ cation of the ADP/ATP translocator (Figure 6.10 ). These inhibitors do

not bear much resemblance to each other or to ATP, except for the presence

of three or four negative charges. Binding sites for the inhibitor and ATP are

similar or overlapping, since binding of one precludes the binding of the other.

Carboxyatractyloside is derived from plants, and bongkrekic acid has been

isolated from certain Pseudomonas strains. Their structures are shown in

Figure 6.10 ). They differ in their ability to penetrate the inner membrane.

Figure 6.10 (A,B) Two highly speciﬁ c inhibitors employed in the puriﬁ cation of the

ADP/ATP translocator.

atractyloside R = H

carboxyatractyloside R = COOH

COOH

bongkrekic acid

COOH

CH CH

OCH

CH C

COOH

334 METABOLIC PATHWAYS INSIDE MITOCHONDRIA

Atractyloside is restricted to the outside and hence binds to the cytoplasmic

side of the carrier, whereas the bongkrekic acid can enter the mitochondria,

and it binds only to the matrix side. The use of radiolabeled or ﬂ uorescent

inhibitors can help in estimating the number of carrier sites per mitochondrion

(it is generally a very abundant protein — up to 15% of total protein). With the

atractyloside bound, the carrier can be stabilized during puriﬁ cations and

protected against enzymatic degradation (see reference 53 for review and

earlier references). This form has also been successfully used for the crystalli-

zation and X - ray diffraction experiments (56, 57) (Figure 6.9 ).

Most interesting and revealing, however, has been the observation that no

ternary complexes have been found; that is, the inhibitors not only prevent the

binding of ATP, or ADP, but both inhibitors cannot bind at the same time, even

though they bind to opposite sides of the carrier. The interpretation by Klin-

genberg was that the carrier is alternately open for the substrate on one side

or the other. It is postulated to exist in two conformational states; and during

the transition from one state to the other, the substrate is translocated

across the membrane. In fact, it was further hypothesized that the transition

between the two states (c state and m state) is possible only with the substrate

(either ATP or ADP) present ( “ induced transition ﬁ t theory of carrier trans-

port catalysis ” ). According to this model, the inhibitors can bind to the open

faces of two opposing “ ground states ” of the carrier, stabilizing a conformation

that is incapable of binding the adenine nucleotide. The crystal structure is

consistent with this model, but it would clearly be strengthened considerably,

if a crystal structure of the alternate conformation binding a nucleotide or

bongkrecic acid on the matrix side could be solved.

6.9.4 The Mitochondrial Carrier Protein Family

The previous sections have emphasized a variety of reactions and pathways

localized in the mitochondrial matrix. It is clear that many intermediates have

to transported into mitochondria, and others have to be exported. Because

metabolism is thus compartmentalized and because the inner membrane must

maintain a membrane potential, the transport of metabolites must be tightly

controlled. Mechanisms for transporting various metal ions have already been

described. Most/all metabolites are transported by a large group of mitochon-

drial transport proteins that can be grouped into 37 subfamilies with a total of

75 members in humans (54) . Several prominent family members have already

been introduced and discussed in a different context: The adenine nucleotide

transporter (ANT) and its isoforms are responsible for the import of ADP and

the export of ATP. A phosphate transporter brings phosphate from the cytosol

into the matrix for ATP synthesis. Multiple isoforms of uncoupling proteins

(UCP1, UCP2, . . . ) were identiﬁ ed after the ﬁ rst member of the group was

found to be a key factor in thermogenesis in brown adipose tissue. The role of

the others appears to be more subtle (see page 266 or see section 5.6.2 ). It is

now clear that most of these transporters have a structure very similar to that

TRANSPORT OF SMALL SOLUTES INTO AND OUT OF MITOCHONDRIA 335

of the ANT: a molecular mass of 30 kDa, three degenerate repeats of ∼ 100

amino acids, six transmembrane helices, N and C terminals in the intermem-

brane space, two short loops in the IMS, and three larger loops in the matrix

(Figure 6.11 ). This topology, originally predicted from the sequence and hydrop-

athy proﬁ les, has been conﬁ rmed by a high - resolution structure (2.2 Å ) deter-

mined from X - ray diffraction (see page 332; also see references 55 and 56 ) A

blast of the human genome with 23 residue long sequences from 12 different

regions of 20 known human mitochondrial transporters has predicted 75 such

transporters (54) . Some of these are signiﬁ cantly larger due to N - terminal

extensions forming a calcium binding site. Others are much smaller, with partial

sequences and truncations at either the N terminal or the C terminal. Their

functions, if any, remain to be established, although at least one has been found

to be expressed in mitochondria. Highly conserved sequence and structure

elements are constrained by the need to form homodimers for their activity.

Only 21 of the human mitochondrial transport proteins have been identiﬁ ed

with regard to their speciﬁ city and function. There is likely to be a certain

degree of redundancy, leading to an estimate of a possible total of 39 functions

(60) . Known and expected functions include the import/export of amino acids

including citrulline and ornithine for the urea cycle, keto acids (pyruvate),

di - and tricarboxylic acids (malate, citrate, α - ketoglutarate), carnitine and acyl

carnitine (fatty acid oxidation), nucleotides for DNA and RNA metabolism,

coenzyme A, NAD

, FAD, S - adenosyl methionine, and more (see Table 6.1 ).

The speciﬁ city of the transporter for a metabolite is not predictable from the

amino acid sequence, and it has to be established from genetic experiments

(gene knockouts) or biochemical experiments. The latter typically involve

overexpression of a carrier in bacteria or yeast, followed by puriﬁ cation and

reconstitution in liposomes.

6.9.5 Cation Transport

The most signiﬁ cant cations other than protons are Na

, K

, Ca

, Mg

, and

the trace metals Fe

, Cu

, Zn

, and others. Because of its special importance

Figure 6.11 Structure of a phosphate transporter. (Adapted from reference 54 .)

I M

N-ter

C-ter

M1 M2 M3

Matrix

IMS

336 METABOLIC PATHWAYS INSIDE MITOCHONDRIA

in speciﬁ c regulatory mechanisms, the uptake and release of Ca

will be

considered under its own heading below. One primary mechanism of cation

uptake/exchange involves antiporters, and several such proteins with distinct

speciﬁ cities have been described (61, 62) . Na

ions can traverse the inner

membrane via an apparently unregulated Na

antiport, or a slightly more

complex Na

/Ca

antiport (see below). The Na

ion concentration is generally

low in the cytosol, and it is also kept low in the mitochondrial matrix. The

presence of the Na

antiport keeps the electrochemical Na

gradient equal

to the H

gradient. K

ions have a major role in maintaining mitochondrial

volume, and their uptake is achieved by three mechanisms: an electroneutral

antiporter that may be regulated by free Mg

in the matrix, a K

channel activated by ATP (63) , and a leakage driven by the high electrochemi-

cal gradient across the inner membrane (64) . The puriﬁ ed mitK

ATP

channel has

been studied in synthetic lipid bilayer to prove that regulatory nucleotides

(and also acyl - CoA esters) bind to the cytosolic side of the membrane (65, 66) .

Thus, a balance must be maintained between (a) an outward ﬂ ux via the anti-

porter driven by the proton gradient and (b) an inward ﬂ ux through the leak

and channel driven by the electrochemical gradient. The precise physiological

signiﬁ cance of the ATP - sensitive channel remains to be fully explained, but

it has been suggested that it serves in regulated energy dissipation at times

of low ATP demand by the cell, when the ﬂ ow and metabolism of reducing

equivalents between matrix and cytosol has to be maintained. Mitochondrial

TABLE 6.1 Carrier Proteins in the Inner Mitochondrial Membrane

Substrate 1 Substrate 2

Antiporters

Malate Phosphate

ATP ADP

ATP dNTP

S - Adenosyl - methionine S - Adenosyl - homocysteine

Carnitine Acyl - carnitine

Ornithine Citrulline

Glutamate Aspartate

Malate

α - Ketoglutarate

Malate Citrate

Uniporters (or antiporters with one proton)

Pyruvate

Glutamate

Phosphate

Glutamine

NAD+

FAD

Coenzyme A

TRANSPORT OF SMALL SOLUTES INTO AND OUT OF MITOCHONDRIA 337

volume has been shown to inﬂ uence the rate of electron transport and respira-

tion (61, 67) . The mitK

ATP

channel has received particular attention in relation

to cardiac function (64, 68, 69) . Based on co - puriﬁ cation and re - constitution

experiments Garlid ’ s group has proposed a possible signaling mechanism from

the observed association of the mitK

ATP

channel with the protein kinase C ε .

Various activators of this kinase could also activate K

ﬂ ux in liposomes with

these two components (70) .

Uniports for Na

and K

have also been identiﬁ ed, but their characteristics

and functional importance are still under investigation. Similarly, the passage

of Mg

ions through the inner mitochondrial membrane is still poorly under-

stood (71) . It appears that mitochondria take up Mg

through membrane

leakage channels driven by the membrane potential, rather than by a speciﬁ c

transporter (72) . Interesting speculations have centered on the possibility of

hormonal control of [Mg

] in both the cytosol and matrix.

6.9.5.1 Transport of Calcium and Its Physiological Role Intramito-

chondrial free calcium [Ca

]

, and the mechanisms associated with its control

have been examined extensively, since its recognition as a potential mediator

of a variety of metabolic activities. Additional importance is derived from the

recognition that certain pathological conditions may be associated with abnor-

mal Ca

levels in mitochondria (49, 73) . The nature of the transporters, the

bioenergetics of transport, and the regulation of Ca

transport all have been

investigated since the discovery of mitochondrial Ca transport in the early

1960s (74) . The existing membrane potential favors inﬂ ux, and efﬂ ux is ener-

getically uphill. An expert discussion of the energetics of Ca

transport is pre-

sented in a review by Gunter and Gunter (48) . From typical values for ΔΨ and

internal and external [Ca

], it is estimated that ∼ 29 kJ/mole Ca

is required

to get Ca

out of mitochondria. From another point of view, this considerable

free energy difference explains why mitochondria have been thought of as

intracellular “ sinks ” or reservoirs for Ca

, leading to the challenge to under-

stand how they contribute to the control to the overall cytosolic level of this

ion. This idea has been challenged in the past decade from the discovery that

intracellular Ca

is largely found in the ER, from where it is mobilized by

inositol triphosphate (IP

). Mitochondria became the target for Ca

signaling

rather than the origin of regulatory Ca

(75) .

A major player in Ca

transport in vertebrate mitochondria is a uniporter

that can also transport Mn

, Ba

, Fe

, Pb

in decreasing order of selectivity,

although, curiously, Mg

uptake may not involve this route. The conclusion

that it is a uniporter rests on a variety of experiments summarized in reference

48 . The membrane potential (negative inside) is essential, favoring the net

charge transfer of 2; that is, Ca

diffuses down the Ca

electrochemical gra-

dient, thereby diminishing the membrane potential. Detailed kinetic studies

have been interpreted to show the existence of two sites for Ca

on the uni-

porter: one for activation and the second for the actual transport. In other

words, Ca

itself modulates the activity of this uniporter and can inhibit its

338 METABOLIC PATHWAYS INSIDE MITOCHONDRIA

own uptake; experiments show that it is a calmodulin - mediated process (see

reference 75 for a further discussion and references). Another interesting

observation is that this uniporter does not appear to allow Ca

ions to pass

down a chemical gradient alone; that is, a membrane potential is required to

induce a conformation necessary for transport.

Two distinct Ca

efﬂ ux mechanisms in vertebrate mitochondria are respon-

sible for protection against the buildup of toxic levels of Ca

in the matrix.

It was a challenge to prove that the Na

- independent mechanism was not

an artifact or simply a reversed uniport (reviewed by in reference 48 ). It now

appears established that Ca

efﬂ ux is accompanied by proton inﬂ ux, and a

ratio 1Ca

/2H

appears to be favored. However, bioenergetic considerations

conclude that it cannot be a simple passive exchanger, and some additional

input of energy from the electron transport chain must be postulated. The

energy coupling mechanism is still obscure. An Na

- dependent mechanism can

be distinguished on the basis of certain ion selectivities (Sr

but not Mn

stimulation by Na

, spermidine, and spermine, and inhibition by a variety of

inhibitors. Current thinking favors an electrogenic Ca

/3 Na

exchanger with

the evidence pointing to a ΔΨ - dependent electrophoretic antiport (76) .

Our understanding of the physiological role of mitochondrial Ca

transport

is still expanding (49, 75, 77) . As stated above, mitochondria can constitute a

buffering compartment with a role in modulating cytosolic calcium, [Ca

]

cytosol

A relevant question was whether transport in and out of mitochondria was

sufﬁ ciently rapid to keep up with rapid, pulsed changes in cytosolic [Ca

found, for example, in beating cardiomyocytes or in nerve cells engaged in

active signaling (exocytosis triggered by elevated [Ca

]

cytosol

. A second major

insight was that two prominent mitochondrial dehydrogenases (pyruvate

dehydrogenase and α - ketoglutarate dehydrogenase) are activated by [Ca

] m .

It should be noted that the critical concentrations are those of the free Ca

Estimates are that only one part of 10

of total mitochondrial Ca

is free, and

this fact must also be considered in the context of the buffering capacity of

mitochondria. The stimulation of the dehydrogenases increases the NADH/

NAD

ratio and hence oxidative phosphorylation. One can consider a plausi-

ble scenario in which increased [Ca

]cytosol in stimulated skeletal or cardiac

muscle is rapidly coupled to increased [Ca

]

and hence increased rates of

oxidative phosporylation. Such a mechanism could serve to signal increased

energy demand in the cytosol to the OXPHOS machinery.

The use of ﬂ uorescent Ca

indicators (e.g., fura - 2) developed by Tsien and

colleagues (77a) has revolutionized the measurement of [Ca

]

cytosol

, and has

even been applied to isolated mitochondria. Subsequently, recombinant aequo-

rin has been targeted to mitochondria to serve as a sensitive indicator of

[Ca

]

(78) . The combination of fura - 2 and mitochondrial aequorin in single

cells permitted simultaneous measurements of [Ca

]

cytosol

and [Ca

]

, with

intriguing results (reviewed in reference 79 and 80 ). Raising [Ca

]

cytosol

inﬂ ux from extracellular Ca

affected [Ca

]

less and at a slower rate than

an increase in [Ca

]

cytosol

by a release (IP

- induced) from intracellular stores

TRANSPORT OF SMALL SOLUTES INTO AND OUT OF MITOCHONDRIA 339

(ER). These observations can now be explained by observations made in the

light/ﬂ uorescent microscope showing that the ER and mitochondria can form

intracellular “ synaptic ” regions. At these junctions the IP

- stimulated Ca

release channel in the ER is physically close to the mitochondrial Ca

uni-

porter, and the ion can be taken up directly into the mitochondria without

signiﬁ cant spreading into the cytosol (81, 82) . An interesting paper by Rapizzi

et al. (83) demonstrates that overexpression of VDAC can enhance the trans-

fer of Ca

from the ER to mitochondria. As a consequence, cytoplasmic Ca

microdomains are generated within the cellular architecture that ﬂ uctuate

both in space and in time. Subcellular Ca

concentrations are dynamically

heterogeneous from one location to another. They can thus regulate a diverse

group of cellular events stimulated by various extracellular factors (77) .

6.9.6 The Mitochondrial Permeability Transition

The mitochondrial permeability transition (MPT) was originally deﬁ ned as

sudden increase in the inner membrane permeability to solutes of molecular

mass less than ∼ 1500 Da. It is now believed to be due to the opening of a mega

channel, which is referred to in some publications as the mitochondrial perme-

ability transition pore (MTP). It has received particular attention in recent

years because pore opening is sensitive to cyclosporin A (see below) and

because it has been implicated in the pathway of apoptosis (Chapter 8 ). The

transition is detectable and measured with isolated mitochondria in protein -

free buffers by an uptake of radioactive molecules such as sucrose, by a change

in optical density due to swelling, and even by patch - clamp techniques (84,

85) . The original observations were made with animal mitochondria (beef

heart, rat liver), but recent publications suggest that yeast mitochondria are

also capable of forming a transition pore. However, the regulation of this pore

in yeast may be different (86) .

Several broad questions can be raised, and answers to these questions are

beginning to accumulate. The discussion will focus on the following issues: (1)

What components of the inner membrane constitue the pore? (2) What mech-

anisms regulate the opening and closing of the pore? (3) What is the physio-

logical signiﬁ cance of the operation of this megachannel? To some extent the

answers are interrelated.

The exact composition of the pore is still not completely clear, except for

a general agreement that the adenine nucleotide translocase (ANT) is a par-

ticipant in channel formation, perhaps by forming complexes with outer mem-

brane proteins to create contact sites and hence to stabilize and/or enlarge the

channel. The peripheral benzodiazepine receptor and VDAC in the outer

membrane have been implicated, and there is speculation about the inclusion

of Bcl2, cardiolipin synthase, 3 - b - hydroxysteroid isomerase, and phospholipid

hydroperoxide glutathione peroxidase (87) . On the cytosolic side, hexokinase

may also be associated, and on the matrix side the participation of cyclophilin

D is suggested by the sensitivity to cyclosporin.

340 METABOLIC PATHWAYS INSIDE MITOCHONDRIA

The channel can be opened when Ca

is accumulated in the matrix by

exposure of mitochondria to high concentrations of Ca

(generally above

physiological levels), and the sensitivity to Ca

is heightened under various

conditions including oxidative stress, adenine nucleotide depletion, elevated

phosphate concetrations, and low membrane potenial, ΔΨ ? The role of

nucleotides points to an involvement of ANT, but a more direct indication is

the effect of inhibitors of ANT on the MPT. Inhibitors that stabilize the “ c ”

conformation of ANT (e.g., atractyloside) favor the MPT, while inhibitors

stabilizing the “ m ” conformation (e.g., bongkrekic acid) also inhibit pore

opening. A high ΔΨ or a low pH prevents the transition (88) . Of special inter-

est was the observation that submicromolar concentrations of cyclosporin A

are very effective in keeping the pore from opening. The target for cyclosporin

is the matrix peptidyl - proline cis – trans isomerase, a member of the cyclophilin

family. One interpretation was that the (soluble) matrix isomerase may inter-

act with an integral membrane protein, and when triggered by Ca

, the pore

opening is induced. Cyclosporin A would interfere with this required interac-

tion. A further speculation was that the relevant IMP was the ANT (89) . The

idea was supported by the observed attachment of cyclophilin to the inner

membrane under certain conditions (oxidative stress, thiol reagents, increased

matrix volume), which favor or enhance pore opening (90, 91) . Even though

the MPT makes an appearance in a large fraction of publications related

to apoptosis, necrosis, oxidative stress, and other pathologies, there is still no

clear distinction between the “ MPT ” and the “ MTP. ” The ﬁ rst describes a

phenomenon, while the second implies the existence of a structure that has

never been isolated and whose composition is hypothesized on the basis of

indirect studies with various inhibitors. The suggestion has even been made

that the “ pore ” is constituted of unfolded integral membrane proteins damaged

by ROS (e.g., oxidation of sulfhydryls, etc.) (92) .

The combined observations have served to support a variety of potential

functions for the MTP, which may not be exclusive (87) : (1) a voltage sensor,

(2) a matrix pH sensor, (3) a divalent cation sensor (role in calcium homeo-

stasis (e.g., 93 )), (4) a sensor of adenine nucleotide concentrations, (5) a sensor

of the redox state of the pyridine nucleotide pool, and (6) a thiol sensor (redox

status of glutathione). The context in which the MPT is most commonly

encountered is the triggering and progression of apoptosis, a subject that will

be expanded in a following chapter.

It is clear that a prolonged opening of a number of pores, or the simultane-

ous opening of many pores, would deenergize and inactivate mitochondria

rapidly, and it may be that such an event is associated with the irreversible

breakdown of ΔΨ observed in the course of apoptosis (see Chapter 8 ).

However, at this time it is still a matter of much speculation as to whether the

MTP makes a signiﬁ cant contribution to mitochondrial activities under phy-

siological conditions. It could be imagined as a safety valve under stressfull

conditions, but the lack of speciﬁ city for any small - molecular - weight solute

precludes any ﬁ ne - tuning for speciﬁ c ions or solutes.