Scheffler Immo E. Mitochondria

Подождите немного. Документ загружается.

tion of damage in rapidly dividing stem cell populations, or, alternatively,

metabolically deﬁ cient cells are eliminated even more rapidly than by normal

turnover. The underlying assumption, that mtDNA turns over even in nondi-

viding cells, is based on an observation made as long ago as 1969 (183) , but

there is a dearth of information on how fast this turnover occurs in such cells.

(3) Different deletions may accumulate in different tissues of the same indi-

vidual. Such an observation lends strong support to the hypothesis that the

observed mutations are the result of stochastic processes in somatic tissue. (4)

The range and nature of the deletions vary from individual to individual;

however, as discussed in a different context, the presence of short repeats in

the mt genome may be responsible for the mechanism generating the deletions

by recombination or “ slip - replication ” (184) , and therefore the deletions are

not entirely random. (5) The tissues of an individual accumulate deletions at

different rates; that is, they do not “ age ” uniformly. (6) Even within a particular

tissue such as skeletal muscle or cardiac muscle, a tissue mosaic pattern has

been observed by histochemical staining reactions. For example, staining for

cytochrome oxidase activity in cardiomyocytes in a large collection of speci-

mens has shown rather uniform staining in samples from young subjects, but

in samples from older individuals the incidence of weakly stained or even

unstained myocytes was roughly related to the age of the individual (185) .

While provocative and consistent with present thinking, it should be stressed

that the variable levels of cytochrome oxidase activity have not been related

directly to mtDNA mutations.

Sequencing of PCR - ampliﬁ ed segments (clones) of mtDNA from a given

tissue is a tedious and time - consuming process, and hence the accumulation

of point mutations in aging tissues is less well documented and the rate of

generation of such mutations is still more difﬁ cult to determine (186) .

Apart from errors resulting from illegitimate recombination or replication,

mtDNA damage is very likely also caused by chemical reactions, and it has

been suggested by numerous authors that reactive oxygen radicals generated

as a byproduct of respiration may make a very substantial contribution to

mtDNA damage. Free radicals are known to be potent mutagens that can

induce base changes and modiﬁ cations. It has been suggested that three factors

make mtDNA particularly vulnerable: (1) The mtDNA is localized close to

the inner membrane, where free radicals are produced; (2) mtDNA is not

extensively condensed and protected by histones; and (3) the DNA repair

activity is limited (see Chapter 4 ). The resulting point mutations (after replica-

tion) may be difﬁ cult to ﬁ nd (see above), but the reaction with free radicals

(superoxide and peroxide) may also leave modiﬁ ed bases in the DNA which

can be detected by sensitive biochemical means. Several reviews on the subject

have been written by Ames and colleagues (187 – 189) . A representative

example is 8 - hydroxy - guanosine, which has been shown to accumulate in

mtDNA of the human diaphragm in individuals of advanced age (190) .

Approximately 0.5% of the guanine residues had been converted to 8 - hydroxy

guanine in an 85 - year - old individual. In a related study, the fraction of deriva-

MITOCHONDRIAL DNA AND AGING 381

382 MITOCHONDRIAL MUTATIONS AND DISEASE

tized guanines was 0.87% in the brain of a 90 - year - old. Covalent DNA altera-

tions termed I - compounds have also been described by Gupta et al. (191) in

the liver of aged rats.

Covalent modiﬁ cations of DNA in the nucleus are readily detected by DNA

repair systems, but there is agreement that some types of DNA repair cannot

be found in mammalian mitochondria. A continuous exposure to free radicals

is therefore expected to be damaging over the long run, and the extent of such

damage would be expected to be more severe in actively respiring tissue com-

pared to less active tissue. This idea has been extended in an investigation of

the effects of physical activity and age on mitochondrial function (192) , and

speciﬁ cally in the muscle from elderly athletes (193) . These authors make the

important comment that the observed decline in respiratory chain function in

human skeletal muscle from older individual is signiﬁ cantly greater than the

observed level of actual mutations. In a carefully selected and matched group

of individuals, there was no correlation between oxidative metabolism and

chronological age. Instead, the results suggested that physical activity contrib-

utes greatly to the maintenance of OXPHOS capacity; in other words, exercise

may counteract and mask the effect of accumulated mtDNA damage.

In order to assess the contribution of mutated mtDNA molecules in the

population present in “ old cells ” (i.e., cells derived from older individuals),

several informative experiments have been performed with the help of ρ

mutant mammalian cells (194, 195) . Such cells do not harbor any mtDNA of

their own, and they can be repopulated with mitochondria under investigation

by fusion with an enucleated donor cell. Such an experiment has the additional

advantage that it removes the nucleus from the old donor cell, and hence

eliminates the inﬂ uence of nuclear genes. The results of an elaborate series of

fusions and analyses have recently been published by Laderman et al. (195) .

Donor cytoplasts were from individuals ranging from 20 weeks of age (fetal)

to more than 100 years. Almost 200 clones were investigated with respect to

growth rate and respiration rate. Every clone was also analyzed to determine

the mtDNA levels per cell, since parallel experiments had shown that the

“ transformation ” of ρ

cells with mtDNA is quite variable and slow; and

oxygen consumption measurements had to be compared with, if not normal-

ized to, the mtDNA content of the cells. Thus, several clones derived from the

same mtDNA donor were also compared, and considerable variation in respi-

ratory capacity as well as mtDNA levels were observed. The ﬁ nal evaluation

of the data required careful statistical analysis, because at least two distinct

age - related phenomena were observed. First, the number of mtDNA mole-

cules in the cybrids was lower, when the enucleated donor cells were from

older individuals. Second, the same subgroup of cybrids also exhibited a lower

respiration rate. Notably, however, there was no correlation between oxygen

consumption and mtDNA content of individual clones. The authors speculated

about an increase in age - dependent mtDNA mutations that could affect the

functional properties of the complexes, the translational machinery of the

mitochondria (rRNAs or tRNAs), or even promoter sequences controlling

transcriptional efﬁ ciency. Such damage would not necessarily affect DNA

replication. On the other hand, mutations in a region crucial for DNA replica-

tion could have affected the efﬁ ciency of transforming the ρ

cells with

mtDNA.

These results clearly support the hypothesis of a functional deterioration

of mtDNA with age, since presumably all mitochondrial proteins encoded by

nuclear genes and lipids from the old enucleated cell will have been replaced

at the time of the analysis. One may also wonder why not all transformants

derived from old cells were partially respiration - deﬁ cient, and the answer is

most likely that aging is accompanied by stochastic mtDNA mutations with

some cells being much less affected than others. Finally, it should be noted that

any such in vitro experiments require the establishment of clonal cell popula-

tions of sufﬁ cient size to be able to perform the analyses, and hence the cells

must grow, albeit at more or less reduced rates. Selections and segregation

were considered by the authors by following the long - term behavior of some

of the clones, which had initially all been analyzed 8 – 18 weeks after fusion.

Additional changes were observed in some cases.

Parenthetically, it should be mentioned that the ρ

cells were derived from

143B.TK

–

cells, an immortal human osteoblastoma cell line. Do such tumor cell

lines accumulate mtDNA mutations after extensive subculture in vitro , or are

such mutations continuously subject to counterselection? A comparison of the

mtDNA sequence of a 143B.TK

–

cell with the mtDNA sequence in the original

tumor and in normal somatic tissue would perhaps be interesting. Alterna-

tively, mtDNA sequences from HeLa cell strains kept for decades in different

laboratories would also be worthy of investigation to learn why mtDNA muta-

tions do not seem to accumulate signiﬁ cantly in rapidly proliferating cells.

7.4.3 Neurodegenerative Diseases

A large literature exists describing mitochondrial dysfunction and oxidative

stress as a signiﬁ cant factor in neurodegenerative diseases, many of which

develop with age or have a pronounced delay of onset (196, 197) . The most

prominent among them include Alzheimer ’ s disease. Parkinson ’ s disease,

Huntington ’ s disease, and amyotrophic lateral disease (ALS). Since many

mitochondrial diseases, especially those due to mutations in mtDNA, also

have a delayed onset, it was tempting to look for an excess of mtDNA damage

in those patients, or even for preexisting mtDNA mutations in pedigrees with

a familial incidence of these diseases (198 – 201) . The subject has remained

controversial and several large surveys have failed to ﬁ nd mtDNA mutations

as a primary cause of these diseases (202) . There is a strong emerging consen-

sus that mitochondrial energy metabolism, oxidative stress, ROS production,

cell death (apoptosis), and neurodegeneration are all part of a larger picture,

where the major uncertainties may be the following: What are the primary

defects? What is a deﬁ nitive cause - and - effect relationship leading to the par-

ticular pathology? Why is the observed neurodegeneration frequently restricted

MITOCHONDRIAL DNA AND AGING 383

384 MITOCHONDRIAL MUTATIONS AND DISEASE

to a limited domain of the nervous system? Are there speciﬁ c mtDNA haplo-

types that predispose toward or increase the risk of initiating a neurodegenera-

tive disease in combination with nuclear mutations?

7.4.3.1 Parkinson’s Disease Much remains to be learned about this rela-

tively common disorder, affecting approximately 1 in 40 individuals over a

lifetime. Clinical symptoms include bradykinesia, rigidity, and tremor. The

most speciﬁ c and characteristic pathological condition is the loss of dopami-

nergic cells in the substantia nigra of the brain. It can be stated at the outset

that there is no evidence for maternal inheritance in PD, and the analysis of

mtDNA in brain has not revealed any major abnormalities such as deletions

or rearrangements, or consistent point mutations that can be associated with

PD (see below).

One of the most intriguing observations hinting at a mitochondrial involve-

ment in PD was made serendipitously. An illegally produced “ designer drug, ”

MTPT (1 - methyl - 4 - phenyl - 1,2,3,6 - tetrahydropyridine), an injectable pethi-

dine analogue, was found to be neurotoxic and to induce parkinsonism in users

(203, 204) . Further experimental support for this analysis came from the induc-

tion of very similar symptoms in primates. It was discovered that the drug

speciﬁ cally destroyed dopaminergic cells in the substantia nigra. MPTP is

converted to the active metabolite 1 - methyl - 4 - phenylpyridinium (MPP

) by

monoamine oxidase in glial cells, and MPP

is transported into dopaminergic

neurons by the dopamine re - uptake pathway. It can be concentrated in mito-

chondria at millimolar concentrations by a process dependent on a membrane

potential, and there it acts as a speciﬁ c inhibitor of complex I. A more detailed

discussion of MPP

and its inhibition of complex I can be found in Chapter 5 .

For the present discussion it sufﬁ ces to state that inhibition of complex I activ-

ity has been postulated to be responsible for dopaminergic cell death.

A second promising approach has employed the inhibitor rotenone at sub-

lethal concentrations to create a rat model with many of the typical symptoms

of Parkinsonism (205 – 207) .

What makes drug - induced parkinsonism so relevant is the substantial body

of evidence showing that in the substantia nigra of PD patients, complex I

activity is reduced 30 – 40% relative to controls, but this reduction is not found

in other brain regions (reviewed in reference 208 ). Mitochondrial abnormali-

ties, and speciﬁ cally reductions in complex I activity, have also been investi-

gated in muscle and platelets of PD patients. The results and their interpretation

are still incomplete in establishing the primary cause (209, 210) .

Since 7 of 45 polypeptides of complex I are encoded by mtDNA, looking

for mitochondrial mutations in PD patients became imperative. A few reports

on more detailed mtDNA sequence analyses have appeared, including at least

one with an examination of the entire mtDNA sequence of several patients

(211) . One was a PD patient, but three others had overlapping symptoms for

PD and Alzheimer ’ s disease (AD). The conclusion was that there was no single

mtDNA mutation shared by all the patients. In particular, earlier studies at

Emory University and by Hutchin and Cortopassi (212) had found an

apparent prevalence (7 - to 200fold increase) of a mutation at position 4336

(tRNA

Gln

) in AD or AD + PD patients, compared to race - and age - matched

controls, and phylogenetic analysis. Some patients were found to be variants

in ND1 or ND5. As the sample size increased, the risk factor for AD associ-

ated with the 4336 mutation actually decreased (212) . Howell et al. (202) have

critically reviewed the evidence for mtDNA mutations in PD (and AD), with

the conclusion that there is no convincing evidence for a (primary) role of

mtDNA mutations in these neurodegenerative disorders. Other large - scale

studies have provided evidence that certain mtDNA haplogroups may be

responsible for a reduction in risk of idiopathic PD (213) .

Some very promising leads have originated from the genetic analysis of

pedigrees (often consanguineous) with familial Parkinsonism (see references

214 and 215 for reviews). A series of nuclear genes (PARK1, PARK2, . . . PARK8)

are being recognized to be associated with the familial form of the disease,

and evidence is accumulating that an interaction of environmental toxins with

the products of these genes may create oxidative damage and mitochondrial

dysfunction leading to cell death (apoptosis). The function of the proteins

encoded by these genes is very suggestive. A hallmark of the disease is the

accumulation of Lewy bodies. The main component of these Lewy bodies is a

protein, α - synuclein (PARK1/4), normally involved in synaptic vesicle forma-

tion. Certain dominant point mutations yield a protein that is postulated to be

toxic in the substantia nigra when it forms protoﬁ brils. There are claims that

a mutated α - synuclein, when overexpressed, can interact with and damage

mitochondria directly (216) . Furthermore, α - synuclein null mice are resistant

to the toxic effect of MPTP (217) . The protein parkin (PARK2) is an E3 ligase,

implicating a defect in ubiquitin - mediated protein turnover in the proteasome

as a step in the development of the disease. For example, parkin - dependent

ubiquitination may be required for the removal of oxidatively damaged pro-

teins. It has been localized to the outer mitochondria membrane, but is also

placed in the matrix in order to account for its interaction with the PINK1

kinase (see below). DJ1 is another gene/protein identiﬁ ed in two consanguine-

ous families with early onset familial PD. Its function is suggested by its

membership in a large superfamily, but it remains to be more precisely deﬁ ned

as either a redox sensor or an antioxidant protein, a chaperone, or a protease.

It has also been tentatively localized in the mitochondrial matrix. The focus

was also clearly shifted to mitochondria when several pedigrees revealed a

role of the phosphatase and tensin homologue (PTEN) - induced kinase

(PINK1) (218, 219) . This serine – threonine kinase has been localized in mito-

chondria (220) .

Several of the above genes have been studied using mice and Drosophila

as model systems. One can observe symptoms such as reduced life span, ﬂ ight

muscle degeneration, apoptosis in various tissues, and male sterility. Up -

regulation of genes related to oxidative stress is observed, and in ﬂ ies a parkin

deﬁ ciency causes a reduced expression of mitochondrial complexes such as

MITOCHONDRIAL DNA AND AGING 385

386 MITOCHONDRIAL MUTATIONS AND DISEASE

complex I and IV. An in vivo Drosophila model with the Drosophila equiva-

lent of the PINK1 gene knocked out showed phenotypes similar to those

observed with a parkin deﬁ ciency including defects in mitochondrial morphol-

ogy (221) . In summary, one can identify two pathways to parkinsonism: protein

misfolding and a defective (or overwhelmed) ubiquitin - proteasome apparatus,

and mitochondrial dysfunction. How do they interact or converge to cause the

speciﬁ c dopaminergic degeneration? Some more detailed speculations can be

found in the reviews by Abou - Sleiman et al. (214) , Gandhi and Wood (222) ,

and Kwong et al. (197) . One plausible idea is that oxidative stress and excess

ROS production leads to an abundance of abnormal proteins that overwhelm

the cellular garbage disposal machinery (the proteasome); alternatively, the

proteasome could be starved for ATP. In either case, abnormal proteins then

form the aggregates characteristic of such degenerating neurons.

7.4.3.2 Alzheimer’s Disease Alzheimer ’ s disease (AD) is an increasingly

common form of dementia in our population in which the life expectancy has

been raised signiﬁ cantly over the past decades. It is a progressive disease that

is linked to the premature death of cholinergic neurons in the brain. The

primary event responsible for this destruction has remained elusive, although

it is likely that multiple genetic and environmental factors can either trigger

it or inﬂ uence the age of onset and the rate of progression. The neuropathol-

ogy of AD includes loss of neurons and synapses as well as the extracellular

deposition of plaques (made up mostly of amyloid β peptide, A β ) and accu-

mulation of intracellular neuroﬁ brillary tangles (consisting of hyperphospho-

ryted microtubule - associated protein tau). In the pedigrees where there is

early onset and a familial pattern of inheritance, the products of several

nuclear genes have been implicated. Autosomal dominant mutations have

been mapped on chromosomes 21 (APP), 14 (PSEN - 1), and 19 (ApoE), and

the β - amyloid precursor protein (APP) as well as the presenilin 1 (PSEN - 1)

and 2 (PSEN - 2) proteins are currently subject to intense research efforts. APP

is cleaved sequentially by β and γ secretases to produce A β . The presenilins

are constituents of the secretase complex. The identiﬁ cation of these genes

from familial AD in humans has directed efforts to the study of the corre-

sponding genes (and mutations) in transgenic mice (see reference 223 ) for a

critical review of mouse models for AD). Such mouse models do not recapitu-

late the complete AD phenotype. The most interesting observation is that

amyloid β peptide accumulation alone does not cause apoptosis (loss) of neu-

ronal cells, but may initiate a secondary processes with pathological conse-

quences. One earlier review (224) had concluded that overproduction and

increased aggregation of the amyloid β peptides may be an obligatory early

step in the development of the disease, followed by deposition of diffuse

plaques which, in turn, activate glia and astrocytes to release cytokines and

acute - phase proteins. Direct neurotoxicity of the amyloid peptides or the

induced “ inﬂ ammatory ” changes lead to numerous metabolic and cytoskeletal

alterations, including changes in calcium homeostasis, and injury to mitochon-

drial functions. A fundamental question is therefore whether mitochondrial

(OXPHOS) dysfunction is a secondary effect or one of the primary causes. A

number of authors have reported a direct effect of amyloid β on mitochondrial

functions ( (225, 226) , but one has to reconcile such claims with the fact that

amyloid β plaques are extracellular. In a recent review of AD, Lin and Beal

(196) summarize a bewildering array of observations that suggest either that

oxidative stress causes A β accumulation in cells or that it alters APP process-

ing. Alternatively, APP has been reported to have a dual ER/mitochondrial

targeting sequence, and overexpression can clog the mitochondrial protein

import machinery. Finally, A β is claimed to be able to inhibit a variety of

enzymes in the mitochondrial matrix ( α - ketoglutarate dehydrogenase and

cytochrome oxidase), thus causing free radical production. How A β gets into

the mitochondrial matrix is left unanswered.

In the case of the more common, late - onset AD a genetic component has

been indicated by some tantalizing observations, but the uncertainty is indi-

cated by the statement that “ the lack of a family history is a negative risk factor

for AD. ” One possible risk factor identiﬁ ed to date is the prevalence of the

apolipoprotein E type 4 allele among affected individuals (see reference 224

for a review). The delayed onset and age - related incidence of the disease has

made it very tempting to establish a connection to mitochondrial mutations.

Indeed, the existence of a maternal relative with the disease is associated with

an increased risk of AD, indicative of a link to maternal inheritance and mito-

chondria. Numerous studies have attempted to solidify the basis for a role of

the mitochondrial genotype in AD. The results are intriguing, but the evidence

is still inconclusive or negative. As mentioned above, a recent critical review of

the relevant literature by Howell et al. (202) found little evidence in support

of a role of mtDNA mutations in the development of PD and AD. A clear dis-

tinction should be made between (a) mtDNA mutations inherited from the

mother and leading to a predisposition toward the development of AD (or an

early onset) and (b) mtDNA mutations acquired by somatic cells over a life-

time. Arguments in favor of a role of the acquired mutations have been made

emphatically by Wallace and colleagues (201, 227) , by Beal (164) , and by others

(228, 229) . Much attention was once attracted by a study correlating mutations

in two mitochondrial cytochrome oxidase genes with late - onset AD (230) .

Several results were striking and even puzzling. A follow - up study in two inde-

pendent laboratories concluded that the presumed COX mutations were found

in a pseudogene of mtDNA transferred to the nucleus (231, 232) . In a related

but less extensive study the mitochondria from 33 AD patients and 30 age - and

sex - matched controls were examined. The activity of four complexes (I – IV)

were assayed in mitochondria obtained directly from lymphocytes. None of the

patients had respiratory chain enzyme activities below the normal range. Thus,

the relationship of cytochrome oxidase activity to AD remains controversial.

7.4.3.3 Huntington’s Disease Huntington ’ s disease is well - characterized at

the genetic level as an autosomal dominant disease with 100% penetrance

(233) . The responsible gene (IT - 15) has been mapped on chromosome 4, and

the mutation belongs to the class of trinucleotide repeat polymorphisms

MITOCHONDRIAL DNA AND AGING 387

388 MITOCHONDRIAL MUTATIONS AND DISEASE

(CAG), ﬁ rst identiﬁ ed in fragile X individuals (and also found in Friedreich ’ s

ataxia; see above). The resulting abnormal protein, huntingtin (HTT), with an

expanded polyglutamine tract near the N terminal, must represent a gain in

function to explain the dominant phenotype. Its normal function is still not

completely clear, and the mechanism of the gain in function (or enhancement

of a normal function) is also obscure at this time. It appears to interact with

calmodulin in the presence of calcium. Inappropriate apoptosis triggered by

abnormal huntingtin has been proposed (234) . Its expression in a wide variety

of tissues including brain, testes, liver, heart and skeletal muscle, and lung,

among others, suggests that it has a very general function outside of the

nervous system, but selective neuronal cell death remains to be explained. It

is interesting and revealing that transgenic mice expressing only exon I with

the polyglutamine tract display most of the neurological symptoms. Thus, it

has been suggested that huntingtin is proteolytically processed to produce the

pathogenic peptide when the glutamine tract exceeds a certain length.

A possible relationship to mitochondrial dysfunction is deduced very indi-

rectly from the late onset of the disease, and possibly from the neurodegenera-

tion observed in affected individuals. Measurements of biochemical activities

in mitochondria of HD patients have also been suggestive, but here, even more

than in PD and AD, one has to approach the problem from the point of view

that the mitochondrial deﬁ ciencies are clearly secondary. Studies measuring

OXPHOS enzymatic activities in selected brain tissue have to be interpreted

with care, since the problem of normalization is not a trivial one. If a subpopu-

lation of cells degenerate and disappear, the remaining cells may have normal

activities. A more reliable determination is the measurement of an intrinsic

property, namely the fraction of mtDNAs in a given tissue with point muta-

tions or deletions. Seeking to ascertain whether the progression of HD is cor-

related with an accelerated accumulation of somatic mtDNA mutations, and

speciﬁ cally the mtDNA4977 deletion (the “ common deletion ” ), Horton et al.

(235) indeed found that HD patients have an - order - of - magnitude - higher

levels of this deletion in the frontal and temporal lobes of the cerebral cortex

compared to age - matched controls. There was no statistically meaningful dif-

ference in the corresponding comparison for the occipital lobe and the

putamen. The latter tissue is atrophied in the most advanced stages of the

disease. The apparent paradox is hypothesized to be explained by the death

of the neurons containing the highest levels of mutations. With such a scenario,

any analysis must be made at a time when the integrity of mtDNA or the

activity of mitochondrial enzymes is already in decline, but before the damage

is sufﬁ cient to kill the cells.

A biochemical link between mutated huntingtin and energy metabolism is

indicated by the following intriguing observations. The polyglutamine tract of

huntingtin has been shown to bind to glyceraldehyde - 3 - phosphate dehydro-

genase, a key enzyme in glycolysis. It is not yet clear whether this interaction

is directly responsible for the hypometabolism of glucose in HD brains (236) .

In conjunction with decreased activities of complexes II and III in mitochon-

dria of HD brain (caudate), this excess metabolism of glucose causes lactic

acidosis and elevated lactate/pyruvate ratios in HD cerebrospinal ﬂ uid. Animal

models resembling HD have been induced by the complex II inhibitors malo-

nate or 3 - nitropropionic acid (237) . Most recently, a study describes how the

pathogenic polyglutamate tract induced preferential loss of complex II by a

post - transcriptional mechanism (238) . Others have claimed that the patho-

genic polyQ peptide increases ROS production by mitochondria (oxidative

stress) (239) or an impairment of ATP production (240) without a reduction

in OXPHOS complexes. Closely related studies with an additional emphasis

on NMDA signaling and the role of Ca

in mitochondrial energy metabolism

in striatal cells have been published by Seong et al. (241) . Finally, abnormal

huntingtin and p53 have been demonstrated to form a “ deadly alliance ” (242,

243) and perhaps link activities in mitochondria with activities in the nucleus

leading to the HD pathology. Speculations have centered around mutant HTT

binding p53, up - regulation of BAX, mitochondrial membrane depolarization.

and apoptosis (196) .

7.4.3.4 Amyotrophic Lateral Sclerosis (ALS) ALS (also known as Lou

Gehrig ’ s disease) results from the degeneration of upper and lower motor

neurons in the cortex, brainstem, and spinal cord. The clinical conseuences

are progressive weakness, atrophy, and spasticity of muscle tissue. It is a fatal

neurodegenerative disease with a characteristic delayed onset. The vast major-

ity of cases ( ∼ 90%) are sporadic, and very little is understood about the

mechanism triggering the onset of the disease. About 10% of the cases are

familial, and 20% of these (i.e., 2% of the total) are caused by speciﬁ c muta-

tions in the SOD1 gene encoding Cu/Zn - superoxide dismutase. This scavenger

enzyme is generally found in the cytosol, while the Mn - superoxide dismutase

(SOD2) is located in the mitochondrial matrix. Recent observations from

several laboratories have suggested that (speciﬁ c) mutant forms of SOD1 can

be found associated with the outer mitochondrial membrane and/or with the

intermembrane space, or even with the mitochondrial matrix, somehow leading

to aberrant ROS production, and so on. Cleveland and colleagues (244) attri-

bute its effect on a clogging of the import machinery, but in a recent review

they hypothesize that a combination of several mechanisms (protein mis-

foldiong, oxidative damage, defective axonal transport, excitotoxicity, insufﬁ -

cient growth factor signaling, and inﬂ ammation) contribute to the motoneuron

damage that can spread to neighboring non - neuronal cells (245) . Binding to

Bcl - 2 and aggregation in spinal cord mitochondria could also explain the death

of these neurons (246) .

7.5 MITOCHONDRIA AND APOPTOSIS

Although cell death can occur as the result of a variety of causes, including

accidental failure of an incubator, it is “ programmed cell death ” that has

MITOCHONDRIA AND APOPTOSIS 389

390 MITOCHONDRIAL MUTATIONS AND DISEASE

attracted enormous attention in recent years. The program referred to in this

context is a genetically controlled pathway and mechanism that can eliminate

unwanted cells under three major circumstances: (1) development and homeo-

stasis, (2) as a defense mechanism against genetically damaged and hence

potentially tumorigenic cells, and (3) natural senescence and aging. The phe-

nomenon is primarily found in multicellular, eukaryotic organisms, although

it has been argued that cell death may have evolved in single - celled organisms

as a defense strategy (247) . By committing suicide, a cell may eliminate itself

as a host for a parasitic invader and hence protect related, neighboring cells

from the spread of the pathogen. In higher organisms, programmed cell death

or apoptosis is now recognized as an essential aspect of development, and

many examples could be cited. Many neurons die within a developing brain,

self - reactive T - lymphocytes are eliminated during the development of the

immune system (248) , the morphogenesis of limbs requires cell death as part

of the shaping of the ﬁ nal limb such as a hand, and a most detailed description

of the cells destined to die has been worked out for all the lineages in a devel-

oping worm, C. elegans . The immune system can take advantage of and activate

an existing program to kill infected cells. In the adult, apoptosis assures that

there is turnover and repair in many tissues without continued growth, and

certain other tissues may expand or contract reversibly in response to hor-

mones and cytokines. Another aspect of apoptosis is that this program is

inhibited or inactivated in tumor cells. Hence some genes playing a role in this

mechanism were initially identiﬁ ed as oncogenes in tumor cells. Overexpres-

sion of bcl - 2 in hematopoietic tumor cells, for example, inhibits cell death

normally induced in growth factor - deprived B - cells and others. Inducing apop-

tosis speciﬁ cally in tumor cells has become a major new strategy in the ﬁ ght

against cancer, as opposed to inhibition of tumor cell proliferation (249) .

A number of isolated systems were initially studied without an obvious

connection, but now the emphasis is on (a) the common mechanisms found

in many cells across a broad spectrum of species and (b) homologous genes

encoding functions associated with apoptosis. For example, treatment of lym-

phocytes with glucocorticoids has long been known to initiate a suicide

program, but the generality of the mechanism was not recognized until later.

A major boost to the ﬁ eld also came from the observation that very speciﬁ c

cells in the developing nematode C. elegans would die during development

and that speciﬁ c mutations could identify genes ( ced = cell death abnormal)

required for this process. This observation stimulated the search for homolo-

gous genes in other organisms and made the nematode a powerful model

system for studies of programmed cell death. A detailed description of the

physiological processes in different organisms is beyond the scope of this

treatise. It is to be reemphasized that the programmed cell death is orches-

trated by genes within the affected cell, and that common genetic pathways

may have evolved over a wide phylogenetic range. Nevertheless, since most of

this research has been performed with mammalian cells, the emphasis in the

following discussion will be on mammalian systems. A most relevant aspect of