Scheffler Immo E. Mitochondria

Подождите немного. Документ загружается.

MOLECULAR GENETICS OF HUMAN MITOCHONDRIAL DISEASES 361

tion immediately raises two questions, one less challenging, the other difﬁ cult

to answer at this time. How can such individuals survive? And how does

homoplasmy for the mutation come about?

The ﬁ rst question can be answered theoretically and experimentally simply

by showing that the point mutation leads to a “ leaky ” phenotype, a protein or

a tRNA that is still partially functional. Another possibility is that the mutation

in a tRNA affects the rate or efﬁ ciency of maturation of individual mRNAs

from the polycistronic transcript with no effect on tRNA function. While this

answer may satisfy temporarily, it immediately generates more questions,

especially in the case of homoplasmy. First, one of the outstanding and most

challenging questions in the whole ﬁ eld is to explain why only certain tissues

are affected; in the most extreme case, why, for example, do LHON patients

suffer the central vision loss (although other symptoms have also been

observed)? Second, the optic atrophy (similar to other syndromes discussed

here) is delayed in onset that is variable among affected individuals in a pedi-

gree. Third, the severity of the vision loss is also variable. Finally, pedigrees

harboring the more severe mutations in a homoplasmic state can include

family members that are not affected at all.

The distinction between severe and mild mutations is in part based on an

assessment of the amino acid substitution and the effect it might have on the

residual activity of the peptide. A Leu - to - Pro change observed in one pedigree

with homoplasmy is expressed in 80% of the individuals of the family, presum-

ably because of a drastic alteration in the conformation of the peptide. Where

amino acids are conserved in evolution, functional or spatial restraints are

TABLE 7.1 Mitochondrial Diseases Resulting from Mutations in mtDNA

Clinical Description Inheritance

Ragged

Red

Fibers Defect in mtDNA

Kearns – Sayre

syndrome (KSS)

Sporadic + Single deletion or insertion

Incomplete KSS Sporadic + Single deletion

Pearson syndrome Sporadic + Single deletion or insertion

Leber ’ s hereditary

optic neuropathy

(LHON)

Maternal − Point mutation in ND4,

ND1, CYTB, COI

Neuropathy, ataxia,

retinitis pigmentosa

(NARP)

Maternal − Point mutation in ATPase 6

MELAS Maternal + Point mutation tRNA

Leu(UUR)

MERRF Maternal + Point mutation tRNA

Lys

MIMyCa Maternal + Point mutation tRNA

Leu(UUR)

Sensorineural hearing

loss

Maternal 12 S rRNA, tRNA

362 MITOCHONDRIAL MUTATIONS AND DISEASE

indicated, and changes at such positions are expected to be more serious.

Changes found in nonconserved regions of the peptide are difﬁ cult to evaluate

in the absence of a complete structure of the complex.

In more formal genetic terminology, one can speak of the expressivity of

the mutation in individuals within affected families and the penetrance of the

mutation in the general population. Penetrance is deﬁ ned as the probability

of detecting a particular combination of genes (mutations) when they are

present (in the population or in a pedigree). This may depend on the criteria

and means of observation (measurement). In other words, the degree of pen-

etrance of a gene is a measure of the fraction of individuals possessing the

mutation and showing manifestations of the mutation. Expressivity refers to

the degree to which a particular phenotype due to a particular mutation mani-

fests its presence. “ Expressivity refers to the type of manifestation of a gene

that is penetrant ” ( 82a , page 290) . Both parameters are most likely dependent

on other genes, and in the present discussion the expression of mitochondrial

mutations will depend on a combination of nuclear genes, often referred to as

modiﬁ er genes.

While the possibility of the existence of homoplasmy in viable individuals

can be explained on the basis of a leaky mutation, the development of homo-

plasmy for a mutation in a rare pedigree in the human population is more

difﬁ cult to explain. In almost all normal individuals, ∼ 99.9% of mtDNAs are

identical with no harmful mutations. In its simplest form the problem can be

stated as follows: A rare point mutation occurs in a mtDNA in germ - line cells

of a female. At the time of origin the mutated mtDNA is one among hundreds,

if not thousands, of normal mtDNAs. How many generations does it take to

accumulate a signiﬁ cant fraction of mutated mtDNAs (heteroplasmy)? How

quickly is homoplasmy achieved?

There are indications that a signiﬁ cant change in the mitochondrial geno-

type can occur within a few generations, or even a single generation, in

mammals. Hauswirth, Laipis, and colleagues (83, 84) have taken advantage of

a common polymorphism in a herd of Holstein cows to demonstrate that a

switch between alleles differing by point mutations was frequently almost

complete from one generation to the next. In other words, homoplasmy with

respect to a speciﬁ c base deﬁ ning the alleles was achieved in the change and

transmission from mother to offspring.

7.3.5 Mitochondria and Oogenesis

To escape from this conundrum, investigators have proposed the idea of a

“ bottleneck ” : a very small number of mtDNAs, or even a single mtDNA, serve

as templates for replication to populate the oocytes from which the mtDNA

is transmitted to the offspring (85 – 87) . A critical review has challenged the use

of the term “ bottleneck ” because it can be confusing in its meaning and has

been used inconsistently by different authors (88) . Instead, the term “ sampling

and ampliﬁ cation ” has been suggested by these authors. They also point out

MOLECULAR GENETICS OF HUMAN MITOCHONDRIAL DISEASES 363

that while either random drift or a complex sampling process can readily

explain a rapid segregation of mitochondrial genomes, the real puzzle is the

stable heteroplasmy observed in some pedigrees. A female with a relatively

small fraction of mutated mtDNAs can have offspring with a signiﬁ cantly

higher proportion of mutated mtDNAs, if the mutated DNA is preferentially

ampliﬁ ed. But unless there is bias in the sampling, in the majority of her off-

spring the heteroplasmy should be reduced.

At what stage in gametogenesis or embryogenesis is the sampling and

ampliﬁ cation mechanism operational? A priori , several stages could be con-

sidered (Figure 7.1 ). One should be aware that the human egg has an estimated

200,000 mtDNAs (77) and that as much as one - third of the total DNA of an

oocyte is mtDNA. There is certainly no “ bottleneck ” at this level, in the sense

that only a small number of mitochondria are transmitted by the oocyte. After

fertilization and prior to implantation, a number of rapid zygotic (mitotic)

divisions occur (step 1 in Figure 8.1 ), apparently without any further mtDNA

replication, in which the pool of oocyte mitochondria is distributed between

the daughter cells and the number of mtDNAs per cell is reduced to the

normal range of ∼ 1000 or less (89) . At a relatively early stage in development

a primordial germ cell lineage is set aside (step 2) to migrate to the ovary, and

within the ovary a subpopulation of diploid cells (oogonia) is formed by

repeated mitotic divisions (step 3). The oogonia increase signiﬁ cantly in size,

and a multiplication in the number of mitochondria could accompany this

growth (step 4). Meiosis and maturation leads to one ovum (and polar bodies)

from one oogonium. (It should be noted that in mammals, meiosis I is arrested,

possibly for years, and completed only at periodic intervals when eggs mature.

Furthermore, meiosis II is subsequently arrested and completed only if fertil-

ization by a mature sperm takes place.) It has also been suggested that a series

of zygotic divisions with nuclear but no mtDNA replication could reduce the

number of mtDNAs to a much smaller number ( “ the bottleneck ” ), accompa-

nied by large stochastic variations in mitochondrial genotype between differ-

ent cells. In this hypothesis, segregation is a postzygotic event. Thus, after 12 – 15

zygotic divisions without mtDNA replication, oogonia could have mtDNAs

numbering in the range of 50 – 56 . If the subsequent ampliﬁ cation of mtDNA

molecules during the maturation of the primary oocyte from an oogonial

precursor cell started from such a small population, or it involved only a small

subpopulation of a larger number of mtDNA templates ( “ sampling and ampli-

ﬁ cation ” ), shifts in the distribution of “ alleles ” could be achieved quite

rapidly.

To restate the two plausible hypotheses, sampling and ampliﬁ cation could

occur in the process of formation of an oocyte from an oogonial cell, or a

reduction in mtDNA copy number could occur by simple dilution during the

formation of the germ cell lineage in the early embryo. A third hypothesis has

proposed that the sampling may occur during a series of mitotic divisions

leading to a large population of oogonia (step 3). Finally, it is conceivable that

the partitioning of the embryonic cells between the inner cell mass and the

364 MITOCHONDRIAL MUTATIONS AND DISEASE

zygotic divisions

cleavage

blastula

gastrulation

somatic cells

primordial germ cells

migration to the

ovary

oogonia

primary oocyte

secondary oocyte

+ 1st polar body

ovum

+ 2nd polar body

multiple mitotic

divisions

growth of oogonia

large increase

in cell volume

increase in number

of mitochondria

meiosis I

unequal cell division

meiosis II

Fertilization

2 Pro-nuclei

Fertilized

Egg

Figure 7.1 (A) Mitochondrial DNA ampliﬁ cation and relationship to oogenesis. The

fertilized egg contains an estimated 100,000 mtDNAs. These may be distributed during

the early zygotic divisions (step 1) to yield cells with considerably smaller populations

of mtDNA. Early in development, germ cells are differentiated from somatic cells (step

2). A very limited number of mtDNAs could be replicated during step 3 (expansion of

the precursor pool of oogonia by mitotic divisions), or in step 4 when oogonia increase

in size and prepare for meiosis. (B) A model for the origin of the bottle neck based on

the assumption that the ﬁ rst n zygotic divisions occur without any mtDNA replication,

diluting the population of mtDNA molecules to a small number per cell at the time

when germ line cells are set aside. This small number then expands again during meiosis

and formation of the mature oocyte.

MOLECULAR GENETICS OF HUMAN MITOCHONDRIAL DISEASES 365

extraembryonic tissues is accompanied by an unequal partitioning of mtDNAs

which could be responsible for the segregation and reduction of mtDNA

genotypes in the ICM. A review by Chinnery et al. (90) presents these hypoth-

eses with more references to the original literature, and it also illuminates

some interesting differences in oogenesis between mice and humans. They

illustrate that precise quantitative predictions for genetic counseling based on

mouse models may be difﬁ cult.

Several experimental systems have been developed in recent years to

achieve “ cracks in the bottleneck ” (85) Heteroplasmic mouse strains have now

been created to be used to understand the mechanisms for segregation of

genetically marked mtDNAs (48) (49, 91, 92) . Advances in the manipulation

of mouse embryos have been instrumental in creating the kind of experimental

conditions allowing a direct observation of the segregation of mtDNAs in

heteroplasmic mice. Jenuth et al. (93) produced mouse embryos by the fusion

of cytoplasts containing mitochondria of one haplotype with one - cell embryos

of a strain with another mitochondrial haplotype. For simplicity the mtDNA

genotypes will be referred to as NZB and BALB, respectively. The cytoplasts

were obtained from one cell embryos by using a glass pipette to draw up and

pinch off cytoplasm surrounded by plasma membrane. These were injected

under the zona pellucida of recipient one - cell embryos and electrofused.

Transplantation into pseudopregnant females produced female and male

fertilized egg (zygote)

100-500,000 mtDNA

(

of maternal origin)

n zygotic divisions

with no

mtDNA replication

1 - 5 x 10

cells

= m = mtDNAs/cell

somatic cell

lineages

early germ line cells

large pool of oogonia

oocyte

1 - 5 x 10

mtDNAs/cell

expansion of mtDNA

pool from m to > 100,000

Figure 7.1 (Continued)

366 MITOCHONDRIAL MUTATIONS AND DISEASE

animals whose mtDNA could be analyzed from tail biopsies. Five founder

females were produced with 3.1 – 7.1% of the donor mtDNA genotype. When

a female with 5% donor (NZB) mtDNA was crossed to a BALB male, 28

offspring were analyzed and found to contain, on average, 6.7% NZB mtDNA.

However, in individual offspring the proportion of donor (NZB) mtDNA

varied from 0% to 29.6%. A rigorous statistical analysis of 191 F

offspring

from all ﬁ ve crosses, along with 176 offspring from several more back - crosses

with F

females, indicated a large variance of haplotype frequencies. It was

concluded that the results supported the idea that germ - line segregation is a

stochastic process capable of producing signiﬁ cant ﬂ uctuations from an origi-

nal distribution of genotypes.

In mice approximately 50 primordial germ cells can be identiﬁ ed in 7.5 - day

mouse embryos. When several (3 – 12) such primordial germ cells from a given

embryo were analyzed, the coefﬁ cient of variation was very small, suggesting

that a given proportion of donor mtDNAs received in the zygote was main-

tained during the earliest stages up to the establishment of the germ cell

lineage. These primordial germ cells, when properly located on the genital

ridge, divide and give rise to oogonial cells, from which approximately 25,000

primary oocytes are derived in a series of synchronous divisions. Based on

these experimental results, Jenuth et al. (93) propose that segregation of

mtDNA variants must occur during the mitotic cell divisions that precede

meiosis in the ovary.

At this point, one can only speculate about details of the process (86, 93) .

Several aspects of mtDNA replication must be considered (see Chapter 4 ).

Unlike the case of the nuclear genome, mtDNA replication is relaxed; that is,

in a given population of mtDNAs, some may replicate multiple times during

the cell cycle, while others do not replicate at all. This still leaves the question,

How many tDNA molecules are selected for replication, and is there any kind

of selection for a subpopulation? There is evidence that localization within a

cell can inﬂ uence mtDNA replication, with proximity to the nucleus being a

determining parameter (94) . Employing mathematical modeling using popula-

tion genetic theory, along with assuming that segregation is the result of

genetic drift, one can calculate the effective number of segregating units, N ,

from the observed variance in the nth generation, if the initial distribution of

genomes is known. With an estimated 15 cell divisions required to produce the

total population of primary oocytes in the mouse ovary, Jenuth et al. (93) cal-

culated that N is in the range of 200.

7.3.6 Clinical Aspects of Mitochondrial DNA Mutations

In humans, the distinction between variants with point mutations and variants

with large deletions is an important one. The fact that changes are occurring

post - fertilization is clearly indicated by the differences in heteroplasmy

observed between muscle tissue and blood cells in patients with mtDNA dele-

tions. Homoplasmic cells with deleted mtDNA would not be expected to be

MOLECULAR GENETICS OF HUMAN MITOCHONDRIAL DISEASES 367

viable. Whatever the starting position in the fertilized egg (zygote), an unre-

solved issue is whether there is genetic selection in somatic cells in speciﬁ c

tissues, or simply random genetic drift due to relaxed mtDNA replication.

7.3.6.1 MtDNA Deletions: Kearns –Sayre Syndrome and Pearson Syn-

drome It need not be stressed that homoplasmic null mutations would be

lethal. Two distinct mechanisms can lead to partial deﬁ ciency. In heteroplas-

mic cells the most severe mutation, a complete deletion of one or more genes,

can be partially compensated by the presence of normal mitochondrial

genomes. In homoplasmic cells a mutation must be less severe — for example,

by causing a missense substitution in a protein which reduces its activity. In

either case, one expects the tissue to suffer from an inadequate production of

ATP by oxidative phosphorylation; and one may now ask, What is it that dis-

tinguishes the various “ mitochondrial diseases ” from each other. A priori , one

would expect the severity of the disease to depend on the degree to which a

particular function is compromised. In the same vein, one would expect the

deﬁ ciency to be most deleterious in tissues such as the brain and muscle which

are energetically the most demanding. In practice, the situation is much more

complicated in ways that continue to challenge our understanding of the basic

phenomena.

Large - scale rearrangements or deletions of mtDNA are associated with two

clinical conditions that share overlapping symptoms. The ﬁ rst, recognized by

Kearns and Sayre in 1958 (95) includes among the symptoms retinitis pigmen-

tosa, external ophthalmoplegia, a complete heart block. Morphological altera-

tions in the brain were also noted at autopsy, and these have since been

conﬁ rmed as the rule rather than an exception or artifact. The disease is char-

acterized by a childhood onset, progressive ophthalmoparesis, pigmentary

retinopathy, increased propensity of complete heart block (which could lead

to sudden death), and neurological symptoms including incoordination, mental

retardation, and episodic coma. Growth retardation and hearing loss are

common, and ragged red ﬁ bers are found in almost all affected patients. An

unbalanced and insufﬁ cient energy metabolism is indicated by elevated lactate

and pyruvate levels in plasma. Named after its original discoverers, the syn-

drome is now referred to as the Kearns – Sayre syndrome (KSS). As explained

earlier, the occurrence of KSS is sporadic not familial, tissues are heteroplas-

mic with respect to wild type and rearranged mtDNA, and both sexes are

affected at about equal frequencies.

A second syndrome named after its discoverer is Pearson syndrome (96) .

The affected infants had pancytopenia and altered pancreatic exocrine func-

tion, and it was fatal in some cases, but others survived into adulthood and

developed clinical features of KSS. When molecular genetic analyses became

possible, lesions in mtDNA were found very similar to those found in KSS.

How is this syndrome survivable, only to be converted into the second severe

condition? Speculation centers around cell population dynamics in the bone

marrow, where cells with the greater proportion of mutated mtDNAs have a

368 MITOCHONDRIAL MUTATIONS AND DISEASE

disadvantage during rapid cell divisions and may be eliminated; that is, in this

tissue there may be a selection for cells with the highest proportion of normal

mtDNAs. At the same time, in tissues with nondividing cells like muscle or the

brain there may be a selection at the level of mtDNA replication which favors

the deleted mtDNAs, and ultimately an oxidative phosphorylation deﬁ ciency

is developed concomitant with encephalomyopathy. Patients diagnosed in

infancy with Pearson syndrome do not have ragged red ﬁ bers, whereas muscle

biopsies showing an abundance of ragged red ﬁ bers also show deleted mtDNAs

in excess.

Patients with Pearson syndrome and those KSS patients with the most

severe symptoms have detectable heteroplasmy not only in muscle, but also

in ﬁ broblasts and leukocytes, tissues easily available from biopsies. At autopsy,

brain and liver tissue (and other tissues) become available, with the brain

showing pronounced heteroplasmy or an abundance of mutated mtDNAs,

while liver may have a smaller proportion. On the one hand, the ubiquitous

presence of mutated mtDNA is an indication that the mutation was present

in the oocyte before fertilization, but the variability in the ratio of normal to

mutated DNA in different tissues is indicative of perhaps two selection mecha-

nisms based on (1) the ﬁ tness of individual cells in rapidly dividing or renewing

tissue (blood, liver), and the replicative advantage of mutated mtDNA mole-

cules in nondividing tissue. As a result, these distributions may vary between

patients and, most signiﬁ cantly, may shift with age. Delayed onset and variable

expressivity may be explained quite satisfactorily in general terms, but more

deﬁ nite predictions at the time of the ﬁ rst diagnosis may not be possible. Even

a determination of the proportion of mutated DNA in a muscle biopsy (which

muscle?) may not have any predictive value, since there is no correlation

between this fraction and the clinical severity.

It is commonly assumed that there are several mtDNAs per mitochondrion,

and that intra - mitochondrial complementation occurs. Continuous ﬁ ssion and

fusion of mitochondria would be expected to contribute to a mixing of the two

genomes. However, such a picture, derived from observations in tissue culture,

may be misleading in tissues such as muscle, where the highly structured cyto-

plasm and the cytoskeleton (sarcomeres) may prevent extensive diffusion and

mixing of mitochondria.

The severity and the precise combination of symptoms may also be depen-

dent on the nature of the rearrangement and the speciﬁ c segment of mtDNA

missing in heteroplasmic cells. Nevertheless, for reasons explained above,

patients with identical deletions may exhibit different clinical abnormalities.

Almost any deletion of some length will cause the loss of several genes, includ-

ing structural genes encoding essential subunits for complexes I, II, and IV,

alone or in combination, and almost always the loss of at least one tRNA gene.

Thus, very speciﬁ c predictable biochemical activities will be compromised, but

the loss of one or more tRNA genes will also diminish the capacity for mito-

chondrial protein synthesis, and hence the level of all complexes of the elec-

tron transport chain may be affected. A more detailed discussion of several

MOLECULAR GENETICS OF HUMAN MITOCHONDRIAL DISEASES 369

speciﬁ c cases can be found in the review of Harding (73) . Superimposed on

the effects of the speciﬁ c deletions are (a) mechanisms of post - transcriptional

regulation such as mRNA stability and the control of steady - state levels of

individual mRNAs and (b) translational control mechanisms (initiation?) that

may inﬂ uence the half - life of an mRNA. If such mechanisms are tissue - speciﬁ c,

a simple consideration of heteroplasmy and gene dose is deﬁ nitely not sufﬁ -

cient to predict the full scope of clinical symptoms and their severity.

7.3.6.2 Familial Mitochondrial DNA Depletion A rare disorder with

an inevitably fatal outcome in early childhood has been studied in several

families. Healthy parents have produced children suffering a severe depletion

of mtDNA in speciﬁ c tissues, most frequently liver or muscle. First recognized

by Moraes et al. (97) as a novel genetic abnormality, the defect appears to

reside in a nuclear gene, since no mitochondrial mutations could be

detected.

7.3.6.3 Point Mutations Table 7.1 lists ﬁ ve distinct clinical conditions

resulting from point mutations in mtDNA which are maternally inherited.

They are: LHON, NARP, MELAS, MERRF, MIMyCa, and sensorineural

hearing loss. A genetically interesting and challenging feature of these diseases

is the frequent ﬁ nding of homoplasmy (see discussion in Section 7.3.5 ); and a

perhaps oversimpliﬁ ed aspect of each mutation is that it is “ leaky, ” which in

biochemical terms implies that some function is retained. A detailed con-

sideration of each condition and patients having the mutation leave many