Scheffler Immo E. Mitochondria

Подождите немного. Документ загружается.

with minimal harmful effects on mitochondria were combined to study mito-

chondrial distributions and dynamics in living cells. A landmark paper describ-

ing the laser dye rhodamine 123 was published by Chen ’ s group in 1980 (30) ,

illustrating the distribution of mitochondria in a variety of mammalian cell

types. It also already gave descriptions of alterations observed in cell transfor-

mation, and as the consequence of the disruption of microtubules by colchi-

cine. Many studies followed and have been expertly evaluated and summarized

at various times (31 – 34) . A following discussion (Chapter 5 ) will address in

more detail the use of lipophilic cationic dyes to measure the membrane

potential and the possibility of heterogeneity within a mitochondrial popula-

tion of a cell with respect to staining, hence membrane potential, reﬂ ecting a

difference in metabolic activity. The present discussion will deal with the dis-

tribution of mitochondria within cells and the mechanisms by which charac-

teristic or functionally meaningful distributions are achieved or maintained.

The review by Bereiter - Hahn (34) , in particular, includes many historical refer-

ences and is comprehensive in its coverage of all aspects of the dynamics of

mitochondria in living cells.

Like in so many other areas of cell biology, the green ﬂ uorescent protein

(GFP) has been used as a versatile reporter to observe the dynamic behavior

and ﬂ uctuating distribution of mitochondria in various cell types. When

produced as a chimeric protein with a mitochondrial targeting sequence, it

has been localized into the mitochondrial matrix of yeast cells (35) and HeLa

cells (36) and in the intermembrane space (37) . Combined with powerful

new imaging technology and computers, high - resolution 3 - D images can be

recorded almost continuously with live cells. The conclusions on the dynamic

aspects of mitochondria will be presented below. In studies of HeLa cells with

differential labeling of mitochondria and the endoplasmic reticulum (ER), the

authors reported close appositions between mitochondria and the ER, and

they estimated that about 5 – 20% of the mitochondrial surface was involved

in such contacts (38) . A possible signiﬁ cance is that at such sites, microdomains

are created where Ca

is released from stores in the ER mediated by IP

However, caution is in order, because both mitochondria and the ER network

can become attached to microtubules via kinesin or dynein - like motors, and

thus their physical proximity may be coincidental. Furthermore, the density of

these subcellular structures also makes apparent contacts quite likely. A

broader discussion of the interaction of the smooth endoplasmic reticulum

with mitochondria has been presented recently (39) .

It is clear that a consideration of a single generic cell does not sufﬁ ce in

discussions of the broad range of observations made on mitochondrial distri-

butions in cells, and the reader is referred to specialized reviews on the subject

(33, 34) . Some general principles will be emphasized.

Mammalian cells in tissue culture (ﬁ broblasts, cancer cells, etc.) have been

popular for study because of ease of access, and important deductions have

been made. When certain highly differentiated vertebrate or invertebrate cells

are investigated, special considerations have to be included. Unique situations

INTEGRATION INTO THE CELL 31

32 STRUCTURE AND MORPHOLOGY. INTEGRATION INTO THE CELL

may be encountered in cells of lower eukaryotes, with the kinetoplast of try-

panosomes being one speciﬁ c example. The discussion of events associated

with cell divisions and the cell cycle usually ignores mitochondria, but it is

obvious that there must be some mechanism assuring that each daughter cell

obtains half of the mitochondria of the dividing cell. This creates a special

problem for the budding yeast, Saccharomyces cerevisiae , to give one example,

which will be addressed below.

What dictates the distribution of mitochondria within a cell? In large and

highly asymmetrical cells, mitochondria may have to be localized in regions

where the ATP supply has to be high. Thus, mitochondria wrapped around

or packed along the axonemes of sperm tails can be easily rationalized.

(Figure 3.6 ).

In neurons, mitochondria have to be moved and distributed along the often

considerable length of axons and into the growth cone (40, 41) . Skeletal muscle

cells require a very uniform distribution of mitochondria in the myoﬁ brils

along their entire length to provide a uniform supply of ATP to all the sarco-

meres (Figure 3.7 ). A particularly impressive example of a highly regular

arrangement of mitochondria in striated muscle cells is found in insect ﬂ ight

muscles.

Rod cells in the retina appear to have their mitochondria concentrated in

the cell body closest to the outer segment. In round yeast cells or in a gamete

of Chlamydomonas the mitochondria are distributed around the cell peri-

phery, perhaps to maximize their access to oxygen. Thus it has been speculated

Figure 3.6 Mitochondria aligned along an axoneme of a sperm tail. (From The Cell ,

by D. W. Fawcett, 1981, W. B. Saunders Co.)

Figure 3.7 (A) Mitochondria distributed in skeletal muscle between sarcomeres.

(From The Cell , by D. W. Fawcett, 1981, W. B. Saunders Co.) (B) Cross section of cardiac

muscle showing uniform distribution of mitochondria. (From The Cell , by D. W. Fawcett,

1981, W. B. Saunders Co.)

INTEGRATION INTO THE CELL

34 STRUCTURE AND MORPHOLOGY. INTEGRATION INTO THE CELL

that ATP (and/or ADP) gradients might be responsible for inﬂ uencing mito-

chondrial distributions. Alternatively, the clustering of mitochondria within a

cell might create gradients of other metabolites and ions such as calcium and

cause transient, highly localized perturbations, including pH gradients.

3.2.2 Interaction with Cytoskeleton

Mitochondria often tend to have an elongated, thread - like appearance, and

when viewed in live cells they appear to be roughly aligned along their long

axes. When they were colocalized with cytoskeletal structures, their association

with microtubules became immediately apparent in several cell types. In fact,

mitochondria surrounding the meiotic spindle in insect cells were noted before

the composition of the mitotic and meiotic spindles was known. Observations

in the laboratory of Singer (28, 29) (Figure 3.8 ) and of Chen (30) provided

deﬁ nite proof, and many more observations in diverse cell types have followed

these early studies.

One interpretation is that mitochondria are simply trapped and partially

aligned within this cytoskeletal array of microtubules and are further dis-

persed by their own “ intrinsic motility ” (shape changes, ﬁ ssion, and fusion; see

below). At the other end of the spectrum of interpretations, it has been argued

that molecular motors attached to the outer membrane of mitochondria can

become engaged with microtubules. In other words, dynein - and kinesin - like

molecules have been looked for, initially based on the differential sensitivity

of these motors to speciﬁ c inhibitors. The evidence clearly points to an ATPase

activity bridging the outer mitochondrial membrane and microtubules (34) .

Before discussing the locomotion of mitochondria along cytoskeletal struc-

tures, it is relevant to consider whether the elongated shape of mitochondria

may be related to or even dependent on their attachment to ﬁ lamentous or

long tubular structures, in the same way as the ﬂ attening of cells on a substrate

is dependent on interactions between integral membrane proteins in the

plasma membrane and extracellular matrix elements. In other words, is the

sausage shape of mitochondria observed in vivo retained when mitochondria

are isolated and highly puriﬁ ed in vitro ? Unfortunately, the isolation of mito-

chondria tends to break up mitochondria, and the resulting mitochondrial

preparations typically consist of smaller, spherical or only slightly elongated

vesicles. A more gentle test of the dependence of mitochondrial morphology

on intact microtubules can be made by treating intact cells with colchicine, and

one of the earliest experiments of this type showed clearly that the distribution

or orientation of mitochondria was perturbed, but their threadlike appearance

was not altered by such a treatment. Therefore, some internal mechanism must

be active that controls the shape. In the absence of an internal mitochondrial

skeleton, the answer has to be found in understanding the maintenance of the

varying curvature of the membranes. Under some conditions, however, deﬁ ned

by speciﬁ c mutations in extra - mitochondrial proteins, giant spherical mito-

chondria can be observed in yeast (see below).

Figure 3.8 Correlation of microtubule alignment and mitochondria (27, 28) . Anti-

bodies against beef heart cytochrome oxidase and against chick tubulin were used in

indirect immunoﬂ uorescence microscopy to localize microtubules and mitochondria

in permeabilized cells. (A) Normal rat kidney (NRK) cell stained with antitubulin

antibody and rabbit anti - chick tubulin IgG. (B) NRK cell stained with rabbit anti -

cytochrome oxidase antibody and goat anti - rabbit IgG. (C, D) NRK cells double -

stained with anti - tubulin and anti - cytochrome oxidase antibodies. The axis of elongated

mitochondria ( arrow ) is always parallel to the microtubule axis.

A B

C D

Neurons, axons, and axonal transport have received the maximum of atten-

tion, since both retrograde and anterograde movement of mitochondria

in axons has been observed by video - enhanced microscopy in intact axons

as well as in extruded axoplasms (41) . Closely related studies with systems

INTEGRATION INTO THE CELL 35

36 STRUCTURE AND MORPHOLOGY. INTEGRATION INTO THE CELL

of highly aligned microtubules of uniform polarity have been possible in

unsheathed nutritive tubes of insect ovaria and in permeabilized reticulopodia

of the giant amoeba Reticulomyxa . (The primary references to these studies

are found in reference 34 .) A characteristic aspect of such movements is the

saltatory nature of the displacements typical for motor - driven movements on

cytoskeletal elements, and not compatible with simple Brownian motion.

Unresolved problems are related to the bidirectional displacements, the spe-

ciﬁ c speeds observed, the effectiveness (or lack of it) of known inhibitors such

as vanadate (a dynein inhibitor in vertebrates), and the possible contribution

of additional mechanisms (e.g., an actin - based motility).

A series of reports from the laboratory of Hollenbeck serves to illustrate

the experimental approaches and the conclusions drawn from the observations

(42 – 48) . In one study, chick sympathetic neurons were cultured in the continu-

ous presence of either cytochalasin E to eliminate microﬁ laments (F - actin,

MF), or in the presence of vinblastine or nocodazole to eliminate microtubules

(MTs). Mitochondrial movements were followed by computer - aided video

analysis after staining the organelle with rhodamine 123. In such neurons,

mitochondria continued to move with saltatory movements in both directions

in neurites with MTs, but no MFs, but failed even to enter the neurites in the

absence of MTs but presence of MFs. The dependence of mitochondrial move-

ment on MTs was conﬁ rmed, and it appeared that MFs by themselves were

insufﬁ cient to support the movement of mitochondria into the neurites. Inter-

esting and partially contrasting results were obtained when the neurons were

ﬁ rst cultured in the absence of any drug to allow for the normal development

of axons. Subsequently, the mature neurons were treated with either cytocha-

lasin E or with vinblastine. The drugs disrupted the cytoskeletal structures as

before, but in this case the axons already contained mitochondria. Now the

average mitochondrial velocity increased in both directions in axons lacking

only MFs, but net directional transport decreased. Surprisingly, in axons with

MFs but no MTs, mitochondria also moved in both directions at a reduced

average velocity and excursion length; net retrograde transport was favored

under these conditions. When both drugs were added to eliminate both MTs

and MFs, but leaving neuroﬁ laments, no movement was observed. The studies

clearly support an actin - based motility of mitochondria over shorter distances

and a faster MT - based motility over long distances. They suggest that mito-

chondrial movement in neurites is a complex process involving several motors

(kinesins, dyneins, and myosins), each with characteristic intrinsic properties

and presumably able to ﬁ ne - tune mitochondrial distributions for optimal func-

tioning of the axon (47) . The discussion must also include the nature of various

linkers, adaptors, or anchors on the surface of the organelle as reviewed by

Hollenbeck (46) . Finally, attention has now shifted to cellular signals modulat-

ing the activity of the motors and their attachment to mitochondria. An assort-

ment of kinases, a GTPase (Miro in Drosophila ), G proteins, and growth

factors (NGF) have been shown in a variety of model systems to inﬂ uence

axonal transport of mitochondria (see reference 47 for a recent collection of

THE DYNAMICS OF MITOCHONDRIAL MORPHOLOGY 37

references). Miller and colleagues have made the very interesting observation

that mitochondria with high membrane potentials are transported primarily

toward the growth cone, while mitochondria with low potential were trans-

ported toward the cell body (49) . This may reﬂ ect an aspect of quality control

by turnover (mitophagy; see below).

There have been numerous investigations aimed at establishing an interac-

tion between mitochondria and intermediate ﬁ laments. A co - distribution has

been observed in a few instances (see reference 34 for references), but since

there are few systems with highly organized, regular arrays of such ﬁ laments,

model systems and especially cell - free systems are lacking. Some observations

point to a role of desmin in anchoring mitochondria in skeletal muscle cells

to preserve their relative positions during repeated contraction – relaxation

cycles, but a more deﬁ nite proof is lacking. In non - muscle cells deliberate

attempts to demonstrate an alignment of mitochondria with intermediate ﬁ la-

ments have failed.

Further impetus to relate mitochondria and intermediate ﬁ laments has

come from pionering studies in Yaffe ’ s laboratory (50) with the yeast Saccha-

romyces cerevisiae . Taking a genetic approach, this laboratory has isolated

mutants defective in mitochondrial inheritance; that is, these mutants fail to

transfer mitochondria into daughter buds. A gene for an essential protein was

thus identiﬁ ed ( MDM1 ), and the Mdm1 protein was found to have modest

sequence similarity to animal intermediate ﬁ lament proteins. In vitro this

protein can form intermediate ﬁ lament - like (10 nm) structures, while in vivo

immunochemical staining reveals a punctate distribution of the antigen. A

further discussion will be deferred to one of the following sections.

3.3. THE DYNAMICS OF MITOCHONDRIAL MORPHOLOGY

3.3.1 Mitochondrial Shape Changes

The phenomenon referred to as “ intrinsic mitochondrial motility ” (34) has

been observed in animal cells, in fungal hyphae, in algal cells, and in the cyto-

plasms of plant cells where cytoplasmic streaming is absent to allow longer -

term observations. A mitochondrion appears to be able to slither, creep, or

slide at rates ranging from 2 to 30 μ m/min in the direction of its long axis. The

displacement is accompanied by subtle shape changes: slight thickening or

contractions along the axis, possibly reﬂ ecting localized changes in the internal

arrangement of the cristae. Microinjection studies with ATP and/or ADP have

shown that these reagents can immobilize mitochondria; and in general, the

inﬂ uence of metabolic conditions on mitochondrial motility have been inter-

preted in favor of intrinsic activities driving their displacements (34) .

Intrinsic dynamic changes in mitochondria are most strongly indicated by

the well - documented studies on mitochondrial fusions and ﬁ ssions. Such events

can be followed in the phase - contrast microscope by time - lapse photography

38 STRUCTURE AND MORPHOLOGY. INTEGRATION INTO THE CELL

(34) ; and in some cases, observations on live cells have been elegantly related

to follow - up examinations with the electron microscope (see Figure 3.9 ) .

As a result of the combination of the power of genetics with the in vivo

visualization of mitochondria in the yeast Saccharomyces cerevisiae , an explo-

sion of papers has occurred in the past few years, contributing to an under-

standing of mitochondrial shape changes (fusion, ﬁ ssion, and tubulation). A

variety of novel proteins were identiﬁ ed, and their characterization was fre-

quently extended to mammalian cells by “ homology cloning. ” In other words,

genetic studies in yeast have generally provided guidance for studies in

mammalian cells. Many, but not all, proteins characterized in yeast have been

identiﬁ ed in mammalian cells, either because they cannot yet be recognized

or because there are mechanistic differences between organisms. The growth

of the relevant literature dictates that reference is made primarily to review

articles (51 – 58) . The recent reviews by Okamoto and Shaw ( 59 ) and by Hoppins

et al. ( 59a ) are noteworthy for being up-to-date, authoritative, comprehensive,

and illustrated with highly informative drawings. Table 3.1 and Figures 3.11

and 3.12 represent summaries of information from these reviews.

Historically, fusion - defective mitochondria were ﬁ rst discovered in sterile

Drosophila males (see below). Subsequently, the discovery and observations

of yeast mutants with defective mitochondrial morphologies deﬁ ned three

major mechanisms/pathways in genetic and biochemical terms. When fusion

is interrupted, ongoing ﬁ ssion produces a large number of mitochondrial frag-

ments. When ﬁ ssion is defective, ongoing fusion leads to extensive intercon-

nected nets. A third pathway that so far is unique to fungi has been called

tubulation; defects in this pathway cause tubular mitochondria to form large

spheres.

Serial sections and image reconstructions have revealed that in the yeast

Saccharomyces cerevisiae a cell contains a small number of mitochondria exist-

ing in a highly branched, reticulated shape, such that a section viewed in the

electron microscope gives the appearance of multiple mitochondria (Figure

3.10 ) . One to ten such reticular structures are expanded below the cortex of

the cell (22, 23) . A state of the art reexamination of this feature and its dynamic

aspects in live cells was pioneered by Nunnari and colleagues (35) , who

modiﬁ ed a green ﬂ uorescent protein (GFP) for import into mitochondria to

be able to use wide - ﬁ eld ﬂ uorescence microscopy for the acquisition of three -

dimensional images over long time intervals. The use of GFP avoids the prob-

lems of photobleaching. Beautiful 3 - D images were obtained, but more

signiﬁ cantly, time - resolved images from single optical sections yielded quanti-

tative data on fusion and ﬁ ssion events. While the authors are cautious to state

that the resolution does not unambiguously allow the distinction between

fusion and close juxtaposition, many events leading to stable connections were

documented. During vegetative growth of wild - type cells, 0.39 ± 0.14 ﬁ ssion

events per minute and 0.40 ± 0.08 fusion events per minute were observed,

indicative of a highly dynamic situation. The matching ﬁ gures indicate a steady

state in which a continuous network undergoes constant fragmentation and

THE DYNAMICS OF MITOCHONDRIAL MORPHOLOGY 39

Figure 3.9 (A) Dividing mitochondrion. (Original photograph of T. Kanaseki; from

The Cell , by D. W. Fawcett, 1981, W. B. Saunders Co.) (B) Successive stages in a dividing

mitochondrion. (Original photograph of T. Kanaseki; from The Cell , by D. W. Fawcett,

1981, W. B. Saunders Co.)

re - assembly. They conﬁ rmed previous observations that at least one tip (end)

was involved, where the putative fusion machinery may be located.

A similar study with GFP in HeLa mitochondria (36) also obtained high -

resolution 3 - D images at 30 - s intervals, and it was concluded that these mito-

chondria also formed a connected, continuous, and highly dynamic network

40 STRUCTURE AND MORPHOLOGY. INTEGRATION INTO THE CELL

exhibiting growth and retraction, as well as fusions to other portions of the

network. Photobleaching of a small area within a cell was followed by a rapid

recovery (5 – 20 min) of GFP ﬂ uorescence within that area.

3.3.1.1 Fission Fission is the postulated mechanism for mitochondrial

proliferation. The insertion of new components into the matrix and all the

Figure 3.10 Mitochondria in living yeast cells (35) . Green ﬂ uorescent protein was

modiﬁ ed to be imported into mitochondria. Wide - ﬁ eld ﬂ uorescence microscopy was

used for the acquisition of three - dimensional images over long time intervals. (A)

Three - dimensional images of yeast mitochondria in wild - type cells grown on galactose.

(B) Fission and fusion of mitochondria. Time - resolved images from a single optical

section of 0.2 μ m. Time intervals from left to right in 3 - min intervals.