NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Alberts B, Johnson A, Lewis J, et al. Molecular Biology of the Cell. 4th edition. New York: Garland Science; 2002.

  • By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.
Cover of Molecular Biology of the Cell

Molecular Biology of the Cell. 4th edition.

Show details

The Genetic Systems of Mitochondria and Plastids

It is widely accepted that mitochondria and plastids evolved from bacteria that were engulfed by nucleated ancestral cells. As a relic of this evolutionary past, both types of organelles contain their own genomes, as well as their own biosynthetic machinery for making RNA and organelle proteins. Mitochondria and plastids are never made from scratch, but instead arise by the growth and division of an existing mitochondrion or plastid. On average, each organelle must double in mass in each cell generation and then be distributed into each daughter cell. Even nondividing cells must replenish organelles that are degraded as part of the continual process of organelle turnover, or produce additional organelles as the need arises. The process of organelle growth and proliferation is complicated because mitochondrial and plastid proteins are encoded in two places: the nuclear genome and the separate genomes harbored in the organelles themselves (Figure 14-50). In Chapter 12, we discuss how selected proteins and lipids are imported into mitochondria and chloroplasts from the cytosol. Here we describe how the organelle genomes are maintained and the contributions they make to organelle biogenesis.

Figure 14-50. Mitochondrial and nuclear DNA stained with a fluorescent dye.

Figure 14-50

Mitochondrial and nuclear DNA stained with a fluorescent dye. This micrograph shows the distribution of the nuclear genome (red) and the multiple small mitochondrial genomes (bright yellow spots) in a Euglena gracilis cell. The DNA is stained with ethidium (more...)

Mitochondria and Chloroplasts Contain Complete Genetic Systems

The biosynthesis of mitochondria and plastids requires contributions from two separate genetic systems. Most of the proteins in mitochondria and chloroplasts are encoded by special genes devoted to this purpose in nuclear DNA. These proteins are imported into the organelle from the cytosol after they have been synthesized on cytosolic ribosomes. Other organelle proteins are encoded by organelle DNA and synthesized on ribosomes within the organelle, using organelle-produced mRNA to specify their amino acid sequence (Figure 14-51). The protein traffic between the cytosol and these organelles seems to be unidirectional, as no known proteins are exported from mitochondria or chloroplasts to the cytosol. An exception occurs under special conditions when a cell is about to undergo apoptosis. The release of intermembrane space proteins (including cytochrome c) from mitochondria through the outer mitochondrial membrane is part of a signaling pathway that is triggered in cells undergoing programmed cell death (discussed in Chapter 17).

Figure 14-51. The production of mitochondrial and chloroplast proteins by two separate genetic systems.

Figure 14-51

The production of mitochondrial and chloroplast proteins by two separate genetic systems. Most of the proteins in these organelles are encoded by the nucleus and must be imported from the cytosol.

The processes of organelle DNA transcription, protein synthesis, and DNA replication (Figure 14-52) take place where the genome is located: in the matrix of mitochondria and the stroma of chloroplasts. Although the proteins that mediate these genetic processes are unique to the organelle, most of them are encoded in the nuclear genome. This is all the more surprising because the protein-synthesis machinery of the organelles resembles that of bacteria rather than that of eucaryotes. The resemblance is particularly close in chloroplasts. For example, chloroplast ribosomes are very similar to E. coli ribosomes, both in their structure and in their sensitivity to various antibiotics (such as chloramphenicol, streptomycin, erythromycin, and tetracycline). In addition, protein synthesis in chloroplasts starts with N-formyl methionine, as in bacteria, and not with the methionine used for this purpose in the cytosol of eucaryotic cells. Although mitochondrial genetic systems are much less similar to those of present-day bacteria than are the genetic systems of chloroplasts, their ribosomes are also sensitive to antibacterial antibiotics, and protein synthesis in mitochondria also starts with N-formyl methionine.

Figure 14-52. An electron micrograph of an animal mitochondrial DNA molecule caught during the process of DNA replication.

Figure 14-52

An electron micrograph of an animal mitochondrial DNA molecule caught during the process of DNA replication. The circular DNA genome has replicated only between the two points marked by red arrows. The newly synthesized DNA is colored yellow. (Courtesy (more...)

Organelle Growth and Division Determine the Number of Mitochondria and Plastids in a Cell

Mitochondria and plastids are large enough to be observed by light microscopy in living cells. For example, mitochondria can be visualized by expressing a genetically engineered fusion of a mitochondrial protein linked to the green fluorescent protein (GFP) in cells, or cells can be incubated with a fluorescent dye that is specifically taken up by mitochondria because of the electrochemical gradient across their membranes. From such images, the mitochondria in living cells are seen to be very dynamic—frequently dividing, fusing, and changing shape (Figure 14-53), as mentioned previously. Division (fission) and fusion of these organelles are topologically complex processes, because the organelles are enclosed by a double membrane and the integrity of the separate mitochondrial compartments must be maintained (Figure 14-54).

Figure 14-53. Dynamic mitochondrial reticulum.

Figure 14-53

Dynamic mitochondrial reticulum. (A) In yeast cells, mitochondria form a continuous reticulum underlying the plasma membrane. (B) A balance between fission and fusion determines the arrangement of the mitochondria in different cells. (C) Time-lapse fluorescent (more...)

Figure 14-54. Mitochondrial fission and fusion.

Figure 14-54

Mitochondrial fission and fusion. These processes involve both outer and inner mitochondrial membranes. (A) During fusion and fission, both matrix and intermembrane space compartments are maintained. Different membrane fusion machines are thought to operate (more...)

The copy number and shape of mitochondria vary dramatically in different cell types and can change in the same cell type under different physiological conditions, ranging from multiple spherical organelles to a single organelle with a branched structure (or reticulum). The arrangement is controlled by the relative rates of mitochondrial division and fusion, which are regulated by dedicated GTPases that reside on mitochondrial membranes. The regulation of mitochondrial morphology and distribution is important for cell differentiation and function. As an example, mutations in Drosophila that impair mitochondrial fusion, and hence cause extensive mitochondrial fragmentation, block sperm development and produce infertility.

There can be many copies of the mitochondrial and plastid genomes in the space enclosed by each organelle's inner membrane. How many of these genomes are present in a single organelle depends on the degree of organelle fragmentation; frequently, many genomes are housed in the same compartment (Table 14-2). In most cells, the replication of the organelle DNA is not limited to the S phase of the cell cycle, when the nuclear DNA replicates, but occurs throughout the cell cycle—out of phase with cell division. Individual organelle DNA molecules seem to be selected at random for replication, so that in a given cell cycle, some may replicate more than once and others not at all. Nonetheless, under constant conditions, the process is regulated to ensure that the total number of organelle DNA molecules doubles in every cell cycle, as required if each cell type is to maintain a constant amount of organelle DNA. When conditions change, the total organelle mass per cell can be regulated according to need. A large increase in mitochondria (as much as five- to tenfold), for example, is observed if a resting skeletal muscle is repeatedly stimulated to contract for a prolonged period.

Table 14-2. Relative Amounts of Organelle DNA in Some Cells and Tissues.

Table 14-2

Relative Amounts of Organelle DNA in Some Cells and Tissues.

In special circumstances, organelle division can be precisely controlled by the cell. In some algae that contain only one or a few chloroplasts, the organelle divides just before the cell does, in a plane that is identical to the future plane of cell division.

The Genomes of Mitochondria and Chloroplasts Are Diverse

The multiple copies of mitochondrial and chloroplast DNA contained within the matrix or stroma of these organelles are usually distributed in several clusters, called nucleoids. Nucleoids are thought to be attached to the inner mitochondrial membrane. Although it is not known how the DNA is packaged, the DNA structure in nucleoids is likely to resemble that in bacteria rather than that in eucaryotic chromatin. As in bacteria, for example, there are no histones.

The size range of organelle DNAs is similar to that of viral DNAs. Mitochondrial DNA molecules range in size from less than 6000 nucleotide pairs in Plasmodium falciparum (the human malaria parasite) to more than 300,000 nucleotide pairs in some land plants (Figure 14-55). Like a typical bacterial genome, most mitochondrial DNAs are circular molecules, although linear mitochondrial DNA exists as well. In mammals, the mitochondrial genome is a DNA circle of about 16,500 base pairs (less than 0.001% of the size of the nuclear genome). It is nearly the same size in animals as diverse as Drosophila and sea urchins. The chloroplast genome of land plants ranges in size from 70,000 to 200,000 nucleotide pairs, and it is circular in all organisms examined thus far.

Figure 14-55. Various sizes of mitochondrial genomes.

Figure 14-55

Various sizes of mitochondrial genomes. The complete DNA sequences for more than 200 mitochondrial genomes have been determined. The lengths of a few of these mitochondrial DNAs are shown to scale as circles for circular genomes and lines for (more...)

In mammalian cells, mitochondrial DNA makes up less than 1% of the total cellular DNA. In other cells, however, such as the leaves of higher plants or the very large egg cells of amphibians, a much larger fraction of the cellular DNA may be present in mitochondria or chloroplasts (see Table 14-2), and a large fraction of RNA and protein synthesis takes place there.

Mitochondria and Chloroplasts Probably Both Evolved from Endosymbiotic Bacteria

The procaryotic character of the organelle genetic systems, especially striking in chloroplasts, suggests that mitochondria and chloroplasts evolved from bacteria that were endocytosed more than 1 billion years ago. According to this endosymbiont hypothesis, eucaryotic cells started out as anaerobic organisms without mitochondria or chloroplasts and then established a stable endosymbiotic relation with a bacterium, whose oxidative phosphorylation system they subverted for their own use (Figure 14-56). The endocytic event that led to the development of mitochondria is presumed to have occurred when oxygen entered the atmosphere in substantial amounts, about 1.5 × 109 years ago, before animals and plants separated (see Figure 14-69).

Figure 14-56. A suggested evolutionary pathway for the origin of mitochondria.

Figure 14-56

A suggested evolutionary pathway for the origin of mitochondria. Microsporidia and Giardia are two present-day anaerobic single-celled eucaryotes (protozoans) without mitochondria. Because they have an rRNA sequence that suggests a great deal of evolutionary (more...)

Plant and algal chloroplasts seem to have been derived later from an endocytic event involving an oxygen-producing photosynthetic bacterium. To explain the different pigments and properties of the chloroplasts found in present-day higher plants and algae, it is usually assumed that at least three independent endosymbiotic events occurred.

Most of the genes encoding present-day mitochondrial and chloroplast proteins are in the cell nucleus. Thus, an extensive transfer of genes from organelle to nuclear DNA must have occurred during eucaryote evolution. In contrast, present organelle genomes are stable, indicating that a successful transfer is a rare evolutionary process. This is expected, because a gene moved from organelle DNA needs to change to become a functional nuclear gene: it must adapt to the nuclear and cytoplasmic transcription and translation requirements, and also acquire a signal sequence so that the encoded protein can be delivered to the organelle after its synthesis in the cytosol.

The gene transfer hypothesis explains why many of the nuclear genes encoding mitochondrial and chloroplast proteins resemble bacterial genes. The amino acid sequence of the chicken mitochondrial enzyme superoxide dismutase, for example, resembles the corresponding bacterial enzyme much more than it resembles the superoxide dismutase found in the cytosol of the same eucaryotic cell. Further evidence that such DNA transfers have occurred during evolution comes from the discovery of some noncoding DNA sequences in nuclear DNA that seem to be of recent mitochondrial origin; they have apparently integrated into the nuclear genome as “junk DNA.”

Gene transfer seems to have been a gradual process. When mitochondrial genomes encoding different numbers of proteins are compared, a pattern of sequential reduction of encoded mitochondrial functions emerges (Figure 14-57). The smallest and presumably most highly evolved mitochondrial genomes, for example, encode only a few inner-membrane proteins involved in electron-transport reactions, plus ribosomal RNAs and tRNAs. Mitochondrial genomes that have remained more complex contain this same subset of genes, plus others. The most complex genomes are characterized by the presence of many extra genes compared with animal and yeast mitochondrial genomes. Many of these genes encode components of the mitochondrial genetic system, such as RNA polymerase subunits and ribosomal proteins; these genes are instead found in the cell nucleus in organisms that have reduced their mitochondrial DNA content.

Figure 14-57. Comparison of mitochondrial genomes.

Figure 14-57

Comparison of mitochondrial genomes. Less complex mitochondrial genomes encode subsets of the proteins and ribosomal RNAs that are encoded by larger mitochondrial genomes. The five genes present in all known mitochondrial genomes encode ribosomal RNAs (more...)

What type of bacterium gave rise to the mitochondrion? From sequence comparisons, it seems that mitochondria are descendants of a particular type of purple photosynthetic bacterium that had previously lost its ability to perform photosynthesis and was left with only a respiratory chain. It is not certain that all mitochondria have originated from the same endosymbiotic event, however.

Mitochondrial Genomes Have Several Surprising Features

The relatively small size of the human mitochondrial genome made it a particularly attractive target for early DNA-sequencing projects, and in 1981, the complete sequence of its 16,569 nucleotides was published. By comparing this sequence with known mitochondrial tRNA sequences and with the partial amino acid sequences available for proteins encoded by the mitochondrial DNA, all of the human mitochondrial genes were mapped on the circular DNA molecule (Figure 14-58).

Figure 14-58. The organization of the human mitochondrial genome.

Figure 14-58

The organization of the human mitochondrial genome. The genome contains 2 rRNA genes, 22 tRNA genes, and 13 protein-coding sequences. The DNAs of many other animal mitochondrial genomes have also been completely sequenced. Most of these animal mitochondrial (more...)

Compared with nuclear, chloroplast, and bacterial genomes, the human mitochondrial genome has several surprising features:


Dense gene packing. Unlike other genomes, nearly every nucleotide seems to be part of a coding sequence, either for a protein or for one of the rRNAs or tRNAs. Since these coding sequences run directly into each other, there is very little room left for regulatory DNA sequences.


Relaxed codon usage. Whereas 30 or more tRNAs specify amino acids in the cytosol and in chloroplasts, only 22 tRNAs are required for mitochondrial protein synthesis. The normal codon-anticodon pairing rules are relaxed in mitochondria, so that many tRNA molecules recognize any one of the four nucleotides in the third (wobble) position. Such “2 out of 3” pairing allows one tRNA to pair with any one of four codons and permits protein synthesis with fewer tRNA molecules.


Variant genetic code. Perhaps most surprising, comparisons of mitochondrial gene sequences and the amino acid sequences of the corresponding proteins indicate that the genetic code is different: 4 of the 64 codons have different “meanings” from those of the same codons in other genomes (Table 14-3).

Table 14-3. Some Differences Between the “Universal” Code and Mitochondrial Genetic Codes*.

Table 14-3

Some Differences Between the “Universal” Code and Mitochondrial Genetic Codes*.

The observation that the genetic code is nearly the same in all organisms provides strong evidence that all cells have evolved from a common ancestor. How, then, does one explain the few differences in the genetic code in many mitochondria? A hint comes from the finding that the mitochondrial genetic code is different in different organisms. In the mitochondrion with the largest number of genes in Figure 14-57, that of the protozoan Reclinomonas, the genetic code is unchanged from the standard genetic code of the cell nucleus. Yet UGA, which is a stop codon elsewhere, is read as tryptophan in mitochondria of mammals, fungi, and invertebrates. Similarly, the codon AGG normally codes for arginine, but it codes for stop in the mitochondria of mammals and codes for serine in the mitochondria of Drosophila (see Table 14-3). Such variation suggests that a random drift can occur in the genetic code in mitochondria. Presumably, the unusually small number of proteins encoded by the mitochondrial genome makes an occasional change in the meaning of a rare codon tolerable, whereas such a change in a large genome would alter the function of many proteins and thereby destroy the cell.

Animal Mitochondria Contain the Simplest Genetic Systems Known

Comparisons of DNA sequences in different organisms reveal that the rate of nucleotide substitution during evolution has been 10 times greater in mitochondrial genomes than in nuclear genomes, which presumably is due to a reduced fidelity of mitochondrial DNA replication, inefficient DNA repair, or both. Because only about 16,500 DNA nucleotides need to be replicated and expressed as RNAs and proteins in animal cell mitochondria, the error rate per nucleotide copied by DNA replication, maintained by DNA repair, transcribed by RNA polymerase, or translated into protein by mitochondrial ribosomes can be relatively high without damaging one of the relatively few gene products. This could explain why the mechanisms that perform these processes are relatively simple compared with those used for the same purpose elsewhere in cells. The presence of only 22 tRNAs and the unusually small size of the rRNAs (less than two-thirds the size of the E. coli rRNAs), for example, would be expected to reduce the fidelity of protein synthesis in mitochondria, although this has not yet been tested adequately.

The relatively high rate of evolution of mitochondrial genes makes a comparison of mitochondrial DNA sequences especially useful for estimating the dates of relatively recent evolutionary events, such as the steps in primate evolution.

Some Organelle Genes Contain Introns

The processing of precursor RNAs has an important role in the two mitochondrial systems studied in most detail—human and yeast. In human cells, both strands of the mitochondrial DNA are transcribed at the same rate from a single promoter region on each strand, producing two different giant RNA molecules, each containing a full-length copy of one DNA strand. Transcription is therefore completely symmetric. The transcripts made on one strand are extensively processed by nuclease cleavage to yield the two rRNAs, most of the tRNAs, and about 10 poly-A-containing RNAs. In contrast, the transcript of the other strand is processed to produce only 8 tRNAs and 1 small poly-A-containing RNA; the remaining 90% of this transcript apparently contains no useful information (being complementary to coding sequences synthesized on the other strand) and is degraded. The poly-A-containing RNAs are the mitochondrial mRNAs: although they lack a cap structure at their 5′ end, they carry a poly-A tail at their 3′ end that is added posttranscriptionally by a mitochondrial poly-A polymerase.

Unlike human mitochondrial genes, some plant and fungal (including yeast) mitochondrial genes contain introns, which must be removed by RNA splicing. Introns have also been found in plant chloroplast genes. Many of the introns in organelle genes consist of a family of related nucleotide sequences that are capable of splicing themselves out of the RNA transcripts by RNA-mediated catalysis (discussed in Chapter 6), although these self-splicing reactions are generally aided by proteins. The presence of introns in organelle genes is surprising, as introns are not common in the genes of the bacteria whose ancestors are thought to have given rise to mitochondria and plant chloroplasts.

In yeasts, the same mitochondrial gene may have an intron in one strain but not in another. Such “optional introns” seem to be able to move in and out of genomes like transposable elements. In contrast, introns in other yeast mitochondrial genes have also been found in a corresponding position in the mitochondria of Aspergillus and Neurospora, implying that they were inherited from a common ancestor of these three fungi. It is possible that these intron sequences are of ancient origin—tracing back to a bacterial ancestor—and that, although they have been lost from many bacteria, they have been preferentially retained in some organelle genomes where RNA splicing is regulated to help control gene expression.

The Chloroplast Genome of Higher Plants Contains About 120 Genes

More than 20 chloroplast genomes have now been sequenced. The genomes of even distantly related plants (such as tobacco and liverwort) are nearly identical, and even those of green algae are closely related (Figure 14-59). Chloroplast genes are involved in four main types of processes: transcription, translation, photosynthesis, and the biosynthesis of small molecules such as amino acids, fatty acids, and pigments. Plant chloroplast genes also encode at least 40 proteins whose functions are as yet unknown; in addition, about twice that many genes of unknown function are present in the chloroplasts of some algae. Paradoxically, all of the known proteins encoded in the chloroplast are part of larger protein complexes that also contain one or more subunits encoded in the nucleus. We discuss possible reasons for this paradox later.

Figure 14-59. The organization of the liverwort chloroplast genome.

Figure 14-59

The organization of the liverwort chloroplast genome. The chloroplast genome organization is very similar in all higher plants, although the size varies from species to species—depending on how much of the DNA surrounding the genes encoding the (more...)

The similarities between the genomes of chloroplasts and bacteria are striking. The basic regulatory sequences, such as transcription promoters and terminators, are virtually identical in the two cases. The amino acid sequences of the proteins encoded in chloroplasts are clearly recognizable as bacterial, and several clusters of genes with related functions (such as those encoding ribosomal proteins) are organized in the same way in the genomes of chloroplasts, E. coli, and cyanobacteria.

Further comparisons of large numbers of homologous nucleotide sequences should help clarify the exact evolutionary pathway from bacteria to chloroplasts, but several conclusions can already be drawn:


Chloroplasts in higher plants arose from photosynthetic bacteria.


The chloroplast genome has been stably maintained for at least several hundred million years, the estimated time of divergence of liverwort and tobacco.


Many of the genes of the original bacterium are now present in the nuclear genome, where they have been integrated and are stably maintained. In higher plants, for example, two-thirds of the 60 or so chloroplast ribosomal proteins are encoded in the cell nucleus; these genes have a clear bacterial ancestry, and the chloroplast ribosomes retain their original bacterial properties.

Mitochondrial Genes Are Inherited by a Non-Mendelian Mechanism

Many experiments on the mechanisms of mitochondrial biogenesis have been performed with Saccharomyces cerevisiae (baker's yeast). There are several reasons for this preference. First, when grown on glucose, this yeast has an ability to live by glycolysis alone and can therefore survive with defective mitochondria that cannot perform oxidative phosphorylation. This makes it possible to grow cells with mutations in mitochondrial or nuclear DNA that interfere with mitochondrial function; such mutations are lethal in many other eucaryotes. Second, yeasts are simple unicellular eucaryotes that are easy to grow and characterize biochemically. Finally, these yeast cells normally reproduce asexually by budding, but they can also reproduce sexually. During sexual reproduction two haploid cells mate and fuse to form a diploid zygote, which can either grow mitotically or divide by meiosis to produce new haploid cells.

The ability to control the alternation between asexual and sexual reproduction in the laboratory greatly facilitates genetic analyses. Mutations in mitochondrial genes are not inherited in accordance with the Mendelian rules that govern the inheritance of nuclear genes. Therefore, long before the mitochondrial genome could be sequenced, genetic studies revealed which of the genes involved in yeast mitochondrial function are located in the nucleus and which in the mitochondria. An example of non-Mendelian (cytoplasmic) inheritance of mitochondrial genes in a haploid yeast cell is shown in Figure 14-60. In this example, we follow the inheritance of a mutant gene that makes mitochondrial protein synthesis resistant to chloramphenicol.

Figure 14-60. The difference in the patterns of inheritance between mitochondrial and nuclear genes of yeast cells.

Figure 14-60

The difference in the patterns of inheritance between mitochondrial and nuclear genes of yeast cells. For nuclear genes (Mendelian inheritance), two of the four cells that result from meiosis inherit the gene from one of the original haploid parent cells, (more...)

When a chloramphenicol-resistant haploid cell mates with a chloramphenicol-sensitive wild-type haploid cell, the resulting diploid zygote contains a mixture of mutant and wild-type genomes. The two mitochondrial networks fuse in the zygote, creating one continuous reticulum that contains genomes of both parental cells. When the zygote undergoes mitosis, copies of both mutant and wild-type mitochondrial DNA are segregated to the diploid daughter cell. In the case of nuclear DNA, each daughter cell receives exactly two copies of each chromosome, one from each parent. By contrast, in the case of mitochondrial DNA, the daughter cell may inherit either more copies of the mutant DNA or more copies of the wild-type DNA. Successive mitotic divisions can further enrich for either DNA, so that subsequently many cells will arise that contain mitochondrial DNA of only one genotype. This stochastic process is called mitotic segregation.

When diploid cells that have segregated their mitochondrial genomes in this way undergo meiosis to form four haploid daughter cells, each of the four daughters receives the same mitochondrial genes. This type of inheritance is called non-Mendelian, or cytoplasmic inheritance, to contrast it with the Mendelian inheritance of nuclear genes (see Figure 14-60). When non-Mendelian inheritance occurs, it demonstrates that the gene in question is located outside the nuclear chromosomes.

Although clusters of mitochondrial DNA molecules (nucleoids) are relatively immobile in the mitochondrial reticulum because of their anchorage to the inner membrane, individual nucleoids occasionally come together. This occurs frequently, for example, at sites where the two parental mitochondrial networks fuse during zygote formation. When different DNAs are present in the same nucleoid, genetic recombination can occur. This recombination can result in mitochondrial genomes that contain DNA from both parent cells, which are stably inherited after their mitotic segregation.

Organelle Genes Are Maternally Inherited in Many Organisms

The consequences of cytoplasmic inheritance are more profound for some organisms, including ourselves, than they are for yeasts. In yeasts, when two haploid cells mate, they are equal in size and contribute equal amounts of mitochondrial DNA to the zygote (see Figure 14-60). Mitochondrial inheritance in yeasts is therefore biparental: both parents contribute equally to the mitochondrial gene pool of the progeny (although, as we have just seen, after several generations of vegetative growth, the individual progeny often contain mitochondria from only one parent). In higher animals, by contrast, the egg cell always contributes much more cytoplasm to the zygote than does the sperm. One would therefore expect mitochondrial inheritance in higher animals to be nearly uniparental—or, more precisely, maternal. Such maternal inheritance has been demonstrated in laboratory animals. When animals carrying type A mitochondrial DNA are crossed with animals carrying type B, the progeny contain only the maternal type of mitochondrial DNA. Similarly, by following the distribution of variant mitochondrial DNA sequences in large families, it has been shown that human mitochondrial DNA is maternally inherited.

In about two-thirds of higher plants, the chloroplasts from the male parent (contained in pollen grains) do not enter the zygote, so that chloroplast as well as mitochondrial DNA is maternally inherited. In other plants, the pollen chloroplasts enter the zygote, making chloroplast inheritance biparental. In such plants, defective chloroplasts are a cause of variegation: a mixture of normal and defective chloroplasts in a zygote may sort out by mitotic segregation during plant growth and development, thereby producing alternating green and white patches in leaves. The green patches contain normal chloroplasts, while the white patches contain defective chloroplasts (Figure 14-61).

Figure 14-61. A variegated leaf.

Figure 14-61

A variegated leaf. In the white patches, the plant cells have inherited a defective chloroplast. (Courtesy of John Innes Foundation.)

A fertilized human egg carries perhaps 2000 copies of the human mitochondrial genome, all but one or two inherited from the mother. A human in which all of these genomes carried a deleterious mutation would generally not survive. But some mothers carry a mixed population of both mutant and normal mitochondrial genomes. Their daughters and sons inherit this mixture of normal and mutant mitochondrial DNAs and are healthy unless the process of mitotic segregation by chance results in a majority of defective mitochondria in a particular tissue. Muscle and nervous tissues are most at risk, because of their need for particularly large amounts of ATP.

An inherited disease in humans caused by a mutation in mitochondrial DNA can be recognized by its passage from affected mothers to both their daughters and their sons, with the daughters but not the sons producing grandchildren with the disease. As expected from the random nature of mitotic segregation, the symptoms of these diseases vary greatly between different family members—including not only the severity and age of onset, but also which tissue is affected.

Consider, for example, the inherited disease myoclonic epilepsy and ragged red fiber disease (MERRF), which can be caused by a mutation in one of the mitochondrial transfer RNA genes. This disease appears when, by chance, a particular tissue inherits a threshold amount of defective mitochondrial DNA genomes. Above this threshold, the pool of defective tRNA causes a decrease in the synthesis of the mitochondrial proteins required for electron transport and production of ATP. The result may be muscle weakness or heart problems (from effects on heart muscle), forms of epilepsy or dementia (from effects on nerve cells), or other symptoms. Not surprisingly, a similar variability in phenotypes is found for many other mitochondrial diseases.

Because of the unusually high rate of mutation observed in mitochondria, it has also been suggested that mutations that accumulate in mitochondrial DNAs may contribute to many of the medical problems of old age.

Petite Mutants in Yeasts Demonstrate the Overwhelming Importance of the Cell Nucleus for Mitochondrial Biogenesis

Genetic studies of yeasts have had a crucial role in the analysis of mitochondrial biogenesis. A striking example is provided by studies of yeast mutants that contain large deletions in their mitochondrial DNA, so that all mitochondrial protein synthesis is abolished. Not surprisingly, these mutants cannot make respiring mitochondria. Some of these mutants lack mitochondrial DNA altogether. Because they form unusually small colonies when grown in media with low glucose, all mutants with such defective mitochondria are called cytoplasmic petite mutants.

Although petite mutants cannot synthesize proteins in their mitochondria and therefore cannot make mitochondria that produce ATP, they nevertheless contain mitochondria. These mitochondria have a normal outer membrane and an inner membrane with poorly developed cristae (Figure 14-62). They contain virtually all the mitochondrial proteins that are specified by nuclear genes and imported from the cytosol—including DNA and RNA polymerases, all of the citric acid cycle enzymes, and most inner membrane proteins—demonstrating the overwhelming importance of the nucleus in mitochondrial biogenesis. Petite mutants also show that an organelle that divides by fission can replicate indefinitely in the cytoplasm of proliferating eucaryotic cells, even in the complete absence of its own genome. It is possible that peroxisomes normally replicate in this way (see Figure 12-34).

Figure 14-62. Electron micrographs of yeast cells.

Figure 14-62

Electron micrographs of yeast cells. (A) The structure of normal mitochondria. (B) Mitochondria in a petite mutant. In petite mutants, all the mitochondrion-encoded gene products are missing, and the organelle is constructed entirely from nucleus-encoded (more...)

For chloroplasts, the nearest equivalent to yeast mitochondrial petite mutants are mutants of unicellular algae such as Euglena. Mutant algae in which no chloroplast protein synthesis occurs still contain chloroplasts and are perfectly viable if oxidizable substrates are provided. If the development of mature chloroplasts is blocked in higher plants, however, either by raising the plants in the dark or because chloroplast DNA is defective or absent, the plants die.

Mitochondria and Plastids Contain Tissue-specific Proteins that Are Encoded in the Cell Nucleus

Mitochondria can have specialized functions in particular types of cells. The urea cycle, for example, is the central metabolic pathway in mammals for disposing of cellular breakdown products that contain nitrogen. These products are excreted in the urine as urea. Nucleus-encoded enzymes in the mitochondrial matrix perform several steps in the cycle. Urea synthesis occurs in only a few tissues, such as the liver, and the required enzymes are synthesized and imported into mitochondria only in these tissues.

The respiratory enzyme complexes in the mitochondrial inner membrane of mammals contain several tissue-specific, nucleus-encoded subunits that are thought to act as regulators of electron transport. Thus, some humans with a genetic muscle disease have a defective subunit of cytochrome oxidase; since the subunit is specific to skeletal muscle cells, their other cells, including their heart muscle cells, function normally, allowing the individuals to survive. As would be expected, tissue-specific differences are also found among the nucleus-encoded proteins in chloroplasts.

Mitochondria Import Most of Their Lipids; Chloroplasts Make Most of Theirs

The biosynthesis of new mitochondria and chloroplasts requires lipids in addition to nucleic acids and proteins. Chloroplasts tend to make the lipids they require. In spinach leaves, for example, all cellular fatty acid synthesis takes place in the chloroplast, although desaturation of the fatty acids occurs elsewhere. The major glycolipids of the chloroplast are also synthesized locally.

Mitochondria, in contrast, import most of their lipids. In animal cells, the phospholipids phosphatidylcholine and phosphatidylserine are synthesized in the endoplasmic reticulum and then transferred to the outer membrane of mitochondria. In addition to decarboxylating imported phosphatidylserine to phosphatidylethanolamine, the main reaction of lipid biosynthesis catalyzed by the mitochondria themselves is the conversion of imported lipids to cardiolipin (bisphosphatidylglycerol). Cardiolipin is a “double” phospholipid that contains four fatty acid tails (Figure 14-63). It is found mainly in the mitochondrial inner membrane, where it constitutes about 20% of the total lipid.

Figure 14-63. The structure of cardiolipin.

Figure 14-63

The structure of cardiolipin. Cardiolipin is an unusual lipid in the inner mitochondrial membrane.

We discuss the important question of how specific cytosolic proteins are imported into mitochondria and chloroplasts in Chapter 12.

Why Do Mitochondria and Chloroplasts Have Their Own Genetic Systems?

Why do mitochondria and chloroplasts require their own separate genetic systems, when other organelles that share the same cytoplasm, such as peroxisomes and lysosomes, do not? The question is not trivial, because maintaining a separate genetic system is costly: more than 90 proteins—including many ribosomal proteins, aminoacyl-tRNA synthases, DNA and RNA polymerases, and RNA-processing and RNA-modifying enzymes—must be encoded by nuclear genes specifically for this purpose (Figure 14-64). The amino acid sequences of most of these proteins in mitochondria and chloroplasts differ from those of their counterparts in the nucleus and cytosol, and it appears that these organelles have relatively few proteins in common with the rest of the cell. This means that the nucleus must provide at least 90 genes just to maintain each organelle's genetic system. The reason for such a costly arrangement is not clear, and the hope that the nucleotide sequences of mitochondrial and chloroplast genomes would provide the answer has proved to be unfounded. We cannot think of compelling reasons why the proteins made in mitochondria and chloroplasts should be made there rather than in the cytosol.

Figure 14-64. The origins of mitochondrial RNAs and proteins.

Figure 14-64

The origins of mitochondrial RNAs and proteins. The proteins encoded in the nucleus and imported from the cytosol have a major role in creating the genetic system of the mitochondrion, in addition to contributing most of the organelle's other proteins. (more...)

At one time, it was suggested that some proteins have to be made in the organelle because they are too hydrophobic to get to their site in the membrane from the cytosol. More recent studies, however, make this explanation implausible. In many cases, even highly hydrophobic subunits are synthesized in the cytosol. Moreover, although the individual protein subunits in the various mitochondrial enzyme complexes are highly conserved in evolution, their site of synthesis is not (see Figure 14-57). The diversity in the location of the genes coding for the subunits of functionally equivalent proteins in different organisms is difficult to explain by any hypothesis that postulates a specific evolutionary advantage of present-day mitochondrial or chloroplast genetic systems.

Perhaps the organelle genetic systems are an evolutionary dead-end. In terms of the endosymbiont hypothesis, this would mean that the process whereby the endosymbionts transferred most of their genes to the nucleus stopped before it was complete. Further transfers may have been ruled out, for mitochondria, by recent alterations in the mitochondrial genetic code that made the remaining mitochondrial genes nonfunctional if they were transferred to the nucleus.


Mitochondria and chloroplasts grow in a coordinated process that requires the contribution of two separate genetic systems—one in the organelle and one in the cell nucleus. Most of the proteins in these organelles are encoded by nuclear DNA, synthesized in the cytosol, and then imported individually into the organelle. Some organelle proteins and RNAs are encoded by the organelle DNA and are synthesized in the organelle itself. The human mitochondrial genome contains about 16,500 nucleotides and encodes 2 ribosomal RNAs, 22 transfer RNAs, and 13 different polypeptide chains. Chloroplast genomes are about 10 times larger and contain about 120 genes. But partly functional organelles form in normal numbers even in mutants that lack a functional organelle genome, demonstrating the overwhelming importance of the nucleus for the biogenesis of both organelles.

The ribosomes of chloroplasts closely resemble bacterial ribosomes, while mitochondrial ribosomes show both similarities and differences that make their origin more difficult to trace. Protein similarities, however, suggest that both organelles originated when a primitive eucaryotic cell entered into a stable endosymbiotic relationship with a bacterium. A purple bacterium is thought to have given rise to the mitochondrion, and (later) a relative of a cyanobacterium is thought to have given rise to the plant chloroplast. Although many of the genes of these ancient bacteria still function to make organelle proteins, most of them have become integrated into the nuclear genome, where they encode bacterialike enzymes that are synthesized on cytosolic ribosomes and then imported into the organelle.

Image ch14f69
Image ch12f34

By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.

Copyright © 2002, Bruce Alberts, Alexander Johnson, Julian Lewis, Martin Raff, Keith Roberts, and Peter Walter; Copyright © 1983, 1989, 1994, Bruce Alberts, Dennis Bray, Julian Lewis, Martin Raff, Keith Roberts, and James D. Watson .
Bookshelf ID: NBK26924