Background

The use of mouse ES (mES) cells to generate transgenic animals has revolutionized the study of gene action in vivo, and has provided animal models for studying the pathophysiology of human genetic disease and therapies thereof. Since the discovery that inherited mitochondrial DNA (mtDNA) mutations can cause age-related degenerative diseases (1), it has been realized that inherited mtDNA mutations are also associated with human adaptation to different environments, a broad spectrum of degenerative diseases, and with cancer. Somatic mtDNA mutations can accumulate during the lifetime of the individual, and have been implicated in the onset and progression of age-related diseases, and of ageing and cancer. Indeed, the importance of somatic mtDNA mutations in ageing has been affirmed by the expression in mice of a mutant form of mtDNA polymerase g subunit A, which contains a D257A mutation that inactivates the 3'-5' endonuclease. These animals have an increased mtDNA mutation rate and age prematurely (2). These findings highlight the pressing need for animal models of mitochondrial disease to better understand the role of the mitochondrion in the pathophysiology of disease, and in human genetics.

To date, the vast majority of transgenic mouse models of mitochondrial disease have been generated by modifying nuclear-encoded genes whose products contribute to mitochondrial function (3-8). These strains of mutant mice exhibit certain phenotypes characteristic of mitochondrial disease. However, inherently, they cannot recapitulate the unique genetic features of mtDNA mutations, such as maternal inheritance, bioenergetic threshold expression (see Section 1.1.1), and heteroplasmy, which are central to understanding mitochondrial diseases (9, 10). Heteroplasmy is defined as a mixture of two mtDNA genotypes, for example mutant and wild-type (wt) mtDNA, within a cell; and homoplasmy, as the presence of one genotype only.

Several different experimental approaches have been used in attempts to generate trans-mitochondrial mice, that is animals whose cells contain mitochondria with altered genomes. Microinjection of exogenous cytoplasm containing mitochondria with different mtDNA sequences into oocytes produced trans-mitochondrial embryos, but the mutation was lost by segregation during pre-implantation development (11). Heteroplasmic mice were also generated by fusion of enucleated cells (termed 'cytoplasts') with one-cell stage embryos, permitting introduction of mtDNAs with either naturally occurring polymorphisms (12, 13) or deletions (14). A third experimental approach to produce trans-mitochondrial mice involved microinjection of mES cells harbouring mitochondria with mutations in their DNA into blastocyst-stage embryos. Initial efforts to manipulate whole-animal systems involved the mtDNAs from cultured mouse cells resistant to the mitochondrial ribosome inhibitor, chloramphenicol (CAP). In a number of independent mouse cell lines, CAP resistance (CAPR) results from a T to C transition at nucleotide pair (np) 2433 (nucleotide (nt) #2433 T > C) near the 3' end of the 16S rRNA gene, among others (15, 16). Several decades ago, chimeric mice were produced using mouse embryonal carcinoma (mEC) cells into which CAPR mitochondria had been introduced by fusion of the mEC cells with cytoplasts from CAPR mouse cells; these CAPR mEC cells were then introduced into blastocysts, and the chimeric embryos transferred to the reproductive tract of pseudopregnant females (17). However, this experimental approach was limited by the extremely low efficiency with which mEC cells contribute to the germline, and there was a lack of data as to whether the CAPR mtDNAs were incorporated into the tissues of the chimeric animals.

More recently, these limitations have been overcome using mES cells that can contribute to the germline with high efficiency in combination with methods for identifying molecular polymorphisms to trace the origin and fate of the CAPR mtDNAs. In the contemporary experimental approach, female mES cells are fused to cytoplasts carrying a CAPR mtDNA, thereby generating cytoplasmic-hybrid mES cells, or 'mES cybrids.' These mES cybrids are tested to confirm that they harbour the mutant mtDNA, and those cybrids with mutant mtDNAs used to generate chimeras (18,19). By combining this approach with the selective depletion of the mES cells' endogenous mitochondria and mtDNAs prior to fusion, through treatment with rhodamine 6G (R6G) (20, 21), we succeeded in causing the CAPR mitochondria to be transmitted by chimeric females, through their oocytes, into subsequent generations (22). This has permitted us to make a more detailed analysis of the transmission of both heteroplasmic and homoplas-mic mutations, and to assess their in vivo consequences.

Our strategy should enable introduction of a wide variety of mouse mtDNA genotypes into the mouse female germline via somatic cell genetics. These include naturally occurring variants from different strains or species of mice and rodents, as well as mutants recovered from cells resistant to inhibitors of mitochondrial oxidative phosphorylation (OXPHOS), such as rotenone, antimy-cin A, and mycidin (23-28). In addition, the capture of somatic mtDNA mutations that accumulate during normal ageing, by clonal expansion of brain mtDNAs in synaptosome cybrids (29), may permit the introduction of naturally occurring, deleterious somatic mtDNA mutations into mice. Such trans-mitochondrial mice would allow the exploration of mtDNA changes in complex genetic processes, such as ageing.

In this chapter we present a brief primer to mitochondrial genetics. We then describe our protocols for the introduction of mtDNA mutations into female mES cells via cytoplast fusion to generate mES cybrids, in order to achieve maternal germline transmission.

1.1 Mitochondrial genetics and biochemistry

Mitochondria generate much of the cellular energy by OXPHOS. This process uses five multi-subunit protein complexes that are assembled from a mixture of gene products derived from both (cytoplasmic) mtDNA and nuclear (chromosomal) DNA (nDNA). Hence, mitochondrial energy production involves an interplay of Mendelian and mitochondrial genetics.

1.1.1 The mammalian mitochondrial genome

Mitochondria are derived from a symbiosis between bacteria and protoeukaryo-tic cells that occurred approximately two to three billion years ago. The ancestral bacterial genome contained all the genes required by a free-living organism that generated its energy by oxidizing fats and carbohydrates from its environment with oxygen, to generate water and ATP. However, early in the consolidation of this symbiotic relationship, many of the bacterial genes were transferred to the nucleus where, in the mammalian cell, they now reside, are replicated, and transcribed. The remainder of the ancestral bacterial genes are compartmentalized within the mitochondria. Messenger RNA from these nuclear-encoded genes is translated on cytosolic ribosomes and the resulting proteins are selectively imported into the mitochondrion, in some cases directed by an amino-terminal mitochondrial-targeting peptide. Today, the mammalian mitochondrial proteome is comprised of approximately 1500 gene products, 37 of which are encoded by the mtDNA (Figure 1), while the remainder are encoded by the nDNA. The mitochondrial genome is circular and encodes: seven (ND1, 2, 3,4L, 4, 5, and 6) of the 46 polypeptides of OXPHOS complex I (NADH dehydrogenase, or 'NDH'); none of the four subunits of complex II (succinate dehydrogenase, SDH); one (cytochrome b, or 'cytb') of the 11 subunits of complex III (bc1 complex); three (COI, II, III) of the 13 subunits of complex IV (cytochrome c oxidase, or 'COX'); and two (ATP6 and 8) of the 16 polypeptides of complex V (ATP synthase). All

DEAF 1555 F I T

MERRF 8344

NARP 8993

Leigh's 8993

Figure 1 Mapofthe human mitochondrial genome. D-loop = control region. Letters around the outside perimeter indicate cognate amino acids of the tRNA genes. Other gene symbols are defined in the text. Arrows followed by names of continents, and associated letters on the inside of the circle, indicate the position of polymorphisms that define region-specific mtDNA lineages (i.e. haplogroups). Arrows associated with abbreviations followed by numbers, around the outside of the circle, indicate representative pathogenic mutations, the number being the nucleotide position of the mutation. DEAF = deafness; MELAS = mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; LHON = Leber's hereditary optic neuropathy; ADPD = Alzheimer's and Parkinson's disease; MERRF = myoclonic epilepsy and ragged red fibre disease; NARP = neurogenic muscle weakness, ataxia, and retinitis pigmentosum; LDYS = LHON plus dystonia; PC = prostate cancer. (http://www.mitomap.org).

Control region mutations (Mutations: Somatic & Inherited ?)

DEAF 1555 F I T

LHON 14484

LDYS 14459

LHON 11778

LHON 10663

MERRF 8344

NARP 8993

Leigh's 8993

Figure 1 Mapofthe human mitochondrial genome. D-loop = control region. Letters around the outside perimeter indicate cognate amino acids of the tRNA genes. Other gene symbols are defined in the text. Arrows followed by names of continents, and associated letters on the inside of the circle, indicate the position of polymorphisms that define region-specific mtDNA lineages (i.e. haplogroups). Arrows associated with abbreviations followed by numbers, around the outside of the circle, indicate representative pathogenic mutations, the number being the nucleotide position of the mutation. DEAF = deafness; MELAS = mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; LHON = Leber's hereditary optic neuropathy; ADPD = Alzheimer's and Parkinson's disease; MERRF = myoclonic epilepsy and ragged red fibre disease; NARP = neurogenic muscle weakness, ataxia, and retinitis pigmentosum; LDYS = LHON plus dystonia; PC = prostate cancer. (http://www.mitomap.org).

remaining mitochondrial gene products are encoded by the nDNA, including the DNA and RNA polymerases, ribosomal proteins, and all enzymes (30).

The mtDNA genome also encodes a 12S and a 16S rRNA, as well as 22 tRNAs whose genes punctuate the larger genes (Figure 1). In addition, the mtDNA contains a 'control region' (CR) that harbours the promoters for transcription of the C-rich light- (L-) and G-rich heavy- (H-) strands, as well as the origin of H-strand replication. The origin of L-strand replication is located two-thirds of the way in a clockwise direction from the CR.

Mitochondrial DNA is inherited exclusively through the mother (31). In part this is because the mammalian oocyte contains up to 200 000 copies of the mitochondrial genome while the sperm contains relatively few (on the order of tens to hundreds). Moreover, the sperm mitochondria, which are located in the mid-piece of the flagellum, are ubiquitinated on the prohibitin isoforms of the

Figure 2 (see Plate 1) Morphology of mitochondria in endothelial cells in culture. Endothelial cells were fixed and stained with DAPI for the presence of DNA (blue), BODIPY FL phallacidin for filamentous actin (green), and Mitotracker Red (CMXRos) for mitochondria (red). In some cells the mitochondria appear as single organelles that can be elongated (arrowheads), while in other cells the mitochondria have a networked appearance (arrow).

mitochondrial membrane, and thus targeted for destruction by the oocyte ubiquitin-proteosome pathway (32-36).

Each somatic cell contains hundreds of mitochondria and thousands of mtDNAs. Mitochondria are frequently found as interconnected organelles within the cytosol of the cell (Figure 2). Hence, when a mtDNA mutation arises (whether spontaneously or experimentally), it creates an intracellular mixture of mutant and wt mtDNAs, that is heteroplasmy. During cell division the mutant and wt mtDNAs can be randomly distributed between the daughter cells, resulting in intracellular genetic drift toward either pure mutant or wt genotypes (i.e. homo-plasmy), a process referred to as replicative segregation. As the percentage of mutant mtDNAs increases, the energy output available to the cell declines until the cell has insufficient energy to function properly (i.e. the bioenergetic threshold), and ultimately dies (30).

1.1.2 Somatic-cell mitochondrial genetics

The first proof that mtDNA encodes gene products that are important for the mammalian phenotype came from somatic-cell genetic analysis of CAPR mouse and human cell lines, isolated following treatment with the mutagen, ethidium bromide (EtBr), and selection in CAP. The hereditary factor that carried the CAPR phenotype was subsequently shown to be cytoplasmic, as resistance could be transferred to a sensitive (CAPS) cell by fusion with a cytoplast derived from the CAPR cell. The resulting trans-mitochondrial cybrids could be cultured indefinitely with CAP, demonstrating that they had acquired an altered, heritable trait that was independent of the nucleus (37-39). Subsequent studies linked the transfer of CAPR with that of polymorphisms in mtDNA restriction endonuclease sites (40) and of mtDNA-encoded proteins (41); and that the CAPr phenotype was the result of mutations in the mtDNA-encoded 16S rRNA (15, 16, 42).

The availability of the CAPR marker enabled elucidation of the cellular principles of mitochondrial genetics. Using the CAPR and CAPS phenotypes as surrogate markers for the mtDNA genotypes of different cells, the fate of the corresponding mtDNAs could be monitored following the mixing that occurred during cell-cell and cytoplast-cell fusions to create heteroplasmy in hybrid cells and cybrids, respectively. These studies revealed that when mtDNAs originating from the same cell line, but with distinct and introduced genetic markers, were mixed by cell fusion, the two mtDNA genotypes could be propagated in culture as a heteroplasmic mixture for a considerable period of time. Eventually, however, the two forms segregated to homoplasmy. If the mtDNAs used were derived from different cell types of the same species, then incompatible combinations could arise and one of the mtDNA genotypes would selectively be excluded (43-45). Hence, it was inferred that the cells could discriminate between different mtDNAs within a mixed cytoplasm. As the genetic difference between the two mtDNA types increased, the selective exclusion of one of the mtDNAs became increasingly pronounced (39, 46-49).

Studies using different types of human cells with CAPR and CAPS phenotypes linked to distinctive mtDNA ND3 polymorphisms permitted analysis of the interaction between different mtDNAs. This revealed that mtDNAs that were present within a cell could complement each other in trans, suggesting that these organelles can fuse to allow mixing of the mtDNA-encoded products (50, 51). This mixing phenomenon is potentially problematical for studies involving the introduction of heteroplasmy into somatic cells (by embryo reconstitution). To obviate this, two methods have been developed to eliminate mtDNAs from the recipient cells prior to cell fusion. Cells can be exposed to an acute treatment with the dye, rhodamine 6G (R6G), which is toxic to mitochondria (20, 21). Alternatively, the mtDNAs of the recipient cell can be permanently removed by relatively long-term culture in EtBr (in the presence of increased concentrations of glucose, plus pyruvate and uridine), resulting in mtDNA deficient cells, commonly referred to as p° ('rho zero') cells (52).

The use of p° cells as recipients in cell-fusion experiments has permitted the production of xeno-mitochondrial cybrids, that is cultured cells in which the nucleus and mtDNAs are derived from different species. Using this approach, mtDNAs of chimpanzee and gorilla have been combined with the human nucleus. This resulted in cells with a partial complex I deficiency, presumably because of interspecies incompatibility of the mtDNA- and nDNA-encoded complex I polypeptides (47, 48). Similarly, xeno-mitochondrial cybrids with rat mtDNA and mouse nuclei have been prepared, which resulted in significant reduction in activities of complexes I-IV and in respiration. This illustrates the close functional coevolution of the nuclear and mitochondrial genomes. By contrast Mus spretus mtDNAs were functionally compatible with Mus musculus nuclei (48, 49).

1.2 Deleterious mtDNA mutations in hereditary disease

Since the late 1980s, many independent pathogenic mtDNA mutations have been identified and linked to a plethora of degenerative diseases. Mutations involving both mtDNA rearrangements and base substitution are common, and all are associated with diseases having a delayed onset and progressive course. Due to the organization of the mtDNA genome most rearrangement mutations remove at least one functional tRNA, which results in defects in synthesis of mitochondria-encoded proteins. Base-substitution mutations can either alter a tRNA or rRNA molecule, and thereby result in a protein synthesis defect, or alter a polypeptide and cause a specific OXPHOS complex defect. Representative pathogenic mtDNA mutations are shown in Figure 1. The tissues most commonly affected by mtDNA mutations involve those with a high-energy demand, including the central nervous system, leading to various forms of blindness, deafness, dementias, movement disorders, etc. Other affected tissues may include the heart, skeletal muscle, renal, and endocrine tissues, with dysfunction of the latter commonly presenting as Type II diabetes. Hence, mtDNA diseases typically affect the same tissues, and give the same pathology as seen in the common age-related diseases and in the ageing process per se (30). This concept has been amply confirmed by generating mice that are heteroplasmic for mtDNA mutations.

1.3 Generation, identification, and recovery of mtDNA mutations from mice

To study the effects of mtDNA mixing in mammals with the mouse as the experimental paradigm, methods had to be developed for introducing hetero-logous mtDNAs into the whole animal. This, in turn, necessitated the identification of variant mouse mtDNAs, as well as the development of procedures to recover the mutant mtDNAs and introduce them into the female germline. Functional mtDNA mutations can be generated by several methods, which are summarized as follows.

1.3.1 Capture of naturally occurring mtDNA variants from ageing animals, and from mice around the world

In addition to variants available in cell lines and mutant mice, the wild mouse population harbours an extensive reservoir of mtDNA sequence variation. Many of the inbred mouse strains were derived from the same female lineage, and thus have very similar mtDNAs (53). In contrast, the genetic diversity of wild mice is great. Mus musculus encompasses five subspecies: domesticus found in Europe, Africa, Australia, and the Americas; musculus in Asia; bactrianus in India; castaneus in Southeast Asia; and molossinus in Japan (54). The corresponding strain differences in mouse mtDNAs have recently been exploited to generate homoplasmic and xeno-mitochondrial mice (55).

That naturally occurring mtDNA variants can be functionally relevant in different mouse strains has been demonstrated by the introduction of New Zealand

Black (NZB/BinJ) mouse mtDNA into BALB/c embryos, or C57BL/6 mtDNA into NZB embryos. This was achieved by fusion of cytoplasmic vesicles from donor oocytes with recipient, fertilized oocytes (12). In the resulting heteroplasmic mice, the NZB mtDNA was invariably selected for in the liver and kidney, while the BALB/c mtDNA was selected for in the blood and spleen; and the mtDNAs did not segregate in postmitotic tissues (13). The molecular basis for this selection is not yet known. Studies of single hepatocytes indicated that the NZB mtDNA had a 14% selective advantage over the BALB/c mtDNA, which apparently was not associated with a difference in mitochondrial respiration. Interestingly, in vitro culture of heteroplasmic hepatocytes results in the reversal of the segregation pattern to favour BALB/c-derived mtDNA (56). This same directionality in mtDNA segregation was observed when heteroplasmic, female BALB/ c mice were crossed with male Mus musculus castaneus. The segregation rate of the NZB and BALB/c mtDNAs in liver, kidney, and spleen was found to vary among the F2 generation progeny, which facilitated mapping of nDNA loci that modulated the tissue-specific segregation patterns. Accordingly, quantitative trait loci (QTL) were mapped to chromosome 5 for liver segregation, chromosome 2 for kidney, and chromosome 6 for spleen. The chromosome 5 locus was associated with marker D5Mit25 with a LOD score of 31.5, clearly demonstrating that the tissue-specific segregation patterns can be associated with specific chromosomal loci (57).

Further evidence that the NZB mtDNA sequence variation was of phenotypic relevance was obtained by backcrossing the NZB mtDNA onto a CBA/H nuclear background, and vice versa. This resulted in mice with differences in cognitive capacity and neuroanatomy, indicating that mtDNA polymorphisms can affect cortical development and function (58).

1.3.2 Generation of cultured cells resistant to inhibitors of mitochondrial respiration

A variety of mouse cell lines have been isolated that have mutations in specific mtDNA genes that encode polypeptides. Lines resistant to the complex I inhibitor, rotenone, were derived with frame-shift mutations in both the ND5 (59) and ND6 (60) genes. Similarly, lines resistant to a number of inhibitors of cytb, including Antimycin A (24, 26), HQNO (25, 27), myxothiazol (23), and stimatellin (26), have been reported. Lines carrying cytb mutations have been isolated using negative selection procedures (61), while spontaneous mutations have been identified in mtDNA COX genes (62).

Thus, diverse procedures have been used to isolate mitochondrial mutations in cultured cells. Those with cytoplasmic - as opposed to nuclear - mutations are then identified using the cybrid transfer technique. Cells that harbour the mtDNA 16S rRNA gene CAPR mutation have a mitochondrial protein synthesis defect similar to those seen in patients with mtDNA, rRNA, and tRNA mutations. The CAPR mutation reduces mitochondrial protein synthesis, leading to a partial inhibition of complexes I, III, and IV (18).

1.3.3 Generation of random mtDNA mutations in cells or animals using various mutator gene strategies

An alternative method for selecting mtDNA mutations in cultured cells involves rescue of somatic mutations from either ageing mice or mice harbouring mtDNA mutator genes. Age-related somatic mtDNA mutations have been retrieved from both human and mouse adult brain cells by fusion of mitochondria-containing synaptosomes with p0 cells. Fresh brain tissue is isolated and homogenized, and the synaptosomes isolated using Percoll™ gradients (Protocol 4). The synaptosomes are then fused with p0 cells generated from cell lines of the same species, and the cybrids selected for growth in the absence of pyruvate and/or uridine, which selects for cells that have acquired functional mitochondria. Such synap-tosome cybrids can then be screened for defects in OXPHOS, and for mutations in the mtDNA (29).

To increase the probability of isolating desirable mutations, cytoplasts or synaptosomes may be generated from cells or mice, respectively, that display increased rates of mtDNA mutation. For cultured cells this may arise from expression of a mitochondrial polymerase g subunit A transgene that encodes an inactivated 3'-5' exonuclease, for example the D198A missense mutation in human polymerase g (63). Human cell lines expressing such a polymerase g mutation display very high mtDNA mutation rates, resulting in the accumulation of 1:1700 mtDNA point mutations during 3 months of continuous culture. Further culture results in the inactivation of the mutant polymerase (64). Such induced mtDNA mutations may be recovered by fusion of the cytoplasts of the mutator cells with p0 cells. Similarly, somatic mtDNA mutations can be rescued from the brains of transgenic mice in which the endogenous polymerase g has been replaced by a polymerase lacking the proofreading capacity (2).

The somatic mtDNA mutation rate can also be elevated by increasing production of mitochondrial reactive oxygen species (ROS). This has been achieved in our laboratories either through inactivation of the ANT1 and 2 genes in the brains of mice (4, 65, 66), or through inactivation of the mitochondrial manganese superoxide dismutase (MnSOD) (3, 7, 67).

1.3.4 Genetic engineering of desirable mtDNA mutations for introduction into cells

Ideally, experimental strategies would enable the introduction of specific genetic alterations into the mtDNAs of mice. Various attempts have been made to accomplish mtDNA transformation, but no practical procedure for mtDNA-mediated gene transfer currently exists (68). However, one promising approach is to generate a peptide nucleic acid (PNA), in which a peptide backbone substitutes for a sugar phosphate backbone, conjugated with an amino-terminal mito-chondrial-targeting peptide. The PNA can then be annealed to an oligonucleotide homologous to the mtDNA sequence of interest. When cells are exposed to PNA-oligonucleotide, the complex is taken up and rapidly targeted to the mitochondrion, in whose matrix the oligonucleotide is deposited (69). By engineering a base substitution within the oligonucleotide to constitute the desired mutation, the oligonucleotide may serve as a primer for mtDNA replication and thereby generate the desired mutation. Additional studies are required to demonstrate the practicality of this method.

In summary, the above experimental approaches have been successful to varying degrees in generating, identifying and recovering mtDNA mutations, and introducing them into cultured cells. Next, methods were required for introduction of these mutations into mES cells with subsequent production of lines of trans-mitochondrial mice.

Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook


Post a comment